Trans-Acting Genotypes Drive mRNA Expression Affecting Metabolic And Thermal Tolerance Traits

Abstract

AO_SCPLOWBSTRACTC_SCPLOWEvolutionary processes driving physiological trait variation depend on the underlying genomic mechanisms. Evolution of these mechanisms depends on whether traits are genetically complex (involving many genes) and how gene expression that impact the traits is converted to phenotype. Yet, genomic mechanisms that impact physiological traits are diverse and context dependent (e.g., vary by environment or among tissues), making them difficult to discern. Here we examine the relationships between genotype, mRNA expression, and physiological traits to discern the genetic complexity and whether the gene expression effecting the physiological traits is primarily cis or trans-acting. We use low-coverage whole genome sequencing and tissue specific mRNA expression among individuals to identify polymorphisms directly associated with physiological traits and expressed quantitative trait loci (eQTL) driving variation in six temperature specific physiological traits (standard metabolic rate, thermal tolerance, and four substrate specific cardiac metabolic rates). Not surprisingly, there were few, only five, SNPs directly associated with physiological traits. Yet, by focusing on a select set of mRNAs belonging to co-expression modules that explain up to 82% of temperature specific (12{degrees}C or 28{degrees}C) metabolism and thermal tolerance, we identified hundreds of significant eQTL for mRNA whose expression affects physiological traits. Surprisingly, most eQTL (97.4% for heart and 96.7% for brain) of eQTL were trans-acting. This could be due to higher effect size or greater importance of trans versus cis acting eQTLs for mRNAs that are central to co-expression modules. That is, we may have enhanced the identification of trans-acting factors by looking for SNPs associated with mRNAs in co-expression modules that are known to be correlated with the expression of 10s or 100s of other genes, and thus have identified eQTLs with widespread effects on broad gene expression patterns. Overall, these data indicate that the genomic mechanism driving physiological variation across environments is driven by trans-acting tissue specific mRNA expression. AO_SCPLOWUTHORC_SCPLOW SO_SCPLOWUMMARYC_SCPLOWIn the salt marsh killifish Fundulus heteroclitus there is amazingly large variation in physiological traits assumed to be under stabilizing selection, which should reduce their variation. To discern the heritability of this physiological variation we took an innovative approach to define the DNA variation that drives mRNA expression linked to physiological variation. This indirect approach revealed many DNA sequence variants associated with physiological variation via their effect on mRNA expression. Surprisingly, these changes were not in the mRNAs themselves, but in unlinked distant genes which regulate mRNA expression. That is, the vast majority (>95%) were trans-acting. This is surprising because trans-acting effects are found less often than DNA variants within or close to mRNA expression genes. Our results are likely related to the select subset of mRNAs across environments that are linked to physiological variation.