Background

Stroke is a group of diseases with brain cells death and loss of related neurologic function due to an acute interruption of blood flow or a sudden rupture of blood vessels to the brain, including ischemic and hemorrhagic strokes. Ischemic stroke is the most common type of stroke, accounting for 87% of all stroke patients. At present, stroke ranks the fifth leading cause of death and one of the main causes of disability according to American Heart Association (AHA), increasing the burden on families and communities. Its high morbidity and disability are due to primary and secondary brain damage, including neuronal death, oxidative stress damage, and brain edema [1].

Lack of blood supply deprives neurons of essential glucose and oxygen, disturbing neuronal homeostasis, which leads to oxidative stress, inflammatory response, apoptosis, necrosis, and other pathological processes and culminates in cell death [2]. To date, reperfusion therapy, which aims to restore blood flow and oxygen before neuronal damage, is still the most effective treatment for acute ischemic stroke. Tissue plasminogen activator (tPA) is the only US Food and Drug Administration approved thrombolytic agent for patients with acute ischemic stroke [3]. This therapy has led to a significant reduction in the mortality rate from ischemic stroke; however, there is an optional treatment time window of only 4.5 h of current tPA therapy, and patients also need to meet multiple criteria [4]. Moreover, it is observed that tPA given outside the therapeutic window could increase the risk of intracranial hemorrhage, which may further damage the brain tissues [5]. In addition, tPA increases the neurotoxicity induced by hemoglobin and causes neuronal cells damage [6]. Therefore, it is significant and urgent to find potential therapeutic targets for ischemic stroke.

There is increasing evidence that protective mitochondrial function is essential to improve neuronal survival and neurological function after ischemic stroke [7,8,9]. When an ischemic stroke occurs, mitochondrial dysfunction impairs energy generation, increases reactive oxygen species (ROS) production, and stimulates cytochrome C (Cyt c) release into the cytosol to induce cell death. Although reperfusion can restore blood flow to salvage brain tissue, it may also cause mitochondria to produce excessive ROS, exacerbating cell damage [10]. Therefore, control of the quality and quantity of mitochondria is vital to protect neurons from the pathological effects of dysfunctional mitochondria. In this regard, timely and efficient elimination of damaged mitochondria by autophagy, termed mitophagy, seems to be essential for cell survival in ischemic stroke. Therefore, the development of novel drugs targeting mitochondria may be a promising treatment option for ischemic stroke.

This review summarizes recent advances on the role of mitochondria in ischemic stroke, the understanding of mitochondrial structure and function, and the crucial role of mitophagy in ischemia and reperfusion injury. We discussed the therapeutic potential of mitophagy regulation and evaluated the potential clinical applications of several mitophagy-related drug candidates, which provide valuable insights for clinical treatment.

Mitochondrial Quality and Quantity Control with Ischemic Stroke

As a central cell organelle, mitochondria are involved in numerous cellular biological processes ranging from the synthesis of ATP to being intimately involved in programmed cell death, comprised of at least six compartments: outer membrane, inner boundary membrane, intermembrane space, cristal membranes, intracristal space, and matrix. Among them, the outer membrane is permeable for ions and small molecules, while the inner mitochondrial membrane is almost impermeable, which is equipped with various ion channels and transporters. Several important enzymes, such as creatine kinase and Cyt c, are located in the intermembrane space. Notably, mitochondria are not static. In contrast, increasing evidence has shown that mitochondria constantly undergo fission and fusion, forming either interconnected mitochondrial networks or separated fragmented mitochondria, referred to as mitochondrial dynamics [11]. Their transient and rapid morphological adaptations are crucial for mitochondrial quality control and maintaining an effective adaptation of the mitochondrial compartment to the cellular metabolism [12].

Mitochondria dynamics and mitophagy co-operate in controlling the quality and quantity of mitochondria to maintain function when facing external and internal stress [13]. In addition, these alterations have been observed to participate in several pathology processes, including neurodegenerative diseases [14, 15].

Mitochondria Dynamics and Ischemic Stroke

The fission is the process where a single mitochondrion is divided into two independent daughter mitochondria; it mainly depends on dynamin-related protein 1 (Drp1), mitochondrial fission factor (MFF), and mitochondrial fission protein 1 (Fis1) [16]. Current evidence suggests that Drp1 is recruited to the outer mitochondrial membrane (OMM) by Fis1 during pathological fission, whereas MFF plays a predominant role in recruiting Drp1 during physiological fission [17]. Furthermore, once Drp1 gets onto the OMM, it multimerizes and forms a rings structure and is gradually compressed until the mitochondria rupture [18]. Meanwhile, mitochondrial fusion is regulated at both mitochondrial membranes: OMM and the inner mitochondrial membrane (IMM). Mitofusin (Mfn) 1 and Mfn2 are GTPases required to mediate fusion of the outer membrane by forming homodimers or heterodimers [19,20,21]. The protein optic atrophy 1 (Opa1) mainly participated in IMM fusion [22, 23]. The process of mitochondria fusion is observed to run in a certain order: (1) mitochondrial trans tethering; (2) mitochondrial outer membrane fusion; and then (3) mitochondrial inner membrane fusion [24].

Disturbance of mitochondrial fission and fusion is an essential phenomenon in cerebral ischemia/reperfusion (I/R) injury. Studies have indicated that mitochondrial fission elevated neuronal death after ischemic stroke [25]. Meanwhile, advanced studies have shown that Drp1 inhibition rescued superoxide production, which could attenuate the severity of cerebral ischemic injury, and inhibition of Fis1 could be protective to the ischemic brain [26, 27]. Furthermore, deletion of mitochondrial uncoupling protein 2 exacerbated cerebral I/R injury by breaking mitochondrial dynamic balance towards fission and mitophagy [28, 29]. In contrast, during I/R injury, the expression of Mfn2 and Opa1 was significantly reduced [30,31,32].

Mitophagy

Autophagy is a biological process by which cells can degrade their unnecessary or dysfunctional cytosolic contents through autophagosome and delivery into the lysosomes [33]. Mitophagy, a special form of autophagy occurring in the mitochondrion, is a crucial process that can selectively remove superfluous or dysfunctional mitochondria, to maintain mitochondrial quality and function [34]. Over the past few years, the important role of mitophagy in cancer [35], neurodegenerative diseases [36], and metabolic disorders [37] have gradually been reported. For example, the accumulation of dysfunctional mitochondria due to impaired mitophagy was observed in Alzheimer’s disease [38].

Mitophagy starts with the formation of phagophore, which is a cup-shaped double-membrane structure from the endoplasmic reticulum (ER). The phagophore then is triggered by the interaction between ubiquitinated substrates and LC3-membranes, thereby engulfing damaged mitochondria for degradation. Mitochondria are mainly responsible for energy production. Mitophagy is damaged under internal and external stress accompanied by ROS and other oxidants into the cytoplasm, such as H2O2 and peroxynitrite, which could be harmful to the proteins, nuclear acid, and membranes [39]. Meanwhile, when mitochondria face severe damage, Cyt c, a mitochondrial intermembrane space protein, will be released and finally trigger a caspase cascade and apoptosis [40]. Therefore, appropriate mitophagy is necessary to depredate damaged mitochondria for cell survival.

More specifically, lack of oxygen and nutrients will lead to the depolarization of mitochondria, thereby inducing ROS production, reducing ATP production, mediating the accumulation of putative kinase 1 (PINK1), consequently triggering mitophagy [41]. Under ischemic injury, there is a lack of glucose and oxygen, oxidative phosphorylation is thus inhibited, and cell function is seriously impaired, which eventually leads to cell death [42]. After ischemic injury, the lack of energy causes the accumulation of ER and extracellular free Ca2+, which is then transferred to the mitochondrial matrix through the associated mitochondrial membrane [43]. The accumulation of free Ca2+ in the mitochondrial matrix leads to activating and opening the mitochondrial permeability transition pore (mPTP) (Fig. 1) [44]. Furthermore, Cyt c and other pro-apoptotic factors pass through the opened mPTP and eventually lead to cell apoptosis [45].

Fig. 1
figure 1

Mitochondrial consequences of ischemia–reperfusion. Under conditions of ischemia, mitochondria are depolarized (i.e., ΔΨm is decreased) and ATP stores are depleted. Currently, the mPTP remains closed and the OMM keeps intact. Reintroduction of oxygen results in the rapid normalization of PH, mitochondrial ROS overproduction, exacerbated calcium overload, and opening of the mPTP

Full size image

Interplay Between Mitophagy and Mitochondria Dynamics

Fission and fusion are two independent processes but also interfere with each other. Depending on the quality of mitochondria, when one process is down-regulated, the frequency of another process will increase [46]. There are interactions and mutual regulation between mitochondrial dynamics and mitophagy, which are essential for maintaining mitochondrial homeostasis and function.

In general, fission is characterized by the division of mitochondria into polarized and depolarized mitochondria. Depolarized mitochondria which include damaged proteins and mitochondrial DNA are removed by mitophagy. Consequently, fission and mitophagy co-operate in segregating and eliminating dysfunctional mitochondria. Meanwhile, fission could preserve the healthy part of mitochondria and reduce unnecessary loss during mitophagy [47]. It is observed that mitochondrial fragmentation could trigger mitophagy and mitochondrial apoptosis in response to mitochondrial oxidative stress injuries [48]. Moreover, fission can generate ROS, facilitate mitophagy, and accelerate cell apoptosis.

More specifically, Drp1-mediated mitochondrial fission is a prerequisite for mitophagy. It is observed that mitophagy receptor FUN14 domain containing 1 (FUNDC1) regulates mitochondrial division-fusion and mitochondrial autophagy by interacting with mitochondrial mitotic protein, Drp1, and optic nerve dystrophin 1 or with Drp1 and calcitonin [49]. Mdivi-1, the highly selective and efficient inhibitor of Drp1, was observed to suppress mitophagy because this agent could inhibit the formation of mitochondrial fragments [50,51,52]. Meanwhile, the inhibitor of Drp1 was consistently reported to attenuate neurological dysfunction and had a trend to decrease oxidative injury induced by the phosphorylation of Drp1 [53]. In the middle cerebral artery occlusion (MCAO) model mice, Mdivi-1 can significantly increase caspase-9 without affecting the expression of caspase-3, caspase-8, and Cyt c to inhibit mitophagy, indicating that inhibition of mitochondrial fission reduces mitophagy-mediated cell death without affecting apoptotic receptor-mediated apoptosis [54]. Targeting and selective inhibition of Drp1 contribute to maintaining normal mitochondrial morphology and save neurons from neurite outgrowth defects and cell apoptosis [55]. Notably, some studies indicated that pretreatment of Mdivi-1 could prevent Drp1-dependent excessive mitochondrial fission and mitophagy and mitigated neuro-cell death [56].

In contrast to fission, mitochondria fusion forms an elongated organelle [57]. Notably, mitochondrial fusion could maintain normal proliferation and selectively distribute the damaged components of mitochondria to the offspring, which is regarded as reversible damaged mitochondria. The fusion between reversibly damaged and healthy mitochondria is accomplished by diluting the accumulated mutational mitochondrial DNA and oxidized proteins to repair these slightly damaged mitochondria functionally [58]. However, the membrane potential of severely damaged mitochondria cannot be restored by fusion. Consequently, the severely damaged mitochondria could not participate in fusion with healthy ones and could be cleared by mitophagy. These processes resulted in the separation of healthy mitochondria and severely damaged mitochondria [59]. These observations revealed that mitochondrial dynamics are closely related to mitophagy in cerebral I/R injury (Fig. 2).

Fig. 2
figure 2

The dynamic changes of mitochondria and mitochondria-related mechanisms during ischemia/reperfusion. Under ischemia conditions (left), Drp1-mediated fission upregulated and subsequent mitophagy co-operate in eliminating dysfunctional mitochondria. Reperfusion of blood flow (right) results in an excessive increase in Drp1-mediated fission but a decrease in Opa1 and Mfn1/2-mediated fusion. This is accompanied by mitochondrial outer membrane permeabilization mediated by Bax/Bak oligomers and the release of apoptogenic species into the cytoplasm, and, eventually, neuronal death. When mitophagy is insufficient or excessive, it can also lead to cell death

Full size image

Mitophagy Signaling Pathway

The mechanisms of mitophagy generally involve Parkin-dependent as well as Parkin-independent pathways (Fig. 3). The current study of mitophagy primarily focuses on PINK1/Parkin, BNIP3/NIX, and FUNDC1 pathways.

Fig. 3
figure 3

Three signaling pathways regulating mitophagy. Mitophagy depends on Parkin-dependent and Parkin-independent pathways. In PINK1/Parkin pathway, PINK1 accumulates on OMM, recruiting Parkin from the cytoplasm to the mitochondria. Then PINK1 phosphorylates and activates the ubiquitin ligase activity of Parkin, and LC3 adaptors on autophagosomes recognize the ubiquitinated mitochondria; BNIP3, NIX, and dephosphorylated FUNDC1 can also connect to LC3, thereby inducing mitophagy

Full size image

PINK1/Parkin-Mediated Pathway

The most common and extensive mechanisms of mitophagy are the E3 ubiquitin ligase Parkin and the protein kinase PINK1-mediated classic pathways. PINK1 is a nuclear-encoded mitochondrial serine (Ser)/threonine kinase [60, 61], which contains N-terminal mitochondrial targeting signal, α-helical transmembrane domain, Ser/threonine kinase domain, and C-terminal mitochondrial outer membrane retention signal peptide sequence [62]. PINK1 was observed to mediate the ubiquitination of substrates and could regulate protein degradation as well as signal transduction [63,64,65].

In physiological conditions, PINK1 produced in the cytoplasm is transported to IMM, then it is cut and degraded rapidly by the ubiquitin-protease system in its transmembrane domain to 52 kDa PINK1. The processed PINK1 is released and rapidly removed [66]. However, the IMM is depolarized when the intracellular mitochondrial membrane is damaged. The transport of PINK1 into the mitochondria would be blocked, and the subsequent processing of PINK1 is disrupted, leading to the accumulation of full-length PINK1 in the mitochondrial outer membrane [67]. And it leads to Parkin activation and ubiquitin phosphorylation [63, 68, 69].

PINK1/Parkin pathway mediates mitophagy through the ubiquitination process. PINK1 can phosphorylate Ser65 on the N-terminal ubiquitin-like domain of Parkin to recruit and activate Parkin [70, 71]. Meanwhile, phosphorylated Ser65 activates its ubiquitin ligase activity, which could connect the polyubiquitin chain to the mitochondrial outer membrane protein [72]. The more ubiquitin chains, the more PINK1 substrates, the more Parkin will undoubtedly be recruited. Ubiquitin on the surface of mitochondria sends out the “EAT ME” signal, triggering the activation of the ubiquitin–proteasome system and autophagy [73,74,75]. LC3, the autophagy related protein binds to ubiquitinated mitochondria, and the damaged mitochondria will finally be degraded [76,77,78].

BNIP3/NIX-Mediated Pathway

Two vital receptors in the turnover of mitochondria located in the outer membrane of mitochondria are named B Lymphoma-2 gene/adenovirus E1B interacting protein 3 (BNIP3) and BNIP3-like (BNIP3L, also known as NIX) [79]. They have approximately 56% amino acid sequence identity [80, 81], both of them are pro-apoptotic mitochondrial proteins, and they are also essential participants in the process of autophagy and even mitophagy.

Under stress conditions, the phosphorylation of the LC3 interaction region sequence further increases the interaction with LC3 proteins [82]. It helped to form an autophagosome to clear the damaged mitochondria. In addition, the competitive binding of BNIP3 and NIX to the anti-apoptotic protein, Bcl-2, will dissociate the Bcl-2-Beclin-1 complex and release Beclin-1, subsequently activating autophagy and mitophagy [79, 83]. Furthermore, under hypoxia conditions, the expression of NIX and BNIP3 is regulated by forkhead box O3 and hypoxia-inducible factor [84, 85], thereby inducing or inhibiting mitochondrial fission, mitochondrial fusion, and mitophagy.

BINP3 is the target gene of hypoxia-inducible factor 1α. BINP3 could directly trigger autophagy under hypoxic conditions [86]. A study revealed that BNIP3 could promote the release of Cyt c. Meanwhile, treatment of mitochondrial oxidative phosphorylation uncoupling agent on Hela cells showed a significant increase in the interaction between NIX and LC3, proving that NIX is involved in the mitophagy process [87]. Another stroke study observed that upregulation of NIX could not restore mitochondrial function when mitophagy defects mediated by the deletion of BINP3 [88]. It may be related to the differences in conditions. NIX is transcriptionally upregulated during erythrocytes maturation to clear mitochondria by increasing the production of ROS, so it mainly works on mitochondrial autophagy under physiological conditions [89, 90].

FUNDC1-Mediated Pathway

FUNDC1 is an essential mitophagy receptor, which mediates mitophagy of most mammalian cells under hypoxic conditions [91]. In myocardial ischemia/reperfusion, studies have found that hypoxic preconditioning could induce FUNDC1-dependent mitophagy to resist I/R injury [92].

FUNDC1 is a novel tertiary transmembrane protein on the outer mitochondrial membrane, which contains three transmembrane domains, as well as a LC3 binding conformation face to the cytoplasm and a C-terminal domain inserted into the mitochondrial inner membrane [93].

FUNDC1 activity is modulated by its phosphorylation state. Under non-stress conditions, FUNDC1 can stably exist on the outer mitochondrial membrane without mediating mitophagy. This is due to the phosphorylation of LC3 interaction region motif of sarcoma gene kinase at Tyr18 and casein kinase II at Ser13, which lead to the inability of LC3 recruit and decrease of the interaction between FUNDC1 and LC3 [94, 95]. However, when mitochondria are damaged or dysfunctional during hypoxia, sarcoma kinase and casein kinase II are inactivated. Moreover, Ser17 on FUNDC1 is dephosphorylated by ULK1, and Ser13 is dephosphorylated by phosphoglycerol mutase family 5, enhancing FUNDC-LC3 interaction and inducing mitophagy [96]. In addition, FUNDC1 can modulate mitochondrial division-fusion and mitophagy by recruiting mitochondrial mitotic protein, Drp1, and disrupting its physical interaction with Opa1 under stress [94].

Mitophagy Under Different Pathological Injury States of Ischemia Stroke

When cerebral ischemia occurs, blocked blood flow leads to the lack of oxygen and nutrients, triggering a cerebral ischemic cascade [2]. Therefore, restoration of blood flow before tissue irreversibly injury is regarded as a foundational treatment. However, sudden reperfusion of blood supply can sometimes be harmful to ischemic or anoxic tissues, contributing to the so-called reperfusion injury [5]. Moreover, mitochondria play an important role in maintaining brain tissue function, which is highly energy-dependent. Until now, the specific mechanism of mitophagy in ischemic stroke still remains unclear.

Mitophagy‐Related Pathological Mechanisms in Ischemic Stroke

The cell death mechanism of the ischemic brain mainly involves ion imbalance, excitotoxicity, oxidative stress, and inflammation. These mechanisms develop in a relatively definite order and become the main events in different pathological injury states of ischemic stroke. Notably, oxidative stress reaches its peak at the beginning of reperfusion, meanwhile, inflammatory damage can last for several days or weeks after reperfusion. The activation of these mechanisms affects the function of cell membranes and organelles, such as mitochondria, endoplasmic reticulum (ER), lysosomes, and cell nuclei [97].

More specifically, an ischemic injury occurs when blood flow to brain tissue is reduced or blocked. Cell alterations caused by oxygen and glucose deprivation occurs in a few minutes after ischemia, leading to the dysfunction of mitochondrial oxidative phosphorylation and overproduction of ROS. After the oxygen and nutrients supply is restored, however, a series of pathological processes happens, such as the loss of mitochondrial membrane potential, generation of ROS, and activation of inflammation [98,99,100,101]. Meanwhile, mPTP opens due to an overload of free Ca2+ and ROS in the mitochondrial matrix, which can trigger different cellular responses, ranging from the physiological regulation of mitophagy to the activation of apoptosis or necrosis [102]. A variety of solutes and water will enter the mitochondrial matrix, which contributes to rupture of the outer membrane and swelling of the inner membrane, leading to the release of Cyt c and other proteins. Cyt c binds to apoptotic protease activating factor 1, and then, caspase-9-mediated apoptosis is consequently triggered, which causes cell deaths [103].

The Activation of Mitophagy in Ischemic Stroke

The ischemic and I/R injury reduced the expression of Siah2 or Drp1, which could significantly lead to mitochondria fragmentation and fission. Moreover, enhanced mitochondria fragmentation was observed following oxygen glucose deprivation (OGD) in primary rat neurons, accompanied by Opa1 releasing [104]. An advanced study showed that at the early stage of hypoxic/ischemic injury, a Drp1-dependent mitophagy was triggered, which participated in the removal of damaged mitochondria and cellular survival. It was also found that 30-min transient focal ischemia induced mitochondrial dynamic imbalance towards fission may favor mitochondrial fragmentation and mitophagy [54].

A rapid decrease in ATP production during the ischemic phase triggers AMPK pathways to initiate autophagy [105, 106]. AMPK activates the phosphorylation of unc-51-like kinase 1 (ULK1), which will then trigger the class III PI3K complex (Beclin 1, VPS34, and VPS15) that initiate the nucleation of the phagophore [105]. Meanwhile, ULK1 could bind to the FUNDC1 receptor, indirectly activating mitophagy [107].

Resupply of nutrients and oxygen sometimes could accelerate ROS production and accumulation, leading to excessive oxidative stress and local inflammation [108]. ROS has been observed to activate BNIP3 and PINK1/Parkin mediated mitophagy in cerebral ischemia and reperfusion [109, 110]. Meanwhile, another study showed that the mTOR pathways are inhibited in response to ROS during the reperfusion phase, promoting autophagy [111]. Ischemia and reperfusion significantly boosted mitophagy activated by PINK1/Parkin pathway [112].

Different Roles of Mitophagy During Ischemic Stroke

Under physiological conditions, autophagy could remove the abnormally aggregated proteins and degenerated subcellular organelles, while excessive autophagy may lead to massive cell death. As for mitophagy, many studies revealed that it could remove the damaged or dysfunctional mitochondria to prevent the generation of excessive ROS and the subsequent cell death [113]. In numerous clinical and experimental studies, mitophagy protects brain cells from ischemic injury after ischemic stroke occurs; however, mitophagy plays its double-edged sword role when brain suffers reperfusion injury [113,114,115,116].

In the ischemic phase of ischemic stroke, carnosine could attenuate ischemic injury by protecting against mitophagy [117]. Moreover, acidic postconditioning in the early stage of ischemic stroke (within 6 h) was observed to increase mitophagy, which could render the brain tissue resistant to ischemia in mice [118].

Studies have indicated that rapamycin treatment could protect against cerebral I/R injury by enhancing mitophagy and attenuating mitochondrial dysfunction in transient MCAO (tMCAO) rats. And these protective effects could be reversed by 3-methyladenine, an autophagy inhibitor, which mediated the induction of LC3-II and Beclin-1 in mitochondria [119]. Similarly, knockout of the mitophagy-related gene NIX showed aggravated neuronal apoptosis after I/R injury, and overexpression of NIX could rescue neurons [120]. An in vitro experiment found that H2 had a neuroprotective effect on OGD/R damaged neurons through elevating mitophagy mediated by PINK1/Parkin signaling pathway [121]. Besides, Parkin-dependent mitophagy has been reported to negatively modulate the activation of NLRP3 inflammasome to reduce excessive inflammatory responses induced by blood reperfusion [122].

Various studies have already confirmed the protective role of mitophagy in attenuating brain injury during the reperfusion phase [54, 123,124,125]. However, mitophagy has been found to be harmful in some studies. An in vitro study indicated that small nucleolar RNA host gene 14 was upregulated in OGD/R-damaged mouse hippocampal neurons, contributing to mitophagy which induced severe cell apoptosis [126]. In the MCAO of ischemic stroke, studies have found that mitophagy-related proteins were increased in brain tissues [127]. Similarly, inhibiting mitophagy can protect against cerebral I/R injury in the MCAO rats, which may be related to the inhibition of excessive mitochondrial autophagy [109, 128]. In addition, rehmapicroside inhibited mitophagy by preventing the accumulation of mitophagy-associated proteins in mitochondria to improve neurological deficit scores [127].

Therapeutic Potential of Mitophagy Regulation for Ischemic Stroke

In recent years, numerous experimental studies on the association of ischemic stroke and mitophagy aimed at interfering with autophagy to treat cerebral diseases [129, 130]. Advanced studies indicated that modulating mitophagy may benefit mitochondria function under I/R injury. These mitophagy regulation interventions could be proposed as adjunctive approaches for ischemic stroke management.

At present, existing agents, such as chloroquine, bafilomycin A1, and 3 methyladenine, have achieved pharmacologic inhibition of mitophagy [125]. However, although these agents are currently used to investigate the effect of mitophagy on I/R injury, they suffer from the same limitation-a lack of selectivity and specificity [131,132,133]. We reviewed some drug candidates which could modulate mitophagy and analyzed their possibilities in clinic application for ischemic stroke.

Given the low toxicity and safe pharmacokinetics as well as high utility rate, drugs derived from natural compounds are of high therapeutic value in the clinic for ischemic stroke treatment and exhibited considerable antioxidant, anti-inflammatory, and neuroprotective properties [134]. For example, resveratrol is a type of natural phenol found in various plants. Notably, a recent study revealed that administration of resveratrol at the reperfusion phase decreased cerebral I/R injury by promoting mitophagy [135]. Moreover, it was observed that resveratrol inhibited the activation of NLRP3 inflammasome through Sirt1-dependent autophagy to protect the brain from NLRP3 damage [136, 137].

Curcumin was observed to protect the brain from I/R injury by improving mitophagy and preserving mitochondrial function [113]. Curcumin also reduced ROS levels, which prevent the impairment of mitochondrial function from cerebral I/R injury. Moreover, curcumin elevated the co-localization of LC3B and mitochondrial marker VDAC1 as well as the ratio of LC3-II to LC3-I.

Additionally, it has been confirmed that natural compounds, such as ginkgetin, chrysophanol, and ginsenoside Rg1, can elevate mitophagy to exert an anti-ischemic and neuroprotective effect in tMCAO models [138, 139]. For example, ginkgetin can attenuate cerebral I/R injury induced autophagy and apoptosis by inhibiting the NF-κB/p53 signaling pathway [140]. Meanwhile, ginkgolic acid was revealed to decrease neuronal damage mediated by impaired mitophagy. The efficacy of ginkgolic acid was accompanied by increased autophagosomes as well [141]. Similarly, luteolin showed neuroprotection via enhanced oxygen free radicals scavenging and the activity of SOD in mitochondria as well as reversing the dysfunction of mitochondria [142]. On the contrary, a most recent study indicated that rehmapicroside could attenuate infarct size and improve neurological functions by inhibiting peroxynitrite-mediated mitophagy activation in vitro and in vivo. Moreover, rehmapicroside reduced the translocations of PINK1, Parkin, and Drp1 into the mitochondria for mitophagy in the ischemia-reperfused rat brains [127].

Like the double-edged sword role of mitophagy, we think that the regulation of mitophagy by natural compounds may also be “bidirectional,” inhibiting excessive mitophagy to prevent unnecessary cell death or promoting proper mitophagy to eliminate dysfunctional mitochondria. Further research is needed on the more precise mechanism and precise regulation of mitophagy.

Conclusions

The ischemic area of the brain can be generally divided into the infarct core and the ischemic penumbra, according to the severity of the blood flow reduction. There is a rapid decrease in ATP levels and energy stores and severe ionic disruption in the infarct core, which contributes to cell death within a few minutes. In comparison, the blood flow reduction in the ischemic penumbra is less severe due to collateral blood vessels. Therefore, the ischemic penumbra is the focus of clinical intervention on ischemic stroke, and reperfusion therapy within the time window can rescue the brain tissue of penumbra to a certain extent. There are multiple milder cell death mechanisms, such as inflammation and apoptosis in the ischemic penumbra; however, the ischemic penumbra changes constantly, and the infarct core will expand during cerebral ischemia on the base of ischemic penumbra [97]. Hence, early and timely reperfusion is the most effective way to treat and improve the prognosis of patients with ischemic stroke [143]. However, the re-supply of blood can cause excessive activation of enzymes which previously inhibited by ischemia-induced ATP deficiency, leading to increased production of ROS and destruction of calcium homeostasis in the cytoplasm and mitochondria, eventually triggering apoptosis-related factors release and initiating cell death [10]. Hence, controlling the quantity and quality of mitochondria through mitophagy may be a possible solution to protect neurons and prevent cell death for ischemic stroke. Meanwhile, to apply it to the clinical treatment for ischemic stroke, some scientific problems must be solved.

Firstly, under different pathological injury states of ischemic stroke, the role of mitochondria is dissimilar, and the mitophagy results will also be different. In a tMCAO model with 1-h ischemia and 23-h reperfusion, mitophagy was inhibited at the onset of reperfusion, which increased neuronal death [125]. Under mild ischemia or tolerable hypoxic stress, studies suggested that selective removal of dysfunctional mitochondria by mitophagy is protective for the neurons. On the contrary, when subjected to long-term ischemic injury or severe I/R damage, dysfunctional mitochondria increased; thus, destructive mitophagy was induced, and vulnerable brain cells began to die, leading to permanent and irreversible injury [144]. Therefore, selective treatment strategies may be required for different pathological injury states of ischemic stroke. On the other hand, a particular theme in mitochondrial quality control is recurring, which refers to balance. Moderate mitophagy could effectively eliminate damaged mitochondria and reduce excessive ROS generation. However, enhanced mitochondrial fragmentation and fission as well as excessive mitophagy cause a deficit in ATP production and energy supplementation for neuronal survival. At the same time, inefficient or inappropriate inhibition of mitophagy would also lead to mitochondria dysfunction and neuronal apoptosis. Hence, mitophagy needs to be monitored with more sophisticated methods.

Mitophagy could be regulated by enhancing or inhibiting factors, such as Drp1 that mediated fission, thereby reducing I/R injury. For example, the inhibition of Drp1 has been shown to have an excellent neuroprotective effect during the transient cerebral ischemia and reperfusion injury; however, long-term application of Drp1 inhibitors may disrupt the dynamic changes of mitochondria and affect the normal cell cycle. Therefore, mitochondrial fission inhibitors may only be used in the acute phase of cerebral ischemia, and its adverse effects and deeper molecular mechanisms require further observation. Furthermore, we summarized several natural compounds which could regulate mitophagy through multiple targets and exhibit significant neuroprotective effects in experimental ischemic stroke. The application of these natural compounds may provide reliable treatments for ischemic stroke, especially for the long recovery stage. However, with poor water solubility and the blood–brain barrier permeability as well as the limited bioavailability of these compounds, the use for biological applications would be seriously limited, such as curcumin [145]. For those phytochemicals, modification of chemical structures or development of efficient drug delivery systems is required to enhance their bioavailability and blood–brain barrier permeability.

Data obtained over the past two decades have investigated compelling evidence of inextricable links between mitochondrial quality and quantity control with cell fate in neurons subjected to ischemia–reperfusion injury. We believe that the in-depth study of the molecular biological mechanism of mitophagy provides new perspectives, potential therapeutic targets for ischemic stroke and diseases related to the neural system. However, further studies will be required to (1) develop novel and targeted, mechanisms-based pharmacologic therapies to improve prognosis in patients with I/R injury and to (2) identify the interaction between mitochondrial dynamics and mitophagy at different ischemia periods.