Comparative linkage mapping uncovers recombination suppression across massive chromosomal inversions associated with local adaptation in Atlantic silversides

Abstract

The role of recombination in genome evolution has long been studied in theory, but until recently empirical investigations had been limited to a small number of model species. Here, we compare the recombination landscape and genome collinearity between two populations of the Atlantic silverside (Menidia menidia), a small fish distributed across the steep latitudinal climate gradient of the North American Atlantic coast. We constructed separate linkage maps for locally adapted populations from New York and Georgia and their interpopulation laboratory cross. First, we used one of the linkage maps to improve the current silverside genome assembly by anchoring three large unplaced scaffolds to two chromosomes. Second, we estimated sexspecific recombination rates, finding 2.3-fold higher recombination rates in females than males-one of the most extreme examples of heterochiasmy in a fish. While recombination occurs relatively evenly across female chromosomes, it is restricted to only the terminal ends of male chromosomes. Furthermore, comparisons of female linkage maps revealed suppressed recombination along several massive chromosomal inversions spanning nearly 16% of the genome. These inversions segregate between locally adapted populations and coincide near perfectly with blocks of highly elevated genomic differentiation between wild populations. Finally, we discerned significantly higher recombination rates across chromosomes in the northern population compared to the southern. In addition to providing valuable resources for ongoing evolutionary and comparative genomic studies, our findings represent a striking example of structural variation that impacts recombination between adaptively divergent populations, providing empirical support for theorized genomic mechanisms facilitating adaptation despite gene flow.

Temperature-dependent gene regulatory divergence underlies local adaptation with gene flow in the Atlantic silverside

“…DEGs, inheritances classes, specific modules) are enriched inside one or multiple polymorphic inversions. First, we confirmed that known major inversions are polymorphic in our dataset by performing principal components analyses using SNPrelate (Zheng et al 2012) based on SNPs within known inversions (Tigano et al 2021;Akopyan et al 2022). Inversion boundaries were defined based on linkage mapping data as described in the Supplementary text.…”

“…To examine underlying gene regulatory differences, we sequenced whole-body transcriptomes of 42 individuals (6-8 from each population x temperature treatment, Fig 1a) and aligned reads to a linkage-map anchored version of the Atlantic silverside genome (Tigano et al 2021; Akopyan et al 2022). After filtering, we retained sufficient information for 18,517 annotated genes.…”

“…These relatively small proportions suggest that inversions are not the predominant source of regulatory variation, at least not directly. However, the inversions only cover a relatively small proportion of the genome (∼16%) (Akopyan et al 2022), so we tested if the proportion of DE and misexpressed genes was higher within individual inversions compared to the collinear genome. Indeed, DE genes were significantly enriched within inv4 and inv18 at 20°C, but only within inv18 at 26°C (Table 1).…”

“…and aligned reads to a linkage-map anchored version of the Atlantic silverside genome (Tigano et al 2021;Akopyan et al 2022). After filtering, we retained sufficient information for 18,517 annotated genes.…”

“…fecundity, metabolic rate, swimming performance, and foraging behavior (Yamahira and Conover 2002; Conover et al 2009; Hice et al 2012). Multiple large inversions (0.4 to 12.5 Mb) display fixed differences for opposite orientations (co-linear vs. inverted) across the latitudinal cline (Wilder et al 2020; Tigano et al 2021; Akopyan et al 2022), with otherwise minimal genetic background differentiation between populations, suggesting that chromosomal rearrangements that locally reduce recombination (Akopyan et al 2022) play a key role in maintaining co-adapted alleles despite gene flow.…”