Loss of self-tolerance is fundamental to autoimmunity. A clear autoimmune component in IgAN is the development of autoantibodies directed against the undergalactosylated IgA, leading to formation of immune complexes that deposit on mesangial cell triggering an autoinflammatory cascade. The fine details of the genetic and molecular alterations that lead to this autoimmune cascade remain unknown. However the GWASs have revealed a significant number of shared loci with other autoimmune conditions. A deeper analysis of this pathway could shed light on the shared autoimmune alteration and clinical features between these heterogeneous conditions. GWASs have indicated that HLA–DQA1, HLA–DQB1, HLA–DRB1, and HLA–DP are associated with IgAN and DQA1*0101 and DQB1*0301 emerged as risk alleles and DQA1*0102 and DQB1*0201 [22, 24] as protective alleles after imputation analysis. A significant number of MHC loci identified in IgAN are shared with other autoimmune diseases such as rheumatoid arthritis [193], systemic sclerosis [194], alopecia areata [195], Graves’ disease [196], type I diabetes [193], and celiac disease [197, 198]. Interestingly, a discordant effect for these loci has been found in SLE [199], multiple sclerosis [200], and ulcerative colitis [201], suggesting that some MHC alleles may enhance recognition of autoantigen for a specific disease but have reduced affinity for the autoantigen for another autoimmune disease and ultimately result in protection.

As mentioned above, a common deletion in CFHR1 and CFHR3 genes has been associated to IgAN [22]. The same deletion identified is protective in age-related macular degeneration [22, 24, 164, 168, 202] but increases risk in SLE [169] and atypical hemolytic uremic syndrome (aHUS). In aHUS, the CFHR3,1-del has been associated to the development of inactivating anti-FH antibodies [170] that would be the final effectors of the increased risk. Similar to MHC, CFHR loci, the IgAN risk alleles at ITGAM, and ITGAX loci have an opposite effect in SLE [203].

Significant overlap with pathways for IBD was also detected. CARD9 is a pro-inflammatory molecule which is responsible for both innate and adaptive immune responses [204]. The IgAN risk allele at the CARD9 locus is direction consistent with associations reported on the risk of ulcerative colitis and Crohn’s disease [179, 180, 205–207]. α-Defensins 5 and 6 (DEFA5 and DEFA6) are constitutively produced by the intestinal Paneth cells of the intestine [208]. Deficiencies in α-defensins 5 and 6 have been previously associated with Crohn’s disease [209, 210]. Finally, the LIF/OSM locus is also associated with IBD. Additional, suggestive associations with IBD loci were also reported and the pathway analysis revealed significant enrichment for networks involved intestinal IgA productions [24]. These shared associations between IgAN and IBD disease highlight common pathways in the pathophysiology of the inflammation of mucosal barrier and may explain co-occurrence of these two disorders [211].

TNFSF13 has not been previously related to specific autoimmune disease; however, mutations in the TNFSF13 receptor (TACI) produce IgA deficiency or combined variable immunodeficiency, with increased propensity to mucosal infections [212]. TNFSF13 encodes APRIL, a powerful B cell-stimulating cytokine that is induced by intestinal bacteria and promotes CD40-independent IgA class switching [213]. APRIL levels are elevated in patients with IgAN [174] and the IgAN risk allele is associated with increased IgA levels [23].

The recent GWAS of IgAN demonstrated a significant association of the LIF/OSM and the TNFSF13 loci with IgA levels (additive effects). In addition, the 15-SNP genetic risk score is associated with the age of presentation of disease; there was a 20-year difference in the age of presentation between individuals with the most and least number of risk alleles [24]. While the effect size is too modest to be clinically useful, these data clearly demonstrate that genetic factors influence clinically important variables. Identification of novel loci combined with careful genotype–phenotype correlations may reveal association of individual SNPs or the genetic risk score with important variable such as proteinuria, progression, or histopathological severity. These data also suggest that detection of more common variants with GWAS or identification of rare alleles with large effect will delineate stronger clinical correlations.

Another approach is to improve phenotypic characterization. The role of undergalactosylated IgA in the pathogenesis is well established and has been the object of more than 30 years of investigation [214]. Combined measurement of galactose-deficient IgA1 and autoantibody levels may provide more refined tools for characterization of patients and relatives. The parameters are highly associated with development of disease and may enhance the resolution of genetic studies.

In addition, examination of disease mechanisms requires a strong functional assays and animal models. Cellular models to date include the production of patient-derived IgA1-producing cell lines that recapitulate defects in IgA production and allow examination of molecular defects leading to aberrant IgA1 O-glycosylation. In addition, patient-derived lymphocytes producing autoantibodies targeting undergalactosylated IgA1 have also been developed, allowing examination of pathways leading to breach in tolerance. Challenges in developing animal models include the differences in the composition and structure of IgA between rodents and humans. However, novel animal models of disease recapitulate many clinical aspects of disease and will likely prove useful for studying pathogenesis [174, 215, 216]. The recent report describing the development of IgAN in mice overexpressing BAFF [174, 217] shows convergence with detection of association of TNFSF13 with IgAN in recent GWASs [23], indicating that these mice would be suitable for studying IgAN pathogenesis.

While GWAS represents an important advance in the evaluation of genetic component of complex diseases, it can only detect relatively frequent variants with small effect. Hence in the majority of cases, the GWAS loci explain only a fraction of the overall heritability. While fine-mapping studies may uncover additional common and rare risk alleles that together explain a larger fraction of the disease variance, there is additional “missing heritability” that may be attributable to either genetic or nongenetic factors [218–220]. In particular, the missing heritability may be due to the presence of low-frequency variants with large effect that are not detectable by GWAS methodology. For example, recent studies have also shown that CNVs with large effect contribute to many neuropsychiatric, developmental, and immune-mediated phenotypes [221–224]. In addition, next-generation sequencing (NGS) has revolutionized genomics, achieving superlative speed and accuracy in sequencing [225, 226]. Large-scale sequencing studies have now identified rare single nucleotide variants predisposing to many complex phenotypes such as autism, dyslipidemia, idiopathic pulmonary fibrosis, or ALS [227–231]. These data suggest that CNV analysis, whole genome analysis, or exome analysis of IgAN patients may similarly uncover rare genetic variants predisposing to familial and sporadic disease. Moreover these genetic studies can be combined in clinical trials of IgAN, to detect interaction of genetic factors with progression or response to treatment.