M-type phospholipase A2 receptor as target antigen in idiopathic membranous nephropathy

Abstract

Background: Idiopathic membranous nephropathy, a common form of the nephrotic syndrome, is an antibody-mediated autoimmune glomerular disease. Serologic diagnosis has been elusive because the target antigen is unknown.

Methods: We performed Western blotting of protein extracts from normal human glomeruli with serum samples from patients with idiopathic or secondary membranous nephropathy or other proteinuric or autoimmune diseases and from normal controls. We used mass spectrometry to analyze the reactive protein bands and confirmed the identity and location of the target antigen with a monospecific antibody.

Results: Serum samples from 26 of 37 patients (70%) with idiopathic but not secondary membranous nephropathy specifically identified a 185-kD glycoprotein in nonreduced glomerular extract. Mass spectrometry of the reactive protein band detected the M-type phospholipase A(2) receptor (PLA(2)R). Reactive serum specimens recognized recombinant PLA(2)R and bound the same 185-kD glomerular protein as did the monospecific anti-PLA(2)R antibody. Anti-PLA(2)R autoantibodies in serum samples from patients with membranous nephropathy were mainly IgG4, the predominant immunoglobulin subclass in glomerular deposits. PLA(2)R was expressed in podocytes in normal human glomeruli and colocalized with IgG4 in immune deposits in glomeruli of patients with membranous nephropathy. IgG eluted from such deposits in patients with idiopathic membranous nephropathy, but not in those with lupus membranous or IgA nephropathy, recognized PLA(2)R.

Conclusions: A majority of patients with idiopathic membranous nephropathy have antibodies against a conformation-dependent epitope in PLA(2)R. PLA(2)R is present in normal podocytes and in immune deposits in patients with idiopathic membranous nephropathy, indicating that PLA(2)R is a major antigen in this disease.

2009 Massachusetts Medical Society

Figures

Figure 1. Results of Western Blotting of Glomerular Proteins with Serum from Patients with Idiopathic Membranous Nephropathy

The top of Panel A shows the results of Western blotting of extract of human glomerular proteins with serum samples from each of five patients with idiopathic membranous nephropathy (MN1 through MN5) and five patients with other proteinuric conditions (two with focal and segmental glomerulosclerosis [FGS1 and FGS2] and three with diabetic nephropathy [DN1, DN2, and DN3]). Serum samples from the five patients with membranous nephropathy all recognized a band of approximately 185 kD, whereas the samples from the patients with other diseases did not. The bottom of Panel A shows the results of Western blotting, with reactive serum samples from the five patients with membranous nephropathy, of glomerular proteins that were deglycosylated with peptide N-glycosidase F (PNGase F+) or not deglycosylated (PNGase F−). All five samples showed the 185-kD native antigen and a deglycosylated protein of approximately 145 kD. Panel B shows the specificity of the reactivity of serum samples to the 185-kD antigen.

Figure 2. Identification of the 185-kD Antigen in Human Glomeruli

Panel A shows the results of Western blotting of samples of human glomerular extract (HGE) and recombinant phospholipase A2 receptor (rPLA2R), deglycosylated with peptide N-glycosidase F (PNGase F+) or not deglycosylated (PNGase F−), with either a reactive serum sample from a patient with membranous nephropathy (MN) or a polyclonal antibody raised against the M-type phospholipase A2 receptor (PLA2R). The recombinant protein migrated to a slightly lower position than the native glomerular protein, although deglycosylation with peptide N-glycosidase F caused both the recombinant and native isoforms to migrate to the same position. Panel B shows that immunoprecipitation of native PLA2R occurred in a reactive serum sample from each of five patients with membranous nephropathy (MN1, MN2, MN3, MN6, and MN7) but not in a nonreactive sample (MN8) and not in nonreactive serum samples from normal controls (NC1 and NC2) or in the absence of serum. The immunoprecipitates were then electrophoresed under reducing conditions and subjected to Western blotting with antibodies against PLA2R (top) or against human IgG (bottom). In the reactive samples from all five patients with membranous nephropathy, PLA2R was immunoprecipitated from human glomerular extract, whereas there was no immunoprecipitation in nonreactive samples from one patient and from the controls. No PLA2R was detected when human serum was omitted from the reaction. The bottom image shows that the immunoprecipitates of samples from controls and patients with membranous nephropathy contained at least equivalent amounts of IgG.

Figure 3. Characterization of the Antibody Response to Reduced and Nonreduced Phospholipase A2 Receptor (PLA2R)

Panel A shows a Western blot in which equal amounts of human glomerular extract (HGE) and equal amounts of recombinant PLA2R (rPLA2R) were electrophoresed under reducing and nonreducing conditions. Western blotting was performed with the use of a reactive serum sample from a patient with membranous nephropathy (MN5) or a polyclonal anti-PLA2R antibody and detected with appropriate secondary antibodies. Both the serum sample from the patient with membranous nephropathy and the polyclonal anti-PLA2R antibody reacted with the native and recombinant PLA2R in the nonreduced state, but the reduced proteins were reactive with only the anti-PLA2R antibody. Panel B shows the IgG-subclass specificity to PLA2R. HGE was blotted initially with serum samples from six patients with membranous nephropathy (MN1 through MN6), followed by sheep antibodies specific for each human IgG subclass (1 through 4), and was detected with peroxidase-conjugated antisheep IgG antibody. The arrowheads indicate the fully glycosylated native PLA2R. The predominant IgG subclass that reacted with the native antigen was IgG4, with varying amounts of reactivity seen for IgG1, IgG2, and IgG3. Identical results were obtained with the use of recombinant PLA2R instead of HGE (not shown).

Figure 4. Expression of the M-Type Phospholipase A2 Receptor (PLA2R) in Normal Kidney Tissue and Glomeruli

Serial cryosections of the cortex of normal human kidneys were immunostained with anti-PLA2R antibody (1:150 dilution), followed by Cy3-conjugated anti–guinea pig IgG antibody (1:500 dilution). Panel A shows two positively stained glomeruli (arrowheads) and scattered tubular staining (arrow). Panel B shows an adjacent section in which the anti-PLA2R antiserum was preabsorbed with a recombinant fragment of PLA2R containing C-type lectinlike domains (CTLDs) 4, 5, and 6 (Panel H). The glomerular PLA2R staining is completely blocked by the recombinant fragment (arrowheads), whereas the scattered tubular staining (arrow) is not blocked and is therefore nonspecific. Panel C shows a human kidney-tissue specimen treated with nonimmune guinea pig serum (1:200 dilution), followed by Cy3-conjugated anti–guinea pig IgG antibody, and stained with Hoechst 33342 (a nuclear stain); the inset shows a glomerulus exposed instead to anti-PLA2R antibody (1:200 dilution) but otherwise assessed under identical conditions. Next, cryosections of normal human kidney cortex were costained with antiagrin antibody, followed by an Alexa Fluor 488–conjugated antirabbit secondary antibody, to label the glomerular basement membrane (Panels D and E); in addition, PLA2R staining with guinea pig anti-PLA2R antibody was either left unblocked (Panel D) or was blocked (Panel E), as described above. The PLA2R signal is clearly present outside the glomerular basement membrane and localizes to both the cell body and the processes of the podocyte. The PLA2R staining is markedly reduced when the antibody is preabsorbed with the recombinant PLA2R fragment (Panel E). Panel F shows an enlargement of the left lower portion of the glomerulus shown in Panel D, with demarcation of the capillary lumina (asterisks) and urinary space (US) to highlight the location of the PLA2R-stained podocytes. Hoechst-stained nuclei of parietal epithelial cells in the Bowman’s capsule are indicated by the arrow. Panel G shows a portion of a human glomerulus stained with anti-PLA2R antibody (as described above), anti–Wilms’ tumor 1 antibody (WT1) (1:100 dilution), and Alexa Fluor 488–conjugated anti–rabbit IgG antibody (1:500 dilution) to label podocyte nuclei. The inset shows the image of the entire glomerulus from which the enlarged view was taken. Panel H illustrates the domain structure of PLA2R, which is composed of an N-terminal cysteine-rich domain (Cys-R), a fibronectin type II domain (FNII), eight CTLDs, a transmembrane domain (TM), and a short intracellular C-terminal tail (IC). The blocking fragment used in the experiments described above consists of CTLDs 4, 5, and 6 of a recombinant rabbit PLA2R.

Figure 5. Colocalization of the M-Type Phospholipase A2 Receptor (PLA2R) and IgG4 and Reactivity of Eluted IgG4

Confocal microscopic analysis of cryosections of a kidney-tissue specimen from a patient with membranous nephropathy revealed the presence of PLA2R (Panel A) and IgG4 (Panel B), which were colocalized in the peripheral capillary walls and the glomerular basement membrane (Panel C). Panels D, E, and F are enlarged images of the boxed areas in Panels A, B, and C, respectively. The asterisks in Panel F indicate the capillary lumina. The anti-PLA2R antibody was blocked with the recombinant fragment (Panel G), with staining virtually eliminated, despite the continued presence of IgG4 (Panel H). Panel I is a confocal image of representative glomerular capillary loops in a biopsy specimen from a patient with lupus membranous nephropathy; PLA2R and IgG4 are not colocalized. IgG4 was detected with sheep anti–human IgG4 antibody (1:500 dilution) and rabbit anti–sheep IgG antibody (1:500 dilution). IgG was eluted from biopsy cores from patients with membranous nephropathy (MN), lupus membranous nephropathy (LMN), or IgA nephropathy (IgAN). This eluted IgG was used to immunoblot human glomerular extract (HGE) or recombinant PLA2R (rPLA2R) (Panel J). Only IgG eluted from the MN samples identified the native and recombinant PLA2R. Panel K shows that the IgG eluted from the MN3 biopsy sample recognized only those bands corresponding to PLA2R.

Figure 6. Antibody against the M-Type Phospholipase A2 Receptor (PLA2R) and Disease Activity in a Patient with Membranous Nephropathy

Serum samples were collected serially from a patient with membranous nephropathy who had clinical remission after receiving immunosuppressive treatment with cyclophosphamide and prednisone (see the Clinical Vignettes section in the Supplementary Appendix). Panel A shows a decline in the urinary protein level (as measured by the protein-to-creatinine ratio) and an increase in the serum albumin level. The results of Western blotting of serum samples taken from the patient at the same times as the laboratory measurements (Panel B) show reactivity to PLA2R, either the 185-kD native form not deglycosylated by peptide N-glycosidase F (PNGase F−) or the 145-kD deglycosylated form (PNGase F+) in the initial serum sample (from December 2005) only. Equal loading is verified by the nonspecific detection of a 98-kD band. Equal volumes of the four serial serum samples were electrophoresed and blotted with anti–human IgG antibody (bottom), with a progressive increase in total IgG after the patient entered a period of remission from membranous nephropathy.