Evolutionarily Conserved Concepts in Glial Cell Biology

Abstract

The evolutionary conservation of glial cells has been appreciated since Ramon y Cajal and Del Rio Hortega first described the morphological features of cells in the nervous system. We now appreciate that glial cells have essential roles throughout life in most nervous systems. The field of glial cell biology has grown exponentially in the last ten years. This new wealth of knowledge that has been aided by seminal findings in non-mammalian model systems. Ultimately, such concepts help us to understand glia in mammalian nervous systems. Rather than summarizing the field of glial biology, I will first briefly introduce glia in non-mammalian models systems. Then, highlight seminal findings across the glial field that utilized non-mammalian model systems to advance our understanding of the mammalian nervous system. Finally, I will call attention to some recent findings that introduce new questions about glial cell biology that will be investigated for years to come.

Glia are essential for nervous system functionality. Research in recent years have identified evolutionarily conserved cell biological concepts that help us understand the mammalian nervous system[1]. Here, I introduce glia in non-mammalian model systems and then outline a few discoveries in non-mammalian systems that advanced our current understanding of astroglia and oligodendrocytes in mammals (Figure 1).

Figure 1.

Evolutionarily Conserved Glial Concepts. A. Schematic showing Draper/Megf10 dependent astrocytic phagocytosis of neuronal components. Inset that is outlined by blue box shows neuronal corpses (green) and synapses (blue) that are cleared by astrocytes. B. Schematic showing microdomain (blue) and cell soma (green) calcium transients in astrocytes. Such calcium transients require distinct Trp channels. C. Schematic demonstrating FGF receptors are required for the complex morphology of astrocytes. Note the reduction of branches (blue) in FGFR knockouts (FGFR KO). D. Illustration of two oligodendrocyte progenitor cells in the process of contact-dependent repulsion, which is used to space oligodendrocytes in the nervous system. The depiction of evolution is an illustration and does not necessary represent the evolutionarily path of depicted organisms.

Introduction to Glia in Non-Mammalian Systems

C. elegans have been an important model system to investigate the basic cell biology of the nervous system. Ever since the complete reconstruction from electron microscopy and lineage map of all cells, it has been appreciated that C. elegans have glia-like cells[2]. We now know that the C. elegans nervous system is composed of 302 neurons and 50 glial cells[2,3]. Single glial cells in C. elegans can serve multiple roles which would be individually performed by multiple different cells in the vertebrate nervous system. The genetic tractability of C. elegans and near identical cellular-anatomy between individuals, has made the model system invaluable for interrogating basic cell biological concepts in glial populations like phagocytosis, homeostatic support and neurodevelopment [3–5].

Drosophila melanogaster have also been integral in our understanding of the basic cell biology of glia. In part, this is because the Drosophila model system is genetically tractable and has an excellent set of tools (e.g. UAS(upstream activation sequence):RNAi libraries) to test the cell-autonomous roles of genes in a biological processes[6]. Glia account for likely 5%−10% of the cells in the fly central nervous system (CNS)[7] and we know that distinct glial subtypes are present in Drosophila (e.g. astrocytes, cortex glia, ensheathing glia), many of which share conserved functional roles with mammalian glia. The astrocytic subtype, for example, shares many properties with mammalian astrocytes[7–9]. Unlike in the mammalian nervous system, Drosophila glia do not myelinate. However, investigations of Drosophila wrapping glia have proven to be a powerful system to probe the basic mechanisms of glial ensheathment that is essential for myelination [10,11].

Finally, Danio rerio (zebrafish) have been a key model system for understanding glial cell biology[1,12]. Zebrafish have all the major glial cell-types that are present in mammals including Schwann cells, satellite glia, enteric glia, microglia, astrocytes and oligodendrocytes[12–17]. Myelinating glia are also abundant and essential in the zebrafish nervous system. Thus, the cell biology of myelination can be pursued in zebrafish[12]. The zebrafish system is highly amenable to time-lapse imaging, providing valuable information on the temporal and spatial dynamics of glia over the course of development. Transgenes that label every glial subtype in vertebrates are also now available in zebrafish, and combined with genetic screens, the zebrafish model system has been a major source for new concepts in glial biology.

Here, I outline examples of historical studies in non-mammalian systems that revealed some of the new principles of glial cell biology in mammalian systems (Figure 1).

Historical Discoveries that Provided Insight about Mammalian Glia

Astrocyte phagocytosis.

One near universal principle of nervous system development is that cells or parts of cells are overproduced and then pruned back. During this pruning processes, neural debris is present in the nervous system. The basic cell biology of clearance of debris likely extends beyond the presence of debris itself, as synapses for example are pruned. Neural debris is also cleared after injury. To investigate the clearance of neural debris and potential role of glia in that process, several different processes have been studied in Drosophila. First, expression analysis of genes enriched in glial cells of Drosophila and functional analysis identified Draper is highly expressed in glia and necessary for glia-dependent clearance of neuronal debris[9]. Insight from an injury model in which olfactory receptor neurons are severed, further revealed that astrocytes phagocytose severed cell corpses after injury, a process that also required Draper[18]. Additionally, by tracking astrocytes during metamorphosis, it was identified that synaptic material is specifically cleared by astrocytes with Crk/Mbc/dCed-12 and Draper. Interestingly, Crk/Mab/dCed-12 is required for clearance of axonal debris but Draper appears to be dispensable, revealing that astrocytes utilize distinct genetic mechanisms to phagocytose specific debris (Figure 1A)[19,20].

It is now appreciated that mammalian astrocytes utilize similar molecular processes to phagocytose neuronal debris. The investigation of this, ten years after the first Drosophila findings, was undertaken in the visual system of mice, which undergoes synaptic pruning during development. Megf10 is expressed in astrocytes in the dorsal lateral geniculate nucleus where synaptic removal occurs. Genetic mutants of Megf10 (the homologue of Draper) and Mertk resulted in a substantial reduction in the amount of synaptic debris engulfed by astrocytes[21]. This lack of clearance has functional consequences to the visual circuit, resulting in abnormal innervation. Megf10 and Jedi1 (another engulfment receptor) also function in peripheral nervous system (PNS) glia to clear neuronal corpses[22]. Interestingly, C. elegans AMsh glial cells also utilize a GEF complex (Ced2/CrkII, Ced-12) and CED-1/Draper/Megf10 to clear neuronal debris[4]. Collectively, this work highlights an evolutionarily conserved process of neuronal debris clearance by glia that is dependent on Ced-1/Draper/Megf10.

Astrocyte Calcium Transients.

Recent years have shown that glial cells, like neurons, exhibit calcium transients (Figure 1B), but the biological relevance of these transients remains unclear. To understand the importance of astrocytic Ca²⁺ transients, Ma et al. investigated them in Drosophila using a library of RNAi targeting 500 Ca²⁺ signaling-related genes with astrocyte-specific Gal4 drivers. This investigation revealed that the TRP channel, Wrtw, was important[23]. The functional importance of these Ca²⁺ transients became clear when Wrtw knockdown in astrocytes caused changes to neuronal chemotactic responses. Genetic analysis demonstrates that Tyr receptors, which respond to neuronal activity, and Wrtw function in the same genetic pathway in astrocytes. These genes regulate the neuronal activity-dependent astrocytic Ca²⁺ transients that control dopaminergic neurons critical to the behavioral response[23].

We now also know that astrocytes exhibit distinct subtypes of Ca²⁺ transients including cell soma and microdomain transients. More Drosophila studies have demonstrated that these subtypes are regulated by distinct genetic components. Wtrw regulates cell soma transients, but microdomain activity can be modulated by a different Trp channel, TrpML[24]. While the relevance of these microdomain transients is not entirely clear, their disruption in Drosophila via manipulation of TrpML caused structural defects in tracheal filopodia growth.

Mammalian astrocytes also exhibit soma and microdomain Ca²⁺ transients and in vitro studies of astrocytes demonstrated that Ca²⁺ transients are also mediated by a Trp channel, TRPA1. In slice culture, modulating astrocytic Ca²⁺ via TRPA1 caused changes in mIPSC amplitudes in interneurons, suggesting that the transients have functional relevance[25]. Recent work also revealed that zebrafish astrocytes experience similar microdomain and cell soma Ca²⁺ transients[13]. The functional roles of these different Ca²⁺ transients and the unique genetic components that modulate them are still an area of ongoing research. Work in Drosophila has already been integral to our molecular understanding of these transients and likely will drive further exploration in the future.

Astrocyte Morphogenesis.

Astrocytes have ramified morphologies that interact with neuronal synapses, cell bodies, and dendritic branches. Like mammalian protoplasmic astrocytes, Drosophila astrocytes in the larval ventral nerve cord display a dense network of processes. Analysis of these astrocytes demonstrated that they require the FGF receptor (Htl) to develop their mature arbors (Figure 1C)[26]. Experiments support the idea that Htl functions as a mostly permissive cue, but there could also be an instructive aspect that directs specific branches or cell bodies during astrocyte morphogenesis. Overall, the work in Drosophila demonstrates that FGF is important for astrocytic morphology.

Work in zebrafish now also supports the hypothesis that FGF controls astrocyte morphology. With a new appreciation that zebrafish have astrocytes comparable to other model systems, Chen et al. investigated the cellular, developmental, and molecular attributes of vertebrate astrocytes. Like Drosophila astrocytes, the morphogenesis of zebrafish astrocytes also requires FGF receptors (Fgfr3 and Fgfr4)[13]. While the specific role of FGF in mammalian astrocytes in vivo is not known, it is clear that FGF can be used in cell culture to direct stem cells (and maybe other cell types in vivo) to generate astrocytes[27,28]. Research also suggests that FGF signaling is important in reactive astrocytes to influence morphology[29]. This conserved role of FGF signaling in astrocyte morphology from Drosophila to zebrafish supports the idea that additional discoveries of molecular cascades controlling astrocyte morphology in non-mammalian model systems can help build a blueprint for how mammalian astrocytes are constructed.

Contact Dependent Repulsion of Oligodendrocytes.

It is widely accepted that glia and neurons exhibit an evolutionarily conserved phenomenon known as tiling, allowing glia to be spaced in non-overlapping areas in the nervous system. This applies to astrocytes, oligodendrocytes and microglia in the CNS[30,31]. Seminal work in zebrafish revealed that contact-dependent repulsive mechanisms promote the spacing of oligodendrocytes in the CNS (Figure 1D). Timelapse imaging of labeled oligodendrocytes progenitor cells (OPCs) showed that when two OPCs contact each other, they repel in opposite directions[32]. Additionally, neighboring OPCs extended into regions where OPCs were experimentally ablated, a process that mimics historical data gathered from studies of contact-dependent repulsion of neurons[30]. This observation of dynamic OPC interactions during development demonstrated that oligodendrocyte organization is dependent not only on neurons but also other oligodendrocytes.

The idea that OPC growth is dependent on interactions with other OPCs was confirmed in mice 6 years later. With timelapse imaging of reporters that labeled OPCs in the mouse brain, it was revealed that even in adult brains, oligodendrocyte processes are dynamic[33]. Like in the zebrafish spinal cord, adult cortical OPCs exhibit repulsion after contact with neighboring OPCs. These observations also revealed that processes of the same OPC also repel each other, a process similar to the self-avoidance behavior seen between neighboring neuronal dendrites[31]. Finally, ablation of OPCs caused neighboring cells to invade and divide. These results are consistent with findings in zebrafish and point to an evolutionarily conserved process controlling spatial organization of oligodendrocytes. While it is now widely accepted that contact-dependent mechanisms space glia, the molecular underpinnings of this process still remain relatively unknown[31,34,35]. Given that glial tiling is seen throughout evolution, non-mammalian models will be critical to investigate the molecular components of glial tiling.

Temporal Restriction of Myelination.

Oligodendrocytes are the myelinating cells of the central nervous system. The field collectively agrees that myelination continues throughout life and that this continued myelination is critical for learning. Continued myelination is also relevant in the context of demyelinating diseases, where remyelination can be a critical factor in disease progression and prognosis. New myelin sheaths in the nervous system could be generated by either differentiating oligodendrocytes or by existing oligodendrocytes. The idea that oligodendrocyte sheath formation could be regulated at different stages of maturation was supported by results demonstrating that cultured immature oligodendrocytes make fewer myelin sheaths than oligodendrocytes produced from oligodendrocyte progenitor cells[36]. These results imply that oligodendrocyte maturation is inversely correlated with their ability to generate new myelin sheaths but the temporal scale of this limitation needs more investigation.

In the zebrafish spinal cord, such temporal limitation of new sheath formation can be observed with time-lapse imaging paradigms in transgenic animals because the entirety of the oligodendrocyte differentiation can be visualized with impressive temporal resolution. Consistent with the idea that the differentiation state of oligodendrocytes impacts new myelin sheath formation, the number of myelin sheathes per each zebrafish oligodendrocytes remains constant after ~5 hours from the initial formation of myelin sheathes[37]. The authors support this idea by showing that despite increased myelination in the animal through manipulation of Fyn kinase, the temporal restriction of myelin sheath formation persisted[37]. In the future, it will be important to understand the molecular mechanisms that control this evolutionarily conserved restriction of sheath formation.

Emerging Directions

Astrocyte dependent critical periods.

Critical periods are intervals of time during development in which neural circuits can be altered by activity. After a critical period has closed, dendritic branches can no longer be modulated by changes in neuronal activity. It has been known for years that critical periods are important for proper neural circuit assembly but the underlying mechanisms that regulate critical periods, as well as the role of glial cells, are poorly understood. Using Drosophila, Ackerman et al. first identified a critical period during the formation of the Drosophila motor circuit [38]. Supporting the role of glia in closing the critical period, ablation of astrocytes resulted in an extension of the critical period. A RNAi knockdown screen identified a mechanism by which astrocytic Nlg controls the critical period via signaling to Nrx-1 (Neurexin) on the neurons, stabilizing microtubules to alter dendrite morphology[38]. Work in the mouse visual cortex also supports the idea that astrocytes function in closing critical periods[39]. Whether astrocytes (or other glia) function in all critical periods remains to be determined, but the tools in non-mammalian models will provide more targeted approaches that can be tested in other vertebrates.

Paranodal Bridge.

It is well known that oligodendrocytes generate multiple sheathes. These sheathes were thought to extend from branches that connect to the cell body. However, recent evidence from a preprint suggests that this idea may need to be updated. Visualizing oligodendrocytes in the mouse and then later confirming the observation in zebrafish, researchers have revealed small cellular bridges that connect adjacent myelin sheathes, termed paranodal bridges[40]. The description of paranodal bridges opens up additional questions in the field of glial biology. For example, the underlying cell biology and molecular mechanisms that allows one myelin sheath to create a paranodal bridge and the mechanism by which a paranodal bridge could aid in the maintenance of the distal sheaths are still unknown. Imaging of both mammals and zebrafish can help to identify the cell biology underlying these cellular bridges.

Nexus Glia.

It is well known that astrocytes, oligodendrocytes and microglia are integral to vertebrate CNS development and function[1]. Despite the important specialized roles of each of these glia, current research understands the peripheral nervous system to be largely devoid of cells like astrocytes and microglia. Recent work, however, indicates that astroglia-like cells in the PNS could be more abundant than previously thought. In zebrafish, gfap⁺/glast⁺ cells, termed nexus glia, populate the outflow tract of the heart during embryonic stages and develop complex morphologies that mimic astrocytes in the zebrafish CNS[41]. Distinct from astrocytes, nexus glia are derived from the neural crest. Analysis of single cell RNA sequencing analysis revealed that transcripts typically present in astrocytes are also found in human and mouse heart glial cells. The abundance of nexus glia is dependent on Metrn, a gene implicated in astrocyte development[41]. The discovery of nexus glia, along with previous studies of enteric glia[42], points to a potential critical role of astroglia-like cells in the PNS. Such astroglia-like cells have also been described in the spleen[43]. The presence of astroglia-like cells in other organs, and their biological relevance, is still unclear. However, these populations, at least nexus glia, can now be investigated in zebrafish to develop more targeted approaches in mammalian systems.

Concluding Remarks

Non-mammalian model systems have been critical in our understanding of glial cell biology. Further investment in these systems will be important to reveal emerging concepts that aid in the understanding of the mammalian nervous system.