Increased excitatory connectivity and epileptiform activity in thrombospondin1/2 knockout mice following cortical trauma.

ABSTRACT

Thrombospondins (TSPs) are astrocyte-secreted extracellular matrix proteins that play key roles as regulators of synaptogenesis in the central nervous system. We previously showed that TSP1/2 are upregulated in the partial neocortical isolation model ("undercut" or "UC" below) of posttraumatic epileptogenesis and may contribute to abnormal axonal sprouting, aberrant synaptogenesis and epileptiform discharges in the UC cortex. These results led to the hypothesis that posttraumatic epileptogeneis would be reduced in TSP1/2 knockout (TSP1/2 KO) mice. To test the hypothesis, we made UC lesions at P21, and subsequent experiments were conducted 14d later at P35. Ex vivo extracellular single or multi-electrode field potential recordings were obtained from layer V in cortical slices at P35 and in vivo video-EEGs of spontaneous epileptiform bursts were recorded to examine the effect of TSP1/2 deletion on epileptogenesis following cortical injury. Immunohistochemical experiments were performed to assess the effect of TSP1/2 KO + UC on the number of putative excitatory synapses and the expression of TSP4 and HEVIN, other astrocytic proteins known to up-regulate excitatory synapse formation. Unexpectedly, our results showed that, compared with WT + UC mice, TSP1/2 KO + UC mice displayed increased epileptiform activity, as indicated by 1) increased incidence and more rapid propagation of evoked and spontaneous epileptiform discharges in UC neocortical slices; 2) increased occurrence of spontaneous epileptiform discharges in vivo. There was an associated increase in the density of VLUT1/PSD95-IR colocalizations (putative excitatory synapses) and significantly upregulated TSP4- and HEVIN-IR in TSP1/2 KO + UC versus WT + UC mice. Results suggest that TSP1/2 deletion plays a potential epileptogenic role following neocortical injury, associated with compensatory upregulation of TSP4 and HEVIN, which may contribute to the increase in the density of excitatory synapses and resulting neural network hyperexcitability.