Abstract
Every year, millions of people worldwide will have a stroke. For those who survive, many will live on with persistent disability, including cognitive impairments. To better understand the mechanisms underlying cognitive deficits that arise from stroke, our laboratory generated a preclinical stroke model that targets the prefrontal cortex (PFC) bilaterally in mice using the photothrombosis method, that results in delayed-onset spatial memory impairment. The overall aim of this thesis was to further characterise this stroke model in aged animals in an attempt to better represent the clinical population disproportionately affected by stroke. To do so, we employed a battery of behavioural tasks, immunohistochemical labelling and real-time quantitative polymerase chain reaction techniques to investigate age- and PFC stroke-mediated cognitive impairments and underlying mechanisms.
Relative to younger counterparts, aged sham animals displayed increased anxiety-like behaviour and an age-related decline in cognitive flexibility as assessed on the elevated plus maze (EPM) and visual discrimination (VD) reversal tasks, respectively. Moreover, age-related reductions in neurotrophic support were evident in the PFC of sham animals, including a reduction in the genetic expression of brain-derived neurotrophic factor (BDNF) and one of its receptors (Ntrk2.fl). Similarly, aged stroke animals displayed dampened stroke-induced elevations in Bdnf mRNA within peri-infarct PFC tissue at four-weeks post stroke. Despite this, young and aged stroke animals presented similar stroke-induced spatial memory impairments in the object location recognition task (OLRT) at four-weeks post-stroke. Tissue collected from all stroke animals (day-56 post-stroke) revealed a significant stroke-induced loss thalamo-PFC connectivity, evidenced by a reduction in cholera toxin subunit-B (CTB)+ve cells in the nucleus reuniens (RE) and mediodorsal thalamus (MD), five-days after injections of CTB into the prelimibic cortex next to the stroke. Lastly, stroke failed to affect VD reversal task performance at 2-3 weeks post-stroke in young animals, but facilitated an improved performance in aged animals. Treatment with the BDNF receptor decoy, TrkB-Fc, dampened this effect, suggesting a BDNF-mediated reopening of a critical window for post-stroke recovery in aged animals.
Given the fundamental role of BDNF in functional recovery following stroke, we investigated the therapeutic potential of BDNF signalling in our aged stroke model. Hydrogel administration of recombinant human BDNF (from day-5 post-stroke) prolonged stroke-induced elevations in Bdnf mRNA at four-weeks post-stroke, but failed to influence performance across all cognitive tasks. Chronic treatment with systemic CX1837, a high-impact AMPAkine that modulates AMPA receptor activity and also elevates BDNF expression, afforded significant improvements to VD reversal learning in aged sham animals and significantly alleviated post-stroke spatial memory impairments in aged animals performing the OLRT. However, CX1837 treatment failed to affect VD reversal learning or thalamo-PFC connectivity in aged stroke animals. The therapeutic potential of two novel glycomimetics (compound A and G) with or without co-administration with rhBDNF, was then assessed in young stroke mice. Both glycomimetics alone failed to affect OLRT performance in stroke animals, despite exerting mild effects of thalamo-PFC connectivity and reactive astrogliosis at day-35 post-stroke (determined by glial-fibrillary acidic protein (GFAP)+ve staining). However, when given in combination with rhBDNF, compound G significantly improved OLRT performance at four-weeks post-stroke, dampened stroke-induced reductions in CTB+ve cells (RE and MD thalamus), and attenuated GFAP+ve staining in the peri-infarct, corpus callosum and infralimbic regions. These data highlight the need for a combinatorial approach for treating strokes occurring in older populations.
Lastly, we aimed to further investigate post-stroke spatial memory impairments observed in our young stroke model using the translational, touchscreen-based TUNL task. By varying the spatial separation and delay between stimuli, we were able to demonstrate novel stroke-induced impairments to both spatial and working memory components of the TUNL task. Moreover, we reported significant positive correlations between the degree of TUNL impairment and GFAP+ve staining in the IL and CC regions at day-56 post-stroke, providing further support for the involvement of reactive astrogliosis in post-stroke cognitive impairment. Overall, this thesis has addressed the importance of using an aged stroke model to better assess age-related differences in stroke pathology. Further, we highlight the importance of neurotrophic support to the ischaemic brain, and identify a novel glycomimetic/rhBDNF treatment that showed promise for promoting functional recovery in our PFC stroke model.