White matter injury after brain ischemia

Abstract

Ischemic white matter injury (WMI) is a significant concern in various clinical settings, particularly affecting the elderly and those with comorbidities such as hypoxia and hyperglycemia. Ischemic WMI can be caused either by primary WM ischemia, or secondary axonal injury as a result of the loss of the “remote” axonal parent neurons or the depletion of adhesive and neurotrophic molecules following post-synaptic neuronal death in the grey matter after focal ischemia. While neurons are highly vulnerable to ischemia, potentially undergoing cell death after an ischemic duration longer than a just few minutes, oligodendrocyte (OLG) lineage cells and axons show somewhat greater initial resistant to ischemia but ultimately suffer various degrees of damage after prolonged ischemia exposure. WMI typically progresses over time from initial demyelination to OLG death and axonal injury, followed by varying degrees of OLG biogenesis, remyelination, axonal sprouting, and synaptic reconnections. The severity of ischemic WMI is a key predictor of poor long-term outcome in brain ischemia-related conditions. Despite its clinical importance, ischemic WMI remains significantly understudied in preclinical animal models, creating a weak link in the field of ischemic brain injury animal research. Many molecular and pathological mechanisms underlying post-ischemic WMI remain poorly understood, and currently, no specific treatments exist for WMI patients. This review aims to stimulate WMI research by providing updated knowledge, emerging hypotheses, and future research directions to foster understanding the mechanisms and to guide the development of the effective therapeutic strategies.

1. Introduction

The central nervous system (CNS) is made up of grey matter (GM) and white matter (WM). GM, found in the outer cortical portion of the brain, mainly consists of neuron cell bodies, dendrites as well as axons, and is responsible for processing information. WM, mostly distributed in the inner subcortical portion of the brain, is composed of axons, oligodendrocyte (OLG) lineage cells, including myelin producing OLGs and OLG precursors cells (OPCs), and other common brain cell types (microglia, astrocytes, and blood vessels). Most WM axons are wrapped with myelin sheaths, providing electrical insulation and increasing neural processing speed and efficiency. These axons form networks interconnecting neurons, transmitting information from GM to other parts of the nervous system.

Ischemic WM injury (WMI) can be caused by focal ischemia (stroke) affecting either the GM, WM, or both, as well as from a global reduction in blood flow affecting multiple brain regions, including both GM and WM. The incidence of WM stroke in subcortical regions of the brain can be as high as 20% of all stroke subtypes and 25% of all ischemic strokes (Gouw et al., 2008; Bailey et al., 2012).^1,2 Even if a stroke primarily affects GM, the surrounding WM can experience various degrees of damage (Wang et al., 2016).³ Ischemic WMI can lead to cognitive impairments, motor dysfunction, and other neurological deficits depending on the severity and ischemic brain regions (Wang et al., 2016).³

WMI involves damage to various WM components, including axons, myeline sheaths, OLGs, and OPCs. WMI is a key predictor of poor long-term functional outcomes in human ischemic stroke, both large and small (Wang et al., 2016; Marin and Carmichael, 2018; Marin et al., 2023; Bir et al., 2021; Li et al., 2021).^3–7 In untreated stroke patients with large vessel occlusion, an estimated 1.9 million neurons, 14 billion synapses, and 12 km (7.5 miles) of myelinated fibers are destroyed every minute (Saver, 2006).⁸ Roth et al. (2024)⁹ showed that post-stroke cognitive outcomes were associated with the preservation of long-range WM fibers beyond the lesion. Pendlebury et al. (1999)¹⁰ reported that WMI in the descending motor pathways strongly correlated with motor deficits. Moreover, Velly et al. (2018)¹¹ found that WMI detected by MRI was the strongest predictor of unfavorable neurological outcomes following cardiac arrest. Although WMI likely occurs in the acute phase in most brain ischemia patients at the cellular levels, only severe WMI is visible on magnetic resonance imaging (MRI) (Li et al., 2013; Fu et al., 2021; Chen et al., 2013; Xu et al., 2020; Sharma, 2024).^12–16 With the advent of advanced imaging techniques such as artificial intelligence (AI)-based tractography, WMI will likely be detected in varying severities of stroke (Kuwabara et al., 2024).¹⁷

Despite its high incidence and clinical importance in humans, WMI remains significantly understudied in animal brain ischemia models. At the writing of this review (December 2024), PubMed searches reveal only 1,194 publications for “animal ischemic white matter injury” compared to 54,350 for “animal ischemic brain injury,” highlighting the limited focus on WMI in preclinical studies. Consequently, many pathological mechanisms underlying ischemic WMI remain unknown. No viable therapy for WMI currently exists (Marin et al., 2023).⁵

2. Ischemic White Matter Injury (WMI)

Ischemic WMI can exhibit initial demyelination, OLG death, and axonal injury (Fig. 1), which is followed by inflammatory response. Ischemic axonal injury typically manifests as bulbous or swelling axonal segments (Fig. 1). Axonal injury can be primary, due to an ischemic insult in WM, or secondary to the neuronal death, OLG death, or loss of synaptic connections to post-synaptic neurons (Hu et al., 1998).¹⁸ Axonal synaptic connections are necessary for maintaining axonal health, providing vital adhesive and neurotrophic support (Weidner et al., 1999; Poplawski et al., 2018; Hinman, 2014).^19–21

Fig. 1.

Histopathological and schematic illustration of WMI after brain hypoxia-ischemia (HI) in adult mice. The mouse brain HI was induced by a combination of ligation of the left common carotid artery and systemic hypoxia (8% O₂). Mice (3-months old) were subjected to 30 min brain HI followed by 2-day recovery. In this model, WMI occurs in the ipsilateral corpus callosum (CC) and typically progresses from initial demyelination to OLG death and bulbous axonal (injury) segment formation. A: Histopathology of hematoxylin and eosin (H&E)-stained brain section at the striatal level to show the contralateral healthy and ipsilateral damaged WM in the CC area (squares). B: Schematic illustration of healthy myelin sheath, OLG, and axons. C: Schematic illustration of demyelination, OGL death, and axonal bulbous injury (arrows).

After global brain ischemia, WMI is often moderate and reversible in animal models. Chen et al. (2013)¹⁴ observed decreased myelination levels 7–14 days after ischemia, with recovery to normal levels by 28 days in a rat four-vessel occlusion (4VO) model, where both carotid and vertebral arteries are occluded for 10 minutes to simulate transient global brain ischemia. In focal brain ischemia such as rodent middle cerebral artery occlusion (MCAO) models, WMI often occurs in varying degrees. Pantoni et al. (1996)²² reported OLG swelling at 0.5-hour, significant OLG loss at 3 hours, and additional segmental myelinated axon swelling at 24 hours following permanent MCAO. Animal studies suggest that WM components (OLG lineage cells and axons) may be relatively tolerant to ischemia compared with GM neurons. Irreversible WMI typically occurs after prolonged or permanent focal ischemia in animal models.

Chronic cerebral hypoperfusion (CCH) is another condition that can preferentially lead to WMI. It occurs when the brain’s blood flow is chronically reduced, often due to narrowing or partial closure of cerebral blood vessels. This can be caused by small cerebral vessel disease or arteriosclerosis in larger vessels supplying the brain, such as the carotid arteries. CCH is a significant contributor to cognitive impairment, largely due to its tendency to cause preferential WMI (Rajeev et al., 2023).²³

Brain ischemia in combination with comorbidities such as hypoxia or hyperglycemia can also preferentially leads to WMI (Ferdinand et al., 2016; Ma et al., 2018; Jang et al., 2024).^24–26 Brain ischemia is often accompanied by hypoxia due to conditions such as pneumonia, aspiration and respiratory muscle dysfunction, and cardiac failure (Ferdinand et al., 2016).²⁴ Brain hypoxia-ischemia (HI) can result in a laminar pattern of cortical necrosis and delayed WMI (Hu et al., 2000; Takahashi et al., 1993).^27,28 MRI studies show that severe hypoxia-ischemia usually damages both GM and WM, while milder insults preferentially affect periventricular WM in both adults and neonates (Huang et al., 2008; White et al., 2013; Jang et al., 2024).^29,30,26 Pre-ischemic hyperglycemia may promote microglia/macrophage polarization toward a pro-inflammatory phenotype and impair white matter repair (Ma et al., 2018).²⁵ Fig. 1 schematically illustrates WMI after brain hypoxia-ischemia (HI) in adult mice. In this model, HI is induced by ligation of the left common carotid artery combined with systemic hypoxia (7.5% O2). Mice (3-months old) were subjected to 30 min brain HI followed by a 2-day recovery. WMI occurs in the ipsilateral corpus callosum (CC) and typically progresses from initial demyelination to OLG death and bulbous axonal (injury) segment formation. Modern resuscitation and advanced life support have increased survival rates among patients experiencing hypoxia-ischemia events, but many survivors still experience varying degrees of neurological deficits (Takahashi et al., 1993; Jang et al., 2024).^28,26 WMI is becoming increasingly recognized as a predictor of poor neurological outcomes in cardiac arrest patients (Hoiland et al. 2022, Velly et al. 2018).^31,11 These findings highlight the desperate need for understanding the underlying mechanisms and developing effective therapies against WMI.

3. OLG Lineage Cell Response to Brain Ischemia:

OLGs are terminally differentiated cells that have lost their ability to proliferate. They are responsible for producing myelin sheaths, which wraps most axons in WM, facilitates electrical impulse propagation, and protects axons from injury (Fig. 1). Each OLG can wrap multiple axons (Fernandez-Castaneda et al., 2016).³² OLG damage leads to demyelination (Martín-Lopez et al., 2024).³³ Li et al. (2021)⁷ reported that axonal fibers were relatively preserved in monkey WM infarct areas, while myelin was damaged in the early stages of permanent ischemic stroke, suggesting that myelin and OLG damage precedes axonal loss in WM ischemic stroke. Consistently, OLGs were found to be more susceptible to WM focal ischemia than axons (Sozmen et al., 2019; Martín-Lopez et al., 2024).^34.33 Current available studies support the hypothesis that OLG injury and demyelination in the early post-ischemic phase can be either reversible or irreversible, with OLGs appearing more vulnerable to WM focal ischemia than their associated axons.

Dead or dysfunctional OLGs can be replaced by newly generated OLGs through the differentiation of OPCs (El Waly et al., 2014).³⁵ OPCs comprise approximately 3–4% of cells in GM and 8–9% in WM throughout the brain (Fernandez-Castaneda et al., 2016; Beiter et al., 2022, Martín-Lopez et al., 2024).^32,36,33 Martín-Lopez et al. (2024)³³ reported that myelin basic protein and Black Gold II-stained WM were reduced at 2 and 5 days post-focal brain ischemia but returned to control levels by 21 days in a rat permanent MCAO model. Simultaneously, OPCs numbers increased significantly at 2 days and, then gradually declined from 5 to 21 days after permanent MCAO (Martín-Lopez et al., 2024).³³ Similar findings were reported by Sozmen et al. (2019)³⁴ in a mouse model of subcortical WM stroke.

Marin et al. (2023)⁵ found that an increase in newly differentiated OLGs in the peri-infarct area corresponded to the functional recovery in young mice (2–3 months old) but not in older mice (18–22 months old) in a mouse subcortical WMI stroke model. Sabo et al. (2017)³⁷ demonstrated that OLG Transcription Factor 1 (Olig1) is required for the OPC differentiation into mature OLGs after WMI, as olig-null neonatal mice showed significant hypomyelination after moderate neonatal hypoxia-ischemia. Increased OPCs after WMI may originate from the local parenchyma or the subventricular zone (SVZ) (Xing et al., 2014; Menn et al., 2006).^38,39 These studies suggest that newly generated OLGs can replace damaged OLGs through OPCs differentiation, potentially facilitating or supporting axonal and functional recovery after brain ischemia. This process may partially explain the recovery of sensorimotor function observed following initial deficits in rodent models of focal brain ischemia (Martín-Lopez et al., 2024).³³

4. Axonal Injury, Remodeling, and Reconnections After Brain Ischemia:

Axons are long, slender projections of neurons and can reach volumes up to ten thousand times greater than their parent neuronal somata. There are two types: myelinated and unmyelinated axons in the nervous system. Axons without myelin sheath, known as an unmyelinated axons, are more prevalent in the gray matter, particularly in the cerebral cortex and the autonomic nervous system. In contrast, myelinated axons, which wrapped in a myelin sheath, are the main component of WM, occupying approximately 87% of its space (Wang et al., 2008).⁴⁰ Neurons without functional axons are unable to transmit signals. Therefore, axonal injury may be the most critical element of WMI. Ischemic axonal injury can be caused by a direct ischemic insult within the WM, presynaptic axonal parent neuronal death, loss of synaptic connections due to post-synaptic neuronal death, or damage to OLGs or other axon-supporting cells such as astrocytes after brain ischemia. Additionally, inflammatory cells may contribute to WM integrity by engulfing or remodeling axons and other WM components, or by influencing axonal re-myelination (Lundgaard et al., 2014; Hemati-Gourabi et al., 2022).^41,42 However, the mechanisms underlying ischemic axonal injury remains poorly understood.

Axonal injury typically manifests as bulbous or swelling axonal segments, observed in the penumbral area after permanent MCAO in rats (Yam et al., 1998).⁴³ Similar bulbous axonal swellings were observed in WM without Purkinje neuronal loss after prolonged (24–48 hours) hypoxia in an organotypic cerebellar slice culture model (Cui et al. 2020).⁴⁴ These axonal swellings might often be accompanied by neurofilament light chain (NF-_l) and alpha-spectrin degradation, accumulation of shortened mitochondria and endolysosome-like structures, and activation of calpain mediated proteolysis (Cui et al., 2020).⁴⁴

Even though TBI is not the main focus of this review, it is relevant to mention that axonal injury is a central feature of TBI, as understanding common and different pathophysiological processes of axon injury among various diseases, such as between stroke and TBI, may provide better opportunities for developing interventions against WMI. Unlike stroke, which primarily disrupts cellular machinery and metabolism in cell bodies and their processes, TBI primarily damages cell processes, mostly long axons, through shearing and tearing forces, leading to widespread axotomy or diffuse axonal injury (DAI). A key feature of axonal injury after TBI is the impairment of axonal transport of organelles between the cell body and the axon terminals. This disruption causes organelles accumulation, resulting in bulbous axon segments known as “retraction balls” or “axonal bulbs,” at the injury site. TBI-induced axonal injury also initiates Wallerian-like degeneration via a self-destructing process, eventually resulting in degeneration of parent cell bodies (Armstrong et al., 2024).¹¹⁸ Furthermore, TBI, particularly its severe form, is often accompanied by tissue bleeding and swelling, resulting in secondary ischemic injury. Therefore, stroke and TBI may share some common features of cellular dysfunctions (Park et al., 2015; Park et al., 2013; Truettner).^119,123,124

Hinman (2014)²¹ described stroke-induced axonal injury in three phases: initial rapid axons loss, delayed axonal degeneration, and a recovery phase characterized by axonal sprouting and reconnection. Various molecular mechanisms have been implicated in axonal remodeling and synaptic plasticity after ischemic brain injury. Zhang et al. (2008)⁴⁵ found that new axonal projections may arise from neurogenesis in the subventricular zone (SVZ) and subgranular zone (SGZ). Wang et al. (2017)⁴⁶ observed increased levels of repulsive guidance molecule b and synaptophysin, suggesting their involvement in synaptic remodeling. Hayden et al. (2019)⁴⁷ noted upregulation of Mark4 in cortical neurons following axonal injury. Overman et al. (2019)⁴⁸ demonstrated that blocking ephrin-A5, combined with forced limb use, led to widespread axonal projection. Bu et al. (2021)⁴⁹ showed Rac1’s role in cognitive and sensorimotor recovery and noted increased GAP-43 expression promoting synaptogenesis. These studies collectively demonstrate the complex interplay of molecular mechanisms, axonal remodeling or reconnections, and neurogenesis in synaptic plasticity after brain ischemia.

While loss of axonal function may be partially compensated through axonal sprouting and synaptic reconnection, current studies are mostly observational or fragmented, lacking cohesive and mechanistic understanding. It is critical to investigate how to manage post-ischemic axonal sprouting and remodeling to promote functional and beneficial synaptic reconnections. Unregulated or abnormal reconnections can lead to undesirable outcomes such as post-stroke seizure or psychological disturbances (Myint et al., 2006).⁵⁰

5. Primary and Secondary Axonal Injury:

Compared to neuronal somata, axons, particularly those that are derived from ischemic penumbral or non-ischemic regions, appear more tolerant to acute ischemic events (Hinman, 2014).²¹ While the initial ischemic event may lead to directly and immediately damage to axons, i.e., primary axonal injury, many axons can survive the initial ischemic events but undergo delayed damage or degeneration, referred to as secondary or delayed axonal injury. Neuronal death in the ischemic zone interrupts transport of cargoes from the neuronal somata to the axons. Axons that are derived from dead neuronal somata after ischemia may survive for up to 36 hours, suggesting they may have some self-supporting materials needed for temporary survival (Beirowski et al., 2005).⁵¹ Synaptic terminals and their axons may become disconnected from their post-synaptic counterparts due to ischemic neuronal death. Some disconnected synaptic terminals and their axons may survive for days, weeks, or even months but may eventually degenerate, likely due to the lack of synaptic adhesive proteins and neurotrophic factors support (Park et al., 2013; Hinman, 2014).^52,21 Without appropriate treatment, this secondary axonal injury often worsens with time, contributing to long-term functional deficits.

The molecular pathways leading to secondary axonal injury after brain ischemia remains poorly understood. Most knowledge of secondary axonal injury is derived from abundant research of traumatic brain injury (TBI) or spinal cord injury (SCI), which supports that the secondary axonal injury can be caused by the deprivation of adhesive molecules and neurotrophic factors, disruption of axonal transport and axoglial contact, as well as secondary inflammatory response (Hinman, 2014; Hill et al., 2016).^21,53

Deprivation of neurotrophic factors and synaptic adhesion molecules (SAMs):

Neurotrophic factors and synaptic adhesion molecules (SAMs) play key roles in sustaining axonal survival, growth, function and regeneration (Fig. 2). SAMs are diverse families of proteins acting as molecular “glue,” bridging the gap or synaptic cleft between pre- and post-synaptic membranes. They play a crucial role in organizing synaptic signaling and scaffold complexes, facilitating synaptic transmission and the formation of new synapses. SAMs include neural cell adhesion molecule (NCAM), cadherin, cell synaptic cell adhesion molecule (SynCAM), neurexin, neuroligin, eph receptor and its binding ligand ephrin, receptor-like protein tyrosine phosphatase (RPTP), interleukin-1 receptor accessory protein-like 1 (IL1RAPL1), amyloid precursor protein (APP) and APP-like protein 1 (APLP1), etc. Each family of SAMs comprises several subtypes. Some SAMs are also present at axonal and glial contact sites or on other types of cells (Danos et al., 2025; Coleman, 2011).^54,55 The importance of SAMs is highlighted by the fact that their deficiency leads to neurological deficits, as seen in stroke, TBI, and various other neurological disorders.

Fig. 2.

Schematic diagram of (i) presynaptic vesicle (SV) cycle, (ii) trans-synaptic adhesion molecules, and (iii) peri-synaptic neurotrophins systems. (i) The SV cycle involves exocytic neurotransmitter release into the synaptic cleft, followed by endocytic SV recycling for subsequent SV release in a calcium-dependent manner. Fusion of SV with presynaptic membrane is mediated by, among others, the core fusion machinery consists of N-ethylmaleimide-sensitive factor (NSF), soluble NSF attachment protein (SNAP), and SNAP receptor (SNARE). After fusion and release of neurotransmitter, SNAREs on SV are dissociated by NSF-SNAP ATPase activity. The SV membrane is then endocytosed to form an endosome that are eventually converted into a new SV; (ii) Synaptic adhesion molecules (SAMs) include family members of neural cell adhesion molecule (NCAM), cadherin, synaptic cell adhesion molecule (SynCAM), neurexin, neuroligin, eph and its binding ligand ephrin, receptor protein tyrosine phosphatase (RPTP), interleukin-1 receptor accessory protein-like 1 (IL1RAPL1), as well as amyloid-beta precursor protein (APP) and APP-like protein 1 (APLP1), etc. (iii) Neurotrophic factor receptors include tropomyosin receptor kinases-A (TrkA), -B (TrkB), and -C (TrkC) on neuronal membranes, particularly in peri-synaptic membrane areas. Binding of neurotrophic factors (also known as neurotrophins) to their receptors activate several signaling pathways to regulate cellular and axonal growth and synaptic plasticity.

Neurotrophic factor receptors like the tropomyosin-related kinases-A (TrkA), -B (TrkB), and -C (TrkC) are protein tyrosine kinases on cell membranes of neurons, particularly in peri-synaptic regions. Upon binding to their corresponding neurotrophic factors (e.g., TrkA with nervous growth factor (NGF), TrkB with brain-derived neurotrophic factor (BDNF) or neurotrophin-4/5 (NT4/5), or TrkC with NT-3), these receptors regulate cellular and axonal growth and synaptic plasticity via various signaling pathways. Deficiency in neurotrophic signaling impairs neuronal viability, axonal growth and regeneration, and neurotransmission. For example, BDNF is crucial for maintaining axon survival and regeneration after TBI and SCI (Nagahara and Tuszynski, 2011; Li et al., 2022; Tuszynski, 2024).^56–58 Brain BDNF levels are typically reduced after stroke and are associated with poorer functional outcomes and increased risk of post-stroke depression (Chaturvedi et al., 2020).⁵⁹ Exercise has been shown to elevate circulating BDNF level in stroke survivors (Ashcroft et al., 2022).⁶⁰ Delivery of BDNF via adeno-associated virus serotype 2 (AAV2)-BDNF vector promotes axon plasticity and restore diaphragm function following SCI (Charsar et al., 2019).⁶¹ In animal models of Alzheimer’s disease, BDNF exhibits potent therapeutic effects that include prevention of cell death, stimulation of neuronal function, improvement in synaptic markers, and improvements in learning and memory (Tuszynski, 2024).⁵⁸

Disruption of axonal transport and synaptic vesicle (SV) cycle:

Axonal survival and functions rely on bidirectional transport of new or recycle damaged materials and organelles between cell bodies and axonal terminals. Anterograde transport is mediated by a wide range of kinesin superfamily motor proteins, while the retrograde transport is driven by the dynein motor complex. For example, synaptic vesicle precursors (SVPs) are synthesized in the neuronal soma and transported along microtubules to their axonal terminals, where they are converted into synaptic vesicles (SVs) for storing and releasing neurotransmitters. Another example is the retrograde trafficking of endosomes containing neurotrophic factors and their receptors from axonal terminals to the soma, which is required for neurotrophic signaling (Berth et al., 2023).⁶² Significant insights into dysfunctional axonal transport and its role in axonal injury have come from studies of TBI, SCI, aging, and chronic neurodegenerative disorders. However, disruption of axonal transport can be initiated by different mechanisms, affecting varied proteins and cargoes among TBI, SCI, and different neurodegenerative diseases (Berth et al., 2023).⁶²

Axonal bidirectional transport of endosomes and autophagosomes is essential for axonal survival and function and is operated by several protein complexes. For example, Rab5 and Rab7 regulate early and late endosome trafficking, respectively (Osterli et al., 2025).⁶³ Axonal LAMP1-positive structures, particularly those located distally from the soma, are likely early or late endosomes, as indicated by their relatively higher luminal pH and low level of degradative enzymes. These early or late endosomal structures are mainly responsible for retrograde transport of damaged materials and organelles to the soma-surrounding area where they fuse with degradation-active lysosomes for degradation (Osterli et al., 2025; Roney et al., 2022; Hu et al., 2021).^63–65 Adaptor proteins ARL8B and SKIP help anterograde transport, whereas adaptor proteins JIP3 and JIP4 assist retrograde transport (Berth et al., 2023).⁶²

Axonal SVPs and endosomes travel along axonal microtubules and, upon reaching their destinations, must fuse with their target membranes to exert functions (Osterli et al., 2025; Hu et al., 2021).^63,65 This bilayer lipid fusion between the SVP or endosomal membrane and the target membrane is mediated by several protein complexes. The core protein complex is commonly composed of N-ethylmaleimide-sensitive factor (NSF), soluble NSF attachment proteins (SNAPs), and SNAP receptors (SNAREs). SNAREs are membrane fusion proteins and can be divided into vesicle (v)-SNAREs and target (t)-SNAREs. Interactions between v- and t-SNAREs from two opposite lipid membranes form a trans-SNARE complex, bringing the two membranes nearby for their phospholipid membranes to merge into a single membrane (Fig. 2). After fusion, the SNAREs convert into an inactive stable cis-complex, which must be reactivated via NSF’s adenosine triphosphatase (ATPase)-mediated dissociation, restoring them to individual, active trans-conformations for the next round of membrane fusion. During this SNARE reactivation process, SNAP acts as both an adaptor to bridge SNAREs with NSF ATPase, and stimulator of NSF ATPase (Fig. 2) (Osterli et al., 2025; Hu et al., 2021).^63,65 Mammalian cells express only a single form of NSF (Whiteheart et al., 1994).⁶⁶ Therefore, deficiency in NSF halts membrane fusion. In comparison, there are three subtypes of SNAPs isoforms - alpha, beta, and gamma, and approximately 60 SNAREs in humans. While NSF and SNAPs are relatively universal, SNAREs exhibit subcellular organelle-specific for intracellular membrane fusion events. For example, syntaxin-1 and synaptosome-associated proteins 25kDa or 29 kDa (SNAP-25 or −29) are known t-SNAREs on presynaptic membrane, and synaptobrevin-1 or −2, also known as vesicle-associated membrane protein-1 or −2 (VAMP-1 or −2), are known v-SNAREs on SV membrane (Fig. 2).

In addition to the conversion of SVPs into SV as described above, SV can also be recycled via exocytic neurotransmitter release into the synaptic cleft, followed by endocytic SV cycle for the SV reuse. Exocytic release is mediated by protein complexes in a calcium-dependent manner (Fig. 2). Specifically, interactions between v-SNAREs (synaptobrevin-1 or −2) on SV membrane and t-SNAREs (syntaxin-1 and SNAP-25 or −29) on the presynaptic membrane bring SV and presynaptic membranes together (Fig. 2). The SV membrane protein synaptotagmin is a calcium sensor, and, upon binding calcium ions, triggers conformational changes in the synaptic SNARE complex to execute membrane fusion. After fusion, SNAREs form inactive complexes that must be dissociated or reactivated by NSF ATPase activity via ATP hydrolysis to ADP (Fig. 2). Once SNARE reactivation, the segment containing SV proteins is endocytosed by clathrin/AP2-dependent or -independent events to form an endosome that eventually converts into recycled SV.

Axonal spheroids:

Axonal spheroids are abnormal bubble-like swellings appearing along axonal trucks, branches, and terminals. They are a common feature of axonal injury seen in various brain diseases but may arise through different mechanisms. These mechanisms often involve alteration in microtubule dynamics, disruption of neurofilament organization, and modifications in actin filament networks. Axonal spheroids contain different disease-associated proteins and organelles among various neurodegenerative diseases, as well as among TBI, SCI, and stroke (Berth et al., 2023).⁶² For example, synaptophysin and amyloid precursor protein (APP) are significantly accumulated in axonal spheroids after TBI (Gudi et al., 2017).⁶⁷ The accumulation of endosomal and autophagosome intermediates in axonal spheroids is a characteristic feature seen in various neurodegenerative diseases, such as Alzheimer’s disease or some lysosomal (deficient) storage diseases, such as Niemann-Pick type C (NPC) disease (Berth et al., 2023).⁶² These neurodegenerative axonal spheroids are often referred to as axonal dystrophy or dystrophic neurites. Axonal transport falls dramatically with aging, affecting fast and slow transport of cargoes both in retrograde and anterograde directions, which can also lead to axonal spheroid formation (Berth et al., 2023).⁶² Understanding the mechanisms underlying the formation of disease-specific axonal spheroids may facilitate the development of corresponding therapeutics among axonal injury related diseases.

Dysfunctional axonal signaling pathways:

Axonal injury can trigger cellular responsive signaling pathways, leading to changes in gene expression, axonal degeneration, and axonal remodeling. An axonal survival pathway involving nicotinamide mononucleotide adenylyltransferase 2 (NMNAT2) and sterile α and Toll/interleukin-1 receptor motif–containing protein 1 (SARM1), which has drawn significant interests recently (Coleman, 2011).⁵⁵ NMNAT2 is synthesized in the soma and transported into the axon. In glutamatergic neurons, NMNAT2 is required for axonal survival (Coleman, 2011).⁵⁵ When NMNAT2 axonal transport is interrupted, NMN levels rise while NAD levels decrease in the axon, resulting in activation of the NAD degrading enzyme SARM1 and subsequent axonal degeneration (Colman, 2011).⁵⁵ Activation of SARM1 leads to axonal degenerative changes whereas SARM1 deficiency in mice shown to reduce axonal damage and improves cognitive outcome after traumatic brain injury (Geisler et al., 2019; Maynard et al., 2020; Wang et al., 2025).^68–70 Recently, the study of Wang et al (2025)⁷⁰ show that SARM1 deficiency in mice confers robust axonal and neuronal cell body protection in two different animal models of ischemic stroke.

In addition to the SARM1 pathway, several other signaling cascades play important roles in axonal survival and function, including the phosphatase and tensin homolog deleted on chromosome 10 (PTEN), extracellular signal-regulated kinase (ERK), phosphoinositide 3-kinase (PI3K)/Akt, dual leucine zipper-bearing kinase (DLK, also known as MAP3K12), c-Jun N-terminal kinase (JNK), Rho families of guanine triphosphatase (GTPase) such as GEF-H1 and Cool-2/αPix, etc., (Goodwani et al. 2020; Hausott et al., 2022; Sabirzhanova et al., 2013; Luo et al., 2014).^71–74 PI3 kinases, Erk1/2, Erk5, DLK, JNK, GEF-H1, and Cool-2/αPix are activated when axons are exposed to stress or stimulated by neurotrophins signaling, supporting axonal survival or repair (Coleman, 2011).⁵⁵ Although these pathways are also activated after brain ischemia, their roles and underlying mechanisms remain to be further studied (Luo et al., 2014; Hinman 2015).^74,75

6. WMI Mechanisms:

Mechanisms underlying ischemic brain injury may vary across different brain cell types and various conditions. Common mechanisms underlying ischemia-induced cell death, including mitochondrial dysfunction, adenosine triphosphate (ATP) depletion and release, oxidative stress, excitotoxicity, inflammation, and various forms of cell death (necrosis, apoptosis, etc.), have been postulated to play key roles in WMI (Wang et al., 2016; Baltan, 2014; Sozmen et al., 2019; Brunelli et al., 2023).^3,76,34,77 The complex interactions among glial cells, axons, and axon-associating neurons add several layers of complexities of WMI mechanisms (Baltan, 2016).⁷⁸ Although these post-ischemic WMI hypotheses may be viable, they are mostly postulated based on common cell injury mechanisms and lack thorough and specific confirmations for WMI with animal models.

Mitochondrial dysfunction, ATP depletion and release, and oxidative stress:

Mitochondria travel along axons between the soma and presynaptic terminals and are relatively concentrated in the presynaptic nerve terminal for supporting synaptic activities (Li et al., 2022).⁷ Axons, particularly presynaptic terminals, consume a significant amount of ATP to sustain neurotransmission, posing exceptional challenges in maintaining presynaptic energy homeostasis. Ischemic conditions reduce ATP in axon and presynaptic terminal, resulting in widespread membrane depolarization (Li et al., 2022).⁷ Persistent ATP depletion can lead to irreversible axonal injury. Following ischemia, reperfusion achieved either through re-cannulation of the occluded vessel after transient ischemia with reperfusion or through the establishment of collateral circulation after permanent or prolonged ischemia, can restore membrane polarization and energy status to varying degrees. Glucose transporter (GLUT) 3, the primary glucose transporter on axonal and presynaptic membranes for the local ATP production, is significantly upregulated in penumbral regions after focal brain ischemia (Li et al., 2022; Ginsberg, 1990).^7,79

Mitochondrial biogenesis, trafficking, and quality control are essential for maintaining healthy mitochondria in axonal trunks, branches, and terminals. New axonal and synaptic terminal mitochondria are primarily generated in the neuronal soma and transported anterogradely to axons, while damaged mitochondria are retrogradely transported to the soma for degradation in lysosomes (Li et al., 2022; Osterli et al., 2025).^7,63 However, animal studies of axonal mitochondrial trafficking under simulated ischemic conditions are limited, with most conducted in cell cultures. For example, Zheng et al. (2019)⁸⁰ reported that, after oxygen and glucose deprivation (OGD), axonal mitochondria in cultured cortical neurons lost anterograde motility but showed increased retrograde transport to the neuronal soma for mitophagy.

Mitochondrial dysfunction also causes overproduction of reactive oxygen species (ROS), leading to oxidative stress. Although oxidative stress likely impacts all post-ischemic brain cells, it particularly affects ischemia susceptible populations like neurons and OLGs. While neuronal oxidative stress has been well studied, WMI after brain ischemia-specific studies remain sparse (Siesjö et al., 1999).⁸¹ Stahon et al. (2016)⁸² found that WM in aged mice exhibits fewer axonal mitochondria, lower ATP levels, higher levels of protein nitration and lipid peroxidation, and altered levels of mitochondrial fusion and fission proteins. Some studies have suggested that interventions targeting oxidative stress might preserve axonal mitochondrial function and promote brain functional recovery after brain ischemia (Stahon et al., 2016; Baltan et al., 2011; Baltan, 2012).^82–84

Dysfunctional intracellular trafficking and degradation systems:

Dysfunction of intracellular degradation systems may contribute to WMI after brain ischemia. Eukaryotic cells have two major cellular degradation mechanisms: the ubiquitin-proteasome system (UPS) and autophagy. Both systems are essential for maintaining cellular homeostasis (Hu et al., 2021; Osterli et al., 2025).^65,63 The UPS degrades intracellular soluble ubiquitin-tagged proteins, while macroautophagy, often simply known as autophagy, eliminates insoluble protein aggregates and damaged organelles (Hu et al., 2021; Osterli et al., 2025).^65,63

The dysfunctional UPS has been widely reported in post-ischemic cells, particularly in neuronal soma and apical dendritic areas (Hu et al., 2021; Osterli et al., 2025).^65,63 However, its role in post-ischemic WMI is less studied. Some studies have suggested that UPS dysfunction might contribute to WMI in animal stroke models. For instance, Minis et al. (2019)⁸⁵ reported that deficiency in Proteasomal Inhibitor of 31kD (PI31), a proteasome-binding protein, causes axon degeneration, neuronal loss, and neurological dysfunction in mice. Liu et al. (2019)⁸⁶ found that increase activity of ubiquitin c-terminal hydrolase L1 (UCHL1), which regulates redox state and misfolded proteins degradation in neurons and axons, contributed to WMI in a mouse stroke model.

Dysfunctional autophagy is widely reported in post-ischemic neurons (Hu et al., 2021; Osterli et al., 2025).^65,63 Brain ischemia causes protein misfolding and aggregation, and damages virtually all subcellular organelles in both GM and WM cells (Hu et al., 2021; Osterli et al., 2025).^65,63 Damaged components may be either locally packed into autophagosomes along axons or directly transported to the soma for lysosomal degradation (Yang et al., 2022).⁸⁷ Under normal conditions, autophagosomes are mostly transported to the soma for degradation, but this process is impaired after axonal injury (Luo et al., 2024).⁸⁸ Zheng et al. (2019)⁸⁹ found that axonal mitochondrial elimination persisted in Atg7fl/fl;nes-Cre mice, suggesting that damaged mitochondria might not necessarily form axonal autophagosome to be transported retrogradely to the soma for degradation. The deletion of Atg5 in OLGs impairs functional recovery after brain ischemia (Saraswat et al., 2018).⁹⁰ Inhibition of general autophagy reduces both the number and length of myelin segments on axons (Belgrad et al., 2020).⁹¹ While evidence suggests that dysfunctional autophagy may play a key role in WMI after brain ischemia, detailed in vivo studies remain limited, highlighting an important area of future research.

Apoptosis and necrosis:

Ischemia, simulated ischemia (e.g., OGD), or ischemia-like conditions (e.g., hypoxia) may trigger both caspase-dependent apoptosis and cytotoxic edema-induced necrosis in OLG, contributing significantly to WMI (Shibata et al., 2000; Dewar et al., 2003).^92,93 OLGs are particularly sensitive to energy deprivation due to their high energy demands and role in axonal energy supply (López-Muguruza et al., 2023).⁹⁴ Severe and prolonged ischemia likely leads to OLG ATP reduction, depolarization, cellular edema, and eventually cell rupture or necrosis (Dewar et al., 2003).⁹³ In moderate or transient ischemia, OLGs may either recover or undergo regulated cell death (López-Muguruza et al., 2023; Rosko et al., 2019; Huang et al., 2023).^94–96 Furthermore, overactivation of glutamate receptors may induce OLG apoptosis involving caspases 3, 8, or 9 (Huang et al., 2023).⁹⁶ Cell culture studies showed that ATP release into the interstitial extracellular space may trigger OLG apoptosis through P2X7 receptors activation in a Ca²⁺-dependent manner (Matute et al., 2007; Schädlich et al., 2023).^97,98 However, detailed investigations into OLG death mechanisms in animal brain ischemia models remains limited, despite being crucial for developing targeted therapeutic interventions to mitigate WMI.

Excitotoxicity and calcium permeability:

α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl D-aspartate (NMDA) receptors are two major glutamate receptors expressed not only on synaptic membranes but also on axons and OLG lineage cells (Kolodziejczyk et al., 2010).⁹⁹ Axonal ion channels, such as calcium channels, play a key role in action potential initiation, propagation, and neurotransmitter release (Stirling et al., 2010).¹⁰⁰ Calcium signaling promotes actin-dependent myelin sheath extension, myelin formation, OLG maturation, and OLG communication with axons (Iyer et al., 2024).¹⁰¹

During energy deprivation, dysfunctional glutamate transporters on axonal and OLG membranes may increase non-vesicular glutamate release and calcium influx into axoplasm, exacerbating WMI (Baltan, 2016).⁷⁸ Activation of glutamate receptors can lead to an acute increase in intracellular calcium after TBI, which has been suggested contributing to axonal injury (Hill et al., 2016).⁵³ Studies using OGD treatment in isolated mouse optic nerve preparations shown that WMI can be reduced by removing extracellular Ca²⁺ or blocking reverse Na⁺/Ca²⁺ exchange in young mice (<3 months), but not in older mice (>12 months) (Baltan et al., 2008 and 2016).^102,78 Similarly, immature OLG cultures were sensitive to OGD injury, which was preventable by removing Ca²⁺ or using AMPA/kainate receptor antagonists (Fern et al., 2014).¹⁰³ Al-Griw et al. (2021)¹⁰⁴ hypothesized that OGD damages WM via AMPA and NMDA receptor signal activation, enhanced OPC proliferation and maturation, and suppressed inflammation in the ex vivo cerebellar slice models. However, the role of excitotoxicity in WMI during in vivo brain ischemia remains controversial. For example, Yam et al. (2000)¹⁰⁵ showed that the NMDA receptor antagonist MK-801 failed to significantly reduce axonal injury in a cat stroke model. This discrepancy highlights the need for further investigation into the role of excitotoxicity in WMI using animal brain ischemia models, as the simulated ischemia studies in ex vivo or cell culture systems have yet to be proven in animal studies.

Inflammation:

Inflammation is widely observed in WMI areas following brain ischemia, mirroring the secondary inflammatory response observed in other post-ischemic brain regions. Choi et al. (2016)¹⁰⁶ reported that chronic cerebral hypoperfusion in Wistar rats led to diverse cognitive impairments, WM inflammation, loss of OLGs and myelin density, structural derangement at the nodes of Ranvier, and disintegration of white matter tracts. Shkirkova et al. (2024)¹⁰⁷ found that toll-like receptor 4 (TLR4) knockout mice exhibited attenuated WM degradation after chronic cerebral hypoperfusion. Jalal et al. (2012)¹⁰⁸ reported decreased myelin basic protein level, OLG apoptosis, increased inflammatory cells densities, and higher levels of tumor necrosis factor-α in the WM of spontaneously hypertensive rats following unilateral carotid artery occlusion. Additionally, Moxon-Emre et al. (2010)¹⁰⁹ observed activated microglia/macrophages surrounding axon bundles after brain ischemia. Ma et al. (2018)²⁵ observed that pre-ischemic hyperglycemia may increase microglia/macrophage polarization toward a pro-inflammatory and impair white matter repair, thus augmenting post-ischemic WMI. While current evidence clearly demonstrates the presence of inflammation in WMI areas, its exact role of inflammation, whether detrimental or beneficial, remains unclear. Inflammation likely plays a dual role in WMI following brain ischemia, potentially contributing to both damage and repair or remodeling processes. However, our current understanding of these complex interactions is far from complete.

7. WMI Treatments:

WMI is a common clinical feature of stroke and is associated with significantly more severe function deficits. However, few therapeutic studies have been specifically designed to target stroke-induced WMI. One key reason may be that the mechanism underlying stroke-related WMI remain poorly understood. In comparison, several animal studies have been conducted on WMI in models of neonatal hypoxia-ischemia injury and TBI.

A recent study demonstrated that a chemokine C-X-C motif receptor-2 (CXCR2) antagonist SB225002 reduced WMI and mitigated functional deficits in a rat model of cerebral palsy (Kitase et al., 2025).¹²⁰ Furthermore, cell-based therapies, such as neural stem cells (NSCs; both adult and embryonic) and mesenchymal stromal cells (MSCs), have reported efficacy in pre-clinical models of TBI. For example, MSC treatment has been reported to improve myelin basic protein (MBP) expression and white matter fiber integrity through anti-inflammatory, immunomodulatory, and trophic effects across adult stroke, multiple sclerosis, and neonatal brain injury in animal models (Vaes et al., 2019; Tan et al., 2022; Otero-Ortega et al., 2015).^113–115 In addition, transplantation of oligodendrocyte progenitor has shown promise in enhancing remyelination in mouse neonatal hypoxia models (Wang et al., 2013).¹²¹ A nature compound, oxysterol 20α-hydroxycholesterol, found in human breast milk, was shown to promote oligodendrogenesis via Sonic Hedgehog signaling and to improve motor outcomes in a neonatal WMI mouse model (Lehnerer et al., 2024; Chao et al., 2023).^116,117

Combination therapies that target different mechanisms of WMI has also been explored in preclinical TBI models (Wang et al., 2023).¹²² For example, combined treatment with apocynin, tert-butylhydroquinone (tBHQ), and Salubrinal reduced lesion volume and improved functional outcomes in a rat TBI model. Taken together, although several promising results have been reported in animal models of TBI and neonatal hypoxia-ischemia, clinical trials targeting WMI have not yet demonstrated clear benefits, indicating that further research is needed (Armstrong et al., 2024).¹¹⁸

8. Concluding Remarks:

The translation of scientific discoveries from preclinical animal research to clinical intervention is an ultimate goal of biomedical research. However, this translation in preclinical stroke research has been particularly challenging, with no brain cell protective intervention successfully moving from animal studies into the clinical application (Lyden, 2021).¹¹⁰ Neurological deficits remain a primary concern of stroke patients. WMI has shown a stronger correlation with post-ischemic neurological deficits than GM injury (Baltan et al., 2016).⁷⁸ Furthermore, WMI is increasingly recognized as a predictor of poor neurological outcomes in stroke and cardiac arrest patients (Hoiland et al. 2022, Velly et al. 2018).^31,11 However, although rodent models of transient MCAO with reperfusion have been widely used to investigate stroke mechanisms and interventions, they produce only limited WMI (Hinman, 2014).²¹ Despite its clinical significance, post-ischemic WMI has been significantly understudied in preclinical animal models, resulting in a significant scientific knowledge gap.

Several challenges exist in studying WMI in animal brain ischemia models. One obstacle is the relative tolerance of WM to ischemia, i.e., with limited WMI, in the commonly used mouse monofilament MCAO model, where OLG loss and axonal damage become apparent only after prolonged or permanent focal ischemia (Hinman, 2014; Pantoni et al., 1996).^21,22 However, the mortality rate can be significantly high that impedes the analysis of WMI after prolonged or permanent focal ischemia in the mouse MCAO model (Yuan et al., 2018).¹¹¹ Another challenge is the significantly smaller ratio of WM and GM in rodents compared to humans. For example, this ratio is 10:90 in mice versus 60:40 in humans (Krafft et al., 2012).¹¹² Furthermore, WMI involves diverse cellular components and injury mechanisms (Marin et al., 2023)⁵ and can occur as either a primary or secondary to the death of the parent and connecting neurons after brain ischemia. Different types of ischemic insults also lead to varying degrees of WMI. For instance, WMI may be less severe in young healthy mice, subjected to MCAO stroke, compared to a local WM stroke, MCAO stroke in the elderly mice, MCAO stroke in young healthy mice with comorbidities of pre-ischemic hypoxia, hyperglycemia, CCH, etc.

To overcome these challenges and advance our understanding of WMI, researchers should consider: (i) selecting appropriate animal models that better reflect human WM proportions and vulnerability; (ii) utilizing state-of-the-art, high resolution, multi-channel, and AI-based imaging analysis to track spatial and temporal changes in various WM components, such as 3-dimensional (3D) confocal microscopy, 3D block face electron microscopy (SBF-SEM), and 3D magnetic resonance imaging (MRI) tractography; and (iii) studying various aspects of WMI in concert to understand the processes and interactions of ischemic WMI components in animal models.

Understanding the underlying mechanisms of ischemic WMI is critical for developing targeted interventions. One key manifestation of WMI after brain ischemia is the formations of bulbous axons, suggesting axonal transport dysfunction. These structures, known as axonal spheroids or dystrophic axons or neurites, are associated with axonal degeneration in numerous neurodegenerative diseases, including Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, hereditary spastic paraplegia, amyotrophic lateral sclerosis, and Charcot-Marie-Tooth disease (Berth et al., 2023).⁶² However, it remains unclear whether the mechanisms underlying bulbous axons formation in ischemic WMI are similar to or different from those in chronic neurodegenerative diseases. Another example is the loss of adhesion molecules due to post-synaptic neuronal death after stroke. Potential therapeutic approaches could include the administration of adhesion or neurotrophic molecules or the use of cell-based therapies to establish new synaptic connections and neuroplasticity. However, caution may be necessary, as aberrant new synaptic connections might lead to unwanted neuronal functions, such as post-stroke seizure or mental health disorders.

In summary, ischemic WMI remains a critical concern in various clinical settings. Although WMI has emerged as a superior predictor of worse long-term outcomes following ischemic events, it remains significantly understudied in preclinical animal brain models. Recent progress in preclinical WMI studies has provided valuable insights into appropriate animal models and advanced our ability to analyze the temporal and spatial contributions of various WM components and signaling pathways related to WMI. As our understanding of WMI expands, it holds the potential to bridge the gap between preclinical findings and clinical applications, addressing a critical need in the treatment of brain ischemia. These new advancements are expected to foster more robust mechanistic research and facilitate the development of targeted and effective interventions against ischemic stroke injury.

Highlights.

• Ischemic white matter injury (WMI) occurs in many clinical settings, can be caused either by primary WM ischemia or secondary axonal injury due to cell injury after focal ischemia in the grey matter.

• Compared to grey matter injury, WMI is more strongly associated with worse functional outcome, remains significantly understudied in preclinical brain ischemia-related animal models, and thus represents a weak link of ischemic brain injury research.

• A typical sequence of ischemic WMI events includes initial demyelination, OLG death, and then axonal injury, which are often followed by varying degrees of OLG biogenesis, remyelination, axonal sprouting, and synaptic reconnections.

• Ischemic WMI occurs preferentially in aged animals after brain ischemia and ischemic animals with comorbidities such as hypoxia and hyperglycemia.

Abbreviations

APP

amyloid precursor protein

APLP1

APP-like protein 1

ARL8B

ADP-ribosylation factorlike protein 8B

BDNF

brain-derived neurotrophic factor

CCH

chronic cerebral hypoperfusion

DLK

dual leucine zipper-bearing kinase

EE

early endosome

EM

electron microscopy

ERK

extracellular signal-regulated kinase

GAP-43

Growth Associated Protein 43

GLUT

glucose transporter

GM

grey matter

HI

hypoxia-ischemia

IL1RAPL1

interleukin-1 receptor accessory protein-like 1

JIP3

c-Jun N-terminal kinase-interacting protein 3

JIP4

c-Jun-amino-terminal kinase-interacting protein 4

JNK

c-Jun N-terminal kinase

L

lysosome (terminal)

LAMP

lysosomal associated membrane protein

NCAM

neural cell adhesion molecule

LE

late endosome

NF-L

neurofilament light chain

NMNAT2

nicotinamide mononucleotide adenylyltransferase 2

NSF

N-ethylmaleimide sensitive factor

NT-3

neurotrophin-3

NT4/5

neurotrophin-4/5

MCAO

middle cerebral artery occlusion

OGD

oxygen and glucose deprivation

OLG

oligodendrocyte

OPC

oligodendrocyte precursor cell

PI31

proteasomal inhibitor of 31k

PTEN

phosphatase and tensin homolog deleted on chromosome 10

Rab5

rasrelated protein 5

Rab7

ras-associated protein 7

ROS

reactive oxygen species

RPTP

receptorlike protein tyrosine phosphatase

SAM

synaptic adhesion molecules

SARM1

Sterile alpha and Toll/Interleukin-1 receptor motif containing 1

SBF-SEM

Serial Block-Face Scanning Electron Microscopy

SCI

spinal cord injury

SGZ

subgranular zone

SKIP

SifA and kinesin-interacting protein

SNAP

soluble NSF attachment protein

SNARE

SNAP receptor

SV

synaptic vesicle

SVP

synaptic vesicle precursors

SVZ

subventricular zone

SynCAM

synaptic cell adhesion molecule

TBI

traumatic brain injury

TLR4

toll-like receptor 4

TrkA

tropomyosin-related kinases-A

TrkB

tropomyosin-related kinases-B

TrkC

tropomysin-related kinases-C

UCHL1

ubiquitin c-terminal hydrolase L1

UPS

ubiquitin-proteasome system

WM

white matter

WMI

White matter injury

WML

white matter lesion