Application of genomic tools to study and potentially improve the upper thermal tolerance of farmed Atlantic salmon (Salmo salar)

Background The Atlantic salmon (Salmo salar) aquaculture industry must mitigate the impacts of rising ocean temperatures and the increased prevalence/severity of marine heat waves. Therefore, we investigated the genetic architecture and gene expression (transcriptomics) responsible for determining a salmon's upper thermal tolerance. Results A genome-wide association study (GWAS) was conducted using fin clips of salmon from a previous incremental thermal maximum (ITMax) challenge (n = 251) and the North American 50 K SNP chip. ITMax was a highly polygenic trait with low/moderate heritability (mean SNP-based h(2) = 0.20 and pedigree-based h(2) = 0.25). Using data from the same fish, a separate GWAS assessed thermal-unit growth coefficient (TGC). Five significant SNPs were detected on chromosomes three and five, and high heritability estimates were calculated for TGC measured as fish grew from 12 to 20 degrees C (mean SNP-based h(2) = 0.62 and pedigree-based h(2) = 0.64). RNA-seq analyses of liver samples (n = 5-6 family(-1) temperature(-1)) collected from the four most and four least tolerant families at 10 and 20 degrees C were also used to provide insights into potential mechanisms modulating this species' thermal tolerance. Between the top and bottom families, 347 and 175 differentially expressed transcripts (FDR-adjusted p < 0.01; fold-change >=|2.0|) were identified at 10 and 20 degrees C, respectively. GO term enrichment analysis revealed unique responses to elevated temperature between family rankings (e.g., 'blood coagulation', 'sterol metabolic process' and 'synaptic growth at neuromuscular junction'). qPCR analyses further confirmed differences pertaining to cholesterol metabolism (lpl), inflammation (epx, elf3, ccl20), apoptosis (htra1b, htra2, anxa5b), angiogenesis (angl4, pdgfa), nervous system processes (insyn2a, kcnj11l) and heat stress (serpinh1b-1, serpinh1b-2). Three differentially expressed transcripts (i.e., ppp1r9a, gal3st1a, f5) were located in close proximity (+/- 120 kbp) to near-significant SNPs from the GWAS. Interestingly, ppp1r9a and gal3st1a have putative neurological functions, while f5 regulates blood coagulation. Conclusions These analyses provide several putative biomarkers of upper thermal tolerance in salmon that could prove valuable in helping the industry develop more temperature-tolerant fish. Further, our study supports previous reports that ITMax has low/moderate heritability in this species, and suggests that TGC at elevated temperatures is highly heritable.