Abstract book - Conventus Congressmanagement & Marketing GmbH

Proteomic Forum 2015
2
Table of contents
Pre–Congress Workshops Sunday, 22 March 2015 ............................................................. 3
HUPO - Workshop Cardiovascular Disease ............................................................................................................................. 4
Mechanism and understanding complexity of protein modifications .............................................................................................................4
Biomarkers towards clinical relevance in CVD ................................................................................................................................................5
In the road to translation in CVD .....................................................................................................................................................................7
Educational Day ..................................................................................................................................................................... 8
Session I ...........................................................................................................................................................................................................8
Session II ........................................................................................................................................................................................................10
Session III .......................................................................................................................................................................................................10
Scientific Program 22 – 25 March 2015 ........................................................................... 11
Sunday, 22 March 2015 ........................................................................................................................................................ 12
News Corner ..................................................................................................................................................................................................12
Opening Lecture ............................................................................................................................................................................................18
Monday, 23 March 2015 ...................................................................................................................................................... 19
Plenary Session • New Technologies.............................................................................................................................................................19
Plenary Session • Post-Translational Modifcations and Pathways ................................................................................................................20
Symposium • Affinity Proteomics .................................................................................................................................................................20
Symposium • Gene Regulation & Epigenetics ...............................................................................................................................................22
Symposium • Plant & Agricultural Proteomics I ............................................................................................................................................24
Symposium • Disease Proteomics .................................................................................................................................................................26
Symposium • Imaging ...................................................................................................................................................................................28
Symposium • Microbial Proteomics ..............................................................................................................................................................30
Tuesday, 24 March 2015 ...................................................................................................................................................... 32
Plenary Session • Interaction Proteomics .....................................................................................................................................................32
Plenary Session • Subcellular Proteomics .....................................................................................................................................................33
Symposium • Host-Pathogen Interaction & Antibiotical Resistance .............................................................................................................34
Symposium • Plant & Agricultural Proteomics II ...........................................................................................................................................36
Symposium • Quantitative Strategies ...........................................................................................................................................................38
Young Investigator Session ............................................................................................................................................................................40
Wednesday, 25 March 2015 ................................................................................................................................................. 46
Plenary Session • Systems Biology ................................................................................................................................................................46
Plenary Session • Proteomics in Health and Disease ....................................................................................................................................47
Symposium • Computational Omics I............................................................................................................................................................48
Symposium • Extracellular Vesicles ...............................................................................................................................................................50
Symposium • Post-translational Modifcations ..............................................................................................................................................51
Symposium • Computational Omics II...........................................................................................................................................................53
Symposium • Structural Proteomics .............................................................................................................................................................54
Closing Lecture ..............................................................................................................................................................................................56
Poster Sessions ................................................................................................................ 57
Poster Session A • 22 March 2015, 19:00–21:00 .................................................................................................................. 58
Bioinformatics (P014–P030) ..........................................................................................................................................................................58
Free topics (P032–P049)................................................................................................................................................................................69
Proteomics for Human Health I (P123–P145) ...............................................................................................................................................83
Proteomics in Biotechnology (P192–P198) ...................................................................................................................................................96
Quantitative Proteomics I (P199–P215) ......................................................................................................................................................101
Poster Session B • 23 March 2015, 15:00–16:30 ................................................................................................................. 112
Affinity Proteomics (P001–P013) ................................................................................................................................................................112
Imaging (P050–P055) ..................................................................................................................................................................................120
Mass Spectrometry Technologies (P056–P065) ..........................................................................................................................................123
Plant and Microbial Proteomics I (P066–P079) ...........................................................................................................................................129
Post-translational Modifications I (P092–P107) ..........................................................................................................................................137
Proteomics for Human Health II (P146–P169) ............................................................................................................................................146
Poster Session C • 24 March 2015, 15:00–16:30 ................................................................................................................. 164
Plant and Microbial Proteomics II (P080–P091) ..........................................................................................................................................164
Post-Translational Modifcations II (P108–P122) .........................................................................................................................................172
Proteomics for Human Health III (P170–P190) ...........................................................................................................................................180
Quantitative Proteomics II (P216–P232) .....................................................................................................................................................195
Structural Proteomics (P233–P238) ............................................................................................................................................................205
Subcellular Proteomics (P239–P251) ..........................................................................................................................................................210
Proteomic Forum 2015
3
Pre–Congress Workshops
Sunday, 22 March 2015
HUPO - Workshop Cardiovascular Disease ............................................................................................................................. 4
Mechanism and understanding complexity of protein modifications .............................................................................................................4
Biomarkers towards clinical relevance in CVD ................................................................................................................................................5
In the road to translation in CVD .....................................................................................................................................................................7
Educational Day ..................................................................................................................................................................... 8
Session I ...........................................................................................................................................................................................................8
Session II ........................................................................................................................................................................................................10
Session III .......................................................................................................................................................................................................10
Proteomic Forum 2015
4
HUPO - Workshop Cardiovascular Disease
Mechanism and understanding complexity of protein modifications
09:00–10:30
Chair: Merry Lindsey (Jacksonville/US)
HUPO 1
Exploiting antibody and capture reagents for diseases affecting the cardiovasclular system
1
1
1
1,2,3,4
1
1
J. Odeberg , M. Iglesias , E. Birgersson , . VinnCARDIO Consortium
, P. Nilsson , M. Uhlen , J. Schwenk
1
SciLifeLab, KTH Royal Institute of Technology, Affinity Proteomics, Solna, Sweden
2
Karolinska Institutet, Stockholm, Sweden
3
Uppsala University, Uppsala, Sweden
4
Karolinska University Hospital, Solna, Sweden
1
For a systematic exploration of proteins in larger sets of plasma samples, we developed an affinity proteomics strategy based on
suspension bead arrays assays [1] and antibodies from the Human Protein Atlas (www.proteinatlas.org). Allowing to perform
hypothesis-free disease studies, this antibody-based workflow was established to reach beyond a pure discovery and to build
sandwich assays with indicate antibodies and towards a clinical implementation downstream. For target validation, we employ
additional study sets, several antibodies towards a common target, mass spectrometry for the identification of proteins
captured from plasma [2] as well as sandwich assays [3]. This single-binder bead array has yet revealed potential candidates for
prostate cancer [1,3], neuroendocrine tumors [4], multiple sclerosis [5], ALS [6], childhood malaria [7], neuromuscular
dystrophies [8] and renal impairment [9].
The presentation will describe an affinity-based strategy applied to discover and validate proteins in plasma related to diseases
of the cardiovascular system. In a project, denoted VinnCARDIO, we initially screened 384 patient samples from four distinct
clinical study sets using more than 10,000 antibodies. We then selected 720 candidates for subsequent targeted analysis in >
3,000 retrospective plasma samples in the context of myocardial infarction and unstable angina. The dependency of the
discovered candidates was assessed for established risk factors (BMI, smoking, age) as used for clinical management. Candidate
targets were then validated by immune-capture mass spectrometry and sandwich assays with plasma. Presently sandwich tests
are being implemented into a clinical routine laboratory to evaluate marker performance in prospective samples and in relation
to troponin and other clinical risk factors. This work describes a translational study on affinity-based plasma proteomics that
provided candidates that, upon further clinical evaluation, may support and improve current tools in clinical routine.
References: [1] Schwenk, J.M. et al. (2010) Mol Cell Proteomics; [2] Neiman, M. et al (2013) Proteomics; [3] Qundos, U. et al
(2014) Transl. Proteomics; [4] Darmanis, S. et al. (2013) PLOSone; [5] Byström et al (2014) JRP; [6] Häggmark et al (2014) Ann Clin
Transl Neurol.; [7] Bachmann et al (2014) PLOS Pathog; [8] Ayoglu et al (2014) EMBO Mol Med; [9] Neiman M. et al. (2011) JPR
HUPO 2
Age-, sex- and disease dependent myocardial protein patterns
1
V. Regitz-Zagrosek
1
Charité, Berline, Germany
Cardiovascular diseases are known to change gene expression and the protein composition in the myocardium and both occurs
in a sex specific manner. Since protein expression is closer to function we focused our effort on analyzing disease-, sex- and agedependent protein expression in the human heart and also in animal models. We studied the heart proteome in young and old
male and female mice at 14 and 100 weeks of age using two-dimensional electrophoresis and mass spectrometry and compared
with patterns in young and old humans. Protein pattern comparison in young and old mice revealed 7 and 22 protein spots with
sex- and age related expression profiles. The variant protein spots were identified and revealed 10 distinct proteins and several
isoforms, thereof: α1-antitrypsin, apolioproteins, carbonic anhydrase, peroxiredoxin and others. Surprisingly, more sex-related
proteins were detected in old than in the young mice. Of note, six of the variant proteins detected in mice were also observed to
change in age and sex dependent manner in the human heart. Moreover these proteins are associated with functional pathways
that have been suggested to be affected by myocardial diseases in genomic studies. Some of them are closely related to free
radical production and defense against free radical induced damage. These proteins have been described as potentially
protective in female human hearts. We conclude that relevant myocardial protein changes can be detected by two-dimensional
Proteomic Forum 2015
5
electrophoresis in human and mouse hearts and some of these protein changes are conserved throughout species. They may
offer particularly interesting therapeutic targets.
HUPO 3
Chamber specific protein expression in the human heart
Anthony Gramolini
Toronto, Canada
TBA.
HUPO 4
Functional plasticity of cardiac interfibrillary mitochondria (IFM) as response mechanism to stress
Juliane Heidler
Frankfurt am Main, Germany
TBA.
Biomarkers towards clinical relevance in CVD
11:00–12:30
Chair: Anthony Gramolini (Toronto/CA)
HUPO 5
Biomarkers to cardiac extracellular matrix
Merry Lindsey
Jacksonville, FL, United States
TBA.
HUPO 6
Immunoassays: High Sensitivity, High Specificity, High Speed for, any Type of Analyte, and Clinical Specimen
1
O. Pötz
1
NMI Naturwissenschaftliches und Medizinisches Institut an der Universität Tübingen , Protein Analytics, Reutlingen, Germany
The clinical demand for immunoassays targeting protein biomarkers or biomarker candidates is still increasing. To date many
requirements in terms of analyte concentration, type of clinical sample, clinical specificity and multiplex capabilities can be
addressed by choosing the right assay platform or assay set-up. An overview of classical and emerging immunoassays ranging
from sandwich-, mass spectrometry-, bead array-, and digital array-based immunoassays will be given. Advantages of the
different methods will be demonstrated by their application in drug-induced injury biomarker research, toxicology, and
pharmacological research.
HUPO 7
Journey Toward Application: Multiplex MRM assays, Kits, and High Throughput Applications
1
C. Borchers
1
University of Victoria - Genome BC Proteomics Centre, Biochemistry & Microbiology, Victoria, Canada
Precise and accurate protein quantitation is essential for screening biomarkers for risk stratification, disease prognostication,
and therapeutic monitoring. The most promising analytical strategy for quantifying unverified biomarkers therein relies on
targeted MRM/MS with isotopically labeled standards. Using that general strategy, we have developed a number of highly
reproducible and multiplexed panels for quantifying candidate protein disease biomarkers in biofluids (plasma, cerebrospinal,
and urine) and dried blood spots (DBS). The methods collectively utilize a bottom-up sample prep workflow with a complex
mixture of our in-house synthesized peptide standards (for normalization) and standard-flow LC-MRM/MS analysis (for
heightened robustness and sensitivity). In these workflows, data analysis and results interpretation were facilitated by our
developed software tool - Qualis-SIS. To date, we have robustly quantified 192 (76 CVD-linked), 136 (45 CVD-linked), 130 (55
Proteomic Forum 2015
6
CVD-linked), and 103 (40 CVD-linked) endogenous proteins (all spanning at least 5 orders of magnitude in concentration) in
human plasma, urine, cerebrospinal fluid, and DBS, respectively. In mouse, we have quantified 93 proteins (35 CVD-linked) in
plasma and found 211 proteins (77 CVD-linked) to qualify for quantitation from preliminary interference screening of heart
tissue. To aid standardization, these developments are being translated into biomarker assessment kits (BAKs) for the
quantitative proteomics field. Using these BAKs in conjunction with our established quality control (QC) kits should help improve
method reproducibility and transferability between laboratories, leading to a more rapid and accurate evaluation of putative
protein biomarkers in a biological sample of interest. This presentation will first provide an overview of our latest developments
and applications in MRM quantitative proteomics (human and mouse) then highlight our MRM kits (i.e., BAK for discovery and
CVD) for assessing instrument/method performance and protein biomarker utility.
HUPO 8
Comparative lipidomic and proteomic analysis of the plasma elimination rate of current lipid apheresis methods
1
2
3
2
4
1,5
S. Wallner , J. Gräßler , C. Borchers , U. Julius , H. Meyer , G. Schmitz
1
University Hospital Regensburg, Institute for Clinical Chemistry and Laboratory Medicine, Regensburg, Germany
2
Technical University Dresden, Pathological Biochemistry, Dresden, Germany
3
University of Victoria, Proteome Center, Victoria, Canada
4
Leibniz-Institut für Analytische Wissenschaften(ISAS), Dortmund, Germany
5
Lipoconsult GmbH, Havixbeck, Germany
Lipid apheresis is a procedure that allows the elimination of lipoproteins from plasma or whole blood. It is a therapeutic option
in high-risk patients, who do not respond satisfactorily to drug therapy. Functional principles include (i)filtration (MDF, MONET),
(ii)adsorption (Liposober D, Therasorb and DALI) and a (iii)precipitation based approach using heparin (HELP). Currently no data
on a method specific modulation of the plasma lipidome and proteome are available. Therefore the aim of this study was to
perform lipidomic and proteomic analysis before and after apheresis and identify differences between the different apheresis
approaches.
79 patients with elevated LDLc or Lp(a) were enrolled at the TU Dresden. Each patient was randomized to one of six different
treatment groups. Lipidomic analysis was performed in Regensburg and proteomic analysis in Victoria. Additionally routine lipid
parameters were determined.
LDLc was lowered effectively by all apheresis methods with a mean decrease of 70% (110mg/dl to 34mg/dl). Lp(a) levels
decreased on average 76% (55mg/dl to 16mg/dl). Triglycerides showed a reduction rate of 40% to 67% (208mg/dl to 100mg/dl).
HDLc was most affected by the filtration methods MDF and MONET. In lipidomics CE species were lowered by all methods while
PC, PC ether and LysoPC showed a lower reduction. PE species showed a slight decrease with the strongest reduction induced by
MONET. PE plasmalogens, PI and ceramides were lowered equally by all methods.
In proteomics Therasorb led to a stronger reduction in ApoD than the other methods. It also had the lowest impact on the
complement system, while treatment with DALI or LIPO D had the strongest effect. Fibrinogen was reduced in patients treated
with HELP, MDF and MONET, whereas lipid adsorption had no effect. DALI, HELP and LIPO D diminished coagulation factor IX and
XII. These levels remained constant with MDF, MONET and Therasorb. Gelsolin and Kallistatin, proteins associated with
atherosclerosis development, were reduced by DALI and LIPO D. Apheresis with HELP selectively lowered CD44 levels. The
transporter haptoglobin was selectively reduced by MDF and Therasorb. The high molecular weight protease inhibitor α-2macroglobulin was most effectively eliminated by filtration methods. In contrast Insulin-like growth factor-binding protein 3
(IGF-BP3) was most strongly lowered by the adsorption techniques DALI and LIPO D. Only Therasorb was able to eliminate the
Alzheimer related proteins Tau and pTau. Currently analysis of the elimination rate of microparticle bound proteins is
performed.
In summary all techniques lowered standard lipid parameters comparably. Still there were method specific differences in the
impact on individual lipid species. Proteomic data show that the methods clearly differ in their elimination profile which might
account for differences in therapeutic efficacy and frequency of side effects.
Proteomic Forum 2015
7
In the road to translation in CVD
14:00–15:30
Chair: Jochen Schwenk (Stockholm/SE)
HUPO 9
Metabolomics profiling in late stage heart failure
Peipei Ping
Los Angeles, CA, United States
TBA.
HUPO 10
Proteomic toolset to elucidate context-dependent disease pathways in aortic aneurysm: divergent TGF beta signaling
1
1
2
1
J. Van Eyk , S. Parker , K. Raedschelders , H. Dietz
1
Johns Hopkins, Genetic Medicine, Baltimore, United States
2
Cedar Sinai Medical Center, Heart Institute, Los Angeles, United States
A cells reception and response to signaling messages from the tissue microenvironment is dependent on the internal and
external context. Pathology often emerges from aberrations in these interactions. The overarching goal of this project is to
delineate the molecular mechanisms underlying aneurysm of the thoracic aorta. We are developing and implementing mass
spectrometry based pipelines to define these context specific biological events and predict how they influence the disease
process. By assembling a library of proteins specific to aortic vascular smooth muscle cells (VSMCs), under basal conditions and
in response to stimulation by multiple different factors relevant to aortic disease, we are building a resource for multiple
avenues of inquiry. For instance, we have identified the underlying factors intrinsic to VSMC lineage of origin (i.e., second heart
field and neural crest) that may mediate the cellular response to transforming growth factor b (TGFb). We hypothesize that
these subtle molecular differences may be exacerbated in the context of the genetic mutations that are known to cause aortic
aneurysm. In quantitative MS experiments using both iTRAQ or SWATH-MS approaches we quantified expression of > 2,300
unique proteins between VSMC lineages. Uncovering 188 soluble and 88 nuclear or membrane proteins differentially expressed
between disparate embryonic lineages of origin. In other work using the same reference library, we have mapped how
endocrine activation within the renin-angiotensin system interacts with TGFb signaling in the wall of the ascending aorta to
promote pathological remodeling. Data independent acquisition using SWATH-MS allowed for confident quantification of over
2,900 non-redundant proteins. Treatment with TGFb induced significant changes in expression among 227 VSMC proteins.
Surprisingly, the AT1R blocker, telmisartan, attenuated TGFb-induced responses in more than half (n=162) of these proteins. The
patterns of altered expression highlighted a handful of upstream transcriptional regulators with starkly different responses to
TGFb in the presence of AT1R inhibition, including RICTOR, HIF1alpha, and MYC. We are integrating this downstream expression
analysis with the rapid upstream phosphopeptide response to TGFb, as well as interactome mapping of AT1R and TGFb signaling
effectors, to pinpoint the precise molecular intersections between these two signaling systems. Our ultimate goal is to apply
these molecular discoveries toward the identification of novel candidates for therapeutic interventions that will allay the
progression of aneurysm in aortic disease.
HUPO 11
Moving from inventory to personalized medicine: reducing cardiovascular mortality
1
1
1
1
A. Sickmann , J. M. Burkhart , R. Ahrends , R. P. Zahedi
1
Leibniz-Institut für Analytische Wissenschaften - ISAS - e.V., Dortmund, Germany
The concept of personalized medicine is not novel: Scientists and clinicians have long observed that patients with similar
symptoms may have different diseases, with diverse causes; and likewise, that therapeutic interventions may work well in many
patients with a disease but not in others with obviously the same disease. Personalized medicine is the idea that diagnoses and
treatments will be tailored by knowing those molecular differences of a patient that might influence disease progression and/or
the efficacy of drug treatment. The ISAS is aims to drive the underlying analytical science forward that will assist the
determination and interpretation of genetic variation in near future. The ISAS recently demonstrated that thousands of proteins
as well as their posttranslational modification can be analyzed from platelets for the distinction of different diseases states. In
particular, quantitative phosphoproteomic studies will pave the way for a refined understanding of platelet properties, beyond
the classical biochemical studies which focused on the description of linear pathways. Although these approaches undeniably
have made and will make important discoveries, the past few years clearly showed that signaling is much more complicated,
Proteomic Forum 2015
8
branched, and nonlinear than originally anticipated, including a considerable level of cross talk, in platelets for instance between
inhibitory and activatory pathways.
The combination of different omics technologies will therefore provide a powerful instrument to prevent, detect, understand
and to treat disease of different subtypes which would otherwise clinically not addressed.
HUPO 12
Workshop summary: making it count, saving lives by reducing heart disease
J. Yates
San Diego, CA, United States
TBA.
Educational Day
Session I
09:00–10:30
Chair: Merry Lindsey (Jacksonville, FL/US)
ED1
Protein quantification by SWATH MS: Data-independent acquisition combined with targeted data analysis
1
C. Ludwig
1
ETH Zurich, Institute of Molecular Systems Biology, Zurich, Switzerland
During the past few years, targeted proteomics has emerged as a complement to the more widely used discovery or shotgun
proteomic methods. The major benefit of targeted proteomics is its capability to generate high-quality quantitative data at a
high level of reproducibility, sensitivity and accuracy. The significance of targeted proteomics is increasingly being recognized,
highlighted by its selection as the method of the year 2012 by Nature Methods. The most established MS approach supporting
targeted proteomics is Selected Reaction Monitoring (SRM), however, several novel targeted methods have emerged recently,
one of them being SWATH MS.
In SWATH MS data are acquired in a sequential windowed data-independent acquisition mode, which allows an unbiased
recording of all proteins, peptides and fragment ions present in a sample above the detection limit. In turn, SWATH MS data
consist of composite MS/MS spectra, which are highly complex and can no longer by analyzed using conventional MS/MS
identification software. Therefore in SWATH MS a targeted data analysis workflow has been established in which prior
knowledge derived from shotgun proteomic measurements gets exploited. Important MS coordinates, like the detectable
peptides per protein, the fragmentation pattern of each peptide and the expected chromatographic retention time, are
gathered in form of an assay library and are extracted in silico from the SWATH data in a targeted fashion. SWATH MS vastly
extends the number of measurable peptides and proteins in a single MS injection compared to SRM, and at the same time
manages to maintain the major performance advantages of targeted proteomics. It is therefore optimally suited to acquire highquality proteome-wide data in a reproducible way over large sample cohorts.
In this lecture I will demonstrate how SWATH MS has been developed and benchmarked over the last years, how it performs in
comparison to SRM and how reproducible SWATH data can be generated across laboratories worldwide. In order to significantly
simplify the application of SWATH MS, several organism-specific comprehensive SWATH assay libraries have been generated and
are freely available, including Mycobacterium tuberculosis, Saccharomyces cerevisiae and Homo sapiens. Finally, I will highlight
the potential of SWATH MS based on specific biological applications, for example to study the absolute proteome composition
and proteome dynamics in Mycobacterium tuberculosis, for applications in clinical cancer proteomics, to investigate dynamics in
protein complex formation or to elucidate post-translational modifications.
In conclusion, SWATH MS offers unprecedented possibilities for the quantitative analysis of proteomes and their modification
states and it provides new possibilities to re-mine the once-and-forever acquired SWATH data sets to test new hypotheses and
to discover new features of the proteome.
Proteomic Forum 2015
9
Figure 1
ED2
Label-free quantification using ion-mobility enhanced data-independent acquisition (UDMSE)
1
1
1
1
U. Distler , J. Kuharev , P. Navarro , S. Tenzer
1
Universitätsmedizin Mainz, Institute for Immunology, Mainz, Germany
With the arrival of high-resolution instruments in proteomics laboratories, precursor-intensity based label-free quantification is
rapidly gaining popularity. In addition to classical data-dependent
acquisition (DDA) workflows for protein identification, dataE
independent acquisition (DIA) schemes, including MS , SWATH and All Ion Fragmentation are increasingly used. These do not
suffer from inherent limitations of
DDA approaches, such as irreproducible precursor ion selection and
under-sampling.
Recently, traveling wave based ion mobility separation
(IMS) has been introduced into commercially available high-resolution
instruments, providing an additional dimension of separation without
increasing acquisition time. However, the significant
increase in system peak capacity so far failed to deliver the promise of deeper proteome
coverage. We have developed an
optimized IMS-enhanced DIA-based discovery
proteomics workflow, which employs drift-time specific collision energy profiles
that enhance precursor fragmentation efficiency, thereby substantially improving proteome coverage in data-independent
acquisition proteomics. To evaluate performance, we used tryptically digested
HeLa lysate, which was separated by nanoUPLC
and analyzed on a Waters
Synapt G2-S mass spectrometer. Rawdata processing and protein identification was performed by
PLGS3.0. Data were then processed using ISOQuant, a software pipeline for processing and analysis of IMS-based DIA data,
which integrates retention time alignment, clustering, multidimensional normalization and TOP3-based absolute quantification,
which increased identified proteins by >20%. Notably, cross-annotation increased the percentage of proteins identified in 3/3
replicates to 97%. Using a 180 min nanoUPLC gradient, our IMS-DIA-based workflow enabled the identification of >4400
proteins from
only 300 ng of HeLa tryptic digest (at <1% FDR, requiring 2 peptides per
protein). Additionally, high precision
and accuracy of DIA-based
label-free quantification were verified using a hybrid proteome sample.
a
Distler U, Kuharev J, Navarro P, Levin, Y, Schild H, Tenzer S (2014), “Drift-time specific collision energies enable deep coverage
data-independent acquisition proteomics”, Nat Methods 11(2): 167-70.
Proteomic Forum 2015
10
Session II
12:00–13:30
Chair: Carina Ramallo Guevara (Bochum/DE)
ED4
Targeted quantification by SRM and PRM
1
B. Domon
1
L.I.H., Luxembourg Clinical Proteomics Center, Strassen, Luxembourg
The targeted analysis of proteins in biological samples are routinely performed on triple quadrupole mass spectrometers
operated in selected reaction monitoring (SRM) mode. The technique is sensitive, selective and offers a wide dynamic range.
However, the low resolution of quadrupole mass filters have limited selectivity, which is an issue for the analysis of complex
samples, where the background interferes with the signals of the analytes. Thus, hybrid mass spectrometers with high resolution
and accurate mass (HRAM) capabilities overcome this limitation, and have opened new avenues in quantitative proteomics.
The targeted analyses of biological samples carried out using the parallel reaction monitoring (PRM) technique implemented on
a quadrupole-orbitrap mass spectrometer have demonstrated a significant gain in selectivity, while the assignment of the
fragment ions by accurate mass increased the confidence. The analyses in PRM mode showed better quantification performance
for peptides present in low amount in bodily fluids. This translated in more consistent quantitative data for the different
peptides of the same protein, and a clear discrimination between the control and patient samples. Furthermore, optimal PRM
performance was obtained through the control of the acquisition by monitoring the analytes in real-time and dynamically
adjusting the parameters.
Session III
14:00–15:30
Chair: Christine von Törne (Munich/DE)
ED5
The nuts and bolts of performing top down MS and top down proteomics
N.Kelleher
Evanston, IL/US
TBA.
ED6
Bioinformatic challenges in quantitative proteomics
O. Kohlbacher
Tübingen, Germany
TBA.
Proteomic Forum 2015
11
Scientific Program
22 – 25 March 2015
Sunday, 22 March 2015 ........................................................................................................................................................ 12
News Corner ..................................................................................................................................................................................................12
Opening Lecture ............................................................................................................................................................................................18
Monday, 23 March 2015 ....................................................................................................................................................... 19
Plenary Session • New Technologies .............................................................................................................................................................19
Plenary Session • Post-Translational Modifcations and Pathways ................................................................................................................20
Symposium • Affinity Proteomics .................................................................................................................................................................20
Symposium • Gene Regulation & Epigenetics ...............................................................................................................................................22
Symposium • Plant & Agricultural Proteomics I ............................................................................................................................................24
Symposium • Disease Proteomics .................................................................................................................................................................26
Symposium • Imaging ...................................................................................................................................................................................28
Symposium • Microbial Proteomics ..............................................................................................................................................................30
Tuesday, 24 March 2015 ....................................................................................................................................................... 32
Plenary Session • Interaction Proteomics .....................................................................................................................................................32
Plenary Session • Subcellular Proteomics .....................................................................................................................................................33
Symposium • Host-Pathogen Interaction & Antibiotical Resistance .............................................................................................................34
Symposium • Plant & Agricultural Proteomics II ...........................................................................................................................................36
Symposium • Quantitative Strategies ...........................................................................................................................................................38
Young Investigator Session ............................................................................................................................................................................40
Wednesday, 25 March 2015 ................................................................................................................................................. 46
Plenary Session • Systems Biology ................................................................................................................................................................46
Plenary Session • Proteomics in Health and Disease ....................................................................................................................................47
Symposium • Computational Omics I............................................................................................................................................................48
Symposium • Extracellular Vesicles...............................................................................................................................................................50
Symposium • Post-translational Modifcations ..............................................................................................................................................51
Symposium • Computational Omics II...........................................................................................................................................................53
Symposium • Structural Proteomics .............................................................................................................................................................54
Closing Lecture ..............................................................................................................................................................................................56
Proteomic Forum 2015
12
Sunday, 22 March 2015
News Corner
16:00–17:30
Chair: Roland Kellner (Darmstadt/DE)
NC1
A Universal Method for Peptide Identification on Orbitrap Fusion MS
1
P. Saliba
1
Thermo Fisher Scientifics, Dreieich, Germany
Introduction: Achieving the maximum identifications from different peptide samples requires optimization of MS methods.
Optimizations are time/sample intensive to determine the best balance of scan rate and number of ions per spectrum. This is
particularly true when accurate sample concentration, complexity, and dynamic range are unknown, often the case following
fractionation/enrichment. Unfortunately, optimizations are often not performed due to sample/time restraints. Here we present
a universal method which adjusts parameters “on-the-fly” according to spectral complexity/intensity, eliminating the
requirement for optimization.
Methods: Here, we analyzed various samples including HeLa digests and immunoprecipitations. Analysis was performed on an
Orbitrap Fusion MS. The resulting LC-MS/MS data were searched using Proteome Discoverer, matches were filtered to 1% FDR.
Each sample was analyzed with varying ion targets and maximum injection times to determine optimal parameters.
Novel instrument control software, now implemented on Orbitrap Fusion MS, was used to develop a Universal Method which
makes “on-the-fly” decisions about length of injection time per precursor based on the ion flux, complexity of full scan and
available cycle time without user input.
Results: Maximum identifications are obtained by reaching a balance between scan rate and quality of spectra. With 1
microgram HeLa digest, Orbitrap Fusion achieves maximal identifications (~25,000 unique peptide identifications (UPI)) using 35
ms maximum injection time and 1e4 ion target. At 1 ng, however, maximal identifications are achieved (~700 UPI) using 500 ms
maximum injection time and 1e4 ion target. Depending on the sample load, complexity, and dynamic range, optimal values
change dramatically. A single Universal Method achieved maximal identifications in all sample types.
Conclusions: Our results show that it is possible to achieve maximal peptide identifications from samples with unknown
concentrations without method optimization and lengthy reanalysis, thereby, increasing the throughput of the instrument while
simultaneously improving the quality of the data acquired.
NC2
Systematic Development of Targeted Proteomics Assays using Expressed Prostatic Secretions: Liquid Biopsies to Identify
Extracapsular Prostate Cancers
1
T. Kislinger
1
Princess Margaret Cancer Center, Toronto, Canada
Question: Current prostate cancer (PCa) prognostic factors stratify patients into risk groups, but are inaccurate in predicting
outcome, resulting in over-treatment of many men with indolent disease. In addition, men on active surveillance are required to
undergo repeated needle biopsies, subjecting them to associated risks. A pressing need in PCa management is the development
of improved prognostic factors that enable follow-up of men with low-risk disease in a non-invasive manner.
Methods: We have systematically developed a proteomics strategy utilizing non-invasively collected prostate-proximal fluids,
known as expressed prostatic secretions (EPS), as a novel source of PCa biomarkers. Initially, clinically annotated EPS samples
were analyzed by various shotgun proteomics strategies to develop an atlas of detectable proteins. Selected candidates were
then quantified using multiple reaction monitoring mass spectrometry (MRM-MS).
Results: Based on comprehensive proteomics profiling of stratified EPS fluids, we selected 232 proteotypic peptides for the
systematic development of MRM-MS assays to be tested in post-DRE urines (i.e. voided urine following a digital rectal
examination - EPS-urine). Rigorous selection criteria enabled us to develop MRM-MS assays for 147 peptides using crude stable
Proteomic Forum 2015
13
isotope-labelled peptides. MRM assays were applied to a test cohort of EPS-urines, stratified as fluids from controls (n=26),
organ-confined (n=42) and extracapsular (n=16) patients. Using statistical evaluations we selected 40 peptides for further assay
optimization (i.e. calibration curves, LOD and LOQ determination, multiplexed scheduled assays) using highly purified AQUA
peptides. As a final step of this project we are currently applying this multiplexed MRM assay to an independent verification
cohort of EPS-urines stratified as biopsy negative controls (n=50), BPH (n=75), organ-confined (n=70) and extracapsular (n=42)
patients.
Conclusion: The current program combined comprehensive proteomic profiling of expressed prostatic secretions with the
systematic development of MRM-MS assays using clinically relevant fluids. Our ultimate goal is to develop signatures, to be used
as liquid biopsies in the context of active surveillance, that distinguish organ-confined from extracapsular prostate cancers.
References:
Principe S, et al. In-depth proteomic analyses of exosomes isolated from expressed prostatic secretions in urine. Proteomics.
2013;13(10-11):1667-71.
Kim Y, et al. Identification of differentially expressed proteins in direct expressed prostatic secretions of men with organconfined versus extracapsular prostate cancer. Mol Cell Proteomics. 2012;11(12):1870-84.
Principe S, et al. Identification of prostate-enriched proteins by in-depth proteomic analyses of expressed prostatic secretions in
urine. J Proteome Res. 2012 6;11(4):2386-96.
Drake RR, et al. In-depth proteomic analyses of direct expressed prostatic secretions. J Proteome Res. 2010 7;9(5):2109-16.
NC3
Biophysical analysis of molecular interactions with electro-switchable DNA nanolevers on a chip
1
U. Rant
1
Dynamic Biosensors GmbH, Planegg, Germany
Measurements in stationary or mobile phases are fundamental methods in protein analysis. Although the immobilization of
molecules on solid supports allows for the parallel analysis of interactions, properties like size or shape are usually inferred from
the molecular mobility under the influence of external forces. However, as these principles are mutually exclusive, a
comprehensive characterization of proteins usually involves a tedious multi-step workflow.
Here we show how these measurement modalities can be reconciled by tethering proteins to a surface via dynamically actuated
nanolevers. Short DNA strands, which are switched by alternating electric fields, are employed as capture probes to bind target
proteins. By swaying the proteins over nanometer amplitudes and comparing their motional dynamics to a theoretical model,
the protein diameter can be quantified with Angstrom accuracy. Alterations in the tertiary protein structure (folding) and
conformational changes are readily detected, and even post-translational modifications are revealed by time-resolved molecular
dynamics measurements.
We demonstrate the analysis of interactions involving proteins, nucleic acids, and small molecules with exceptional sensitivity,
i.e. the quantification of dissociation constants (KD) in the femto-molar concentration regime, and the real-time measurement of
association and dissociation rate constants (kon, koff). High-affinity and complex interactions are characterized employing variable
capture molecule densities and DNA nanostructures. Thermodynamic energies (G, H, S) and melting temperatures are quantified
from temperature dependent binding studies.
We discuss examples including different antibody formats, protein-protein interactions, DNA-binding proteins like polymerases,
DNA/RNA detection with single mismatch specificity, and conformation changes in proteins induced by small molecule binding.
References: Nat Comm 4:2099 (2013) | JPC B 118:597 (2014) | JACS 134:15225 (2012) | Bioanal Rev 4:97 (2012)
NC4
Individual pharmacokinetics of thiol-containing drugs can be estimated by their interaction with the serum protein
transthyretin
1
1
2
2,3
1
1
E. N. Curo Gutierrez , J. Raila , M. Kemper , A. F. Pfeiffer , F. J. Schweigert , A. Henze
1
University of Potsdam, Institute of Nutritional Science, Physiology and Pathophysiology of Nutrition, Nuthetal, Germany
2
German Institute of Human Nutrition, Department of Clinical Nutrition, Nuthetal, Germany
3
Charité-University Medicine, Department of Endocrinology, Diabetes and Nutrition, Berlin, Germany
Proteomic Forum 2015
14
Introduction: In terms of personalized medicine the knowledge of the individual substance-specific pharmacokinetic is desirable.
However, determination of pharmacokinetics is time-consuming and material-intensive. For thiol-containing drugs these
problems might be overcome by the use of their interaction with thiol groups of serum proteins. Therefore, we analyzed the
interaction of the thiol-containing drugs D-penicillamine, N-acetylcysteine (NAC), and captopril with the serum protein
transthyretin (TTR) in vitro and for NAC also in vivo.
Patients & Methods: Human serum was incubated with different concentrations of D-penicillamine, NAC, and captopril for 2h at
37°C. Subsequently, TTR was isolated by immunoprecipitation and modifications were analyzed by MALDI-TOF-MS using the
linear positive mode. The AUC of TTR modified by thiol-containing drugs was determined and expressed as percentage of the
total AUC of TTR. For the analysis of the NAC-TTR interaction in vivo 19 healthy volunteers (18-54 years of age, 7 male and 12
female) consumed 1200 mg NAC dissolved in water per os. Before as well as 1h, 2h, and 4h after NAC intake blood was drawn
and TTR was analyzed as for the in vitro incubations. Additionally, NAC serum concentration was determined in these samples by
HPLC and the pharmacokinetic parameters maximal serum concentration (Cmax) and mean residence time (MRT) were estimated.
Results: TTR provides as single cysteine residues, which is not involved in any intra- or interprotein disulfide bond, but which is
excessively modified by naturally occurring thiols such as cysteine, cysteineglycine and glutathion, resulting in the formation of
specific adducts (Figure 1A). The incubation of serum with D-penicillamine, NAC, and captopril resulted in the formation of
additional specific adducts with a mass shift in comparison to the unmodified TTR of +149 Da, +163 Da and +217 Da,
respectively. The AUC of the specific adducts correlated with the amount of D-penicillamine, NAC, and captopril used for the
incubation (Figure 1B-D).
The oral intake of 1200 mg NAC resulted in an increase of NAC serum concentration with an average C max of 37.7±17.4 µmol/L
and an average MRT of 5.39±0.83 h. The oral intake of NAC also resulted in the formation of a specific adduct with TTR in a timedependent manner (Figure 2). The AUC of the NAC-TTR adduct formed 1h (3.02±1.25%), 2h (2.92±1.41%), and 4h (2.25±0.79%)
after oral intake was correlated with the Cmaxof NAC (r=0.647, r=0.753, and r=0.547, respectively, p
Conclusion: The interaction of thiol-containing drugs with TTR is a potent, simple and fast tool to estimate the individual
pharmacokinetics of these substances and might be useful for therapy optimization.
Figure 1
Figure 2
Proteomic Forum 2015
15
NC5
A Novel 6x5 Peptide Mixture for LC-MS/MS performance monitoring
1
1
2
M. Rosenblatt , E. Strauss , C. End , M. Uhr
1
Promega Corp., Madison, United States
2
Promega, Mannheim, Germany
1
Performance monitoring of LC-MS/MS instrumentation continues to be challenging and has yet to be standardized across all
Mass Spectrometry laboratories. In addition there is a specific need for a reliable reagent to aid in method development and
optimization. Towards that end, we have developed a peptide mixture containing 30 peptides: 6 peptide sets of 5 isotopologues
of the same peptide sequence. The 6 different peptide sequences are tryptic peptides having a different degree of
hydrophobicity to monitor the performance of an LC. The isotopologues differ 1. by mass due to a different number of stable,
heavy-labeled amino acids incorporated into the sequence and 2. by molar abundance in the mixture. In fact the isotopologues
of each peptide are present over 5 logs in molar abundance. This range (5 logs) allows assessment of the mass spectrometer
instrument’s dynamic range and sensitivity. To assist in data processing, we developed as well a complementary software tool
TM
(PReMiS ) that report on LC data (retention time, peak height, peak width) and on MS data (sensitivity, dynamic range, mass
accuracy). In addition to reporting on instrument parameters, the software can also report on parameter history, compare
instruments, and display XIC traces for the peptides. This is the first example of a peptide mixture designed to give a report of all
critical LC and MS parameters in a single run.
Figure 1
NC6
Application of Oncoproteomics in Changing the Treatment Paradigm in Acute Lymphoblastic Leukemia (ALL)
1
1
2
3
E. Lopez Villar , L. Madero , W. Cho , X. Wang
1
Hospital Universitario Infantil Niño Jesús, Oncohematology, Clinical proteomics Research, Madrid, Spain
2
Department of Clinical Oncology, Queen Elizabeth Hospital, Hong Kong, Hong Kong, China
3
Biomedical Research Centre, Fudan University Zhongshan Hospital, Shanghai,Department of Respiratory Medicine, Zhongshan
Hospital Fudan University School of Medicine, Shanghai Respiratory Research Institute, Shanghai, China and Lund University,
Sweden,, China, China
ALL treatments tend to be personalized via proteomic innovations Current adjustment ALL parameters include single
nucleotide polymorphisms available with the completion of the human genome project and the potential of such markers (Table
1) in predicting patient responses (1) (2). A shotgun oncoproteomic strategy, space and time-based, using sequential elution
from immobilized metal affinity chromatography (3) and isobaric tag for relative and absolute quantitation (4) coupled to liquid
chromatography, electrospray ionization, tandem mass spectrometry (5, 6) is proposed to be applied in our research team by
the Spanish Health System in collaboration with international research teams to unravel signatures of ALL patients who do not
respond well to treatments compared to those who do. We expect this research strategy will help to improve current ALL
therapies, especially for ALL patients with bad prognoses (Figure 1). Conclusion Network-proteins with distinctive characteristics
for ALL patients, unrevealed through cytogenetics, are currently an important source of novel therapeutic targets that emerge
from shotgun-proteomics. The ALL evolution per each patient can be studied via shotgun oncoproteomics, thus ALL patients can
be benefited (7)
Proteomic Forum 2015
16
1. López Villar E, et al.Proteomics-based discovery of biomarkers for paediatric acute lymphoblastic leukaemia: challenges and
opportunities. J Cell Mol Med. 2014 Jul;18(7):1239-46.
2. Cunningham, L.; Aplenc, R. Pharmacogenetics of acute lymphoblastic leukemia treatment response. Expert Opin.
Pharmacother. 2007, 8 (15), 2519-31.
3. Thingholm TE, Jensen ON, Robinson PJ, Larsen MR.SIMAC (sequential elution from IMAC), a phosphoproteomics strategy for
the rapid separation of monophosphorylated from multiply phosphorylated peptides. Mol Cell Proteomics. 2008;7(4):661-71
4. Pflieger D, Jünger MA, Müller M, Rinner O, Lee H, Gehrig PM, Gstaiger M,Aebersold R. Quantitative proteomic analysis of
protein complexes: concurrent identification of interactors and their state of phosphorylation. Mol Cell Proteomics.
2008;7(2):326-46.
5. Choudhary C, Mann M.Decoding signalling networks by mass spectrometry-based proteomics. Nat Rev Mol Cell Biol. 2010
Jun;11(6):427-39.
6. Sharma K, D'Souza RC, Tyanova S, Schaab C, Wiśniewski JR, Cox J, Mann M.Ultradeep human phosphoproteome reveals a
distinct regulatory nature of tyr and ser/thr-based signaling.Cell Rep. 2014 Sep 11;8(5):1583-94.
7. Application of Oncoproteomics to Aberrant Signaling Networks in Changing the Treatment Paradigm in Acute Lymphoblastic
Leukemia. E Lopez Villar*, et al; Journal: Journal of Cellular and Molecular Medicine; DOI XX98SSBTK PUBMED 2015 (accepted
Nov 2014; in Pubmed January 2015) († Dedicated to Prof Juan Pablo Albar; His valuable suggestions, high professional and human level, contributed greatly for this important
scientific work PI13/02475)
Figure 1
Figure 2
Proteomic Forum 2015
17
NC7
®
Capillary LC-MSMS on TripleTOF mass spectrometry systems for proteomic applications
1
2
N. Morrice , C. Baumann
1
AB Sciex, Phoenix House, Warrington, United Kingdom
2
AB Sciex, Darmstadt, Germany
For the past 15 years proteomic analyses by LC-MS has relied on nano-HPLC coupled to an optimised nanospray source. This
setup has either involved the use of packed nanospray emitters (20-75mm i.d) or packed columns with an internal frit coupled to
TM
a fused silica or metal nanospray emitter. Recent developments with microfluidic devices such as the Eksigent cHiPLC have
made nano-HPLC more user friendly and reliable. However, nano-HPLC systems are run at low flow rates (200-300nl/min),
resulting in long column equilibration and sample loading times, so the technique is not ideally set up for high sample
throughput.
Capillary bore columns (0.2-0.5mm) offer an attractive alternative at flow rates between 1-15ml/min, but conventional ESI
sources have too much internal dead volume causing excessive peak broadening, which compromises the separations obtained.
This has meant that capillary HPLC columns have been largely ignored for LC-MS applications. However Eksigent have developed
TM
TM
low diameter metal probes to fit the AB-Sciex TurboV and Duospray ESI sources and these were tested on 5600 and 6600
TripleTOF mass spectrometers coupled to either an Eksigent MicroLC200 or ekspert™ nanoLC 425 systems. In LC-MSMS mode
with a 150x0.3mm column at 5ml/min, protein digests could be successfully analysed to sub-fmol column loading and narrow
peak widths (5-15s) were obtained. When the system was compared to the nano-HPLC setup with a nanospray source, very
similar results were obtained with simple mixtures and total cell lysate digests. For example over 2000 proteins were identified
from the analysis of 1mg of a HeLa cell lysate digest by microflow LCMSMS using a 60min gradient.
The main advantage of this system was the total analysis time, which could be reduced from 45min (nanoHPLC) to 10 min
(capillary HPLC) for simple mixtures such as in-gel digests of colloidal coomassie stained bands. Complex cell lysate digests could
be analysed with a total runtime of less than 1hr as compared to 2.5hrs when using nano-HPLC. All the analyses using capillary
HPLC were performed with direct injection and the observed retention time reproducibility was very high, making this an
TM
attractive HPLC platform for label free quantification experiments such as Swath or scheduled SRM analyses. Setting up
microflow LC-MS is very simple and it’s performance doesn’t suffer from small dead volumes (such as between the column and
the electrospray source) that have such a detrimental effect on the performance of nano-LCMSMS . In a busy proteomics core
facility, this capillary LC-MS set-up would improve throughput dramatically and has potential to be used for large scale clinical
analyses in a realistic timeframe.
NC8
Early warning signals of aging in the hematopoietic stem cell niche.
1
1
2
2
1
M. Hennrich , F. Ye , P. Horn , A. Ho , A.- C. Gavin
1
EMBL, Structural and Computational Biology, Heidelberg, Germany
2
University Medicine V, Heidelberg, Germany
Introduction: Aging and the limitation of life has always been in the interest of mankind. Aside the socio-cultural aspect of aging,
the medical interest on the process of getting older will touch almost everyone in our society, as certain common diseases are
predominantly arising when we are aged.
Objectives: In the study we are conducting at the moment; we are interested in the aging of the hematopoietic system in
humans. The hematopoietic system is susceptible to certain types of cancers, when we reach a certain age. Myelodysplastic
syndromes (MDS or myelodysplasia) or B-cell chronic lymphocytic leukemia (B-CLL), also known as chronic lymphoid leukemia
(CLL) are two examples of hematopoietic diseases that are predominantly diagnosed at ages above 60 years. Our objective is to
first characterize the normal perturbations of the hematopoietic system during aging and subsequently compare perturbations
in certain diseases with the regular aging.
Material and Methods: We analyze cells originating from human bone marrow, sampled from the iliac crest. The age of the
donors spans from 20 to 60 years. Isolated cells are lysed and the proteins are digested with Lys-C followed by trypsin. The
resulting peptides are labeled with the tandem mass tag (TMT) and five donor samples and one internal standard are mixed
Proteomic Forum 2015
18
together. This mix is separated by reversed phase chromatography at basic pH and fractions are collected. The fractions are
analyzed via a nanoflow reversed phase liquid chromatography system directly coupled to a high resolution mass spectrometer.
The identification of peptides is conducted with standard software and the results are loaded into a database. Quality controls,
filtering, quantification and further analysis are performed based on the structured query language (SQL).
Results: Several protein groups in the hematopoietic system change in abundance during our lifespan. This leads to certain
changes in the bone marrow itself and has an impact on the outcome of cells in number and cell type.
Conclusions: The hematopoietic system is subject to molecular changes during our lifespan. The changes in the proteome are
not random, but many of the proteins whose abundances change during aging are described to interact with each other. Our
study will unravel some of the important mechanisms in the human hematopoietic system. This knowledge will allow future
scientists to focus on the mechanisms important in the aging of the human hematopoietic system.
NC9
Multiplexed mass spectrometric analysis of proteostasis elucidates the mode-of-action of cancer drugs
1
1
1
1
1
1
1
M. M. Savitski , N. Zinn , M. Faelth-Savitski , D. Poeckel , S. Melchert , H. Franken , G. Bergamini , M. Bantscheff
1
Cellzome GmbH, Heidelberg, Germany
1
Most targeted (anti-cancer) therapies aim at modulating the activity of dysregulated cellular pathways by inhibiting the signaling
or enzymatic activities of key pathway components. Alternative strategies directly affect proteostasis of target proteins by
down-regulation of expression or by enhancing degradation, e.g. HSP90 inhibitors, Lenalidomide. In some cases enzymatic
inhibitors have been modified to also enhance degradation with potential implications on efficacy in e.g. cancer therapies.
Transcriptional profiling has become a standard tool for studying and potentially optimizing the effects of bioactive compounds
on protein expression, however, activities on translational regulation and regulation of protein degradation are only observed on
proteome level and difficult to distinguish when only analyzing protein abundances. Here we introduce a quantitative mass
spectrometry-based proteome profiling method that for the first time allows differentiating regulation on protein synthesis level
from protein degradation in a time, stimulus and compound-dependent way. The combination of pulsed-SILAC labeling with
isobaric mass tagging enables the multiplexed analysis of up to 5 conditions in two full biological replicates in a single LC-MS/MS
experiment. This enabled a comprehensive analysis of HSP90 client proteins using the ATP-competitive drug candidate 17-AAG
and differentiation of clients requiring the chaperone on nascent level from those clients permanently dependent on interaction
with the chaperone in different cell lines and cultured primary cells. Further, we investigated synergistic effects of HSP90
inhibition with a set of estrogen receptor antagonists in the breast cancer cell line MCF-7.
Opening Lecture
17:30 –19:00
Chair: Albert Sickmann (Dortmund/DE)
PL01
Understanding diseases using mass spectrometry
J. Yates
San Diego, CA, United States
TBA.
Proteomic Forum 2015
19
Monday, 23 March 2015
Plenary Session • New Technologies
09:00–10:30
Chairs: Oliver Pötz (Reutlingen/DE), Christian Huber (Salzburg/AT)
PL02
Capillary electrophoresis for bottom-up proteomics
N. J. Dovichi
Notre Dame University, Indiana, United States
TBA.
PL03
On the value of proteoform resolved measurements
1
N. Kelleher
1
Northwestern University, Evanston, United States
In a manner similar to the Human Genome Projects cataloging of all the genes in the human body the Proteome Project seeks to
identify the proteins present in the human population. To that end it has become essential to optimize the way we handle the
complexity of the variety of the human proteome, both in terms of proteoforms as well as protein complexes. Here we describe
recent advances to develop new approaches to measure protein complexes by Top Down tandem MS, whose varied structures
are foundational to their ability to perform biological functions. We have optimized the use of native gels to produce samples
compatible with downstream analyses, including native, multistage mass spectrometry (nMS/MS). We have found that a
continuous system (i.e., no stacking gel) with a gradient in its extent of cross-linking and use of the clear native buffer system
performs well for both fractionation and native mass spectrometry of extracts from heart mitochondria and a fungal secretome.
This integrated advance in separations and nMS/MS, offers the prospect of untargeted proteomics at the next hierarchical level
of protein organization in biology. A computational approach to identify specific isoforms within multi-protein assemblies will
also be described. We combined the CORUM and UniProt databases to create candidates for an error-tolerant search engine
designed for top-down analyses of protein complexes by native mass spectrometry.
PL04
Mass Spectrometry Strategies for Rapid and Artifact-Free Antibody and Hemoglobin Structure Analysis
1
Y. Tsybin
1
EPFL, Lausanne, Switzerland
Top-down and middle-down mass spectrometry and proteomic approaches are currently under validation as complementary
techniques to bottom-up proteomics for protein qualitative and quantitative analysis. The primary rationale behind their
development is in the unique information they can provide - proteoform-level proteome analysis. Furthermore, avoiding
enzymatic digestion or performing it quickly and at proper (physiological buffer) conditions reduces the risks of digestionintroduced artifacts present in bottom-up approach. Here, we will overview the current state-of-the-art in middle-down and topdown approaches and showcase their development and application via monoclonal antibody and hemoglobin analysis.
Biological sample complexity requires further innovations in mass spectrometry methods and techniques. Particularly, increased
speed of high-resolution data acquisition and improved mass accuracy are needed to further advance top-down and middledown approaches. We will thus present some of our recent developments targeting these limitations.
Proteomic Forum 2015
20
Plenary Session • Post-Translational Modifcations and Pathways
11:00–12:30
Chairs: Albert Sickmann (Dortmund/DE) , Sasha Singh (Boston, MA/US)
PL05
Computational analysis of mitochondrial protein dynamics in heart failure
P.Ping
Los Angeles, United States
TBA.
PL06
Comprehensive strategy for the study of PTMs
M. Røssel Larsen
Odense, Denmark
TBA.
PL07
Exploring alternative proteases and fragmentation methods for proteomics
1
A. Heck
1
Utrecht University, Utrecht, Netherlands
Proteome analysis heavily relies on a broad mixture of analytical techniques, starting from sample preparation, to separation
and enrichment and last but certainly not in the least mass spectrometry. Through developments in these enabling technologies
MS-based proteomics has matured and start to deliver biological relevant information. Still our measurements are still far from
comprehensive, and the optimal matured workflow does not (yet) exist for proteomics. In my laboratory we try to explore
alternative complementary methods to explore better the richness of the proteome
In more detail I will describe methods enabling 1) enhanced peptide coverage based on a combination of HCD and ETD that
1
2
leads to improved analysis of phosphopeptides, MHC peptides and protein disulfide bridges , 2) enhanced protein and
proteome coverage, by using multiple proteases and 3) improved analysis of intact proteins and protein complexes enabling to
visualize all proteoforms of a given protein at once and following kinase phosphorylation in real-time using top-down
3
approaches .
1. Mommen et al. PNAS 111 (2013) 4507-4512
2. Liu et al. Mol Cell Prot 13 (2014) 2776-2786
3. Van den Waterbeemd et al. Angew Chemie 53 (2014) 9660-9664
Symposium • Affinity Proteomics
14:00–15:00
Chair: Matthias Mann (Martinsried/DE)
O01
Quantitative chemical proteomics identifies the interactome of PDE3A in cancer cells
1
1
1
1
1
E. Corradini , G. Klaasse , U. Leurs , N. I. Martin , A. Scholten , A. J. R. Heck
1
Utrecht University, Utrecht, Netherlands
1
The second messengers cAMP and cGMP mediate in transmitting a wide variety of external signals. Both cyclic
nucleotides are degraded by the superfamily of phosphodiesterases (PDEs) consisting of more than 50 different
isoforms. Several of these are implicated in disease processes and therefore a plethora of selective PDE inhibitors
have been synthetized over the past two decades. Unfortunately, many of these inhibitors failed in clinical trials. This
may be partially caused by their pharmacological action. At the same time, accumulating data suggests that even
Proteomic Forum 2015
21
small differences between different PDE isoforms results in specific tissue distributions, cellular localization and
different involvement in higher order signal protein complexes. This latter effect has only been marginally addressed
and no screening methodology is available to address this in a more comprehensive way.
Affinity based chemical proteomics is a relatively new tool to identify specific protein-protein interactions. Here, to
study the interactome of PDEs, we synthesized a broad spectrum PDE-capturing resin based on the non-selective
PDE inhibitor 3-isobutyl-1-methylxanthine (IBMX). Chemical proteomics characterization of this resin in HeLa cell
lysates revealed capture of several different PDE families.
Combining the IBMX-resin with in-solution competition with the more selective PDE inhibitors, cilostamide and
papaverine, revealed the interactome of PDE3A in HeLa cells. Besides known interactors such as the 14-3-3 proteins,
a PP2A complex composed of a regulatory, scaffold and catalytic subunit was identified.
The design, synthesis, optimization and applicability of this novel IBMX-resin shows great potential for the future
elucidation of signaling nodes of different PDEs in a variety of cell and tissue types.
O02
Deciphering preferential interactions within supramolecular protein complexes: the proteasome case
1
1
1
1
2
1
1
1
2
B. Fabre , T. Lambour , L. Garrigues , F. Amalric , N. Vigneron , T. Menneteau , A. Stella , B. Monsarrat , B. Van den Eynde , M.1
P. Bousquet-Dubouch
1
IPBS CNRS - University of Toulouse, Structural Biology and Biophysics, Toulouse, France
2
Ludwig Institute for Cancer Research, Brussels, Belgium
Question:One challenge of systemic biology is to characterize the complex nature of supramolecular protein machineries.
Proteasome particles display high heterogeneity because they are formed by the dynamic association of several sub-complexes,
a 20S core particle (20S CP), either single or associated to one or two regulatory particles (RPs) of identical or different protein
composition. The 20S CP, which presents anα7β7β7α7 barrel-like structure, can itself be found in the eukaryotic cell as four
different sub-types, the major forms being the standard proteasome and the immunoproteasome which differ in their catalytic
beta subunits composition.
Identifying preferential associations within proteasome sub-complexes would therefore help to better characterize the
functional relevance of proteasome diversity.
Methods: Recently, protein correlation profiling associated to mass spectrometry, based on the mass separation of protein
complexes and their analysis by quantitative MS (1), was used for the determination of protein complexes and their dynamics
(2). We developed an integrated proteomic workflow based on proteasome immunopurification, label-free quantification using
high resolution MS on LTQ Orbitrap Velos (3), protein correlation profiling, and statistical methods, to resolve the various
complexes the different 20S CPs might be involved in.
Results: From a wide set of nine different human cell lines exhibiting different proteasome compositions (4) (Figure 1, Panel A),
we could highlight previously unreported preferential associations within proteasome sub-complexes. In particular, we showed
for the first time that the two main proteasome sub-types, standard proteasome and immunoproteasome, interact with a
different subset of important regulators. This trend was observed in very diverse human cell types (Figure 1, Panel B) and was
confirmed by changing the relative proportions of both 20S proteasome forms using interferon-γ (Figure 1, Panel C) (5).
Conclusions: This integrated proteomic workflow provides a valuable tool to better understand the dynamic and heterogeneous
nature of proteasomes and more generally of other biologically relevant supramolecular protein complexes.
(1) Andersen JS et al.(2003) Nature 426(6966):570-574.
(2) Kristensen AR et al.(2012) Nat Methods 9(9):907-909.
(3) Fabre B et al. (2013) Mol Cell Proteomics 12:687.
(4) Fabre B et Lambour T. (2014) J Proteome Res, 2014, 13:3027-37.
(5) Fabre B et al. (2014) Mol Syst. Biol In Press.
Proteomic Forum 2015
22
Figure 1
PL10
Unravelling global RNA-binding protein architectures by quantitative proteomics
C. Frese
Heidelberg, Germany
TBA.
Symposium • Gene Regulation & Epigenetics
14:00–15:00
Chairs: Brian T. Chait (New York, NY/US), Kathryn Lilley (Cambridge/UK)
O03
Characterization of histone PTM crosstalk by middle-down mass spectrometry and data integration
1
1,2
1
V. Schwämmle , S. Sidoli , O. N. Jensen
1
University of Southern Denmark, Odense, Denmark
2
University of Pennsylvania , Department of Biochemistry and Biophysics, Philadelphia, United States
Post-translational modifications (PTMs) of histone proteins play an important role in maintaining chromatin structure and in the
dynamic regulation of DNA replication and repair, transcription of genes and propagation of epigenetic traits. One of the largest
challenges in current chromatin biology is to characterize the relationships between co-existing histone PTMs, the order and
hierarchy of their deposition and their distinct biological functions. We developed a work flow to analyze the co-existing marks
1,2
as revealed by "middle-down" MS experiments of histone proteins . The implemented new method quantifies positive and
negative interplay between pairs of methylation and acetylation marks in proteins. Many of the detected features were
conserved within different cell lines, thereby revealing general rules for crosstalk between histone marks. The observed features
are not only in accordance with previously reported examples of crosstalk but also revealed novel types of interplay. Integration
with data coming from ChIP-seq allowed aligning the frequency of histone PTMs with MS-measured levels. Further investigation
of the involved pathways reveals a well-defined hierarchy of biological functions within single and binary histone marks. These
results show that we gather deeper insight into chromatin function by consideration of multiple marks e.g. measured with mass
spectrometry approaches.
[1] Sidoli, S. and Schwämmle, V. and Ruminowicz, C. and Hansen, T. A. and Wu, X. and Helin, K. and Jensen, O. N. Middle-down
hybrid chromatography/tandem mass spectrometry workflow for characterization of combinatorial post-translational
modifications in histones Proteomics, 2014, 14, 2200-2211
[2] Schwämmle, V.; Aspalter, C.-M.; Sidoli, S. & Jensen, O. N. Large-scale analysis of co-existing post-translational modifications
on histone tails reveals global fine-structure of crosstalk Mol Cell Proteomics, 2014, 13, 1855-1865
Proteomic Forum 2015
23
Figure 1
O04
Chromatin factors mapped by co-regulation patterns in proteomic experiments
1
1
1
1
1
G. Kustatscher , K. Wills , C. Furlan , J.- C. Bukowski-Wills , J. Rappsilber
1
Wellcome Trust Centre for Cell Biology, University of Edinburgh, Edinburgh, Germany
We present a new way of analysing complex cellular structures based on co-regulation patterns in proteomics experiments,
using chromatin as an example. Chromatin proteins mediate replication, regulate expression and ensure integrity of the
genome. A comprehensive inventory of these factors could boost our understanding of these fundamental processes. However,
chromatin has remained refractory to proteomic analyses due to its heterogeneous and dynamic composition, which makes
conclusive biochemical purification difficult, if not impossible. It defies classical organelle proteomics and cannot be described by
a single and ultimate list of protein components. Instead we propose a new approach that provides a quantitative assessment of
a protein’s probability to function in chromatin (EMBO Journal, 2014). Our analysis followed a three-stage process. (1) We
developed Chromatin Enrichment for Proteomics (ChEP), a simple biochemical procedure that allows for optimal detection of
chromatin factors (Nat Protoc, 2014). (2) We used machine learning to capture co-variation of proteins with known chromatin
factors across various biochemical and biological conditions. (3) From these data we derived the probability for each protein to
have a chromatin-related function. This resulted in interphase chromatin probabilities for 7635 human proteins, including 1840
previously uncharacterized proteins. We demonstrated the power of chromatin probabilities to predict novel chromatin factors
during analyses of CDK regulation (with Helfrid Hochegger, University of Sussex) and DNA replication (with Anja Groth,
University of Copenhagen; Nature Cell Biology, 2014). We envisage that this approach can be applied to any organelle,
compartment or other complex biological structure.
Figure 1
Proteomic Forum 2015
24
O05
Novel approaches to investigate global and locus-specific chromatin composition
1
1
J. Krijgsveld , M. R. Rafiee
1
EMBL, Heidelberg, Germany
One of the key questions in genome biology is to understand how transcription of every single gene in the genome is regulated,
and how this depends on cellular context (e.g. development or disease). Most of these functions are performed by specific and
general transcription factors in a gene-specific manner, in close interaction with a wide range of other regulatory proteins e.g.
involved in chromatin modification and/or acting downstream of signaling pathways. While several DNA-centric methods are
available to probe the localization of a protein of interest in the genome (such as ChIP-seq), currently no method exists that can
approach this question from the protein perspective. In particular, proteomics approaches are essential when aiming to
understand which proteins bind to chromatin, and with which other proteins they interact at defined genomic regions.
To address this, we have developed 3 novel and complementary methods to characterize the composition and dynamics of
chromatin in a global and locus-specific manner. In the first, we have developed an approach to specifically isolate and identify
the collective set of proteins that associate with the genome (SICAP, Selective identification of chromatin-associated proteins).
In this way we could demonstrate that chromatin in stem cells consists of >3000 proteins, the composition of which depends on
pluripotency state. In the second approach we combine SICAP with Chromatin immuno-precipitation (ChIP) to define the protein
interactions of a bait protein while being associated to DNA. This has enabled us to identify the protein network around the
pluripotency facors Oct4, Sox2 and Nanog, recapitulating the crucial players in what is known as the core circuitry of
pluripotency while extending this to a range of novel candidates. The third approach entails the identification of proteins binding
to a single genomic locus of interest (TiGr, Targeted isolation of genomic regions). Applied to the promoter region of Nanog, we
have identified many of the known proteins involved in transcriptional regulation of this pluripotency gene, as well as many
novel candidates. This includes several proteins that differentially bind to this region when stem cells are grown in the presence
of serum or in 2i-conditions. Since all three methods (SICAP, ChIP-SICAP and TiGr) are highly generic without the need for genetic
engineering, they can be applied to any other cell type, biological question, and genomic region. Thereby, they will prove
powerful tools in many areas of genome biology for the discovery of proteins conferring functionality to chromatin (transcription
regulation, structure, etc).
Symposium • Plant & Agricultural Proteomics I
14:00–15:00
Chair: Hans-Peter Braun (Hannover/DE)
PL08
Mining post translational modifications in Arabidopsis proteomics data
1
1
1
2,1
G. Mann , S. Fernández-Niño , A. Carroll , J. Heazlewood
1
Lawrence Berkeley National Laboratory, Joint BioEnergy Institute, Berkeley, Australia
2
The University of Melbourne, School of Botany, Melbourne, Australia
Introduction: The past decade has seen the development of multiple online proteomics resources in the reference plant
Arabidopsis, reflecting multiple large-scale studies. These resources existed independently and lacked any type of integration. In
an effort to rationalize these data, the Multinational Arabidopsis Steering Committee, Proteomics (MASCP) developed a
proteomics aggregation portal, MASCP Gator (http://gator.masc-proteomics.org/) (1). The portal provides a summary of
proteomics and protein information aggregated directly from ten online resources.
Objectives: The development of the MASCP Gator provided access to many millions of pieces of data for Arabidopsis proteins.
We were interested in assessing whether collectively these data could be used to uncover features within a protein. Virtually all
large-scale proteomics analyses in Arabidopsis have matched MS/MS data without modifications (2) and thus we have protein
coverage skewed towards the unmodified regions.
Materials: & Methods: The ability to locate and identify post translational modifications experimentally by mass spectrometry
is extremely challenging and there is a requirement for complementary techniques. We have used the matched MS/MS data,
known Arabidopsis PTMs, non-synonymous SNPs (nsSNP) from related accessions, orthologous regions between species and
predicted tryptic peptides to develop a bioinformatics technique to identify and display likely regions of modification in proteins
of Arabidopsis.
Proteomic Forum 2015
25
Results: Integration and display of these new data sources into the MASCP Gator was undertaken to develop a new portal where
this information could be visualized (3). Verification of data types and their role in identifying regions of modification was next
addressed. For example, an analysis of known phosphorylation sites from Arabidopsis revealed that the substitution of
experimentally derived phosphorylated Ser/Thr (nsSNPs) in a protein was statistically less likely to occur than nonphosphorylated Ser/Thr (4). Finally, using a recently published large-scale N-linked glycosylation survey as a test set, we could
demonstrate that unmatched regions represent modification hotspots in proteins.
Conclusion: We have now developed a method to locate putative regions of protein modification by exploiting mass spectral
data in the public domain. We are currently developing this into a functional portal for the assessment modifications in
proteomic datasets (http://modhunter.masc-proteomics.org/).
1. Joshi HJ, et al. (2011) Plant Physiol 155:259-270.
2. Mann GW, et al. (2013). J Proteomics 79:195-199.
3. Mann GW, et al. (2013). Front Plant Sci 4:411.
4. Joshi HJ, et al. (2012) Bioinformatics 28:1303-1306.
PL09
Understanding plant biology through functional proteomics
1
S. Baginsky
1
Plant Biochemistry, Institute of Biochemistry, Halle (Saale), Germany
Our research is focused on the characterization of basic plant biology with functional proteomics. Two lines of research are
currently prevalent: First, it is our goal to understand the assembly of organellar proteomes with a special emphasis on the
plant-specific plastids. To this end, we establish methods for absolute protein quantification (e.g. Helm et al., 2014, J Proteomics
98, 79-89) and quantify proteins in chloroplasts actively performing photosynthesis. Furthermore, we characterize the plastid
protein import machinery, assess its dynamic subunit composition and aim at identifying new interaction partners (e.g. Bischof
et al., 2011, Plant Cell 23, 3911-28; Grimmer et al., 2014, Front Plant Sci, 5, 258). On the other hand, our research aims at
deciphering the chloroplast phosphoproteome network entailing the identification of protein kinase substrates and the analysis
of phosphorylation dynamics (e.g. Reiland et al., 2011, PNAS 108, 12955-60; Lu et al., 2014, J Exp Bot, PMID: 25316064). The
tools we are using comprise comparative quantitative phosphoproteomics, targeted analyses by means of accurate inclusion
mass scanning and the peptide chip ChloroPhos1.0 that allows the multiparallel analysis of in vitro phosphorylation activity on
more than 900 substrate peptides (Schönberg et al., 2014, PlosONE, 9, e108344).
O06
Geolocalization of chloroplast proteins: an important feature for a better understanding of chloroplast metabolism
1
1
1
1
1
1
1
2
1
2
1
M. Ferro , G. Curien , M. Tomizioli , C. Lazar , S. Brugiere , T. Burger , D. Salvi , L. Gatto , L. Moyet , K. Lilley , C. Bruley , D.
1
1
1
1
Berny , Y. Vandenbrouck , G. Finazzi , N. Rolland
1
Institute for Technology and Life Sciences, Grenoble, France
2
University of Cambridge, Cambridge, France
Quantitative proteomics approaches, together with controlled organelle fractionation, allows subcellular localization to be
determined. Using such a strategy we provided the subplastidial localization of chloroplast proteins in the stroma, the
thylakoids, and envelope membranes compartments (AT_Chloro database; Ferro et al., 2010). Recently we went a step further in
the characterization of subplastidial localization by investigating the thylakoids subcompartments. The thylakoids of plants and
some green algae are structurally inhomogeneous, consisting of two main domains: the grana, which are piles of membranes
gathered by stacking forces, and the stroma-lamellae, which are unstacked thylakoids connecting the grana. The major
photosynthetic complexes are unevenly distributed, within these compartments, due to steric and electrostatic constraints, as
indicated by biochemical and spectroscopic analysis. Thus, we combined experimental approaches to perform a complete survey
of the protein composition of these thylakoid sub-compartments using thylakoid membrane fractionation. We employed
quantitative proteomics coupled with a data analysis pipeline and manual annotation to differentiate genuine grana and stromalamellae proteins from possible contaminants. Using this approach, we were able to improve the dynamic range detection of
minor thylakoid proteins. About 300 thylakoid (or potentially thylakoid) proteins were shown to be enriched in either the grana
or the stroma-lamellae fractions. Besides confirming previous information on the localization of photosynthetic proteins, these
experiments have provided some unexpected results, and have allowed us to revisit the composition of the thylakoid
microdomains (Tomizioli et al., 2014).
Proteomic Forum 2015
26
The accurate localization of proteins is a key feature in the context of functional investigations, especially to get insight over
metabolic pathways occurring in the chloroplast. Indeed the chloroplast is a complex and integrated network that produces a
high number of metabolites (e.g sugars, vitamins, lipids and pigments). One way to improve our knowledge of such a metabolic
factory is to build metabolic pathways with highly curated and integrated knowledge. As current knowledge on chloroplast
metabolism is still dispersed we have initiated the building of a chloroplast knowledge base (ChloroKB) dedicated to chloroplast
metabolic pathways. The aim of Chloro-KB is to build curated metabolic maps, to visualize these maps and to integrate “omics”
qualitative and quantitative data required for modeling purposes. In that context we have integrated in Chloro-KB proteomics
quantitative data that provide information over the subplastidial localization, a critical information to infer functional
annotations and to help data curation.
Symposium • Disease Proteomics
16:30–17:30
Chair: Andrea Urbani (Rome/IT)
O07
The oncogenic EGFR network hairball: proteomic approach for finding the tangled knot.
1
1
2
2
3
1
M. A. Jarboui , L. dernayka , C. Raso , S. Kennedy , K. Bryan , K. Boldt , M. Ueffing
1
Medical Proteome Centre, University Klinikum Tübingen, Tübingene, Germany
2
Sysytem biology Ireland, University College Dublin, Dublin, Ireland
3
Lynn Group in Biomedical Informatics,, EMBL Australia, Adelaide, Australia
1
During multicellular evolution, signaling networks adapted to allow cells to integrate cues and make cooperative decisions
toward a synchronized specific cellular outcome. Cancer cells, through mutations and alterations, hijack these mechanisms and
therefore escape cooperative rules. Accordingly, proteins interactions in signaling network, such as in the EGFR pathway, are key
elements to integrate stimuli to direct the information flow into a cellular specific decision-making response. Network-attacking
mutations dysregulate signaling networks by perturbing the proteins interaction landscape effectively creating new cancerspecific attractors. Uncovering these changes represent a challenge in cancer system biology. Addressing this challenge requires
new system-based strategies that capture intrinsic properties of cancer signaling networks. Thus, providing deeper
understanding of the process by which specific mutation (KRASG13D) perturbs the EGFR network and lead to disease phenotype
by rewiring the signaling landscape.
Using a quantitative proteomic approach based on stable isotope labeling and affinity purification in tandem with mass
spectrometry, we established a comprehensive map of the EGFR pathway interactions alteration in a KRAS oncogenic cellular
model. By investigating driver nodes in the interaction map we identified key edges. In depth driven network topology and
information flow analysis identify a set of potential proteins and proteins interactions affected by the KRAS mutation. We
further investigated dynamic changes of these proteins interactomes following EGFR signaling network perturbations.
Here we present multiple novel evidences that provide new insights into how a single mutation can create a cascade of
functional changes to circumvent signaling information flow. Accordingly, we experimentally validate interaction modulations of
selected driver nodes and investigate the potential role they play in usurping the EGFR molecular signal processing machines and
ultimately carcinogenesis. Subsequently, we explore molecular targeted approaches to circumvent interaction modulation in
KRAS mutation-driven oncogenic models as possible therapeutic strategies.
Furthermore, our data support new therapeutic potential targets for combinatory drug design based on integrative analysis of
the oncogenic EGFR interaction pathway in KRASG13D driven colorectal cancer.
O08
Altered phosphorylation and Calpain-dependent cleavage profiles of Scott syndrome platelets revealed by quantitative
proteomics
1
2
1
2
3
2
1
2
F. A. Solari , N. Mattheij , J. Burkhart , F. Swieringa , P. W. Collins , J. M. E. Cosemans , A. Sickmann , J. W. M. Heemskerk , R. P.
1
Zahedi
1
Leibniz – Institut für Analytische Wissenschaften - ISAS - e.V., Systems Analysis, Dortmund, Germany
2
Cardiovascular Research Institute Maastricht , Department of Biochemistry, Maastricht, Netherlands
3
Arthur Bloom Haemophilia Centre, Cardiff, United Kingdom
Introduction: The Scott syndrome is a rare bleeding disorder associated with a mutation in anoctamin-6 (TMEM16F). After
2+
stimulation of Ca -mobilizing agonists, syndromatic platelets show a reduced phosphatidylserine (PS) exposure and do not form
Proteomic Forum 2015
27
membrane blebs. The latter is considered to be a consequence of calpain-dependent degradation of the membrane actin
cytoskeleton and changes in protein phosphorylation.
Objective: Quantitative comparison of neo N-terminal peptides generated from endogenous calpain and quantification of
differential phosphorylation patterns between Scott patients and healthy donors.
Patients & Methods: Washed blood platelets were activated with the Ca2+-mobilizing agonists, convulxin/thrombin or
ionomycin. N-terminal ChaFRADIC (1) and iTRAQ labeling at the protein level were used to quantify N-terminal peptides.
Phosphopeptide quantification was conducted by iTRAQ labeling at the peptide level followed by titanium dioxide enrichment
(2) and HILIC fractionation.
Results: >800 N-terminal peptides and >200 endogenously acetylated N-terminal peptides were quantified from only 20 µg per
condition. Calpain specific cleavage sites were identified based on an in vitro platelet digestion derived consensus motif. From
this we deduced e.g. an enrichment of hydrophobic amino-acids N-terminal of the Calpian cleavage site. Indeed, decreased
calpain cleavage was observed in Scott platelets after convulxin/thrombin treatment, compared to controls. For instance, Src
and talin-1 showed a reduced clevage in Scott patient platelets, as confirmed by Western blot.
In addition >1500 phosphopeptides were quantified. Scott patient and control platelets showed differential phosphorylation
profiles after convulxin/thrombin treatment. Instead, this effect was more pronounced after ionomycin treatment. The top 50
altered phosphoproteins of activated Scott platelets were assigned to the following function or structure classes: actin-myosin
cytoskeleton (20%), platelet adhesion (22%), signaling/adapter proteins (24%), receptor-linked cytoskeleton (8%), and
microtubule cytoskeleton (8%).
Conclusion: Our data demonstrate that the impaired PS exposure and membrane blebbing in Scott patient is accompanied by
changes in phosphorylation patterns and likely linked to calpain-dependent cleavage patterns. Furthermore, studying changes in
Scott patients compared to healthy controls allowed us to gain novel insights into the procoagulant pathway in platelets.
(1) Venne A, et al. (2013).Novel highly sensitive, specific, and straightforward strategy for comprehensive N-terminal proteomics
reveals unknown substrates of the mitochondrial peptidase Icp55. J Proteome Res.
(2) Larsen M.R. et al (2005) Highly Selective Enrichment of Phosphorylated Peptides from Peptide Mixtures Using Titanium
Dioxide Microcolumns. Mol. Cell. Proteomics.
O09
A renal Fanconi-syndrome caused by mistargeting of a peroxisomal protein
1
2
1
1
3
2
2
1
N. Assmann , C. Broeker , K. Dettmer , P. J. Oefner , R. Kleta , R. Warth , M. Reichold , J. Reinders
1
University Regensburg, Functional Genomics, Regensburg, Germany
2
University Regensburg, Medical Cell Biology, Regensburg, Germany
3
University College London, Center of Nephrology and Institute of Child Health, London, United Kingdom
The renal Fanconi-syndrome is a disorder of the proximal tubule of the kidney to reabsorb small molecules causing urinary loss
of amino acids, glucose, electrolytes, phosphate, and low-molecular-weight proteins leading to the clinical consequence of
rickets and compromised growth. While most inherited Fanconi-syndromes are due to mitochondriopathies, we identified yet a
new form of an autosomal-dominantly inherited Fanconi-syndrome caused by mutation of the peroxisomal bifunctional enzyme
[1]. The single point mutation results in a mis-localization of the protein to the mitochondria, thereby disturbing mitochondrial
energy production. Interaction with the mitochondrial trifunctional protein leads to impaired β-oxidation and hampered
assembly of respiratory chain supercomplexes. Consequently, ATP-production in the proximal tubule cells is diminished resulting
in an energy under-supply of the cellular transporters, thus causing the renal Fanconi-syndrome. We used a combination of
genetic, proteomic and metabolomic approaches to elucidate the underlying molecular mechanism and cellular consequences of
this disease. Thereby, we identified the first pathomechanism of a mitochondriopathy caused by mis-targeting of an intrinsically
non-mitochondrial protein.
[1] Klootwijk, E.D., et al., “Mistargeting of Peroxisomal EHHADH and Inherited Renal Fanconi's Syndrome”, New England Journal
of Medicine 370 (2), 129-138 (2014)
Acknowledgements: This work was partially funded by the Deutsche Forschungsgemeinschaft, Sonderforschungsbereich 699,
Teilprojekt A11.
Proteomic Forum 2015
28
Symposium • Imaging
16:30–17:30
Chair: Gerd Schmitz (Regensburg/DE)
O10
Peptide MALDI Imaging - Searching For The Best Protocol
1
2
1
3
1
1
1
2
2
H. Diehl , B. Beine , J. Elm , D. Trede , K. Marcus , M. Ahrens , M. Eisenacher , H. E. Meyer , C. Henkel
1
Ruhr-University Bochum, Medizinisches Proteom-Center, Bochum, Germany
2
Leibniz-Institut für Analytische Wissenschaften - ISAS - e.V., Biomedical Research, Dortmund, Germany
3
SCiLS GmbH, Bremen, Germany
Objective: Mass spectrometry imaging (MSI) has become a powerful and successful tool in the context of biomarker detection
especially in recent years. This emerging technique is based on the combination of histological information of a tissue and its
corresponding spatial resolved mass spectrometric information. The identification of differentially expressed protein peaks
between samples is still the methods bottleneck. Therefore peptide MSI compared to protein MSI is closer to the final goal of
identification since peptides are easier to fragment than proteins. Nevertheless the processing of peptide imaging samples is
challenging due to experimental complexity.
Method: To address these issues a method development study for peptide MSI using cryoconserved and formalin fixed paraffin
embedded (FFPE) rat brain tissue is provided. In total 13 different experimental set ups were performed. Different digestion
times, matrices and proteases were tested to define a standard workflow for peptide MSI. All experiments were done in
triplicates and analyzed by the SCiLS Lab software, using structures derived from myelin basic protein (MBP) peaks, principle
component analysis (PCA) and probabilistic latent semantic analysis (pLSA) to rate the experiments quality. Blinded experimental
evaluation in case of defining countable structures in the data sets was done by three individuals.
Results: We have successfully demonstrated that the choice of matrix has a great impact on the outcome of the results.
Additionally the digestion time seems to be not as important as expected; nevertheless overnight digestion showed the highest
amount of MBP structures summarizing all experiments. One trypsin led to the best results in comparison to other brands and a
LysC/trypsin mixture.
Conclusion: Such an extensive method improvement study for peptide MALDI imaging is a valuable contribution to the
community. These kinds of experiments have not been performed so far and the resulting problems and consequences were
analyzed and discussed.
.
Reference: Diehl H.C., Beine B., Elm J., Trede D., Ahrens M., Eisenacher M. , Marcus K., Meyer H.E. and Henkel C. MALDI-MSI - a
comparison of different protocols to improve imaging experiments, Anal Bioanal Chem. in press
Figure Legend (out of Diehl H.C. et al)
Example of experimental setup: Comparison of matrices DHB vs. HCCA (II) using FFPE tissue digested for 2 h. Overview of the
statistic and structure analysis. (a) pLSA, only components with at least two clearly visible structures are displayed. (b) Mean of
counted structures for all visible m/z values of theoretically digested MBP. The three numbers for each experimental condition
are derived from counts of three different researchers (R1, R2, and R3). Color coding for (c) and (d): HCCA (II) in red and DHB in
blue. (c) PCA of the mean spectra and (d) PCA of the spectra group
Figure 1
Proteomic Forum 2015
29
O11
MALDI Mass Spectrometry Imaging Revealed Specific Drug Distribution of Vemurafenib against BRAF V600E in Malignant
Melanoma Tissues
1
1
1
1
1
Y. Sugihara , î Végvári , C. Welinder , T. Fehniger , B. Baldetorp , G. Marko-Varga
1
Lund University, Lund, Sweden
1
Introduction: Malignant melanoma (MM) is the leading cause of death among diseases of the skin. Vemurafenib treatment is
one of the most effective therapeutic strategies for metastatic melanoma with BRAF V600E mutation. However, the drug
efficacy depends on the patient adherence as well as the biological activity of the drug. Therefore, technology development is
still needed to predict the drug efficacy for individual patient. The presence of effective drug concentration at the target sites is
an important finding for its efficacy. MALDI-MS imaging (MALDI-MSI) is a powerful tool to evaluate the drug dose and
distribution in the tissues. Especially, in vitro experimental model with MALDI-MSI is the ideal model, which can compensate for
the insight from in vivo drug administration model. To date, specific distribution of vemurafenib has not been reported, although
evaluation of the drug localization in tumor tissue specimen has the potential for prediction of drug efficacy.
Objectives: The aim of this study was to distinguish the non-specific drug distribution from specific one on the MM tissues by
MALDI-MSI.
Patients & Methods: Six frozen MM tissues resected from lymphoid nodes were used. Nonspecific binding was determined at
100 nM concentration of unlabeled vemurafenib by co-incubation of sections with a 1000-fold excess of stable isotope labeled
vemurafenib. The mass spectra of vemurafenib were obtained using a MALDI LTQ-Orbitrap XL instrument (Thermo Scientific,
Germany). Image analysis was conducted using ImageQuest (Thermo Scientific, Germany). For histological analysis, both H&E
staining and immuno-histochemical analysis using monoclonal antibody against BRAF V600E were conducted.
Results: Vemurafenib exhibited intense distribution confined to cancer cells, that correlation was evident mostly with cancer
regions as defined by H&E staining. Non-specific binding in the tumors represented less than 10% of total binding in MM tumors.
Immuno-histochemical analysis showed that the BRAF V600E protein expression level was heterogeneous in MM tissues. There
was an excellent correlation between high intensity areas seen on the MALDI-MSI and histologically determined tumor tissue
with immuno-staining.
Conclusion: In MM tissue specimens, MALDI-MSI revealed vemurafenib binding in cancer cells and specific area; there was high
BRAF V600E immuno-staining at regions of vemurafenib binding. Our approach will give new insights into the evaluation of drug
efficacy. Further studies will be needed to investigate the heterogeneity of tumor tissue toward drug efficacy.
PL11
Targeted and Highly Multiplexed Imaging with Subcellular Resolution by CyTOF Mass Cytometry
1
2
1
1
2
1,2
2
1
1
B. Bodenmiller , J. Buhmann , R. Catena , C. Giesen , D. Grolimund , D. Günther , B. Hattendorf , H. Jackson , D. Schapiro , P.
2
3
2
3
Schueffler , Z. Varga , H. Wang , P. Wild
1
University of Zurich, IMLS, Zurich, Switzerland
2
ETH Zurich, Zurich, Switzerland
3
University Hospital Zurich, Zurich, Switzerland
Introduction: Tissues are complicated assemblies of multiple interacting cell types that communicate with each other to achieve
physiological states. In cancer, malignant cells and normal cells of the tumor microenvironment (TME) facilitate tumor
progression and drug resistance. For an understanding of the underlying processes it is important to comprehensively
investigate the components and their relationship within the TME. This necessitates imaging approaches that can simultaneously
measure dozens of biomarkers in a targeted and reproducible manner to define cell types, their functional and signaling states,
and spatial relationships.
Methods: For highly multiplexed tissue imaging at subcellular resolution, we have coupled immunohistochemical (IHC) methods
with high-resolution laser ablation and mass cytometry (1). In mass cytometry, rare earth metals are used as reporters on
antibodies. Analysis of metal abundances using the mass cytometer - an inductively coupled plasma mass spectrometer - allows
determination of biomarker expression. In the approach presented here, tissue sections were prepared for antibody labeling
using IHC protocols. Rare-earth-metal isotope tagged antibodies were selected to target proteins and protein modifications
relevant to breast cancer. After antibody staining, the sample was positioned in a laser ablation chamber developed by Wang et
al. (2) to minimize aerosol dispersion for high-resolution, high-throughput and highly sensitive analyses. The tissue was then
ablated spot by spot, and the ablated material was transported to the CyTOF mass cytometer using a gas stream. After data
Proteomic Forum 2015
30
preprocessing, the 32 transient, isotope signals were plotted using the coordinates of each single laser shot, and a highdimensional image of the sample was generated. Single-cell features were computationally segmented and the single cell
marker expression data were extracted for downstream bioinformatics analyses.
Results: Imaging mass cytometry provides targeted, high-dimensional analysis of cell type and state at subcellular resolution to
study tissues and adherent cells. The novel imaging approach enabled the simultaneous visualization of 32 proteins and protein
modifications, with the potential to map up to 120 markers on a single tissue section with the availability of additional isotopes.
Application of imaging mass cytometry to breast cancer samples allowed delineation of cell subpopulations and cell-cell
interactions, highlighting tumor heterogeneity and enabling new routes to patient classification. As such it has the potential to
yield novel insights of the TME by exploiting existing large collections of FFPE tumor samples and associated clinical information.
Conclusions: Imaging mass cytometry complements existing imaging approaches and will support the transition of medicine
toward individualized molecularly-targeted diagnosis and therapies.
(1) Giesen C. et al. Nat Methods. 2014 Apr;11(4):417-22.
(2 ) Wang HA et al. Anal Chem. 2013 Nov 5;85(21):10107-16.
Symposium • Microbial Proteomics
16:30–17:30
Chair: Uwe Völker (Greifswald/DE)
O12
Metaprotemic anylsyis of taxonomic and functional signatures in several grana padano experimental caseification models
1
1
2,3
2,3
2,3
1
1,4
A. Soggiu , C. Piras , S. Levi Mortera , A. Cavola , A. Urbani , L. Bonizzi , P. Roncada
1
University of Milan, Department of Veterinary Sciences and Public Health (DIVET), Milan, Italy
2
Università Tor Vergata, Dipartimento di Medicina Sperimentale e Chirurgia, , Roma, Italy
3
IRCCS Fondazione Santa Lucia, Roma, Italy
4
Istituto Sperimentale Italiano "Lazzaro Spallanzani", Milan, Italy
Introduction: During cheese production different types of bacteria activate a complex network of metabolic and fermentative
processes . In particular, during Grana Padano production the butyric fermentation during the ripening could cause the so-called
“late-blowing” phenomenon. Butyric fermentation is made by clostridia, in particular from vegetative forms of C. tyrobutyricum
and other species deriving from spore naturally present in raw milk used for the production. To avoid this phenomenon, egg
lysozyme is added to the cheese to inhibit clostridial growth. The use of this addictive could be a serious problem in people
allergic to egg proteins. Metaproteomics applied to the food expecially, is essential to investigate specific microbial functions in
complex microbial communities, to assess their effects in term of safety, in product transformation and to improve also the
quality and the taste.
Objectives: Apply metaproteomic techniques to several experimental caseification models of Grana Padano to obtain global
microbial taxonomic and metabolic signatures associated with a low or high number of clostridial spore in relation also to the
use of lysozyme.
Materials & Methods: Investigations were made on experimental caseification models of Grana Padano cheese. Eight samples
with a known number of clostridial spore (low or high) and lysozyme (with or without) were processed. Bacterial proteins were
extracted from samples and subjected to metaproteomics using a bidimensional chromatography approach (SCX-C18) coupled
to an Amazon Speed ETD Ion-trap mass spectrometer (Bruker Daltonics). Raw data were processed with Compass Data Analysis
1.3 (Bruker Daltonics) while protein identifications and 2D fractions data combining was performed with Compass ProteinScape
2.1 (Bruker Daltonics). Peptide list were parsed using a custom pipeline coupled to MEGAN[1] for taxonomic and functional
analysis and STAMP[2] for statistical analysis.
Results: A deep taxonomic and functional view associated to clostridial spore number and lysozyme treatment was obtained
using a comprensive metaproteomic approach.
Conclusion: The resulting microbial composition was useful to better understand taxonomic composition and metabolic features
that are able to characterize each sample in each type of treatment. Starting from these results we could use selected microbial
taxa and relative functional properties to counteract clostridial community growth within the cheese microbiota.
Proteomic Forum 2015
31
References: 1. Huson, D.H. and N. Weber, Microbial community analysis using MEGAN. Methods Enzymol, 2013. 531: p. 465-85.
2. Parks, D.H., et al., STAMP: statistical analysis of taxonomic and functional profiles. Bioinformatics, 2014. 30(21): p. 3123-4.
Work supported by FILIGRANA project, financed from MiPAAF D.M 25741/7303/11 - 01/12/2011
PL12
Misincorporation of nonproteinogenic amino acids in bacterial proteome analyzed by global, unbiased quantitative MS of
protein modifications
1
B. Macek
1
Proteome Center Tübingen, Tübingen, Germany
To study incorporation of nonproteinogenic amino acids in bacterial proteome we performed a global, unbiased protein
modification analysis of the E. coli K12 strain under various growth and stress conditions. We employed classical shotgun
proteomics approach based on the Orbitrap MS and processed the data using “dependent peptide” analysis integrated in the
MaxQuant platform. As a proof of principle we analyzed a strain with defective editing mechanism of the leucyl tRNA
synthetase, which in addition to leucine incorporates a nonproteinogenic amino acid norvaline. We measured widespread loss of
methyl groups at leucine positions, indicative of norvaline incorporation. We performed a Super-SILAC experiment and
established that under microaerobic conditions norvaline misincorporation reaches a maximum approximately 10 hours after
entry into stationary phase, up to 30% of all measured leucines are substituted with norvaline. We are currently conducting a
proteogenomics analysis to address misincorporation dependence on codon usage and we are extending this approach to
analysis of other amino acids. Our study presents the first global quantitative analysis of incorporation of nonproteinogenic
amino acids in a bacterial proteome.
PL13
Metaproteomics – novel insights into old questions in medical microbiology and microbial ecology
K. Riedel
Greifswald, Germany
TBA.
Proteomic Forum 2015
32
Tuesday, 24 March 2015
Plenary Session • Interaction Proteomics
09:00–10:30
Chairs: Julia Burkhardt (Dortmund/DE), Friedrich Lottspeich (Stockdorf/DE)
PL14
Chemical Cross-Linking and Mass Spectrometry for Mapping Protein-Protein Interactions
1
A. Sinz
1
Institute of Pharmazie, Department of Pharm. Chem. & Bioanalytics, Halle/Saale, Germany
During the last 15 years, chemical cross-linking combined with mass spectrometry (MS) and computational modeling has
advanced from investigating 3D-structures of isolated proteins to deciphering protein interaction networks [1-3]. Chemical crosslinking relies on the introduction of a covalent bond between functional groups of amino acids within one protein, to gain insight
into the conformation of a protein, or between interaction partners to elucidate interfaces in protein complexes. Based on the
distance restraints derived from the chemical cross-linking data, three-dimensional structural models of proteins and protein
complexes can be constructed. Most commonly, homobifunctional amine-reactive cross-linkers, such as N-hydroxysuccinimide
esters, are used for studying protein-protein interactions. One of our goals is to extend the arsenal of existing cross-linkers to
obtain complementary structural information. To facilitate the identification of cross-linked products, we have designed novel
MS/MS cleavable cross-linkers creating characteristic marker ions [4]. A direct way to probe in vivo protein-protein interactions
is by site-specific incorporation of genetically encoded photo-reactive amino acids or by non-directed incorporation of photoreactive amino acids. In my talk, I will give an overview of different cross-linking strategies and illustrate them for two examples:
(i) Mapping the interaction sites between the basement membrane proteins laminin and nidogen [5] and (ii) elucidating the
topology of the tetrameric tumor suppressor protein p53 [6].
Literature References: [1] Sinz, A., Expert Rev. Proteomics 2014, 11, 733. [2] Herzog et al., Science 2012, 337, 1348. [3]
Rappsilber, J., J. Struct. Biol. 2011, 173, 663. [4] Müller, M.Q. et al., Anal. Chem. 2010, 82, 6958. [5] Lössl et al., PLoSOne 2014,
9:e112886. [6] Artl et al., Proteomics 2015
PL15
Lipid-protein interactions
1,2
A.- C. Gavin
1
EMBL, Heidelberg, Germany
2
European Molecular Biology Laboratory, Heidelberg, Germany
Eukaryotic cells use membrane-bounded organelles with unique lipid and protein compositions to regulate and spatially
organize cellular functions and signalling. As part of this tight control, many proteins are regulated by lipids. In humans, the
importance of these regulatory circuits is evident from the variety of disorders arising from altered protein-lipid interactions,
which constitute attractive targets for pharmaceutical drug development. However, the full repertoire of interactions remains
poorly explored and exploited because their detection is still difficult to achieve on a large, systematic scale. I will describe a
series of chemical biology approaches to characterize in vivo assembled, stable protein-lipid complexes(1) and to study lipid
interactions with peripheral membrane proteins(2). Data from yeast and human cell lines reveal surprising insights, such as the
discovery of a new family of oxysterol-binding protein, conserved in humans (where it has been linked to several diseases) with
unexpected specificities for an important signaling lipid, phosphatidylserine. The assays are scalable to the proteome and/or
lipidome levels and are easily adapted to the study of small-molecules that disrupt protein-lipid interactions.
1. K. Maeda et al., Nature 501, 257 (Sep 12, 2013).
2. A. E. Saliba et al., Nat. Methods, doi: 10.1038/nmeth.2734 (2013).
Proteomic Forum 2015
33
PL16
The proteome of the synapse
1
H. Urlaub
1
Max-Planck-Insitute for Biophysical Chemistry, Bioanalytical Mass Spectrometry, Goettingen, Germany
In the synapse (the terminal nerve ending), synaptic vesicles are the storage organelles for neurotransmitters. These are
released upon stimulation of the nerve terminus, which leads to depolarization of the plasma membrane and the subsequent
fusion of synaptic vesicles with the membrane. In the mammalian CNS, an active pool of synaptic vesicles is attached to
specialized structures, termed active zones, in the presynaptic plasma membrane. In this presentation I will describe the
biochemical, microscopic and quantitative proteomic approaches that have been undertaken in recent years - in collaboration
with the laboratories of Reinhard Jahn and Silvio Rizzoli in Göttingen, and very recently together with Carla Schmidt in the
laboratory of Carol Robinson in Oxford - to elucidate the functions of proteins in synaptic vesicles and synpatosomes. I shall
present results on relative and absolute protein quantities, post-translational modifications, protein-protein contacts and
protein localization, and will show ways in which modern high-resolution microscopy techniques can be combined with
quantitative mass-spectrometric measurements.
Plenary Session • Subcellular Proteomics
11:00–12:30
Chairs: Steven Verhelst (Munich/DE), Anne-Claude Gavin (Heidelberg/DE)
PL17
15
Protein turnover by progressive N labelling for analysis of plant mitochondrial biogenesis and assembly
1
1
1
1
H. Millar , C. Nelson , L. Li , S. Huang
1
The University of Western Australia, ARC Centre of Excellence in Plant Energy Biology, PERTH, Australia
Shotgun approaches or targeted SRM analyses have dominated proteome studies as tools to find changes in the proteome.
These approaches, however, focus just on the proteins that are changing in abundance in order to find biological insights. They
also require statistically significant changes in the total accumulated protein pool size in order to identify that ‘anything has
occurred’. Analysing protein synthesis and degradation rates with progressive stable isotope labelling provides a new window on
the control of protein abundance. With this approach we can determine the ‘relative age’ of the proteins that we see and define
15
the energetic effort employed by the cell to build or maintain particular activities. We are using progressive N labelling of plant
cells from nitrate and ammonium salts and modelling incorporation fits, to calculate the rate at which proteins which are often
static in abundance in the proteome are turning over. We have developed pipelines to undertake these studies in plant cells,
plant leaves and in whole plants through the use of hydroponics. Through combining such labelling with separation of protein
complex and subcomplexes by native electrophoresis, we can observe the in vivo turnover rate of assembly intermediates of
protein complexes in plant mitochondria. We are also using this approach to study the role of mitochondrial proteases in
mitochondrial maintenance. These approaches provide new avenues for peptide mass spectrometry to provide answers to a
wide range of questions in plant mitochondrial biology.
PL18
Hybrid Methods for Defining the Structure and Function of Cellular Machines
1
B. Chait
1
The Rockefeller University, New York, United States
Mass spectrometry (MS) together with a variety of associated proteomics tools have become increasingly important
components of hybrid structure determination strategies. Here I will discuss how such tools can enable key steps in the process
of elucidating the structure of multiprotein complexes.
At the very earliest stage of analysis, it is essential to gain insight into the precise ensemble of macromolecular players that
make up the complex of interest. MS has become the tool of choice for this task because of its sensitivity and its facility for
characterizing the primary structures of endogenous proteins along with modifications that may be present. Once the relevant
ensemble is established, it is essential to determine their stoichiometry. Here, native MS can provide accurate direct measures
of stoichiometry, and where this tool is not applicable, a variety of stable heavy isotope methodologies are proving useful. Next,
proteomic approaches are being applied to increasingly good effect to gain insight into the architectures of protein complexes,
especially those that too large or too poorly behaved for traditional X-ray crystallographic diffraction, NMR spectroscopic or
electron cryomicroscopy analyses alone. Here the approaches range from simple affinity isolation-based association studies
Proteomic Forum 2015
34
(sometimes with selected domain ablation), to chemical crosslinking studies (both in vitro and in vivo), to MS-based dissociation
studies (monitoring dissociations both in solution and in the mass spectrometer; all these appropriately integrated with high
resolution structure and shape determination, electron microscopy, and modeling. I will also discuss limits of the present
proteomics technologies and prospects for the future.
I acknowledge NIH Grants GM103314 and GM109824
PL19
Subcellular protein localization by mass spectrometry
K. Lilley
Cambridge, United Kingdom
TBA.
Symposium • Host-Pathogen Interaction & Antibiotical Resistance
14:00–15:00
Chair: Paola Roncada (Rome/IT)
O13
AP-MS and BioID analysis of the HBx Viral-Host interactome
1
1
2
E. Milani , B. Wollscheid , J. Gao
1
ETHZ, 8093 Zurich, Switzerland
2
Hoffmann-La Roche, Infectious Diseases DTA, Roche pRED, Basel, Switzerland
Question: Worldwide, 2 billion individuals have been infected with hepatitis B virus (HBV). Despite advances in HBV prevention
chronic hepatitis B remains an unsolved problem and a prime target for novel antiviral therapies. The HBV X protein (HBx) is a
non-structural, regulatory viral protein with pleiotropic activities which is thought to play a central role in HBV persistence and
HBV-associated hepatocellular carcinoma. Here we investigated the HBx interactome with two complementary Interactomic
strategies in order to elucidate network structure and potential targets for pharmaceutical intervention.
Methods: Systematic reciprocal affinity purification mass spectrometry (AP-MS) and a proximity-dependent biotinylation of
living cells coupled to affinity purification and mass spectrometry (BioID) strategy was used to investigate the HBx-host
interaction network.
Results and Discussion: Combination of AP-MS and BioID strategies revealed a HBx-interacting network consisting of defined
subnetworks. The identified subnetworks confirm and extend current knowledge on the pleiotropic activities of HBx pointing
towards an involvement of HBx in multiple distinct cellular processes. Identified interactors show HBx involvement and
potentially modulation of proteasome-dependent degradation, regulation of cellular proliferation, mRNA processing and nuclear
translocation. Using two complementary approaches for network reconstruction not only enabled the informed filtering of
commonly identified interaction partners, but also the retrieval of new, presumably lower affinity, interaction partners mainly
due to the BioID proximity-tagging strategy.
Conclusions: Together, the HBx-Host interactome revealed new leads for pharmaceutical intervention and show the benefit of
using a combined AP-MS and BioID strategy for probing biological network structure.
Proteomic Forum 2015
35
O14
Physiology of Staphylococcus aureus isolated from human macrophages and the role of CodY in host pathogen interaction
1
1
2
3
1
1
4
2
5,6,1
S. Reiß , N. Köhler , J. Mostertz , P. Hildebrandt , S. Fuchs , M. Hecker , C. Wolz , F. Hochgräfe , S. Engelmann
1
Universität Greifswald, Institut für Mikrobiologie, Greifswald, Germany
2
Universität Greifswald, ZIK Functional Genomics, Junior Research Group Pathoproteomics , Greifswald, Germany
3
UniverstätsmedizinGreifswald, Funktionelle Genomforschung, Greifswald, Germany
4
Universität Tübingen, Institut für Medizinische Mikrobiologie und Hygiene, Tübingen, Germany
5
Institute for Microbiology, Microbial Proteomics, Braunschweig, Germany
6
Helmholtzzentrum für Infektionsforschung, Mikrobielle Proteomik, Braunschweig, Germany
Staphylococcus aureus is a human pathogen of strong clinical significance due to increasing infections with multi-resistant
isolates. A better understanding of its infection biology is urgently required to combat this pathogen also in the future. The
fitness of this pathogen, which is based on its ability to successfully adapt to host conditions, is crucial for full virulence. To get
valuable insights into staphylococcal physiology after phagocytosis by professional phagocytes, THP-1 human macrophages were
infected with green fluorescence protein (GFP) expressing S. aureus Newman cells. Phagocytized GFP positive bacteria were
separated via FACS sorting and the intracellular and surface-associated proteome was analyzed by mass spectrometry. For
quantitative in vivo proteomics we applied dual analysis strategy by combining Super-SILAC and Label-free approaches.
Altogether we identified almost 700 S. aureus proteins. The amount of 104 of the S. aureus proteins was significantly and at least
1.5 fold changed. Among them 64 proteins which accumulated after phagocytosis. We found evidence for the induction of
different staphylococcal virulence factors such as IsaA, IsdA, Sbi, Scn. Moreover we showed that the global regulator CodY plays
as substantial role in gene expression in phagocytized S. aureus cells. Interestingly, codY mutants kill macrophages more
efficiently than the corresponding wild type strains possibly due to an increased expression of secreted virulence factors such as
lipases, proteases, nucleases and toxins.
PL21
Host pathogen interactions in the pig, and the relevance to antimicrobial resistance.
1
E. Bendixen
1
Aarhus University, Molecular Biology and Genetics, Aarhus, Denmark
Human health is closely linked to the health of animals, and to the state of the environment we share with all living beings. This
connectivity is currently most alarmingly seen as the rise in antimicrobial resistance to antibiotics, which is directly related to the
massive use of antibiotics in livestock production.
Despite a European ban on antibiotics in animal feed since 2006, the need for antibiotics medication of farm animals is steadily
increasing. With currently 80 % of all antibiotics being given to farm animals, it is clear that improving farm animal health is
mandatory for counteracting the global challenges of antimicrobial resistance.
This talk will present our research of host pathogen interaction in pig. We study how variants of the porcine FUT1 gene influence
health, growth and the gut bacteria. Moreover, we use use mass spectrometry to study the gut proteome, and to characterize
the glycosylation of potential E.coli receptor proteins in the pig gut.
Our research is directly aimed to improve gut health in pig production, but the knowledge we create is valuable also for
developing pig models aimed at human biomedical research. In particular, pig models for human metagenomics, nutrigenomics
and host-pathogen interaction research is gaining popularity because of the close similarities between porcine and human
metabolism and gut physiology. This talk will present some of the relevant pig model models we have studied in the recent
years.
Proteomic Forum 2015
36
Symposium • Plant & Agricultural Proteomics II
14:00–15:00
Chair: Harvey Millar (Crawley/AU)
PL20
Evaluation of plant genetic resources for stress defence mechanisms using proteomic approaches
1
2
1
H.- P. Mock , K. Witzel , A. Matros
1
Leibniz-Institut für Pflanzengenetik und Kulturpflanzenforschung, Gatersleben, Germany
2
Leibniz-Institut für Gemüse- und Zierpflanzenbau Großbeeren, Großbeeren, Germany
Question: Crop plants are the basis for food and feed. Plant breeding aims at improving traits of crop plants such as yield and
the resistance against biotic and abiotic stresses. Changing environments will request the adaptation of current elite crop lines.
The genetic diversity contained in plant collections such as in the genebank of the IPK Gatersleben. This genebank harbours in
total 150 000 accessions, with a strong focus on cereals. Both barley and wheat are each represented with about 25 000
accessions, including landraces, wild relatives as well as modern cultivars. Evaluation of the diversity within the collection is a
pre-requisite to support future breeding strategies for improved cereal crops.
Methods: We are using proteomic approaches to evaluate barley accessions from the genebank collection in the context of
stress tolerance and seed development. Barley is used as a crop, but also as a model for wheat having higher agronomic
significance. Identification of proteins benefits from the recent sequencing of the barley genome. Evaluation of the plant
samples on the proteome level are complemented by metabolomic approaches and by targeted analysis of proteins, metabolites
and transcript. Data avaluation is performed with the help of bio-informatic ressources including tools such as VANTED and
MAPMAN to allow mapping of candidate genes onto biochemical pathways.
Results: We have explored barley mapping populations with respect to contrasting salt stress tolerance of the parental lines
Steptoe and Morex. The analysis was performed on different plant tissues and developmental stages. In addition to whole
proteome analysis, we performed analysis of fractions enriched by two-phase partitioning for plasma membrane vesicles. At
present, the study has revealed a large number of candidates associated with the higher salinity tolerance of the parent line
Morex. For the functional validation of candidates, we used transgenic approaches and extended the analysis for selected
offspring lines from the Steptoe x Morex mapping population. In addition we made us of a salt sensitive yeast strain by screening
and testing barley genes providing higher tolerance under high salt concentrations to yeast transformants. Several of the
candidates were confirmed both in the yeast and in the plant system and are studied with priority to reveal the underlying
mechanisms providing tolerance.
Conclusions: Our studies in cereal crops provide evidence that proteomic approaches are a valuable approach to study the
genetic diversity of genebank collections with the potential to reveal novel mechanisms associated with stress tolerance.
Information on novel candidates and insights into networks associated with stress tolerance will provide novel opportunities for
crop breeding.
O15
Quantitative proteomic analysis of lysine acetylation in Arabidopsis leaves.
1
I. Finkemeier
1
Max Planck Institute for Plant Breeding Research, Plant Proteomics, Cologne, Germany
Acetylation of the ε-amino group of lysine is a reversible post-translational modification recently discovered to be widespread,
occurring on proteins outside the nucleus, in most sub-cellular locations in various organisms. Until recently, almost nothing was
known about this modification in plants beyond the well-studied acetylation of histone proteins in the nucleus. We developed a
protocol based on high resolution mass spectrometry for the identification and relative quantification of lysine-acetylated
peptides in the plant species Arabidopsis thaliana. Our results indicate that lysine acetylation could be important in the
regulation of key metabolic enzymes and protein complexes, including a large proportion of photosynthetic proteins. Central
enzymes of the Calvin-cycle, as for example RuBisCO, are specifically and dynamically acetylated at various sites and
deacetylation in vitro has a strong impact on enzyme activity. One of the main questions of our research is to identify the
acetylation sites which are improtant for the regulation of enzyme functions and to quantify changes in lysine aceytlation under
different environmental conditions using LC-MS/MS-based proteomics.
Proteomic Forum 2015
37
O16
Mass spectrometry identification of free flow separated protein complexes from the thylakoid membrane of Arabidopsis
thaliana
1
2
3
4
1
5
6
4
L. Eichacker , S. Tenzer , U. Sukop-Köppel , G. Fucile , K. Kmiec , A. Christophoru , M. Goldschmidt-Clermont , J.- D. Rochaix , G.
3
Weber
1
University of Stavanger, CORE, Stavanger, Norway
2
University Mainz, Immunology, Mainz, Germany
3
FFE Service GmbH, Feldkirchen, Germany
4
University of Geneva, Dept. Mol. Biology, Geneva, Switzerland
5
University of Cambridge, Centre for Proteomics, Cambridge, United Kingdom
6
University of Geneva, Dept.Mol. Biology, Geneva, Switzerland
Photosynthetic organisms adapt the photosynthetic machinery to changes in the light environment. Two complementary kinasephosphatase pairs corresponding to STN7/PPH1 and STN8/PBCP have been shown to determine the phosphorylation status of
the light-harvesting proteins of photosystem II and its core proteins in Arabidopsis thaliana (Rochaix, 2013). The antagonistic pair
formed by the STN7 kinase and the PPH1/TAP38 phosphatase was found to be required for state transition of photosynthesis via
phosphorylation/dephosphorylation of the LHCII subunits LHCB 1 and LHCB 2 (LHCII/I). In addition to a relocalization of LHCII and
transition in the phosphorylation state, the two distinct assembly states of photosystem complexes in Chlamydomonas
reinhardtii have been correlated with a transition of the photosynthetic electron flow from cyclic to linear (state 2 to 1)
(Minagawa, 2013). The assembly states in higher plants are less clear.
Here, free-flow interval zone electrophoresis is shown to be an excellent technology for high resolution separation of membrane
protein complexes. The assembly states of the photosynthetic complexes from Arabidopsis thaliana are utilized to demonstrate
the charge based separation of protein complexes. Within 4-6 min, distinct assembly states of the membrane complexes from
mutant plants lacking either STN7 or PPH1 and hence locked in state 1 or 2 were analyzed. Subunit compositions were identified
by mass spectrometry and assembly states determined by hierarchial cluster analysis.
References: Minagawa, Jun (2013) Dynamic reorganization of photosynthetic supercomplexes during environmental acclimation
of photosynthesis». Frontiers in Plant Science 4, 513.
Rochaix, Jean-David. (2013) Redox Regulation of Thylakoid Protein Kinases and Photosynthetic Gene Expression, Antioxidants &
Redox Signaling 18, 2184-2201.
Figure 1, Hierarchial cluster analysis of thylakoid membrane proteins
Iz-FFE fractions tryptically digested and proteins identified by mass spectrometry. For hierarchial cluster determination of
protein complex assembly, Mascot search results with a minimum cut-off value of 20 were used. The number of peptidespectrum matches for each identified protein was counted in each of the analyzed fractions, and compiled into a matrix. For
each identified protein, the number of spectral counts across the analyzed fractions was normalized to a total of one. A
Euclidean distance matrix was calculated for all proteins identified with a minimum of 20 counts, and visualized as a dendrogram
using R, and cut at an euclidean distance of 0.6. Complexes were labelled according to the hierachial cluster assignment of
proteins as ATPase, Cytb6f, PSI, LHCI, PSII, LHCII, LHCII/I and assembly status of complexes indicated by a dash between complex
units.
Proteomic Forum 2015
38
Figure 1
Symposium • Quantitative Strategies
14:00–15:00
Chair: Roland Kellner (Darmstadt/DE)
O17
Quantitative proteomics meets highly sensitive phosphoproteomics - a simple workflow that combines both for low sample
amount experiments
1
1
2
2
1
1,3
S. Loroch , O. Pagel , T. Schommartz , W. Brune , R. Zahedi , A. Sickmann
1
Leibniz-Institut für Analytische Wissenschaften - ISAS e.V., Systems Analysis, Dortmund, Germany
2
Heinrich-Pette-Institut – Leibniz-Institut für Experimentelle Virologie, Hamburg, Germany
3
University of Aberdeen, School of Natural & Computing Sciences, Aberdeen, Germany
Introduction: In clinical and biomedical research limited sample amount is a major bottle neck especially when working with
primary cell culture or patient tissue. If the sample material is restricted, phosphoproteome analysis is often not promising since
phosphoproteomics workflows are insufficient sensitive. Currently sensitive phosphoproteomics is based on affinity
4+
chromatography (such as TiO2-MOAC or Ti -IMAC) which is difficult to establish and hard to automate. Thus, sensitive and
robust workflows are required to enable efficient analysis of the proteome and the phosphoproteome with the least sample
material possible.
Objectives: Our aim was to develop a highly sensitive strategy solely based on HPLC-based fractionation and robust solid-phase
extraction to enable sensitive protein phosphorylation analysis and a global proteome analysis in parallel.
1
Material and Methods: We combined electrostatic repulsion-hydrophilic interaction chromatography (ERLIC) with strong cation
2
exchange solid-phase extraction to efficiently enrich and fractionate phosphorylated peptides from tryptic digests . All nonphosphorylated peptides were recovered and fractionated in a second dimension. Fractions were analysed by nanoHPLC-ESIMS/MS. Database search was performed using MASCOT implemented in Proteome Discoverer 1.4 and phosphorylation sites
were validated using phosphoRS 3.1. We evaluated our workflow in a quantitative experiment using a 2-plex SILAC sample of
cytomegalovirus infected mouse fibroblasts (5h infection) against a non-infected control.
Results: With only 150 µg protein per condition and 33 h of total LC-MS time, we were able to reproducibly quantify >5,000
phosphopeptides covering ~5,500 phosphorylation sites of ~2,000 phosphoproteins. The recovery of non-phosphorylated
peptides enabled a parallel quantification of ~20,000 non-phosphorylated peptides allowing the reproducible quantification
2
(R =0.92) of ~4,000 proteins in 21 h additional LC-MS time. Further, two third of all phosphoproteins could be quantified by at
least two non-phosphorylate peptides.
Conclusion: Our highly sensitive ERLIC-based strategy enabled the reproducible quantification of ~33 phosphopeptides per µg
protein. Thus, it is among the most sensitive strategies for quantitative phosphoproteomics. A quantitative proteome analysis
Proteomic Forum 2015
39
can be done in parallel without the need to employ additional sample material. The workflow is easy-to establish and requires
just one HPLC and 1.5 SPE tips per fraction (besides the LC-MS instrumentation). The simplicity and sensitivity render its future
promising for low-sample amount experiments in biological, biomedical and clinical research.
[1]. A. J. Alpert, Analytical chemistry 2008, 80. 62-76, DOI: 10.1021/ac070997p.
[2] S. Loroch, R. P. Zahedi, A. Sickmann, Analytical chemistry 2014. DOI: 10.1021/ac502708m.
O18
Extensive Proteome Remodeling and Protein Aggregation in C. elegans Revealed by SILAC-based Quantitative Proteomics
1
2
2
2
3
3,4
4
3
3
D. Walther , P. Kasturi , M. Zheng , S. Pinkert , G. Vecchi , P. Ciryam , R. Morimoto , C. Dobson , M. Vendruscolo , M. Mann
2
U. Hartl
1
Max-Planck Institute of Biochemistry, Proteomics and Signal Transduction, Martinsried, Germany
2
Max-Planck Institute of Biochemistry, Cellular Biochemistry, Martinsried, Germany
3
University of Cambridge, Department of Chemistry, Cambridge, United Kingdom, United Kingdom
4
Northwestern University, Department of Molecular Biosciences, Evanston, IL, United States
1
Proteostasis (or protein homeostasis) is a tightly controlled cellular process which regulates synthesis, maintenance and
degradation of proteins. While organisms age, proteostasis becomes increasingly compromised, resulting in a higher
susceptibility of humans to protein folding diseases such as Alzheimer’s, Parkinson’s and Huntington’s diseases. However, the
underlying mechanisms which lead to proteostasis collapse are at present poorly understood. Here we investigated proteomic
changes in the nematode C. elegans during aging. We not only characterized the total proteome of wild type (wt) and longevity
mutant strains but further developed a method to biochemically isolate and analyze insoluble protein aggregates.
While SDS lysis was used for total proteome analysis, the aggregated sub-proteome was recovered from native-like lysates by
ultracentrifugation followed by mild detergent extraction. To achieve highly accurate quantification, we made use of metabolic
13
15
C6- N2 lysine (SILAC) labeling by feeding nematodes with labeled cells of a lysine auxotroph E. coli strain. Lysates from SILAClabeled animals were spiked into biological samples as an internal standard. After digestion and fractionation, peptides were
separated by nHPLC and analyzed with Orbitrap Velos or Orbitrap Elite mass spectrometers in HCD mode, followed by data
processing using MaxQuant and Andromeda.
We quantified more than 5,000 proteins at multiple time points during life span and found that C. elegans undergoes extensive
remodeling, with approximately one third of the proteome changing at least twofold in wt animals. Among the components of
the proteostasis system, we observed a major impact of aging on ribosomal and proteasomal subunits. While proteasomal
proteins increased, ribosomal proteins decreased in abundance. These changes were accompanied by elevated proteasomal
activity and reduced protein synthesis rates, respectively. In accordance with previous studies, our analysis of the aggregated
proteome yielded that the vast majority of proteins accumulated significantly in the insoluble fraction between young and aged
nematodes. However, most of our more than 3,300 quantified aggregation-prone proteins have not previously been reported,
presumably owing to the ability of our isolation procedure to capture weakly aggregating species. Unexpectedly, the abundance
of aggregates in the long-lived daf-2 mutant strain was strongly increased over wt. Moreover, aggregates were associated with
members of the small heat shock protein family, suggesting a mechanism of aggregate sequestration that is beneficial for
survival.
In conclusion, we have quantified both the total and the aggregated proteomes in C. elegans in a time-resolved manner at high
depth and accuracy. These data provide valuable insights into the cellular deterioration caused by proteostasis decline during
aging.
Proteomic Forum 2015
40
PL22
HR-PRM for quantification of low abundance plasma proteins implicated in cardiovascular diseases
1
S. Singh
1
Brigham and Women's Hospital, Cardiovascular Sciences, Boston, United States
Introduction: Atherosclerosis causes coronary heart disease (CHD), a global health burden. Dyslipidemia such as low levels of
high-density lipoprotein (HDL) and arterial calcification contribute to the onset of acute thrombotic complications of
atherosclerosis. Modulation of the size distribution or apolipoprotein component of HDL particles may prevent CHD.
Characterization of instigators of smooth muscle cell mineralization such as sortilin may help to develop much needed therapies
for calcification. High resolution parallel reaction monitoring (HR-PRM) serves as a powerful tool to accomplish these goals.
Objectives: To refine kinetics of human HDL particles and their protein components, and to explore the mechanism by which
sortilin promotes calcification using stable isotope labeling strategies and high resolution parallel reaction monitoring (HR-PRM).
Methods: HDL apolipoprotein studies - Three participants with similar lipoprotein profiles were infused with a bolus of D3-Leu
tracer and blood was collected up to 70 hours. ApoA-I immunopurified HDL particles were separated by non-denaturing-PAGE,
and five molecular size window fractions were in-gel digested. HR-PRM quantified tracer enrichment (2HM3/M0) for seven
apolipoproteins across the five fractions. Sortilin studies - Standard DDA MS analyses of cellular and in vitro synthesized/labeled
sources of sortilin provided a spectral library including a C-terminal phosphorylated peptide. HR-PRM measured the unmodified
and phosphorylated forms of the C-terminal peptide in the atherosclerotic carotid arteries of apoE-deficient mice (Apoe-/-) and
human endarterectomy samples.
Results: Seven apolipoprotein D3-Leu tracer enrichment profiles were determined across five HDL size fractions. HR-PRM
enabled reproducible quantification the D3-derived 2H M3 ion as low as 0.02 %. The enrichment curves for each protein varied
from the other by slope and time of peak enrichment. In contrast, the curves across HDL sizes for a single protein showed
smaller, but likely meaningful, differences in either slope or time of peak enrichment. The protein with the highest enrichment
was apoE (8 %) and the lowest was apoD (0.2 %). Interestingly, these unique enrichment curves observed for each protein was
conserved in the three participant samples. Sortilin - Relative quantification of fragment ion peaks for unmodified vs.
phosphorylated SGYHDD(pS)DEDLLE revealed significantly higher phospho-peptide in calcifying arteries compared to non-/calcifying tissue dissected from Apoe mice. Calcified human carotid endarterectomy samples contained up to 36% phosphopeptide, whereas non-calcifying carotid artery specimens had 8% of phospho-peptide.
Conclusion: Our in vivo evidence indicate that HR-PRM provides us with a highly sensitive and specific means to monitor
proteins implicated in CHD; thereby providing a foundation for the development of new therapies.
Young Investigator Session
16:30–18:00
Chair: Roland Kellner (Darmstadt/DE)
YIA1
GRP78 phosphorylation dynamics during sperm maturation
1
1
1
1
1
V. Lobo , P. Rao , R. Gajbhiye , V. Kulkarni , P. Parte
1
National Institute for Research in reproductive health, Mumbai, India
Question: Glucose regulated protein 78 (GRP78), an ‘ER resident chaperone’ plays an important role in protein folding and
degradation of misfolded or unfolded proteins. As spermatozoa lack a recognizable ER, GRP78 presence in mature spermatozoa
and its relevance to sperm function is an enigma. Previous work by our lab has revealed reduced GRP78 phosphorylation in
asthenozoospermatozoa. The present study proposes to investigate in-depth this differential phosphorylation of GRP78.
Methods: Differential phosphorylation of GRP78 in asthenozoospermatozoa was determined by Differential in Gel
Electrophoresis (DIGE). The presence of GRP78 in rat and human spermatozoa was studied by Western blot (WB) analysis and its
localization by Indirect Immunofluorescence (IIF). Enzyme linked Immunosorbent assay (ELISA) was optimized to determine
surface localization of GRP78 in intact rat spermatozoa. Immunoprecipitation (IP) experiments were done to determine the type
of amino acid residues phosphorylated in GRP78. The extent of GRP78 phosphorylation in rat mature- and immature sperm and
its differential phosphorylation in asthenozoospermatozoa vis a vis normal spermatozoa was discerned by Nanofluidic proteomic
immunoassay (NIA) and Phosphatase assays.
Proteomic Forum 2015
41
Results: Using Differential in Gel Electrophoresis (DIGE) phosphorylated GRP78 was observed to be > 2 fold less in
asthenozoosperm vis-ẚ-vis normozoosperm. WB showed presence of GRP78 in rat testis and spermatozoa as well as in human
spermatozoa. IIF indicated surface localization of GRP78 on the equatorial region in case of rat- and in equatorial, midpiece and
neck region in case of human- spermatozoa. Post permeabilization it was observed in the neck and ventrally at tip of sperm head
for rat and in midpiece of human spermatozoa. Surface localization of GRP78 was also validated by ELISA. GRP78 was observed
to be phosphorylated at serine, threonine and tyrosine residues by IP studies. NIA coupled with Phosphatase assays revealed
three phosphoforms GP4.96, GP4.94 and GP4.85 in mature rat spermatozoa versus two phosphoforms GP 4.96 and GP4.94 in testicular
spermatozoa. In human spermatozoa, 3 phosphoforms GP5.04, GP4.96 and GP4.94 were observed, of which, GP4.94 and GP5.04 were
significantly reduced in asthenozoospermatozoa.
Conclusions: Our study provides novel information on the spatial distribution of phosphorylated GRP78 during sperm
maturation. Based on our observations we have proposed a schematic demonstrating the putative dynamic changes in GRP78
phosphorylation in spermatozoa during its maturation. The observation of reduced phosphorylation of GRP78 in the
asthenozoosperm and identification of the phosphoforms reduced, suggests relevance of GRP78 phosphorylation to sperm
function.
YIA2
Exposure of Mycobacterium tuberculosis to thioridazine alters the cell envelope permeability; an alternative mechanism for
the synergistic effects with other antibiotics
1
2
1
2
3
2
J. de Keijzer , P. de Haas , A. de Ru , E. Heerkens , L. Amaral , D. van Soolingen , P. van Veelen
1
Leiden University Medical Centre, Leiden, Netherlands
2
National Institute for Public Health and the Environment (RIVM), Bilthoven, Netherlands
3
Universidade Nova de Lisboa, Lisboa, Portugal
1
The emergence of multidrug resistant tuberculosis mandates the search for alternative drugs. Several neuroleptics, such as the
phenothiazine thioridazine, exhibit in vitro activity against Mycobacterium tuberculosis. Thioridazine has already been
successfully used in combination with antibiotics for compassionate therapy of patients presenting with extensively drug
resistant infections of tuberculosis. Using an unbiased proteomic approach, we set out to unravel the molecular mechanism of
this potential new anti-tuberculosis component by examining the impact of continuous thioridazine exposure on the proteome
of M. tuberculosis. We discovered that under the influence of thioridazine several proteins involved in the maintenance of the
cell wall permeability barrier are differentially regulated. By assessing accumulation of fluorescent dyes in M. tuberculosis over
time, we demonstrated that long-term thioridazine exposure of M. tuberculosis affected the mycobacterial cell envelope and
increased the permeability towards both hydrophilic and hydrophobic fluorescent compounds, thereby the enhancing the
uptake of compounds. Furthermore, we demonstrated that treatment of M. tuberculosis with thioridazine altered the
composition of the plasma membrane. These observations may explain the previously reported synergistic effects between
thioridazine and other anti-TB drugs. The knowledge of thioridazines mode of action is a major step forward and could facilitate
further development of this class of drugs for therapy of multidrug resistant pulmonary infections of tuberculosis. In fact, due to
the increased permeability of the cell envelope to old drugs and new anti-tuberculosis antibiotics, not only would the old drug(s)
be restored to an effective therapeutic level, the dose levels of new but toxic anti-tubercular compounds may be significantly
reduced.
YIA3
Improving Proteotypic Peptide Prediction Using Information Retrieval Based Measures
1,2
1,2
A. Argentini , L. Martens
1
VIB , Department of Medical Proteins Research, Ghent, Belgium
2
Ghent University, Department of Biochemistry, Ghent, Belgium
Introduction: The prediction of proteotypic peptides is a relevant problem in targeted proteomics, where the right choice of the
selected peptide for the proteins of interest improves the quality of the result. Several Machine Learning based predictors for
proteotypic peptides exist, and these can be divided into two main groups: (I) the problem is solved using binary classifiers
(Fusaro et al. 2009)-(Eyers et al. 2011), or (II) a ranking function is learned from the data (learn-to-rank methods) (Qeli et al.
2014).
Objectives: The work described here focused on using an intermediate approach that combines the easy interpretability of the
binary classifier approach with the extra information obtained from ranking all peptides for a given protein. Moreover, our
approach avoids the necessity of specifying an a priori relevance for each peptide in typical learn-to-rank methods (Qeli et al.
2014). Our approach is shown to outperform the best binary classifiers currently available.
Proteomic Forum 2015
42
Material& Methods: The CPTAC dataset (Paulovich et al. 2010) of 43 MS experiments from 5 yeast samples with different
concentrations of spiked-in UPS proteins was used. MS1 apex peak intensities were extracted using an in-house tool. 569
sequence-based features (aa composition, aaIndex (Kawashima et al. 1999)) were calculated for each peptide. As training and
test sets we have used the yeast and the UPS spiked proteins, respectively. Cross validation was used during training, and the
nd
th
average precision for the 2 and the 5 position ([email protected] and [email protected]) of the ranked list of peptides on the test set. As predictor we
have trained both a fast linear SVM, and a Random Forest regression model that predicts peptide intensity.
Results: Our predictor based on a linear SVM outperforms all competitors evaluated both in terms of [email protected] and [email protected] (Figure 1).
The regression model shows better results with respect to the competitors but does not outperform the linear SVM. The
optimization of the evaluated classifiers using [email protected] and [email protected] in the cross-validation thus allows our approach to obtain a well
ranked list of the best ionized peptides.
Conclusions: In this work we showed that the use of apex peak intensities is really effective to improve the prediction of
proteotypic peptides. Moreover, we highlight that addressing the problem as a ranking problem instead of a classification
provides a more reliable, more useful prediction, in which candidate peptides for SRM are reported in a ranked order of
proteotypicity.
nd
th
Figure 1: The [email protected] and [email protected] (average precision at the 2 and the 5 position of the ranked list of peptides) result of our
predictors and the best binary classifiers available (ESP, CONSeQuence) on the test set. Our linear SVM outperforms all the
competitors in both the metrics evaluated
Figure 1
YIA4
Label-free quantification of GeLC-MALDI data with a novel software reveals a subtype-specific protein signature for
pancreatic ductal adenocarcinoma
1,2
1,2
3
1,2
1,2
W. Nadler , A. Kerner , B. Klaus , L. Kuhlmann , C. Rösli
1
German Cancer Research Center, Heidelberg, Germany
2
Heidelberg Institute for Stem Cell Technology, Heidelberg, Germany
3
European Molecular Biology Laboratory, Genome Biology Unit, Heidelberg, Germany
Introduction: In 2011, Collisson and co-workers defined three different subtypes (classical, exocrine-like and quasimesenchymal) of pancreatic ductal adenocarcinoma (PDAC) based on a gene signature study using microdissected human
tumors. Patients with distinct subtypes showed significant differences in clinical outcome and therapeutic response (Collisson et
al., Nat Med. 2011). A proteomic signature of the different subtypes could yield biomarkers for diagnosis and targeted therapies.
Objectives: In order to identify biomarkers for cancer diagnosis and patient stratification, 12 primary human PDAC cell lines from
9 different patients were proteomically characterized with a label-free GeLC-MALDI approach. The novel software suite MSQBAT
was developed to quantify the GeLC-MALDI data and results were evaluated using multiple reaction monitoring of biological
markers.
Methods: The proteome of 12 patient-matched PDAC cell lines (grown under serum-free conditions) and 2 control cell lines
were analyzed in a GeLC-based bottom-up approach (Figure 1). Peptides from each of the 12 gel fractions per cell line were
separated into 600 fractions with a 110 min gradient (11-40% ACN, 0.1% TFA). The fractions were spotted on a MALDI target
Proteomic Forum 2015
43
plate using a SunChrom robot and analysed with a MALDI TOF/TOF™ 5800 mass spectrometer (AB SCIEX). Peak lists (Data
Explorer) and identifications (ProteinPilot™, both AB SCIEX) of proteotypic peptides (PepSir, build in-house) were loaded into
MSQBAT and the gel sample sets were quantified with respect to a super-alignment of all cell lines.
Results: A total of 5627 proteins were identified based on proteotypic peptides and 4492 of these were quantified within all 14
cell lines. Unsupervised clustering robustly separated the exocrine subtype and the controls from the other cell lines.
Comparison of the proteomic data with data from an Illumina® Gene Expression Array showed highly consistent results for
clusters of regulated gene families. Within the exocrine subtype proteins involved in biotransformation, e.g. members of the
CYP450 and AKR families were significantly up-regulated (Figure 2: Heat maps of the log-fold changes of quantified proteins
involved in biotransformation). Several markers were validated with MRM, confocal microscopy and classical proteomics
techniques. Using the resulting MRMs, the performance of the MS QBAT software was evaluated. Despite different sample
preparation techniques (quantification along 12 gel slices for MSQBAT or shotgun for MRM quantification), both methods yielded
a highly consistent and stable quantification.
Conclusions: MSQBAT, a novel software suite, enables the label-free quantification of GeLC-based MALDI data, resulting in the
identification of a subtype-specific protein signature for PDAC and the accession of regulated protein families with potential
clinical relevance.
Figure 1
Figure 2
YIA5
Proteomic analysis of barley cell nuclei
Proteomic Forum 2015
1
1
1
2
2
44
1
I. Chamrad , J. Urinovska , R. Lenobel , H. Jerabkova , B. Petrovska , M. Sebela
1
Centre of the Region Haná for Biotechnological and Agricultural Research, Palacky University, Department of Protein
Biochemistry and Proteomics, Olomouc, Czech Republic
2
Institute of Experimental Botany AS CR, Department of Proteomics of Cell Structures, Olomouc, Czech Republic
Introduction: In eukaryotes, a majority of the hereditary information is localized in the nucleus. Therefore, almost all DNArelated processes occur in this organelle, including DNA repair, replication, recombination, and transcription as well as
appropriate regulatory circuits. These important biological phenomena cannot be completely understood without a secure
characterization of the nuclear proteins. Given the (potentially) high number of individual protein species to be analyzed,
modern high-throughput proteomic approaches are ideally suitable for comprehensive study of nuclear proteomes. Indeed,
many works have been published on proteomic analysis of animal nuclei over the past decade. On the other hand, only sparse
proteomic reports have dealt with nuclear proteins from plants. This clear imbalance is presumably due to troublesome sample
preparation caused by high integrity of plant cells, presence of many interfering secondary metabolites and high endogenous
protease activity. To shed new light on “the protein dark matter” of plant nuclei a protocol comprising flow cytometry sorting of
formaldehyde-fixed nuclei coupled to protein/peptide separation and MS analysis was proposed. Hereby, preliminary results
from its application to a study of barley cell nuclei in G2 phase are presented.
Objectives: The main objective of this study was to test the compatibility of plant nuclei flow cytometry sorting and subsequent
proteomic analysis of nuclear proteins.
Materials and methods: The methodology involved the following steps: barley seedling cultivation, formaldehyde fixation,
homogenization of seedling root tips, preparation of crude nuclei and their flow cytometry sorting, protein extraction, SDSPAGE, in-gel digestion of proteins, nanoLC-ESI-MS and nanoLC-MALDI-MS and bioinformatic analysis of acquired data.
Results: Using the proposed protocol, over 2000 proteins were identified covering three orders of protein abundances. A GO
term analysis confirmed that the vast majority of the proteins were nucleus-related with the most enriched processes connected
to chromatin assembly and organization, ribosome biogenesis and RNA processing. To our knowledge, this dataset represents
the largest complex of plant nuclear proteins identified in a single proteomic study so far.
Conclusion: Successful identification of thousands of proteins indicates that the presented protocol has several advantages.
Firstly, the results suggest that fast and simple sample preparation combined with formaldehyde fixation helps to preserve the
nuclear proteome from unwanted artificial changes (protein degradation etc.). Secondly, flow cytometry allows the separation
of G1, S and G2 phase nuclei and provides fractions with high purity. Thirdly, the protocol can be easily extended to any type of
plant tissue. Fourthly, as the intact plant chromosomes can be readily isolated employing the same flow sorting procedure, the
protocol can be adapted for the study of chromatin-associated proteins. Altogether, it seems that coupling flow cytometry with
proteomic techniques provides an elegant and powerful means for mapping of the plant nuclear proteome.
YIA6
Proteomic evaluation of nucleocytoplasmic partitioning and the CRM1-dependent nuclear exportome
1
2
2
1,2
S. Karaca , K. Kirli , D. Görlich , H. Urlaub
1
MPI for Biophysical Chemistry, Bioanalytical Mass Spectrometry Group, Goettingen, Germany
2
University Medical Center Goettingen, Department of Clinical Chemistry, Goettingen, Germany
Question: The defining eukaryotic organelle is nucleus, which is separated from cytoplasm by the nuclear envelope. Passive
diffusion and importin- as well as exportin-mediated active nuclear transport determine how proteins partition between the
cytoplasm and the nucleus. Despite the importance of this partition in eukaryotic life, it has not been thoroughly characterized
due to the challenges to obtain cytoplasmic and nuclear proteomes without cross-contamination. Therefore, we worked with a
model system (giant X. leavis oocytes) that allows manual micro-dissection to obtain cytoplasmic and nuclear proteomes with
unmatched purity. Further, to complement the nucleocytoplasmic partitioning, we employed affinity chromatography approach
to find "exportome" of most promiscuous exportin (CRM1).
Methods: Enucleation of X. laevis oocyte: Oocytes were manually micro-dissected to obtain cytoplasmic and nuclear proteomes.
Universal Proteomics Standard 2 was spiked into extracts to estimate absolute protein concentration.
CRM1/Exportin-1 affinity chromatography: CRM1 immobilized beads were incubated with X. leavis, HeLa and yeast extract in the
absence or presence of RanQ69L.
Proteomic Forum 2015
45
LC-MS/MS and data analysis: X. leavis cytoplasmic and nuclear extracts and eluted proteins from CRM1 affinity chromatography
were separated by SDS-PAGE and in-gel digestied with trypsin. Extracted peptides were analyzed LC system coupled to either
LTQ-Orbitrap Velos or Q-Exactive (both Thermo Scientific) as triplicates. Raw MS data were analyzed with MaxQuant (1).
Results: We quantified the nucleocytoplasmic distribution of ~7500 proteins that revealed trimodal distribution; ~2500 of these
showed an exclusive nuclear, ~2000 an exclusive cytoplasmic and the rest an intermediate localization (Figure 1.C). These
numbers had interesting implications e.g. for RNA metabolism, protein degradation, and organization of the cytoskeleton.
Further, we identifid hundreds of new CRM1 export cargoes (Figure 2.C) in each model organism. These include nearly all
translation factors, certain cytoskeletal proteins and vesicle coat proteins. We also validated several candidates by in vivo
fluorescence imaging.
Conclusions: We present a valuable resource for the nucleocytoplasmic distribution of a vertebrate proteome, provide insights
into eukaryotic compartmentation, and unprecedentedly expand the known CRM1 cargo repertoire. Majority of the these CRM1
cargoes relate to cytoplasmic activities that should be (at least temporarily) suppressed inside nuclei.
References: 1. Cox, J. & Mann,M., Nat. Biotechnol., 26, 1367-1372 (2008)
Figure 1 A. Workflow of spatial proteomics of X. leavis oocyte. B. SDS-PAGE analysis of cytosolic and nuclear extract. C.
Histogram plot shows log2 nuclear- to-cytosolic concentration ratio distribution of quantified proteins.
Figure 2 A. Workflow of CRM1 affinity chromatography (CAC). B. SDS-PAGE analysis of eluted proteins with and without
RanQ69L incubation. C. Scatterplot represents label free quantification of CAC. Blue-colored dots represent potential CRM1
cargoes.
Figure 1
Figure 2
Proteomic Forum 2015
46
Wednesday, 25 March 2015
Plenary Session • Systems Biology
09:00–10:30
Chairs: Marius Ueffing (Tübingen/DE) , René P. Zahedi (Dortmund/DE)
PL23
Genotypic Variability and the Quantitative Proteotype
1,2
R. Aebersold
1
ETH Zurich , Department of Biology, Institute of Molecular Systems Biology, Zurich, Switzerland
2
University of Zurich, Faculty of Science, Zurich, Switzerland
The question how genetic variability is translated into phenotypes is fundamental in biology and medicine. Powerful genomic
technologies now determine genetic variability at a genomic level and at unprecedented speed, accuracy and (low) cost. To date
the effects of genomic variability on the expressed information of the cell has been mainly studied by transcript profiling.
In this presentation we will discuss emerging computational and quantitative proteomic technologies to relate genotypic
variation to the proteome. Proteomic data to support such correlations need to be quantitatively accurate, highly reproducible
across multiple measurements and samples and generated at high throughput. Ideally, the data also would provide information
about spatial arrangement of proteins in the cell. Data with these qualities can now be generated by the targeted proteomic
methods selected reaction monitoring (SRM) and, at higher throughput, by SWATH-MS (1).
We will discuss the principles of these mass spectrometric methods, discuss the computational challenged they pose for data
analysis and demonstrate with selected applications, using genetic reference strain compendia, their ability to determine the
effect of genetic variability on the quantitative proteome, thus functionally connecting the genome to the proteome (2,3).
1. Gillet LC, Navarro P, Tate S, Roest H, Selevsek N, Reiter L, Bonner R, Aebersold R. (2012) Targeted data extraction of the
MS/MS spectra generated by data independent acquisition: a new concept for consistent and accurate proteome analysis
(2012).Mol Cell Proteomics 11:O111.016717.
2. Picotti P, Clément-Ziza M, Lam H, Campbell DS, Schmidt A, Deutsch EW, Röst H, Sun Z, Rinner O, Reiter L, Shen Q, Michaelson
JJ, Frei A, Alberti S, Kusebauch U, Wollscheid B, Moritz RL, Beyer A, Aebersold R. (2013) A complete mass-spectrometric map of
the yeast proteome applied to quantitative tranalysis.Nature. 494(7436):266-70.
3. Multilayered genetic and omics dissection of mitochondrial activity in a mouse reference population (2014). Wu Y, Williams
EG, Dubuis S, Mottis A, Jovaisaite V, Houten SM, Argmann CA, Faridi P, Wolski W, Kutalik Z, Zamboni N, Auwerx J, Aebersold R.
Cell. 158(6):1415-30. doi: 10.1016/j.cell.2014.07.039.
PL24
Genotypic Variability and the Quantitative Proteotype
M. Mann
Martinsried, Germany
TBA.
PL25
Protein abundance evolution and its systems-level constraints
,
C. von Mering
Zurich, Switzerland
TBA.
Proteomic Forum 2015
47
Plenary Session • Proteomics in Health and Disease
11:00–12:30
Chairs: Stefanie Hauck (Munich/DE) , Helmut E. Meyer (Bochum/DE)
PL26
Casting doubt on conventional approach to cancer biomarker development by proteomics
1
T. Kondo
1
National Cancer Center Research Institute, Division of Rare Cancer Research, Tokyo, Japan
Casting doubt on conventional approach to cancer biomarker development by proteomics
Biomarker development is a major subject in cancer proteomics. Biomarkers are required for better clinical outcome, and
proteomics was developed to address to clinical requirement. Unfortunately, beside substantial technology development and
investment in proteomics, no biomarker reported by proteomics was proven to be beneficial for cancer patients. The time has
come to reconsider the conventional proteomic approach to biomarker development. Technical limitation for comprehensive
study has been considered as a bottle neck of proteomic approach, and more efforts will be developed to technology
development. In global mRNA expression study, almost all gene products have been observed by array technology since more
than a decade ago. However, there are only few cases that comprehensive approach at mRNA identified biomarker genes for
clinical practice. Thus, we cannot expect practical biomarkers just by increasing the number of observable proteins by novel
proteomics technologies. In the study of screening biomarkers for early diagnosis, prior probability of disease is always ignored,
and the case and control samples of similar numbers are always compared. In reality, the number of cancer patients is very small
at the screening step, and such an unrealistic study design will never result in practical biomarkers. Comprehensive approach
often results in multiple biomarkers. Heterogeneous mechanisms underlying single cancer types or clinical observations may
justify the use of multiple biomarkers which may need initial investment for novel device, and thus should be considerably more
expensive than conventional biomarker examination. However, components of multiple biomarkers developed by proteomics
usually don’t fully explain the heterogeneity of diseases, and it is a bold hypothesis that the components of multiple biomarkers
can be a novel hub of complex mechanisms of diseases. Comparison of patients with different prognosis has been performed to
identify novel prognostic biomarkers. Many prognostic factors have been reported by clinical and pathological observations, and
the patients in the prognostic biomarker study should be well stratified according to them. In conclusion, novel comprehensive
proteomics technology will not generate promising biomarker by itself, and realistic theme setting and knowledge of cancer
biology and clinical oncology are fundamentally necessary for successful biomarker development.
Reference: T Kondo, BBA 1844 (2014) 861-865
PL27
N-terminomics in immunity: Mechanisms of disease instead
C. M. Overall
Vancouver, BC, Canada
TBA.
PL28
MRM and iMALDI for Clinical Research and Diagnostic
1
1
C. Borchers , C. Borchers
1
University of Victoria - Genome BC Proteomics Centre, Biochemistry & Microbiology, Victoria, Canada
Dried blood spot (DBS) methodology has improved sample collection workflows for clinical diagnostics targeting small molecule
biomarkers. To extend these benefits to protein targets, we have recently developed a multiple reaction monitoring mass
spectrometry (MRM-MS) assay for quantifying 103 proteins using stable isotope-labeled standards and standard curves. High
precision was achieved as demonstrated by an average intra-assay CV 7.1% and inter-assay CV of 10.7%. The overall protein
dynamic range spans endogenous protein concentration across almost 5 orders of magnitude, from 15 mg/mL down to 176
ng/mL. Furthermore, the majority of proteins are stable in the DBS samples (within 20%) when stored at room temperature for
at least 4 weeks. These results demonstrate that DBS-MRM holds considerable potential for clinical proteomics applications in
advancing biomarker screening and personalized medicine.
Proteomic Forum 2015
48
To quantify lower abundance protein targets we have also developed immuno-MALDI (iMALDI) technology which combines the
sensitivity of immunoaffinity capture with the specificity of MS detection. We have now taken a multifaceted approach for
translating our iMALDI technology into clinical laboratories for routine protein quantification. First, we have automated the
sample preparation using the Agilent Bravo liquid handling robot for improved sample throughput. Secondly, we have optimized
iMALDI assays for the Bruker Microflex MALDI-TOF, a bench-top instrument that is already widely used in regulated healthcare
environments. Here we demonstrate iMALDI technology for the clinical measurement of plasma renin activity (PRA), an
established biomarker for primary aldosteronism. The current method automates 192 iMALDI captures and analysis requires
only seconds per sample. Initial validation indicates a strong correlation between iMALDI-PRA values and a clinical LC-MS/MS
method (R2 = 0.95, n=188). This iMALDI approach to protein quantification can be multiplexed and is applicable to a wide array
of clinical peptide and protein biomarker targets.
Symposium • Computational Omics I
14:00–15:00
Chair: Lennart Martens (Ghent/BE)
O19
Proteogenomic RNA-Seq Data Integration beyond six-frames and single nucleotide variations
1
1
F. Zickmann , B. Y. Renard
1
Robert Koch Institute, NG4 Bioinformatics, Berlin, Germany
Introduction and Objectives: Ongoing advances in high-throughput technologies have facilitated accurate mass spectrometry
measurements and offer a wealth of information on genomic and transcript level. In proteogenomics this multi-omics data is
combined to allow more accurate sample-specific predictions. Although proteogenomic studies become more and more
popular, existing analysis methods still mainly depend on six-frame translations or reference protein databases that are
extended with transcriptomic information or known single nucleotide polymorphisms (SNPs). Since (i) six-frame analyses can
result in bias for peptide identification due to an artificial six-fold increase of the database size, and (ii) extended protein
databases might not contain all information necessary to identify novel or mutated genes, sample-specific approaches
independent from databases are desirable.
Methods: We present MSProGene (Mass Spectrometry and RNA-Seq based Protein and Gene Identification) a novel
computational method that integrates genomic, transcriptomic and proteomic information to infer and verify genes and their
protein products including all their variations. In particular, MSProGene is independent from existing protein reference
databases by constructing customized transcript databases. First, an RNA-Seq read mapping is analyzed to infer gene models,
which are updated by including variations present in the RNA-Seq reads. These sequences are translated to protein sequences
for a peptide search with tandem mass spectra. MSProGene performs protein inference by reassigning shared peptides using a
linear program approach based on RNA-Seq information. Shared peptides are not ignored, but rather all identifications are
represented by a network that is resolved using the peptide spectrum match score and the reliability of each gene model.
Finally, peptide identifications are controlled with regard to their false discovery rate (FDR) by estimating the distribution of true
and false positives with an EM algorithm.
Results: We applied MSProGene on two data sets of B. henseale and L. sigmodontis and compared it to a standard six-frame
translation and existing proteogenomic analysis methods. MSProGene shows a favorable accuracy in protein identifications by
finding more than 700 additional proteins at the same annotation-based FDR of 1%. In addition, it detected novel proteins,
confirmed by manual inspection of RNA-Seq reads and tandem mass spectra.
Conclusion: MSProGene is a novel method for proteogenomic analysis that performs data integration beyond extension of
protein databases and six-frame translations. It includes RNA-Seq to provide accurate protein identifications and to approach
shared peptide protein inference.
O20
Summarization vs. Peptide-Based Models in Label-free Quantitative Proteomics : Performance, Pitfalls and Data Analysis
Guidelines
1,2,3
2,3
2,3
1
G. Ludger , A. Argentini , L. Martens , L. Clement
1
Ghent University, Department of Applied Mathematics, Computer Science and Statistics, Ghent, Belgium
2
Ghent University, Department of Biochemistry, Ghent, Belgium
3
VIB, Department of Medical Protein Research, Ghent, Belgium
Proteomic Forum 2015
49
Quantitative label-free mass spectrometry is increasingly used to analyze the proteomes of complex biological samples.
However, the choice of appropriate data analysis methods remains a major challenge. Objective comparisons and general
guidelines for the practitioner are still lacking, which impedes the dissemination of more efficient data analysis pipelines into the
proteomics community. We therefore provide a rigorous comparison between methods that directly adopt statistical models to
the peptide intensities and more popular modular data analysis strategies that first summarize the peptide intensities into
protein expression values prior to the statistical analysis.
We assessed three peptide-based methods and four modular approaches, including two popular Perseus-MaxQuant workflows,
in our study. We exploited the availability of benchmark data sets to provide insight into performance differences between the
different methods and to pinpoint technological artifacts that often arise in label free proteomics experiments. We show that
peptide-based methods outperform aggregation-based pipelines in terms of sensitivity, specificity, accuracy and precision. We
also demonstrate that predefined FDR cut-offs can become problematic when differentially expressed proteins are highly
abundant in one or more samples. Care should therefore be taken when data is interpreted from samples with spiked-in internal
controls, and from samples that contain a few very highly abundant proteins. We do however, show that specific diagnostic plots
can be used for assessing differentially expressed proteins, and the overall quality of the obtained fold-change estimates. Finally,
our study also illustrates that imputation under the missing by low abundance assumption is beneficial for the detection of
differential expression in low abundant proteins, but that it negatively affects moderate to high abundant proteins. Hence,
imputation strategies that are commonly implemented in standard proteomics software should be used with care.
Figure 1
PL29
Applications of predictions of peptide properties
L. Käll
Solna, Sweden
TBA.
Proteomic Forum 2015
50
Symposium • Extracellular Vesicles
14:00–15:00
Chair: Corinna Henkel (Dortmund/DE)
O21
Analysis of extracellular vesicles(EVs)in diagnostics and quality control of blood products in transfusion medicine
1
1
1
1
1
2
A. Pienimäki-Römer , K. Kuhlmann , G. Liebisch , H. Meyer , E. Orso , G. Schmitz
1
Uni-Klinik Regensburg, Institut für Laboratoriumsmedizin und Transfusionsmedizin, Regensburg, Germany
2
Universitätsklinik Regensburg, Institut für Laboratoriumsmedizin und Transfusionsmedizin, Regensburg, Germany
Background: Extracellular vesicles (EVs) are cell-derived secretory and apoptotic vesicular structures, shed into the extracellular
milieu. Recently, EV research rapidly expands due to available novel technologies and biomarkers for their analysis. Three types
of human EVs have been characterized by multi-omics analysis: (i) five subsets (F1-F5) of platelet-derived EVs (PL-EVs) isolated
by density gradient centrifugation from platelet concentrates (PCs) during platelet senescence, (ii) erythrocyte-derived EVs (RBCEVs) in erythrocyte concentrates (ECs) and PCs, and (iii) neutrophil-derived EVs (PMN-EVs) of isolated cells, exposed to
atherogenic lipoproteins or inflammatory stimuli.
Methods: The EVs were analyzed by nanoparticle tracking analysis, advanced high sensitivity flow cytometry (Apogee 50M),
mass spectrometry-based proteomics and lipidomics and microarrays for micro-RNA (miRNA) expression.
Results: PL-EVs represent the major part of all EVs in PCs, ECs and plasma. The subsets F1-F2 have the highest content of ESCRTs,
RNA-binding proteins, CD62P, the Parkinson´s disease (PD)-related alpha-synuclein (aSyn), and free cholesterol (FC). Major
biomarkers in F2-F4 include CD63, CD107a, the Alzheimer’s disease (AD)-associated amyloid-precursor protein (APP) and 7ketocholesterol. Minor lipid classes, such as lysophosphatidic acid, phosphatidylserine, ceramide, phosphatidylglycerol and
cardiolipin, predominate in F5. The subsets F3-F5 are enriched in caveolin-1 and apolipoproteins, whereas F5 accumulate
mitochondrial proteins. PL-EVs increase during in vitro platelet senescence, parallel with enrichment of AD- and PD-related
miRNAs. In comparison to PL-EVs, RBC-EVs are particularly enriched in aSyn, sphingomyelin and FC, parallel with diminished
content phosphatidylcholine and plasmalogens.
PMN-EVs dramatically shed in response to atherogenic lipoproteins and LPS. PMN-EVs are chemotactic for monocytes and
induce apoptosis/necrosis, prior to splenic clearance or entering vascular lesions.
Conclusion: Qualitative and quantitative changes in blood cell-derived EVs are paralleled with disease pathologies, and their
biomarkers may reflect the quality and senescence of hemotherapeutic products and open up new avenues towards EVapheresis therapy.
O22
Development of a blood test for Alzheimer's disease by proteomic identification a specific phenotype in platelets
1
2
2
3
1
2
M. Zellner , M. Veitinger , E. Umlauf , G. Mitulovic , R. Oehler , E. Rappold
1
Medical Universit of Vienna, Surgical Research Laboratories, Vienna, Austria
2
Medical University of Vienna, Center of Physiology and Pharmacologz, Vienna, Austria
3
Medicla University of Vienna, Clinical Institute of Laboratory Medicine, Vienna, Austria
Background: Alzheimer’s disease (AD) is a chronic, neurodegenerative disorder and one of the major health care challenges of
the 21st century. Due to still unknown aetiology, ‘definite’ AD diagnosis relies on post mortem brain examination, while ante
mortem only a diagnosis of ‘probable’ AD is possible; this is achieved either by neuropsychological assessment or brain scans. On
the molecular level, AD-specific cerebrospinal fluid biomarkers are promising, however, the invasive lumbar puncture renders
their routine application impossible. Consequently, alternative sample materials and minimally invasive diagnostic procedures
are urgently needed. Based on the shared biological similarities of platelets and neurons, these peripheral cell fragments were
used to characterize AD-specific biomarkers.
Methods: The study was approved by the local ethics committee and blood was collected from 62 patients with clinically
suspected AD and 63 age/sex-matched controls. Their platelet proteomes were resolved in two independent discovery and
verification phases using fluorescence 2D-DIGE; AD-regulated proteins were tryptically digested and MS-identified. APOE 4 and
GSTO1*A140 genotyping was performed. For high-throughput analysis with the novel multiplex protein biochip, platelet-rich
plasma (PRP) was prepared from whole blood (20 min, 80 x g) and stored at -80 ˚C. After thawing, platelets were isolated and
lysed with SDS before addition of 2% BSA/PBS to bind excess SDS. Lysates were applied to the protein biochip (immunoassay
sandwich principle) and protein biochips subsequently imaged on the Evidence Investigator analyser.
Proteomic Forum 2015
51
Results: Four proteins were highly significantly up-regulated in LOAD samples: monoamine-oxidase-B (MaoB), tropomyosin-1
(Tm1), apolipoprotein E4 (ApoE4), and glutathione S-transferase omega1 isoform A140 (GSTO1*A140). While both isoforms
result from single nucleotide polymorphisms, the latter was highly predominant in APOE 4-negative patients: genotyping
revealed significantly more APOE ε4 carriers in the AD (66%) than in the control (11%) group and presence of exclusively two
GSTO1*A140 alleles in non-APOE ε4 AD patients (n=20) relative to 38% in controls (30% in non-APOE ε4 controls) and 32% in
APOE ε4-positive AD patients. Biochip analysis correctly identified 98% of all samples of the GSTO1*A140 and 100% of the APOE
ε4 genotype by normalisation with either ERK2 or panApoE concentrations. Biochip quantification of Tm1 and MaoB (ERK2normalised) also replicated the higher expression of these two proteins in AD patients relative to controls. An algorithm utilising
these four biomarkers yielded the highest separation power for AD and control samples with a ROC AUC of 0.969 (95% CI=0.9440.994).
Discussion: The combination of four AD-regulated platelet proteins, including the most powerful genetic risk factor of late-onset
AD, APOE ε4, enabled identification of AD patients with a sensitivity of 94% and a specificity of 89%. This demonstrates the
utility of this innovative multiplex device as reliable peripheral diagnostic tool to aid ante mortem AD diagnosis in a routine
blood-based clinical screening.
PL30
Role of extracellular vesicles in neurological diseases associated with protein misfolding
1
A. Hill
1
University of Melbourne, Biochemistry and Molecular Biology, Parkville, Australia
Neurodegenerative diseases affecting humans, such as Alzheimer's, Parkinson's and prion diseases, are associated with the
misfolding and spreading of neurotoxic forms of the pathogenic proteins involved in each of these disorders. Prion diseases arise
by the misfolding of the prion protein (PrP) which gains a unique infectious property and has been shown to transfer prion
infection in association with exosomes. Our research has examined several facets of exosome and extracellular vesicle biology,
including: (i) using cryo-electron microscopy with tomography to analyse the ultrastructure of highly purified vesicles; (ii) using
functional assays to determine the components of vesicles required to transfer prion infectivity between cells; (iii)
demonstrating novel processing of prion proteins and the amyloid precursor protein (APP) in exosomes for both prion and
Alzheimer’s diseases; (iv) using next generation sequencing to compare exosomal miRNA profiles in normal and disease
associated exosomes; (v) analysing the protein content of extracellular vesicles from neurons and brain tissues. Together, our
multi-omics approach to studying the cargo contained within these vesicles has revealed potential mechanisms for disease
pathogenesis and also identified disease biomarkers which have utility in the diagnosis of these diseases.
Symposium • Post-translational Modifcations
14:00–15:00
Chair: Martin Røssel Larsen (Odense/DK)
O23
Chemical proteomics highlights protein kinases involved in macrophage differentiation and function
1
1
1
1
E. Richter , M. Harms , K. Ventz , J. Mostertz , F. Hochgräfe
1
University of Greifswald, Greifswald, Germany
1
Introduction and objectives: Protein kinase alterations during macrophage differentiation have not been interrogated
comprehensively, despite fundamental links between kinase activity and cell development and specialization. We hypothesize
that this approach may identify key protein kinases involved in the regulation of differentiation and macrophage-specific
functions including cytokine/chemokine production and defense against pathogens.
Methods: We applied stable isotope labeling with amino acids in cell culture (SILAC) combined with small molecule based kinase
purification, phosphopeptide enrichment and high resolution mass spectrometry to study the kinome of phorbol myristate
acetate (PMA)-induced differentiation of the human monocytic cell line THP-1. Macrophage-specific kinases were further
characterized by ELISA, fluorescence microscopy and bacterial phagocytosis assays.
Results: We quantified 168 protein kinases encompassing all families of the human kinome. Interestingly, whereas more kinases
were decreased (63) than increased (27) in their amount following differentiation, a general increase in overall phosphorylation
was observed. Based on our expression and phosphorylation data, a number of protein kinases including MAPK13, MerTK,
CaMK2b and MAP3K7 (TAK1) showed dramatic regulation suggesting vital roles for macrophage function or even differentiation.
Functional analysis using the TAK1 specific inhibitor 5Z-7-oxozeaenol uncovered TAK1 as key regulator of PMA-induced
Proteomic Forum 2015
52
differentiation in the THP-1 background. In addition, we also got indications that TAK1 is involved in efficient phagocytosis and
killing of the human bacterial pathogen Staphylococcus aureus.
Conclusions: With help of chemical proteomics we could describe the reorganization of the kinome during monocyte-tomacrophage transition and highlight protein kinases potentially essential for differentiation and specialization of macrophages.
In this regard, TAK1 activity could already been identified as key regulator.
O24
Identifying Kinase Substrates via a Heavy ATP Kinase Assay and Quantitative Mass Spectrometry (HAKA-MS)
1
2
1
3
4
3
A. Müller , R. Giambruno , P. Májek , A. Hofer , J. Bigenzahn , H. Jessen , K. Bennett
1
CeMM, Mass Spectrometry, Vienna, Austria
2
San Raffaele Scientific Institute, Division of Regenerative Medicine, Milan, Italy
3
University of Zurich, Department of Chemistry, Zurich, Switzerland
4
CeMM, Molecular Networks and Systems Pharmacology, Vienna, Austria
1
Introduction: Tyrosine kinases play a central role in numerous cellular signal transduction pathways and are frequently found
de-regulated in many human pathologies such as cancer. Several approaches have been developed to correctly identify tyrosine
kinase substrates, although the results are often biased by false positive data.
Objectives: In order to improve confidence in in vitro kinase substrate identification, we have developed a novel kinase assay
18
coupled to quantitative mass spectrometry. In the presence of stable, isotope-labelled γ-[ O2]-ATP, we were able to monitor the
de novo phosphorylation of kinase substrates. The non-receptor tyrosine kinase ABL1 was chosen as a model because of the
extensive knowledge that is available on ABL1 substrates.
Materials & Methods: Our kinase assay was performed on HEK293 cytoplasmic protein extracts that were pre-treated with the
pan kinase inhibitor 5'-[p-(fluorosulfonyl)benzoyl]adenosine (FSBA) to abolish endogenous cellular kinase activity. Extracts were
incubated with either immunopurified constitutively active ABL1-PP, or the catalytically inactive ABL1-Kin in the presence of
18
exogenous γ-[ O2]-ATP. The reactions were monitored at 30, 90 and 150 minutes to assess the phosphorylation kinetics. After
quenching, the proteins were digested, the tyrosine phosphorylated peptides enriched with anti-phosphotyrosine antibodies,
and analysed by quantitative mass spectrometry using TMT 6-plex reagents.
Results: The 4 Da mass increments marked sites of de novo phosphorylation and enabled dissection of new phospho-sites from
the pre-existing ‘light’ phosphoproteome. Thus, 105 unique ‘heavy’ phosphotyrosine (pY) sites from 73 phosphoproteins were
identified. Of these, 16% are known ABL1 substrates and DDX3X was confirmed as a new in vivo substrate. Additionally, 22
proteins were phosphorylated at the preferred ABL1 YxxP consensus motif. For 17 proteins, multiple in vitro pY-sites were
identified. Whilst the inactive controls remained relatively stable, phosphorylation levels for active ABL1-PP followed
characteristic reaction progress kinetics.
Conclusion: HAKA-MS increases the reliability and confidence of in vitro substrate identification by not only dissecting new
phosphorylation sites from the basal phosphoproteome; but also by taking the kinetic profile of substrate phosphorylation into
account. Our method preserves the physiological environment of the kinase and retains protein-protein interactions and protein
post-translational modifications. This is contrary to kinase assays performed on heat-inactivated or phosphatase-treated extracts
using tag-purified recombinant kinases. Thus, HAKA-MS is highly beneficial as readily-available technologies are merged and
many pitfalls of MS-based in vitro kinase assays are circumvented.
O25
Hydrophilic interaction chromatography for the enrichment of sialylated glycopeptides
1
1
1
L. Drici , M. Ibanez Vea , M. R. Larsen
1
Southern University of Denmark, Biochemistry and molecular biology, Odense, Denmark
Introduction: Sialic acids constitute a large subset of the even larger family of nine-carbon a-keto acids called nonulosonic acids.
They are attached to glycoproteins, proteoglycans and glycolipids as terminal entities of larger glycan structures. Proper sample
handling and well-designed analytical strategies including pre-LC-MS methodologies is a necessity for obtaining high quality
information of glycoprotein sialylation. The lability of the sialic acid residues together with its negative charge and its structural
heterogeneity demands dedicated workflows.
Proteomic Forum 2015
53
Objectives: In the present study we are testing the effect of sialic acid on glycopeptides for purification by HILIC materials. We
speculate that the sialic acids significantly influence the retention of sialylated peptides on HILIC material similarly to
multiphosphorylated peptides.
Materials and methods: In order to evaluate the behaviour of sialylated glycopeptides on 3 different HILIC stationary phases,
bovine fetuin was tryptic digested and the sialylated glycopeptides were enriched using titanium dioxide and subjected to
dimethyl labeling. The heavily labeled fractions were loaded onto HILIC microcolumns, and the glycopeptides were selectively
eluted using a two step sequential elution using 0.1% TFA and 2,5-dihydroxybenzoic acid (DHB) in 60% ACN, 0.1% TFA. The
fractionated samples (heavy) and the light labeled samples were then mixed (1:1), deglycosylated and subjected to mass
spectrometry analysis. Furthermore, fetuin sialylated glycopeptides were treated with sialidase A and subjected to the same
experiment in order to investigate whether the degree of sialylation is the main reason of the retention behavior.
Results: Our preliminary results have showed that sialic acid has not an influence on the behavior of the sialylated glycopeptides
when Polyhdroxyethyl A resin was used as HILIC microcolumn. On the contrary, several higher mass glycopeptides appeared in
the second DHB elution when ZIC-HILIC and TSK amide-80 resins were used; these could be associated with the glycopeptides
bearing longer peptide moieties and more sialic acids than those eluted with 0.1%TFA. Moreover, the elution of desialylated
glycopeptides was found to be more efficient using only 0.1% TFA without further elution with DHB.
Conclusion: The length of the peptide moiety might influence the binding affinity of glycopeptides to ZIC-HILIC and TSK amide-80
sorbent and solvent additives, such as DHB, influence the glycopeptide elution after HILIC chromatography. Moreover, sialylated
glycopeptides bind ZIC-HILIC and TSK amide-80 sorbents stronger than non-sialylated glycopeptides under commonly used acidic
conditions.
Symposium • Computational Omics II
15:30–16:30
Chair: Lukas Käll (Solna/SE)
O26
MS/MS-free protein identification in complex mixtures using multiple enzymes with complementary specificity
1,2
1
3
1
1,2
1,2
1,2
1,2
3
M. Ivanov , I. Tarasova , K. Srzentić , M. Pridatchenko , L. Levitsky , A. Lobas , E. Solovyeva , J. Bubis , Y. Tsybin , M.
1,2
Gorshkov
1
V.L. Talrose Institute for Energy Problems of Chemical Physics, Russian Academy of Sciences, Moscow, Russian Federation
2
Moscow Institute of Physics and Technology (State University), Dolgoprudny, Russian Federation
3
Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
Introduction: Peptide mass fingerprinting of complex mixtures regains interest recently with the progress in mass spectrometry
technologies. It potentially gives higher dynamic range compared with tandem mass spectrometry. However, identification and
scoring of peptide features detected in MS1 spectra are still challenging if not impossible.
Objectives: In this work, we present a workflow based on the multi-enzyme digestion strategy which allows MS1-based protein
identification in “shotgun” applications. In the proposed strategy, several cleavage reagents of different specificity are used for
parallel digestion of the protein sample followed by MS1-based database search.
Methods&Materials: We have performed the workflow evaluation in silico for the human protein database. Theoretical MS1
spectra of proteolytic peptides were generated for 2000 random proteins extracted from the database. In this evaluation we
estimated the number of identified proteins for different mass measurement and retention time prediction accuracies, as well
as the fraction of unidentifiable peptide-like features in the theoretical MS1 spectra. In the follow-up proof-of-principle
validation of the proposed strategy, experimental data were obtained for the model mixture of 7 proteins using high resolution
mass spectrometry and employing trypsin and a number of chemical digestion reagents.
Results: We found that up to 78% of the proteins can be unambiguously identified for the routinely achievable 1 ppm and 3 min
mass measurement and retention time prediction accuracies, respectively. All proteins from the experimental mix were
identified with the proposed strategy and compared with the results of standard database search using tandem mass
spectrometry data.
Conclusion: Potentially, the MS/MS free proteome analysis can become a strong alternative to currently employed ”shotgun”
strategies based on tandem mass spectrometry. Experimental implementation of the proposed approach is in progress.
Proteomic Forum 2015
54
PL31
Speaking to spectra: what we can learn without using search engines
L. Martens
Ghent, Belgium
TBA.
PL32
Advances in pulse-chase experiments using mass spectrometry to reveal protein turnover rates - a case study on the highly
encumbered photosynthetic protein complexes.
1
1
J. Barth , C. Fufezan
1
University Muenster, IBBP, Münster, Germany
Pulse-chase experiments analyzed by mass spectrometry allow deeper insights into protein turnover rates, metabolic label
incorporation and metabolic fluxes. However, the analysis of pulse-chase experiments MS data is currently not straightforward.
For example, peptide to spectrum matches (PSMs) become very difficult to resolve as soon the chase has commenced and the
chase isotope is built into amino acids, peptides and proteins. Naturally, this depends on the isotope that is used for the pulse
and chase. Nevertheless partially labeled peptides are a challenge since they are not supported by current search engines or
quantification tools.
Our aim was to overcome these limitations. For this we have developed piqDB, which is, at its core, a mass spectrometry data
analysis and storage platform. PiqDB (protein identification and quantification database) is based on mongoDB and hadoop and
includes our Python libraries W14 and pyQms for job distribution and mass spectrometry data quantification, respectively. Since
hadoop and mongoDB support current state-of-the-art Big Data analysis techniques, we are able to analyze experiments with
1000+ 2h LC-MS/MS runs on our 5 node developer cluster with ease. The complete framework allows a) high quality retention
time alignments, b) quantification of partially labeled peptides - the very basic prerequisite for pulse-chase experiments analysis
- and c) enhancement of the data. We have used our tools to investigate protein turnover rates in C. reinhardtii, a green algae
which shows high protein turnover rates of photosynthetic protein complexes due to the nature of oxidative photosynthesis. A
technical description of piqDB and the results of light / oxidative stress induced increased protein turnovers will be presented.
Symposium • Structural Proteomics
15:30–16:30
Chair: Christoph Borchers (Victoria, BC/CA)
O27
Using cross-linking coupled to mass spectrometry and integrated modeling to study the molecular architecture of the
40S•eIF1•eIF3 translation initiation complex
1
2
3
2
2
2
3
2
3
F. Stengel , J. P. Erzberger , R. Pellarin , S. Zhang , T. Schaefer , C. H. S. Aylett , P. Cimermančič , D. Boehringer , A. Sal , R.
1
2
Aebersold , N. Ban
1
ETH Zurich, Institute of Molecular Systems Biology HPT E53, Zurich, Switzerland
2
ETH Zurich, Institute of Molecular Biology and Biophysics, Zurich, Switzerland
3
University of California, 3Department of Bioengineering and Therapeutic Sciences, Department of Pharmaceutical Chemistry,
and California Institute for Quantitative Biosciences, San Francisco, United States
Question: The integration of hybrid methods to interpret complex structural assemblies has been a longstanding goal in
molecular biology. Emerging methods in structural mass spectrometry (MS) can pose an exciting alternative to classical methods
for the determination of the structure of protein complexes, due to their low sample requirements and comparatively high
measuring speed.
Methods: We combined cross-linking MS with integrated modeling in order to generate a generic hybrid approach for the
accurate modeling of large proteasomal assemblies. This integrated platform was subsequently used to elucidate the structure
of eIF3 bound to the ribosome. EIF3 is the largest initiation factor and has evaded crystallization efforts for two decades.
Proteomic Forum 2015
55
Results: Our hybrid approach permits for a large number of low resolution restraints to be effectively integrated and optimized,
allowing us to position and orient all eIF3 components on the 40S•eIF1 complex, revealing an extended, modular arrangement
of eIF3 subunits on the ribosome.
Conclusions: This approach takes advantage of the fact that many proteasomal assemblies, in particular highly heterogeneous,
very large or dynamic systems, remain refractory to structure determination by a single more conventional approach, but are in
principle addressable by such a hybrid approach. Thus we believe that this approach will become a generic pairing in the future.
*
*
*
References: Jan P. Erzberger , Florian Stengel , Riccardo Pellarin , Suyang Zhang, Tanja Schaefer, Christopher H. S. Aylett, Peter
Čimermančic, Daniel Boehringer, Andrej Sali, Ruedi Aebersold and Nenad Ban, Molecular Architecture of the 40S⋅eIF1⋅eIF3
Translation Initiation Complex, Cell, 2014 Aug 28;158(5):1123-35. [OA]
*equal contribution
PL33
Scalable HX-MS methods to characterize ultra-large protein complexes
1
D. Schriemer
1
University of Calgary, Calgary, Canada
The integration of biophysical data from multiple sources is critical for developing accurate models of large multiprotein
systems. Mass spectrometry can be used to measure the insertion location for a wide range of chemical probes. The insertion
data can provide a rich but disparate set of modeling restraints, useful for developing domain orientations and interfaces, as well
as modes of conformational dynamics. Hydrogen/deuterium exchange (HX) chemistry is an obvious choice for all but the
orientational data, as it reports on the H-bonding environment for every residue in a protein. Unfortunately, data complexity
and the labile nature of the chemistry and have limited the method to single proteins and smaller complexes. Here we describe
a comprehensive set of solutions to support the profiling of ultralarge multiprotein systems, and we apply them to regulatory
mechanisms of the mitotic spindle. New digestion reagents, labeling routines, MS data acquisition methods and software are
combined to generate a pipeline from bench to protein modeling. We introduce a new set of proteases for protein mass
spectrometry, with cleavage specificities that are ideal for both bottom-up and middle-down strategies in HX-MS. Specific
2
attention will be given to new data-independent HX-MS modes that enhance sequence coverage, and simplifies the allimportant HX data validation exercise for ultra-large complexes.
O28
Crosslinking-MS analysis combined with molecular modelling provides novel insight into the architecture of the Parkinson’sassociated protein kinase LRRK2
1,2
3
4
5,6
2
2
7
5,6
G. Guaitoli , F. Raimondi , B. K. Gilsbach , P. Kumar Jagtap , K. Boldt , F. von Zweydorf , K. Gotthardt , A. Geerlof , M.
5,6
4
1,2
1,2,8
Sattler , A. Kortholt , M. Ueffing , C. J. Gloeckner
1
Helmholtz Zentrum München, Research Unit Protein Science, Neuherberg, Germany
2
Eberhard Karls University Tübingen, Medical Proteome Center, Tübingen, Germany
3
University of Modena and Reggio Emilia, Modena, Italy
4
Rijksuniversiteit Groningen, Groningen, Netherlands
5
Helmholtz Zentrum München, Institute of Structural Biology , Neuherberg, Germany
6
Technical University Munich, Chair Biomolecular NMR-Spectroscopy at TUM Department Chemie, Garching, Germany
7
Max Planck Institute of Molecular Physiology, Dortmund, Netherlands
8
DZNE -- German Center for Neurodegenerative Diseases, Tübingen, Germany
Leucine-rich repeat kinase 2 (LRRK2) is a complex multiple domain protein, member of the Roco protein family of G-domain
proteins and it also exerts kinase activity. Mutations in the LRRK2 gene are associated with familial and sporadic Parkinson's
disease (PD). PD-associated mutations, at least in its kinase domain, have been reported to increase kinase activity. Besides a
kinase, a Roc GTPase and a COR dimerization domain, LRRK2 contains four solenoid-like repeat domains (from N-terminal to Cterminal, Armadillo, Ankyrin, Leucine-rich repeat and WD40) predicted to be involved in protein-protein interactions or
intramolecular scaffolding.
Given that PD-associated LRRK2 mutations lead to an increased kinase activity, LRRK2 has emerged as an important drug target.
As a prerequisite for rational drug design a characterization of the molecular mechanisms of activation are of key importance.
Detailed understanding of how kinase activity and GTPase activity are cross-regulated as well as insight into the structurefunction relationship of inter- as well as intramolecular domain-domain interactions are important to advance pharmacological
Proteomic Forum 2015
56
targeting of this protein. Although, first high resolution structures for orthologous bacterial proteins are available, including the
Roc-COR interface of a Roco protein form Chlorodium tepidum, the three-dimensional orchestration of the Roco protein
domains, including those of LRRK2, and their activity-dependent conformations are largely uncharacterized.
To gain such mechanistic and structural insights we combined classical biochemical approaches with structural proteomics and
molecular modeling towards a low-resolution model of the LRRK2 quaternary structure. By purification via a Strep/FLAG affinity
tag we were able to produce highly pure full-length human LRRK2 from HEK293T cell culture, which is of suitable quality for lowresolution structural investigation. Characterization by Blue-Native PAGE and DLS revealed that LRRK2 forms constitutive dimers.
By combining chemical crosslinking via isotope labeled NHS-based cross linkers of different length (DSS and DSG), mass
spectrometry and subsequent data analysis by XQuest/XProphet we were able to generate a crosslink map which we used as a
restraint in an integrative modeling approach.
The low-resolution model obtained suggests a compact folding of the LRRK2 homodimer. By providing insights into the
mechanistic interplay between the catalytically active domains and the scaffolding domains, including the N-terminal ankyrin
and leucine-rich repeats, our model adds mechanistic insight to a wealth of biochemical data on the regulation of LRRK2 kinase
activity by the Roc GTPase domain.
Closing Lecture
16:45-17:45
Chair: Marius Ueffing (Tübingen/DE)
PL 34
Metabolic profiling technologies and applications
E. Holmes
London, United Kingdom
TBA.
Proteomic Forum 2015
57
Poster Sessions
Poster Session A • 22 March 2015, 19:00–21:00 .................................................................................................................. 58
Bioinformatics (P014–P030) ..........................................................................................................................................................................58
Free topics (P032–P049)................................................................................................................................................................................69
Proteomics for Human Health I (P123–P145) ...............................................................................................................................................83
Proteomics in Biotechnology (P192–P198) ...................................................................................................................................................96
Quantitative Proteomics I (P199–P215) ......................................................................................................................................................101
Poster Session B • 23 March 2015, 15:00–16:30 ................................................................................................................. 112
Affinity Proteomics (P001–P013) ................................................................................................................................................................112
Imaging (P050–P055) ..................................................................................................................................................................................120
Mass Spectrometry Technologies (P056–P065) ..........................................................................................................................................123
Plant and Microbial Proteomics I (P066–P079) ...........................................................................................................................................129
Post-translational Modifications I (P092–P107) ..........................................................................................................................................137
Proteomics for Human Health II (P146–P169) ............................................................................................................................................146
Poster Session C • 24 March 2015, 15:00–16:30 ................................................................................................................. 164
Plant and Microbial Proteomics II (P080–P091) ..........................................................................................................................................164
Post-Translational Modifcations II (P108–P122) .........................................................................................................................................172
Proteomics for Human Health III (P170–P190) ...........................................................................................................................................180
Quantitative Proteomics II (P216–P232) .....................................................................................................................................................195
Structural Proteomics (P233–P238) ............................................................................................................................................................205
Subcellular Proteomics (P239–P251) ..........................................................................................................................................................210
Proteomic Forum 2015
58
Poster Session A • 22 March 2015, 19:00–21:00
Bioinformatics (P014–P030)
P014
Robust peptide-based models in quantitative proteomics
1,2,3,4
2,3
4
1
L. Goeminne
, K. Gevaert , K. Vandepoele , L. Clement
1
Ghent University, Department of Applied Mathematics, Computer Science and Statistics, Ghent, Belgium
2
VIB, Ghent University, Department of Medical Protein Research, Ghent, Belgium
3
Ghent University, Department of Biochemistry, Ghent, Belgium
4
VIB, Ghent University, Department of Plant Systems Biology, Ghent, Belgium
Tandem mass spectrometry coupled to liquid chromatography (LC-MS/MS) allows for large-scale protein identification and
differential quantification in complex proteomes. However, data analysis remains a challenging task. Certain peptides are more
easily generated by digestion than others, basic and apolar peptides tend to be more easily ionized with electrospray, peptides
compete for ionization and co-elute, and elution is continuous while MS sampling is a discrete process. This leads to highly
variable intensities with many outliers and a vast amount of missing values when assessing data from multiple runs; i.e. only a
limited number of peptides are identified in all samples.
Popular strategies analyze the data protein-by-protein by (a) discarding peptides from the analysis when they are unidentified in
one of the samples or (b) summarizing the intensities over the available peptides for each protein in a sample thus ignoring
missingness. Discarding peptides leads to a substantial loss of valuable information. Strategy (b) however, can introduce a
considerable bias and is suboptimal as the summarized protein intensities are often based on different peptides as well as
different numbers of peptides. Summarization procedures also do not correct for differences in peptide characteristics. Finally,
many existing methods are sensitive to outliers, especially when proteins are detected by few peptides.
Here, we present a method in which normalized log 2 intensities are modeled directly on the peptide level, avoiding the need for
a separate summarization step. It accounts for differences in peptide characteristics, within and between sample variances as
well as within sample correlation between peptides of the same protein. Priors are put on model parameters, which allow for
borrowing information across peptides, impose shrinkage on differential expression estimators and provide more stable
variance estimators. Finally, M-estimation makes the method robust against outliers. Our method performs on par with
conventional methods for proteins identified by many peptides and it outcompetes them for proteins with few identified
peptides as it still detects strong effects while being robust against outliers.
P015
Analysis of Labeled and Non-Labeled DIA and DDA Proteomic Data using Progenesis QI for Proteomics
1
1
1
1
2
3
3
1
L. A. Gethings , G. Atwal , M. Palmer , M. Kipping , J. Pope , A. Collins , A. Jones , J. I. Langridge , J. P. Vissers
1
Waters Corporation, Health Sciences, Wilmslow, United Kingdom
2
Nonlinear Dynamics, Newcastle upon Tyne, United Kingdom
3
University of Liverpool, Department Functional and Comparative Genomics, Liverpool, United Kingdom
1
LC-MS is routinely applied for the qualitative and quantitative analysis of complex proteomes to characterize biological
processes and understand disease states. However,experiments can readily generate large and complex data sets with the
analysis and interpretation of the results readily evolving into the rate determining steps. In turn, this has led to a demand for
improved data analysis systems, including efficient and accurate data compression routines, intuitive software interfaces with
menu-guided workflows, flexible experimental designs without sample number restrictions, consistent peak detection for
improved accuracy and precision, complete data matrices without missing values for reliable statistics, and the ability to analyze
fractionated samples. These features are illustrated for novel informatics for the quantification and identification of example
isotopically-labeled and label-free proteomics datasets.
E
A number of experimental designs and cases were investigated, including ion mobility assisted DIA HDMS experiments for the
label-free quantitative analysis of samples with known differential protein expression values to assess quantification accuracy,
SILAC and dimethyl labeled cases to assess peak pair detection efficiency and quantitative reproducibility across multiple
technical replicates, and application examples to demonstrate identified biological relevance. Data from DDA based label-free
studies were analyzed in detail to investigate peak matching efficiency across a complete experiment, indicating that co-
Proteomic Forum 2015
59
detection affords near 100% feature intensity value measurement within individual runs. The identification of significantly
regulated peptides and proteins and the classification of samples are demonstrated through the use of embedded multivariate
analysis tools, alongside analysis of variance scoring, which include unsupervised principal component analysis and hierarchical
clustering techniques. The software also enables easy export of differentially regulated features to pathway analysis tools and
flexible reporting options to produce publication-ready reports.
P016
PRIDE Cluster: How to take advantage of spectrum clustering in a heterogeneous repository of MS proteomics data
1
1,2
1
1
1
1
1
R. Wang , J. Griss , S. Lewis , J. A. Dianes , N. D. Toro , H. Hermjakob , J. A. Vizcaino
1
European Bioinformatics Institute (EMBL-EBI), Proteomics Services, Cambridge, United Kingdom
2
Medical University of Vienna, Department of Dermatology, Vienna, Austria
Introduction: The PRIDE database (http://www.ebi.ac.uk/pride) at the European Bioinformatics Institute, part of the
international ProteomeXchange consortium, is one of the main public repositories for mass spectrometry (MS)-based
proteomics data. As PRIDE is an archival database, it represents the data as submitted, resulting in highly heterogeneous data
content. Identifying reliable peptide identifications across the database can be challenging. We therefore developed the ‘PRIDECluster’ algorithm (Griss et al., 2013, Nat Methods 10(2):95-96) which was used to cluster all identified spectra in PRIDE. Based
on these results we were able to separate reliable from unreliable identifications.
Objectives: Because of the large increase of data stored in PRIDE (due to the success of ProteomeXchange), the original
algorithm could no longer be run in a feasible time frame. Therefore, we needed to develop an updated version of the ‘PRIDECluster’ algorithm that would be more efficient and able to continuously cluster all data submitted to PRIDE without impairing
clustering quality.
Material & Methods: The new ‘PRIDE-Cluster-H’ algorithm has been designed to run in the Hadoop open-source software
framework (http://hadoop.apache.org/). Hadoop is developed for the distributed storage and processing of ‘Big data’ by
implementing the ‘Map reduce’ algorithm. The same three test datasets were used to assess the quality of the clustering
algorithm as previously shown in the original publication.
Results: The new ‘PRIDE-Cluster-H’ algorithm is able to cluster all identified spectra in PRIDE in a fraction of the time used by the
original version. The original PRIDE Cluster application took 14 days (601 CPU days) on a compute farm to cluster 20 million
spectra, and could not scale with the huge increase of data in PRIDE. The updated Hadoop-based implementation currently takes
2 days (760 CPU days) to cluster 55 million spectra. In addition, the new algorithm’s clustering quality is at least as good as the
original one. The new ‘PRIDE Cluster’ web (http://www.ebi.ac.uk/pride/cluster/) provides a way to access the clustering results.
All available spectral libraries have been updated based on the new results.
Conclusion: Spectral clustering is an approach that can be used to efficiently assess the quality of the identifications in public
proteomics repositories. A framework like Hadoop is needed to handle the huge data volumes present in PRIDE. The updated
Hadoop-based implementation opens new opportunities for quality control in PRIDE through frequent-re-clustering, and rapid
assessment of newly submitted large-scale datasets.
P017
Selection of SRM-compatible peptides to detect single amino acid variations
1
1
1,2
2
O. Kiseleva , A. Lisitsa , G. Krasnov , A. Kudryavtseva
1
Orekhovich Institute of Biomedical Chemistry, Bioinformatics, Moscow, Russian Federation
2
Engelhardt Institute of Molecular Biology, Moscow, Russian Federation
High-throughput sequencing technologies generate huge amounts of information about genetic variations. Interpretation of
RNAseq data has become a bottleneck for impactful use of genomic data. The major problem is to identify variations (single
amino acid changes, alternative splicing and other molecular events) responsible for diseases. Bottom-up shotgun massspectrometry was a basic procedure of studying single amino acid polymorphisms (SAPs), until a novel MS-technique for
targeted protein quantification - the selected reaction monitoring (SRM) - has emerged. Thus, the new step is to match SRM
spectra against mutations derived from RNAseq data of cancer-associated single nucleotide polymorphisms.
Development of SRM method for detecting SAPs requires a number of clinically-relevant matches against the cancer cell line and
normal tissue genomes. The bioinformatic pipeline for SAPs research was developed as follows:
Proteomic Forum 2015
60
(i) Genomes of normal liver tissue sample and hepatocellular cancer cell line (HepG2) were comprehensively characterized using
Illumina GA IIX.
(ii) Variations in nucleotide sequence in comparison with the reference genome were selected (169 thousand and 68 thousand
variants for HepG2 and liver respectively).
(iii) Non-synonymous single nucleotide polymorphisms with high quality of mapping (21 thousand and 8 thousand for HepG2 and
liver respectively) were translated in silico into protein sequence library (14 thousand and 5 thousand SAPs with high reading
quality for HepG2 and liver respectively).
(iv) Variant protein sequences were virtually digested by trypsin. Theoretical peptides were mapped onto one of human
chromosomes (18) to obtain the pool of unique tryptic peptides with a length appropriate for MS (6-25 aa).
As result 39 and 17 proteotypic peptides for HepG2 and liver respectively were selected out of the initial pool of 574 SAVs for
chr.18. We conclude that less than 7% of predicted tryptic SAP-specific proteotypic peptides are theoretically available for
targeted MS analysis.
P018
Estimating the Limits of Detection in Metaproteomics
1
1
M. Kuhring , B. Y. Renard
1
Robert Koch Institute, Research Group Bioinformatics (NG 4), Berlin, Germany
Introduction: In recent years, there has been an increasing interest in using mass spectrometry for metaproteomic studies. In
contrast to a single organism sample, the composition of microbial communities is more complex and heterogeneous.
Additionally, present organisms are often only inaccurately covered or even entirely missing in available reference databases. As
a consequence, the identification of spectra from metaproteomic experiments with protein database searches is not completely
achievable, leaving an unknown number of organisms undetected. Due to constant evolution and the large number of microbial
organisms, even ongoing sequencing efforts will not overcome this challenge.
Objectives: Some research has been carried out on improving the identification of MS/MS spectra in general by the use of error
tolerant hybrid approaches integrating de novo sequencing and database searches or by metaproteogenomic approaches.
However, no studies have been found which investigate the influence of error tolerant searches on the ratio of spectra identified
within different taxonomic levels and unidentified spectra. In this contribution, we estimate those detection limits of error
tolerant database searches. In addition, we investigate to what extent metagenomic supported spectra identification affects the
amount of unidentified spectra.
Methods: To estimate the quantity and ratio of the detection levels of a protein database or phylogenetic subgroup we designed
a study motivated by a cross validation approach. One by one, a single organism or species branch is removed from a protein
database. The extracted proteins are then in silico digested and searched in the altered database with a specific number of
errors allowed. Additionally, the database is enhanced by genome sequencing data to allow for even higher error-tolerance.
Finally, the given origin of the artificial spectra allows the classification of the search results.
Results: We applied the approach on databases composed of NCBI RefSeq bacteria proteins of selected phyla. We provide a
cross section view of detection levels by extracting each genus once. Additionally, we highlight the variance of detection levels
by focusing on particular organisms with varying popularity. To validate the estimations of protein database detection limits we
make use of real, microbial MS/MS experiments as ground truth data. The results show favorable similarities between
estimations and real spectra identifications, in particular for the ratios of identified and unidentified spectra.
Conclusions: While providing a comprehensive overview across organisms, we also provide a pipeline for experimental design,
allowing researchers to decide in the planning stage of an experiment which benefit to expect from various strategies, e.g.
whether a parallel metagenomic analysis is recommendable.
Proteomic Forum 2015
61
P019
Using public proteomics data to find the blind spots in mass spectrometry
1,2
K. Verheggen
1
Ghent University, Medical Protein Research, Ghent, Belgium
2
VIB, Ghent, Belgium
In proteomics, the scenario of mass spectrometry (MS) failing to detect certain peptides still raises questions. Firstly, there is the
influence of actual sequence length and composition. Secondly, it is plausible that some protein products are not abundant
enough, escaping detection. Degradation may even precede detection by MS. This study aims to investigate these hypotheses by
1
sequence database searches and large-scale reprocessing of the PRIDE database.
2
3
4
An in-house pipeline (ReSpin) incorporating pride-asap , SearchGUI and PeptideShaker was used to automatically process
PRIDE assays. To speed up processing, an in-house GRID engine (Pladipus) was used to manage all sequence database searches.
5, 6
6
7
Identified proteins were mapped to abundance and half-life (murine only). In addition, Instability Index values (II) were
calculated for all human and murine proteins.
The composition of identified peptides strongly resembles the distribution of the in sillico peptides, save for the digestion
-1
enzyme related N-termini. Protein abundance can cause detection issues, yet the lowest abundance was found at 0.68 ng*ml .
Bubble plots for protein abundance are shown in figure 1. The size of the bubble reflects the amount of assays the protein was
identified in. Degradation of proteins can potentially be of influence. Considering a neutral stability index value between 40 and
50; both stable and instable proteins are identified by MS, presented in figure 2.
Reflecting on the results in this study, large scale reprocessing of public proteomics data promotes rehash of historical topics and
consequentially brings forth novel insights.
References:
1. Vizcaíno, J. A. et al. The PRoteomics IDEntifications (PRIDE) database and associated tools: status in 2013. Nucleic Acids Res.
41, D1063-9 (2013).
2. Hulstaert, N. et al. Pride-asap: Automatic fragment ion annotation of identified PRIDE spectra. J. Proteomics 95, 89-92 (2013).
3. Vaudel, M., Barsnes, H., Berven, F. S., Sickmann, A. & Martens, L. SearchGUI: An open-source graphical user interface for
simultaneous OMSSA and X!Tandem searches. Proteomics 11, 996-999 (2011).
4. Barsnes, H. et al. compomics-utilities: an open-source Java library for computational proteomics. BMC Bioinformatics 12, 70
(2011).
5. Anderson, L. & Hunter, C. L. Quantitative mass spectrometric multiple reaction monitoring assays for major plasma proteins.
Mol. Cell. Proteomics 5, 573-88 (2006).
6. Schwanhäusser, B. et al. Global quantification of mammalian gene expression control. Nature 473, 337-42 (2011).
7. Guruprasad, K., Reddy, B. V & Pandit, M. W. Correlation between stability of a protein and its dipeptide composition: a novel
approach for predicting in vivo stability of a protein from its primary sequence. Protein Eng. 4, 155-61 (1990).
Figure 1
Proteomic Forum 2015
62
Figure 2
P022
Peakjuggler - A Proteome Discoverer Node for label-free MS1 Quantification
1
1
J. Doblmann , F. Dusberger , K. Mechtler
1
IMP, Protein Chemistry, Vienna, Austria
1
Introduction: Label-free quantification is becoming increasingly popular in proteomics. The availability of high resolution mass
spectrometers allows quantification of peptides on MS1 level at high selectivity. The method offers high dynamic range for
shotgun proteomics samples without requiring additional sample processing steps or assay development. Although the
comparison of large sample numbers is possible missing values still constitute a major problem. Therefore we describe a novel
algorithm for the quantification of peptides in shotgun proteomic experiments.
Objectives: The new software package should allow the accurate and reproducible quantification of peptides from shotgun
proteomics samples. As Proteome Discoverer is widely used the software should be available as a Node within the Proteome
Discoverer Software suite. Another goal is to perform match-between-runs, in order to transfer identification onto non
identified features.
Materials & Methods: Peakjuggler is developed in C# (requires .Net Framework 4.5) and utilizes the open-source math library
MathNet.Numerics (version 2011.4.17.0) for sophisticated calculations. For manual validation of extracted peak areas Thermo
Qual Browser version 2.0 was used. Data analysis and visualization is performed in R (version 3.1.1) using the ggplot2 and shiny
packages.
Results: The software was developed to be compatible with the current version of Proteome Discoverer (version 1.4) as well as
the upcoming version 2.0. Our novel area calculation algorithm can quantify 8% more identified peptides than the inbuilt
Precursor Ions Area Detector node. Furthermore, these areas correspond well to manually extracted monoisotopic peak areas
with Thermo XCalibur Qual Browser. The ratios between different samples are also presented in a seperate table.
Conclusion: We present a new, easy-to-use software for label-free quantification of peptides integrated into the widely used
Proteome Discoverer platform. It calculates more areas than the inbuilt software and additionally performs matching between
runs.
Proteomic Forum 2015
63
P023
GelMap: annotation and evaluation of gel-based proteomic data
1
1
1
J. Klodmann , M. Senkler , H.- P. Braun
1
Leibniz Universität Hannover, Pflanzenproteomik, Hannover, Germany
When the GelMap project was initiated in 2010, it was originally established for easy and quick annotations of two dimensional
gels, which actually was an important advantage in comparison to other available annotation tools. In the following years, the
software has developed into a multifunctional tool, providing a broad variety of options to annotate and analyze gel-based
proteomic datasets [1-3].
The integration of GelMap into meta protein platforms, like the GATOR [4], further enhances its value for systematic data
mining.
A major purpose of GelMap is the data exchange between researchers. Therefore, it is freely accessible via the internet
(www.gelmap.de). However, if data sets should be kept confidential, projects can be protected by a password.
Here we present the most recent proteomic projects in GelMap as well as the latest updates of the GelMap software which
enhance the browsing capabilities in established reference maps. Furthermore, new features are available which allow users to
transfer identification data from published GelMap projects to newly prepared gel images. This allows rapid protein
identification without the necessity to perform MS.
GelMap is a flexible tool for data annotation and evaluation. As it has only been used for plant projects so far, linked databases
and tools are from the plant field. The integration of further external sources is always possible. The structure of the software is
easily expendable and allows it to be used for a multitude of future purposes.
[1] Rode C, Senkler M, Klodmann J, Winkelmann T, Braun HP. (2011) GelMap-a novel software tool for building and presenting
proteome reference maps. J Proteomics 74: 2214-9.
[2] Klodmann J, Senkler M, Rode C, Braun HP. (2011) Defining the protein complex proteome of plant mitochondria. Plant
Physiol. 157: 587-98.
[3] Senkler M, Braun HP. Functional Annotation of 2D Protein Maps: The GelMap Portal. (2012) Front Plant Sci. 3: 87.
[4] Joshi, H.J., Hirsch-Hoffmann, M., Baerenfaller, K., Gruissem, W., Baginsky, S., Schmidt, R., Schulze, W.X., Sun, Q., van Wijk, K.,
Egelhofer, V., Viennakoop, S., Weckwerth, W., Bruley, C., Rolland, R., Toyoda, T., Nakagam, H., Jones, A., Briggs, S.P., Castleden,
I., Tanz, S., Millar, A.H., and Heazlewood, J.L. (2011) MASCP Gator: An aggregation portal for the visualization of Arabidopsis
proteomics data. Plant Physiol. 155, 259-270.
P024
Combining SWATH/DIA proteomics data with large scale RNA expression data enables cross-omics quantitative biology
1
1
1
1
2
3
3
O. Rinner , C. Escher , R. Bruderer , O. Bernhardt , S. Messner , S. Bleuler , P. Zimmermann , L. Reiter
1
Biognosys AG, Schlieren, Switzerland
2
Insphero AG, Schlieren, Switzerland
3
Nebion AG, Zurich, Switzerland
1
Question: Recently the concept of proteogenomics, the combination of proteomics and genomics, or more specifically protein
and RNA expression data, has gained increased attention. It has long been appreciated that both omics methods provide
complementary information which, combined, could provide a more comprehensive picture of cellular dynamics. The advances
in next-gen sequencing on one side and the recent availability of large scale proteomics datasets on the other has made such a
cross-omics approach conceivable. So far, however, these two fields are still mostly separated. We believe that incompatibility of
basic data structure and analytical peculiarities on the proteomics side, such as the focus on peptides vs. protein quantities or
missing data in large proteomics datasets, have been the main obstacles for researchers outside the proteomics community.
Methods: Here, we show that quantitative protein expression data generated from SWATH/DIA acquisition provide almost
complete data matrices with a structure very similar to RNA-Seq data. Therefore, they are very amenable to cross-omics
analysis. We developed methods for data normalization and condensation of peptide ion-currents into protein quantities, which
are implemented in the Spectronaut software. We were able to seamlessly integrate large proteomics datasets into
Genevestigator , the most widely used curated transcriptomics resource with more than 125,000 datasets from microarray and
Proteomic Forum 2015
64
RNA-Seq experiments. Furthermore, we introduce the concept of a proteomics platform akin to the microarray platforms, which
is based on comprehensive spectral libraries and allows comparisons across datasets of different origin.
Results and Conclusion: We provide 2 publicly available datasets via Genevestigator: A toxicological study with liver spheroid
tissue applying different doses of acetaminophen (Paracetamol) and a large reference dataset of HEK 293 cells. All tools in
Genevestigator, although originally conceived for the analysis of RNA abundance data, can be directly used with proteomics
data. We show examples for highly correlated protein/RNA expression. We furthermore show that the dynamic range of
expression in the proteomics platform is higher compared to RNA expression. By contrast, the range of regulation is dampened
in proteomics data compared to corresponding RNA expression differences. The analysis of normalized SWATH/DIA proteomics
data can easily be transfered from Spectronaut to Genevestigator and allows direct comparison with curated public proteomic
or transcriptomic experiments. The availability of large curated expression compendia from a wide range of experimental
conditions facilitates results interpretation and verification of biological hypotheses.
P025
MetaProteomeAnalyzer - a software suite for the functional and taxonomic characterization of (meta)proteome data
1
2
2
2
2
1,2
3,4
1
T. Muth , A. Behne , R. Heyer , F. Kohrs , M. Hoffmann , U. Reichl , L. Martens , E. Rapp
1
Max Planck Institute for Dynamics of Complex Technical Systems, Bioprocess Engineering, Magdeburg, Germany
2
Otto von Guericke University, Magdeburg, Germany
3
Department of Medical Protein Research VIB, Ghent, Belgium
4
Ghent University, Department of Biochemistry, Ghent, Belgium
In nature, microorganisms typically live in complex communities, competing or cooperating with each other. Hence, researchers
investigate biological processes of bacteria in the human gut or anaerobic digestion in biogas plants on a mixed culture level [1].
Proteomics of microbial communities (so-called “metaproteomics”) constitutes one approach to monitor changes in the protein
composition of samples, for example, to characterize their metabolic complexity, to investigate the impact of environmental
parameters on strain abundance, or to identify markers for disease. In addition to the high complexity and heterogeneity of the
samples, the most severe challenges of metaproteomics research reside in the analysis and interpretation of the data. This is
mainly due to the lack of protein sequence information and the correct assignment of shared peptides between different
proteins and species [2]. To overcome limits of existing solutions, we developed a dedicated software suite for the functional
and taxonomic characterization of metaproteome data called “MetaProteomeAnalyzer” (MPA). Our software significantly
facilitates in-depth metaproteomics data analysis and interpretation. The pipeline is built around multiple search algorithms for
reliable peptide and protein identification. In addition, we included various features for the comprehensive analysis of the
taxonomic diversity in a microbial community, and for the functional assignment (KEGG pathways and E.C. classification). In
order to handle protein redundancy, the software includes a set of common useful rules for protein grouping. Finally, we provide
an innovative graph database system in the back end for a detailed and user-defined result analysis. Thus, the MPA constitutes a
protein analysis platform specialized on metaproteomics research, however, with all its included features (e.g. multiple search
engine support and functional assignment), it also represents a valuable software suite applicable to single/few species
proteomics samples, in particular with the full integration of an automated KEGG pathway analysis.
REFERENCES
[1] Heyer, R., Kohrs, F. , Benndorf, D., Rapp, E. , Kausmann, R., Heiermann, M., Klocke, M., Reichl, U. “Metaproteome analysis of
the microbial communities in agricultural biogas plants.”, N Biotechnol., 30(6):614-22 (2013).
[2] Muth, T., Benndorf, D., Reichl, U., Rapp, E., Martens, L. “Searching for a needle in a stack of needles: challenges in
metaproteomics data analysis.”, Mol Biosyst., 9(4):578-85 (2013).
Proteomic Forum 2015
65
Figure 1
Figure 2
P026
Informatics Support for Isomeric Separation and the Structural Identification of Labeled N-Glycans from Proteins
1
2
3
3
Ningombam Sanjib Meitei ; Arun Apte ; Udayanath Aich ; Julian Saba ; Madalina Oppermann
1PREMIER Biosoft, Indore, India
2PREMIER Biosoft, Palo Alto, United States
3Thermo Fisher Scientific, CA, United States
4Thermo Fisher Scientific, Stockholm, Sweden
4
The separation of glycans by chromatography prior to MS analysis can provide number of benefits. Primarily, separation can
reduce sample complexity, minimize ion suppression, increase dynamic range of analysis and provide separation of structural
isomers. Recent developments in mixed -mode column chemistries and faster scanning mass spectrometers have increased the
number of glycans resolved and identified by LC-MS workflows.However, the increase in peaks resolution and detected
compounds also leads to large data sets. Additionally, chromotograms of isomeric glycans are complex with some isomers co-
Proteomic Forum 2015
66
eluting under a single peak. Manual deconvolution of such complex chromatograms, identification of isotopic peaks
components, identifying MS/MS scans for detected compounds and selection of correct precursor m/z values from the isotope
cluster for MS/MS data analysis is time consuming task. Therefore, we have developed a software tool (SimGlycan) to streamline
this process.
Software modules were developed for automatic detection of compounds, deconvolution of chromatograms to separate glycan
isomers, identification of isotope clusters and MS/MS scans corresponding to detected compounds and precursor m/z selection.
2AB-labeled glycans from various proteins were separated and analyzed on a Thermo Scientific GlycanPac AXR-1 column coupled
to Orbitrap Fusion Tribrid mass spectrometer. LC-MS and MS/MS data was subjected to the program. The accuracy of the results
were also tested on Q Exactive mass spectrometer.
All isomeric glycans correctly detected, separated and identified by the program were manually validated for these experiments.
P027
On the Use of Multiple Databases for DB Search in Mass Spectrometry-Based Proteomics
1
1
2
1
C. Has , M. D. Mungan , C. Ciftci , J. Allmer
1
Izmir Institute of Technology, Molecular Biology and Genetics, Urla Izmir, Turkey
2
Izmir Institute of Technology, Biotechnology, Urla Izmir, Turkey
Question: Research in Proteomics is currently driven by mass spectrometry (MS). Especially for the identification of proteins
from complex samples, MS is an indispensable tool. Since proteins don’t lend themselves well to MS analysis, they are usually
cleaved into shorter peptides which are then analyzed by MS and MS/MS. Computational analysis of the resulting data then
determines the peptide sequences of the recorded tandem-MS spectra and integrates identified peptides into proteins. To
establish a peptide sequence match (PSM) two general approaches are possible. De novo sequencing directly derives the
sequence from the provided MS/MS spectra while database search algorithms need a list of possible amino acid sequences that
can give rise to the MS spectra. Many database search algorithms have been proposed and consensus scoring using multiple
algorithms has been performed. To be able to compare among possible results false discovery rate and other statistical
measures have been discussed, but no consensus has been reached so far. While the comparison among results from different
algorithms is interesting, there has been no attempt to integrate the results from multiple databases for any of the given
algorithms. This is however important since it poses technical problems when all databases, needed for a study, are simply
concatenated and searched at once. We, and others, have shown previously, that databases of different size influence scoring
and prohibit the direct comparison of results even when using just one algorithm.
Methods: Here we analyzed four algorithms (OMSSA, X!Tandem, MS-GF+, and MassWizz) on databases of the same size and of
increasing difference in size as well as in respect to the sequence redundancy within the databases.
Results. According to our findings, truly non-redundant databases, of same size, may be searched and the resulting scores may
be directly compared without any further transformation. However, difference in size or difference in sequence redundancy
prohibits such an approach.
Conclusions: This is especially important for proteogenomics studies where the translation of the whole genome must be
searched but the individual chromosomes are of different size and redundancy and concatenated database usually lead to
runtime problems with existing algorithms.
P028
A search strategy for the identification of cross-linked peptides based on their CID behavior
1,2
2
1,2
S. Giese , L. Fischer , J. Rappsilber
1
Technische Universität Berlin, Department of Bioanalytics, Berlin, Germany
2
University of Edinburgh, Wellcome Trust Centre for Cell Biology, School of Biological Sciences, Edinburgh, Germany
Questions: Recent developments in structural biology employ cross-linking/mass spectrometry (CLMS) as a low-resolution
1-3
method to resolve protein-protein interactions or protein folds by help of distance constraints . To improve the analysis of
cross-link experiments we formulize the following questions:
(Q1) How can cross-link (CL) acquisitions be optimized?
Proteomic Forum 2015
67
(Q2) What are the mass spectrometric differences of linear (LN) and cross-linked peptides?
(Q3) How can we use knowledge about CL and LN peptides to build an efficient, yet sensitive identification algorithm?
Methods: We analyzed 157 database searches from our in-house search engine Xi. All data was acquired on LTQ Velos or QExactive with CID fragmentation. We performed in-depth statistical analysis and data mining of precursor and fragment
properties such as mass, charge and intensity to develop an integrated search strategy.
Results: We observe that cross-linked peptides are generally larger and higher charged than linear peptides. Our data suggests
that theoretically, removing up to 60% of linear peptides is possible by using charge and mass based filters (Q1). Moreover,
relative fragment mass and charge are very effective predictors to identify CL fragments of cross-linked peptides (Q2). Using the
fact that a CL peptide comprises two individual peptides can be exploited to largely reduce the search space. Once the first
peptide is identified the second peptide is as easy to identify as it is to identify a modified linear peptide. With additional
implementation tricks an integrative, well-adopted search strategy was developed (Q3).
Conclusion: CL-MS is an important field of on-going research. The above described results have the potential to improve the
quality and quantity of biological knowledge that can be drawn from CL experiments. This is feasible due to our improved and
4,5
integrated search strategy for cross-linked peptides. Not only alleviate our results the need for special chemistry , they also
simplify the application for quantitative studies. This was possible by carefully investigating the CID behavior of CL peptides.
1. Rappsilber, J. The beginning of a beautiful friendship: cross-linking/mass spectrometry and modelling of proteins and multiprotein complexes. J. Struct. Biol. 173, 530-40 (2011).
2. Chen, Z. A. et al. Architecture of the RNA polymerase II-TFIIF complex revealed by cross-linking and mass spectrometry. EMBO
J. 29, 717-26 (2010).
3. Leitner, A. et al. Probing native protein structures by chemical cross-linking, mass spectrometry, and bioinformatics. Mol. Cell.
Proteomics 9, 1634-49 (2010).
4. Götze, M. et al. StavroX--a software for analyzing crosslinked products in protein interaction studies. J. Am. Soc. Mass
Spectrom. 23, 76-87 (2012).
5. Rinner, O. et al. Identification of cross-linked peptides from large sequence databases. Nat. Methods 5, 315-8 (2008).
P029
xiFDR - False Discovery Rate Estimation in Cross-Linking/Mass Spectrometry
1
1,2
L. Fischer , J. Rappsilber
1
University of Edinburgh, Wellcome Trust Centre for Cell Biology, Edinburgh, United Kingdom
2
Technische Universität Berlin, Department of Bioanalytics, Institute of Biotechnology, Berlin, United Kingdom
Introduction: There is a number of work-flows to match cross-linking/mass spectrometry (CLMS) - spectra against a search
database. But assigning confidence to your matches is not straight forward. We provide a search-engine independent tool, that
can be used to filter CLMS results to a given FDR.
Objectives: Provide an application, that can be used to estimate an FDR for cross-linked peptide spectrum matches (PSM). Not
only at the level of PSMs can a confidence value be assigned but also at the level of cross-linked peptides, unique linked residue
pairs (links), and unique pairs of proteins. In this way a confidence value can be assigned at the level of information that is
actually of interest to the researcher.
Beyond that it also can maximize the number of unique links or protein pairs that can be returned while maintaining the desired
confidence of the results.
Methods: xiFDR is a java application, that can be used to filter CLMS results to a given FDR at different levels of information PSMs, linked peptide pairs, links, and protein pairs. To do so, the PSMs are folded up to higher levels of information. For this the
scores of PSMs that support e.g. a unique linked residue pair are combined into a score for that link. This is done in a stepwise
manner, meaning that we first do an FDR at the level of PSMs and then turn the resulting PSMs into unique peptide pairs. Next,
the peptide pairs are filtered to a specified FDR and folded up into linked residue pairs (links). And again the same for unique
protein pairs. Each level can be filtered to a different FDR. Finally the results can be written out as a set of CSV-files.
xiFDR accepts a CSV-file as input with a minimum set of information. The columns can be manually mapped to the required
information - thus making it search-engine agnostic.
Proteomic Forum 2015
68
Alternatively xiFDR also accepts an mzIdentML-file as input. In this case the results can also be exported as a new mzIdentMLfile.
Results: We validated xiFDR with a large-scale HSA analysis. This enabled us to compare the results with the crystal structure of
HSA and shows that estimated FDR fits in terms of expected over-length cross-links according to the crystal structure.
Also using an FDR at the lower levels of information (PSMs and peptide pairs) enables us to boost the number of unique links
that can be reported at a given link-level FDR.
Conclusion: With xiFDR we provide a tool, that can be used to filter PSMs at the level of actual interest and maximize the results
for a given FDR. It can be used with any search - engine that provides the results either as a CSV-file with the required
information or as a mzIdentML file according to the upcoming version 1.2.
P030
Complexation of proteins with cationic porphyrins and significance of fatty acids
1
A. Gyulkhandanyan
1
Institute of Biochemistry of NAS of Armenia, Yerevan, Armenia
In photodynamic therapy of tumor (PDT) binding of photosensitizers (PS) (generally porphyrins) and their delivery to tumor is
one of the important problems of this method. Delivery of PS realized through the blood and such proteins as serum albumen
and hemoglobin play a key role in this process. At PDT cytochrome c provides a crucial role in apoptosis and its possible binding
to porphyrins is also important in the study of this problem. The presence of ligands (including fatty acids) greatly complicates
the solution of tasks in relation to a possible competition between these two ligands (porphyrins and fatty acids) for binding
sites on the proteins-carriers. Earlier in Armenia were developed methods and synthesized a series of new cationic porphyrins
with various peripheral groups. In the interaction of an array of porphyrins and several possible carrier proteins these tasks are
further divided into many tasks on optimization of binding conditions of each pair of protein-porphyrin. One of the new effective
methods to solve such tasks of multidimensional biology is the method of small molecule microarrays. Another method absorption spectroscopy with high accuracy allow to determine the binding constants of ligands (porphyrins) with proteins. Via
these two independent experimental methods we have determined the most promising porphyrins for use in photodynamic
therapy and the effect of fatty acids on the complexation protein-porphyrin.
Investigations by computer simulation method (molecular docking) substantially complement experimental methods, allow to
predict the possibility of protein-ligand interactions and to conduct screening of ligand binding sites. By the method of molecular
docking is shown that the main site for the binding of ligands (porphyrins, fatty acids and their complexes) is an internal cavity of
the macromolecule Hb (Fig. 1, porphyrin is presented in blue color), and that complexes [porphyrin-fatty acid] can displace free
porphyrins from the internal cavity of the macromolecule of Hb. Molecular docking confirmed results that in the molecule of
serum albumin porphyrins and fatty acids compete for a common binding site FA1, and that the meso-substituted functional
groups of porphyrins play an important role in modulating of conformational rearrangements of protein. Molecular docking
method is a convenient and available tool for study of ligand binding sites on the protein macromolecule and can significantly
supplement the data of the experimental methods about structural features of proteins.
Figure 1
Proteomic Forum 2015
69
P031
Speaking to spectra: what we can learn without using search engines
1,2
L. Martens
1
VIB, Department of Medical Protein Research, Ghent, Belgium
2
Ghent University, Department of Biochemistry, Ghent, Belgium
Question: Proteomics data analysis usually hinges on the identification of the peptides, and from there the proteins, in a sample.
Yet in many cases, identification is decidedly not trivial. Notable examples include unsequenced organisms as well as spectra
from peptides carrying unexpected or complex modifications. The interpretation of the data becomes the bottleneck in these
cases, because the mass spectrometer is perfectly capable of acquiring meaningful spectra of such analytes, yet the
bioinformatics tools to make sense of those spectra are lacking. We here present two tools that seek to address these challenges
in well-defined use cases.
Methods: Here, two tools are presented that allow unidentified fragmentation mass spectra to be used to perform differential
proteomics analyses, and to find signatures of complex modifications in large data sets. These tools take as input only the
acquired MS2 spectra from one or more runs, enabling these raw data to be interpreted within a specific context without prior
identification. The tools are built in the Java programming language, and are open source under the permissive Apache2
software license.
Results: The first tool is a spectral comparison pipeline that was applied to the specific case of finding unique signature peptides
that differentiate different Taenia species, which are important parasites in the developing world. This tool is sufficiently
sensitive to generate useful candidate differentiating spectra, while remaining very specific. The second tool, dubbed Spectrawl,
is based on an analysis of the pair-wise peak distances across MS2 spectra. Signature peak m/z as well as peak m/z deltas can be
combined into useful signatures that can detect spectra that contain certain characteristic neutral losses or immonium ions,
allowing complex modifications to be detected without identification.
Conclusions: Working with hard to identify spectra poses a steep yet often-encountered challenge in proteomics. When focusing
on specific questions however, the identification process can be bypassed and meaningful results can be obtained directly from
the fragmentation mass spectra. It should be noted that these approaches still carry limitations compared to similar analyses on
readily identifiable spectra, but the methods shown here do demonstrate that even without identification, mass spectrometry
based proteomics data can be put to good use.
Free topics (P032–P049)
P032
Computational and mass spectrometry-based workflow for the discovery and validation of human missing proteins:
application to chromosomes 2 and 14
1
2
3
1
2
4
4
4
Y. Vandenbrouck , C. Carapito , L. Lane , M. Benama , A. Opsomer , E. Mouton-Barbosa , L. Garrigues , A. Gonzalez de Peredo ,
2
1
3
4
1
4
2
1
A. Burel , C. Bruley , A. Gateau , D. Bouyssie , M. Jaquinod , O. Burlet-Schiltz , A. Van Dorsselaer , J. Garin
1
CEA, EDyP, Grenoble, France
2
CNRS UMR7178, Strasbourg, France
3
Swiss Institute of Bioinformatics , Geneva, France
4
CNRS UMR5089, Toulouse, France
In the framework of the C-HPP, our Franco-Swiss consortium has adopted chromosomes 14 and 2, coding for a total of 382
missing proteins for which evidence is currently lacking at protein level. Over the last four years, the French proteomics
infrastructure has collected high quality datasets from 40 human samples, including a series of rarely studied cell lines, tissue
types and sample preparations. Here, we described the use of bioinformatics screening and subsequent mass spectrometry
(MS)-based validation to identify what were up to now missing proteins in these datasets. Screening database search results
(85,288.dat files) returned identifications of 85 missing proteins by 126 unique peptides. These peptides were evaluated by
applying two different criteria: peptide-level False Discovery Rate (FDR) calculation and expert MS/MS spectral quality
assessment. Synthetic peptides were then produced for 76 proteins (103 peptides) and used to generate reference MS/MS
spectra. Spectral similarity score was calculated for each pair of reference-endogenous spectra. Finally, LC-SRM assays were
developed to target proteotypic peptides from 4 missing proteins detected in tissue/cell samples which were still available and
for which sample preparation could be reproduced. These LC-SRM assays unambiguously detected the endogenous unique
peptide for 3 proteins. For 2 of these, identifications were confirmed by additional proteotypic peptides. Thus, a rigorous step-
Proteomic Forum 2015
70
by-step approach combining bioinformatics screening and MS-based validation assays is particularly suitable for confirmation of
missing proteins. All MS/MS data have been deposited in ProteomeXchange under identifiers PXD000889 and PXD001178.
P033
Odorant Binding Proteins in the field mouse (Apodemus sylvaticus)
1
2
1
2
2
2
Z. Zdrahal , L. Havrdova , O. Sedo , P. Klempt , D. Vinkler , R. Stopkova , P. Stopka
1
Masaryk University, CEITEC-MU, Brno, Czech Republic
2
Charles University, Department of Zoology, Prague, Czech Republic
2
Introduction: The X-linked Odorant Binding Proteins (OBP) are globular lipocalins with typical beta barrel structure and CxxxC
motif. The Hamster Aphrodisin is an ortholog of the mouse OBPs and serves signal excretion by transporting volatile ligands
(pheromones) in their beta barrel with vaginal fluid. However, mouse OBPs are highly abundant in nasal mucosa and saliva. In
nasal mucosa they are believed to transport volatiles to the vicinity of chemosensory receptors, whilst in saliva they may have
potential to transport volatiles to the fur during self-grooming and thus contribute to amplify specific body odor signature.
Objectives: We have studied a species of murine rodent (Apodemus sylvaticus) that is known for extensive selfgrooming and
allogrooming behaviour in order to identify lipocalins that are abundant in saliva and may have capacity to transport, and
release pheromones when deposited in the fur.
Materials & Methods: Salivary samples were collected repeatedly with 50ul distilled water using pipets. The obtained samples
were subjected to transcriptomic and proteomic analysis. For transcriptome analysis 454 (FLX) method of sequencing of orofacial tissue mRNA[U1] was used to generate cDNA library for A. sylvaticus. PCR amplification, 5’race and 3’race and Sanger
methods were applied to obtain longer sequences to provide a pool of transcripts that were serviceable for further MALDIMS/MS structure elucidation. For proteomic analysis, proteins were precipitated by acetone and separated by 2D gel
electrophoresis. Corresponding 2D gel spot were digested by trypsin and analysed by MALDI -MS/MS.The identification was
based on transcript derived sequence database or manual interpretation.
Results: We have identified a population of highly expressed OBPs and Major Urinary Proteins (MUP) in saliva of wild living field
mice (A. sylvaticus). The OBP members included 11 positively identified protein spots mapped on 3 unique mouse Obp
transcripts and 4 MS-identical MUP spots for which the RNA seq approach revealed just a single transcript.
Conclusions: In comparison with the house mouse, the field mouse salivary proteome is overpopulated with Odorant Binding
Proteins whilst the production of Major Urinary Proteins is marginal. The fact that 11 OBP spots was successfully mapped on just
3 transcripts and 4 MUPs was mapped on a single transcript suggests that gene duplication and/or protein modification events
took place in the evolution of these chemical signals in field mice (Apodemus sylvaticus).
The study was supported by Czech Science foundation project (P206-12-G151), by the European Regional Development Fund
(CEITEC (CZ.1.05/1.1.00/02.0068) and “BIOCEV (CZ.1.05/1.1.00/02.0109)) and by the European Social Fund
(CZ.1.07/2.3.00/20.0189).
P034
How and when, came into existence, the first stem cells on Earth?
1
R. Amirmardfar
1
Sabz Andishan, Tabriz, Iran, Islamic Republic of
If a bacterial cell divided more than hundred times, the similar bacteria arise finally.If a Euglena cell divided more than hundred
times, the similar Euglena arise finally.
In a cells colony, the cell divides and produces the similar cells in colony.
Bacteria, unicellular and colonies are the first cells on earth. These cells can only, create the similar cells.
But a Sponge’s zygote (stem cell) can produce three types of cells during cell division (Amoeboid cell, Flattened surface cell and
Collar cell).
This article answers the following questions:
Proteomic Forum 2015
How a cell has found this ability for first time that produces the various cells?
At what times these cells appeared on earth?
Are stem cells specific, multicellularity and coloniality?
Are stem cells found only in the multicellularity and coloniality?
Are stem cells found in the unicellular and protozoa?
Figure 1
71
Proteomic Forum 2015
72
P035
Comparative Metabolomics Analysis of Prostate Cancer Cells with Different Ethnic Backgrounds
1
1
2
3
3
J. Nyalwidhe , T. Burch , J. Rhim , J. I. Langridge , A. Baker , G. Isaac
1
Eastern Virginia Medical School, Norfolk, United States
2
Center for Prostate Disease Research, Bethesda, United States
3
Waters Corporation, Health Sciences, Wilmslow, United Kingdom
3
African American (AA) men have disproportionately high incidence and mortality rates of prostate cancer (PCa) when compared
to Caucasian American (CA) men and other ethnic groups in the USA. Untargeted metabolite profiling may serve as surrogates
for disease stratification and potentially prognostic and diagnostic biomarkers. Metabolomics of PCa is currently being studied to
screen for biomarkers with high sensitivity and specificity. However, to date no comparative metabolomic analyses of disease
stratified African and Caucasian Americans matched prostate clinical samples and cell lines has been done. Here, we provide
comprehensive metabolomics profiling data from CA and AA cell lines that include a novel set of African American cell lines.
Prostate cancer cells with different ethnic backgrounds and tumorigenic phenotypes were homogenized and the lysates
extracted using Bligh and Dyer method. The upper phase containing water soluble metabolites was collected for polar
metabolite analysis and lower organic phase containing hydrophobic metabolites was collected for lipid analysis. The analysis
was performed using an ACQUITY i-Class system coupled to a SYNAPT G2-S MS. For the lipid separation an CSH C18 (2.1 x 100
mm, 1.7 µm) column was used with a flow rate of 400 µl/min. For the polar metabolites BEH amide (2.1 x 150 mm, 1.7 µm) was
used in both basic and acidic conditions.
Early detection and screening of PCa relies heavily on the measurement of prostate specific antigen (PSA) in serum. The utility of
PSA as a biomarker for prostate cancer has been challenged with respect to its sensitivity and specificity, and its inability to
distinguish indolent from aggressive forms of the disease. This inability to accurately predict the aggressiveness of PCa based
solely on standard clinicopathologic features clearly underscores the need to explore the ability of novel biomarkers to enhance
outcome prediction at biopsy and to understand the molecular basis of PCa metastasis. Therefore, additional biomarkers with
high sensitivity and specificity, and preferably obtained minimally invasively are urgently needed for PCa diagnosis and
prognosis. Data was collected using DIA data collection on a time of flight MS which allowed the characterization of lipids and
polar metabolites by precursor and product ion alignment. The LC/MS metabolomic profiles of the different PCa cell lines were
investigated using MVDA to determine the pattern and composition of the different metabolites classes. PCA)was performed
and three groups were separated, reflecting the pathological features of these cell lines. Differential analysis of results across
the different PCa cell lines can quickly be performed, thereby facilitating identification and quantitation of potential biomarkers.
Potential bio-markers that discriminate the different PCa cell lines were identified and quantified.
P036
Influence of column- and gradient-length on protein identification by nanoLC-MS/MS of whole cell shotgun proteome analysis
of marine bacteria
1
1
2
2
L. Wöhlbrand , R. Rabus , B. Blasius , C. Feenders
1
Institute for Chemistry and Biology of the Marine Environment (ICBM) University Oldenburg, General and Molecular
Microbiology, Oldenburg, Germany
2
Institute for Chemistry and Biology of the Marine Environment (ICBM), Mathematical Modelling, Oldenburg, Germany
Introduction: Shotgun proteomics by online-coupled nanoLC ESI-MS/MS is a commonly applied method to analyze bacterial
proteomes. While the mass spectrometer accounts for mass accuracy and determines time constraints for successive MS and
MS/MS cycles, the nanoLC setup and especially the applied gradient determine the number of peptides available for mass
spectrometric analysis per time interval. The latter directly affects the amount of detectable peptides and, hence, the number of
identifiable proteins. Shotgun proteomics is aimed at the detection of as many proteins as possible, which may be achieved by
application of long LC gradients (e.g. ≥480min). Such long gradient times, however, reduce the number of samples analyzed per
time.
Objective: Investigation of the influence of (i) analytical column-length coupled to (ii) the applied gradient-length on
peptide/protein identification by shotgun proteomics with two different marine bacteria, Phaeobacter inhibens DSM 17395 and
Desulfobacula toluolica Tol2.
Material & Methods: Analytical columns (C18, 2µm bead size, 75µm id) of 15, 25, and 50 cm length were used in combination
with linear gradients of 120, 240, 360, 480, and 600 min. The nanoLC was online-coupled to a 3D ion-trap mass spectrometer.
Proteomic Forum 2015
73
Three biological replicate samples were analyzed per bacterium and column/gradient setup yielding a total of 90 shotgun
analyses.
Results & Conclusion: For each column, the number of identified peptides/proteins increased with longer gradients (e.g. from
390 to 880 proteins; 25 cm column, P. inhibens) with a concomitant increase in identification score (mean Mascot score from
280 to 325). While identification improved only marginally with the 15 cm column when applying long gradients (≥ 480 min),
performance of the 50 cm column was impaired by too short gradients (≤ 240 min). Out of the identified proteins, 66% were
detected in all replicate analyses per column/gradient setup and >98% of these proteins were also detected with longer
gradients and/or columns. Hence, a certain core set of proteins is covered by all analysis types. Compilation of peptides from
replicate analyses (same column/gradient setup) increased the number of identified proteins by 21% as compared to the mean.
Furthermore, compilation of peptides originating from different gradient-lengths analyses often increased the number of
identified proteins, with similar total analysis time. E.g., compilation of triplicate analysis of P. inhibens with a cummulative time
of 1080 min (25 cm column, 3×360 min gradient) identified 805 proteins, while compilation of 120, 360, and 600 min gradients
(total time 1080 min) yielded 919 proteins (14% increase). Notably, the latter compilation identified even more proteins as
compared to the triplicate analysis of 3×480 min gradients (876 proteins; total time 1440 min), thus saving 360 min of
instrument time.
P037
MALDI-TOF MS profiling of spider venoms
1
1
2
1
O. Bocanek , O. Sedo , S. Pekar , Z. Zdrahal
1
CEITEC Masaryk University, Research Group Proteomics, Brno, Czech Republic
2
Masaryk University, Department of Botany and Zoology, Brno, Czech Republic
Question: Spider venoms are highly complex mixtures with high amount of biologically active substances with potential use in
biotechnology or pharmacology. Composition of venom varies from species to species, and also between genders. MALDI-TOF
MS profiling can be used to distinguish these differences. Currently, there is no standard procedure for analysis of spider venoms
using MALDI-TOF MS profiling. The main objective of this study is to find and to establish a standard sample preparation
procedure for venom profiling by MALDI-TOF MS.
Methods: We selected the most frequently used conditions for MALDI-TOF MS analysis of spider venom from already published
studies. Various matrices, matrix solvents, as well as sample spotting methods, were tested to find the optimum conditions. As a
model sample, venom from Brachypelma albopilosa was analyzed in weekly intervals.
Results: The evaluation of obtained mass spectrometric data proved that the sample preparation conditions have significant
impact on the quality of obtained data, especially in terms of the signal resolution, relative intensity, or resolution. Apart from
the signal quality, also the reproducibility of the mass spectra was assessed as a crucial factor for selection of the optimum
conditions.
Conclusions: The optimized sample preparation protocol will provide the best results for MALDI-TOF MS profiling in terms of
informativeness and reproducibility. This protocol will be proposed as a standard procedure bringing the possibility of valuable
inter-laboratory comparison of the MALDI-TOF MS profiles of spider venoms.
This study was supported by CEITEC (Central European Institute of Technology) (CZ.1.05/1.1.00/02.0068 funded from the
European Regional Development Fund).
P038
Low temperature stress has an impact on the regulation of sexual development in Aspergillus nidulans
1
1
1
1,2
1,2,3
B. Hanf , T. Krüger , D. Mattern , A. A. Brakhage , O. Kniemeyer
1
Leibniz Institute for Natural Product Researchand Infection Biology – Hans-Knöll-Institute, Molecular and Applied Microbiology,
Jena, Germany
2
Friedrich Schiller University, Department of Microbiology and Molecular Biology, Institute of Microbiology, Jena, Germany
3
University Hospital, Integrated Research Treatment-Center, Center for Sepsis Control and Care (CSCC), Jena, Germany
Fungi naturally encounter sharp temperature shifts in their environment. In this context, some fungi have evolved protective
mechanisms to be more resistant against deep temperatures. Moreover, silent gene clusters can be induced due to the initiation
of stress factors. To identify the key processes induced at low temperatures, the proteome of the well-studied filamentous
Proteomic Forum 2015
74
fungus Aspergillus nidulans was analyzed under defined cold stress conditions in a comparative gel-based (DIGE) and gel-free
(LC-MS/MS) proteomic approach. Furthermore, the formation of secondary metabolites was identified by LC-MS/MS. Another
focus of our study is the investigation of post-translational changes, particularly of the acetylome, which is currently completely
unknown for Aspergillus. The comparative protein analysis led to the identification of proteins involved in cellular transport,
transcription, cell cycle and protein modification as a stress response to low temperatures. We also identified indirectly by
proteomics and directly by metabolic analyses factors that influence the sexual development in A. nidulans and compounds that
exhibit toxic activity. Besides providing first insights into the regulation of sexual development in A. nidulans upon cold stress,
this work shows the successful combination of different genomic-based methods to improve the identification of activated
secondary metabolite biosynthesis gene clusters.
P039
Exploring the potential of a last generation UHR-Q-TOF for rapid generation of accurate information on proteoforms mass,
distribution and relative abundancy.
1
2
2
2
P.-O. Schmit , W. Jabs , K. Meyer , C. Albers , S. Kaspar
1
Bruker Daltonique S.A., WISSEMBOURG, France
2
Bruker Daltonik GmbH, Bremen, Germany
2
Introduction: Proteins undergo various transformations that can alter their functions while keeping a part of their primary
sequence intact. These multiplication of PTM patterns, alternative splicing forms or products of proteolitic processing cannot be
simply resolved with a bottom-up approach. However, the information relative to the distribution of different proteoforms is
encoded in their intact masse. Being able to catch this information efficiently will drive the biologist into a new analytical
dimension, far beyond the traditional gel-based approaches.
Objectives: In this study we were willing to evaluate how the latest generation UHRQ-Tof could help to quickly resolve
proteoforms in complex mixtures separated with fast LC, in order to map their distribution and relative abundancies.
Material and Methods: Undigested protein mixtures of E.Coli Yeast , and non-depleted plasma have been separated on a
150x2.1mm C4 mm column, using 20 min or 45 minutes chromatographic methods. Separation were performed on an Ultimate
RSLC system coupled to an impact II benchtop UHR-Q-Tof or a floor-standing maXis II UHR-Q-Tof (Bruker).
Mixtures of yeast intact protein (Promega) spiked with various ratios of UPS2 (Sigma) have been separated on a 250 mmX100
µm monolithic column (ThermoFischer Scientific), coupled to an impact II benchtop UHR-Q-Tof (Bruker) via a CaptiveSpray
nanoBooster ion source (Bruker).
All data have been automatically processed in Data Analysis 4.2 (Bruker Daltonics). Statistical analyses have been performed on
a slightly modified version of the Profile Analysis 2.1 Software. Identifications have been performed using BioTools 3.2 (Bruker
Daltonics) and Mascot 2.4(Matrix Science).
Results: Using High-Quality threshold protein detection, we could easily detect more than 800 proteoforms out of the E.Coli
mixture, and more than 1500 proteoforms out of 50µg of the Yeast mixture with a 20 minutes method. A 45 minutes method
enabled to distinguish more than 1000 proteoforms from a 1µl injection of non-depleted plasma.The spectral quality usuallay
observed for single compounds was preserved while measuring in these highly complex mixtures : the average mass error for
the monoisotopic peak of the RS 19 protein (10,2 Kda) over 4 injections was 0,07 ± 0,2 ppm. The isotopic fidelity was always
better that 2%, and the protein could be identified in Auto LC-MS/MS using CID fragmentation. We have been able to separate
the yeast extracts spiked with various amount of the UPS II mixture, suggesting that the approach enables to give a relative
quantitation information for the compounds present in the initial mixture.
Conclusions: The last generation UHRQ-TOF, by combining a large spectral dynamic range to the capacity of preserving a high
spectral quality over a large mass range in complex mixtures, are now capable of delivering rapidly a high-quality proteoform
distribution information that has the potential to complement the information delivered by the bottom-up approaches.
Proteomic Forum 2015
75
P040
Pinpointing Protein-Protein Interactions of the Yeast Mediator Complex by Glutaraldehyde Cross-linking and GeLCMS
1
1
1
1
M. Bach , H. Uthe , J. T. Vanselow , A. Schlosser
1
Uni Würzburg, Rudolf-Virchow-Zentrum, Würzburg, Germany
Question: Detailed information about protein-protein interactions is essential in order to understand how proteins and protein
complexes fulfill their cellular functions. Mass spectrometry (MS) in combination with co-immunopurification (co-IP) has become
the method of choice for the identification of protein interaction partners. One limitation of this common strategy is that weak,
transient and rapidly exchanging interaction partners are typically lost during Co-IP. Another limitation is that the list of specific
interaction partners comes without any spatial information. If the bait is a single protein this means that it is not possible to
distinguish between direct and indirect interaction partners. If the bait is a protein complex this means that an interaction
partner cannot be pinpointed to a specific subunit of the complex. Recently, Subbotin et al. demonstrated that glutaraldehyde
cross-linking has the potential to overcome these limitations [1]. Here we combined glutaraldehyde cross-linking with GeLCMS
to pinpoint protein-protein interactions of the yeast mediator complex.
Methods: HA-tagged Med2 was used for co-IP of the mediator complex from yeast. Cross-linking was performed on the beads
with 0.01 % glutaraldehyde for 5 min on ice. The reaction was quenched with Tris-HCl buffer. Proteins were separated with TrisAcetate gels (Invitrogen). Each Lane was cut into 30-60 bands and analyzed by nanoLC-MS/MS on an Orbitrap Velos (Thermo)
using a HCD TOP5 method. MS data were processed with MaxQuant and protein intensities were used to create profiles for all
identified mediator subunits.
Results: Having optimized the conditions for on-beads cross-linking with glutaraldehyde for the mediator complex, we used TrisAcetate gels to separate proteins. The complete gel lane was analyzed by GeLCMS and an intensity profile was calculated for
each identified mediator subunit. All subunits showed an intense peak at their expected molecular mass, and most subunits
showed a number of additional peaks with lower intensity at higher molecular mass, corresponding to cross-linked species of the
same subunit. All proteins identified in the same band as the cross-linked species are potential cross-link partners. To narrow
down the list of potential cross-link partners we applied additional criteria, such as the protein stoichiometry calculated from the
iBAC values of the proteins [2]. This allowed us to retrieve proximity information for several subunits of the mediator complex.
Conclusions: Glutaraldehyde cross-linking in combination with GeLCMS is a promising strategy for retrieving proximity
information from co-IP experiments.
References:
[1] R. I. Subbotin, B. T. Chait, Molecular & Cellular Proteomics 2014, 13: 2824-35, DOI: 10.1074/mcp.M114.041095.
[2] Smits et al. 2013, Nucl. Acids Res.
P041
LysargiNase mirrors trypsin for protein C-termini and methylation site identification
1,2
2
2
2
2
3
4
4
P. F. Huesgen , P. F. Lange , L. D. Rogers , N. Solis , U. Eckhard , O. Kleifeld , T. Goulas , F. X. Gomis-Ruth , C. M. Overall
1
Forschungszentrum Jülich, ZEA-3: Analytics, Jülich, Germany
2
University of British Columbia, Centre for Blood Research, Vancouver, Canada
3
Monash University, Department of Biochemistry and Molecular Biology, Clayton, Australia
4
Molecular Biology Institute of Barcelona, Department of Structural Biology, Barcelona, Spain
2
Question: Bottom-up-proteomics relies on proteolytic digestion of complex proteomes into peptides for large-scale
identification by tandem mass spectrometry. The cleavage specificity of the chosen digestion enzyme determines peptide
length, m/z, sequence and charge position. This affects the likelihood of sequence identification and localization of posttranslational modification sites. Hence, proteases with new cleavage site selectivity are of great interest. Here we identified an
enzyme that complements commonly used trypsin by selectively cleaving before arginine and lysine residues.
Methods: We tested a recombinant form of the archaea metalloproteinase LysargiNase (formerly named ulilysin) for proteomics
applications. Cleavage site specificity and fidelity was characterized by shotgun analysis of LysargiNase- or trypsin-digested
human MDA-MB-231 breast cancer cell lysates, by proteomic identification of protease cleavage sites (PICS) and with
biochemical assays using commercial and synthetic substrates. We further analyzed TiO2-enriched phosphopeptides from
LysargiNase and tryptic digests of MDA-MB-231 cell lysates.
Proteomic Forum 2015
76
Results: LysargiNase was compatible with conditions used in bottom-up proteomics experiments and tolerated temperatures up
to 55C. Shotgun proteomics showed that LysargiNase cleaved selectively before arginine and lysine residues. Fragmentation of
LysargiNase-generated peptides with N-terminal basic residues in ion-trap lead to pronounced b-type fragment ions and
complementary sequence coverage compared to the peptides from tryptic digests. In phosphoproteomics experiments,
phosphorylation motifs with a basic residue N-terminal of the phosphorylation site were overrepresented in LysargiNase digests
in comparison to those with trypsin. PICS and biochemical assays with synthetic peptide substrates further revealed a surprising
ability to cleave at methylated and dimethylated basic residues, which facilitated the identification of peptides carrying these
important post-translational modifications.
Conclusion: We characterized the thermophilic metalloprotease LysargiNase (formerly named ulilysin) as a novel digestion
enzyme for bottom-up proteomics. LysargiNase mirrored trypsin specificity by selectively cutting N-terminal of arginine and
lysine residues, including their methylated and dimethylated forms. LysargiNase improved the identification of C-terminal
peptides, specific phosphorylation motifs and peptides with methylated or dimethylated basic residues. Overall, LysargiNase is a
broadly applicable, powerful addition to the protease toolbox for bottom-up proteomics.
Figure 1
P042
Optimization of capillary electrophoresis method for monitoring of RepA protein - DNA binding reaction
1
1
1
1
1
P. Baran , P. Zaleski , P. Wawrzyniak , A. Bierczynska-Krzysik , G. Plucienniczak , A. Plucienniczak
1
The Institute of Biotechnology and Antibiotics, Bioengineering Department, Warszawa, Poland
1
Question: Rep A protein is a plasmid replication protein expressed in Escherichia coli that binds specifically to the determined
DNA sequence. The goal of this study was to develop and optimize the capillary electrophoresis (CE) method enabling
monitoring of interactions between the RepA protein and pIGWZ12 plasmid derived DNA sequences. The cryptic plasmid
isolated from an Escherichia coli strain was described in 2006 as a completely new molecule with only two small regions of
nucleotide homology to already known plasmids (Zaleski et al., 2006). Determination of the DNA sequence fragment recognized
by the RepA protein has a substantial role in understanding the replication control mechanisms in pIGWZ12.
Methods: A ProteomeLab™ PA 800 system (Beckman-Coulter, Fullerton, CA) with laser induced fluorescent (LIF) detection was
employed. Fluorescence was excited with the 488-nm line of a 3mW argon-ion laser and detected at 520 nm. The DNA sample
was electrokinetically injected for 5 s at 5kV voltage. The samples were separated by applying 10 kV voltage. Data were collected
and analyzed with Beckman System Gold software.
Results: A series of ten DNA fragments covering the whole pIGWZ12Kan plasmid were analyzed separately and in the presence
of the RepA protein. The fragments overlapped by stretches of several nucleotides. Based on the capillary electrophoresis results
two DNA fragments were indicated to interact with the RepA protein. Both fragments contained respectively one and four
specific nucleotide sequences enabling the protein binding. In the latter case, the presence of multiple protein binding sites
(PBS) resulted in enhancement of the RepA protein - DNA signal on the electropherogram. Moreover, for higher RepA protein
concentrations additional peaks corresponding to the bound DNA fragments arised.
Conclusions: It is important to notice that it is not the DNA sequence length, but the size of the whole DNA-protein complex that
determines the choice of the CE separation method. The method advised for analysis of 1000 base pairs (bp), as in our case, did
not allow to obtain satisfactory results. Taking into account only the DNA length may lead to false-negative results. Application
and optimization of the CE method for analysis of 20 000 DNA bp permitted to achieve recurrent and readable results.
Interestingly, the approach enabled identification of the multiple RepA protein - DNA complexes.
Proteomic Forum 2015
77
P043
Comparative venomic characterization of four medicinally important Turkish vipers (Macrovipera lebetina, Montivipera
xanthina, Vipera ammodytes, V. kaznakovi)
1
2
3
2
4
1
N. Igci , D. Petras , B. Gocmen , R. Süssmuth , M. Z. Yildiz , F. D. Ozel-Demiralp
1
Ankara University, Biotechnology Institute, Ankara, Turkey
2
Technische Universitat Berlin, Institüt für Chemie / FG Organische Chemie, Berlin, Germany
3
Ege University, Department of Biology, Izmir, Turkey
4
Adiyaman University, Department of Biology, Adiyaman, Turkey
Introduction: Exploring venom proteomes is of great interest due to their potential as sources of bioactive peptides and proteins
with novel pharmacological properties.
Objectives: Our aim is to characterize the venom proteome of four Turkish vipers in order to evaluate their potential as a source
for bioactive molecules, as well as to determine specific venom variations.
Materials and Methods: Snakes were collected in various regions of Turkey and pooled venom samples were freeze-dried after
extraction. Bottom-up proteomics experiments were carried out on a Thermo Orbitrap XL mass spectrometer (MS) equipped
with an Agilent 1100 HPLC system. Crude venom proteins were separated by reverse phase chromatography (RPC) using C18
column, under increasing acetonitrile gradient. Each fraction were collected and evaporated in a vacuum centrifuge. Selected
fractions were loaded onto 12% SDS-PAGE gels and stained with Coomassie Brilliant Blue. Protein bands were excised and in gel
digestion was performed with trypsin. The resulting peptides were submitted to the same LC-MS/MS system using C18-RPC for
prior peptide separation. MS/MS fragmentation spectra were obtained for selected peptides using CID and HCD methods in
combination. De novo peptide sequences were obtained either by manual data interpretation or by Peaks de novo software tool.
Thereafter, sequences were submitted to BLASTP to identify the proteins.
Results:We identified phospholipase A2 (PLA2), metalloproteinase (SVMP), serine proteinase (SVSP), cysteine-rich secretory
protein (CRISP), disintegrin (DISI), c-type lectin (CLP), L-amino acid oxidase (LAAO), vascular endothelial growth factor (VEGF),
nerve growth factor (NGF), proteinase inhibitor and bradykinin-potentiating peptides (BPP) totally in all venoms. Hovewer,
distribution and abundance of these proteins, which were calculated, based on the corresponding peak areas of UV 214-RPC
among studied venoms shows variation and is demonstrated in the Figure 1.
Conclusions: In this work, the first detailed proteomic insight into the Turkish viper venoms was achieved, including totally
unexplored venoms like M. xanthina and V. kaznakovi. Our results show that these venoms contain biotechnolgically important
proteins and peptides that will be a guide for our further studies on the purification of bioactive proteins. Furthermore, the
determination of the compositional variations of different Turkish viper venoms is of great importance for developing much
more potent antivenoms and our comprehensive study gives the first insight into the venom variation of Turkish vipers.
References: Calvete JJ, Sanz L, Angulo Y, Lomonte B, Gutierrez JM (2009) Venoms, venomics, antivenomics. FEBS Letters, 583:
1736-1743.
Figure 1: Comparative protein/peptide compositions of Macrovipera lebetina, Montivipera xanthina, Vipera ammodytes and V.
kaznakovi
Figure 1
Proteomic Forum 2015
78
P044
Metaproteome analysis of the salivary and tongue proteome
1
1
2
3
3
1
A. Rabe , M. Gesell Salazar , S. Fuchs , H. Senkbeil , T. Kocher , U. Völker
1
Interfaculty Institute for Genetics and Functional Genomics, University Medicine Greifswald, Department of Functional
Genomics, Greifswald, Germany
2
Ernst-Moritz-Arndt-University Greifswald, Institute of Microbiology, Greifswald, Germany
3
Department of Restorative Dentistry - Unit of Periodontology, Periodontology and Endodontology, University Medicine
Greifswald, Greifswald, Germany
Background: In recent years, the contribution of the human microbiome to the health status of humans has become a new focus
of research. Metaproteomics enable insights into the behavior of microbial communities in their natural habitat.
Aim: The aim of the present study is to establish a community proteomics approach for different habitats in the human oral
cavity. Different sampling protocols were tested. The focus was on the optimization of coverage of proteins and characterization
of protein functions and taxonomy.
Materials and Methods: Saliva (split into pellet and supernatant) and tongue swabs were taken from ten healthy subjects. The
oral samples were analyzed by a gel-free approach and measured via the QExactive. The analyzed spectra were searched against
the "Human Oral Microbiome Database" (HOMD) combined with a human swissprot database. Taxonomic and functional
assignment was performed via the Prophane bioinformatics pipeline. Proteins were considered to be present if they were
covered by at least two peptides. To find indications for a possible core-metaproteome, proteins had to be identified in at least
50 % of all biological and technical replicates.
Results: The highest number of human proteins could be identified for the saliva - pellet followed by the tongue samples. Most
bacterial proteins could be analyzed for the tongue and for the saliva - pellet whereas in comparison the identification rate for
the saliva - supernatant was low. Analyses of the bacterial protein functions revealed that most proteins could be assigned to
protein syntheses and energy metabolism. Taxonomical analysis demonstrated that Firmicutes, Actinobacteria, Bacteriodetes,
Proteobacteria and Fusobacteria were the most prominent phyla in saliva - pellet as well as in the tongue swabs.
Conclusions: The applied protocols enable the creation of a workflow for community proteomics analysis of the bacterial and
human composition of the tongue and salivary proteome. The obtained data provides an insight into the functional and
taxonomical composition of different microbial communities in the human oral cavity.
P045
Protein expression patterns in ileal tissue of sheep infected with Mycobacterium avium subsp. paratuberculosis: in search for
disease mechanisms and biomarkers
1
2
2
1
1
S. Pisanu , T. Cubeddu , S. Rocca , S. Uzzau , M. F. Addis
1
Porto Conte Ricerche, Alghero, Italy
2
University of Sassari, Veterinary Medicine, Sassari, Italy
Introduction: Mycobacterium avium subspecies paratuberculosis (MAP) is the causative agent of Johne’s disease (JD), or
paratuberculosis, a chronic granulomatous enteritis of ruminants. The disease is mostly studied in cattle, but it severely affects
also other economically relevant ruminants, including sheep. In view of its relevance, and of the difficulties in early diagnosis,
new markers and a better understanding of the disease are highly needed.
Question: To characterize protein expression changes in intestinal tissues of sheep at two different levels of infection,
paucibacillary and multibacillary, with the aim of further understanding disease processes and to find molecules with prognostic
and diagnostic potential.
Methods: Ileal sheep tissues were examined for presence of Acid-Fast bacilli and by quantitative PCR (qPCR) for IS900 and F57 of
MAP, and classified into MAP-free (K), paucibacillary (P) and multibacillary (M). Tissue samples were then subjected to protein
extraction, two-step filter-aided sample preparation (FASP), single run Orbitrap Velos MS, and label-free quantitation by the
spectral counting approach. Then, differential proteins were subjected to pathway analysis by STRING.
Results: A total of 3643 sheep protein identifications were obtained (at least two peptide-spectrum matches and at least two
biological replicates). Among these, 473 showed a differential abundance in the three experimental groups (RSC ≥ ±1.5, p ≤ 0.05,
Proteomic Forum 2015
79
adjusted by FDR). Specifically, 121 and 413 proteins were differential in P vs K and in M vs K, respectively, with 70% more
proteins changed in M vs K (Fig 1). The differential protein patterns obtained produced a statistically significant clustering of K, P,
and M upon Principal Component Analysis (Fig 2). Concerning MAP proteins, 13 were identified in P and 30 in M. STRING
highlighted a significant association of differential host proteins with disease process pathways. In general, defense response
and innate immune response were the two functions with the strongest association with increased proteins. Interesting
differences and similarities emerged between the proteins changed in P and M when compared to K.
Conclusions: The study carried out on sheep at two levels of infection enabled to highlight several processes occurring in
infected ileal tissues upon MAP infection. In particular, the ability to detect changes associated either with the P or M forms may
contribute to understand the differences between these two levels of infection, with the perspective of developing sensitive and
early prognostic and diagnostic tools for detecting infected animals before they develop the clinical manifestations of Johne's
Disease.
Fig 1. Clustering of experimental samples by PCA.
Fig 2. Volcano plots for the samples Paucibacillary/Control (Top, N= 4 vs 3) and Multibacillary/Control (Bottom, N= 4 vs 3).
Figure 2
Figure 1
P046
Novel findings on innate immunity of the mammary gland and their implications for diagnosis, as inferred from the sheep
mastitis model
1
2
1
1
3
2
1
1
S. Pisanu , T. Cubeddu , S. Ghisaura , D. Pagnozzi , G. Marogna , S. Rocca , S. Uzzau , M. F. Addis
1
Porto Conte Ricerche, Alghero, Italy
2
University of Sassari, Veterinary Medicine, Sassari, Italy
3
Istituto Zooprofilattico Sperimentale della Sardegna G. Pegreffi, Sassari, Italy
Introduction: Mastitis is one of the most impacting problems in the dairy industry, being a direct cause for significant
economical losses due to drops in milk production and quality, animal culling, and treatment costs. Typically, dairy farms are
monitored by the Somatic Cell Count (SCC) in milk. Although easy to measure and well established, SCC presents significant
specificity problems, especially in sheep and goats for which variability can sometimes reach impressive levels. New mastitis
biomarkers are therefore highly needed, and further elucidation of the events leading to release in milk of marker molecules and
to the increase in SCC is also required.
Objectives: To characterize the molecules released in mastitic milk, to identify new mastitis markers, and to gather novel
information on the immune response mechanisms taking place in the infected mammary gland.
Proteomic Forum 2015
80
Methods: Milk and tissues were collected from sheep affected by bacterial mastitis due to Mycoplasma agalactiae or
Streptococcus uberis. Proteins associated to the milk fat fraction were analyzed by 2D-DIGE/MS, by Gel-MS/MS, and by filteraided sample preparation (FASP) followed by Orbitrap Velos MS. Gel-MS/MS and shotgun proteomics data were processed by
label-free quantitation followed by spectral counting, and differential proteins were subjected to Ingenuity Pathway Analysis.
Then, results were investigated and validated by PCR, culture, and western blotting of milk, and by fluorescent in-situ
hybridization (FISH), immunohistochemistry, and confocal immunomicroscopy of mammary tissues.
Results: Several proteins involved in inflammation, chemotaxis, and antimicrobial defense were identified in milk of infected
animals. Then, immunohistochemistry (Fig 1) and confocal immunomicroscopy (Fig 2) of mammary tissues provided the first
proteomic evidence that at least two of these proteins are produced and released by mammary epithelial cells (MECs),
highlighting their contribution to the innate immune response to pathogens in the first stages of infection and in subclinical
conditions. In acute infection, these proteins were mainly released by neutrophils recalled inside the alveolus lumen. Then, both
proteomic and microscopic findings demonstrated the presence of NETs.
Conclusion: This study characterized the events occurring in the mammary gland during infection in vivo, providing the first
evidences on the contribution of MECs to the innate immune response. In addition, useful novel diagnostic markers for
mammary infections were identified.
Fig 1. IHC results for two mastitis markers in healthy tissues (top), acutely mastitic tissues (middle) and subclinically mastitic
tissues
(bottom).
Fig 2. Confocal Immunomicroscopy results for two mastitis markers in subclinically mastitic tissues (left) and acutely mastitic
tissues (right).
Proteomic Forum 2015
Figure 1
81
Figure 2
P047
Differential peptidomics of raw and pasteurized ovine milk cheese
1
2
1
1
2
1
1
S. Pisanu , M. Pes , R. Anedda , D. Pagnozzi , A. Pirisi , S. Uzzau , M. F. Addis
1
Porto Conte Ricerche, Alghero, Italy
2
AGRIS Sardegna, Department of Animal Science, Olmedo (SS), Italy
Introduction: During cheese ripening, milk proteins undergo progressive degradation due to residual rennet activity,
endogenous enzymes from milk, microbial enzymes, and enzymes added to improve cheese maturation. Therefore, production
variables can lead to changes in the final cheese peptide profile. One of these variables is pasteurization, which has the ability to
produce changes in milk proteins and enzymes. Investigating the impact that the pretreatment of milk has on the final product
can have important consequences in terms of valorization and preservation of traditional foods, as well as for the consumers’
health in terms of food bioactivity.
Objectives: To characterize the peptide profile of 9 month ripened ovine cheese prepared with raw (R) and pasteurized (P) milk,
and to identify possible differences due to milk pasteurization.
Materials and Methods: Ripened cheese produced using raw or pasteurised ovine milk (R and P) was subjected to extraction of
the water soluble components followed by differential precipitation or FPLC. Mass spectrometry characterization of peptides
was then carried out by ESI-Q-TOF-MS/MS or by Orbitrap Velos MS. Data were then processed by bioinformatic tools to evaluate
differences in peptide abundance, enzyme activities, and bioactive potential.
Results: The specific sequences identified and their abundances in R and P cheese showed statistically significant differences.
Bioinformatic analysis by EnzymePredictor provided insights into the different susceptibility of R and P cheese proteins to
proteolytic enzymes. In addition, BIOPEP analysis predicted a significantly higher abundance in R cheeses of peptides having
bioactive properties.
Conclusion: Exposure of milk to high temperatures likely impacts also on suitability of its proteins to enzymatic degradation,
adding to its influence on the indigenous enzymes. In addition to sensory and textural features, the bioactive potential of the
final product appears to be influenced. The identification of peptide molecules with beneficial properties in sheep cheeses made
from raw milk can provide important opportunities for their valorization and preservation.
Proteomic Forum 2015
82
P048
Omics Analysis in the Brain: How Similar are Transcriptome and Proteome?
1
1
1
2
1
C. Moritz , T. Mühlhaus , T. Schulenborg , S. Tenzer , E. Friauf
1
University of Kaiserslautern, Department of Biology, Kaiserslautern, Germany
2
University Medical Center Mainz, Institute of Immunology, Mainz, Germany
BACKGROUND AND HYPOTHESIS: The degree of (post-)translational regulation strongly depends on the sample. The proteome
of the brain is more complex than that of other organs and is highly regulated by various types of post-translational
modifications (PTMs), such as phosphorylation, ubiquitination, glycosylation, and oxidation. Translational and post-translational
regulation affects the degree of the linear transcriptome-proteome correlation. Besides PTMs, a further potential cause of
nonlinear-correlating gene products in the brain is the asymmetric morphology of neurons: proteins may be spatially separated
from their transcripts, e.g. by axonal transport. We expect that these factors - and thus the transcriptome-proteome correlation
- vary between brain regions.
METHODS: To address this issue, we here analyzed the highly specialized cochlear nuclear complex (CN), superior olivary
complex (SOC), and inferior colliculus (IC) in the rat auditory brainstem. As a fourth and more general region, we included the
rest of the brain (Rest). We compared these brain regions to each other regarding their transcriptome and their proteome,
enclosed by microarray analysis and lable-free mass spectrometry, respectively.
RESULTS: Regarding the principal component analysis, the arrangement of the protein samples reflects the physical location of
the different brain regions suggesting that the protein sets of these regions determine their specialization. The protein level
2
2
2
clearly correlated with the corresponding transcript level in all four brain regions (CN: R =0.47; SOC: R =0.54; IC: R =0.44; Rest:
2
2
R =0.34). The R values ranging from 0.34 to 0.54 indicate that the four brain regions comprise different degrees of (post)translational regulation or axonal protein transport. To further specify the correlation, we identified 29 gene products which
showed significantly higher protein and transcript levels in the auditory brainstem nuclei CN, SOC, and IC in comparison to Rest.
For example, the synaptic proteins Synaptotagmin-2, Extended Synaptotagmin-1, and Synaptobrevin-1 belong to this set.
Further, we found 17 gene products (e.g., Synaptotagmin-1) that showed significantly lower protein and transcript levels in the
auditory brainstem nuclei compared to Rest.
CONCLUSION: Together, our findings emphasize the importance of proteomic research, especially as structures (e.g. brain
regions) differ regarding their correlation between transcripts and corresponding proteins.
P049
In vivo proteomics approaches unravel adaptation strategies of the pathogen S. aureus HG001 and host defense mechanisms
of human alveolar epithelial cells upon internalization
1
1
1,2
1,2
1
3
1
4
5
K. Surmann , M. Simon , P. Hildebrandt , S. Michalik , H. Pförtner , S. Stentzel , V. M. Dhople , R. Schlüter , P. Gierok , M.
5
3
1,2
1
Lalk , B. Bröker , F. Schmidt , U. Völker
1
University Medicine Greifswald, Functional Genomics, Greifswald, Germany
2
University Medicine Greifswald, ZIK-FunGene Applied Proteomics, Greifswald, Germany
3
Institute of Immunology and Transfusion Medicine, Greifswald, Germany
4
Institute for Microbiology, Greifswald, Germany
5
Institute of Biochemistry, Greifswald, Germany
Questions: Staphylococcus aureus causes life-threatening diseases such as pneumonia or septicaemia. During persistence inside
non-professional phagocytic cells it survives elimination by immune cells or antibiotics. To develop more efficient therapies and
vaccination strategies, knowledge on the host-pathogen interactions is indispensable. Here we simultaneously analyzed
adaptation of S. aureus HG001 and human lung alveolar epithelial A549 cells to internalization. Since proteome analyses of
infections are often hampered by the low number of bacteria or infected cells available, enrichment strategies are necessary to
investigate adaptation reactions in more detail.
Methods: A549 cells were infected with S. aureus HG001 pMV158GPF and proteome changes were investigated hourly in the
first 6.5 h. Using GFP, bacteria could be enriched from lysed host cell debris by fluorescence activated cell sorting. In addition,
intact A549 cells, which carried the fluorescent bacteria after infection could be sorted from those, which did not. Furthermore,
we conducted a mock-infection of A549 cells with sterile infection mix to investigate the impact of even pure contact to bacteria
on the host proteome. Proteins of bacteria or A549 cells were proteolytically digested and analyzed in a gel-free shotgun
Proteomic Forum 2015
83
proteomics approach. Additionally, metabolome measurements of the host were performed to determine the intracellular
nutrient reservoir at the starting point of infection.
Results: S. aureus HG001 replicated inside A549 cells during the first 6.5 h p.i. but only 20% of the A549 cells actually carried
bacteria after contact. We quantified more than 800 S. aureus HG001 proteins upon internalization and observed reduced
protein biosynthesis, altered levels of various metabolic enzymes, increased response to cell wall and oxidative stress, and
adaptation to microaerobic conditions. The proteome profile of A549 cells which contained S. aureus HG001 after contact
differed from that which did not. Interestingly, also the non-infected A549 cells showed different protein abundance compared
to the mock-infected A549 cells. Infected A549 host cells produced cytokines, showed altered amounts of mitochondrial
proteins, and activated early caspase-9 mediated apoptosis.
Conclusion: To our knowledge, we provide for the first time a comprehensive and time-resolved picture of adaptation
mechanisms of internalized S. aureus and defense strategies of infected epithelial cells in one simultaneous infection assay.
Thus, this study contributes to a better understanding of the pathogenicity and virulence of this important pathogen.
Proteomics for Human Health I (P123–P145)
P123
Concentration gradients in human hippocampal regions of interest (CA1, CA2, CA3, fascia dentata) - relevance for
neurodegenerative diseases ?
1
2
3,4
1
A. Schrötter , H. Heinsen , L. T. Grinberg , H. E. Meyer
1
Leibniz-Institut für Analytische Wissenschaften - ISAS - e.V., Biomedical Research , Dortmund, Germany
2
Laboratory of Morphological Brain Research, University of Würzburg, Department of Psychiatry, Würzburg, Germany
3
UCSF School of Medicin, Neurology, San Francisco, United States
4
University of Sao Paulo , Medical School, Sao Paulo, Brazil, United States
Alzheimer’s disease (AD) and Parkinson’s disease (PD) are the most common forms of dementia. Regarding the last 40 years of
scientific research we conclude that the reason for AD and PD might be more complex then known so far. Braak et al. (2000)
observed in AD that neuronal damage of the different regions of the human hippocampus (CA1, CA2, CA3, fascia dentata) occurs
in a time dependent matter. Studies reporting hippocampal volumes have shown that hippocampal atrophy is also seen in nondemented and non-medicated patients with early stage PD (Brück et al. 2004, Camicioli et al. 2003). Taken all this together we
decided to analyze the content of the human hippocampal regions of interest by performing a differential proteomic study (label
free LC-MS/MS approach) combined with a couple of functional analyses. We used the latest laser-microdissection technique to
separate the human hippocampal regions of interest. Afterwards we identified and quantified the hippocampal proteoms of
these regions of interest. By using 6 biological human post mortem replicates we was able to identify specific hippocampal
protein concentration gradients of prominent candidate proteins. These candidate proteins might be the reason for the
prograding hippocampal atrophy. In addition we confirmed our findings by using immunohistochemistry.
P124
Proteome-wide analysis reveals an age-associated cellular phenotype of in situ aged human fibroblasts
1
2
3
1
1
4
2
5
6
D. M. Waldera-Lupa , F. Kalfalah , S. Sass , F. Kruse , V. Rieder , J. Tigges , F. Boege , F. Theis , G. Reifenberger , K. Stühler
1
Heinrich-Heine-University, Molecular Proteomics Laboratory, Düsseldorf, Germany
2
Heinrich-Heine-University, Institute of Clinical Chemistry and Laboratory Diagnostics, Düsseldorf, Germany
3
Helmholtz Center Munich, Institute of Computational Biology, Neuherberg, Germany
4
Leibniz Research Institute for Environmental Medicine, Düsseldorf, Germany
5
Technical University Munich, Department of Mathematics, Garching, Germany
6
Heinrich-Heine-University, Department of Neuropathology, Düsseldorf, Germany
1
It has been suggested that the aging process of the skin is mainly determined by alterations of the dermis. The dermis is a
stationary tissue compartment of mostly quiescent cells that cannot be removed quickly in case of dysfunction. Thus, the
homeostasis of the dermis is primarily based on cellular adaptation. As the major cell source in the dermis and a long-lived cell
system, fibroblasts are able to accumulate aging-associated alterations and adapt their cellular functions. Therefore, fibroblasts
are a favored cell model to analyse skin aging. Several studies have been applied analyzing aged fibroblasts focusing on single
genes or pathways, while comprehensive investigations of aging-associated alterations at the transcriptome-, miRNAome- and
proteome-wide levels are sparse.
Proteomic Forum 2015
84
The goals of this study were to identify and quantify age-dependent proteins in primary, ex vivo cultured human dermal
fibroblasts obtained from donors of different ages. By performing combined bioinformatic analyses of proteome, transcriptome
and miRNAome data we aimed to characterize aging-associated molecular changes and to identify relevant biological processes
and regulatory mechanisms.
Dermal fibroblasts of 15 female donors of three different age groups (20-26, 40-49, 60-67) were cultivated under standardized
conditions. For LC-MS analysis the cells were prepared by standard cell lysis and digestion protocol. LC-MS analyses were carried
out using a LC-MS system composed of an RSLCnano and an Orbitrap Velos. MASCOT was used for protein identification.
MS/MS-spectra were searched against the UniProtKB database. Label-free quantification of proteins was carried out using
Progenesis. Network analysis was carried out using Cytoscape and ClueGo.
We analyzed an ex vivo model of in situ aged human dermal fibroblasts, obtained from 15 donors from three different age
groups using an unbiased quantitative proteome-wide approach applying label-free mass spectrometry. Statistical analysis
revealed 43 proteins with a significant age-associated expression. Most of the differentially abundant proteins have not been
described in the context of fibroblasts’ aging before, but the deduced biological processes confirmed known hallmarks of aging.
The exhaustive analysis of protein and mRNA data revealed that 77 % of the age-associated proteins were not linked to
expression changes of the corresponding transcripts. This is in line with miRNA data and led us to the conclusion that most of
the age-associated alterations at the proteome level are likely caused post-transcriptionally rather than by differential gene
expression.
The present study led to the characterization of novel proteins potentially associated with fibroblast aging and revealed that
primary cultures of in situ aged fibroblasts are characterized by moderate age-related proteomic changes comprising the
multifactorial process of aging.
P125
Beyond the hairball, uncovering the quantitative EGFR interactome in oncogenic K-Ras Colorectal cancer model
1
1
2
2
3
1
L. Dernayka , M. A. Jarboui , C. Raso , S. Kennedy , K. Bryan , K. Boldt , M. Ueffing
1
Universitaetsklinikum Tübingen, Medical Proteom Center, Tübingen, Germany
2
System Biology Ireland, University College Dublin , Dublin, Ireland
3
EMBL Australia , SAHMRI, Adelaide, Australia
1
The EGFR signalling pathway is involved in the regulation of crucial biological processes such as proliferation, survival,
differentiation and growth. Additionally, alterations in this pathway have been associated with human malignancies such as
colorectal, pancreatic, lung and breast cancer, which makes it of great interest for targeted therapy and drug design. Yet, poor
response rate and development of resistance to EGFR targeted therapy exemplify the need for an in-depth knowledge of this
pathway to efficiently design proficient treatments.
In the PRIMES project, our goal is to untangle, using a quantitative proteomic strategy, a static and dynamic interaction map of
the EGFR pathway in a colorectal cancer (CRC) model. Our experimental approach is based on stable isotope labelling coupled
with affinity purification in tandem with quantitative mass spectrometry.
We successfully identified specific interactors of 93 selected baits of the EGFR core pathway. Furthermore, our analysis depicts
quantitative changes in the bait/preys interaction dependent on the K-Ras G13D mutation. Moreover, using in-silico Data
integration and analysis, we isolated cluster interaction landmarks within the static interactome that provide insights for future
quantitative dynamic analysis.
Our data set the ground for a systems biology-based model of the EGFR-pathway to further integrate signalling dynamics and
cancer-associated perturbations.
Proteomic Forum 2015
85
P126
Protein alterations in women with chronic widespread pain - a proteomic study of the trapezius muscle
1
1
1
1
1
1
P. Olausson , D. Sjöström , E. Blixt , N. Ghafouri , B. Ghafouri , B. Gerdle
1
IMH, Rehabilitation medicine, Linköping, Sweden
Methods: This study aims to investigate the proteome of trapezius muscle biopsies of women with chronic widespread pain
(CWP) compared to female pain-free controls (CON) and to what extent proteins correlate to different pain thresholds using
proteomic tools with multivariate modeling. CWP has a high prevalence resulting in prominent negative individual and societal
consequences. There is no clear consensus concerning the etiology behind CWP although alterations in the central processing of
nociception maintained by peripheral nociceptive input has been suggested. Two-dimensional gel electrophoresis (2-DE) analysis
and nano-liquid chromatography/mass spectrometry (nLC/MS) was used to study the proteome differences of biopsies from the
trapezius muscle between 18 female patients diagnosed with CWP and 19 healthy female controls. Multivariate statistics were
used to detect significant proteins.
Results: Seventeen proteins were significantly altered. 10 were up-regulated and 7 were down-regulated in the CWP group
compared to healthy controls. By multivariate modeling 9 proteins correlated with the patients self-appointed numeric rating
scale values (NRS) of pain. Furthermore, 14 proteins correlated with pressure point thresholds (PPT). Many of the proteins were
important enzymes in metabolic pathways like the glycolysis and gluconeogenesis. Other proteins were associated with muscle
damage and muscle recovery. The contractile and structural proteins have been associated with various types of myopathies.
Conclusions: The changes in expression levels of proteins belonging to the metabolic, structural, stress and inflammatory and
contractile groups, reflects abnormalities and protein alterations in the myalgic trapezius muscle. Our findings give further
support that the periphery maintains and is a part of the activated pain mechanisms in chronic pain conditions.
P127
Investigating Biological Variation in Human Hepatocytes of Phase I and II drug Metabolism Enzymes
1
1
1
2
3
C. Baumann , J. Dojahn , M. Macht , C. Hunter , H. Zhang , X. Wang
1
AB SCIEX, Darmstadt, Germany
2
AB SCIEX, Redwood City, United States
3
Pfizer, Groton, United States
4
AB SCIEX, Framingham, United States
4
Introduction & Objectives: Measurement of drug metabolizing enzymes responsible for phase I and II biotransformations is a
fundamental aspect of assessing drug-drug interactions, and evaluating drug safety and efficacy. In this work, we used SWATH
Acquisition, a data independent acquisition method, to analyze protein expression levels of many of the enzymes involved in the
drug metabolism.
Materials & Methods: A spectral ion library containing more than 2000 proteins was generated from data dependent analysis of
®
a pooled sample. In the SWATH data generated using a TripleTOF 5600+ system, an average of 1987 proteins, including 19 CYP
proteins, 12 UGT proteins, and 7 GST proteins, were quantified across the 13 samples. The quantitative interpretation of SWATH
data was achieved by automatic extracting characteristic fragment ions for each identified peptide from high resolution TOF
MS/MS spectra. A set of protein/peptides of interest obtained from SWATH acquisition was then further analyzed with MRM
®
using a QTRAP 6500 System.
Results & Conclusion: Quantitative comparison of two phase II metabolism enzymes, EST1 - liver carboxylesterase 1 and EST2 cocaine esterase, showed 40% variations across 13 samples. The relative intensities of 4 individual peptides of each protein
showed very good agreement (
1. Lin Y. S. et al., Mol. Pharmacol. 62: 162-172, 2002.
Proteomic Forum 2015
86
P128
Comparative proteomics analysis reveals that small GTPase signaling pathway may playimportant roles in cadmium induced
colon cell maligant transformation
1
J. Lu
1
Jiangsu University, School of Medicine, Zhenjiang, China
Question: Cadmium is a toxic heavy metal which is environmentally and occupationally relevant. The mechanisms underlying
cadmium-induced cell malignant transformation are not yet completely understood,How many proteins involve in this
progression and what are their roles?
Methods: 2D DIGE, Maldi TOF/TOF MS ,Western Blot and some bioinformatic methods (such as Go , KEGG and iHOP analysis)
were employed to reveal the mechanism.
Results: This study shows that small GTPase and G-protein associated signaling pathways may play important roles in cadmium
induced cell transformation. 2D-DIGE and subsequent Maldi TOF/TOF MS and western blot analysis reveal that a set of proteins
linking to G-protein signaling pathways may involved in this progression, such as the small GTPase Rho and Ral, Rho GDPdissociation inhibitor 1, G protein pathway suppressor 1, G protein-regulated inducer of neurite outgrowth 1. Proteins that
involve in mitochondrial transcription , such as mitochondrial transcription factor A, G elongation factor were abnormally
expressed, proteins that associate with DNA repair, such as DNA excision repair protein ERCC-1, Isoform 1, DNA mismatch repair
protein MSH2 isoform 2 were remarkably increased. Proteins that participate in protein metabolism were also notably elevated,
such as Eukaryotic translation initiation factor 4E, Protein disulfide-isomerase A6, Proteasome activator complex subunit 3
isoform 1, Ubiquitin-conjugating enzyme E2C. Go , KEGG and iHOP analysis reveal an interesting connection between small
GTPase, Rho GDP-dissociation inhibitor, COX2 and NEDD9. Evidences show that activation of small GTPase associated signaling
pathway may play important roles in cadmium induced colon cell transformation.
Conclusions: This study demonstrate some important proteins,such as small GTPase RhoA and RalA, Rho GDP-dissociation
inhibitor, COX2 and NEDD9 and associated signaling pathway may play important roles in cadmium induced colon cell
transformation.
Figure 1
Figure 2
Proteomic Forum 2015
87
P129
Lipidomic and proteomic characterization of platelet derived extracellular vesicles
1
2
3
2
1
F. El Magraoui , A. Pienimaeki Roemer , K. Kuhlmann , G. Schmitz , H. E. Meyer
1
Leibniz-Institut für Analytische Wissenschaften -ISAS- e.V., Biomedical Research, Dortmund, Germany
2
University Clinic of Regensburg, Institute for Clinical Chemistry and Laboratory Medicine, Regensburg, Germany
3
Ruhr University Bochum, Medical Proteome Center, Bochum, Germany
Platelets are the smallest of the three major types of blood cells. They are produced by megakaryocytes and contribute to
hemostasis. During their lifetime of 8 to 10 days they release so called extracellular vesicles (EVs) which make up till 90 % of the
circulating EVs in the bloodstream. These platelet derived extracellular vesicles (PL-EVs) are involved in various cellular processes
like autoimmunity, chronic inflammation, thrombosis and neurodegeneration. Function, activity and composition of the PL-EVs
are highly heterogeneous due to their subcellular origin.
To analyze this heterogeneity we used differential centrifugation, filtration and density gradient ultracentrifugation, ending with
five distinct microvesicle (PL-MV) fractions. Further analysis regarding the protein and lipid composition revealed significant
differences between the PL-MV fractions. The detection of the amyloid precursor protein APP, the hallmark protein of
Alzheimer’s disease and the Parkinson's disease related protein alpha-synuclein indicates a contribution of PL-EVs in the
manifestation of neurodegenerative disorders.
P130
MALDI imaging for determining the protein network associated with atrial fibrillation
1
2
2
2
2
3
H. THIELE , S. Mohamed , H. H. Sievers , J. Yan , T. Hanke , O. Klein
1
FRAUNHOFER MEVIS, Inst. Medical Image Computing, Lübeck, Germany
2
Department of Cardic and Thoracic Vascular Surgery, UKSH Campus Lübeck, Lübeck, Germany
3
Charite - Universitätsmedizin Berlin, Center for Regenerative Therapies, Berlin, Germany
Question: Atrial fibrillation (AF) is associated with increased risks of stroke, cardiac failure, and mortality. Since, due to the lack
of sufficient markers, the discrimination and classification of AF subtypes (paroxysmal,persistent, and long-standing persistent
different) is inadequate, the underlying mechanisms and pathology of AF remain elusive. The aim of this study is to
proteomically analyze the left atrial appendage tissue obtained from patients suffering from different subtypes of AF.
Methods: MALDI Imaging mass spectrometry (MALDI-IMS) will be used for differentiation and classification of
pathophysiological AF subtypes, through the direct (in situ) analysis of formalin-fixed paraffin embedded (FFPE) left atrial
appendage tissue. FFPE left atrial appendage tissue resected routinely during MAZE procedure in surgery were collected of
patients with paroxysmal (n = 9, mean age 69.0±3.1 years), persistent (n = 18, mean age 67.0±2.7 years), and long-standing
persistent (n = 19, mean age 71.0±2.0 years) arrhythmia. Sections were dewaxed and trypsin solutions were applied directly
onto the section using an automated spraying device. Spectra were acquired at a mass range of m/z 800-3500Da and lateral
resolution of 80 µm. Two hundred laser shots were acquired per pixel and random walk of 50/position. Data analyses were
performed using SCiLS Lab software (SCiLS GmbH, Bremen).
Results: Component analysis of MALDI Imaging data by probabilistic latent semantic analysis (pLSA) results in a clear
discrimination in the first 3 components of atrial fibrillation. By using receiver operating characteristic (ROC) analysis (AUC > 0.7)
characteristic intensity distribution of given m/z values, which are discriminative for the considered cluster, was determined to
distinguish between paroxysmal vs persistent AF, and persistent vs long-persistent AF. Follow-up with case-controlled
assessment of neurological events it was shown, that persistent AF showed its vulnerability for neurological attacks, compared
to paroxysmal and long-persistent AF (56% vs 13% and 42%, p = 0.001).
Conclusions: The tissue-based proteomic approach provides clinically relevant information to be beneficial in improving risk
stratification in AF patients. In the future the obtained information might be considered as new biomarker to support the
diagnosis of the severity of AF status. They also suggest a new criteria to determine the most appropriate surgery for each AF
subtype to improve postoperative outcomes.
Proteomic Forum 2015
88
P131
Dynamic analysis of differential global phosphorylation changes in Lung Cancer by Discovery Proteomics
1
1
1
2
1
D. Treue , J. Lechner , M. Bockmayr , B. Thomas , F. Klauschen
1
Charité, Institute of Pathology, Berlin, Germany
2
Oxford University, Sir William Dunn Pathology School, Oxford, United Kingdom
Lung cancer is the leading cause of cancer-related deaths worldwide. While patients with advanced lung cancer hardly benefit
from conventional chemotherapy, hopes are high that novel so-called personalized or precision therapies targeting mutationally
altered kinases, such as the epidermal growth factor receptor EGFR, may offer a solution. However, the frequent development
of resistance to targeted drugs still limits their success. Analysis of the underlying mechanisms is therefore pivotal to proposing
more efficaceous combination therapies to overcome such resistances.
To this end, we here present a mass spectrometry-based discovery proteomics approach to analyze dynamic global protein
phosphorylation changes following network perturbation with small molecular inhibitors. Our method is capable of detecting
and quantifying 3,000 phosphopeptides (without fractionation). We compare signaling dynamics in different lung cancer cell
models treated with the EGFR inhibitor Gefitinib. A subsequent bioinformatic analysis of differentially regulated pathways
improved overall robustness and allowed for the identification of potential combination therapy targets.
P132
A mass spectrometry method to identify naturally processed and presented HPV epitopes
1
1
1
2
1
R. Blatnik , S. Hoppe , M. Wühl , C. Rösli , A. B. Riemer
1
German Cancer Research Center (DKFZ), Immunotherapy and -prevention, Heidelberg, Germany
2
German Cancer Research Center (DKFZ), Biomarker Discovery, Heidelberg, Germany
Introduction: Detailed knowledge about human papillomavirus (HPV) CD8+ T cell epitopes that are naturally presented on the
surface of HPV-transformed cells is essential for rational design of a therapeutic vaccine. CD8+ T cell epitopes are 8- to 11-mer
peptides which are produced by the antigen processing machinery, resulting in peptides that do not always contain charged
amino acids. Because of low viral protein expression levels, the quantities of surface-presented peptides originating from HPV E6
and E7 proteins are minimal. Low abundance and challenging peptide ionization complicate the MS-based identification of HPV
epitopes.
3
Objectives: We aimed to develop a nano-UPLC-ESI-MS methodology employing selected-reaction monitoring (SRM) for the
direct detection of low abundant HPV16 T-cell epitopes in the whole cell epitome.
3
Methods: We applied nano-UPLC-ESI-MS with SRM scanning for direct detection of low abundant T cell epitopes from HPVtransformed cells. Epitopes were predicted for their binding affinities and tested for binding to HLA-A2 in vitro. Verified binders
3
were used for MS reference spectra generation. As E6 and E7 are cysteine (Cys)-rich proteins, we ascertained that Cyscontaining peptides were reduced and alkylated. HLA-A2-epitope complexes were immunoprecipitated (IP) from HPV163
transformed cells and analyzed by nano-UPLC-ESI-MS . To confirm the presence of target peptides in IP samples, reference
peptide spectra were compared to the ones of IP samples.
Results: HLA-A2-restricted HPV16 E6 and E7 epitopes were identified by high-throughput in silico epitope predictions. Peptides
scoring high within a self-defined cut off were synthesized and tested for their actual binding affinity to HLA-A2 molecules in
cellular binding assays, resulting in the identification of 25 HLA-A2 binding peptides. Verified binders were used for optimization
3
of MS reference spectra. Next, HLA-A2-peptide complexes were immunoprecipitated from HPV16-transformed cells. Several IP
parameters such as the composition of the lysis buffer, incubation time and additional sonication were optimized, using
continuous feedback from the MS results. Following the IP, epitopes were eluted from HLA-peptide complexes by pH change,
3
subjected to enrichment and analyzed by scheduled nano-UPLC-ESI-MS . Resulting spectra were compared to reference spectra
of synthetically produced target peptides. Positive control peptides were readily detected in all tested HLA-A2+ HPV16+ cell
lines. As expected from known viral interference in antigen processing, HPV epitopes were not constantly present.
Proteomic Forum 2015
89
Conclusions: The presented methodology can be applied for the identification of any low abundant epitopes binding to various
HLA types. Epitopes identified in this project will be further checked for immunogenicity and considered for therapeutic HPV
vaccine design.
P133
Identification of Proteins and Molecular Pathways Regulating Tumor-Initiating Pancreatic Cancer Stem Cells Employing
Proteomics and Transcriptomics
1
1
2
2
1
1
L. Weilnböck , A. Loipetzberger , M. Rurik , O. Kohlbacher , F. Aberger , C. G. Huber
1
University of Salzburg, Department of Molecular Biology, Salzburg, Austria
2
Universtity of Tübingen, Applied Bioinformatics, Center for Bioinformatics, Quantitative Biology Center, and Dept. of Computer
Science, Tübingen, Germany
Cancer stem cells (CSCs) represent a very small subset of the tumor bulk with highly malignant properties. CSCs are responsible
for tumor initiation, growth and metastasis, making them promising targets for innovative cancer therapies. In order to reveal
molecular markers and determinants of CSCs, we employed in this study two highly metastatic pancreatic cancer cell lines
(L3.6sl and L3.6pl). Pancreatic CSCs were enriched in three-dimensional spheroid cultures and compared to bulk tumor cells
using (PTX) and transcriptomics (TCX). For the proteomic approach, isotope-labeled tryptic peptides from CSCs and non-CSCs
samples were analyzed by capillary ion-pair reversed-phase high-performance liquid chromatography and identified and
quantified by linear ion trap-Orbitrap mass spectrometry. Differential transcriptomics was performed by means of microarray
analysis on the Affymetrix Gene Atlas system.
In this study we identified about 350 target proteins differentially expressed in pancreatic CSCs and bulk tumor cells. By
combining the unbiased PTX and TCX analysis with a candidate-based functional approach we identified two calcium binding
proteins S100A8 and S100A9 as well as galactin-3-binding protein LGALS3BP as significantly upregulated in the CSCs population.
In silico pathway analysis followed by RNA interference mediated functional analysis revealed the critical role of S100A8, S100A9
and LGALS3BP as molecular determinants of CSCs proliferation, migration and in vivo tumor initiation. Our study highlights the
power of combining unbiased -omics approaches with focused functional analyses for the identification of key regulators of
CSCs, an approach that warrants further application to identify CSCs proteins amenable to drug targeting.
P134
Integrated OMICS reveal organ-specific proteome deterioration during aging in rat
1
2
3
1
3
2
A. Ori , B. Toyama , M. Harris , T. Bock , N. Ingolia , M. Hetzer , M. Beck
1
EMBL, Heidelberg, Germany
2
Salk Institute, La Jolla, United States
3
University of California, Berkeley, United States
1
Aging is a multifaceted process that occurs simultaneously at the genomic, proteomic and organism level. Genetic screens and
OMICS studies have previously identified various factors involved in aging, however, little is known about how the molecular
events occurring at the different regulation levels integrate within complex biological systems to ultimately cause organ-specific
deterioration. To systematically study the molecular basis that makes old individuals distinct from young, we combined OMICS
techniques to analyze brain and liver in aging rat. We measured changes in transcript abundance and translation rates by next
generation sequencing and ribosome profiling, and determined alterations of protein abundance and post-translation
modifications by shotgun proteomics. The integration of genomic and proteomic data obtained from the same set of samples
allowed us to identify a previously unappreciated relationship between changes in translation output and protein abundance
that occur during aging. We also show that changes in protein localization, levels of protein phosphorylation and alternative
splicing act as secondary mechanisms. Our integrated systems analysis points to alterations of mediators of neuronal
communication and protein kinases in brain and modification of metabolic processes in liver, and reveals that specific protein
complexes are modulated in their abundance and composition during aging. We describe hundreds of previously unknown
alterations, which build a rich resource for the scientific community to further study the molecular mechanisms underlying
aging.
Proteomic Forum 2015
90
P135
Deorphanizing the matrix metalloprotease MT5-MMP via proteomic substrate identification in the secretome
1,2,3
1,2,3
J. Herber , S. Lichtenthaler
1
German Center for Neurodegenerative Diseases, Neuroproteomics, Munich, Germany
2
Technical University Munich, Munich, Germany
3
Munich Cluster for Systems Neurology, Munich, Germany
As a membrane-type matrix metalloprotease (MT-MMP) MT5 (also knows as MMP-24) can cleave its substrates just outside of
the membrane and thus can engage in ectodomain shedding of membrane proteins. This cell surface proteolysis of membrane
proteins has an essential role in cellular processes as diverse as cytosolic signalling, transcriptional activation, and
communication between cells but is also linked to diseases, such as cancer, inflammatory responses and Alzheimer’s disease
(AD).
MT5 is primarily expressed in the nervous system and only shows low expression levels in other tissues. It is known as an
essential mediator of inflammation in the peripheral nervous system, where it is involved in neuropathic pain and in neurite
outgrowth of sensory peripheral neurons. This indicates a potential pathophysiological relevance of MT5 proteolytic activity on
brain proteins in the central nervous system. For example, in post-mortem brain tissue of AD patients, a co-localization of MT5
immunoreactivity with amyloid plaques was already demonstrated. The specific role and mode of action of MT5 in AD is still
unclear and needs to be characterized further. Moreover, the substrates and thus the detailed molecular function of MT5 are
largely unknown.
Here, we applied the proteomic SPECS (secretome protein enrichment with click sugars) technique and studied the secretome of
primary wildtype and MT5 knockout neurons to identify MT5 substrates. As a result of cell surface proteolysis, ectodomains of
membrane proteins are released into conditioned medium, which makes this compartment an attractive source to identify
membrane protein substrates. SPECS selectively enriches glycosylated proteins, present in cell culture supernatants, and thereby
allows performing a proteome-wide identification of secreted proteins from cells, including primary cells. Membrane proteins
whose ectodomains are significantly reduced in the MT5 knockout secretome are considered as candidate MT5-MMP
substrates.
This study focuses on the characterization of MT5-MMP to elucidate its biological context of action and its involvement in
physiological and thus pathophysiological mechanisms.
References: Kuhn, P. H. et al. Secretome protein enrichment identifies physiological BACE1 protease substrates in neurons.
EMBO J 31, 3157-3168 (2012).
Folgueras, A. R. et al. Metalloproteinase MT5-MMP is an essential modulator of neuro-immune interactions in thermal pain
stimulation. Proc Natl Acad Sci USA 106, 16451-16456 (2009).
P136
Ribosomal proteins generate protein signatures in health & disease
1
1
1
1
1
1
2
3
1
2
A. Friedrich , C. Brandl , T. Karl , B. Wimmer , P. Radler , J. Teufl , A. Klausegger , S. Klauck , M. Breitenbach , H. Hintner , J. W.
2
1
Bauer , H. Breitenbach-Koller
1
University of Salzburg, Dept. of Cell Biology, Salzburg, Austria
2
SALK, Department of Dermatology, Paracelsus Private Medical University Salzburg, Salzburg, Austria
3
German Cancer Research Center (DKFZ), Division of Molecular Genome Analysis, Heidelberg, Germany
Introduction: Proteomics has arrived in functional analysis of gene expression. A first map of the human proteome has been
published (1) and it becomes apparent that disease associated genetic lesions on a cellular level can now be characterized by
differentially expressed protein signatures. This elevates the question of genetic cause and cellular consequence to a level of
analysis with the potential to identify regulatory hot spots within differentially expressed protein signatures. Therefore
innovative concepts targeting these hot spots, modulating selectively the expression level of an individual disease mRNA or
modulating complete protein signatures.
Objectives: We aimed to develop a specialized ribosome screen, which is able to identify individual ribosomal proteins (RPs)
which selectively increase the expression of reporter mRNAs/disease mRNAs while leaving bulk translation unaltered.
Proteomic Forum 2015
91
Furthermore we wish to investigate differential protein expression in autistic patients (ASD) and patients with intellectual
disability (ID).
Patients & Methods: Mutations in the LAMB3 gene, inducing one of the most severe forms of blistering skin disease
Epidermolysis bullosa (EB), were screened for translational repair with a luciferase-based protein expression reporter system,
operating in the presence of different specialized ribosomes. In a second approach, comparative proteome analysis of patients
with ASD and ID was performed.
Results: We have identified individual RPs, which selectively increase readthrough of mutant LAMB3-PTC (premature
termination codon). Topologically they line the protein exit tunnel, a domain of the ribosome, known to respond to
manipulation for PTC readthrough (2).
In a complementary investigation, we identified differentially expressed protein signatures in ASD patients with and without
mutations in RpL10/RpuL16, as well as in patients with ID. The protein signatures for the ASD patients form three co-expression
networks, operating in energy-, redox- and protein metabolism. Interestingly the ASD signatures in a subset overlap with the ID
signatures in three enzymatic core functions of the ASD networks (3).
Conclusion: We conclude that RPs are promising targets for therapeutic interventions to repair protein expression of human
disease mRNAs. Specifically small molecules can be employed that, by binding to RPs, alter their ribosomal functions to
favorably increase (decrease) protein production in a given disease state. Furthermore we propose that monitoring of protein
signatures observed in the presence of a given genetic lesion will inform on the cellular consequences both of the lesion and a
therapeutic intervention.
We gratefully acknowledge DEBRA Austria for funding this project.
(1) Wilhelm et al., 2014, doi:10.1038/nature13319
(2) Bauer et al., 2013, PLoS One 8, e67609
(3) Chiocchetti et al., 2014, Molec. Autism 5:10
P137
Investigations into the mechanism of action of the neurotrophic drug Cerebrolysin
1
2
1
F. Füssl , S. Winter , C. Huber
1
University of Salzburg, Molecular Biology, Salzburg, Austria
2
Ever Neuro Pharma GmbH, Unterach, Austria
Cerebrolysin is a neurotrophic drug used for the treatment of dementia, stroke and traumatic brain injury. Already available for
6 decades, the drug derived from enzymatic breakdown of porcine brain is known to be a complex mixture of amino acids and
neurotrophic peptides which are considered to trigger the drug’s neuroprotective and neuroregenerative effects. Despite the
availability of various clinical data and a remarkably good safety profile, the exact mechanism of action is still incompletely
understood. To encounter this situation we are focusing on gaining experimental information on multiple stages, by analysing
the drug itself as well as treating cells with Cerebrolysin and observing the effects on multiple molecular levels. A subsequent
integration of the data achieved may offer novel insights into the molecular events behind the observable effects of
Cerebrolysin.
For generating a list of peptides, 3 different batches of Cerebrolysin were analyzed in quintuplicate. Each sample was measured
by the application of a capillary ion-pair reversed-phase nano-HPLC system hyphenated to a quadrupole-Orbitrap mass
spectrometer. A polystyrene-divinylbenzene monolithic column was used for separation and a 120 min gradient of acetonitrile
in 0.050 % trifluoroacetic acid was applied. Peptide identification was forced by tandem mass spectrometry and the application
of the Thermo Proteome Discoverer 1.4. For studying the drug’s effect, human neuronal progenitor cells (NTERA-2) were treated
for six durations in triplicate ranging from 1 up to 28 days. After cell harvest and lysis, protein, metabolite and RNA contents
were extracted. The protein samples were labelled with Tandem Mass Tags (TMT) and subsequently analyzed in triplicate by the
application of exclusion lists using a capillary C18 column and a linear gradient of acetonitrile in 0.1 % acetic acid. Data analysis
was conducted with the OpenMS software [1] and data interpretation is accomplished by the application of the Ingenuity
Pathway Analysis Software (IPA) [2]. RNA analyses were performed using the Affymetrix GeneChip microarray technology.
For each of the 15 Cerebrolysin samples analyzed, the majority of identified peptides were found in all runs or at least in all 3
batches. The analysis of the proteomics samples revealed a total number of over 6,000 identified proteins of which 463 showed
Proteomic Forum 2015
92
significant regulation. The top molecular networks represented by the proteomics and transcriptomics data are involved in cell
growth, cell death and survival, cell signalling and inflammation. Moreover several molecules were shown to be significantly
regulated that are known to participate in neuroregenerative or -degenerative processes.
References:
[1] Sturm, M., et al., “OpenMS - an open-source software framework for mass spectrometry”, BMC Bioinformatics, 163 (1), 2008
[2] IPA ®, QIAGEN Redwood City, www.quiagen.com/ingenuity
P138
Effect of ascitic fluid from ovarian cancer patients on the proteome of transformed cells
1
1
1
1
2
A. Toledo-Leyva , J. C. Villegas-Pineda , J. C. Osorio-Trujillo , V. Hernández-Ramírez , D. Gallardo-Rincón , P. Talamás-Rohana
1
CINVESTAV-IPN, Departamento de Infectómica y Patogénesis Molecular, Distrito Federal, Mexico
2
Instituto Nacional de Cancerología, Departamento de Oncología Médica, Distrito Federal, Mexico
1
Introduction: Ovarian cancer is a very complex disease difficult to detect, diagnose and treat. About 30% of the patients in
advanced stages of the disease produce significant amounts of ascitic fluid that correlates with a bad prognostic for the patient
and raising the mortality rates. Our research group is interested in characterizing the effect that this ascitic fluid produce on the
proteome of ovarian cancer (OvCa) transformed cells to propose new biomarkers and/or therapeutic targets that may help to
control this disease.
Objective: To analyze the effect of the ascitic fluid of OvCa patients on the proteome of the SKOV-3 cell line.
Material and methods: The SKOV-3 cell line was treated, individually with 33 ascitic fluid samples from OvCa patients in
different clinical stages of the disease. Then, total extract from the respective cell cultures were analyzed by SDS-PAGE. To learn
about specific changes in the protein expression pattern, a monoclonal antibody against integrin-like molecules (3C10, Sengupta
et al., 1998) was used for western blot and immunofluorescence assays. Also some biopsies from tumoral tissues were analyzed
with this 3C10 MAb.
Results: The ascitic fluid from OvCa patients induced several changes in the SKOV-3 cell line, at the morphological level and at
the proteomic level. In 80% of the treated cultures, an approximately 25 kDa protein was detected with the 3C10 MAb. This
protein was also present in OvCa biopsies.
Conclusion: Ascites from OvCa patients are able to induce changes in the proteomic profile of cells; particularly a 25 kDa protein
with unidentified identity and function that is present only in the condition that resembles the one in the microenvironment of
the disease.
P139
TXNL1 characterize aneuploid subtypes of colorectal and endometrial cancer
1
1
2
3
4
5
4
1
6
T. Gemoll , U. Roblick , S. Szymczak , T. Braunschweig , S. Becker , U. Hellman , C. Lundgren , J. Lahmann , M. Difilippantonio ,
6
7
4
1
T. Ried , H. Jörnvall , G. Auer , J. Habermann
1
University of Lübeck, Section for Translational Surgical Oncology & Biobanking, Department of Surgery, Lübeck, Germany
2
University of Lübeck, Institute for Medical Biometry and Statistics, Lübeck, Germany
3
University Clinic RWTH Aachen, Institute for Pathology, Aachen, Germany
4
Karolinska Institutet, Karolinska Biomic Center, Stockholm, Sweden
5
Ludwig Institute for Cancer Research, Uppsala, Sweden
6
Genetics Department, Center for Cancer Research, NCI/NIH, Bethesda, United States
7
Karolinska Institutet, Department of Medical Biochemistry and Biophysics, Stockholm, Sweden
Background: DNA aneuploidy has been identified as prognostic factor for epithelial malignancies. Further understanding of the
translation of DNA aneuploidy into protein expression will help to define therapies, prognosis and prevention. We therefore
aimed at identifying aneuploidy-associated protein expression in colorectal and endometrial cancer.
Methods: We compared two-dimensional gel electrophoresis-based protein expression patterns of diploid and aneuploid
colorectal and endometrial carcinoma. Proteins were identified by mass spectrometry. TXNL1 as one target protein was
validated by Western Blot and clinically evaluated by immunohistochemistry using a tissue microarray (TMA).
Proteomic Forum 2015
93
Results Two independent statistical analyses revealed 64 and 121 ploidy-associated protein spots that were differentially
expressed between the diploid and aneuploid colorectal cell lines and diploid and aneuploid endometrial samples, respectively.
68 protein spots could be identified by mass spectrometry, whereupon an up-regulation of TXNL1 in diploid samples was found
in both colorectal cell lines and endometrial cancer tissue. TXNL1 was significantly validated also in clinical samples discerning
aneuploid from diploid colorectal carcinomas (p < 0.05).
Conclusion Distinct protein expression patterns discern aneuploid from diploid colorectal and endometrial carcinomas. TXNL1
expression alterations characterize aneuploid subtypes of colorectal and endometrial cancers that are known to present poor
prognosis.
P140
MALDI-imaging reveals TYB4 to classify diploid from aneuploid colon cancer
1
1
1
2
1
T. Gemoll , S. Strohkamp , K. Schillo , C. Thorns , J. Habermann
1
University of Lübeck, Section for Translational Surgical Oncology & Biobanking, Department of Surgery, Lübeck, Germany
2
University Medical Center Schleswig-Holstein, Institute of Pathology, Lübeck, Germany
Background: DNA aneuploidy has been identified as a prognostic factor for epithelial malignancies. In this study, we compared
diploid and aneuploid colon cancer tissues against normal mucosa of the colon by means of matrix-assisted laser
desorption/ionization (MALDI) imaging mass spectrometry (IMS).
Material and Methods: DNA image cytometry determined the ploidy status of tissue samples that were subsequently subjected
to MALDI-IMS. After obtaining protein profiles through direct analysis of tissue sections, a discovery and a validation set were
used to predict ploidy and disease status by applying proteomic classification algorithms [Supervised Neural Network (SNN) and
Receiver Operating Characteristic (ROC)]. Clinical target validation was performed by immunohistochemistry using tissue
microarrays (TMA) comprising healthy controls as well as diploid and aneuploid colorectal carcinomas.
Results: SNN algorithm categorized 99% of normal mucosa and 90% of colon carcinoma as well as 99% of diploid and 94% of
aneuploid colon cancers correctly. Validation of both comparisons showed a correct classification of normal mucosa in 92%,
tumors in 96%, and diploid and aneuploid colon cancers in 92% and 78%, respectively. Five peaks (m/z 2,396 and 4,977 for the
diploid vs. aneuploid comparison and m/z 3,375, 6,663, 8,581 for the normal mucosa vs. carcinoma comparison) reached
significance in both SNN and ROC analysis. Among these, m/z 4,977 was identified as thymosin beta 4 (TYB4). TYB4 showed
expression differences also in clinical samples using a tissue microarray of normal mucosa, diploid and aneuploid colorectal
carcinomas and could serve to predict overall survival.
Conclusion: Our data underscore the potential of MALDI-IMS proteomic algorithms to reveal significant molecular details from
distinct tumor subtypes such as different ploidy types.
P141
A Data Independent Strategy for a Multi-Omic Approach to Investigate Obesity Treatment within a Mouse Model
1
1
2
3
3
2
2
G. Kramer , N. Dekker , L. A. Gethings , V. Lee , R. Beynon , J. I. Langridge , J. P. Vissers , J. M. Aerts
1
Academic Medical Centre, Amsterdam, Netherlands
2
Waters Corporation, Health Sciences, Wilmslow, United Kingdom
3
University of Liverpool, Centre for Proteome Research, Liverpool, United Kingdom
1
Obesity is one of the risk-factors associated with metabolic syndrome, causing excess body fat to be accumulated to the extent
that it adversely affects health and life expectancy. It has previously been demonstrated that glycosphingolipids play a crucial
part in such metabolic syndromes. The manipulation of the function of glycosphingolipids with small molecule drug compounds
within mouse models has shown that symptoms can be negated. Knowledge relating to the proteome, metabolome and
lipidome during development is still to be fully explored. The work presented here is to provide a multi-omic analysis of protein
and lipid liver extracts from control and obese mouse models undergoing treatment to prevent or revert obesity.
Lipid and protein extracts were generated from liver tissue originating from 3 control and 3 obese mice models. Protein extracts
were proteolysed with trypsin and the resulting peptides separated over a 90 min linear RP nanoscale LC gradient, whilst the
extracted lipids were separated over a 20 min RP LC gradient. Data were acquired using a DIA approach, whereby the collision
energy was switched between a low and elevated energy state during alternate scans. Proteomic acquisitions also utilized IM in
Proteomic Forum 2015
94
the acquisition scheme. The acquired data were processed and searched using Progenesis QI and dedicated protein sequence
and lipid compound databases, providing normalized label-free quantitation results for both datasets.
Proteomic samples were based on 100 ng loadings and analyzed as triplicate technical replicates in a randomized order.
Processing and searching the data resulted in over 1250 proteins being identified, across all technical replicates and biological
conditions. Over 70 proteins exhibit a fold change greater than 2 with significant analysis of variance. Unique peptides were
used for relative quantitation with median abundance normalization performed across all samples. Lipid extracts were prepared
using 500 µL IPA/water (50:50), of which 2 µL were injected on-column and analyzed in triplicate. Samples were also acquired in
a random order with a QC comprised of all samples in equal amounts and injected every 5 injections. Lipid data analysis was
conducted with Progenesis QI used for processing and searching. Interrogation of the data revealed over 500 potential
identifications for combined positive and negative ion acquisitions with mass errors less than 2 ppm. Compound searches
provided a range of lipid classes including free fatty acids, ceramides, triglycerides, sphingomyelins and glycosphingolipids.
Unsupervised multvariate analyses showed clear distinction between obese and control groups in both proteomic and lipidomic
experiments. Pathway analysis tools were used to review the complimentary datasets and hence provide an understanding of
the underlying biology of differentially expressed proteins and lipids.
P142
New Proteomic Biomarkers for Psoriasis and Psoriasis Arthritis
1
1
1,2
1
1
1
3
1
1
3
R. Büchler , J. Reindl , P. Muckova , S. Wendler , S. Opitz , T. Krüger , N. Krieg , J. Pesek , S. Nemitz , J. Norgauer , H. Rhode
1
Universitätsklinikum Jena, Biochemie I, Jena, Germany
2
Universitätsklinikum Jena, Hans Berger Department of Neurology, Jena, Germany
3
Universitätsklinikum Jena, Klinik für Hautkrankheiten, Jena, Germany
1
Introduction: Psoriasis (Ps) is a chronic, immune-mediated systemic disease that impairs skin and causes several comorbidities.
It affects 2-4 % of the population worldwide. 20-30 % of Ps-patients develop psoriatic arthritis (PsA), in part years before skin
signs. Yet, despite several nominations no biomarker for diagnosis of Ps and PsA has been validated for reliable use in clinical
practice.
Objectives: The aim of this study is to look for biomarkers in blood plasma of patients with Ps and PsA suitable for diagnostics.
These still unavailable parameters should be identified by a comprehensive and precise proteomic approach.
Material & Methods: Samples: Blood samples were obtained by SOP and building a comprehensively characterized biobank.
Inclusion criteria were age >18 years and clinical diagnosis of Ps vulgaris according to [1]. All individuals were examined for PsA
by fluorescence imaging.
Proteomic search: Six couples of EDTA-plasma from patients and controls were selected and applied to a native 2-dimensional
fractionation in microplates [2]. From the resulting ~4000 fractions per sample in total about 3000 fractions with high protein
content were selected for analysis. After tryptic digestion we subject the fractions to capillary UHPLC-LTQ-Orbitrap tandem mass
analysis. The obtained data are processed by Proteome Discoverer, Sieve 2.0 (Thermo Scientific), and MetaboAnalyst 2.0.
Validation: Some biomarker candidates were quantified by ELISA using 8-23 new samples per group. These concentrations have
been statistically analyzed.
Results: Proteomic results: In summary, up to 130 increased and 160 decreased biomarker candidates have been obtained in
each couple. Out of them 109 candidates are considered as valuable with respect to available literature data. Yet only fifteen
could be verified preliminary by immunoquatitation.
Validation: Out of the verified biomarker candidates, four individual proteins, fourteen combinations with two and one with
three parameters show significant discrimination between the groups and some of them strong correlation with several clinical
data.
Conclusion: The application of an automated multidimensional analytical platform for high-throughput proteomic sample
preparation yields valuable biomarker candidates for non-invasive diagnostics of Ps and PsA.
[1] W. Taylor, et al. CASPAR Study Group. Arthritis Rheum. (2006) 54, 2665-73.
Proteomic Forum 2015
95
[2] S. Wendler, et al., in: Automation Systems of the 21th Century. Nova Science Publishers, NY, 2013, ISBN: 978-1-62948-262-0,
Chapter 1, 1-50
P143
Acute respiratory infection - biomarker search in an animal model disease
1,2
1
1
3
1
P. Muckova , R. Buechler , S. Wendler , P. Reinhold , H. Rhode
1
University Hospital , Institute of Biochemistry 1, Jena, Germany
2
University Hospital , Hans-Berger-Klinik für Neurologie, Jena, Germany
3
Friedrich-Loeffler-Institute (Federal Research Institute for Animal Health), Institute of Molecular Pathogenesis , Jena, Germany
Introduction: Chlamydia psittaci (Cp) causes psittacosis which is transmitted from birds to mammals. Human psittacosis, also
known as parrot fever, manifests in respiratory symptoms, headache, myalgia etc. and can result in liver and neurological
1
dysfunctions. Cp infection of cattle is a good model of an acute respiratory infection in humans suitable to study the
aetiopathology and to search for biomarkers.
Objectives: Our goal is the search for proteomic biomarkers of iatrogenic psittacosis in blood plasma. Comparing the protein
profiles of three samples per calf, before (healthy), 24h, and 48h after bronchial infection we look for proteins which could help
to explain the severe systemic disease, to identify non-antibiotic therapeutics, and which could be used as early and specific
diagnostic markers of infection.
Materials and Methods: A 2-dimensional fractionation method is applied combined with mass spectrometry analysis which is
able to identify the complete proteome of plasma. Our workflow starts with a native 1D-size exclusion chromatography followed
2
by a semiautomatic parallelized 2D-anion exchange chromatography in microplates . So, each of the 1 ml plasma (EDTA)
samples of 6 biological replicates yield ~ 4000 2D-fractions that are analyzed in technical duplicates. After reduction, IAA
alkylation and tryptic digestion we subject the fractions to capillary UHPLC-LTQ-Orbitrap tandem mass analysis. The obtained
data are processed by Proteome Discoverer and Sieve 2.0 (Thermo Scientific).
Results: So far, we were able to identify 1151 non-redundant proteins within one bovine serum sample. At that time (2012) the
available bos taurus database entries in UniProtKB with ruminants were much less than for humans. We will present data on the
whole proteome profile of bovine EDTA-plasma obtained by our method and the currently available database. Moreover,
concise differential expressed plasma constituents identified by multivariate statistical analysis of expression profiles in samples
of the three time points will serve as biomarker candidates. This work is still under progress.
Conclusions: The applicability, comprehensiveness, and high resolving power of the automated multi-dimensional analytical
platform have been shown in several diseases. It is applied here to search for biomarker candidates of Cp infection of cattle.
Inevitably, subsequent validation has to prove the applicability in veterinary diagnostics.
1
P. Reinhold, et al., PlosOne (2012), 7, e30125
S. Wendler, et al., in: Automation Systems of the 21th Century. Nova Science Publishers, NY, 2013, ISBN: 978-1-62948-262-0,
Chapter 1, 1-50
2
P144
Characterisation of regenerative phenotype of unrestricted somatic stem cells from human umbilical cord blood by functional
secretome analysis
1
1
2
3
4
1
5
3
K. Stühler , J. Schira , H. Falkenberg , M. Hendricks , G. Kögler , D. Waldera-Lupa , H. E. Meyer , H.- W. Müller
1
Heinrich Heine University, Molecular Proteomics Laboratory, Düsseldorf, Germany
2
HHU, MPL, Düsseldorf, Germany
3
Heinrich Heine University, Molecular Neurobiology Laboratory, Düsseldorf, Germany
4
Heinrich Heine University , Institute for Transplantation Diagnostics and Cell Therapeutics, Düsseldorf, Germany
5
Leibniz-Institut für Analytische Wissenschaften, Biomedical Research, Dortmund, Germany
Introduction: Stem cell transplantation is a promising therapeutic strategy to enhance axonal regeneration after spinal cord
injury (SCI). Unrestricted somatic stem cells (USSC) isolated from human umbilical cord blood is an attractive stem cell source
which could be isolated with GMP grade without ethical concerns. It has been shown that USSC transplantation into acute
Proteomic Forum 2015
96
injured rat spinal cords leads to axonal regrowth, enhanced tissue sparing and significant locomotor recovery, but without
differentiation. Instead, USSC secrete trophic factors which enhance neurite growth of primary neurons in vitro.
Objectives: Characterization of the regenerative phenotype of unrestricted somatic stem cells with a workflow comprising
secretome analysis by quantitative mass spectrometry and neurite outgrowth in vitro assays.
Materials & Methods: Serum-free USSC secretome was analyzed by mass spectrometry (LTQ Orbitrap Velos or LTQ Orbitrap
Elite mass spectrometer) after TCA precipitation or lyophilization, 1D SDS-gel-electrophoresis and trypsin digestion. Proteins
were identified using proteome discoverer (Version 1.4, Thermo Fisher Scientific) including Mascot search engine (Version 2.4.1,
Matrix Science). Predictions of secreted proteins have been performed with SignalP, SecretomeP algorithms and UniProt
database. Neurite outgrowth assays with primary rat cortical neurons were performed to validate candidate proteins.
Results: With our proteomic approach, we identified 1278 USSC secreted proteins including 385 proteins secreted classically,
471 non-classically secreted proteins and 360 proteins predicted to be localized extracellular. These proteins were assigned to
Gene Ontology biological processes including blood vessel formation, regulation of cell motion and cell adhesion. Moreover,
USSC express proteins influencing precursor cell differentiation and the immune response. Altogether, these processes have to
be tightly regulated after SCI and seems to be influenced by transplanted USSC. Additionally, comparison with literature
revealed that USSC express at least 33 well-known neurite growth promoting factors, e.g.,Neuregulin-2/-3, Neuronal growth
regulator 1, NDNF, PEDF and SPARC. Subsequent performed functional validations with primary neuronal cell cultures suggest
that SPARC and PEDF are involved in USSC mediated neurite growth whichlikely to be involved in significant locomotor recovery
after transplantation thereby supporting the concept of paracrine functions of USSC in regeneration of spinal cord.
Conclusion:Secretome analyses with subsequent functional validation reveal a powerful tool to identify factors involved in
neurite growth and neural regeneration processes.
P145
Identification of Low Molecular Weight Biomarkers of Prostate Cancer by 2-D Gel Electrophoresis
1
J. Alruwaili
1
Northern Border University, Biochemistry, Arar, Saudi Arabia
Proteins biomarkers are used for diagnosis, screening and predicting therapy responses. 2-D gel electrophoresis (2DPAGE) was
used to examine the proteome of fresh radical prostatectomy specimens from 29 malignant and 10 benign tumor samples. The
methodology was to identify proteins present in the serum of Prostate Cancer patients. It was noticed that Albumin and
Immunoglobulins pose a major obstacle in the success of 2D‐PAGE for the identification of candidate proteins. So, combinatorial
peptide ligand library (CPLL) approach was used prior to 2DPAGE to identify less abundant proteins. This approach was
successful in isolating four spots in the mass range <14 kDa.
Proteomics in Biotechnology (P192–P198)
P192
Evaluating sample preparation for mesenchymal stem cell secretome analysis - a 2D gel electrophoretic approach
1
2
2
1
A.- K. Koch , V. Charwart , C. Kasper , M. Marchetti-Deschmann
1
Technical University of Vienna, Institute of Chemical Technologies and Analytics, Vienna, Austria
2
University of Natural Resources and Life Science Vienna, Institute of Applied Microbiology, Vienna, Austria
Question: Normoxia and hypoxia are tools to control cell differentiation in biotechnology [1]. In the present study we are
focusing on the
1) characterization and detection of changes in the secretome of mesenchymal stem cells
2) analysis of low abundant secretome proteins after removal of high-abundance proteins from the cell medium (affinity
enrichment and depletion)
3) comparison of the secretome after normoxic and hypoxic cultivation using a DIGE approach
Methods: Mesenchymal stem cells were cultivated under different oxygen supply levels (5%, 21% oxygen) in a bioreactor in fetal
calf serum. 500mL samples were collected and stored at -20°C. Proteins were precipitated (trichloroacetic acid/acetone (1:8))
Proteomic Forum 2015
97
and dissolved in IPG-buffer (7M Urea, 2M Thiourea, 2% CHAPS). High-abundance protein depletion and low-abundance protein
enrichment were applied using a combinatorial peptide library (ProteoMiner, Bio Rad) and depletion columns (Top 12, Pierce).
Proteins were separated on 12.5% gels (HPE Tower, 255x200x0.65 mm, Serva), and visualized by fluorescence (Serva Purple,
Serva) and silver staining. For DIGE, samples were minimal labeled with G-Dyes (NH DyeAGNOSTICS) and separated the same
way.
Results and Conclusion: For method evaluation protein precipitation and 2D-GE had to be established. Sample preparation
could be shown to be highly reproducible. However, it was difficult to detect and to visualize low-abundance proteins relevant
for secretome analysis. Therefore, high-abundance serum proteins had to be significantly removed without losing valuable
secreted proteins. By using Top12 depletion spin columns an effective depletion of serum proteins was achieved. An efficient
reduction of high-abundance proteins was observed resulting in a reduced number of protein bands in 1D and a lower number
of spots in 2D PAGE. The high possibility of co-depletion of low abundant proteins was considered to be very likely. Therefore
the combinatorial peptide library was also tested, but showed a lower efficiency for serum protein removal. Serum albumin was
still observed, however less concentrated. We assumed that co-depletion is reduced under these conditions and a more
complete coverage of the secretome will be achieved. We therefore further enhanced and evaluated the latter approach to
show reproducibility and robustness of sample preparation. A reliable sample preparation was achieved for both oxygen supply
levels. Finally DIGE analysis was introduced to point out proteome differences between the normoxic and hypoxic secretome of
mesenchymal stem cells. Overall it can be said that both methods efficiently reduced high abundance serum proteins giving
access to the interesting secretome of mesenchymal stem cells.
Future work will focus on the identification of differentially expressed proteins by mass spectrometry.
References: [1] Skalnikova-Kupcova H., et al. Biochimie 95 (12), 2196-2211 (2013)
P193
Comparative proteomic analysis of human osteosarcoma and pulmonal metastasis
1
1
1
1
1
2
3
4
1
5
T. Gemoll , F. Epping , L. Heinrich , B. Fritzsche , U. Roblick , S. Szymczak , S. Hartwig , R. Depping , H.- P. Bruch , C. Thorns , S.
3
6
1
Lehr , A. Paech , J. Habermann
1
University of Lübeck, Section for Translational Surgical Oncology & Biobanking, Department of Surgery, Lübeck, Germany
2
University of Lübeck, Institute for Medical Biometry and Statistics, Lübeck, Germany
3
German Diabetes Center (DZD), Institute of Clinical Biochemistry and Pathobiochemistry, Düsseldorf, Germany
4
University of Lübeck, Institute for Physiology, Lübeck, Germany
5
University Medical Center Schleswig-Holstein, Institute of Pathology, Lübeck, Germany
6
University Hospital Schleswig Holstein, Department of Orthopaedics and Traumatology, Lübeck, Germany
Background: Cancer proteomics provide a powerful approach to identify biomarkers in screening for alterations in protein levels
and posttranslational modifications that are associated with tumors. Particularly, biomarkers for early detection, prognosis and
therapeutic intervention of bone cancers, especially osteosarcomas, are missing.
Material and Methods: Protein expression pattern between cell lines of fetal osteoblasts, osteosarcoma and pulmonal
metastasis derived from osteosarcoma were compared using two-dimensional gel electrophoresis (2-DE). Mass spectrometry
served for identification of differential expressed protein spots. Target validation was performed by Western Blot on cell lines
and by immunohistochemistry using tissue microarrays (TMA) on clinical samples, respectively.
Results: Statistical analysis revealed 34 differentially expressed protein spots (p<0.05). 17 (50%) Proteins were identified by
mass spectrometry that interact in pathways of Gene Expression, Cell Death and Cell-To-Cell Signaling and Interaction. Ran/TC4binding protein (RANBP1) and Cathepsin D (CTSD) were validated by Western Blot in cell lines while the latter one showed
higher expression differences also in clinical samples using tissue microarrays.
Conclusion: This study showed significant differences in protein expression between fetal osteoblasts, osteosarcomas, and
pulmonal metastasis. Particularly, CTSD was validated in clinical material and could trigger malignant transformation in bones.
P194
A comparative study on the quantification of non-biological protein modifications using mass spectrometry
Proteomic Forum 2015
1
1
98
1
E. Soher , C. Stephan , M. Marchetti-Deschmann
1
Vienna University of Technology, Institute of Chemical Technologies and Analytics, Vienna, Austria
Introduction: Synthetically modified proteins can be used in a wide range of biotechnological applications. Since it is always a
goal to produce modified proteins at the lowest possible cost, it was of interest whether different synthesis strategies can lead
to a reduction in the use of chemicals while the same results in terms of modification efficiency on the protein can be achieved.
Objectives: Therefore the goal was to establish a method to quantify artificially introduced protein modifications.
Materials and Methods: Conalbumin (CON) was modified in various syntheses to carry different amounts of the mycotoxin
zearalenon (ZON). To gain information on the modification localization sites, the proteins were digested and analyzed by LCMS/MS. For digestion the proteins were reduced and alkylated. Trypsin (Thermo Scientific) and a Trypsin/LysC Mix (Promega)
were used as proteases. The samples were desalted using C18 spin columns (Thermo Scientific). Detailed protein
characterization was carried out on a MALDI-TOF/TOF instrument (UltrafleXtreme, Bruker) and a nLC-ESI-Iontrap instrument
(Ultimate 3000, Dionex; Acclaim PepMap100 column, Thermo Scientific; Esquire HCT, Bruker) performing CID fragmentation for
peptide identification and modification localization. Method development for quantification was performed on an UPLC-ESITriplequadrupole instrument (Nexera System, LC-MS-8030plus, Shimadzu; Acquity UPLC Peptide BEH column, Waters). Biotools
(Bruker) was used to evaluate MALDI-TOF/TOF and nLC-Iontrap data, LabSolutions (Shimadzu) was used for UPLC-ESI-MS/MS
results.
Results: In order to maximize the protein sequence coverage, protein digestion was optimized. Sole trypsin was compared to a
trypsin/LysC mix. Latter gave better results showing 48.3 % sequence coverage in comparison to 33.2 % for trypsin. Finally it was
possible to identify modification sites and it was found that most of the protein modifications were located on the protein’s
surface. Furthermore it was possible to clearly distinguish between samples carrying a high number of modifications and those
showing little to no modifications at all. According to the synthesis, ZON modifications were most likely to be present on Lys, but
also possible for Arg, Ser or Thr. Modifications were not only found on Lys but also on Arg or Ser. Yet, MS/MS spectra gave
insight on the fragmentation mechanisms of ZON modified peptides, exhibiting indicator fragment ions which were further used
to develop a quantification method applying triplequadrupole MS technology. MRM, neutral loss scan and precursor ion scan
were compared in respect to quantification efficiencies.
Conclusion: It can be said that a targeted LC-MS/MS approach is a quick and reliable tool to assess the modification level of
peptides.
Acknowledgement Modified protein was provided by Romer Labs (B. Cvack, G. Häubl and P. Ansari)
P195
Cattle subfertility and bacterial infection: uterine epithelial cells response after LPS exposure
1
1
2,3
4
1
2,3
4
1,5
C. Piras , A. Soggiu , V. Greco , Y. Guo , L. Bonizzi , A. Urbani , P. Humblot , P. Roncada
1
University of Milan, Department of Veterinary Science and Public Health, Milan, Italy,
2
Fondazione Santa Lucia – IRCCS, Rome, Italy
3
Università degli Studi di Roma “Tor Vergata”, Dipartimento di Medicina Sperimentale e Chirurgia, Rome, Italy
4
Swedish University of Agricultural Sciences, SLU, Sweden, Division of Reproduction, Departement of Clinical Sciences, Faculty of
Veterinary Medicine and Animal Science, Uppsala, Sweden, Sweden
5
Istituto Sperimentale Italiano L. Spallanzani, Milan, Italy
Introduction: Uterine infections could represent the concause of cattle infertility. Uterine epithelial cells play e a key role in
protection against bacterial pathogens infection forming a mucosal barrier that avoids their infiltration and promotes the
movement of IgA and IgG from the submucosa to the lumen. This cells express toll-like receptors useful to engage microbes
detection and to stimulate immune-response (Fahey, Wallace et al. 2006).
Objectives: The different response of uterine epithelial cells to a microbial challenge could be the cause of the differences in
cattle fertility and/or susceptibility to uterine infections. This proteomics study will shed light on the mechanisms and suggest
solutions to overcome this problem.
Proteomic Forum 2015
99
Materials & Methods: Bovine epithelial cell cultures were exposed to 8 µg and 16 µg per ml of LPS. Cell number growth rate was
measured. Afterwards, cells were collected and the proteomic profile was evaluated through 2D electrophoresis followed by
MALDI TOF-MS and SHOTGUN-MS approaches (fig. 1).
Results: Cell growth rate was found to be increased according to LPS stimulation in the group challenged with 8 µg LPS in
comparison to control group.
Preliminary proteomic results highlighted the differential protein expression of Interferon-induced dynamin-like GTPase, Protein
disulfide-isomerase A3, and of transketolase that were increased after LPS exposure. Interferon-induced dynamin-like GTPase
plays a potential role in pathogens response has an antiviral activity (Kim, Shenoy et al. 2012), but its role in bacterial invasion
and protection of uterine tissue still needs to be elucidated (Haller and Kochs 2011).
Also the presence of differential expression of Histone H2B could (fig. 2) be related to the mechanism of defense of epithelial
cells against bacterial infection (Liechti and Goldberg 2013).
Conclusion: Obtained results uncover the mechanisms that are behind the response of uterine epithelial cells during bacterial
infection such as oxidative stress, interferon production and cell proliferation. Obtained knowledge suggests new strategies to
improve cattle subfertility.
Figure legend:
Figure 1. Workflow used for differential proteomics evaluation.
Figure 2. Results of Shotgun MS analysis
Aknowledgement
Work supported by EU7FP Prolific Project
Figure 1
Proteomic Forum 2015
100
Figure 2
P196
The Separation and Identification of Glycans Using an Ultra High Resolution Column and an Ultra High Resolution Mass
Spectrometer
1
2
Julian Saba , Kristina Hempel , Madalina Oppermann
1
Thermo Fisher Scientific, United States
2
Thermo Fisher Scientific, Europe
2
Understanding, measuring, and controlling glycosylation in glycoprotein-based drugs and thorough characterization of
biosimilars has become increasingly important. Liquid chromatography (LC) coupled to mass spectrometry (MS) has emerged as
one of the most powerful tools for the structural elucidation of glycans. Here we present a novel stationary phase that provides
superior selectivity and ultra fast resolution, with isomeric separation, as compared to commercially available columns. The
column is based on novel mixed-mode column chemistry, combining both weak-anion exchange (WAX) and reversed-phase (RP)
retention mechanisms. The WAX functionality provides retention and selectivity for negatively charged glycans, while RP mode
facilitates the separation of the same charge according to their isomeric structure, polarity, and size. As a result this column
provides resolution with more than 4 times the glycan structures identified compared to existing commercial columns, for
bovine fetuin N- glycans. The ability to separate isomers reveals a greater complexity of the glycan population from a given
glycoprotein. Namely, far more MS/MS spectra need to be triggered in a single analysis. Additionally, wider dynamic range and
2
sensitivity are needed to detect and generate good quality MS spectra not only for the most abundant glycans but the low
abundant species as well. Orbitrap Fusion with it’s wide dynamic range and ultrahigh mass resolution was selected for looking
deeper into the glycome and confidently identifying low-abundance glycans. Overall, 135 unique glycan structures were
identified using the mixed-mode column and Orbitrap Fusion. To our knowledge this is the largest number of glycans identified
for bovine fetuin in a single analysis. Similar column performance was demonstrated for 2-AA labeled N-glycans from antibodies.
P197
Understanding organic matrix proteins of Oyster Shell in context with Climate Change and Biomineralization
Proteomic Forum 2015
1
1
101
1
A. Upadhyay , D. Ramadoss , V. Thiyagarajan
1
University of Hong Kong, Ecology and Biodiversity, Hong Kong, Hong Kong
Concomitants of Climate Change such as ocean acidification (OA), flooded fresh water and global warming are threatening the
shellfishes. An economically and ecologically important marine invertebrate Oyster also contains shell, a highly adept biomaterial with incredible mechanical properties i.e. >3000 times harder than limestone. Shell, a blend of inorganic crystals (~
95%) and organic matrix (which consists <5 wt% of the shell), is a calcareous exoskeleton which encloses, supports and protects
the soft parts of an animal. It is very vulnerable to be damaged mechanically. Indeed, successive effects were contemplated to
influence organic matrix proteins (OMP), which has recently been observed in OMP of another calcifying animal Mussel (Mytilus
edulis) at the outset. Oyster shells of even evolutionarily closely related Crassostrea species (C. hongkongensis-Hong Kong, C.
angulata-Portuguese) do naturally show different hardness. Specifically, production of OMP to assemble CaCO3 crystals into an
incredibly strong oyster shell may be affected by climate-related stressors, particularly OA.
In our Study, we use quantitative proteomics and bioinformatics tools to understand the response of OMP under projected near
future climate change scenarios and mechanism involved in modulation of bio-mineral properties/shell hardness due to the
altered proteome in these oyster species. After visualizing proteins with SDS-PAGE and identification of proteins by off-gel
fractionation and in-gel digestion methods with the help of MALDI-TOF-TOF Mass Spectrometry, comparative analysis of OMPs
of concerned oyster species is being done using complete genome and shell proteins database of the Pacific oyster (C. gigas) as a
reference database.
Quantitative Proteomics I (P199–P215)
P199
Impact of conventional cigarette smoking and cessation on the lung proteome of ApoE-/- mice quantified by Tandem Mass
Tag (TMT) LC MS/MS approach
1
1
1
1
1
1
1
2
1
1
C. Nury , T. Schneider , S. Dijon , B. Titz , A. Elamin , F. Martin , M. Cabanski , B. Phillips , P. Vanscheeuwijck , N. V. Ivanov , J.
1
1
Hoeng , M. C. Peitsch
1
Philip Morris International, Neuchatel, Switzerland
2
Philip Morris International, Singapore, Singapore
Introduction: Cigarette smoking is a strong causative element to the development of many diseases, including cardiovascular
disease (CVD) and chronic obstructive lung diseases such as COPD/emphysema. Significant health benefits have been associated
with smoking cessation. However, the underlying mechanisms of both disease progression and the impact of cessation are
largely unknown.
Objectives: Using a quantitative proteomic approach based on Tandem Mass Tag (TMT), this study examines both COPD and
-/CVD development in Apolipoprotein E-deficient (ApoE ) mice exposed to either cigarette smoke (CS) or fresh air over an 8
month period. In addition, the reversibility of changes induced by CS exposure was evaluated in a cessation group after 2
months of CS exposure.
Materials & Methods: Mice were exposed to either cigarette smoke (targeting 29.9 µg/L of nicotine, 3R4F) or fresh air for 3
hours per day, five days per week, for up to 8 months. After 2 month of exposure to cigarette smoke (CS), the cessation group
was exposed to fresh air for 6 months. Lung samples were collected at month 1, 2, 3, 6 and 8, proteins extracted from tissues
and labelled using a TMT 6-plex procedure. Tryptic peptides were separated on an Easy nanoLC 1000 instrument connected
online to a Q-Exactive mass spectrometer.
Results: The TMT approach allowed to identify ~5400 and quantify ~ 2000-2300 proteins from mouse lungs.
-/-
An increasing biological effect over time on the lung proteome of ApoE mice exposed to 3R4F was detected as, for example,
highlighted by an increase in number of differentially expressed proteins. The major affected biological functions included an
“immune-system”, a “detoxification of reactive oxygen species (ROS)”, a “metabolism”, a “xenobiotic phase I”, and an “ER
stress” cluster.
Finally, cessation groups showed a rapid decrease in the number of differentially regulated proteins.
Proteomic Forum 2015
102
Conclusions: CS exposure demonstrated a major impact on the murine lung proteome. Most of the differentially expressed
proteins after 2 months exposure appeared to be reversible or stabilized after 6 month cessation. This results complement those
obtained from other endpoints such as traditional toxicology and transcriptomics analyses that were conducted within PMI.
P200
Proteasome isoforms exhibit only quantitative differences in cleavage and epitope generation.
1
M. Mishto
1
Charite, Institut für Biochemie, Berlin, Germany
Questions: The 20S proteasome is the central proteolytic machinery of the ubiquitin proteasome system, being responsible for
the main part of extra-lysosomal protein degradation and generation of MHC class I-restricted epitopes. Upon inflammatory
stimuli a specific proteasome isoform, i.e. the immunoproteasome, is assembled in cells. Immunoproteasomes are considered to
be optimized to process antigens and to alter the peptide repertoire by generating a qualitatively different set of MHC-class I
epitopes. However, the question still remains whether, at the biochemical level, immunoproteasomes influence the quality
rather than the quantity of the immunogenic peptide pool. Methods: In last years we developed new methods for the
quantitative analysis of the entire spectrum of proteasomal digestion products by mass spectrometry (1, 2). These methods
provided higher detection sensitivities to our study. Therefore, we quantified cleavage site usage by human standard- and
immuno-proteasomes and proteasomes from immuno-subunit deficient mice as well as the peptides generated from model
polypeptides.
Results (3): We show that different proteasome isoforms can exert significant quantitative differences in cleavage site usage and
MHC class I-restricted epitope production. However, independent of the proteasome isoform and substrates studied, no
evidence was obtained for the abolishment of specific cleavage site usage, or for differences in the quality of the peptides
generated. Furthermore, our quantitative analysis also proved that the hypothesis of “destructive cleavages” proposed by
others could not be applied to MHC class I epitopes.
Conclusions: Thus, we conclude that observed differences in MHC class I-restricted antigen presentation between standard- and
immuno-proteasomes are due to quantitative differences in the proteasome-generated antigenic peptides.
Reference: 1. Mishto, M., Goede, A., Taube, K. T., Keller, C., Janek, K., Henklein, P., NieViennada, A., Kloss, A., Gohlke, S.,
Dahlmann, B., Enenkel, C., and Michael Kloetzel, P. (2012) Driving Forces of Proteasome-catalyzed Peptide Splicing in Yeast and
Humans. Mol Cell Proteomics 11, 1008-1023.
2. Liepe, J., Mishto, M., Textoris-Taube, K., Janek, K., Keller, C., Henklein, P., Kloetzel, P. M., and Zaikin, A. (2010) The 20S
Proteasome Splicing Activity Discovered by SpliceMet. PLOS Computational Biology 6, e1000830.
3. Mishto, M., Liepe, J., Textoris-Taube, K., Keller, C., Henklein, P., Weberruss, M., Dahlmann, B., Enenkel, C., Voigt, A.,
Kuckelkorn, U., Stumpf, M. P., and Kloetzel, P. M. (2014) Proteasome isoforms exhibit only quantitative differences in cleavage
and epitope generation. Eur J Immunol.
P201
Integrated Metabolome and Proteome Profiling of ρ0 Cells
1
1
2
I. Gielisch , D. Meierhofer , I. Wittig
1
Max Planck Institute for Molecular Genetics, Berlin, Germany
2
Goethe University , Funktionelle Proteomics, SFB 815 Core Unit , Frankfurt am Main, Germany
Introduction: Respiratory chain diseases represent a subset of mitochondrial disorders, which are biochemically characterized
by oxidative phosphorylation (OXPHOS) defects. Mitochondrial disorders are caused by mutations in the mitochondrial (mtDNA)
0
or the nuclear DNA, which can result in erroneous OXPHOS complexes. Human ρ cells have been depleted of mtDNA by
1
prolonged incubation with ethidium bromide . They lack core subunits of the respiratory chain and cannot support oxidative
0
2
phosphorylation. Thus, ρ cells are dependent on ATP solely derived from glycolysis . This cell model represents a very effective
tool to study the consequences of an absent OXPHOS system.
0
Objectives: The study aims to characterize the metabolome and proteome of ρ cells in comparison to the wild-type parental
cell line 143B.TK (thymidine kinase deficient osteosarcoma cell line). Identification of molecular mechanisms and dynamic
regulations in mtDNA depleted cells will be investigated by system biology approaches. Further functional studies and
validations will contribute to a deeper understanding of molecular mechanisms of the OXPHOS system.
Proteomic Forum 2015
103
Material & Methods: Proteome profiling is performed by nanoLC-MS/MS (Dionex Ultimate 3000 online coupled to a Q Exactive
Plus, Thermo Scientific) with strong anion exchange chromatography as additional fractionation step. Quantitative analysis is
investigated by SILAC labeling including a label-switch and accompanied by monitoring post transcriptional modifications, such
as phosphorylation and ubiquitination.
Metabolome profiling is achieved by targeted LC-MS/MS (Agilent 1290 Infinity UHPLC online coupled to a QTrap 6500, ABSciex),
based on multiple reaction monitoring (MRM). Specific findings will further be validated by biochemical methods.
Results: Preliminary results show a significant down-regulation of the pyruvate metabolism, the citric acid cycle and the OXPHOS
0
system. The electron carrier cytochrome c (CYCS) was almost 2-fold up-regulated in ρ cells, although there is no functional
respiratory chain and no significant regulation of the apoptosis pathway. Furthermore, proteins belonging to the G12/G13-family,
which regulate the actin cytoskeletal remodeling, were up-regulated. Additionally, the proteasome and mitochondrial ribosomes
0
were highly regulated in ρ cells.
Conclusion: This study will improve the understanding of OXPHOS impairments in mitochondrial pathologies by providing new
insights in altered pathways, networks and their molecular mechanisms.
1. King, M. P. & Attardi, G. Isolation of human cell lines lacking mitochondrial DNA. Methods Enzymol. 264, 304-13 (1996).
2. Chandel, N. S. & Schumacker, P. T. Cells depleted of mitochondrial DNA (rho0) yield insight into physiological mechanisms.
FEBS Lett. 454, 173-6 (1999).
P202
Label free MS1 based quantitative analysis of mouse skeletal muscle after 30 days in zero-gravity on the BIOn-M1 biosatellite
1,2,3
1
2
3
2
G. Tascher , S. Blanc , A. Van Dorsselaer , G. Gauquelin-Koch , F. Bertile
1
CNRS, Université de Strasbourg, Département Ecologie, Physiologie et Ethologie (DEPE), Strasbourg, France
2
CNRS, Université de Strasbourg, Laboratoire de spectrométrie de masse bio-organique (LSMBO), Strasbourg, France
3
Centre national d’études spatiales (CNES), Paris, France
Introduction: Weightlessness, as well as physical inactivity (e.g. during immobilization, unloading, ageing), trigger rapid decline
in skeletal muscle mass/force and alteration in its phenotype. Ultimately, we hypothesize that muscle changes may lead to
insulin resistance and development of a metabolic syndrome. Exercise countermeasures provide only minor protection, and
maintaining good nutrition may not be enough to cure the physiological problems of long-term space flights or physical
inactivity. Therefore, by feeding our understanding of muscle atrophy-related mechanisms, unravelling changes in muscle
proteome in response to zero-gravity would provide possible new targets towards which innovative countermeasures and/or
treatments against muscle atrophy could be designed.
Objectives: Using label-free MS-based proteomics, this study investigated the impact of zero-gravity (Bion-M1 flight) on mice
skeletal muscles.
Materials & Methods: After spaceflight during 30 days, C57/BL6 mice (n=5) were sacrificed and dissected muscles snap-frozen in
liquid nitrogen. Two control groups on earth were used (n=5 each), with mice housed either in the same confined conditions
used during spaceflight or in standard cages. After protein extraction, separation on a SDS-PAGE stacking gel, digestion with
Trypsin, and nanoLC-MS/MS analysis (nanoAcquity UPLC coupled to a Q-Exactive), protein identification was achieved using
Mascot and validation using Scaffold. MS1-based protein quantification was achieved using Skyline.
Results: More than 1300 proteins were identified (FDR <1%) of which 860 remained for quantification after manual validation.
Considering all samples in the experiment, very narrow retention times and low CV (<12%) for intensities of iRT peptides
indicated a very good stability of our instrument system. Moreover, very low CVs (<10%) for abundances of all proteins in a
reference sample repeatedly analyzed during the whole experiment indicated very good quantitative reproducibility. Statistical
analysis then revealed that relative abundance of 40 muscle proteins, notably involved in energy metabolism and inflammation
process, was significantly different between groups.
Conclusion: Using a robust quantitative proteomics workflow, we show that zero-gravity induces changes in the muscle
proteome of mice that notably support our hypothesis that weightlessness may drive to muscle lipid metabolism alteration and
development of insulin resistance.
Proteomic Forum 2015
104
P203
Absolute protein quantification of clinically relevant metabolizing enzymes by mass spectrometry-based targeted proteomics
1
1
1
1
D. Busch , C. Gröer , S. Oswald , W. Siegmund
1
University Greifswald, Clinical Pharmacology, Greifswald, Germany
Background: The pharmacokinetics of most drugs on the market is markedly influenced by intestinal and hepatic
biotransformation enzymes such as cytochrome P450 (CYP450) enzymes and UDP-glucuronosyltransferases (UGT). In order to
predict their impact on drug disposition as well as drug-drug interactions, data on their absolute intestinal and hepatic
abundance are required. Therefore, it was the aim of this study to develop and validate a LC-MS/MS method for the absolute
quantification of these proteins.
Methods: A LC-MS/MS method has been developed for the simultaneous detection of nine CYP (CYP1A2, CYP2B6, CYP2C8,
CYP2C9, CYP2C19, CYP2D6, CYP2E1, CYP3A4 and CYP3A5) and four UGT enzymes (UGT1A1, UGT1A3, UGT2B7 and UGT2B15)
which are considered to be of clinical relevance in human drug metabolism. Protein quantification was done by liquid
chromatography-tandem mass spectrometry (LC-MS/MS)-based targeted proteomics using enzyme specific tryptic peptides and
stable isotope-labelled peptides as internal standard. The MS/MS detection of all peptides was performed simultaneously with a
scheduled multiple reaction monitoring (sMRM) method in the positive mode by monitoring in each case three mass transitions
per enzyme specific peptides and the corresponding isotope-labelled internal standards. The assays were validated according to
current bioanalytical guidelines with respect to specificity, linearity, within-day and between-day accuracy and precision,
digestion efficiency as well as stability. Finally, the developed method was applied to determine the CYP and UGT protein
amount in human liver (HLM) and intestinal microsomes (HIM).
Results: A LC-MS/MS method has been developed that allowed the simultaneous quantification of all aforementioned enzymes
in one analytical run (60 min). The method was shown to be selective for the respective enzymes. The calibration span reaches
from 0.25 to 50 nmol/L for each enzyme. Within-day as well as between-day accuracy ranged from -13.1 to 12.5 % (relative
error) and precision from 1.1 to 14.8 % (coefficient of variation). All peptides featured stability during preparation, storage in the
autosampler (24 h at 4°C), overnight digestion (16 h at 37 °C) and up to three freeze-thaw cycles. Significant matrix effects could
not be observed. Using this validated method, the following relative percentage contribution of metabolizing enzymes was
observed in pooled HLM: CYP2E1, 24%; CYP2C9, 13.8%; CYP3A4, 12.1%; UGT2B15, 10.3%; CYP2C8, 10%; UGT2B7, 8.9%; UGT1A1,
7.5%; CYP1A2, 4.7%; CYP2B6 & CYP2D6, 3.5%; CYP2C19, 0.8%; CYP3A5, 0.7% and UGT1A3, 0.2%. In HIM it was: CYP3A4, 47.1%;
UGT1A1, 19.7%; CYP2C9, 10.8%; UGT1A3, 8.3%; UGT2B7, 7.1% and CYP2C19, 7%.
Conclusion: The developed method was shown to possess sufficient specificity, sensitivity, accuracy, precision and stability to
measure clinically relevant metabolizing enzymes in human tissue preparations. These absolute expression data may allow more
precise prediction of drug disposition using PBPK modelling-based approaches.
P204
Verification of a Parkinson's Disease Protein Signature by Multiple Reaction Monitoring LC-MS
1,2
3
3
3
3
4
5
5
3
T. Alberio , K. McMahon , M. Cuccurullo , L. A. Gethings , G. Weibchen , C. Lawless , M. Zibetti , L. Lopiano , J. P. Vissers , M.
1,2
Fasano
1
University of Insubria, Department of Theoretical and Applied Sciences, , Insubria, Italy
2
University of Insubria, Center of Neuroscience, Insubria, Italy
3
Waters Corporation, Health Sciences, Wilmslow, United Kingdom
4
University of Manchester, Faculty of Live Sciences, Manchester, United Kingdom
5
University of Torino, Department of Neuroscience,, Torino, Italy
Diagnosis of Parkinson's disease is based on the appearance of motor symptoms. A 2-DE analysis panel of proteins in the Tlymphocyte proteome was recently proposed as a disease signature. Here, an LC-MS based method was used to quantitatively
evaluate this signature by MRM in T-lymphocytes and peripheral blood mononuclear cells. A discriminant function was applied
to MRM data from T-lymphocytes protein extracts, assigning seven controls out of nine as true negatives and nine patients out
of nine as true positives. Good discriminant power was obtained by selecting a subset of peptides from the protein signature
(GELS, MOES, SEPT6, TWF2, LSP1, VIME, TALDO), with an ROC AUC of 0.877. The signature was not able to classify subjects by
analyzing whole mononuclear cells. The results suggest portability of the method to large cohort validation using alternative
technologies such as LC-MS.
P205
Proteomic Forum 2015
105
High quantification efficiency for discovery and validation approaches on a Q-TOF platform
1
1
S. Kaspar , U. Schweiger-Hufnagel , P.- O. Schmit
1
Bruker Daltonik GmbH, Bremen, Germany
2
Bruker Daltonique S.A., Wissembourg, France
2
Introduction: Shotgun mass spectrometry is sophisticated method for protein biomarker discovery. It is a global profiling
method, typically performed on high-resolution instruments, which is used to gain qualitative and quantitative information of
several thousands of peptides in complex proteomic digests. Subsequent validation and verification of detected biomarkers
requires accurate measurements of protein target abundances in biological samples. These tasks are typically performed
through a targeted quantitative proteomic approach involving MRM and MRM-based approaches. These methods require a
priori knowledge of the target precursor/product ion pairs (referred to as transitions) and time-consuming method preparation,
but prevent post-analysis data mining. Actual high-resolution systems address these limitations while providing comparable
selectivity. We report here the evaluation of a benchtop Q-TOF system for exact discovery quantification as well as for targeted
quantification of peptides.
Objectives: Detailed evaluation of a Q-TOF instrument for quantitative approaches
Material & Methods: To evaluate quantification performance for proteomics, different complex tryptic digests were mixed with
stable isotope labeled peptides or digests of standard proteins at known concentrations. Peptides were separated on a nanoflow UHPLC and analyzed on a impact II Q-TOF instrument equipped with a CaptiveSpray ion source. For peptide identification
and label-free quantitative analysis the MaxQuant software package was used (Nature Biotechnology 26, 1367 - 1372 (2008)).
Targeted quantification data was processed using Skyline.
Results: For the label-free quantification experiment a defined model system for complex proteomics samples, consisting of a
mixture of 48 standard proteins spanning a concentration range of five orders of magnitude (UPS-2, Sigma) was spiked into
samples of 500ng Yeast digest at a concentration of 1:2. During separation in 2h gradients, the UPS-2 peptides could be
quantified based on the MS full scans at levels from 500fmol down to the low attomole range.
For targeted quantification plasma tryptic digests samples were spiked with stable isotopically labelled standard peptides,
spanning a 10000 fold concentration range. Samples were measured using a high resolution extracted ion chromatogram (HRXIC) or Data Independent Analysis (DIA) mode. Results clearly show high quantification efficiency in undepleted plasma covering
a dynamic range of 4 orders of magnitude. Comparison with MRM-based approach revealed excellent correlation displaying high
versatility of the Q-TOF instrument for targeted quantification.
Conclusions: The evaluated high-resolution Q-TOF mass spectrometer platform clearly provides high efficiency both for
discovery and targeted approaches.
P206
Performance Investigation of a Novel Integrated Microfluidics Platform in High-Throughput LC-MS MRM Disease Protein
Marker Verification
1
1
1
1
1
C. Hughes , D. Turkovic , J. P. Vissers , L. A. Gethings , J. I. Langridge
1
Waters Corporation, Health Sciences, Wilmslow, United Kingdom
Biomarker discovery and validation are the first steps in understanding disease and drug development. Validation is technology
challenged since it requires analyzing a large number of samples with high-throughput, but also requires high sensitivity, high
resolution, large dynamic range and excellent selectivity. Targeted LC-MS based assays afford protein quantification with the
reproducibility and throughput required in order to improve marker acceptance and MRM, using tandem quadrupole MS being
one of the enabling technologies. Here, the application of a novel microfluidics platform for the quantification of peptides and
proteins is presented, considering speed, sensitivity and selectivity.
The MS Qual/Quant mixture was used for evaluation and contains 14 peptides, present as light and isotopically labeled analogs.
This was spiked into an E. Coli background such that loads for a 1 µl injection ranged from 32 amol to 40 fmol peptides in the
presence of 100 ng E.Coli. The sample was injected 3 times at 4 different loadings and separated using a 45 min RP gradient at a
flow rate of 1.2 µl/min. The mass spectrometer was operated at two different quad resolutions settings with three transitions
per peptide.
Proteomic Forum 2015
106
MRM chromatograms for each peptide showed that the peptides are resolved from the background matrix over the entire
dynamic range present within the sample. A particular example showed that 16 amol on-column for peptide NLSVEDAA[R] to 20
fmol for peptide GGPFSDSYR are present with no background matrix interference. MRM transitions were inspected ensuring
that a minimum of 3 peptides per protein and 3 transitions per peptide were detected and could be analyzed. Typical retention
time reproducibility for the monitored peptides show a standard deviation of 0.02 min for peptide EGHLSPDIVAEQK and
technical reproducibility for the three monitored transitions of Light (L)/Heavy (H) analogs of peptide GGPFSDSYR were
calculated as y5 (L 3.3%, H 5.7%), y6 (L 0.8%, H 1.2%) and y7 (L 5.4%, H 2.9%).
Quantitative measurement consistency between technical replicates was achieved. Ratio measurements for peptide
AVQQPDGLAVLGIFLK from Carbonic Anhydrase II are in agreement with an expected value of 8.4 vs. 8.5±0.8 determined. The
experiments were repeated at unit and elevated quadrupole resolution settings and measured ratios contrasted with the
expected values from four different injection volumes, ranging from 0.1 to 1 µl. The results show that unit quadrupole resolution
afforded sufficient quantitative accuracy, with precision not noticeably affected by resolution. As an indication of the LOD that
can be achieved, peptide NLSVEDAA[R] from Catalase is observed at the lowest sample injection amount, 0.1 µl and equates to
3.2 amol on column.
P207
Method Development for the Quantification of Cardiac Biomarkers by Targeted Mass Spectrometry
1
1
1
F. Polten , K. Wollert , A. Pich
1
Medical School Hannover, Hannover, Germany
Cardiovascular disease is the most common cause of death in the western world. Especially the acute myocardial infarction
poses a major threat and leads to severe permanent tissue damage, often followed by death. A new medical approach focuses
on exploiting the intrinsic regeneration processes to minimize the loss of function. Therefore it is crucial to understand the
mechanisms of how paracrine acting factors, such as proteins, control the regeneration to enable better diagnostics and the
foundation for new therapies. However, it often poses a major challenge to detect these low-abundance proteins within the
complex matrix of blood.
To identify such regulatory proteins the triple-quadrupol-based MRM-analysis is especially qualified. The relative concentration
of target protein can be calculated through the measurement of several specific precursor-fragment-transitions, which are
selected based on a bioinformatics approach or other MS-generated data. Within this work we established state-of-the-art MRM
assays for the detection and quantification of several potential biomarkers of cardiovascular diseases.
We quantified the relative levels of these biomarkers in clinical plasma samples of patients with an acute myocardial infarction
and, as a control group, patients with a treated stable cardiovascular condition. Furthermore homologous mouse proteins of
mice with a ligation of the coronary artery were compared to plasma of only operated mice (control) by utilizing an adapted
MRM-assay to analyze potential cardiac biomarker. Knock-out mice lacking the respective biomarker were used as a negative
control. Furthermore, absolute quantification of biomarker was achieved using an internal standard approach based on heavy
stable isotope labeled peptides.
P208
A new TMT-SRM 16plex neurodegeneration assay to measure biomarkers in cerebrospinal fluid
1
2
3
3
3
C. Lößner , M. Ward , H. Zetterberg , J. Gobom , K. Blennow , I. Pike
1
Proteome Sciences R&D, Frankfurt am Main, Germany
2
Proteome Sciences plc, London, United Kingdom
3
University of Gothenburg, Gothenburg, Sweden
2
Selected Reaction Monitoring (SRM) using triple quadrupole mass spectrometers is the preferred method for consistent
quantitative measurement of multiple protein markers across numerous samples and over time. Nevertheless, SRM method
development is time consuming and this results in high costs. As a cost-efficient alternative, we have developed a TMT-SRM
assay based on TMT discovery data and using bulk material of the respective biological matrix (universal reference). We have
compared this novel universal reference method to a conventional internal standard approach and obtained favourable
Proteomic Forum 2015
107
performance metrics. We have further evaluated the stability of the universal reference material using a sophisticated testing
regime.
Based on this, a TMT-SRM assay for neurodegeneration protein markers in cerebrospinal fluid (CSF) has been developed. These
analytes have been selected based on our prior proteomics discovery projects, review of the literature and consultation with key
opinion leaders. The developed assay has been quality controlled by using preparative as well as analytical replicates. The final
assay consists of 30 peptides representative of 16 proteins and has been applied to Alzheimer’s disease patient samples as well
as age and sex matched non-demented controls. To demonstrate the performance of the developed assay format, we are
presenting here for the first time the most recent data obtained for a sample cohort comprising of 96 samples.
P209
Extracellular matrix and intracellular proteome of primary skin fibroblasts in 2D compared to 3D tissue culture
1
R. Tölle
1
University Medical Center Freiburg, Dermatology, Freiburg im Breisgau, Germany
Matrix based cell culture is commonly used as tissue model. When growing in a 3D matrix, many cells differentiate and adopt
tissue characteristics, leading to a different gene expression pattern as compared to 2D. A commonly used matrix for 3D tissue
culture is matrigel. It displays a basement membrane matrix in tissue culture and models the cellular microenvironment. To
study the cellular response to 3D culture conditions we employed stable isotope labeling by amino acids in cell culture (SILAC)based quantitative mass spectrometry (MS) and compared cells grown in 2D and 3D.
The 3D matrix is made of growth factor reduced matrigel and collagen I from rat tail supplemented with insulin-transferrinselenium. To mimic physiological conditions, the main part of the matrix consists of collagen type I. Primary normal human
fibroblasts (NHF) are seeded directly into the gel before polymerization. The cells are cultured in SILAC media for 5 days before
cells or their ECM is extracted from the gel. By using two different sets of SILAC labels for 2D and 3D tissue culture, both
intracellular and extracellular proteomes can be quantitatively compared. For efficient cell extraction from the matrix and
subsequent cell lysis a protocol has been set up. We identified characteristic changes in protein abundances of cells grown in 2D
and 3D and currently study the roles of single proteins only present in 3D culture conditions.
Taken together, we established a protocol for quantitative MS analysis of cells cultured in a 3D matrix which also allows the
analysis of ECM secreted by cells. Comparing the intra- and extracellular proteome of cells cultured in 2D and in 3D molecular
pathways only observable in 3D culture conditions were identified.
P210
Secretomics of the endometrium during diapause and reactivation in the tammar wallaby (Macropus eugenii)
1
2
2
1
1
1
F. C. Martin , C.- S. Ang , N. Williamson , M. Renfree , D. Gardner , G. Shaw
1
Department of Zoology, University of Melbourne, Melbourne, Australia
2
Mass Spectrometry and Proteomics Facility, Bio21 Molecular Science and Biotechnology Institute, University of Melbourne,
Melbourne, Australia
Introduction: Embryonic diapause, an arrest in development at the blastocyst stage, occurs in around 140 species of mammals
representing most of the major orders. In the tammar wallaby, a marsupial, diapause is maintained for 11 months with no
embryonic growth or cell death, first by lactational inhibition and later by seasonal influences. During the lactational phase,
reactivation can be induced reliably by removing the sucking pouch young (RPY). Since the blastocyst is surrounded by an
acellular shell coat, control of diapause and reactivation must be mediated by soluble factors in the uterine environment,
regulated by changes in secretory activity of the endometrial glands and luminal epithelium.
Objective: The objective of the study is to investigate the protein composition of the uterine environment during diapause and
reactivation.
Method: Tammar endometrium and uterine flushings from diapause (d0 RPY), day 4, day 5, day 6 and day 8 after reactivation by
RPY were analysed by LC-MS/MS and iTRAQ labelling.
Results & Conclusion: Quantitation of endometrial proteins during diapause and reactivation stages by iTRAQ labelling showed
that heat shock, structural and transport proteins were differentially expressed. Comparison of the endometrial proteome and
Proteomic Forum 2015
108
the proteins from uterine flushings identified certain proteins that were distinct to reactivation stages. For example, hepatoma
derived growth factor and Insulin-like growth factor binding protein-1, which are known to stimulate embryonic growth and cell
proliferation in cattle, were present in the tammar uterine flushings of the reactivation stages. Enzymes involved in amino acid
metabolism such as Carnosine dipeptidase-1 were present only in the reactivation stages, suggesting that they may be necessary
to alter the milieu surrounding the blastocyst, thereby affecting metabolic processes and reactivation. These dynamic changes
suggest that reactivation from embryonic diapause in the tammar involves a coordinated sequence of secretion of uterine
proteins.
P211
Poly(ADP-ribose) glycohydrolase silencing down-regulates TCTP and Cofilin-1 associated with metastasis in benzo(a)pyrene
carcinogenesis
1
1
1
1
2
1
1
H. Huang , X. Li , Z. Zhuang , J. Liu , X. Li , W. Gao , Y. Liu
1
Shenzhen Center for Disease Control and Prevention, Key Laboratory of Modern Toxicology of Shenzhen, Shenzhen, China
2
Guangxi Medical University, School of Public Health, Nanning, China
QUESTION: Benzo(a)pyrene (BaP) is a ubiquitously distributed environmental pollutant. BaP is a known carcinogen and can
induce malignant transformation of rodent and human cells. Many evidences suggest that inhibitors of poly(ADP-ribose)
glycohydrolase (PARG) is potent anticancer drug candidate. However, the effect of PARG on BaP carcinogenesis remains unclear.
METHODS: We explored this question in a PARG-deficient human bronchial epithelial cell line (shPARG cell) treated with various
concentration of BaP for 15 weeks. Soft agar assay was used to examined BaP-induced cell malignancy of human bronchial
epithelial cells and shPARG cell. Mechanistic investigations were used by 2D-DIGE and mass spectrometry. Western blot analysis
and Double immunofluorescence detection were used to confirm some of the results obtained from DIGE experiments.
RESULTS: We found that PARG silencing could dramatically inhibit BaP-induced cell malignancy of human bronchial epithelial
cells in soft agar assay. Altered levels of expression induced by BaP were observed within PARG-deficient cells for numerous
proteins, including proteins required for cell mobility, stress response, DNA repair and cell proliferation pathways. Among these
proteins, TCTP and Cofilin-1 involved in malignancy, were validated by western blot analysis and immunofluorescence assay.
PARG inhibition contributed to down-regulation of TCTP and Cofilin-1.
CONCLUSION: This is the first experimental demonstration of a link between PARG silencing and reduced cell migration after
BaP exposure. We propose that PARG silencing might down-regulate TCTP and Cofilin-1 associated with metastasis in BaP
carcinogenesis.
Figure 1. DIGE analysis of cells after treated with BaP (PH=3-11, NL, 240 mm). Overlay of Cy3 and Cy5 derived from a single gel,
highlighting differentially expressed protein spots.
Figure 2. The possible molecular mechanisms of BaP carcinogenesis and the role of PARG silencing.
Acknowledgments: This work was supported by NSFC (No.81001261, No.81370080), the Science and Technology Program of
Shenzhen (No. 201302238), and the Shenzhen Science and Technology Development Fund Project (No.
JCYJ20130329103949642).
Figure 1
Proteomic Forum 2015
109
Figure 2
P212
Extending the limits of quantitative proteome profiling with data-independent acquisition and application to acetaminophen
treated 3D liver microtissues
1
1
1
1
2
3
1
1
1
R. Bruderer , O. Bernhardt , T. Gandhi , S. M. Miladinovic , L.- Y. Chen , S. Messner , T. Ehrenberger , V. Zanotelli , Y. Butscheid ,
1
2
1
1
C. Escher , O. Vitek , O. Rinner , L. Reiter
1
BiognoSYS AG, Schlieren, Switzerland
2
Purdue University, Department of Statistics, West Lafayette, United States
3
Inphero AG, Schlieren, Switzerland
Data-independent acquisition (DIA) is a novel mass spectrometric method for proteomics which can be used for similar
experiments as data-dependent acquisition or shotgun proteomics, an established technology. No systematic evaluation of the
two technologies has been reported yet.
We implemented a novel DIA method on the widely used Orbitrap platform and used retention time normalized (iRT) spectral
libraries for targeted data extraction. We call this combination hyper reaction monitoring (HRM). Using a controlled sample set,
we show that HRM outperformed shotgun proteomics in peptide detection and quantification of differential protein abundance.
Utilizing HRM, we profiled acetaminophen treated 3D human liver microtissues. An early onset of relevant proteome changes
was revealed at subtoxic doses of acetaminophen. Further, we detected for the first time six human NAPQI-protein adducts that
might be relevant for the toxicity of acetaminophen.
Our findings imply that HRM should be the preferred method for quantitative protein profiling.
P213
Clinical Proteomic Assays Based on High-Resolution Mass Spectrometry
1
1
1
1
D. Bruno , S. Gallien , Y. J. Kim , G. Berchem
1
Luxembourg Clinical Proteomics, CRP-Sante, Strassen, Luxembourg
Introduction: Targeted proteomics biomarker studies are routinely performed on triple quadrupole instruments. The low
resolution of quadrupole mass filters limits the selectivity of the analysis in bodily fluids. Hybrid mass spectrometers with high
resolution and accurate mass (HR/AM) capabilities can overcome this limitation.
Analysis of clinical samples performed in parallel reaction monitoring (PRM) on a quadrupole-orbitrap mass spectrometer
showed definite gain in selectivity, translating in more robust results.
Proteomic Forum 2015
110
Objectives: The objective was to demonstrate the benefit of using high-resolution, accurate mass targeted analyses to improve
the selectivity and the precision of the measurements in clinical samples.
Method: Plasma samples were collected from patients diagnosed with lung cancer (at various disease stages), and healthy
controls. Samples were depleted of their two most abundant proteins albumin/IgG, then reduced, alkylated, digested with
trypsin. Isotopically labeled peptides were added before LC-MS/MS analysis. A nano-HPLC system coupled to a quadrupoleorbitrap instrument (Q-Exactive HF, Thermo Scientific, Bremen). The quantification was performed by extracting post-acquisition
the ion chromatograms, and the peptide identities were confirmed by spectral matching.
Results: The targeted LC-MS/MS analyses were performed in parallel reaction monitoring mode (PRM) on plasma samples. The
quadrupole-orbitrap instrument was operated in parallel reaction monitoring mode, acquiring full MS/MS spectra of predefined
precursor ions corresponding to the peptides of interest. The actual experimental design was greatly simplified compared to a
triple quadrupole analysis as only m/z values of selected precursor ions were required The signals of fragment ions were
extracted post-acquisition; narrow mass windows (10 ppm) were used, which reduced background interferences. The trapping
capability of the instrument allowed to enrich for precursor ions in tiny amounts. The width of the quadrupole isolation window
impacts on the selectivity of measurements.
A total of 80 peptides corresponding to 40 proteins, as potential lung cancer markers, were monitored in triplicated analyses of
80 patient and 40 control blood samples. The PRM analyses showed quantification performance translating in consistent results
for the different peptides of the same protein, a clear discrimination between the control and patient samples, and the
differentiation of the disease stages.
Conclusion: The migration of assays from triple quadrupole instruments to hybrid high-resolution quadrupole orbitrap mass
spectrometers increases the selectivity and the precision of the measurements. In addition, the systematic generation of full
product ion spectra corroborates the identity of the analytes. Assays based on HR/AM are thus applicable to the analysis of
bodily fluids and tissue samples.
P214
DIA and Spectronaut: Comprehensive and precise proteome profiling
1
1
1
1
L. Reiter , O. Bernhardt , T. Gandhi , R. Bruderer
1
Biognosys AG, Research & Development, Schlieren, Switzerland
Introduction: In recent years the proteome coverage using shotgun proteomics has steadily increased. Low complex proteomes,
such as the E. coli or yeast proteomes, can be measured almost completely in a single LC-MS measurement. For relative
quantitation, the label free approach has recently gained in popularity mainly due to its simplicity. However, this approach has
been limited by the semi-stochastic nature of shotgun proteomics which leads to a large number of missing values, especially if
many conditions are measured. Even though MS1 alignment attenuates the missing value problem, it is difficult to control the
reliability of identification with this approach. Further, low intensity signals in MS1 often show interferences which lowers the
precision of relative quantitation. Data independent acquisition (DIA) has promised to solve the missing value problem. By using
wide precursor windows, DIA consistently measures all precursors that are above the limit of detection. For DIA data analysis we
have developed the Spectronaut software. Using Spectronaut, we get more identifications in a single LC-MS measurement as
compared to shotgun proteomics. Further, this approach resulted in quasi gap-free quantitation matrices without alignment and
higher precision of quantitation as compared to shotgun proteomics.
Methods: Spectronaut performs a targeted analysis of DIA data. It uses the information in a spectral library to identify peptide
signals. Among the information in the spectral library are the precursor and fragment ion m/z, the indexed retention time of the
peptide (iRT) (Escher et al. 2012) and relative fragment ion intensities. False discovery rates are estimated using the mProphet
method (Reiter et al. 2011). To increase the precision of quantitation we have implemented an interference detection algorithm.
This algorithm detects interferences which are often unstable by comparing each extracted fragment ion current of a peptide to
the consensus elution profile. Fragment ion signals not following the consensus elution profile likely contain an
interference[RB1] [C2] and hence will be consistently removed from quantitation across all conditions.
Results: Label free proteome profiling in human HEK-293 cell lines using Spectronaut resulted in median coefficients of variation
below 8%. Also, we show that applying our interference detection algorithm further reduced the coefficients of variation.
Proteomic Forum 2015
111
Conclusions: Spectronaut is a fast and easy to use software for the targeted analysis of DIA data and quantitative proteome
profiling. The hallmarks of the resulting data are high proteome coverage, very precise proteome profiling and nearly gap free
quantitation matrices[C3] without the use[RB4] of alignment algorithms.
References: Escher, C. et al., 2012. Using iRT, a normalized retention time for more targeted measurement of peptides.
Proteomics, 12(8), pp.1111-21.
Reiter, L. et al., 2011. mProphet: automated data processing and statistical validation for large-scale SRM experiments. Nature
methods, 8(5), pp.430-5.
P215
Quantitative Strategies for Multiplexed High-throughput Protein Analysis
1
1
A. Bhangu-Uhlmann , W. Deininger
1
PolyQuant GmbH, Bad Abbach, Germany
Fast-paced innovation in analytical mass spectrometry and its inherent methodological refinements for modern life sciences
research have significantly impacted discovery and collection of valuable information at the cellular state. Hence, this hand-inhand progress has been actively involved in underpinning comprehensive scientific knowledge particularly in the diagnostic and
pharmaceutical framework. With proteins being the major cellular executers, advanced proteomics tools have gained
tremendous importance in deciphering pathological issues while facilitating solid data validation along with gene expression at
the transcriptome level. New avenues such as personalized medicine in clinical laboratory settings require robust platforms to
assess quantitative read-outs for individualized therapies. In this respect, use of quantotypic peptides unparallels protein
quantification by absolute means in a singular LC-MS run. Quantotypic peptides are represented as an array of concatenated
proteotypic peptides, which are synthesized from tryptic fragments, and termed as a QconCAT protein. Customizable stable
isotope labelling options enable for the concomittant detection of up to 100 peptides. The ease of multiplexing absolute protein
quantification as high-throughput fashion will be presented highlighting current reference examples applied for instance in
human health or plant proteomics.
Proteomic Forum 2015
112
Poster Session B • 23 March 2015, 15:00–16:30
Affinity Proteomics (P001–P013)
P001
DNA-Affinity purification followed by mass spectrometry (AP-MS): application to the analysis of transcriptional regulators
interacting with a defined DNA sequence
1
1
1
A. Tacheny , M. Dieu , P. Renard
1
University of Namur, Research Unit in Cell Biology, Namur, Belgium
Transcription factors are targets of intense researches to better understand how cells regulate gene expression. Nonetheless,
most regular assays investigating the binding of transcription factors to a DNA sequence require pre-existing knowledge about
proteins supposed to bind the sequence of interest. Therefore, unbiased methods, called DNA-Affinity Purification followed by
Mass Spectrometry (AP-MS) are being developed to identify the transcription regulators interacting with a defined DNA
sequence.
DNA-AP-MS has the following general scheme: a long DNA bait synthesized by PCR is immobilized on a chromatographic
support, and then incubated with cell nuclear extracts. The DNA-protein complexes are then eluted, digested and identified by
mass spectrometry. Although conceptually simple, this represents a technological challenge due to the low abundance of
regulatory proteins compared to the highly abundant proteins binding to nucleic acids in a non sequence-specific manner and to
the proteins adsorbed on the chromatographic support. The key parameters to increase the chances of success will be
discussed, such as the requirement for a specific elution step, the need for a decomplexification of the peptide sample and an
adapted MS-MS protocol to focus on low abundant proteins. The crucial question of the relevant controls will also be discussed.
These considerations will be illustrated through the analysis of the proteins interacting with a 226bp regulatory sequence from
HIV LTR5’. Over 100 proteins were identified, among which >50 % of transcription factors and co-activators/co-repressors. Most
of the transcription factors known to control this promoter region were identified, as well as transcriptional regulators not yet
described as HIV-1 regulators, including the Meis transcription factor. Based on DNA AP-MS experiments performed with the
HIV LTR5’ sequence mutated in the Meis binding site, we have able to propose a model of Meis-dependent recruitment of the
Sin3a co-repressor complex.
P002
Unbiased allele-specific quantitative proteomics unravels molecular mechanisms influenced by cis-regulatory genomic
variations
1
2
1
2
3
4
4
2
2
S. Hauck , H. Lee , C. von Toerne , M. Claussnitzer , H. Grallert , I. Dahlman , P. Arner , H. Hauner , H. Laumen
1
Helmholtz Zentrum München, Research Unit Protein Science, Neuherberg, Germany
2
Technische Universität München, Else Kroener-Fresenius-Centre for Nutritional Medicine, Chair of Nutritional Medicine, 85350
Freising-Weihenstephan, Germany
3
Helmholtz Zentrum München, Institute of Epidemiology II, 85764 Neuherberg, Germany
4
Karolinska Institutet, Department of Medicine, Huddinge, 14186 Stockholm, Sweden
Genome-wide association studies identified numerous risk loci for various diseases. Recent large scale next generation
sequencing approaches based open chromatin and novel bioinformatics address a major challenge of human genetics, the
identification of cis-regulatory variants. However, to further enhance our understanding of gene regulation mediated by cisregulatory variants, it is essential to delineate the precise molecular mechanism underlying cis-regulatory variants and disease
risk. Here, we introduce a label-free, quantitative DNA protein interaction approach which enables simple identification of allelespecific binding proteins at cis-regulatory variants. We analyzed allele-specific differential protein binding at two predicted
complex and two non-complex SNPs at the PPARG locus conferring risk for development of type 2 diabetes and found overall
increased protein binding at the predicted complex regions confirming the recently developed integrated approach to identify
causal variants (Claussnitzer et al., Cell 2014). Among the differential binders, we identified allele-specific binding of the
transcription factor YY1 and its coregulator RYBP (RING1 and YY1 binding protein) at the rs7647481 A-allele. Specific binding was
validated in competition and supershift experiments and additionally in reporter gene assays where luciferase constructs of all
tested variants demonstrated that overexpression of YY1 significantly activates solely the rs7647481 non risk allele. The
rs7647481 non risk variant also significantly increased transcriptional activity in different cell types. In subjects carrying the
rs7647481 nonrisk allele, adipose mRNA levels of RYBP correlate with improved insulin sensitivity.
Proteomic Forum 2015
113
Thus, the efficient identification of protein complexes containing TFs and their co-regulators by sensitive label-free mass
spectrometry improves understanding of mechanisms underlying genetic associations. We have confirmed feasibility of this
label-free proteome DNA binding approach in two additional complex diseases, for restless legs syndrome (RLS) where we could
identify binding of Creb1 to intronic region of Meis1 regulating its enhancer function in the developing brain (Spieler et al.,
Genome Research 2014) and for the RAD50 locus associated to atopic diseases, where we identified differential bind of Smad3
and SP1 that impacted transcriptional activity (Kretschmer et al., Allergy 2014). Future work will concentrate on uncovering
molecular effects of genetic variations in age-related macular degeneration.
P003
Label-free quantitative proteomic analysis of LRRK2-associated protein interaction networks in neuronal cell model systems
1
2,3
2,3
2,3
1,4
1,4
A. Meixner , D. C. Schöndorf , M. Deleidi , T. Gasser , M. Ueffing , C. J. Gloeckner
1
University of Tübingen, Medical Proteome Center, Tübingen, Germany
2
German Center for Neurodegenerative Diseases (DZNE), Department of Neurodegenerative Diseases, Tübingen, Germany
3
University of Tübingen, Hertie-Institute for Clinical Brain Research, Tübingen, Germany
4
Helmholtz Zentrum München, Research Unit for Protein Science, Munich, Germany
Parkinson’s disease (PD) is the most prevalent age-related neurodegenerative movement disorder, typically developed in a
sporadic fashion but rare monogenic inherited forms exist as well. Mutations in the leucine-rich repeat kinase 2 (LRRK2) are
collectively the most common cause of familial and idiopathic PD yet discovered, whereas the protein’s involvement in disease
pathogenesis is uncertain. Although previous studies revealed a variety of biological processes to be affected by LRRK2, there is
little consensus on the underlying molecular mechanisms and impact of pathogenic mutations on the physiological function
which still remain elusive.
Belonging to the ROCO protein superfamily, LRRK2 features a catalytically active GTPase as well as a MAPK homologous kinase
domain flanked by several protein interaction domains. This strongly suggests that LRRK2 is an integral part of a functional
protein network and a molecular hub for signaling complexes. To gain knowledge on normal LRRK2 protein interactions as well
as deviations occurring as a consequence of pathogenic mutations in a neuronal context we applied quantitative affinity-based
proteomic approaches and compared LRRK2 variant-specific protein interactions. We utilized neuronal model systems that base
on established cell lines or human induced pluripotent stem cells (hiPSC) and represent the endogenous situation as close as
possible. Stable, inducible SH-SY5Y cell lines were generated, recombinantly expressing different (N)SF-TAP-tagged LRRK2
variants. Using the Flp-in T-Rex system, the created cell lines are isogenic, hence assuring constant protein expression across the
cell population from a single site of chromosomal integration. In addition, we employed a LRRK2 WT hiPSC line, reprogrammed
from dermal fibroblasts of a healthy individual, as well as a heterozygous G2019S mutant line derived thereof by targeted
genome editing, constituting an isogenic disease model with the PD-relevant mutation being the single genetic difference
against the same genetic background as the control. A LRRK2-specific (N)SF-TAP tag was integrated into these hiPSC lines via
additional site-directed engineering.
Applying affinity-based purification of LRRK2-associated protein complexes from the model systems followed by high-resolution
mass spectrometric analysis and label-free quantification, we were able to confidently identify and verify LRRK2-specific protein
interactions that converge in relevant pathways, including vesicular transport and cytoskeletal function. The integration of our
data with preexisting knowledge by computational approaches will ultimately assist the generation of a high confident
interaction network on which PD-relevant mutation act, causing pathophysiological deviations that affect the development and
progression of the disease.
P004
Stoichiometry Determination of the intraflagellar transport complex A by Absolute Quantification
1
2
3
1
4
1,5
Y. Wissinger , J. Fuchs , N. Mischerikow , A. Meixner , K. Mechtler , M. Ueffing , K. Boldt
1
University of Tübingen, Medical Proteome Center, Tübingen, Germany
2
IMBA, Mass Spectrometry & Protein Chemistry, Vienna, Austria
3
Boehringer Ingelheim, Vienna, Austria
4
Research Institute for Molecular Pathology, Vienna Biocenter, Vienna, Austria
5
Helmholtz Zentrum München, Research Unit Protein Science, Neuherberg, Germany
1
Cilia protrude from the surface of virtually any polarized cell in the human body and are involved in many sensing and signalling
processes. Furthermore they are conserved from chlamydomonas to humans. Genetic disorders caused by mutations, affecting
ciliary functions are summarized under the term ciliopathies. One of the essential ciliary machineries is the intraflagellar
Proteomic Forum 2015
114
transport (IFT). In collaboration with dynein and kinesin motors, it enables a bidirectional movement of protein cargo along the
ciliary axoneme. The IFT machinery consists of two subcomplexes (IFT-A and IFT-B) which represent the link between the motor
proteins and the cargo. To understand and to generate models of the underlying molecular mechanism of ciliary function, as
well as their dysfunction in disease, determining the constitution and stoichiometry of involved protein complexes is of major
significance.
As a first proof of principle and due to the fact that malfunction of IFT-A proteins results in cilia deformation that causes the
ciliopathy Sensenbrenner Syndrome, this protein complex was chosen.
To purify the protein complex IFT-A, the FLP-In System was used to generate stable cell lines. Three proteins of interest were
chosen as baits: IFT122, an integral component of the IFT-A, TULP3 which is described to be associated with the protein complex
of interest and LCA5, a protein that interacts with the IFT-A in an indirect and rather transient way. Purification was performed
using one-step affinity purification.
For the Absolute Quantification a standard mix consisting of an equimolar amount of standard peptides was created using the
“Equimolarity through Equalizer Peptide” approach. Afterwards a known amount of unlabelled standard mix containing known
amounts of standard peptides was spiked into the purified and stable isotope labeled sample. Quantification was performed
using Selected Reaction Monitoring (SRM) on a QTrap 5500 and Parallel Reaction Monitoring PRM on a Q-Exactive plus.
The results show that this is a highly repetitive protein complex that consists of multiple copies of each component, resulting of
a molecular weight of 4.2 MDa. The successful determination of the stoichiometry is the first step for generating a molecular
model of the IFT-A complex and will be supplemented by cross-linking. Furthermore, this approach will be adapted to other
ciliary complexes, aiming towards a molecular understanding of ciliopathies.
P005
Analysis of Virus Attachment Complexes by
combining native Gel Electrophoresis and nLC-MS/MS
with Cluster Analysis of the protein co-migration behaviour
1
1
J. Doellinger , A. Nitsche
1
Robert Koch-Institute, Centre for Biological Threats and Special Pathogens, Berlin, Germany
Introduction & Objectives: Virus attachment to its host cell is the first prerequisite for permissive infections. The identification
of protein complexes formed between virus particles and host cells during the attachment is therefore important for the
understanding of viral pathogenesis.
This study aimed to evaluate the use of combining native gel electrophoresis and nLC-MS/MS with cluster analysis of the protein
co-migration behavior for the analysis of protein complexes between virus and cell surfaces during attachment.
As a model we choose to analyze the attachment of three cowpox viruses (CPXV) strains to HEp-2 cells. CPXV is a human
pathogenic large dsDNA virus, which causes most orthopoxvirus (OPV) zoonotic infections in Europe and Northern and Central
Asia.
Materials & Methods: Three strains of purified CPXV particles were resuspended in cultivation media and incubated at 4°C on a
confluent cell layer of HEp-2 cells. At 4°C CPXV particles bind to cells but cannot fuse with the membrane. Viral particles
attached to cells were purified by zonal sucrose gradient centrifugation and caesium chloride density gradient centrifugation.
Membrane complexes were solubilized and separated on a Blue native PAGE. Entire gel lanes between ~ 100 -1,000 kDa were
cut into slices, proteins were digested in gel and identified by nLC-MS/MS. Proteins were further clustered according to their comigration profiles in the native gel separately for each virus using the tiling with overlap approach in WEKA (University of
Waikato,https://weka.waikato.ac.nz).
Results: The protein clusters from the three samples were compared, and only proteins that co-occurred in at least one cluster
in each sample were regarded as forming a complex and were visualized in a network. The network consists of 27 proteins, 17
viral and 10 human, with 20 of them known to be membrane associated or secreted. The core of the network is formed by three
viral membrane proteins. The human membrane-associated proteins cluster together with a high degree of connection and are
separated from presumably cytoplasmic contaminants within the network. Six membrane-bound cellular proteins are identified
to form one complex at the virus surface.
Proteomic Forum 2015
115
Conclusions: The data show that purification of native complexes formed during the attachment of virus and host is possible
from repurified CPXV and is reproducibly across different strains. The combination of native gel electrophoresis and nLC-MS/MS
with cluster analysis could be especially useful for complex viruses, whose attachment is not initiated by only one specific
protein.
P006
Chemical proteomic analysis of cytokinin molecular partners in Arabidopsis
1
1
1
2
1
3
R. Simerský , V. Skalický , R. Lenobel , V. Mik , I. Chamrád , M. Srnad
1
Centre of the Region Haná for Biotechnological and Agricultural Research, Department of Protein Biochemistry and Proteomics,
Olomouc, Czech Republic
2
Centre of the Region Haná for Biotechnological and Agricultural Research, Department of Chemical Biology and Genetics,
Olomouc, Czech Republic
3
Palacký University & Institute of Experimental Botany AS CR, Laboratory of Growth Regulators, Olomouc, Czech Republic
Introduction: Principle mechanisms of cytokinin mode of action have been thoroughly characterized within last years;
nevertheless, our knowledge in the cellular physiology remains rather enigmatic. In this study, we report the development of
chemical proteomics methodology to dig in deeper in this interesting subject.
Objectives: The aim of this study was to prepare affinity matrices with immobilized analogs of isopentenyladenine (iP) and use
them consequently for the isolation of cytokinin-interacting proteins from the Arabidopsis thaliana suspension culture cells.
Material and Methods: Two iP-derived ligands with short linkers situated in C2 and N9 position, respectively, were synthesized.
The ligands were immobilized through its primary amino group on commercially-supplied NHS-activated beads. Beads blocked
with ethanolamine served as a blank matrix.
The affinity experiment consisted of three pull-downs: a blank pull-down (incubation of blank beads with protein extract), a
regular pull-down (incubation of affinity beads with protein extract), and an iP-competitive pull-down (association of affinity
beads with protein extract in the presence of free iP)
An aliquot of the A.t. protein extract was diluted into incubation buffer containing a mixture of protease inhibitors, combined
with the affinity beads and incubated overnight at 4 °C. The captured proteins were eluted, digested and analysed utilising a
tandem mass spectrometer UHR-QTOF coupled to a nanocapillary liquid chromatography system and an on-line nanoESI source.
Functional annotation and enrichment analysis were performed using publicly available bioinformatic annotation tools and
databases.
Results:197 proteins were identified following the affinity purification. The group of proteins interacting with an iP ligand
immobilized at C2 position (193) was significantly larger in comparision to the group of iP(N9)(31) . The competition experiment
revealed 35 proteins specifically interacting with iP ligands (25 with iP(C2) and 10 with iP(N9)). The proteins purified using iP(C2)
matrix were clustered in consonance with their functional similarities. These proteins were predominantly involved in response
to abiotic stress, cytoskeleton formation and carbohydrate metabolism.
Conclusion: Our chemical proteomic strategy that integrates cytokinin affinity chromatography with proteomic analysis should
contribute to deeper and more thorough understanding of the relations between CK structure and their biological function and
lead to the identification of their novel cellular targets potentially usable for various agricultural applications.
P007
Comparison of the compounds VI16832 and KiNet-1 for small molecule affinity chromatography of protein kinases
1
1
1
1
1
K. Morgenroth , E. Richter , M. Harms , J. Mostertz , F. Hochgräfe
1
Competence Center Functional Genomics, Pathoproteomics, Greifswald, Germany
Protein kinases are essential key nodes in cellular signal transduction networks. Their dysfunction, e.g. by aberrant alterations in
abundance or regulatory posttranslational modifications, is implicated in a variety of diseases. System-wide approaches for the
characterization of protein kinases therefore not only help to gain a better understanding of the molecular biology of the cell but
can also lead to the development of novel pharmaceutical intervention strategies.
Proteomic Forum 2015
116
A powerful chemical proteomics strategy for a comprehensive study of protein kinases in biological samples involves a
combination of affinity chromatography with small molecule broad-band kinase inhibitors (SMAC) and high accuracy LC-MS/MS
analysis. Here, we tested the compounds VI16832 (Evotec) and KiNet-1 (SYNkinase), either alone or in combination, for their
usage with SMAC in human THP-1 macrophages and 16HBE14o- human bronchial epithelial cells.
In total, we identified 152 protein kinases with at least two independent peptides in THP-1 cells with an Orbitrap Velos Pro
hybrid mass spectrometer. Of those, 129 protein kinases were identified by VI16832 and 106 by KiNet-1 in single inhibitor
assays. Interestingly, 46 protein kinases were exclusively identified with VI16832 and 23 with KiNet-1. Comparative classification
of the identified kinases revealed a greater binding capacity for VI16832 for kinases of the CMGC class, whereas KiNet-1 binds
more kinases of the atypical class. Our results indicate that the combined usage of both substances will result in maximized
kinome coverage. Eventually, we outline our quantitative strategy for the global analysis of thiol-redox modifications in protein
kinases by a combination of SILAC, SMAC and specific labeling of reversibly oxidized thiols with thiol-reactive affinity tags.
P008
On Terminal Alkynes That Can React with Active-Site Cysteine Nucleophiles in ProteasesTitle of the contribution
1
2
2
1
1
2
S. Goerdayal , R. Ekkebus , S. van Kasteren , A. Scholten , A. Heck , H. Ovaa
1
University of Utrecht, Biomolecular Mass spec and Proteomics, Utrecht, Netherlands
2
the Netherlands Cancer Institute, Division Cell Biology, Amsterdam, Netherlands
Active-site directed probes are powerful instudiesof enzymatic function. For capturing proteases such as deubiquitinases we
report an active-site directed probe based on a warhead so far considered to be unreactive. By replacing the C-terminal
carboxylate of ubiquitin (Ub) with an alkyne functionality, a selective reaction with the active-site cysteine residue of
deubiquitinating enzymes (DUBs) was observed. The resulting product was shown to be a quaternary vinyl thioether, as
determined by X-ray crystallography.
Using mass spectrometry based quantification by stable isotope dimethyl labelling, proteins specifically binding to the
immobilized Ub alkyne probe could be monitored. This experiment confirmed the selectivity towards a wide variety of DUBs.
This methodology proved to be a valuable tool in the study of DUBs in both cellular and tissue contexts.
P009
Chemoproteomics: Does kinase abundance after enrichment reflect kinase activity?
1,2
1,3,4
1,3,4
1
1,5
1,3,4,2
B. Ruprecht , J. Zecha , S. Heinzlmeir , G. Medard , S. Lemeer , B. Kuster
1
Technische Universität München, Chair of Proteomics and Bioanalytics, Freising, Germany
2
Center for Protein Science Munich (CIPSM), München, Germany
3
German Cancer Consortium (DKTK), Heidelberg, Germany
4
German Cancer Research Center (DKFZ), Heidelberg, Germany
5
Utrecht University, Bijvoet Center for Biomolecular Research and Utrecht Institute of Pharmaceutical Sciences, Utrecht,
Germany
Protein kinases are important mediators of intracellular signaling and are reversibly activated by phosphorylation. As aberrant
kinase activity is often associated with diseases like cancer, kinases are prime targets for the design of small molecule inhibitors.
Chemoproteomic approaches like Kinobeads or MIBs have been proven applicable to enrich these low abundant species and to
identify targets of kinase inhibitors. It has been argued that kinase abundance after enrichment is not only dependent on
expression and affinity to the immobilized inhibitors, but also on kinase activity. In order to examine this in more detail, we used
tyrosine phosphatase inhibition by pervanadate to activate kinases before enrichment with single and mixed inhibitor resins.
Quantitative phosphoproteomics indicated global signaling induction and activation of key kinases which was not fully reflected
by chemoproteomic kinase binding. In general, kinases showed three distinct modes of capture upon activation: kinases that are
lost, such with increased binding and unaffected ones. Moreover, activity based binding was a function of the applied inhibitor
resin and was often distorted by the mixture thereof. As a consequence, we argue that binding behavior has to be examined for
each kinase and the respective inhibitor resin individually in order to distinguish activity from expression. We further compiled a
resource of activity resolved, single inhibitor resins that can be used for: (i) rational composition of bead mixtures for activity
profiling, (ii) conformation selective capture and screening of kinases, (iii) identification of blind spots within affinity kinase
resins and (iv) prioritization of kinases that can adopt an inactive-like conformation. Finally, we used the gathered information to
show that Lapatinib preferentially targets an active-like conformation of ERBB2 and that direct pathway activity read out is only
applicable to a priori characterized kinases.
Proteomic Forum 2015
117
P010
Chemically cleavable linkers for target identification
1
S. Verhelst
1
Leibniz Institut für Analytische Wissenschaften, Dortmund, Germany
Question: Covalent chemical probes are crucial tools in chemical biology for the study of enzymes, mechanisms of covalent
drugs and post-translational modifications. Mass spectrometry plays a crucial role in the identification of the probe targets and
modification sites, but enrichment of probe targets is often accompanied by background.
Methods: We have embarked upon the development of chemically cleavable linkers in order to reduce background protein
identifications and to enable selective elution of probe-modified tryptic peptides.
Results: We will here present our latest results on the design, synthesis and evaluation of cleavable trifunctional tags,
incorporating an azide for click chemistry, a fluorophore for sensitve detection and a biotin for enrichment purposes.
Conclusion: Cleavable trifuntional tags are versatile reagents for the in-gel detection and mass spectrometry-based
identification of target proteins.
Figure 1
P011
Presenting AFFIRM - a novel immunoaffinity platform that combines recombinant antibody fragments and LC-SRM analysis
for sensitive multiplexed protein analysis
1
1
S. Waldemarson , A. Säll , H. Persson
1
Lund University, Lund, Sweden
1
Introduction: Targeted measurements of low abundant proteins in complex mixtures are highly demanded in many areas, not
the least in clinical research and analysis. Developing assays for measuring these however remains challenging.
Objectives: We have developed the novel platform AFFIRM (AFFInity sRM) that utilizes the power of antibody fragments (scFv)
to efficiently enrich for target proteins from a complex background and the exquisite specificity of an SRM-MS based detection
(1). Such recombinant antibody fragments represent a sustainable resource that allows fast production and is amendable to
high-throughput generation of specific binders. Novel multiplexed AFFIRM assays can be implemented with short lead-time
through the continuous build of a resource of antibodies and SRM assays (Figure 1). Protein level affinity enrichment allows in
addition to protein quantification for the exploration of protein sequence variants. These might be the ultimate discriminative
biomarkers.
Materials & Methods: The AFFIRM platform has been developed for a high-throughput workflow in a 96-well format using a
magnetic bead processor for complete automation of capture and wash steps (Figure 1). The ability of AFFIRM has been
demonstrated through generating response curves in single- and multiplexed format for three target proteins of interest,
Proteomic Forum 2015
118
Keratin 19, BRCA1 and MUC1 in a serum background (1). The platform is being further developed through the generation of
biotin tagged antibody fragments for directed coupling to streptavidin coated magnetic beads.
Results: Linear responses have been demonstrated down to low ng/ml concentrations with high reproducibility (1). The
platform allows for high throughput measurements in 96-well format and all steps are amendable to automation and scale-up.
Development of biotin-tagged antibody fragments allows for oriented coupling of the antibodies to the magnetic beads that
provides further AFFIRM assay sensitivity as well as multiplexing capacity. The resource of antibodies and SRM assays is
continuously being built and assay LOD and LOQ is being explored.
Conclusion: We believe the use of recombinant antibody technology in combination with SRM MS analysis provides a powerful
way to reach sensitivity, specificity and reproducibility as well as the opportunity to build resources for fast on demand
implementation of novel assays.
(1) J Proteome Res. 2014 Nov 25. AFFIRM-A Multiplexed Immunoaffinity Platform That Combines Recombinant Antibody
Fragments and LC-SRM Analysis.
1
Säll A , Carlsson F, Olsson N, Wingren C, Ohlin M, Persson H, Waldemarson S.
Figure 1
P012
Abduction of protein complexes from mammalian cells by Virotrap
1
1
1
1
1
1
1
1
S. Eyckerman , K. Titeca , E. Van Quickelberghe , E. Cloots , A. Verhee , N. Samyn , L. De Ceuninck , E. Timmerman , D. De
1
1
2
1
1
Sutter , S. Lievens , S. Van Calenbergh , J. Tavernier , K. Gevaert
1
VIB/UGhent, Medical Protein Research, Ghent, Belgium
2
UGhent, Laboratory for Medicinal Chemistry, Ghent, Belgium
Proteins typically exert their function within supra-molecular complexes. In current mass spectrometry (MS)-based strategies to
map protein-protein interactions, preparation of a cell or tissue homogenate is an inevitable step. However, the associated lysis
and purification steps are known to induce loss of interactors (false negatives) as well as artificially gain false interactors (false
positives) due to the loss of cell integrity. Virotrap is a novel approach for mapping intracellular protein interactions based on
fusion of a bait protein to the HIV-1 GAG protein so that interaction partners become trapped in virus-like particles (VLPs) that
bud from mammalian cells (see Figure 1). It obviates the need for cell homogenization and protects the abducted protein
complexes during purification. Using an efficient VLP enrichment protocol, Virotrap was applied to known binary interactions, to
the MS-based identification of novel proteins involved in inflammation signalling (including the ubiquitin-editing enzyme A20 as
Proteomic Forum 2015
119
bait and its newly identified preys), to the identification of stimulus-dependent interactions (LPS-induced protein complex
dynamics) as well as interactions with small molecules (drug target screens).
Figure 1
P013
In vivo interaction mapping in C. elegans embryos
1
2
2
2
2
J.- X. Chen , P. G. Cipriani , D. Mecenas , K. C. Gunsalus , F. Piano , M. Selbach
1
Max Delbrück Center for Molecular Medicine, Berlin, Germany
2
New York University, New York, United States
1
Question: Mapping protein interactions in vivo is fundamental to our understanding of the molecular mechanisms underlying
development.
Methods: Here, we used a method that combines in vivo expressed GFP-tagged proteins with quantitative proteomics to
identify protein-protein interactions in developing C. elegans embryos. This strategy is generic and can be used with any GFPtagged protein.
Results: To test our approach we focused on key proteins involved in embryogenesis and built an initial in vivo interaction map
composed of 561 interactions among 484 proteins. Our network reflects known biology and is highly enriched in functionally
relevant interactions. We further demonstrate the utility of the map by looking for new regulators of P granule (ribonuclear
protein particles required for germline development) dynamics. We find that a previously uncharacterized protein, GEI-12,
interacts with the kinase MBK-2 and is required for P granule maintenance. The interaction between GEI-12 and MBK-2 appears
conserved in mammalian cells and leads us to propose a model in which the phosphorylation state of GEI-12 controls P granule
assembly and disassembly in development.
Conclusions: In summary, our results show that in vivo interactome mapping is a powerful approach that provides unique
insights into animal development.
Proteomic Forum 2015
120
Imaging (P050–P055)
P050
3D Molecular Histology
- Correlation of 3D Histology and 3D Molecular Imaging 1
1
1
1,2
1
3
2
4
H. THIELE , J. Lotz , J. Lotz , F. Hoffmann , S. Heldmann , D. Trede , G. Ernst , O. Guntinas-Lichius , F. von Eggeling
1
FRAUNHOFER MEVIS, Inst. Medical Image Computing, Lübeck, Germany
2
Friedrich Schiller University Jena, Inst. Physical Chemistry, Jena, Germany
3
SCiLS GmbH, Bremen, Germany
4
Klinik Hals-, Nasen- und Ohrenheilkunde Universitätsklinikum Jena, Jena, Germany
4
Question: New molecular imaging techniques have arisen in the last year. One is the matrix-assisted laser desorption/ionization
mass spectrometry imaging (MALDI MSI) which has the potential to detect and characterize tumor cells and their environment
in a spatial context. Here, 3D data analysis helps to better understand structures (vessels, tumors) of the tissue. The concept of
3D multimodal imaging combining molecular imaging methods with standard histology techniques is a promising attempt to
meet this challenge. Therefore we analyzed an oral squamous cell carcinoma (OSCC) with up to 60 consecutive sections with
H&E staining, immunohistochemistry (IHC) against CD31 and MALDI MSI.
Methods: Serial sections of an oral squamous-cell carcinoma (OSCC) were analyzed by MALDI MSI, H&E and IHC. H&E and IHC
stained sections were digitized with a high-resolution scanner. Microscopic resolution leads to large images to big to keep in
memory. Concepts have been developed to handle these images for visualization and computation needed for the 3D
reconstruction of the tumor. To recapture the spatial correspondence between neighboring slides in the serial sections, the used
algorithms for fully automatic registration compute either a rigid/affine transformation or a nonlinear, elastic transformation.
Especially elastic transformations allow for local deformations in the individual sections. For 3D MALDI MSI data processing a
spatial segmentation of the entire 3D MALDI MSI data was obtained by performing a clustering analysis, to detect different
morphological regions in the tissue. Co-registration of the high resolution H&E and IHC images and the MALDI MSI data united
anatomical, histological and molecular information in 3D. Data fusion of multimodal images yields information that cannot be
delivered by one isolated technique.
Results: We developed an interactive tool that visualizes the co-registered data in a 3D volume rendering, allowing simultaneous
analysis, orthogonal views to axial, coronal, and sagittal direction, and virtual cutting along a free plane positioned with arbitrary
angle and offset. MALDI MSI cluster results representing defined regions of functional difference can be rendered into the 3D
volume and allow a better understanding of tissue borders and the spatial context. We present a new MALDI-imaging data
analysis approach that clusters m/z-images according to their spatial similarity. In summary, our approach creates new
capabilities for understanding and analysis of the complex course of tumorigenesis and tumor progression.
Conclusion: The integration of 3D MALDI MSI data with H&E and IHC scans allows a correlation between histological and
molecular information in 3D. Moreover, this model is open for additional image modalities which will further enrich the
information available in combined datasets.
P051
MALDI-Imaging mass spectrometry:
Discrimination of pathophysiological processes in traumatized skeletal muscle
1
2
3
3
2,3
2,3
4
H. Thiele , O. Klein , G. Nebrich , K. Strohschein , G. N. Duda , T. Winkler , D. Trede
1
Fraunhofer Mevis, Inst. Medical Image Computing, Lübeck, Germany
2
Charite-Universitätsmedizin Berlin, Berlin-Brandenburg, Center for Regenerative Therapies, Berlin, Germany
3
Charite - Universitätsmedizin Berlin, Julius Wolff Inst. Center f. Musculoskeletal Surgery, Berlin, Germany
4
SCiLS GmbH, Bremen, Germany
Questions: Due to formation of fibrosis and the loss of contractile muscle mass, severe muscle injuries result in insufficient
healing marked by a significant reduction of muscle force and motor activity. We used the approach of matrix-assisted laser
desorption/ionization (MALDI) mass spectrometry imaging (MSI) to map the spatial distribution of peptide signatures in injured
skeletal muscle tissue from rat. By using an unsupervised segmentation, peptide signatures were analyzed, which are
characteristic for trauma muscle region.
Proteomic Forum 2015
121
Methods: Optical images of the muscle sections were taken using standard flatbed scanner. MALDI-Imaging data acquisition was
performed on an Autoflex III MALDI-TOF/TOF (Bruker daltonics). After MALDI-Imaging experiments the tissue sections were
stained with hematoxylin and eosin (H&E). The SCiLS Lab software was used for unsupervised characterization of the
traumatized and healthy rat muscle data. The denoised data was spatially segmented using the bisecting k-means algorithm. For
each of the two clusters, we searched for spatially co-localized m/z-values by calculating Pearson correlations between the
spatial masks of the cluster and each m/z-value and taking m/z-values with highest correlation values.
Results: This study shows that tissue-based proteomic profiling by MALDI-IMS and follow “bottom up” LC−MS/MS analysis is a
suitable method to to discriminate pathophysiological alterations between the primary trauma and trauma adjacent muscle. The
discriminative m/z-values for injured muscle regions were performed by using an unsupervised segmentation (SCiLS Lab).
Conclusions: We identified and validated by immunohistochemistry alpha skeletal muscle actin (Acts) and carbonic anhydrase 3
(Ca3) as promising novel biomarkers to discriminate these pathophysiological regions. Further potential markers, tenascin-C
(Tnc), filamin C (Flnc) and nascentpolypeptide-associated complex alpha (Naca), could be localized in primary trauma region. It is
shown that MALDI-Imaging technique represents a valuable tool to describe the spatial characteristic of the pathophysiological
muscle regions.
References:
[1] Matziolis, G., Winkler, T., Schaser, K., Wiemann, M., et al., Autologous bone marrow-derived cells enhance muscle strength
following skeletal muscle crush injury in rats. Tissue engineering 2006, 12, 361-367.
[2] Winkler, T., von Roth, P., Matziolis, G., Mehta, M., et al., Dose-response relationship of mesenchymal stem cell
transplantation and functional regeneration after severe skeletal muscle injury in rats. Tissue engineering. Part A 2009, 15, 487492
[3] Klein, O. et. al., MALDI imaging mass spectrometry: Discrimination of pathophysiological regions in traumatized skeletal
muscle by characteristic peptide signatures, J. Proteomics 2014, DOI. 10.1002/pmic.20140088
P052
Lipid Visualization and Identification Through Collision Cross Section Aided Correlation of MALDI Imaging and MS/MS
Fragmentation Data Sets
1
1
1
1
M. Towers , E. Claude , I. Edwards , J. P. Vissers
1
Waters Corporation, Health Sciences, Wilmslow, United Kingdom
MALDI mass spectrometry imaging (MSI) is becoming an established technique within lipidomics research. Using MSI, a large
number of species can be visualized within a tissue section. Subsequent identification can be extremely challenging. Lipids can
be identified by extracting them from the same or consecutive tissue sections and performing MS/MS. However, due to the
large number of (near) isobaric isobaric species, confidence can be lost when assigning identifications to peaks within the
imaging data set by accurate mass alone. Here, we demonstrate the use of ion mobility to differentiate ions and calculate
collision cross sections (CCS) along with high accurate mass, adding confidence to peak assignment when relating fragmentation
information from extracted lipids to imaging data.
CCS calibration was conducted prior to imaging. During the imaging experiment, an external lock mass was acquired at set time
points to maintain high accuracy mass measurement. Post imaging, lipids were extracted from the same tissue and deposited
onto a stainless steel target plate. A second CCS calibration was acquired and the extracted lipids analyzed by MS/MS.
Identifications were assigned to imaging information using accurate mass and calculated CCS values.
P053
Classification of prostate cancer tissue samples into prognostic subgroups Testing the potential of MALDI MSI as a clinical tool
1
2
2
1
3
1
B. Beine , D. Dietrich , G. Kristiansen , H. E. Meyer , B. Sitek , C. Henkel
1
Leibniz-Institut für Analytische Wissenschaften -ISAS- e.V., Biomedical Research, Dortmund, Germany
2
University Hospital Bonn, Institute of Pathology, Bonn, Germany
3
Ruhr-Universität Bochum, Medizinisches Proteom-Center, Bochum, Germany
Introduction: Prostate cancer (PCa) is today’s third most common cancer diagnosed in Europe (Ferlay et al. 2013). In Germany
PCa was the most prevalent cancer in men in 2012 (67.000 men diagnosed) and the third-leading cause of cancer death (Robert
Proteomic Forum 2015
122
Koch Institute, 2012). For diagnosis and treatment decisions, histological findings based on needle biopsy material are the
clinical standard (Heidenreich et al. 2012). At the time of diagnosis prediction for individual tumor progression, aggressiveness
and PCa related death is limited (Bangma et al. 2007). Due to this fact risk assesment of patients is one major problem in PCa
management (Klotz L. 2012). Early detection, simultaneously risk assessment and PCa progression are highly important for
successful therapy.
Objective: The presented study aims the identification of different subgroups of PCa patients by matrix assisted laser
desorption/ionization (MALDI) mass spectrometry imaging (MSI) using tissue microarrays (TMAs) of 220 patients. A workflow for
data analysis will be developed and validated using SCiLS Lab. By using MALDI MSI the progression of individual tumors should
be assessed. The goal is to develop a classification tool for the identification of patients with aggressive PCa using MALDI MSI.
Method: Tissue microarrays of 220 PCa patients with either a recurrence or not were analyzed using peptide MALDI MSI. Each
patient was represented by 2 to 6 tissue cores of tumor tissue. Preparation of formalin fixed paraffin embedded tissue (FFPE)
was established and transferred to TMA analysis. A data analysis workflow was developed using the software SCiLS Lab.
Classification was based on the results of training sets for statistical analysis. The data of the training set was used to classify the
TMA into different prognostic patient subgroups.
Results: Data analysis using SCiLS Lab allowed the classification of the different tissue cores into patient subgroups. The strategy
for data analysis was validated.
Conclusion: Peptide imaging using MALDI MSI is a suitable tool for answering clinically relevant questions. The successful
analysis of TMAs allows the risk assessment of many patients simultaneously and is therefore a powerful tool.
P054
xiNET - a visualization tool for exploring cross-linking/mass spectrometry results
1
2
2,1
C. Combe , C. Forbrig , J. Rappsilber
1
University of Edinburgh, Wellcome Trust Centre for Cell Biology, Edinburgh, United Kingdom, Germany
2
Technische Universität Berlin, Institute of Biotechnology, Department of Bioanalytics, Berlin, Germany
xiNET is a visualization tool for exploring cross-linking/mass spectrometry results. The interactive maps of the cross-link network
that it generates are a type of node-link diagram. In these maps xiNET displays: (i) residue resolution positional information
including linkage sites and linked peptides; (ii) all types of cross-linking reaction product; (iii) ambiguous results; and, (iv)
additional sequence information such as domains. xiNET runs in a browser and exports vector graphics which can be edited in
common drawing packages to create publication quality figures. Availability: xiNET is open source, released under the Apache
version 2 license. Results can be viewed by uploading data to http://crosslinkviewer.org/ or by downloading the software from
https://github.com/colin-combe/crosslink-viewer and running it locally.
P055
Multiple mass spectrometric imaging analyses from single tissue section with Desorption Electrospray Ionization (DESI) on a
oa-TOF mass spectrometer
1
1
1
I. Edwards , E. Claude , E. Jones
1
Waters Corporation, Health Sciences, Wilmslow, United Kingdom
Introduction: Over the years Desorption Electrospray Ionization (DESI) has been applied to Mass Spectrometry Imaging (MSI) for
direct analysis of surfaces at atmospheric pressure. Here we demonstrate repeat analysis of the same tissue section without
modification or exhaustion of the surface molecules, enabling dual polarity analysis giving rise to a wealth of molecular
information from a single tissue section. Furthermore a low spatial resolution image can be acquired followed with a higher
spatial resolution image, specific to a region of interest.
Methods: Snap frozen tissues of porcine liver were sectioned on a cryo-microtome to 15µm thickness and thaw mounted onto
conventional glass slides; the samples were stored at -80°C. Immediately prior to analysis the samples were brought to room
temperature and placed onto the stage, no further sample preparation is required. The 2D DESI source was mounted onto a
SYNAPT HDMS G2-Si. DESI spray conditions were set at 1.5µl/min, 90:10 MeOH: water at 100psi N2 gas pressure and a voltage of
5kV for both polarities. The pixel size was determines in the X-direction by the speed of the stage movement and acquisition
rate of mass spectra. The Y-direction was defined by the distance between two lines of acquisition.
Proteomic Forum 2015
123
Results: To test the potential of performing on the same tissue section dual polarity DESI-MS imaging experiments, the same
section of porcine liver tissue was imaged in both positive and negative mode. In both modes of ionization, an abundance of
lipids and endogenous metabolites were detected, giving intense peaks. A second experiment was designed to evaluate whether
in "dual polarity" mode the first passage of the DESI spray alters the chemical information obtained from the same section in
subsequent acquisitions. Spectra from a single DESI imaging experiment in positive mode when compared to spectra generated
from a consecutive tissue section, also in positive mode after a first experiment was carried out in negative mode show identical
peaks with very similar relative intensities.
Differential spatial resolution experiments on the same tissue section were also performed. In the first DESI imaging experiment
a raster pattern was defined over the whole porcine liver tissue section with a pixel size of 150 microns. A second experiment
was carried out on a specific region of the same section at 50 microns. The relative lipid intensities obtained from spectra from
each experiment were comparable.
Further experiments will be carried out using human clinical liver samples, followed by the Hematoxylin and Eosin (H&E) staining
for accurate characterization of the tissue regions.
Conclusions: Multiple imaging analyses on the same tissue section can be performed using DESI increases the amount of
information obtainable from a single tissue section. The non destructive nature of the technique enables subsequent H&E
staining.
Mass Spectrometry Technologies (P056–P065)
P056
Increasing Depth of Coverage in Data Independent Acquisition
1
1
1
2
2
D. Waidelich , J. Dojahn , M. Macht , B. Collins , R. Aebersold , C. Hunter
1
AB SCIEX, Darmstadt, Germany
2
ETH Zurich, Zurich, Switzerland
3
AB SCIEX, Redwood City, United States
3
Introduction & Objectives: Data independent acquisition (DIA) strategies have been used to increase the comprehensiveness of
data collection while maintaining high quantitative reproducibility. In DIA, larger fixed-width Q1 windows are stepped across the
mass range in an LC timescale, transmitting populations of peptides for fragmentation, and high resolution MS and MS/MS
spectra are acquired. Previous work has shown that using more narrow Q1 windows can improve peptide detection and increase
sample coverage. Here both method and instrumentation advancements will be explored to continue to increase depth of
sample coverage.
Materials & Methods: The MS analysis was performed on a modified quadrupole time of flight instrument equipped with an
ADC detection system. DIA data collection was done using SWATH™ acquisition with prototype acquisition software to explore a
variety of acquisition strategies. The DIA data was interrogated using a comprehensive yeast spectral library created from many
data dependent experiments. Results assessment was performed using Excel tools.
Results & Conclusion: Original work exploring variable window size and more narrow windows demonstrated that increasing the
number of total Q1 windows from 24 to 60 windows provided an increase in confident peptide detections with good
quantitative reproducibility of ~15%. To enable higher sample loads, the dynamic range of the detection system was extended
by switching from a TDC based detection system to an ADC based system on a modified TripleTOF system. At the higher sample
loads, we next applied increasingly narrow Q1 windows during SWATH acquisition to continue to improve the S/N in MS/MS.
The number of windows was extended from 60 to 100 windows across the 400-1250 precursor m/z range while maintaining a
cycle time of 3.2 secs. This provided a 20% increase in confident peptide detections with 20% or better CVs across replicates.
Further optimization of longer chromatography and investigations of impact on other proteomes will be discussed.
Proteomic Forum 2015
124
P057
Exploring Impact of Dynamic Accumulation for Improving MS/MS Quality of QqTOF Data
1
1
1
2
2,3
D. Merkel , J. Dojahn , M. Macht , N. Bloomfield , S. Seymour , C. Hunter
1
AB SCIEX, Darmstadt, Germany
2
AB SCIEX, Redwood City, United States
3
Seymour Data Science, San Francisco, United States
2
Introduction & Objectives: Recent innovations in QqTOF instrumentation has resulted in a large increase in MS and MS/MS
®
acquisition speed providing deeper coverage of complex proteomes. Some workflows, such as iTRAQ reagent quantitation or
PTM characterization, benefit more from higher spectral quality than traditional data-dependent workflows. Here, a QqTOF
acquisition strategy that uses precursor intensity to adapt the MS/MS accumulation time (dynamic accumulation) was explored
for its utility in improving these proteomic datasets.
®
Materials & Methods: Analysis of complex protein digests was performed using nanoflow LC/MS analysis on a TripleTOF
system. Data collection was done in data dependent mode with prototype acquisition software to explore a range of acquisition
rates and precursor intensity combinations for optimal coverage and spectral quality. Protein identification data was processed
using ProteinPilot™ Software and results assessment was performed using Excel tools. A number of areas of improvement were
investigated, impact on MS/MS quantitation for iTRAQ reagents, effect on number of acquired spectra and therefore
subsequent processing time, impact on protein identification rates, and impact on the generation of SWATH™ acquisition
spectral ion libraries.
®
Ecoli lysate was labeled with 8plex iTRAQ regents and mixed with equal loading in all channels. The sample was analyzed using
three different acquisition strategies and the identification yields were characterized for both ID and quantitation.
Results & Conclusion: The dynamic accumulation approach provided a small increase in total protein/peptide identifications and
significant improvements in the quantitation quality. The median reporter ion intensity was shifted higher by 34%, and the
variance of protein ratio distributions was reduced (16% improvement in quality). The peptide variation about the protein was
constant across the peptide intensity range, indicating improved quantitation of lower signal peptides.
P058
Characterization of an Improved Ultra-High Resolution Quadrupole Time of Flight (UHR-TOF) Instrument for Proteomics
Applications
1
1
1
1
S. Kaspar , M. Lubeck , A. Michalski , O. Raether , C. Gebhardt
1
Bruker Daltonik GmbH, Bremen, Germany
1
Introduction: In shotgun proteomics it is desirable to identify and quantify a large number of individual peptides from complex
samples, such as tryptic digests of human plasma samples or whole cell lysates in the shortest possible time. Complexity and
concentration range, however, pose a great challenge to the MS instrumentation in terms of sensitivity, resolution and dynamic
range. Several hardware modifications of a bench-top UHR-TOF instrument were carried out and evaluated addressing these
particular performance aspects.
Objectives: Evaluation of a bench-top UHR-Q-TOF instrument for proteomics applications
Methods & Materials: To test the impact of these modifications on proteomics performance, different complex tryptic digests
were analyzed with nano-flow UHPLC and a CaptiveSpray ion source connected to the impact II instrument (Bruker Daltonik). For
peptide identification the MaxQuant software package was used (Nature Biotechnology 26, 1367 - 1372 (2008)).
Results: For higher sensitivity at fast acquisition speed, ion extraction from the collision cell into the orthogonal acceleration of
the TOF-analyzer was improved by using a novel collision cell design. Increased resolution without changing the effective flight
path could be achieved with a modified reflectron. In addition, a faster detector (reduced width of individual ion signals) led to
further improvements in resolving power. Comparison with previous design clearly shows an increased resolving power resulting
in a full sensitivity resolution (FSR) of 50,000.
Using an optimized detector digitizer combination, a threefold higher dynamic range was observed. However in complex
samples, the dynamic range is also limited by the capability of the instrument to resolve nearly isobaric compounds. The
Proteomic Forum 2015
125
dynamic range detectable in proteomics samples has been investigated. For one of the most difficult proteomics samples undepleted plasma - we could cover a dynamic range of 4 orders of magnitude. We will show further results evaluating the
impact of the novel hardware features on performance improvements for proteomics applications.
Conclusion: Improvements to several hardware components allow identification and quantification of complex proteomics
samples with very high dynamic range.
P059
Identification of complex I assembly intermediates in TMEM126A and TMEM126B deficient cells
1,2
2
1,2
1,2
3
3
2,4
5
1,2
V. Strecker , L. Bleier , M. Steger , H. Heide , B. Schwamb , M. Zörnig , S. Dröse , U. Brandt , I. Wittig
1
Goethe University, Functional Proteomics, SFB815 Core Unit, Frankfurt am Main, Germany
2
Goethe University, Molecular Bioenergetics Group, Cluster of Excellence Frankfurt "Macromolecular Complexes” , Frankfurt am
Main, Germany
3
Chemotherapeutisches Forschungsinstitut Georg-Speyer-Haus, Frankfurt am Main, Germany
4
Goethe University, Hospital Frankfurt am Main, Department of Anesthesiology, Intensive-Care Medicine and Pain Therapy,
Frankfurt am Main, Germany
5
Radboud University Medical Center, Nijmegen Center for Mitochondrial Disorders, Nijmegen, Netherlands
The well-established method of quantitative BNE is suitable to measure steady state levels of OXPHOS complexes in specimens
from patient and model organisms. However, application of standard BNE is rather limited for the identification of labile or
scarce sub-complexes and assembly intermediates. To overcome these limitations, we recently introduced complexome
profiling, an unbiased bottom up method that is suitable to detect low abundant protein complexes, to identify known and
unknown interaction partners and to study dynamics, protein turnover and posttranslational modifications of macromolecular
complexes. By complexome profiling, we identified and characterized the mitochondrial membrane protein TMEM126B as an
essential assembly factor of complex I that forms a stable mitochondrial complex I assembly complex (MCIA) with the other
known assembly factors CIA30, Ecsit and Acad9 [1]. Here we studied the impact of TMEM126A, a homologue to TMEM126B that
was recently found to be associated with mitochondrial disorders. TMEM126A knockdown caused a complex I assembly defect
and respiration deficiency with NADH-linked substrates. TMEM126A and TMEM126B knockdown cells showed an attenuated
assembly of the membrane arm of complex I. Furthermore, building blocks of the peripheral arm of complex I bound to
assembly factors were found to accumulate. This indicates that the assembly of the N- and Q-module is not affected.
Interestingly, in TMEM126A knockdown cells MCIA seems to associate with different assembly intermediates since its
components are found at a lower mass range compared to parental cells. In summary our results suggest that TMEM126A could
act in a close sequence with MCIA during membrane arm assembly of complex I.
[1] Heide H, Bleier L, Steger M, Ackermann J, Dröse S, Schwamb B, Zörnig M, Reichert AS, Koch I, Wittig I and Brandt U,
Complexome profiling identifies TMEM126B as a component of the mitochondrial complex I assembly (MCIA) complex. Cell
Metab. 2012;16:538-549
P060
A Complete Workflow for Achieving High Accuracy Multiplexed N-Terminomics Data Using TMT10plex
1
P. Saliba
1
Thermo Fisher Scientifics, Dreieich, Germany
Proteolytic cleavage is a critical PTM that can significantly affect the behavior and function of a protein and its interaction
partners. However, identifying, quantifying, and confirming the substrates and specificities of the known proteases, many of
1
which are only vaguely categorized, has proved to be a challenging task . A breakthrough methodology in this line of analysis by
LC/MS takes advantage of a negative enrichment strategy coupled with amine-reactive labeling of newly formed (neo) N-termini
2
to yield relative abundance information . This study presents a complete workflow for this analysis using TMT10plex tags with
3
accurate Synchronous Precursor Selection (SPS) MS quantitation and a new data analysis pipeline, Proteome Discoverer 2.0,
3
employing the Byonic search engine.
The Byonic searches of MS spectra from two replicate analyses yielded identifications for nearly 3000 protein groups and 18000
unique peptides before enrichment and 2500 and 5500 respectively following enrichment, a two-fold increase when compared
to Mascot search results. In the enriched sample, over 800 TMT10plex N-terminally labeled peptides were detected with a
quantitation rate exceeding 90%, These represent high confidence sites of protease activity. The high-multiplexing capabilities of
TMT10plex enabled the use of additional tags as technical replicates within the same LCMS experiment enhancing quantitative
Proteomic Forum 2015
126
3
reproducibility and confidence. The magnitude of observed fold changes increased using the SPS MS method (compared to data
collected using MS2 based quantitation) due to elimination of interference, co-isolated with the parent ion. Using Proteome
Discoverer 2.0 with Byonic, enabled facile identification of biologically relevant substrates showing significant changes in
abundance. In conclusion, combining negative N-terminal peptide enrichment and higher multiplexing TMT with high
3
resolution SPS MS and an advanced data analysis pipeline made possible a much deeper characterization of the N-termini
proteome.
P061
MALDI in-source decay for characterization of engineered proteins
1
2
3
1
2
M. Michalak , S. Fiedler , H.- J. Gabius , J. Kopitz , M. Schnölzer
1
Institute of Pathology, University Hospital Heidelberg , Department of Applied Tumor Biology, Heidelberg, Germany
2
German Cancer Research Center (DKFZ), Functional Proteome Analysis, Heidelberg, Germany
3
Ludwig-Maximilians University, Institute of Physiological Chemistry, Faculty of Veterinary Medicine, Munich, Germany
Introduction: In-source decay (ISD) coupled with matrix-assisted laser desorption/ionization (MALDI) time-of-flight (TOF) mass
spectrometry is a fragmentation technique for structural analysis of intact proteins and peptides, among other biomolecules.
Fragmentation along the protein backbone allows rapid N- and/or C-terminal sequence determination. ISD can be applied to
confirm the sequence of recombinant engineered proteins, including structural variants of growth and differentiation regulatory
members of the galectin family, that are attractive means in protein structure-activity studies.
Objectives: Application of top-down sequencing using MALDI-TOF ISD in primary structure analysis of engineered variants of
human galectins.
Materials & Methods: ISD spectra of the proteins were acquired by performing MALDI-TOF mass spectrometry, both in reflector
and linear mode, on a dried droplet preparation with sinapinic acid or super 2,5-dihydroxybenzoic acid (sDHB) as matrices.
Results: ISD fragmentation induced stepwise sequencing from the smallest obtained c and (z+2) ions, which defined N- and Cterminal peptides. Thus, the overall ISD analysis of the protein resulted in N- and C-terminal sequence determination.
Conclusions: Top-down sequencing by ISD is an effective way to characterize an intact protein at low picomole levels. To provide
comprehensive primary structure information the ISD data can complement traditional peptide mass fingerprinting where Nand C-terminal peptides are often absent in the mass spectrum.
P062
Comparison of human whole saliva samples of patients suffering from chronic periodontitis using different mass
spectrometric approaches
1
2
1
1
1
3
1
A. Murr , M. Kipping , M. Gesell Salazar , E. Hammer , V. M. Dhople , T. Kocher , U. Völker
1
University Medicine Greifswald, Interfaculty Institute for Genetics and Functional Genomics, Department of Functional
Genomics, Greifswald, Germany
2
Waters GmbH, Eschborn, Germany
3
University Medicine Greifswald, Department of Restorative Dentistry, Periodontology and Endodontology, Unit of
Periodontology, Greifswald, Germany
Introduction: The discovery of disease-specific biomarkers in various human biofluids using proteomic techniques has received
major interest over the last decade. Quantitative as well as qualitative fluctuations of a proteomic pattern can reflect the
pathological state or disease-associated changes in patients.
Objectives: In order to compare different mass spectrometric data acquisition technologies, we analyzed the given sample set
E
once in a data-dependent mode and on the other hand with a data independent MS approach. Both methods were applied for
the screening for biomarker candidates in whole saliva from patients suffering from periodontitis in comparison to healthy
subjects.
Proteomic Forum 2015
127
Patients & Methods: The study included 20 patients suffering from chronic periodontitis and 20 healthy age matched subjects.
All 40 subjects were recruited from the population-based cross-sectional Study of Health in Pomerania (SHIP) where a
standardized medical and dental examination was carried out.
LC-MS data was collected on a LTQ Orbitrap Velos (Thermo Fisher Scientific) and a Synapt G2-S instrument (Waters). Data
analysis regarding qualitative and relative quantitative differences was performed with Progenesis QI and Genedata Analyst
software.
Results: In total more than 1600 proteins with at least two proteospecific peptides were identified in whole saliva and verified in
at least 60% of the whole sample set. The overlap of proteins covered by both MS approaches was estimated at roughly 40 % of
all proteins covered and the remaining larger portion of protein data were generated with either of the two LC-MS methods
E
only. In general, with the data-independent MS approach a higher number of peptides was detected and thereby a better
sequence coverage per protein could be achieved.
Quantification of differences in protein intensities between patients and healthy subjects revealed a number of protein
candidates identified in both datasets, but again an even larger number was determined specifically with one of the instrument
used.
Conclusion: Both MS instrument settings revealed a comprehensive set of proteins which displays the complex composition of
whole saliva. The complementary results obtained by the application of different MS methods provide a more extensive base for
the disclosure of disease specific biomarker candidates. . With the approaches used, established periodontitis biomarkers such
as matrix metalloproteinases were covered, but also protein candidates which have not been described so far.
P063
A Quality Control Tool for LCMS based Proteomics: MQQC
1
1
H. Zauber , M. Selbach
1
MDC, Berlin, Germany
In the last decade mass spectrometry-based proteomics became a powerful technique for high-throughput identification and
quantification of proteins. The yield in identified proteins per sample depends on the performance of the system. Regular
system performance checks are therefore crucial for every proteomics lab and are ideally done automatically. However,
available quality control tools either cannot be automated or require expert IT knowledge. Here, we developed MQQC
(MaxQuant Quality Control) as an automated and easy to handle quality control system which takes advantage of MaxQuant. It
provides live post-run reports on critical system metrics from LC, MS and MSMS levels. The performance of individual
instruments is presented in simple status reports and can be traced over time. MQQC is compatible with all vendor output
formats supported by MaxQuant.
P064
Comparison of data-dependent and data-independent acquisition techniques for relative protein quantification
1
1
1
2
1
1,3
A. Dörrbaum , G. Dürnberger , S. Opravil , M. Schutzbier , Z. Demianova , K. Mechtler
1
Research Institute of Molecular Pathology - IMP, Mass Spectrometry and Protein Chemistry, Vienna, Austria
2
Gregor Mendel Institute of Molecular Plant Biology - GMI, Vienna, Austria
3
Institute of Molecular Biotechnology of the Austrian Academy of Sciences - IMBA, Vienna, Austria
Question: Proteomics requires high-throughput technique for quantitative protein analysis. Beside the well-established datadependent acquisition technique shotgun-MS, data-independent acquisition (DIA) techniques, such as SWATH and DIA-MSX,
were described for relative protein quantification. The principle of DIA is based on the usage of wide precursor ion isolation
windows that enable the fragmentation of all precursor ions independent from their abundance. For SWATH-MS, these
precursor isolation windows are sequentially applied over the m/z range of the quadrupole and the resulting fragment ions are
separated and detected by a TOF mass analyzer. DIA-MSX combines precursor ions of two (or more) non-adjacent isolation
windows for fragmentation. The resulting fragment ions are analyzed by an Orbitrap mass analyzer. The aim of this project is to
compare the quantitative results achieved by SWATH, DIA-MSX as well as MS1 based quantification of shotgun data.
Methods: Samples of different complexity were used to compare the sensitivity, accuracy and repeatability of the different
techniques. As simple testing mixtures, two different concentrations (25 ng and 75 ng per injection) of a tryptic digest of UPS 2
Proteomic Forum 2015
128
(Sigma) were used. As complex testing mixtures, the UPS 2 digest was spiked into a digested yeast cell lysate. The samples were
separated by reversed phase nano liquid chromatography prior to the mass spectrometric analysis. SWATH acquisition was
conducted on an AB Sciex Triple TOF 6500 instrument using 10 m/z wide isolation windows at a resolving power of 15,000. For
DIA-MSX (Thermo Q Exactive Plus), two 10 m/z wide isolation windows were multiplexed and fragment ion spectra were
recorded at a resolving power of 35,000. Shotgun data (Thermo Q Exactive Plus) was acquired using a top 12 data-dependent
method with an MS1 resolving power of 70,000. Data processing and quantification was performed with MaxQuant and Skyline.
Results: First preliminary data shows that shotgun-MS provides the highest number of detectable and quantifiable peptides. On
the other hand, the investigated DIA techniques benefit from better repeatability and quantitative accuracy. Thereby, SWATH
enables faster scan rates compared to DIA-MSX. Furthermore, higher numbers of unique peptides were quantified by SWATH
and a higher accuracy and better repeatability was achieved.
Conclusion: The investigated data-dependent as well as data-independent acquisition techniques are suitable for relative
protein quantification. The comparison presents the strengths and limitations of each technique.
P065
Detection of stress responses in lactic acid bacteria Lactobacillus paracasei ssp. paracasei and Lactobacillus brevis by MALDITOF mass spectrometry
1
1
1
1
A.- S. Schott , B. C. Schurr , J. Behr , R. F. Vogel
1
Technical University Munich, Technische Mikrobiologie, Freising, Germany
Lactic acid bacteria (LAB) play important roles in the food industry, as starter cultures or causing microbial food spoilage. Upon
preparation and in foods, LAB are exposed to harsh environmental conditions (extreme pH-values, aridity and osmotic stress).
These extreme conditions and the presence of weak acid preservatives endanger the survival and fitness of microorganisms. In
order to keep their viability, it is of prime importance for LAB to maintain cellular homoeostasis. This is achieved by the
elicitation of sublethal stress and subsequent exploitation of stress response mechanisms.
Proteomic approaches have been used to analyse stress responses, to detect differences between stress-tolerant and -sensitive
LAB followed by the identification of biomarkers and stress proteins related to fitness and survival. Regarding a fast screening of
stress responses and changes in low molecular weight (lmw) proteins, the currently mostly been used approach of 2D gel
electrophoresis has its limitations and forces the evaluation of an easily manageable tool for high sample throughput. Therefore,
we have chosen MALDI-TOF MS analysis as a tool for the fast screening of stress responses along the accumulation of lmw
proteins to identify most responsive conditions as candidates for more elaborate analyses.
In this study Lactobacillus (L.) paracasei ssp. paracasei TMW 1.1434 (F19), a starter culture for the fermentation of dairy
products, was subjected to different responsive conditions (osmotic stress, oxidative stress, temperature stress, pH stress and
starvation), and two beer spoiling L. brevis strains (hop-sensitive and hop-tolerant) were exposed to acid and hop stress. For F19,
it was possible to monitor the expression of stress proteins, identify a specific time point of 60 minutes when the expression of
stress proteins reached its maximum and statistically differentiate types of stress responses into groups. For beer spoiling L.
brevis, the comparison of a hop-sensitive and a hop-tolerant strain revealed differences in acid and hop shock induced stress
responses. In this case the fatty acid biosynthesis-associated acyl carrier protein was identified as most responsive indicating
that the bacterium wants to adjust its cellular membrane composition.
Above the specific results of F19 and beer spoiling L. brevis, these results demonstrate the discriminatory power of MALDI-TOF
MS to characterize stress response dynamics of bacteria and enable a knowledge-based focus for the laborious identification of
biomarkers and stress proteins. Consequently, we generally propose MALDI-TOF MS as an easy and quick method to
characterize responses of microbes to different environmental conditions, to focus efforts of more elaborate approaches on
responsive time points and stress dynamics.
Parts of this work were federally funded by projects BLE 2817400111 and AiF16124N.
Proteomic Forum 2015
129
Plant and Microbial Proteomics I (P066–P079)
P066
Mechanism of drug resistance in Candida albicans
1
C. YingYing
1
The Second Military Medical University, Department of Pharmacy, Shanghai, China
Question: Candida albicans is one of the most important human fungal pathogen, causing diseases varying from superficial
mucosal infections to life-threatening systemic disorders. However, with the increasing use of antifungal agents, drug resistance
occurs and the mechanisms of which remains elusive.
Methods: In this study, a metabonomic method using gas chromatography-mass spectrometry (GC/MS) was developed to
characterize the metabolic profiles of C. albicans cells exposed to AmB. Also, the metabolic profiles of the whole biofilm
developmental phases compared to the planktonic mode in C. albicans were characterized.
Results: Thirty-one differentially produced metabolites between AmB-treated and untreated groups were identified, among
which ten metabolites were upregulated and twenty-one metabolites were downregulated. These differentially produced
metabolites were mainly involved in polyamines synthesis, tricarboxylic acid (TCA) cycle, oxidative stress, glutathione
metabolism, lipid synthesis and glycolysis. Polyamines are polybasic molecules which have diverse roles within cells. Our study
showed a noticeable increase of intracellular polyamines in C. albicans cells upon AmB treatment. Further experiment
demonstrated that polyamines could regulate C. albicans cell sensitivity to AmB. The survival rate of the cells treated by the
combination of polyamines and AmB was significantly higher than that of the cells treated by AmB alone. Presence of ornithine,
the primary source for polyamines biosynthesis, could also attenuate AmB killing of the cells. When AmB and inhibitors of
ornithine decarboxylase were used in combination, the survival rate was remarkably decreased. Combination of AmB and
spermidine resulted in a dramatical reduction of ROS production. Consistant with this, the expression level of SPE1, which
encodes the ornithine decarboxylase that is required for polyamine biosynthesis, was significantly increased upon AmB
treatment. We further constructed the SPE1 deleted and reintroduced strains and found that the spe1Δ mutant cultures showed
high degree of cell killing upon exposure to AmB. These results indicated that polyamines play an important role in protecting C.
albicans cells from AmB killing. The differentially produced metabolites between the biofilm and planktonic specimens at each
time point were mainly involved in lipid synthesis, amino acid metabolism, glycolysis and oxidative stress. More important, we
found that trehalose was accumulated in biofilms. Biofilm formed by tps1D mutant failed to form normal biofilm and was more
sensitive to antifungal agents. These results revealed that trehalose played an important role in biofilm formation and drug
resistance.
Conclusions: Multi-component, phase-specific mechanisms are operative in the drug resistance of C. albicans.
P067
Comparative proteome analysis of differentially expressed proteins in rubber latex of Hevea brasiliensis from Phytophthora
resistant (BPM24) and susceptible (RRIM600) clones
1
1
2
1
P.- O. Havanapan , N. Pongthanom , A. Bourchookarn , C. Kriitanai
1
Institute of Molecular Biosciences, Mahidol University, Nakhonpathom, Thailand
2
Songkla University, Science and Technology, Pattani, Thailand
Introduction: Para rubber tree, Hevea brasiliensis, is the most important species producing commercially viable natural rubber.
The worldwide demand of natural rubber was increasing. However, several high-yielding clones are susceptible to various fungal
species leading to a yield loss in rubber plantations. The widely cultivated clone BPM24 shows highly tolerant to Phytophthora
species, whereas the RRIM600 is found to be susceptible. The rubber latex contains numerous biological-active molecule
discarded as waste in the rubber industry. Especially, some abundant proteins found in non-rubber constituents of clone BPM24
compared to RRIM600 might be involved in defense mechanism and antifungal activity.
Objective: To construct the comparative proteome of rubber latex between clone BPM24 and RRIM600 and protein
identification. To classify the biological functions of differentially expressed protein.
Materials and Methods: Comparative proteomic analysis of Hevea brasiliensis rubber latex from Phytophthora resistant
(BPM24) and susceptible (RRIM600) clones was performed by two-dimensional gel electrophoresis. Relative quantification
Proteomic Forum 2015
130
analysis and tandem mass spectrometry (nanoLC-ESI-MS/MS) were utilized to identify proteins. 1-D and 2-D Western blot
analysis was used to validate mass spectrometric data. Moreover, the functional activity assay of beta-1,3-glucanase and
chitinase was investigated by AZCL-beta-glucan and 4-Nitrophenyl N-acetyl-b-D-glucosaminide as suitable substrates,
respectively.
Results: Quantitative intensity analysis coupling with nano-LC-ESI-MS/MS revealed 16 forms of 12 proteins that were
significantly up-regulated (more than 2.0-fold), whereas 20 forms of 16 proteins were significantly down-regulated in the
tolerant clone BPM24. The altered proteins play important roles in plant defense and carbohydrate metabolism including beta1,3-glucanase and chitinase which was found to be glycoprotein. 1-D and 2-D Western blot analysis confirmed the up-regulation
of beta-1,3-glucanase and chitinase. Moreover, endoglucanase and exochitinase (beta-N-acetylglucosaminidase) activity of clone
BPM24 were found higher than that of RRIM600.
Conclusion: Based on mass spectrometric data coupling to the functional activity, the induction and differential expression of
several proteins in rubber latex may be associated with the tolerance and response of clone BPM24 to infection of Phytophthora
spp. The activity of beta-1,3-glucanase and chitinase may synergistically contribute to enhance fungal tolerance in para rubber
tree.
P068
Comparative proteomic analysis reveals involvement of the Tor pathway in iron regulation of the human-pathogenic fungus
Aspergillus fumigatus
1
1,2
1
3
1
1,2
T. Krüger , C. Baldin , V. Valiante , H. Haas , O. Kniemeyer , A. A. Brakhage
1
Leibniz Institute for Natural Product Research and Infection Biology - Hans Knöll Institute Jena, Molecular and Applied
Microbiology, Jena, Germany
2
Institute for Microbiology, Friedrich Schiller University Jena, Jena, Germany
3
Division of Molecular Biology, Biocenter, Innsbruck Medical University, Innsbruck, Austria
Introduction: Aspergillus fumigatus is an opportunistic human-pathogenic fungus which is the major cause of death among
immunocompromised patients as a result of invasive aspergillosis. The ability of the fungus to sense and to adapt to
2+
environmental stresses is crucial for its virulence. Conserved intracellular signalling pathways including Ca signalling, MAP
Kinase cascades, the cyclic AMP, and the Tor pathway are driving forces of its adaptivity. Sequence alignments revealed that A.
fumigatus possesses a unique gene encoding a putative Tor kinase that can act in two structurally and functionally distinct
multiprotein complexes TORC1 and TORC2.
Objectives: As the role of the putative Tor pathway has not been elucidated in filamentous fungi until now, it was the aim of this
study to gain deeper insights into the impact of the tor signalling pathway in A. fumigatus applying a comparative proteomic
analysis (wild-type vs. tor mutant strain).
Material and Methods: Since the deletion of the putative tor gene is lethal for A. fumigatus, a conditional lethal mutant was
generated by replacing the endogenous tor promoter by an xylose-inducible promoter. Both wild-type and xylp-tor mutant were
incubated for 4 h in AMM without xylose after a pre-cultivation of 16 h in AMM with 1% (w/v) xylose. To identify and relatively
quantify differentially regulated proteins, two distinct proteomic techniques have been applied: (i) 2D-gel electrophoresis and
MALDI-TOF/TOF analysis. (ii) LC-MS/MS analysis of TMT 6plex-labeled peptides for relative reporter ion quantification (3
biological and 3 analytical replicates).
Results: 2D-gel analysis under tor repressing conditions revealed fifty-three differentially regulated protein spots (>1.5 fold). In
the gel-free LC-MS/MS approach using reporter ion quantification, 2095 proteins were identified (≥2 peptides, <1% FDR) of
which 29 were up- and 55 were significantly downregulated (>1.5 fold). Besides proteins that are known to be affected by the
Tor kinase in other organisms, we surprisingly found proteins involved in iron sensing and storage to be affected by Tor such as
the ornithine/acetylornithine aminotransferase ArgD and the coproporphyrinogen III oxidase HemF.
Conclusion: In A. fumigatus Tor represents a central regulatory node controlling proteins involved in nutrient sensing, stress
response, cell cycle progression, protein biosynthesis as well as degradation and especially low iron adaptation independent of
the transcription factor and iron regulator HapX. Our data also indicate that Tor is a major kinase regulating protein import into
the mitochondria, and consequently respiration and ornithine metabolism which is in turn involved in siderophore biosynthesis.
Proteomic Forum 2015
131
P069
Metaproteomics: analysing the active fraction of the gut and rumen associated microbiota
1
1
1
J. Seifert , B. Tilocca , S. Deusch
1
University of Hohenheim, Feed-Gut Microbiota Interaction, Stuttgart, Germany
The microbiota of the animal’s gastrointestinal tract plays an essential role for the growth and health of the animal. The
identification and assessment of the protein inventory of the microbiota is named metaproteomics and displays a valuable
interface between gene-based analysis and metabolomics studies. Metaproteomic studies of gut samples or feces are
challenged by a high fraction of eukaryotic proteins, which originate from feed particles and host cells and are co-extracted
during the sample preparation procedure. Samples from the crop and ceca of broilers and rumen samples were used to improve
sample preparation strategies and to give a first impression of the active microbial fraction in these gut sections.
Protein extracts from digesta samples from crop and ceca from broilers originated from a dietary treatment based on a two
maize-soybean meal basal diets varying in the amount of monocalcium phosphate (MCP) and phytase were analysed by gelbased shotgun metaproteomic approach. Solid (RS) and liquid (RF) rumen samples were obtained from a fistulated HolsteinFrisian cow. Two protocols were tested to separate prokaryotic cells from feed particles in the RS samples and one of the RF
sample. The resulting proteins were analysed by a gel-based shotgun metaproteomic approach.
The protein identifications of the broiler samples showed that active fraction of the microbiota of the crop had a low bacterial
diversity with Lactobacillaceae as the dominant family. The ratio of the prokaryotic to eukaryotic proteins of the caecal samples
was 3:1 and the results showed that the composition of the microbiota changed among diets.
The analyses of the rumen metaproteome showed a useful application of the cheesecloth gauze filtration for the RS samples to
achieve a suitable protein identification ratio. About 2300 prokaryotic proteins could be identified in the RS samples showing a
distinct phylogenetic distribution compared to the 1500 prokaryotic proteins identified in the RF sample.
The data showed the added value of metaproteomic datasets to obtained insights in the active fraction of the microbiota of the
gastrointestinal tract and to detect possible changes regarding feeding strategies. Metaproteomic data of the broiler study
highlighted the effect of dietary MCP and phytases on the composition of the chicken microbiota. The optimization of the
protein extraction procedures of rumen samples was successfully in improving the identification ratio of prokaryotic to
eukaryotic proteins although the total number of identifications still only depicts a sparse part of the actually active microbiota.
Anyhow, technical progress in mass spectrometry and an increasing availability of reference sequences enhance appropriate
application of metaproteomic approaches for microbiota research in animal nutrition.
P070
Mining cross-reactive allergens from Coconut and Date palm pollen through homology driven proteomics.
1
1
B. Saha , S. Gupta Bhattacharya
1
Bose Institute, Division of Plant biology, Kolkata, India
Introduction: Allergies to pollen grains are common. Coconut and Date Palm pollen grains have been found to cause severe
allergic symptoms in susceptible individuals in Kolkata, India. Identification of allergenic proteins is thus required for structural
diagnosis as well as development of hypoallergenic vaccines.
Objectives: Identification of allergenic proteins from two palm pollen grains through mass spectrometry, cross reactive allergen
purification and characterization.
Patients and materials and methods: Ten patients with positive skin prick test and high specific IgE with coconut and date palm
pollen were selected for the study abiding ethical rules. Crude pollen extract was profiled in 2D gels of broad as well as narrow
range pH gradients. 2D immunoblot with patient specific sera was performed. Periodic acid Schiff staining and deglycosylation
study was done to ascertain involvement of glycosylation in allergenecity. Mass spectrometric identification of allergic spots by
MALDI was done using a layered proteomic approach involving manual de novo sequencing and homology based search.
Common allergen amongst the two pollen grains was purified through affinity, ion exchange and gel filtration chromatography.
Purified allergen further confirmed by MALDI Tof/Tof. ELISA inhibition, Immunoblot inhibition, Histamine release assay, basophil
marker tests elucidated atopic potency of purified allergen.
Proteomic Forum 2015
132
Results: 2D profiles showed better resolution in 4-7pH gradient in both samples. Immnublot with each patient sera helped to
recognise IgE reactive spots which was identified through MALDI Tof/Tof. Manual de novo sequencing and homology search
helped to ascertain proteins which could not be found by traditional MASCOT based search. A heat map generated with absence
and presence of IgE reactive spots categorized the allergens into clusters distinguishing major and minor allergens. Beta
conglycinin has been found to be present in both the species as major allergen and hence may be responsible for cross
reactivity. Other allergens identified were enolase, ascorbate peroxidase, isoflavone reductase in coconut and carbonic
anhydrase, beta galactosidase in date palm. Beta conglycinin, a glycoprotein was purified by chromatographic methods showed
two bands at 22 and 25 Kd respectively. Homotrimeric nature(~150Kd) of this protein is ascertained through SDS PAGE and Gel
filtration chromatography. A small allergenic isoform of this protein is observed in 2D gel. ELISA inhibition and blot inhibition
predicts presence of both linear and conformational epitopes. 9ug of beta conglycinin from date palm was able to inhibit 50% of
IgE reactivity in coconut pollen. CD63 marker assay by FACS showed an increase after challenge with betaconglycinin.
Conclusion: De novo sequencing coupled with MASCOT search helped to increase the sequence coverage and thus confidence.
Beta conglycinin is reported to be an allergen from pollen grains for the first time and could be used as for vaccine development
against palm pollen allergies.
Figure 1
P071
Physiological role of ETHE1 for A. thaliana seed metabolism
1
1
2
1
1
C. Lorenz , T. Busch , D. Gerhard , H.- P. Braun , T. Hildebrandt
1
Leibniz University of Hannover, Plant Proteomics, Hannover, Germany
2
University of Canterbury, School of Mathematics and Statistics, Christchurch, New Zealand
The sulfur dioxygenase ETHE1 (ETHYLMALONIC ENCEPHALOPATHY PROTEIN1) catalyzes the oxidation of persulfides in the
mitochondrial matrix. Mutations in the ETHE1 gene cause the fatal metabolic disease ethylmalonic encephalopathy affecting the
brain, the gastrointestinal tract, and the peripheral vessels [1]. In plants, ETHE1 is involved in sulfur and branched-chain amino
acid catabolism [2]. The metabolic function of the sulfur dioxygenase during seed production is still unknown, however it is
clearly essential since ETHE1 knockout is embryo lethal in the model plant Arabidopsis thaliana [3]. We investigated the
physiological role of ETHE1 in seeds using an Arabidopsis knockdown mutant with a residual sulfur dioxygenase activity of about
1 %, which was sufficient for plant viability but led to pronounced growth defects throughout the life cycle. Embryo
development was severely delayed in the mutant plants and resulted in smaller seeds with a decreased germination rate.
Proteomic Forum 2015
133
Shading of the siliques further enhanced the defects in seed production, which is in agreement with the postulated function of
ETHE1, since amino acid catabolism serves as an alternative energy source when photosynthesis is disrupted. We will present a
detailed analysis of the proteome adaptations in ETHE1 deficient seeds during early embryo development based on a mass
spectrometry approach. In addition, the role of amino acid catabolism in seeds will be discussed.
[1] Tiranti V., D’Adamo P., Briem E., Ferrari G., Mineri R., Lamantea E., Mandel H., Balestri P., Garcia-Silva M. T., Vollmer B. et al.
(2004) Ethylmalonic encephalopathy is caused by mutations in ETHE1, a gene encoding a mitochondrial matrix protein. Am J
Hum Genet 74: 239-252
[2] Krüßel, L., Junemann, J., Wirtz, M., Birke, H., Thornton, JD, Browning, L.W., Poschet, G., Hell, R., Balk, J., Braun, H.P. and
Hildebrandt, T.M. (2014): The mitochondrial sulfur dioxygenase ETHE1 is required for amino acid catabolism during
carbohydrate starvation and embryo development in Arabidopsis thaliana. Plant Physiol. 165, 92-104
[3] Holdorf M. M.,Owen H.A., Lieber S.R., Yuan L., Adams N., Dabney-Smith C., Makaroff C.A. (2012) Arabidopsis ETHE1 encodes
a sulfur dioxygenase that is essential for embryo and endosperm development. Plant Physiol 160: 226-236
P072
Reduction of the signal suppression effects in MALDI
1
2
2
1
O. Sedo , A. Nemec , L. Krizova , Z. Zdrahal
1
CEITEC Masaryk University, Research Group Proteomics, Brno, Czech Republic
2
National Institute of Public Health, Laboratory of Bacterial Genetics, Prague, Czech Republic
Question: MALDI MS analysis is accompanied with undesirable suppression effects. These complicate the detection of several
components of mixtures, especially those with relatively higher molecular weight. In this study, we assessed a matrix solution
capable of substantial modulation of the suppression effects.
-1
Methods: The matrix solution based on ferulic acid (12.5 mg.mL solution in water:acetonitrile:formic acid 50:33:17, v/v),
adapted from the paper of Madonna et al. (Rapid Commun. Mass Spectrom. 14, 2220-2229), was used for the analysis of
mixtures of synthetic proteins, whole-cell extracts of closely related bacterial strains from the Acinetobacter calcoaceticusAcinetobacter baumannii (ACB) complex, and protein extracts from cereals. MALDI-TOF mass spectra were obtained by using an
UltrafleXtreme instrument (Bruker Daltonik).
Results: The reduction of signal suppression effects was observed for all types of tested samples, where the detection of several
high-mass components of the analyzed mixture was enabled by using the selected matrix. The acidity of the solvent was
determined as the crucial factor for the substantial change in the ionization properties of ferulic acid. The discrimination
between strains of the ACB complex was improved owing to the detection of more species-specific signals, compared to those
obtained by using standard sample preparation protocol for bacteria.
Conclusions: Strongly acidified solution of ferulic acid is an excellent matrix for overcoming the suppression effects, which can
be used for the detection of high-mass signals from protein mixtures.
This work was supported by the Czech Science Foundation, project No. 13-26693S, and the project “CEITEC - Central European
Institute of Technology” (CZ.1.05/1.1.00/02.0068) from European Regional Development Fund.
P073
Nitrogen deficiency responses of Solanum tuberosum genotypes revealed by proteomic and phosphoproteomic approaches.
1
2
1
1
A. Jozefowicz , S. Kaspar , A. Matros , H.- P. Mock
1
IPK- Gatersleben, Physiology and Cell Biology, Gatersleben, Germany
2
Bruker Daltonics, Bremen, Germany
Introduction: Nitrogen is the major macronutrient limiting crop plant yield worldwide, necessitating fertilization for crop plants
cultivation. Nitrogen as a fertilizer is one of the most expensive nutrients, and its use can result in nitrate leaching and
eventually strong environmental pollution. Therefore, breeding of crop plants with higher nitrogen use efficiency (NUE) is of
great importance for future agronomical performance.
Objective: In our study we are focused on two aspects of N-limitation responses. Our main goal is an investigation of early stress
signaling in potato roots by the use of phosphoproteomic approaches. Secondly, proteome studies with the emphasis on plasma
Proteomic Forum 2015
134
membrane (PM) composition, of two contrasting in NUE genotypes, are performed to investigate adaptational mechanism of
plants towards N- deficiency.
Material and Methods: Plants are cultivated in an in vitro system, allowing growth under controlled and reproducible
conditions. For investigation of early signaling changes, plants are subjected to N-starvation medium. Roots are collected after
10 min and 1h and also after 24 and 72h to study later responses. Phosphopeptides are enriched using TiO 2 beads and analyzed
by nanoLC-ESI-QTOF MS. To study further adaptational mechanisms, PM of root tissue is enriched using aqueous two-phase
partitioning. PM proteins together with those of crude extract, soluble and microsomal fractions are analyzed by mass
spectrometry.
Results: In our preliminary experiment label- free quantification of phosphopeptides revealed different regulation patterns
between short- and long-time treatment with about three times more phosphopeptides showing significant regulation in the 10
min then in the 72h treatment. Further studies are under way to identify candidate phosphoproteins involved in NUE.
In our second approach full proteome comparison of the N efficient and inefficient cultivars revealed significant difference in
abundance of more than 100 proteins. Many proteins associated with primary metabolism were affected. Most of the proteins
taking part in C- metabolism were more abundant in the efficient genotype. Proteins connected to glutamate metabolism were
up- regulated under stress conditions in both genotypes, whereas increase in glutamine synthetase was observed only for
efficient one. N-limitation was found to have a high impact on the several enzymes of secondary metabolism. A
phenylpropanoid glucosyltrasferase and a putative cinnamyl alcohol dehydrogenase were higher abundant in the tolerant
genotype.
Conclusions: Our preliminary results revealed various reaction patterns for the short- and long-term N deficiency treatment.
Additional, protein pattern comparison of nitrogen use efficient and inefficient genotypes revealed significant differences in the
abundance of enzymes connected to primary and secondary metabolism.
P074
Host targets of luteoviral protein in plants using BioID
1
1
M. Kraner , U. Sonnewald
1
Friedrich-Alexander-Universität Erlangen-Nürnberg, Biology, Chair of Biochemistry, Erlangen, Germany
Introduction: In higher plants the symplastic communication between neighbouring cells occurs through plasmodesmata (PD).
PD are plasma membrane-lined channels that traverse the cell walls and facilitate the trafficking of small metabolites as well as
macromolecules involved in local and systemic signalling. This system is also exploited by plant viruses, which spread from cell to
cell by encoding for specialized movement proteins (MP).
Objectives: Focus of our ongoing research is the interaction between the luteoviral movement protein (MP17) of the Potato
leafroll virus (PLRV) with host factors enabling the targeting and accumulation of the protein at PD. This hopefully will lead to the
identification of host factors required for PD targeting of MP17 and PD intrinsic proteins which are essential for progressive viral
infection.
Material & Methods: Due to the strong binding of MP17 to PD, harsh conditions have to be chosen to solubilise this protein.
This renders immuno-precipitation of putative interacting partners difficult. To circumvent this problem we made use of the
biotin ligase BirA from E.coli (Roux et al., 2012). This technique (called BioID ) allows the biotinylation of proximal proteins in
vivo. In order to identify proteins in the vicinity of MP17, the protein was fused to BirA carrying a point mutation (R118G). The
mutated BirA enzyme will convert biotin to reactive biotinyl-AMP. Activated biotinyl-AMP will immediately react with free
epsilon amino groups of accessible lysine residues. Therefore, proteins which are close to MP17 will be preferentially
biotinylated in vivo. Since neighbouring proteins are covalently tagged, subsequent protein isolation can be done under
denaturing conditions. Biotinylated proteins are subsequently purified by streptavidin-magnetic beads and bound proteins are
identified by LC-MS/MS (See Attachment 1).
Results: To validate putative interaction partners of MP17, resulting from MS data, we chose the dominant candidate Heat
shock protein 70 for further investigations. For intracellular localization of GFP-HSP70 we used transient expression with and
without MP17-His in Nicotiana benthamiana leaves. In the presence of MP17 the cytosolic HSP70-GFP is recruited to
plasmodesmata (see Attachemt 2).
Proteomic Forum 2015
135
Conclusion:Using the intrinsic function of MP17 to target PD in living plant cells, we were able to bring the promiscuous biotin
ligase BirA* to PD. In the presence of ATP, BirA* converts biotin to reactive biotinyl-AMP which covalently labels plant proteins
in close proximity of the BirA*-fuison protein. This lead to the identification of cytosolic HSP70 which is recruited to PDs by
MP17. Further investigations and optimized protocols will advance our knowledge of PD proteins.
Figure 1
Figure 2
P078
Semi-in vivo tobacco pollen tube secretome
1,2
3
3
3
1
1
3,4
1,2
D. Potesil , S. Hafidh , J. Fila , V. Capkova , E. Srbova , K. Stejskal , D. Honys , Z. Zdrahal
1
Masaryk University, CEITEC-MU, Brno, Czech Republic
2
Masaryk University, Faculty of Science, Brno, Czech Republic
3
Academy of Sciences of the Czech Republic, Institute of Experimental Botany, Prague, Czech Republic
4
Charles University in Prague, Faculty of Science, Prague, Czech Republic
The delivery of non-motile male sperm cells for fertilization depends on precise signaling between the growing pollen tube and
the female gametophyte conveyed by secreted ligand molecules. We used semi-in vivo technique to reconstruct tobacco pollen
tube growth through the pistil enabling us to characterize semi-in vivo tobacco pollen secretome.
Biological replicates of media with secreted proteins (and negative controls from un-pollinated pistils) were subjected to filter
aided sample preparation (FASP) technique using 10kDa cut-off membranes. Resulting peptides after trypsin digestion were
separated using nanoscale reverse phase liquid chromatography (RSLCnano system, Thermo) and online analyzed in Orbitrap
ELITE mass spectrometer (Thermo). MS/MS data were searched against concatenated protein database containing in-house
protein tobacco database and selected taxonomies from UniRef100 database. Quantitative comparison was made using top3
label-free algorithm.
On average more than 800 protein groups were identified in positive samples. Majority of those (>500) were found not to be
present in negative control samples or were present in quantitatively higher amounts and are likely to be secreted by the pollen
tube during in the course of ovule targeting and fertilization. Their analysis provided evidence that pollen tube growth through
the pistil endures diverse signaling pathways including small GTPase-mediated signal transduction, membrane fused secretion
Proteomic Forum 2015
136
and cell wall matrix remodeling. Proteins belonging to plant defensin-like family (DEFL), LORELEI-like GPI-anchored protein 3
(LLRG3), Thionin-like protein, plant lipid transfer proteins (LPTs), small GTPases as well pectinases were among the identified
pollen tube secreted proteins.
This work was supported by the project CEITEC (CZ.1.05/1.1.00/02.0068) from European Regional Development Fund, by Czech
Science Foundation (projects No. GBP206/12/G151 and P501/11/1462) and Czech Ministry of Education, Youth and Sports
(LD14109).
References:
Hafidh S, Potěšil D, Fíla J, Feciková J, Capková V, Zdráhal Z, et al. In search of ligands and receptors of the pollen tube: the
missing link in pollen tube perception. Biochem Soc Trans. 2014;42:388-94.
Johnson MA. Plant Reproduction: Teaching a New Language of Love. Current Biology. 2012;22:R528-9.
Okuda S, Tsutsui H, Shiina K, Sprunck S, Takeuchi H, Yui R, et al. Defensin-like polypeptide LUREs are pollen tube attractants
secreted from synergid cells. Nature. 2009;458:357-61.
Grossmann J, Roschitzki B, Panse C, Fortes C, Barkow-Oesterreicher S, Rutishauser D, et al. Implementation and evaluation of
relative and absolute quantification in shotgun proteomics with label-free methods. Journal of Proteomics. 2010;73:1740-6.
P079
FOOD SAFETY: IMPROVE KNOWLEDGE OF SECRETOME DURING BACTERIAL COMPETITION
1,2
2
2
3,4
2
5
3,4
2
P. Roncada , I. Alloggio , C. Piras , V. Greco , A. Soggiu , N. Losio , A. Urbani , L. Bonizzi
1
istituto sperimentale italiano lazzaro spallanzani, Milan, Italy
2
University of Milan, Department of Veterinary Science and Public Health, Milan, Italy
3
Fondazione Santa Lucia – IRCCS, , Roma, Italy
4
Università di Roma Tor Vergata, Dipartimento di Medicina Sperimentale e Chirurgia,, Roma, Italy
5
Istituto Zooprofilattico Sperimentale della Lombardia e dell' Emilia, Brescia, Italy
Introduction: Most microorganisms face a constant battle for resources. A great numbers of microbes are present in all but the
most rarefied environments. New molecular methodologies such as metagenomic sequencing and deep microbial tag
sequencing have revealed incredible microbial diversity. Therefore, the potential pool of microbial competitors is vast, and a
wide range of mechanisms can be responsible for the emergence and radiation of dominant microbial populations. (1) The same
happens in food, and it could be possible to explore this social communication to improve food safety.
Objectives: Explore the molecular mechanisms of bacterial competition of Lactococcus Lactis vs Listeria monocytogenes using
several in-vitro experiments.
Materials & Methods: Two strain of Lactococcus lactis and one strain of Listeria monocytogenes were cultivated in appropriate
medium cultures (BHI), alone and together (competion). Filtrated coltures (SECRETOME)(2) were liophylized and resuspended
for proteomics analysis. In order to characterize each secretome, shotgun analysis was performed on nano UPLC-MS system
(Waters). LC-MS data were processed using the software ProteinLynx Global Server and protein identifications were obtained
searching against curated databases (UniProtKB / Swiss-Prot Protein Knowledgebase) with taxonomical restrictions: respectively
Listeria Monocytogenes and Lactococcus lactis for monocoltures, Listeria + Lactococcus together in condition of competition.
This type of method has allowed, with the use of an internal standard of known concentration added to the samples, the
identification and the quantification of secreted proteins from each medium.
Results: Shotgun MS analysis (and also 2-DE coupled to Maldi-TOF/TOF) highlighted, during competition, the expression by
Listeria of the moonlighting protein Enolase and some proteins involved into the replication of bacteria. We also report the
down regulation of some listeria virulence factors during competition.
Conclusion: The adaptation of listeria and lactococcus secretome in medium cultures, added to the information derived from
the different composition of secretome of Lactococcus, wild type and isolated from cheese, are very interesting and give new
insights of biology of signaling during bacteria competition. These results are the first step of a targeted validation ‘in vivo’,
directly in cheese.
References: 1.Hibbing ME, Fuqua C, Parsek MR, Peterson SB. Bacterial competition: surviving and thriving in the microbial
jungle. Nature reviews. Microbiology 2010;8(1):15-25. doi:10.1038/nrmicro2259. 2. Ranganathan S, Garg G. Secretome: clues
into pathogen infection and clinical applications. Genome Medicine 2009;1(11):113. doi:10.1186/gm113.
Proteomic Forum 2015
137
Work supported by CCM expo 2015project- Ministry of HEALTH-LB
Post-translational Modifications I (P092–P107)
P092
Analysis of posttranslational cysteine modifications in a neural differentiation model by Redox Proteomics
1
1
1
K. Nau , K. Stühler , G. Poschmann
1
Molecular Proteomics Laboratory, Heinrich-Heine-Universität Düsseldorf, Düsseldorf, Germany
Dynamic changes of cellular redox status during neurogenesis allow a tight and rapid control of cell proliferation, differentiation
and death. Regulatory cysteine switches are key players in redox sensitive pathways, as the functional thiol moiety occurs in a
number of chemotypes and oxidation states. Although in the last years enormous progress has been made in the detection of
functional relevant cysteine oxidation events, still the analysis of cysteines’ redox state is challenging and there is a need to
improve bioanalytical tools and model systems.
Here, we evaluated different proteomic strategies for the analysis of a model system to identify cysteine switches relevant for
early events of neurodifferentiation. We examined global changes in the proteome of human neuroblastoma cell line SH-SY5Y at
five different time points (0, 1, 3, 8, 48 hours) in the early stages of retinoic acid induced neurodifferentiation using label-free
liquid chromatography mass spectrometry (LC-MS). A more detailed view on the total reversible cysteine oxidations three hours
post-induction was gained by a modified two-dimensional fluorescence difference gel electrophoresis (2-D DIGE) approach in
combination with LC-MS.
In the global proteome analysis of stimulated SH-SY5Y cells, 114 out of 2055 quantified proteins showed retinoic acid treatment
dependent abundance changes whereby the most pronounced changes occurred at later time points. At three hours postinduction three proteins with increased and two with decreased abundance could be observed, these include proteins involved
in neuronal adhesion and the redox sensitive pentose-phosphate shunt and MAPK pathway. Using the 2-D DIGE approach, we
were able to identify eleven proteins with changes in the cysteine oxidation state at the same time point. These proteins were
involved e.g. in the MAPK signaling pathway, glycolysis and redox regulation of the cell. This suggests that redox processes are
involved in early stages of retinoic acid induced neurodifferentiation.
P093
Biosensing of Extracellular Complexes of Glycosylphosphatidylinositol-Anchored Proteins and Phospholipids as Biomarkers for
Stress-Induced Diseases
1
1
G. Müller , M. Tschöp
1
Helmholtz Center Munich, Institute for Diabetes and Obesity, Garching-Hochbrück, Germany
Question: A novel “phenomenological” approach may lead to biomarkers for the prediction and stratification relevant of stressinduced disorders, such as type 2 diabetes (T2D) with higher informative value than traditional phenotypic, peptidic and
genotypic ones. It relies on the demonstration and biophysical characterization of extracellular complexes in plasma, urine or
saliva which harbor glycosylphosphatidylinositol-anchored membrane proteins (GPI-APs) and phospholipids and possibly
additional proteins and lipids (ECGAPP): (i) ECGAPP have been shown to be released from the surface of metabolically relevant
cells through non-classical secretory mechanisms in response to metabolic stress as is prevalent during T2D [1,2]; (ii) ECGAPP are
assumed to differ in level, morphology (see Figure), structure and biophysical properties between distinct (pre-)diabetic states;
(iii) GPI-APs have been shown to exhibit high susceptibility for release in ECGAPP from the surface of mammalian cells in
response to metabolic stress [1,2]. However, the presence of ECGAPP in body fluids of T2D patients has not been studied so far,
presumably due to conceptual restrictions (i.e. reductionistic and holistic thinking) and technological challenges.
Methods: To overcome these hurdles, a novel type of chip-based biosensor will be used for the detection and biophysical
characterization of ECGAPP which relies on the generation of horizontal surface acoustic waves (SAW) within the gold surface of
a microfluidic four-channel chip. Any interaction of (macro)molecules with the gold surface will result in corresponding changes
in the shape of the SAW, altering both their frequency and amplitude.
Results: These alterations reflect mass loading (i.e. binding of ECGAPP) to and biophysical properties (i.e. size/shape depending
on protein composition as well as viscoelasticity/rigidity depending on lipid composition and open/closed configuration) at the
chip surface. The major advantages of the SAW vs. SPR biosensor rely on the possibility of measurement of large (lipid-
Proteomic Forum 2015
138
containing) macromolecules (micelles, nanodiscs, vesicles) even in the presence of serum [3]. Preliminary data indicate that
signatures recorded in course of successive binding of ECGAPP to the chip surface via specific interaction with alpha-toxin (for
detection of GPI-APs) and annexin (for detection of phospholipids) significantly differ between plasma from normal and (pre-)
diabetic mice.
Conclusions: Albeit SAW biosensing per se does not enable the delineation of the type of ECGAPP, the SAW signatures will be
characteristic for the overall contents of all ECGAPP, either as summation or subtraction signals of high informative value.
References
[1] Müller G, Jung C, Wied, S, Bimer-Daub, G, Frick, W (2010) Br J Pharmacol 160, 878-891.
[2] Müller, G, Schneider, M, Biemer-Daub, G, Wied, S (2011) Cell Signal 23, 1207-1223.
[3] Gronewold, T (2007) Anal Chim Acta 603, 119-128.
Figure Legend: Morphology of Already Identified and Hypothetical ECGAPP
Figure 1
P094
Using SWATH™ Acquisition for Characterization and Quantification of the Epigenetic Histone Modifications
1
2
2
2
2
3
J. Dojahn , S. Mollah , C. Hunter , E. Johansen , L. Xiong , B. A. Garcia
1
AB SCIEX, Darmstadt, Germany
2
AB SCIEX, Redwood City, United States
3
University of Pennsylvania School of Medicine, Philadelphia, United States
Introduction & Objectives: Histone post-translational modification (PTMs) sites regulate gene transcription, thus making reliable
quantification a high priority in epigenetic studies. Quantification of histone PTMs is performed by shotgun proteomics favoring
discovery of novel/low level PTMs, or targeted analyses (SRM) of known PTM sites. Each workflow has strengths and
weaknesses for PTM quantification. Data independent acquisition (DIA) for comprehensive data generation combined with
targeted data processing has recently been demonstrated to provide very high quality quantitative data. The advantages of this
approach for targeted PTM quantification include no upfront assay development, quantitative data on all analytes and no
dynamic exclusion of isobaric peptides. In this study, we develop a SWATH™ acquisition platform for quantitating histone.
Materials & Methods: Histones were acid extracted from HeLa cells. A custom-made synthetic heavy labeled modified peptide
library was spiked into the sample. SWATH™ acquisition of the sample was performed using a QqTOF system interfaced to a
nanoflow source and a chip based LC system operated at various flow rates. Replicate analysis was performed for these samples
to obtain characterization and quantitation statistics. Data was processed using either commercial or in-house software.
Results & Conclusions: Results from both the asynchronous non-treated and butyrate treated samples run in a normal datadependent mode (DDA) was used to generate a spectral ion library for the histone samples. A key issue in analyzing histone
peptide mixture is the presence of various isoforms of peptides, primarily from PTMs on the same peptide, but localized on
different amino acids. Quantitation of these isoforms in a data-dependent mode using MS1 quantitation has been very
challenging due to co-elution of isobaric forms. The peptide, GKGGKGLGKGGAKR from histone H4 is found to be acetylated at 4
Proteomic Forum 2015
139
sites: Lys2, Lys5, Lys9 or Lys13. Extracted ion chromatogram (XIC) of its precursor mass of m/z 768.9465 resulted in 2 peaks and
a third that partially overlapped. This made it difficult to distinguish and quantify each isoform based on MS1. However, with
SWATH acquisition, all the MS/MS fragments are collected in one data set, so unique/combination of MS/MS fragments of each
isoform were extracted post-acquisition and used for sequence assignment and quantification even if they co-eluted. Narrowing
Q1 Window thru Variable Window SWATH™ Acquisition helps to increase specificity, which leads to even better quantitation,
especially for low abundant compounds. Based on the SWATH analysis, Lys5 and Lys9 peaks co eluted and partially overlapped
with Lys13 peak, which would make it impossible to be quantified by MS1 analysis. Replicate SWATH analysis resulted on
average a 10% CV for the peptide fragments, quality rivaling an MRM approach. We will present results from more of these
histone isoforms. This work illustrates the advantage of using SWATH analysis for characterization of histone modifications.
P095
PFP-based separation improves characterization of histone modification status by mass spectrometry
1
2
2
1
G. Lochmanová , T. Panić-Janković , A. Fichtenbaum , Z. Zdráhal , G. Mitulović
1
Masaryk University, Brno, Czech Republic
2
Medical University of Vienna, Vienna, Austria
2
Question: Histone post-translational modifications (PTMs) are essential epigenetic regulators of gene expression, replication and
repair. As the imbalance or disturbance in the modification status often leads to the disease formation, the development of
methods for histone characterization is of critical importance. Purification and fractionation of intact proteins is routinely
accomplished using RP-HPLC, although the resolution is usually limited. Similarly, histone peptides have been separated on RP
prior MS/MS analysis. In the present study, we were searching for an alternative stationary phase to hydrocarbon aliphatic
chains in order to improve separation of intact histones as well as histone peptides prior LC-MS/MS.
Methods: Pentafluorophenyl (PFP) phase bound to core-shell silica was used for separation of intact histone proteins extracted
from MEC-1 chronic lymphocytic leukemia (CLL) cell-line as well as for histone peptides after trypsin and Glu-C digestion.
Enrichment of histone phosphopeptides was performed on monolithic column with immobilized nano-TiO2 particles (BIASeparations). LC-MS/MS analysis of histone peptides was performed using LTQ Velos Ion trap (Thermo Fisher Scientific) or
Amazon-Speed ion ETD trap (Bruker Daltonik) mass spectrometers.
Results: We demonstrate the impact of separation before MS on final results. Using PFP stationary phase, higher efficiency of
separation at fast retention time of main histone variants has been achieved compared to conventional RP-HPLC. High sequence
coverage has been obtained for histone tryptic peptides of all histones tested showing greater propensity of PFP to retain short
peptides. In addition, the impact of separation before MS on detection of phosphopeptides was examined by comparison of PFP
residues and hydrocarbon aliphatic chains.
Conclusions: Application of PFP-based separation provides a powerful tool for enhancing feasibility of histone modification
status characterization by MS.
Acknowledgment: This work was supported by program CZ.1.07/2.3.00/30.0009 co-financed from European Social Fund and the
state budget of the Czech Republic, by CZ.1.05/1.1.00/02.0068 financed from European Regional Development Fund, by
CZ.1.07/2.3.00/20.0189 and P206-12-G151; from European Union (FP7, THEME [HEALTH.2011.1.1-3], Prot-HiSPRA), Grant
agreement no: 282506.
P096
Comparison of ERLIC with a new mixed-mode strategy using WCX-RPC for phosphopeptide separation prior to mass
spectrometry
1
1
1
C. Regl , L. Weilnböck , C. Huber
1
Universität Salzburg, FB Molekulare Biologie, Salzburg, Austria
Protein phosphorylation is a transient and low abundant post-translational modification, thus specific and sensitive enrichment
strategies are required to perform phosphoproteome analysis. In addition to affinity based methods using metal ions or metal
oxides, electrostatic repultion liquid chromatography (ERLIC) has been established for the fractionation of phosphorylated
peptides [1].
Proteomic Forum 2015
140
For ERLIC we used a PolyWAX LP mixed-mode column (100 x 2.1 mm i.d.), consisting of a cross-linked coating of linear
polyethyleneimine on silica gel [2]. In comparison we tested a combination of weak cation-exchange chromatography (WCX)
with reversed-phase chromatography (RPC), using a mixed-mode WCX-1 column (150 x 2.1 mm i.d.), composed of terminal
carboxyl groups on octadecyl chains, derivatized onto silica gel.
The idea of this WCX-RPC strategy is to start the separation at a low pH and low salt and organic solvent concentrations. At these
conditions hydrophobic interactions as well as electrostatic interactions with both, phosphorylated and non-phosphorylated
peptides are enabled, as the carboxylic groups of the stationary phase should be fully protonated. An increase in the pH leads to
a de-protonation of the carboxylic groups on the stationary phase, causing on the one hand amplification of electrostatic
interactions due to the positively charged N-terminus and C-terminal lysine or arginine of tryptic peptides with the stationary
phase, and on the other hand provokes an electrostatic repulsion of phosphopeptides. A linear gradient in both, increasing salt
and organic solvent concentration allows an effective separation of phosphorylated from non-phosphorylated peptides.
We fractionated HeLa whole cell lysates using ERLIC and WCX-RPC, and subsequently separated the fractions in an additional
orthogonal dimension using ion-pair reversed-phase chromatography on a monolithic polystyrene divinylbenzene column (150 x
0.2 mm i.d.) and analysed via a nanoESI-coupled linear ion trap orbitrap mass spectrometer. Executing three replicate analyses
each, we were able to identify an average of 368 phosphorylated peptides in 20 µg of a Hela lysate using the new WCX-RP
strategy, which is very similar to the 398 phosphopeptides that were identified using the established ERLIC method.
P097
Regulatory signaling pathways underlying the dynamic translocation of mucins upon carbachol stimulation.
1
1
1
1
E. Berger , H. Schneider , T. Pelaseyed , G. C. Hansson
1
University of Gothenburg, Medical Biochemistry, Gothenborg, Sweden
The intestinal border is covered by a layer of large secreted glycoproteins which are called gel-forming mucins. Another type of
mucins, the membrane-tethered mucins, is anchored to the apical membrane of enterocytes and builds the glycocalyx. Both
groups of glycoproteins serve as nutrient source for commensal bacteria, and protect the epithelium from mechanical stress as
well as pathogenic bacteria.
Our group has shown before that secretion of the gel-forming mucin MUC2, and shuttling of the membrane-tethered mucin
MUC17 is induced upon stimulation with the cholinergic agent carbachol. Thus, this study aims to identify and characterize
signaling pathways activated by carbachol to understand common mechanisms triggering mucin translocation.
Human intestinal cell lines are stimulated with carbachol, proteins digested into peptides and isotope labeled. Afterwards,
phosphopeptides are enriched by MOAC using TiO 2 beads. Quantitative phosphoproteomics measurement is performed by LCMS/MS using a Q Exactive instrument. In parallel, the proteome is analyzed to serve as basis for normalization of the
phosphoproteomics data.
2+
Preliminary results show regulated pathways involved in Ca signaling and endocytosis. More intense analyses of data will give
new insights into the spatial regulation of mucins.
P098
Using multiple collision energies to improve N- and O-glycopeptide identification by LC-MS/MS
1
1
2
2
U. Schweiger-Hufnagel , S. Pengelley , H. Hinneburg , D. Kolarich
1
Bruker Daltonik GmbH, Bremen, Germany
2
Max Planck Institute of Colloids and Interfaces, Potsdam, Germany
Introduction: In most naturally occurring glycoproteins, pools of glycans are attached to one or more glycosylation sites. For
proteins with multiple sites of glycosylation, the analysis of each individual attachment site becomes a challenging task. Reverse
phase (RP) liquid chromatography separation in combination with quadrupole TOF mass spectrometric detection is highly suited
for the analysis of glycopeptides due to its high mass accuracy, fast duty cycles, the high m/z range, and the usability of multiple
collision energies. This approach allows the full characterization of glycopeptides, whereas the information about the
relationship between the glycans and their glycosylation sites remains preserved. This is particularly important for the
investigation of complex glycoprotein samples.
Proteomic Forum 2015
141
Objective: Identification of the glycan and the peptide part of N- and O-glycopeptides.
Material & Methods: Tryptic digests of fetuin and other standard glycoproteins were separated by RP-LC, and QTOF mass
spectrometric data were acquired on an impact II (Bruker Daltonics) using different collision energies. Data were analyzed in
ProteinScape 3.1 (Bruker Daltonics), where they were classified to (1) filter glycopeptide spectra based on diagnostic fragments,
and (2) determine the masses of the peptide moieties in a single step. Based on this, the compositions and putative structures of
the glycan moiety of the glycopeptides were identified using the integrated GlycoQuest search engine. The peptide moieties
were then identified using Mascot (Matrix Science).
Results: In the present approach, the comprehensive investigation of glycopeptides is based on (1) the use of different collision
energies for the differential fragmentation of the glycan and the peptide part of glycopeptides. (2) The peptide mass moiety was
automatically determined using protein- and glycan-related functionalities integrated in one software package, which allowed
the interactive validation of glycan and peptide ID. The approach is similarly applicable to N- and O-glycopeptides (example in
Fig. 1), and it proves particularly useful in the fields of glycoproteomics research as well as biopharmaceutical development and
QC.
Conclusion: Applying different collision energies on a QTOF instrument, the peptide and the glycan part of glycopeptides were
fragmented. For these data, the peptide mass moieties were determined automatically using an integrated bioinformatic
platform, which allowed the identification of the complete glycopeptides.
Figure 1
P099
Phosphorylation-mediated host signalling during interaction with staphylococcal alpha-toxin or Staphylococcus aureus
1
1
1
1
1
1
E. Richter , M. Harms , R. Nölker , K. Ventz , J. Mostertz , F. Hochgräfe
1
Competence Center Functional Genomics, Pathoproteomics, Greifswald, Germany
Reversible protein modifications are characteristic sub-cellular companions of host-pathogen interactions - promoting or
suppressing infectious processes. Our group characterizes alterations in phosphorylation-mediated signal transduction in airway
epithelial cells during interaction with individual staphylococcal virulence factors or following infection with S. aureus.
We quantified the dynamics of protein phosphorylation during the first hours of infection using phosphoproteomic enrichment
techniques and liquid chromatography-high accuracy mass spectrometry. Fuzzy c-means clustering allowed differentiation of
Proteomic Forum 2015
142
distinct temporal profiles of phosphorylation with associated enriched biological processes and signaling pathways. In
combination with the analysis of consensus sequence logos for substrate phosphorylation sites and kinase inhibitor assays, we
highlight individual protein kinases as major players during the different stages of the infection.
Furthermore, we characterized the influence of staphylococcal alpha-toxin (Hla) on host cell signalling in more detail. Hla is a
major pore-forming cytotoxin produced by most S. aureus strains and a key factor in the pathogenesis of S. aureus diseases,
including pneumonia. Interestingly, responsiveness of cells to alpha-toxin (Hla) from S. aureus appears to occur in a cell-type
dependent manner. We compared Hla-susceptible 16HBE14o- and Hla-resistant S9 cells, by a quantitative multi-omics strategy
for a better understanding of Hla-induced cellular programs. Phosphoproteomics revealed a substantial impact on
phosphorylation-dependent signaling in both cell models and highlights alterations in signaling pathways associated with cellcell and cell-matrix contacts as well as the actin cytoskeleton as key features of early rHla-induced effects. Along comparable
changes in down-stream activity of major protein kinases significant differences between both models were found upon rHlatreatment including activation of EGFR and MAPK1/3 signaling in S9 and repression in 16HBE14o- cells. System-wide transcript
and protein expression profiling indicate induction of an immediate early response in either model. In addition, EGFR and
MAPK1/3-mediated changes in gene expression suggest cellular recovery and survival in S9 cells but cell death in 16HBE14ocells. Strikingly, inhibition of the EGFR sensitized S9 cells to Hla indicating that the cellular capacity of activation of the EGFR is a
major protective determinant against Hla-mediated cytotoxic effects.
In summary, the system-wide analysis of protein phosphorylation in host cells during infection-related processes does not only
contribute to a better understanding of the biology of the infection but also holds great potential for the development of novel
host-centric anti-infectives.
P100
Quantitative Separation of Protein Isoforms by Free Flow Electrophoresis
1
1
1
G. Weber , R. Wildgruber , U. Sukop-Köppel , L. Eichacker
1
FFE Service GmbH, Feldkirchen, Norway
1
Protein isoforms are defined as variants of a single polypeptide which generally alter its function. More than 90% of naturally
occurring isoforms arise from post-translational modifications and less than 10% from mRNA splice variations. In many cases,
post-translational modifications change the biological activity of proteins.
Recent developments in capillary electrophoresis and especially imaged capillary isoelectric focusing have emerged as powerful
charge based separation techniques within the biopharmaceutical industry. However, the very small volumes that enable rapid
and high-resolution separations also prevent capillary methods from being used for variant isolation and further identification.
Using a novel very fast workflow for IEF-Free Flow Electrophoresis with ultra-flat pH gradients, resulting in a resolution of 0.02
∆pI between single fractions and combined with rapid UV detection, we separated and characterized multiple isoforms of
different samples in quantitative amounts.
Here, we show the isolation of up to 100 milligram of individual protein isoforms of mABs under native conditions with Free
Flow Electrophoresis. Resulting fractions were well suited for direct use in further studies such as enzyme- and/or immuneassays.
P101
Multiplex quantitation of reversible thiol modifications in proteins with iodoTMT reagents
1
1
1
1
1
S. Blankenburg , R. Nölker , D. Ulbrich , J. Mostertz , F. Hochgräfe
1
Competence Center Functional Genomics, Pathoproteomics, Greifswald, Germany
Thiol-modifications at cysteine residues can be crucial for the functionalization of proteins, e.g. disulfide bonds for structural
integrity or in the catalytic cycle. Increased intracellular level of reactive oxygen or nitrogen species, on the other hand, can also
result in thiol oxidation and modulate the activity of proteins. System-wide quantification of thiol-redox modifications in
proteins can therefore uncover redox-sensitive cysteine residues that may serve as regulatory molecular switches.
Proteomic Forum 2015
143
To characterize reversible thiol oxidations in cells at a global scale, we established two workflows utilizing thiol-reactive
iodoTMTsixplex reagents, which are a set of isobaric tags that can be quantified with high-resolution LC-MS/MS mass
spectrometry.
We either differentially label unmodified and reversibly oxidized cysteines with iodoTMT reagents or chemically label
unmodified cysteines with iodoacetamide before the exclusive labeling of the reversibly oxidized cysteines with iodoTMT
reagents. Following proteolytic digestion, iodoTMT-labelled peptides are enriched by immunoprecipitation using an anti-TMT
specific antibody. Whereas differential thiol-labeling allows direct quantification of the redox-status of protein thiols under
various conditions, the second approach allows upscaling of the initial protein amounts and thus potentially increases the
sensitivity for the detection of oxidized cysteine residues.
We have tested both strategies with 16HBE141o- human bronchial epithelial cells that were treated with the oxidative reagents
diamide, hypochlorite or hydrogen peroxide. The presented results include the description of the general thiol-redox status of
16HBE14o- under control or oxidative stress conditions as well as the identification of redox-sensitive cysteines. The pros and
cons of both approaches, e.g. in terms of sensitivity and thiol-proteome coverage, will be discussed.
P102
Mass spectrometry-based analysis of thiol-redox and phosphorylation cross talk in human bronchial epithelial cells
1
1
1
1
1
1
1
J. Mostertz , R. Nölker , A. M. Klingebeil , D. Ulbrich , S. Blankenburg , E. Richter , F. Hochgräfe
1
Greifswald University, ZIK Fungene, Pathoproteomics, Greifswald, Germany
In recent years it has become apparent that reactive oxygen intermediates including hydrogen peroxide serve as essential
secondary messengers in signal transduction. In the respiratory epithelium, signal transduction is modulated by hydrogen
peroxide-mediated oxidation of cysteine-thiol groups in protein kinases and protein tyrosine-phosphatases, e.g. after
stimulation of hydrogen peroxide production by Duox1/2. However, in epithelial cells it is largely unknown which proteins
function as thiol switches and to which target proteins they do cross talk to thereby mediating alterations at the level of
phosphorylation. Here, we used cysteine-reactive tandem mass tags for differential redox-labeling of proteins and enrichment of
phophorylated peptides in combination with high-resolution mass spectrometry to identify potential mediators of oxidation and
their targets that show alterations at the level of phosphorylation after hydrogen peroxide signaling events. Based on our results
we aim to highlight the significance of cross talk between phosphorylation and thiol-redox- modifications and the central role of
hydrogen peroxide as a second messenger.
P103
MALDI and ESI identify orthogonal parts of the phosphoproteome
1
2
1,3
1
B. Ruprecht , C. Roesli , S. Lemeer , B. Kuster
1
Chair of Proteomics and Bioanalytics, TUM, Freising, Germany
2
HI-STEM - Institute for Stem Cell Technology and Experimental Medicine, German Cancer Research Center DKFZ, Heidelberg,
Germany
3
Biomolecular Mass Spectrometry and Proteomics, Bijvoet Center for Biomolecular Research and Utrecht Institute of
Pharmaceutical Sciences, Utrecht University, Utrecht, Netherlands
Advances in phosphopeptide enrichment methods enable the routine identification of thousands of phosphopeptides out of
complex lysates. Yet the mapping of the human phosphosite complement is far from complete which is why orthogonal
identification methods are still highly desirable. Here we used a Fe-IMAC column connected to an HPLC to enrich
phosphopeptides out of a cell line digest. The same sample was then measured with both LC-MALDI and LC-nESI-MS/MS. The
trypsin-based MALDI ionization workflow was finally benchmarked using orthogonal digestion enzymes and a phosphotyrosine
immunoprecipitation (pY-IP).
Our results indicate that the phosphosite overlap of MALDI and nESI based LC-MS/MS is only 50% in 4h measurements of the
same sample. Even upon boosted sampling depth due to phosphopeptide fractionation by hydrophilic strong anion exchange,
this overlap increased to only 64%. MALDI clearly favored the identification of acidic amino acids which translates into the
enrichment of acidophilic kinase substrate motifs. We benchmarked the MALDI phosphopeptide identification workflow against
the use of five different digestion enzymes (Asp-N, Arg-C, Chymotrypsin, Glu-C and Lys-C) and found a comparable orthogonality
to a standard trypsin-based nESI-ionization workflow. Strikingly, the MALDI based workflow identified an unusually high number
and percentage of phosphotyrosine sites (357 sites, 20% of the dataset). Even compared to the selective enrichment of
Proteomic Forum 2015
144
phosphotyrosine containing peptides by immunoprecipitation (pY-1000) and LC-nESI-MS/MS, LC-MALDI-MS/MS identified 43%
(155) unique phosphotyrosine sites.
In conclusion, we show that MALDI offers a phosphosite identification orthogonality which is comparable to the use of different
digestion enzymes. In addition MALDI is beneficial for both the identification of acidophilic phosphopeptides and
phosphotyrosine sites. We envision that the combination of digestion enzymes and pY-IPs with MALDI based ionization will
result in an additional increase in phosphosite coverage.
P104
Generation of biomarkers for the detection of ADP-ribosylated proteins during cellular stress
1
1
2
V. Bilan , M. O. Hottiger , G. Hopfgartner
1
University of Zurich, Institute of Veterinary Biochemistry and Molecular Biology, Zurich, Switzerland
2
University of Geneva, Life Sciences Mass Spectrometry, School of Pharmaceutical Sciences, Genève, Switzerland
Introduction: Protein ADP-ribosylation is a reversible post-translational modification (PTM) in which the ADP-ribose (ADPr)
moiety of NAD+ is transferred onto specific amino acids of acceptor proteins. Intracellular ADP-ribosylation is catalyzed by
diphtheria toxin-like ADP-ribosyltransferases (ARTDs/PARPs). Some members of the ARTD family (e.g. ARTD1 and ARTD2) extend
mono-ADP-ribosylation further to poly-ADP-ribose (PAR) chains up to 200 units in length.
Under steady state conditions, PAR levels are generally low and hardly detectable. The induction of PAR synthesis occurs in
response to different stress conditions such as oxidative stress, DNA damage and ionizing radiation. Recently, several studies
aimed to identify the cellular ADP-ribosylome (all ADP-ribosylated proteins) upon variable stress conditions. This for example led
to identification of 235 ADP-ribosylated proteins in cells exposed to oxidative stress (Jungmichel, Rosenthal et al. 2013).
Objectives: We hypothesize that ADP-ribosylation can be used as biomarker for the detection of cellular stress conditions and
that analysis of specific ADP-ribosylated peptides can be used as quantitative sensor of oxidative stress. Thus, the aims of the
presented work are to:


Identify the cellular ADP-ribosylome upon oxidative stress and identify candidate proteins that can be used for
quantitative analysis.
Develop reliable assays for the quantification of defined ADP-ribosylated peptides using a LC-SRM MS/MS approach.
Materials and Methods: Cell culture (HeLa, MEFs); shotgun proteomics; AF1521 enrichment; LC-SRM/MS.
Results: By spiking in vitro automodified ARTD10 (i.e. mono-ARTD-ribosylated) spiked into cell lysate we show that the low
abundance of ADP-ribosylation in the cell proteome prevents identification of ADP-ribosylated peptides without prior
enrichment.
Applying an enrichment strategy based on the ADP-ribose binding domain AF1521 covalently bound to GST-beads, we identified
117 ADP ribosylated proteins in HeLa cells treated with 1mM H 2O2.
We identified ADP-ribosylated candidate peptides for quantitative analysis by exposing HeLa cells to different concentrations of
H2O2, which models different levels of oxidative stress. Preliminary results of LC-SRM MS/MS measurements of these candidate
peptides will be presented.
Conclusions: Our enrichment strategy allows us to reliably identify and assign protein ADP-ribosylation in cells exposed to
various degrees of oxidative stress.
References: Jungmichel, S., F. Rosenthal, et al. (2013). "Proteome-wide identification of poly(ADP-Ribosyl)ation targets in
different genotoxic stress responses." Molecular cell52(2): 272-285.
Proteomic Forum 2015
145
P105
Development of a phosphoproteomic workflow for the analysis of crosstalk between insulin signaling and selected metabolic
pathways
1
1
1
1
M. Rahm , A. Schaefer , M. Ueffing , S. M. Hauck
1
Helmholtz Zentrum München, Research Unit Protein Science, Oberschleißheim, Germany
Introduction: In healthy individuals, insulin stimulates cellular effects through a complex network of signaling molecules.
Perturbations in this signaling network can lead to inadequate signal transmission to effectors and consequently to resistance of
cells to insulin.
Quantitative LC-MS/MS-based proteomics performed by our group served to identify promising candidate pathways implicated
in high-fat diet-induced hepatic insulin resistance in mice, which is considered an early hallmark in the development of type 2
diabetes. Since some of these novel pathways may likely play a role in the development of insulin resistance, analyzing the
influence of these pathways on insulin signaling is of great interest.
Objectives: Protein phosphorylation is one of the most important regulatory mechanisms of intracellular signaling. Therefore,
quantitative analysis of protein phosphorylation as integral part of the insulin signaling network will be performed after
treatment of cells with effector molecules of the identified metabolic pathways. In this way, affected downstream signaling
cascades and molecular mechanisms of crosstalk between insulin signaling and these pathways can be identified.
Materials & Methods: As a first step, we are currently developing an analytical strategy combining TiO 2 affinity chromatography,
SCX and immunoprecipitation for the separation and enrichment of phosphopeptides from complex cell lysates. Afterwards,
relative quantification of phosphopeptides using SILAC or stable isotope dimethyl labeling in combination with non-targeted
phosphoproteomic analysis by LC-MS/MS will be performed in order to study the impact of the selected metabolic pathways on
insulin signaling.
Results: During preliminary studies using TiO 2 affinity chromatography for the selective enrichment of phosphopeptides from
lysates of cells stimulated with insulin, we were able to identify about 3000 phosphopeptides in a single run from only 400 µg of
total protein as starting material. While some important phosphosites implicated in insulin signaling have already been covered
by this approach, other major components of this signaling pathway are yet to be detected. The results of the different
enrichment techniques with respect to coverage of the known intracellular insulin signaling molecules will be discussed in detail.
In addition, the quantitative influence of effector molecules will be presented.
Conclusion: In order to further improve the coverage of phosphosites relevant for insulin signaling, the establishment of an SCX
liquid chromatography method for further separation of phosphopeptides and an approach using immunoprecipitation for the
enrichment of tyrosine phosphosites are in preparation. Application of the final method to studying the effects of previously
selected metabolic effector molecules on insulin signaling could then provide insight into novel mechanisms of hepatic insulin
resistance.
P106
Analysis of post translational modifications using DIA with high resolution MS1 and retention time prediction
1
1
1
R. Bruderer , T. Gandhi , O. Bernhardt , L. Reiter
1
BiognoSYS AG, Schlieren, Switzerland
1
Introduction: Shotgun proteomics is well suited to identify post translational modifications (PTMs) on peptide sequences. The
direct correlation of precursor mass and charge to its fragmentation spectrum is predestined for the identification of the
modifications. The mass difference of the precursor ion and the detected fragment ions can enable the identification and
location of the modification. For data-independent acquisition (DIA) this direct association of MS1 and MS2 is lost; in SWATH
type DIA, large precursor windows are fragmented, resulting in complex fragmentation spectra of multiple precursors, which
may contain mixed spectra of the modified and the unmodified precursor. We addressed the accessibility of peptides with PTMs
in DIA data using accurate retention time prediction to de-convolute spectra from multiple modified precursors in SWATH data.
Materials and Methods: We used a DIA workflow, termed HRM, containing high resolution MS1 spectra and high resolution
retention time prediction (iRT) (Escher et al. 2012) to analyze PTMS in cell line samples. We generated spectral libraries
optimized for peptides with PTMs iRT information.
Proteomic Forum 2015
146
Results: Using Spectronaut, the information on the precursor ion from the MS1 spectra combined with correlation of elution of
fragment ions and precursor ion enabled confident identification of peptides with PTMs. Fragment ions containing information
on the nature and location of the modification were essential. Modified peptides eluted at different retention times, when using
high resolution chromatography serving as very useful information in the spectral library.
Discussion: Using a SWATH like DIA approach, we show that the automated identification and quantification of peptides with
PTMs can be achieved with high reproducibility and precision. Remarkably, HRM enables identification of substoichiometric
modified peptides, which are frequently missed by shotgun proteomic analyses.
Escher, C. et al., 2012. Using iRT, a normalized retention time for more targeted measurement of peptides. Proteomics, 12(8),
pp.1111-21. Available at: http://www.ncbi.nlm.nih.gov/pubmed/22577012
P107
Proteomic Analysis of Cysteine Persulfides
1
2
1
2
1
S. Longen , F. Richter , K.- F. Beck , I. Wittig , J. Pfeilschifter
1
Pharmacenter Frankfurt, Institute of Pharmacology and Toxicology, Frankfurt am Main, Germany
2
Medical School, Johann Wolfgang Goethe-University, Functional Proteomics, SFB815 core unit, Frankfurt am Main, Germany
Cysteine residues (thiols) of proteins are preferred targets for modifications mediated by reactive oxygen species (ROS), reactive
nitrogen species (RNS) and other signaling molecules such as H 2S. These modifications can lead to dramatic changes in the
stability, function, activity and localization of proteins. Interestingly, one cysteine residue can be modified by different molecules
1
leading to different cellular consequences. E.g. modification of Cys38 of the NFκB subunit p65 by NO leads to its inhibition . Vice
2
versa, modification of Cys38 by H2S leads to an activation of NFκB . H2S was discovered recently as an important signaling
3
molecule with high therapeutic potential in inflammation, in the cardiovascular system and in cell growth . However, its exact
4
mode of action is poorly understood. It is believed that not the gaseous form of H 2S but rather polysulfides can interact with
cysteine residues forming persulfides (R-S-SH) and thereby changing the properties of a protein. Here, we show a SILAC based
redox proteomic approach for LC-MS/MS in order to identify possible proteins affected by H 2S and persulfides on peptide level.
1
Matthews JR, Botting CH, Panico M, Morris HR, Hay RT. Inhibition of NF-kappaB DNA binding by nitric oxide. Nucleic Acids Res.
1996 24:2236-42.
2
Sen N, Paul BD, Gadalla MM, Mustafa AK, Sen T, Xu R, Kim S, Snyder SH. Hydrogen sulfide-linked sulfhydration of NF-κB
mediates its antiapoptotic actions. Mol Cell. 2012 45:13-24.
3
Vandiver SM, Snyder SH. Hydrogen sulfide: A gasotransmitter of clinical relevance. J. Mol. Med. 2012 90:255-263
4
Greiner R, Pálinkás Z, Bäsell K, Becher D, Antelmann H, Nagy P, Dick TP. Polysulfides link H2S to protein thiol oxidation. Antioxid
Redox Signal. 2013 19:1749-65.
Proteomics for Human Health II (P146–P169)
P146
Coupling SPRi and MALDI MS for Interaction Analysis
1
2
3
3
4
5
2
U. Anders , J. Schaefer , C. Frydman , F. Hibti , C. Schuhmacher , D. Suckau , A. Plückthun , R. Zenobi
1
ETH Zurich, D-CHAB, Zurich, Switzerland
2
University of Zurich, Zurich, Switzerland
3
Horiba Jobin Yvon, Palaiseau, France
4
Horiba Jobin Yvon, Unterhaching, Germany
5
Bruker Daltonics, Bremen, Germany
1
In Pharma and Life Science research, there is a lot of interest for fast and accurate screening methods. Surface plasmon
resonance (SPR) is suitable to analyze biomolecules in a label-free fashion. Rapid and high-throughput analysis of different
interactions in parallel can be achieved by working in an array format (SPR imaging). Coupling SPRi with MALDI MS enables a
multiplexed detection and quantification of interactions by SPRi on one hand, and on the other hand, identification of
Proteomic Forum 2015
147
interacting ligands by MALDI MS directly on chip. In addition to the possibility for real time monitoring a major advantage of
SPRi-MALDI MS is its ability to measure directly on the plate with no need of eluting the binding partner. [1]
The aim of this work is to investigate off-target binding specificity of various DARPins. DARPins (designed ankyrin repeat
proteins) rival antibodies for target binding, but are genetically engineered. They are more robust and can be tagged for an
aligned chemical coupling to solid supports. It is known that the DARPin off7 interacts specifically with maltose binding protein
(MBP) [2]. MBP is part of the maltose/maltodextrin system, which is a sophisticated regulatory and transport system involving
many proteins. The binding specificity of DARPins with a mixture of important proteins will be determined while DARPins are
immobilized.
The platform used for the measurements is a functionalized gold chip based on NHS-esters for immobilizing ligands. For SPRi
measurements a SPRi-Plex II from Horiba Jobin Yvon was used. Before MALDI MS analysis the captured proteins have to be
treated by reduction, tryptic digestion and matrix deposition. Then the gold chip is introduced using an adapter target from
Bruker Daltonics in the Ultraflex II MALDI-TOF for identification and characterization of the specific interacting analytes.
In the sensorgram (the output from SPRi) we are able to see the specific interaction between the DARPins and several proteins,
like off7 and MBP. In addition to this after a successful digestion a mass spectrometric identification of interacting partners is
also possible.
The coupling of SPRi and MALDI MS technologies meet the strong demand for high-throughput analysis and identification of
biomolecular interactions. In addition it provides information on binding kinetics and binding affinity in real time.
[1] F. Remy-Martin et al. Anal. Bioanal. Chem. 2012, 404:423-432
[2] H. K. Binz et al. Nature Biotechnol. 2004, 22:575-582
figure legend: Scheme of SPRi-MALDI
Figure 1
Proteomic Forum 2015
148
P147
Proteomic profiling of patient-derived CD4+ T-cells for biomarker discovery in type 1 diabetes
1
1
2
2
1
M. Lepper , C. von Törne , E.- M. Sedlmeier , A.- G. Ziegler , S. Hauck
1
Helmholtz Zentrum München, Research Unit Protein Science, Neuherberg, Germany
2
Helmholtz Zentrum München, Institute of Diabetes Research, Neuherberg, Germany
Introduction: Type 1 diabetes (T1D) is a chronic autoimmune disease resulting from the destruction of ß-cells in pancreatic
islets. The pathogenesis of type-1-diabetes is hallmarked by two major events: the initiation of islet autoimmunity and the onset
of clinical diabetes. While clinical diabetes manifests itself as an end-stage insulitis at the time of diagnosis, the preclinical period
of islet autoimmunity can vary considerably in time between multiple autoantibody positive subjects. Currently, knowledge of
cellular markers characterizing different autoimmune phenotypes is insufficient.
Objectives: Autoreactive CD4+ T-cells play a key role in the development and progression of T1D. We hypothesize that CD4+ Tcells exhibit distinct proteomic (surface) signatures which are decisive for their function; therefore, correlation of protein
expression levels with disease status may offer a unique strategy for discovering novel cellular biomarkers. We will explore
different proteomic approaches for proteomic phenotyping of human CD4+ T-cells and ultimately profile a set of clinical samples
in a discovery-driven approach.
Materials & Methods: Different sample preparation strategies for proteomic profiling of human CD4+ T-cells such as Glyco-PAL,
FASP and in-StageTip were assessed with focus on optimizing protein identification rates in quantity-limited samples.
Quantitation of protein expression levels were performed in a label-free manner. Clinical samples of biobanked PBMCs were
obtained from clinical studies conducted within the Helmholtz Zentrum München. The discovery set is composed of a T1D
subject group and a control group.
Results: Processing as few as 1 million CD4+ T-cells with Glyco-PAL resulted in the identification of more than 100 proteins
encompassing 60 % plasma membrane associated proteins as revealed by Gene Ontology. When assessing whole-proteome
profiling approaches using approximately 250.000 CD4+ T-cells in-StageTip based sample preparation resulted in the
identification of 1434 proteins corresponding to 3843 total unique peptides, while FASP achieved a total of 1110 proteins with
2417 unique peptides, respectively. Proteomic profiling of the discovery set by in-StageTip sample preparation using SDP-RPS
teflon embedded beads followed by LC-MS/MS is currently underway.
Conclusion: After the discovery set is profiled correlation of protein expression levels with disease status will eventually reveal
protein biomarkers that monitor different autoimmune phenotypes.
P148
Platelet proteome changes in stable angina pectoris patients induced by aspirin treatment
1
1
1
2
3
1
1
P. Majek , K. Pecankova , Z. Riedelova-Reicheltova , M. Maly , M. Oravec , J. Suttnar , J. Dyr
1
Institute of Hematology and Blood Transfusion, Biochemistry, Prague, Czech Republic
2
Central Military Hospital, 1st Medical Faculty, Charles University, Prague, Czech Republic
3
Complex Cardiovascular Center for Adult Patients - Cardiology Clinic of the 2nd Faculty of Medicine of Charles University and
University Hospital Motol, Prague, Czech Republic
Blood platelets, an essential part of hemostasis, play an important role in the development of cardiovacular diseases which are
the main cause of death in the developed countries. Platelet proteome is known to be affected by patients’ treatment and so is
platelet function. It is known that aspirin (ASA) treatment affects platelet function, their activation and aggregation. The goal of
this work was to identify the proteome changes in platelets of patients with stable angina pectoris (SAP) with aspirine (ASA)
treatment when compared to the control groups of patients with and without ASA treatment. Platelets were isolated by
centrifugation from patients with SAP (n=12) and controls with (n=6) or without (n=6) ASA treatment. Platelet proteins were
separated by 2D SDS-PAGE, gels were stained with colloidal Coomassie blue and evaluated using Progenesis SameSpots
software. Following the proteomes comparison, there were 6 different spots found to differ significantly (ANOVA p<0.05) among
the studied groups. Spots of interest were excised, digested by trypsin, and further analyzed using nanoLC-MS/MS. Preliminary
results show affection (modification) of cytoskeletal proteins in ASA treated patients. The differences will be further analysed
using immunodetection methods and mass spectrometry-based quantification.
Proteomic Forum 2015
149
This study was supported by the Czech Science Foundation P205/12/G118 and by the Ministry of Health of the Czech Republic
project for the conceptual development of the research organization (Institute of Hematology and Blood Transfusion).
P149
Platelet-protein-profiling as a screening method for colorectal cancer
1
1
1
2
2
3
3
1
1
3
S. Strohkamp , T. Gemoll , S. Johanning , S. Hartwig , S. Lehr , S. Becker , B. Franzén , U. J. Roblick , T. Keck , G. Auer , J. K.
1
Habermann
1
University Hospital Schleswig-Holstein, Section for Translational Surgical Oncology and Biobanking, Department of Surgery,
Lübeck, Germany
2
German Diabetes Center, Institute of Clinical Biochemistry and Pathobiochemistry, Düsseldorf, Germany
3
Karolinska Institute, Karolinska Biomics Center, Department of Oncology and Pathology, Stockholm, Sweden
Introduction: Colorectal cancer (CRC) is one of the most frequent malignancies in the western world. Early tumor detection and
intervention are important determinants on CRC patient survival. During tumor dissemination, growth and angiogenesis,
platelets actively and selectively store and segregate proteins. Hence, platelet proteins could potentially serve as screening
markers for early malignancy.
Material and Methods: We compared protein profiles of platelets between healthy volunteers (n=12) and patients with early(n=7) and late-stage (n=5) CRCs by using multiplex-fluorescence two-dimensional gel electrophoresis (2-DE). Results with
®
differences in protein levels between these groups were analyzed with SameSpots software followed by Principle Component
Analysis (PCA). Proteins of interest were identified by mass spectrometry (MS) and subjected to Ingenuity Pathway Analysis
(IPA). Target proteins were validated and confirmed by multiplex-fluorescence-based Western Blot analyses using an
independent cohort of platelet protein samples [healthy controls (n=24), early CRCs (n=16) late CRCs (n=16)].
Results: By inter-group comparison, 39 differentially expressed protein spots were detected (p < 0.05) and revealed a good
separation between all three groups in the PCA-based cluster analysis. 31 (79 %) proteins were identified by MS and IPA resulted
in one high-ranked network maintaining 14 proteins. Of those, Caspase 3 (CASP3), Clusterin (CLU) and Vinculin (VCL) were downregulated, whereas Cofilin 1 (CFL1) and Glutathione Synthetase (GSS) were present at higher levels in platelets from CRC
patients compared to control individuals. Expression characteristics of these five proteins were confirmed by Western Blot
analyses in an independent cohort.
Conclusion: In this study, we identified differentially regulated proteins within platelets distinguishing healthy controls from
patients with early- and late-stage CRCs. The potential of CASP3, CLU, VCL, CFL1 and GSS as biomarkers for CRC screening should
be confirmed in a prospective multicenter trial.
P150
Peripheral blood mononuclear cell proteome changes in patients with myelodysplastic syndrome
1
1
1
1
K. Pecankova , P. Majek , J. Cermak , J. Dyr
1
Institute of Hematology and Blood Transfusion, Biochemistry, Prague, Czech Republic
Mononuclear cells are round-nucleated blood cells involved in immune system processes. Myelodysplastic syndrome (MDS)
encompasses a diverse range of oncohematological diseases characterized by ineffective hematopoiesis, by blood cytopenias,
and by a progression to acute myeloid leukemia. Our aim was to search for proteome changes in the peripheral blood
mononuclear cells of MDS patients with refractory cytopenia with multilineage dysplasia.
Peripheral blood mononuclear cells were isolated from a total of 12 blood samples (6 healthy donors and 6 MDS patients) by
gradient centrifugation. The proteins were separated by 2D SDS-PAGE (pI 4-7) and double-stained. The proteomes were
compared and statistically processed with Progenesis SameSpots software, then proteins were identified by nanoLC-MS/MS.
Protein functional association and expression profiles were analyzed using the EnrichNet application and Progenesis SameSpots
hierarchical clustering software, respectively.
By comparing the two groups, 178 significantly (p<0.05, ANOVA) differing spots were found, corresponding to 139 unique
proteins. Data mining of the Reactome and KEGG databases using EnrichNet highlighted the possible involvement of the
identified protein alterations in apoptosis, proteasome protein degradation, heat shock protein action, and signal transduction.
Proteomic Forum 2015
150
The comparison of this study with results of previous MDS plasma proteome studies suggests that the alterations in plasma
proteins (possibly secreted by PBMCs) are caused by posttranslational modifications.
To the best of our knowledge, this is the first time that proteome changes have been identified in the mononuclear cells of MDS
patients.
This study was supported by the Czech Science Foundation’s P205/12/G118 and by the state project (Ministry of Health, Czech
Republic) for the conceptual development of the research organization (Institute of Hematology and Blood Transfusion).
P151
Drug-dependent regulation of B-Raf containing macromolecular protein complexes
1
2
3
3
1
B. Diedrich , K. Rigbolt , M. Röring , T. Brummer , J. Dengjel
1
University Medical Center of Freiburg, Dermatology, Freiburg, Germany
2
University of Freiburg, ZBSA, Freiburg, Denmark
3
University of Freiburg, Institute of Molecular Medicine and Cell Research, Freiburg, Germany
The majority of human cancers are caused by the malignant transformation of epithelial cells, one of the most frequently
occurring malignancies being colorectal carcinoma (CRC). Activating mutations of the oncogene BRAF contribute decisively to
the development of carcinomas. Recent studies suggest that the use of B-Raf inhibitors in BRAF mutant CRC cells yields less
uniform responses than in melanoma reflecting the heterogeneous character of CRC. These studies highlight the need for a
better understanding of normal and mutant B-Raf signalling at the molecular level, in particular a solid understanding of B-Rafdependent interaction networks, their structure and regulation.
To study B-Raf-dependent alterations in protein-protein interactions potentially driving tumor development we used human
epithelial colorectal adenocarcinoma cells (CaCo-2) inducibly expressing oncogenic or wild type BRAF as CRC model system. We
studied B-Raf as well as global alterations in protein interactions upon oncogene expression. (1) Combining affinity purification
WT
V600E
with quantitative proteomics differences between interaction partners of B-Raf and B-Raf
were characterized. (2) To
further expand our search we applied a combination of size-exclusion chromatography and high-accuracy, high-resolution
quantitative mass spectrometry for unbiased high-throughput screening of changes in protein interactions. (3) Pharmacological
interference of clinically approved kinase inhibitors with B-Raf complexes was analyzed via BN PAGE.
We identified B-Raf itself as exhibiting the most robust changes in protein interactions depending (a) on the variant of the
oncogene being expressed and (b) on the used drug. B-Raf was present in two distinct protein complexes of different molecular
WT
V600E
weight. The majority of B-Raf was found in a complex of 200 kDa, whereas the majority of B-Raf
localized to a complex of
600 kDa. Interestingly, both complexes consisted of different proteins and the use of several kinase inhibitors indicated a
V600E
WT
differential regulation of B-Raf
and B-Raf containing complexes. Inhibitor-dependent protein dynamics could be linked to
observed drug responses in cancer patients.
This global unbiased investigation of the B-Raf-dependent interactome elucidates new mechanisms in malignant cell
transformation and gives new insight to the mechanism of action of approved drugs used to treat B-Raf driven tumors.
P152
A Kit for Mass-Spectrometry based Absolute Quantification of Metabolic Enzymes
1
2
2
2
1
1
1
K. Schnatbaum , S. Kempa , G. Mastrobuoni , F. Bindel , P. Ensle , J. Avramova-Nehmer , H. Wenschuh , U. Reimer
1
JPT Peptide Technologies GmbH, Berlin, Germany
2
Berlin Institute for Medical Systems Biology at the MDC Berlin-Buch, Berlin, Germany
1
Question: Metabolic transformation is a major event during cancer development. Specifically the de-regulation of the central
energy metabolism, a phenomenon already observed a century ago, contributes to the cancer phenotype. These alterations are
thought to induce metabolic vulnerabilities and thus are of therapeutic interest.
TM
Recently, SpikeTides have been described as a low cost peptide standard source for development of MRM assays as well as for
absolute and relative quantification of protein expression levels (1). For absolute protein quantification the peptides are
produced in a quantified format via a highly efficient procedure based on a UV-active tag (2).
Proteomic Forum 2015
151
The goal of the presented study was the development of a kit containing absolutely quantified peptides for multiplexed absolute
quantification of metabolic enzymes in biological fluids and tissues to support diagnosis and stratification of cancer.
Methods and Results: Proteotypic peptides of major metabolic enzymes were selected based on experimental data and
bioinformatics predictions. The peptides were synthesized and absolutely quantified with the help of a UV-active tag that is
removed after quantification by tryptic digestion. For all peptides MRM-based assays were set up and LOD and LOQ determined.
For 50 peptides spanning the known deregulated central metabolic enzymes highly reproducible MRM assays could be
established.
The quantitative determination of enzyme abundance from biological samples revealed a concentration range spanning several
orders of magnitudes.
Conclusions: We describe the first commercially available kit of a large number of absolutely quantified stable isotope labeled
peptides. The efficient assembly of the kit was made possible by the high efficiency of the peptide quantification approach. The
kit enables the multiplexed absolute quantification of 30 central metabolic enzymes from complex protein mixtures.
(1) (a) Picotti, P. et al., Nat. Methods 2010, 7, 43-46. (b) Simicevic, J. et al., Nat. Methods 2013, 10, 570-576. (c) Wilhelm, M., et
al., Nature 2014, 509, 582-587.
(2) Schnatbaum, K. et al. non-peer-reviewed application note in Nature Methods 2011, 8.
P153
Analysis of Cerebrospinal Fluid using high resolution mass spectrometry
1
1
1
S. Galozzi , K. Barkovits , K. Marcus
1
Ruhr-Universität Bochum, Bochum, Germany
Nowadays, the most common neurodegenerative disorders are Alzheimer’s and Parkinson’s disease (AD, PD). Although these
disorders are extensively investigated from several decades, an exact diagnosis of AD and PD can only be confirmed postmortem. To facilitate therapeutic intervention, an early diagnosis as well as the possibility to monitor disease progression is very
important. Since cerebrospinal fluid (CSF) is the body fluid that surrounds the brain it is a promising source for biomarker
discovery of neurodegenerative disorders. For this reason multiple proteomic studies of CSF were performed to investigate
differences between healthy individuals and patients suffering from neurodegenerative diseases.
Here we show a label-free mass spectrometry (MS) approach to analyze CSF, which provides the great advantage of fast and
cheap sample preparation compared to stable-isotope labelling methods. Our workflow consists of an in-solution tryptic
digestion, determination of protein concentration followed by LC-MS data acquisition and data analysis/evaluation. Hence, the
digested peptides of non-processed CSF (in terms of non-depleted or pre-fractionated) are analyzed to get a profound
understanding of the CSF proteomic content. By using a three hour LC gradient and a high resolution orbitrap mass
spectrometer we are able to identify over 600 proteins from 500 ng CSF. In order to reach a deeper coverage of the CSF
proteome and to enable the detection of low abundant proteins we use several alternative strategies concerning sample
preparation, LC separation and MS method. This label-free MS approach was utilized to compare CSF proteome profiles of
healthy individuals with PD patients in order to identify potential biomarkers.
P154
Molecular and proteomic characterization of two primary cultures recovered from the ascites of Mexican ovarian
adenocarcinoma patients.
1
1
1
1
1
O. L. Garibay-Cerdenares , V. Hernández Ramírez , J. C. Osorio Trujillo , B. Chávez Munguía , A. González-Robles , M. Hernández
2
2
3
1
Ortíz , S. M. Encarnación Guevara , D. Gallardo Rincón , P. Talamás Rohana
1
CINVESTAV, Infectomics and Molecular Pathogenics, Mexico DF, Mexico
2
Center of Genomic Sciences of UNAM, Proteomic Laboratory, Cuernavaca, Morelos, Mexico
3
National Cancer Institute, Medical Oncology, México, D.F., Mexico
Introduction: Ovarian cancer ranks second in incidence among gynecological cancers worldwide; however, it represents 47% of
deaths. Usually diagnosis occurs at later stages, when the tumor has spread to other organs. In Mexico, little is known about the
proteomic and molecular features of transformed ovary cells. Objectives. To perform a molecular and proteomic
characterization of primary cells recovered from two ovarian cancer patients with a high and low degree of ovarian carcinoma
Proteomic Forum 2015
152
Patients: Samples were obtained from the Mexico National Cancer Institute under the signature of informed consent and with
the approval of bioethics and scientific committees.
Methods: Confocal and electron microscopy, and bidimensional electrophoresis and Mass Spectrometry (MALDI-TOF/TOF)
assays were developed.
Results: Primary cultures were recovered from ascites samples of patients who have not undergone chemotherapy treatment,
diagnosed with ovarian endometrioid and serous papillary adenocarcinomas. Culture cells, identified as INCan017 and
INCan019, respectively, were free of contamination, of epithelial origin and had a replication rate of 48 h. Ultrastructural assays
showed cells with adherent morphology with large actin protrusions (Fig. 1), without contact inhibition, proliferating and highly
metastatic. Proteomic analysis resulted in the definition of at least 2000 proteins with several regions of differential protein
expression pattern (Fig. 2); of these, 27 proteins were identified and related with processes such as metabolism, cell cycle and
signal transduction pathways (Table 1).
Conclusions: Primary cell cultures from two Mexican patients with ovarian adenocarcinoma were obtained. These primary
cultures could be a replicable model that may be widely used for studying the mechanisms involved in cell metastasis or drug
resistance processes among others.
References: 1. Siegel R et al., Cancer statistics, 2012. CA: A Cancer J. for Clinicians, 62:10-29. 2. Tan DSP et al., Mechanisms of
transcoelomic metastasis in ovarian cancer. The Lancet Oncology 2006, 7:925-934. 3. Burleson KM, et al., Ovarian carcinoma
ascites spheroids adhere to extracellular matrix components and mesothelial cell monolayers. Gynecol Oncol 2004, 93:170-181.
Figure Legends: Fig. 1. Cellular morphology of ovarian epithelial cancer cells. Fig. 2. Selection of a differential expression profile
zone. Table 1. Protein identification by Mass Spectrometry (MALDI/TOF).
Figure 1
Figure 2
Proteomic Forum 2015
153
155
Proteomic changes of Caco-2 cells after silver nanoparticle treatment
1
1
1
2
1
3
1
4
3
S. Juling , D. Lichtenstein , L. Böhmert , S. Selve , A. Niedzwiecka , M. Schümann , H. Sieg , A. Thünemann , E. Krause , A.
1
1
Braeuning , A. Lampen
1
Federal Institute for Risk Assessment, Unit Food Safety, Berlin, Germany
2
Zentraleinrichtung Elektronenmikroskopie (ZELMI) , Transmissionselektronenmikroskopie, Berlin, Germany
3
Leibniz-Institut für Molekulare Pharmakologie , Mass Spectrometry, Berlin, Germany
4
Federal Institute for Materials Research and Testing, Polymers in Life Science and Nanotechnology, Berlin, Germany
Due to the rising application of silver nanoparticles in food-associated consumer products the uptake of silver nanoparticles by
the oral route has become a serious scenario. Therefore, knowledge about silver nanoparticle-induced effects should be studied.
In order to achieve this a comparative proteomic analysis of Caco-2 cells which are exposed to nanosilver and ionic silver was
performed. The study design provides the possibility to differentiate between a ionic release of the particles as well as the
influence of the nanoparticle-coating (stabilizer).
The characterization of the nanoparticles was performed via DLS and SAXS measurements. Differentiated Caco-2 cells were
incubated 24 h with silver nanoparticles, silver ions and the nanoparticle stabilizer. Cellular proteins were isolated and 1D LC ESI18
Orbitrap MS/MS analysis with O stabile isotope quantification was performed to identify the deregulated proteins.
It was observed that the silver nanoparticles, silver ions and stabilizer itself showed no reduction of cell viability in the
concentration used. After expression analysis, 50 - 100 deregulated proteins for each nanoparticle, ion and matrix/stabilizer
treatment were identified. Mass spectrometry results show that the deregulated proteins may have a relation to lipid
metabolism and are associated with tight junction.
We can conclude that different proteins were significantly deregulated for nanoparticles, ions and the stabilizer treatment. The
hypothesis that the combination of ions and matrix/ stabilizer effects to proteins may result in the same effect that is observed
for nanoparticles, cannot be asserted on the basis of the experimental results. Depending on the first results a nanoparticle
specific effect cannot be ruled out and should be taken into account for the risks posed by nanoparticles.
P156
Comparison of two different Iron Oxide Nanoparticle Protein Corona after Isolation from complex HepG2 cell lysate
1
1
1
1
1
1
1
S. Juling , D. Lichtenstein , J. David , H. Sieg , C. Rozycki , B. Niemann , A. Braeuning , A. Lampen
1
Federal Institute for Risk Assessment, Unit Food Safety, Berlin, Germany
1
There is a widespread use of different materials in nanometer range. In medicine iron oxide nanoparticles are used as a contrast
media to improve the visibility of internal body structures in magnetic resonance imaging (MRI). The food industry use iron oxide
and hydroxide as food additive (E172). Indeed there is a reason to believe that the protein corona, the interface between a
nanoparticle and the environment has an influence on absorption, distribution and cellular downstream effects. Therefore,
knowledge about the corona is of pivotal importance. In this study, the protein corona of two iron oxide nanoparticles from
HepG2 cell lysate were isolated and compared. The investigated particles were similar in core size but differed in their coating
(carboxydextran and polyacrylic acid) as well as their zeta potential.
Iron oxide nanoparticles with two different coatings were incubated with whole lysates from HepG2 human hepatocellular
carcinoma cells. The nanoparticles were isolated via magnetization, along with their protein corona. After four washing steps for
the removal of unspecific bound proteins, the protein corona was separated from the nanoparticles via SDS-PAGE. Finally, the
corona building proteins were identified the by 1D LC-QTOF MS/MS.
It was observed that both iron oxide nanoparticles were taken up by HepG2 cells. No reduction in cell viability was observed for
the particles in the used concentrations. After duplicate protein corona analysis, between 10 and 20 proteins were identified as
a part of the corona for each nanoparticle. Only two proteins were found in both corona.
We can conclude that it is possible to identify the protein corona of iron oxide incubated with cell lysate. The results suggest that
the corona composition of nanoparticles depends on their coating.
Proteomic Forum 2015
154
P157
A first proteome wide prediction of the dynamics of insulin resistance
unveils systems wide mechanisms behind type II diabetes
1
1
1
1
R. Magnusson , E. Nyman , M. Gustavsson , P. Strålfors , G. Cedersund
1
Linköping University, Linköping, Sweden
1
Introduction: Type 2 diabetes (T2D) is an increasingly costly and rapidly growing disease, as the living standards in the world are
increasing. At the heart of T2D lies insulin resistance, a malfunction in the complex insulin signaling network. Because of its
complexity, our understanding is still limited, and the development of new drugs is slow. To truly makeprogress in the
understanding of such a complex system, an approach involving mathematicalmodeling is necessary. In a series of papers, we
have already developed the basis for such anapproach [1]. This approach is based on time-series and dose-response data for the
mainplayers in the network. The resulting model explains insulin resistance for these main players (Fig 1), and is already today
used by pharmaceutical companies to aid their drug developmentprocess. However, the model [1] still only accounts for a small
fraction (proteins and phosphorylation sites that responds to insulin.
Objective: To make a quantum leap to a new type of a more systems-wide model, new systems-level data are necessary. Last
year, simultaneous measurements of >30 000 phosphorylation sites, where >5000 of these respond to insulin, were achieved by
mass spectrometry [2]. In this poster, we present our new and developing approach to analyzing these large scale data, in a way
that makes use of our already developed model [1] and other prior knowledge.
Materials and methods: To derive a network from the data, the regression based method LASSO is used. In this way, large
systems can be approximated using statistical algorithms. We propose a method to merge the well-studied developed model [1]
with such a data driven statistical network, generating an ordinary differential equation based model comprising thousands of
connected states (Fig 2).
Results: We present a first proteome-wide dynamic model of insulin signaling. This model can be used for system-wide
predictions of insulin resistance.
Conclusion: This project leads the way to a new generation of mathematical models, where high-quality knowledge is
embedded as a core in statistical networks (Fig 2). This project thus also starts to bridge the two main communities of systems
biology: those based on high-quality hypothesis-driven mechanistic models, and those based on omics data and statistics.
[1] Nyman et al, J Biol Chem, 2014, 289:33215-30
[2] Humphrey et al, Cell Metal, 2013, 17:1009-20
Figure 1
Figure 2
Proteomic Forum 2015
155
158
Proteomic approach to study a couple of monozygotic twins discordant for the chronic fatigue syndrome: focus on
mitochondria involvement
1,2
1
3
3
4
5
5
2,4
3
F. Ciregia , L. Giusti , L. Kollipara , R. Zahedi , M. Ronci , L. Bazzichi , C. Giacomelli , A. Urbani , A. Sickmann , A. Lucacchini
1
University of Pisa, Department of Pharmacy, Pisa, Italy
2
Fondazione Santa Lucia, Proteomic and Metabonomic Laboratory, Rome, Italy
3
Leibniz-Institut für Analytische Wissenschaften-ISAS , Dortmund, Germany
4
University of Rome Tor Vergata, Department of Experimental Medicine and Surgery, Rome, Italy
5
University of Pisa, Division of Rheumatology, Department of clinical and experimental medicine, Pisa, Italy
1
Introduction: Chronic fatigue syndrome (CFS) is a severe, systemic illness characterized by persistent, debilitating and medically
unexplained fatigue not improved by rest. CFS is accompanied by several symptoms that differ among patients (e.g. headaches;
sleep disturbances; difficulties with concentration). Etiology is unknown and diagnosis is formulated through the patient’s
history and exclusion of other medical and psychiatric causes with the same symptoms. Thereby, the availability of biomarkers
for CFS could be of great usefulness for clinical research.
Objectives: CFS has been proposed as a multi-factorial disease, and, since genetics is presumed to have a role in the etiology of
CFS, one area of investigation is the twin studies. Moreover, it is increasingly recognized that CFS can be related to
mitochondrial dysfunctions. Therefore, we decided to analyze mitochondria extracted from platelets of a couple of monozygotic
twins discordant for CFS.
Materials & Methods: Blood was collected from the twins at the same time, and mitochondria have been isolated from platelets
by differential centrifugation. In order to test mitochondria enrichment, electron microscopy studies, western blot analysis and
screening of the activity of cytochrome oxidase have been performed. The combination of two proteomic approaches was used
to highlight the specific pathways involved in CFS: 2DE (coupled to MALDI TOF/TOF) and label-free LC-MS/MS.
Results: With 2DE we detected 317 spots in the enriched mitochondrial fraction, and 82 were differentially expressed. The label
free approach identified 1008 proteins, and 484 were differentially expressed between CFS and control. Proteins differentially
expressed obtained from the two proteomic approaches, were functionally analyzed using the Ingenuity Pathways Analysis (IPA)
software with the aim to determine the interaction network involved. IPA revealed a specific association of our proteins with
Inflammatory Response, metabolism of reactive oxygen species (ROS) and viral infection.
These data support the idea that persistent infection may lead to chronic activation of the immune system involving excessive
production of ROS. The raised levels of ROS damage mitochondria, leading to an alteration of oxidative phosphorylation, to a
diminished electron transfer, and thus to a diminished ATP production, which in turn leads to energetic decline. In fact, if we
consider only mitochondrial proteins, 69 were down-regulated in CFS (e.g, cytochrome c oxidase, ATP synthase, NADH
dehydrogenase) while only 18 were up-regulated.
Conclusion: Our observations strongly correlate the mitochondrial dysfunction with CFS symptoms: a metabolic alteration at
mitochondrial level entails an energy deficiency and explains the peculiar fatigue. However, our results should be confirmed in a
properly planned study in a different cohort of patients affected by CFS.
P159
A refined mass spectrometry based draft of the human proteome
1,2
2,3
3
2
3
4
5
3
H. Hahne , M. Wilhelm , J. Schlegl , A. Moghaddas Gholami , M. Lieberenz , M. M. Savitski , B. Hallström , E. Ziegler , T.
4
2
2
3
3
2
2
6
6
6
Mathieson , L.- H. Li , D. Wang , L. Butzmann , S. Gessulat , H. Marx , S. Lemeer , K. Schnatbaum , U. Reimer , H. Wenschuh , A.
7
2
2
3
3
4
7
5
3
Asplund , M. Mollenhauer , J. slotta-Huspenina , J.- H. Boese , A. Gerstmair , M. Bantscheff , F. Pontén , M. Uhlen , F. Faerber ,
2
B. Kuster
1
OmicScouts GmbH, Freising, Germany
2
TU München, Freising, Germany
3
SAP SE, Walldorf, Germany
4
Cellzome GmbH, Heidelberg, Germany
5
Karolinska Institutet, Stockholm, Sweden
6
JPT Peptide Technologies GmbH, Berlin, Germany
7
Uppsala University, Uppsala, Sweden
Proteomic Forum 2015
156
Proteomes are characterized by large protein abundance differences, cell type and time dependent expression patterns and
post-translational modifications all of which carry biological information complementary to and not accessible by genomics or
transcriptomics. Despite significant ongoing efforts, it is still not entirely clear which human genes exist as proteins. We
addressed this challenge using mass spectrometry based proteomics to answer one of the most fundamental scientific questions
of the post-genome era.
We recently published a mass spectrometry based draft map of the human proteome and ProteomicsDB, a public in-memory
database for real time analysis of big data (https://www.proteomicsdb.org) [1]. The information assembled from hundreds of
human tissues, cell lines and body fluids representing >40,000 hours of LC-MS/MS time allowed us to re-estimate the size of the
protein coding genome, to identify hundreds of translated lincRNAs, to define the core proteome of human cells and to perform
whole organism protein expression analysis along with the identification of organ-specific proteins of hitherto unknown
function. Analysis of mRNA and protein expression profiles of human tissues revealed conserved control of protein abundance,
integration of drug sensitivity data allowed the identification of proteins predicting resistance or sensitivity for kinase inhibitor
drugs and proteome profiles also hold considerable promise for analysing the composition and stoichiometry of protein
complexes. ProteomicsDB thus enables navigation of proteomes, provides biological insight and fosters the development of
proteomic technology.
Having established a baseline proteome profile of the human body, we are now in the process of refining this draft map. The
comprehensive and deep proteome analyses of 44 healthy human tissues profiled in the Human Protein Atlas project
(www.proteinatlas.org) [2] and the integration with histological information, antibody-based staining of >15,000 human proteins
and sample-matched RNA Seq information will reveal unprecedented insights into tissue- and cell specific biology covering, for
instance, proteome complexity, isoform expression and transcriptional and translational regulatory mechanisms.
References:
[1] Wilhelm et al. (2014) Nature 509, 582-587
[2] Fagerberg et al. (2014) Mol Cell Proteomics 13, 397-406
P160
Systematic exploration of the ciliary protein landscape by large-scale affinity proteomics.
1
2,3
4
5
1
2,3
6
2,3
K. Boldt , J. van Reeuwijk , Q. Lu , K. Koutroumpas , N. Horn , S. van Beersum , Y. Texier , M. T. Nguyen , The Syscilia
7
5
4
1,8
2,3
Consortium EU , F. Kepes , R. B. Russell , M. Ueffing , R. Roepman
1
Eberhard-Karls Universität Tübingen, Medical Proteome Center, Tübingen, Germany
2
Radboud University Medical Center , Department of Human Genetics, Nijmegen, Netherlands
3
Radboud University Nijmegen, Radboud Institute for Molecular Life Sciences, Nijmegen, Netherlands
4
University of Heidelberg, Heidelberg, Cell Networks, Bioquant, Cluster of Excellence, Heidelberg, Germany
5
Université d'Evry, Institute of Systems and Synthetic Biology, Evry, France
6
Helmholtz Zentrum München, Research Unit Molecular Epigenetics, München, Germany
7
http://www.syscilia.org/, Brüssel, Belgium
8
Helmholtz-Zentrum München, Research Unit Protein Science, München, Germany
Objective: Mutations in different ciliopathy-associated genes often result in overlapping clinical phenotypes, which can in part
be explained by disruption of overlapping functional protein modules. In this study we conducted large-scale affinity proteomics
in a systems biology-based approach to boost insights into the assembly of these ciliary modules, and their connectivity in larger
functional protein networks: the ciliary protein interaction landscape. This provides an important framework to deconvolute the
pathways and processes that drive ciliopathies, and to understand the general importance of ciliary function for cellular
homeostasis.
Methods: Using more than 220 known and potential ciliary proteins as baits, fused to the Strep/FLAG-tandem affinity
purification tag (SF-TAP), we purified protein complexes from human embryonic kidney cells (HEK293T), which were analysed by
mass spectrometry. In parallel, specific modules were scrutinized for binary interactions by yeast two-hybrid analyses. Existing
and newly developed bioinformatic algorithms were employed to validate the confidence of the identified interactions and to
define functional modules.
Results: We obtained low, medium and high confidence sets of protein interactions and modules. From this data we could
assign novel components to known ciliary modules such as the anterograde and retrograde intraflagellar transport modules and
the dynein-2 module. Due to the strong focus on ciliary proteins as baits and the integration of data from various sources, we
could also identify several new modules, potentially with cilia-associated functions in health and disease.
Proteomic Forum 2015
157
Conclusion: Our systems oriented approach, employing affinity proteomics to define the ciliary network has resulted in a
comprehensive description of known and candidate ciliary protein networks and modules, which can serve as a resource for
candidate ciliopathy proteins and our understanding of pathogenic mechanisms underlying ciliopathies.
P161
Clostridium difficile toxins A and B glucosylate small GTPases and alter the proteome of human cells
1
1
1
1
1
1
J. Junemann , J. Erdmann , J. Zeiser , N. Jochim , I. Just , R. Gerhard , A. Pich
1
Hannover Medical School, Institute of Toxicology, Hannover, Germany
1
The anaerobic bacterium Clostridium difficile is one of the most common nosocomial pathogens and triggers antibioticassociated gastrointestinal infections ranging from mild diarrhea to life-threatening pseudomembranous colitis. Toxins TcdA and
TcdB are the two major virulence factors of C. difficile that specifically glucosylate and inactivate small GTPases. The
consequences are reorganization of the cytoskeleton, loss of cell-cell contacts, and finally cell death.
Comprehensive shotgun data-dependent proteome analyses were conducted using human epithelial cell lines that were treated
with wild type or mutant toxins and quantified by labelling with heavy stable isotopes. Long (24 h) and short term (5 h) effects of
TcdA on the colonic proteome in Caco-2 cells were analyzed. The impact of TcdB was investigated in HEp-2 cells after 8 h of
intoxication. Proteins were analysed by gel-LC-MS. Results were verified by western blot and MRM analysis. The activity of
clostridial glycosylating toxins was evaluated in vivo by determining the glucosylation rate of target GTPases.
Wild type toxins induced considerable changes in the protein profile of epithelial cells. In case of TcdA more than 800 proteins of
the over 6000 identified proteins were altered in their abundance. Proteins that were differently abundant were primarily
involved in regulation, metabolic processes, endocytosis, organelle function, cell-cycle and cytoskeleton organisation.
Glucosyltransferase deficient TcdA induces only changes after short incubation periods. After treatment with TcdB 183 of 5320
identified proteins were affected. These proteins are involved in signal transduction, cytoskeleton organisation, cell-cycle and
cell death. The mutant TcdB revealed a pyknotic effect, which is glucosyltransferase independent but triggered the response of
129 proteins. Besides known targets of TcdA, glucosylation was additionally identified in Rap1(A/B), Rap2(A/B/C), Ral(A/B), and
(H/K/N)Ras which had not been considered as TcdA targets before. This proteome analysis demonstrates that clostridial
glucosylating toxins affect several cellular processes that have not been considered before.
P162
Time-resolved quantitative phosphoproteomics to study platelet signaling in health and disease
1
2
3
1
4
5
6
7
1,7
F. Beck , J. Geiger , S. Gambaryan , S. Loroch , O. Pötz , C. Fufezan , J. Heemskerk , U. Walter , A. Sickmann , R. Zahedi
1
Leibniz-Institut für Analytische Wissenschaften - ISAS - e.V., Systems Analysis, Dortmund, Germany
2
Interdisciplinary Bank of Biomaterials and Data, Würzburg, Germany
3
Sechenov Institute of Evolutionary Physiology and Biochemistry, S. Petersburg, Russian Federation
4
Natural and Medical Sciences Institute at the University of Tübingen, Reutlingern, Germany
5
Universität Münster, Institut für Biologie und Biotechnologie der Pflanzen, Müsnter, Germany
6
CARIM, Department of Biochemistry, Maastricht, Netherlands
7
Center for Thrombosis & Hemostasis, Mainz, Germany
1
More than 130 years ago it was recognized that platelets are key mediators of hemostasis. Nowadays, we know that platelets
participate in further physiological processes and contribute to the genesis and progression of cardiovascular diseases.
In circulating platelets the equilibrium between activation and inhibition is well-balanced to prevent uncontrolled platelet
aggregation and vessel clotting. Using quantitative time-resolved phosphoproteomics we studied two central pathways that
regulate platelet activation (ADP/P2Y12) as well as inhibition (prostacyclin/cAMP-PKA) and represent major targets of current
anti-platelet drug therapy.
We reasoned that a more detailed understanding of these central signaling pathways will help to identify critical hubs and
sensitive components that may serve as (i) predictors of patient susceptibility to medication, and (ii) potential targets for novel
anti-platelet treatment strategies.
Methods:Platelets were purified from healthy donors from fresh blood donations, using optimized and quality controlled
protocols. For each patient, purified platelets were then separated into 100 µg aliquots that either served as controls (resting
Proteomic Forum 2015
158
platelets) or were stimulated with ADP (activation), Iloprost (inhibition), or a combination of both, for different time points (10 s,
30 s, 60 s). After lysis and proteolytic digestion, samples were labeled with iTRAQ and phosphopeptides were enriched using
TiO2 with subsequent HILIC fractionation. Samples were analyzed by nano-LC-MS/MS and data analyzed using Proteome
Discoverer 1.4.
Results: We provide unprecedented quantitative insights into time-resolved changes for more than 4,797 high confidence
phosphopeptides. We identified a significant regulation in hundreds of proteins, some of which are involved in well-established
activatory functions such as degranulation and cytoskeletal re-organization, but also less well-understood pathways involving
ubiquitin ligases and GTPase exchange factors/GTPase-activating proteins (GEF/GAP).
Conclusion: Our data indicate for the first time that both platelet activation and inhibition involve numerous previously
unknown or unanticipated pathways. Many relevant signaling events seem to occur already 10 s after stimulation and include
other PTMs, indicating a role of PTM crosstalk. We can demonstrate a substantial crosstalk of pathways and an overlap in
signaling events after platelet activation and inhibition, and furthermore pinpoint previously unknown central signaling nodes
that show reciprocal regulation. Thus, quantitative time-resolved phosphoproteomics provides unprecedented insights into the
very processes that govern platelet homeostasis and improvements in sensitivity enabled us to start comparing
phosphoproteomic signatures between healthy donors and patient suffering from cardiovascular/bleeding disorders.
P163
Proteomic study of SET-mediated phosphorylation and regulatory analysis of nucleolin in trichloroethylene-induced hepatic
cytotoxicity
1
1
1
1
1
J. Liu , X. Ren , B. Xia , W. Liu , X. Yang
1
Shenzhen center for disease control and prevention, Key Laboratory of Modern Toxicology of Shenzhen, Shenzhen, China
Background: Trichloroethylene (TCE) is an environmental and occupational toxicant which can cause serious hepatotoxicity.
However, the toxic mechanisms of TCE remain poorly understood. We have previously reported that knockdown of SET
attenuated TCE-induced hepatic cytotoxicity. Moreover, SET was also reported as the inhibitor of phosphatase 2A. Hence to
further analyze the characteristic of SET in protein phosphorylation, we conducted a phosphoproteomic study in TCE-induced
cytotoxicity.
Methods: Proteins were extracted, digested and labeled with iTRAQ. They were then enriched by Immobilized Affinity
Chromatography (IMAC). Protein phosphorylation was identified and quantified by LC-MS/MS. Functional studies of SETmediated phosphorylation of nucleolin were conducted by Western-blot and co-IP analysis.
Results: In this study, 20 sites of 16 proteins were differentially phosphorylated in TCE-treated L-02 cells. Phosphorylation of two
apoptosis-related proteins 4E-BP1 and MCM2 were validated by Western-blot analysis.
Further analysis suggested that differential phosphorylation at 14 sites of 13 proteins were mediated by SET in TCE-induced
hepatic cytotoxicity. Interestingly, SET-mediated alteration of overall phosphorylation on a tumor progression related protein,
nucleolin, was observed in TCE-treated L-02 cells. Additional studies showed that the SET-mediated alteration of overall
phosphorylation of nucleolin could impact its self-regulation by modulating the binding and inhibition to c-myc. These findings
suggested a novel role of SET in nucleolin’s regulatory process, and they further revealed the molecular mechanism of SETmediated hepatic cytotoxicity of TCE.
Conclusions: Two apoptosis-related phosphorylation protein sites were validated in TCE-treated L-02 cells,which further
supported our previous observation of TCE toxicity. SET-mediated overall phosphorylation of nucleolin impacts the selfregulation of nucleolin by modulating its binding and inhibition to c-myc in TCE-induced hepatic cytotoxicity(Fig.1).
Acknowledgements: This work was supported by the NSFC[81273126], the Key Project of the Guangdong Natural Science
Foundation [S2012020010903], the Project of Shenzhen Basic Research Plan [JCYJ20130329103949656,
JCYJ20130329161137543,201201027], the Upgrade Scheme of the Shenzhen Municipal Key Laboratory
[JCYJ20130401102255980].
Proteomic Forum 2015
159
Figure 1
P164
A plasma proteomics study of experimental thyrotoxicosis in humans
1
2
1
3
4
1
2
1
B. Engelmann , A.- L. Dirk , E. Hammer , N. Friedrich , D. Führer , G. Homuth , G. Brabant , U. Völker
1
University Medicine Greifswald, Interfaculty Institute for Genetics and Functional Genomics, Greifswald, Germany
2
University of Luebeck, Experimental and Clinical Endocrinology, Med Clinic I, Luebeck, Germany
3
University Medicine Greifswald, Institute for Clinical Chemistry and Laboratory Medicine, Greifswald, Germany
4
Universitätsklinikum Essen, Clinic for Endocrinology and Metabolic Disorders, Essen, Germany
Introduction: Hyperthyroidism is a highly prevalent disease with a need for early diagnostic and prognostic markers to improve
therapy options. So far most of the studies have been performed with patients with endogenous hyperthyroidism with a wide
variability in the underlying pathogenesis and severity of the disease. Experimental thyrotoxicosis in healthy volunteers might
provide new insights into the impact of thyroid hormone changes at subclinical levels.
Objectives: The effects of experimentally induced thyrotoxicosis on the human plasma proteome using oral doses of
levothyroxine were tested. Using an unsupervised proteomic approach we assessed alterations in the amounts of potential new
marker proteins in comparison to well known parameters changed in thyrotoxicosis.
Patients & Methods: Plasma of 16 male healthy volunteers was sampled at baseline, 4 and 8 weeks under 0.25 mg/d
levothyroxine, as well as 4 and 8 weeks after discontinuation. Plasma proteins were analyzed after depletion of 6 high-abundant
proteins (MARS6) by LC-ESI-tandem mass spectrometry. MS raw data were processed using a label free intensity based
workflow. Time-dependent alterations in the protein pattern correlating with FT4 levels were identified according to a linear
regression based mixed model.
Results: All subjects developed biochemical thyrotoxicosis associated with strong TSH suppression and increased FT4 and FT3
plasma levels, which quickly returned to pretreatment levels after discontinuation of levothyroxine exposure. None of them
noticed any subjective symptoms of thyrotoxicosis. Plasma protein profiling covered 439 proteins, out of which 148 proteins
correlated significantly with FT4. Among those proteins, sex-hormone binding globulin (SHBG) could be quantified, a commonly
used marker of thyrotoxicosis. The positive correlation of FT4 with complement system proteins and coagulation factors is
pointing to an influence of hyperthyroid conditions on the coagulation system. Also for proteins of the acute phase response
Proteomic Forum 2015
160
signaling a linear association with FT4 could be found. The inverse effect was observed for a number of apolipoproteins for
which a higher abundance in the presence of low fT4 levels was detected.
Conclusion: Data underline the influence of only subclinical thyroid hormone levels on the plasma protein composition. Results
show that combining global plasma protein profiling with an experimental hyperthyroid model is a promising approach for the
identification of molecular markers of hyperthyroidism.
P165
Disease proteomics to study the deregulation of molecular mechanisms in inherited skin fragility
1
K. Thriene
1
Universitätsklinikum Freiburg, Molekulare Dermatologie I, Freiburg im Breisgau, Germany
Dystrophic epidermolysis bullosa (DEB) is an inherited skin disorder which leads to blister formation, abnormal wound healing,
and excessive scarring. It often results in aggressive skin cancer, which is the main reason for death in adult patients. The genetic
causes are mutations in the gene COL7A1 encoding the protein collagen VII, which is located in the extracellular matrix (ECM). As
a major component of anchoring fibrils collagen VII plays a crucial role in epidermal-dermal adhesion. Currently, there is no
effective therapy available and there is a high unmet clinical need.
To determine how the loss of collagen VII affects molecular mechanisms in primary human keratinocytes, we investigate the
composition and regulation of the cellular transcriptome, proteome, as well as the microenvironment using RNA-Seq and
unbiased, quantitative mass spectrometry (MS)-based proteomics analyses.
Loss of collagen VII has global implications on the cellular microenvironment. Comparing transcriptome and proteome of normal
human and DEB keratinocytes yields disease-specific alterations in gene transcription and protein translation. The latter is
currently being investigated by a pulse SILAC labeling approach. The combinations of transcriptome and proteome analyses as
well as saturation and pulse labeling approaches addressing protein abundance changes and turnover, respectively, enable us to
provide a detailed but also global image of the deregulation of molecular mechanisms in DEB. Loss of collagen VII leads to more
complex alterations as so far anticipated and our analyses suggest new disease-relevant targets for further analyses.
P166
Investigation of platelet proteome changes in human endotoxemia model
1
G. Schmidt
1
Center of Physiology and Pharmacology, Vienna, Austria
In endotoxemia and sepsis, where the blood coagulation system is hyperactivated due to systemic inflammation, platelets are
considered to play a major player. It has been described that platelet functions are influenced in sepsis, a state with systemic
inflammation and thrombotic dysfunction initiated by bacteria [1;2]. Lipopolysaccharide (LPS), which is part of the outer
membrane of gram negative bacteria, is recognized in platelets via toll-like-receptor 4 (TLR4) and induces secretion of proinflammatory and pro-coagulant proteins stored in the α-granula [3;4].
We analyse the platelet proteome after LPS challenge of 15 healthy male volunteers in a defined time course (-50min before and
after 2h, 6h and 24h) by two dimensional differential gel-electrophoresis (2D-DIGE) to characterize the role of platelets in
endotoxemia. In parallel the activation of platelets are validated with flow cytometry by the surface expression CD62P
expression which reflects degranulation degree of platelets.
First results showed the strongest change in the platelet proteome of LPS-challenged volunteers by a 40% up-regulation of
calmodulin (P < 0.01) a calcium-binding messenger protein and a highly significant up-regulation of seven isoforms from
Tropomyosin 1 was found. A significant change in the CD62P surface expression confirmed the activation of platelet in this
human endotoxemia model.
In the future we plan to confirm these LPS-related changes in the platelet proteome in vitro models with thrombin and LPS
primed platelets. Finally we will confirm these findings in patients with sepsis.
Proteomic Forum 2015
161
References:
1. Levi M. Platelets at a crossroad of pathogenic pathways in sepsis. Journal of thrombosis and haemostasis : JTH. 2004
Dec;2(12):2094-5. PubMed PMID: 15613011. Epub 2004/12/23. eng.
2. Vincent JL, Yagushi A, Pradier O. Platelet function in sepsis. Critical care medicine. 2002 May;30(5 Suppl):S313-7. PubMed
PMID: 12004253. Epub 2002/05/11. eng.
3. Stahl AL, Svensson M, Morgelin M, Svanborg C, Tarr PI, Mooney JC, et al. Lipopolysaccharide from enterohemorrhagic
Escherichia coli binds to platelets through TLR4 and CD62 and is detected on circulating platelets in patients with hemolytic
uremic syndrome. Blood. 2006 Jul 1;108(1):167-76. PubMed PMID: 16514062. Pubmed Central PMCID: PMC1895830. Epub
2006/03/04. eng.
4. Zhang G, Han J, Welch EJ, Ye RD, Voyno-Yasenetskaya TA, Malik AB, et al. Lipopolysaccharide stimulates platelet secretion and
potentiates platelet aggregation via TLR4/MyD88 and the cGMP-dependent protein kinase pathway. Journal of immunology
(Baltimore, Md : 1950). 2009 Jun 15;182(12):7997-8004. PubMed PMID: 19494325. Pubmed Central PMCID: PMC2787095. Epub
2009/06/06. eng.
P167
Investigation of the Platelet Proteome in a human endotoxemia model
1
1
1
2
1
G. Schmidt , A. Assinger , W. Schrottmaier , B. Jilma , M. Zellner
1
Medical University of Vienna, Center of Physiology and Pharmacologz, Vienna, Austria
2
Medical University of Vienna, Clinical Pharmacology, Vienna, Austria
In endotoxemia and sepsis, where the blood coagulation system is hyperactivated, platelets are considered to play a major role.
Lipopolysaccharide (LPS), which is part of the outer membrane of gram negative bacteria, is recognized in platelets via toll-likereceptor 4 (TLR4) and induces secretion of pro-inflammatory and pro-coagulant proteins stored in their α-granula.
We are analysing the platelet proteome after LPS challenge of 15 healthy male volunteers in a defined time course (-50min
before and after 2h, 6h and 24h) by two dimensional differential gel-electrophoresis (2D-DIGE) to characterise the function of
platelets in endotoxemia. In parallel the activation of platelets are validated with flowcytometry by the surface expression
CD62P expression which reflects degranulation degree of platelets.
First results showed the strongest change in the platelet proteome of LPS-challenged volunteers by a 40% upregulation of
calmodulin (P < 0.01) a calcium-binding messenger protein after 6h. Additionally, a highly significant upregulation of seven
isoforms from Tropomyosin 1 was found (P < 0.01), beside many other changes of different protein spots which will be identified
by mass spectrometry and further evaluated by pathway analysis. A significant change in the CD62P surface expression
confirmed the activation of platelet in this human endotoxemia model.
In the next steps these LPS-related changes will be validated in vitro models with thrombin- and LPS-primed platelets. Finally we
will confirm these findings in patients with sepsis.
P168
Identification of breast cancer biomarker signature for risk classification using a targeted proteomics approach
1
1
2
2
U. Korf , J. Sonntag , A. Wachter , T. Beissbarth
1
DKFZ, Molecular Genome Analysis (B050), Heidelberg, Germany
2
University of Göttignen, Institute for Medical Statistics, Göttingen, Germany
Reverse phase protein arrays (RPPA) present a quantitative platform for targeted proteomics with characteristic technical
features such as high sensitivity, high sample capacity, and high reproducibility [1].
RPPA was therefore used as large-throughput tool to screen for proteins associated with the risk of cancer progression in
patients with estrogen receptor-positive (ER+) breast cancer. Bioinformatic analysis of RPPA data involved different classification
approaches (PAM, SVM, RF) and identified a protein signature (CAV1, RPS6, Ki-67, NDKA) that resolved the intermediate risk
group into a low and a high risk group, potentially relevant for clinical decision-making [2] (figure 1). Perspectives of RPPA for
cancer biomarker discovery and cancer subtyping will be discussed, focusing on breast cancer as example.
Figure 1: RPPA-based biomarker discovery applied to risk classification in breast cancer.
Proteomic Forum 2015
162
[1] Akbani R, Becker KF, Carragher N, Goldstein T, de Koning L, Korf U, Liotta L, Mills GB, Nishizuka SS, Pawlak M, Petricoin EF
3rd, Pollard HB, Serrels B, Zhu J. Realizing the promise of reverse phase protein arrays for clinical, translational, and basic
research: a workshop report: the RPPA (Reverse Phase Protein Array) society (2014). Mol Cell Proteomics13, 162543.http://www.ncbi.nlm.nih.gov/pubmed/24777629
[2] Sonntag J, Bender C, Aulmann S, von der Heyde S, Nietert M, Lier A, Burwinkel B, Wiemann S, Beissbarth T, Sinn HP,
Schneeweiss A, Korf U. Protein biomarker signature for risk classification of hormone receptor positive breast cancer patients
identified by reverse phase protein array based tumor profiling (2014). Translational Proteomics 2, 5259.http://www.sciencedirect.com/science/article/pii/S2212963414000023
Figure 1
P169
Longitudinal Study of Urinary Proteome Using Targeted LC-MS Approach
1
2
1
N. Khristenko , I. Larina , B. Domon
1
CRP-Sante, Strassen, Luxembourg
2
Institute for BioMedical Problems, Moscow, Russian Federation
Introduction: The systematic and reproducible large-scale proteomic analysis of biological samples remains challenging. In
particular, urine samples, which are easily accessible for clinical investigation, but suffer from a high biological variability. The
analysis of such samples can benefit by applying targeted methods based on high resolution and mass accuracy mass
spectrometers.
Objectives: The aim of this study is to determine the inter- and intra- variability of protein abundances in urine samples
collected from well-controlled healthy donors, participating in the cosmonaut training program, using a targeted LC-MS
approach.
Proteomic Forum 2015
163
Materials and Methods: Urine samples were collected from healthy donors subjected to well controlled diet and sport activities.
The sample preparation protocol included protein precipitation using trichloroacetic acid, trypsin digestion and desalting using
C18 cartridges. The analyses of 480 proteins in urine samples were performed in a single LC-MS/MS run using nano -LC (Thermo
Scientific) coupled to a quadrupole-orbitrap mass spectrometer (Thermo Scientific) operated in the parallel reaction monitoring
(PRM) mode. The identity of the peptides was confirmed by matching the acquired MS/MS spectra with the spectra from a
curated reference library. The relative quantification of the peptide abundances was performed on the fragment ion intensity
signal using the Skyline software.
Results: The urine samples of six individuals living under controlled conditions were collected over 15 weeks and analyzed using
a targeted LC-MS approach. The proteotypic peptides monitored (1315) were selected from 480 proteins. As the truncated form
of the proteins often appears in urine due to internal proteolysis, typically three proteotypic peptides were used to capture the
full protein length. Diversity of ratio between the peptides corresponding to one protein across samples was indicative of the
truncated form.
The study was focused on detecting the relative changes in protein abundances between urine samples using the fragment ion
intensity of the corresponding peptides. The proteins detected in these samples covered concentrations of at least three orders
of magnitude, for example, the highly concentrated serotransferrin (P02787) with 640 ng/mL and the low concentrated Kunitztype protease inhibitor 1 (O43278) with 4.5 ng/mL.
Conclusion: The methodology developed for the systematic analysis of urine samples allowed to select a set of proteins which
show relatively low inter- and intra- individual variability. In contrast, the profile of other proteins in the longitudinal study was
used to monitor the difference upon changes of the environmental conditions (e.g., food, sport).
Proteomic Forum 2015
164
Poster Session C • 24 March 2015, 15:00–16:30
Plant and Microbial Proteomics II (P080–P091)
P080
Quantitative ChaFRADIC reveals that mitochondrial Icp55 cleavage specificity is conserved between yeast and Arabidopsis
1
2
2
1
S. Venne , C. Carrie , J. Soll , R. Zahedi
1
Leibniz Institute for analytical Science -ISAS e.V.-, Systems Analysis, Dortmund, Germany
2
Biozentrum derLMU München, Biology I - Botany, Planegg-Martinsried, Germany
Introduction: Most mitochondrial proteins contain an N-terminal targeting signal which is removed by specific proteases
following import. In yeast, we recently identified the novel peptidase Icp55 that removes N-terminal amino acids in accordance
with the N-end rule. In plants, however, so far only the mitochondrial processing peptidase (MPP) has been characterized.
Objectives: We used our novel ChaFRADIC strategy to determine the substrates and cleavage sites of the Arabidopsis thaliana
homologs to the yeast Icp55 and Oct1 proteins.
Materials & Methods: Recently we developed a novel method for enrichment of N-terminal peptides, called charge-based
fractional diagonal chromatography (ChaFRADIC) [1]. Here, after blocking of primary amines on the protein level, strong cation
exchange chromatography (SCX) is used to separate peptides into distinct charge state fractions at pH 2.7 (see figure 1). These
fractions are separately deutero-acetylated, inducing a charge state reduction of internal peptides whereas N-terminal peptides
remain unchanged and can be enriched in a second dimension SCX. We used quantitative ChaFRADIC to compare mitochondria
from wildtype Arabidopsis versus Icp55 and Oct1 knock out mutants (see figure 2).
Results: We could identify 92 Icp55 and 7 putative Oct1 substrates. Indeed, Arabidopsis Icp55 has an almost identical cleavage
site when compared to yeast. However, in Arabidopsis it can remove a greater range of amino acids. The putative Oct1
substrates from Arabidopsis displayed no clear consensus cleavage motif, and do not contain the classical -10R motif identified
in other eukaryotes. Notably, Oct1 seems to cleave presequences independently, without the prior cleavage of MPP [2].
Conclusion: We could use quantitative N-terminal ChaFRADIC to demonstrate that Icp55, having a direct impact on the stability
of the mitochondrial proteome, most likely evolved early in the evolution and shows a conserved consensus motif.
[1] Venne et al, Novel Highly Sensitive, Specific, and Straightforward Strategy for Comprehensive N-Terminal Proteomics Reveals
Unknown Substrates of the Mitochondrial Peptidase Icp55. J Proteome Res 2013, 21, 21.
[2] Carrie et al, Identification of Cleavage sites and Substrate Proteins for Two Mitochondrial Intermediate Peptidases in
Arabidopsis thaliana. 2014 (submitted).
Figure 1
Proteomic Forum 2015
165
Figure 2
081
Efficient purification of dilute samples for quantitative proteomics studies
1
1
2
2
1
1
F. Bonn , J. Bartel , K. Büttner , M. Hecker , A. Otto , D. Becher
1
EMAU Greifswald, Department of Microbial Proteomics and Mass Spectrometry, Greifswald, Germany
2
EMAU Greifswald, Institute for Microbiology, Greifswald, Germany
In sample preparation for proteomics a common encountered issue is a low concentration of the analyte in a dilute sample.
Therefore it is required to deal with highly reproducible and effective enrichment of proteins especially in quantitative
proteomics. To tackle this, we have set up a protocol for protein enrichment with StrataClean beads. Highly efficient protein
enrichment was achieved by optimizing several steps of the protocol, including a priming step at elevated temperature with
hydrochloric acid. The optimized protocol is compatible with common mass spectrometry based proteomic workflows like GeLCMS as well as with digest on beads for direct LC-MS/MS measurements.
To test the performance of the new protocol, StrataClean enrichment was compared with classical methods demonstrating
reproducible enrichment of the whole mass range of proteins contained in a bacterial cell extract at full advantage of the onepot/ fast protocol of SPE by Strataclean beads. Superior efficiency of SPE by Strataclean beads was confirmed for highly dilute
samples containing 20µg protein extract in 200ml of sample comparing Strataclean beads with detergent assisted TCA
precipitation as gold standard. This represents an enrichment of factor 10,000. GeLC-MS analysis of the enriched samples
revealed higher absolute identification rates as well as higher reproducibility of StrataClean enrichment compared to the
classical detergent assisted TCA precipitation. Unimpaired, bias free enrichment was demonstrated by label-free absolute
quantification after on-bead digest of enriched proteins. In addition we suggest a novel storage/ shipping workflow for trapped
proteome samples at ambient temperature.
In conclusion, StrataClean enrichment represents a universal sample preparation method for diluted proteome samples. The
workflow combines superior efficiency with the advantage of easy handling and a minimal use of hazardous chemicals.
P082
Characterization of the protein complexes associated to the phosphatidylinositol 3-kinase during Phaseolus vulgaris-rhizobia
interaction
1
1
1
1
C. V. Dorantes Torres , G. Estrada Navarrete , F. E. Sanchez Rodriguez , N. D. J. Cruz Mireles
1
Biotechnology Institute of National Autonomous University of Mexico, Molecular Biology of Plants, Cuernavaca, Mexico
Introduction: Phaseolus vulgaris (common bean) as other legumes is able to establish symbiotic relationships with bacteria of
the genus Rhizobium. The symbiosis between common bean and Rhizobium is a process that requires a constant molecular
Proteomic Forum 2015
166
exchange of signalling molecules between these two organisms, thus allowing the onset of biological nitrogen fixing nodules.
The phosphatidylinositol 3-kinase class III (PI3KC3) can scaffold protein multicomplexes that are involved in vesicle trafficking,
autophagy and phagocytosis in mammals and yeast. This work unravels that this kinase has a key role during nodule
organogenesis and the onset of the symbiotic nitrogen fixation process
Objectives:
1) To design and obtain the molecular tools to perform tandem affinity purification (TAP) of the PI3KC3 asssociated protein
complexes.
2) To isolate by TAP the protein complexes associated to PI3KC3 from root-nodule symbiosomes of P. vulgaris transgenic roots
inoculated with R. tropici.
3) To analyse by LC-MS the protein complexes associated to PI3KC3.
Methods: Isolate protein complexes associated to PvPI3KC3 of bean transgenic roots during the symbiotic process.
Results: We generated transgenic roots of P. vulgaris inoculated with R. tropici that express a TAP-PI3KC3 construction under the
regulation of its own promoter.
Conclusion: These work aims to obtain an outlook about the protein complexes associated to PI3KC3 in nodule ontogeny and
the nitrogen fixation process.
The work was funded by CONACyT No. 177744 and DGAPA IN-222012.
P083
Impact of Matrix-Assisted Laser Desorption Ionization-Time of Flight (MALDI-TOF) in the routine diagnostic Microbiology
Laboratory in Alexandria University, Egypt
1
1
2
N. Abouseada , M. Moheb , P. Moez
1
Alexandria Faculty of Medicine , Microbiology & Immunology, Alex, Egypt
2
Alexandria Faculty of Medicine , Clinical Pathology , Alex, Egypt
In the last few years, matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) has emerged as a very useful tool for
routine identification of microorganisms in clinical microbiology laboratories. Identification of the various bacterial genera and
species has been found to be challenging using the routine phenotypic tests or commercially available automated systems
relying on the biochemical identification of bacteria. In this study, we aimed to evaluate the possible role of the MALDI-TOF
system with its Biotyper software, which is dedicated to microbial identification, as an accurate and rapid method of bacterial
identification. We also evaluated its superior role in identification of rare bacterial species, which were difficult to diagnose or
were misdiagnosed using the conventional techniques.
P084
Why database matters: impact of sequence databases on metaproteome analysis of a mock microbial mixture
1
1
2
1
1
1
1
1
A. Tanca , A. Palomba , M. Deligios , T. Cubeddu , C. Fraumene , G. Biosa , D. Pagnozzi , M. F. Addis , S. Uzzau
1
Porto Conte Ricerche, Alghero, Italy
2
Università di Sassari, Dipartimento di Scienze Biomediche, Sassari, Italy
1,2
Introduction: Metaproteomics allows the investigation of the protein repertoire expressed by complex microbial communities.
To exploit its full potential, improvement in bioinformatic approaches for data analysis are still needed. In this context, sequence
database selection represents a key challenge.
Objectives: This work evaluated the impact of different databases in metaproteomic investigations by analyzing a mock
microbial mixture.
Materials & Methods: The microbial mixture included 7 prokaryotes and 2 eukaryotes. DNA from the 9 individual strains and
from the mixture were subjected to Illumina sequencing, to generate genome- and metagenome-derived protein databases,
which were used along with public databases (namely, NCBI, UniProtKB/SwissProt and UniProtKB/TrEMBL, parsed at different
taxonomic levels) to analyze the metaproteomic dataset achieved by shotgun LTQ-Orbitrap MS analysis. Taxonomic attribution
was carried out according to the lowest common ancestor approach using MEGAN and Unipept.
Proteomic Forum 2015
167
Results: The quantitative comparison demonstrated that only 35% of peptides identified were common to all database classes;
moreover, genus/species-specific databases provided up to 17% more identifications compared to databases with generic
taxonomy, while the metagenomic database enabled a minor increment in respect to public databases (Fig. 1). In addition,
public databases with generic taxonomy exhibited a markedly different behavior in terms of false discovery rate and peptide
degeneracy compared to the counterparts. Furthermore, the level of taxonomic attribution misassignments varied among the
different databases, and specific thresholds based on the number of taxon-specific peptides were established to minimize false
positives.
Conclusion: This study reveals that database selection plays a pivotal role in metaproteomics, and suggests the use of iterative
searches and suitable filters for taxonomic assignments to improve depth and trustworthiness of metaproteomic results.
Fig. 1. Comparison of metaproteomic data obtained with different databases. A) Peptide sequences (left) and peptidespectrum matches (PSMs, right) identified in the 9MM using different sequence databases. B) Venn diagrams illustrating the
peptide distribution among four different DB classes (left), NCBI-, TrEMBL- and SwissProt-based DBs (center), DBs with generic
microbial taxonomy (BFV), genus-specific taxonomy (G), and species-specific taxonomy (S) (right).
Figure 1
P085
A straightforward and efficient analytical pipeline for metaproteome characterization
1
1
1
2
1
2
1
1
A. Tanca , A. Palomba , S. Pisanu , M. Deligios , C. Fraumene , V. Manghina , D. Pagnozzi , M. F. Addis , S. Uzzau
1
Porto Conte Ricerche, Alghero, Italy
2
Università di Sassari, Dipartimento di Scienze Biomediche, Sassari, Italy
1,2
Introduction: The massive characterization of microbial communities is a recent breakthrough in life sciences. Metaproteomics
specifically enables to make the transition from assessing the genomic potential to actually measuring the functional expression
of a microbiome. However, significant research efforts are still required to develop analysis pipelines optimized for
metaproteome characterization.
Objectives: This study was aimed at developing an efficient and time-effective analytical pipeline for shotgun metaproteomic
analysis. Mock microbial mixtures were used to test efficiency, sensitivity and dynamic range of the pipeline. Then, the pipeline
Proteomic Forum 2015
168
was applied to the mouse fecal metaproteome (MFM), with the aim of validating its ability to provide reliable, reproducible and
deep taxonomic and functional information from a complex microbiome.
Materials & Methods: The pipeline combines bead-beating/freeze-thawing for protein extraction, filter-aided sample
preparation (FASP) for clean-up and digestion, and single-run liquid chromatography-tandem mass spectrometry for peptide
separation and identification. Sequest-HT, matched experimental metagenomic sequences and Percolator were employed as
search engine, database and peptide validation tool, respectively.
Results: Over 15,000 non-redundant peptide sequences were identified per run in the mock mixtures, with a linear dynamic
range from 10^4 to 10^8 colony-forming units. When applied to the MFM, the pipeline led to the detection of over 13,000 nonredundant microbial peptides (FDR <1%), belonging to over 600 different microbial species and 250 functionally relevant protein
families. An extensive mapping of the main microbial metabolic pathways actively functioning in the gut microbiome was also
achieved (Fig. 1).
Conclusion: The analytical pipeline presented here may be successfully used for the in-depth and time-effective characterization
of complex microbial communities, such as the gut microbiome, and represents a useful tool for the microbiome research
community.
Fig. 1. Metabolic pathway analysis of the murine fecal metaproteome. Top: distribution of the identified proteins belonging to
6 microbial kingdoms/phyla (see color legend) into metabolic pathways using iPATH. Bottom: detail of the Wood-Ljungdahl
pathway (left) and of the dissimilatory sulfate reduction pathway (right). Bold type: identified proteins; brackets: peptidespectrum matches (PSMs); italic: taxonomic assignments.
Figure 1
Proteomic Forum 2015
169
P086
Enrichment or depletion? The impact of stool pretreatment on metaproteomic characterization of the human gut microbiota
1
1
1
1
1,2
A. Tanca , A. Palomba , S. Pisanu , M. F. Addis , S. Uzzau
1
Porto Conte Ricerche, Alghero, Italy
2
Università di Sassari, Dipartimento di Scienze Biomediche, Sassari, Italy
Introduction: Up to date, most metaproteomic studies of the gut microbiota employ stool sample pretreatment methods to
enrich for microbial components. However, a specific investigation aimed at assessing if, how and to what extent this may
impact on the final taxonomic and functional results is still lacking.
Objectives: The aim of this work was to assess the impact of stool pretreatment (i.e., differential centrifugation) on the
metaproteomic results, concerning both taxonomic and functional data.
Materials & Methods: Here, stool replicates were either pretreated by differential centrifugation (DC) or not centrifuged (NC).
Protein extracts were then processed by filter-aided sample preparation, single-run LC and high-resolution MS, according to a
metaproteomic pipeline recently optimized by our group (Tanca A. et al., Microbiome 2014) and the metaproteomic data were
compared by spectral counting.
Results: DC led to a higher number of identifications, a significantly richer microbial diversity, as well as to reduced information
on the non-microbial components (host and food) when compared to NC. Nevertheless, dramatic differences in the relative
abundance of many gut microbial taxa were also observed, including a significant change in the Firmicutes/Bacteroidetes ratio.
Furthermore, several microbial functional categories, including cell surface enzymes, membrane-associate proteins, and flagella,
were significant reduced after DC.
Conclusion: In conclusion, this work underlines that a critical evaluation is needed when selecting the appropriate stool sample
processing protocol in the context of a metaproteomic study, depending on the specific target to which the research is aimed.
Fig. 1. A) Cladogram showing a hierarchical representation of the taxa identified in this study. Taxa with significantly different
abundance between NC and DC are colored. B) Bar graph illustrating the top 12 GO biological processes according to the NSAF
abundances of the related proteins. For each GO-BP category, the abundance values corresponding to the two main phyla
(Firmicutes and Bacteroidetes) are reported. Asterisks indicate a statistically significant difference between groups (p < 0.01).
Figure 1
Proteomic Forum 2015
170
P087
Functional proteomics on Arabidopsis MAP3K mutants and MAP2K overexpressor line
1
1
1
2
1
1
1
1
T. Takac , P. Vadovic , S. Bekesova , T. Pechan , I. Luptovciak , V. Smekalova , M. Ovecka , J. Samaj
1
Centre of the Region Hana, Faculty of Science, Palacky University , Department of Cell Biology, Olomouc, Czech Republic
2
Mississippi State University, Institute for Genomics, Biocomputing, Mississippi, United States
Introduction: Along with transcriptomic studies, proteomic and phosphoproteomic approaches provide significant functional
information about mitogen-activated protein kinases (MAPKs) in plants. Here we provide a shot-gun proteomic analyses of
Arabidopsis thaliana plants overexpressing Medicago sativa stress-induced MAPKK (SIMKK) and about loss- and gain-of-function
Arabidopsis yda1 and ΔNyda1 mutants in MAPKKK4 gene called YODA.
Objectives: Our aim was to utilize shot-gun proteomic analysis for characterization of these three transgenic lines and to reveal
which molecular mechanisms determine the aberrant root growth of yda1 and ΔNyda1 as well as increased salt stress sensitivity
of SIMKK-YFP overexpressors.
Materials and Methods: Fourteen days old Ler, yda1, ΔNyda1, Col0 and seedlings overexpressing SIMKK-YFP construct were
used for proteomic analyses. Preparation of trypsin-digested extracts (phenol extraction followed by in solution trypsin
digestion) was performed as described by Takáč et al. (2011). Ler, yda1, and ΔNyda1 samples were separated and analyzed on a
nanoAcquity 2D-UPLC system directly coupled to Xevo G2-S Q-TOF tandem mass spectrometer and roots of SIMKK-YFP
constructs were analyzed by 1D nanoACQUITY ultraperformance liquid chromatography (UPLC).
Results: Comparative proteomic analyses of SIMKK-YFP overexpressor line versus wild type showed decreased abundance of
proteins involved in salt-induced oxidative stress, which may determine the lower tolerance of SIMKK-YFP line to salt stress.
Similarly, the auxin-related phenotypes of yoda and ΔNyda1 mutants were accompanied by changes in abundances of auxin
biosynthesis proteins. The proteomes of the transgenic and mutant plants will be presented by means of protein function.
Conclusion: When supported by appropriate biochemical and cell biological validation we consider this proteomic approach
feasible for functional characterization of respective kinases including molecular interpretation of phenotypes and stress
responses of such transgenic and mutant plants.
References: Smékalová V, Luptovčiak I, Komis G, Šamajová O, Ovečka M, Doskočilová A, Takáč T, Vadovič P, Novák O, Pechan T,
Ziemann A, Košútová P, Šamaj J (2014) New Phytologist 203, 1175-1193.
Ovečka M, Takáč T, Komis G, Vadovič P, Bekešová S, Doskočilová A, Smékalová V, Luptovčiak I, Šamajová O et al. (2014) Journal
of Experimental Botany 65(9), 2335-2350.
Acknowledgements: This work was funded by National Program for Sustainability I (grant no. LO1204) provided by the Czech
Ministry of Education and by grant No. P501/11/1764 from the Czech Science Foundation GAČR, and P.V. was supported by the
grant CZ.1.07/2.3.00/30.0041 POSTUP II at Palacký University, Olomouc, Czech Republic.
P088
Quantifying Protein Synthesis and Degradation in Arabidopsis
1
1
2
1
P. Ihmor , W. Gruissem , M. Robinson , K. Baerenfaller
1
ETH Zurich, Plant Biology, Zurich, Switzerland
2
University of Zurich, Institute of Molecular Life Sciences, Zurich, Switzerland
Question: Recent time-resolved studies integrating transcriptomics and proteomics data have shown that a substantial part of
the proteome of higher eukaryotes displays dynamics that are decoupled from the corresponding transcript dynamics. Although,
substantial knowledge on various regulatory mechanism are known, it is currently unclear how they interact to determine the
final protein amounts.
Methods: We are approaching this question by quantifying transcription, translation, protein quantities and protein degradation
rates in Arabidopsis suspension cell cultures. For this, we employ polysome profiling and dynamic SILAC (stable isotope labeling
by amino acids in cell culture) experiments in combination with protein and transcript quantifications. These four genome-wide
assessments together will allow us to obtain rates for mRNA translation and protein degradation. Based on this data we will
Proteomic Forum 2015
171
develop statistical and mathematical ODEs (ordinary differential equations) to model the dynamics of protein level
determination.
Results: Until now we have completed establishing the required methods for plant suspension cultures. SILAC experiments for
instance require high labeling efficiencies of externally supplied, labeled amino acids, which were yet not obtainable in
autotrophic plant systems, and dynamic SILAC experiments have not been reported for plant systems. With our protocol, we
manage to reach labeling rates of over 85% and preliminary data on protein turnover assessments show that the rates
determined for Arabidopsis housekeeping genes correspond with those reported in small-scale studies employing 15N metabolic
labeling and DIGE (difference gel electrophoresis)
Conclusion: With the methodology set, we aim to assess shortly the impact of protein degradation and translational efficiency
on protein dynamics in plant cells after stimulation by the pathogen-derived flagellin peptide.
P089
Global and quantitative study of protein complexes in Arabidopsis thaliana.
1
1
1
M. Górka , A. Graf , D. Walther
1
Max-Planck-Institut für molekulare Pflanzenphysiologie, Potsdam, Germany
Most proteins are not functional as isolated polypeptides but form protein complexes. The size of such complexes ranges from
simple homodimers to large heteromeric structures. The majority of cellular processes are carried out by protein complexes and
their composition and abundance affects metabolic fluxes. Therefore, global and quantitative analysis of protein complexes is
needed to understand how the organization of proteins into functional units is connected to the regulation of metabolic
pathways.
In this project we aim to establish a map of protein complexes which can be reliably detected in the model plant Arabidopsis
thaliana. We use size exclusion chromatography (SEC) or native PAGE for the fractionation of protein complexes. Fractions are
prepared for mass spectrometry analysis and measurements are performed by nanoHPLC MS/MS. Following the identification of
proteins, distribution profiles along the molecular weight gradient are subjected to deconvolution and clustering analysis.
Statistical analysis is supported by previously published data on protein complexes which serve as true positive interactions.
As proof of concept, a focused analysis of the Arabidopsis proteasome was performed. Data presented here shows, that the
regulatory and the core particle of the proteasome were identified with high coverage. Moreover, RPN5B which was previously
thought to be part of the regulatory proteasome subunit was shown to be part of the core particle. Extending our analysis to a
range of metabolic pathways we were able to identify several new putative protein complexes. Of special interest is the
observed interaction between three enzymes of the Arabidopsis starch degradation pathway which might be important for the
regulation of the pathway in response to changes in environmental conditions.
In the future we will implement our established sample preparation and data analysis pipeline to quantitatively compare the
abundance of protein complexes on a proteome wide scale in different environmental conditions or genetic backgrounds.
P090
A redox dependent interaction between ROC4 and GWD might play a role in regulating starch degradation in plants.
1
2
C. Swart , A. Skeffington
1
Max-Planck Institute for Molecular Plant Physiology, Ag-Graf, Proteomics, Potsdam-Golm, Germany
2
Max-Planck Institute for Molecular Plant Physiology, Potsdam-Golm, Germany
Introduction: During the day plants synthesize carbon rich molecules by photosynthesis. Partial storage of the carbon takes
place as starch molecules that are accumulated in the chloroplasts of photosynthetically active leaves. At night, when
photosynthesis is not possible, the starch is linearly degraded to support metabolism and growth of the plant. Tight regulation is
required to ensure that a supply of carbon is available for the entire period of darkness. Glucan water dikinase (GWD) has been
shown to play an essential role in starch degradation in Arabidopsis thaliana, where it functions by catalyzing the
phosphorylation of the starch granule surface. Loss of function mutants for GWD have a strong starch excess phenotype,
suggesting it has the potential to be a key regulator of the pathway. We have recently identified the cyclophilin ROC4 as a
potential GWD interactor. Known to have a function in the redox state of the chloroplast, ROC4 could very well function as a
redox regulator of GWD.
Proteomic Forum 2015
172
Objectives: We aim to investigate the interaction between GWD and ROC4 and its role for the regulation of the starch
degradation pathway in the model plant A.thaliana.
Materials and methods: Proteins were extracted from A. thaliana wild type and mutant plants. Following fractionation by SDSPAGE and in-gel digestion, samples were analysed by nanoHPLC MS/MS. Label-free quantitative analysis of the obtained data
was performed using the Progenesis LCMS software (Nonlinear Dynamics). To study whether the interaction between GWD and
ROC4 is affected by the redox state of the proteins, native protein extracts were treated with different concentrations of DTT or
H2O2 for 5 minutes and protein abundances in the high molecular weight fraction (> 100 kDa) was analysed as described above.
Results and discussion: Proteomic analysis of wild type A. thaliana extracts led to the interesting observation that a small
protein, ROC4 (28 kDa) was present in the high molecular weight fraction of the gel. Silencing of GWD by RNAi and sequential
investigation showed that not only GWD, but ROC4 also disappeared from the upper fraction of the gel. This demonstrates that
ROC4 presence in the upper fraction is reliant on that of GWD, revealing a potential interaction. This behavior was not observed
for any other proteins in the dataset. Initial experimental data also seems to indicate that the relative abundance of ROC4 in the
high molecular weight range is based on the redox state of the proteins, suggesting a redox-dependent interaction with GWD.
Conclusions: Preliminary findings point to an interaction between GWD and ROC4. The redox state of the proteins appear to
influence this interaction. Redox regulation of GWD could play a role in regulating starch degradation in plants, thereby ensuring
a continued carbon supply for metabolism and growth throughout the night.
P091
Investigation of Hülle cells from the filamentous fungus Aspergillus nidulans
1
B. Dirnberger
1
University of Göttingen, Göttingen, Germany
Hülle cells are unique cell types of the Aspergillus genus and have crucial functions during sexual development. These globose
thick- walled cells develop from hyphae and associate with the developing closed sexual fruiting bodies named cleistothecia (1).
Hülle cells have different nuclei and they can fuse to form a macronucleus (2). The absence of Hülle cells results in the formation
of significantly smaller cleistothecia (3). This fact provides a possible auxiliary and a nursing function of Hülle cells during the
development of the cleistothecium. To explore the molecular switches from filamentous growth to single Hülle cells we are in
the process of analyzing the transcriptome, proteome and metabolome of these cells and therefore we have developed
methods to isolate them from the remaining fungal structures and we have compared these results to other fungal structures
such as the hyphae. In an integrated approach the metabolic secretome of Hülle cells will be investigated and compared with
data of other fungal structures such as the hyphae.
Post-Translational Modifcations II (P108–P122)
P108
A proteomic investigation of proteasome malfunctioning.
1
J. Demmers
1
Erasmus University Medical Center, Proteomics Center, Roterdam, Netherlands
In cancer therapeutics, the induction of apoptosis by inhibition of the proteasome using chemical agents is widely used, for
instance for the treatment of myeloma patients. However, such proteasomal inhibitors block catalytic subunits of the
proteasome and may cause severe side-effects. Therefore, more selective proteasome inhibitors that affect the regulatory
subunit of the proteasome (‘third-generation’ inhibitors) are being developed. Here, we aim to dissect the functional modules of
the proteasome by profiling both the dynamic proteome and ubiquitinome by targeted inactivation of specific subunits. This
way, better targeting strategies for proteasome inhibition can be developed.
The effect of proteasome inhibition and malfunctioning on the global Drosophila S2 proteome and ubiquitinome is monitored
using a SILAC proteomics approach. Inhibition of the proteasome is accomplished either by using chemical agents
(MG132/lactacystin) or by using selective RNAi knockdown of subunits. For the identification and relative quantitation of
ubiquitinated proteins we use recently developed protocols based on immunoprecipitation of proteolytic diGly peptides.
Proteomic Forum 2015
173
In these SILAC-MS screens, >5,000 proteins are identified and quantified routinely. After relatively short incubation times with
inhibitors, proteins that show severe and relatively fast upregulation and/or accumulation are associated with functional
categories such as stress response, cell cycle regulation, apoptosis and the ubiquitin-proteasome system. After longer incubation
times, and as an effect of simultaneous RNAi of knockdown of several proteasomal subunits, the abundances of several 100s of
proteins are altered. In addition, the extent of protein ubiquitination is dramatically increased upon inactivation of the
proteasome. Currently, the effects of more subtle perturbations, such as the selective knockdown of individual regulatory and
catalytic proteasomal subunits, are investigated using this approach. This could shed more light on the yet unknown roles of the
three proteasome-bound DUBs. Overall, these results are expected to give more insight into the mechanism and specificity of
proteasome functioning.
P109
A Strategy for high throughput lysine methylation screening
1,2
2
1
F. Richter , G. Mittler , I. Wittig
1
Goethe University Frankfurt, Medical School, Functional Proteomics, SFB815 Core Unit, Frankfurt am Main, Germany
2
Max Planck Institute of Immunbiology and Epigenetics, Proteomics Facility, Freiburg i. Br., Germany
Each type of posttranslational modification require adapted and advanced strategies for enrichment, identification and
quantitative analysis. Mass Spectrometry based identification of lysine and arginine methylation and its integer multiplicity by
detecting a 14.015650 a.m.u. mass shift is hampered by misinterpretation of (i) amino acid exchanges (e.g. alanine to glycine),
(ii) false annotation of two modifications (iii) generation of in gel digestion artefacts at aspartate, glutamate and threonine.
Application of methanol-D4 facilitates [IW1] identification of glutamate and aspartate artefacts in data base search. The most
straightforward technique is heavy methyl-SILAC using heavy methionine as substrate for intra cellular generation of heavy
biological methyl donor S-adenosylmethionine (SAM)[1]. For samples SILAC is not suitable we here introduce a strategy to use
heavy dimethyl labelling [2] to facilitate in vivolysine methylation screening. Proteins are methylated prior to digestion as an
additional step in the common filter-aided sample preparation (FASP) protocol [3] to ensure complete removal of the labelling
reagents. Thus we introduce peak triplets of 0, 2 or 4 a.m.u. for each lysine depending on the intrinsic-, heavy di- and mixed dimethylation. Applying this strategy to histone proteins confirmed methylation sites and unmasked artefacts. In few cases
histone methylations were detected by the dimethyl labelling not identified earlier by heavy methyl SILAC indicating SAM
independent methylation events. This example points to requirements of high throughput methods that enable unambiguous
identification of methylation events beyond the epigenetic signalling on tissue samples that cannot be analysed by heavy methyl
SILAC. An addition open question is to which extent methylation is present in eukaryotic subcellular compartments such as the
mitochondria. A recent report discovered a first mitochondria oriented methyltransferase METTL20 [4]. We applied the dimethyl
labelling strategy to study the impact of lysine methylation in mammalian mitochondria.
[1] Ong SE, Mittler G, Mann M. Identifying and quantifying in vivo methylation sites by heavy methyl SILAC. Nat Methods. 2004,
1:119-26.
[2] Boersema PJ, Raijmakers R, Lemeer S, Mohammed S, Heck AJ. Multiplex peptide stable isotope dimethyl labeling for
quantitative proteomics. Nat Protoc. 2009,4:484-94.
[3] Wiśniewski JR, Zougman A, Nagaraj N, Mann M. Universal sample preparation method for proteome analysis. Nat Methods.
2009,6:359-62.
[4] Rhein VF, Carroll J, He J, Ding S, Fearnley IM, Walker JE. Human METTL20 methylates lysine residues adjacent to the
recognition loop of the electron transfer flavoprotein in mitochondria. JBC 2014, 289: 24640-24651.
P110
Identification of targets of the Ser/Thr kinase PrkC from staphylococcus aureus by quantitative phosphoproteomics
1
1
M. Steger , A. Schlosser
1
University of Würzburg, Rudolf-Virchow Center for Experimental Biomedicine, Würzburg, Germany
Introduction: The phosphorylation of serine, threonine and tyrosine in proteins is a widespread occurring posttranslational
modification used by eukaryotes to regulate enzyme activity and in signaling networks underpinned by the fact that protein
kinases, catalyzing the phosphorylation constitute a large protein family containing about 500 members [1]. Several studies shed
light on the fact that this modification also plays an important role in prokaryotes [2][3].
Objectives: In this study we applied quantitative phosphoproteomics in order to identify the cellular target proteins and target
phosphorylation sites of the bacterial Ser/Thr kinase PrkC. For this purposes we knocked out PrkC, the phosphatase Stp or both
Proteomic Forum 2015
174
and used dimethyl-labeling to quantify changes in the degree of phosphorylation on specific sites compared to untreated
wildtype cells.
Materials and methods: Wild type and transfected staphylococcus aureus cells were grown in cell culture, cells were lysed in the
presence of protease and phosphatase inhibitors and the proteins digested with LysC and trypsin. After dimethyl- labeling of the
derived peptides a small aliquot was used to quantify protein abundance. The major part of the sample was used for the
enrichment of phosphopeptides by Phos-TiO-cartridges (GL sciences).
Peptides were separated on an Easy-nLC HPLC system and analyzed with an Orbitrap Velos Pro mass spectrometer coupled via
an electrospray interface. Peptide and protein identification and quantitation was done with MaxQuant software.
Results: In a qualitative approach about 110 phosphopeptides were identified, leading to a comprehensive map of 125
phosphorylation sites on 64 different proteins in wild type cells. The serine/threonine protein kinase PrkC itself that was
knocked out in subsequent quantitative studies shows complex clusters of phosphorylation sites in different regions.
In the subsequent quantitative approach 740 proteins were quantified with at least 3 peptides identified in every sample with a
small set of proteins showing a change in abundance. The quantitation of the phosphopeptides showed distinct changes in
abundance of phosphorylation sites with the majority down regulated in PrkC knockdown cells compared to the wild type.
Conclusion: The combination of dimethyl-labeling and phosphopeptide enrichment with titanium dioxide allows for the
comprehensive qualitative and quantitative analysis of the phosphoproteome of staphylococcus aureus and can help to
elucidate the function of the bacterial kinase PrkC.
[1] Manning, G., et al., Science 298, Nr. 5600: 1912-34. doi:10.1126/science.1075762.
[2] Macek, B., et al., Molecular & Cellular Proteomics 6, Nr. 4: 697-707. doi:10.1074/mcp.M600464-MCP200.
[3] Pereira, et al., Microbiology and Molecular Biology Reviews 75, Nr. 1: 192-212. doi:10.1128/MMBR.00042-10.
P111
Subtilisin - an alternative enzyme for PTM research
1
1
1
1
H. Gonczarowska-Jorge , M. Dell`Aica , C. Dickhut , R. P. Zahedi
1
Leibniz – Institut für Analytische Wissenschaften - ISAS - e.V., Dortmund, Germany
Introduction: In recent years, mass spectrometry-based phosphoproteomics has propelled our knowledge about the regulation
of cellular pathways. Nevertheless, typically applied bottom-up strategies have several limitations. Trypsin, the preferentially
used proteolytic enzyme shows impaired cleavage efficiency in the vicinity of phosphorylation sites. Moreover, depending on
the frequency and distribution of tryptic cleavage sites (Arg/Lys), generated peptides can be either too short or too long for
confident identification using standard LC-MS approaches.
Objective: To overcome these limitations, we evaluated the use of an alternative and simple approach based on the usage of the
non-specific serine protease subtilisin, which should provide access to ‘hidden’ areas of the proteome.
Material and Methods: We used 125 µg of non-stimulated HeLa to (i) evaluate the reproducibility of subtilisi-based
phosphoproteomics and (ii) compare the results with a typical trypsin-based workflow. Phosphopeptide enrichent was
conducted using TiO2, samples were analyzed by nano-LC-MS/MS and data analyzed using Proteome Discoverer 1.3. Only
phosphopeptides with phosphoRS probabilities>99 % were considered.
Results: Comparing three independent subtilisin digests of HeLa, we obtained more than 80% overlap considering peptide
sequence and phosphorylation site, compared to 87% for trypsin. Despite the increased search space and thus higher probability
for false-positive identifications we could identify a similar number of highly confident phosphorylation sites with subtilisin
(1797 considering just two out of three replicates and 2559 considering all identifications) and trypsin (2143 considering just two
out of three replicates and 2841 considering all identifications) at 1% FDR on the PSM level. Notably, on the level of
phosphorylation sites the overlap between subtilisin and trypsin was less than 20%. When considering the 1251 phosphorylation
sites exclusively found with subtilisin, 313 of those are theoretically not accessible via trypsin.
Proteomic Forum 2015
175
Conclusion: Under optimized conditions subtilisin can be used for reproducible digestion and furthermore provides access to
new phosphorylations sites that are concealed from LC-MS identification after tryptic digestion. Consequently, subtilisin may
serve as an alternative approach to considerably increase the coverage of the phosphoproteome.
P112
A MS/MS Based Workflow for the Detection of Unknown Stable Protein Adducts
1
2
2,3
1
4,3
1
M. Mueller , A. Acosta-Martin , P. Antinori Malaspina , O. Horlacher , A. Scherl , F. Lisacek , P. Lescuyer
1
Swiss Institute of Bioinformatics, Geneva, Switzerland
2
University of Geneva, Geneva, Switzerland
3
Swiss Centre of Applied Human Toxicology, Geneva, Switzerland
4
Geneva University Hospitals, Geneva, Switzerland
4
Introduction: Covalent interactions between reactive small molecules and proteins are important in many scientific fields. Often
little is known about which proteins a molecule binds to and even if the binding partners are known the exact form of the
chemical interaction and adduct formation remains elusive. The unbiased detection of protein adducts is of paramount
importance. In contrast to standard MS/MS searches open modification searches (OMS ) do not require prior knowledge of the
modifications present in the sample, but extract the modification mass shifts directly from the MS/MS data. OMS use spectral
alignment algorithms, which match the spectra of modified peptides in the query MS/MS data to their non-modified
counterparts in the database.
Methods: Here we present an experimental and computational workflow, which allows for the unbiased identification of
covalent or strongly bound small molecule adducts on proteins. Firstly, the small molecule under investigation is added to a
biological sample, where it reacts with its target proteins. Then after standard proteomics sample processing steps peptide
MS/MS spectra are measured by a LC-MS/MS system. In parallel a control sample without the addition of the small molecule is
processed in the same way.
We developed a novel bioinformatics workflow for comparative modification count analysis. First we perform an OMS, which
finds the detectable modified forms of the peptides. For both target and control samples we retrieve all detected modification
mass shifts and count in how many spectra they were found. After estimating the statistical variation in these counts we retrieve
those mass shifts that are more frequently detected in the target sample than in the control sample. All the spectral alignments
corresponding to these mass shifts are then subjected to further analysis to exclude potential artifacts that are often present in
OMS strategies.
Results: Firstly, we investigated the modifications induced by formalin fixation and paraffin embedding of human tissues. By
comparison to frozen tissue control samples, we were able to identify methylation (+14Da) on lysine as the major modification.
Secondly, we identified modifications that occur when albumin and hemoglobin are incubated with high concentrations of the
toxic drug Busulfan, an alkylating drug used in leukemia treatment. Very little was known about the reaction of this molecule
with blood proteins, but using our approach we were able to report several previously unknown adducts. These adducts formed
on on aspartic and glutamic acid as well as histidine residues. A primary adduct of 150Da, which further transformed into
secondary adducts, and a cross link adduct of 54Da were found. Thirdly, butyl methanesulfonate (BMS) a chemical for which no
reactions with proteins were reported showed a clear adduct signal at +56Da on aspartic and glutamic acid residues of the
abundant blood proteins.
P113
Identification of interaction networks and posttranslational modifications of mutant and wild type IDH1 in glioma cell lines
1
1
2
2
1
N. Overbeck , A. Stefanski , V. Scherbaum , C. B. Knobbe-Thomsen , K. Stühler
1
Heinrich-Heine-University, Molecular Proteomics Laboratory, Düsseldorf, Germany
2
Heinrich-Heine-University, Department of Neuropathology, Düsseldorf, Germany
Diffuse gliomas are the most frequentprimary brain tumors in humans and have an unpredictable rate of progression. It has
been shown that malignant transformation of glial cells is associated with a number of genetic events, including p53 mutations,
1p/19q codeletion/translocation, as well as amplification of epidermal growth factor receptor (EGFR)(Purow & Schiff, 2009).
Recently, mutations in the isocitrate dehydrogenase genes 1 and 2 (IDH1 and IDH2) have been described in 85-90% of grade II-III
astrocytomas, as well as in a minority of glioblastoma. The most frequent mutation in IDH1 is an amino acid exchange in codon
132 (R132H). Mutations in IDH1 appear to occur at an early stage during tumor development and it has also been shown that
patients with gliomas carrying IDH mutations have a better overall prognosis.
Proteomic Forum 2015
176
The goal of the project is the characterization of the interaction network as well as of the posttranslational modifications of wild
type and mutant IDH1 in glioma cells. Protocols for the enrichment of IDH1 both from non-modified glioblastoma cell lines as
well as from glioblastoma cell lines overexpressing wild-type or mutant IDH1 and a protocol for phosphopeptide enrichment
using titanium dioxide chromatography were established. For quantitative labelfree analysis an Orbitrap Elite mass spectrometer
equipped with a nanoUHPLC was used.
In summary we obtained the first detailed characterization of the interaction partners of as well as of the posttranslational
modifications of mutated und wild type IDH1. The combination of immunoprecipitation and mass spectrometry has proven to
be a powerful tool to dissect the intractome of IDH1 and PTM status.
Purow, B., & Schiff, D. (2009). Advances in the genetics of glioblastoma: are we reaching critical mass? Nat Rev Neurol, 5(8), 419426. doi: 10.1038/nrneurol.2009.96
P114
Global analysis of the impact of protein phosphorylation on mRNA-binding ability
1
1
1
K. Imami , M. Milek , M. Landthaler , M. Selbach
1
MDC, Berlin, Germany
1
Introduction: Protein phosphorylation has been shown to regulate the interaction between particular RNA-binding proteins
(RBPs) and RNA (Jammi et al., NAR 2001, Li et al., Nature 2003). In addition, recent phosphoproteomic studies suggest that
phosphorylation levels in RBPs are extensively modulated in response to stimuli such EGF (Olsen et al., Cell 2006) and DNA
damage (Matsuoka et al., Science 2007). Biochemically, phosphorylation is expected to result in a more negative charge of the
RBP which might decrease binding to the negatively charged RNA[MS1] . To understand how protein phosphorylation affects
RNA binding ability on a global scale, we have performed a proteome-wide measurement of protein phosphorylation and mRNAbinding ability using quantitative proteomics.
Methods: We sought to examine the correlation between phosphorylation levels in RBPs and their RNA-binding abilities. To do
this we performed a SILAC experiment where light- and heavy-labeled HEK cells were treated with a phosphatase inhibitor to
boost phosphorylation level or mock control. We measured >10,000 phosphorylation sites using TiO2 chromatography and in
parallel we quantified ~1,500 proteins bound to mRNA using UV crosslinking-oligo dT pulldown. The (phospho)peptides were
analyzed by nanoLC-MS/MS (Q-Exative plus) and data analysis was done with MaxQuant.
Results and Conclusions: The parallel measurement revealed that several RBPs showed increased phosphorylation and a
concomitant reduction in mRNA-binding. Thus, phosphorylation might indeed impair the interaction between RBPs and RNA. On
the basis of the correlation between phosphorylation levels and mRNA-binding abilities, we identified some RBPs, including CCRNOT4 complex and 40S ribosomal proteins, whose specific phosphorylation sites could regulate RNA binding. We are currently
investigating how such phosphorylation events affect mRNA binding as well as gene expression.
P115
Post-translational modification analysis of A. nidulans histone 3
1
2
2
1
C. Gruber , A. Gacek-Matthews , J. Strauss , F. Altmann
1
University of Natural Resources and Life Sciences, Chemistry, Vienna, Austria
2
University of Natural Resources and Life Sciences, Applied Genetics and Cell Biology, Tulln, Austria
Chromatin modification is widely recognized as an important mechanism in gene and genome regulation, conferring stability
and flexibility to expression patterns, cellular memory and positional controls. The core packaging proteins, histones, are subject
to a wide variety of regulatory covalent modifications, most commonly involving attachment of methyl-, acetyl-, phosphate- or
ubiquitin groups to particular amino acid residues of histone N- termini at the nucleosome exterior.
So far, a systematic biochemical analysis of histone post-translational modifications present in A. nidulans has not been carried
out and we thus performed mass spectrometry to determine the extent of methylation at histone H3 lysine 9 and lysine 36, the
putative targets of KdmA (a histone H3 demethylase with bipartite function). LC-MS/MS analysis of histone H3 obtained from
acidic nuclear extracts prepared from wild type, kdmAΔ, and clrDΔ cells (devoid of H3K9 methylation activity) grown for 17
hours (active growth, primary metabolism) revealed peptides (K27AAPSTGGVK36K37PHR) in which di- and tri-methylated K36
was dominant. MS/MS analysis of the clrDΔ control strain lacking the histone H3 lysine K9 methyltransferase revealed that
Proteomic Forum 2015
177
methylation on the H3 peptides containing K27 and K36 was identical to WT. As KdmA is predicted to also target lysine 9 at
histone H3, we investigated H3 peptides containing this crucial residue. In the K9STGGK14APR peptide tri- methylation of H3K9
was found (between 1.5% to 4% of the analyzed K9 peptides). This H3K9 methylation level of bulk histones is significantly below
the levels found in higher eukaryotes and around half the level found in N. crassa (Xiong et al., 2010).
For MS analysis relevant histone H3 protein bands were cut out and digested in gel. The proteins were S-alkylated with
iodoacetamide and digested with ArgC (Roche). The peptide mixture was analysed using a Dionex Ultimate 3000 system directly
linked to a Q-TOF MS (Bruker maXis 4G ETD) equipped with the standard ESI source in the positive ion, DDA mode.
P116
Identification of the Myofibrillar Z-Disc as a Nodal Point in Skeletal Myocyte Signaling by Large-Scale Phosphoproteomics
1
1
1
2
2
1
2
2
L. Reimann , H. Wiese , A. L. Fricke , Y. Leber , Z. Orfanos , G. Radiziwill , P. van der Ven , D. O. Fürst , B. Warscheid
1
University of Freiburg, Insitute of Biology II and BIOSS Centre for Biological Signalling Studies, Freiburg, Germany
2
University of Bonn, Institute for Cell Biology, Bonn, Germany
1
The myofibrillar Z-disc is an essential, structure-bearing component of muscle fibers which directly interacts with thin filaments,
intermediate filaments and titin. While originally considered to bea passive structural unit, its role in signal transduction has
become apparent in the last decade. To improve our understanding of signaling processes in and out of the Z-disc, we
delineated the phosphorylation landscape of cultured skeletal myotubes and studied the role of Z-disc proteins in hypertrophic
signaling processes.
To characterize kinase-mediated phosphorylation events in contracting C2C12 cells, we employed large-scale
phosphoproteomicswith SCX chromatography and TiO 2 affinity enrichment of phosphopeptides prior to LC/MS analysis.
Following up the results of our basal phosphoproteome study, we analyzed the PI(3)K/Akt dependent signaling pathway in a
SILAC approach using CID, MSA and HCD fragmentation on an Orbitrap Elite system.
We delineated the phosphorylation landscape of skeletal myocytes comprising more than 10.000 phosphorylation sites.
Remarkably, we show that the vast majority of Z-disc-specific and Z-disc-associated proteins are phosphoproteins accounting for
more than 60% of all localized sarcomeric phosphorylation sites. We identified filaminC (FLNC) and its interaction partners as a
group of highly phosphorylated proteins indicating a key role for these constituents in Z-disc signaling. In total, we found 12
distinct FLNC phosphorylation sites including several potential PKC and Akt motifs. Using radioactive and MS-based in vitro
kinase assays, we identified murine FLNC Ser-2625 as novel PKC-alpha substrate site. Through Western blot analysis and topdown mass spectrometry, we succeeded in demonstrating the importance of this phosphorylation site for the calpain-mediated
cleavage of FLNC in vitro. Moreover, pathway analysis showed that the majority of myofibrillar proteins involved in Z-disc
assembly and maintenance are targets downstream of PI(3)K/Akt. In our work, we continued to dissect the IGF-1-activated
PI(3)K/Akt pathway leading to skeletal myotubehyperthrophy using SILAC and different kinase signaling pathway inhibitors. First
results show that Akt-mediated phosphorylation of FLNC Ser-2234 has a significant effect on FLNC localization in fluorescence
recovery after photobleaching (FRAP) experiments.
To conclude, we characterized the myofibrillar Z-disc as a nodal point of skeletal myocyte signaling. We propose that FLNC acts
as a signaling hub whose localization, protein interactions are tightly controlled by cytosolic kinases.
P117
Advances in antibody-based proteomic analysis
1
1
1
1
1
1
M. P. Stokes , B. Textor , H. Gu , C. L. Farnsworth , J. M. Ren , K. A. Lee , J. C. Silva
1
Cell Signaling Technology Inc., Danvers, MA, United States
1
Immunoaffinity purification (IAP) of proteins and peptides is a powerful tool to enrich samples for subsequent LC-MS/MS
analysis. These antibody-based methods have long been used to probe for post-translationally modified (PTM) peptides that
normally would not be detected due to their lower relative abundance in samples. Antibodies are generated using degenerate
peptide libraries that recognize a particular PTM selectively but are agnostic to amino acids surrounding the PTM. A limited
number of amino acids in addition to the modified residue can also be fixed in the library to generate antibodies that recognize a
particular sequence motif, such as a consensus substrate sequence for protein kinases. The method has been successfully
applied to a number of PTM’s, including phosphorylation, acetylation, and ubiquitination.
Proteomic Forum 2015
178
Recently, a number of advances have allowed for significantly improved performance and increased Proteome coverage using
IAP methods. New antibodies have been developed to cover more PTMs, including methyl-arginine, methyl-lysine, and succinyllysine. Existing antibodies have also been improved, allowing identification of even more post-translationally modified peptides
in a single LC-MS/MS run. These include an updated phosphotyrosine antibody, as well as a new monoclonal acetyl-lysine
reagent (consisting of seven monoclonal antibodies) that provides the highest number of acetylated peptide identifications of
any available acetyl-lysine antibody. PTMScan Direct, an IAP LC-MS/MS method utilizing cocktails of site-specific antibodies to
phosphosites on critical signaling proteins, has been updated to include more reagents (thus covering more pathways), and to
cover more proteins/sites using a single affinity reagent. The IAP LC-MS/MS protocol has also been adapted for use on robotic
platforms such as the Agilent AssayMAP Bravo system, allowing automation of enrichment protocols for larger scale
experiments to probe cellular signaling.
P118
One for all? TiO2 and IMAC enrichment of phospho- and glycopeptides from serum
1
1
1
1
E. Klement , T. Raffai , Z. Darula , E. Hunyadi-Gulyas , K. F. Medzihradszky
1
Biological Research Centre, Laboratory of Proteomics, Szeged, Hungary
1
Introduction: Glycosylation and phosphorylation are the two most prominent and widely studied posttranslational modifications
(PTMs) of proteins. Their importance is justified by their diverse roles in the different biological processes. Glycosylation is
predominantly found on secreted and membrane proteins and promotes proper folding of the proteins, increases their stability
and facilitates cell-cell adhesion. Phosphorylation is the main regulatory modification on cellular proteins, however, Golgiresiding kinases have been identified recently and were implicated in some diseases. Thus, the role of extracellular
phosphorylation is also an exciting question.
Objectives: Phosphopeptides are usually enriched using metal affinity chromatography, i.e. TiO 2 or IMAC. It was demonstrated
that TiO2 also binds sialylated glycopeptides. We investigated whether IMAC could also be used for glycopeptide isolation and if
the selectivity could be shifted towards either of the two modifications.
Materials & Methods: Tryptic digests of human and fetal calf serum (SigmaAldrich, Germany) were applied to FeNTA-agarose
(Qiagen, Germany) beads or TiO2 (GL Sciences, Japan) under different binding conditions to isolate different subsets of the
modified peptides. Combined HCD/ETD fragmentation on an LTQ-Orbitrap Elite (Thermo, Germany) mass spectrometer was
used for the analysis of the enriched samples to identify the modified peptides and determine the site of modification.
Results: We found that glycopeptide capture on TiO2 or FeNTA occurs only at high acetonitrile content most likely via
hydrophilic interaction. Lowering the acetonitrile content to 50-60 % results in the selective isolation of phosphopeptides.
Conclusion: Changing the enrichment conditions the phosphopeptides can be selectively enriched or together with the
glycopeptides. When captured together, sequential elution of the glyco- and phosphopeptides from the beads enables parallel
analysis of these two important posttranslational modifications.
Acknowledgements: This work was supported by the Hungarian Scientific Research Fund #105611. E. Klement was also
supported by the Janos Bolyai Fellowship of the Hungarian Academy of Sciences.
P119
N- and O-glycopetide analysis from HCD data
1
1
2,3
1,4
Z. Darula , E. Hunyadi-Gulyas , M. Bern , K. F. Medzihradszky
1
HAS Biological Research Centre, Laboratory of Proteomics Research, Szeged, United States
2
Protein Metrics Inc., San Carlos, CA, United States
3
Palo Alto Research Center, Palo Alto, CA, United States
4
University of California San Francisco, Department of Pharmaceutical Chemistry, San Francisco, CA, United States
Introduction: There is a growing understanding of the biological significance of extracellular glycosylation. Thus, deciphering
site-specific glycosylation in various biological systems is of major importance.
The term glycosylation covers a wide variety of modifications. Different amino acid side-chains may be modified with single
sugar units, small oligosaccharides or huge linear or branched structures. Consensus motifs were identified for N-glycosylation,
Proteomic Forum 2015
179
and the N-glycans have a common core structure. O-glycosylation is a lot more diverse, modifying Ser, Thr and Tyr featuring a
series of different core structures.
Mass spectrometric characterization of N- and O-glycopeptides represent a similar challenge. Glycosidic bonds are weaker than
peptide bonds, collisional activation usually reveals a lot more about the oligosaccharide structures than the peptides. Thus,
electron-transfer dissociation is recommended for glycopeptide analysis. However, this activation is efficient for high chargedensity precursors while glycan addition usually results in significant mass increase without proportionally increasing the charge
state. Thus, frequently one has to rely on data acquired using collisional activation.
Objectives: In order to supplement ETD-based glycopeptide identification we tested the performance of Byonic (Protein
Metrics) - a search engine with special support for glycopeptide analysis - for the identification of glycopeptides from HCD data.
Materials & Methods: Glycopeptides were enriched from the tryptic digest of human serum and mouse synaptosome using
lectin affinity chromatography applying Jacalin, wheat germ agglutinin and Concanavalin A. Samples were analyzed by
LC/MS/MS applying HCD product ion dependent ETD acquisition. Database searches were performed using Byonic.
Results and Conclusion: Unique features of Byonic, including glycopeptide fragmentation based scoring algorithm, assignment
of carbohydrate oxonium ions, neutral losses and Y ions along with user-settable or predefined glycan databases enabled the
identification of numerous glycopeptides from HCD data. Examples from complex N- and O-linked glycopeptide-containing
mixtures, highlighting the advantages and limitations of this approach will be presented.
Acknowledgements: The authors thank financial support to NIGMS 8P41GM103481, Howard Hughes Medical Institute (to the
Bio-Organic Biomedical Mass Spectrometry Resource at UCSF, Director: A.L. Burlingame), OTKA 105611 (ZD), GOP-1.1.1-112012-0452 (KFM) and BAROSS-DA07-DA-ESZK-07-2008-0036 (ZD). Z. Darula was supported by the Janos Bolyai Fellowship of the
HAS.
P120
Wound fluid degradomics: system-level insight into protease dynamics in cutaneous wound healing
1
1
1
1
1
1
2
3
U. auf dem Keller , F. Sabino , O. Hermes , F. Egli , T. Kockmann , P. Schlage , P. Croizat , J. Kizhakkedathu , H. Smola
1
ETH Zurich, Zurich, Switzerland
2
Paul Hartmann AG, Heidenheim, Germany
3
University of British Columbia, Vancouver, Canada
2
Introduction: Proteases are major contributors to all phases of cutaneous wound healing by facilitating recruitment of immune
cells to the wound site, migration of keratinocytes and fibroblasts, angiogenesis and scar tissue remodeling. However, protease
function in the healing skin wound has been only partially unraveled and many cleavages remain to be elucidated.
Objectives: In this study, we aimed at recording proteolytic signatures at multiple time points after wounding to better
understand protease activity and dynamics in cutaneous wound healing and to establish novel protease-substrate relations in
vivo.
Methods: We applied iTRAQ-Terminal Amine Isotopic Labeling of Substrates (iTRAQ-TAILS) to the analysis of wound fluids from
a clinically relevant pig wound-healing model. Using five animals per time point wound fluids were extracted from polyurethane
foams that had been placed into the wound site at 1, 3 or 7 days after wounding. Selected cleavage events were validated by
selected reaction monitoring (SRM) and further assessed in cell-based assays.
Results: By iTRAQ-TAILS we identified more than 650 proteins and almost 1300 N-terminal peptides in ~450 proteins.
Abundance clustering of these proteins allowed time-resolved discrimination of major phases of the healing process based on
distinctive abundances of markers of inflammation, granulation tissue formation and re-epithelialization. Pathway mapping of
predicted cleavage sites showed temporal interdependent activation of blood coagulation, fibrinolysis and complement
cascades as well as processing of complement C3 at distinct time points after wounding and by different proteases.
Using bioinformatics we assigned neo-N termini with help of known cleavage specificities to responsible candidate proteases.
This suggested a prevalence of members of the trypsin-like S1 family in wound fluids, which includes components of the
coagulation and complement systems. The second most abundant specificity cluster corresponded to matrix metalloproteinase
(MMP) like cleavages, while a third prominent group of cleavages was related to caspases. Of these, we validated a caspase-like
Proteomic Forum 2015
180
processing event in the integrin activator protein kindlin-3 by SRM and identified caspase-3 as the responsible protease in
apoptotic monocytes.
Conclusions: Exploiting the power of iTRAQ-TAILS we quantitatively assessed the wound proteome and the activity of distinct
protease groups along the healing process. TAILS mapped pivotal proteolytic pathways with unprecedented depth in vivo and
established novel protease-substrate relations with important implications for this highly complex tissue response.
Reference: Sabino F*, Hermes O* et al., and auf dem Keller U (2014). In Vivo Assessment of Protease Dynamics in Cutaneous
Wound Healing by Degradomics Analysis of Porcine Wound Exudates. Mol Cell Proteomics, mcp.M114.043414
P121
Quantitative MS-based analysis of histone modification patterns as markers for response to HDAC inhibitors in breast cancer
1
2
2
2
2,3
R. Noberini , C. Miccolo , G. Pruneri , S. Chiocca , S. Minucci , T. Bonaldi
1
Istituto Italiano di Tecnologia, Milan, Italy
2
European Institute of Oncology, Milan, Italy
3
University of Milan, Milan, Italy
2
Question: Abnormalities in histone post-translational modification (hPTM) patterns are frequently implicated in the
development of cancers and could represent biomarkers for drug response and patient stratification. Histone deacetylases
(HDACs) are a class of enzyme that deacetylate histone lysine residues and have emerged as attractive targets for the therapy of
various diseases, including breast cancer. Starting from our observation that primary breast cancer cells in culture display
different sensitivities to HDAC inhibitors we sought to identify the epigenetic biomarkers that determine cellular responses to
inhibition of HDAC in breast cancer by using mass spectrometry approaches.
Methods: We performed proliferation assays to identify groups of breast cancer cell lines that are either sensitive or resistant to
HDAC inhibitors and employed a novel analytical platform that combines ultra-high pressure liquid chromatography with high
resolution mass spectrometry analysis on a Q Exactive instrument to carry out a comprehensive analysis of their hPTMs
patterns. This approach was combined with stable isotope labeling with amino acids in cell culture (SILAC), using a super-SILAC
set up where a mix of heavy-labelled breast cancer cells served as spike-in reference for comparative analysis with unlabelled
cells.
Results: We profiled 24 distinct modifications at 14 different sites on histone H3 and H4, covering well-characterized histone
marks, in addition to identifying novel or poorly characterized modifications. Comparison of breast cancer cells that are sensitive
or resistant to HDAC inhibitors in the presence and in the absence of the compounds revealed hPTM patterns that could be
associated with the cellular response to the drugs and pinpointed the modifications affected by HDAC inhibitors with different
specificities.
Conclusions: The robust method that we have developed for the identification of hPTM allowed the profiling of modification
patterns linked to drug sensitivity in breast cancer cells. Currently, we are applying this strategy to the analysis of clinical
samples, such as fresh-frozen and formalin-fixed-paraffin embedded tissues, for which we have developed ad hoc histone
extraction protocols.
Proteomics for Human Health III (P170–P190)
P170
Proteome Analysis and Comparative Study of Normal Endometrial Stem cells, Endometriosis Stem Cells
1
2
3
S. heidari-keshel , M. Rezaei-tavirani , M. Ebrahimi
1
SPU, farabi eye hospital, Tehran, Iran, Islamic Republic of
2
beheshti university, proteomics research center, Tehran, Iran, Islamic Republic of
3
tehran university of medical sciences, Tissue engineering, tehran, Iran, Islamic Republic of
Purpose: The aim of this study is Isolation and characterization of sub population stem cell in vitro condition from healthy and
endometriosis donors and proteomics evaluation for identification of putative markers. Method: Endometriosis endometrial
stem cells (EESCs), normal endometrial stem cell (ESCs), endometrial lesions stem cell (ELSCs) and bone marrow Mesenchymal
stem cell (MSCs) were isolated. Flowcytometry and Real time-PCR were utilized to detect the cell surface marker and expression
Proteomic Forum 2015
181
pattern of stemness genes. Proliferation of all stem cells was observed by MTT assay. The differentiation potential was
evaluated by alizarin red, oil red O and RT-PCR method. The karyotyping was performed on EESCs and ELSCs at passage 20.
Furthermore, we systematically tested the immunophenotype of cultured MSCs, EESCs, ESCs and ELSCs by flowcytometry
analysis using a wide variety of markers. Direct comparison of this phenotype to the one derived from cultured BM-MSCs
demonstrated that cultured MSCs from both sources exhibit similar expression patterns. Using the two-dimensional gel
electrophoresis and matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) approach, we
have generated for the first time the protein map of cultured endometrial stem cell by identifying 6 proteins, and we compared
it directly to that of cultured BM-MSCs. Results: Stem cell characteristics of single cells isolated from normal endometrial and
endometriosis tissues from women ages 28 ± 8 years (n = 18) were assessed. Proteomics finding show that high expression level
for GAPDH and Glutathione-s transferase in Normal endometrial stem cell and high level expression for HSP-27, KRT-1, Albumin,
Calumenin in endometriosis endometrial stem cell. These findings suggest that spatially all of protein have role in migration,
stemness, cancer stem cell. In the fowling of work gene expression evaluation for migration, stemness and CSC was performed.
Expression of CAPN7, MMP2, MMP9 genes significantly higher than in ELSC group and CXCR4, FGFR3, SDF1 gens have a high
expression level in EESC sample. Meanwhile, compared with EESC-CM, ELSC-CM had greater effects for promoting MSCs and
ESCs migration and enhancing proliferation ability of cells (p < 0.01). Although these cells (EESCs, ESCs, ELSCs and MSCs) have
background expression of stemness-related genes, the unique patterns was detected. Spindle-like morphology, normal
karyotype, Adipogenic and osteogenic potential, expression of Oct4, Nanog, Klf4 and ERAS genes and CD44, CD105, CD90, CD73
and CD146 specific surface markers, in EESCs and ELSCs was observed. Furthermore, rate of proliferation and expression of
these genes like SALL4, DPPA2, TCL1 and Sox2 were significantly higher than ESCs and MSCs. Conclusion: This study illustrates
the presence of CD146 positive human Endometrial Somatic stem cell in the human endometrium and reveals their potential
use in regenerative medicine and taming stem cell heterogeneity for cell therapy and Tissue engineering such as in the
treatment of related infertility diseases.
Figure 1
Figure 2
Proteomic Forum 2015
182
P171
Moroccan cases with Alzheimer disease: Genetic and proteomic aspects
1
1
N. El Kadmiri , N. El Kadmiri
1
HASSAN II University, Faculty of Medicine and Pharmacy, Laboratory of Genetics and Molecular Pathology, Casablanca,
Morocco
Introduction: In Morocco, Alzheimer disease (AD) has emerged as a serious public concern with the number of people suffering
from AD expected to increase as the elderly population continues to grow. Genes predisposing individuals to AD and predicting
disease incidence remain elusive and prevent health care professionals from identifying AD in its early stages, with the goal of
slowing down the progression.
The purpose of the present study is to:
- Evaluate the clinical and para clinical aspects of AD in familial and sporadic cases;
- Evaluate the genetic contribution of mutations in in the presenilin-1 (PS1) and presenilin-2 (PS2) genes to familial early-onset
AD cases (FAD) and sporadic late-onset AD cases;
- Elucidate the critical role of GAPDH in the blood of FAD cases carrying presenilin mutations and its interaction with Aβ amyloid.
Methods: Seventeen sporadic late-onset AD cases and eight familial early-onset AD cases were seen at the memory clinic of the
University of Casablanca Neurology Department. These patients underwent standard somatic neurological examination,
cognitive function assessment, brain imaging and laboratory tests. Direct sequencing of PS1 and PS2 genes was performed on
genomic DNA of AD patients. Activity assay of GAPDH, Western blot analysis, Dot Blotting and Electron microscopy (EM) were
performed on blood of FAD cases carrying presenilin mutations.
Results: All patients had hippocampal atrophy. A higher level of atrophy reflects a decrease in neuropsychological performance.
we identified 1 novel frameshift mutation in the PS1 gene and 2 novel frameshift mutations in the PS2 gene. The activity of
GAPDH in FAD cases, was significantly decreased as compared to sporadic cases and healthy controls. The expression of GAPDH
in blood samples from Mutant tau transgenic mice and FAD cases was decreased as compared to sporadic cases and healthy
controls. The Dot blotting examination showed an incresead Aβ accumulation in the blood samples from FAD cases.EM
examination showed an increase in amyloid fibrils both in brain in both the blood and brain samples.
Conlusion: Our mutational analysis report a correlation between clinical symptoms and genetic factors in our cases and
suggesting that these mutations increase the risk for developing AD. Our proteomic analysis, report the involvement of GAPDH
in AD that may influence the pathogenesis of neurodegenerative disease.
P172
Analysis of Protein-Protein Interactions in Living Cell using NanoBRET™ Technology
1
2
1
1
3
1
1
1
1
1
T. Machleidt , C. Woodroofe , J. Mendez , N. Murphy , C. End , K. Zimmermann , P. Otto , M. Schwinn , M. Robers , D. Daniels ,
2
1
T. Kirkland , K. Wood
1
Promega Corporation, Madison, United States
2
Promega Biosciences LLC, San Luis Obispo, United States
3
Promega, Mannheim, Germany
Protein-protein interactions (PPI) play an intricate part in the network of signaling pathways that governs every aspect of cellular
physiology, which could make PPI an appealing target for the development of novel therapies. The dynamic and often transient
nature of protein-protein interactions poses a considerable challenge for the development of physiologically relevant, HTS
compatible assays. Resonance energy transfer technologies especially bioluminescence resonance energy transfer (BRET) has
been extensively used for the interrogation of PPI. BRET relies on the proximity dependent, non-radiative transfer of energy
between a donor and an acceptor. The most commonly used version of BRET is BRET1 which utilizes a combination of Renilla
luciferase and YFP as donor/acceptor pair. Despite its widespread use, BRET1 suffers from a number of intrinsic limitations due
to the relatively weak luminescence of its donor and narrow spectral separation between donor and acceptor which result in
relatively poor sensitivity and limited dynamic range. These limitations often require the massive overexpression of both BRET
partners in a permissive cellular background, which can potentially lead to experimental artifacts. Here we present NanoBRET, a
novel BRET platform which addresses the current limitations of BRET technology. NanoBRET employs NanoLuc® luciferase, a
small (19kDa) and extremely bright luciferase in combination with HaloTag® (HT) labeling technology. The brightness of
NanoLuc® luciferase allows the use of a red fluorescent HaloTag® ligand as acceptor resulting in a donor/acceptor pair with a
spectral separation greater than 150 nm. The combination of brightness and spectral separation result in substantially increased
Proteomic Forum 2015
183
detection sensitivity and dynamic range compared to BRET1. Using a number of well-established in vitro and cell based models
we show that these improvements considerably expand the application of BRET, including the analysis of PPI at low expression
levels, imaging of PPI and the use of difficult to transfect cell backgrounds and high-density assay formats. The potential of the
NanoBRET™ platform is further demonstrated by the development of the first HTS compatible cell based assay for the analysis of
Histone/Bromodomain interactions.
P173
Cigarette smoke treatment induces alterations in protein networks of AMD risk genes
1
1
1
2
2
1
J. Merl-Pham , F. Gruhn , N. Senninger , G. John-Schuster , A. Yildirim , S. M. Hauck
1
Helmholtz Zentrum Muenchen GmbH, Research Unit Protein Science, Neuherberg, Germany
2
Helmholtz Zentrum Muenchen GmbH, Comprehensive Pneumology Center, Neuherberg, Germany
Introduction: Age-related macular degeneration (AMD) is a medical condition usually affecting older adults and resulting in a
loss of vision in the macula, the center of the visual field. For the majority of patients affected with the dry form of the disease,
the hallmark pathology is atrophy of the retinal pigment epithelium (RPE), resulting in the detachment of the retina and loss of
photoreceptors. Apart from age, cigarette smoke is the main risk factor for AMD and increases the risk for developing the
disease by two to three fold.
Objectives: We aimed to increase understanding of the tobacco smoke-induced molecular mechanisms underlying RPE
destruction in the context of AMD. Therefore, we performed proteomic profiling of cellular reactions to cigarette smoke
treatment, firstly in vitro in subfractionated primary RPE cells and secondly in vivo in isolated RPE cells from cigarette smoke
treated mice.
Materials & Methods: Primary RPE cells were isolated from porcine eyes and cultivated for a maximum of 2 weeks. Those cells
were treated with different concentrations of cigarette smoke extract (CSE) for 24 h and then fractionated in secretome, cell
surface, nucleus and cytoplasm. These fractions were analyzed separately by label-free quantitative LC-MS/MS on an
OrbitrapXL. In order to translate in vitro findings to the in vivo situation, we next exposed mice (2 and 12 month old) to cigarette
smoke over 4 months. The eyes of the smoked mice and of the corresponding air-treated controls were dissected, and whole
cell lysates of the RPE cells were analyzed again by LC-MS/MS.
Results: In primary porcine RPE cells, 2045 proteins were identified with high confidence (> 2 peptides) and supplementation of
cigarette smoke extract (CSE) resulted in differential expression of 40 proteins for 1% CSE, 125 proteins for 5% CSE and 239
proteins for 10% CSE. Cluster analyses of protein expression changes revealed perturbation in protein networks overrepresented
with AMD risk genes even at low dosages. Subsequently, in the RPE of cigarette smoke treated mice, we identified altogether
2425 proteins in a total of 16 samples (four replicates per condition: young +air, young +smoke, old +air, old +smoke) and
compared their respective abundances between the four sample groups. We found 234 proteins significantly altered in aged
mice in comparison to younger ones. Furthermore, in young mice, 54 proteins were significantly differentially abundant after
cigarette smoke treatment; in old mice, even 274 proteins. In line with the in vitro results, we found several already described
AMD risk genes among the proteins significantly altered with both age and/or cigarette smoke treatment, like extracellular
matrix proteins and complement components.
Conclusion: We conclude, that cigarette smoke treated mice could serve as a potential mouse model resembling AMD-like
phenotypes in the affected eyes.
P174
Key event in human allergic contact dermatitis - adduct formation of the common xenobiotic human contact sensitizer
methylisothiazolinone
1,2
1,3
1,3
1,4
H. Regine , S. Nicolai , A. Luch , H.- J. Thierse
1
German Federal Institute for Risk Assessment, Chemicals and Product Safety, Berlin, Germany
2
Humboldt Universität, Berlin, Germany
3
Freie Universität Berlin, Berlin, Germany
4
Ruprecht-Karls Universität Heidelberg, Heidelberg, Germany
Question and Methods: Allergies are worldwide increasing, and molecular and immunological checkpoints of human allergic
contact dermatitis (ACD) comprise interactions of small reactive chemicals with extracellular or cellular skin self proteins or
Proteomic Forum 2015
184
peptides, isochronic innate inflammatory cell responses, antigen processing and presentation, and the chemical-specific
(re)activation of human T cells. In this mass spectrometric study we analyzed first key event in ACD, posttranslational adduct
formation of the common xenobiotic human contact sensitizer methylisothiazolinone (MI) with both, an all amino acidscontaining model peptide (+/- Cys) and a human model protein.
Results: Mass spectrometric (MS; MALDI, Bruker UltrafleXtreme) detection of MI-haptenated peptides revealed specific thioldependent peptide adduct formation at all pH values tested and physiologically found in human skin layers. Surprisingly, best
modification rates were obtained at more acidic pH conditions, also present in human epidermis.
Conclusion: Molecular and mechanistic understanding of this first key event in human ACD supports the development of novel
predictive MS-based in vitro assays, thereby potentially also replacing animal experiments in chemical regulation and risk
assessment
P175
Mass Spectrometry Assays of Plasma Biomarkers to Predict Radiographic Progression of Knee Osteoarthritis
1,2
1
3
3
1,2
1,2*
Susan Y. Ritter , Jamie Collins , Mary Lopez , Bryan Krastins , Elena Losina , Antonios O. Aliprantis
1
2
3
Department of Medicine, Brigham and Women’s Hospital, United States; Harvard Medical School, United States Thermo
Fisher Scientific - BRIMS Center, United States.
1,2
2
M. Oppermann , M. Lopez
1
Thermo Fisher Scientific, Germany, Sweden
2
BRIMS, Boston, United States
Background: Biomarkers to identify osteoarthritis (OA) patients at risk for disease progression are needed. As part of a
proteomic analysis of knee synovial fluid from normal and OA patients, differentially expressed proteins were identified that
could represent potential biomarkers for this disease. The aim of this study was to use mass spectrometry assays to identify
representative peptides from several of these proteins in synovial fluid and peripheral blood, and to assess their plasma levels as
biomarkers of OA progression.
Methods: Multiplexed high throughput selected reaction monitoring (SRM) assays were developed to measure tryptic peptides
representative of 23 proteins in matched serum and synovial fluid samples from late OA subjects at the time of joint
replacement. Subsequently plasma samples from the baseline visit of 173 subjects in an observational OA cohort were tested by
SRM for peptides from 9 of these proteins: afamin, clusterin, cartilage oligomeric matrix protein, hepatocyte growth factor,
kallistatin, insulin-like growth factor binding protein, acid labile subunit, lubricin, lumican, and pigment epithelium-derived
factor. Linear regression was used to determine the association between the peptide biomarker level at baseline and maximal
joint space narrowing (JSN) from baseline to 30 months, adjusting for age and sex.
Results: In the matched cohort, 17 proteins could be identified in synovial fluid and 15 proteins were detected in serum. For the
progression cohort, the average age was 62 and average JSN over 30 months was 0.68 mm. A high correlation between different
peptides from individual proteins was observed, indicating our assays correctly measured their target proteins. Peptides
representative of clusterin, lumican and lubricin showed statistically significant associations with joint space narrowing after
2
adjustment for age and sex. Partial R values showed the clusterin FMETVAEK and lubricin LVEVNPK peptide biomarkers explains
about 2-3% of the variability of JSN, similar to that explained by age. A biomarker score combining normalized data for both
2
lubricin and clusterin peptides increased the model R to 0.079.
Conclusions: Our results suggest that when combined, levels of peptides representative of clusterin and lubricin in plasma are as
predictive of OA progression as age. Replication of these findings in other prospective OA cohorts is planned.
Proteomic Forum 2015
185
P176
Host cell protein synthesis response to infection with different influenza A virus strains
1
1
2
2
2
K. Eichelbaum , B. Bogdanow , A. Sadewasser , K. Paki , T. Wolff , M. Selbach
1
MDC, Berlin, Germany
2
RKI, Berlin, Germany
1
Influenza A virus depends on the host cellular machinery for successful replication. The quantification of protein synthesis
differences after viral infection with different virus strains could therefore help to understanding their species-specificity and
pathogenicity.
In this project, we quantitatively compared newly synthesized proteins in human lung epithelial cells (A549) after the infection
with a permissive and a non-permissive influenza A strain at different time points. For the enrichment of newly synthesized
proteins, the non-natural amino acid azidohomoalanine (AHA), an azido-analogue of methionine, was added to the methioninedepleted growth medium of cells, and therefore incorporated in all newly synthesized proteins. Using a copper-catalyzed
cycloaddition the azidohomoalanine labeled proteins were covalently coupled to an alkyne-resin and released by tryptic
digestion. We selectively captured the newly synthesized proteins from stimulated and control SILAC labeled cells and analyzed
the proteins by quantitative LC-MSMS. Up to 5500 newly synthesized proteins were quantified in a time-course spanning 16
hours of viral infection. Viral protein synthesis was strongly induced in the investigated time frame, while a global reduction in
host cellular protein synthesis was observed. Differential protein synthesis after infection with two virus strains was determined
for several proteins (e.g. involved in viral defense) and points to steps where replication of the non-permissive strain is blocked.
Our data set demonstrates the immense impact of the influenza A virus on host cellular protein synthesis and identifies new
candidates potentially contributing to the difference in permissiveness of the two investigated viral strains.
P177
Proteomic alterations in human renal epithelial cells exposed to nephrotoxins
1
2
1
2
L. Kollipara , A. Limonciel , A. Sickmann , P. Jennings
1
Leibniz – Institut für Analytische Wissenschaften - ISAS - e.V., Dortmund, Germany
2
Medical University of Innsbruck, Division of Physiology, Department of Physiology and Medical Physics, Innsbruck, Austria
Proteomic analysis holds great promise for biomarker discovery. Protein detection, as opposed to transcriptomic approaches, is
more widely applicable for developing downstream assays. Also such assays can be readily applied to quantification in biological
fluids, such as urine, and thus may transpire to be useful biomarkers in clinical settings.
Differentiated transporting monolayers of the human renal proximal tubule cell line RPTEC/TERT1 were exposed to a single
bolus of a sub-cytotoxic concentration of the immunosuppressant cyclosporine A (CsA, 15 µM), the cosmetic ingredient and
oxidant potassium bromate (KBrO3, 0.8 mM) and the mycotoxin ochratoxin A (OTA, 0.13 µM) for 24h. Cell lysates were prepared
for proteomic studies (4 replicates) at the end of the 24h exposure. Samples were processed with FASP [1] protocol and the
resulting tryptic peptides were analyzed using nano-LC coupled to the Orbitrap Fusion mass spectrometer. Label free
quantitation of the acquired MS data was performed by the Progenesis LC-MS software.
Nephrotoxin exposure caused a significant (P <0.01) alteration in the abundance of hundreds of proteins. CsA resulted in 332
alterations, KBrO3 192 and OTA 53. The largest similarities were between CsA and KBrO3, both which have been previously
show to cause a strong Nrf2 oxidative response in proximal tubular cells. Interestingly, while OTA had a very large impact on the
transcriptome after 24h, it was the compound that induced the least alterations in the proteome. Ingenuity pathway analysis of
the proteomic changes indeed identified the Nrf2 response as a major effect of CsA and KBrO3. Additionally, CsA impacted on
the unfolded protein response and KBrO3 on cell cycle regulation. Of the 26 proteins that were altered by CsA and KBrO3 in the
same direction had the same direction the Nrf2 regulated hemeoxygenase-1 (HO-1) was the highest increased protein. Urinary
HO-1 has been previously associated with acute and chronic kidney disease. The data show that the approach used is an
excellent strategy to uncover mechanistic biomarker of compound induced renal injury.
Abbreviations: FASP: filter-aided sample preparation; LC-MS: liquid chromatography-mass spectrometry.
[1] Wisniewski, J. R., Zougman, A., Nagaraj, N., Mann, M., Universal sample preparation method for proteome analysis. Nat
Meth 2009, 6, 359-362.
Proteomic Forum 2015
186
P178
Identification of widespread alterations in RNA metabolism in sporadic inclusion-body myositis using SILAC-based
quantitative proteomics
1,2
2
2
1
2
T. Opialla , M. Berger , A. Marg , S. Kempa , S. Spuler
1
Berlin Institute for Medical Systems Biology/Max Delbrück Center for Molecular Medicine, Integrative Metabolomics and
Proteomics Platform, Berlin, Germany
2
Experimental and Clinical Research Center, a joint cooperation between Charité Medical Faculty and Max Delbrück Center for
Molecular Medicine, Muscle Research Unit, Berlin, Germany
Introduction: sIBM is the most common acquired muscle disease of the elderly. Symptoms are progressive muscle weakness and
atrophy often accompanied by dysphagia. Loss of ambulation and artificial nutrition are frequent. The cause of sIBM is unknown.
The pathogenesis is complex and includes inflammatory-immunological and degenerative aspects. Since there is currently no
therapy available, the progressive course of sIBM cannot be modified.
We prepared a human SILAC reference for muscle and provide an insight into sIBM-affected muscle that may show new options
for a beneficial intervention.
Objectives: To compare the muscle proteome in inclusion body myositis (sIBM) and healthy controls and to characterize
proteome changes over time in longitudinal studies.
Patients & Methods: For accurate quantification in our proteome analysis, we created a human muscle-specific SILAC-reference
from myotubes generated from primary human myoblasts. Frozen muscle specimens from sIBM (n=9) and age matched healthy
control (n=8) were subjected to urea extraction with Lys-C and trypsin-digestion followed by high-performance liquid
chromatography coupled to tandem mass spectrometry. From two patients muscle biopsy specimens were analysed that were
obtained at two different time points (time point A and B) seven years apart allowing insight into longitudinal changes. Data
were analyzed with MaxQuant, Perseus and R.
Results: Shotgun proteomics using human muscle-specific SILAC-reference resulted in identification of 2500 proteins, of which
2000 were quantified. Protein expression differed fundamentally between sIBM muscle and healthy muscle. Many pathways are
affected, especially those which are associated with RNA processing, RNA metabolism, splicing and translation. The longitudinal
study revealed that the proteomes were very similar between time point A and B, although, histologically, at time point A only
features of polymyositis without increase in connective tissue, rimmed vacuoles or congophilic deposits were present. This
applies in particular to the disruption in RNA metabolism, translation and mitochondrial abnormalities.
Conclusion: sIBM implicates severe dysregulation of about 500 muscle proteins in particular widespread alterations of RNA
metabolism. The longitudinal study revealed that proteome changes precede histopathology by years.
The proteome analyses of polymyositis muscle could be used in the future to reach a very early diagnosis of sIBM, thus avoiding
unnecessary potentially harmful immunosuppressive therapies.
P179
Measuring Wnt-pathway activity in cancer models
1
2
1
2
1
O. Poetz , C. Esdar , C. Sommersdorf , R. Schneider , T. O. Joos
1
NMI Naturwissenschaftliches und Medizinisches Institut an der Universität Tübingen , Protein Analytics, Reutlingen, Germany
2
Merck Serono, Darmstadt, Germany
The Wnt-signaling pathway is often found to be hyperactive in colorectal, hepatocellular, and other cancers. Therefore, its
modulation is in the focus of basic and pharmaceutical research. The activity of the pathway is reflected by concentration
changes of beta-catenin. The homeostasis of beta catenin is regulated by the ubiquitin proteasome system. Mutations encoding
either proteins of the degradation machinery or the N-terminal region of beta-catenin lead to accumulation of beta-catenin in
the cell. Accumulated cytosolic beta-catenin is translocated into the nucleus and its function as transcriptional co-activator is
elevated. Since the protein plays also a role as adaptor molecule in cell adhesion processes by bridging cadherins with cell
skeleton proteins the analysis of the pathway activity on protein level is challenging.
Proteomic Forum 2015
187
We have combined three biochemical methods - sandwich immunoassay, co-immunoprecipitation and protein-protein
interaction assay - in one suspension bead assay panel (10 plex) for the relative quantification of total beta-catenin, the extent
of phosphorylation at multiple sites and the ratio of complexed and free beta-catenin. We investigated the dynamic effects of
chemical inhibitors on proteins abundance, on function and on posttranslational modifications. A detailed quantitative and timeresolved analysis of the pathway activity is possible and gives insights into the mechanisms of natural Wnt-ligands and pathway
inhibitors.
The assay was used to test the impact of Wnt-signalling inhibitors on pathway activity in an in vivo PK/PD study. Human colon
tumor cell lines were subcutaneously injected into athymic mice and mice were treated with Wnt pathway inhibitors.
Modulation of the Wnt pathway activity in the tumor xenografts was tested by using the beta-catenin assay panel with ex vivo
lysates.
P180
Comprehensive analysis of acute- and late- cellular response to photon and carbon irradiation using quantitative proteomics
1,2
1,2
2
2
1
1
2
1
M. Winter , T. Rackwitz , P. Seidel , C. Zhou , R. Mayer , U. Warnken , A. Abdollahi , M. Schnölzer
1
German Cancer Research Center (DKFZ), Functional Proteome Analysis, Heidelberg, Germany
2
German Cancer Research Center (DKFZ), Molecular & Translational Radiation Oncology, Heidelberg, Germany
Introduction: Pulmonary microvascular endothelial cells are critically involved in development of radiation induced acute
inflammation and late fibrosis often limiting the delivery of sufficient dose to optimally treat lung cancer. Moreover, tumorendothelium communication has emerged as a key target to augment radiotherapy by resensitizing endothelial specific survival
signals.
Objectives: We aimed to employ proteome and phosphoproteome analysis to uncover the molecular mechanisms governing
tumor-vessel communication and radiation induced normal tissue toxicity. Conventional photon radiotherapy was compared to
cutting-edge carbon ion radiotherapy as it is employed at the Heidelberg Ion-Beam Therapy Center (HIT). To investigate timedependent cellular response we included two different timepoints (acute and late) in our study.
Methods: Human pulmonary microvascular endothelial cells were irradiated with 2 Gy carbon and 6 Gy photon (radiobiological
equivalent doses). Cells were harvested two hours and six days after irradiation and proteins were extracted. Mass spectrometry
was performed on an LTQ-Orbitrap XL and label-free quantification (LFQ) was employed for relative quantification of the protein
expression and phosphorylation status. For the phosphoproteome analysis a combination of IMAC and TiO 2 phosphopeptide
enrichment steps was applied.
Results: At the early timepoint a strong induction of proteins involved in apoptosis and DNA-damage response such as
GADD45A, PARP3 and DDB2 was identified. In contrast to acute response cell adhesion (Integrins), inflammatory (Interleukines),
angiogenesis and matrix-remodelling associated proteins (MMPs, PDGFs, VEGFs) were differentially regulated late.
Conclusions: Integrative analysis of protein expression and phosphorylation status revealed novel insights to an intricate protein
regulatory network underlying late radiation response in microvascular endothelium. The identified proteins involved in late
radiation response constitute attractive targets for modulation of radiotherapy effects in tumor-endothelium and to potentially
prevent undesired radiotherapy induced late side effects.
P181
Critical comparison of sample preparation strategies for shotgun proteomic analysis of formalin-fixed, paraffin-embedded
samples
1
1,2
1
1
1,2
A. Tanca , M. Abbondio , S. Pisanu , D. Pagnozzi , S. Uzzau , M. F. Addis
1
Porto Conte Ricerche, Alghero, Italy
2
Università di Sassari, Dipartimento di Scienze Biomediche, Sassari, Italy
1
Introduction: Hospital repositories store a huge number of formalin-fixed paraffin-embedded (FFPE) tissue samples, associated
with retrospective information regarding diagnosis, prognosis, and response to therapy, with important implications for protein
biomarker discovery and validation. Direct tissue trypsinization (DT) and protein extraction followed by in solution digestion
(ISD) or filter-aided sample preparation (FASP) are the most common workflows for shotgun proteomic analysis of FFPE samples,
but a critical comparison of the different methods is currently lacking.
Proteomic Forum 2015
188
Objectives: This study was aimed to compare 3 methods for FFPE shotgun proteomics: direct tissue trypsinization (DT) and
protein extraction followed by in solution digestion (ISD) and filter-aided sample preparation (FASP).
Materials & Methods: The 3 workflows were applied to tissue sections from an FFPE liver block. DT was preceded by
homogenization in ammonium bicarbonate, while ISD and FASP comprised protein extraction in SDS based-buffer, followed by
SDS depletion with spin columns and MW cut-off filters, respectively. Peptide mixtures were separated by long-gradient LC,
analyzed by Orbitrap MS and quantified according to a spectral counting approach.
Results: DT showed lower reproducibility, good preservation of high-MW proteins, a general bias towards hydrophilic and acidic
proteins, much lower keratin contamination, as well as higher abundance of non tryptic peptides. Conversely, FASP and ISD
proteomes were depleted in high-MW proteins and enriched in hydrophobic and membrane proteins; FASP provided higher
identification yields, while ISD exhibited higher reproducibility. Data concerning MW, pI and hydrophobicity are shown in Fig. 1.
Conclusion: These results highlight that diverse sample preparation strategies provide significantly different proteomic
information, and present typical biases that should be taken into account when dealing with FFPE samples. When a sufficient
amount of tissue is available, the complementary use of different methods is suggested to increase proteome coverage and
depth.
Fig. 1. Quantitative protein distribution according to physicochemical features. Quantitative protein distribution according to
MW (A), pI (B), number of transmembrane domains (TMD, C) and hydrophobicity (GRAVY score, D). (* = p < 0.05).
Figure 1
P182
Host-Adaptation of Burkholderia pseudomallei Alters Metabolism and Virulence Factors: a Global ComparativeProteomic
Analysis
1
1
1
V. Mariappan , K. M. Vellasamy , J. Vadivelu
1
University of Malaya, Medical Microbiology, Kuala Lumpur, Malaysia
Introduction: Burkholderia pseudomallei is the causative agent of melioidosis, a potential life threatening disease contracted
through inhalation, skin abrasions or direct inoculation from contaminated soil and water. The exact pathogenesis mechanisms
of melioidosis are still not completely understood, especially the interaction between host and pathogen that results during
initial infections.
Proteomic Forum 2015
189
Objective: It is of our interest to analyse the potential B. pseudomallei protein factors involved in the pathogen colonisation
progression using 2-DE and MALDI-TOF/TOF MSto perform a global comparative proteomic analysis from soil B. pseudomallei
and its counterpart passaged through host immune system.
Materials & Methods: B. pseudomallei MARAN (BpMARAN),was isolated from soil during a melioidosis outbreak in Malaysia
was injected intra-peritoneally into mice and the passagedcounterpart strain was recovered from the spleen (Bpmice-adapted).
A gel-based proteomic profiling approach was used to map and identify the differentially expressed proteins (fold change ≥2; pvalue ≤0.05).
Results: A total of 730 and 685 protein spots were visualised from BpMARAN and Bpmice-adapted, respectively. Of the 730
protein spots detected from the BpMARAN, 87 were differentially regulatedcompared to Bpmice-adapted (44 significantly upregulated and 43 significantly down-regulated). Overall, 44 up-regulated and 39 down-regulated spots, corresponding to 37 and
25 proteins, respectively, were identified using MS and database search. It was noticeable that five proteins (tuf, DnaK, hsp,
hypothetical protein BPSS1107 and BPSS1158) were found to be absent fromBpMARAN, but present in Bpmice-adapted. In
contrast, 12 proteins (clpB, katG, alededh, potF, pcaF, ompW, ppa, pixA, groEL,metX, argC and extracellular ligand binding
protein) weremissing from Bpmice-adapted. We predicted that majority of thedifferently up-regulated proteins were involved in
metabolism(60.66%), cellular processing and signal (24.59%), information storage and processing (6.56%) and poorly
characterised proteins (8.20%). In contrast, 38.30% down-regulated proteins were involved in cellular processing and signal,
23.40% ininformation storage and processing, 19.15% in metabolism andidentified as poorly characterised proteins,
respectively. The results presented in this study indicated that there were subsets of up-regulated proteins in BpMARAN,
particularly the adhesins, virulence factors, and stress- response proteins which are likely to contribute to colonisation.
Conclusion: Our study highlights the understanding of theevolution and adaptation of B. pseudomallei within the host during
infection and importance of proteins involved in the establishment of pathogenesis of melioidosis. This may providea novel
insight of bacterial factors which aid survivability of B. pseudomallei in the host.
P183
Investigation of the effect of cigarette smoke exposure on changes in the mitochondrial proteome of cultured human retinal
pigment epithelium cells
1
1
2
2
1
U. Ohmayer , J. Merl-Pham , M. Gorza , F. Perocchi , S. M. Hauck
1
Helmholtz Zentrum München, Research Unit Protein Science, Neuherberg, Germany
2
Gene Center Munich, Department of Biochemistry, Munich, Germany
Introduction: Age related macular degeneration (AMD) is a complex disease and the leading cause of blindness in elderly adults
in the developed world. It is characterized by degeneration of retinal pigment epithelium (RPE), the formation of extracellular
deposits, the thickening of Bruch’s membrane and results in irreparable loss of central vision. Beside genetic abnormalities it is
caused by several environmental influences among which long-term exposure to cigarette smoke is thought to result in the most
severe damages. These smoke induced damages are mainly mediated by complement activation and oxidative stress. Many
studies investigated the effects of cigarette smoke induced oxidative stress on different aspects of RPE vitality including
composition and functionality of mitochondria, whose integrity is essential for cell survival. However, a comprehensive mass
spectrometry-based approach investigating the effect of cigarette smoke on the mitochondrial proteome in the context of AMD
is missing.
Objectives: We aim to comprehensively characterize smoke exposure-induced changes in the mitochondrial proteome of
cultured human RPE cells in order to better understand major aspects of cigarette smoke induced damages on the integrity of
RPE cells on a molecular level.
Methods: Human retinal pigment epithelium (ARPE-19) cells were incubated with increasing amounts of cigarette smoke extract
(CSE) for 24 h. After high pressure mediated cell lysis, mitochondria of these cells and untreated control cells were purified and
subjected to proteomic analyses using a label free quantitative mass spectrometry based approach.
Results & Conclusions: In contrast to the analysis of the proteome of whole cell extracts, in which only a minor fraction (~15%)
of the identified proteins derive from mitochondria, we achieved a high enrichment of mitochondrial proteins. GO term analyses
indicated that more than 500 of the around 1200 identified proteins in these preliminary experiments were indeed of
mitochondrial origin. This underlines the applicability of the used fast and easy purification procedure for proteomic studies.
Strongly increased levels of heme oxigenase 1 (HMOX1) after CSE exposure confirmed the oxidative stress effect of cigarette
Proteomic Forum 2015
190
smoke on the used cultured cells. Changes in the levels of identified mitochondrial proteins after treatment with CSE, as
observed by differences in the abundances of the respective peptides, could be observed and will be discussed.
P184
Identification of galectin-specific glycoprotein ligands on RPE and RMG cells
1
1
2
J. Obermann , S. Hauck , C. Alge-Priglinger
1
Helmholtz Zentrum München, PROT, München, Germany
2
Ludwig-Maximilians-University Munich, Department of Ophthalmology, München, Germany
Introduction: Proliferative vitreoretinopathy (PVR) is a blinding disease frequently occurring after retinal detachment surgery. It
is characterized by the formation of sub- and epiretinal fibrocellular membranes, which contract and lead to repetitive tractional
retinal detachment. Adhesion, migration and epithelial-to-mesenchymal transition of retinal pigment epithelial (RPE) and retinal
Müller Glia (RMG) cells are the key cellular events in the onset of PVR.
Objectives: It is known that there is a glycomic shift on the cell surface of RPE and RMG cells associated with acquisition of a
myofibroblastic phenotype and that Galectin-1 (Gal-1), Gal-3, Gal-8 and Gal-9 - different types of carbohydrate binding proteins show a highly specific binding to dedifferentiated but not native RPE. Importantly, we have shown that galectins functionally
interfere with PVR-associated cellular events such as adhesion and proliferation. However, the cell surface proteins targeted by
specific galectins on RPE and RMG cells are largely unknown. We aim here at identification of galectin-specific glycoprotein
ligands on RPE and RMG cells as a prerequisite to unravel the functional effects of galectins on cellular behavior.
Material&Methods: For this purpose galectin pull-down experiments followed by mass spectrometry and western blot analysis
are established. For the galectin pull down, galectins are coupled on cyanogen bromide-activated Sepharose 4B. After incubation
of RPE and RMG cell lysates with these galectin-conjugated agarose beads and unconjugated control beads, beads were
sequentially eluted with sucrose and lactose. Eluates were analysed by label-free quantitative LC-MSMS and identified proteins
quantitatively clustered to identify those selectively enriched in lactose eluates.
Results: During previous studies Gal-3 counterreceptors on RPE were identified: integrinβ-1 and CD147. Those known
interactors were confirmed with this approach and additional >75 candidates for cell surface interactors could be detected. The
results of this glycoprotein enrichment technique will be discussed in detail and the specific glycoprotein ligands for all types of
galectins will be presented.
Conclusions: After the identification of galectin-specific glycoprotein ligands on RPE and RMG cells, novel glycophenotype-based
prognostic markers for PVR can be determined and new insights in the highly specific binding of Gal-1, Gal-3, Gal-8 and Gal-9 to
dedifferentiated but not native RPE and the following prevention of PVR-associated cellular events can be revealed.
P185
Proteome Analysis and Comparative Study of Normal Endometrial Stem cells, Endometriosis Stem Cells and Bone Marrow
Mesenchymal Stem Cells
1
S. heidari-keshel
1
SPU, farabi eye hospital, Tehran, Iran, Islamic Republic of
Background Information: The endometrial proliferation related diseases leads to endometrial hyperplasia, endometriosis.
Endometriosis associated with an increased risk of malignant transformation in approximately 1.0% of affected women, with the
involvement of multiple pathways of development.
Purpose: The aim of this study is Isolation and characterization of sub population stem cell in vitro condition from healthy and
endometriosis donors and proteomics evaluation for identification of putative markers.
Subject population & Method: Endometriosis endometrial stem cells (EESCs), normal endometrial stem cell (ESCs), endometrial
lesions stem cell (ELSCs) and bone marrow Mesenchymal stem cell (MSCs) were isolated. Flowcytometry and Real time-PCR
were utilized to detect the cell surface marker and expression pattern of stemness genes. Proliferation of all stem cells was
observed by MTT assay. The differentiation potential was evaluated by alizarin red, oil red O and RT-PCR method. The
karyotyping was performed on EESCs and ELSCs at passage 20. Furthermore, we systematically tested the immunophenotype of
cultured MSCs, EESCs, ESCs and ELSCs by flowcytometry analysis using a wide variety of markers. Direct comparison of this
Proteomic Forum 2015
191
phenotype to the one derived from cultured BM-MSCs demonstrated that cultured MSCs from both sources exhibit similar
expression patterns. Using the two-dimensional gel electrophoresis and matrix-assisted laser desorption ionization time-of-flight
mass spectrometry (MALDI-TOF-MS) approach, we have generated for the first time the protein map of cultured endometrial
stem cell by identifying 6 proteins, and we compared it directly to that of cultured BM-MSCs.
Findings: Stem cell characteristics of single cells isolated from normal endometrial and endometriosis tissues from women ages
28 ± 8 years (n = 18) were assessed. Fifteen of 18 samples contained a population of clonogenic cells with no significant
difference in property of morphology between the abnormal and normal endometrial stromal samples. Expression of CAPN7,
MMP2, MMP9 genes significantly higher than in ELSC group and CXCR4, FGFR3, SDF1 gens have a high expression level in EESC
sample. Meanwhile, compared with EESC-CM, ELSC-CM had greater effects for promoting MSCs and ESCs migration and
enhancing proliferation ability of cells (p < 0.01). Although these cells (EESCs, ESCs, ELSCs and MSCs) have background
expression of stemness-related genes, the unique patterns was detected. Spindle-like morphology, normal karyotype,
Adipogenic and osteogenic potential, expression of Oct4, Nanog, Klf4 and ERAS genes and CD44, CD105, CD90, CD73 and CD146
specific surface markers, in EESCs and ELSCs was observed. Furthermore, rate of proliferation and expression of these genes like
SALL4, DPPA2, TCL1 and Sox2 were significantly higher than ESCs and MSCs.
Conclusion: This study illustrates the presence of CD146 positive human Endometrial Somatic stem cell in the human
endometrium and reveals their potential use in regenerative medicine and taming stem cell heterogeneity for cell therapy and
Tissue engineering such as in the treatment of related infertility diseases.
P186
Proteomic profiling of Moroccan cobra Naja haje legionis venom and screening of its pharmacological activities
1
2
1
3
I. Malih , M. R. Ahmad Rusmili , T. Y. Tee, R. Saile , N. Ghalim , I. Othman
1
University Hassan II Mohammedia-Casablanca, Casablanca, Morocco
2
Monash University Sunway Campus, Malaysia, Selangor, Malaysia
3
Pasteur Institute of Morocco, Casablanca, Morocco
2
Moroccan cobra Naja haje legionis is one of the most dangerous snakes implicated in the pathogenesis of snakebites in
Morocco. The lack of information on the composition of its venom, as well as the clinical manifestations of envenomation by this
species, represents an obstacle for the medical management of bitten victims in the country.
To identify the essential components of the venom involved in the pathophysiology of envenomation by the cobra Naja haje
legionis, we have initially performed a proteomic characterization of the venom of Naja haje legionis, using a combination of
chromatographic techniques that includes gel filtration chromatography, reverse phase high performance liquid
chromatography, SDS-PAGE, tryptic digestion, tandem mass spectrometry and interrogation of the Swiss-Prot database.
Secondly, the pharmacological properties of the venom of Naja haje legionis were also studied by using tissue preparations from
rodents and chicks.
The obtained results revealed a very complex venom proteome and yielded to the identification of a total of 76 proteins from
Swiss-Prot database that can be classified into 9 protein families. The majority of venom families were those of three fingers
toxins, metalloproteinases and cysteine-rich secretory proteins representing up to 81 % of crude venom proteins. Moreover,
Naja haje legionis venom venom and its toxic fractions F3and F4 were found to exhibit neurotoxic, myotoxic and cardiotoxic
activities.
In conclusion, we have conducted for the first time and with success the proteomic and pharmacological characterization of the
venom of the Moroccan cobra Naja haje legionis. The results confirm the complexity and potency of the studied venom.
Proteomic Forum 2015
192
Figure 1
P187
Striatal global phosphoproteomics analysis of the zQ175 Huntington’s Disease mouse model following PDE10A inhibition
1
2
1
1
1
2
3
J. N. Dybowski , V. Beaumont , C. Schaab , K. Grundner-Culemann , S. Elschenbroich , D. Lavery , C. Schmidt , I. Munoz-San
2
Juan
1
Evotec (München) GmbH, Martinsried, Germany
2
CHDI Foundation, Los Angeles, California, United States
3
Pfizer Neuroscience Research Unit, Cambridge, Massachusetts, United States
Question: Currently, Huntington’s disease (HD), a hereditary, neurodegenerative disease, can only be managed but not cured.
Expression of Phosphodiesterase 10A (PDE10A), an enzyme promoting the hydrolysis of cyclic adenosine and guanosine
monophosphate, is dramatically reduced in Huntington’s Disease (HD) patients and models. Nevertheless, PDE10A inhibition has
been proposed as a viable target for the treatment of HD, as several studies showed improvement in motor and cognitive
function in rodent models.
Methods: To further elucidate the signaling pathways and target proteins under PDE10A regulation in an HD context, we
investigated the effect of treatment with the selective PDE10A inhibitor PF-02545920 on the proteome and phosphoproteome
of early symptomatic zQ175 knock-in mice. We applied quantitative mass spectrometry to globally profile the proteome and
phosphoproteome of striata of 6-month old wild type, heterozygous and homozygous zQ175 knock-in mice.
Results: Significant genotype-related differences in the basal level of 154 proteins were identified. Most prominently, protein
levels of Huntingtin, PDE10A and c-AMP producing Adenylyl cyclase 5 (ADCY5) were found to be reduced in zQ175 models.
Enrichment analysis revealed an up-regulation of proteasome and translation-related proteins in zQ175 models, while ion
channel and phosphodiesterase activity were found down-regulated.
Significance testing of PDE10A inhibition induced changes in the phosphoproteome of wild type, heterozygous and homozygous
groups revealed regulations in 690, 262 and 123 phosphorylation sites, respectively. Analysis of significantly enriched gene
ontology terms points to similar processes activated in wild type, heterozygous and homozygous zQ175 knock-in models.
However, the effects of PDE10A inhibition were less pronounced in hetero- and homozygous zQ175 models, as anticipated from
the reduced basal expression level of PDE10A and ADCY5 in zQ175.
Conclusions: Our in-depth global proteomics and phosphoproteomics profiling of PDE10A inhibitor treated zQ175 knock-in mice
revealed effects on signaling pathways and identified canonical and novel mechanisms downstream of PDE10 signaling which
may underlie the beneficial physiological changes observed on treatment.
Proteomic Forum 2015
193
P188
Clinical interest of Human Intestinal Mucus proteomics
1
1
1
2
1
P.- A. Gómez Buitrago , C.- A. Gonzalez-Correa , G. Taborda-Ocampo , R. Acuña-Sornoza , M. Santacoloma-Osorio , M.- A. Zezzi3
Arruda
1
University of Caldas, Manizales, Colombia
2
Colombian National Coffee Research Center , Manizales, Colombia
3
State University of Campinas, Campinas, Brazil
Introduction: Human intestinal epithelia is covered by Intestinal Mucus (HIM), a viscous fluid acting as the first line of physical
and biochemical defense, but in different ways along the gastrointestinal tract. Chemically, it consists of a network of the
glycoprotein Mucin 2 (Muc2) with proteins, water, electrolytes, acids, cells and microbiota embedded in it. In the colon, HIM
serves both as habitat for the microbiota, and, at the same time, as a protection of the host against microbial invasion. Analysis
of this bio fluid can give interesting clues for the study of different diseases, because it seems to behave a kind of written
memory of the patients´ clinical history, especially in relation to microbiota evolution, alterations of the immune system and the
intake of xenobiotics. In the future, it can be useful in the implementation of personalized medicine. The study of HIM is difficult
because it requires a sample of good quality and quantity, and because Muc2 is a complex and poor soluble mega protein.
Objective: To report about a methodology for the identification of proteins of the HIM with possible clinical interest.
Methodology: HIM was collected from six patients admitted to a leaning program for overweight people, which included
supplemented fasting, consumption of montmorillonite, colon cleansing and recto-sigmoidoscopy. Protein extraction was
carried out with a denaturation buffer and separation by electrophoretic gels, using SDS-PAGE of 13 cm and 2D-PAGE in pH
ranges of 3-10 and 4-7. A protein shotgun with desalination of the solution by SPE and albumin depletion was also carried out.
Digested peptides were analyzed in a chromatograph NanoAcquity UPLC coupled to a mass spectrometer HDMS equipped with
a NanoESI source. Samples were injected to capillary column C 18 with a 5 μL/min flow of water/acetonitrile (97:3) with 0.1%
formic acid and a subsequent transfer to C 18 analytical column and elution with a flow of 1 μL/min.
Results: 1) A methodology to analyze HIM is reported, from the preparation of patients, through the collection of samples and
up to the characterization of HIM proteins by mass spectrometry, using three different techniques: 2D-PAGE, SDS-PAGE, and
shotgun. 2) 14 proteins of possible clinical interest were identified: annexin A2, calcium-activated chloride channel regulator,
complement C3, disulfide isomerase, galectin-3 and 4, immunoglobulin programmed cell death 1 ligand, lamtor3, muc2,
nucleolar protein 8, ruvb like 2, tyrosine-protein phosphatase non-receptor type 11, zymogen granule membrane protein 16. 3)
Residues of cyclosporin were identified as human-calcineurin-complexed-with-cyclosporin, in a person who had consumed it 25
years ago.
Conclusion: proteomic analyzes of HIM can contribute to a better understanding of the pathophysiology and evolution of
different diseases, especially those associated with the human intestinal tract.
P189
H2S role in Amyotrophic Lateral Sclerosis: unravelling new cellular and molecular mechanisms
1,2
3
4
5
4,6
3
1,2
V. Greco , A. Spalloni , C. Piras , N. B. Mercuri , P. Roncada , P. Longone , A. Urbani
1
Santa Lucia Foundation, Proteomics and Metabonomics laboratory, Rome, Italy
2
University of Rome “Tor Vergata”, Department of Experimental Medicine and Surgery, Rome, Italy
3
Santa Lucia Foundation, Molecular Neurobiology Unit, Experimental Neurology, Rome, Italy
4
Università Degli Studi Di Milan, Dipartimento di Scienze Veterinarie e Sanità Pubblica , DIVET, Milan, Italy
5
University of Rome “Tor Vergata”, Neurological Clinic, Rome, Italy
6
Istituto Sperimentale Italiano Lazzaro Spallanzani, Milan, Italy
Introduction: Amyotrophic lateral sclerosis (ALS) is a lethal disease characterized by a progressive motor neuron degeneration.
Several etiologic factors are implicated in the pathogenesis such as glial inflammation, mitochondrial swelling, engulfation by
misfolded proteins and excitotoxicity. Hydrogen sulphide (H 2S) is an essential body product synthesized and recognized as an
endogenous neuromodulator in the central nervous system. In the brain, H 2S is mainly produced by astrocytes and microglia
through the cystathionine-β-synthase (CBS), a cytoplasmatic enzyme that accumulates in mitochondria under oxygen sensitive
conditions. Higher concentrations of H2S hamper the mitochondrial respiratory chain by inhibiting cytochrome C oxidase.
G93A
Moreover, impairments in the complex IV-driven respiration have been described in SOD1
mice.
Proteomic Forum 2015
194
Objectives: Our recent data show poisonous liquoral levels of H 2S in ALS patients and in the familial ALS mouse model
G93A
SOD1 [1]. Therefore, the aim of this study is to investigate the involvement of H2S in ALS and to unravel the cellular and
molecular mechanisms through which H2S contributes to the ALS-related neurodegeneration.
G93A
Methods: we developed specific HPLC test to measure H 2S levels in tissues and in spinal cord cultures of SOD1
mice. In order
to highlight key molecular pathways and putative protein-protein interactions, these data were extended to a deeper
proteomics analysis by a shotgun nLC-MS/MS profiling. Bioinformatics analysis was performed with Ingenuity Pathways Analysis
software. A targeted proteomics analysis was carried out using the Skyline software to validate the expression of proteins
involved in ALS related mechanisms.
G93A
Results: We found increased H2S levels in tissues and in spinal cord cultures of SOD1
mice. We showed that endogenous H2S
is released by astrocytes and microglia and that added H 2S is toxic for motor neurons. An accumulation of CBS has been shown
significant in the mitochondrial-enriched fraction, probably linked to hypoxic stress and to the increased production of H2S.
Shotgun proteomics analysis, supported by a targeted approach, showed some key proteins related to H 2S and ALS;
bioinformatics study allowed to highlight cellular and molecular networks in which these proteins might be involved.
Conclusions: Based on these data, the H2S toxic effects seem to associate with phenotype development in ALS. We may infer
that H2S over-production, due to the reactive gliosys, contributes to the motor neuron degeneration. In particular, we
hypothesize that the increased H2S amount reaching toxic levels could further distress an already compromised mitochondrial
function.
References: [1] Davoli A., Greco V., Spalloni A., et al. "Evidence of Hydrogen Sulphide involvement in Amyotrophic Lateral
Sclerosis” Ann Neurol. 2015 Jan 27. doi: 10.1002/ana.24372.
P190
Survival and proliferation of acute myeloid leukemia cells is triggered by FLT3-ITD and TLR9 activating different signaling
pathways via Bruton’s tyrosine kinase
1
2,3,4
2
2
2
2
5
3,4
6
7
J. Corso , T. Oellerich , S. Mohr , C. Doebele , S. Münch , G. Bug , J. Beck , J. Wicht , M. Oellerich , H. Bohnenberger , C.
7
5
2,3,4
8,9
Perske , E. Schütz , H. Serve , H. Urlaub
1
MPI for Biophysical Chemistry, Bioanalytical Mass Spectrometry Group, Göttingen, Germany
2
Goethe University, Hematology/Oncology, Department of Medicine II, Frankfurt am Main, Germany
3
German Cancer Consortium, Heidelberg, Germany
4
German Cancer Research Center, Heidelberg, Germany
5
Chronix Biomedical, San Jose, CA 95138, United States
6
Goethe University, Cardiology, Department of Medicine III, Frankfurt am Main, Germany
7
Georg August University, Department of Pathology, Göttingen, Germany
8
MPI für biophysikalische Chemie, Bioanalytische Massenspektrometrie, Göttingen, Germany
9
Georg August University, Bioanalytics, Göttingen, Germany
Acute myeloid leukemia (AML) is a genetic heterogeneous malignancy of myeloid progenitor cells which is characterized by fast
growth of leukemic blasts in the bone marrow leading to suppression of hematopoiesis. The receptor-type tyrosine-protein
kinase FLT3 with an internal tandem duplication (FLT3-ITD) and the spleen tyrosine kinase Syk have been already identified to
promote cell survival and proliferation in AML precursors. In this study we identified Bruton’s tyrosine kinase as constitutive
active kinase and analyzed its activation mechanisms in FLT3-ITD-negative (KG1) and FLT3-ITD-negative (MV4-11) positive AML
cell lines by interactome analysis, quantitative phosphoproteomics and transcriptome sequencing.
Methods: For proteomic analysis, the AML cell lines MV4-11 and KG1 were labeled with light (Arg+0/Lys+0) and heavy
(Arg+10/Lys+8) SILAC amino acids. For the characterization of the phospho-proteome a combination of strong cation exchange
chromatography (SCX) and titanium dioxide (TiO2) enrichment was applied. Additionally, antibody-based phosphotyrosine
enrichment was performed to specifically enrich for low abundant phosphorylated tyrosine containing peptides. Interactome
and phosphoproteomic samples were analyzed by a hybrid quadrupole-Orbitrap (Q Exactive, Thermo Scientific) mass
spectrometer coupled to a nanoflow liquid chromatography system. Resulting raw files were processed with MaxQuant (Version
1.3.0.5) and further downstream analysis was performed with Perseus. For transcriptomic analysis, sequencing libraries were
prepared and single-read sequencing was conducted on an Illumina HiSeq2000.
Proteomic Forum 2015
195
Results: Bone marrow biopsies from AML patients showed elevated Btk expression and activation levels in 26 of 30 AML cases.
Inhibition of Btk with the small-molecule inhibitor Ibrutinib leads to reduced cellular proliferation and apoptosis in different
primary AML cultures and cell lines.
Transcriptome analysis showed that Btk signaling is context dependent. In FLT3-ITD-positive AML, Btk drives FLT3-ITDdependent Myc- and STAT5 activation whereas in FLT3-ITD-negative AML, Btk couples Toll-like receptor 9 (TLR9) activation to
NFkB and STAT5. Both Btk-dependent transcriptional programs are involved in cell cycle progression and apoptosis.
Our quantitative phosphoproteomic analysis of FLT3-ITD-positive and FLT3-wildtype AML cell lines comparing Btk inhibition
against no inhibition identified different survival-inducing effectors. These appeared differentially regulated on the
phosphorylation level up- as well as downstream of Btk.
Conlusions: Taken together, in our study we identified context-dependent signaling pathways that point to subtype-specific
treatment strategies. Additionally, Btk is a potential new targetable kinase that was identified within this project and clinical
evaluation of Btk inhibitors in AML seems to be promising.
Quantitative Proteomics II (P216–P232)
P216
Quantitative Mass Spectrometric analysis Reveals the Target Acetylome of Epigenetic Inhibitor UVI5008
1
1
1
1
V. Sharma , A. de lera , E. Lasonder , H. G. Stunnenberg
1
Plymouth University, Biomedical Sciences, Plymouth, Netherlands
Introduction: Lysine acetylation has emerged as a key post-translational modification that exerts its effects on a wide range of
cellular pathways including tumourigenesis and cancer cell proliferation. However, there are few studies reporting in-depth
proteomic analysis and comparison of the target acetylome of KDAC inhibitors (KDACi).
Objectives: We aimed to quantify the changes in the acetylome in response to treatment with UVI5008, SAHA and Ex527 and to
reveal the similarities and differences in the mode of action of these inhibitors
Materials and Methods: For SILAC, the cells were labelled with either L-arginine and L-lysine or arginine-U-13C6-15N4 and Llysine-U-13C6-15N2 as described elsewhere (2). The light population of cells was treated with 5mM of SAHA or UVI5008 and
10uM EX527for 4 hours. Heavy labelled cell population was treated with DMSO as vehicle control. Cells were processed for
isolation of nuclear soluble, chromatin and cytosolic protein extraction, in-solution digestion and immunoenrichment (3). All
nLC-MS/MS-experiments were performed on nano-LC easy system from Proxeon connected to a 7-T LTQ-FTICR Ultra (Thermo
Electron, Bremen, Germany) mass spectrometer.
Results: We identified more than 2000 lysine acetylation sites on over 1000 proteins and quantified site-specific acetylation
changes in response to treatment with the KDAC inhibitors. Motif analysis revealed a strong preference for arginine downstream
to acetylated lysines upregulated by both Ex527 and UVI5008. UVI5008, SAHA and Ex527 treatments also result in an increase in
histone H4K5 and H4K16 acetylation. Furthermore, majority of the acetylation sites upregulated by SAHA show a motif
preference similar to sites upregulated by UVI5008.
Conclusions: Our quantitative data suggests putative SIRT1 specific regulation of acetylation sites on constituents of MOZ/MORF
and HBO1 protein complexes. Taken together, our data reveals a wide range of substrates affected by UVI5008 thereby
projecting this compound as a potent and a novel epidrug.
References:


1. Nebbioso, A., Pereira, R., Khanwalkar, H., Matarese, F., Garcia-Rodriguez, J., Miceli, M., Logie, C., Kedinger, V.,
Ferrara, F., Stunnenberg, H. G., de Lera, A. R., Gronemeyer, H., and Altucci, L. Death receptor pathway activation and
increase of ROS production by the triple epigenetic inhibitor, UVI5008. Molecular cancer therapeutics.
2. Ong, S. E., and Mann, M. (2006) A practical recipe for stable isotope labeling by amino acids in cell culture (SILAC).
Nature protocols.
Proteomic Forum 2015

196
3. Kim, S. C., Sprung, R., Chen, Y., Xu, Y., Ball, H., Pei, J., Cheng, T., Kho, Y., Xiao, H., Xiao, L., Grishin, N. V., White, M.,
Yang, X. J., and Zhao, Y. (2006) Substrate and functional diversity of lysine acetylation revealed by a proteomics survey.
Molecular cell.
P217
Identification of a novel neurotrophic factor from primary retinal Müller cells using SILAC
1
1
1
1
2
1
C. von Törne , J. Menzler , A. Ly , N. Senninger , M. Ueffing , S. M. Hauck
1
Helmholtz Zentrum München, Research Unit Protein Science, Neuherberg, Germany
2
University Medical Centre, Centre of Ophthalmology, Tübingen, Germany
Introduction: Retinal Muller glial cells (RMG) have a primary role in maintaining the homeostasis of the retina. In pathological
situations, RMG execute protective and regenerative effects, but can also contribute to neurodegeneration. Cultured primary
RMG have recently been recognized to secrete pro-survival factors for retinal neurons for up to two weeks in culture, but this
ability is lost when RMG are cultivated for longer durations.
Objective: The aim of our study was to identify novel neuroprotective factors, using a quantitative proteomics approach. By
comparing supernatants of RMG that demonstrated neuroprotection to supernatants from cells which had already lost their
protective capacity, we confirmed known factors and identified C-X-C motif chemokine 10 (CXCL10) a novel neuroprotective
factor.
Materials and Methods: Stable isotope labeling by amino acids in cell culture (SILAC) was used on primary porcine RMG.
Supernatants from cells cultivated for three weeks were compared to supernatants from cells cultivated for two weeks.
Photoreceptor survival assays were performed investigating the neuroprotective capacity of the supernatants as well as the
protein candidates differential in the SILAC analysis. Leukemia inhibitory factor (LIF) and identified C-X-C motif chemokine 10
(CXCL10) were additionally tested on porcine retinal explants. To illustrate the effect of candidates on pro-survival signaling,
western blot analyses on photoreceptors and explants were performed.
Results: We detected established neurotrophic factors such as transferrin, osteopontin (SPP1), and leukemia inhibitory factor
(LIF), and identified C-X-C motif chemokine 10 (CXCL10) as a novel candidate neuroprotective factor. All factors prolonged
photoreceptor survival in vitro. Ex-vivo treatment of retinal explants with LIF or CXCL10 demonstrated a neuroprotective effect
on photoreceptors (PR). Western blots on CXCL10 and LIF stimulated explanted retina and PR lysates indicated activation of prosurvival Signal Transducer and Activator of Transcription (STAT) signaling and B-cell lymphoma (BCL) pathways.
Conclusion: These findings suggest that CXCL10 contributes to the supportive potential of RMG towards retinal neurons.
P218
Using Study Design to streamline data processing of highly multiplexed quantitative experiments
1
1
1
2
1
M. Scigelova , T. N. Arrey , E. Damoc , R. Viner , B. Delanghe
1
Thermo Fisher Scientific, Bremen, Germany
2
Thermo Fisher Scientific, San Jose, United States
Purpose: The outcomes of quantitative experiments performed with a high resolution LC-MS system were used to assess the
usability and reporting capabilities of a new software platform.
Methods: Data files were obtained by LC-MS analysis of a standard proteome digest sample using different parameter settings
TM
for the isolation width, collision energy, resolving power and maximum ion fill time used for MS2 detection. Thermo Scientific
TM
Proteome Discoverer 2.0 software was used to compare the outcomes, relying extensively on the new capabilities of the
software platform enabling a complex study design, accommodation of multiplexed quantitative experiments with several study
factors (variables), results confidence filtering based on peptide and protein probability, and enhanced graphical output.
Results: The Study Design, a novel feature in Proteome Discoverer 2.0 software, streamlined performing numerous pair-wise as
well as multi-faceted comparisons. Using the example of an experimental study assessing the potential impact of several
acquisition method parameters deemed crucial for a multiplexed quantitation experiment, the advanced results filtering options
and graphical output of the sw package made it rather easy to evaluate the effect of several variables. The figures of merit used
for the evaluation were: number of confidently identified PSM (1%FDR), number of confidently identified protein groups (1%
Proteomic Forum 2015
197
FDR), number of quantifiable peptides (peptides containing the signal of all expected reporter ions), the extent of co-isolation
interference, and the protein ratio variability for experimental replicates. The outcome of the study provided a sound basis for
designing an optimized LC-MS acquisition method for the reporter ion-based quantification.
P219
Relative Quantitation of N-linked Glycans Using Carbonyl-Reactive Tandem Mass Tag™ (TMT™) Reagents
1
1,2
M. Oppermann , P. Suemchen
1
Thermo Fisher Scientific, Stockholm, Sweden
2
Thermo Fisher Scientific, Germany
Introduction and Objectives: Protein glycosylation is one of the most abundant and, arguably, one of the most difficult posttranslational modifications to study. Accurate quantitation of glycans remains elusive due to the lack of a comprehensive
selection of standards, poor ionization efficiency of carbohydrates relative to other classes of biomolecules, and broad structural
heterogeneity of glycomic samples. Recently, we have introduced a set of isobaric carbonyl-reactive Tandem Mass Tag (TMT)
reagents, Thermo Scientific™ aminoxy TMT™ Reagents, which enable efficient relative quantitation of carbohydrates, improve
ionization efficiency and increase analytical throughput. In this work we further explore the use of these reagents in quantitative
glycomics by combining our multiplexed TMT-based approach with HILIC LC-MS technique to enable more sensitive analysis with
improved glycome coverage.
Methods: N-glycan pools from several monoclonal antibodies and human serum samples were labeled with a set of isobaric
aminoxyTMT reagents and the samples were analyzed by ESI mass spectrometry in multiplex experiments. Samples were
analysed using a Thermo Scienitific™ Velos Pro™ dual-pressure linear ion trap and a Thermo Scientific™ Orbitrap Fusion™
Tribrid™ mass spectrometers. Combination of multiple-stage trap-HCD/CID fragmentation of the fully and partially protonated
precursor ions was used to obtain both quantitative and structural information for sample glycans.
Results and Discussion: We have established a protocol for complete derivatization of N-glycan mixtures (up to 100 μg total
sample), including all necessary quenching and clean-up steps. At least 20-fold improvement in signal intensity at the MS-level
was observed for native labeled glycans ions for all precursor types. A combination of trap-HCD and CID MS analysis enabled
2
both the quantitation by measuring reporter ion peak intensities at MS level and confirmation of the corresponding glycan
structures using the characteristic diagnostic ions. Both neutral and acidic glycoforms are amenable to this strategy.
Conclusions: The use of isobaric aminoxyTMT reagents enabled more sensitive relative quantitation of N-glycans with better
precision and increased throughput.
P220
Quantifying Safety Protein Biomarkers for Drug Induced Organ Injury in Animal Models and Humans
1
1
1
1
2
1
B. H. J. van den Berg , H. Planatscher , F. Weiss , H. Hammer , T. O. Joos , O. Pötz
1
NMI Naturwissenschaftliches und Medizinisches Institut an der Universität Tübingen , Protein Analytics, Reutlingen, Germany
2
NMI Naturwissenschaftliches und Medizinisches Institut an der Universität Tübingen , Biochemistry, Reutlingen, Germany
Organ toxicity accounts for the failure of many drug candidates. During the drug development process, drug induced organ
injury (DIOI) is assessed by pathology in animal models to predict possible toxicity in humans. Only a limited number of protein
biomarkers are in use to test drug-induced organ injury in humans and animals. Currently, more than 50 potential protein
biomarkers are under validation by two large joint academic-industrial consortia, the Predictive Safety Testing Consortium
(http://c-path.org/programs/pstc) and the Safer and Evidence-based Translation Consortium (http://www.imi-safe-t.eu).
However, these candidate markers could not be analyzed in all pharma-relevant animals, as established sandwich immunoassays
often reveal no cross-reactivity in all species. Therefore, the cross-validation in other species is challenging so far.
We developed quantitative MS-based immunoassays applicable for screening plasma and urine for liver and kidney injury
biomarkers across 5 species; human, cynomolgus, mouse, rat and dog. Our assays employ antibodies with short C-terminal
peptide epitopes (triple X proteomics antibody) present in tryptic fragments of the potential biomarkers in all 5 species. The
enrichment of peptides that share a C-terminal motif allows highly sensitive and selective mass spectrometry-based
quantification of proteotypic peptides from multiple proteins in tissue, plasma or urine with a single immunoprecipitation step.
Based on the short antibody epitopes the antibodies can be used as capture for MS-based immunoassays across species.
Proteomic Forum 2015
198
We demonstrate i) the capability to identify multiple biomarker in plasma and urine across species using a single TXP-antibodies,
ii) the validation (dynamic range, LOD, LOQ, reproducibility) of five safety protein biomarker in human, rat, and cynomolgous
urine and iii) the quantitative screening of healthy controls, DIKI patients, and animal studies on five safety protein biomarker
urine levels. Applying our TXP strategy allows rapid monitoring of drug induced organ injury across humans and preclinicalrelevant model organisms.
P221
Quantification of palmitoylated proteins in T cells using acyl-biotin-exchange
1
1,2
3
2
1
E. Morrison , B. Kuropka , S. Kliche , E. Krause , C. Freund
1
Freie Universitaet Berlin, Chemie und Biochemie, Berlin, Germany
2
Leibniz-Institut fuer Molekulare Pharmakologie, Berlin, Germany
3
Otto-von-Guericke University, Molecular and Clinical Immunology, Magdeburg, Germany
Palmitoylation, the reversible addition of 16-carbon palmitate groups to cysteine residues, has recently been implicated as a
general cellular mechanism for the localization of both cytoplasmic and membrane proteins to and within ordered lipid
microdomains in the plasma membrane. While this posttranslational modification has been relatively well studied in neurons, its
role in T-cell signaling generally remains an open question, despite the well-established identification of proteins like LAT, whose
palmitoylation is essential for sorting from the Golgi to lipid rafts in the PM, leading to proper immune response (Hundt et al., J
Immunol, 2009). Central to this question is the identification of proteins that are palmitoylated in T cells. A previous study used
metabolic labeling with a clickable palmitate analogue to identify ~300 palmitoylated proteins in Jurkat T cells (Martin and
Cravatt, Nat Meth, 2009). Here we report the use of acyl-biotin exchange, a sequential exchange of palmitoyl groups for biotin
followed by enrichment with streptavidin, coupled with quantitative LC-MS/MS, to identify 275 palmitoylated T cell proteins,
including 128 novel candidates. While providing much-needed insight into the role palmitoylation plays in T cells, this technique
also opens the door for further quantitative studies of global palmitomes in primary cells and organotypic tissue cultures.
Figure 1
P222
MSQBAT: A platform-independent software suite for the label-free quantification of proteins from (Ge)LC-MALDI-MS and
(Ge)LC-ESI-MS data
1,2,3
1,2,3
1,2
A. Kerner , W. Nadler , C. Rösli
1
DKFZ, Junior Research Group Biomarker Discovery, Heidelberg, Germany
2
HI-STEM, Biomarker Discovery, Heidelberg, Germany
3
Helmholtz Int. Grad. School for Cancer Research, Heidelberg, Germany
Introduction: Accessing the relative changes in protein abundance is essential for a proper understanding of the various
processes underlying disease progression and development.Mass spectrometry-based proteomics allows today for the
identification of several thousand proteins in a single analysis.To measure quantitative changes in protein abundance, biological
samples need either to be labeled using stable isotopes or protein abundances have to be computed using so called label-free
techniques.Today, no software exists for the quantification of data from state-of-the-art LC-MALDI-MS systems. LC-MALDI-MS is
Proteomic Forum 2015
199
a widely used system for routine applications in proteomics and as well in other fields of diagnostic. In contrast to LC-ESI-MS, the
1
decoupling of liquid chromatographic sample separation and mass spectrometric sample analysis allows to separate MS and
2
MS in time, thereby providing unique possibilities for advanced and dynamic data acquisition mostly unused today.
Methods: In order to quantify data acquired by LC-MALDI-MS and to take full advantage of the mentioned LC-MALDI-MS specific
features, a novel software suite termed MS QBAT was developed and evaluated. MSQBAT is a platform independent software
1
suite for MS -based, label-free protein quantification. Quantification capabilities were evaluated using spike-in experiments
analyzed using both different proteomic workflows and instruments. Human proteins (UPS2, Sigma-Aldrich) were spiked in
variable concentrations into a complex E.coli background proteome and processed both using a LC-MS and GeLC-MS approach.
Samples were chromatographically separated on a nanoACQUITY UPLC using a 110 minutes and 60 minutes gradient
respectively and subsequently analyzed by an AB SCIEX TOF/TOF 5800 System and an AB SCIEX QTRAP 6500 System.
Results: MSQBAT can be applied to quantify data deriving from both LC/GeLC-MALDI-MS and LC/GeLC-ESI-MS workflows with
high accuracy.
Conclusion: In contrast to other software solutions, MS Q BAT is highly flexible and suited for the quantification of mass
spectrometric data from various instrumental setups and proteomic workflows, such as (Ge)LC-MALDI-MS and (Ge)LC-ESI-MS.
MSQBAT therefore has a range of application outperforming currently available solutions.
Figure 1
Figure 2
P223
Phosphoserine-mediated protein interactions of the T cell adaptor protein ADAP
1
1
1,2
A. Manns , E. Krause , C. Freund
1
Leibniz-Institut für Molekulare Pharmakologie im Forschungsverbund Berlin e.V. (FMP) , , Berlin, Germany
2
Freie Universität Berlin, Chemie und Biochemie, Berlin, Germany
Reversible protein phosphorylation is an important feature in T cell signaling, regulating many of the signal transduction
pathways required for proper T cell functioning. While tyrosine phosphorylation events immediately following T cell receptor
(TCR) stimulation have been described in great detail, the role of phosphorylation at serine and threonine residues is less well
described for many proteins involved in TCR proximal signaling. The human adhesion and degranulation promoting adaptor
protein (ADAP) plays a central role in T cell signaling since it becomes tyrosine phosphorylated at multiple sites and thereby
serving as a hub for SH2 domain containing effector proteins [1]. More recently, phosphoproteomics has identified several
serine and threonine phosphorylation sites within the N-terminal domain of ADAP. However, the role of these phosphorylation
events and of serine/threonine kinases in T cell signaling events mediated by ADAP has not been investigated. Here, we use
peptide pull-down assays and quantitative (SILAC) mass spectrometry to identify interaction partners of serine-phosphorylated
ADAP sequences. We demonstrate for the first time that distinct phosphorylated serine residues in the N-terminal part of ADAP
bind specifically to 14-3-3 proteins. This interaction may yield a new insight into T cell signaling pathways steered by these type
of interactions.
Proteomic Forum 2015
200
1 Sylvester M, Kliche S, Lange S, Geithner S, Klemm C, et al. (2010) Adhesion and Degranulation Promoting Adapter Protein
(ADAP) Is a Central Hub for Phosphotyrosine-Mediated Interactions in T Cells. PLoS ONE 5(8)
P224
Oyster shotgun proteomics provide new insights into various cellular pathways of multiple climate-related stressors
1
1
D. Ramadoss , T. V
1
The University of Hong Kong, Biological Science, Hong Kong, Hong Kong
Edible oysters have an incredibly complex life-cycle with metamorphic process that transforms swimming larva to sessile adult in
minutes. This transition is perhaps the most challenging step because larva must choose the right place, switch their molecular
machinery to suit sessile life and start building calcite shells. In addition to global warming and extreme weather patterns, the
high CO2 levels have been leading to progressive Ocean Acidification (OA). OA compounded with other stressors will hinder the
adaptability of the marine organisms to the ensuing changes and might impact human dependence on oceans as a source of
food. High temperature, low salinity and low pH resulting from anthropogenic CO 2 induced climate change are detrimental to
edible and economically important oyster populations. Mechanistic studies to understand the tolerance responses of closely
related species assume significance in this context, but have not been reported thus far.
Using the newly annotated complete genome information of the Pacific oyster and iTRAQ quantitative proteomics approach, the
interactive effect of multiple climate stressors P, T and S on larval proteome was examined in the current study. Over 1369
proteins were identified in three reproducible iTRAQ experiments. Interaction of three stressors (TSP) up-regulated protein
processing, cell signalling, energy production, growth and developmental pathways. The warming activated metabolic and
cellular stress responses whereas freshening activated ionic and oxidative stress response pathways. Interestingly this study also
identified, that temperature showed a rescuing effect as fewer pathways were altered at TP when compared to P.
This study identified species- and stressor-specific tolerance responses, and survival mechanisms that preserved calcification, in
oyster larvae. The study reveals the existence of tolerance mechanisms in oysters that help them adapt to stresses resulting
from ecosystem fluctuations. This study pioneered the use of a proteomics approach to understand the impact of multiple
stressors on oyster larvae and the molecular mechanisms underlying their successful adaptation to them.
P225
Global protein oxidation profiling of Podospora anserina mitochondria suggests efficient removal of damaged proteins even
at high age
1
2
2
2
1
1
C. Ramallo Guevara , O. Philipp , A. Hamann , H. D. Osiewacz , S. Rexroth , M. Rögner , A. Poetsch
1
Ruhr-University Bochum, Bochum, Germany
2
Goethe University Frankfurt , Frankfurt am Main, Germany
1
Introduction: The free radical biology of aging is based on the idea that reactive oxygen species (ROS) damage cellular
components and should lead to the accumulation of age-related protein oxidation [1]. While most studies target either a
fraction of the oxidized or the non-oxidized protein species, we argued that simultaneous and unbiased detection would be
most suitable to interpret the biological consequence of oxidative protein damage during the course of ageing. This study with
the aging model Podospora anserina [2] focuses on the mitochondrial proteome as main target for ROS-induced damage.
Objectives: The objective of this study was the development of a new proteomics workflow for global, quantitative and untargeted analyses of oxidative protein modifications to scrutinize the ROS theory.
Methods: To account for the low amount of modified proteins and the danger of artificial protein oxidation, a workflow
consisting of gel-free sample preparation and iTRAQ quantification was established. Mitochondrial proteins from four age stages
were digested with trypsin using the slightly modified FASP protocol [3] and iTRAQ-labeled peptides were separated on a 25 cm
column with a NanoAcquity system. A four hour linear solvent gradient was chosen to obtain high proteome coverage and to
reduce quantitation interference from co-eluting peptides. For optimal peptide identification and quantitation, MS analysis was
done on an LTQ Orbitrap Velos using CID for identification and HCD for quantitation. A statistical framework was developed that
allowed to quantify oxidized and non-oxidized protein species, accounting for identification rate, variance, regulation trend, and
correlation in regulation profiles between modified and un-modified protein species.
Proteomic Forum 2015
201
Results: This first large scale, unbiased oxidative modification-centric study for mitochondrial aging allowed the comprehensive
quantification of 2352 protein species and 23 different oxidative amino acid modifications. For 746 proteins unmodified and
oxidatively modified species were detected. For the majority of proteins a positive correlation of changes in protein amount and
oxidative damage, and rarely age-related increases in protein oxidation were noticed. Our data suggest that P. anserina is
efficiently capable to compensate ROS-induced protein damage during aging as long as protein de-novo synthesis is functioning,
ultimately leading to an overall constant relationship between damaged und undamaged protein species.
Conclusion: In contradiction to the ROS theory, our results do not confirm massive increase in protein oxidation during aging
and rather suggest a functional protein homeostasis mechanism even at high age.
[1] Groebe et al. (2007) Experimental Gerontology, 42: 887-898
[2] Osiewacz (2007) Biologie in unserer Zeit, Volume 37: 164-172
[3] Wisniewski et al. (2011) Analytical Biochemistry, Volume 410: 307-309
P226
High resolution quadrupole TOF for quantitative proteomics
1
1
2
2
2
2
1
3
1
S. Beck , F. Meier , A. Michalski , M. Lubeck , S. Kaspar , O. Raether , I. Paron , J. Cox , M. Mann
1
Max-Planck-Institute of Biochemistry, Proteomics and Signal Transduction, Martinsried (near Munich), Germany
2
Bruker Daltonik GmbH, Bremen, Germany
3
Max-Planck-Institute of Biochemistry, Computational Systems Biochemistry, Martinsried (near Munich), Germany
Fundamental challenges of shotgun proteomics comprise the very large number of peptides that elute over a relatively short
period of time and peptide abundances that vary by many orders of magnitude. To address these critical challenges
developments in mass spectrometry technology towards higher sensitivity, sequencing speed and resolution are needed. Hybrid
quadrupole time-of-flight (QTOF) instruments are one of the two major mass spectrometers used in proteomics. Over the last
decade, the QTOF technology improved significantly in terms of achievable resolution, mass accuracy and dynamic range.
Recently, we described and optimized the impact II, a benchtop QTOF instrument, as an instrument for shotgun proteomics
(Beck et al., manuscript submitted). Here, we further evaluate the quantitative capabilities of the impact II.
We used an Easy nLC-1000 coupled on-line to the Bruker impact II with a modified CaptiveSpray ion source. Peptide mixtures
were loaded onto in-house packed columns (50 cm, 75 µm ID) filled with C 18 material (ReproSil-Pur C18 AQ, 1.9 µm).
Chromatographic separation was carried out using a linear gradient of 5-30% buffer B (80% ACN and 0.5% acetic acid) at a flow
rate of 250 nL/min over 90 min. All data were analyzed with the MaxQuant software, which was further developed to optimally
handle QTOF data.
The Bruker impact II is equipped with a new reflectron and detector, which improved the mass resolution up to 40,000 at m/z
1222. Adaption of our heated LC system yielded very narrow peptide elution peaks.
To evaluate the label-free quantitative capabilities of the impact II we spiked the Universal Protein Standard 2 in two different
concentrations into the yeast proteome. The instrument achieved accurate fold change determination in this spike-in
2
experiment over four orders of magnitude. Moreover, we demonstrated high quantitative reproducibility (R > 0.99) by
analyzing technical replicates of complex peptide mixtures derived from a mammalian cell line.
We also developed workflows for the analysis of chemical labeled samples. We investigated the capability of the instrument to
perform “transfer time stepping”, which refers to the selective extraction of low mass ions without the loss of the higher mass
fragment range. We reasoned that this feature might be valuable for experiments that rely on quantification of low mass
reporter ions, e.g. iTRAQ or TMT labeling strategies. Evaluating the ion current, we found that these ions are transferred with
more than 60% efficiency during the low m/z extraction phase.
We conclude that the impact II is a high resolution quadrupole time-of-flight instrument that allows high quantitative
reproducibility and accurate fold change determination. It can be used for label-free quantification as well as for chemical
labeled strategies. We expect additional method developments to further improve these results.
Proteomic Forum 2015
202
P227
pyQms - a versatile and fast python module for accurate quantitative mass spectrometry
1
1
1
1
J. Barth , A. Niehues , M. Hippler , C. Fufezan
1
Westfälische Wilhelms-Universität, Münster, Germany
Introduction: The importance of high throughput quantitative mass spectrometry is evidenced by its wide applications in
Proteomics as well as Metabolomics, Lipodomics or Modomics. However, a computational tool to find, score and quantify the
large variety of potential molecules with different modifications and labelling states and types is missing. Here we present
pyQms, a Python module which allows precise quantification of molecules measured by high-accuracy mass spectrometry
Objectives: pyQms is able to quantify molecules with any labelling and any combination of modifications. Its main advantages
are accuracy, speed and adaptability. Modern experiments devise approaches employing novel labelling methods that require
tailored tools. pyQms has the unprecedented ability to quantify molecules in pulse-chase experiments i.e. molecules with a
15
partial labelling (e.g. N from 0 to 100% in 1% steps). It enables most accurate pulse-chase experiment analysis, even with
multiple chase elements, which is beyond current state of the art.
Materials & Methods: With pyQms anyone can seamlessly incorporate a powerful platform independent Python open source
library into workflows and analysis pipelines, which allows any potential labelling combination to be quantified with ease.
pyQms uses the mzML file standard and data processing speed is enabled by using pymzML (1).
Results. pyQms can correctly detect partial labelled peptides and determine their amounts in mixed samples in a gold standard
data set. The labelling state of each strain can be identified unambiguously. The general overall matching speed advantage of
pyQms is reflected in around 590.000 ± 60.000 matches / sec (single i7 core). pyQms has already been successfully employed in
different studies including large scale proteomics studies (2-4).
Conclusion. pyQms offers a fast, accurate and varied Python module for quantitative mass spectrometry data with no limits on
the complexity of molecule labelling and modifications. Novel questions about protein turn-over or metabolite stability can be
addressed with an unmatched high level of detail.
References
1. Bald, T. et al. (2012) Bioinformatics 28, 1052-3
2. Höhner, R. et al. (2013) MCP 12, 2774-90
3. Barth, J. et al. (2014) MCP 13, 969-89
4. Kukuczka, B. et al. (2014) Plant Physiol. 165, 1604-1617
P228
In vivo cell type specific analysis of protein-protein interactions
1
2
2
1
D. Vasiljevic , S. Seelk , B. Tursun , M. Selbach
1
Max-Delbrück Centrum, Proteome Dynamics, Berlin, Germany
2
Max-Delbrück-Centrum, Berlin, Germany
Protein-protein interactions (PPIs) are involved in almost every biological process. Therefore, different methods have been
developed and employed in order to map PPIs systematically. However, our knowledge about the cell type specificity of PPIs is
very limited since available approaches cannot provide this information.
Here, we are developing a mass spectrometry-based method to study cell type specific PPIs in vivo. The method is based on
tagging a target protein with a composite tag that allows in vivo biotinylation by the bacterial biotin ligase BirA in a cell type
specific manner. Sequence specific enzymatic elution of streptavidin pulled-down complexes, the possibility of second
purification step and mass spectrometry-based absolute quantification are additional advantages.
Although the project is still in an early phase, preliminary results show that the protein of interest can be biotinylated in vivo in
C. elegans and efficiently enriched. Nevertheless, further optimization is required in order to stabilize native protein complexes
to also capture weak and transient interactions and to reduce unspecific background.
Proteomic Forum 2015
203
P229
Development of a gel-based quantitative mass spectrometry assay for determining the fragmentation of cardiac troponin T in
serum
1
1
2
2
1
1
A. Streng , D. de Boer , F. Bouwman , E. Mariman , M. van Dieijen-Visser , W. Wodzig
1
Maastricht University Medical Centre, Central diagnostic laboratory, Maastricht, Netherlands
2
Maastricht University, Human Biology, Maastricht, Netherlands
Introduction: Identifying and quantifying low-abundant proteins from a serum matrix using mass spectrometry (MS) is a
daunting and challenging task. A typical proteomics workflow includes the identification of proteins of interest and the
subsequent quantification of selected peptides representing those proteins using one of several targeted proteomics
approaches. A critical assumption in this workflow is that the representing peptides do not undergo posttranslational
modifications such as degradation or phosphorylation, since this may influence signal intensity. Following this line of reasoning;
if there exists a constant intensity ratio between an unmodified peptide of interest and a quantotypic peptide used as an
internal standard, it should be possible to quantify the extent of modification within a protein based upon the change of this
ratio.
Objectives: The aim of this study is to develop a gel-based targeted proteomics assay to quantify the degree of modification of a
protein. Fragmentation of the protein cardiac Troponin T (cTnT) in serum is used to demonstrate the functionality of this assay.
cTnT is an important biomarker which is routinely used in the diagnosis of acute myocardial infarctions but due to
fragmentation, its exact molecular structure in serum varies.
Materials & methods: Intact and fragmented cTnT was isolated from serum using immunoprecipitation employing the cTnTspecific M11.7 monoclonal antibody (Roche Diagnostics) and fractionated with SDS-PAGE. Coomassie stained bands at 37, 27, 18
and 16 kDa were excised and digested with MS-grade porcine trypsin (Promega) following accepted protocols. Digests were
analysed on a Q Exactive instrument (Thermo Scientific) set on targeted Selected Ion Monitoring (t-SIM) mode with data
dependent tandem-MS (dd-MS2) for identification. Retention times and optimal collision energies were determined by
measuring synthesised cTnT peptides of interest using a data-dependent method. Database searching was performed using
SEQUEST and intensity chromatograms were analysed with the targeted proteomics software platform Skyline.
Results: Table 1 shows the targeted precursor ions and their properties. The linearity, sensitivity and reproducibility of the tSIM/dd-MS2 assay were first validated for each peptide by creating a dilution series of purified cTnT spiked in human serum.
Next, we incubated intact cTnT in human serum at 37°C for up to 48 h to induce fragmentation. Using our assay, we have been
able to identify clear differences in the ratios of several N- and C-terminal peptides, indicative of proteolytic cleavage, and
identified the cleavage sites of 4 different cTnT fragments.
Conclusion: A quantitative targeted mass spectrometric method was developed to identify the extent of fragmentation of cTnT
in human serum. In principle, this method can be expanded to include other proteins and modifications.
Figure 1
Proteomic Forum 2015
204
P230
specL - An R/Bioconductor package to prepare peptide spectrum matches for use in targeted proteomics
1
1
1
1
C. Panse , J. Grossmann , C. Trachsel , R. Schlapbach
1
Functional Genomics Center Zurich, Zurich, Switzerland
Targeted data extraction methods are attractive ways to obtain quantitative peptide information from a proteomics experiment.
SWATH and DIA methods increase reproducibility of acquired data because the classical precursor selection is omitted and all
present precursors are fragmented. However, especially for targeted data extraction, MS coordinates, i.e., retention time
information precursor and fragment masses, are required for the particular entities (peptide ions). These coordinates are usually
estimated in a so called discovery experiment earlier on in the project if not available in public spectral library repositories. The
quality of the assay panel is crucial to ensure appropriate downstream analysis.
specL provides methodes for generating spectra libraries which can be used for MRM SRM MS workflows in proteomics. The
package provides a BiblioSpec reader, a function which can add the protein information using a FASTA formatted amino acid file,
and an export method for using the created library in the Spectronaut software
P231
High resolution proteomics
1
C. Koehler
1
UIO, Oslo, Norway
The proteomics field has shifted during the last years from two-dimensional gel electrophoresis (2-DE)-based approaches to SDSPAGE or gel free workflows. However, 2-DE still offers the highest resolution in protein separation. We combined SILAC of
controls and apoptotic HeLa cells with 2-DE and subsequent analysis of tryptic peptides by nano-LC coupled to an LTQ-Orbitrap
mass spectrometer to obtain quantitative data using the methods with the highest resolving power on all levels of the
proteomics workflow. More than 1,200 proteins with more than 2,700 protein species were identified and quantified from 816
Coomassie Brilliant Blue G-250 stained 2-DE spots. About half of the proteins were identified and quantified only in single 2-DE
spots. The majority of spots revealed one to five proteins, however, up to 23 proteins were identified in a single 2-DE spot. Only
half of the 2-DE spots represented a dominant protein with more than 90% of whole protein amount. Consequently,
quantification based on staining intensities in 2-DE gels would in approximately half of the spots be imprecise and minor
components could not be quantified. These circumstances were circumvented by quantification using SILAC.
P232
Multilayered Genetic and Omics Dissection - A new age for biomedical researches
1
2
2
Y. Wu , E. Williams , J. Auwerx , R. Aebersold
1
ETH Zurich, Zurich, Switzerland
2
EPFL, Lausanne, Switzerland
1
The manner by which genotype and environment coordinately influence complex traits is one of the fundamental questions in
biology. On the one hand, the proteome is more than a mere translation of a genome. DNA sequence is not an accurate proxy
for its transcript and protein product, and transcript levels generally have modest correlation with the levels of their
corresponding proteins (Ghazalpour et al., 2011; Gygi et al., 1999). On the other hand, genome-wide association analysis
(GWAS) and quantitative trait locus (QTL) analysis have been successfully applied to identify genes driving phenotypic variations
and have provided valuable evidence on disease-driven SNPs on gene expression. However, it has become evident that
regulation of complex traits cannot be fully explained by reductive gene-by-gene or protein-by-protein theory, and genotypephenotype association will become apparent only with system-scale measurement and multilayered dissection of gene products
in the form of dynamic networks.
In the previous study, we combined transcriptomics, targeted proteomics and metabolomics strategies, and applied it to
investigate mitochondrial activity in a mammalian genetic reference population (Wu et al., 2014). We have identified dozens of
transcript and protein QTLs, several of which were linked to phenotypes. Most significantly, we identified Dhtkd1 as a diabetic
gene shared in mice and human. Furthermore, these multilayered data sets allowed further characterization of the
mt
mitochondrial unfolded protein reponse (UPR ), which shows significantly variant responses at the transcript and protein level.
Proteomic Forum 2015
205
Currently, we expanded this integrated omics approach to a more system-wide scale using Sequential Window Acquisition of all
THeoretical Mass Spectra (SWATH-MS) and obtained complete coverage of 2,622 proteins in the livers of 80 distinct BXD mice,
providing an unprecedentedly comprehensive view of cellular processes in a population study. These findings indicate that the
integrated multilayered omics approach is ready to tackle complex traits from an entirely new angle, and a new age for
biomedical researches has just begun.
Structural Proteomics (P233–P238)
P233
FtsZ Protofilaments Use a Hinge-Opening Mechanism for Constrictive Force Generation
1
Y. Li
1
Zhejiang University, Life Science Institute, Hangzhou, Zhejiang Province, China
Questions: The essential bacterial protein FtsZ is a guanosine triphosphatase that self-assembles into a structure at the division
site termed the “Z ring”. During cytokinesis, the Z ring exerts a constrictive force on the membrane by using the chemical energy
of guanosine triphosphate hydrolysis. Although multiple crystal structures of FtsZ in different nucleotide states and as
monomers and protofilaments are available, little is known about the structural mechanism underlying a hydrolysis-mediated
conformational change.
Methods: We determined the crystal structure of Mycobacterium tuberculosis FtsZ (MtbFtsZ)-guanosine diphosphate (GDP)
which reveals an intersubunit interface formed by highly conserved residues. To confirm the physiological relevance of the
observed intersubunit interface, firstly, we created many mutants in E.coli FtsZ (EcFtsZ) based on our crystallographic
observations and used a complementation system in E.coli to characterize their division phenotypes. Then we complemented
the GTPase assay and EM study in vitro to confirm that the conserved residues dramatically affect the longitudinal assembly.
Finally, we compared the structural of the curved MtbFtsZ-GDP dimer with the straight SaFtsZ-GDP dimer and then proposed
the“hinge-opening” model.
Results: In the E.coli complementary assay , all hydrophobic residues within the intersubunit interface are physiologically
relevant, and these mutants have a dramatically fall in the GTPase activity. In the EM study, these mutants cannot assemble into
long filaments while the wildtype can assmeble into filaments in bundlings. The comparison betweent the curved and straight
structures reveals a “hinge-opening” model and in this model, a hydrolysis-dependent conformational switch at the T3 loop
leads to longitudinal bending between subunits, which could generate sufficient force to drive cytokinesis.
Conclusion: We concluded the molecular basis of the energy transfer through the structure comparison between the curved and
straight structures. The GTP g-phosphate stabilizes the T3 loop conformation in a compact state (tension or Tstate). The same
Tstate conformation is maintained in all GTPbound and some GDP-bound FtsZ structures and is necessary for the longitudinal
assembly of a straight FtsZ protofilament, in which the T3 loop interacts extensively with the T7 loop of the top subunit. In
contrast, in our GDP-boundMtbFtsZ structure, the T3 loop adopts a relaxed conformation (R state) in the absence of the gphosphate, indicating that it is flexible and can be in either state. Because both β- and γ-phosphates are negatively charged, the
release of g-phosphate by hydrolysis may trigger the T3 loop transition from a T to an R state. This would weaken the FtsZ
longitudinal interactions between the T3 and T7 loops and further drive the hinge-opening event around the pivot point.
P234
Quantitative XL-MS and different crosslinker reveal parallel orientation of HOP2 and MND1
1
1
2
3
4
2
1,4
E. Rampler , T. Stranzl , Z. Orban-Nemeth , D. Hollenstein , O. Hudecz , P. Schloegelhofer , K. Mechtler
1
IMP, Institute of Molecular Pathology, Mass Spectrometry and Protein Chemistry, Vienna, Austria
2
Max F. Perutz Laboratories, Department of Chromosome Biology, Vienna, Austria
3
Max F. Perutz Laboratories, Department of Biochemistry and Cell Biology, Vienna, Austria
4
IMBA, Institute of Biotechnology of the Austrian Academy of Sciences, Mass Spectrometry and Protein Chemistry, Vienna,
Austria
In recent years, chemical crosslinking of proteins combined with mass spectrometry (XL-MS) matured into a state-of-the-art
technique to obtain structural protein information. Despite wide-spread use of the technology in basic research, several critical
issues have not been addressed sufficiently for the design of broadly applicable protocols. Among these issues are (1)
optimization of workflows in terms of protein concentration needed, (2) choice of appropriate crosslinkers and (3) quantitation.
Proteomic Forum 2015
206
To address these aspects, we investigated the Arabidopsis thaliana HOP2/MND1 heterodimer, a highly conserved protein
complex important for meiosis. The optimized crosslinking mass spectrometry (XL-MS) workflow encompassed the use of
crosslinkers with different distances, quenching, digestion, enrichment and HCD based LC-MS detection prior to data evaluation.
We introduced a novel methodology for size exclusion allowing considerably less protein starting material than reported in the
literature. Moreover, we applied and systematically tested three different homobifunctional amine-reactive crosslinkers (DSS11.4 Å; BS2G and BS2GH/L, 7.7 Å) as well as a heterobifunctional zero-length crosslinker (EDC). We evaluated the different
crosslinkers with respect to conformational information on the HOP2/MND1 complex. The majority of the crosslinks were
identified in the coiled-coil domains of HOP2/MND1, known to be sufficient for complex formation. Importantly, the gained
information led us to propose that HOP2 and MND1 are arranged in a parallel orientation within the complex. Finally, we
performed first steps towards quantitative crosslinking using differentially labeled crosslinker to compare heated versus nonheated states of HOP2/MND1.
P235
Enrichment of Cross-Linked Peptides using charge-based fractional diagonal chromatography
1
1
1
1
V. Tinnefeld , S. Verhelst , R. Zahedi , A. Sickmann
1
Leibniz-Institut für Analytische Wissenschaften ISAS e.V., Systemanalyse, Dortmund, Germany
Introduction: Protein complexes are the basis for cellular signaling, energy transfer, motion, stability and cellular metabolism.
Understanding these complex interactions on the molecular level is an essential step to obtain a comprehensive insight into
cellular mechanisms. Protein cross-linking mass spectrometry is an emerging approach to study protein structures and
complexes. Here, proteins are covalently fixed to preserve their native states and interactions and to make those accessible for
mass spectrometric analysis.
Objectives: Especially within complex samples, cross-linked peptides are very low-abundant. Hence there is a need for efficient
enrichment techniques that can improve their analysis by MS. An obvious choice for enrichment is strong cation exchange
chromatography (SCX), since on average cross-linked tryptic peptides have higher charge states than non-linked peptides, owing
to the presence of two N-termini and tryptic cleavage sites.
We evaluated whether our recently developed ‘charge-based fractional diagonal chromatography (ChaFRADIC) approach 1 can
be adopted for the enrichment of cross-linked peptides prior to LC-MS.
Materials & Methods: The approach is based on a two dimensional separation of peptides by SCX. In the first dimension
peptides are separated into fractions representing different peptide charge states at pH 2.7. These fractions are subjected to
chemical derivatization inducing a charge state shift that can be used to enrich cross-linked peptides in a second dimension.
Results: By inducting a specific charge shift of cross-linked peptides, we could achieve an effective enrichment within model
systems of low complexity, thus reducing the huge excess of non-cross-linked and so-called ‘mono-link’ peptides in order to
facilitate the subsequent analysis by LC-MS. Furthermore we applied our method to more complex samples to demonstrate the
possibilities of this approach.
Conclusion: Our novel ChaFRADIC approach allows the effective enrichment of cross-linked peptides from low and medium
complex samples and may open new avenues for analyzing protein complexes and protein-protein interaction in the future.
References: (1) Venne, A. S., Vogtle, F. N., Meisinger, C., Sickmann, A., Zahedi, R. P., Novel highly sensitive, specific, and
straightforward strategy for comprehensive N-terminal proteomics reveals unknown substrates of the mitochondrial peptidase
Icp55, J Proteome Res, 12, 2013
P236
CASP 11: A blind study towards CLMS as a structure determination tool
1
2
1
2
1
M. Mendes , A. Belsom , M. Schneider , L. Fischer , O. Brock , J. Rappsilber
1
TU Berlin, Berlin, Germany
2
University of Edinburgh, Edinburgh, Germany
1,2
Introduction: Cross-linking combined with mass spectrometry (CLMS) has become a powerful technique to determine proteinprotein interactions, protein conformational changes and low-resolution structure determination. In our lab, we were able to
Proteomic Forum 2015
207
determine the structure of human serum albumin (HSA), in its native environment, with a resolution of up to 3.8 Å deviation
th
from the X-Ray structure. To further test our approach, the lab embarked on the 11 CASP experiment. In a blind study, we
provided cross-linking data for four proteins of unpublished structure. 14 laboratories worldwide used the data to predict the
structure of these proteins.
Objectives: The objective of this work was to blind test the current value of cross-linking data in guiding high-resolution
prediction of protein structures.
Materials & Methods: Cross-linking and in gel digestion
Proteins were cross-linked with sulfosuccinimidyl 4,4’-azipentanoate (sulfo-SDA) using different protein/cross-linker ratios
(w/w). Cross-linked proteins were separated by SDS PAGE and the bands corresponding to cross-linked proteins were in gel
digested using trypsin. Peptides were desalted using C18-StageTips and analysed by mass spectrometry.
Mass spectrometry and data analysis
Peptides were separated by reverse-phase chromatography and directly sprayed into an LTQ-Orbitrap Velos (Thermo Fisher
Scientific). Acquisition was performed using a high-high resolution strategy. The 8 most intense peaks of each cycle were
selected for fragmentation in the LTQ and both precursor and fragmented ions were detected in the Orbitrap mass analyser.
Charge states +1 and +2 were excluded.
Identification of cross-linked peptides was performed using the in-house software, Xi (ERI, Edinburgh), with an FDR at the link
level of 5%.
Results: We identified around 300 links for each of the cross-linked proteins. The results of CASP11 confirmed that CLMS data
can assist de novo protein folding. In addition, CASP11 revealed to us the current limitations of our methodology. We can now
work on the optimisation of the methodology and transform CLMS into a routine tool for determining high-resolution protein
models.
Conclusion: CLMS already helped to predict low-resolution protein structures. In our lab we optimized the technology and
predicted a high-resolution structure for HSA in its native environment. CASP11 confirmed this success in a blind study and also
showed us the current limitations of our methodology. We will now work on these limitations to transform CLMS into a
structure determination tool.
P237
Protein structure determination in biological matrices using 3D proteomics and computational biology
1
2
1
2
1,3
A. Belsom , M. Schneider , L. Fischer , O. Brock , J. Rappsilber
1
University of Edinburgh, Wellcome Trust Centre for Cell Biology, Edinburgh, Germany
2
Technische Universität Berlin, Robotics and Biology Laboratory, Berlin, Germany
3
Technische Universität Berlin, Department of Biotechnology, Berlin, Germany
Introduction: Cross-linking coupled with mass spectrometry (CLMS) has now advanced as a powerful method for yielding lowresolution data regarding protein-protein interactions and protein tertiary structure. This has been compounded with recent
successes from our lab with cross-linking analyses of Ndc80c, Pol II and the complement system. However, the low resolution of
cross-linking has so far precluded its use in determining structures de novo.
Objectives: We have employed the use of highly reactive and promiscuous cross-linking reagents, which allow a much higher
number of cross-links to be identified across a protein. We have capitalised on this high-resolution cross-linking data, by
combining it with conformational space search, allowing us to investigate the structure of proteins in their native environment.
Methods: Reaction conditions for cross-linking human serum albumin (HSA) using the photo-activatable cross-linker
sulfosuccinimidyl 4,4’-azipentanoate (sulfo-SDA) were optimised. Mixing ratios of SDA to HSA were determined using different
protein-to-cross-linker weight-to-weight ratios. HSA was mixed with sulfo-SDA in cross-linking buffer to initiate lysine reaction
with the NHS ester component of the cross-linker. The diazirine group was then activated using UV irradiation. The resulting
cross-linked mixture was separated using an SDS-PAGE gel, bands were excised and the proteins reduced, alkylated and digested
Proteomic Forum 2015
208
using trypsin following standard protocols. Cross-linked peptides were fractionated using SCX-StageTips and desalted using C-18
StageTips prior to MS analysis by ES-LCMS using an LTQ-Orbitrap and identification of cross-links using Xi.
Results: The advantage of using the heterobifunctional cross-linker sulfo-SDA, which contains only one NHS-active ester, is that
only one lysine is required for reaction. In contrast, homobifunctional NHS-ester based BS3 requires the availability of two lysine
residues that are <11.4 Å apart (the maximal length of the BS3 spacer). Once lysine has reacted with the NHS-ester of sulfo-SDA,
the secondary photoactive functionality is activated by UV irradiation. The highly reactive carbene generated inserts
indiscriminately into any amino acid. The result is the creation of a much richer internal network map of residue proximities, in
comparison with the result obtained with utilisation of the homobifunctional cross-linker BS3.
Conclusion: Combining high-resolution cross-linking with conformational space search in the context of human serum led to
production of an albumin domain model, with RMSD to crystal structure of 3.38 Å. Combining CLMS with computational biology
is emerging as an exciting new approach to study the structure of proteins in their native environments.
Figure 1: Sulfo-SDA cross-links shown in the crystal structure of HSA.
Figure 2: Fragmentation spectrum of a cross-linked peptide pair.
Figure 1
Figure 2
Proteomic Forum 2015
209
P238
Quantitative cross-linking/mass spectrometry illuminates protein conformational changes that drive activation of the
complement
1
1
2
1
1
3
1,4
Z. Chen , L. Fischer , J. Cox , S. Tahir , J.- C. Bukowski-Wills , P. Barlow , J. Rappsilber
1
University of Edinburgh, Wellcome Trust Centre for Cell Biology, Edinburgh, United Kingdom
2
Max-Planck Institute of Biochemistry, Computational Systems Biochemistry, Martinsried, Germany
3
University of Edinburgh, Schools of Chemistry and Biological Sciences, Edinburgh, United Kingdom
4
Technische Universität Berlin, Department of Bioanalytics, Institute of Biotechnology, Berlin, United Kingdom
Introduction: Quantitative cross-linking/mass spectrometry (QCLMS) using isotope labelled cross-linkers is emerging as a new
1,2
strategy to study protein conformational changes in solution . However its application is limited by both experimental
2
challenges and lacking of computational tools . Here we developed a workflow to address these challenges. The workflow was
3, 4
tested using the complement proteins C3 and its active form C3b as a structurally well-described reference system . We then
used QCLMS to interrogate the unknown structure of another active form of C3, iC3, a key player in complement activation.
Objective: To provide a tool for studying protein conformational changes in solution, we developed a workflow for QCLMS
analysis, which



presents a code of good practice for QCLMS analysis;
is well evaluated;
is ready and easy to be followed by other researchers
Methods and Materials: To assess the abilities of QCLMS to reflect conformation changes, purified C3 and C3b were separately
cross-linked in solution using cross-linker (BS3) and its deuterated analogue BS3-d4 in four separated protein-cross-linker
combinations. After trypsin digestion, a 1:1 mixture of C3+BS3 and C3b+BS3-d4 digests and a label-swap replica (C3+BS3-d4 and
C3b+BS3 digests) were analyzed by LC-MS/MS. Cross-linked peptides were identified by database searching and then quantified
based on their MS signals (Fig.1). Quantitation for cross-linked peptides was automated by using Pinpoint (Thermo) and
5
MaxQuant .
To interrogate the unknown structure of iC3, it was compared pair-wise against both C3 and C3b. The QCLMS analysis for both
protein combinations (iC3 vs. C3 and iC3 vs. C3b) was conducted as described above.
Results: QCLMS revealed differences and similarities between structures of C3 and C3b.



QCLMS results highly agree with the C3-to-C3b transition as revealed by their crystal structures (Fig.2).
Comparing QCLMS data with crystal structures revealed the relationships between conformational changes and crosslink yields.
QCLMS analysis revealed C3-like, C3b-like and iC3-unique features in the iC3 structure and results in a model for iC3
domain architecture.
Conclusion: As demonstrated and evaluated in the benchmark study, our QCLMS workflow can provide reliable quantitation for
cross-link data, correctly revealing protein conformational changes. The automated quantitation process lowers the barrier to
entry for any research group wishing to explore this technique. More important, it means QCLMS is now feasible for studying
conformational changes in large protein assemblies or complex protein network.
References
1. Fischer, L. et al. Journal of proteomics 88, 120-128 (2013)
2. Schmidt, C. et al. Nature communications 4, 1985 (2013)
3. Janssen, B.J. et al. Nature 444, 213-216 (2006)
4. Janssen, B.J. et al. Nature 437, 505-511 (2005)
5. Cox, J. & Mann, M. Nat Biotechnol 26, 1367-1372 (2008)
Proteomic Forum 2015
210
Figure 2
Figure 1
Subcellular Proteomics (P239–P251)
P239
Hitting the target: Novel reagents for the chemical-proteomics based identification of vascular accessible biomarkers
1,2,3
1,2,3
4
1,2
S. Hanke , A. Kerner , Y. Zhang , C. Rösli
1
Deutsches Krebsforschungszentrum (DKFZ), Junior Research Group Biomarker Discovery, Heidelberg, Germany
2
HI-STEM gGmbH, Biomarker Discovery, Heidelberg, Germany
3
Helmholtz Int. Grad. School for Cancer Research, Heidelberg, Germany
4
B CUBE Center for Molecular Bioengineering, Dresden, Germany
Introduction: A promising approach for the development of novel therapeutics with fewer side effects in healthy tissues is the
targeted delivery of bioactive molecules directly to the site of disease. Therefore, vascular accessible, tissue specific target
proteins have to be identified. To analyze the vascular accessible proteome by MS-based approaches, the respective proteins
(e.g. membrane proteins, proteins of the extracellular matrix) have to be specifically enriched.
Objectives: This project focuses on the development and multi-step validation of two novel peptide-based biotinylation
reagents with high reactivity and selectivity for vascular accessible proteins.
Proteomic Forum 2015
211
Material and Methods. Two novel peptide-based biotinylation reagents were synthesized and evaluated. Their reactivity was
tested by BSA coupling in different ratios and analysis with linear MALDI MS. Biotinylation success on HeLa cells was assessed by
FACS analysis and confocal laser scanning microscopy (CLSM). Biotinylated proteins were captured on Streptavidin-Sepharose
and analyzed following tryptic digest by LC-MALDI-MS. Perivascular biotinylation in different tissues was demonstrated by in vivo
perfusion of mice and analysis of CD31-Biotin costaining by CLSM. Mass spectrometric analysis of the enriched surface proteome
fraction from kidney tissue was performed in comparison to the commercial reagents Sulfo-NHS-LC-Biotin and NHS-PEG12-Biotin.
Results. An increase in size and polarity decreases the reagent’s membrane crossing potential. Two novel multiple charged,
peptide-based reagents were synthesized. Activity tests on BSA revealed increased reactivity by site-specific NHS-activation for
protein coupling. Biotinylation efficacy on HeLa cells was demonstrated by FACS analysis, reaction with membrane proteins
proven by CLSM. In total, 1093 proteins could be quantified with at least 2 proteotypic peptides. The surface biotinylation
approach led to a significant decrease of high abundant, intracellular proteins, whereas the identification rate of low-abundant
surface protein was increased. In vivo perfusion of mice and analysis of CD31-Biotin costaining by CLSM demonstrated efficient
biotinylation of proteins on vascular endothelial cells and in the subendothelial extracellular matrix in different tissues (Figure:
kidney tissue perfused with novel peptide reagent, blue: nuclei, red: blood vessels, green: biotinylated proteins). Mass
spectrometric data was evaluated by label-free quantification using the in house developed software MS QBAT. The data
confirms enrichment of vascular accessible proteins.
Conclusion. Two novel biotinylation reagents for biomarker discovery were designed and synthesized. Validation demonstrates
selective enrichment of the vascular accessible proteome.
Figure 1
P240
Membrane microdomain dynamics in P. falciparum infected erythrocytes: proteome, lipidome and interactome.
1
1
2
1
F. FRATINI , G. Sferra , C. Ferreri , M. Ponzi , E. Pizzi
1
ISS, MIPI, ROME, Italy
2
CNR ISOF , Bologna, Italy
1
Introduction: Erythrocyte cholesterol-rich membrane microdomains are largely reorganized upon Plasmodium falciparum
infection. These specialized membrane compartments, called detergent resistant membranes (DRMs), are extremely dynamic
and play a key role in parasite development and infection success.
Objectives: To characterize DRMs at different time-points of the parasite erythrocytic life-cycle (asexual and sexual stages) by
identifying and relatively quantifying associated proteins and lipid/fatty acids. Further, to gain insights on protein-protein
interactions, which occur in these specific microdomains, by reconstructing a probabilistic global interactome of P. falciparum.
Methods: We obtained DRMs fractions by sucrose gradient centrifugation. We then extracted both proteins and lipids from 7
fractions collected along the flotation profile. Lipid and fatty acids were characterized by TLC, and GC-MS. Proteins were
identified by LC/MS-MS and quantified using the Top3 method generating for each protein an abundance profile (PAP). Stagespecific PAPs were submitted to cluster analysis for comparative purposes. The global interactome of P. falciparum was
constructed using genomic, transcriptomic and proteomic data applying a Bayesian approach.
Proteomic Forum 2015
212
Results: Clustering of PAPs provides a sort of barcode characteristic for each analysed developmental stage. Furthermore, we
mapped DRM associated protein components on the predicted global interactome of P. falciparum and the resulting subnetworks were dynamically analysed.
Conclusions: This work provides insight on lipid and protein composition of DRMs in P. falciparum and their dynamics during the
intraerythrocytic parasite development.
P241
Reshaping of the autophagosomal proteome by influenza A virus
1
1
1
1
J. Dengjel , A. Becker , C. Muenz , M. Gannage
1
University of Freiburg, Dep. of Dermatology, ZBSA, Freiburg, Switzerland
Autophagy is responsible for lysosomal degradation of whole organelles and protein complexes. It also plays an important role in
anti-viral or -microbial defense mechanisms. It has been shown that influenza A virus interferes with autophagosomal
degradation. To better understand influenza A virus-autophagy crosstalk and elucidate potential new anti-viral therapies
targeting autophagy we selectively monitored changes in the cellular protein composition as well as in the composition of
autophagosomes during influenza A infection. Using imaging and protein biochemistry approaches we were able to show an
accumulation of autophagosomes in infected cells compared to control cells and a block of functional autophagy. We also
identified viral proteins inside autophagosomes. Interestingly, viral infection had global implications on the autophagosomal
protein composition. Next to viral proteins, an accumulation/depletion of various cellular proteins inside autophagosomes could
be detected. Thus, it appears that the virus reshapes the organellar proteome to create a viral permissive environment by
selective targeting/exclusion of protein complexes from autophagosomes. Currently, we analyze the virus-induced targeting of
host proteins into autophagosomes.
P242
Identification of six cell surface proteins for specific liver targeting
1
1
2
1
S. Kux van Geijtenbeek , P. Cutler , S. Simon , A. Ducret
1
F. Hoffmann-La Roche Ltd, Roche Innovation Center Basel, Basel, Switzerland
2
F. Hoffmann-La Roche Ltd, Drug Disposition and Safety, Basel, Switzerland
Introduction: Cell surface proteins are the primary means for a cell to sense and interact with its environment and their
dysregulation has been linked to numerous diseases. In particular, the identification of proteins specific to a single tissue type or
to a given disease phenotype may enable the targeting of therapeutics to an organ of interest. We tested here the feasibility of a
cell surface proteomics approach to identify pertinent markers directly in a clinically relevant tissue. We use a strategy to
chemically tag and immunocapture cell-surface glycosylated proteins[1] at a very high level of specificity to the primary human
Small Airways Epithelial Cell (SAEC) line, the primary human Lung Fibrobasts (LuFi) cell line, and tissue-derived human primary
hepatocytes to identify lineage-specific cell surface receptors.
7
Method: Cell lines (3x10 cells per condition) were grown to confluence using their respective growth culture medium. Human
primary hepatocytes were purchased from Abcellulate tissue bank (UK) and were plated for 2 h prior to labeling. The
biotinylation of the glycoproteins with hydrazide-biocytin, crude membrane preparation, protein tryptic digestion, and
glycocopeptide enrichment using streptavidin beads were performed according to Wollscheid’s protocol [1] with few
adjustments. Flow-throughs and eluates were analyzed by tandem mass spectrometry using a nanoLC/LTQ-Orbitrap mass
spectrometer (Thermo Scientific, Germany). The MS/MS data were processed using Sequest V.27 (UniProt human database,
70’212 forward and reverse entries)
and the protein scoring algorithm Clotho [2] using a protFDR < 1%. Candidate proteins were further validated bioinformatically
taking into account protein structure (extracellular domain, transmembrane domain or membrane anchors) and mRNA
expression across all human tissue.
Results: The glycopetides’ labeling and isolation was identically performed for the SAEC, LuFi and the human hepatocytes.
Approximately 1050-1150 different proteins were identified from each experiment (66-75% reproducibility within replicate),
including 116-163 glycoproteins. Glycoproteins were defined as a protein with a glycopeptide showing the “Asn-Xxx-Ser/Thr”
motif whereas the Asn residue was specifically converted to Asp due to PGnase treatment.
Proteomic Forum 2015
213
The selection for lineage-specific cell surface candidates was performed in several steps. Using primary lung epithelial cell
cultures as negative controls, we identified 32 hepatocyte-specific cell surface proteins candidates. We used mRNA expression
to select six markers that may provide adequate specificity for therapeutic intervention in the liver.
Conclusions: This result clearly emphasizes the importance of direct analysis in a clinically relevant tissue for cell surface markers
combined with meaningful additional bioinformatics.
1) Wollscheid B. et al. Nat Biotech. 2009, 378-386
2) Berntenis N. et al. HUPO conference 2009, poster C522
P243
Secretome and proteome analysis of monocyte-derived dendritic cells stimulated with the major birch pollen allergen Bet v 1
1
1
1
1
L. Strasser , C. Ramsauer , J. Horejs-Hoeck , C. Huber
1
University of Salzburg, Department of Molecular Biology, Salzburg, Austria
Dendritic cells (DCs) play a key role as antigen presenting cells in the immune system. Thus, allergy research is interested in their
proteome and secretome to obtain knowledge of the immunological mechanisms involved in allergic responses. Especially
secreted proteins of dendritic cells are essential for cell-cell communication with other cells of the immune system. We prepared
monocyte-derived dendritic cells (moDCs) from human blood for stimulation experiments under serum-free cell culture
conditions employing recombinant Bet v 1, the major allergen of birch pollen and BM4, a hypoallergenic variant thereof. After
protein extraction and tryptic digestion, cell lysates were analyzed via ion-pair reversed-phase (IP-RP) high-performance liquid
chromatography (HPLC) at pH 2 followed by identification using quadrupole-Orbitrap mass spectrometry (MS). For supernatants,
tryptic peptides were fractionated in two dimensions at pH 10 applying RP-HPLC and IP-RP-HPLC at pH 2 before MS-analysis.
Data analysis revealed high variation between the blood-donors. Up to 1900 protein groups were identified in the differentially
stimulated cell lysate samples. For example, CD44 antigen and MAP kinase-activated protein kinase 3 were detected after
stimulation with BM4. Additionally, proteins from the MHC-II complex of the antigen presentation pathway were revealed with
ingenuity pathway analysis (IPA). High variations within protein group identifications were also observed in the supernatants.
Nevertheless, among 563 protein groups identified in the supernatant, immune-relevant proteins like N-sulfotransferase 1 or
SPARC were clearly found to be upregulated after stimulation. The combination both of cellular proteome and secretome
analysis is shown to be a viable tool for studying the response of dendritic cells to stimulation by allergens.
P244
A Plasma Membrane Proteomic Analysis Of
Mouse And Human Cardiovascular Proteins
1
A. Gramolini
1
University of Toronto, Physiology, Toronto, Canada
Introduction: Analyzing cardiac membrane proteins is crucial to a better understanding of heart function and development. A
clear understanding of these proteins is crucial when we consider disturbances in conduction from malfunctioning membrane
channels lead to several human pathologies that result in cardiac arrhythmias
Materials and Methods: We employed cationic silica-bead coating coupled with shotgun proteomics to enrich for and identify
cell-surface associated proteins from primary mouse neonatal and human fetal ventricular cardiomyocytes, endothelial cells,
and smooth muscle cells. Human coronary artery smooth muscle cells, human coronary artery endothelial cells, and human
cardiac muscle derived cardiomyocytes or fibroblasts were cultured for 2-3 weeks. Freshly dissociated human fetal ventricular
myocytes were acutely dissociated and analyzed. Membrane proteins were cross-linked to cationic silica beads to isolate
cytosolic proteins and a membrane fraction attached to the beads. Samples were analyzed by LC-MS MuDPIT strategies on a
Thermo LTQ or LTQ Orbitrap.
Results: Shotgun proteomics identified >3,000 mouse proteins and >2,500 human proteins. Enrichment of membrane protein
and depletion of cytosolic proteins was confirmed by immunoblotting. Mapping of orthologous proteins between mouse and
human resulted in 1717 proteins. Focussing on the cardiomyocytes, QSpec statistical analysis calculated differential spectral
counts between proteins found in the membrane enriched and membrane depleted fraction and provided a dataset of 555
cardiomyocytes proteins including many known membrane proteins. Bioinformatic integration with transmembrane helix
Proteomic Forum 2015
214
predictions, Phenotype Ontology (PO), and publically available microarray data sets, identified a rank ordered set of cardiacenriched surface proteins; select examples of which the subcellular location were further confirmed using high resolution
confocal microscopy, immunogold electron microscopy, and sucrose density gradient biochemistry. For several of the highly
2+
ranked membrane proteins, lentiviral-based shRNA knock-down demonstrated significantly altered Ca transient amplitudes,
2+
2+
Ca release rates, and Ca uptake rates in both neonatal mouse cardiomyocytes and in human embryonic stem cell derived
cardiomyocytes. Initial knockouts in Danio rerio (zebrafish) support the cultured myocyte studies.
Conclusion: We have provided the first comprehensive analysis of membrane cell surface-associated proteins in the cardiac cell,
and provided a pipeline to validate surface proteins in cardiomyocytes that might be involved in excitation-contraction.
P245
Protein determination in RNAlater fixed cells returned from the ISS
1
1
2
3
3
3
4
1
J. Pietsch , S. Riwaldt , A. Sickmann , J. Segerer , A. Schwarzwälder , C. Bruderrek , J. Bauer , M. Wehland-von Trebra , M.
1
1,5
Infanger , D. Grimm
1
University Clinic Magdeburg, Clinic for Plastic, Aesthetic and Hand Surgery, Magdeburg, Denmark
2
Leibniz-Institut for analytic sciences, system analysis, Dortmund, Germany
3
Airbus Defense and Space GmbH, Friedrichshafen, Germany
4
Max Planck Institute of Biochemistry, Martinsried, Germany
5
Arhus University, Institute of Biomedicine, Arhus, Denmark
Space travelers face many different health problems and various diseases observed in astronauts so far seem to be based on
cellular dysfunction. To elucidate the mechanisms, cells are exposed to microgravity (µg) for varying time periods. Alteration of
gene expression patterns and protein content as well as the transition from a 2 to a 3 dimensional growth were observed until
now in comparison to control cells being incubated under normal gravity (1g) and laboratory conditions. There are several ways
of exposing cells to µg. One indispensable way is sending cells into the Earth orbit for example to the ISS thus removing the
Earth gravity forces almost completely (real-µg). However, opportunities of transporting cells to the ISS or to Space are rare and
expensive. They require special equipment, which are very small sized, light weighty and capable to perform an automatic cell
conservation during the flight. Due to the rarity of Spaceflights, each single experiment should be exploited in a way which
enables the experimenter to obtain as much information as possible. For this purpose, RNAlater fixation proved to be a
reasonable way of cell conservation. Cells fixed in Space by RNAlater and returned to Earth several days later, provided
information about the cells’ growth behavior and could successfully be used to analyze gene expression patterns. In this study,
we isolated the proteins in parallel to the RNA out of RNAlater fixed thyroid cancer cells (FTC-133). We then applied Mass
Spectroscopy to identify the proteins, because we aimed to identify even unforeseeable proteins which are differently regulated
in real-µg.
180 different types of proteins were found. They include rather abundant housekeeping proteins of glycolytic enzymes,
cytoskeletal and heat shock proteins, but also a few proteins such as the inter-alpha-trypsin inhibitor heavy chain H1 (ITIH1) or
the cartilage oligomeric matrix protein (COMP), an anti-apoptotic caspase 3 inhibitor, which have not been seen before in
comprehensive proteome analyses of thyroid cancer cells. 180 types of proteins is a small number considering the sensitivity of
mass spectrometry. Therefore, we reason that the RNAlater fixation may not be the fixation method of choice for protein
analysis but offers a possibility to analyze at least proteins of rather high abundance in parallel to RNA in rare samples. Our
analysis points to the possibility that the protein types detected in the Space samples but not in earlier samples of thyroid cells
have been highly up-regulated during this specific way of culturing.
P246
Functional plasticity of cardiac interfibrillary mitochondria (IFM) as response mechanism to stress
1
1
2
3
4
4
5
J. Heidler , I. Wittig , D. Henze , C. Krüger , S. Kostin , T. Braun , M. Szibor
1
Goethe University Medical School, Functional Proteomics, Frankfurt am Main, Germany
2
Henze & Partner PartG, Praxis für Anästhesiologie, Halle/Saale, Germany
3
Medical Affairs Fa. Cellpharm, Bad Vilbel, Germany
4
Max Planck Institute for Heart and Lung Research, Dept. I Cardiac Development and Remodeling, Bad Nauheim, Germany
5
Research Program of Molecular Neurology, University of Helsinki, Helsinki, Finland
A morphological hallmark of the failing human heart is a devastative autophagic degradation of cellular structures starting from
the perinuclear region, proposed to
Proteomic Forum 2015
215
actively shift the heart into a decompensated state [1]. We studied heart samples from three species, i.e. a mouse model of
cardiac specific expression of MCP1 that autonomously develops heart failure [2], hibernating Syrian hamsters [3] and a pig
model of mitochondrial dysfunction exposed to hyperbaric oxygen. Our data reveal an age-dependent increase of perinuclear
degradation in mouse hearts that occurred prior to the onset of cardiac dysfunction. These center core-like lesions in the
myofibrillar compartment are most likely the end-stage result of a vicious cycle that starts with a physiological response to
lowered levels of cardiac workload. Accordingly we found that in hibernating Syrian hamsters under conditions of depressed
metabolism interfibrillary mitochondria are reversibly silenced whilst subsarcolemmal mitochondria remain more active. With
blue native electrophoresis (BNE) we analyzed the hearts of physical inactive pigs and found an isolated impairment of the
interfibrillary compartment due to a non-functional assembly of the respiratory chain that can be fully re-activated upon
treatment of the pigs with hyperbaric oxygen. For a deeper understanding of the molecular mechanism behind we performed
complexome profiling [4] to analyze the composition of large mitochondrial complexes in these mitochondrial sub-populations.
Comparing eight complexome profiles we identified proteins and protein complexes that show dynamic expression and
assembly in these conditions.
We conclude from our studies that differential compartment regulation by switching the activity status of mitochondrial subpopulations from on to off and vice versa might provide a hitherto unnoticed flexible on-demand plasticity in cardiomyocytes.
Such alterations make proper myofibril contraction in the silenced compartment unlikely. Silenced mitochondria can be reactivated on demand. Only long-lasting mitochondrial silencing, e.g. upon chronic cardiac overload, might increase the risk of
adverse cardiomyocyte remodeling.
[1]Hein S, Arnon E, Kostin S, Schönburg M, Elsässer A, Polyakova V, Bauer EP, Klövekorn W-P,Schaper J (2003) Circulation 107:
984-991
[2]Kolattukudy PE, Quach T, Bergese S, Breckenridge S, Hensley J, Altschuld R, Gordillo G, Klenotic S, Orosz C, Parker-Thornburg J
(1998) Am J Pathol 152: 101-111
[3]Toole L, Belai A, Shochina M, Burnstock G (1999) Cell Tissue Res 296: 479-487
[4]Heide H,Bleier L,Steger M,Ackermann J,Dröse S,Schwamb B,Zörnig M,Reichert AS,Koch I,Wittig I,Brandt U(2012) Cell Metab
16(4):538-49
P247
Redox complexome profiling: a method to determine the redox dependent dynamics of protein-protein interactions in a
whole organelle.
1,2
3
2
1
2,4
2,3
I. Wittig , E. Bonke , L. Bleier , F. Richter , U. Brandt , S. Dröse
1
Goethe University, Medical School, Functional Proteomics, Frankfurt am Main, Germany
2
Goethe-University, Molecular Bioenergetics Group, Frankfurt am Main, Germany
3
Goethe-University, Department of Anesthesiology, Intensive-Care Medicine and Pain Therapy, Frankfurt am Main, Germany
4
Radboud University Medical Centre, Nijmegen Centre for Mitochondrial Disorders, Nijmegen, The Netherlands, Netherlands
We recently introduced the method complexome profiling that combines classical blue-native electrophoresis with modern
proteomic methods of quantitative mass spectrometry and hierarchical clustering to investigate the composition of
macromolecular complexes. The method so far enabled the identification of unknown subunits and the dynamic assembly of
mitochondrial protein complexes. Here we extend complexome profiling to determine the thiol redox state of mitochondrial
protein complexes upon induction of ROS production from respiratory chain complexes I and III. In organello thiol labelling with
NEM/d5-NEM enabled the quantitative identification of redox sensitive targets of isolated mouse heart mitochondria in their
physiological environment and native protein conformation. After solubilization of the mitochondria with the mild detergent
digitonin and advanced complexome profiling we were able to analyse the thiol redox states within native protein complexes,
supercomplexes and single proteins. In summary redox complexome profiling can serve as a powerful bottom-up approach to
identify redox dependent dynamics of protein-protein interactions and redox sensitive sites.
Proteomic Forum 2015
216
P248
Hydrazide-based protein-centered glycoproteomics for the analysis of LPS-mediated membrane protein expression changes in
THP-1 cells
1
1
1
2
M. Müller , R. Lehmann , T. Conrad , D. Driesch , H. Slevogt
1
UK Jena, ZIK Septomics, Jena, Germany
2
BioControl Jena, Jena, Germany
1
Introduction: Exposure of human monocytes to lipopolysaccharide (LPS) induces temporary monocytic insensitivity to
subsequent LPS challenge. The underlying mechanism of this tolerance could have important clinical impact in preventing
and/or treating severe infections like sepsis. To date, the only marker characteristic for non-responsive monocytes is downregulated HLA-DR expression. A “tolerant” signature of plasma membrane spanning cell surface receptors would be helpful to
classify patients monocyte functionality or tolerance in severe infections and sepsis.
Objective: Analysis of LPS-induced membrane protein expression changes in monocytic THP-1 cells to test a label-free
Hydrazide-based protein-centered glycoproteomic approach for performance and robustness.
Methods: THP-1 cells were stimulated with LPS for 4h, 24h and 48h. Cells were solubilized by SDS and glycoproteins enriched by
Hydrazide-based chemistry after oxidation of glycans. Tryptic and PNGaseF released peptides were identified by LC-MS/MS on
an Orbitrap Fusion and quantified.
Results: We identified a total of 2681 integral membrane proteins (GO:31224). 421 membrane spanning proteins showed an
expression change > 2 after LPS treatment. Subcellular localization analysis revealed no organelle-specific response to LPS,
glycoproteins from all membrane compartments within the cell and the plasma membrane were affected.
Conclusion: Hydrazide-based enrichment of glycoproteins from cell lines is a powerful approach to study expression changes of
membrane proteins. When used for the identification and quantification of human primary monocytes the findings could be of
significance to reveal a tolerance-specific receptor expression signature and in understanding the mechanism of LPS tolerance.
Newly identified plasma membrane receptors of tolerant cells will provide values for designing new approaches for regulating
monocyte functionality in patients with immunosuppression in sepsis.
P249
The value of subcellular fractionation in ProteoGenomics Studies
1
1
1
1
1
1
1
P. Díez , M. Gonzalez-Muñoz , N. Ibarrola , R. M. Degano , C. Droste , A. Orfao , J. De Las Rivas , M. Fuentes
1
Cancer Research Center, Proteomics Unit, Salamanca, Spain
1
After having deciphered the human genome, the characterization of the whole proteome is the next main challenge. Proteins
vary between samples, cellular compartments or mutational stages, making difficult its study. Furthermore, post-translational
modifications, such as oxidations, acetylations or methylations, change the properties, not being able to be considered as the
same protein. Here, it is proposed an extensive comparison between 7 subcelular fractionation studies and total extraction
protocols for deep proteome characterization of a human Burkit´s Lymphoma cell line. Our studies provide information about >
7000 human proteins specificically identified by nanoLC-MS/MS assays. In adddition, a strong correlation with transcriptomics
assays performed in Affymetrix platform.
P250
Proteomics of alternatively activated macrophages reveals novel signalling pathway downstream of Scavenger Receptor A
1
1
1
1
2
M. Guo , M. Gierliński , A. Prescott , B. Dill , D. Russell , M. Trost
1
University of Dundee, MRC PPU, Dundee, United Kingdom
2
Cornell University, Ithaca, New York, United States
1
Macrophages activated by IL4 or IL13 enter a state called alternative activation which is thought to be important during wound
healing and helminth infection. We performed a thorough proteomics analysis of the total cell proteome and phagosome
proteome of alternatively activated (IL4-treated) bone marrow-derived macrophages (BMDMs). Our data indicates that
alternative activation leads to phagosomal recruitment of proteins in favour of apoptotic cell clearance, enhanced fusion with
early endosomes and the lysosome as well as with parts of the endoplasmic reticulum (ER). Furthermore, we show that the
Proteomic Forum 2015
217
TAK1/MKK7/JNK complex is specifically recruited to phagosomes from AAMΦs by increased K63-polyubiquitylation. We
identified many polyubiquitylated phagosomal proteins, including macrophage scavenger receptor 1 (MSR1/SRA), whose
interaction with the TAK1/MKK7/JNK complex we validated. When we inhibited pharmacologically JNK, phagosome fusion with
the ER was decreased reducing translocation of ER-resident lipases, indicating a role of MSR1 ubiquitylation and
TAK1/MKK7/JNK signalling to regulate lipid metabolism on the phagosome.
P251
A large-scale approach to study mitochondrial protein complexes of Arabidopsis thaliana
1
1
S. Martinez Jaime , M. Gorka , A. Graf
1
MPIMP, Golm-Potsdam, Germany
1
Introduction: Proteins are involved in all aspects of live from enzymatic reactions to structural support. However, most enzymes
are not active as monomer and most process in the cell are performed by multimeric protein complexes. So far, the most
common approaches to study protein-protein interactions (PPIs) are biochemical purification followed by mass spectrometry,
genetic engineering of cellular systems or microscopy techniques. These approaches have two disadvantages: they do not allow
high throughput analysis (low number of PPIs can be studied per experiment), and a lack of quantitative data.
Objectives: To avoid these shortfalls we use a large-scale approach to identify and quantify protein complexes. We focus our
work on the mitochondrial proteome of Arabidopsis thaliana. The aim of the project is to compare the abundance and
composition of organellar protein complexes under different environmental conditions or following in vitro treatments of
isolated mitochondria. Conditions and treatments will be selected such that significant adjustment of metabolic fluxes in
mitochondria can be expected. These experiments should provide insights into the regulation of plant metabolism on the level
of protein complexes.
Materials & Methods: To achieve that purpose, native protein extracts of isolated and fully functional mitochondria are
fractionated using native PAGE. Following electrophoresis the gel is cut in 24 fractions and in-gel digestion is performed to elute
peptides. Samples are analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS) on a Q-Exactive (Thermo
Scientific) connected to a Easy n-LC II or 1000 (Thermo Scientific). Raw data is analyzed by MaxQuant (Max Planck Society) and
subcellular localization of identified proteins is obtained from the SUBA3 database (Tanz et al., (2013). Nucleic Acids Res. 41:
D1185-91).Protein complex formation is studied using in house created R scripts.
Results: In an initial “proof of concept” study, quantitative data on 756 mitochondrial proteins was obtained. This equals
coverage of 56% of the A. thaliana mitochondrial proteome. By comparing the theoretical molecular weight of the quantified
proteins with their distribution profile in the native gel we calculated that approximately 60% these proteins are part of protein
complexes. The data also reveals a clear tendency for the interaction between enzymes which catalyze neighboring reactions in
metabolic pathways.
Conclusion: These encouraging results are the basis for future studies to investigate changes in protein complex composition
and/or abundance under different environmental conditions. Furthermore, we believe that our methods can be used to reveal
putative PPIs on a proteome-wide scale. Candidate PPIs can be verified by standard methods like CoIP, Y2H or BiFC.
Proteomic Forum 2015
A
Abbondio, M.
Abdollahi, A.
Aberger, F.
Abouseada, N.
Acosta-Martin, A.
Acuña-Sornoza, R.
Addis, M. F.
Aebersold, R.
Aerts, J.
Ahmad Rusmili, M. R.
Ahrens, M.
Alberio, T.
Albers, C.
Alge-Priglinger, C.
Allmer, J.
Alloggio, I.
Alruwaili, J.
Altmann, F.
Amalric, F.
Amaral, L.
Amirmardfar, R.
Anders, U.
Anedda, R.
Ang, C.-S.
Antinori Malaspina, P.
Argentini, A.
Arner, P.
Arrey, T. N.
Asplund, A.
Assinger, A.
Assmann, N.
Atwal, G.
Auer, G.
auf dem Keller, U.
Auwerx, J.
Avramova-Nehmer, J.
Aylett, C. H. S.
P181
P180
P133
P083
P112
P188
P045, P046, P047, P181
P084, P085, P086
O27, P232, PL23, P056
P141
P186
O10
P204
P039
P184
P027
P079
P145
P115
O02
YIA2
P034
P146
P047
P210
P112
YIA3, O20
P002
P218
P159
P167
O09
P015
P139, P149
P120
P232
P152
O27
B
Bach, M.
Baerenfaller, K.
Baginsky, S.
Baker, A.
Baldetorp, B.
Baldin, C.
Ban, N.
Bantscheff, M.
Baran, P.
Barkovits, K.
Barlow, P.
Bartel, J.
Barth, J.
Bauer, J. W.
Baumann, C.
Bazzichi, L.
Beaumont, V.
Becher, D.
Beck, K.-F.
P040
P088
PL09
P035
O11
P068
O27
P159, NC9
P042
P153
P238
P081
P227, PL32
P136, P245
P127, NC7
P158
P187
P081
P107
Beck, M.
Beck, F.
Beck, J.
Beck, S.
Becker, S.
Becker, A.
Behne, A.
Behr, J.
Beine, B.
Beissbarth, T.
Bekesova, S.
Belsom, A.
Benama, M.
Bendixen, E.
Bennett, K.
Berchem, G.
Bergamini, G.
Berger, E.
Berger, M.
Bern, M.
Bernhardt, O.
Berny, D.
Bertile, F.
Beynon, R.
Bhangu-Uhlmann, A.
Bierczynska-Krzysik, A.
Bigenzahn, J.
Bilan, V.
Bindel, F.
Biosa, G.
Birgersson, E.
Blanc, S.
Blankenburg, S.
Blasius, B.
Blatnik, R.
Bleier, L.
Blennow, K.
Bleuler, S.
Blixt, E.
Bloomfield, N.
Bocanek, O.
Bock, T.
Bockmayr, M.
Bodenmiller, B.
Boege, F.
Boehringer, D.
Boese, J.-H.
Bogdanow, B.
Böhmert, L.
Bohnenberger, H.
Boldt, K.
Bonaldi, T.
Bonizzi, L.
Bonke, E.
Bonn, F.
Borchers, C.
Bourchookarn, A.
Bousquet-Dubouch, M.-P.
Bouwman, F.
218
P134
P162
P190
P226
P139, P149
P241
P025
P065
P053, O10
P168
P087
P236, P237
P032
PL21
O24
P213
NC9
P097
P178
P119
P106, P212, P214, P024
O06
P202
P141
P215
P042
O24
P104
P152
P084
HUPO 1
P202
P101, P102
P036
P132
P247, P059
P208
P024
P126
P057
P037
P134
P131
PL11
P124
O27
P159
P176
P155
P190
P004, O07, P160, O28, P125
P121
P079, O12, P195
P247
P081
HUPO 8, HUPO 7, PL28
P067
O02
P229
Proteomic Forum 2015
Bouyssie, D.
Brabant, G.
Braeuning, A.
Brakhage, A. A.
Brandl, C.
Brandt, U.
Braun, H.-P.
Braun, T.
Braunschweig, T.
Breitenbach, M.
Breitenbach-Koller, H.
Brock, O.
Broeker, C.
Bröker, B.
Bruch, H.-P.
Bruderer, R.
Bruderrek, C.
Brugiere, S.
Bruley, C.
Brummer, T.
Brune, W.
Bryan, K.
Bubis, J.
Büchler, R.
Bug, G.
Buhmann, J.
Bukowski-Wills, J.-C.
Burch, T.
Burel, A.
Burger, T.
Burkhart, J.
Burlet-Schiltz, O.
Busch, T.
Busch, D.
Butscheid, Y.
Ahrends, R.
Büttner, K.
Butzmann, L.
P032
P164
P155, P156
P068, P038
P136
P247, P059
P071, P023
P246
P139
P136
P136
P236, P237
O09
P049
P193
P106, P212, P214, P024
P245
O06
P032, O06
P151
O17
P125, O07
O26
P142, P143
P190
PL11
O04, P238
P035
P032
O06
O08, HUPO11
P032
P071
P203
P212
HUPO11
P081
P159
C
Cabanski, M.
Capkova, V.
Carapito, C.
Carrie, C.
Carroll, A.
Catena, R.
Cavola, A.
Cedersund, G.
Cermak, J.
Chait, B.
Chamrád, I.
Charwart, V.
Chávez Munguía, B.
Chen, J.-X.
Chen, L.-Y.
Chen, Z.
Chiocca, S.
Cho, W.
Christophoru, A.
Ciftci, C.
Cimermančič, P.
Cipriani, P. G.
P199
P078
P032
P080
PL08
PL11
O12
P157
P150
PL18
YIA5, P006
P192
P154
P013
P212
P238
P121
NC6
O16
P027
O27
P013
219
Ciregia, F.
Ciryam, P.
Claude, E.
Claussnitzer, M.
Clement, L.
Cloots, E.
Collins, P. W.
Collins, A.
Collins, B.
Combe, C.
Conrad, T.
Corradini, E.
Corso, J.
Cosemans, J. M.
Cox, J.
Croizat, P.
Cruz Mireles, N. d. J.
Cubeddu, T.
Cuccurullo, M.
Curien, G.
Curo Gutierrez, E. N.
Cutler, P.
P158
O18
P052, P055
P002
P014, O20
P012
O08
P015
P056
P054
P248
O01
P190
O08
P226, P238
P120
P082
P045, P046, P084
P204
O06
NC4
P242
D
Dahlman, I.
Damoc, E.
Daniels, D.
Darula, Z.
David, J.
de Boer, D.
De Ceuninck, L.
de Haas, P.
de Keijzer, J.
De Las Rivas, J.
de lera, A.
de Ru, A.
De Sutter, D.
Degano, R. M.
Deininger, W.
Dekker, N.
del Toro, N. .
Delanghe, B.
Deleidi, M.
Deligios, M.
Dell`Aica, M.
Demianova, Z.
Demmers, J.
Dengjel, J.
Ping, P.
Depping, R.
Dernayka, L.
Dettmer, K.
Deusch, S.
Dhople, V. M.
Dianes, J. A.
Dickhut, C.
Diedrich, B.
Diehl, H.
Dietrich, D.
Dietz, H.
Dieu, M.
Díez, P.
P002
P218
P172
P118, P119
P156
P229
P012
YIA2
YIA2
P249
P216
YIA2
P012
P249
P215
P141
P016
P218
P003
P084, P085
P111
P064
P108
P151, P241
HUPO 9, PL05
P193
O07, P125
O09
P069
P049, P062
P016
P111
P151
O10
P053
HUPO 10
P001
P249
Proteomic Forum 2015
Difilippantonio, M.
Dijon, S.
Dill, B.
Dirk, A.-L.
Dirnberger, B.
Distler, U.
Doblmann, J.
Dobson, C.
Doebele, C.
Doellinger, J.
Dojahn, J.
Domon, B.
Dorantes Torres, C. V.
Dörrbaum, A.
Drici, L.
Driesch, D.
Dröse, S.
Droste, C.
Ducret, A.
Duda, G. N.
Dürnberger, G.
Dusberger, F.
Dybowski, J. N.
Dyr, J.
P139
P199
P250
P164
P091
ED2
P022
O18
P190
P005
P127, P056, P057, P094
P169, ED4, P213
P082
P064
P122
P248
P247, P059
P249
P242
P051
P064
P022
P187
P148, P150
E
Ebrahimi, M.
P170
Eckhard, U.
P041
Edwards, I.
P052, P055
Egli, F.
P120
Ehrenberger, T.
P212
Eichacker, L.
P100, O16
Eichelbaum, K.
P176
Eisenacher, M.
O10
Ekkebus, R.
P008
El Kadmiri, N.
P171
El Magraoui, F.
P129
Elamin, A.
P199
Elm , J.
O10
Elschenbroich, S.
P187
Encarnación Guevara, S. M. P154
End, C.
NC5, P172
Engelmann, S.
O14
Engelmann, B.
P164
Ensle, P.
P152
Epping, F.
P193
Erdmann, J.
P161
Ernst, G.
P050
Erzberger, J. P.
O27
Escher, C.
P212, P024
Esdar, C.
P179
Estrada Navarrete, G.
P082
Eyckerman, S.
P012
F. Medzihradszky, K.
P118, P119
F
Fabre, B.
Faelth-Savitski, M.
Faerber, F.
Falkenberg, H.
Farnsworth, C. L.
O02
NC9
P159
P144
P117
220
Fasano, M.
Feenders, C.
Fehniger, T.
Fernández-Niño, S.
Ferreri, C.
Ferro, M.
Fichtenbaum, A.
Fiedler, S.
Fila, J.
Finazzi, G.
Finkemeier, I.
Fischer, L.
Forbrig, C.
Franken, H.
Franzén, B.
Fratini, F.
Fraumene, C.
Freund, C.
Friauf, E.
Fricke, A. L.
Friedrich, A.
Friedrich, N.
Fritzsche, B.
Frydman, C.
Fuchs, S.
Fuchs, J.
Fucile, G.
Fuentes, M.
Fufezan, C.
Führer, D.
Furlan, C.
Fürst, D. O.
Füssl, F.
P204
P036
O11
PL08
P240
O06
P095
P061
P078
O06
O15
P028, P236, P237, P029, P238
P054
NC9
P149
P240
P084, P085
P221, P223
P048
P116
P136
P164
P193
P146
O14, P044
P004
O16
P249
P162, P227, PL32
P164
O04
P116
P137
G
Gabius, H.-J.
Gacek-Matthews, A.
Gajbhiye, R.
Gallardo Rincón, D.
Gallien, S.
Galozzi, S.
Gambaryan, S.
Gandhi, T.
Gannage, M.
Gao, J.
Gao, W.
Garcia, B. A.
Gardner, D.
Garibay-Cerdenares, O. L.
Garin, J.
Garrigues, L.
Gasser, T.
Gateau, A.
Gatto, L.
Gauquelin-Koch, G.
Gavin, A.-C.
Gebhardt, C.
Geerlof, A.
Geiger, J.
Gemoll, T.
Gerdle, B.
P061
P115
YIA1
P154, P138
P213
P153
P162
P106, P212, P214
P241
O13
P211
P094
P210
P154
P032
P032, O02
P003
P032
O06
P202
NC8, PL15
P058
O28
P162
P139, P140, P149, P193
P126
Proteomic Forum 2015
Gerhard, D.
Gerhard, R.
Gerstmair, A.
Gesell Salazar, M.
Gessulat, S.
Gethings, L. A.
Gevaert, K.
Ghafouri, B.
Ghafouri, N.
Ghalim, N.
Ghisaura, S.
Giacomelli, C.
Giambruno, R.
Gielisch, I.
Gierliński, M.
Gierok, P.
Giese, S.
Giesen, C.
Gilsbach, B. K.
Giusti, L.
Gloeckner, C. J.
Gobom, J.
Gocmen, B.
Goeminne, L.
Goerdayal, S.
Goldschmidt-Clermont, M.
Gómez Buitrago, P.-A.
Gomis-Ruth, F. X.
Gonczarowska-Jorge, H.
Gonzalez de Peredo, A.
Gonzalez-Correa, C.-A.
Gonzalez-Muñoz, M.
González-Robles, A.
Górka, M.
Görlich, D.
Gorshkov, M.
Gorza, M.
Gotthardt, K.
Goulas, T.
Graf, A.
Grallert, H.
Gramolini, A.
Gräßler, J.
Greco, V.
Grimm, D.
Grinberg, L. T.
Griss, J.
Gröer, C.
Grolimund, D.
Grossmann, J.
Gruber, C.
Gruhn, F.
Gruissem, W.
Grundner-Culemann, K.
Gu, H.
Guaitoli, G.
Gunsalus, K. C.
Günther, D.
Guntinas-Lichius, O.
Guo, Y.
Guo, M.
P071
P161
P159
P044, P062
P159
P015, P204, P206, P141
P014, P012
P126
P126
P186
P046
P158
O24
P201
P250
P049
P028
PL11
O28
P158
P003, O28
P208
P043
P014
P008
O16
P188
P041
P111
P032
P188
P249
P154
P089, P251
YIA6
O26
P183
O28
P041
P251, P089
P002
P244, HUPO 3
HUPO 8
P079, P195, P189
P245
P123
P016
P203
PL11
P230
P115
P173
P088
P187
P117
O28
P013
PL11
P050
P195
P250
Gu Pta Bhattacharya, S.
P070
Gustavsson, M.
Gyulkhandanyan, A.
H
Haas, H.
Habermann, J. K.
Hafidh, S.
Hahne, H.
Hallström, B.
Hamann, A.
Hammer, E.
Hammer, H.
Hanf, B.
Hanke, T.
Hanke, S.
Hansson, G. C.
Harms, M.
Harris, M.
Hartl, U.
Hartwig, S.
Has, C.
Hattendorf, B.
Hauck, S.
Hauner, H.
Havana Pan, P.-O.
Havrdova, L.
Heazlewood, J.
Heck, A. J. R.
Hecker, M.
Heemskerk, J. W. M.
Heemskerk, J.
Heerkens, E.
Heidari-Keshel, S.
Heide, H.
Heidler, J.
Heinrich, L.
Heinsen, H.
Heinzlmeir, S.
Heldmann, S.
Hellman, U.
Lindsey, M.
Hendricks, M.
Henkel, C.
Hennrich, M.
Henze, A.
Henze, D.
Herber, J.
Hermes, O.
Hermjakob, H.
Hernández Ortíz, M.
Hernández Ramírez, V. I.
Hetzer, M.
Heyer, R.
Hibti, F.
Hildebrandt, P.
Hildebrandt, T.
Hill, A.
Hinneburg, H.
Hintner, H.
Hippler, M.
Ho, A.
221
P157
P030
P068
P139, P140, P193, P149
P078
P159
P159
P225
P164, P062
P220
P038
P130
P239
P097
P099, P007, O23
P134
O18
P149, P193
P027
PL11
P002, P147, P184, P105, P217
P173, P183
P002
P067
P033
PL08
O01, P008, PL07
O14, P081
O08
P162
YIA2
P170, P185
P059
HUPO 4, P246
P193
P123
P009
P050
P139
HUPO 5
P144
P053, O10
NC8
NC4
P246
P135
P120
P016
P154
P154, P138
P134
P025
P146
O14, P049
P071
PL30
P098
P136
P227
NC8
Proteomic Forum 2015
Hochgräfe, F.
Hoeng, J.
Hofer, A.
Hoffmann, M.
Hoffmann, F.
Hollenstein, D.
Homuth, G.
Honys, D.
Hopfgartner, G.
Hoppe, S.
Horejs-Hoeck, J.
Horlacher, O.
Horn, P.
Horn, N.
Hottiger, M. O.
Huang, H.
Huang, S.
Huber, C.
Hudecz, O.
Huesgen, P. F.
Hughes, C.
Humblot, P.
Hunter, C.
Hunyadi-Gulyas, E.
P099, P007, P101, P102
O23, O14
P199
O24
P025
P050
P234
P164
P078
P104
P132
P243
P112
NC8
P160
P104
P211
PL17
P096, P137, P243, P133
P234
P041
P206
P195
P127, P056, P057, P094
P118, P119
I
Ibanez Vea, M.
Ibarrola, N.
Igci, N.
Iglesias, M.
Ihmor, P.
Imami, K.
Infanger, M.
Ingolia, N.
Isaac, G.
Ivanov, M.
Ivanov, N. V.
P122
P249
P043
HUPO 1
P088
P114
P245
P134
P035
O26
P199
J
Jackson, H.
Jaquinod, M.
Jarboui , M. A.
Jennings, P.
Jensen, O. N.
Jerabkova, H.
Jessen, H.
Jilma, B.
Jochim, N.
Johanning, S.
Johansen, E.
John-Schuster, G.
Jones, A.
Jones, E.
Joos, T. O.
Jörnvall, H.
Jozefowicz, A.
Juling, S.
Julius, U.
Junemann, J.
Just, I.
PL11
P032
O07, P125
P177
O03
YIA5
O24
P167
P161
P149
P094
P173
P015
P055
P220, P179
P139
P073
P155, P156
HUPO 8
P161
P161
K
Kalfalah, F.
Karaca, S.
Karl, T.
Kaspar, S.
Kasper, C.
Kasturi,
Keck, T.
Kelleher, N.
Kempa, S.
Kemper, M.
Kennedy, S.
Kepes, F.
Kerner, A.
Khristenko, N.
Kim, Y. J.
Kipping, M.
Kirkland, T.
Kirli, K.
Kiseleva, O.
Kislinger, T.
Kizhakkedathu, J.
Klaasse, G.
Klauck, S.
Klaus, B.
Klauschen, F.
Klausegger, A.
Kleifeld, O.
Klein, O.
Klement, E.
Klempt, P.
Kleta, R.
Kliche, S.
Klingebeil, A. M.
Klodmann, J.
Kmiec, K.
Kniemeyer, O.
Knobbe-Thomsen, C. B.
Koch, A.-K.
Kocher, T.
Kockmann, T.
Koehler, C.
Kögler, G.
Dovichi, N. J.
Kohlbacher, O.
Köhler, N.
Yates, J.
Kohrs, F.
Kolarich, D.
Kollipara, L.
Kondo, T.
Kopitz, J.
Korf, U.
Kortholt, A.
Kostin, S.
Koutroumpas, K.
Kramer, G.
Kraner, M.
Krasnov, G.
Krause, E.
Krieg, N.
Kriitanai, C.
Krijgsveld, J.
222
P124
YIA6
P136
P039, P058, P205, P073, P226
P192
P. O18
P149
ED5, PL03
P152, P178
NC4
P125, O07
P160
P239, YIA4, P222
P169
P213
P015, P062
P172
YIA6
P017
NC2
P120
O01
P136
YIA4
P131
P136
P041
P051, P130
P118
P033
O09
P221
P102
P023
O16
P068, P038
P113
P192
P044, P062
P120
P231
P144
PL02
ED3, P133
O14
HUPO12, PL01
P025
P098
P158, P177
PL26
P061
P168
O28
P246
P160
P141
P074
P017
P155, P221, P223
P142
P067
O05
Proteomic Forum 2015
Kristiansen, G.
Krizova, L.
Krüger, T.
Krüger, C.
Kruse, F.
Kudryavtseva, A.
Kuharev, J.
Kuhlmann, K.
Kuhlmann, L.
Kuhring, M.
Kulkarni, V.
Kumar Jagtap, P.
Kuropka, B.
Kustatscher, G.
Kuster, B.
Kux van Geijtenbeek, S.
P053
P072
P068, P142, P038
P246
P124
P017
ED2
P129, O21
YIA4
P018
YIA1
O28
P221
O04
P103, P159, P009
P242
L
Lahmann, J.
Lalk, M.
Lambour, T.
Lampen, A.
Landthaler, M.
Lane, L.
Frese, C.
Riedel, K.
Lange, P. F.
Langridge, J. I.
Larina, I.
Larsen, M. R.
Lasonder, E.
Laumen, H.
Lavery, D.
Lawless, C.
Lazar, C.
Leber, Y.
Lechner, J.
Lee, H.
Lee, K. A.
Lee, V.
Lehmann, R.
Lehr, S.
Lemeer, S.
Lenobel, R.
Lepper, M.
Lescuyer, P.
Leurs, U.
Levi Mortera, S.
Levitsky, L.
Lewis, S.
Li, L.-H.
Li, X.
Li, Y.
Li, L.
Lichtenstein, D.
Lichtenthaler, S.
Lieberenz, M.
Liebisch, G.
Lievens, S.
Lilley, K.
Limonciel, A.
Lisacek, F.
P139
P049
O02
P155, P156
P114
P032
PL10
PL13
P041
P015, P206, P141, P035
P169
P122, PL06
P216
P002
P187
P204
O06
P116
P131
P002
P117
P141
P248
P149, P193
P103, P159, P009
P006, YIA5
P147
P112
O01
O12
O26
P016
P159
P211
P233
PL17
P155, P156
P135
P159
O21
P012
O06, PL19
P177
P112
223
Lisitsa, A.
Liu, W.
Liu, Y.
Liu, J.
Lobas, A.
Lobo, V.
Lochmanová, G.
Loipetzberger, A.
Longen, S.
Longone, P.
Lopez, M.
Lopez Villar, E.
Lopiano, L.
Lorenz, C.
Loroch, S.
Losio, N.
Lößner, C.
Lotz, J.
Lotz, J.
Lu, J.
Lu, Q.
Lubeck, M.
Lucacchini, A.
Luch, A.
Ludger, G.
Ludwig, C.
Lundgren, C.
Luptovciak, I.
Ly, A.
P017
P163
P211
P211, P163
O26
YIA1
P095
P133
P107
P189
P175
NC6
P204
P071
P162, O17
P079
P208
P050
P050
P128
P160
P058, P226
P158
P174
O20
ED1
P139
P087
P217
M
Macek, B.
Machleidt, T.
Macht, M.
Madero, L.
Magnusson, R.
Majek, P.
Májek, P.
Malih, I.
Maly, M.
Manghina, V.
Mann, M.
von Mering, C.
Mann, G.
Manns, A.
Marchetti-Deschmann, M.
Marcus, K.
Marg, A.
Mariappan, V.
Mariman, E.
Marko-Varga, G.
Marogna, G.
Holmes, E.
Martens, L.
Martin, N. I.
Martin, F.
Martin, F. C.
Martinez Jaime, S.
Marx, H.
Mastrobuoni, G.
Mathieson, T.
Matros, A.
PL12
P172
P127, P056, P057
NC6
P157
P148, P150
O24
P186
P148
P085
PL24, O18, P226
PL25
PL08
P223
P192, P194
P153, O10
P178
P182
P229
O11
P046
PL34
YIA3, O20, PL31, P025, P031
O01
P199
P210
P251
P159
P152
P159
P073, PL20
Proteomic Forum 2015
Mattern, D.
Mattheij, N.
Mayer, R.
McMahon, K.
Mecenas, D.
Mechtler, K.
Medard, G.
Meier, F.
Meierhofer, D.
Meixner, A.
Melchert, S.
Mendes, M.
Mendez, J.
Menneteau, T.
Menzler, J.
Mercuri, N. B.
Merkel, D.
Merl-Pham, J.
Messner, S.
Meyer, H.
Meyer, H. E.
Meyer, K.
Miccolo, C.
Michalak, M.
Michalik, S.
Michalski, A.
Mik, V.
Miladinovic, S. M.
Milani, E.
Milek, M.
Millar, H.
Minucci, S.
Mischerikow, N.
Mishto, M.
Mittler, G.
Mitulović, G.
Mock, H.-P.
Moez, P.
Moghaddas Gholami, A.
Mohamed, S.
Moheb, M.
Mohr, S.
Mollah, S.
Mollenhauer, M.
Monsarrat, B.
Morgenroth, K.
Morimoto, R.
Moritz, C.
Morrice, N.
Morrison, E.
Mostertz, J.
Mouton-Barbosa, E.
Moyet, L.
Muckova, P.
Mueller, M.
Muenz, C.
Mühlhaus, T.
Müller, A.
Müller, G.
Müller, H.-W.
Müller, M.
P038
O08
P180
P204
P013
P004, P234, P022, P064
P009
P226
P201
P003, P004
NC9
P236
P172
O02
P217
P189
P057
P173, P183
P212, P024
O21, HUPO 8
P123, P129, P053, P144, O10
P039
P121
P061
P049
P058, P226
P006
P212
O13
P114
PL17
P121
P004
P200
P109
O22, P095
PL20, P073
P083
P159
P130
P083
P190
P094
P159
O02
P007
O18
P048
NC7
P221
P099, P007, O23, O14
P101, P102
P032
O06
P142, P143
P112
P241
P048
O24
P093
P144
P248
224
Münch, S.
Mungan, M. D.
Munoz-San Juan, I.
Murphy, N.
Murr, A.
Muth, T.
P190
P027
P187
P172
P062
P025
N
Nadler, W.
Nau, K.
Navarro,
Nebrich, G.
Nelson, C.
Nemec, A.
Nemitz, S.
Nguyen, M. T.
Nicolai, S.
Niedzwiecka, A.
Niehues, A.
Niemann, B.
Nilsson, P.
Nitsche, A.
Noberini, R.
Nölker, R.
Norgauer, J.
Nury, C.
Nyalwidhe, J.
Nyman, E.
YIA4, P222
P092
P. ED2
P051
PL17
P072
P142
P160
P174
P155
P227
P156
HUPO 1
P005
P121
P099, P101, P102
P142
P199
P035
P157
O
Obermann, J.
Odeberg, J.
Oefner, P. J.
Oehler, R.
Oellerich, M.
Oellerich, T.
Ohmayer, U.
Olausson, P.
Opialla, T.
Opitz, S.
Oppermann, M.
Opravil, S.
Opsomer, A.
Oravec, M.
Orban-Nemeth, Z.
Orfanos, Z.
Orfao, A.
Ori, A.
Orso, E.
Osiewacz, H. D.
Osorio Trujillo, J. C.
Käll, L.
Oswald, S.
Othman, I.
Otto, A.
Otto, P.
Ovaa, H.
Ovecka, M.
Overall, C. M.
Overbeck, N.
Ozel-DemiralP, F. D.
P184
HUPO 1
O09
O22
P190
P190
P183
P126
P178
P142
P026, P219, P196, P175
P064
P032
P148
P234
P116
P249
P134
O21
P225
P138, P154
PL29
P203
P186
P081
P172
P008
P087
PL27, P041
P113
P043
Proteomic Forum 2015
P
Paech, A.
Pagel, O.
Pagnozzi, D.
Paki , K.
Palmer, M.
Palomba, A.
Panić-Janković, T.
Panse, C.
Parker, S.
Paron, I.
Parte, P.
Pecankova, K.
Pechan, T.
Peitsch, M. C.
Pekar, S.
Pelaseyed, T.
Pellarin, R.
Pengelley, S.
Perocchi, F.
Perske, C.
Persson, H.
Pes, M.
Pesek, J.
Petras, D.
Petrovska, B.
Pfeiffer, A. F.
Pfeilschifter, J.
Pförtner, H.
Philipp, O.
Phillips, B.
Piano, F.
Pich, A.
Pienimäki-Römer, A.
Schmit, P.-O.
Pietsch, J.
Pike, I.
Pinkert, S.
Piras, C.
Pirisi, A.
Pisanu, S.
Pizzi, E.
Planatscher, H.
Plucienniczak, A.
Plucienniczak, G.
Plückthun, A.
Poeckel, D.
Poetsch, A.
Polten, F.
Pongthanom, N.
Pontén, F.
Ponzi, M.
Pope, J.
Poschmann, G.
Potesil, D.
Pötz, O.
Prescott, A.
Pridatchenko, M.
Pruneri, G.
P193
O17
P046, P047, P181, P084, P085
P176
P015
P084, P085, P086
P095
P230
HUPO 10
P226
YIA1
P148, P150
P087
P199
P037
P097
O27
P098
P183
P190
P011
P047
P142
P043
YIA5
NC4
P107
P049
P225
P199
P013
P207, P161
O21, P129
P039
P245
P208
O18
P079, O12, P195, P189
P047
P045, P046, P047, P181
P085, P086
P240
P220
P042
P042
P146
NC9
P225
P207
P067
P159
P240
P015
P092
P078
P162, P220, HUPO 6, P179
P250
O26
P121
R
Rabe, A.
Rabus, R.
Rackwitz, T.
Radiziwill, G.
Radler, P.
Raedschelders, K.
Raether, O.
Raffai, T.
Rafiee, M. R.
Rahm, M.
Raila, J.
Raimondi, F.
Ramadoss, D.
Ramallo Guevara, C.
Rampler, E.
Ramsauer, C.
Rant, U.
Rao, P.
Rapp, E.
Rappold, E.
Rappsilber, J.
Raso, C.
Regine, H.
Regitz-Zagrosek, V.
Regl, C.
Reichl, U.
Reichold, M.
Reifenberger, G.
Reimann, L.
Reimer, U.
Reinders, J.
Reindl, J.
Reinhold, P.
Reiß, S.
Reiter, L.
Ren, J. M.
Ren, X.
Renard, B. Y.
Renard, P.
Renfree, M.
Rexroth, S.
Rezaei-Tavirani, M.
Rhim, J.
Rhode, H.
Richter, E.
Richter, F.
Ried, T.
Riedelova-Reicheltova, Z.
Rieder, V.
Riemer, A. B.
Rigbolt, K.
Rinner, O.
Riwaldt, S.
Robers, M.
Robinson, M.
Roblick, U. J.
Rocca, S.
Rochaix, J.-D.
Roepman, R.
Roesli, C.
Rogers, L. D.
225
P044
P036
P180
P116
P136
HUPO 10
P058, P226
P118
O05
P105
NC4
O28
P197, P224
P225
P234
P243
NC3
YIA1
P025
O22
O04, P054, P028, P236
P237, P029, P238
P125, O07
P174
HUPO 2
P096
P025
O09
P124
P116
P152, P159
O09
P142
P143
O14
P106, P212, P214, P024
P117
P163
O19, P018
P001
P210
P225
P170
P035
P142, P143
P099, P007, P102, O23
P247, P107, P109
P139
P148
P124
P132
P151
P212, P024
P245
P172
P088
P149, P139, P193
P045, P046
O16
P160
P103
P041
Proteomic Forum 2015
Rögner, M.
Rolland, N.
Roncada, P.
Ronci, M.
Röring, M.
Rosenblatt, M.
Rösli, C.
Rozycki, C.
Ruprecht, B.
Rurik, M.
Russell, R. B.
Russell, D.
P225
O06
P079, O12, P195, P189
P158
P151
NC5
P239, P132, YIA4, P222
P156
P103, P009
P133
P160
P250
S
Sabino, F.
Sadewasser, A.
Saha, B.
Saile, R.
Sal, A.
Saliba, P.
Säll, A.
Salvi, D.
Samaj, J.
Samyn, N.
Sanchez Rodriguez, F. E.
Santacoloma-Osorio, M.
Sass, S.
Sattler, M.
Savitski, M. M.
Schaab, C.
Schaefer, T.
Schaefer, A.
Schaefer, J.
Schapiro, D.
Scherbaum, V.
Scherl, A.
Schillo, K.
Schira, J.
Schlage, P.
Schlapbach, R.
Schlegl, J.
Schloegelhofer, P.
Schlosser, A.
Schlüter, R.
Schmidt, F.
Schmidt, G.
Schmidt, C.
Schmit, P.-O.
Schmitz, G.
Schnatbaum, K.
Schneider, H.
Schneider, R.
Schneider, T.
Schneider, M.
Schnölzer, M.
Scholten, A.
Schommartz, T.
Schöndorf, D. C.
Schott, A.-S.
Schriemer, D.
Schrötter, A.
Schrottmaier, W.
P120
P176
P070
P186
O27
P060, NC1
P011
O06
P087
P012
P082
P188
P124
O28
P159, NC9
P187
O27
P105
P146
PL11
P113
P112
P140
P144
P120
P230
P159
P234
P040, P110
P049
P049
P166, P167
P187
P205
P129, O21, HUPO 8
P152, P159
P097
P179
P199
P236, P237
P061, P180
O01, P008
O17
P003
P065
PL33
P123
P167
Schueffler, P.
Schuhmacher, C.
Schulenborg, T.
Schümann, M.
Schurr, B. C.
Schütz, E.
Schutzbier, M.
Schwamb, B.
Schwämmle, V.
Schwarzwälder, A.
Schweiger-Hufnagel, U.
Schweigert, F. J.
Schwenk, J.
Schwinn, M.
Scigelova, M.
Sebela, M.
Sedlmeier, E.-M.
Sedo, O.
Seelk, S.
Segerer, J.
Seidel, P.
Seifert, J.
Selbach, M.
Selve, S.
Senkbeil, H.
Senkler, M.
Senninger, N.
Serve, H.
Seymour, S.
Sferra, G.
Sharma, V.
Shaw, G.
Sickmann, A. HU
Sidoli, S.
Sieg, H.
Siegmund, W.
Sievers, H. H.
Silva, J. C.
Simerský, R.
Simon, M.
Simon, S.
Singh, S.
Sinz, A.
Sitek, B.
Sjöström, D.
Skalický, V.
Skeffington, A.
Slevogt, H.
slotta-Huspenina, J.
Smekalova, V.
Smola, H.
Soggiu, A.
Soher, E.
Solari, F. A.
Solis, N.
Soll, J.
Solovyeva, E.
Sommersdorf, C.
Sonnewald, U.
Sonntag, J.
Spalloni, A.
226
PL11
P146
P048
P155
P065
P190
P064
P059
O03
P245
P205, P098
NC4
HUPO 1
P172
P218
YIA5
P147
P033, P072, P037
P228
P245
P180
P069
P114, P013, P176, P063, P228
P155
P044
P023
P217, P173
P190
P057
P240
P216
P210
PO11, P245, P158, P235
P162, O08, O17, P177
O03
P155, P156
P203
P130
P117
P006
P049
P242
PL22
PL14
P053
P126
P006
P090
P248
P159
P087
P120
P079, O12, P195
P194
O08
P041
P080
O26
P179
P074
P168
P189
Proteomic Forum 2015
Spuler, S.
Srbova, E.
Srnad, M.
Srzentić, K.
Stefanski, A.
Steger, M.
Stejskal, K.
Stella, A.
Stengel, F.
Stentzel, S.
Stephan, C.
Stokes, M. P.
Stopka, P.
Stopkova, R.
Strålfors, P.
Stranzl, T.
Strasser, L.
Strauss, E.
Strauss, J.
Strecker, V.
Streng, A.
Strohkamp, S.
Strohschein, K.
Stühler, K.
Stunnenberg, H. G.
Suckau, D.
Suemchen, P.
Sugihara, Y.
Sukop-Köppel, U.
Surmann, K.
Süssmuth, R.
Suttnar, J.
Swart, C.
Swieringa, F.
Szibor, M.
Szymczak, S.
P178
P078
P006
O26
P113
P110, P059
P078
O02
O27
P049
P194
P117
P033
P033
P157
P234
P243
NC5
P115
P059
P229
P140, P149
P051
P092, P124, P144, P113
P216
P146
P219
O11
P100, O16
P049
P043
P148
P090
O08
P246
P139, P193
T
Taborda-Ocampo, G.
Tacheny, A.
Tahir, S.
Takac, T.
Talamás-Rohana, P.
Tanca, A.
Tarasova, I.
Tascher, G.
Tavernier, J.
Tee, T. Y.
Tenzer, S.
Teufl, J.
Texier, Y.
Textor, B.
The Syscilia Consortium EU
Theis, F.
Thiele, H.
Thierse, H.-J.
Thiyagarajan, V.
Thomas, B.
Thorns, C.
Thriene, K.
Thünemann, A.
Tigges, J.
P188
P001
P238
P087
P138, P154
P181, P084, P085, P086
O26
P202
P012
P186
O16, P048, ED2
P136
P160
P117
P160
P124
P050, P051, P130
P174
P197, P224
P131
P140, P193
P165
P155
P124
Tilocca, B.
Timmerman, E.
Tinnefeld, V.
Titeca, K.
Titz, B.
Toledo-Leyva, A.
Tölle, R.
Tomizioli, M.
Towers, M.
Toyama, B.
Trachsel, C.
Trede, D.
Treue, D.
Trost, M.
Tschöp, M.
Tsybin, Y.
Turkovic, D.
Tursun, B.
U
Ueffing, M.
Uhlen, M.
Uhr, M.
Ulbrich, D.
Umlauf, E.
Upadhyay, A.
Urbani, A.
Urinovska, J.
Urlaub, H.
Uthe, H.
Uzzau, S.
V
Vadivelu, J.
Vadovic, P.
Valiante, V.
van Beersum, S.
Van Calenbergh, S.
van den Berg, B. H. J.
Van den Eynde, B.
van der Ven, P.
van Dieijen-Visser, M.
Van Dorsselaer, A.
Van Eyk, J. HU
van Kasteren, S.
Van Quickelberghe, E.
van Reeuwijk, J.
van Soolingen, D.
van Veelen, P.
Vandenbrouck, Y.
Vandepoele, K.
Vanscheeuwijck, P.
Vanselow, J. T.
Varga, Z.
Vasiljevic, D.
Vecchi, G.
Végvári, Á.
Veitinger, M.
Vellasamy, K. M.
Vendruscolo, M.
227
P069
P012
P235
P012
P199
P138
P209
O06
P052
P134
P230
O10, P050, P051
P131
P250
P093
O26, PL04
P206
P228
P125, P003, P004, O07, P160
O28, P105, P217
HUPO 1, P159
NC5
P101, P102
O22
P197
P158, P079, O12, P195, P189
YIA5
YIA6, P190, PL16
P040
P045, P046, P047, P181
P084, P085, P086
P182
P087
P068
P160
P012
P220
O02
P116
P229
P032, P202
PO 10
P008
P012
P160
YIA2
YIA2
P032, O06
P014
P199
P040
PL11
P228
O18
O11
O22
P182
O18
Proteomic Forum 2015
Venne, S.
P080
Ventz, K.
Verhee, A.
Verheggen, K.
Verhelst, S.
Vigneron, N.
Villegas-pineda, J. C.
Viner, R.
Vinkler, D.
Vissers, J. P. C.
Vitek, O.
Vizcaino, J. A.
Vogel, R. F.
Völker, U.
von Eggeling, F.
von Toerne, C.
von Törne, C.
von Zweydorf, F.
P099, O23
P012
P019
P235, P010
O02
P138
P218
P033
P015, P204, P052, P206, P141
P212
P016
P065
P164, P044, P062, P049
P050
P002
P147, P217
O28
W
Wachter, A.
Waidelich, D.
Waldemarson, S.
Waldera-Lupa, D. M.
Wallner, S.
Walter, U.
Walther, D.
Wang, X.
Wang, R.
Wang, X.
Wang, D.
Wang, H.
Ward, M.
Warnken, U.
Warscheid, B.
Warth, R.
Wawrzyniak, P.
Weber, G.
Wehland-von Trebra, M.
Weibchen, G.
Weilnböck, L.
Weiss, F.
Welinder, C.
Wendler, S.
Wenschuh, H.
Wicht, J.
Wiese, H.
Wild, P.
Wildgruber, R.
Wilhelm, M.
Williams, E.
Williamson, N.
Wills, K.
Wimmer, B.
Winkler, T.
Winter, S.
Winter, M.
Wissinger, Y.
Wittig, I.
Witzel, K.
Wodzig, W.
P168
P056
P011
P124, P144
HUPO 8
P162
O18, P089
NC6
P016
P127
P159
PL11
P208
P180
P116
O09
P042
P100, O16
P245
P204
P133, P096
P220
O11
P142, P143
P152, P159
P190
P116
PL11
P100
P159
P232
P210
O04
P136
P051
P137
P180
P004
P201, P246, P247, P107, P109
P059
PL20
P229
228
Wöhlbrand, L.
Wolff, T.
Jabs, W.
Wollert, K.
Wollscheid, B.
Wolz, C.
Wood, K.
Woodroofe, C.
Wu, Y.
Wühl, M.
P036
P176
P039
P207
O13
O14
P172
P172
P232
P132
X
Xia, B.
Xiong, L.
P163
P094
Y
Yan, J.
Yang, X.
Ye, F.
Yildirim, A. Ö.
Yildiz, M. Z.
YingYing, C.
P130
P163
NC8
P173
P043
P066
Z
Zahedi, R. P.
Zaleski, P.
Zanotelli, V.
Zauber, H.
Zdráhal, Z.
Zecha, J.
Zeiser, J.
Zellner, M.
Zenobi, R.
Zetterberg, H.
Zezzi-Arruda, M.-A.
Zhang, S.
Zhang, H.
Zhang, Y.
Zheng, M.
Zhou, C.
Zhuang, Z.
Zibetti, M.
Zickmann, F.
Ziegler, A.-G.
Ziegler, E.
Zimmermann, P.
Zimmermann, K.
Zinn, N.
Zörnig, M.
HUPO11, O08, O17, P162
P080, P111, P158, P235
P042
P212
P063
P033, P072, P037, P078, P095
P009
P161
O22, P167
P146
P208
P188
O27
P127
P239
O18
P180
P211
P204
O19
P147
P159
P024
P172
NC9
P059