Investigating genome-wide transcriptional and methylomic consequences of a balanced t(1;11) translocation linked to major mental illness
Item statusRestricted Access
Embargo end date08/07/2018
McCartney, Daniel Lawrence
MetadataShow full item record
Schizophrenia, bipolar disorder and major depressive disorder are devastating psychiatric conditions with a complex, overlapping genetic and environmental architecture. Previously, a family has been reported where a balanced chromosomal translocation between chromosomes 1 and 11 [t(1;11)] shows significant linkage to these disorders. This translocation transects three genes: Disrupted in schizophrenia- 1 (DISC1) on chromosome 1, a non-coding RNA, Disrupted in schizophrenia-2 (DISC2) antisense to DISC1, and a non-coding transcript, DISC1 fusion partner-1 (DISC1FP1) on chromosome 11, all of which could result in pathogenic properties in the context of the translocation. This thesis focuses on the genome-wide effects of the t(1;11) translocation, primarily examining differences in gene expression and DNA methylation, using various biological samples from the t(1;11) family. To assess the genome-wide effects of the t(1;11) translocation on methylation, DNA methylation was profiled in whole-blood from 41 family members using the Infinium HumanMethylation450 BeadChip. Significant differential methylation was observed within the translocation breakpoint regions on chromosomes 1 and 11. Downstream analysis identified additional regions of differential methylation outwith these chromosomes, while pathway analysis showed terms related to psychiatric disorders and neurodevelopment were enriched amongst differentially methylated genes, in addition to more general terms pertaining to cellular function. Using induced pluripotent stem cell (iPSC) technology, neuronal samples were developed from fibroblasts in a subset of individuals profiled for genome-wide methylation in whole blood (N = 6) with an aim to replicate the significant findings around the breakpoint regions. Here, methylation was profiled using the Infinium HumanMethylation450 BeadChip’s successor: the Infinium MethylationEPIC BeadChip. The results from the blood-based study failed to replicate in the neuronal samples, which could be attributed to low statistical power or tissue-specific factors such as methylation quantitative trait loci. The differences in methylation in the most significantly differentially methylated loci were found to be driven by a single individual, rendering further interpretation of the findings from this analysis difficult without additional samples. Cross-tissue analyses of DNA methylation were performed on blood and neuronal DNA from these six individuals, revealing little correlation between cell types. DISC1 is central to a network of interacting protein partners, including the transcription factor ATF4, and PDE4; both of which are associated with the cAMP signalling pathway. Haploinsufficiency of DISC1 due to the translocation may therefore be disruptive to cAMP-mediated gene expression. In order to identify transcriptomic effects which may be related to the t(1;11) translocation, genome-wide expression profiling was performed in lymphoblastoid cell line RNA from 13 family members. No transcripts were found to be differentially expressed at the genome-wide significant level. A post-hoc power analysis suggested that more samples would be required in order to detect genome-wide significant differential expression. However, imposing a fold-change cut-off to the data identified a number of candidate genes for follow-up analysis, including SORL1: a member of the brain-expressed Sortilin gene family. Sortilin genes have been linked to multiple psychiatric disorders including schizophrenia, bipolar disorder and Alzheimer’s disease. Follow-up analyses of Sortilin family members were performed in a Disc1 mouse model of schizophrenia, containing an amino acid substitution (L100P). Here, developmental gene expression profiling was performed with an additional aim to optimise and validate work performed by others using this mouse model. However, results from these experiments were variable between two independent batches mice tested. Additional investigation of Sortilin family genes was performed using GWAS data from human samples, using machine learning techniques to identify epistatic interactions linked to depression and brain function, revealing no statistically significant interactions. The results presented in this thesis suggest a potential mechanism for differential DNA methylation in the context of chromosomal translocations, and suggests mechanisms whereby increased risk of illness is conferred upon translocation carriers through dysregulation of transcription and DNA methylation.