Neuroprotective κ-Opioid Receptor Agonist BRL 52537 Attenuates Ischemia-Evoked Nitric Oxide Production In Vivo in Rats
Background and Purpose— κ-Opioid receptors (KOR) have been implicated in neuroprotection from ischemic neuronal injury. We tested the effects of a selective and specific KOR agonist, BRL 52537 hydrochloride [(±)-1-(3,4-dichlorophenyl)acetyl-2-(1-pyrrolidinyl) methylpiperidine], on infarct volume and nitric oxide production after transient focal ischemia in the rat.
Methods— With the use of the intraluminal filament technique, halothane-anesthetized male Wistar rats (weight, 250 to 300 g) were subjected to 2 hours of focal cerebral ischemia confirmed by Doppler flowmetry. In a blinded randomized fashion, rats were treated with intravenous saline or 1 mg/kg per hour BRL 52537 infusion, initiated 15 minutes before occlusion and maintained until 2 hours of reperfusion. In a second experiment, rats were treated during reperfusion with saline or 1 mg/kg per hour BRL 52537, initiated at onset of reperfusion and continued for 22 hours. In a final experiment, in vivo striatal nitric oxide production was estimated via microdialysis by quantification of citrulline recovery after labeled arginine infusion in striatum of intravenous BRL 52537– or saline-treated rats.
Results— In rats treated with BRL 52537 during ischemia and early reperfusion, infarct volume was significantly attenuated in cortex (16±6% versus 40±7% of ipsilateral cortex in saline group) and in caudoputamen (30±8% versus 66±6% of ipsilateral caudoputamen in saline group). Infarct volume was also reduced by treatment administered only during reperfusion in cortex (19±8% in BRL 52537 group [n=10] versus 38±6% in saline group) and in caudoputamen (35±9% versus 66±4% in saline group). BRL 52537 treatment markedly attenuated NO production in ischemic striatum compared with saline-treated controls.
Conclusions— These data demonstrate that (1) the selective KOR agonist BRL 52537 provides significant neuroprotection from focal cerebral ischemia when given as a pretreatment or as a posttreatment and (2) attenuation of ischemia-evoked nitric oxide production in vivo may represent one mechanism of ischemic neuroprotection.
The opioidergic system in brain has been implicated in the pathophysiology of cerebral ischemia.1 Three receptor subtypes have been identified, μ, κ, and δ, all of which have been shown to produce antinociceptive effects.2 κ-Opioid receptor (KOR) agonists can also act as neuroprotectants in some animal models of cerebral ischemia. For example, dynorphin A3 and other selective nonpeptide KOR agonists such as U-62,066E and U-50,488H and their analogues reduce mortality and ameliorate hippocampal CA1 neuronal necrosis after transient global ischemia.4,5⇓ In permanent focal ischemia, GR89696 and CI-977 reduce cortical damage in rodents and in cat.2,6–8⇓⇓⇓ Furthermore, KOR agonists improve postischemic recovery of complex neurobehavior.9–11⇓⇓ Behavioral recovery is facilitated with KOR agonist PD117302 after transient global ischemia and correlates with hippocampal CA1 neuronal protection.12
However, the signaling mechanisms utilized by KOR under normal conditions or during neuroprotective actions are not well established in vivo. In vitro,13–16⇓⇓⇓ KOR agonists modulate glutamate excitotoxicity by inhibiting presynaptic glutamate release17 via closure of N-type calcium channels18 and restriction of calcium entry into presynaptic terminals. KOR agonists also inhibit excitatory postsynaptic potentials through similar presynaptic mechanisms involving reduced glutamate release.13,16,17⇓⇓ Accordingly, we hypothesized that neuroprotection associated with KOR agonists could be linked to dampening of excitotoxic injury or cell death mechanisms downstream, such as N-methyl-d-aspartate (NMDA)–evoked nitric oxide (NO) production. The over-elaboration of NO generated from type 1 and II NO synthase (NOS) located in neurons and inflammatory cells and conversion to peroxynitrite is a well-described mechanism of ischemic cell damage.19,20⇓ Pharmacological inhibitors of NO toxicity are well known to reduce infarction volume after experimental stroke.20,21⇓
In the present study we investigated whether a highly selective KOR agonist, BRL 52537 hydrochloride [(±)-1-(3,4-dichlorophenyl)acetyl-2-(1-pyrrolidinyl) methylpiperidine],22–25⇓⇓⇓ reduces infarct volume in a well-characterized model of transient middle cerebral artery occlusion (MCAO) when given as an intravenous pretreatment after 2 hours of MCAO. Furthermore, we determined whether BRL 52537 neuroprotection is sustained when treatment is restricted to the reperfusion period. Histological consequences of prolonged continuous treatment with BRL 52537 were examined in naive rats to exclude potential toxicity to neurons or glia. Finally, we tested the hypothesis that BRL 52537 decreases ischemia-evoked NO production as a potential mechanism of neuroprotection.
Materials and Methods
General Preparation and Animal Surgery
All experimental protocols were approved by the Institutional Animal Care and Use Committee and conformed to the National Institutes of Health guidelines for the care and use of animals in research. All techniques are as previously described.26–30⇓⇓⇓⇓ In brief, adult male Wistar rats (weight, 250 to 300 g) were anesthetized with halothane (1.0% to 2.0%) in oxygen-enriched air and allowed to ventilate spontaneously. With the use of aseptic surgical techniques, the right femoral artery was cannulated to monitor arterial blood pressure and arterial blood gases, and the femoral vein was cannulated for vascular access. After cannulation, both catheters were tunneled subcutaneously and exteriorized through a swivel that was sutured over the posterior mid-thorax that allows for the rat to move freely in its cage after emergence from anesthesia. Rectal and temporalis muscle temperatures were maintained with a heating lamp throughout surgical procedures.
Focal Ischemia and Reperfusion
Cortical perfusion was measured with laser-Doppler flowmetry (LDF) as previously described26–30⇓⇓⇓⇓ (model MBF3D, Moor Instruments Ltd). Transient focal ischemia (2 hours) was produced by MCAO with the use of an intraluminal suture technique as previously described28 with modifications.27,29,30⇓⇓ Briefly, the right common carotid artery (CCA) was exposed through a lateral neck incision, and the external carotid artery (ECA) was ligated. The occipital artery branch of the ECA was coagulated, and the internal carotid artery (ICA) was carefully separated from the vagus nerve. The pterygopalatine artery was ligated with a 4-0 silk suture close to its origin. Ischemia was produced by advancing a 4-0 monofilament nylon suture, with its distal tip rounded by application of heat, into the ICA through a puncture in the CCA until the LDF signal displayed a significant reduction. After placement, the intraluminal suture was secured with a 6-0 silk suture tied around the ICA. Reperfusion was produced by withdrawal of the intraluminal suture; this was associated with rapid restoration of the LDF signal. Rats that did not demonstrate significant reduction of the LDF signal (≤40% of baseline) during MCAO or rapid restoration of the LDF signal during reperfusion were excluded from the study. LDF measurements were averaged over 5-minute periods at 5, 15, 30, 60, 90, and 120 minutes of MCAO and at 15 minutes of reperfusion.
All experiments were performed in a blinded, randomized fashion. In the first set of experiments, rats were treated with either saline or 1 mg/kg per hour BRL 52537 as a continuous intravenous infusion starting 15 minutes before MCAO and continued until 2 hours of reperfusion. In a second set of experiments, rats were treated with saline or 1 mg/kg per hour BRL 52537 starting at onset of reperfusion and continued until 22 hours of reperfusion. All infusions were at a rate of 0.5 mL/h. In all experiments rats were allowed to emerge from anesthesia at 15 minutes of reperfusion and were provided free access to food and free water. At 22 hours of reperfusion, rats were deeply anesthetized with 5% halothane and decapitated. The brain was harvested and sliced into seven 2-mm-thick coronal sections for staining with 1% triphenyltetrazolium chloride (TTC) in saline at 37°C for 30 minutes, as previously described.29,30⇓ Infarction volume was measured by a blinded observer using digital imaging (Digital Camera 40, Eastman Kodak Co) and image analysis software (SigmaScan Pro, Jandel). The infarcted area was numerically integrated across each section and over the entire ipsilateral hemisphere. Infarct volumes were measured separately in cerebral cortex and caudoputamen and expressed as a percentage of the ipsilateral structure volume, as previously described.29,30⇓
Cerebral Microdialysis, MCAO, and Estimation of NO Production
Cerebral microdialysis experiments were performed in a separate cohort of rats, as described previously.26,27⇓ Briefly, the rat’s head was placed in a Kopf stereotaxic frame, and microdialysis cannulas were placed into the striatum bilaterally. Under microscopic observation to minimize trauma to cortex, cannulas were advanced to predetermined coordinates (0.5 mm anterior and 2.5 mm lateral to the bregma; depth 6 mm from the dura) with a micromanipulator and fixed in position with dental cement. The animals were then removed from the stereotaxic apparatus and allowed a 60-minute postsurgical equilibration period before the experiment began. The cannulas were then perfused with 3 μmol/L [14C]l-arginine in artificial cerebrospinal fluid at 1 μL/min (2 hours of “preloading” before MCAO, 2 hours of MCAO, and 3 hours of reperfusion). Rats were randomized to receive either an intravenous infusion of 1 mg/kg per hour BRL 52537 or saline that was initiated 30 minutes before MCAO. As in the previous experiments, brains were harvested at 22 hours of reperfusion and sectioned for confirmation of microdialysis probe position and for analysis of infarction volume.
NO production in the dialysate collection was measured as originally described by Bredt et al,31 with modifications.26,27,32–34⇓⇓⇓⇓ The method is based on the premise that the NOS substrate arginine is converted to equimolar concentrations of citrulline and NO.31 Briefly, 20-μL effluent dialysate samples were collected in 20-minute epochs, assayed for [14C]l-citrulline content, and compared in a paired manner (right versus left striatum in each animal). Dialysate samples were poured over 0.5-mL resin AG-50WX8 (Na+ form, pH 7.0) 400-mesh columns, and radioactivity of column flow was quantified by liquid scintillation spectroscopy. Specific activity was corrected for background activity and expressed as femtomoles per minute of perfusion. The efficiency of arginine trapping is approximately 99%, and citrulline flow-through is 92% to 96%.
To evaluate potential toxicity of the KOR agonist, separate naive nonischemic rats were treated with a continuous intravenous infusion of 1 mg/kg per hour BRL 52537 or saline at 0.5 mL/h for 4 days. Rats were then killed by decapitation under 5% halothane anesthesia and perfused intracardially with 10% neutral buffered formalin. The brains were postfixed in formalin and cut coronally at 2-mm intervals from the frontal pole. The coronal slices were embedded in paraffin, cut at 10-μm–thick sections, and stained with hematoxylin and eosin, hematoxylin and eosin with Luxol fast blue counterstain for myelin, and an antibody to glial fibrillary acidic protein (DAKO), as previously described.30 Histopathological evaluation was performed by a neuropathologist blinded to treatment group.
BRL 52537 HCl was obtained from Research Biochemical International.
Infarction volume and physiological data between and within groups were analyzed by 2-way ANOVA. For microdialysis experiments, citrulline effluent within groups was analyzed by 2-way ANOVA; citrulline from the 2 striata (ischemic versus nonischemic) was analyzed as 1 within-subjects factor, and the 20-minute collections were analyzed as a second within-subjects factor. If the overall effect of treatment or the treatment versus time interaction was significant, comparisons of mean values between the 2 treatments at individual time points were made by orthogonal contrasts. A value of P<0.05 was considered significant. Data are presented as mean±SEM.
Mean arterial blood pressure, Paco2 and Pao2, pH, and temporalis muscle temperature were within normal physiological ranges in all animals at baseline, during MCAO, and at early reperfusion (Table). In the first series of experiments with preischemic treatment, residual LDF signal was similar in rats pretreated with control saline (n=10) or 1 mg/kg per hour BRL 52537 (n=10) (Figure 1). One rat in the saline-treated group died before completion of the experimental protocol because of subarachnoid hemorrhage that was apparently produced by inadvertent suture penetration of the MCA. TTC-determined infarct volume at 22 hours of reperfusion was significantly attenuated in BRL 52537–treated rats in the cortex (16±6% of ipsilateral structure) compared with saline-treated controls (40±7%; P<0.05) as well as in the caudoputamen complex compared with saline-treated controls (30±8% versus 66±6%; P<0.05) (Figure 2).
In a second series of experiments with treatment at reperfusion, physiological parameters (data not shown) were again equivalent between treatment groups (n=10 per group), and residual LDF signal averaged for 2 hours during MCAO was similar in the 1 mg/kg per hour BRL 52537–treated group (38±2% of preischemic baseline) and saline-treated controls (36±2%). Two rats in the saline-treated group died before completion of the experimental protocol. There was a significant attenuation in infarct volume in the cortex in 1 mg/kg per hour BRL 52537–treated rats (19±8% of ipsilateral cortex versus 38±6% in saline-treated rats) and in caudoputamen (35±9% of ipsilateral caudoputamen versus 66±4% in saline-treated rats) (P<0.05) (Figure 3).
Effect of BRL 52537 on NO Production During MCAO and Reperfusion
Physiological parameters and intraischemic LDF were not different in BRL 52537 versus control groups (n=7 per group; data not shown). Microdialysis probe tracts were confirmed to be intrastriatal in all rats. As expected, there was a time-dependent increase in labeled citrulline recovery bilaterally in all rats (Figure 4). Compared with paired nonischemic striatum, labeled citrulline recovery increased throughout occlusion and continued to increase during reperfusion in saline-treated animals, but less so in BRL 52537–treated rats. Infarction volume at 22 hours of reperfusion was significantly attenuated in BRL 52537– versus saline-treated rats in cortex (20±8% versus 59±5% of ipsilateral cortex; P<0.05) as well as in the caudoputamen complex (26±7% versus 69±4% of ipsilateral caudoputamen; P<0.05).
Light microscopy on brains harvested from nonischemic rats 4 days after continuous intravenous infusion (0.5 mL/h) with 1 mg/kg per hour BRL 52537 or control saline (n=3 per group) revealed no gross histological differences among treatment groups. There was no mortality in any of the treatment groups before the desired end point. There was no evidence of hypoxic-ischemic injury to gray or white matter, macrophage infiltration or myelin pallor, or sponginess to suggest white matter injury or edema. There was no evidence of reactive gliosis in hematoxylin and eosin–stained sections. Immunohistochemistry for glial fibrillary acidic protein did not reveal reactive astrocytes.
This study demonstrates 3 important findings. First, intravenous administration of a selective KOR agonist, BRL 52537, provides significant ischemic neuroprotection when administered as either a pretreatment or a posttreatment initiated at the onset of reperfusion. Second, prolonged intravenous infusion up to 4 days does not result in gross neuropathology or myelin injury. Third, ischemia-evoked NO production in the striatum is attenuated by BRL 52537 at the neuroprotective dose, suggesting that one mechanism of this KOR agonist in focal stroke is via reduction of early NO toxicity.
Pharmacologically defined subtypes for the opioid receptors are well established on the basis of relative affinity and rank order of agonist potencies.35,36⇓ Autoradiographic studies have shown 2 functionally distinct KOR (κ1 and κ2) in rat brain,35,36⇓ and both full-length and truncated KOR mRNA transcripts have been characterized in brain.37 We utilized a water-soluble agent, BRL 52537 hydrochloride,24–27⇓⇓⇓ which is highly specific for the κ-receptor (eg, Ki κ, 0.24 nmol/L; Ki μ, 1560 nmol/L) with 16 times the potency of standard KOR ligands such as U-60488 as a pharmacological tool. Intravenous administration did not have any significant effects on the physiological parameters evaluated within our study. The relative κ1-κ2 affinity of BRL 52537 is not well described. At present, receptor subtype–specific KOR agonists are not available.
The finding that BRL 52537 reduces damage from experimental stroke is consistent with several previous animal studies. Baskin et al8 demonstrated ischemic neuroprotection in cat with 3 different κ-agonists when drugs were started 6 hours after vascular occlusion and continued with slow-release subcutaneous osmotic pumps for continuous drug delivery for 7 days. In mice, opioid receptor binding studies using focal cerebral ischemia suggest that KOR binding is preserved for long time periods (12 to 48 hours), more so than δ- and μ-opioid receptors,38 suggesting a potentially long therapeutic window. However, a recent randomized clinical trial with a selective KOR antagonist, nalmefene hydrochloride (Cervene), in stroke patients did not improve functional outcomes at 3 months.39 Nevertheless, in our study the finding that a short treatment duration (2 hours as well as 22 hours) yielded significant neuroprotection may reflect the long duration of action of BRL 52537. Prolonged treatment did not produce gross evidence of neuronal death, myelin injury, or gliosis. We specifically failed to observe morphological changes in cingulate gyrus and retrosplenial cortex, areas that are commonly injured in toxicity seen with NMDA receptor antagonists.40
The principal novel finding of this study is that BRL 52537 attenuates ischemia-evoked striatal NO production, and this action was accompanied by amelioration of regional infarct volume. Others have postulated that KOR agonists have interactions with NO. For example, repeated injections of a KOR agonist, U-50,488H, in mice under nonischemic conditions increase NOS activity in cortex but not in the cerebellum, midbrain, or spinal cord.41 Although we cannot distinguish the cellular source of NO in our microdialysis sampling technique, an indirect method of assessing NO activity in vivo, there was a clear reduction of NO production in drug-treated brain. With the use of our technique, labeled citrulline recovery is a reflection of the sum total of NO from both endothelial (type III NOS) and neuronal (type I NOS) sources. However, because the major source of NO during early ischemia is from neurons,19–21⇓⇓ we speculate that this source contributes largely to our measurements. We27 and others42 have previously demonstrated significant reduction in injury volume with selective neuronal NOS inhibitors in the rat model of MCAO. The inducible isoform of NOS (iNOS or type II NOS) is also expressed by neurons, endothelial cells, and microglia during cerebral ischemia19,20⇓ and clearly plays a role in stroke outcome, as evidenced by its selective inhibition or its genetic deletion.21 Effects of KOR agonists on iNOS activity are unknown at present, and we cannot distinguish interactions between BRL 52537 and iNOS. However, it seems unlikely that iNOS activity contributed to labeled citrulline recovery in the rat experiments. We monitored in vivo NOS activity for up to 5 hours after occlusion in these animals by microdialysis and detected large increases in labeled citrulline recovery within 1 hour of MCAO. The steady production of labeled citrulline persisted throughout the 5-hour monitoring period. Such an acute time frame of NO production is probably too early to reflect iNOS activity. Iadecola et al43–45⇓⇓ reported that induction of iNOS became detectable at 12 hours but not at 6 hours after focal ischemia. However, we cannot exclude that placement of the probe may have caused release of inflammatory mediators that more rapidly induce iNOS.46
Furthermore, additional antiexcitotoxic mechanisms may be important in the neuroprotection provided by KOR agonists in cerebral ischemia. For example, in vitro13–16⇓⇓⇓ use of KOR agonists modulates glutamate toxicity by presynaptic inhibition of glutamate release, possibly by decreasing or by closing N-type Ca2+ channels.17 KOR agonists may also inhibit excitatory postsynaptic potentials by attenuating presynaptic Ca2+ influx and subsequent glutamate release,13–16⇓⇓⇓ as well as by modulating NMDA-induced dopamine release.35 Other alternative neuroprotective mechanisms include amelioration of cerebral edema.5,47⇓
We chose the striatum as our region of study because it is highly vulnerable to cerebral ischemia48 and has abundant synthetic mechanisms for NO in the rat.22 MCAO enhances labeled citrulline recovery in the ischemic striatum, which is attenuated by local infusion of NOS inhibitor l-nitroarginine, thereby reflecting increased NO production.26,27,32–34⇓⇓⇓⇓ The time-dependent increase in recovery of labeled citrulline with control perfusion of artificial cerebrospinal fluid is presumed to reflect a dynamic and complex kinetic process involving diffusion of labeled arginine across the dialysis membrane, its cellular uptake and efflux, and diffusion of labeled citrulline back to and across the dialysis membrane.32–34⇓⇓ We used a bilateral microdialysis perfusion in a paired experimental design to reduce interanimal variability arising from physiological factors such as arterial blood pressure, blood gases, and depth of anesthesia. However, variability may also arise from (1) differences in the efficiency of arginine trapping by the Dowex column; (2) biological variability secondary to differences in infarction volume in the striatum; and (3) tissue injury resulting from placement of microdialysis probe.
In conclusion, these data demonstrate that a continuous intravenous infusion of the potent KOR agonist BRL 52537 attenuates early stroke damage when given as a pretreatment or during reperfusion and acts, in part, by a mechanism that may be linked to attenuation of ischemia-evoked NO toxicity.
This work was supported in part by US Public Health Service, National Institutes of Health grants NS20020, NS33668, and NR03521. Dr Bhardwaj is supported in part by an Established Investigator Grant from the American Heart Association.
- Received July 3, 2002.
- Revision received December 4, 2002.
- Accepted January 2, 2003.
- ↵Zabramski JM, Spetzler RF, Selman WR, Roesmann UR, Hershey LA, Crumrine RC, Macko R. Naloxone therapy during focal cerebral ischemia evaluation in a primate model. Stroke. 1984; 15: 621–627.
- ↵Baskin DS, Widmayer MA, Browning JL, Heizer ML, Schmidt WK. Evaluation of delayed treatment of focal cerebral ischemia with three selective κ-opioid agonists in cats. Stroke. 1994; 25: 2047–2054.
- ↵Wagner JJ, Caudle RM, Chavkin C. κ-Opioids decrease excitatory neurotransmission in the dentate gyrus of the guinea pig hippocampus. J Neurosci. 1992; 12: 132–141.
- ↵Gross RA, Macdonanld RL. Dynorphin A selectively reduces a large transient (N-type) calcium current of mouse dorsal root ganglion neurones in cell culture. Proc Natl Acad Sci U S A. 1987; 84: 5469–5473.
- ↵Samdani AF, Dawson TM, Dawson VL. Nitric oxide synthase in models of focal ischemia. Stroke. 1997; 28: 1283–1288.
- ↵Iadecola C, Zhang F, Casey R, Nagayama M, Ross ME. Delayed reduction of ischemic brain injury and neurological deficits in mice lacking inducible nitric oxide synthase gene. J Neurosci. 1997; 17: 9157–9164.
- ↵Vecchietti V, Clarke GD, Colle R, Giardina G, Petrone G, Sbacchi M. (1S)-1-(Aminomethyl)-2-(arylacetyl)-1,2,3,4-tetrahydroisoquinoline and hetrocycle-condensed tetrahydropyridine derivatives: members of a novel class of very potent kappa opioid analgesics. J Med Chem. 1991; 34: 2624–2633.
- ↵Vecchietti V, Clarke GD, Colle R, Dondio G, Giardina G, Petrone G, Sbacchi M. Substituted 1-(aminomethyl)-2-(arylacetyl)-1,2,3,4-tetrahydroisoquinolines: a novel class of very potent antinociceptive agents with varying degrees for kappa and mu opioid receptors. J Med Chem. 1992; 35: 2970–2978.
- ↵Bhardwaj A, Sawada M, London ED, Koehler RC, Traystman RJ, Kirsch JR. Potent δ1-receptor ligand 4-phenyl-1-(4-phenylbutyl)-piperidine modulates basal and N-methyl-d-aspartate–evoked nitric oxide production in vivo. Stroke. 1998; 29: 2404–2411.
- ↵Goyagi T, Goto S, Bhardwaj A, Dawson VL, Hurn PD, Kirsch JR. Neuroprotective effect of δ1-receptor ligand 4-phenyl-1-(4-phenylbutyl)-piperidine (PPBP) is linked to reduced neuronal nitric oxide production. Stroke. 2001; 32: 1613–1620.
- ↵Longa EZ, Weinstein PR, Carlson S, Cummins R. Reversible middle cerebral artery occlusion without craniectomy in rats. Stroke. 1989; 20: 84–91.
- ↵Alkayed NJ, Murphy SJ, Traystman RJ, Hurn PD. Neuroprotective effects of female gonadal steroids in reproductively senescent female rats. Stroke. 2000; 31: 161–168.
- ↵Bhardwaj A, Harukuni I, Murphy SJ, Alkayed NJ, Crain BJ, Koehler RC, Hurn PD, Traystman RJ. Hypertonic saline worsens infarction volume following transient focal ischemia in rat. Stroke. 2000; 31: 1694–1701.
- ↵Bhardwaj A, Northington FJ, Koehler RC, Steifel T, Hanley DF, Traystman RJ. Adenosine modulates N-methyl-d-aspartate–stimulated hippocampal nitric oxide production in vivo. Stroke. 1995; 26: 1627–1633.
- ↵Bhardwaj A, Northington FJ, Ichord RN, Hanley DF, Traystman RJ, Koehler RC. Characterization of ionotropic glutamate receptor–mediated nitric oxide production in vivo in rats. Stroke. 1997; 28: 850–857.
- ↵Boutin H, Dauphin F, MacKenzie ET, Jauzac P. Differential time-course decreases in nonselective μ-, δ-, and κ-opioid receptors after focal cerebral ischemia in mice. Stroke. 1999; 30: 1271–1277.
- ↵Clark WM, Raps EC, Tong DC, Kelly RE, for the Cervene Stroke Study Investigators. Cervene (nalmefene) in acute ischemic stroke: final results of a phase III efficacy study. Stroke. 2000; 31: 1234–1239.
- ↵Olney JW, Labuyere J, Price MT. Pathological changes induced by phencyclidine and related drugs. Science. 1989; 244: 1360–1362.
- ↵Bhargava HN, Kumar S. Effect of multiple injections of U50,488H, a kappa-opioid receptor agonist, on the activity of nitric oxide synthase in brain regions and spinal cord of mice. J Pharmacol. 1997; 29: 397–399.
- ↵Escott KJ, Beech JS, Haga KK, Williams SC, Meldrum BS, Bath PM. Cerebroprotective effect of the nitric oxide synthase inhibitors, 1-(2-trifluromethylphenyl) imidazole and 7-nitro indazole, after transient focal cerebral ischemia in the rat. J Cereb Blood Flow Metab. 1998; 18: 281–287.
- ↵Iadecola C, Zhang F, Casey R, Clark HB, Ross E. Inducible nitric oxide synthase gene expression in vascular cells after transient focal cerebral ischemia. Stroke. 1996; 27: 1373–1380.
- ↵Brian JE Jr, Faraci FM. Tumor necrosis factor-α–induced dilatation of cerebral arterioles. Stroke. 1998; 29: 509–515.