Monomeric human IgA1 has two heavy chains and two light chains connected by disulfidic bridges [37]. Each heavy chain has three constant (C_α) domains and one variable (V) domain; each light chain has one C and one V domain (Fig. 4.2a). Unlike IgA2, the other IgA subclass in humans, IgA1 has a unique hinge region between C_α1 and C_α2 that consists of two octapeptide repeats (TPPTPSPS) and is the site of attachment of O-glycans [38]. On circulatory IgA1, usually three to six of the nine potential O-glycosylation sites are glycosylated (Fig. 4.2a) [39–42]. There are also two N-glycosylation sites on each heavy chain containing complex glycans [43–47]. Monomers of IgA1 can be covalently associated during their production in plasma cells with a ~15-kDa J-chain [48–50] and thus the immunoglobulin can be present as dimers and higher oligomers, termed polymeric IgA1. In the circulation, most of IgA1 is monomeric [51, 52].

O-glycans of serum IgA1 of healthy individuals are usually core 1 glycans, i.e., disaccharides consisting of Ser/Thr-linked N-acetylgalactosamine with a β1,3-linked galactose (Fig. 4.2b). The disaccharide may be sialylated on either sugar or on both sugars [45, 53]. Sialic acid (in humans, N-acetylneuraminic acid) is attached to N-acetylgalactosamine of IgA1 by an alpha2,6-linkage and to galactose by an alpha2,3-linkage. Lectins (proteins that bind to particular carbohydrates) serve as useful tools to determine presence of terminal N-acetylgalactosamine or the disaccharide consisting of N-acetylgalactosamine with β1,3-linked galactose (Fig. 4.2b), but careful assessment of lectin specificities should always be performed using well-characterized IgA proteins [40–42, 54–57].

O-glycosylation of IgA1 is a stepwise process beginning with attachment of N-acetylgalactosamine to Ser/Thr residues of the hinge region; it is mediated by N-acetylgalactosaminyl-transferase 2 (GalNAc-T2) [58, 59], but other GalNAc-Ts, including GalNAc-T1, GalNAc-T11, and GalNAc-T14, may contribute to the process [60, 61]. The O-glycan chain can be then extended by attachment of galactose to the N-acetylgalactosamine residue (Fig. 4.3); this process is mediated by core 1 beta1,3-galactosyltransferase (C1GalT1) [62]. The stability of this enzyme during biosynthesis depends on interaction with its chaperone (Cosmc) for assistance with protein folding. Without this association with Cosmc, the C1GalT1 nascent protein is rapidly degraded [63–65]. The core 1 structures of IgA1 may be further modified by sialyltransferases that attach sialic acid to the galactose residues (mediated by an ST3Gal enzyme) and/or N-acetylgalactosamine residues (mediated by ST6GalNAc-II, as the usual ST6GalNAc-I is not expressed in IgA-producing cells) [61, 66, 67].

Serum IgA1 of healthy individuals had been thought to contain few or no galactose-deficient O-glycans [45], but it has been shown that galactose-deficient O-glycans may be present at some sites in the IgA1 hinge region [40]. Glycosylation studies have revealed aberrancies in the O-glycans of IgA1 in the circulation of patients with IgAN. Specifically, most patients with IgAN have elevated serum levels of IgA1 with some O-glycans deficient in galactose [68–73]. Thus, a fraction of circulatory IgA1 molecules has some hinge-region O-glycans without galactose, i.e., consisting of terminal N-acetylgalactosamine or sialylated N-acetylgalactosamine (Fig. 4.2). This galactosylation defect appears to be specific for IgA1, as other O-glycosylated serum proteins such as C1 inhibitor and IgD do not exhibit galactose deficiency [40, 41, 56, 57, 74].

Development of immortalized IgA1-secreting cells derived from cells in the circulation of patients with IgAN and healthy and disease controls provided a new tool for studies of normal and aberrant O-glycosylation of IgA1 [103]. IgA1 secreted by the cells from patients with IgAN had higher degrees of galactose deficiency of the O-glycans than that from cells of healthy controls. Notably, the relative degree of galactose deficiency of the IgA1 secreted by IgA1-producing cells correlates with that of serum IgA1 from the corresponding donor [103]. Thus, these cells served as an excellent model system to study enzymatic pathways and the actual O-glycosylation of secreted IgA1. Initial studies showed that the polymeric IgA1 form is the most affected by galactose deficiency. Furthermore, this phenotype is related to decreased expression and activity of C1GalT1 and elevated expression and activity of ST6GalNAc-II [103]. Moreover, the expression of C1GalT1-specific chaperone Cosmc [63, 64], that is necessary for stability of the nascent C1GalT1 protein, is decreased [103]. More recently, the roles of C1GalT1 and ST6GalNAc-II in production of Gd-IgA1 have been confirmed, using siRNA knockdown and biochemical approaches [67, 117, 118].

IgA1 secreted by our established cell lines was isolated and its O-glycosylation characterized by using high-resolution mass spectrometry. The IgA1 secreted by cells from patients with IgAN has more galactose-deficient O-glycans, consisting of terminal or sialylated N-acetylgalactosamine, compared to IgA1 from the cells from healthy controls [119]. Moreover, IgA1 from patients with IgAN also had more O-glycans per heavy chain. Together, these data revealed another type of glycosylation abnormality, an elevated number of O-glycans on IgA1 from patients with IgAN. This finding implicated a central role for an O-glycosylation-initiating enzyme, a GalNAc-T, in production of Gd-IgA1.

Several GalNAc-Ts are abundantly expressed in IgA1-producing cells [61, 120], including GalNAc-T2 [103]. Of this family of enzymes, only GalNAc-T14, the closest structural relative of GalNAc-T2, was expressed at several-fold greater levels in the cells from patients with IgAN compared to cells from healthy controls [120]. Conversely, the expression of GalNAc-T2 and other GalNAc-Ts did not differ between patients and healthy controls [103]. These findings thus indicate that overexpression of GalNAc-T14 may contribute to enhanced O-glycosylation of IgA1 in IgAN [9, 61].

Notably, the dysregulated expression and activities of several key glycosyltransferases are further enhanced by some cytokines, including IL-6 [117]. The signaling process through the IL-6 receptor/gp130 complex is mediated by STAT3 [121, 122] and increases activity of ST6GalNAc-II and decreases activity of C1GalT1 (Fig. 4.4) [117]. This process leads to greater production of Gd-IgA1 [117]. Another type of signal involved in cellular differentiation and survival that may be relevant to Gd-IgA1-producing cells includes B-cell activating factor (BAFF) [122]. Notably, overexpression of human BAFF in mice leads to overproduction of IgA and development of IgA glomerular deposits [123].

These observations may together explain why the clinical onset or a flare in the activity of IgAN often coincides with an active mucosal infection of the upper respiratory tract and/or digestive system. It can be speculated that such environmental factors exert their activity through cytokines (e.g., IL-6) and growth factors (e.g., BAFF, APRIL) (Fig. 4.4) to upregulate cellular production of Gd-IgA1 and/or cell survival in susceptible individuals [9, 94, 122].

Multiple lines of evidence allow us to characterize IgAN as an autoimmune disease in which Gd-IgA1, produced in increased amounts and leading to increased levels in the blood, is bound by unique autoantibodies. This sequence of events leads to formation of pathogenic circulating Gd-IgA1-containing immune complexes. Some of these complexes deposit in the kidney and induce a mesangioproliferative glomerular injury [124]. Elevated serum levels of Gd-IgA1 are genetically codetermined and can be further increased by some cytokines. Moreover, serum levels of Gd-IgA1 may be predictive of disease progression. Gd-IgA1 is produced by IgA1-secreting cells of patients with IgAN due to dysregulation of several key glycosylation enzymes. A combination of approaches with use of IgA1-producing cell lines and high-resolution mass spectrometry will likely provide a better understanding at cellular and molecular levels of the heterogeneity of Gd-IgA1 in patients with IgAN. We believe that through a detailed understanding of the disease processes, biomarkers specific for IgAN can be identified and developed into clinical assays to aid in the diagnosis, assessment of prognosis, and monitoring of the disease progression. Moreover, characterization of specific pathogenetic pathways, such as those involved in Gd-IgA1 production, will identify targets for future disease-specific therapies.