Exome array analysis suggests an increased variant burden in families with schizophrenia
Introduction
Schizophrenia (lifetime risk ~ 1%) is a potentially disabling disorder with a multifactorial aetiology comprising a major genetic contribution (broad-sense heritability estimates range from 60 to 80% (Lichtenstein et al., 2009, Sullivan et al., 2003) and a host of environmental factors, ranging from neurodevelopmental insults to psychosocial adversity (Matheson et al., 2013). The genetic underpinnings of schizophrenia are likely to be complex, and involve genetic variation across the frequency spectrum - from rare, highly penetrant variants such as copy number variants (CNVs) to common single nucleotide polymorphisms (SNPs) with individually small effect sizes.
The largest genome wide association study (GWAS) to date, conducted by the Psychiatric Genetic Consortium, identified 108 loci associated with schizophrenia at genome-wide significance (Ripke et al., 2014). Studies from the same consortium have shown that schizophrenia is highly polygenic (Bulik-Sullivan et al., 2015b), and composite risk profile scores account for ~ 7% of variation on the liability scale.
Sequencing studies of schizophrenia have so far been underpowered for the detection of very rare variants showing study-wide significant association. Exome sequencing of unaffected parent-singleton affected offspring trios has identified a higher burden of de novo variants in sporadic cases relative to expected rates, or rates in unaffected relatives (Fromer et al., 2014, Girard et al., 2011, Xu et al., 2011). Furthermore, case-control studies have revealed an increased burden of rare, likely pathogenic variants in cases versus controls (Kenny et al., 2014, Need et al., 2012, Purcell et al., 2014). Sequencing studies in multiplex families have focussed on rare, highly penetrant causative variants segregating with schizophrenia and reported protein-altering variants in 3 genes associated with function of the NMDA receptor – GRM5, PPEF2 and LRP1B (Timms et al., 2013); in the SHANK2 post-synaptic density scaffolding protein and the SMARCA1 transcriptional regulator (Homann et al., 2016) and in UNC13B, (Egawa et al., 2016), which encodes a protein involved in the priming step of synaptic vesicle exocytosis.
An intermediate step between GWAS and sequencing studies is to investigate so-called ‘goldilocks’ variants (Price et al., 2010) – variants common enough to be detected in population samples of a reasonable size and rare enough to be under purifying selection (minor allele frequency, MAF > 0.5% & ≤ 2%). Simulations for complex traits suggest that this class of variant may be quite frequent, with at least one variant present in up to one-eighth of genes affecting the trait (Price et al., 2010). The exome array was designed to target likely deleterious, recurrent variants of this type, and includes 97–98% of the nonsynonymous variants detected in an average genome through exome sequencing.
Two case-control studies of schizophrenia have been published recently using the exome chip, both of which implicated these variants in schizophrenia. Loohuis et al. (2015) described an increased burden of variants with MAF ≥ 0.01% in cases versus controls, and enrichment of rare variants in neurodevelopmental genes. Richards et al. (2016) reported that the rare (MAF < 0.1%) variant association signal was enriched in genes previously implicated in both common and rare variant studies of schizophrenia.
We hypothesised that families with multiple affected individuals may be enriched for disease-associated variants of this class compared to unrelated populations. Here, we report the results of an exome chip study in a sample comprising 99 pedigrees from the Western Australian Family Study of Schizophrenia (WAFSS), all of which included a proband with a diagnosis of schizophrenia and at least one first degree relative with schizophrenia. We assessed the association of variants with schizophrenia both individually and collectively, to ascertain whether exome chip variant burden is enriched in cases compared to their unaffected relatives. We looked for enrichment of variants in the schizophrenia-related gene-sets reported by Richards et al. (2016) and also conducted agnostic pathway analyses. Finally, we investigated whether rare variant burden is increased in this multiplex family dataset as a whole, compared to a large unselected sample of similar ethnicity.
Section snippets
Materials and methods
A full description of methods is provided in Supplementary Methods. Unless otherwise specified, all statistical analyses were performed in R version 3.2.0.
Association with individual variants
Quantile-Quantile plots (Supplementary Figs. 9 and 10) indicate that there was a slight deflation of the test statistics for association with schizophrenia in all 56,398 variants compared to the null distribution, which disappears when only common (MAF > 5%) variants are included. None of the variants were individually associated with schizophrenia beyond Bonferroni threshold (P < 8.8E-07). The top ten most significant associations are shown in Supplementary Table 9. Most of these ten variants were
Discussion
This study set out to investigate for the first time whether there is evidence that rare-but-recurrent, likely-deleterious variants assayed on the exome array play a role in schizophrenia in multiplex families which we hypothesised might be enriched for rare, schizophrenia-associated variants.
Minor allele frequencies of rare variants were on average higher in the WAFSS compared to the DIAGRAM, suggesting that these multiplex families may harbour a greater variant burden compared to a
Conflict of interest
All other authors declare that they have no conflicts of interest.
Contributors
N.S.M., E.K.M., P.M, A.J., J.C.B., J.B., V.A.M. were involved in the design of the study.
N.S.M. carried out all analysis with input from S.V.W., S.M.A., M.D., M.L.C., B. M and J.P.R.
N.S.M. wrote the manuscript which was subsequently revised by all authors.
All authors contributed to and have approved the final manuscript.
Funding
This work was supported by National Health and Medical Research Council Australia (APP1064582 to AJ, EKM, NSM, PEM, JCB, VAM 2014–2016), and by The Society for Mental Health Research (SMHR) Postdoctoral Fellowship [to NSM, 2016] and the Medical Research Foundation, Royal Perth Hospital (Small Project Grant to NSM 2015). Support from the Cooperative Research Centre for Mental Health Australia is also gratefully acknowledged.
Acknowledgements
We would like to thank all of the participants in the Western Australian Family Study of Schizophrenia.
References (35)
- et al.
Multipoint quantitative-trait linkage analysis in general pedigrees
Am. J. Hum. Genet.
(1998) - et al.
Genetic evidence for a distinct subtype of schizophrenia characterized by pervasive cognitive deficit
Am. J. Hum. Genet.
(2005) - et al.
Common genetic determinants of schizophrenia and bipolar disorder in Swedish families: a population-based study
Lancet
(2009) - et al.
Neural ECM in addiction, schizophrenia, and mood disorder
Prog. Brain Res.
(2014) - et al.
Exome sequencing followed by large-scale genotyping suggests a limited role for moderately rare risk factors of strong effect in schizophrenia
Am. J. Hum. Genet.
(2012) - et al.
Pooled association tests for rare variants in exon-resequencing studies
Am. J. Hum. Genet.
(2010) - et al.
A missense mutation in PPP1R15B causes a syndrome including diabetes, short stature, and microcephaly
Diabetes
(2015) - et al.
Merlin-rapid analysis of dense genetic maps using sparse gene flow trees
Nat. Genet.
(2002) - et al.
Translational control in synaptic plasticity and cognitive dysfunction
Annu. Rev. Neurosci.
(2014) - et al.
An atlas of genetic correlations across human diseases and traits
Nat. Genet.
(2015)
LD score regression distinguishes confounding from polygenicity in genome-wide association studies
Nat. Genet.
The diagnostic interview for psychoses (DIP): development, reliability and applications
Psychol. Med.
Robust inference of population structure for ancestry prediction and correction of stratification in the presence of relatedness
Genet. Epidemiol.
Diagnostic exome sequencing in persons with severe intellectual disability
N. Engl. J. Med.
Rare UNC13B variations and risk of schizophrenia: whole-exome sequencing in a multiplex family and follow-up resequencing and a case-control study
Am. J. Med. Genet. B
De novo mutations in schizophrenia implicate synaptic networks
Nature
Increased exonic de novo mutation rate in individuals with schizophrenia
Nat. Genet.
Cited by (15)
Psychiatric symptoms and syndromes transcending diagnostic boundaries in Indian multiplex families: The cohort of ADBS study
2021, Psychiatry ResearchCitation Excerpt :The large extended Scottish pedigree that detected the link between a translocation 1:11 (DISC1) and schizophrenia, bipolar disorder and recurrent mood disorder (but not major depression or addiction) (St Clair et al., 1990), suggested that syndromes often overlapped in families.Individuals who transit from one syndrome to the other are often observed (Kendler and Tsuang, 1982); and whether these are distinct syndromes, has long been debated (Tsuang, 1975). In the contemporary genomics era, these families have provided a wealth of data, as valuable leads have been provided by whole exome or genome sequencing (Homann et al., 2016), in specific families (Anjanappa et al., 2020; John et al., 2020); and to also probe the interface between ancestry and familial aggregation of mental illness (McCarthy et al., 2017; Sul et al., 2020). In addition, the heritability of neuropsychological traits, psychological symptoms, minor physical anomalies or soft neurological signs, with genetic and environmental influences, may also be useful (Blokland et al., 2017).
Genetic studies of psychosis
2020, Risk Factors for Psychosis: Paradigms, Mechanisms, and PreventionNo association between common genetic variation in FOXP2 and language impairment in schizophrenia
2019, Psychiatry ResearchCitation Excerpt :Schizophrenia cases without hallucinations were included in the analysis with the same phenotypic coding as the healthy controls, all of whom were free from hallucinations. The WAFSS cohort has previously undergone genotyping for genome wide association studies (GWAS) (n = 592 unrelated individuals) and whole genome sequencing (WGS) (n = 317 individuals from the familial sample, described in (McCarthy et al., 2017). These genetic data were cleaned using standard genotype quality control (QC), including removal of highly missing (> 5%) SNPs and individuals.
Rare variant analysis in multiply affected families, association studies and functional analysis suggest a role for the ITGΒ4 gene in schizophrenia and bipolar disorder
2018, Schizophrenia ResearchCitation Excerpt :Follow up studies in other samples provide additional support for the involvement of this gene in the aetiology of SCZ and bipolar disorder. Further support comes from data from 872 subjects from the Western Australian Family Study of Schizophrenia and 36,355 controls who reported two SCZ case only ITGB4 variants (McCarthy et al., 2017). However, it should be noted that for rs147480547 and rs145976111 the MAF (minor allele frequency) observed in both sets of controls is lower than in the Non-Finnish European samples in the ExAC database of aggregated exome sequencing data from large-scale sequencing projects (Lek et al., 2016).