Comparative genomics of cetartiodactyla: energy metabolism underpins the transition to an aquatic lifestyle

Abstract

Foraging disruption caused by human activities is emerging as a key issue in cetacean conservation because it can affect nutrient levels and the amount of energy available to individuals to invest into reproduction. Our ability to predict how anthropogenic stressors affect these ecological processes and ultimately population trajectory depends crucially on our understanding of the complex physiological mechanisms that detect nutrient availability and regulate energy metabolism, foraging behavior and life-history decisions. These physiological mechanisms are likely to differ considerably from terrestrial mammalian model systems. Here, we examine nucleotide substitution rates in cetacean and other artiodactyl genomes to identify signatures of selection in genes associated with nutrient sensing pathways. We also estimated the likely physiological consequences of adaptive amino acid substitutions for pathway functions. Our results highlight that genes involved in the insulin, mTOR and NF-ĸB pathways are subject to significant positive selection in cetaceans compared to terrestrial artiodactyla. These genes may have been positively selected to enable cetaceans to adapt to a glucose-poor diet, to overcome deleterious effects caused by hypoxia during diving (e.g. oxidative stress and inflammation) and to modify fat-depot signaling functions in a manner different to terrestrial mammals. We thus show that adaptation in cetaceans to an aquatic lifestyle significantly affected functions in nutrient sensing pathways. The use of fat stores as a condition index in cetaceans may be confounded by the multiple and critical roles fat has in regulating cetacean metabolism, foraging behavior and diving physiology.

Keywords: Cetaceans; NF-ĸB; energy metabolism; evolution; insulin; mTOR.

Figure 1

Species cladogram of the cetacea ingroup (blue) and artiodactyla, human and mouse outgroups (black and red), derived from 532 genes selected from six key nutrient-sensing pathways. Branch labels display posterior branch probabilities.

Figure 2

Ratios of non-synonymous versus synonymous nucleotide substitution rates (dN:dS ratios; ω) in 532 genes estimated from codon evolution models in PAML. (A) baseline estimates for whole alignments from null models (ω₀), estimates for foreground (cetacea; ω_f) and background (all others; ω_b) branches from branch models, and estimates for codons under purifying (ω₀) and positive (ω₂) selection in foreground branch (cetacea) from branch-site models. (B) Foreground dN:dS ratio (ω_f) and statistical significance (P value) from likelihood-ratio tests between free-ratio branch models and neutral branch models. Significant tests after FDR correction (q ≤ 0.05) are highlighted in blue. (C) Foreground dN:dS ratio of positively selected codons (ω₂) and statistical significance (P value) from likelihood-ratio tests between positive-selection branch-site models and neutral branch-site models. Significant tests after FDR correction (q ≤ 0.05) are highlighted in orange. The red solid lines in all plots represent neutral evolution (ω = 1).

Figure 3

Distribution of PROVEAN scores of 1936 amino-acid substitutions under positive selection in cetacea. Functional effects of substitutions are colour-coded according to standard PROVEAN thresholds: positive scores indicate neutral effects and scores below −2.5 indicate biologically significant functional changes.

Figure 4

Functional interactions between genes under positive selection in the insulin signaling pathway (ingenuity pathway analysis, IPA). Genes with functionally relevant amino-acid substitutions (PROVEAN score ≤ −2.5) coloured in yellow. Possible downstream effects of these genes were visualized using the molecule activity predictor (MAP) tool in IPA (see prediction legend). The interactions highlight evolutionary changes in glucose metabolism.

Figure 5

Functional interactions between genes under positive selection in the SIRT signaling pathway (ingenuity pathway analysis, IPA). Genes with functionally relevant amino-acid substitutions (PROVEAN score ≤ −2.5) coloured in yellow. Possible downstream damaging effects of these genes were visualized using the molecule activity predictor (MAP) tool in IPA (see prediction legend). These highlight changes to gluconeogenesis and fatty acid oxidation.

Figure 6

Functional interactions between genes under positive selection in the NF-ĸB signaling pathway (ingenuity pathway analysis, IPA). Genes with functionally relevant amino-acid substitutions (PROVEAN score ≤ −2.5) coloured in yellow. Possible downstream damaging effects of these genes were visualized using the molecule activity predictor (MAP) tool in IPA (see prediction legend). These highlight changes in responses to hypoxia and inflammation.

Figure 7

Functional interactions between genes under positive selection in the mTOR signaling pathway (ingenuity pathway analysis, IPA). Genes with functionally relevant amino-acid substitutions (PROVEAN score ≤ −2.5) coloured in yellow. Possible downstream damaging effects of these genes were visualized using the molecule activity predictor (MAP) tool in IPA (see prediction legend). These highlight important changes to key biological decisions about energetic investment.