Systematic review of melatonin in cerebral ischemia-reperfusion injury: critical role and therapeutic opportunities

Cerebral ischemia-reperfusion (I/R) injury is the predominant causes for the poor prognosis of ischemic stroke patients after reperfusion therapy. Currently, potent therapeutic interventions for cerebral I/R injury are still very limited. Melatonin, an endogenous hormone, was found to be valid in preventing I/R injury in a variety of organs. However, a systematic review covering all neuroprotective effects of melatonin in cerebral I/R injury has not been reported yet. Thus, we perform a comprehensive overview of the influence of melatonin on cerebral I/R injury by collecting all available literature exploring the latent effect of melatonin on cerebral I/R injury as well as ischemic stroke. In this systematic review, we outline the extensive scientific studies and summarize the beneficial functions of melatonin, including reducing infarct volume, decreasing brain edema, improving neurological functions and attenuating blood-brain barrier breakdown, as well as its key protective mechanisms on almost every aspect of cerebral I/R injury, including inhibiting oxidative stress, neuroinflammation, apoptosis, excessive autophagy, glutamate excitotoxicity and mitochondrial dysfunction. Subsequently, we also review the predictive and therapeutic implications of melatonin on ischemic stroke reported in clinical studies. We hope that our systematic review can provide the most comprehensive introduction of current advancements on melatonin in cerebral I/R injury and new insights into personalized diagnosis and treatment of ischemic stroke.


Introduction
Ischemic stroke is one of the main causes of mortality and the leading cause of adult disability worldwide.It affects the lives of millions of patients and imposes a heavy financial burden on the society (Regenhardt et al., 2018;Zhou et al., 2018;Datta et al., 2020).According to estimates in 2020, the average costs of hospitalization on ischemic stroke per person was amounted to $18,154 and the 3-year follow-up cost was $44,347 in the United States (Yousufuddin et al., 2020).Currently, therapeutic options for ischemic stroke are still very limited and the only drug authorized by FDA is recombinant tissue plasminogen activator (t-PA), which can accelerate clot dissolution and blood reperfusion.However, administration of t-PA is only suitable for about 5% of ischemic stroke patients due to the narrow therapeutic window (Zhou et al., 2018;Powers et al., 2019).Moreover, it may paradoxically result in cerebral I/R injury owing to blood flow restoration, which can lead to reactive oxygen free radical accumulation and thus cause brain edema, neurological deficits as well as individual death through a series of mechanisms, including oxidative stress, neuroinflammation, apoptosis, excessive autophagy, glutamate excitotoxicity, and mitochondrial dysfunction (Duehrkop and Rieben, 2014;Sarmah et al., 2018;Datta et al., 2020).Nowadays, a valid treatment for cerebral I/R injury has not been determined.Despite several substances have shown neuroprotective effects on cerebral I/R injury models, consistent results are not be obtained in further clinical trials.Thus, more potentially effective substances are being actively studied, among which melatonin is the most concerned.
Melatonin (N-acetyl-5-methoxytrytamine), an amine hormone discovered by Aaron Lerner half a century ago, is mostly secreted by the pineal gland and exists in extra-pineal sites including the retina, gastrointestinal tract, and bone marrow as well (Zawilska JolantaB. et al., 2009;Samantaray et al., 2009;Venegas et al., 2012).The timing of melatonin secretion adapts to the light-dark cycle (Zawilska, Skene, and Arendt, 2009a).Therefore, it is originally considered to be responsible for controlling circadian rhythms, including sleepwake rhythm, neuroendocrine rhythm as well as seasonal response, by providing night information and mediating dark signals (Claustrat, Brun, and Chazot, 2005).Moreover, melatonin is involved in lowering core body temperature at night and promotes sleep tendency, whereas its secretion is disrupted upon exposure to light, thus causing wakefulness (Akerstedt et al., 1979).Additionally, melatonin plays important physiological roles in multiple other systems: cardiovascular system, respiratory system, reproductive system, immune system, and endocrine system, etc., For example, recent studies have indicated that melatonin could take participate in the regulation of blood pressure and heart rate via endothelium-dependent vasodilation and sympathovagal autonomic modulation (Girouard et al., 2001;Baker and Kimpinski, 2018).Besides, when combing with melatonin receptors on pancreatic islets, melatonin could induce the production of insulin growth factors, and subsequently promote insulin receptor tyrosine phosphorylation, thereby modulating insulin sensitivity and glucose homeostasis (Sharma et al., 2015).
To date, melatonin is thought to exert various physiological effects primarily through its action on two specific membrane receptors, MT1 and MT2.As members of the seventransmembrane G protein-coupled receptor family, these two MT receptors are widely distributed throughout the central nervous system, including cerebral cortex, hippocampus, cerebellum, midbrain and especially hypothalamus and suprachiasmatic nucleus (Ekmekcioglu, 2006;Ng et al., 2017).Moreover, MT1 receptors are more abundant than MT2 receptors in most brain regions.Recent researches have demonstrated that in the various CNS regions, melatonin may exert multiple neuroprotective effects by direct binding to MT1 and MT2 receptors.In mouse primary neurons, Liu et al. observed that melatonin supplementation could activate Akt/GSK-3β/ CRMP-2 signaling by acting on MT2 receptors, subsequently enhance excitatory synaptic transmission and thus promote functional and synaptic formation (Liu et al., 2015).Similarly, in APP/PS1 transgenic mice, melatonin administration could reverse mitochondrial dysfunction and decrease abnormal Aβ deposition in the cortex, striatum, and hippocampus brain region via MT2 receptor signaling, eventually improving cognitive and behavior deficits (Dragicevic et al., 2011).Additionally, in a transgenic mouse model of Huntington's disease (HD), MT1 receptor levels in mice brain were much lower compared with wild-type controls.However, melatonin treatment could inhibit mutant huntingtin-induced caspase-3 activation and preserve MT1 receptor expression.Furthermore, melatonin would delay disease onset and mortality of HD mice, which was dependent on the presence and activation of MT1 receptor (Wang et al., 2011).
More importantly, emerging studies have confirmed that melatonin also plays an important effect in preventing I/R injury in a variety of organs including heart, liver, and brain (Tai et al., 2010;Kang et al., 2011;Zhang et al., 2019).Zhang et al. observed that melatonin administration contributed to improve mitochondrial fusion/mitophagy through inhibiting I/R-mediated Optic atrophy 1 (OPA1) downregulation, thereby maintaining myocardial function and cardiomyocyte viability in myocardial I/R injury (Zhang et al., 2019).Additionally, Kang et al. confirmed that melatonin could protect against hepatic I/R injury via mitigating Toll-like receptor 3 (TLR3) and Toll-like receptor 4 (TLR4) overexpression, which further inhibited TLR-mediated downstream inflammatory signaling cascades, including myeloid differentiation primary response 88 (MyD88)-dependent and tollreceptor-associated activator of interferon (TRIF)-dependent pathways (Kang, Koh, and Lee, 2011).As for cerebral I/R injury, Tai et al. demonstrated that melatonin reduced endothelial damage and preserved blood-brain barrier (BBB) integrity via attenuating matrix metalloproteinase-9 (MMP-9) protein expression and activity, thereby exerting neuroprotective effects (Tai et al., 2010).Nowadays, growing studies have demonstrated that melatonininduced protective effects exert an important role in the brain.Thus, melatonin is expected to become a latent therapeutic agent in brain I/R injury.However, almost no systematic review that covering all protective effects of melatonin in cerebral I/R injury has been reported.Based on existing evidence, we conduct this comprehensive review aiming to study the neuroprotective effect of melatonin and the underlying molecular mechanisms, and explore the latent clinical application prospects of melatonin in cerebral I/R injury together with ischemic stroke.

Article search strategy process and study selection
A total of 920 articles are identified in PubMed database by using the following terms: melatonin and cerebral ischemia/brain ischemia/cerebral infarction/brain infarction/stroke/ischemiareperfusion.Among them, by screening the titles and abstracts, 503 apparently irrelevant articles are excluded.293 articles are further removed after reviewing the full-text articles for eligibility, among which 142 articles are reviews and metaanalyses, 8 articles are not related to melatonin, 90 articles are  3 Protective effects of melatonin in experimental I/R injury models During ischemic stroke, blood supply for the brain is interrupted, which can cause irretrievable tissue damage.Ischemic tissue injury is commonly attributed to tissue hypoxia and the subsequent cellular ATP depletion, which may in turn cause death of neurons and glia and then lead to the brain tissue infarction as well as neurological deficits.After reperfusion, the generation of reactive oxygen species (ROS) is at an accelerated rate in tissues (Granger and Kvietys, 2015).The central nervous system (CNS) is particularly sensitive to ROS because it contains high concentrations of substances that promote oxidative processes, including iron, ascorbic acid, and polyunsaturated fatty acids, while low amounts of endogenous antioxidants (Watson et al., 2016).ROS accumulation can cause reperfusion injury, including programmed cell death and necrosis in the brain through triggering a cascade characterized by ionic imbalance, mitochondrial disturbances, oxidative stress and endoplasmic reticulum (ER) stress (Murphy and Steenbergen, 2008;Bagheri et al., 2016;Lin et al., 2018).Thus, protecting CNS cells from I/R injury is considered as an important therapeutic strategy for ischemic stroke treatment.For investigations of cerebral I/R injury, the preferred in vivo model is middle cerebral artery occlusion (MCAO), including transient MCAO and permanent MCAO (Lemmerman et al., 2022).Besides, neurons and glial cells subjected to oxygen-glucose deprivation/reoxygenation (OGD/R) are common in vitro models, allowing us to investigate the specific feature of different cell types.
4 The mechanism of melatonin's neuroprotective effect in cerebral I/R injury Melatonin could exert neuroprotective effects through various mechanisms in cerebral I/R injury (Figure 2).

Oxidative stress
Oxidative stress is characterized by excessive ROS and reactive nitrogen species (RNS) production under harmful stimuli.During cerebral I/R injury, ROS overproduction causes immense damage to brain (Manzanero et al., 2013).Compelling evidence suggests that, as a powerful scavenger against ROS/RNS, melatonin can inhibit oxidative stress in cerebral I/R injury via multiple mechanisms, among which nuclear factor E2-related factor-2 (Nrf2) occupies a vital position.
Melatonin can exert neuroprotective effects based on the upregulation of Nrf2 and Nrf2-related antioxidants during cerebral oxidative stress injury.Specifically, melatonin could reduce the ubiquitination of Nrf2 via modifying Kelch-like ECHassociated protein 1 (Keap1) cysteine residues, thereby reducing Nrf2 degradation by proteasome and enhancing Nrf2 phosphorylation and nuclear translocation to the nucleus Frontiers in Pharmacology frontiersin.org(Zhao et al., 2018).In addition, melatonin can promote Ca 2+ influx and induce phosphorylation of protein kinase C (PKC), an upstream mediator of Nrf2, which further leads to the downstream Nrf2 nuclear translocation (Romero et al., 2010).Moreover, melatonin can directly serve as a proteasome inhibitor to decrease the degradation of Nrf2 (Vriend and Reiter, 2014).Nrf2, an antioxidative transcription factor, can positively regulate the expression of a series of antioxidant response element (ARE)-dependent genes.Melatonin induced Nrf2 translocation and upregulation could increase antioxidant enzymes expression levels that comprise glutathione (GSH), glutathione s-transferase (GST), superoxide dismutase (SOD), catalase (CAT), heme oxygenase-1(HO-1) and glutathione peroxidase (GPx), etc (Rodriguez et al., 2004;Ma, 2013;Zhang et al., 2013).These upregulated enzymes effectively scavenge free radicals as well as related reactants and ultimately reduce the cerebral damage during the oxidative stress process (Deng et al., 2015;Liu et al., 2019).
As shown in a former study by Parada et al., in hippocampal cultures exposed to OGD/R, administration with melatonin could induce the Nrf2 nuclear accumulation as well as the combination of Nrf2 and ARE sequences within HO-1 promoter, thereby elevating HO-1 expression and decreasing the production of OGD-induced ROS to basal levels (Parada et al., 2014).During heme metabolic process, HO-1 can reduce the ROS by degrading free heme.Meanwhile, the by-products of heme catabolism such as carbon monoxide (CO) and biliverdin (BV)/bilirubin (BR) can not only directly scavenge OGD-induced ROS production, but also inhibit NADPH oxidase activity, further alleviating oxidative stress (Kim et al., 2011).Additionally, Nrf2α subunit was found to bind with sirtuin-3 (SIRT3) promoter directly, leading to the upregulation of SIRT3 expression.And the elevation of SIRT3 further activated SOD through promoting SOD deacetylation and then reduced mitochondrial oxidative stress (Satterstrom et al., 2015;He et al., 2019).Moreover, another possible mechanism by which melatonin suppresses oxidative stress involves increasing GSH content (Paterniti et al., 2016).Nrf2 can promote GSH biosynthesis enzymes and GSH reductase (GSR) expression, and thus exerts a valid effect in preserving the mitochondrial GSH pool (Ryoo and Kwak, 2018).High concentrations of GSH are oxidized to glutathione disulfide (GSSG) by GPx, thereby reducing peroxides to non-toxic substances and alleviating oxidative stress (Harvey et al., 2009).
In addition to activating Nrf2, studies have also demonstrated that administration with melatonin can play protective effects via inhibiting NADPH oxidase, which can produce ROS upon activation (Wilkinson and Landreth, 2006;Li et al., 2014).Specifically, in a rat MCAO/R model, by hindering phosphatidylinositol 3-kinase (PI3K)/Akt signaling pathway, melatonin can suppress the p47phox subunit phosphorylation, and thus block the combination of p47phox with gp91phox subunit.This process disturbs the assembly and activation of Protective mechanisms of melatonin in cerebral ischemia-reperfusion injury.

Neuroinflammation
Neuroinflammation is the inflammatory response concentrated in the CNS and might be initiated from ischemia.It is featured as the accumulation of pro-inflammatory factors, comprise IL-1β, IL-6, TNF-α and inducible nitric oxide synthase (iNOS) that are generated from activated microglia and astrocytes after ischemic stroke (Jin et al., 2010;DiSabato et al., 2016;Yang and Zhou, 2019).Melatonin can hinder microglia and astrocytes phenotype switching through multiple mechanisms, further attenuating neuroinflammation.
During the I/R process, two different forms of microglia are involved, including pro-inflammatory microglia (M1) and antiinflammatory microglia (M2) phenotypes.Under the ischemic conditions, M2 phenotype transforms into the active M1 phenotype, which promotes further pro-inflammatory cytokines release and eventually accelerates stroke-induced secondary brain injury (Jin, Yang, and Li, 2010).The switch and balance between the M1/M2 phenotype can be modulated by melatonin through various mechanisms, including inhibiting nuclear factor kappa-B (NF-κB), activating signal transducer and activator of transcription 3 (STAT3) and interaction with MT2 receptor.Detailly, Zhao et al. demonstrated that in MCAO/ R rats, melatonin could inhibit TLR4 expression, further hindering downstream NF-κB activation.NF-κB inactivation can initiate the transformation of microglia from M1 to M2 phenotype, thereby down-regulating the release of M1 phenotype-related proinflammatory factors (Y.Zhao et al., 2019;J. Sun et al., 2019;Ji et al., 2017).In addition, treatment with melatonin was also confirmed to significantly upregulate the expression of phosphorylated STAT3 in microglial cells (Z.J. Liu, Ran, et al., 2019).The phosphorylation of STAT3 has been demonstrated to exert a crucial effect in IL-10-induced anti-inflammatory effects.Specifically, removing the docking sites of STAT3 from IL-10R1 would render the receptor unable to transduce the signal for suppressing cytokine synthesis.Thus, the activated STAT3 may contribute to the anti-inflammatory cytokines expression which induces microglia transforming from M1 to M2 phenotype and further inhibits neuroinflammation (Williams et al., 2004).Besides, in a rat MCAO/R model, melatonin was found to act on MT2 and increase the ratio of triggering receptor expressed on myeloid cells 2 (TREM2)/iNOS, a marker for the transformation of M1 to M2, which eventually resulted in the suppression of neuroinflammation (Azedi et al., 2019).
Melatonin also exerts anti-inflammatory effect through suppressing the proliferation and activation of astrocytes.Regarding astrocytes, they are involved in BBB formation and maintenance, synaptogenesis, neurotransmission, and metabolic regulation.Under ischemic conditions, cytokines produced by activated microglia cause astrocyte reactivity hyperplasia and transform into neurotoxic reactive astrocytes, which further result in cerebral damage (Liu and Chopp, 2016;Liddelow et al., 2017;Yang et al., 2020).Chen et al. reported that administration with melatonin attenuated TRIF expression, an adapter molecule of TLRs, and ultimately prevented the conversion of astrocytes from anti-inflammatory (A2) to pro-inflammatory (A1) phenotype (Chen et al., 2020).In addition, in rats subjected to MCAO/R, melatonin administration could effectively decrease glial fibrillary acid protein (GFAP) expression, C3, and S100A10, suggesting that melatonin inhibited reactive astrogliosis and A1 astrocyte polarization (Yawoot et al., 2022).Besides, melatonin was also shown to reduce astrocytemediated inflammatory response by inhibiting glycogen synthase kinase-3 beta (GSK-3β) expression levels and receptor-interacting serine/threonine-protein 1 kinase (RIP1K) activities, consequently enhanced axonal regeneration and promoted neurobehavioral recovery (Yawoot et al., 2022).Mechanistically, suppression of GSK-3β can reduce NF-κB nuclear translocation and upregulate cyclic AMP response element-binding protein (CREB) transcription.These cause a relative increase in CREB compared to NF-κB in the nucleus, resulting in increased binding of CREB Binding Protein (CBP) with CREB and decreased interaction between CBP and NF-κB p65.The increased combination of CBP and CREB produces high levels of IL-10, a critical antiinflammatory factor (Maixner and Weng, 2013).Meanwhile, the reduced interaction of CBP with NF-κB p65 inhibits the proinflammatory factors, and thus alleviates neuroinflammation (Martin et al., 2005).

Apoptosis
Apoptosis is an ATP-dependent death program characterized by fragmentation of chromosomal DNA, degradation of cytoskeletal, and generation of apoptotic bodies (Elmore, 2007;Xu et al., 2019).Regarding the apoptosis pathways, two major patterns are widely accepted including the extrinsic pathway and the intrinsic pathway.The extrinsic pathway refers to receptor-mediated apoptosis, and the activation of death receptors involves in two main ligands, including tumor necrosis factor (TNF) and Fas.Ischemia can induce the combination of the ligands with death receptors which lead to apoptosis and eventually cerebral injury.While in the intrinsic pathway, ischemia could directly induce the generation of intracellular signals and the subsequent formation of mitochondrial permeability transition pore, induce cytochrome C release into cytoplasm and activate caspase-3, and eventually lead to apoptosis (Broughton et al., 2009).In addition, members of Bcl-2 protein family, such as Bim, Bad, Bid, Bcl-x (L), and Bcl-2, are responsible for regulation of apoptosis via modifying the permeability of mitochondrial membrane (Cory and Adams, 2002;Elmore, 2007).
Previous studies have reported that melatonin could play antiapoptotic functions during cerebral I/R injury through multiple biological mechanisms, including activation of the PI3K/Akt, Raf/ MEK/extracellular-regulated kinase (ERK) signaling pathway as well as inactivation of Jun N-terminal kinases (JNK) pathway (Koh, 2008c).In 2008, Koh et al. elucidated that pretreatment with melatonin in the rat model of MCAO-induced cerebral ischemia could significantly reduce the infarct volume by activating Akt.Specifically, melatonin administration can promote the phosphorylation of specific residues on forkhead rhabdomyosarcoma transcription factors (FKHR) by activating PI3K/Akt.Then, 14-3-3 protein, an anti-apoptotic factor, would directly interact with the phosphorylated FKHR and anchor the phosphorylated FKHR within the cytoplasm.These FKHR are blocked from further translocation into the nucleus, therefore inhibiting downstream target genes transcription, including Fas ligand (Koh, 2008d).In 2021, Ran et al. elucidated that both in MCAO/R mice model and N2a cells exposed to OGD/R, treatment with melatonin could significantly reduce the brain infarct volume and neuronal cells apoptosis by activating Akt.Specifically, melatonin administration can decrease phosphatase and tensin homolog (PTEN) deleted on chromosome 10 activity through promoting its phosphorylation.PTEN is a major up-stream negative regulator of the PI3K/Akt signal transduction which can promote the dephosphorylation of PIP3.In the presence of melatonin, phosphorylated PTEN levels were increased, thereby decreasing PTEN activity and activating PI3K/Akt pathway (Ran et al., 2021).Similarly, in a MCAO/R mouse model, melatonin was shown to activate the phosphorylation of Akt at Thr308 via PI3K/ pyruvate dehydrogenase kinase-1 (PDK-1) signaling, which further inactivate GSK-3α/β, two critical players in the activation of apoptosis (Kilic et al., 2017).
Besides, in the cerebral MCAO model, melatonin was found to increase the phosphorylation of pro-apoptotic protein Bad through activating PI3K/Akt, further preventing Bad from interacting with Bcl-x(L) and blocking the activation of downstream apoptosis pathway (Koh, 2008b).
Apart from activating PI3K/Akt pathway, Koh et al. also demonstrated that in a rat MCAO model, treatment with melatonin can protect brain tissue from ischemic injury through Raf/MEK/ERK signaling pathway by modulating the pro-apoptotic protein phosphorylation.To some extent, melatonin can increase the phosphorylation of Raf-1 as well as MEK.And the activation of these two proteins kinases can phosphorylate ERK1/2 which results in the phosphorylation of its downstream target 90 ribosomal S6 kinase (90RSK) (Koh, 2008a).The phosphorylated 90RSK may further phosphorylate the Ser112 residue of Bad and inactivate the pro-apoptotic effect of Bad (Zha et al., 1996;Koh, 2008a).Moreover, Chen et al. reported that melatonin can effectively suppress fork-head box O3a (FoxO3a) activity by inhibiting JNK, thereby hindering the binding of FoxO3a to Bim promoters and the subsequent expression of Bim proapoptotic protein (Chen et al., 2022).Besides, JNK has been shown to regulate cytochrome C release from mitochondria through inducing Bid cleavage.Melatonin could attenuate JNK activation, thereby hindering cytochrome C release and suppressing cytochrome C-induced apoptosis cascades (Madesh et al., 2002;Song et al., 2014).

Autophagy
Autophagy is a lysosome-mediated self-digestion and recycling process that contributes to clearing folded or aggregated proteins and promoting the degradation of damaged organelles.It plays a key role in preserving cellular homeostasis under physiological conditions (Glick et al., 2010;Boga et al., 2019).However, under pathological circumstances including cerebral I/R injury, ischemiamediated excessive autophagy activation may be harmful and further lead to cell death.Autophagy begins with nucleation of the phagophore induced by the unc-51-like autophagy activating kinase 1/2 (ULK1/2) complex.Then, the phagophore elongates with the aid of autophagy-related (ATG) proteins and finally the expanding membrane forms an autophagosome which closes around its cargo.The outer membrane of the autophagosome subsequently fuses with lysosomal membrane, and further generate an autolysosome, lead to autophagic cargo degradation (Itakura and Mizushima, 2010;Wong et al., 2013;Parzych and Klionsky, 2014).
Accumulating evidences have demonstrated that melatonin could promote physiological autophagy, while inhibiting excessive autophagy to exert neuroprotective effects.On the one hand, in hippocampal HT22 cells exposed to OGD/R, melatonin was shown to increase the expression of both Ras-related protein 7 (Rab7) and transcription factor Forkhead box class O1 (FoxO1), subsequently promoting autophagosome maturation and attenuating ischemiclike injury (Luchetti et al., 2022).Furthermore, both in a MCAO/R model and HT22 cell line exposed to OGD/R, treatment with melatonin upregulated SIRT1, a histone deacetylase that can cause the deacetylation of brain and muscle ARNT-like protein 1 (BMAL1) and decrease its degradation.The upregulation of BMAL1 could promote the expression of ATG14 through directly binding to the E-box elements in the ATG14 promoter, subsequently increasing the level of Beclin-1 and the ratio of LC3II/ LC3I, enhancing autophagy and improving neurological function (Liu et al., 2021).On the other hand, melatonin could activate mammalian target of rapamycin complex1 (mTORC1) through multiple pathways, thereby hindering excessive autophagy and exerting neuroprotective effects.Specifically, in a rat MCAO/R model, melatonin was observed to decrease the phosphorylation of tuberous sclerosis complex 2 (TSC2) through PI3K/Akt signaling pathway.Therefore, the downstream protein complex TSC1/ TSC2 was disrupted, leading to stabilization of the downstream Rheb-GTPase and subsequent activation of mTORC1 (Yang and Klionsky, 2010;Zheng et al., 2014).The mTORC1 can inactivate the autophagy initiators ATG13 and ULK1/2 through combination and phosphorylation, regulate the class III PI3K complex, and thus exerts an important effect in inhibiting autophagy (Rabanal-Ruiz, Otten, and Korolchuk, 2017).Likewise, mTORC1 can also phosphorylate transcription factor EB (TFEB) and TFE3 which facilitates the interaction between these two transcription factors and cytosolic chaperone 14-3-3 and retains them in cytoplasm (Martina et al., 2012).Therefore, the expression of several autophagy-related genes including PIK3C3 and ATG16L1, etc., are inhibited, thereby suppressing excessive autophagy and reducing cerebral injury (Settembre et al., 2011).
In addition to activating mTORC1 signaling pathway, melatonin could also attenuate excessive autophagy through hindering JNK signaling pathway, according to previous studies.Melatonin can inactivate JNK1, which further inhibits the separation of Bcl-2 from Beclin-1.Beclin-1 acts as a part of class III PI3K complex, and the combination of Bcl-2 and Beclin-1 can hinder class III PI3K complex activation, thereby inhibiting autophagy (H.D. Xu and Qin, 2019;Zheng et al., 2014).Similarly, in a rat MCAO/R model, Feng et al. reported that pretreatment with melatonin could inhibit pancreatic ER kinase (PKR)-like kinase (PERK)/Inositol-requiring enzyme 1 (IRE1) pathway, further suppressing the activation of the JNK, subsequently preventing the uncoupling of Bcl-2 with Beclin-1 and excessive autophagy (Feng et al., 2017).Moreover, in a rat MCAO/R model and PC12 cell OGD/R model, pretreatment with melatonin was shown to inhibit excessive autophagy through upregulating miR-26a-5p expression and then downregulating downstream neuron-restrictive silence factor (NRSF) expression (Yang et al., 2020).

Glutamate excitotoxicity
Glutamate acts as the predominant excitatory neurotransmitter in the CNS (Teichberg et al., 2009).During cerebral I/R injury, ischemia can lead to the excessive glutamate release, which triggers the rapid influx of calcium into the cell cytoplasm through the overstimulation of N-methyl-D-aspartate receptors (NMDARs), thereby resulting in a succession of harmful signaling cascades and subsequent CNS cells death (Tehse and Taghibiglou, 2019;M. Zhou and Baudry, 2006;Sattler and Tymianski, 2000).Moreover, glutamate can also cause the overproduction of nitric oxide (NO) and the decrease of downstream mediator protein phosphatase 2A (PP2A) which result in neurotoxicity and cerebral injury (Koh, 2012a;Jia et al., 2015).Melatonin can reduce glutamate excitotoxicity during cerebral I/R injury through modulating glutamate as well as its receptor pathways.
Detailly, Patino et al. confirmed that in rat hippocampal slices exposed to OGD/R, melatonin could directly reduce the release of glutamate and prevent a persistent activation of NMDARs, thereby reducing intracellular Ca 2+ levels (Patino et al., 2016).Similarly, in hippocampal slices exposed to OGD/R, melatonin reduced a surge of synaptic glutamate release and neuronal cell death after ischemiareperfusion, which further strongly suppressed the increase in the intracellular Ca 2+ concentration by downregulating the NMDAR activity (Furuta et al., 2022).Besides, in hippocampal HT22 cell lines exposed to glutamate excitotoxicity, melatonin pretreatment could prevent glutamate-mediated reduction in calcium-buffering proteins such as parvalbumin and hippocalcin and thus downregulate Ca 2+ levels (Koh, 2012b).Similarly, in Sprague-Dawley rat model subjected to MCAO, melatonin has also been documented to reduce intracellular Ca 2+ levels through increasing the expression of parvalbumin and hippocalcin (Koh, 2012b).Additionally, melatonin can directly bind with calreticulin and further reduce the level of Ca 2+ by hindering the release of Ca 2+ from endoplasmic reticulum, in rat C6 astroglial cells exposed to glutamate excitotoxicity (Das et al., 2008;Venkatesan et al., 2021).With the decrease of Ca 2+ levels, calpain expression and caspase-3 activation are inhibited.Calpain inhibition subsequently reduces the degradation of cytoskeletal proteins as well as axon-myelin structural unit, thus maintaining the structural integrity of CNS cells and preventing glutamate excitotoxicity (Samantaray et al., 2008;Doshi and Lynch, 2009).
Moreover, in a MCAO rat model, melatonin was found to interact with calmodulin, further preventing the binding of calmodulin to NOS, and thus inhibited calmodulin-dependent NOS activation (Koh, 2008e).Apart from inhibiting NOS activity, Koh et al. suggested that in a rat focal cerebral ischemia model, treatment with melatonin could mitigate glutamate-induced decrease of PP2A subunit B, a critical subunit in facilitating various functions of PP2A.As an essential protein phosphatase, PP2A can promote DNA repair, cellular proliferation and differentiation.From this point of view, melatonin can maintain PP2A levels, thereby promoting neuronal survival and alleviating glutamate excitotoxicity (Koh, 2012a).

Mitochondrial dysfunction
Mitochondria serve as essential organelles that exert a crucial effect in maintaining energy metabolism and cellular homeostasis (Anderson and Sims, 1999;Lesnefsky et al., 2017).Under normal circumstances, mitochondria produce ATP through the electron transport in the respiratory chain to keep cellular function and integrity (Newmeyer and Ferguson-Miller, 2003).Meanwhile, mitochondria could release pro-apoptotic factors, and thus control cell survival (Ott et al., 2002).However, under pathological conditions such as cerebral I/R injury, ischemia can lead to mitochondrial dysfunction, including impaired ability to generate ATP, accumulation of ROS, transition of the mitochondrial permeability and increased release of pro-apoptotic factors (Sims and Anderson, 2002;Ham and Raju, 2017).Melatonin has been found to play neuroprotective effects by improving ischemiainduced disturbance in mitochondrial redox state, fusion and fission, biogenesis, mitophagy and mitochondrial transfer.
Firstly, as a potent free radical scavenger, melatonin has the capacity to directly counteract mitochondrial oxidative injury.In a MCAO/R mice model, melatonin pretreatment significantly increased SIRT1 expression and thus reduced the expression of acetylated p53 and NF-κB.Subsequently, the deacetylated p53 and NF-κB maintained mitochondrial membrane potential, elevated the activity of mitochondrial Complex I and reduced mitochondrial ROS levels (Yang et al., 2015).Apart from alleviating mitochondrial oxidative damage, in N2a neuroblastoma cells exposed to OGD/R as well as a MCAO/R rat model, melatonin was shown to upregulate OPA1 expression, one of the mitochondrial fusion-related proteins, by activating Hippo/Yap pathway.Then, the mitochondrial cristae junctions are tightened and cytochrome c release is restricted, thereby improving OPA1-related mitochondrial fusion and inhibiting mitochondrial fission (Wei et al., 2019;Nasoni et al., 2021).Moreover, Nasoni et al. reported that in hippocampal HT22 cells with excessive mitochondrial oxidative stress, melatonin was found to enhance the levels of peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC-1α) expression.Subsequently, PGC-1α can further enhance the combination of estrogen-related receptor (ERR) α with ERR binding element.Thus, SIRT3 expression is initiated which further induces mitochondrial biogenesis through activating nuclear respiratory factors-1 (NRF-1) and nuclear respiratory factors-2 (NRF-2) and subsequently increasing expression of mitochondrial transcription factor A (mtTFA) (Kong et al., 2010;Nasoni et al., 2021).Besides, in HT22 cells exposed to OGD/R, melatonin was shown to promote mitochondrial transfer and reshape mitochondrial network via increasing the number of tunneling nanotubes (TNTs).TNTs are dynamic structures that connect cells, and intercellular mitochondria could transfer through these structures, which further compensate for damaged organelles and promote cell recovery (Nasoni et al., 2021).Lastly, in HT22 neuronal cell line with glutamate injury, melatonin could upregulate the levels of Bcl-2 expression and downregulate the levels of Beclin-1 expression, thereby further reducing glutamate-induced mitophagy and restoring mitochondrial function (Wang et al., 2019).In addition, in PC12 cell line treated with ropivacaine, melatonin was shown to reduce the expression of two mitophagy-related proteins, PTEN-induced kinase I (PINK1) and Parkin, which may hinder the occurrence of mitophagy (Yu et al., 2018;Wang et al., 2019;Zeng et al., 2022).

Clinical prospects
In recent years, numerous clinical investigations exploring the influence of melatonin on ischemic stroke and the sequelae of ischemic stroke have been reported.Firstly, a series of preclinical studies have shown melatonin was an extremely safe neurotherapeutic agent even at high concentrations (Romero et al., 2016).So far, melatonin has been applied to treat a variety of diseases, involving sleep disorders, depression, etc., and no clinical studies on melatonin therapy have shown serious side effects (Cardinali et al., 2012;Hassell et al., 2015;Riha, 2018).Secondly, pineal calcifications diagnosed by applying multi-spiral computer and/or magnetic resonance tomography has been described as having a significant impact on the risk of ischemic stroke (Kitkhuandee et al., 2014).Besides, subsequent studies confirmed a positive relationship between pineal calcifications and decrease of melatonin synthesis (Schmid et al., 1994).Thus, human clinical trials of ischemic stroke to evaluate potential clinical applications of melatonin seemed to be necessary.Current studies have found that endogenous melatonin, mainly produced by the pineal gland, might be a good predictor for the prognosis of ischemic stroke.In a prospective observational study published in 2018, Lorente et al. found that in patients with severe middle cerebral artery infarction, serum melatonin concentrations were associated with total antioxidant capacity as well as malondialdehyde levels (used to assess lipid peroxidation).Furthermore, results of this research also illustrated a positive correlation between serum melatonin levels and the severity or even mortality of stroke patients during the 30-day follow-up period (Lorente et al., 2018).For this point of view, we speculated that patients with more severe ischemic stroke would produce high levels of ROS, which might result in higher concentrations of serum melatonin to compensate for the increase in oxidation products.Among them, when the attempts to maintain a balance between oxidation and antioxidant status were insufficient, the patients with severe ischemic stroke would eventually die.Additionally, a case-control study involving 42 patients with ischemic stroke demonstrated that the levels of melatonin in the urine was remarkably decreased in contrast to the control value.It was reasonable to suspect that due to the excessive free radical production, the catabolism of melatonin might be speeded up during acute ischemic stroke, indicating the vital role of melatonin in scavenging free radicals and neuroprotection (Ritzenthaler et al., 2013).
A previous systematic review focusing on clinical trials of melatonin on various brain injury provided evidence that melatonin treatment improved sleep disturbance following traumatic brain injury and increased survival rate in intubated patients with hemorrhagic stroke and asphyxiated newborns (Ramos et al., 2020).Regarding the clinical potential of exogenous melatonin therapy on cerebral I/R injury and ischemic stroke, a double-blind randomized controlled trial recruited 60 patients, who orally received melatonin at 6 mg/day for 3 days before and after carotid endarterectomy, to explore the therapeutic effect of melatonin on cerebral I/R injury (Yang et al., 2018).The circulating levels of inflammatory cytokines in patients of the melatonin group were significantly reduced, implying that melatonin could decrease the inflammatory damage and ameliorate subsequent brain I/R injury.Furthermore, Mehrpooya et al. recruited 65 patients with acute ischemic stroke who were not eligible for reperfusion treatment in a double-blind placebocontrolled trial.They observed that melatonin supplementation within 24 h after stroke at a dose of 20 mg once daily for 5 days had a higher reduction in modified Rankin Scale score (mRS) at 30 and 90 days after treatment, indicating melatonin had beneficial effects on neurological recovery after ischemic stroke (Mehrpooya et al., 2022).As one of the holy grails for acute ischemic stroke treatment, neuroprotection could theoretically improve the neurological disability of stroke survivors, which makes it attract widespread attention.According to previous studies, therapeutic hypothermia can maintain organ vitality and is one of the most powerful neuroprotective treatments (Perlman et al., 2010;Kurisu and Yenari, 2018).Although no clinical researches involving in the neuroprotective effect of melatonin combined with hypothermic therapy in cerebral I/R injury treatment following ischemic stroke, a clinical trial recruiting 30 newborns with confirmed hypoxicischemic encephalopathy demonstrated that early use of melatonin combined with hypothermia could lower plasma free radicals generation, protect against the subsequent CNS injury and ultimately improve survival rate of patients (Aly et al., 2015).Additionally, as a common and severe complication of ischemic stroke, poststroke delirium (PSD) has attracted wide attention from researchers due to it was related to increased mortality, longer hospital stays, and lower functional outcome (Siddiqi et al., 2006;Shaw et al., 2019).A propensity score-matched analysis involving 573 patients with acute ischemic stroke showed that patients who received prophylactic treatment with melatonin within 24 h after stroke onset had a lower risk of PSD compared with patients receiving standard treatment (Mengel et al., 2021).Although the underlying mechanism of melatonin for preventing PSD was still unclear, the evidence of this propensity score-matched analysis implied that supplementation with melatonin could reduce PSD susceptibility, and further supported the neuroprotective effect of melatonin on acute ischemic stroke.Taken together, recent clinical studies have partially clarified the predictive and therapeutic potential of melatonin for ischemic stroke.Nevertheless, further studies, ideally enrolling more clinical subjects, are still needed to evaluate the clinical safety and effectiveness of melatonin to better understand its medical effects in cerebral I/R injury and ischemic stroke.

Conclusion
In the past decades, ischemic stroke has been one of the most threatening diseases for human health.Due to its high morbidity and mortality, efforts have been made for detecting appropriate Frontiers in Pharmacology frontiersin.org13 alternative or complementary drugs on ischemic stroke therapy.Currently, accumulated lines of evidence suggested that melatonin could reduce cell death and increase cell viability, and display an obvious neuroprotective effect on various CNS cells in vitro.Meanwhile, experimental stroke models in vivo have confirmed that melatonin would reduce infarct size and brain edema, and improve neurological function.Subsequently, we also outlined the extensive studies and performed a comprehensive introduction of various functions of melatonin in cerebral I/R injury, including regulating oxidative stress, neuroinflammation, apoptosis, autophagy, glutamate excitotoxicity and mitochondrial dysfunction.Meanwhile, some key signaling pathways has been demonstrated to play important roles in melatonin neuroprotection, such as Nrf2, NF-κB, PI3K/Akt, Raf/MEK/ERK, JNK, and Hippo/ Yap pathways.In future studies, the biomolecular mechanism of melatonin in cerebral I/R damage still needs in-depth research, which is necessary for its clinical application in ischemic stroke.
Given its multiple roles against cerebral I/R damage, it is not surprisingly that melatonin has been considered as a promising candidate, both as a diagnostic biomarker for stroke prognosis and as a drug target for stroke therapy.Recent studies have demonstrated a potential relationship between circulating melatonin levels and the severity and mortality of stroke patients.Regarding exogenous melatonin therapy, oral melatonin in patients undergoing carotid endarterectomy could reduce serum pro-inflammatory cytokines levels and alleviate cerebral I/R injury to a certain extent.Moreover, early treatment with melatonin following stroke may help improve functional recovery in patients ineligible for reperfusion therapy.However, there are still some limitations to the clinical application of melatonin, and some potential issues should be considered in future clinical trials.First, high-quality randomized controlled trials with larger sample sizes and more ethnic groups are needed to further explore the clinical efficacy and side effects of melatonin in patients with ischemic stroke.Second, inappropriate treatment time windows, insufficient dosage or treatment duration may be responsible for the lack of positive results in clinical trials.Thus, the dosage, timing, and duration of melatonin require further reconsideration to determine the optimal application strategy in stroke patients.Third, melatonin might be more effective when combined with other types of stroke treatments, such as hypothermia or hyperbaric oxygen therapy, and the effectiveness of combination therapy should be fully investigated in the future.
In summary, based on the information reviewed above, as a low toxicity and well-tolerated agent, melatonin supplement was supposed to be a new and prospective treatment method for cerebral I/R injury.More randomized and multiple-center clinical trials were need to be designed to confirm predictive and therapeutic role of melatonin in ischemic stroke.

FIGURE 1
FIGURE 1Flow diagram of study identification and selection.

TABLE 1
Protective effects of melatonin on in vitro cerebral I/R injury model.

TABLE 1 (
Continued) Protective effects of melatonin on in vitro cerebral I/R injury model.

TABLE 2
Protective effects of melatonin on in vivo cerebral I/R injury model or ischemic stroke model.

TABLE 2 (
Continued) Protective effects of melatonin on in vivo cerebral I/R injury model or ischemic stroke model.

TABLE 2 (
Continued) Protective effects of melatonin on in vivo cerebral I/R injury model or ischemic stroke model.