Renal Phospholipase A₂ Receptor in Hepatitis B Virus-Associated Membranous Nephropathy

Since M-type phospholipase A₂ receptor (PLA₂R) was characterized as a major target antigen of idiopathic membranous nephropathy (MN) in 2009 [1], numerous studies have been conducted on renal PLA₂R and its associated autoantibodies [2,3,4,5,6,7]. Serum anti-PLA₂R antibodies (anti-PLA₂R-Ab) are detected in the majority of patients with this disease, and the antibody titer is associated with disease activity and long-term outcome [5,8,9,10]. Patients with detectable serum PLA₂R-Ab also exhibit strong positive staining for PLA₂R in the kidney [11,12]. PLA₂R is a glycoprotein, and its expression has been reported in the podocytes of the kidney [1]. A recent study using a specific antibody shows that PLA₂R is detected in the kidney in the majority of patients with idiopathic MN, but not in non-MN patients or normal control patients [11]. It is hypothesized that the antigenicity of PLA₂R on podocytes may be altered in these PLA₂R positive patients. Although several studies showed that the polymorphisms of PLA₂R and HLA complex class II HLA-DQ alpha chain 1 (HLA-DQA1) were associated with idiopathic MN [13,14], the cause of the change in renal PLA₂R and the production of PLA₂R-Ab remain unclear.

It has been reported that positive staining for PLA₂R in glomeruli strongly correlates with the presence of PLA₂R-Ab in the serum [11]. Secondary causes of MN were not found in any of the patients with positive renal PLA₂R but were found in more than half of the patients with negative renal PLA₂R, suggesting that renal PLA₂R staining may be useful for discriminating idiopathic from secondary MN [11]. However, other studies showed that positive renal PLA₂R could be found in secondary MN, such as hepatitis virus C- or malignant tumor-associated MN [15] and that serum PLA₂R-Ab could also be detected in a small number of these patients [6]. Therefore, whether renal PLA₂R and serum PLA₂R-Ab are biomarkers for idiopathic MN remain controversial.

Hepatitis B virus (HBV) infection is prevalent worldwide, especially in China, and HBV-associated MN (HBV-MN) is one of the most common causes of secondary MN [16,17,18,19]. Therefore, the aim of our study was to (1) examine whether renal PLA₂R could discriminate between primary and secondary MN or aid in determining the secondary cause of MN, (2) investigate renal PLA₂R in HBV-MN patients and (3) compare the pathological characteristics of the renal PLA₂R positive HBV-MN patients and idiopathic MN patients. We report that almost two-thirds of the HBV-MN patients were found to be renal PLA₂R positive, and serum PLA₂R-Ab was also detected in the renal PLA₂R positive patients who had the serum available. The association of HBV with MN may also provide clues to aid our understanding of the pathogenesis of MN.

During 2008-2013, 179 patients with biopsy-proven MN from Huashan Hospital (a tertiary hospital in Shanghai, China) were included in this retrospective study. Detailed clinical data including medical history, physical examination, serological analysis (anti-nuclear, extractable nuclear antigen, double stranded DNA and anti-neutrophil cytoplasmic antibodies; anti-HBV, hepatitis C virus and human immunodeficiency virus antibodies; and tumor markers), chest X-ray and abdominal ultrasound, in addition to renal biopsies, were collected. Among these patients, 102 patients without identifiable secondary causes at the time of biopsy were diagnosed with idiopathic MN, 38 patients were diagnosed with class V lupus nephritis (LN-V) and 39 patients were diagnosed with HBV-MN based on the following criteria: (1) evidence of serum positivity for HBsAg, (2) presence of glomerulonephritis, excluding LN and other secondary glomerular diseases and (3) presence of HBV antigens, including HBsAg or HBcAg, or HBV-DNA detected in glomeruli [20]. In addition, non-MN subjects including 40 patients with primary glomerulonephritis, 31 patients with focal or diffuse LN and 17 patients with HBV-associated membranoproliferative glomerulonephritis or IgA nephropathy were also included as controls. This study was approved by the Ethics Committee of Huashan Hospital, Fudan University.

A standard renal biopsy procedure including light, immunofluorescence (IF) and electron microscopy was used for renal pathological diagnosis. Direct IF was used for IgG, IgA, IgM, C3, C4 and C1q measurements on frozen sections of fresh tissue and scaled from 0 to 4+. Electron microscopy were used to identify dense deposits and scaled from 0 to 3+. HBV biomarkers, including HBsAg (Monoclonal Mouse Anti-HBsAg from Gene Tech, GT202429) and HBcAg (Polyclonal Rabbit Anti-HBcAg from Gene Tech, GB058629), were regularly detected in the kidney of the patients with positive serum HBsAg using frozen sections by indirect IF.

Renal PLA₂R detection and the co-staining of PLA₂R and HBsAg in the glomeruli were performed on paraffin-embedded renal biopsy samples. Citrate buffer of pH 6.0 and microwaving at 100% power for 8 min were used for antigen retrieval. Three percent bovine serum albumin was used for blocking. The anti-PLA₂R-Ab (Sigma, HPA012657) was diluted at 1:500 and incubated overnight at 4°C. The secondary antibody was a fluorescein Cy3-conjugated donkey anti-rabbit IgG antibody (Jackson, 711-165-152) and diluted at 1:200. Positive PLA₂R was characterized as granular staining along the capillary loops and scaled from 0 to 3+ (fig. 1). The anti-HBsAg antibody (Gene Tech, GT202429) was diluted at 1:50 and co-incubated with anti-PLA₂R-Ab overnight at 4°C, and then detected by an Alexa-green G488-conjugated donkey anti-mouse IgG antibody (Jackson, 715-545-150). Negative control (secondary antibody only) was used in every case to exclude the cross-reactions between secondary antibody and human IgG.

Circulating PLA₂R-Ab was detected by indirect IF. HEK293 cells were cultured on glass coverslips that had been pretreated for cell culture. Twenty-four hours later, the cells were transiently transfected with a plasmid containing complementary DNA encoding a full-length PLA₂R isoform 1 (origene, accn: NM_007366.3). Forty-eight hours after transfection, cells cultured on coverslips were fixed with cold acetone. Serum samples were diluted 1:10, 1:100 and 1:1,000 in phosphate-buffered saline, applied to the fixed transfected cells and incubated overnight at 4°C to determine whether the serum was positive for PLA₂R-Ab. A fluorescein isothiocyanate (FITC)-conjugated anti-human IgG antibody from mice (Gene Tech, GF020229) with 1:100 dilution was used for detection of bound IgG antibodies. Transfected cells with specific cytomembrane fluorescence were considered to be positive. Commercial anti-PLA₂R-Ab was used for positive control, and normal serum and secondary antibodies only were used for negative control. Inter-group control was run by the same serum sample (aliquots) each time to make sure the assay is stable and consistent. All the slides were read by a single person who was blinded to the nature of samples (fig. 2).

Chi-square analysis was performed to compare the difference between PLA₂R positive and PLA₂R negative idiopathic MN patients, between PLA₂R positive and PLA₂R negative HBV-MN patients and between PLA₂R positive idiopathic MN patients and PLA₂R positive HBV-MN patients. The differences with a p value <0.05 were considered significant.

In a series of 102 idiopathic paraffin-embedded MN biopsy tissues, 86 (84%) showed granular staining of PLA₂R along the capillary loops. To further determine whether PLA₂R positivity is specific to idiopathic MN, we also examined PLA₂R in paraffin-embedded kidney samples from patients with non-MN primary glomerular disease, including IgA nephropathy (n = 24), focal segmental glomerulosclerosis (n = 8) and minimal change disease (n = 8), and secondary MN, including HBV-MN (n = 39) and LN (class V, n = 38; table 1).

None of the patients with non-MN primary glomerular diseases (n = 40) exhibited positive renal PLA₂R. In LN patients with membranous histology (class V lesions), only 1 of the 38 patients exhibited positive staining of renal PLA₂R. However, in the HBV-MN patients, positive renal PLA₂R was present in 64% of patients (25 of 39), and there was no significant difference in PLA₂R intensity between PLA₂R positive idiopathic MN and PLA₂R positive HBV-MN (table 2). In addition, IF co-localization study showed that renal PLA₂R staining was overlapped with HBsAg along the capillary loop in HBV-MN patients, while HBsAg staining was negative in the idiopathic MN kidneys (fig. 3).

Among the HBV-MN patients, 76.9% (30 of 39) had mesangial electron dense deposits (EDs) and all the patients had positive HBsAg and/or HBcAg staining. In 79.5% (31 of 39) of HBV-MN patients, HBsAg exhibited granular staining along the capillary loop. IgA and IgM IF were more often detected in HBV-MN patients compared with idiopathic MN patients, a feature of HBV-associated nephropathy. Compared with PLA₂R negative kidney biopsies from HBV-MN patients, stronger IgG and weaker C1q and C4 staining was found in PLA₂R positive renal tissues. There was no significant difference in the subepithelial EDs between these 2 groups. We then compared the renal histology between PLA₂R positive HBV-MN patients and PLA₂R positive idiopathic MN patients. The results showed that the rates of IgA, IgM and mesangial EDs positivity were significantly higher in HBV-MN than in idiopathic MN patients, supporting the secondary type of HBV-MN. There were no differences in IgG, C3 and subepithelial EDs staining between these 2 renal PLA₂R positive groups (table 2).

Serum samples were available in 41 of the 102 idiopathic MN patients, 6 of the 39 HBV-MN patients, 13 of the 38 LN-V patients and 40 of the 88 non-MN patients. All the serum samples were collected at the same time as the renal biopsy, and neither glucocorticoids nor immunosuppressants had been used. To examine the association between serum PLA₂R-Ab and renal PLA₂R expression, the 41 idiopathic MN cases with both serum PLA₂R-Ab and renal biopsy PLA₂R (paraffin-embedded) data available were used. Among them, 24 were serum-positive (59%, 5 with 1:10+, 8 with 1:100+, 11 with 1:1,000+) and 17 were serum-negative (41%) for PLA₂R-Ab. Of the 24 serum PLA₂R-Ab positive cases, all showed positive staining for renal PLA₂R. Among the 17 serum PLA₂R-Ab negative cases, 9 exhibited positive staining for kidney PLA₂R (table 3). Interestingly, in 6 HBV-MN patients with serum available, all showed positive PLA₂R-Ab (2 with 1:100+ and 4 with 1:1,000+) and positive staining for kidney PLA₂R. In addition, the 13 LN-V patients and the 40 non-MN patients (including 15 IgA nephropathy, 8 focal segmental glomerulosclerosis, 5 minimal change disease and 12 non-MN LN) tested negative for PLA₂R-Ab.

The positivity of serum PLA₂R-Ab was significant associated with the severity of nephrotic syndrome. Compared with serum PLA₂R-Ab negative individuals, patients with serum PLA₂R-Ab positive had lower serum albumin (p < 0.001), higher urine protein excretion (p = 0.006) and higher serum creatinine (p = 0.161). In the 9 patients with positive renal PLA₂R, but negative serum PLA₂R-Ab, serum albumin levels were >2.5 g/dl. However, in the 24 patients with both positive renal PLA₂R and serum PLA₂R-Ab, serum albumin levels were <2.5 g/dl in 18 patients (75%) and <2 g/dl in 13 patients (54%; fig. 4).

Circulating PLA₂R-Ab has been reported to be a valuable biomarker for idiopathic MN [1,6]. The present study characterized the staining of PLA₂R in the kidney and showed that renal PLA₂R was detectable in 84% of patients with idiopathic MN but not in the patients with non-MN diseases; among the patients with secondary MNs, renal PLA₂R was detected in 64% of patients with HBV-MN but only in 2.6% of patients with LN-V. Although renal PLA₂R is not exclusively detected in primary glomerular disease, it appears to be able to discriminate between HBV-MN and lupus-associated MN, both of which could be present together and need to be differentiated. In addition, renal PLA₂R positive was present in all the patients with positive serum PLA₂R-Ab, which suggested that renal PLA₂R positivity was associated with the production of serum PLA₂R-Ab that reflected the disease activity.

The finding of renal PLA₂R positivity in idiopathic MN but not in lupus-associated MN was consistent with an earlier study in which 75% (64 of 85) of idiopathic MN, but only 1 of the 46 MN patients with connective tissue disorders were PLA₂R positive [15]. The previous reports regarding PLA₂R in HBV-MN were limited. We could only identify 3 publications with a total of 6 HBV-MN patients whose renal PLA₂R had been examined [11,15,21]. Both of the 2 HBV-MN cases from the publication by Svobodova et al. [21] showed renal PLA₂R positive staining. The other 2 publications including 1 and 3 HBV-MN cases did not find positive renal PLA₂R [11,15]. China is one of the countries with high infection rate of HBV. It has been reported that about 10% of the population developed chronic HBV infection and the total infection (infected with the virus at some time during their lives) was as high as 60% [22]. In our study, we had a total of 39 patients with HBV-MN and 25 (64%) showed positive renal PLA₂R. Interestingly, this positive rate was similar to that in HCV-associated MN patients (6 of 11) reported by Larsen et al. [15]. Importantly, we also found that 6 of the 6 HBV-MN patients with positive renal PLA₂R and serum available were serum-positive for PLA₂R-Ab. This result strongly suggests an association between PLA₂R-Ab and HBV-MN. The patients with HBV-MN in the present study do not seem to coincidentally have HBV infection and idiopathic MN for the following reasons: (1) HBV infection has been well documented to be associated with MN [18,19,23], (2) in addition to typical membranous changes in renal biopsy histology, HBV antigen deposits (HBsAg, HBcAg and HBeAg) were found in the kidneys, particularly overlapping with renal PLA₂R along the capillary loop, in all of these patients and (3) the finding of mesangial EDs and more profound C1q, C4, IgA and IgM IF, in addition to the distribution of IgG and C3 along capillary loops, supports the presence of a secondary type of MN in the kidneys of these patients. All these findings suggest that the secondary MN caused by HBV infection may be a PLA₂R-associated disease. Studies on how the virus leads to the production of serum PLA₂R-Ab will be helpful for understanding the mechanism of idiopathic MN.

The mechanism by which HBV-MN develops is not completely understood. PLA₂R is a glycoprotein constituent of normal human glomeruli and is located on podocytes [1]. However, of particular interest was the finding in other studies that renal PLA₂R was positive in idiopathic MN patients but negative in non-MN or normal controls using a PLA₂R-specific antibody [11]. In our study, positive renal PLA₂R was found in both HBV-associated and idiopathic MN patients, which suggested that these 2 different diseases may share the same mechanism. In PLA₂R-associated HBV-MN, PLA₂R IF exhibited a granular staining along the capillary loops and was co-localized with HBsAg, which suggests that PLA₂R and HBV antigen may have cross-talk in the glomerular capillary wall and then lead to podocyte injury. However, the mechanism by which HBV induces the PLA₂R autoantibody production or/and positive renal PLA₂R remain to be explored.

The present study shows that all the patients with positive serum PLA₂R-Ab have positive renal tissue PLA₂R staining. However, we also have 9 patients with positive renal PLA₂R but negative serum PLA₂R-Ab. This may suggest to us that podocyte PLA₂R exposure be a signature of PLA₂R-associated MN. The negative serum PLA₂R-Ab may reflect altered immune response induced by immunosuppressive treatment or an unknown mechanism that is responsible for automatic remission frequently seen in patients with MN. The present study also shows that positivity of serum PLA₂R-Ab is associated with the severity of nephrotic syndrome. This result is consistent with previous studies [5,8]. These results may support the hypothesis that the change in the PLA₂R in the glomeruli is likely the initiating event, with the serum autoimmune antibodies being a consequence of the podocyte PLA₂R antigen expansion and the antibody concentration reflecting disease activity. Whether PLA₂R conformational change or autoantibody binding may alter its function leading to podocyte injury seen in MN remains to be investigated.

In conclusion, we found that renal PLA₂R was positive in two-thirds of patients with HBV-MN as well as in the majority of patients with idiopathic MN, while the PLA₂R was not detectable in most of the lupus-associated MN. The presence of serum PLA₂R-Ab was associated with the severity of nephrotic syndrome in PLA₂R-associated MN. These results provide valuable information for understanding the mechanism of PLA₂R-associatied MN.

This study was supported by a grant from the Health Industry Special Scientific Research Projects, Ministry of Health (ID: 201002010), and the National Science and Technology Ministry (ID: 2011BAI10B05). Q. Xie was supported by Natural Science Foundation of China (NSFC: 81300569). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Investigators declare no other conflicts of financial interest for participation in this research.

All authors had access to the data and played a role in writing this article.

Q.X. and Y.L. contributed equally to the study.