Concurrent mucous membrane pemphigoid and membranous glomerulonephritis in a patient with autoantibodies targeting the 1080 region of collagen XVII

Dear Editor, Mucous membrane pemphigoid (MMP) is a heterogeneous autoimmune disease with subepithelial blistering of mucous membranes and occasionally skin. It is characterized by autoantibodies targeting epidermal basement membrane zone (BMZ) proteins: BP180 (collagen XVII), BP230 (BPAG1), laminin‐332, collagen VII and α6β4 integrin. Reactivity to BP180 is predominantly outside the noncollagenous domain 16A (NC16A) region known to be pathogenic in bullous pemphigoid. A documented link exists between pemphigoid disorders and immune‐mediated renal abnormalities,1 2 but a common mechanism has not been identified. To our knowledge, this is the first report of simultaneous presentation of MMP and membranous glomerulonephritis (MGN). A 68‐year‐old man presented with a 1·5‐year history of oral erosions and some cutaneous involvement. Direct immunofluorescence (IF) of skin biopsy showed linear deposits of IgG and C3 along the BMZ. Indirect IF revealed linear IgG (titre 1:1280), as well as IgA, reacting exclusively with epidermal side of 1M NaCl‐split normal human skin (Fig. 1a). Circulating antibody concentrations were 11·6 U mL−1 for BP180 and negative for BP230 (MBL International, Japan, positive test > 9 U mL−1). The patient's serum reacted positively with recombinant integrin α6 on immunoblot, consistent with MMP. Screening labs were significant for non‐nephrotic proteinuria and haematuria. Direct IF of renal biopsy revealed deposition of IgG and C3 in glomeruli. Electron microscopy demonstrated electron‐dense deposits along the subepithelial region of glomerular BMZ. Serum and tissue PLA2R was negative. A diagnosis of MGN was made. Immunofluorescence staining of skin and identification of the target antigen by immunoblot. (a) Indirect immunofluorescence shows linear deposits of IgG4 along the membrane basement zone of 1M NaCl‐split healthy human skin substrate. (b) Patient serum reacts with lysates made from healthy human skin and kidney; note that reactivity includes bands in the 100–150 kDa region, similar to the bands produced against recombinant AA 490–1497 region of BP180 or sec180. (c) Patient serum shows reactivity against the AA1080–1107 region of BP180, preabsorption of serum with recombinant 1080 protein ablates 1080 reactivity, while mock‐absorption with glutathione S‐transferase (GST) has no effect. As the commercial BP180 enzyme‐linked immunosorbent assay (ELISA) utilizes only the NC16A region, antibodies to other regions are not detected. We evaluated serum reactivity to BP180 using immunoblot of recombinant proteins.3 Strong serum IgG, but not IgA, reactivity was observed to recombinant secreted collagen XVII (sec180) [amino acids (AA) 490–1497] and mild reactivity to NC16A (AA 490–566), consistent with the ELISA results. A similar reactivity pattern was observed in immunoblot against extracts from healthy human skin and kidney; this includes the 100–150 kDa bands produced against the recombinant sec180 (Fig. 1b). To define the regions of specificity, we performed epitope mapping against the following recombinant fragments of BP180: AA 1080–1107, AA 1280–1315 (NC4), AA 1331–1404 and AA 1365–1458. Strong IgG reactivity against AA 1080–1107 and a moderate IgA reactivity against NC4 region was observed. Preabsorption of serum with recombinant 1080 resulted in specific loss of immunoblot reactivity (Fig. 1c) as well as 50% reduction in indirect IF signals. Indirect IF with patient serum against healthy human or mouse kidney did not produce any glomerular staining. At the time of writing, the patient had three mucous membrane/cutaneous flares, each accompanied by non‐nephrotic proteinuria and haematuria. He received three cycles × 4 doses of 375 mg m−2 rituximab. His mucosa and kidney responded each time and he remained in remission for approximately 1·5 years. The parallel course in severity and treatment response of cutaneous and renal manifestations suggest that in our patient the two‐organ involvement may be a result of primary events, although the possibility of secondary MGN cannot be completely ruled out. Consistent with this, his immunoreactants were observed as electron‐dense deposits only along the glomerular basement membrane subepithelium on electron microscopy. Evidence suggests that in about 10% of patients with negative PLAR2 (our patient), primary MGN is the result of a still unknown antipodocyte antibody.4 Thus, we hypothesized that our patient's autoantibodies cross‐react with a common BP180 epitope in both the skin and kidney. Indeed, we identified a similar pattern of reactivity to skin and kidney lysates and using epitope mapping we show IgG reactivity specific to AA 1080–1107 of BP180. These findings are consistent with the recently demonstrated BP180 expression in podocytes of normal mouse and human kidney.5 It is possible that our patient had additional renal‐specific antibodies, for example his serum reacted strongly with recombinant integrin α6, expressed in both kidney tubules and skin.6 Alternatively, the concurrence of mucocutaneous and renal involvement here may be because of patient‐specific renal antigen not present on healthy tissue. Anti‐collagen IV antibodies have been implicated in the pathogenesis of subepidermal blisters and renal pathology.1,7,8 In our patient there was not a reaction with collagen IV recombinant proteins and his glomerular lesions did not contain crescents, the expected presentation with collagen type IV antibodies. In summary, the parallel onset, remission and relapses of concurrent MMP and MGN support a common mechanism of mucocutaneous and renal involvement in this patient. Although the precise nature of this association remains to be elucidated, we suggest a common or cross‐reactive antibody plays a role. Possible candidates include autoantibodies to the AA 1080–1107 region of BP180 protein, which are associated with mucosal pathogenicity. We would like to thank the Genito‐Urologic Tissue Repository (GUMER), an Institutional Review Board‐approved biospecimen repository at the University of Iowa, for providing us with healthy human kidney samples. Funding sources: none. Conflicts of interest: none to declare.