NEW YORK (GenomeWeb News) – A team reporting online today in Science started tallying up sites in archaic hominin genomes where methylation marks differed from those in modern humans — regions that appear to be replete with disease-associated and developmental genes.
Researchers from Israel, Germany, and Spain used degradation profiles in ancient Neanderthal and Denisovan DNA to map methylated and unmethylated cytosines across the genomes of the two archaic hominins. By then comparing these methylation patterns with those found in modern humans, they narrowed in on roughly 2,000 regions in the genome that were differentially methylated in the archaic hominins.
"We reconstruct[ed] a full high-resolution map of the methylation along archaic humans (Neanderthals and Denisovans)," co-corresponding author Liran Carmel, a genetics researcher at the Hebrew University of Jerusalem, told GenomeWeb Daily News in an email message, explaining that this made it possible to compare epigenomic patterns in Homo sapiens and our closest evolutionary relatives for the first time.
"We hope that the differences in methylation discovered here will help to uncover the epigenetic basis for phenotypic differences between present-day and archaic humans," he and his co-authors wrote, "and shed light on the role of epigenetics in the recent evolution of our lineage."
Those differentially methylated sites included genes known for their role in development, such as genes from the HOXD cluster. The team also saw an over-representation of genes implicated in human disease amongst the differentially methylated regions.
The latter finding hints that epigenetic shifts in recent evolutionary history that produced functional changes involving certain genes could have unwittingly contributed to the emergence of some diseases through pleiotropic effects. But there are other potential explanations too, study authors noted, including the possibility that disease-associated genes might be more apt to undergo methylation shifts.
Several studies have been done to profile Neanderthal and Denisovan genome sequences and to pick up the genetic differences between present-day humans and archaic hominins, which belong to a branch that split from the modern human lineage an estimated 550,000 to 765,000 years ago.
But far less is known about the epigenetic and regulatory profiles present in the extinct hominins — something Carmel and his colleagues set out to do for the current study.
Using a DNA degradation-based methylation profiling method similar to that described in Genome Research late last year by researchers at the University of Copenhagen and elsewhere, the group reconstructed cytosine methylation patterns across the Neanderthal and Denisovan genomes.
In particular, the researchers took advantage of the fact that methylated versions of cytosine degrade to form thymine, while unmethylated forms of the DNA base eventually become uracil bases. Consequently, sequence differences in degraded DNA can provide clues to methylation modifications once found in the genome.
The team started by identifying methylation-prone CpG sites in the human genome where cytosine and guanine are frequently found together. From there, it looked at the proportion of these sites showing cytosine-to-thymine transitions in existing Neanderthal and Denisovan genome sequences, which were generated using DNA from bone samples from female representatives of each species.
"[W]e expect to see an appreciable fraction of [thymines at CpG positions] in methylated regions, but a negligible faction of [thymines] in unmethylated regions," Carmel said. "In fact, we found that the fraction of [thymines at CpG positions] is an excellent … proxy for the archaic methylation level."
The researchers also considered experimentally ascertained methylation profiles from bone-precursor cells of a human woman in the same age range as the deceased archaic humans.
After taking steps to verify their degradation-based methylation profiling method using data for a 4,000-year-old Eskimo individual, investigators went on to reconstruct methylation maps using the Neanderthal and Denisovan data.
In the process, they uncovered methylation profiles in the ancient hominin genomes that largely overlapped with those found in contemporary human DNA. But there were some differences as well.
Each of the archaic hominin genomes contained roughly 1,100 sites with methylation marks that differed from those found in modern humans. Though some of these differentially methylated regions, or DMRs, are expected to represent variations between individuals, others consistently differed between species.
When the researchers looked at the location of the DMRs — which were grouped into Neanderthal-specific, Denisovan-specific, human-specific, or unclassified DMRs — they found several in and around a set of HOXD genes known for their role in limb development. DMRs also turned up around a transcription factor known for regulating HOXD genes, amongst other regulatory sites in the genome.
Given that the inactivation of certain HOXD genes has been linked to phenotypes that resemble the limb features described in Neanderthal fossils, researchers suspect that the DMRs in the HOXD region may be behind at least some of the notable morphological differences between Neanderthals and modern humans.
Likewise, the over-representation of disease genes in DMRs is expected to continue offering clues about human evolution.
While the current analysis is based on data for individual Neanderthal and Denisovan representatives, Carmel noted that additional insights should become available as researchers continue sequencing ancient samples.
"As of today, there is only a single individual from each type of [archaic] human with a high-quality genome," he explained. "But, given the recent advances in ancient DNA sequencing, we indeed anticipate many more high-quality genomes coming out in the very near future."