Behavioural and neurophysiological effects of schizophrenia risk genes: a multi-locus, pathway based approach

Schizophrenia is a major mental disorder, which affects between 0.5 and 1% of the population and has substantial impact on patients' wellbeing and life expectancy. Although medical and psychosocial interventions can bring some relief, the causes of the disease are poorly understood and no causal treatments are available. Because schizophrenia is highly heritable one hope is that identification of genes that contribute to disease risk will help elucidate the causal mechanisms and thus enable the development of new treatments. Over the last years, large scale genetic studies, many of which were led by investigators from Cardiff, have identified hitherto unknown genetic variants associated with schizophrenia. It has emerged from this work that most patients with schizophrenia probably carry many risk variants, each of which only contributes a small amount to disease risk, and different genes and thus different aspects of the brain's biology may be affected in different patient groups. Understanding the effects of these genes on the brain will likely reveal more about the causal mechanisms of schizophrenia and also whether subgroups of patients need to be diagnosed and treated in different ways ("stratification"). However, very little is yet known about the effects of these variants on brain and behaviour and how they contribute to clinical symptoms. We therefore plan to investigate these effects, bringing together expertise in the molecular genetics of schizophrenia and in the investigation of brain substrates of cognition and psychological symptoms.
In order to fulfil our aim of elucidating the effects of schizophrenia risk variants, alone and in combination, we will conduct brain scanning with magnetic resonance imaging (providing high spatial resolution) and magnetoencephalography (providing high temporal resolution) on 200 healthy individuals. The reason for investigating healthy individuals rather than patients in the first place is that these genetic variants are common in the healthy population as well, and here they can be studied without the confounding effects of medication and other consequences of suffering from an illness. We will recruit participants from an existing cohort of 10000 young adults in the Bristol area (the ALSPAC cohort) who are representative of the general population and have been followed up from birth. Genetic information is already available from this cohort, and we can thus recruit 100 individuals each with high and low genetic risk for schizophrenia. We will investigate imaging measures that have previously been associated with abnormal brain processes in schizophrenia and tap into core cognitive functions such as learning and memory. We will also obtain detailed cognitive profiles of these individuals. Our expectation is that individuals with high genetic risk scores for schizophrenia will show similar abnormalities to patients on these measures. We also expect to see genes that affect specific biological pathways, for example brain development, to be associated with particular imaging measures, for example those probing the connections between brain areas. This information would help us identified biological subtypes of schizophrenia. The next step would be to validate these measures in patient groups, for which we have lined up collaborations with groups who already have relevant data from patient studies.
Potential applications of this work include the stratification of patients with schizophrenia into different biological groups, which could be tested for different responses to particular drugs in clinical trials. This could improve the targeting of existing medication. Another application, based on the better understanding of the biological pathways affected by the genetic risk variants would be to suggest brain mechanisms that might be targeted by new drugs.

Our research is based on the recent discovery of several genetic risk variants for schizophrenia and the observation that most cases of schizophrenia carry several (and only partly overlapping) common variants that each only contribute a small amount to disease risk. Because little is know about the downstream effects of these variants, we combine neuroimaging, neuropsychology and molecular genetics to elucidate their effects on brain and behaviour. Because power has been an issue in previous studies in this area, we have developed a new approach using healthy participant groups pre-selected for their polygenic risk scores from an existing genotyped cohort (ALSPAC). We will compare high and low genetic risk individuals on imaging and behavioural measures that have been proposed as candidate markers for schizophrenia. Examples include measures of reward sensitivity, associative learning, oscillatory brain activity, cognitive inhibition and maturation of white matter. We hypothesise that high risk individuals will show changes similar to those observed in schizophrenia patients on these measures. More specifically, though, we expect risk variants that cluster along specific biological pathways, for example those of synaptic plasticity or myelination, to be associated with some of these intermediate phenotypes but not with others. Our study is powered to detect these associations. We will also probe, in an exploratory fashion, the effects of single risk variants on the full range of neural and behavioural phenotypes acquired. Our research will provide an answer to the question whether intermediate phenotypes proposed by the current biological models can index genetic risk of schizophrenia, and whether they provide evidence for different neural effects of genes on different pathways. If the expected results are obtained, this could pave the way for a future biological stratification of schizophrenia, and for the identification of new pathways for drug development.

In addition to the academic beneficiaries, which include basic and clinical researchers in the wider fields of neuropsychiatric genetics, bioinformatics, psychiatric neuroimaging and biological models of schizophrenia, we expect interest into our research and its further application to patient populations from
a) drug companies working on the translation of genetic findings into biological models for the discovery and screening of new compounds for psychotic disorders
b) clinicians and policy makers working on new classification systems and diagnostic guidelines (especially with a view towards incorporating biological/ aetiological criteria) for psychiatric disorders
c) clinicians and clinical researchers seeking biomarkers for treatment prediction and stratification for clinical trials.
Any association between risk variants/ polygenic scores confirmed in our study can be taken into patient research, and, if replicated, will yield biomarkers that can tap into different components of biological risk and thus contribute to the identification of disease pathways. Specifically, future patient studies can use these biomarkers to test for their diagnostic value in comparison between patients and controls, for their prognostic value in high-risk individuals (combining genetic and clinical risk according to present prodromal criteria), and for their value in treatment prediction in stratification studies. Ultimately, our findings can thus contribute to drug development (a), and provide a basis for biological classification (b) or stratification (c). In addition to these downstream impacts on industry and health services, we will also make immediate contributions to the UK's skills base in biomedical research, by developing novel bioinformatics tools (e.g. for polygenic genetic imaging and behavioural genetics analysis) and training a new generation of investigators in their use.