Exploring the Epi-signature
Dr. Bekim Sadikovic updates us on the progress of DNA methylation as a diagnostic tool
By Luke Turner
“Do patients with intellectual disabilities have identifiable and specific changes in their epigenomes – and can we find the answer in peripheral blood?” This is the question that Bekim Sadikovic, Head of Molecular Genetics at London Health Sciences Centre (LHSC) in Canada, posed in a recent article on the potential diagnostic power of DNA methylation (1). At the time, evidence showed that significant genome-wide DNA methylation alterations could be detected in peripheral blood, providing hope that DNA methylation tests could determine whether a variant of unknown significance is pathogenic. But, 18 months on, how much progress has Sadikovic – and the field as a whole – made?
His latest research applies the technique to DNA samples from patients with neurodevelopmental and congenital anomalies who lacked a definitive diagnosis (2). “Currently, the more sophisticated genetic analyses produce a 30–35 percent diagnostic yield for patients who present with developmental disabilities, leaving many cases unexplained,” Sadikovic says. “When we applied genome-wide DNA methylation analysis, we were able to diagnose a significant number of additional patients in whom we identified an underpinning genetic condition based on evidence of an epigenetic defect.”
The team has demonstrated that genetic defects can be manifested either as a result of, or in association with, genome-wide DNA methylation patterns. When these methylation defects occur across multiple loci, they are referred to as epi-signatures. “We can compare these epi-signatures to our reference database, which contains data for various conditions, to identify underlying genetic defects,” explains Sadikovic.
“We’re reaching a plateau in terms of what we can do using genomics and DNA sequencing – but what we’re now doing is moving the technology from research into the clinic. ”
In the new study, a computational model facilitated the concurrent detection of 14 syndromes with over 99 percent accuracy. Across 965 undiagnosed patients, the model identified 15 subjects with syndromic Mendelian disorders, 12 with imprinting and trinucleotide repeat expansion disorders, and 106 with rare epi-variants. “For the last couple of decades, genomics has been at the forefront of molecular diagnostics. We’re reaching a plateau in terms of what we can do using genomics and DNA sequencing – but what we’re now doing is moving the technology from research into the clinic.”
As the first site in the world to offer this form of genetic testing, Sadikovic hopes that the work at LHSC sets a precedent. “The technology is easily accessible and most labs already have the ability to run genomic testing. The challenge will be interpreting the data. We have already built large reference databases to map epi-genetic signatures – and hopefully these continue to grow as the technology becomes more widespread.”