CELL MEMBRANE COMPLEX 451 fi ber. Masukawa did not list amounts for total covalently bound fatty acids, only for 18- MEA. Therefore, I will use the data for Wertz and Downing (35) for total covalently bound fatty acids. Since Masukawa et al. (57) found 14.3 mg/gm total fatty acid, but did not determine the total covalently bound fatty acids, and Wertz and Downing found 4 mg/gm total cova- lently bound fatty acids, then the Masukawa et al. fi nding most likely represents or is closer to the total amount of non-covalently bound fatty acids in human hair. Therefore, if we assume human hair has approximately 14 mg/gm of non-covalently bound fatty acids and we assume about ½ the equivalent amount of free lipid in the cuticle relative to cova- lently bound fatty acid, this provides 2 mg/gm of free fatty acid in cuticle layers and leaves about 12 mg/gm of non-covalently bound fatty acids. If we then assume 2 mg/gm of fatty acid as intracellular lipid, that leaves 10 mg/gm of fatty acids in the cortex–cortex CMC. Therefore, with these approximations, about 10 mg/gm of fatty acids will exist in the “bilayers” of the CMC of the cortex of human hair along with cholesterol, cholesterol sulfate, and ceramide, (see Figure 3). Wertz and Downing (35) also found cholesterol (0.6 mg/gm), cholesterol sulfate (2.9 mg/gm), and ceramides (0.5 mg/gm) in their alkaline hydrolysates from human hair after removal of all free lipids by chloroform–methanol extraction. These same scientists also found these same lipid components in hair from sheep, dogs, pigs, cattle, and humans: cholesterol (0.3–1.4 mg/gm), ceramides (0.6–1.4 mg/gm), and cholesterol sulfate (0.7– 3.3 mg/gm) (36). Examination of these data from different laboratories suggests the fol- lowing ingredients in these approximate ratios as the principal components of the bilayers of the cortex–cortex CMC for human hair: Lipid component Approximate amount Approximate relative amount Fatty acids 10 mg/gm hair 10 Cholesterol sulfate 0.7 to 3.3 mg/gm 2 Cholesterol 0.6 to 1.2 mg/gm 1 Ceramides 0.6 to 1.4 mg/gm 1 After shampooing an appreciable amount of free lipid remains on the hair surface. Shaw (62) sug- gested that washing hair with ether or shampoos in a one-step application leaves virtu- ally the entire hair surface free of lipid (removable by surfactants or solvent) and that differences in cleaning effi ciencies of surfactants relate to the amounts of internal lipid removed. Recent XPS data show that shampooing does remove some free lipid from the surface of hair, but even after shampooing a signifi cant amount of free lipid remains in the surface layers, that is in the top 3 to 5 nm (58). Free lipid in surface layers affects the isoelectric point of wool and hair. Capablanca and Watt (63) examined wool fi ber that had been washed with detergent (Lissapol) and extracted with various solvents using a streaming potential method to estimate the effect of free-lipid (non-covalently bound fatty acids) in the surface layers on the isoelectric point of wool fi ber. These scientists found an appreciable effect of free lipid on the isoelectric point. The surfactant-washed wool (containing the most free lipid) provided an isoelectric point of These ratios are most likely not exact, but they show a large amount of fatty acid followed by cholesterol sulfate and smaller amounts of cholesterol and ceramide.

JOURNAL OF COSMETIC SCIENCE 452 3.3. The isoelectric point of wool increased as the effectiveness of the solvent system increased, with the most effective lipid solvent providing wool with an isoelectric point of 4.5. These data show that the true isoelectric point of the surface proteins of wool fi ber is closer to 4.5 than 3.5 and that free fatty acids in the surface layers are an important and essential component of the surface of animal hairs. Thus, the more free lipid that is present in these surface layers, the lower the isoelectric point of keratin fi bers. Therefore, all free lipid is not totally removed and should not be totally removed from the surface layers by the shampooing of hair or surfactant scouring of wool fi ber. In addition, free fatty acids are important to the isoelectric point of animal hair fi bers. Furthermore, the amount of free lipid on the surface of hair fi bers will infl uence hair friction, surface energy, and a whole range of important properties, including the adsorption of surfactants and other ingredients onto human hair and wool fi bers. PROTEINS OF THE CMC The schematic in Figure 2 depicts cell membrane proteins and multiple layers of pro- teins in the delta layer of the cuticle–cuticle CMC analogous to the delta layer of the cortex (5,9). The structures and composition of the proteins of the CMC are still not adequately characterized. The primary reason for this gap is that it is extremely diffi cult to isolate proteins from only the cell membranes or only the delta layer, and this diffi - culty has been the primary obstacle to our understanding the composition and structure of the proteins of this important region of the fi ber, but nevertheless I will review this area in this section. To date, much more scientifi c attention has been given to the analy- sis of cuticle cell membranes than to those of the cortex therefore, I will begin this discussion on the proteins in the cuticle cell membranes. PROTEINS IN THE CUTICLE CELL MEMBRANES The proteins of the cuticle cell membranes are associated with the Allworden reaction (14) as described earlier. The membranous epicuticle supports 18-MEA and is attached to the A-layer on the top of cuticle cells and has been isolated by shaking animal hair fi bers dur- ing Allworden sac formation with chlorine water and has subsequently been analyzed for amino acids. Perhaps the most quoted and “reliable” amino acid analysis of the Allworden membrane has been provided by Allen and coworkers (30), summarized in Table III. Since the attachment of 18-MEA to hair proteins is through thioester linkages and the cuticle cell membrane protein is cross-linked by cystine bridges, Negri et al. (12) proposed that the lipid layer must be attached to an ultra-high sulfur protein (UHSP) that can pro- vide attachment sites at approximately 1-nm spacings along the top of its folded chains. Zahn et al. (31) have provided indirect evidence by a multiple regression technique for the presence of approximately 51% UHSP, 42% loricrin, and 7% involucrin in the All- worden membrane as analyzed by Allen et al. (30), that is, in the cuticle cell membranes of wool fi ber or the 13–15 nm of the hair fi ber surface underlying the F-layer or 18-MEA. See Table III describing the amino acid analyses of these and other important proteins adapted from the paper by Zahn, et al. (31). From the results of this work and from previous work on the cell envelope of stratum corneum by Steinert and Marekov (68) and Jarnik et al. (69), Zahn et al. concluded that