The anti-PLA2R antibody in membranous nephropathy: What we know and

what remains a decade after its discovery.

Anne-Els van de Logt¹, Maryline Fresquet², Jack F. Wetzels¹*, Paul Brenchley3*.

1 Radboud university medical center, Radboud Institute for Health Sciences, Department of Nephrology, Nijmegen, The Netherlands.

2 Welcome Centre for Cell-Matrix Research, Faculty of Biology, Medicine and Health,University of Manchester, UK. 3 Institute of Cardiovascular Sciences, Faculty of Medicine, Biology and Health, University of Manchester, Oxford Road, Manchester M13 9PW, UK.

*These authors contributed equally

Word count: 5972

Corresponding author:

E mail: Anne-Els.vandelogt@radboudumc.nl

1

1

2

3

456789

10111213141516171819

20

21

22

23

24

25

26272829303132333435363738

12

Abstract:The discovery in 2009 of the M-type phospholipase A2 receptor (PLA2R) as the primary target in membranous nephropathy (MN) greatly advanced basic and clinical research. Primary MN is now considered a renal-limited autoimmune disease, with antibodies against PLA2R (aPLA2Rab) identified in 70-80 % of patients of various ethnic groups. Although the use of aPLA2Rab as a diagnostic and prognostic biomarker is now widely accepted, many questions related to the development of the auto-immune response, the role of IgG subclasses and antigenic epitopes, and the pathways to podocyte injury remain unresolved. PLA2R-associated MN most likely develops governed by factors such as genetic susceptibility, loss of tolerance, alterations in antigen expression with a role for environmental factors like air pollution, smoking, and infections. More detailed knowledge of genetic factors, the relevant B and T cell epitopes, and the mechanisms of podocyte injury is needed to identify patients at risk for disease progression and to develop optimized, targeted treatment strategies. In this review we highlight unresolved issues, addressing initiation of antibody formation, the timeline of antibody production, the role of IgG subclass, and the pathogenicity of the antibodies in concert with complement to produce glomerular pathology and proteinuria.

Keywords: glomerulonephritis, membranous nephropathy, nephrotic syndrome.

Introduction:Membranous nephropathy is an important cause of nephrotic syndrome1. In about one-third of patients an underlying cause, such as an infection, malignancy, systemic autoimmune disease or use of drugs such as nonsteroidal anti-inflammatory agents, can be identified (secondary membranous nephropathy)2. In the remaining 70 % of patients no underlying cause is found and the term primary (idiopathic) membranous nephropathy (MN) is used. The discovery in 2009 of the M-type phospholipase A2 receptor (PLA2R) as the primary target in MN greatly advanced basic and clinical research3. MN is now considered a renal-limited autoimmune disease, with antibodies against PLA2R (aPLA2Rab) identified in 70-80 % and antibodies against thrombospondin type-1 domain-containing 7A (THSD7Aab) in 2-5 % of patients4,5. Assays for the measurement of aPLA2Rab are commercially available, and there is consensus that aPLA2Rab are a useful biomarker in establishing the diagnosis of MN4, predicting the prognosis6-9 and guiding treatment10,11. The initiation of the aPLA2Rab response and pathogenesis of PLA2R-associated MN remain an enigma. In this review we highlight unresolved issues, addressing initiation of antibody formation, the subclass and epitope specificity of the antibodies and the mechanisms of podocyte injury and proteinuria.

1. Anti-PLA2R antibodies in membranous nephropathy

2

3940414243444546474849505152535455565758596061626364656667686970717273747576777879808182

83

34

In their landmark paper, Beck et al. described the discovery of aPLA2Rab3. Serum of patients with MN was analyzed by Western blot technique, using extracts of normal human glomeruli under non-reducing conditions. Reactive serum contained IgG that recognized a protein, identified as PLA2R. Western blotting of reduced proteins did not show a band, suggesting that the antibodies recognized a conformational epitope.

1.1 Diagnostic assays

The introduction of a commercial immunofluorescent assay (IFT) (EUROIMMUN Lübeck, Germany) allowed widespread clinical research. Anti-PLA2R antibodies were reported present in the serum of 52 % to 86 % of patients with MN12. The wide range is explained by the heterogeneity of the studies. Studies used either Western Blot (WB) or IFT technique, and the sensitivity of the IFT test is lower than the sensitivity of the WB technique13. The studies differed with respect to ethnicity, and often included patients with longstanding disease, sometimes treated and in remission. It is also known that initially aPLA2Rab can be absent in the serum of patients with PLA2R-associated MN. Debiec et al studied a cohort of 42 patients and identified 10 patients with positive staining of PLA2R antigen in immune complexes in the kidney biopsy, with no measurable antibodies in serum14. Subsequent studies showed that serum aPLA2Rab indeed can be absent at disease onset, and become measurable during follow-up15,16. We hypothesized that the antibodies bind to the antigen in the kidney with high affinity, and only become detectable when binding sites in the kidney are saturated15. Of note, Debiec et al also described three patients with a high circulating level of aPLA2Rab that did not have detectable glomerular PLA2R14. This phenomenon has received little attention.

The IFT technique only allows semiquantitative assessment of aPLA2Rab, by reporting dilution steps. Better quantitation of aPLA2Rab was possible with the development of ELISA assays. Some centers have used “in house” developed ELISA assays, but its widespread use in clinical practice came with the introduction of a commercially available ELISA assay (EUROIMMUN Lübeck, Germany)17. This ELISA assay measures total aPLA2Rab IgG. In the development phase, a patient serum was used as calibrator to construct a standard curve that allowed quantitation in relative units (RU)/ml. When using the ELISA assay a cut-off value is used to define aPLA2Rab positivity. The literature is equivocal, with some studies using 14RU/ml and others 20RU/ml as cut-off value. Even with using the lower cut-off value the ELISA technique is less sensitive than IFT or WB13. In a recent study, the authors advised to use 2 RU/ml as cut-off value and to perform confirmatory IFT in all patients with ELISA values >2 and < 20 RU/ml18.Uncertainty regarding sensitivity of methodology remains in the absence of an international evaluation of a large number of sera in a blinded controlled comparative study using the three methods.

1.2 Role of antibody measurement

The presence of aPLA2Rab is highly specific for a diagnosis of MN. In a meta-analysis specificity was calculated at 98-100 %, although the authors admit study limitations4. In a recent study, Bobart et al.18 showed that all 132 patients who tested positive for aPLA2Rab, had a primary diagnosis of MN in their native kidney biopsy. Moreover, in the patients with a baseline eGFR of >60 ml/min/1.73m2 there were no findings that affected patient care or

3

84858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130

56

treatment. Therefore, the presence of aPLA2Rab may be considered sufficient for a diagnosis of membranous nephropathy19.

Qualitative assessment of aPLA2Rab during the disease course already demonstrated that disappearance of antibodies preceded onset of remission10. Persistence of aPLA2Rab at the end of immunosuppressive therapy predicted unfavourable outcome8. These data suggest that immunological remission could become the goal of therapy.

The quantitative analysis confirmed the temporal association between clearance of antibodies and reduction of proteinuria. In patients treated with Rituximab a 50% reduction in aPLA2Rab level preceded a similar reduction of proteinuria by approximately 10 months9. Anti-PLA2Rab levels also predicted response. High levels of antibodies were associated with a lower likelihood of remission20. In multivariable analysis, using data of the GEMRITUX cohort, a randomized trial comparing Rituximab with conservative treatment, aPLA2Rab levels < 275 RU/ml at baseline were associated with an increased likelihood of remission at six months. In Rituximab treated patients the probability at 24 months of achieving a partial or complete remission decreased from 80% in patients in the lowest tertile of aPLA2Rab to 30% in the highest tertile9. A subsequent study confirmed the low efficacy of standard dose Rituximab in inducing an immunological remission in patients with high antibody titers11. In these patients cyclophosphamide or extended Rituximab therapy are likely more effective11,21. Monitoring antibody levels at baseline and at regular time intervals after start of immunosuppressive therapy may pave the way toward personalized treatment strategies.

1.3 Characteristics of anti-PLA2R antibodies

Although the ELISA assay provides some useful information, the predictive accuracy for clinical outcome is limited. Therefore, investigators have questioned the added value of more detailed analysis of aPLA2Rab. Since MN is considered an IgG4 dominant disease, subclass specific assays were introduced6,17,22. One study suggested that the sensitivity for diagnosing MN was slightly better when using a IgG4-subclass specific assay compared to total IgG (98.5% versus 96.5%)17. Hofstra et al studied 82 patients with PLA2R-associated MN, and subclass specific aPLA2Rab were measured using in house ELISA’s. IgG4-aPLA2Rab levels were significantly correlated to total IgG aPLA2Rab, supporting IgG4 as the dominant subclass. Spontaneous remission as outcome was significantly associated with tertiles of total IgG aPLA2Rab. In additional analysis, outcome was only significantly associated with the IgG4 subclass6. Subsequent studies, using IgG4 subclass specific aPLA2Rab ELISA’s, confirmed the correlation between IgG4 aPLA2Rab levels and total IgG aPLA2Rab levels, although the variation remained high17. In a retrospective study, IgG4 aPLA2Rab predicted spontaneous remission23. In contrast, in another study of the same group, using a different patient population, there was no association between prognosis and levels of IgG4 antibodies directed at human PLA2R24. The discrepancies between the studies are likely related to the heterogeneity between the studies with respect to patient characteristics, treatment, and the inclusion of aPLA2Rab negative patients.

In their latter study24, Seitz-Polski et al not only used human PLA2R, but also rabbit PLA2R and mouse PLA2R as antigen. The cross-reactivity of 53 sera from MN patients already known to be positive for human PLA2R was investigated. All sera recognized rabbit PLA2R

4

131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177

78

and only 51% of sera recognized mouse PLA2R. Interestingly, the patients who had antibodies that recognized mouse PLA2R had worse outcome. This study suggested that sera of patients contained aPLA2Rab directed at different epitopes. In subsequent studies, these investigators identified aPLA2Rab with reactivity against different domains of PLA2R protein, namely the cysteine rich domain (CysR), the C-type lectin domains 1 (CTLD1) and CTLD722,25. The CysR domain was already defined as an epitope containing domain by other groups26,27, and is considered the immunodominant epitope. In the first study of Seitz Polski et al, including 69 patients with aPLA2Rab associated MN22, antibodies against CysR were present in all patients. In 23 patients (33%), antibodies only recognized CysR, 14 patients (20%) presented with antibodies reactive to CysR and CTLD1, and 32 patients (47%) were reactive with CysR, CTLD1 and CTLD7. In this small retrospective study, which included patients with non-nephrotic proteinuria, with a wide range of eGFR, and large heterogeneity in treatment, the patients with reactivity against two or three epitopes (so called ‘spreaders’) were older, and had more severe proteinuria. Of note, spreaders had higher levels of aPLA2Rab against CysR, although no significant differences were observed in IgG4 aPLA2Rab levels. During follow-up poor prognosis was associated with epitope spreading, and in multivariable analysis both epitope spreading and antibody levels were independently associated with outcome. The authors have validated their findings, using data of the GEMRITUX study25. Treatment was started after an observation period of at least 6 months, and the study included nephrotic patients only. The analysis included 58 aPLA2Rab positive patients (measured using the EUROIMMUN assay). Of these patients, 29 received rituximab. In 34.5% of patients antibodies only reactive to CysR were found (non-spreaders). In this cohort, and in contrast to the previous study, spreading was not associated with age, but clearly correlated with total IgG aPLA2Rab titer. During follow-up, spreading was an independent predictor of outcome. The authors suggested that spreading might be particularly relevant in patients with antibody levels > 50 RU/ml. They warned for too early conclusions and point to the need for confirmation and validation, in view of the small number of patients, and the inclusion of treated and untreated patients.

2. What triggers anti-PLA2R antibody production?

PLA2R-associated MN is a renal-limited auto-immune disease. We refer the reader to the literature for a detailed and extensive description of the normal immune-response and the deviations leading to auto-immunity. Figure 1 illustrates the various pathways that may be involved in the development of any auto-immune disease, using PLA2R-associated MN as an example. The development of autoreactive antibodies and thus of auto-immune disease can result from loss of central or peripheral tolerance, altered expression of the antigen (antigen modification, increased expression), and intermolecular epitope spreading (molecular mimicry). In the process of B cell activation several processes can be discerned, e.g. immunoglobulin class switching, increased antibody production and affinity maturation (higher levels of more avid antibodies), and epitope spreading.

2.1 Genetic association with human leukocyte antigen (HLA) class II genes

5

178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208

209210211212213214215216217218219220221222

910

Most auto-immune diseases are associated with major histocompatibility complex (MHC) genes28. This is not unexpected, since peptide fragment (epitopes) are presented to T cells in the context of the MHC molecule. Already in 1979, a association between HLADR3 and MN was reported29. This was confirmed in subsequent studies based on serotyping, showing associations with HLA class I (B8, B18) and class II (DR 2, DR3, DR7)28. New advances came with the introduction of molecular genotyping, which refined the locus to DQA130. Most recently genome-wide association (GWAS) studies were introduced. In the first GWAS that included 556 patients of European ancestry, the highest association was between MN and single-nucleotide polymorphisms (SNPs) in the HLA-DQA1 region31. This association was confirmed in another European cohort32. This study disclosed additional associations with DRB1*03:01 (HLA-DR3) and HLA-DQA*05:01-HLA-DQB1*02:01 (HLA-DQ2). The association with HLA DQA1 was also confirmed in Asian populations33-36. In the Han Chinese population a significant association was found not only with HLA-DRB1*03:01, but also with HLA-DRB1*15:0135,36. In one study, HLA-DRB3*02:02, which shares a haplotype with HLA-DRB1*03:01, was considered as risk allele36. Cui et al.35 studied, by means of structural modelling, the role of amino-acid substitutions and pointed to certain amino-acid positions in the peptide-binding pocket of the MHC-beta chain, which may determine epitope specificity. Currently, the Class II genetic restriction is best interpreted as controlling which peptides of PLA2R are presented to T cells to provide the necessary T cell help for high affinity autoantibody production. Unfortunately, most studies have evaluated a limited number of patients, without sufficient fine-mapping of the HLA-genes, and unable to perform conditional analyses. Thus, the causal alleles remain unknown.

2.2 Loss of tolerance

It is not known if central tolerance is involved in preventing an autoreactive response to PLA2R. The presence of PLA2R in the thymus is debated. Peripheral tolerance could be the controlling mechanism in preventing an autoreactive response to PLA2R, maintained by the low level of soluble PLA2R (sPLA2R) that is reported to be present in the circulation of healthy individuals at 50-5000ng/ml by western blotting37. Our unpublished data suggests a much lower level of sPLA2R at approximately 300pg/ml by ELISA (PEC Brenchley personal communication). Importantly, sPLA2R may maintain peripheral tolerance by controlling function of regulatory T cells to suppress B cells with the potential to generate aPLA2Rab.

In a recent study, using blood samples collected regularly in US military, the presence of aPLA2Rab in the circulation long before onset of MN was documented (Joshi et al, J Am Soc Nephrol 29, 2018: P823). This observation might be explained by the natural presence of auto-reactive B cells. Peripheral tolerance is largely dependent on the function of Tregs. Roccatello38 studied changes in B and T cell phenotypes by flow cytometry in rituximab treated patients. Rituximab induced B cell depletion, which coincided with remission of proteinuria and an increase in Tregs. In two of the non-responders patients, no increase in Tregs was observed. Rosenzwajg et al39 showed that patients with active MN had lower number of Tregs. Importantly, an increase in Tregs at day 8 after start of rituximab predicted remission after 3-6 months follow-up. Two recent studies confirmed that Tregs are decreased in patients with active MN (Ramachandran et al, J Am Soc Nephrol 29, 2018: P823 and Cantarelli et al, J Am Soc Nephrol 29, 2018: P824), with a significant increase after immunosuppressive therapy (Ramachandran et al, J Am Soc Nephrol 29, 2018: P823).

6

223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269

1112

2.3 Activation of antigen presenting cells: the role of the lung in MN

Antigen presenting cells (APCs) are critical in determining the immune response. They process antigen and present small peptides fragments to the T cell. The subsequent response, varying from T cell anergy to T cell activation, is dependent on the presence of costimulatory molecules on the surface of the APCs, a sign of APC activation. APC activation occurs through cytokines, produced in inflammatory states. In recent years, the lung has emerged as a potential contributor to allergies and autoimmune disease40. The lung is a large surface area, exposed to the external world. Infection, smoking and air pollution all cause oxidative stress and inflammation, and the resulting cytokines may spill over into the systemic circulation and exert distant effects. Also, in the inflammatory milieu APCs become activated. Indeed, the lung has been implied in the development of autoimmune diseases such as type 1 diabetes, lupus nephritis and rheumatoid arthritis40.

An interesting finding is the association between air pollution and MN that has been suggested41,42. Xu et al. studied the incidence of glomerular disease over an 11 year period in China41. They observed an increasing incidence in MN cases. There was a clear association between air pollution and MN. This was particularly related to the changes in fine particulate matter < 2.5 µm. In another study, this association was confirmed42. Clearly, the observational nature of these studies is an important limitation, as there are many confounders that co-occur with air quality. Smoking could be another factor, also acting by inducing pulmonary inflammation. Indeed, smokers are at increased risk of developing autoimmune disease such as multiple sclerosis, rheumatoid arthritis, type 1 diabetes and systemic lupus erythematosus40. Smoking was associated with a more rapid decline of estimated glomerular filtration rate in MN43,44. However, an association between smoking and the development of MN has yet to be proven.

Another clue towards involvement of the lung is the association of MN with sarcoidosis. In this group of patients a high prevalence of PLA2R-related disease has been described45.

Because of this indirect evidence the lung has been implicated in MN (Figure 2). Although intriguing, the role of the lung has not been firmly established thus far.

2.4 Change in conformation or expression of PLA2R

2.4.1 Genetic association with the PLA2R gene

In 2010 Liu et al. were the first to report an association between single nucleotide polymorphisms in the gene that encodes PLA2R and the development of MN46. These findings were confirmed in genetic association studies in Korean47 and European cohorts31. Genetic studies of PLA2R so far lack sufficient detail to identify how the genetic risk is translated through a pathological mechanism to cause disease. It remains uncertain which polymorphism(s) is the culprit and how this affects the protein sequence/structure.Genetic variants alone or in cooperation may affect the molecular conformation of the antigen exposing new conformational B cell epitopes or allowing cryptic enzyme cleavage sites to become accessible facilitating the processing of linear T cell peptides. In this respect,

7

270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316

1314

some polymorphisms identified were in regions that were recently identified as B cell epitope containing domains48. However, nothing is known of a link between SNP markers and T cell epitopes. Genetic variants, especially in the intronic regions may also increase expression of the antigen49, another possible initiator of disease. Some of these explanations are explored in greater detail in a global second GWAS in MN involving 3782 cases to be reported in 2019 (in submission Kiryluk 2019).

All studies have confirmed the strong interaction and synergy between genetic variants in PLA2R and HLA class II genes. Thus far, no study has separately analyzed genetic associations in aPLA2Rab positive versus aPLA2Rab negative patients.

2.4.2 Altered or increased expression of PLA2R antigen

In normal and non-MN diseased kidneys PLA2R is expressed only faintly in a granular pattern on podocytes. Hoxha et al.50 observed increased staining of PLA2R, in a coarse granular pattern lining the glomerular basement membrane (GBM) in patients with MN only, tightly related to the presence of aPLA2Rab in serum. The authors evaluated mRNA expression in kidney biopsies of patients and controls, and found no evidence for increased transcription of the receptor in patients. Thus, there is no proof of increased production of PLA2R antigen due to increased mRNA transcription. We hypothesize this increased staining for PLA2R antigen in the glomeruli in the absence of increased production might be caused by a process where circulating aPLA2Rab captures PLA2R, normally shuttling between the endoplasmic reticulum and the membrane. Similar mechanisms have been described for other proteins51. The resulting immune complexes will be shed from the membrane and form the larger subepithelial deposits.

Obviously, it is plausible that conformational changes or increased expression of PLA2R outside the kidney could trigger the autoimmune response. In this respect, we hypothesize that the lung might be relevant (Figure 2). Expression of PLA2R in human lung alveolar macrophages was first described in 200552. More recently, PLA2R expression was also observed in bronchiolar tissue (von Haxthausen et al, J Am Soc Nephrol 29, 2018: P822).

The role of antigen presentation in pulmonary tissue in autoimmunity has been suggested in other autoimmune diseases40. In anti-GBM disease two major epitopes of the α3(IV)NC1 molecule have been identified, EA and EB53,54. These antibodies are present at low levels and with low affinity in healthy individuals55, but the epitope is hidden and conformational changes are necessary for antibodies to bind56. The α3(IV)NC1 molecule is present in both the lung and kidney. It has been suggested that these epitopes are exposed after a primary insult to the lung (as might for instance occur as a consequence of smoking)57. Is MN another example of an autoimmune disease in the kidney in which processing of the antigen might start in the lung, as in anti-GBM disease? And if this is the case, why do MN patients not present with symptoms of lung pathology? Although not firmly established, one explanation might be that the α3(IV)NC1 molecule is present as a permanent structural component of the basement membrane in the lung in contrary to PLA2R antigen. PLA2R is present on macrophages52, which are mobile and designed to clear immune complexes efficiently. This might explain the renal-limited character of MN.

8

317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363

1516

2.5 Molecular mimicry

Autoimmunity can be initiated when epitopes on microbes are presented to the immune system which share homologies with self-antigen (molecular mimicry)58. Li et al studied the occurrence of epitope mimicry by microbial peptides in the initiation of anti-GBM disease 59. These authors defined a critical amino acid motif on human α3(IV)NC1, based on known T- and B cell epitopes. They have searched the Uniprot database for microbial peptides that mimic this critical motif. In total 23826 peptides were identified, seven of them were related to human infections. Subsequently, the frequency of serum antibodies recognizing these seven microbial peptides were measured. Microbe-derived peptides were recognized in the majority of anti-GBM patients. An inhibition ELISA demonstrated that these antibodies showed cross-reactivity with the B cell epitope and α3(IV)NC1. Of note, a low frequency of antibodies against 20 linear peptides unrelated to the critical amino acid motif were measured (3.9 %).

Also in PLA2R-related MN, molecular mimicry may contribute to disease initiation27,60. There is a partial homology between a part of the major CysR epitope in PLA2R and an antigenic cell wall component of some clostridium species27. Natural exposure to such infections could increase the frequency of B cells with a B cell immunoglobulin receptor able to bind PLA2R or PLA2R fragments containing this CysR epitope. However, other (epidemiological and experimental) evidence for a role of molecular mimicry in MN is lacking.

3. Time course of antibody production and subclass switching

Beck et al already suggested that aPLA2Rab were predominantly of the IgG4 subclass3. Indeed, MN is considered an IgG4 dominant disease. The mechanisms of production of antibodies of the IgG4 subclass are only recently better understood. It is suggested that there is a programmed order of the immune response, starting from IgM producing B cells, followed by class switching to IgG producing B cells, with production of IgG subclass antibodies in a fixed order of IgG3 > IgG1> IgG2 > IgG461. However, IgG4 production is also considered the consequence of an allergic response, where an allergenic peptide (which induced a Th2 cytokine profile) initially triggers the production of IgE antibodies62,63. With prolonged exposure, and increased production of interleukin 10, the IgE response mitigates and is replaced by steadily rising IgG4. In the allergic response, IgG4 blocks the action of IgE, thus limiting the allergic reaction. In clinical practice, this advantageous suppression of IgE action is used in the setting of desensitisation therapy.

Auto-antibodies of the IgG4 subclass are predominantly present in diseases such as pemphigus vularis or some forms of myasthenia gravis64. In these diseases, the IgG4 antibodies were pathogenic, as proven by passive transfer experiments or the induction of disease by using Fab fragments. In anti-GBM disease most patients have anti-GBM antibodies of the IgG1 or IgG3 subclass65-67. In few patients, antibodies of the IgG4 subclass dominated, and these patients presented with an unusually mild disease course67. Thus, in anti-GBM disease the IgG4 antibodies are likely less pathogenic.

9

364365366367368369370371372373374375376377378379380381382383384385386387

388389390391392393394395396397398399400401402403404405406407408409

1718

Although MN is considered an IgG4 dominant disease, and some investigators focus on measurement of IgG4 subclass aPLA2Rab, there is no proof that IgG4 aPLA2Rab are responsible for disease. Indeed, aPLA2Rab of other IgG subclasses are present in the majority of patients. The percentage of patients with aPLA2Rab of IgG1 subclass was 76% in an European cohort6 and 92% in a Chinese cohort68. The studies did not mention the percentage of patients with either IgG1 or IgG3 aPLA2Rab. Still, in 5-10% of patients antibodies of IgG4 subclass were not detectable, and in these patients always IgG1 or IgG3 antibodies were present6. Although both IgG4 aPLA2Rab and IgG1 aPLA2Rab correlated with total IgG aPLA2Rab levels, we reanalyzed the data and the relation between IgG4 aPLA2Rab and IgG1 aPLA2Rab in this study was less tight (Pearson’s r=0.451). No study closely evaluated the time course of subclass specific aPLA2Rab. Still, histological studies suggest that whereas IgG4 is the dominant subclass in patients with longstanding MN (EM stage III/IV), IgG1 is the predominant subclass in early MN (EM stage I/II)69,70. These findings suggest that the onset of MN starts with the formation of aPLA2Rab of the IgG1 (or IgG2/3) subclass, with increased formation of IgG4 during prolonged disease activity. No study provided information on the presence of IgE aPLA2Rab.

In clinical practice the onset of disease is not easily recognized by patients with MN. Thus, referral to nephrology care and kidney biopsy are done many months after the onset of edema. At the time of first presentation at nephrology care, aPLA2Rab are of high affinity27, which is compatible with maturation of the immune response.

The clinical course of MN is quite variable, some patients developing a spontaneous remission, and others with disease progressing to end-stage renal disease. The immunological response patterns mimic the clinical course: aPLA2Rab disappear spontaneously in the patients with remission, and persist, often at high levels, in patients with progressive disease. The reasons for these differences have not been elucidated.

4. Pathogenicity of aPLA2R antibodies

Although there is a close temporal relationship between aPLA2Rab and clinical parameters (proteinuria), formal proof for the pathogenicity of the antibodies is lacking. Proof of the pathogenicity would need studies that show that the disease/podocyte injury and proteinuria can be induced after passive transfer of the aPLA2Rab. Animal models are not yet available, since wild type mice and rats lack podocytic PLA2R. Mice with a podocyte knock-in of the N-terminal domains of human PLA2R –CysR-CTLD3 which contains the major B cell epitope (P28mer) have been constructed (Rhoden et al, J Am Soc Nephrol 29, 2018: P 823) for passive transfer experiments of human IgG1, IgG3 and IgG4 subclass specific aPLA2Rab and mouse anti-human aPLA2Rab to test for induction of proteinuria.

4.1 Complement mediated podocyte injury

10

410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441

442443444445446447448449450451452453454

1920

The almost universal presence of C3 in the subepithelial deposits in membranous nephropathy71,72 and the strong association between the intensity of glomerular C3 staining and disease severity72 supports a role for complement in the development of proteinuria. However, the role of complement in MN is undetermined. IgG4, the dominant IgG subclass present in membranous nephropathy, does not bind complement and is unable to activate the classical complement pathway. It was suggested that complement may be activated through the mannose binding lectin (MBL) pathway73-76. Indeed, the presence of complement components such as C4d, factor B and properdin in the glomerular deposits is compatible with activation of the MBL pathway76. In one study IgG4 binding to MBL was suggested (Ma et al, J Am Soc Nephrol 22, 2011: P62A). However, in an interesting case study, Bally et al. demonstrated the development of MN in a patient with proven MBL deficiency but revealed complement activation via the alternative pathway, indicating that the MBL pathway is not required for membranous nephropathy to develop77. Some authors suggest that IgG4 is not the pathogenic subclass. Indeed, in most patients with membranous nephropathy antibodies of the complement fixing IgG1 and IgG3 subclasses have been found50. Still, the regular absence of C1q deposits71 argues against activation or involvement of the classical complement pathway. A recent study pointed to another mechanism of complement activation, independent of immune complex formation. In a mouse model of MN, investigators showed loss of glomerular heparan sulphate proteoglycans from the glomerular basement membrane, which bind complement factor H (cFH), causing local dysregulation of the alternative pathway 78. The group of Dr. Ronco et al described patients with MN and an unexplained and fast deterioration of kidney function in whom antibodies against cFH were present79.

4.2 Non-complement mediated podocyte injury

PLA2R is thought to mediate some of the effects of the soluble phospholipases. These molecules play a role in cell proliferation, airway constriction, smooth muscle contraction and inflammatory responses80. mRNA expression studies have documented the presence of mRNA of PLA2R most abundantly in the kidney; other tissues with mRNA expression were muscle, lung and placenta. In the kidney PLA2R is observed in the podocyte. The function of the PLA2R in the podocyte is unknown. PLA2R antibodies bind to PLA2R with high affinity27. Thus, it is possible that upon binding of the aPLA2Rab the function of the podocyte changes. Indeed aPLA2Rab (in the absence of complement activation) were shown to modulate podocyte cell biology. Anti-PLA2R antibodies from 5 MN patients induced oxidative stress in differentiated human podocytes, increased cell apoptosis and compromised the permeability of the cell monolayer (Fresquet et al, J Am Soc Nephrol 27, 2016). How aPLA2Rab induces these changes and disrupts PLA2R function remains however unknown.

In kidney tissue two different mRNA transcripts are present, one encoding the membrane bound PLA2R, the other encoding a soluble form of the receptor37. Indeed, Watanabe et al. recently showed that soluble PLA2R can be found in normal healthy volunteers37. PLA2R binds to collagen type I and type IV through the FNII domain. The membrane bound receptor mediated the interaction between collagen I and beta1integrin, affecting podocyte mobility. Interestingly, the soluble receptor blocked the interaction between collagen and integrins, and thus suppressed a migratory response of the podocyte. Theoretically antibodies against PLA2R1 on the cell membrane could affect podocyte function. Annexin2/S100A10 (A2t)

11

455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501

2122

complex was identified as a novel binding partner of PLA2R in podocytes81. This PLA2R/A2t complex was found on the podocyte plasma membrane and in secreted extracellular vesicles. This implies that PLA2R may be at the heart of actin cytoskeleton reorganization and tight junction assembly, two functions known to be modulated early in proteinuric membranous nephropathy.

The possibility that anti-podocyte antibodies induce podocyte injury or alter podocyte function independent from complement is illustrated by studies that evaluated thrombospondin type 1 domain-containing 7A (THSD7A). Tomas et al. studied the pathogenicity of thrombospondin type 1 domain-containg 7A autoantibodies (aTHSD7A ab)82,83. Injection of serum and purified IgG in mice induced proteinuria within three days. Proteinuria increased further, which was paralleled by the development of mouse anti human IgG antibodies and the development of immune complexes containing human and mouse IgG, characteristic for the autologous phase. Interestingly, in this model C3 deposits were seen in this later phase, but not in the initial heterologous phase82. In a subsequent study, they developed rabbit anti THSD7A antibodies. Again, injection of rabbit serum induced proteinuria which was complement independent. In vitro, incubation of podocytes with anti THSD7A antibodies caused cytoskeleton rearrangement and activation of focal adhesion signaling. Of note, when using zebra fish larvae knockdown of THSD7A caused podocyte changes with injury of the glomerular filtration barrier and development of edema, suggesting an important role of THSD7A in glomerular filtration barrier integrity83. These findings clearly demonstrate that proteinuria in models of MN may not be dependent on complement. In a recent study evidence was found that THSD7A functions as a foot process protein and stabilizes the membrane of the podocytes84.

5. Current and future therapies for removing and suppressing anti-PLA2R levels

Current treatment strategies in patients with MN are delineated in the 2012 KDIGO guidelines85, with added guidance formulated in a recent consensus document86. All patients should receive optimal conservative therapy for at least 6 months or until eGFR decreases. Patients at high risk for disease progression, identified by duration and severity of proteinuria, eGFR, urine excretion of low molecular weigth proteins or aPLA2Rab levels, are eligible for immunosuppressive therapy. Physicians and patients may consider cyclophosphamide, rituximab or a calcineurin inhibitor dependent on expected side effects and efficacy. Recent data suggest that measurement of aPLA2Rab may enable to adjust the duration of therapy or even lead to prefer one therapy over the other11.

The identification of PLA2R as the dominant antigen in MN has been a ‘game changer’. The discovery of antibodies in the serum opened the way toward novel therapies. Since patients with high PLA2Rab have most severe disease and are at increased risk of progression, it can be questioned if removal of antibodies by immunoadsorption might be beneficial. Currently a small pilot study of aPLA2Rab immunoadsorption is being undertaken87 to establish the feasibility of rendering patients seronegative for aPLA2Rab in the absence of immunosuppression.

Recognition of the B cell epitopes would enable the development of specific, targeted

12

502503504505506507508509510511512513514515516517518519520521522523524525526527528

529530531532533534535536537538539540541542543544545546547

2324

therapies as reviewed for pemphigus vulgaris88. Removal of pathogenic aPLA2Rab by immunoadsorption may reduce disease activity, but it will not cure the disease. It will remain necessary to block antibody production. Identification of the pathogenic B cell epitopes may enable the development of treatment strategies that specifically target the autoreactive B cells, causing antigen specific B cell depletion. One option is the targeted cell death of autoreactive B cells by infusion of a toxin-conjugated epitope-containing fragment of the PLA2R antigen that binds to the anti-PLA2R B cell receptor. A more durable elimination of aPLA2R reactive B cells is expected from cellular immunotherapy, also known as chimeric autoantibody receptor (CAAR) T cell therapy. In this approach the patient’s T cells are ex vivo modified, with incorporation of CD28 and/or CD137 costimulatory domains and of the pathogenic epitope of PLA2R in the T cell receptor. These T cells will specifically kill PLA2R-specific B cells, and their persistence as memory CAR-T cells will provide continuous surveillance against re-emergence of PLA2R-positive B cells.

On the other hand, identification of T cell epitopes might allow development of immune vaccination, which should enable the development of antigen-specific immune tolerance. In this situation, the goal is immunization of MN patients using a pool of T cell peptides to restore specific Tregs allowing suppression of B cell production of anti-PLA2R.

It remains important to decipher the mechanisms of pathogenicity of the aPLA2Rab. If IgG4 antibodies cause podocyte injury by binding to PLA2R, independent from complement or Fc receptor mediated neutrophil activation, it would provide opportunities to develop treatment strategies using small peptides that would interfere with the binding of the antibodies to the PLA2R antigen. On the other hand, if podocyte injury is dependent on complement activation, we must consider the use of anti-complement therapies.

In conclusion, in the ten years after the discovery of PLA2R as the major auto-antigen in MN a great progress has been made. Still, many questions remain unsolved (Table 1), and we expect the answers in the next decade.

AcknowledgementsPB and MF acknowledge support from Kidney Research UK grant RP31/2015, the Manchester Academic Healthcare Science Centre (MAHSC) (186/200) and core funding from the Wellcome Trust (203128/Z/16/Z) to the Wellcome Centre for Cell-Matrix Research at the University of Manchester. PB acknowledges support from the Greater Manchester Local Clinical Research Network.JW and AL acknowledge support from a grant from the Dutch Kidney Foundation.The work leading to these results received funding from European union Seventh Framework Programme FP7/2007-2013 grant 305608: European Consortium for High-Throughput Research in Rare Kidney Diseases.

DisclosureNo conflicts of interest.

13

548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594

2526

References1. Ponticelli, C. & Glassock, R.J. Glomerular diseases: membranous nephropathy--a modern view. Clin J Am Soc Nephrol 9, 609-16 (2014).2. Glassock, R.J. Secondary membranous glomerulonephritis. Nephrol Dial Transplant 7 Suppl 1, 64-71 (1992).3. Beck, L.H., Jr. et al. M-type phospholipase A2 receptor as target antigen in idiopathic membranous nephropathy. N Engl J Med 361, 11-

21 (2009).4. Du, Y. et al. The diagnosis accuracy of PLA2R-AB in the diagnosis of idiopathic membranous nephropathy: a meta-analysis. PLoS One 9,

e104936 (2014).5. Tomas, N.M. et al. Thrombospondin type-1 domain-containing 7A in idiopathic membranous nephropathy. N Engl J Med 371, 2277-87

(2014).6. Hofstra, J.M. et al. Antiphospholipase A2 receptor antibody titer and subclass in idiopathic membranous nephropathy J Am Soc Nephrol

23, 1735-1743 (2012).7. Hoxha, E., Harendza, S., Pinnschmidt, H., Panzer, U. & Stahl, R.A. M-type phospholipase A2 receptor autoantibodies and renal function

in patients with primary membranous nephropathy. Clin J Am Soc Nephrol 9, 1883-90 (2014).8. Bech, A.P., Hofstra, J.M., Brenchley, P.E. & Wetzels, J.F. Association of anti-PLA(2)R antibodies with outcomes after immunosuppressive

therapy in idiopathic membranous nephropathy. Clin J Am Soc Nephrol 9, 1386-92 (2014).9. Ruggenenti, P. et al. Anti-Phospholipase A2 Receptor Antibody Titer Predicts Post-Rituximab Outcome of Membranous Nephropathy. J

Am Soc Nephrol 26, 2545-58 (2015).10. Beck, L.H., Jr. et al. Rituximab-induced depletion of anti-PLA2R autoantibodies predicts response in membranous nephropathy. J Am Soc

Nephrol 22, 1543-50 (2011).11. van de Logt, A.E. et al. Immunological remission in PLA2R-antibody-associated membranous nephropathy: cyclophosphamide versus

rituximab. Kidney Int 93, 1016-1017 (2018).12. Hofstra, J.M. & Wetzels, J.F. Anti-PLA(2)R antibodies in membranous nephropathy: ready for routine clinical practice? Neth J Med 70,

109-13 (2012).13. Hoxha, E. et al. An Indirect Immunofluorescence Method Facilitates Detection of Thrombospondin Type 1 Domain-Containing 7A-

Specific Antibodies in Membranous Nephropathy. J Am Soc Nephrol 28, 520-531 (2017).14. Debiec, H. & Ronco, P. PLA2R autoantibodies and PLA2R glomerular deposits in membranous nephropathy. N Engl J Med 364, 689-90

(2011).15. van de Logt, A.E., Hofstra, J.M. & Wetzels, J.F. Serum anti-PLA2R antibodies can be initially absent in idiopathic membranous

nephropathy: seroconversion after prolonged follow-up. Kidney Int 87, 1263-4 (2015).

14

595596597598599600601602603604605606607608609610611612613614615616617618619620621622623

2728

16. Ramachandran, R., Kumar, V., Nada, R. & Jha, V. Serial monitoring of anti-PLA2R in initial PLA2R-negative patients with primary membranous nephropathy. Kidney Int 88, 1198-9 (2015).

17. Dahnrich, C. et al. Development of a standardized ELISA for the determination of autoantibodies against human M-type phospholipase A2 receptor in primary membranous nephropathy. Clin Chim Acta 421, 213-8 (2013).

18. Bobart, S.A. et al. Noninvasive diagnosis of primary membranous nephropathy using phospholipase A2 receptor antibodies. Kidney Int 95, 429-438 (2019).

19. De Vriese, A.S., Glassock, R.J., Nath, K.A., Sethi, S. & Fervenza, F.C. A Proposal for a Serology-Based Approach to Membranous Nephropathy. J Am Soc Nephrol 28, 421-430 (2017).

20. Dahan, K. et al. Rituximab for Severe Membranous Nephropathy: A 6-Month Trial with Extended Follow-Up. J Am Soc Nephrol (2016).21. Dahan, K. et al. Retreatment with rituximab for membranous nephropathy with persistently elevated titers of anti-phospholipase A2

receptor antibody. Kidney Int 95, 233-234 (2019).22. Seitz-Polski, B. et al. Epitope Spreading of Autoantibody Response to PLA2R Associates with Poor Prognosis in Membranous

Nephropathy. J Am Soc Nephrol 27, 1517-33 (2016).23. Jullien, P. et al. Anti-phospholipase A2 receptor antibody levels at diagnosis predicts spontaneous remission of idiopathic membranous

nephropathy. Clin Kidney J 10, 209-214 (2017).24. Seitz-Polski, B. et al. Cross-reactivity of anti-PLA2R1 autoantibodies to rabbit and mouse PLA2R1 antigens and development of two novel

ELISAs with different diagnostic performances in idiopathic membranous nephropathy. Biochimie 118, 104-15 (2015).25. Seitz-Polski, B. et al. Phospholipase A2 Receptor 1 Epitope Spreading at Baseline Predicts Reduced Likelihood of Remission of

Membranous Nephropathy. J Am Soc Nephrol 29, 401-408 (2018).26. Kao, L., Lam, V., Waldman, M., Glassock, R.J. & Zhu, Q. Identification of the immunodominant epitope region in phospholipase A2

receptor-mediating autoantibody binding in idiopathic membranous nephropathy. J Am Soc Nephrol 26, 291-301 (2015).27. Fresquet, M. et al. Identification of a major epitope recognized by PLA2R autoantibodies in primary membranous nephropathy. J Am Soc

Nephrol 26, 302-13 (2015).28. Robson, K.J., Ooi, J.D., Holdsworth, S.R., Rossjohn, J. & Kitching, A.R. HLA and kidney disease: from associations to mechanisms. Nat Rev

Nephrol 14, 636-655 (2018).29. Klouda, P.T. et al. Strong association between idiopathic membranous nephropathy and HLA-DRW3. Lancet 2, 770-1 (1979).30. Vaughan, R.W., Demaine, A.G. & Welsh, K.I. A DQA1 allele is strongly associated with idiopathic membranous nephropathy. Tissue

Antigens 34, 261-9 (1989).31. Stanescu, H.C. et al. Risk HLA-DQA1 and PLA(2)R1 alleles in idiopathic membranous nephropathy. N Engl J Med 364, 616-26 (2011).

15

624625626627628629630631632633634635636637638639640641642643644645646647648649650651652

2930

32. Sekula, P. et al. Genetic risk variants for membranous nephropathy: extension of and association with other chronic kidney disease aetiologies. Nephrol Dial Transplant 32, 325-332 (2017).

33. Lv, J. et al. Interaction between PLA2R1 and HLA-DQA1 variants associates with anti-PLA2R antibodies and membranous nephropathy. J Am Soc Nephrol 24, 1323-9 (2013).

34. Ramachandran, R. et al. PLA2R antibodies, glomerular PLA2R deposits and variations in PLA2R1 and HLA-DQA1 genes in primary membranous nephropathy in South Asians. Nephrol Dial Transplant 31, 1486-93 (2016).

35. Cui, Z. et al. MHC Class II Risk Alleles and Amino Acid Residues in Idiopathic Membranous Nephropathy. J Am Soc Nephrol 28, 1651-1664 (2017).

36. Le, W.B. et al. HLA-DRB1*15:01 and HLA-DRB3*02:02 in PLA2R-Related Membranous Nephropathy. J Am Soc Nephrol 28, 1642-1650 (2017).

37. Watanabe, K. et al. Human soluble phospholipase A2 receptor is an inhibitor of the integrin-mediated cell migratory response to collagen-I. Am J Physiol Cell Physiol (2018).

38. Roccatello, D. et al. New insights into immune mechanisms underlying response to Rituximab in patients with membranous nephropathy: A prospective study and a review of the literature. Autoimmun Rev 15, 529-38 (2016).

39. Rosenzwajg, M. et al. B- and T-cell subpopulations in patients with severe idiopathic membranous nephropathy may predict an early response to rituximab. Kidney Int (2017).

40. Ritz, S.A. Air pollution as a potential contributor to the 'epidemic' of autoimmune disease. Med Hypotheses 74, 110-7 (2010).41. Xu, X. et al. Long-Term Exposure to Air Pollution and Increased Risk of Membranous Nephropathy in China. J Am Soc Nephrol 27, 3739-

3746 (2016).42. Li, J. et al. Primary glomerular nephropathy among hospitalized patients in a national database in China. Nephrol Dial Transplant (2018).43. Yamaguchi, M. et al. Smoking is a risk factor for the progression of idiopathic membranous nephropathy. PLoS One 9, e100835 (2014).44. Stengel, B., Couchoud, C., Cenee, S. & Hemon, D. Age, blood pressure and smoking effects on chronic renal failure in primary glomerular

nephropathies. Kidney Int 57, 2519-26 (2000).45. Stehle, T., Audard, V., Ronco, P. & Debiec, H. Phospholipase A2 receptor and sarcoidosis-associated membranous nephropathy. Nephrol

Dial Transplant 30, 1047-50 (2015).46. Liu, Y.H. et al. Association of phospholipase A2 receptor 1 polymorphisms with idiopathic membranous nephropathy in Chinese patients

in Taiwan. J Biomed Sci 17, 81 (2010).47. Kim, S. et al. Single nucleotide polymorphisms in the phospholipase A2 receptor gene are associated with genetic susceptibility to

idiopathic membranous nephropathy. Nephron Clin Pract 117, c253-c258 (2011).

16

653654655656657658659660661662663664665666667668669670671672673674675676677678679680681

3132

48. Coenen, M.J. et al. Phospholipase A2 Receptor (PLA2R1) Sequence Variants in Idiopathic Membranous Nephropathy. J Am Soc Nephrol (2013).

49. Hanson, G. & Coller, J. Codon optimality, bias and usage in translation and mRNA decay. Nat Rev Mol Cell Biol 19, 20-30 (2018).50. Hoxha, E. et al. Enhanced expression of the M-type phospholipase A2 receptor in glomeruli correlates with serum receptor antibodies in

primary membranous nephropathy. Kidney Int 82, 797-804 (2012).51. Mentzel, S., Dijkman, H.B., van Son, J.P., Wetzels, J.F. & Assmann, K.J. In vivo antibody-mediated modulation of aminopeptidase A in

mouse proximal tubular epithelial cells. J Histochem Cytochem 47, 871-80 (1999).52. Granata, F. et al. Activation of cytokine production by secreted phospholipase A2 in human lung macrophages expressing the M-type

receptor. J Immunol 174, 464-74 (2005).53. Hellmark, T., Burkhardt, H. & Wieslander, J. Goodpasture disease. Characterization of a single conformational epitope as the target of

pathogenic autoantibodies. J Biol Chem 274, 25862-8 (1999).54. Borza, D.B. et al. The goodpasture autoantigen. Identification of multiple cryptic epitopes on the NC1 domain of the alpha3(IV) collagen

chain. J Biol Chem 275, 6030-7 (2000).55. Yang, R. et al. Natural anti-GBM antibodies from normal human sera recognize alpha3(IV)NC1 restrictively and recognize the same

epitopes as anti-GBM antibodies from patients with anti-GBM disease. Clin Immunol 124, 207-12 (2007).56. Pedchenko, V. et al. Molecular architecture of the Goodpasture autoantigen in anti-GBM nephritis. N Engl J Med 363, 343-54 (2010).57. Salama, A.D. & Levy, J.B. Tolerance and autoimmunity in anti-GBM disease. J Am Soc Nephrol 14, 2988-9 (2003).58. Rojas, M. et al. Molecular mimicry and autoimmunity. J Autoimmun 95, 100-123 (2018).59. Li, J.N. et al. Plasma from patients with anti-glomerular basement membrane disease could recognize microbial peptides. PLoS One 12,

e0174553 (2017).60. Beck, L.H., Jr. The dominant humoral epitope in phospholipase A2 receptor-1: presentation matters when serving up a slice of pi. J Am

Soc Nephrol 26, 237-9 (2015).61. Collins, A.M. & Jackson, K.J. A Temporal Model of Human IgE and IgG Antibody Function. Front Immunol 4, 235 (2013).62. Qian, Y. et al. Non-infectious environmental antigens as a trigger for the initiation of an autoimmune skin disease. Autoimmun Rev 15,

923-30 (2016).63. Trampert, D.C., Hubers, L.M., van de Graaf, S.F.J. & Beuers, U. On the role of IgG4 in inflammatory conditions: lessons for IgG4-related

disease. Biochim Biophys Acta Mol Basis Dis 1864, 1401-1409 (2018).64. Koneczny, I. A New Classification System for IgG4 Autoantibodies. Front Immunol 9, 97 (2018).65. Zhao, J., Yan, Y., Cui, Z., Yang, R. & Zhao, M.H. The immunoglobulin G subclass distribution of anti-GBM autoantibodies against

rHalpha3(IV)NC1 is associated with disease severity. Hum Immunol 70, 425-9 (2009).

17

682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711

3334

66. Qu, Z., Cui, Z., Liu, G. & Zhao, M.H. The distribution of IgG subclass deposition on renal tissues from patients with anti-glomerular basement membrane disease. BMC Immunol 14, 19 (2013).

67. Ohlsson, S. et al. Circulating anti-glomerular basement membrane antibodies with predominance of subclass IgG4 and false-negative immunoassay test results in anti-glomerular basement membrane disease. Am J Kidney Dis 63, 289-93 (2014).

68. Cheng, G. et al. Serum phospholipase A2 receptor antibodies and immunoglobulin G subtypes in adult idiopathic membranous nephropathy: Clinical value assessment. Clin Chim Acta 490, 135-141 (2019).

69. Huang, C.C. et al. IgG subclass staining in renal biopsies with membranous glomerulonephritis indicates subclass switch during disease progression. Mod Pathol 26, 799-805 (2013).

70. Hihara, K. et al. Anti-Phospholipase A2 Receptor (PLA2R) Antibody and Glomerular PLA2R Expression in Japanese Patients with Membranous Nephropathy. PLoS One 11, e0158154 (2016).

71. Larsen, C.P., Messias, N.C., Silva, F.G., Messias, E. & Walker, P.D. Determination of primary versus secondary membranous glomerulopathy utilizing phospholipase A2 receptor staining in renal biopsies. Mod Pathol 26, 709-15 (2013).

72. Zhang, X.D. et al. Clinical implications of pathological features of primary membranous nephropathy. BMC Nephrol 19, 215 (2018).73. Yang, Y. et al. IgG4 anti-phospholipase A2 receptor might activate lectin and alternative complement pathway meanwhile in idiopathic

membranous nephropathy: an inspiration from a cross-sectional study. Immunol Res 64, 919-30 (2016).74. Hayashi, N. et al. Glomerular mannose-binding lectin deposition in intrinsic antigen-related membranous nephropathy. Nephrol Dial

Transplant 33, 832-840 (2018).75. Lhotta, K., Wurzner, R. & Konig, P. Glomerular deposition of mannose-binding lectin in human glomerulonephritis. Nephrol Dial

Transplant 14, 881-6 (1999).76. Segawa, Y. et al. IgG subclasses and complement pathway in segmental and global membranous nephropathy. Pediatr Nephrol 25,

1091-9 (2010).77. Bally, S. et al. Phospholipase A2 Receptor-Related Membranous Nephropathy and Mannan-Binding Lectin Deficiency. J Am Soc Nephrol

27, 3539-3544 (2016).78. Luo, W. et al. Alternative Pathway Is Essential for Glomerular Complement Activation and Proteinuria in a Mouse Model of

Membranous Nephropathy. Front Immunol 9, 1433 (2018).79. Seikrit, C., Ronco, P. & Debiec, H. Factor H Autoantibodies and Membranous Nephropathy. N Engl J Med 379, 2479-2481 (2018).80. Nicolas, J.P., Lambeau, G. & Lazdunski, M. Identification of the binding domain for secretory phospholipases A2 on their M-type 180-kDa

membrane receptor. J Biol Chem 270, 28869-73 (1995).81. Fresquet, M. et al. PLA2R binds to the annexin A2-S100A10 complex in human podocytes. Sci Rep 7, 6876 (2017).

18

712713714715716717718719720721722723724725726727728729730731732733734735736737738739740

3536

82. Tomas, N.M. et al. Autoantibodies against thrombospondin type 1 domain-containing 7A induce membranous nephropathy. J Clin Invest 126, 2519-32 (2016).

83. Tomas, N.M. et al. A Heterologous Model of Thrombospondin Type 1 Domain-Containing 7A-Associated Membranous Nephropathy. J Am Soc Nephrol 28, 3262-3277 (2017).

84. Herwig, J. et al. Thrombospondin Type 1 Domain-Containing 7A Localizes to the Slit Diaphragm and Stabilizes Membrane Dynamics of Fully Differentiated Podocytes. J Am Soc Nephrol 30, 824-839 (2019).

85. Kidney Disease: Improving Global Outcomes (KDIGO) Glomerulonephritis Work Group. KDIGO Clinical Practice Guidelines for Glomerulonephritis. Kidney Int Suppl 2012;2: 139-274.

86. Floege, J. et al. Management and treatment of glomerular diseases (part 1): conclusions from a Kidney Disease: Improving Global Outcomes (KDIGO) Controversies Conference. Kidney Int 95, 268-280 (2019).

87. Hamilton, P. et al. Peptide GAM immunoadsorption therapy in primary membranous nephropathy (PRISM): Phase II trial investigating the safety and feasibility of peptide GAM immunoadsorption in anti-PLA2 R positive primary membranous nephropathy. J Clin Apher 33, 283-290 (2018).

88. Ellebrecht, C.T. & Payne, A.S. Setting the target for pemphigus vulgaris therapy. JCI Insight 2, e92021 (2017).

Legend to the figures

Figure 1: Autoimmunity in membranous nephropathy

We propose two potential pathways leading to the clonal expansion of B cells into plasma cells secreting aPLA2R antibodies and generating memory B cells: 1. PLA2R antigen binds to autoreactive B cells, gets internalized and fragmented then presented to the helper T cells through the MHC class II receptor. Stimulated T cells release cytokines feeding back to the B cells to stimulate division and differentiation to plasma cells and antibody production and memory B cells. These cytokine signals can also direct class switch recombination to distinct IgG subclasses resulting in IgG4 antibodies. 2. Antigen presenting cells (APCs) recognize PLA2R antigen or microbial agent (via molecular mimicry), then process and present PLA2R/microbes fragments on MHC class II molecules on their surface. When APC’s present PLA2R in the presence of co-stimulators, self-reactive T cells are activated rather than rendered tolerant. B cells are activated outside of follicles by the combination of antigen and T cells.

19

741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770

3738

After migration to germinal centers B cells interact with helper T cells, leading to proliferation and survival. Some of the B cells undergo differentiation to either memory B cells or plasma cells and leave the germinal center. If regulatory T cells encounter their self antigen on an APC, they secrete inhibitory cytokines such as Il-10 and TGF-βeta that inhibit surrounding autoreactive T cells.

Figure 2: The lung as primary insult in membranous nephropathy: hypothetical role of pulmonary inflammation and infections Air pollution can create oxidative stress in the airways, causing airway epithelium cells and alveolar macrophages to express pro-inflammatory cytokines. These cytokines can have local effects, includinga specific stimulation of resting antigen presenting cells (APCs). In the pulmonary tissue, PLA2R or microbial agent get endocytosed and presented, triggering the maturation of airway-resident APCs. Activated APCs can in turn stimulate and activate T cells and subsequently B cells. In an individual in whom self-reactive lymphocytes have escaped mechanisms of central tolerance, these APCs initiate an autoimmune response. Plasma cells secrete and release autoantibodies in the circulation which can then reach the kidney. Accumulation of anti-PLA2R antibodies in the glomerulus leads to kidney damage and the pathogenesis of MN.

20

771772773774775776777778779780781782783

3940