Circadian genes and prostate cancer risk: Results from the EPICAP study

Abstract

Background Night work has been associated with cancer risk, including prostate cancer, in several studies. It has been suggested that disruption of circadian rhythm, induced by night work, may play a role in carcinogenesis. Indeed, circadian rhythms regulate several physiological functions and genes controlling the circadian rhythm were found to regulate cell proliferation, cell cycle and apoptosis. Very few studies have investigated the role of those circadian genes in prostate cancer occurrence. This study aims to analyze the association between circadian genes polymorphisms and prostate cancer risk based on data from the EPICAP study. Methods EPICAP is a French population-based case-control study including 819 incident prostate cancer cases and 879 controls frequency matched for age, among which 777 cases and 823 controls provided blood or saliva sample, allowing the constitution of a DNA bank. Blood and salivary samples were used for DNA extractions and genotyping using the Infinium® Oncoarray-500 K BeadChip (Illumina). This chip contains 276,000 single nucleotide polymorphisms (SNPs) assuring a genome-wide coverage and was completed by 234,000 SNPs selected for their potential relevance to the most common cancers including prostate cancer. Overall, the EPICAP genetic database included 1515 subjects (732 cases/783 controls) and 447,896 SNPs that passed stringent quality controls. For the purpose of this study, 31 circadian clock genes including 872 SNPs with a minor allele frequency (MAF) greater than 0.01 were selected. Genes from the circadian pathway was additionally divided into 2 sub-pathways: the well-established core circadian genes that include 9 genes (CLOCK, NPAS2, ARNTL, CSNKIE, CRY1, CRY2, PER1, PER2, and PER3) and the other 22 genes. Odds ratios (ORs) for association between prostate cancer and each SNP were estimated using unconditional logistic regression assuming a log-additive model. We also used a gene-based and pathway-based approach such as the Adaptive Rank Truncated Product (ARTP) method which combines association signals from the SNPs in a given gene (or from the genes in a pathway) to provide a P-value at the gene (or pathway level). Separate analyses were conducted by prostate cancer aggressiveness according to the Gleason score (low or intermediate score ≤ 7 [including 3 + 4], high score ≥ 7 [including 4 + 3]). Results No SNP was significantly associated to prostate cancer after correction for multiple testing using the false discovery rate (FDR) method. At the gene level, we reported 2 genes significantly associated to prostate cancer risk: NPAS2 (P = 0.046) and PER1 (P = 0.046). The whole circadian pathway was significantly associated to prostate cancer (P = 0.023) and this association was mainly supported by the circadian core-genes sub-pathway (P = 0.0002). Similar results were observed either in aggressive (P = 0.04 for the whole pathway and P = 0.01 for the core-genes sub-pathway) or less aggressive (P = 0.004 for the whole pathway and P = 0.001 for the core-genes sub-pathway) prostate cancer. Conclusion Our results support the hypothesis of a potential link between genetic variants in circadian genes and prostate cancer risk. Further investigation is warranted to confirm these findings and to better understand the biological pathways involved.