Genome-wide association study of toxic metals and trace elements reveals novel associations

Abstract

The accumulation of toxic metals in the human body is influenced by exposure and mechanisms involved in metabolism, some of which may be under genetic control. This is the first genome-wide association study to investigate variants associated with whole blood levels of a range of toxic metals. Eleven toxic metals and trace elements (aluminium, cadmium, cobalt, copper, chromium, mercury, manganese, molybdenum, nickel, lead and zinc) were assayed in a cohort of 949 individuals using mass spectrometry. DNA samples were genotyped on the Infinium Omni Express bead microarray and imputed up to reference panels from the 1000 Genomes Project. Analyses revealed two regions associated with manganese level at genome-wide significance, mapping to 4q24 and 1q41. The lead single nucleotide polymorphism (SNP) in the 4q24 locus was rs13107325 (P-value = 5.1 × 10(-11), β = -0.77), located in an exon of SLC39A8, which encodes a protein involved in manganese and zinc transport. The lead SNP in the 1q41 locus is rs1776029 (P-value = 2.2 × 10(-14), β = -0.46). The SNP lies within the intronic region of SLC30A10, another transporter protein. Among other metals, the loci 6q14.1 and 3q26.32 were associated with cadmium and mercury levels (P = 1.4 × 10(-10), β = -1.2 and P = 1.8 × 10(-9), β = -1.8, respectively). Whole blood measurements of toxic metals are associated with genetic variants in metal transporter genes and others. This is relevant in inferring metabolic pathways of metals and identifying subsets of individuals who may be more susceptible to metal toxicity.

Figure 1.

Manhattan and quantile–quantile plot of GWAS of serum manganese. Each point corresponds to a SNP passing quality control (QC), plotted according to genomic position on the x-axis and the strength of association (−log₁₀ P-value) on the y-axis. The horizontal lines indicate genome-wide significance threshold (5 × 10⁻⁸) and also a nominal threshold of 10⁻⁰⁵.

Figure 2.

Manhattan and quantile–quantile plot of GWAS of serum cadmium. Each point corresponds to a SNP passing QC, plotted according to genomic position on the x-axis and the strength of association (−log₁₀ P-value) on the y-axis. The horizontal lines indicate genome-wide significance threshold (5 × 10⁻⁸) and also a nominal threshold of 10⁻⁰⁵.

Figure 3.

Manhattan and quantile–quantile plot of GWAS of serum mercury. Each point corresponds to a SNP passing QC, plotted according to genomic position on the x-axis and the strength of association (−log₁₀ P-value) on the y-axis. The horizontal lines indicate genome-wide significance threshold (5 × 10⁻⁸) and also a nominal threshold of 10⁻⁰⁵.