Aim:To identify risk variants associated with gene expression in peripheral blood and to identify genes whose expression change may contribute to the susceptibility to periodontitis.
Material and Methods:We systematically integrated the genetic associations from a recent large-scale periodontitis GWAS and blood expression quantitative trait loci (eQTL) data using Sherlock, a Bayesian statistical framework. We then validated the potential causal genes in independent gene expression data sets. Gene co-expression analysis was used to explore the functional relationship for the identified causal genes.Results: Sherlock analysis identified 10 genes (rs7403881 for MT1L, rs12459542 for SIGLEC5, rs12459542 for SIGLEC14, rs6680386 for S100A12, rs10489524 for TRIM33, rs11962642 for HIST1H3E, rs2814770 for AIM2, rs7593959 for FASTKD2, rs10416904 for PKN1, and rs10508204 for WDR37) whose expression may influence periodontitis. Among these genes, AIM2 was consistent significantly upregulated in periodontium of periodontitis patients across four data sets. The cis-eQTL (rs2814770, ~16 kb upstream of AIM2) showed significant association with AIM2 (p = 6.63 × 10 -6 ) and suggestive association with periodontitis (p = 7.52 × 10 -4 ). We also validated the significant association between rs2814770 and AIM2 expression in independent expression data set. Pathway analysis revealed that genes co-expressed with AIM2 were significantly enriched in immune-and inflammation-related pathways.
Conclusion:Our findings implicate that AIM2 is a susceptibility gene, expression of which in gingiva may influence periodontitis risk. Further functional investigation of AIM2 may provide new insight for periodontitis pathogenesis.
K E Y W O R D SAIM2, eQTL, GWAS, periodontitis, Sherlock