The Novel Monoacylglycerol Lipase Inhibitor MJN110 Suppresses Neuroinflammation, Normalizes Synaptic Composition and Improves Behavioral Performance in the Repetitive Traumatic Brain Injury Mouse Model

Modulation of the endocannabinoid system has emerged as an effective approach for the treatment of many neurodegenerative and neuropsychological diseases. However, the underlying mechanisms are still uncertain. Using a repetitive mild traumatic brain injury (mTBI) mouse model, we found that there was an impairment in locomotor function and working memory within two weeks post-injury, and that treatment with MJN110, a novel inhibitor of the principal 2-arachidononyl glycerol (2-AG) hydrolytic enzyme monoacylglycerol lipase dose-dependently ameliorated those behavioral changes. Spatial learning and memory deficits examined by Morris water maze between three and four weeks post-TBI were also reversed in the drug treated animals. Administration of MJN110 selectively elevated the levels of 2-AG and reduced the production of arachidonic acid (AA) and prostaglandin E2 (PGE2) in the TBI mouse brain. The increased production of proinflammatory cytokines, accumulation of astrocytes and microglia in the TBI mouse ipsilateral cerebral cortex and hippocampus were significantly reduced by MJN110 treatment. Neuronal cell death was also attenuated in the drug treated animals. MJN110 treatment normalized the expression of the NMDA receptor subunits NR2A and NR2B, the AMPA receptor subunits GluR1 and GluR2, and the GABAA receptor subunits α1, β2,3 and γ2, which were all reduced at 1, 2 and 4 weeks post-injury. The reduced inflammatory response and restored glutamate and GABA receptor expression likely contribute to the improved motor function, learning and memory in the MJN110 treated animals. The therapeutic effects of MJN110 were partially mediated by activation of CB1 and CB2 cannabinoid receptors and were eliminated when it was co-administered with DO34, a novel inhibitor of the 2-AG biosynthetic enzymes. Our results suggest that augmentation of the endogenous levels of 2-AG can be therapeutically useful in the treatment of TBI by suppressing neuroinflammation and maintaining the balance between excitatory and inhibitory neurotransmission.


Introduction
Traumatic brain injury (TBI) is one of the major causes of mortality and disability in modern industrialized societies [1], and has a significant impact on behavioral impairment, emotional disturbance and long term medical complications among the victims. It is estimated that there are nearly 1.7 million TBI cases each year in the United States, among which around 50,000 people died and 275,000 people were hospitalized [2]. Recently, attention has been focused on the repetitive mild traumatic brain injury (mTBI), due to its high incidence in military personnel deployed in the combat fields, and professional Committee. Repetitive closed head injury was performed in mice using an electromagnetic controlled stereotaxic impact device as we described previously [18]. Briefly mice were anesthetized with 3% isoflurane prior to stabilizing the head using ear bars in a mouse stereotaxic frame (Stoelting Co., Wood Dale, IL, USA). A midline sagittal incision was made to expose the skull. The stereotaxic electromagnetic impactor with a 3.0 mm steel tip impounder was used to deliver a single controlled cortical impact, delivered at coordinates of 1.8 mm caudal to bregma and 2.0 mm left to midline with a controlled velocity at 3.0 m/s, an impact depth of 2.0 mm and a dwell time at 100 milliseconds. Mice with depressed skull fracture or visible hemorrhage were excluded from the study. Sham mice underwent identical surgical procedures as the injured group, but no impact was delivered. After impact, the skin was sutured and the animals were allowed to recover from anesthesia and then returned to their home cages. A second and third identical closed head injury procedures were performed on day 2 and day 3 after the initial injury. The skull fracture rate was less than 2.0% during the surgical procedures.
For drug treatment, mice were given an intraperitoneal injection of various doses of MJN110 (0.5, 1 and 2.5 mg/kg), a selective inhibitor of MAGL, at 30 min after each impact and then once a day for 5 additional days (eight injections in total). At the selected concentrations, MJN110 has been shown to dose-dependently increase the brain levels of 2-AG and alleviate neuropathic pain in mice [24,25]. To determine the cannabinoid receptor dependency, MJN110 (2.5 mg/kg) was co-administered with the CB1R antagonist AM281 (3 mg/kg) and the CB2R antagonist AM630 (3 mg/kg), respectively. To examine the importance of 2-AG molecule in this TBI mouse model, MJN110 was co-administered with DO34 (30 mg/kg), a selective inhibitor of the 2-AG synthetic enzymes diacylglycerol lipase alpha and beta (DAGLα and β) to block the 2-AG synthesis [29]. The individual or combined pharmacological inhibitors were dissolved in 1:1:18 ratio of DMSO-cremophorsaline that served as a vehicle control.

Behavioral Assay
The beam-walk, Y-maze and Morris water maze were performed at various time points post-injury to assess fine motor movement [13,18,30], working memory [13,18] and spatial learning and memory [18,31]. The timeline for all the behavioral tests, and molecular, biochemical and immunohistochemistry assessments is illustrated in Figure 1.

Animal Model and Drug Treatment
Male, 8-12 weeks old C57BL/6J mice were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). All animal procedures were performed in accordance with the guidelines established by the National institutes of Health (NIH, Bethesda, MD, USA) and approved by the Uniformed Services University Institutional Animal Care and Use Committee. Repetitive closed head injury was performed in mice using an electromagnetic controlled stereotaxic impact device as we described previously [18]. Briefly mice were anesthetized with 3% isoflurane prior to stabilizing the head using ear bars in a mouse stereotaxic frame (Stoelting Co., Wood Dale, IL, USA). A midline sagittal incision was made to expose the skull. The stereotaxic electromagnetic impactor with a 3.0 mm steel tip impounder was used to deliver a single controlled cortical impact, delivered at coordinates of 1.8 mm caudal to bregma and 2.0 mm left to midline with a controlled velocity at 3.0 m/s, an impact depth of 2.0 mm and a dwell time at 100 milliseconds. Mice with depressed skull fracture or visible hemorrhage were excluded from the study. Sham mice underwent identical surgical procedures as the injured group, but no impact was delivered. After impact, the skin was sutured and the animals were allowed to recover from anesthesia and then returned to their home cages. A second and third identical closed head injury procedures were performed on day 2 and day 3 after the initial injury. The skull fracture rate was less than 2.0% during the surgical procedures.
For drug treatment, mice were given an intraperitoneal injection of various doses of MJN110 (0.5, 1 and 2.5 mg/kg), a selective inhibitor of MAGL, at 30 min after each impact and then once a day for 5 additional days (eight injections in total). At the selected concentrations, MJN110 has been shown to dose-dependently increase the brain levels of 2-AG and alleviate neuropathic pain in mice [24,25]. To determine the cannabinoid receptor dependency, MJN110 (2.5 mg/kg) was co-administered with the CB1R antagonist AM281 (3 mg/kg) and the CB2R antagonist AM630 (3 mg/kg), respectively. To examine the importance of 2-AG molecule in this TBI mouse model, MJN110 was co-administered with DO34 (30 mg/kg), a selective inhibitor of the 2-AG synthetic enzymes diacylglycerol lipase alpha and beta (DAGLα and β) to block the 2-AG synthesis [29]. The individual or combined pharmacological inhibitors were dissolved in 1:1:18 ratio of DMSO-cremophor-saline that served as a vehicle control.

Behavioral Assay
The beam-walk, Y-maze and Morris water maze were performed at various time points post-injury to assess fine motor movement [13,18,30], working memory [13,18] and spatial learning and memory [18,31]. The timeline for all the behavioral tests, and molecular, biochemical and immunohistochemistry assessments is illustrated in Figure 1. The red arrows represent the days when the first, second, and third TBI was delivered. The repetitive closed head injury was established using an electromagnetic controlled stereotaxic impact device. BW, beam walk; qPCR, quantitative polymerase chain reaction; LC-MS/MS, liquid chromatography with tandem mass spectrometry; WB, western blot; IHC, immunohistochemistry; ELISA, enzyme-linked immunoassay; MWM, Morris water maze. The red arrows represent the days when the first, second, and third TBI was delivered. The repetitive closed head injury was established using an electromagnetic controlled stereotaxic impact device. BW, beam walk; qPCR, quantitative polymerase chain reaction; LC-MS/MS, liquid chromatography with tandem mass spectrometry; WB, western blot; IHC, immunohistochemistry; ELISA, enzyme-linked immunoassay; MWM, Morris water maze.

Beam Walk Test
The beam-walk test was performed to assess fine motor movement as we previously described [13]. Briefly, the beam-walk apparatus consists of a wooden beam measuring 6 mm in width, 120 cm in length, and suspended 30 cm above a table. Mice were trained to walk on the beam for 3 days before surgery and at the end of the third day the baseline values (i.e., the number of foot faults per 50 steps) were collected. The beam-walk balance test was performed at days 4, 6 and 10 after the initial impact and the number of foot faults per 50 steps was recorded.

Spontaneous Alternation Y-Maze Test
At 13 days after repetitive closed head injury, mice were tested in a Y-maze apparatus. The Y-maze device (Stoelting Co., Wood Dale, IL, USA) has a symmetrical plastic Y shape with arms measuring 25 cm long, 8 cm wide and 15 cm high. Mice were placed at the end of a randomly chosen arm as a starting point, and allowed to explore the maze for 5 min. Each arm visit was counted as a mouse moving all four paws into the arm, and each alternation was defined as a consecutive entry into three different arms. The sequence and the number of entries to each arm were video recorded. The percentage alternation was calculated based on the formula: total number of alternations/(total number of arm entries-2) × 100 as previously described [13,14]. The spontaneous alternation Y-maze test measures hippocampus-based working memory [32], and was used to evaluate the effect of drug treatment on TBI-induced working memory deficits.

Morris Water Maze Test
The Morris water maze test was performed at 24-29 days post-TBI to examine spatial learning and memory as described previously with some modifications [31]. Briefly, a circular water tank with a diameter of 1.5 m was used to perform the Morris water maze. For extra-maze visual cues, the four walls around the tank were permanently hanged with different colors and shapes for orientation. A round transparent platform (with a diameter of 11.5 cm) was hidden 1 cm beneath the surface of the water at the center of a given quadrant of the water tank. Mouse received training in the Morris water maze for 5 days and each session was composed of four trials, plus a probe trial 24 h after the last hidden platform test. For each trial, the mouse was released facing the wall of the tank and allowed to search, find, and stand on the platform for 15 s within the 60-s trial period. The time it took to find the platform was calculated as Latency. In each training session, the starting quadrant point and sequence of the four quadrants from where the mouse was released into the tank were randomly chosen so that it was varied among the separate sessions for each animal and was different for individual animals. In the latter case, the mouse who couldn't find the platform was guided to the platform and remained there for 15 s before being returned to the home cage. The probe trial was conducted by removing the platform and releasing the mouse facing the wall exactly from the opposite or parallel side where the platform was hidden previously. The task performance was recorded for 60 s. Tracking of animal movement was achieved with the ANY-maze video-tracking system.

Quantitative Real-Time PCR
Mice were sacrificed on day 4 after three impacts to extract total RNA from the ipsilateral cortex using TRIzol (Sigma, St Louis, MO, USA) following the manufacturer's instructions. RNA yield and purity were evaluated using a NanoDrop spectrophotometer. RNA (1 µg) was reverse transcribed to cDNA. First-strand cDNA was synthesized with Maxima First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, USA). Amplified cDNAs were diluted at 1:10 in ultrapure water and subjected to SYBER green master mix (Applied Biosystems, Grand Island, NY, USA) based real-time PCR on a light cycler 480 II Roche System with primers IL-1β, IL-6, TNF-α, iNOS, CB1R, CB2R, COX1, COX2, and GAPDH. PCR reactions were conducted as follows: 95 • C for 15 s, 75 • C for 1 min, and 60 • C for 30 s for a total of 40 cycles, followed by a melting point determination or dissociation curves. All samples were run in triplicate. Relative levels of gene expression were determined by the 2 −∆∆Ct method normalized to the expression of GAPDH.

PGE 2 Assay
For PGE 2 assay, the ipsilateral cortex was isolated from day 4 post-TBI and subjected to homogenization with 40 µL of 0.02% trifluoroacetic acid (TFA) and 100 µL of acetonitrile on ice. To extract maximal lipid, 140 µL of tissue homogenate was dispersed in 1 mL acetonitrile by vortex and left at 4 • C overnight. On the next day the homogenate-acetonitrile mixture was centrifuged at 2000× g for 5 min to remove the debris and the supernatant was transferred to a silanized glass tube. The supernatant was evaporated under the nitrogen gas streaming in a water bath (approx. 35 • C) and then reconstituted with acetonitrile. The levels of PGE 2 in the lipid extract were measured using PGE 2 enzyme-linked immunoassay (ELISA) kit following the manufacturer's protocol (Cayman Chemical, Ann Arbor, MI, USA).

Western Blot
To determine the levels of phosphorylated ERK, phosphorylated AKT, NMDA and AMPA receptor subunits, GABA A subunits, GAD67, VGLUT2, COX1 and COX2, mice were sacrificed at 8, 14 and 30 days post-TBI. The ipsilateral cortical tissues adjacent to the site of injury were removed and homogenized in RIPA buffer supplemented with 1x protease & phosphatase inhibitor cocktail (Cat# 1861280, Thermo Fisher Scientific, Rockford, IL, USA). Protein concentration was determined for each sample, and equal amount of proteins were run on 4-15% SDS-PAGE (Bio-Rad). Thereafter, proteins were transferred onto nitrocellulose membranes, blocked for 1 h with 5% BSA at room temperature, and then incubated overnight at 4 • C with the primary antibodies. These antibodies included antirabbit phosphorylated Membranes were subsequently probed for the expression of β-actin as a loading control. Images were acquired with a Chemidoc Touch image system (BioRad) and analyzed using ImageJ software.

LC-MS/MS Analysis
The cortical tissues isolated from day 4 TBI mice were subjected to liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) for the measurement of 2-AG, AEA and AA. The tissue was homogenized with 40 µL of 0.02% TFA, 250 µL of acetonitrile, and 250 picomoles of 2-AG-d5 and AA-d8 and 0.5 picomoles of AEA-d4 (Cayman Chemical) using a Potter homogenizer at 4 • C. The homogenate was dissolved completely in 2.5 mL acetonitrile by vortex and kept at 4 • C overnight. The homogenate was subjected to centrifugation at 2000× g for 5 min to remove the debris, and then the supernatant was evaporated under nitrogen gas streaming in a water bath (approx. 35 • C). The lipid was re-suspended with 100 µL of acetonitrile and stored at −80 • C until use.
An HPLC system (1200 Series, Agilent Technologies, Santa Clara, CA, USA) was used with a reverse phase guard column (Wide Pore C18 (ODS), 4 × 2 mm ID; Phenomenex, Torrance, CA, USA), and the column (Sephasil Peptide C18, 5 µ, ST, 100 × 4.6 mm ID; Pharmacia Biotech, Piscataway, NJ, USA) was maintained at 40 • C. The mobile phase was composed of solvent A (0.2% formic acid in water) and solvent B (0.2% formic acid in methanol), and the following gradient was used: 62% A/38% B isocratic for 30 s, ramp to 90% B in 60 s, isocratic at 90% B for 18.5 min, ramp back to 62% A/38% B in 60 s, and re-equilibrate at 62% A/38% B for 8 min. The flow rate was 0.4 mL/min. The HPLC output was directed into the TurboV electrospray ionization (ESI) source of a Q-Trap 4000 mass spectrometer (AB Sciex, Framingham, MA, USA). The injection volume was 20 µL. LC-MS/MS analysis was performed in a positive mode with the ion source temperature of 600 • C, a spray voltage of 5.5 kV, and a declustering potential of 45 V. Multiple reactions monitoring (MRM) was performed on the transitions m/z 379→287 for 2-AG, 348→62 for AEA, and 304→121 for AA. The concentrations of 2-AG, AEA and AA were determined by calculating the corresponding peak area ratio to the internal standard (IS) using a linear fit weighting to the calibration curve.

Immunohistochemistry
On day 8 post-TBI, mice were anesthetized using a combination of ketamine and xylazine (90 mg/kg ketamine/10 mg/kg xylazine in a volume of 10 µL/g body weight, i.p.) and then intracardially perfused with ice-cold PBS followed by 4% paraformaldehyde. The brains were quickly removed from the skulls and post-fixed in 4% paraformaldehyde at 4 • C overnight. On the following day, brains were cryoprotected in 30% sucrose in PBS at 4 • C until sinking, embedded in Tissue Tek OCT (Sakura Finetek, Torrance, CA, USA) and stored at −80 • C until use. Coronal sections of the brain were cut at 30 µm using SM2000R microtome (Leica, Buffalo Grove, IL, USA) and the series sections (10 to 12 slices) were collected in a 24-well plate and stored in cryoprotectant solution at −20 • C until the assay for immunostaining. Primary antibodies used for immunostaining included the anti-goat Iba1 (1:300; Cat# ab48004, Abcam), and anti-mouse GFAP (1:500; Cat# 3670, Cell Signaling) antibodies that were used to determine whether there was an altered expression of microglia and astrocytes, respectively. Free floating sections were immunostained using the specific antibodies listed above, followed by incubation with the corresponding fluorescent-labeled secondary antibodies. The sections were washed with 1× PBS twice, mounted on slides and then covered with fluoroshield mounting medium with DAPI. Immunofluorescence images were obtained with a fluorescence microscope (Eclipse TE-2000U, Nikon, Melville, NY, USA). The cells with both DAPI and expected fluorescence were defined as positively stained cells. Negative controls were routinely performed in which the primary antibodies were omitted. All immunofluorescence data were obtained in a minimum of 5-7 serial sections from the brain tissues of each animal. GFAP and Iba1 immunostaining images were taken, analyzed and quantified using the NIH ImageJ software.

Fluoro-Jade C Staining
Fluoro-Jade C (FJC) staining was applied to discern the degenerated neurons. Free floating sections mounted on slides were incubated in the solution with FJC (0.0004% solution, EMD Millipore, Temecula, CA, USA) and 4-6-diamidino-2-phenylindole (0.5 µg/mL) for 20 min, followed by 3 × 1 min wash with distilled water. The slices were dried naturally at room temperature without light. The images were taken using a fluorescence microscope (Nikon Eclipse TE-2000U) and the FJC positive cells were counted and expressed as mean cell numbers per mm 2 .

Statistical Analysis
Statistical analysis was performed using Prism 7 software (GraphPad Software, Inc., San Diego, CA, USA). For multiple comparison, one-way or two-way analysis of variance (ANOVA) followed by Tukey-Kramer's post-hoc tests were used. Differences in significance were defined at p < 0.05 and data were expressed as means ± SEM.

Locomotor and Working Memory Deficits Associated with Repetitive mTBI Were Mitigated by the MAGL Inhibition
To determine the therapeutic effect of augmentation of 2-AG through pharmacological inhibition of its hydrolysis, different doses of MJN110 (0.5, 1.0 and 2.5 mg/kg) were given intraperitoneally at 30 min after each impact and then once daily for 5 additional days. Treatment with MJN110 at 1.0 and 2.5 mg/kg, but not at 0.5 mg/kg significantly reduced the number of foot faults in beam-walk test at 4 and 6 days post-TBI (Figure 2A). The number of spontaneous alternations in the Y-maze was also significantly reduced on day 13 post-TBI and increased by MJN110 treatment ( Figure 2B). Administration of MJN110 to the sham animals did not impact the performance in both beam-walk and Y-maze tests (data not shown). Since MJN110 at 2.5 mg/kg had greater therapeutic effects to mitigate TBI-induced locomotor and working memory deficits, this dose was mostly used for the subsequent studies.

Statistical Analysis
Statistical analysis was performed using Prism 7 software (GraphPad Software, Inc., San Diego, CA, USA). For multiple comparison, one-way or two-way analysis of variance (ANOVA) followed by Tukey-Kramer's post-hoc tests were used. Differences in significance were defined at p < 0.05 and data were expressed as means ± SEM.

Locomotor and Working Memory Deficits Associated with Repetitive mTBI Were Mitigated by the MAGL Inhibition
To determine the therapeutic effect of augmentation of 2-AG through pharmacological inhibition of its hydrolysis, different doses of MJN110 (0.5, 1.0 and 2.5 mg/kg) were given intraperitoneally at 30 min after each impact and then once daily for 5 additional days. Treatment with MJN110 at 1.0 and 2.5 mg/kg, but not at 0.5 mg/kg significantly reduced the number of foot faults in beam-walk test at 4 and 6 days post-TBI (Figure 2A). The number of spontaneous alternations in the Y-maze was also significantly reduced on day 13 post-TBI and increased by MJN110 treatment ( Figure 2B). Administration of MJN110 to the sham animals did not impact the performance in both beam-walk and Ymaze tests (data not shown). Since MJN110 at 2.5 mg/kg had greater therapeutic effects to mitigate TBI-induced locomotor and working memory deficits, this dose was mostly used for the subsequent studies.

MAGL Inhibition Selectively Elevated the Endogenous 2-AG Levels in the TBI Mouse Brain
To determine the selectivity of MJN110 on 2-AG metabolism in the TBI mouse brain, the levels of 2-AG, AEA and AA in the ipsilateral cortex at 4 days post-TBI were examined using LC-MS/MS. Although there were no differences in the brain levels of 2-AG, AEA and AA between the TBI and sham control animals ( Figure 3A-C), administration of MJN110 (2.5 mg/kg) selectively elevated the brain levels of 2-AG ( Figure 3A), but not AEA ( Figure 3B). The levels of AA were significantly reduced in the MJN110 treatment group ( Figure 3C). These results further support the notion that MAGL is a principal enzyme to regulate AA synthesis in the brain [29].

MAGL Inhibition Selectively Elevated the Endogenous 2-AG Levels in the TBI Mouse Brain
To determine the selectivity of MJN110 on 2-AG metabolism in the TBI mouse brain, the levels of 2-AG, AEA and AA in the ipsilateral cortex at 4 days post-TBI were examined using LC-MS/MS. Although there were no differences in the brain levels of 2-AG, AEA and AA between the TBI and sham control animals ( Figure 3A-C), administration of MJN110 (2.5 mg/kg) selectively elevated the brain levels of 2-AG ( Figure 3A), but not AEA ( Figure 3B). The levels of AA were significantly reduced in the MJN110 treatment group ( Figure 3C). These results further support the notion that MAGL is a principal enzyme to regulate AA synthesis in the brain [29].

Inhibition of MAGL Attenuated TBI-Induced Neuroinflammation and Neuronal Cell Death
TBI is commonly associated with enhanced inflammatory response which is characterized by activation of glial cells, infiltration of peripheral leukocytes and release of a variety of inflammatory mediators including cytokines and chemokines [33]. It is possible

Inhibition of MAGL Attenuated TBI-Induced Neuroinflammation and Neuronal Cell Death
TBI is commonly associated with enhanced inflammatory response which is characterized by activation of glial cells, infiltration of peripheral leukocytes and release of a variety of inflammatory mediators including cytokines and chemokines [33]. It is possible that such changes in cellular and molecular events can initiate neuropathological process in the animal model of TBI. We measured the mRNA levels of several pro-inflammatory mediators and the expression of CB1 and CB2 cannabinoid receptors. Quantitative RT-PCR analysis showed that the mRNA expression of IL-6, IL-1β, TNFα, iNOS and COX-1 was significantly increased in the ipsilateral cortices of TBI mice and all the changes were reversed by MJN110 (2.5 mg/kg) treatment ( Figure 4A-E). However, the expression of COX-2 was not altered between the sham and TBI groups, though a significant reduction was found in the MJN110 treated group ( Figure 4F). There was no difference in the levels of CB1 receptor among the sham, TBI/vehicle and treatment groups ( Figure 4G), whereas the expression of CB2 receptor was significantly elevated in the TBI/vehicle and reduced by MJN110 treatment ( Figure 4H). In harmony with these findings, immunostaining using GFAP and Iba1 revealed that the accumulation of astrocytes and microglia in the ipsilateral cortex and dentate gyrus of the TBI mouse brain was significantly reduced by MJN110 treatment (Figure 5A,B). Neuronal cell death in the TBI mouse ipsilateral cortex and hippocampal dentate gyrus examined by Fluoro-Jade C staining was also reduced in the MJN110 treated animals ( Figure 5C).

Alteration of the Ionotropic Glutamate and GABA Receptor Components and the Survival Signaling Molecules in the TBI Mouse Hippocampus Was Restored by MAGL Inhibition
TBI-induced cognitive deficits can be caused by the loss of hippocampal neurons and the alteration of receptor molecules involved in the process of learning and memory such as GABA and glutamate receptors [34]. Using hippocampal tissue lysates, we examined the expression of glutamate and GABA receptor subunits at 8 days post-TBI. The expression of GABAA receptor subunits α1, β2, 3 and γ2 ( Figure 6A,B) and glutamate receptor subunits GluR1, GluR2, NR2A and NR2B ( Figure 6A,D) were significantly reduced in the TBI vehicle group, and these changes were restored in the MJN110 (2.5 mg/kg) treated animals. Conversely, NMDAR1 ( Figure 6D) and the GABA synthetic enzyme glutamate decarboxylase (GAD) 67 ( Figure 6C) were increased in the TBI/vehicle and reduced by MJN110 treatment. Notably, there was no difference in the expression of the vesicular glutamate transporter 2 (VGLUT2) in all the experimental groups ( Figure 6E). The phosphorylation of ERK and AKT ( Figure 6F) were significantly down regulated in the TBI mouse brain and reversed by MJN110 treatment. Similar to the changes in the mRNA expression, Western blot also revealed the increased expression of COX-1, but not COX-2 Figure 5. MJN110 treatment reduced glial cell accumulation and neuronal death in the ipsilateral cortex and dentate gyrus of TBI mouse brain. At 8 days post-injury, the (A) GFAP-, (B) Iba1-and (C) FJC-positive immunostainings were significantly increased in the ipsilateral cortex and dentate gyrus (DG) of the TBI mouse brain, and these changes were attenuated by MJN110 (2.5 mg/kg) treatment. (*** p < 0.001 and **** p < 0.0001; n = 3 animals in each group. Bars represent means ± SEM, and the scale bar is equal to 50 µm). DAPI, 4 ,6-diamidino-2-phenylindole; GFAP, glial fibrillary acidic protein; FJC, Fluoro-Jade C.

Alteration of the Ionotropic Glutamate and GABA Receptor Components and the Survival Signaling Molecules in the TBI Mouse Hippocampus Was Restored by MAGL Inhibition
TBI-induced cognitive deficits can be caused by the loss of hippocampal neurons and the alteration of receptor molecules involved in the process of learning and memory such as GABA and glutamate receptors [34]. Using hippocampal tissue lysates, we examined the expression of glutamate and GABA receptor subunits at 8 days post-TBI. The expression of GABA A receptor subunits α1, β2, 3 and γ2 ( Figure 6A,B) and glutamate receptor subunits GluR1, GluR2, NR2A and NR2B ( Figure 6A,D) were significantly reduced in the TBI vehicle group, and these changes were restored in the MJN110 (2.5 mg/kg) treated animals. Conversely, NMDAR1 ( Figure 6D) and the GABA synthetic enzyme glutamate decarboxylase (GAD) 67 ( Figure 6C) were increased in the TBI/vehicle and reduced by MJN110 treatment.
Notably, there was no difference in the expression of the vesicular glutamate transporter 2 (VGLUT2) in all the experimental groups ( Figure 6E). The phosphorylation of ERK and AKT ( Figure 6F) were significantly down regulated in the TBI mouse brain and reversed by MJN110 treatment. Similar to the changes in the mRNA expression, Western blot also revealed the increased expression of COX-1, but not COX-2 in the TBI mouse hippocampal tissues, and treatment with MJN110 attenuated the expression of both enzymes, although only the expression of COX-2 was significantly reduced when compared to the TBI/vehicle group ( Figure 6G).

Reversal of Locomotor and Short-Term Working Memory, but Not Long-Term Cognitive Function by MJN110 Was Mediated by the CB1 Receptor Activation
To determine whether the action of MJN110 is mediated through cannabinoid receptors, we administered MJN110 together with the CB1 receptor antagonist AM281 or the CB2 receptor antagonist AM630 in mice undergoing repetitive mTBI. On day 4 post-injury, the locomotor impairment was significantly reduced in the MJN110 (2.5 mg/kg) treated animals, and the therapeutic effect of MJN110 was greatly diminished in the AM630 (3 mg/kg) co-treatment group. Despite that treatment with AM281 (3 mg/kg) also led to increased numbers of foot faults, there was no significant difference when it was compared with the MJN110 treatment alone ( Figure 7A). Interestingly at 6 and 10 days post-TBI, the locomotor deficits were completely reversed by co-treatment with AM281 ( Figure 7A). Similarly, the improvement of MJN110 on spontaneous alternation in Y-maze at 13 days post-TBI was also significantly reversed by AM281 (Figure 7B), suggesting the major contribution of CB1 receptor in improving working memory. Next, we performed Morris water maze to examine spatial learning and memory at 24-29 days post-TBI (Figure 7C-E). TBI animals travelled longer distance and spent extended latency to find the The expression of COX-1, but not COX-2 was increased in the TBI/vehicle group (G). (** p < 0.01, *** p < 0.001, and **** p < 0.0001 comparison between sham and TBI/vehicle; # p < 0.05, ## p < 0.01 and ### p < 0.001 comparing TBI/vehicle vs. TBI + MJN110 groups. n = 5 animals in each group). Bars represent means ± SEM).

Reversal of Locomotor and Short-Term Working Memory, but Not Long-Term Cognitive Function by MJN110 Was Mediated by the CB1 Receptor Activation
To determine whether the action of MJN110 is mediated through cannabinoid receptors, we administered MJN110 together with the CB1 receptor antagonist AM281 or the CB2 receptor antagonist AM630 in mice undergoing repetitive mTBI. On day 4 post-injury, the locomotor impairment was significantly reduced in the MJN110 (2.5 mg/kg) treated animals, and the therapeutic effect of MJN110 was greatly diminished in the AM630 (3 mg/kg) co-treatment group. Despite that treatment with AM281 (3 mg/kg) also led to increased numbers of foot faults, there was no significant difference when it was compared with the MJN110 treatment alone ( Figure 7A). Interestingly at 6 and 10 days post-TBI, the locomotor deficits were completely reversed by co-treatment with AM281 ( Figure 7A). Similarly, the improvement of MJN110 on spontaneous alternation in Y-maze at 13 days post-TBI was also significantly reversed by AM281 ( Figure 7B), suggesting the major contribution of CB1 receptor in improving working memory. Next, we performed Morris water maze to examine spatial learning and memory at 24-29 days post-TBI ( Figure 7C-E). TBI animals travelled longer distance and spent extended latency to find the submerged platform at training days 25-28 compared to the sham animals. MJN110 treatment significantly reduced the travel distance and latency compared to the TBI/vehicle group (Figure 7C,D).
Similarly, in the probe trial on day 29 post-TBI, TBI animals had fewer entries to the hidden platform than the sham and MJN110 treated animals ( Figure 7E). The effect of MJN110 was not significantly altered by co-administration of either AM281 or AM630, despite a tendency of improvement in the AM281 treated animals ( Figure 7C-E).  Figures 7C,D). Similarly, in the probe trial on day 29 post-TBI, TBI animals had fewer entries to the hidden platform than the sham and MJN110 treated animals ( Figure  7E). The effect of MJN110 was not significantly altered by co-administration of either AM281 or AM630, despite a tendency of improvement in the AM281 treated animals (Figure 7C-E).

Inhibition of MAGL Normalized the Expression of Ionotropic Glutamate and GABA Receptors at 30 Days Post-TBI
To determine whether alteration of the glutamate and GABA receptor components is persistent and thus contributes to the long-term memory and cognitive deficits [35,36], the expression of the NMDA, AMPA and GABA receptors was also examined at 30 days post-TBI. Similar to the findings at 8 days post-TBI, the hippocampal expression of GABA A -α1, -β2,3 and -γ2, GluR1, GluR2, NR2A, NR2B at 30 days post-TBI was also significantly reduced compared to the sham control animals, and all the changes were restored in the MJN110 (2.5 mg/kg) treatment group ( Figure 8A,B,D). Conversely, the expression of the NMDAR1 subunit ( Figure 8D) and GAD67 ( Figure 8C) was dramatically increased in the TBI group and reduced by MJN110 treatment. There were no changes in the expression of VGLUT2, COX1 and COX2 in all the experimental groups ( Figure 8E,F). Co-treatment with AM281 or AM630 did not alter the effect of MJN110 on the expression of these receptors.

Inhibition of MAGL Normalized the Expression of Ionotropic Glutamate and GABA Receptors at 30 Days Post-TBI
To determine whether alteration of the glutamate and GABA receptor components is persistent and thus contributes to the long-term memory and cognitive deficits [35,36], the expression of the NMDA, AMPA and GABA receptors was also examined at 30 days post-TBI. Similar to the findings at 8 days post-TBI, the hippocampal expression of GABAA -α1, -β2,3 and -γ2, GluR1, GluR2, NR2A, NR2B at 30 days post-TBI was also significantly reduced compared to the sham control animals, and all the changes were restored in the MJN110 (2.5 mg/kg) treatment group ( Figure 8A,B,D). Conversely, the expression of the NMDAR1 subunit ( Figure 8D) and GAD67 ( Figure 8C) was dramatically increased in the TBI group and reduced by MJN110 treatment. There were no changes in the expression of VGLUT2, COX1 and COX2 in all the experimental groups ( Figure 8E,F). Co-treatment with AM281 or AM630 did not alter the effect of MJN110 on the expression of these receptors.

The Therapeutic Effects of MJN110 Were Eliminated by Simultaneous Blockade of 2-AG Synthesis in the TBI Mouse Brain
To further validate the therapeutic effect of MJN110 is attributable to the elevated brain levels of 2-AG, we utilized a recently developed DAGL inhibitor, DO34, to block the 2-AG synthesis [29]. TBI animals were treated with DO34 (30 mg/kg), MJN110 (1 mg/kg) and their combination once a day for 7 days. Inhibition of DAGLα and DAGLβ by DO34 reversed the therapeutic effect of MJN110 on locomotor function and working memory in both beam walk and Y-maze tests, whereas treatment with DO34 alone did not have any effect (Figure 9A,B). To assess the selectivity of systematically administered DO34 in the mouse brain, 2-AG, AEA and AA levels were measured in the ipsilateral cortex at 4 days post-TBI using LC-MS/MS. We found that DO34 treatment almost completely blocked the synthesis of 2-AG and the elevated levels of 2-AG by MJN110 were significantly reduced ( Figure 9C). Notably, there was no change in the levels of AEA among the groups ( Figure 9D). TBI per se did not affect the levels of AA compared to the sham group, but treatment with DO34 or its combination with MJN110 significantly reduced AA levels ( Figure 9E). The levels of PGE 2 in the TBI mouse brain was also significantly reduced by DO34 and its combined treatment with MJN110 ( Figure 9F). Treatment with MJN110 also normalized the expression of GABA A -β2, 3, GABA A -γ2, GluR1, GluR2, NR2A, NR2B and NMDAR1 receptor subunits in hippocampal tissue lysates at 14 days post-TBI and these changes were mostly reversed by co-treatment with DO34. Treatment with DO34 alone did not have any effect on the altered expression of these inotropic receptor subunits in the TBI mouse hippocampus ( Figure 9G). Cells 2021, 10, x 15 of 22

Discussion
Accumulating evidence supports that augmentation of endocannabinoids appears to be an effective strategy for the treatment of many neurological and neuropsychiatric diseases [6,7]. However, the underlying therapeutic mechanisms remain unclear. In this study, we found that selective inhibition of MAGL with MJN110 elevated the brain levels of 2-AG and reduced the production of AA and PGE 2 . Treatment with MJN110 suppressed neuroinflammation, normalized the expression of glutamate and GABA receptor subunits and improved motor and cognitive function. The improvement in locomotor function and working memory in the TBI animals seems to be mediated by activation of cannabinoid receptors. Furthermore, we found that the therapeutic effect of MJN110 was diminished when the 2-AG synthetic pathway was blocked suggesting that the production of 2-AG is critically involved in neuroprotection.
Neuroinflammation immediately following TBI is a major stepping-stone to initiate the pathological processes and may cause long-term neurological deficits and behavioral abnormalities [33,37]. Numerous studies have shown that TBI is accompanied by infiltration of peripheral immune cells and activation of resident glial cells in the brain through cytokine and chemokine signaling, and the release of many inflammatory mediators [38]. Consistently, we observed the accumulation of astrocytes and microglia and the increased production of pro-inflammatory mediators IL-6, iNOS, IL-1β, TNF-α and COX-1 in the ipsilateral cerebral cortex in our repetitive mTBI mouse model. Neuronal cell death was also observed in the cortex and hippocampal dentate gyrus of the TBI mouse brain. We and others have previously shown that enhancement of ERK and AKT signaling may improve motor and cognitive function after TBI [13,18,39,40]. Here we found that inhibition of MAGL increased the brain levels of 2-AG and reduced the production of pro-inflammatory mediators and neuronal cell death, possibly through activation of ERK and AKT, two widely studied pathways known for their roles promoting survival and improving neurological outcomes after injury to the brain and spinal cord [13,14,30,41]. Despite that 2-AG can exert its beneficial effects via cannabinoid receptors [38], the involvement of CB1 and CB2 receptors in the action of the inhibitors of 2-AG hydrolytic enzymes remains uncertain. Using the CB1 and CB2 receptor antagonists, we found that the locomotor function recovery is mediated by both CB1 and CB2 receptors in the first several days following brain injury, and that activation of CB1 receptor contributes to working memory recovery at the relatively late time point. However, the improvement of MJN110 treatment on TBI-induced spatial learning and memory deficits is independent on cannabinoid receptors. Although co-administration of the CB1 receptor antagonist tends to reverse the improved performance of MJN110 in both the Morris water maze acquisition phase and the subsequent probe trial, there were no statistically significant differences in the MJN110 alone and MJN110 plus AM281 treatment groups. These findings are in agreement with the study using the experimental ischemic stroke model treated with the MAGL inhibitor JZL184 [27]. These results indicate that the behavioral improvements of MAGL inhibition in this TBI mouse model are attributable to both cannabinoid receptor dependent and independent mechanisms.
Neuronal cell death following brain injury could directly or indirectly influence synaptic plasticity, neuronal circuitry and signaling transmission [42][43][44]. Our recent study revealed that the reduced expression of several SNARE proteins including synaptosomeassociated protein of 25 kDa (SNAP25), and cysteine string protein alpha (α-CSP) at one month post-injury is associated with TBI induced cognitive deficits [18]. Maintenance of proper synaptic integrity is crucial for brain development, synaptic function as well as mitigation of neuropathology [45][46][47]. Exaggerated activation of NMDA receptors causes deregulated Ca ++ influx and neuronal cell death in stroke, ischemia, TBI and other neurodegenerative diseases such as Alzheimer's, Parkinson's and Huntington's diseases [35,48]. NMDA receptor activation in hippocampal neurons controls the induction of long-term potentiation (LTP) which is associated with memory formation [49,50]. Consistent with previous reports in the mild repetitive closed head injury mouse model [15,51], our cur-rent study using hippocampal tissue lysates from 8, 14 and 30 days post-TBI revealed a reduced expression of NMDA receptor subunits (NR2A, NR2B), and the AMPA receptor subunits (GluR1, GluR2), and these changes might contribute to the reduction of LTP and the impaired cognitive function.
Examination of the postmortem brain tissues from Alzheimer's patients showed a decrease in the expression of NR2A and NR2B mRNA and protein levels in the hippocampus and entorhinal cortex [52,53]. Deletion of the NR2A subunit of the NMDA receptors resulted in significant reduction of hippocampal LTP and spatial learning deficits [54], whereas overexpression of the NR2B receptor subunits enhanced synaptic potentiation as well as learning and memory [55]. The reduced expression of NR2A and NR2B in our current TBI mouse model might be due to the cleavage of their intracellular c-termini by the increased activity of calpain, and thus resulting in the uncoupling of the NMDA receptors from downstream survival pathways as reported in the animal model of focal cerebral ischemia [56]. Despite the expression of both NR2A and NR2B was reduced, it is likely that these NMDA receptor subunits could be differentially regulated. The increased ratio of the NR2A and NR2B subunits was demonstrated to cause long-term memory deficits [57], suggesting that the proper balance between these two subunits is also critical for the maintenance of normal synaptic neurotransmission. Given that NR2A and NR2B are the dominant NMDA receptor components expressed at the respective synaptic and extrasynaptic sites, the composition and function of both synaptic and extrasynaptic NMDA receptors are likely to be impaired following repetitive TBI, and thus are attributable to the reduced cognitive function. It has been reported that AMPA receptor subunits can control the trafficking and delivery of NMDA receptors to the plasma membrane and synaptic compartments through their intercellular domains [58]. Reduced GluR2 receptor expression in the glutamate tetrameric receptor complex was linked to excitotoxicity [59] and found in several neurological diseases like ischemia [60,61], epilepsy [62,63] and brain trauma [64]. Studies in schizophrenia patients and Parkinson's disease animal models also revealed that alteration of NMDA receptor trafficking, and its decreased expression might contribute to the cognition and behavioral deficits [15,65,66]. All these studies suggest that LTP dependent memory formation and recognition is dependent on synaptic strength and plasticity which is functionally maintained by the NMDA and AMPA receptors at the synaptic terminals. The improvement of MJN110 on neurological and behavioral deficits is likely attributable to the restoration of TBI-induced loss of hippocampal NMDA and AMPA receptor subunits.
Another important synaptic components are GABA A receptors, that are essential for GABAergic stimulation and the synchronization of oscillatory activities required for the normal cognition [67]. Several studies in schizophrenia patients found an altered expression of GABA A receptor subunits and GABA synthesis in the cerebral cortex, and electrophysiological dysfunction in parvalbumin and somatostatin GABAergic interneurons in the cortex and hippocampus [68][69][70]. Blockade of the fast inhibitory GABA A receptors by bicuculline, pentylenetetrazol or picrotoxin causes severe motor seizures in experimental animals [71]. A rat controlled cortical impact-induced TBI model study showed a longlasting working memory impairment associated with increased levels of GAD67 one month after injury [72]. The loss of GABAergic interneurons and the reduced surface expression of GABA A receptors (GABA A -α1, -β2 and -γ2) in hippocampus and amygdala of TBI animals lead to epileptic seizure and long-lasting anxiety disorders [36,[73][74][75][76][77]. Our current study demonstrated that the expression of GABA A -α1, -β2, 3 and -γ2 receptor subunits in the ipsilateral hippocampus was reduced at 8, 14 and 30 days post-TBI, and treatment with MJN110 normalized the expression of GABA A receptor subunits in hippocampus. Among the GABA A β receptor subunits, β3 is shown to play a crucial role in GABAergic inhibitory transmission [78]. Moreover, a recent study found that 2-AG can interact with GABA A β2 receptor subunit in the presence of low concentration of GABA to modulate locomotion behavior [79]. Taken together, our results suggest that MAGL inhibition can benefit TBI treatment through elevation of the endogenous 2-AG levels which can restore the dys-regulated excitatory and inhibitory neurotransmission and maintain the homeostasis of neuronal signaling.
The endogenous levels of 2-AG are tightly regulated by its synthesis and degradation. 2-AG in brain is chiefly synthesized by DAGLα, and to a lesser extent, by DAGLβ. Pharmacological inhibition or targeted gene deletion of DAGLα lowers brain levels of 2-AG, impairs synaptic plasticity, hippocampal dependent learning and memory, fear extinction and triggers trauma-like behaviors [80][81][82][83]. However, recent studies demonstrated that DAGLβ is the major enzyme for 2-AG synthesis in microglia/macrophage, and genetic deletion of DAGLβ suppressed inflammatory and neuropathic pain [84][85][86]. Consistently, the dual DAGLα and DAGLβ inhibitor DO34 reversed LPS-induced allodynia in mice with DAGLα deletion but had no effect on the antinociceptive effect of DAGLβ knockout mice [87]. Interestingly, a recent study showed that DAGLβ knockout male mice are resistant to TBI-induced mortality, but had no effect on motor and cognitive function, suggesting that blockade of AA and the subsequent prostaglandin production in reactive microglia/macrophages by targeting DAGLβ is unable to exert neuroprotective effects [88]. In our current study, treatment with DO34 significantly reduced the brain levels of 2-AG, AA and PGE 2 , and reversed the MJN110 mediated beneficial effects in the TBI animals. These results suggest that the production of 2-AG rather than the reduced eicosanoid signaling is imperative for the therapeutic effects of the MAGL inhibitors.