Lateral fluid percussion injury of the brain induces CCL20 inflammatory chemokine expression in rats

Animals were anesthetized using a mixture of ketamine (90 mg/kg)/xylazine (10 mg/kg) (IP). To deliver LFPI, a 1 mm diameter craniotomy was performed centered at 2 mm lateral and 2.3 mm caudal to the bregma on the right side of the midline. A female luer-lock hub was implanted at the craniotomy site and secured with dental cement. The FPI device was then fastened to the luer-lock. All tubing was checked to ensure that no air bubbles had been introduced, after which a mild impact ranging from 2.0-2.2 atm. was administered [14]. Impact pressures were measured using a transducer attached to the point of impact on the fluid percussive device. The luer-lock was then detached, the craniotomy hole was sealed with bone wax and the scalp was sutured. Ketoprofen (5 mg/kg) was administered to minimize postsurgical pain and discomfort. Rats were then replaced in their home cages and allowed to recover for 24-48 h prior to subsequent experiments. Animals were excluded from further tests if the impact did not register between 2.0 and 2.2 atm. or if the dura was disturbed during the craniotomy prior to impact. In sham (control) animals, craniotomy was performed at the same coordinates as the TBI animals but no impact was delivered.

Animals were deeply anesthetized with ketamine (75 mg/kg) and xylazine (7.5 mg/kg) 24 or 48 hours after TBI. Thymuses and spleens were removed and immediately snap frozen on dry ice. Animals were then perfused with 0.9% saline followed by 4% paraformaldehyde in phosphate buffer (pH 7.4). The brains were harvested, post-fixed in 2% paraformaldehyde and saturated with increasing sucrose concentrations (20% to 30%) in phosphate-buffered saline (PBS, pH 7.4). Brains were then frozen, sectioned coronally at 30 μm thickness using a cryostat, thaw-mounted onto glass slides and stored at -20°C prior to staining. In the initial studies 80% of the injured neurons were found in the brain region between 3.5 and 5.5 mm caudal to the bregma. Therefore, for all subsequent staining experiments, three sections from each brain corresponding to 3.5, 4.5, and 5.5 mm caudal to the bregma were selected for analysis.

Total RNA was extracted from 50 mg of frozen spleen tissue using TRIZOL reagent (Invitrogen, Carlsbad, CA). Briefly, the samples were homogenized with 1 ml of TRIZOL, incubated at room temperature for 5 minutes and phase-separated by chloroform. Total RNA was precipitated by isopropyl alcohol, collected by centrifugation and purified using an RNeasy mini kit (Qiagen, Valencia, CA). The RNA concentration and purity was determined by spectrophotometry at 260/280 nm and 260/230 nm. First strand cDNA was synthesized from the isolated RNA using the Superscript III system (Invitrogen).

A panel of proinflammatory cytokines and chemokines and their receptors was analyzed using a SYBR green-optimized primer assay (RT² Prolifer PCR Array) from SA bioscience (Frederick, MD). Briefly, cDNA was synthesized from fresh frozen spleens as stated above. cDNA was mixed with the RT2 qPCR master mix and the mixture was aliquoted across the PCR array. The PCR was done in a CFX96 Real-Time C1000 thermcycler (BioRad) for 5 min at 65C, 50 min at 50C and 5 min at 85C. Control gene expression was normalized and target gene expression was expressed as fold increase or decrease compared to control. PCR data were analyzed using the SA Bioscience Excel program.

Spleen tissue lysates were prepared from 5 mg of fresh frozen tissue using protein lysis buffer containing NP-40. CCL20 was estimated by ELISA using the DuoSet ELISA Development kit for CCL20 from R & D systems (Minneapolis, MN). Briefly, 96 well sterile ELISA microplates were coated with anti-rat CCL20α antibody overnight at room temperature. Next day, the plates were washed and blocked with bovine serum albumin (BSA). Plates were incubated sequentially with standards or samples for 2 h, detection antibody (biotinylated goat anti-rat CCL20α antibody) for 2 h, streptavidin-HRP for 20 minutes and substrate solution (1:1 mixture of H₂O₂ and tetramethylbenzidine) for 20 minutes. Reactions were stopped with 2N H₂SO₄. All incubations were performed at room temperature and the microplate was thoroughly washed after each incubation. The absorbance of each well was determined at 450 nm using a Synergy H4 Hybrid reader (BioTek). Total protein concentrations from the same samples were determined by BCA protein assay (Pierce). CCL20 was expressed as pg per μg of total protein in the tissue.

Fluoro-Jade (Histochem, Jefferson, AR) staining was performed to label degenerating neurons. This method was adapted from that originally developed by Schmued et at [15] and subsequently detailed by Duckworth [16]. Thaw-mounted sections were placed in 100% ethanol for 3 minutes followed by 70% ethanol and deionized water for 1 minute each. Sections were then oxidized using a 0.06% KMnO₄ solution for 15 minutes followed by thee rinses in ddH2O for 1 minute each. Sections were then stained in a 0.001% solution of Fluoro-Jade in 0.1% acetic acid for 30 min. Slides were rinsed, dried at 45°C for 20 min, cleared with xylene, and cover-slipped using DPX mounting medium (Electron Microscopy Sciences, Ft. Washington, PA).

Nuclear DNA fragmentation, a marker of apoptotic cells was measured using the DeadEnd Fluorimetric TUNEL system (Promega, Madison, WI). Fixed cryosections (30μ thick) were permeabilized with 20 μg/ml proteinase K at room temperature for 8 minutes followed by 4% PFA in PBS for 5 minutes. The sections were washed in PBS and equilibrated with 200 mM potassium cacodylate, pH 6.6; 25 mM Tris-HCl, pH 6.6; 0.2 mM DTT; 0.25 mg/ml BSA and 2.5 cobalt chloride (equilibration buffer) for 10 minutes at room temperature. The sections were then incubated at 37°C for 1 hour with incubation buffer containing equilibration buffer, nucleotide mix and rTdT enzyme mix, covered with plastic cover slip and placed away from exposure to light. The cover slips were removed and the reactions were stopped with 2X SSC. The sections were then washed with PBS and mounted with VectaShield mounting medium containing DAPI. The green fluorescence of fluorescein-12-dUTP was detected in the blue background of DAPI under the fluorescence microscope. Images were taken and apoptotic nuclei were quantified using the Image J quantitation program.

Spleen, thymus or brain tissue sections were washed with PBS for 5 min, incubated in 3% hydrogen peroxide for 20 min and washed 3 times in PBS. They were then heated in antigen unmasking solution (1:100; Vector Laboratories Inc., Burlingame, CA) for 20 min at 90°C, incubated for 1 h in permeabilization buffer (10% goat serum, 0.1% Triton X-100 in PBS) and incubated overnight at 4°C with either rabbit anti-CCL20 primary antibody (1:1000) or mouse monoclonal anti-CD11b antibody (1:400) (Abcam, Cambridge, MA) in antibody solution (5% goat serum, 0.05% Triton X-100 in PBS). The following day, sections were washed with PBS and incubated 1 h at room temperature with secondary antibody (biotinylated goat anti-rabbit, 1:400, Vector Laboratories Inc., Burlingame, Ca or Alexafluor 594 conjugated antimouse antibody, 1:50 or DyLight 594 conjugated antirabbit antibody, 1:50) in antibody solution. Sections incubated with biotinylated antirabbit antibody were then washed in PBS, incubated in avidin-biotin complex mixture (ABC,1:100; Vector Laboratories Inc, Burlingame, Ca) for 1 h, washed again and visualized using DAB/peroxide solution (Vector Laboratories Inc). After three washes, sections were dried, dehydrated with increasing concentrations of ethanol (70%, 95%, 100%), cleared with xylene and cover-slipped with Vectamount mounting medium. Sections incubated with mouse anti-CD11b antibody followed by alexafluor 594-conjugated antimouse antibody were washed three times with PBS and used for double staining with IB4. Some of the anti-CCL20 antibodies followed by DyLight 594-conjugated antirabbit antibody treated sections were incubated with Alexa fluor 488-conjugated mouse antineuronal nuclei (NeuN) monoclonal antibody (1:100; Millipore, Temecula, CA) 3 hours at room temperature, washed with PBS, dried and cover slipped with vectamount mounting medium with DAPI.

Slide mounted sections were washed in PBS and CCl20 immunostaining was performed as described above and developed with DyLight 594 conjugated anti rabbit antibody. Sections were then incubated in acidic 0.0001% FJ solution for 20 min on shaker. Slides were washed, dried and cover slipped with Vecta Shield mounting medium.

Brain sections were washed with modified PBS (PBS with 0.5mM CaCl₂, pH 7.2) and permeabilized with buffer containing 10% goat serum, 3% lysine, 0.3% triton X-100 in modified PBS for 1 hour at room temperature. Brain sections already immunostained were transferred to modified PBS. Sections were then incubated overnight at 4°C with 5 μg/ml Alexa 488-conjugated isolectin IB4 (Molecular Probes) dissolved in modified PBS with 0.3% triton X-100 and 2% goat serum. Stained sections were washed with modified PBS, mounted with Vecta-Shield mounting medium with DAPI and viewed with an Olympus IX71 fluorescent microscope using the FITC filter. Images were taken using the Olympus DP70 imaging system and IB4-positive cells were quantified using the Image J quantitation program.

All quantitation was performed using the NIH Image J software. For immunohistochemical analysis, images were acquired using an Olympus IX71 microscope controlled by DP70 manager software (Olympus America Inc., Melville, NY). Photomicrographs captured at 200x magnification with an Olympus DP70 camera were used for quantification. Images were taken at the same exposure and digital gain settings for a given magnification to minimize differential background intensity or false-positive immunoreactivity across sections. The channels of the RGB images were split and the green channel was used for quantitation of the FJ, IB4 and TUNEL staining images. The CCL20 images were converted to gray-scale before quantitation. The single channel or gray-scale images were then adjusted for brightness and contrast to exclude noise pixels. The images were also adjusted for the threshold to highlight all the positive cells to be counted and a binary version of the image was created with pixel intensities 0 and 255. Particle size was adjusted to exclude the small noise pixels from the count. Circularity was adjusted to between 0 and 1 to discard any cell fragments, processes or tissue aggregates resulting in false labelling from the quantitation. The same specifications were used for all sections. Cell counts of sections from 3.5, 4.5 and 5.5 mm caudal to the bregma were summed to represent the number of positive cells from each brain. The results for the FJ, TUNEL, IB4 and CCL20 immunoreactivity were expressed as mean number of positive cells ± S.E.M. CCL20 immunoreactivity of the thymus or the spleen was expressed as mean area of immunoreactivity ± S.E.M.

Inconsistencies in injury assessment across laboratories and lack of a reliable, quantitative approach to assessing neural injury have impeded efforts to develop novel treatments for TBI pathology. Therefore, a detailed investigation throughout the brain was sought to determine which regions show consistent, prominent neurodegeneration in rats subjected to mild LFPI (Figure 1). A consistent profile emerged in which the majority of Fluoro-Jade (FJ)-positive cells were found within the cerebral cortex (Figure 1), hippocampus (Figure 1), and thalamus (Figure 1). Cortical Fluoro-Jade was ubiquitous and was present at various levels throughout the brain. Hippocampal FJ staining was localized to the pyramidal cell layers (Figure 1), while some diffuse labelling throughout the general structure was also evident. The thalamic staining was diffuse and sparsely distributed. Quantitation revealed that the neurodegeneration in these regions significantly increased at both 24 and 48 h post-impact relative to sham-operated controls. Additionally, data showed that FJ-stained degenerating hippocampal neurons were restricted to the ipsilateral hemisphere, whereas few cortical and thalamic FJ-positive neurons were also detected in the contralateral hemisphere in some animals.

Internucleosomal DNA fragmentation, an important marker for apoptotic cells, was assessed by terminal deoxynucleotidyl transferase biotin-dUTP nick end labelling (TUNEL) histochemistry. Few TUNEL-positive cells were detected in the contralateral hemisphere, and, while the ipsilateral thalamus showed sparse TUNEL staining in some sections, this was not a consistent finding throughout the experiment (data not shown). The majority of TUNEL-stained nuclei were detected at 24 h post-TBI in the ipsilateral cortex (Figure 2A) and hippocampus (Figure 2B), while sections from sham-operated controls were predominantly devoid of TUNEL staining in these regions (Figure 2A, Figure 2B) and showed only background levels of fluorescence. By 48 h after TBI, sections showed very few TUNEL-positive cells in the cortex and hippocampus and resembled sham-operated controls. Quantitation revealed a significant increase in TUNEL-positive cells in both cortex and hippocampus 24 h post TBI as compared to sham-operated control groups (Figure 2C).

Isolectin-IB4, a 114 kD protein isolated from the seeds of the African legume, Griffonia simplicifolia has been shown to have a strong affinity for resident microglia in the central nervous system and peripheral macrophages that are activated in response to neural injury. To assess the local inflammatory response following mild TBI, Alexa-Fluor 488-conjugated IB4 was used to label microglia/macrophages in the brain tissue. While IB4 labelling was primarily restricted to the ipsilateral hemisphere, sparse labelling was detected within the contralateral hippocampus (data not shown). IB4-positive cells were abundant in the hippocampus, especially in the dentate gyrus (Figure 3A). Microglia were also found in the cortex and thalamus (data not shown) following TBI. CD11b, an activated microglial marker, was also found in the cells of the cortex and hippocampus (dentate gyrus, Figure 3A) of the ipsilateral side. Confocal microscopy revealed that most but not all IB4⁺ cells in the cortex or hippocampus were also CD11b⁺ (Figure 3A). Quantitation showed that the number of IB4-positive cells was significantly increased in each of these brain regions 24 h after TBI, while number of IB4⁺ cells in these regions 48 h post-TBI did not significantly differ from sham-operated controls (Figure 3B). These observations indicate that an inflammatory response was mounted within the brain parenchyma as early as 24 h after the injury involving microglial activation/ migration to the site of injury.

Several studies have suggested that in addition to the local response, activation of the systemic inflammatory response is critical in inducing TBI-associated neuropathies. Although a number of cytokines and chemokines have been studied, the key systemic inflammatory molecules have not yet been identified. Because the spleen has been shown to be involved in the systemic inflammatory response in various injury models, SuperArray analysis was performed on spleen RNA from three separate experiments to identify alterations in the expression of genes associated with pro-inflammatory signalling after LFPI (Figure 4). SuperArray data indicates that more genes were down-regulated (Figure 4B) than were up-regulated (Figure 4A). Among the genes that were up-regulated, CCL20 was uniquely up-regulated by five-fold compared to controls (Figure 4A) 24 h after TBI. These studies led to the identification of CCL20 as a potentially important pro-inflammatory, systemic marker of TBI. To confirm this observation as well as to determine whether alterations in CCL20 mRNA paralleled protein expression, ELISAs and immunohistochemistry were performed on spleen tissues. Immunohistochemistry on spleen tissues indicated significant up-regulation of CCL20 expression at 24 h after TBI as indicated by the increase in mean area of CCL20 intensity. Significant expression of the protein was also observed 48 h after impact (Figures 5A, B). The immunohistochemical observation was further supported by the data obtained from ELISA of spleen tissues showing at least two-fold up-regulation of CCL20 protein expression 24 h after TBI (Figure 5C). In addition to spleen, the thymus also expressed CCL20 at 24 h after TBI as evident from the immunohistochemical labelling of thymus (Figure 5A and 5B) and ELISA for CCL20 of thymic tissues (Figure 5C). These observations support the notion that CCL20 chemokine signalling contributes to the systemic inflammatory response, and that the spleen and thymus respond as early as 24 h after TBI.

Data from the regional injury distribution experiments showed that mild TBI resulted in highly reproducible cellular injury within the cortex as well as the hippocampus. Because splenic CCL20 expression was increased in the acute phase of TBI injury (24 h post-insult) and the splenic inflammatory response is known to exacerbate neural injury [10, 17, 18] experiments were performed to determine whether CCL20 expression is associated with neural injury. Brain sections from animals subjected to mild TBI or sham-TBI were immunostained for CCL20 expression using an antibody generated against the same CCL20 antigen that was used to immunostain the spleen and thymus sections (Figure 6).

CCL20 immunoreactivity was observed in the cortex and hippocampus 48 h after TBI. In the cortex CCL20 was expressed in the ipsilateral as well as contralateral sides. The immunoreactivity was observed in the CA1 and CA3 hippocampal pyramidal cell layers and was restricted to ipsilateral side of the brain. CCL20 immunoreactivity was absent in the 24 h group. Additionally, CCL20-positive neuronal cell bodies displayed pyknotic morphology and were surrounded by areas devoid of tissue (Figure 6A; Figure 7A). The immunohistochemical observation was further supported by the quantitation of the CCL20-positive cell bodies which showed a significant increase in CCL20-positive neurons in the cortex and hippocampus of rats euthanized 48 h post-TBI compared to 24 h or sham control rats (Figure 6B). It is noteworthy that although CCL20 immunoreactivity was not seen in the damaged neurons at 24 h, it was expressed by the neurons of cortex and hippocampus (Figure 7A), including the degenerating ones in these regions at 48 h after impact as evident by the co-localization of FJ and CCL20 stainings (Figure 7B). Importantly, CCL20 expressing cells in the cortex (Figure 8) and hippocampus (data not shown) were mostly neurons as they were also NeuN positive. Taken together, these observations demonstrate that CCL20 expression is increased in the brain due to TBI-induced neuronal injury at a later time point than the systemic increase of the same chemokine in response to mild TBI and may play a role in the neural injury and inflammatory reaction in the brain.

To evaluate the significance of the spleen in LFPI-induced neurodegeneration, splenectomy was performed immediately after the induction of TBI. FJ histochemistry and CCL20 immunostaining were performed to evaluate the extent of damage in splenectomised animals. It was observed that in splenectomised rats the number of FJ-positive cells was significantly reduced compared to non-splenectomised animals at the same time points, while within the splenectomy group the number of FJ-positive cells was significantly increased after TBI compared to splenectomised shams (Figure 9A). Splenectomy also reduced CCL20 expression in the cortex 48 h after TBI. In splenectomised rats, CCL20 expression increased significantly when compared to splenectomised sham animals; but the CCL20 expression was reduced significantly when the spenectomised TBI rats were compared to the non-splenectomised TBI group. These observations indicate that the spleen plays a role in TBI induced neurodegeneration and CCL20 expression in the rat brain after mild TBI.