Autoantibodies in Lichen Planus Pemphigoides React with aNovel Epitope within the C-Terminal NC16A Domain ofBP180

Detlef Zillikens,*† Frederic Caux,†‡ Jose M. Mascaro Jr,† Ulrich Wesselmann,* Enno Schmidt,*Catherine Prost,‡ Jeffrey P. Callen,§ Eva-B. Brocker,* Luis A. Diaz,†¶ and George J. Giudice†**Departments of Dermatology, *University of Wurzburg, Wurzburg, Germany; †Medical College of Wisconsin, Milwaukee, Wisconsin, U.S.A.;‡Hopital St. Louis, Paris, France; §University of Louisville, Kentucky, U.S.A.; ¶VA Medical Center, Milwaukee; ** Department of Biochemistry,Medical College of Wisconsin, Milwaukee, Wisconsin, U.S.A.

Lichen planus pemphigoides is an autoimmune sub-epidermal blistering disease. The finding of immuno-globulin G antibodies directed against the basementmembrane zone differentiates it from bullous lichenplanus. The aim of this study was to identify the targetantigen of lichen planus pemphigoides autoantibodies.Sera from lichen planus pemphigoides patients (n J 4)stained the epidermal side of NaCl-split human skinin a pattern indistinguishable from that produced bybullous pemphigoid sera. In bullous pemphigoid, theautoimmune response is directed against BP180, ahemidesmosomal transmembrane collagenous glyco-protein. We previously demonstrated that bullouspemphigoid sera predominantly react with a set offour epitopes (MCW-0 through MCW-3) clusteredwithin a 45 amino acid stretch of the major noncolla-genous extracellular domain (NC16A) of BP180. Byimmunoblotting and enzyme-linked immunosorbentassay, lichen planus pemphigoides sera were alsostrongly reactive with recombinant bullous pem-

The term ‘‘lichen ruber pemphigoides’’ or lichen planuspemphigoides (LPP) was first used by Kaposi in 1892for a disease characterized by lichen planus (LP) lesionsand numerous bullae (Kaposi, 1892). In addition toclinical and histologic criteria, immunofluorescence

microscopic findings are crucial in distinguishing between bullousLP and LPP. In bullous LP, blisters are confined to LP lesions andare a consequence of severe basal cell degeneration, as found onhistopathology. In contrast, bullous lesions in LPP follow theappearance of papules and plaques and may be found on both LPlesions and previously unaffected skin. In LPP, light microscopyshows the typical features of LP in papular lesions and changes ofbullous pemphigoid in lesional biopsies of blistered skin. Most

Manuscript received November 24, 1998; revised February 22, 1999;accepted for publication March 9, 1999.

Reprint requests to: Dr. Detlef Zillikens, Department of Dermatology,University of Wurzburg, Josef-Schneider-Str. 2, 97080 Wurzburg,Germany.

Abbreviations: BP, bullous pemphigoid; GST, glutathione S-transferase;IB, immunoblot; LP, lichen planus; LPP, lichen planus pemphigoides.

0022-202X/99/$14.00 · Copyright © 1999 by The Society for Investigative Dermatology, Inc.

117

phigoid 180 NC16A. The lichen planus pemphigoidesepitopes were further mapped using a series of over-lapping recombinant segments of the NC16A domain.All lichen planus pemphigoides sera reacted withamino acids 46–59 of domain NC16A, a protein seg-ment that was previously shown to be unreactivewith bullous pemphigoid sera. Two lichen planuspemphigoides sera, in addition, reacted with theimmunodominant antigenic region associated withbullous pemphigoid. In conclusion, there are now fivebullous diseases that are associated with an auto-immune response to BP180: bullous pemphigoid;pemphigoid/herpes gestationis; cicatricial pem-phigoid; linear immunoglobulin A disease; and lichenplanus pemphigoides. In addition, we have identifieda novel epitope within the BP180 NC16A domain,designated MCW-4, that appears to be uniquely recog-nized by sera from patients with lichen planus pem-phigoides. Key words: autoantigen/collagen/epitope/hemidesmosome. J Invest Dermatol 113:117–121, 1999

importantly, the finding of linear deposits of immunoglobulin (Ig)Gand/or C3 at the basement membrane zone (BMZ) in perilesionalskin differentiates LPP from bullous LP (Stingl and Holubar, 1975;Hintner et al, 1979; Lang and Maize, 1983; Willesteed et al, 1991).

The clinical and immunologic observations associated with bullaformation in LPP are similar to findings in bullous pemphigoid(BP), an autoimmune subepidermal blistering disease of the elderlythat is also characterized by the presence of autoantibodies to theBMZ (reviewed by Zillikens et al, 1996a). Antibodies in BP aredirected to two hemidesmosomal proteins, BP180 and BP230(Stanley et al, 1981, 1988; Labib et al, 1986; Diaz et al, 1990).BP230 is located within the intracellular hemidesmosomal plaque(Tanaka et al, 1990), whereas BP180 is a transmembrane glycopro-tein with a long extracellular tail that spans the lamina lucida ofthe BMZ (Giudice et al, 1992; Li et al, 1993; Bedane et al, 1997;Masunaga et al, 1997; Zillikens and Giudice, 1999). Data from apassive transfer mouse model strongly suggest the pathogenicrelevance of antibodies to BP180 (Liu et al, 1993; Giudice et al,1995). Epitope mapping studies have shown that the NC16Adomain of human BP180 harbors the major extracellular antigenicsites recognized by BP sera (Giudice et al, 1993; Zillikens et al,1997a). Further, NC16A was shown to contain four distinct BP-

118 ZILLIKENS ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Table I. Clinical and immunofluorescence findings in LPP patientsa

LPP-KN LPP-NO LPP-DA LPP-PA

Age 52 43 61 34Sex F M F FRace Caucasian Caucasian African CaucasianDirect IF C3 , IgG C3 IgG, C3 IgG, C3Indirect IF titerb 1:80 1:80 1:320 1:160Clinical features Generalized lichenoid papules; Verrucous papules and plaques Pruritic violaceus verrucous Violaceus papules on trunk,

blisters on dorsal aspects of on the dorsum of the hands papules disseminated over lower arms, legs, and feet;hands and feet and wrists; vesiculae and bullae trunk, buttocks, thighs; vesicles bullae on thighs and feet

on thighs, lower legs, abdomen on thighsand arms

Treatment Clearance of blisters with oral Clearance with oral Clearance with topical Clearance with tapering oralprednisone—initially 100 mg, prednisone—initially 60 mg, corticosteroids corticosteroids over 8 wktapering over 2 wk; treatment tapering over 4 wkof LP with oral psoralenultraviolet A

aAll patients showed white streaks on buccal mucosa.bIndirect IF was performed on 1 M NaCl-split human skin.

associated epitopes – all of which are clustered within the N-terminal 45 amino acid stretch of this protein domain (Zillikenset al, 1997a). Using the NC16A domain as the target antigen,sensitive immunoblot (Matsumura et al, 1996) and enzyme-linkedimmunosorbent assay (ELISA) techniques (Zillikens et al, 1997b)for the detection of anti-BP180 autoantibodies in the sera of BPpatients have recently been developed. In addition to the NC16Adomain, autoantibodies may react with other antigenic sites on theextra-and intracellular domains of BP180 (Balding et al, 1996;Murakami et al, 1998; Pierrard et al, 1999; Nie and Hashimoto,1999). Whereas an argument can be made that LPP is merely thecoexistence of two diseases, LP and BP, there are several clinicalobservations that support the concept that LPP represents a clinicalentity distinct from either LP or BP. The mean age of our LPPpatients is 48 y, which is in contrast to the mean age of 75 y inthe 115 BP patients treated at the University of Wurzburg overthe past 10 y (Kippes et al, 1999). In contrast to BP, bullous lesionsin LPP are preceded by papules and plaques. In addition, LPPappears to show a considerably better response to treatmentcompared with BP (Bouloc et al, 1998). Finally, as reported in thispaper, we now have immunologic data that distinguishes LPP fromBP at the epitope level.

Previous studies to characterize the autoantigen in LPP haveyielded conflicting results. Okada et al (1986) reported LPP autoanti-bodies to a 240 kDa protein also recognized by antibodies in BP.Subsequently, LPP sera were found to react with the two majorBP antigens, BP180 and BP230, as well as a unique band at200 kDa (Davis et al, 1991; Tamada et al, 1995; Ogg et al, 1997;Bouloc et al, 1998). In this study, we used recombinant forms ofBP180 to determine whether this autoantigen was indeed a targetof LPP autoantibodies. We found that LPP sera, like BP sera, labelthe epidermal side of salt-split skin and react with the NC16Adomain of BP180. In contrast to BP, however, LPP sera recognizeda novel epitope located on the C-terminal portion of the NC16Adomain of BP180.

MATERIALS AND METHODS

Patients and sera Skin and serum samples were obtained from 4 wellcharacterized LPP patients. Clinical features of the patients are summarizedin Table I. In all patients, LP lesions preceded the formation of blisterswhich occurred on both involved and uninvolved skin. Biopsies frompapular lesions showed typical changes of LP histologically and the presenceof cytoid bodies by direct immunofluorescence (IF). Histopathologicalexamination of biopsies from blistered skin revealed subepidermal vesicula-tion. Direct IF microscopy of skin biopsies taken from sites immediatelyadjacent to bullae showed linear deposits of IgG and/or C3 at the BMZ.Indirect IF analysis on 1 M NaCl-separated human skin was performed asdescribed previously (Zillikens et al, 1996b).

Figure 1. Schematic diagram of BP180 and recombinant forms ofthe NC16A domain used in this study. TM, transmembrane domain;the NC16A domain has been subdivided into five regions each ofapproximately 15 amino acids in length. Each of the fusion proteinscontains an N-terminal GST moiety. Amino acid residue numbers areshown above the boxes.

Preparation of recombinant proteins DNA segments encoding theentire NC16A domain and various fragments thereof (Fig 1) were obtainedby polymerase chain reaction amplification using a previously clonedhuman BP180 cDNA as template (Giudice et al, 1992, 1993; Zillikenset al, 1997a). The amplification products were subcloned into the BamHI/EcoRI sites of pGEX-2T (Pharmacia, Piscataway, NJ). The resultingglutathione S-transferase (GST)-BP180 fusion proteins and recombinantGST were expressed in Escherichia coli strain DH5α and purified byglutathione–agarose affinity chromatography, as described (Giudice et al,1993; Zillikens et al, 1997a).

Immunoblotting of recombinant proteins and skin extractsSodium dodecyl sulfate–polyacrylamide gel electrophoresis and immunob-lotting were performed as reported (Zillikens et al, 1996b). Epidermal anddermal extracts were prepared as described (Labib et al, 1986; Zillikenset al, 1996b) and fractionated on 6% and 5% protein gels, respectively,whereas affinity-purified bacterial fusion proteins were fractionated on 15%gels. To eliminate reactivity with the GST moiety of the fusion proteins,sera of patients and controls were preadsorbed with a bacterial cell lysatecontaining recombinant GST prior to labeling of immunoblots.

ELISA for recombinant GST and GST-NC16A To determine ELISAreactivity with BP180 NC16A, we used a recently developed procedure(Zillikens et al, 1997b). Briefly, each well of a standard 96-well microtiterplate was coated with 400 ng of purified recombinant GST-NC16A1–5.Each well was then incubated with 50 µl of primary serum (LPP, BP, or

VOL. 113, NO. 1 JULY 1999 LPP AUTOANTIBODIES TARGET NC16A OF BP180 119

Figure 2. LPP sera react with the epidermal side of NaCl-splithuman skin. One mole per liter NaCl-separated human skin was reactedwith serum of patient LPP-KN, followed by a second incubation withfluorescein isothiocyanate-labeled goat anti-human IgG. Scale bar: 100 µm.

control) diluted 200-fold in phosphate-buffered saline, pH 7.2, containing1% bovine serum albumin and 0.05% Tween-20 (Sigma, Diesenhofen,Germany), followed by an incubation with horseradish peroxidase-labeledgoat anti-human IgG (Kirkegaard and Perry, Gaithersburg, MD) diluted10,000-fold in phosphate-buffered saline, with 1% bovine serum albuminand 0.05% Tween-20. After substrate addition, the optical density (OD)of the reaction solution was measured at a wavelength of 492 nm. Eachserum was assayed in triplicate for reactivity with both GST-NC16A1–5and recombinant GST. For each serum tested, the mean OD readingobtained with GST was subtracted from the mean reading obtained withGST-NC16A1–5. To minimize plate-to-plate variability, each ELISA plateincluded the same four internal controls as described (Zillikens et al, 1997b).

Immunoadsorption procedures Immunoadsorptions were performedusing a liquid phase protocol as described previously (Zillikens et al, 1997a).Diluted sera (1:100) were incubated overnight with recombinant GST orwith one of the GST fusion proteins containing various segments of theNC16A domain. The mixtures were then centrifuged at 10,000 3 g for15 min at 4°C. Immunoadsorptions were verified to be complete by theabsence of immunoblot reactivity with the recombinant protein that hadbeen used as the immunoadsorbent.

RESULTS

LPP sera react with the epidermal side of 1 M NaCl-splitskin All LPP sera included in this study revealed circulating IgGautoantibodies binding to the epidermal side of NaCl-split skin(Fig 2). Antibody titers ranged from 1:80 to 1:320 (Table I). Nocirculating IgA or IgM antibodies were detected. The pattern ofindirect IF reactivity was the same as in two control BP sera. Twosera from patients with epidermolysis bullosa acquisita reacted withthe dermal side of the split (data not shown).

Three of four LPP sera react with BP180 extracted fromepidermis To characterize further these anti-BMZ antibodies inLPP sera, we performed immunoblot (IB) studies using epidermaland dermal extracts. Three LPP sera reacted with BP180 extractedfrom epidermis. No reactivity was found with BP230. By IBanalysis against a dermal extract, the two control EBA sera labeledtype VII collagen whereas, LPP and normal control sera showedno specific reactivity (data not shown).

LPP sera react with recombinant BP180 NC16A by immuno-blot analysis and ELISA To study reactivity of LPP sera withthe BP180 NC16A domain, we preadsorbed the sera againstrecombinant GST and assayed them by IB analysis against affinity-purified GST-NC16A. All four LPP sera, like the BP control sera,reacted with this region of BP180 (Fig 3). This reactivity wasconfirmed by ELISA. Autoantibodies to BP180 NC16A weredetected at levels above the cut-off of the ELISA (mean of controlsera plus 3 SD) in all LPP sera (Table II).

Figure 3. LPP sera react with the immunodominant NC16A domainof BP180. GST-NC16A1–5 was fractionated by 15% sodium dodecylsulfate–polyacrylamide gel electrophoresis, transferred to nitrocellulose, andreacted with a BP serum (lane 1), sera from LPP-KN, LPP-NO, LPP-DA,and LPP-PA (lanes 2–5), and a normal control (NC) serum (lane 6). Allsera were completely preadsorbed against recombinant GST. The migrationpositions of molecular weight markers of 45, 31, and 21 kDa are shownat the right.

Table II. LPP sera react with an epitope in region 4 of theNC16A domain of BP180

LPP Patients

LPP-KN LPP-NO LPP-DA LPP-PA

BP180NC16A1–5 1 1 1 1 1 1 1ELISAa 0.82b 0.75 0.47 0.9IB reactivity with:

NC16A1c – – – –NC16A2c – – 1 –NC16A2.5c – 1 1 –NC16A3c – 1 – –NC16A1–3c – 1 1 –NC16A4d 1 1 1 1 1 1NC16A5d – – – –

a1:200 dilutions of LPP sera were assayed in triplicate against affinity-purifiedforms of GST-NC16A and recombinant GST. The mean OD492 reading wassubtracted from the mean reading for GST-NC16A1–5.

bNumbers indicate the results of the OD492 reading. The cut-off for the ELISAwas set at an OD492 reading of 0.43 (mean of controls plus 3 SD).

c1:100 dilutions of LPP sera were reacted by immunoblot analysis against GST fusionproteins containing NC16A1, NC16A2, NC16A2.5, NC16A3, and NC16A1–3.

dImmunoblot reactivity with NC16A4 and NC16A 5 was assessed indirectly asexplained in the text and demonstrated in Fig 5. After preadsorption of 1:100dilutions of LPP sera against NC16A1–3, the sera continued to react withNC16A1–5. This reactivity was completely abolished after preadsorption of the LPPsera with NC16A2–4, demonstrating the presence of an epitope within NC16A4and the absence of such in NC16A5.

Within the BP180 NC16A domain, LPP sera show a differentpattern of reactivity compared with BP sera To characterizethe antigenic sites recognized by LPP sera within the NC16Adomain, we performed an IB analysis using the following sixsubregions of NC16A (Fig 4): region 1 (GST-NC16A1; lane 2),region 2 (GST-NC16A2; lane 3), the junctions of regions 2 and 3(GST-NC16A2.5; lane 4), region 3 (GST-NC16A3; lane 5), regions1 through 3 (GST-NC16A1–3; lane 6), and the entire NC16Adomain (GST-NC16A1–5; lane 7). Equal mass amounts of thevarious recombinant proteins were used for this analysis. All sera

120 ZILLIKENS ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 4. LPP sera show different reactivity with recombinantforms of BP180 NC16A compared with BP sera. Recombinantproteins were fractionated by 15% sodium dodecyl sulfate–polyacrylamidegel electrophoresis, transferred to nitrocellulose, and labeled with BP (BP-VI) and LPP sera (LPP-KN and LPP-PA). The three immunoblots shownin this figure contained the same loadings of recombinant, affinity-purifiedproteins: GST (lane 1) and fusion proteins NC16A1, NC16A2, NC16A2.5,NC16A3, NC16A1–3, and NC16A1–5 (lanes 2–7, respectively). Themigration positions of molecular weight markers of 45, 31, 21, and 14 kDaare shown at the right.

Figure 5. Autoantibodies in LPP sera react with an epitope locatedin region 4 of BP180 NC16A. Recombinant proteins were fractionatedby 15% sodium dodecyl sulfate–polyacrylamide gel electrophoresis,transferred to nitrocellulose, and labeled with two LPP sera [LPP-NO in(a)–(c) and LPP-PA in (d)–(f)]. Parts (a) and (d) contained same proteinloadings, likewise (b) and (e) and (c) and (f). Fusion protein GST-NC16Awas loaded in lanes 2, 4, 6, 8, 10, and 12. Lanes 1 and 3 contained GST-NC16A1–3, lanes 5 and 7 GST-NC16A2–4, and lanes 9 and 11 GST-NC16A-2–5. Lanes 1, 2, 5, 6, 9, and 10 were incubated with LPP serapreadsorbed with recombinant GST. Lanes 3 and 4 were incubatedwith sera preadsorbed against GST-NC16A1–3, lanes 7 and 8 with serapreadsorbed against GST-NC16A2–4, and lanes 11 and 12 with serapreadsorbed against GST-NC16A2–5. The migration positions of molecularweight markers of 45 and 31 kDa are shown at the right.

were preadsorbed against recombinant GST to eliminate reactivitywith the GST moiety of the fusion protein. GST was included ineach immunoblot assay (lane 1) to demonstrate that this adsorptionwas complete. All LPP sera, like BP control sera, reacted withGST-NC16A1–5. No serum reacted with region 1, one serumwith region 2, two sera with region 2.5, and one serum withregion 3. Interestingly, in contrast to BP sera, sera from LPP-KNand LPP-PA did not react with GST-NC16A1–3, and a serumfrom LPP-NO revealed lower reactivity with NC16A1–3 comparedwith reactivity with NC16A1–5 as demonstrated in Table II. Serafrom LPP-KN and LPP-PA showed no reactivity with any of theN-terminal fragments of NC16A (Fig 4).

LPP sera react with a novel epitope within the NC16Adomain of BP180 To identify further the binding sites of LPPsera, we assayed for reactivity against the NC16A domain afterdepleting the sera of reactivity with certain subfragments of thisBP180 domain. The immunodepletion was accomplished by aliquid phase preadsorption of the sera with affinity-purified BP180fusion proteins. Representative results of this analysis are shown inFig 5. After preadsorption with the control protein, GST, all four

of the LPP sera showed reactivity with the fusion protein containingthe full-length NC16A domain (GST-NC16A1–5), as well as withGST-NC16A2–4, and GST-NC16A2–5; two LPP sera reactedwith GST-NC16A1–3 (lanes 1, 2, 5, 6, 9, 10). After depletion ofreactivity with GST-NC16A1–3, all four of the LPP sera retainedtheir reactivity with full-length NC16A (lanes 3 and 4), indicatingthat the LPP sera reacted with one or more epitopes located onregions 4 and/or 5 of NC16A. After preadsorption with eitherGST-NC16A2–4 or GST-NC16A2–5, however, all four of theLPP sera lost the ability to bind to the full-length NC16A domain(lanes 7, 8, 11, 12). Thus, there was no detectable reactivity of anyof the LPP sera with either regions 1 or 5. Taken together, theseresults indicate that all four of the LPP sera included in this studycontained antibodies that reacted with the 14 amino acid proteinsegment designated region 4 of the BP180 NC16A domain.

DISCUSSION

The initial phase of this investigation into the autoimmune responsein LPP involved the analysis of sera from four well characterizedLPP patients by indirect IF on human skin sections and by IBusing a human epidermal extract as the antigen source. By both ofthese techniques, the LPP sera showed reactivity patterns similarto those exhibited by BP sera, i.e., linear IgG staining of the BMZand reactivity with a 180 kDa epidermal antigen which comigratedwith BP180. To determine more rigorously whether BP180 wasindeed the target antigen of LPP autoantibodies, these sera wereassayed for reactivity with recombinant forms of human BP180.Like sera from most BP patients, LPP sera were shown to reactwith the immunodominant NC16A domain of BP180. As reportedpreviously for BP sera (Zillikens et al, 1997b), we found that IBand ELISA protocols using recombinant BP180 NC16A as thetarget antigen were more sensitive than IB with epidermal extractsfor detecting anti-BP180 autoantibodies.

Interestingly, when the fine specificities of LPP autoantibodieswere further defined by epitope mapping analysis, it was uncoveredthat LPP sera exhibited a reactivity pattern distinct from those ofBP sera. We previously demonstrated that autoantibodies in BPreact with a tightly clustered set of four epitopes within the N-terminal 45 amino acids of the NC16A domain of BP180 (designatedMCW-0, MCW-1, MCW-2, and MCW-3; Zillikens et al, 1997b).These BP-related antigenic sites correspond to regions 1 through3 of NC16A. No reactivity with regions 4 or 5 have been detectedin BP sera. Importantly, all LPP sera included in this study reactedwith an epitope in region 4. This region comprises amino acids46–59 of the NC16A domain. We propose to designate this novelepitope MCW-4. In two LPP patients, MCW-4 was the onlyantigenic site targeted within NC16A. The remaining two patients,in addition, recognized BP-related epitopes in regions 2 and 3.

One can only speculate on reasons for the differences in reactivitypatterns of autoantibodies in BP and LPP. An interesting hypothesishas been that the lichenoid infiltrate leading to damage anddegeneration of basal keratinocytes in LP may induce the productionof antibodies to the BMZ (Stingl and Holubar, 1975). In fact, theinflammatory process in LP lesions, possibly on a certain humanleukocyte antigen background, may trigger an autoimmune responseto a different site on BP180 compared with the site recognized byautoantibodies in BP. It remains possible that the autoimmuneresponses to BP180 that are associated with BP and LPP involvedistinct pathogenic mechanisms. LPP patients of this study camefrom a similar ethnic background as BP patients included in aprevious epitope mapping study (Zillikens et al, 1997a). Therefore,it is unlikely that ethnic differences account for the differentreactivity with BP180 in LPP compared with BP.

In contrast to previously published reports (Davis et al, 1991;Ogg et al, 1997; Bouloc et al, 1998), we found no evidence ofreactivity of LPP autoantibodies with antigens of 200 kDa, 230 kDa,or 240 kDa. These observed immunologic differences could bedue to a heterogeneity of the autoimmune response in this disease.It is also quite possible, however, that these differences are merely

VOL. 113, NO. 1 JULY 1999 LPP AUTOANTIBODIES TARGET NC16A OF BP180 121

due to technical variation, e.g., differences in the methods used toprepare the antigen extracts or differences in the conditions of theimmunologic assays. One recently reported case of LPP doesprovide evidence for heterogeneity of the autoimmune responseamong LPP patients (Bouloc et al, 1998). The serum of this LPPpatient bound to the dermal side of NaCl-split skin; whereas, LPPsera in this study, as well as those previously reported, label theepidermal side of the split. The target antigen in this recentlydescribed patient remained uncharacterized, but from the indirectIF result, it is highly unlikely that the BP180 NC16A domain istargeted in this patient. It was suggested that the clinical phenotypeof LPP might be caused by antibodies to different components ofthe BMZ (Bouloc et al, 1998). This is similar to the situation foundin patients with the clinical phenotype of cicatricial pemphigoid,where autoantibodies may be directed to BP180 or laminin5 (Bernard et al, 1992; Domloge-Hultsch et al, 1992; Baldinget al, 1996).

In summary, we demonstrate that BP180 is the target of LPPautoantibodies binding to the epidermal side of NaCl-split skin.Like sera from most BP patients, LPP sera react with the immuno-dominant NC16A domain of BP180. Within the NC16A domain,LPP sera recognize a novel epitope located at the C-terminalportion of NC16A. Altogether, there are now five autoimmunesubepidermal bullous diseases in which BP180 has been shownto be the target: BP; pemphigoid/herpes gestationis; cicatricialpemphigoid; linear IgA disease (Zone et al, 1998); and LPP.Future investigations will address the hypothesis that fine specificitydifferences in autoantibodies could account for the clinical hetero-geneity seen in patients with these various diseases.

This work was supported by U.S. Public Health Service Grants R01-AR40410(G.J.G.), RO1-AR32599, R37-AR32081 (L.A.D.), by Deutsche Forschungs-gemeinschaft Grant Zi 439/2–1 and Grant 98.073.1 from the WilhelmSander-Stiftung, Munich, Germany (D.Z.). We thank Heike Krenig and IakovChimanovitch, University of Wurzburg, and Shannon M. Ewing, Medical Collegeof Wisconsin, for technical assistance.

REFERENCESBalding SD, Prost C, Diaz LA, Bernard P, Bedane C, Aberdam D, Giudice GJ:

Cicatricial pemphigoid antibodies react with multiple sites on the BP180extracellular domain. J Invest Dermatol 106:141–146, 1996

Bedane C, McMillan JR, Balding SD, et al: Bullous pemphigoid and cicatricialpemphigoid autoantibodies react with ultrastructurally separable epitopes onthe BP180 ectodomain: evidence that BP180 spans the lamina lucida. J InvestDermatol 108:901–907, 1997

Bernard P, Prost C, Durepaire N, Basset-Seguin N, Didierjean L, Saurat JH:The major cicatricial pemphigoid antigen is a 180-kD protein that showsimmunologic cross-reactivities with the bullous pemphigoid antigen. J InvestDermatol 99:174–179, 1992

Bouloc A, Vignon-Pennamen M-D, Caux F, et al: Lichen planus pemphigoides is aheterogeneous disease: a report of five cases studied by immunoelectronmicroscopy. Br J Dermatol 138:972–980, 1998

Davis AL, Bhogal BS, Whitehead P, Frith P, Murdoch ME, Leigh IM, WojnarowskaF: Lichen planus pemphigoides: its relationship to bullous pemphigoid. Br JDermatol 125:263–271, 1991

Diaz LA, Ratrie III H, Saunders WS, Futamura S, Squiquera HL, Anhalt GJ, GiudiceGJ: Isolation of a human epidermal cDNA corresponding to the 180-kDautoantigen recognized by bullous pemphigoid and herpes gestationis sera.Immunolocalization of this protein to the hemidesmosome. J Clin Invest 86:1088–1094, 1990

Domloge-Hultsch N, Gammon WR, Briggaman RA, Gil SG, Carter WG, YanceyKB: Epiligrin, the major human keratinocyte integrin ligand, is targeted inboth an acquired autoimmune and an inherited subepidermal blistering disease.J Clin Invest 90:1628–1633, 1992

Giudice GJ, Emery DJ, Diaz LA: Cloning and primary structural analysis of thebullous pemphigoid autoantigen, BP180. J Invest Dermatol 99:243–250, 1992

Giudice GJ, Emery DE, Zelickson BD, Anhalt GJ, Liu Z, Diaz LA: Bullouspemphigoid and herpes gestationis autoantibodies recognize a common non-collagenous site on the BP180 ectodomain. J Immunol 151:5742–5750, 1993

Giudice GJ, Liu Z, Diaz LA: An animal model for bullous pemphigoid. What canit teach us? Proc Assoc Am Phys 107:237–241, 1995

Hintner H, Tappeiner G, Honigsmann H, Wolff K: Lichen planus and bullouspemphigoid. Acta Derm Venereol (Stockh) 59 (Suppl.):71–76, 1979

Kaposi M: Lichen ruber pemphigoides. Arch Dermatol Syphilol (Berl) 24:343–346, 1892Kippes W, Schmidt E, Roth A, Rzany B, Brocker EB, Zillikens D:

Immunopathological changes in 115 patients with bullous pemphigoid.Hautarzt, 1999, in press

Labib RS, Anhalt GJ, Patel HP, Mutasim DF, Diaz LA: Molecular heterogeneity ofthe bullous pemphigoid antigens as detected by immunoblotting. J Immunol136:1231–1235, 1986

Lang PG, Maize JC: Coexisting lichen planus and bullous pemphigoid or lichenplanus pemphigoides. J Am Acad Dermatol 9:133–140, 1983

Li K, Tamai K, Tan EML, Uitto J: Cloning of type XVII collagen. J Biol Chem268:8825–8834, 1993

Liu Z, Diaz LA, Troy JL, Taylor AF, Emery DJ, Fairley JA, Giudice GJ: A passivetransfer model of the organ-specific autoimmune disease, bullous pemphigoid,using antibodies generated against the hemidesmosomal antigen, BP180. J ClinInvest 92:2480–2488, 1993

Masunaga T, Shimizu H, Yee C, Borradori L, Lazarova Z, Nishikawa T, Yancey B:The extracellular domain of BPAG2 localizes to anchoring filaments and itscarboxyl terminus extends to the lamina densa of normal human epidermalbasement membrane. J Invest Dermatol 109:200–206, 1997

Matsumura K, Amagai M, Nishikawa T, Hashimoto T: The majority of bullouspemphigoid and herpes gestationis serum samples react with the NC16a domainof the 180-kDa bullous pemphigoid antigen. Arch Dermatol Res 288:507–509, 1996

Murakami H, Hashimoto T, Bhogal BS, et al: Analysis of cicatricial pemphigoidantigens. J Dermatol Sci 17:39–44, 1998

Nie Z, Hashimoto T: IgA antibodies of cicatricial pemphigoid sera specifically reactwith the C-terminus of BP180. J Invest Dermatol 112:254–255, 1999

Ogg GS, Bhogal BS, Hashimoto T, Coleman R, Barker JN: Ramipril-associatedlichen planus pemphigoides. Br J Dermatol 136:412–414, 1997

Okada N, Kitano Y, Miyagawa S, Sakamoto K, Steinberg ML: Expression ofpemphigoid antigen by SV-40 transformed human keratinocytes. J InvestDermatol 86:399–401, 1986

Perriard J, Jaunin F, Favre B, Budinger L, Hertl M, Saurat J-H, Borradori L: IgGautoantibodies from bullous pemphigoid (BP) patients bind antigenic sites onboth the extracellular and the intracellular domains of the BP antigen 180.J Invest Dermatol 112:141–147, 1999

Stanley JR, Hawley-Nelson P, Yuspa SH, Shevach EM, Katz SI: Characterization ofbullous pemphigoid antigen: a unique basement membrane protein of stratifiedsquamous epithelia. Cell 24:987–903, 1981

Stanley JR, Tanaka T, Mueller S, Klaus-Kovtun V, Roop D: Isolation ofcomplementary DNA for bullous pemphigoid antigen by use of patients’autoantibodies. J Clin Invest 82:1864–1870, 1988

Stingl G, Holubar K: Coexistence of lichen planus and bullous pemphigoid. Animmunopathological study. Br J Dermatol 93:313–320, 1975

Tamada Y, Yokochi K, Nitta Y, Ikeya T, Hara K, Owaribe K: Lichen planuspemphigoides: Identification of 180 kd hemidesmosome antigen. J Am AcadDermatol 32:883–887, 1995

Tanaka T, Korman NJ, Shimizu H, Eady RAJ, Klaus-Kovtun V, Cehrs K, StanleyJR: Production of rabbit antibodies against carboxy-terminal epitopes encodedby bullous pemphigoid cDNA. J Invest Dermatol 94:617–623, 1990

Willesteed E, Bhogal BS, Das AK, Wojnarowska F, Black MM, McKee PH: Lichenplanus pemphigoides: a clinicopathological study of nine cases. Histopathology19:147–154, 1991

Zillikens D, Giudice G: BP180/type XVII collagen: Its role in acquired and inheriteddisorders of the dermal-epidermal junction. Arch Dermatol Res 291:187–194, 1999

Zillikens D, Giudice GJ, Diaz LA: Bullous pemphigoid. An autoimmune blisteringdisease of the elderly. J Geriatr Dermatol 4:35–41, 1996a

Zillikens D, Kawahara Y, Ishiko A, et al: A novel subepidermal blistering diseasewith autoantibodies to a 200-kD antigen of the basement membrane zone.J Invest Dermatol 106:1333–1338, 1996b

Zillikens D, Rose PR, Balding SD, Liu Z, Olague-Marchan M, Diaz LA, GiudiceGJ: Tight clustering of extracellular BP180 epitopes recognized by bullouspemphigoid autoantibodies. J Invest Dermatol 109:573–579, 1997a

Zillikens D, Mascaro JM, Rose PR, et al: A highly sensitive enzyme-linkedimmunosorbent assay for the detection of circulating anti-BP180 autoantibodiesin patients with bullous pemphigoid. J Invest Dermatol 109:679–683, 1997b

Zone JJ, Taylor TB, Meyer LJ, Petersen MJ: The 97 kDa linear IgA bullous diseaseantigen is identical to a portion of the extracellular domain of the 180 kDabullous pemphigoid antigen, BPAg2. J Invest Dermatol 110:207–210, 1998