Taken together, our data demonstrate chronic microglial activation after experimental TBI, which may contribute to the observed progressive neurodegeneration and tissue loss that is associated with functional impairments. Furthermore, they demonstrate that the temporal window for neuroprotective intervention after TBI is significantly longer than generally believed. These observations are consistent with prior experimental work demonstrating progressive cortical damage and increased NFκB activation in macrophages and/or microglia up to one year after trauma [12, 38, 39]. Prolonged microglial activation has been demonstrated many months after TBI in humans  and increased microglial activation has also been observed up to four years after TBI in human post-mortem tissue .
We previously reported that CHPG treatment, acting through mGluR5, reduces microglial activation and the associated release of free radicals and pro-inflammatory cytokines in microglial cell culture models (both primary cultures and a mouse microglial cell line) after stimulation with the classical activators lipopolysaccharide or interferon-γ [19, 21]. CHPG treatment also abolished the neurotoxic potential of activated microglia and reduced NADPH oxidase activity in such in vitro models. These effects of CHPG were blocked by knockout of the mGluR5 receptor or by addition of the selective mGluR5 antagonist MTEP, and reduced by co-incubation with siRNAs directed against either of the two membrane subunits of NADPH oxidase . The current study supports these findings, as MTEP administration prior to CHPG treatment blocked its protective actions.
Examining the injured cortex one month after TBI, we demonstrated that mGluR5 expression was up-regulated in activated microglia that co-expressed the phagocytic marker, ED1, and exhibited a hypertrophic or bushy cellular phenotype. The observed post-traumatic expression of mGluR5 in activated microglia is consistent with previous in vivo studies demonstrating microglial mGluR5 expression at the lesion site in spinal cord injury and excitotoxic brain injury models [20, 40]. In response to trauma, there is both microglial proliferation and activation, along with migration to the site of injury [41, 42]. In contrast to the ramified appearance of resting microglia, activated microglia undergo substantial morphological changes, with reduction and thickening of processes leading to a hypertrophic or bushy appearance . Although microglia are believed to have both neurotoxic and neuroprotective properties [3, 5, 43], considerable experimental data suggest that post-traumatic inflammation can contribute to delayed cell and tissue loss [35, 44, 45]. Indeed, microglial activation and release of associated inflammatory factors has been proposed as an important contributing factor for many acute and chronic neurodegenerative disorders [7, 46].
Here we demonstrate that delayed CHPG administration significantly reduced the number of microglia showing the reactive bushy or hypertrophic phenotypes associated with a pro-inflammatory and potentially neurotoxic state . Moreover, activated microglia expressed the NADPH oxidase sub-unit, gp91phox four months post-injury, suggesting the chronic expression of NADPH oxidase in these cells; these data are consistent with the chronic up-regulation of expression of NADPH oxidase sub-units in a microglial-associated gene cluster six months after spinal cord injury . Activated microglia cause neuronal cell death in culture through mechanisms that involves NADPH oxidase activation, and this process is inhibited by CHPG treatment [19, 21]. Activated microglia also have been implicated in chronic functional deficits after TBI in humans . Recently, inhibition of NADPH oxidase has been shown to reduce microglial activation in the post-ischemic brain . Notably, single dose CHPG administration one month after TBI reduced the expression of NADPH oxidase in reactive microglia at four months post-injury. Together, these data suggest a positive activation feedback loop for neuroinflammation that contributes to delayed neurodegeneration and related functional deficits. Interruption of this positive feedback loop may explain why even a single injection of CHPG administered one month post-injury inhibits chronic neuroinflammation and limits functional loss in our model.
mGluR5 is also expressed in other cell types of the CNS, such as neurons, astrocytes and oligodendrocytes [49, 50]. In addition to their anti-inflammatory effects, group I mGluR agonists also reduce neuronal apoptosis  as well as oligodendrocyte cell death . Although it is possible that anti-apoptotic effects of CHPG on neurons or oligodendrocytes may have contributed to the recovery observed in the current study, such apoptotic processes so late after injury are likely limited and, given the half life of the compound, modulation of such events would probably contribute minimally at best to the improved outcome observed. In order to clarify the mechanism underlying mGluR5-mediated neuroprotection after TBI, future studies will require cell specific mGluR5 knockout in neurons, astrocytes, oligodendrocytes and microglia.
It is widely accepted that the therapeutic window for limiting post-traumatic neurodegeneration after acute brain injury is limited . Indeed, most experimental treatment studies for TBI have focused on the first hours after injury . With the recognition that more delayed apoptotic mechanisms may contribute to injury [54, 55], the potential therapeutic window has been expanded to perhaps 24 to 72 hours. Although it has been shown experimentally and, more recently, clinically that tissue loss after TBI may progress for months or longer [12, 14, 15, 56, 57], there have been few attempts to pharmacologically modify such markedly delayed neurodegeneration. Here we demonstrate that single dose treatment one month after trauma significantly reduced both histological changes and behavioral dysfunction over a subsequent period of months. Considerable tissue sparing was observed in hippocampal and cortical regions after CHPG treatment, which was associated with significant improvements in both sensorimotor (beam walk) and cognitive (MWM) function.
Chronic neurological deficits are characteristic of moderate to severe TBI, although such deficits may stabilize or improve over time - likely reflecting endogenous plasticity. The DTI studies extend the observations from T2-weighted MRI by demonstrating significantly better preservation of white matter tracks in the brain at four months in the CHPG-treated TBI group as compared to controls. DTI detects directionality of water diffusion. After TBI, FA increases, whereas mean diffusivity, or the directed diffusion of water, typically along white matter tracks, is reduced [58, 59]. These techniques have been used to reflect the integrity of white matter tracks (tractography) in the CNS . FA has been shown to be negatively correlated with deficits in memory performance , consistent with our observations that CHPG treatment reduced this measure while improving cognitive performance.