Detection of the 170-kDa Bullous Pemphigoid Antigen by Immunoprecipitation

Tamotsu Ebihara, Takashi Hashimoto, Jun Kudoh, Shinobu Gamou, Nobuyoshi Shimizu, and Takeji Nishikawa Departments of Dermatology (TE, TH, TN) and Molecular Biology OK, SG, NS), Keio University School of Medicine, Tokyo, Japan

There has been controversy concerning the nature of the bullous pemphigoid (BP) antigen: immunoprecipitation identified BP antigen as a single, unique 230-kOa protein, whereas immunoblot analysis showed multiple antigen mole­cules, mainly 230- and 170-kOa proteins. In this study, to further characterize the 170-kOa protein, we have examined whether the 170-kOa protein is detected by immunoprecipi­tation. Extracts of human squamous cell carcinoma cells re­vealed the 170-kOa protein with immunoblot analysis. Al­though the conventional immunoprecipitation detected only the 230-kOa protein, some BP sera that detected the 170-

Bullous pemphigoid (BP) is an autoimmune skin disease characterized by tense blisters due to cleft formation at the basement membrane zone (BMZ) [1]. In most cases, circulating autoantibodies against the epidermal basement membrane zone (BMZ) are detected [2] .

These autoantibodies are considered to play an important role in the pathogenesis of BP [3 -6]. There is uncertainty about the BP anti­gens that are defined immunologically by anti-BMZ antibodies. Some studies have suggested that BP antibodies recognize a single antigen [7 -10], whereas other studies have indicated the molecular heterogeneity in BP antigens [11-13]. In general, BP antigen was identified as a uni<J.ue 230-kDa protein by means ofimmunoprecipi­tation technique l7 -10]; however, immunoblot analysis showed several different BP antigens, including 230- and 170-kDa proteins [12-13].

To clarify this discrepancy, we examined whether the 170-kDa protein is detected by immunoprecipitation technique. Immuno­blot analysis of human squamous cell carcinoma cells detected both the 230- and 170-kDa proteins. Immunofluorescence study suggested that the two proteins are located in hemidesmosomes.

Manuscript received March 18, 1992; accepted for publication October 6, 1992.

Reprint requests to: Dr. Takashi Hashimoto, Department of Dermatol­ogy, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160, Japan.

Abbreviations: BMZ: basement membrane zone BP: bullous pemphigoid EDT A: ethylenediaminetetraacetic acid Ig: immunoglobulin(s) IlF: indirect immunofluorescence MoAb: monoclonal antibody P AGE: polyacrylamide gel electrophoresis PBS: phosphate-buffered saline, pH 7.4 PMSF: phenylmethylsulfonyl fluoride SDS: sodium dodecyl sulfate TBS: Tris-HCI- buffered saline, pH 8.0

kOa protein with immunoblotting also precipitated the 170-kOa protein with our modified immunoprecipitation, in which the cells were extracted with 1 % sodium dodecylsul­fate (SOS) buffer and reacted with the sera under reduced SOS concentration. The 170-kOa protein-specific BP sera clearly showed hemidesmosomal plaque staining with immu­nofluorescence of cultured cells. These results indicate that the 170-kOa protein is indeed one of the BP antigens and that the 230- and 170-kOa BP antigens are integrated in different ways in hemidesmosomes.] Invest DermatoI100:176-179, 1993

Furthermore, the 170-kDa protein was clearly demonstrated by our modified immunoprecipitation using 1 % sodium dodecyl sulfate (SDS) for cell extraction. These results suggest that the two BP antigens are present in different ways in hemidesmosomes.

MATERIALS AND METHODS

Sera Nine sera were obtained from clinically, histologically, and immunohistologically typical patients with BP (BP1-BP9). Six sera from healthy volunteers were used as normal controls. All sera were stored at-20 °C and at 4°C in the presence of 0.1 % NaN) for less than 1 month during these experiments.

Monoclonal Antibody (MoAb) A mouse MoAb, D-20, was a generous gift from Dr. K. Owaribe. The MoAb was produced by standard hybridoma technique using the SDS-polyacrylarnide gel electrophoresis (PAGE)-eluted 170-kDa protein of hemides­mosomal preparation obtained from bovine corneal epithelium [14,15].

Cell Line KU8 cell used in this study was a human squamous cell carcinoma cell line derived from a penile carcinoma. Char­acteristics and culture condition of KU8 cells were previously de­scribed [16].

Immunoblot Analysis Immunoblot assay using ethylenedia­minetetraacetic acid (EDT A) - separated normal human ef.idermal extracts was performed as described previously [12,13, except that peroxidase - conjugated anti-mouse immunoglobulins (Igs) (DAKO, Copenhagen, Denmark) was used as a second antibody for MoAb D-20. To detect the expression of BP antigens in KU8 cells by immunoblotting, we used a modification of the method for nor­mal human epidermis. Briefly, confluent cultures ofKU8 cells were scraped in 0.01 M Tris-HCI buffer, pH 6.8, containing 1% SDS, 5% p-mercaptoethanol, 2 mM EDTA, 2 mM phenylmethylsul­fonyl fluoride (PMSF), and 5 mg/liter each oHour proteinase inhibi­tors (Ieupeptin, antipain, chymostatin, and pepstatin A; Sigma Chemical Co., St. Louis, MO), boiled for 5 min, centrifuged at 10,000 X g for 40 min, and supernatant was harvested as an antigen

0022-202X/93/$06.00 Copyright © 1993 by The Society for Investigative Dermatology, Inc.

176

VOL. 100, NO. 2 FEBRUARY 1993 IMMUNOPRECIPIT A TION FOR 170-kDa BP ANTIGEN 177

B c

230 KD­ 230 KD-

I 230 KD-

170 KD - 170 KD-170 KD-

t f

BBBBBBBBBC PPPPPPPPPO 1 2 3 4 5 6 7 .8 9 N

T

B B D P P I

8 6 2 o

BBBBBBB PPPPPPP 3 4 5 6 789

Figure 1. Immunoblot analysis. (A - C) The positions of the 230- and 170-kOa proteins are indicated on the lift; bars on the right indicate the positions of standard molecular weight markers, from top to bottom: 200-, 116-, and 97-kOa. (A) Normal human epidermal extracts. (B) Normal epidermal extracts for MoAb 0-20. The MoAb showed a little weak but clear reactivity with the 170-kOa protein. (C) KUB cell extracts. BP3-BP6 specifically labeled doublet bands of approximately 170-kOa . (CONT) control.

source. SDS-PAGE and immunostaining of the blots were per­formed by the same method described previously [1 3].

Indirect Immunofluorescence (llF) To detect BP antigen ex­pression in cultured cells by IIF, KU8 cells were cultured on glass covers lips to near confluence, fixed with acetone at -20°C for 20 min, and extracted with 0.5% Triton X-I00 in phosphate-buf­fered saline (PBS), pH 7.4, for 15 min at room temperature. The coverslips were first blocked with PBS containing 1 % bovine serum albumin and 0.1 % NaN3 for 30 min at room temperature and then incubated with sera or MoAb appropriately diluted in the same buffer for 1 h at room temperature. After washing for 15 min with PBS, the coverslips were incubated with 1: 800 diluted fluorescein isothiocyanate-conjugated anti-human IgG rabbit antiserum or anti-mouse Igs rabbit antiserum (DAKO) for human sera or MoAb D-20, respectively , as second antibodies. After extensive washing w ith PBS, the covers lips were examined by fluorescence micro­scopy (Optiphot EFD2, Nikon, Tokyo, Japan).

In some experiments, the extraction step with 0.5% Triton X -IOO was omitted to determine the possibility that this procedure may remove some pool of BP antigen present in the cytoplasm.

IrnUlUnoprecipitation Analysis Immunoprecipitation was per­formed mainly according to the method described by Stanley et al [7]. Briefly, KU8 cells were radiolabeled with 35S-methionine (0.925 X 106 Bqjml) in methionine-free medium containing 10% fe tal calf serum and lysed with 0.5% Nonidet P-40 in the Tris-Hcl ­buffered saline (TBS), pH 8.0, containing 2 mM PMSF. After cen­trifugation at 10,000 Xg at 4°C for 40 min, the supernatant was dialyzed against 0.3% Nonidet P-40 and 2 mM PMSF in TBS over­night and stored at - 80°C as aliquots. To reduce nonspecific back­ground, the 200-lll antigen sample was first preabsorbed with 20 III of combined sera from three normal volunteers at 4 ° C overnight and precipitated with 60 III protein G sepharose slurry (Pharmacia, Uppsala, Sweden) at 4°C for 30 min. Fifteen microliters of preab­sorbed antigen sample was incubated with 3 III BP or normal sera or 20 ,ull : 20 diluted MoAb D-20 ascites at 4 °C overnight. Antigen­antibody complexes were precipitated by incubation with 3 ,ul pro­tein G sepharose slurry at 4°C fo~ 30 min. After the precipitates were washed six times with TBS containing 0.3% Nonidet P-40, 0.3% sodium deoxycholate, and 2 mM PMSF, protein G, Igs, and antigens were dissociated by boiling in Laemmli's sample buffer under the reducing condition (0.00625 M Tris-HCl, pH 6.8, con­taining 2% SDS and 5% p-mercaptoethanol), and protein G was

removed by centrifugation. After proteins were fractionated by SDS-PAGE, gels were dried and processed for fluorography .

To release proteins more extensively, we modified the extraction procedure according to the immunoprecipitation for desmocollins I and II described by Parrish et at [17] . Briefly, the cells were extracted with the same extraction buffer as that used for immunoblot analysis (i.e., Tris-HCI ~uffer suppleme~ted ~ith. 1.% SDS, 5% p-mercap­toethanol, and SIX different protemase mhlbltors). After centrifuga­tion at 10,000 X g at 4°C for 40 min, the supernatant was dialyzed at 4 °C overnight against TBS containing 0.1 % SDS and 2 mM PMSF to reduce the SDS concentration in the sample solution. Procedures fo llowing this were the same as those for the conven­tional immunoprecipitation.

RESULTS

Immunoblot Analysis Immunoblotting of the normal human epidermal extracts with the sera from BP patients BP1 and BP2 detected both the 230- and 170-kDa proteins, whereas the sera from patients BP3 - BP6 reacted only with the 170-kDa protein, and the sera from patients BP7 - BP9 reacted only with the 230-kDa pro­tein. However, the SIX control s~ra reacted with neither protein (Fig lA). MoAb D-20 detected a sll1gle 170-kDa protein band in the human epidermal extracts that showed the same mobility as the 170-kDa protem recogl1lzed by some BP sera (Fig IB). Immuno­blotting of the extracts of KU8 cells with the BP sera showed these two proteins in the same pattern, although the 170-kDa protein appeared to be doublet bands (Fig 1 C). This confirmed that both 230- and 170-kDa proteins are also expressed in KU8 cells.

IIF When KU8 cells were fixed with acetone and treated with 0.5% !riton X-~OO , all BP sera displayed numerous punctate arrays in various densities and patterns at the lower cell surface (Fig 2A,B). These staining patterns are characteristic of the hemidesmosomal plaque seen in cultured epithelial cells [15,18,19]; however, no apparent difference of reactivity was seen·between the BP sera reac­ting only with the 230-kDa protein on immunoblotting (Fig 2A) and the BP sera reacting only with the 170-kDa protein (Fig 2B). MoAb D-20 also exhibited the same pattern of reactivity (Fig 2D); however, none of the six control sera showed this positive staining (Fig 2C).

When the 0.5% Triton X-lOO extraction step was omitted, all BP sera sti ll showed specific hemidesmosomal plaque staining, al­though coarse, granu lar, nonspecific fluorescence increased slightly

178 EBIHARA ET AL

Figure 2. Immunofluorescence using acetone fixed- and Triton X-IOO treated-KUB cells. (A) BP3 serum. (B) BP7 serum. (C) Normal control serum did not show this specific staining. (D) MoAb 0-20 also showed the similar reactivity. Bar, 20 11m.

(data not shown); however, no apparent difference of reactivity was seen between the 230-kDa protein- specific BP sera and the 170-kDa protein - specific BP sera.

hnrnunoprecipitation The conventional immunoprecipitation revealed that all the BP sera reactive with the 230-kDa protein by immunoblotting precipitated the 230-kDa protein from the cell extract with 0.5% Nonidet P-40 buffer, although the 170-kDa protein was not precipitated by any sera (Fig 3A).

When cell extracts were prepared in the presence of 1 % SDS and dialyzed against 0.1 % SDS buffer, the 230-kDa protein was de­tected by all of the BP sera reactive with the 230-kDa protein by immunoblotting, although the 170-kDa antigen was not detected (data not shown) . This concentration ofSDS does not interfere with

A

230kD-- 230kD--170kD--

B

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

the binding of desmoglein I or desmocollins I and II to their specific MoAb [17] and does not interfere with the binding of the 230-kDa antigen to the autoantibody. To avoid the possibility that the bind­ing of the 170-kDa protein to its antibody is prevented by this SDS concentration, we further diluted the dialyzed SDS extract 10 times with TBS and subsequently incubated it with BP sera. All of the BP sera reactive with the 170-kDa protein by immunoblotting precipi­tated the 170-kDa antigen from this antigen sample (Fig 3B). Nei­ther the two BP sera reactive with only the 230-kDa protein nor the three control sera that were not used for pre-clearance of the cell extract precipitated the 170-kDa protein. Although the signal was weak from the same antigen sample, MoAb D-20 also precipitated the 170-kDa protein, which showed the same mobility as that pre­cipitated by BP sera (Fig 3C).

DISCUSSION

In the present study, we showed for the first time that the 170-kDa protein can be detected by immunoprecipitation. We first con­firmed that the BP sera detect both the 230- and 170-kDa proteins with immunoblot analysis by using extracts of cultured human squamous cell carcinoma cells (KU8 cells) in a pattern similar to that produced by normal human epidermal extract. That both the 230-and 170-kDa proteins are expressed in KU8 cells was confirmed by the result of IIF, in which characteristic hemidesmosomal plaque staining was clearly detected by both the 230- and the 170-kDa protein - specific BP sera. The pattern of reactivity was indistin­guishable from that by MoAb D-20.

Like previous reports [7 - 10], with conventional immunoprecipi­tation, only the 230-kDa protein was detected from Nonidet P-40 extracts; however, if the radiolabeled cells were extracted with 1% SDS, and antigen-antibody complex was produced in the presence of a reduced SDS concentration, both the 230- and 170-kDa pro­teins were detected. The pattern of reactivity with the two proteins was the same as that with immunoblot analysis . That MoAb D-20 also precipitated the 170-kDa protein in SDS extracts further con­firmed the specificity of this reactivity. These results indicate that the 170-kDa protein may be extracted in the presence of 1 % SDS but not by 0.5% Nonidet P-40 buffer because the 170-kDa protein is integrated in hemidesmosomes more strongly than the 230-kDa protein. A less likely possibility is that the 170-kDa protein is ex­tremely sensitive to proteases. In our modified immunoprecipita­tion using 1 % SDS for cell extraction, SDS may inactivate proteases by denaturation, and therefore the 170-kDa protein can be detected. Another possibility is that non-ionic detergents, such as Nonidet P-40, conceal the epitopes of the 170-kDa protein.

Mueller et al [11] reported that a very faint 166-kDa protein band

- 230kD--

170kD--

c

BBBBBBBBBC PPPPPPPPPO 123456789N

T

P P 8 5 2

o Figure 3. lmmunoprecipitation. (A) Conventional nonidet P-40 extracts; ., the position of nonspecific band of approximately 200 kOa, which is constantly observed in this irnrnunoprecipitation, even by control (CONT) serum. (B) SOS extracts. (C) SOS extracts for MoAb 0-20. (See Fig I for explanation of the arrows and IJars.)

VOL. 100, NO.2 FEBRUARY 1993

was occasionally detected with the conventional immunoprecipita­tion. This 166-kDa protein may be the same as the 170-kDa pro­tein; however, because this protein was always associated with the more prominent 230-kDa protein, they considered that the 166-kDa protein may be a degradation product of the 230-kDa protein. Contrary to this report, our immunoprecipitation using 1 % SDS buffer for extraction clearly detected a 170-kDa protein that was not necessarily associated with the 230-kDa protein. Moreover, the pattern of reactivity on immunoprecipitation was the same as that on immunoblotting for each serum. These results suggest that the 170-kDa protein is not a degradation product of the 230-kDa pro­tein but is an intrinsic molecule.

Although the 230-kDa protein was immunoprecipitated from the 1 % SDS extract of KU8 cells after dialysis against 0.1 % SDS­TBS, the 170-kDa antigen was not immunoprecipitated from the same sample but could be detected after the sample was tenfold further diluted. The 230-kDa protein may be bound by its corre­sponding antibody under the condition containing 0.1 % SDS, whereas the 170-kDa protein is capable of making complexes with its antibody at an SDS concentration of less than 0.1 %. This differ­ence of affinity for the antibodies may also suggest the different characteristics between the 230- and 170-kDa proteins.

Because the BP sera reactive only with the 170-kDa antigen stained the hemidesmosomal plaque ofKU8 cells in the same man­ner as the 230-kDa antigen-specific BP sera and MoAb D-20, the 170-kDa BP antigen is most likely located in the hemidesmosome. Recently, cDNA encoding the 170-kDa BP antigen has been cloned, although the cDNA does not cover the complete molecule [20]. The 230-kDa BP antigen has been extensively investigated using its cDNA [21-23]; however, the real nature of the 170-kDa BP antigen has not been revealed . So far, only four molecules have been identified as hemidesmosomal components, namely, the 230-and 170-kDa BP antigens [20-23], the a6fJ4-integrin complex [24,25], and epiligrin [26]. In the future, studies of these proteins, including isolation and characterization of cDNA encoding entire molecules of the 170-kDa BP antigen, will unravel the structure and function of hemidesmosomes, the most important apparatus of cell- substrate adhesion of epidermis.

We are grateful to Dr. K. Owaribe, Departmwt of Molewlar Biology, FaCIlity of Science, Nagoya U"iversity, for providi"g us with MoAb D-20, a/Jd to Drs. K. Katsube a"d Y. I/Jokuclzi for help with the immu/Joprecipitatio/J assay. We also thank Dr. Y. Kitajima, Departlllwt of Dermatology, Jichi Medical College, for advice on immunojluoresceIJce.

REFERENCES

1. Lever WF: Pemphigus. Medicine (Baltimore) 32:1-123, 1953 2. Jordon RE, Beutner EH, W itebsky E, Blumental G, Hale WL, Lever

WF: Basement zone antibodies in bullous pemphigoid. JAMA 200:91- 96, 1967

3. Jordon RE, Day NK, Sams WMJr, Good RA: The complement sys­tem in bullous pemphigoid. I. Complement and component levels in sera and blister fluids. J Clin Invest 52:1207 -1214,1973

4. Gammon WR, Levis DM, Carls JR, Sams WM Jr, Wheeler CE Jr: Pemphigoid antibody mediated attachment of peripheral blood leu­kocytes at the dermoepidermal junction of human skin. J Invest DermatoI75:334-339,1980

5. Naito K, Morioka S, Ogawa H: The pathogenic mechanisms of blister formation in bullous pemphigoid.J Invest DermatoI79:303-306, 1982

6. N aito K, Morioka S, Ikeda S, Ogawa H: Experimental bullous pem­phigoid in guinea pigs: the role of pemphigoid antibodies, comple­ment and migrating cells. J Invest Dermatol 82:227 - 230, 1984

7. Stanley JR, Hawley-Nelson P, Yuspa SH, Shevach EM, Katz SI: Char­acterization of bullous pemphigoid antigen: a unique basement

IMMUNOPRECIPITATION FOR 170-kDa BP ANTIGEN 179

membrane protein of stratified squamous epithelia. Cell 24:897-903,1981

8. Stanley JR, Woodley DT, Katz SI: Identification and partial character­ization of pemphigoid antigen extracted from normal human skin. J Invest Dermatol 82: 1 08 - 111, 1984

9. Thivolet CH, Hinter HH, Stanley JR: The effect of retinoic acid on the expression of pemphigus and pemphigoid antigens in cultured human keratinocytes. J Invest Dermatol 82:329-334,1984

10. Stanley JR: A specific antigen-antibody interaction triggers the cellular pathophysiology of bullous pemphigoid. Br J Dermatol 113:67-73', 1985

11. Mueller S, Klaus-Kovtun V, Stanley JR: A 230-kD basic protein is the major bullous pemphigoid antigen. J Invest Dermatol 92:33 - 38, 1989

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

13. Sugi T, Hashimoto T, Hibi T, Nishikawa T: Production of human monoclonal anti-basement membrane zone (BMZ) antibodies from a patient with bullous pemphigoid (BP) by Epstein-Barr virus trans­formation . J Clin Invest 84:1050-1055,1989

14. Owaribe K, Nishizawa Y, Franke WW: Isolation and characterization of hemidesmosomes from bovine corneal epithelial cells. Exp Cell Res 192:622-630, 1991

15. Owaribe K, KartenbeckJ, Stumpp S, Magin LM, Krieg T, Diaz LA, Franke WW: The hemidesmosomal plaque. I. Characterization of a major constituent protein as a differentiation marker for certain forms of epithelia. Differentiation 45:207 - 220, 1990

16. Tsukamoto T: Establishment and characterization of a cell line (KU-8) from squamous cell carcinoma of the penis. Keio J Med 38:277-293,1989

17. ParrishEP, MarstonJE, Mattey DL, Measures HR, Venning R, Garrod DR: Size heterogeneity, fhospholyla~ion and transmembra.ne or­gamzatlon of desmosoma glycoprotem 2 and 3 (desmocollms) in MDCK cells. J Cell Sci 96:239 - 248, 1990

18. Schwarz MA, Owaribe K, KartenbeckJ, Franke WW: Desmosomes and hemidesmosomes: constitutive molecular components. Annu Rev Cell Bioi 6:461-491, 1990

19. Riddelle KS, Green KJ, Jones JCR: Formation ofhemidesmosomes in vitro by a transformed rat bladder cell line. J Cell Bioi 112:159-168, 1991

20. Diaz LA, Ratrie. H III, Saunders . WS, Futamura S, Squiquera HL, Anhalt GL~ GIUdice GJ: Isolation of a human epidermal eDNA corresp.ondlllg to the 180-kD .autoantlgen recognized by bullous pemphigOId and herpes gestatloms sera. J Clin Invest 86:1088-1094, 1990

21. Stanley JR, Tanaka T, Muller S, Klaus-Kovtun V, Roop D: Isolation of complementary DNA for b~llous pemphigoid antigen by use of patients' autoantibodies. J Clm Invest 82:1864-1870,1988

22. Amagai M, Hashimoto T, Tajima S, Inokuchi Y, Shimizu N, Saito M, Miki K, Nishikawa T: Partial cDNA cloning of the 230-kD mouse bullous pemphigoid antigen by use of a human monoclonal anti­basement membrane zone antibody. J Invest Dermatol 95:252-259, 1990

23. Tanaka T , Parry DAD, Klaus-Kovtun V, Steinert PM, Stanley JR: Comparison of molecularly cloned bullous pemphigoid antigen to desmoplakm I confirms that they define a new family of cell adhe­sion junction plaque proteins. J Bioi Chem 266, 12555 -12559, 1991

24. Kajiji S, Tamura RN, Quaranta V: A novel integrin (aE,84) from human epithelial cells suggest a fourth family of integrin adhesion receptors. EMBO J 8:673-680,1989

25. Sonnenberg A, Calabat J , Janssen H, Daams H, Helmer ME, Falcioni R, Kennel SJ, AplinJD, Barker J, Loizidou M, Garrod D: Integrin ('(6/,84 complex is localized in hemidesmosomes, suggesting a major role in epidermal cell-basement membrane adhesion. J Cell Bioi 113:907 -917, 1991

26. Carter WG, Ryan MC, Gahr PJ: Epiligrin, a new cell adhesion ligand for integrin ('(3,81 in epithelial basement membranes. Cell 65:599-610, 1991