May 27, 2014. The latest exome sequencing findings in schizophrenia cast suspicion upon SETD1A, a gene involved in chromatin remodeling. Published May 21 in Neuron, the study looked for the rare “de novo” mutations that spontaneously arose in 231 people with schizophrenia, but not in their parents. This revealed two protein-disrupting mutations in SETD1A in two different people with schizophrenia—a convergence almost unheard of for rare variants.
“It’s priceless,” said Maria Karayiorgou, who led the study with her colleague Joseph Gogos at Columbia University, New York City. Statistically, the two hits were unlikely to have happened by chance, and they represent only the second time two hits in the same gene were deemed significant. Earlier this year, two loss-of-function (LOF) mutations in TAF13, which encodes transcription machinery, were found (see SRF related news report).
Genes involved in chromatin remodeling were also fingered in an exome study published last month (see SRF related news report).
The new study builds on Karayiorgou and Gogo’s previous studies of exome sequencing in family trios, which have made a case for the involvement of de novo mutations in schizophrenia risk (see SRF related news report). The new study mines their sequence data again, looking specifically for “indels”—small insertions and/or deletions of a few nucleotides of DNA that can have damaging consequences for the resulting protein. Accurately identifying indels is tricky, however, because they are hard to correctly align to a reference sequence.
Karayiorgou’s group analyzed their sequences with new methods that more reliably identified indels and found five of them. “We always knew we had just scratched the surface for indels in the previous papers,” she said.
Two of the new de novo indels hit SETD1A in two different people. In one, a single nucleotide went missing, resulting in a frameshift; in another, two nucleotides were deleted at a splice site.
“I think this paper’s big strength is that it was able to identify two hits in one gene, which makes a much stronger case for that particular gene, and [is] worth pursuing at the biological level,” said SRF advisor Francis McMahon, a researcher at the National Institute of Mental Health in Bethesda, Maryland. McMahon was not involved in the study.
SETD1A encodes a catalytic subunit of an enzyme that adds methyl groups to histone proteins, which mark genes for transcription. This and other de novo LOF mutations from the sample strengthen the case for a role for genes involved in chromatin regulation in schizophrenia. In chromatin, DNA is tightly wound around protein spools called histones. When it’s time to transcribe a gene, the DNA is unlooped from the spool in a highly coordinated process to give transcription machinery access to a gene’s DNA.
Problems with chromatin regulation could affect the transcription of many different genes, which is particularly important early in brain development. In addition, they could alter the way that environmental influences impinge upon gene expression at any point in life. Autism studies also have turned up genes involved in chromatin remodeling.
In contrast to other studies, however, Karayiorgou’s group did not find that their LOF mutations preferentially hit genes encoding synaptic proteins (see SRF related news report). This discrepancy may stem from differences in the types of cases in the samples studied, or may reflect too small sample sizes.
“As the sample sizes grow, I expect we will start to see more convergence, maybe even across studies, with the same genes hit multiple times.” McMahon said, noting that the same thing has happened in autism and epilepsy studies.
Adding the five de novo indels to the point mutations reported in their previous studies, the researchers measured a 3.7-fold enrichment of de novo LOF variants in schizophrenia compared to controls; however, this was not statistically significant. They did find that cases with poor school performance in childhood were significantly enriched for this type of mutation.
The study also addressed, for the first time, the role of rare LOF variants in the inherited component of schizophrenia risk. With the exome sequences of both parents and offspring in hand, the researchers could track the transmission of LOF variants from parent to offspring. Though any given variant has a 50-50 chance of being transmitted, one that is transmitted to someone with schizophrenia more often than expected may be a culprit in the disorder.
The researchers found that LOF variants as a group were transmitted to cases and controls less than expected by chance—probably a sign of negative selection weeding these variants out. But this under-transmission was less pronounced for schizophrenia cases than controls, which suggests that cases are receiving more than their fair share of deleterious mutations.
“This burden of loss-of-function mutation could be a path toward the disease,” Karayiorgou said.—Michele Solis.
Takata A, Xu B, Ionita-Laza I, Roos JL, Gogos JA, Karayiorgou M. Loss-of-function variants in schizophrenia risk and SETD1A as a candidate susceptibility gene. 2014 Neuron 82, 773-780. Abstract