**2. m6A and Neurogenesis**

#### *2.1. Writers*

During embryonic neurogenesis, Mettl14 displays the highest expression in radial glia cells, and *Mettl14* knockout (KO) in embryonic mouse brains extends the cell cycle of radial glia cells and induces aberrant cortical neurogenesis. Similar defects were induced by Mettl3 knockdown [11]. Mettl14 also regulates the cell cycle of human cortical neuronal progenitor cells [11]. The deletion of *Mettl14* in embryonic neural stem cells (eNSCs) led to a remarkable decrease in proliferation and immature differentiation in vitro and in vivo [16]. In addition, *Mettl3* knockdown reduced the proliferation and skewed the differentiation of adult neural stem cells (aNSCs) towards neuronal lineage, while the newborn neurons displayed immature morphology [12]. Transcriptome analysis revealed that the deficiency of either *Mettl3* or *Mettl14* affected the expression of transcripts related to neurogenesis, the

cell cycle, and neuronal development [11,12,16]. *Mettl3* conditional-knockout mice showed severe developmental defects of the cerebellum and cell death [17]. These results suggest an essential and conserved function of m6A in maintaining normal neurogenesis in the mammalian brain (Figure 2A). histone H3 trimethylation at lysine 27 (H3K27me3) in aNSCs [12]. Ectopic Ezh2 could rescue *Mettl3*-knockdown-induced deficits in aNSCs [12]. These findings suggest a crosstalk between RNA modification and transcriptional regulation and reveal a new layer of the mechanism regulating neurogenesis (Figure 2C).

neurons displayed immature morphology [12]. Transcriptome analysis revealed that the deficiency of either *Mettl3* or *Mettl14* affected the expression of transcripts related to neurogenesis, the cell cycle, and neuronal development [11,12,16]. *Mettl3* conditional-knockout mice showed severe developmental defects of the cerebellum and cell death [17]. These results suggest an essential and conserved function of m6A in maintaining normal

m6A regulates gene expression not only through regulating RNA metabolism but also via modulating mRNAs encoding histone modifiers and transcription factors [25]. In mouse eNSCs, transcripts for histone acetyltransferases CBP (CREB binding protein) and p300 are m6A-modified [16]. In addition, transcripts for histone methyltransferase Ezh2 are also m6A-modified, and *Mettl3* knockdown reduces the level of Ezh2 and consequent

*Cells* **2021**, *10*, x 3 of 10

neurogenesis in the mammalian brain (Figure 2A).

**Figure 2.** m6A modification in neural development and neurological disorders. (**A**). Schematic representation of neurogenesis. Neural stem cells have the capability to self-renew and differentiate into neural cells, such as neurons, astrocytes, and oligodendrocytes. (**B**). Loss of m6A modification affects histone modifications, including H3K27me3 and H3K27ac, which regulate the expression of genes related to the proliferation and differentiation of neural stem cells. (**C**). The modulation of m6A modification machinery contributes to neurodegenerative diseases, including Alzheimer's disease and Parkinson's disease, through the regulation of multiple pathways, such as mTOR. AD, Alzheimer's disease; PD, Parkinson's disease; TSC1, tuberous sclerosis 1; TSC2, tuberous sclerosis 2. **Figure 2.** m6A modification in neural development and neurological disorders. (**A**). Schematic representation of neurogenesis. Neural stem cells have the capability to self-renew and differentiate into neural cells, such as neurons, astrocytes, and oligodendrocytes. (**B**). Loss of m6A modification affects histone modifications, including H3K27me3 and H3K27ac, which regulate the expression of genes related to the proliferation and differentiation of neural stem cells. (**C**). The modulation of m6A modification machinery contributes to neurodegenerative diseases, including Alzheimer's disease and Parkinson's disease, through the regulation of multiple pathways, such as mTOR. AD, Alzheimer's disease; PD, Parkinson's disease; TSC1, tuberous sclerosis 1; TSC2, tuberous sclerosis 2.

> *2.2. Erasers*  The fat mass and obesity-associated (Fto) gene was originally referred to as an obesity-risk gene and is the first identified m6A demethylase [26]. The loss-of-function mutation of the *Fto* gene caused growth retardation and severe neurodevelopmental disorders, including microcephaly, functional brain defects, and delayed psychomotor activity in humans [27–29]. *Fto*-deficient mice showed increased postnatal mortality, significant loss m6A regulates gene expression not only through regulating RNA metabolism but also via modulating mRNAs encoding histone modifiers and transcription factors [25]. In mouse eNSCs, transcripts for histone acetyltransferases CBP (CREB binding protein) and p300 are m6A-modified [16]. In addition, transcripts for histone methyltransferase Ezh2 are also m6A-modified, and *Mettl3* knockdown reduces the level of Ezh2 and consequent histone H3 trimethylation at lysine 27 (H3K27me3) in aNSCs [12]. Ectopic Ezh2 could rescue *Mettl3*-knockdown-induced deficits in aNSCs [12]. These findings suggest a crosstalk between RNA modification and transcriptional regulation and reveal a new layer of the mechanism regulating neurogenesis (Figure 2C).

#### *2.2. Erasers*

The fat mass and obesity-associated (Fto) gene was originally referred to as an obesityrisk gene and is the first identified m6A demethylase [26]. The loss-of-function mutation of the *Fto* gene caused growth retardation and severe neurodevelopmental disorders, including microcephaly, functional brain defects, and delayed psychomotor activity in humans [27–29]. *Fto*-deficient mice showed increased postnatal mortality, significant loss of adipose tissue and body mass, and disordered energy homeostasis [27,30]. The constitutive loss of *Fto* decreased brain size and body weight, impaired the pool of adult neural stem

cells (aNSCs), and impaired the learning and memory of mice [15]. Specific ablation of *Fto* in aNSCs also inhibited neurogenesis and neuronal development [13]. In addition, specific deletion of *Fto* in lipids led to decreased neurogenesis and increased apoptosis [14]. These findings indicate that Fto regulates neurogenesis through diverse pathways, including affecting brain-derived neurotrophic factor (BDNF) signaling, the expression of plateletderived growth factor receptor (Pdgfra) and suppressor of cytokine signaling 5 (Socs5), and adenosine levels [13–15].

Another m6A demethylase, Alkbh5, is primarily localized in the nuclear speckles. Alkbh5-mediated demethylation activity affects nuclear RNA export and RNA metabolism and, consequently, regulates gene expression. The cerebellum of *Alkbh5*-deficient mice did not show detectable changes in weight and morphology, but *Alkbh5*-KO mice were more sensitive to hypoxia and showed a significantly reduced size of whole brain and cerebellum compared to control littermates [18]. In addition, the number of proliferating cells was significantly increased, but mature neurons were reduced in the cerebellum of *Alkbh5*-deficient mice [18], which suggests that *Alkbh5* deficiency affects the proliferation and differentiation of neuronal progenitor cells.

#### *2.3. Readers*

Ythdf1 is preferentially expressed in the hippocampus of mouse brains. Genetic deletion of *Ythdf1* impaired the learning and memory of mice, whereas it did not affect gross hippocampal and cortical histology, neurogenesis, and motor abilities [31]. Electrophysiological data showed that *Ythdf1*-deficient neurons had reduced spine density and decreased amplitude and frequency of miniature excitatory postsynaptic currents, which could be rescued by ectopic Ythdf1 [31]. This study further showed that Ythdf1 facilitates learning and memory by promoting the translation of target transcripts, including Gria1, Grin1, and Camk2a induced by neuronal stimulation.

m6A reader Ythdf2 is critical for embryonic development and has a lethal effect in mice [32]. *Ythdf2*-deficient mice embryos were alive at embryonic day 12.5, 14.5, and 18.5 but displayed abnormal brain development, including reduced cortical thickness and decreased proliferation of neural stem/progenitor cells (NSPCs) [32]. In addition, *Ythdf2* deficiency skewed the differentiation of NSCs towards neuronal lineage, but newborn neurons had fewer and shorter neurites [32].

Fragile X mental retardation protein FMRP can bind mRNAs, and FMRP target mR-NAs are significantly enriched for m6A modification [22]. The loss of the FMRP coding gene *Fmr1* altered the m6A landscape and reduced the expression of FMRP-targeted long mRNAs in the cerebral cortex of adult mice. In addition, FMRP can interact with Ythdf2 [22]. This study provides a new layer of mechanism that specifies how FMRP regulates neuronal development and brain function.
