Glial progenitor heterogeneity and plasticity in the adult spinal cord

Glial progenitor cells were reported to have the capacity to generate various types of cells in both the central nervous system (CNS) and peripheral nervous system. Glial progenitor cells can respond to diverse environmental signals and transform into distinct populations, each serving specific functions. Notably, the adult spinal cord hosts various populations of glial progenitors, a region integral to the central nervous system. During development, glial progenitors express glial fibrillary acidic protein (GFAP; Dimou and Gotz, 2014). However, the specific identities of GFAP-expressing progenitors in the adult spinal cord were not thoroughly investigated. Glial progenitors in the adult spinal cord may hold more heterogeneity and plasticity than previously thought. GFAP-expressing cell responses in the course of spinal cord injury: Traditionally, glial progenitors were thought to only produce glial cells but not neurons, while radial glial cells possessing the capacity to differentiate into neuronal and macroglial precursors including astrocyte progenitor cells, oligodendrocyte progenitor cells and ependymal cells. Recent findings in the adult spinal cord provided evidence for remarkable plasticity and heterogeneity of glial progenitors in the adult CNS. Astrocyte activation plays a pivotal role in the response to CNS injuries and the progression of neurodegenerative diseases. The analysis of RNA sequencing (RNA-seq) datasets obtained from both spinal cord tissue and purified GFAP-expressing cells provides valuable insights into the genome-wide changes in gene expression within the injury site and sheds light on the specific reactions and contributions of astrocytes in this context (Wei et al., 2021). Recent research in the brain and spinal cord has unveiled a greater level of heterogeneity among GFAP-expressing cells than previously anticipated (John Lin et al., 2017; Milich et al., 2021). However, the understanding of the dynamics of these cells following spinal cord injury (SCI) remains limited. In recent studies, single-cell RNA sequencing (scRNA-seq) was performed using GFAP-expressing cells from sub-chronic spinal cord injury models to uncover various subpopulations, and the data were compared with those observed in acute-stage data (Milich et al., 2021; Wei et al., 2023). These subpopulations exhibit specific molecular signatures, morphological features, and functional enrichments and are characterized by subpopulation-specific regulons that define their identities. In the sub-chronic stages, cells displaying both astrocyte (e.g., Gfap) and oligodendrocyte progenitor cells (e.g., Pdgfra) markers (referred to as intermediate cells) comprised five distinct subpopulations. Functional enrichment revealed that these cells were enriched in neuronal, proliferative, inflammatory, oligodendrocyte, or myeloid cell markers. Additionally, cells expressing both astrocyte and ependymal markers were referred to as astrocyte-ependymal cells, showing functional enrichment such as motile cilia and response to interferon gamma. One subpopulation in the sub-chronic stages only expressed astrocyte marker genes (referred to as cells with only astrocyte markers). Its functional enrichments include astrocyte projection and calcium ion transmembrane import into the cytosol. All the subpopulations detected in sub-chronic stages were also present in acute stages, although the percentages of cells in each subpopulation varied between acute and sub-chronic stages. For instance, astrocyte-ependymal cells were detected in both stages, but those with Nes were more prevalent in the acute stages. These studies have also revealed that the cellular heterogeneity of astrocyte-lineage cell subpopulations was driven by transcription factors and regulons (Wei et al., 2023). For instance, in the sub-chronic stages, the astrocyte-ependymal cell population was defined by astrocyte, ependymal and stemness -related regulons, such as Sox2, Sox9, Foxj1, and Pax6. Proliferating intermediate cells contained proliferation-related regulons (e.g., E2f1). Potential of intermediate cells in transitioning to other cell types: A recent study has documented the conversion of astrocytes into neurons within the adult spinal cord (Tai et al., 2021). Although many have relied on the ectopic expression of specific key factors or pharmacological approaches to induce the transformation of resident astrocytes into neurons, this raises the question of whether astrocyte lineage inherently contains subpopulations capable of differentiating into other cell types. Intermediate cells, as identified in the recent study (Wei et al., 2023), expressed both astrocyte and oligodendrocyte progenitor cell markers (Gfap+/Pdgfra+), distinguishing them from NG2 (Cspg4 expressing) glia. There may be an overlap between Intermediate cells and NG2 population but they are not identical. During brain development, a progenitor population known as B cells expresses PDGFRa and GFAP in the subventricular zone. These B cells have the capacity to give rise to transient amplifying neural progenitor cells that produce neuroblasts and differentiate into both neurons and glial cells in vivo. However, a similar cell type was not reported in the adult spinal cord. Intermediate cells may have the potential to differentiate into various cell types based on the above-mentioned functional enrichment analysis. Trajectory and velocity analysis revealed that these intermediate cells exhibited increased dynamic behavior following injury and had the potential to transition into other cellular states. Histological analysis demonstrated that intermediate cells were present in Naive, Sham, and injured groups, although they were relatively rare in Naive samples, and were distributed in both gray matter and white matter. These cells displayed distinct morphological characteristics. In Naive or Sham samples, they presented with a progenitor-like shape, featuring a round soma and short processes, or exhibited a stellate astrocyte- or fibrous astrocyte-like morphology with distended bodies in some cases, and long filamentous processes after injury. Astrocyte-ependymal cell heterogeneity: Emerging from radial glia during development, ependymal cells are ciliated cells forming the ependyma—an epithelial barrier. Ependymal cells within the spinal cord play a pivotal role in the CNS. They were reported to line the cavities of the CNS, encompassing the central canal of the spinal cord and the brain's ventricles (Walker and Echeverri, 2022). In the context of spinal cord injury, ependymal cells can become actively involved in the response to injury, contributing to various cellular and molecular changes within the injured area. A previous study has explored the potential of ependymal cells to participate in the regeneration and repair of spinal cord tissue following injury (Llorens-Bobadilla et al., 2020). Investigations have revealed that ependymal cells are not a homogenous population and contain several distinct subpopulations. For example, Llorens-Bobadilla et al. identified three astrocyte-ependymal subpopulations (AE1, AE2, and AE3) (Llorens-Bobadilla et al., 2020). Their findings demonstrated the potential of astrocyte-ependymal cells to differentiate into astrocytes and oligodendrocytes. This group also reported that ependymal cells possess multipotent characteristics, becoming activated and proliferating after SCI. A recent study identified astrocyte-ependymal cells marked by the colocalization of FOXJ1, SOX9, and GFAP markers. Four astrocyte-ependymal cell subpopulations with varying proliferation potentials after acute SCI (Wei et al., 2023) resulted from scRNA-seq analyses. The S+G2M cell-cycle score of astrocyte-ependymal subpopulations indicated that only astrocyte-ependymal-Neshigh were significantly increased, and trajectory analysis suggested that these astrocyte-ependymal-Neshigh had a high potential for differentiation into other subtypes. A previous study predominantly focused on examining ependymal cells within the central canal. Notably, recent research (Wei et al., 2023) revealed that these cells could be detected in both gray matter and white matte. The results also indicated that astrocyte-ependymal cells were not exclusive to SCI at 7 days post-injury (7 dpi) and 1-month post-injury but could also be found in Naive and Sham groups, albeit in low numbers. The abundance of these subpopulations exhibited dynamic changes during the injury process. For example, in Naive samples, astrocyte-ependymal-Neshigh cells were scarce in the spinal cord and increased in number after Sham surgery. Intriguing, these cells are more abundant in white matter compared to gray matter (excluding central canal) in Naive samples, and at the rostral region of Sham samples. The stereological analyses confirm that at 7 dpi, there is a significant increase in astrocyte-ependymal cells observed in both white matter and gray matter. These cells showed similar density in white matter and gray matter, possibly also migrated from the central canal. Furthermore, a panel of 30 factors reported to negatively or positively regulate axon regeneration and synaptogenesis were analyzed. The results indicated that astrocyte-ependymal cells expressed genes of known neurotrophic growth factors, growth promoting extracellular matrix factors, and few growth inhibitors after injury. Cells with only astrocyte markers: ScRNA-seq identified a distinct cluster solely expressing astrocyte markers (cells with only astrocyte markers). Cell proliferation assay showed these cells were significantly increased in 3dpi and 7dpi. Another study reported that cortical astrocytes can transit to an neural stem cell-like state and further to a self-amplifying progenitor-like state after stab brain injury (Zamboni et al., 2020). Basak et al. (2018) observed that quiescent neural stem cells expressing the Troy+ marker (official gene symbol Tnfrsf19) showed an enriched expression of astrocyte markers. These astrocyte markers include Slc1a3, Fgfr3, Agt, Aqp4, and Gja1 etc. Additionally, individual cells with high levels of TroyGFP demonstrated multipotency, indicating their ability to differentiate into neurons, astrocytes, and oligodendrocytes. Consistent with the above-mentioned study, Troy cells expressed Nes and Gfap at 3 dpi after injury (Wei et al., 2023). Examining the expression of Troy in all astrocyte lineage cells showed that cells with only astrocyte markers had the highest Troy expression in acute and sub-chronic stages. Troy is expressed, but to a lesser degree, in astrocyte-ependymal cells. The purified brain cell RNA-seq data (http://jiaqianwulab.org/braincell/RNASeq.html; https://www.brainrnaseq.org/) also showed that Troy was highly enriched in astrocytes (Zhang et al., 2014). RNA-velocity analysis showed high velocity vectors originating from the subpopulation with only astrocyte markers (cluster 3), pointing to cluster 0, which is an intermediate population. These observations align with the hypothesis that astrocytes in the adult spinal cord could possibly undergo dedifferentiation processes and acquire stem-like properties (Gabel et al., 2016). To validate this hypothesis, future in vivo lineage mapping experimentation will be required. Conclusions and future directions: Glial cells in the context of SCI are of particular importance as they play significant roles in the injury response and recovery processes. Bulk RNA-seq provides average gene expression values of the entire sample. Even when isolating cell types using specific markers, the resulting sample can still contain a mixture of different cell types. Recent studies have shown that relying on one or two marker genes is often insufficient for accurate cell type identification. In contrast, the robust technology of scRNA-seq enables the characterization of gene expression at the individual cell level and allows distinguishing cell types by utilizing multiple cell markers and more sophisticated molecular signatures. After spinal cord injury, cells expressing Gfap comprise many subpopulations, and their dynamics change during the injury and recovery process (Figure 1). It is important to note that the Gfap expression is not exclusive to astrocytes and is also expressed in progenitor cells. The analysis of the scRNA-seq data has revealed that these progenitor cell subpopulations exhibit distinct functional enrichments, with their identities defined by subpopulation-specific transcription factors and regulons. Nevertheless, there are still many unanswered questions about these subpopulations, particularly regarding their physiological functions and their communication with other cell types. The isolation of specific subpopulations using multiple markers is essential for studying their specific functions. However, current techniques such as fluorescence-activated cell sorting and genetic mouse models can only accommodate a limited number of markers which makes the downstream functional characterization challenging. Confocal microscopy for visualization also has a limited number of channels. Future technological developments are in need to overcome these limitations and better characterize the cell types/subtypes in finer resolution. Spatial transcriptomics is a new technology that allows researchers to study the gene expression profiles of individual cells within their specific anatomical locations in tissues. Particularly, spatial transcriptomics technology, if done at a single-cell resolution, enables the investigation of cell-cell communication in vivo. Trajectory and velocity analyses showed that the cell types described in this perspective article have a potential to differentiate into other cell types. It is exciting to uncover resident stem cell/progenitor subpopulations in the adult spinal cord, as the GFAP expressing cells sorted from adult spinal cord samples were able to generate O4+ progenitor cells, early neurons, and astrocytes in the presence of differentiation signals (Figure 1). However, the physiological functions of these cell types after SCI are not yet dissected because of some of the above-mentioned changes. Lineage tracing experiments and eliminating or blocking the differentiation of specific subtypes can offer further insights into the sources of heterogeneity of the subpopulations and their derived progenies and functions. Finally, studies to enhance the number or promote dedifferentiation toward neural stem cell/neural progenitor cell-like cells after chronic injury, and combined with the modulation of the signaling pathways that inhibit endogenous neurogenesis in the adult spinal cord could have important implications for enhancing regeneration and plasticity.Figure 1: Heterogeneity and plasticity of GFAP-expressing cells in the adult spinal cord.Single-cell RNA sequencing of GFAP-expressing cells isolated from spinal cord injury models revealed multiple subpopulations (Wei et al., 2023). Notably, these subpopulations exhibited dynamic proliferation potential at different injury stages (uninjured, 1, 3, and 7 days post injury). Additionally, GFAP-expressing cells isolated from adult spinal cord samples demonstrated the ability to generate O4+ progenitor cells, early neurons, and astrocytes in the presence of differentiation signals. The physiological functions of these cell types following spinal cord injury are the subjects of further investigation. Created with BioRender.com. FACS: Fluorescence-activated cell sorting; GFAP: glial fibrillary acidic protein; scRNA-seq: single-cell RNA sequencing.We thank Drs. Jyotirmoy Rakshit and Faiz Ul Amin from The University of Texas Health Science Center at Houston (UTHealth), USA for helpful discussions. This work was supported by grants from the NIH, United States (R01AG078728-01 and R21 NS113068), Amy and Edward Knight Fund-the UTHSC Senator Lloyd and B.A. Bentsen Center for Stroke Research (to JQW). C-Editors: Zhao M, Sun Y, Qiu Y; T-Editor: Jia Y