Neuronal Proteins as Antigenic Targets in Membranous Nephropathy

Context: The discovery of new target antigens in membranous nephropathy (MN) has revealed new disease phenotypes and, in some cases, has suggested mechanisms of disease shared by two concurrent autoimmune diseases. Subject of Review: Several recent reports and an accompanying editorial describe the association of anti-contactin-1 (CNTN1) autoantibodies of the IgG4 subclass with a novel subtype of MN that co-occurs with a form of chronic inflammatory demyelinating polyradiculoneuropathy caused by anti-CNTN1 antibodies. CNTN1, the cellular source of which is still undetermined, is identified as the target antigen in the kidney since it is present within glomerular subepithelial deposits and anti-CNTN1 IgG4 antibodies can be eluted from the corresponding kidney biopsy tissue. Second Opinion: These new reports reinforce recent findings that many proteins targeted in several other types of primary and secondary MN are proteins whose expression is shared by podocytes and neurons. While complement-mediated podocyte damage represents a well-established paradigm in the pathogenesis of MN, interference with the normal functions of these shared proteins by autoantibodies should be considered as another potential mechanism of glomerular injury to be explored in future research.

© 2022 S. Karger AG, Basel

Antigen discovery in adult membranous nephropathy (MN) began in 2009, with the identification of the phospholipase A2 receptor (PLA2R) as a commonly targeted and podocyte-expressed autoantigen in primary MN [1]. This was followed 5 years later by the identification of a second podocyte protein, thrombospondin type 1 domain-containing 7A (THSD7A), as a minor antigen in primary MN and a potential etiologic agent in malignancy-associated MN [2, 3]. Over the last 3 years and due to a methodologic change that has accelerated antigen discovery, there has been exponential growth in our identification and understanding of novel antigens in MN [4] (see Table 1). In some cases, the autoantigen may point to particular demographic groups or associations with specific disease states, such as the finding of semaphorin 3B (SEMA3B) as target antigen in a predominantly pediatric form of disease [5], or neural cell adhesion molecule 1 (NCAM1) as the protein targeted in cases of class V lupus nephritis, some of which were associated with neuropsychiatric disease [6]. Recent reports by Le Quintrec et al. [7] and Plaisier et al. [8] and an accompanying editorial by Debiec and Ronco [9] shed light on another rare but important association: MN that occurs in the setting of an autoimmune form of demyelinating neuropathy.

Table 1.

The known autoantigens in various subtypes of membranous nephropathy and what is known about their expression by the podocyte and their function in the nervous system

/WebMaterial/ShowPic/1480896

Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) is a rare, progressive autoimmune disease of the peripheral nervous system that can be likened to a chronic form of the related and more acute Guillain-Barré syndrome. Like primary MN, the median age of onset is in the 50s and 60s, and men are more often affected than women. In a small proportion of patients, IgG4-predominant autoantibodies target and appear to functionally inhibit proteins expressed in the node of Ranvier that assist in binding and organization of the Schwann cell-derived myelin sheath to the neuronal axon [10]. Autoantibodies targeting nodal and paranodal proteins such as neurofascin (NF) 155, contactin-1 (CNTN1), and contactin-associated protein 1 (CASPR1) have been found in 1%, 0.7%, and 0.2% of cases, respectively, in a CIDP bioregistry [11]. There have also been very rare cases in which autoantibodies to another NF isoform, NF186, have been detected [12]. CIDP associated with anti-CNTN1 antibodies has been known to be associated with concurrent MN in a significant proportion of cases [13]. A common antigen shared between podocytes and neurons had previously been suggested, but its identity had not been revealed until now.

In a recent report of 5 cases with CIDP that were associated with both anti-CNTN1 antibodies and MN, found by virtue of a larger CIDP bioregistry [11], Le Quintrec and colleagues identify CNTN1 by immunofluorescence within the glomerular subepithelial immune deposits in the biopsies of these individuals and are able to elute IgG4-predominant anti-CNTN1 antibodies from the biopsy tissue [7], implicating these autoantibodies in the pathogenesis of the kidney disease. They demonstrate that IgG eluted from CNTN1-associated, but not PLA2R-associated, MN cases is able to bind to the paranodal regions of mouse sciatic nerve. Although podocyte expression of CNTN1 has not been convincingly shown (see below), the authors demonstrate by immunoblotting with a commercial antibody that CNTN1 can be detected in extracts from both human glomeruli and brain, and they additionally confirm the presence of CNTN1 in the glomerular band by mass spectrometric analysis. The report stops short of determining whether the serum-derived IgG4 autoantibodies can also detect the antigen within glomerular extract.

In several additional case reports that were published in the same time period, Santoro et al. [14] and Plaisier et al. [8] also demonstrate that CNTN1 is present within immune deposits in this rare form of MN associated with CIDP. Other recent reports continue to demonstrate the association of MN with cases of CIPD specifically associated with anti-CNTN1 antibodies [15, 16]. The B cell depleting anti-CD20 agent rituximab was used as therapy in several of these cases and led to improvement in both neurological features of the CIDP and remission of the nephrotic syndrome from MN [8, 14]. As rituximab is currently a first-line treatment for MN and can lead to the decline and disappearance of circulating anti-podocyte antibodies such as anti-PLA2R, it is not surprising that it could have a similar effect on the autoantibodies in CIDP.

Expressed by myelinated nerve fibers in the paranodal region of the node of Ranvier, CNTN1 is a glycosylphosphatidylinositol (GPI)-linked protein that partners with another axonal membrane protein CASPR1 to bind to NF155, which is expressed by the adjacent Schwann cell (Fig. 1). As a component of septate-like junctions, this cell-cell adhesion complex plays a crucial role in the attachment of myelin loops to the axonal surface to maintain distinct molecular domains at the nodes of Ranvier [17]. Deficiency in contactin results in disturbance of septate-like junctions and an aberrant organization of the paranodal region, which causes a severe reduction in nerve conduction velocity [18].

Fig. 1.

Potential pathomechanisms of anti-CNTN1 antibodies in the nerve and glomerulus. a In the paranodal region of the peripheral nerve, neurofascin (NFSC) 155 on the terminal myelin loop binds to axonal contactin-1 (CNTN1) complexed with CASPR1 (not shown) to link the myelin sheath to the axon. In CIDP, anti-CNTN1 antibodies disrupt this adhesive interaction, leading to demyelination. b Possible mechanisms by which anti-CNTN1 antibodies may lead to subepithelial immune complex formation and complement activation (CA) in membranous nephropathy. (1) Autoantibodies may target CNTN1 that becomes expressed on the podocyte foot processes to form in situ immune complexes. More speculatively, anti-CNTN1 could interfere with a normal biological function of CNTN1 in the podocyte, such as binding to NFSC on adjacent foot processes or tenascin C (TNC) in the glomerular basement membrane. (2) If the source of CNTN1 turns out to be exogenous to the podocyte, the protein may arrive via the circulation and become planted in a subepithelial position (possibly binding to NFSC or TNC) and later become the target of anti-CNTN1 antibodies. (3) Finally, there may be preformed immune complexes of CNTN1 and anti-CNTN1 that preferentially deposit in the subepithelial position. In all cases, there is likely complement activation (CA) that damages the podocyte and filtration barrier.

/WebMaterial/ShowPic/1480894

CNTN1 is a multidomain extracellular-facing protein composed of six N-terminal-type C2 immunoglobulin domains and four fibronectin type III-like domains, as well as multiple sites for Asn-linked glycosylation [19, 20]. It shares a similar domain structure with the NF isoforms (NF186, NF155, and NF140), which also have six N-terminal immunoglobulin-type domains and a variable number of C-terminal fibronectin type-3 domains. In contrast to our knowledge of the specific domains of PLA2R targeted by circulating anti-PLA2R autoantibodies [21], the particular epitope(s) or domains targeted by anti-CNTN1 are not yet known.

Not all patients afflicted with CIDP and anti-CNTN1 antibodies develop MN, and those that do develop both diseases tend to be younger, more often presenting before age 60 [13]. This calls into question whether there could be differences in the epitopes targeted, the exposure of these epitopes, potentially different post-translational modifications of the antigen in each tissue, or variability in presentation of the antigen by inherited class II HLA molecules. While CIPD has been associated with DR3, DQ2, and HLA-DR15:01 [22, 23], similar to MN in general [24], whether or not the CNTN1-associated forms of each disease specifically share these or other HLA haplotypes is not known. In their Supplemental Information, Le Quintrec et al. [7] show that the anti-CNTN1 titer is higher in those with MN and CIDP when compared to those with CIDP alone. This could potentially mirror the situation in MN, in which epitope spreading to additional domains in the PLA2R molecule has been associated with both higher titer and more severe disease [21, 25].

Why is this neuro-renal syndrome associated with anti-CNTN1 autoantibodies of interest to the nephrology community? It has long been observed that neurons and podocytes have similarities that transcend their disparate host organs and functions. Both are post-mitotic cell types with profound and complicated cytoarchitectural structures that form the polarized and highly arborized neuronal dendritic tree or podocyte foot processes. The collection of podocyte-expressed proteins (e.g., synaptopodin and synaptotagmin) whose names reference synaptic function exemplify the shared proteins and mechanisms in these two otherwise distinct cell types [26, 27].

It is intriguing that among the set of proteins shared between podocytes and neurons, quite a few have been demonstrated to be target antigens in MN (Table 1). To date, THSD7A [28], SEMA3B [29], protocadherin 7 (PCDH7) [30], protocadherin FAT1 [31], netrin G1 [32], neural EGFL-like 1 (NELL1), and NCAM1 have been shown to play roles in neuritogenesis or axon guidance. Exostosin 1 (EXT1), serine protease HTRA1, and TGF-beta receptor 3 (TGFBR3 or betaglycan) may additionally have peripheral roles in normal neuronal physiology. For example, the genetic loss of EXT1 causes severe axon guidance errors of the major commissural axon tracts [33]. However, the most commonly targeted podocyte antigen in primary MN, PLA2R, shows no connection to neuronal physiology or cell biology to the best of our knowledge. The report by Le Quintrec and colleagues and the related case reports add CNTN1 to the list of shared molecules, and in this case, the humoral autoimmune attack appears to target both organs, causing CDIP and MN.

Studies decades ago in the experimental rat model of Heymann nephritis shifted the paradigm view of the source of the subepithelial immune deposits in MN from circulating complexes to in situ formation, due to recognition of the fact that the rat podocyte expresses megalin/LRP2 [34]. This was supported by the identification of the first target antigens in human MN: neutral endopeptidase, PLA2R, and THSD7A, all of which are expressed by the podocyte. However, with several of the more recently identified target antigens, the question about the source of the antigen (intrinsic vs. exogenously delivered as a planted antigen or circulating immune complex) has surfaced once again.

CNTN1 has not been clearly established as a podocyte-expressed protein. An early report showed only a weak signal in total kidney mRNA by Northern blot [35]. The weak immunoblot signal demonstrated by Le Quintrec et al. within glomerular protein extract is supported only by the identification of several CNTN1 peptides cut from the gel band by mass spectrometry, but the source of the protein is unknown. While the authors mention that cultured podocytes yield a signal by PCR, these data are not shown. Public databases such as the Kidney Precision Medicine Project Atlas Explorer (www.atlas.kpmp.org/explorer) document very low levels of CNTN1 mRNA in only a fraction of the podocytes, but whether this message is translated into detectable amounts of protein is unclear. The Human Protein Atlas (www.proteinatlas.org) does not suggest that there is any substantial glomerular or podocyte expression of CNTN1 protein.

It has been noted as an alternative that CNTN1 can be found in the circulation [9], raising the possibility that the protein could arrive in the form of a circulating immune complex or a planted antigen (see Fig. 1). While one can think of mechanisms by which a circulating antigen might be directed to an exclusively subepithelial position, the absence (as is the case in anti-CNTN1-associated MN) of mesangial or subendothelial deposits has been previously used as an argument for in situ complex formation at the basal surface of the podocyte. As an example of how a CNTN1 could be captured and retained in the glomerulus, circulating CNTN1 could potentially bind to tenascin C in the glomerular matrix [36]; however, tenascin C is felt to be a mesangial- and not a podocyte-expressed protein.

Other possibilities may exist to explain the apparent absence of a target antigen in normal glomeruli. Rare genetic mutations could cause altered tissue expression of the gene in susceptible individuals. The target antigen could be upregulated with age, podocyte stress, or the gradual accumulation of “replacement” podocytes (regenerated from parietal or hilar sources; this has been suggested for LRP2 in anti-LRP2 nephropathy [37]). Very low or nearly undetectable levels of protein expression in the presence of pre-existing autoantibodies could lead to immune deposits that accumulate slowly over time, leading to more podocyte injury and further upregulation of the target protein. The discovery of target antigens such as CNTN1 or NELL1, not typically felt to be expressed by podocytes, makes it evident that lessons learned from the prototype podocyte target antigen PLA2R cannot be universally applied in all subtypes of MN and that more creative thinking and research is required in the field.

Heymann nephritis has also established the paradigm that the major pathologic mechanism of humoral autoimmunity in MN is through complement-mediated injury to the podocyte [34, 38]. This is supported by the consistent finding of relevant complement factors such as C3 and C4 within human MN immune deposits. Although there is minimal evidence for classical complement pathway activation in human disease, anti-PLA2R autoantibodies of the IgG4 subclass that are modified by hypogalactosylated glycans have been shown capable of initiating the lectin pathway of complement activation [39]. In the current cases of CNTN1-associated MN reported by Le Quintrec and colleagues, C3 was found in immune deposits in all five cases, although at relatively weak intensity [7].

The ability of autoantibodies, in other contexts, to block protein function raises the question of whether anti-CNTN1 (or other recently identified) autoantibodies might interfere with specific podocyte functions important for stabilization of foot process and slit diaphragm integrity. In vitro evidence suggests that anti-THSD7A antibodies can affect cultured podocyte adhesion and viability [40], and localization of THSD7A immediately beneath the slit diaphragm [41] suggests tantalizing mechanisms by which such autoantibodies could lead to proteinuria without necessarily invoking the traditional mechanism of complement activation. Of note, NF has been identified as a novel component within podocyte major processes [27] and public databases such as the Human Protein Atlas show clear expression of NF in the human podocyte. Yet the nephrotic syndrome that occurs in CIDP in conjunction with anti-NF155 antibodies has been more closely linked with minimal change disease or FSGS than with MN [42], suggesting that autoantibodies targeting a podocyte protein may not universally lead to subepithelial immune deposits and MN. Anti-nephrin autoantibodies may stimulate disruption of the glomerular filtration barrier in a similar manner, rather than from discrete immune deposits [43].

As is so often the case, exciting new findings and associations only serve to generate more unanswered questions. Where is CNTN1 initially exposed to the immune system? Are the epitopes targeted on CNTN1 in the nodes of Ranvier the same as those targeted in the kidney? While the authors show that human serum can bind to the paranodal region of mouse sciatic nerve, they have not provided evidence that the IgG4 anti-CNTN1 antibodies can bind to a (presumedly) glomerular form of the protein present in glomerular extract. The timing of onset of the two diseases relative to each other may be dependent on whether injury in the glomerulus is primarily complement-mediated (in which case it would take months to develop before significant proteinuria and nephrosis begin [44]) or due to a more rapid antibody-mediated inhibition of some physiologic pathway needed to sustain the integrity of the podocyte foot processes. We look forward to the future research that will help address these questions.

The work reported by Le Quintec and colleagues has capitalized on the existence of large disease biorepositories from which infrequent disease types can be found and studied at the molecular level [7, 11]. This is reminiscent of the large archives of MN kidney biopsy tissue from which cases without a known target antigen were selected for laser capture microdissection and protein G tissue immunoprecipitation followed by mass spectrometric analysis, yielding new antigens and new associations [4]. Similarly, it is now feasible to assemble a small cohort of individuals with rare concurrences of diseases and to study these cases at the molecular level. In this way, the protocadherin FAT1 was identified as the predominant target antigen associated with a form of MN that occurs following allogeneic hematopoietic stem cell transplantation [45].

Although both diseases are extremely rare, the convergence of CIPD and MN in the context of circulating anti-CNTN1 antibodies brings to the surface questions in MN that have existed for decades (e.g., the source of the subepithelial deposits) and reinvigorates our thinking about the neuronal and axon guidance pathways at play in normal podocyte and glomerular filtration barrier physiology. It is important to continuously challenge existing disease paradigms in the face of new clinical information, no matter how rare, such that we can better understand the true underpinnings of these fascinating autoimmune diseases and develop precision and more rationally applied clinical therapies.

Acknowledgment

We would like to acknowledge Dr. Alfred Cheung for his continuous support of the studies of membranous nephropathy during his tenure as Division Chief.

Conflict of Interest Statement

Dr. Beck reports having consultancy agreements with or advisory board payments from Alexion, Ionis, Novartis, and Visterra; reports receiving honoraria from UpToDate, Inc.; reports patents and inventions as coinventor on, and receiving royalties related to, the US patent “Diagnostics for Membranous Nephropathy”; and reports being a scientific advisor or member of Kidney Medicine editorial board. Dr. Al-Rabadi has no conflicts of interest to declare.

Funding Sources

There are no funding sources to report.

Author Contributions

This commentary was jointly written by Dr. Laith Al-Rabadi and Dr. Laurence Beck.

Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.

留言 (0)

沒有登入
gif