Forty-eight male Sprague-Dawley rats (Charles River Laboratory, Wilmington, MA) were used in this study. Rats were approximately 120 days (400–500 g) of age at the commencement of experimentation. The animals were housed in a room with controlled temperature and a 12 h light/dark cycle (lights on at 0600) with ad-libitum access to food and water. All experimental protocols were approved by the Wright Patterson Air Force Base Institutional Animal Care and Use Committee, and conformed to the guidelines published in the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals.

Two housing conditions were used: environmentally enriched condition (EC) and standard control (ST), both for 15 days before injury (see Figure 1 for experimental timeline). The ST group was housed in standard-sized polycarbonate rat cages, two per cage. Animals in both housing conditions were subjected to a medial prefrontal TBI via CCI. An additional group of animals received sham injury and were maintained under standard housing conditions. Sixteen animals were placed in each cohort: EC-TBI, ST-TBI, and Sham.

The EC group was housed eight animals per cage in large wire mesh cages (1 m³) with two levels and access to running wheels and toys (e.g., ropes, plastic cars and trucks, plastic tunnels, ladders and running wheels) (Briones et al., 2004; Lippert-Gruener et al., 2007) (Figure 2). The objects inside the EC cages were replaced by new toys twice weekly (Briones et al., 2004) in order to provide the animals with opportunities for learning and sensory stimulation. In addition, twice a day each EC animal underwent individual motor skills training (MST), an enriching activity that involves learning a complex motor task and combines coordination, balance, concentration, and learning components. MST consisted of an elevated obstacle course of parallel bars, balance beams, ropes, ladders, and chains (Jones et al., 1999). Animals were trained on the MST course between 1300 and 1500. EC animals were also exposed to daily olfactory stimulation consisting of variously perfumed paper strips presented in the cage for 10 min (Maegele et al., 2005). The scents used each day were chosen randomly and consisted of peppermint, lavender, or cinnamon. ST-TBI and Sham animals were handled daily, yet otherwise left undisturbed in their group cages without any additional stimulating enrichment.

After 15 days of housing in either EC or ST conditions, animals were subjected to TBI in the prefrontal cortex by means of CCI. This region in the rat is homologous to the anterior cingulate cortex in humans, implicated in frontal lobe syndrome (Leonard, 1969). The CCI model of brain injury in rats results in pathophysiology similar to human TBI including contusion and axonal injury (Dixon et al., 1996). Animals were anesthetized with isoflurane (5% induction, 2% maintenance) and mounted in a stereotaxic apparatus with heads fixed in a horizontal position. Blood oxygen saturation (SpO₂) was monitored and maintained at levels >90% and body temperature was maintained at 37°C with a homeothermic heating blanket system. A midline incision was made and the scalp and fascia were retracted to expose the cranium. Using a 6 mm diameter trephan, a craniectomy was made over the medial frontal cortex, 3 mm anterior to bregma. Once the dura was exposed, a bilateral contusion was induced using a 5 mm diameter anvil with an electric contusion impactor (Custom Design and Fabrication, Virginia Commonwealth University Medical Center, Richmond, Virginia). The injury was made with a velocity of 2.25 m/s, depth of 3 mm, and 50 ms of brain contact. Bleeding was stopped with gauze and cool saline, and the scalp was subsequently sutured with 7 mm surgical staples. Sham animals received craniectomy but no cortical impact. Following recovery from anesthesia, animals were returned to their original cages with cage mates. The EC animals were returned to their original cages, but no longer received additional enrichment, and the levels and toys were removed from the cages. The animals recovered for 1 week before the commencement of behavioral tests.

The sensory neglect test is used to observe responsivity to focal somatosensory stimuli (Hoffman et al., 2003). All animals were tested for sensory neglect at baseline and post-injury. Baseline measurements of sensory neglect were taken one day prior to CCI or sham surgery. Further testing was conducted at 8 and 25 days post-surgery. The test consisted of a 2 cm diameter sticker being placed on the distal-radial area of both forelimbs. The animal was then placed in a clear empty shoebox cage and the latency to remove both stickers was recorded. Each animal received 3 trials with approximately 5 min intervals. The maximum length of a trial was 2 min, at which time any remaining stickers were removed.

The elevated plus maze (EPM) is a standard test of anxiety-like behaviors in rodents (Benaroya-Milshtein et al., 2004). All animals were tested at day 7 post-injury. Animals were placed in the center of the maze, consisting of two open (10.2 × 102 cm) and two enclosed arms (10.2 × 102 × 30.5 cm), elevated 53.3 cm above the floor. At the start of the test, animals were placed facing one of the enclosed arms. Each animal was tested for one 5 min trial and the maze was cleaned with 70% ethanol after each trial. Ethovision software was used to record the behavior of the animal and time spent in each area of the maze. Behaviors which were recorded included rearing, grooming, and head dipping.

The Morris water maze (MWM) is an established test of spatial learning and memory (Buccafusco, 2009). All animals were tested on the MWM beginning 11 days post-surgery. Animals were placed in a 178 cm diameter dark circular tank filled with water (approximately 37 cm deep). The tank was divided evenly into four quadrants (A, B, C, D) with appropriate marking of the quadrants on the top edge of the tank. A 10.2 cm diameter clear plexiglass platform was submerged to a depth of 2 cm below waterline and placed approximately 28 cm from the wall of the pool in quadrant C. Latency to find the hidden platform was recorded using Ethovision software that tracked and timed the animals in the tank until they reached the platform. The position of the platform remained the same throughout the experiment.

The MWM consisted of 10 tests performed over 12 days (two 5-day blocks with a 2-day break between the blocks to test memory retention). Days 1–11 consisted of acquisition testing and were comprised of 2 trials per day. For each trial, the animal was placed in the pool at random quadrants, facing the wall. If the animal did not find the platform in 90 s, they were physically guided to it. Upon reaching the platform the animal remained there for 10 s and then was removed for a 30 s interval before the start of the second trial. MWM day 12 consisted of only one probe trial with the platform removed and time spent in quadrant C was recorded.

Following the behavioral tests and stress challenge, animals were briefly anesthetized with 5% isofluorane and decapitated via guillotine. The brains from half of each group were removed and split in half sagittally (eight brains in each cohort: EC-TBI, ST-TBI, and Sham). The remaining brains were taken for biochemical determinations, however, freezer failure resulting in the loss of approximately half of the samples compromised our ability to perform biochemical and more detailed histolgical analyses that had been planned. Coronal sections of the intact brain hemispheres (14 μm) through the prefrontal cortex and hippocampus were cut on a cryostat and stored at −20°C. Every fourth section was used for lesion analysis using Nissl staining to identify cell bodies and grossly visualize damage from the CCI. Slides were fixed in 4% paraformaldehyde for 20 min. After being rinsed in 1× phosphate buffered saline (PBS) and dried overnight, slides were placed in cresyl violet for 4–5 min. Racks were rinsed 5 times with H₂O and dehydrated using a variety of alcohol concentrations and xylenes. The slides were coverslipped and imaged using an Axio Imager.Z1 (Carl Zeiss Microimaging, Thornwood, NY). The Nissl-stained coronal sections of the prefrontal cortex were visually analyzed for gross differences in lesion size. In both ST-TBI and EC-TBI animals, there was substantial loss of Nissl neuronal profiles in the dorsal prefrontal cortex, typically compromising the anterior cingulate cortex, with occasional damage seen in the prelimbic cortex. Qualitatively, the extent of cortical damage was similar between the two groups. Due to use of unfixed tissue, we were unable to perform detailed quantitative mapping of the extent of prefrontal cortex damage. Note that prefrontal cortical damage produced by this CCI approach has been previously documented (Hoffman et al., 2003).

All results are expressed as mean ± standard error (SE), unless depicted differently in the figures. The data were tested for normality and equal variance using the Kolmogorow-Smirnov test and F-test, respectively. Outliers for each data set were identified using Grubb's test with α = 0.05 (Barnett and Lewis, 1994). The significance level for all tests was 0.05. A One-Way analysis of variance was used to compare groups ST-TBI, EC-TBI, and Sham. If day was included as a factor, a mixed-design ANOVA was used. Since the specific hypotheses tests were identified a priori, planned comparisons were performed regardless of the outcome of the ANOVA. Post-hoc paired comparisons of groups used Tukey simultaneous confidence intervals with a 0.05 experimentwise error level.

One rat (Sham) was removed from all analyses for chewing the sticker instead of trying to remove it. Prior to TBI, there was not a significant effect of group on latency to sticker removal, indicating that housing conditions did not affect sensory neglect. Following TBI, there were significant effects of group on sticker removal latency on day 8 [F_{(2, 40)} = 6.33, p < 0.01] and day 25 [F_{(2, 40)} = 6.13, p < 0.01]. Post-hoc analysis indicated that both Sham and EC-TBI removed the stickers significantly faster than the ST-TBI group at both post-injury time points (Figure 3). A mixed design ANOVA using group and day as factors did not have a significant group/day interaction. A scatter graph of individual rats indicated an elevated mean for the ST-TBI group at baseline was due to the contribution of only two rats. Since mean differences during baseline were not considered meaningful, it was concluded that housing conditions did not affect sensory neglect and TBI-induced sensory neglect deficits were attenuated when enrichment was applied prior to injury, particularly on day 8.

Percent time in the open arm, head dipping, rearing, and grooming were tested as indicators of anxiety-related behaviors in the EPM [One-Way analysis of variance (ANOVA)]. There were significant effects of groups on open arm time [F_{(2, 40)} = 8.08, p < 0.01] and head dipping [F_{(2, 40)} = 7.23, p < 0.01], but no significant group effects on rearing or grooming (Figure 4). Post-hoc paired comparisons indicated that both ST-TBI and EC-TBI groups spent more time in the open arm and more time head dipping than the Sham group. Time spent in the open arm and head dipping can be associated with reduced anxiety-like behavior. However, it is important to note that prefrontal damage can increase risk-taking behavior (Bechara et al., 1994; Floden et al., 2008) which may result in increased open arm time via a non-anxiety related mechanism (Pandey et al., 2009).

Five separate trials throughout the MWM testing were considered outliers and removed from analysis (1 Sham, 1 ST-TBI, 3 EC-TBI). A mixed-design ANOVA on latency to reach the submerged platform was performed for each trial separately using days 8–11, following a two-day break when animals were not tested. For trial 1 there was a significant difference among the groups [F_{(2, 41)} = 4.56, p = 0.01]. Post-hoc analysis indicated that in trial 1, following the 2-day break, both Sham and EC-TBI located the platform significantly faster than the ST-TBI group (Figure 5). For trial 2 there was not a significant difference among the groups. The EC-TBI animals performed similarly to the Sham controls, demonstrating no injury-induced deficits in spatial memory. ST-TBI animals, however, demonstrated impaired MWM performance, especially during trial 1 after the two-day break, indicative of memory recall deficits.

No significant differences in swim speed (range = 25.3 ± 1.1 to 28.0 ± 1.1 cm/s) were observed among the groups on the first or final day, confirming that the differences observed in escape latencies could not be attributed to motor deficits affecting swimming speed. On the last day of MWM testing, the platform (originally in quadrant C) was removed and time spent in quadrant C was analyzed for memory retention. A One-Way ANOVA revealed no significant difference among the groups for time in quadrant or distance swam.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.