NEW YORK (GenomeWeb) – Using a screen they developed, researchers from New York University and elsewhere have uncovered mosaic mutations associated with two early-onset genetic diseases.
Mosaic mutations arise spontaneously at low rates during development and may then be passed on to a person's children. As mosaic mutations haven't undergone purifying selection — much like de novo mutations — the researchers led by NYU's Orrin Devinsky suspected that they might be enriched for pathogenicity.
Devinsky and his colleagues developed a screen for such mutations, which they then applied to two cohorts: nine cases of sudden unexplained death in childhood (SUDC) and 338 cases of epileptic encephalopathy. As the researchers reported this week in Genetics in Medicine, they uncovered mosaic mutations in both cohorts, two of which appeared to be likely disease causing.
"These results highlight explicit screening for mosaic mutations as an important complement to the established approach of screening for de novo mutations," Devinsky and his colleagues wrote in their paper.
One of the nine SUDC families consisted of two healthy parents, an asymptomatic son, a daughter who was the SUDC proband, and a son with Dravet syndrome. The researchers suspected that the proband and the son with Dravet syndrome — both of whom suffered febrile seizures — might share a causal variant. But screens for recessive variants shared by the siblings came up empty.
Instead, Devinsky and his colleagues turned to a mosaic mutation transmission screen that, using next-generation sequencing data from a parent-child pair, identifies nonsynonymous variant calls that are heterozygous in the proband and observed among a parent's reads, yet absent in controls. These mosaic variants show an about 50 percent alternative allele ratio in the proband, but less than 50 percent in the mosaic parent of origin.
Applying this approach to the SUDC families led the researchers to uncover 606 nonsynonymous singleton variant transmissions in 17 parent-child pairs. Two of these variants — in SCN1A and NTNG1— met the researchers' expectations for allele read ratios in the proband and parent.
SCN1A, Devinsky and his colleagues noted, encodes a voltage-gated ion channel protein that is involved in generating and propagating action potentials in the nervous and neuromuscular systems, while NTNG1 encodes a protein that affects axonal guidance and signaling.
The SCN1A variant, they added, was likely a mosaic mutation transmitted paternally. Of the 167 aligned reads from the normal father at this position, 42 — about 25 percent — carried the variant, less than would be expected for a heterozygote. The SUDC proband and her brother with Dravet syndrome, however, carried this variant in 47 percent and 46 percent of reads, respectively, indicating they are heterozygotes.
The researchers confirmed the father's mosaicism, the affected siblings' heterozygosity, and the mother's and unaffected sibling's lack of this variant using Sanger sequencing.
Evidence of mosaicism of the NTNG1 variant, meanwhile, was insufficient, and its discovery in a healthy sibling cast doubt on its pathogenicity, the researchers added.
The death of the SUDC proband was likely seizure-related, the researchers surmised. Most febrile seizures are convulsive, they noted, and convulsive seizures precede most sudden epilepsy deaths.
Devinsky and his colleagues also applied their screen to a larger cohort of 676 parent-child pairs from the Epilepsy Phenome/Genome Project to find 13,142 nonsynonymous singleton variant transmissions, four of which met the researchers' criteria for being potential mosaic mutations. One of those variants, they noted, is located in an intolerant gene, SLC6A1. The variant was confirmed in the proband and confirmed as mosaic in the parent using Sanger sequencing.
SLC6A1 encodes a Υ-aminobutyric acid transporter, and knocking out the gene in mice leads to absence seizures. The researchers also noted that recurrent mutations in this gene had previously been found in a large case cohort of myoclonic-atonic seizures and that mutagenesis at the mutation site reduces transporter activity to some 10 percent of the wild type.
This, they said, suggests that SLC6A1 dysfunction contributes to epileptic encephalopathy pathogenesis.
Within the 693 parent-child pairs the researchers examined, they uncovered six mosaic mutation transmissions, a mutation incidence of some 0.009 detectable transmissions per pair or about 0.018 per trio.
"[O]ur yield suggests that a formal screen, alongside screens for de novo mutations, could become a routine part of identifying candidate pathogenic variants in clinical genetics environments," Devinsky and his colleagues wrote.