Resting state functional magnetic resonance imaging in Fragile X syndrome

Abstract

Autism Spectrum Disorders (ASD) and Intellectual Disabilities (ID) are two co-occurring neurodevelopmental disorders that affect approximately 100 million individuals worldwide, making them leading causes of disability in children. Current therapeutic interventions are restricted to managing symptoms and behavioural therapy. There is therefore an urgent unmet need for rational treatments targeting underlying causes. The study of animal models of monogenic disorders associated with a high prevalence of ASD/ID, such as Fragile X Syndrome (FXS) can help us gain a better understanding of the pathophysiology of ASD/ID and hence develop new disease-modifying therapies. However, a key requirement for effective therapeutic development is the ability to directly compare findings from basic neuroscience in rodent models with human studies. Methodological approaches that can be applied to both human and rodents would greatly facilitate this endeavour. In this thesis, I used resting state functional magnetic resonance imaging (rsfMRI) to study functional brain connectivity in rat models of FXS and affected humans. RsfMRI is a non-invasive imaging technique that is directly translatable across species, therefore providing a powerful tool to investigate how brain activity is modulated across development and in disease. First, I established and validated a methodology to acquire and analyse rsfMRI data from anaesthetised rodents. I verified my methods by successfully identifying resting state networks (RSNs) in both humans and rats. I additionally demonstrated that anaesthesia concentration, but not duration, affect RSNs connectivity in wild type (WT) rats. Second, a parallel imaging approach was used in individuals with FXS and in the Fmr1 −/y rat model of the disorder to investigate the functional connectivity of FXS. I observed a global decrease in connectivity in Fmr1 −/y Long-Evans Hooded (LEH) rats, as well as a specific decrease in their Default Mode Network (DMN) connectivity. To determine how this change in functional brain connectivity in the rat model related to the human condition, I next examined whether similar alterations in functional connectivity occurred in individuals with FXS. I observed that cortico-cortical connectivity was reduced at a whole-brain scale and that the DMN was underconnected in FX individuals, indicating that the Fmr1 −/y rat is a good model for global changes in brain connectivity found in FXS. Several other interesting findings observed in the rat data may be relevant to the human condition. First, I found that Fmr1 −/y LEH rats were less sensitive to isoflurane anaesthesia than WT rats. I also observed strain-specific functional connectivity impairments. In comparison with Sprague-Dawley (SD) rats, LEH rats showed a relative decrease in hypothalamic connectivity. Additionally, genetic deletion of Fmr1 in SD rats resulted in a selective hypoconnectivity of the cerebellum and caudate putamen. Finally, I examined the developmental trajectory of DMN connectivity and tested the hypothesis that reducing ERK activity with lovastatin would restore DMN connectivity in Fmr1 −/y rats, as lovastatin was previously found to restore behavioural, electrophysiological and molecular impairments in rodent models of FXS. I found that DMN connectivity between Fmr1 −/y rats and WT littermates is comparable at 4 weeks of age, but diverges by 8 weeks, leading to an underconnected DMN in Fmr1 −/y rats. In addition, I demonstrated that lovastatin treatment during this developmental period does not prevent the emergence of DMN underconnectivity in Fmr1 −/y rats. In summary, this thesis provides the first account of altered functional brain connectivity in a rat model of FXS, adding to the existing body of literature reporting altered connectivity in ASD. In addition, it provides evidence that functional connectivity abnormalities observed in a rat model of FXS directly translate to individuals with FXS. It further shows that while effective in preventing some of the pathophysiology resulting from the loss of FMRP, early lovastatin treatment was unsuccessful at restoring DMN connectivity in Fmr1 −/y rats. Taken together, these findings highlight the potential of using rsfMRI as a diagnostic tool and in assessing therapeutic efficacy in rodent models of neurodevelopmental disorders.