TGFβ-activated Kinase 1 (TAK1) Inhibition by 5Z-7-Oxozeaenol Attenuates Early Brain Injury after Experimental Subarachnoid Hemorrhage*

Background: Role of TGFβ-activated kinase 1 (TAK1) in the pathogenesis of early brain injury after subarachnoid hemorrhage (SAH) has not been reported. Results: TAK1 inhibition attenuates early brain injury and improves neurological deficits after SAH. Conclusion: TAK1 inhibition exhibits neuro-protective effects possibly through anti-apoptotic function. Significance: These results provide a novel target for SAH treatment. Accumulating evidence suggests that activation of mitogen-activated protein kinases (MAPKs) and nuclear factor NF-κB exacerbates early brain injury (EBI) following subarachnoid hemorrhage (SAH) by provoking proapoptotic and proinflammatory cellular signaling. Here we evaluate the role of TGFβ-activated kinase 1 (TAK1), a critical regulator of the NF-κB and MAPK pathways, in early brain injury following SAH. Although the expression level of TAK1 did not present significant alternation in the basal temporal lobe after SAH, the expression of phosphorylated TAK1 (Thr-187, p-TAK1) showed a substantial increase 24 h post-SAH. Intracerebroventricular injection of a selective TAK1 inhibitor (10 min post-SAH), 5Z-7-oxozeaenol (OZ), significantly reduced the levels of TAK1 and p-TAK1 at 24 h post-SAH. Involvement of MAPKs and NF-κB signaling pathways was revealed that OZ inhibited SAH-induced phosphorylation of p38 and JNK, the nuclear translocation of NF-κB p65, and degradation of IκBα. Furthermore, OZ administration diminished the SAH-induced apoptosis and EBI. As a result, neurological deficits caused by SAH were reversed. Our findings suggest that TAK1 inhibition confers marked neuroprotection against EBI following SAH. Therefore, TAK1 might be a promising new molecular target for the treatment of SAH.

Subarachnoid hemorrhage (SAH) 3 is a devastating neurological injury associated with significant patient morbidity and mortality. Despite the severe impact of SAH on public health, no specific pharmacological therapy for SAH is available that would improve the outcome of these patients. A growing body of evidence suggests that early brain injury (EBI) may contribute to the poor outcome and early mortality following SAH (1-3); However, elucidation of the complex cellular mechanisms underlying EBI remains a major challenge. Recently, multiple signaling pathways that are activated within minutes after the initial bleeding, which leads to early brain injury, are identified (4). Among them, activation of mitogen-activated protein kinases (MAPKs), particularly c-Jun-N-terminal kinases (JNK) and p38, exacerbates SAH injury by provoking proapoptotic and proinflammatory cellular signaling (5). Additionally, NF-B signaling activation also raises SAH-induced inflammatory responses and leads to worse SAH outcomes (6). Because multiple mechanisms contribute to EBI pathophysiology, it is thought that a drug specific for a single target may not give rise to meaningful improvements in neurobehavioral outcome. Alternatively, inhibition of a target involved in multiple EBI pathways may be a successful strategy. For this reason, it is appreciated that those agents, which have the ability to intervene at more than one critical pathway in the early brain injury after SAH, will have greater advantage over other single-target agents. TGF␤-activated kinase 1 (TAK1) may be such a target.
TAK1 is a member of the mitogen-activated protein kinase (MAPK) kinase kinase family and was initially found to function in TGF-␤-mediated MAPK activation (7). Previous study (8,9) demonstrated that TGF-␤ is a fibrogenic factor involved in the etiology of post-SAH pathology and plays an important role in generating communicating hydrocephalus after SAH. Several other known stimuli of TAK1 signaling, including IL-1 ␤, TNF-␣, and TLR4 have also been shown to be activated and to play a detrimental role in SAH (10 -12). The activated TAK1 in turn activates the NF-B and MAPK pathways (13,14). Recent investigations also demonstrate that inhibition of TAK1 provides neuroprotection in ischemia and traumatic brain injury (15)(16)(17). Taking into account this background, the present study aimed to evaluate whether TAK1 is activated after SAH. Moreover, the possible role of TAK1 in the regulation of SAHinduced neuronal apoptosis was also analyzed by means of TAK1 pharmacological modulation with its specific inhibitor 5Z-7-oxozeaenol (OZ).

Experimental Procedures
Animals and Subarachnoid Hemorrhage Model-Male Sprague-Dawley rats weighing 300 -350 g were used in this study. Rats were housed in a reversed 12-h light/12-h dark cycle controlled environment with free access to food and water. All procedures were approved by Nanjing University Animal Care and Use Committee and in accordance with the Guide for the Care and Use of Laboratory Animals by the National Institutes of Health (NIH).
Experimental SAH was performed as described previously in our laboratory and other laboratories (18 -21). First, the rats were intraperitoneally anesthetized with 10% chloral hydrate (400 mg/kg body weight) and placed in a stereotaxic head frame. An insulin injection needle (BD Biosciences) was tilted 45°in the sagittal plane, and placed 8-mm anterior to bregma in the midline, with the hole facing the right side. It was lowered until the tip reached the base of the skull, 2-3 mm anterior to the chiasma (ϳ10 -12 mm from the brain surface), and retracted 0.5 mm. Loss of cerebrospinal fluid and bleeding from the midline vessels were prevented by plugging the burr hole with bone wax before inserting the needle. A total of 300 l of non-heparinized fresh autologous arterial blood was slowly injected into the prechiasmatic cistern for 3 min under aseptic techniques. The heart rate was monitored and the rectal temperature was kept at 37 Ϯ 0.5°C by using physical cooling (ice bag) when required throughout experiments. Arterial blood samples were analyzed intermittently to maintain pO 2 , pCO 2 , and pH, parameters within normal physiological ranges. To maintain fluid balance, all rats were supplemented with 2 ml of 0.9% NaCl administered subcutaneously. After recovering from anesthesia, rats were returned to their cages with free access to food and water provided ad libitum. In the present study we observed that the basal temporal lobe was always stained with blood as described before. Therefore, the brain tissue adjacent to the clotted blood was used for analysis in our study (Fig. 1D).
Experimental Design 1-A total of 64 rats were used in this experiment. Of them, six rats with SAH were excluded later from the study because of little blood in prechiasmatic cistern but lots of blood clot in the frontal lobe instead, and 10 SAH rats died before the intended sacrifice. The animals were randomly assigned to six groups post-SAH: animals surviving SAH for 2 (n ϭ 6), 6 (n ϭ 6), 12 (n ϭ 6), 24 (n ϭ 9), 48 (n ϭ 6), and 72 h (n ϭ 6). Control (n ϭ 9) animals underwent the exact same procedure as described above with the exception that no blood was injected intracisternally. Tissue was harvested at the designed time points for the evaluation of TAK1 and p-TAK1 expression (n ϭ 6 each). Cellular distribution of TAK1 was also determined by double-staining immune fluorescence at 24 h post-SAH and control groups (n ϭ 3 each).
Experimental Design 2-To determine the role of TAK1 in SAH, rats were randomly allocated into the following groups and survived 24 h. 1) Control group (n ϭ 18) animals were injected with 0.3 ml of saline; 2) SAH group (n ϭ 12); 3) SAH ϩ vehicle group (n ϭ 18), rats were subjected to SAH plus intracerbroventricular administration of 5 l of dimethyl sulfoxide; and 4) SAH ϩ OZ group (n ϭ 18), rats were subjected to SAH plus intracerbroventricular administration of the TAK1 inhibitor OZ (Tocris Bioscience; 25 g of 10 min post-SAH). This dose was based on our prior published work using OZ in the traumatic brain injury model (15). Coordinates for the injection placement were 1.0 mm posterior to bregma, 1.4 mm lateral to midline, and 4.4 mm below the skull surface, and the injection duration was 10 min. Following the behavioral test, brain tissue of half the number of rats was harvested for biochemical and histopathological analysis at 24 h post-SAH. A total of 85 rats were used in this experiment (seven rats with SAH were excluded later from the study for the same reason mentioned in experiment 1, 12 SAH rats died before the intended sacrifice).
The neurological scoring was performed with a separate cohort of rats at 24 and 72 h after surgery. Finally, the rats were sacrificed for histological analysis. A total of 56 rats were used in this experiment (four rats with SAH score of 0 were excluded from the study for low SAH grade, eight SAH rats died before the intended sacrificed).
Experimental Design 3-To determine whether administration of the TAK1 inhibitor OZ confers brain protection with a wide therapeutic window against SAH, OZ or vehicle was administered 4 h post-SAH. Neurological scores were evaluated at 24 and 72 h post-SAH, and histological analysis was performed at 72 h. A total of 30 rats were used in this experiment (three rats with SAH score of 0 were excluded from the study for low SAH grade at 24 h post-SAH, six SAH rats died before the intended sacrifice).
Histological Examination-Rats were anesthetized and transcardially perfused with saline and 4% paraformaldehyde. The brains were removed and postfixed in paraformaldehyde for 24 h. For immunofluorescence microscopy, the brains were frozen in O.C.T. media, sectioned (10 m), and mounted onto slides. Immunofluorescence staining was performed as described previously (15). The specificity of the immunofluorescence reaction was evaluated by replacement of the primary antibody with rabbit IgG.
Terminal dUTP nick-end labeling (TUNEL) assay was conducted by using a TUNEL detection kit according to the manufacturer's instructions (In situ Cell Death Detection Kit, Roche Applied Science). In brief, each section was incubated with 50 l of TUNEL reaction mixture (contains 5 l of enzyme solution and 45 l of Label solution) for 1 h at 37°C in the dark and then washed with PBS. Slides were then counterstained with 4,6-diamidino-2-phenylindole (DAPI), washed, cover slipped with a water-based mounting medium, and sealed with nail polish. For the negative control, the sections were only incubated with Label solution. The positive cells were identified, counted, and analyzed under the light microscope by an investigator blinded to the grouping. The extent of brain damage was evaluated by the apoptotic index, defined as the average percentage of TUNEL-positive cells in each section counted in 10 cortical microscopic fields (at ϫ400 magnification). A total of four sections from each animal were used for quantification. The final average percentage of TUNEL-positive cells of the four sections was regarded as the data for each sample.
Formalin-fixed brains (72 h post-SAH) were dehydrated, embedded in paraffin, and sliced into 4-m thick sections that were stained with cresyl violet. The sampled region for each subfield was demarcated in the inferior basal temporal lobe and cresyl violet neuronal cell bodies were counted. To quantify the amount of nissl staining, 10 random high-power fields (ϫ400) in each coronary section were chosen, and the mean number of intact neurons in the 10 views was regarded as the data of each section. A total of four sections from each animal were used for quantification. The final average number of the four sections was regarded as the data for each sample. The averages of 10 different fields of view were calculated for each animal. A total of five sections (with a minimum of 100 m from the next) from each animal were used for quantification. The final average number of the five sections was regarded as the data for each sample.
Brain Water Content (Brain Edema)-Brains were removed 24 h after surgery in experiment 2 and weighed immediately (wet weight) and reweighed after drying in 100°C for 72 h (dry weight). The percentage of water content was calculated as [(wet weight Ϫ dry weight)/wet weight] ϫ 100%. . D, representative photographs of rat brains after surgery. In the sham-operated rats, there is no blood clotting throughout the brain. In the SAH group, the inferior basal temporal lobe was always stained by blood. Hence, the brain tissue adjacent to the clotted blood was taken to the analysis. The site of the obtained sample was indicated with ovals. ##, p Ͻ 0.01; #, p Ͻ 0.05 versus control; n ϭ 6 in each group. Neurologic Scoring-Three behavioral activity examinations to record appetite, activity, and neurological deficits (Table 1) were performed by an investigator blinded to the study groups at 24 and 72 h after SAH using the scoring system reported previously (22,23).
Statistical Analyses-Data are expressed as mean Ϯ S.E. One-way ANOVA followed by Tukey test was used to analyze differences between groups except for the neurobehavioral scores, which were analyzed with nonparametric tests (Kruskal-Wallis, followed by Dunn's post-hoc test). SPSS 16.0 was used for the statistical analysis (SPSS, Inc., Chicago, IL). Differences were determined to be significant with p Ͻ 0.05.

Results
The Mortality and General Observation-No significant changes in body temperature or injected arterial blood gas data were detected in any of the experimental groups. Intracerebroventricular injection of vehicle or OZ did not significantly alter arterial blood gas and heart rates of rats. The mortality rate was 0% (0/41) in the control group and 18.6% (36/194) in the SAH group. The mortality among SAH, vehicle, and OZ treatment groups was not significantly different in experimental designs 2 and 3 (data not shown).
TAK1 Was Activated in Rats Post-SAH-To determine whether TAK1 is activated after SAH, tissue extracts from basal temporal lobe were examined for the levels of TAK1, p-TAK1 (Thr-187), and p-TAK1 (Ser-439) by Western blotting. Oneway ANOVA showed a significant difference in p-TAK1 (Thr-187) expression among the seven groups (F ϭ 5.88, p ϭ 0.003). Unexpectedly, the level of total TAK1 and p-TAK1 (Ser-439) were kept constant after SAH (Fig. 1, A and B, both p Ͼ 0.05). However, the level of p-TAK1 (Thr-187) showed a substantial increase at 24 (Fig. 1B, p Ͻ 0.01) and 48 h (Fig. 1B, p Ͻ 0.05) after SAH. Double immunofluorescence also demonstrated that TAK1 in the basal temporal lobe mainly co-localized with NeuN, a neuronal marker, without difference between control and SAH groups (Fig. 1C).
OZ Reduced the Levels of TAK1 and p-TAK1 in SAH Rats-To evaluate the role of TAK1 in early brain injury following SAH, the TAK1-specific inhibitor OZ was used to treat SAH and control rats. OZ treatment did not affect levels of p-TAK1 and TAK1 in control rats as compared with normal rats (data not shown), which is in line with data from a previous study (17). Statistical significance was revealed in TAK1 (F ϭ 7.928, p ϭ 0.001) and p-TAK1 (F ϭ 17.5, p Ͻ 0.001) expression among the four groups. OZ treatment significantly reduced the levels of TAK1 ( Fig. 2A, p Ͻ 0.01) and p-TAK1 (Thr-187) (Fig. 2B, p Ͻ 0.001) compared with vehicle-treated rats. Double immunofluorescence staining further confirmed the dramatic reduction of TAK1 by OZ administration in neuronal cells (Fig. 2C).
Neuronal Apoptosis Was Attenuated by TAK1 Inhibition-TAK1 inhibition was shown to play a neuroprotective effect in animal models of some diseases (15,17). Neuronal apoptosis is the main pathological process in EBI (1,24). Thus, TUNEL staining was performed and cleaved caspase-3, an indicator in the early phase of apoptosis (25), was detected. The results revealed a significant difference in the apoptosis index (F ϭ 43.5, p Ͻ 0.001) and cleaved caspase-3 expression (F ϭ 6.849,  AUGUST 7, 2015 • VOLUME 290 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 19903 p ϭ 0.002) among the groups. SAH increased the number of TUNEL positive cells in the basal temporal lobe at 24 h after SAH (Fig. 3, A and B, p Ͻ 0.001). Blockage of TAK1 decreased the amount of TUNEL positive cells in SAH rats compared with untreated animals (Fig. 3, A and B, p Ͻ 0.01). Likewise, we found a significant increase of cleaved caspase-3 in SAH and SAH ϩ vehicle groups compared with the control group (Fig.  3C, p Ͻ 0.05). Notably, OZ treatment significantly decreased the level of cleaved caspase-3 when compared with the SAH ϩ vehicle-treated samples (Fig. 3C, p Ͻ 0.05).

TAK1 Inhibition Attenuates Early Brain Injury after SAH
TAK1 Inhibition Down-regulated SAH-induced Activation of MAPK Signaling Pathway-To investigate the cellular mechanisms by which treatment with OZ may attenuate apoptosis, and further terminate the development of early brain injury, we evaluated activation by Western blot analysis of the MAPK family such as p-JNK, p-ERK1/2, and p-p38. One-way ANOVA revealed significant differences in p-p38 (F ϭ 15.34, p Ͻ 0.001), p-ERK1/2 (F ϭ 4.745, p ϭ 0.012), p-JNK (F ϭ 4.369, p ϭ 0.016), and p-c-Jun (F ϭ 10.12, p Ͻ 0.001) expression among the four groups. A significant increase in p-ERK1/2 (Fig. 4B, p Ͻ 0.05) and p-p38 (Fig. 4C, p Ͻ 0.001) expression was observed in SAH rats. Although the expression of p-JNK was found to increase in the SAH group as compared with the control group, this difference was not statistically significant (Fig. 4A, p Ͼ 0.05). Singledose intracerebroventricular administration of OZ prevented the SAH-induced phosphorylation of JNK (Fig. 4A, p Ͻ 0.05) and p38 (Fig. 4C, p Ͻ 0.05), but not ERK (Fig. 4B, p Ͼ 0.05). c-Jun, a downstream target for phosphorylation by JNK and an important factor in neuronal apoptosis induced by SAH, was evaluated. The result showed that OZ treatment reduced the level of activated c-Jun (Fig. 4D, p Ͻ 0.05), confirming the neuroprotective effect of OZ after SAH.
Effect of TAK1 Inhibition on SAH-induced Nuclear Translocation of NF-B p65 and Degradation of IB␣ Protein-In addition to MAPK signaling pathways, NF-B also plays a key role in the pathogenesis of early brain injury after SAH (6,26). To determine the effect of TAK1 inhibition on NF-B signaling in EBI in this study, we investigated the effect of OZ on IB, an inhibitor of NF-B. Our results revealed significant differences in IB␣ (F ϭ 13.4, p Ͻ 0.001) and p65 (F ϭ 10.21, p Ͻ 0.001) expression among the four groups. A basal level of IB␣ was high in the basal temporal lobe from rats of the control group, whereas it was substantially reduced in SAH rats and reversed in OZ treatment group (Fig. 5A), suggesting that OZ administration prevented the SAH-induced IB␣ degradation. Then, nuclear fractions were isolated from the basal temporal lobe and probed for NF-B subunit p65 via Western blot. The results demonstrated that the level of p65 significantly increased at 24 h after SAH compared with those in the control group and returned to the basal level after OZ treatment (Fig.  5B). To further analyze downstream NF-B and MAPKs signaling, the levels of TNF-␣ and IL-1␤ were also measured. Likewise, a statistically significant difference was revealed in TNF-␣ (F ϭ 4.97, p ϭ 0.009) and IL-1␤ (F ϭ 4.852, p ϭ 0.011) expression among the four groups. The levels of both IL-1␤ (Fig. 6A, p Ͻ 0.05) and TNF-␣ (Fig. 6B, p Ͻ 0.05) increased following SAH as has been reported previously (18). However, TAK1 inhibition had no effect on TNF-␣ and IL-1␤ levels (Fig. 6, A and B, both p Ͼ 0.05).
The Cerebral Edema after SAH Were Not Improved by OZ Administration-Brain edema, along with the development of neuronal apoptosis, is the main pathological process in EBI (1). To determine whether TAK1 inhibition reduces the cerebral edema, brain water content was evaluated at 24 h post-SAH. One-way ANOVA revealed significant differences in water content (F ϭ 4.444, p ϭ 0.015), ZO-1 (F ϭ 4.478, p ϭ 0.015), and occludin expression (F ϭ 20.03, p Ͻ 0.001) among the four groups. Although a significant increase in water content of the whole brain at 24 h after SAH was revealed compared with that in the control group (Fig. 7A, p Ͻ 0.05), the mean value was not decreased by TAK1 inhibition (Fig. 7A, p Ͼ 0.05). The results suggested that TAK1 inhibition could not attenuate brain edema in this rat SAH model. To confirm this at molecular levels, we assessed the effect of TAK1 inhibition on the expression levels of the tight junction proteins occludin and ZO-1, which play important roles in maintaining the function and blood-brain barrier integrity. Western blot analysis was performed using tissue homogenates from cortex harvested 24 h after SAH. The levels of occludin and ZO-1 were significantly lower in SAH and vehicle groups compared with the control group (Fig. 7, B and C, both p Ͻ 0.05). TAK1 inhibition failed to attenuate SAH-induced decreases of the protein levels of ZO-1 and occludin (Fig. 7, B and C, p Ͼ 0.05).
TAK1 Inhibition Ameliorated Clinical Neurological Function and Brain Tissue Damage after SAH-To further justify the protective effect of OZ on behavioral improvement, mean neu-rological scores were evaluated with a separate cohort of rats. Nonparametric tests revealed significant differences in neurological scores among different groups (Fig. 8A, both 24 and 72 h post-SAH: p Ͻ 0.001). SAH induced prominent impairment of the clinical behavioral function at 24 and 72 h (Fig. 8A, p Ͻ 0.001 versus control group). There were no significant differences between SAH and SAH ϩ vehicle-treated rats at both time points (Fig. 8A, p Ͼ 0.05). However, OZ-treated rats   exhibited significant improvement in clinical behavioral function at both 24 and 72 h after injury when compared with the vehicle-treated rats (Fig. 8A, p Ͻ 0.05). Furthermore, OZ also ameliorated the clinical behavioral function even when administered up to 4 h post-SAH (Fig. 8B). Finally, we stained the brain sections with cresyl violet (Nissl stain) to count the number of intact versus non-intact neurons 72 h post-SAH with and without OZ treatment. One-way ANOVA showed a significant difference in the number of intact neurons among the four groups (F ϭ 10.65, p Ͻ 0.001). A large proportion of neurons in the vehicle-treated group exhibited pyknotic or fragmented nuclei, whereas OZ treatment significantly increased the number of intact neurons compared with the vehicle group even when OZ administration was delayed by 4 h after SAH (Fig. 9).

Discussion
The main findings of this study are summarized as follows. First, TAK1 was activated by SAH although the levels of total TAK1 had no difference between the control and SAH groups. Second, direct intracerebroventricular delivery of OZ, a TAK1 inhibitor, reduced the severity of experimental SAH. Third, delayed treatment with a clinical relevant time window improved the neurologic deficits. Fourth, selectively reduced levels of p-p38, p-JNK, and its downstream target p-c-Jun and prevented nuclear translocation of NF-B p65 may be involved in the mechanism of neuroprotection of TAK1 inhibition. TAK1 can be activated by many upstream cytokines, such as TGF-␤, IL-1␤, TNF-␣, and Toll-like receptor ligands as illustrated in Fig. 10. Stimulation of IL-1R and Toll-like receptors activates myeloid differentiation primary response gene 88 (MyD88)-dependent pathways and MyD88 recruits TNF receptor-associated factor 6 (TRAF6), IL-1 receptor-associated kinase (IRAK1), and IRAK4. TRAF6 acts with ubiquitin-conjugating enzyme 13/ubiquitin E2 variant 1a to catalyze lysine (K)63 polyubiquitination of the TAK1 protein kinase complex. For full activation, TAK1 forms a heterotrimeric complex with TAB1 and TAB2 or TAB3, which strongly interact with K63linked polyubiquitin chains to drive TAK1 activation by ubiquitination and autophosphorylation. The activated TAK1 complex triggers phosphorylation and activation of MAPKs and NF-B pathways (27). The downstream effects of TAK1 are mediated via phosphorylation of multiple residues in its activation loop such as Ser-192, Thr-178, Thr-184/187, and Ser-412 (28 -30). Despite the expression of TAK1, no prominent alternation after SAH in the basal cortex, where acute brain injury was the most severe, was shown. Our Western blot analysis revealed that phosphorylation of p-TAK1 (Thr-187) was significantly increased after SAH, whereas the phosphorylation of p-TAK1 (Ser-439) showed no significant alteration. Taking into account that phosphorylation of the Thr-187 residue in TAK1 could simultaneously induce the activation of NF-B and MAPK signaling pathways (28), we speculated that Thr-187 phosphorylation might play an important role rather than Ser-439 in this animal model of SAH. It has been well demonstrated that OZ acts as a potent and selective inhibitor of TAK1 (Fig.  10), which displays Ͼ33and Ͼ62-fold selectivity over MEKK1 and MEKK4, respectively (31). Mechanistically, OZ forms a covalent complex with TAK1 and inhibits both the kinase and ATPase activity of TAK1 following a biphase kinetics (32). Importantly, OZ is a relatively safe drug found in vivo and has been previously reported to cross the blood-brain barrier (17,33,34). The target dose used in the current study was chosen based on our published study (15) and treatment of SAH rats with 25 g of OZ did not induce a significant difference in body temperature, injected arterial blood gas data, and mortality rate compared with the SAH vehicle-treated groups. Our data also showed that OZ did not only suppress the expression of p-TAK1 but also TAK1, which was consistent with our previous and other studies (15,35,36). However, the exact mecha-

TAK1 Inhibition Attenuates Early Brain Injury after SAH
nism that OZ decreased in TAK1 expression needs to be elucidated in the future.
Several studies have reported that TAK1 is a critical regulator of apoptosis, which is mainly mediated by the downstream pathways of p38 MAPKK, JNK, ERK-1/2, and NF-B p65 (15,16,37). Evidence shows that activation of MAPK and NF-B pathways contributes to EBI after SAH, and inhibition of these pathways offer neuroprotective effects against SAH (5,24,26).
To identify the molecular mechanisms by which TAK1 inhibition provides neuroprotection after SAH, we first focused on activation of the MAPK pathway. Inhibition of TAK1 was found to be selectively prevented by the SAH-induced phosphorylation of p38 and JNK, but not ERK1/2. Then we found that inhibition of TAK1 resulted in less NF-B activation and less neuronal apoptosis in vivo (15). NF-B has emerged as one of the most promising molecular targets in the prevention of EBI (6,26). NF-B resides in the inactive state in the cytoplasm as a heterotrimer consisting of p50, p65, and IB␣ subunits. An IB␣ kinase, IKK, phosphorylates serine residues in IB␣ at position. Upon phosphorylation and subsequent degradation of IB␣, NF-B activates and translocates to the nucleus, where it binds to DNA and activates the transcription of various genes (38). In this study, we report that SAH was associated with significant IB degradation as well as an increased nuclear localization of p65 in the basal temporal lobe at 24 h after SAH. TAK1 inhibition significantly reduced IB degradation as well as NF-B p65 nuclear translocation. Together, the above results demonstrate a selective involvement of TAK1 in the various signaling pathways mobilized in the pathological process of EBI following SAH.
TAK1 activation is also known to induce proinflammatory gene expression leading to the production of proinflammatory cytokines and chemokines and activation of immune cells (14). Importantly, conditional depletion of TAK1 in microglia only significantly reduced CNS inflammation and diminished axonal and myelin damage by cell-autonomous inhibition of the NF-B, JNK, and ERK1/2 pathways (39). SAH has long been known to induce an inflammatory response in blood vessels and neuronal tissues (40 -42). Elevation of TNF-␣ and IL-1␤ were reported in the early stage after SAH and leads to an exacerbation of EBI (10,43). Consistent with previous work, our data demonstrate that SAH induced a significant increase of TNF-␣ and IL-1␤ levels. However, there were no differences in TNF-␣ and IL-1␤ levels between vehicle and drug-treated SAH animals, a possible explanation for the phenomenon is that TAK1  . Schematic illustrating the proposed mechanism of neuroprotective role of TAK1 inhibition by OZ. TAK1 plays a central role in several signaling pathways involved in early brain injury after SAH. As illustrated, several toxic stimuli (e.g. TNF-␣ and IL-1␤) induced by SAH cause an increase in TAK1 activity in neurons, leading to NF-B, JNK, and p38 activation, which further activate the apoptotic pathway. Inhibition of TAK1 by OZ will block the signal transduction to protect the neurons from apoptosis, eventually prevent the EBI. AUGUST 7, 2015 • VOLUME 290 • NUMBER 32 was expressed mainly in neurons but TNF-␣ and IL-1␤ are generally thought to be released from glia (44). Thus, we speculate that the anti-inflammatory properties of OZ may play a less significant role in neuroprotection.

TAK1 Inhibition Attenuates Early Brain Injury after SAH
Increasing evidences support that SAH significantly decreases the levels of tight junction proteins, occludin and ZO-1 (45). Altered expression of these tight junction proteins could cause blood-brain barrier breakdown following SAH leading to brain edema (45). Global edema is an independent risk factor for mortality and a poor outcome after SAH (46). In the present study, our data demonstrated that SAH degraded occludin and ZO-1, whereas TAK1 inhibition could not restore these proteins. As a result, there was no difference between vehicle and the OZ treatment group with a single treatment in brain water content at 24 h after SAH. These data suggest that inhibition of TAK1 by OZ could not alleviate brain edema and blood-brain barrier disruption after SAH.
In conclusion, the present study identified TAK1 as a novel upstream mediator of the apoptosis in EBI following SAH (Fig.  10). Selective inhibition of TAK1 by OZ reduced the activation of JNK, p38, and NF-B, effectively prevented neuronal apoptosis, and attenuated neurological deficits in SAH. These data shed new light on the treatment of SAH, and suggest that OZ may be an effective drug therapy for EBI after SAH. However, we used only one single-dose intracerbroventricular treatment to evaluate the neuroprotective role of OZ in the current study. Whether systemic injection of OZ provided similar beneficial effects remains to be further investigated.