4D confocal microscopy method for drug localization in the skin

4D confocal microscopy method for drug localization in the skin
Ulf Maeder*a, Thorsten Bergmanna, Jan Michael Burga, Sebastian Beera, Peggy Schluppb, Thomas
Schmidtsb, Johannes T. Heverhagenc, Frank Runkelb, Martin Fiebicha
Institute of Medical Physics and Radiation Protection, Technische Hochschule Mittelhessen University of Applied Sciences, Germany
Institute of Biopharmaceutical Technology, Technische Hochschule Mittelhessen - University of
Applied Sciences, Germany
Department of Diagnostic Radiology, Philips University Marburg, Germany
*[email protected]; phone ++49 641 3092645; imps.th-mittelhessen.de
A 4D confocal microscopy (xyzλ) method for measuring the drug distribution in skin samples after a permeation study is
investigated. This approach can be applied to compare different drug carrier systems in pharmaceutical research studies.
For the development of this detection scheme phantom permeation studies and preliminary skin measurements are
carried out. The phantom studies are used to detect the permeation depth and the localization of the external applied
fluorescent dye naphthofluorescein that is used as a model agent. The skin study shows the feasibility of the method for
real tissue.
For the differentiation of tissue/phantom and the dye, spectral unmixing is performed using the spectral information
detected by a confocal microscope. The results show that it is possible to identify and localize external dyes in the
phantoms as well as in the skin samples.
Keywords: transdermal drug transport, skin permeation study, confocal microscopy, spectral unmixing, drug permeation
depth, local drug distribution
The development of drug carrier systems (DCS) for the treatment of skin diseases is emerging in recent years [1].
Modern DCS like multiple emulsions [2] and solid lipid nanoparticles [3] are developed to increase the drug-uptake of
the skin. Regarding the very poor uptake-rate compared to the amount administered to the skin for conventional DCS,
improved systems to reduce costs and dose are needed.
The evaluation of the developed DCS is an important step in designing new systems. A common method to investigate
the performance of a DCS is a permeation study using Franz diffusion cells. This method measures the total amount of a
drug that permeates through and the total amount that is located within a skin sample using high-performance liquid
chromatography. Unfortunately, no information about the local drug distribution in the skin is available. Furthermore the
method is error-prone and time-consuming.
In this study we introduce a method for measuring the drug distribution after a permeation study using 4D confocal
microscopy (xyzλ). We use fluorescent dyes as model agents to evaluate the drug transport and the local distribution.
Measuring the emission spectra allows the separation into skin auto-fluorescence and the dye.
For validating the method a tissue phantom that resembles the optical properties of the dermis skin layer of human and
porcine skin [4] is used to perform phantom permeation studies. Inclusions of the fluorescent dye naphthofluorescein are
detected inside a fluorescein phantom and the depth of these inclusions is determined.
Additionally, measurements of a skin permeation study are presented. In this study naphthofluorescein dissolved in water
is applied topically onto a pig skin sample for several hours. The results show the location of the penetrated dye in the
2.1 Fluorescent dyes
The fluorescent dyes fluorescein natrium and naphthofluorescein (both Sigma-Aldrich Chemie GmbH, Taufkirchen,
Germany) are used. The excitation/emission maxima of fluorescein natrium and naphthofluorescein are 485 nm/514 nm
and 594 nm/663 nm respectively. Both dyes are soluble in water and therefore suitable for incorporation in tissue
phantoms. The fluorescence of naphthofluorescein is pH dependent and can be observed at a pH higher than 8. Therefore
pH adjustment is necessary.
2.2 Fluorescein phantom
The phantom consists of agarose (200 mg), aqua dest (19 ml), lipid emulsion (1 ml) and fluorescein natrium dissolved in
aqua dest (30 µl, 10 mM solution). The ingredients are heated up to 90°C on a heating plate and stirred using a magnetic
stirrer. Afterwards the fluid phantom is poured out into a petri dish and cooled down using a thermal pack.
2.3 Skin Samples
Porcine skin [5] extracted from the ear is used in this study. The subcutis tissue is softly removed with a scalpel. For the
permeation study skin samples (20mm diameter) are taken and stored in phosphate buffered saline at room-temperature.
2.4 Confocal Microscope
A TCS SP5 II (Leica Microsystems, Mannheim, Germany) providing 496 nm, 514 nm and 561 nm excitation lines is
used for the measurements. The image evaluation is performed with the accompanying software LAS AF Lite.
3.1 Phantom permeation study
For the permeation study 50 µl of a 1 mM naphthofluorescein solution (dissolved in aqua dest, adjusted to pH 9) is
applied onto the fluorescein phantom for 4 hours. After removing the residue a 4D scan is performed recording the
fluorescence intensity distribution in dependency of the excitation and emission wavelengths. Table 1 lists the spectral
scanning parameters. The excitation wavelengths 496nm and 561nm are nearest to the excitation maximum of
fluorescein and naphthofluorescein respectively.
Table 1: 4D scanning parameters
excitation [nm]
detection start/end [nm]
527 / 727
550 / 725
586 / 763
detection steps/width
10 / 20nm
10 / 20nm
10 / 20nm
3.2 Skin measurements
In this preliminary study the skin is treaded with dimethyl sulfoxid (DMSO) dissolved in water. DMSO is known to
degrade the skin barrier and therefore helps incorporating substances into the skin. Naphthofluorescein (50 µl of a 1mM
solution) is applied topically onto the skin samples as a model drug to simulate a skin permeation study. After 5 minutes
the sample is cleaned with a tissue to eliminate rough residues but to keep small amounts of the dye on the skin.
Afterwards the 4D measurements are accomplished.
Figure 1: Phantom penetration measurements with varying excitation wavelengths (496nm, 514nm, 561nm). Left:
Overview. Right: Detailed view of the emission region of interest.
4.1 Phanton measurements
Figure 1 shows spectra of selected ROIs recorded from the fluorescein phantom treaded with the naphthofluorescein
solution excited with 496nm, 514nm and 561nm respectively. The ROIs are defined to cover the naphthofluorescein
containing areas (see Figure 2 red box). A detailed view of the wavelengths of interest when identifying
naphthofluorescein is given on the right. Figure 2 shows the images recorded of the phantom at different emission and
excitation wavelengths. The inclusions of naphthofluorescein are clearly visible at the 561nm excitation line and 660nm
emission wavelength. On the left side a circular structure can be identified (red circle) at the 590nm emission images that
seems to be a naphthofluorescein inclusion as well. Taking the information of the 660nm emission images into account it
becomes obvious that this structure does not contain naphthofluorescein and is possibly an air-inclusion.
The spectral information is further used for the spectral unmixing technique to differentiate the naphthofluorescein
inclusions from the phantom. The intensity depth profile of the phantom and an inclusion are shown in figure 3. It is
possible to evaluate the area and the permeation depth of naphthofluorescein.
4.2 Skin measurements
The same 4D scans are performed for the prepared skin samples. Figure 4 shows images and the spectra of a sample
excited at 514 nm and detected at 562 nm and 659 nm (±10 nm bandwidth) respectively. The appearance of the sample
surface changes depending on the regarded emission band. This is due to naphthofluorescein being present in the
observed layer. The spectral data shows that naphthofluorescein clearly permeated into the sample at ROI 2 but does not
show a strong influence on the signal detected at ROI 1. Figure 4 also depicts the spectral unmixing results (red: skin
autofluorescence; green: naphthofluorescein; using the spectral data shown) indicating that naphthofluorescein has
permeated through the sample except the boundary. This might be explained by the topographical surface of the skin
sample. The topically applied naphthofluorescein does not reach the small skin folds. The structure described by ROI 2
in figure 4 can be interpreted as a local accumulation of naphthofluorescein.
Recorded at
590 ± 10nm
Recorded at
660 ± 10nm
Figure 2: Images of the fluorescein phantom treaded with naphthofluorescein showing dye inclusions. Top row: Images
recorded at 590 ± 10nm emission wavelength. Button row: Images recorded at 660 ± 10nm. Excitation wavelengths:
496nm (left), 514nm (middle), 561nm (right). Circle: possibly an air inclusion. Box: naphthofluorescein inclusion
Figure 3: Depth profile of the sample and the naphthofluorescein inclusion (left) and the fluorescence image (right) after
performing spectral unmixing. The inclusion is shown in red, the phantom is colored green
Emission [nm]
(ROI 1)
(ROI 2)
Figure 4: Skin sample detected at 562 nm and 659 nm (±10 nm bandwidth) respectively and the spectral unmixing results
(left). Spectral unmixing was performed using the spectral data (shown on the right) of ROI 1 and ROI 2. The skin
autofluorescence is shown in red, naphthofluorescein is shown in green.
The phantom measurements show the possibility to localize external dyes and to evaluate their permeation depth.
Because of the broad autofluorescence spectrum of tissue that superimposes the signal of external dyes spectral
information is important to differentiate the signal origins. It is possible to include dyes like melanin and riboflavin and
to add oil-water emulsions to the phantoms to resemble the skins autofluorescence and to produce realistic scattering
behavior [4]. In this way standardized measurements can be performed to validate the method.
The presented preliminary skin permeation study also shows results indicating that the method is applicable for
comparing DCS. It has to be mentioned that the sample was only roughly cleaned to keep residues of the dye on the
surface. In this way it was assured to detect the dye and to test the method on tissue.
In the next studies DMSO as permeation enhancer that destroys the skin barrier will be replaced by a submicroemulsion
that is specially designed to increase permeation rates of drugs into the skin. Naphthofluorescein will also be replaced
because of the strong pH value dependence on the fluorescence intensity. It is suitable for phantom measurements where
the pH value is easily adjusted, but as the skin has varying pH values in different skin layers [6,7] the dye is not adequate
for these studies. Furthermore the following permeation studies will be performed for different time periods to
investigate the influence on the dye uptake.
Assuming that the skin autofluorescence is equally distributed in z-direction and therefore the signal decreases only due
to the extinction coefficient of the sample, the autofluorescence intensity can be used as internal normalization reference
for the dye intensity. In this way we can qualitatively describe the signal loss of the dye due to the decreasing dye
concentration along the z-direction and it is therefore possible to evaluate the drug distribution in the sample.
The main problem with depth dependent measurements is the low achievable imaging depth of confocal microscopes in
tissue. For measuring depth profiles of structures in phantoms with low light scattering and absorption it is well suited,
but skin in contrast is highly scattering and therefore strongly reduces the imaging depth. Two-photon microscopy as
used by other groups [7,9] might solve this problem or at least enhance the imaging depth to get useful information of the
dye distribution in deeper tissue layers.
Accompanying these studies, Monte Carlo models are designed to calculate calibration factors for quantitative
measurements [8]. Using these factors it becomes possible to convert the detected signal intensities into the dye
concentration. Keeping in mind that the dyes are used as model agents this is a step towards the quantitative description
of the transdermal drug transport.
We would like to thank the Hessen State Ministry of Higher Education, Research and the Arts for the financial support
within the Hessen initiative for scientific and economic excellence (LOEWE-Program).
[1] Prausnitz, M. R., Langer, R., "Transdermal drug delivery" Nat Biotechnol 26(11), 1261-1268 (2008).
[2] Schmidts, T., Dobler, D., Schlupp, P., Nissing, C., Garn, H., Runkel, F.,“ Development of multiple W/O/W
emulsions as dermal carrier system for oligonucleotides: Effect of additives on emulsion stability”, International
Journal of Pharmaceutics 398, 107-113 (2010).
[3] Steinle, T., Ebrahimi, M., Schmidts, T., Runkel, F., Czermak, P.,“ Membrane Assisted Production of
Sphingosine-1-Phosphate (S1P) Loaded Solid Lipid Nanoparticles (SLN) for Treatment of Acne vulgaris”,
Desalination 250, 1132-1135 (2010).
[4] Bergmann, T., Beer, S., Maeder, U., Burg, J.M., Schlupp, P., Schmidts, T., Runkel, F., Fiebich, M.,“
Development of a skin phantom of the epidermis and evaluation by using fluorescence techniques” Proc. SPIE
7906, 79060T (2011).
[5] Herkenne, C., Naik, A., Kalia, Y.N., Hadgraft, J., Guy, R.H., "Pig ear skin ex vivo as a model for in vivo
dermatopharmacokinetic studies in man" Pharm Res 23(8), 1850-1856 (2006).
[6] Wagner, H., Kostka, K.H., Lehr, C.M., Schaefer, U.F.,” pH profiles in human skin: influence of two in vitro test
systems for drug delivery testing” European Journal of Pharmaceutics and Biopharmaceutics 55(1), 57-65
[7] Hanson, K.M., Bardeen, C.J.,“ Application of nonlinear optical microscopy for imaging skin”, Photochemistry
and Photobiology 85(1), 33-44 (2009).
[8] Maeder, U., Schmidts, T., Avci, E., Heverhagen, J.T., Runkel, F., Fiebich, M.,“Feasibility of Monte Carlo
simulations in quantitative tissue imaging” Int J Artif Organs 33(4), 253-259 (2010).
[9] Ericson, M.B., Simonsson, C., Guldbrand, S., Ljungblad, C., Paoli, J., Smedh, M., “Two-photon laser-scanning
fluorescence microscopy applied for studies of human skin.” J Biophotonics 1(4), 320-330 (2008).