Stratum Corneum Structure and Function Correlates with Phenotype in Psoriasis

Ruby Ghadiall y, Jeffrey T. Reed, and Peter M. Elias Dermatology Service, Veterans Administration Medica l Center, San Francisco, and Department of Dermatology, University of Ca lifomia School of Medicine, San Francisco, Cali forn ia, U.S.A.

Psoriatic epidermis demonstrates a defective pro cellular domains largely devoid of lalnellae. In con gram of growth and differentiation, including an trast, patients with chronic plaque psoriasis and abnormal permeability barrier. Despite the fact that sebopsoriasis displayed a lesser increase in transepi damage to the epidermis often initiates the disease, dermal water loss, normal numbers of lamelIar bodies psoriasis is commonly viewed as triggered by aber with only a few retained organelles, and abundant rant immune phenomena in deeper skin layers. Per extracellular lamellar material (although a normal meability barrier homeostasis requires the formation unit bilayer pattern did not form). Thus, both func and secretion of lamellar body contents, as well as the tionally and structurally, permeability barrier ho extracellular processing of lamellar body contents meostasis was more disrupted in erythrodermic and into lamellar bilayers. To address the hypothesis that active plaque psoriasis than in chronic plaque psori psoriasis is triggered by exogenous rather than inter asis and sebopsoriasis; i.e., the extent of defective nal factors, we assessed permeability barrier func barrier function correlated with abnorlnalities in the tion, lamellar body structure, and extracellular la known mechanisms of barrier repair, including la mellar bilayer formation in untreated patients with mellar body production and extracellular bilayer for different psoriatic phenotypes. Subjects with erythro mation. These findings are consistent with the hy derma and active plaque phenotypes displayed ele pothesis that both the initial appearance of psoriasis vated trans epidermal water loss levels, increased (Koebner phenomenon) and changes in disease phe numbers of epidermal lamellar bodies (many of notype are driven by alterations in barrier function. which failed to be secreted); i.e., corneocytes dis Key lVol,ds: bal'riel' f"llctioll/"ltmstl'llct"l'e/lamellal' bodies. played retained cytosolic lamellar bodies, and extra- ] Invest Derlllatol 107:558-564, 1996

n.volved psoriatic skin reportedlY.disP.lays de fective perme the stratum granulosum (S G)-SC interface (Elias and Menon, ability barrier function with transepidermal water loss 1991). Altho ugh little is known about permeability barrier ho (TEWL) levels up to 20 times normal (Felsher and Roth m eostasis in humans, in rodents barrier abrogation is followed by man, 1945; Grice and Bettley, 1967; Frost and Weinstein, (1) rapid secretion of performed epidermal LB, (2) replenishment of 1968; Grice el aI, 1973; Tagami and Yoshikuni, 1985) . the cytosol with n ewly fo rmed LB with further secretion of LB IC urr ent dogm a suggests that the epidermal changes in psonasls contents, and (3) formation of lam ell ar bilayers in the SC inter result from primary trigger factors released by abnormal circulating stices, leading to barri er recovery (Menon el aI, 1992a). Epidermal T cell s (Bos, 1988; Griffiths and Voorhees, 1992) and possibly other lipid syn thesis is required to sustain this process (Menon el aI, 1985; cell types in deeper skin layers (Sa iag et aI, 1985; Schubert and Grubauer el aI, 1989), and epidermal DNA synthesis also m ay C hristophers, 1985). Although this view could explain the rela contribute to barrier homeostasis (Proksch ci aI, 1991). tively uncomm o n guttate Aares that fo ll ow upper respiratory Because a primary barrier insult could be the m echanism that illnesses, and perhaps psoriatic erythroderm a, it igno res the impor underlies the Koehner phenom eno n , we have re- evaluated the tant role o f primary insults to the epidermis (Koebner phenome relatio nship betw een epidermal barrier function and epidermal non) as the trigger of m ost typical psoriatic lesio ns (Eyre and structure leading to lamell ar bi layer fo rmatio n in psoriasis. Prior Krueger, 1982). attempts to characterize the structural basis of psoriasis provide few Epidermal permeabili ty barrier integrity requires the organization insights about this relationship: both increased (Mottaz and Zelick of stratum corneum (SC) Li pids into extracellular lam ellar bilayers son, 1975; Lupulescu el aI, 1979) and normal (Bonneville el ai, fo ll owing the secretion of epidermal lam ell ar body (LB) contents at 1968) numbers of LB have been reported, all reportedly with normal internal structure (Mottaz and Zeli ckson , 1975; Bonneville e/ aI, 1968). Unfortunately, prio r studies did not correlate differ M'1I1uscript received February 12, 1996; revised April '12, 1996; accepted ences in LB m orphology with lesion phenotype. M oreover, they for publication May 21, 1996. did not correlate barrier function of the psoriatic lesions with the Rcprint requests to: Ruby Ghadially, M.D., Dcrmatology Serv ice (190), V.A. Medi cal Center, 4150 Clement Street, San Fran cisco, CA 94121. morphology of the LB secretory system . Finally, because of the lack Abbreviations: TEWL, transcpidermal water loss; EFAD, essential fatty of suitable post-fIxation m ethods, prior studies have not addressed acid deficiency; [,(3. lame ll ar body; SC, stratum corneum; SG, stratum the in tegrity of the SC extracellular lamellar bilayers in psoriasis. gral1ui oslIIn. Using ruthenium tetroxide (RuO.) post-fixation, it is now possible

558 VOL. 107, NO. 4 OCTOBER 1996 EPIDERMAL OAR.R.1ER IN PSORIAS IS 559

to obtain reproducible, accurate, ultrastructural images of the SC Table I. The Abnormality in Barrier Function in interstices (Hou et ai, 1991). In this study, we correlated perme Psoriasis Correlated with the Sev erity of the Psoriatic ability barrier status with ultrastructure of the LB secretory system Phenotype" and SC extracellular bilayers, in relation to disease p henotype. Our results suggest that disease p henotype correlates w ith abnorm ali ties Involved Skin Uninvolved Skin TEWL (g/ m2/ h) Significance TEWL (g/ m2/h) in LB structure, exocytosis, and the extracellular processing of secreted LB contents. These structural abnormalities, in turn, Erythroderma (n = 3) 36.4 ± 2.26 p ';; 0.001 3.5 ± 1. 0 correlate with the exten t of the barrier abnormali ty in psoriasis. Our P ,;; 0.005 results support the hypoth esis that primary barrier insults trigger the Active Plaque (n = 8) 16.1 ± 0.97 p :=; 0.001 3.9 ± 0.4 acute psoriatic phenotype. p ,;; 0.005 Chronic Plaque (n = 12) 9.0 ± 1.9 p < 0.05 4.1 ± 0.5 MATERIALS AND METH ODS Normal (n = 8) 4.8 ± 0.4 Sources of Tissue After obtaining informed consent, skin biopsies were (/ TE\VL was measured in patients w ith diffe rent psoriatic phenotypes . Measure obtained from the involved and uninvolved, adjacent sacral skin of 17 mentS were taken fro lll the ;lflcc tcd area and ;m adjacent area of uninvolved skin. T he patients with psoriasis (untreated for at least 2 wk), as well as 8 normal sex most severe phenotype (erythroderm a) displays the highest TEWL, approximiltely '10 and age-matched controls (homologous skin sites). T he psoria tics included times the uninvolved skill. R esults are presented ti S mean :t: SEM. 1 patient with sebopsariasis (involved sample fro m scalp rather than sacrum); i.e., psoriasiform scalin g of the scalp with minimal ski n involve ment elsewhere; 8 patients with chronic plaque psoria sis (l esions stationary for over 6 months); 3 patients 'with active, inAalll111a tory plaque psori asis; Uninvolved Epidertnis in Psoriasis Displays Normal Archi and 5 wi th acute, generalized erythroderma. Of these patients, 7 had tecture Post-fixation with standard osmium tetroxide fixation functional studies performed, as weB (sec below). provides excel.l ent preservation of LB structure. As in previo us studies (Ghadially et ai, 1992), cross-sectional images ofLB in both F unctional Studies TEWL meas urcments were performed 0 11 Icsions fram a separate cohort of untreated psoriatic subjects using a Servomed normal and uninvo lved psoriatic epidermis dem onstrated a trilam evaparimeter (Pinnagada " I nl, 1990). Meas urements were made on psori inar limiting m embrane and internal lam ellar disc-like structures aric lesions of 3 patients with acute erythroderma, 8 patients with active consisting of prominent dense lamellae separated by an electron inAa mmatory plaque psoriasis, and 12 patients with chroni c plaq ue psorias is. lucent band and divided cen trally by a mino r, striated electron TEWL al so was measured on contralateral or adjacent uninvolved skin in dense band (not shown; see Ghadiall y e/ ai, 1992). In no ne of the each of the patients. Biopsies also were obtaincd from 7 of these patiell ts (2 control samples examined did we find retained LB structures in with acute erytlu'oderma, 1 with active plaq ue psoriasis, and 4 with chronic com eocytes. Lamel.l ar bilayers w ithin the intercellular spaces dis plaque psoriasis). T he data Ii-om these patients were pooled witl, ti, e data p layed the characteristic unfurled pattern in the lower one o r two from those for whom functional or morphologic studies donc were 'lVail able, because the results were sil11jiar. T hree 111 caSUrCITICn ts ,vcre taken fro nl SC layers w ith m ature lamell ar bilayer basic unit structures filJing each site and the mean value then determined. Data from cach group were the interstices of the mid- to o uter SC (Fig lA, inset). T hus, these analyzed and the mean ± SEM determined. Statistical differcnces were studies show that th e uninvolved epidermis in psoriasis displays calculated usin g the Studcnt's tvvo-tail ed t tes t. normal ultrastructure (see also Bonneville e/ ai, 1968; Mottaz and Zeli ckson, 1975; Lupulescu e/ ai, 1979). Ultrastructural Methods Biopsy samples were minced to ';; 1 mm' pieces; fixed ovcrnight at 4°C in 2% glu taraldehyde, 2% pa.raformaldchydc Erythrodermic Psoriasis Displays Severe Ultrastructural with 0.06% calcium chl oride in 0.1 M sodium cacodylate buffer, pH 7.3; and Abnormalities (Figs I, 2; Tables II, III) T he number of LB placed in 0.1 M sodium cacadylate buffer prior to further processing. within the cytosol of granular cel.l s appeared to be increased Portions of each tissue sample were placcd in 0.2% RuO. (Polyscicnccs, dramatically in erythrodermic lesions (Fig IB,q, resembling the W arrington, PAl with 0.5% potassium fe rrocyanide in 0. 1 M sodium numbers detected 3- 6 h after acetone- w iping to removc the cacodylate, pH 7.4, at room temperature in the dark for 0.5 h (Hou el nl, epidermal p ermeabi.lity barrier (Menon e/ ai, 1992a), as well as in 1991), whereas other portions were post-fixed in 1% osmium tetroxidc (OsO.) with potassium fcrrocyanide in 0.1 M sodium cacodylate at room essential f,1tty acid deficient (EFAD) epidermis (Eli as and Brown, temperature in the dark far 1 h. After a rinsc in buffcr, all tissuc samples 1978). Yet both the dimensions and internal stru cture ofindividual were dehydrated in a graded ethanol seri es and subsequently embedded in LB appeared no rmal (Table III and Fig IB, C; quantitative data a low- viscosity epoxy resin . T hin sections were eX:lllljne d under a Zeiss lOA not shown). electron microscope operating at 60 kV after sta ining with lea d citrate U ltrastructural examinatio n of th e SC in erythrodermic psoriasis and/ o r uranyl acetate. A blinded observer pcrformed all mcasurements of revealed the presence of numerous intracellular lipid droplets (Fig membrane and organ ell e dimcnsions from codcd mi crographs. l A ). Moreover, retained lamell ar structures also were visible frequently within the cytosol of corneocytes (Fig lA,B, o)Jm RESULTS a/1"Ol/Is). Furtherm ore, even with Ru04 post-fixation, the intercel Barrier Function Is Compromised to an Extent Propor lular spaces appeared stdkingly devoid oflam ell ar bilayers (Fig lA, tional to the Severity of the Psoriatic Phenotype In order to solid arrol/ls). Finally, desmosomes appeared to be increased in compare the extent of the barrier abnormali ty in the different number and to persis t to higher levels above the stratum granulo psoriatic phenotypes, we performed TEWL m easurem ents on sum-stratum corneum interface than in normal skin (i.e., > 6 layers) different types of p soriatic patients (Table I). TEWL levels of (Fig tAl· T hese studies reveal a dramatic increase in the number of psoriatic plaques in erythrodermic subjects were approximately 10 LB in the most severe psoriatic phenotypes, w ith a strik ing reduc times rugher than those of uninvolved skin (p < 0.001) and were tio n in extracellular bilayers, apparently due to retentio n oflam ell ar also significantly higher than both active and chronic plaque lesions bodies intracellularly in corneocytes. (Table I). Similarly, TEWL levels of active plaque psoriasis were In Chronic Psoriatic Phenotypes a Lesser Disturbance of increased about 4- fo ld over uninvolved skin (p :=; 0.005) and were LB and Intercellula r Membrane Structure Occurs (Fig 2; significantly hig her than chro nic plaque lesio ns, as well (Table I). Tables II, III) Although TEWL levels also were more than do ubled in chro nic plaque psoriasis , the difference compared to uninvolved skin was Ch ollic Plaq ll e Psoriasis In contrast to erythrodermic psot;asis, in not as great (2- to 3- fo ld; p :=; 0.02). Finally, there was no stable plaque lesio ns, both the number and size of the LB appeared statisticall y significant difFerence between the uninvolved skin of relatively no rmal (Table III; microscopic data not shown) (mea psoriatic patients and normal skin of nonpsoria tic patients (Table surem ents of 120 LB in psoriatics and 60 LB o f normal controls I). These studies show that barrier perturbatio n is more extensive in showed a m ean LB diameter of 172.4 ± 7.2 nm (mean ± SEM) in the more severe psoriatic phenotypes (erythroderma and active psoriatics and 177.2 ± 6.0 in no rmal controls; not significant) . plaque psodasis) than in chro nic plaque pso ri asis. Moreover, lesions of chro nic plaque psoriasis displayed fewer 560 GHADIALLY liT ilL THE JOURNAL OF I NVESTIGATI VE DEI~ATOLO GY

Figurc 1. Erythrodcrmic psoriasis di.plays cxtcnsivc abnormalitics ill thc lamcllar body sccrctory systcm. (AJ Stratum corneum in erythrodennic pSOl-ia sis. Note the virtual absence of intercellul ar lamell ae (1I4, ile nr"""," cn dsJ- III addition , numerous lipid droplets (LD) and retained lamellar structures (OpCII nrro ll'''ends) arc present. illSel shows normal intercellular bilayer structure, displayi ng the basic unit pattern (1I,"ile nrro ll'''end). 8 and C) Stratum grallulosum in erythrodeJ'll1 ic psori asis. T he number oflamellar bodies w ithin the cytosol appeared to be dramaticall y increased, resembling the appearance of mice 3-6 h after acetone wipi ng to remove the epiderm al permeabili ty barri er (15)_ Lamellar bodies were of llo rmal size and internal structure (8, C, -r Senle I",rs: (A) 20 f-tm; (ill Sci) 22.5 f-tm; (8) 17_ 5 f-t 111 ; (C) 22_5 f-ttl1 . VOL. 107. NO.4 OCTOBER 1996 EPIDERMAL BARRIER IN PSORIASIS 561

Figure 2. Chronic plaque psoriasis displays less extensive abnorlnalities in the lamellar body secretory system. (A and B) Stratum corneum in chronic plaque psoriasis . A paucity of intercellular lamellae are observed throughout the stratum corneum interstices (ml'n'''s). T he m embrane structures present retain the unfurled pattern of the lower stratum corne um, and the mature pattern of bilaycrs usuall y o bscrved in thc upper str:!tum corneum is not ev;dent (c.f. Fig lA, inset). Lipid droplets and rctained Iamc ll ar structures still occur (OpCII nrro ",/r cnd), but less fi'cqucntly than in erytllrodcrmic psoriasis. (C) The stratU111 cornClIIll in scbopsoriasis. T he intercellular spa ces contain 111 0 rc bilaycrs th an in crythrodcnnic psoriasis (mnl l/Js) . but even in the upper stratlllll corneum, the membranes do not reveal a basic lamellar unit pattern. (D) The stratum granulosum in erythrodermic psori as is. Multivesicular bodies arc detected in aU forms of psoriasis (Tables II, III). but only rarely observed in normal s!ci n . In erythrodermic psoriasis. m any lTIul tivesicular bodies often can be o b served in a single section through o ne stratum granulosum cell. Senlc bnrs: (A, 0) 20 J.LlTI; (B, C) 25 J.Lm. 562 GHADIA LL Y ET A I- T H E JOURNAL OF IN V ESTIG ATI VE DEltMATOLOGY

Table II. Increased Lamellar Body Numbers Suggest an Attempt to Repair the Barrier Abnormality, Which Is Most Prominent itl Erythrodermic Patients" Erythroderm a Active Plaque Chronic PI"quc Sebopsoriasis

Case #: 2 3 2 2 5 6 7 8 Lamell ar Body # it it i ti ti ti ti i N N N N i N N N N (it. j,N) Lamell ar Body N N N N N N N N N N N N N N N N N Structure Size N N N N N N N N N N N N N N N N N

,/ Com pendium of ultrastructural studies of rite lalllcll:u body secretory system in diffe re n t psoriatic phenotypes. All asseSSl1I ents and m C:l su rC IllCllts were m ade ill ruthenium tetroxide post-fixated tiss li es. T he most severe phenotype (erythroderma) displays non-n il l lamcll nr bodies ill gr cOl tl y incrcascdllumbcr. suggestin g a ll attelll p t by the epidermis to repair the pcrmc:lbility barrier. Abbreviations: it. greatl y increased; i , moderatel y in creased; N, normal. retained LB w ithin the corneocyte cytosol than were presen t in OCCUlTed in active plaque lesions, whereas chl·onic lesions of acute eryth rodermic psoriasis (Fig 2A,B, opell arrows). Further psoriasis displayed TEWL levels intermediate between these and mo re, the numbers of extracellular lamell ar bilayers were corre uninvolved skin. Moreove r, epidernl'll m orphology in worsening spondingly greater than in erythrodermic lesions (Fig 2A-C IJcr SIIS active plaque psoria sis was comparable to that in acute erythro lA). Mature lamell ar basic unit structures, as observed in the derma. In contrast, epidermal structure in the stable erythrodem1a mid-to-outer stratum corneum of both normal and uninvolved patients closely resembled that in active plaque psoriasis. These psori ati c epidermis, were found and were of normal dimensio ns results suggest that barrier repair m echanisms are operative in (in tcrl amellar dimensions: normal = 13.2 nm; psorias is = 12.8 nm psoriasis, and they appear to be in part successful at normali zing (not significant), but the lamell ae often maintained the unfurled, barrier functi on, pe rhaps leading to more chronic p SOl·iatic pheno elongated patte rn characteristic of secreted lamell ar body contents types. T he presence of e ffe ctive barrier repair mechanisms in in the lower stratum corneum (Table II; microscopic data not clinica lJ y involved skin of psoriasi s would be diffi cul t to study; shown; see Mcnon ef ai , 1992a). T hese studies show that in the however, in clinically uninvolved skin, recovery rates w ere identi more chronic psori atic phenotype, a lesser disturbance of LB and cal to no rmal con tro l skin (den Arend and van de Kerkhof, 1987). in tercellular m embrane structure occurs. Likewise, with artificial barrier restoration by occlusion alone, psori,nic lesions usually regress (Baxter and Stoughton , 1970; S c/)()psoriasis T he sin gle biopsy from a patient w ith sebopsorias is Shore, 1985; Friedman, 1987; Gri ffit hs cf ai, 1995). displayed changes comparable to the subjects with chronic plaque Acute perturbations of the permeability barrie r, e.g., solvenr psori asis, although to a lesser extent (Tables II, III; Fig 2e). applications or tape stripping, stimulate a sequence of homeostatic Active Plaque Psoriasis Displays an Interm.ediate Ultra mechanisms, (Meno n el ai , 19923). Because the en tire sequence. structure T hree of the fo ur patients with actively evolving including (1) rapid secretion o f pre-formed LB contents, (2) plaques of psorias is exhibited changes siqlilar to erythrodermic generation o f nascent LB, (3) accelerated in tercellular deposition of psoriatic. subjects, whereas one patient exhibited changes compa LB conten ts, (4) extracellular processing of lamell ar body contents rable to chronic pl aque psori asis (Tables II, III; microscopic data by co-locali zed hydrolyti c enzym es into lam ell ar basic unit struc not shown). tures (Holleran el ai, 1993), is inhibited w hen acutely perturbed skin is occluded w ith a vapor-impermeable wrap (G rubauer el ai, 1989), DISC USSION tl1ese changes all appear to be linked to barrier function (Feingold, Barrier function is abnormal in scali.ng dermatoses including psori 1991). Likewise, in the EFAD m o use (a chronic, hyperproliferative as is , in which TEWL levels are increased 1- to 20-fold (Felsher and model often used as an analog fo r psoriasis), increased numbers of R o thman, 1945; Grice and Bettl ey, 1967; Frost and Weinstein, defective LB, decreased extracell ul ar lamell ar bilayers (Eli as and 1968; Grice el ai , 1973; T agami and Yoshikuni, 1985; M otta ef ai, Brown, 1978; Hou el ai, 1991), and increased lipid synthesis 1994). Y et few prior studies have correlated TEWL with lesion (Feingold cl ai, 1986) occur, abnormalities that again are reversed phenotype. Whereas Grice and Bettley (1967) fOl1J1d no significant by occlusio n (Feingold el ai, 1986; Feingold, 1991). Fi nally, in both functional differences between erythrodermic and plaque psoriasis, the acute and chronic murine models, barrier perturbation is TEWL levels d ecreased as the disease became less active, consistent associa ted w ith increased generation of a family of cytokines (32). with o ur data that barrier fimction correlates better with di sease potential mediators of portions of the homeostatic response de activity than w ith lesion phenotype; i.e., the highest TEWL levels scribed above. R ecently, we and others have proposed that epider-

Table III. SC Structural Abnormalities, Corresponding to Decreased Barrier Function, Are Most Prominent in Erythrodermic Psoriasis" Erythroderma Active Plaquc Chronic Plaque Sebopsoriasi

CASE #: 2 3 2 2 3 4 () 7 8

R.ctaincd Lamcllar l3 0di es +++ + NO ++ ++ + +++ + ++ + + NO + + NO + + + (- .+,+ +,+ ++) # of In tercellu lar Lam ella e t t NO t R ! ! ! ! NO N I! ! NO N I! N (! ,N, 1') Pattern of R MUI' ND R R I ~ I ~ R MUP MUP NO MUP MUP NO MUP N MUP Intercellular La mell ae

" Compendiu11l o f uitrastru cillral studies of ril e st"r protein (Wood et ai, 1992, 1994; Lli pul escll AP. C hadwick JM. Do wnhalll TF: U ltrastru ctural and cell su rf.1 cc changes of human psori:ILic skill fo ll owing Goeckerl'l1 ;"C n therapy . .I C I/(flIH'lHIS Pmlwl Tsai et nl, 1994). Finall y, N ickoloff and Naidu (1994) showed that 6:347-363. 1979 epidermal production of cytokines fo llowing barri er abrogation Mell o n GK, Feingold KR.. , Elia s PM: L:l1 11 e ll ar body secretory response ro barrier precedes m ovem en t of inA all1matory cell s fi'om the circul atio n in to disruption_ J hlll('sl /)erlllnt,,1 98:279-289, 1992a 564 GH ADIAL LY ET IlL T HE J O UR N AL OF IN VESTIGAT IVE DERMATOLOGY

Menon GK, Feingold KR., Moser AH. BrowlI DO. Eli as PM: De IWIIO stcro logcll csis in th e Saiag p. Coulo mb B, Lebreton C, Be ll E. DlIbc rtre t L: Psoria tic fibro blas ts induce skin . II. Regulation by cuta ncous ba rri er requirements.) Upid g "s 26 :4 18- 427, 1985 hyperpro life ra tio l1 of no rl11 ;d ke ra tin ocytes in a skin equivalcnt mo del ;tt llitro. Menon CK. W illinrn s ML. Glwdi::dl y R . Eli as PM : Ltl l1l c ll ar bodies as delivery systems SciCll ce 230:669-672. 1985 of hydrolytic enzymes: implica ti ons for lIon nal cohesion and abnonnni desqua Schubert C, Christophers E: Mas t cells :1I1d macropha ges in earl y relapsin g psoriasis. mation. Br) 0"1"111 " /11 /1 26:337- 345. 1992b A rc/, 01'1'111 ,,101 Res 277:352- 358. 1985 Mon a S. Mo nti M. SCS:lll i1 S. Mcll csi L. C hido ni R . Capu to R : Abno nnniity o f wate r Sho re R..M: C learin g of psoria tic lesio ns afte r th e applicatio n of tnpe . N r::IIS' J iVIed barri e r fun ctio n in psori as is: ro le of ccramidc frac ti o ns. Arch Dcrlll llto l '130:452- 312:246, 1985 456. 1994 Sundberg JP. Dunstan R W. Roop D R . l3 en mer WG: Full-thickncss skin g rnfts from Mot[:IZ JH . Z clicksOIl AS: KcratinoSOIlH!S in psoriatic skin . D er",fl tcweller (S/orkb) fl aky skin m ice to nude mice: mailltcnance o f the psoriasifonn phcnotype.) JUl/esl 55:8 1-85. 1975 Ocrtll"IfI/I 02 :78 1-788 . 1994 Nicko lo ffUJ, Naidu Y: Perturbation o f cpidcnllil l hanier functioll correlates w ith ini tia ti o n Tag:llni 1-1 . Yoshiklllli K: Inccrre latio nship between w atcr h;;1 rri e r and reservoir of cyto killC c;\scnd c in IlUlllall skin . J Jl III Arm/ Dall/fllo/ 30:4 :535-546. ·1994 Pinllagodn H , T upkcr RA, Agner T, Scru p J: Guide Lin es for rrn llsepiciermal w ater loss fun ctio ns o f patho logic stratum corn cum. Arc" Derlll tlwi 12 1:6 42-645. 1985 (TEWL) measurement. COII/l/el Oel'l" ",ilis 22: I 64 -1 78. 1990 T,ai J-C , Fe ingold KR . C rumrine D. W ood LC . Grunfe ld C, Eli as PM : Permeabilit\' Pri cstl ey GC. Lo rd R: Pibroblast keratinocycc in terncri o ll s in psorinsis: failure of ba rri er disruptio n alters the IOc:llizatio n il nd expression of T NF-nlpha protein ~ psori ati c fi bro bl nsts to stirllul ate keratin ocytc pro lifcrn ri on ;/1 I/i/r" , HI" J D Cflll(lW/ th e epidermis. A rc" Oel"l" n"" 2H6:24 2-248, 1994 123:467-472. 1990 W ood LC, Fein gold KR. Sequeira- Martin SM. Elias PM, Grunfeld C: Barrier function Pro ksch E. Fein gold K It . Mao-Qiang M. Elias PM : Barricr fU ll ctio n regulates coordinatel y regulates epid e r1l1 al lL-1 nn d lL- II"tA I1lI ~A Icvels. £xp Denuaeol epidermal DNA synthesis.) CIi" III lies' 87: 1668-1673. 199'1 3:56-60, 1994