Mild TBI comprises almost 80% of clinical TBI. Despite continuing research and accumulated knowledge, an effective treatment for mild TBI is still not available. In the present study, we have adopted the LFPI model of TBI originally characterized by McIntosh et. al. [19
] to develop a methodology that results in quantifiable reproducible injury. Because pressure pulses within the range used here (2.0-2.2 atm.) are generally considered to reflect mild injury in the rat model [14
], this paradigm is particularly attractive in that it lends relevance to the clinical population suffering from mild injury. However, conflicting data in the literature regarding the regional and temporal injury distribution prompted us to conduct a comprehensive investigation throughout the brain to determine where approximately 80% of the injury is found at 24 and 48 h post-impact. These endpoints were selected because they represent a delayed window within the secondary injury phase that can be targeted by novel therapeutics.
Consistent with previous studies using the LFPI [20
] FJ and TUNEL staining in this study showed that the predominant areas of neurodegeneration and apoptosis include the cerebral cortex, hippocampus, and thalamus. In a previous study using similar TBI methodology, Sato et al. [21
] showed Fluoro-Jade and TUNEL staining that persisted from 3 hours to 7 days and included cerebellar damage in addition to damage in those regions identified here. Furthermore, we have demonstrated that LFPI-induced neurodegeneration paralleled with increase in activated microglia in the injured brain. Also, in accordance with the observation of Stahel et al., the dying neurons showed characteristics of both necrosis and apoptosis [22
]. These observations indicate that substantial inflammation takes place in the brain parenchyma in response to mild LFPI in this study.
The nature and progression of TBI-induced brain pathology limits the goals of treatment to either blocking the secondary injury phase or facilitating plasticity and repair at some point after the initial impact. The secondary phase is largely a result of the migration of activated microglia towards the site of injury, secreting toxic cytokines and oxygen radicals and thereby causing further neuronal damage [13
]. Lunemann et. al. [25
] have shown that following the formation of a brain lesion, microglia invade the damaged brain tissue after maturing and becoming activated by producing macrophage activating factor (MAF) in a CD11b-positive pathway. In agreement with this finding, we have also observed that within 24 h of initial damage the brain parenchyma is invaded by activated microglial cells. This indicates that an active inflammatory reaction is generated locally in the brain as early as 24 hours after injury.
The spleen is a reservoir of peripheral macrophages and other immune cells in the body, and it is now well known that splenic signalling contributes to injury of various tissues after ischemic insult. For example, splenectomy prior to insult protects both the liver [26
] and brain [8
] from ischemic damage. In a recent study, Li et al. have shown that splenectomy immediately after TBI in rats decreased[18
] proinflammatory cytokine production and mortality rate and improved cognitive function. In our study, we observed that splenectomy immediately after induction of TBI attenuated TBI-induced neurodegeneration and CCL20 expression in the brain. Although it is not clear how this spleen-brain interaction takes place, Lee et al. [27
] suggested that vagal nerve stimulation may reduce immune cell infiltration and consequent decrease in brain inflammation and edema while Stewart and McKenzie [28
] suggested a role of sympathetic stimulation in causing the release of immune cells from spleen and subsequent infiltration into the brain tissues. Regardless of the neural mechanism, removal of the spleen immediately after the insult would remove the largest pool of immune cells, resulting in decreased infiltration and consequent neuroinflammation. Our study clearly shows that reduction in the splenic immune cell population reduced neuronal damage and CCL20 production.
Interestingly, CCL20 is a unique chemokine known to interact specifically with CC chemokine receptor 6 (CCR6) and induce chemotaxis of dendritic cells, T cells and B cells [29
], all of which reside in the spleen and have the potential to promote neuroinflammation. Several lines of evidence support this hypothesis. Ohta et al. [30
] have shown that CCL20 was up-regulated under normothermic conditions in a rat middle cerebral artery occlusion (MCAO) model. CCL20 is also expressed in inflamed epithelial cells [31
] and in the synovial tissues of rheumatoid arthritis patients [32
], while up-regulation of CCL20 along with other cytokines has been observed in human subjects one day after severe traumatic brain injury [34
]. Furthermore, a recent study identified CCL20 as a dual-acting chemokine with the potential for inhibiting immune reactions and more importantly in attracting inflammatory effectors and activators [35
]. Although a great deal of investigation has recently been done to elucidate the relationship between brain trauma and the immune system, very little is known about the function of the thymus after brain trauma. Since the thymus is the major source of mature circulating T cells, CCL20 expression in the thymus as observed in this study seems significant, although further investigation is needed to identify the specific function of thymus after TBI in adult rats.
Because CCL20-CCR6 signalling is now known to facilitate the immune response in pathological circumstances, data from the present study demonstrating up-regulated CCL20 in spleen and thymus 24 h post-LFPI likely reflects the initiation or persistence of a systemic signal that drives neural inflammation and cell death. Mouse models of autoimmune encephalomyelitis (EAE) have provided some evidence that T cells may be targeted by the splenic signal. A recent knockout study demonstrated that CCR6 modulates the infiltration of T cells into the brains of EAE-infected mice, although reduced infiltration of Treg in CCR6-/- mice was associated with increased neurological damage [36
]. Despite evidence of a protective role for CCR6 activation, CCL20 signaling through CCR6 on Th1 or Th17 cells, rather than Treg cells, would be expected to promote inflammation. CCR6 is constitutively expressed in the choroid plexus of mouse and human and there are data showing that the binding of CCL20 to CCR6 on Th17 cells is critical for T cell infiltration into the CNS through the choroid plexus [37
]. Indeed, T cells are well known for infiltrating the brain in neural injury models characterized by a compromised BBB. Because BBB degradation is also a critical component of TBI [19
], peripheral CCL20 signalling may be an important initiator of T cell chemotaxis and extravasation into the brain parenchyma.
Data presented in this report also show that CCL20 was not expressed in degenerating cortical or hippocampal cell layers until 48 h after the impact. This raises the question of why cortical and hippocampal neurons expressed CCL20 at 48 h, which is 24 h after the systemic expression of the same chemokine and the neurodegeneration in the same areas of the injured brain. Although CCL20 is produced by astrocytes in response to bacterial infections [38
] and EAE [39
], to the best of our knowledge, ours is the first report to demonstrate neuronal expression of CCL20. One possibility is that cellular injury induces expression of CCL20 as a signal for peripheral or local immune cell recruitment to the injured site. If so, it is also likely that the neuronal cells that expressed CCL20 were in the immediate vicinity of those cells undergoing neurodegeneration. However, another possibility is that neuronal CCL20 expression is a 'tombstone' marker in cells that are beyond repair and need to be removed from the surrounding viable tissues. This latter explanation is supported by the pyknotic morphology that was observed in CCL20-expressing neurons, as well as the fact that the areas surrounding the cell bodies appeared to be devoid of tissue. The morphological analysis, anatomical localization and colocalization with FJ and NeuN protein of CCL20-positive cells strongly suggest that neurons represent the predominant cell type expressing this chemokine following TBI. Preliminary observations from this laboratory indicate downregulation of peroxysome proliferator-activated receptor γ (PPARγ) in neuronal cells (data not shown) after TBI; however, a causal role of PPARγ in regulating CCL20 signalling and/or expression in these cells remains to be established.
While results here demonstrate a link between CCL20 expression and LFPI-induced injury and indicate involvement of peripheral immune organs like the spleen in this response, further experiments are required to define the precise mechanisms by which CCL20 signalling contributes to cell death and the exact role played by spleen and thymus in inducing neuronal death. Furthermore, if CCL20 exerts direct actions on neurons, the 11 kDa protein could easily enter the CNS from the systemic circulation and promote injury even in the absence of peripheral leukocyte recruitment. If this latter scenario is the case, plasma CCL20 levels could be utilized as an important biomarker indicating the presence and severity of TBI.