Whole genome sequencing study of identical twins discordant for psychosis

Schizophrenia, schizoaffective disorder, and bipolar disorder are substantially heritable psychiatric disorders often exhibiting psychotic symptoms. While twin studies suggest high heritability (60-80%), genome-wide association studies (GWAS) have only explained a fraction of this heritability. This suggests that rare variants, not detectable by GWAS, may play a significant role. Monozygotic (MZ) twins, despite sharing nearly identical genomes at birth, often exhibit phenotypic discordance, highlighting the potential contribution of post-zygotic mutations. These mutations, occurring after fertilization, could be responsible for the differences in disease manifestation between twins. Whole exome sequencing (WES) and whole genome sequencing (WGS) offer a powerful approach to identify these rare variants. This study leveraged the MZ twin design to minimize the confounding effects of shared genetics and environment, focusing on identifying rare, post-zygotic variants present only in the affected twin of discordant pairs. The researchers hypothesized that rare, post-zygotic genetic variation, present in the affected twin, increases their disease risk, even when a common genetic and environmental risk is already shared by both twins. This approach drastically reduces the search space for candidate causal variants compared to case-control studies.

Existing literature highlights the significant heritability of schizophrenia and related psychotic disorders, but also the limitations of GWAS in capturing the full genetic landscape. Previous studies have demonstrated the presence of post-zygotic mutations in MZ twins discordant for various disorders, suggesting a potential role in phenotypic differences. Research employing WES or WGS has identified de novo post-zygotic variants in MZ twins discordant for a range of disorders, however, similar investigations in psychotic disorders remain limited. The use of MZ twin pairs discordant for disease phenotypes offers a powerful approach to discover rare variants contributing to disease risk, by controlling for genetic and many shared environmental effects. While the concordance rate for schizophrenia in MZ twins is approximately 50%, substantially higher than the population incidence rate, understanding the genetic basis for discordance remains crucial.

The study used data from the Swedish Twin Register (STR) and the National Patient Register (NPR) to identify MZ twin pairs discordant for schizophrenia, schizoaffective disorder, or bipolar disorder. Initially, 19 pairs were selected from the STAR cohort, with one twin diagnosed with one of the target disorders and the other unaffected (or having MDD without psychotic symptoms). Blood samples were collected, and DNA extracted using a Puregene extraction kit. WGS was performed using a HiSeqX sequencer, aiming for 30x coverage. Data quality control included checks for DNA contamination, degradation, and sex discordance. Variant calling involved using BWA-MEM for alignment, and GATK for variant calling and quality control. Peddy was used to assess relatedness and ancestry. SNVs and Indels were filtered based on quality scores (QUAL <100, GQ <20, DP <10), and prioritized based on predicted deleteriousness (SIFT, PolyPhen-2, CADD), functional impact (VEP), and allele frequency (<1% in 1000 Genomes and gnomAD). CNV calling was performed using a family-based consensus approach across four algorithms, removing low-confidence calls and those overlapping common CNVs. Prioritization was based on rarity, presence in known pathogenic databases (ClinGen), and overlap with genes previously associated with psychosis. Somatic CNVs were assessed using MoChA, and repeat expansions using ExpansionHunter.

After quality control, data from 17 MZ twin pairs were analyzed. The analysis revealed six rare, predicted deleterious, discordant SNVs in four genes: FOXN1 (three variants), FLOT2, NUTM2G, and KRTAP10-6. These variants were present only in the affected twin. Further analysis of regulatory variants did not reveal significant differences in regulatory feature overlap between affected and unaffected twins. The CNV analysis identified four rare, discordant CNVs in affected individuals. Notably, a duplication on chromosome 3q29, previously implicated in autism and developmental delay, was found exclusively in one affected twin. An additional 14 CNVs with clinical impact were identified, but only the 3q29 duplication was private to an affected twin. Analysis of repeat expansions did not identify any significant findings.

This study identified rare, discordant SNVs and CNVs in affected twins from MZ pairs discordant for psychotic disorders. The findings suggest that rare post-zygotic variants, potentially arising early in development, could contribute to the phenotypic discordance observed in MZ twins with psychotic disorders. While the sample size limits the ability to definitively establish statistical significance, the identification of variants in genes previously implicated in schizophrenia or bipolar disorder (through other studies like SCHEMA and BipEx), provides support for the role of rare genetic variation in disease etiology. The 3q29 duplication, while previously linked to other neurodevelopmental disorders, raises questions about the potential pleiotropic effects of this region. The lack of significant findings with respect to repeat expansions is notable. Future research could investigate the potential interaction between identified variants and other factors, such as environmental effects and epigenetic modifications, to gain a more comprehensive understanding of the pathophysiology of psychotic disorders.

This large-scale WGS study of MZ twins discordant for psychotic disorders provides evidence supporting the role of rare, post-zygotic genetic variants in disease susceptibility. The identification of specific SNVs and CNVs, some in genes already associated with these disorders, warrants further investigation. Future studies should focus on larger cohorts to achieve greater statistical power and include parental genome information to confirm de novo status and examine shared de novo variants. Investigating other tissue types beyond blood and incorporating epigenetic analyses will further enhance our understanding of the complex interplay of genetic and environmental factors contributing to the phenotypic diversity of psychotic disorders.

The study's relatively small sample size limits the statistical power to definitively assess the burden of rare, discordant, protein-coding variants. The lack of parental genome data prevents confirmation of de novo status for all identified variants. The use of blood samples may not fully capture somatic mutations present in brain tissue, which might be more directly relevant to disease pathogenesis. The potential for later-onset disease in unaffected twins is acknowledged, which could influence interpretation of the findings. While comprehensive diagnostic procedures were employed, residual uncertainty about phenotypic discordance always exists in these studies.