Abnormality of Water Barrier Function in Psoriasis

Role of Ceramide Fractions

Stefania Motta, PhD; Marcello Monti, MD; Silvia Sesana, PhD; Luisa Mellesi, PhD; Riccardo Ghidoni, PhD; Ruggero Caputo, MD

Background and Design: In psoriasis the formation of the cornified layer is deranged and water flux is reportedly increased. We investigated three different forms of psoriasis: transepidermal water loss was measured on uninvolved skin and psoriatic plaques; lipids from plaques were extracted; and ceramide distribution in scale vs normal stratum corneum was compared. Moreover, the lipid biochemical results were compared with transepidermal water loss rates in the same lesions. To assess potential alteration in ceramide distribution, lipids from both psoriatic scale and normal stratum corneum were extracted by the Bligh-Dyer method, separated o n high performance thin layer chromatography plates, and quantified by computerized densitometry. Water flux was measured as transepidermal water loss using a n evaporimeter; results between uninvolved and involved psoriatic

skin and age-matched control skin were statistically evaluated.

Results: In comparison with normal stratum corneum, psoriatic plaque ceramides showed a different distribution; in particular, ceramide 1 was significantly decreased. The increased transepidermal water loss values of psoriatic plaques vs control skin and between psoriatic involved vs uninvolved skin were significant. Conclusion: Our findings indicate that in psoriasis the

altered ceramide distribution can be linked specifically to the defect in keratinocyte differentiation; the defect in skin barrier function may be attributed largely or in part to ceramide 1 reduction.
(Arch Dematol. 1994;130:452-456)

From the Department of Dermatology, University of Milano and IRCSS, Ospedale Maggiore (Drs Motta, Monti, Sesana, Mellesi, and Caputo) and Department of Medical Chemistry and Biochemistry, University o Milano f (Dr Ghidoni), Italy.

SORIASIS 1s characterized by abnormal epidermal maturation leading to an incomplete process of differentiation. The main pathophysiologic event in psonasis is an increase in keratinocyte proliferation.' Excess production of cells, coupled with abnormal cohesiveness, leads to excessive stacking of psonatic parakeratotic scales and thickening of the stratum corneum to about 10 times greater than normal. The stratum corneum provides the permeability barrier that regulates the transcutaneous water loss, a necessity for terrestrial life. In psoriasis, as in several other scaling dermatoses,* these permeability functions are i m ~ a i r e d ~ . ~ an causing increase i n transepidermal water loss (TEWL) . Three intercellular lipids, sphingolipids, free sterols, and free fatty acids, are implicated in the permeability bamer prop-

erties of the stratum c ~ r n e u m Both bar.~,~ rier and water retention properties' require sphingolipids; in fact they are the largest lipid fraction by weight8; their removal is associated with an increase in water 1oss9;and they are enriched in linoleic acid, a fatty acid essential for permeability barrier function.1 Epidermal sphingolipids comprise a heterogeneous class of ceramides and glucosylceramides. Such heterogeneity is due to (1) different Nacyl chain lengths; (2) differences in type and extent of hydroxylation; ( 3 ) presente of either the sphingosine or phytosphingosine base; and (4) presence of additional acylation at the omega terminus

ARCH DERMATOWOL 130.APR 1994

452

MATERIALS AND METHODS

SOURCE OF SKIN SAMPLES Fresh abdominal skin samples of six normal subjects of both sexes were obtained from plastic surgery. After scraping off the subcutaneous fat with a surgical blade, the full-thickness skin was incubated at 37OC dermisside downward for about 4 hours in a solution of 10 mM sodium edetate in phosphate-buffered saline, calcium. and magnesium free. Once dermoepidermal separation was obtained, the epidermis was washed twice in O.gO/ sodium chloride solution for 10 minutes and soaked with 0.5% trypsin in phosphate-buffered saline for 1 hour at 37"C.15 After overnight incubation at 4C in a fresh 0.5% trypsin solution, the stratum corneum was cleaned of residual granular cells with filter paper. The stratum comeum sheets were dried overnight in a silica gel dryer and weighed. Abdominal psoriatic scale samples were obtained from untreated (>2 weeks), neither systemically nor topically, lesiona1 areas of six male psoriatic subjects. Psoriatic phenotypes were clinically identified as exfoliative/erythrodermic psoriasis,two cases; annular-plaque psoriasis, three cases; and rupioid psoriasis, one case. The scales were removed from the skin with tweezers, collected in a Petri dish, stored overnight in a silica gel dryer, and weighed. Lipid Extraction Normal stratum comeum sheets and psonatic scales were extracted by che Bligh-Dyer method,16 filtered (Whatman No. l ) , and split into aqueous and organic phases. The organic phases were dned under nitrogen gas and stored in chloroform at -40C until used. Thin-Layer Chromatography The whole lipid extract from each of the normal stratum corneum and psonatic scale samples was separated on precleaned, high-performance, thin-layer plates, 1 0 x 2 0 cm (HPTLC, Pre-Coated Plates Silica Gel, Merck, Darmstadt, Germany). Thin-layer chromatography (TLC)-plates were prewashed in a solvent system consisting of c h l o r o f o d

methanoVwater/glacial acetic acid (120/70/9/1 [v/v/v/v]) and developed in chambers saturated with the appropriate solvent system. Al1 separations were carried out at room temperature. The major neutral lipid species were separated petroleum etherldiethyl ether/glacial acetic acid (80/20/1 [v/v/v])." Then, the material located at the origin of the TLC plate, after visualization with 0.25% aqueous 8-anilino- 1-naphthalene sulfonic acid under UV illumination, was scraped, eluted from the gel, desiccated, and resuspended in chloroform. One hundred micrograms of the polar lipid fraction (10 to 20 pi.) were spotted o n a new TLC plate. Elution was performed twice with chloroform/methanoVglacial acetic acid (1 90/ 9/1 [v/v/v]) to separate the ceramide species. Lipid standard mixtures (20 pg each) and reference ceramide fract i o n s ( 1 0 p g each) were a p p l i e d i n parallel for identification and calibration purposes. Ceramide standards, ceramide type 111 (CER 3), and ceramide type IV (CER 41, were purchased from Sigma Chemical Co (St Louis, Mo). Seven human ceramide fractions (CER 1 to CER 611 ) were purified in our laboratory by preparative TLC. Structural analysis of individual ceramides was performed by gas-liquid chromatography to establish the composition in fatty acids and elution, the plates were air-dried, long-chain bases. Af~er sprayed with 10% copper sulfate, 8% phosphoric acid aqueous solution, and charred on a 180C hot plate.I8 Quantification of individual ceramide species was performed by densitometry, using a Camag TLC Densitometer equipped with a computerized image analyzer. Student's t test was performed for statistical purposes. TEWL MEASUREMENTS The TEWL measurements were performed using an evaporimeter (EP1, Servomed, Sweden). The results were read 30 seconds after probe application to the skin, as described previo~sly.'~ studied nine psoriatic patients in We whom we correlated TEWL with scaling and fissuring, as well as TFWL after scale removal using cellophane tape stripping. For each patient the result was expressed as TEWL (mean+SD of 3 to 5 measurements). As a control, we used the uninvolved forearm skin of the same nine psoriatic patients and the forearm skin of 10 age-matched control subjects. Student's t test between these two groups was performed for statistical purposes.

of the N-acyl group." This last modification results in the so-called acylceramides and acylglucosylceramides.12 In particular the acylceramides and their glycosylated precursors, acylglucosylceramides,have been postulated to carry out membrane-organizing functions in the stratum corneum and membrane-stacking functions in the lamellar body,13,14 respectively. These observations are consistent with a requirement for ceramides in the control of the epidermal

permeability barrier. We are proposing that in psoriasis, in which the formation of the cornified layer is deranged and in which water flux is reportedly increased, an alteration in ceramide content or distribution may occur. For these reasons we compared ceramide distribution in scale derived from different forms of psoriasis vs normal stratum corneum, and compared the lipid biochemical results with T E W rates in the same lesions.

ARCH DERMATOWOL 130, APR 1994 453

CER 1 CER 2 CER 2' CER 3" CER 4 CER 5 CER CER 6 CER 611

amide distribution was observed in al1 of the psonatic vanants considered in this study. The quantitative results are summarized in Figure 2 in which lipid weight mean percent, SDs, and statistical significances between the ceramides are shown. TRANSEPIDERMAL WATER LOSS
9

10

11

Figure 1. Ceramides (CER) of the normal human stntum corneum and the psoriatic scale. Lanes 1 to 7 indicate isolated ceramides from normal human ,; stratum corneum: CER 1 to CER 6 ,lane 8, ceramides from normal human stratum corneum; lane 9, ceramides from psoriatic scale; and lanes 10 and 11. reference ceramides from Sigma Chemical Co (CER 11 and CER IV). 1

.. . 51 -. Normal
Corneum CER 1 CER 2, CER 3 CER 4 4 CER 51 CER 6, CER 61,

Designation of lsolaited Ceramiides of the Psoriatic S and the IUormal Hurnan Straturn Corneum
rsoriaric Scale CER 1 CER 2, CER 211 CER 3 CER 4 CER 51 CER 51 1 CER 61
CFR C., -,,

Rf*

O 69 0.61 0.56

CEF CEF

The mean TEWL value of scaling psoriatic plaques was 11.5, while the fissuring plaque and the scaling psonatic plaque after scale removal displayed mean TEWL values of 20.9 and 29.2, respectively. The mean TEWL value of the age-matched control subjects was 4.3; while uninvolved psoriatic skin displayed a mean TEWL value of 6.3. No statistical significance (Student's t test) between mean TEWL values of healthy subjects and the uninvolved skin of psoriatic patients was observed. In contrast, the TEWL values of psoriatic plaques vs healthy skin and between psoriatic involved vs uninvolved skin were significant. These results are summarized in Figure 3.

0.39 0.29 O 17 0.14

CERS

CEF CEF
CFI., -,,

--. .

"-.

* Rf indicates retention factor.

CERAMIDE CONTENT AND COMPOSITION Total ceramides in psoriatic scale and in normal stratum corneum, quantified by densitometry, ranged from 75% to 80% of the total polar lipids. Thin-layer chromatography separation of psonatic scale lipids generated nine different fractions of ceramides (Figure 1), while normal stratum corneum displayed seven distinct species. The additional species,after isolation and gas-liquid chromatography analysis, was found to contain phytosphingosine and CI6-Cl8 nonhydroxy fatty acids. This spot was named CER 4. Previously CER 4/5 was confirmed to contain sphingosine and a-hydroxy fatty acids and was renamed as CER 511 (Table). In psonasis two new fractions appeared, which were designated as CER 2,, and CER 5, as shown in the Table. The densitometric quantification of the individual ceramide fractions showed certain consistent differences between the fractions in psoriatic scale vs control stratum corneum. In al1 types of psoriatic scale, CER 1, 3,4,5,,, and 61 content was decreased, while CER 2,, 211, and 5, content was increased. This same pattem of cer-

The bamer function of the stratum corneum is impaired in parakeratotic psoriatic plaques. In our patients, the water flux of psoriatic plaque was a mean value of 11.5 g/m2 per hour. Although this value is significantly higher than normal, it is lower than that obtajnedwith 1%sodium lauryl sulfatepatch-tests after 24 hours in the absenceof clinical signs in a comparable control group. Even psonatic fissured plaques andscaling psonatic plaques, after scale removal, gaveTEWLvalues of 20.9 and29.2 g/m2perhour, respectively, which, in our expenence, are still lower than those produced with the 1% sodium lauryl sulfate patch test when it produced an erythematous scaly reaction. Marks et al2' studied TEWL in psonasis and demonstrated increasedwater loss acrosspsoriatic plaques in comparison with the uninvolved skin. Moreover, Takenouchi et al2'demonstratedan inverserelationship betweenTEWL and skin surface hydration, evaluated in terms of conductance, in psonatic lesions.Thus, in this disease the pathologic homy layer is unable to bind water but instead passes water through it. SerupandBlichmannUalcocompared the defects in epidermalbamer function with clinical signs of scaling in psoriasis. Moreover, these authors attnbuted the reduced hydration (low conductance) of psonatic skin to an abnormal proportion of polypeptides present in psonaticscale. Ina highly derangedstratum comeum, asin psonatic plaque, the low conductance may be due to many factors other than the ability to retain water by stratum corneum proteins, ie, the organization of the intercellularlipid compartment. In light of the great degree of homy layer derangement in psoriatic plaques and the increment of epidermal water flux caused by anionic detergents, the nearnormal bamer functionin psonasis may be due to compen-

ARCH DERMATOWOL 130, APR 1994 454

sation by other factorssuchasstratum comeum thickness, excessive number of stratum comeum cell layers, or, alternatively, an alteration in the content andlor distribution of intercellular, lamellar lipids. Of these factors,lipids desewe more attention because their removal provokesan increase in TEWL.3.4Many studies haveshown thatwater retention is mainly due to the presence of intercellular l i p i d ~ ~ ~ and that ceramides togetherwithcholesteroland fatty acidsplay Abundant india critica1role in this important f ~ n c t i o n . ~ ~ rect evidence points to a special role for ceramides in the bamer.12This hypothesis is supported by the presence of large amounts ofsaturated,very-long-chain N-acyl fattyacids in ceramidess."; by the presence of linoleic acid in selected cera mide^,'^.^^ and by Elias andcow~rkers'studies~~ on the epidermisof manne cetaceans,which, having much less stringent bamer requirements, display much shorter chain-length unsaturated N-acylated fatty acids in their sphingolipids. Conversely, recent investigations stressed the relationship between bamer functionandlipidsynthesis.Damage to the barrier with either detergents, tape stripping, or lipid solvents increases epidermal sterologenesis and fatty acid synthesis proportional to the degree of barrier disruption, and when the hyperpermeability is corrected with animpermeable film, the cholesterologenesisretums to n o r m a l v a l u e ~ . ~ ~ ~ ~ ~ A l s o , recentstudies have shown that animmediateburst inbothepidermalcholesteroland fatty acid synthesis occurs after artificiavacute barrier disruption, which normalizes over 6 hours in parallel with the rapid early phase of re~overy.~'-~~ In essential fatty aciddeficient hairless mice, animals with both defective barner functionand epidermalhyperplasia, anincrease in epidermal cholesterol, sphingolipid, and total nonsaponifiable lipid synthesis has been ~ b s e r v e d . ~ ~ lovastatin, When an inhibitor of cholesterol synthesis, is applied topically to rodents, it produces an alteration in fatty acid and cholesterol synthesis coincident with barrier di~ruption.~' In this condition, as in essentialfatty acid deficiency, epidermal hyperplasia occurs, presumably as part of the repair r e ~ p o n s eConsequently, it is likely that psoriasis, a dis.~~ ease model charactenzed by both a deranged barrier and epidermal hyperplasia, also will result in an alteration in lipid synthesis. In fact, psoriatic scale reportedly displays and free an increase in free ste1-01~~~ fatty a c i d ~ . ~ ~ According to recent studies by Holleran and co~ o r k e r sin~ ~ , which barrier disruption with acetone or tape stripping or in essential fatty acid deficiency resulted in increased sphingolipid synthesis and activity of its rate-iimiting enzyrne, serine palmitoyl transferase,we would expect that in psonasis there also would be an increase in sphingolipids.However, we found the same relative content of ceramides in psoriatic scale as in normal stratum corneum, although the procedures used in homy layer hawesting were different. However, chromatographic separation of ceramide species displayed a differentpattern of ceramide distribu-

25

20

15
C

f 1o
5

o
CER 1 CER 21 CER 211 CER 3 CER 4 CER 51CER 511CER 61 CER 611

Figure 2. Ceramides (CER) in the psoriatic scale and normal human stratum corneum. Data are expressed as percentage mean (sk samples), SD, and statistical significancy (Student's t test). NS not significant.

35 30

2
S E
Cd

25 20 15

N+ s

lo 4.311.2

( J* ?
Psoriatic
Skin

ro1 0
29.1t9.8 20.9*8.02

Healthy Subjects

Uninvolved

Psoriatic
Plaque

Psoriatic Plaque After Scale Rernoval

Fissured Plaque

Figure 3. Transepidermal water loss (TEWL) values in psoriatic subjects. Dagger indicates Student's t test; asterisks, mean values. NS indicates not significant.

tion in psoriatic vs normal specimens. In psoriatic scale, regardless of clinical phenotype, the phytosphingosinecarrying ceramides (CER 3 and CER 61) showed a statistically significant decrease vs normal stratum corneum; some sphingosine-carrying ceramides (CER 2,, CER 2,,, and CER 5,) increased, while others (CER 1,CER 511, and CER 4) decreased. Interestingly, this new pattem of ceramide fractions in psonatic scale was obsewed in al1 psoriasis studied independent of patient age, topographic cite of hawesting, and, as noted above, clinicalpsoriatic type. Thus, it is possible that thisceramide pattemis linked specificallyto the defect in differentiation in psoriasis, or, alternatively, it could reflect epidermal hyperproliferation. However, in atopic eczema, another scalinginflammatorydisorder investigated for ceramide content, a comparable abnormality in ceramide distnbution has not been o b s e ~ e d . ' ~ Current knowledge about epidermalceramidesynthesis does not permit speculationaboutthe cause ofvanations found in psoriatic scale,apart from the significant relative

ARCH DERMATOWOL 130,APR 1994

455

13. Downing DT, Stewart ME, Wertz PW, Colton SW, Abraham W, Strauss JS. Skin decrease when compared with normal stratum corneum in lipids: an update. J lnvest Dermatol 1987;88:2s-6s. CER 1fraction, an o-ac~lcerarnide ester ennchedinlinoleic 14. Landmann L, Wertz PW, Downing DT. Acylglucosylceramide causes fiattening acid. These data stand in strong contrast to the great availand stacking of liposomes: an analogy for assembly of the epidermal permeability barrier. Biochim Biophys Acta. 1984;778:412-418. ability oflinoleic acid in psoriatic ~ c a l e . ~ ~linoleic acid Since 15. Elias PM, Brown BE, Fritsch PO, et al. Localization and composition of lipids levels are increased, and other esterified derivatives of linoleic acid are present in large abundance in psoriatic ~ c a l e , ~ ~ in neonatal mouse stratum granulosum and stratum corneum. J lnvest Dermatol. 1979;73:339-348. we can hypothesize that theimpairment for CER 1produc16. Bligh EG. Dyer WJ. A rapid method of total lipid extraction and purification. Can tion could be at the level of the enzyme linking N-acyl J Biochem Physiol. 1959;37:911-917. 17. Lampe MA, Burlingame AL, Whitney JA, et al. Human stratum corneum lipids: o-hydroxy fatty acids with fatty acyl COA. characterization and regional variations. J Lipid Res. 1983;24:120-130. Ceramide type 1is believed to be important for wa18. lmokawa G, Abe A, Jin K, Higaki Y, Kawashima M , Hidano A. Decreased level ter retention as deduced from the highly permeable and of ceramides in stratum corneum of atopic dermatitis: an etiologic factor in scaly skin of essential fatty acid-deficient animals in which atopic dry skin? J lnvest Dermatol. 1991;96:523-526. 19. Blichmann C, Serup J. Reproducibility and variability of transepidermal water the defective barrier is attributable to substitution of oleloss measurement: studies on the Servo Med evaporimeter. Acta Derm Veneate for lin~leate.~' of the linoleate in epidermal sphinMost reol. 1987;67:206-210. golipids is ester linked to o-hydroxy-acids, which in turn 20. Marks J, Rogers S, Chadkirk B, Shuster S. Clearance of chronic piaque psoare amide linked to sphingosine (CER 1).It has been sugriasis by anthralin-subjective and objective assessment and comparison with photochemotherapy. Br J Dermatol. 1981;105:96. gested that this molecule is critica1 for lamellar bilayer 21. Takenouchi M. Suzuki H, Tagami H. Hydration characteristics of pathologic formation in the straturn corneum, and hence for mainstratum corneum-evaluation of bound water. J lnvest Dermatol. 1986;87:574tenance of the permeability barrier.38'' Similarly, it is pos576. sible that the defect in skin barrier function in psoriasis 22. Serup J, Blichmann CW. Epidermal hydration of psoriasis plaques and the relation to scaling. Acta Derm Venereol. 1987;67:357-366. may be due largely or in part to CER 1reduction. In any 23. lmokawa G, Akasaki S, Hattori M, Yoshizuka N. Selective recovery of deranged case, further studies will be needed to clarify the speciwater-holding properties by stratum corneum lipids. J lnvest Dermatol. 1986; ficity of these findings. 87:758-761.

Accepted for publication May 28, 1993. W e gratefully acknowledge Peter M. Elias, MD, for giving useful suggestions during the experimental work and for suggestions in writing the manuscript. Repnnt requests to the First Department of Dermatology, University of Milano, Via Pace 9, 20122 Milano, Italy (Dr Motta).

1 . Rothberg S, Crounse RG, Lee J. Glycine: C14 incorporation into the proteins of normal stratum corneum and the abnormal stratum corneum of psoriasis. J lnvest Dermatol. 1961;37:497. 2. Williams ML. Lipids in normal and pathological desquamation. Adv Lipid Res. 1991;24:211-262. 3. Marks J, Rogers S, Chadkirk B, Shuster S. Clearance of chronic plaque psoriasis by anthralin: subjective and objective assessment and comparison with photochemotherapy. Br J Dermatol. 1981:105:96-99. 4. Serup J, Blichmann C. Epidermal hydration of psoriasis plaques and the relation to scaling. Acta Derm Venereol. 1987;67:357-366. 5. Elias PM. Lipids and the epidermal permeability barrier. Arch Dermatol Res. 1981;270:95-117. 6. Elias PM. Epidermal lipids, barrier function, and desquamation. J lnvest Dermatol. 1983;80:44s-49s. 7. Elias PM. Dynamics of the epidermal barrier: new implications for percutaneous drug delivery, topical therapeutics and disease pathogenesis. Prog Dermato1 1992;26:1-8. 8. Gray GM, Yardley HJ. Different populations of pig epidermal cells: isolation and lipid composition. J Lipid Res. 1975;16:441-447. 9. Grubauer G, Elias PM, Feingold KR. Transepidermal water loss: the signal for recovery of barrier structure and function. J Lipid Res. 1989;30:323-333. 10. Gray GM. White RJ. Glycosphingolipids and ceramides in human and pig epidermis. J lnvest Dermatol. 1978;70:336-341. 11. Wertz P, Downing D. Acylglucosylceramides of pig epidermis: structure determination. J Lipid Res. 1983;24:753-758. 12. Wertz PW, Downing DT. Glycolipids in mammalian epidermis: structure and function in the water barrier. Science. 1982;217:1261-1262.

24. Holleran W, Feingold K, Man MQ. Gao W. Lee J, Elias PM. Reguiation of epidermal sphingolipid synthesis by permeability barrier function. J Lipid Res. 1991;32:1151-1158. 25. Bowser PA, Nugteren DH, White RJ, Houtsmuller UMT, Prottey C. Identification, isolation and characterization of epidermal lipids containing linoleic acid. Biochim Biophys Acta. 1985;834:419-428. 26. Elias PM, Menon GK, Grayson S, Brown BE, Rehfeld SJ. Avian sebokeratocytes and marine mammal lipokeratinocytes: structural, lipid biochemical and functional considerations. Am J Anat. 1987;180:161-177. 27. Menon GK, Feingold KR, Moser AH, Brown BE, Elias PM. De novo sterologenesis in the skin, II: regulation by cutaneous barrier requirements. J Lipid Res. 1985;26:418-427. 28. Grubauer G, Feingold KR. Elias PM. Relationship of epidermal lipogenesis to cutaneous barrier function. J Lipid Res. 1987;28:746-752. 29. Feingold KR, Man MQ, Menon GK, Cho SS, Brown BE, Elias P. Cholesterol synthesis is required for cutaneous barrier function in mice. J Clin Invest. 1990; 8611738-1745. 30. Feingold KR, Brown BE, Lear SR, Moser AH, Elias PM. Effect of essential fa& acid deficiency on cutaneous sterol synthesis. J lnvest Dermatol. 1986;87:588-591. 31. Feingold KR, Man MQ, Proksch E. Menon GK, Brown BE, Elias PM. The lovastatintreated rodent: a new model of barrier disruption and epidermal hyperplasia. J lnvest Dermatol. 1991 ;96:201-209. 32. Proksch E, Feingold KR, Elias PM. Barrier function regulates epidermal DNA synthesis. Clin Res. 1990;38:638A. Abstract. 33. Cooper MF, McGrath H, Shuster S. Epidermal lipid metabolism in psoriasis and lichen simplex. Br J Dermatol. 1976;94:369-378. 34. Wertz P, Madison K, Downing D. Covalently bound lipids of human stratum corneum. J lnvest Dermatol. 1989;92:109-111. 35. Holleran WM, Man MQ, Wen NG, Gopinathan KM, Elias PM, Feingold KR. Sphingolipids are required for mammalian epidermal barrier function. J Clin Invest. 1991;88:1338-1345. 36. Baer AN, C o s t e l l o PB, Green FA. Free a n d e s t e r i f i e d 1 3 ( R , S ) hydroxyoctadecadienoic acids: principal oxygenase products in psoriatic skin scales. J Lipid Res. 1990;31:125-130. 37. Wertz PW, Cho ES, Downing DT. Effect of essential fatty acid deficiency on the epidermal sphingolipids of the rat. Biochim Biophys Acta. 1983;753:350-355. 38. Wertz PW, Swartzendruber DC, Abraham W, Madison KC, Downing DT. Essential fatiy acids and epidermal integrity. Arch Dermatol. 1987;123:1381-1384. 39. Wertz PW, Miethke MC, Long SA. Strauss JS. Downing DT. The composition of the ceramides from human stratum corneum and from comedones. J lnvest Dermatol. 1985;84:410-412.

ARCH DERMATOUVOL 130, APR 1994 456