Germline lung cancer SNPs dysregulate known (LATS1) and novel (ADCY2) oncogenes through distal, spatially-constrained eQTLs.

Understanding how genomic variants contribute to lung cancer (LC) risk is key to better understanding the molecular mechanisms underlying that risk. While genome-wide association studies (GWAS) have identified numerous LC risk loci, most single nucleotide polymorphisms (SNPs) reside in non-coding regions, making the interpretation of their function challenging. We accounted for lung-specific chromatin interactions and allele-specific gene expression levels in lung tissue to identify novel interactions between LC GWAS SNPs and distal genes. Pathway enrichment analysis implicated eight target genes (CYP2A6, ADCY2, CHRNA3, CHRNA5, LATS1, RAD52, RIF1, TP53BP1) in functional networks involving caffeine metabolism, DNA ionizing radiation (IR)-double strand breaks and cellular response, and nicotine effect on dopaminergic neurons. Novel findings include a role for rs2853677 in ADCY2 dysregulation (previous attribution to TERT) and rs9322193 in targeting tumour suppressor gene LATS1 (previous attribution to RPS18P9/KATNA1). By linking germline variants to more biologically relevant gene targets and somatic processes, our results align more closely with established epidemiological and environmental risk factors for lung cancer, including a potential genetic explanation for the environmental interaction of caffeine and smoking in LC risk. This highlights the value of integrating 3D genome architecture and tissue-specific expression to refine our understanding of cancer susceptibility.