Anti-Mycobacterium leprae monoclonal antibodies cross-react with human skin: an alternative explanation for the immune responses in leprosy.

A panel of 17 mouse monoclonal antibodies (MoAb) raised against Mycobacterium leprae (M. leprae) antigens was used to detect antigenic determinants in normal human skin. An indirect immunoperoxidase technique was used. Eight of the MoAb detected epidermal antigens similar to patterns well known for human sera. Five of these MoAb detected determinants in the dermis, too. These observations may indicate a certain degree of similarity between the antigenic determinants occurring in M. leprae and in the human host. We propose that such a similarity on the one hand may facilitate the survival of M. leprae in the human host when the antigens are not recognized as "non-self," a situation which seems to occur in lepromatous leprosy, when the patients' tissues are loaded with bacteria virtually without any immune response. On the other hand, M. leprae antigens which mimic host antigens may induce an auto-immune reaction against the host's own antigens, which could explain the immune reaction in tuberculoid leprosy and during a "reversal reaction" when M. leprae is not observed in the host tissues, but extensive granuloma formation occurs.     

Delipidified cell components of Mycobacterium leprae and its applications.

Delipidified cell components (DCC) of Mycobacterium leprae obtained as an insoluble material consist of several proteins. This preparation, DCC, has ability to differentially bind to sera from lepromatous leprosy patients and antibodies to this complex get reduced as patients improve under chemotherapy. The antigenic complex has no ability to bind to proteins of sera from normal healthy individuals or tuberculoid leprosy patients. The DCC is antigenic and is recognised by immune deficient cells of lepromatous leprosy patients, leading to lymphocyte proliferation, production of Interleukin II and interferon gamma, and resulting in activation of the phagocytes to initiate killing of endocytosed M.leprae through reactive oxygen intermediates, primarily superoxide. The DCC has also immunomodulatory properties to protect mice against M.leprae infection. Experiments with mice and isolated peripheral blood cells from patients have indicated the probable molecular mechanism of immunomodulation by DCC.     

Serum samples from patients with mycobacterial infections cross-react with HIV structural proteins Gp41, p55 and p18

BACKGROUND: Infection with Mycobacterium leprae is associated with a high frequency of false positive results in a variety of serological assays. Our studies have found cross-reactivity to HIV structural proteins in serum samples from leprosy patients, irrespective of the type of disease, treatment duration, age and gender and from a few patients with active TB disease. METHODS: Western blot (WB) analysis revealed that sera from HIV negative leprosy patients across the spectrum showed high reactivity with p18, Gp41 and p55 and lower reactivity with other HIV proteins. The reactivity appeared to be specific; western blot-positive samples were negative in ELISA and in several rapid tests for HIV. Cross-reactivity was not found in sera from patients with leishmaniasis or from normal healthy individuals. RESULTS: None of the WB reactive leprosy patients seroconverted to HIV positivity within 6 months to 1 year after Western blot testing. BLAST analysis revealed that envelope antigens of HIV (Gp41, Gp120 and Gp160) contained amino acid sequences similar to M. leprae ML0470, putative integral membrane protein, Rv0740, mmpL9 (M. tuberculosis). Core (gag) antigens (p18) had similarities to ML0406, but polymerase antigens (p52) had similarities to PE_PGRS (M. tuberculosis, H37Rv). Nucleotide sequence analysis, on the other hand, did not reveal any significant homology between M. leprae or M. tuberculosis and HIV. CONCLUSIONS: The occurrence of these high false-positive rates in M. leprae-infected individuals suggests a possible complication of serodiagnosis of HIV in regions where mycobacterial infections are endemic. There is a need for caution in reporting HIV infection among leprosy patients. Our observations emphasise the value of the various rapid assay kits for HIV, where this false positivity is not observed.     

Two groups of bullous pemphigoid antigens are identified by affinity-purified antibodies

The bullous pemphigoid antigen was originally described as a 240-kD protein extracted from human epidermis, but a subsequent report has described patients' sera which react with epidermal proteins of molecular masses 240, 200, 180, 97, and 77 kD. We have evaluated the heterogeneity of the pemphigoid antigens identified by the sera of 10 patients with clinically typical bullous pemphigoid. We used indirect immunofluorescence and Western immunoblots of epidermal extracts prepared from epidermis separated by either 1 M salt or 20 mM EDTA to characterize the reactivity of both crude sera and affinity-purified antibodies. Affinity purification of antibodies was performed with either normal human epidermis or protein bands blotted onto nitrocellulose as immunoabsorbents. The anti-basement membrane antibody titers determined by indirect immunofluorescence on the saline- and EDTA-separated epidermis were identical. Despite this, Western blots of extracts prepared from EDTA-separated epidermis demonstrated greater amounts of the 240-kD antigen than saline split skin. Multiple antigens were recognized in epidermal extracts on Western blots by most crude BP sera, including bands at 240, 200, 160, and 100 kD. Different sera reacted with these antigens with a markedly different intensity, falling into two major groups, those bearing antibodies to the 240-200-kD antigens and those with antibodies to the 160-100-kD components. When epidermis was used as a substrate for affinity purification of bullous pemphigoid antibodies, the eluted antibodies reacted with multiple bands on Western blots, demonstrating the reactivity of anti-basement membrane zone antibodies with multiple proteins. Antibodies eluted from several individual bands of immunoblots were found to react with the basement membrane on indirect immunofluorescence. When these nitrocellulose-purified antibodies were reapplied to Western blots, they cross-reacted within two groups, the 240-200 kD antigens and the 160-100 kD antigens. We conclude that pemphigoid antigens are best demonstrated when EDTA-split skin is used for extraction and that different pemphigoid sera may contain antibodies to two separate groups of basement membrane zone antigens.     

Expression of common cytoplasmic antigens by epithelial cells of thymus and epidermis

Summary The expression of common cytoplasmic antigens by thymic epithelial cells and keratinocytes was analyzed by immunolabelling on cryostat sections of human and animal thymus and epidermis. Experimental sera to human epidermal keratin subunits of molecular weight (MW) 67 K, 63 K and 55 K were used, as well as human sera with antibodies to epidermal cytoplasmic antigens (KCA), reacting either with U-Cyt antigens expressed by cells from the upper compartment of the epidermis or with BLC antigens in cells from the basal layer. It was demonstrated that keratin subunits of MW 67 K and 63 K were detected only in the upper layers of human epidermis, whereas 55 K was present in all epidermal layers. On rabbit lip a labelling was obtained throughout the epidermis by the three immune sera. On thymus sections a cytoplasmic staining of epithelial cells was observed with the three immune sera, suggesting an antigenic similarity between cytokeratins of thymic epithelial cells and keratins of the upper layers of human epidermis. Two human sera with U-Cyt antibodies and one with BCL antibodies labelled human thymic epithelial cells but were negative with mice thymus. Ten other sera with KCA were negative on both thymus. These results show that the expression of keratins and cytoplasmic antigens, defined by human KCA, differed in immunofluorescence on normal human epidermis, rabbit lip and thymic epithelial cells. They suggests a heterogeneity among human keratinocyte cytoplasmic antigens defined by KCA and the presence in man of common antigenic constituents different from keratins in epidermal cells and thymic epithelial cells.Supported by INSERM CRL 812045 and UER Grange Blanche     

Evaluation of four semi-synthetic Mycobacterium leprae antigens with sera from healthy populations in endemic and non-endemic areas.

In order to determine the frequency of occurrence of antibodies to semisynthetic antigens of Mycobacterium leprae in clinically healthy nonpatient populations and to establish a 'baseline' for comparison with antibody frequencies in both patients with a history of leprosy and their contacts, ELISAs were conducted using representative sera from two areas: a leprosy endemic area, Cebu City, Philippines and a nonendemic area for leprosy Chicago, Illinois, USA. These sera were tested, by an indirect IgM ELISA, for the presence of antibodies reacting with four semisynthetic antigens based on the phenolic glycolipid I antigen of M. leprae: ND-O-BSA (natural disaccharide with octyl linkage to bovine serum albumin), NT-O-BSA (natural trisaccharide with octyl linkage to BSA), ND-P-BSA (natural disaccharide with phenolic ring linkage to BSA) and NT-P-BSA (natural trisaccharide with phenolic ring linkage to BSA). Using an OD reading > or = 0.16 as positive, the antigen with the lowest background seroreactivity was ND-O-BSA, which reacted with 5/398 (1.3%) sera from Cebu, and 3/426 (0.7%) sera from Chicago. A total of 10 (2.5%) of 398 sera from the endemic area reacted with at least one antigen and 5 (1.3%) sera reacted with all four semisynthetic antigens. Of the 426 sera from Chicago, 12 (2.8%) were reactive with at least one antigen and 3 (0.7%) were reactive with all four semisynthetic antigens. Mean ELISA values for the 22 positive sera for each antigen ranged from 0.17 to 0.3 OD units, while the mean values for all sera in each area ranged from 0.01 to 0.04 OD units for all four antigens. Reactivity of 14 of the positive sera to some antigens, but not all four semisynthetic antigens, indicated that the carrier and linker arms might be associated with this background reactivity. Investigation of alternative linker arms and carriers is warranted. We conclude that nonspecific background reactivity to the semisynthetic antigens representing the PG-I molecule of M. leprae is 0.7-1.3%, based on a > or = 0.16 OD cutoff value. From these data it was concluded that reactivity in individuals free of leprosy was low enough to warrant use of these antigens in a diagnostic setting, such as screening household contacts and highly endemic populations. When incidence and prevalence of leprosy are low, testing with these antigens would not be cost effective, unless applied to high risk individuals.(ABSTRACT TRUNCATED AT 400 WORDS)     

Identification of basement membrane zone antigens defined by antibodies that react to both the epidermal and dermal side of 1 M sodium chloride split skin.

Some individuals have basement membrane zone (BMZ) antibodies that react to both the epidermal and dermal side of skin split with 1 M NaCl. To examine the significance of this combined staining pattern, we tested sera from 185 different, sequential, patients with BMZ antibodies for reactivity to normal human skin split with 1 M NaCl. Six sera (3.2%) stained both the epidermal and dermal sides of split skin, 173 (93.5%) stained only the epidermal side, and 6 (3.2%) only the dermal side. All six sera with a combined staining pattern yielded the same pattern when tested with three different specimens of skin, indicating that this pattern is reproducible. By immunoblot analysis, five (83%) of the six combined staining sera reacted to a 160-kD antigen present only in epidermal extracts of normal skin, one reacted in addition to a 230-kD epidermal antigen, and one did not react to either epidermal or dermal extracts. In contrast, five (83%) of the six sera with dermal staining reacted to a 290-kD antigen present only in dermal extracts. Eighteen (90%) of twenty representative epidermal staining sera reacted to a 230-kD epidermal antigen and seven (35%) sera (five with the 230-kD antibody and two without) also reacted to the 160-kD epidermal antigen. Affinity purified antibody to the 160-kD antigen defined by combined staining sera reacted to the BMZ of normal human skin. These results indicate that the combined staining pattern on 1 M NaCl split skin is due to the presence of a distinctive antibody response directed predominantly to a 160-kD BMZ antigen located on the epidermal side of the split skin and to an as yet unidentified BMZ antigen located on the dermal side.     

M. leprae recombinant antigens important for T-cell reactivity.

Identification of M. leprae antigens recognized by T-cell is important for specific diagnosis, vaccine development and understanding the basic mechanisms involved in protection against and pathogenesis of leprosy. Screening of an M. leprae recombinant DNA library with antibody probes led to the identification of half a dozen M. leprae antigens recognized by B-cells. When tested for T-cell reactivity, all the antigens recognized by antibodies were shown to have T-cell reactivity. However, among these antigens 18 kDa, 65 kDa and 70 kDa heat shock proteins (hsps) were most frequently recognized by T-cell lines and clones established from healthy donors vaccinated with killed M. leprae. A 24 kDa secreted antigen of M. leprae with T-cell epitope specific for M. leprae and M. tuberculosis complex was identified by direct screening of the recombinant DNA library with T-cell clones. The recombinant T-cell antigens of M. leprae were recognized by memory T-cells of Th1 type in association with multiple HLA-DR molecules. Epitope mapping with synthetic peptides identified M. leprae-specific as well as cross-reactive T-cell epitopes on the 18 kDa, 65 kDa and 70 kDa hsp antigens. In conclusion, our studies suggest that the recombinant antigens of M. leprae could be useful as reagents for specific diagnosis as well as in subunit and recombinant vaccine design against leprosy.     

Challenges presented by nerve damage in leprosy

The basis of nerve damage in leprosy is the unique tendency of Mycobacterium leprae to invade Schwann cells. alphaBeta-Dystroglycan on the basement membrane of Schwann cells binds to laminin alpha2, in turn binding to receptors on the M. leprae surface, comprising a histone-like protein and phenoglycolipid-1. When nerve damage during reversal reactions was found to be associated with an abrupt increase in delayed type hypersensitivity against M. leprae antigenic determinants released from Schwann cells, it suggested that the nerve is damaged as an innocent bystander during the immune response. This strongly influenced the introduction of therapy based on immunosuppression combined with continued anti-mycobacterial medication. Lysis of Schwann cells presenting M. leprae antigenic determinants by activated CD4+ T cells and interaction of M. leprae with Toll-like receptors on Schwann cells are additional mechanisms implicated in nerve damage. Persistence of M. leprae antigen in local lesions after regular multiple drug therapy (MDT) is an important risk factor for late reactions. In spite of significant advances in the provision of MDT globally, early diagnosis, together with effective treatment of the disease and associated nerve damage at initial presentation remains a major challenge for the health services. Reduced prevalence as a result of MDT should not be taken to indicate that the challenges of leprosy control are diminished as long as nerve damage is not controlled and new case detection rates are not declining.     

Characterization of bullous pemphigoid antibodies by use of recombinant bullous pemphigoid antigen proteins

The seroreactivity of patients with bullous pemphigoid (BP) to recombinant proteins representing sequences in the carboxyl domain of the murine 230-kD BP antigen (BPA) was determined. Sera from 133 patients with BP, 20 patients with pemphigus, and 21 normal subjects were examined by Western blotting by using two recombinant proteins: RP120 (MW = 120 kD), representing the C-terminal half of the 230-kD BPA, and RP60 (MW = 60 kD), representing the C-terminal quarter. These RP120 and RP60 were recognized by 84% and 61%, respectively, of the BP sera that reacted with the 230-kD BPA in epidermal extract, and not by any of pemphigus or normal sera. Furthermore, these RP120 and RP60 were not recognized by any BP sera that reacted only with the 170-kD BPA, which is known to be another major BPA. These findings indicate that one or more of the major antigenic regions localizes in the carboxyl-half domain of the 230-kD BPA, and also suggest that the 230-kD BPA may be distinct from the 170-kD BPA.     

Demonstration of Mycobacterium leprae specific antigens in leprosy lesions using monoclonal antibodies

Cryostat sections of dermal lesions from 13 untreated patients of leprosy were studied by indirect immunoperoxidase using monoclonal antibodies (MLO4 & MLO6), defining M. leprae specific antigens. The lymphocytes and macrophages in both the tuberculoid and lepromatous granulomas showed membranous staining with the above antibodies. M. leprae organisms in the lepromatous granulomas and the cells in the section of lymph nodes of patients with tuberculosis, or sections of normal skin or psoriatic lesions did not show any staining with these antibodies. These observations suggest that M. leprae specific antigens are present and expressed on the cells infiltrating the granulomas of leprosy lesions.     

Use of non-conventional antigen: M. habana in detecting M. leprae antibodies from leprosy patients and contacts in FLA-ABS test

An indirect immunofluorescent (FLA-ABS) test has been developed to detect M. leprae specific antibodies in the active and subclinical cases of leprosy. An antigenically related mycobacterium, M. habana, was used as an antigen to detect M. leprae specific antibodies in the sera samples of leprosy patients. A comparison was made with M. leprae antigen using same set of sera samples. M. habana is capable of detecting anti-M. leprae antibodies in the serum samples of leprosy patients, previously absorbed with various mycobacterial antigens, cardiolipin and lecithin, almost to the same percentage as M. leprae. Possible use of M. habana antigen as an alternative to M. leprae, in the serodiagnosis of leprosy, has been discussed.     

Monoclonal antibodies cross-reactive with group A streptococci and normal and psoriatic human skin

Infection with group A streptococci has been implicated as a factor capable of exacerbating psoriasis. In order to explore the possibility of cross-reactivity between streptococcal antigens and human skin in this phenomenon, skin from psoriatic patients and control subjects was reacted with 3 monoclonal antibodies against group A streptococci and antibody binding was estimated by the indirect immunofluorescence technique. Monoclonal antibody 54.2.8 stained the nuclei and cytoplasm of cells within the epidermis and epidermal appendages, as well as cells scattered throughout the dermis. In contrast, monoclonal antibodies 49.8.2 and 36.2.2 labeled the cytoplasm of epidermal cells and epidermal appendages but did not react with nuclei. No difference in the staining patterns of control skin and uninvolved skin from patients with psoriasis was observed. However, skin from psoriatic lesions contained large amounts of cross-reactive skin component(s). Sera from patients with guttate psoriasis did not react differently with normal or psoriatic skin when compared with normal sera. Western immunoblots of skin extracts demonstrated that monoclonal antibody 54.2.8 reacted with a family of proteins in the molecular weight range of 60-70K. The results indicate that component(s) in human skin share cross-reactive epitopes with group A streptococci. Immunologic cross-reactions between group A streptococci and human skin may play an important role in the exacerbation of certain skin disorders following streptococcal infections.     

Development of a mouse food pad model for detection of sub clinical leprosy

Early diagnosis of leprosy and a multi-drug therapy (MDT) regimen will block the trajectory of nerve damage, disability and deformity that are the hallmarks of this chronic disease. However, the diagnosis of leprosy is made solely by recognition of clinical signs and symptoms, requiring special expertise. These limitations also result in the under reporting of worldwide prevalence and incidence rates for leprosy. Sorely needed is an objective laboratory test for detecting early leprosy. As the antigenic burden of M. leprae can be virtually undetectable in early clinical leprosy, cell mediated immunity and antibody responses will likely be weak. So the sensitivity of new diagnostic tests is as important as specificity. Major efforts are underway employing recombinant M. leprae antigens and synthetic peptides, to develop diagnostic assays for early leprosy infection, using in vitro T cell reactivity or serological tests. We have used the initial phase of the mouse foot pad model as an 'early' model of leprosy infection to screen T cell responses against M. leprae specific antigens and synthetic peptides. Unlike human disease in animal models we can control infection progress and monitor bacillary growth relative to time course of development of T cell response to specific M. leprae antigens. The study employed splenic T cells instead of draining lymph node T cells to model the systemic response as opposed to a local one. We found that 10(5) live M. leprae is the minimum dose required for any meaningful and consistent in vitro splenic IFN-gamma response against M. leprae antigens 3 months after foot pad inoculation. Using this model we found that several M. leprae recombinant proteins, ML0840, ML2028, ML2307, ML2346, ML2478, and ML2532, induced significant levels of IFN-gamma secretion. By controlling for variables that can be confounding factors in the sensitivity of human testing, this mouse model provides an interface between M. leprae diagnostic antigen development and the screening of these antigens in humans under field conditions.     

Identification of a 160-kD molecule as a component of the basement membrane zone and as a minor bullous pemphigoid antigen.

The antigens in normal human skin defined by antibodies in patients with bullous pemphigoid (BP) were studied by Western immunoblots. Eighteen (90%) of 20 BP sera reacted to a 230-kD antigen. Seven (35%) of the sera reacted to a 160-kD antigen. Two of these reacted only to the 160-kD antigen and five also reacted to the 230-kD antigen. Antibodies to the 160-kD antigen were not present in 25 control sera obtained from normal individuals or patients with other bullous diseases. The 160-kD antigen was present in epidermal extracts of four different specimens of normal human skin but not in dermal extracts or extracts of control cells including melanoma, fibroblasts, lung carcinoma, and colon carcinoma. Monospecific sera with antibodies to either the 230-kD or to the 160-kD antigen reacted solely to their respective target antigens, but not to both, in extracts of epidermis that contained both antigens. The 160-kD antigen broke down to a 140-kD fragment, while the 230-kD antigen was unchanged in the absence of protease inhibitors. Western blot affinity purified antibody to the 160-kD antigen bound only to the basement membrane zone on the epidermal side of 1M NaCl split skin. These results indicate that a 160-kD antigen is a normal component of the basement membrane zone of human skin. The antigen is located on the epidermal side of skin split with 1M NaCl. It is a minor BP antigen, antibodies to which are present in some patients with BP.     

Antigenic components of Malassezia species for immunoglobulin E antibodies in sera of patients with atopic dermatitis

Antigenic components of Malassezia furfur, M. globosa, M. restricta, M. slooffiae, and M. sympodialis were studied for immunoglobulin E antibodies in sera of patients with atopic dermatitis (AD). Antigenic components were extracted from Malassezia cells by treatment with beta-mercaptoethanol, referred to as 2-ME extract. CBB staining and lectin blots using Con A, LCA, PHA-E4, PNA or RCA120 showed that the 2-ME extracts contained several species-dependent components that differed quantitatively and qualitatively among the Malassezia species at the protein level. In the Western blot with the 2-ME extracts, of 54 sera of the patients with AD (54 patients), the patients' IgE antibodies most frequently recognized components showing molecular weights of 43-46 kDa for M. slooffiae, 12-22 kDa for M. sympodialis, 35-40 kDa for M. restricta, 45-50 kDa for M. globosa, and of 67-72 kDa for M. furfur, respectively. In the correlative study, in which the total band intensities generated for each extract in Western blot were compared among the Malassezia species, the intensity for M. globosa was well correlated with that for M. sympodialis (r=0.756). In the Western blot inhibition test, the 2-ME extract of M. globosa partially inhibited the reaction of the antigenic components of other Malassezia species with the patient's IgE antibodies. These results indicated that Malassezia species contained both species-specific and common antigenic components at the IgE antibody level.     

Cytokeratins in human basal and squamous cell carcinomas: biochemical, immunohistological findings and comparisons with normal epithelia

The nature of epithelial cell cytokeratins from epidermal basal cell carcinomas (BCC) (8 cases) and squamous cell carcinomas (SCC) (5 cases) was investigated by biochemical and immunological analysis. Cytokeratin proteins were extracted with high salt buffer and triton X 100 and were comparatively analyzed by SDS (sodium dodecyl sulphate) polyacrylamide gel electrophoresis. Both types of tumor showed either an absence or a very low amount (5% of the total protein) of the major protein band (MW 67000) present in normal human epidermis. This correlated well with results of immunolabelling showing that 67000 keratin antisera, only reacted with some dyskeratotic cells in sections of these tumors. Gel electrophoresis showed in BCC and SCC, three distinct groups of predominant polypeptide bands of apparent relative MW: (1) 60–62000 (2) 54–56000 and (3) 49000, representing respectively about 43.0%, 31.0% and 20.4% of the total proteins.
Antibodies raised in animals against polypeptide bands C₁ (MW 62000), C₂ (MW 56000) and C₃ (MW 49000) from SCC, strongly labelled (indirect immunofluorescence) all malignant cells present in the 2 kinds of tumors. These antisera showed a preferential reaction with the basal epithelial cells, in sections of human and animal epidermis and mucosa thus, suggesting numerous common antigenic determinants between epithelial cells from diverse origins. On the other hand, strong differences between mucosal and epidermal upper layers were noted with C₁ C₂, C₃, and 67000 antisera. These results are further evidence for the existence of different pathways of keratinization in epidermis and mucosa.     

High-molecular-weight human epidermal transglutaminase

Human stratum corneum was extracted in Tris-HCl containing EDTA and phenylmethylsulfonyl fluoride, separated on sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transblotted to nitrocellulose papers and reacted with rabbit antihuman epidermal transglutaminase (ETG) antibody. Protein-bound antibody was detected with a multistep peroxidase procedure. Proteins with a molecular weight of 50,000 (50kDa) and 72,000 daltons (72kDa) were stained when anti-ETG was used and not when second antibody alone or sera from nonimmunized animals were used. When ETG was treated with trypsin or organic solvents, there was no alteration in the mobility of the 50kDa ETG band, but there was complete disappearance of the 72kDa band. Antibody that bound 72kDa protein, when eluted from the blot, reacted with both 50kDa and 72kDa proteins; similarly, antibody that bound to the 50kDa protein, when eluted from the blot, reacted with both the 50kDa and 72kDa proteins. Partially purified 72kDa ETG activity was increased (3 to 16 times control levels) after heating at 56 degrees C in the presence of calcium and dithiothreitol or by treatment with trypsin. These studies, in conjunction with the previous studies of ETG activation, are consistent with there being two forms of ETG. The different forms may play a role in regulating enzyme activity.     

Antigen specificities of antibasement membrane zone antibodies: immunofluorescence and Western immunoblotting studies

Summary To clarify the antigen specificities of autoantibodies in sera and blister fluids from patients diagnosed as bullous pemphigoid (BP) by routine histology and immunofluorescence (IF) methods, indirect IF studies using the salt split-skin technique were performed. In addition, to detect the BP antigen(s) in human epidermal extracts, Western immunoblotting analyses were carried out. Of 41 sera, 39 (95%) showed a linear pattern of fluorescence along the epidermal side of the separation. Two (5%) sera showed a linear pattern of fluorescence along the dermal side. Blister fluids produced IF staining patterns identical with those of serum samples. These fluorescence patterns were not related to the BP antigen expression of the skin used as substrates. In Western immunoblotting analyses, selected sera showing an epidermal pattern on separated skin primarily reacted with 240 kD, 220 kD, 180 kD, and 150 kD proteins extracted from normal human epidermis. Two sera showing a dermal pattern on separated skin revealed no specific bands. The protein bands recognized by blister fluids were indentical with those of serum samples. These results indicated that blister fluids are also available in immunological analysis, and that BP antibodies have more than one antigenic specificity. Moreover, it is suggested that differential diagnosis between BP and other bullous diseases may be more important than previously recognized, particularly in patients with epidermolysis bullosa acquisita (EBA).