Item

Abstract

One of the advantages of sexual reproduction is the possibility of forming new combinations of alleles through crossovers (COs) that
swap portions of the maternal and paternal homologous chromosomes during meiosis, making new trait combinations available for
natural selection. The rate of CO formation and the locations of CO events can therefore affect the co-inheritance of traits. Although
several factors that influence CO rates and distributions have been identified, detailed knowledge of how the CO landscape is established
remains superficial. This is largely because characterizing the CO landscape has traditionally been laborious and imprecise. In this study,
we engineered cost-effective methods for performing high-throughput sequencing on large populations of recombinant individuals to
generate precise CO maps. Using this approach, we examined the influence of large- and small-scale genomic structural variations on CO
frequency and positioning by generating a CO map from over 2000 individuals of an F2 population derived from two Arabidopsis thaliana
accessions with high-quality reference genomes: Col and Ler. With these data, we were able to characterize a landscape of over 15,000
CO events within a single F2 cross, representing the densest CO map available for a higher eukaryote. We examined the frequency of COs
within and around inversions, insertions, deletions, translocations, and tandem copy number variations. COs occurred rarely within these
structural variants, but CO rates were often slightly elevated in the flanking regions. Other hypervariable regions of the genome, such as
disease resistance gene clusters, exhibited both high and low CO rates. COs were strongly associated with regions of open chromatin. We
conclude that COs are generally suppressed within regions containing structural variation, but that this effect does not depend on the
size of the variant region and is only marginally affected by the variant type.