Glycerol Regulates Stratum Corneum Hydration in Sebaceous Gland Deficient (Asebia) Mice

Home , Stratum corneum

CORE Metadata, citation and similar papers at core.ac.uk Provided by Elsevier - Publisher Connector ORIGINAL ARTICLE See related Commentary on page iv Glycerol Regulates Stratum Corneum Hydration in Sebaceous Gland De¢cient (Asebia) Mice

Joachim W. Fluhr,nz1 Man Mao-Qiang,n1 Barbara E. Brown,n Philip W.Wertz,w Debra Crumrine,n John P. Sundberg,y Kenneth R. Feingold,nz and Peter M. Eliasn nDermatology and zMedical (Metabolism) Services,Veterans Administration Medical Center, and Departments of nDermatology and zMedicine, University of California, San Francisco, California, USA; wDows Institute, University of Iowa College of Dentistry, Iowa City, Iowa, USA; yThe Jackson Laboratory, Bar Harbor, Maine, USA; zDepartment of Dermatology, Friedrich-Schiller University, Jena, Germany

The only known function of human sebaceous glands glands, normalized stratum corneum hydration, and is the provocation of acne. We assessed here whether the glycerol content of asebia stratum corneum was sebum in£uences stratum corneum hydration or per- 85% lower than in normal stratum corneum. In con- meability barrier function in asebia J1 and 2 J mice, trast, another potent endogenous humectant (urea) did with profound sebaceous gland hypoplasia. Asebia J1 not correct the abnormality.The importance of glycerol mice showed normal permeability barrier homeostasis generation from triglyceride in sebaceous glands for and extracellular lamellar membrane structures, but stratum corneum hydration was demonstrated further they displayed epidermal hyperplasia, in£ammation, by (i) the absence of sebaceous-gland-associated lipase and decreased (450%) stratum corneum hydration, activity in asebia mice, whereas abundant enzyme ac- associated with a reduction in sebaceous gland lipids tivity was present in the glands of control mice; and (wax diesters/monoesters, sterol esters). The triglyceride (ii) the inability of high concentrations of topical trigly- content of both asebia and control stratum corneum ceride to correct the hydration abnormality, despite the was low, consistent with high rates of triglyceride hy- presence of abundant lipase activity in asebia stratum drolysis within the normal pilosebaceous apparatus, corneum. These results show that sebaceous-gland-de- despite high rates of triglyceride synthesis. Although a rived glycerol is a major contributor to stratum cor- mixture of synthetic, sebum-like lipids (sterol/wax neum hydration. Key words: asebia mice/barrier function/ esters, triglycerides) did not restore normal stratum glycerol/hydration/lipase/SCD1/sebum/triglycerides. J Invest corneum hydration to asebia skin, topical glycerol, the Dermatol 120:728 ^737, 2003 putative product of triglyceride hydrolysis in sebaceous

utaneous sebaceous glands deliver sebum to the skin skin, such as that in prepubertal children, exhibits normal basal surface through a process of continuous, holocrine barrier function, it has been assumed that sebum does not secretion. Sebum comprises a complex lipid mix- in£uence epidermal permeability barrier function (Kligman, ture that is enriched in wax monoesters and diesters, 1963). But the barrier recovery kinetics of sebaceous-gland-poor with substantial species-speci¢c di¡erences in trigly- versus -enriched skin after acute insults has not been examined, Ccerides (TGs), sterol esters, free fatty acids (FFAs), and squalene and this form of stress test is a more sensitive indicator of perme- (Thody and Shuster, 1989; Stewart and Downing, 1991). Although ability barrier status than are basal assessments (Feingold and a variety of functions have been proposed for cutaneous sebac- Elias, 1999). Moreover, products of Meibomian glands (modi¢ed eous-gland-derived lipids (Nicolaides, 1974; Thody and Shuster, sebaceous glands) in£uence barrier function in conjunctival 1989), the predominant view holds that the principal role of epithelia (Bron and Ti¡any, 1998). Furthermore, after sebum is sebum in humans is negative, i.e., its well-accepted pathophysio- deposited on the surface of the SC, it inspissates into the inter- logic role in the provocation of acne (Kligman, 1963; Stewart stices (Kligman and Shelley, 1958) and interacts with the extra- and Downing, 1991). Because sebaceous-gland-impoverished cellular lamellar bilayer system (Sheu et al, 1999). There, its constituent fatty acids, which lack essential fatty acids, could di- lute or replace the linoleic acid in SC acylceramides, as proposed Manuscript received September 14, 2002; revised November 19, 2002; for follicular epithelia (Stewart et al, 1986). As a net result of one accepted for publication December 17, 2002 or more of these processes, sebum could in£uence permeability 1Each contributed equally, and therefore they should be considered co- barrier homeostasis. ¢rst authors. To assess two potential functions of cutaneous sebaceous Reprint requests to: Peter M. Elias, M.D., Dermatology Service (190), glands, we compared permeability barrier homeostasis and SC Veterans A¡airs Medical Center, 4150 Clement Street, San Francisco, CA hydration in two closely related models where sebaceous glands 94121; Email: [email protected] Abbreviations: abJ/abJ, asebia J; ab2J/ab2J, asebia 2 J; AQP3, aquaporin 3; are largely absent, the asebia mouse. Asebia mice display not only FFA, free fatty acid; PPARa, peroxisomal proliferator activator a; ScdabJ/ profound sebaceous gland hypoplasia, but also other cutaneous ScdabJ, asebia; SCD1, stearyl CoA desaturase 1 (protein); SCD1, stearyl CoA abnormalities, including mild scaling, patchy scarring alopecia, desaturase 1 (gene); TEWL, transepidermal water loss; TG, triglyceride. and epidermal hyperplasia (Gates and Karasek, 1965; Josefowicz

0022-202X/03/$15.00 . Copyright r 2003 by The Society for Investigative Dermatology, Inc.

728 VOL. 120, NO. 5 MAY 2003 ABNORMAL STRATUM CORNEUM HYDRATION IN ASEBIA MICE 729 and Hardy, 1978; Sundberg and King, 1996). Although recent aqueous toluidine blue for light microscopy and to visualize mast cell studies have shown that mutations in the gene that encodes stear- metachromasia. yl CoA desaturase 1 (SCD1) are responsible for both the asebia J1 and asebia 2 J phenotypes (Zheng et al, 1999; Sundberg et al, Lipid content analysis Full-thickness mouse skin was excised from 2000), the basis for sebaceous gland hypoplasia in these models is shaved asebia J1 and control animals, and incubated with 10 mM ethylenediamine tetraacetic acid (EDTA), pH 7.4, at 371C for 30 min to not known. We report here that, despite the presence of epider- separate epidermis from dermis (Grubauer et al, 1989). Nucleated mal hyperplasia, asebia J1 mice display normal permeability bar- epidermal cells were separated from SC by further incubation with 0.5% rier homeostasis but a profound abnormality in SC hydration. trypsin in phosphate-bu¡ered saline for 2 h, followed by vortexing. SC The hydration abnormality in asebia skin could be attributed sheets were freeze-dried, weighed, and total lipids were extracted (Bligh further to decreased glycerol generation due to a lack of endogen- and Dyer, 1959), dried, weighed, and stored at ^701C until analyzed. ous, sebaceous-gland-derived TG and associated lipase activity. Neutral lipids were separated from polar lipids by high performance The hydration abnormality in asebia mice could explain the in- thin layer chromatography, using a CAMAG linomat IV Autospotter creased propensity of sebum-depleted skin sites of humans to de- (CAMAG Scienti¢c, Washington, NC), extracted, and solubilized in velop eczematous skin disorders, particularly when subjected to chloroform:methanol (2:1 vol). Nonpolar lipids were then fractionated and quantitated (Law et al, 1995). Brie£y, 10 10 cm glass-backed TLC plates, xeric stress. coated with 0.25 mm thick silica gel G (Adsorbosil-plus-1; Alltech Associates, Deer¢eld, IL), were washed with chloroform:methanol (2:1) and activated in a 1101C oven; the adsorbent was scored into 6 mm wide METHODS lanes. Calibrated glass capillary tubes were used to apply 5 ml samples and chromatograms were developed in n-hexane:benzene (1:1 vol) to Animals and experimental procedures Male asebia J1 mice (ABJ/Le- 95 cm. Finally, chromatograms were developed to 5 cm with hexane:ethyl ScdabJ/ScdabJ) (The Jackson Laboratory, Bar Harbor, ME) and their normal ether:acetic acid (70:30:1), charred, and quantitated by photodensitometry. heterozygote ( þ /abJ) and wild-type ( þ / þ or þ /?) littermates, between Lipid standards were obtained from Sigma (St. Louis, MO). 6 and 10 wk of age, were utilized for most studies. Asebia 2 J (ab2J/ab2J, (The Jackson Laboratory) mice were utilized in some experiments, where Glycerol content analysis Asebia J1 and 2 J mice as well as wild-type indicated. All mice were maintained in the Animal Care Facility of the littermates were carefully shaved 24 h before collecting total SC by pooling Veterans A¡airs Medical Center, San Francisco, in a temperature- and sequential tape strips down to the glistening layer in all groups (D-Squame humidity-controlled room, and fed standard laboratory chow (Teklad 22/5 disks, CuDerm Corporation, Dallas, TX). Because of di¡erences in SC Rodent Diet) and acidi¢ed tap water ad libitum. All functional studies were thickness and cohesion ( ¼ protein per stripping) between the two groups, performed in the Animal Care Facility, and mice were anesthetized with data re£ect total glycerol/total SC protein from equivalent surface areas. chloral hydrate and/or iso£orane. SC hydration was quantitated as changes The disks from each site were pooled (four to six disks each) and soaked in electrical capacitance in arbitrary units (Corneometer CM 820, Courage in 500 ml 1% Triton-X100 (Sigma) in water, sonicated 5 min, and then & Khazaka, Cologne, Germany) in asebia J1 and 2 J and wild-type mice, stored overnight at 41C. Blank D-Squame disks were treated similarly as a over skin sites (mean of three measurements in each animal) that had background control in all assays. Samples were vortexed thoroughly and 50 been shaved 24 h previously. Corneometry (capacitance measurements) ml aliquots of each solution were taken to assay glycerol content, using a provides a reliable and reproducible measure of SC hydration (e.g., Fluhr previously described radiometric method (Newsholme, 1974). Glycerol standard curves ranging from 0.0 to 8.8 nmol per ml were generated. et al, 1999; Gloor et al, 2002). The slight £akiness of untreated asebia SC did 14 not alter the reproducibility of measurements taken from or adjacent to Each assay volume contained 0.05 mCi of glycerol [ C(U)], 160 mCi per previously shaved sites. mmol (American Radiolabeled Chemicals, St. Louis, MO), 0.002 U of In some animals, immediately after capacitance measurements, basal glycerol kinase from Bacillis stearothermophilus (85 units per mg at 251C with adenosine-50 -triphosphate and glycerol as substrate), and 0.3 mg transepidermal water loss (TEWL) was measured with an electrolytic 0 water analyzer (MEECO, Warrington, PA). The barrier was then adenosine-5 -triphosphate (Roche Molecular Biochemicals, Indianapolis, abrogated over three separate sites in each animal by sequential cellophane IN). A ¢nal concentration of 0.5% Triton X-100 was maintained in (Scotch-types, 3M, Minneapolis, MN) tape strippings until TEWL levels standard curve as well as samples. Incubation times were 4 min at 271C, exceeded 2 mg per cm2 per h. Permeability barrier recovery rates were and protein assays were performed on the remaining disks and soak compared in asebia J1 (n ¼17) versus wild-type ( þ /?) (n ¼10) mice 3 and solutions, as described previously (Fluhr et al,2001). 6 h after acute disruption. Statistical analyses were performed with Prism 3 software (GraphPad, To assess the potential role of sebum and its products in SC hydration, San Diego, CA), using paired and unpaired Student’s t tests. Where a non- we performed the following replenishment studies: (i) a synthetic Gaussian distribution occurred, a nonparametric Wilcoxon test was used. sebaceous lipid mixture, containing a TG (tripalmitate, 250 mg), a wax monoester (stearyl stearate, 125 mg), and a cholesterol ester (cholesterol palmitate, 25 mg), or (ii) tripalmitate alone (1%, 5%, and 10%) in RESULTS propylene glycol:ethanol (7:3 vol) versus vehicle alone was applied to 2.5 or 4 cm2 areas on contralateral £anks of asebia J1 or 2 J and wild-type Asebia skin displays normal permeability barrier function control mice (n ¼ 5 or 6 animals each) twice daily for 3 d; (iii) 0.05%, but abnormal SC hydration As described previously (Gates 0.1%, 1.0%, 2.5%, or 10% aqueous glycerol or water alone was applied 2 and Karasek, 1965; Josefowicz and Hardy, 1978), asebia J1 mice twice daily to 4 cm areas on the £anks of asebia J1 or 2 J mice (n ¼ 4 displayed marked sebaceous gland hypoplasia, hyperkeratosis, animals in each glycerol group) for 3 d. Finally, SC hydration was compared after 10% glycerol versus 10% urea applications to contralateral epidermal hyperplasia with a normal-appearing granular layer £anks of asebia 2 J animals (n ¼ 4) each twice daily for 4 d, as above. In by light microscopy, and increased numbers of mast cells in the all studies, SC hydration was measured daily by corneometry immediately dermis (Fig 1A vs 1B). Basal TEWL levels were comparable in before the next topical application (E12 h after the last prior topical mutant versus control animals (data not shown). Following acute application). At the end of all studies, tape strippings or biopsies were barrier abrogation by cellophane tape stripping, the kinetics of obtained for light microscopy, lipid analysis, and/or lipase cytochemistry. barrier recovery were similar in asebia J1 versus control mice at 3 and6h(Fig 3A). Extracellular lamellar membrane structures, a Morphologic studies Biopsies were obtained from asebia J1 versus further parameter of permeability barrier status, also appeared control mice under basal conditions from skin sites that had been shaved completely normal in RuO4 post¢xed asebia J1 SC (Fig 2). 24 h earlier. Skin pieces were minced into pieces of 0.5 mm3, and processed Likewise, there were no di¡erences in the lamellar body for light and electron microscopy. Electron microscopy samples were secretory system in asebia J1 versus control mice (not shown). pre¢xed in half-strength Karnovsky’s ¢xative, post¢xed in either osmium In contrast, asebia J1 mice displayed E50% reduction in tetroxide (OsO4) or ruthenium tetroxide (RuO4), and embedded in an SC hydration assessed by electrical capacitance (p 0.0001) epoxy resin (Hou et al, 1991). To assess potential sites where TG to glycerol o hydrolysis could occur, we performed ultrastructural cytochemistry for (Fig 3B). Together, these studies demonstrate normal epidermal acid lipase, as described previously (Rassner et al, 1997). Ultrathin sections barrier function and lamellar membranes in asebia J mice, were viewed in a Zeiss 10 A electron microscope, operated at 60 kV, despite the presence of sebaceous gland hypoplasia, evidence of after further contrasting with lead citrate and uranyl acetate. One-half epidermal hyperplasia, and mast cell proliferation and profound micron sections of epoxy-embedded samples were stained with 1% abnormalities in SC hydration. 730 FLUHR ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure1. Asebia J1 skin displays hyperkeratosis, epidermal hyperplasia, sebaceous gland hypoplasia, and mast cell activation. Sebaceous glands (seb) are seen adjacent to follicles (large arrows) in wild-type (W) controls (B), but not in asebia J1 (As) skin (A). Asebia epidermis displays not only a thickened SC, but also acanthosis (increased numbers of nucleated cell layers; B, small arrows). Asebia J skin also displays increased numbers of mast cells, with various degrees of degranulation (A, arrowheads). One-half micron, plastic-embedded sections, toluidine-blue stained. Bar:200mm.

major nonpolar lipid species, namely, wax diesters, wax esters, sterol esters, was decreased signi¢cantly (Fig 4A, B), consistent with the absence of sebaceous glands (Nikkari, 1974). FFA levels also appeared to be reduced in asebia SC, but the di¡erence did not achieve statistical signi¢cance (FFA would continue to be generated in asebia epidermis from catabolism of epidermal phospholipids). Yet, substantial sterol/wax esters and some TGs were present not only in control but also in the SC of asebia J1 mice (Fig 4B). Finally, there was also a slight but statistically signi¢cant increase in free sterols (Fig 4B), as noted in previous studies (Wilkinson and Karasek, 1966; Arndt, 1996; Sundberg et al, 2000). These studies demonstrate a signi¢cant depletion, but not an elimination, of nonpolar lipids of presumed sebaceous gland origin in the SC of asebia J1 mice.

Glycerol is the sebum constituent that speci¢cally increases hydration of asebia sc The abnormality in hydration in asebia J1 SC suggests that sebaceous-gland-derived products could in£uence this process. Hence, we next applied a mixture of synthetic, sebum-like lipids, i.e., a wax monoester (stearyl stearate), a sterol ester (cholesterol palmitate), and a TG (tripalmitate), to the skin of asebia J1 and age/sex-matched wild-type control mice. Despite repeated applications, this mixture did not correct the hydration abnormality in asebia Figure 2. Asebia J1 SC reveals normal extracellular membrane J1 mice (Tab l e I ). structures. Comparable SC extracellular lamellar membranes and desmo- The inability of topical sebaceous lipids to correct the somes are seen in the SC of asebia J1 (A) and control (N) mice (arrows). hydration abnormality suggested that glycerol, a well-known RuO -post¢xed. Bar:0.25mm. humectant, generated as a result of the high rates of TG 4 synthesis and hydrolysis that occur within the pilosebaceous follicle (Nicolaides and Wells, 1957; Thody and Shuster, 1989; Nonpolar lipids are depleted, but not absent, in asebia SC Stewart and Downing, 1991), could be the responsible hydrating We next assessed the lipid biochemical basis for the abnormal component. Indeed, endogenous glycerol levels were profoundly SC hydration in asebia J mice. Whereas the total lipid content reduced in the SC of asebia J1 mice (control, 0.823370.1991 nmol was not altered in asebia J versus control SC, the content of the glycerol per mg SC protein, versus asebia, 0.128470.053 nmol VOL. 120, NO. 5 MAY 2003 ABNORMAL STRATUM CORNEUM HYDRATION IN ASEBIA MICE 731

that the abnormal hydration in asebia J skin is largely reversed by topical glycerol, but not by sebaceous lipids or another potent endogenous humectant, urea.

Absent sebaceous-gland-associated lipase accounts for the asebia hydration abnormality We next determined whether reduced generation of glycerol from TG within the pilosebaceous follicle accounts for the hydration abnormality. To evaluate potential sites of glycerol generation in the skin, we assessed lipase activity in the SC and pilosebaceous follicles of asebia J1 versus control mice by ultrastructural cytochemistry (Menon et al, 1986; Rassner et al, 1997). Control mice revealed abundant lipase activity (i) in sebaceous glands; (ii) in the outer SC; (iii) around nascent hair shafts; and (iv) within the cytosol of the follicular epithelium (Fig 6A^C). Asebia J1 skin revealed similar, abundant levels of enzyme activity in follicles (Fig 7A, B) and in the outer SC (Fig 7C, D), but with the absence of sebaceous glands no lipase activity could be demonstrated in such structures. As these results demonstrate comparable lipase activity in both the follicles and SC of asebia versus control mice, the defect in glycerol generation in asebia mice can be attributed to an absence of sebaceous-gland-associated lipase activity. As SC contains abundant lipase activity, glycerol could be generated, in theory, by hydrolysis of either surface-deposited (from sebaceous glands) or epidermal-derived TG. Hence, to assess further whether the hydrating fraction of glycerol derives speci¢cally from sebaceous-gland-derived TG, we next applied increasing concentrations of tripalmitin versus vehicle to asebia 2 J SC. SC hydration did not change after repeated applications of TG, at concentrations up to 10% twice daily for up to 5 d (data not shown). Together, these results demonstrate (i) the requirement for sebaceous-gland-associated lipases in the generation of the hydrating fraction of glycerol in normal skin; and (ii) that glycerol generation occurs primarily within the pilosebaceous follicle, rather than at the skin surface.

Figure 3. Barrier recovery is normal, but asebia J1 mice display DISCUSSION abnormal SC hydration. Panel A: Permeability barrier was disrupted by sequential tape stripping in asebia J1 (n ¼ 6) versus wild-type (n ¼ 4) mice Numerous speculations have been advanced about the functions (three to four sites were assessed on each animal, yielding a total n of of cutaneous sebaceous glands and their holocrine product, X 2 10^12). TEWL rates were raised to 4 mg per cm per h, and then percent sebum (Nikkari, 1974; Thody and Shuster, 1989). In nonhuman barrier recovery was compared 3 and 6 h after acute disruption. The di¡er- vertebrates, there is evidence (i) that sebaceous glands deposit vi- ences were not statistically signi¢cant at either time point. Panel B:Hydra- tamins D and E on the skin surface, and (b) that sebum possesses tion of SC was measured by electrical capacitance with a corneometer, 7 antimicrobial, olfactory/pheromone, thermoregulatory, and other and expressed in arbitrary units SEM. The di¡erences between asebia J1 functions. Moreover, the functional importance of modi¢ed se- (n ¼12) and control (n ¼ 8) mice are highly signi¢cant (po0.0001). baceous glands, such as the Meibomian glands, is also generally acknowledged (Ti¡any, 1985; Bron and Ti¡any, 1998; Mathers and Lane, 1998). Yet, sebum is commonly believed to subserve glycerol per mgSCprotein;p¼ 0.0141). Based upon the no positive functions in the epidermis of humans (Kligman, demonstrated de¢ciency of this humectant in asebia SC, we next 1963; Thody and Shuster, 1989; Webster, 1999). Although the link assessed whether glycerol applications alone could correct the between the genetic mutation and sebaceous gland hypoplasia is asebia hydration abnormality. Four days of twice daily 10% not known (see below), asebia mice o¡er unique models to ad- glycerol applications completely reversed the hydration abnorm- dress unresolved issues about the potential functions of sebaceous alities in asebia J1 SC, and signi¢cantly, but incompletely, reversed glands and their products. For example, the association of scar- the abnormality in asebia 2 J animals (Tab l e I I ). Lower con- ring alopecia with sebaceous gland hypoplasia in asebia J1 and 2 centrations (i.e., from 0.1% to 5% glycerol) signi¢cantly J mice led to the recent hypothesis that sebaceous gland products increased SC hydration in asebia 2 J animals, but only concen- are critical for the dissociation of the hair shaft from the follicular trations above 1% exceeded the impact of the aqueous vehicle epithelium (Stenn et al, 1999; Sundberg et al,2000). alone (Tab l e I I ). In contrast, glycerol concentrations above 10% Using asebia mice we examined here two other potential func- did not increase SC hydration further (not shown). The water tions of sebum, i.e., in epidermal permeability barrier homeosta- vehicle alone slightly, but signi¢cantly, improved hydration in sis and in SC hydration. Sebum is not thought to be important both asebia J1 and 2 J mice (Tab l e I I ). In contrast, topical for the permeability barrier, because both prepubertal children glycerol slightly but signi¢cantly reduced, rather than increased, (Thody and Shuster, 1989) and the extrafacial skin of the aged SC hydration in normal wild-type mice (by E10%; po0.01). (Ghadially et al, 1995) exhibit normal barrier function under basal Finally, to ascertain whether the requirement for glycerol is conditions. Moreover, repeated extraction of skin surface lipids speci¢c or replaceable by another potent polar, endogenous, but with a nonpolar solvent depletes both the SC and subjacent se- chemically unrelated humectant, we compared SC hydration baceous glands of all stainable, nonpolar lipids, but produces only after applications of 10% glycerol versus 10% urea to contralateral a minor barrier abnormality (Grubauer et al, 1989). Furthermore, £anks on asebia 2 J mice. As seen in Fig 5, only glycerol when sebaceous glands are involuted pharmacologically, with su- improved SC hydration in the asebia mice. These results show praphysiologic doses of isotretinoin (Gomez, 1981; Geiger, 1995), 732 FLUHR ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 4. Asebia J1 SC displays a signi¢cant reduction in nonpolar lipid fractions. Panel A: Full-thickness mouse skin was excised from shaved asebia J1 (n ¼ 4) and wild-type (n ¼ 6) mice. The skin samples were incubated with EDTA, followed by trypsin, as described in Methods. Isolated nonpolar lipids were then fractionated, charred, and quantitated as described in Methods. Equal amounts of lipids were applied to each lane. Panel B: Isolation, fractio- nation, and quantitation of SC lipids were the same as described for Fig 5. Each species is presented as mean (mg per mg SC)7SEM. Whereas controls (n ¼ 6) are represented by open bars, the results for asebia J mice (n ¼ 4) are indicated by the shaded bars. The di¡erences for wax diesters, wax monoesters, and sterol esters are highly signi¢cant. both barrier function and SC lamellar membranes remain unal- Table I. E¡ects of a sebaceous lipid-like mixture on SC tered (Elias et al, 1981). Yet, true functional status is best assessed 7 using a dynamic measure, e.g., the kinetics of barrier recovery hydration (arbitrary units SEM) after a standardized, acute insult (Feingold, 1999). Using such Animal Pretreatment 1 d 3 d measures, important functional abnormalities have been un- earthed and quantitated in aged, photoaged, ichthyotic, testoster- Control (n ¼ 6) 49.872.3 52.674.7 57.874.5 one-replete, and psychologically stressed skin (Ghadially et al, Asebia J1 (n ¼ 6) 29.872.8 24.571.6 26.571.6 1995; Reed et al, 1995; Zettersten et al, 1998; Denda et al,2000; To assess the potential role of sebaceous gland lipids in SC hydration, a lipid Kao et al, 2001).We showed here, however, that permeability bar- mixture (40 ml) containing TG (250 mg), stearyl stearate (125 mg), and cholesterol rier homeostasis is normal, i.e., neither accelerated nor delayed, in palmitate (25 mg) in a 40 ml propylene glycol:ethanol (7:3 vol) vehicle versus vehicle the presence of sebaceous gland hypoplasia in asebia J1 mice. The alone was applied twice daily to 4 cm2 areas on opposite £anks of asebia J and persistence of normal barrier function in asebia J1 mice can be control mice (n ¼ 6 each) for 3 d. During these studies, SC hydration was measured daily by corneometry immediately before and 12 h after the last prior appli- explained by the unaltered levels of the three key barrier lipids, cation. There were no signi¢cant changes in hydration with lipid applications to ceramides, free sterols, and FFAs, as well as the persistence of nor- either asebia or control mice. The di¡erences between asebia and control mice re- mal SC extracellular membranes in asebia J1 SC, as shown here. mained signi¢cant at all time points (po0.001). VOL. 120, NO. 5 MAY 2003 ABNORMAL STRATUM CORNEUM HYDRATION IN ASEBIA MICE 733

Table II. E¡ects of glycerol on SC hydration in asebia J1 and 2 J versus control mice (arbitrary units7SEM) Groups Pretreatment Post-treatment Signi¢cance versus baseline

Experiment 1 (asebia J1 mice) Control þ vehicle (water) (n ¼ 8) 46.070.7 56.871.9 po0.0001

Control þ10% glycerol ^ 50.971.5 po0.01 Asebia þvehicle (water) 24.171.4 31.472.3 po0.01 Asebia þ10% glycerol (n ¼ 8) ^ 52.071.3 po0.0001

Experiment 2 (asebia 2 J mice) Control þ vehicle (water) (n ¼16) 48.771.1 46.472.2 NS

Control þ1% glycerol ^ 45.671.5 NS Control þ 5% glycerol ^ 47.171.6 NS Control þ10% glycerol ^ 48.471.8 NS Asebia þvehicle (water) (n ¼16) 15.271.1 19.371.3 po0.05 Asebia þ1% glycerol ^ 21.372.5 po0.05 Asebia þ 5% glycerol ^ 25.871.9 po0.001 Asebia þ10% glycerol ^ 31.171.6 po0.0001 10% aqueous glycerol or water alone was applied twice daily to 4 cm2 areas on the £anks of mice for 4 d, followed by measurement of SC hydration (see Methods) (n ¼16 sites each on four mice or as shown). Experiments 1 and 2 were performed in asebia J1 and 2 J mice respectively versus wild-type littermates.

and Rosen¢eld, 2000). Indeed, fatty acid metabolites also activate PPARa and PPARd, and some endogenous fatty acids are pro- miscuous activators of more than one PPAR. Hence, it is possible that a more profound depletion of functional Scd1, with reduced generation of one or more endogenous PPAR activators, could account for the abnormality in barrier function in asebia 2 J mice. Second, SCD1 is necessary for the normal formation of TGs and secretion of VLDL from the liver (Attie et al,2002).In the absence of SCD1 (transgenic knockout), not only are sebaceous glands involuted (Miyazaki et al, 2001), but a substantial portion of the FFA pool, destined for TG production, is instead oxidized (Ntambi et al, 2002). Thus, more profound FFA depletion could occur in the asebia 2 J animals, accounting for their relatively severe phenotype. In contrast to barrier function, our results demonstrate that sebum plays an important and speci¢c role in SC hydration. In fact, the hydration of the SC of both asebia J1 and asebia 2 J mice was Figure 5. The hydration abnormality in asebia SC is not corrected E by another humectant. Contralateral £anks of asebia 2 J mice were trea- markedly reduced (by 50% and 70%, respectively) in compar- ted with 10% glycerol or 10% urea versus water alone, as described in Meth- ison to both heterozygote and wild-type mice. The residual hy- n ods (n ¼ 5^6 sites each). Results are expressed as mean7SEM. po0.05; dration of the SC in both asebia strains could be due to either nn po0.001. the known ability of more polar SC lipids (e.g., ceramides) to in£uence SC hydration (Imokawa and Hattori, 1985), and/or the persistence of ¢laggrin breakdown products with potent humec- Together, these results show that sebum plays neither an essential tant properties (Scott and Harding, 1986). Nevertheless, these re- nor a destabilizing role in permeability barrier homeostasis. sults show that an intact intercellular membrane bilayer system, Interestingly, the closely related asebia 2 J (ab2J/ab2J) mutant although su⁄cient for permeability barrier homeostasis, does mouse displays a permeability barrier abnormality (Zheng et al, not always su⁄ce for normal SC hydration. 1999; Sundberg et al, 2000). Both the asebia J1 and asebia 2 J phe- Further experiments attempted to shed light on the relation- notypes result from mutations in the gene that encodes the en- ship between the sebaceous gland hypoplasia and abnormal SC zyme Scd1 (Zheng et al, 1999; Sundberg et al, 2000), the only hydration.We showed that asebia J1 SC displays a dramatic reduc- isoform that is expressed in sebaceous glands. Why the Scd1 ab- tion in the classic sebaceous lipids (wax mono/diesters and sterol normality provokes a barrier abnormality in asebia 2 J mice, esters), as described previously (Wilkinson and Karasek, 1966; whereas a related mutation in asebia J animals does not, is unclear. Sundberg et al, 2000). Yet, asebia SC still displays low but signi¢- Two possibilities should be considered. First, Scd1 catalyzes the cant levels of classic sebaceous lipids, i.e., wax diesters, wax formation of unsaturated 18 and 16 carbon fatty acids, such as monoesters, cholesterol esters, and TGs, even with profound se- oleic acid (C18:1) and palmitoleic acid (C16:1) from stearic acid baceous gland hypoplasia. Our observation suggests that epider- (C18:0) and palmitic acid (C16:0), respectively. Oleic and palmito- mis retains the capacity to form these lipids, con¢rming a recent leic acids, as well as other unsaturated FFAs, are endogenous acti- report of wax ester production by cultured keratinocytes.1 vators of the peroxisomal proliferator activator (PPAR) family of Although lipid staining of frozen sections has shown that sebum nuclear hormone receptors. PPARa regulates lipid metabolism in forms a ¢lm of neutral lipids on the surface of normal SC (‘‘skin extracutaneous tissues, but it is also a potent regulator of epider- surface lipid’’) (Kligman and Shelley, 1958; Grubauer et al,1989; mal homeostasis. Speci¢cally, PPARa regulates cutaneous epider- Sheu et al, 1999), normal hydration is not restored when asebia J mal proliferation, di¡erentiation, and permeability barrier SC is replenished with a mixture of sebum-like lipids alone. homeostasis (Hanley et al, 1997; 1999), whereas another PPAR Although ideally this mixture would also have included wax die- (i.e. PPARa), regulates sebaceous gland di¡erentiation (Deplewski sters, these are not commercially available. Hence, our results do 734 FLUHR ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 6. Abundant lipase activity is present in sebaceous glands and the outer SC of wild-type skin. (A) Normal sebaceous glands display abundant lipase activity (arrows) both around and in association with lipid droplets (LD). (B) Normal skin incubated for lipase with the added speci¢c inhibitor tetrahydrolipstatin (thls). (C) A band of intense lipase activity (arrows) is present in the outer SC (OSC) of wild-type animals. Low levels of activity also are present in the underlying lower SC (LSC) and stratum granulosum (SG). Bars:0.75mm. not exclude a role for other lipid species, also present in sebum absence of AQP3 in the basal layer of the epidermis in association (e.g., wax diesters), as contributors to SC hydration. with abnormal SC hydration and decreased SC glycerol levels. It Because the sebum-like lipid mixture did not increase hydra- should be noted, however, that whereas these studies suggest that tion, we hypothesized that its ine¡ectiveness might instead re£ect glycerol of blood or epidermal origin could contribute to SC hy- a requirement for glycerol, a well-known humectant and a poten- dration, AQP3 expression in sebaceous glands was not assessed. tial catabolic product of sebaceous-gland-derived TG (Nicolaides Therefore, the absence of sebaceous gland AQP3 could, at least and Wells, 1957; Freinkel and Shen, 1969; Marples et al, 1971). The in theory, contribute to the hydration abnormality in AQP3- loss of epidermal TG during terminal di¡erentiation implies that knockout animals. Our studies nevertheless demonstrate the im- most TGs are utilized as an energy source within the epidermis portance of the sebaceous gland as the source of a critical hydrat- (Nicolaides, 1974); therefore TGs do not persist into the SC, and ing fraction of surface-deposited glycerol, which after absorption therefore they are not normally available as a source of glycerol into the SC interstices, regulates SC hydration. for SC hydration. Although very little TG of epidermal origin is The known high rates of TG turnover in sebaceous glands available to serve as an endogenous substrate, it is possible that (Nicolaides and Wells, 1957; Thody and Shuster, 1989; Stewart there are sources of glycerol in SC other than those derived from and Downing, 1991), and low levels of TG in normal SC (Schurer sebaceous glands. Indeed, phospholipid catabolism, which gener- and Elias, 1991), are consistent with the hypothesis that most se- ates a family of nonessential FFAs required for the barrier (Mao- baceous-gland-derived TGs are hydrolyzed to glycerol prior to Qiang et al, 1995; 1996), could simultaneously generate glycerol delivery to the skin surface. Indeed, we showed that the glycerol that would be available in the SC interstices. In addition, a poten- content of asebia SC is markedly reduced, and that topical glycer- tial role for either blood- or epidermal-derived glycerol in SC hy- ol, but not TG-containing sebaceous lipids, restores normal hydration (after prior phospholipid synthesis/degradation) is further dration to asebia J1 SC. Similar treatment largely, but suggested by recent work on aquaporin 3 (AQP3) knockout ani- incompletely, reversed the hydration abnormality in 2 J mice. mals (Ma et al, 2002; Hara et al, 2002). These animals display an Whether prolongation of glycerol treatment would have further VOL. 120, NO. 5 MAY 2003 ABNORMAL STRATUM CORNEUM HYDRATION IN ASEBIA MICE 735

Figure 7. Asebia SC and follicles reveal normal lipase activity. Lipase activity is present around corneocytes in nascent hair shafts (A, arrows), and at the interface of the follicular epithelium with the nascent hair shaft in asebia J1 skin (B, arrows). Lipase activity also is present focally in the mid to outer layers of asebia SC (D, closed arrows), and at the SG^SC interface (D, open arrows). (C) Asebia tissue sample, incubated for lipase with added speci¢c enzyme inhibitor tetrahydrolipstatin (thls). H, hair shaft. Bars:0.75mm. normalized hydration in 2 J mice was not assessed. Alternatively, the hydration abnormality in asebia J mice. (ii) The available pool some applied glycerol could be absorbed across 2 J SC and di- of epidermal TG is utilized in the nucleated layers below the SC, verted to other pathways, because of its associated barrier as indicated by the low levels of TG in normal SC (these studies abnormality (see above). It is also noteworthy that high concen- and Schurer and Elias, 1991). Moreover, topical TG, even at high trations of topical TG did not improve the hydration of asebia J concentrations, did not restore hydration, suggesting that TG per- SC, despite the presence of abundant lipase activity of presumed sistence into the SC could not hydrate SC normally. Together, epidermal origin in the SC. Instead, we provide here evidence these results are consistent with the hypothesis that the absence for the origin and importance of glycerol, derived from of TG hydrolysis to glycerol by endogenous sebaceous gland the lipolysis of sebaceous-gland-derived TG, for normal SC lipases accounts for the SC hydration abnormality in asebia J1 hydration. Although lipase activity could be demonstrated cyto- mice. chemically in asebia SC and hair follicles, abundant sebaceous- Our studies also demonstrate a speci¢c requirement for glycer- gland-associated lipase activity was present only in control, but ol in SC hydration, because topical urea, another potent endo- not in asebia, skin (due to the absence of sebaceous glands). The genous humectant, did not restore SC hydration to asebia SC. importance of sebaceous glands as the source of glycerol can be These results should not be interpreted as indicating the ine¡ec- explained as follows. (i) Lipolysis reportedly proceeds more e⁄- tiveness of urea as a moisturizer. Rather, they demonstrate that ciently within the pilosebaceous apparatus (Nicolaides and Wells, urea cannot substitute for the hydrating capacity of sebum-de- 1957; Kligman, 1963; Freinkel and Shen, 1969) than within the SC, rived glycerol. Why glycerol exhibits such unique e⁄cacy in re- thereby also accounting for the inability of topical TG to correct storing SC hydration, however, remains unknown. Glycerol is 736 FLUHR ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY not only a potent humectant, but topical glycerol also accelerates Bron AJ,Ti¡any JM:The Meibomian glands and tear ¢lm lipids. Structure, function, barrier recovery after acute external perturbations (Fluhr et al, and control. Adv Exp Med Biol 438:281^295, 1998 Brown WR, Hardy MH: A hypothesis on the cause of chronic epidermal hyperpro- 1999). Moreover, at least in theory, if glycerol is absorbed into liferation in asebia mice. Clin Exp Dermatol 13:74^77, 1988 the nucleated layers, it could serve as either a building block for Denda M, Wood LC, Emami S, Calhoun C, Brown BE, Elias PM, Feingold KR: the assembly of complex glycerolipids, and/or a potential sub- The epidermal hyperplasia associated with repeated barrier disruption by ace- strate for ATP production. Whether topical glycerol exhibits any tone treatment or tape stripping cannot be attributed to increased water loss. of these additional properties, or whether still other characteris- Arch Dermatol Res 288:230^238, 1996 tics account for its ability to modulate SC hydration, remains Denda M, Sato J,Tsuchiya T, Elias PM, Feingold KR: Low humidity stimulates epidermal DNA synthesis and ampli¢es the hyperproliferative response to barrier unknown. disruption: Implication for seasonal exacerbations of in£ammatory dermatoses. The reduced SC hydration in asebia J mice is associated not J Invest Dermatol 111:873^878, 1998a only with abnormal hydration, but also with excess scale produc- Denda M, Sato J, MasudaY, et al: Exposure to a dry environment enhances epidermal tion, epidermal hyperplasia, and microscopic evidence of in£am- permeability barrier function. J Invest Dermatol 111:858^863, 1998b mation (mast cell hyperplasia) (see also Gates and Karasek, 1965; Denda M, Tsuchiya T, Elias PM, Feingold KR: Stress alters cutaneous permeability Brown and Hardy,1988). Although it has been proposed that the barrier homeostasis. Am J Physiol Regul Integr Comp Physiol 278:367^R372, 2000 Deplewski D, Rosen¢eld RL: Role of hormones in pilosebaceous unit development. in£ammation in asebia mice drives the epidermal hyperplasia, Endocr Rev 21:363^392, 2000 leading to excess scale (Brown and Hardy, 1988), recent studies Elias PM, Fritsch PO, Lampe M,Williams ML, Brown BE, Nemanic M, Grayson S: favor an alternative pathogenic sequence; i.e., sustained exposure Retinoid e¡ects on epidermal structure, di¡erentiation, and permeability. Lab of normal skin to a dry external environment, although improv- Invest 44:531^540, 1981 ing permeability barrier homeostasis (Denda et al, 1998b), never- Feingold KR, Elias PM: The environmental interface: Regulation of permeability theless leads to decreased SC hydration, as well as epidermal barrier homeostasis. In: Loden M, Maibach HI, eds. Dry Skin and Moisturizers: Chemistry and Function. Boca Raton, FL: CRC Press, 1999; p 45^58 hyperplasia, in£ammation, and excess scale (Denda et al, 1998a). Fluhr JW, Gloor M, Lehmann L, Lazzerini S, Distante F, Berardesca E: Glycerol Moreover, restoration of normal SC hydration by a variety of accelerates recovery of barrier function in vivo. Acta DermVenereol 79:418^421,1999 maneuvers, including topical glycerol, normalizes epidermal pro- Fluhr JW, Kao J, Jain M, Ahn SK, Feingold KR, Elias PM: Generation of free fatty liferation and reduces mast cell hypertrophy (a marker of in£am- acids from phospholipids regulates stratum corneum acidi¢cation and integ- mation) in the face of ongoing xeric stress (Denda et al, 1998a). rity. JInvestDermatol117:44^51, 2001 Thus, it is likely that decreased SC hydration contributes to the Freedberg IM: Fitzpatrick’s Dermatology in General Medicine. New York: McGraw-Hill, 1997 scaling phenotype in asebia J mice. It is important to note that Freinkel RK, Shen Y:The origin of free fatty acids in sebum. II. Assay of the lipases barrier function, and even barrier recovery, remain normal in of the cutaneous bacteria and e¡ects of pH. JInvestDermatol53:422^427, 1969 the face of epidermal hyperplasia, con¢rming prior studies that Gates AH, Karasek M: Hereditary absence of sebaceous glands in the mouse. Science epidermal hyperplasia per se does not necessarily provoke a barrier 14 8:1471^1473, 1965 abnormality (Denda et al, 1996; 1998a; 1998b). Yet, the signal se- Geiger JM: Retinoids and sebaceous gland activity. Dermatology 191:305^310, 1995 quence that is responsible for the low hydration-induced epider- Ghadially R, Brown BE, Sequeira-Martin SM, Feingold KR, Elias PM: The aged epidermal permeability barrier. Structural, functional, and lipid biochemical mal hyperplasia has not been identi¢ed. abnormalities in humans and a senescent murine model. JClinInvest95:2281^ In the light of these observations, decreased SC hydration, even 2290, 1995 in the face of normal barrier function, could have important clin- Gloor M, Fluhr J, Lehmann L, Gehring W,Thiero¡-Ekerdt R: Do urea/ammonium ical consequences. Indeed, decreased hydration might be expected lactate combinations achieve better skin protection and hydration than either in any situation where sebaceous glands are absent or involuted. component alone? Skin Parmacol Appl Skin Physiol 15:35^43, 2002 For example, prepubertal children (42yand 9 y) commonly Gomez EC: Di¡erential e¡ect of 13-cis-retinoic acid and an aromatic retinoid o (Ro 10^9359) on the sebaceous glands of the hamster £ank organ. JInvestDer- display eczematous patches (pityriasis alba) on the face and trunk matol 76:68^69, 1981 that disappear coincidentally with the onset of sebaceous gland Grubauer G, Feingold KR, Harris RM, Elias PM: Lipid content and lipid type activation. Moreover, the distal extremities of aged skin clinically as determinants of the epidermal permeability barrier. JLipidRes30:89^96, often displays moderate to severe xerosis. Furthermore, the xero- 1989 sis in patients receiving systemic isotretinoin, with attendant se- Hanley K, Jiang Y, Crumrine D, et al: Activators of the nuclear hormone receptors baceous gland involution (Landthaler et al, 1980; Geiger, 1995), is PPARa and FXR accelerate the development of the fetal epidermal permeability barrier. JClinInvest100:705^712, 1997 most-accentuated on the face, back, and chest, all sebaceous- Hanley K, Komuves LG, Bass NM, et al: Fetal epidermal di¡erentiation and barrier gland-enriched sites (Peck et al, 1979). Further, atopic dermatitis, development in vivo is accelerated by nuclear hormone receptor activators. nummular eczema, as well as a variety of other eczematous pro- J Invest Dermatol 113:788^795, 1999 cesses, show a predilection for the extremities (Arndt, 1996; Hara M, Ma T, Verkman AS: Selectively reduced glycerol in skin of aquaporin-3 Freedberg, 1997), which are sebaceous-gland-depleted. These clin- de¢cient mice may account for impaired skin hydration, elasticity and barrier ical observations create a compelling therapeutic rationale for op- recovery. JBiolChemM2:09003, 2002 Hou SY, Mitra AK, White SH, Menon GK, Ghadially R, Elias PM: Membrane timizing SC hydration as a part of the therapy and prevention of structures in normal and essential fatty acid-de¢cient stratum corneum: Char- eczematous skin conditions. Finally, asebia mutant mice could acterization by ruthenium tetroxide staining and X-ray di¡raction. JInvestDer- provide a useful model to test the e⁄cacy of agents that are in- matol 96:215^223, 1991 tended speci¢cally as SC hydrating/moisturizing agents, and to Imokawa G, Hattori M: A possible function of structural lipids in the water-holding further dissect the role of sebaceous glands in normal skin biology. properties of the stratum corneum. J Invest Dermatol 84:282^284, 1985 Josefowicz WJ, Hardy MH: The expression of the gene asebia in the laboratory mouse. I. Epidermis and dermis. Genet Res 31:53^65, 1978 Kao JS, Garg A, Mao-Qiang M, Crumrine D, Ghadially R, Feingold KR, Elias PM: Sue Allen, Jerelyn Magnusson, and Jennifer C. Quevedo capably prepared the manu- Testosterone perturbs epidermal permeability barrier homeostasis. JInvestDer- script. These studies were supported by NIH grants AR 19098, AR 39448 (PP), matol 116:443^451, 2001 RR00173, and the Medical Research Service, Department of Veterans A¡airs. Kligman AM:The use of sebum. In Advances in Biology of the Skin: the Sebacous Glands. 1963 Kligman AM, Shelley WB: An investigation of the biology of the human sebaceous gland. J Invest Dermatol 30:99^125, 1958 Landthaler M, Kummermehr J, Wagner A, Plewig G: Inhibitory e¡ects of 13-cis- REFERENCES retinoic acid on human sebaceous glands. Arch Dermatol Res 269:297^309, 1980 Arndt KA: Cutaneous Medicine and Surgery: An Integrated Program In Dermatology. Phila- Law S, Wertz PW, Swartzendruber DC, Squier CA: Regional variation in content, delphia, PA: W.B. Saunders, 1996 composition and organization of porcine epithelial barrier lipids revealed by Attie AD, Kraus RM, Gray-Keller MP, et al: Relationship between stearoyl-CoA de- thin-layer chromatography and transmission electron microscopy. Arch Oral saturase activity and plasma triglycerides in human and mouse hypertriglycer- Biol 40:1085^1091, 1995 idemia. JLipidRes43:1899^1907, 2002 Ma T, Hara M, Sougrat R, Verbavatz JM, Verkman AS: Impaired stratum corneum Bligh EG, Dyer WJ: A rapid method of total lipid extraction and puri¢cation. Can J hydration in mice lacking epidermal water channel aquaporin-3. JBiolChem Biochem Physiol 37:911^917, 1959 277:17147^153, 2002 VOL. 120, NO. 5 MAY 2003 ABNORMAL STRATUM CORNEUM HYDRATION IN ASEBIA MICE 737

Mao-Qiang M, Feingold KR, Jain M, Elias PM: Extracellular processing of Scott IR, Harding CR: Filaggrin breakdown to water binding compounds during phospholipids is required for permeability barrier homeostasis. JLipidRes36: development of the rat stratum corneum is controlled by the water activity of 1925^1935, 1995 the environment. Dev Biol 115:84^92, 1986 Mao-Qiang M, Jain M, Feingold KR, Elias PM: Secretory phospholipase A2 Sheu HM, Chao SC, Wong TW,Yu-Yun Lee J, Tsai JC: Human skin surface lipid activity is required for permeability barrier homeostasis. JInvestDermatol106: ¢lm: An ultrastructural study and interaction with corneocytes and intercellu- 404^411, 1996 lar lipid lamellae of the stratum corneum. Br J Dermatol 140:385^391, 1999 Marples RR, Downing DT, Kligman AM: Control of free fatty acids in human sur- Stenn KS, Sundberg JP, Sperling LC: Hair follicle biology, the sebaceous gland, and face lipids by Corynebacterium acnes. J Invest Dermatol 56:127^131, 1971 Mathers WD, Lane JA: Meibomian gland lipids, evaporation, and tear ¢lm stability. scarring alopecias. Arch Dermatol 135:973^974, 1999 Stewart ME, Downing DT: Chemistry and function of mammalian sebaceous lipids. Adv Exp Med Biol 438:349^360, 1998 Menon GK, Grayson S, Elias PM: Cytochemical and biochemical localization of li- Adv Lipid Res 24:263^301, 1991 pase and sphingomyelinase activity in mammalian epidermis. J Invest Dermatol Stewart ME, Grahek MO, Cambier LS, Wertz PW, Downing DT: Dilutional e¡ect 86:591^597, 1986 of increased sebaceous gland activity on the proportion of linoleic acid in Miyazaki M, ManWC, Ntambi JM:Targeted disruption of stearoyl-CoA desaturase 1 sebaceous wax esters and in epidermal acylceramides. JInvestDermatol87: gene in mice causes atrophy of sebaceous and Meibomian glands and depletion 733^736, 1986 of wax esters in the eyelid. J Nutr 131:2260^2268, 2001 Sundberg JP, King LE Jr: Mouse models for the study of human hair loss. Dermatol Newsholme EA: Glycerol. In: Bergmeyer HU, (ed). Methods of Enzymatic Analysis, Clin 14:619^632, 1996 2nd edn. San Francisco, CA: Academic Press, 1974:pp 1409^1414 Sundberg JP, Boggess D, Sundberg BA, Eilertsen K, Parimoo S, Filippi M, Stenn K: Nicolaides N: Skin lipids: Their biochemical uniqueness. Science 186:19^26, 1974 Asebia-2J (Scd1 (ab2J)): A new allele and a model for scarring alopecia. Am J Nicolaides N, Wells GC: On the biogenesis of the free fatty acids in human skin Pathol 156:2067^2075, 2000 surface lipids. J Invest Dermatol 29:423^433, 1957 Thody AJ, Shuster S: Control and function of sebaceous glands. Physiol Rev 69: Nikkari T: Comparative chemistry of sebum. J Invest Dermatol 62:257^267, 1974 383^416, 1989 Ntambi JM, Miyazaki M, Stoehr JP, et al: Loss of stearoyl-CoA desaturase-1 function Ti¡any JM: The role of Meibomian secretion in the tears. Tr a n s O p h t h a l m o l S o c U K protects mice against adiposity. Proc Natl Acad Sci USA 99:11482^11486, 2002 104:396^401, 1985 Peck GL, Olsen TG, Yoder FW, et al: Prolonged remissions of cystic and conglobate Webster GF: Acne vulgaris: State of the science. Arch Dermatol 135:1101^1102, 1999 acne with 13-cis-retinoic acid. NEnglJMed300:329^333, 1979 Wilkinson DI, Karasek MA: Skin lipids of a normal and mutant (asebic) mouse Rassner UA, Crumrine DA, Nau P, Elias PM: Microwave incubation improves lipolytic enzyme preservation for ultrastructural cytochemistry. Histochem J 29: strain. J Invest Dermatol 47:449^455, 1966 387^392, 1997 Zettersten E, Man MQ, Sato J, et al: Recessive x-linked ichthyosis. Role of cholester- Reed JT, Ghadially R, Elias PM: Skin type, but neither race nor gender, in£uence ol-sulfate accumulation in the barrier abnormality. J Invest Dermatol 111: epidermal permeability barrier function. Arch Dermatol 131:1134^1138, 1995 784^790, 1998 Schurer NY, Elias PM: The biochemistry and function of stratum corneum lipids. Zheng Y, Eilertsen KJ, Ge L, et al: Scd1 is expressed in sebaceous glands and is Adv Lipid Res 24:27^56, 1991 disrupted in the asebia mouse. Nat Genet 23:268^270, 1999