~n.t:ioxidant Enzyntes in Psoriatic Fibroblasts and Erythrocytes

patrice Therond,* Pascale Gerbaud,t Stephanie Dimon,i" Wayne B. Anderson,t Daniele Evain-Brion,i" and franc;oise Raynaudt ~Service d e Biochimie. H8 pital Bice tre , 71 rue du Gelleral Leclerc, 94275 Le Kremlin Bicetre Cedcx, France; t lNSERM U 427 ,

peve10ppe m ellt HUl11ain, Croissance et Difle renciation. Faculte des Sciences Pharmaceutiques et Biologiques de Paris, Universite R ene pescartes, Paris V, 4 avenue de l'Observatoire, 75270 Paris Cedex 06, France; :j:Laboratory of Cellular Oncology, National Cancer

lnstit:ute, National Institutes of H ealth , Bethesda, M aryland 20892, U.S .A .

~----------------------------------------------------------------------------------------------------------------------------

~t:ioxidant enzyme activities in fibroblasts and eryt:hrocytes prepared from normal and psoriatic pat:ient:s were Ineasured and compared. The most significant differences were noted in superoxide disrnut:ase (SOD) activities. A dramatic (5.2-fold) in­crease in Mn-SOD activity along with a lesser (1.8-fold) in.crease in CuZn-SOD activity was observed in fibroblasts from lesional and nonlesional psoriatic skin.. The increase ofMn- SOD activity was correlated wit:h an increase of both protein and mRNA. A slight (1 .2-fold) increase in CuZn-SOD activity was also found in psoriatic as compared to normal red blood

. cells, "",hile Mn-SOD activity was not present in these cells. In contrast, both glutathione peroxidase and

Reactive o""ygen species such as the superoxide radical (O~-), hydrogen peroxide (H2 0 2 ) , and hydrm .... yl radical (OH,) have been linked to skin cancers, cutaneous aging, and m~ny inflammatory dJsorders (Cross e/ ai, 1987; Nlwa et ai, 1987).

Reactive oxygen species are generated by fibroblasts in response to stilnulation by various hormones and cytokines (Meier e/ ai, 1989) and have been reported to act as second m essengers (Schreck and Bae u erle, 1991) . In turn, the oxidative m odification of several key proteins lllcluding pro tein kinases involved in cellular s i~nal trans­duction has been described (Whisler et ai, 1995). PrevIously, we have established that cAMP-dependent protein kinase activiry as

I well as 8-azido-eZPlcycl.ic AMP binding to the RI and RII regulatory subunits are decreased in cell s from psoriatic patients compared to cells fi:olll normal subjects (Evain-Brion et ai, 1986; Rayna ud c/ ai, 1989) . The abili ry of retinoic ac id treatment of intact cells to rapidly reverse the oxidatively m odified state of protein kinase C (Gundimala et ai, 1993) as well as to reverse the altered

I state of protein kinase A noted in psoriatic cells (Raynaud : / ~/ , 1987, 1993), coupled with the inflammatory evolution of psonaSlS,

MaI;1uscripr received April 19, 1995; revised February 27,1996; accepted for publication March 5. 1996.

Reprint requcs ts to: Or. Franl'oisc R aynalld . IN SERM U 427, Facul te des Sciences Phannacelltiqucs, et Biologiques de Paris, 4 Avenuc de I'O b servatoire, 75270 Paris. Ccdex 06.

A bbreviations: SOD, superoxide d ismutasc; EDTA, ethylenediamine [etraacetic acid; INT. 2-(4-iodophenyl)-3-( 4-nitrophenol)-5-phcnyltetrazo­lium c hl o l;de; HPLC. high-pcrfo rl113ncc liq uid ch.romatography; GSH . reduced form of glutathionc; GPX. glutathione peroxidase; IPS, Invo lved psoriatic Skin; UIPS, Uninvo lved Psoriatic Skin.

catalase activities were only slightly (1.3-fold) in­creased in psoriatic fibroblasts, with no appreciable change noted in psoriatic erythrocytes. Likewise, glutathione levels were observed to be similar in normal and psoriatic cells. The increases in SOD activities did not appear to correlate with the severity of the disease as expressed by the Psoriatic Area Severity Index score or with plasma inflammatory markers. These results demonstrate that antioxidant enzyme activities, particularly Mn-SOD in fibroblasts and CuZn-SOD in erythrocytes, are significantly el­evated in cells from psoriatic patients. Key words: psoriasislsJJperoxide dismJJtases/free l,adicalslfibroblastsl el1'throcytes. ] Invest Derfllatol 106:1325-1328, 1996

suggested the possibility that the re may be chan ges in the oxidative state of psoriatic cell s. The cellular oxidative sta tus is related to a balance between the production of o" .. ygen-derived free radicals and their destruction by enzym es that m etabolize oxygen-reduction products (antioxidant enzym es). T hus, w e carried out studies to de termine whether the activities of enzym es involved in modulat­ing the redox sta te of the cell, such as superoxide di sl11utases (SOD) , catalase, and glutathione peroxidase, might be altered in fibroblasts and erythrocytes from psoriatic patients. T he results indicate a selective increase in SOD activiry in these psoriatic cells .

MATELUALS AND METHODS

Cells Human fib roblasts were isolated from eight normal and e ight untreated adult pso riatic pati cn ts (Psoriasis vlligaris in Rare for app roximately 3 months) by enzymatic digestion of lesional and nonlesional buttock and abdomin al skin punch bio psies (4 mm) as prev io usly describcd (Raynaud ct

ai, 1995). T hese white subjects were age- and sex-matched. T llis study was approvcd by the etllical committee of the H 8pital Cochin (Paris). Cells were grown as previously described (Raynaud ef ai, 1995) and were uscd subconRuently (106 cells/dish) at the fifth passage . All data w ere obtained w ith cells propaga ted for the same number of passages (fiftll passage) under identical conditions (culture medium . serum . etc). For each donor tile diller"nt assays were performed on the same cell extracts prepared fro m cells arising from the sam e cell seeding. Erythrocytes were isolated fronl h ealthy subjects and from age- and sex-matched untreated patients as previously described (R aynaud et ai, 1993 ).

Preparation of Cellular Fractions Fibroblasts were washed three tim es with icc-cold phosphate-buffered salinc and harvested by scrap ing into ice-co ld buffer [0.25 M sucrose. 20 111M T ris (PH 7.4), 1 mM MgCI21 w ith a cell scraper. The cell s were recovered by centrifugation at 1000 X g for 5 min and the cell pellet was then frozen at - 80 c C. For antioxidant enzyme activi ty measurements. the control and p soriatic ceI.l pellets were di srupted

0022-202X/96/S10.50 • Copyright © 1996 by T he Society for Investigative D ermatology, Inc.

1325

1326 THEROND fiT AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Table I. Modulation of CuZn-SOD, Glutathione Peroxidase, and Catalase Activities in Erythrocytes From Normal and Psoriatic Subjects"

Erythrocytes

N orma l

Psoriatic

C uZn-SOD (U/g Hb)

790 :!: 30.94 (508 - 1216) (n = 28)

983.77 :!: 29.59 (7 54 -1236) (n = 25)

Antioxidant activi ties

GPX (U/g I-Ib)

27 .49 :!: 1.28 (19.33- 52.33) (n = 28)

32.22 :!: 1.68 (18.5- 46.62) (n = 22)

Cata lase (U/mg protein)

9.2 1 :!: 1.26 (5.6-16.8) (n = 8)

8. 11 :!: 0.52 (6.5-1 0 .7) (n = 7)

" T he <ll1ciox:idanl enzymatic acrivirics were nss:l ycd as described under Mtlrc/ials tllU/ J\1cthods. C uZn-S OD and G PX ac ti vities arc expressed 'IS uni ts per IlliUigrarn of hen 'lOg lob in (I-Ib), and CilLaJase :H; tiv ity is given as uoi ts per milligr:ll11 of mcmb r:lI1c prote in . Data ::i rc expressed as the mean :!: SEM of triplicate dctcrlniJ13tions carried out with cells trom each donor w ith the range given be low in p:lfcnthcscs. 11 = number of indi vidual donors included in the study .

by sonication in 500 J.11 of 1 0 mM sodium phosphate burre r, pH 7 , and used in the same experiment. Erythrocyte hemolysates were prepared by sus­pending the cell s in 50 mM sodium phospha te burrer, pH 7.2, containing 1 mM ethylenediamine tetraacetic acid (EDTA), and then freez ing at - 80°C. H emoglo bin levels were d etermined for each sample by the method of Drabkin ;Ind Austin (1955).

Erythrocyte membranes (ghosts) were prepared by suspending erythro­cytes into 10 mM phosphate burrer, pH 7.2, containing 1 mM EDTA for 30 min on icc and then harvesti ng by centrifugation at 17,000 X g for 30 min . Ce ll membranes were then washed three times (hemoglobin free) in the same buffer prior to freezing at - 80°C.

SOD Activity X anthine- xanthine oxidase was uti lized to genera te a 0 '" 2 Au x and the red uctio n of 2-(4-iodophenyl)-3-(4-nitrophenol)-5-phen yltetrazo lium chlo ride (INT) to red formaza n by 0 '" 2 was fo llowed at 505 nm at 30°C . T he rate of INT reduction in the absence of samples was used as the reference rate (0.02 :!: 0 .005 absorbance/min) . The data were plotted as percentage inhibition of tNT red uction IIcrS IiS protein concentra­ri on. One unit of activity was defined as the amount o f protein necessary to reduce the rate ofI N T reduction to 50% of maximum inhibirion. Each assay tube con ta ined 50 mM 3-(cyclo hexylamino)- 'I-propa nesulfonic acid burrer pH 10.2, to determine C uZn-SOD activi ty (Bannister et ai, 1987) or 50 mM phosphate buffer, pH 7.8, to determine tota l SOD activity (O'Neill et ai, 1988), along w ith 1 111M EDTA, 25 J.1M INT, 50 J.1M xanthine, 1 U / ml cata lase, 0 .05 111M bathocuproine di sulfo nate disodium sa lt, 0.13 mg/ml bovine SCrU ITI a lbulninc, and sufficient xa nthine oxidase to achieve the required 100% level of non inhibitio n . Bathocuproinc disul fo na te disodium sa lt and boville se rum album ine we re added to inhibit nonenzym atic scavenging of 0 '" 2 (Spitz and OberI ey , 1989). Mn-SOD activity leve ls, m easured by adding cyanide (2 mM) to the assay mixture at pH 7.8 to inbibit any C uZn-SOD activity or determined by substractin g the CuZn­SOD activ ity determined at pH 10.2 from the tota l SOD activity measured at pl-l 7.8, were found to be simi lar (data not shown) . All data were expressed in uni ts of SOD acti vity per milligram of protein for fibroblasts or hemoglobin for erythrocytes.

Catalase Activity Cata lase activity was measured as described by J ohans­son and Borg (1988). Cata lase is able to decompose 1-1, 0 2 by two types of reaction. Both reacrions include a first step of formation o f an intermediate consisti ng of the enzyme and H 2 0 2 . In a second step the cata lase activity was measured by reacrion of this in termediate with a h ydrogen donor other than 1-120 2 , Methano l was used as the hydrogen donor, and we measured the productio n of forma lde hyde spectropho tometri call y with 4-amino-3-hydra zino- 5-mercapto-1,2,4-triazo le (Purpald Sigma, St. Quentin Fallavier, France) as a chromogen . Interferences due to oxidatio n of methanol by other enzyme activ iries were not obse rved for the present method .

Glutat hione Dete rlnination Cellular levels of the red uced form of glu tathione (GSH) we re determined by ion-pairing reversed-phase high­performance liquid chro ma tograph y (HPLC ) coupled to a Coulometri c electrochemica l detector. A 200-J.1l aliquot of the cell suspensions was mixed with 200 I.d of co ld 10°;', t ri chlo roacetic acid . T he solu tion was kept on icc for 15 min , and precipitated proteins were then removed by centrifugation . Fifty microli ters o f the supernatant was then ana lyzed by I-IPLC. A Spheriso rb C-18 reversed-phase ODS-2 column (5 J.1m; 250 X 4.6 mm) was used . Isoc ratic elutio n was performed employing the HPLC so lvent 10 111M NaH 2 PO., adjusted to p H 2. 7 w ith 85'1., H JPO. containjng 0.05 mM o f the ion-pairing reagen t octylsu lfilte and 2'1'0 aceto nitrile (v/v) . Separations were performed at room temperature w ith a Aow rate of 1.0 Il1 I/ mill. GS I-I was detected fo llowing iso lation by HPlC w ith a Coulochem detec tor equipped with a model 5010 dual analyti ca l ce ll and a model 5020 g uard cell. T he app li ed electrode poten tia ls fo r detector 1, detector 2, and guard cdl worki ng electrodes we re set at + 0.6, -I- 1.2, and + 1.4 vo lts, respectively. A standard GS I-] was prepared to produce a standard curve

within the bio logic concentra tion range. The standard curve was derived from dilutions o f known concentrations of gluta thio ne (1 - 20 J.1mol/ liter).

Glutathione Peroxidase (GPX) Activity GPX was assayed by au adaptation of the method of Pagli a and Valentine, 1967 with cumene hydroperoxide as substrate. Both sample and refere nce tubes contained 0.05 M phosphate buffer, pH 7.2, 4.3 mM EDTA, 0.28 mM NADPH, 0.5 U of glutathione reductase, 4 mM glutathione, and the appropriate amount of red blood cell hemolysates o r of fibroblasts. T he oxidation of NADPH b y cumene hydroperoxide (0.1 8 mM) added to the sample tube was followed spectrophotometrically at 340 nm at 30°C. An additional blank, containing all components except the sample, was used to co rrect for nonenzym atic oxidation of GSH and NADPH by cumene hydroperoxide . One unit of GPX was defined as 1 J.1M substrate con verted in 1 min .

Western Blot Analysis Cell pellets were prepared as described above. After measurem ent of protein levels, lysa tes were heated fo r 5 min at 95°C in laemmli sample buffer and ana lyzed by electrophoresis on 10% poly­acrylamide ge ls (50 J.1g/sample). Proteins were electrotransferred to nirro­cellulose and weste rn blots were performed as described in the ECl kit (Amersham , Les Ulis, Fnm ce). Proteins were detected with a monoclonal anribody raised agai nst human Mn-SOD, diluted 100-fold (Bender, Vienna. Austria).

Slot Blot Analysis of the Mn-SOD mRNA Tota l RNA was extracted fro m normal (N), lesional (I1>S, Involved Pso riati c Skin), and nonlesional (UIPS, Uninvolved Psoriatic Skin) cultured fibroblasts (Chomczynski and Sacchi, 1987). Decreasing amounts (20, 10, 5, 2 .5 J.1g) of total RNA were blotted onto a nylon sheet (Amersham, les Uli s, France) and were hybrid­ized with a 32P-labeled human Mn-SOD oligonucleotide antisense probe (5' -GAACTTCAGTGCAGGCTGA)(Frankenne et ai, 1992) and with GAPDH (glycerald hehyde-3-phosphate deh ydrogenase) as control. The specificity of the Mn-SOD oligonucleo tide pro be was assessed by northern blot anal ysis. Two transcripts of 4 .0 and 1.5 kb were de tected (data not shown).

Plasmatic Inflatnmatory Markers Orosomucoid and ceru loplasmin measurements were performed using a nephelom etric method with th e Behring N ephelometer Ana lyzer (BNA ; Behringwerke , AG, Marburg, Germany) as described by the manuf.,cturer.

Analysis of Statistical Significance Data were analyzed by Student's t

test method. Data fro m the les ional and non les ional skin of psoriatic patients were analyzed in a paired Student's t tes t.

RESULTS

Psoriatic Erythrocytes: Increased Antioxidant Enzymes T h e activities of the three major anti oxidant e n zym es were m ea­

Sllre d in red blood cells from normal and psoriatic subjects. In these cell s t h e only active SOD is the C uZn-SOD (Nakano ef ,,/,1990). As presente d in Table I a sig nificant increase (1.2-fold) (p < 0 .001) in CuZn-SOD ac tivity was o b served in psor iatic red blood cells as

compared to n ormal cells. N o corre lation was found betwee n C u Z n-SOD ac ti v ity and the age of th e normal and psoriatic donors. G lu tathio n e peroxidase activ ity was also som ewh a t increased (1.2-

fold) (p < 0.05), w h ereas n o sign ifi cant diffe r e n ces were observed in catalase activity b etween normal and psoriatic erythrocytes.

Psoriatic Fibroblasts: Antioxidant Enzymatic Activities and Glutathione Levels T h e results presente d in Table II show that

in normal fibroblasts antioxidant enzyme activities va.ri ed ove r a

VOL. 106. NO.6 J UNE 19% ANTIOXIDANT EN Z YMES IN PSOR IAS IS 1327

Table II. CuZn- and Mn-SOD Activities, GPX, and Catalase Activities in Cultured Fibroblasts Prepared Front Skin of Norntal Subjects and Front Lesional and Nonlesional Skill o f Psoriatic Patients"

Antioxidant activities

CuZn-SOD Mn-SOD GPX Catalase GSI-I Fibroblasts (mU/mg prot) (rnU/ mg prot) (mU/ mg prot) (U/ mg prot) (nmollmg prot)

Normal 210 = 38 57 .6 = 8.8 5.98 = 0.51 4.49 = 0,41 16.25 = 0.99 (88 - 367) (22-90) (3.8 - 8) (2.7- 5.9) (12.5- 21.3)

Psoriatic 387 = 38 297 = 42 9.0 = 1.1 6 6.42 = 1.11 14.08 = 1.20 N onlesional (275- 601 ) (149-503) (4 -1 3) (4.2-10.3) (12-17.5)

Psoriatic 372 = 25 243 = 21 7.63 = 0.84 6.20 = 0.46 14.30 = 1.15 Lesional (295-478) (139 - 345) (2.3- 9.8) (5 .2-7.8) (12.1-16)

a Prilnary fibrob last cultures prepared from eight !lonnal and e igh t' age- and sex- matched psori:ltic patients were assayed for the i1l1tioxid~l1t enzyma tic :lc tivitics indicated. as \ve ll as for reduced g iutarhjoll c (GSH) levels. as described under l\t/ tl fer ja/s " lid {v!cflwds. Data 'Ire e xpressed as meall :!: SEM of triplicate determinations .md as the range in parcnrhescs of trip lica te de te rminations in the cell extracts fro m eight fibrob last primary cultures.

wide range with the different donors. N o correlation was fo und between the variable activities and th e ages of the subjects (data not sho-wn) . In psoriatic subj ects no significant differences were ob­served fo r any of the antioxidant enzym e activities determined with fib roblasts prepared fro m lesional and nonlesional skin of the sam e patient. GPX ac tivity was not significantly different between pso­riatic fibrobl as ts cultured from lesional skin and norm al fi broblasts and "W'as found to be slightly increased (1.5- fold) (p < 0.05) in

• fibroblasts fi'om nonlesional skin w hen compared to norm al fibro­blasts. Catalase activity was also slightly enhanced (1.4- fo ld) (p < 0.05 ) i.n. psoriatic fibrob lasts from lesio nal and nonlesional skin as cOlnpare d to normal cells, w hil e glutathione levels were similar in nonnal and psoriatic ce]]s.

Interestingly, significantly eleva ted levels (1.8-fold) of the CuZn-SOD activi ty (p < 0.005 for nonlesional and p < 0.003 for

, lesional skin), and particularly of the Mn-SOD activ ity [p < 0.000 1 for both nonlesional (5.2- fo ld) and les ional skin (4 .2-fold)] were fo und in psoriatic fibrobla sts compared to normal fibrobla sts. T his very high Mn-SOD activ ity was associated with an increase in Mn-SOD mRNA (Fig 1) and in protein expression as determined by -western blot analysis (Fig 2). Densito metric scanning of the Mn-SOD band on the autoradiograph revealed a signifi cant in­crease (p < 0.05) in the amount of Mn-SOD protein in psoriatic fi broblas ts (n = 5) (mcan ± SEM, 1814 ± 473 and 1743 ± 548

,.. arbitrary units, respectively, in les ional and nonlesion al skin) as cOITlpare d to norm al fibroblasts (n = 5) (1 107 ± 287 arbitrary

• units)_

Plasmatic Infiamntatory Markers N o signi ficant differences were tound in the levels of plasmatic orosomucoid and ceruloplas­min b etween psoriatic patients (mean ± SEM, 0.55 ± 0.19 g/ li ter and 310 ± 69 g/liter, respectively) and normal donors (0 .45 ± 0.13

j , g/ liter a nd 356 ± 105 glliter , respectively) .

'-I

DISCUSSION

In tins study we present e vidence that SOD is significantly in­crease d in ce tl s (fibroblasts and circulating erythrocytes) prepared froD;l psoriatic patients . SOD plays a major antioxidant role in cells by converting the superoxide io n (0 '" 2) into H 2 0 2 . In m ost cell s, SOD includes both the soluble C uZn-SOD and the m.itochondrial Mn- SOD. In fibrobl asts, CuZn-SOD has been reported to be localized in the peroxisomcs (Kell er et ai, 1991), wherc there is additional cooperation w ith the catalase and GPX activities. Cata­lase ac t s to neutralize high peroxisoma l H 2 0 2 concen trations, although som e H 2 0 2 m ay escape. T his excess H 2 0 2 in turn is

I normally neutra lized by cytoso Li c GPX. Any no n-neu tra lized H 2 0 2

i can lead to the formation of the highly reactive and damaging OH' by the iron-catalyzed reaction of 0 '" 2 and H 20 2 (Wciss, 1986). Taking into account the effect of elevated levels o f SOD activity on free r adical production , o ur data suggest that both 0 '" 2 and H 2 0 2

levels might be significan tly increased in cul tured psoria tic fibro­blas ts. Direct m easurem ent of fi:ec radicals in psoriatic fibroblasts is required to determine ifhighly reactive 01''1' are also prod uced even with th e slight increase in catalase ac tivity observed.

Antioxidant enzym.e activities have been reported to vary from one cell type to another. Fibro blasts have higher levels of catalase, GPX, and SOD activity than ke ratinocytes (Yohn et ai, 1991). Red b lood cells were fou nd to have high levels of catalase activity (Goth , 1989). T he va lues obtained in tlli s study for catalase and GPX activ ities in normal human fibroblasts and in red bloo d cells are in agreem ent with pre vious reports (M oysan el ai, 1993; Shindo el ai, 1994), but the SOD levels determined here in fibrobl asts differ from a previous report (Moysan cl ai, 1993). This is apparently due to the differences in the assays employed , as SOD valu es have been shown to vary significantly depending on the type of assay (Spitz and O berley, 1989) . Variations in SOD activ ity have also been observed in different disease states. For example, SOD activity has been reported to be decreased in various cancer tissues (Hamanaka el ai, 1991). Previous studies that examined psoriatic epidermi s reported either normal total SOD or decrcased C uZ n- SOD ac tivity (Van Baal' el ai, 1987; liz uka cl ai, 1993 ) and no rmal o r reduced am o un ts of the Mn-SOD en zym e detennined by immunochemical staining of frozen sections of skin (Ko bayashi el ai, 1991) . A recent report by Lentz ct al (1995) , however, describes an increase ill Mn-SOD mRNA expression ;I S determin ed by reverse transcription polym erase chain reaction in frozen-involved epidermis of PSOl;atiC patients.

Interestingly, the increase in pSOl'iatic fi brobl ast SOD activ ity was found to be unrelated to passage number in fibroblasts cultured

A

IPS UIPS N

B

Figure 1. Increased MIl-SOD mRNA in cultured psoriatic fibro­blasts. Slot blot analysis of tota l RNA from normal (IV). lesiona.! (IPS, Involved Psoriatic Skin) and nonlesiollal (UIPS, Uninvolved Psoriatic Skin) cu'!cured fibroblasts hybridized with an Mn-SOD oligonucleotide antisense probe (A) as described under M aretinls nlld 1\1lclllOd-,. Hybridization was performed with a GA PDH probe as control (B).

1328 T H EROND ET A L

25kD ~

UIPS IPS N

Figure 2. Immunoblot comparison of Mn-SOD protein levels in normal and psoriatic fIbroblasts. Fifty microgram s of Iysates from norm al fibrob lasts (N) . lI o nlesional (UIPS), or lesion,,1 (rPS) psoriatic fIbroblasts wcre analyzed by sodium dodccyl sul f.1te ge l e lectropho res is, o'''lIs fe rred to nitroce llulose, and immulloblotted with a monoclo nal anti­human M n-SOD antibod y. T he figure shows sa mples with the highest differell ce ill Mn-SOD expression bc",'Vecn 1I 0 rmal and pso riati c cell s.

from both les ioll al and nonl esional ski n (Gerbaud, personal data) . T his correlates w ith our previous observation that the alteration ill cA MP binding to the R[ and R.II regulatory subunits of cAMP­depende nt protein kinase is also o bserved in both lesio nal and no nlesion al skin (Raynaud e/ ai, 1987). T his suggests that alter­ations observed in psoriatic fibroblasts ma y be independent of the inAamma tory lymphocyte in fi ltratio n noted in lesional tissue. An in crea se in 0 -:- 2 pl'oduction could initiate an in crease in the cellul ar antioxid ant defen se mechanism , i. e., an increase in SOD ac tivity in psoria t ic fibrobl asts. Eleva tio n of SOD activity alone, howe ver, could result in in creased levels of 1-1 2 0 2 in psoriatic fibroblasts, w hi ch migh t act to modify oxidatively important protein s within the cell including enzym es involved in transmembrane signal transduction such as the cAMP-dependent protein kinase sys tem. T hese abno rmal bio logic prope rties of p soriatic fibroblasts mi ght be invo lved in the inductive role of these cells in th e appea rance of psoriatic les ions (Saiag e/ ai, 1985; Krueger and J o rgensen, 1990).

T his increase in SOD activity also is obse rved in circulating cells (i .e. , el-ythrocytes). To de te rmin e whether an increase in the inA ammato ry process might correlate with the increase in SOD activity o bserved in circul atin g red blood cells, we measured prim ary markers of the inAamm atory process (orosomucoid and ceruloplasm in) in plasm a obtained frol11 th e psoriatic do no rs . No change in the level of these m arke rs wa s observed. T hus, the in crease noted in SOD activities in psori ;lti c ceUs appea rs to be independent of the general inA am111atory process of the disease and of th e severi ty of the psoriasis expressed as determined by the Psoriatic Area Severity Ind ex Score (Rayn<lud r/ ai, 1987) (data not shown) .

T ";s 'I'ork ",as -'"PI""1ed li y a c~ra ll/ fro", I'IIIS/;III/ C llr;" ,,"(1);''' ''' I 'Associa /;,," I){II11' la R,'r!wl'r"e coIIIl'e Ie Ca llcer. 11Ie 1110111.1 B"'IlI;1 £ '''' ;11 alld C IIOr/,,/le Pall;"/ ) ,)1' 1,'r!III;cal a-,s;s/allce alld Dr. Maral Dm"lOd;,," )ill" ,,;s "dp '1';/" slal;s/;cal ,,"alys;.<.

THE J OURNAL OF INVESTIGATIVE DERMATOLOGY

REFERENCES

Bannister JV ~ Bannister WH , R otilio G: Aspc c.: ts of the strucrurc, fun crion and i.'ppiiciltio ns of superoxide d isllIlIt<lSC. C rit Rf~1I Bitld/l'/II 22: 11 1-180. 1987

C holllczynski P. Silcchi N: Si ng le-ste p method or RN A iso )<ttion by acid g U31lidi um t11iocyanatc-pl1Cllol ch loroform extractio n. Jlll nt lJiocl,cIII -162 : 156- t 59 , 1987

Cross CC. Halliwell B , 130rish IT. Pryor WA. Ames BN . Sa ul RL. McCord JM. I-farman 0: D:wis co nfe re llce: oxygen rad i4.: als ;1110 human disease. A I/" 1I11/'I1I l\Jcd 107:526-545. 1987

Drilhkin DL. Austill JlI : Spectrop hotoll lCtric studi es: prcparl1tion fro m washed blood cell s: nil-ric ox ide hcm oglo bin alld sulhclllogio bill.) BioI CI/{~I/I 1 '12:5 1-65. 1955

EVl1 in-l3rio ll D. R :lyn:1url F. Pi er A, La ure ll! P. Ledtl c 0 , A ndcrson Wl3: De fi ciency o f cyclic AM P-depcndent pro tein kinases ill human psoril1sis . Prof Nafl Awd Sci USA 83:5272-5276, 1')86

Fr;lI1kclIlIC 1=. Alsar E. Scippo ML. Igo ut A. He lll1c n G. EVl1 in-B rio n 0: Eviden ce for the expressio ll or g rowth hormonc receptors ill hum an placenta. 13iochclII Biop/I}'s /"{t 'S C OllI III 1111 182:48 1-486. 1992

Goth L: Hum;1Il e rythrocyte cata ll1sc. isolat io n with :III improved m cthod , chl1 ractcr­izatio n and comparison to bovine Ii vcr c:ltalasc. c lI.:'yllle 4 1 :19"1- 1 99. 1989

C lill dil11~b U . 1-i:lr:J. SK, Andc nmn W13. Gop:lbkrishn :J. R: Retinoid inhibit the ox idOltivc mo dification of protcin k inasc C indu ced by oxidant tumor promo rers. A rrll 13iClrll clI' Bi(ll'lI),s 300:526-530 .1 993

H il ll101,wka 1-1. Miy:u.:hi Y. Tzchiball il T. Im:Hnura S: Lowered ell Z n supcroxidc di!lmu tasc activity in hUTl1 ,1I11l1illig l1am skin tUllI o rs. ) D(',."IIIloI 18:258 - 26 1. 199 1

li zuk:1 H . Asaga H, Ko ike K. lizuka S: Dccre;lscd e u Zn supcrox ide disl11urasc acrivity in psori ill:ic.: hype rpro liferative cpidermis. Ell,.) D':I'IIltrtol 3:56-58. 1993

J ohanssOII LH. Borg HLA: A spectropilotollle tri c m ethod for determination ofca[aln sl' activity in sl11all tissue sample. 11 11 111 13i"rllclIf 174 :33.1-33 6. 1988

Ke ll er G A, W arner TG, Steim er KS. I--I:illc \vell R.A: e ll Z n SlIpc ro x idc dis1l111t:1sC is il pcroxisomal c nzym c in human fibro bl;lsts :md hcpatom a cells. PrtJc N tHl A(ad Sci

US!l 88:738 1-7385. 199 1 Ko haY;lshi T. Ma[sumo to M . lisuka H. Suzuki K. T 'l11iguschi N: Supcro x.ide dislllur;tsc

in psoria sis squamo us cell CotrCinOlll il and basal c pithe lio m 'l: An i,nmuno histo­che inieal scud )'. 13r) Den",,/ol 124:555-559. 1991

Kniegcr GG , J orgcnsen C M : Experimen ta l models for psori:J. sis. ) I/II!(~S f D I'I1IIt1(ol

95 :565-585. 1990 Lontz W , Sirsjo A, Liu W , Lindberg M. R olllllan 0 , Torl11a H : Increased II1I~A

expre ssio n o f man gancse supcroxidc dislllumsc in psoria sis skin Icsio ns :md in cultured human kcr~ltil1ocytes ex posed to IL- I{3 and TNF-a. Fret' Rodic B;ol J\'1t.~d 18:349-355, 1995

Meier 131-11-1. Radeke S, Se Ue M. YOlln g H. Sics K. Rcsh 1<. Habermehl GG: Hllman fibro blasts rele ase reac ti ve oxygcn species ill rcspo nsc to intc r! euk iJl-1 o r tumor necrosis fac tor. LJiorllt' /JI) 263 :539-545. '1989

Mu),s'lII A. M ;II'quis I. G:J.horiau P, S:lIl1'lI S R, D lI bertrel L. Mo rlil.-re P: U lcrl1vio lct A-induced lipid peroxidotrio ll <l lId ·an tiox idan t defe nsc s),S tC1l1 ill culturcd human skin fib roblasts. J I,",cs/ Den"" /,, /l 00:692-698 , 1993

Naka no M . Killiura 1-1 , I-la ra M. J{ uro iw:1 M , K:IO M . TotSlIll C K. Yoshikawil T: A hi g:hl y sensit ive rnethod fo r determining both Mn- and Cu-Z n Sllpcroxidc dislllut;lse acrivities in ti ssucs ilnd hlood cells. A lwl Bifl( I!CIII I H7:277- 280. 1990

N iwa Y. K:J. ll o h T. Sa kane T . Soh H . I(awai S. Mi )'ilchi Y: Detection o rcllhanccd lipid peroxide levels in patients wi th infl;u11Inaro ry skin diseases. ) eli" 13io(ilCIII N ur,. 2:245-25 1. 1987

O ' Neill P. Dav ies S, Fic lde n EM: T he dfecr of p H and v"rious sal ts upo n rhe ac tivity of a se ries of supc rox ide dislllllta scs . BioclH'''' ) 24"1 :4 1-46. 1988

Pag lia DF. Va len t in e WN: Studies 0 11 th e ql.J:llitative ;Ind q ll ~lI1ti tativc characwrizatio l1 or cryrhroc)ltc g luta thio llc perox idase. J tflb Clill iVIed 70: 158-169. 1967

I"taynaud F. Lcd uc C. Anderson WO. Evain- l3rion D: Retino id treatmcll( of hUlllan pso ri:.ltic fibrob lasts in creascs cyclic AMP-dc pcndc ll( proteill k inase levels.) II/ flcs t

0" .... '," ,,1 H9 : I U5-1 10. 1987 Ra yn:llId F. Gerbaud p. Enjo lras O. Gorin I , DOllnadie u M . Anderson WH, Evain­

Urio n 0: A cAM P binding abno rma lity in psoriasis. Lm/ft'f 11 53- 11 56. 1989 R:lyn:1un F. Gerhaud P. llo ul oc A. Gorin I. Andcrson WOo Evnin-Brio n D: R apid

effec t of treatmcnt of psori:l tic e rythrocytes w ith th e. synthetic retino id acitrctin to inc reOisc 8-:tzido cyclic AMP hindillg Lo the III reglll ;ltory subunir. ) I llIlesr

Dr .... '"'ol 100:77-8 1. 1993 TOllrnicr S. Gcrbaud p. And erson WB. Lohmanll SM. EV:lin-13rio n D , R aYl1:1ud F:

POS(- tT'II"l.sla tional abno rmality or thc typc II c.:ycl ic.: AMP-dependent pro(cin ki ll :lSe in psorias is: 111Odulario n by rClino ic ac id. ) C ell Bille/II'III 57:647- 654. 1995

S:liag P. Coulo m b n. Lebreton C. l3c il E, D lIhcrtret L: Psoriatic fi bro blas[s induce hype rpro li fe ra t io ll of no rllla l kc ratinocy tcs ill the sk ill cquiv;llent m odcl ill IJ;fro. Sr;" "n' 230:669-672. 1985

Schreck R . Bac ucrlc PA.: A ro le ror oxyge ll r ildi c..:als as second m essengers. Trends Cell /Ji,,1 1:39-4 2. 1991

Shilldo Y. W itt E. H:lI"l D. Epstein 'IA/ . Pa cker L: Enzymic .md no n-cn zymic :1l1tioxiei;11Its ill epidermis and den llis or human sk i, l.) IlI lJCst O Cr/H(Jrtll 102: 122-124, 1994

Spitz DR. O hcrl cy L W : An ass;\y fo r supcl'ox ide d isl11l1t'lse acti vity in m.unmaHan tissue ho m ogenarcs. A /wi U;(lc/Il.' 111 179:8- 1 R. 198()

Va ll Baar I-I. Van dc KerkhofP, SclwlwUkJ . Micr P: C UtilIH:OUS supcro x_idc d isl1lutase 'lc ti vit), i ll psorins is. Br ) DCI"I/wfo/ 1 16:462, 1987

Wciss Sj: O,,-')'gcn . ischelll_ia ,lIlei illfl am l1la tion. ilcttl PII)/s hll SC'lIId Suppl 548 :9 - 37. 1986

W hisler RL. Goyette MA , Grallts IS. Newhouse YG: Subl ethallc"c ls of oxidant stress stimulate l11 ultip le serine / threonin e k.ill ases Olnd suppress protein phospha [otscs iu Jurkat T cells. A rc" 13ille" c", /J i"p l, ),s 3 19:23-35, 1995

Yohll JJ. Norri s DA. Yrastorza D G, BUll o IJ . LefrJ A. H ake SS. R.epill c J E: D isparate :1111"ioxid:IIH enzym c activities in cu1 rLl l'ed hurn :1 n C lIt<iIl C OU S fibrob lasts. kcrat ino­c)'tes alld melanocytes.) I, ,,,,·s/ Dcmlnt,,1 97:405- 409. '1 99 '1

"...---- - -

LETTER TO THE EDITOR.

~ack of Induction of IL- l0 Expression in H um an :f{.eratinocytes

fa th e E di tor: We !:"ead wi th interest the report by Grewe et a/ [4] discussing the juducti o n of IL-I0 fo llowin g ultrav iolet B (UV -B) and ultrav iolet ,.1 irradiation of human keratin ocytes (KC) in primal")' cul ture. We fOO have perform ed simil ar UV-B experiments, and have also f!ea t ed hum an KC and A43 1 epidermoid ca rcino m a ceUs wi th lensitizing and irri tant chemi cals. We were unable, however , to pe tect an y l L-l 0 protein in the cul ture supernatants by an enzyl11e­linked in1111unosorben t assay (ELI SA) w ith a sensitivity of l .5 pgl l11l pefo!:"e o r after an y o f the above treatm ents. Furtherm ore, we could pot detect IL-I 0 ml~A except 24 h afte r UV-B doses of200 and ~ OO J hT1 2 . Generall y, our results agree wi th those o f Teunisse n et ai, who "",ere un abl e to detect in te rl eu k.in 10 (IL-l0) mR.NA (35 thennal cycles) o r IL- l [) pro te in in T(C after UV -B (15- 200 J/ m2), or treatment w ith ni ckel sul fate, TL-l alj3, tum or necros.is factor-a, jnterferon-)I, lipopolysacchm;de, phorbol ester, or supcl11atants fi'om phytoh e m agglutinin-stimulated T cells (Teunissen ME M , Koomen CW, de W aal Ma.let)'t R , Bos JD: Human kera tin ocytes arc unable to produ ce IL- IO. J 1II III'sl Derlllato/l 02:632, 19?~ (abstL). .

T h e experim ents th at we perfo rm ed ut1hzed KC deri ved fi'om eith e r fo reskin o r abdominal skin fi'o m five di lfcren t donors. T he sensitive reverse transcrip tion po lym erase chain reaction [5] was empl oyed using cthidium bromide to detect amplified p roducts resolved o n 1. 5'1., aga rose gels and sil ver stainin g w ith a comm ercial kit (Bi o-R ad La bo ratories Ltd) , or 33 P- labeled p rimers to visualize produ cts o n 12.5% polyacrylamide ge ls. We also carried o ut south en1 blotting of th e agarose gels and hybridized IL-I 0 ampli­con~ "",ith a digoxigenin-I abeled human l L- lO probe, w hich was detecte d w ith an an ti- digoxi genin antibody conj ugated to alkalin e phosph atase. T he p1"imers were obtained fi'om C lontech and .success­fully detected abundant LL-l 0 mRNA in RN A de1"ived fi'o m lipopoly­saccharide- or phorbol-ester-stimulated pet;pheral blood monocytes.

The KC fi'om fo ur dillc rcn t do no rs were negativc for TL- 10 rnRNA for up to 24 h fo ll owing irradiation w ith 100 ]1 m

2 of

broadband UV -B, (280-320 nm). JL-I 0 m R N A was detected 24 h afte!:" UV-B irradiatio n wi th 200 and 300 J/ m2 onl y; however, we had to extend the therma l cyclin g to 40 cycles to detect even a weak band. Tn Grewe's study, 28 cycles w e re typically used . In contrast to the results presented in G rewe's stu dy, 100 J / m

2 and

timepoin ts ea rli er th an 24 h di d no t induce IL-1 0 mRNA. Densi­tometry of the IL-1 0 southern b lots probed with digoxigenin showed a 24-fo ld increase in IL-10 ml~A w ith 300jlm

2, but no

de tectable IL-1 0' prote in was fo und in the cul ture supernatan ts . Although no IL-I 0 mllN A inductio n was observed , a 4.5- fo ld

induction of IL - 8 mRNA and an 8- fo ld increase in LL- 8 protein (meas u red by ELISA) w as noted 24 h afte r irradiation with 100 J/m 2 and a 30-fold inductio n of m RNA for the p40 subu nit of JL-12 was seen 3 h after 100 J/ m2. T hu s, thc cells were clearly respon s ive to UV-B. C hanging the calcium conten t oftllc m edium from 0.06 mM to 2 mM fo r 48 h prio r to irradi ation or stimulatin g the KC w ith 10 ng/ml IL- 1a, 200 U/ml IL-2, 0.0 1% sodium dodecyl sulf.,te, o r 200 f.Lg/ml nickel sul fate had no elfect on the ind u ction of IL-I 0 ml~A. Furtherm orc, treatm en t of A43 1 cells with i rritan ts [sod iu m dodeeyl sul fate (0.000 1-0.01 %) o r nonan oic aci d (0.0001 - 0.0 1%)] , sensitizers Inickel sul fate (10- 200 f.Lg/ ml) o r dip h e n y lcypro nc (0.000003 - 0.003%) ]. o r pho rbol ester (1 0 ng/ml) were al l ine lfective in inducing IL-1 0 ml~NA o r protein . Phorbol

ester trea t1nen t, however, did induce tumor necrosis factor-a mRN A, ,m d irritan ts induced IL-S ml~A showin g that the cells were capable of responding to stimulation with c)'tokine gene expression .

O ur observations and those ofTeun.i ssen 1'111/ ra ise several points. First, if IL-I 0 ml~A is in d uced bu t no protein is detccted in the cul ture supern ata nt, is the IL-I 0 protein being sequ cstered, possibly by a solubl e receptor , o r is this anoth er case that illustrates th c p itfa ll s of m easuring o nly mR.NA abundance? Second, the induc­tion of IL-l 0 m RNA [1] an d pro te in [2] by sensitizc rs and IL-l in murine KC could not be reproduced in hum an KC, e ither in o ur studies o r in those of Teunissen ct at. T his suggests that TL- l 0 may be <!n other exam ple, like IL-3 in which hum an and mlll; ne KC express diffe re n t cytokines [7] . T he ind uction of IL- 10 in m urine KC and epidermis by sensitizers bu t not irritan ts has b een dem on­strated both ill Ilit ro [1,2] an d ill Il i ll() 161- It has been proposed that th.i s property m ay be a usefi.I1 means of distin g uishing sensitizers 11'0111 irritan ts ill I/il ro . C learly, this test m ay be inappropriate if human KC do not respond to sensitizers. Finally, is the discrepancy in resul ts between di lferen t stud ies d ue to dilferences in the p rim ers or ELISA kits used (Stratagene p"imers and ELISA kits fro m Genzym e and Laboserve in [4] and C lontech prime rs and ELISA kits tiom R &D system s in our study) , or does it represent true pheno typic di fterences in the ex pressio n of IL-1 0 in human cell s? In suppo rt of the latter is a recen t stud y 13], w hich showed that o nl), five o u t of nin e KC cultures fro m 11lIman subjects showed consti­tu tive IL-l 0 mRNA expressio n and that cul tur ' supen1atan ts had to be concen trated 10- fo ld be fore low «30 pg/ml) levels of IL- 1 0 p rotein were detected , in fo ur irradiate d cul tures. T hese q uestions require resolu tion before epidermal I L-l 0 exp ression can be eftcc­tively integrated into the skin cytokinc n etwork in human subjects.

Melan y Jackson Depar tmen t of Medical M icro bio logy

U niversity of Edinburgh Edi nburgh EH8 9AG

Scotland, UK

KelTY E.Thomson Department of De rmatology

U ni versity of Edinburgh Edinburgh EI-13 9YW

Scotland, UK

Rich ard Laker Departmcn t of Dermatology

U ni versity of Edinburgh Ed inburgh EH3 9YW

Scotland, UK

Mary Norval Department of Medical M icrobiology

University of Edinbu rgh Edinburgh EH8 9AG

Scotland, U1(

0022-202X/ 96 / S·10.50 • Copyrigh t tid J 996 by T he Sociery fo r Investigative Dermatology, Inc .

1329

1330 LETTER TO THE EDITOR

J ohn A. A.Hunter Department of Dermato logy

University ofEdiJlburgh Edinburgh EH3 9YW

Scotland, UK

Roderick C . Mckcnzic Department of Dermatology

Univers ity of Edinburgh Edinburgh EH3 9YW

Scotland, UK

NOTE IN PROOF Another report of lack o flL-1 0 induction in hUlll an KC has appeared (I<.ied C, Michel G, Bectz A, Kemeny L, R uzicka T: Lack of ill duct io n of lL-l 0 in human keratinocytes by illAallll11atory cytokines and UVB. J 111 11".<1

D c rll/tllo/I 03 :443, 1994, abstr.).

REFERENCES

I . Elik AH. I{ ;Hz SI: Idcmi fi catiOIl ~lIld ind uc tio n of kc rati ll oc}' I'c-dcri vcd IL- IO. J /1111/111/1 0 ' 142:92-96. 1992

2 . Ellk AI-!: ill tc rl cukilJ- IO. In : Luger TA. Schwarz T (cds.). Epiderlll fl l e/'illll'" Fa{fOfJ

II//(/ C yfo}..·;w's. Marcel Dekker In c .. New York, pp J 13- 130, L t) t)4

3. Enk C D. Srcdni D. Ulauvcl t A, Katz 51: Il1dtl crion of IL- IO ge lle expression in hUllIan kc r.lt illocytcs by UV- I3 exposure.) 1111111 1111(1/1 54:4 85 1-48 56, '1IJIJS

4 . Grewe M. Gyufko K. KrlltlllaTlIl J: IIll:c rl eukil1- IO prod uction by culrurcd hlll11;lII kcrnti nocytcs: rcgll iatio ll by ul travio le t 13 and u ltr;1 v io lct A I.) 11I !! ('S f 1)cn // nf(l /

J 0'1 :3-(,. 1995 S. K Olldo S. KOllo T, Sauder ON. M ckellz ie ItC: IL-H g ene exprcssif) lI and produ Clion

in hunwn keratiTl o cytcs a lld their IIlo dui;1 tion hy UV-B. J l'III('sl OCr/ // nlol

10 1:69U-694. '1993 6. KOll do S. P:'lstorc S. Shi vj i G M. Mcken zie ltC , Slimie r. DN: Char<lc l:cri zatio ll o f

c picie rllla l cyrokiJlc pro fi les in sensiti zati on and e licitatio ll phases of 'lllcrgic cOlnn e r: de rl11:niri s as we ll ,IS irrit 'lIl t d Crllw titi s in lI10use skin. L}'/IIl'ltnkillL' alltl C )""kill" Res 13:%7-3 7(1 . 1994

7. Kondo S, C iarle rra A, Turner I<.J . Sauder DN. Mckenzie R C : H uman kcra tin ocytcs do 1I0t produce IL-3: :lhsC l1 cc or l L-3 rnRNA and prote in . An tibo dies 1.0

GM-CS F and IL-() (but no t. IL-3) neutralise " IL-3 " bioactivity. J lI l11esl

DL'I'IIIa',,/ J(J4.:335- 33 9. 1995

Repl y: "W' e rcad with su.rpri se the Ictter by Jackson cl ai, in which the authors report on thcir m ethodological probl ems in detection of I L -10 mRNA and, in particu lar lL-I O protein production, by UV -irradiated human kera tinocytes.

The fa ilure to dctect increascd IL-J 0 ml~A cxpress ion in human keratinocytes cxposed to 100 J/ m2 ultra vio lct-B (UV-B) radia tion may be duc to lack of sensitivity of the revcrsc-tran­scriptasc-polymerase cha in reaction (RT-PCR) method used. N ot only sclectio n of primers, but also optimization of reversc transcrip­tion and PCR is essentia l for the detection of low amo unts of tra nscripts . In our hands, reverse transcription of tota l RNA was optimal with the use of random h examcr primers resul ting in highcst amplification produ cts for several keratinocyte-derived cytokin es including IL-l0 . For the sa kc of spccifi c rcversc tran­scription of mRNA , the data published 111 were genera ted using oligo-dT I ~ primers . T hese findings have been corroboratcd in an independent study cmploying R T -PC R methods, dem onstrating that exposure of culturcd human kcratinocytcs to doses as low as 50 J / m 2 UVB radiation significantl y indu ccs I L- 1 0 mR.NA expression [2]_ Since RT-PCTI.. has some pitfa ll s, we have additionally pe r­fo r m ed Northern-blot experiments (Fig 1). Increased IL-10 l",RNA steady state levels could bc detected in cultured normal human keratinocytes following UVB o r UVA 1 irrad iation . More­ove r, essen tially identical data have been obtain ed by N orthern blot ana lysis in an independent stud y condu cted by Y. Araganc and T. Schwarz, employing cultured human keratinocytes as well as 1m cel1s (data not shown). T hese if( (lilm findings are supported by l-ecent if( lIillo studies demonstrating that UV irradiation of human skin leads to increased exp rcssion of IL-10 ml~A cxpression in epidermaJ keratinocytes [3 J. In aggregate, these studies provide furth e r proof that UVB radiation at physiologically releva nt doses is

2B s RNA

185 RNA

THE JOURNAL OF INVESTIGATI VE DERMATOLOGY

<D > ::>

<D > ::>

< IL - l O m RNA

Figure 1. Northern blot analysis of IL-10 mRNA expression in UV-irradiatcd l10rtnal hUluan kcrati11ocytcS. Normal human kcrati-1I 0cytes we re left unirradi ated o r exposed to '100 J / m 2 VV-B and 32 kJlcm" UV-A 1 radiatio n, respectiv el y. Afte r 24 h , extra ction of total RNA w ith subsequcnt Northern blot anal ysis usi ng a digoxigcnin-labclcd, IL-l 0-specifi c cDNA W ;IS carri ed o ut by standard techniqucs. L~/i pnffei, ethidiu m bromidc sta in of RN A agarose ge l electrophoresis; R(~lrl f'arrc/ , IL-I O I11I~ A-spccifi c signal.

capable of inducing IL- 10 ml~A exprcssion in human keratino­cytes both ill "itm and ill "i ll().

W c have previo usl y rcported tbat UYB-irxadia ted normal human. ke ratinocytes secre te signifi cant amounts of IL-10 protein into culturc supernatants 11] . Parallel to their f.1 ilure to detect increased IL-10 mRNA expression , Jackson ct ai, by employing enzyme­linked immlln osorbent assay (ELISA) techniques, also could not measure increascd IL-10 protein secretion by hum an keratinocytes. In our study. IL-1.0 protein secretion could be detec ted by means of two difle rent ELISAs, both of which differed fi'om that cmployed by J ackson e/ at [1] . Onc ELISA was bascd on a po lyclonal anti-hul11an­IL-1 0 an tiserum , w hcrcas thc o the r used a non-lJcutralizing anti­IL-l 0 antibody. Since nonc of the ELiSAs successfully uscd by li S

was based on neutraLiz ing antibodies , it may be that IL-I0 released into superna tants of stimula ted hl1l11 '111 keratinocytes binds to a. carri er protein, e.g. , soluble IL-1.0 receptors, which may compete w ith binding of neu trali zin g anti-I L-I0 antibodies and t11ereb y result in lack of I L-:I 0 protein detection. Our rcsu lts have agau"l bcen corroborated and extended in an independent study, in which IL-10 proteins could be measured by ELISA afte r cmde sup em a­tants wcre conccn tra ted 10-fold [2]. In our hands, concentration Of culture supernatan ts was not necessary to de tect sufi-icien t amounts of IL-10 protein . It should be noted, how ever, that our culture conditions were optimal in reducing the volul11c of supernatants to 1. 5 m ilS X 106 ce lls in orde r to avoid dilution of lL-10 .

Esscn tiaUy identical data were recently o b tained in independent experimcnts conducted by M . Duvic and S. Ullrich (Fig 2). [n these experim ents, UV -B radiation dose-dcpendent IL-10 protcin release into supcrnatants of human skin equivalcnts [4] was assessed by employing an ELISA diffcrent from those used in our sr.udy [1] . In. thcse experiments, cul ture supernatants did no t ha ve to bc concen_ trated for detcction of IL-1 0 protein . In addi tion, irradiation with, UV-B induced IL-] 0 protein secre tion to a level similar to that observed in our study [1] . It co uld also bc e legantly demonstrated tha t increased IL-I0 protein secretion of UV-B-irradiated humat~ skin equivalents was keratinocyte-dcrived, since immunohisto­chemica l studies revea led strong JL-I O proteiJ) expression witllir) th e cytoplasm of keratinocytes following UV -B radiation exposure of ski.n eq uival ents (Fig 3).

In summary, wc and othcrs have indepcndently provided defin_ itivc proof both at thc 11l[~A and protcin level that hUI11 <U1 keratinocytes are capablc of syn thes izing and relcasing increased amounts o f [L-I0 upon exposure to UV-B radiation. N egative experimcntal results, as dcscribed by Jackson el ai, need to b e intcrpreted with great care in order to avoid neglect ofkeratinocyte fun ctions important for the cytokine network in hUl11an skin .

VOL. 106. NO . 6 JUNE 1996

Markus Grewe C lini cal and Experimental Photodennatology

Department of Dermatology Heinrich-Heine-University, Dusseldorf, Germany

Madeleine Duvic University of Texas Medical School and

M. D. Anderson Cancer Center Houston, Texas, U .S.A.

Y oshinori Aragane Department of Dermatology

University of Munster, Munster, Germany

Thomas Schwarz Department of Dermatology

University of Munster, Munster , Germany

Stephen E. Ulbich Department of Immunology

M. D. Anderson Cancer Center Bouston, Texas, U .S.A.

Jean Krutmann Clinical and Experimental Photodermatology

Department of Dermatology Heimich-Beine-University, Dusseldorf, Gen11any

B

LETTER TO THE EDITOR 1331

150,--------------------------,

100

50

o~----,_----_,----_.----~

o 500 1000 1500 2000

uv (JIm 2)

Figure 2. IL-l0 protein secretion into the supernatant of human skin equivalents. Slcin equivalents, constructed by plating 5 X 105 human foreslci n keratinocytes over type I collagen gels containing human foreskin fibroblasts, were raised to the air inte rface and grown for 14 days in FAD media as described [4]. The slcin equiva lents were then irradiated with UV radiation supplied by a singek unfiltered FS-40 sunlamp (0-1600 J/ m2) as meas ured with a lL-700 research radiometer. The cul ture fluid was co llected and the IL-I0 concentration determined by ELISA usin g reagents suppli ed by PharMingen Inc. (San Diego, CAl. For each treatment. the mean value of triplicate samples was detelwined. Error bars. SD. A representative experinlent is shown.

'. . ~,"

'.­. (

Figure 3. Keratinocyte IL-l0 protein expression after UVB irradiation of hun Ian skin equivalents. The ski n equi val ents were treated with a UV light at varyi ng amounts 'lI1d 24 h later the SE were harvested and fixed for inllllullohistochemistr)' [4]. Anti-human IL-I0 antibody (rat IgG l, clone ]£S3-9D7) was used at 1 :10 dilution and detected with DAKO LSAJ3 2° antibody and countcrstained with hematoxylin and eos in. A. non-irradia ted s\...; n equivalent-basa l layer (an"lII) (40 X); B, skin equivalent receiving 1600 J / m 2-dennis below (40 X) ; C . negative contro l ofSE in B without primary lL-IO antibody (40 X).

REFERENCES

1. G r e\NC M . Gyurko K. Krutl1lann J: In tcrlcukin 10 production by cultured hum an keratinocytcs: rcgularioll by ultravio let 13 and ultr~violct A I radi;}tiOIl. J JII Pcst

DtTHIClIt1f 104 :3-6, 1995 2. Enk C D, Srcdni D. Blauvelt A. Katz SI: Induction or IL-IO gene expression in

human kCfatinocytcs by UVB exposure ill vivo and ill vitro. J 1111111///10/

154:4851-4856. 1995

3 . K:mg K, !-I ,111111lCrocrg C. Mell nie r L. Cooper KD: CD t Ib + mn croph;1gcs that infilrr;ltc hUll1 <1 n e pidermis afte r in vivo ulrravio lct exposure potently produce IL- \ U ;l lld rcprCScll t the 1ll ;~j or sccrccofY Soun:c of cpidcrrnal IL- IO prote in . J /IIIIIIIIIW/ 153:5256- 5264. 1994

4. Duvic M. Nelsoll DC . AIlI1 <lfCIJ:l M. C lio M, Esglcycs-Ilibol. T. RCll1cllyik E. ULmer R.. Rapini R.I'. Sacks ['G. C layman GL. Davi es PJ A. Thacher S: Kcratinoc), tc transgillt<llllin:l sc expression v;lril!s in squam olls ce ll carcino mas .

.J ltlllcsr J)cn ll tl(o/ 102:462-4 61) , 11)94