Skip to main content
NCBI home page
Genome Research logoLink to Genome Research
. 2023 Oct;33(10):1818–1832. doi: 10.1101/gr.277881.123

Single-nucleus transcriptomic mapping uncovers targets for traumatic brain injury

Qiuyun Yang 1,2,8, Lingxuan Zhang 3,8, Manrui Li 1,8, Yang Xu 1, Xiaogang Chen 4, Ruixuan Yuan 3, Xiaofeng Ou 5, Min He 5, Miao Liao 1, Lin Zhang 6, Hao Dai 4, Meili Lv 7, Xiaoqi Xie 5,✉, Weibo Liang 1,✉, Xiameng Chen 4,✉
PMCID: PMC10691476  PMID: 37730437

Abstract

The subventricular zone (SVZ) is a neurogenic niche that contributes to homeostasis and repair after brain injury. However, the effects of mild traumatic brain injury (mTBI) on the divergence of the regulatory DNA landscape within the SVZ and its link to functional alterations remain unexplored. In this study, we mapped the transcriptome atlas of murine SVZ and its responses to mTBI at the single-cell level. We observed cell-specific gene expression changes following mTBI and unveiled diverse cell-to-cell interaction networks that influence a wide array of cellular processes. Moreover, we report novel neurogenesis lineage trajectories and related key transcription factors, which we validate through loss-of-function experiments. Specifically, we validate the role of Tcf7l1, a cell cycle gene regulator, in promoting neural stem cell differentiation toward the neuronal lineage after mTBI, providing a potential target for regenerative medicine. Overall, our study profiles an SVZ transcriptome reference map, which underlies the differential cellular behavior in response to mTBI. The identified key genes and pathways that may ameliorate brain damage or facilitate neural repair serve as a comprehensive resource for drug discovery in the context of mTBI.

Mild traumatic brain injury (mTBI) is the most common type of traumatic brain injury (Silverberg et al. 2020). A single mTBI can trigger pathophysiological changes in the brain, resulting in acute neurological dysfunction and a range of subsequent postacute or chronic sequelae (Mychasiuk et al. 2014; Collins-Praino et al. 2018). Nevertheless, specific and effective mTBI biomarkers are currently lacking, and the clinical outcomes of patients with mTBI are unsatisfactory.

The subventricular zone (SVZ) is a neurogenic niche in the adult brain responsible for maintaining tissue homeostasis and recovery after brain stimulation (Chavali et al. 2018). This neurogenic niche is an intricate microenvironment comprising morphologically and molecularly distinct subtypes of neural stem cells (NSCs) that give rise to specific progenitor cell types committed to either neuronal or glial differentiation (Doetsch et al. 1999; Buono et al. 2013). Upon brain injury, NSCs in the SVZ undergo reprogramming to induce neurogenesis (Llorens-Bobadilla et al. 2015). Although progress has been made in characterizing NSC proliferation and differentiation during neurogenesis, their diversity, responsive heterogeneity, and governing factors are not entirely clear. In addition, the number of newly generated neurons is insufficient to fully compensate for the loss of neurons caused by the injury (Cao et al. 2002). Therefore, it is essential to comprehend the diversity and potential of niche cells, their dynamics upon injury, and the underlying principles of their regulation to elucidate mTBI pathogenesis and identify relevant diagnostic or therapeutic targets.

Although recent advancements in snRNA-seq have contributed to unraveling the heterogeneity of the SVZ under physiological circumstances (Zywitza et al. 2018; Xie et al. 2020), the impact of mTBI on SVZ remains unexplored, leaving a critical void in the understanding of the postinjury dynamics within this region. The primary purpose of our study is to use droplet-based snRNA-seq to examine cellular and molecular heterogeneity within the SVZ niche under both normal and mTBI conditions. The data obtained from this study will serve as a crucial resource for investigating adult neurogenesis and its governing mechanisms. Additionally, we aim to leverage this data resource to explore and validate novel targets that have the potential to enhance neurogenesis after mTBI.

Results

Single-cell profiling of the entire adult SVZ niche

In this study, we investigated how the adult SVZ changes at the single-cell level under mTBI (Fig. 1A). The neurological severity score (NSS) (Fig. 1B) revealed a significant difference between the CCI and sham-operated group 1 h postinjury. The NSS results, consistent with previous mTBI studies, indicated mild brain impairment (Fig. 1B; Huynh et al. 2020). To elucidate the specific gene expression features of each cell type in the SVZ niche under CCI insult, we profiled single-nucleus sequencing of matched SVZs from three mTBI mice and three sham-operated controls (Fig. 1C). Specifically, the lateral ventricular wall of each mouse was microdissected and digested into individual cells, which were then suspended using an optimized protocol without cell sorting to best preserve in vivo conditions (Biancalani et al. 2021). We used a chromium controller chip to load the pooled samples, barcode each cell, prepare an mRNA library, and perform sequencing. Following quality-control filtering, 15,265 single cells with an average of 6948 detected unique molecular identifiers and an average of 2542 genes in each cell were retained for subsequent analysis (Supplemental Fig. S1A).

Figure 1.

Figure 1.

Open in a new tab

Major cell types in adult SVZ and cell type–specific markers based on single-cell profiling. (A) Workflow of sample preparation, single-nucleus RNA-seq, and the downstream analysis. (B) Neurological severity score (NSS) of the control group and mTBI group. Three mice in each group were included. The t-test was used to determine statistical significance, (**) P ≤ 0.01. (C) Bright-field image of the SVZ; the sampling area is circled by the dotted line. (D) Uniform manifold approximation and projection (UMAP) visualization of 15,265 cells from the SVZ of the control and the mTBI group; 17 cell clusters were detected. (E) Heat map showing differentially expressed genes in each cell type; dot plot showing the gene expression level in the indicated GO terms in each cell type. (F) Violin plots showing the expression level of representative genes that were explicitly expressed in each cell population.

Sequencing data were normalized using SCTransform. Subsequent data dimension reduction and visualization were achieved via PCA and UMAP using the Seurat algorithm (Fig. 1A; Supplemental Fig. S1B; Stuart et al. 2019; https://satijalab.org/seurat/). The 17 designated clusters were classified into seven major cell types using previously known markers (Supplemental Table S1; Luo et al. 2015; Shah et al. 2018; Zywitza et al. 2018; Mizrak et al. 2019). Considering that SVZ astrocytes share many hallmarks with NSCs and are generally believed to function as NSCs in this neurogenic niche (Llorens-Bobadilla et al. 2015), we grouped clusters 4, 5, and 10 together as “NSCs & astrocytes” (Fig. 1D,E; Supplemental Fig. S1C; Supplemental Table S2). Gene Ontology (GO) analysis of differentially expressed genes (DEGs) across cell types revealed biological processes consistent with previously known functions (Fig. 1E), for instance, the terms “dendrite development” in neurons, “myelination” in oligodendrocytes, and “neural precursor cell proliferation” in NSCs & astrocytes. These results further support our classification of the cell clusters.

Based on cell clustering and classification, we identified genes exclusively expressed in individual cell types, which were previously unknown. These cell type–specific genes included Prr5l in oligodendrocytes, Kank1 in oligodendrocyte precursor cells (OPCs), Rorb in NSCs & astrocytes, Fgfr2 in microglia, Mecom in endothelial and mural cells, and Dnah12 in ependymal cells (Supplemental Fig. S2A), which could potentially serve as novel markers for each cell type. We used the SVZ single-cell data set from previous studies to validate these markers and confirm their accuracy (Supplemental Fig. S2B,C; Zywitza et al. 2018; Batiuk et al. 2020). Collectively, this profile represents the first single-nucleus transcriptome map of rodent SVZ under mTBI.

Each SVZ cell type shows transcriptome alterations in response to mTBI

To elucidate the molecular mechanisms underlying mTBI pathogenesis, we compared the cellular and transcriptional differences between the mTBI and control groups (Fig. 2A,B). First, we examined the proportion of each cluster in both groups and validated the results using scCODA (Supplemental Table S3; Büttner et al. 2021). Compared with the control group, scCODA suggested the proportion of neurons in the mTBI group largely decreased, whereas OPCs, oligodendrocytes, and microglia were more abundant (Fig. 2A). These shifts are consistent with reports of neuronal loss (Bu et al. 2016; Holden et al. 2021; Jamjoom et al. 2021) and glial responses postinjury, such as neuroinflammatory responses induced by microglia and demyelination related to oligodendrocyte pathology (Chiu et al. 2016; Furtado et al. 2021; Matson et al. 2022). Notably, clusters 13 and 15, two neuron subtypes, were only observed in the mTBI group (Fig. 2B). Cluster 13 highly expressed the genes Shox2, Slc17a6, and Grid2ip (Supplemental Fig. S3A; Supplemental Table S1), most of which were synapse related (Supplemental Fig. S3A; Dougherty et al. 2013; Matson et al. 2022). The enriched genes in cluster 15 mainly participated in various metabolic and biosynthetic processes (Supplemental Fig. S3B), which suggested a metabolically active state. However, whether these novel mTBI-induced subtypes are newly generated neurons or previously uncharacterized neuronal states/identities still needs further validation.

Figure 2.

Figure 2.

Open in a new tab

Transcription characteristics and heterogeneity of each cell type in response to the mTBI. (A) Bar plot showing the cell count proportion of seven major cell types in the control and mTBI group. (B) Pie chart showing the proportion of cells in 16 clusters in the control and mTBI group. (C) Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of the major cell types. (D) GO analysis showing enriched terms in seven major cell types: (left) up-regulated in the mTBI, (right) down-regulated in the mTBI. (C,E) Enrichment of representative signaling pathways on UMAP plot.

The mTBI-associated transcriptional changes in individual cell types were investigated based on the regional gene set defined above (Fig. 2C,D; Supplemental Table S4). Identified DEGs were evaluated and annotated using GO and the Kyoto Encyclopedia of Genes and Genomes (KEGG), which helped explain the fundamental mechanisms of mTBI pathogenesis (Fig. 2C,D; Supplemental Fig. S3C). Diverse biological processes were activated after mTBI, including “learning” in neurons, “postsynapse organization” in oligodendrocytes, and “autophagy” and “receptor-mediated endocytosis” in microglia. In addition, we plotted certain signaling pathways using UMAP, achieving an intuitive visualization of cell-specific enrichment and alterations (Fig. 2E). For instance, “long-term potential” was down-regulated in neurons postinjury and “nucleotide excision repair” was activated in certain NSC & astrocyte subtypes, whereas the “PPAR signaling pathway” was augmented in other subtypes. Altogether, this cell type–specific transcriptome atlas elucidates the cellular basis of mTBI pathogenesis, providing a resource for future investigations into SVZ pathophysiological processes at the cell-type resolution.

mTBI induces extensive reorganization of cell–cell interactions in the SVZ

Cellular function and behavior can also be influenced by cell–cell interactions. Therefore, we conducted a survey of the expression of various ligand–receptor pairs to predict cellular cross talk in the SVZ using CellChat (Supplemental Fig. S4A; Jin et al. 2021; Zhao et al. 2022). To explore ligand–receptor interaction patterns, cell types and pathways were clustered according to the similarity of incoming or outgoing signaling (Supplemental Fig. S4A). For instance, the incoming communication of the targeted cells encompasses four different patterns. Microglia incoming communication comprises a pattern that includes the epigenetic regulator GAS; colony-stimulating factor (CSF); chemokines, such as CX3C; and TGFB (Supplemental Fig. S4A). OPCs and NSCs & astrocytes were assigned to another pattern, involving growth factors, such as PDGF, EGF, and the neurogenesis-associated factor NRG.

By calculating the numbers and weights of these ligand–receptor pairs (Fig. 3A–C), we found that the total number of inferred interactions increased, and interaction strength was enhanced in response to mTBI (Fig. 3B). The most flexible cell types were oligodendrocytes, NSCs & astrocytes, and neurons (Fig. 3C). Notably, the communication between neurons and NSCs & astrocytes was relatively more enhanced among all cell types after mTBI (Fig. 3A), consistent with the reports of significant astrocyte–neuron cross talk under adverse stimulation (Volterra and Meldolesi 2005; Barreto et al. 2011). However, communication between neurons and NSCs has not been studied. Here, we found that these two cell types interact via the TGFA-EGFR and ERG1/2/3-ERBB4 pairs. These ligand receptors are known to participate in progenitor cell proliferation (Seshadri et al. 2010), indicating that insulted neurons may release signals to NSCs following mTBI to promote neurogenesis via these pairs. Additionally, we characterized the major pathways of the affected ligand–receptor pairs (Fig. 3D,E) to visualize the detailed reorganization of cell-to-cell signaling. For instance, CSF signaling, a pathway known to induce inflammatory responses and microglia proliferation after brain injury (Wyss-Coray 2006; Luo et al. 2013), was significantly enhanced between microglia and endothelial cells in the mTBI SVZ (Fig. 3D). Whether endothelia play a role in microglial behavior postinjury remains to be elucidated (Fig. 3A). Taken together, the breadth of our study enables us to disclose elaborate alterations in the molecular network after mTBI, pinpointing potential targets for its treatment or diagnosis.

Figure 3.

Figure 3.

Open in a new tab

Alterations of the cell–cell cross talk in the SVZ after mTBI. (A) Cell–cell communication signaling network among major cell types analyzed with CellChat. The width of the lines indicates the number of pairs; different colors represent different signal sources. (B) Bar plot showing the number of inferred interactions and the proportion of interaction strength in the control and the mTBI. (C) The location of major cell types based on the count of their significant incoming (y-axis) or outgoing (x-axis) signaling pattern. (D) Signal pathway network and bar plot showing the expression of specific signaling pathways in the control and the mTBI SVZ. The width of the lines indicates signal strength. (E) Dot plot showing the pairing of ligands (red) and receptors (blue) in the control and the mTBI SVZ.

Niche NSCs & astrocytes show strong heterogeneity within the neurogenic lineage

To comprehensively characterize NSC heterogeneity within the SVZ, we conducted further resolution of the original “NSCs & astrocytes” cell type (Fig. 4A; Supplemental Table S5) into five subtypes based on unbiased clustering and previously known markers (Llorens-Bobadilla et al. 2015; Shah et al. 2018). By comparison, we identified several potential novel markers specific to individual subtypes (Fig. 4B). For instance, Mertk and Grin2c were especially enriched in niche astrocytes but barely expressed in quiescent NSC (qNSC) (Fig. 4B); thus, they can be potentially used as markers to distinguish these two cell types, which has long been a challenge. Other markers, such as Kank1 and Ntsr2 for qNSC (Shah et al. 2018; Kim et al. 2022), Adamts18 and Thbs4 for early actively dividing NSCs (aNSCs) (Beckervordersandforth et al. 2010), Egfr and Mms22l for late aNSCs (Pluchino and Nicaise 2021), and Cpne4 and Dcx for neuroblasts (Fig. 4B; Kim et al. 2022), also showed specific expression and deserve future downstream validation.

Figure 4.

Figure 4.

Open in a new tab

NSC & astrocyte diversities in the niche and their alteration post-mTBI. (A) UMAP plot showing five subtypes of NSCs & astrocytes. (B) Dot plot showing the representative markers for each subtype. A gradient of light purple to yellow indicates inhibition to activation of the term. Size of the bubble indicates the P-value. (C) Bar plot showing the proportion of each cell types in the control and mTBI group. (D, top) Volcano plot showing DEGs in the control and mTBI group. (Bottom) Violin plot showing up-regulated and down-regulated genes in each cell type. The t-test was used to determine statistical significance: (****) P ≤ 0.0001. (E) SCENIC results of the control and mTBI group. Major regulators were shown; their corresponding enriched DNA-binding motifs are shown in right column. (F) Ridge plot showing the expression of Tcf7l1. (G) Protein–protein interaction (PPI) networks of TFs and miRNAs in neuroblast (upper) and astrocyte, qNSC, and early aNSC (lower).

Based on the above subclustering, we found that the proportion of late aNSCs and qNSCs increased in the acute postinjury stage, whereas the proportion of early aNSCs, astrocytes, and neuroblasts decreased (Fig. 4C); the significant change of qNSCs has also been validated using scCODA (Supplemental Table S3; Büttner et al. 2021). This shift indicated that mTBI may promote SVZ cell differentiation from early to late aNSCs and their descendants to counterbalance cell loss caused by the sustained damage. Nevertheless, the descendants failed to further differentiate into the neuronal lineage; instead, they entered a quiescent status. This hypothesis may provide an explanation for the limited ability of the adult brain to supply neuronal loss after injury. Therefore, transcriptomic events in each subtype, detailed reprogramming lineage trajectories, and their governing factors need to be addressed.

Here, we identified DEG details and their annotated functions for each NSC & astrocyte subtype (Fig. 4D; Supplemental Table S6), creating an ideal resource for subsequent studies of mTBI pathophysiology and therapeutic targets to promote neurogenesis after brain injury. We focused on transcription factors (TFs), critical factors regulating gene expression. Single-cell regulatory network inference and clustering (SCENIC) analysis was used to establish a regulatory network for TFs and their downstream genes (Fig. 4E,F; Aibar et al. 2017). We used SCENIC analysis to binarize TF enrichment and generate a regulon activity matrix (Fig. 4E). We found that TF behavior (on/off) varied across different cell clusters, which can be traced back to the subtypes characterized above. We detected five separate modules in the SCENIC map and listed the TFs of interest in Figure 4E. For instance, ESRRG, RFX4, GRHL, and EPAS1 were especially active in astrocytes and qNSCs; KLF9, GlLS3, TEF, EPAS1, and TCF7L1 were “switched on” in astrocytes, qNSCs, and early aNSCs; BRCA1, KDM5A, TAF1, and E2F1 were highly active in early aNSCs; and FOXP2, PBX3, and SOX4 were enriched in late aNSCs and neuroblasts (Fig. 4E). In addition, we found that most DEGs were regulated by a combination of TFs, thus forming various regulatory networks grouped in different modules (Fig. 4G). Particularly, we found that the Tcf7l1, which was enriched in astrocytes, qNSCs, and early aNSCs (Fig. 4F), has been demonstrated to maintain spinal cord precocious progenitor population by inhibiting its differentiation (Kim and Dorsky 2011).

We hypothesize that the Tcf7l1 might play a similar differentiation-inhibiting role in subsets of SVZ NSCs as well. Consequently, in the downstream investigation, we explored this potential function of Tcf7l1 in NSC differentiation and neurogenesis. In summary, our study revealed the significant heterogeneities among SVZ NSCs in response to mTBI, and these responses were regulated by a complex network of TFs. Elucidating the roles of these key genes may facilitate the recovery of brain tissues.

To elucidate the reprogramming lineage trajectories and their governing factors under each condition, we used Monocle2 to uncover and show pseudotime trajectory ordering of all the NSC & astrocyte subtypes (Fig. 5A). Five states were detected along the route, with two bifurcations: branch 1 from state 1 to states 2 and 5, and branch 2 from state 2 to states 3 and 4 (Fig. 5C). We plotted the cell types against pseudotime and found that niche astrocytes took up the tip of the pseudotime tree (state 1) (Fig. 5B), indicating that niche astrocyte priming may be the initial event of neurogenesis. This result challenges the traditional notion of neurogenesis but agrees with a recent study on striatal astrocytes, which showed that astrocytes were located upstream of NSCs in the neuronal lineage and may be targeted to guide neuronal differentiation (Magnusson et al. 2020). Further experimental validation and more scientific evidence are required to clarify this issue. This process was followed by qNSC expansion (states 2 and 5) (Fig. 5B). Upon branching, the vast majority of early aNSCs formed the end of state 2 and dormant state 4. Another small group of early aNSCs gradually developed into late aNSCs and then into neuroblasts (Fig. 5B). During this period, the mTBI group showed an ∼10% decrease in the number of cells in state 3, suggesting limited differentiation toward late aNSCs and neuroblasts (Fig. 5D). The changes in cell-type proportions have been validated using the scCODA tool (Supplemental Table S3; Büttner et al. 2021). This might partially explain why the newly generated neurons in the adult brain are insufficient to fully compensate for the loss of neurons caused after the injury.

Figure 5.

Figure 5.

Open in a new tab

Pseudotime analysis reveals reprogramming trajectory of NSCs after injury. (A, top) Colored cells plotted onto the UMAP according to the pseudotime. (Bottom) Monocle2-generated pseudotemporal trajectory of NSCs & astrocytes. (B) Different astrocyte & NSC subtypes along the pseudotime trajectory. (C) Cell ordering from different differentiation stages along the pseudotime trajectory. (D) Bar plot showing the proportion of each differentiation stage in the control and the mTBI. (E, left) Heat map showing gene expression patterns along the pseudotime (1000 genes). (Right) Three clusters were identified based on the expression dynamics; the GO term for each cluster was displayed on the right column. (F) Expression pattern of representative genes during differentiation from aNSCs to neuroblasts. States are shown in colors, and branches are indicated by lines.

To identify the factors and targets that govern neurogenesis, we analyzed gene expression patterns along the reprogramming trajectories and grouped them into three clusters according to their expression dynamics (Fig. 5E). Genes responsible for NSC proliferation, differentiation, and neuronal function showed largely different expression when cell identity switched from early aNSCs to late aNSCs and neuroblasts (Fig. 5E). These changes included genes involved in “nuclear division,” “axonogenesis,” “dendrite development,” “regulation of neurogenesis,” and “neuron projection guidance” (Fig. 5E). In addition, NSC maintenance factors, such as genes from the “Hippo signaling pathways,” “Rap1 signaling pathways,” and “Wnt signaling pathways,” increased in certain remaining qNSCs, which might underlie their maintained cell fate (Fig. 5E). Tcf7l1, Fars2, Wdr17, Csmd1, Prex2, Adamts18, Erbb4, Maml3, and Nrxn3 showed varying degrees of increase or decline during aNSC differentiation to neuroblasts (Fig. 5F). Considering the aforementioned alterations of the regulon Tcf7l1, we are strongly interested in its role in neurogenesis following mTBI.

Tcf7l1 knockdown in NSCs promotes differentiation toward the neuronal lineage

Based on the above findings, as well as literature review (Kim and Dorsky 2011), we hypothesized that Tcf7l1 might play a differentiation-inhibiting role in subsets of SVZ-NSC. To investigate this, we used the loss-of-function method to study the function of Tcf7l1 in NSCs through siRNA-induced Tcf7l1 knockdown (KD) in NE-4C cells (an immortalized mouse NSC line) (Fig. 6A). After validating the KD efficiency (Fig. 6B,C), we harvested NE-4C cells at different time points following Tcf7l1_siRNA treatment (1 d and 3 d) and analyzed gene expression. Tcf7l1 deficiency decreased the expression of marker genes for qNSCs (Fabp7) and early aNSCs (Adamts18 and Cd9) and augmented the neuronal lineage markers Dcx and Meg3 and the late aNSC marker Egfr, suggesting differentiation toward late aNSCs and neuroblasts (Fig. 6D; Supplemental Fig. S4C). We further verified the gene expression results using immunofluorescence staining (Fig. 6E,F). Consistently, the qNSC marker Thbs4 decreased, and the neuronal lineage marker Dcx increased after KD. Furthermore, Tcf7l1-deficient cells gathered together in a neurosphere-like manner. This morphological change has also been observed in previous investigations of neuronal differentiation from NSCs (Fig. 6E,F). Overall, these results suggest that Tcf7l1 is critical in maintaining cell stemness, and Tcf7l1 deficiency could change the NSC programming trajectory, promoting differentiation toward neuronal lineage cells.

Figure 6.

Figure 6.

Open in a new tab

Knockdown of Tcf7l1 induces differentiation of NE-4C cells toward neuron lineage cells in vitro. (A) Schematic showing the siRNA knockdown approach on the NE-4C cell line. (B) Cy3 signal indicating successful transfection of cells using siRNA. (C) qPCR results of Tcf7l1 in each group. Expression levels were normalized to Gapdh; three technical duplicates were performed in individual qPCR experiments. The t-test was used to determine statistical significance: (*) P ≤ 0.05, (**) P ≤ 0.01, (***) P ≤ 0.001, (****) P ≤ 0.0001, (ns) not significant. (D) qPCR assay of marker genes of different NSCs, progenitors, and neuroblasts under different treatment. Expression levels were normalized to Gapdh; three technical duplicates were performed in individual qPCR experiments. The t-test was used to determine statistical significance: (*) P ≤ 0.05, (**) P ≤ 0.01, (***) P ≤ 0.001, (****) P ≤ 0.0001, (ns) not significant. (E,F) Immunofluorescence staining against THBS4 (E) and DCX (F) in cultured cells under different treatments. Sections are counterstained with DAPI.

Tcf7l1knockdown enhances neurogenesis following mTBI

To further validate the critical role of Tcf7l1, we performed Tcf7l1-KD in an experimental mouse model of mTBI. We stereotactically injected adenoviruses loaded with vectors encoding green fluorescent protein (GFP) and scrambled sh-Tcf7l1 into the SVZ of mice 3 d before mTBI (Fig. 7A). On the third day after mTBI, the mice were euthanized, and the SVZ was dissected (Fig. 7A). GFP signals were observed in a large number of cells in the SVZ niche, suggesting successful injection and viral infection of the SVZ (Fig. 7B). In vivo Tcf7l1-KD was validated via gene expression analysis of the SVZ (Fig. 7C). Immunostaining revealed a significantly increased ratio of MKI67 and THBS4 costained/THBS4-positive cells in the SVZ after mTBI, further elevated in mTBI mice pretreated for Tcf7l1-KD. These results suggest that mTBI promoted aNSC proliferation in this neurogenic niche, and suppressing THBS4 augmented this alteration (Fig. 7D,E). Notably, the total number of aNSCs (indicated as THBS4+) did not change significantly under Tcf7l1-KD following mTBI. Overall, the above results raised the question whether TCF7L1 deficiency promotes the differentiation of proliferated aNSCs toward their descendants. To trace the fate of these proliferated cells, we treated mice with BrdU injections for five consecutive days after mTBI and costained BrdU with DCX (Fig. 7A,F). We observed a significant increase in the percentage of BrdU+ and DCX+/BrdU+ in the mTBI + Tcf7l1-KD group, suggesting that neuronal lineage differentiation of proliferated NSCs is enhanced under TCF7L1 deficiency (Fig. 7G). Together, our in vitro and in vivo results demonstrate that Tcf7l1 plays a critical role in maintaining NSC quiescence and suppressing neuronal lineage differentiation. Targeting Tcf7l1 represents a potential direction in regenerative medicine for brain injury.

Figure 7.

Figure 7.

Open in a new tab

Suppressing Tcf7l1 promotes neuronal differentiation of the SVZ NSCs after mTBI. (A, top) Schematic diagram of in vivo experiment. (Bottom) Schematic representing the adenovirus injection, along with its genetic sequence. (B) GFP signal indicating successful SVZ injection of the adenovirus. (C) qPCR assay of Tcf7l1 indicating successful knockdown of Tcf7l1 in the SVZ. Expression levels were normalized to Gapdh. Three technical duplicates were performed in individual qPCR experiments. (D,E) Immunofluorescence staining (D) and THBS4+MKI67+/THBS4+ cell counting results (E) for THBS4 and MKI67 in the SVZ region from the sham, mTBI + control vector, and mTBI + Tcf7l1 vector. Tissues are counterstained with DAPI. The t-test was used to determine statistical significance: (*) P ≤ 0.05. (F,G) Immunofluorescence staining (F) and DCX+BrdU+/BrdU+ cell counting results (G) for DCX and BrdU in the SVZ region from the sham, mTBI + control vector, and mTBI + Tcf7l1 vector. Tissues are counter-stained with DAPI. The t-test was used to determine statistical significance: (**) P ≤ 0.01.

Discussion

Through high-throughput single-cell profiling, we (1) demonstrate that our comprehensive snRNA-seq strategy generates a cell atlas of the entire SVZ niche; (2) report significant information about the cells, genes, and pathways vulnerable to mTBI, and reveals cell type–specific dynamics along the neurogenic lineage; (3) identify key genes like Tcf7l1 that have the potential to ameliorate cell damage in the brain or facilitate neural repair, making them potential targets for developing therapeutics for mTBI; and (4) provide a resource for studying adult neurogenesis in the SVZ under both physiological and pathological conditions. To support our data-driven strategy, we conducted experimental validations and demonstrated, for the first time, that a deficiency of the TF TCF7L1, a regulator of cell cycle genes, switches the NSC reprogramming directions and promotes NSC differentiation toward neuronal lineage, providing a promising target for future regeneration medicine approaches.

Previous studies have described the processes involved in cell proliferation and differentiation during neurogenesis. Recent efforts continue to unveil the diversity of NSCs and progenitors harbored in the heterogeneous pool, but their lineage trajectories and delicate intercellular communications remain poorly understood. Using single-cell technology, Llorens-Bobadilla et al. (2015) unveiled distinct cell types or states within the SVZ. Their investigation successfully identified four subtypes of NSCs and elucidated the factors implicated in the transition of NSCs from a quiescent state to an activated state under forebrain ischemia. Subsequently, Zywitza et al. (2018) characterized and classified six subtypes of NSCs in the SVZ based on molecular features and RNA velocity analysis. Additionally, they demonstrated the pivotal role of LRP2 as a key regulator of neurogenesis in the SVZ. Recently, Xie et al. (2020) conducted an analysis on the heterogeneity of SVZ stem cells in mice across different age groups, spanning from neonates to aged individuals. Their study successfully identified four distinct subtypes of qNSCs and three subpopulations of progenitor cells. Although these studies differ in terms of the precision in subtyping SVZ stem cells, they share similarities in classifying major NSC types such as aNSCs, qNSCs, and NBs, which aligns with the findings of our study.

Using a well-established mTBI model and single-nucleus transcriptomics, we conducted a pioneering investigation into the molecular and transcriptional alterations specific to cell types within the SVZ stem cell pool following mild brain trauma. In addition, we uncovered a number of novel genes that have the potential to mark cell type in the SVZ niche and each NSC/progenitor cell subtype, thus filling this current knowledge gap. For instance, distinguishing qNSCs from niche astrocytes effectively with a specific marker has been challenging; hence, the likelihood ratio has been applied to identify them (Llorens-Bobadilla et al. 2015). Our data detected several genes exclusively enriched in qNSCs, not in niche astrocytes (Supplemental Table S5). Specifically, Kank1 showed exclusive expression in qNSCs, whereas Grin2c and Mertk showed high expression in astrocytes. These genes, not previously reported, offer potential markers for identifying these cell types. Another previous notion challenged by our data is that neurogenesis lineage priming is initiated by qNSC activation. Using pseudotime analysis, a powerful approach for revealing lineage trajectories (Holden et al. 2021; Jamjoom et al. 2021), we discovered that niche astrocytes triggered the neurogenic process, rather than qNSCs. Our findings align with a recent study suggesting that striatal astrocytes act as latent NSCs, triggering neuronal lineage cell genesis upstream of NSCs (Magnusson et al. 2020). In addition, our analysis revealed that qNSCs differentiated into two branches along the pseudotime line. One branch of qNSC transitioned into a dormant state, mainly comprising early aNSCs. These cells shared similar activated TFs with astrocytes and qNSCs (Fig. 4E), suggesting that their transcriptional regulatory network did not undergo significant changes during differentiation. Another branch differentiated into late aNSCs and subsequently into neuroblasts. In comparison, this branch showed a completely different regulatory network (Fig. 4G), suggesting a distinct switch in cell identity. Conducting in-depth studies to validate these analytical results and further exploring targets involved in NSC homeostasis and neurogenesis would be of significant interest. Our data also offer insights into the mechanisms underlying the limited replacement of lost neurons after brain injury and potential targets to address this issue. In addition, we clarified the molecular events orchestrating each neurogenic stage at single-cell resolution and identified potential genes that may govern the fate of each NSC/progenitor subtype. This discovery is expected to facilitate the development of regenerative medicines.

Our single-nucleus SVZ transcriptome atlas and lineage trajectory regulation analysis revealed Tcf7l1 as a key gene in NSC differentiation. Among the four TCF/lymphoid enhancer factor family members, Tcf7l1 is the only one expressed in undifferentiated mouse embryonic stem cells (Yi et al. 2011) and in the developing forebrain primordium during the presomitic and early somite stages (Galceran et al. 1999; Merrill et al. 2004). A literature review revealed that Tcf7l1 can maintain the spinal cord precocious progenitor cell population and inhibit its differentiation in zebrafish (Andoniadou et al. 2011; Kim and Dorsky 2011). However, its role in the SVZ of adult mice and its effects on neurorepair after mTBI have not been addressed. In vitro and in vivo loss-of-function experiments demonstrated that suppressing Tcf7l1 promotes qNSC differentiation toward neuronal lineage cells. This experiment validates our single-nucleus data-driven strategy. In addition, these results aid in understanding the pathophysiological process of mTBI, and our fate-switching approach represents a significant advancement in neuroregeneration, offering a potential therapeutic target for enhancing neuronal supply after brain injury.

To further explore the potential therapeutic effects of Tcf7l1, we conducted the NSS assessment in mice 1 h after mTBI. Mice with Tcf7l1-KD showed significantly lower NSS scores compared with those of the mTBI group treated with a control vector (Supplemental Fig. S5A), indicating a reduced level of neurological impairment in the Tcf7l1-KD group. To evaluate the long-term impact of Tcf7l1 KD in mice, we conducted additional behavioral experiments on sham, mTBI + Control vector, and mTBI + Tcf7l1 KD mice 1 mo after injury. These experiments included the Rotarod test, Morris water maze test (MWM), Y-maze test, and novel object recognition test (Supplemental Fig. S5). In the Rotarod test, we found that the motor and coordination abilities of mTBI + Tcf7l1 KD mice were significantly better than those of the mTBI + Control vector group in the first three trials (Supplemental Fig. S5B). From the second day onward (after the fourth trial), there was no significant difference among the two groups. The Y-maze test results showed that the spontaneous alternation rate of mTBI + Tcf7l1 KD mice was significantly higher than that of mTBI + Control vector mice 1 mo after injury (Supplemental Fig. S5C). The MWM results showed that the latency of mTBI + Tcf7l1 KD mice on the fourth and fifth days was significantly shorter than that of mTBI + Control vector mice, and their overall spatial learning and localization abilities were close to those of the sham mice (Supplemental Fig. S5D). In the spatial exploration, mTBI + Tcf7l1 KD mice explored the quadrant where the original platform was located for a longer time compared with mTBI + Control vector mice, but the difference was not significant (Supplemental Fig. S5E). The novel object recognition test showed that there was no significant difference among the groups of mice in exploring old objects, but mTBI + Tcf7l1 KD mice spent significantly more time exploring new objects compared with mTBI + Control vector mice (Supplemental Fig. S5F). The above results indicate that mTBI + Tcf7l1 KD mice have better motor, memory, and learning abilities compared with mTBI + Control vector mice 1 mo after injury. However, it is essential to acknowledge that the transition from SVZ neurogenesis to cognitive function in post-TBI mice is a complex and multifaceted process. This journey involves not only the proliferation and differentiation of NSCs but also the migration of neural cells, the establishment of the surrounding microenvironment, and the construction of neural circuits. Consequently, confirming a direct causal relationship between the enhancement of cognitive function and SVZ neurogenesis regulated by Tcf7l1 would necessitate a considerable volume of subsequent and comprehensive studies. As such, we recognize that further research is required to fully unravel the exact mechanisms that underpin the observed enhancements in cognitive function.

Acknowledging the limitations of our study is imperative. It has been observed that following TBI, female animals experience an elevation in estrogen levels at the site of trauma, which has potential to mitigate neuroinflammatory responses mediated by microglia and astrocytes (Wang et al. 2021). This increase in estrogen plays a vital role in reducing intracranial pressure and mitigating brain edema, ultimately leading to a better prognosis (Brotfain et al. 2016; Lu et al. 2018). Consequently, many studies investigating brain injury opt to restrict the use of mice to a specific sex in order to minimize variability and eliminate the influence of sex hormones on experimental outcomes (Llorens-Bobadilla et al. 2015; Arneson et al. 2018; Park et al. 2018; Somebang et al. 2021). In the present study, in order to observe the molecular level changes after mTBI more clearly, we have solely used male mice for single-nucleus sequencing, which inherently restricts our ability to draw conclusions regarding potential sex disparities in response to mTBI. Thus, we have outlined future research directions that include conducting comparative studies between male and female mice using the single-cell transcriptomic approach to investigate sex-specific differences in mTBI response. We firmly believe that such investigations will yield valuable insights into the underlying mechanisms of sex disparities in mTBI and provide guidance for potential sex-specific therapeutic interventions.

Methods

Ethics approval and consent to participate

Animal experiments were performed in accordance with the institutional guidelines approved by the animal care and experimental committee of Sichuan University, China.

Animals

All animal experiments were performed in accordance with the institutional guidelines following approval by the animal care and experimental committee of Sichuan University, China. C57BL/6J male mice, aged 8–10 wk, were group-housed conventionally in standard conditions (12-h light/dark cycle; temperature: 22°C–25°C; relative humidity: 40%–60%) with ad libitum food and water for at least 1 wk before the experiment. To ensure proper blinding, each mouse was assigned a unique alphanumeric code unrelated to the treatment group for identification purposes. For NSS assessment and single-cell sequencing, a total of six SVZ samples were used. In the in vivo experiments, a cohort of 60 mice was used, with 20 mice allocated to each respective group. These mice underwent immunofluorescence staining, BrdU injection, and behavioral testing procedures. The study implemented specific exclusion criteria for mouse selection in subsequent experimentation. Mice meeting any of the following criteria were excluded: (1) age <8 wk or >10 wk, (2) body weight <18 g or >25 g, (3) presence of illness or disease, (4) technical issues during experimental procedures, (5) specific breeding or housing conditions introducing confounding variables or impacting study outcomes, (6) prior treatments or interventions, and (7) certain genetic or phenotypic backgrounds.

mTBI

To ensure consistent and standardized anesthesia, we used an inhalation anesthesia machine with isoflurane for our study as isoflurane anesthesia offers advantages such as rapid induction and recovery (Bielefeld et al. 2017; Cho et al. 2019; Huynh et al. 2020). Mice were anesthetized with 4% isoflurane and were maintained on anesthesia with 1.5% isoflurane until after impact and suturing. Mice were placed in an anesthetic mask and a stereotaxic apparatus, with a foam pad placed underneath to provide cushioning for the impact (Xu et al. 2016). First, a 10-mm midline incision was made in the mouse's scalp, and the skin and fascia were cleared. Then mice received an impact by the CCI machine eCCI-6.3 (Custom Design and Fabrication) directly onto the skull, with an impact diameter of 3 mm, impact velocity of 3.5 m/sec, depth of 1 mm, and duration of 150 msec. The impact position was 0.5 mm posterior to the bregma and 2 mm lateral to the midplane. These injury parameters are consistent with those used in previous studies of mTBI (An et al. 2016; Xu et al. 2021). The sham group of mice did not receive an impact but was otherwise treated identically to the mTBI group.

Neurological severity score

We used the NSS assessment to validate the severity of brain injury in mice 1 h post injury as previously described in detail (Flierl et al. 2009). The NSS evaluation comprises a comprehensive range of assessments to evaluate neurological function, encompassing motor coordination, balance, and reflexes. According to the established criteria (Beni-Adani et al. 2001), a score of three to four was considered indicative of mTBI (Henninger et al. 2016). According to the NSS protocol in the work of Flierl et al. (2009), the mice were exposed to isoflurane inhalation for a limited duration of 5–10 min to prevent prolonged interference with NSS evaluation. To demonstrate that mice are fully awake from the anesthetic state 1 h after injury, we conducted an NSS comparison between mice that received anesthesia-only and naive mice (Supplemental Fig. S5A). The results clearly demonstrated that the NSS scores of anesthetized mice at 1 h post-TBI were comparable to those of naive mice, indicating no significant differences in neurological function.

Intracranial injections, immunofluorescence staining, real-time quantitative polymerase chain reaction (qPCR)

Please refer to our previous publications (Xiao et al. 2019; Cao et al. 2022; Long et al. 2022).

Antibodies used were as follows: anti-THBS4 (Abcam ab263898, RRID: AB_2922811), anti-DCX (Abcam ab207175, RRID: AB_2894710), and anti-Ki67 (Abcam ab279653, RRID: AB_2934265).

For primer sequences, please see Supplemental Table S7.

BrdU injection and immunostaining

We initiated BrdU injections immediately after mTBI and continued daily injections for five consecutive days to label proliferative cells. Mice were received daily single i.p. injections of BrdU (Abcam ab142567, 50 mg/kg). Sections were incubated with 2 N HCl for 1 h before immunostaining against BrdU (Thermo Fisher Scientific MA3-071, RRID: AB_10986341).

SVZ dissection

Please refer to previous studies (Kim and Dorsky 2011; Li et al. 2023).

Nuclei isolation

The mice SVZ was dissected according to the method of a previous study (Walker and Kempermann 2014). NLB buffer (0.2 U/μL RNase inhibitor [Takara], 250 mM sucrose, 10 mM Tris-HCl, 3 mM MgAc2, 0.1% Triton X-100 [Sigma-Aldrich], 0.1 mM EDTA) was used to homogenize the frozen tissue. Different concentrations of sucrose were applied to purify the nuclei using sucrose density gradient centrifugation, and the nuclei were inspected for visual appearance and cell lysis using trypan blue and quantified with a hemocytometer before being adjusted to a concentration of 1000 nuclei/μL.

Single-nucleus RNA sequencing

The snRNA-seq libraries were built using the 10× Genomics Chromium Controller Instrument (PN-120223, 10×Genomics) and chromium single-cell 3′ V3.1 reagent kits (PN-1000121, 10× Genomics). Please refer to manufacturer's instructions.

Cell culture and transfection

NE-4C cell lines were cultured in a 37°C incubator under 5% CO2 with Dulbecco's Modified Eagle Medium (DMEM) containing 10% fetal bovine serum and 1% double antibiotics (penicillin/streptomycin). For transfection, cells were transfected with siRNA-Tcf7l1 or the control siRNA using Lipofectamine 3000 according to the manufacturer′s instructions. Briefly, 250 pmol of each siRNA was transfected using 2.5 µL of Lipofectamine 2000 per dish in Opti-MEM. Following transfection, cells were incubated at 37°C in a 5% CO2 humidified atmosphere for 48 h before being harvested for the assays. The sequences corresponding to the siRNA-Tcf7l1 were sense 5′-GGAAGAAGAAGAAGAGGAAGAGAGA-3′.

Single-nucleus RNA sequencing quality control

Expression matrixes were loaded into R (v.4.1.3) using the function Read10× in Seurat (v.1.1) (Stuart et al. 2019; R Core Team 2023) and then merged together by column (detailed data set descriptions of the sham group have been reported in our recent data descriptor) (Li et al. 2023; https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE19807). This resulted in a total of 16,754 cells. We performed quality control to remove doublets, dead cells, and empty droplets that could impact the accuracy and reliability of our sequencing results. The quality-control criteria are as follows: (1) 1000 < total UMI counts < 20,000, (2) 500 < gene numbers < 7500, and (3) mitochondrial gene percentage < 10%. After quality control, we filtered out 1489 low-quality cells that did not meet our filtering criteria for analysis. This step was crucial in ensuring the quality of our data and minimizing any potential sources of bias or artifacts in our analysis. The expression level of each gene in each cell was normalized using the function NormalizeData with the default parameters to remove the influence of sequencing library size, which converted expression values from UMI counts to ln[10,000 × UMI counts/total UMI counts in cell + 1]. The Seurat R package (v4.1.1) and SCTransform (v0.2.1) R packages were used for the downstream analysis of single-cell RNA-seq data (Hafemeister and Satija 2019; Stuart et al. 2019). Briefly, the expression matrix was processed using SCTransform with default parameters, followed by dimensionality reduction (principal component analysis [PCA]). Then clusters were visualized using uniform manifold approximation and projection for dimension reduction (UMAP). Cell clustering was further performed using the FindNeighbors and FindClusters functions in Seurat. Cell populations were defined based on the DEGs identified between clusters.

Marker gene identification

Marker genes were identified with a Wilcoxon rank-sum test implemented in the FindAllMarkers function with the following criteria: logFC > 0.25, P < 0.05, and min.pct > 0.1. The expression of single genes was depicted using custom R scripts, either per cell cluster as distribution of normalized UMI counts (violins) or per cell as color gradient in UMAP.

Functional annotation and pathway analysis

GO annotations (The Gene Ontology Consortium et al. 2000) were downloaded from NCBI (http://www.ncbi.nlm.nih.gov/), GO knowledgebase (http://www.geneontology.org/), and UniProt (http://www.uniprot.org/). Pathway analysis was performed using the KEGG database (https://www.genome.jp/kegg/). A Fisher's exact test was applied to identify significant GO categories and pathways. A false-discovery rate (FDR) was used to correct the P-values, and P-values < 0.05 were considered significant (Draghici et al. 2007).

Cell–cell communication analysis

Communications between cells were analyzed using CellChat (Jin et al. 2021). Please refer to GitHub (https://github.com/sqjin/CellChat). P-values < 0.05 were selected for revealing relationships between cell types.

Single-cell regulatory network inference and clustering

For SCENIC (Aibar et al. 2017), we created the motif databases the allow to use RcisTarget and SCENIC (v1.3.1) on the mouse. Briefly, we used GENIE3 to construct a coexpression network and RcisTarget to identify the direct binding by DNA-motif analysis. After constructing regulons for each TF using motif data set (mm9-tss-centered-10 kb-7-species.mc9nr.feather, mm10-refseq-r80-500 bp-up-and-100 bp-down-tss.mc9nr.feather), the activity score of the regulons in each cell was quantified using AUCell. The activity scores of regulons generated in NSCs & astrocytes in mTBI and control mice were averaged, scaled, and visualized via heat map. The network of the regulon and its targets was visualized using Cytoscape (v3.8.0) (Shannon et al. 2003). Only the connections of a specific regulon and downstream targets that were coregulated by multiple TFs were shown.

Pseudotime analysis

The reprogramming trajectory analysis was performed using Monocle2 (v2.22.0; http://cole-trapnell-lab.github.io/monocle-release), with DDRTree (v0.1.5) and default parameters. A heat map was produced to display the series of genes with certain expression patterns along the pseudotime. Using the differential gene test function of Monocle2, a Q value < 0.01 was set to identify significantly changed genes.

scCODA

To identify statistically significant changes in cell population composition, we used the scCODA v.0.1.9 Python package to perform compositional analysis of the single-cell data (Büttner et al. 2021). For Figures 2A, 4C, and 5D, we selected ependymal, astrocyte, and state 1 cells as the reference cell types, respectively. To ensure the detection of subtle yet biologically relevant changes, we set the FDR value to 0.4, as recommended by Büttner et al. (2021) in their documentation.

Statistical analyses

The data analysis was performed by a blinded investigator, unaware of the treatment groups. The data were labeled with unique alphanumeric codes, unlinked to the treatment groups. In the statistical analysis, FDR was used to correct the P-values. We defined P-values < 0.05 as significant. The P-values are reported in the figures and figure legends: (*) P ≤ 0.05, (**) P ≤ 0.01, (***) P ≤ 0.001, (****) P ≤ 0.0001, and (n.s.) not significant. The error bars in the figures represent the mean ± SEM. For the statistical analysis of snRNA-seq data, we used R (v4.1.3). The Wilcoxon rank-sum test was used for marker gene identification, as well as for the selection of DEGs for pseudotime analysis, cell–cell communication analysis, and other related analyses. A Fisher's exact test was applied to identify significant GO categories and pathways. Additionally, the t-test was used to determine statistical differences between samples. Detailed analysis methods and relevant parameter indicators have been described previously. For NSS, qPCRs, and immunostaining statistic data, we performed unpaired two-tailed t-tests using GraphPad Prism (v.8.0) to determine statistical differences between samples.

Data access

All raw and processed sequencing data of mTBI generated in this study have been submitted to the NCBI Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) under accession number GSE198074. The analysis code is available at GitHub (https://github.com/limanrui/scseq.git) and as Supplemental Code. A matrix of samples in this study is available at Figshare (https://doi.org/10.6084/m9.figshare.21907497.v2) and as Supplemental Data.

Supplementary Material

Supplement 1
Supplement 2
Supplement 3
Supplement 4
Supplemental_Table_S4.docx (721.7KB, docx)
Supplement 5
Supplemental_Table_S5.docx (601.9KB, docx)
Supplement 6
Supplemental_Table_S6.docx (757.2KB, docx)
Supplement 7
Supplement 8
Supplement 9

Acknowledgments

The study was funded by the National Natural Science Foundation of China to Xiameng Chen (no. 82202077) and the Natural Science Foundation of Sichuan Province to Xiameng Chen (no. 23NSFSC4762).

Author contributions: Q.Y., Lingxuan Z., Manrui L., Y.X., Xiaogang C., and R.Y. performed the statistical and computational analysis. X.O., M.H., and Miao L. performed CCI and tissue sampling. Q.Y., Lingxuan Z., Manrui L., Y.X., and H.D. performed the PCR validation, immunofluorescence staining, cell counting, and SVZ injection. Q.Y. and Xiameng C. wrote the manuscript. Lin Z., Xiaogang C., and Meili L. revised the manuscript. The study was designed and managed by Xiameng C., W.L., and X.X.

Footnotes

[Supplemental material is available for this article.]

Article published online before print. Article, supplemental material, and publication date are at https://www.genome.org/cgi/doi/10.1101/gr.277881.123.

Competing interest statement

The authors declare no competing interests.

References

  1. Aibar S, González-Blas CB, Moerman T, Huynh-Thu VA, Imrichova H, Hulselmans G, Rambow F, Marine JC, Geurts P, Aerts J, et al. 2017. SCENIC: single-cell regulatory network inference and clustering. Nat Methods 14: 1083–1086. 10.1038/nmeth.4463 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. An C, Jiang X, Pu H, Hong D, Zhang W, Hu X, Gao Y. 2016. Severity-dependent long-term spatial learning-memory impairment in a mouse model of traumatic brain injury. Transl Stroke Res 7: 512–520. 10.1007/s12975-016-0483-5 [DOI] [PubMed] [Google Scholar]
  3. Andoniadou CL, Signore M, Young RM, Gaston-Massuet C, Wilson SW, Fuchs E, Martinez-Barbera JP. 2011. HESX1- and TCF3-mediated repression of Wnt/β-catenin targets is required for normal development of the anterior forebrain. Development 138: 4931–4942. 10.1242/dev.066597 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Arneson D, Zhang G, Ying Z, Zhuang Y, Byun HR, Ahn IS, Gomez-Pinilla F, Yang X. 2018. Single cell molecular alterations reveal target cells and pathways of concussive brain injury. Nat Commun 9: 3894. 10.1038/s41467-018-06222-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Barreto GE, Gonzalez J, Torres Y, Morales L. 2011. Astrocytic-neuronal crosstalk: implications for neuroprotection from brain injury. Neurosci Res 71: 107–113. 10.1016/j.neures.2011.06.004 [DOI] [PubMed] [Google Scholar]
  6. Batiuk MY, Martirosyan A, Wahis J, de Vin F, Marneffe C, Kusserow C, Koeppen J, Viana JF, Oliveira JF, Voet T, et al. 2020. Identification of region-specific astrocyte subtypes at single cell resolution. Nat Commun 11: 1220. 10.1038/s41467-019-14198-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Beckervordersandforth R, Tripathi P, Ninkovic J, Bayam E, Lepier A, Stempfhuber B, Kirchhoff F, Hirrlinger J, Haslinger A, Lie DC, et al. 2010. In vivo fate mapping and expression analysis reveals molecular hallmarks of prospectively isolated adult neural stem cells. Cell Stem Cell 7: 744–758. 10.1016/j.stem.2010.11.017 [DOI] [PubMed] [Google Scholar]
  8. Beni-Adani L, Gozes I, Cohen Y, Assaf Y, Steingart RA, Brenneman DE, Eizenberg O, Trembolver V, Shohami E. 2001. A peptide derived from activity-dependent neuroprotective protein (ADNP) ameliorates injury response in closed head injury in mice. J Pharmacol Exp Ther 296: 57–63. [PubMed] [Google Scholar]
  9. Biancalani T, Scalia G, Buffoni L, Avasthi R, Lu Z, Sanger A, Tokcan N, Vanderburg CR, Segerstolpe A, Zhang M, et al. 2021. Deep learning and alignment of spatially resolved single-cell transcriptomes with tangram. Nat Methods 18: 1352–1362. 10.1038/s41592-021-01264-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bielefeld P, Sierra A, Encinas JM, Maletic-Savatic M, Anderson A, Fitzsimons CP. 2017. A standardized protocol for stereotaxic intrahippocampal administration of kainic acid combined with electroencephalographic seizure monitoring in mice. Front Neurosci 11: 160. 10.3389/fnins.2017.00160 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Brotfain E, Gruenbaum SE, Boyko M, Kutz R, Zlotnik A, Klein M. 2016. Neuroprotection by estrogen and progesterone in traumatic brain injury and spinal cord injury. Curr Neuropharmacol 14: 641–653. 10.2174/1570159X14666160309123554 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bu W, Ren H, Deng Y, Del Mar N, Guley NM, Moore BM, Honig MG, Reiner A. 2016. Mild traumatic brain injury produces neuron loss that can be rescued by modulating microglial activation using a CB2 receptor inverse agonist. Front Neurosci 10: 449. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Buono KD, Vadlamuri D, Gan Q, Levison SW. 2013. Leukemia inhibitory factor is essential for subventricular zone neural stem cell and progenitor homeostasis as revealed by a novel flow cytometric analysis. Dev Neurosci 34: 449–462. 10.1159/000345155 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Büttner M, Ostner J, Müller CL, Theis FJ, Schubert B. 2021. scCODA is a Bayesian model for compositional single-cell data analysis. Nat Commun 12: 6876. 10.1038/s41467-021-27150-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Cao Q, Benton RL, Whittemore SR. 2002. Stem cell repair of central nervous system injury. J Neurosci Res 68: 501–510. 10.1002/jnr.10240 [DOI] [PubMed] [Google Scholar]
  16. Cao S, Li M, Sun Y, Wu P, Yang W, Dai H, Guo Y, Ye Y, Wang Z, Xie X, et al. 2022. Intermittent fasting enhances hippocampal NPY expression to promote neurogenesis after traumatic brain injury. Nutrition 97: 111621. 10.1016/j.nut.2022.111621 [DOI] [PubMed] [Google Scholar]
  17. Chavali M, Klingener M, Kokkosis AG, Garkun Y, Felong S, Maffei A, Aguirre A. 2018. Non-canonical Wnt signaling regulates neural stem cell quiescence during homeostasis and after demyelination. Nat Commun 9: 36. 10.1038/s41467-017-02440-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Chiu CC, Liao YE, Yang LY, Wang JY, Tweedie D, Karnati HK, Greig NH, Wang JY. 2016. Neuroinflammation in animal models of traumatic brain injury. J Neurosci Methods 272: 38–49. 10.1016/j.jneumeth.2016.06.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Cho C, Michailidis V, Lecker I, Collymore C, Hanwell D, Loka M, Danesh M, Pham C, Urban P, Bonin RP, et al. 2019. Evaluating analgesic efficacy and administration route following craniotomy in mice using the grimace scale. Sci Rep 9: 359. 10.1038/s41598-018-36897-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Collins-Praino LE, Arulsamy A, Katharesan V, Corrigan F. 2018. The effect of an acute systemic inflammatory insult on the chronic effects of a single mild traumatic brain injury. Behav Brain Res 336: 22–31. 10.1016/j.bbr.2017.08.035 [DOI] [PubMed] [Google Scholar]
  21. Doetsch F, Caille I, Lim DA, García-Verdugo JM, Alvarez-Buylla AJC. 1999. Subventricular zone astrocytes are neural stem cells in the adult mammalian brain. Cell 97: 703–716. 10.1016/s0092-8674(00)80783-7 [DOI] [PubMed] [Google Scholar]
  22. Dougherty KJ, Zagoraiou L, Satoh D, Rozani I, Doobar S, Arber S, Jessell TM, Kiehn O. 2013. Locomotor rhythm generation linked to the output of spinal shox2 excitatory interneurons. Neuron 80: 920–933. 10.1016/j.neuron.2013.08.015 [DOI] [PubMed] [Google Scholar]
  23. Draghici S, Khatri P, Tarca AL, Amin K, Done A, Voichita C, Georgescu C, Romero R. 2007. A systems biology approach for pathway level analysis. Genome Res 17: 1537–1545. 10.1101/gr.6202607 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Flierl MA, Stahel PF, Beauchamp KM, Morgan SJ, Smith WR, Shohami E. 2009. Mouse closed head injury model induced by a weight-drop device. Nat Protoc 4: 1328–1337. 10.1038/nprot.2009.148 [DOI] [PubMed] [Google Scholar]
  25. Furtado ABV, Gonçalves DF, Hartmann DD, Courtes AA, Cassol G, Nunez-Figueredo Y, Argolo DS, do Nascimento RP, Costa SL, da Silva VDA, et al. 2021. JM-20 treatment after mild traumatic brain injury reduces glial cell pro-inflammatory signaling and behavioral and cognitive deficits by increasing neurotrophin expression. Mol Neurobiol 58: 4615–4627. 10.1007/s12035-021-02436-4 [DOI] [PubMed] [Google Scholar]
  26. Galceran J, Farinas I, Depew MJ, Clevers H, Grosschedl R. 1999. Wnt3a-/--like phenotype and limb deficiency in Lef1−/−Tcf1−/− mice. Genes Dev 13: 709–717. 10.1101/gad.13.6.709 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. The Gene Ontology Consortium, Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, et al. 2000. Gene ontology: tool for the unification of biology. Nat Genet 25: 25–29. 10.1038/75556 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hafemeister C, Satija R. 2019. Normalization and variance stabilization of single-cell RNA-seq data using regularized negative binomial regression. Genome Biol 20: 296. 10.1186/s13059-019-1874-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Henninger N, Bouley J, Sikoglu EM, An J, Moore CM, King JA, Bowser R, Freeman MR, Brown RH Jr. 2016. Attenuated traumatic axonal injury and improved functional outcome after traumatic brain injury in mice lacking Sarm1. Brain 139: 1094–1105. 10.1093/brain/aww001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Holden SS, Grandi FC, Aboubakr O, Higashikubo B, Cho FS, Chang AH, Forero AO, Morningstar AR, Mathur V, Kuhn LJ, et al. 2021. Complement factor C1q mediates sleep spindle loss and epileptic spikes after mild brain injury. Science 373: eabj2685. 10.1126/science.abj2685 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Huynh LM, Burns MP, Taub DD, Blackman MR, Zhou J. 2020. Chronic neurobehavioral impairments and decreased hippocampal expression of genes important for brain glucose utilization in a mouse model of mild TBI. Front Endocrinol (Lausanne) 11: 556380. 10.3389/fendo.2020.556380 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Jamjoom AAB, Rhodes J, Andrews PJD, Grant SGN. 2021. The synapse in traumatic brain injury. Brain 144: 18–31. 10.1093/brain/awaa321 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, Myung P, Plikus MV, Nie Q. 2021. Inference and analysis of cell-cell communication using CellChat. Nat Commun 12: 1088. 10.1038/s41467-021-21246-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kim HS, Dorsky RI. 2011. Tcf7l1 is required for spinal cord progenitor maintenance. Dev Dyn 240: 2256–2264. 10.1002/dvdy.22716 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Kim H, Cho B, Park H, Kim J, Kim S, Shin J, Lengner CJ, Won KJ, Kim J. 2022. Dormant state of quiescent neural stem cells links Shank3 mutation to autism development. Mol Psychiatry 27: 2751–2765. 10.1038/s41380-022-01563-1 [DOI] [PubMed] [Google Scholar]
  36. Li M, Chen X, Yang Q, Cao S, Wyler S, Yuan R, Zhang L, Liao M, Lv M, Wang F, et al. 2023. Single-nucleus profiling of adult mice sub-ventricular zone after blast-related traumatic brain injury. Sci Data 10: 13. 10.1038/s41597-022-01925-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Llorens-Bobadilla E, Zhao S, Baser A, Saiz-Castro G, Zwadlo K, Martin-Villalba A. 2015. Single-cell transcriptomics reveals a population of dormant neural stem cells that become activated upon brain injury. Cell Stem Cell 17: 329–340. 10.1016/j.stem.2015.07.002 [DOI] [PubMed] [Google Scholar]
  38. Long X, Yang Q, Qian J, Yao H, Yan R, Cheng X, Zhang Q, Gu C, Gao F, Wang H, et al. 2022. Obesity modulates cell-cell interactions during ovarian folliculogenesis. iScience 25: 103627. 10.1016/j.isci.2021.103627 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Lu H, Ma K, Jin L, Zhu H, Cao R. 2018. 17β-estradiol rescues damages following traumatic brain injury from molecule to behavior in mice. J Cell Physiol 233: 1712–1722. 10.1002/jcp.26083 [DOI] [PubMed] [Google Scholar]
  40. Luo J, Elwood F, Britschgi M, Villeda S, Zhang H, Ding Z, Zhu L, Alabsi H, Getachew R, Narasimhan R, et al. 2013. Colony-stimulating factor 1 receptor (CSF1R) signaling in injured neurons facilitates protection and survival. J Exp Med 210: 157–172. 10.1084/jem.20120412 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Luo Y, Coskun V, Liang A, Yu J, Cheng L, Ge W, Shi Z, Zhang K, Li C, Cui Y, et al. 2015. Single-cell transcriptome analyses reveal signals to activate dormant neural stem cells. Cell 161: 1175–1186. 10.1016/j.cell.2015.04.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Magnusson JP, Zamboni M, Santopolo G, Mold JE, Barrientos-Somarribas M, Talavera-Lopez C, Andersson B, Frisén J. 2020. Activation of a neural stem cell transcriptional program in parenchymal astrocytes. eLife 9: e59733. 10.7554/eLife.59733 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Matson KJE, Russ DE, Kathe C, Hua I, Maric D, Ding Y, Krynitsky J, Pursley R, Sathyamurthy A, Squair JW, et al. 2022. Single cell atlas of spinal cord injury in mice reveals a pro-regenerative signature in spinocerebellar neurons. Nat Commun 13: 5628. 10.1038/s41467-022-33184-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Merrill BJ, Pasolli HA, Polak L, Rendl M, García-García MJ, Anderson KV, Fuchs E. 2004. Tcf3: a transcriptional regulator of axis induction in the early embryo. Development 131: 263–274. 10.1242/dev.00935 [DOI] [PubMed] [Google Scholar]
  45. Mizrak D, Levitin HM, Delgado AC, Crotet V, Yuan J, Chaker Z, Silva-Vargas V, Sims PA, Doetsch F. 2019. Single-cell analysis of regional differences in adult V-SVZ neural stem cell lineages. Cell Rep 26: 394–406.e5. 10.1016/j.celrep.2018.12.044 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Mychasiuk R, Farran A, Esser MJ. 2014. Assessment of an experimental rodent model of pediatric mild traumatic brain injury. J Neurotrauma 31: 749–757. 10.1089/neu.2013.3132 [DOI] [PubMed] [Google Scholar]
  47. Park J, Shrestha R, Qiu C, Kondo A, Huang S, Werth M, Li M, Barasch J, Suszták K. 2018. Single-cell transcriptomics of the mouse kidney reveals potential cellular targets of kidney disease. Science 360: 758–763. 10.1126/science.aar2131 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Pluchino S, Nicaise AM. 2021. NSCs: sentinel cells of the brain. Cell Stem Cell 28: 177–179. 10.1016/j.stem.2020.11.016 [DOI] [PubMed] [Google Scholar]
  49. R Core Team. 2023. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. https://www.R-project.org/. [Google Scholar]
  50. Seshadri S, Kamiya A, Yokota Y, Prikulis I, Kano S, Hayashi-Takagi A, Stanco A, Eom TY, Rao S, Ishizuka K, et al. 2010. Disrupted-in-Schizophrenia-1 expression is regulated by β-site amyloid precursor protein cleaving enzyme-1–neuregulin cascade. Proc Natl Acad Sci 107: 5622–5627. 10.1073/pnas.0909284107 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Shah PT, Stratton JA, Stykel MG, Abbasi S, Sharma S, Mayr KA, Koblinger K, Whelan PJ, Biernaskie J. 2018. Single-cell transcriptomics and fate mapping of ependymal cells reveals an absence of neural stem cell function. Cell 173: 1045–1057.e9. 10.1016/j.cell.2018.03.063 [DOI] [PubMed] [Google Scholar]
  52. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T. 2003. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 13: 2498–2504. 10.1101/gr.1239303 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Silverberg ND, Duhaime A-C, Iaccarino MA. 2020. Mild traumatic brain injury in 2019-2020. JAMA 323: 177–178. 10.1001/jama.2019.18134 [DOI] [PubMed] [Google Scholar]
  54. Somebang K, Rudolph J, Imhof I, Li L, Niemi EC, Shigenaga J, Tran H, Gill TM, Lo I, Zabel BA, et al. 2021. CCR2 deficiency alters activation of microglia subsets in traumatic brain injury. Cell Rep 36: 109727. 10.1016/j.celrep.2021.109727 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Stuart T, Butler A, Hoffman P, Hafemeister C, Papalexi E, Mauck WM III, Hao Y, Stoeckius M, Smibert P, Satija R. 2019. Comprehensive integration of single-cell data. Cell 177: 1888–1902.e21. 10.1016/j.cell.2019.05.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Volterra A, Meldolesi J. 2005. Astrocytes, from brain glue to communication elements: the revolution continues. Nat Rev Neurosci 6: 626–640. 10.1038/nrn1722 [DOI] [PubMed] [Google Scholar]
  57. Walker TL, Kempermann G. 2014. One mouse, two cultures: isolation and culture of adult neural stem cells from the two neurogenic zones of individual mice. J Vis Exp e51225. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Wang J, Hou Y, Zhang L, Liu M, Zhao J, Zhang Z, Ma Y, Hou W. 2021. Estrogen attenuates traumatic brain injury by inhibiting the activation of microglia and astrocyte-mediated neuroinflammatory responses. Mol Neurobiol 58: 1052–1061. 10.1007/s12035-020-02171-2 [DOI] [PubMed] [Google Scholar]
  59. Wyss-Coray T. 2006. Inflammation in Alzheimer disease: driving force, bystander or beneficial response? Nat Med 12: 1005–1015. [DOI] [PubMed] [Google Scholar]
  60. Xiao X, Jiang Y, Liang W, Wang Y, Cao S, Yan H, Gao L, Zhang L. 2019. miR-212-5p attenuates ferroptotic neuronal death after traumatic brain injury by targeting Ptgs2. Mol Brain 12: 78. 10.1186/s13041-019-0501-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Xie XP, Laks DR, Sun D, Poran A, Laughney AM, Wang Z, Sam J, Belenguer G, Fariñas I, Elemento O, et al. 2020. High-resolution mouse subventricular zone stem-cell niche transcriptome reveals features of lineage, anatomy, and aging. Proc Natl Acad Sci 117: 31448–31458. 10.1073/pnas.2014389117 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Xu L, Nguyen JV, Lehar M, Menon A, Rha E, Arena J, Ryu J, Marsh-Armstrong N, Marmarou CR, Koliatsos VE. 2016. Repetitive mild traumatic brain injury with impact acceleration in the mouse: multifocal axonopathy, neuroinflammation, and neurodegeneration in the visual system. Exp Neurol 275 Pt 3: 436–449. 10.1016/j.expneurol.2014.11.004 [DOI] [PubMed] [Google Scholar]
  63. Xu X, Cowan M, Beraldo F, Schranz A, McCunn P, Geremia N, Brown Z, Patel M, Nygard KL, Khazaee R, et al. 2021. Repetitive mild traumatic brain injury in mice triggers a slowly developing cascade of long-term and persistent behavioral deficits and pathological changes. Acta Neuropathol Commun 9: 60. 10.1186/s40478-021-01161-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Yi F, Pereira L, Hoffman JA, Shy BR, Yuen CM, Liu DR, Merrill BJ. 2011. Opposing effects of Tcf3 and Tcf1 control Wnt stimulation of embryonic stem cell self-renewal. Nat Cell Biol 13: 762–770. 10.1038/ncb2283 [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Zhao L, Han S, Su H, Li J, Zhi E, Li P, Yao C, Tian R, Chen H, Chen H, et al. 2022. Single-cell transcriptome atlas of the human corpus cavernosum. Nat Commun 13: 4302. 10.1038/s41467-022-31950-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Zywitza V, Misios A, Bunatyan L, Willnow TE, Rajewsky N. 2018. Single-cell transcriptomics characterizes cell types in the subventricular zone and uncovers molecular defects impairing adult neurogenesis. Cell Rep 25: 2457–2469.e8. 10.1016/j.celrep.2018.11.003 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1
Supplemental_Table_S1.docx (2.2MB, docx)
Supplement 2
Supplemental_Table_S2.docx (1.2MB, docx)
Supplement 3
Supplemental_Table_S3.docx (23.3KB, docx)
Supplement 4
Supplemental_Table_S4.docx (721.7KB, docx)
Supplement 5
Supplemental_Table_S5.docx (601.9KB, docx)
Supplement 6
Supplemental_Table_S6.docx (757.2KB, docx)
Supplement 7
Supplemental_Table_S7.docx (12.9KB, docx)
Supplement 8
Supplemental_Code.rar.zip (32.1KB, zip)
Supplement 9
Supplemental_Figures.pdf (14MB, pdf)

Articles from Genome Research are provided here courtesy of Cold Spring Harbor Laboratory Press

RESOURCES