Modelling perinatal stroke and stem cell grafting in a neonatal rat : $$b modelling and treating perinatal stroke in an animal model

Abstract

Perinatal ischemic stroke (PIS) in humans occurs mostly at full term and is a significant cause of hemiplegic cerebral palsy. A suitable animal model is needed to test early intervention therapies. Firstly, we developed and compared two PIS Models; middle cerebral artery occlusion (MCAO) at the level of the temporal bone, and the injection of reversible vasoconstrictor endothelin-1 (ET-1) into the sensorimotor cortex (SMC), at postnatal day (P) 12. Appropriate sham procedures were performed. Animals underwent behavioural testing (cylinder and grid walking tests) at P35-40 followed by immunohistological examination of the brain for markers of inflammation and hypoxia. We found that MCAO is a poor stroke model in P12 rat pups with minimal involvement of the SMC, a major site of damage in human neonates, however ET-1 models did induce focal ischemia in the SMC. However, no significant behavioural deficits were detected in either model although there was a trend towards a deficit in the ET-1 animals. In a second study, ET-1 or sham treated animals received a unilateral injection of retrograde tracer into the contralesional cervical spinal cord at P40. There was depletion of corticospinal (CS) neurons in the ipsilesional hemisphere but an increase in labelled CS neurons in the contralesional cortex. An upregulated ipsilateral corticospinal pathway is a feature of human hemiplegic CP again suggesting that ET-1 model is an appropriate model. We found preliminary evidence that this nervous system plasticity could rescue behavioural performance. Finally, we attempted an intervention therapy to repair the corticospinal tract by grafting human neural stem cells into the SMC of ET-1 lesioned animals. Cells were dispersed in a semi-synthetic extracellular matrix (1x106 cells per μl) and either cultured in vitro or transplanted in vivo to the lesioned SMC at P14. Rosettes of cells resembling cerebral organoids and expressing human specific NCAM were observed in vivo at the graft site after a month but not in in vitro cultures. The organoids comprised a dense cellular layer expressing neural progenitor cell markers arranged around a small lumen surrounded by more loosely packed cells with neurites expressing markers for post-mitotic neurons. Some host cells and blood vessels infiltrated the graft and after three months the organoids had broken down and very few transplanted neurons had integrated with host tissue. ii Animal modelling of hemiplegic cerebral palsy continues to present a significant challenge but some progress has been made. The feasibility of using neural stem cells for perinatal brain repair has been tested with unexpected results.