How to Bring the ‘‘Unseen’’ Proteome to the Limelight via

Bioscience Reports, Vol. 25, Nos. 1/2, February/April 2005 ( 2005)
DOI: 10.1007/s10540-005-2844-2
How to Bring the ‘‘Unseen’’ Proteome to the Limelight via
Electrophoretic Pre-Fractionation Techniques
Pier Giorgio Righetti,1,4 Annalisa Castagna,1 Ben Herbert,2 and Giovanni Candiano3
The present review reports a panoply of electrophoretic methods as pre-fractionation tools
in proteomic investigations in preparation for mass spectrometry or two-dimensional electrophoresis map analysis. Such electrophoretic pre-fractionation protocols include all those
electrokinetic methodologies which are performed in free solution, most of them relying on
isoelectric focusing steps (although some approaches based on gels and granulated media
are also discussed). Devices associated with electrophoretic separations are multi-chamber
apparatuses, such as the multi-compartment electrolyzers equipped with either isoelectric
membranes or with isoelectric beads, Off-Gel electrophoresis in a multi-cup device and the
Rotofor, an instrument also based on a multi-chamber system but exploiting the
conventional technique of carrier-ampholyte-focusing. Other free-flow systems, as well as
miniaturized chambers, are also described.
KEY WORDS: Proteomics; pre-fractionation; isoelectric focusing; two-dimensional maps.
ABBREVIATIONS: FFE: free-flow electrophoresis; FF-IEF: free-flow isoelectric focusing;
IPG: immobilized pH gradients; CBI: codon bias index; EOF: electroendoosmotic flow;
2-DE: two-dimensional electrophoresis.
Although, at the latest count, the total number of coding genes in humans would
appear to oscillate between only 25,000 and 30,000 (Southan, 2004), the complexity
of the human proteome could, nevertheless, be overwhelming. An example on such a
vast complexity and thus on the stringent need for pre-fractionation in proteome
analysis comes from some recent articles on the plasma proteome (Anderson and
Anderson, 2002; Pieper et al., 2003). If one assumes that there are just 500 true
‘‘plasma proteins’’, each present in 20 variously glycosylated forms and in five different sizes, one would end up with 50,000 molecular forms. If one further
hypothesizes that the ca. 30,000 gene products in the human proteome exist on
average as 10 splice variants, cleavage products and post-translational modifications,
Department of Industrial and Agricultural Biotechnologies, University of Verona, Strada Le Grazie 15,
37134, Verona, Italy.
Proteome Systems, 35 Waterloo Rd, North Ryde, 2113, Sydney, NSW, Australia.
Laboratory of Physiopathology of Uremia, G. Gaslini Children’s Hospital, 16148, Genova, Italy.
To whom should be addressed. E-mail: [email protected]
0144-8463/05/0400-0003/0 2005 Springer Science+Business Media, Inc.
Righetti, Castagna, Herbert, and Candiano
this would yield some additional 300,000 protein forms. Rammensee (2004) reported
that, by splicing proteins into small pieces, stitching different portions together, and
then cutting out amino acids sequences from the melded pieces, cells can manufacture a very large variety of new forms from the original gene coded proteins.
Moreover single proteins such as antibodies might contain more than 1,000,000
different epitopes sequences which add to the complexity of the serum content. To
this very intricate situation, additional difficulties come from the fact that the
dynamic range, at least in serum, might be more than 10 orders of magnitude.
In essence it is clear that a natural way to analyze the content of a proteome or
even checking phenotyping differences, is to pre-fractionate the proteome into discrete groups and then analyze separately each group. Some recent papers suggest
that a problem of this vast complexity could be overcome not by pre-fractionation
but rather by running a series of narrow-range IPG strips (covering no more than 1
pH unit). Hoving et al. (2000) and Westbrook et al. (2001) note that ‘‘zoom’’ and
‘‘ultra-zoom’’ gels are quite important for avoiding or at least minimizing the
problem of spot overlapping in 2-DE, an ever present hazard in 2-DE maps (see
Pietrogrande et al., 2002, 2003; Campostrini et al., 2005). The use of large-size gel
slabs (18-cm or longer in the first dimension, 18 · 20 cm, or larger, in the second
dimension), would even dramatically increase the resolution, as reported by Corthals
et al. (2000) and Wildgruber et al. (2000). By that way it seems possible to make
portions of bi-dimensional mappings without the necessity of pre-fractionating the
sample. The entire, wide-range map would then be electronically reconstructed by
stitching together the narrow-range maps. In practice, however, it remains the fact
that, even when using very narrow IPG strips, they have to be loaded with the entire
tissue lysate with consequent massive precipitation, along with the additional
drawback that the proteins which should focus in the chosen narrow-range IPG
interval will be strongly under-represented because they will be only a small fraction
of the entire sample loaded (Herbert et al., 2004). These serious issues have been
debated in a recent work by Gygi et al. (2000), who argued that, in 2D gels, proteins
from genes with codon bias values of <0.1 (low-abundance species), large-size
proteins (>100 kDa) and most membrane proteins could not be found.
Several published papers highlight major limitations of available technologies
for proteome investigations. Current approaches are qualified as incapable of having
a whole vision of the proteome, even limited to structural aspects. For instance
strongly alkaline proteins are poorly represented when using classical two-dimensional electrophoresis, as underlined by Bae et al. (2003), and highly hydrophobic
proteins cannot be properly solubilized and consequently not analyzed and/or
identified. Electrophoresis-based methods taken alone (still the most commonly used
to date) are neither appropriate for polypeptides of masses lower than 5000 Da, nor
effective for very alkaline proteins. Only mass spectrometry contributes significantly
to the analysis of low sized polypeptides. To this panel it is to be added that posttranslational modifications and especially glycosylations are still part of the nonresolved dilemmas. In this situation authors estimate that only about 20 30% of
expressed proteins are detectable by standard methods to date.
Pre-fractionation, in all of its possible variants, as here reviewed, appears to be
the logical way to follow in the attempt to make a step in the right direction. As
elegantly stated by Pedersen et al. (2003), in fact, pre-fractionation could be a
How to Bring the ‘‘Unseen’’ Proteome to the Limelight
formidable tool for ‘‘mining below the tip of the iceberg to find low abundance and
membrane proteins’’. So, let us wear our proper mining tools, a pickaxe, acetylene
lamp, helmet and goggles and descend deep inside the mine to start our digging.
Although several types of pre-fractionation tools exist, we will limit this review only
to electrophoretic techniques. Centrifugal pre-fractionation, for isolation of cell
organelles, has been surveyed in a number of reports (Cordwell et al., 2000; Corthals
et al., 2000; Jung et al., 2000; Dreger, 2003; Huber et al., 2003; Stannard et al.,
2004). Chromatographic techniques have also been covered by several authors
including us (Lopez, 2000; Isaaq et al., 2002; Righetti et al., 2003a, 2003b; Lescuyer
et al., 2004).
Although plain zone electrophoresis in gel cylinders would appear to be just
about the last resort, it was recently applied by Fountoulakis and Juranville (2003) to
the enrichment of low-abundance brain proteins, by eluting some 80 fractions (each
of 10 ml) from a 11%T gel equilibrated with 0.1% lithium dodecyl sulphate (LDS).
Interestingly, they could enrich relatively low molecular mass proteins, such as
hippocalcin, visinin, 14-3-3 proteins. By the same token, one could perhaps exploit
the electrosmotic-pump-driven apparatus of Hayakawa et al. (2003), although gelbased systems do not appear to be popular in pre-fractionation protocols.
Over gel phases, this technique has the advantage that much higher sample
loads can be applied, coupled to the absence of possible protein modifications induced by free monomers always present in gel phases (Chiari et al., 1992; Bordini
et al., 2000). Present equipment derives from the concepts and instrumentation of
Hannig (1967, 1982), by which the electrolyte solution flows in a direction normal to
the lines of forces of the electric field and the mixture to be separated is added
continuously at a small spot in the flowing medium. Components of the mixture are
deflected in diagonal trajectories according to their electrophoretic mobility and can
be collected at the bottom of the chamber into as many as 96 fractions. Free-flow
electrophoresis (FFE) was born as a technique for purifying cells and sub-cellular
organelles, which could be recovered highly purified as thin zones, due to their very
low diffusion coefficients. The Hannig apparatus went through successive designs
and improvements, from an original liquid descending curtain to the present commercial version, dubbed Octopus, exploiting an upward liquid stream (Kuhn and
Wagner, 1989). Instrumental to the success of the method, especially when run for
long periods of times, is the constancy of the elution profile at the collection port, so
that the same protein species is always collected into the same test tube. Thus,
electroendoosmotic flow should be suppressed via a number of ways, including glasswall silanol deactivation and addition of polymers (e.g., 0.1% hydroxypropylmethyl
cellulose), providing dynamic wall coating and proper liquid viscosity. FFE in a
ProTeam apparatus was recently reported by Zischka et al. (2003) for purification of
S. cerevisiae mitochondria, previously purified by fractional centrifugation. These
authors claimed identification of many more proteins (n = 129) from FFE-purified
Righetti, Castagna, Herbert, and Candiano
mitochondria as compared with mitochondrial protein extracts isolated by differential centrifugation (n = 80). In addition, a marked decrease of degraded proteins
was found in the FFE-purified mitochondrial protein extracts, suggesting that the
organelles were contaminated by lysosomes.
FFE would not be ideal for protein pre-fractionation, though, due to their
higher diffusion coefficients, as compared with cells and organelles. However,
Kobayashi et al. (2003) reported a microfabricated FFE device useful for continuous
separation of proteins. Their separation chamber is barely 66 · 70 mm in size, with a
gap between the two Pyrex glass plates of only 30 lm (see Fig. 1a and b). The liquid
curtain and the sample are continuously injected from five and one holes, respectively, at the top. At the bottom of the chamber, a micromodule fraction collector,
consisting of 19 stainless steel tubes, is connected perpendicular to the chamber and
liquid stream. FFE, for protein separation, would work much better in the isoelectric
focusing mode (FF-IEF), due to built-in forces impeding entropic peak dissipation.
The first report on the use of FF-IEF for pre-fractionation of total cell lysates from
HeLa and Ht1080 cell lines, in view of a subsequent 2-DE map, is perhaps the one of
Burggraf et al. (1995), who collected individual or pooled fractions for further 2-DE
analysis. Hoffman et al. (2001) proposed FF-IEF as the first dimension of a 2-DE
map, the eluted fractions being directly analyzed by orthogonal SDS-PAGE. In turn,
individual bands in the second SDS-PAGE dimension were eluted and analyzed by
electrospray ionization, ion-trap MS. By this approach, they could identify a number
of cytosolic proteins of a human colon carcinoma cell line. One advantage of FFIEF is immediately evident from their data: large proteins (e.g., vinculin, Mr
116.6 kDa) could be well recovered and easily identified; on the contrary, recovery of
large Mr species has always been problematic in IPG gels. Weber et al. (2004) have
also adopted FF-IEF for the efficient separation and analysis of peroxisomal
membrane proteins. Their success was documented by the detection of PMP22, the
most hydrophobic and basic protein (pI > 10) of peroxisomal membranes. Perhaps
pre-fractionation of proteins could also be attempted by IEF in the recently-revived
vortex-stabilized, free-flow electrophoretic device (Ivory, 2004; Tracy and Ivory,
Behind this remarkable invention by M. Bier, there is a long history going back
to the doctoral thesis of Hjerte´n (1967): as he was trained in astrophysics, his
apparatus was a ‘‘Copernican revolution’’ in electrokinetic methodologies. Hjerte´n
was the first one to propose electrophoretic separations in a free zone (i.e., in the
absence of anticonvective, capillary media, such as polyacrylamide and agarose gel
networks) but he had to fight the noxious phenomenon of electrodecantation,
induced by gravity. He thus devised rotation of the narrow-bore tubes used as
electrophoretic chambers around a horizontal axis, mimicking celestial planet motions! This must have spurred the fantasy of Svensson-Rilbe, in those days a colleague at the Uppsala University, who finally described a large multi-compartment
electrolyzer, capable of fractionating proteins in the gram range (Jonsson and Rilbe,
1980). The cell was assembled from 46 compartments, accommodating a total
sample volume of 7.6 l, having a total length of 1 m, hardly user-friendly! Cooling
How to Bring the ‘‘Unseen’’ Proteome to the Limelight
Fig. 1. Scheme of the miniaturized FFE apparatus. Reservoir R1 is used for the electrophoresis buffer, whereas R2 (0.1 M aqueous NaOH ethanol, 50:50, v/v), R3 (0.01 M
HCl) and R4 (80% aqueous ethanol) are used for washing cycles. S1 is the sample
reservoir, P1 P3 are peristaltic pumps for pouring the solutions into the separation
chamber, the electrode reservoirs and the sample port, respectively. D is a dumper, DRN
a drain duct, PF is a fuse unit for excessive pressure. B: Scheme of the micromodule
fraction separator (MFS). A cross-section of the separation chamber’s bottom with upper
and lower parts of Pyrex glass is shown (from Kobayashi et al., 2003, by permission).
Righetti, Castagna, Herbert, and Candiano
and stirring were affected by slow rotation of the whole apparatus in a tank filled
with cold water. Bier’s 50-year-long love affair with preparative electrophoresis in
free solution produced, as a last evolutionary step, a remarkable device, the Rotofor
(Egen et al., 1988; Bier, 1998). A preparative-scale Rotofor is capable of being
loaded with up to 1 g of protein, in a total volume of up to 55 ml. A mini-Rotofor,
with a reduced volume of about 18 ml, is also available. The device is assembled
from 20 sample chambers, separated by liquid-permeable nylon screens, except at the
extremities, where cation- and anion-exchange membranes are placed against the
anodic and cathodic compartments, respectively, so as to prevent diffusion within
the sample chambers of noxious electrodic products. At the end of the preparative
run, the 20 focused fractions are collected simultaneously by piercing a septum at the
chambers’ bottom via 20 needles connected to a vacuum source. The narrow-pI
range fractions can then be used to generate conventional 2-DE maps. This is the
original approach described by Hochstrasser et al. (1991). In recent times, this
methodology has taken another turn: the Rotofor is used directly as the first
dimension of a peculiar 2-DE methodology, in which each fraction is further analyzed by hydrophobic interaction chromatography, using non-porous reversed-phase
HPLC (Zhu et al., 2003). Each peak collected from the HPLC column is then digested with trypsin, subjected to MALDI-TOF MS analysis and MSFit database
searching. By this approach, Wall et al. (2000) have been able to resolve a total of ca.
700 bands from a human erythroleukemia cell line. It should be stated, though, that
the pI accuracy of this methodology which is still based on conventional carrier
ampholyte-isoelectric focusing, CA-IEF, is quite poor: it ranges from ±0.65 to
±1.73 pI units, a large error, indeed. On a similar line of thinking, Davidsson et al.
(2001) have sub-fractionated human cerebrospinal fluid and brain tissue, whereas
Wang et al. (2002) have mapped the proteome of ovarian carcinoma cells. More
recently, Xiao et al. (2004) have reported a novel application of the Rotofor, not just
for fractionation of intact proteins in presence of carrier ampholytes, but for fractionation of peptide digests of an entire proteome (in this case, human serum) in an
ampholyte-free environment. The peptides themselves would act as carrier ampholyte-buffers and create a pH gradient via an ‘‘autofocusing’’ process (with a caveat,
though: the pH gradient will be quite poor, since only a few peptides have good
buffering power and conductivity in the pH 5 8 range).
The Gradiflow is a multi-functional electrokinetic membrane apparatus that can
process and purify protein solutions based on differences of mobility, pI and size
(Margolis et al., 1995; Horva`th et al., 1996). Its interfacing with 2-DE map analysis
was demonstrated by Corthals et al. (1997), who adapted this instrument for prefractionation of native human serum and enrichment of protein fractions. In a more
recent report (Locke et al., 2002), this device was also shown to be compatible, in the
pre-fractionation of bakers yeast and Chinese snow pea seeds total cellular extracts,
with the classical denaturing/solubilizing solutions of 2-DE maps, comprising urea/
thiourea and surfactants. Whereas, in the case of size fractionation, this can be
achieved with polyacrylamide coated membranes at different %T and %C for
sieving of macromolecules in given Mr ranges, its use for separating approximate pI
How to Bring the ‘‘Unseen’’ Proteome to the Limelight
fractions is more complex and has to adopt low-conductivity buffers, such as those
devised by Bier et al. (1984), so as to allow reasonably high voltage gradients and
relatively short separation times. It should be remembered, though, that the Gradiflow operates on the principle of binary fractionations, so that, in general, only two
populations can be collected during each run, the so called ‘‘upstream’’ and
‘‘downstream’’ fractions. More recently, Bae et al. (2003) adopted the Gradiflow for
pre-fractionation of alkaline proteins from Helicobacter pylori, although the terminology ‘‘extremely basic fraction’’ seems an exaggeration, considering that all the
species identified by MALDI-TOF MS hardly reached pI values as high as pH 10.
We have already mentioned multi-compartment electrolyzers (MCE; Jonsson
and Rilbe, 1980; Bier, 1998), as a class of instruments based on conventional IEF in
presence of soluble, amphoteric buffers (carrier ampholytes, CA). However, the
MCEs based on Immobiline membranes represent a quantum jump over the previous technique (Righetti et al., 1989, 1990, 1992). This method relies on isoelectric
membranes, fabricated with the same acrylic monomers adopted in IPG fractionations (Wenger et al., 1987; Righetti, 1990). Advantages of such a procedure are
immediately apparent: (i) such a device offers a method that is fully compatible with
the subsequent first dimension separation in 2-DE maps, a focusing step based on
Immobiline technology. Thus, protein mixtures harvested from the various chambers
of this apparatus can be loaded onto IPG strips without any need for further
treatment, in that they are isoelectric and isoionic; (ii) it permits harvesting a population of proteins having pI values precisely matching the pH gradient of any
narrow (or wider) IPG strip; (iii) as a corollary of the above point, much reduced
chances of protein precipitation will occur, as compared to loading onto a narrow
IPG strip an unfractionated sample composed of a much wider pI spectrum (in the
latter case, proteins non-isoelectric in the given pH range will massively precipitate
towards the ends of the IPG strips, most often co-precipitating neighbouring species); (iv) due to the fact that only proteins co-focusing in the same IPG interval will
be present, much higher sample loads can be operative, permitting detection of lowabundance proteins.
The original apparatus, as miniaturized by Herbert and Righetti (2000),
Righetti et al. (2001) and Herbert et al. (2004), is shown schematically in Fig. 2a. In
this exploded view, two terminal electrodic chambers are used to block, in between,
three sample chambers. Fig. 2b is a schematic diagram of the MCE for initial plasma
fractionation. The four disks in the upper part are the isoelectric membranes inserted
in between the various chambers. In this particular set-up, the couple pI 5.0 and pI
6.0 is used as a trap for capturing albumin. By properly exploiting this pre-fractionation device, Pedersen et al. (2003) have been able to capture and detect a large
number of the ‘‘unseen’’ yeast membrane proteome.
Figure 3 gives an example of the large number of membrane proteins detected,
via this pre-fractionation protocol, in the pH 7 10.5 range, an interval that
cartographers of the 16 century would have described as ‘‘terra incognita’’, devoid of
any landmarks, and stamped inside the contour of the map the inscription:
Righetti, Castagna, Herbert, and Candiano
Fig. 2. (a) Exploded view of the miniaturized multi-compartment electrolyzer operating with
isoelectric membranes. An assembly with only 5 chambers is shown (3 sample chambers and the
two termini electrodic reservoirs). (b) Schematic diagram of the MCE for initial plasma fractionation. The four upper disks represent the isoelectric membranes to be sandwiched in between
each chamber (by courtesy of Proteome Systems).
‘‘hic sunt leones’’, just as they did with the African continent. These data fully
misspell the notion expounded by Gygi et al. (2000) (see Introduction) that 2-DE
maps cannot detect membrane and low abundance proteins (see also Herbert et al.,
2003); the key for ferreting them out is to use appropriate pre-fractionation methods
associated with concentration.
A number of additional approaches have been described such as miniaturized
devices (Zuo and Speicher, 2002) and the Rotofor accommodating isoelectric
membranes (Shang et al., 2003). Zhu and Lubman (2004) have modified the IsoPrime device from Hoefer so as to lessen run volumes significantly; additionally, the
protein content captured in each chamber was further fractionated via non-porous
reversed-phase HPLC. The notion that isoelectric membrane-based devices could
How to Bring the ‘‘Unseen’’ Proteome to the Limelight
Fig. 3. Coomassie brilliant blue stained 2-D gels of the alkaline MCE fraction from a membrane
preparation of log phase yeast. The alkaline fraction was separated in the first dimension IPG
using 2% ASB 14 detergent. The excess detergent has combined with SDS to form mixed micelles
in the second dimension gel and caused the smearing observed in the low molecular mass part of
the gel. The 2-D gel is the display of a 1.0 mg membrane protein preparation using an 11 cm pH
7 10.5 IPG for the first dimension and GelChipTM 8 18% T second dimension gel. The gel is
annotated with 237 proteins, representing 93 unique gene products (from Pedersen et al., 2003, by
not capture very high pI proteins has been recently misspelled by Lalwani et al.
(2004a). These authors used high pI membranes fabricated with quaternary
ammonium derivatives of cyclodextrins and poly(vinyl alcohol), cross-linked with
glycerol-1,3-diglycidyl ether and demonstrated compatibility with catholytes as
caustic as 1 M sodium hydroxide. By the same token, this same group (Lalwani
et al., 2004b) has produced hydrolytically stable, low-pI isoelectric membranes from
low-pI ampholytic components, poly(vinyl alcohol), and a bifunctional cross-linker,
glycerol-1,3-diglycidyl ether. The low-pI ampholytic components used contain one
amino group and at least two weakly acidic functional groups. These new, very lowpI isoelectric membranes have been successfully used as anodic membranes in isoelectric trapping separations with pH <1.5 anolytes and have been found to be a
good replacement for the hydrolytically less stable polyacrylamide-based isoelectric
membranes. Now the circle is closed and no one can any longer claim that IPGs
cannot capture very low and very high pI proteins!
Perhaps, though, one of the limiting steps in fractionating samples with these
devices is the length of time needed to capture a given protein population in a given
chamber, due to the sieving properties of the isoelectric membranes. A remedy to the
Righetti, Castagna, Herbert, and Candiano
slow migration of proteins in MCEs due to the sieving effect of isoelectric membranes has been recently proposed by Cretich et al. (2003) who suggested using
hydrogel beads, in lieu of membranes, as pI barriers sandwiched in between the
various chambers. Although this approach greatly reduced the focusing time, it was
plagued by the presence of electroendoosmotic flow (EOF), as the beads did not
provide a flow-tight system. Aware of that, we designed new amphoteric beads,
composed of ionic acrylamide derivative monomers co-polymerized within the pores
of a central ceramic hard core, minimizing thus mass transfer resistance of proteins
that are transiently adsorbed onto the beads (Fortis et al., 2005a). Additionally,
these beads exhibit a much reduced EOF, thus permitting fast separations in MCE
devices (Fortis et al., 2005b) with minimal liquid flux from chamber to chamber. It is
anticipated that isoelectric beads will find a role in proteomic applications as a result
of a rapid separation in miniaturized devices.
An interesting variant to the use of the MCE apparatus could be a kind of direct
2D method, as depicted in Fig. 4, by which the surface charge fractionation, as
obtained in the various MCE fractions, is coupled to size discrimination by running
directly the content of each MCE chamber into an SDS gel. The bands eluted from
the latter step would then be analyzed directly by MS (Cottingham, 2003).
Although in the previous section we have mentioned ‘‘miniaturized’’ multichamber instruments, in reality these approaches still accommodate sizable sample
volumes in each chamber, of the order of 0.5 2 ml. Smaller devices have been
recently described. In one approach Tan et al. (2002) have built a device consisting in
96 mini-chambers (~75 ll each) arranged in eight rows. Neighbouring chambers in a
given row are separated by short glass tubes (4 mm innerdiameter, ID, 3 mm long),
within which isoelectric hydrogels of specific pH values are polymerized. During
focusing, the device is sandwiched between blocks incorporating reservoirs for
catholyte and anolyte. This device has been used not for the fractionation of proteins, but rather of their digests. With the described set up, however, some reservations are to be underlined; one of them is that peptides will collect mostly around
acidic and basic pI values, leaving preciously little few (mostly those containing His
residues) in the pH 5 8 range. In the 2-DE map displayed in Fig. 2a of Tan et al.
(2002) the peptides deriving from the digest of four protein markers are visibly
grouped into vertical pillars centered around the following pIs: 4.5, 6.0, 8.8 and 10.0.
In another approach, Zilberstein et al. (2003) proposed parallel processing in isoelectric focusing chips. The main separation tool, here, is a dielectric membrane
(chip) with conducting channels that are filled by isoelectric hydrogels of varying pH
values. The membranes are held perpendicularly to the applied electric field and
proteins are trapped in the channels whose pH values are equal to the pI of the
proteins. Further progress in these parallel, miniaturized devices has been reported
by the same group (Zilberstein et al., 2004a, 2004b).
In yet a third approach, a system called ‘‘Off-gel IEF’’ has been described by
Ros et al. (2002). Just like the multi-compartment separation technique, the system
has been devised for the separation of proteins according to their pI and for their
direct recovery in solution without adding buffers or ampholytes. The principle is to
How to Bring the ‘‘Unseen’’ Proteome to the Limelight
Fig. 4. Schematic diagram of a 2-D method interfacing MCE fractions with SDS-PAGE.
Upper drawing: MCE instrument assembled with seven sample and two electrodic chambers
(the numbers on top refer to the pIs of the various isoelectric membranes). Lower drawing:
loading of the content of each chamber directly into SDS-PAGE gels. The bands resolved
after the second step will then be analysed by MS (Courtesy of D. Speicher, Winstar
place a sample in a liquid chamber which is positioned on top of an IPG gel.
Theoretical calculations and modelling have shown that the protonation of an
ampholyte occurs in the thin layer of solvation closed to the IPG gel/solution
interface (Arnaud et al., 2002). Upon application of a voltage gradient, perpendicularly to the liquid chamber, the electric field penetrates into the channel and extracts
all charged species (those having pI values above and below the pH of the IPG gel),
thus vacating them from the sample cup. After separation, only the globally neutral
species (pI = pH of the IPG gel) remain in solution. In a further extension of this
initial work, the system was improved and adapted to a multi-well device, composed
Righetti, Castagna, Herbert, and Candiano
Fig. 5. Experimental set-up used to perform Off-Gel separations with multi-cup devices, composed of either 10 (a) or 22 (b and c) wells in different pH intervals, as specified under the gels
(from Michel et al., 2003, by permission).
of a series of compartments of small volume (100 300 ll) and compatible with
current instruments for separation (Michel et al., 2003) (see Fig. 5).
Contrary to all above-described methods Go¨rg et al. (2002), reported a technique of sample pre-fractionation with neutral beads of dextran (Sephadex) to isolate
proteins by isoelectric focusing prior to 2-DE map analysis. This is in fact a rediscovery of the well known ‘‘Radola (1973, 1975) technique’’, described in the 70s.
When the method was first introduced it became quite popular and as a consequence
commercial products were designed to ‘‘guillotine’’ out the entire Sephadex cake into
20 pieces after migration (for a more thorough description of the principle and the
set-up, see Righetti, 1983). The focusing process is of course induced by the presence
of carrier ampholytes, the Sephadex beads being exploited only as an anticonvective
medium. Proteomic analysis that follows is not often compatible with the presence of
carrier ampholytes, which are difficult to remove. Mass spectrometry is, for instance,
not compatible, since also carrier ampholytes would be detected and not easily
distinguished from peptides of similar molecular masses.
PGR is supported by FIRB 2001 (No. RBNE01KJHT), PRIN 2005 (MURST,
Rome), Fondazione Cassa di Risparmio di Verona (Bando 2002) and by the
How to Bring the ‘‘Unseen’’ Proteome to the Limelight
European Community (proposal No. 12793, Allergy Card, 2005). GC is supported
by TELETHON (GP0019Y01), the ‘‘Foundation for Renal Disease in Children’’
and FFC#13/2003 ‘‘Proteomics of the airway surface liquid: implication for cystic
Anderson, L. N. and Anderson, N. G. (2002) Mol. Cell. Proteomics 1:845 867.
Arnaud, I. L., Josserand, J., Rossier, J. S., and Girault, H. H. (2002) Electrophoresis 23:3253 3261.
Bae, S. H., Harris, A. G., Hains, P. G., Chen, H., Garfin, D. E., Hazell, S. L., Paik, Y. K., Walsh, B., and
Cordwell, S. J. (2003) Proteomics 3:569 579.
Bier, M. (1998) Electrophoresis 19:1057 1063.
Bier, M., Mosher, R. A., Thormann, W., and Graham, A. (1984) In: Electrophoresis ’83 (H. Hirai, ed.),
de Gruyter, Berlin, pp. 99 107.
Bordini, E., Hamdan, M., and Righetti, P. G. (2000) Rapid Commun. Mass Spectrom. 14:840 848.
Burggraf, D., Weber, G., and Lottspeich, F. (1995) Electrophoresis 16:1010 1015.
Campostrini, N., Areces, L. B., Rappsilber, J., Pietrogrande, M. C., Dondi, F., Pastorino, F., Ponzoni, M.,
and Righetti, P. G. (2005) Proteomics 5: 2385 2395.
Chiari, M., Righetti, P. G., Negri, A., Ceciliani, F., and Ronchi, S. (1992) Electrophoresis 13:882 884.
Cordwell, S. J., Nouwens, A. S., Verrills, N. M., Basseal, D. J., and Walsh, B. J. (2000) Electrophoresis
21:1094 1103.
Corthals, G. L., Molloy, M. P., Herbert, B. R., Williams, K. L., and Gooley, A. A. (1997) Electrophoresis
18:317 323.
Corthals, G. L., Wasinger, C. V., Hochstrasser, D. F., and Sanchez, J. C. (2000) Electrophoresis
21:1104 1115.
Cottingham, K. (2003) J. Proteome Res. 2:574 574.
Cretich, M., Pirri, G., Carrea, G., and Chiari, M. (2003) Electrophoresis 24:577 581.
Davidsson, P., Paulson, L., Hesse, C., Blennow, K., and Nilsson, C. L. (2001) Proteomics 1:444 452.
Dreger, M. (2003) Eur. J. Biochem. 270:589 599.
Egen, N. B., Bliss, M., Mayersohn, M., Owens, S. M., Arnold, L., and Bier, M. (1988) Anal. Biochem.
172:488 494.
Fortis, F., Girot, P., Brieau, O., Boschetti, E., Castagna, A., and Righetti, P. G. (2005a) Proteomics 5:
620 628.
Fortis, F., Girot, P., Brieau, O., Castagna, A., Righetti, P. G., and Boschetti, E. (2005b) Proteomics 5:
629 638.
Fountoulakis, M. and Juranville, J. F. (2003) Anal. Biochem. 313:267 282.
Go¨rg, A., Boguth, G., Ko¨pf, A., Reil, G., Parlar, H., and Weiss, W. (2002) Proteomics 2:1652 1657.
Gygi, S. P., Corthals, G. L., Zhang, Y., Rochon, Y., and Aebersold, R. (2000) Proc. Natl. Acad. Sci. USA
97:9390 9395.
Hannig, K. (1967) In: Electrophoresis: Theory, Methods and Applications, Vol. 2 (M. Bier, ed.), Academic
Press, New York, pp. 423 471.
Hannig, K. (1982) Electrophoresis 3:235 242
Hayakawa, M., Hosogi, Y., Takiguchi, H., Shiroza, T., Shibata, Y., Hiratsuka, K., Kiyama-Kishikawa,
M., Hamajima, S., and Abiko, Y. (2003) Anal. Biochem. 313:60 67.
Herbert, B. R. and Righetti, P. G. (2000) Electrophoresis 21:3639 3648.
Herbert, B., Pedersen, S. K., Harry, J. L., Sebastian, L., Grinyer, J., Traini, M. D., McCarthy, J. T.,
Wilkins, M. R., Gooley, A. A., Righetti, P. G., Packer, N. H., and Williams, K. L. (2003) PharmaGenomics 3:22 36.
Herbert, B., Righetti, P. G., McCarthy, J., Grinyer, J., Castagna, A., Laver, M., Durack, M., Rummery,
G., Harcourt, R. and Williams, K. L. (2004) In: Purifying Proteins for Proteomics (R. J. Simpson,
ed.), Cold Spring Harbor Lab. Press, Cold Spring Harbor, pp. 431 442.
Hjerte´n, S. (1967) Chromatogr. Rev. 9:122 219.
Hochstrasser, A. C., James, R. W., Pometta, D., and Hochstrasser, D. (1991) Appl. Theor. Electrophor.
1:333 337.
Righetti, Castagna, Herbert, and Candiano
Hoffmann, P., Ji, H., Moritz, R. L., Connolly, L. M., Frecklington, D. F., Layton, M. J., Eddes, J. S., and
Simpson, R. J. (2001) Proteomics 1:807 818.
Horva`th, Z. S., Gooley, A. A., Wrigley, C. W., Margolis, J., and Williams, K. L. (1996) Electrophoresis
17:224 226.
Hoving, S., Voshol, H., and van Oostrum, J. (2000) Electrophoresis 21:2617 2621.
Huber, L. A., Pfaller, K., and Vietor, I. (2003) Circulation Res. 92:962 968.
Ivory, C. F. (2004) Electrophoresis 25:360 374.
Issaq, H. J., Conrads, T. P., Janini, G. M., and Veenstra, T. D. (2002) Electrophoresis 23:3048 3061.
Jonsson, M. and Rilbe, H. (1980) Electrophoresis 1:3 14.
Jung, E., Heller, M., Sanchez, J. C., and Hochstrasser, D. F. (2000) Electrophoresis 21:3369 3377.
Kobayashi, H., Shimamura, K., Akaida, T., Sakano, E., Tajima, N., Funazaki, J., Suzuki, H., and
Shinohara, E. (2003) J. Chromatogr. A 990:169 178.
Kuhn, R. and Wagner, H. (1989) J. Chromatogr. 481:343 350.
Lalwani, S., Shave, E., Fleisher, H. C., Nzeadibe, K., Busby, M. B., and Vigh, G. (2004a) Electrophoresis
25:2128 2138.
Lalwani, S., Shave, E., and Vigh, G. (2004b) Electrophoresis 25:3323 3330.
Lescuyer, P., Hochstrasser, D. F., and Sanchez, J. -C. (2004) Electrophoresis 25:1125 1135.
Locke, V. L., Gibson, T. S., Thomas, T. M., Corthals, G. L., and Rylatt, D. B. (2002) Proteomics
2:1254 1260.
Lopez, M. F. (2000) Electrophoresis 21:1082 1093.
Margolis, J., Corthals, G., and Horva`th, Z. S. (1995) Electrophoresis 16:98 100.
Michel, P. E., Reymond, P., Arnaud, I. L., Josserand, J., Girault, H. H., and Rossier, J. S. (2003)
Electrophoresis 24:3 11.
Pedersen, S. K., Harry, J. L., Sebastian, L., Baker, J., Traini, M. D., McCarthy, J. T., Manoharan, A.,
Wilkins, M. R., Gooley, A. A., Righetti, P. G., Packer, N. H., Williams, K. L., and Herbert, B. (2003)
J. Proteome Res. 2:303 312.
Pieper, R., Gatlin, C. L., Makusky, A. J., Russo, P. S., Schatz, C. R., Miller, S. S., Su, Q., McGrath, A.
M., Estock, M. A., Parmar, P. P., Zhao, M., Huang, S. T., Zhou, J., Wang, F., Esquer-Blasco, R.,
Anderson, N. L., Taylor, J., and Steiner, S. (2003) Proteomics 3:1345 1364.
Pietrogrande, M. C., Marchetti, N., Dondi, F., and Righetti, P. G. (2002) Electrophoresis 23:283 291.
Pietrogrande, M. C., Marchetti, N., Dondi, F., and Righetti, P. G. (2003) Electrophoresis 24:217 224.
Rammensee, H.G. (2004) Nature 427:203 204.
Righetti, P.G. (1983) Isoelectric Focusing: Theory, Methodology and Applications. Elsevier, Amsterdam
129 136.
Righetti, P. G., Wenisch, E., and Faupel, M. (1989) J. Chromatogr. 475:293 309.
Righetti, P. G., Wenisch, E., Jungbauer, A., Katinger, H., and Faupel., M. (1990) J. Chromatogr.
500:681 696.
Righetti, P. G. (1990) Immobilized pH Gradients: Theory and Methodology. Elsevier, Amsterdam 1 400.
Righetti, P. G., Faupel, M. and Wenisch, E. (1992) In: Advances in Electrophoresis, Vol. 5 (A. Chrambach,
M. J. Dunn and B. J. Radola, eds.), VCH, Weinheim, pp. 159 200.
Righetti, P. G., Castagna, A., and Herbert, B. (2001) Anal. Chem. 73:320A 326A.
Righetti, P. G., Castagna, A., Herbert, B., Reymond, F., and Rossier, J. S. (2003a) Proteomics
3:1397 1407.
Righetti, P. G., Castagna, A., and Hamdan, M. (2003b) In: Advances in Chromatography, Vol. 42 (P. R.
Brown and E. Grushka, eds.), M. Dekker, New York, pp. 270 321.
Ros, A., Faupel, M., Mees, H., Oostrum, J. V., Ferrigno, R., Reymond, F., Michel, P., Rossier, J. S., and
Girault, H. H. (2002) Proteomics 2:151 156.
Shang, T. Q., Ginter, J. M., Johnston, M. V., Larsen, B. S., and McEwen, C. N. (2003) Electrophoresis
24:2359 2368.
Southan, C. (2004) Proteomics 4:1712 1726.
Stannard, C., Brown, L. R., and Godovac-Zimmermann, J. (2004) Curr. Proteomics 1:13 25.
Tan, A., Pashkova, A., Zang, L., Foret, F., and Karger, B. L. (2002) Electrophoresis 23:3599 3607.
Tracy, N. I. and Ivory, C. F. (2004) Electrophoresis 25:1748 1757.
Wall, D. B., Kachman, M. T., Gong, S., Hinderer, R., Parus, S., Misek, D. E., Hanash, S. M., and
Lubman, D. M. (2000) Anal. Chem. 72:1099 1111.
How to Bring the ‘‘Unseen’’ Proteome to the Limelight
Wang, H., Kachman, M. T., Schwartz, D. R., Cho, K. R., and Lubman, D. M. (2002) Electrophoresis
23:3168 3181.
Weber, G., Islinger, M., Weber, P., Eckerskorn, C., and Volkl, A. (2004) Electrophoresis 25:1735 1747.
Wenger, P., de Zuanni, M., Javet, P., Gelfi, C., and Righetti, P. G. (1987) J. Biochem. Biophys. Methods
14:29 43.
Westbrook, J. A., Yan, J. X., Wait, R., Welson, S. Y., and Dunn, M. J. (2001) Electrophoresis
22:2865 2871.
Wildgruber, R., Harder, A., Obermeier, C., Boguth, G., Weiss, W., Fey, S. J., Larsen, P. M., and Goerg,
A. (2000) Electrophoresis 21:2610 2616.
Xiao, Z., Conrads, T. P., Lucas, D. A., Janini, G. M., Schaefer, C. F., Buetow, K. H., Issaq, H. J., and
Veenstra, T. D. (2004) Electrophoresis 25:128 133.
Zhu, K., Yan, F., O’Neil, K. A., Hamler, R., Lin, L., Barder, T. J., and Lubman, D. M. (2003) Curr. Prot.
Protein Sci. :2331 23328.
Zhu, Y. and Lubman, D. M. (2004) Electrophoresis 25:949 958.
Zilberstein, G. V., Baskin, E. M., and Bukshpan, S. (2003) Electrophoresis 24:3735 3744.
Zilberstein, G. V., Baskin, E. M., Bukshpan, S., and Korol, L. E. (2004a) Electrophoresis 25:3643 3651.
Zilberstein, G. V., Korol, L. E., Bukshpan, S., and Baskin, E. M. (2004b) Proteomics 4:2533 2540.
Zischka, H., Weber, G., Weber, P. J. A., Posch, A., Braun, R. J., Buhringer, D., Schneider, U., Nissum,
M., Meitinger, T., Ueffing, M., and Eckerskorn, C. (2003) Proteomics 3:906 916.
Zuo, X. and Speicher, D. W. (2002) Proteomics 2:58 68.