NEW YORK — Identical twins harbor an epigenetic signature into adulthood that may reflect the twinning process, a new study has found.
Monozygotic twins, which represent less than 2 percent of births, do not tend to run in families, suggesting that the process is not genetic and may instead be random. But as twinning tends to take place early in development around the time of major epigenetic reprogramming events, researchers from Vrije Universiteit Amsterdam suspected that DNA methylation changes could be involved in the process.
Using DNA methylation array data from six different twin studies, the researchers compared methylation patterns in monozygotic (MZ) and dizygotic (DZ) twins. As they reported in Nature Communications on Tuesday, they found hundreds of sites that were differentially methylated between the groups, including a number of CpG sites within Polycomb-repressed regions or heterochromatin involved in cell adhesion or cell fate, hinting at a potential twinning mechanism.
"We interpret the findings as representing a molecular signature of the MZ twinning event that has been propagated, through many rounds of mitosis, to adult somatic tissues," Jenny van Dongen, an assistant professor at VU Amsterdam, wrote in an email. "This is a major step forward, but the exact mechanisms that lead to the formation of MZ twins require further functional studies."
Using a discovery cohort of 1,957 individuals — 924 monozygotic and 1,033 dizygotic twins — from the Netherlands Twin Register, the researchers conducted an epigenome-wide association study that uncovered 243 differentially methylated positions between the two types of twins. Following replication in the TwinsUK, Finnish Twin Cohort, Brisbane Systems Genetics Study, and the Environmental Risk Longitudinal Twin study cohorts, a meta-analysis of 5,723 individuals uncovered 834 differentially methylated positions, of which 497 had decreased and 337 had increased methylation.
Van Dongen added that even though there were suggestions that epigenetics could influence MZ twinning, "we were astonished to find such a strong and replicating epigenetic signature in somatic tissues from adult MZ twins and MZ [twin] children."
These differentially methylated positions were not randomly distributed, the researchers noted. For instance, hypomethylated sites were enriched near telomeres as well as in Polycomb-repressed regions. Meanwhile, hypermethylated sites were enriched near centromeres as well as in heterochromatin and ZNF genes and repeats. Both hypo- and hypermethylated sites were also found around CpG islands and intergenic regions.
Additionally, by combining these sites into a trained classifier, the researchers could predict whether or not someone was a monozygotic twin with an accuracy of about 80 percent.
Some of these differentially methylated positions overlap with transcription factor binding sites of genes involved in embryonic development and cell adhesion pathways. This finding suggested to the researchers that cell adhesion could be involved in the monozygotic twinning process, as it could affect the ability of an embryo's cells to dissociate and form twins.
However, they noted that additional functional research is needed. "We yet have to find out which of these methylation differences represent a cause, effect, or byproduct of the twinning event," van Dongen said, adding that her team also doesn't yet know how the methylation differences arise.
She and her colleagues are now examining additional DNA methylation data on MZ twins and triplets to see whether there is any link between the epigenetic signature they uncovered and chorionicity, or the number of placentas formed in a pregnancy. They further plan to examine when the epigenetic signature in MZ twins arises — as well as whether it segregates within pedigrees of families with MZ twins — and the steps in the MZ twinning process.