IgG anti-laminin-332 autoantibodies are present in a subset of patients with mucous membrane, but not bullous, pemphigoid

Abstract

Background: Antiepiligrin cicatricial pemphigoid is a mucosal-predominant subepidermal blistering disease associated with an increased relative risk of cancer. In contrast to prior reports showing that anti-laminin (L)-332 autoantibodies are a reliable marker for patients with antiepiligrin cicatricial pemphigoid, a recent report suggested that as many as 40% of patients with bullous pemphigoid (BP) have IgG reactive with this laminin isoform.

Objective: We sought to determine whether patients with BP possess circulating IgG anti-L-332 autoantibodies.

Methods: Sera from 100 adults with BP were analyzed by indirect immunofluorescence testing of intact skin, immunoblot studies of human keratinocyte (HK) extracts, and a new L-332 enzyme-linked immunosorbent assay. Sera showing reactivity suggestive of anti-L-332 autoantibodies in these assays were further analyzed in immunoblot studies of HK extracellular matrix and immunoprecipitation studies of biosynthetically radiolabeled HK extracts.

Results: IgG from all patients with BP bound intact epidermal basement membrane by indirect immunofluorescence microscopy and immunoblotted bullous pemphigoid antigen-1, -2, or both in HK extracts. None of these sera immunoblotted L-332 in HK extracts, although 13 did score above the cut point of a new IgG(4) L-332 enzyme-linked immunosorbent assay (sensitivity = 0.91, specificity = 0.98, Youden index = 0.89). Further analysis of sera from these 13 patients found: (1) all had IgG that bound the epidermal side of 1 mol/L NaCl split skin by indirect immunofluorescence microscopy; (2) none immunoblotted L-332 purified from HK extracellular matrix; and (3) none immunoprecipitated L-332 from biosynthetically radiolabeled HK extracts.

Limitations: The basis of false-positive enzyme-linked immunosorbent assay determinations for anti-L-332 IgG among patients with BP is unknown.

Conclusion: Anti-L-332 autoantibodies remain a reliable marker for patients with antiepiligrin cicatricial pemphigoid.

Figure 1

The IgG₄ anti-L-332 ELISA detects autoantibodies in patients with AECP with great sensitivity and specificity. Scatter plot representation of ELISA results concerning the performance of sera from 32 patients with AECP and 173 controls. All sera were tested in duplicate; plotted dots represent the average of the OD₄₉₀ reading obtained for each sample. The dashed line indicates the cut point score determined by ROC methodology. The insert depicts the ROC curve of the assay (sensitivity 0.91 [confidence interval 0.75 – 0.98], specificity 0.98 [confidence interval 0.95 – 0.99], Youden index 0.89).

Figure 2

Sera from patients with BP do not contain IgG autoantibodies directed against L-332. a. Sera from 13 BP patients scoring above the cut point in the IgG₄ anti-L-332 ELISA (lanes 1-13) immunoblotted either BPAG1 or BPAG2 in HK extracts. None of these samples displayed reactivity against L-332. Lane L represents a positive control where L-332 was immunoblotted with a well characterized rabbit antiserum and developed with alkaline phosphatase-conjugated goat anti-rabbit IgG (Biosource); lane N represents a negative control where L-332 was immunoblotted with normal human serum and developed with the same second-step antibody used in immunoblot studies of sera from patients with BP. Ticks in the left margin correspond to BPAG1 and BPAG2; brackets in the right margin correspond to various L-332 subunits (specifically, the unprocessed and processed α subunit, the unprocessed and processed γ subunit, and the β subunit). b. Sera from 13 BP patients scoring above the cut point in the IgG₄ anti-L-332 ELISA (lanes 1-13) showed no evidence of reactivity to L-332 in HK ECM (the most sensitive antigen source for detection of autoantibodies in patients with AECP) . Lanes L and N as well as the brackets in the right margin are the same as noted in the description of panel a. c. Sera from all BP patients scoring above the cut point in the IgG₄ anti-L-332 ELISA (lanes 1-13) immunoprecipitated either BPAG1 and/or BPAG2 from extracts of biosynthetically radiolabeled HKs. None of these samples immunoprecipitated L-332 from this antigen source (the most sensitive known technique for detection of autoantibodies in patients with AECP) . Lane L represents a positive control where L-332 was immunoprecipitated with a well characterized rabbit antiserum. Annotations in the panel margins are the same as those depicted in panel a.