Translating genetics into molecular pathways: revision of the 4p locus linkage for bipolar affective disorder and studying the effects of a copy number variation in DLG2 associated with schizophrenia

Abstract

The field of psychiatric genetics has identified several genes and molecular pathways as potentially implicated in the pathophysiology of major mental illness. This thesis describes two projects based on findings from genetic linkage and association studies- the 4p15-p16 locus implicated in bipolar affective disorder (BD) and copy number variants (CNVs) in DLG2 associated with schizophrenia (SCZ). By taking these putative risk variants with a relatively large effect size as starting points, my thesis aimed to explore potential cellular processes and molecular pathways that might contribute to the aetiology of these two psychiatric disorders.

Genome-wide linkage study in a large Scottish family (SBF2) identified strong linkage in a region on chromosome 4p (4p15-p16 locus) for affective disorders (BD and major depressive disorder, MDD). My initial aim was to investigate the cellular and molecular consequences of the disease-linked haplotype in iPSC-derived neurons from SBF2 family members. Surprisingly, review of the clinical records led to the re-evaluation of a significant number of BD cases and their subsequent revision to ‘non-BD’ diagnoses. This created the need for a new linkage study of SBF2. The new linkage analysis identified suggestive linkage in two novel loci- 2q36-q37 and 18q11-q12, for BD and MDD. Unlike previous linkage studies, the linkage on the 4p15-p16 locus for BD and MDD was not confirmed. However, a suggestive signal in a region overlapping with this locus was detected in a model including all individuals with a psychiatric diagnosis in SBF2. All three regions contain genes genetically and/or functionally implicated in major mental disorders.

A genome-wide association study conducted by the International Schizophrenia Consortium showed association of CNVs within DLG2 with SCZ. Moreover, recent studies have identified deficits in cognitive functions in individuals carrying intra-DLG2 CNVs. DLG2 encodes PSD-93- a member of the DLG-Membrane associated guanylate kinases (DLG-MAGUKs) subfamily of proteins. PSD-93 is a scaffold protein located at the post-synaptic density (PSD) of excitatory synapses, where it interacts with other molecules (e.g., scaffold proteins, adhesion and cytoskeletal molecules, and enzymes) and forms a functional bridge between the pre- and post-synaptic membrane and the cytoskeleton. Importantly, PSD-93 binds to the ionotropic AMPA and NMDA glutamate receptors. During synaptic activity, PSD-93 participates in the trafficking of AMPARs to and from the PSD, a process underlying long-term potentiation and depression (LTP and LTD). In addition, PSD-93 regulates the production of nitric oxide by coupling nitric oxide synthetase with NMDARs. Up to this day, little is known of how patient-associated DLG2 CNVs affect these functions, especially in the dopaminergic system.

To address this, I modelled an intra-DLG2 deletion found in a patient with SCZ in the human dopaminergic neuronal cell line LUHMES using CRISPR/Cas9 genome editing. I then explored the effects of the mutation on AMPAR trafficking during chemically-induced LTP, as well as on neuronal viability before and after inducing oxidative stress. Compared to wild-type (WT), dopaminergic neurons carrying the intra-DLG2 CNV showed reduced AMPAR trafficking during the glycine-induced LTP protocol, suggesting an impairment in this form of synaptic plasticity. Meanwhile, the presence of the antipsychotic haloperidol appeared to rescue this phenotype. The CNV mutation was also associated with reduced viability following H2O2-induced oxidative stress in immature dopamine neurons, but not in neuronal precursors. Technical difficulties and time constraints prevented me from assessing the effect of oxidative stress on neuronal viability at later differentiation stages. In addition, the CNV-carrying neurons displayed lower extracellular and intracellular dopamine levels compared to WTs.

Overall, the work described in this thesis builds on previous findings in psychiatric genetics. It addresses caveats of the contemporary classification system of psychiatric disorders by highlighting ways for minimising clinical diagnostic uncertainty and bias. Moreover, it comments upon best practice in the diagnostic process during patient recruitment for linkage studies. To my knowledge, this is the first study to report possible linkage on 2q36-q37 and 18q11-q12 for BD and MDD. It is also the first attempt to assess the effect of a SCZ-associated intra-DLG2 CNV in human dopaminergic neurons. It provides some initial evidence of impaired LTP, as well as reduced neuronal viability after oxidative stress and altered extra- and intra-cellular dopamine levels. Further work is needed to validate both the results from the new linkage analysis in SBF2 and the potential molecular consequences of intra-DLG2 CNVs in the dopaminergic system.