Heritable genetic variants in key cancer genes link cancer risk with anthropometric traits

Abstract

Background: Height and other anthropometric measures are consistently found to associate with differential cancer risk. However, both genetic and mechanistic insights into these epidemiological associations are notably lacking. Conversely, inherited genetic variants in tumour suppressors and oncogenes increase cancer risk, but little is known about their influence on anthropometric traits.

Methods: By integrating inherited and somatic cancer genetic data from the Genome-Wide Association Study Catalog, expression Quantitative Trait Loci databases and the Cancer Gene Census, we identify SNPs that associate with different cancer types and differential gene expression in at least one tissue type, and explore the potential pleiotropic associations of these SNPs with anthropometric traits through SNP-wise association in a cohort of 500,000 individuals.

Results: We identify three regulatory SNPs for three important cancer genes, FANCA, MAP3K1 and TP53 that associate with both anthropometric traits and cancer risk. Of particular interest, we identify a previously unrecognised strong association between the rs78378222[C] SNP in the 3' untranslated region (3'-UTR) of TP53 and both increased risk for developing non-melanomatous skin cancer (OR=1.36 (95% 1.31 to 1.41), adjusted p=7.62E^-63), brain malignancy (OR=3.12 (2.22 to 4.37), adjusted p=1.43E^-12) and increased standing height (adjusted p=2.18E^-24, beta=0.073±0.007), lean body mass (adjusted p=8.34E^-37, beta=0.073±0.005) and basal metabolic rate (adjusted p=1.13E^-31, beta=0.076±0.006), thus offering a novel genetic link between these anthropometric traits and cancer risk.

Conclusion: Our results clearly demonstrate that heritable variants in key cancer genes can associate with both differential cancer risk and anthropometric traits in the general population, thereby lending support for a genetic basis for linking these human phenotypes.

Keywords: cancer: CNS; cancer: dermatological; clinical genetics; complex traits; developmental.

Figure 1

A circos plot indicating the functional SNPs that significantly associate with anthropometric traits and cancer risk. SNPs are located on the top half of the plot while cancer types and anthropometric traits are on the bottom half. Significant associations (Bonferroni corrected p values <1E⁻⁵) are shown as solid lines. Blue lines indicate significant associations with anthropometric traits. Red lines depict significant associations with differential cancer risk.

Figure 2

A gene map of the chromosomal regions containing the pleiotropic functional SNPs. Lead SNPs are highlighted with red arrowhead. Tables summarise salient genome-wide association study (GWAS) Catalog, expression quantitative trait loci (eQTL) and Caner Gene Census information for each variant.

Figure 3

Detailed association results for cancer eSNPs with anthropometric traits and cancer risk in the UK Biobank cohort. A heatmap (A) depicting the significance of the associations of the three lead pleiotropic SNPs with differential cancer risk. The colour scale represents the −log10 adjusted p: with the darkest red on the scale being 7.62E⁻⁶³ and the lightest yellow being 1E⁻⁵. A radar plot (B) illustrating the level of significance of the associations of the three lead pleiotropic SNPs with anthropometric traits (radial axis: −log10 adjusted p). The darkest two pink categories are bioelectrical impedance measures (darkest being measures of fat). Traditional bedside anthropometric measures are in blush. SNPs related to FANCA (rs1805007†) and MAP3K1 (rs889312†) associate with standing height. The lead functional SNP related to TP53, rs78378222†, markedly associates with multiple measures of height and lean body mass, as well as basal metabolic rate. An error bar plot (C) of the beta coefficients (Y-axis) of the significant associations with anthropometric traits (X-axis). Error bars denote the 95% CI of the beta value. †Lead cancer eSNP.