Article Text


Nitric oxide in acute ischaemic stroke: a target for neuroprotection
  1. Department of Geriatric Medicine, University of Birmingham, Clinical Investigation Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK
  2. Department of Clinical Pharmacology, University of Birmingham, Clinical Investigation Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK
  1. Dr O’Mahony, Department of Geriatric Medicine, University of Birmingham, Clinical Investigation Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK. email: d.omahony{at}
  1. Department of Geriatric Medicine, University of Birmingham, Clinical Investigation Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK
  2. Department of Clinical Pharmacology, University of Birmingham, Clinical Investigation Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK
  1. Dr O’Mahony, Department of Geriatric Medicine, University of Birmingham, Clinical Investigation Unit, Queen Elizabeth Hospital, Edgbaston, Birmingham B15 2TH, UK. email: d.omahony{at}

Statistics from

Stroke illness remains the single greatest cause of serious physical disability in western countries and is a major consumer of scarce medical resources. In the United Kingdom, it is estimated that the management of stroke patients accounts for 4%–5% of the National Health Service budget.1 Clearly, any intervention that reduces cerebral infarct size will have a major impact on the personal and health budget costs of stroke illness. At present, low dose aspirin and early referral to a dedicated “stroke unit” seem to be the two positive things that physicians can do for patients with acute ischaemic stroke that significantly reduce mortality and dependency levels, for which there is a firm evidence base. Treatment with thrombolytic drugs may be beneficial in those with acute stroke who are fortunate enough to obtain treatment within 3 hours of onset of symptoms.2 However, with such a narrow therapeutic time window, early thrombolysis is applicable in only a small minority of patients. Similarly, human trials of several neuroprotective agents which work in animals have so far proved disappointing.

In the midst of this therapeutic gloom, scientists and clinicians have been forced to re-evaluate the potential therapeutic targets in the cerebral ischaemic cascade. Perhaps the reason for the failure of neuroprotective interventions is that trials have concentrated on inhibiting the early components of the cascade which are triggered within minutes of onset of ischaemia. Attempts to block these processes after several hours, particularly glutamate activation of N-methyl-D-aspartate receptors and intracellular calcium influx may be a futile exercise. Or, it may be that potentially beneficial drugs simply do not reach their site of action because of the ischaemia affecting the penumbral zone during the putative therapeutuc time window. Is there then scope for effective therapeutic intervention beyond the first few hours of onset of ischaemia? Given recent evidence indicating that the ischaemic penumbra in humans remains metabolically viable for up to 24 hours after onset of cerebral ischaemia,3 penumbral salvage and infarct volume reduction should, in theory, be feasible. Information about the delayed neurodestructive pathways has increased rapidly in recent years. Free radical oxidative damage, cell energy depletion, and altered gene expression leading to apoptotic cell death seem to be particularly important processes in penumbral deterioration. The purpose of this article is to focus on nitric oxide release as a delayed neurodestructive mechanism in acute cerebral ischaemia and the pharmacological methods of inhibiting it that may be applicable in humans.

Nitric oxide and other oxidative free radicals in cerebral ischaemia

By virtue of their unpaired electron, oxidative free radicals (for example, hydroxyl and superoxide anions) are highly reactive chemical species. Under normal circumstances, free radicals that are produced as byproducts of cellular metabolism are removed by free radical scavenging enzymes, such as superoxide dismutase, catalase, glutathione peroxidase, and other scavenging substances such as α-tocopherol and ascorbic acid. Free radicals are produced in high concentration after mitochondrial failure from cellular deprivation of oxygen and glucose. Scavenger systems become rapidly saturated, allowing free radicals to react avidly with and damage complex nucleic acid, lipid, carbohydrate, and protein molecules. This leads to major disruption of normal cellular processes and eventual cell death. Further, oxidative free radical release occurs on reperfusion of an ischaemic area of brain, contributing to socalled “reperfusion injury”. The importance of free radical cytotoxicity is further seen in transgenic mice underexpressing the superoxide dismutase gene. These animals develop significantly larger cerebral infarcts than wild-type mice.4 Conversely, intravascular administration of superoxide dismutase bound to polyethylene glycol results in smaller cerebral infarcts in mice subjected to ischaemia-reperfusion compared with vehicle treated animals.5 Also, mice genetically altered to overexpress superoxide dismutase develop significantly smaller cerebral infarcts.6 Superoxide dismutase-polyethylene glycol complex might be beneficial in acute ischaemic stroke, given its efficacy in patients with severe head injury,7 but clinical trials in stroke are awaited.

Nitric oxide is a free radical substance with several well established physiological properties including functional regulation of vascular smooth muscle cells, neurons, macrophages, platelets, and neutrophils when produced in very small, regulated concentrations. Nitric oxide synthase activity and nitric oxide release are greatly increased in the acutely ischaemic brain.8 Nitric oxide reacts rapidly with superoxide to form peroxynitrite, which is a powerful oxidant and highly cytotoxic (figure). Nitric oxide in high concentration is indirectly neurotoxic through other mechanisms including iron-mediated lipid peroxidation (nitric oxide liberates iron from cell stores) and cell energy depletion by disruption of mitochondrial enzymes and nucleic acids. Release of nitric oxide may also trigger neuronal apoptotic cell death.9

Causes and consequences of nitric oxide release in acute cerebral ischaemia. NMDA/Ca2+=N-methyl-D-aspartate/calcium ion membrane complex; nNOS=neuronal nitric oxide synthase; iNOS=inducible nitric oxide synthase; NO*=nitric oxide; O2*=superoxide anion; ONOO=peroxynitrite.

Endothelial and neuronal nitric oxide synthase isoforms in cerebral ischaemia

Nitric oxide release in acute cerebral ischaemia may have positive as well as negative effects (table). Increased endothelial nitric oxide synthase (eNOS) activity seems to be neuroprotective, probably through its cerebral vasodilatory effects as well as by inhibition of platelet aggregation and leucocyte endothelial adhesion. Mouse models of stroke in which the eNOS gene is not expressed (eNOS knockout) develop significantly larger cerebral infarcts after permanent middle cerebral artery occlusion (MCAO).10 Conversely, enhanced eNOS mediated nitric oxide release with intravascular eNOS substrate, L-arginine and nitric oxide donor drugs such as sodium nitroprusside results in smaller cerebral infarcts than those in vehicle treated animals.9

Nitric oxide synthase isoforms and their pharmacological properties in relation to acute cerebral ischaemia

By contrast with the neuroprotective effect of enhanced endothelial nitric oxide, neuronal release of nitric oxide plays a major part in ischaemic neurotoxicity. After cerebral ischaemia, glutamate release in high concentration activates several postsynaptic glutamate receptor/ion channel complexes, most notably N-methyl-D-aspartate (NMDA)/Ca2+. Intracellular Ca2+ influx activates several calcium dependent cytodestructive enzymes, including neuronal nitric oxide synthase (nNOS), leading to release of nitric oxide in high local concentrations with resultant free radical damage. Studies involving transgenic mouse models and selective nNOS inhibitors further highlight the importance of nNOS in the pathophysiology of cerebral ischaemia (table). Huanget al 11 have recently shown an average infarct volume reduction of 38% in nNOS knockout mice compared with wild-type mice. Similarly, treatment with selective inhibitors of nNOS, 7-nitroindazole and ARL17477, produces significant infarct volume reduction in rats (up to 27% with 7-nitroindazol.12 From these studies, it would seem that enhancement of eNOS activity using nitric oxide donors combined with selective nNOS inhibitors is an appropriate therapeutic strategy for ischaemic stroke in humans.

Inducible nitric oxide synthase in cerebral ischaemia

The situation is not quite as straightforward as this, however. In rodents, the neuroprotective effect of L-arginine and nitric oxide donor drugs is lost about 2 hours after onset of cerebral ischaemia. A greater cause for concern is the finding that intravascular L-arginine is associated with increased infarct volume when its administration is delayed by 24 hours in rats.13These time dependent opposite effects of L-arginine are thought to result from the appearance in the ischaemic penumbra of the third nitric oxide synthase isoform, inducible or immunological NOS (iNOS), after a time lag of 6–12 hours (table).9 Inducible NOS is known to contribute to delayed neuronal injury in stroke, as iNOS knockout mice develop significantly smaller cerebral infarcts after focal ischaemia than wild type mice.14 Inducible NOS differs fundamentally from eNOS and nNOS as its activation is not calcium dependent—that is, iNOS may be fully activated at basal intracellular concentrations of calcium, whereas eNOS and nNOS only become upregulated at high calcium concentrations.15 The appearance of iNOS after cerebral ischaemia is delayed because, unlike eNOS and nNOS, iNOS is not a constitutive enzyme and is only produced after cytokine stimulation of neutrophils resulting from local ischaemia.9 Inducible NOS messenger RNA is detectable 12 hours after ischaemia, reaches peak concentrations at 48 hours, and takes about 7 days to return to baseline.16 The major source of delayed nitric oxide release in the ischaemic brain, therefore, is thought to be iNOS. The delayed appearance of iNOS would also explain the dual early neuroprotective and later neurotoxic action of L-arginine, as L-arginine is the principal natural substrate for each of the three NOS isoforms.

Neuronal nitric oxide synthase as a therapeutic target in cerebral ischaemia

Neuroprotection from selective nNOS inhibitors in rodents seems to come from early intervention—that is, in the first 2 hours after focal cerebral ischaemia. In a mouse model of traumatic brain injury, the selective nNOS inhibitor 7-nitroindazole significantly reduced the neurological deficit when given at 5, 30, and 60 minutes after injury but not at 2 hours.17 Neuronal NOS has been shown to have a major role in the cerebral hyperaemic response to hypoxia18 and therefore inhibition of nNOS could, in theory, disrupt an important compensatory mechanism in cerebral ischaemia. In addition, there may be risks from inhibiting an enzyme that contributes to such major neurological functions as synaptic plasticity and neuronal signalling. There is also concern that the finding of disturbed, aggressive behaviour in nNOS knockout mice19 could mean that comparable adverse effects with selective nNOS inhibitors may occur in humans. Finally, inhibition of nNOS activates nuclear factor-κβ that leads to induction of iNOS,20 which could, in theory, indirectly increase tissue damage after cerebral ischaemia. For these reasons, there are serious doubts whether selective nNOS inhibitors are applicable in acute stroke in humans.

Inhibition of inducible nitric oxide synthase: aminoguanidine

In many respects, iNOS seems to be the most appropriate target for pharmacological manipulation of nitric oxide activity in acute cerebral ischaemia, as its appearance is delayed for several hours after onset of ischaemia and its effects are protracted over several days, allowing time for clinical intervention.16 Cerebral infarcts after permanent MCAO are 28% smaller and motor deficits significantly less in iNOS gene knockout mice compared with wild-type mice.14Furthermore, aminoguanidine, a relatively selective iNOS inhibitor, shows time and dose dependent neuroprotective effects—that is, the earlier the drug is given and the higher the dose, the greater the reduction of infarct volume. Reductions of 88% and 85% in infarct volume have been reported when aminoguanidine is given 1 and 2 hours respectively after acute cerebral ischaemia in mice.21Moreover, unlike other neuroprotective agents studied to date, aminoguanidine still affords significant neuroprotection even when administration is delayed by 24 hours after onset of ischaemia, with reported infarct volume reduction of 33%.22 The degree of reduction in infarct volume with aminoguanidine correlates with the level of residual motor function—that is, the neuroprotective effect of aminoguanidine is functionally significant. Of interest, in one recent study in rats the reduction of cerebral infarct volume was not significantly less when aminoguanidine was given 24 hours after onset of ischaemia compared with starting treatment at 12 hours.23

Importantly, aminoguanidine inhibits pathways other than nitric oxide which may be important in mediating cytotoxicity after acute cerebral ischaemia—that is, it is not entirely selective for iNOS. For example, it also inhibits polyamine oxidase in the brain. Polyamine oxidase mediated release of 3-aminopropanal in high concentrations after cerebral ischaemia is directly neurotoxic and induces glial apoptotic cell death. After cerebral ischaemia, there is a marked increase in polyamine oxidase activity and aminoguanidine inhibits the consequent overproduction of 3-aminopropanal with associated reduction of cerebral infarct volume.24 Through its inhibitory effect on advanced glycation end product formation, aminoguanidine is also known to be beneficial in preventing complications related to small vessel disease in diabetes mellitus (retinopathy, neuropathy, and nephropathy) when given for long periods—that is, months to years.25However, this property is of doubtful relevance to neuroprotection from acute cerebral ischaemia.

The ability of aminoguanidine to disrupt a major delayed neurodestructive pathway in acute cerebral ischaemia is an important difference from the neuroprotective therapies so far tried in humans. From the evidence of animal studies, this property could be highly valuable in the treatment of virtually all patients presenting with acute ischaemic stroke, particularly those presenting to hospital outside the 6 hour time window for most other neuroprotective drugs. Aminoguanidine is already known to be safe and well tolerated in humans25 and it may be given intravenously as well as orally. Assuming that there is similar upregulation of brain iNOS in humans as in rodents after acute cerebral ischaemia (although this remains to be proved), aminoguanidine could be a highly effective neuroprotective drug. The therapeutic potential of this and other selective iNOS inhibitors in acute stroke should now be pursued with clinical trials.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.