In the cerebellum, both astrocytes and oligodendrocytes are characterized by a peculiar heterogeneity. Astrocytes include Bergmann glia, granular layer, and white matter astrocyte types, each endowed with specific morphological and functional features; oligodendrocytes derive from multiple embryonic sources and establish neuron type-specific interactions. Nevertheless, while the mechanisms of neuronal diversification have been deeply investigated and in part elucidated, gliogenesis in the cerebellum remains poorly explored. Here, we critically discuss the available knowledge regarding the ontogenesis of the repertoire of glial cells in the cerebellum, their relationships with cerebellar neurons, and the processes regulating the acquisition of their mature morphological and molecular traits, pointing out the open questions that still seek elucidation.
KeywordsAstrocytes Oligodendrocytes Cerebellum Development Morphological and functional diversity
External granular layer
Purkinje cell layer
Prospective white matter
The cerebellum is characterized by a remarkable anatomical and functional complexity, mirrored not only by the variety of its neuronal phenotypes but also by a notable heterogeneity of glial cells. Specific morphological features of astrocytes are associated with defined functional properties, unique among the astroglia of the entire central nervous system. Cerebellar oligodendrocytes, by myelin deposition along defined axon subsets, critically shape cerebellar circuit activity. Further, both glia types heavily influence cerebellar circuit development. However, in the last decades, the interest in lineage, differentiation, heterogeneity, and functions of cerebellar glia has lagged behind studies on cerebellar neurons and circuits. As a result, while mechanisms on neuronal diversification have been widely investigated and partially clarified, astrogliogenesis remains poorly explored. Given the increasing relevance of glia in developmental processes and in the correct functioning of mature circuitries, filling this gap is crucial in order to achieve a deeper understanding of the complexity of cerebellar structure and functions. Here we review the current status of research on gliogenesis and highlight crucial questions that remain to be addressed.
Origin and Differentiation of Cerebellar Astrocytes
The Phenotypic Heterogeneity of Cerebellar Astrocytes
Glial cells of the cerebellum were first described by Ramόn y Cajal (1911), who distinguished three main categories, according to their morphology and position in cerebellar layers. In the white matter, (i) glial cells (including both astrocytes and oligodendrocytes) display processes oriented along the direction of axons; the cerebellar cortex, on the other side, hosts (ii) the astrocytes of the granular layer, with star-shaped bushy processes, and (iii) the so-called neuroepithelial cells with Bergmann fibers (commonly known as Bergmann glia, BG), with cell bodies aligned to the Purkinje cell (PC) somata, several ascending processes spanning radially into the molecular layer, and terminal end feet contacting the pial surface (Fig. 1). Afterward, electron microscopy analyses confirmed this classification and unveiled another rare type of granular layer astrocytes, called protoplasmic, with thin processes devoid of lamellae (Fig. 1; Palay and Chan-Palay 1974). These studies further showed veil-like appendages emanating from the processes of the major type of granular layer astrocytes (thereby named velate astrocytes; Fig. 1) and described their relationship in the local neuropil with groups of granule cells and glomeruli (see chapter “Development of Glutamatergic and GABAergic Synapses”). Similarly, BG processes were shown to ensheath the entire dendritic surface of PCs and their synapses, thereby suggesting that both kinds of cortical astrocytes actively participate in the arrangement of cerebellar cortical synapses and exert specialized functions in the related circuitries. On the other hand, the closer inspection of white matter astrocytes led to their classification as fibrous astrocytes (Fig. 1). A special staining procedure, the Cajal’s gold sublimate technique (Cajal 1926; Globus 1927), also allowed, early in 1916, the identification of a second astroglial phenotype located in the Purkinje cell layer (PCL) and intermingled with BG, with “feathers” of cytoplasmic extensions shorter than processes of BG: these cells were called feathered cell of Fañanas, after the name of their discoverer (Fañanas 1916). Nevertheless, due to their morphological similarity to BG, their identification remained difficult, explaining why these cells were essentially forgotten. Only recently, these neglected cells found a first confirmation in a study that identified a subset of astrocytes in the cerebellar molecular layer showing specific expression of defined potassium channel-related peptides (Goertzen 2018).
Origin of Cerebellar Astrocytes
Ramόn y Cajal, early in 1911, investigated cerebellar development of different species and concluded that all cerebellar glia derive from the ventricular zone (VZ) (Ramon y Cajal 1911). This evidence was later demonstrated in mouse by fate-mapping analyses using inducible reporter genes expressed in ventricular radial glia (RG) (Hoshino et al. 2005; Mori et al. 2006; Sudarov et al. 2011). Ramόn y Cajal also argued that BG progenitors result from the retraction of the apical processes of RG lining the VZ of the cerebellar primordium that, keeping their end feet attached to the pial surface, subsequently delaminate radially and translocate into the nascent parenchyma, where they settle in the PCL (Ramόn y Cajal 1911). Successively, anatomical investigations on neurochemically identified glia provided evidence consistent with this interpretation (Yuasa 1996; Yamada and Watanabe 2002). In further support of this notion, abrogation of protein tyrosine phosphatase Shp2-dependent extracellular signal-regulated kinase (ERK) signaling in ventricular RG perturbed their transformation into BG (Li et al. 2014). Moreover, Gdf10 (growth and differentiation factor 10), a glial marker specific for both developing and mature BG, was shown to be expressed in the VZ according to a precise pattern. In the lateral cerebellar primordium, Gdf10 is expressed in the posterior portion of the VZ, whereas it appears diffusely produced throughout the whole extension of the VZ in the medial cerebellum (Mecklenburg et al. 2014). These observations support a direct derivation of BG from RG and suggest that a restricted subset of RG may generate BG, although the compartmentalization of gliogenic ventricular progenitors remains to be demonstrated.
Altman and Bayer (1997) also noted that some delaminating cells continue to divide in the overlying tissue and, therefore, first proposed that cerebellar astrocytes may be generated by a second wave of progenitors that proliferate within the developing cerebellar parenchyma. Indeed, astroglial-like progenitors were successively described to delaminate during late embryonic development and to reach the prospective white matter (PWM), where they proliferate postnatally to produce parenchymal astrocytes and GABAergic interneurons (Yamada and Watanabe 2002; Parmigiani et al. 2015). Thus, these studies suggested that astrocytes of the granular layer and white matter derive from intermediate progenitors amplifying in the PWM, yet without clarifying whether these progenitors could also produce BG. Recent in vivo clonal analyses of astroglial cerebellar lineages shed a new light on these issues (Cerrato et al. 2018a), unveiling that postnatal progenitors in the PWM can produce all the three main astrocyte types. Moreover, this study also disclosed an unprecedented bi-potency of BG progenitors in the postnatal PCL, that generate both BG and granular layer astrocytes, but not white matter astrocytes. Therefore, two distinct populations of intermediate progenitors can be found in the postnatal cerebellar parenchyma: PWM progenitors, capable of giving rise to all main astrocyte types, and BG progenitors in the PCL, producing BG and granular layer astrocytes.
The delamination of RG and their translocation toward the nascent parenchyma were known to occur soon after the completion of PC genesis and their emigration from the VZ, at around embryonic day (E) 14 in mice, when this germinal layer exclusively gives rise to non-neuronal elements (Yuasa 1996; Yamada and Watanabe 2002). Hence, this developmental time point has always been defined as the start of gliogenesis in the mouse cerebellum. Nevertheless, recent studies on individual embryonic RG unveiled that these cells are gliogenic as early as E12, a developmental phase previously thought to be fully neurogenic, and highlighted a well-defined spatiotemporal pattern of astrocytes generation, characterized by a time-dependent allocation of the cells first to the hemispheres and then to the vermis, depending on their origin at E12 or E14, respectively (Cerrato et al. 2018a). This same study also unveiled that astrogliogenesis in the cerebellum occurs according to a remarkably orderly developmental program, where embryonic ventricular progenitors produce either only a single astrocyte type or more types and go through a decline over time in their amplification and lineage potentials.
Whether or not also the rhombic lip (RL) and its derivatives populating the external granular layer (EGL) contribute to cerebellar astrogliogenesis still remains controversial (Buffo and Rossi 2013 and references therein). This possibility was first suggested based on 3H-thymidine labeling and immunolabeling for astrocyte markers, but other studies did not confirm this view. However, manipulations of morphogens (Notch, Sonic Hedgehog (Shh), and bone morphogenetic protein 2 (BMP2)) indicated that RL and EGL progenitors have the potential to generate astrocytes. Consistently, the BG-specific gene Gdf10 (see above) was recently detected also in the embryonic RL (Mecklenburg et al. 2014). Moreover, EGL progenitors generate both neurons and glia ex vivo. In turn, a possible relationship between RL-derived neurons and BG was recently highlighted by the new evidence that a subset of nestin-expressing astroglial progenitors in the PCL (i.e., arguably BG progenitors) can switch their fate and regenerate granule cells in mice after ablation of the perinatal EGL (Wojcinski et al. 2017). In summary, a minor contingent of cerebellar astrocytes may derive from RL precursors. However, production of astrocytes from EGL progenitors remains to be unequivocally confirmed in vivo.
Lineage Relationship Between Cerebellar Astrocytes and Neurons
The notion that cerebellar neurons and astrocytes are generated from the same germinative niches (i.e., the VZ and, possibly, the RL) raises the hypothesis that they might be clonally related. Several fate-mapping analyses, in which embryonic progenitors were targeted and followed based on the expression of specific markers, suggested a common origin between the two lineages. When classical RG markers such as BLBP (brain lipid-binding protein), Glast (glutamate aspartate transporter), and TenascinC were used, the labeled progenies comprised cerebellar inhibitory neurons as well as BG and parenchymal astrocytes (Anthony et al. 2004; Anthony and Heintz 2008; Fleming et al. 2013). Moreover, in vivo targeting of ventricular RG through injections of lentiviral vectors displaying a tropism toward astroglial cells demonstrated that RG can generate astrocytes and interneurons during late embryonic development (Parmigiani et al. 2015). In parallel, in fate-mapping studies exploiting the regulatory regions of transcription factors known to be necessary for the specification toward the neuronal lineage (i.e., pancreatic transcription factor 1a (Ptf1a), Neurogenin 2, achaete-scute family BHLH transcription factor 1, Ascl1), some astrocytes could be detected among the offspring cells (Hoshino et al. 2005; Florio et al. 2012; Sudarov et al. 2011). Nevertheless, the low frequency of astroglial cells observed in these studies questions the actual origin of most cerebellar astrocytes from the tagged progenitor populations. Common ancestors for cerebellar neurons and glia were also suggested by functional studies performed at embryonic stages. Manipulation of both Notch/BMP signaling and proneural genes (Ascl1) resulted in an altered balance between the numbers of neurons and astrocytes, with macroscopic cerebellar defects (Buffo and Rossi 2013), indicating lineage contiguities and that proneural genes may act by suppressing default gliogenic differentiation programs in ventricular multipotent precursors.
Further, the existence of bipotent precursors for neurons and glia in the postnatal cerebellum was supported by studies reporting the isolation of neurosphere-forming cells from perinatal cerebella and describing their capability to differentiate into neurons and astrocytes both in vitro and following transplantation in postnatal cerebella (Klein et al. 2005; Lee et al. 2005). Consistent with these reports, genetic fate-mapping analyses of cerebellar postnatal GFAP-expressing progenitors (Silbereis et al. 2009) and stem cell-like PWM precursors (Fleming et al. 2013) indicated common progenitors for cerebellar astroglia and inhibitory interneurons. However, the most conclusive evidence of bipotent neuroglial progenitors was recently provided by Parmigiani and colleagues (2015). In this study, by in vitro, ex vivo, and in vivo clonal analyses at the single progenitor level, the authors showed that GABAergic interneurons and white matter astrocytes share a common ancestor residing in the postnatal PWM. Examining whether such progenitors also exist at earlier developmental ages and, if so, what neuron and astroglia subpopulations they might generate will be of particular interest in the future.
Postnatal Amplification of Intermediate Astrocyte Precursors
Several birthdating studies reliably reported that most cerebellar astrocytes are generated during late embryonic and postnatal development (Miale and Sidman 1961; Altman and Bayer 1997; Sekerková et al. 2004). Three discrete sites of intense proliferation can be identified in the postnatal cerebellum, comprising the EGL, the PWM, and the PCL (Yuasa 1996; Altman and Bayer 1997; Parmigiani et al. 2015). Not considering the possible, minor, origin of glial cells from the EGL (see above), it is likely that the bulk of cerebellar astrocytes derive from the amplification of intermediate precursors homed in the PWM and PCL, that, as discussed above, represent independent progenitor pools with distinct fate potencies (Altman and Bayer 1997; Yamada and Watanabe 2002; Parmigiani et al. 2015, Cerrato et al. 2018a). Interestingly, recent proliferation and birthdating analyses unveiled distinct layer-dependent rhythms of amplification and cell cycle exit in the postnatal cerebellum. This translates into astrocyte clones with highly stereotyped architectures, resulting from the combination of recurrent modules (i.e., subclones) typically composed of constant relative numbers of different astrocyte types. Namely, white matter astrocytes leave the cell cycle early, while proliferation in cortical layers is more intense and lasts longer, especially in the PCL, in parallel with the tangential expansion of the cerebellar surface (Cerrato et al. 2018a). Distinct molecular machineries could sustain different rhythms, as suggested by the nonoverlapping distribution of distinct cyclins isoforms between BG progenitors in the PCL (expressing cyclinD1) and PWM precursors (expressing cyclinD2) (Leto et al. 2011; Parmigiani et al. 2015), which associates with higher proliferative activity and faster cell cycle reentry of BG compared to PWM cells (Parmigiani et al. 2015). As a further level of heterogeneity between the two progenitor pools, bFGF (basic fibroblast growth factor) and PC-derived Shh are known to promote the proliferation of PWM precursors (Lee et al. 2005; Fleming et al. 2013), while neither of the two appear to affect BG progenitor expansion (Dahmane and Ruiz i Altaba 1999). However, available data on the role of Shh in regulating the proliferation of BG progenitors are conflicting (Wojcinski et al. 2017). Thus, the same factors may influence at various extent and with different outcomes astrocyte progenitors in different layers due to either different intrinsic cell sensitivities or layer-dependent variations in factor concentration, or both.
Taken together, these observations indicate that, although all cerebellar astrocyte phenotypes originate from the same germinative neuroepithelium, they have distinct natural histories. BG progenitors initially derive from the morphological transformation of RG and, later during postnatal development, extensively proliferate to match the concomitant expansion of the cerebellar tissue. On the other side, parenchymal astrocytes are posited to derive from proliferative events of RG, whose daughter cells migrate to the overlying parenchyma and here continue to divide.
Differentiation of Cerebellar Astrocytes
Early co-culture studies showed that cerebellar neurons profoundly influence the morphology, antigenic profile, and proliferative rates of cerebellar astroglia (Hatten 1985; Nagata et al. 1986). Subsequent studies especially focused on disclosing the cellular and molecular mechanisms of BG differentiation (reviewed in Leung and Li 2017), likely because this astroglial type displays a highly specialized morphology, exerts crucial functions in supporting cerebellar development at both the cellular and structural levels, and regulates PC synaptic activity (possible cross-refs: see chapters “Granule Cell Migration and Differentiation,” “Foliation,” and “Purkinje Cell Migration and Differentiation”). Hereafter, we will present available evidences on the acquisition of BG typical morphology and layering and subsequently address BG molecular maturation. Of note, the processes of maturation of the other cerebellar astroglial phenotypes remain essentially undetermined.
Maturation of Morphological Features
After withdrawal of their apical process, RG displace their cell body toward the cerebellar cortex and intermingle with PCs. Here, they undergo several morphological changes to produce BG and, together with developing PCs, form a multilayered structure, which later gradually transforms into a monolayer. Different from RG that possess a single basal branch, during maturation, BG extend multiple ascending processes (usually three to six per cell in mice) that cross the molecular layer. These multiple branches emerge during postnatal development: they increase in number until the end of the first postnatal week and then decrease (Yamada and Watanabe 2002), in parallel with the expansion and reduction of the EGL (Shiga et al. 1983), therefore suggesting that granule cells contribute to regulate the formation of BG processes (see below). At early maturation stages, BG fibers are rather smooth with small enlargements and excrescences that progressively grow to bushy expansions covering most of the radial process (Shiga et al. 1983) and establishing tight interactions with PC synapses (Yamada and Watanabe 2002).
For the induction of the proper morphological phenotype, BG progenitors need to interact with three components of the cerebellar environment: the subpial basement membrane, PCs, and granule cells (Buffo and Rossi 2013). The basement membrane lies on the surface of the cerebellar cortex, and it is deposited by meningeal elements during early development; for BG correct polarization, process outgrowth, and positioning in the PCL, BG end feet need to be tightly anchored to it. If this anchorage is defective, the fibers fail to form and orient, and BG cell bodies move to the molecular layer, resulting in a severe disruption of cerebellar foliation and layering (Sievers et al. 1994; Li et al. 2014; He et al. 2018). Several receptors and intracellular transduction pathways are implicated in this anchorage, mediating a bidirectional cross talk with components of the extracellular matrix. They include β1-integrins (Graus-Porta et al. 2001; Frick et al. 2012) and other molecules associated with their intracellular domains (integrin-linked kinase ILK) (Belvindrah et al. 2006), the ABL tyrosine kinases (Qiu et al. 2010), RIC-8A (Ma et al. 2012), and members of the dystrophin-dystroglycan complex (Moore et al. 2002; Qu and Smith 2005; Satz et al. 2008). Consistent with their physical interactions with developing BG, PCs strongly impact on BG morphological maturation. This action is exerted through Shh secretion (Dahmane and Ruiz-i-Altaba 1999) and juxtacrine signals based on the Notch pathway (Stump et al. 2002; Eiraku et al. 2005; Weller et al. 2006; Komine et al. 2007; Hiraoka et al. 2013). Moreover, all the transformation events occurring in the shape and number of BG fibers are intimately associated with the growth of PC dendrites and afferent fibers rewiring (Yamada et al. 2000). Further influence on BG development is exerted by granule cells, which act through Notch signaling (Stump et al. 2002), the NRG pathway, and the secretion of FGF9, thereby providing additional regulators of BG differentiation (Schmid et al. 2003; Lin et al. 2009).
In addition to the extrinsic influence of the surrounding neurons on astrocyte differentiation, several cell-intrinsic determinants are known to be involved in the morphological maturation of BG. Among Sox proteins of group C, known to be important modulators of the maturation of both neurons and glia in the central nervous system, regulated Sox4 levels were demonstrated to be crucial for RG migration into the position normally taken by BG and maintenance of radial fibers projecting toward the pial surface (Hoser et al. 2007), while Sox2 was recently implicated in the correct organization and localization of BG after birth (Cerrato et al. 2018b). Conversely, stellate astrocyte morphologies did not appear to be affected. Moreover, the transcription factor Zeb2 was recently shown to promote BG formation and differentiation likely through alteration of FGF, Notch, and TGFβ (transforming growth factor beta)/BMP downstream targets (He et al. 2018). In addition, PTEN (phosphatase and tensin homolog) deletion in hGFAP-positive cells also induced the disappearance of cells with the typical morphology of BG without affecting other parenchymal astrocytes, thus revealing that active PTEN signaling is intrinsically required for correct BG differentiation and for the maintenance of a polarized phenotype (Yue et al. 2005). Moreover, a misalignment of BG and abortive formation of radial fibers that often do not contact the pial surface was observed in the absence of the ubiquitin ligase Huwe1 (D’Arca et al. 2010). As formerly mentioned, the deletion of Ptpn11, coding for the tyrosine phosphatase Shp2, was also shown to block the transformation of RG in BG and to affect cerebellar foliation, pointing to the critical role of ERK signaling in BG induction (Li et al. 2014). Manipulations of FGF ligands/receptors known to activate ERK or of the transcription factors Etv4 and Etv5 known as FGF targets/mediators also led to BG severe defect demonstrating that the FGF-ERK-ETV axis is important for the induction of BG (Leung and Li 2017 and references therein). Finally, after deletion of the tumor suppressor gene adenomatous polyposis coli (APC) in GFAP-expressing cells, BG differentiation proceeded normally during the first postnatal week, but at later stages, their cell bodies translocated to the molecular layer, then losing contacts to the pial surface and acquiring a stellate morphology (Wang et al. 2011). This indicates that the radial phenotype not only has to be actively promoted but also continuously maintained. Collectively, impairment of these regulatory mechanisms results in BG malpositioning and/or the acquisition of a stellate morphology, which may thus represent a default differentiation pathway for cerebellar astroglial precursors. Yet, it is likely that the refinement of the variety of multipolar morphologies in the granule cell and white matter is instructed by local cues.
Maturation of Molecular Profiles
But how are these molecular profiles achieved? The mechanisms of molecular maturation of cerebellar astrocytes, which underlie their functional specialization, have been poorly addressed and are mostly unknown. A recent study unveiled the role for PC-derived Shh in regulating the molecular and functional profile of cerebellar cortical astrocytes (Farmer et al. 2016). In BG, Shh signaling sustains the expression of GLAST, Kir4.1, and the AMPA subunits GluA1 and GluA4, thereby maintaining AMPA receptor-mediated currents. Moreover, it prevents AQP4 expression. However, when Shh signaling was constitutively activated in velate astrocytes, these cells could acquire the same molecular and functional signature of BG. Thus, by blocking Shh signaling in BG and by constitutively activating it in granular velate astrocytes, which are physiologically exposed to lower amounts of Shh, the authors showed that the two subtypes are interchangeable under both the molecular and functional point of view. However, these molecular and functional changes were not mirrored by morphological modifications, indicating the role of additional intrinsic/extrinsic cues in the regulation of astrocyte morphological heterogeneity.
Importantly, first single cell RNA-seq studies on cerebellar development have recently sketched out distinct astrocyte clusters whose molecular identities and regulatory transcriptional networks still remain largely unexplored (Carter et al. 2018; Gupta et al. 2018). The datasets of these studies, therefore, represent very valuable sources to be carefully examined in the future to unveil the still unknown molecular identity of the distinct astrocyte progenitors, and to clarify what regulatory factors for gene expression are implicated both in the precursors’ fate decision and in the specification, maintenance, and function of each cerebellar astrocyte type.
On the whole, the available data indicate that BG maturation requires tight and timely regulated stimuli coming from the surrounding environment as well as intrinsically defined mechanisms. An impairment of these regulatory mechanisms often results in a failure in acquiring and/or maintaining a radial morphology and translates in the acquisition of a stellate multipolar shape. On the other side, possible instructors of specific phenotypic traits of the other stellate astroglial phenotypes still need to be deeply investigated.
Origin and Differentiation of Cerebellar Oligodendrocytes
Oligodendroglial cells in the mature cerebellum comprise both oligodendrocyte progenitors and fully differentiated myelinating oligodendrocytes. Oligodendrocyte progenitors are dispersed throughout the cerebellar white matter and granular and molecular layers. They maintain a degree of proliferation also at adult stages and are capable of differentiating into oligodendrocytes if myelin is damaged. Moreover, they likely sustain a certain degree of myelin remodeling throughout life (Nishiyama et al. 2009; Young et al. 2013). Myelin deposition is restricted to PC axons and afferent climbing fibers in the white matter and granular layer (Palay and Chan-Palay 1974; Reynolds and Wilkin 1988). Furthermore, despite the abundant presence of oligodendrocyte precursors, the only myelinated processes present in the molecular layer are occasional branches of the recurrent supraganglionic plexus of PC axons (Palay and Chan-Palay 1974; Rossi et al. 2007). This peculiar pattern reflects type-specific interactions between oligodendrocytes and different categories of cerebellar cortical neurons that still have to be elucidated.
Origin of Cerebellar Oligodendrocytes
Despite the apparent largely independent origins of cerebellar oligodendrocytes and astrocytes, a few evidences support the presence of bipotent gliogenic progenitors. Oligodendrocyte precursors expressing the chondroitin sulfate proteoglycan NG2 were described to generate astrocytes besides oligodendrocytes in cerebellar postnatal ex vivo explants (Leoni et al. 2009). However, these data were contradicted by in vivo genetic fate mapping of NG2-expressing cells showing exclusive production of oligodendrocytes (Zhu et al. 2008). Notably, upon deletion of the polycomb group protein Bmi1 leading to an aberrant and ectopic activation of the BMP pathway, an increased generation of astrocytes was found, associated with a decrease in oligodendrocytes (Zhang et al. 2011). These findings suggested the existence of bipotent progenitors that, however, were not characterized by the authors.
Taken together, these studies indicate a major extracerebellar source for cerebellar oligodendrocytes and the presence in vivo of distinct gliogenic subsets for astrocytes and oligodendrocytes.
Differentiation of Cerebellar Oligodendrocytes
Once in the cerebellar parenchyma, oligodendrocyte precursors first occupy the deepest region surrounding the cerebellar nuclei and progressively enter the developing lobules, where they settle both in the intralobular white matter and in the cortical layers. Afterward, they undergo maturation according to the same centrifugal spatiotemporal pattern of dispersion in the cerebellar tissue (Reynolds and Wilkin 1988). Accordingly, myelin is deposited from the deepest to the outermost layers (Reynolds and Wilkin 1988; Gianola et al. 2003) until the end of the second postnatal week in rodents. Despite the mechanisms responsible for this peculiar centrifugal pattern still need to be clarified, it is likely that this precise spatiotemporal schedule results from the proximo-distal gradient of PC maturation along the outgrowing cortical lobules (Gianola et al. 2003).
As in other regions of the central nervous system, both cell-autonomous mechanisms and environmental factors cooperate in oligodendrocyte maturation in the cerebellum. Yet, some factors specifically impact on cerebellar oligodendrogenesis. Thyroid hormones, for instance, at early postnatal stages, promote oligodendrocyte progenitor cell cycle exit and commitment to differentiation through indirect actions exerted by PCs and astrocytes. Conversely, at later mature phases, through cell-autonomous mechanisms, they directly restrain cell proliferation of oligodendroglial progenitors (Picou et al. 2012). Furthermore, PCs influence oligodendrocyte functioning by release of Shh. During early postnatal development, high amounts of PC-derived Shh stimulate the proliferation of oligodendrocyte progenitors. Then, by the end of the first postnatal week, PCs downregulate Shh and start producing vitronectin, which promotes oligodendrocyte maturation (Bouslama-Oueghlani et al. 2012). Moreover, recent evidence highlighted the role of GABAergic signaling from interneurons in limiting proliferation and stimulating differentiation of oligodendrocyte precursors in the postnatal cerebellum (Zonouzi et al. 2015). These effects were attributed to GABAA receptor-mediated synaptic inputs to oligodendroglial cells. Whether and how other neurotransmitters participate in oligodendrocyte differentiation in the cerebellum remains to be established.
Conclusions and Future Directions
The findings discussed above witness the increasing interest of the scientific community toward the understanding of the mechanisms of astroglial ontogenesis and differentiation in the cerebellum. Nevertheless, many features of this process still need clarification. Indeed, while it is now well accepted that cerebellar astrocytes are primarily generated from the embryonic VZ through an intermediate phase of amplification in the PCL and PWM, the precise contribution of these postnatal germinative site to astroglial diversity still needs to be elucidated. Similarly, possible lineage relationships between cerebellar astroglial phenotypes have not been addressed. Clonal analyses at single progenitor level may clarify both these points. Furthermore, much remains to be understood on the regulatory mechanisms, likely including epigenetic modifications, that drive the specification and maintenance of the diverse astrocyte phenotypes. In this respect, studies on the well-established heterogeneity of cerebellar astroglia can reveal fundamental mechanisms, exploitable as a reference point to further chart astrogliogenesis in other brain areas, where diversity in astroglia is far less evident and understood.
On the other side, many are the open questions on cerebellar oligodendrocytes. The extra-/intracerebellar origin of oligodendrocytes waits to be ultimately and unequivocally demonstrated by time-lapse experiments and/or through abrogation of oligodendrogenesis in regions outside the cerebellum. Moreover, possible selective preference of cerebellar or extracerebellar oligodendrocytes for defined axon types or compartments have not been investigated so far. Similarly, it remains to be understood whether oligodendrocyte progenitors in the molecular layer, where very little myelin is present, engage in functions different from being a source of myelinating oligodendrocytes through the interaction with neurons and BG (Boda and Buffo 2014).
This work was funded by local grants of the University of Turin. VC was partly supported by a FENS fellowship.
- Altman J, Bayer SA (1997) Development of the cerebellar system in relation to its evolution, structure and functions. CRC, Boca RatonGoogle Scholar
- Bouslama-Oueghlani L, Wehrlé R, Doulazmi M, Chen XR, Jaudon F, Lemaigre-Dubreuil Y, Rivals I, Sotelo C, Dusart I (2012) Purkinje cell maturation participates in the control of oligodendrocyte differentiation: role of sonic hedgehog and vitronectin. PLoS One 7:e49015. https://doi.org/10.1371/journal.pone.0049015CrossRefPubMedPubMedCentralGoogle Scholar
- Cerrato V, Parmigiani E, Figueres-Oñate M, Betizeau M, Aprato J, Nanavaty I, Berchialla P, Luzzati F, de’Sperati C, Lopez-Mascaraque L, Buffo A (2018a). Multiple origins and modularity in the spatiotemporal emergence of cerebellar astrocyte heterogeneity. PLOS Biology 16(9):e2005513. https://doi.org/10.1371/journal.pbio.2005513CrossRefGoogle Scholar
- Cerrato V, Mercurio S, Leto K, Fucà E, Hoxha E, Bottes S, Pagin M, Milanese M, Ngan CY, Concina G, Ottolenghi S, Wei CL, Bonanno G, Pavesi G, Tempia F, Buffo A, Nicolis SK (2018b). Sox2 conditional mutation in mouse causes ataxic symptoms, cerebellar vermis hypoplasia, and postnatal defects of Bergmann glia. Glia 66(9):1929–1946. https://doi.org/10.1002/glia.23448CrossRefGoogle Scholar
- D’Arca D, Zhao X, Xu W, Ramirez-Martinez NC, Iavarone A, Lasorella A (2010) Huwe1 ubiquitin ligase is essential to synchronize neuronal and glial differentiation in the developing cerebellum. Proc Natl Acad Sci U S A 107:5875–5880. https://doi.org/10.1073/pnas.0912874107CrossRefPubMedPubMedCentralGoogle Scholar
- Fañanas JRY (1916) Contribucion al estudo de la neuroglia del cerebelo. Trab Lab Invest Biol (Madrid) 14:163–179Google Scholar
- Farmer WT, Abrahamsson T, Chierzi S, Lui C, Zaelzer C, Jones EV, Bally BP, Chen GG, Théroux J-F, Peng J, Bourque CW, Charron F, Ernst C, Sjöström PJ, Murai KK (2016) Neurons diversify astrocytes in the adult brain through sonic hedgehog signaling. Science 351:849–854. https://doi.org/10.1126/science.aab3103CrossRefPubMedGoogle Scholar
- Florio M, Leto K, Muzio L, Tinterri A, Badaloni A, Croci L, Zordan P, Barili V, Albieri I, Guillemot F, Rossi F, Consalez GG (2012) Neurogenin 2 regulates progenitor cell-cycle progression and Purkinje cell dendritogenesis in cerebellar development. Development 139:2308–2320. https://doi.org/10.1242/dev.075861CrossRefPubMedPubMedCentralGoogle Scholar
- Frick A, Grammel D, Schmidt F, Poschl J, Priller M, Pagella P, von Bueren AO, Peraud A, Tonn J-C, Herms J, Rutkowski S, Kretzschmar HA, Schuller U (2012) Proper cerebellar development requires expression of beta1-integrin in Bergmann glia, but not in granule neurons. Glia 60:820–832. https://doi.org/10.1002/glia.22314CrossRefPubMedGoogle Scholar
- Gupta I, Collier PG, Haase B, Mahfouz A, Joglekar A, Floyd T, Koopmans F, Barres B, Smit AB, Sloan SA, Luo W, Fedrigo O, Ross ME, and Tilgner HT (2018) Single-cell isoform RNA sequencing characterizes isoforms in thousands of cerebellar cells. Nature Biotechnology 36:1197–1202. https://doi.org/10.1038/nbt.4259CrossRefGoogle Scholar
- Hashimoto R, Hori K, Owa T, Miyashita S, Dewa K, Masuyama N, Sakai K, Hayase Y, Seto Y, Inoue YU, Inoue T, Ichinohe N, Kawaguchi Y, Akiyama H, Koizumi S, Hoshino M (2016) Origins of oligodendrocytes in the cerebellum, whose development is controlled by the transcription factor, Sox9. Mech Dev 140:25–40. https://doi.org/10.1016/j.mod.2016.02.004CrossRefPubMedGoogle Scholar
- Hoshino M, Nakamura S, Mori K, Kawauchi T, Terao M, Nishimura YV, Fukuda A, Fuse T, Matsuo N, Sone M, Watanabe M, Bito H, Terashima T, Wright CVE, Kawaguchi Y, Nakao K, Nabeshima YI (2005) Ptf1a, a bHLH transcriptional gene, defines GABAergic neuronal fates in cerebellum. Neuron 47:201–213. https://doi.org/10.1016/j.neuron.2005.06.007CrossRefPubMedGoogle Scholar
- Leto K, Bartolini A, Di Gregorio A, Imperiale D, De Luca A, Parmigiani E, Filipkowski RK, Kaczmarek L, Rossi F (2011) Modulation of cell-cycle dynamics is required to regulate the number of cerebellar GABAergic interneurons and their rhythm of maturation. Development 138:3463–3472. https://doi.org/10.1242/dev.064378CrossRefPubMedGoogle Scholar
- Parmigiani E, Leto K, Rolando C, Figueres-Oñate M, López-Mascaraque L, Buffo A, Rossi F (2015) Heterogeneity and bipotency of astroglial-like cerebellar progenitors along the interneuron and glial lineages. J Neurosci 35:7388–7402. https://doi.org/10.1523/JNEUROSCI.5255-14.2015CrossRefPubMedGoogle Scholar
- Ramon y Cajal S (1911) Histologie du Système Nerveux de l’Homme et des Vertébrés. Maloine, ParisGoogle Scholar
- Ramon y Cajal S (1926) Beitrag zur Kenntnis der Neuroglia des Groß- und Kleinhirns bei der progressiven Paralyse mit einigen technischen Bemerkungen zur Silberimprägnation des pathologischen Nervengewebes. Z Neurol 100:738–793Google Scholar
- Satz JS, Barresi R, Durbeej M, Willer T, Turner A, Moore SA, Campbell KP (2008) Brain and eye malformations resembling Walker-Warburg syndrome are recapitulated in mice by dystroglycan deletion in the epiblast. J Neurosci 28:10567–10575. https://doi.org/10.1523/JNEUROSCI.2457-08.2008CrossRefPubMedPubMedCentralGoogle Scholar
- Schmid RS, McGrath B, Berechid BE, Boyles B, Marchionni M, Sestan N, Anton ES (2003) Neuregulin 1-erbB2 signaling is required for the establishment of radial glia and their transformation into astrocytes in cerebral cortex. Proc Natl Acad Sci U S A 100:4251–4256. https://doi.org/10.1073/pnas.0630496100CrossRefPubMedPubMedCentralGoogle Scholar
- Sekerková G, Ilijic E, Mugnaini E (2004) Time of origin of unipolar brush cells in the rat cerebellum as observed by prenatal bromodeoxyuridine labeling. Neuroscience 127:845–858. https://doi.org/10.1016/j.neuroscience.2004.05.050CrossRefPubMedGoogle Scholar
- Wang X, Imura T, Sofroniew MV, Fushiki S (2011) Loss of adenomatous polyposis coli in Bergmann glia disrupts their unique architecture and leads to cell nonautonomous neurodegeneration of cerebellar Purkinje neurons. Glia 59:857–868. https://doi.org/10.1002/glia.21154CrossRefPubMedPubMedCentralGoogle Scholar