In this study, we identified multiple genes showing significantly pleiotropic association with IOP through a SMR approach. Our results not only confirmed findings from previous studies on the association of some genes with IOP, but also revealed some novels genes underlying the genetic mechanisms of IOP regulation and/or POAG.
In our study, except for ABO and SGTB, most genes showing significantly pleiotropic association with IOP were in or adjacent to 17q21.31 (chr17: 40900001-44900000, GRCh37/hg19), one of the most structurally complex and evolutionarily dynamic regions of the genome[33–35]. This region contained a ~970 kb inversion of the microtubule-associated protein tau (MAPT) locus in populations with European ancestry[36]. MAPT was associated with both ganglion cell inner plexiform layer (GCIPL) and retinal nerve fiber layer (RNFL) in a large multi-ethnic meta-analysis of GWAS, indicating that it might impact glaucoma pathogenesis through modulation of retinal thickness[37]. The MAPT locus has two divergent haplotypes, H1 (direct orientation) and H2 (inverted orientation), which have distinct functional impacts[33]. Although no probes tagging MAPT were included in the SMR analysis using GTEx and CAGE eQTL data, some of the genes identified in our SMR analysis were reported to be associated with MAPT haplotypes. For example, the H1 haplotype of MAPT was associated with an increased expression of LRRC37A4, the top hit gene in the SMR analysis using CAGE eQTL data[36]. Moreover, several other identified genes in the 17q21.31 region were either reported to be associated with IOP or act collectively in influencing IOP or associated traits. For example, a GWAS of 68,423 participants from the UK Biobank cohort identified 139 genetic loci associated with macular thickness (MT), including genetic variants in KANSL1, LRCC37A4P-MAPK8IP1P2 and NSF[38]. In addition, KANSL1-AS1 (identified in GTEx), LRRC37A2 (identified in CAGE) and OR7E14P were found to form a regulatory cluster in influencing both IOP and MT[39]. Together, these findings indicate that 17q21.31 might be an essential region for regulating IOP. Further investigation is needed to elucidate the exact functions of this region and examine its role in influencing IOP and the pathogenesis of glaucoma.
We found that ABO (Alpha 1-3-N-Acetylgalactosaminyltransferase and Alpha 1-3-Galactosyltransferase) showed significantly pleiotropic association with IOP using GTEx eQTL data. ABO, located on chromosome 9q34.2, encodes proteins related to the first discovered blood group system[40]. It has seven exons and six introns[41], and variation in ABO forms the basis of the ABO blood group[42]. Genetic variants in ABO have been found to be associated with diabetes, thromboembolism, myocardial infraction, atherosclerosis, and stroke[43, 44]. A previous meta-analysis, performed by the International Glaucoma Genetics Consortium (IGGC) on 18 population cohorts comprising 35,296 multi-ancestry participants, revealed a novel genetic polymorphism in ABO (rs8176693) in association with IOP[45]. Later, a large-scale meta-analysis found that the genetic polymorphism rs8176741 in ABO showed significant association with not only IOP, but also vertical cup-disc ratio (VCDR) and cup area[46]. Despite these encouraging findings, the exact mechanisms underlying the observed association between genetic variants in ABO and IOP remains to be elucidated. More studies are needed to explore the exact functions of this gene and examine its role in influencing IOP.
In our study, some genes, such as AFAP1 (Actin Filament Associated Protein 1), showed nominal pleiotropically/potentially causal association with IOP, although the association was not significant after correction for multiple comparison using FDR (FDR Q-value =0.051 and 0.098 for GTEx and CAGE eQTL data, respectively). AFAP1, located on 4p16.1, encodes a protein that binds to actin filaments and allows their crosslinking[47, 48]. Actin cytoskeleton-modulating signals have been shown to be involved in the regulation of aqueous outflow and IOP[49, 50]. Two SNPs in AFAP1 (rs4619890 and rs11732100) were reported to be associated with POAG in GWAS studies[51, 52]. In European-ancestry populations, the two SNPs were moderately associated with another SNP in AFAP1 (rs28795989) which was found to be associated with IOP in a large multi-ethnic GWAS study[22]. Another GWAS using data from UK Biobank found that rs6816389 in AFAP1 was associated with IOP in European-ancestry participants[53]. Moreover, the expression of AFAP1 was detected in the trabecular meshwork, retina (including RGCs) and optic nerve of normal human eye and glaucomatous eye. Together, existing evidence implies potential involvement of AFAP1 in the pathogenesis of glaucoma[52].
There are limitations in our study. The eQTL data were based on limited sample size and as such we may have insufficient statistical power. Moreover, the eQTL data have limited eligible probes. Together, it is possible that some important genes could have been missed in our analysis. Our SMR analysis used eQTL data from the blood because usable eQTL data from the eye is unavailable. Future studies using eQTL data from the eye are needed to explore whether our findings still hold. The SMR analysis was done using data from participants of European ancestry. However, since the prevalence of POAG is ethnic-specific, it is reasonable to postulate that the GWAS results might also be ethnic-specific. Therefore, our results might not be generalized to other ethnic populations.