NEW YORK (GenomeWeb News) – Stretches of DNA with especially low levels of methylation are prone to genomic instability, providing fodder for structural changes that can contribute to evolutionary change and human disease, according to a study appearing online last night in PLoS Genetics.
Researchers from the Baylor College of Medicine and Warsaw University assessed methylome maps for a handful of sperm cells, together with information on population-derived human structural variation data and findings from past studies of human disease and genome evolution studies.
Consistent with previous findings, their analyses indicated that low-copy repeat regions of the genome are hotbeds for structural mutations occurring through a mechanism called non-allelic homologous recombination. But epigenomic features appeared to be even more closely linked to structural mutability, they found.
In particular, the team saw an enrichment of structural alterations in parts of the genome with low methylation in the sperm methylation maps. These regions also tended to contain rearrangements found in humans but not chimpanzees or other related primates, as well as loci linked to tissue-specific gene expression.
In addition, the team found that many of the regions of the genome previously reported to contain rare or de novo structural variations in schizophrenia, bipolar disorder, autism, and developmental delay overlap with the so-called methylation deserts detected in the sperm methylome maps.
Based on such patterns, they speculated that these stretches of muted methylation might mark more evolutionarily fertile parts of the genome.
"We have found that these rare and de novo structural variants, as well as changes of the human genome that have accumulated the fastest since the branching chimpanzee are significantly concentrated within methylation deserts," senior author Aleksandar Milosavljevic, a molecular and human genetics researcher at Baylor College of Medicine, said in a statement.
"[W]e found that hypomethylation, an epigenomic mark, forms a kind of 'highlight' on top of the genome, marking an evolutionary 'work in progress,'" he added. "What is in retrospect not so surprising is that these 'epigenomically highlighted' genes appear to code for brain development or function."
Going into the study, the researchers were aware of hotspots in the human genome where structural mutations are especially likely to occur. Some, but not all, of these hotspots are characterized by non-allelic homologous recombination involving low copy repeats, they explained, though some studies have hinted that these low copy repeats might sometimes be a consequence rather than a cause of structural mutability.
To look at whether epigenetic features might also contribute to hotspots of genomic instability, the researchers started by using low coverage Illumina GAII bisulfite sequencing to assess methylation profiles in two sperm cells.
They then came up with genome-wide sperm cell methylation maps based on the newly generated bisulfite sequence data, along with methylation information for two more sperm cells that had previously been bisulfite sequenced to 15 times coverage by another group.
With these sperm methylome maps in hand, the team was able to use computational methods to start classifying structural variation-prone sites in the genome based on whether they were associated with low methylation, non-allelic homologous recombination/low copy number repeats, or both.
For instance, in their subsequent analyses of copy number information on hundreds of individuals tested by custom array comparative genomic hybridization, as well as publicly available structural variation information, researchers found that structural mutations were more common in parts of the human genome with low methylation in the sperm cells tested.
These hypomethylated regions appeared to be more closely tied to structural mutability and human-specific rearrangements than the low copy number repeats, study authors explained, and follow-up analyses indicated that the hypomethylated stretches of sequence tend to house loci influencing tissue-specific gene expression.
Meanwhile, the team's analysis of CNV data generated by members of the International Schizophrenia Consortium suggested that disease-associated variants were enriched in parts of the genome with low levels of methylation.
Similarly, rare or de novo alterations found in previous bipolar disorder, autism, and developmental delay studies were more common in regions that the researchers had classified as hypomethylated in the sperm methylomes.
The study's authors argued that their results "suggest a connection between the epigenome, selective mutability, evolution, and human disease."
Still, they noted that more research is needed to tackle questions related to this process, such as potential variability in the hypomethylation-related process, the mechanism behind the enhanced mutability, and so on.