The outermost surface layer of the human skin is the stratum corneum (SC), which serves as a physicochemical barrier between inner body and outer environment. The SC is mainly composed of dead epidermal cells (corneocytes) embedded in an intercellular lipid-rich matrix, providing our body with a highly sophisticated barrier against not only excess water evaporation from inner body but also penetration of foreign substances from outside [1
]. Recent X-ray, neutron, and electron diffraction studies have revealed that intercellular lipids are organized in multi-lamellar structures with repeat distances of about 6 nm and 13 nm, and arranged in orthorhombic (Ort) and hexagonal (Hex) lattices within the lamellae in healthy SC [4
]. Moreover, in addition to these two kinds of ordered phases, FTIR studies have indicated that a small amount of disordered liquid (Liq) phase coexists with the Ort and Hex phases [11
The water content in the whole SC of healthy skin has been estimated to be about 20–30% w
(weight of water/weight of SC) [12
]. Water is considered to be constantly replaced in SC. Namely, the water content of 20–30% w
is maintained by balancing the inflow from viable epidermis with evaporation from the skin surface [14
]. As the extent of water evaporation from SC is quantified as transepidermal water loss (TEWL), a TEWL sensor is usually used to evaluate the barrier function of skin [15
Generally, impairment of the barrier function of SC reduces the SC ability to prevent water from over-evaporating, resulting in an increase in TEWL. Recent studies have revealed that the organization of intercellular lipids plays a crucial role for the skin barrier [18
]. For example, Pilgram et al. reported that the Hex-to-Ort ratio in SC as well as TEWL is higher in the atopic dermatitis and lamellar ichthyosis skins than in the healthy skin, suggesting that the Ort phase is more water-tight than the Hex phase [20
]. However, as well as we know, there is no unequivocal quantitative evidence showing the causality between lipid organization in the intercellular matrix and TEWL. Direct evidence should be required to elucidate the role of intercellular lipid organization for the skin barrier properties.
In this study, we measured temperature dependence of the X-ray diffraction (XD) pattern and TEWL in the isolated human SC to clarify how the lipid organization affects the water permeability. These two kinds of measurements were carried out simultaneously using the same SC sample because the component and structure of SC are slightly different from preparation to preparation and even in the same sample its physicochemical state is affected by various factors such as preparation process, temperature hysteresis, and hydration level [6
]. For this purpose, we newly developed a sample holder for simultaneous measurements of structure and water permeability in an isolated human SC sheet. Detailed comparison of XD and TEWL data made it possible to evaluate the effects of fine structural changes on the water permeation processes.
We succeeded in acquiring X-ray diffraction (XD) patterns and TEWL values simultaneously in human skin stratum corneum (SC) during a temperature scan and obtained a direct evidence for the relationship between the lipid organization in SC and the skin barrier function. The new method we developed in this study provides the fundamental layout to measure additional physicochemical parameters during X-ray diffraction measurements. We designed a new sample holder which holds a single spread SC sheet between donor and acceptor chambers for the diffusant molecule at an angle of 45° with respect to the incident synchrotron X-ray beam (Figure 1
). The slanted SC sheet was able to give Bragg reflections strong enough for quantitative analysis in the wide-angle region. We did not take the undulation state of lipid layers into account in the interpretation of the reflection intensity, assuming that the undulation changed little due to the high bending energy of the ordered phases dominant in the temperature range measured.
The water permeability of the SC sheet was evaluated with an open-chamber type TEWL sensor (Tewameter) placed on the top of the acceptor chamber (Figure 1
). In the Results section, we showed the results in which the TEWL values were significantly lower than those for the membrane filter and exhibited similar thermal behavior in the heating and cooling scans. In some cases, however, the TEWL value increased almost linearly as temperature increased as in the filter membrane probably because the SC sheet adhered poorly to the sample support plate. In addition, we plotted TEWL as relative values because it was hard to obtain reliable absolute values due to indefinite effective area of the sample, inhomogeneous temperature distribution and water condensation in the acceptor chamber. Although further improvement is required in the TEWL measurement to obtain more suitable data for quantitative analysis, the relative TEWL data qualitatively reflect the state of water permeation in the SC.
In this study we showed only the best data to show the correlation between SC structures and water permeability because the reproducibility of TEWL data was low and the scope of this study is limited to establishing the new method for acquiring two kinds of data simultaneously using a spread SC sheet. As it is known that structures and physicochemical properties of SC significantly vary from sample to sample, a subject for future studies will be the detailed analysis of the difference of correlation pattern between samples.
Comparison of TEWL and XD data provides an insight into the mechanism of water diffusion in SC. It suggested that the temperature dependence of TEWL showed two stepwise changes at the beginning of structural transitions. In heating new hexagonal phases, HexH1 and HexH2, started to form at ~40 °C (Figure 5
) and the fluid phase might start to form at ~60 °C as described previously [23
]. We speculate that the stepwise change is related to the structural change at the boundary between structural domains with different orientations, considering that the phase transition may start from the domain boundary, which is a plausible pathway of water diffusion due to lattice defects. The subsequent transition inside the domain may have a much smaller effect on TEWL because it may hardly affect the orientation (direction of the symmetry axis of lipid lateral packing) of the domain. If it is the case, the water diffusion through the defects at the domain boundary may cause only a secondary effect (the extent of the stepwise change in TEWL is relatively small).
The fact that the nonlinear increase in TEWL above 40 °C except for the stepwise change around 60 °C occurred in parallel with the linear increase (thermal areal expansion) in lattice spacings of HexH1 and HexH2 suggests that the lateral molecular packing density in the intercellular lipid layer may be the key factor for the water diffusion in SC. In the temperature range from 40 °C to 60 °C variation of the ratio between hexagonal phases with different lattice spacings makes the situation more complicated. Further discussion should be postponed until more information on the properties of each hexagonal phase is available. Incidentally, it should be noted that temperature dependence of diffusion constant may not be the dominant factor to determine the thermal behavior of TEWL because the overall temperature dependence of TEWL was not the Arrhenius type with a constant activation energy.
In contrast, the increase in TEWL below 40 °C seemed linear, suggesting that the determinant factor for the thermal behavior of TEWL may be different from that above 40 °C. The structural characteristic below 40 °C is definitely different from that above 40 °C and consists of the coexistence of the Ort phase with high lateral packing density and the Hex phase with low lateral packing density. If the lateral packing density is the key factor for water diffusion in SC as discussed above, the Ort phase must be a much higher barrier against water permeation than the Hex phase. Therefore, we infer that the existence ratio of the Ort phase rather than the thermal areal expansion may determine the thermal behavior of TEWL below 40 °C.
In summary, we speculate that the thermal areal expansion of the hexagonal phase may give nonlinearity to the thermal behavior of TEWL above 40 °C whereas that below 40 °C may be closely related to the existence ratio of the Ort phase. However, several problems remain to be solved: (1) The behaviors of TEWL and XD in cooling are slightly different from those in heating, especially below 40 °C, (2) details of phase transition processes (from which phase to which phase) are still in debate, and (3) the TEWL measurement should be improved as discussed above. Here, we put forward a possible explanation of the thermal behavior of TEWL on the basis of the change in the intercellular lipid organization. However, many other explanations may be possible, considering that the skin SC is a multi-component complex system containing various materials other than lipid layers (e.g., keratin fibers). Thus, further study is needed to confirm the above speculations.