IgA Antibodies of Linear IgA Bullous Dermatosis Recognize the 15th Collagenous Domain of BP180
3
Nevertheless the TNF238.2 polymorphism on the B57.1 haplotype could characterize psoriasis patients with a particular disease course or be related to an impaired clearance of skin infections with candida or steptococci, and thus predispose individuals to the development of psoriasis. Thomas Ho ¨hler Medizinische Klinik und Poliklinik, Johannes Gutenberg-Universita ¨t Mainz, Mainz, Germany, REFERENCES Arias AI, Giles B, Eiermann TH, Sterry W, Pandey JP: TNF alpha gene polymorphism in psoriasis. Clin Exp Immunogenet 14:118–122, 1997 Ho ¨hler T, Kruger A, Schneider PM, et al: A TNF-alpha promoter polymorphism is associated with juvenile onset psoriasis and psoriatic arthritis. J Invest Dermatol 109:562–565, 1997 Imanishi T, Akaza T, Kimura A, Tokunaga K, Gojobori T: Allele and haplotype frequencies for HLA and complement loci in various ethnic groups. In: Tsuji K, Aizawa M, Sasazuki T, eds. HLA 1991. Proceedings of the Eleventh International Histocompatibility Workshop and Conference. Vol. 1. Oxford: Oxford Science Publications, 1992; 1065–1220. Jenisch S, Westphal E, Nair RP, et al: Linkage disequilibrium analysis of familial psoriasis: identification of multiple disease associated MHC haplotypes. Tissue Antigens 53:135–146, 1999 Kaluza K, Reuss E, Grossmann S, Schopf RE, Galle PR, Maerker-Hermann E, Ho ¨hler T: Influence of tumor necrosis factor-alpha (TNF-a)-promoter polymorphisms on transcriptional and in vitro TNF-a-production in patients with psoriatic arthritis and psoriasis. J Invest Dermatol 114:1180–1183, 2000 Nair RP, Stuart P, Henseler T, et al: Localization of psoriasis-susceptibility locus PSORS1 to a 60-kb interval telomeric to HLA-C. Am J Hum Genet 66:1833– 1844, 2000 IgA Antibodies of Linear IgA Bullous Dermatosis Recognize the 15th Collagenous Domain of BP180 To the Editor: Linear IgA bullous dermatosis (LAD) is an autoimmune blistering disease characterized by IgA antibodies against epidermal autoanti- gens. The antigen for the major type of LAD (lamina lucida type) was reported to be a 97 kDa protein (Zone et al, 1990; Dmochowski et al, 1993; Ishiko et al, 1998) or 120 kDa protein (Marinkovich et al, 1996). The 97 kDa/120 kDa protein was found to be a degradation product of the 180 kDa bullous pemphigoid antigen (BP180) (Marinkovich et al, 1997; Pas et al, 1997; Hirako et al, 1998; Zone et al, 1998). To further characterize the antigenic sites in LAD, we performed immunoblot studies using various bacterial recombinant proteins of BP180. Fourteen sera were collected from clinically typical LAD patients and showed IgA antibodies reactive with epidermal side of 1M NaCl-split skin sections with a titer 1:10–1:40 in immunofluorescence. Ten BP sera and 10 normal sera were used as controls. A monoclonal antibody D-20, which reacts with the 120 kDa LAD antigen (Hirako et al, 1998) and an antiglutathione-S-transferase (GST) rabbit antiserum (Sigma, St. Louis, MO) were also included as controls. Normal human epidermal extracts (Dmochowski et al, 1993) and concentrated supernatant of cultured DJM-1 cells (Hirako et al, 1998) and six different GST-fusion proteins of BP180, i.e., NC16A, BP1050, BP963, BP915, BPC15, and BP144, were used as antigen sources for immunoblotting. The first four fusion proteins were reported previously (Matsumura et al, 1996; Nie and Hashimoto, 1999). BP1050, BP963, and BP915 encode N- terminus, central part, and C-terminus of BP180 extracellular domain, respectively. By the similar method described pre- viously (Matsumura et al, 1996; Nie et al, 1999), we prepared two new recombinant proteins; BPC15 encoding the entire 15th collagenous domain, present just downstream of the NC16a domain; and BP144 encoding 38 amino acids, which is further downstream of BPC15, corresponding to the C- terminal region of BP1050. NC16A, BPC15, and BP144 cover the entire BP1050, although BP1050 lacks the N-terminus of the NC16a domain. The constructs for BPC15 and BP144 were confirmed by DNA sequencing and their correct sizes (about 46 kDa for BPC15 and about 31 kDa for BP144) as revealed by anti-GST antiserum in immunoblotting. The positions of all the recombinant proteins are depicted in Fig 1. All the results of immunoblotting were summarized in Table I. In immunoblotting of epidermal extracts, none of the 14 LAD sera showed specific IgA reactivity either with BP180 or with protein bands in 97 kDa/120 kDa areas, whereas BP180 was detected by IgG antibodies of 60% BP sera. In contrast, in immunoblotting of concentrated culture supernatant of DJM-1 cells, IgA antibodies of 12 LAD sera clearly reacted with a 120 kDa protein (Fig 2, panel 1). This protein was also reacted by IgG antibodies of six BP sera and the monoclonal antibody D-20. Because BP1050, BP963, and BP915 roughly cover the whole extracellular domain of BP180, we first examined IgA reactivity of the 14 LAD sera with these fusion proteins. Seven LAD sera reacted clearly with BP1050 (Fig 2, panel 2), and four sera reacted relatively weakly with BP915 (data not shown), whereas no LAD sera reacted with BP963. BP1050 is a large fragment containing 350 amino acids. To further characterize the epitope(s) within BP1050, we examined IgA reactivity of LAD sera with three other fusion proteins, i.e., NC16A, BPC15, and BP144, which in total cover BP1050. Interestingly, five of the 11 LAD sera examined reacted with BPC15 (Fig 2, panel 3), whereas no sera reacted with either NC16A (Fig 2, panel 4) or BP144. The monoclonal antibody D-20 also recognized BP1050 and BPC15, but not other fusion proteins. As expected, IgG antibodies of the control BP sera reacted with Manuscript received May 25, 2000; revised September 6, 2000; accepted for publication September 7, 2000. Reprint requests to: Dr. Takashi Hashimoto, Department of Dermatol- ogy, Kurume University School of Medicine, 67 Asahimachi, Kurume, Fukuoka 830–0011, Japan. Email: [email protected] Figure 1. Schematic structure of BP180 and the positions of the recombinant fusion proteins used in this study. An oval in the left indicates the transmembrane domain, gray columns indicate the collage- nous domains, and open columns indicate noncollagenous domains of BP180. Recombinant proteins of C-terminal extracellular region of BP180 are aligned in the corresponding positions of BP180. 1164 LETTER TO THE EDITOR THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
Transcript of IgA Antibodies of Linear IgA Bullous Dermatosis Recognize the 15th Collagenous Domain of BP180
Nevertheless the TNF238.2 polymorphism on the B57.1haplotype could characterize psoriasis patients with a particulardisease course or be related to an impaired clearance of skininfections with candida or steptococci, and thus predisposeindividuals to the development of psoriasis.
Thomas HoÈhlerMedizinische Klinik und Poliklinik,
Johannes Gutenberg-UniversitaÈt Mainz, Mainz, Germany,
REFERENCES
Arias AI, Giles B, Eiermann TH, Sterry W, Pandey JP: TNF alpha genepolymorphism in psoriasis. Clin Exp Immunogenet 14:118±122, 1997
HoÈhler T, Kruger A, Schneider PM, et al: A TNF-alpha promoter polymorphism is
associated with juvenile onset psoriasis and psoriatic arthritis. J Invest Dermatol109:562±565, 1997
Imanishi T, Akaza T, Kimura A, Tokunaga K, Gojobori T: Allele and haplotypefrequencies for HLA and complement loci in various ethnic groups. In: TsujiK, Aizawa M, Sasazuki T, eds. HLA 1991. Proceedings of the EleventhInternational Histocompatibility Workshop and Conference. Vol. 1. Oxford: OxfordScience Publications, 1992; 1065±1220.
Jenisch S, Westphal E, Nair RP, et al: Linkage disequilibrium analysis of familialpsoriasis: identi®cation of multiple disease associated MHC haplotypes. TissueAntigens 53:135±146, 1999
Kaluza K, Reuss E, Grossmann S, Schopf RE, Galle PR, Maerker-Hermann E,HoÈhler T: In¯uence of tumor necrosis factor-alpha (TNF-a)-promoterpolymorphisms on transcriptional and in vitro TNF-a-production in patientswith psoriatic arthritis and psoriasis. J Invest Dermatol 114:1180±1183, 2000
Nair RP, Stuart P, Henseler T, et al: Localization of psoriasis-susceptibility locusPSORS1 to a 60-kb interval telomeric to HLA-C. Am J Hum Genet 66:1833±1844, 2000
IgA Antibodies of Linear IgA Bullous Dermatosis Recognizethe 15th Collagenous Domain of BP180
To the Editor:
Linear IgA bullous dermatosis (LAD) is an autoimmune blisteringdisease characterized by IgA antibodies against epidermal autoanti-gens. The antigen for the major type of LAD (lamina lucida type)was reported to be a 97 kDa protein (Zone et al, 1990;Dmochowski et al, 1993; Ishiko et al, 1998) or 120 kDa protein(Marinkovich et al, 1996). The 97 kDa/120 kDa protein was foundto be a degradation product of the 180 kDa bullous pemphigoidantigen (BP180) (Marinkovich et al, 1997; Pas et al, 1997; Hirakoet al, 1998; Zone et al, 1998). To further characterize the antigenicsites in LAD, we performed immunoblot studies using variousbacterial recombinant proteins of BP180.
Fourteen sera were collected from clinically typical LADpatients and showed IgA antibodies reactive with epidermal sideof 1 M NaCl-split skin sections with a titer 1:10±1:40 inimmuno¯uorescence. Ten BP sera and 10 normal sera wereused as controls. A monoclonal antibody D-20, which reactswith the 120 kDa LAD antigen (Hirako et al, 1998) and anantiglutathione-S-transferase (GST) rabbit antiserum (Sigma, St.Louis, MO) were also included as controls. Normal humanepidermal extracts (Dmochowski et al, 1993) and concentratedsupernatant of cultured DJM-1 cells (Hirako et al, 1998) and sixdifferent GST-fusion proteins of BP180, i.e., NC16A, BP1050,BP963, BP915, BPC15, and BP144, were used as antigensources for immunoblotting. The ®rst four fusion proteins werereported previously (Matsumura et al, 1996; Nie andHashimoto, 1999). BP1050, BP963, and BP915 encode N-terminus, central part, and C-terminus of BP180 extracellulardomain, respectively. By the similar method described pre-viously (Matsumura et al, 1996; Nie et al, 1999), we preparedtwo new recombinant proteins; BPC15 encoding the entire15th collagenous domain, present just downstream of theNC16a domain; and BP144 encoding 38 amino acids, whichis further downstream of BPC15, corresponding to the C-terminal region of BP1050. NC16A, BPC15, and BP144 coverthe entire BP1050, although BP1050 lacks the N-terminus ofthe NC16a domain. The constructs for BPC15 and BP144were con®rmed by DNA sequencing and their correct sizes
(about 46 kDa for BPC15 and about 31 kDa for BP144) asrevealed by anti-GST antiserum in immunoblotting. Thepositions of all the recombinant proteins are depicted in Fig 1.
All the results of immunoblotting were summarized in Table I.In immunoblotting of epidermal extracts, none of the 14 LAD serashowed speci®c IgA reactivity either with BP180 or with proteinbands in 97 kDa/120 kDa areas, whereas BP180 was detected byIgG antibodies of 60% BP sera. In contrast, in immunoblotting ofconcentrated culture supernatant of DJM-1 cells, IgA antibodies of12 LAD sera clearly reacted with a 120 kDa protein (Fig 2, panel 1).This protein was also reacted by IgG antibodies of six BP sera andthe monoclonal antibody D-20.
Because BP1050, BP963, and BP915 roughly cover thewhole extracellular domain of BP180, we ®rst examined IgAreactivity of the 14 LAD sera with these fusion proteins. SevenLAD sera reacted clearly with BP1050 (Fig 2, panel 2), and foursera reacted relatively weakly with BP915 (data not shown),whereas no LAD sera reacted with BP963. BP1050 is a largefragment containing 350 amino acids. To further characterizethe epitope(s) within BP1050, we examined IgA reactivity ofLAD sera with three other fusion proteins, i.e., NC16A,BPC15, and BP144, which in total cover BP1050. Interestingly,®ve of the 11 LAD sera examined reacted with BPC15 (Fig 2,panel 3), whereas no sera reacted with either NC16A (Fig 2,panel 4) or BP144. The monoclonal antibody D-20 alsorecognized BP1050 and BPC15, but not other fusion proteins.As expected, IgG antibodies of the control BP sera reacted with
Manuscript received May 25, 2000; revised September 6, 2000; acceptedfor publication September 7, 2000.
Reprint requests to: Dr. Takashi Hashimoto, Department of Dermatol-ogy, Kurume University School of Medicine, 67 Asahimachi, Kurume,Fukuoka 830±0011, Japan. Email: [email protected]
Figure 1. Schematic structure of BP180 and the positions of therecombinant fusion proteins used in this study. An oval in the leftindicates the transmembrane domain, gray columns indicate the collage-nous domains, and open columns indicate noncollagenous domains ofBP180. Recombinant proteins of C-terminal extracellular region of BP180are aligned in the corresponding positions of BP180.
1164 LETTER TO THE EDITOR THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
both BP1050 and NC16A, which contain major epitopes forBP; however, neither IgG nor IgA antibodies of any BP ornormal control sera reacted with BPC15.
These results suggest that major epitope(s) for IgA antibodiesof LAD sera locate(s) in the ®fteenth collagenous domain ofBP180. This ®nding, however, is different from that in theprevious study (Zillikens et al, 1999), which showed that someLAD sera reacted with NC16a domain. This discrepancy maybe due to differences in subsets of LAD patients between the
two studies; however, our conclusion is supported by thefollowing observations; (i) the epitope(s) for LAD seradisappeared by collagenase treatment (Pas et al, 1997); (ii) theepitope(s) for LAD sera localized in the lamina lucida betweenNC16a and C-terminal domain of BP180 indicated byimmunoelectron microscopy (Ishiko et al, 1998); (iii) the97 kDa LAD antigen is a truncated form of BP180 startedfrom the amino acid 566, 41 amino acids downstream from thetransmembrane domain, which includes the 15th collagenousdomain (Zone et al, 1998); and (iv) some BP sera reactive withthe 97 kDa LAD antigen reacted with a mid-portion of BP180extracellular domain,1 or with epitope(s) present in the areadistal from the NC16a domain (Egan et al, 1999).
The most interesting and mysterious issue for LAD is why IgAantibodies of LAD sera recognize the degraded 97 kDa/120 kDaLAD antigen (Marinkovich et al, 1996; Pas et al, 1997) and smallfusion proteins, BP1050 and BPC15 (in this study), but not withthe intact BP180. One possible explanation is that the proteins maynot be totally degenerated and may still possess some conformationeven under the presence of SDS, a potent detergent. In thissituation, amino acid sequences in either N-terminus or C-terminus of BP180 may produce some difference in conformationalstructures between the 97 kDa/120 kDa LAD antigen and intactBP180; however, further studies should be necessary to unravel thisissue.
This work was supported by a Grant-in-Aid for Scienti®c Research from the
Ministry of Education, Science and Culture of Japan, and a grant from the Ministry
of Health and Welfare of Japan.
Zhuxiang Nie, Yoshiko Nagata, Sohaola Joubeh,Yoshiaki Hirako,* Katsushi Owaribe,*
Yasuo Kitajima,² Takashi HashimotoDepartment of Dermatology, Kurume University School of
Medicine, Fukuoka, Japan*Unit of Biosystems, Graduate School of Human Informatics,
School of Science, Nagoya University, Aichi, Japan²Department of Dermatology,
Gifu University School of Medicine, Gifu, Japan
REFERENCES
Dmochowski M, Hashimoto T, Bhogal BS, Black MM, Zone JJ, Nishikawa T:Immunoblotting studies of linear IgA disease. J Dermatol Sci 6:194±200, 1993
Egan CA, Taylor TB, Meyer LJ, Petersen MJ, Zone JJ: Bullous pemphigoid sera thatcontain antibodies to BPAg2 also contain antibodies to LABD97 that recognizeepitopes distal to the NC16A domain. J Invest Dermatol 112:148±152, 1999
Hirako Y, Usukura J, Uematsu J, Hashimoto T, Kitajima Y, Owaribe K: Cleavage ofBP180, a 180 kDa bullous pemphigoid antigen, yield a 120-kDa collagenousextracellular polypeptide. J Biol Chem 273:9711±9717, 1998
Ishiko A, Shimizu H, Masunaga T, Yancey KB, Giudice GJ, Zone JJ, Nishikawa T:97 kDa linear IgA bullous dermatosis antigen localizes in the lamina lucidabetween the NC16A and carboxyl terminal domains of the 180 kDa bullouspemphigoid antigen. J Invest Dermatol 111:93±96, 1998
Marinkovich MP, Taylor TB, Keene DR, Burgeson RE, Zone JJ: LAD-1, the linearIgA dermatosis autoantigen, is a novel 120-kDa anchoring ®lament proteinsynthesized by epidermal cells. J Invest Dermatol 106:734±738, 1996
Marinkovich MP, Tran HH, Rao SK, et al: LAD-1 is absent in a subset of junctionalepidermolysis bullosa patients. J Invest Dermatol 109:356±359, 1997
Matsumura K, Amagai M, Nishikawa T, Hashimoto T: The majority of bullouspemphigoid and herpes gestationis serum samples react with the NC16Adomain of the 180-kDa bullous pemphigoid antigen. Arch Dermatol Res288:507±509, 1996
Nie Z, Hashimoto T: IgA antibodies of cicatricial pemphigoid sera speci®cally reactwith C-terminus of BP180. J Invest Dermatol 112:254±255, 1999
Pas HH, Kloosterhuis G, Heeres K, van der Meer JB, Jonkman MF: Bullouspemphigoid and linear IgA keratinocyte collagenous glycoprotein withinantigenic cross-reactivity to BP180. J Invest Dermatol 108:423±429, 1997
Zillikens D, Herzele K, Georgi M, et al: Autoantibodies is a subgroup of patients with
Figure 2. Results of immunoblotting using various antigen sourcesindicated that LAD sera react with ®fteenth collagenous domain ofBP180. Panel 1 (concentrated culture supernatant of DJM-1 cells): IgAantibodies of ®ve representative LAD sera (lanes 3±7), but not a LAD serum(lane 8), reacted with the 120 kDa LAD antigen, which is indicated by anarrow in the left. The same protein band was also recognized by IgGantibodies of a BP serum (lane 1) and the monoclonal antibody D-20 (lane2), but not by IgA antibodies of a normal serum (lane 9).In panels 2±4, thelane with the same number showed the reactivity of the same sera orantibodies, and an arrow in the left indicates the position of intact GST-fusion protein. Panel 2 (BP1050 GST-fusion protein): three of the ®verepresentative LAD sera (lanes 3±7) showed IgA antibodies reactive withthis fusion protein, which was also reacted by the anti-GST antiserum (lane1) and the monoclonal antibody D-20 (lane 2), but not by IgA antibodies ofa normal serum (lane 8). Panel 3 (GST-fusion protein BPC15): the resultsare the same as those of panel 2. Panel 4 (NC16A GST-fusion protein): thisfusion protein was not reacted by IgA antibodies of any LAD sera or normalserum. The anti-GST antiserum, but not the monoclonal antibody D-20,reacted with this fusion protein.
1Egan CA, Reddy D, Z Nie, et al: IgG anti-LABD antibodies in bullouspemphigoid patients' sera react with the mid-portion of the BPAg2ectodomain. Submitted for publication.
VOL. 115, NO. 6 DECEMBER 2000 LETTER TO THE EDITOR 1165
linear IgA disease react with the NC16A domain of BP180. J Invest Dermatol113:947±953, 1999
Zone JJ, Taylor TB, Kadunce DP, Meyer LJ: Identi®cation of the cutaneousbasement membrane zone antigen and isolation of antibody in linearimmunoglobulin A bullous dermatosis. J Clin Invest 85:812±829, 1990
Zone JJ, Taylor TB, Meyer LJ, Petersen MJ: The 97 kDa linear IgA bullousdisease antigen is identical to a portion of the extracellular domain of the180 kDa bullous pemphigoid antigen, BPAg2. J Invest Dermatol 110:207±210, 1998
Mechanisms other than Shunting are Likely Contributingto the Pathophysiology of Erythromelalgia
To the Editor:
We read with interest the article by Mùrk et al, ``Microvasculararteriovenous shunting is a probable pathogenetic mechanism inerythromelalgia,'' in the April 2000 issue of the Journal ofInvestigative Dermatology, in which the authors proposed that thecommon pathogenetic mechanism for erythromelalgia is increasedthermoregulatory arteriovenous shunt ¯ow. We have also notedincreased local perfusion during symptoms of heat, redness, andpain (Sandroni et al, 1999) and have observed a concomitant localhypoxemia during symptoms, which is consistent with thehypothesis of shunting.
Wehave been intrigued by the following additional observations inpatients we have studied (Sandroni et al, 1999): (1) the temperature ofthe symptomatic extremity occasionally exceeds core temperature;shunting alone would not explain this ®nding; (2) of the patients
studied, 63% had evidence of small-®ber neuropathy, mostcommonly that of a severe postganglionic pseudomotor impairment,implying that neuropathy is prevalent. Whether the observedneuropathy led to erythromelalgia, or vice versa, is unclear.
Therefore, we submit that erythromelalgia is a complex disease,and shunting alone does not explain the ®ndings observed duringsymptoms. Increased local metabolism may explain the increase inlocal heat, above that of core temperature. The contribution ofsmall-®ber neuropathy to the syndrome needs to be delineatedfurther.
Mark D. P. Davis, Thom W. Rooke,* Paola Sandroni²Department of Dermatology, *Gonda Vascular Center,
²Department of Neurology, Mayo Clinic and Mayo Foundation,Rochester, Minnesota, U.S.A.
REFERENCES
Mùrk C, Asker CL, Salerud EG, Kvernebo K: Microvascular arteriovenous shuntingis a probable pathogenetic mechanism in erythromelalgia. J Invest Dermatol114:643±646, 2000
Sandroni P, Davis MDP, Harper CM Jr, Rogers RS III, O'Fallon WM, Rooke TW,Low PA: Neurophysiologic and vascular studies in erythromelalgia: aretrospective analysis. J Clin Neuromuscular Dis 1:57±63, 1999
Manuscript received July 4, 2000; accepted for publication September 4,2000.
Reprint requests to: Dr. Mark Davis, c/o Section of Scienti®cPublications, Mayo Clinic, 200 First Street SW, Rochester, Minnesota55905. Email: [email protected]
Table I. Summary of immunoblotting using various antigen sources
Recombinant GST-fusion proteins
Cases Epidermala DJM-1b BP1050 BP963 BP915 NC16A BPC15 BP144
1 ± + + ± ± ± + ±2 ± + + ± + ± + ±3 ± + + ± + ± + ±4 ± + + ± ± ± + ±5 ± + + ± ± ± + ±6 ± + + ± + ± ± ±7 ± + + ± ± ± ndc nd8 ± + ± ± ± ± ± ±9 ± + ± ± + ± ± ±10 ± + ± ± ± ± ± ±11 ± + ± ± ± ± ± ±12 ± + ± ± ± ± ± ±13 ± ± ± ± ± ± nd nd14 ± ± ± ± ± ± nd nd
Total 0 12 7 0 4 0 5 0
D-20 ± + + ± ± ± + ±GST-Ab not done not done + + + + + +
Normal(IgG) 0/10 0/10 0/10 0/10 0/10 0/10 0/10 0/10Normal(IgA) 0/10 0/10 0/10 0/10 0/10 0/10 0/10 0/10BP(IgG) 6/10 6/10 10/10 0/10 0/10 10/10 0/10 0/10BP(IgA) 0/10 0/10 0/10 0/10 0/10 0/10 0/10 0/10
aEpidermal, epidermal extracts; the results show the reactivity with BP180.bDJM-1, concentrated culture supernatant of DJM-1 cells. (+) means that the LAD serum reacted with the 120 kDa LAD antigen.cnd, not done due to shortage of the serum.
1166 LETTERS TO THE EDITOR THE JOURNAL OF INVESTIGATIVE DERMATOLOGY
