Neuroprotective effects of FK866 against traumatic brain injury: Involvement of p38/ERK pathway.

Abstract Objective FK866 is an inhibitor of nicotinamide phosphoribosyltransferase (NAMPT), which exhibits neuroprotective effects in ischemic brain injury. However, in traumatic brain injury (TBI), the role and mechanism of FK866 remain unclear. The present research was aimed to investigate whether FK866 could attenuate TBI and clarified the underlying mechanisms. Methods A controlled cortical impact model was established, and FK866 at a dose of 5 mg/kg was administered intraperitoneally at 1 h and 6 h, then twice per day post‐TBI until sacrifice. Brain water content, Evans blue dye extravasation, modified neurological severity scores (mNSS), Morris water maze test, enzyme‐linked immunosorbent assay (ELISA), immunofluorescence staining, and western blot were performed. Results The results demonstrated that FK866 significantly mitigated the brain edema, blood–brain barrier (BBB) disruption, and ameliorated the neurological function post‐TBI. Moreover, FK866 decreased the number of Iba‐1‐positive cells, GFAP‐positive astrocytes, and AQP4‐positive cells. FK866 reduced the protein levels of proinflammatory cytokines and inhibited NF‐κB from translocation to the nucleus. FK866 upregulated the expression of Bcl‐2, diminished the expression of Bax and caspase 3, and the number of apoptotic cells. Moreover, p38 MAPK and ERK activation were significantly inhibited by FK866. Interpretation FK866 attenuated TBI‐induced neuroinflammation and apoptosis, at least in part, through p38/ERK MAPKs signaling pathway.


Introduction
Traumatic brain injury (TBI) is defined as damage to the brain induced by external mechanical impact. The severity of TBI is predicated on consciousness alteration, organized as mild, moderate, or severe. 1 It has been estimated that the annual number of patients suffering from TBI around the world has exceeded 50 million, which costs about 400 billion dollars each year. 2 TBI is grouped into two stages: the primary injury which is caused by external impact, and secondary injury which is caused by a series of changes in neurochemistry, metabolism, cells, and molecules. 3 The mechanisms involved in secondary injury include excitotoxicity, oxidative stress, neuroinflammation, mitochondrial dysfunction, axon degeneration, brain edema, blood-brain barrier damage, and cell death. 4 Exploring novel agents that target secondary injury is still urgent for TBI clinical treatment.
Nicotinamide phosphoribosyltransferase (NAMPT) is a crucial enzyme in the cyclic biosynthetic pathway of nicotinamide adenine dinucleotide (NAD), serving as the catalyzer of the generation from nicotinamide (NM) to nicotinamide mononucleotide (NMN). 5 Indeed, it has been postulated that NAMPT may participate in the regulation of inflammation and is responsible for many inflammatory diseases. 6,7 NAMPT was also involved in the regulation of apoptosis. [8][9][10] FK866 is an inhibitor of NAMPT. [11][12][13] Recently, it has been demonstrated that FK866 protects ischemic neuronal injury and spinal cord injury, which were achieved by inhibition of neuroinflammation and apoptosis. 14,15 However, whether NAMPT inhibitor FK866 attenuates neuroinflammation and apoptosis after TBI remains to be clarified.
This study aims to illuminate the role and possible underlying mechanisms of NAMPT inhibitor FK866 in neuroinflammation and apoptosis post-TBI.

Animals
Adult male and female Sprague-Dawley rats weighing 300-330 g from SLAC Laboratory Animal Co., Ltd. (Shanghai, China) were adopted as experimental animals. Housing conditions of the rats were unified by temperature and humidity control and a 12 h light/dark shift. Consent to all animal experiments was given by the Institutional Animal Care and Use Committee of Zhejiang University and all the procedures complied with the Guide for the Care and Use of Laboratory Animals by National Institutes of Health.

TBI model
The TBI model was established by controlled cortical impact (CCI) as described previously. 16 The CCI model was shown in Figure S1. The rats were fixed on a stereotaxic apparatus. A midline incision was made to create a 5-mmdiameter craniotomy. The marked skull was cut and removed by a drill without damaging the dura mater. The CCI was accomplished vertically to the brain surface by a PinPoint TM Precision Cortical Impactor (Cary, NC, USA) according to the following parameters: 4-mm-diameter impact tip, impact velocity of 3 m/sec, an impact duration time of 120 ms, and 2 mm displacement of the brain. The rat was placed on a warming pad with a rectal thermometer to maintain its body temperature at 36.0-36.5°C. After the injury, the bone flap was immediately replaced, sealed, and closured the skin. Sham-operated animals underwent the same procedures without performing CCI. The rats were put back into cages until complete recovery from anesthesia in a heated chamber. The rats in our study were carefully taken care, especially after surgery.

Animals grouping and drug administration
One hundred and Fifty-nine male rats and twenty-four female rats were involved and assigned into three groups in random: sham group, TBI + vehicle group, and TBI+ FK866 group. The experimental outline was shown in Figure S2. The rats sacrificed in each group were shown (Fig. S3). FK866 (Sigma-Aldrich, F8557) was dissolved in dimethyl sulfoxide (DMSO) and diluted in saline before use. FK866 (5 mg/kg) was administrated intraperitoneally at 1 h and 6 h after TBI. And the same dose of FK866 was given twice a day until sacrificed. The dose and the administration time of FK866 were selected based upon a previous publication. 14

Brain water content
The extent of brain edema was quantified by the wet weight/dry weight method. Twenty-four hours post-TBI, rats were euthanatized, and the brains were weighed for wet weight at once after removal. The dry weight of the brains was obtained after 24 hours of drying at 105°C. The brain water content was calculated as: [(wet weight À dry weight)/wet weight] 9 100%.

Evaluation of neurological deficits
The neurological deficits after TBI were assessed by the modified neurological severity scores (mNSS) including motor, sensory, reflex, and balance tests. 17 The score was conducted and analyzed by an observer who was blinded to the experimental cohorts.

Morris water maze test
Morris water maze test was conducted as previously described. 18 Briefly, the rat was trained to find the platform before TBI or sham operation. For each trial, the rats were randomly placed into a quadrant start point facing the wall of the pool and were allowed a maximum of 60 sec to escape to the platform. Rats that failed to escape within 90 sec were placed on the platform for a maximum of 20 sec and returned to the cage for a new trial. Maze performance was recorded using a video camera suspended above the maze and interfaced with a videotracking system. The average escape latency of a total of five trials was calculated. The test was conducted at 3 and 7 days after TBI or sham operation. The test was conducted and analyzed by an observer who was blinded to the experimental cohorts.

Evans blue dye extravasation
The Evans blue dye extravasation was conducted as previously described. 19 Evans blue dye (2%, 5 ml/kg) was administrated via the left femoral vein and circulated for 1 hour. Under deep anesthesia, rats were sacrificed by cardiac perfusion. Then, removed and separated the brain   . The effect of FK866 on brain edema, and BBB disruption at 1 day after TBI or sham operation. (A) Quantification of brain water content. The bar represents the mean AE SD. n = 6. The bar represents mean AE SD, **P < 0.01 versus sham, ## P < 0.01 versus vehicle control. (B) Quantification of Evans blue dye extravasation. The bar represents the mean AE SD. n = 6. **P < 0.01 versus sham. ## P < 0.01 versus vehicle control. (C) Representative western blots and Densitometric quantification of albumin. The bar represents the mean AE SD. N = 6. **P < 0.01 versus sham group; ## P < 0.01 versus TBI + vehicle group. (D) Representative western blots and densitometric quantification of occluding. The bar represents the mean AE SD. n = 6. **P < 0.01 versus sham group; ## P < 0.01 versus TBI + vehicle group.
to get the ipsilateral pericore parietal cortex immediately. Subsequently, weighted the brain samples and homogenized with 3 ml of 50% trichloroacetic acid, then centrifuged at 15,000g for 30 min. The supernatant was mixed with an equal volume of trichloroacetic acid with ethanol. After overnight incubation (4°C), the samples were centrifuged again (15,000 g, 30 min) and measured by spectrofluorophotometer (excitation wavelength 620 nm and emission wavelength 680 nm).

Enzyme-linked immunosorbent Assay
The levels of inflammatory cytokines were quantified using enzyme-linked immunosorbent assay (ELISA) kits specifically for rats, following the instruction from manufacturers (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). Briefly, the brain samples were homogenized in 1 ml of buffer containing 1mmol/L of phenylmethylsulfonyl fluoride, 1 mg/L of pepstatin A, 1 mg/L of aprotinin, and 1 mg/L of leupeptin in PBS solution with a glass homogenizer and then centrifuged at 12,000g for 20 min at 4°C. The supernatant was then collected and total protein was determined. The cytokine contents in the brain samples were expressed as a pictogram of antigen per milligram protein.
We selected at least three sections per rats with the ipsilateral pericore parietal cortex and three fields with a magnification of 200 9 per section. For quantification of Iba-1 and GFAP-positive cells, the number from fields was averaged and expressed as positive cells per millimeter. For quantification of apoptotic neurons and AQP4positive cells, the percentage of TUNEL-positive neurons or AQP4-positive cells was calculated as follows: (the number of TUNEL-positive neurons or AQP4-positive cells/total number of neurons or cells) 9100%. Tissue sections were analyzed by an observer who was blinded to the experimental cohorts.

Statistical analysis
Band density values of the target proteins were normalized to the control group to enable comparisons between the different groups. Data were presented as mean AE SD or median (interquartile range) based on whether satisfying normal distribution or not. For data satisfying normal distribution, a significant difference between groups was analyzed with one-way ANOVA followed by Tukey's multiple comparison tests. For data satisfying non-normal distribution, comparisons between groups were performed by Mann-Whitney test. P < 0.05 was regarded as statistical significance.

FK866 administration improved neurological function following TBI
The mNSS scores in TBI + vehicle group were significantly increased at day 1, day 3, and day 7 compared with that in sham operation (P < 0.01, Fig. 1A), FK866 administration treatment significantly decreased mNSS scores at day 1, day 3, and day 7 compared with the vehicle administration (P < 0.05, Fig. 1A). On day 3 and day 7, there was a significant increase in escape time among the TBI + vehicle group compared with sham controls (P < 0.01, Fig. 1B). However, the TBI-induced deficits could be significantly attenuated by FK866 administration (P < 0.05, Fig. 1B). In addition, similar neuroprotective effects were also observed in the female rats after TBI (Fig. S4).
FK866 alleviated brain edema and bloodbrain barrier (BBB) following TBI Brain water content in the TBI + vehicle group was significantly increased compared with the sham control (P < 0.01, Fig. 2A), which was significantly reduced by FK866 administration (P < 0.01, Fig. 2A). Consistently, administration of FK866 led to a significant amelioration of Evans blue dye extravasation at 24 h post-TBI (P < 0.01, Fig. 2B). Besides, the expression of albumin in the pericore cortex was significantly increased compared with the sham controls (P < 0.01, Fig. 2C), which was opposed to occluding protein expression (P < 0.01, Fig. 2D). FK866 administration significantly downregulated the albumin expression (P < 0.01, Fig. 2C) and upregulated the occluding expression compared with vehicle administration (P < 0.01, Fig. 2D).  The percentage of AQP4-positive cells was significantly higher in the TBI + vehicle group than the control group (P < 0.01, Fig. 3A and B), whereas FK866 administration significantly decreased the percentage of AQP4-positive cells (P < 0.01, Fig. 3A and B).

FK866 attenuated the nuclear translocation of NF-jB p65
Compared with the sham group, the cytoplasmic expression and the nuclear expression of NF-jB p65 display opposite alteration: the former decreased significantly (P < 0.01, Fig. 6A and B) while the latter increased significantly (P < 0.01, Fig 6A and C) at 24 h post-TBI. However, the administration of FK866 significantly Figure 7. The effect of FK866 on neuronal neurons in the peri-injured cerebral cortex at 24 h after TBI. In the rats subjected to TBI, the percentage of TUNEL-positive neurons was higher when compared with the sham group, while FK866 decreased neuronal apoptosis. Scar bar = 50 lm. n = 5/group. The bar represents mean AE SD, **P < 0.01 versus sham group; ## P < 0.01 versus TBI + vehicle group. upregulated its cytoplasmic expression (P < 0.05, Fig 6A  and B) and decreased its nuclear expression (P < 0.01, Fig 6A and C).

FK866 decreased neuronal apoptosis
The percentage of TUNEL-positive neurons was significantly higher in the TBI + vehicle group than the control group (P < 0.01, Fig. 7A and B), whereas FK866 administration significantly decreased the percentage of TUNEL-positive neurons (P < 0.01, Fig. 7A and B). The content of Bcl-2, Bax, and caspase 3 was examined to further confirm the antiapoptotic effects of FK866. By contrast with the sham group, the expression of Bcl-2 was significantly reduced at 24h post-TBI (P < 0.01, Fig. 8A and B), while FK866 administration upregulated its expression (P < 0.01, Fig. 8A and B). On the other hand, the expression of Bax and caspase 3 exhibited a significant increment in comparison with the sham group (P < 0.01, Fig. 8A, C and D), while the administration of FK866 significantly decreased their expressions (P < 0.01, Fig. 8A, C and D).

FK866 inhibited p38 MAPK and ERK activation
To explore the underlying mechanism, western blot was utilized to examine phosphorylated p38 MAPK and ERK at the protein level. Compared with the sham group, the expression of phosphorylated p38 MAPK and ERK increased significantly at 24 h post-TBI (P < 0.01, Fig. 9A and B), while administration of FK866 significantly decreased their expressions (P < 0.01, Fig. 9A and B).

Discussion
In this study, several novel founding presented as follow: (a) FK866 prevented brain edema, BBB disruption, and improved neurological function after TBI. (b) FK866 reduced the number of microglia/macrophages and astrocyte, downregulated the level of proinflammatory cytokines and proapoptotic proteins, and upregulated the levels of antiapoptotic protein. (c) FK866 attenuated TBI-induced neuroinflammation and apoptosis, and the potential mechanisms, at least in part, involved p38/ERK MAPKs signaling pathway (Fig. 10).
TBI refers to the impairment of brain function resulting from an external force such as impact or penetration. 20 Brain edema following TBI is an important factor which contributes to the evolution of brain injury and is associated with significant morbidity and mortality. 21,22 It is classified as vasogenic or cytotoxic edema according to BBB disruption or dysfunction of cellular ionic pumps. Both vasogenic edema and cytotoxic edema were found in TBI patients by MRI. 23 AQP4 plays an important role in maintaining brain water homeostasis, mediating the flux of water at the interface between the cerebral vasculature. 24 AQP4 is known to predominantly contribute to cytotoxic edema after TBI. 25 In our study, we found that FK866 not only prevented the BBB disruption but also downregulated the expression of AQP4. In addition, neuroinflammation is closely connected with BBB disruption: BBB disruption fuels neuroinflammation, and on the other hand, neuroinflammation aggravate BBB disruption. Astrocytes and microglia/macrophages are often the primary cell types to initiate an inflammatory cascade upon sensing damage. 26 In this study, we found that astrocytes and microglia/macrophages increased significantly after TBI, which were reduced by FK866 treatment. In addition, proinflammatory cytokines were also decreased by FK866 administration. NF-jB is a classic transcription factor that exerts regulatory effects on proinflammatory cytokines such as TNF-a, IL-1b, and IL-6 at the transcriptional and translational levels. 27 Previous studies indicated that NF-jB activation participated in the secondary brain injury post-TBI. 28,29 In our study, we found NF-jB activation after TBI, which was inhibited by FK866 administration. In addition, neuronal apoptosis increased, which was also decreased by FK866 treatment.
The mitogen-activated protein kinases (MAPKs) pathway is an important factor regulating the release of proinflammatory cytokines and apoptosis. 30 MAPK is a family of serine/threonine protein kinases, including ERK, JNK, and p38. Previous studies indicated that activation of NF-jB was closely associated with MAPK signaling pathway. 31,32 The activation of MAPK led to the upregulation of caspase 3 and the decline in Bcl-2, thus engendering apoptosis. 33 It was observed that phosphorylated ERK and p38, rather than JNK, augmented rapidly after TBI in a cortical impact model of mice, 34 while the findings of Chen et al. indicated that the activation of JNK and p38, instead of ERK, was enhanced in the cortical neurons after traumatic injury. 35 Nevertheless, in a lateral fluid percussion model, phosphorylated ERK and JNK, other than p38, demonstrated an increase in the cortex and hippocampus of the TBI rats. 36 Our results showed that the phosphorylation of ERK and p38 was intensified at 24 h after TBI. And, we also found that FK866 reduced phosphorylated levels of p38 and ERK. Taken together, these observations suggest that FK866 attenuated TBI-induced neuroinflammation and apoptosis, at least in part, through p38/ERK MAPKs signaling pathway.
There did exist some limitations in this study. First, Fk866 exhibited neuroprotective effects in female TBI rats, but the potential mechanisms are still needed. Second, we examined the effects and possible mechanisms of FK866 in a focal TBI model, studies in global events such as those mimicked by fluid percussion brain injury are still required.
This study demonstrated that NAMPT inhibitor FK866 attenuated neuroinflammation, apoptosis, BBB disruption, and brain edema, and improved neurological function following TBI. The p38/ERK MAPK signaling pathway may be involved in the anti-inflammatory and antiapoptotic effects of FK866 following TBI.

Supporting Information
Additional supporting information may be found online in the Supporting Information section at the end of the article. Figure S1. The typical image of traumatic brain injury model. Figure S2. The experimental outline in this study. Figure S3. The number of rats sacrificed in this study. Figure S4. The effect of FK866 on neurological function after TBI in female rats.