Long non-coding RNAs mediate cerebral vascular pathologies after CNS injuries

Abstract

Central nervous system (CNS) injuries are one of the leading causes of morbidity and mortality worldwide, accompanied with high medical costs and a decreased quality of life. Brain vascular disorders are involved in the pathological processes of CNS injuries and might play key roles for their recovery and prognosis. Recently, increasing evidence has shown that long non-coding RNAs (lncRNAs), which comprise a very heterogeneous group of non-protein-coding RNAs greater than 200 nucleotides, have emerged as functional mediators in the regulation of vascular homeostasis under pathophysiological conditions. Remarkably, lncRNAs can regulate gene transcription and translation, thus interfering with gene expression and signaling pathways by different mechanisms. Hence, a deeper insight into the function and regulatory mechanisms of lncRNAs following CNS injury, especially cerebrovascular-related lncRNAs, could help in establishing potential therapeutic strategies to improve or inhibit neurological disorders. In this review, we highlight recent advancements in understanding of the role of lncRNAs and their application in mediating cerebrovascular pathologies after CNS injury.

1. Introduction

Central nervous system (CNS) injuries, such as stroke, traumatic brain injury (TBI), and spinal cord injury (SCI), are one of the leading causes of morbidity and mortality worldwide, accompanied with high medical costs and a reduced quality of life (Gittleman H et al., 2017;Huang H et al., 2018;Pham TM et al., 2018). A series of events are involved in the pathogenesis of CNS injuries, such as excitotoxicity, oxidative stress, endoplasmic reticulum stress, inflammation, apoptosis, autophagy, and blood-brain barrier (BBB) dysfunction, thus resulting in neurovascular damage in the brain (Boguszewska-Czubara A et al., 2019;Lu XC et al., 2004;Sharp FR et al., 2011;Zhang L and Wang H, 2019). Secondary brain damage and neurological dysfunction after TBI, SCI, and subarachnoid hemorrhage (SAH) are associated with these pathological mechanisms as well (Boguszewska-Czubara A, et al., 2019). Noticeably, patients with severe CNS injury experience unsatisfactory outcomes. Vascular homeostasis, which plays a dominant role in maintaining physiological functions of the brain, is initially disturbed in CNS pathologies, thus leading to irreversible injury to the neural network (Michalik KM et al., 2014;Ozduman K et al., 2010). Consequently, functional remodeling of the cerebrovasculature tends to be a crucial and promising therapeutic target in CNS injury. Hence, a better understanding of the molecular mechanisms of cerebrovascular injuries and dysfunction associated with CNS disorders, which remain largely unexplored, will help in developing effective therapeutic strategies.

Proteins serving as the primary protagonists of normal cellular function has been the dominated principle of the molecular biology landscape for half a century. RNAs (such as transfer RNAs and ribosomal RNAs) have been thought to have infrastructural importance without coding for proteins (Mattick JS, 2011). A large number of non-coding transcripts have been regarded as junk DNA and transcript noises that implies total non-functionality for a period. However, an important milestone in our understanding of the complex role of RNAs in genome regulation was achieved with sequencing of the human genome. Remarkably, despite the fact that protein-coding genes account for only about 2% of the genome, increasing evidence has shown that more than 90% of the human genome is actively transcribed into functional RNA molecules during one’s lifetime (Anastasiadou E et al., 2018;Hulshoff MS et al., 2019;Jensen TH et al., 2013;Katayama S et al., 2005;Wilusz JE et al., 2009).

Long non-coding RNAs (lncRNAs) consist of a very heterogeneous group of non-protein-coding RNAs with lengths more than 200 nucleotides (Jaé N et al., 2019). The majority of lncRNAs, similar to protein-coding messenger RNAs (mRNA), are 5′ capped, spliced, and 3′ polyadenylated. However, some of the most abundant lncRNAs are processed in non-canonical ways or seem to lack processing (Jaé N, et al., 2019;Wilusz JE, 2016). Based on their nearby protein-coding genes, lncRNAs can be sub-categorized into sense, antisense, intronic, intergenic (lincRNA), bidirectional promoter lncRNAs and enhancer lncRNAs (Goff LA et al., 2015;Ma L et al., 2013). Increasing evidence indicates that lncRNAs participate in the regulation of various physiological and pathophysiological processes, such as angiogenesis by modulating the stability and nuclear retention of their target genes (Lee KT and Nam JW, 2017). LncRNAs regulate gene expression at the levels of epigenetics, transcription, post-transcription, and chromatin remodeling (Batista PJ and Chang HY, 2013;Dahariya S et al., 2019). The expression of target genes can be activated or inhibited by direct binding to the lncRNAs or the recruitment of transcription factors (Li Z et al., 2018;Zhang L and Wang H, 2019). Expression profiles of lncRNAs are associated with specific neuroanatomical regions, cell types and developmental processes, indicating the potential functional roles of lncRNAs in the nervous system (Hong SH et al., 2018;Ramos AD et al., 2013). Previous studies have shown that altered expression and function of lncRNAs are involved in mediating vascular pathology that leads to diverse neurological disorders following CNS injury (Das S et al., 2020;Dharap A et al., 2012;Qureshi IA and Mehler MF, 2012;Riva P et al., 2016;Schmitt AM and Chang HY, 2016;Yin KJ et al., 2014;Zhang X et al., 2017a;Zhang X et al., 2019a;Zhang X et al., 2018;Zhang X et al., 2017b).

To date, the exact roles and molecular mechanisms of lncRNAs still remain to be elucidated thoroughly. Currently in the neuroscience field, the functional significance of lncRNAs has been gradually explored in-depth to underline their importance in various CNS injuries (Chandran R et al., 2017). In this review, we will discuss recent progress in examining the role and regulatory mechanisms of lncRNAs associated with vascular pathology after CNS injury.

2. Cerebral vascular pathophysiology after CNS injuries

In CNS injury, normal cerebrovascular structure and function are disturbed and damaged, thus contributing to varying degrees of neurological dysfunction.

2.1. Cerebral vascular pathophysiology in ischemic stroke

An intact brain-blood barrier (BBB) is essential for maintaining physiological activities of the brain. The structure of the BBB is based on tightly connected vascular endothelial cells, which can restrict the entry of molecules and immune cells into the CNS.

As the major component of the brain microvasculature, cerebral vascular endothelial cells play a vital role in maintaining BBB integrity and cerebral homeostasis under physiological conditions (Sandoval KE and Witt KA, 2008). Evidence from experimental animal studies and clinical practices have indicated that structural and functional changes of the cerebral endothelium during cerebral ischemia contribute to BBB disruption, vascular inflammation, edema, and angiogenesis, which directly lead to brain infarction, neurological deficits and post-stroke neurovascular remodeling (Fisher M, 2008). Several mechanisms are attributed to post-stroke cerebrovascular pathophysiology. First, reactive oxidants are the main regulators for reperfusion-induced cerebrovascular endothelial injury that cause membrane damage and cell death of endothelial cells (ECs) (Yin KJ, et al., 2014). The NADPH oxidase (NOX) family, as the main source of intracellular reactive oxygen species (ROS), contributes principally to post-ischemia oxidative stress (Yang C et al., 2019). Second, disruption of tight junctions between endothelial cells results in alterations of BBB permeability and BBB breakdown following stroke. In fact, ischemic stroke induces post-transcriptional modification, translocation, and degradation of tight junction proteins, which change the structure of tight junction complexes and weaken the BBB (Jiang X et al., 2018). Third, with upregulated inflammatory mediators in endothelial cells, inflammation is also pivotal to ischemia-induced endothelial injury and subsequent enhanced leukocyte-endothelial interactions. The three mostly upregulated chemokines after hypoxia/ischemia are monocyte chemoattractant protein-1 (MCP-1/CCL2), SDF-1 (CXCL12), and MIP-1α (CCL3)(Yang C, et al., 2019). On the other hand, ECs play an important role in angiogenesis following ischemic stroke through proliferation and by also increasing the expression of endogenous pro-angiogenic molecules that may be beneficial for recovery of ischemic stroke (Hayashi T et al., 2003). Stroke patients with increased microvessel density seemed to survive longer and have a better prognosis than those with lower microvessel density (He Z et al., 2020;Jensen TH, et al., 2013). Thus, promoting angiogenesis of the injured brain tends to be a potent therapeutic strategy for the treatment of CNS injuries in the clinical application.

2.2. Cerebral vascular pathophysiology in hemorrhagic stroke

Hemorrhagic stroke causes cerebrovascular injury through multiple mechanisms and is regulated by many signal transduction pathways. Primary and the secondary neurovascular injuries are involved in the pathological process of intracerebral hemorrhage (ICH). Primary neurovascular injury is caused by the occupying effect of the hematoma. At the same time, more serious secondary neurovascular damage is induced by toxic effects of blood and decomposition products from blood cells such as hemoglobin, enzymes, and iron that lead to a variety of molecular, cellular, and biochemical reactions, such as inflammation, cell apoptosis, lipid peroxidation, free radical damage, and glutamate excitatory toxicity (Zheng H et al., 2016).

Subarachnoid (SAH), which is caused by rupture of intracranial aneurysms (IAs), accounts for 85% of all cases and is a main cause of severe disability and death (D’Souza S, 2015). Complex mechanisms such as excessive production of ROS, vascular smooth muscle cell (VSMCs) phenotypic alteration, immune responses, and inflammation all contribute to deterioration of the vascular wall that eventually can result in IA formation (Jin D et al., 2019). In addition, vascular endothelial growth factor (VEGF), which is a molecule promoting angiogenesis, plays a dual role in the pathophysiology of SAH. Pathologically increased VEGF could be observed within 24h in the brain stem after SAH, and downregulation of VEGF is related to reduction of SAH-induced brain edema. Either VEGF expression or injury could be decreased by hyperbaric oxygen (HBO) treatment, indicating that HBO may have positive effect on the prognosis of SAH patients (Ostrowski RP et al., 2005).

The integrity of the basement membrane (BM), whose main components include collagen IV and laminin, is of great significance for normal structure and function of the BBB. There is a positive correlation between basement membrane destruction and hemorrhagic stroke, and studies showed that loss or degradation of basement membrane was associated with the occurrence of both ICH and SAH (Yao Y, 2019).

Alterations in the cellular composition of the vessel wall are also important for the pathological changes caused by hemorrhagic stroke. Degradation of the extracellular matrix (ECM), along with activation and dysfunction of endothelial cells (ECs), occurs during vascular insult (Gareev I et al., 2020). In response to vascular injury, VSMCs may migrate into the intima and switch their phenotype, contributing to restoration of the vessel wall (Gareev I, et al., 2020). Inflammatory factors are involved in the response of the cerebrovascular wall to injury as well. MMPs and tumor necrosis factor (TNF) promote the influx of macrophages and permanent degradation of collagen and elastin fibers (Gareev I, et al., 2020).

2.3. Cerebral vascular pathophysiology in traumatic brain injury

The disrupted cerebral vasculature is involved in the pathogenesis of TBI, which leads to hemorrhage, edema, blood flow abnormalities, and BBB disruption followed by hypoperfusion, altered delivery of metabolic substrates as well as hypoxic and ischemic tissue damage (Salehi A et al., 2017). The primary pathological alternations occur promptly after TBI, which include increased BBB permeability, diffuse axonal injury, and brain contusions. TBI patients always suffer from BBB disruption (Ho KM et al., 2014). In fact, reactive oxygen species generated after TBI ultimately contribute to BBB dysfunction. Thereafter, neuroinflammation and the formation of ROS can be enhanced by BBB destruction itself (Abdul-Muneer PM et al., 2015). Persistent inflammation will eventually lead to myelin loss, thus inhibiting functional recovery (Glushakova OY et al., 2014). In particular, BBB breakdown may give rise to calcium perturbations of brain cells, thus inducing secondary injury (Logsdon AF et al., 2015). Targeting BBB restoration and vascular remodeling after injury have been emphasized in the development of therapeutic strategies for TBI.

Emerging research suggests that the injured cerebral vasculature undergoes repair in which angiogenesis and vasculogenesis are the two primary processes (Salehi A, et al., 2017). Studies showed the microvascular density in the traumatic cortex recovered to near-sham levels in moderately injured animals by 14 days after injury; whereas the microvascular density in severely injured animals at 14 days after injury was reduced compared to the contralateral hemisphere with an abnormally repaired vasculature. Although neovascularization and recovery of microvascular tone is initiated as a compensatory response following injury, severe injury still tended to result in substantial loss of microvascular function (Park E et al., 2009). That is, abnormal repair of the vasculature may also be an important contributing factor to the poor outcomes in rodent models and patients with TBI.

2.4. Vascular pathophysiology in spinal cord injury

Studies have shown that SCI-induced damage to motor function was related to abnormal CNS capillary contraction, which was affected by the release of serotonin and norepinephrine from brainstem neurons. After spinal cord injury, although the synthesis of 5-hydroxytryptamine (5-HT1) is reduced, altered activity of pericyte monoamine receptors leads to local contraction of capillaries and reduced blood flow, ultimately resulting in local ischemia and hypoxia. Inhibition of monoamine receptors or reduction of trace amine synthesis can significantly alleviate motor function after SCI (Cheng J et al., 2018;Li Y et al., 2017).

The loss of vessel wall integrity is another mechanism of CNS injury caused by SCI. Studies reported that the release of elastase by neutrophils from SCI injury sites not only played a key role in vascular injury, but also affected the production of angiopoietin. The inhibition of neutrophil elastase by sivelestat could effectively promote angiogenesis and be a potential strategy for SCI treatment (Kumar H et al., 2018).

Pathological processes of spinal cord-related neurovascular diseases are correlated with vascular alterations, which are initiating factors of sustained impairment of nerve function. By enhancing proliferation and migration of endothelial cells, VEGF is a key factor for angiogenesis in the CNS and is also critical for promoting neuroprotection and neuronal proliferation (Storkebaum E et al., 2004). Therefore, upregulating VEGF expression to promote angiogenesis after SCI is one of the potential strategies conducive to improving the prognosis of SCI patients.

3. Functional properties of lncRNAs in cerebrovascular biology and pathology

3.1. Overview lncRNAs

Long non-coding RNAs (lncRNAs) are a class of large and diverse transcriptional RNA molecules with over 200 nucleotides in length that do not encode proteins, and thus distinct from protein-coding genes and microRNAs. Transcribed by RNA polymerase II, most of lncRNAs undergo alternative splicing, polyadenylation, and 5’-capping (Schmitz SU et al., 2016). The functions of lncRNAs are of great variety, and their specific mechanisms are associated with their subcellular localization. In addition to the well-known role of nuclear lncRNAs in epigenetic regulation of genes, they also participate in regulation of both gene transcription and pre-mRNA splicing. LncRNAs in the cytoplasm mainly act as regulators at the post-transcriptional level, by sponging microRNAs or modulating mRNA stability and translation (Jaé N, et al., 2019;Kok FO and Baker AH, 2019).

3.2. LncRNAs in brain vascular biology and pathology

Brain vascular endothelium, as an important part of the blood-brain barrier, is essential in maintaining cerebrovascular homeostasis by adapting to dynamical alterations in the microenvironment. It has been revealed that a variety of non-coding RNAs play crucial roles in vascular biology and retaining physiological function, including angiogenesis, vascular repair, senescence, and inflammatory signaling responses (Suárez Y and Sessa WC, 2009;Wang S and Olson EN, 2009;Yin KJ et al., 2015).

Recently, increasing studies have shown that lncRNAs in vascular endothelial cells appear to be key regulators of cerebrovascular biology (Figure 1). Different lncRNAs are expressed in vascular endothelial cells at different differentiation stages. The expression of AGAP2-AS1 (PUNISHER) is only detected in terminally differentiated ECs (Kok FO and Baker AH, 2019). Expression of PUNISHER is correlated with transcripts involved in vascular development, but negatively associated with modificatory factors related to the cell cycle, chromatin, and DNA damage-response genes. Silencing PUNISHER in vivo could lead to severe vascular defects and impaired vascular endothelial function (Jaé N, et al., 2019). Another important lncRNA functioning in the early phase of vessel development is SENCR, which is an abbreviation of Smooth muscle and Endothelial cell enriched migration/differentiation-associated long Non-Coding RNA. SENCR upregulation during EC differentiation enhances early mesodermal and endothelial cell commitment. In HUVECs, it was positively associated with proliferation, migration, and angiogenesis (Boulberdaa M et al., 2016). For example, it was reported that SENCR improves angiogenesis through regulating two proangiogenic genes C-C chemokine ligand 5 (CCL5) and Fractalkine/CX3C chemokine ligand 1 (CX3CL1)(Boulberdaa M, et al., 2016). Studies also showed that SENCR affected the integrity of the plasma membrane via restraining CKAP4 to the cytosol.Then CKAP4 bonded to the adherence junction protein VE-cadherin (CDH5) to trigger CDH5 internalization and subsequently increased endothelial permeability (Lyu Q et al., 2019). A reduced level of SENCR was found both in human critical limb ischemia and coronary artery disease samples compared to controlled subjects (Boulberdaa M, et al., 2016). In addition, SENCR may participate in cellular motility and biological differentiation of human coronary artery smooth muscle cells (HCASMCs), since genetic deletion of SENCR led to actin cytoskeleton reorganization and lamellipodia formation (Bell RD et al., 2014). Moreover, SENCR downregulation resulted in the decreased level of myocardin (MYOCD), which is a key transcriptional regulator for contractile gene in smooth muscle cells (Bell RD, et al., 2014), resulting in the upregulation of several pro-migratory genes. Consistently, SENCR was identified as a VSMC migratory inhibitor by loss-of-function studies (Jaé N, et al., 2019). Tyrosine kinase containing immunoglobulin and epidermal growth factor homology 1 (Tie1), a receptor tyrosine kinase specifically expressed in vascular endothelial and hematopoietic cells, is critical for the regulation of angiogenesis and maintenance of vascular integrity (Chowdhury TA et al., 2018;Puri MC et al., 1995). LncRNA tie-1 antisense (AS) downregulated the Tie-1 level in embryonic zebrafish by directly binding Tie-1 mRNA, leading to mislocalization of CDH5 and zona occludens 1 (ZO-1) and resulting in impaired tight junctions of vascular endothelium both in vivo and in vitro. lncRNA Tie-1-AS downregulated VEGF-induced endothelial tube formation in cultured HUVECs. Decreasing the level of the Tie-1 protein damages vascular endothelial cell junctions and subendothelial network stability, (Li K et al., 2010). Antisense transcript of GATA binding protein 6 (GATA6-AS), increased in ECs during hypoxia, was reported as another angiogenetic regulator (Neumann P et al., 2018). GATA transcription factors stand for a family of zinc-finger DNA-binding proteins that target consensus DNA sequence (T/A)GATA(A/G)(Cantor AB, 2018). Silencing of GATA6-AS could lead to decreasing endothelial-mesenchymal transition induced by TGF-β2 in vitro and promote vascular formation in mice. Mechanistically, GATA6-AS negatively regulated the function of nuclear epigenetic regulator lysyl oxidase-like 2 (LOXL2) to alter histone methylation and modulate gene expression in ECs (Neumann P, et al., 2018). Spliced-transcript endothelial-enriched lncRNA (STEEL) is a set of EC-enriched lncRNAs highly expressed in ECs. STEEL participates in the regulation of angiogenesis, shear stress responsiveness and macrovascular/microvascular identity. STEEL facilitates angiogenesis in vitro, as well as promotes vascular growth and maintains vascular integrity in vivo. Precisely, STEEL modulates vascular structure and function by regulating various genes in ECs. It interacts with PARP1 (poly [ADP-ribose] polymerase 1) and forms a complex to regulate expression of multiple genes. Both endothelial nitric oxide synthase (eNOS) (an essential regulator of vasomotor tone), and the transcription factor Kruppel-like factor 2 (KLF2, a significant sensor of hemodynamic forces), were upregulated by STEEL (Man HSJ et al., 2018).

Figure 1.

The role and regulatory mechanisms of lncRNAs involved in cerebrovascular biology and pathology. Abundant lncRNAs in vascular endothelial cells and vascular smooth muscle cells participate in the modulatory process of cerebral angiogenesis, vascular development, vascular integrity and migration via various signaling cascades.

Dysregulation of vascular function will result in consequential pathological alterations in CNS injuries. Therefore, thorough understanding of vascular modulation is significant for developing potential therapeutic approaches. In recent years, cumulative studies have also begun to focus on examining the role of lncRNAs’ pathological expression in vascular cells, indicating that abnormal lncRNA expression could be associated with vascular lesion in atherosclerosis, coronary artery disease, CNS injuries, and cancer (Ghafouri-Fard S et al., 2020;Wang Y et al., 2018;Zhang L and Wang H, 2019)(Figure1). In our previous studies, we have identified lncRNA expressional profiles in oxygen-glucose deprivation (OGD)-treated mouse BMECs via RNA-sequencing technology. After 16h of OGD exposure, the expression levels for 362 of the 10,677 lncRNAs analyzed changed significantly. In particular, a total of 147 lncRNAs increased and 70 lncRNAs decreased by more than 2-fold. The most significantly upregulated lncRNAs were small nucleolar RNA host gene 12 (SNHG12), Metastasis-associated lung adenocarcinoma transcript 1 (MALAT1), and lnc-OGD 1006, while the most highly downregulated lncRNAs consisted of 281008D09Rik, Peg13, and lnc-OGD 3916, which were further verified both in vitro and in vivo experiments. These findings revealed that endothelial lncRNAs are available to function as principal regulators in mediating cerebrovascular pathologies after ischemic stroke. (Zhang J et al., 2016). MALAT1 regulatory functions targeting endothelial cells occur via modulating the expression of cell cycle regulators. Hypoxia is an inducer of MALAT1 expression and silencing of MALAT1 in endothelial cells inhibits proliferation and functional sprouting angiogenesis (Michalik KM, et al., 2014). Maternally expressed gene 3 (MEG3) regulates endothelial cell proliferation by interacting with other non-coding RNAs, such as miR-21, and inhibiting EC proliferation and migration (Wu Z et al., 2017a). Researchers suggested that MEG3 might be an essential regulator in cerebral vascularization and might inhibit tumor growth partially by suppressing angiogenesis. The expression levels of genes involved in the VEGF pathway including VEGFA, VEGF receptors 1 (VEGFR1), Neuropilin 1 (Nrp1), Wiskott-Aldrich syndrome-like (Wasl), IQ motif containing GTPase activating protein 1 (Iqgap1), were significantly upregulated in MEG3-knockout brains of mouse embryos, resulting in increased microvascular density (Gordon FE et al., 2010). Additionally, lncRNA MANTIS is a scaffolding lncRNA in a chromatin-remodeling complex and its expression is mediated by the H3K4 lysine-specific demethylase 5B (JARID1B), which is a histone demethylase. MANTIS can modulate gene expression of ECs by interacting with Brahma Related Gene 1 (BRG1), which is the catalytic subunit of the switch/sucrose nonfermentable chromatin-remodeling complex, and subsequently affect the expression of pro-angiogenetic genes SMAD6, Sex determining region Y-box 18 (SOX18), or chicken ovalbumin upstream promoter-transcription factor II (COUP-TFII). MANTIS is positively correlated with angiogenesis, acting as a mediator for transcription of efficient endothelial genes (Leisegang MS et al., 2017). MIAT overexpression was observed in cerebral microvascular endothelial cells (CMECs) after cerebral ischemia. It serves as a competitive endogenous RNA (ceRNA) to sponge miR-204-5p and simultaneously upregulate high-mobility group box 1 (HMGB1). Inhibition of MIAT increased the miR-204-5p level and decreased HMGB1 expression, which promoted angiogenesis and reduced cerebral ischemia-induced CMEC damage (Deng W et al., 2020). Antisense non-coding RNA in the INK4 locus (ANRIL) overexpression increased the VEGF level and enhanced angiogenesis via activation of the NF-κB signaling pathway in a rat model with diabetes mellitus and cerebral infarction (Zhang B et al., 2017). Moreover, taurine upregulated 1 (TUG1) was demonstrated to promote tumor-induced angiogenesis by upregulating vascular endothelial growth factor A (VEGFA) and inhibiting miR-299 in human glioblastoma. TUG1 knockdown in a xenograft glioblastoma model in vivo resulted in suppression of tumor growth (Cai H et al., 2017). In the next section, the role of lncRNAs and their correlation to vascular pathologies in CNS injury will be discussed thoroughly.

4. LncRNAs associated with CNS injuries

To date, most studies have focused on the regulatory effects of lncRNAs on neurons in CNS injuries, including neuronal autophagy, apoptosis and neuroinflammation. As summarized in Figure 2, lncRNAs actively participate in the pathological process of CNS injuries. For example, MALAT1 could modulate the pathological process of ischemic stroke by interacting with other non-coding RNAs. A study demonstrated a role of the Malat1/miR-181c-5p/HMGB1 axis in mice after middle cerebral artery occlusion (MCAO), in which MALAT1 acted as a competing endogenous RNA (ceRNA) and sponge that targets miR-181c-5p and regulates berberine-induced anti-inflammation effects in ischemic stroke (Cao DW et al., 2020). Wang et. al. found that by competitively binding to miR-145 and affecting aquaporin-4 (AQP4) expression, MALAT1 promoted cerebral ischemia-reperfusion injury (Wang H et al., 2020). Additionally, a previous study showed that MALAT1, as a potent autophagy inducer, protected BMECs during cerebral ischemic injury by sponging miR-26b as well as increasing ULK2 expression (Li Z et al., 2017)., suggesting that MALAT1 might play multiple roles through different signaling pathways. MALAT1 also downregulated the miR-30a level via sponging, which subsequently led to de-repression of Beclin-1 and promotion of neuronal cell death induced by autophagy. Inhibition of MALAT1 in mice could decrease these effects and reduce infract volumes and neurological dysfunction (Guo D et al., 2017).

Figure 2.

Summary of lncRNAs in the regulation of CNS injury. Various lncRNAs actively participate in the pathological processes of ischemic stroke, hemorrhagic stroke, TBI and SCI. LncRNAs in the red frame prompt brain or spinal cord damage and inhibit functional recovery, while lncRNAs in the green frame exert protective and/or restorative effects. These regulatory lncRNAs may function as promising pharmacological or molecular targets in further developing novel lncRNA-based therapeutic strategies in CNS injury.

It was reported that lncRNA MEG3 and mRNA Sema3A were upregulated in ischemic stroke, accompanied by a decrease of miR-424-5p. Sema3A is a downstream target of miR-424-5p, which participates in various biological process such as apoptosis and outgrowth of axons. Downregulated levels of phosphorylated JNK and p38 were identified after genetic deletion of Sema3A, indicating that it could activate the MAPK pathway. By upregulating Sema3A and activating the MAPK pathway, MEG3 increased cell apoptosis and repressed cell viability, while inhibiting MEG3 could protect against cerebral ischemia injury (Xiang Y et al., 2020). MEG3 also activates p53 and promotes ischemic neuronal death in vitro by directly binding to the p53 DNA binding domain. Blocking MEG3-p53 binding could increase neuronal survival in ischemic stroke both in vitro and in vivo (Yan H et al., 2016). MEG3 also modulates ischemic neuronal death as a ceRNA by targeting the miR-21/PDCD4 signaling pathway, while knockdown of MEG3 improved overall neurological function and reduced ischemic damage in vivo (Yan H et al., 2017). In addition, MEG3 serves a ceRNA for miR-181b. An increased MEG3 level subsequently downregulated miR-181b and upregulated 12/15-LOX (lipoxygenase) in both HT22 cells and brains with hypoxia, while silencing MEG3 or elevating the miR-181b level could separately reduce hypoxia-induced HT22 cell apoptosis and improved mouse motor function after MCAO (Liu X et al., 2016). It has been reported that lncRNA H19 was upregulated by hypoxia and could induce cerebral ischemia reperfusion injury by activating autophagy, while H19 knockdown could promote axon sprouting and improve functional recovery. Silencing H19 could protect SH-SY5Y cells from OGD-induced death. Also, mice with H19 knockdown showed significant improvement of motor task completion following ischemic stroke (Hu S et al., 2020;Wang J et al., 2017). Increasing lncRNA TUG1 was found under ischemia, and may participate in promoting neuronal apoptosis, regulating microglial polarization, and production of inflammatory cytokines. Enhanced TUG1 expression was observed in mouse brains after ischemic stroke and cultured SH-SY5Y cells after OGD, leading to a larger brain infarction volume or a higher apoptotic rate. Mechanistically, studies revealed that TUG1 sponged miR-9 in ischemic injury. Inhibition of miR-9 by TUG1 could cause increased expression of Bcl2l11, resulting in a cytotoxic effect, whereas knockdown of TUG1 had protective effects (Chen S et al., 2017;Wang H et al., 2019). LncRNA N1LR, an I/R-induced lncRNA, acts as a protector in neurons against ischemic stroke and also reduces OGD/R-induced apoptosis in N2a cells via inhibiting p53 activation. (Wu Z et al., 2017b;Wu Z et al., 2017c). It was reported that lncRNA FosDT, which was upregulated in rat brains after transient focal ischemia, aggravated behavior disorders and brain injury after stroke by interacting with chromatin modification proteins and the subsequent increase of repressor element-1 silencing transcription factor (REST) downstream genes glutamate receptor ionotropic NMDA 1(GRIN1), NFκB2, and AMPA receptor subunit 2 (GRIA2) in the postischemic brain (Mehta SL et al., 2015). Silencing FosDT by intracerebral injection of FosDT siRNA effectively improved motor function and reduced brain infarct volume in rats after MCAO. Therefore, lncRNA-mediated epigenetic remodeling can affect the prognosis of stroke and targeting FosDT could be a potential therapeutic to reduce brain damage after stroke (Dharap A et al., 2013;Mehta SL, et al., 2015). Recently, the inhibition of autophagy and neuronal apoptosis were found to be related to the interference of lncRNA small nucleolar RNA host gene 3 (SNHG3) by upregulating miR-485 and downregulating ATG7 in brain ischemia-reperfusion (I/R) injury (Lee KT and Nam JW, 2017). In MCAO mice, knockdown of SNHG6 could reduce brain infarct size and promote the recovery of neurological function. Consistently, the inhibition of SNHG6 increased cell viability and inhibited cell apoptosis in OGD-induced neuronal cells through targeting the miR-181c-5p/BIM signaling pathway (Zhang X et al., 2019b). The expression of SNHG14 was increased after OGD in BV-2 cells and ischemic cerebral tissues, which suppressed miR-145-5p and upregulated pro-inflammatory factor cPLA2 group IVA (PLA2G4A), eventually promoting apoptosis in BV-2 cells. Downregulation of SNHG14 by use of si-SNHG14 could reverse these detrimental effects (Qi X et al., 2017). LncRNA C2dat1 was significantly overexpressed in both OGD-treated N2a cells and ischemic mouse brains, accompanied by the upregulation of calcium/calmodulin-dependent kinase II delta (CAMK2D), which may promote neuronal survival (Wilusz JE, 2016). Also, C2dat1 and C2dat2 protected neurons from ischemic injury by upregulating CaMKIIδ and IKKα/β and further activating the NF-κB signaling pathway (Ma L, et al., 2013;Wilusz JE, 2016;Xu Q et al., 2016). Epigenetically-induced lncRNA 1 (EPIC1) was found to increase the expression of MYC targets (Cyclin A1, CDC20 and CDC45) by ectopic overexpression, thus significantly attenuating neuronal cell death and apoptosis induced by H₂O₂ (Goff LA, et al., 2015).

It is well recognized that secondary brain injury following ICH is caused by activated parenchymal inflammatory responses, and also both necrotic and apoptotic neuronal death. A recent study by Chen et al. suggested the overexpression of lncRNA Mtss1 could aggravate inflammatory brain injury following ICH through increasing inflammatory cytokine secretion and targeting miR-709 (Chen JX et al., 2020). H19 was observed to be significantly upregulated in ICH brains up to 7 days after hemorrhage. H19 enhances neuroinflammation by promoting histone deacetylase 1-dependent microglial polarization (Kim JM et al., 2019). Another lncRNA Ptprj-as1, the antisense strand of Ptprj gene, was highly expressed in ICH-induced inflammation in tissues. MN9D cells with overexpressed Ptprj-as1 that were treated with lipopolysaccharide (LPS) showed higher apoptotic death compared to the LPS only treatment group, indicating that Ptprj-as1 might be associated with apoptosis. Ptprj-as1 is also involved in inflammatory injury by activating the NF-κB pathway in microglia and upregulating the secretion of inflammatory cytokines (Wen J et al., 2018). Early brain injury (EBI) is another main cause of disability and mortality in patients with SAH, accompanied with acute pathophysiological events, such as BBB disruption, inflammation, and neural cell death. Previous studies showed that lncRNAs were involved in EBI after SAH by regulating broad signaling pathways. In SAH patients, the expression of lncRNA MEG3 was significantly higher, which was positively associated with its severity. Neuronal activity was decreased and apoptotic cell death was increased due to the overexpression of MEG3 via the PI3k/Akt pathway (Liang Z et al., 2018). Yang et al. demonstrated that the H19-miR-675-P53 and H19-let-7a-neural growth factor (NGF) apoptotic signaling pathways were involved in EBI after SAH. Additionally, melatonin treatment exerted protective effects by upregulating the expression of H19 in EBI after SAH. H19 was identified to promote the expression of miR-675, and miR-675 was a negative regulator of P53. Meanwhile, H19 was revealed to be a competing non-coding RNA for let-7a, while NGF was a target gene of let-7a. P53 and NGF eventually modulated cell apoptosis, thus affecting progression of EBI after SAH (Yang S et al., 2018).

Gm4419, a lncRNA targeting miR-4661, increased in astrocytes of TBI mice, and tended to alleviate brain inflammation (Zhang L and Wang H, 2019). Additionally, lncRNA NEAT1 was identified as a crucial inhibitory factor for apoptosis after TBI (Hirose T et al., 2014). Both NEAT1 and Gm4419 played neuroprotective roles in TBI and promoted the recovery of neurological function (Zhang L and Wang H, 2019). NEAT1 and MALAT1 secreted from human adipose-derived stem cells (hADSCs) are beneficial to motor, cognitive function, and contribute to decreased secondary lesion volume following TBI in young rats (Tajiri N et al., 2014). LncRNA colorectal neoplasia differentially expressed (CRNDE) is highly expressed in multiple cancerous diseases, especially in brain cancers (Zheng J et al., 2016). The expression of CRNDE was also found to be upregulated in a rat TBI model. CRNDE downregulation induced differentiation of neurons and directional growth and regeneration of nerve fibers by promoting the expression of glial fibrillary acidic protein, BrdU, nerve growth factor, nestin, and neuronal nuclei. Besides, downregulation of CRNDE could improve neurobehavioral function, elevate the number of Nissl bodies, and inhibit neuroinflammatory factors as well as neuronal apoptosis and autophagy in TBI models (Yi M et al., 2019). In patients with TBI, the expression of MEG3 was significantly lower than that of healthy controls, and the levels of inflammatory cytokines in plasma (TNF-α, IL-1β, IL-6, and IL-8) were significantly higher than the control group, indicating a negative correlation between MEG3 and inflammatory cytokines. In addition, patients with high MEG3 and low inflammatory cytokine levels had a better prognosis (Shao HF et al., 2019). A recent study by Cheng et al. showed that lncRNA homeobox transcript antisense RNA (HOTAIR) was abnormally highly expressed in activated microglia in TBI. Knockdown of HOTAIR could inhibit microglia activation and the release of inflammatory factors by promoting Nrdp1-mediated ubiquitination of MYD88 protein, indicating that HOTAIR might be a promising therapeutic target for TBI (Cheng S et al., 2021).

Previous studies suggested the lncRNA brain-derived neurotrophic factor antisense (BDNF-AS)/miR-130b-5p/PRDM5 axis might be a promising therapeutic target for acute SCI (ASCI), since BDNF-AS knockdown could suppress neuronal cell apoptosis via miR-130b-5p/PRDM5 axis (Zhang H et al., 2018a). Besides, hydrogen sulfide was indicated as an important neuroprotective agent in spinal cord ischemia-reperfusion injury (SCII) rat model by upregulating the expression of lncRNA CasC7 (Liu Y et al., 2018). LncSCIR1 was reported to be downregulated in rats following SCI. Its inhibition contributed to gliosis by targeting various molecules and modulating inflammation, axonal growth, neural regeneration, and cell death (Chandran R, et al., 2017). In ASCI model rats, lncRNA DGCR5 was found to suppress neuronal apoptosis and ameliorate ASCI by directly binding or negatively regulating PRDM5 (Zhang H et al., 2018b). Overexpression of the lncRNA, TCTN2 contributed to enhanced autophagy via targeting the miR-216b-Beclin-1 pathway to ameliorate neuronal apoptosis and relieve SCI (Zhang H, et al., 2018b). In addition, Jiang et al. showed that lncRNA SNHG5 could enhance astrocyte and microglia viability by upregulating KLF4 in SCI. Increased SNHG5 had no effect on neurological function in rats after SCI (Jiang ZS and Zhang JR, 2018). LncRNA Airsci was the most upregulated lncRNA associated with the NF-κB signaling pathway in SCI. Motor function recovery in SCI rats could be achieved through lncRNA Airsci-siRNA application to alleviate the inflammatory response via inhibiting the NF-κB signaling pathway (Zhang T et al., 2021). LncRNA ZNF667-AS1 was gradually downregulated with the prolongation of SCI. As an anti-inflammatory molecule, ZNF667-AS promoted SCI recovery via inhibiting the JAK-STAT pathway (Li JW et al., 2018).

5. LncRNAs mediating cerebral vascular pathology in CNS injuries

Previous studies of lncRNAs after CNS injury have mainly focused on lncRNA-based neuronal restoration and mechanisms. The roles and regulatory mechanisms of vascular-associated lncRNAs in CNS injury has attracted less attention. Noticeably, vascular protective strategies are of vital importance in the repairment of CNS injury on account that blood vessels nourish neurons and provide prerequisites for neural restoration. Meanwhile, the destruction and reconstruction of vascular homeostasis are also the core of pathogenesis in the context of CNS injury. Recently, the association between lncRNAs and CNS injury in mediating neurovascular pathologies has gained great attention. Herein, we concentrate on a class of vascular-related lncRNAs and provide a framework of lncRNA functional significance correlating to cerebral vascular pathologies after CNS injuries (Table 1).

Table 1.

The expression, functions, and molecular targets of lncRNAs in cerebral vascular pathologies after CNS injury.

CNS injuries	LncRNAs	Animals or Cells	Models	Expression	Function	Molecular Targets	Reference
Ischemic stroke	ANRIL	Rats	MCAO combining diabetes	increased	alleviate inflammation, promote angiogenesis	NF-κB,VEGF	(Zhang B et al., 2017)
	DANCR	BMECs	OGD	increased	promote survival of OGD-injured BMECs and angiogenesis	miR-33a-5p/XBP1	(Zhang M et al., 2020)
	MALAT1	bEnd.3 cells	OGD	increased	suppress apoptosis, increase the cell viability	miR-205-3p, PTEN	(Gao Q and Wang Y, 2020)
		BMEC, mice	OGD, MCAO		inhibit apoptosis and inflammation	Bim/IL-6, MCP-1, E-selectin	(Zhang X et al., 2017)
		Plasma	Acute ischemic stroke			IL-10	(Ren H et al., 2020)
						IL-6, IL-8, IL-22, CRP and TNF-α
	MEG3	Mice, hBMEC	MCAO, OGD	increased	promote apoptosis		(Wang M et al., 2020)
	MIAT	Rats, BMEC	MCAO, OGD	increased	inhibit angiogenesis, increase damage	MIAT/miR-204-5p/HMGB1	(Deng W et al., 2020)
	NEAT1	BMECs	OGD	increased	promote the angiogenesis and survival of BMECs	miR-377	(Zhou ZW et al., 2019)
	SNHG1	Mice	MCAO	increased	improve brain infarct size and neurological scores	miR-18a, HIF-1α/VEGF	(Zhang L et al., 2018)
		BMEC, hCMEC/D3 cells	OGD/R		promote the angiogenesis of BMECs	miR-199a	(Wang Z et al., 2018)
			OGD		alleviated apoptosis and inflammation	miR-376a	(Lv L et al., 2020)
			OGD		reduced cell apoptosis, alleviated OGD induced injury	miR-338/HIF-1α	(Yang X and Zi XH, 2019)
	SNHG12	BMEC	OGD/R, I/R	increased	reduced inflammatory response and cell death	MiR-199a	(Long FQ et al., 2018)
		bEnd.3 cells	OGD		promote angiogenesis	MiR-150/VEGF	(Zhao M et al., 2018)
Hemorrhagic stroke	SNHG3	BMECs	ICH	increased	inhibit BMEC proliferation and migration, and promote cell apoptosis and monolayer permeability	TWEAK/Fn14/STAT3	(Zhang J et al., 2019)
	FENDRR	Mice	HICH	increased	promoted the apoptosis of HBMECs	miR-126, VEGFA	(Dong B et al., 2018)
	MALAT1	Rats	SAH	increased	reduce endothelial cell apoptosis and endothelial cell viability	MALAT1/miR-143/VEGFA	(Gao G et al., 2020)
TBI	MALAT1	HUVEC, mice	hypoxia/hindlimb ischemia	increased	promote cell proliferation and vascularization	VEGF	(Michalik KM et al., 2014)
SCI	Xist	Rats	CCSCI	increased	promote angiogenesis and microvessel density	miR-32-5p/Notch-1	(Cheng X et al., 2020)

5.1. LncRNAs in cerebral vascular pathology after ischemic stroke

Using new technologies such as RNA-seq, deep sequencing, and microarrays, researchers have revealed a number of differentially expressed lncRNAs, including MALAT1, MEG3, lncRNA-H19, taurine upregulated gene1(TUG1), N1LR, Fos downstream transcript (FosDT), small nucleolar RNA host gene 14 (SNHG14), and CaMK2D-associated transcript 1 (C2dat1), that affect functions such as neurogenesis, angiogenesis, and inflammation through regulating gene expression in ischemic stroke (Bao MH et al., 2018). Herein, we mainly discuss lncRNAs regulating vascular pathological processes in ischemic stroke.

MALAT1 is one of the most extensively studied lncRNAs, which was initially identified as promoting proliferation and metastasis of different tumors (Zhang X, et al., 2017a). MALAT1 is highly expressed in vascular endothelial cells, cardiomyocytes, and skeletal muscle, contributing to pathological angiogenesis and myogenesis. The genetic deletion of MALAT1 in mice reduced retinal endothelial proliferation and vascular growth, which might alleviate diabetic retinopathy (Liu JY et al., 2014). In our previous studies, we found that inhibition of MALAT1 by using locked nucleic acid (LNA) GapmeRs significantly reduced tube formation, cell migration, and cell proliferation in cultured mouse primary skeletal muscle microvascular endothelial cells. Genetic deficiency of Malat1 led to reduced blood vessel formation and local blood flow perfusion in mouse hindlimbs at one to four weeks after hindlimb ischemia. We documented that MALAT1 regulates cell-autonomous angiogenesis through direct regulation of VEGFR2 (Zhang X, et al., 2018). Furthermore, we revealed that MALAT1 also played important roles in the pathogenesis of ischemic stroke. The expression of MALAT1 was upregulated in brain microvascular endothelial cells (BMECs) after OGD and in cerebral microvessels isolated from mice after transient focal cerebral ischemia (Zhang J, et al., 2016). Genetic ablation of MALAT1 resulted in a worsened prognosis with larger infract size and impaired sensorimotor and neurological functions in mice after MCAO. Moreover, compared to wild-type controls, we demonstrated that the expression levels of pro-apoptotic and pro-inflammatory cytokines in the cerebral cortex were significantly increased in MALAT1 KO mice after ischemic stroke, Silencing of MALAT1 significantly upregulated the pro-apoptotic factor Bim and pro-inflammatory cytokines such as interleukin-6 (IL-6), monocyte chemotactic protein-1 (MCP-1) and E-selectin in ischemic stroke models both in vitro and in vivo. Furthermore, MALAT1 could bind to Bim and E-selectin directly to act as a protective mediator in ischemic stroke, confirming that MALAT1 plays an anti-apoptotic and anti-inflammatory role in brain microvasculature to reduce cerebrovascular and cerebral parenchymal injury caused by ischemic stroke (Zhang X, et al., 2017b). Consistent with our findings, one group reported that MALAT1 expression was positively correlated with the anti-inflammatory factor, IL-10, but negatively correlated with pro-inflammatory factors, such as IL-6, IL-8, IL-22, CRP, and TNF-α (Ren H et al., 2020). Studies from another group also confirmed that expression of MALAT1 was decreased in OGD-induced apoptosis of bEnd.3 cells, which acted as the competing endogenous RNA (ceRNA) for miR-205-3p to affect miR-205-3p expression, thus further regulating phosphatase and tensin homolog deleted on chromosome ten (PTEN) expression, whereas knockdown of MALAT1 inhibited the activation of caspase-3 (Gao Q and Wang Y, 2020). Of note, lncRNA MALAT1 may participate in different stages of ischemic stroke. Although there was no statistical correlation between the expression of MALAT1 and recurrence-free survival (RFS), a trend for longer RFS in patients with a high expression of MALAT1 compared to those with low expression was observed (Ren H, et al., 2020). Thus, MALAT1 is considered as an important therapeutic target for ischemic stroke.

Studies found the expression of MEG3, a lncRNA promoting apoptosis of human brain microvascular endothelial cells (hBMECs), was significantly upregulated in patients with ischemic stroke within 48 hours of onset. In vivo experiments showed a decreased survival time of ischemic stroke mice with higher MEG3 expression, which was consistent with the notion that high MEG3 expression in patients often indicates poor prognosis (Wang M et al., 2020).

The expression of lncRNA SNHG1 was remarkably increased in mouse brains after MCAO and in cultured brain microvascular endothelial cells after OGD, which was negatively correlated with miR-18a. Inhibition of SNHG1 contributed to a greater brain infarction size and worsened neurological scores in stroke mice along with increased OGD-induced caspase 3 activity and apoptosis in cultured brain microvascular endothelial cells. SNHG1, functioning as a ceRNA for miR-18a, upregulated the expression of its endogenous target, HIF-1α and promoted BMEC survival through the HIF-1α/VEGF pathway (Zhang L et al., 2018). SNHG1 also targeted the miR-338/HIF-1α axis by sponging miR-338, which resulted in inhibition of BMEC apoptosis and alleviation of OGD-induced cell injury. Silencing SNHG1 and overexpression of miR-338 could aggravate the injury (Yang X and Zi XH, 2019). Next, miR-199a is another target for SNHG1 to promote angiogenesis in cultured BMECs after OGD (Riva P, et al., 2016). It was also shown that SNHG1 overexpression in human cerebral microvascular endothelial cells (hCMEC/D3 cells) could reduce apoptosis and inflammation by directly targeting miR-376a (Lv L et al., 2020). An increased level of SNHG12 was reported in mouse BMECs after OGD and ischemia/reperfusion (I/R). Elevated SNHG12 significantly reduced the inflammatory response and BMEC death, as well as stimulated post-ischemic angiogenesis by directly binding and inhibiting miR-199a, while genetic deletion of SNHG12 abolished these beneficial effects (Long FQ et al., 2018). SNHG12 also functioned as a ceRNA of miR-150 and upregulated VEGF expression to promote angiogenesis in bEnd.3 cells after OGD. Consistently, SNHG12 also increased bEnd.3 cell viability and migration after exposure to OGD (Zhao M et al., 2018).

In addition, lncRNA ANRIL was also increased in rat brain infarct regions and overexpression of ANRIL significantly upregulated VEGF and promoted post-ischemic angiogenesis via the NF-κB pathway (Schmitt AM and Chang HY, 2016). This was consistent with a later report that ANRIL expression was negatively correlated with the National Institutes of Health Stroke Scale (NIHSS) score, and downregulation of circulating lncRNA ANRIL was associated with higher stroke risk and disease severity, along with elevated inflammation in acute ischemic stroke patients (Feng L et al., 2019). CARD8, as one of the downstream target genes, was positively correlated with ANRIL. Single nucleotide polymorphism (SNP) rs2043211 in CARD8 is associated with ischemic stroke. Increased ANRIL may inhibit NF-κB via the elevation of CARD8, and thus alleviate inflammation (Bai Y et al., 2014).

Studies showed the expression of MIAT, which plays a protective role in regulating vascular function, was remarkably increased in vascular endothelial cells in response to hypoxic and oxidative stress..MIAT sponged miR-204-5p as a ceRNA and upregulated the level of high-mobility group box 1 (HMGB1). Inhibition of MIAT in rats with MCAO and cerebral microvascular endothelial cells (CMECs) with OGD promoted angiogenesis and reduced damage via MIAT/miR-204-5p/HMGB1 axis (Deng W, et al., 2020). MIAT might be a promising pharmacological target for treating several neurovascular-related diseases (Jiang X, et al., 2018;Shimizu F et al., 2018). Increased expression of MIAT was observed in peripheral blood leukocytes from patients with ischemic stroke, and overall survival analysis showed that increased MIAT expression was correlated with a relatively poor prognosis. Moreover, MIAT was identified as an independent prognostic marker of death and neurological dysfunction in patients with ischemic stroke by multivariate analysis (Zhu M et al., 2018). Furthermore, recent research has found that lncRNA DANCR expression was significantly increased in BMECs after OGD and promoted post-ischemic cell viability, migration, and angiogenesis through the miR-33a-5p/XBP1s axis, indicating that DANCR might play a protective role in BMECs after ischemic stimuli (Zhang M et al., 2020). LncRNA NEAT1 also promoted survival and angiogenesis in BMECs after OGD via targeting miR-377 and upregulating VEGFA, SIRT1, and BCL-XL, suggesting NEAT1 may be a promising neurorestorative target in cerebral ischemia (Zhou ZW et al., 2019).

5.2. LncRNAs in cerebral vascular pathology after hemorrhagic stroke

Hemorrhagic stroke is characterized by bleeding from the rupture of cerebral vessels, which can be divided into intracerebral hemorrhage (ICH) and subarachnoid hemorrhage (SAH). Hemorrhagic stroke is caused by a complex biological mechanism that includes hypoxia, inflammatory reactions, apoptosis, and autophagy, leading to cerebrovascular and neuronal damage and eventually manifesting as the loss of normal neurological function. Recent evidence suggested that lncRNAs play essential roles in the automatic regulation of cerebral blood flow in hemorrhagic stroke (Fan J et al., 2020). Studies showed that SNHG3 expression was increased in OGD plus hemin-treated primary BMECs, an in vitro model of ICH. Overexpression of SNHG3 inhibited BMEC proliferation and migration, and promoted cell apoptosis and monolayer permeability. Downregulation of SNHG3 siRNA in ICH rats effectively improved neurobehavioral and histological outcomes such as the number of right turns, limb placement scores, integrity of BBB, brain water content, and cell apoptosis. Mechanistically, SNHG3 improvement of cerebral microvascular function in ICH rats is mediated by activating the TWEAK/Fn14/STAT3 pathway (Zhang J et al., 2019). LncRNA fetal-lethal non-coding developmental regulatory RNA (FENDRR) was an endothelial regulator essential for vascular development. Vascular endothelial growth factor A (VEGFA) and miR-126 participated in the physiological process of cerebral vascular endothelial cells. Studies by Dong et al. revealed increased levels of FENDRR and VEGFA, as well as decreased miR-126 in the brain vascular endothelium of the hypertensive intracerebral hemorrhage (HICH) mice and in cultured human brain microvascular endothelial cells (HBMECs) treated with thrombin. Of note, FENDRR overexpression promoted brain endothelial apoptosis via miR-126 regulating VEGFA in HICH mice (Dong B et al., 2018), whereas FENDRR downregulation could relieve HICH injury.

As one of the main causes of non-traumatic SAH, occurrence of intracranial aneurysms (IAs) has a close association with lncRNAs. Researchers found that MALAT1 was upregulated in tissue samples from rats with IAs and may competitively bind to miR-143, thereby promoting VEGFA expression. As a result, vascular endothelial cell apoptosis was induced, leading to decreased vascular endothelial cell viability in IA (Gao G et al., 2020). In addition, the increased MALAT1 level was not only a diagnostic marker in SAH, but also could be used to predict the prognosis of patients with SAH. Cerebral vasospasm (CVS) frequently occurs following SAH and is a major cause of associated mortality and morbidity. It has been reported that ~90% CVS cases exhibited enhanced LncRNAs MALAT1 and LINC01619, being potential biomarkers for prognosis prediction in patients with SAH. High expression of MALAT1 and low expression of LINC01619 indicated poor prognosis (Pan CY et al., 2020). In addition, several SNPs of another lncRNA ANRIL which was also named as CDKN2BAS, including rs1333040, rs10757272, rs10733376 and rs6475606, were reported to be positively correlated with the risk of IA, but more in-depth studies are warranted in the future (Yu T et al., 2020).

5.3. LncRNAs in cerebral vascular pathology after traumatic brain injury

Recent studies have shown altered expression of mRNAs and lncRNAs in TBI. Among them, mRNAs were mainly related to inflammation, immunity, and neuronal and vascular networks, while lncRNAs were mainly related to transcriptional regulation of some mRNAs (Zhong J et al., 2016). MALAT1, a significant lncRNA related to cerebrovascular function, also plays a role in TBI in the regulation of neurovascular regeneration. An RNA-sequencing study revealed the abundance of MALAT1 in the mouse cerebral cortex 24 h after TBI (Zhong J, et al., 2016). Ischemia generally occurred after TBI onset and resulted in hypoxia, which could induce the upregulation of MALAT1 in vascular endothelial cells. Silencing MALAT1 in cultured HUVECs can inhibit cell proliferation and promote angiogenic sprouting as well as cell migration, and genetic deletion of MALAT1 inhibited proliferation and vascularization in vivo (Michalik KM, et al., 2014). Therefore, it is reasonable to infer that MALAT1 can reduce post-traumatic brain damage by promoting angiogenesis. Consistently, treatment with exosomes containing MALAT1 can effectively reduce cortical damage and promote functional recovery after TBI (Patel NA et al., 2018).

5.4. LncRNAs in vascular pathology after spinal cord injury

Similar to TBI, recent studies have also advanced our understanding of the roles and mechanisms of lncRNAs in vascular pathology after SCI (Chandran R, et al., 2017). Ischemia/hypoxia-induced angiogenesis is of great importance in endogenous repairment after chronic compressive spinal cord injury (CCSCI). LncRNA Xist was found to be upregulated in cervical spinal lesions and increased angiogenesis and microvessel density through the miR-32-5p/Notch-1 axis in rats after CCSCI. Overexpression of Xist promoted endogenous neurological recovery in rats after CCSCI, while downregulation of Xist reduced angiogenesis and human umbilical vein endothelial cell sprouting and migration (Cheng X et al., 2020).

6. LncRNA-based therapeutics in CNS diseases

Recent research advancements of the roles and mechanisms of non-coding RNAs in neurologic disorders have uncovered potential novel therapeutic strategies to treat vascular pathologies in CNS injury. In particular, microRNA-based therapeutics have recently gained great attention for clinical application due to better understanding of their functions. MicroRNA-based therapy includes two categories, microRNA inhibitors and microRNA mimics. MicroRNA mimics can be used to restore the activity of down-regulated miRNA under pathological conditions, while microRNA inhibitors or antimiRs inhibit pathologically overexpressed microRNAs (Sun P et al., 2018). The delivery efficiency is attributed to both modifications of RNA and their vehicles. For the purposes of obtaining better efficiency and safety, oligonucleotides based on different chemical modifications have been developed to promote the specificity and stability of microRNA inhibition. These modifications consist of oligodeoxynucleoside phosphorothioate (Campbell JM et al., 1990;van Rooij E, 2011), the addition of 2’-O-methyl (2’-O-methyl) to phosphorothioate nucleotides, locked nucleic acid (LNA) modified antimiRs (Wahlestedt C et al., 2000), 2’-O-methoxyethyl-Oligonucleotides (2’-O-MOE), peptide nucleic acid modified antimiRs (Davis S et al., 2006;Oh SY et al., 2010), fluorine derivatives (2’deoxy-2’-fluoro-RNA)(Pallan PS et al., 2011), or a mixture of the above mentioned modification.

In addition to oligonucleotide modification, delivery vehicles and routes are also important factors for delivery efficiency, which may facilitate RNA uptake. Effective delivery of non-coding RNA-based therapy in CNS injury can be achieved via vehicles such as viral vectors, nanoparticles, exosomes, and stem cells and though various routes of administration including intracerebroventricular (ICV) injection, continuous ICV infusion with osmotic minipumps, intrathecal administration, intravenous (IV) injection, and intranasal administration. Delivery molecules include miRNA mimic, LNA anti-miR, GapmeR, and shRNA among others (Hung J et al., 2018;Zhang J et al., 2020).

There are several clinical investigations regarding the clinical application of microRNA-based therapy, most of which are observational studies focusing on the expression profiles or diagnostic efficacy of microRNAs collected from circulating blood or cerebrospinal fluid. However, only a few interventional clinical trials with microRNAs of interest have been conducted to date. A clinical study, performed by researchers in the Xuanwu Hospital in Beijing, China, aimed at verifying the specific mechanisms of miR-494 to alleviate brain injury after ischemic stroke (https://www.clinicaltrials.gov/ identifier: NCT03577093). Another study performed by Ohio State University in the United States evaluated the differential expression and roles of exosomal microRNAs in patients with a stroke due to acute large vessel occlusion (LVO) as compared to healthy controls. In addition, the investigators also evaluated the differential expression of exosomal microRNA in patients with good versus poor collateral grades (https://www.clinicaltrials.gov/ identifier: NCT03905434). Other studies are also being conducted to identify microRNA markers of subarachnoid hemorrhage (https://www.clinicaltrials.gov/, identifier: NCT03344744, NCT02389634)(Lu G et al., 2017), intraventricular hemorrhage (https://www.clinicaltrials.gov/, identifier: NCT02386228), and traumatic brain injury (https://www.clinicaltrials.gov/, identifier: NCT02639923) (Balakathiresan N et al., 2012). In the future, more in-depth clinical investigations are needed to provide new insights for further development of vascular restorative therapies after CNS injury.

LncRNA-based therapeutics may share similar principle and implementation modes with microRNAs. Through further studies of lncRNAs, manipulation of endogenous lncRNAs by pharmacological and genetic approaches may provide novel insight into the precise treatment of specific CNS injuries. LncRNAs associated with CNS injuries act partly through regulation of the vascular structure and function as described above. For example, some pro-angiogenic lncRNAs may prompt neovascularization remodeling and are beneficial to lesion repair and tissue remodeling in some diseases related to ischemia and hypoxia.

LncRNA-targeting therapy in CNS injuries includes repression and activation of specific lncRNA activities (Ma J et al., 2020). To suppress a lncRNA of interest, several approaches have been developed to achieve therapeutic silencing of lncRNAs by using aptamers or small interfering RNA/short hairpin RNA (siRNA/shRNA), antisense oligonucleotides (ASOs), GapmeRs, transcriptional repression, gene editing tools like CRISPR, and genetic deletion (Zuo X et al., 2020). GapmeRs are a more suitable approach for targeting nuclear lncRNAs (Kok FO and Baker AH, 2019). Our previous studies have documented that silencing of lncRNA MALAT1 by GapmeRs significantly reduced cell proliferation, cell migration, and tube formation in cultured mouse primary skeletal muscle microvascular endothelial cells (SMMECs) after OGD (Zhang X, et al., 2018). Besides, lncRNA MALAT1 silencing by GapmeR increased BMEC death following OGD (Zhang X, et al., 2018). Both genetic and pharmacological strategies are involved in in vivo studies to achieve deletion or inhibition of lncRNA activity. For instance, our group previously reported that genetic deletion of Malat1 in mice worsened neurological outcomes through proapoptotic and proinflammatory pathways after MCAO, indicating Malat1 plays a crucial protective role in ischemic stroke (Zhang X, et al., 2017b). Despite that genetic deletion and genome editing are considered as the gold standard to prevent unpredictable or off-target effects of oligonucleotides, it should be taken into consideration that the locus itself rather than the transcribed RNA may have biological relevance (Jaé N and Dimmeler S, 2020).

On the other hand, overexpression of vascular protective lncRNAs is more complicated than lncRNA silencing. To increase the expression of lncRNAs, viral vectors (lentivirus, adenovirus, adeno associated virus) or non-viral vectors (lipid or polymeric nanoparticles) can be used to deliver lncRNAs to the CNS lesion area (Zuo X, et al., 2020). Although adeno-associated virus (AAV) delivery methods show great promise with enhanced protein coding transcripts and microRNAs (Cao Y et al., 2020;Meng X et al., 2019;Yang Q et al., 2020;Zhang J, et al., 2019), its efficacy on gain-of-lncRNA function in the CNS still needs to be investigated in-depth. Due to the immunogenicity induced by viral approaches, it is anticipated that non-viral vectors will be successfully applied in clinical trials with diverse chemical modifications (Jaé N and Dimmeler S, 2020). Additionally, the CRISPR/Cas9 gene editing approach is emerging as a novel and potent tool to manipulate lncRNA expression through specific gain- or loss-of function manners (Simion V et al., 2019). In combination with a viral-based delivery platform, it holds great promise to achieve precise treatment in CNS injury.

Comparatively, lncRNA intervention is more straightforward and specific than protein intervention since lncRNAs are differentially expressed in specific tissues or diseases in the context of different pathologies. Unlike proteins which usually have multiple regulatory roles, the functions of lncRNAs are relatively specific. As a result, lncRNAs can be targeted for more precise treatment in CNS injury with less side effects. This places an emphasis on targeting lncRNAs in a specific tissue or lesion region precisely and efficiently. Noticeably, many lncRNAs play functional roles in other tissues and physiopathological processes besides the CNS systems. Additionally, lncRNAs are usually not evolutionarily conserved, thus creating a greater gap between experimental animal research and clinical applications. Therefore, it is critical to elucidate important lncRNA-specific targets in different tissues to fully delineate possible side effects (Bao MH, et al., 2018).

Despite cumulative experimental animal studies of lncRNAs in CNS injury, there are no lncRNA-based therapeutic clinical trials in this field. The main challenges and obstacles come from the more complex or diversified functions and mechanisms of lncRNAs compared with other non-coding RNAs, as well as low delivery efficacy to the brain vasculature, difficulties targeting lncRNAs using siRNA/shRNA, and the accuracy of lncRNA delivery (Bhan A et al., 2017;Lepoivre C et al., 2013;Ma SC et al., 2018;Uthaya Kumar DB and Williams A, 2020).

7. Future Perspectives and Conclusions

To date, the molecular mechanisms of CNS injuries such as stroke, TBI, and SCI have not yet been completely elucidated. LncRNAs, an important class of protein non-coding RNA transcripts, play critical regulatory roles in the vascular pathogenesis and progression in CNS injuries and lncRNA-based therapy may hold optimal potential for the treatment of vascular pathologies in neurologic disorders. However, we still have limited understanding of the role and regulatory networks of lncRNAs mediating vascular pathogenesis in CNS injury. Future emphasis should be laid on the precise identification of lncRNA downstream target genes and correlative signaling pathways to fully understand the interactive network and mechanisms of lncRNAs in the regulation of cerebrovascular dysfunction in CNS injury. It is also worth noting that the cross-talks of lncRNAs between the vascular and neural networks in CNS injury still need further in-depth exploration.

At present, lncRNA-based therapeutics is still at the nascent stages of testing in experimental CNS injury models. It is expected that more genetically manipulated animals, especially vascular cell-specific lncRNA transgenic or knockout mice, will be developed for experimental application, thus aiding us in gaining better understanding of lncRNA signaling mechanisms. On the other hand, several clinical studies have recently shown the promising potential of lncRNAs as diagnostic or therapeutic biomarkers in CNS injuries. A clinical trial conducted by Nanjing First Hospital has focused on the differential expression pattern of circular RNA (circRNA), miRNA and lncRNA between patients with acute ischemic stroke and healthy controls. The candidate non-coding RNAs (ncRNAs) would be further validated as biomarkers for the detection and prognosis of acute ischemic stroke and may serve as a novel diagnostic or predictive method (https://www.clinicaltrials.gov/, identifier: NCT04175691). Researchers in Nanjing First Hospital have also investigated the differential expression pattern of ncRNAs in acute ischemic stroke patients before and/or after endovascular treatment, which may provide the evaluation basis for the outcome of endovascular treatment (https://www.clinicaltrials.gov/, identifier: NCT04230785). Another clinical study in National Taiwan University Hospital also proposed to establish a set of new, lncRNA-based diagnostic and prognostic biomarker in stroke patients with cognitive dysfunction or dementia, which may benefit clinical preventive and therapeutic care (https://www.clinicaltrials.gov/, identifier: NCT03152630). In fact, none of therapeutic clinical trials based on lncRNA intervention is being conducted in CNS injuries. With the development of RNA therapeutics, it is not too far-fetched to apply novel lncRNA-based pharmaceutical strategies for the treatment of CNS injuries in the near future.

A better understanding of the functional significance of lncRNAs in the vascular pathophysiology of CNS injuries will pave the way to develop novel pharmaceutical therapeutic strategies that target and regulate lncRNA activity in vascular pathology and intervene against cerebrovascular-related disorders. However, current lncRNA-related discoveries in research are mainly from experimental CNS injury animal models. A greater number of translational and clinical investigations of lncRNA-based diagnostics or therapeutics are expected in the future. Despite the challenges, lncRNA-based therapeutics will become promising strategies for the treatment of CNS injuries.

Highlights.

1. A greater emphasis on the structural and functional changes in the cerebral vasculature following CNS injuries.

2. LncRNAs have emerged as important mediators in the regulation of cerebral vascular homeostasis after CNS injuries.

3. This review provides clear insights into the roles and regulatory mechanisms of lncRNAs in cerebral vascular pathologies after CNS injuries.

4. In-depth investigations on lncRNA-mediated cerebrovascular responses to CNS injuries have opened novel therapeutic strategies for stroke, traumatic brain injury, and spinal cord injury.