Regulation of brain endothelial barrier function by microRNAs in health and neuroinflammation

Brain endothelial cells constitute the major cellular element of the highly specialized blood‐brain barrier (BBB) and thereby contribute to CNS homeostasis by restricting entry of circulating leukocytes and blood‐borne molecules into the CNS. Therefore, compromised function of brain endothelial cells has serious consequences for BBB integrity. This has been associated with early events in the pathogenesis of several disorders that affect the CNS, such as multiple sclerosis, HIV‐associated neurologic disorder, and stroke. Recent studies demonstrate that brain endothelial microRNAs play critical roles in the regulation of BBB function under normal and neuroinflammatory conditions. This review will focus on emerging evidence that indicates that brain endothelial microRNAs regulate barrier function and orchestrate various phases of the neuroinflammatory response, including endothelial activation in response to cytokines as well as restoration of inflamed endothelium into a quiescent state. In particular, we discuss novel microRNA regulatory mechanisms and their contribution to cellular interactions at the neurovascular unit that influence the overall function of the BBB in health and during neuroinflammation.—Lopez‐Ramirez, M. A., Reijerkerk, A., de Vries, H. E., Romero, I. A. Regulation of brain endothelial barrier function by microRNAs in health and neuroinflammation. FASEB J. 30, 2662‐2672 (2016). www.fasebj.org


INTRODUCTION
Brain endothelial cells (BECs) constitute the major cellular element of the highly specialized blood-brain barrier (BBB) that is essential for CNS homeostasis. The cerebral microvasculature formed by BECs constitutes an elaborate network of vessels that is composed of arterioles (10-100 mmin diameter), capillaries (4-10 mm in diameter), and venules (10-100 mm in diameter), which allows transport of nutrients and gases and removal of waste products throughout the brain and spinal cord (1). Of note, BECs play an important role in maintaining blood flow and limiting entry of circulating leukocytes and blood-borne molecules into the CNS. In mammals, BECs are unique and differ from endothelial cells that are present in peripheral tissues in a number of ways, which ensures specific brain endothelial barrier properties. First, BECs have few pinocytotic vesicles and lack fenestrations that are typical of peripheral tissue capillaries. Second, BECs have a high number of mitochondria, which suggests elevated metabolic activity. Third, BECs form a metabolic barrier by containing several catalytic membrane-bound and cytosolic enzymes that are capable of regulating metabolism of endogenous and xenobiotic molecules (2,3). Fourth, BECs allow immune surveillance of the CNS, thereby playing a fundamental role in host defense with minimal inflammation (4). Fifth, BECs are the gatekeeper of the CNS as a result of an elaborate system of transport proteins and ion channels that are present in the plasma membrane and are asymmetrically distributed on the luminal (blood-facing) and abluminal (brain-and spinal cord-facing) sides, which allows bidirectional transport of selected substances (1,2,5). Of importance, BECs are interconnected by complex tight junctions (TJs) between lateral plasma membranes, which results in a polarized insertion of lipids and proteins (fence function). In turn, this feature restricts paracellular diffusion of nonionic hydrophilic molecules and ionic molecules, and, consequently, confers BECs a high electrical resistance (;3000 V·cm 2 in vivo; barrier function) (6,7).
The BBB is regulated and induced by the CNS-specific microenvironment, which leads to the emerging concept that the complex cellular structure known as the neurovascular unit is what determines its specialized function. The neurovascular unit includes dynamic communication and feedback networks between various cell types, such as BECs, pericytes, astrocytes, perivascular microglia, smooth muscle cells, and neurons, as well as the extracellular matrix, so that it may respond to physiological changes and function as a biological barrier (8,9). It is noteworthy that gene expression programs that underline the unique brain endothelial phenotype and function are highly regulated and under the control of the CNS microenvironment under physiological conditions (10). Indeed, the complexity of the cerebrovascular network requires coordinated genetic programs that, in part, are controlled by transcriptional activity [for a review, see Minami and Aird (11)] and, in part, by chromatin remodeling [for a review, see Matouk and Marsden (12)]. Recent studies have proposed posttranscriptional gene regulation, mediated via small noncoding RNAs termed microRNAs, as a possible mechanism that integrates the CNS milieu and BEC phenotype and function. In this context, microRNAs would contribute to the fine tuning of the complex regulatory gene networks of BECs under physiological and pathological conditions (13,14). Moreover, it has become evident that many aspects of BEC biology, such as angiogenesis, neuroinflammation, and barrier function, are regulated by microRNAs. Recent research has been carried out in the context of stroke (15), and this is extensively reviewed in Yin et al. (16).Itisalso worth mentioning that endothelial microRNAs regulate the permeability of the blood-tumor barrier (17,18) and destabilize TJs in CNS tumor metastasis (19). In this review, we discuss brain endothelial microRNA regulatory mechanisms and their contribution to BBB function under both physiological and pathological states, in particular, those related to neuroinflammation.

BRAIN ENDOTHELIAL microRNA FAMILIES AND THEIR ROLES IN NEUROINFLAMMATION
To better understand the physiology of microRNAs, bioinformaticians have grouped microRNAs into broadly conserved families on the basis of sequence and structure similarity of pre-microRNAs; therefore, microRNA families can exhibit full or partial conservation among animal species on the 2-8 mature microRNA sequence (20,21). Experimental studies demonstrate that microRNA family members execute conserved biological functions across species (22).
Our recent studies of microRNA expression in the brain endothelium, using conditions known to negatively modulate BBB function (13,23,24), revealed profound changes in the brain endothelial microRNA profile ( Table 1) (13). Of these changes, several microRNA families with roles in cardiovascular biology seemed to be critically affected by cytokine treatment ( Table 2). For instance, microRNA-30, microRNA-125, and let-7 families were remarkably negatively regulated by proinflammatory cytokines (Table 2). It is noteworthy that decreased expression of microRNA-30 family members has been associated with loss of cell adhesion, changes in cell morphology, and increased cell migration in epithelial cells by modulation of snai-1 (25), a transcription factor that silences genes that are important for TJ and adherens junction formation and stability by interacting with E-box sequences (26)(27)(28). In endothelial cells, members of the microRNA-30 family regulate vascular development and sprouting in angiogenesis by targeting the notch ligand, delta-like ligand 4 (29). Thus, members of the microRNA-30 family represent potential targets, with roles in modulation of cell-to-cell contacts, and, as a consequence, possibly barrier function in the brain endothelium; however, the role of microRNA-30 family in the regulation of BBB function in vivo remains to be elucidated.
Members of microRNA-125 family have been shown to play common roles that are associated with induced cell proliferation and decreased apoptosis by targeting proapoptotic genes; however, other biological effects by members of microRNA-125 family have also been reported (30). In the context of brain endothelium, we recently reported that microRNA-125a-5p supports immune quiescence and endothelial barrier function (13). Our study demonstrated that during neuroinflammation, expression of brain endothelial microRNA-125a-5p is suppressed, resulting in increased monocyte migration as a result of endothelial up-regulation of intercellular cell adhesion molecule 1 (ICAM-1) (13). In addition, members of microRNA-125 family can modulate angiogenesis by targeting transcription factors. Repression of microRNA-125a/b promotes angiogenesis, whereas overexpression of microRNA-125a/b impairs angiogenesis by targeting related transcriptional enhancer factor-1 and VEcadherin (31,32) in endothelial cells. MicroRNA-125b has also been shown to regulate glioma-associated angiogenesis. Mechanistically, brain endothelial microRNA-125b is suppressed by the glioblastoma microenvironment, and VEGF stimuli resulted in increased expression levels of its target myc-associated zinc finger protein, which is a transcription factor that regulates the expression of VEGF (33).
The let-7 family is another microRNA family that has been shown to support brain endothelial barrier function and quiescent state (34). The proinflammatory cytokine, TNF-a, decreased levels of let-7 family members microRNA-98 and let-7g-3p (let-7g*) in brain endothelium, which in turn increases monocyte cell adhesion and migration in vivo (34).
In summary, proinflammatory cytokines affect several families of brain endothelial microRNAs that have important roles in BBB function and in angiogenesis; however, it remains to be elucidated whether these families of microRNAs cooperate during neuroinflammation and whether they form a link between neuroinflammation and angiogenesis in diseases that affect the CNS.

BRAIN ENDOTHELIAL microRNA CLUSTERS AND THEIR REGULATION BY PROINFLAMMATORY CYTOKINES
More than half of all currently known microRNA genes are linked as clusters on chromosomes and are often transcribed as a single polycistronic transcript, possibly to provide the opportunity for different microRNAs to simultaneously target different genes in specific cellular pathways (35,36). Moreover, clustered microRNAs may have similar seed sequences, thereby sharing common hCMEC/D3 cells were stimulated with TNF-a + IFN-g (10 ng/ml concentration of each cytokine) for 24 h. Total RNA was isolated from 3 independent biological replicates, and microarray analysis (Agilent-021827 human microRNA microarray G4470C; Agilent Technologies, Santa Clara, CA, USA) was performed to determined microRNA levels (13). miR, microRNA.
gene targets (37). Our previous study shows that proinflammatory cytokines alter the levels of 2 important microRNA clusters in BECs, microRNA-17 and microRNA-221 clusters, both of which have been associated with several biological processes in endothelium ( Table 2) (13). Certainly, one of the most studied clusters is the microRNA-17 cluster, which is formed by 6 members situated on chromosome 13 (37)(38)(39)(40). Several members of the microRNA-17 cluster may have related functions and act as proangiogenic signals by silencing antiangiogenic genes, such as thrombospodin-1 and connective tissue growth factor, thereby leading to enhanced neovascularization (41). In addition, microRNA-17 cluster has also been shown to regulate cell sprouting and proliferation by inhibiting tissue inhibitor of metalloproteinase-1 (40,42,43); however, using ischemia animal models, it has been demonstrated that endothelial-specific conditional and forced overexpression of microRNA-19 and microRNA-92a, both members of the microRNA-17 cluster, negatively regulates ischemiainduced arteriogenesis (44,45). These studies suggest that the microRNA-17 cluster might act either as a positive or negative regulator of angiogenesis depending on the cell environment.
Finally, microRNA-221 cluster and family are highly conserved in vertebrates (37). The microRNA-221 cluster regulates endothelial cell proliferation, migration, and capillary tube formation by targeting c-Kit (46). In addition, microRNA-221 cluster has been shown to indirectly modulate eNOS expression levels (47,48), and might be implicated in endothelial cell fate commitment and cytoskeleton remodeling by targeting the Cip/Kip family of proteins (39). It should be stated that microRNA-125 and microRNA-221 might have complementary actions. Depletion of the microRNA-221 cluster indirectly reduced microRNA-125 expression in HUVECs (49). Moreover, ICAM1 is targeted by the microRNA-221 cluster, and decreased levels of this microRNA cluster result in up-regulation of ICAM1 protein expression in epithelial cells (50,51).
Although proinflammatory cytokines induce alterations in brain endothelial microRNA cluster expression (13), whether this is mediated at the transcriptional level or by modulation of microRNA processing remains unclear.

BRAIN ENDOTHELIAL microRNAs REGULATE DIFFERENT PHASES OF THE NEUROINFLAMMATORY RESPONSE
The cerebral microvasculature plays a major role in the regulation of several phases of the neuroinflammatory response, including vascular permeability, leukocyte adhesion and migration, and restoration of activated endothelium back to a quiescent state. Indeed, the process of neuroinflammation requires a coordinated activation of immune cells, brain resident cells, and CNS vasculature to avoid collateral damage and unnecessary chronic inflammation. Depending on the strength and duration of inflammatory stimuli, brain endothelial microRNAs may either change the genetic program toward an activated inflamed vascular endothelium or regulate genes to counteract cytokine signaling to return to a quiescent state. Details of brain endothelial microRNAs that have been shown to regulate inflammation as well as their possible implication in regulating phases of the neuroinflammatory response have been summarized in the next and following sections.
Families and clusters were considered deregulated when levels of microRNA members changed $2 (italics) in the presence of TNF-a + IFN-g (13). [Analysis was performed by using miRbase (http://www.mirbase.org) and miRfocus (http://pepcyber.org/mirfocus/ ) databases.] Some microRNAs with high sequence similarity (underlined) could not be distinguished by the array analysis and were considered as one. miR, microRNA; NA, not applicable; NE, not expressed.

BRAIN ENDOTHELIAL microRNAs THAT PROMOTE CELL ACTIVATION IN NEUROINFLAMMATION
MicroRNA-155 promotes inflammation and activates both immune cells and brain endothelium by silencing hundreds of transcripts with modest effects, a characteristic of fine tuning regulation (14,52,53). MicroRNA-155 originates from an exon in a noncoding transcript, B-cell integration cluster, which is located on chromosome 21q21 (54). MicroRNA-155 is highly conserved in the animal kingdom, including mammals, chicken, lizards, zebrafish, and frogs (21,55). Inflammatory mediators, such as bacterial LPS, polyriboinosinic-polyrobocytidylic acid, and TNF-a, induce expression of microRNA-155 in several cell types, including the brain endothelium (14,(56)(57)(58). Inhibitors of JNK, PKC, and RhoA prevent microRNA-155 up-regulation in macrophages, B cells, T cells, epithelial cells, and BECs, and prevent their ability to respond to immune challenge (59-61) (unpublished observation). In the brain endothelium, microRNA-155 seems to be an important control node for expression and organization of adhesion molecules as well as for inducing a gene expression program characteristic of inflammation ( Table 3) (14). For instance, both treatment with TNF-a and IFN-g and overexpression of microRNA-155 in BECs induce expression of genes associated with immune response, major histocompatibility complex, cell adhesion molecules, and the complement system, pathways that are known to contribute to the inflammatory response (14,62,63). It is possible that microRNA-155 up-regulation in both immune cells and brain endothelium during neuroinflammation contributes to the harmonization of cell activation as a mechanism of defense of the CNS against pathogens; however, sustained and prolonged increases in microRNA-155 levels in the brain endothelium may provoke brain endothelial barrier breakdown by altering components of the cell-cell and cell-extracellular matrix adhesion pathways, both important cellular processes that stabilize barrier function (14). Indeed, we demonstrated that brain endothelial microRNA-155 is upregulated by proinflammatory cytokines in a time-and dose-dependent manner (14). This response coincides with cytokine-mediated increases in brain endothelial barrier permeability that is associated with redistribution of ZO1 at cell-cell contacts and with focal adhesion reorganization. Functional analysis of microRNA-155 in the brain endothelium shows that this microRNA dampened the expression of 2 interendothelial junctional complex molecules, annexin-2 and claudin-1, and 2 focal adhesion components, dedicator of cytokinesis-1 and syntenin-1, which suggests an important mechanism by which microRNA-155 exacerbates BBB breakdown (14).
Nevertheless, protective roles for microRNA-155 have also been suggested to occur in different cell systems. Two independent research groups have proposed that, in resting conditions, microRNA-155 targets caspase-3, Fasassociated death domain-containing protein, IKKe,a n d the receptor-interacting serine-threonine kinase-1 to maintain homeostasis and low levels of cell death (64,65). These results suggest that microRNA-155 might not only act as a positive regulator of inflammation but can negatively regulate activation of inflammatory pathways as well, which may be dependent on the cellular context and/or strength and duration of the stimuli. In the context of endothelial cells, microRNA-155 may indeed modulate a number of biological processes. For instance, microRNA-155 has been shown to be involved in regulation of blood pressure and endothelial-dependent vasorelaxation by modulating expression of angiotensin II type 1 receptor and eNOS expression (59), respectively (66-68). Other groups recently described effects of microRNA-155 on endothelial cells, including modulation of the microcirculatory vascular tone by targeting endothelin-1 (69) and actomyosin contractility by targeting RhoA and myosin light-chain kinase (70).
In prostate carcinoma cells, microRNA-146 regulates the expression of Rho kinase 1 (78), a key mediator in endothelial stress fiber formation and contractility that precedes cell-cell junction delocalization and increased barrier leakage during inflammation (79). However, brain endothelial microRNA-146 does not affect Rho kinase 1 protein levels during neuroinflammation (74), which suggests that microRNA-146 confers accuracy in response in a cell type-specific manner. Moreover, consistent with a role for endothelial microRNA-146 as a negative regulator of NF-kB signaling by repressing multiple targets, overexpression of microRNA-146 in endothelial cells negatively regulates the thrombin-induced increase in leukocyte adhesion to the endothelium by targeting caspase recruitment domain family 10 (80, 81), a key scaffold protein of GPCR-mediated NF-kB signaling (80). In addition, a recent study demonstrated that endothelial microRNA-146 participates in the resolution of cytokine-induced activated endothelium. Indeed, Cheng et al. (75) demonstrated that whereas microRNA-146a and microRNA-146b transcript (pri-microRNA-146a/b) expression occurs early after treatment with proinflammatory cytokine IL-1b, the mature forms, microRNA-146a/b, did not increase until after long-term treatment with IL-1b, which coincides with the resolution of inflamed endothelium into a quiescent state. Mechanistically, endothelial microRNA-146 post-transcriptionally regulates RNA binding protein HuR, a protein that is involved in endothelial activation and leukocyte recruitment in response to the proinflammatory cytokine IL-1b (75).

MicroRNA-126/microRNA-126-5p
MicroRNA-126 and microRNA126-5p are highly enriched in endothelial cells and play key roles in endothelial integrity and function (82,83). Cytogenic location of microRNA-126 and microRNA-126-5p is on chromosome 9q34.3, and both originate from an intron of the gene that codes for epidermal growth factor-like-7 (Egfl7) (82)(83)(84). Expression of microRNA-126 has also been detected in the hematopoietic system, including bone marrow stem cells, bone marrow samples from patients with acute promyelocytic leukemias (85), and CD4 + cells of patients with multiple sclerosis (MS) (86). MicroRNA-126 has been more extensively studied than microRNA-126-5p and has been implicated in several important aspects of vascular biology, including cell migration, cytoskeletal organization, capillary formation, and vascular inflammation (43,87). Functional roles of microRNA-126 were first reported by using knockout animal models that showed their essential role in blood vessel formation and integrity (82,83).Indeed,before these studies, it was observed that Egfl7-knockout phenotype was associated with failure in angiogenesis and blood vessel sprouting and formation, and only 50% of embryos survived in utero (88); however, selective floxed Egfl7 D (removal of exon 5-7 without disruption of microRNA-126) and microRNA-126 D (removal of a segment of intron 7 without disruption of Egfl7) alleles demonstrated that Egfl7 D/D mice were phenotypically normal, whereas microRNA-126 D/D mice presented abnormalities in the vasculature (83,84) similar to those previously reported after complete removal of Egfl7 (88). Phenotypic abnormalities in microRNA-126-knockout mice include lethality in ;5 0 %o fe m b r y o sb e c a u s eo fl o s so fv a s c u l a ri n t e g r i t y (remarkably in the brain) accompanied by edema and hemorrhage (83,84). Surviving mutant neonates showed defects in endothelial sprouting, proliferation, radial migration, and angiogenesis, leading to defective retinae vascularization and cardiac neovascularization after myocardial infarction (83,84). It has been suggested that the partial lethality of the microRNA-126 mutant mice might be a result of modulation of the gene expression program in endothelial vascularization, rather than a complete shut-down of the angiogenic process (83). Indeed, microRNA-126 has been shown to regulate responses of endothelial cells to VEGF and basic fibroblast growth factor via MAPK and PI3K activity (82)(83)(84). The proangiogenic actions of microRNA-126, in part, may be a result of reduced expression of negative regulators of angiogenesis, including sprouty-related EVH1 domaincontaining protein 1 that binds and inhibits RAF-1 involved in the Ras/RAF-1/MAP/ERK signaling pathway, and by silencing PIK3R2/p85-b, a regulatory subunit of PI3K that results in reduced AKT/PKB activity (83,84).
MicroRNA-126 is also involved in vascular inflammation and is highly regulated by proinflammatory cytokines (Table 1). Indeed, microRNA-126 is an inhibitor of inflammation by negative modulation of VCAM-1 expression in the resting endothelium (89). Moreover, BEC activation by cytokines decreases levels of microRNA-126, which is associated with loss in VCAM-1 posttranscriptional control and promotes leukocyte adherence to the luminal side of brain endothelium (89) (unpublished observation). Furthermore, Ets-1 and Ets-2, important regulators of immune response and of angiogenesis, interact with Ets binding elements upstream of Egfl7/ microRNA-126 (90) and mediate microRNA-126 expression. In addition, Ets-1 can be induced by TNF-a in the endothelium and promotes transcription of genes that are involved in vascular inflammation, including CCL2, VCAM-1, and matrix metallopeptidase 9 (87). Harris et al. (89,90) have proposed that microRNA-126 might act as a negative feedback loop in TNF-a signal transduction. Furthermore, the net effect of Ets-1 on vascular inflammation might, in part, depend on the balance between Ets-1-induced proinflammatory factors, such as CCL2 and VCAM-1, and Ets-1-induced anti-inflammatory factors, such as microRNA-126.

OTHER microRNAs WITH ROLES IN CEREBRAL VASCULATURE HOMEOSTASIS AND PATHOPHYSIOLOGY
High-throughput studies indicate that the brain endothelial microRNA milieu facilitates the capacity of cells to respond to stress and healthy environmental factors. Indeed, lupus serum and complement protein C5a, which are known to negatively regulate brain endothelial integrity (91), as well as caloric restriction, which is known to positively regulate brain endothelial integrity (92), induce changes in the levels of brain endothelial microRNAs (91,92). This suggests that brain endothelial microRNA expression not only orchestrates accurate tuning of gene expression that contributes to cell homeostasis and BBB integrity, but also might enforce new gene expression patterns that can influence the pathophysiology of disorders that affect CNS vasculature. In this context, when BECs are treated with homocysteine, a factor known to disrupt the BBB in vitro (93) and to play a role in brain damage (94), expression of microRNA-29 family members is significantly increased (93). Mechanistically, the authors suggested that microRNA-29b suppresses DNA (cytosine-5)-methyltransferase 3b,whichleadstoincreased levels of matrix metallopeptidase 9, a known factor that disrupts BBB integrity (93). In addition, a recent study demonstrated that HIV-1 Tat protein up-regulates expression of brain endothelial microRNA-101, leading to suppression of VE-cadherin protein expression and to increased brain endothelial barrier permeability (95). microRNA-150 is another microRNA with roles in brain endothelial barrier function. Enhanced BBB breakdown is observed by overexpression of microRNA-150 during permanent middle cerebral artery occlusion, an animal stroke model (15). It was reported that microRNA-150 induces enhanced BBB breakdown by directly targeting brain endothelial tyrosine-protein kinase receptor TIE-2, a factor involved in maintaining vascular homeostasis and barrier function (15). Moreover, intracerebroventricular injection of microRNA-150 inhibitor significantly ameliorates BBB disruption in middle cerebral artery occlusion (15). Of note, brain endothelial microRNAs might also contribute to molecular processes that contribute to brain vascular normalization after brain injury. Ge et al. (96,97) suggested that increased levels of microRNA-21 could exert protective roles during BBB damage as a result of a traumatic brain injury animal model by modulating several pathways concurrently. During traumatic brain injury, increased levels of microRNA-21 promoted both VEGF and angiopoietin-1/TIE-2 expression simultaneously, factors known to increase angiogenesis and cerebrovascular integrity (96,97); therefore, microRNA-21 could be used as a therapeutic strategy to promote neovascularization with barrier properties during brain injury.

THERAPEUTIC POTENTIAL OF BRAIN ENDOTHELIAL MICRORNAS FOR CNS DISEASES
MS is a condition with a central neuroinflammatory component in which brain endothelial barrier function is compromised. Furthermore, MS is a chronic autoimmune CNS disease that is characterized by demyelination, axonal degeneration, and, ultimately, brain and spinal cord atrophy. Studies using gadolinium imaging and MRI analysis show that BBB disruption occurs in localized brain areas and is an early event that precedes development of MS lesions and disease progression (98,99); however, the critical molecular events in the initiation of cerebrovascular endothelial dysfunction during MS are largely unknown.
It is noteworthy that recent work demonstrates that microRNA-155 is focally increased in the active inflammatory MS plaques at the neurovascular unit (14,56) and activated infiltrative leukocytes (100-102) and might contribute to the pathophysiology of MS. Our pioneering work showsthatmicroRNA-155ishighlyexpressedinconfined areas with MS lesions in the cerebrovascular endothelium.
Our study further suggests that increased levels of brain endothelial microRNA-155 contribute to early BBB impartment observed during neuroinflammation. Indeed, microRNA-155 levels are rapidly up-regulated in experimental autoimmune encephalomyelitis (EAE), an animal model for MS, during the relapsing-remitting paralysis course, which are clinical stages with compromised BBB integrity (14). Moreover, mice that are deficient in microRNA-155 show partial resistance to the development of relapsing-remitting EAE and its associated increases in BBB permeability, observations that are consistent with the effect of microRNA-155 as a microRNA that promotes inflammation in T cells, astrocytes, and brain endothelium in the CNS (14,56,100,101). However, it still remains elusive whether the functional consequences of inflammationassociated microRNA-155 are specific to each of the cell types involved or whether it represents a conserved generalized feedback mechanism of the neurovascular unit leading to cellular uncoupling during neuroinflammation.
Moreover, a recent study further supports brain endothelial microRNA-155 as a potential molecular target for improving BBB breakdown (103). The authors investigated therapeutic manipulation of microRNA-155 by using a specific anti-microRNA-155 that was injected systemically and observed that microRNA-155 inhibition at an early subacute stage in an animal model of stroke led to preservation of brain endothelial TJs and barrier integrity in vivo (103). These studies suggest that brain endothelial microRNA-155 plays an essential role in neuroinflammatory cerebrovascular pathologies that are associated with disruption of the BBB, and modulation of microRNA-155 in brain resident cells and brain endothelium may constitute a novel therapeutic approach for CNS neuroinflammatory vascular diseases.
Another microRNA therapeutic target for CNS neuroinflammatory disorders is brain endothelial microRNA-146. This microRNA was found predominately in the cerebral microvasculature of active lesions of MS and is highly abundant in the brain endothelium that is surrounded by perivascular inflammatory cuffs in the spinal cord of an EAE relapsing-remitting mice model (74). Mice that are deficient for microRNA-146 are developmentally normal at birth but acquired chronic inflammation around 5-6 mo of age (75) and had enhanced expression of several inflammatory genes, including Vcam-1, Icam-1, Sele, Mcp-1, and Egr1/3 after challenge with the proinflammatory cytokine IL-1b (75). Wu et al. (74) demonstrated that microRNA-146 negatively modulated the NF-kB signaling pathway in cultured human brain endothelium via stifling multiple genes, including IL-1 receptor-associated kinase 1, TNF receptor-associated factor 6 protein, nuclear factor of activated T cells 5, and RhoA, which regulates VCAM1 and CCL2 protein expression, and, therefore, directly resulted in decreased leukocyte adhesion to inflamed blood vessels.
Another study has analyzed the therapeutic manipulation of brain endothelial microRNA-98 and Let-7g-3p to prevent BBB dysfunction in neuroinflammation. In this study, the authors identified 2 potential targets, CCL2 and CCL5, by which microRNA-98 and Let-7g-3p may regulate leukocyte adhesion (34). These studies suggest that a brain endothelial microRNA therapeutic strategy aimed at ameliorating BBB dysfunction during neuroinflammation may be beneficiated as a result of microRNAs concomitantly modulating several molecular and cellular processes implicated in neuroinflammation.

CONCLUSIONS
Cytokines regulate the expression of brain endothelial microRNAs that either promote or inhibit inflammatory pathways to orchestrate neuroinflammation at the cerebrovascular bed. Depending on the strength and duration of the inflammatory stimuli, brain endothelial microRNAs may either regulate gene expression to counteract cytokine signaling to maintain the quiescent state or promote a gene expression profile toward an activated cerebrovascular endothelium that couldleadtoBBBbreakdown.Here,wepresentedthatsome cytokine-responsive brain endothelial microRNAs can be grouped into coexpressed microRNA clusters and families previously implicated in molecular and cellular processes of angiogenesis. Nevertheless, it remains to be determined whether these altered microRNA clusters and families might participate in increased angiogenesis that has been observed in neurologic conditions affected by neuroinflammation.
Further research in brain endothelial microRNAs will no doubt increase our understanding of the molecular processes that regulate BBB integrity in health and disease, which will tremendously benefit the development of microRNA therapeutic strategies aimed at ameliorating BBB dysfunction during neuroinflammation. Indeed, several preclinical and clinical trials have started microRNAbased therapeutics for many types of disease, including cancers that affect the CNS such as glioblastomas (104). Use of systemic administration of locked nucleic acid-modified oligonucleotide might be a promising form for the delivery of inhibitory microRNAs to the CNS (105), including the brain vasculature (103), which could be used to ameliorate or prevent CNS disorders in which microRNAs play a key pathogenic role (106). In the near future, a major challenge will be to define the spatio-temporal activities of brain endothelial microRNAs and whether microRNA therapeutic strategies are targeted to ameliorate BBB dysfunction during transient or chronic CNS disorders.