Parvalbumin immunoreactivity and expression of GABAA receptor subunits in the thalamus after experimental TBI

- Only 64% of neurons are left in the ipsilateral reticular nucleus after TBI.
- Ventroposterior complex has increased PARV-immunolabeling after TBI.
- In controls seizure susceptibility associates with thalamic PARV immonolabeling.
- Correlation between seizure susceptibility and PARV labeling was lost after TBI.

Traumatic brain injury (TBI) causes 10-20% of acquired epilepsy in humans, resulting in an ictogenic region that is often located in the cerebral cortex. The thalamus provides heavy projections to the cortex and the activity of thalamocortical pathways is controlled by GABAergic afferents from the reticular nucleus of the thalamus (RT). As rats with TBI induced by lateral fluid-percussion injury (FPI) undergo epileptogenesis, we hypothesized that damage to the parvalbumin (PARV)-immunoreactive (ir) neurons in the RT is associated with seizure susceptibility after lateral FPI. To address this hypothesis, adult Sprague-Dawley rats (n = 13) were injured with lateral FPI. At 6 months post-TBI, each animal underwent a pentylenetetrazol (PTZ) seizure susceptibility test and 2 weeks of continuous video-electroencephalography (EEG) monitoring for detection of the occurrence of spontaneous seizures. Thereafter, the brain was processed for PARV immunohistochemistry. We (a) estimated the total number of PARV-ir neurons in the RT using unbiased stereology, (b) measured the volume of the ventroposteromedial (VPM) and ventroposterolateral (VPL) nuclei of the thalamus, which receive PARV-ir inputs from the RT and project to the perilesional cortex, (c) quantified the density of PARV-ir terminals in the VPM-VPL, and (d) studied the expression of GABAA receptor subunits in a separate group of rats using laser-dissection of the thalamus followed by Real-Time polymerase chain reaction (RT-PCR) array studies. At 6 months post-TBI, only 64% of PARV-ir neurons were remaining in the RT ipsilaterally (p < 0.001 as compared to controls) and 84% contralaterally (p < 0.05). Accordingly, the volume of the ipsilateral RT was 58% of that in controls ipsilaterally (p < 0.001) and 90% contralaterally (p > 0.05). Also, the volume of the VPM-VPL was only 51% of that in controls ipsilaterally (p < 0.001) and 91% contralaterally (p < 0.05). The density of PARV-ir axonal labeling was remarkably increased in the lateral aspects of the VPM and VPL (both p < 0.001). Expression of the ε- and θ-subunits of the GABAA receptor was down-regulated (0.152, p < 0.01 and 0.302, p < 0.05, respectively), which could relate to the inclusion of the hypothalamus into the tissue analyzed with RT-PCR arrays. In controls, the lower the number of PARV-ir neurons in the RT, the higher the seizure susceptibility in the PTZ test. Rats with TBI showed seizure susceptibility comparable to that in controls with the lowest number of PARV-ir neurons in the RT. Our data show that the RT and VPM-VPL undergo remarkable degeneration after lateral-FPI which results in reorganization of PARV-ir terminals in the VPM-VPL. The contribution of RT damage to seizure susceptibility and post-traumatic epileptogenesis deserves further studies.