O-glycosylation of IgA1 and the pathogenesis of an autoimmune disease IgA nephropathy

Abstract

IgA nephropathy is a kidney disease characterized by deposition of immune complexes containing abnormally O-glycosylated IgA1 in the glomeruli. Specifically, some O-glycans are missing galactose that is normally β1,3-linked to N-acetylgalactosamine of the core 1 glycans. These galactose-deficient IgA1 glycoforms are produced by IgA1-secreting cells due to a dysregulated expression and activity of several glycosyltransferases. Galactose-deficient IgA1 in the circulation of patients with IgA nephropathy is bound by IgG autoantibodies and the resultant immune complexes can contain additional proteins, such as complement C3. These complexes, if not removed from the circulation, can enter the glomerular mesangium, activate the resident mesangial cells, and induce glomerular injury. In this review, we briefly summarize clinical and pathological features of IgA nephropathy, review normal and aberrant IgA1 O-glycosylation pathways, and discuss the origins and potential significance of natural anti-glycan antibodies, namely those recognizing N-acetylgalactosamine. We also discuss the features of autoantibodies specific for galactose-deficient IgA1 and the characteristics of pathogenic immune complexes containing IgA1 and IgG. In IgA nephropathy, kidneys are injured by IgA1-containing immune complexes as innocent bystanders. Most patients with IgA nephropathy progress to kidney failure and require dialysis or transplantation. Moreover, most patients after transplantation experience a recurrent disease. Thus, a better understanding of the pathogenetic mechanisms is needed to develop new disease-specific treatments.

Introduction

IgA nephropathy (IgAN), initially called Berger disease in the honor of the discoverer, was described in 1968 based on pathological analyses of kidney-tissue biopsies obtained from a group of patients with recurrent hematuria (Berger and Hinglais 1968). Specifically, the use of immunofluorescence microscopy with antibodies specific for human immunoglobulins and complement C3 proteins revealed that these patients had glomerular immunodeposits consisting of IgA, IgG, and C3. Moreover, these features were associated with glomerular mesangioproliferative pathological changes. IgA vasculitis with nephritis (IgAVN), previously called Henoch-Schönlein purpura with nephritis, is thought to be a disease related to IgAN (Davin et al. 2001; Davin 2011). Patients with IgAVN also exhibit glomerular IgA-containing immunodeposits and similar types of renal pathology (Lau et al. 2010). Thus, IgAN and IgAVN are often considered similar diseases with a different spectrum of clinical manifestations (Waldo 1988; Wyatt et al. 2006; Kiryluk et al. 2011; Kamei et al. 2016; Hastings et al. 2022). IgAN has been recognized as the most common primary glomerulonephritis in the world (Julian et al. 1988).

Patients with IgAN often exhibit a synpharyngitic hematuria at disease onset or at the time of disease activity (Wyatt and Julian 2013), thus linking upper-respiratory tract infections with IgAN. Proteinuria may be part of the initial disease manifestation. Diagnosis of IgAN requires pathological assessment of kidney biopsy specimens by routine immunofluorescence, light microscopy with histological stains (e.g. hematoxylin and eosin and periodic acid Schiff stains), and electron microscopy. IgA is detected by routine immunofluorescence as the dominant/codominant immunoglobulin, with variable amounts of IgG and/or IgM. Complement C3 is usually detected, whereas C1q is absent, suggesting a participation of the alternative complement pathway (Maillard et al. 2015). These IgA-containing immunodeposits are usually detected by electron microscopy as electron-dense deposits in the mesangial regions. Light-microscopic features usually reveal mesangioproliferative changes. Patients with IgAN have increased mortality, and life expectancy is reduced by 6–10 years (Hastings et al. 2018; Jarrick et al. 2019). A recent study showed that most IgAN patients progress to kidney failure within 10–15 years from diagnosis and that only few patients may avoid kidney failure in their lifetime (Pitcher et al. 2023). IgAN patients with kidney failure require renal replacement therapy, i.e. dialysis or transplantation (Wyatt and Julian 2013). However, more than 50% of patients with kidney transplant develop recurrent IgAN within 5 years (Berger 1988; Odum et al. 1994; Floege 2004; Chandrakantan et al. 2005). Thus, a better understanding of the disease pathogenesis is needed to inform development of disease-specific therapies to reduce IgAN-associated morbidity and mortality.

IgA in the glomerular immunodeposits of patients with IgAN is exclusively of IgA1 subclass (Conley et al. 1980). Furthermore, these glomerular immunodeposits are enriched for glycoforms of IgA1 with some O-glycans deficient in galactose, termed galactose-deficient IgA1 (Gd-IgA1) (Allen et al. 2001; Hiki et al. 2001). This finding is consistent with earlier observations of the abnormal O-glycosylation of IgA1 in the circulation of patients with IgAN [for review see (Novak et al. 2001; Julian and Novak 2004; Kiryluk et al. 2013; Kiryluk and Novak 2014; Lai et al. 2016; Novak et al. 2018; Reily et al. 2019; Hansen et al. 2021)] and the proposed role of these Gd-IgA1 glycoforms in the pathogenesis of the disease (Mestecky et al. 1993). Current understanding of the pathogenesis defines IgAN as an autoimmune disease with aberrantly O-glycosylated IgA1. These abnormal IgA1 glycoforms, i.e. Gd-IgA1, are recognized as autoantigens by autoantibodies, mainly of IgG isotype, resulting in the formation of pathogenic immune complexes in the circulation (Suzuki et al. 2011). Some of these circulating immune complexes deposit in the glomeruli and induce kidney injury. Here, we review current knowledge of IgA1 O-glycosylation pathways, immune responses to aberrantly glycosylated IgA1, formation of nephritogenic Gd-IgA1-containing immune complexes, and the subsequent processes resulting in a mesangioproliferative glomerular injury. As environmental factors are associated with IgAN, we also discuss the origin of natural anti-glycan antibodies recognizing N-acetylgalactosamine (GalNAc) and the corresponding microbial glycoconjugates.

IgA immunoglobulins in humans

IgA, together with IgM, represent the main antibody isotypes in mucosal secretions, including for example tears, saliva, colostrum, milk, nasal secretions, and intestinal fluids (de Sousa-Pereira and Woof 2019). In the circulation, IgA is the second most abundant isotype; healthy individuals have approximately 2 mg of IgA per ml of serum. IgA-producing cells secrete IgA as a monomeric or polymeric molecular form. Monomeric IgA has two heavy chains and two light chains; polymeric IgA has two or more of the IgA monomers connected by disulfide bridges between the joining chain (J-chain) and the tail segment of the IgA heavy chain (Mestecky et al. 1974; Brandtzaeg 1981; Yoo et al. 1999; Yoo and Morrison 2005). The main form of IgA in mucosal secretions is secretory IgA. This molecular form contains the secretory component, a glycoprotein derived by proteolytic cleavage from the polymeric immunoglobulin receptor during transcytosis of polymeric IgA through mucosal epithelial cells (Brandtzaeg and Prydz 1984). Mucosal surfaces of the respiratory, gastrointestinal, and genitourinary tracts contain high amounts of secretory IgA that can neutralize and eliminate pathogens to protect the host against infections (Woof and Kerr 2006).

Humans have two subclasses of IgA, IgA1 and IgA2, that share a high degree of amino-acid sequence identity (Putnam et al. 1979; Mestecky and Russell 1986; Putnam 1989). However, IgA1 and IgA2 have substantial differences in glycosylation (Baenziger and Kornfeld 1974a; Tomana et al. 1976; Mattu et al. 1998; Hansen et al. 2021). Heavy chains of IgA1 have two N-glycans and those of IgA2 have up to five N-glycans (Baenziger and Kornfeld 1974b). Moreover, IgA1 has clustered O-glycans in the hinge region of the heavy chains (Frangione and Wolfenstein-Todel 1972; Baenziger and Kornfeld 1974a; Mattu et al. 1998). These O-glycans, usually 3 to 6 per hinge region, are typically core 1 glycans, i.e. consisting of GalNAc with β1,3-linked galactose, that can be sialylated (Mattu et al. 1998; Renfrow et al. 2005; Takahashi et al. 2010; Takahashi et al. 2012). Together, these distinct glycosylation features contribute to the functional differences of human IgA subclasses (Moura et al. 2004; Moura et al. 2005; Matysiak-Budnik et al. 2008; Papista et al. 2015; Steffen et al. 2020).

Circulatory IgA is predominantly a monomer (90% of total serum IgA), with approximately 10% being contributed by polymeric forms (Mestecky et al. 1986; Novak et al. 2001). Most circulatory IgA is of IgA1 subclass (90% of serum IgA). It is thought that the monomeric form of circulatory IgA is produced mainly by plasma cells in the bone marrow. Conversely, polymeric IgA is thought to be of mucosal origin, being produced by plasma cells close to the mucosal epithelium [for review see (Novak et al. 2001; Novak et al. 2018)]. As polymeric IgA1 is the main molecular form of IgA1 in the immunodeposits in IgAN (Conley et al. 1980; Monteiro et al. 1985), it thus raises a question whether this IgA1 is produced by plasma cells in mucosal tissues and whether upper-respiratory tract infections during synpharyngitic hematuria may be involved in the process (Suzuki et al. 2014; Novak et al. 2018).

Biosynthesis of IgA1 O-glycans and aberrant O-glycosylation in IgAN

IgA1 hinge region is comprised of two octapeptide repeats (Fig. 1A) (Frangione and Wolfenstein-Todel 1972), and carries 3 to 6 O-glycans (Baenziger and Kornfeld 1974a). Circulatory IgA1 contains core 1 glycans, i.e. GalNAc with β1,3-linked galactose (Mattu et al. 1998). This disaccharide can be modified by one or two sialic acid residues, galactose can have α2,3-linked sialic acid and GalNAc can have α2,6-linked sialic acid (Takahashi et al. 2012). IgA1 O-glycans can also be without galactose, representing the so-called galactose-deficient glycans (Allen 1995; Allen et al. 1995; Tomana et al. 1997; Tomana et al. 1999; Moldoveanu et al. 2007). Those galactose-deficient glycans contain either GalNAc or GalNAc with α2,6-linked sialic acid. Notably, secretory IgA1 can contain additional O-glycan types, such as core 2 O-glycans with additional modifications not present in the O-glycans of circulatory IgA1 (Royle et al. 2003).

Fig. 1.

Hinge-region amino-acid sequence and O-glycosylation of circulatory IgA1. A) IgA1 usually has three to six sites with attached O-glycans. The six commonly used sites include T225, T228, S230, S232, T233, and T236 (marked by stars in the amino-acid sequence). B) Biosynthesis of O-glycans of circulatory IgA1. Glycosylation of IgA1 begins in the Golgi apparatus with the addition of N-acetylgalactosamine (GalNAc) to S and T residues by GalNAc-transferase (GalNAc-T). GalNAc can be then modified by core 1 synthase (C1GalT1) with a β1,3-linked galactose (Gal). This enzyme requires expression of C1GalT1-specific chaperone 1 (C1GalT1C1). In the follow-up reaction, one or both sugars can be modified by sialic acid: Galactose by α2,3-linked sialic acid and GalNAc by α2,6-linked sialic acid. These reactions are catalyzed by ST3Gal and ST6GalNAc2 enzymes, respectively. Alternatively, GalNAc can be modified before galactosylation by α2,6-linked sialic acid; this modification blocks any other glycosylation and thus yields a Gal-deficient site, also termed sialyl-Tn antigen. If GalNAc remains unmodified, it is called Tn antigen.

IgA1 hinge-region O-glycans are synthesized in a stepwise process (Fig. 1B), similarly as other mucin-type O-glycans (Bennett et al. 2012; Revoredo et al. 2016; Brockhausen et al. 2022). The initiation step, addition of GalNAc to a Ser or Thr residue, is catalyzed by a GalNAc-transferase (GalNAc-T) in the Golgi apparatus (Gerken et al. 1997; Gerken 2004; Gerken et al. 2011; Bennett et al. 2012; de Las Rivas et al. 2019; Daniel et al. 2020). Among the 20 human GalNAc-T isoenzymes, GalNAc-T2 is thought to be the main enzyme responsible for the initiation of O-glycosylation of IgA1 in IgA1-producing cells (Iwasaki et al. 2003). In addition, several other GalNAc-Ts can initiate O-glycosylation of IgA1 (Wandall et al. 2007). This conclusion is based on testing various peptide acceptors in enzymatic reactions with different GalNAc-Ts. In the next step, GalNAc can be modified by addition of β1,3-linked galactose. This step is catalyzed by core 1 synthase (glycoprotein-N-acetylgalactosamine 3-β-galactosyltransferase 1; C1GalT1) (Ju et al. 2002). Production of active C1GalT1 enzyme requires C1GalT1-specific chaperone 1 (C1GalT1C1) (Ju and Cummings 2002). After galactosylation of GalNAc, one or both sugars can be modified with sialic acid in reactions catalyzed by ST3Gal and ST6GalNAc2 enzymes (Fig. 1B) (Novak et al. 2001; Stuchlova Horynova et al. 2015; Stewart et al. 2021). If GalNAc is sialylated by ST6GalNAc2 before galactosylation, this disaccharide cannot be galactosylated (Suzuki et al. 2014). GalNAc can also remain without galactose or sialic acid; this structure is called Tn (“T nouvelle”) antigen (Dausset et al. 1959; Ju and Cummings 2005).

The above-described step-wise synthesis of IgA1 O-glycans is simplified in a sense that it considers a single site of glycosylation at a time. However, the IgA1 hinge region has a cluster of nine potential sites of O-glycosylation, thus the hinge-region amino-acid sequence serves as the initial unmodified template. The addition of a monosaccharide at one of the nine potential sites generates a new template for further glycan addition (Stewart et al. 2019; Daniel et al. 2020; Ballard et al. 2023). Thus, this process of sequential attachment of glycans at additional sites has implications for the subsequent addition of sugars at the remaining sites. As an example of the complexity of clustered O-glycosylation, we can consider the initial steps, attachment of GalNAc residues catalyzed by GalNAc-T enzymes. Most of the GalNAc-Ts, including GalNAc-T2, have two domains, a catalytic domain and a lectin domain. In general, the lectin domains exhibit glycan-binding specificity for GalNAc and are required for high-density glycosylation of GalNAc-containing glycopeptides. Moreover, lectin domains of some GalNAc-Ts exhibit an extrinsic activity that can control activity of glycosyltransferases involved in O-glycosylation pathways. This regulatory activity, described as a regulatory scaffold, involves protein–protein interactions that in some instances are substrate specific, and can exert inhibitory activity on other enzymes (Lorenz et al. 2016). The two domains of GalNAc-T2 have distinct and complementary roles in the biosynthesis of multiple sites of O-glycosylation in close proximity in the IgA1 hinge region. While the selection of the first site of addition is driven by the catalytic domain, the subsequent site-selection is influenced by the lectin domain (Stewart et al. 2019). The final glycan density, i.e. the number of GalNAc residues added to the hinge-region segment, can also be affected by the expression and activity of other GalNAc-T isoenzymes that can glycosylate IgA1 (Wandall et al. 2007). Still, there is consistent fidelity to the final distribution of IgA1 O-glycans ranging from three to six O-glycans per hinge region. Notably, selection of the initial site of GalNAc attachment impacts the utilization of sites for follow-up glycosylation (Gerken et al. 2011; de Las Rivas et al. 2019; Stewart et al. 2019; Ballard et al. 2023). In vitro studies of GalNAc-T2 enzyme reactions with IgA1 hinge-region peptide acceptor revealed that the addition of the first three sites of GalNAc is rapid and driven mostly by the catalytic domain (Stewart et al. 2019). The main initial sites utilized by GalNAc-T2 in the hinge-region peptide are Thr225 and Thr236, and the minor sites include Ser228 and Ser230. The selection of the second site is codetermined by the localization of the first site and involves both the catalytic and lectin domains of GalNAc-T2. Thus, the multiple initial sites of glycosylation and follow-up pathways determine the number of glycans added. Notably, these in vitro reactions generate glycopeptides with three to six O-glycans per hinge region, which correspond to the pattern of O-glycosylated forms of serum IgA1 (Takahashi et al. 2010; Takahashi et al. 2012; Ohyama et al. 2020b; Dotz et al. 2021; Ohyama et al. 2022). It is possible that the unique structural features of the IgA1 restrict the maximal number of O-glycans to six per hinge region. Specifically, the hinge-region segment is flanked on the N-terminal side by the globular antigen-binding fragment (Fab) and on the C-terminal side by cysteine residues involved in the formation of disulfide bonds between the component chains of IgA1 and the Fc region. Additionally, the distribution of IgA1 O-glycoforms is impacted by the presence of galactose and sialic-acid modifications on existing GalNAc residues. The addition of GalNAc moieties creates a template for addition of these monosaccharides and also generates a competition for enzyme activity and space in the hinge-region segment (Stewart et al. 2021). Ser230, Thr236, and Thr233 are the sites that commonly contain mixtures of different O-glycans, including those deficient in galactose. The process of clustered O-glycosylation and glycan density of the hinge region is likely influenced by the steric hindrances, concurrent enzyme activities, and other factors such as formation of heteromeric enzyme complexes and pH in the Golgi apparatus (Hassinen et al. 2011). Thus, many factors are likely impacting the biosynthesis of IgA1 O-glycans in IgA1-producing cells and the resultant glycan content and heterogeneity of circulatory IgA1.

Early studies used lectins to assess O-glycosylation of IgA1 in the circulation of patients with IgAN and healthy controls (for review see (Novak et al. 2018)). Two types of lectins were especially useful: jacalin, a lectin specific for core 1 glycans (Kabir 1998; Bourne et al. 2002; Jeyaprakash et al. 2002; Wu et al. 2003; Jeyaprakash et al. 2003), and GalNAc-specific lectins from Helix aspersa and Helix pomatia.  Figure 2 shows structures of jacalin and H. pomatia lectins, and illustrates the features that impart the glycan selectivity. Jacalin (Fig. 2A–D) can bind to GalNAc, galactose, and the disaccharide T antigen, although T antigen has a 36- and 100-fold greater affinity by comparison to GalNAc and galactose, respectively (Sastry et al. 1986). Crystal structures of jacalin show that the core structure is formed by a complex of the α and β chains. This heterodimer then tetramerizes to form the larger oligomer, a hetero-octamer. Structures of jacalin bound to various sugars (Sankaranarayanan et al. 1996; Jeyaprakash et al. 2002; Jeyaprakash et al. 2003) show that each monosaccharide and GalNAc of the T antigen binds to the primary glycan binding side (Fig. 2C). Figure 2A–D shows the T antigen bound to jacalin (Jeyaprakash et al. 2002). The T antigen complex reveals that GalNAc has an additional hydrogen bond mediated by the aspartic acid 125 (D125) to the O4 oxygen of GalNAc and additional interactions are contributed between the O3 oxygen of the glycosidic bond connecting the two sugars and the amine of the amino-terminal glycine in the lectin β chain (Jeyaprakash et al. 2002). Several water-mediated interactions lock the galactose residue into place. Collectively, these interactions explain why binding to the disaccharide is up to two orders of magnitude greater than binding to the monosaccharide.

Fig. 2.

Jacalin and HPA lectin structures and glycan-binding specificity. Panels (A–D) show jacalin, a lectin from Artocarpus integer, with preference for binding to O-linked Gal(β1-3) α-GalNAc. Panel (A) shows a cartoon representation of the hetero-octameric structure of jacalin composed of four chains each of the α- and β-chains [PDB ID: 1M26; (Jeyaprakash et al. 2002)] with Gal(β1-3) α-GalNAc shown in space-filling model. In (A–C), each chain is shown in alternating colors with α- and β-chains shaded in lighter and darker shades, respectively. In (B), a monomeric version of the jacalin α/β heterodimer is shown with glycan in stick model. Panel (C) shows a closeup of the glycan-binding site with amino-acid residues within five angstroms of the glycan shown as sticks and labeled. Panel (D) shows the surface rendering of jacalin with bound glycan. The surface is colored by electrostatic potential. The surface of jacalin provides an extensive pocket to accommodate the disaccharide with high specificity. Panels (E–H) show the lectin from Helix pomatia, which has a preference to bind terminal GalNAc. Panel E shows the hexameric H. pomatia complex [PDB ID: 2CVV; (Sanchez et al. 2006)], where two trimeric assemblies are covalently linked together through disulfide bonds. Each monomer harbors an N-glycan (shown as sticks) and bound GalNAc (shown in space filling model). Panel F shows the monomeric model with the GalNAc shown in stick model. Panel (G) shows a closeup of the GalNAc-binding site with amino-acid residues and water molecules within five angstroms of the glycan shown as sticks and labeled. The GalNAc binding site is generated by two adjacent lectin monomers in the trimeric assembly. Panel H features a surface rendering of the GalNAc-binding pocket colored according to electrostatic potential. For both jacalin and H. pomatia lectins, hydrogen bonding to the O4 atom of the GalNAc moiety (labeled in D and H) provides specificity to accommodate GalNAc. An (*) denotes the point of attachment of GalNAc to serine or threonine residues of a glycopeptide form of Tn antigen. Figures 2A–C and E–G were generated with PyMol (DeLano 2002). Figures 2D and H were generated with ChimeraX (Meng et al. 2023).

The lectins from H. aspersa and H. pomatia have also been studied by X-ray crystallographic methods, revealing their unique structure and specificity of glycan binding (Sanchez et al. 2006; Lescar et al. 2007; Pietrzyk et al. 2015). Figure 2E–H show the structure of the H. pomatia lectin bound to GalNAc. The active lectin is formed by six monomeric subunits assembled as two trimeric structures that are covalently bonded together through disulfide bonds. The binding site for GalNAc is comprised of amino-acid residues donated from two adjacent monomers within the trimeric substructure. Eight hydrogen bonds are formed between the GalNAc and the lectin. Three of the bonds are formed between the O4 oxygen of GalNAc and the three individual amino acids contributed from each monomer that generate the binding pocket [R63 and W83, and D55 (from the adjacent subunit)]. Atom O4 is buried deep within the pocket (Fig. 2D) and is believed to contribute to high affinity for the monosaccharide (Sanchez et al. 2006).

Studies using jacalin (Fig. 2) suggested that serum IgA1 from patients with IgAN may be differently glycosylated compared to that of healthy controls (Andre et al. 1990). However, jacalin properties may vary, depending on the origin of the lectin (Pineau et al. 1991), and additional lectins and alternative analytical techniques were explored (for review see (Novak et al. 2012; Hansen et al. 2021)). Many follow-up studies used GalNAc-specific lectins, monosaccharide compositional analysis, and mass spectrometry (Tomana et al. 1997; Allen 1999; Tomana et al. 1999; Allen et al. 2001; Hiki et al. 2001; Dotz et al. 2021). These studies determined that patients with IgAN have elevated levels of circulatory IgA1 with defects in the hinge-region O-glycans, specifically with some O-glycans missing galactose; i.e. Gd-IgA1 (Mestecky et al. 1993; Allen 1995; Tomana et al. 1997; Odani et al. 2000; Odani et al. 2010; Ohyama et al. 2020b; Ohyama et al. 2022). Furthermore, the glomerular immunodeposits of patients with IgAN are enriched for Gd-IgA1 (Allen et al. 2001; Hiki et al. 2001), thus directly linking aberrant O-glycosylation of IgA1 to the pathogenesis of IgAN.

Table 1 summarizes the findings from mass spectrometric profiling of IgA1 O-glycosylation. Multiple studies determined several features of O-glycans of circulatory IgA1 that are typical for patients with IgAN, including the reduced number of O-glycans in the hinge region and reduced galactosylation and sialylation (see specific references in Table 1). These characteristics and the elevated abundance of glycoforms with three O-glycans correlate with reduced kidney function. One study detailed O-glycans of polymeric molecular form of circulatory IgA1 and another study characterized IgA1 secreted in culture by tonsillar cells. These findings are consistent with the characteristics of glomerular IgA1 that is enriched for galactose-deficient asialo-glycoforms (Allen et al. 2001; Hiki et al. 2001). In terms of glycan attachment sites, the information is much more scarce, with current studies focused on IgA1 isolated from healthy individuals. Future studies will need to focus on the sites of glycan attachment and define the sites that are predominantly galactose deficient in IgA1 glycoforms in the circulating immune complexes. Nevertheless, the emerging information suggests that IgA1 in the circulating immune complexes is enriched for galactose-deficient asialo-glycoforms of polymeric IgA1 that is bound by IgG autoantibodies and associated with complement C3 (Hall et al. 2023).

Table 1.

O-glycosylation of IgA1 in IgA nephropathy: Examples of findings from mass spectrometric analyses.

IgA1 source a	Findings for hinge-region glycopeptides b	References
Glomerular immunodeposits (IgAN)	Reduced number of O-glycans, decreased galactosylation, mostly not sialylated (4:0:0, 4:2:0, 4:2:1, 4:4:0, 6:2:0)b	(Hiki et al. 2001)
Serum, plasma (IgAN vs. HC)	Reduced number of O-glycans, decreased sialylation and galactosylation (e.g. elevated in IgAN: 5:3:3, 5:3:0, 5:2:1, 4:2:2, 3:1:1, 3:1:0)b Decreased sialylation correlated with faster decline in eGFR	(Odani et al. 2000; Hiki et al. 2001; Odani et al. 2010; Dotz et al. 2021; Chen et al. 2022)
Serum (HC)	Sites with variable O-glycosylation, positional isomers: S230, T233, T236. Sites often deficient in galactose: T236 followed by S230, T233, T228, and S232.	(Takahashi et al. 2012; Ohyama et al. 2020b)
Serum, plasma (IgAN vs. HC)	Reduced number of O-glycans, elevated abundance of glycoforms with 3 O-glycans; both features associated with elevated blood pressure and reduced eGFR	(Ohyama et al. 2022)
Plasma (IgAN vs. DC vs. HC)	Reduced number of O-glycans, reduced galactosylation	(Zhang et al. 2022)
Plasma (IgAN vs. HC) Serum (IgAVN vs. IgAN vs. DC vs. HC) Tonsillar cells (IgAN, DC)	Reduced number of O-glycans in polymeric IgA1, associated with a severe IgAN phenotype (crescents) Reduced number of O-glycans, elevated abundance of glycoforms with 3 O-glycans, reduced galactosylation Reduced galactosylation and sialylation	(Yu et al. 2021) (Nakazawa et al. 2019a; Nakazawa et al. 2019b) (Horie et al. 2003)

^aAbbreviations: IgAN, IgA nephropathy; HC, healthy controls; DC, disease controls; IgAVN, IgA vasculitis with nephritis; eGFR, estimated glomerular filtration.

^bx : y : z, number of GalNAc : galactose : sialic acid residues.

GalNAc-specific lectins from H. aspersa and H. pomatia (Fig. 2) have played an important role in Gd-IgA1 assays. After discovering a lectin specific for binding to GalNAc-containing IgA1 (Moore et al. 2007), the first quantitative lectin ELISA for Gd-IgA1 was developed (Moldoveanu et al. 2007). This assay revealed that most patients with IgAN have elevated serum levels of Gd-IgA1 (Moldoveanu et al. 2007; Shimozato et al. 2008). Moreover, serum levels of Gd-IgA1 predict disease progression of IgAN (Zhao et al. 2012; Maixnerová et al. 2019). Serum levels of Gd-IgA1 determined by the lectin ELISA test have also served as a quantitative phenotype for genome-wide association studies. Results of these studies identified two loci, corresponding to genes encoding the galactosyltransferase C1GalT1 and its chaperone C1GalT1C1 (Gale et al. 2017; Kiryluk et al. 2017), that are associated with serum levels of Gd-IgA1. Despite the progress of high-resolution analytical techniques (Odani et al. 2000; Renfrow et al. 2005; Renfrow et al. 2007; Takahashi et al. 2010; Wada et al. 2010; Takahashi et al. 2012; Ruhaak et al. 2013; Ohyama et al. 2020a; Ohyama et al. 2020b; Dotz et al. 2021; Ohyama et al. 2022), there are many advantages that these lectin assays provide, including quantitation, high-throughput processing, and ease of use.

Furthermore, GalNAc-specific lectins enabled detection of Gd-IgA1 in different molecular forms of IgA1 in serum samples from patients with IgAN or Gd-IgA1 secreted by IgA1-producing cell lines (Gomes et al. 2010; Reily et al. 2018). These approaches revealed that circulatory Gd-IgA1 is predominantly bound in immune complexes with IgG (Tomana et al. 1997; Tomana et al. 1999). The discovery of IgG autoantibodies specific for Gd-IgA1 (Suzuki et al. 2009) thus defined IgAN as an autoimmune disease wherein the nephritogenic IgG-IgA1 immune complexes are formed in the circulation (Suzuki et al. 2011).

Natural anti-glycan antibodies

Antibodies to a wide array of glycans and glycoconjugates, including those containing GalNAc, are abundant in human sera (Bovin et al. 2012; Schneider et al. 2015; Sterner et al. 2016; Purohit et al. 2018; Kappler and Hennet 2020; Marglous et al. 2024). These antibodies are generally considered to be part of the natural antibody repertoire, i.e. antibodies present in the sera of animals or humans in the absence of intentional vaccination or infection (Holodick et al. 2017).

Anti-glycan antibodies were first detected in connection with ABO blood-group antigens discovered by Karl Landsteiner (see e.g. (Landsteiner 1945)). The ABO blood groups are defined by the presence of a specific sugar on the surface of red blood cells, added to the trisaccharide called H antigen. If the H antigen is left unmodified, the resulting blood group is O, whereas A antigen has GalNAc added and B antigen has addition of D-galactose [biosynthetic pathways reviewed in (Green 1989)]. As the A and B antigens are immunogenic, individuals naturally develop antibodies against the ABO antigens they do not have. For example, individuals with blood group A will have anti-B antibodies, and individuals with blood group O will have both anti-A and anti-B antibodies.

The origin of natural antibodies is commonly ascribed to innate-like B cell populations, predominantly B1 cells. In mice, these antibodies are derived from germline immunoglobulin variable genes and are devoid of somatic mutations. However, these characteristics do not translate to humans, as the phenotypic markers that define this population in mice are more broadly expressed by human B cells (Griffin et al. 2011; Covens et al. 2013; Weisel et al. 2020). Furthermore, human immunoglobulins exhibiting the specificities of natural antibodies, such as for phosphorylcholine (Fiskesund et al. 2014) and carbohydrates, including αGal (Wang et al. 1995) or gangliosides (Sterner et al. 2017), underwent affinity maturation. Compared to antibodies against protein antigens, natural antibodies are commonly described as polyreactive and of low affinity, typically in the micromolar range. However, many of the characterized anti-carbohydrate antibodies often display exquisite selectivity and may achieve nanomolar affinity (Wang et al. 1995; Haji-Ghassemi et al. 2015; Gildersleeve and Wright 2016; Sterner et al. 2017).

There is substantial evidence that anti-carbohydrate antibodies, also called anti-glycan antibodies, arise from immune responses to the commensal microbiota. In both humans and mice, the emergence of serum anti-carbohydrate antibodies occurs postnatally, and the levels of these antibodies increase with age (Hamanova et al. 2015; Muthana and Gildersleeve 2016; Bello-Gil et al. 2019). Germ-free mice exhibit reduced levels of anti-carbohydrate antibodies which are induced following colonization. Moreover, the reactivity profiles of serum anti-carbohydrate antibodies differ between animals depending on bacteria present in gnotobiotic systems (Khasbiullina et al. 2018). Development of B cells producing antibodies reacting with a single anti-carbohydrate specificity, such as N-acetylglucosamine (GlcNAc), depends on the commensal microbiota (New et al. 2020). This conclusion was based on the analyses of murine B cells producing antibodies recognizing GlcNAc-containing carbohydrates of Streptococcus pyogenes. Compared to specific-pathogen free (SPF) mice, germ-free animals did not have serum GlcNAc-reactive antibodies; this finding was consistent with the significant reduction in the number of GlcNAc-binding B cells. Single-cell-based sequencing of these antibodies revealed a near lack of canonical gene rearrangements that are associated with GlcNAc-reactive B cells in SPF mice. Conversely, colonization of germ-free mice with a conventional microbiota restored both the GlcNAc-reactive B cells and antibody production, including IgA-producing plasma cells in the small intestine (New et al. 2020).

In humans, sera of individuals exhibiting differences in both the carbohydrates targeted and the levels of anti-carbohydrate antibodies likely reflect differences in the composition or timing of colonization of members of the microbiota (Bello-Gil et al. 2017; Luetscher et al. 2020; New et al. 2020). Carbohydrate-reactive antibody specificity appears to be directed to the terminal sugars of complex carbohydrates (Schneider et al. 2015), although glycans that share similar structures are differentially targeted. This is indicative of restricted fine specificities within anti-carbohydrate natural antibodies (Schneider et al. 2015; Luetscher et al. 2020), while precise specificities of some anti-carbohydrate antibodies, such as those recognizing ABO blood-group antigens, are genetically determined. The enzymes that generate these antigens are differentially expressed among individuals thereby defining the ABO blood-group types. As ABO blood-group structures expressed within an individual represent autoantigens, ABO-reactive B cells are differentially censored by negative selection during B-cell development. Thus, ABO blood-group types, ethnicity, and age, contribute to variations in levels of anti-carbohydrate antibodies (Muthana and Gildersleeve 2016; Luetscher et al. 2020). These factors that influence anti-glycan antibody diversity should be considered in studies of disease-specific biomarkers, such as anti-carbohydrate antibodies.

Human anti-carbohydrate antibodies are present as IgM as well as IgG and IgA isotypes with overlapping specificities (Muthana et al. 2015). The mechanisms that direct carbohydrate-reactive B cells to undergo affinity maturation and class switching are not clear. In part due to the association with mucosal responses cited above, anti-carbohydrate antibodies are commonly attributed to extra-follicular responses, and several models have been proposed to explain these observations [reviewed in (Bemark et al. 2024)]. One model postulates a participation of glycan-reactive B cells in the germinal-center reactions. Glycans covalently linked to proteinaceous antigens function as haptens; this property has been exploited as a vaccine strategy designed to induce protective immunity against the capsular polysaccharides of pathogenic bacteria, such as Streptococcus pneumoniae, where conjugate vaccines as well as non-conjugate polysaccharides are used (Davies et al. 2022).

Tumor-associated Tn antigens occur on glycoproteins, such as mucin MUC-1, and the abnormal glycosylation induces humoral immune responses (Tarp et al. 2007; Storr et al. 2008; Pinheiro et al. 2010; Stuchlova Horynova et al. 2013; Taylor-Papadimitriou et al. 2018; Bermejo et al. 2024). Tumor-associated Tn-antigen-induced antibodies are considered protective and their epitopes include GalNAc, GalNAc-glycopeptide, or GalNAc-induced glycopeptide conformations (Martinez-Saez et al. 2015; Sanz-Martinez et al. 2023). In IgAN, the precise epitopes of Gd-IgA1-specific autoantibodies are not known, but these autoantibodies are required for the formation of pathogenic immune complexes.

Microbial glycoconjugates containing GalNAc

GalNAc-containing glycoconjugates occur in some viruses and bacteria. For example, some enveloped viruses can have one or more glycoproteins with O-glycans that can contain GalNAc (Dausset et al. 1959) or sialylated GalNAc (also termed sialyl Tn antigen) (Brockhausen et al. 2022). Some of these glycoconjugates may induce corresponding antibodies. In early studies of HIV-neutralizing antibodies targeting these epitopes, it was proposed that these glycan-containing epitopes may be of clinical relevance for vaccine development (Hansen et al. 1990; Hansen et al. 1991; Hansen et al. 1996). Other enveloped viruses with O-glycans on their envelope glycoproteins include herpes simplex virus, varicella zoster virus, human cytomegalovirus, Epstein–Barr virus, as well as influenza virus and SARS-Co-V2 (Bagdonaite et al. 2016; Olofsson et al. 2016; Mayr et al. 2018; Bagdonaite et al. 2021).

Bacteria can express multiple different GalNAc-containing glycoconjugates. These structures include glycoproteins from both Gram-negative and Gram-positive bacteria, such as those found in Campylobacter jejuni (multiple proteins) (Bernatchez et al. 2005), Streptococcus mutans (Chia et al. 2001), and Streptococcus parasanguinis (fimbrial adhesin) (Stephenson et al. 2002). In Gram-negative bacteria, GalNAc is frequently found in various serotypes of the O-antigens of lipopolysaccharides, including those from Escherichia coli (Stenutz et al. 2006), Salmonella (Liu et al. 2014), and Haemophilus influenzae (Shao et al. 2002), and in the shorter lipooligosaccharides from C. jejuni (Bernatchez et al. 2005) and Neisseria gonorrhoeae (Gotschlich 1994). In Gram-positive bacteria, GalNAc is an important component of the wall teichoic and lipoteichoic acids of S. pneumoniae (Fischer et al. 1993; Seo et al. 2008), the wall teichoic acids (group antigens) of the group C, D, F, and G streptococci (Pritchard et al. 1981), and the wall teichoic acids of some Staphylococcus aureus strains (Mnich et al. 2019). In both Gram-positive and Gram-negative bacteria, GalNAc can be found in capsular polysaccharides [e.g. multiple serotypes of S. pneumoniae (Bentley et al. 2006), group 4 capsules of E. coli (Sande and Whitfield 2021)], other surface polysaccharides [e.g. those of the oral streptococci S. mitis, Streptococcus gordonii, Streptococcus oralis, S. sanguis (Koga et al. 1983; Abeygunawardana et al. 1991; Reddy et al. 1994; Cisar et al. 1997), S. aureus (Lei et al. 2024)], and in exopolysaccharides (e.g. Pel in Pseudomonas aeruginosa (Le Mauff et al. 2022) and possibly numerous other Gram-negative and Gram-positive bacteria (Bundalovic-Torma et al. 2020).

Autoantibodies specific for Gd-IgA1 in IgAN

Serologic studies showed that IgA-deficient individuals develop anti-IgA antibodies (Sennhauser et al. 1988; Koskinen et al. 1995). Furthermore, healthy individuals also produce anti-IgA antibodies and these antibodies bind to the Fab portion of IgA (Jackson et al. 1987b). These antibodies are of IgM and IgG isotypes. Additional studies performed with IgAN patients revealed that their sera have anti-IgA antibodies of IgG isotype and circulating immune complexes consisting of IgA and IgG (Jackson et al. 1987a; Jackson 1988). These results were further extended by the finding that these immune complexes in sera of IgAN patients contain IgA1 with O-glycans deficient in galactose (Tomana et al. 1997), thus linking abnormal glycosylation of IgA1 with formation of Gd-IgA1-IgG immune complexes in IgAN. Follow-up mechanistic studies determined that these immune complexes in sera of IgAN patients consist of polymeric Gd-IgA1 with IgG antibodies that recognize Gd-IgA1 in a glycan-specific context (Tomana et al. 1999), i.e. bind Gd-IgA1 with exposed GalNAc. Smaller quantities of such antibodies were found also in patients with non-IgA proliferative glomerulonephritis and in healthy controls (Tomana et al. 1999). Fab fragment that contained part of the O-glycosylated hinge-region segment with GalNAc was recognized in ELISA by antibodies of IgG isotype and, to a lesser degree, also by antibodies of IgA and IgM isotypes (Tomana et al. 1999). Removal of O-glycans from this Fab fragment by O-glycanase reduced binding of IgG and IgA1 but not of IgM. These findings suggest that abnormal O-glycosylation of IgA1 results in the generation of GalNAc-containing antigenic determinants that are recognized by IgG and IgA1 antibodies. The role and specificity of IgM antibodies in IgAN are not well understood. It is possible that IgM antibodies bind to Gd-IgA1 due to their anti-glycan activity (Matsumoto et al. 2022).

Circulatory IgG autoantibodies specific for Gd-IgA1 are found in an uncomplexed form as monomeric IgG or in immune complexes bound to Gd-IgA1 [for review see (Knoppova et al. 2016; Knoppova et al. 2021)]. These IgG autoantibodies are produced by B cells and can be detected in blood. More information about these antibodies emerged from studies using immortalized B cells derived from peripheral blood of patients with IgAN and healthy controls. These cells were used for single-cell-based cloning and sequencing of the variable segments of the heavy and light chains (VH and VL) of IgG antibodies specific for Gd-IgA1. The expanded cultures of the single-cell clones enabled characterization of the secreted IgG antibodies (Suzuki et al. 2009). Moreover, some of the VH-VL paired sequences were cloned and expressed as recombinant IgG. The IgG autoantibodies from patients with IgAN are specific for Gd-IgA1 with GalNAc residues in the hinge-region O-glycans (Suzuki et al. 2009). In vitro modification of these GalNAc residues with galactose or sialic acid blocks IgG binding. These IgG autoantibodies from IgAN patients exhibit an unusual feature: substitution of Ala₉₇ for Ser₉₇. This residue is located in the framework 3, near the third complementarity determining region 3 (CDR3) of the heavy chain and is a co-determinant for binding to Gd-IgA1 (Suzuki et al. 2009). This alteration originates from somatic mutations of the germline sequence rather than from rare variants of VH genes in IgAN patients (Huang et al. 2016). Notably, IgG derived from healthy controls can also bind Gd-IgA1, albeit weakly, and these clones do not contain the Ser₉₇ substitution found in IgAN-derived IgG autoantibodies. However, when a change of Ala₉₇ to Ser₉₇ was introduced to recombinant IgG derived from a healthy control, this modified IgG bound Gd-IgA1 more strongly (Suzuki et al. 2009), showing 80% binding compared to the recombinant IgG derived from the IgAN patient. Follow-up studies identified additional regions of IgG autoantibodies that are essential for the recognition of Gd-IgA1. These regions include CDR1, CDR3, and framework regions of the heavy chains and framework region 3 of the light chain (Green et al. 2023). A more detailed structure–function characterization of the IgG autoantibodies will define features critical for the formation of pathogenic Gd-IgA1-containing immune complexes.

IgG autoantibodies specific for Gd-IgA1 play an important role in the clinical expression of IgAN. In sera collected at the time of kidney biopsy, serum IgG autoantibody levels correlated with proteinuria (Suzuki et al. 2009), a commonly accepted surrogate outcome biomarker in clinical trials (Thompson et al. 2019). Elevated serum levels of IgG autoantibodies are predictive of disease progression (Berthoux et al. 2012; Maixnerová et al. 2019). Furthermore, serum levels of Gd-IgA1 and the corresponding IgG autoantibodies correlate in IgAN patients, but not in disease or healthy controls (Yanagawa et al. 2014; Placzek et al. 2018).

Immune complexes consisting of IgG autoantibodies and Gd-IgA1 in IgAN

Multiple lines of evidence support the hypothesis that immune complexes consisting of Gd-IgA1 and IgG autoantibodies play a key role in IgAN (Fig. 3) (Suzuki et al. 2011; Novak et al. 2012; Wyatt and Julian 2013; Kiryluk and Novak 2014; Knoppova et al. 2016; Lai et al. 2016). This conclusion is further supported by the finding that IgG is in the immunodeposits in all biopsies of patients with IgAN, even those in kidney biopsies without detection of IgG by routine immunofluorescence (Rizk et al. 2019; Rizk et al. 2023). Furthermore, this IgG, but not IgG extracted from immunodeposits of patients with membranous nephropathy or lupus nephritis, is specific for Gd-IgA1 (Rizk et al. 2019).

Fig. 3.

Pathogenesis model of IgA nephropathy. Pathogenesis of the autoimmune disease IgA nephropathy (IgAN) has been described as a multi-step process of IgA1 and IgG antibody generation, immune-complex formation, and deposition in the glomeruli, leading to kidney injury (Suzuki et al. 2011). Polymeric IgA1 (dimeric in this figure) harboring O-glycans in the hinge region, with an incomplete repertoire of galactose moieties [galactose (Gal)-deficient IgA1 (Gd-IgA1)], is found elevated in the circulation of most IgAN patients and some of these glycoforms are recognized as an autoantigen. IgG autoantibodies with a specificity for the glycosylated hinge region of Gd-IgA1 bind Gd-IgA1 (antigen-binding sites of IgG marked by *), and pathogenic immune complexes are formed. Larger complexes are formed as additional proteins are added, including complement C3. C3 or activated C3b covalently binds to IgA1 or IgG through the thioester bond found in the thioester-containing domain. Some of these immune complexes pass through the fenestration of glomerular endothelial cells, enter glomerular mesangial space, and activate mesangial cells. A cascade of effects is associated with mesangial-cell activation, including cellular proliferation and overproduction of extracellular matrix (ECM), cytokines, chemokines, some of which and can activate podocytes. These events can lead to glomerular injury. Glycan content at up to six Ser/Thr residues (225, 228, 230, 232, 233, 236; shaded red) and the IgA1 hinge-region conformation distinguishes whether the IgA1 molecule is viewed as normal IgA1 or is identified as an autoantigen. Space filling models of IgA1, IgG, and complement C3 were generated from coordinates published in (Person et al. 2022), (PDB ID: 1IGT; (Harris et al. 1997)), and (PDB ID: 2I07; (Janssen et al. 2006)), respectively. Component chains of Gd-IgA1 and IgG are shown in different colors. Complement C3b is colored by domain as in (Rajasekaran et al. 2023).

Experimental proof of the pathogenic potential of IgG autoantibodies in vivo was obtained by using a passive mouse model of IgAN. In this model, immunodeficient mice are injected with immune complexes formed in vitro from human Gd-IgA1 and human IgG autoantibodies. These i.v.-injected immune complexes induce mesangioproliferative glomerular injury, whereas individual components alone (i.e. Gd-IgA1 or IgG) do not (Moldoveanu et al. 2021). Notably, this animal model of IgAN can be used for pre-clinical testing and assessment of mechanisms of drugs for IgAN treatments, such as those that are kidney-targeting (Reily et al. 2024).

Analyses of serum samples from patients with IgAN indicated that the amounts of IgA-IgG immune complexes are elevated at the time of disease activity as compared to those from the time of disease quiescence (Coppo et al. 1982; Novak et al. 2011b). Using a model of cultured primary human mesangial cells enables additional testing, such as assessment of the influence of various factors on cellular proliferation and production of cytokines, chemokines, and components of extracellular matrix (Chen et al. 1994; Novak et al. 2002; Novak et al. 2005; Novak et al. 2008; Novak et al. 2011a; Novak et al. 2011b; Ebefors et al. 2016; Liu et al. 2017; Bian et al. 2021). Based on the use of size-exclusion chromatography for isolation of immune complexes, it was determined that large-molecular-mass IgA-IgG complexes (Mr >700 kDa) stimulate cellular proliferation of quiescent primary human mesangial cells in culture (Hall et al. 2023). These immune complexes contain IgA1, IgG, and complement C3. IgA1 in these complexes is enriched for galactose-deficient and minimally sialylated O-glycoforms. Stimulatory activity of these immune complexes is removed by depleting the serum of total IgA1 by jacalin affinity chromatography. This process removes IgA1-containing immune complexes with IgG, and C3, suggesting that these components are associated in the stimulatory IgA1-containing immune complexes. Furthermore, C3 in these complexes is covalently associated with IgG as well as IgA1, presumably through the thioester bond of C3 (Fig. 3). Molecular forms of C3 in these complexes include C3 as well as its activated and inactivated forms, i.e. C3b and iC3b, respectively (Hall et al. 2023).

These results together indicate that the pathogenic immune complexes in IgAN contain Gd-IgA1 that is bound by IgG autoantibodies, and that complement C3 is associated with these complexes, with at least some C3 covalently bound to the immunoglobulins. Future studies will determine the function and stoichiometry of individual components and the mechanism by which these complexes activate mesangial cells.

Emerging and future treatment approaches for IgAN

The basic approach in IgAN therapy has been based on the control of blood pressure. With the improved understanding of the pathogenetic pathways in IgAN, new opportunities for treatment emerged and current therapies and clinical trials have been summarized in a recent review (Lim et al. 2024). These approaches include gut-released budesonide to reduce gut-produced IgA, B-cell targeting therapies to reduce production of IgA and IgG, and thus Gd-IgA1 and IgG autoantibodies, complement-targeting reagents to reduce complement activation, and kidney-specific treatments to mitigate responses of resident glomerular cells. While all these approaches have a good rationale, the issue is which of the treatments can sufficiently reduce the eGFR slope to prevent the progressive loss of kidney function. With the new and emerging knowledge on complement C3, IgA1 O-glycoforms, and IgG autoantibodies that are all needed to form nephritogenic complexes, it is hoped that additional potential targets emerge that may enable removal of Gd-IgA1-containing immune complexes from the circulation or mitigate complement C3 attachment to the complexes. Other approaches that can prevent formation of nephritogenic immune complexes may utilize blockade or removal of autoantibodies or removal of the cells that produce them. A better understanding of the pathophysiology of this complex disease may inform development of such new approaches in the future.

Concluding remarks

The last two decades witnessed significant advances in the understanding of the pathogenesis of IgAN. IgAN is now recognized as an autoimmune disease wherein Gd-IgA1 is the main autoantigen bound by IgG autoantibodies. These Gd-IgA1-IgG complexes formed in the circulation can then associate with other proteins, such as complement C3. Some of these complexes that are not cleared from the blood may enter the mesangial space in the glomeruli, activate the resident mesangial cells, and induce glomerular injury (Fig. 3). This hypothesis provides a framework for developing preclinical tests, designing clinical trials, and selecting biomarkers (Lim et al. 2024). However, despite the progress, a recent study revealed that most IgAN patients progress to kidney failure within 10–15 years from diagnosis and that only few patients may avoid kidney failure in their lifetime (Pitcher et al. 2023). It is therefore important to determine the key biochemical, immunological, and genetic factors involved in pathogenesis of IgAN to develop a knowledge base for new disease-specific therapeutic approaches. Defining molecular mechanisms of aberrant IgA1 O-glycosylation leading to the production of autoantigen and the recognition by IgG autoantibodies are among the key issues. New mechanistic studies may provide insights to inform development of therapeutic approaches that target abnormal IgA1 O-glycosylation, production of IgG autoantibodies, or immune-complex formation and catabolism (Knoppova et al. 2021). Furthermore, determining the role of C3 in the pathogenicity of IgA1 immune complexes will inform the utility of complement modulation in IgAN (Rajasekaran et al. 2023).

Author contributions

Jan Novak (Conceptualization [equal], Funding acquisition [equal], Resources [lead], Writing—original draft [lead], Writing—review & editing [lead]), R. Glenn King (Conceptualization [equal], Writing—original draft [equal], Writing—review & editing [equal]), Janet Yother (Conceptualization [equal], Writing—original draft [equal], Writing—review & editing [equal]), Matthew B. Renfrow (Conceptualization [equal], Writing—original draft [equal], Writing—review & editing [equal]), and Todd J. Green (Conceptualization [equal], Funding acquisition [equal], Visualization [equal], Writing—original draft [equal], Writing—review & editing [equal])

Funding

This work was supported in part by grants from the National Institutes of Health DK078244, AI149431, DK082753, and DK105124; a gift from IGA Nephropathy Foundation; and by UAB Research Acceleration Funds. The authors apologize to their colleagues in the field whose work is not adequately discussed or cited in this article due to space limitations.

 

Conflict of interest statement. JN is a co-founder and co-owner of and consultant for Reliant Glycosciences, LLC. MBR is a co-founder and co-owner of Reliant Glycosciences, LLC. JN and MBR are co-inventors on US patent application 14/318,082 (assigned to the UAB Research Foundation [UABRF] and licensed by UABRF to Reliant Glycosciences, LLC).