Introduction

The 5th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) defines neurodevelopmental disorders (NDDs) as a range of abnormal brain developmental conditions characterized by an early childhood onset of cognitive, behavioral, and functional deficits that can result in memory and learning problems, emotional instability, and a lack of self-control (American Psychiatric Association 2013). Common NDDs include autism spectrum disorder (ASD), cerebral palsy (CP), attention-deficit/hyperactivity disorder (ADHD), Tourette syndrome (TS), fragile X syndrome (FXS), and Down’s syndrome (DS) (American Psychiatric Association 2013). Children are being increasingly diagnosed with these disorders (Boyle et al. 2011). A recent population-based study in India reported NDDs to be a significant public health burden, with the probability of a clinical diagnosis being as high as 1:8 (Arora et al. 2018). There is a higher prevalence of comorbidities such as epilepsy (Reilly et al. 2014; Gillberg et al. 2017; Alabaf et al. 2019), asthma (Chen et al. 2013; Kotey et al. 2014), headaches (Parisi et al. 2014), migraines (Fasmer et al. 2011; Sullivan et al. 2014), autoimmune disorders (Zerbo et al. 2015; Frye et al. 2017), and gastrointestinal problems (Chaidez et al. 2014; Ferguson et al. 2017; Li et al. 2017) in individuals with NDDs. A higher mortality rate has also been observed among persons with NDDs (Dalsgaard et al. 2015; Hirvikoski et al. 2016; Schendel et al. 2016). A recent study in sub-Saharan Africa revealed that although child mortality rates have dropped from 3.4 million to 2.7 million between the years 1990 and 2016, the mortality rate for children aged < 5 years with NDDs has increased from 8.6 to 14.7 million (Olusanya et al. 2018). The prevalence of ASD in the USA has increased over time, with 1 in 150 being reported in 2000–2002, 1 in 59 reported in 2014, and 1 in 54 reported in 2016 (Maenner et al. 2020). Similarly, the incidence rate of DS in the USA has also increased from 49,923 in 1950 to 206,366 in 2010 (de Graaf et al. 2016). The global burden of ADHD and TS has been reported to be 5.29% (Smith 2017) and 0.3–1% respectively (Cath et al. 2011). Comparatively, FXS has a global incidence rate of 1/4000 in males and 1/8000 in females (Peprah 2011); for in-depth population-based reports, refer to Peprah (2011). Studies have also shown that siblings of individuals with CP have a higher chance of developing NDDs and have higher mortality rates (Tollånes et al. 2016). A lack of education combined with the stigma and fear surrounding these disorders often results in children with NDDs being abandoned or institutionalized, which can cause further trauma and exacerbate psychiatric symptoms (Dumaret and Rosset 1993; Solomon and Peltz 2008). Unless tested for during pregnancy, expectant mothers are often unaware that their fetus has an NDD and are unprepared for raising such an individual. On the other hand, couples may worry about having a severely affected child with NDD and may choose to terminate the pregnancy upon detection; for example, the termination rate in Denmark following a prenatal diagnosis of DS is high (> 95%) (Lou et al. 2018). However, these tests are not always reliable (Sainz et al. 2012; Alldred et al. 2017; Santorum et al. 2017; Wiechec et al. 2017); benign variations in maternal DNA have been shown to increase false-positive screening results (Snyder et al. 2015). Similarly, these tests have also been known to generate false-negative results, which could in turn have negative psychological effects on the parents and their child (Hall et al. 2000; Simonsen et al. 2013; Simionescu and Stanescu 2020). One factor that may often be overlooked by parents is that NDDs tend to vary in their severity, range, and duration of symptoms. Hence, a positive diagnosis does not always mean that the fetus will develop the most severe form of the condition. Therefore, better counseling and a more robust screening procedure are necessary to ensure that parents can make informed decisions.

The Need for an Early Diagnostic Biomarker

While some NDDs such as DS have prenatal screening options, several other NDDs are diagnosed much later after birth. Individuals with DS or FXS may exhibit certain facial phenotypes specific to their condition, which increases the likelihood of early detection (Cornejo et al. 2017; Ciaccio et al. 2017). However, this is not true for NDDs such as ADHD, ASD, and TS that lack specific diagnostic tests and rely heavily on a behavioral diagnosis (Wilens and Spencer 2010; Randall et al. 2018; Novotny et al. 2018). Studies have shown that compared with males, females with ASD have a higher risk of remaining undiagnosed or being diagnosed later into adulthood, with some individuals diagnosed after the age of 40 years (Lehnhardt et al. 2015; Bargiela et al. 2016; Young et al. 2018; Leedham et al. 2019; Green et al. 2019). The diagnosis of ADHD is often contentious, with some studies claiming ADHD to be an overly diagnosed and over-treated condition, whereas other researchers propose the opposite (Hamed et al. 2015). Patients with ASD are often misdiagnosed with ADHD in their early years only to be correctly diagnosed much later in their adolescence (Kentrou et al. 2018). In the case of TS, although the usual age at symptom onset is around 3 years, individuals are typically not diagnosed until they are aged 8 years (Mol Debes et al. 2008; Shilon et al. 2008). This late diagnosis usually has a negative impact on individuals with NDDs as they can go through life being misunderstood, self-critical, and often suffer from psychosocial difficulties, identity crises, addiction, and poor mental health, leading to suicidal ideations (Huntley et al. 2012; Agnew-Blais et al. 2018; Leedham et al. 2019). Therefore, there is a vital need for early diagnosis and intervention for individuals with NDDs. An early diagnosis would provide more adjustment time for families to educate and prepare themselves for raising an individual with special needs. This would also ensure that lifestyle changes are made according to the child’s needs, including the use of applied behavior analysis, and other interventions such as the Early Start Denver Model, Picture Exchange Communication Systems, Discrete Trial Training, and Pivotal Response Treatment are used to educate children with NDDs instead of relying on the conventional education system (Yu et al. 2020). Early childhood intervention and treatment have been shown to reduce symptoms and complications of NDDs in adulthood, ensuring an overall better quality of life (Angulo-Barroso et al. 2008; Dawson et al. 2012; O’Neill et al. 2012; McPhilemy and Dillenburger 2013; Anderson et al. 2014; Sullivan et al. 2014; Ornoy and Spivak 2019; Smith et al. 2020). However, the current diagnostic tools are not precise enough to detect NDDs in their early stages, thus prompting the search for a more accurate and reliable biomarker that can aid in confirming the diagnosis and facilitate personalization of treatment regimens. MicroRNA (miRNA) has recently been discovered to play a role in several NDDs. This review aims to highlight its potential role as a biomarker.

miRNA: an Overview

miRNAs are small non-coding RNAs (~ 22 nucleotides in length) that are involved in the regulation of gene expression through RNA interference or gene silencing (O’Brien et al. 2018). Their function is critical to survival, as each cell in the human body possesses identical copies of DNA and cells differentiate and mature in response to the set of genes that are expressed or silenced. The process of miRNA synthesis begins in the nucleus, where RNA polymerase II transcribes miRNA genes into a hairpin loop structure referred to as primary miRNA (Yin et al. 2015). The RNA-binding protein DiGeorge Syndrome Critical Region 8 (DGCR8) then detects specific motifs such as the N6-methyladenylated GGAC region of the primary miRNA and binds to it (Alarcon et al. 2015). This is followed by DGCR8 coupling with a ribonuclease III enzyme called Drosha, which cleaves the primary miRNA into a smaller precursor miRNA (Denli et al. 2004). This precursor miRNA is then exported into the cytoplasm through a nuclear pore via the GTP-binding nuclear protein Ran/exportin-5 complex (Bohnsack et al. 2004). Upon its release into the cytoplasm, this precursor miRNA is recognized by another enzyme, the RNase III endonuclease Dicer, which further cleaves the terminal loop of this precursor to generate a shorter double-stranded mature miRNA (Liu et al. 2015). The protein Argonaute-2 (AGO-2) then associates with Dicer to bind to the mature miRNA, which initiates the unwinding of the double-stranded structure (Bossé and Simard 2010). This single guide strand binds to AGO-2 couples with other proteins to form the miRNA-induced silencing complex (RISC), which carries out the crucial function of gene silencing (Kawamata and Tomari 2010). Upon completion of transcription, this structure can interact with complementary sequences on the newly generated messenger RNA (mRNA) and can inactivate it either by cleaving the mRNA strand (Park and Shin 2014) or by hindering mRNA-ribosomal complex interaction (Antic et al. 2015) (all steps are illustrated in Fig. 1), thus inhibiting translation and protein generation. Studies have revealed that miRNAs regulate gene silencing by interacting with either the promoter region, the 5′ untranslated region (UTR), or the 3′ UTR of their target mRNA (Huntzinger and Izaurralde 2011; Xu et al. 2014; Ipsaro and Joshua-Tor et al. 2015). The interaction with the 5′ UTR results in gene silencing (Zhang et al. 2018), whereas the binding of miRNA to the promoter region results in the upregulation of transcription (Dharap et al. 2013). Although miRNAs are usually known for gene silencing, several studies have reported their paradoxical role in the upregulation of gene expression and translational activation (Vasudevan and Steitz 2007; Orom et al. 2008; Truesdell et al. 2012; Bukhari et al. 2016). For further information, readers can refer to the comprehensive reviews detailing the regulation and function of miRNAs by Gebert and MacRae (2019) and O’Brien et al. (2018). Numerous studies have confirmed the involvement of miRNAs in embryogenesis (Tang et al. 2007; Feng et al. 2015; Yuan et al. 2016), apoptosis (Chang et al. 2007; Asuthkar et al. 2012; Adams et al. 2016), cell adhesion, and intracellular signaling (Harris et al. 2008; Cuman et al., 2015; Lui et al. 2015; Cheng et al. 2020). miRNAs are highly expressed in the central nervous system and have been shown to play a role in synaptic plasticity and memory formation (Busto et al. 2015; Ryan et al. 2015; Kremer et al. 2018; Smith and Kenny 2018), axonal development of retinal ganglion cells (Marler et al. 2014; He et al. 2018; Mak et al. 2020), differentiation of oligodendrocytes (Dugas et al. 2010; Letzen et al. 2010; Buller et al. 2012; Santra et al. 2014) and astrocytes (Neo et al. 2014; Shenoy et al. 2015; Zhou et al. 2017), and the differentiation, proliferation, migration, and axonal growth of neural progenitor cells (Otaegi et al. 2011; Dajas-Bailador et al. 2012; Bian et al. 2013; Zhang et al. 2013; Radhakrishnan and Alwin Prem Anand 2016), thus highlighting their critical role in neurodevelopment. Due to its diverse physiological roles, the dysregulation of miRNA has been implicated in numerous diseases including cancer (Peng and Croce, 2016), with more recent studies reporting altered miRNA activity in NDDs.

Fig. 1
figure 1

The biosynthesis of microRNA (miRNA). The process begins in the nucleus of the cell where RNA Polymerase II transcribes the miRNA gene of interest to generate a primary miRNA. The primary miRNA is then cleaved by DiGeorge Syndrome Critical Region 8 (DGCR8) and Drosha to create a shorter precursor miRNA which is transported into the cytoplasm via Exportin 5 in a RanGTP-dependent manner. The terminal loop of this precursor miRNA undergoes further cleavage via Dicer resulting in the formation of a mature double stranded miRNA which proceeds to couple with Argonaute-2 (AGO-2), initiating the unwinding of the double stranded structure. This structure is recognized by several other proteins that associate with it to form the miRNA-induced silencing complex (RISC). The RISC carries out gene silencing either through messenger RNA (mRNA) cleavage or translational repression

Full size image

miRNA Alterations in Neurodevelopmental Disorders

Autism Spectrum Disorder

The term ASD has been assigned to individuals suffering from a set of heterogenous NDDs characterized by social communication deficits and repetitive behavior, with intellectual function ranging from severely impaired to exceptionally gifted (Nakata et al. 2019). Studies have suggested various causes and risk factors for ASD, including genetics, drug usage (e.g. anti-depressants), and environmental factors (e.g., toxins, maternal infection) (Amaral 2017). Numerous studies based on children with ASD have reported alterations in their miRNA levels (Vasu et al. 2014; Hicks et al. 2016; Kichukova et al. 2017; Hicks et al. 2020; Sehovic et al. 2020). Their saliva samples revealed downregulation of miR-23a-3p, miR-32-5p, miR-140-3p, and miR-628-5p, and an upregulation of miR-7-5p (Sehovic et al. 2020). The predicted target genes for these miRNAs (ZBTB20, GAS7, NTRK2, and SCN2A) (Table 1) have previously been shown to be involved in critical neural processes and have also been linked to ASD (Garbett et al. 2008; Chandley et al. 2015; Zhang et al. 2016; Jones et al. 2018; Spratt et al. 2019). Expression levels of 14 other miRNAs were also shown to be altered in the serum of children with ASD, with expression levels of five miRNAs (miR-181b-5p, miR-320a, miR-572, miR-19b-3p, and miR-130a-3p) being significantly different enough to distinguish ASD individuals from the control group (Vasu et al. 2014). Furthermore, higher expression levels of serum miR-365a-3p, miR-619-5p, and miR-664a-3p were detected in children with ASD, along with the downregulation of miR-197-5p, miR-328-3p, miR-424-5p, miR-500a-5p, and miR-3135a (Kichukova et al. 2017). Studies have identified Fragile X Mental Retardation 1 (FMR1) and Forkhead Box Protein P2 (FOXP2) mRNAs to be common targets for several of these altered miRNAs (Table 1) (Hicks et al. 2016; Sehovic et al. 2020). FMR1 is essential for the regulation of synapses, and alterations in its resulting protein can cause FXS (Wang et al. 2012), which is known to account for 20% of female and 40–60% of male patients with ASD (Chaste et al. 2012). Additionally, FOXP2 has been previously implicated in speech development and language disorders (Kurt et al. 2012); thus, both FMR1 and FOXP2 are strong candidates for generating ASD phenotypes. Children with ASD have also been shown to have specifically altered miRNAs compared with peers with non-ASD NDDs (Hicks et al. 2020). A test that screens for these miRNAs could therefore be more accurate in diagnosing ASD and distinguishing it from other NDDs relative to behavioral analysis alone. Another study identified overexpression of serum miR-486-3p in individuals with ASD resulting in the downregulation of AT-Rich Interaction Domain 1B (ARID1B) protein expression (Table 1) (Yu et al. 2018). ARID1B is crucial for neuronal differentiation and maturation (Ka et al. 2016; Ronzoni et al. 2016), and an insufficiency of this protein can lead to alterations in the expression levels of c-Fos and Arc resulting in the abnormal development of dendrites and synapses (Yu et al. 2018). These neural aberrations have previously been linked to cognitive deficits and autistic behavior (Curatolo et al. 2014; Tang et al. 2014). In addition, mutations in ARID1B have also been associated with the development of ASD (Pinto et al. 2014; D’Gama et al. 2015; Alvarez-Mora et al. 2016). The upregulation of miR-23a and downregulation of miR-106b have also been correlated with ASD and its associated comorbidities (Abu-Elneel et al. 2008; Sarachana et al. 2010). Altered pre-miRNA and miRNA expression levels have also been also detected in the superior temporal sulcus (miR-4753-5p and miR-1) and in the primary auditory cortex (miR-664-3p, miR-4709-3p, miR-4742-3p and miR-297) of patients with ASD (Ander et al. 2015). The target genes of these miRNAs have been reported to be functionally involved in the cell cycle, various neural processes, immune pathways, and canonical signaling pathways such as the PI3K-Akt signaling pathway, all of which have been associated with the development of ASD (Ander et al. 2015). Similarly, another study reported that patients with ASD exhibited higher expression levels of miR34c-5p, miR92a-2-5p, miR-145-5p, and miR199a-5p, as well as lower expression levels of miR27a-3p, miR19-b-1-5p, and miR193a-5p (Vaccaro et al. 2018). These miRNAs were shown to regulate SIRT1, HDAC2 (Table 1), immune system development and response, and the PI3K/Akt/TSC/mTOR signaling pathways (Vaccaro et al. 2018). Previous studies have reported similar expression levels of these miRNAs in patients with Alzheimer’s disease (AD) (Wang et al. 2014). Additionally, MECP2, a common target of miR-199a-5p (Table 1), has also been linked to Rett syndrome (RS), which could explain the similarities observed between ASD and RS (Vaccaro et al. 2018). A recent study identified overlapping expression levels of miR-19a-3p, miR-361-5p, miR-3613-3p, miR-150-5p, miR-126-3p, and miR-499a-5p in both animal and human models of ASD (Ozkul et al. 2020). Several studies have also observed altered miRNA function in association with ASD: miR-34b and miR-103a-3p (Huang et al. 2015), hsa-miR-21-3p and hsa_can_1002-m (Wu et al. 2016), miR-6126 (Nakata et al. 2019), miR486-3p (Popov et al. 2012), and miR-142-5p, miR-142-3p, miR-451a, miR-144-3p, and miR-21-5p (Mor et al. 2015). The genes regulated by these miRNAs have all been predicted to be directly or indirectly involved in the pathophysiology of ASD and its associated comorbidities. Age-dependent alterations in miRNA expression levels have also been observed in patients with ASD (Stamova et al. 2015); this could explain the lack of a coherent trend in previous studies.

Table 1 Alterations in the expression levels of miRNAs and their corresponding target genes in neurodevelopmental disorders
Full size table

Attention-Deficit/Hyperactivity Disorder

Commonly observed in children, ADHD is a complex NDD characterized by hyperactivity and an inability to pay attention or control impulses, which can hinder educational progress and cause symptoms that persist into adulthood (Wu et al. 2015). This disorder can present with comorbidities such as ASD and other behavioral disorders in association with common risk factors such as genetics, early childhood trauma, prenatal or postnatal exposure to lead, and premature birth (Thapar et al. 2013). Children with ADHD have been shown to have higher serum expression levels of hsa-miR-101-3p, hsa-miR-130a-3p, hsa-miR-138-5p, and hsa-miR-195-5p as well as lower expression levels of hsa-miR-106b-5p (Zadehbagheri et al. 2019). Hsa-miR-101 is responsible for various neurological processes and in particular neuronal excitation, synapse formation, and dendritic growth. By supervising the expression levels of NKCC1, KIF1A, and ANK2 (Table 1), it can modulate gamma aminobutyric acid (GABA) signaling and protect the brain from hyperexcitability and memory impairment (Lippi et al. 2016). In hippocampal neurons, this miRNA is known to regulate Ran-binding protein (RANBP9) (Table 1) expression levels and hence control the metabolism of the amyloid precursor protein (APP), which plays a critical role in AD (Barbato et al. 2014; Zadehbagheri et al. 2019). Hsa-miR-130a controls neurite outgrowth and dendritic density by repressing the MECP2 gene (Table 1), a dysfunction that can lead to RS and ASD (Zhang et al. 2016). It has been suggested that ADHD and ASD may share a common molecular mechanism (Zadehbagheri et al. 2019). hsa-miR-138-5p has been shown to monitor dendritic spine formation in hippocampal neurons (Siegel et al. 2009) by upregulating glycogen synthase kinase-3β (GSK-3β) (Table 1) and downregulating retinoic acid receptor alpha (RARA) (Table 1). This enhances the phosphorylation of tau protein, which is involved in synaptic plasticity and memory formation (Schröder et al. 2014; Wang et al. 2015). Animal model studies have revealed that hsa-miR-195-5p inhibits dendritic degeneration and neuronal death by downregulating APP, death receptor 6 (DR-6), and BACE1 (Table 1) in the hippocampal and cortical brain regions (Ai et al. 2013; Chen et al. 2017). On the other hand, hsa-miR-106b orchestrates cell cycle arrest during neuronal lineage differentiation by modulating cyclin D1 (CCND1), E2F1, CDKN1A (p21), PTEN, RB1, RBL1 (p107), and RBL2 (p130) (Table 1) expression levels (Trompeter et al. 2011). Previous studies have also reported polymorphisms in the gene for hsa-miR-106b to be linked to ASD (Toma et al. 2015). Significantly higher serum levels of miRNA-let-7d have been detected in children with ADHD relative to their healthy counterparts (Wu et al. 2015; Cao et al. 2019). miRNA-let-7d plays a crucial role in cellular differentiation and pluripotency, somatic reprogramming, neurogenesis, and synaptic plasticity (Andolfo et al. 2010; Wong et al. 2012; Wu et al. 2015); animal model studies have identified that overexpression of this miRNA in the rat prefrontal cortex results in suppression of the TH gene (Table 1), which is essential for dopamine metabolism (Wu et al. 2010; Weaver et al. 2012). Altered expression levels of miR-652-3p, miR-148b-3p, and miR-942-5p were also detected in patients with ADHD in association with the downregulation of their target genes (B4GALT2, SLC6A9, TLE1, ANK3, TRIO, TAF1, and SYNE1) (Table 1) (Nuzziello et al. 2019). These genes are crucial for neuron survival (Dastidar et al. 2012), cognitive functioning (Iqbal et al. 2013), neuronal migration (Zong et al. 2015), axon guidance (Zong et al. 2015), memory formation (Nuzziello et al. 2019), and regulation of glutamate and GABAA receptors at the postsynaptic membrane (Zhang et al. 2017; Rathje et al. 2019). Several other miRNAs have also been reported to alter the expression of ADHD-associated genes such as SNAP-25, BDNF, DAT1, HTR2C, and HTR1B (Wu et al. 2010; Németh et al. 2013; Sánchez-Mora et al. 2013; Kandemir et al. 2014; Ye et al. 2016; Wu et al. 2017; Srivastav et al. 2018; Tian et al. 2019; Huang et al. 2019). Recent studies have linked the SLC1A3 gene with ADHD susceptibility (Turic et al. 2005; Huang et al. 2019). This gene encodes the excitatory amino acid transporter 1 (EAAT1), which is essential for astrocytic glutamate reuptake upon synapse, preventing spill-over into neighboring synapses; it is predominantly found in the cerebellum, which has a role in motor movements (Parkin et al. 2018). Mutations in this gene can alter the hsa-miR-3171 (Table 1) binding site (Huang et al. 2019). Knockdown of the EAAT1 analogue in mice has been shown to generate ADHD phenotypes (Karlsson et al. 2008; Karlsson et al. 2009). Furthermore, methylphenidate, the conventional drug used to treat ADHD, has been proven to control the striatal levels of EAAT1 (Cavaliere et al. 2012), further emphasizing the crucial regulatory role of miRNAs in NDDs.

Tourette Syndrome

Characterized by the early childhood onset of involuntary motor and vocal tics, TS can be present with other comorbid disorders such as obsessive-compulsive disorder, anxiety, ASD, ADHD, behavioral disorders, and learning disabilities (Pagliaroli et al. 2020). However, unlike other NDDs, little is currently known about the molecular mechanisms underlying TS. Several studies have detected genetic variants of the Slit protein and Trk-like family member 1 (SLITRK1) gene in brains of individuals affected by TS (Abelson et al. 2005; O’Roak et al. 2010). Similar expression levels of mutant SLITRK1 mRNA (containing var321) and hsa-miR-189 have been noted in the neocortex, hippocampus, and cerebellum of TS mice models (Abelson et al. 2005). SLITRK1 is crucial for dendritic growth in neurons, and its expression levels are regulated by miR-189 (Table 1) (Abelson et al. 2005). However, the var321 mutation upregulates the miR-189 silencing function, which results in the neuro-morphological abnormalities seen in patients with TS (Abelson et al. 2005; O’Roak et al. 2010). Several studies have detected polymorphisms in the SLITRK1 gene that contributes to the TS phenotype (Miranda et al. 2009; Karagiannidis et al. 2012; Inai et al. 2015); however, other studies have reported no SLITRK1 gene variants in certain populations (Deng et al. 2006; Fabbrini et al. 2007; Zimprich et al. 2008; Yasmeen et al. 2013), suggesting genetic heterogeneity in TS. Lower expression levels of miR-429 have also been reported in patients with TS relative to healthy controls (Rizzo et al. 2015). miR-429 belongs to the miR-200 family that is known to modulate the epithelial-mesenchymal transition by silencing ZEB1 and ZEB2 (Table 1) (Korpal et al. 2008; Ragusa et al. 2010). These miRNAs have been shown to induce differentiation of midbrain dopaminergic neurons as well as ventral midbrain and hindbrain neural progenitor cells by regulating SOX2 and E2F3 (Table 1) expression levels (Peng et al. 2012). Animal model studies have also reported that miR-429 plays a crucial role in dendrite formation and synaptic plasticity (Eipper-Mains et al. 2011; Rizzo et al. 2015), thus further supporting the morphological alterations observed in brains of individuals with TS (Steeves et al. 2008; Patel et al. 2011; Tossell et al. 2011; Rizzo et al. 2015). miR-429 shares the same genetic locus (1p36.33) as several copy number variants have been implicated in ASD (Vaishnavi et al. 2013; Marrale et al. 2014). Genetic variants of the nicotinic acetylcholine receptor alpha 7 subunit (CHRNA7) gene and Netrin 4 (NTN4) gene (Table 1) have also been reported to alter the seed sequence of miRNA-106b and miRNA-198b, resulting in a five-fold reduction in their expression levels in brains of individuals affected by TS (Pagliaroli et al. 2017). CHRNA7 is essential for the release of various neurotransmitters (Sinkus et al. 2015) and for cognitive processes (Wallace and Porter 2011; Lendvai et al. 2013; Wallace and Bertrand 2013), whereas NTN4 has been shown to play a role in cell migration, tissue morphogenesis, angiogenesis, and apoptosis (Xu et al. 2017). Lower expression levels of CHRNA7 have also been previously reported in ASD brains, suggesting a possible link between the two (Allen-Brady et al. 2010; Yasui et al. 2011). Similarly, a more recent study reported genetic variants of the LIM homeobox 6 (LHX6), inner mitochondrial membrane peptidase subunit 2 (IMMP2L), and arylacetamide deacetylase (AADAC) genes were associated with altered miRNA function in patients with TS (Pagliaroli et al. 2020). LHX6 is critical for the regulation of striatal and cortical interneurons (Grigoriou et al. 1998; Liodis et al. 2007; Zhao et al. 2008; Kreitzer et al. 2009) as well as the differentiation and formation of neural and lymphoid cells (Tepper et al. 2007). The protein encoded by IMMP2L ensures that transit proteins in the mitochondria are correctly processed, thus indirectly regulating apoptosis (Pagliaroli et al. 2020). Although not much is known about the function of the AADAC protein in the brain, studies have detected AADAC expressed in the hippocampus, corpus callosum, and caudate nucleus, which are regions commonly affected in TS (Plessen et al. 2009). The mutant AADAC also occurs in close proximity to the binding site of miR-4263, which has been reported to play a role in the differentiation of embryonic and neural stem cells (Goff et al. 2009; Pagliaroli et al. 2020), thus suggesting a possible role in TS.

Fragile X Syndrome

FXS is a common X-linked inherited intellectual disorder that arises due to hypermethylation of the upstream CpG islands and CGG repeats in the 5′ UTR of the FMR1 gene, resulting in the downregulation of fragile mental retardation protein (FMRP) expression, which is an essential translational regulator during synapsis (Jin et al. 2004a; Wang et al. 2012). FXS has been shown to be a monogenic cause of ASD (Wang et al. 2012). FMRP can reversibly block ribosomal progression on mRNA during translation, thus regulating the levels of several presynaptic and postsynaptic proteins that have been associated with ASD (Darnell et al. 2011). Early studies in Drosophila have revealed that the miRNA protein AGO-1 predominantly regulates FMRP function during neurodevelopment and synaptogenesis (Jin et al. 2004b), whereas modulation via the bantam miRNA has been shown to be involved in primordial germ cell differentiation and maintenance (Yang et al. 2009). miR-125 and miR-132 (Table 1) are known to interact with FMRP to regulate group 1 metabotropic glutamate receptors (mGluR1) and N-methyl-d-aspartate receptor (NMDAR) signaling during normal neurodevelopment (Lin 2015). An in vivo study in zebrafish that utilized anti-FMR1 miRNA detected synaptic deformities and increased long-term depression (LTD) via mGluR1 (Lin et al. 2006). Another zebrafish study reported that miRNA overexpression resulted in hypermethylation of the FMR1 5′-r(CGG) region causing neurite deformations and synaptic dysfunction, suggesting a crucial role in FXS (Chang et al. 2008). The 3′ UTR of the FMR1 gene was also shown to be targeted by miR-19b, miR-302b*, and miR-323-3p (Table 1), causing gene repression and generating the FXS phenotype (Yi et al. 2010). A knockout study in mice revealed overexpression of several pre-miRNAs and miRNAs, with miR-34b, miR-340, and miR-148a (Table 1) shown to significantly downregulate reporter gene expression levels through the Met 3′ UTR (Liu et al. 2015). The miRNA miR-302 has been shown to play a physiological role by inhibiting FMR1 translation and regulating neurodevelopment during the early stages of embryogenesis (Lin 2015). Following the blastocyst stage, miR-302 expression levels are downregulated, thus freeing FMR1 to synthesize FMRP and initiate neural development. However, in FXS, there is an accumulation of mutated miRNAs that results in hypermethylation of the promoter regions of the FMR1 gene, causing gene repression and a lack of FMRP production (Lin 2015). During spermatogenesis, the accumulation of these mutated miRNAs results in the downregulation of FMRP expression, causing hyperproliferation of Sertoli cells and spermatogenic defects (Ramaiah et al. 2019). These miRNAs have also been reported to modulate the levels of the eukaryotic translational initiation factor (eIF4E) and the cytoplasmic FMR1 interacting protein 1 (CYFIP1) that, when coupled with FMRP, create a translational regulatory complex (Ramaiah et al. 2019). A more recent in vitro study reported that the genes involved in the differentiation of neurons and axon guidance were abnormally expressed in FXS-derived neurons (Halevy et al. 2015). These genes were shown to be targeted by the RE-1 silencing transcription factor (REST), which was overexpressed in FXS cells. A further investigation revealed lower expression levels of hsa-mir-382 in these cells, and the subsequent reintroduction of an hsa-mir-382 (Table 1) analog resulted in REST suppression and upregulation of its target genes, thus highlighting the importance of miRNAs in normal neurodevelopment. In addition, recent urine samples from children with FXS revealed the overexpression of several miRNAs involved in the regulation of developmental processes, homeostasis, and neuronal function (Putkonen et al. 2020). Of these miRNAs, miR-125a was shown to have a significantly higher expression level compared with the control group. miR-125a has been associated with normal mGluR1 regulation, thereby hindering synaptic plasticity in FXS (Putkonen et al. 2020). This further supports the notion that miRNAs can be used as potential biomarkers for FXS.

Down’s Syndrome

DS or trisomy 21 is a chromosomal abnormality that arises due to an extra copy or fragment of the 21st chromosome. It is associated with severe comorbidities including cognitive deficits, mental retardation, immune system disorders, leukemia, congenital heart defects, hypotonia, dementia, and early onset AD (Lim et al. 2015a). Due to the high mortality rate associated with this disorder, expectant women are widely offered screening and diagnostic testing as optional components of prenatal care. The prevalence of DS has been directly linked to maternal age (Wu and Morris 2013). A recent genome-wide microarray study in DS placental samples reported overexpression of miRNA on chromosome 21 (miR-99a, miR-125b, and let-7c) as well as other chromosomes (miR-542-5p, miR-10b, miR-615, and miR-654) (Lim et al. 2015b). The placental samples also exhibited higher expression levels of mir-1973 and mir-3196, which have been suggested to modulate the genes involved in neurodevelopment. Another study reported up to 20 genes being upregulated in DS placental samples, with BACE2, CBS, RUNX1, and CRYAA showing significant associations with DS comorbidities (Lim et al. 2015a); miR-504, miR-320c, miR-1298, miR-1297, miR-891b, and miR-513a-3p (Table 1) have been reported to be involved in upregulation of these genes. This study also found miR-133b, miR-188-3p, and miR-1301 to be responsible for the downregulation of five other genes linked to complications of DS such as mental retardation, behavioral problems, and congenital abnormalities. Several human and animal model studies have reported altered miRNA expression levels in fetal samples affected by DS (Xu et al. 2013b; Lin et al. 2016; Arena et al. 2017; Karaca et al. 2018; Zbucka-Kretowska et al. 2019). A recent study in pregnant women with fetal DS reported elevated levels of hsa-miR-15a, hsa-let-7d, hsa-miR-142, hsa-miR-23a, hsa-miR-199, and hsa-miR-191, as well as lower levels of hsa-miR-1290, hsa-miR-1915, hsa-miR30e, hsa-miR-1260, hsa-miR-483, hsamiR-548, and hsa-miR-590, with their target genes being implicated in generating DS phenotypes (Zbucka-Kretowska et al. 2019). Another study reported higher levels of miR-125b-2, miR-155, and miR-3156 in the amniotic fluid of pregnant women with fetal DS relative to controls (Karaca et al. 2018). Overexpression of miR-99a-5p, miR-155-5p, and let-7c-5p has also been implicated in cardiac anomalies in fetal DS (Izzo et al. 2017). The prevalence of leukemia and immune system disorders in patients with DS could also be attributed to alterations in miRNA expression levels (Xu et al. 2013a; Shaham et al. 2015). Overexpression of miR-155, miR-802, miR-125b-2, let-7c, and miR-99a has been reported in DS with downregulation of their target mRNA (Elton et al. 2010; Elton et al. 2013; Siew et al. 2013; Alexandrov et al. 2018). The complement factor H (CFH) mRNA that is targeted by miR-155 (Table 1) is critical for neuroprotection, the immune response, complement opsonization, and leukocyte infiltration (Griffiths et al. 2009; Li et al. 2012). Thus, the overexpression of miR-155 may silence CFH function, possibly resulting in the pathological phenotype observed in patients with DS (Li et al. 2012). Both in vivo and in vitro studies have reported higher expression levels of miR-155 and miR-802 (Table 1) to result in decreased levels of methyl-CpG binding protein 2 (MECP2), causing developmental defects and neuronal malformations (Elton et al. 2010; Keck-Wherley et al. 2011; Lu et al. 2013). Lower levels of MECP2 in DS have also been associated with poor synaptic strength, which could result in cognitive deficits and reduced synaptic plasticity (Chao et al. 2007). Intellectual deficits due to the downregulation of enhancer of zeste homolog 2 (EZH2) by miR-138-5p (Table 1) have also been associated with DS (Shi et al. 2016). Therefore, altered miRNA function could play a significant role in neurodevelopment and DS-related comorbidities.

Conclusion

Although our current knowledge regarding the role of miRNAs in NDDs is limited, it is a promising area for future research. Several of the abovementioned studies detected altered miRNA expression levels using liquid biopsy samples (e.g., blood, urine, saliva) from patients with NDDs, thus eliminating the necessity for invasive procedures while also providing a more reliable diagnostic method than behavioral diagnosis alone. Northern blotting, in situ hybridization, reverse transcription qPCR, microarray, and next-generation sequencing are few of the existing assays that can detect altered miRNA levels in liquid biopsies (Dave et al. 2018). However, each method has its advantages and disadvantages (summarized by Dave et al. 2018), and researchers have yet to discover a full-proof miRNA-based screening test for NDDs. Currently, the Food and Drug Administration (FDA) has authorized a number of miRNA-based tests for cancer diagnosis which have been proven to be more sensitive and specific as compared with the conventional screening methods (Dave et al. 2018). Therefore, there is a possibility for miRNA to be used as a potential diagnostic marker for NDDs in the near future. The utilization of miRNAs to facilitate early diagnoses would ensure that individuals with NDDs are provided with the right kind of care beginning in their developmental years. They could be introduced to special enrichment classes designed for their condition, allowing them to learn and develop in a safe environment. An early diagnosis would also help prepare families and caretakers well in advance. Current treatment options for NDDs are limited. However, identification of target miRNAs could facilitate the development of new treatment options.

Availability of Data and Material

Data sharing is not applicable to this article as no new data were created or analyzed in this review work.