Genomes are rarely homogeneous aggregations of Gs, As, Ts, and Cs. Indeed, variation in basepair frequency can have important implications for how genomes, and the organisms they generate, evolve. Regions with relatively homogenous GC content that extend for more than 300 kb known as isochores are prominent features of previously sequenced amniote genomes. Isochores are associated with a range of important variables, including gene density, intron length, DNA replication timing, and gene expression. GC-rich isochores also tend to experience high rates of recombination, resulting in elevated effective population sizes and increased efficiency of purifying selection relative to drift.
Although the abundance of isochores in birds and mammals indicates their presence in the ancestral amniote genome, reptile genomes have not been available to thoroughly test this prediction. Now, Fujita et al.‘s surveys of “GC composition in the Anolis genome reveal the most homogeneous amniote genome yet known, with no “classical” (>300 kb) GC-rich isochores and the most narrow GC content distribution of any amniote sequenced thus far.” Overall, anole isochores are rarer (just 15% of genome) and smaller (50% reduction) than in birds and mammals. The shorter introns and higher gene density associated with GC-rich isochores in other amniotes have also largely disappeared from anoles. As expected by fact that purifying selection is less effective in regions with low crossing over, lineage specific dN/dS in anoles is significantly lower in high GC regions.
To account for the relatively meager contribution of isochores to the anole genome, Fujita et al. infer erosion of GC-rich isochores in the squamate lineage at some point after divergence from the common ancestor of amniotes. The reasons for this erosion remain poorly understood, but Fujita et al. suggest that the GC-biased gene conversion that is thought to maintain isochores in other taxa may be weakened or reversed in anoles. Fujita et al.’s work is a good example of the insight offered by comparative genome sequencing; as the number of available genomes expands, this work is sure to continue to challenge overly simplistic assumptions about genome architecture and evolution derived from biased sampling of the tree of life.