• biomarker
• reprogramming
• stroke

Stroke is one of the major causes of adult disability worldwide. While some subjects display a significant restoration of function, the majority of stroke patients still suffer permanent deficits, especially in the motor domain. Currently, there are no ways to determine either the extent or time-course of recovery in individual subjects. Great is therefore the need for reliable biomarkers predictive of spontaneous recovery and responsive to rehabilitation to allow a better patient stratification in clinical trials and to personalize therapies maximizing the final outcome. In our study we took advantage of a mouse model of middle cerebral artery occlusion (MCAO), which is the main cause of ischemic insult in humans, to reproduce an highly variable sensorimotor deficit of the mouse forelimb. Spontaneous recovery was assessed at different time points using gridwalk and skilled reaching tests, while forelimb strength performance was measured with a robotic device, the M-platform. Mice were also implanted with chronic electrodes in the caudal forelimb area (CFA) to record local field potentials (LFPs) from both hemispheres during the retraction test in the M-platform and in freely moving condition. In this context, I set up novel and quantitative methods to evaluate lesion size and location in histological brain sections obtained 30 days after injury. Specifically, I calculated the fraction of the CFA affected by the lesion, the overall volume of damage and the relative shrinkage of the affected hemisphere. These structural changes, together with electrophysiological parameters, will be then correlated with the extent of spontaneous recovery, to predict motor outcome and response to therapy after stroke.
From the therapeutical point of view, one fascinating approach in this field is represented by transdifferentiation of resident reactive glia. Between different cortical cells, astrocytes are particularly promising for reprogramming into neurons as they maintain patterning information from their radial glial ancestors. The group of Magdalena Gotz in Munich, who is collaborating with us on this project, has recently obtained data that support this idea. They demonstrated that forced expression of Neurog2 and Nurr1 genes in astrocytes following a stab wound injury leads their reprogramming towards pyramidal neurons of all different layer positions. Reprogrammed cells display long distance axonal projections through the callosum and to subcortical targets, supporting the layer-specific identity of neurons achieved by reprogramming of astrocytes at different layer positions. However, if a functional recovery occurs and if the same approach could be used with different types of brain injury is still unknown.
We investigated these two aspects using mice which underwent a photothrombotic focal cortical stroke, highly reproducible in term of position and size. We injected flexed AAV vectors coding for Neurog2, Nurr1 and GFP into the perilesional cortex of GFAP-Cre mice (reprogramming condition), 3 days after photothrombosis. Flexed AAV vectors coding for GFP only were injected in stroke controls. Our results demonstrate that over 20% of the GFP-labelled cells expressed NeuN in the reprogramming condition, indicative of a successful reprogramming toward a neuronal phenotype. Less than 5% of GFP/NeuN double labeled cells were observed in controls (t-test, p < 0.05, control vs reprogramming). In addition, the reprogrammed cells displayed consistent GFP-positive projections to distinct corticospinal targets, including the internal capsule and the spinal cord. At the behavioral level, I used the gridwalk and Schallert cylinder tests to ascertain the impact of cell reprogramming on motor recovery. 60 days after the lesion, I found no significant differences in motor outcome between the reprogrammed and control groups, probably due to an incomplete integration of the newly neurons. Our current investigation are aimed at clarify if this integration could be guided by optogenetic or chemogenetic stimulations in phase with robotic training during the post- stroke rearrangement. Due to Hebbian plasticity, this protocol is expected to lead to strengthening and stabilization of connections between the newly generated cells and the host circuitry with more effective neuronal integration, thus improving functional recovery.