Identification of the immunodominant epitope region in phospholipase A2 receptor-mediating autoantibody binding in idiopathic membranous nephropathy

Abstract

Membranous nephropathy (MN) is a common cause of nephrotic syndrome in adults. Recent clinical studies established that >70% of patients with idiopathic (also called primary) MN (IMN) possess circulating autoantibodies targeting the M-type phospholipase A2 receptor-1 (PLA2R) on the surface of glomerular visceral epithelial cells (podocytes). In situ, these autoantibodies trigger the formation of immune complexes, which are hypothesized to cause enhanced glomerular permeability to plasma proteins. Indeed, the level of autoantibody in circulation correlates with the severity of proteinuria in patients. The autoantibody only recognizes the nonreduced form of PLA2R, suggesting that disulfide bonds determine the antigenic epitope conformation. Here, we identified the immunodominant epitope region in PLA2R by probing isolated truncated PLA2R extracellular domains with sera from patients with IMN that contain anti-PLA2R autoantibodies. Patient sera specifically recognized a protein complex consisting of the cysteine-rich (CysR), fibronectin-like type II (FnII), and C-type lectin-like domain 1 (CTLD1) domains of PLA2R only under nonreducing conditions. Moreover, absence of either the CysR or CTLD1 domain prevented autoantibody recognition of the remaining domains. Additional analysis suggested that this three-domain complex contains at least one disulfide bond required for conformational configuration and autoantibody binding. Notably, the three-domain complex completely blocked the reactivity of autoantibodies from patient sera with the full-length PLA2R, and the reactivity of patient sera with the three-domain complex on immunoblots equaled the reactivity with full-length PLA2R. These results indicate that the immunodominant epitope in PLA2R is exclusively located in the CysR-FnII-CTLD1 region.

Keywords: immunology and pathology; membranous nephropathy; pathophysiology of renal disease and progression.

Figure 1.

Structure model of PLA₂R. PLA₂R protein consists of a larger extracellular portion, a single transmembrane region, and a short cytoplasmic tail. The large extracellular portion contains 10 domains: a CysR (dark gray hexagon), an FnII (gray square), and 8 repeated CTLD (light gray ovals) domains. The polymorphisms that may link to the occurrence of IMN are indicated as white circles. The positions of amino acids used for designing experimental constructs are indicated by arrows.

Figure 2.

Design and expression of truncated PLA₂R extracellular domains. (A) The cartoon shows the constructs containing the truncated PLA₂R extracellular domains. Dark gray hexagon, CysR; gray line, 1D4 tag; gray square, FnII; light gray oval, CTLD. (B) The expressed truncated PLA₂R extracellular domains were concentrated and resolved on a 4%–20% SDS-PAGE under nonreducing conditions. Protein samples were then transferred to a nitrocellulose membrane and detected with a mouse anti-1D4 mAb. The experiment was performed at least three times.

Figure 3.

Characterization of patient serum containing anti-PLA₂R autoantibodies. HEK 293 cells expressing the full-length PLA₂R protein were lysed in the lysis buffer containing protease inhibitors and resolved under the nonreducing and reduced conditions. Protein samples were transferred to a nitrocellulose membrane and probed with the patient serum at 1:1000 dilutions in the immunoblotting buffer. The membrane was then stripped and reprobed with a rabbit anti-PLA₂R antibody (Sigma-Aldrich) to determine the level of protein in each sample. The assay was performed at least three times.

Figure 4.

Immunodetection of truncated PLA₂R extracellular domains with the characterized patient serum containing anti-PLA₂R autoantibodies. (A) The transferred nitrocellulose membrane was incubated with the characterized patient serum at 1:1000 dilutions in immunoblotting buffer for 2 hours at room temperature. The membrane was then washed with TBST and incubated with HRP-conjugated rabbit anti-human secondary antibody (1:10,000) in immunoblotting buffer. (B) The nitrocellulose membrane was stripped and then reprobed with a mouse anti-1D4 antibody. The experiment was performed three to five times.

Figure 5.

Characterization of the isolated PLA₂R 1–3 construct. The isolated 1–2 and 1–3 constructs were resolved under nonreducing and reduced conditions (2% β-ME) on a 4%–20% SDS-PAGE. The 1–2 construct was included to serve as a negative control. The membrane was then probed with the characterized patient serum and reprobed with the anti-1D4 antibody as described in Figure 4. The experiment was performed three to five times.

Figure 6.

Identification of the domains in isolated PLA₂R 1–3 construct responsible for anti-PLA₂R autoantibody recognition. (A) The cartoon indicates the locations of introduced thrombin digestion sites in the 1–3 construct. 1–1T is the thrombin digestion site that was introduced between CysR and FnII domains, and 1–2T is the thrombin digestion site that was introduced between FnII and CTLD1 domains. (B and C) The thrombin untreated and treated 1–3, 1–1T, and 1–2T constructs were resolved under nonreducing conditions on a 4%–20% SDS-PAGE. The 1–3 construct was included to serve as a positive control. The membrane was (B) first probed with the characterized patient serum and (C) then reprobed with the anti-1D4 antibody as described in Figure 4. Asterisks indicate thrombin-treated protein samples. (D) The thrombin untreated and treated 1–3, 1–1T, and 1–2T constructs were resolved under reducing conditions (2% β-ME) on a 4%–20% SDS-PAGE and probed with the anti-1D4 antibody. Each experiment was performed three to five times.

Figure 7.

Immunoblocking of the autoantibody reaction with the full-length PLA₂R by the 1–3 construct. (A) The 1–3 construct was stepwise diluted in the TBS buffer and incubated with an anti-PLA₂R serum for 2 hours at room temperature before application on a blot transferred with nonreduced full-length PLA₂R protein. Control, 1–10 construct (undiluted) blocked anti-PLA₂R serum; 1, TBS buffer only; 2–10, 1–3 construct at dilutions (vol/vol) of 1:600, 1:300, 1:150, 1:75, 1:37.5, 1:18.75, 1:9.38, 1:4.69, and 1:2.35, respectively. (B) Sera from 10 patients with high levels of anti-PLA₂R autoantibodies were first incubated with the 1–3 construct at 1:9 dilutions (vol/vol) for 2 hours at room temperature and then applied on the full-length PLA₂R protein on a blot (−, without the 1–3 construct; +, with the 1–3 construct). (C) Summary of the 1–3 construct blocked sera reactivity with the full-length PLA₂R. Error bars represent the means±SEMs (n=5).

Figure 8.

Immunodetection of the 1–3 construct or CTLD1 domain-deleted PLA₂R constructs with anti-PLA₂R sera from patients with IMN. (A) The cartoon shows the constructs with the 1–3 construct or CTLD1 domain deletion and a full-length PLA₂R with an introduced thrombin digestion site between CTLD1 and CTLD2 domains. Deleted domains are indicated by unfilled shapes. (B) Truncated PLA₂R extracellular domains 1–2, 1–3, 1–8 with 1–3 deletion (1–8*), 1–8 with CTLD1 deletion (1–8**), 1–10 with 1–3 deletion (1–10*), and 1–10 with CTLD1 deletion (1–10**) were (left panel) resolved under nonreducing condition and probed with the characterized patient serum. The 1–2 and 1–3 constructs were used as a negative and a positive control, respectively. (Right panel) The level of protein in each sample was determined by the anti-1D4 antibody. (C) Full-length PLA₂R, PLA₂R-T, and thrombin-digested PLA₂R-T (PLA₂R-T*) were resolved under the nonreducing condition and processed as described above. (Left panel) Anti-PLA₂R serum. (Right panel) Anti-PLA₂R antibody (Sigma-Aldrich). (D) Summary of anti-PLA₂R sera reactivity with PLA₂R, PLA₂R-T*, 1–8*, and 1–10* (+++, strongly recognized; −, background level). Each experiment was performed three to five times.

Figure 9.

Immunodetection of PLA₂R with polymorphism M292V and H300D. (A) Whole-cell lysates expressing wild-type PLA₂R (wt-PLA₂R), PLA₂R-M292V, or PLA₂R-H300D were resolved under nonreducing conditions and probed with the characterized patient serum at three dilutions in immunoblotting buffer. (B) The level of antibody binding to each sample was quantified by densitometry. In each experiment, the level of antibody binding was compared with that of wt-PLA₂R, which had antibody binding set to 100%. Error bars represent the means±SEMs (n=3–5).

Figure 10.

Representative results of immunoscreen of patient samples with the PLA₂R 1–3 construct and the full-length PLA₂R. (A) HEK 293 cell lysate containing heterologously expressed full-length PLA₂R protein and the PLA₂R 1–3 construct was resolved under nonreducing condition and transferred to nitrocellulose membranes. After 30 minutes of blocking with immunoblotting buffer, the membranes were assembled into a multiscreen apparatus and individually incubated with each of the patient samples at 1:100 dilutions for 2 hours at room temperature. (B) Representative results of full-length PLA₂R and the 1–3 construct immunoblotted with the characterized patient serum at three dilutions: 1:100, 1:1000, and 1:10,000 (vol/vol). The assay was performed three times.