The Genomic Signature and Transcriptional Response of Metal Tolerance in Brown Trout Inhabiting Metal-Polluted Rivers

Abstract

Industrial pollution is a major driver of ecosystem degradation, but it can also act as a driver of contemporary evolution. As a result of intense mining activity during the Industrial Revolution, several rivers across the southwest of England are polluted with high concentrations of metals. Despite the documented negative impacts of ongoing metal pollution, brown trout (Salmo trutta L.) survive and thrive in many of these metal-impacted rivers. We used population genomics, transcriptomics, and metal burdens to investigate the genomic and transcriptomic signatures of potential metal tolerance. RADseq analysis of six populations (originating from three metal-impacted and three control rivers) revealed strong genetic substructuring between impacted and control populations. We identified selection signatures at 122 loci, including genes related to metal homeostasis and oxidative stress. Trout sampled from metal-impacted rivers exhibited significantly higher tissue concentrations of cadmium, copper, nickel and zinc, which remained elevated after 11 days in metal-free water. After depuration, we used RNAseq to quantify gene expression differences between metal-impacted and control trout, identifying 2042 differentially expressed genes (DEGs) in the gill, and 311 DEGs in the liver. Transcriptomic signatures in the gill were enriched for genes involved in ion transport processes, metal homeostasis, oxidative stress, hypoxia, and response to xenobiotics. Our findings reveal shared genomic and transcriptomic pathways involved in detoxification, oxidative stress responses and ion regulation. Overall, our results demonstrate the diverse effects of metal pollution in shaping both neutral and adaptive genetic variation, whilst also highlighting the potential role of constitutive gene expression in promoting metal tolerance.

Keywords: RADseq; RNAseq; adaptation; freshwater; pollution; toxic metals.

FIGURE 1

Schematic of collection sites and experimental design. (A) Location within the UK of the Cornwall and West Devon mining region and location of sampling sites. Metal‐impacted sites are coloured in red and control sites are coloured in blue. Filled coloured shapes represent sites sampled for tissue‐metal burden analysis before and after a period of depuration. The heatmap shows the scaled metal concentrations of each river (averaged over a 15‐year period; full data available in Table S2). (B) Timepoint 1: River. For population genomics, n = 20 brown trout were sampled from three metal‐contaminated rivers: Hayle (red triangle); Red River (red circle); Gannel (red square) and three control rivers: Camel (blue square); Fowey (blue circle) and Carrick Roads (blue triangle). (C) Timepoint 1: River. For tissue‐metal burden analysis, metal‐impacted trout were sampled from the Hayle (red triangle; n = 10) and control trout were sampled from the Fowey (blue circle; n = 10). (D) Timepoint 2: Depuration. Trout were sampled after 11 days of depuration in control water for tissue burden analysis (n = 10 per group) and transcriptomic analysis of the gill and liver using RNAseq (n = 6 per group).

FIGURE 2

Population genomic data describing the structure and diversity of metal‐impacted and control trout. (A) Heatmap of pairwise F _ST calculated between brown trout sampled from all six populations. (B) Principal component analysis (PCA) showing PC1 (~19% of the variance) against PC2 (~9% of the variance) and PC1 against PC3 (~8% of the variance). Individuals are coded by sampling site as in panel A. (C) Neighbour‐joining tree from aligned SNPs, computed using the K80 substitution model. (D) Proportion of polymorphic sites for each individual sampled from each of the six sampling sites. (E) Estimates of the effective population size (N _e) for each of the six sampling sites. (F) Quantification of fixed alleles calculated at each of the six sampling sites. (G) Redundancy Analysis (RDA; p = 0.007, R ² = 0.2145) showing RDA1 (p = 0.037) and RDA2 (p = 0.1). Red vectors represent RC1 (water chemistry features; Table S7) and RC2 (Cd, Cu, Ni, Zn). In all panels, sampling sites are coloured by metal‐impacted (red) or control (blue). For panels A, B, C, and F the sampling sites are coloured and shaped as: Hayle (red triangle); Red River (red circle); Gannel (red square); Camel (blue square); Fowey (blue circle) and Carrick Roads (blue triangle).

FIGURE 3

Tissue‐metal burden data of cadmium (Cd), copper (Cu), nickel (Ni), and zinc (Zn) measured in metal‐impacted (in red) and control (in blue) trout measured at Timepoint 1 (River) and Timepoint 2 (Depuration). Metal burden is measured as ng/mg of dry weight tissue for the (A) Gill; (B) Gut; (C) Kidney; (D) Liver; and (E) Muscle. Significance was measured using a Two‐way ANOVA (full details in Table S11). Only significant results from pairwise Tukey tests are included where **** indicates p ≤ 0.0001, *** indicates p ≤ 0.001, ** indicates p ≤ 0.01 and * indicates p ≤ 0.05.

FIGURE 4

Differential gene expression between metal‐impacted and control trout in the gill and liver. (A) Principal Components Analysis (PCA) of the variance stabilising transformation (VST) normalised counts in the gill. (B) Principal Components Analysis (PCA) of the VST normalised counts in the liver. Coloured dots represent the transcriptomic profile of individual fish from the metal‐impacted (red) and control (blue) rivers. (C) Volcano plot of the 28,261 genes measured in the gill. (D) Volcano plot of the 21,820 genes measured in the liver. Coloured dots in the volcano plots correspond to differentially expressed genes (DGEs) with FDR ≤ 0.05 and a log2foldchange of 1 (red) or −1 (blue), meaning genes are over‐expressed, or under‐expressed in metal‐impacted trout, respectively.