Author response: Creating an atlas of the bone microenvironment during oral inflammatory-related bone disease using single-cell profiling
Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Oral inflammatory diseases such as apical periodontitis are common bacterial infectious diseases that may affect the periapical alveolar bone tissues. A protective process occurs simultaneously with the inflammatory tissue destruction, in which mesenchymal stem cells (MSCs) play a primary role. However, a systematic and precise description of the cellular and molecular composition of the microenvironment of bone affected by inflammation is lacking. In this study, we created a single-cell atlas of cell populations that compose alveolar bone in healthy and inflammatory disease states. We investigated changes in expression frequency and patterns related to apical periodontitis, as well as the interactions between MSCs and immunocytes. Our results highlight an enhanced self-supporting network and osteogenic potential within MSCs during apical periodontitis-associated inflammation. MSCs not only differentiated toward osteoblast lineage cells but also expressed higher levels of osteogenic-related markers, including Sparc and Col1a1. This was confirmed by lineage tracing in transgenic mouse models and human samples from oral inflammatory-related alveolar bone lesions. In summary, the current study provides an in-depth description of the microenvironment of MSCs and immunocytes in both healthy and disease states. We also identified key apical periodontitis-associated MSC subclusters and their biomarkers, which could further our understanding of the protective process and the underlying mechanisms of oral inflammatory-related bone disease. Taken together, these results enhance our understanding of heterogeneity and cellular interactions of alveolar bone cells under pathogenic and inflammatory conditions. We provide these data as a tool for investigators not only to better appreciate the repertoire of progenitors that are stress responsive but importantly to help design new therapeutic targets to restore bone lesions caused by apical periodontitis and other inflammatory-related bone diseases. Editor's evaluation Data from scRNA-Seq analysis demonstrated that acute inflammation stimulates periodontal stem cells to differentiate into osteoblast lineage cells to protect the alveolar bone. In murine models and patients with apical periodontitis, the genes and proteins associated with osteogenesis were enriched. The studies help us understand how MSCs respond to inflammation during apical periodontitis disease progression. https://doi.org/10.7554/eLife.82537.sa0 Decision letter Reviews on Sciety eLife's review process Introduction Oral diseases, particularly dental caries and periodontal diseases, affect 3.5 billion people worldwide (Disease et al., 2018). Untreated dental caries can directly lead to pulp necrosis and periapical lesions, resulting in apical periodontitis (AP). Individuals with at least one tooth affected by AP comprise up to 52% of cases, indicating that AP is a highly prevalent disease (Tibúrcio-Machado et al., 2021). Inflammation in the oral cavity can lead to destruction of surrounding periapical tissues and resorption of hard tissues, a consequence of the unbalanced interaction between infection and the immune response (Gazivoda et al., 2009; Márton and Kiss, 2014). Restoring and regenerating the destroyed periapical alveolar bone structures have always been a challenging task in clinical practice. Active inflammation, the tissue injury and the protective process all occur simultaneously in the setting of chronic AP (Márton and Kiss, 2014). Importantly, there is a complex assemblage of immune cell types involved in the pathogenesis, highlighting the importance of polymorphonuclear leukocytes, lymphocytes, and monocyte/macrophages in periapical defense (Braz-Silva et al., 2019; Nair, 2004). Notably, an increasing number of studies report the involvement of mesenchymal stem cells (MSCs) in the protective action that occurs during oral inflammatory diseases, whereby MSCs exert immunomodulatory effects and have regenerative potential (Li et al., 2014; Márton and Kiss, 2000; Nair, 2004). MSC markers such as CD44, CD73, CD90, CD106, and STRO-1 have been observed in human periapical inflammatory tissues (Estrela et al., 2019; Liao et al., 2011). Cells isolated from the inflamed periapical region were able to produce colony-forming unit-fibroblasts (CFU-Fs) with high-osteogenic capacity. It is also reported that interference with MSC mobilization toward the periapex region in an AP mouse model led to enlargement of lesions, accompanied by decreased wound healing markers and increased inflammatory cytokines (Araujo-Pires et al., 2014). These findings indicate the involvement of MSCs in the repair and regeneration of oral inflammatory-related bone lesions. They also suggest that MSCs present promising targets for treating bone lesions, with great potential for modulating inflammation and promoting tissue regeneration. However, most studies have surveyed whole tissues to understand the transcriptomic and cellular profile of these diseases. Specific cell populations and their regulatory molecules, as well as the interaction among different cell populations, remain far from clear. The advances in single-cell technologies offer an unbiased approach for identifying heterogeneous cell subsets, tracking the trajectories of distinct cell clusters and uncovering regulatory relationships between genes (Hwang et al., 2018; Tang et al., 2009). In this study, we collected mandibular alveolar bone samples from control and AP in mice and subjected them to single-cell RNA sequencing (scRNA-seq). The atlas of the mandibular alveolar bone explored the distinct cell subsets and their expression profiles relevant to AP. We also investigated the relationship between MSCs and immune cell subsets. The results reveal the role of a subset of MSCs in inflammation, which showed increased frequency and which formed a self-supporting network. Moreover, MSCs exhibited upregulated osteogenic potential, which was confirmed in transgenic mouse models and human patients with chronic AP. These results advance our understanding of heterogeneity and interactions of alveolar bone cells in the pathogenesis of inflammatory-related bone diseases. Defining key cellular subsets such as MSCs and their biomarkers in inflamed tissue will be important for identifying new therapeutic targets for oral inflammatory-related bone diseases. Results Single-cell transcriptional profiling identified 15 discrete populations in homeostasis and chronic AP samples Individual cells were isolated from alveolar bone of healthy mice and mice with AP. We modeled AP using a well-established AP mouse model in which the mandibular first molar pulp was exposed and subsequently developed chronic AP over a 3 wk period (Taira et al., 2019). Bar-coded cDNA libraries from individual cells were obtained using the 10× Genomics Chromium Controller platform (Zheng et al., 2017; Figure 1A). The combined libraries from healthy and AP alveolar bone contained 15,148 individual cells. The median value of feature_RNA was between 1000 and 2000 (Figure 1—figure supplement 1B). After quality control filtering and removal of the batch effect between batches, the t-stochastic neighbor embedding (t-SNE) method was applied to reduce the dimensionality. Seurat's unbiased cluster detection algorithm identified 15 distinct cell populations (Figure 1B and C). Cluster-specific transcripts were utilized to annotate cell types with classic markers as described in a previous study (Lin et al., 2021). These included B cell (Cd79a), hematopoietic stem cell (HSC) (Cd34), MSC (Col1a1), natural killer (NK) cell (Klrd1), T cell (Cd3g), dendritic cell (Siglech), epithelial cell (Epcam), erythrocyte (Hbb-bt), macrophage (Adgre1), mast cell (Fcer1a), megakaryocyte (Gp1bb), monocyte (Ly6c2), myeloid progenitor (Mpo), neutrophil (S100a8), and pre-B cell (Vpreb1; Figure 1E and F). The top 20 enriched genes in each defined cluster were identified and compared (Figure 1G). Figure 1 with 1 supplement see all Download asset Open asset Identification of the single-cell atlas of alveolar bone using scRNA-sequencing (scRNA-seq) and unbiased clustering. (A) Schematic diagram of the experimental design. (B–C) t-Stochastic neighbor embedding (t-SNE) representation of aligned gene expression data in single cells extracted from mandibles of control mice (n=8340) and apical periodontitis (AP) mice (n=6808) showing 15 distinct clusters and cellular origin. (D) Relative abundance of 15 cell populations composing alveolar bone under healthy and AP conditions. (E) Expression of gene markers in distinct cell types. (F) Gene expression patterns projected onto t-SNE plots of marker genes. Scale: log-transformed gene expression. (G) Heatmap showing the 20 most upregulated genes (ordered by decreasing Padj value) in each cluster defined in B. Scale: log2 fold change. AP led to significant changes in frequency and transcriptional expression of cell populations All the identified cell clusters were present in both AP and control samples, but there were significant differences in the cellular compositions of particular clusters. T cell, B cell, NK cell, macrophage, epithelial cell, and MSC had significantly increased frequency in AP samples. Neutrophil, myeloid progenitor, monocyte, megakaryocyte, mast cell, HSC, and dendritic cell were markedly decreased (Figures 2A and 1D). Figure 2 Download asset Open asset Changes in frequency and transcriptional expression pattern in each cell population from control and apical periodontitis (AP) groups. (A) Bar plot of cells per cluster (AP versus control). Normalization to overall number of inputs per condition. Fisher's exact test with Bonferroni correction was used. *p<0.05, **p<0.01, and ****p<0.0001. All data were exhibited as mean ± SEM. (B) Violin plots of cluster-specific expression of representative genes. (C–G) Violin plots showing genes that significantly changed in each cluster from control and AP. AP is a complex inflammatory process involving innate and adaptive immune responses (Cotti et al., 2014). A variety of inflammatory cells such as neutrophils, mast cells, monocytes, macrophages, and lymphocytes are involved in periapical lesions, highlighting the direct involvement of the immune response in the pathogenesis of AP (Nair, 2004). Neutrophils are important components in the acute phase of AP as a first line of defense. But they are also important in the progression of AP by interacting with microorganisms, leading to tissue damage and chemotaxis (Braz-Silva et al., 2019). Single-cell differential expression analysis revealed that the most significantly enriched genes in neutrophils were various proinflammatory chemokines and cytokines. These included C-X-C motif chemokine ligand 2 (Cxcl2), C-C motif chemokine ligand 6 (Ccl6), NLR family pyrin domain containing 3 (Nlrp3), and Interleukin-1β (Il1b). Notably, we found that C-C motif chemokine receptor like 2 (Ccrl2) was upregulated in neutrophils during AP (Figure 2D). It is responsible for the innate defense against pathogens and is also involved in the regulation of neutrophil migration (Del Prete et al., 2017; Kolaczkowska and Kubes, 2013; Mantovani et al., 2011). Mast cells, monocytes, and macrophages have critical roles in the inflammatory infiltrate during chronic AP (Braz-Silva et al., 2019). The production of Interleukin-6 (Il6) was present in these cell populations with the highest expression level in mast cells. The pro-inflammatory cytokine IL-1β is a key mediator of host response to microbial infection and is associated with the persistence of AP (Morsani et al., 2011; Ng et al., 2008). We found Il1b transcripts in a series of cell types, such as monocyte, macrophage, mast cell, and neutrophil. Of these, macrophages had the highest Il1b expression. Another major cytokine, tumor necrosis factor (Tnf; Cotti et al., 2014), was detected in immunoresponsive cell clusters, such as monocyte, macrophage, mast cell, myeloid progenitor, neutrophil, and HSC, with the highest expression observed in the monocyte population (Figure 2B). Furthermore, gene signatures from monocytes showed that the interferon-induced transmembranes (IFITMs) protein 1 and 2 (Ifitm1 and Ifitm2; Figure 2E) were upregulated the most during AP. These factors have been associated with signal transduction of anti-inflammation activity in the immune system (Yánez et al., 2020). We also detected upregulated expression levels of Ccl9 in the monocyte population from AP. Ccl9 is an important cytokine and is involved in the survival of osteoclasts during the destruction of the periapical bone (Silva et al., 2007). Also, genes coding for pro-inflammatory calcium-binding S100 family proteins such as S100a9 and S100a11 had increased expression in the AP monocyte cluster (Figure 2E). Previous reports demonstrated that macrophages are capable of secreting pro- and anti-inflammatory substances which act on the development and repair of the AP lesions (Italiani and Boraschi, 2014; Shapouri-Moghaddam et al., 2018). Indeed, several genes encoding pro-inflammatory mediators, including Cxcl2, Cxcl16, Il1a, and Ptgs2, were upregulated in macrophages from AP samples (Figure 2C). Expression of anti-inflammatory-associated genes such as Ifitm1 and Ifitm2 was significantly increased in the AP macrophage cluster compared to control cells. Furthermore, Fcγ receptor IIB (Fcgr2b) was markedly upregulated in macrophages from AP samples. Fcgr2b is expressed in most tissue-resident macrophages (Gautier et al., 2012) and functions to inhibit Fcγ-dependent phagocytosis. It also inhibits release of cytokines such as IL-6, TNF-α, IL-1α, as well as neutrophil chemotactants (Clatworthy and Smith, 2004; Espéli et al., 2016). In addition, expression of Apolipoprotein E (Apoe), which can suppress the pro-inflammatory response (Jofre-Monseny et al., 2007), was significantly increased in the macrophage population (Figure 2C). These data indicated the activation of anti-inflammatory factors by macrophages during local inflammation by AP. The major classes of lymphocytes are T lymphocytes, B lymphocytes, and the NK cells. T and B lymphocytes comprise the majority of the inflammatory infiltrate in AP (Graunaite et al., 2012). A significant increase in the expression of inflammatory-associated genes, such as Srgn, Emb, Ctla4, and Il7r could be observed in the AP T cell population (Figure 2F). In the AP B lymphocytes cluster, inflammation-responsive genes (Fth1, Ftl1, Ebf1, and mt-Nd5) were upregulated (Figure 2G). Interestingly, Lars2, the gene encoding a mitochondrial leucyl tRNA synthase (Carminho-Rodrigues et al., 2020; 't Hart et al., 2005), was significantly upregulated in T and B lymphocytes, indicating changes in mitochondrial metabolism in both clusters. Inflammation induces osteoclasts differentiation leading to periapical alveolar bone destruction Bone destruction is a major pathological factor in chronic inflammatory diseases such as AP. Various cytokines including TNF-α, IL-1α, and IL-6 were released by immunocytes to recruit the osteoclast precursors and induce the maturation of osteoclasts (Lyu et al., 2022). We have detected osteoclast markers including Ctsk, Acp5, Mmp9, and Nfatc1 by scRNA-seq. Moreover, Csfr1, Cx3cr1, Itgam, and Tnfrs11a were used to identify osteoclast precursors. Markers of osteoclast and osteoclast precursors were highly expressed in the clusters of monocyte and macrophage (Figure 3A and B). Gene Ontology (GO) analysis showed that inflammation related immune reactions and bone resorption activity were significantly enriched in macrophage cluster (Figure 3C). To further study the differential trajectory of osteoclasts, pseudotime analysis was performed for the clusters of macrophage and monocyte. Two independent branch points were determined, and five monocyte/macrophage subclusters were scattered at different branches in the developmental tree (Figure 3D and G). The results showed that the monocyte cluster differentiated into the macrophage cluster (Figure 3E). During this trajectory, the gene expression pattern across pseudotime showed that osteoclastic genes, such as Ctsk, Acp5, Mmp9, Atp6v0d2, and Dcstamp, were progressively elevated (Figure 3F). Of note, we have observed a branch which was highly positive for Ctsk and Acp5 (Figure 3H), indicating the mature osteoclasts were differentiated from monocyte/macrophage lineage and contributed to inflammatory bone resorption during AP. We have also analyzed the expression of osteoclast related genes using the bulk RNA-seq library built on mandibular samples extracted from mice with AP. Markers of osteoclast and osteoclast precursors were significantly upregulated, confirming the osteoclasts activity in the inflammatory-related bone lesion (Figure 3I). Figure 3 with 1 supplement see all Download asset Open asset Inflammatory-related bone resorption under apical periodontitis (AP) situation. (A) The expression levels of markers of osteoclasts and osteoclast precursors. (B) Violin plots of the expression of osteoclastogenesis genes. (C) Gene Ontology (GO) enrichment analysis of the biological functions of macrophage cluster. (D) Trajectory order of the monocyte/macrophage populations by pseudotime value. (E) The differentiation trajectory of monocyte and macrophage clusters presented on a t-stochastic neighbor embedding (t-SNE) visualization. (F) The expression patterns of osteoclast markers during the trajectory of monocyte/macrophage populations. (G) Distribution of monocytes/macrophages on the developmental tree by clusters. (H) Heatmap of differential genes of three states. (I) Heatmap of genes associated with osteoclastogenesis in bulk RNA-seq analysis. AP leads to reduced transcriptionally inferred cellular interactions with an increased self-supporting network in MSCs We next sought to characterize the cell-cell communication related to the perturbation of signaling pathways detected in the AP samples by employing CellphoneDB (Efremova et al., 2020; Nagai et al., 2021). We identified a close interaction among MSC, macrophage, and dendritic cells under homeostasis conditions (Figure 4A). A similarly close communication was found among MSC, macrophage, and dendritic cells under inflammatory conditions (Figure 4B). Next, we compared the differential cell-cell interaction (CCI) network between AP and control samples using CrossTalkeR (Nagai et al., 2021). The results suggested that AP is associated with an overall decrease in cellular interactions. However, MSC intercellular communication with mast cells and monocytes is upregulated and accompanied by the highest number of interactions within MSCs themselves (Figure 4C). These results indicate that, although cell populations lose their normal physiological interactions, MSCs were able to establish a self-interacting network and coordinate with certain types of cells during chronic AP. Next, we ranked the individual ligands by the number of their interactions. Inflammatory-related proteins (Lgals9, Tnf, and Ccl4), extracellular matrix protein (Fn1), and protein involved in biomineralization (Spp1) were among the highest interactions. Also, Tgfb1, Vegfb, and Vegfa were in the top 10 most abundant ligands (Figure 4D). Bar plots were also generated to display the top 10 upregulated gene/cell pairs, showing that inflammation associated genes (Tnf/neutrophil, Ccl3/mast cell, Ccl3/monocyte, and Il1b/macrophage) and matrix related genes (Sele/MSC, Fn1/MSC, and Fn1/monocyte) were the most influential ligands during AP when compared to control state (Figure 4E). We used a Sankey plot to further focus on MSC-related interactions (Figure 4F). The results indicated that Sele was primarily directed by MSC toward the MSC cluster via multiple receptors including Glg1, Selplg, and Cd44. Moreover, Fn1 was secreted by MSC, monocyte, macrophage, and mast cells toward MSCs (Figure 4G). Of note, Sele and Fn1 are important cell adhesion molecules that mediate cell homing and migration (Frenette et al., 1998; To and Midwood, 2011). This was in accordance with the increased frequency of the MSC cluster (Figure 2A) as well as the upregulated cellular interaction among MSCs and other cell populations. Figure 4 Download asset Open asset Apical periodontitis (AP) suppressed transcriptionally inferred cellular interactions and increased a self-supporting network within the mesenchymal stem cell (MSC) cluster. (A–C) Network plot of ligand-receptor activity in control (A), AP (B), and AP versus control (C). (D) Bar plot of top 10 most abundant ligands in all inferred ligand-receptor interactions. (E) Ranking of ligand/source regarding communication gains in AP state. (F and G) Sankey plot listing all predicted source, receptor, and receiver interactions associated with Fn1 and Sele. scRNA-seq based identification of AP-associated MSC population MSCs represented the non-immune cell population in alveolar bone, constituting 1.76% of total identified cells. This cell population could be decomposed into four subclusters (Figure 5A). The most dominant subcluster was characterized by high expression of Prrx1, platelet-derived growth factor receptor β (Pdgfrb), and hematopoiesis supporting factors such as C-X-C motif chemokine 12 (Cxcl12) and angiopoietin (Angpt1). It was also characterized by osteogenic-related markers, such as Runt-related transcription factor 2 (Runx2), Sp7, and was thus classified as the MSC_osteolineage cells (OLCs). The other subclusters were identified as MSC_endothelial and Figure and B). The cell composition of the MSC cluster indicated an of subcluster during the subcluster was reduced (Figure Of note, we performed lineage tracing using and transgenic mouse models to identify stem cells in alveolar bone and the surrounding periodontal tissues et al., 2020; et al., et al., 2020; et al., 2020). The results suggested increased of and as well as upregulated and periodontal stem cells in AP lesions (Figure supplement 1A). We compared the MSC marker genes, identified among the four enrichment analysis confirmed that there are four MSC populations. were enriched for tissue system and osteoblast MSC_endothelial cells enrichment for and cells were enriched for of and suggested highly of cells, including regulation of protein from mitochondrial and (Figure supplement Figure with 3 see all Download asset Open asset Identification and of apical periodontitis mesenchymal stem cell (MSC) population and (A) t-Stochastic neighbor embedding (t-SNE) representation an of single cells within the MSC cluster. (B) Violin plots of MSC expression of representative genes. (C) The of four subclusters of MSC population were in control and AP groups. (D) Violin plots the changes in the expression of top upregulated genes in the MSC cluster. and of Sparc (E) and (F) in (G) of in periodontal stem cells (H) tracing analysis of and AP stimulates MSC differentiation toward osteoblast lineage cells We next compared the data from the MSC cluster between AP and control groups. the top upregulated genes, we observed a significant in the single cell expression levels of such as secreted protein that is and in and accompanied by a toward increased expression (Figure We their by and bulk RNA analysis. The results revealed that and were significantly upregulated, and to increase during AP (Figure supplement B). In confirmed the upregulated expression of osteogenic-related markers, such as Sparc and in the AP were increased of cells in the bone the protective of MSC to differentiate into and under AP conditions (Figure and F). tracing analysis further revealed that to AP lesions and differentiated into (Figure Moreover, using mouse we identified cells in the alveolar bone, confirming that inflammation could the progenitor cells in the bone and toward thus to the protective during AP (Figure We investigated the heterogeneity of MSCs by branch expression analysis and pseudotime analysis in 2 et al., The results revealed three distinct states. Of these, state 2 and state 3 represented differentiated cell populations Figure We the of the MSCs with to their states. We the pseudotime of each cluster cell in the state and the results indicate that with MSC_endothelial cells, MSCs were able to differentiate into and cells. cells may to an state (Figure Interestingly, in state as osteogenic exhibited the highest toward differentiation with upregulated of and (Figure Notably, this MSC population increased significantly in the AP (Figure A of the of the gene expression pattern across pseudotime revealed of osteogenic genes such as and during inflammation when compared to the state (Figure the MSC subcluster exhibited gene expression levels of osteogenic markers that increased during AP. These results were in accordance with our previous single-cell analysis of gene expression in the AP (Figure Figure 6 Download asset Open asset Apical periodontitis (AP) stimulates mesenchymal stem cell (MSC) differentiation toward osteoblast lineage cells. (A) analysis of the MSC (B) cell differentiation trajectory of MSC populations in control and AP groups. (C) Heatmap of differential genes of three states. (D) Bar plot of changes between AP and control in as identified in pseudotime analysis. Fisher's exact test with Bonferroni correction was used. ****p<0.0001. All data were exhibited as mean ± SEM. (E) between the pseudotime gene trajectories of MSC showed of and line and indicate marker levels were identified in alveolar bone from patients with AP We next osteogenesis within the MSC population was in human patients with AP. The of and were significantly enhanced in alveolar bone from patients with accompanied by in and (Figure Moreover, detected a of and in AP alveolar bone confirming higher osteogenic potential under AP conditions (Figure of markers in AP lesions from patients is with our previous an of osteogenic within the MSC subcluster in alveolar bone under AP conditions. Figure Download asset Open asset alveolar bone in apical periodontitis (AP) is associated with higher subcluster (A) Gene expression of and in samples of healthy and AP patients showed an upregulated expression in AP in healthy and in AP All data are as the mean ± SEM. (B–C) of and in bone of human alveolar bone exhibited high osteogenic protein levels in AP. 10 In summary, the current study the of MSCs and immunoresponsive cells under healthy and chronic AP including heterogeneity in their of We that this analysis a
Read more
18
