Complement is Complimentary in Membranous Nephropathy

The complement system provides an ancient, highly conserved, cascade of interacting proteins orchestrating innate and adaptive immune responses. In addition to protecting the host from infections, the complement contributes to the pathogenesis of various disorders, including kidney diseases.1 The exploration of complement-directed treatment strategies in these diseases, including several glomerulonephritides, makes eminent sense. ANCA vasculitis provides the latest example in which preclinical models implicated the complement as an important disease mediator and where the clinical application was not "lost in translation." Experimental studies in cell and murine vasculitis models identified the interaction of the anaphylatoxin C5a with the activating C5a receptor-1 (C5aR1 or CD88) as a central disease process.2 The human C5aR1 was then introduced into mice, allowing the development of a small molecule that blocked the human C5aR1.3 This complement-directed strategy was recently explored in a randomized controlled trial.4 The oral C5aR blocker, avacopan, was superior to prednisone taper with respect to sustained remission at 52 weeks. Avacopan is now on its way into routine clinical practice, and it will be our task to position this new tool within existing ANCA vasculitis treatments. Membranous nephropathy (MN) has the potential to become an additional example for a complement-targeted approach. MN is an autoimmune disease featuring autoantibodies to several podocyte proteins, the formation of immune complexes along the glomerular basement membrane, glomerular complement deposition, and nephrotic range proteinuria.5 Most patients with primary MN harbor autoantibodies to the M-type phospholipase A2 receptor 1 (PLA2R1).6 More than 40 years ago, complement depletion by means of cobra venom factor protected rats from proteinuria in a Heymann nephritis model of MN.7 In this issue of JASN, Gao et al.8 explored the interaction of the anaphylatoxin C3a with the C3a receptor as a disease mechanism in MN and as a potential treatment target. First, the investigators showed that patients with MN, almost all with PLA2R1-positive kidney biopsy specimens, had increased plasma C3a levels together with strong C3a receptor expression on podocytes that correlated with proteinuria and serum creatinine. When human podocytes were exposed to plasma from patients with MN in vitro, PLA2R1 increased and podocyte injury occurred. Both effects were reduced with C3a receptor blockade. Using again the Heymann nephritis rat model, Gao et al. observed that both plasma C3a and the podocyte C3a receptor were increased, similarly to what they observed in their cohort of patients with primary MN. Importantly, a C3a receptor antagonist significantly reduced proteinuria and glomerular injury in early and late treatment protocols. Glomerular rat Ig deposition also decreased with this treatment, as did the deposition of C1q, factor B, and the C5b-9 complex. Plasma from the experimental rats upregulated C3a receptors on rat podocytes in vitro and caused cell injury that was prevented by the C3a receptor blocker, recapitulating the findings from the human podocyte model. Together, these findings support the notion that C3aR-dependent effects were at work in patients with MN and rats, leading to podocyte injury and proteinuria. Megalin is the receptor serving as the antigen causing Heymann nephritis in rats; however, megalin is not expressed in human glomeruli. In contrast, the majority of patients with primary MN harbor autoantibodies against PLA2R1, an antigen that is not present in glomeruli of rodents. Moreover, additional antigens have been characterized in human MN. Thus, rat and mouse models do not capture all human disease features, including differences in the inflicting antigens and Fcγ receptors. Better model systems are required, for example, to define the initial trigger for complement activation in PLA2R1 autoantibody–associated MN.9 PLA2R1 antibodies are found along the glomerular basement membrane, together with complement cleavage products. However, PLA2R1 antibodies are predominantly of the IgG4 subclass that is incapable of classic complement pathway activation via C1q. Is there an alternative complement activation trigger used by PLA2R1 IgG4 antibodies—independent of C1q binding? In the absence of a suitable PLA2R1 animal model for MN, Haddad et al.10 recently studied complement activation in a human podocyte cell culture model using PLA2R1 antibodies containing patient sera or isolated IgG4 preparations. The investigators found that the PLA2R1 IgG4 antibodies activated the lectin pathway and induced injury in PLA2R1-expressing podocytes in a manner dependent on MBL serine proteases (MASPs). Patient IgG4, but not control IgG4, strongly bound MBL and subsequently activated the complement cascade. The lectin-binding property of the IgG4 antibodies and, therefore, the activation MASPs, depended on the glycosylation status of the IgG4 antibodies. Patients with MN had glycosylation alterations in their IgG4 that fixed more lectin, correlating with podocyte injury. Lectin pathway activation resulted in generation of C5b-9 and the anaphylatoxins C3a and C5a. C6 deficiency and combined gene silencing or pharmacologic inhibition of C3aR1 and C5aR1 strongly reduced podocyte injury, underscoring the treatment potential of these findings. However, it is important to point out that, in contrast to the data obtained by Gao et al. in Heymann nephritis, silencing the C3aR1 alone did not provide protection in the human cell culture model. Moreover, glomerular MBL deposition was not detected by Gao et al., suggesting a different complement trigger in the rat model compared with PLA2R1 antibody–positive MN in patients. Nevertheless, both studies underscore the importance of complement as a disease mechanism and as a promising treatment target (Figure 1). Conceivably, more than one activation mechanism is at work (e.g., MBL and alternative pathway) and their specific contribution may differ between the distinct human MN forms and even between individual patients with their specific genetic background. Humanized rodent and minipig models that resemble primary MN (e.g., podocyte PLA2R1 expression) are currently being developed and will be invaluable for further evaluating complement-directed treatments. Moreover, clinical studies in patients with MN are already on the way, exploring the inhibition of the lectin pathway by MASP-2 and the alternative pathway using a factor B inhibitor. Nephrologists should stay tuned as additional glomerulonephritides will be tested for complement-targeted treatment strategies in the near future.Figure 1.: Complement as a disease mediator and treatment target in primary MN. Circulating IgG4 autoantibodies to PLA2R1 bind their target antigens on the podocyte cell membrane (red circles). These subepithelial immune complexes contain galactose-deficient IgG4 that strongly binds MBL, thereby activating MASPs that, in turn, generate the C3 convertase. Cleavage of C3 generates the C3a split product that finds the C3aR1 on podocytes. C3aR1 engagement causes C3aR and PLA2R1 upregulation, contributing to podocyte injury. Potential complement-directed treatment strategies derived from these findings include MASP and C3aR1 inhibition, possibly together with C5aR blockade. EC, endothelial cell; GBM, glomerular basement membrane.Disclosures R. Kettritz reports serving on advisory boards for Boehringer Ingelheim, Insmed Incorporated, and Vifor Pharma; and receiving honoraria from med update (medical education) and Vifor Pharma. A. Schreiber reports serving in an advisory or leadership role for Alexion, Hansa Biopharm, Otsuka, Sanofi, and Travere; having consultancy agreements with Alexion, Hansa Biopharm, Otsuka, Sanofi, Travere, and Vifor Pharma; having other interests in, or relationships with, Deutsche Gesellschaft für Nephrologie; receiving research funding from Eleva GmbH; and receiving honoraria from, and serving on the advisory board for, Vifor Pharma. Funding None.