XBP1 (X-Box–Binding Protein-1)–Dependent O-GlcNAcylation Is Neuroprotective in Ischemic Stroke in Young Mice and Its Impairment in Aged Mice Is Rescued by Thiamet-G
Background and Purpose—Impaired protein homeostasis induced by endoplasmic reticulum dysfunction is a key feature of a variety of age-related brain diseases including stroke. To restore endoplasmic reticulum function impaired by stress, the unfolded protein response is activated. A key unfolded protein response prosurvival pathway is controlled by the endoplasmic reticulum stress sensor (inositol-requiring enzyme-1), XBP1 (downstream X-box–binding protein-1), and O-GlcNAc (O-linked β-N-acetylglucosamine) modification of proteins (O-GlcNAcylation). Stroke impairs endoplasmic reticulum function, which activates unfolded protein response. The rationale of this study was to explore the potentials of the IRE1/XBP1/O-GlcNAc axis as a target for neuroprotection in ischemic stroke.
Methods—Mice with Xbp1 loss and gain of function in neurons were generated. Stroke was induced by transient or permanent occlusion of the middle cerebral artery in young and aged mice. Thiamet-G was used to increase O-GlcNAcylation.
Results—Deletion of Xbp1 worsened outcome after transient and permanent middle cerebral artery occlusion. After stroke, O-GlcNAcylation was activated in neurons of the stroke penumbra in young mice, which was largely Xbp1 dependent. This activation of O-GlcNAcylation was impaired in aged mice. Pharmacological increase of O-GlcNAcylation before or after stroke improved outcome in both young and aged mice.
Conclusions—Our study indicates a critical role for the IRE1/XBP1 unfolded protein response branch in stroke outcome. O-GlcNAcylation is a prosurvival pathway that is activated in the stroke penumbra in young mice but impaired in aged mice. Boosting prosurvival pathways to counterbalance the age-related decline in the brain’s self-healing capacity could be a promising strategy to improve ischemic stroke outcome in aged brains.
Ischemia activates many pathological processes that have been targeted for neuroprotection in ischemic stroke. Such neuroprotection strategies are based on the assumption that stroke outcome will be improved by blocking these pathological processes because the brain’s self-healing capacity will then be sufficient to restore impaired cellular functions. However, the brain’s capacity to activate prosurvival pathways when challenged by ischemic stress dramatically declines with age.1 Importantly, preclinical neuroprotection studies are performed primarily in young rodents, and thus, it is conceivable that the success in these animals is limited in elderly stroke patients. To improve ischemic stroke outcome in aged brains, we tested a novel strategy for neuroprotection that boosts a prosurvival pathway in aged brains after stroke to restore neurological function impaired by ischemic stress. We considered the unfolded protein response (UPR) as a promising target because UPR is activated to restore endoplasmic reticulum (ER) function that is impaired in a variety of stress conditions including stroke2–9 and because the ability to activate UPR declines with age.10–12
UPR comprises 2 prosurvival pathways that are controlled by stress sensor proteins in the ER membrane: ATF6 (activating transcription factor 6) and IRE1 (inositol-requiring enzyme-1).13 After activation, IRE1 converts to an active endonuclease that cleaves a fragment of 26 bases from the coding region of Xbp1 (X-box binding protein-1) mRNA.14 This cleavage triggers a frameshift and formation of XBP1s, a new protein and transcription factor. XBP1s regulates expression of genes coding for ER chaperons, as well as proteins that are involved in ER-associated degradation and the hexosamine biosynthetic pathway (HBP).14–16 HBP links the IRE1 UPR branch to O-linked β-N-acetylglucosamine modification of proteins (O-GlcNAcylation), which protects cells in a variety of stress conditions.16 Here, we report that the IRE1-regulated signal transduction pathway critically defined acute ischemic stroke outcome, that O-GlcNAcylation was activated in the stroke penumbra in young but not in aged brains, and that boosting O-GlcNAcylation pharmacologically provided neuroprotection in ischemic stroke in young and aged mice.
A detailed Methods section is available in the online-only Data Supplement.
Animal experiments were approved by the Duke University Animal Care and Use Committee. All studies were conducted in accordance with the United States Public Health Service’s Policy on Humane Care and Use of Laboratory Animals. C57BL/6 mice were purchased from The Jackson Laboratory (Maine). Xbp1f/f mice (C57BL/6 background) were kindly provided by Dr Laurie Glimcher (Weill Cornell Medical College).17 To delete Xbp1 in forebrain neurons, we first crossed Xbp1f/f mice with Emx1Cre/Cre mice (JAX stock #005628; C57BL/6 background),8 to generate Xbp1f/+;Emx1-Cre mice. These mice were then mated with homozygous floxed Xbp1f/f mice to generate experimental animals: Xbp1f/f;Emx1-Cre (Xbp1-cKO) and Xbp1f/f (control). Xbp1-cKO mice are healthy and show no overt phenotype. Inducible XBP1s transgenic mice (TRE-XBP1s; C57BL/6 background) were kindly provided by Dr Joseph Hill (Texas Southwest Medical Center).16 In TRE-XBP1s transgenic mice, XBP1s expression is driven by 7 tetracycline-responsive elements (TRE), which can be activated by binding of tTA (transactivator protein). To generate mice with inducible expression of XBP1s in forebrain neurons, we crossed TRE-XBP1s mice with Camk2a-tTA mice expressing the tetracycline-controlled tTA under the control of the neuron-specific Camk2a promoter (Tet-off system; JAX stock #007004; C57BL/6 background), resulting in double transgenic TRE-XBP1s;Camk2a-tTA mice (XBP1s-TG). In XBP1s-TG mice, XBP1s expression is suppressed by adding doxycycline to the drinking water. When doxycycline is removed, XBP1s expression is activated. Primers for genotyping are listed in Table I in the online-only Data Supplement.
Animal surgery was performed on young male mice (2–3 months) and aged male mice (22–24 months). Mouse models of transient middle cerebral artery occlusion (tMCAO) and permanent MCAO (pMCAO) were used. The online tool Quickcalcs (http://www.graphpad.com/quickcalcs/) was used to randomly assign animals to groups. tMCAO was performed as described previously.8 Briefly, anesthesia was induced in mice with 5% isoflurane in 30% O2 balanced with N2O. Mice were then orally intubated and mechanically ventilated with 1.8% isoflurane during surgery. The rectal temperature was maintained at 37±0.2°C using a heating pad and a heating lamp throughout the entire procedure. Transient focal brain ischemia was induced by inserting a 6-0 monofilament (Doccol) into the right internal carotid artery and anterior cerebral artery via the external carotid artery and temporary ligation of the right common carotid. Laser-Doppler flowmetry (Moor) was used to monitor regional cerebral blood flow in the MCA territory during the procedure. After 30-minute MCAO (Xbp1-cKO experiments) or 45-minute MCAO (Thiamet-G experiments), mice were supplemented with 0.5 mL saline and were then placed into a temperature-controlled incubator for 4 hours before returning to their home cages. Animals with brain hemorrhage and those that did not show a reduction in regional cerebral blood flow >80% during MCAO and recovery of regional cerebral blood flow >70% after 5-minute reperfusion were excluded.
In this study, we also used a modified mouse model of pMCAO that is based on our rat pMCAO model.18 Briefly, mice were anesthetized with isoflurane, intubated, and ventilated. The rectal temperature was maintained at 37±0.2°C throughout the procedure. Mice were placed in left lateral position, and a small skin incision was made between the eye and ear. The lowest part of the temporal muscle was cut slightly using a high temperature loop, and a 3-mm segment of zygomatic arch was removed. After exposing the skull base and trigeminal nerve branch, a small window (1–2 mm2) was drilled on the skull above the MCA. The MCA trunk was lifted with an 8-0 needle and permanently ligated with silk suture proximal to the cortical branch to the rhinal cortex. The muscle and skin were then closed separately. Animals that did not show an infarct lesion were excluded. Animals excluded for analysis are listed in Table II in the online-only Data Supplement.
For pretreatment experiments and post-treatment experiments, mice were dosed with thiamet-G 18 hours before tMCAO or 30 minutes after onset of pMCAO.
Neurological Deficit and Infarct Volume
A 48-point scoring system was used to evaluate neurological deficits at 24 hours after tMCAO and 3 days after pMCAO. Infarct volume was measured using the 2,3,5-triphenyltetrazolium chloride–staining method. All evaluations were performed by observers who were blinded to genotype and group assignment.
Immunofluorescence Staining and Microscopy
Immunofluorescence staining was performed on frozen tissues as described previously.19 ImageJ software (NIH) was used to analyze fluorescence intensity of O-GlcNAcylation.
Reverse-Transcription Polymerase Chain Reaction, Quantitative PCR, and Western Blotting
Reverse-transcription polymerase chain reaction, quantitative PCR, and Western blot analysis were performed as previously described.1
All data analyses were performed with Prism 6 (GraphPad Software). The primary outcome for stroke experiments was infarct volume, which was used to determine the group size for each experiment based on our previous studies or pilot experiments. Statistical analysis was assessed by unpaired Student t test (infarct volumes, mRNA and protein levels, and fluorescence intensities) or Mann–Whitney U test (neurological scores). Data are presented as mean±SEM, mean±SD, or the median (neurological score). The individual data of infarct volume and neurological score are also shown in the figures. The level of significance was set at P<0.05.
Deletion of Xbp1 in Forebrain Neurons Worsens Stroke Outcome
To achieve deletion of Xbp1 predominantly in forebrain neurons, we crossed conditional Xbp1f/f mice with Emx1-Cre mice to generate Xbp1f/f;Emx1-Cre (Xbp1-cKO) mice.8 In Xbp1f/f mice, exon 2 of the Xbp1 gene is floxed. When deleted by Cre, a frameshift occurs, and the mRNA is translated into a nonfunctional short XBP1 fragment. Reverse-transcription PCR indicated that exon 2 of Xbp1 was deleted in the Xbp1-cKO mouse brain (Figure 1A), which was further confirmed by quantitative PCR analysis using an Xbp1 exon 2–specific primer (Figure I in the online-only Data Supplement). We then measured the diameter of cerebral arteries in Xbp1-cKO and Xbp1f/f littermates (control) and compared laser-Doppler measurements of regional cerebral blood flow between control and Xbp1-cKO mice during ischemia and early reperfusion. Our data demonstrated that Xbp1-cKO mice have normal cerebrovascular anatomy and cerebral blood flow response to MCAO (Figure II in the online-only Data Supplement).
Next, we confirmed that splicing of Xbp1 mRNA was activated in our MCAO model (Figure III in the online-only Data Supplement), which is consistent with previous reports.3,8 Then, to clarify the role of the IRE1/XBP1 UPR branch in stroke outcome, we subjected Xbp1-cKO and control (Xbp1f/f) littermate mice to 2 stroke models, tMCAO and pMCAO. Deletion of Xbp1 in forebrain neurons resulted in worse stroke outcome in both stroke models (Figure 1B and 1C); compared with control mice, Xbp1-cKO mice exhibited larger infarct volumes after tMCAO (control: 29.23±6.21 versus Xbp1-cKO: 64.24±7.02 mm3) and pMCAO (control: 35.19±4.18 versus Xbp1-cKO: 47.91±3.19 mm3). This suggests that activation of the IRE1/XBP1 UPR branch is neuroprotective in ischemic stroke.
XBP1s Activates O-GlcNAc Modification in Neurons
Recent data demonstrate that XBP1s activates expression of genes coding for enzymes of the hexosamine biosynthetic pathway (HBP; Figure 2A).16 This enhances flux through the HBP to generate uridine diphosphate N-acetylglucosamine (UDP-GlcNAc), the substrate for O-GlcNAcylation (Figure 2A). O-GlcNAcylation is a post-translational modification that protects cells from many stress conditions.16,20 The O-GlcNAcylation is controlled by O-GlcNAc transferase, which adds O-GlcNAc to target proteins, and O-GlcNAcylase, which removes O-GlcNAc from modified proteins (Figure 2A).
To determine whether the XBP1/HBP/O-GlcNAc axis is functional in neurons, as it is in cardiomyocytes,16 we used TRE-XBP1s;Camk2a-tTA (XBP1s-TG) mice in which neuron-specific expression of transgene XBP1s is controlled using the Tet-off system. After 8 days on regular water to activate XBP1s expression, XBP1s-TG exhibited strong XBP1s immunostaining signals in forebrain regions, which were colocalized with neuronal marker neuronal nuclear antigen staining (Figure IV in the online-only Data Supplement).
To determine whether overexpression of XBP1s in neurons upregulates enzymes of HBP, thereby increasing levels of O-GlcNAcylated proteins, we first verified by quantitative PCR that mRNA levels of Xbp1s were markedly increased in brains of XBP1s-TG mice and that XBP1s upregulated expression of P58IPK, an XBP1s-dependent gene (Figure 2B).15 We also confirmed an increase in mRNA abundance of glutamine:fructose-6-phosphate aminotransferase (GFAT1), the rate-limiting enzyme of HBP, glucosamine-phosphate N-acetyltransferase 1 (Gnpnat1), and UDP-glucose 4-epimerase (GaIE) in brains of XBP1s-TG mice (Figure 2B). Furthermore, Western blot analysis demonstrated the increased protein level of GFAT1 (Figure 2C). Finally, we showed that O-GlcNAc modification of proteins was increased in brains of XBP1s-TG mice (Figure 2D). Together, our data indicate that the XBP1/HBP/O-GlcNAc axis is functional in neurons.
Stroke Activates O-GlcNAcylation in the Ischemic Penumbra in an Xbp1-Dependent Fashion
Because the IRE1/XBP1 UPR branch is activated in experimental stroke (Figure III in the online-only Data Supplement),3,8 we hypothesized that stroke activates O-GlcNAcylation. To test this hypothesis, we exposed Xbp1-cKO and control (Xbp1f/f) littermate mice to tMCAO and evaluated changes in O-GlcNAcylation using immunofluorescence (Figure 3). In brains of control mice, O-GlcNAcylation was indeed activated in neurons located in the penumbra, the brain region adjacent to the ischemic core that showed low MAP2 staining (Figure 3A and 3C). In contrast, in brains of Xbp1-cKO mice, activation of O-GlcNAcylation in neurons of the stroke penumbra was markedly less pronounced than that in control mice, indicating severe impairment of this pathway (Figure 3B and 3D). These findings demonstrate that stroke activates O-GlcNAcylation in the penumbra and that this activation is predominantly Xbp1 dependent.
Increased O-GlcNAcylation Improves Stroke Outcome in Young Mice
Results from in vitro studies show that under conditions of long-lasting ER stress, UPR triggers apoptosis that can be inhibited when the IRE1 UPR branch is artificially sustained.21 Furthermore, XBP1s and O-GlcNAc are regulators of stress resistance,10,22 and O-GlcNAcylation attenuates ER stress–induced cell death.23 We, therefore, hypothesized that activation of IRE1/XBP1/HBP/O-GlcNAcylation axis in the penumbra of ischemic stroke is a neuroprotective stress response and that boosting O-GlcNAcylation improves stroke outcome. To test this hypothesis, we administered thiamet-G, a potent and highly specific O-GlcNAcylase inhibitor (Figure 2A). Notably, thiamet-G has been proven to cross the blood–brain barrier, trigger a marked increase in brain levels of O-GlcNAc modified proteins, and produce no obvious adverse effects in mice dosed for many weeks.24 We first determined the optimal route of thiamet-G administration and found that brain levels of O-GlcNAc–modified proteins were increased to a similar extent by IP injection or IV injection of thiamet-G in mice (Figure V in the online-only Data Supplement). To determine the optimal dose for this study, we injected mice with 10 to 100 mg/kg thiamet-G IP and analyzed brain levels of O-GlcNAc–modified proteins 9 hours later. Mice treated with thiamet-G exhibited a robust increase in O-GlcNAc–modified proteins, and a dose of 30 mg/kg was sufficient to produce this effect (Figure 4A). Thus, we chose this dose for all of our experiments. Next, we evaluated the time-dependent effect of thiamet-G on O-GlcNAcylation in the brain. O-GlcNAc–modified protein levels were increased at 24 hours post-injection of thiamet-G, and levels peaked at ≈9 to 12 hours (Figure 4B). To determine whether increased O-GlcNAcylation improves stroke outcome, we used 2 experimental paradigms: pretreatment with thiamet-G at 18 hours before tMCAO (Figure 4C) and post-treatment with thiamet-G at 30 minutes after onset of pMCAO (Figure 4D). Infarct volumes were significantly reduced in mice treated with thiamet-G in both tMCAO (vehicle: 87.14±4.19 mm3 versus thiamet-G: 58.93±5.71 mm3) and pMCAO (vehicle: 41.24±5.09 mm3 versus thiamet-G: 22.59±4.65 mm3). These findings indicate that increased O-GlcNAcylation is neuroprotective in ischemic stroke.
Increased O-GlcNAcylation Improves Stroke Outcome in Aged Mice
Age is a key risk factor for poor outcome after ischemic stroke. Recently, we reported that transient forebrain ischemia activates O-GlcNAcylation in young but not in aged mice.1 This suggests that the inability of the brain to activate O-GlcNAcylation on ischemic stress could be a critical factor in postischemic recovery of neurological function at advanced age. To determine whether this earlier finding in transient forebrain ischemia is also valid in ischemic stroke, we subjected young and aged mice to pMCAO and analyzed O-GlcNAcylation by immunostaining (Figure 5A and 5B). O-GlcNAcylation was activated in neurons located in the penumbra adjacent to the ischemic core in young mice, but this activation was impaired in aged mice (Figure 5A and 5B). This suggests that aging is associated with a loss in the brain’s capacity to activate O-GlcNAcylation under conditions associated with impaired ER function.
To test whether we can pharmacologically boost O-GlcNAcylation in aged brains, we dosed young and aged mice with thiamet-G and analyzed brain levels of O-GlcNAcylated proteins at 6 hours post-injection. The increase in levels of O-GlcNAcylated proteins was similar in brains of young and aged mice (Figure 5C). This indicates that the O-GlcNAcylation pathway controlled by O-GlcNAc transferase and O-GlcNAcylase is functional in aged brains. We, therefore, hypothesized that thiamet-G is a promising tool to rescue the impaired activation of O-GlcNAcylation in ischemic brains of aged mice and thereby improve stroke outcome. To test this hypothesis, we subjected aged mice to pMCAO and dosed the animals with thiamet-G or saline 30 minutes after ischemia onset. On day 3 post-surgery, we found that neurological scores were significantly improved, and infarct volumes were significantly reduced in thiamet-G–treated mice (Figure 5D; vehicle: 48.96±4.56 mm3 versus thiamet-G: 35.05±4.78 mm3). Together, these data provide convincing evidence that aging is associated with impaired activation of O-GlcNAcylation in the stroke penumbra and that boosting O-GlcNAcylation pharmacologically improves stroke outcome in aged mice.
Here, we report results from an experimental study designed to evaluate the potential of the IRE1/XBP1/HBP/O-GlcNAc axis as a target for neuroprotection in ischemic stroke. Using genetic and pharmacological approaches, we provide evidence that the IRE1/XBP1 UPR pathway critically defines ischemic stroke outcome. Indeed, mice in which Xbp1 was deleted in forebrain neurons had worse stroke outcome (Figure 1). Furthermore, our results suggest that the IRE1/XBP1/HBP/O-GlcNAc axis is activated in stroke. Transgenic mice expressing XBP1s in forebrain neurons showed increased levels of GFAT1, the rate-limiting enzyme of HBP, and increased O-GlcNAcylation. In ischemic stroke, O-GlcNAcylation was activated in an Xbp1-dependent fashion in neurons of the penumbra in young brains (Figures 2 and 3) but was markedly impaired in aged brains, and boosting O-GlcNAcylation pharmacologically significantly improved stroke outcome in young and aged brains (Figures 4 and 5). Together, this suggests that O-GlcNAcylation is a key component of the prosurvival function of the IRE1 UPR branch. Recently, it was reported that thiamet-G improved acute stroke outcome in young mice by suppressing ischemia-induced inflammation.25 Results presented here suggest that the IRE1/XBP1 UPR branch plays a major role in postischemic activation of O-GlcNAcylation in the brain. This implies that restoration of protein homeostasis (proteostasis) is an important component of thiamet-G–induced neuroprotection in stroke.
Boosting O-GlcNAcylation provided strong neuroprotection even when stroke was induced 18 hours after dosing mice with thiamet-G (Figure 4C). This long time window of protection against ischemia-induced brain damage could be of interest for the clinical setting. Notably, various major pediatric and adult cardiovascular surgeries involve cardiopulmonary bypass procedures that require a period of cardiac arrest. To protect the brain and other organs from ischemic damage, surgery is usually performed under moderate-to-deep hypothermic conditions.26,27 However, deep hypothermia is associated with adverse effects and requires extra time to cool down and rewarm patient. Thus, a pharmacological approach with similar neuroprotection properties would be of major interest.
Importantly, the current study revealed a novel approach to neuroprotection in acute ischemic stroke that takes into consideration, for the first time, the age-dependent decline in the brain’s capacity to respond to ischemic stress. Age is a key risk factor for a variety of brain pathologies that are also associated with ER dysfunction, including stroke, traumatic brain injury, and degenerative diseases, suggesting stress-induced impaired proteostasis. Indeed, in the aged brains, stroke induces more pronounced oxidative protein damage but with decreased activation of prosurvival pathways compared with the young brain.28 Furthermore, aging is associated with impaired capacity to restore proteostasis after stress.11,29 When challenged by ischemic stress, activation of stress response pathways, including ubiquitin conjugation and O-GlcNAcylation is impaired,1 suggesting reduced capacity of aged brains to clear unfolded/misfolded proteins and to restore proteostasis.
Our knowledge of the critical role of UPR to ensure proteostasis is predominantly based on results derived from model systems including cell culture and Caenorhabditis elegans (C. elegans). UPR has 3 response branches that are activated when proteostasis is disturbed. The PERK (protein kinase RNA-like ER kinase) branch reduces the load of newly synthesized proteins that must be folded, and the ATF6 and IRE1 branches activate genetic programs to restore ER function. However, if ER stress persists, UPR triggers apoptosis (PERK branch). This shift from prosurvival to proapoptosis is blocked when activation of the IRE1 UPR branch is experimentally prolonged.21 Notably, impaired ER proteostasis associated with age can be reversed by neuronal XBP1s expression,10 and stress resistance is improved by a genetic program involving the interaction of O-GlcNAc with promoters of genes that regulate stress response pathways.22 Importantly, neuron-specific XBP1s expression induces a global stress resistance that is not restricted to neurons.10 Together, these observations and our results reported here point to a critical role for the IRE1-regulated signal transduction pathway in restoring proteostasis in stressed cells of the stroke penumbra and supports a notion that the inability of aged brains to activate O-GlcNAcylation is a pivotal factor in stroke outcome.
We show here that the IRE1/XBP1/HBP/O-GlcNAc axis is functional in neurons and that stroke activates O-GlcNAcylation in young but not in aged brains. Previously, we have shown that stroke activates the IRE1 UPR pathway, as indicated by the splicing of Xbp1 mRNA,3,8 and we have provided evidence that age does not play a major role in the extent to which Xbp1 splicing is activated by ischemic stress.1 The observation that thiamet-G increased O-GlcNAcylation to a similar extent in young and aged mice suggests that the flux through the HBP and the O-GlcNAcylation reaction controlled by O-GlcNAc transferase and O-GlcNAcylase are not impaired in aged brains. Therefore, the inability of aged brains to activate O-GlcNAcylation in the penumbra of ischemic stroke may be the result of impaired protein synthesis. Indeed, translation is required for the formation of the XBP1s protein from spliced Xbp1 mRNA and for expression of HBP enzymes, which is activated by XBP1s. In contrast, thiamet-G does not require protein synthesis because the thiamet-G–induced increase in O-GlcNAcylation is triggered by blocking the O-GlcNAcylase that removes O-GlcNAc from proteins, not by activating flux through the HBP. Notably, protein synthesis is sensitive to even a moderate reduction in blood flow,30 and acute stroke triggers spreading depression waves that are associated with a prolonged and more pronounced drop in perfusion in aged than in young animals.31 These findings are important steps toward defining the mechanisms underlying the impaired activation of O-GlcNAcylation in aged brains exposed to ischemic stress.
The results presented here are a first proof of concept for a novel strategy to increase neuroprotection in acute ischemic stroke by boosting a prosurvival pathway that is impaired in aged brains. This strategy is expected to be most effective when boosting a pathway that is activated in young but not in aged brains and that is required to restore function impaired by ischemic stress. In future work, we will further evaluate the potentials of boosting O-GlcNAcylation pharmacologically to improve stroke outcome, and we will also consider other aspects, such as the therapeutic window, sex of animals, and long-term outcome, and include a second species. To further develop this novel approach, we propose a global search for prosurvival pathways that are activated in the penumbra in young but not in aged brains, to define the role of identified pathways in stroke outcome, and to then develop strategies to boost those pathways pharmacologically. Finally, a comprehensive approach to neuroprotection should be considered that combines the novel neuroprotection strategy described here with the traditional strategies that interfere with pathological processes.
We thank Pei Miao for her excellent technical support and Kathy Gage for her excellent editorial contribution. Paschen and Yang were responsible for conception and design of the study. Dr Jiang, Dr S. Yu, Dr Z. Yu, Dr Sheng, Dr Li, Dr Liu, Dr Warner, and Dr Yang were responsible for acquisition and analysis of data. Dr Jiang, Dr Paschen, and Dr Yang were responsible for drafting a significant portion of the manuscript and figures.
Sources of Funding
This study was supported by American Heart Association grant 16GRNT30270003 (to Dr Yang) and NIH R01 grants NS099590 (to Dr Yang) and NS097554 and NS081299 (to Dr Paschen).
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.117.016579/-/DC1.
- Received January 1, 2017.
- Revision received March 9, 2017.
- Accepted March 27, 2017.
- © 2017 American Heart Association, Inc.
- Liu S,
- Sheng H,
- Yu Z,
- Paschen W,
- Yang W
- Dickhout JG,
- Carlisle RE,
- Austin RC
- Martindale JJ,
- Fernandez R,
- Thuerauf D,
- Whittaker R,
- Gude N,
- Sussman MA,
- et al
- Yu Z,
- Sheng H,
- Liu S,
- Zhao S,
- Glembotski CC,
- Warner DS,
- et al
- Taylor RC
- Llorente IL,
- Burgin TC,
- Pérez-Rodríguez D,
- Martínez-Villayandre B,
- Pérez-García CC,
- Fernández-López A
- Lee AH,
- Iwakoshi NN,
- Glimcher LH
- Yang W,
- Sheng H,
- Thompson JW,
- Zhao S,
- Wang L,
- Miao P,
- et al
- Lin JH,
- Li H,
- Yasumura D,
- Cohen HR,
- Zhang C,
- Panning B,
- et al
- Love DC,
- Ghosh S,
- Mondoux MA,
- Fukushige T,
- Wang P,
- Wilson MA,
- et al
- Ngoh GA,
- Hamid T,
- Prabhu SD,
- Jones SP
- He Y,
- Ma X,
- Li D,
- Hao J
- Camboni D,
- Philipp A,
- Schebesch KM,
- Schmid C
- Takami Y,
- Tajima K,
- Terazawa S,
- Okada N,
- Fujii K,
- Sakai Y
- Ben-Zvi A,
- Miller EA,
- Morimoto RI
- Xie Y,
- Mies G,
- Hossmann KA