Functional network study on the wild type and DISC1 transgenic mice

Abstract

Major mental disorders such as schizophrenia (SZ), bipolar disorder and major depression create a heavy social burden and remain poorly treated. Recent reports examining the genetic and molecular structure of brain disorders discovered the gene Disrupted-in-Schizophrenia 1 (DISC1) to be significantly associated with these disorders and plays a substantial role in influencing a range of shared endophenotypes underlying these disorders. This has led to a powerful gateway to explore the underlying pathology and to achieve better diagnosis and treatments of major mental diseases. Characterising the phenotypes of DISC1 in brain functions would provide valuable insights into the mechanismof how this gene influences the brain.Resting-state fMRI (rs-fMRI) is a powerful technique being applied to more than 30 different kinds of brain disorders and different species. Findings from rs-fMRI studies on humans provide ample evidence of abnormal functional patterns in patients with mental illnesses. Specifically, altered functional connectivity (FC) and network topologies are proposed to be plausible imaging endophenotypes of psychotic disorders. However, little is known about the direct effects of gene dysfunction, such as DISC1 mutations, on resting-state brain activations. One aim of this thesis is to investigate the impacts of DISC1 on variations of resting-state functional networks (RSNs) in anaesthetized mouse brains usingDISC1 transgenic mice and rs-fMRI.Before characterizing disease phenotypes in mouse models, it is necessary to understand how the mouse brain functions in normal states. Anaesthesia is currently an integrated part of most mouse rs-fMRI experiment. However, it influences blood-oxygenation-level dependent (BOLD) fMRI signals via modulation of neurovascular coupling and brainmetabolism, and therefore alters FC and topologies of RSNs. Investigating FC and its changes associated with genes therefore initially requires an understanding of the anaesthetics. Furthermore, various anaesthetic protocols are adopted by different labs leading to difficulties in result comparisons.The brain is a complex and hierarchical system, whereby its RSN architecture can be characterised from mutiple perspectives at regional or global levels. Previous mouse rs-fMRI studies have focused on long-distance FC, however, given the non-uniform influence of anaesthesia on the brain, mapping functional activations across scales in mice underdifferent anaesthetics remain sparsely explored. Establishing this relationahsip would provide a reliable reference to study mouse brain functions in abnormal states and facilitate comparisons across laboratories. In the first aim of this thesis, I investigated the local connectivity in wide type (WT) mice under six commonly used anaesthetic regimens. I identified that the regional connectivity altered across the brain as a function of the expression density of receptors relative to specific agents. In addition, the depth of anaesthesia influences local synchronization.In the second aim, I investigated how large-scale network topologies changed under different regimens. Small-worldness was observed to be different between medetomidine and a combination of medetomidine and isoflurane, which was driven by altered clustering coefficients. This indicated that local alterations played important roles on global topologies.Similar functional modules were observed under most regimens, but the anaesthetic dosage influenced module structures significantly.In the third aim, I investigated the potential link between RSNs and the DISC1 gene. Isoflurane was chosen based on results from the previous aims and experimental conditions available. Eleven DISC1 transgenic mice (Tg) and ten WT littermate controls (Wt) were first sent to Y maze and fear conditioning behaviour tests before the acquisition of rs-fMRI data from all mice. Independent component analysis (ICA), regional connectivity and complex network analysis were performed to detect the link between FC and the DISC1 gene across analytical scales. No significant differences were revealed by behaviour assays, ICA, regional analysis or global graph parameters. However, functional modules were different between the two groups. Specifically, the amygdala, a region closely related to the fear conditioning, was assigned to a different module in Tg mice compared to Wt mice.Furthermore, some discrepancies in modular structures in Wt mice in this aim and WT mice under isoflurane in the second aim were observed. This implies that there might be other factors influencing community identification using mouse rs-fMRI, for example, factors relating to experimental pipeline used.The results from the first two aims are the first to establish regional connectivity and large scale network topology in mice across different anaesthetics. Methods used in these aims are widely applied in human studies, therefore, the results presented here will serve as preliminary references for comparisons across labs and help translate observations frommice to humans. Results from the third aim are the first to investigate FC alterations associated with the DISC1 gene across scales using mouse rs-fMRI. The abnormal functional modules in DISC1 mice provide insights into how this gene may influence brain functions in amygdala-related emotional processing. Finally, differences in modular organizations of WT mice between the second and the third aims suggest that experimental pipelines and other factors may influence module identification.