Plasticity is a critical factor enabling stem cells to contribute to the development and regeneration of tissues. In the mammalian central nervous system (CNS), neural stem cells (NSCs) that are defined by their capability for self-renewal and differentiation into neurons and glia, are present in the ventricular neuroaxis throughout life. However, the differentiation potential of NSCs changes in a spatiotemporally regulated manner and these cells progressively lose plasticity during development. One of the major alterations in this process is the switch from neurogenesis to gliogenesis. NSCs initiate neurogenesis immediately after neural tube closure and then turn to gliogenesis from midgestation, which requires an irreversible competence transition that enforces a progressive reduction of neuropotency. A growing body of evidence indicates that the neurogenesis-to-gliogenesis transition is governed by multiple layers of regulatory networks consisting of multiple factors, including epigenetic regulators, transcription factors, and non-coding RNA (ncRNA). In this review, we focus on critical roles of microRNAs (miRNAs), a class of small ncRNA that regulate gene expression at the post-transcriptional level, in the regulation of the switch from neurogenesis to gliogenesis in NSCs in the developing CNS. Unraveling the regulatory interactions of miRNAs and target genes will provide insights into the regulation of plasticity of NSCs, and the development of new strategies for the regeneration of damaged CNS.
The discovery of neural stem cells (NSCs) in the developing and adult brain of mammals, including human, has raised expectations for their clinical application in cell replacement therapies for central nervous system (CNS) disorders.1,
In mammals, the initial transcript containing miRNAs is transcribed mostly by RNA polymerase II from the exons and introns of coding genes, or from intergenic regions encoding long ncRNAs containing one or more miRNAs. A single strand of the mature miRNA duplex, is incorporated into the RNA-induced silencing complex (RISC) containing Dicer, a ribonuclease III enzyme, and Argonaute 2 protein (Ago2). The miRNA–RISC complex recognizes mRNA targets through partial base pairing to the seed sequence, which is 6–8 nucleotides mostly located at positions 2–8 from the 5′-end of miRNA, and mediates their repression through translational repression, mRNA degradation, or mRNA deadenylation. Each miRNA recognizes multiple target transcripts and each mRNA transcript is targeted by multiple miRNAs. Details on the biogenesis and functions of miRNA, including non-canonical miRNA, have been reviewed elsewhere.12,13
Hundreds of miRNAs are detected in the developing and adult mammalian CNS, and some of these have been shown to have critical roles in CNS development.14,
Regulation of gliogenic competence by miR-153 and miR-17/106
The initiation of gliogenesis by NSCs proceeds in two steps: acquisition of gliogenic competence and induction of differentiation.5,21 In the developing mammalian CNS, NSCs acquire gliogenic competence at midgestation, and subsequently oligodendrogliogenesis begins.22 Astrocytic differentiation can be detected only in the perinatal period.21 The gliogenic competence is defined as the competence to respond to gliogenic signals that induce glial differentiation in NSCs. NSCs in the early-neurogenic phase before midgestation cannot differentiate into glia, even in the presence of these signals.22,
We also identified miR-17 and 106a/b (miR-17/106), belonging to the miR-17 family of miRNAs that share a common seed sequence as critical regulators for neurogenesis-to-gliogenesis switch in NSPCs downstream of chicken ovalbumin upstream promoter-transcription factor I and II (COUP-TFI and COUP-TFII, also known as NR2F1 and NR2F2), which belong to the orphan nuclear hormone receptor family (Figure 2).43 Ahead of this discovery, we had found that the transient increase of COUP-TFI/II expression in NSPCs at midgestation is essential for their temporal specification, including acquisition of gliogenic competence.22 The double knockdown (KD) of Coup-tfI/II in ESC-derived NSPCs and the developing mouse forebrain resulted in sustained neurogenesis and the prolonged generation of specific types of neurons, which are normally born only at the early neurogenic stage. Initially, miR-17/106 was found to be expressed in NSPCs specifically at the early-neurogenic stage and increased in response to the Coup-tfI/II KD in ESC-derived NSPCs. Our functional analyses revealed a time-dependent decrease of miR-17/106 expression in developing NSPCs, leading to an increase of one of their targets, mitogen-activated protein kinase 14 (MAPK14, also known as p38α) in turn, which is essential for the acquisition of gliogenic competence. Importantly, miR-17-mediated OE or LOF of MAPK14 in stage-progressed highly gliogenic NSPCs resulted in robust recovery of neuropotency, which is literally an instance of ‘rejuvenation’. However, miR-17 OE did not alter the temporal change in the epigenetic status of Gfap promoter.
These results suggest that the acquisition of gliogenic competence involves at least two distinct regulatory layers, regulation of the expression of glial-specific genes, including epigenetic regulation, which involves regulation by the Notch-NFIA and miR-153-NFIA/B axes, and regulation of NSPC neuropotency by the miR-17/106-MAPK14 axis. The former is currently only observed in the regulation of astrogliogenesis. In fact, miR-153 OE does not alter oligodendrogliogenesis in ESC-derived NSPCs (Tsuyama et al., unpublished result). In the latter, miR-17/106 may be essential for the maintenance of neuropotency in NSPCs, as the neurogenic phenotype induced by miR-17 OE in developing and highly gliogenic NSPCs is extremely pronounced compared with that induced by miR-153 OE, and even resembles to that of early-neurogenic NSPCs (Figure 2a,b). Moreover, miR-17 has been shown to suppress astrocytic differentiation of mouse cortical progenitors via repression of BMPR2, which is acquired in cortical VZ after midgestation and is essential for BMP2/4 signaling, which inhibits neurogenesis and promotes astrogliogenesis.44 If that is the case, then MAPK14 may be required for the competence change of NSCs to allow glial differentiation in the presence of sufficient amounts of gliogenic factors, including NFIA/B and gliogenic signals. The miR-17 OE in ESC-derived NSPCs does not alter Nfia/b expression (Naka-Kaneda et al., unpublished result). Alternatively or additionally, miR-17/106 may suppress differentiation of NSCs into glial lineages independent of the epigenetic regulation of glial-specific genes (Figure 2c). In this case, the expression level of MAPK14 may determine the frequency of the commitment into glial lineages and NFIA/B may simply be required for astrocytic differentiation of glial-restricted precursors (GRPs), because NFIA OE suppresses differentiation of oligodendrocyte progenitor cells into oligodendrocytes. It is possible that the majority of undifferentiated NSPCs increased as a result of miR-153 OE41 are GRPs and/or astrocyte precursors. To date, there is no reliable marker to distinguish these glial progenitor cells from NSCs, although GRPs can be isolated in vitro.45,46 We also do not know whether astrocyte and oligodendrocyte lineages can be differentiated directly from NSCs, or must necessarily pass through the GRP state. Future studies regarding glial development should thus include a focus on identification of definitive NSCs and/or GRP markers.
Regulation of neuronal and glial differentiation by miRNAs
miRNAs are also involved in the terminal differentiation of neurons and glia at appropriate time points. Initial analyses of Dicer conditional knockout (CKO) mice using Nestin, Emx1-, Foxg1- and Olig1-Cre drivers revealed that the contribution of miRNAs to neuronal and glial differentiation in the developing CNS is more pronounced after mid-gestation.14,18,
Among several dozens of specific miRNAs involved in neurogenesis and gliogenesis, the let-7 family of miRNAs seems to have a key role in the timing of glial differentiation (Figure 3a). let-7 miRNAs are heterochronic genes encoding components of machinery underlying timing mechanisms in animal development, and were first identified as genes that control the timing of cell-fate decisions in the larval development of Caenorhabditis elegans.53 In mammals, let-7 miRNAs have been shown to be involved in a wide range of biological processes, including differentiation and temporal specification of several types of stem cells.54 In NSC development, let-7b was initially found to be involved in the age-dependent decline of stem cell function by targeting high-mobility group-AT-hook 2 (HMGA2) a member of the high-mobility group A (HMGA) family, which encodes a small non-histone chromatin-associated protein that modulates transcription of many genes by altering chromatin structure.55 Age-dependent increase of mature let-7 miRNAs in NSPCs after midgestation seems to contribute to the age-dependent decline of HMGA2, which induces upregulation of the cell cycle inhibitors p16INK4a and p19ARF. let-7b has also been shown to induce differentiation of NSPCs into neurons and glia by targeting an orphan nuclear receptor TLX and cyclin D1 (Figure 3a).56 Moreover, deletion of Insulin-like growth factor two mRNA binding protein 1 (IGF2BP1; also known as IMP-1, ZBP1), another target of let-7 miRNAs in mice, reduces proliferation of NSPCs and induces premature differentiation of neurons and astrocytes along with a reduction of HMGA2 expression in the developing cortex, while let-7g OE exhibits a similar phenotype with a significant reduction of IGF2BP1 expression.57 Gain-of-function (GOF) and LOF phenotypes of HMGA2, however, do not completely fit in the expected phenotypes as a target of let-7 miRNAs. In the mouse embryonic cortex and human NSPC cultures derived from hPSCs, HMGA2 appears to suppress astrocyte differentiation but be required for neurogenesis.58,59 This is unsurprising, as let-7 miRNAs regulate multiple target genes, as described above. Overexpression of let-7 miRNAs suppresses neuronal differentiation, depending on the cellular context. In the neurogenesis from hESC-derived NSPCs and zebrafish retinal injury, let-7 miRNAs are likely to target a neurogenic basic helix-loop-helix transcription factor Achaete-scute homolog 1 to inhibit neurogenesis60 or prevent premature Müller glia dedifferentiation upon retinal injury.61 Intriguingly, Cimadamore et al.60 reported that let-7 activity was not detected in hESC-derived NSPCs owing to the high expression level of LIN28, a well-known inhibitor of let-7 biogenesis driven by SOX2. This finding suggests that these NSPCs were in an early-neurogenic phase because the temporal patterns of LIN28 and mature let-7 miRNA expression in developing NSPCs are inversely correlated. SOX2 and LIN28 are downregulated along the neuronal differentiation of hESC-derived NSPCs, thereby let-7 is derepressed in differentiating neurons. Phenotypes induced by the ectopic expression of particular genes are often inconsistent with their physiological functions. Recently, let-7 and miR-125 have been shown to promote astrocytic differentiation of GRPs derived from mouse ESCs through the regulation of multiple targets in parallel with JAK/STAT pathway (Figure 3a).62 Moreover, let-7 OE in hPSC-derived NSPCs enhanced oligodendrogliogenesis.56 Thus, let-7 miRNAs are involved in gliogenesis at multiple levels. The developmental increase of let-7 miRNAs in NSPCs thus seems to be integrated into the timing mechanisms of gliogenesis (Figure 3b).
A recent large body of work on miRNAs has revealed their significant contribution to a wide range of biological processes, including cell-fate determination events. One might also argue that the major function for miRNAs may be buffering stochastic noise in the gene expression,63 as LOF of individual miRNAs often results in mild or no phenotype under normal physiological conditions.64 This may be due in part to compensation by other miRNAs targeting the same mRNAs.65 In contrast, GOF of a miRNA causing a significant reduction of the expression of multiple target often results in a clear phenotypic change.66 However, as described above, several miRNAs have critical roles in the regulation of neurogenesis-to-gliogenesis switch in the developing CNS (Figure 3b). Similarly, let-7, miR-125, and miR-9 have also been found to have major roles in the temporal specification of retinal progenitors.67 Inhibition of all these or let-7 activity leads to the prolonged generation of ganglion cells from progenitors stuck in an early state. Moreover, evidence is accumulating for the involvement of specific miRNAs in the age-related changes in the properties of several tissue stem cells such as hematopoietic stem cells and mesenchymal stem cells in addition to NSCs.54,68 Aging usually causes reduction of regenerative capacity of stem cells.69
The limited regenerative capacity of adult tissues in mammals compared with that of other lower vertebrates, such as urodele amphibians and teleost fish, may largely be due to the limited number and plasticity of tissue stem cells. Therefore, rejuvenation of adult stem cells represents an important target in regenerative medicine and aging research. Given the robust contribution of miRNAs to the temporal specification of several tissue stem cells including NSCs, controlling stem cell plasticity via LOF and GOF of specific miRNAs along with the mobilization of stem cells may serve as a possible therapeutic approach to various degenerative diseases. Further studies to elucidate how miRNAs regulate plasticity of stem cells during aging will provide important information for better understanding of stem cell aging and development other therapeutic approaches including use of small molecules that modulate functions and/or expressions of specific targets of key miRNAs.
We are grateful to Dr D. Sipp for his critical reading of the manuscript. TS and HO are supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan. This work was also partly supported by Research Center Network for Realization of Regenerative Medicine from the Japan Science and Technology Agency (JST) and Japan Agency for Medical Research and Development (AMED) to HO, and by the Program for Intractable Disease Research utilizing disease-specific iPS Cells from JST and AMED to HO.
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/