Background: Chronic exposure to certain metals plays a role in disease development. Integrating untargeted metabolomics with urinary metallome data may contribute to better understanding the pathophysiology of diseases and complex molecular interactions related to environmental metal exposures. To discover novel associations between urinary metal biomarkers and metabolism networks, we conducted an integrative metallome-metabolome analysis using a panel of urinary metals and untargeted blood metabolomic data from the Strong Heart Family Study (SHFS).
Methods: The SHFS is a prospective family-based cohort study comprised of American Indian men and women recruited in 2001-2003. This nested case-control analysis of 145 participants of which 50 developed incident diabetes at follow up in 2006-2009, included participants with urinary metal and untargeted metabolomic data. Concentrations of 8 creatinine-adjusted urine metals/metalloids [antimony (Sb), cadmium (Cd), lead (Pb), molybdenum (Mo), selenium (Se), tungsten (W), uranium (U) and zinc (Zn)], and 4 arsenic species [inorganic arsenic (iAs), monomethylarsonate (MMA), dimethylarsinate (DMA), and arsenobetaine (AsB)] were measured. Global metabolomics was performed on plasma samples using high-resolution Orbitrap mass spectrometry. We performed an integrative network analysis using xMWAS and a metabolic pathway analysis using Mummichog.
Results: 8,810 metabolic features and 12 metal species were included in the integrative network analysis. Most metal species were associated with distinct subsets of metabolites, forming single-metal-multiple-metabolite clusters (|r|>0.28, p-value < 0.001). DMA (clustering with W), iAs (clustering with U), together with Mo and Se showed modest interactions through associations with common metabolites. Pathway enrichment analysis of associated metabolites (|r|>0.17, p-value < 0.1) showed effects in amino acid metabolism (AsB, Sb, Se and U), fatty acid and lipid metabolism (iAs, Mo, W, Sb, Pb, Cd and Zn). In stratified analyses among participants who went on to develop diabetes, iAs and U clustered together through shared metabolites, and both were associated with the phosphatidylinositol phosphate metabolism pathway; metals were also associated with metabolites in energy metabolism (iAs, MMA, DMA, U, W) and xenobiotic degradation and metabolism (DMA, Pb) pathways.
Conclusion: In this integrative analysis of multiple metals and untargeted metabolomics, results show common associations with fatty acid, energy and amino acid metabolism pathways. Results for individual metabolite associations differed for different metals, indicating that larger populations will be needed to confirm the metal-metal interactions detected here, such as the strong interaction of uranium and inorganic arsenic. Understanding the biochemical networks underlying metabolic homeostasis and their association with exposure to multiple metals may help identify novel biomarkers, pathways of disease, potential signatures of environmental metal exposure.
Keywords: American Indians; Diabetes; Integrative omics; Metabolomics; Metal-mixtures; Metallomics; Metals.
Copyright © 2021 The Authors. Published by Elsevier Ltd.. All rights reserved.