Contrasting GC-content dynamics across 33 mammalian genomes: relationship with life-history traits and chromosome sizes

Both size and GC-content of minimal introns are selected in human populations

BACKGROUND

We previously have studied the insertion and deletion polymorphism by sequencing no more than one hundred introns in a mixed human population and found that the minimal introns tended to maintain length at an optimal size. Here we analyzed re-sequenced 179 individual genomes (from African, European, and Asian populations) from the data released by the 1000 Genome Project to study the size dynamics of minimal introns.

PRINCIPAL FINDINGS

We not only confirmed that minimal introns in human populations are selected but also found two major effects in minimal intron evolution: (i) Size-effect: minimal introns longer than an optimal size (87 nt) tend to have a higher ratio of deletion to insertion than those that are shorter than the optimal size; (ii) GC-effect: minimal introns with lower GC content tend to be more frequently deleted than those with higher GC content. The GC-effect results in a higher GC content in minimal introns than their flanking exons as opposed to larger introns (≥125 nt) that always have a lower GC content than that of their flanking exons. We also observed that the two effects are distinguishable but not completely separable within and between populations.

CONCLUSIONS

We validated the unique mutation dynamics of minimal introns in keeping their near-optimal size and GC content, and our observations suggest potentially important functions of human minimal introns in transcript processing and gene regulation.

Substitution patterns are under different influences in primates and rodents

There are large-scale variations of the GC-content along mammalian chromosomes that have been called isochore structures. Primates and rodents have different isochore structures, which suggests that these lineages exhibit different modes of GC-content evolution. It has been shown that, in the human lineage, GC-biased gene conversion (gBGC), a neutral process associated with meiotic recombination, acts on GC-content evolution by influencing A or T to G or C substitution rates. We computed genome-wide substitution patterns in the mouse lineage from multiple alignments and compared them with substitution patterns in the human lineage. We found that in the mouse lineage, gBGC is active but weaker than in the human lineage and that male-specific recombination better predicts GC-content evolution than female-specific recombination. Furthermore, we were able to show that G or C to A or T substitution rates are predicted by a combination of different factors in both lineages. A or T to G or C substitution rates are most strongly predicted by meiotic recombination in the human lineage but by CpG odds ratio (the observed CpG frequency normalized by the expected CpG frequency) in the mouse lineage, suggesting that substitution patterns are under different influences in primates and rodents.