110 JOURNAL OF COSMETIC SCIENCE IDENTIFICATION OF ANOTHER BARRIER IN HUMAN STRATUM CORNEUM: THE WATER BARRIER Johann W. Wiechers', Ph.D., Joke A Bouwstra 2, Ph.D. and Anko de GraafF •Uniqema, PO Box 2, 2800AA Gouda, The Netherlands 2Leiden/Amsterdam Center of Drug Research, PO Box 9502, 2300 RA Leiden, The Netherlands Introduction: For more than two decades the intercellular lipid composition has been identified as the barrier to the influx of chemicals from the outside environment. But whilst this barrier does perform an excellent job in keeping the external world outside, we still lose considerable amounts of water every day via Transepidermal water loss. How can this barrier in one direction be so tough to cross but from the other direction not be watertight? Why do we not completely dehydrate in a dry environment? In order to address these rather fundamental questions, knowledge of the water distribution in human stratum comeurn (SC) at various relative humidities is required. In previous studies it has been hypothesized that water is partly absorbed in the corneocytes and remains partly in the intercellular regions in separate domains, i.e., phase separated from the lipid lamellar domains. Absorption of water into the comeocytes is expected to increase the volume of the comeocytes. But, since comeocytes are entirely surrounded by a crystalline monolaye r of lipids (lipids bound to the cornified envelope), a change in the lipid density of this monolayer is not expected to occur. Therefore, it is hypothesized that swelling of the comeocytes has to occur without inducing a strong change in the total surface area of the comeocytes. Methods: A new technique was developed that allowed the water distribution in the SC at various hydration levels to be studied: cryo-planing in combination with cryo scanning electron microscopy (Cryo- SEM) to obtain flat cryo-sections across the entire stratum comeum. This is an excellent method to determine both the distribution of water as a function of localization in the SC and the shape (that is the cross section) of the comeocytes as a function of the hydration level from one single image. SC was cut into sheets of 8 x 8 mm 2 and equilibrated over various salt solutions with known relative hurrfidities. After folding and embedding in tissue-freezing medium, the samples were rapidly frozen in liquid propane and sectioned in an ultramicrotome at a sample temperature of-90øC and freeze-dried for 3 minutes at the same temperature to obtain contrast and sputtered with a 5nm thick layer of platinum. Thereafter, they were subjected to SEM at a temperature of-190øC. Water that was present in the original SC sectioned samples will have sublimated during the freeze-drying procedure from the flat surface, creating a relief (contrast) that can be identified as a lighter surface in the cryo-SEM's. The technique is illustrated in Figure I. Numerical parameters such as average cell thickness, circumference as well as the enclosed surface area could be calculated from the digital cryo- SEM's. Figure 1 Schetnatic presentation of the sequential events of the Cryo-Scan ing electron microscolby technique Cryo-planing in combination with cryo-SEM V._•.•Freshlobtainedsurfaces tissue fr. Embedding eezing C•yo-freezing planing Dehydration Cryo-SEM Results: Using this new method, it was discovered that water in the SC is neither homogeneously distributed, nor gradually increasing with depth. Dry skin and skin hydrated to 17% w/w was characterized by low contrast and cells show many undulations, particularly close to the comeocyte cell ends. At 57- 87%w/w water content the hydration level in the central part of SC is higher than in the superficial and deeper cell layers. Water domains are mainly present within the comeocytes and not in the intercellular regions (see Figure 2). At a very high hydration level (300 %w/w), the comeocytes are strongly swollen except for the deepest cell layers adjacent to the viable epidermis. The comeocytes in these layers are not 17%w/w and 300%w/w the average cell thickness increases linearly with the hydration level suggesting that swelling of cells mainly occurs in the direction perpendicular to the skin surface. At an increased hydration level, the comeocyte envelope surrounds the cell content more efficiently (see Figure 3) compensating for the increased cell volume. The changes in stratum corneum morphology with increasing water level have also been observed in dermatomed skin indicating that the observations are not simply an artifact due to the lack of underlying tissue.

2002 ANNUAL SCIENTIFIC MEETING 111 Figure 2A SC hydrated to 70%w/w reveals slightly swollen cells with a lighter appearance in the central part (see black asterisks) indicating the presence of water. In the upper and lower part the appearance of the SC is similar to that of dry skin (white asterisks). The white arrows indicate undulations. Figure 2B Two SC sheets hydrated to 90%w/w. The white arrow indicates the interface between the sheets. Both sheets show an increased hydration level in their central regions (lighter appearance, see black asterisks). Note that these cells are also strongly swollen. Discussion: Tape-stripping studies always show an exponentially decreasing distribution of inward penetrating chemical when going from the outside to the inside. Because the same intercellular lipid composition that acts as a barrier for the inwardly penetrating molecules is also anticipated to act as the harrier for the outwardly diffi•sing water molecules, an exponentially increasing wamr concentration was expected when going from the outside to the inside of the skin. However, a different distribution profile was found favoring the presence of water in the central regions with only minimal amounts of water in the upper and lower part (the stratum compactum) of the stratum corneum. This profile coincides with the presence of the Natural Moisturizing Factor (NMF). The sudden reduction in water levels in the stratum compactum prevents a complete dehydration whereas the presence of elevated amounts of water in the central layers allows the stratum corneum its flexibility and biochemistry {e.g., desquamation). The composition of the intercellular lipids in the stratum compacturn does not have to be different from that in the central and upper parts of the stratum corneum to cause a lower water concentration. We postulate that the intercellular lipids affect the permeation rates of water and other inwardly penetrating chemicals whereas intracellular polar ingredients like NMF affect the water quantities. Figures 3A and B Mean thickness of cells as a function of SC hydratlon 3.5 2.5 1.5 1 0.,5 0 100 200 300 Hydrat•on %w/w 100 90 80 70 60 50 40 30 20 10 0 400 (Itot)2/A as function of hydratlon level of stratum comeurn 18-26% w/w 57-87% w/w 292-332% w/w (NEBO (Na2C03} (H20} (/1) The mean cell thickness plotted as a ftruction of SC hydration. /1 linear relation is observed (B) The plotted as fitnction of SC hydration. A tlecrease indicates an increase in efficiency of the cornified envelope to surround the cell content