0022-202X/84/8202-0 139$02.00/ 0 THE JOURNAL O F INVESTIGATIVE Dlli!MATO LOGY, 82:139- 144 , 1984 Vo l. 82, No.2 Copyri ght © I 984 by The Williams & Wilkins Co. Printed in U.S.A. Differentiating Anti-Lamina Lucida and Anti-Sublamina Densa Anti­ BMZ Antibodies by Indirect Immunofluorescence on 1.0 M Sodium Chloride-Separated Skin

W. RAY GAMMON, M.D., ROBERT A . BRIGGAMAN, M.D. ALFRED 0. INMAN ,III, M.S., LAURINDA L. QUEEN, M .D. AND CLAYTON E. WHEELER, M.D. Department of Dermatology, University of North Carolina School of Medicine, Chapel Hill, North Carolina, U.S.A.

Sixty-one bullous disease sera containing lgG anti­ in EBA, t hey bind along t he lower edge or just below t he lamina BMZ antibodies were examined by indirect immunoflu­ densa (a nti-sublamina densa antibodies) [5-12]. Anti-lamina orescence on intact skin and skin separated through the Iucida and anti-sublamina densa ant ibodies cannot be reli ably lamina Iucida by incubation in 1.0 M NaCl. All sera distinguished using standard direct and indirect immunofl uo­ produced an indistinguishable pattern of linear immu­ rescence procedures and substrates because of t heir identical nofluorescence on intact skin at dilutions of 1:10 or staining patterns. higher. On separated skin, antibodies bound to e ither Due to the limited availability of immunoelectron micros­ the epidermal (epidermal pattern), dermal (dermal pat­ copy, t he ultrast ructural binding sites of anti-BMZ antibodies tern), or epidermal and dermal (combined pattern) sides are seldom defined in bullous diseases, and differential diag­ of the separation. The binding patterns were consistent nosis is based primarily on the clinical and hi stologic features on separated skin from several donors and titers of anti­ of t he diseases. T hese features may be misleading, especially in basement membrane zone antibodies on separated skin patients with BP, CP, and EBA in which features considered were comparable to those on intact skin. Sera from 3 characteristic of one disease may be present in another. Since patients with herpes gestationis (HG), 36 patients with BP and CP a re characterized by anti-lamina Iucida antibodies bullous pemphigoid (BP), and 1 patient with clinical and and EBA by anti-sublamina densa antibodies, precise localiza­ histologic features of epidermolysis bullosa acquisita tion of anti-BMZ antibody binding sites could be of value in (EBA) showed an epidermal pattern. Sera from 9 pa­ differential diagnosis. Furthermore, routine or more frequent tients with BP showed a combined pattern and sera from determination of antibody binding sites could provide a better 6 patients with EBA and 6 patients with clinical and understanding of t he cl inical, pathologic, a nd immunoul tra­ histologic features of BP showed a dermal pattern. In­ structural correlations in all bullous diseases associated with direct immunoelectron microscopy of selected sera ant i- BMZ antibodies. s howed antibodies producing the epidermal and com­ In search of an alternative to immunoelectron microscopy bined patterns were anti-lamina Iucida antibodies and for differentiating between anti-lamina Iu cida and anti-sublam­ those producing the dermal pattern were anti-sublamina ina densa antibodies, we have investigated indirect immunof1u­ densa antibodies. These results show indirect immuno­ orescence using NaCl-treated human skin as substrate. One fluorescence on separated skin is a dependable method molar NaCI has previously been shown to separate the BMZ for differentiating bullous disease anti-lamina lucida through the lamina Iucida leaving BP antigen on t he epidermal and anti-sublamina densa antibodies and that differen­ side and components of t he lamina densa, type IV coll agen, on t iating between the antibodies is essential for accurate the dermal side [1 3]. Using t his substrate, we examined sera diagnosis in some patients. The results also suggest BP containing a nti-BMZ antibodies from 61 patients t hought to anti-lamina Iucida antibodies may have more than one have BP, HG, or EBA based on clinical, hi stologic, and routine a ntigenic specificity. immunohistologic criteria. The results showed that anti-lamina Iucida antibodies from patients with BP and HG bound to t he Circulating and tissue-deposited lgG antibodies to t he cuta­ epidermal or both epidermal a nd dermal sides of separated skin neous basement membrane zone (BMZ) are characteristic of while anti-sublamina densa antibodies from patients with EBA patients with bullous pemphigoid (BP), cicatri cia l pemphigo id bou nd only to the derma l side. Of interest was t he fi nding that (CP), herpes gestationis (HG), and epidermolysis bullosa ac­ 6 patients thought to have BP and 1 patient thought to have quisita (EBA) [1-8]. By immunoelectron microscopy, anti­ EBA by routine clinical, histologic, a nd direct immunoflu ores­ BMZ antibodies can be divided into two groups based on t heir cence criteri a had anti-subla mina densa and anti-lamina Iu cida ultrastructural binding sites. In BP, CP, and HG t he antibodies antibodies, respectively. In addition, 9 BP sera had anti-lamina bind to the lamina Iu cida (anti-lamina Iucida a ntibodies) and Iucida antibodies which consistently bound to both epidermal and dermal sides of separated skin while 36 BP sera had a nt i­ Manuscript received July 1, 1983; accepted for publication Septem­ lamina Iu cida a ntibodies which consistently bou nd only to t he ber 29, 1983. epidermal side. This work was supported by NIH grants 5-R01 -A M30475-02, 5-K04- AM00956-02, 5-T32-AM07369-04, and 5- R01-AM10546- 18. MATERIALS AND METHODS Reprin t requests to: W. Ray Gammon, M.D. , Department of Der­ matology, Room 137, North Ca rolina Memorial Hospital, Chapel Hill, I mm.u.noreagents North Carolina 275 14. Abbreviations: Rabbit. anti-laminin serum (specific immunoglobu lin concentration BMZ: basement membrane zo ne = 0. 25 mg/ml) wa s purchased from Bethesda Research Labs, Ga ithers­ BP: bullous pemphigo id burg, Maryland and used at a di lution of 1: 100 in 0.15 M NaC l buffered CP: cicatricial pemphigoid with 0.01 M Na 1HP0.1 and 0.1 M NaH1 PO.,, pH 7.2 (PBS). Fluorescein EBA: epidermolys is bu ll osa acq ui sita isothiocyanate (FITC)-conjugaLed fgG fraction of goat anti human fgG FITC: flu orescein isothiocyanate and ant.irabbit lgG were purchased from Cappel Labo ratories, Coch­ HG: herpes gestationis ranville, Pennsylvania. Spec ifi c antibody and mo lar flu oresce in t.o NHS: normal human serum protein ratios of these reage nts were 2. 1 and 2.6, respectively for 140 GAMMON ET AL Vol. 82, No.2 antihuman IgG a nd 3.0 a nd 2.8 fo r a ntirabbit IgG. T hese conjugates results were similar in all 3 specimens. In each, dermis and were used at a dilution of 1:10- 1:20 in PBS. Bullous pemphigoid anti­ e~idermis were separated primarily through the lamina Iucida lamina Iucida and EBA anti-sublamina densa reference sera were w1th recogmzable basal cell plasma membrane on t he epidermal obtained from 2 patients diagnosed by direct lgG immunoelectron s1de and lamma densa_ on_ the dern~al_ s1de. The_ lamina densa microscopy. Specificity of these anti-BMZ antibodies for the lamina Iucida and sublamina densa zone was confirmed by indirect lgG im ­ showed some foca l t hmnmg and mfrequent disruptions but munoelectron microscopy at a dilution of 1:10. Titers of lgG anti-BMZ only rarely were fragments of lamina densa adherent to the antibodies by indirect immunofluorescence on intact human skin were epidermal side of the separation. T he basal cell plasma mem­ 1:1280 for BP and 1:320 for EBA. Reference sera were used at a dilution brane also showed degenerative changes and interruptions but of 1:10 in PBS. was only rarely seen attached to the dermal side of the sepa­ ration. Portions of lamina Iucida could be seen adherent to Bullous Disease and Control Sera both plasma membrane and lamina densa. Control normal human sera (NHS) and bullous disease sera were obtained from 8 normal volunteers a nd 61 bullous disease patients with Localization of BMZ Antigens in NaCl-Separated Skin circulat ing lgG anti-BMZ antibodies at a t iter of at least 1:10 by Frozen specimens of separated skin from each of t he 3 donors standard indirect immunofluorescence testing. Bullous disease patients were diagnosed as BP (51), HG (3), and EBA (7) by accepted clini cal, were examined by indirect immunofluorescence using antibod­ histologic, and im munohistologic features. Six patients with EBA and ies to laminin and reference sera to BP and EBA ant igens. The 3 with BP were also diagnosed by direct lgG immunoelectron micros­ copy. Most patients were diagnosed by one of the authors; however, several sera were generous gifts from Drs. E. H. Beutner, L. A. Diaz, S. 1. Katz, and W . M. Sams, Jr. All sera were heat-inactivated at 56· c for 30 min a nd stored frozen at -7o· c.

Shin Strips of adult human skin approximately 2.0 em wide and 10.0 em long were obtained from 3 fresh cadavers or surgical patients wi thout skin disease and subcutaneous fat scraped away. Skin was rinsed in PBS and a 1 cm1 piece was cut from each strip, quick -frozen in liquid

N 2, and stored frozen at -7o· c. T hese pieces were used as intact skin. Sodium Chloride Separation of Shin The remainder of the s kin strips were treated with 1.0 M NaCl using a previously described method [14 ]. Briefly, each strip was immersed in 50.0 ml of cold (4.C) 1.0 M NaCI in a 50-ml capped polystyrene tube and slowly rotated for 72 h at 4 ·c. Sodium chloride was exchanged at 24 a nd 48 h. At the end of 72 h, strips were blotted on fil ter paper, placed epidermis side up in Petri di shes, and t he epidermis gently dislodged from the dermis by lateral traction appli ed to the skin with a tongue blade. T he epidermis was left in place and strips cut into LO­ cm2 pieces with a razor blade. Some pi eces were qui ckly frozen in liquid

N 2, mounted in Ames O.C.T. compound (Ames Co., E lkhart, Indiana), a nd stored frozen at -7o· c. Several pieces were immediately fixed in 2.5 % gluteraldehyde in 0.2 M Soresen's NaPO, buffer, pH 7.4

Electron Microscopy Glutaraldehyde-fixed pieces of each of the 3 strips of NaCI-separated skin were embedded a nd processed for routine electron microscopy according to previously described methods [15].

I ndirecL I mmtmo(luorescence Indirect immunoflu orescence was performed on intact and separated human skin using established methods [1 6] . Specimens of separated s kin from each of the 3 donors were examined to localize laminin and antige ns reactive with t he BP and EBA reference sera. All 61 bullous disease se ra we re examined by lgG indirect immunofluorescence on intact and separated skin from l donor at a dilution of 1:10. Thirty­ eight of the BP-labeled sera and all HG- and EBA-labeled se ra were examined on separated skin from 3 donors at a titer of 1:10. Twenty­ one of the BP- and EBA-Iabeled sera we re titered by doubling dilutions with PBS from 1:10 to 1:20,480 a nd examined on one specimen of separated skin.

Indirect lmmu.noelecLron Microscopy Indirect immunoelectron microscopy was performed on selected BP­ and EBA-labeled sera at a titer of 1:10 using 10- 12 l'm-thick cryostat sections of frozen intact or sepa rated skin and a previously described peroxidase-antiperoxidase method (6] .

RESULTS FIG 1. a, Epidermal pattern of immunol1uorescence on LO M NaCl­ Ultrastructure of NaCl-Separated Skin separated sk in produced by most BP and all HG sera (X250). b, Combined epidermal and dermal pattern of immonotluorescence on Glutaraldehyde-fixed specimens of separated skin from each separated skin produced by a few BP sera (X250). c, Dermal pattern of of 3 donors were examined by electron microscopy to determine immunofluorescence on separated skin produced by HBA- and BP­ the site of separation and preservation of BMZ structures. T he labeled anti-sublamina densa antibodies (X400). Feb. 1984 ANTI-BMZ ANTIBODIES STUDIED ON NACL-SEPARATED SKIN 141

results of all 3 specimens were identical. Antilaminin produced summarized in Table I. Of interest was the remarkably low a sharp uninterrupted linear band of fluorescence along the background fluorescence observed with separated skin com­ dermal edge of the separation with no staining on the epidermal pared to intact skin. side. Antibody to EBA antigen also produced a sharp, uninter­ To determine the reproducibility of the staining patterns, 27 rupted linear band of fluorescence along the dermal edge of the sera producing the epidermal pattern (23 labeled BP, 1 labeled separation which was slightly brighter and wider than that seen EBA, and 3 labeled HG) 9 sera producing the combined pattern, with antilaminin but otherwise indistinguishable. Antibody to 12 sera producing the dermal pattern (6 labeled BP, 6 labeled BP antigen produced a sharp ge nerally continuous band of EBA), and 8 NHS were examined on separated skin from each fluorescence along the epidermal edge of the separation. Occa­ of 3 donors. On all 3 specimens, t he patterns of flu orescence sionally, irregularities and small interruptions in BP staining were consistent for each bullous disease sera and no flu ores­ were seen but in no case was fluorescence seen on the dermis. cence was observed with NHS. To determine t he sensit ivi ty of separated skin as a substrate Indirect IgG Immunofluorescence of Bullous Disease and NHS for detecting anti-BMZ antibodies and further analyze BP­ on Intact and Separated Shin labeled sera producing the combined pattern of fluorescence, Each of the 61 bullous disease and 8 NHS was examined at 21 BP- and EBA-Iabeled sera were titered in doubling dilutions a titer of 1:10 by indirect immunofluorescence on intact and from 1:10 to 1: 20,480 on in tact and separated skin. These results separated skin fro m 1 donor. All 61 bullous disease sera pro­ are shown in Table II. Of 5 BP-Iabeled sera producing the duced a linear pattern of flu orescence at t he BMZ on intact epidermal pattern, titers on separated skin were equal to or 1 skin. On separated skin bullous disease sera produced a linear dilution greater than on intact skin. Of 4 BP-Iabeled sera pattern of fluorescence either along the epidermal side (epider­ producing the dermal pattern, titers on separated skin were mal pattern), dermal side (dermal pattern), or both epidermal consistently 1- 2 dilutions greater than on intact skin. Of 4 and dermal sides (combined pattern) of t he separation (Fig 1) . EBA-labeled sera producing the dermal pattern, t iters were 1- Of 51 BP-labeled sera, 36 (71 %) showed an epidermal pattern, 2 dilutions greater on separated skin. These results show sep­ 9 (17%) showed a combined pattern, and 6 (12%) showed a arated skin is at least as sensitive a substrate as intact human dermal pattern. All 3 HG-labeled sera showed an epidermal skin for titering anti-BMZ antibodies. In no case did diluting pattern. Of 7 EBA-labeled sera, 1 showed an epidermal pattern sera cause a change in pattern from epidermal to dermal or and 6 showed a dermal pattern. None of the NHS showed vice versa. fluorescence on intact or separated skin. These results are Of 8 sera producing the combined pattern, dermal and epi­ dermal staining ant ibody titers were the same in 3, higher on epidermis in 4, and higher on dermis in 1. In several cases, the TABLE I. Resu.lts of lgG 1:n.direc t ,:mmun.ofluorescence of bul.lous titer of ant i-BMZ on either the epidermal or dermal sides of disease and control sera on sodium chloride-separated skin t he separation was equal to or greater than titers in intact skin. Total Serum Ep idermal Combined Dermal no. groupa,b patte rn pattern pattern sera Indirect IgG Immunoelectron. Microscopy Bullous pemphigoid 51 36 (71%) 9 (17%) 6 (12 %) To correlate immunoflu orescent patterns on separated skin Herpes gestationis 3 3 (14 %) 0 0 with ultrastructural localization of antibody binding sites, se­ Epidermolys is bullosa 7 1 (14 %) 0 6 (86%) lected sera were examined by indirect lgG immunoelectron acquisita microscopy on separated or intact skin. The resul ts are shown Normal human 8 0 0 0 in Table TIL Two BP and 1 EBA-labeled sera producing an a Sera diluted 1:10 in PBS. epidermal pattern on separated skin showed IgG deposits on b Diagnosis based on clinical, hi stologic, and routine direct. immu ­ the epidermal side of the separation within the lamina Iucida. nofluorescence criteria. These sera produced no deposits on the dermis. Two BP sera

TABLE II. !gG an.ti-BMZ antibody titers of bu.llau.s dL~ ease sera on. intact and ~odium chloride-separated skin Titers Patterns Disease on Se rum Separated skin groupn separated sk in In tact sk in Epidermal Dermal Bullous pemphigo id Epidermal on ly BEl 10,240 20,480 LYN 5,128 10,240 MIL 10,240 10,240 WIL 10,240 20,480 MAR 5,128 10,240 Bullous pemphigo id Combined epi- FAL 20 40 40 dermal-der- MECI 80 320 80 mal CAN 40 320 80 FOR 160 320 320 WAL 640 320 40 ROL 10 40 80 BOL 20 20 20 WAT 40 40 20 Bullous pemphigoid Dermal only DKI 320 1280 DEV 40 80 CONL 160 320 CON 40 80 Epidermolysis bullosa Dermal MAN 320 1280 acquisita 560 80 160 815 80 160 CHO 40 80

a Diagnosis based on clinical, hi stologic, and direct immunofluorescence criteria. 142 GAM MON ET AL Vo l. 82, N o.2

TABL E Ill. Correlation. of ultrastruc tural binding sites and binding diffe rent iated from EBA anti-sublamina densa ant ibodies by patterns of sodium chloride-separated skin. binding patterns o n separated skin. Patterns Since anti-lamina Iucida antibodies we re not ava ilable fr orn on No. sera Ultrastructural Group" bi ndin g a patient with classical CP, we cannot be certain they could be sepa rated tested differentiated b y t he method. However, it is likely that Cp skin site a n t ib o d~ es would. produce a pattern. lik.e ~ h at ~ee n wit h BP a n d Bullous pem ­ Epidermal 2 L amina lucida H G ant tbodtes smce all three are md1stmgUtshable by irnrnu­ phigo id noelectron microscopy [7- 12]. The value of t he method i s of Combined epi- 2 Lamina Iucida course, limited by the in cidence of a nt i-BMZ ant ibodies. In 'cp dermal-der- the incidence is probabl mal y less than 20%; however, in EBA a nd HG it a ppear Dermal 4 Subbasal lamina s to be at least 50% and greater than 70% in Bp Epidermolysis Epidermal 1 Lamina Iucida [3,4,6,17-21] . In patients wi thout serum antibodies, a dou ble bullous ac­ immunofluorescence method has been d escribed f or differen­ qui sita tiating between in vivo deposited anti-lamina Iucida and a n ti­ Dermal 5 Subbasal lamina sublamina densa antibodi es [22 ]. "Diagnosis based on clinical, histologic, and rout ine immunoflu ores­ One molar NaCI-separated skin appeared to be at least as cence criteri a. sensitive as intact skin for d etecting a nti-BMZ antibodies. I n some cases, end-point t iters were 1-2 dilutions g reater in sep­ arated skin. This a pparent increase in sensitivity may have with a co mbined pattern also showed deposits on t he epidermal been due to improved exposure of antigen b y t he sepa ration side of separated skin but in addi tion showed d eposits on that process and/ or the low background fluorescence. The low b ack. portion of the lamina Iucida which h ad r emained attached to ground l1uorescence is presumably due to removal of t issue the lamina densa (Fig 2). Four BP-labeled sera and 5 EBA sera immunoglobulin during t he NaCl incubation. The sensit ivity producin g a de rmal pattern showed an identical deposition of of separated skin also suggests there is li ttle, if a ny, loss of or im munoreactant o n and just beneath the lamina densa (Fig 3) . alteration in BP or EBA antigens during the separation process. No deposits were seen in t he lamina Iucida; however, occasion­ Since the materials and reagents are a vail able to any immu­ ally deposits we re seen b eneath the basal l amina, anchoring noflu orescence laboratory, t he method can b e used in t he fibril complex. These resul ts confirmed the ul trastructural routine e valuation of anti-BMZ antibody binding s ites. Once bindin g sites of a nti-BMZ ant ibodies producing t he epidermal, skin is separated, it can b e stored at - 7o·c and in our experi­ dermal, and combined staining patterns on separated skin and ence is sui table as substrate for at least 3 months. Al t hough we showed t hat the "EBA serum" with anti-lamina Iucida antibod­ used electron microscopy to document t he site of separation ies and t he BP-labeled sera wi th anti-sublamina densa ant ibod­ this can be easily confirmed by immunoflu orescence alone u sin~ ies were fro m patients who were inco rrectly diagnosed. reference bullous disease sera or commercially availa ble a n t i~ bodies to Jaminin. The fact that DISCUSSION laminin is a lamina lucida co nstituent a nd separates to t he dermal side o f 1 M NaCJ­ Im munoelectron microscopy is the o nly method currently treated skin does raise the possibility t hat bullous disease anti­ avail able for precisely determining t he ultrastructural binding lamina Iucida antibodies which produce a dermal pattern m icrht sites of ant i-BMZ antibodies. However, it is technically diffi ­ be encountered [1 3,23]. However, t his was not observed in ~ur cul t, t ime-consuming and of limited availability to practicing series of 49 bullous disease sera containing a nt i-lamina lucida and academic dermatologists. Because of these limitations, we antibodi es. Ant ilaminin autoantibodi es have not been r eported examined sk in separated t hrough the lamina Iucida by 1 M in bullous disease but e ven if they were, they should be distin­ NaCl as a substrate for differentiating between anti-lamina gu ishable by t heir binding to vascular as well as cutaneous Iucida and a nti-sublamina densa antibodies by indirect immu­ basement membranes [23). nol1uo rescence. The resul ts showed that HG and BP anti­ Discriminating between anti-sublamina densa and ant i-lam­ lamina Iucida antibodies co uld be reliably a nd rep roducibly ina Iucida a ntibodies may be more important t han previously

F1r. 2. Im mu noelectron m ic rograph showing the dermal side of 1.0 M NaCI­ separated skin treated wi th co mbined pattern BP serum. Note immune depos­ its (D) on the sepa ration (S) side of the lam ina densa (L). Bar = 0.5 JLm .

; Feb. 1984 ANTI-BMZ ANTIBODIES STUDIED ON NACL-SEPARATED SKI N 143

FIG 3. lmmunoelectron micrographs showin g the BMZ of 1.0 M NaC I-sepa­ rat.ed skin treated with anti-sublamina densa antibody. Immune deposits (D) are present on and just beneath the lam­ ina densa (L ). The site of separation (S) is seen between the lamina densa and the basal cell plasma membrane (P). Bar = 0.5 IJm .

realized, part icula rl y in patients with BP, CP, and EBA. In t he "combined staining" sera have been examined by Dr. John these diseases overlapping clinical, histologic, and immuno­ Stanley, and shown to immunoprecipitate a 220,000 dalton histologic features occur. For example, patients with EBA have protein identical to t hat immunoprecipitated by other BP sera been described with features considered characteristic of BP (personal communication) [33 ]. T his would suggest t hat all BP [24-26]. These include a genera lized inflammatory bullous di s­ anti-lamina Iucida an tibodies share an identical specificity a nd ease with flexura l involvement, spontaneous blisters, healing t hat t hose producing t he combined pattern may have addit ional of some lesions without scarring or milia, a nd inflammatory specific it ies. subepidermal blisters. Patients have been reported as BP wi t h REFERENCES mechanical fragility of skin, trauma-induced blisters, and cu­ 1. Jordon RE, Beutner EH , Witebsky ET, Blumental G, Hale WL. taneous and mucous membrane lesions t hat hea l with scarring Lever WF: Basement zo ne antibodies in bullous pemphigoid. and milia [27- 29 ]. Both EBA and CP are diseases which JAMA 200:91- 96, 1967 2. Jordon RE, Triftshauser CT, Schroeter AL: Direct immunofl u· typically heal with scarring and both may have mucous mem­ orescent studies of pemphigus and bullous pemphi goid. Arc h bran e lesions including lesions of ocular mucous membranes Dermatol 103:486- 591, 1971 (17 ,30-32]. It is conceivable t hat failure to define the ult rastruc­ 3. Bean SF: Cicatricial pemphigoid. Arch De rm atol110:552-555, 1974 t ural locali zation of anti-BMZ ant ibodies in t hese diseases may 4. Lawley TJ, Sting! G, Katz SI: Fetal and matern al ri sk factors in herpes gestationis. Arch Dermatol 114:552- 555, 1978 have led to misdiagnoses a nd to the proposals t hat CP and 5. Nieboer C, Boorsma OM, Woerdeman MJ, Kalsbeek GL: Epider­ EBA may be the same disease and t hat BP and CP may evolve molysis bullosa acquisita: immunoiluorescence, electron micro­ into EBA [31,32]. scopic and immunoe lectron microscopic studi es in four patients. The results of t his study showed that misdiagnoses can be Br J Dermatol 102:383-392, 1980 6. Yaoita H, Briggaman RA , Law ley TJ, Provost TT, Katz Sl: Epi· made if t he ultrastructural locali zation of anti-BMZ a nt ibodies dermolysis bullosa acquisita: ultrastructural and immunological is not determined. Sera from 6 pat ien ts suspected of having BP studi es. J In vest Dermato l 76:288- 292, 1981 a nd 1 suspected of having EBA on t he basis of clinical, histo­ 7. Holubar K, Wolff K, Konrad K, Beutner EH: Ultrastructural logic, a nd rout ine immunofluorescence criteria were shown to loca li zation of immunoglobulins in bullous pemphigoid skin: employment of a new peroxidase-antiperoxidase mul tistep con tain anti-sublamina densa and anti-lamina Iucida antibod­ method. J In vest Dermatol 64:220- 227, 1975 ies, respectively. The patients init ially thought to have BP will 8. Schaumburg-Lever G, Rule A, Schmidt-U llrich B, Lever WF: Ul­ be discussed in more detail in a separate communication but, trastructural loca li zation of in vivo bound immunoglobulins in in b rief, t hey presented eit her with clinical features indistin­ bullous pemphigoid. A preliminary report. J Invest Dermatol 64:47- 49, 1975 guish able from classic BP or combined features of both BP and 9. Honigsmann H, Stin g! G, Holubar K, Wolff-Schreiner E, Konrad EBA. Because these patients have immunoelectron microscopic K, Wolff K: Autoa ntibodies and immune co mpl exes in immu­ features indistinguishable from classic EBA, we currently re­ nodermatoses. Mapping of fin e structural binding sites (abstr). gard t hem as being a variant of t hat disease. In t he 1 patient J Invest Dermatol 66:263, 1976 10. Albini B, Holubar K, Shu S, Wolff K: Enzyme antibody methods with clinical and histologic fea tures of EBA and a nti-lamina in immunode rmato logy, Immunopathology of the Skin , 2d ed. Iucida antibodies t he diagnosis is not clear but the locali zation Edited by EH Beutner, TP Chorze lski , SF Bean. New York, of antibodies to t he la mina Iucida suggests a diagnosis of CP. J ohn Wiley & Sons, 1979, pp 93-95 T he patient had a scarring, blistering eruption localized to 11. Honigs mann H, Sting! G, Holubar K, Wolff 1<: Herpes gestationis: fin e structural pattern of immunoglobulin deposits in the sk in t rauma-susceptible sites of skin and oral mucous membranes in vivo. J Invest Dermato l 66:389- 392, 1976 but without ocula r involvement. 12. Harringto n CI, Blee hen SS: Herpes gestationis: immunopatholog­ In t his study, 9 (20%) of 45 BP sera produced a combined ical and ul trastru ctura l studies. Br J Derma to! 100:389- 399, 1979 pattern of immunofluorescence on separated skin from several 13. Woodley 0, Sauder D, Talley MJ, Silver M, Grotendorst G, Qwarn · strom E: Loca li zation of basement membrane co mponents after donors. These results suggest some BP sera have anti-BMZ dermal-epidermal junction separation. J Invest Dermatol specificities different from or in addition to t hose BP sera 81:149- 153, 1983 producing only epidermal deposits on separated skin. Two of 14. Scaletta LJ, Occhino JC, MacCallum OK, Lillie JH: Iso lation and 144 GAMMON ET AL Vol. 82, No_ 2

immunologic identification of basement membrane zone a nti­ 23. Foidart JM, Bere EW_Jr, Yaar M,_R ennard SI, Gulli noM, Martin gens from human skin. Lab Invest 39:1- 9, 1978 GR, Katz S l: D1stnbut1on and 1mmunoelectron microscopic lo­ 15. Briggaman RA , DalldorfFG, Wheeler CE Jr: Formation and origin ca li zation of laminin, a noncollagenous basement membrane of basal lamina a nd anchoring fibril s in adult human skin. J Cell glycoprotein. Lab Invest 42:336-342, 1980 Bioi 51:384-395, 1971 24. Gammon WR, Briggaman RA, Wheeler CE Jr: Epidermolysis 16. Beutner EH, Nisengard RJ: Defined immuno11uorescence in clini­ bullosa acquisita presenting as an inflammatory bullous disease cal im munopathology, Immunopathology of t he Skin: Labeled J Am Acad Dermatol 7: 382-387, 1982 · Antibody Studies. Edited by EH Beutner, TP Chorzelski, SF 25. Provost TT, Maize JC, Ahmed AR, Strauss J S, Dobson RL: U nu­ Bean, RE Jordon. Stroudsburg, Penna, Dowdon, Hutchinson & sual subepidermal bullous diseases with immunologic features of Ross, 1973, pp 197- 247 bullous pemphigoid. Arch Dermatol115:156- 160, 1979 17. Griffit h MR, Fukuyama K, Tuffanelli D, Si lverman S: Immunollu­ 26. Capu to R, Berti E, Monti _M: Pemphigoide bulleuse: a type d 'epi­ orescent studies in mucous membrane pemphigoid. Arch D er­ dermolyse bulleure acqlllse. Les Gournees de Dermatologiques a matol 109:195-199, 1974 Paris 114:1127, 1982 18. Rogers RS LII , Perry HO, Bean SF, Jordon RE: Immunopathology 27. Person JR, Rogers RS Ill: Bullous pemphigoid responding to of cicatricial pemphigoid: studies of complement depos1t10n. J sulfapyridine and the sulfone. Arch Dermatol 113:610- 615, 1977 Invest Dermatol 68:39-43, 1977 28. Chorzelski TP, J ablonska J, Beutner EH: Pemphigoid, Immuno­ 19. Ahmed AR, Maize JC, Provost TT: Bullous pemphigoid clinical pathology of t he Skin: Labelled Antibody Studies. Edited by EH and immunologic foll ow- up after successful t herapy. Arch Der­ Beutner, TP Chorzelski , SF Bean, RE Jordon. Stroudsburg matol 113:1043- 1046, 1977 Penna, Dowdon, Hutchinson & Ross. 1973, pp 243- 255 ' 20. Hodge L, Marsden RA, Black MM, Bhogal B, CorbettMF: Bullous 29. Dahl MGC, Cook LJ: Lesions induced by !.rauma in pemphigoid pemphi goid: the frequency of mucosal involvement and concur­ Br J Dermatoll01:469-473, 1979 · rent malignancy related to indirect immunoflu orescence find - 30. Dantzig P: Circulating antibodies in cicatri cial pemphigoid. Arch ings. Br J Dermatol 105:65- 69, 1981 . Dermatol 108:264 - 266, 1973 21. Dabelsteen E Ullman S, Thomsen K, Rygaard J: DemonstratiOn 31. Palestine RF, Kossard S, Dicken CH: Epidermolysis bullosa ac­ of basemen't membrane autoantibodies in patients wit h b enign quisita: a heterogeneous disease. J Am Acad Dermatol 5:43-53 mucous membrane pemphigoid. Acta Derm Venereol (Stockh) 1981 ' 54 :189-192, 1974 32. Dahl MGC: Epidermolysis bullosa acquisita-a sign of cicatricial 22. Gammon WR, Robinson T, Briggaman RA, Whee ler CE: Double pemphigoid: Br J Dermatol 101:475- 484 , 1979 immunofluorescence microscopy: a method for localizing im ­ 33. Stanley JR, Hawley-Nelson P, Yuspa SH, Shevach EM, Katz SI· mune deposits in skin diseases associated with lamina basement Characterization of bullous pemphigoid antige n: a unique base: membrane zo ne immunofluorescence. J Invest Dermatol 79:312- ment membrane protein of stratified squamous epit heli a cell 317,1982 Ce ll 24:897- 903, 1981 ·