Researchers ID Disease-Causing Mutations in the Exome That Alter Gene Splicing

June 5, 2017
Researchers ID Disease-Causing Mutations in the Exome That Alter Gene Splicing
Source: dem10/ Getty Images

Diana Kwon, Contributing Editor

In recent years, DNA sequencing has become faster, easier, and cheaper. However, pinpointing disease-causing variants in an individual’s genome remains a challenge. “Our validation technologies—our means of analyzing variance—is actually very low throughput,” said William Fairbrother, Ph.D., a biology professor at Brown University. “So we were looking to develop technology that would match the throughput of the genotyping technologies.”

To help identify disease-causing variants in the genome, Dr. Fairbrother and his colleagues created MaPSy (Massively Parallel Splicing Assay), a tool that works by producing artificial “minigenes” that either contain disease-causing mutations or the corresponding “normal” versions of the sequence. These synthesized genes are then tested both in vivo in living cells and in vitro by examining the effects of splicing on synthesized RNA.

“The main strength of the MaPSy approach is that they couple both in vivo and in vitro massively parallel reporter assays to identify which part of the spliceosome is disrupted by each tested mutation,” said Alex Rosenberg, Ph.D., a postdoctoral researcher in bioinformatics at Washington University who was not involved in the work.

In a study published April in Nature Genetics, the team used MaPSy to examine 4,964 diseases-causing exonic mutations from the Human Gene Mutation Database, a collection of genetic variants that lead to inherited diseases. The analysis revealed that around 10 percent (around 500 alleles) altered splicing both in vivo and in vitro. “That’s by pretty stringent criteria—we demand that there’s at least 1.5-fold difference in the representation of an allele to call it a splicing mutation,” Dr. Fairbrother noted. “It [also] has to happen in two different systems. So we might be undercounting that number.”

Some exons, the researchers found, had a higher proportion of splicing-associated mutations than others. “Some [exons] were robust and everything worked great, but others were sort of borderline—so if there was no mutation, everything was fine and you’d splice normally, but many mutations would disrupt splicing,” Dr. Fairbrother said. In addition, he added, there were certain features of those exons made them easier to disrupt than others. “It was [generally] things that made the exon a worse substrate for splicing,” he said. They also found that splicing mutations were more likely to appear in haploinsufficient genes, where losing on functional copy leads to disease. 
“In addition to identifying many disease-causing mutations that likely act through altered splicing, the work also provides strong evidence that some exons are especially prone to splice-altering mutations,” Dr. Rosenberg noted.

To further validate MaPSy’s results, the team also tracked down patient tissues samples from the original clinical studies that reported the genetic mutations. Though they were only able to track down samples linked to 32 of the MaPSy-detected exonic splicing mutations, they were able to validate 81% of those variants by conducting RNA splicing analyses on the tissues. 
“The type of massively parallel reporter assay used in this work is a powerful tool towards uncovering pathogenic variants that alter splicing,” Dr. Rosenberg said. “I expect that this approach and similar ones will be scaled up to test even larger sets of potential splice-altering variants.”

A few years ago, Dr. Fairbrother and his colleagues created Spliceman, a freely available web-based tool for predicting how likely mutations in DNA sequences are to cause splicing errors.

Currently, the team is looking for ways to improve MaPSy. One of their goals, Dr. Fairbrother noted, is to synthesize longer regions of DNA—presently, the system can only test mutation in exons that are less than 100 nucleotides long. They also hope to expand that assay to test more mutations at once. “In this study, we tested 5,000 mutations,” he said. “Now, we’re testing around 20 to 30 thousand variants.”