IgA Nephropathy: Pleiotropic impact of Epstein-Barr virus infection on immunopathogenesis and racial incidence of the disease

Abstract

IgA nephropathy (IgAN) is an autoimmune disease in which poorly galactosylated IgA1 is the antigen recognized by naturally occurring anti-glycan antibodies, leading to formation of nephritogenic circulating immune complexes. Incidence of IgAN displays geographical and racial disparity: common in Europe, North America, Australia, and east Asia, uncommon in African Americans, many Asian and South American countries, Australian Aborigines, and rare in central Africa. In analyses of sera and cells from White IgAN patients, healthy controls, and African Americans, IgAN patients exhibited substantial enrichment for IgA-expressing B cells infected with Epstein-Barr virus (EBV), leading to enhanced production of poorly galactosylated IgA1. Disparities in incidence of IgAN may reflect a previously disregarded difference in the maturation of the IgA system as related to the timing of EBV infection. Compared with populations with higher incidences of IgAN, African Americans, African Blacks, and Australian Aborigines are more frequently infected with EBV during the first 1-2 years of life at the time of naturally occurring IgA deficiency when IgA cells are less numerous than in late childhood or adolescence. Therefore, in very young children EBV enters "non-IgA" cells. Ensuing immune responses prevent infection of IgA B cells during later exposure to EBV at older ages. Our data implicate EBV-infected cells as the source of poorly galactosylated IgA1 in circulating immune complexes and glomerular deposits in patients with IgAN. Thus, temporal differences in EBV primo-infection as related to naturally delayed maturation of the IgA system may contribute to geographic and racial variations in incidence of IgAN.

Keywords: Epstein-Barr virus; IgA nephropathy; IgA system maturation; age of infection; galactose-deficient IgA1; virus spread.

Figure 1

Comparison of systemic and mucosal IgA compartments. Systemic and mucosal compartments differ in proportions of the IgA subclasses, amount of IgA produced daily, dynamics of IgA production relative to normal adult values, the proportion of individual Ig isotype-positive cells, and tissues with IgA-secreting plasma cells. IgA1 heavy chain (α1) has, in contrast to IgA2 (α2), a unique hinge region with an additional 13 amino acids that include Thr and Ser residues which may be glycosylated. Red-highlighted amino acids may be O-glycosylated (, –70).

Figure 2

Differences in glycosylation pathways in EBV-infected or non-infected IgA1-producing cells. In the healthy conditions, IgA1-producing plasma cells generate IgA1 with hinge-region O-glycans; the prevailing form consists of the N-acetylgalactosamine (GalNAc) with β1,3-linked galactose (Gal) forming the Core 1 structure (also called T antigen) and its mono- and di-sialylated forms. O-glycosylation is a highly complex process involving about 50 glycosyltransferases and occurs in the Golgi complex. O-glycosylation is initiated by one of several N-acetylgalactosaminyltransferases (GalNAc-Ts) which catalyze the transfer of GalNAc to the Ser or Thr residues (S/T), leading to formation of Tn antigen. GalNAc-T2 is probably an essential enzyme responsible for galactosylation of IgA1; however, other GalNAc transferases are also expressed in B cells and could participate in this process (, , –73). Formation of Tn antigen is followed by the addition of Gal catalyzed by only one known Core1 β1,3-galactosyltransferase 1 (C1GalT1) and its chaperon Cosmc. Core 1 can be expanded with sialic acid(s) attached, by several sialyltransferases to Gal, GalNAc, or both. The process is catalyzed by Galβ1,3GalNAc α2,3-sialyltransferase (ST3Gal) (72) or/and a α2,6-sialyltransferase (ST6GalNAc-I or ST6GalNAc-II), respectively (74). Replicating EBV-infected IgA1⁺ cells can produce EBV gp350 and IgA1. C1GalT1 participates in the parallel formation of Core1 on gp350 and IgA1, leading potentially to a relative C1GalT1 deficiency and generation of O-glycans with terminal GalNAc with or without α2,6 attached sialic acid. Preterminal sialylation of Tn antigen increases formation of Gd-IgA1 (75). .

Figure 3

T cell-dependent and T cell-independent IgA isotype switching. Surface IgM-positive (sIgM⁺) B cells are induced to undergo isotype switch by i) T cell-dependent manner in Peyer’s patches and other mucosa-associated lymphoid follicles (left panel) or by ii) T cell-independent manner in the vicinity of mucosal epithelial surfaces in diffuse lymphoid tissues of lamina propria mucosae (right panel) (–134). In Peyer’s patches, B cells differentiation depends only partially on the cognate help signals from Tfh exposing CD40L and secreting TGF-β and IL-21. In concert, FDCs secrete BAFF, APRIL, RA, and TGF-β in response to DAMP stimulation. Furthermore, FDCs present native antigens to B cells to support crosslinking by BCR. Peyer’s patches contain also a TipDCs which render B cells more sensitive to TGF-β due to NO-induced enhancement in the expression of TGF-β receptor. Peyer’s patches contain also pDC secreting BAFF and APRIL upon stimulation by type I interferon from the ISC (128, 134). Diffuse lymphoid tissues of the intestinal lamina propria contribute mostly to the T cell-independent Ig switching in B1 cell subset. Besides several subsets of MC and DC, local PCs are involved in this process. TipDC are typical for lamina propria and they act similarly to their counterpart in Peyer’s patches. In addition, DC expressing TLR5 member of the DAMP family are stimulated by flagellin to secrete RA and IL-6. DAMP-activated DC co-stimulated by TSLP from epithelial cell produce BAFF and APRIL. Some DC extend their dendrites through the epithelial cell junction or across the M cells into intestinal lumen to sample antigens for recycling and presentation in unprocessed form to B cells, as the T cell-independent antigens. PC maturation and survival could be supported by mast cells producing IL-4, IL-5, IL-6, and BAFF. Furthermore, Eo could contribute to PC survival by secretion of IL-6. Finally, local IgA-secreting PC were identified as a producers of TNF and iNOS (128, 129, 135, 136). APRIL, a proliferation-inducing ligand; BAFF, B cell activating factor; BCR, B cell receptor; DAMP, danger-associated molecular pattern; DC, dendritic cells; Eo, eosinophils; FDC, follicular dendritic cells; iNOS, TNF-inducible nitric oxide synthase; ISC, intestinal stromal cells; MC, mast cells; NO, nitric oxide; PC, plasma cells; pDC, plasmacytoid dendritic cells; RA, retinoic acid; Tfh, follicular T helper cells; T_FR, follicular regulatory T cells; TipDC, TNF- and iNOS-producing DC; TSLP, thymic stromal lymphopoietin.

Figure 4

The impact of EBV infection on sIgA1⁺ B cells. After the initial mucosal infection, the virus remains in resident memory B cells; upon activation, Igs and EBV are produced in plasma cells (–143). EBV-infected plasma cells secrete J-chain-containing Gd-pIgA1 with preferentially λ light chains. Such cells also display homing receptors involved in the selective population of the upper respiratory mucosa. In addition, vIL-10 is likely to support the differentiation of cells into IgA producers and probably suppresses the cytotoxic activity of CTLs (141, 143, 144). CTL, cytotoxic T lymphocytes; Gd-IgA1, galactose-deficient IgA1; J, joining; L light; URT, upper respiratory tract.

Figure 5

Waldeyer’s ring. Waldeyer’s ring is comprised of the nasopharyngeal tonsils (adenoids) attached to the roof of the pharynx, the tubal tonsils (adenoids) located at the pharyngeal aperture of the Eustachian tubes, the palatine tonsils in the oropharynx, and the lingual tonsils on the posterior third of the tongue. Tonsils are lymphoreticular and lymphoepithelial organs. Tonsillar epithelium invaginates and lines the tonsillar crypts enhancing the surface for direct contact with exogenous antigens to a surface of 350 cm², predominantly in the palatine tonsils (162).