Activation of a Latent Epitope Causing Differential Binding of Antineutrophil Cytoplasmic Antibodies to Proteinase 3

Abstract

Objective: Proteinase 3 (PR3) is the major antigen for antineutrophil cytoplasmic antibodies (ANCAs) in the systemic autoimmune vasculitis, granulomatosis with polyangiitis (GPA). PR3-targeting ANCAs (PR3-ANCAs) recognize different epitopes on PR3. This study was undertaken to study the effect of mutations on PR3 antigenicity.

Methods: The recombinant PR3 variants, iPR3 (clinically used to detect PR3-ANCAs) and iHm5 (containing 3 point mutations in epitopes 1 and 5 generated for epitope mapping studies) immunoassays and serum samples from patients enrolled in ANCA-associated vasculitis (AAV) trials were used to screen for differential PR3-ANCA binding. A patient-derived monoclonal ANCA 518 (moANCA518) that selectively binds to iHm5 within the mutation-free epitope 3 and is distant from the point mutations of iHm5 was used as a gauge for remote epitope activation. Selective binding was determined using inhibition experiments.

Results: Rather than reduced binding of PR3-ANCAs to iHm5, we found substantially increased binding of the majority of PR3-ANCAs to iHm5 compared to iPR3. This differential binding of PR3-ANCA to iHm5 is similar to the selective moANCA518 binding to iHm5. Binding of iPR3 to monoclonal antibody MCPR3-2 also induced recognition by moANCA518.

Conclusion: The preferential binding of PR3-ANCAs from patients, such as the selective binding of moANCA518 to iHm5, is conferred by increased antigenicity of epitope 3 on iHm5. This can also be induced on iPR3 when captured by monoclonal antibody MCPR2. This previously unrecognized characteristic of PR3-ANCA interactions with its target antigen has implications for studying antibody-mediated autoimmune diseases, understanding variable performance characteristics of immunoassays, and design of potential novel treatment approaches.

Fig. 1 –. Cartoon model of human proteinase 3 (PR3)*.

The left panel shows PR3 in the standard orientation facing the active site pocket; the right panel shows PR3 with an approximately 90 degree rotation. The catalytic triad of this neutrophil serine protease comprises His57, Asp102, and Ser195. Ala146, Trp218, and Leu223 in PR3 shown here are replaced by their hydrophilic murine counterparts (Thr146, Arg218, and Gln223) in iHm5. Shown in red is Epitope 3 on the opposite side of the PR3 structure from the side where the mutations were introduced. (*Generated using PyMOL v. 1.7.0.3, Schroedinger, LLC; amino acid numbering is based on Fujinaga M. et al. J. Mol. Biol. 1996; 261:267-78)

Fig. 2 –. Reactivity of PR3-ANCA positive serum samples with iHm5 and iPR3.

Representative examples of serum samples from patients with PR3-ANCA associated vasculitis showing higher (A and B) or equal (C and D) binding to iHm5 compared to iPR3 in the anchor ELISA.

Fig. 3 –. PR3-ANCA reactivity with iHm5 and iPR3 determined by anchor ELISA using serum from patients included in the WGET and in the RAVE trials.

Scatter plot of the net absorbance values of iHm5 recognition plotted against iPR3 recognition using the anchor ELISA described with plasma and sera from patients of two AAV cohorts, the WGET (A) and the RAVE (B), respectively. There is a shift of the net absorbance values towards the binding of patient PR3-ANCA to iHm5. (Abbreviations: OD - optical density; PR3 - proteinase 3; RAVE – rituximab vs. cyclophosphamide for remission induction in AAV; WGET - Etanercept vs. placebo for remission induction in AAV)

Fig. 4 –. Identification of Epitope 3 of iHm5 as a target of moANCA518 and PR3-ANCA.

A. Inhibition of moANCA518 binding to iHm5 using different moAbs to capture antigen iHm5. Binding of moANCA518 to iHm5 was impaired significantly when the moAbs MCPR3-3 and WGM2 (recognizing Epitope 3) were used to capture the antigen, compared to PR3G-2 which binds (recognizing Epitope 1). This previously shown data (32) is included again here to illustrate similarities with the PR3-ANCA serum results (panel B). B. Serum PR3-ANCA from the same patient and study visit from which moANCA518 was derived from plasmablasts yields similar differential reactivity of PR3-ANCA with iHm5 captured by the different moAbs Results for purified iHm5 antigen [black bars] or crude iHm5 contained in media supernatant from cells expressing the recombinant antigen [white bars] were similar. C. Fabs from epitope-specific moAbs as competitors of binding of moANCA518 to iHm5 using the anchor ELISA. Compared to control, binding of moANCA518 to iHm5 was reduced significantly by moANCA518 Fab (p=0.005) and Fabs MCPR3-3 (p=0.005) and WGM2 (p=0.01) but not Fabs from MCPR3-7 (recognizing Epitope 5). D. Binding of PR3-ANCA to iHm5 in the anchor ELISA was similarly inhibited by Fabs recognizing Epitope 3 (MCPR3-3 and moANCA518). Comparison in panel C and D are with control mouse IgG1 Fab.

Fig. 5 –. Binding to monoclonal antibodies facilitates iPR3 recognition by the moANCA518.

A. Detection of the binding of moANCA518 to iHm5 and iPR3 using anchor ELISA and MCPR3-2 capture ELISA. Binding of the moANCA518 to iHm5 was detected by anchor ELISA, whereas no binding to iPR3 could be detected (p=0.018). In contrast, on the MCPR3-2 capture ELISA, the moANCA518 bound to both constructs, iHm5 and iPR3, but the signal of the binding to iHm5 was higher when compared with the signal of the binding to iPR3 (p=0.005). B. Control experiments documenting comparable coating of iPR3 and iHm5, on the anchor ELISA, and comparable coating of MCRP3-2 bound to iPR3 and iHm5, on the MCPR3-2 capture ELISA (p=ns) detected by polyclonal rabbit anti-PR3 antibody. (Abbreviations: ELISA - enzyme-linked immunosorbent assay; PR3 - proteinase 3)