CELL MEMBRANE COMPLEX 449 t-butanol and heptane sequentially (36,56) have been used to extract “external lipids” such as wool wax or sebaceous matter from animal hairs. Such lipids are sometimes called ex- ternal, extrinsic, or even exogenous (57) and are primarily sebaceous in origin in humans. A large percentage of these lipids are not believed to be involved in the intercellular structure of animal hairs, but there is evidence (58) that some of this solvent-extractable lipid may be part of the structure of the surface lipids of hair fi bers. In the case of human hair, external lipids are normally removed by shampoo or sodium lauryl sulfate washing, by a combination of shampoo followed by incubation in hexane for only fi ve minutes (57), or in some cases by other solvents that most likely do not remove substantial amounts of internal hair lipids (e.g, ether or heptane). Removal of internal lipids not covalently bound to hair. Hair-swelling organic solvents alone or in combination with a second lipid solvent are used to remove internal lipids that are part of the internal structure of hair fi bers (of the CMC) but not covalently bonded to the hair protein structures. Solvents such as chloroform/methanol (44,56), methanol (52), ethanol (6), formic acid (50), n-propanol/water (50), or acetone (52) have been used to extract internal matter from animal hairs. The most frequently used solvent for removal of inter- nal lipids has been chloroform/methanol (70/30), although other mixtures have been used. Normally soxhlet extraction is employed however, multiple room temperature extractions have also been employed (57). Although formic acid and n-propanol/water (generally 1:1) do remove some internal lipids, these two solvents also remove some hair proteins of the CMC (most likely from the delta layers and possibly from other regions of the fi bers), as will be described in the section entitled “Proteins of the CMC.” Removal of covalently bound hair lipids plus salts insoluble in lipid solvents. Alkaline hydrolysis or methanolic alkali is used to remove covalently bound hair lipids. This technique can be used to remove total hair lipids, but it is generally used after extraction of external and internal lipids that are not covalently bound to the fi bers. Those lipids that are covalently bound at or near the fi ber surface in the cuticle are generally removed with potassium t- butoxide in t-butanol (bulky cleaving agent in a bulky solvent) (59), and total covalently bound lipids are generally removed with potassium hydroxide in methanol because alkali in a swelling solvent like methanol penetrates well into hair. In addition to covalently bound lipids, Wertz and Downing (34) have suggested that salts of cholesterol sulfate bind ionically to cationic groups on the hair proteins and will be insoluble in chloroform/methanol and therefore will remain in the fi brous residue after extraction with organic solvents. Korner et al. (48) used a solution of chloroform/metha- nol/aqueous potassium chloride to extract CMC lipids from wool and human hair and formed liposomes from the extracts. Since cholesterol sulfate is essential for liposome for- mation of keratin fi ber extracts, this solvent system must extract cholesterol sulfate too. TOTAL LIPIDS IN HAIR FIBERS The total amount of lipid extractable from hair is generally 1% to 9% of the weight of the hair (57,60). Masukawa et al. (57) have provided a relatively thorough analysis of hu- man hair lipids. In their study, the total hair lipid composition from 44 Japanese women, ages 1 to 81, was examined. The lipids were extracted/removed from hair in varying pro- cedures, to allow for analysis of several lipids and covalently bound 18-MEA. In this study, total fatty acids and 18-MEA were determined, but other important fatty acids,

JOURNAL OF COSMETIC SCIENCE 450 both covalently bound and non-covalently bound, such as palmitic, stearic, oleic and palmitoleic acids, were not quantitated and were found in signifi cant quantities in other studies (13,34,35,42,56). Cholesterol sulfate was also not determined in this effort by Masukawa et al. Logan et al. (56) analyzed human hair by extracting it with a chloroform/methanol azeo- trope for fi ve hours after surface lipids had been removed with t-butanol and heptane, and they found 23% palmitic, 25% palmitoleic, 4% stearic, and 13% oleic acids, as well as other fatty acids. These are all non-covalently bound fatty acids, with 39% of the total fatty acids being unsaturated (palmitoleic and oleic acids, but it is possible that other unsaturated fatty acids were present). Weitkamp et al. (61), by analysis of solvent-extracted lipids from pooled adult Caucasian human hair clippings, found 51% of the total fatty acids to be unsaturated, with palmitoleic and oleic acids as the principal unsaturated fatty acids, but other unsaturated fatty acids were found in these extracts. Masukawa et al. (57) initially shampooed the hair and then washed it with hexane, allow- ing a fi ve-minute incubation time. The hexane wash was determined by plotting the amount of lipid extracted versus the square root of the time of the hexane wash. The time that diffusion of lipids from the interior of the fi ber began was determined graphically, a reasonable approach to removing external lipid soils from the fi bers, presumably leaving most of the internal and structural lipids in the hair. The lipids removed were separated into eight groups by Masukawa et al., and their data are summarized in Table II. These data show that approximately 58% of the total lipids in hair under these conditions are fatty acids, some covalently bonded, with others exist- ing as free and non-covalently bound fatty acids. The total fatty acids found were 14.4 mg/gm of hair, but only 0.3 mg/gm hair of 18-MEA were found. Wertz and Downing (34) found 1.31 to 2.1 mg/gm of 18-MEA in four different human hair samples (three from individuals and one pooled hair sample, presumably Caucasian hair) and in a later paper (35) cited 4.0 mg/gm total integral (covalently bound) fatty acids with 40.5% as 18-MEA for human hair or 1.6 mg/gm 18-MEA. Since most 18-MEA estimates in wool fi ber are close to 1 mg/gm or higher and human hair contains more cuticle layers than wool fi ber, one would expect more covalently bound fatty acids in human hair than in wool Table II Lipids in Human Hair, from Masukawa et al. (57) and Wertz and Downing (35) Type of lipid mg/gm hair Percentage of total lipid Hydrocarbons 2.4 9.7 Squalene 0.7 2.8 Wax easters 4.9 19.8 Triglycerides 0.5 2.0 Total fatty acids 14.4 58.1 (58)** Total covalent fatty acids — (4.0)* Cholesterol 1.3 (0.6)* 5.2 Cholesterol sulfate — (2.9)* Ceramides 0.29 (0.5)* 1.2 18-MEA 0.30 (1.6 )* 1.2 Totals 24.79 100% *Data in parenthesis by Wertz and Downing (35), not in parenthesis by Masukawa et al. (57). **See Logan et al. (56) for a breakdown of the actual fatty acids in human hair.