22 Mar 2016
March 23, 2016. Two studies have converged on the same gene, SETD1A, as a major risk factor for schizophrenia. One study surveyed the protein-coding exomes of over 60,000 people and found that disabling mutations in SETD1A escalated risk for schizophrenia 38-fold. The other study implicates synonymous, or "silent," mutations, which alter the DNA sequence but not an amino acid in the resulting protein, in schizophrenia. One of these silent mutations hit SETD1A.
The results highlight transcriptional regulation as a wellspring of risk for schizophrenia and other neurodevelopmental disorders. SETD1A (also known as KMT2F) encodes an enzyme involved in methylating the histone proteins around which DNA is coiled. Changes to histone methylation can alter which genes are ultimately exposed for transcription. Problems with SETD1A, then, could result in genes transcribed at the wrong time, in the wrong place, as the wrong version, thus disrupting the brain's developmental program.
Pathway analysis of clues from an earlier version of the Psychiatric Genome Consortium's genomewide association study (GWAS) of schizophrenia has also pointed to histone methylation as a process of note (see SRF related news report). This suggests that although SETD1A mutations affect only a tiny fraction of people with schizophrenia, they may have relevance to the other cases.
"The fact that the same bit of biology is coming up in the relatively mild perturbations seen in GWAS and the really catastrophic perturbations in SETD1A in exome sequencing suggests that the gene might be relevant in thinking about new treatment avenues for everyone with schizophrenia," said Jeffrey Barrett of the Wellcome Trust Sanger Institute in Cambridge, UK, who led the large-scale exome sequencing project published on March 14 in Nature Neuroscience. "That's the hope."
This sentiment was echoed by Bin Xu, who was involved in the silent mutation study published in Neuron on March 6, which was led by Maria Karayiorgou at Columbia University, New York City.
"These findings altogether suggest that abnormalities in epigenetic regulation could be important for [the] etiology of schizophrenia in general," Xu wrote in an email to SRF.
Karayiorgou and colleagues first fingered SETD1A as a risk factor for schizophrenia in 2014 (see SRF related news report). They reported two protein-disabling "loss-of-function" mutations in two different people with schizophrenia in a screen of de novo mutations, which occur spontaneously rather than being inherited from a parent. Finding mutations in the same gene was notable, because such a "pile up" had mostly eluded researchers in earlier exome sequencing studies (see SRF related news report).
This suggested that loss-of-function mutations in schizophrenia are very rare and distributed across many genes, necessitating much larger studies to find multiple hits to the same gene.
"People had done some good and big studies before, but none of them were quite big enough to really push any one gene into the absolutely statistically certain camp," Barrett told SRF.
Largest exome screen so far
Barrett's study combined data from new exomes sequenced by his group—about 2,000 with schizophrenia and 7,000 controls—and other previously published sequences, including some de novo datasets. First author Tarjinder Singh and colleagues found 10 loss-of-function mutations in SETD1A out of nearly 8,000 exomes from people with schizophrenia. In contrast, they found only two loss-of-function mutations in nearly 60,000 controls, including 45,000 from the Exome Aggregation Consortium (ExAC) database. The resulting odds ratio was 38, and although the confidence interval was wide, the effect was definitely big, on par with 22q11.2 deletions associated with schizophrenia.
Xu said that he and his colleagues were "quite happy" to see their earlier SETD1A finding supported on such a large scale.
The 10 people with SETD1A mutations had the classic combination of positive and negative symptoms found in schizophrenia. Intellectual functioning or educational attainment data were available for seven of these subjects, and all had evidence of some intellectual disability or limited educational achievement, again paralleling people with 22q11.2 deletions.
These cognitive effects prompted the researchers to look for SETD1A mutations in children with undiagnosed developmental delay, where they found another six mutations. These children may be at risk for schizophrenia as they grow older.
Splice site suspects
The second paper suggests that synonymous mutations could be at work in schizophrenia and autism alike. While these do not alter the amino acid sequence of a protein, they can change how a gene is transcribed: Transcripts may become less stable, or splice sites that affect how RNA is processed may change, or transcription factors may not get a good grip onto a gene. Exome sequencing covers DNA at the boundaries between exons and introns, revealing sites acted upon by regulatory elements. Evidence for synonymous mutations in human disease has so far come from cancer (Supek et al., 2014).
First author Atsushi Takata and colleagues found that silent mutations near splice sites were particularly enriched in 1,043 autism cases (OR = 1.96) and in 1,021 schizophrenia cases (OR = 1.53) compared to 731 controls; in contrast, silent mutations far from the splice sites were not enriched. Narrowing in on mutations landing in regions known to be involved in splicing gave an even stronger enrichment (OR = 2.52 for autism and OR = 1.89 in schizophrenia), which was confirmed in two larger autism datasets.
Synonymous mutations also hit genes already implicated by de novo loss-of-function mutations. This convergence just missed genomewide significance for RAB2A in autism but reached criteria for SETD1A in schizophrenia.
"The additional support for the role of SETD1A in psychiatric disorders was particularly satisfying," Xu wrote. "Of course, we need to proceed with caution because the effect of SETD1A mutations has not been directly tested experimentally."
Karayiorgou's group estimates that these silent mutations have a similar impact as the de novo protein-disabling mutations they had found before. However, Barrett is not yet convinced of this. He noted that a preliminary analysis of his dataset did not find a similar enrichment of splice site-synonymous variants. On the other side of the coin, they found that control exome data from the ExAC database had slightly fewer than expected—something consistent with a role for this type of mutation in disease, in general at least.
"This does not mean that there is no effect of synonymous mutations, but the extent of their influence may not be as substantial as reported here," Barrett said.—Michele Solis.
Takata A, Ionita-Laza I, Gogos JA, Xu B, Karayiorgou M. De Novo Synonymous Mutations in Regulatory Elements Contribute to the Genetic Etiology of Autism and Schizophrenia. Neuron. 2016 Mar 2;89(5):940-7. Abstract
Singh T, Kurki MI, Curtis D, Purcell SM, Crooks L, McRae J, Suvisaari J, Chheda H, Blackwood D, Breen G, Pietilainen O, Gerety SS, Ayub M, Blyth M, Cole T, Collier D, Coomber EL, Craddock N, Daly MJ, Danesh J, DiForti M, Foster A, Freimer NB, Geschwind D, Johnstone M, Joss S, Kirov G, Korkko J, Kuismin O, Holmans P, Hultman CM, Iyegbe C, Lonnqvist J, Mannikko M, McCarroll SA, McGuffin P, McIntosh AM, McQuillin A, Moilanen JS, Moore C, Murray RM, Newbury-Ecob R, Ouwehand W, Paunio T, Prigmore E, Rees E, Roberts D, Sambrook J, Sklar P, Clair DS, Veijola J, Walters JT, Williams H; Swedish Schizophrenia Study; INTERVAL Study; DDD Study; UK10 K Consortium, Sullivan PF, Hurles ME, O'Donovan MC, Palotie A, Owen MJ, Barrett JC. Rare loss-of-function variants in SETD1A are associated with schizophrenia and developmental disorders. Nat Neurosci. 2016 Mar 14. Abstract