UCLA

Stroke is the leading cause of adult disability and leaves millions of patients worldwide with chronic deficits in sensory, motor, and cognitive function. Although there are currently no pharmacological treatments for brain recovery after stroke, clinical trials have shown that a limb overuse paradigm, known as constraint-induced movement therapy (CIMT), can lead to significant and lasting motor improvements. This functional improvement is strongly associated with increased activity in the ipsilesional premotor cortex, but the exact neuronal connections and molecular processes that underlie this plasticity process are unknown.

The body of work within this dissertation identifies one potential mechanism of cortical axonal sprouting after limb overuse in a mouse model of stroke and forelimb overuse that approximates human CIMT. Via quantitative neuroanatomical mapping, we find that limb overuse drives the novel formation of a connection from retrosplenial cortex to premotor cortex (RSC-PMC). We then isolate the very neurons that comprise this connection by FACS (Fluorescence-Activated Cell Sorting) purification for RNA-Seq transcriptional profiling. The RSC-PMC transcriptome is characterized by expression of activity-induced and growth-related genes, several of which are new to the context of CNS injury and plasticity. Pathway analyses indicate that limb overuse induces signaling in cell growth and tissue development, calcium signaling, and Ephrin A pathways. Upstream transcriptional regulators of the differentially expressed genes within RSC-PMC include NeuroD1, NeuroG3, YY1, and Otx2. Subsequently, top candidates from the 160 differentially regulated genes by limb overuse were functionally screened using CRISPR/cas9 in an in-vitro neuronal outgrowth assay. Two molecular hits were Otx2 and NeuroG3, genes previously associated with critical period onset/closure and neurogenesis, respectively. Finally, whole transcriptome comparison studies of the RSC-PMC across various plasticity paradigms indicate relationship to activity-dependent/learning transcriptomes and divergence from developmental growth programs. In parallel, we find that GDF10 (growth and differentiation factor 10) is an early trigger of axonal sprouting after stroke. Even in the absence of limb overuse, we find that in-vivo delivery of GDF10 increases axonal peri-infarct sprouting, forms synaptic contacts in premotor cortex, and results in improved behavioral recovery. Together, these studies identify potentially synergistic mechanisms to target in future therapeutic strategies for stroke.