- Email: [email protected]
Neuronal damage and functional deficits are ameliorated by inhibition of aquaporin and HIF1α after traumatic brain injury (TBI)
Journal of the Neurological Sciences 323 (2012) 134–140
Contents lists available at SciVerse ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Neuronal damage and functional deficits are ameliorated by inhibition of aquaporin and HIF1α after traumatic brain injury (TBI) Mohammed Shenaq a, c, Hassan Kassem a, Changya Peng a, Steven Schafer b, Jamie Y. Ding d, Vance Fredrickson a, Murali Guthikonda a, Christian W. Kreipke b, José A. Rafols b, Yuchuan Ding a,⁎ a
Department of Neurological Surgery, Wayne State University School of Medicine and College of Engineering, Detroit, MI, United States Department of Anatomy and Cell Biology, Wayne State University School of Medicine and College of Engineering, Detroit, MI, United States Department of Biomedical Engineering, Wayne State University School of Medicine and College of Engineering, Detroit, MI, United States d Butler College, Princeton University, Princeton, NJ, United States b c
a r t i c l e
i n f o
Article history: Received 1 March 2012 Received in revised form 27 July 2012 Accepted 31 August 2012 Available online 3 October 2012 Keywords: Marmarou's TBI model Cerebral edema Water channels Cellular injury FluoroJade Functional deficits
a b s t r a c t The present study, using a rodent model of closed-head diffuse traumatic brain injury (TBI), investigated the role of dysregulated aquaporins (AQP) 4 and 9, as well as hypoxia inducible factor −1α(HIF-1α) on brain edema formation, neuronal injury, and functional deficits. TBI was induced in adult (400–425 g), male Sprague–Dawley rats using a modified Marmarou's head impact-acceleration device (450 g weight dropped from 2 m height). Animals in each treatment group were administered intravenous anti-AQP4 or -AQP9 antibodies or 2-Methoxyestradiol (2ME2, an inhibitor of HIF-1α) 30 min after injury. At 24 h post-TBI, animals (n = 6 each group) were sacrificed to examine the extent of brain edema by water content, as well as protein expression of AQP and HIF-1α by Western immune-blotting. At 48-hours post-TBI, neuronal injury (n = 8 each group) was assessed by FluoroJade (FJ) histochemistry. Spatial learning and memory deficits were evaluated by radial arm maze (n=8 each group) up to 21 days post-TBI. Compared to non-injured controls, significant (pb 0.05) increases in the expression of AQP4 and −9 were detected in the brains of injured animals. In addition, significant (pb 0.05) brain edema after TBI was associated with increases (p b 0.05) both in neuronal injury (FJ labeling) and neurobehavioral deficits. Selective inhibition of either AQP4 or −9, or HIF-1α significantly (pb 0.05) decreased the expression of the proteins. In addition, inhibition of the AQPs and HIF-1α significantly (pb 0.05) ameliorated brain edema, as well as the number of injured neurons in cortical layers II/III and V/VI, striatum and hippocampal regions CA1/CA3. Finally, compared to the non-treated TBI animals, AQP or HIF-1α inhibition significantly (pb 0.01) improved neurobehavioral outcomes after TBI. Taken together, the present data supports a causal relation between HIF-AQP mediated cerebral edema, secondary neuronal injury, and tertiary behavioral deficits post-TBI. The data further suggests that upstream modulation of the molecular patho-trajectory effectively ameliorates both neuronal injury and behavioral deficits post-TBI. © 2012 Published by Elsevier B.V.
1. Introduction Traumatic brain injury (TBI) is a major cause of disability and death in the first four decades of life, accounting for 15% of deaths in the 15–45 year age group with a high prevalence of long-term disability in those who survive [1]. Diffuse TBI, resulting from head concussions, accounts for many of the TBI's seen in the same age group. Cerebral edema, a common pathology in TBI, develops within hours of impact and remains a critical problem in the emergency room. Edema brings about an elevated intracranial pressure (ICP) and may lead to brain ⁎ Corresponding author at: Department of Neurological Surgery, Wayne State University, School of Medicine, 550 E Canfield, Detroit, MI 48201, United States. Tel.: + 1 313 577 0038; fax: + 1 313 993 9269. E-mail address: [email protected] (Y. Ding). 0022-510X/$ – see front matter © 2012 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.jns.2012.08.036
herniation, which increases morbidity and carries a poor prognosis among survivors [2–4]. Edema develops as a consequence of breakdown of the blood brain barrier (BBB), and is thought to be mediated by specific water channels in certain cell membranes called aquaporins (AQPs) [5]. AQPs are a family of water channels embedded in the plasma membrane of astrocytes and endothelial cells, and they are responsible for the flow of water in and out of these cells [6,7]. AQPs were first discovered in 1992 by a team of investigators led by Dr. Peter Agre [7]. Subsequent research revealed a dozen more AQP family members in addition to those initially found [6]. In brain tissue, AQP4 and 9 have been shown to be localized to ependymal cells, retinal Muller cells, foot processes of astrocytes, as well as luminal and abluminal membranes of endothelium; with the last two cell types forming integral components of the blood brain barrier (BBB) [8]. Recent TBI and ischemia research has shown
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
that brain edema is associated specifically with AQP4 and 9, and these channels are thought to be responsible for tissue regulation of homeostasis and osmolarity [6,7]. However, although dysfunctional AQP4 and 9 have been thought to be related to edema formation following ischemia or brain trauma, treatments targeting these channels in an attempt to improve neurological outcome have not been forthcoming. Post TBI, upstream events within the molecular patho-trajectory leading to edema appear to precede dysregulation of the AQPs and the resulting failure of the BBB. Thus, TBI is thought to up-regulate hypoxia inducible factor-1 (HIF-1), a transcription factor that responds to hypoxic conditions. This event leads to increased expression of both AQP4 and 9 [9]. In abnormally hypoxic conditions such as brain ischemia and TBI, HIF-1α, a short-lived (5 min) highly oxygen sensitive subunit of HIF-1, is prevented from being hydroxylated and degraded, resulting in higher cellular levels [10]. In turn, this results in activation of the HIF-1 complex with up-regulation of certain target genes, inducing the transcription of mRNA for certain proteins such as AQPs [10]. However, whether upstream inhibition of HIF-1α may lead to a reduction in neuronal damage and improvement in functional outcome, and whether the same outcomes might be causally associated with reduced expression of AQPs and mitigation of edema have not been investigated to date. Here, using a rodent model of diffuse TBI, we address whether AQP inhibition shortly after trauma causes a reduction in brain edema and neuronal damage in the cerebral cortex, striatum and hippocampus, as well as an improvement in functional outcomes. We further address whether upstream HIF-1α inhibition might ultimately impact both neuro-histologic and behavioral outcomes via direct modulation of AQP4 and 9 expressions. 2. Materials and methods 2.1. Marmarou's rat acceleration-impact model Adult (400–425 g), male Sprague–Dawley rats (Charles River, USA) were used. Animals were divided into a control (sham-operated) group and four TBI groups.TBI groups included: 1) non-treatment, or treatment with: 2) anti-AQP4 antibody; 3) anti-AQP9 antibody; or 4) 2-Methoxyestradiol (2ME2, an inhibitor of HIF-1α). A modified Marmarou's rat acceleration-impact model was used to induce diffuse TBI [11]. Unlike other TBI models that directly impact the cerebral cortex (e.g., fluid percussion and cortical impact), the Marmarou model is a closed-head TBI model that induces significant cerebral edema and increased intracranial pressure [12]. This model is more representative of concussive-type TBIs, which rarely involve penetration of the brain [11]. Briefly, the animals were anesthetized with halothane and were placed prone on a thick, foam-covered platform. A 450 g weight was first aligned with the surface of a 10 mm diameter stainless steel helmet, which was directly attached to the skull between bregma and lambda sutures, and then dropped directly onto the helmet from a height of 2 meters. Placement of the helmet ensures both that the skull would not be fractured by the weight impact and that the brain would not be directly contused. Control (sham-operated) animals were anesthetized and had the helmet attached to the skull, but they did not receive the weight drop. All animal experimental procedures were approved by the Institutional Animal Investigation Committee of Wayne State University and were in accordance with National Institutes of Health Guidelines for Care and Use of Laboratory Animals. 2.2. Inhibition of AQP4, AQP9 and HIF-1α Animals from appropriate TBI groups were given intravenous anti-AQP4 IgG (1 μg/kg, Sigma Aldrich, USA, Cat# A5971), anti-AQP9 IgG (1 μg/kg, Lifespan Biosciences, USA, Cat# LS-C20770) [5,13] or 2-Methoxyestradiol (2ME2, 2.5 mg/kg, Sigma Aldrich, USA, Cat# M6383-50MG) dissolved in 0.25 ml of isotonic saline 30 minutes
135
after injury. 2ME2 is a naturally occurring metabolite of estradiol known to post-transcriptionally down-regulate the expression of HIF-1α [10]. Animals from the non-treatment TBI group were administered intravenous 0.25 ml of isotonic saline. For both brain edema (n = 6 per group) and protein analyses (n= 6 per group), animals were sacrificed at 24 h after injury. At 48 h post-TBI, neuronal injury (n = 8 each group) was assessed by FluoroJade (FJ) histochemistry. For behavior tests, animals were allowed to survive up to 21 days following injury. 2.3. Assessment of brain edema Edema was measured directly by assessing water content in brains. Briefly, each brain in its full form was harvested and immediately weighed after extraction to determine its wet weight before being dried in an 80 °C oven for 72 hours, after which it was weighed again to obtain the dry weight. The formula ((wet weight - dry weight)/wet weight × 100) was used to calculate the water content and expressed as a percentage of the wet weight [14]. 2.4. Expression of AQP4, AQP9 and HIF-1α Western blot analysis was used to investigate the protein expression of AQP4, AQP9, and HIF-1α [5]. Samples containing both cerebral hemispheres were processed in a lysis buffer which included protease inhibitors on ice. Equal volumes (10 ul) of tissue extracts, normalized by protein concentration, were mixed with a sodium dodecyl sulfate (SDS) sample buffer. The samples were separated by electrophoresis through a 10% polyacrylamide gel (Bio-Rad, Hercules,CA) and then transferred to a nitrocellulose membrane (Bio-Rad, Hercules, CA). Supernatants were used as whole tissue lysates and the protein concentration was determined using the Bradford assay (Bio-Rad, CA, USA). Three different primary antibodies were used: 1) polyclonal rabbit anti-AQP4 (1:500, Santa Cruz, Cat# sc-20812), 2) polyclonal anti-AQP9 (1:2000, Santa Cruz, Cat# sc-28623) and 3) polyclonal rabbit anti-HIF-1α (1:2000, Santa Cruz, Cat# sc-10790). Protein equal loading was confirmed by the intracellular protein β-actin (goat polyclonal anti-β-actin antibody, dilution 1:1,000, Santa Cruz, Cat# sc-1616). These antibodies were incubated with the membrane at 25°C for 1 hour. After three wash cycles, the membrane was then incubated with a secondary antibody conjugated to horseradish peroxidase (Sigma) for 30 min. Finally, the targeted antigens were visualized using the standard chemical luminescence methods (ECL, Amersham Pharmacia Biotech). In quantifying the relative levels of target protein expression, blot images were analyzed using an image analysis program (Image J 1.42, National Institutes of Health, USA) and the expression intensity of the proteins from different groups was statistically compared. 2.5. FluoroJade (FJ) staining and stereological analysis At 48 h post-TBI, animals underwent intra-cardiac perfusion with saline (0.9% NaCl in dH2O) and 4% paraformaldehyde, and the brain was subsequently extracted. FJ histochemistry was performed as originally described [11,15]. Briefly, 40 μm coronal sections containing the sensorimotor cortex (smCx) and dorsal hippocampus (hipp), as well as striatum, were washed in 80% ethanol with 1% NaOH, 70% ethanol in distilled water. Next, tissue sections were washed in 0.06% KMnO4 for 10 min and then washed again in distilled water. The FJ solution (2 ml) was diluted in 48 ml 0.1% acetic acid and the tissue was incubated for 30 min. Sections were then rinsed in distilled water, dehydrated, air dried and mounted with a water-based mounting medium on poly-L-Lysine coated slides. Six sections were chosen from each brain for analysis: two anterior (containing the rostral striatum/smCx), two middle (including the rostral dorsal hipp), and two posterior (representing the caudal dorsal hipp). Labeled FJ neurons within
136
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
randomly selected areas were counted using stereological procedures with a computer-assisted microscope and an image analysis system (AxioVision 4.5TM, Zeiss) [16]. Briefly, an unbiased counting frame (0.18 ×0.1 mm2) digitally superimposed 20 non-overlapping areas (at 400×) in the II/III layers of smCx, the CA1/CA3 regions of the hippocampus, and the dorsolateral striatum. The numbers of labeled cells crossing points on the grid, as well as the total number of points (up to 96) falling within a sample, were counted. The number of FJ-labeled neurons per unit area was estimated by the ratio of the number of points falling upon cells to the total number of points in the sample area. 2.6. Spatial behavior testing in the eight-arm radial arm maze As previously described by us, [17] animals were first placed on the central platform of the maze and allowed to move about freely in order to become familiar with their surroundings (n = 8 each group). Animals were maintained on a regimen of two food pellets per day to encourage maze foraging. During behavioral testing, the maze was enclosed within four black linen walls. A white paper triangle (15 cm sides) was placed on one linen wall 10 cm above the base of radial arm No. 3. A white paper square with bisecting black lines was placed on the same linen wall 10 cm above the base of radial arm No. 5. The researcher was positioned in front of radial arm No. 8 during the study. The three visual cues provided spatial guidance as to the location of the baited arms (n = 4, alternating with unbaited arms). Animals were tested for the time latency taken to find the bait (half of a Fruit Loop Cereal) placed in a plastic cup. Each rat was tested daily for three consecutive trials from day one through day 21, post-TBI. The maximum time a rat was allowed to spend in the maze was 10 min. Averages of these trials were calculated and recorded. 2.7. Statistical analysis All the data was described as mean ± SE. Statistical analysis was performed with SPSS for Windows (version 13.0, SPSS, Inc.). Differences among multiple groups were assessed using one-way analysis of variance (ANOVA) with a significance level at p b 0.05. Post-hoc comparisons between groups were detected using the least significant difference (LSD) method. 3. Results 3.1. Brain Edema Brain edema was determined as a mean percentage change of water content of brain tissue. ANOVA analysis indicated that as compared to the sham-operated control (no TBI, no treatment), TBI significantly (F(4, 25) = 2.993, p b 0.05) caused brain edema with an increase in water content at 24 h (Fig. 1). Furthermore, inhibition of AQP4 or 9 caused a significant (p b 0.05) reduction in brain water content post-TBI. In addition, TBI animals treated with 2ME2 (the HIF-1α inhibitor) also exhibited a significant (pb 0.05) reduction in brain water content. The data suggests that all three treatments have beneficial effects on brain edema formation. 3.2. AQP and HIF-1αprotein expression The effect of AQP inhibition with antibodies on protein expression was measured by Western immunoblotting in the following groups: 1) control, 2) TBI, 3) TBI with AQP4 antibody injection and 4) TBI with AQP9 antibody injection (Fig. 2). In our previous study, [5] we found a temporal co-relation between brain edema and up-regulation of AQPs. Here, we selectively used either AQP4 or 9 antibodies to block the function of AQP4 or AQP9, respectively. We observed a significant (F(3, 20) = 7.01, p b 0.01) 87% decrease in AQP4 expression and an 84% decrease (F(3, 20) = 13.126, p b 0.01) in AQP 9 expression, using their respective
Fig. 1. Graph displaying the brain water content following various treatments post-TBI. TBI significantly (p b 0.05) increases water content in the brain parenchyma. Inhibitions of AQP4, AQP9, and HIF-1α significantly (pb 0.05) reduce brain edema.
antibodies at 24 h post-TBI. Post-hoc analysis further revealed a significant (pb 0.01) reduction of 63% in AQP4 with AQP9 antibody treatment; and a reduction of 66% in AQP9 with AQP4 antibody treatment. In order to confirm the presence of the antibodies after systemic administration, we used secondary antibodies and immunocytochemistry in brain slices from animals that had received TBI. We found that antibodies were expressed both in the wall of blood vessels and adjacent perivascular regions (Fig. 3). The biochemical and histological data taken from animal groups treated with antibody administration suggests the following: 1) AQP antibodies cross the BBB, 2) each antibody binds to its respective target antigens in brain, 3) receptor/ligand antibody interaction reduces AQP expression, and 4) a synergistic interaction of the receptors may exist, as selective blocking of either AQP with its antibody blunts the expression of the other. Although we did not perform co-staining of anti-Aquaporin antibodies with markers for endothelial cells and astrocytes, expression of AQPs on vessels was based on morphology. In our future study, we will further clarify the binding sites of antibodies in astrocytes and endothelial cells. The effect of HIF-1α inhibition on its expression was measured in sham-operated control (no TBI, no treatment), TBI (without 2ME2 treatment), and TBI with 2ME2 groups (Fig. 4). As compared to the control group (1.0 unit of relative protein expression of HIF1α), post-TBI animals exhibited a significant, 100% increase in HIF-1α (F(2, 15) = 11.065, pb 0.01) at 24 h. In addition, 2ME2 administration significantly reduced the post-TBI over-expression of HIF-1α to 15% of its normal level. This suggests effective inhibition of TBI-induced HIF-1α up-regulation by 2ME2.
3.3. Neuron cell injury and quantification In another set of animals, neuronal injury after TBI was assessed by FluoroJade (FJ) labeling at 48 h post injury. FJ has been used as a marker of cell membrane injury which may not necessarily lead to cell death [11]. Compared to sham controls (Fig. 5A), an increase in the number of FJ-labeled neuronal perikarya were detected in the superficial layers (II-III) of the sensorimotor cortex (smCx), dorsal hippocampus (hipp), and striatum (Fig. 5B). In all regions, the numbers of FJ labeled cells were significantly reduced by inhibition of AQP4 (Fig. 5C), AQP 9 (Fig. 5D), or HIF-1α (Fig. 5E). Compared to the TBI animals, stereological analyses in layers II/III of the injured smCx (Fig. 6A) (F(3, 28) = 40.204, p b 0.01), CA1/CA3 of dorsal hipp (Fig. 6B) (F(3, 28) = 17.2598, p b 0.01), or striatum (Fig. 6C) (F(3, 28) = 17.699, p b 0.01) demonstrated a significant reduction in neuronal injury after inhibition of either AQP4, or AQP9, orHIF-1α. Post-hoc statistical analysis did not further demonstrate significant differences in neuronal injury among the different treatments.
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
137
Fig. 2. A: AQP4 expression is greatest in TBI with no treatment, indicating a cascading up-regulation effect after TBI. Treatment with AQP4 antibody demonstrates a significant (p b 0.05) decrease in AQP-4 expression. B: Similar to AQP4, AQP9 expression was significantly (p b 0.05) decreased following the administration of the aquaporin-9 antibody.
3.4. Behavior testing
4. Discussion
Cognitive behavioral analyses were assessed using standard measures consisting of time latency in completing the radial arm maze. These functional outcomes provided insight on the efficacy of the treatments. After TBI, animals consistently took nine or more minutes (upper limit set at ten minutes, as described in Methods) to complete the radial arm maze task, and did so with a significantly higher level of errors (data not shown), this level of performance was maintained for a duration of nine days (Fig. 7A). By 21 days, the average latency time remained above four minutes. Animals with inhibition of AQP4 post-TBI showed a reduction in task completion times by the end of behavioral testing. The average latency time for completion of the maze task on day one was 8.7 min. By the 21st day post-TBI, the average latency time was 1.5 min. The anti-AQP 9 treatment group showed similar results. The average latency time on day one was 9.0 min and by day 21 was 2.2 min. The group treated with 2ME2 had an average latency time of 8.2 min on day one, and 1.8 min by day 21. ANOVA indicated that TBI animals showed significantly (F(3, 26) =69.3 p b 0.01) poorer performance as compared to any treatment animal group. The data further showed a reduction in task time post-TBI, which is evidence of improved behavioral outcome using the described methods of treatment.
The present study found that a concurrently elevated expression of HIF-1α, AQP4 and AQP9 in brain tissue post-TBI may induce brain edema formation associated with neuronal injury and functional deficits. Edema, as determined by brain water content, was reduced by selective inhibition of HIF-1α, AQP4 or AQP9. Our data therefore supports the concept that cerebral edema after TBI is mediated through the AQPs and that edema formation is part of a molecular signaling cascade involving both HIF-1α and the AQPs. Inhibition of HIF-1α, AQP4 or AQP9 decreased brain edema. In addition, this inhibition ameliorated neuronal damage as determined by FJ histochemistry, and improved functional outcomes as determined by the cognitive and motor behavior tests. Taken together, our data suggests that the molecular cascade involving HIF-1α and the AQPs is part of a pathotrajectory that leads to secondary cell injury and ultimately to neurological dysfunction post-TBI. Brain edema is one of the major factors contributing to cell damage post-TBI [7,18]. It is widely known that edema causes a rise in intracranial pressure, which contributes significantly to neuronal cell damage, [19] as well as increases in morbidity and mortality of TBI patients [2–4]. The present study provides insight into the development of brain edema from dysfunctional AQPs via an HIF mediated mechanism which could lead both to neuronal damage and functional deficits following TBI. While the role of AQPs in TBI-induced cerebral edema has been previously reported, [5,6,19–21] AQP modulation to ameliorate neuronal injury and functional outcome has not been thoroughly investigated. Here, we have first attempted to elucidate whether inhibition of AQPs might be effective in reducing brain edema post-TBI. As observed previously in a stroke model, [13] the present TBI data supports the idea of AQP4 and 9 channels being synergistically coupled, as blocking one channel blunts the performance of the other. In addition, the changes in AQP expression post-TBI appear to be causally related to the amount of cell damage, as the numbers of FJ-labeled cells increased in parallel with the increased expression of AQP 4 and 9. Again, these effects were blunted following the administration of AQP antibodies. Conversely, reduction in AQP expression after TBI led to reduced neuronal damage and improved functional outcome. Therefore BBB dysfunction and AQP dysregulation appear to directly impact both on secondary cell injury and neurologic outcome post-TBI. A correlation between neuronal damage, as revealed by FJ staining, and the extent of edema was also revealed by this study. FJ is an anionic fluorochrome that diffuses through the cell membrane of damaged neurons [22]. Our stereological analyses showed that TBI-induced neuronal
Fig. 3. Graph showing the expression of HIF-1α following TBI. HIF-1α was significantly (p b 0.05) increased in expression following TBI. 2ME2 administration suppressed the expression of HIF-1α.
138
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
A
B
C
Fig. 4. Fluorescence microscopy photomicrographs showing Alexa Fluo647 staining of brain tissue from TBI rats (A) and TBI rats administered anti-AQP4 (B) or anti-AQP9 antibodies (C). The increased fluorescent labeling in the parenchyma around blood vessels suggest that systemic administration of AQP antibodies cross the BBB, bind to their target along endothelial cells and adjacent astrocyte end feet.
damage in the cerebral cortex, dorsal hippocampus and striatum could be ameliorated following the inhibition of AQPs or HIF-1α. It is also known that TBI often leads to both focal and diffuse cell damage, [23] leading to physical, cognitive, and motor impairment. Neuronal damage involves a disruption in the structural integrity or the normal physiological and biochemical processes of the cell membrane [6,24]. Brain edema is essentially a massive influx and accumulation of water within the brain parenchyma, and can result from trauma, stroke or hydrocephalus, as well as systemic infections such as childhood malaria and meningitis [7,19]. Immediately following TBI, inflammatory factors are activated and the influx of water begins, resulting in a build-up of water in cellular compartments of the brain, including the cerebral cortex [7,19,20,25]. Concurrently, there is an osmotic imbalance resulting in an abnormal flow of water into the brain [7,19,20]. The accumulation of water causes increased intracranial pressure which presses on the meninges and brain matter itself. As a result of the rigid bone encasement, compression of the brain ensues [7,19,20,25]. This compression affects not only the neurons but also astrocytes, the main cell type that swells during edema. The pathological effects of edema on astrocytes further impair their ability to protect and support the neurons [7,20]. In addition to cell injury, behavioral performance is highly impacted by TBI [26,27]. In this study, behavioral outcome was correlated with the amount of cell damage after TBI. Radial arm maze measurements
have been routinely used to assess cognitive function [17,28,29]. The present data confirms that the functional performance of TBI animals was impaired, and further demonstrated an improved functional outcome after inhibition of AQP4, 9, and HIF-1α. This suggests a neuroprotection by blocking any part of HIF-AQP cascade. The HIF-1α cascade has been implicated in the formation of brain edema. Thus, increased HIF-1α expression induces the expression of vascular endothelial growth factor (VEGF) which can in turn lead to increased permeability of blood vessels [30]. In addition, recent studies on hypoxic rats have demonstrated that HIF-1α up-regulation is associated with an increase of brain AQP4 [30]. Studies also show that the up-regulation of HIF-1α and its subsequent BBB disruption may be linked to an up-regulation of AQP4, [13] and brain edema [9]. HIF-1α up-regulation leading to over expression of AQP4 and 9 after closed head TBI, was previously reported by us [21]. The present study has revealed that inhibition of HIF-1α with 2ME2 ameliorates TBI-induced neuronal damage. Similarly, 2ME2 treatment reduced AQP expression in association with reduced brain edema. Our previous studies have demonstrated this, and in specific that 2ME2 reduces the expression of both AQP4 and AQP9 [5,21]. We found that after administration of 2ME2 to TBI rats, mRNA and protein expression of AQP4 and AQP9 was decreased. These results provide a therapeutic strategy for the resolution of edema, reduction of neuronal injury and improved behavior outcome, if administered shortly after TBI.
Fig. 5. A-E: Levels II/III of cortex; A′–E′: CA1/CA3 of the Hippocampus; A″–E″: Striatum. Immunohistochemical images of various sections of the brain using Fluoro Jade markers. Fluorescent labeling demonstrates neuronal injury. In the cortex, hippocampus and striatum, inhibition of AQP4, AQP9 and HIF-1α all demonstrated decreased fluorescent cells as compared to TBI without treatment.
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
139
Raidal Arm Maze 12 Aqp4 average Aqp9 average
10
2ME2average ShamAVERAGE
Latency
8
TBIAVERAGE
6 4 2 0 1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20
Days Fig. 7. Graph showing time latency of rats in the radial arm maze. TBI animals showed significantly (p b 0.01) poorer performance than treatment groups. Inhibition of AQP4 and 9 demonstrated nearly the same significantly (p b 0.05) lower latency time over the 21 days. In addition, 2ME2 demonstrated a similar neuroprotective effect to those of AQP inhibition groups.
edema and neuronal damage as well, in association with AQP reduction, leading to functional recovery. Conflict of interest There is no conflict of interest. Acknowledgments This work was supported partially by the WSU Neurosurgery Fund (YD), NIH NS 39860 (JAR), and NIH NS 064976, VA RRD (CWK). References
Fig. 6. Graphs showing FJ labeled cells in each area of the brain. The neuronal damage post TBI was significantly (*p b 0.05) higher than the control. Inhibition of AQP4, AQP9 and HIF-1αall significantly (p b 0.05) contributed to the reduction in neuronal cell damage in the cortex (II/III, V/VI), the hippocampus (CA1/CA3), and the striatum.
In addition, a possible interaction may exist between AQP4 and 9, lending support to the possible benefit of a combined treatment of anti-AQP4 antibody, anti-AQP9 antibody, and 2ME2. These beneficial effects will be investigated in our future studies and we will prove the benefit of the treatment combination. In summary, TBI and TBI-induced brain edema remain one of the world's major causes of death and disability. This study demonstrates that reduction in TBI-induced AQP expression ameliorates cerebral edema and neuronal damage. HIF-1α inhibition blunts the TBI-induced
[1] Belli A, Sen J, Petzold A, Russo S, Kitchen N, Smith M. Metabolic failure precedes intracranial pressure rises in traumatic brain injury: a microdialysis study. Acta Neurochir 2008;150(5):461-9 [discussion 70]. [2] Feldmann H, Klages G, Gartner F, Scharfenberg J. The prognostic value of intracranial pressure monitoring after severe head injuries. Acta Neurochir Suppl 1979;28(1):74-7. [3] Sarabia R, Lobato RD, Rivas JJ, Cordobes F, Rubio J, Cabrera A, et al. Cerebral hemisphere swelling in severe head injury patients. Acta Neurochir Suppl 1988;42:40-6. [4] Marmarou A. A review of progress in understanding the pathophysiology and treatment of brain edema. Neurosurg Focus 2007;22(5):E1. [5] Higashida T, Kreipke CW, Rafols JA, Peng C, Schafer S, Schafer P, et al. The role of hypoxia-inducible factor-1alpha, aquaporin-4, and matrix metalloproteinase-9 in blood–brain barrier disruption and brain edema after traumatic brain injury. J Neurosurg 2011;114(1):92–101. [6] Badaut J, Lasbennes F, Magistretti PJ, Regli L. Aquaporins in brain: distribution, physiology, and pathophysiology. J Cereb Blood Flow Metab 2002;22(4):367-78. [7] Papadopoulos MC, Verkman AS. Aquaporin-4 and brain edema. Pediatr Nephrol 2007;22(6):778-84. [8] Sun MC, Honey CR, Berk C, Wong NL, Tsui JK. Regulation of aquaporin-4 in a traumatic brain injury model in rats. J Neurosurg 2003;98(3):565-9. [9] Kaur C, Sivakumar V, Zhang Y, Ling EA. Hypoxia-induced astrocytic reaction and increased vascular permeability in the rat cerebellum. Glia 2006;54(8):826-39. [10] Lundby C, Calbet JA, Robach P. The response of human skeletal muscle tissue to hypoxia. Cell Mol Life Sci 2009;66(22):3615-23. [11] Rafols JA, Morgan R, Kallakuri S, Kreipke CW. Extent of nerve cell injury in Marmarou's model compared to other brain trauma models. Neurol Res 2007;29(4):348-55. [12] Marmarou A, Foda MA, van den Brink W, Campbell J, Kita H, Demetriadou K. A new model of diffuse brain injury in rats. Part I: Pathophysiology and biomechanics. J Neurosurg 1994;80(2):291-300. [13] Higashida T, Peng C, Li J, Dornbos III D, Teng K, Li X, et al. Hypoxia-inducible factor-1alpha contributes to brain edema after stroke by regulating aquaporins and glycerol distribution in brain. Curr Neurovasc Res 2011;8(1):44-51. [14] Kimelberg HK, Macvicar BA, Sontheimer H. Anion channels in astrocytes: biophysics, pharmacology, and function. Glia 2006;54(7):747-57. [15] Schmued LC, Albertson C, Slikker Jr W. Fluoro-Jade: a novel fluorochrome for the sensitive and reliable histochemical localization of neuronal degeneration. Brain Res 1997;751(1):37-46.
140
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
[16] Swain RA, Harris AB, Wiener EC, Dutka MV, Morris HD, Theien BE, et al. Prolonged exercise induces angiogenesis and increases cerebral blood volume in primary motor cortex of the rat. Neuroscience 2003;117(4):1037-46. [17] Kreipke CW, Morgan R, Kallakuri S, Rafols JA. Behavioral pre-conditioning enhances angiogenesis and cognitive outcome after brain trauma. Neurol Res 2007;29(4): 388-94. [18] Guo Q, Sayeed I, Baronne LM, Hoffman SW, Guennoun R, Stein DG. Progesterone administration modulates AQP4 expression and edema after traumatic brain injury in male rats. Exp Neurol 2006;198(2):469-78. [19] Griesdale DE, Honey CR. Aquaporins and brain edema. Surg Neurol 2004;61(5): 418-21. [20] Papadopoulos MC, Krishna S, Verkman AS. Aquaporin water channels and brain edema. Mt Sinai J Med 2002;69(4):242-8. [21] Ding JY, Kreipke CW, Speirs SL, Schafer P, Schafer S, Rafols JA. Hypoxia-inducible factor-1alpha signaling in aquaporin upregulation after traumatic brain injury. Neurosci Lett 2009;453(1):68-72. [22] Poirier JL, Capek R, De Koninck Y. Differential progression of Dark Neuron and Fluoro-Jade labelling in the rat hippocampus following pilocarpine-induced status epilepticus. Neuroscience 2000;97(1):59-68. [23] Kawai N, Maeda Y, Kudomi N, Yamamoto Y, Nishiyama Y, Tamiya T. Focal neuronal damage in patients with neuropsychological impairment after diffuse traumatic brain injury: evaluation using (1)(1)C-flumazenil positron emission tomography with statistical image analysis. J Neurotrauma 2010;27(12):2131-8.
[24] Loane DJ, Faden AI. Neuroprotection for traumatic brain injury: translational challenges and emerging therapeutic strategies. Trends Pharmacol Sci 2010;31(12): 596-604. [25] Agre P, Nielsen S, Ottersen OP. Towards a molecular understanding of water homeostasis in the brain. Neuroscience 2004;129(4):849-50. [26] Cernak I, Chang T, Ahmed FA, Cruz MI, Vink R, Stoica B, et al. Pathophysiological response to experimental diffuse brain trauma differs as a function of developmental age. Dev Neurosci 2010;32(5–6):442-53. [27] Ding Y, Yao B, Lai Q, McAllister JP. Impaired motor learning and diffuse axonal damage in motor and visual systems of the rat following traumatic brain injury. Neurol Res 2001;23(2–3):193-202. [28] Timofeeva OA, Eddins D, Yakel JL, Blackshear PJ, Levin ED. Hippocampal infusions of MARCKS peptides impair memory of rats on the radial-arm maze. Brain Res 2010;1308:147-52. [29] Langdon KD, Granter-Button S, Corbett D. Persistent behavioral impairments and neuroinflammation following global ischemia in the rat. Eur J Neurosci 2008;28(11): 2310-8. [30] Witt KA, Mark KS, Huber J, Davis TP. Hypoxia-inducible factor and nuclear factor kappa-B activation in blood–brain barrier endothelium under hypoxic/reoxygenation stress. J Neurochem 2005;92(1):203-14.
Contents lists available at SciVerse ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Neuronal damage and functional deficits are ameliorated by inhibition of aquaporin and HIF1α after traumatic brain injury (TBI) Mohammed Shenaq a, c, Hassan Kassem a, Changya Peng a, Steven Schafer b, Jamie Y. Ding d, Vance Fredrickson a, Murali Guthikonda a, Christian W. Kreipke b, José A. Rafols b, Yuchuan Ding a,⁎ a
Department of Neurological Surgery, Wayne State University School of Medicine and College of Engineering, Detroit, MI, United States Department of Anatomy and Cell Biology, Wayne State University School of Medicine and College of Engineering, Detroit, MI, United States Department of Biomedical Engineering, Wayne State University School of Medicine and College of Engineering, Detroit, MI, United States d Butler College, Princeton University, Princeton, NJ, United States b c
a r t i c l e
i n f o
Article history: Received 1 March 2012 Received in revised form 27 July 2012 Accepted 31 August 2012 Available online 3 October 2012 Keywords: Marmarou's TBI model Cerebral edema Water channels Cellular injury FluoroJade Functional deficits
a b s t r a c t The present study, using a rodent model of closed-head diffuse traumatic brain injury (TBI), investigated the role of dysregulated aquaporins (AQP) 4 and 9, as well as hypoxia inducible factor −1α(HIF-1α) on brain edema formation, neuronal injury, and functional deficits. TBI was induced in adult (400–425 g), male Sprague–Dawley rats using a modified Marmarou's head impact-acceleration device (450 g weight dropped from 2 m height). Animals in each treatment group were administered intravenous anti-AQP4 or -AQP9 antibodies or 2-Methoxyestradiol (2ME2, an inhibitor of HIF-1α) 30 min after injury. At 24 h post-TBI, animals (n = 6 each group) were sacrificed to examine the extent of brain edema by water content, as well as protein expression of AQP and HIF-1α by Western immune-blotting. At 48-hours post-TBI, neuronal injury (n = 8 each group) was assessed by FluoroJade (FJ) histochemistry. Spatial learning and memory deficits were evaluated by radial arm maze (n=8 each group) up to 21 days post-TBI. Compared to non-injured controls, significant (pb 0.05) increases in the expression of AQP4 and −9 were detected in the brains of injured animals. In addition, significant (pb 0.05) brain edema after TBI was associated with increases (p b 0.05) both in neuronal injury (FJ labeling) and neurobehavioral deficits. Selective inhibition of either AQP4 or −9, or HIF-1α significantly (pb 0.05) decreased the expression of the proteins. In addition, inhibition of the AQPs and HIF-1α significantly (pb 0.05) ameliorated brain edema, as well as the number of injured neurons in cortical layers II/III and V/VI, striatum and hippocampal regions CA1/CA3. Finally, compared to the non-treated TBI animals, AQP or HIF-1α inhibition significantly (pb 0.01) improved neurobehavioral outcomes after TBI. Taken together, the present data supports a causal relation between HIF-AQP mediated cerebral edema, secondary neuronal injury, and tertiary behavioral deficits post-TBI. The data further suggests that upstream modulation of the molecular patho-trajectory effectively ameliorates both neuronal injury and behavioral deficits post-TBI. © 2012 Published by Elsevier B.V.
1. Introduction Traumatic brain injury (TBI) is a major cause of disability and death in the first four decades of life, accounting for 15% of deaths in the 15–45 year age group with a high prevalence of long-term disability in those who survive [1]. Diffuse TBI, resulting from head concussions, accounts for many of the TBI's seen in the same age group. Cerebral edema, a common pathology in TBI, develops within hours of impact and remains a critical problem in the emergency room. Edema brings about an elevated intracranial pressure (ICP) and may lead to brain ⁎ Corresponding author at: Department of Neurological Surgery, Wayne State University, School of Medicine, 550 E Canfield, Detroit, MI 48201, United States. Tel.: + 1 313 577 0038; fax: + 1 313 993 9269. E-mail address: [email protected] (Y. Ding). 0022-510X/$ – see front matter © 2012 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.jns.2012.08.036
herniation, which increases morbidity and carries a poor prognosis among survivors [2–4]. Edema develops as a consequence of breakdown of the blood brain barrier (BBB), and is thought to be mediated by specific water channels in certain cell membranes called aquaporins (AQPs) [5]. AQPs are a family of water channels embedded in the plasma membrane of astrocytes and endothelial cells, and they are responsible for the flow of water in and out of these cells [6,7]. AQPs were first discovered in 1992 by a team of investigators led by Dr. Peter Agre [7]. Subsequent research revealed a dozen more AQP family members in addition to those initially found [6]. In brain tissue, AQP4 and 9 have been shown to be localized to ependymal cells, retinal Muller cells, foot processes of astrocytes, as well as luminal and abluminal membranes of endothelium; with the last two cell types forming integral components of the blood brain barrier (BBB) [8]. Recent TBI and ischemia research has shown
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
that brain edema is associated specifically with AQP4 and 9, and these channels are thought to be responsible for tissue regulation of homeostasis and osmolarity [6,7]. However, although dysfunctional AQP4 and 9 have been thought to be related to edema formation following ischemia or brain trauma, treatments targeting these channels in an attempt to improve neurological outcome have not been forthcoming. Post TBI, upstream events within the molecular patho-trajectory leading to edema appear to precede dysregulation of the AQPs and the resulting failure of the BBB. Thus, TBI is thought to up-regulate hypoxia inducible factor-1 (HIF-1), a transcription factor that responds to hypoxic conditions. This event leads to increased expression of both AQP4 and 9 [9]. In abnormally hypoxic conditions such as brain ischemia and TBI, HIF-1α, a short-lived (5 min) highly oxygen sensitive subunit of HIF-1, is prevented from being hydroxylated and degraded, resulting in higher cellular levels [10]. In turn, this results in activation of the HIF-1 complex with up-regulation of certain target genes, inducing the transcription of mRNA for certain proteins such as AQPs [10]. However, whether upstream inhibition of HIF-1α may lead to a reduction in neuronal damage and improvement in functional outcome, and whether the same outcomes might be causally associated with reduced expression of AQPs and mitigation of edema have not been investigated to date. Here, using a rodent model of diffuse TBI, we address whether AQP inhibition shortly after trauma causes a reduction in brain edema and neuronal damage in the cerebral cortex, striatum and hippocampus, as well as an improvement in functional outcomes. We further address whether upstream HIF-1α inhibition might ultimately impact both neuro-histologic and behavioral outcomes via direct modulation of AQP4 and 9 expressions. 2. Materials and methods 2.1. Marmarou's rat acceleration-impact model Adult (400–425 g), male Sprague–Dawley rats (Charles River, USA) were used. Animals were divided into a control (sham-operated) group and four TBI groups.TBI groups included: 1) non-treatment, or treatment with: 2) anti-AQP4 antibody; 3) anti-AQP9 antibody; or 4) 2-Methoxyestradiol (2ME2, an inhibitor of HIF-1α). A modified Marmarou's rat acceleration-impact model was used to induce diffuse TBI [11]. Unlike other TBI models that directly impact the cerebral cortex (e.g., fluid percussion and cortical impact), the Marmarou model is a closed-head TBI model that induces significant cerebral edema and increased intracranial pressure [12]. This model is more representative of concussive-type TBIs, which rarely involve penetration of the brain [11]. Briefly, the animals were anesthetized with halothane and were placed prone on a thick, foam-covered platform. A 450 g weight was first aligned with the surface of a 10 mm diameter stainless steel helmet, which was directly attached to the skull between bregma and lambda sutures, and then dropped directly onto the helmet from a height of 2 meters. Placement of the helmet ensures both that the skull would not be fractured by the weight impact and that the brain would not be directly contused. Control (sham-operated) animals were anesthetized and had the helmet attached to the skull, but they did not receive the weight drop. All animal experimental procedures were approved by the Institutional Animal Investigation Committee of Wayne State University and were in accordance with National Institutes of Health Guidelines for Care and Use of Laboratory Animals. 2.2. Inhibition of AQP4, AQP9 and HIF-1α Animals from appropriate TBI groups were given intravenous anti-AQP4 IgG (1 μg/kg, Sigma Aldrich, USA, Cat# A5971), anti-AQP9 IgG (1 μg/kg, Lifespan Biosciences, USA, Cat# LS-C20770) [5,13] or 2-Methoxyestradiol (2ME2, 2.5 mg/kg, Sigma Aldrich, USA, Cat# M6383-50MG) dissolved in 0.25 ml of isotonic saline 30 minutes
135
after injury. 2ME2 is a naturally occurring metabolite of estradiol known to post-transcriptionally down-regulate the expression of HIF-1α [10]. Animals from the non-treatment TBI group were administered intravenous 0.25 ml of isotonic saline. For both brain edema (n = 6 per group) and protein analyses (n= 6 per group), animals were sacrificed at 24 h after injury. At 48 h post-TBI, neuronal injury (n = 8 each group) was assessed by FluoroJade (FJ) histochemistry. For behavior tests, animals were allowed to survive up to 21 days following injury. 2.3. Assessment of brain edema Edema was measured directly by assessing water content in brains. Briefly, each brain in its full form was harvested and immediately weighed after extraction to determine its wet weight before being dried in an 80 °C oven for 72 hours, after which it was weighed again to obtain the dry weight. The formula ((wet weight - dry weight)/wet weight × 100) was used to calculate the water content and expressed as a percentage of the wet weight [14]. 2.4. Expression of AQP4, AQP9 and HIF-1α Western blot analysis was used to investigate the protein expression of AQP4, AQP9, and HIF-1α [5]. Samples containing both cerebral hemispheres were processed in a lysis buffer which included protease inhibitors on ice. Equal volumes (10 ul) of tissue extracts, normalized by protein concentration, were mixed with a sodium dodecyl sulfate (SDS) sample buffer. The samples were separated by electrophoresis through a 10% polyacrylamide gel (Bio-Rad, Hercules,CA) and then transferred to a nitrocellulose membrane (Bio-Rad, Hercules, CA). Supernatants were used as whole tissue lysates and the protein concentration was determined using the Bradford assay (Bio-Rad, CA, USA). Three different primary antibodies were used: 1) polyclonal rabbit anti-AQP4 (1:500, Santa Cruz, Cat# sc-20812), 2) polyclonal anti-AQP9 (1:2000, Santa Cruz, Cat# sc-28623) and 3) polyclonal rabbit anti-HIF-1α (1:2000, Santa Cruz, Cat# sc-10790). Protein equal loading was confirmed by the intracellular protein β-actin (goat polyclonal anti-β-actin antibody, dilution 1:1,000, Santa Cruz, Cat# sc-1616). These antibodies were incubated with the membrane at 25°C for 1 hour. After three wash cycles, the membrane was then incubated with a secondary antibody conjugated to horseradish peroxidase (Sigma) for 30 min. Finally, the targeted antigens were visualized using the standard chemical luminescence methods (ECL, Amersham Pharmacia Biotech). In quantifying the relative levels of target protein expression, blot images were analyzed using an image analysis program (Image J 1.42, National Institutes of Health, USA) and the expression intensity of the proteins from different groups was statistically compared. 2.5. FluoroJade (FJ) staining and stereological analysis At 48 h post-TBI, animals underwent intra-cardiac perfusion with saline (0.9% NaCl in dH2O) and 4% paraformaldehyde, and the brain was subsequently extracted. FJ histochemistry was performed as originally described [11,15]. Briefly, 40 μm coronal sections containing the sensorimotor cortex (smCx) and dorsal hippocampus (hipp), as well as striatum, were washed in 80% ethanol with 1% NaOH, 70% ethanol in distilled water. Next, tissue sections were washed in 0.06% KMnO4 for 10 min and then washed again in distilled water. The FJ solution (2 ml) was diluted in 48 ml 0.1% acetic acid and the tissue was incubated for 30 min. Sections were then rinsed in distilled water, dehydrated, air dried and mounted with a water-based mounting medium on poly-L-Lysine coated slides. Six sections were chosen from each brain for analysis: two anterior (containing the rostral striatum/smCx), two middle (including the rostral dorsal hipp), and two posterior (representing the caudal dorsal hipp). Labeled FJ neurons within
136
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
randomly selected areas were counted using stereological procedures with a computer-assisted microscope and an image analysis system (AxioVision 4.5TM, Zeiss) [16]. Briefly, an unbiased counting frame (0.18 ×0.1 mm2) digitally superimposed 20 non-overlapping areas (at 400×) in the II/III layers of smCx, the CA1/CA3 regions of the hippocampus, and the dorsolateral striatum. The numbers of labeled cells crossing points on the grid, as well as the total number of points (up to 96) falling within a sample, were counted. The number of FJ-labeled neurons per unit area was estimated by the ratio of the number of points falling upon cells to the total number of points in the sample area. 2.6. Spatial behavior testing in the eight-arm radial arm maze As previously described by us, [17] animals were first placed on the central platform of the maze and allowed to move about freely in order to become familiar with their surroundings (n = 8 each group). Animals were maintained on a regimen of two food pellets per day to encourage maze foraging. During behavioral testing, the maze was enclosed within four black linen walls. A white paper triangle (15 cm sides) was placed on one linen wall 10 cm above the base of radial arm No. 3. A white paper square with bisecting black lines was placed on the same linen wall 10 cm above the base of radial arm No. 5. The researcher was positioned in front of radial arm No. 8 during the study. The three visual cues provided spatial guidance as to the location of the baited arms (n = 4, alternating with unbaited arms). Animals were tested for the time latency taken to find the bait (half of a Fruit Loop Cereal) placed in a plastic cup. Each rat was tested daily for three consecutive trials from day one through day 21, post-TBI. The maximum time a rat was allowed to spend in the maze was 10 min. Averages of these trials were calculated and recorded. 2.7. Statistical analysis All the data was described as mean ± SE. Statistical analysis was performed with SPSS for Windows (version 13.0, SPSS, Inc.). Differences among multiple groups were assessed using one-way analysis of variance (ANOVA) with a significance level at p b 0.05. Post-hoc comparisons between groups were detected using the least significant difference (LSD) method. 3. Results 3.1. Brain Edema Brain edema was determined as a mean percentage change of water content of brain tissue. ANOVA analysis indicated that as compared to the sham-operated control (no TBI, no treatment), TBI significantly (F(4, 25) = 2.993, p b 0.05) caused brain edema with an increase in water content at 24 h (Fig. 1). Furthermore, inhibition of AQP4 or 9 caused a significant (p b 0.05) reduction in brain water content post-TBI. In addition, TBI animals treated with 2ME2 (the HIF-1α inhibitor) also exhibited a significant (pb 0.05) reduction in brain water content. The data suggests that all three treatments have beneficial effects on brain edema formation. 3.2. AQP and HIF-1αprotein expression The effect of AQP inhibition with antibodies on protein expression was measured by Western immunoblotting in the following groups: 1) control, 2) TBI, 3) TBI with AQP4 antibody injection and 4) TBI with AQP9 antibody injection (Fig. 2). In our previous study, [5] we found a temporal co-relation between brain edema and up-regulation of AQPs. Here, we selectively used either AQP4 or 9 antibodies to block the function of AQP4 or AQP9, respectively. We observed a significant (F(3, 20) = 7.01, p b 0.01) 87% decrease in AQP4 expression and an 84% decrease (F(3, 20) = 13.126, p b 0.01) in AQP 9 expression, using their respective
Fig. 1. Graph displaying the brain water content following various treatments post-TBI. TBI significantly (p b 0.05) increases water content in the brain parenchyma. Inhibitions of AQP4, AQP9, and HIF-1α significantly (pb 0.05) reduce brain edema.
antibodies at 24 h post-TBI. Post-hoc analysis further revealed a significant (pb 0.01) reduction of 63% in AQP4 with AQP9 antibody treatment; and a reduction of 66% in AQP9 with AQP4 antibody treatment. In order to confirm the presence of the antibodies after systemic administration, we used secondary antibodies and immunocytochemistry in brain slices from animals that had received TBI. We found that antibodies were expressed both in the wall of blood vessels and adjacent perivascular regions (Fig. 3). The biochemical and histological data taken from animal groups treated with antibody administration suggests the following: 1) AQP antibodies cross the BBB, 2) each antibody binds to its respective target antigens in brain, 3) receptor/ligand antibody interaction reduces AQP expression, and 4) a synergistic interaction of the receptors may exist, as selective blocking of either AQP with its antibody blunts the expression of the other. Although we did not perform co-staining of anti-Aquaporin antibodies with markers for endothelial cells and astrocytes, expression of AQPs on vessels was based on morphology. In our future study, we will further clarify the binding sites of antibodies in astrocytes and endothelial cells. The effect of HIF-1α inhibition on its expression was measured in sham-operated control (no TBI, no treatment), TBI (without 2ME2 treatment), and TBI with 2ME2 groups (Fig. 4). As compared to the control group (1.0 unit of relative protein expression of HIF1α), post-TBI animals exhibited a significant, 100% increase in HIF-1α (F(2, 15) = 11.065, pb 0.01) at 24 h. In addition, 2ME2 administration significantly reduced the post-TBI over-expression of HIF-1α to 15% of its normal level. This suggests effective inhibition of TBI-induced HIF-1α up-regulation by 2ME2.
3.3. Neuron cell injury and quantification In another set of animals, neuronal injury after TBI was assessed by FluoroJade (FJ) labeling at 48 h post injury. FJ has been used as a marker of cell membrane injury which may not necessarily lead to cell death [11]. Compared to sham controls (Fig. 5A), an increase in the number of FJ-labeled neuronal perikarya were detected in the superficial layers (II-III) of the sensorimotor cortex (smCx), dorsal hippocampus (hipp), and striatum (Fig. 5B). In all regions, the numbers of FJ labeled cells were significantly reduced by inhibition of AQP4 (Fig. 5C), AQP 9 (Fig. 5D), or HIF-1α (Fig. 5E). Compared to the TBI animals, stereological analyses in layers II/III of the injured smCx (Fig. 6A) (F(3, 28) = 40.204, p b 0.01), CA1/CA3 of dorsal hipp (Fig. 6B) (F(3, 28) = 17.2598, p b 0.01), or striatum (Fig. 6C) (F(3, 28) = 17.699, p b 0.01) demonstrated a significant reduction in neuronal injury after inhibition of either AQP4, or AQP9, orHIF-1α. Post-hoc statistical analysis did not further demonstrate significant differences in neuronal injury among the different treatments.
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
137
Fig. 2. A: AQP4 expression is greatest in TBI with no treatment, indicating a cascading up-regulation effect after TBI. Treatment with AQP4 antibody demonstrates a significant (p b 0.05) decrease in AQP-4 expression. B: Similar to AQP4, AQP9 expression was significantly (p b 0.05) decreased following the administration of the aquaporin-9 antibody.
3.4. Behavior testing
4. Discussion
Cognitive behavioral analyses were assessed using standard measures consisting of time latency in completing the radial arm maze. These functional outcomes provided insight on the efficacy of the treatments. After TBI, animals consistently took nine or more minutes (upper limit set at ten minutes, as described in Methods) to complete the radial arm maze task, and did so with a significantly higher level of errors (data not shown), this level of performance was maintained for a duration of nine days (Fig. 7A). By 21 days, the average latency time remained above four minutes. Animals with inhibition of AQP4 post-TBI showed a reduction in task completion times by the end of behavioral testing. The average latency time for completion of the maze task on day one was 8.7 min. By the 21st day post-TBI, the average latency time was 1.5 min. The anti-AQP 9 treatment group showed similar results. The average latency time on day one was 9.0 min and by day 21 was 2.2 min. The group treated with 2ME2 had an average latency time of 8.2 min on day one, and 1.8 min by day 21. ANOVA indicated that TBI animals showed significantly (F(3, 26) =69.3 p b 0.01) poorer performance as compared to any treatment animal group. The data further showed a reduction in task time post-TBI, which is evidence of improved behavioral outcome using the described methods of treatment.
The present study found that a concurrently elevated expression of HIF-1α, AQP4 and AQP9 in brain tissue post-TBI may induce brain edema formation associated with neuronal injury and functional deficits. Edema, as determined by brain water content, was reduced by selective inhibition of HIF-1α, AQP4 or AQP9. Our data therefore supports the concept that cerebral edema after TBI is mediated through the AQPs and that edema formation is part of a molecular signaling cascade involving both HIF-1α and the AQPs. Inhibition of HIF-1α, AQP4 or AQP9 decreased brain edema. In addition, this inhibition ameliorated neuronal damage as determined by FJ histochemistry, and improved functional outcomes as determined by the cognitive and motor behavior tests. Taken together, our data suggests that the molecular cascade involving HIF-1α and the AQPs is part of a pathotrajectory that leads to secondary cell injury and ultimately to neurological dysfunction post-TBI. Brain edema is one of the major factors contributing to cell damage post-TBI [7,18]. It is widely known that edema causes a rise in intracranial pressure, which contributes significantly to neuronal cell damage, [19] as well as increases in morbidity and mortality of TBI patients [2–4]. The present study provides insight into the development of brain edema from dysfunctional AQPs via an HIF mediated mechanism which could lead both to neuronal damage and functional deficits following TBI. While the role of AQPs in TBI-induced cerebral edema has been previously reported, [5,6,19–21] AQP modulation to ameliorate neuronal injury and functional outcome has not been thoroughly investigated. Here, we have first attempted to elucidate whether inhibition of AQPs might be effective in reducing brain edema post-TBI. As observed previously in a stroke model, [13] the present TBI data supports the idea of AQP4 and 9 channels being synergistically coupled, as blocking one channel blunts the performance of the other. In addition, the changes in AQP expression post-TBI appear to be causally related to the amount of cell damage, as the numbers of FJ-labeled cells increased in parallel with the increased expression of AQP 4 and 9. Again, these effects were blunted following the administration of AQP antibodies. Conversely, reduction in AQP expression after TBI led to reduced neuronal damage and improved functional outcome. Therefore BBB dysfunction and AQP dysregulation appear to directly impact both on secondary cell injury and neurologic outcome post-TBI. A correlation between neuronal damage, as revealed by FJ staining, and the extent of edema was also revealed by this study. FJ is an anionic fluorochrome that diffuses through the cell membrane of damaged neurons [22]. Our stereological analyses showed that TBI-induced neuronal
Fig. 3. Graph showing the expression of HIF-1α following TBI. HIF-1α was significantly (p b 0.05) increased in expression following TBI. 2ME2 administration suppressed the expression of HIF-1α.
138
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
A
B
C
Fig. 4. Fluorescence microscopy photomicrographs showing Alexa Fluo647 staining of brain tissue from TBI rats (A) and TBI rats administered anti-AQP4 (B) or anti-AQP9 antibodies (C). The increased fluorescent labeling in the parenchyma around blood vessels suggest that systemic administration of AQP antibodies cross the BBB, bind to their target along endothelial cells and adjacent astrocyte end feet.
damage in the cerebral cortex, dorsal hippocampus and striatum could be ameliorated following the inhibition of AQPs or HIF-1α. It is also known that TBI often leads to both focal and diffuse cell damage, [23] leading to physical, cognitive, and motor impairment. Neuronal damage involves a disruption in the structural integrity or the normal physiological and biochemical processes of the cell membrane [6,24]. Brain edema is essentially a massive influx and accumulation of water within the brain parenchyma, and can result from trauma, stroke or hydrocephalus, as well as systemic infections such as childhood malaria and meningitis [7,19]. Immediately following TBI, inflammatory factors are activated and the influx of water begins, resulting in a build-up of water in cellular compartments of the brain, including the cerebral cortex [7,19,20,25]. Concurrently, there is an osmotic imbalance resulting in an abnormal flow of water into the brain [7,19,20]. The accumulation of water causes increased intracranial pressure which presses on the meninges and brain matter itself. As a result of the rigid bone encasement, compression of the brain ensues [7,19,20,25]. This compression affects not only the neurons but also astrocytes, the main cell type that swells during edema. The pathological effects of edema on astrocytes further impair their ability to protect and support the neurons [7,20]. In addition to cell injury, behavioral performance is highly impacted by TBI [26,27]. In this study, behavioral outcome was correlated with the amount of cell damage after TBI. Radial arm maze measurements
have been routinely used to assess cognitive function [17,28,29]. The present data confirms that the functional performance of TBI animals was impaired, and further demonstrated an improved functional outcome after inhibition of AQP4, 9, and HIF-1α. This suggests a neuroprotection by blocking any part of HIF-AQP cascade. The HIF-1α cascade has been implicated in the formation of brain edema. Thus, increased HIF-1α expression induces the expression of vascular endothelial growth factor (VEGF) which can in turn lead to increased permeability of blood vessels [30]. In addition, recent studies on hypoxic rats have demonstrated that HIF-1α up-regulation is associated with an increase of brain AQP4 [30]. Studies also show that the up-regulation of HIF-1α and its subsequent BBB disruption may be linked to an up-regulation of AQP4, [13] and brain edema [9]. HIF-1α up-regulation leading to over expression of AQP4 and 9 after closed head TBI, was previously reported by us [21]. The present study has revealed that inhibition of HIF-1α with 2ME2 ameliorates TBI-induced neuronal damage. Similarly, 2ME2 treatment reduced AQP expression in association with reduced brain edema. Our previous studies have demonstrated this, and in specific that 2ME2 reduces the expression of both AQP4 and AQP9 [5,21]. We found that after administration of 2ME2 to TBI rats, mRNA and protein expression of AQP4 and AQP9 was decreased. These results provide a therapeutic strategy for the resolution of edema, reduction of neuronal injury and improved behavior outcome, if administered shortly after TBI.
Fig. 5. A-E: Levels II/III of cortex; A′–E′: CA1/CA3 of the Hippocampus; A″–E″: Striatum. Immunohistochemical images of various sections of the brain using Fluoro Jade markers. Fluorescent labeling demonstrates neuronal injury. In the cortex, hippocampus and striatum, inhibition of AQP4, AQP9 and HIF-1α all demonstrated decreased fluorescent cells as compared to TBI without treatment.
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
139
Raidal Arm Maze 12 Aqp4 average Aqp9 average
10
2ME2average ShamAVERAGE
Latency
8
TBIAVERAGE
6 4 2 0 1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20
Days Fig. 7. Graph showing time latency of rats in the radial arm maze. TBI animals showed significantly (p b 0.01) poorer performance than treatment groups. Inhibition of AQP4 and 9 demonstrated nearly the same significantly (p b 0.05) lower latency time over the 21 days. In addition, 2ME2 demonstrated a similar neuroprotective effect to those of AQP inhibition groups.
edema and neuronal damage as well, in association with AQP reduction, leading to functional recovery. Conflict of interest There is no conflict of interest. Acknowledgments This work was supported partially by the WSU Neurosurgery Fund (YD), NIH NS 39860 (JAR), and NIH NS 064976, VA RRD (CWK). References
Fig. 6. Graphs showing FJ labeled cells in each area of the brain. The neuronal damage post TBI was significantly (*p b 0.05) higher than the control. Inhibition of AQP4, AQP9 and HIF-1αall significantly (p b 0.05) contributed to the reduction in neuronal cell damage in the cortex (II/III, V/VI), the hippocampus (CA1/CA3), and the striatum.
In addition, a possible interaction may exist between AQP4 and 9, lending support to the possible benefit of a combined treatment of anti-AQP4 antibody, anti-AQP9 antibody, and 2ME2. These beneficial effects will be investigated in our future studies and we will prove the benefit of the treatment combination. In summary, TBI and TBI-induced brain edema remain one of the world's major causes of death and disability. This study demonstrates that reduction in TBI-induced AQP expression ameliorates cerebral edema and neuronal damage. HIF-1α inhibition blunts the TBI-induced
[1] Belli A, Sen J, Petzold A, Russo S, Kitchen N, Smith M. Metabolic failure precedes intracranial pressure rises in traumatic brain injury: a microdialysis study. Acta Neurochir 2008;150(5):461-9 [discussion 70]. [2] Feldmann H, Klages G, Gartner F, Scharfenberg J. The prognostic value of intracranial pressure monitoring after severe head injuries. Acta Neurochir Suppl 1979;28(1):74-7. [3] Sarabia R, Lobato RD, Rivas JJ, Cordobes F, Rubio J, Cabrera A, et al. Cerebral hemisphere swelling in severe head injury patients. Acta Neurochir Suppl 1988;42:40-6. [4] Marmarou A. A review of progress in understanding the pathophysiology and treatment of brain edema. Neurosurg Focus 2007;22(5):E1. [5] Higashida T, Kreipke CW, Rafols JA, Peng C, Schafer S, Schafer P, et al. The role of hypoxia-inducible factor-1alpha, aquaporin-4, and matrix metalloproteinase-9 in blood–brain barrier disruption and brain edema after traumatic brain injury. J Neurosurg 2011;114(1):92–101. [6] Badaut J, Lasbennes F, Magistretti PJ, Regli L. Aquaporins in brain: distribution, physiology, and pathophysiology. J Cereb Blood Flow Metab 2002;22(4):367-78. [7] Papadopoulos MC, Verkman AS. Aquaporin-4 and brain edema. Pediatr Nephrol 2007;22(6):778-84. [8] Sun MC, Honey CR, Berk C, Wong NL, Tsui JK. Regulation of aquaporin-4 in a traumatic brain injury model in rats. J Neurosurg 2003;98(3):565-9. [9] Kaur C, Sivakumar V, Zhang Y, Ling EA. Hypoxia-induced astrocytic reaction and increased vascular permeability in the rat cerebellum. Glia 2006;54(8):826-39. [10] Lundby C, Calbet JA, Robach P. The response of human skeletal muscle tissue to hypoxia. Cell Mol Life Sci 2009;66(22):3615-23. [11] Rafols JA, Morgan R, Kallakuri S, Kreipke CW. Extent of nerve cell injury in Marmarou's model compared to other brain trauma models. Neurol Res 2007;29(4):348-55. [12] Marmarou A, Foda MA, van den Brink W, Campbell J, Kita H, Demetriadou K. A new model of diffuse brain injury in rats. Part I: Pathophysiology and biomechanics. J Neurosurg 1994;80(2):291-300. [13] Higashida T, Peng C, Li J, Dornbos III D, Teng K, Li X, et al. Hypoxia-inducible factor-1alpha contributes to brain edema after stroke by regulating aquaporins and glycerol distribution in brain. Curr Neurovasc Res 2011;8(1):44-51. [14] Kimelberg HK, Macvicar BA, Sontheimer H. Anion channels in astrocytes: biophysics, pharmacology, and function. Glia 2006;54(7):747-57. [15] Schmued LC, Albertson C, Slikker Jr W. Fluoro-Jade: a novel fluorochrome for the sensitive and reliable histochemical localization of neuronal degeneration. Brain Res 1997;751(1):37-46.
140
M. Shenaq et al. / Journal of the Neurological Sciences 323 (2012) 134–140
[16] Swain RA, Harris AB, Wiener EC, Dutka MV, Morris HD, Theien BE, et al. Prolonged exercise induces angiogenesis and increases cerebral blood volume in primary motor cortex of the rat. Neuroscience 2003;117(4):1037-46. [17] Kreipke CW, Morgan R, Kallakuri S, Rafols JA. Behavioral pre-conditioning enhances angiogenesis and cognitive outcome after brain trauma. Neurol Res 2007;29(4): 388-94. [18] Guo Q, Sayeed I, Baronne LM, Hoffman SW, Guennoun R, Stein DG. Progesterone administration modulates AQP4 expression and edema after traumatic brain injury in male rats. Exp Neurol 2006;198(2):469-78. [19] Griesdale DE, Honey CR. Aquaporins and brain edema. Surg Neurol 2004;61(5): 418-21. [20] Papadopoulos MC, Krishna S, Verkman AS. Aquaporin water channels and brain edema. Mt Sinai J Med 2002;69(4):242-8. [21] Ding JY, Kreipke CW, Speirs SL, Schafer P, Schafer S, Rafols JA. Hypoxia-inducible factor-1alpha signaling in aquaporin upregulation after traumatic brain injury. Neurosci Lett 2009;453(1):68-72. [22] Poirier JL, Capek R, De Koninck Y. Differential progression of Dark Neuron and Fluoro-Jade labelling in the rat hippocampus following pilocarpine-induced status epilepticus. Neuroscience 2000;97(1):59-68. [23] Kawai N, Maeda Y, Kudomi N, Yamamoto Y, Nishiyama Y, Tamiya T. Focal neuronal damage in patients with neuropsychological impairment after diffuse traumatic brain injury: evaluation using (1)(1)C-flumazenil positron emission tomography with statistical image analysis. J Neurotrauma 2010;27(12):2131-8.
[24] Loane DJ, Faden AI. Neuroprotection for traumatic brain injury: translational challenges and emerging therapeutic strategies. Trends Pharmacol Sci 2010;31(12): 596-604. [25] Agre P, Nielsen S, Ottersen OP. Towards a molecular understanding of water homeostasis in the brain. Neuroscience 2004;129(4):849-50. [26] Cernak I, Chang T, Ahmed FA, Cruz MI, Vink R, Stoica B, et al. Pathophysiological response to experimental diffuse brain trauma differs as a function of developmental age. Dev Neurosci 2010;32(5–6):442-53. [27] Ding Y, Yao B, Lai Q, McAllister JP. Impaired motor learning and diffuse axonal damage in motor and visual systems of the rat following traumatic brain injury. Neurol Res 2001;23(2–3):193-202. [28] Timofeeva OA, Eddins D, Yakel JL, Blackshear PJ, Levin ED. Hippocampal infusions of MARCKS peptides impair memory of rats on the radial-arm maze. Brain Res 2010;1308:147-52. [29] Langdon KD, Granter-Button S, Corbett D. Persistent behavioral impairments and neuroinflammation following global ischemia in the rat. Eur J Neurosci 2008;28(11): 2310-8. [30] Witt KA, Mark KS, Huber J, Davis TP. Hypoxia-inducible factor and nuclear factor kappa-B activation in blood–brain barrier endothelium under hypoxic/reoxygenation stress. J Neurochem 2005;92(1):203-14.