Identifying polymorphic cis-regulatory variants as risk markers for lung carcinogenesis and chemotherapy responses in tobacco smokers from eastern India

Abstract

Aberrant expression of xenobiotic metabolism and DNA repair genes is critical to lung cancer pathogenesis. This study aims to identify the cis-regulatory variants of the genes modulating lung cancer risk among tobacco smokers and altering their chemotherapy responses. From a list of 2984 SNVs, prioritization and functional annotation revealed 22 cis-eQTLs of 14 genes within the gene expression-correlated DNase I hypersensitive sites using lung tissue-specific ENCODE, GTEx, Roadmap Epigenomics, and TCGA datasets. The 22 cis-regulatory variants predictably alter the binding of 44 transcription factors (TFs) expressed in lung tissue. Interestingly, 6 reported lung cancer-associated variants were found in linkage disequilibrium (LD) with 5 prioritized cis-eQTLs from our study. A case-control study with 3 promoter cis-eQTLs (p < 0.01) on 101 lung cancer patients and 401 healthy controls from eastern India with confirmed smoking history revealed an association of rs3764821 (ALDH3B1) (OR = 2.53, 95% CI = 1.57-4.07, p = 0.00014) and rs3748523 (RAD52) (OR = 1.69, 95% CI = 1.17-2.47, p = 0.006) with lung cancer risk. The effect of different chemotherapy regimens on the overall survival of lung cancer patients to the associated variants showed that the risk alleles of both variants significantly decreased (p < 0.05) patient survival.

Figure 1

Pathway analysis for identifying regulatory genetic loci as susceptibility markers conferring risk towards tobacco smoke-induced lung carcinogenesis. (A) After a detailed literature review on epidemiological reports, association studies, expression studies, and case studies followed by ONCOMINE and TCGA validation, important xenobiotic metabolism and DNA repair genes were selected in cigarette smoke-induced lung cancer. (B) The selection and screening of genes, their DNase I hypersensitive sites, and SNVs within the DNase I hypersensitive sites by step-wise use of in silico tools and databases to prioritize potential regulatory SNVs as susceptibility loci in lung carcinogenesis with replications in case–control cohorts. DHS, DNase I hypersensitive sites, eQTL, expression quantitative trait loci, XMG, xenobiotic metabolism gene group, DRG, DNA repair gene group, rSNP, regulatory single nucleotide polymorphism, MAF, minor allele frequency.

Figure 2

Genotyping of (A) rs3764821 of ALDH3B1, (B) rs3748523 of RAD52 and (C) rs5742926 of PMS1 by PCR–RFLP method with a representative chromatogram of Sanger sequencing for each genotype of the rSNPs. For rs3764821, Gel 1: Lane 1,2,4,6 depicts AG genotypes with cut patterns as 240 bp, 198 bp, 42 bp; Lane 3, 7, 8 depicts GG genotypes with 198 bp and 42 bp fragments. For rs3748523, Gel 2: Lane 1,2,3,7 & 8 depicts CC genotypes with 228 bp and 24 bp (not visible) fragments; Lane 4 & 6 depicts GG genotypes as uncut (252 bp) fragments; Lane 8 depicts CG genotype with 252 bp, 228 bp and 24 bp fragments. For rs5742926, Gel 3: Lane 1, 2 & 5 depicts GG genotypes as uncut (340 bp) fragments and Lane 4 depicts GT genotype with 340 bp, 231 bp, 109 bp. A representative chromatogram for heterozygous peak is also provided for each rSNP.

Figure 3

Kaplan–Meier curves depict the association between overall survival in lung cancer patients and the cis-regulatory polymorphic variants in the eastern Indian population. It shows significantly lower overall survival in lung cancer patients with (A) rs3764821 (ALDH3B1); a combination of heterozygous and homozygous variant genotypes (AG + GG), and (B) rs3748523 (RAD52); the combination of heterozygous and homozygous variant genotypes (CG + GG). Significance at log-rank p < 0.05*.

Figure 4

Kaplan–Meier curves depict the association between the polymorphic cis-regulatory variants and overall survival in lung cancer patients treated with different chemotherapy regimens in the eastern Indian population. It shows significantly lower overall survival in lung cancer patients with (A) rs3764821 (ALDH3B1); treated with gemcitabine-cis/carboplatin in the second regimen and paclitaxel-cis/carboplatin in the first regimen, and (B) rs3748523 (RAD52); treated with paclitaxel-cis/carboplatin in the first regimen. Significance at log-rank p < 0.05*.