Multivariate statistical analysis of metabolomics profiles in tissues of polar bears (Ursus maritimus) from the Southern and Western Hudson Bay subpopulations
Polar bears (Ursus maritimus) are apex predators of the Arctic, which exposes them to an array of natural and anthropogenic stress factors. Metabolomics analysis profiles endogenous metabolites that reflect the response of biological systems to stimuli, and the effects of multiple stressors can be assessed from an integrated perspective. A targeted, quantitative, liquid chromatography–mass spectrometry-based metabolomics platform [219 metabolites including amino acids, biogenic amines, acylcarnitines, phosphatidylcholines (PCs), sphingomyelins, hexoses (Hex), and fatty acids (FAs)] was applied to the muscle and liver of polar bears from the Southern and Western Hudson Bay (Canada) subpopulations (SHB and WHB, respectively). Multivariate statistics were then applied to establish whether bears were discriminated by sex and/or subpopulation. Five metabolites identified by variable importance projection (VIP) discriminated the hepatic profiles of SHB males and females (Hex, arginine, glutamine, one PC, one sphingomyelin), while fifteen metabolites (primarily PCs along with leucine) contrasted the livers of males from SHB and WHB. Metabolite profiles in the muscle of male and female bears could not be differentiated; however, the muscles of SHB and WHB males were discriminated primarily by PCs and FAs. Stable isotope ratios (δ13C and δ15N) were variably related to metabolites; δ13C was correlated with some VIP metabolite concentrations, particularly in comparisons of male bears from SHB and WHB, suggesting an influence of dietary differences. However, δ15N and age exhibited few, relatively weak correlations with metabolites. The metabolite profiles discriminating the sexes and subpopulations may have utility for future assessments regarding the effects of specific stressors on the physiology of Hudson Bay polar bears.
|Keywords||Hudson Bay, Metabolomics, Multivariate analysis, Partial least squares discriminant analysis (PLS-DA), Polar bears (Ursus maritimus)|
Morris, A.D. (A. D.), Letcher, R.J, Dyck, M. (M.), Chandramouli, B. (B.), Fisk, A.T. (A. T.), & Cosgrove, J. (J.). (2017). Multivariate statistical analysis of metabolomics profiles in tissues of polar bears (Ursus maritimus) from the Southern and Western Hudson Bay subpopulations. Polar Biology, 1–17. doi:10.1007/s00300-017-2200-6