REST in the Road Map of Brain Development

Abstract

Repressor element-1 silencing transcription factor (REST) or also known as neuron-restrictive silencing factor (NRSF), is the key initiator of epigenetic neuronal gene-expression modification. Identification of a massive number of REST-targeted genes in the brain signifies its broad involvement in maintaining the functionality of the nervous system. Additionally, REST plays a crucial role in conferring neuroprotection to the neurons against various stressors or insults during injuries. At the cellular level, nuclear localisation of REST is a key determinant for the functional transcriptional regulation of REST towards its target genes. Emerging studies reveal the implication of REST nuclear mislocalisation or dysregulation in several neurological diseases. The expression of REST varies depending on different types of neurological disorders, which has created challenges in the discovery of REST-targeted interventions. Hence, this review presents a comprehensive summary on the physiological roles of REST throughout brain development and its implications in neurodegenerative and neurodevelopmental disorders, brain tumours and cerebrovascular diseases. This review offers valuable insights to the development of potential therapeutic approaches targeting REST to improve pathologies in the brain.

Graphical Abstract

The important roles of REST as a key player in the nervous system development, and its implications in several neurological diseases.

Introduction

In brain development, the neural progenitor cells must undergo several essential stages to form the fully matured functioning neurons (Alberts et al. 2002). At the same time, it is important that there is a neuroprotective mechanism capable of protecting the neurons from various stressors. A zinc finger repressor, known as the repressor element-1 silencing transcription factor (REST) or neuron-restrictive silencing factor (NRSF), plays an inevitable role throughout this process.

Discovered in the mid-1990s by two independent research groups, REST is a multizinc finger transcription factor that was first known as the master regulator of neurogenesis (Chong et al. 1995; Schoenherr & Anderson 1995). REST is found abundantly in the non-neural cells and the immature neuronal cells but not in the differentiated neurons. Its downregulation is also necessary for proper neuronal differentiation (Chong et al. 1995; Mori et al. 1992; Schoenherr & Anderson 1995). Hence, it is not surprising that REST is postulated to play a role in neuronal gene repression. Nonetheless, a minimal expression of REST is still required during the neuronal differentiation process to ensure a high neuronal viability (Covey et al. 2012; Nechiporuk et al. 2016; Wang et al. 2021). Previous studies of forced REST expression in differentiating neurons have caused higher tendency of having axon guidance error in the differentiating neurons (Paquette et al. 2000) and a significant reduction in neuronal formation (Ravanpay et al. 2010). In regard to this, the regulation of REST expression during the neuronal differentiation process is a key factor to achieving the optimal balance between the neural progenitor proliferation and differentiation processes. This review paper will further discuss the cellular and molecular mechanism of REST expression that intricates the balance of the brain cell population.

The neuroprotective potential of REST was highlighted in a hallmark paper by Lu et al.. Their discoveries have achieved a new milestone in understanding how REST functions primarily in the ageing brain. In brief, the expression of REST in hydrogen peroxide- and oligomeric amyloid-beta (Aβ)-challenged neurons has increased the neuronal resilience against oxidative stress and oligomeric Aβ, whereas neurons without REST show significant elevated neurodegeneration and cell death (Lu et al. 2014). The significance of nuclear localisation of REST to confer neuroprotection is also emphasised in the same study. Subsequently, emerging studies presented evidence on how REST dysregulation and impaired REST nuclear translocation contribute to the pathogenesis of various neurodevelopmental and neurodegenerative disorders (Alsaqati et al. 2022; Katayama et al. 2016; Kawamura et al. 2019; Lim et al. 2023; Meyer et al. 2019). These studies also investigated the potential therapeutic approaches targeting REST, in which all of these will be further discussed in this review paper.

Most of the recent review articles are focusing on the mechanism of REST action in certain aspects of the brain function or specific stages in brain development. There is limited attention dedicated to the negative impacts of impaired nuclear localisation of REST seen in neurodevelopmental and neurodegenerative diseases. We provide a comprehensive summary and highlights recent updates on the essential role of REST throughout the brain development, from embryogenesis, neurogenesis to the ageing brain covering physiological roles of REST throughout the brain development and its implications in neurodegenerative and neurodevelopmental disorders, brain tumours and cerebrovascular diseases. Hence, this review offers valuable insights to the development of potential therapeutic approaches targeting REST to improve pathologies in the brain.

Physiological Aspects of REST in the Brain Development

REST as an Important Neuronal Gene Regulator

REST is well known for its transcriptional inhibitory function. Its expression and function are highly dependent on the cell type, subcellular localisation, stimulus and developmental stage of the cell. To fully understand how this transcription factor regulates global gene expression, it is necessary to identify and comprehend the functionality of all its target genes. The genome-wide analysis identified 1892 human and 1894 mouse genes containing REST-binding site that contains a 21-base pair consensus motif (TTCAGCACCatGGACAGcgcC) (Bruce et al. 2009). Surprisingly, among these genes, at least 40% of them were identified to be involved in the nervous system (Bruce et al. 2004). The wide spectrum of genes targeted by REST inhibition encodes proteins essential for neuronal functions, strengthening the role of REST as a neuronal gene-expression regulator. For instance, REST targets the synaptic vesicle proteins [synaptophysin (Lietz et al. 2003) and synapsin I (Li et al. 1993)], neurotransmitter receptors [glutamate ionotropic receptor AMPA type subunit 2 (Calderone et al. 2003) and D3 dopamine (Bruce et al. 2004)] and voltage-gated ion channels [NaCh II (Palm et al. 1998)]. Sun and colleagues further identified 89 REST-targeted genes through the chromatin immunoprecipitation-based cloning strategy. These genes encompass various aspects of neuronal functions including neuronal differentiation, axonal guidance, neurite outgrowth, neurotrafficking and ionic conductance (Sun et al. 2005). These processes ensure the proper development of neuronal phenotypic characteristics, neuronal synaptic plasticity and remodelling. Therefore, it is undeniable that REST serves as an essential neuronal gene regulator.

Role of REST in Embryogenesis

In embryogenesis, REST represses a subgroup of neuron-specific genes in the embryonic stem cells (ESC) to prevent neuronal gene development in non-neuronal precursor cells. REST deletion in chicken embryos is reported to cause embryonic lethality with upregulation of neuronal marker tubulin in non-neural cells (Chen et al. 1998). Moreover, REST shares 18–31% of its target genes with transcription factors for pluripotency, OCT4, SOX2, and NANOG. This indicates integration of the REST regulatory network with the ESC pluripotency network (Johnson et al. 2008). REST-deficit ESCs demonstrated a 50% decrease in self-renewal via downregulation of pluripotency markers OCT4, NANOG, SOX2 and c-MYC, with the ensuing elevated expression of differentiation markers (Singh et al. 2008).

In contrast, Jørgensen and colleagues reported contradicting results that ESCs retain the ability to self-renew despite abolishing REST. The REST-deficit ESCs maintained their pluripotency before undergoing differentiation into multiple germ layers, concluding that REST is not a critical determinant for the multi-lineage development of ESCs (Jørgensen et al. 2009). Yamada and colleagues subsequently provided evidence that REST was not required to maintain the pluripotency of ESCs. They also observed that REST is involved in the early differentiation of ESCs through the suppression of a pluripotent marker (NANOG) and the forced expression of REST also accelerated the differentiation process in ESCs (Yamada et al. 2010). However, the expression of pluripotency markers does not reflect the actual pluripotency in ESCs. During an in vivo teratoma formation assay, the REST-null ESCs can form teratoma and generate the three germ layers (Covey et al. 2012). These findings further support the argument that REST does not play a role in maintaining the ESCs pluripotency.

Nevertheless, when genetically modified ESCs (with and without REST) were cultured in different conditions (presence or absence of feeder cells and laminin), it resolved the discrepancies by inferring that REST regulation towards ESC pluripotency is highly dependent on the culture condition and the cell line (Singh et al. 2012). This finding is comparable to a previous study mentioning that the REST-null phenotypes include reduced neural progenitor cells, faulty cell adhesion and increased apoptotic cells. These phenotypes were contributed by the dysregulated REST-targeted genes Lama1, Lama2, Lamb1 and Lamc1, in which the dysregulated extracellular matrix can be improved by laminins (Sun et al. 2008). Furthermore, Singh and colleagues investigated the underlying mechanism behind REST regulation towards ESCs pluripotency. They reported that microRNA-21 (miR-21) is a REST-targeted gene, and it has a direct association with SOX2. Forced expression of miR-21 into ESCs reduced SOX2 expression and eventually diminished the ESCs self-renewal capability. Hence, it was proposed that the REST-miR-21-SOX2 axis is involved in the REST regulation towards maintaining the ESCs pluripotency (Singh et al. 2015).

Role of REST in the Developing Brain

Functioning as a neuronal gene repressor, the expression of REST is highly dependent on the neuronal cell development stages. In the central nervous system, abundant REST was detected in the undifferentiated neural stem cells (NSCs) and differentiated non-neuronal cell types, but not in the differentiated neurons (Chong et al. 1995; Mori et al. 1992; Schoenherr & Anderson 1995). A recent study also reported elevated REST expression in the embryonic developing brain, in which the regulation of REST is achieved through the corticotropin-releasing hormone receptor 1 (CRHR1)-induced cAMP-response element binding protein (CREB) activity (Kwon et al. 2023). Hence, it is postulated that REST maintains the stemness of NSCs through neuronal gene repression. In fact, protein expression studies on cortical progenitors and neurons supported the inverse relationship between the expression of REST and TUJ1, a marker for differentiated neurons (Ballas et al. 2005). In contrast, high expression of REST was detected in both the hippocampal developing progenitors and hippocampal matured neurons (Sun et al. 2005). It seems that REST functions differentially throughout the neural developmental stages and also among different populations of neurons.

REST downregulation is a necessity for proper neuronal development. As a repressing factor against neuronal differentiation, REST must be inactivated or silenced to de-repress the differentiation process. The inactivation of REST upon neuronal differentiation is regulated by NeuroD2 protein as part of its neurogenic programme. NeuroD2 activates a transcription repressor Zfhx1a, which in turn represses REST and eventually initiates the neuronal differentiation process (Ravanpay et al. 2010). Forced REST expression into differentiating neurons caused a significantly increased frequency of axon guidance error in the commissural neurons (Paquette et al. 2000) and interference with the neuronal differentiation in the hippocampus (Ravanpay et al. 2010). Upregulated REST expression in neural progenitor cells (NPCs) also resulted in the cells having disrupted radial migration during neurogenesis (Mandel et al. 2011).

Although the loss of REST leads to neuronal differentiation into different lineages, REST plays a central role in promoting the self-renewal and pluripotency of NPCs. Deletion of REST in E12.5 NPCs has generated lesser and smaller neurospheres compared to control NPCs (Covey et al. 2012). When REST was eliminated prematurely from NPCs, the cells tended to have catastrophic DNA damage during the S-phase of the mitosis cycle, resulting in premature cell cycle exit. This phenomenon, better known as precocious neuronal differentiation, is characterised by abnormal chromosome division, cell apoptosis and a depleted progenitor pool (Nechiporuk et al. 2016). Mice born with such a condition were presented with significantly smaller brains and reduced cortical thickness. REST-null mice exhibited reduced hippocampal size, disoriented cell arrangement in the subgranular zone and spinal defects, emphasising the importance of REST for proper postnatal development (Kim et al. 2015; Wang et al. 2021). Hence, REST is still necessary during the neurogenesis process to confer protection for genomic integrity until the terminal neuronal differentiation process. In short, REST downregulation is an important criterion to initiate neuronal differentiation. At the same time, it is also required at minimal levels to maintain neuronal integrity and survival. Therefore, regulating REST expression levels is key to achieving an optimal balance between neural progenitor proliferation and differentiation processes.

It is of great interest to researchers to investigate the underlying mechanism that governs REST regulation in neurogenesis. The proposed mechanism that decides whether REST is to be degraded or stabilised, relies on two motifs: (1) the C-terminal domain small phosphatase 1 (CTDSP1) recruited by REST to neuronal genes that stabilises REST in stem cells to promote self-renewal, whereas (2) the ERK-dependent phosphorylation along with peptidylprolyl cis/trans isomerase (Pin1) activation, triggers REST degradation in NPCs towards neuronal differentiation (Burkholder et al. 2018; Nesti et al. 2014). Moreover, a complex formation between insulinoma-associated 1 (INSM1) and REST corepressors RCOR1 and RCOR2 is also associated with REST regulation during neurogenesis. INSM1 is a transcription repressor that leads to cell cycle arrest. Transcript profiling analysis revealed that Rest is included in the list of dysregulated genes involved in the removal of both INSM1 and RCOR1/2 complexes, where Rest is upregulated in the telencephalon and the REST-targeted genes were repressed. Consequently, the mice lacking both RCOR1 and RCOR2 died prenatally, due to morphological defects in brain development such as deepened interganglionic sulcus, enlarged ventricular and subventricular zones, diminished corpus callosum, reduced axonal fasciculation in the striatum, defected hippocampal formation, and hypoplastic thalamus and hypothamalus. There is also unusual abundance of neural progenitor cells with reduced matured neurons (Monaghan et al. 2017). Furthermore, other microRNAs such as the miR-26, miR-124 and miR-9 family, were also reported to play a role in regulating REST expression in modulating the neuronal differentiation process (Birtele et al. 2019; Sauer et al. 2021). Taken together, findings from these studies widen the possibility of modulating REST expression in the developing brain by targeting regulatory factors.

Role of REST in the Adult Brain

REST is implicated in the adult and ageing stages, mainly for two functions: adult neurogenesis and neuroprotection. Adult neurogenesis is the process where slow-dividing quiescent neural progenitors (QNPs) in the lateral subventricular zone (SVZ) and the dentate gyrus (DG) of the hippocampus undergo self-renewal to generate either proliferating QNPs or fast-dividing, transient-amplifying progenitors (TAPs) before differentiating into granule neurons (Ming and Song 2011; Urbán and Guillemot 2014). REST plays a crucial role in this process by maintaining the quiescence and undifferentiated state of the neural progenitor cells, thereby conserving the neural progenitor pool in the adult brain to ensure continuous tissue replenishment. Precocious activation of QNPs was evident upon conditioned knockout of REST leading to increased neurogenesis over time, eventually the neural progenitor pool will lose its functionality and results in the loss of granule neurons (Gao et al. 2011; Mukherjee et al. 2016).

Another neuroprotective role of REST in the adult brain is demonstrated through the reciprocal regulation of insulin-like growth factor- and protein kinase C-dependent upregulation of the µ-opioid receptor expression. This regulatory process helps enhance the survival of maturing neurons, for instance, during stressful events or injury towards neuronal cells (Bedini et al. 2008, 2010). Furthermore, REST level increased in the prefrontal cortex and hippocampus of the healthy ageing brain, along with the upregulation of stress response genes and the downregulation of genes that promote cell death. Upon treatment with hydrogen peroxide, neural SH-SY5Y cells with REST knockout had high levels of reactive oxygen species (ROS) and increased oxidative DNA damage (Lu et al. 2014). In contrast, when REST is overexpressed up to 3- to 20-folds, there was significantly reduced neuronal cell death, and the REST expression was consolidated within the nucleus. Hence, nuclear REST expression is crucial to ensure neuronal survival against oxidative damage and preserve cognitive function in an ageing brain (Lu et al. 2014). In fact, the dysregulated REST expression in neurodegenerative diseases such as Alzheimer disease (Lu et al. 2014), Parkinson disease (Kawamura et al. 2019) and Huntington disease (Zuccato et al. 2007) further explains the inevitable role of REST as a neuroprotector against stressors in ageing brains.

The expression of REST in a healthy human brain varies throughout a human’s life cycle. Transcriptome profiling analysis of REST expression data extracted from the BrainSpan database (https://www.brainspan.org/rnaseq/search/index.html) reveals the expression profile of REST in different human developmental stages. Embryonic samples ranging from post-conception weeks (pcw) represent the prenatal stage, followed by a 4-month-old infant (infancy), 3-year-old toddler (early childhood), 13- and 18-year-old teenagers (adolescence), and 30- and 40-year-old adults (adulthood). Overall, gene-expression data revealed a decreasing trend of REST expression from prenatal to adulthood that correlates with the functional aspects of REST in each developmental stage (Fig. 1).

Fig. 1.

REST expression profiling in the brain throughout a human’s life cycle. The heatmap expression profiling of REST in different brain regions was adapted from BrainSpan: Atlas of the Human Developing Brain (https://www.brainspan.org/rnaseq/search/index.html), presented in log2 expression. The overall decreasing trend from the developing (prenatal) brain to the adult brain, correlates with the function of REST in each developmental stage. pcw, post-conception week

Transcriptional Regulation of REST

REST and its Cofactors

The regulation of REST towards the neurogenic lineage program cannot be achieved by REST on its own. It relies on the recruitment of corepressors and the assistance of regulatory factors. The full-length protein structure of REST comprises three main functional domains: two repressor domains located at the N-terminal and C-terminal respectively and a DNA binding domain that contains a cluster of eight zinc fingers located near the N-terminal (Chong et al. 1995; Tapia-Ramírez et al. 1997). The interaction between REST and its RE-1-binding site triggers the repressor domains to recruit corepressors and epigenetic regulatory factors which eventually leads to a series of transcriptional inhibitory mechanisms of REST. Upon a specific binding of REST to its binding site, the N-terminal repressor domain recruits a corepressor mSin3A, forming a platform for the binding of histone deacetylase (HDAC). Consequently, the localised histone deacetylation at the promoter region results in tightly packed nucleosomes, thereby prohibiting the transcriptional activation machinery resulting in transcriptional repression (Grimes et al. 2000; Naruse et al. 1999).

On the other hand, the C-terminal repressor domain recruits another corepressor coREST, which attracts regulatory factors important for the transcriptional epigenetic modification process. The cofactors are: HDAC, methyl-CpG binding proteins (MECP2), histone methyltransferase G9A, lysine-specific histone demethylase 1A (LSD1), carboxy-terminal binding protein 1 (CTBP1) and the ATP-dependent chromatin-remodelling enzyme brahma-related gene 1 (BRG1) that functions to stabilise the REST-RE-1 interaction. These epigenetic regulatory cofactors form a complex that is essential in facilitating the transcriptional repressive function of REST (Ballas et al. 2001; Mari´a et al. 1999).

Nuclear Localisation of REST

The nuclear localisation of REST is a key determinant for REST to perform its transcriptional regulatory function. Nuclear translocated REST is a prerequisite for REST to exert neuroprotection in ageing brains. When compared with healthy ageing brains, both Alzheimer and Parkinson disease brains lacked the nuclear REST indicating a loss of REST neuroprotection and stress resilience (Kawamura et al. 2019; Lu et al. 2014; Mampay et al. 2021; Meyer et al. 2019). On the contrary, nuclear REST was abundantly expressed in ageing brains with Huntington disease due to the mutant huntingtin proteins in the cytoplasm that could not sequester REST in the cytoplasm. The highly expressed REST in the nucleus had caused the over-repression of neuronal genes such as the brain-derived neurotrophic factors (BDNF), thereby restricting neuroprotection (Zuccato et al. 2003).

To study the mechanism of nuclear trafficking and how REST binds to its binding site, researchers initially proposed a nuclear localisation sequence (NLS) that resides in the C-terminal of REST (Chong et al. 1995; Grimes et al. 2000). However, REST4, a C-terminally truncated splice variant of REST, also retains the ability to translocate into the nucleus. Further investigation shows that the zinc fingers 2 to 5 might contain signals for nuclear trafficking and facilitate translocation machinery (Shimojo et al. 2001). The same group of researchers reported another splice variant of REST, REST1, that contains only 4 of the zinc fingers. Interestingly, REST1 is localised in the cytosol and not the nucleus suggesting that the NLS for nuclear targeting is located within the zinc finger 5 of REST (Shimojo 2006). Furthermore, the discovery of a nuclear membrane protein named REST/NRSF-Interacting LIM Domain Protein (RILP) added value to the fundamental understanding of the REST nuclear trafficking mechanism. This RILP located at the outer nuclear membrane, serves as a nuclear translocation receptor for REST and REST4. It comprises nuclear localisation signals, protein kinase A (PKA) phosphorylation sites and the farnesylation motif of RILP, in which mutation of any of the motifs leads to a malfunctioned RILP. A REST-RILP interaction would lead to nuclear localisation of REST, while a mutant RILP results in cytosol localisation of REST (Shimojo and Hersh 2003, 2006).

Role of REST in Epigenetic Remodelling

The role of REST in neural synaptic plasticity was proposed in earlier studies through the target repression of GRIN2 (Calderone et al. 2003; Huang et al. 1999). The developmental switch of N-methyl-D-aspartate receptors (NMDAR) subunit expression from GRIN2A to GRIN2B, is a process during postnatal development essential for synaptogenesis and early brain plasticity development (Liu et al. 2004). This development switch is also important to enhance long-term potentiation (LTP) (Xu et al. 2009; Yashiro and Philpot 2008) and promote hippocampus-dependent learning (von Engelhardt et al. 2008).

In 2012, an interesting association between REST and the brain cognitive function was first presented by Rodenas-Ruano et al. in 2012. REST knockout was performed on rats at different postnatal ages to investigate the impact of REST on NMDAR-mediated transmission. It was discovered that the postnatal day-10 (P10) rats exhibited greater sensitivity towards the GRIN2B-selective antagonist compared to those P24 rats. This indicates that REST contributes to the developmental switch of NMDAR only at a specific critical time window of development. Moreover, in those rat pups that were isolated from their mother (maternal deprived) during the first week of postnatal life, there was a significantly reduced REST activation and interference with the NMDAR switch. The purpose of the maternally deprived setting was to investigate the effect of an adverse experience during early life on REST activation. It was denoted that the repression of REST towards Grin2b via epigenetic remodelling facilitates the NMDAR development switching process, and that REST activation is experience-dependent (Rodenas-Ruano et al. 2012).

Another study described the novel essential role of the cellular prion protein (PrP^c) in mediating the REST-dependent NMDAR switching (Song et al. 2018). In the wildtype C57BL/6 mice, both PrP^c and REST are localised in the cytoplasm under normal conditions. REST is found to be nuclear localised upon treatment with lithium chloride (LiCl) which is a REST agonist. Conversely, the deletion of PrP^c resulted in impaired REST nuclear translocation, thus disrupting the REST-dependent NMDAR switching (Song et al. 2018). Taken together, it is clear that REST facilitates the experience-dependent fine-tuning of genes, and that PrP^c is a key player in this epigenetic remodelling. These findings enhance our understanding of the epigenetic mechanism of REST and its contributioning to neuronal synaptic plasticity, remodelling and plasticity.

REST and Its Implications in Neurological Diseases

REST and Neurodegenerative Diseases

Huntington Disease (HD)

One of the common neurodegenerative diseases associated with REST dysregulation is Huntington Disease (HD). In a normal healthy brain, the wildtype huntingtin protein (HTT) acts by sequestering REST in the cytoplasm which will prevent the nuclear translocation of REST, thereby preventing the silencing activity of REST. The interplay between wild type HTT and REST proteins is important to regulate the availability of REST to its binding site. However, this interaction is disorientated in the pathology of HD in which the mutated HTT can no longer sequester REST in the cytoplasm, resulting in anomalous accumulation of nuclear REST (Fig. 2). The loss of cytoplasmic retention of REST eventually leads to concomitant over-repression of REST-targeted neuronal genes and hence loss of neuroprotection (Zuccato et al. 2003, 2007). Ravache et al. proposed that the upregulation of REST and its over-repression might be due to the impaired Sp1 protein, a transcription activator triggered by the mutant HTT. To seek possible treatments, chromatin immunoprecipitation (ChIP) was employed to identify the REST-targeted genes that were dysregulated in HD (Ravache et al. 2010). Those dysregulated genes encompassed neuronal genes that are essential for neuronal survival (genes encoding for BDNF, FGF-1, IL-6R, CDK5R1 and CX3CL1) and neuronal synaptic connectivity (Cplx2, Syp, Syn1 and Snap25). All these genes were essential for maintaining a healthy nervous system development, which they are repressed by REST in HD pathology. However, rescue can be made through REST knockout thus it serves as a potential therapeutic intervention in HD (Soldati et al. 2013).

Fig. 2.

Nuclear localisation of REST in healthy ageing neuron and diseased ageing neurons. A proper nuclear localisation of REST is essential for it to confer neuroprotection. Stress events in a healthy ageing neuron trigger REST to translocate into the nucleus to perform transcriptional regulation and confer neuroprotection. Whereas in the diseased ageing neurons, REST is either aberrantly accumulated in the nucleus causing over-repression of neuronal genes (HD), or REST failed to translocate into the nucleus causing loss of neuroprotection (AD and PD)

To date, several possible approaches have been proposed to combat the devastating effect of REST dysregulation in HD. These approaches aim to improve the REST dysregulation from different aspects. For instance, Conforti and colleagues disrupted the REST function by overexpression of a dominant-negative form of REST protein (DN:REST) in the brain of a HD mouse model. REST availability was reduced and resulted in the restoration of BDNF and other REST-regulated genes (Conforti et al. 2013a). The same group of researchers then discovered quinolone-like compound 91 (C91), a small molecule that mimics HTT and interfered with the REST-mSIN3b complex formation. In the presence of mutant HTT, C91 can mimic HTT, interfere with the recruitment of REST cofactors and stimulate the expression of REST-targeted genes, evident through an increased level of BDNF in HD cells (Conforti et al. 2013b). The over-repressive effect of REST in HD can also be reversed by targeting its degradation through a benzoimidazole-5-carboxamide derivative, X5050, which is a potent REST inhibitor. This chemical compound has been proven to upregulate the expression of BDNF in HT iPSC-derived neurons by promoting the degradation of REST (Charbord et al. 2013). Moreover, another approach that combats the REST nuclear trafficking, can be achieved by inducing exon 3 skipping through the antisense oligo treatment into HD cells (Chen et al. 2017). This causes the loss of a critical motif needed for REST nuclear trafficking. When REST nuclear translocation is reduced, the expression of REST-targeted neuronal genes can be de-repressed. Furthermore, heat shock protein 90 (Hsp90) was found to regulate REST and mutant HTT directly. Knockout of endogenous Hsp90 resulted in significantly reduced REST and mutant HTT in cells. Therefore, an interaction between REST-Hsp90 and HTT-Hsp90 serves as a potential therapeutic target in HD (Orozco-Díaz et al. 2019).

Alzheimer Disease (AD)

The association between REST and Alzheimer Disease (AD) was first described by (Lu et al. 2014), who emphasised the significance of REST in ageing brains. In REST-deficient mice, neuronal cells begin to degenerate and undergo apoptosis when the mice were eight months old. Similarly, REST level was found to be significantly upregulated in protein and RNA extracts from the human ageing prefrontal cortex (Lu et al. 2014). This explains the inevitable neuroprotective function of REST in ageing neurons. Unfortunately, REST is found to be diminished in AD patients particularly neurons in the prefrontal cortex and hippocampus (Lu et al. 2014) and also in the sporadic AD patient’s iPSC-derived neural progenitor cells (NPCs) (Meyer et al. 2019) thus resulted in loss of neuroprotection for the vulnerable neuronal population in AD. In addition, the AD NPCs which lack REST expression also demonstrated reduced proliferative capability alongside premature differentiation, and this phenomenon can be reversed by REST overexpression (Meyer et al. 2019). Interestingly, another study on spatiotemporal immunolocalisation of REST in the brains of healthy ageing and AD transgenic rats found no significant differences in cortical REST levels between both models. However, the fact that the neurons failed to induce nuclear REST expression in the AD rat model, could be equally detrimental (Mampay et al. 2021).

The underlying mechanism of REST dysregulation in the neuropathology of AD remains controversial. Firstly, the depletion of nuclear REST often comes along with its localisation to autophagosomes containing pathologic misfolded amyloid-beta in the cytoplasm. In other words, activation of autophagy has led to the failure of REST nuclear translocation in AD brains (Fig. 2). In fact, the diminished nuclear REST accompanied by activation of autophagy were evident in other dementing diseases (Lu et al. 2014). Using the gene editing techniques, AD NPCs and organoids derived from isogenic AD-iPSC line homozygous in APOE allele E4 (APOE4) was created. These cells recapitulated the neuropathology seen in AD brains such as premature differentiation and increased tau phosphorylation. Nuclear REST level was significantly reduced in APOE4 NPCs. The affected APOE4 NPCs exhibit disrupted nuclear membrane characterised by invaginations of the nuclear lamina and intranuclear circular structures. Besides, it is also presumed that APOE4 might alter the metabolism of phospholipid and cholesterol, thereby affecting the nuclear lamina structure and prevent the nuclear translocation of REST (Fig. 2) (Levi et al. 2005; Lin et al. 2018). Thus, APOE4 and abnormal nuclear lamina morphology in AD-affected neurons, are significant factors for impaired REST nuclear translocation leading to AD neuropathology (Meyer et al. 2019).

Improper REST regulation towards the glutamate receptors and its immediate early genes also causally linked to impaired cognitive function in AD brains. Xu and colleagues observed that REST dysfunction (caused by the amyloid plaques) had led to an imbalance in neuronal excitation and inhibition, suggesting REST fails to maintain synaptic plasticity (Xu et al. 2021). On the other hand, aberrant activation of microglia around the amyloid plaques is also a prominent feature in AD aetiology (Chen and Colonna 2021; Hansen et al. 2018). In mouse microglia-like BV2 cells treated with Aβ_1-42 oligomers, both REST protein and RNA expression were elevated. REST nuclear translocation was also evident, and REST was found to suppress the expression and secretion of proinflammatory cytokines TNF-α, IL-1β and IL-6 by microglia (Yu et al. 2020a). Moreover, the conditional deletion of REST in BV2 cells induced cell migration, indicating that REST acts as a repressor for BV2 cell migration. A chemoattractant of microglia, progranulin (PGRN) can promote BV2 cell migration, potentially containing REST-binding sites in its promoter sequence. Hence, it was concluded that REST is a repressor for microglia migration by regulating PGRN expression (Yu et al. 2020b).

Parkinson Disease (PD)

There is an association between REST dysregulation and the neuropathology of Parkinson disease (PD), the second most common neurodegenerative disease (Mhyre et al. 2012). REST was found in the nucleus of normal ageing dopaminergic neurons but not in the diseased PD neurons. Diminished nuclear REST was detected in dopaminergic neurons associated with PD and there is abundant REST expression in the pathologic Lewy bodies and pale bodies (Kawamura et al. 2019). Conditional deletion of REST in PD mice models demonstrated increased vulnerability towards the dopaminergic neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). These mice were presented with lower locomotor ability and are more prone to neuroinflammation evident by a more intense activation of astrocytes and microglia (Li et al. 2020; Yu et al. 2013; Zhou et al. 2019).

Kawamura and colleagues reported the accumulation and sequestration of REST in Lewy bodies of the autophagy-dysfunction PD mice model (Fig. 2) (Kawamura et al. 2019). The cytoplasmic aggregates are comprised of p62 and ubiquitin proteins. Therefore, it is proposed that the Lewy pathology and dysregulated cellular homeostasis had caused diminished REST nuclear translocation, resulting in loss of neuroprotection for PD neurons. In the transgenic PD mice model, the misfolded alpha-synuclein aggregates have caused mitochondria dysfunction. REST can ameliorate dopaminergic neuronal-specific dysfunction (Ryan et al. 2021). Besides, REST could reverse the detrimental effects of ferroptosis which is also a significant causal factor for PD pathogenesis, in the erastin-induced LUHMES cells (Ma et al. 2022). In short, these studies emphasised the significance of REST neuroprotection for dopaminergic neurons in PD pathogenesis.

Since the dopaminergic neurons are the primary affected neuronal type in PD (Parkinson 2002), the neuroprotective potential of REST in dopaminergic neurons became a popular research focus. Dopaminergic neurons treated with the neurotoxin manganese (Mn), can lead to parkinsonian-like symptoms (Kwakye et al. 2015). REST can reverse the Mn-induced repressive effect towards the expression of the tyrosine hydroxylase (TH) enzyme that synthesises dopamine, by interacting with the RE-1-binding site in the TH promoter region. Ultimately, REST expression protects the dopaminergic neurons against Mn toxicity, characterised by improved Mn-induced oxidative stress, neuroinflammation and apoptosis (Pajarillo et al. 2020). In addition, astrocytic REST is also neuroprotective of dopaminergic neurons against Mn, by reversing the Mn-induced excitatory neuronal injury caused by dysregulation of the excitatory amino acid transporter 2 (EAAT2) (Pajarillo et al. 2021).

REST and Neurodevelopmental Disorders

Intellectual disability (ID) is categorised as one of the neurodevelopmental disorders that usually manifest during early childhood. Individuals with ID are commonly challenged with limited intellectual functioning and adaptive behaviour caused by genetic abnormalities or environmental exposure (Fidler et al. 2022). Since ID is closely related to dysfunction in the development of the nervous system, its association with REST dysregulation is worth investigating and will be discussed here.

Down Syndrome (DS)

Down syndrome (DS) is one of the most common causes of intellectual disability (Roizen and Patterson 2003). The association between DS and REST dysregulation was first described by Bahn et al. who reported a significant downregulation of REST and its spliced variants in the DS NPCs. The differentiated neurons had increased apoptosis and were presented with abnormal morphology including shortened and convoluted neurites, and excessive side branching (Bahn et al. 2002). Similarly, the downregulated REST expression was reported in two DS mice models, Tc1 and Ts1Cje (Canzonetta et al. 2008). However, compared to wildtype mice, REST was upregulated in the embryonic 152F7 DS mouse model but downregulated in the postnatal mouse (Lepagnol-Bestel et al. 2009) The expression of REST in different DS models is inconsistent across these studies (Table 1). Recently, Lim and colleagues (Lim et al. 2023) performed a comprehensive spatiotemporal REST profiling analysis in the Ts1Cje DS mice model and revealed a significantly downregulated REST expression in both embryonic DS cultured neurospheres and adult DS mice brains. To the best of our knowledge, this study also first reported the impaired REST nuclear translocation in the DS model (Lim et al. 2023).

Table 1.

Expression of REST in different Down syndrome models

DS Model	Timepoint	REST Expression (compared to wildtype)	Citation
DS patient-derived neural progenitor cells	Embryonic	Downregulated	Bahn et al. 2002
Tc1 DS mice model	Adulthood	Downregulated	Canzonetta et al. 2008
Ts1Cje DS mice model	Adulthood	Downregulated
152F7 DS mouse model	Embryonic	Upregulated	Lepagnol-Bestel et al. 2009
	Adulthood	Downregulated
Ts1Cje DS mouse model	Embryonic	Downregulated	Lim et al. 2023
	Adulthood	Downregulated

The REST dysregulation in DS was due to a dosage imbalance of a gene known as the dual specificity tyrosine phosphorylation regulated kinase 1A (DYRK1A) located in the human chromosome 21 (Canzonetta et al. 2008; Lepagnol-Bestel et al. 2009). There is an interaction between DYRK1A and a SWI/SNF complex known to be recruited by REST. This interaction is proposed to be in a strict stoichiometry manner because both the overdose and inhibition of Dyrk1a resulted in Rest suppression that occurred persistently in all the DS adult mice models. Interestingly, this observation was not seen in the DS embryonic samples, suggesting the interrelation between the chromatin status (which is expected to be distinct between embryonic and adult mice) and the expression of Rest (Lepagnol-Bestel et al. 2009). The regulation of DYRK1A towards REST was further explained in another study by Lu et al. (2011) who proposed a negative feedback loop between DYRK1A and REST: REST activates DYRK1A by interacting with its binding site in the DYRK1A promoter, whereas imbalance dosage of DYRK1A deactivates REST by facilitating in the ubiquitination and subsequent degradation of REST. The close relationship between DYRK1A and REST is hence crucial in the development of the nervous system (Lu et al. 2011).

Autism Spectrum Disorder (ASD)

Apart from DS, REST is implicated in other neurodevelopmental disorders as well. In the case of autism spectrum disorder (ASD), REST is upregulated in the Chd8 mutant C57BL/6 ASD mice model. Gene set enrichment analysis also showed that the set of genes mostly affected by mutated Chd8 is the REST-targeted genes. Chd8 knockout in the ASD mice resulted in reduced expression of the REST-targeted genes and it could be the cause of neurodevelopmental delay in ASD pathogenesis. REST dysregulation in ASD is proposed to be induced by the haploinsufficiency of CHD8 which acts as a chromatin-remodelling factor (Katayama et al. 2016). Agreeing that REST was overexpressed in ASD, corrective action to inhibit REST expression would be beneficial to combat ASD pathogenesis. Therefore, Kawase and colleagues (2019) developed ms-11 that mimics the mSin3-binding helix in REST to inhibit endogenous interaction of mSin3-REST in ASD mice model. Eventually, the chronic treatment of ms-11 had successfully improved the ASD-related symptoms in the ASD mice models (Kawase et al. 2019).

Kleefstra Syndrome (KS)

Kleefstra Syndrome (KS) is a rare neurodevelopmental disorder characterised by ID, autistic behaviour, childhood hypotonia and distinct facial appearance. It is an autosomal dominant genetic disorder caused by the haploinsufficiency of euchromatic histone methyltransferase-1 (EHMT1) (Kleefstra and Leeuw 2022). REST is interrelated with EHMT1, where EHMT1 is an epigenetic repressor that indirectly regulates the expression of REST through the repression of microRNA (miRNA) that interferes with REST protein synthesis. MiR-142, miR-153-1, miR-26a-2 and miR-548f-1 are known to be inhibited by EHMT1 and at the same time targeting REST mRNA. The expression of these miRNAs was upregulated in the KS patient-derived induced pluripotent stem cells (KS-iPSCs) (Alsaqati et al. 2022). The downregulation of EHMT1 has led to decreased REST expression in the KS cell model. Consequently, the KS-iPSCs demonstrated accelerated neuronal differentiation and elevated caspase-3 activity, indicating increased apoptosis and aberrant neuronal function. These findings reflect the detrimental effects of REST depletion which is postulated to contribute to neuronal gene dysregulation and impaired neuronal development in KS-iPSCs. In order to overcome the devastating effect of reduced EHMT1, the group of researchers used the CRISPR technique to create REST mRNA that lacks the miRNA target region, with the intention to eliminate the EHMT1-mediated miRNA regulation towards REST, eventually restore REST levels and ultimately improve the neuropathologies in KS (Alsaqati et al. 2022).

REST and the Development of Neurological Tumours

The development of tumours in the brain is one of the most pathogenic conditions among neurological diseases. Unlike neurodevelopmental and neurodegenerative diseases where low expression of REST is pathogenic in most conditions, the overexpression of REST is oncogenic and contributes to the development of brain tumours. Previous studies have reported the overexpression of REST and its oncogenic properties in several types of brain tumours, such as neuroblastoma (Liang et al. 2014; Singh et al. 2011), medulloblastoma (Fuller et al. 2005; Lawinger et al. 2000; Su et al. 2006; Taylor et al. 2012) and glioblastoma (Conti et al. 2012; Wang et al. 2023; Zhang et al. 2021). REST’s expression or activity is significantly higher in the tumour cells than the neighbouring non-cancerous cells. A common finding in all these studies is that the overexpression of REST is associated with a bad prognosis and poor overall survival of the patients.

According to Conti et al. (2012) who attempted to knock down REST in glioblastoma (GB) cells, an increased immunoreactivity of active-Caspase-3 in the GB cells was reported after REST deletion. When the transfected GB cells were transplanted into immunocompromised SCID mice, there was no sign of tumourigenesis in the animals up to 50 days. Hence, it was postulated that REST contributes to the self-renewal property of GB cells and that the deletion of REST can interfere with tumourigenesis through the activation of apoptosis (Conti et al. 2012). In addition, several studies also reported that REST knockout in the brain tumour cells resulted in elevated expression of neuronal differentiation markers such as β-tubulin III and Synapsin (Conti et al. 2012; Singh et al. 2011; Taylor et al. 2012). Since REST is well known as a transcription repressor, it was suggested that the overexpression of REST in neuronal cells could repress the neuronal differentiation capability, thereby leaving the cells in a pre-differentiated state and hence contributes to the tumourigenicity of neuronal cells (Conti et al. 2012; Fuller et al. 2005).

Since REST is a potential biomarker of poor prognosis in brain tumour-related diseases, further actions to combat REST overexpression are warranted to discover more possibilities of therapeutic options targeting REST. Singh et al. (2011) treated neuroblastoma cells with retinoic acid. The treatment caused a reduction in REST protein expression. Following that, neuronal markers β-tubulin III and Synapsin expression were upregulated. Interestingly, the effect of REST protein reduction was counteracted upon deletion of the proteasomal activity. Hence, it was suggested that retinoic acid could promote REST proteosomal degradation, which will de-repress the neuronal genes, initiate the neuronal differentiation process and further reduce the tumourigenicity of cells in the brain. Moreover, since histone deacytelases (HDACs) will be recruited for the functional repression of REST, thus inhibitors to block the activity of HDACs are expected to combat the overexpression of REST. As reported by Taylor et al. (2012), HDAC inhibitors (HDACi) such as benzamides (MS-275) and suberoylanilide hydroxamic acid (SAHA) were able to enhance the expression of REST-targeted neuronal gene Syn1, and at the same time causing the least upregulation of REST in medulloblastoma tissues. Exposure to SAHA also reduced cell growth in the samples (Taylor et al. 2012). Similarly, in a recent in-silico study by Wang et al. (2023), an enrichment analysis of REST in glioblastoma revealed “chromatin organization” and “histone modification” as the most significant enriched terms. These findings have strengthened the therapeutic potential of HDACi to treat brain tumours by targeting REST.

REST and Cerebrovascular Diseases

The association between REST and cerebrovascular diseases has been described in which REST was found to be highly expressed in animal models of ischaemic stroke. The elevated expression of REST was accompanied by upregulated expression of its corepressors, and the post-ischaemic REST-mediated dysregulation of several neuroprotective genes was able to be rectified by knocking down REST (Cheng et al. 2022; Formisano et al. 2013, 2015; He et al. 2022; Morris-Blanco et al. 2019; Noh et al. 2012). Furthermore, REST deletion also reduced apoptosis and mitochondrial damage evident through the downregulated expression of cleaved-caspase-3 and phosphorylated DRP1 (Morris-Blanco et al. 2019). The mechanism behind the overexpression of REST causing cell death was proposed due to the repressing effect of REST towards the cocaine- and amphetamine-regulated transcript (CART) transcription, which further antagonises the CREB signalling pathway, eventually augmenting cell death (Zhang et al. 2012).

Formisano et al. (2013) discovered an inverse relationship between REST and a Na⁺–Ca²⁺ exchanger 1 (NCX1), a bidirectional transporter actively involved in cerebral ischaemia. Deletion of REST restored the expression of NCX1. It resulted in positive outcomes whereby neuronal injury and infarct volume had been greatly reduced in the mice subjected to middle cerebral artery occlusion (MCOA). Moreover, the interaction between microRNA-124 (miR-124) and REST is also worth investigating as miR-124 has been reported to enhance neurovascular remodelling, leading to increased neuroangiogenesis 8 weeks post-stroke. Neuroprotection by miR-124 under ischaemic conditions was proposed to be associated with Usp14 (a deubiquitinating enzyme)-dependent REST degradation (Doeppner et al. 2013). In addition, a recent study by He et al. (2022), who investigated the effects of diabetes-induced REST elevation in ischaemic stroke, reported that REST knockout protected microvasculature against diabetic-worsened ischaemic brain injury. It was found that the expression of neuropilin-1 (NRP-1), which is a transmembrane protein mainly found in neurons and endothelial cells, was significantly increased upon REST deletion. NRP-1 contains RE-1-binding site in the promoter region in which its expression can be regulated by REST (Kurschat et al. 2006). Notably, vascular endothelial growth factor (VEGF) expression was upregulated upon REST deletion. Collectively, these findings elucidated the involvement of dysregulated NRP-1/VEGF signalling in REST overexpression-mediated vasculature defects in diabetic-worsened ischaemic brain injury (He et al. 2022).

Conclusion

Many studies have proven the significance of REST in the central nervous system. Its primary role as an epigenetic transcriptional regulator is crucial in brain development. Moreover, it is clear that REST in its dysregulated form is indeed involved in many neurological disorders throughout the brain development. Several approaches targeting REST have been proposed to rescue the pathological effects and this have raised potential in REST to become a pharmacotherapeutic target. For instance, the fact that REST is regulated by the Wnt signalling pathway (Nishihara et al. 2003), has made lithium a potential pharmacotherapeutic drug against REST dysregulation as lithium has the ability to activate the Wnt signalling pathway (Lu et al. 2014; Zhang et al. 2019). Additionally, it is noteworthy that there have been successful cases reporting the effectiveness of lithium to improve neurogenesis, synaptic plasticity and memory in Down syndrome mice models (Bianchi et al. 2010; Contestabile et al. 2013). Another study also reported enhanced antitumour effect of lithium against glioblastoma cells (Han et al. 2017). However, the involvement of REST is not implicated in these studies. Hence, future studies are needed to investigate further about the detailed underlying mechanism between REST and lithium.

In conclusion, this review paper has summarised the significance of REST throughout the road map of brain development, from embryogenesis to neurogenesis and lastly to ageing in the brain. It is important to take note that REST dysregulation can end up with severe consequences that can cause pathogenesis in the brain development. Despite emerging evidence on several approaches to rescue REST dysregulation, the detailed underlying molecular mechanism remains elusive. Additionally, the fact that REST can be a silencer and also an activator, has created a big challenge in developing REST-targeted interventions. The expression of REST varies in different diseases depending on the timepoint, the type of tissues involved and the nature of the disease pathogenicity. In regard to this, it would be beneficial if there is a comprehensive database that includes REST expression across different disease models at different timepoints. Therefore, future studies are warranted to provide better understanding of REST function in the brain and by expanding the data collection on REST expression in different disease models, the barrier to develop REST-targeted therapeutic interventions can be overcome.

Author Contributions

All authors have substantial contributions to the conception of the review article. XJL performed literature search, data analysis and drafted the manuscript; SM, PSC and KHL critically revised the manuscript; PSC gave final approval of the version to be published.

Funding

This work was supported by funding from the Malaysia Ministry of Higher Education (MOHE) Fundamental Research Grant Scheme (FRGS/1/2021/SKK06/UPM/02/4; 04-01-21-2388FR) awarded to PSC.

Data Availability

This manuscript has no associated data in a data repository.

Declarations

Competing interests

The authors have no competing interests to declare that are relevant to the content of this article.

Ethical Approval

No ethical approval is required.