The phenomics and expression quantitative trait locus mapping of brain transcriptomes regulating adaptive divergence in lake whitefish species pairs (Coregonus sp.)

Abstract

We used microarrays and a previously established linkage map to localize the genetic determinants of brain gene expression for a backcross family of lake whitefish species pairs (Coregonus sp.). Our goals were to elucidate the genomic distribution and sex specificity of brain expression QTL (eQTL) and to determine the extent to which genes controlling transcriptional variation may underlie adaptive divergence in the recently evolved dwarf (limnetic) and normal (benthic) whitefish. We observed a sex bias in transcriptional genetic architecture, with more eQTL observed in males, as well as divergence in genome location of eQTL between the sexes. Hotspots of nonrandom aggregations of up to 32 eQTL in one location were observed. We identified candidate genes for species pair divergence involved with energetic metabolism, protein synthesis, and neural development on the basis of colocalization of eQTL for these genes with eight previously identified adaptive phenotypic QTL and four previously identified outlier loci from a genome scan in natural populations. Eighty-eight percent of eQTL-phenotypic QTL colocalization involved growth rate and condition factor QTL, two traits central to adaptive divergence between whitefish species pairs. Hotspots colocalized with phenotypic QTL in several cases, revealing possible locations where master regulatory genes, such as a zinc-finger protein in one case, control gene expression directly related to adaptive phenotypic divergence. We observed little evidence of colocalization of brain eQTL with behavioral QTL, which provides insight into the genes identified by behavioral QTL studies. These results extend to the transcriptome level previous work illustrating that selection has shaped recent parallel divergence between dwarf and normal lake whitefish species pairs and that metabolic, more than morphological, differences appear to play a key role in this divergence.

Figure 1.—

Genome locations of significant eQTL (α = 0.05), phenotypic QTL, and outlier AFLP loci from a genome scan. Shown separately are (a) combined, (b) female-only, and (c) male-only data sets. Linkage group numbers are shown and successive linkage groups are separated by a vertical line. Numbers on the x-axis correspond to successively numbered genetic markers on each linkage group and each interval on the x-axis corresponds to a 17.6-cM bin. Phenotypic QTL correspond to Rogers and Bernatchez (2007) and are shown with 2.0-LOD intervals, represented by boxes. Phenotypic QTL are abbreviated as follows: A, activity; B, burst swim; C, condition factor; De, depth preference; Di, directional swimming change; G, growth rate; Gr, gill raker number; L, life history (age at maturity). Locations of outlier AFLP loci from a previous genome scan (Rogers and Bernatchez 2007) are shown with arrows. Shading of histograms, boxes, and arrows reflects colocalization with eQTL: histogram bars that colocalize with either phenotypic QTL or outlier loci are solid; otherwise, they are shaded. Arrows (representing outlier loci) that colocalize with eQTL are solid; otherwise, they are shaded. Similarly, phenotypic QTL and their 2.0-LOD interval boxes that overlap with eQTL are solid and otherwise are shaded. The number of species pairs for which AFLP loci were outliers is not shown (see Table 4 for those that colocalize with eQTL). Hotspots are defined as bins containing five or more eQTL and are shown numbered successively within each of the analyses according to Table 2.