A multilayered post-GWAS assessment on genetic susceptibility to pancreatic cancer

Abstract

Background: Pancreatic cancer (PC) is a complex disease in which both non-genetic and genetic factors interplay. To date, 40 GWAS hits have been associated with PC risk in individuals of European descent, explaining 4.1% of the phenotypic variance.

Methods: We complemented a new conventional PC GWAS (1D) with genome spatial autocorrelation analysis (2D) permitting to prioritize low frequency variants not detected by GWAS. These were further expanded via Hi-C map (3D) interactions to gain additional insight into the inherited basis of PC. In silico functional analysis of public genomic information allowed prioritization of potentially relevant candidate variants.

Results: We identified several new variants located in genes for which there is experimental evidence of their implication in the biology and function of pancreatic acinar cells. Among them is a novel independent variant in NR5A2 (rs3790840) with a meta-analysis p value = 5.91E-06 in 1D approach and a Local Moran's Index (LMI) = 7.76 in 2D approach. We also identified a multi-hit region in CASC8-a lncRNA associated with pancreatic carcinogenesis-with a lowest p value = 6.91E-05. Importantly, two new PC loci were identified both by 2D and 3D approaches: SIAH3 (LMI = 18.24), CTRB2/BCAR1 (LMI = 6.03), in addition to a chromatin interacting region in XBP1-a major regulator of the ER stress and unfolded protein responses in acinar cells-identified by 3D; all of them with a strong in silico functional support.

Conclusions: This multi-step strategy, combined with an in-depth in silico functional analysis, offers a comprehensive approach to advance the study of PC genetic susceptibility and could be applied to other diseases.

Keywords: 3D genomic structure; Genetic susceptibility; Genome-wide association analysis; Local indices of genome spatial autocorrelation; Pancreatic cancer risk.

Fig. 1

Study flowchart: overview of the complementary approaches adopted in this study to identify new pancreatic cancer susceptibility regions

Fig. 2

Three-dimensional genome organization in healthy and PANC-1 cells and association results corresponding to the genomic region around XBP1 using the standard GWAS and 2D approaches. a Coverage-normalized Hi-C maps of healthy samples and PANC-1 cells at 40 Kb resolution. Green ellipses highlight the interaction between the region harboring four Local Moran’s Index (LMI)-selected SNPs and the XBP1 promoter. b Tracks of the ChromHMM Chromatin for 8 states in healthy pancreas, PANC-1 cells, and a Pancreatic Intraepithelial Neoplasia 1B. Promoters are colored in light purple, strong enhancers in dark green, and weak enhancers in yellow. Note that the strong enhancer in the target region is lost in the PANC-1 and PanIN-1B samples, compared to the healthy samples. c UCSC tracks of H3K27ac, an enhancer-associated mark, and arcs linking significant interactions called by HOMER. Interactions in healthy pancreas samples are in green and those in PANC-1 and in the PanIN-1B sample are in purple. Red arc represents the interaction between LMI-prioritized SNPs and the XBP1 promoter (highlighted region in Hi-C map in a). d Scatterplots of SNPs in region chr22:28,400,000-29,600,000 (hg19) and their –log10 (p value), LMI, and odds ratio. Bait and target chromatin interaction regions are highlighted in yellow and blue, respectively

Fig. 3

Zoom plot of the 8q24.21 CASC8 (cancer Susceptibility 8) region and linkage disequilibrium pattern of the PanGenEU GWAS prioritized variants. Red and green points indicate OR < 1 and OR > 1, respectively

Fig. 4

Scatterplots of the –log10 p values, Local Moran’s Index (LMI) values, and odds ratios (OR) for three genomic regions prioritized based on their LMI value. Highlighted regions show the hits identified in the 2D, but not in the 1D approach

Fig. 5

Network of traits in the GWAS Catalog enriched with the genes prioritized in the 1D approach of PanGenEU GWAS. Twelve densely connected subgraphs identified via random walks are displayed in different colors