NEW YORK (GenomeWeb) – A Norwegian team has garnered evidence suggesting schizophrenia-associated variants tend to fall in parts of the genome that have been subject to DNA methylation changes since humans split from archaic hominins.
The researchers compiled summary statistics from genome-wide association studies focused on a dozen traits or conditions, including schizophrenia. When they set the GWAS association-enriched sites alongside the human, Neanderthal, and Denisovan genomes, they found that SNPs implicated in schizophrenia were over-represented in parts of the genome with differential methylation in the human genome relative to the archaic hominins.
That enrichment in human-specific differentially methylated regions (DMRs) was evident for variants contributing to schizophrenia, the team reported. The analysis also pointed to a similar pattern for variants involved in height, though the DMR enrichment detected there appeared somewhat more limited. The results were published today in BMC Evolutionary Biology.
"Our results suggest that SNPs in regions of the human genome that have undergone recent changes in DNA methylation status are enriched for association with schizophrenia, and to a lesser extent with height," senior author Stephanie Le Hellard, a clinical science researcher affiliated with the University of Bergen and Haukeland University, and her co-authors wrote.
For the analysis, the team tapped into published datasets from GWAS focused on schizophrenia, bipolar disorder, attention deficit hyperactivity disorder, rheumatoid arthritis, blood cholesterol profiles, height, body mass index, and other human traits or diseases.
That data was considered in parallel with DMRs identified by comparing methylation profiles in the human genome with DNA methylation maps for Neanderthals and Denisovans that researchers in Israel, Germany, and Spain reported in Science in 2014.
"While [99 percent] of the methylation maps were identified in the three hominids, nearly 2,000 differentially methylated regions were identified, which give the first clues about the role of epigenomics evolution in generating anthropometric differences between modern humans and their ancient cousins," Le Hellard explained.
Indeed, she and her colleagues found that the schizophrenia-associated SNPs were more apt to fall in human-specific DMRs — a pattern that persisted even when their analysis left out variants in the major histocompatibility complex and adjusted for the number of schizophrenia-associated SNPs that have been detected.
Conversely, investigators did not see ties between variants involved in schizophrenia risk and Neanderthal or Denisovan DMRs. And there was no clear overlap between human-specific DMRs and SNPs implicated in other diseases, though height-related variants showed a slight enrichment in these regions.
When they dug into the related genes falling in the human-specific DMRs, meanwhile, the researchers found that the regions were prone to contain schizophrenia risk variants in and around genes involved in neuronal signaling, nervous system function, and synaptic processes.
"Our results concur with previous genomic studies demonstrating that methylation changes in Homo sapiens have had the greatest impact on the nervous system," the authors wrote, "and provide evidence that epigenomics evolution plays a role in conferring a high risk of schizophrenia on humans."