IN VIVO CONFOCAL MICROSCOPY 99 Figure 9. Semi-thin histological section of a Negroid skin sample exhibiting melanin clusters above the nucleus of basal cells. Courtesy of F. Fiat. (23). The abrupt mismatch in the refractive index between the cytoplasm (-1.33) and the protein content of the melanosomes (-1.55) (24) is large enough to generate a strong reflection at the surface of the plasma membrane of the melanosome. Thus, melanosomes could act as myriads of nanomirrors reflecting sunbeams and protecting the nucleus against injury. This hypothesis suggests that the wavelength independent reflection process would be effective in the whole light spectrum and could contribute to the limiting of the penetration of light into the nucleus. The residual part penetrating into the melanosome would then be scattered and absorbed by the melanin. The residual protection by absorption is indeed much more efficient in the UV than in the visible part Figure 10. SEM pictures showing the localization and morphology of melanosomes. (a) In a basal cell, they form tightly packed clusters within the cytoplasm. (b) At a higher magnification, their shape and size can be evaluated. White bar = I pm. Courtesy of A.M. Minondo.

100 JOURNAL OF COSMETIC SCIENCE of the sun spectrum. It is inconceivable to limit the protection to absorption only. This would lead to a colossal storage of energy that would need to be dissipated by the cell. Such a process still remains unknown. The effects of sun exposure on the physiology of the epidermis have largely been reported in the literature. The well-known hyperkeratosis (25,26) has been measured i, vivo for the first time by confocal microscopy (Figure 5). This demonstrates the unique perfor- mance of this device in measuring noninvasively the thickness of the epidermal layers. Today, confocal microscopy remains the only technology providing rapid thickness measurements of the epidermal layers with a precision on the order of the micrometer. This avoids the need to perform painful biopsies on human volunteers. A 25% increase in stratum corneum thickness on each of the two subjects was recorded for a period of one month post-exposure. In this experiment, the epidermal hyperplasia of the epidermis i, vivo could not be determined. The only explanation for this is that such a measurement needs a marker formed by the bright signal of the basal keratinocytes that was lacking after exposure. Sun-exposed forearms led to unusual images through the confocal microscope. These unusual images were not observed before exposure and disappeared one month after exposure. These transient changes lasted for at least three weeks and consequently could not be attributed to a random phenomenon. The bright inclusions in the nuclei of numerous granular and spinous keratinocytes (Figure 6) can be interpreted as local condensations of chromatin. This would generate an index mismatch and consequently would lead to high reflections at the interface. Condensation of the chromatin seems to be in accordance with the histologic description of sunburn cells (27-29). Brightly spotted filaments in the intercellular spaces between keratinocytes (Figure 7) could correspond to aligned melanosomes migrating towards the surface via the den- drites of melanocytes (20,22). It has been stated that in normal conditions the dendritic epidermal cell lines could not be clearly identified with the i, vivo confocal microscope. This experiment has demonstrated, to the contrary, that only the dendrites but not the body of the melanocytes can be observed in exposed skin. The most striking transient observation turned out to be the total lack of melanosome caps in basal keratinocytes (Figure 8b) for a period of three weeks following sun expo- sure. A possible first explanation could be that photodegradation (30) or redox mecha- nisms (31) liberate the melanin contents of the melanosomes. Their geometry would thus not permit any further optical signal except for a diffuse absorption that is respon- sible for the tanned aspect of the upper layers. A second explanation could be related to the slow formation of melanosomes by the melanocytes, limiting both their increased migration towards the skin surface and the renewal of the reservoir in the basal kera- tinocytes (19-20). CONCLUSION The new phototype of CLSM dedicated to the exploration of human skin i, vivo provides major improvements when compared to previous systems. Sharper images were recorded with the blue laser line (488 nm), while the red line (647 nm) provided better pen-