We have examined the quantity and composition of melanin in both photoprotected (volar upper arm) and chronically photoexposed (dorsal forearm) skin from a range of different ethnic skin types including African, Indian, Mexican, Chinese and European. The most lightly pigmented (European, Chinese and Mexican) skin types have approximately half as much epidermal melanin as the most darkly pigmented (African and Indian) skin types. However, the composition of melanin in these lighter skin types is comparatively more enriched with lightly coloured, alkali-soluble melanin components (up to three-fold). Regardless of ethnicity, epidermal melanin content is significantly greater in chronically photoexposed skin than it is in corresponding photoprotected skin (up to two-fold). However, by comparison there is only a modest enrichment of lightly coloured, alkali soluble melanin components in photoprotected skin (up to 1.3-fold). Analysis of melanosomes extracted from the epidermis in these subjects indicates that the proportion of spheroidal melanosomes is low in all skin types examined (<10%). This suggests that in human skin, pheomelanin is a very minor component of epidermal melanin, even in the lightest (European) skin types. Analysis of melanosome size revealed a significant and progressive variation in size with ethnicity: African skin having the largest melanosomes followed in turn by Indian, Mexican, Chinese and European. On the basis of these findings, we propose that variation in skin pigmentation is strongly influenced by both the amount and the composition (or colour) of the melanin in the epidermis. Variation in melanosome size may also play a significant role. However, the data also suggest that in human skin there are subtle differences in the mechanisms associated with the maintenance of constitutive pigmentation and facultative hyperpigmentation, respectively.

Objective measurements of human skin colour were made with a tristimulus (L*a*b*) chromameter in a range of different ethnic skin types. These were compared with biochemical measurements of melanin content, melanin composition and melanosome size in skin biopsies obtained from the same sites. L*, a* and b* values were found to vary significantly with ethnicity. In general, constitutively dark skin types have lower L* values, higher a* values and higher b* values than constitutively light skin types. Total epidermal melanin content appears to be the primary determinant of L* values in human skin (r = -0.88; P < 0.00001), whilst melanosome size also has a significant but more subtle influence on L* values (r = -0.73; P < 0.00001). There is also a strong positive contribution to a* values from epidermal melanin (r = 0.66, P < 0.00001), which accounts for the ethnic variation in a* values observed in this study. Melanin is also a major contributor to b* values in lighter skin types (r = 0.71, P < 0.00001). However, this relationship breaks down in darker skin types where b* values actually reach a maximum and then decrease as the concentration of melanin in the skin increases. This appears to be because of optical masking of yellow light by high concentrations of melanin in the epidermis. Analysis of the relationships between L*, a* and b* values in human skin indicate that they are very closely interrelated, and suggest that the optical properties of melanin in the epidermis are very similar to those of a dye on a fabric substrate.

A combination of techniques, including high-performance liquid chromatography (HPLC), spectrophotometric measurements, and a novel method for quantifying melanosome morphology, were applied to the analysis of melanin content and composition in highly pigmented (Fitzpatrick type V and VI) human skin. We found that total epidermal melanin content is significantly elevated in photoexposed type V and VI skin (approximately 1.6 x), while analysis of individual melanin components suggests that pheomelanin content increases only slightly, whereas 5,6-dihydroxyindole-2-carboxylic acid (DHICA)-eumelanin and to a greater extent 5,6-dihydroxyindole (DHI)-eumelanin content are both markedly elevated. Analysis of the relative composition of epidermal melanin in these subjects revealed that DHI-eumelanin is the largest single component (approximately 60-70%), followed by DHICA-eumelanin (25-35%), with pheomelanin being a relatively minor component (2-8%). Moreover, there was a comparative enrichment of DHI-eumelanin at photoexposed sites, with a corresponding decline in the relative contributions from DHICA-eumelanin and pheomelanin. There was also a good correlation and close agreement between the concentration of spheroidal melanosomes determined by morphological image analysis and the concentration of pheomelanin determined by a combination of HPLC and spectrophotometric analysis (r = 0.89, P < 0.02). This study demonstrates the usefulness of melanosome morphology analysis as a sensitive new method for the quantification of melanin composition in human skin. The data also suggest that DHI-eumelanin formation is the dominant pathway for melanin synthesis in heavily pigmented (Fitzpatrick V and VI) skin types in vivo, and is the favoured pathway when melanin production is increased in chronically photoexposed skin.

The relative expression of a number of key mediators of human pigmentation including tyrosinase, tyrosinase related protein-1 (TYRP1), endothelin-1 and adrenocorticotrophic hormone (ACTH) proteins were analysed and quantified in immunohistochemically stained skin sections using semiquantitative computer assisted image analysis. Comparisons were made between a range of different ethnic skin types including European, Chinese, Mexican, Indian and African at both chronically photoexposed and photoprotected sites. Melanocyte number varied little with ethnicity except in the European group which had 60-80% more melanocytes than other skin types (P < 0.01, n = 10; Student Neuman-Keuls). However, melanocyte number was increased approximately twofold in chronically photoexposed skin of all ethnic groups (P < 0.001, n = 48; paired t-test). Tyrosinase protein expression in melanocytes did not vary with ethnicity, but TYRP1 protein was significantly elevated (approximately 2.6-fold) in darkly pigmented African and Indian skin types compared with lightly pigmented Mexican, Chinese and European skin types. In melanocytes from chronically photoexposed skin, there was a modest but significant increase in the expression of tyrosinase protein (approximately 1.2-fold, P < 0.001, n = 48; paired t-test), together with a significant and slightly larger increase in the expression of TYRP1 protein (approximately 1.6-fold, P < 0.005, n = 48; paired t-test). In contrast, the expression of endothelin-1 and ACTH showed no significant variation with either ethnicity or photoexposure. These data are consistent with the view that maintenance of a chronically hyperpigmented phenotype in chronically photoexposed human skin is largely the result of a stable increase in the number of tyrosinase positive melanocytes at these sites. Moreover, the observed ethnic variation in TYRP1 protein expression suggests that TYRP1 may play a significant role in mediating ethnic differences in melanogenesis and constitutive skin pigmentation in vivo.

No abstract