May 19, 2014. People with schizophrenia carry 3.5 times more spontaneously occurring point mutations than healthy controls do, according to exome sequencing results published online April 29 in Molecular Psychiatry. The researchers, led by Dick McCombie of Cold Spring Harbor Laboratory, New York, and Aiden Corvin of Trinity College Dublin, Ireland, were also able to show that the mutations they identified were likely to be damaging to the genes involved, rather than being innocuous.
The study implicates genes involved in chromatin remodeling (CHD8, MECP2, HUWE1), which suggests that the processes by which genes are made available for transcription go awry in schizophrenia. Several genes hit by mutations in the study have also been implicated in autism and intellectual disability, further underscoring genetic overlaps between schizophrenia and other neurodevelopmental disorders.
The study bolsters a role for new, damaging mutations in the protein-coding part of the genome in schizophrenia. The study’s convergence on chromatin remodeling also fits with the theme that disrupted brain development increases risk for schizophrenia (see SRF related news report; SRF news report). Chromatin is the compact package of DNA wound around proteins, which allows our very long genomes to fit inside a cell’s nucleus. To be expressed, a gene needs to be unwound from these proteins and sufficiently exposed so that transcriptional machinery can get to it. Problems with this process could derail the precise transcriptional choreography that goes on during brain development.
Because it is focused on the protein-coding parts of the genome, exome sequencing can point directly to genes with higher resolution than studies of common variants or copy number variation (see SRF related news report). But the promise of this approach has given way to some dismay, as sequencing turns up plenty of suspicious mutations in cases and controls alike. This means that, after all the hard work of sequencing, geneticists are faced with the task of coming up with principled, unbiased ways of discerning which variants contribute to a disorder (MacArthur et al., 2014).
This problem is compounded for schizophrenia, which is thought to arise through many different genetic variants, some of which may be very rare. To bolster the case for the rare mutations found in the study, first author Shane McCarthy and colleagues used tools that score genes on how sensitive their protein products are to mutation, that is, whether their function would tolerate mutation (Petrovski et al., 2013) or whether loss of a gene copy would be noticed in the phenotype (Huang et al., 2010).
Categorizing de novos
Like previous studies, the new study focused on “de novo” variants, which are more likely to be deleterious because they arise anew in sperm or egg cells and so haven’t yet been weeded out by natural selection. The researchers found these by sequencing the exomes of 57 people with schizophrenia and both their parents. Fifteen of these trios were deemed “familial” cases, in that there was a history of psychosis in first- or second-degree relatives; the remaining majority were termed “sporadic.” With over 90 percent of the exome sequenced 10 times or more, a total of 59 de novo variants turned up, 47 of which occurred in sporadic cases. The mutation rates were within the range reported for de novo mutation in other studies (e.g., 1.62 x 10-8 per base per generation for sporadic cases).
Most of these were “missense” mutations, in that they coded for a different amino acid (28 of 47 of the mutations were found in sporadic cases). But the more damaging kind of “nonsense” mutations were also identified: Five of 47 mutations encoded a premature stop codon, which presumably truncates the resulting protein at a different location. Though these five mutations may not seem like much, they constituted a statistically significant, 3.5-fold increase over the number of de novo, nonsense mutations found in controls in autism studies (see SRF related news report). Nonsense mutations were not identified in the familial trios.
This increased rate of de novo, deleterious mutation looks pretty serious on its own, but the researchers looked for further evidence with methods that score genes on how sensitive they are to mutation. One method scored genes on haploinsufficiency, meaning how necessary it is to have two working copies of a gene to maintain normal function (Huang et al., 2010). The researchers found that the genes hit by nonsense variants were more likely to be rated as haploinsufficient (i.e., unable to get by with only one working copy) than comparable genes in the genome. Another method underscored the sensitivity of these genes to mutation, as four were in the top tier of “intolerant-to-variation” scores (Petrovski et al., 2013).
Of the genes hit by nonsense mutation, the researchers highlighted CHD8, AUTS2, and MLL2. Both CHD8 and MLL2 are chromatin modifiers, and more involvement with chromatin was found among missense mutations in the Rett syndrome genes MECP2 and HUWE1, both transcriptional regulators.
Several of these genes have also been flagged in studies of autism, including CHD8, AUTS2, MECP2, and HUWE1. Using the NEUROCARTA database, which links genes with phenotypes to explore this more systematically, the researchers found that the genes hit by all protein-altering mutations in schizophrenia were over-represented in genes linked to autism and intellectual disability. This echoes other genetic and epidemiology studies, which find shared risk among these disorders (see SRF related news report; SRF news report; SRF news report).
While other exome findings are sure to come in the future, this study shores up a role for new, damaging mutations in schizophrenia. It also provides a meditation on the importance of epigenetic factors in risk. Understanding how transcriptional mishaps during brain development reverberate throughout brain circuitry may yield clues to the beginnings of schizophrenia.—Michele Solis.
McCarthy SE, Gillis J, Kramer M, Lihm J, Yoon S, Berstein Y, Mistry M, Pavlidis P, Solomon R, Ghiban E, Antoniou E, Kelleher E, O'Brien C, Donohoe G, Gill M, Morris DW, McCombie WR, Corvin A. De novo mutations in schizophrenia implicate chromatin remodeling and support a genetic overlap with autism and intellectual disability. Mol Psychiatry. 2014 Apr 29. Abstract