2 Nov 2015
November 3, 2015. On the final day of the World Congress on Psychiatric Genetics, Tuesday, October 23, in Toronto, Canada, the morning plenary session featured a pre-recorded interview with Nobel Laureate Arvid Carlsson, who discovered dopamine's action as a neurotransmitter in the brain. A video conference followed, allowing the Toronto audience to interact with Carlsson, who gamely answered questions from Sweden about career planning, antipsychotics, and women in science. On antipsychotics, he blamed the stall in their development on misguided efforts to find highly potent and specific molecules, which resulted in adverse side effects. With this approach, "We are working like elephants in a porcelain shop," he said. As a counter example, he talked about a dopamine-stabilizing drug called OSU-6162, which normalizes dopamine levels only where they are disrupted. So far, his research on OSU-6162 suggests it as a promising therapy for people with mental fatigue following a stroke, and perhaps for narcolepsy.
A schizophrenia session followed, with six talks spanning a wide range of research. Daniel Howrigan of Massachusetts General Hospital in Boston led with an update on the largest effort to discover rare de novo mutations—those that occur spontaneously rather than being inherited from a parent—in schizophrenia in a large Taiwanese sample (see SRF related news report). Exome sequencing of the sample now includes 2,700 trios, each consisting of a person with schizophrenia and both parents. Howrigan reported that damaging loss-of-function mutations were slightly more frequent in schizophrenia than in controls. Some of these hit the same gene twice: 14 genes had recurrent mutations, which was greater than expected by chance. No single gene, however, was implicated with exomewide statistical significance. As there is not much left to sequence in this cohort, Howrigan suggested that combining data from other de novo studies may implicate single genes. These de novo variants, however, seem to make limited contributions to schizophrenia compared to autism (see SRF related news report).
Tristram Lett of Charite University Hospital in Berlin, Germany, explored the relationship between a genetic variant and gamma-aminobutyric acid (GABA) signaling, brain structure, and working memory, which is impaired in schizophrenia. Lett reported that a variant in GAD1, an enzyme that produces GABA, was associated with white matter structure and working memory performance in people with schizophrenia and controls. Inhibition in the dorsal lateral prefrontal cortex, as measured by a combination of transcranial magnetic stimulation with electroencephalography (TMS-EEG), mediated the association between the genetic variant and behavior, specifically in people with schizophrenia. Lett suggested that GAD1 genotype may eventually be used to stratify people for brain stimulation therapies geared toward normalizing inhibition.
Psychiatric genetics consortia continue to proliferate, and Alex Richards of Cardiff University in Wales described the European Union Gene Environment Interaction (EUGEI) collaboration, which aims to collect detailed phenotype information (symptoms, cognition, social variables, environmental risks) alongside genotype in people with psychosis. So far, the EUGEI dataset consists of about 7,000 samples from eight countries, resulting in a relatively genetically heterogeneous population. Using a custom-designed chip to probe common and rare variation, Richards reported that preliminary data were similar to the results obtained by the giant Psychiatric Genetics Consortium's (PGC) latest GWAS: 78 percent of variants showed the same direction of effect as the PGC's, and polygenic risk scores that sum the effect of all risk variants resulted in larger scores in psychosis cases than in controls.
Ahmed Al Amri of the University of Leeds in England returned to rare variants in his talk describing genetic insights from a consanguineous Pakistani family with multiple cases of schizophrenia. Homozygosity mapping, combined with exome sequencing, zeroed in on a mutation in the DFNB31 gene that seemed to act in a recessive manner: People with only one copy of the mutation were unaffected, but those with two copies had schizophrenia. DFNB31 encodes a scaffolding protein, and functional assays suggested that the mutation disrupts interactions between this protein and its binding partners.
The next talk focused on the rarest of the rare: the so-called "private" mutations that, by definition, are found in a single person and no one else. Because natural selection weeds out mutations of large effect, rarity can be used as a proxy for damaging mutations. Screening exome sequencing data for private mutations in a Swedish sample of over 12,000 people with schizophrenia and controls, Giulio Genovese of the Broad Institute of MIT and Harvard in Cambridge, Massachusetts, found that each person carried 0-100 private variants. Using various classifiers to estimate how damaging the mutations were, Genovese reported that people with schizophrenia had slightly more damaging private mutations than did controls. These mutations were enriched in gene sets already linked to schizophrenia, such as post-synaptic density genes or targets of the Fragile X mental retardation protein (FMRP), but this enrichment could be explained by how many constrained genes (meaning intolerant to mutation) existed in a gene set.
Emma Dempster of the University of Exeter in the UK ended the session with a look at epigenetic variation in schizophrenia in the form of methyl groups that regulate gene expression when added to DNA . Studying 75 monozygotic twin pairs discordant for schizophrenia, Dempster reported different patterns of methylation in subjects with schizophrenia compared to their unaffected twins. The most significant difference occurred over the HDAC4 gene, which encodes a histone deacetylase that controls transcription factor access to DNA. Similar methylation profiles were found in a separate sample comparing cases to controls, and Dempster noted that eight methylation signals overlapped with genomewide-significant loci identified by the PGC's latest genomewide association study, including CACNA1C, a calcium channel gene.—Michele Solis.