Crosstalk between cerebral endothelium and oligodendrocyte

Abstract

It is now relatively well accepted that the cerebrovascular system does not merely provide inert pipes for blood delivery to the brain. Cerebral endothelial cells may compose an embedded bunker of trophic factors that contribute to brain homeostasis and function. Recent findings suggest that soluble factors from cerebral endothelial cells nourish neighboring cells, such as neurons and astrocytes. Although data are strongest in supporting mechanisms of endothelial-neuron and/or endothelial-astrocyte trophic coupling, it is likely that similar interactions also exist between cerebral endothelial cells and oligodendrocyte lineage cells. In this mini-review, we summarize current advances in the field of endothelial-oligodendrocyte trophic coupling. These endothelial-oligodendrocyte interactions may comprise the oligovascular niche to maintain their cellular functions and sustain ongoing angiogenesis/oligodendrogenesis. Importantly, it should be noted that the cell–cell interactions are not static—the trophic coupling is disturbed under acute phase after brain injury, but would be recovered in the chronic phase to promote brain remodeling and repair. Oligodendrocyte lineage cells play critical roles in white matter function, and under pathological conditions, oligodendrocyte dysfunction lead to white matter damage. Therefore, a deeper understanding of the mechanisms of endothelial-oligodendrocyte trophic coupling may lead to new therapeutic approaches for white matter-related diseases, such as stroke or vascular dementia.

Introduction

Brain physiology and pathophysiology are very complex. Several different types of brain cells may work together to maintain, remodel, and repair our brain functions. In this regard, the concept of “neurovascular unit” was raised as a new paradigm to understand the pathology of central nervous system (CNS) diseases, such as stroke [1–5]. This modular concept is defined at an intercellular level that comprises dynamic interactions between cerebral endothelial cells, glia, neurons, and the extracellular matrix. Dysfunctional crosstalk within the neurovascular unit may lead to multiple aspects of acute pathophysiology in CNS diseases. For example, impaired glutamate release-reuptake mechanisms in neurons and astrocytes can amplify excitotoxicity [6]. Perturbed signaling between cerebral endothelium and astrocytes (and sometimes pericytes) can disrupt blood–brain barrier integrity [4]. In addition, dysfunctional coupling between neuronal activation and vascular responses can also accelerate deleterious spreading depression [7]. Moreover, disordered signaling between all neurovascular and gliovascular elements may underlie the evolution of neuroinflammation and cell death [8].

Besides the importance of cell–cell interaction, the neurovascular unit also emphasizes the distinct roles of cerebral endothelium on brain functions. Although cerebral vascular system is the major constituent of the brain, the cerebrovascular system was traditionally thought as a passive conduit for blood stream. However, recent research has proposed that this system plays more active roles in maintaining the CNS homeostasis. As mentioned, cerebral endothelial cells form the blood–brain barrier (BBB) with astrocytes and pericytes. The BBB constitutes anatomical, physiochemical, and biochemical barrier that controls the exchange of materials between blood, brain, and cerebrospinal fluid. BBB breakdown due to endothelial dysfunction is frequently associated with a myriad of neurological pathologies, including chronic CNS diseases [9–11]. Another example for the importance of cerebral vascular system is that cerebral endothelial cells nourish neighboring neurons. Through releasing trophic factors, cerebral endothelial cells guide developing axons [12], protect neurons against stress [13, 14], and provide a niche for supporting neural stem/progenitor cells (NSPCs) [15]. NSPCs were shown to have direct coupling with cerebral endothelial cells [16], and in this so-called neurovascular niche, cell–cell signaling between cerebral endothelial cells and neuronal precursor cells helps mediate and sustain pockets of ongoing neurogenesis and angiogenesis in adult brain [15, 17]. Even under the remodeling phase after brain injury, these close relationships are maintained, and both neurogenesis and angiogenesis occur in the neurovascular niche to promote repairing of the brain. Indeed, angiogenic stimulation enhances neurogenesis after stroke [15, 18], and in turn, neuroblasts migrate along perivascular routes and the promotion of neurogenesis enhances vascular re-growth [19].

For the most part, research that studies mechanisms of trophic coupling in the neurovascular unit has mainly focused on endothelium–neuron and endothelium–astrocyte interactions. However, cell–cell interactions between endothelial cells and oligodendrocytes are likely to be important to maintain brain functions as well, especially in white matter. Recent papers indicate that some populations of oligodendrocyte lineage cells are located close to cerebral endothelial cells [20, 21] (Fig. 1), and these cells may communicate each other via secreting soluble factors. In this mini-review, therefore, we attempt to overview key findings for the crosstalk between cerebral endothelial cells and oligodendrocytes.

Fig. 1.

Oligodendrocyte precursor cells (red: PDGF-R-alpha antibody) are often located closely to cerebral endothelial cells (green: CD31). Adult rat brain, corpus callosum region. Scale bar = 10 μm

Endothelium-oligodendrocyte signaling

Oligodendrocyte generation from neural stem cells

The interaction between brain microvascular endothelium and NSPCs is receiving a lot attention as a key mechanism of NSPC biology [22]. During embryonic development, interaction between endothelium and stem cells occurs in both the CNS and non-nervous tissues [22]. Endothelial cells modulate the behavior of NSPCs in the stem cell niche seen in subventricular zone (SVZ) or the hippocampal subgranular zone (SGZ) [23]. These regions provide a specialized microenvironment where NSPCs persist throughout their life. The SVZ contains a subpopulation of cells with astroglial properties (so-called type B cells), which give rise to dividing intermediate progenitors or transient amplifying progenitors (so-called type C cells) to generate neuroblasts (so-called type A cells). The type B cells can also differentiate into oligodendrocyte lineage cells (i.e., migratory OPCs) through Olig2-expressing type C cells. These type B cells-derived OPCs leave the SVZ and migrate to the corpus callosum, the white matter tract in the striatum and fimbria fornix [24]. Thus far, at least two studies have examined the cell–cell interaction between cerebral endothelium and NSPCs in generating oligodendrocytes. Chintawar et al. [25] investigated the interactions of human fetal neural precursor cells (hfNPCs) with human brain endothelial cells in their in vitro model systems. They found that in the in vitro sub-endothelial niche, NPCs interacted with endothelial cells to promote NPC differentiation into astrocytes, neurons, and even oligodendrocytes. This study proposed that the chemokine CCL2/MCP-1 mediated the interaction between endothelium and neural precursor cells for the differentiation into oligodendrocytes [25]. Another study by Plane et al. also examined the mechanisms of differentiation from postnatal murine forebrain neural stem cells. Interestingly, they showed that conditioned medium from endothelial cells promoted the differentiation of neural stem cells into oligodendrocyte linage cells [26]. These two studies may suggest that cerebral endothelial cells participate in generating oligodendrocyte linage cells from neural precursor cells under some conditions.

Signaling from cerebral endothelial cells to oligodendrocytes

In the developmental stage, most oligodendrocytes are generated from their precursors (oligodendrocyte precursor cells; OPCs). OPCs migrate to colonize the white matter brain (and sometimes gray matter as well) from their sites of origins (e.g., SVZ region) [27, 28], and the differentiation into mature oligodendrocytes occurs in a multiple steps [29]. Many OPCs differentiate into myelinating oligodendrocytes during development, but some populations of OPCs remain at the immature state and are widely distributed within the brain even after adolescence [30]. Several evidences indicate that OPCs persist in the adult SVZ [24], and oligodendrocytes continue to be newly produced in the adult white matter [31–34]. These OPCs in the adult brains may participate in myelin repair after injury [35, 36]. Past studies have demonstrated that adult SVZ progenitors (i.e., type B cells) can generate new OPCs/oligodendrocytes after demyelinating lesions of the corpus callosum [37, 38], seizures [39], or stroke [40–42]. As noted, these SVZ type B cells serve as primary precursors for new oligodendrocytes and give rise to young migrating oligodendrocytes through intermediate Olig2 transit-amplifying cells (so-called type C cells) under normal and injured brains [24]. They are known to move out of the SVZ into the corpus callosum, neighboring striatum, and fimbria fornix to differentiate into non-myelinated and myelinated oligodendrocytes [43]. Compared with the large number of new neurons born in the SVZ, fewer OPCs/oligodendrocytes are normally produced from adult SVZ type B cells. This may be related to the slow turnover of oligodendrocytes in adult brain [34]. However, the number of OPCs/oligodendrocytes derived from SVZ type B cells are increased after a demyelinating lesion. Under the stress conditions, SVZ-generated oligodendrocytes express markers of myelin, suggesting that they contribute to the remyelination process [37, 38, 43, 44].

Recently, the concept of oligovascular niche was proposed to understand the phenomena of trophic coupling between cerebral endothelium and oligodendrocyte precursors [29]. Cerebral endothelial cells may promote the proliferation of OPCs through releasing trophic factors, such as BDNF and bFGF, in vitro [45]. In addition, cerebral endothelial-derived VEGF-A increased the mobility of OPCs without affecting their proliferation [46, 47]. Importantly, sub-lethal oxidative stress decreased the levels of growth factor productions in endothelial cells, and the “sick” endothelial cells seemed no longer supportive for OPCs [45]. These results indicate that endothelium-oligodendrocyte trophic coupling would be disturbed during pathological conditions, which lead to white matter dysfunction. As noted, some OPCs are located close to cerebral endothelial cells in adult white matter (e.g., corpus callosum) [20], and therefore, cerebral endothelial cells may participate in the process of oligodendrocyte remodeling and repair under both normal and pathological conditions.

Signaling from oligodendrocytes to cerebral endothelial cells

Thus far, we have discussed how cerebral endothelial cells affect functions of oligodendrocyte lineage cells, but the trophic coupling between cerebral endothelial cells and oligodendrocytes should be “two-way”. Oligodendrocyte lineage cells could act as growth factor providers for neighboring cells [48]. For example, oligodendrocytes synthesize defined growth factors and provide trophic signals to nearby neurons. Past studies show that oligodendrocyte-derived trophic factors, such as IGF1 and GDNF, promote neuronal survival and axonal outgrowth in vitro [49]. Moreover, oligodendrocytes may serve as a principal metabolic supplier of lactate, which is integral for axonal energy support through monocarboxylate transporter 1 [50]. Although there is little known about the mechanisms by which oligodendrocyte lineage cells affect the vascular system, Pham et al. [51] recently reported that after white matter injury, mature oligodendrocytes secreted a well-known angiogenic factor MMP-9 to promote vascular remodeling. Hence, similar to the neurovascular niche, the oligovascular niche may provide an important mechanism for both angiogenesis and oligodendrogenesis in the adult white matter.

However, it should be noted that oligodendrocyte lineage cells might not always be supportive for the cerebral vascular system. Very recently, OPCs were shown to release MMP-9 under acute phase of white matter injury [20]. This OPC-derived MMP-9 initiated the early onset of BBB breakdown, which would progress to white matter dysfunction at the later time point [20]. These findings suggest that the oligodendrocyte-endothelium interaction is not merely static. As the concept of neurovascular unit emphasizes, depending on the context, these cell–cell interactions can be both beneficial and detrimental. Under normal conditions, cell–cell trophic coupling is necessary for maintaining brain homeostasis. On the other hand, under the acute phase of neurodegenerative diseases, these cell–cell trophic couplings could be disrupted and lead to cellular dysfunction. However, in the chronic phase after brain injury, the cell–cell interaction may be recovered to some extent, and brains would try to repair their damaged function. Hence, understanding the precise mechanisms of oligodendrocyte-endothelium trophic coupling would lead us to find novel therapeutic approaches both in protecting white matter against stress under acute phase and in boosting white matter remodeling/repairing under chronic phase for white matter-related diseases such as stroke and vascular dementia.

Possible mediators for oligodendrocyte-endothelial interaction

As mentioned, crosstalk between the vascular and neuronal compartments in the neurovascular niche is mediated by an exchange of soluble signals [3, 4, 14, 52–56]. Many of these trophic factors may also affect oligodendrocyte lineage cells. Moreover, growing literature suggests that similar to cerebral endothelium, oligodendrocyte lineage cells could work as a “bank” for trophic factors (Table 1). The precise regulatory mechanisms that underlie angiogenesis and oligodendrocytes in the oligovascular niche still require to be elucidated. However, we will try to summarize several candidates of mediators for the crosstalk between cerebral endothelium and oligodendrocytes in this section.

Table 1.

Releasing factors from oligodendrocyte linage cells

Releasing	Factor cellular type
BDNF	Cultured rat oligodendrocytes [67]
	Immortalized oligodendroglial cell line (N19) [131]
	MBP-positive oligodendrocytes in rat brain [68]
IGF-1	Cultured rat OPCs [132]
NGF	Cultured rat OPCs [133]
	Cultured rat oligodendrocytes [67, 133]
	Immortalized oligodendroglial cell line (N19) [131]
	MBP-positive oligodendrocytes in rat brain [68]
NT-3	Cultured rat oligodendrocytes [67]
	MBP-positive oligodendrocytes in rat brain [68]
Neuregulin	cultured rat OPCs [134]
	Cultured adult human oligodendrocytes [135]
GDNF	Cultured rat oligodendrocytes [49, 136]
	Cultured rat OPCs [136]
	Transformed oligodendroglial cell line OLN-93 [136]
	Immortalized oligodendrocyte precursor cell line OLI-neu [136]
TGF-beta	Cultured rat OPCs [82]
HMGB-1	Oligodendrocyte-like cells in rat brain [137]
AM	Human oligodendroglial cell line KG1C [103]
MMP-9	Cultured rat OPCs [20]
	MOBP-positive oligodendrocytes [21]

Brain-derived neurotrophic factor

Brain-derived neurotrophic factor (BDNF) is a member of the “neurotrophin” family of growth factors, which are related to the canonical nerve growth factor (NGF). BDNF is well known to act on neurons to support the survival of existing neurons and encourage the growth and differentiation of new neurons and synapses [57, 58]. BDNF can also modulate non-neuronal cell types. Irrefutably, BDNF is one of the major mediators for endothelium-oligodendrocyte interaction.

BDNF is known to regulate oligodendrocyte function in several ways. Firstly, BDNF promotes OPC proliferation and differentiation into mature oligodendrocytes. Cell culture and in vivo animal studies suggested that Trk-B receptor and ERK pathway mediated the supportive effects of BDNF [59–62]. Secondly, even under pathological conditions, BDNF can be supportive for oligodendrocyte lineage cells. In in vivo white matter injury models (spinal cord injury or cuprizone-treated model), BDNF was shown to play an important role in regulating the numbers of oligodendrocyte lineage cells after demyelination [63, 64]. Thirdly, BDNF may also work on neural stem cells (NSCs) to trigger the differentiation into oligodendrocyte lineage cells. Chen et al. [65] reported that BDNF enhanced the cell commitment of NSCs to neuronal and oligodendrocytic fates by activating Wnt/β-catenin signaling pathway in vitro. Since cerebral endothelial cells are the major BDNF-producing cell type, cerebral endothelial cells may support oligodendrocyte lineage cells via BDNF signaling. As introduced in the previous section, endothelial-derived BDNF was shown to promote the proliferation of OPCs in vitro [45]. But importantly, stressed endothelial cells secrete less BDNF and can no longer support OPCs [45], indicating that endothelial-oligodendrocyte interaction is highly dependent on their cellular conditions.

BDNF also promotes endothelial cell survival and induces angiogenesis in brain [66]. While cell–cell trophic interactions are generally considered as “two-way”, there have been still no direct proofs that BDNF mediates signals from oligodendrocytes to endothelial cells. Nevertheless, several lines of evidence strongly support the idea that oligodendrocyte lineage cells would produce BDNF to support endothelial cells. In situ hybridization and immunocytochemical studies identified expressions of BDNF mRNA/protein in cultured basal forebrain oligodendrocytes [67, 68]. The physiological relevance of these culture works is supported by detection of BDNF in oligodendrocytes in vivo. For example, BDNF mRNAs are localized in subpopulations of myelin-basic-protein-positive mature oligodendrocytes in the basal forebrain, cingulate cortex, and corpus callosum of postnatal day 7 rats [68]. Similarly, BDNF mRNA and protein are also expressed in subsets of APC-positive oligodendrocytes of adult spinal cords [69, 70]. The oligodendrocytic BDNF is bioactive because conditioned medium from oligodendrocyte cultures increased cholinergic neurons and the effect was partially blocked by co-treatment with anti-BDNF neutralizing antibody [68]. Considering the fact that oligodendrocyte linage cells are often located to cerebral endothelial cells, it would be reasonable to think that oligodendrocytic BDNF may regulate cerebral vascular system to some extent.

Fibroblast growth factor-2

Similar to BDNF, fibroblast growth factor-2 (FGF-2/bFGF) can be proposed as an important mediator for the endothelium-oligodendrocyte interaction. FGF-2 is a potent stimulator of endothelial cell migration, proliferation, sprouting, and tube formation. In addition, past studies have substantially revealed the roles of FGF-2 on oligodendrocyte function. FGF-2 by itself stimulates proliferation of late-stage OPCs and blocks their terminal differentiation into mature oligodendrocytes in vitro. Upon removal of FGF-2 from the cell culture medium of OPCs, the cells readily enter the terminal differentiation [71–73]. In addition, FGF-2 cooperates with PDGF to up-regulate the expression of PDGF-receptor-alpha for OPC proliferation [71, 74]. As for OPC migration, the importance of FGF signaling in early stage OPCs has been evaluated both in vivo and in vitro. Using an in vivo transplantation approach, Osterhout et al. [75] demonstrated that OPCs with dominant-negative FGF receptor 1 failed to migrate. Subsequent in vitro study showed that in response to FGF-2 stimulation, OPCs growing in an agarose drop successfully move out from the drop [76]. Another study suggested that FGF-2 also enhanced the migration of pre-OPCs from oligospheres, but might not induce the migration of late-stage OPCs [77]. While still controversial, FGF-2 may affect OPC survival. FGF-2 can prevent OPCs from apoptotic stress [78], but this kind of protective effects were not observed in optic nerve OPCs [79]. Since the cellular localization of FGF-2 is observed in cerebral endothelium both in normal and pathological conditions [80], future studies are warranted to examine how FGF-2 modulate the dynamics of endothelium-oligodendrocyte crosstalk under normal conditions as well as acute/chronic phases after brain injury.

Transforming growth factor-beta

Transforming growth factor-beta (TGF-beta), a prototypic member of a large family of pleiotropic cytokines, is also a potent modulator in the oligovascular niche. Many cells have been reported to produce TGF-beta, and TGF-beta would modulate several cellular functions in most cells [81]. In terms of endothelium-oligodendrocyte interaction, oligodendrocytes are reported to express TGF-beta in vitro cell culture systems [82] and in vivo spinal cords [83]. TGF-beta contributes to angiogenesis by stabilizing newly formed capillary sprouts [84]. Thus far, many studies in mouse and human have demonstrated its pivotal roles in modulating angiogenesis after brain injury, such as stroke [85–90]. In addition, dysregulation of TGF-beta signaling may cause hereditary vascular disorders [81]. For example, mutations in TGF-beta receptors lead to hereditary hemorrhagic telangiectasia [91], and in the cerebral white matter, accumulation of TGF-beta1 due to HtrA1 mutation is associated with a hereditary disorder CRASIL (cerebral autosomal recessive arteriopathy with subcortical infarcts and leukoencephalopathy) [92]. In contrast, the roles of TGF-beta on oligodendrocytic functions are still understudied, but TGF-beta was reported to affect the migration of cultured OPCs when they were maintained on astrocytes [93]. Using the oligodendrocyte precursor cell line OLI-neu, TGF-beta was shown to upregulate a chondroitin sulfate proeoglycan DSD-1-PG on the cell surface [94], indicating that TGF-beta may modulate the cell adhesion property of OPCs to change cell motility.

Adrenomedullin

Adrenomedullin (AM) was originally isolated from pheochromocytoma cells [95]. AM has a variety of actions on the vascular systems, such as endothelial survival/proliferation, vasodilatation, regulation of BBB permeability, and modulation of oxidative stress levels in endothelium [96, 97]. AM is secreted from various organs, and in the CNS, AM is mainly expressed in neurons and cerebral endothelium [95, 98]. AM has been shown to reduce the infarct volumes in transient stroke models [99, 100]. Under stroke conditions, AM expression is markedly increased via the hypoxia-inducible factor-1 signaling [101], indicating that AM could work for vascular remodeling/repairing after brain injury. Interestingly, in the mouse model of prolonged cerebral hypoperfusion (i.e. vascular dementia model), AM was demonstrated to be protective toward white matters [102], where most oligodendrocytes are populated. In addition, Uezono et al. [103] detected AM mRNAs in human oligodendroglial cell line, and showed the potential effects of AM in modulating oligodendrocyte function through AM receptors in oligodendrocytes. Importantly, past studies have demonstrated that AM increases several growth factors such as VEGF and FGF [104–107]. Hence, AM might work as a “master” modulator for growth factors in the oligovascular niche.

Vascular endothelial growth factor

Vascular endothelial growth factor (VEGF) is a primary regulator of angiogenesis by stimulating endothelial cell proliferation, migration, and tube formation [54], but it is now well recognized that VEGF is not solely an endothelial mediator, rather may represent one of the major mediators in the neurovascular unit [108–110]. For example, VEGF signaling plays a key role in neuronal migration and CNS development [111]. In addition, VEGF may also affect oligodendrocyte lineage cells. OPCs were positive for the VEGF receptor, Flk-1 (also known as KDR or VEGF-receptor 2) [47], which is primarily responsible for VEGF-induced angiogenesis. Recent in vitro study suggested that VEGF-A significantly accelerated the motility of OPCs through Flk-1, and this effect was partly mediated by ROS production [47]. Importantly, conditioned media from cerebral endothelial cells promoted both OPC proliferation and migration. However, endothelial-derived VEGF-A may participate in the OPC migration but not proliferation [46]. In addition to VEGF-A, other VEGF families and VEGF receptors may also be involved in oligodendrocyte functions. Le Bras et al. [112] demonstrated that VEGF-C promoted OPC proliferation through VEGF-receptor-3. Taken together, these findings indicate that VEGFs/VEGF-receptors may play a central role in the endothelial-oligodendrocyte trophic coupling.

If VEGF is an important modulator for the oligovascular niche, we may need to consider the “biphasic actions” of VEGF to comprehend the dynamics of trophic coupling between cerebral endothelium and oligodendrocytes. Since VEGF is a primary regulator of angiogenesis, VEGF can trigger remodeling responses in endothelial cells (i.e., accelerating angiogenesis) after brain damage, such as stroke. Infusing VEGF into the lateral ventricles stimulated angiogenesis and decreased infarct volumes in rodent models of focal cerebral ischemia [113]. An increase in angiogenesis by VEGF in rats was also associated with reduced neurological deficits after stroke [114]. Similar effects were also reported in neonatal focal rodent stroke [115]. Moreover, in transgenic mice with overexpressing human VEGF165, brain microvessel density was significantly elevated compared to wild-type mice before ischemia, and the microvessel density was higher 3 days after stroke onset [116]. On the other hand, VEGF increases BBB permeability in the acute phase in stroke. VEGF administration worsens BBB leakage by ischemic insults [117, 118]. Therefore, depending on the context of the microenvironment in the oligovascular nice, VEGF may regulate the endothelium-oligodendrocyte interaction to maintain and remodel white matter function.

Matrix metalloproteinases

Matrix metalloproteinases (MMPs) comprise a family of zinc endopeptidases, and their activities are tightly regulated by tissue inhibitor of metalloproteinases (TIMPs). This larger MMP network plays major roles in the physiology and pathology of the mammalian CNS, including stroke (MMP-2/3/7/9/13, TIMP-1/2 [119]), vascular dementia (MMP-2, 3, 9 [120, 121]), and CADASIL (cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy: TIMP-3 [122]). Similar to VEGF, MMPs are well-known biphasic modulators in the CNS. During the acute phase of stroke, MMP-2/3/9/13 are deleterious, i.e., they degrade the extracellular matrix that comprises the basal lamina, which could directly damage BBB [123]. In addition, proteolysis of the neurovascular matrix by MMP-9 after stroke can trigger anoikis-like neuronal death [124]. However, during delayed phases of stroke recovery, some of these proteinases (MMP-2/3/9) may play beneficial roles [119, 123, 125]. In a mouse stroke model, peri-infarct cortical areas demonstrate a secondary elevation in MMP-9 in endothelial and glial cells within networks of regrowing microvessels [126]. Inhibition of MMPs during this delayed phase disrupted brain repairing with the induction of hemorrhagic and malformed blood vessels. Moreover, signals of MMP-9 were observed in the migrating neuroblasts from the subventricular zone after brain injury, and inhibition of these MMPs also blocked the movement of these neuroblasts originally headed towards damaged brain [127].

As MMPs play multiple roles in the neurovascular unit, MMPs may mediate the crosstalk between cerebral endothelium and oligodendrocytes. Thus far, several studies imply the essential roles of MMP network in oligodendrocytic functions. For example, MMP-9 mediated oligodendrocytic process extension in vitro cell culture system [113] and MMP-9 expression level was increased during myelin formation in the optic nerve in vivo [114]. Moreover, after white matter injury, MMP-9 removed injury-induced deposition of inhibitory NG2 proteoglycan, which is an essential step for OPCs to differentiate into mature oligodendrocytes for remyelination [128]. In addition, MMP-12 was shown to cause demyelination, macrophage infiltration, and motor deficits in a mouse model of virus-induced multiple sclerosis [129]. TIMPs might be also involved in cellular function/survival of oligodendrocyte linage cells. In a mouse model of focal stroke ischemia, TIMP-3 deficiency showed more number in immature oligodendrocytes after injury [130]. Besides roles as a recipient cell for MMPs, oligodendrocyte lineage cells may in turn produce MMPs to send signals to cerebral endothelial cells. As noted, OPCs respond quickly to a stress stimulation after white matter damage and release MMP-9 to induce early BBB leakage in a mouse model of prolonged hypoperfusion model [20]. On the other hand, under the chronic phase of white matter injury, oligodendrocytic MMP-9 may promote vascular remodeling [51]. Hence, as in the neurovascular unit, MMPs contribute to the dynamics of cellular interactions in the oligovascular niche due to their “biphasic property”.

Conclusions

The concept of neurovascular unit emphasizes that cell–cell interaction is critical to maintain normal brain function as well as brain remodeling after injury. Within the conceptual framework of neurovascular unit, cerebral endothelial cells are particularly important in releasing soluble factors to nourish neighboring cells, such as astrocytes or neurons. As we discussed in this mini-review, cerebral endothelial cells may also support oligodendrocyte lineage cells in the white matter. On the other hand, oligodendrocyte-derived factors can modulate cerebral vascular systems under some conditions. In the so-called oligovascular niche (i.e., microenvironment between cerebral endothelium and oligodendrocytes), cerebral endothelium and oligodendrocyte lineage cells may work together for white matter homeostasis (Fig. 2). Although some key mediators for endothelium-oligodendrocyte crosstalk were briefly discussed in this mini-review, the precise underlying mechanisms still remain to be elucidated. Importantly, there may be overlap between factors in the oligovascular niche and the relatively more well-established neurovascular niche. Therefore, a deeper analysis, perhaps using subtractive approaches, may be required in order to rigorously define the regulatory signals that are truly unique to the oligovascular niche. As white matter injury is a key part of most CNS diseases, understanding the cell–cell interaction between cerebral endothelium and oligodendrocytes may lead us effective therapeutic approaches for white matter related diseases, such as stroke and vascular dementia.

Fig. 2.

Proposed model for endothelium-oligodendrocyte crosstalk. In the so-called oligovascular niche (a micro-environment between cerebral endothelial cells and oligodendrocyte lineage cells), these cells exchange soluble factors to maintain white matter homeostasis. However, after brain injury, the trophic coupling would be disrupted, which lead to white matter dysfunction. In addition, some factors such as VEGF and MMP-9 may actively worsen pathological processes (e.g., BBB breakdown), but may promote brain remodeling in a chronic phase after injury