Identifying mechanisms that upregulate the cold shock protein RBM3

Abstract

Synapses are continuously being remodelled in the adult mammalian brain through a process called structural synaptic plasticity. Understanding the mechanisms regulating this could yield important insights into brain function in both health and disease. In particular, the loss of structural plasticity and resultant loss of synapses in neurodegenerative disorders represents an important potential application of these insights. Hypothermia induces dynamic and reversible structural synaptic changes and is an example of structural synaptic plasticity. RNA-binding motif 3 (RBM3) plays a critical role in orchestrating this cold- associated structural plasticity. In neurodegenerative disease models, this structural plasticity is lost in early disease along with the ability to induce RBM3 on cooling. RBM3 overexpression before this period, either by cooling or virally mediated ectopic expression, rescues cold-associated structural plasticity and is neuroprotective in mouse models of neurodegeneration. Very recently, RBM3 upregulation in neurons was achieved without cooling, suggesting the presence of temperature-independent mechanisms that can induce RBM3. The aim of this thesis is to better understand the mechanisms of RBM3 induction. The first hypothesis was that a temperature sensitive potassium channel, TWIK-related K+ channel 1 (TREK-1), acts as a neuronal “cold sensor” that transduces reduced temperature to RBM3 induction. In the hippocampus, TREK-1 knockout was associated with a failure to induce RBM3 on cooling in vitro and in vivo. To mimic the effect of cooling on TREK-1, the effect of a selective blocker, minispadin, was tested and was found to induce RBM3 levels without cooling in vitro and in vivo, via TREK-1. The mechanism of RBM3 induction was then hypothesised to be related to network activity. TREK-1 inhibition was found to suppress network activity, similar to cooling, indicating that TREK-1 is likely to be localised on interneurons. As interneurons release GABA on excitation, the effect of GABAA receptor blockade on RBM3 expression was tested and was found to block RBM3 induction on cooling. To further test the importance of interneuron expression of TREK-1 in inducing RBM3, TREK-1 null mice were injected with virus expressing TREK-1 in excitatory cells and resulted in a suppression of RBM3 on cooling. The second hypothesis was that TREK-1 is a target in neurodegeneration. Minispadin induced hippocampal RBM3 levels in mouse models of Alzheimer’s and prion disease, without cooling. Synapse density in CA1 hippocampus of wild-type mice was significantly increased in mice after treatment with minispadin. In prion diseased mice, minispadin rescued synapse loss, neuronal loss, prolonged survival, and may have improved burrowing behaviour. Finally, I hypothesised that unbiased pharmacological and genetic screens could identify novel targets for RBM3 induction. A high throughput assay suitable for drug screening was developed and validated. Whilst a screen of 33,500 drugs did not reveal any pharmacological inducers, a parallel CRISPR activation screen revealed several classes of proteins that may be druggable candidates. These will be validated in future work.