Regions of open and closed chromatin structure have recently been defined across the human genome [1]. Gilbert et al showed that regions of open chromatin are often gene dense and appear to correlate well with clusters of broadly expressed genes. They suggested that open chromatin fibre domains provide a chromatin environment more conducive to transcriptional activation. However, many genes are also found in regions of closed chromatin structure. This raised the question as to why would genes be maintained in closed chromatin if this meant they were simply less accessible for transcription. One possibility is that they need to be subject to especially tight transcriptional regulation, and that their aberrant or leaky expression in inappropriate cells cannot be tolerated. However, it has also been proposed that open chromatin structure may make the underlying DNA sequence more susceptible to DNA damage [2].
Although some studies have predicted that rates of mutation are relatively constant across mammalian genomes, analysis of human-mouse alignments has suggested that there may be as much as a 3-fold difference in substitution rates across chromosomes [3], with regions containing genes involved in extracellular communication displaying unusually high levels of synonymous substitutions [4]. Previous studies have also shown that, in mammals, genes within close genomic proximity undergo similar rates of neutral divergence and evolution [4-6]. For example, Williams and Hurst showed that the mean difference between the Ka values (substitution rate at non-synonymous sites) of 176 pairs of linked genes was significantly lower than would be expected by chance [5]. Similar results were also observed with Ks (substitution rate at synonymous sites) and Ka/Ks (often used to infer the mode and strength of selection). Consequently they proposed that the murid genome was split into domains of evolution. The reason for this was unknown, but it is possible that some aspect of chromatin structure over different genomic regions influences the rate of DNA damage or its repair.
The availability of a map of long-range chromatin structure across the human genome [1] allows us to assess this idea and, through the comparison of various measures of neutral variation, we have identified those forms of chromatin structure associated with the highest rates of background mutation.
Answers to all your Biology Questions
