Abstract SY25-01: Identification of enhancer elements at kidney cancer susceptibility loci using genome-wide approaches in which post-GWAS functional studies implicate the SWI/SNF DPF3 gene for the 14q24 risk locus

Abstract

Abstract We investigated the set of common single nucleotide polymorphisms (SNPs) that are highly correlated with 20 renal cancer (RCC) susceptibility regions identified by Genome-wide Association Study (GWAS). Together extensive fine-mapping and in silico functional analyses, based on publicly available resources, we used an integrated approach to pursue regulatory elements in which one or more variant allele could confer differential functional activity. To this end, we conducted an initial screening Massively Parallel Reporter Assay (MPRA), followed by three additional approaches: 1. Assay for Transposase-Accessible Chromatin (ATAC-seq); 2. RCC eQTL analysis;3. Capture-HiC. We confirmed the effects reported for three regions (8q24, 11q13 and 12p12). We selected 784 SNPs with an r2>0.4 or D’>0.5 and MAF<0.05 that had evidence for enhancer activity based on available ENCODE transcription factor or histone ChIP-seq, FAIRE, and DNase I hypersensitive data. The MPRA library included 47,461 oligonucleotides of which, 145 bps contained the SNP in either forward or reverse orientation followed by an identification barcode. The set of 784 SNPs tested were tagged 10 separate times for both orientations, while controls were constructed by randomizing the 10 nucleotides centered on the SNP. The MPRA library was the cloned into a pMPRA reporter vector that was transfected into HEK293T and ACHN cell lines, under normoxia and hypoxia conditions; next generation sequence was performed on 5 replicates per condition and analyzed using regression for allele, direction, cell line and condition effects. For 18 of 20 RCC regions, we identified one or more functional variants while Cis-eQTL identified 16 RCC regions that mapped to alterations in expression of 198 genes. Our integrated approach underscores the spectrum of effects of genetic variation on regulation of genes critical for RCC, each notable for determining functional allelic effects on nearby (but not always the closest gene). Our work illustrates the importance of regulatory variants and RCC risk. We further pursued the region on 14q24 and confirmed the strongest signal on MPRA, the GWAS tag SNP, rs4903064, by luciferase assay. In a series of electromobility shift assays (EMSA), we showed higher binding to the rs4903064 T-allele, associated with lower risk (compared to the C-risk allele). Interestingly, the T-allele creates a binding site for IRX2/IRX5, which can act as a transcriptional repressor, whereas the C-allele generates a HIF1a binding site. By eQTL analysis, we observed higher expression of the nearby gene DPF3 (on 14q24) associated with the C-allele in both the TCGA RCC and a normal kidney eQTL data set. In four stable renal cell lines that included a DOX-dependency for overexpression of DPF3 isoforms (DPF3a or DPF3b), we performed RNA-seq analysis and observed over 500 genes deregulated when either isoform is overexpressed. To investigate chromatin accessibility changes due to DPF3 overexpression, we used ATAC-seq and found 86 regions with altered accessibility and 15% of the RNA-seq deregulated genes are located within 2 MB of one of the regions, including the promoter of the top deregulated gene (CEMIP). In growth curve assays, DPF3 increased the growth rate of renal cell lines. Moreover, knockdown of the top two DPF3-deregulated genes (CEMIP and IL23R) by siRNA, reduced the effect of DPF3, suggesting that CEMIP and IL23R could be important cofactors for growth along with DPF3. Since CEMIP is important for apoptosis and in our dox-inducible-DPF3 cell lines, DPF3 overexpression decreased the percentage of apoptotic cells, measured by flow cytometry, and reduced Caspase-3 and PARP cleavage, as determined by Western blot (WB). The second deregulated gene, IL23R, interacts with STAT3, an important RCC oncogene, which has increased phosphorylation following DPF3 overexpression and leads to increased STAT3 phosphorylation. While the SWI/SNF complex genes have been shown to influence response to immunotherapy, we extended our studies to look at co-cultured of tumor cells with stimulated fresh peripheral blood mononuclear cells (PBMCs); DPF3 overexpression in RRC cells leads to increased PBMC-cytotoxicity, mainly due to an increased proportion of T CD8+ cells. In summary, we integrated RCC post-GWAS functional studies to pinpoint notable variants on 14q24 and suggest that altering DPF3 regulation can perturb the SWI/SNF complex by reduction of apoptosis and STAT3 pathway activation. Further confirmation and follow-up studies are required to explain the functional underpinnings of the additional promising regions identified by GWAS and pursued by large-scale screening approaches and fine-mapping. Citation Format: Leandro M. Colli, Lea Jessop J. Chanock, Timothy Myers, Mitchell Machiela, Jiyeon Choi, Mark Purdue, Kevin Brown, Stephen J. Chanock. Identification of enhancer elements at kidney cancer susceptibility loci using genome-wide approaches in which post-GWAS functional studies implicate the SWI/SNF DPF3 gene for the 14q24 risk locus [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr SY25-01.