Depth profiling of Stratum corneum hydration in vivo: a comparison between conductance and confocal Raman spectroscopic measurements

Authors


  • a

    Different values for the measuring depth of the Corneometer have been reported in the literature. The manufacturer claims the value of 10–20 μm below the SC surface (7) and the European Group for Efficacy Measurements on Cosmetics and Other Topical Products (EEMCO) – 60–100 μm below the SC surface (19). A study with inserted isolating sheets between the probe and the SC surface has found that the measured signal drops to 50% of its original value after 12 μm (20).

Mila Boncheva, Corporate R&D Division, Firmenich SA, PO Box 239, Route des Jeunes 1, CH-1211 Geneva 8, Switzerland, Tel.: (41) 22 780 3027, Fax: (41) 22 780 3334, e-mail: mila.boncheva@firmenich.com

Abstract

Abstract:  The high-frequency electrical conductance of tape-stripped human skin in vivo can be used to evaluate the hydration profile of Stratum corneum (SC). Tape-stripping provides access to the underlying SC layers, and the conductance of these layers (as measured by the Skicon instrument) correlates well with their water content, as demonstrated by independent confocal Raman spectroscopic measurements. The correlation shows high inter-individual variance and is not linear over the full measurement range of the instrument, but is helpful to discriminate between dry, normal and highly hydrated SC. The depth profile of hydration in tape-stripped SC corresponds to the one in intact SC only if the barrier function of the skin is not impaired. Thus, conductometry of tape-stripped skin must be used in conjunction with a method that allows to estimate the barrier damage inflicted to SC during the tape-stripping procedure, for example, measurement of the trans-epidermal water loss. The methodology described here is simple, rapid and minimally invasive, and it employs commercially available instrumentation that is cheap, portable and easy to use. This approach is applicable to in vivo estimation of the SC hydration in studies in the areas of dermatology, skin care and transdermal drug delivery.

Introduction

The objective of this work was to investigate the possibility to build depth profiles of Stratum corneum (SC) hydration in vivo using conductance measurements of tape-stripped skin. Using confocal Raman spectroscopy as a reference technique, we demonstrated that the conductance correlates with the water content of tape-stripped SC provided the SC barrier function is not impaired. Despite its considerable inter-individual variance, the SC conductance is useful to discriminate between dry, normal and highly hydrated SC.

The water content of SC is one of the key factors regulating skin health (1). It affects the permeability and flexibility of SC and modulates the activities of several enzymes involved in the processes of barrier formation and desquamation (2,3). SC hydration exhibits a steep gradient from the skin surface to the viable epidermis (4). Knowledge of the water content and distribution throughout the SC is, thus, highly relevant for numerous medical (i.e., physiological health), cosmetic (i.e., perceived appearance) and pharmaceutical (i.e., transdermal drug administration) applications (5–8).

There are only a few techniques suitable for in vivo, non-invasive determination of the SC water content (9). Among the spectroscopic techniques, confocal Raman spectroscopy has quickly gained recognition as a highly reliable method for direct, spatially resolved depth profiling of the skin hydration (10–13). Importantly, the technique clearly distinguishes between water and other molecular species present in SC [e.g., constituents of the natural moisturizing factor (NMF) or topically applied chemicals] (14,15). The commercially available instrument Model 3510 Skin Composition Analyzer allows for accurate and quick (<30 s) determination of molecular profiles within the skin with a depth resolution of 5 μm. Infrared spectroscopy (in particular, ATR-FTIR) has also been used to evaluate the SC hydration (16), and – in combination with tape-stripping – to investigate the water gradients in SC (17,18). The interpretation of the results, however, is not straightforward because of the overlap in the spectral regions typical for water and proteins and because of difficulties in the normalization of the spectra (16).

The electrical properties of SC can, in principle, also be used to evaluate its hydration status (19). There are several commercially available instruments that rely on electrical measurements to assess the epidermal hydration; the instruments most often used are the Corneometer based on low-frequency capacitance and the Skicon based on high-frequency conductance. The measuring depth of the Corneometer reportedly extends well into the SC and possibly further down into the viable epidermis,a while the measuring depth of the Skicon instrument has been estimated to comprise only the topmost surface of SC (19,21); it is, thus, considered to be better suited to evaluate SC hydration than the Corneometer. The relatively low cost and ease of use have made these instruments highly popular in dermatological research. Unlike the spectroscopic techniques, however, the electrical measurements of SC hydration are indirect, and the exact nature of the complex correlation existing in vivo between the actual water content and the measured electrical parameters has not been demonstrated.

In this study, we examined the correlation between the SC conductance and water content in human skin in vivo by measuring independently the two parameters using the Skicon instrument and confocal Raman spectroscopy, respectively. To gain access to the deep SC layers and enable the measurement of SC conductance as a function of depth within SC, we used tape-stripping, the gradual removal of SC layers using adhesive tapes. This technique has been widely used in pharmacological and dermatological studies in vivo because of its minimal invasiveness (22,23). Previous work has shown that the SC conductance increases with the depth of each tape-stripped layer (24,25); this observation was interpreted as an indication for the existence of a hydration gradient in SC, but it was not confirmed by an independent measurement of the water content. To investigate the suitability of using tape-stripping for depth profiling of the hydration gradient existing in native SC, in addition to the Raman water profiles and the SC conductance, we measured also the transepidermal water loss (TEWL) at the intact skin surface and following tape-stripping. Comparing the water profiles obtained by confocal Raman spectroscopy before and after tape-stripping, we determined the influence of tape-stripping (described by the TEWL) on the water distribution within SC and thus, the conditions under which the conductance measurements correctly describe the hydration gradient within intact SC.

Our study demonstrates that, under appropriate conditions, the conductance of tape-stripped SC correlates with the water content and distribution in SC in vivo. This method provides a practical and inexpensive means for a rough estimation of the SC hydration in settings without access to the relatively costly equipment for confocal Raman spectroscopy.

Methods

Human volunteers

Eight healthy volunteers (five female and three male) aged between 17 and 43 years participated in the study after giving their informed consent. The participants have not used skin lotions or soap on their right forearm during the 12 h preceding the study.

Experimental design

We performed all measurements in an air-conditioned room with a temperature of 22 ± 1°C and a relative humidity of 33 ± 2%. The volunteers acclimatized in the room for 15 min before the start of the measurements. We delimited a circular area with diameter of 20 mm on the right volar forearm of the volunteers using adhesive tape, and cleaned it by briefly wiping it with n-hexane (26,27). In this area, we measured successively the TEWL and the SC conductance, and collected Raman spectra. Following tape-stripping, we left the measurement spot exposed to the ambient atmosphere unoccluded and at rest for 2 min, and then repeated the TEWL, conductance and spectroscopic measurements. We repeated this procedure following consecutive tape-stripping with 1, 3, 5, 10, 15 and 20 tapes.

Tape-stripping and protein quantification

For this procedure, we used D-Squame tapes (CuDerm, Dallas, TX, USA) with a diameter of 22 mm. After applying the tape on the delimited spot on the volunteer’s forearm, we rolled a plastic cylinder over the tape 10 times to ensure the homogeneous adhesion of the tape on the skin before removing the tape in one rapid movement. We quantified the amount of protein on the tapes spectrophotometrically using SquameScan 850 (Heiland Electronic, Wetzlar, Germany) as described previously (28).

TEWL

We used the open-chamber TEWAmeter instrument TM 300 (Courage-Khazaka electronic GmbH, Köln, Germany) equipped with an external probe heater. We conducted the measurements following the published guidelines (29). The standard deviation of all measured values was 0.2 ±0.1 g/m2h.

Measurement of SC conductance

We used the Skicon-200EX (IBS Co, Hamamatsu, Japan). All conductance measurements are reported as an average value ± standard deviation from five measurements taken within the delimited spot on the skin.

Confocal Raman spectroscopy

We used the Model 3510 Skin Composition Analyzer (River Diagnostics, Rotterdam, the Netherlands). The system has been previously described in detail elsewhere (14). To quantify the skin hydration, we used the 2500–3800 cm−1 region of Raman spectra collected with a resolution of 4 cm−1 and a laser excitation at 671 nm. We obtained water profiles from the skin surface to 40 μm below it in 4-μm steps by varying the position of the laser focus. We measured each profile on 10 different locations (about 50 μm apart) to average out the inherent variance in the molecular composition of SC that exists on a microscopic scale. For spectral preprocessing, we used the proprietary RiverICon acquisition software of the instrument. We calculated the water content from the OH/CH ratios of the preprocessed Raman spectra as previously described (11).

Data treatment

For data treatment and calculation of the best fits to the experimental data and the Pearson’s correlation coefficients, we used IgorPro 5.01 (WaveMetrics Inc, Lake Oswego, OR, USA).

Results

Correlation between SC conductance and water content in vivo

Figure 1 shows the conductance of SC following tape-stripping as a function of the water content in the corresponding layers determined by confocal Raman spectroscopy. The correlation between the two parameters is clearly not the same throughout the range of water concentrations observed (Fig. 1a): the conductance increases linearly with the water content up to ∼37 mass % (Fig. 1b), and exponentially above ∼37 mass% (Fig. 1c). We hypothesize that these differences reflect the different binding states of the water molecules in the SC. As demonstrated previously using excised human skin (30,31), the water molecules are more or less tightly bound to the SC keratins at low water contents (up to ∼35–40 mass%); at higher water content, the SC water is present also in a bulk, ‘liquid’ state (32). Thus, it is possible that the different proton mobilities in bound and liquid water exert a different influence on the SC conductance [e.g., via the possibility for high conduction exchange of protons along the hydrogen-bonded network of water molecules (19)].

Figure 1.

 Correlation between SC conductance and water content. The two parameters were measured at the SC surface before and after tape-stripping (1, 3, 5, 10, 15 and 20 tapes) using Skicon and confocal Raman spectroscopy, respectively. The data from each of the eight volunteers are shown in a different colour. (a) Full range of the measured values. (b) Expanded view of the conductance values measured for water contents up to 40 mass%. The line represents the average of the linear fits of the data from each of the eight volunteers (Pearson’s coefficients of the individual linear fits: 0.92, 0.83, 0.91, 0.95, 0.88, 0.86, 0.53 and 0.96). (c) Expanded view of the conductance values measured for water contents above 40 mass%. The line represents the best fit (exponential; Pearson’s coefficient 0.88) to the experimental data.

Previous work using a model system had demonstrated a linear correlation between the water content and conductance (21). The model system used in that study, however, lacked the lipids and NMF present in SC in vivo, two components whose contribution to the overall SC conductance cannot be neglected; the gravimetric determination of the water content and the paucity of data points further flawed the correlation.

Influence of tape-stripping on the water
profile in SC

Next, we investigated whether the water profile after tape-stripping corresponds to the one present in the intact SC, before tape-stripping. To enable the comparison of the Raman water profiles before and after tape-stripping and between different volunteers, we first compared the amount of protein removed by each tape with the SC thickness estimated from the Raman water profile collected after removing the tape (Fig. 2). To ensure that the determination of the SC thickness from the Raman water profiles was not biased by perturbation of the existing water gradient within SC during tape-stripping, we used data from those volunteers in which the water profiles before and after tape-stripping were nearly identical (see also Fig. 3a). The linear correlation (Pearson’s coefficient 0.79) between these values indicates that the removal of 100 μg/cm2 protein corresponds to removal of 1.9 ± 0.2 μm from SC. This value is in excellent agreement with the value of 1.7 μm calculated from the content of proteins, lipids and water in SC (ca. 50%, 20% and 30% of the SC weight, respectively) (33), and their specific densities (1.5, 0.9 and 1.0 g/cm3, respectively) (34,35).

Figure 2.

 Estimation of the depth within SC reached by tape-stripping. Correlation between the cumulative amount of protein progressively removed by tape-stripping and the SC thickness estimated from the Raman water profiles collected after removing each tape. Data from five volunteers. The line is the best fit to the experimental points (Pearson’s coefficient 0.79).

Figure 3.

 Influence of tape-stripping on the water profiles in SC. (a, b) Representative examples of Raman water profiles collected before tape-stripping (lines, top axes) and the water content measured by confocal Raman spectroscopy at the newly exposed SC surface after tape-stripping (circles, bottom axes) with 1, 3, 5, 10, 15 and 20 tapes. See text for details.

The water content at the surface of non-occluded, tape-stripped SC depends upon the balance between three factors: (i) the amount of water present in the exposed SC layer [because of the barrier properties of SC, this amount is higher in the deeper layers compared to the surface ones (36)]; (ii) the influx of water from the underlying SC layers and viable epidermis caused by the damage to the SC barrier inflicted by tape-stripping (37); and (iii), the evaporative loss of water (TEWL) from the exposed surface. During the Raman measurement, the third factor is no longer relevant as the skin is in conformal contact with the surface of the fused silica window.

To evaluate the extent of the water influx in tape-stripped SC, we compared the Raman water profiles collected at the SC surface before and after tape-stripping (Fig. 3).The experimental data fell into two distinct groups. In one group (five volunteers), the depth profiles of the water content before and after tape-stripping were very similar (Fig. 3a). Apparently, in these cases, the water content at the tape-stripped SC surface was determined mainly by the amount of water present in these deeper SC layers; the influx of water due to removal of SC layers was relatively low. In the other group (three volunteers), the depth profiles of the water content before and after tape-stripping differed considerably; the tape-stripped SC layers contained more water than expected for the corresponding depth from the Raman profile of the intact SC (Fig. 3b). Apparently, in these cases, the removal of SC layers had caused considerable water influx from the viable epidermis.

These data demonstrate that the water gradient in tape-stripped SC corresponds to the one in native, intact SC only when the stripping procedure causes minimal damage to the SC, i.e., when it provokes only negligible water influx into SC. In principle, such situation is possible when the SC lipids have the appropriate molecular organization (i.e., high content of orthorhombic lipid phases in the lamellae) and/or when SC is thick (38). We are currently investigating the relative importance of these factors.

This finding has important implications for the applicability of the conductance measurement of tape-stripped SC for determination of its hydration profile in vivo. As the water content of tape-stripped SC corresponds to the one of intact SC only if the stripping procedure has not provoked an additional water influx into the exposed deeper layers of SC, the measurement of conductance – unlike confocal Raman spectroscopic measurements – cannot be used as a stand-alone method for determination of SC hydration profile. The measured electrical conductance of the exposed SC layers does not distinguish between the water inherently present in the exposed SC layers and the water influx caused by damaging the SC barrier during tape-stripping; thus, the method has to be combined with a method that allows to estimate the barrier damage inflicted to SC during the tape-stripping procedure, for example, measurement of the TEWL (39). Figure 4 shows the SC conductance of tape-stripped SC as a function of the TEWL of the exposed SC layers. The dependence is not linear throughout the range of measured conductance values, as would be expected from the linear gradient of SC hydration measured non-invasively by confocal Raman spectroscopy (40). The conductance increased linearly with TEWL only up to ∼12 g/m2h, the value considered to indicate an intact SC barrier; above this value, the conductance increased exponentially with TEWL. Thus, once the SC barrier has been damaged by tape-stripping, it is no longer clear whether the conductance increases because of the high water content of the deeper SC layers or because of the water flux induced by tape-stripping.

Figure 4.

 Influence of tape-stripping on the SC conductance. Correlation between the SC conductance and TEWL measured before and after sequential stripping of the skin with 1, 3, 5, 10, 15 and 20 tapes. Data from each volunteer are shown in a different colour. The lines represent the best fits to the experimental points: linear for TEWL ≤ 12 g/m2h, and exponential for TEWL ≥ 12 g/m2h. The individual linear fits of the data for each volunteer had Pearson’s coefficients of 0.91, 0.50, 0.99, 0.88, 0.80, 0.99, 0.79, 0.76, and the exponential fit – 0.91.

In view of this finding, we re-examined the correlation between the conductance and the water content of tape-stripped SC. Figure 5 shows only those data points from Fig. 1 for which the TEWL was below 12 g/m2h. The best fit through the experimental points was linear, with a markedly reduced scatter compared to Fig. 1b (Pearson’s coefficient 0.622). Interestingly, we observed two groups with different pattern of conductance changes: in one group (five volunteers), the average values of slope and intercept were 6.0 ± 0.8 and 108.6 ± 36.2, respectively, and in the other group (three volunteers), these values were, respectively, 17.0 ± 5.2 and 479.2 ± 155.4. We believe that these differences reflect differences in the composition and organization of SC; we are currently investigating this possibility.

Figure 5.

 Correlation between SC conductance and water content of tape-stripped SC with intact barrier (values from Fig. 1 for which TEWL ≤ 12 g/m2h). The line represents the best fit (linear; slope: 6.67 ± 1.56; intercept: −123.7 ± 50.6; Pearson’s coefficient 0.62) of all data points. The individual linear fits of the data for each volunteer had Pearson’s coefficients of 0.92, 0.83, 0.91, 0.97, 0.95, 0.75, 0.88 and 0.89.

Discussion

In this study, we demonstrated for the first time the correlation between SC conductance and water content in vivo. We found that the conductance changes linearly with the SC water content up to 37 mass% and exponentially above this value, reflecting most probably the different influences of bound and free SC water on conductance. Considering the measurement range of the instrument – upper limit of 2000 μs (data from the manufacturer) and lower limit of approximately 20 μs (deduced from the y-intercept in the linear fit of Fig. 1b) – conductometry is a useful technique to measure SC hydration in the range between ca. 20 and 60 mass%.

The use of conductance in conjunction with tape-stripping has several important advantages over the other techniques used to evaluate the SC water content: (i) the method is minimally invasive and, thus, well suited for in vivo studies in the absence of confocal Raman equipment; (ii) in intact skin (i.e., one for which TEWL ≤ 12 g/m2h), the conductance of SC can be used to distinguish between dry SC (i.e., one with water content below 10–15 mass%), normal SC (i.e., one with water content of ca. 25–30 mass%) and highly hydrated SC (i.e., one with water content above 45 mass%); (iii) the method employs commercially available instrumentation that – unlike, for example, the commercially available equipment for confocal Raman or infrared spectroscopy – is cheap, portable and easy to use; (iv) unlike the other commercially available instruments for electrical measurements on SC, for example the Corneometer, the measurement depth of the Skicon comprises only the topmost SC (19,21); it is, thus, suitable for evaluating the hydration gradients within SC.

The method has also several disadvantages: (i) unlike confocal Raman spectroscopy, a method capable of direct, confocal detection, it cannot be used as a stand-alone, fully non-invasive method to determine the depth profile of SC hydration. The measurement of conductance has to be applied to tape-stripped SC in conjunction with measurement of TEWL, as the water content of tape-stripped SC is similar to the one in native, non-stripped SC only if the increased water flux into SC (resulting from removing SC layers) is very low; (ii) because of the huge inter-individual variability of conductance for water contents below 37 mass%, the method does not allow for determination of the water content with the high precision easily achievable by the direct determination using confocal Raman spectroscopy; (iii) the measured electrical conductance is very sensitive to the ambient and skin temperature. Even more importantly, the changes in the SC conductance because of changes in the water content are indistinguishable from those caused by the presence of NMF or exogenous hygroscopic chemicals in SC; any conducting substance present in SC contributes to its conductance (13,41). This last observation has important implications for the use of conductance to evaluate the hydration effect of topically applied substances; it is possible only if the substances themselves are not conducting.

We believe that because of its minimal invasiveness, simplicity and low cost, the methodology we present is well suited for clinical studies of healthy and diseased SC. It is, in principle, applicable also to evaluate the effect of environmental conditions or topically applied (non-conducting) chemicals on the SC hydration. This method is, thus, highly relevant for in vivo studies in the areas of skin biophysics, transdermal drug delivery, dermatology and skin care.

Acknowledgement

This work was financed by Firmenich SA.

Ancillary