0022 -202X/85/8504 -0295$02.00/0 T HE JOU HNAL O F ! NVESTI (:ATI VE 0EHMATOLOGY, 85:295- 298, 1985 Vo l. 85, No.4 Copyri ght 1985 by The Williams & Wilkins Co. Printed in U.S.A. Lipid Composition and Acid Hydrolase Content of Lamellar Granules of Fetal Rat Epidermis*
RUTH K. FREINKEL, M.D. AND THOMAS N. TRACZYK, B.S. Department of Dermatology, Northwestern University M edical School, Chicago, Illinois, U.S.A.
Lipids and acid hydrolases have been characterized in t he intercellular space of stratum corneum, raising the possi a subcellular fraction, enriched with lamellar granules bility that hydrolytic enzymes may originate in LG which can (LG), derived from fetal rat epidermis. This fraction modify t he content of the intercellular space and contribute to contains 23% glycosyl ceramides and ceramides, 15% desquamation and permeability characteristics of the stratum free sterols, and 34% phospholipids. The lipid/protein corneum (9- 11]. ratio is 2.0. The sterols and sphingolipids were present We have previously reported on a subcellular fraction en in proportions similar to those previously reported in riched in LG, bearing acid phosphatase, isolated from fetal rat stratum corneum. These findings provide direct bio epidermis [12]. We have subsequently shown that this fract ion chemical evidence for the widely accepted hypothesis is enriched in glycosphingolipids and ceramides [13]. These that stratum corneum lipids are derived from exocytosis lipids are believed to play a major role in forming the broad of lamellar granules into the intercellular space. The LG intercellular lipid-containing sheets in stratum corneum and fraction was enriched in certain acid hydrolases includ their contribution to the water barrier has been discussed ing glucosidase, acid phosphatase, phospholipases A, and [3,14,15] . The LG fraction was also enriched in acylglucosyl sphingomyelinase; other acid hydrolases, i.e., amino ceramide, a major component of epidermal glycosphingolipids glycosidases, glactosidase and aryl sulfatase (pH 5.5), that has been speculatively implicated in the formation of the and steroid sulfatase were not preferentially localized stacked discs of LG [16,17]. A similar fraction has been more in this fraction. By modulation of phospholipids, glyco recently isolated from newborn mouse skin and also is enriched lipids, and proteins in the membrane regions of stratum in acid phosphatase and sphingolipids [18]. In the present corneum, the acid hydrolases of LG may play a role report we present the first extensive quantification of t he acid relevant to the function and desquamation of stratum hydrolases and the general lipid composition of LG. corneum. MATERIALS AND METHODS It has been inferred from considerable biochemical and cy Tissue Fractions tochemical evidence that t he lipids of stratum corneum largely Skin was obtained from fetal Sprague-Dawley rats at 20 days of reside in the intercellular and membrane regions of t his layer gestation . Epidermis was separated after incubation of whole skin in (1,2]. Their presence is attributed in large measure to t he 0.1 M dithiothreitol in minimum essential medium for 30 min at 37°C. extruded content of lamellar granules (LG) and is presumed to Homogenates of 1.0- 1.5 g epidermis in 0.25 M sucrose (100 mg/ml) were fractionated as previously described [1 2]. Briefly, a pellet was reflect the fusion of the individual lipid-rich lamellae into broad obtained after centrifugation of a 700 g supernatant at 17,000 g. We s heets [1-3]. The lipids of stratum corneum are characteristic have previously shown t hat LG are sedimented at 17,000 g (17K) and of t hat tissue and differ s ignificantly from the lipids of basal cannot be demonstrated in significant quantities in sediments obtained cells and spinous layer cells in their high con tent of sphingo from the 17K supernatant at greater centrifuga l force [1 2]. The 17K lipids and sterols and their low to absent content of phospho pell et thus served as the starting fraction for subsequent purification lipids (3,4]. This general pattern has been demonstrated for of LG . The pell et was resuspended in 0.5 ml 0.25 M sucrose and rodent, pig, and human epidermis [5]. fractionated on a co ntinuous gradient of metri za mide (analytical grade Although it is postulated that t he intercellular lipids of Nyegaard and Co, Oslo), 20- 50%. The 17K fraction and particulate stratum corneum originate in LG, t here have been no direct fractions from the gradient were washed with 0. 25 M sucrose and repell eted at 19,000 g for subsequent assays of enzy me activity and chemical examinations ofLG lipids. Histochemical studies have analysis of protein and lipid content. Morphology of the fractions was s uggested t hey contain sterols, phospholipids, and glycolipids monitored by electron microscopy as prev iously desc ribed [1 2]. (6-8]. More direct examination has been impossible because LG have not been isolated. Furthermore, LG have been classi Lipid Co ntent fied as lysosomes on t he basis of histochemically demonstrated Total li pids were extracted by the Folch method [19] and further acid phosphatase (and possibly other acid hydro lases) [9]. Their fractionated by thin-layer chromatography (TLC) on sili ca ge l 60 (E. extrusion into t he intracellular space appears to be linked to Merck, Darmstadt) using two 2-dimensional systems as fo llows: t he presence of acid phosphatase in this domain. Recent studies System I: Lipids were quantitatively applied to prescored plates and have suggested t hat other hydrolytic processes are ongoing in developed in ethylacetate:benzene (1:1) in the first dimension foll owed by chloroform:methanol :acetic acid:water (104:40:14:6) in the second dimension. This system resolved lysophosphoglycerides, sphingomye Manuscript received June 25, 1984; accepted for publi cation May 21, lin, phosphoglycerides, glycosphingo lipids, ce ramides, and free sterols. 1985. Other neutral lipids were only partially resolved and not further iden This work was supported by grant #AM164 41 of the National Insti tified. tutes of Health. System II: Lipids were fractionated on prescored plates using chl o *Prese nted in part at the Annual Meeting of The Society for roform:methanol:acetic acid:water (104:40:14:6) in the first dimension. Investigative Dermatology, Inc., Washington, D.C., May 6- 8, 1982. After rinsing with acetone, plates were developed in the second dimen Reprint requests to: Ruth K. Freinkel, M.D., Department of Der sion with chloroform:methanol:water:40% methylamine (65:35:5:5). matology, Northwestern Medical Schoo l, 303 E. Chi cago Avenue, Chi This system provided better discrimination of the individual phospho· cago, Illinois 60Gll. lipids found in System I. Abbreviations: Lipid spots were identified by cochromatography of authentic stand LG: lamella r granule(s) ards. Phospholipids, sphingo myelin, sterols, and ce ramides were ob PC: 1,2-di( [1- " CJoleoyl)phosphatidylcholine tained from Sigma. Glucosylceramide and acylglu cosylceramide derived TLC: thin-layer chromatography from pi g epidermis were kindly supplied by Dr. P. W. Wertz. Lipids
295 296 FREINKEL AND TRACZYK Vol. 85, No.4 we re a lso identified by t inctori al properties: chl orox-benzidine for containing small well-preserved mitochondria contaminated by ~ phin go lipid s, Bial's o rcinol r eagent fo r glycolipids, DragendorfCs re very few vesicles and a sm a ll amount of amorphous debris a nd age nt for choline, molybdenum blue for phosphorus-containing lipids denser fraction containing mitochondria, cornified cell enve and ninhydrin for ni trogen-co ntainin g- lipids ( reagents from Applied lopes, and debris (fraction III). Science Co.) For quantitation, lipid spots, identifi ed with iodine vapors, were Lipids sc raped from the plates. Total and individual lipids were quantitated by a modification of the method of Amenta [20 ] in which lipids are Lipids of each fraction were isolated by TLC and identified eluted from silica !(ei wi t h chl oroform methanol (1:1) at 45"C and then as specified in the Materials and Methods. The lipids of the LG reacted with dichromate-H.SO,. Standard curves with cholesterol, cer fraction are represented in Fig 1. Phospholipids, which were amides, ce rebrosides, and phospholipids were run simultaneously since further resolved in TLC system II, included phosphatidylcho each cl ass gives slightly different results. Total lipid conce ntration was line, phosphatidy lserine, phosphatidylinositol, phosphatidy le e~t ima te d aga inst a standard arbitrarily co mposed of 25 % phospholipid, thanolamine, sphingomyelin, and lysophosphatidylcholine plus 25 % cholesterol, 10% cerebroside, 15% ce ramide, and 25 % triglyceride. The method was sensitive over two conveni ent ranges, 10- 60 J.l- g and several minor phospholipids that were not identified (e.g., spot 100- 300 J.L g, and reproducible with an error of less than 10%. Proteins #5 in Fig 1). we re determined by t he method of Lowry et al [21] on samples digested Glycolipids in LG were present as 3 s pots in the TLC system in NaOH. used; 2 had the mobility of glucosylceramides and the least polar (spot #8) cochromatographed with acylglucosylceramide. Enzymes Ceramides were present as 5 separate spots (9-13). We have E nzy mes were as~ayed in aliquots of subcellular fractions containing previously reported resolution of 4 glucosylceramides and 6 20- 200 J.l-g protein. For each enzy me, specifi c activity per mg protein ceramides from LG lipids [13) using several different systems was cal culated from the units of activity. of TLC which provide greater discrimination of each type of Acid hydrolases in subcellular fractions were assayed u sing parani sphingolipid. For purposes of this study we used 2-dimensional trophenyl substrates (Sigma) in co ncentrations and at the pH shown TLC to provide quantitation on a single plate. Free sterols were in T abl e l. Assays were co nducted as previously described in a system present in spot #14. containi ng 0.0.5 % Triton-X 100, 0.05 M acetate buffer in a total volume of 0:5 ml for 30 min at 37"C [22,23]. Reaction products were determined The content of lipid and protein of each fraction was quan spectrophotometrically and ex pressed as arbit rary units of optical titated (Table II). LG had a higher lipid/ protein ratio t han t he density. Appropriate s ubstrate co nce nt rations a nd linearity of the other 2 fractions and the crude starting 17,000 g (17K) fraction. reactions were determined in preliminary experiments. The lipid/ protein ratio of 2.0 was consistent with its low Phospholipase A was assayed as previously described [24]using 1,2- buoyant density in metrizamide (0.8- 1.0) [12). Approximately di([1 -14C ]oleoyl}phosphatidyl choline (PC) (Amersham Searle) as sub 45% of t he lipid in the 17K pellet was recovered in the 3 strate. Assays we re conducted with 0.5 ILCi PC (0.34 mM) in a system part iculate fraction s and of this 14% was present in LG. co ntaining 0.25 % taurodeoxyc holate, 0.05 M acetate buffer, pH 4.6, in Individual lipid components were quantitated in the various a vo lume of 0.5 ml for 45 min at 37"C. Activity of phospholipase A was calcul ated as % substrate hydrolyzed to lysophosphatidyl cho line and subcellula r fractions (Table III). The LG fraction most resem fatty ac id . Sphingomyelinase was assayed by the method of Bowser bled the 17K strating fraction and was rich in sphingolipids and Gray [25] using [N-methyl-14 C]sphingo myelin (Amersham-Searle) including sphingomyelin, glucosylcera mides, and ceramides 1 IJ Ci, 0. 34 mM in 0.25% Triton-X 100, 0.05 M acetate buffer, pH 4.8, which accounted for almost 33% of the total lipids in t his in a volume of 0.5 ml for 1 h at 37•c. Resul ts we re ex pressed as % water soluble radioactivity (l argely [' 4C]phosphorylcholine) liberated from t he s ubstrate. Steroid ~ ul fatase was assaye d b y the method of Sf2 Ruokonen and Oikarinen [26 ] using [7-"H]dehydroepiandrosterone sulfate (New England Nuclear). 500 ,000 dpm (sp act 25 I'Ci/ mmol) C) we re incubated with subcellular fractions in phosphate-buffered saline 13 0 at pH 7.3 in a total vo lume of 0. 7 ml for 2 h at 37•c. Free "H-labeled 0•• 16 steroid was quantitatively extracted in ethylacetate:ethyl ether (1:9) c)D•o 0 14 CJ,5 and r es ult. ~ ex pressed as % of substrate hydrolyzed. ~ 0 c::::::;?709 N c::?6 RESULTS :r: u The lipid composition and enzyme activity were character
TAB LE II. Lipid conl.ent of subcellular fra ctions TABLE V. Steroid sulfatase activity
mg/g rng/rng Perce nt of Subcellular fractions• Perce nt activity• Relative speci fic Fraction (n)" Epi dermis Protein total lipid activity'" ±SO'' ±so• 17,000 g Supern atant 71 .6 ND 17K (6} L8ll ± .188 0.388 ± .020 100 17,000 g Pellet 28.4 1.0 LG (6) 0. 260 ± .1 32 2.000 ± .664 14.3 LG 0. 3 0. 34 Mitochondri a (2) 0.070 0.890 3.8 Mi tochondria 0.5 0.23 Fr III (2) 0.490 0.460 27.0 Fraction lii 5.3 10.7 a n = Number of experim ents, eac h uti li zin g 1200 mg o f fetal rat " Fractions prepared as per text; abbreviations per Tabl e II. Approx epidermis fo r separation of fr actions into a 17,000 g particul ate fraction imately 1.0 g epidermis was used in this represe ntative experim ent. (17K), subsequently subfractionaLed in to lamell ar granul e \LG), mtto b Percent total activity found in each fractio n derived from 700 g cho ndrial, and mix ed mttochondnal, ce ll enve lope (Fr III) fra ctiO ns. supern atant. Assays co nducted as per text with 45-150 1-'g protein in b Lipids a nd proteins were asses ed in washed particulate fracttons eac h of duplicate reaction vessels. Activity expres. ed as % of starti ng obtained a· per text. dp m in l"H]dehydroepi androsterone co nverted to free steroid. c Relative s pec ific activity calculated as in Table III; i.e., ratio of activity/mg protein in each fraction relative to that in 17,000 g pellet. TABLE Ill. Lipid composition of subcellular fractions Perce nt of tot.al lipid ± SO Lipid nase. Of t hese 8 e nzymes, only the p hospholipase could be 17 1< " LG Mit ochondria Fr Ill demonstrated in significant quantities in mitochondria, and acid phospholipase showed some locali zation as well in fraction Phosphoglyce rides" 26.5 ± 1.5 24.5 ± 1.3 60.8 ± 5. 1 10.4± 1.8 III. Otherwise, no enzymes appeared to be specifically locali zed Sphingo mye lin 5.3 ± 0. 2 9.4 ± 0. 3 2.6 ± 1.0 to t his mixed fraction . Glycosphingo li pids 28.8 ± 1.4 23.4 ± 1.0 16. 3 ± 2 .3 We also examined the fractions for presence of steroid sul a nd ce ramides fatase (Table V). Like all other enzy mes it was present in bot h 30.8 ± 8 .2 11.7 ± 3.8 Sterols 14.5 ± 2.4 15.2 ± 2.6 17K p ell et and supernatant. It was not specificall y locali zed in Total 75.0 72.5 91.5 41.0 eit her LG or mitochondria but showed a 10-fold increase 111 a Resul ts represe nt the mean of 4 experim ents as % of total lipid ± specific activity in fraction III as compared to 17K. SD. Fractions 17K, LG, mi toc hondria, Fr Ill are as defi ned in Table II. DISCUSSION b P hosphoglyce rides are all phosphorus-containing lipids except sphingo mye lin . We have recent ly demonstrated t he presence of t he glucosyl ceramides and of acylglucosylceramide typical of stratum cor neum in t he rat epidermal LG [1 3]. In t his report we confirm TABLE IV. Acid hydrulases of subcellular fra ctions t he abundance of lipid and presence of a variety of acid hydro Relative specific activity ± SO" lases in t his fraction. Our data show t hat LG seem to be composed primarily of lipids wi th a low content of protein as Mitochondria Fr III LG would be expected if they are assembled from stacks of fl attened 1 Acid phosphatase 7.1 ± 1.9 ' 0.1 1.0 ± 0. 7 liposomes as has been suggested [16, 17]. Arylsulfar.ase 0.8 ± 0 0 1.0 ± 0.8 T aken together, the above findings provide t he fi rst direct Galactosidase 1.4 ± 0.2 0 L1 ± 0.1 biochemical evidence in support of t he hypothesis t hat stratum 0.9 ± 0.8 Galactosaminidase 0.8 ± 0.1 0 corneum lipids derive from t he exocytosis of LG. Alt hough 2.0 ± 0.4 h 0 0.8 ± 0.8 G Iu cosidase phospholipids are markedly decreased in rodent stratum cor G lucosaminidase 0.8 ± 0.4 NO 0.9 ± 0.3 Acid phospholipase A 3.6 ± 1.3b 2.6 ± 0.2b 1.6 ± .05b neum, their presence in LG is hardly s urprising. LG are Sphingomye lina ·e 4.0 ± 0.8" 0.6 ± 0 0.3 ± 0.2 bounded b y limiting membranes derived from t he Golgi appa ratus and t hus are likely to be rich in phospholipids. These n Resul ts rep resent th e mea n ± SO of 3 experiments a nd are ex limiting membra pressed as the mea n of the ratios of the spec ifi c activity of the enzym_e nes are probably retained in t he granular cell in each fraction/spec ilic activity in J 71< in eac h expenment. Expen when t he granules fuse with plasma membranes for exocytosis ments we re performed as per text usin g was hed, repelleted subcellular of their lamellated contents. Moreover, other phospholipids fractions for enzy me assays and protei n analys is. Abbrevwttons as tn contribut ing to t he stacked discs may be hydrolyzed after Table II. extrusion (see below). If one considers t he composition of t he b Denotes statistica ll y significa nt in creases in relative spec ific activ- LG li pids without phosphoglycerides, free sterols account for ity ( p < .05 by Student's /-test). 21%, glycolipids and ceramides for 31%, and sphingo myelin for 12%; t hese proportions are close to those reported for rodent stratum corneum [3]. fract ion. Mitochondria differed markedly, consisting mainly of It should be noted t hat alt hough LG lipids are very s imilar sterols (30%) and phospholipids (60%). Fraction III, t he mixed in distribution to t hose of t he crude 17K fraction, we recovered fraction containing cornified cell fragments, had smaller only 14% of 17K lipids in t he LG fraction. We believe t hat t his amounts of phospholipids; moreover t he amount of gluc?s;yl is in part due to presence of stratum corneum in 17K but as ceramides and ceramides was greater relative to phospholipids well to t he poor yield of LG after even mild processing. Our t han in the LG fraction. previous studies of the 17K fraction suggested it was g reatly enriched in LG but was contaminated wit h mitochondria a nd Enzyme:; some remnants of cornified cells [23]. The procedure used for A ll of t he e nzymes we assayed were present bot h in the 17K isolating LG removes these contaminants but a pparently a lso pellet and the supernatant fr action. T o assess specific locali causes significant losses of LG. zation of a n enzy me, we compared its spec ific activity in t he These studies have also demonstrated concentrations of cer washed particulate fraction to t hat in 17K (Table IV). In t his tain specific lysosomal type hydrolases in the LG fraction. In convention, a ratio of activity in the fraction vs t hat of 17K addition to acid phosphatase [1 2] we have now shown locali greater t han 1.0 would suggest localization of t he enzy me to zation of glucosidase, acid phospholipase A, and sphingo mye t hat fraction. Of t he 8 acid hydro lases, 4 showed greater specific linase in t he LG fraction. On t he other h and, we fa iled to activity in t he LG fraction t hm1 in 17K, including acid phos demonstrate preferential localization of other acid hydrolases phatase, glucosidase, acid phospholipase A, and sphingo myeli- such as arylsulfatase and other glycosidases. We have not as 298 FRElNKEL AND TRACZYK Vol. 85, No. 4 yet examined the presence of proteolytic enzymes or glucuron 6. Wolff-Schreiner EC: Ultra ·tructural cytochem istry of the epider mi s. Int ,) Dermatol 16:77- 102, 1977 idase. The latte r, at least, is n ot localized in t he starting 7. Landmann L: Lamell ar granules in mammalian, av ian, and repti fraction: l7K pellet [23). li an epidermis. J Ultrastruct Res 72:245- 263, 1980 Arylsulfatase (pH 5.5) has been demonstrated in LG by 8. Wolff K, Schreiner E: Differential enzy matic di gestion of cyto ultrastructural histochemical techniques in mouse epidermis pl asmic components of keratinocytes. J Ultrastruct Res 36:437- 454, 1971 [27]. In other studies, it could not be shown to be present 9. Wolff K, Holubar K: Odland Korper als epiderm al Lysomen. Arch [28] . Our failure to localize it in the G fraction must t herefore Klin Exp Dermatol 231:1- 19, 1967 be viewed somewhat tentatively since it may be present in too 10. Weinstock M, Wil gram GF: Fine structural observations on the few of the LG [27] to allow for biochemical localization in the form ation and enzymatic activity of keratinosomes in mouse !iliform papillae. J Ultrastruct Res 30:262-274, 1970 presence of the enzyme in oth er organelles. 11. Smith WP, Christensen MS, Nacht S, Cans EH: Effect of lipids However, it does not seem fortuitous t hat the enzymes we on the aggregation and permeability of human stratum co rneum. found were glucosidase and acid pbospholipases. It is likely J Invest Dermatol 78:7- 11, 1982 that these enzymes serve specific functions with respect to the 12. Freinkel RK, Traczyk TN: A method for partial purification of lamellar granules from fetal rat epidermis. J Invest Derm atol content of LG. Ceramides are more abundant in relation to 77:478- 482, 1981 glycolipids in stratum corneum than in granular cells of murine 13. Wertz PW, Downing DT, Freinkel RK, Traczyk TN: Sphingolipids epidermis [3]. Simila rly, the relationship between the glucosyl of stratum corneum and lamellar granules of fetal rat epidermi s. ceramides and ceramides of LG vs stratum corneum of rat J Invest Dermatol 83:193- 195, 1984 14. Gray GM, White RJ: Glycosp hingo lipids and ceramides in human epidermis has suggested t hat hydrolysis of the former gives rise and pig epidermis. ,J Invest Dermatol 70:336- 341, 1978 to the latter [13]. Such remodeling may be facilitated by the 15. Wertz PW, Downing DT: Glycolipids in mammalian epidermis: glucosidase and phospholipases that are probably extruded into structure and function of the water barrier. Science 217: 1261- the extracellular space together with the lipid-containing discs. 1262, 1982 16. Landmann L, Wertz PW, Downing DT: Acylglucosylceramides How this correlates with barrier funct ion and desquamation from pig epidermis induce flattening and stacking of liposomes remains to be determined. In this regard, a recent report that (abstr). J Invest Dermatol 82:397, 1984 stratum corneum cell s fai l to reaggregate after treatment with 17. Landmann L: The epidermal perm eability barrier: comparison a and (3 glucosidases may be of interest [ll]. betwee n in vivo and in vitro lipid structures. Eur J Cell Biol 33:258- 264, 1984 No function has yet been ascribed to the acid phosphatase 18. Grayson S, Johnson-Winegar HD, Eli as PM: Isolation of lamellar present in LG and readily demonstrable in t he extracellular bodies from neo natal mouse epidermis by selective sequential spaces of stratum corneum layers. Since acid phosphatase is a fi ltration. Science 221:962- 964, 1983 characteristic enzyme of many lysosomal organelles, it is pos 19. Folch J , Lees M, Sloane-Stanley GH: A simple method for iso lation and purification of total lipids from animal tissues. J Biol Chem sible that in t he case of LG it represents a nonfunctional marker 226:497- 505, 1957 of the lysomal origin of the granule. 20. Amenta JS: A rapid chemi ca l method for quantification of lipids Finally, we were unable to demonstrate localization of steroid separated by thin layer chromatography. J Lipid Res 5:270-272, sulfatase activity among the hydrolases of LG although it was 1964 abundantly present in l7K supernatant and in fraction III 21. Lowry OH, Rose nbrough NJ, Farr AL, Randall RJ: Protein meas urement with Folin phenol reagent. J Biol Chem 193:265-275, which is enriched with stratum corneum membranes. These 1951 findings are con sisten t wit h published evidence t hat the enzyme 22. Ohkawara A, Halprin KM, Taylor .JR, Lev in e V: Acid hydrolases is localized in microsoma l m embranes [29] and stratum cor in human epidermis. Br J Dermatol 83:450- 459, 1972 neum (30]. Although apparent lack of activity does not exclude 23. Freinkel RK, Traczy k TN: Acid hydrolases of the epidermis: sub cellular locali zation and relation to co rnification. J Invest Der the possibility that the enzyme is packaged in LG in an inactive matol 80:441- 446, 1983 form and activated only after extrusion, t he present findings 24 . Freinkel RK, Traczyk TN: The phospholipases A of epidermis. J do not shed any light on the mecha nism by which steroid Invest Dermatol 74:169- 173, 1980 sulfatase might find its way to the membrane regions of stratum 25. Bowse r PA, Gray GM: Sphingomyelinase in pig and human epi dermis. J Invest Derm atol 70:331- 335, 1978 corneum as has been reported [30). 26. Ruokonen A, Oikarinen A: Steroid sulphatase activity in the skin biopsies of various types of ichthyosis. Br J Dermatol 105:291- REFERENCES 295, 1981 27. Gonzales LF, Krawczyk WS, Wilgram GF: Ultrastructural obser 1. Eli as PM, Friend OS: The permeability barrier in mammalian vations on the enzymatic activity of keratinoso mes. J Ultrastruct epidermi s. ,J Cell Biol 65:180- 191, 1975 . . . Res 55:203- 211 , 1976 2. Elias PM, Goerke J , Friend OS: Permeabili ty barner hptds: com 28. Rowden G: Ultrastructural studies of keratinized epithelia of the pos ition and inf1u ence on epiderm al structure. J Invest Dermatol mou se fl. Aryl sulfatase in lysosomes of mouse epidermis and 69:535- 546, 1977 esophageal epithelium. J Invest Dermatol 51:51- 61 , 1968 3. Elias PM, Brown BE, Goerke ,J, Gray GM, White RJ : Localization 29. French AP, Warren JC: Properties of steroid sulfatase and aryl and composition of lipids in neonatal mouse stratum granulosum sulfatase activities of human placenta. Biochem J 174:199- 208, and stratum co rn eum . J In vest Dermatol 73:339- 348, 1979 1976 4. Gray GM, Yardley HJ: Di fferent popul ations of pig epiderm al ce ll s: 30. Williams ML, Grayson S, Bonifas TN, Epstein EH, Elias PM: iso lation and lipid co mposition. J Lipid Res 16:441- 447, 1975 Epidermal cholesterol sulfate and steroid sulfatase activity and 5. Gray GM , Yardl ey HJ: Lipid co mposition of ce ll s isolated from recessive x- linked ichthyosis, Stratum Corneum. Edited by R pi g, huma n, and rat epidermis. J Lipid Res 16:434-44·0, 1975 Marks, G Plewig. Berlin , Springer-Verlag, 1982, pp 79-85