Neuron-Specific Tumor Necrosis Factor Receptor–Associated Factor 3 Is a Central Regulator of Neuronal Death in Acute Ischemic StrokeNovelty and Significance
Neuronal death after ischemic stroke involves multiple pathophysiological events, as well as a complex molecular mechanism. Inhibiting a single therapeutic target that is involved in several ischemic signaling cascades may be a promising strategy for stroke management. Here, we report the versatile biological roles of tumor necrosis factor receptor–associated factor 3 (TRAF3) in ischemic stroke. Using several genetically manipulated mouse strains, we also demonstrated that TRAF3 inhibition can be neuroprotective. TRAF3 expression, which is robustly induced in response to ischemia/reperfusion (I/R) injury, was detected in neurons. Overexpression of TRAF3 in neurons led to aggravated neuronal loss and enlarged infarcts; these effects were reversed in TRAF3-knockout mice. Neuronal TRAF3 also contributed to c-Jun kinase–, nuclear factor κB– and Rac-1–induced neuronal death, inflammation, and oxidative stress. Mechanistically, we showed that TRAF3 interacts with transforming growth factor-β–activated kinase 1 (TAK1) and potentiates phosphorylation and activation of TAK1. Phosphorylated TAK1 sequentially initiated activation of nuclear factor κB, Rac-1/NADPH oxidase, and c-Jun kinase/c-Jun signaling cascades. Using a combination of adenoviruses encoding dominant-negative TAK1 and the TAK1 inhibitor 5Z-7-oxozeaenol, we demonstrated that the TRAF3-mediated activation of ischemic cascades was TAK1-dependent. More importantly, the adverse phenotypes observed in TRAF3-overexpressing mice were completely reversed when the TRAF3–TAK1 interaction was prevented. Therefore, we have shown that TRAF3 is a central regulator of ischemic pathways, including nuclear factor κB, Rac-1, and c-Jun kinase signaling, via its interaction with and activation of TAK1. Furthermore, certain components of the TRAF3–TAK1 signaling pathway are potentially promising therapeutic targets in ischemic stroke.
See Editorial Commentary, pp 472–473
Stroke is a common neurological disorder worldwide and a leading cause of mortality in Western countries. Despite the availability of recombinant tissue-type plasminogen activator, which demands a strictly limited therapeutic window, no drugs are available to spare cerebral tissues from the rapid damage that occurs in stroke.1 Therefore, there is an urgent need for novel therapeutic options. Neuronal death during ischemic stroke involves intricate and complicated signaling networks. Several pathological mechanisms contribute to cell death during this period, including apoptosis, inflammation, oxidative stress, and excitotoxicity. All attempts in clinical trials to inhibit a single pathological event have failed.2 Indeed, the molecular mechanisms responsible for neuronal loss in ischemic stroke are complex and poorly understood. To overcome such difficulties, the use of a cocktail of drugs targeting neuroprotective pathways could be a successful strategy. Transforming growth factor-β–activated kinase 1 (TAK1) is a potential candidate for this type of strategy because TAK1 is an essential upstream modulator of several signaling cascades, including c-Jun kinase (JNK), p38, and nuclear factor κB (NF-κB), in ischemic injury.3 Neubert et al4 recently reported that inhibition of TAK1 is neuroprotective in cerebral ischemia. Nevertheless, although TAK1 can be stimulated by interleukin (IL)-1β, tumor necrosis factor (TNF), and Toll-like receptors (TLRs),3 it is not fully understood how it is activated in ischemic stroke. Thus, identification and, more importantly, substantial insight into the precise molecular regulation of such central regulators in ischemic stroke may be crucial for developing a promising therapeutic strategy.
The tumor necrosis factor receptor–associated factor (TRAF) family, which comprises 7 members (TRAF1–7), is a family of adaptor proteins that mediate signal transduction from TNF receptors, TLRs, and the IL-1 superfamily.5 TRAF3 is ubiquitously expressed in most tissues and cell types, including brain, lung, heart, spleen, and liver.5,6 Because TRAF3 deletion in mice results in early lethality,6 our understanding of its biological function has remained rudimentary for decades compared with other TRAFs. In addition, early studies failed to detect activation of canonical NF-κB and JNK pathways on TRAF3 overexpression as was easily detected after TRAF2 or TRAF6 overexpression.7 After the application of TRAF3 conditional knockout mice in recent studies, TRAF3 has been recognized as a highly versatile regulator of immunity-related signal transduction, exerting various biological functions that are highly dependent on the cell type and the receptor used.5 TRAF3 interacts with multiple receptors, kinases, adaptors, and regulatory proteins to exert its diverse biological functions. Although the canonical NF-κB pathway was unaffected in mice, Perez de Diego et al8 recently showed that R118W-mutant TRAF3 was responsible for impaired TLR-induced nuclear translocation of the NF-κB p65 subunit in simian virus 40–transformed human fibroblasts. Although it acts as a negative regulator of JNK, TRAF3 may have the capacity to activate JNK under certain circumstances, such as TRAF3 myristoylation, which forces it to the membrane.9 The role of TRAF3 in negative regulation of JNK and alternative NF-κB pathways has been extensively studied, but the underlying molecular mechanism and pathophysiological role of TRAF3-regulated classic NF-κB and JNK activation remain poorly understood. Furthermore, NF-κB and JNK are crucial regulators of key pathological events, including inflammation and cell survival, in ischemic stroke.10,11 These studies suggested a potential role for TRAF3 in multiple ischemic signaling cascades, although the biological roles of TRAF3 in neurological diseases remain largely unknown.
This study identified TRAF3 as a central modulator of ischemic signaling cascades, including neuronal death, neuroapoptosis, inflammation, and oxidative stress. We observed that TRAF3 was significantly upregulated in neurons from mouse brains after middle cerebral artery occlusion (MCAO)/reperfusion, even at 2-h postoperation. Comparing neuron-specific TRAF3-transgenic with TRAF3-knockout mice, we demonstrated that the transgenically induced TRAF3 expression in neurons rendered the cerebral tissue more vulnerable to (I/R)-induced inflammation, oxidative stress, and neuronal loss. Mechanistically, we revealed that TRAF3 mediated activation of the JNK prodeath pathway, the NF-κB proinflammatory pathway, and the Rac-1 pro-oxidant pathway. Thus, targeting TRAF3 may exert multiple effects simultaneously, making TRAF3 a promising therapeutic target for stroke management.
Extended Materials and Methods are available in the online-only Data Supplement.
TRAF3flox/flox mice were generated as previously described.12 The TRAF3flox/flox line was backcrossed with C57BL/6J (B6) mice for >9 generations to generate TRAF3flox/flox mice on the C57 genetic background. TRAF3flox/flox homozygous mice were crossed with CaMKIIα-Cre mice, resulting in neuron-specific TRAF3-knockout mice. The neuron-specific TRAF3-transgenic mice were generated by crossing TRAF3-floxed mice and CaMKIIα-Cre mice. Only 11- to 12-week-old (25–30 g) males were used.
TAK1 Inhibitor Treatment
The TAK1 inhibitor 5Z-7-oxozeaenol (5Z-7; O9890-1MG; Sigma, St. Louis, MO) dissolved in dimethyl sulfoxide (0.8 μg/μL); 5Z-7 (2 μL) was intracranially injected into nontransgenic and TRAF3-transgenic mice 30 minutes before MCAO.
Surgical Procedure for MCAO
Focal cerebral ischemia was achieved by transiently occluding the left middle cerebral artery as previously described.13 Regional cerebral blood flow was monitored by Doppler analysis (Periflux System 5010; Perimed, Järfälla, Sweden). Sham control mice received the same procedure without filament insertion.
Neurological Deficit Scores
Quantification of Infarct Volume
Cerebral infarct was determined by 2,3,5-triphenyl-2H-tetrazolium chloride staining of brain sections, and the images were analyzed using Image-Pro Plus 6.0 (Media Cybernetics, Bethesda, MD). The infarct volume (%) was calculated as follows: (volume of the contralateral hemisphere−the volume of the nonlesioned ipsilateral hemisphere)/(contralateral volume×2) ×100%.13
Immunofluorescent, Terminal Deoxynucleotidyl Transferase–Mediated dUTP-Biotin Nick End Labeling, and Fluoro-Jade B Staining
Immunocytochemistry was performed as previously described.16,17 A terminal deoxynucleotidyl transferase–mediated dUTP-biotin nick end labeling assay was performed using the ApopTag Plus In Situ Apoptosis Fluorescein Detection Kit (S7111; Millipore) according to the manufacturer’s protocol. For Fluoro-Jade B staining, the sections were stained with the fluorescent dye Fluoro-Jade B (AG310; Millipore) according to the manufacturer’s protocol.
Measurement of Brain Superoxide Levels Using Dihydroethidium
Superoxide levels were assessed using dihydroethidium (D-23107; Invitrogen). The nuclei were stained with 4′,6-diamidino-2-phenylindole.
Cell Culture and In Vitro I/R Model
Primary cortical neurons were collected from 1-day-old Sprague–Dawley rats as previously described.18 For oxygen and glucose deprivation (OGD), neurobasal medium was replaced with serum-free, glucose-free Locke buffer (154 mmol/L NaCl, 5.6 mmol/L KCl, 2.3 mmol/L CaCl2, 1 mmol/L MgCl2, 3.6 mmol/L NaHCO3, 5 mmol/L HEPES, and 5 mg/mL gentamicin; pH 7.2) and incubated in an experimental hypoxia chamber in a saturated atmosphere of 90% N2 and 5% CO2. The neuronal cells underwent OGD for 60 minutes and were then returned to normal culture conditions at different time points.
Adenoviral Vector Construction and Neuronal Infection
For the in vitro studies, we constructed adenoviruses carrying sequences encoding mouse TRAF3 (AdTRAF3), mutated TRAF3 (AdTRAF3-M), short hairpin RNA targeting TRAF3 (AdshTRAF3), constitutively active TAK1 (Adca-TAK1), and dominant negative TAK1 (Addn-TAK1), as previously described.19 For the in vivo studies, we intracranially injected 2×1010 multiplicity of infection of AdGFP, AdTRAF3, or AdTRAF3-M into wild-type (WT) mice 48 hours before MCAO.
Analyses of Cell Viability and Lactate Dehydrogenase Release
Cell viability and lactate dehydrogenase release were examined using a nonradioactive cell counting kit-8 assay (CK04, Dojindo, Kumamoto, Japan) and a colorimetric lactate dehydrogenase cytotoxicity assay (G1782, Promega, Madison, WI), respectively, according to the manufacturer’s instructions.
The NF-κB–dependent luciferase reporter (pGL3-NF-κB) and pRL-thymidine kinase vectors were purchased from Promega. The NF-κB activity assay was performed as previously described.20
Immunoprecipitation and GST Pull-Down Assay
Cultured HEK293T cells were cotransfected with Flag-TAK1 and enhanced green fluorescent protein–Myc-TRAF3 for 48 hours; cell lysates were precleared and then incubated with 1 μg of antibody and 10 μL of protein A/G-agarose beads on a rocking platform at 4°C overnight. The immunocomplex was collected, washed, and blotted using the indicated primary antibodies. Glutathione s-transferase pull-down is described in the supplementary Methods.
The data were expressed as the means±SD. Differences between groups were assessed by analysis of variance followed by post hoc Tukey test. An unpaired Student t test was used to compare the 2 groups. P<0.05 was accepted as statistically significant. The in vivo and imaging studies were performed in a blinded manner.
TRAF3 was Upregulated in Modeled Ischemia
To investigate the role of TRAF3 in ischemic stroke, we first analyzed TRAF3 expression in mouse cerebral tissues, in which transient ischemia was induced by MCAO for 45 minutes followed by reperfusion for 2 to 72 hours. Interestingly, we observed a time-dependent induction of TRAF3 expression between 2 hours and 24 hours post-I/R (Figure 1A). A similar expression pattern was observed in OGD-challenged cultured primary cortical neurons (Figure 1B), suggesting that neurons may be a potential target of TRAF3 during stroke. Thus, we next studied TRAF3 expression and localization in neurons by costaining for TRAF3 and a neuron-specific marker, NeuN. As expected, TRAF3 was detected in neurons in the hippocampus, cortex, and striatum at baseline, but its expression level was upregulated after MCAO only in the latter 2 regions, where blood flow was interrupted (Figure 1C). Consistent with these observations, in vitro studies using primary cortical neurons further showed that I/R injury enhanced TRAF3 expression in neurons (Figure S1). These data indicate that TRAF3 induction is associated with cerebral damage during ischemic stroke.
TRAF3 Regulates Stroke Outcomes After I/R
To determine the functional role of neuronal TRAF3 induction at stroke onset, we generated four neuron-specific TRAF3-overexpressing transgenic mouse lines driven by the cytomegalovirus enhancer promoter (Figure S2A). Cerebral expression was elevated by 1.64- to 2.18-fold that of NTG control littermates (Figure S2B). Two mouse lines transgenic 3 and transgenic 4 that exhibited moderate TRAF3 overexpression were used for further experiments. Cerebral infarcts were detected by 2,3,5-triphenyl-2H-tetrazolium chloride staining after 24 hours or 72 hours of MCAO/reperfusion (Figure 2A). Interestingly, TRAF3–transgenic 4 displayed an increase in infarct volumes of ≈49.8% and 62.2% at 24-hour and 72-hour poststroke, respectively (Figure 2B). A similar effect on infarct sizes was observed in TRAF3–transgenic 3 mice (Figure 2B). Neurological deficits were also more severe in the transgenic lines at both the time points (Figure 2C), indicating that the induction of neuronal TRAF3 expression after ischemic stroke could be both biologically and functionally deleterious. Therefore, we speculated that TRAF3 ablation may be beneficial for stroke outcomes. To test this hypothesis, we generated neuron-specific TRAF3-knockout mice using the Cre/loxP system (hereafter referred to as TRAF3-knockout mice). The genotype was validated by immunoblotting (Figure S2C). As anticipated, deleting TRAF3 led to a significant reduction in cerebral infarct size compared with WT controls at 24-hour and 72-hour post-I/R, respectively (Figure 2D and 2E). In addition, the infarct sizes were comparable between TRAF3flox/flox mice and WT mice (Figure 2E). Therefore, WT mice were used as controls for the TRAF3-knockout mice in the following studies. Accordingly, the neurological scores were also significantly improved in TRAF3-knockout mice compared with controls (Figure 2F). Neuronal TRAF3 is, therefore, an endogenous positive regulator of cerebral injury in ischemic stroke.
Pro-Death Effects of TRAF3 on Neurons
On the basis of these observations, as well as because neurons are particularly vulnerable owing to their high metabolic demand during ischemic stroke,21 we investigated the biological effect of TRAF3 on neuronal death. The brain was stained with either Fluoro-Jade B for neuronal death or terminal deoxynucleotidyl transferase–mediated dUTP-biotin nick end labeling for neuroapoptosis (Figure 3A). Compared with WT controls, TRAF3-knockout mice showed a significant reduction in Fluoro-Jade B-positive neurons and terminal deoxynucleotidyl transferase–mediated dUTP-biotin nick end labeling–positive neurons after 24 hours of I/R (Figure 3B). Conversely, TRAF3 overexpression rendered neurons more vulnerable (Figure 3B). However, whether TRAF3 directly mediates neuronal death in response to I/R remains unknown. To address this issue, we evaluated the prodeath effect of TRAF3 in cultured rat primary neurons in vitro because of the relative homogeneity within rat strains. Recombinant adenoviruses harboring either WT TRAF3 (AdTRAF3) or a short-hairpin against TRAF3 (AdshTRAF3) were constructed (Figure S2D). Neurons subjected to OGD/reperfusion showed significant neuronal loss, as indicated by decreased cell survival and exaggerated lactate dehydrogenase release (Figure S3A and S3B). This effect, however, was largely ablated by AdshTRAF3 infection and was further enhanced by ≈3.47-fold overexpression of TRAF3 (Figure S3A and S3B). Neuronal apoptosis involves the mitochondrial signaling pathway comprising the proapoptotic factors Fas, FasL, Bax, and Bad, as well as the antiapoptotic factor Bcl-2.22 In TRAF3-knockout brains, the mRNA levels of the proapoptotic genes were significantly downregulated, whereas antiapoptotic Bcl2 was upregulated (Figure S3C). This effect was reversed on TRAF3 overexpression (Figure S3C). In accordance with these data, TRAF3 positively regulated Bax and cleaved-caspase 3 levels and negatively regulated Bcl2 and Bcl-xL levels both in vivo (Figure 3C) and in vitro (Figure S3D). Taken together, these findings indicate that TRAF3 is a direct mediator of neuronal death and apoptotic susceptibility in ischemic stroke.
TRAF3 Activates the Map Kinase Kinase/JNK Cell Death Signaling Pathway
Because TRAF3 had a profound prodeath effect on neuronal cells and manipulated the expression level of a series of apoptotic genes, we hypothesized that it may play a role in the regulation of upstream death and survival factors. The mitogen-activated protein kinase (MAPK) family members, namely p38, extracellular signal-regulated kinase, and JNK, are emerging regulators of neuronal cell death during ischemic stroke.23 Interestingly, we found that ablation and overexpression of TRAF3 had no impact on the expression of phosphorylation of p38 and extracellular signal-regulated kinase both in vivo (Figure S3E) and in vitro (Figure S3F). In contrast to p38 and extracellular signal-regulated kinase, JNK and its major upstream and downstream effectors, map kinase kinase 4/7 and c-Jun, were significantly phosphorylated at 6-hour post-I/R. This effect was further enhanced in TRAF3-transgenic mice but was dampened by TRAF3 deletion (Figure 3D). Thus, TRAF3 may facilitate ischemic brain injury via activation of the prodeath JNK signaling pathway and exacerbation of neuronal loss. Indeed, the activation of the map kinase kinase/JNK/c-Jun pathway was positively regulated by TRAF3 expression in vitro (Figure S3G). This further supports the hypothesis that TRAF3 mediated ischemic-induced neuronal death via JNK activation.
TRAF3 Enhances the NF-κB Proinflammatory Signaling Pathway
Focal ischemia evokes the production of cytokines, such as TNF-α and IL-1β, and further upregulates adhesion molecules that mediate inflammatory cell recruitment, which characterizes the secondary or delayed response to ischemia.1 Thus, we analyzed the mRNA expression profiles of known proinflammatory mediators, including TNF-α, monocyte chemoattractant protein-1, and IL-1β, in the brains of TRAF3-knockout and TRAF3-transgenic mice at 24 hours after MCAO/reperfusion. The expression of these cytokines was significantly lower in TRAF3-knockout mice and significantly higher in TRAF3-transgenic mice than in their respective littermates (Figure 4A), suggesting the involvement of an upstream inflammatory regulator. The transcription factor NF-κB is a central regulator of the inflammatory response, which initiates the induction of the proinflammatory mediators in ischemic stroke. As demonstrated by the potentiated phosphorylation of IκBα, I kappa B kinase, and p65, NF-κB signaling was activated in mouse brains in response to I/R at 6 hours (Figure 4B). TRAF3 expression in neurons positively regulated this effect in mice during I/R (Figure 4B). Accordingly, activation of the NF-κB signaling pathway (Figure S4) and NF-κB luciferase activity (Figure 4C) were positively regulated by TRAF3 expression in vitro, suggesting that NF-κB is also a downstream target of TRAF3 in ischemic stroke.
TRAF3 Promoted Rac-1–Induced Oxidative Stress
Because antioxidant defenses are limited, oxidative stress is a powerful mediator of cerebral damage.1 Superoxide generation was analyzed by dihydroethidium fluorescent staining. TRAF3-transgenic mice exhibited ≈1.9-fold more dihydroethidium fluorescence compared with nontransgenic controls, whereas TRAF3 deletion was antioxidant (Figure 5A). Recent studies showed that NADPH oxidase is a principal enzyme of superoxide generation in ischemic stroke.1 Thus, we measured the mRNA levels of the antioxidant transcriptional factor Nrf2 and its effectors, glutathione peroxidase and manganese superoxide dismutase, as well as the pro-oxidant NADPH oxidase subunits p47-phox, p67-phox, and gp91-phox in mouse brains 24 hours after I/R (Figure S5A). Accordingly, TRAF3 overexpression significantly increased the mRNA levels of NADPH oxidase subunits and decreased the mRNA levels of Nrf2 and its downstream targets (Figure S5A). Consistent with these results, the effects were confirmed using immunoblotting both in vivo (Figure 5B) and in vitro (Figure S5B). A recent study showed that the Rho GTPase Rac-1 plays a crucial role in NADPH oxidase activation in I/R-induced brain injury.24 Interestingly, Rac-1 was phosphorylated and activated in mouse brains in response to I/R insults (Figure 5C). Consistent with TRAF3-dependent NADPH oxidase activation, this effect was enhanced by TRAF3 overexpression but was counteracted by TRAF3 ablation both in vivo (Figure 5C) and in vitro (Figure S5C). Collectively, these data demonstrate that TRAF3-mediated activation of Rac-1/NADPH oxidase may be responsible for adverse oxidative stress during ischemic stroke.
TRAF3 Facilitates Neuronal Death via Phosphorylation of TAK1
On the basis of the observation that TRAF3 promoted JNK-mediated cell death, NF-κB–mediated inflammation and Rac-1/NADPH oxidase–mediated oxidative stress, we hypothesized that TRAF3 may serve as a key upstream regulator of these molecular pathways. TAK1, which is also highly expressed in the brain, has recently emerged as a central activator of JNK and NF-κB in ischemic stroke.3,4,25 However, the regulation of TAK1 activation, particularly during ischemic stroke, is poorly understood. Phosphorylation of TAK1 at Thr-178 and Thr-184 is critical for the activation of JNK and NF-κB.26 Accordingly, we found that mice treated with MCAO/reperfusion showed significantly increased levels of activated TAK1 (Figure 6A). TRAF3 deletion, however, reduced the p-TAK1/TAK1 ratio to 54.9% that of WT controls, whereas TRAF3 overexpression exerted opposite effects both in vivo (Figure 6A), and in vitro (Figure S6A) and in cortical neurons in vivo (Figure S6B). To further confirm that TRAF3-mediated TAK1 activation is vital for TRAF3-mediated neuronal death, we generated adenoviruses harboring WT TAK1 (Adca-TAK1) or a dominant negative form of TAK1 (Addn-TAK1). Cell survival assays were performed with cultured primary neurons, in which the neuroprotective effect of TRAF3 interference (AdshTRAF3) was completely ablated when coinfected with Adca-TAK1 (Figure S6C). Moreover, TRAF3 overexpression failed to initiate neuronal death in the presence of the Addn-TAK1 (Figure S6D). Thus, TAK1 is both sufficient and required for TRAF3-mediated neuronal death. These findings are in accordance with the observation that TRAF3-mediated phosphorylation of JNK, I kappa B kinase, and Rac-1 was TAK1-dependent (Figure S6E). Together, these results indicate that TRAF3 is a critical regulator of TAK1-dependent neuronal death.
TAK1 Inhibition Rescues TRAF3-Mediated Cerebral Injury in Stroke
Having determined that TRAF3-mediated neuronal death is TAK1-dependent, we next examined whether TAK1 is essential for TRAF3-mediated brain damage. We systemically administered nontransgenic and TRAF3-transgenic mice with either 5Z-7 to inhibit TAK1 or dimethyl sulfoxide as a control; 5Z-7 treatment largely abolished TAK1 phosphorylation in TRAF3-transgenic mice (Figure S7A). Prevention of TAK1 activation rendered TRAF3-transgenic mice more resistant to I/R injury, as demonstrated by relatively reduced infarct volumes as well as improved neurological scores in the 5Z-7-treated TRAF3-transgenic mice compared with the dimethyl sulfoxide -treated TRAF3-transgenic control mice (Figure S7B-D). In addition, blocking TAK1 activation also counteracted TRAF3-mediated neuronal death and apoptosis (Figure S7E and S7F). Importantly, TRAF3-transgenic mice treated with 5Z-7 displayed significantly lower levels of phosphorylated JNK/c-Jun (Figure S7G), NF-κB (Figure S7H) and Rac-1/NADPH oxidase (Figure S7I) signaling pathway components. Thus, TAK1 inhibition could be a potentially effective strategy to ameliorate TRAF3-mediated cerebral injury in ischemic stroke.
TRAF3–TAK1 Interaction is Required for TRAF3-Dependent Cerebral Damage
Having demonstrated that TRAF3-mediated cerebral injury is TAK1-dependent, we next investigated whether TAK1 is a direct target of TRAF3. HEK293T cells were transfected with Myc-tagged TRAF3 and Flag-tagged TAK1. TRAF3 coimmunoprecipitated with TAK1 and vice versa (Figure 6B and 6C). The TRAF3–TAK1 interaction was further supported by the glutathione s-transferase pull-down assay results (Figure S8A). Next, we sought to determine which protein domains are essential for the TRAF3–TAK1 interaction. A series of truncated Myc-tagged TRAF3 (Figure S8B) and Flag-tagged TAK1 constructs were generated (Figure S8C) and cotransfected with Flag-tagged TAK1 and Myc-tagged TRAF3, respectively. TRAF3 comprises a C-terminal (TRAF) domain that is responsible for interactions with receptors and other signaling proteins and an N-terminal domain that includes a RING finger domain and several zinc finger motifs.5–7 Co-immunoprecipitations were performed, and the C-terminal domain, but not the N-terminal domain, of TRAF bound to TAK1 (Figure S8B). Interestingly, TRAF3 bound only to the C-terminal TAB2/3-binding domain (TAB2/3 BD) of TAK1, which was also critical for its interaction with TAB2 and TAB3 and for subsequent activation of TAK1 (Figure S8C).27 However, whether the TAK1–TRAF3 interaction is sufficient to mediate TRAF3/TAK1-mediated neuronal death remains unknown. Thus, we constructed an adenoviral vector harboring TRAF3 with a mutant TRAF domain (AdTRAF3-M). As expected, AdTRAF3 but not AdTRAF3-M promoted TAK1 phosphorylation in cultured neurons challenged with OGD/reperfusion (Figure S8D). Furthermore, Ad-TRAF3-M infection failed to increase the susceptibility of neurons to the OGD-induced death that was observed in the AdTRAF3-infected group (Figure S8E). Accordingly, TRAF3-mediated activation of apoptotic signaling as well as of JNK, NF-κB, and Rac-1 signaling was abolished when the TAK1-binding TRAF domain was mutated (Figure S8F-I). To further detect the significance of the TRAF3–TAK1 interaction, we next examined its biological function in vivo. Intracranial injection of AdTRAF3 in mice resulted in an adverse stroke outcome similar to that of TRAF3-transgenic mice (Figure S8J-L). Importantly, this effect was absent in mice injected with AdTRAF3-M (Figure S8K and S8L). Therefore, our data indicate that in response to I/R, TRAF3 expression is upregulated in neurons, resulting in TRAF3–TAK1 interaction and sequential neuronal loss.
Ischemic stroke involves multiple pathophysiological events and a complex molecular mechanism, which to some extent has impeded successful translation of potential therapeutic targets into medicine.1 Nevertheless, cocktails of drugs that each target a single signaling cascade during stroke may not be feasible given difficulties at the industrial, pharmacokinetic, and patent levels.2 Therefore, a single therapeutic target that exerts multiple effects on stroke outcomes may represent a more promising strategy. This study revealed that TRAF3 may be a central regulator of neuronal death, inflammatory responses, and oxidative stress in ischemic stroke. In addition, using genetically manipulated mouse strains, we demonstrated that TRAF3 deletion can be neuroprotective. We showed that TRAF3 expression is upregulated in neurons after I/R injury. Importantly, neuronal TRAF3 interacted with TAK1, thus potentiating phosphorylation and activation of TAK1. TAK1 sequentially activated the JNK/c-Jun, NF-κB, and Rac-1 signaling pathways, leading to enhanced neuronal death, inflammation, and oxidative stress (Figure 6D). These combined adverse effects resulted in enlarged cerebral infarctions and increased neurological deficits poststroke. Notably, this study demonstrated for the first time that TRAF3 functions upstream of ischemic signaling cascades by directly binding to TAK1 and that the TRAF3–TAK1 interaction is a critical step in I/R-induced brain damage. The following key observations support these conclusions: (1) the binding domain of TRAF3 is essential for TRAF3-mediated TAK1 phosphorylation and activation, (2) prevention of TRAF3 from binding to TAK1 abolished the adverse effects of TRAF3 on neuronal death and activation of ischemic cascades, and (3) inhibition of TAK1 activity was sufficient to rescue mouse brains from excessive cerebral injury resulting from TRAF3 overexpression.
One of the most relevant findings of this study is that TAK1 can be activated via direct interaction with TRAF3. Although several stimuli of TAK1–TLR2, TLR4, TNF, and IL-1β are known, the regulation of TAK1 in response to I/R challenge, especially in the neurological system, is poorly understood.4,5,28 Intriguingly, previous studies recognized TRAF3 as a negative regulator of TAK1/MAPK in the immune system as degradation of TRAF3 results in cytoplasmic translocation of the MYD88-associated multiprotein complex and sequential activation of the TAK1/MAPK kinase/MAPK signaling pathway.5,29 These effects are likely because of the highly versatile cell type- and receptor-dependent modes by which TRAF3 functions.5 Indeed, the interaction of TRAF3 with TIR domain-containing adaptor protein–inducing IFNβ also led to activation of the MAPK/JNK pathway, albeit with slower kinetics.5 Using a combination of truncated TRAF3 and TAK1 constructs in coimmunoprecipitation assays, we showed that during ischemic stroke, TRAF3 could bind directly to TAK1 with its N-terminal TRAF domain. The TRAF domain is critical for the interaction of TRAF3 with its downstream effectors. Notably, the C-terminal TAB2/3-binding domain of TAK1 with which TRAF3 interacted was also essential for auto-phosphorylation and activation of TAK1 by TAB2 and TAB3.28 Therefore, we cannot eliminate the possibility that TAB2 or TAB3 may be involved in TRAF3-dependent TAK1 activation. Nevertheless, our data unambiguously showed that TRAF3 with a mutant TRAF domain failed to activate TAK1 and the downstream ischemic signaling cascades, including the NF-κB, Rac-1, and JNK/c-Jun signaling pathways. Thus, the TRAF3 TIR domain-containing adaptor protein inducing IFNβ–TAK1 interaction is essential for neuronal loss and adverse stroke outcomes.
Previous studies indicated that TRAF3 is crucial for B-cell death through activation of the alternative NF-κB pathway.7 In addition, 19% of multiple myeloma patients and 17% of multiple myeloma cell lines have TRAF3 mutations.30 Furthermore, TRAF3 may promote additional prodeath pathways, as the hyper-survival of TRAF3−/− B cells is not entirely NF-κB–dependent.6 Here, we showed that TRAF3–TAK1 axis was necessary for I/R-induced activation of the alternative JNK prodeath pathway. Inhibition of either TRAF3 or TAK1 led to diminished JNK phosphorylation and activation of downstream effectors. Similarly, although TRAF3 was first recognized as a negative regulator of JNK, recent data have demonstrated that TRAF3 could activate JNK under certain conditions. For example, the replacement of several ring finger domains of TRAF3 with those of TRAF5 converts TRAF3 into an activator of JNK. In addition, TRAF3 may lead to JNK activation in a TIR domain-containing adaptor protein inducing IFNβ–dependent manner.5 However, TRAF3 overexpression led to robust JNK phosphorylation, probably as a result of differences in cell types, stimuli, and experimental models. Notably, we and others have shown that JNK expression could be induced as early as 3-hour poststroke and was robustly upregulated at 6 hours.13 A similar expression pattern was observed for TRAF3 during MCAO/reperfusion, suggesting that JNK is a bona fide target of TRAF3–TAK1. Accordingly, TRAF3-knockout mice exhibited a similar beneficial phenotype, as observed in mice treated with JNK inhibitor.10
Until recently, TRAF3 was thought to be involved mainly in the regulation of the alternative, but not the canonical, NF-κB pathway; However, Perez de Diego et al5,8 recently showed that R118W-mutant TRAF3 was responsible for the impaired TLR-induced nuclear translocation of the NF-κB p65 subunit in–transformed human fibroblasts. More recently, Chen et al31 reported that myeloid TRAF3 promotes metabolic inflammation by switching from an anti-inflammatory factor to a proinflammatory one. The underlying mechanism and functional significance of TRAF3-mediated canonical NF-κB signaling, however, remain largely unknown. Our data describe a novel molecular mechanism in which TRAF3 phosphorylated I kappa B kinase, IκBα, and p65 via activation of TAK1. Activation of TAK1-mediated NF-κB signaling has been more intensively studied in the immune system, although Neubert et al4 recently showed that pharmacological inhibition of TAK1 prevented NF-κB activation in cerebral ischemia.3 NF-κB is a central mediator of inflammatory responses during ischemic stroke. Several proinflammatory genes, including cyclooxygenase 2, inducible nitric oxide synthase, TNF, IL-1α and β, IL-6, and matrix metalloproteinases, are induced on NF-κB activation,32 which exacerbates tissue damage. Consistent with our results, pharmacological and genetic inhibition of NF-κB reduced infarct size.32
Notably, our study also identified a novel substrate for TAK1 after I/R. Rac-1, a small GTPase, is critical for activation of NADPH oxidase. The involvement of Rac-1/NADPH pathway in ischemic stroke is well documented. In addition to the known biological functions in other neurological disorders, such as cognitive impairment, subarachnoid hemorrhage, and neuronal oxidative damage, Rac-1 is also responsible for superoxide generation in cardiovascular ischemic diseases, including myocardial infarction.24,33 Interestingly, Matsumoto-Ida et al34 showed that TAK1 can also be activated by ischemia in the heart, suggesting potential significance of the TRAF3–TAK1–Rac-1 signaling pathway in other organs, such as the heart, liver, and kidney, in ischemia-induced diseases.
Notably, cerebral tissue is composed of several cell types, including neurons, glia, astrocytes, and vascular elements. We cannot eliminate that possibility that the influence of TRAF3 on stroke origins from synergistic effects on multiple cells. However, we observed that neuron-specific TRAF3-knockout mice displayed ≈89% reduction in cerebral TRAF3 expression level (Figure S2C), indicating that neurons are the main target of TRAF3 in brain tissue. This may also explain why TRAF3 conditional knockout or overexpression in neurons exhibited profound histological and behavioral outcomes during ischemic stroke.
Our present work reveals a previously unidentified central regulator in ischemic injury. I/R-induced neuronal expression of TRAF3, which bound to TAK1 and thereby potentiated TAK1-dependent activation of the NF-κB, Rac-1 and JNK/c-Jun signaling pathways. The combined effects of these ischemic signaling cascades resulted in profound neuronal loss and aggravated stroke outcomes. In addition, inhibition of either TRAF3 or TAK1 was neuroprotective in an experimental model of acute stroke. Given that TRAF3 is evolutionarily conserved between humans and mice, TRAF3 gene therapy in mice might, therefore, represent a relatively useful animal model for human medicine. Thus, these findings indicate that the TRAF3–TAK1 complex may represent a promising therapeutic target for stroke management.
Sources of Funding
This work was supported by grants from the National Science Fund for Distinguished Young Scholars (No. 81425005), the National Natural Science Foundation of China (No. 81170086 and No. 81270184), National Science and Technology Support Project (No. 2011BAI15B02, No. 2012BAI39B05, No. 2013YQ030923-05, 2014BAI02B01, and 2015BAI08B01), the Key Project of the National Natural Science Foundation (No. 81330005), the National Basic Research Program China (No. 2011CB503902), and Natural Science Foundation of Hubei Province (2013CFB259).
The online-only Data Supplement is available with this article at http://hyper.ahajournals.org/lookup/suppl/doi:10.1161/HYPERTENSIONAHA.115.05430/-/DC1.
- Received March 4, 2015.
- Revision received March 13, 2015.
- Accepted June 5, 2015.
- © 2015 American Heart Association, Inc.
- Hostager BS,
- Bishop GA.
- Jiang X,
- Deng KQ,
- Luo Y,
- Jiang DS,
- Gao L,
- Zhang XF,
- Zhang P,
- Zhao GN,
- Zhu X,
- Li H.
- Huang H,
- Tang QZ,
- Wang AB,
- Chen M,
- Yan L,
- Liu C,
- Jiang H,
- Yang Q,
- Bian ZY,
- Bai X,
- Zhu LH,
- Wang L,
- Li H.
- Wang XA,
- Zhang R,
- She ZG,
- Zhang XF,
- Jiang DS,
- Wang T,
- Gao L,
- Deng W,
- Zhang SM,
- Zhu LH,
- Guo S,
- Chen K,
- Zhang XD,
- Liu DP,
- Li H.
- Broughton BR,
- Reutens DC,
- Sobey CG.
- Vosler PS,
- Chen J.
- Goldmann T,
- Wieghofer P,
- Müller PF,
- Wolf Y,
- Varol D,
- Yona S,
- Brendecke SM,
- Kierdorf K,
- Staszewski O,
- Datta M,
- Luedde T,
- Heikenwalder M,
- Jung S,
- Prinz M.
- Yu Y,
- Ge N,
- Xie M,
- Sun W,
- Burlingame S,
- Pass AK,
- Nuchtern JG,
- Zhang D,
- Fu S,
- Schneider MD,
- Fan J,
- Yang J.
- Chen Z,
- Shen H,
- Sun C,
- Yin L,
- Tang F,
- Zheng P,
- Liu Y,
- Brink R,
- Rui L.
- Matsumoto-Ida M,
- Takimoto Y,
- Aoyama T,
- Akao M,
- Takeda T,
- Kita T.
Novelty and Significance
What Is New?
Tumor necrosis factor receptor–associated factor 3 (TRAF3) is a ubiquitously expressed adaptor protein and processes extensively regulatory capacity on immune responses. However, the impacts of TRAF3 on neurons and ischemic stroke remain largely unknown.
The expression of TRAF3 is significantly upregulated in the brain samples of mice subjected to middle cerebral artery occlusion models.
TRAF3 is an essential promoter of cerebral ischemia/reperfusion-induced brain damage through aggravating cell death, inflammatory response, and oxidative stress.
The transforming growth factor-β–activated kinase 1–dependent activation of JNK, nuclear factor κB, and Rac-1 signalings in response to ischemia/reperfusion injury is required in TRAF3-regulated ischemic stroke.
What Is Relevant?
Applying gain- and loss-of-function approaches, the role of TRAF3 specifically expressed in neuron cells was established, and TRAF3 was identified as a mediator of ischemic injury.
The influence of TRAF3 on cerebral ischemia/reperfusion injury was demonstrated resulting from its powerful influence on cellular behaviors, including death, inflammation, and oxidation.
TRAF3 was found to directly interact with transforming growth factor-β–activated kinase 1 and subsequently lead the activation of c-Jun kinase/c-Jun, nuclear factor κB, and Rac-1 cascades, leading to the exacerbation and acceleration of brain damage induced by ischemia/reperfusion insult.
Our present study demonstrated that TRAF3 functions as a profound regulator of cerebral I/R-induced brain damage deriving from the potently propulsive effects on cell death, inflammatory, and oxidative stress through elevating the activation of JNK, nuclear factor κB, and Rac-1 pathways. The interaction of TRAF3 with transforming growth factor-β–activated kinase 1 is indispensable for TRAF3-mediated aggravation of I/R-induced injury. These observations suggested that targeting TRAF3 or the TRAF3–transforming growth factor-β-activated kinase 1 interaction might represent a promising therapeutic strategy for stroke and other neurological diseases.