Study of HLA-DR Synthesis in CulturedHuman Keratinocytes*

Norman E. Wikner, M.D., J. Clark Huff, M.D., David A. Norris, M.D., Steven T. Boyce, Ph.D.,Matthew Cary, B.S., Mark Kissinger, B.S., and William L. Weston, M.D.Department of Dermatology, University of Colorado School ot Medicine, Denver, Colorado, and Department of Surgery,University of California at San Diego, San Diego, California, U.S.A.

Within the normal human epidermis only Langerhans andindeterminate cells express HLA-DR. Human keratino-cytes (HK), however, may also express HLA-DR in eertaindisease states eharaeterized by mononuelear cell infiltrates.Previous studies have shown that HK synthesize HLA-DRin response to stimulation by interferon gamma (INF-y).The purposes of this study were to define conditions underwhich cultured HK might express HLA-DR and to com-pare the HLA-DR synthesis of HK with that of monocytes.HLA-DR expression by HK as determined by indirect im-munofluorescence of HK cultures was absent under stan-dard low calcium conditions and remained absent with theaddition of calcium, serum, mitogens, and supernatantsfrom Pam-212 cells containing epidermal thymocyte-activating factor. HLA-DR expression in HK was induced

by cocultivation with concanavalin A-stimulated periph-eral blood mononuclear cells (PBMC), btit not unstimu-lated PBMC. This effect was time-dependent and directlyrelated to the nuniber of PBMC. HLA-DR expression wasalso induced in a time- and dose-dependent manner byaddition of supernatant from stimulated PBMC (SS) or byaddition of recombinant INF-y but not by addition of in-terleukin (IL)-l or IL-2. Induction by either SS or INF-ywas blocked by an antiserum to INF-y. As determined bya semiquantitative immunoprecipitation technique, HLA-DR synthesis by HK was directly related to INF-y con-centration. The pattern ot HLA-DR peptides produced byHK was similar to that of monocytes, but the relative quan-tity synthesized was far less than that of monocytes. J IiwcstDermatol 87:559-564, 1986

w ithin the normal buman epidermis, expressionof class II or HLA-DR antigens is confined tothe dendritic Langerhans cell [1|. However, incertain disease states, such as graft-versus-hostdisease, lichen planus, mycosis fungoides, and

skin-graft rejection, the keratinocytes of the epidermis may alsoexpress HLA-DR antigens on their surfaces [2-6]. The diseaseentities associated with HLA-DR expression by keratinocytes areones characterized histologically by inflammatory infiltrates ofmononuclear cells in the skin and arc thought to involve cell-mediated inimune reactions in the skin.

Our understanding of the phenomenon of HLA-DR expressionby cells has advanced considerably in recent years. Several groups[7-10] bave clearly demonstrated that HLA-DR expression bykeratinocytes is due to tbe syntbesis of these proteins by kerat-inocytes and not to passive acquisition of the proteins trom other

cells. Other studies 111-13] have demonstrated that the signal forinduction of HLA-L)R synthesis by keratinocytes is gamma in-terferon (IFN-7).

In an attempt to better understand HLA-DR expression bykeratinocytes, we have studied HLA-DR expression in culturesof human keratinocytes, attempting to stimulate in vitro the con-ditions that lead to HLA-DR expression in vivo. We have con-firmed the role oflFN-y in induction of HLA-DR expression andhave also demonstrated that the HLA-DR peptides synthesizedby keratinocytes arc qualitatively similar to but quantitatively lessthan those of monocytes.

. MATERIALS AND METHODS

Human Keratinoeyte Cultures Human keratinocytes (HK)were isolated from neonatal foreskins and grown in a 0.1 niMcalcium, serum-frcc medium 153, as described by Boycc and Ham

Manuscript received December 3, 1985; accepted for publication May20, 1986.

Supported by grants GM 35068, Al 16637, and AM 26427 from theNational Institutes of Health.

*I'arts of this work were presented at the Annual Meetings of TheSociety for Investigative Dermatology, Inc., May 9, 1984 and May 3,1985.

Reprint requests to: Norman E. Wikncr, M.D., Department of Der-matology B-153, University of Colorado Health Sciences Center, 4200East Ninth Avenue, Denver, Colorado 80202.

Abbreviations:AM: adherent monocytesBSA: bovine serum albuminCon-A: concanavalin Acpm: counts per minute ' .DOC: sodium deoxycholate

ETAF: epidermal thymocyte-aetivatmg factorFCS: fetal calf serumHK: human keratinocytesIFN-y: interferon gammaIIF: indirect immunofluorescenceIL-1: interleukin 1lL-2: interleukin 2NRS: normal (prcinimune) rabbit seruinPBS: pbosphatc-buffered salineI'HA: phytohemagglutininI'BMC: peripheral blood mononuclear cellsSDS: sodium dodecyl sulfateSl'A: staphylococcal protein ASS: superuatants from stimulated PBMCTCA: tricliloroacetic acid

0()22-202X/86/S03.50 Copyright © 1986 by The Society for Inve.stigative Dermatology, Inc.

559

560 WIKNIiR HT AL THE JOURNAL OF INVliSTICAMVL DERMATOLOGY

[14]. For these studies, second-passage keratinocytes were grownon 2-chanibcr Lab-Tek slides. By indirect iniiiitinofluorescencestaining for keratins, all cells in eulture on tbe slides were kerat-inocytes.

Indirect Immunofluorescence Examination of HK for HLA-DR was performed by indirect imniunofluorescence (HF). Cellson slides were washed 3 times in phosphate-buffered saline (PBS)(10 niM Pi, 150 niM NaCl, pH 7.4) fixed and pcrmcabilizcd incold ( —20°C) acetone, and air-dried. Cells fixed on slides wereincubated for 16 h at 4°C with a 1 :100 dilution of a murinemonoclonal antibody in PBS with 0.1 % sodium azide and 5%normal goat serum. Slides were then washed in PBS, incubated2 b in a 1 :100 dilution of goat antimouse IgG F(ab')2 (Tago Inc.,Buriingatne, California), and wasbed again in PBS. Coverslipswere mounted over a glycerol-PBS mounting medium (pH 8.5)containing 0.1% parapbenylenedianiiiie [15]. Slides were viewedunder an Olympus cpifluorescciice microscope equipped with anHBO-lOO mercury light source and a filtcr-reflector system forfiuorescein. The percentage of HK staining positively for HLA-DR on each half of a slide was cicrivcd by counting positive andnegative cells (at least 200) in 10 bigh-power (40 x) fields.

Antibodies and Antisera As positive control antibodies foridentification of HLA antigens on HK in culture, murine mono-clonal antibodies to HLA-A,B,C (Clone MB4().5, Atlantic An-tibodies, Scarborough, Maine) and to (in -microglobulin (CloneL368, Becton-Dickinsoii, Stiiinyvale, C^alitornia) were used. Amonoclonal antibody, C)KT6, was used as a negative control andfor verification tbat Langerbans cells were absent from the cul-tured epidermal cells. For identification of HLA-DR by IIF, anIgCii monoclonal antibody (Clone MCD1/2C3, Atlantic Anti-bodies) was used. For immunoprecipitation of HLA-DR, an IgG2.,nioiioclonal antibody (Clone L243, Becton-Dickinson) was em-ployed. Rabbit antiseruni to IFN-y was from Interferon Sciences,New Brunswick, New Jersey.

Peripheral Blood Mononuclear Cells Peripbcral bloodmononuclear cells (PBMC) from normal donors were isolatedover Ficoll-Hypaquc |16]. Stimulated supernatants (SS) fromPBMC were obtained by eulture of PBMC for 3 days in MCDB153 in the presence of 10% fetal calf serum (FCS) and 10 /xg/mlconcanavalin A (Con-A).

Pam-212 Cells Pam-212 is the designation for a transformedmouse epidermal cell line |17|. These cells were generously pro-vided by Dr. Stuart Yuspa ofthe National Cancer Institute. Tbeywere grown in serial eulture in minimal essential medium (GIBCO,Grand Island, New York) plus 8-10% Hyclone-screencd FCS(Hyclone, Logan, Utab). Conditioned medium from these cul-tures was used as a source of epidermal thymocyte-activatingfactor (ETAF).

Stimulation of HLA-DR Expression by HK After 5-7 daysin eulture on Lab-Tek slides, changes in eulture conditions andaddition of various stimuli were performed. HLA-DR expressionby HK was examined 0-5 days later by IIF. Changes in cultureconditions examined included increases in calcium concentrations(tip to 0.1 HIM), addition of 10% normal human scrum, plasma,or FCS, addition of 10 /U-g/ml Con-A or phytobemagglutiiiin(PHA), and addition of 10% supernatant from cultured Pam-212cells containing ETAF. PBMC were added and coculturcd withHK along with 10% FCS in thc presence or absence of mitogen(10 ;ag/ml Con-A). Other stimulants added included the SS fromPBMC, interleukin 1 (lL-1), interleukin 2 (IL-2; CollaborativeRe.search Inc., Lexington, Massachusetts), and recombinant IFN-y (Genentech, Inc., San Francisco, California). Blocking oftheactivity of IFN-y was achieved by incubating SS or IFN-y withcitber a rabbit antiserum to human IFN-y (K)"* nu/nil, InterferonSciences, New Brunswick, New Jersey) or normal rabbit scrum(NRS). After incubation for 1 h at 37°C, the mixture was addedto cultured HK.

Immunoprecipitation of Human HLA-DR Antigens Sec-ond-passage HK which had grown in 60-niin plastic dishes to80% confluence (approximately 10'' cells/dish) were stimulatedfor 48 h with 0-100 units/ml IFN-y. As positive control cells forHLA-DR expression, adherent human monocytes (AM) wereobtained by incubation of PBMC in plastic dislies contaiiiitigRPMI culture medium and 10% fresb human serum |16]. TheAM were stimulated with 50 units/ml IFN-y and were processedby the immunoprecipitation technique in a manner identical totbe HK. Methionine-deficieiit media were added to cultured cellsfor 4 h. Each plate then received a 2 h pulse of 50 mCi ["'̂ S]-liietliioniiie. The plates were washed 6 times in cold PBS, andthe cells in each plate were Iysed at 4°C in 0.5 ml of a PBS buffercontaining 0.5% Triton X-100, 0.25% sodium deoxycholate(DOC), 2 HIM phenylmethylsulfonyl fluoride, and 10 mM EDTA.Each plate was washed with an additional 0.3 ml of this lysisbuffer and was scraped witb a rubber policeman. The volume oflysate from eacb plate was made 1.3 ml witb addition of lysisbuffer. Tbe lysate was spun at 10,000,i; for 15 min, and 0.3 mlof the supernatant was removed for precipitation with trichlo-roacetic acid (TCA). The remainder was used for immunopre-cipitation.

To each 0.3-aliquot of cell lysate supernatant, TCA was addedto a final concetitration of 10%. After ati overnight incubation at4°C, eacb sample was centrifuged at 10,000 _(; for 15 min. Pelletswere resuspended in 1 ml of 0.6 N NaOH containing 0.1% so-dium dodecyl sulfate (SDS). Samples were placed in BeckmanReady-Solv 566436, and radioactivity was measured in a Beckmanmodel LS-750() /3-counter. The counts per minute (cpm) werecorrected for any dilutions in order to reflect thc cpm in theoriginal lysate.

Immunoprecipitation was performed according to publishedprocedures [18]. A suspension of fixed Cowan strain Stapliylo-coccus aureus was prepared according to the manufacturer's in-struetions and used as a solid phase source of staphylococcal Pro-tein A (SPA) (Enzyme Center, Maiden, Massachusetts). In orderto clear from each lysate any proteins tbat might bind nonspe-cifically to SPA, eacb sample was incubated for 1 b witb 200 pXof SPA. Samples were centrifuged for 5 min, and the supernatantwas removed. SDS was added to the stipernatant to a final con-centration of 1%. To each sample 10 /xl of the undiluted mono-clonal antibody to HLA-DR was added, and samples were in-cubated overnight at 4°C. SPA (100 lA) was added to eacb sample.After 1 h, the SPA was pelleted and washed once with PBScontaining 1% SDS, 1% Triton X-100, 0.5% DOC, and 0.5%bovine scrum albumin (BSA). Five additional washes were per-formed witb the same wash solution without BSA. Immune com-plexes were eluted from the SPA by boiling for 5 min in 20 p\of Laemmli satiiple buffer. Satnplcs were centrifuged, and tbesupernatants were applied to the wells of a 5—20% gradient poly-acrylamidc slab gel containing 1% SDS and with a 5% stackinggel. Tbe gel electropboresis was performed according to standardmetbods ]19]. After electrophorcsis, the separated proteins wereelectroblotted onto nitrocellulose paper (Trans-blot, Biorad). Au-toradiography was performed by exposure of Kodak X-Omat Rfilm (Eastman Kodak, Rochester, New York) to the nitrocellulosepaper for 4 days.

The bands on tbe autoradiograpb corresponding to HLA-DRpeptides were identified by comparison to prestaiiied molecularweight standards on the nitrocellulose paper and to the bandsproduced by the positive control cells (human AM). Thc relativedensities of the bands on the x-ray film were measured for eachsample with a Zeineh Model SL-2D Scanning Laser Densometer(Bio Med Instruments, Fullerton, California). The densities ofthe bands were integrated (Apple Computer Co., Cupertino,California) and expressed as arbitrary units for each sample. Inorder to relate the relative amounts of HLA-DR synthesis to totalprotein synthesis, an HLA-DR synthesis index, the ratio of theHLA-DR band integral to tbe TCA-precipitablc counts, was usedto compare relative HLA-DR synthesis.

VC5L. H7. NO. .S NOVEMBER HLA-DR SYNTHESIS IN HUMAN KERATINOCYTES 561

Figure 1. Indirect immunofluorescence of cultured HK for HLA-DR.HK cultured up to 2 weeks and with the addition of calcium, serum,mitogen, and supernatants with l'ani-212 cells remained negative forHLA-DR.

RESULTS

Second-passage HK stained positively for HLA-A,B,C and /32-microglobulin but did not express HLA-DR or OI'tT6 after cul-ture up to 2 weeks (Fig I). No HLA-DR expression by HK wasnoted with changes in calcium concentration, addition of FCS,normal human scrum, plasma, mitogcns, or ETAF.

When PBMC were added to HK and cocultured in the presenceof Con-A (fO Mg/i"!) for 3 days, HLA-Df̂ expression by HKwas seen (Fig 2). The positive staining by HF was a granularfluorescence that appeared to be accentuated in the perinuclearregion rather than just on the periphery of cells. This is the samepattern of fluorescence that was seen when fixed permeabilizedHK were stained for HLA A,B,C and ^wnicroglobuliii. Al-though HLA-DR '̂ adherent PBMC were also seen in clusters onthe slides, these intensely fluorescent cells were much smaller thanHK and were easily distinguished from the keratinocytes.

The relationship of numbers of PBMC and mitogen to HLA-DR expression by HK is shown in Fig. 3. When the ratio of addedPBMC to HK in each chamber was 50:f or less, the percentageof HK expressing HLA-DR after 3 days of cocultivation was nomore than 2%. With higher cell ratios, HLA-DR expression byHK increased progressively as did the intensity of the fluores-cence. At cell ratios greater than 500:1, HK were destroyed dur-ing the cocultivation procedure. Mitogen was an essential elementof the induction of HLA-DR in HK: without Con-A, no induc-tion was noted.

The time course of HLA-DR induction is shown in Fig 4. Thepercentage of HK expressing HLA-DR and the intensity of thestaining were highest between 2 and 4 days of cocultivation withmitogen-stimulated PBMC. Beyond 4 days the majority of theHK had been destroyed.

When the supernatant from PBMC stimulated with Con-A for3 days was added at a 1 :1 ratio to the HK culture medium, HLA-DR expression was noted in 70-80% of the HK. This findingimplied that a mediator of HLA-f3>R expression was produced bythe stimulated PBMC. Three mediators produced in such cul-

Figurc 2. Indirect imniunofluorescence of HK cocultured with Con-Astimulated PBMC. The granular staining of permeabilized HK with aperinuclear accentuation represents specific staining for HLA-DR. Per-meabilization of the cells does not allow evaluation of their surtace an-tigens. ' ' , • . , • ' :

(0O

Q.

DCQ

<

X

40

30

20

10 (1+)nCon A No

Con A50/1

Con A NoCon A

100/1

Con A NoCon A

500/1Number PBMC/Number HK

Figure 3, Effect of mitogcn-stinuilated PBMC on HLA-DR expressionby HK. HK were cocultured for 3 days with varying numbers of PBMCin the presence of 10% FCS and either 10 /^g/ml Con-A {sluuicd) or noCon-A {solid bars). The ratio of PBMC^ to HK is given on the abscissa.The percent of HK staining positively tor HLA-DR is given on the or-ilinatc. For those conditions in which HK stained positively, the intensityof staining (on a scale of 0 to 3 -I-) is given in parcnthcsa.

tures, IL-1, IL-2, and IFN-y, were tested for their ability to stim-ulate HLA-f.:)R. Neither IL-1 nor IL-2 (50 units/ml) producedHLA-DR expression by FIK, whereas recombinant IFN-y producedHLA-DR expression in a time- and dose-dependent fashion (Fig5). Little expression was induced at IFN-y concentrations of 100units/ml or less, but with higher concentrations, expression wasnoted within 24 h and approached 80% by % h.

In order to evaluate whether IFN-y was the critical mediatorof HLA-DR expression, 2 potent stimulators, 500 units/ml ofIFN-y and the SS, were incubated with rabbit antiserum to IFN-y (10' neutralizing units) or NRS and then added to the culturesfor 3 days. The induction of HLA-DR in HK by IFN.-y and bythe SS was blocked completely by the antiserum to IFN but notby normal rabbit serum.

The restilts of immunoprecipitation and autoradiography areshown in Fig 6. Monocyte lysates produced a characteristic pat-tern of bands clustered between M, 27,000 and 36,000 (lane G).The specificity of this pattern was demonstrated by immunopre-cipitation of adherent monocyte lysates with an impertinent mu-rine monoclonal antibody to lysozyme. Under these circum-stances, no bands in the A-f, 27,000-36,000 K range were produced(Fig 7).

D

(V)

(3*)

I (1')

Time (Days) After Addition of PBMC (500/1)

Figure 4. Time course of induction in HK of HLA-DR expression bycocultured monocytes. PBMC were added to HK cultures at a PBMCto HK ratio of 500:1. The cells were then cocultured tbr 0-5 days. Thepercent of HK staining tor HLA-DR is shown by the vertical bars. Theintensity of this staining is given in parentheses on a scale of 0 to 3-h.

562 WIKNER ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

100-

80 '

8 0 ' •

20' '

•• I

LY DR

A1000

A600

OlOO

• 50

12 24 48 72 96 120

Time (Hours) After Addition of Gamma Interferon

Figure 5. Time and IFN-y dose dependence of HK HLA-DR expression.IFN-y, in the concentrations shown (right panel, units/nil), was added tocultures of HK. After incubation for 12-120 h, the percentage of positivelystaining HK was measured and plotted for each concentration and timepoint.

Immunoprecipitation of HK lysatcs by anti-HLA-DR producedno such bands if the HK had not been stimulated with IFN-y(lane A), or had been stimulated with concentrations of IFN--)/less than 50 units/ml (lanes B-D). The lack of the characteristicpattern produced by monocytes confirms that HLA-1.^R synthesisby normal HK is minimal to absent. When HK were stimulatedwith IFN-y at concentrations of 50 or 100 units/ml, bands ofHLA-DR peptides similar to those of monocytes were seen (lanesE and F). The lower threshold of IFN-y concentrations neededfor HLA-DR synthesis as detected by this method likely reflectsthe greater sensitivity of the method compared with ininuinoflu-orescence.

The similarity of the peptide bands immunoprecipitated fromHK and monocytes is shown by comparison of the densitometerscans (Fig 8). Altliough not absolutely identical, the major peaks

43.

26-

14-

Figure 7. Expanded view of an autoradiograph representing proteinsimmunoprecipitated from lysates of '̂'S-labeled adherent monocytes bymonoclonal antibodies to lysozyme (L'i') or HLA-DR (DR). Migrationpositions of molecular weight markers (in kD) are shown (left).

are present for both cell types. Quantitation of the immunopre-cipitation data is shown in Table I, which presents the IFN-yconcentration, the integral of the densitometer tracings, and theTCA-precipitable counts corresponding to each lane of Fig 6.Only Isackground density is present in HK stimulated with IFN-y at 10 units/ml or less. A comparison of HLA-DR synthesis

MONOCYTES

2 0 0 -?'.

9 7 -

6 8 -

4 3 -

2 6 -

14 —r

A B C D E F GFigure 6. Autoradiograph generated as described in Materials and Methodsrepresenting proteins immunoprecipitated by monoclonal anti-HLA-DRfrom lysates of'^S-labeled cells. Lanes A-P, keratinocytes. Lane G, ad-herent monocytes. INF-y doses for lanes A-G were 0, 1, 5, 10, 50, 100,and 50 units/ml, respectively. Migration positions of molecular weightmarkers (in kD) are shown (left).

LOWMOLECULAR WEIGHT

KERATINOCYTES

HIGH

LOWMOLECULAR WEIGHT

HIGH

Figure 8. Densitometer scans of lanes G (upper scan) and F (lower scan)trom Fig. 6. The abscissa is the scan position along the autoradiograph.The ordinate is the band density.

J

VOL. 87. NO. 5 NOVEMliliR l9Sft HLA-DR SYNTI lliSIS IN HUMAN KERAIINOCYTES 563

Table I. Stimulation of HLA-DR Synthesis in Human Keratinocytes (HK) and Adherent Monocytes (AM)

Cell Type

HK AM

IFN-y (units/ml)"DR band density''TCA-preclpitahle

counts'Ratio''

0.01.5

24.0

0.9

1.1.

12.

018

5.01.2

20.6

10.01.3

11.5

50.011.228.6

1.3 0.9 1.7 5.8

100.018.416.0

17.1

50.017.52.6

100.0

"y-Intcrfcron conccntr.ition.''Band density in arbitr.iry units.'Expressed as tnillions of cpni.''Ratio of band density to TCA-prceipitable counts expressed as a percentage of tlie value for AM.

indices is presented in Fig 9. The dose-dependent relationship ofH K HLA-DR synthesis to IFN-y concentration is again evident.However, compared with total protein synthesis (total TCA-precipitable counts), HLA-DR synthesis by stimulated HK is only17% that of the stimulated monocytes.

DISCUSSION

The initial approach that was taken in these studies of HLA-DRexpression by keratinocytes was to attempt to simulate in vitro,the conditions that are associated with this finding in vivo. Thedisease conditions or experimental situations that are characterizedby HLA-DR expression are those which have mononuclear cellinflammatory infiltrates in the skin, and which are thought toinvolve cell-mediated immune reactions. After determining thatalterations in cultured conditions of HK, such as addition of mi-togen or serum, did not induce HLA-DR, we cocultured PBMCwith HK. Our in vitro finding that cocultivation of PBMC, alongwith mitogen, led to HLA-DR expression by HK, agrees withthe in vivo observation that HLA-DR expression by HK occursduring cell-mediated immune reactions involving nionocyte-macrophages and stimulated T lymphocytes.

Our data are consistent with those of Basham et al [l l] andVolc-Platzer et al [13] regarding the role of IFN-y in this phe-nomenon. The induction of HLA-DR expression by supernatantsfrom stimulated PBMC and by recombinant IFN-y, the blockingof induction by both stimuli with an antibody to IFN-y, and thedose-response relationship of HLA-DR e.xpression to IFN-y con-

MONOCYTES

m

a:a

100-,

90-

80-

70-

60-

50-

40-

30-

20-

10-

0

KERATINOCYTES

I ' ' I • • I ' ^ - r ^ " - I - " — r

0 1 5 10 50 100 50

GAMMA INTERFERON DOSE (units/ml)

Figure 9. The ratio of f-ILA-f5R band density to total protein synthesis(as measured by TCA-precipitable counts) is expressed in arbitrary unitson the ordiiiate. Cell types and IFN-y concentrations are as shown.

centrations, confirm that IFN-y is the mediator produced in thesesystems that leads to HLA-DR expression by other cells. Othermediators, such as ETAF, IL-1, and IL-2 (alone or in combina-tion) were ineffective in this regard.

The results of these studies also confirm by direct analysis thatHLA-DR expression by keratinocytes is due to do novo synthesisof HLA-DR peptides, rather than to passive acquisition of theantigens from other cells. The appearance, in fixed, pcrmeabilizedHK, of HLA-DR staining predominantly within the cells andaccentuated in the perinuclear region implies synthesis within thecells (Fig 2). Furthermore, we were able to induce HLA-DRexpression in HK easily with SS and IFN-y without addition ofany known HLA-DR* cells. Finally, the incorporation of [̂ '̂"Sj-methionine into peptides of the appropriate molecular weightimmunoprecipitated by monoclonal antibody to an H L A - I S Rcommon determinant is the conclusive demonstration that thepeptides of HLA-DR are synthesized by keratinocytes under ap-propriate stimulatory conditions.

According to our analysis of the peptides immunoprecipitatedfrom IFN-y-stimu!ated HK, the pattern of peptides is similar to,it not identical to, that of nionocytes. This is consistent with thework of Morhenn et al. [10] showing that HLA-DR moleculessynthesized by keratinocytes have the same charge and molecularweight as those synthesized by autologous lymphocytes. By IIFstaining, HLA-DR + HK are never as strongly positive as mono-cytes. The results of our attempts to quantitate HLA-DR pro-duction by HK shows that monocytes stimulated with half of theIFN-y concentration as HK, produce 5-10 times as much HLA-DR peptides.

Even though the phenomenon of HLA-DR expression by ke-ratinocytes is relatively well understood at this point, the role ofH L A - I 3 R on keratinocytes is still unknown. Whether HLA-DR +HK serve a role in interaction with the immune system has notyet been clarified. A postulated role for keratinocytes as antigen-presenting cells has not been verified. In fact, it is not yet clearthat HLA-DR* HK may stimulate an allogeneic lymphocyte re-sponse. Studies by Roberts et al. |20] have suggested that onepossible role of the HLA-DR ' epidermis is to facilitate the move-ment ot the Langerhans cell, an antigen-presenting, bone mar-row-derived dendritic cell, into the epidermis. One possibility isthat expression of these histocompatibility antigens by HK indisease states or experimental conditions is an epiphenomenonthat simply serves as a convenient marker of a stimulated cell-mediated immune reaction, and IFN-y production. Now thatHLA-DR expression can be reliably induced in HK by IFN-y,and now that recombinant IFN-y is readily available, it shouldbe possible to understand whether the phenomenon of HLA-DRexpression by HK is important in epidermal-immune systeminteractions.

The cnithors wish lo tlimtk the Gmcntcch Conqnuiy for its f^viicrtlif rccoiiiliiiumt y-iiitcrjcron needed for these experiments.

iis donation of

564 WIKNER liT AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

REFERENCES1. Stingl G, Tamaki K, Katz SI: Origin and function of epidermal

Langerhans cells. Immunol Rev 53:149-174, 19802. Lanipert IA, Smitters AJ, Cliisliolni I'M: Expression of la antigen

on epidermal keratinocytes in graft-versus-host disease. Nature293:149-150, 1981

3. Mason DW, Dallman M, Barclay AN: Graft-versus-host diseaseinduces expression of la antigen in rat epidermal cells and gutepithelium. Nature 293:150-151, 1981

4. Malmnas Tjeinlund U: la-like antigens in lichen planus. Acta DermVenereol (Stockh) 60:309-315, 1980

5. Malmnas-Tjernlund U: (82 microglobulin and HLA-DR-(Ia)-anti-gens in normal and pathological skin. Acta Derm Venereol |Suppl](Stockh) 94:9-41, 1981

6. Dallman MJ, Mason DW: Induction of la antigens on murine epi-dermal cells during the rejection of skin allografts. Transplantation36:222-224, 1983

7. Breathnach SM, Katz SI: Keratinocytes synthesize la antigen in acutecutaneous graft-versus-host disease. J Immunol 131:2741-2743,1983

8. Volc-Platzer B, Majdic O, Knapp W, Wolff K, Hinterberger W,Lechner K, Stingl G: Evidence of HLA-DR biosynthesis by humankeratmocytes in disease. J Exp Med 159:1784-1789, 1984

9. Daynes RA, Emam M, Krueger GG, Roberts LK: Expression of laantigen on epidermal keratinocytes after grafting of normal skinto nude mice. J Immunol 130:1536-1539, 1983

10. Morhenn VB, Charron DJ, Engelman EG: Human skin cells syn-thesize HLA-DR molecules of the same charge and molecularweight as those synthesized by autologous lymphocytes. J InvestDermatol 78:125-127, 1982

11. Basham TY, Nickoloff BJ, Merigan TC, Morhenn VB: Reeombi-

nant gamma interferon induces HLA-DR expression on eulttiredhuman keratinocytes. J Invest Dermatol 83:88-90, 1984

12. Basham TY, Nickoloff BJ, Merigan TC, Morhenn VB: Recombi-nant gamma mterferon differentially regulates class II antigenexpression and biosynthesis on cultured normal human keratino-cytes. J Interferon Res 5:23-32, 1985

13. Vole-l'latzer B, Leibl H, Luger T, Zohn G, Stingl G: Human epi-dermal cells synthesize HLA-DR alloantigen in vitro upon stim-ulation with gamma interferon. J Invest Dermatol 85:16-19, 1985

14. Boyce ST, Ham RG: Calcium-regulated differentiation of normalhuman epidermal keratinocytes in chemically defined cloiial ctil-ture and serum-free serial culture. J Invest Dermatol8 (suppl) :33S-40S, 1983

15. HuffJC, Weston WL, Wanda KD: Enhancement of specific im-munofluorescent findings with use of a para-phenylenediaminemounting buffer. J Invest Dermatol 78:449-450, 1982

16. Boyum A: Isolation of mononuclear cells and granuiocytes fromhuman blood. Scand J Clin Lab Invest 21 (Suppl 97):77-89, 1968

17. Yuspa SH, Hawley-Nelson P, Koehler B, Stanley JR: A survey oftransformation markers in differentiating epidermal cell lines inculture. Cancer Res 40:4694-4703, 1980

18. Tucci A, Goldberger G, Whitehead AS, Kay RM, Woods DE, Col-ten HR: Biosynthesis of postsynthetie processing of human C-reactive protein. J Immunol 131:2416-2419, 1983

19. Tsang VCU: Enzyme-linked inimunoelectrotransfer blot techniquesfor studying the specificities of antigens and antibodies separatedby gel electrophoresis. Methods Enzymol 92:377—391, 1983

20. Roberts LK, Krueger GG, Daynes RA: Correlation between theinducible keratinoeytes expression of la and the movement ofLangerhans cells into the epidermis. J Immunol 134:3781—3784,1985