Genome-Wide Enhancer-Gene Maps Link Genetic Variants To Colorectal Cancer Susceptibility

Objective:Colorectal cancer（CRC）is a major public health problem throughout the world that poses a serious threat to human health.Genome-wide association studies（GWASs）have identified more than 100 common loci associated with CRC susceptibility,and most risk variants lie in non-coding genomic regions,indicating that they might alter regulatory elements and control gene expression to confer CRC risk.However,identifying active regulatory elements and their target genes is still a great challenge.In this study,we created an enhancer-gene map by integrating multi-omics data,and further conducted case-control studies and biological experiments to explore the functional risk variant in CRC.Methods:First,assay for transposase accessible chromatin with high-throughput sequencing（ATAC-seq）,chromatin immunoprecipitation and high-throughput sequencing（Ch IP-seq）was conducted to generate chromatin activity landscape,and RNA-sequencing（RNA-seq）was used to measure global gene expression in CRC tissues.Chromatin interaction frequency was detected using Hi-C data from ENCODE project.Next,by applying an activity-by-contact（ABC）model which integrated multi-omics data,we created a genome-wide enhancer-gene map that predicted active enhancers and their target genes.To characterize the potential function of enhancer variants,we performed epigenetic mark annotation and GWAS enrichment analysis.Similarly,differential expression and functional enrichment analyses were applied to predict the biological function of enhancer-target genes.Then,we utilized GWAS data from db Ga P project to identify the enhancer variants associtated with CRC risk.A two-stage case-control study in Chinese population was further carried out to verify the association between the variant which showed the strongest gene regulation effect and CRC risk.Finally,we explored the underlying molecular mechanisms of the candidate variant through multipronged functional experiments,including luciferase reporter assays,electrophoretic mobility shift assays（EMSA）,chromosome conformation capture（3C）assays and CRISPR-Cas9 mediated single-nucleotide editing.Western-Blot was performed to detect cellcular signaling pathway,and in vitro as well as in vivo assays were conducted to assess the oncogene role of targeted genes.Result:A total of 34,130 enhancer-gene connections were identified by ABC models,including 15,121 enhancers and 12,351 target genes.Variants located in enhancers were found significantly enriched in active chromatin regions and CRC GWAS loci,indicating these variants may play an important role in CRC via regulating gene expression.Functional analyses revealed that the target genes possessed tumor molecular characteristics and might be greatly helpful for investigating potential oncogenic biomarkers.A total of 4,847enhancer variants associated with CRC risk were identified using GWAS data from db Ga P.Among them,rs4810856 showed the strongest regulation effect,which could regulate PREX1,CSE1L and STAU1.A two-stage case-control study in Chinese population further verified that the C>T change at rs4810856 can significantly reduce the risk of CRC(OR=0.90,95%CI=0.86-0.94,P=8.43×10^-6).The luciferase reporter assays and EMSAs showed that the C>T change at rs4810856 could affect the binding affinity to ZEB1,thereby weakening the activity of the enhancer.CRISPR-Cas9 mediated single-nucleotide editing showed the direct effect of rs4810856 on PREX1,CSE1L and STAU1 expression and CRC cellular aggressive phenotype.3C assays showed that rs4810856 could influence the chromatin interaction between the corresponding enhancer region and the promoters of the target genes.In addition,western-Blot suggested that the target genes were jointly involved in the AKT signaling pathway and could exert synergistic effect to activate this signaling pathway.Furthermore,in vitro and in vivo assays revealed that the target genes also exerted synergistic effect on CRC progression.Conclusions:Genome-wide enhancer-gene map was created by integrating multi-omics including ATAC-seq,Ch IP-seq,Hi-C and RNA-seq in CRC.Case-control studies and biological assays indicated that the C>T change at rs4810856 significantly reduced the expression of PREX1,CSE1L and STAU1,which could inhibit the activation of AKT signaling pathway,thereby reducing CRC risk.These findings provided a new insight into genetic susceptibility to CRC and may contribute to the personalized medicine of the tumor.