When Wellcome Sanger Institute geneticist Eugene Gardner set out to look for a specific type of genetic mutation in a massive database of human DNA, he figured it’d be a long shot. Transposons—also known as jumping genes because they can move around the genome—create a new mutation in one of every 15 to 40 human births, but that’s across the entire 3 billion base pairs of nuclear DNA that each cell carries. The sequencing data that Gardner was working with covered less than two percent of that, with only the protein-coding regions, or exons, included. Doing a quick calculation, he determined that, in the best-case scenario, he could expect to find up to 10 transposon-generated variants linked to a developmental disease. And “we really might get zero,” he says. “This whole thing might be for naught.”
But Gardner had recently developed the perfect tool to find the sort of de novo mobile element insertions that come about as a result of transposon movements and are often overlooked in genetic screens and analyses. As a graduate student in Scott Devine’s lab at the University of Maryland, Baltimore’s Institute for Genome Sciences, he had spent many hours making the software for the mobile element locator tool he dubbed MELT. The program was easy to use, so when Gardner moved across the Atlantic for a postdoc in Matthew Hurles’s lab at Sanger near Cambridge and gained access to a database of exomes from 13,000 patients with developmental disorders, he figured running the tool was worth a try.
There is tremendous value for these families that get a diagnosis.—Dan Koboldt, Steve and Cindy Rasmussen Institute for Genomic Medicine, Nationwide Children’s Hospital
Of those 13,000, Gardner focused on 9,738 people in the Deciphering Developmental Disorders (DDD) study whose parents’ exomes had also been sequenced, making it easier to single out variants present in the child but not in mom and dad. And as it turned out, he did get some hits. MELT picked up 40 potentially transposon-generated variants, which Gardner sat down at his computer to review using the raw sequencing data. Nine appeared to be true de novo mobile element insertions. “I remember being in my desk doing the visualization of all the putative de novo variants after I got the first results off the pipeline,” he recalls. “I remember being excited: I think I might have found a diagnostic de novo!”
Discussing the literature on the genes affected by such insertions with clinicians and other colleagues, Gardner narrowed the list down to four insertions found in genes that may be causing or contributing to four different patients’ disorders. He sent these results off to the physicians who had referred each of the patients to the database, and all the doctors confirmed that the results made sense to them given what had been published on those genes and what they knew about other cases involving patients with mutations in the same sequences. In one case, the physician had already linked the patient’s disorder to the gene Gardner had identified; in the other three cases, the patients were still undiagnosed.
“There is tremendous value for these families that get a diagnosis,” says human geneticist Dan Koboldt, who has collaborated with Hurles in the past and has used MELT in his studies of rare disease at the Steve and Cindy Rasmussen Institute for Genomic Medicine at Nationwide Children’s Hospital in Columbus, Ohio, but who was not involved in Gardner’s recent study. A genetic answer not only can help physicians connect patients to appropriate medical and counseling resources; it puts an end to the diagnostic odyssey that families affected by rare disease often endure.
What’s more, the finding of four potentially causative hits out of the nearly 10,000 cases provides first estimate of how commonly such mobile element insertions underlie developmental disorders. “What’s interesting about this study is that it’s taking a very broad approach,” says Ian Adams, a developmental biologist at the University of Edinburgh’s MRC Human Genetics Unit who was not involved in the research. Rather than look for transposon activity in a specific disorder, “it’s casting a much broader net in trying to find what type of diseases this class of mutations could be contributing to.”
This approach is important, agrees Adams’s MRC Human Genetics Unit colleague Jose Garcia-Perez, a transposable elements expert who was also not involved in the new research. In the last few years, two studies have used a tool developed around the same time as MELT to search for de novo mobile elements in people with autism spectrum disorder, but neither identified any that were likely to be responsible for the patients’ symptoms. “[Gardner’s] study shows that, no matter what’s [been found] recently, it’s something that should be explored in further detail in the future,” says Garcia-Perez. “[The study] actually shows a real connection between . . . transposition with that particular [type of] disorder.” Koboldt adds: “The reason this is an important study is that it establishes [that these] variants do occur and [that] they can be pathogenic.”
Gardner says he hopes that his methods can be used to explore other diseases, from both a research and a clinical perspective. Adams says MELT does appear to be “widely applicable to other datasets.” Such a tool could be a boon to research on transposons, given that their movements are often missed by normal screening tools, Adams adds. “I think [MELT is] something that could be readily built into existing pipelines.”
Jef Akst is managing editor of The Scientist. Email her at email@example.com.