4.4 Linking the LDH enzymes to genes
Contrasting homozygosity for the allozyme locus LDH-1 detected the sympatric populations in Lake Bunnersjöarna (Allendorf et al., 1976; Ryman et al., 1979), and we tried to identify the LDH-1 gene in the genome and find the sequence differences that result in the allozyme divergence. We were not fully successful in this, however. We located two LDH-A loci as expected from the genome duplication event. One locus was located on chromosome 7 and the other on chromosome 17. The amino acid composition of the protein products from these loci suggested the LDH locus on chromosome 7 to be LDH-1 due to a slightly more positive charge of the product at this locus as compared to the one on chromosome 17, as expected from allozyme patterns (Allendorf et al., 1984). However, our SNP- and gene-wise analyses revealed that neither copy of LDH-A that we identified in the reference genome shows a fixed or even strong differentiation between the demes.
When investigating the Pool-seq reads mapping to the two gene copies visually in the Integrative Genomics Viewer (IGV), we identified a region in the 3’ UTR of the LDH-A on chromosome 17 that showed strong divergence, albeit not fixed differences, between the two demes, which was confirmed by allele frequency calculations (Figures S7, S9).LDH-A is highly regulated at the transcriptional and posttranscriptional levels (Jungmann, Huang, & Tian, 1998), and the 3’ UTR region is known to be a critical determinant for the mRNA stabilizing activity for this gene (Tian, Huang, Short, Short, & Jungmann, 1998). We hypothesized that the divergence in this region could result in the null allele for Deme II observed in allozyme data, but this would imply that LDH-1 is located on chromosome 17 (and not on 7 as indicated by the amino acid charges). A previous study suggests that regulatory regions may primarily be involved in divergence between the two sympatric forms of whitefish – dwarf and normal – in Cliff Lake in St John River drainage in Maine, USA, since outliers were predominantly found in non-coding regions within genes (Hebert, Renaut, & Bernatchez, 2013).
When we examined the 3´ UTR of the LDH-A on chromosome 17 in the whole-genome sequences from a total of 16 individuals (from 7 lakes, including 2 individuals from each of the demes) scored for allozyme genotype in LDH-1, we found that the majority of the individuals (n =11) showed similar patterns as Deme I in Pool-seq data. However, one individual from Lake Saxvattnet, and all four individuals from Lakes Bunnersjöarna (both demes) showed the Deme II Pool-seq genotype at the 3’ UTR region of chromosome 17 (Figure S9). These results indicate that LDH-1 expression is a complex feature, and we were not able to resolve the genomic background to the contrasting allozyme expressions in the demes. Rather, further studies need to identify the exact genetic control behind the LDH-1 enzyme in brown trout. Our results suggest that contrasting homozygosity of an allozyme (F ST=1) may not result directly from contrasting homozygosity in a DNA sequence. Identifying the genes and the regulatory mechanisms underlying these enzymes is particularly challenging in duplicated genomes.