The utility of melatonin in reducing cerebral damage resulting from ischemia and reperfusion


Address reprint requests to Dr Raymond T. F. Cheung, Division of Neurology, University Department of Medicine, Queen Mary Hospital, The University of Hong Kong, Pokfulam, Hong Kong. E-mail:


Abstract: The brain is highly susceptible to focal or global ischemia. Unless ischemia is promptly reversed, reperfusion produces further cerebral damage. Acute thrombolysis or defibrinogenation is effective only in selective patients with ischemic stroke and carries a significant risk of bleeding complications. Whereas numerous neuroprotectants were shown to be effective in experimental studies, none of them have been shown to work in clinical trials. The major pathogenetic mechanisms of ischemia/reperfusion injury include excitotoxicity, disturbed calcium ion homeostasis, overproduction of nitric oxide and other free radicals, inflammation, and apoptosis. Nitric oxide and other free radicals, the key mediators of excitotoxicity and disturbed calcium ion homeostasis, cause direct injury and also indirectly damage via inflammation and apoptosis. Melatonin is a potent free radical scavenger and an indirect antioxidant. This mini review summarizes the in vivo and in vitro evidence that melatonin protects against ischemia/reperfusion injury. There is convincing evidence from the literature that melatonin treatment is highly effective in different in vivo and in vitro models of excitotoxicity or ischemia/reperfusion in multiple animal species. Melatonin is safe and non-toxic in humans, and its administration via the oral route or intravenous injection is convenient. While more experimental studies should be conducted to further explore the neuroprotective mechanisms and to document any synergistic or additive protection from combining melatonin with thrombolysis, defibrinogenation or other neuroprotectants, interested clinical scientists should consider planning phase II and III studies to confirm the benefit of melatonin as an acute stroke treatment or a preventive measure for stroke patients.


Stroke is the third leading cause of death and a major source of disability in the developed countries and regions, and ischemic stroke is the most common type [1]. Focal ischemia is the predominant pathogenic cause of ischemic stroke, and the prominent histopathological result is cerebral infarction with pannecrosis of neurons, glia, and endothelial cells. Programmed cell death, or apoptosis, also occurs in the surrounding tissue [2, 3]. Thromboembolic occlusion of the supplying artery is the most important cause of focal ischemia in patients [1, 4, 5]. Cardiac arrest, near drowning, carbon monoxide poisoning, massive hemorrhage, and shock constitute the clinical scenarios of global ischemia [4]. The consequence of global ischemia is selective loss of vulnerable neurons by apoptosis in specific brain regions: hippocampal CA1 and CA4 pyramidal neurons, cerebellar Purkinje cells, small and medium-sized striatal neurons, and cortical neurons in layers three, five, and six [6]. Unless ischemia is promptly reversed, reperfusion causes further damage [4, 5, 7, 8].

Intravenous thrombolysis using tissue plasminogen activator within 3 hr of onset, acute defibrinogenation using intravenous modified viper venom within 3 hr of onset, or intra-arterial thrombolysis using prourokinase within 6 hr of onset has been shown to improve functional recovery among highly selected patients with ischemic stroke [9–11]. Nevertheless, acute thrombolysis or defibrinogenation is feasible in a few percent of stroke patients only, and there is a substantial risk of symptomatic hemorrhage into the acute infarct. On the other hand, enhancement of the tolerance of cerebral tissue to ischemia/reperfusion injury has been an attractive concept for three decades, and encouraging results were indeed obtained in experimental studies. Unfortunately, clinical trials have so far failed to confirm any safe and effective neuroprotectants [12]. In this mini review, the published findings relevant to role of melatonin (N-acetyl-5-methoxy-tryptamine) in reducing cerebral damage after ischemia/reperfusion is summarized. To better appreciate the cerebro-protective potential of melatonin, it is important to briefly review the problem of ischemia/reperfusion injury and the pathophysiological mechanisms.

Focal and global ischemia and reperfusion

The adult human brain constitutes only 2% of the total body weight but receives 15% of the total cardiac output and consumes 20% of the oxygen intake. An uninterrupted blood supply of oxygen and glucose is critical because neurons require a high metabolic rate to maintain their transmembrane ionic gradients, electrical activities, synaptic transmission, macromolecular synthesis, efficient intracellular transport, and cytoskeletal integrity [13]. The normal cerebral blood flow (CBF) is about 50–60 mL/100 g/min in primates and 100 mL/100 g/min in rats and gerbils [4, 14]. Fluctuation in CBF with reduction to 50% can be compensated and tolerated without any consequence. When CBF is between 25 and 50%, brain tissue is at risk of ischemic damage from inhibition of protein synthesis, transient potassium ion efflux, transient calcium ion influx, cytotoxic edema and acidosis [13], and increased neuronal apoptosis may occur [2, 3]. Electrical silence with cessation of neural functions develops rapidly when CBF is below 25%. Loss of transmembrane ionic gradients and irreversible damage occur readily if CBF is below 15–20% [13, 15].

During focal ischemia, there is a core of more severe reduction in CBF surrounded by a peripheral zone of milder reduction in CBF, the ischemic penumbra [4, 5, 7, 13]. When there is reperfusion, transient hyperperfusion may occur initially, and later CBF may be depressed or return to normal [7]. Experimentally, focal ischemia can be induced in animals by middle cerebral artery occlusion (MCAO) using ligation of the middle cerebral artery (MCA) with suture threads or clips, or using electrocoagulation, photochemically induced focal thrombosis, endovascular occlusion of the MCA with an intraluminal suture thread, or embolism with microspheres, autologous thrombi or heterologous thrombi [16].

Global ischemia can be achieved experimentally by transient occlusion of both common carotid arteries plus either hypotension or bilateral vertebral artery occlusion [4, 17, 18]. CBF in the forebrain drops to below 10% of normal during global ischemia, whereas CBF in the brain stem remains above 30% [18]. Transient hyperemia develops initially during reperfusion, and a prolonged phase of heterogeneous and variable hypoperfusion (20–80%) follows [7]. Selective ischemic neuronal apoptosis is a delayed phenomenon and develops over 1–7 days [4, 17].

Pathophysiology of ischemia/reperfusion

Ischemic cell death is not instantaneous, and its pathophysiology is complicated (Fig. 1). Severe ischemia rapidly causes necrosis, which is characterized by failure of the cell membrane with swelling of the cell and internal organelles. Immediate mechanisms of ischemic damage are excitotoxicity and disturbed calcium ion homeostasis, free radicals probably mediate ischemic damage during the intermediate stage, and both inflammation and apoptosis cause delayed ischemic damage [19–24].

Figure 1.

A schematic flowchart showing the complex pathophysiology of cerebral ischemia/reperfusion injury and the gray-shaded targets against which melatonin mediates neuroprotection. Major mechanisms of injury include excitotoxicity, disturbed calcium ion homeostasis, overproduction of nitric oxide and other free radicals, inflammation, and apoptosis. Final consequence is cell death and tissue damage by necrosis or apoptosis. Thick arrows represent major pathways, and thin arrows refer to minor pathways. Ca2+, calcium ion; ER, endoplasmic reticulum.

Ischemia causes uncontrolled release of glutamate and aspartate. There is excessive stimulation of different types of glutamate receptors, leading to influx of calcium ion, sodium ion and water into cells [3, 24, 25]. Elevated intracellular calcium ion concentration causes cellular oxidative stress and activation of lipases, proteases and endonucleases, leading to damage of DNA, cell proteins and lipids.

Nitric oxide (NO), a reactive free radical gas, is produced from L-arginine via one of the isoforms of NO synthase (NOS): type I or neuronal NOS (nNOS) encoded on chromosome 12, type II or inducible NOS (iNOS) encoded on chromosome 17, and type III or endothelial NOS encoded on chromosome 7 [26–28]. While NO has a physiological role in neurotransmission, blood pressure regulation and vasodilatation [29, 30], excessive concentration of NO which directly causes lipid peroxidation and depletes cellular energy via disruption of mitochondrial enzymes and nucleic acids [31, 32], reacts with superoxide anion to form peroxynitrite, a highly toxic and reactive nitrogen-based free radical [33], and triggers apoptotic cell death [34].

Depletion of calcium stores in the endoplasmic reticulum during ischemia/reperfusion results in suppression of protein synthesis [19]. Massive influx of calcium ions into mitochondria leads to production of free radicals such as superoxide anion from the mitochondrial respiratory chain, opening of the mitochondrial permeability transition pore, and secondary failure of mitochondrial functions. Superoxide anion can also be produced by cytosolic enzymes such as xanthine oxidase [20, 35]. Hydroxyl radical, which is highly toxic, is readily generated via the metal-mediated Haber–Weiss reaction. The brain is very vulnerable to oxidative damage because of its high rate of oxidative metabolic activity, intense mitochondrial production of reactive oxygen metabolites, relatively low antioxidant capacity, low activity of repair mechanisms, high membrane surface to cytoplasm ratio, and non-replicating nature of neurons [35, 36].

Inflammatory cells accumulate within and around the site of initial injury from ischemia/reperfusion and are capable of producing further damage [21, 23]. As early as 1 hr after ischemia, there is activation and secretion of inflammatory cytokines such as tumor necrosis factor α, interleukin-1, and interleukin-6 [24]. Induction and expression of iNOS and cyclo-oxygenase (COX)-2, at least partly via activation of nuclear factor κB [37], play a critical role in inflammatory reactions [21, 24, 38]. Polymorphonuclear cells elaborate toxic enzymes such as myeloperoxidase to produce further tissue and cellular damage [24]. Free radicals such as NO, superoxide anion, hydroxyl free radical, and peroxynitrite are important mediators of inflammatory tissue damage [39].

Apoptosis may be responsible for up to 50% of cellular death in ischemia [2]. Mitochondrial dysfunctions provide the intracellular signals for apoptosis, and activation of the tumor necrosis factor superfamily receptors constitutes the extracellular signals for apoptosis. Upstream mechanisms involve nuclear factor κB-dependent pathway, P53-dependent pathway and proapoptotic members of the Bcl family [24]. Activation of caspases with release of cytochrome c from mitochondria represents the downstream mechanism [40]. Finally, formation of caspase 3 activates DNA breaking enzymes and energy consuming DNA repair enzymes, leading to DNA breakdown and cell death.

Preclinical evaluation of neuroprotectants

Adequate preclinical evaluation of the potential neuroprotectants has been advocated, including: (1) dose-response information, (2) both permanent and reversible stroke models, (3) determination of treatment time window, (4) monitoring of key physiological parameters, and (5) assessment of infarct volume and functional outcome both acutely and at a later time [3]. A combinational approach in neuroprotection targeting at multiple mechanisms has many theoretical advantages [3], including greater benefit than individual drugs, fewer side effects and longer therapeutic window. An alternative approach is to choose a drug active against multiple pathophysiological mechanisms.

Melatonin protects against excitotoxicity

As a co-treatment added to the culture medium, melatonin was found to protect cultured cerebellar granule neurons of neonatal Sprague–Dawley rats against excitotoxicity from glutamate or kainate but not N-methyl-D-aspartate (NMDA) at various final concentrations [41]. Kainate is an agonist for non-NMDA type of glutamate receptors. Similar results were obtained in a later study in which melatonin treatment was also effective against toxicity because of hydrogen peroxide or NMDA [42]. In addition, melatonin did not affect the binding of glutamate to the receptors in cerebellar membranes. Furthermore, melatonin did not block the kainate-stimulated inward ion currents or kainate-generated increase in cytosolic calcium ion concentration [41]. Pretreatment with melatonin was shown to lessen glutamate-induced apoptotic cell death in the clonal hippocampal cell line HT22 and reduce hydrogen peroxide-induced oxidative cell death in organotypic hippocampal rat brain slice cultures partly via suppression of nuclear factor κB activity [43].

Following these in vitro results, four intraperitoneal (i.p.) injections of melatonin at 2.5 mg/kg, with the first injection made at 20 min before an i.p. injection of kainate at 10 mg/kg and other injections of melatonin at 0, 1, and 2 hr after kainate injection, were shown to reduce neuronal death and apoptosis at day 11 in the hippocampus, amygdala and pyriform cortex as well as behavioral disturbances within 4 hr of kainate injection and biochemical disturbances up to 8 hr after kainate injection in adult male Wistar rats [44]. In addition, melatonin treatment ameliorated kainate-induced apoptosis and loss of Nissl staining at 48 and 72 hr after kainate injection [45]. Furthermore, melatonin treatment reduced the 5-day mortality because of kainate injection from 73% in the vehicle-treated group to 7% in the melatonin-treated group; however, a single injection of melatonin at 10 mg/kg immediately after kainate injection was less effective in reducing the 5-day mortality rate (33%) [46]. Kainate-induced oxidative stress was measured in brain synaptosomes of adult male Wistar rats, and melatonin decreased the kainate-induced lipid peroxidation, suggesting that melatonin ameliorates kainate-induced excitotoxicity via its direct and indirect antioxidant effects [46].

When the kainate model of excitotoxicity and the same regimen of four i.p. injections of melatonin at 2.5 mg/kg were adopted in adult male Sprague–Dawley rats, the benefit of melatonin treatment in reducing the extent of kainate-triggered apoptosis and loss of Nissl staining at 24 and 48 hr after kainate injection were reconfirmed [47]. In addition, these damages were less in the rats treated with kainate at night when the endogenous levels of melatonin are high, suggesting that endogenous melatonin is protective [47]. Following pinealectomy, there is a marked reduction in the serum level of melatonin, and melatonin-deficient adult male Sprague-Dawley rats were more susceptible to kainate-induced excitotoxicity with greater degree of apoptosis and loss of Nissl staining at 72 hr [48].

In an adult mice model of excitotoxicity because of a subcutaneous injection of kainate at 40 mg/kg, severe neurobehavioral disturbances, marked loss of CA3 pyramidal neurons and increased brain lipid peroxidation products would occur [49]. A single i.p. injection of melatonin at 5 mg/kg given 10 min before kainate injection ameliorated these adverse consequences of excitotoxicity. Similar protective results were obtained in another study in which local microinjection of quinolinic acid (an agonist for NMDA receptors) was used to induce hippocampal damage in anesthetized rats and melatonin was given initially as a local co-treatment plus two i.p. injections at 20 mg/kg [50].

In a neonatal mice excitotoxic model of cerebral palsy with periventricular leukomalacia induced by intracerebral injection of ibotenate (an agonist for NMDA receptors) at 10 μg, co-treatment with an i.p. injection of melatonin at 5 mg/kg reduced the lesion size by 82% [51]. In addition, the study provided some evidence that melatonin acts via its membrane receptor and inhibition of cAMP production to mediate its protection against white matter lesion [51]. Thus, there is strong experimental evidence from rat and mice models that melatonin treatment can effectively counteract the downstream harmful effects of excitotoxicity.

Melatonin protects against in vivo focal ischemia/reperfusion

The effects of melatonin treatment on injury as a result of ischemia/reperfusion or free radicals were examined in the hamster cheek pouch microcirculation using in vivo microscopy [52]. Following 30 min of ischemia and 30 min of reperfusion or exposure to free radicals for 2 min in the vehicle-treated hamsters, increased microvascular permeability, adhesion of leukocytes to venules and reduced capillary perfusion were observed. Melatonin as a pre-treatment before ischemia and reperfusion or co-treatment with exposure to free radicals reduced the microvascular disturbances, suggesting that melatonin treatment can protect the endothelial integrity and preserve microvascular perfusion following ischemia/reperfusion in hamster cheek pouch microcirculation [52].

The cerebro-protective role of physiological levels of melatonin may be revealed in melatonin-deficient animals. Pinealectomized adult male Sprague-Dawley rats were more susceptible to focal ischemia with a larger cerebral infarction in both the photothrombotic and MCAO stroke models [48]. The extent of brain injury was determined at 24 hr of photothrombic stroke or at 4 or 6 hr following MCAO for 1 hr using surgical clips plus microsurgical clipping of both common carotid arteries [48]. Infarct volume and degree of apoptosis were measured in a follow-up study at 6 or 24 hr after focal ischemia achieved by MCAO plus occlusion of both common carotid arteries for 90 min in adult male Sprague-Dawley rats [53]. When compared with the sham-operated rats, infarct volume was increased with greater degree of apoptosis in pinealectomized rats. When melatonin at 2.5 mg/kg was given via an i.p. injection at 30 min before onset of ischemia as well as at 0, 1 and 2 hr after reversion of focal ischemia, these adverse effects of pinealectomy were reversed [53].

As pinealectomy may increase arterial blood pressure and influence CBF, hemodynamic parameters, arterial blood gases and hematocrit were also monitored in a subsequent study [54]. The results re-confirmed the opposing effects of pinealectomy and melatonin on brain damage and neurologic disability at 22 hr of reperfusion following focal ischemia induced by endovascular MCAO for 2 hr in adult male Wistar rats [54]. Melatonin at 4 mg/kg given as an i.p. injection before ischemia and repeated before reperfusion reduced the infarct volume by 40% in both pinealectomized and sham-operated rats, and melatonin treatment given only before ischemia or reperfusion was less effective [54].

Adopting a daily regimen of oral melatonin at 20–24 mg/kg with the first dose given before endovascular MCAO for 1 hr in adult male Wistar rats, the treatment reduced cortical infarction by 60% and striatal infarction by 30% at 11 and 19 days after MCAO [55]. In addition, the treatment ameliorated the locomotor deficits and appeared to enhance the survival of glial cells [55]. Melatonin given as four i.p. injections (at onset and 1 hr of ischemia from endovascular MCAO for 2 hr as well as at onset and 1 hr of reperfusion) in adult male Wistar rats was found to decrease the volume of ischemic lesion on diffusion-weighted magnetic resonance imaging at 30 min of reperfusion, neurologic deficits at 24 hr after ischemia, and degree of lipid peroxidation at 2 hr of reperfusion and 72 hr after ischemia [56].

The aforementioned studies have documented the cerebro-protective effects of exogenous melatonin in different rat models of focal cerebral ischemia when melatonin was given as multiple doses with the first one as a pre- or co-treatment. Two recent studies reported that the treatment regimen could be very much simplified. Pretreatment with a single i.p. injection of melatonin at doses between 5 and 15 mg/kg significantly reduced the infarct volume at 72 hr by about 40% without affecting the hemodynamic parameters and regional CBF in both permanent and 3-hr endovascular MCAO stroke models in adult Sprague-Dawley rats [57, 58]. Indeed melatonin treatment was effective when the single injection of melatonin at 5 mg/kg was commenced at 1 hr or less after onset of ischemia because of 3-hr endovascular MCAO in adult Sprague–Dawley rats [59]. Addition of the second and third doses at 24 and 48 hr of ischemia tended to achieve a larger reduction in infarct volume but failed to extend the treatment time window beyond 2 hr of ischemia.

Using electron paramagnetic resonance spectroscopy to measure brain NO concentration at 15 min of endovascular MCAO in adult male Sprague–Dawley rats, elevation in brain NO concentration during focal ischemia was documented in the vehicle-treated rats. A single i.p. injection of melatonin at 1.5, 5, or 50 mg/kg given at 30 min before ischemia significantly suppressed the brain NO concentration during focal ischemia [60].

The COX-2 plays a pathogenetic role in ischemia/reperfusion injury. A combination of melatonin and meloxicam (a COX-2 inhibitor) was found to achieve better results on motor performance and lipid peroxidation in the brain at 24 hr following transient endovascular MCAO for 2 hr than melatonin treatment alone while meloxicam per se was ineffective [61]. Four i.p. injections of melatonin at 20 mg/kg were given at 0, 1.5, 2, and 3 hr after onset of ischemia in this study, and the results suggest synergistic or additive protection from combining melatonin with other neuroprotectants.

Melatonin protects against in vivo global ischemia/reperfusion

Three i.p. injections of melatonin at 10 mg/kg reduced the loss of the CA1 hippocampal neurons in adult male Wistar rats following forebrain ischemia from transient occlusion of both common carotid arteries for 10, 20, or 30 min plus permanent bilateral vertebral artery occlusion only when the first injection was made immediately at reperfusion but not at 30 min before or 1 hr after reperfusion, and the second and third injections were made at 2 and 6 hr of reperfusion [62].

A single i.p. injection of melatonin at 10 mg/kg given at 30 min before forebrain ischemia from bilateral common carotid artery occlusion for 10 min in adult male gerbils were shown to inhibit brain NO production and suppress NOS activity at 5 and 60 min of reperfusion [63]. Multiple indices of tissue damage because of ischemia/reperfusion was examined in another study in which forebrain ischemia for 5 min was induced in adult male gerbils by bilateral common carotid artery occlusion and four i.p. injections of melatonin at 10 mg/kg were given at 30 min before as well as 1, 2, and 6 hr after reperfusion [64]. This treatment regimen reduced NO levels in the plasma at 4 hr, lipid peroxidation in the brain at 1 hr, accumulation of neutrophils in the hippocampus at 4 hr, neurobehavioral disturbances at 24 and 48 hr, cerebral edema at 48 hr, nitrosylation of proteins or activation of nuclear enzyme poly (ADP-ribose) synthetase in the hippocampus at 96 hr, and CA1 neuronal loss at 96 hr after reperfusion [64].

Global ischemia/reperfusion-induced oxidative lipid and DNA damage was assessed in fetal rat brain in a study. An i.p. injection of melatonin at 10 mg/kg given 60 min before bilateral utero-ovarian artery occlusion for 20 min reduced the oxidative damage of lipids and DNA [65].

When global ischemia was induced in adult male cats by cardiopulmonary arrest for 15 min, extensive loss of hippocampal neurons over the CA1 to CA4 fields occurred at day 8 together with significant neurologic deficits from day 1 to 7 [66]. Continuous intravenous infusion of melatonin at 10 mg/kg per hour for 6 hr commencing at 30 min of reperfusion prevented hippocampal neuronal loss and reduced neurologic deficits.

The combination of previous pinealectomy and permanent bilateral carotid artery occlusion was used in adult male Sprague–Dawley rats to assess the additional effects of melatonin deficiency on the consequences of chronic forebrain ischemia [67]. Pinealectomy per se caused a reduction in the hippocampal CA1 and CA4 neurons and increased the immunoreactivity for glial fibrillary acidic protein, and chronic ischemia also reduced the number of hippocampal neurons. Combined pinealectomy and chronic ischemia produced the greatest reduction of hippocampal neurons and the largest disturbance of working memory [67]. Thus, melatonin treatment has been shown to beneficial in models of global ischemia in rats, gerbils and cats.

Neuroprotective effects of melatonin in in vitro studies

Exposure of cultured bovine cerebral endothelial cells to 95 or 100% of oxygen will cause excessive formation of free radicals and in turn activate apoptotic cell death. Melatonin as a co-treatment was found to protect cultured endothelial cells in a dose-dependent manner against DNA damage and cell death because of hyperoxia for 8 hr [68].

Co-treatment with melatonin was shown to protect primary cultures of rat cortical neurons from NMDA-induced excitotoxicity or hypoxia/reoxygenation insult [69]. Moreover, melatonin treatment did not affect the NMDA-induced rise in intracellular calcium ion concentration, and the benefit was not affected by the melatonin membrane receptor antagonist, luzindole [69]. In another study, rat mixed cortical cell cultures were subjected to cyanide- or NMDA-induced toxicity for 24 hr [70]. In addition to NMDA blockade or NOS inhibition, pretreatment with melatonin also reduced the neurotoxicity.

Primary hippocampal cell cultures from embryonic Sprague–Dawley rat fetuses would undergo a gradual and delayed cell death because of enhanced excitatory neurotransmission for 15 min [71]. Blockers of voltage-sensitive sodium channels or NMDA receptor antagonists reduced the cell death at 24 hr when added either during or after the excitotoxic insult. In contrast, a NOS inhibitor, melatonin, or N-tert-butyl-α-phenylnitrone (a free radical scavenger) did not reduce the cell death at 24 hr when added during excitotoxicity but was effective when added up to 4 hr after the excitotoxic insult [71].

Heavy metals such as cobalt can induce oxidative stress and cytotoxicity. Adding cobalt chloride to the culture medium was shown to induce oxidative stress and cell death with increased release of β-amyloid at 24 hr in cultured SHSY5Y human neuroblastoma cells [72]. Pretreatment with melatonin prevented the deleterious effects of cobalt chloride. In an organotypic mouse brain slice model of excitotoxicity, oxidative stress was documented [73]. Co-treatment with melatonin was found to produce significant antioxidant effects.

Oxygen-glucose deprivation (OGD) is a better in vitro model of stroke than hyperoxia or chemical-induced toxicity [74]. OGD simulates ischemia with the removal of glucose and oxygen in the culture medium as well as oxygen in the atmosphere, and a return to the normal culture conditions simulates reperfusion. Neuroprotective potential of melatonin in apoptotic cell death was explored in primary neuronal cell culture from embryonal Wistar rat fetuses and compared with necrosis induced by OGD for 2 hr [75]. Pretreatment with melatonin reduced the OGD-induced necrosis but did not protect against caspase-dependent, free radical-independent apoptotic neuronal cell death. A recent in vitro study reported that co-treatment with melatonin would enhance the survival of SHSY5Y human neuroblastoma cells at 24 hr after OGD for 1 hr in a dose-dependent manner via mechanisms independent of the mt1 and MT2 membrane receptors [76].

Similar to the in vivo results, in vitro studies have confirmed the beneficial effects of melatonin treatment in hyperoxia, excitotoxicity, cyanide-induced toxicity and OGD. Unlike in vivo situations, it is convenient to control and change the extracellular environment, select specific types of cells, and determine the intracellular events of individual cells in in vitro studies. Thus, in vitro studies are useful in examining the neuroprotective mechanisms.

Melatonin – a neuroprotectant with good potential

Very large daily oral doses of melatonin at 300 mg for 4 months have been shown to inhibit ovulation in women without significant side effects [77]. At pharmacological (and possibly physiological) concentrations, melatonin is a free radical scavenger, a direct antioxidant, and an indirect antioxidant [35, 78, 79]. Free radicals, including NO, are the key mediator of excitotoxicity, calcium ion-mediated toxicity, mitochondrial dysfunctions, inflammatory damage, and apoptosis [19–21, 24, 32, 35, 39, 46, 49, 79, 80]. Melatonin has many properties of an ideal neuroprotectant against ischemia/reperfusion injury: excellent tissue diffusion to achieve adequate local concentrations, no serious toxicity even at high doses, and active against multiple pathophysiological mechanisms (Fig. 1); its beneficial actions are scavenger of reactive oxygen species (including NO) indirect inhibition of NOS activity, suppression of nuclear factor κB activity, and anti-apoptotic effects [35, 39, 43, 53, 60, 68, 78–82].

Concluding remarks

There is convincing evidence from the above literature review that melatonin treatment is highly effective against different in vivo and in vitro models of excitotoxicity or ischemia/reperfusion injury in multiple animal species. Melatonin has excellent oral absorption and tissue bioavailability. Although the serum half-life of exogenous melatonin is about 20–40 min, melatonin is selectively concentrated in the brain and perhaps in various subcellular compartments of the brain [83]. Further experimental studies should be conducted to explore the synergistic or additive protection from combining melatonin with thrombolysis, defibrinogenation or other neuroprotectants.

Clinical scientists interested in discovering new therapies for brain damage following ischemia/reperfusion should consider planning the following trials. First, phase II or III clinical trials should be conducted to test the efficacy of melatonin in patients with acute stroke. Because of its safety record and easy administration via injection or oral administration, melatonin can be offered as a prehospital treatment in combination with other proven therapies. Melatonin alone or in combination with other neuroprotectants may improve the treatment effects, reduce the hemorrhagic risk, and/or extend the treatment time window of acute thrombolysis/defibrinogenation. The second type of clinical trials is to test the efficacy of low dose oral melatonin as a preventive measure to reduce the risk and/or severity of cerebrovascular events.


Work of the author cited in the present review was supported by grants (10202138 and 10203847) from the Committee of Research and Conference Grants, The University of Hong Kong, Hong Kong.