Utilizing Identical Twins Discordant for Schizophrenia to Uncover de novo Mutations
We are living in exciting times – the reality of high-resolution Cand individual genome sequencing now offers renewed hope in the search for the causes of complex diseases. When this technology is combined with genetic relationships, individual sequences add unrivaled proficiency.
Our lab is located in London, Ontario, Canada at the University of Western Ontario, and our interest is in elucidating aspects of the underlying genetic mechanisms contributing to complex disease. The project I am working on focuses on the use of identical twins discordant for schizophrenia and their families to uncover de novo mutations that may contribute to one twin having the disease and the other not. Given the nearly equivalent genetic structure of identical twins, any difference between identical twins discordant for a disease will have a likelihood of being involved in disease pathology.
Schizophrenia is a debilitating mental disease that generally onsets in early adult life and is characterized by a disruption in perception and thinking. Across the world, schizophrenia prevalence is approximately 1%. The best predictor of the disease is family history but the inheritance pattern of schizophrenia is complex. Identical twins, who are said to share nearly 100% of their genetic makeup, are concordant for the disease in only 48% of cases. For these reasons, we believe that the complexity of this disease’s inheritance and pathology can be best explained by a mix of environmental, genetic and epigenetic factors.
In an effort to uncover some of these factors, our lab has assessed identical twins using the Affymetrix Human SNP 6.0 Array as well as the Complete Genomics Next Generation Sequencing platform. Our analysis has focused on many different types of variants with special interest placed on Copy Number Variation (CNVs). The identification of chromosomal structural variants in the form of CNVs has opened a new direction in schizophrenia genomics. CNVs are common and widespread in the genome and by virtue of their variable size may directly disrupt multiple genes at once. We have focused on looking at de novo CNVs – that is CNVs that are not shared between identical twins and also not inherited from the mother or father. Many recent studies have found CNVs to be implicated in mental disorders, particularly autism and schizophrenia. A number of candidate regions and genes have been identified to be associated with both of these disorders through CNV studies.
We have also assessed other types of variation in our ongoing project. Our experimental design has allowed us to evaluate the similarities and differences across unrelated individuals, parents and children, as well as between identical twins. Our next-generation sequencing results show that an individual carries approximately 3.7 million Single Nucleotide Variants (SNVs) when compared to the reference, 150 CNVs, 400,000 indels and 220 microRNA variants.
Our next-generation sequencing results also identified that two unrelated individuals differed for 1.5-1.8 million SNVs, that is ~45% of their total SNVs, a parent and child differed for 0.9-1.0 million SNVs (~30%) and a pair of identical twins differed for only about 100,000 (~3%) SNVs. Interestingly, CNV and SNV differences between monozygotic discordant (MZD) twins appear to affect a set of genes enriched in neurodevelopmental genes, as well as genes that have been already implicated in schizophrenia.
We have been fortunate enough to use Golden Helix’s SNP & Variation Suite (SVS) to analyze both our microarray and next-generation sequencing data. This has been incredibly helpful because most days working on a project that outputs a large amount of data, as is the case when working on 6 whole genome sequences and 12 microarray hybridizations, feels like trying to find a rare piece of hay in a giant haystack! The use of the Golden Helix software has allowed us to streamline our analysis and keep all of our data in one place. Specifically, I have used the DNA-seq and CNV workflows provided by Golden Helix to analyze our raw data and compare the genomes of identical twins. All in all, Golden Helix’s software has saved me a lot of time and sped up my research considerably!
The strategy we have used, that is, the use of identical twins discordant for disease, provides a method of assessing variations in individual patients and genetically similar twins in hopes of identifying rare variants. This strategy is contrary to the usual experimental design where a very large number of patients are compared to a very large number of controls. Given the highly heterogeneous nature of schizophrenia, our monozygotic twin approach compliments larger studies by using a more closely matched sample to identify a set of candidate genes that can then be assessed in larger groups of patients. We hope that the final outcomes of this research will provide a prototype for the use of genome-wide twin-based studies in uncovering candidate genes and/or regions in this and other complex diseases.
So far, our findings support our strategy in identifying patient-specific genetic changes that may lead to schizophrenia discordance between identical twins. The results reinforce that individual genomes harbor extensive variability, some inherited and others acquired during meiosis and/or mitosis. There is no single human genome sequence. Even identical twins are not identical and each individual may be a mosaic, potentially carrying different variations in different cells.
Detection of genetic mutations involved in even a small subset of schizophrenia patients will be a major breakthrough. It has the potential to allow for increased genetic understanding of this devastating and life-altering disease which in turn will hopefully have diagnostic and therapeutic implications for those suffering from schizophrenia.