PROTEINS OF HAIR 97 Table II Classification, Protein Extraction Yields, and Electrophoretic Intensity of LSP/HSP of the Different Samples Studied % Extracted Intensity of LSP Classification Sample Hair color proteins and HSP (see text) French Blond 0.6 Belgian Light blond 0.3 German Blond 1.2 Italian Light blond 0.9 Indian Brown 3.0 No LSP a LSP HSP b Caucasian Medium brown 7.2 LSP HSP c South American Black 3.6 Indonesian Black 3.9 Polish Dark brown 6.6 Belgian Dark brown 5.9 Indian Black 18.6 LSP ) HSP d North African Black 19.3 African Black 18.7 French Black 19.5 these observations allowed us to select the middle part of the hair as a model in our study. In terms of molecular weights and mobilities at pH 8.9 we found no influence of the ethnic origin of the hair. In the same way, no significant differences were found in the amino acid compositions (Table III) of hair from different racial origins or according to Table III Amino Acid Composition of Hair Samples (g/100 g) Amino acid European Asiatic African South American Cysteic acid 0.9 1.0 0.5 0.7 Aspattic acid 6.0 6.5 6.7 6.1 Threonine 7.2 7.1 7.2 7.2 Serine 10.0 9.8 10.4 9.8 Glutamic acid 14.9 ! 5.0 15.0 14.7 Proline 7.5 7. ! 6.2 7.5 Glycine 3.6 3.7 3.3 3.9 Alanine 3.1 3.1 3.4 3.4 Valine 5.1 5.2 5.2 5.2 Cystine 13.7 12.6 13.5 12.7 Methionine 0.7 0.5 0.7 0.6 Isoleucine 2.7 2.8 2.6 2.8 Leucine 6.7 6.7 7.1 6.8 Tyrosine 2.4 2.6 2.6 2.7 Phenylalanine 2.3 2.7 2.3 2.6 Lysine 2.8 3.1 3.1 2.9 Histidine 1.2 1.1 1.2 1.1 Arginine 9.2 9.3 8.9 9.1 Lanthionine 0 0 0 0

98 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS donors' ages, suggesting a chemical identity of the proteins. Nevertheless, a surprising fact is that the lighter the hair color, the lesser amount of extracted proteins. The similar electrophoretic patterns of a Caucasian black hair and of an African black hair (Figure 4) illustrate this statement. For a comparable amount of analyzed proteins, intensity of LSP increases with the yield in the protein extraction. The 35-Kd LSP polypeptide is yield-detectable only when this LSP group is very intense. Intermediate patterns (LSP-detectable but less intense that HSP) are obtained with medium brown hair. The highest yields of extracted proteins were obtained from dark brown and black hair samples, and the lowest yields were obtained with blond hair. These data suggest that the low-sulfur and high-sulfur proteins from black and blond hair would behave differently in the reducing buffer and that these proteins are more cross-linked in the blond hair sample than in the black hair sample. Furthermore, this cross-linking is via bonds that are not reducible. The analysis of blond hair proteins provides results similar to that of the weathered Caucasian brown hair. Light and oxygen are well known to significantly affect the protein structure, by promoting the formation of photo-induced cross-links (8). The role of melanin photoprotection is well recognized and thus the less efficient protectiofi of the complex eumelanin-pheomelanin in blond hair might explain the increase in new types of bonds (15). The chemical analysis of weathered hair (16,17) revealed that the sulfur content changed very little but that the cystine content presented a marked loss. Crewther (18) showed that the solubility of weathered hair in urea-bisulfite medium decreased. It can be therefore suggested that in such a case other types of cross-links are initiated to compensate for the loss in disulfide bridges. In blond hair, as in weathered hair, the proteins are non- or hardly extractible, showing that non-reducible cross-links are present but that the exact nature of these bonds remains unknown. In conclusion, we summarize the following results: First, no major differences were found in the expression of the SCM polypeptides of the hair with regard to its racial origin. Second, it is of interest to note that the lowest protein recoveries were from those hair samples with the least pigment (pheomelanin). In the case of weathered hair, this result could be due to photochemical cross-linking or other factors. Sulfur protein composition in light hair and in weathered hair samples show the same electrophoretic pattern changes, suggesting that both of these types of hair contain more non-reducible cross-links than darker hair samples. The differences between dark and light hair seem to be related to the number of reducible cross-links, indicating that the formation of these cross-links may be either under genetic control or in conjunction with pigment production. ACKNOWLEDGMENTS We wish to acknowledge Dr. Bauer and his group for supplying the human tresses, Dr. Bore and Dr. De Labbey for amino acid analysis, and Professor M. Prunieras for helpful discussions and critically reviewing the manuscript. The authors would like to thank Mr. A. Reinhard for the photography. REFERENCES (1) R. C. Marshall and J. M. Gillespie, Comparison of samples of human hair by two dimensional elec- rrophoresis,J. Forensic. Sci. Soc., 22, 377-388 (1982).