Oligodendrocyte Pathophysiology and Treatment Strategies in Cerebral Ischemia

Summary

Oligodendrocytes (OLs), the myelin‐forming cells of the central nervous system, form a functional unit with axons and play a crucial role in axonal integrity. An episode of hypoxia–ischemia causes rapid and severe damage to these particularly vulnerable cells via multiple pathways such as overactivation of glutamate and ATP receptors, oxidative stress, and disruption of mitochondrial function. The cardinal effect of OL pathology is demyelination and dysmyelination, and this has profound effects on axonal function, transport, structure, metabolism, and survival. The OL is a primary target of ischemia in adult‐onset stroke and especially in periventricular leukomalacia and should be considered as a primary therapeutic target in these conditions. More emphasis is needed on therapeutic strategies that target OLs, myelin, and their receptors, as these have the potential to significantly attenuate white matter injury and to establish functional recovery of white matter after stroke. In this review, we will summarize recent progress on the role of OLs in white matter ischemic injury and the current and emerging principles that form the basis for protective strategies against OL death.

Introduction

After an episode of cerebral hypoxia–ischemia (HI), early events include energy crisis, cell depolarization from the breakdown of transmembrane gradients, cytotoxic edema, reactive oxygen species (ROS) production, and endothelial dysfunction 1. These events prompt a complex cascade resulting in neuronal and glial damage and death. OLs, the myelin‐forming cells of the CNS, are acutely damaged by short periods of HI. Cell swelling occurs as early as 30 min after arterial occlusion, and large numbers of OLs die within 3 h 2. It has been reported 3 that 30 min of oxygen–glucose deprivation (OGD) results in the death of 90% of OLs within 9 h. OL pathology results in demyelination and dysmyelination which have profound consequences for axonal function, transport, structure, metabolism, and survival 4, 5, 6. The most devastating effects of HI on these cells occur in premature infants of <32 weeks' gestation, which show pathological symptoms of chronic myelination disturbance, leading to periventricular white matter injury 7. The white matter of these infants is immature and poorly vascularized and contains oligodendrocyte progenitors (pre‐OL) which are sensitive to ischemia and infection.

Research in neurological disorders is progressively embracing the concept of the neurovascular unit, which emphasizes that a successful neurorestorative therapy cannot exclusively target neurons, but must also encompass glial and endothelial cells 8. Thus, therapeutic strategies that target OLs, myelin, and their receptors have the potential to significantly attenuate white matter injury in HI. This review highlights the mechanisms of OL injury and death in HI at all stages of development and focuses on the oligoprotective and oligorestorative therapies that have been investigated thus far.

Intrinsic Susceptibility of Oligodendrocytes to Hypoxic–Ischemic Damage

OLs display a number of features that render them more vulnerable to HI than other CNS glial cells, and in certain brain regions and stages of development, more vulnerable than neurons 9 (Figure 1).

Figure 1.

Features of oligodendrocytes (OLs) and pre‐OLs which render them vulnerable to hypoxia–ischemia (HI). The amplified vulnerability of OLs to HI derives from their high iron content, low reduced glutathione content, high rate of oxidative metabolism, high lipid and sphingolipid content, and high permeability of glutamate receptors. Pre‐OLs are even more vulnerable than their mature counterparts due to low levels of antioxidant enzymes, upregulation of AMPA/kainate receptors and enhanced AMPA/kainate signaling, increased expression of the glutamate transporter EAAT2, and a susceptibility to direct mitochondrial toxicity by reactive nitrogen species. In the event of HI, these properties lead to higher levels of oxidative stress and apoptosis, hence, severe damage, and death to cells of this lineage.

Of all the cell types in the brain, OLs contain the highest levels of immobilized, protein‐bound iron, which is a basic requirement for their function and oxidative metabolism, and for the synthesis of myelin components 10, 11. Apart from its important functional role, ferrous iron (Fe²⁺) can be a potent cytotoxin by catalyzing the conversion of hydrogen peroxide to hydroxyl radicals (OH), via the Fenton reaction 12, 13. In cerebral ischemia, an energy crisis leads to lactic acidosis, which results in mobilization of protein‐bound iron stores. This increases the levels of free cytosolic Fe²⁺ that participates in the Fenton reaction to bring about oxidative stress 14, 15. This effect is further amplified in OLs by their low content of reduced glutathione (GSH) 16, 17 which is an electron donor for the function of glutathione peroxidase, which in turn, scavenges peroxides. OLs contain less than half of the glutathione content of astrocytes and <15% of the glutathione peroxidase activity, which leads to a peroxide‐scavenging deficit 17. OLs also have the highest rate of oxidative metabolism by volume and can support a myelin membrane up to 100 times the weight of their cell bodies 4, 11, 18. This high metabolic activity generates more ROS 18 and requires a correspondingly high consumption of oxygen and ATP, the synthesis of which generates hydrogen peroxide as a by‐product 19, 20, 21.

The subunit composition of glutamate receptors in OLs continues to predispose them to injury during HI. Their AMPA receptors are especially permeable to Ca²⁺ 22, 23, and their NMDA receptors are only weakly blocked by Mg²⁺, enabling them to generate a substantial current even at resting membrane potential 22, 24, 25.

Sphingolipids, constituents of the myelin membrane, may also increase the susceptibility of OLs to damage under pathological conditions 4, 26. The simplest sphingolipid, ceramide, can activate the major pathways that govern cell death 27 and kill cells by limiting access to extracellular nutrients 28. Many apoptotic stimuli activate acid sphingomyelinase, an enzyme that mediates ceramide release from biological membranes 29, 30. Ceramide‐enriched signaling platforms that contain death receptors are formed in the plasmalemma, and these transmit apoptotic signals into the cell 29, 31. Ceramide released intracellularly also acts as a second messenger, leading to caspase‐mediated OL apoptosis within hours 26, 32, 33.

Even more susceptible to injury than mature OLs are the O4⁺/O1⁻ late OL progenitors, which comprises about 90% of all OLs during the high‐risk period for periventricular leukomalacia (PVL) 7, 34. This vulnerability is a consequence of:

1. Amplified oxidative damage that results from a developmental deficit in superoxide dismutases (SOD‐1 and ‐2) and a hydrogen peroxide‐scavenging deficit 35, 36, 37 combined with active iron acquisition 11.

2. Higher vulnerability to reactive nitrogen species attack by direct mitochondrial toxicity with translocation of apoptosis‐inducing factor 38 and formation of peroxynitrite 39, 40.

3. Significant developmental upregulation of non‐NMDA glutamate receptors 41, 42 accompanied by enhanced AMPA‐mediated calcium signaling 43, which increases excitotoxicity. Furthermore, pre‐OLs also exhibit a transiently increased expression of the glutamate transporter (GluT) EAAT2, which may become a source of glutamate under pathological conditions 44.

Mechanisms of Oligodendrocyte Damage in HI

Neurotransmitter‐Mediated Toxicity

OLs express neurotransmitter receptors that allow for axon‐to‐OL signaling and mediate their own development and function. The major excitatory neurotransmitters involved are glutamate and ATP 45, 46. These bind to their respective receptors on the OL plasmalemma and result in an influx of ions, most notably Ca²⁺, which acts as a chemical signal under physiological conditions, triggering OL differentiation and myelination 47.

OLs are extremely sensitive to disruptions in intracellular calcium homeostasis 25. In HI, energy crisis and metabolic stress lead to prolonged overstimulation of neurotransmitter receptors, resulting in a cytosolic Ca²⁺ surge which is worsened by the activation of voltage‐gated calcium channels (VGCC) and the reversal of the Na⁺/Ca²⁺ exchanger (NCX) 48; (Figure 2). This Ca²⁺ is sequestered by mitochondria and leads to mitochondrial bioenergetic dysfunction, which is characterized by impaired oxidative phosphorylation, ROS generation, the release of apoptogenic proteins, such as cytochrome C, and cell death by apoptosis or necrosis 49.

Figure 2.

The major pathways governing neurotransmitter‐mediated oligodendrocyte death in hypoxia–ischemia (HI). The glutamate surge that occurs during HI leads to the overactivation of AMPA/kainate receptors on oligodendrocyte somata and NMDA receptors on myelinating processes. ATP is also released in HI, partly from the oligodendrocyte itself via pannexin hemichannels, leading to the overactivation of purinergic P2X7 receptors and enhanced Ca²⁺ signaling. The Ca²⁺ surge leads to the activation of voltage‐gated calcium channels (VGCC) and reversal of Na⁺/Ca²⁺ exchanger (NCX), further increasing the intracellular Ca²⁺. The glutamate transporter EAAT2 also starts to operate in reverse, contributing to the surge in extracellular glutamate. The excess cytosolic Ca²⁺ is sequestered in mitochondria where it leads to mitochondrial disruption and oxidative stress and eventual death to oligodendrocytes.

Glutamate‐Mediated Toxicity

Glutamate excitotoxicity is one of the major contributors toward ischemic injury in the CNS 50. OLs are sensitive to glutamate‐induced cell death 48 with an EC₅₀ of 200 μM for a 24‐h exposure period 51. The glutamate signaling is governed by ionotropic and metabotropic glutamate receptors (iGluRs and mGluRs, respectively) and GluTs 52, 53. OLs express three main types of iGluRs: the AMPA and kainate receptors, predominantly located on their cell body, and NMDA receptors, clustered on their myelinating processes 25, 54. Pre‐OLs strongly express all three groups of mGluRs, but these are downregulated in mature OLs 55. GluTs are responsible for the uptake of glutamate from the extracellular space and maintenance of low extracellular glutamate levels (1–2 μM). However, under conditions of energy failure that result from HI, GluTs on OLs, astrocytes and microglia operate in reverse, with release of glutamate into the extracellular space 48, 56.

OLs have for a very long time, been known to be vulnerable to AMPA/kainate receptor‐mediated excitotoxicity 57, 58. The AMPA receptors on mature, myelinating OLs contain subunits GluR3 and GluR4, but not GluR1 59, and although mRNA for GluR2 is present, immunoprecipitation experiments indicate that GluR2 does not assemble with the other subunits 22, 60, which renders them highly permeable to Ca²⁺ when activated 23, 61. In fact, Ca²⁺ influx via these receptors alone is enough to induce death of OLs by excitotoxicity in culture, 61 and blockade of AMPA/kainate receptors alone prevents death of OLs by OGD during cerebral ischemic injury 3. Prolonged activation of these receptors leads to caspase‐dependent and caspase‐independent death pathways 62, and this toxicity is dose‐dependent 58. Recent work suggests the involvement of disturbed Zn²⁺ homeostasis in AMPA‐induced excitotoxicity. It has been demonstrated 63 that activation of OL AMPA receptors leads to mobilization of intracellular Zn²⁺ and a surge in cytosolic Zn²⁺, which contributes toward ROS production and mitochondrial depolarization, by a mechanism which is altogether separate from the Fenton reaction.

Oligodendrocytes express mRNA for the kainate subunits GluR6, GluR7, KA‐1, and KA‐2, but not for GluR5 22. Nontoxic concentrations of glutamate can sensitize these cells to complement attack, inducing OL death, in a process that is mediated exclusively by activation of kainate receptors. This complement‐induced death of OLs occurs via formation of the membrane attack complex, which increases membrane conductance and leads to a Ca²⁺ surge 64.

Excitotoxic OL death was previously thought to be exclusively mediated by AMPA and kainate glutamate receptors. Three recent reports 24, 54, 65 have manifestly altered this view by showing in vivo that OLs and myelin possess NMDA receptors and that these are involved in ischemic injury. White matter OLs at all stages of development contain NR1, NR2C, and NR3A NMDA receptor subunits, allowing for inward currents upon binding of glutamate 66. OL NMDA receptors are enriched with NR2C and NR3A subunits, which are blocked weakly by extracellular Mg²⁺ and allow for the generation of a current even at the cell's resting membrane potential 65, 66, 67. As NMDA receptors are clustered on the myelinating processes of OLs, receptor activation leads to a drastic increase in ion concentration because of the small intracellular volume 25, 54, 68 with disruption of myelin structure and action potential propagation 69.

In pre‐OLs, glutamate toxicity also occurs via a non‐receptor‐mediated mechanism, referred to as oxytosis, or oxidative glutamate toxicity 70. This involves system ${\mathrm{x}}_c^{-}$, a plasmalemmal antiport protein that transports cystine into the cytosol in exchange for glutamate to the extracellular space, in a 1:1 ratio 51. Once in the cytosol, the cystine is converted to cysteine, which is used in the production of glutathione 71. High extracellular glutamate concentrations can reverse the direction of this transport, promoting the efflux of cystine with consequent depletion of intracellular glutathione, and enhancement of oxidative stress 51, 72. Although this phenomenon is not exclusive to pre‐OLs, they exhibit enhanced, maturation‐dependent vulnerability because of low levels of glutathione peroxidase and SODs, especially SOD‐2 35, 36, 37.

ATP‐Mediated Toxicity

ATP activates ionotropic P2X and metabotropic P2Y purinoreceptors, both of which are expressed by OLs 73. P2X receptors consist of P2X1‐7 subunits that are most permeable to Ca²⁺ ions 74, 75. During ischemia, ATP‐mediated toxicity to OLs occurs mainly via P2X7 receptor subtypes, the sustained activation of which induces cell death, myelin damage, and white matter injury 53, 76, 77.

During situations of metabolic stress, such as cerebral ischemia, anoxic depolarization causes ATP to be released from glial cells, leading to a surge in the extracellular ATP concentration 78. It has been suggested 76 that OLs may release ATP during ischemia via pannexin hemichannels, resulting in depolarization of mitochondria and release of ROS. ATP released from dying cells can continue to aggravate P2X7‐mediated injury 6. Functional P2Y and P2X receptors are also expressed by pre‐OLs 79, the latter of which exhibit postischemic downregulation 80.

ATP‐mediated toxicity leads to apoptosis or necrosis of OLs, the mode of cell death being determined by the intensity of the Ca²⁺ surge, which, in turn, depends on the intensity of the ischemic insult 6. Prolonged stimulation of P2X7 receptors also leads to several enzyme and secondary messenger cascades, with release of cytokines such as interleukin‐1β and activation of mitogen‐activated protein kinase (MAPK) and nuclear factor‐κB, among others 75, 81.

Mitochondrial Disruption and Oxidative Stress

Oxidative damage is a cardinal consequence of neurotransmitter‐mediated toxicity. HI rapidly causes oxidative stress in OLs, which is characterized by enhanced production of the superoxide radical (${\mathrm{O}}_2^{-}$), lipid peroxidation, and reduction of Fe³⁺ to the oxidant Fe²⁺ 82. The exposure of OLs to systems which generate free radicals, or free radical donors, such as ${\mathrm{O}}_2^{-}$ and NO, leads to their rapid necrosis or apoptosis 5, 83.

The drastic rise in cytoplasmic Ca²⁺ that occurs during HI has profound consequences for mitochondria, which sequester this cation in large amounts and generate ROS at levels dependent on Ca²⁺ uptake 84. The oxidative stress that ensues activates several signaling pathways that modulate the functions of enzymes and transcription factors. These signals cause changes in gene expression that influence the cell's survivability 85. ${\mathrm{O}}_2^{-}$ and NO radicals are particularly toxic to mitochondria as they interact with and block several key proteins of the respiratory chain 18. These radicals also lead to a diffusion‐limited generation of peroxynitrite, which causes death of OLs by lipid peroxidation, release of Zn²⁺, activation of extracellular signal‐regulated kinases and of 12‐lipoxygenase, and formation of additional ROS 86.

Auxiliary Mechanisms

Kinins are peptides produced at sites of tissue injury or inflammation 87. They activate specific B₁ or B₂ receptors, which mediate a number of signaling transduction mechanisms 88. In the CNS, kinins act as neuromediators 89. They also promote the synthesis of other pro‐inflammatory mediators, including cytotoxins and prostanoids, which lead to tissue damage and blood–brain barrier breakdown 88, 90. Functional kinin receptors are expressed by OLs, and their activation leads to a cytosolic Ca²⁺ surge, inflammation, and turnover of phosphoinositide 1, 91. Following ischemia, expression of B1 and B2 kinin receptors is upregulated, and the concentrations of bradykinin and kallidin also increase and result in damage and death of neural and glial tissue. Because of this, B1R and B2R receptor antagonists may be useable as neuroprotective and glioprotective agents during stroke, especially because they target multiple mechanisms that are involved in different stages of brain pathology 1.

The activation of dopamine D2 and D3 receptors 92, GABA_A receptors 22 and adenosine A_2A receptors 93 has also been implicated in ischemic damage of OLs. Moreover, A₁ adenosine receptors are found on pre‐OLs, and their activation in HI inhibits maturation of these cells 94, with consequent shortage of myelinating OLs.

The Role of Neighboring Glia

Neighboring glia cause bystander damage to OLs in HI. Glutamate activates AMPA/kainate receptors in both resting and activated microglia at the site of injury and thereby enhances production and release of the cytokine, tumor necrosis factor‐α 95. This can kill OLs by apoptosis and by potentiation of interferon γ toxicity and is more toxic to pre‐OLs than to mature OLs 96, 97. Reactive microglia also release interleukin‐1β, glutamate 98, and reactive oxygen and nitrogen species, such as peroxynitrite 40, which further inhibit glutamate uptake and amplify excitotoxic damage 99.

Activation of microglia is a major source of damage to pre‐OLs in PVL, especially as the number of microglia in cerebral white matter peaks during the period of highest vulnerability to PVL 100. Reactive astrocytes, microglia, and macrophages also damage pre‐OLs in PVL, by the release of interferon γ 101, which leads to an increase in inducible nitric oxide synthase (iNOS) that becomes upregulated during HI 102. iNOS generates NO, which injures pre‐OLs by peroxynitrite formation and nitrosative damage. Antimicroglial agents, such as minocycline and melatonin, provide promising routes to the attenuation of pre‐OL damage and demyelination in PVL 97.

Recovery from Trauma and Role of Adult Oligodendrocyte Progenitor Cells (OPCs)

An important task of the adult CNS after an episode of HI is the replacement of affected OLs and the remyelination of affected axons, to restore saltatory conduction, improving motor function 103. In vivo rodent models of stroke have demonstrated that a few days following an insult, OLs surrounding the infarct tend to increase in number 104. Axons that have been demyelinated as a result of trauma or disease can be remyelinated by immature cells that “respond to demyelination by differentiating into myelinating OLs” 105. These cells, now referred to as adult OPCs, form part of a larger subtype of glial cells, NG2+ glia, which express the NG2 proteoglycan and platelet‐derived growth factor‐alpha (PDGF‐α) receptors 106. Also known as polydendrocytes, these cells are closely intermingled with other glial cells in the CNS, but nonetheless represent a distinct cell population 107.

Adult OPCs are not pre‐OLs but mature cells which develop after birth. They become activated during axonal inflammation and/or demyelination and develop into mature, myelinating OLs 108. Many chemical signals appear to be responsible for their activation, including axonal signals released on demyelination, growth factors and cytokines from other activated glial cells, as well as other injury‐induced stimuli, such as ATP and glutamate surges 4. It is of interest that, although TNF‐α causes death of OL by apoptosis 96, lack of TNF‐α leads to a delay in remyelination and a reduction in the population of proliferating adult NG2+ OPCs, which is followed by a decrease in the number of myelinating OLs. Apparently, the binding of this cytokine to TNF receptor 2 (TNFR2) is critical for the regeneration of OLs after trauma 109.

Recently, several therapies have been evaluated to target the protection or multiplication of these progenitors and allow for replacement of OLs and remyelination. Sun et al. 110 report that the synthetic cannabinoid agonist WIN55, 212‐2, has been shown to reduce injury to NG2+ glia cells and to promote their multiplication in the stroke penumbra. Adenosine was found to accelerate the maturation of OPCs in culture 111 and erythropoietin to stimulate oligodendrogenesis and maturation in vivo 112. The transplantation of predifferentiated human embryonic stem cells, which develop into myelinating OPCs, has also been proposed 103.

In PVL, pre‐OLs and immature OLs also exhibit a defensive reaction in response to HI. These cells typically take the form of an enlarged soma with elaborate, thickened processes that are not typical of OLs at this stage of development and with a concentrated distribution around areas of injury 34. HI also promotes accelerated maturation of pre‐OLs to immature OLs, which are less vulnerable to ischemia 7.

Protective Strategies for Oligodendrocyte Injury in HI

Numerous neuroprotective agents have been developed and tested for their ability to block specific cell damaging pathways in the ischemic cascade. Although many of these gave promising results in animal models, clinical trials have been, for the most part, disappointing, because of a lack of efficacy and/or clinical safety concerns. This failure may be explained, in part, by the histological and morphological differences between human and rodent brains 113. Ginsberg 114 also suggests that many agents may have been taken to clinical trials without sufficient preclinical evidence of efficacy. More rigorous experimentation is necessary to elucidate efficacious and clinically safe neuroprotective and glioprotective agents, with a focus on targeting multiple biochemical cascades and CNS cell types, and combinatorial therapies. A summary of the agents that have been deemed most promising in conferring protection to OLs is provided in Table 1.

Table 1.

Therapeutic candidates for oligoprotection in hypoxia–ischemia

Mechanism	Oligoprotective agent	Oligodendrocyte maturation stage	Experimental model	References
AMPA antagonist	NBQX	Mature	Brain slices (mouse)	3
		Pre‐OLs	In vivo(rat)	42
	Topiramate	Pre‐OLs	In vivo(rat)	117
	SPD502	Mature	In vivo (rat)	118
	GYKI52466	Mature	Brain slices (mouse)	3
	CNQX	Mature	Optic nerve oligodendrocyte culture	58
NMDA‐antagonist	D‐AP5	Mature	Live adult rat optic nerve	24
		Pre‐OLs, immature, mature	Brain slices (rat)	65
	MK801	Mature	Live rat optic nerve	24
		Pre‐OLs, immature, mature	Brain slices (rat)	65
	Memantine	Mature	Brain slices (rat)	69
		Pre‐ols	In vivo (rat)	120
	7‐CKA	Mature	Live adult rat optic nerve	24
Reverse glutamate transport inhibitor	Dihydrokainic acid	Immature	Cultured rat OLs	119
P2X7 antagonist	BBG	Mature	Rat optic nerve oligodendrocyte culture + isolated optic nerve	53
	Oxidized ATP
P2X antagonist	PPADS
ATP degrader	Apyrase			76
Pannexin hemichannel blocker	Mefloquine
Adenosine receptor antagonist	SCH58261	Mature	In vivo (rat)	93
	Caffeine	Pre‐OLs	In vivo (mouse)	94
Antioxidant/radical scavenger	Mangiferin	Mature	Optic nerve oligodendrocyte culture	84
	Morin
	N‐acetyl cysteine	Pre‐OLs	Rat oligodendrocyte progenitor cultures	124
	Edaravone	Mature	In vivo (rat)	127
		Mature	Clinical trial	128
	α‐phenyl‐tert‐butyl‐nitrone	Mature	In vivo (rat)	126
	Vitamin K	Pre‐OLs	Cultured rat OLs	129
	Ebselen	Mature	In vivo (rat)	130
		Mature	Clinical trial	133
	Erythropoietin	Pre‐OLs	In vivo (sheep)	135
	Melatonin	Pre‐OLs	In vivo (rat)	136, 137
	Estradiol	Mature	In vivo (mouse)	138
Iron chelator	Deferoxamine	Mature	Cultured rat OLs	134
Antiapoptotic agent	IGF‐1	Pre‐OLs	In vivo (rat)	97, 140
		Pre‐OLs	In vivo (lamb)	97, 141
		Mature	In vivo (mouse)	142
	CNTF	Pre‐OLs	In vivo (mouse)	97, 143
	Estradiol	Pre‐OLs	Cultured rat OL + in vivo (rat)	97, 144
Antimicroglial agent	Minocycline	Pre‐OLs	In vivo (rat)	97, 145
Cannabinoid agonist	WIN55, 212‐2	OPCs	In vivo (rat)	110

BBG, brilliant blue‐G; OLs, oligodendrocytes; OPC, oligodendrocyte progenitor cell.

Protection Against Neurotransmitter‐Mediated Injury

Excitotoxic OL, pre‐OL, and neuronal injury can be attenuated by administration of the AMPA antagonist NBQX, which preserves white matter structure and improves motor deficits 3, 42, 115, although this compound may not be clinically safe 116. Topiramate, a clinically safe anticonvulsant, protects pre‐OLs against HI when administered postinsult, as does NBQX 117. SPD 502, a competitive AMPA antagonist, protects both gray and white matter, including OLs, when administered intravenously 15 min before the insult, and for 4 h after the insult 118. Other AMPA antagonists that have been shown to protect OLs against excitotoxic damage include GYKI 52466 3 and CNQX 58. Dihydrokainic acid, an inhibitor of glutamate release via reverse transport, significantly protected immature OLs from ischemic injury in culture 119.

NMDA receptors are excellent targets for antagonists because they contain several sites at which ligands can bind in a subunit‐selective manner, such as glutamate‐binding sites, ion‐channel pores, and allosteric sites on the N‐terminal domain. NMDA receptor antagonists that target NR3A and NR2C subunits have the potential of acting as major therapeutic targets for white matter preservation in stroke 67. The NMDA antagonists, D‐AP5, and MK801 protect OLs and myelin from excitotoxic death, but are not clinically safe 24, 65, 72. Memantine, a clinically safe, uncompetitive NMDA receptor blocker is also effective against injury in both OLs and pre‐OLs 72, 120, and 7‐CKA protects OLs and myelin during chemical ischemia in vitro 24. Of interest is that blockade of NMDA receptors or removal of extracellular Ca²⁺ worsens, rather than improves, functional recovery in aging animals 121, which emphasizes the importance of age‐specific stroke treatment.

Another possible therapeutic route is the upregulation of GluTs, as these allow for ischemic tolerance subsequent to ischemic preconditioning. EAAT2 promoters, such as valproic acid, can protect glia against ischemia by enhanced removal of glutamate from the extracellular space 122, 123, 124.

Ischemia‐induced mitochondrial depolarization and oxidative stress are partially reversed by P2X7 receptor antagonists, by the ATP‐degrading enzyme apyrase, and by pannexin hemichannel blockers such as mefloquine. P2X7 receptor antagonists do not interfere with normal physiological function because of their selective activation 76, 77. The P2X7 antagonists Brilliant Blue‐G (BBG), oxidized ATP (oATP), and the nonselective P2X antagonist PPADS prevent ATP‐mediated OL toxicity 53. The calmodulin antagonist calmidazolium has been shown to inhibit P2X7‐receptor evoked glutamate release and may therefore have potential in oligoprotection during ischemia 125. The administration of the selective adenosine A_2A receptor antagonist SCH58261 also protects OLs against cerebral ischemia by reducing the activation of the MAPK, JNK 93. Caffeine, an adenosine receptor antagonist, was found to be protective in PVL as it promotes the maturation of pre‐OLs after HI 94.

Protection Against Oxidative Stress

Antioxidants are potent therapeutic candidates for oxidative damage to OLs in cerebral ischemia. Mangiferin and morin, two natural antioxidant polyphenols, protect OLs from excitotoxic insult by free radical scavenging and cytosolic Ca²⁺ handling 84. N‐acetyl cysteine also attenuates AMPA/kainate OL cytotoxicity by increasing intracellular glutathione levels 124. Pretreatment with the spin‐trap agent α‐phenyl‐tertbutyl‐nitrone (PBN) reduced the number of damaged OLs by 55%, 40 min after the insult 126. The radical scavenger edaravone protects all components of the neurovascular unit against oxidative stress 8, 127, 128, while Vitamin K prevents oxidative damage to pre‐OLs and neurons during HI, with clinical safety 129. 12‐lipoxygenase inhibitors may also be of protective value to OLs at all stages of development, as 12‐lipoxygenase is a potent generator of ROS 97.

Ebselen, a mimic of glutathione peroxidase and phospholipid hydroperoxide glutathione peroxidase, exerts potent antioxidant effects on OLs and neurons 130, 131, 132. When administered intravenously, 2 h after stroke onset, it can salvage damaged tissue without major side effects 130. In a clinical trial, ebselen demonstrated a significant improvement in stroke patients who started ebselen treatment within 24 h of onset of the insult 133. The iron chelator deferoxamine protects OLs from cytotoxic effects induced by H₂O₂ and suppresses free radical formation 134. In clinical trials for PVL, erythropoietin, an antiinflammatory, antiapoptotic, antioxidant, and neurotrophic agent was found to reduce injury and preserve myelination in infants with moderate damage, without significant adverse effects 135. Melatonin, a free radical scavenger and up‐regulator of SOD, catalase, and glutathione peroxidase, has been found to promote pre‐OL maturation after perinatal brain damage 136 and decreases white matter inflammation, promoting myelination after neonatal stroke 137. The administration of the hormone 17β‐estradiol was recently shown to attenuate OL loss in the corpus callosum of male mice, and results in decreased demyelination and microglial activation 138, by a quinol‐based cyclic antioxidant mechanism 139.

The ability to visualize OLs in living brain through cell type‐selective transfer of genes encoding fluorescent proteins 146 provides new opportunities to understand cell–cell interactions of recovery in diseases of the myelinating unit.

Conclusion

Largely ignored for many years, the importance of OLs in the pathophysiology of a variety of neurological disorders has become evident. We now know that OLs are major targets of cerebral ischemia, both in the case of adult‐onset stroke and especially in PVL, which means that treatment strategies that exclusively target neuronal recovery cannot be optimally successful. This has led and should continue to lead researchers to make new links and explore new pathways of investigation, with the objective of treating cerebral ischemia in a more comprehensive manner.

New, groundbreaking research on oligodendrocyte pathophysiology in ischemia is constantly being made available. A notable example is the relatively recent discovery of functional NMDA receptors on OLs, antagonists of which are now being considered a possibly valid and valuable therapeutic route. Further work should continue to elucidate the exact underlying mechanisms of oligodendrocyte pathophysiology and to shed light on therapies that simultaneously target multiple mechanisms of injury and multiple components of the neurovascular unit. Therefore, it is hoped that future investigations should continue to work toward generating animal models of white matter stroke along with well‐designed clinical trials to extrapolate the findings on experimental animals to human neurological disease. These therapies are expected to be more innovative, more extensive, and more clinically viable.

Conflict of Interest

The authors have no conflict of interest.