Even though there are many challenges in the annotation and analysis of lncRNAs, because of the lack of a clear classification frame, the existing lncRNAs can be divided into several categories based on their function and genomic context.

One way to categorise lncRNAs is according to their function. lncRNAs were reported to be involved in many cellular and molecular processes, like X-chromosome inactivation, imprinting, DNA methylation, transcriptional modulation, and post-transcriptional control (Mercer et al., 2009; Wilusz et al., 2009; Nagano and Fraser, 2011; Geisler and Coller, 2013; Mercer and Mattick, 2013; Zhang and Leung, 2014; Kiang et al., 2015), nuclear-cytoplasmic shuttling, translational inhibition, mRNA degradation, RNA degradation, and regulation of protein activity (Wapinski and Chang, 2011; Yoon et al., 2013; Zimmer-Bensch, 2019). Moreover, many studies have also provided evidence that lncRNAs control gene expression by interaction with DNAs, RNAs, and proteins or chromatin remodelling complexes, and more recent observations suggest that lncRNAs may in fact affect protein-coding directly (Mattick and Gagen, 2001; Khalil et al., 2009; Kiang et al., 2015; Figure 1).

Moreover, with the emergence of advanced biocomputational research tools, a large number of novel lncRNA transcripts have been identified (Amaral et al., 2011; Volders et al., 2013; Yang et al., 2013; Park C. et al., 2014; Kiang et al., 2015). These tools have also helped in predicting lncRNA interaction with various molecules genome-wide. For example, several studies have confirmed lncRNA–EZH2 interaction. EHZ2 is a member of the chromatin-modifying protein PRC2, polycomb repressive complex 2 that plays a role as tumour suppressive/oncogenic regulator. This interaction allowed the lncRNA to act as a guide for the PRC2 complex to the target site (Rinn et al., 2007; Zhao et al., 2008; Jeon and Lee, 2011). This association of lncRNAs with EZH2 is involved in the biology of tumour cells through the up- or downregulation of gene expressions (Bian et al., 2015; Zhang et al., 2015). Moreover, when we look at the downstream molecules of lncRNAs in neurological diseases, we find that they encompass a diverse range of proteins, microRNAs, and other non-coding RNAs, which participate in intricate regulatory networks. Numerous studies have highlighted specific downstream molecules associated with lncRNAs in neurological disorders. For instance, in Alzheimer’s disease (AD), the lncRNA BACE1-AS was found to interact with BACE1, a key enzyme involved in amyloid-β formation and AD pathogenesis (Faghihi et al., 2010). Additionally, the lncRNA HOTAIR was shown to modulate the expression of HOX genes by interacting with PRC2 in glioblastoma (Kciuk et al., 2023; Xin et al., 2023).

The functional diversity of lncRNAs is based on the inherent properties of RNA molecules, such as their modular organisation, ability to fold into different structures, and having many functional domains in their sequence that allow them to interact with different molecules (Wang and Chang, 2011; Zimmer-Bensch, 2019). Furthermore, compared to protein-coding genes, lncRNAs are highly tissue-specific and are usually co-expressed with neighbouring coding genes (Cabili et al., 2011). Diverse expression patterns of lncRNAs have major implications on their regulatory roles. In 2008, Mercer et al. identified hundreds of lncRNAs that are expressed in the brain by the in situ hybridisation method. The expression of these lncRNAs differs according to their anatomical location, cell type, and subcellular location (Mercer et al., 2008). The lncRNA Evf2 is expressed in the ventral forebrain and was shown in an in vivo knockout study to regulate the development of GABAergic neurons. Other examples are the lncRNAs MALAT1 and Neat1. They are both localised in the nucleus and play a role in regulating alternative splicing of pre-mRNA by modulating serine/arginine splicing factor phosphorylation (Tripathi et al., 2010). In addition, the muscle-specific long non-coding RNA, linc-MD1, is expressed in the cytoplasm of myoblasts and plays an important role in muscle differentiation (Cesana et al., 2011). In the next section of the review, we will elaborate more on the known mechanisms of action and functional roles of lncRNAs.

Another way to categorise lncRNAs is according to their genomic location, as in, from where in the genome they are being transcribed. These lncRNAs can be classified into five different groups: Stand-alone lncRNAs, which are transcribed from a specific sequence that does not overlap with a protein-coding gene; antisense transcripts, transcribed opposite to the sense DNA sequence; Pseudogenes, transcribed from genes that lost their coding potential due to a mutation; Intronic, transcribed from an intron sequence; Intergenic, lncRNAs that are promoter-associated and enhancer associated transcripts (Mercer et al., 2009; Kung et al., 2013; Figure 2). It is important to mention that this genomic context categorisation does not provide any information about their function or conservation. In addition, studies showed that the majority of lncRNAs are actually localised in the cytoplasm, instead of the nucleus and associated with ribosomes, where they may help in the evolution of new protein subtypes (Ruiz-Orera et al., 2014; van Heesch et al., 2014).

Even though we do not understand the full functions of lncRNAs, many studies have shown that they play various roles in almost every aspect of biological regulations, from chromatin structure to protein level (Wilusz et al., 2009; Wu et al., 2013). Here, we summarise lncRNAs’ broad mechanisms in regulating gene expression, including chromatin modification, transcription, and post-transcription regulations (Figure 3).

Many lncRNAs that were discovered play a role in mRNA stability in the cytoplasm. The lncRNA BACE1-AS, antisense transcript for β-secretase 1, promotes mRNA stability (Faghihi et al., 2008; Kretz, 2013). While the lncRNA GADD7, growth-arrested DNA-damage inducible gene 7, decreases the stability of mRNAs (Gong and Maquat, 2011; Liu et al., 2012).

Even though lncRNAs are extremely abundant, they were initially classified as transcriptional noise or “junk” DNA. The low sequence conservation and low transcription potential of lncRNAs fuelled the debate about their function and suggested low evolution pressure and biological importance (Mercer et al., 2009; Ponting et al., 2009; Ponting and Belgard, 2010). One study argued that lncRNAs are transcriptional noise and showed that most of the lncRNA transcripts discovered in their sequencing analysis are associated with known genes (van Bakel et al., 2010; Roberts et al., 2014). However, this hypothesis was opposed by other studies that suggested this association between lncRNAs and coding protein loci is consistent with persistent transcription and suggests inadequate sequencing depth in the opposing study (Clark et al., 2011). Moreover, a different finding reported by the GENCODE consortium suggests that most lncRNAs are actually expressed as independent transcripts (Derrien et al., 2012). Nevertheless, many research groups studied lncRNAs and confirmed the wide-ranging functional roles of lncRNAs.

First, lncRNAs are expressed in a tissue-specific manner. Analysing deep sequencing data to understand the transcriptional landscape of different cell lines shows that 29% of lncRNAs were expressed in a cell-specific manner, while only 10% were expressed in all cell types, in contrast to protein-coding genes for which the expression percent were 7 and 53%, respectively (Djebali et al., 2012). Interestingly, 40% of all lncRNAs are expressed in the brain (Derrien et al., 2012).

Moreover, genome-wide expression profiling of lncRNAs in the brain has been carried out using different methods. The Allen Brain Atlas (ABA) is a large study that maps the expression of genes in the developing and adult mouse brain at a genome-wide scale. The ABA used RNA in situ hybridisation (ISH) to visualise the expression of 849 ncRNAs that are expressed in the adult brain and found that most lncRNAs are associated with specific neuroanatomical loci. For example, the lncRNA AK037594 was found to be expressed only in the dentate gyrus and CA1–3 regions of the hippocampus. While Gomafu is expressed only in differentiating neural progenitors and a subset of postmitotic neurons (Sone et al., 2007).

Interestingly, most of the lncRNAs that were identified to be highly expressed in the brain are found to be binding to TFs and involved in the transcription regulation of nearby protein-coding genes that are involved in brain development (Ponjavic et al., 2009; Augoff et al., 2012), which suggests that they have explicit biological role. For example, the ABA shows expression of the lncRNA Evf2, which interacts in trans with the TF Dlx-2 to regulate the expression of the Dlx-6 gene (Feng et al., 2006). Consistent with this function, Evf2 shows a coincident expression profile with Dlx-2, which supports its role in neuronal differentiation (Feng et al., 2006).

Second, lncRNAs showed signs of regulated expression (Ravasi et al., 2006). For example, 174 lncRNAs were differentially expressed during the in vitro 16 days of differentiation of mouse embryonic stem (ES) cells into embryoid bodies (Dinger et al., 2008). Another study showed different expression patterns of four lncRNAs after treatment with retinoic acid to induce neuronal differentiation in mouse ES cells (Sheik Mohamed et al., 2010). Moreover, pluripotency factors, such as Oct4 and Nanog, bind to the promoters of some lncRNA and control their transcription. This suggests that they play a role in the balance between maintenance of pluripotency and lineage commitment. Knockdown and overexpression of these lncRNAs affected the expression of Nanog and Oct4 and promoted lineage-specific differentiation (Sheik Mohamed et al., 2010).

Third, unlike protein-coding genes, which are highly conserved and must preserve their ORF, lncRNAs can only have shorter stretches of sequence that are conserved to sustain their functional domain and structure (Pang et al., 2006). For example, the lncRNA Xist is known for its function in X-chromosome inactivation, but only a short sequence of its length is conserved, and despite this, PRC2 targeting region in Xist is defined in different species. Functionality and high conservation of the sequence may not always correspond to each other (Hendrich et al., 1993; Nesterova et al., 2001; Zhao et al., 2008).

Scientists have used recent high-throughput technologies like sequencing, microarray expression, and high-throughput RNA in situ hybridisation method (Allen Brain Atlas (ABA) study) to visualise the expression of over 20,000 transcripts at cellular resolution (Lein et al., 2007). They observed thousands of lncRNAs expressed in the brain during neural development. Utilising these data for more analysis, we found that 849 lncRNAs examined out of 1,328 exhibit specific expression profiles in distinct neuroanatomical regions, cell subtypes, and subcellular compartments in different adult mouse brain regions (Mercer et al., 2008; Ng et al., 2013b; Shi et al., 2017). In the mice model, lncRNAs showed a different expression pattern across the cortical layers in addition to being specifically expressed in the dentate gyrus, subventricular zone, and olfactory bulb regions of the brain (Belgard et al., 2011; Ramos et al., 2013). In human neocortical brain sections, lncRNAs were shown to be expressed in an age-dependent pattern from infancy to adulthood (Lipovich et al., 2014). This was confirmed in a study done in vivo in mouse of evolutionarily conserved intergenic lncRNAs, in which scientists identified “brain clusters” of lncRNAs (Guttman et al., 2009) that are differentially expressed during development (Mercer et al., 2010). Moreover, in a functional study by Sauvageau et al. (2013) they showed that lncRNAs, in particular BRN1B, are important for organ and brain development using several lncRNA in vivo mice knockout models (Sauvageau et al., 2013). While HAR1, one of the most evolutionary acceleration regions in the human genome, was discovered to belong to a lncRNA gene, HAR1F (HAR1A), which was shown to be specifically expressed in Cajal–Retzius neurons in the developing human neocortex during the period for cortical neuron specification and migration (Pollard et al., 2006). Another example is the lncRNAs Evx1as and Hox5b/6as, which were shown to be specifically associated with genes from the Dlx family that are known to be involved in brain development in mammals and Drosophila (Dinger et al., 2008). In addition, embryonic ventral forebrain-2 (Evf2) is transcribed antisense to the Dlx-6 gene, which is a homeobox-containing TF important in forebrain neurogenesis (Stenman et al., 2003; Feng et al., 2006).

Interestingly, a breakthrough study in in vivo samples obtained surgically from human neocortical tissue by Lipovich et al. (2014) identified lncRNA expression in different ages and found eight lncRNAs with strong statistical associations with ageing and, by so, brain development. Most of these lncRNAs were transcribed antisense from neighbouring protein-coding genes that are known to play a role in neural development. This implies that lncRNAs and protein-coding genes interact and play functional regulatory mechanisms in neural development (Mercer et al., 2010). Importantly, these lncRNAs contain specific anthropoid exons and mRNA processing regions that reside within primate-specific sequences, which confirms their recent evolutionary origins (Lipovich et al., 2014). Furthermore, gene expression analysis studies of the mouse retina at different time points discovered many evolutionary conserved lncRNAs that are expressed in the developing retinal cells, which further confirms that lncRNAs play a functional role in neuron development (Blackshaw et al., 2004). Altogether, these findings suggest the involvement of lncRNAs in the development of the human brain.

Neural stem cells (NSCs) have a significant ability for self-renewal and proliferation, which are important for CNS homeostasis (Hosseinkhani et al., 2013; Zhao et al., 2020). Although the underling regulatory mechanism is still not clear, recent studies confirmed that lncRNA functions as a modulator in NSC self-renewal and proliferation (Zhao et al., 2020). For example, overexpression of TRIM71 interacting long non-coding RNA 1, Trincr, regulated kinase (ERK) signalling pathway, which is essential for cell self-renewal, restrains fibroblast growth factors (FGF) and therefore suppresses NPC self-renewal (Lanner and Rossant, 2010; Li Y. P. et al., 2019). In addition, the long intergenic non-protein-coding RNA regulator of reprogramming, lincRNA-ROR, was shown to enhance the reprogramming of induced pluripotent stem cells (iPSCs) and regulate the maintenance of embryonic stem cell (ESC) pluripotency through forming a regulatory feedback loop with miR-145, Oct4, Sox2, and Nanog (Loewer et al., 2010; Wang et al., 2013). Similar to this is the lncRNA MIAT, myocardial infarction-associated transcript, which functions as a co-activator of Oct4. Loss of MIAT inhibits the expression of Oct4, Sox2, and Klf4 and reduces ESC proliferation (Sheik Mohamed et al., 2010). An interesting example is the Sox2 Overlapping Transcript, Sox2OT, a highly conserved lncRNA that overlaps the Sox2 gene (Fantes et al., 2003). Sox2 is a TF that is important in maintaining self-renewal of neural stem cells (Mizuseki et al., 1998). Sox2OT was shown to be expressed in neural stem cells and is downregulated during differentiation (Amaral et al., 2009).

Furthermore, glioblastoma stem-like cells (GSCs) exhibit the stemness properties of stem cells, like the ability to self-renew and multipotency (Cheng et al., 2013), which were shown to be regulated by lncRNAs. The lncRNA ZNF281 is a newly identified lncRNA that participates in controlling self-renewal and proliferation in GSCs through targeting NF-κB1 signalling pathway (Katsushima et al., 2016; Li X. T. et al., 2019; Tang et al., 2019). Moreover, the lncRNA TALNEC2, tumour-associated lncRNA expressed in chromosome 2, and linc01198 were also found to regulate self-renewal and pluripotency of GSCs (Chen W. L. et al., 2019; Brodie et al., 2021). Thus, lncRNA may become novel potential therapeutic target for glioblastoma therapy.

lncRNAs also play roles in neural cell fate determination, neuronal-glia fate differentiation, and oligodendrocyte expansion. In validation of the above studies, identified lncRNAs using microarray expression analysis in mouse cell lines. They discovered that lncRNAs are differentially expressed between embryonic forebrain-derived neural stem cells (NSCs), GABAergic neuron/oligodendrocyte cells, and the different stages of differentiated neurons and glia. For example, the lncRNA Malat1, metastasis-associated lung adenocarcinoma transcript 1, was downregulated in precursor cells but upregulated in differentiated neuronal and glial cells. When precursor cells were treated with histone deacetylase (HDAC) inhibitor, which is known to suppress the maturation of oligodendrocyte precursor cells and convert them to neuronal pattern, the expression of lncRNAs was also affected, suggesting that they are being regulated by HDAC (Mercer et al., 2010). Another study observed the function of the lncRNA Nkx2.2AS, a natural antisense transcript overlapping the TF gene Nkx2.2, which was shown to regulate oligodendrocyte differentiation. Overexpression of Nkx2.2AS induced oligodendrocyte differentiation and increased Nkx2.2 mRNA expression (Tochitani and Hayashizaki, 2008). In addition, the lncRNA Six3OS, which is transcribed from the opposite strand of gene encoding the homeodomain TF Six3. Six3OS controls Six3 function by acting as a molecular scaffold to recruit histone modification enzymes to the Six3 gene, which results in regulating retinal cell specification (Rapicavoli et al., 2011). Another example, Dlx1AS lncRNA, the antisense transcript of the distal-less homeobox 1 (Dlx-1), was shown to be upregulated during GABAergic differentiation and downregulated during oligodendrocyte differentiation (Mercer et al., 2010). Dlx1AS functions in neuronal differentiation by regulating expression of neighbouring homeobox genes (Ramos et al., 2013). Moreover, the lncRNA Evf2, which regulates GABAergic interneurons formation in the developing mouse and human brain through controlling the expression of Dlx-5, Dlx-6, and glutamate decarboxylase 1 (Gad1), an enzyme responsible for catalysing glutamate to form GABA (Bond et al., 2009; Berghoff et al., 2013; Cajigas et al., 2018). Interestingly, RNA-seq analysis of the expression of lncRNAs in human neurons derived from iPSC found that these lncRNAs dramatically changed during the transition from iPSC to differentiated neurons. Like the lncRNA HOTAIRM1, which is known to be a regulator of several HOXA genes during myelopoiesis, it was shown to be upregulated in differentiated neurons (Lin et al., 2011).

Unlike the above-mentioned examples of lncRNAs that are highly expressed in the brain and known to promote neuronal differentiation, some other lncRNAs were shown to be downregulated in the brain and repress neuronal differentiation. For example, the lncRNA Pnky is known for its role in inhibition of neuronal development, and its expression was decreased during V-SVZ NSC differentiation into neuronal cells. It forms a complex with splicing factor and RNA-binding protein (RBP)-polypyrimidine tract-binding protein (PTBP1) and functions in controlling NSC differentiation in neurons through alternative splicing. Knockdown of Pnky or PTBP1 promoted neurogenesis in cultured postnatal V-SVZ NSC differentiation processes in mature neurons (Ramos et al., 2015; Grammatikakis and Gorospe, 2016). Moreover, the lncRNA lncR492 functions as an inhibitor of neuronal differentiation by binding with mRNA-binding protein HuR and activation of Wnt signalling (Winzi et al., 2018). Additionally, the lncRNA BDNF-AS, brain-derived neurotrophic factor antisense, was shown to control neuronal differentiation. Its overexpression suppressed neurite growth in ketamine-treated mouse embryonic NSC-derived neurons through activation of the potassium uptake system protein (TrkB) signalling pathway (Zheng et al., 2016).

LncRNAs can control neural development by controlling the expression of proximal protein-coding genes. For example, Sox1 and Sox2 are TFs known to maintain the stemness of NSCs. Recently, the lncRNA Sox1ot, Sox1 overlapping transcript, and lncRNA Sox2ot, Sox2 overlapping transcript, were discovered. They are evolutionarily conserved lncRNAs that are highly expressed during neural development and overlap with Sox1 and Sox2 expression, respectively (Kan et al., 2004; Askarian-Amiri et al., 2014; Ahmad et al., 2017; Knauss et al., 2018). The lncRNA Sox2ot suppressed NSC proliferation and neuronal differentiation by associating with the transcriptional factor YY1, which binds to CpG island in the Sox2 locus to suppress Sox2 expression (Knauss et al., 2018).

Another example is the lncRNA RMST, rhabdomyosarcoma 2-associated transcript, which has been shown to be important in neurogenesis. RMST allowed Sox2 binding in the promoter of target genes, and its knockdown led to differential expression of almost 1,000 genes important in neurogenesis (Ng et al., 2013a). Moreover, the lncRNA Kdm2b binds to the Kdm2b gene and increases its expression, which promotes neuronal differentiation in cortical projection neurons (Li et al., 2020). Similarly, the lncRNA Paupar is a CNS-expressed and chromatin-associated lncRNA that is transcribed upstream of the gene encoding the Pax6 TF. It plays a role in neural differentiation by forming RNA multiprotein complex with Pax6 and KAP1 TFs (Vance et al., 2014; Pavlaki et al., 2018). Furthermore, in mammalian eye development, the retina consists of cell-specific subtype neuron layers connected by synapses (Ng et al., 2013b). Six3OS is the long non-coding opposite strand transcript (lncOST) of the homeodomain factor Six3. It controls eye development and postnatal retinal cell specification through its function as a molecular scaffold to regulate Six3 activity (Rapicavoli et al., 2011).

LncRNAs also bind to TFs to regulate neurogenesis, like the TFs SUZ12 (a component of the polycomb repressive complex 2, PRC2), REST, and SOX2 (a pluripotency-associated TFs), in which they act as guides for these proteins. REST is a TF that is known to regulate pluripotency and control neurogenesis. It represses the expression of genes involved in neurogenesis in non-cells (Chong et al., 1995). A study by Johnson et al. discovered two lncRNAs that are regulated by REST in NSCs (Johnson et al., 2009). Similarly, the transcription co-factor CoREST, which also functions in repressing neural genes (Andres et al., 1999). CoREST was identified to associate with 63 lncRNAs in RIP-chip assay; most of them also bind to PRC2, implying that these lnRNAs may also function in regulating neural cell differentiation (Khalil et al., 2009). In addition, Dlx genes encode homeodomain proteins that are known for their function in controlling neuronal differentiation and migration (Anderson et al., 1997a,b). As mentioned above, the lncRNA Evf2 is transcribed from the Dlx locus and interacts with Dlx-4 protein to increase its transcriptional activation functionality in NSCs (Feng et al., 2006). Evf2 knockout mice model showed an abnormal gene expression pattern that led to a decrease in the number of GABAergic interneurons in the mouse hippocampus (Bond et al., 2009).

miRNAs are short non-coding RNAs (approximately 22 nucleotides in length) that are expressed abundantly during brain development and are known to suppress the translation of coding genes in all stages of neural differentiation (Shi et al., 2010). LncRNAs were shown to function as competing endogenous RNAs (ceRNA) and control miRNA to regulate genes important for neural development (Tay et al., 2014; Weng et al., 2018). For example, knockdown of LncRNA 1,604 suppressed neural differentiation by regulating miR-200c to control key TFs zinc finger E-box binding homeobox 1/2 (ZEB1/2) (Weng et al., 2018).

Moreover, lncRNA could be processed to generate several variants that play roles in neurogenesis. The lncRNA C130071C03 Riken variants, Rik-201 and Rik-203, are activated by neurogenesis TF CCAAT/enhancer-binding protein β (C/EBPβ) and therefore modulate brain development. Knockdown of Rik-201 and Rik-203 suppressed the expression of neural differentiation-related gene Sox6, and therefore repressed neural differentiation by function as ceRNAs of miR-96 and miR-467a-3p, respectively (Zhang et al., 2019). Additionally, miR-128-3p is highly expressed in the brain and controls neural differentiation. Overexpression of miR-128-3p suppressed neurons but enhanced gliocyte differentiation. In addition, the LncRNA MEG3 is elevated by the cAMP/response element-binding protein (CREB) pathway. Thus, it enhances neuronal differentiation by acting as a negative regulator of miR-128-3p (Gao et al., 2019).

LncRNAs could also contribute to neural differentiation by being affected by signalling pathway. Neurite outgrowth occurs in the early neuronal differentiation stage. The lncRNA metastasis-associated lung adenocarcinoma transcript 1 (Malat1) was found to play a crucial role in neurite growth. Knockdown of Malat1 prevents neurite outgrowth and advanced cell death by suppressing the mitogen-activated protein kinase (MAPK) signalling pathway and stimulating the peroxisome proliferator-activated receptor (PPAR) and p53 signalling pathways (Chen et al., 2016). Another example is lncRNA-AK053922, which has been shown to control neural cell type specification through regulating sonic hedgehog (Shh) signalling (Hashimoto-Torii et al., 2003; Meyer and Roelink, 2003; Wu et al., 2013).

Synaptogenesis is a critical process during neuronal development, which is altered in many neurodevelopmental disorders (Zoghbi, 2003; Ecker et al., 2013). LncRNAs have been shown to play important and direct roles in regulating genes involved in synaptic plasticity, cognitive function, and memory. GABAergic inhibitory interneurons in the hippocampus are responsible for learning in the embryonic and adult brains. Studies showed that the lncRNA Evf2, which is transcribed from the Dlx-5/6 ultraconserved region, is important for the development of GABAergic neurons. It interacts with the transcription co-activator Dlx-2 to regulate the transcriptional activity of Dlx-5/6 and glutamate decarboxylase 1 (Gad1, necessary for the conversion of glutamate to GABA) (Feng et al., 2006), and then controls the expression of genes that regulate GABAergic interneurons in the developing mouse brain. Knockdown of Evf2 causes abnormal formation of GABAergic circuitry in the hippocampus and dentate gyrus, which affects synaptic activity in mice (Bond et al., 2009). Moreover, gene ontology analysis revealed that genes affected by the lncRNA Malat1 were mostly associated with synaptic function and dendrite development. Knockdown of Malat1 in vitro in primary hippocampal neurons decreased synaptic density and changed the expression of Nlgn1 and SynCAM1 genes that are known to regulate synaptogenesis (Bernard et al., 2010).

LncRNAs also function to support long-term changes in synaptic strength. Nitric oxide (NO) is a signalling molecule that functions as a neurotransmitter and thus is important in learning, long-term potentiation (LTP), and long-term depression (LTD) (Muller, 1996). Nitric oxide synthases (NOSs) are enzymes that function to catalyse the production of NO from l-arginine. Interesting research done in Lymnaea stagnalis snail discovered that an antisense RNA is transcribed from the NOS pseudogene and complements the NOS mRNA. Reduction of NOS pseudo-gene antisense transcript causes upregulation of NOS mRNA levels transiently, and the timing overlapped with the window for memory formation, which suggests that the antisense NOS pseudogene transcripts associate with memory formation by controlling the expression of NOS mRNAs (Korneev et al., 1999; Kemenes et al., 2002).

The rodent-specific BC1 and the non-homologous primate-specific BC200 lncRNAs function to control protein synthesis in postsynaptic dendritic microdomains; therefore, they play important roles in maintenance of synaptic plasticity (Muddashetty et al., 2002). Neurogranin (Nrgn) and calcium/calmodulin-dependent protein kinase II inhibitor 1 (Camk2n1, CaMKIINalpha) are proteins that are expressed in rodents’ brains and have been shown to regulate synaptic long-term potentiation by controlling Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) (Gerendasy and Sutcliffe, 1997; Kennedy, 1998; Lisman et al., 2002). It has been found that transcripts that encode the sense and antisense of the gene locus of these two proteins control their post-transcriptional expression levels during cerebral corticogenesis and synapse function (Ling et al., 2011).

Brain-derived neurotrophic factor (BDNF) is a growth factor that is important for supporting neuronal growth, survival, and synaptic plasticity and is implicated in learning and memory formation (Kang and Schuman, 1995; Figurov et al., 1996; Yamada et al., 2002). The lncRNA BDNF-AS, which is the transcribed antisense of the BDNF gene, interacts with BDNF mRNA in the brain. Silencing of BDNF-AS caused increased BDNF mRNA and protein levels, which resulted in neurite outgrowth and maturation, suggesting BDNF-AS controls the function of the BDNF gene (Lipovich et al., 2012; Modarresi et al., 2012). Moreover, BDNF-AS was also shown to interact with and recruit EZH2 subunit of PRC2 to the BDNF gene promoter and inhibit its expression (Pruunsild et al., 2007; Modarresi et al., 2012). All the above studies confirm that lncRNAs play important functions in synaptic plasticity and cognitive and memory processes on transcriptional and post-transcriptional levels.

Additionally, it acts as a critical determinant in normal brain development and neurogenesis. Recent evidence has confirmed lncRNAs as key regulatory molecules in many neurodevelopmental and neurodegenerative disorders, such as schizophrenia (Scholz et al., 2010; Li et al., 2018), autism spectrum disorder (ASD) (Wang et al., 2015), Alzheimer’s (Faghihi et al., 2010), Huntington’s (Sunwoo et al., 2017), and Parkinson’s (Ni et al., 2017) diseases. We have summarised functional lncRNAs involved in neurological disorders in Table 2.

A glioma is a type of tumour that arises from glial cells in the brain. It is identified by uncontrolled cell growth, necrosis, and dynamic angiogenesis, with symptoms including headache, seizures, impaired neurological function, and, eventually, death (Ferris et al., 2017; Srinivas et al., 2023). Current non-invasive techniques (computed tomography [CT], magnetic resonance imaging [MRI], or positron emission tomography [PET] scans) can help in identification and localisation of these tumours, yet we cannot characterise their pathology by these techniques alone. High-grade gliomas can infiltrate into the extracellular matrix of the brain, which also makes it hard to do surgery and radiotherapy on them (Gwak et al., 2012; Park J. Y. et al., 2014). Hence, identifying the molecular mechanisms and key regulators underlying gliomagenesis is important for the cure of this disease.

Multiple studies have identified some differently expressed lncRNAs that contribute to the pathogenesis of glioblastoma multiform (Ellis et al., 2012). For example, the lncRNA MEG3, a maternally expressed gene 3, has been shown to be downregulated in glioma brain tissues compared to normal ones (Wang et al., 2012). Overexpression of the lncRNA MEG3 in vitro in human glioma cell lines negatively affects cell proliferation and promotes p53-mediated apoptosis. Another example is the lincRNA H19 and its derivative miR-675, which was shown to play a role in glioma cell invasion (Shi et al., 2014). Furthermore, lncRNAs such as FOXD2-AS1, forkhead box D2 adjacent opposite strand RNA 1, HOTTIP, homeobox A (HOXA) distal transcript antisense RNA, and HOTAIR, HOX anti-sense intergenic RNA, have also been identified as regulators of glioma progression as they play roles in cell cycle and epigenetic modifications (Pandey et al., 2014; Vance et al., 2014; Latowska et al., 2020; Chen et al., 2021). All these lncRNAs were confirmed to regulate glioma development and can be considered as potential drug targets in glioma treatment.

lncrnas are associated with various neurological disorders, tumours, and psychiatric conditions, which suggest that they could be used for diagnostic purposes as biomarkers precise to a specific disease and can be sensitive and accurate to early and rapid detection.

Many of the lncRNAs mentioned in this review can be used as biomarkers using minimally invasive methods. For example, in vivo analysis of plasma-derived circulating RNAs has confirmed the lncRNA BACE1-AS as a diagnostic marker of AD (Fotuhi et al., 2019; Srinivas et al., 2023). Moreover, identifying lncRNA as biomarkers with high specificity and sensitivity from the bloodstream at sites distal to the brain has also been done in other neurological disorders, and some of them were confirmed by imaging and tissue biopsies (Srinivas et al., 2023). However, although identifying lncRNA expression in the bloodstream by isolation of serum, plasma, leukocytes, or exosomes is helpful and informative, it may not give a full representation of the molecular changes accruing in the brain. Therefore, samples from the CSF, which are in contact with the brain, have been used to identify circulating lncRNA expression, and it has been considered as a more specific biomarker for diagnosing certain brain pathologies (Hossein-Nezhad et al., 2016; Pan et al., 2020; Whitlock et al., 2022). For example, the lncRNA MALAT1 was found to be elevated in the CSF of patients with AD compared to healthy controls, suggesting its potential as a diagnostic marker (Zhuang et al., 2020). Additionally, the long non-coding RNA activated by TGF-β (lncRNA-ATB) expression was shown to be significantly upregulated in the CSF of AD (Wang et al., 2018).

However, it is important to emphasise that even though the above-mentioned methods are helpful and promising, relying on circulating lncRNAs as biomarkers for brain disorders is difficult. qRT-PCR is used to identify the expression of lncRNAs in the bloodstream, but there is no known reference for lncRNAs from different sources (e.g., plasma vs. serum vs. CSF). In addition, most lncRNAs are expressed at low levels, making it difficult to detect them. Importantly, lncRNAs might not be specifically dysregulated in one disease and could be expressed in more than one. Like the lncRNA NEAT1, which was found to be aberrantly expressed in PD, AD, and ALS (An et al., 2018). Because of that, liquid biopsies can be used in addition to existing diagnostic methods rather than as a defective diagnostic alone. Interestingly, the expression of some lncRNAs that cross the blood–brain barrier (BBB) and become enriched in cerebrospinal fluid (CSF) is high in gliomas ex vivo, making them possible biomarkers for brain tumours (Xu et al., 2021).

In addition to their use as biomarkers, lncRNAs have also been studied as potential in vivo therapeutic targets for the treatment of neurological disorders (Wahlestedt, 2013; Roovers et al., 2018). Many lncRNA-based therapies have been discovered to target lncRNA transcripts for degradation or interreference, and some pharmaceutical companies are also keenly developing lncRNA-targeting therapeutics (Li and Chen, 2013; Ling et al., 2013; Park J. Y. et al., 2014). So far, there are 11 RNA-based therapies that have been approved by the U.S. Food and Drug Administration (FDA) and/or the European Medicines Agency (EMA). The only one that targets the brain is Nusinersen, which is used to treat spinal muscular atrophy, while the others target other tissue types (Winkle et al., 2021; Figure 4).