Bullous pemphigoid antigen 1 isoforms: potential new target autoantigens in multiple sclerosis?

BACKGROUND: The simultaneous occurrence of bullous pemphigoid (BP) and multiple sclerosis (MS), two autoimmune diseases involving the skin and the central nervous system (CNS), respectively, has been described. OBJECTIVES: As the BPAG1 gene encodes the epithelial isoform of BP antigen 1 (BPAG1-e), a major autoantigen of BP, as well as additional variants expressed in the neurones of the CNS and peripheral nervous system and in Schwann cells, we tested the hypothesis that products of the BPAG1 gene act as shared autoantigens in both BP and MS. METHODS: The reactivity of cerebrospinal fluid (CSF) obtained from 18 patients with MS, 10 patients with other inflammatory CNS diseases and 20 normal controls was assayed by immunoblotting against two recombinant fragments of BPAG1-e encompassing regions that are also found in the neuronal variants BPAG1-n and BPAG1-a. RESULTS: The recombinant protein glutathione-S-transferase (GST)-BPAG1-e1-887 was recognized by five of 18 (27%) CSF samples obtained from patients with MS, two of 10 (20%) samples from patients with other inflammatory neurological diseases and five of 20 (25%) samples from normal controls. Furthermore, two of 18 (11%) CSF samples from patients with MS bound to GST-BPAG1-e1880-2649, whereas none of the samples obtained from patients with other inflammatory neurological diseases or from control subjects showed reactivity. CONCLUSIONS: These results raise the possibility that a subset of patients with MS develops an autoantibody response to the neuronal variants of BPAG1. These findings potentially open the avenue of neuronal BPAG1 variants being novel targets of autoantibodies in neurological diseases.

The hemidesmosomal protein bullous pemphigoid antigen 1 and the integrin beta 4 subunit bind to ERBIN. Molecular cloning of multiple alternative splice variants of ERBIN and analysis of their tissue expression.

The bullous pemphigoid antigen 1 (eBPAG1) is a constituent of hemidesmosomes (HDs), cell-substrate adhesion complexes in stratified epithelia. Although its COOH terminus interacts with intermediate filaments, its NH(2) terminus is important for its recruitment into HDs. To identify proteins that interact with the NH(2) terminus of human eBPAG1, we performed a yeast two-hybrid screen, which uncovered a protein belonging to the LAP/LERP (for LRR and PDZ domain) protein family with 16 NH(2)-terminal leucine-rich repeats and a COOH-terminal PDZ domain. The gene for this LAP/LERP protein comprises at least 26 exons located on the long arm of chromosome 5. In most human tissues, several transcripts were detected differing in the coding region situated upstream of or within the PDZ domain. One of the encoded variants was found to correspond to the recently described protein ERBIN. In yeast and in vitro binding experiments, ERBIN was shown to interact not only with eBPAG1 but also with the COOH-terminal region of the cytoplasmic domain of the integrin beta4 subunit, another component of HDs. Antibodies raised against the COOH terminus showed that ERBIN is expressed in keratinocytes. In transfected epithelial cells the protein, however, was not localized in HDs but was either diffusely distributed over the cytoplasm or concentrated at the basolateral plasma membrane. Because ERBIN had been shown previously to interact with the transmembrane tyrosine kinase receptor Erb-B2, which in turn associates with the integrin beta4 subunit, we suggest that ERBIN provides a link between HD assembly and Erb-B2 receptor signaling.

Autoantibodies to the extracellular and intracellular domain of bullous pemphigoid 180, the putative key autoantigen in bullous pemphigoid, belong predominantly to the IgG1 and IgG4 subclasses.

BACKGROUND: Autoantibodies to the extracellular domain (ECD) of bullous pemphigoid (BP) antigen 180 (BP180) are thought to play a crucial part in the pathophysiology of BP. OBJECTIVES: As the various IgG subclasses have different biological properties, we have sought to assess the relative isotype distribution of IgG to BP180 and their reactivity against the ECD and intracellular domain (ICD) of BP180. METHODS: The reactivity of 27 sera from patients with BP was assayed by immunoblotting against recombinant proteins covering the ECD and ICD of BP180. RESULTS: Twenty-seven (100%) and 21 (77%) of 27 BP sera, respectively, contained IgG1 and IgG4 autoantibodies binding to the ECD of BP180. Fourteen (82%) and six (35%) of the 17 BP sera that were reactive with the ICD of BP180 had autoantibodies of the IgG1 and IgG4 subclass, respectively. The profile of the isotype restriction appeared to be similar when the response to the ECD vs. that to the ICD was compared. IgG2 and IgG3 reactivity with BP180 was found less frequently. Patients with BP of longer duration showed a tendency to have, in addition to IgG1, an IgG4 response. CONCLUSIONS: Consistent with prior evidence indicating that subepidermal blister formation in BP is dependent upon complement activation, the frequent finding of complement-fixing IgG1 autoantibodies to both the ECD and ICD of BP180 might have pathogenic relevance in BP. These findings provide new insights relevant for our understanding of the immune response to BP180, the putative key autoantigen in BP.

Plectin, an unusual target antigen in bullous pemphigoid.

BACKGROUND: Bullous pemphigoid (BP) is a blistering disease associated with autoantibodies directed against two components of hemidesmosomes, BP180 and BP230. OBJECTIVES: To assess whether BP patients have autoantibodies targeting plectin, another hemidesmosomal component showing extensive homology to BP230. METHODS: Examination of sera from 16 patients with BP, using immunoprecipitation studies followed by immunoblotting. RESULTS: Serum of one of the 16 (6%) patients with BP contain autoantibodies binding to plectin, while no reactivity was found with sera from three control subjects. Sera from all 16 BP patients immunoprecipitated BP230 from extracts of biosynthetically radiolabelled human keratinocytes. CONCLUSIONS: Our results indicate that sera from BP patients might contain autoantibodies binding to plectin. Although this protein and BP230 are closely sequence-related, the occurrence of autoantibodies binding to plectin is a rare phenomenon in BP.

Fine structural in situ analysis of nascent DNA movement following DNA replication.

Nascent DNA (newly replicated DNA) was visualized in situ with regard to the position of the previously replicated DNA and to chromatin structure. Localization of nascent DNA at the replication sites can be achieved through pulse labeling of cells with labeled DNA precursors during very short periods of time. We were able to label V79 Chinese Hamster cells for as shortly as 2 min with BrdU; Br-DNA, detected by immunoelectron microscopy, occurs at the periphery of dense chromatin, at individual dispersed chromatin fibers, and within dispersed chromatin areas. In these regions DNA polymerase alpha was also visualized. After a 5-min BrdU pulse, condensed chromatin also became labeled. When the pulse was followed by a chase, a larger number of gold particles occurred on condensed chromatin. Double-labeling experiments, consisting in first incubating cells with IdU for 20 min, chased for 10 min and then labeled for 5 min with CldU, reveal CldU-labeled nascent DNA on the periphery of condensed chromatin, while previously replicated IdU-labeled DNA has been internalized into condensed chromatin. Altogether, these results show that the sites of DNA replication correspond essentially to perichromatin regions and that the newly replicated DNA moves rapidly from replication sites toward the interior of condensed chromatin areas.

High resolution analysis of interphase chromosome domains.

Chromosome territories need to be well defined at high resolution before functional aspects of chromosome organization in interphase can be explored. To visualize chromosomes by electron microscopy (EM), the DNA of Chinese hamster fibroblasts was labeled in vivo with thymidine analogue BrdU. Labeled chromosomes were then segregated during several cell cycles to obtain nuclei containing only 2 to 3 labeled chromosomes. Subsequent immunocytochemical detection of BrdU allowed analysis by EM of chromosome territories and subchromosomal domains in well preserved nuclei. Our results provide the first high resolution visualization of chromosomes in interphase nuclei. We show that chromosome domains are either separated from one another by interchromatin space or are in close contact with no or little intermingling of their DNA. This demonstrates that, while chromosomes form discrete territories, chromatin of adjacent chromosomes may be in contact in limited regions, thus implying chromosome-chromosome interactions. Chromosomes are organized as condensed chromatin with dispersed chromatin extending into the interchromatin space that is largely devoid of DNA. The interchromatin space, which is known to be involved in various nuclear functions, forms interconnecting channels running through and around chromosome territories. Functional implications of this organization are discussed.

IgG autoantibodies from bullous pemphigoid patients recognize multiple antigenic reactive sites located predominantly within the B and C subdomains of the COOH-terminus of BP230.

Bullous pemphigoid is a subepidermal bullous disorder characterized by an autoantibody response against the bullous pemphigoid antigen 230 (BP230) and the bullous pemphigoid antigen 180 (BP180), a cytoplasmic component and a transmembrane component, respectively, of hemidesmosomes. Although immunodominant sequences within the extracellular domain of BP180 have been identified, characterization of the antigenic sites on BP230 is still incomplete. To identify autoantibody-reactive sites on BP230 and to examine whether the targeted regions are contained within functionally important domains, recombinant fragments encompassing almost the entire BP230 were used to assess the reactivity of 25 bullous pemphigoid sera by immunoblotting. Our results demonstrate that (i) the region bearing the B and C subdomains of the COOH-terminus of BP230 contains immunodominant sequences recognized by the majority of bullous pemphigoid sera; (ii) additional autoantibody- reactive sites are present over extended regions of the NH2-terminal half of BP230 without evidence for antigenic cross-reactivity between the NH2- and COOH-termini of BP230; and, finally, (iii) autoantibodies reacting with the BP230 tail predominantly belong to the IgG4 and IgG1 subclasses, suggesting that both autoreactive TH2 and autoreactive TH1 cells regulate the autoantibody response to immunodominant sequences of BP230. As the COOH- terminus of BP230 mediates the attachment of keratin intermediate filaments to the hemidesmosomal plaque, whereas its NH2-terminus contains sequences important for its interaction with other constituents of hemidesmosomes, autoantibodies to BP230 might precipitate subepidermal blister formation and perpetuate the disease not only by eliciting an inflammatory reaction but also by interfering with the function of BP230 and thus the stability of hemidesmosomes.

IgG autoantibodies from a lichen planus pemphigoides patient recognize the NC16A domain of the bullous pemphigoid antigen 180.

Lichen planus pemphigoides (LPP) most likely encompasses a heterogeneous group of subepidermal autoimmune blistering disorders occurring in association with lichen planus. We describe the case of a 49-year-old patient with features characteristic of LPP. Direct immunofluorescence microscopy studies demonstrated linear deposits of C3 along the cutaneous basement membrane, while circulating IgG autoantibodies directed against the epidermal side of skin separated by 1 M NaCl were detected. The patient's serum contained IgG autoantibodies immunoblotting a recombinant form of bullous pemphigoid antigen 180 (BP180), but not the COOH-terminus of BP230. By using deletion mutants, it was found that IgG reactivity was restricted to the NC16A domain of BP180, the region harboring the major antigenic sites targeted by IgG autoantibodies from patients with the bullous pemphigoid group of disorders. Our findings provide support to the idea that a subset of patients with LPP have a distinct form of bullous pemphigoid associated with lichen planus.

IgG autoantibodies from bullous pemphigoid (BP) patients bind antigenic sites on both the extracellular and the intracellular domains of the BP antigen 180.

Bullous pemphigoid (BP) and gestational pemphigoid (PG) are subepidermal blistering disorders associated with autoantibodies directed against two components of hemidesmosomes: the BP antigen 180 (BP180) and the BP antigen 230 (BP230). Autoantibodies against the extracellular domain (ECD) of BP180 are thought to play an initiatory role in subepidermal blister formation. To characterize the targeted antigenic sites on BP180, we have assessed the reactivity of sera from BP and PG patients against eukaryotic recombinant proteins encompassing various portions of the ECD and the intracellular domain (ICD) of BP180. Twenty-two of 22 (100%) BP sera that immunoblotted BP180 in keratinocyte extracts, bound a mutant form consisting of the entire ECD of BP180, whereas only three of these 22 sera (14%) reacted against the ECD of BP180 lacking the NC16A membrane proximal region. Thirteen out of the 22 (59%) BP sera recognized the ICD of BP180. Circulating IgG from a representative BP patient that was affinity purified against the ECD of BP180 did not bind the ICD when reblotted, indicating that there was no antigenic cross-reactivity between the ECD and the ICD of BP180. Reactivity against the ICD of BP180 was further ascertained by immunofluorescence microscopy studies showing that nine of the 22 (41%) BP sera stained COS-7 cells expressing the ICD of BP180. Using deletion mutants of the ICD of BP180, the majority of the sera was found to recognize the central region of the ICD of BP180. Specifically, an immunodominant region was localized to an 87-amino acid segment located towards the NH2-terminus of BP180. In contrast to BP sera, five of six (83%) PG sera contained IgG that recognized exclusively the NC16A region, whereas none bound to the ICD of BP180. Together, the results indicate that in BP, autoantibody reactivity to BP180 is not exclusively restricted to the NC16A region, but that additional antigenic determinants exist on the ICD of BP180. The observed heterogeneous immune response against BP180 might reflect intramolecular epitope spreading. Because the ICD ofBP180 harbors functionally important regions, it is possible that autoantibodies against the ICD of BP180 have pathogenic significance for the progression of the disease.

A new immunocytochemical technique for ultrastructural analysis of DNA replication in proliferating cells after application of two halogenated deoxyuridines.

We describe a colloidal gold immunolabeling technique for electron microscopy which allows one to differentially visualize portions of DNA replicated during different periods of S-phase. This was performed by incorporating two halogenated deoxyuridines (IdUrd and CldUrd) into Chinese hamster cells and, after cell processing, by detecting them with selected antibodies. This technique, using in particular appropriate blocking solutions and also Tris buffer with a high salt concentration and 1% Tween-20, prevents nonspecific background and crossreaction of both antibodies. Controls such as digestion with DNase and specific staining of DNA with osmium ammine show that labeling corresponds well to replicated DNA. Different patterns of labeling distribution, reflecting different periods of DNA replication during S-phase, were characterized. Cells in early S-phase display a diffuse pattern of labeling with many spots, whereas cells in late S-phase show labeling confined to larger domains, often at the periphery of the nucleus or associated with the nucleolus. The good correlation between our observations and previous double labeling results in immunofluorescence also proved the technique to be reliable.

Ultrastructural distribution of the death-domain-containing MyD88 protein in HeLa cells.

MyD88, a protein implicated in interleukin-1 signaling, was localized in HeLa cells transiently transfected with an epitope-tagged (flag) version of MyD88. Overexpression of MyD88 can induce apoptosis. We have analyzed the fine structural intracellular distribution of MyD88 using immunoelectron microscopy. MyD88 is localized to the nucleus and to the cytoplasm as revealed by immunofluorescence visualization. Ultrastructural immunocytochemistry shows that, in the cytoplasm, this protein is associated with fibrillar aggregates containing beta-actin. In the nucleus, MyD88 was found in fibrillar domains present only in cells not yet displaying morphological signs of apoptosis. These domains are not derived from nucleoli and do not constitute an accumulation site of splicing factors. We suggest that such structures could be involved in the formation of the apoptotic bodies and/or in the modification of the nuclear structure and of nucleocytoplasmic trafficking during apoptosis.

Cyclic adenosine monophosphate (cAMP) differentially regulates IL-4 in thymocyte subsets.

It was recently reported that cAMP upregulated IL-4 in T cell lines generated in vitro and expressing an unrestricted lymphokine pattern. In order to study the effect of cAMP on IL-4 released by freshly isolated T cell populations, we tested various T cell subsets that have previously been reported to synthesize this lymphokine. We found that cAMP upregulated IL-4 release from in vivo activated single CD4+ peripheral T cells and CD4+CD8-HSAlowNK1.1- thymocytes stimulated with ionomycin and phorbol ester. Furthermore, as in conventional single CD4+ or CD8+ thymocytes, cAMP enhanced IL-4 production in CD4+CD8-NK1.1+ thymocytes. This latter subset has recently been shown to belong to a independent T cell lineage that is positively selected by MHC class I molecules, recognizes CD1 and expresses a restricted T cell receptor repertoire. In contrast, cAMP appeared not to upregulate IL-4 in a third independent T cell lineage. This was suggested by the observation that cAMP did not increase IL-4 in CD3+CD4-CD8- thymocytes, which are thought not to give rise to conventional single CD4+ or CD8+ T cells. It therefore appears that cAMP is coupled differently to IL-4, depending on the T cell lineage. In contrast, in all IL-4 producing stages of conventional T cells, IL-4 is consistently upregulated by cAMP.

Involvement of granulocytes and the adhesion receptors intercellular adhesion molecule-1 and lymphocyte function-associated antigen-1 in tissue inflammation induced by Th2-type helper cells.

We reported recently that subcutaneously injected, anti-CD3 epsilon-pulsed polyclonal Th2 cells mediate interleukin-4-dependent local tissue inflammation. Because a prominent polymorphonuclear infiltrate was observed in the lesions at the time of maximal tissue swelling, we investigated the involvement of polymorphonuclear leukocytes and their adhesion molecules lymphocyte function-associated antigen-1 (LFA-1) and intercellular adhesion molecule-1 (ICAM-1) in Th2-cell-mediated inflammation. Pretreatment of recipient mice with a depleting monoclonal antibody to neutrophils or with blocking antibodies to LFA-1 or to ICAM-1 completely abrogated tissue swelling in Th2-cell-mediated inflammation. Granulocyte infiltration at 6 h was also inhibited by the antibodies to neutrophils and to ICAM-1, but not by that to LFA-1. Tissue swelling mediated by Th1 cells had different kinetics and was not prevented by administration of anti-neutrophil antibody: maximal edema formation occurred at 24-48 h, when the predominant cellular infiltrate was mononuclear. Because the Th1-cell-induced infiltrate at 6 h also consisted of granulocytes but was not associated with pronounced edema, the mere presence of infiltrating polymorphonuclear leukocytes seems not to be sufficient to induce edema. Because edema but not granulocyte infiltration was inhibited by anti-LFA-1 and because anti-LFA-1 antibodies are known to inhibit several functions of neutrophils, our results suggest that, in inflammation mediated by Th2 cells, granulocytes induce edema through their activation and/or degranulation.

Th2 cells mediate IL-4-dependent local tissue inflammation.

We investigated whether polyclonal murine Th1 and Th2 cells obtained after short term culture in vitro were capable of mediating tissue inflammation in vivo. Th cells were pulsed with mAb to the TCR/CD3 complex and injected into the footpads or ears of naive syngeneic recipient mice. Th1 induced delayed swelling that peaked at 24 to 48 h and lasted > or = 5 days. Th2-induced swelling peaked at 6 h and lasted < or = 48 h. Similar responses were also observed in athymic nude mice. Lesions with both Th1 and Th2 cells contained a predominant neutrophilic infiltrate at 6 h, and mainly mononuclear cells at 48 h. The inflammatory response with Th2 was blocked by cyclosporin A, by mAb to IL-4 or by soluble rIL-4R. The requirement for IL-4 was early and transient. Four alloreactive short term Th2 clones induced swelling in allogeneic recipients. mAb- and IL-4-dependent swelling responses were also observed with two long term Th2 clones. Our results demonstrate that Th2 cells mediate IL-4-dependent tissue inflammation, and strengthen the concept that Th2 cells play an important role in some T cell-dependent immune reactions and, possibly, in allergic disorders such as atopic dermatitis and allergic asthma.

Intracellular epidermal interleukin 1-like factors in the human epidermoid carcinoma cell line A431.

Normal human epidermal cells produce, in primary culture, activities which stimulate the release of PGE2 and collagenase by dermal fibroblasts; this factor(s) might play an important role in epidermal-dermal interactions. Since these activities were mainly found in the cell lysates with only little being detected in the conditioned media, we investigated further the problem of cell-associated versus released activity in the model of the human epidermoid carcinoma cell line A431. The activities were consistently found in the cell lysate and in the conditioned media only when the cells were leaky. No membrane-associated activities were identified. Purification of the cytosolic activities were identified. Purification of the cytosolic activities yielded two differently charged species both with a MW of approximately 17K. The copurification of PGE2- and collagenase-stimulating activities with thymocyte comitogenic activity suggests a close physiochemical relation to IL-1. The activities described here might therefore correspond to the intracellular counterpart of epidermal IL-1 formerly described as epidermal cell-derived thymocyte activating factor (ETAF) and identified in the conditioned medium of cultured epidermal cells. These observations are of importance when studying the modulation of these activities.

Interleukin 1 is present in normal human epidermis.

We investigated the presence of interleukin 1 (IL 1)-like molecules in normal unstimulated human epidermal tissue. Epidermis from 21 healthy individuals that was prepared by two different methods showed prostaglandin E2 (PGE2) and collagenase stimulating activity for human dermal fibroblasts. All epidermal extracts tested were positive for thymocyte comitogenic activity (lymphocyte activating factor; LAF). Removal of the horny layer decreased epidermal IL 1-like activity. In contrast to epidermal tissue, freshly isolated peripheral blood mononuclear cells (PBMC) contained no detectable PGE2 stimulatory activity. They could, however, produce PGE2 stimulatory activity after culture and stimulation with phytohemagglutinin (PHA) and concanavalin A (Con A). Little membranous IL 1-like activity could be detected in epidermal extracts when using a method that has previously rendered membranous IL 1 from murine proteose peptone-elicited peritoneal macrophages. Gel filtration chromatography yielded double peaks at m.w. approximately 30,000 and approximately 17,000 for all three activities. High pressure liquid chromatography (HPLC) analysis identified two species with a m.w. of approximately 17,000, and one approximately 30,000 species nondissociable in detergent, all having superposable PGE2 and collagenase stimulatory as well as LAF activity. These results establish the existence of IL 1-like molecules, together with a possible precursor, in normal human epidermis. The release of these preformed epidermal IL 1 stores might be important in vivo.

Normal epidermis contains high amounts of natural tissue IL 1 biochemical analysis by HPLC identifies a MW approximately 17 Kd form with a P1 5.7 and a MW approximately 30 Kd form.

Recently, we demonstrated that normal human epidermis contains IL1. The present work analyses further the significance of this observation by comparing the quantities of epidermal IL1 to those of 7 internal organs in normal rat. Epidermal PGE2 stimulating activity was 200 to 900 fold higher than in the internal organs. Similar results were obtained for the collagenase stimulatory activity and the comitogenic effect on thymocytes. Epidermal IL1 activity in membrane preparations was more than 100 times lower than the activity of the cytosol containing supernatant from the epidermal homogenate. Biochemical analysis by HPLC identified a MW approximately 17 Kd form with a pI 5.7 and a MW approximately 30 Kd form. These results 1) confirm the presence of high amounts of preformed IL1 in normal unstimulated epidermis, 2) show that, compared to 7 internal organs, the epidermis contains the highest activity, 3) suggest that epidermal IL1 is mainly situated in the cytosol, 4) identify a MW approximately 17 Kd form with a pI 5.7, as well as a MW approximately 30 Kd form which might represent an IL1 precursor, 5) demonstrate the copurification of PGE2 and collagenase stimulatory and thymocyte comitogenic activity and 6) give a starting reference for the in vivo study of IL1 activity in pathologic situations.

Cultured human epidermis cells produce cell-associated interleukin 1-like prostaglandin E2- and collagenase-stimulating factors.

In order to identify factors which may regulate the functions of dermal fibroblasts, cell lysates and conditioned media of cultured human epidermal cells were tested on dermal fibroblasts for the stimulation of prostaglandin E2- and collagenase-production. Both prostaglandin E2- and collagenase-stimulating activities appeared during epidermal cell culture: after 2 d they were detected in the cell lysate, and after 4 d of culture they were found also in the conditioned media. Molecular sieving chromatography of epidermal cell lysates led to the detection of two main peaks showing concomitant prostaglandin E2- and collagenase-stimulating activities at Mr approximately equal to 18 000 and Mr approximately equal to 10 000. A single peak of concomitant prostaglandin E2- and collagenase-stimulating activities were seen at Mr approximately equal to 10 000 in the epidermal cell conditioned media. This suggests that the cell-associated concomitant prostaglandin E2- and collagenase-stimulating activities are processed from a common precursor molecule and released. Collagenase-stimulating activity without accompanying prostaglandin E2 was also detected in the range of Mr approximately equal to 30 000-45 000.