This is a fascinating article in Nature by Hurst on genes and evolution, which during a time when Darwin’s work is being honoured is coming up with interesting results
What makes us humans unique, this must be in unique genes, where Darwinian positive selection has occurred. Genes thought to be hotspots for positive selection can be discovered by genome scans that pinpoint especially fast evolutionary change in DNA sequences. Work by Berglund et al. and Galtier et al, however, undermines the assumed connection between fast evolution and positive selection. Instead, it seems that hotspots have probably accelerated evolution by means of a biased DNA repair process, not because the changes were good for us. Indeed, many changes are probably detrimental.
We assume that mutations become common either through conferring an advantage on the organism (positive selection) or through chance (drift).
Berglund et al. report that their methods pinpoint different candidates for positive selection. Any position in a gene is occupied by one of four nucleotides, A, T, C or G, combinations of which code for amino acids – the building blocks of proteins. Curiously, the top candidates for positive selection show a great excess of nucleotide changes that were ancestrally either A or T but became G or C. Galtier et aI. find the same effect, and also show that it applies to hotspots in nonhuman primates.
If positive selection depended upon the choice of amino acids should not so consistently prefer a change of AT to Gc. Moreover, the bias, although highly localized within genes, is not unique to the protein coding parts of the gene, but is seen in the intervening non coding parts as well. Both groups conclude that the hotspot genes are not under positive selection at the protein level.
Some force is driving the transformation of AT to GC. What might the biasing force be? Changing nucleotides at synonymous sites modulates expression of a gene. But why then would the changes be highly localized within genes, and why is the bias also in non coding sequence? A simpler explanation than positive selection on either proteins or expression rate, and one that anticipated the new results, suggests a biased DNA repair process.
During the manufacture of sex cells, a cell with two copies of each of our 23 chromosomes divides to produce cells with just one set of each. During this process, chromosomes can swap DNA (recombination); this involves a break in one chromosome that exposes a single strand of the normally double-stranded DNA. The single strand then finds a complementary strand in its partner chromosome. The two strands pair up to make a new double stranded bit of DNA. The sequence of the two strands may not, however, be perfectly complementary and might break the rules of DNA pairing (G should pair with C, and A with T). Mismatch repair enzymes then correct rule violations. Imagine a C mismatched with an A. There are two choices for repair: replace C with T or replace A with G. The system is biased and more commonly replaces A with G. More generally, it favours Gs and Cs over As or Ts. The repair bias may be an evolved property to cope with a high mutation rate of C toT.
Biased gene conversion” (BGC), as the process is termed, explains a general trend towards higher rates of evolution in chromosome domains that commonly undergo recombination, and correctly predicts the high rates of recombination in the superfast hotspots. Many such sites lie towards the ends of chromosomes, where recombination is common.
Importantly, BGC can drive mutations that are deleterious. Given that BGC can force deleterious mutations to spread through a population, part of the high rate of evolution in the hotspots could be because of the subsequent spread of compensatory mutations.
Hurst 2009 A positive becomes a negative Nature vol 457 pp 54304
- Martin Eastwood