Cellular basis of accelerated whole-tooth regeneration
- School of Biological Sciences, Georgia Institute of Technology, United States
- Institute of Bioengineering and Bioscience, Georgia Institute of Technology, United States
- Department of Medicine, University of California, San Francisco, United States
Figures
Figure 1 with 3 supplements
Tooth replacement is accelerated after plucking in cichlid species with divergent dentitions.
(A) Schematic of the modified pulse-chase method using Alizarin and Calcein to quantify tooth replacement. (B) Representative fluorescence images used to classify teeth by dye incorporation. Teeth positive for both Alizarin and Calcein were scored as pre-existing old teeth, whereas teeth positive for Calcein only were scored as newly formed teeth. Left, whole jaw from MZ Red, with teeth plucked on the right side and the left side serving as the control; the dotted outline marks the region used for plucking and quantification (Figure 1—figure supplement 1B). Right, representative individually classified teeth (Alizarin, magenta; Calcein, green), with the top row showing pre-existing old teeth positive for both dyes and the bottom row showing newly formed teeth positive only for Calcein. (C) Pulse-chase reveals an elevated rate of tooth replacement on the plucked side in CA, MZ Red, and PT. Tooth gain is shown as the normalized replacement rate (new teeth/total teeth) ***p-value<0.0001, paired Student’s t-test; CA (n=8), MZ Red (n=8), and PT (n=9). (D) Rate of tooth gain is consistent across species after plucking. NS, p-value>0.05, one-way ANOVA with Tukey–Kramer post hoc test; CA (n=8), MZ Red (n=8), and PT (n=9).
Figure 1—figure supplement 1
Schematic overview of tooth-plucking experiments.
(A) Following Alizarin Red staining, the majority of teeth on the right side of both the upper and lower jaws were removed. Tooth plucking was restricted to the anterio-lateral region of the jaw, as demarcated by the dotted square. Tooth removal was performed under anesthesia as described in the ‘Materials and methods’. (B) Representative lower jaw from Metriaclima estherae (MZ Red) 15 days post pluck-control experiment. Dotted outlines indicate regions used for quantification: white denotes the plucked side, and red indicates the unmanipulated control side. These regions correspond to the areas used for quantifying newly formed and old teeth on both sides of the jaw. Whole-jaw imaging was acquired using Zeiss Stereo Discovery.V12. Teeth were individually classified and counted using Zeiss Axio Observer Z1 fluorescence.
Figure 1—figure supplement 2
Micro-CT images of cichlid species exhibiting diverse tooth formulae.
(A–C) Cynotilapia afra (CA), (D–F) Metriaclima estherae (MZ Red), and (G–I) Petrotilapia chimba (PT; thick bar). Panels (A), (D), and (G) show combined views of the upper and lower jaws; panels (B), (E), and (F) show upper jaws; and panels (C), (F), and (I) show lower jaws.
Figure 1—figure supplement 3
Pulse-chase reveals elevated rates and number of new tooth formation on the plucked side.
(A–C) Number of newly formed teeth in Cynotilapia afra (CA), Metriaclima estherae (MZ Red), and Petrotilapia chimba (PT), respectively. ***p-value<0.0001, paired Student’s t-test; sample sizes: CA (n=8), MZ Red (n=8), and PT (n=9). (D) Rate of tooth gain at 15 days was similar across species on the control side. (E) Comparison of the total number of newly formed teeth on plucked side across species. (F) Comparison of total number of newly formed teeth on control side across species. (D–F) ***p-value<0.0001, one-way ANOVA with Tukey–Kramer post hoc test; sample sizes: PT (n=9), CA (n=8), and MZ Red (n=8).
Figure 2 with 2 supplements
Single-nucleus RNA-seq of cichlid replacement teeth.
(A) Schematic of the snRNA-seq experimental and processing workflow. (B) Uniform manifold approximation and projection (UMAP) of 27,114 nuclei. Points that are close together represent nuclei with similar gene expression profiles. The color of each point represents the cell type. SI + SR, stellate reticulum and stratum intermedium; OEE, outer enamel epithelium; VEE, ventral enamel epithelium; ES-1, epithelial supporting cells 1; ES-2, epithelial supporting cells 2; Pre-AMB, pre-ameloblasts; CYC-AMB, cycling ameloblasts; M-AMB, maturing ameloblasts; PM-AMB, post maturation ameloblasts; dPLP, distal pulp; DF, dental follicle; DP, dental papilla; Pre-ODO, pre-odontoblasts; ENDO, endothelium; NT, neutrophils; NV, nerves; MP, macrophages; Bone-MP, bone macrophages; NK/T, natural killer and T lymphocytes; B, B lymphocytes; OST, bone; GLIA, glia; PC, pericytes; TB, taste buds; SM, smooth muscle cells. Canonical marker genes used to identify major cell classes include pitx2 (dental epithelium), spp1 (bone), pax9 (dental mesenchyme), and aif1 (immune) (Supplementary file 1). (C) Expression of marker genes used for identification of major cell types in (B).
Figure 2—figure supplement 1
Single nuclei sampling, library metrics, and cluster composition.
(A) Uniform manifold approximation and projection (UMAP) of 27,114 dental cells. Points that are close together represent cells with similar gene expression profiles. The color of each point represents the Seurat clusters (n=50). (B) Cluster composition is consistent across timepoints and between conditions. Individual UMAP plots of nuclei sampled from all 10 test subjects, organized into four pairs with plucked subjects shown in the top row and paired control subjects shown in the bottom row. (C) Sequence metrics across nuclei for each test subject (x-axis), including total number of Unique Molecular Identifiers (UMIs, top panel), total number of unique genes (middle panel), and percentage of genes that were mitochondrial (bottom panel). In box plots, the center line indicates the median, the bounds of the box indicate the upper and lower quartiles, and whiskers indicate 1.5× interquartile range. (D) Sequence metrics are consistent between conditions and across test subjects. Median sequence metrics across test subjects for each experimental condition (x-axis), including median number of UMIs (top panel), median number of unique genes (middle panel), and median percentage of genes that were mitochondrial (bottom panel). (E) Stacked bar chart showing consistent sampling of clusters across test subjects. (F) UMAP embedding of dental nuclei. Each point represents a single nucleus, colored by test subject. (G) Representative differentially expressed genes show consistent expression differences between plucked and control samples across biological replicates. Paired boxplots of average gene expression for representative differentially expressed genes from multiple cell populations at days 0, 1, 3, and 7. Each point represents one biological replicate (test subject), with paired plucked and control samples connected by dashed lines. The y-axis shows average gene expression, and the x-axis indicates the experimental condition. These representative examples illustrate the consistent direction of differential expression across biological replicates, supporting the reproducibility of the single-nucleus RNA-seq dataset.
Figure 2—figure supplement 2
SAMap-based identification of homologous cell types between cichlid dental tissue and mouse incisor atlases.
(A) Similarity scores between cichlid clusters (x-axis) and mouse cell types (y-axis) in Sharir et al., 2019, which profiled only the proximal region of the mouse incisor epithelium. (B) Similarity scores between cichlid clusters (x-axis) and mouse cell types (y-axis) in Krivanek et al., 2020, which sequenced the entire mouse incisor, including both epithelial and mesenchymal components. (A, B) The heatmap scale indicates degree of overlap, with yellow representing higher homology between matched cell types. The dots denote significance determined by permutation testing (qperm <0.01, nperm = 1000).
Figure 3 with 1 supplement
Transcriptional programs of developmental potential in replacement tooth epithelium.
(A) CellRank 2 projected velocity fields showing inferred cell–cell transition dynamics in epithelium using CytoTRACE-derived developmental potential. Arrows indicate the predicted direction of state change in the UMAP embedding. (B) UMAP overlay with SL (left) and ES (right) cell distribution, identified by co-expression of canonical marker genes curated from the literature. (C) The developmental trajectory of dental epithelial cells is analyzed via pseudotime alignment. (D) Gene expression dynamics along pseudo-differentiation reveal distinct transcriptional modules, with corresponding GO enrichment for each module. (E) Schematic diagram summarizing lineage specification from ES/SL cells.
Figure 3—figure supplement 1
Developmental trajectory of the epithelial subpopulation.
(A) Epithelial subtype composition is consistent across timepoints and between conditions. Individual UMAP plots of nuclei sampled from all 10 test subjects, arranged in four columns by timepoint. Each column represents a matched pair, with the plucked subjects displayed in the top row and the corresponding control subjects in the bottom row. (B, C) UMAP embeddings of control (left) and plucked (right) epithelial cells colored by pseudotime. Cells appearing earliest in pseudotime are denoted by dark purple, and those latest in pseudotime are in yellow.
Figure 4 with 2 supplements
Transcriptional programs of developmental potential in replacement tooth mesenchyme.
(A) CellRank 2 projected velocity fields showing cell–cell transition dynamics in mesenchyme using CytoTRACE-derived developmental potential. Arrows indicate the predicted direction of state change in the UMAP embedding. (B) Gene-weighted density estimates overlaid on the UMAP showing expression distributions of twist1, twist2, dnmt1, and runx2 across mesenchymal states. (C) The developmental trajectory of mesenchymal cells is analyzed via pseudotime alignment. (D) Gene expression dynamics along pseudo-differentiation identify distinct transcriptional modules, with corresponding GO enrichment for each module. (E) Schematic diagram summarizing lineage specification from mesenchymal progenitor cells.
Figure 4—figure supplement 1
Developmental trajectory of the mesenchymal subpopulation.
(A) Mesenchymal subtype composition is consistent across timepoints and between conditions. Individual UMAP plots of nuclei sampled from all 10 test subjects, arranged in four columns by timepoint. Each column represents a matched pair, with the plucked subjects displayed in the top row and the corresponding control subjects in the bottom row. (B, C) UMAP embeddings of control (left) and plucked (right) mesenchymal cells colored by pseudotime. Cells appearing earliest in pseudotime are denoted by dark purple, and those latest in pseudotime are in yellow. (D) Gene expression dynamics along pseudo-differentiation identify distinct transcriptional modules, with corresponding GO enrichment for each module.
Figure 4—figure supplement 2
Identification of a DEM-like mesenchymal population and its signaling dynamics.
(A) UMAP of mesenchymal cell subpopulations. (B) Gene-weighted density estimates overlaid on the UMAP showing expression distributions of Lum, Col6a3, Aspn, and Vegfc across mesenchymal states. (C) Dot plot of DEM-like marker gene expression. (D) Heatmap of differential interaction strength among cell populations between plucked and control shows the outgoing and incoming signaling change associated with each cell group at day 0 (where sufficient replicates and cells were available for robust analysis). Red (positive values) represents increased signaling in plucked and blue (negative values) represents decreased signaling in plucked, between cell groups, as compared to control. The color scale indicates the magnitude of differential interaction strength. (E) Signaling pathways between each sender–receiver cell type pair were ranked by differences in communication probability between conditions. Pathways detected exclusively in plucked are shown in red, whereas pathways detected exclusively in control are shown in blue. Pathways shown in black were detected in both conditions but exhibited significant differences in communication probability between conditions.
Figure 5 with 2 supplements
Gene expression dynamics across cell types following tooth plucking.
(A) Differentially expressed genes identified within each cell type, comparing plucked to control across time points. Red and blue dots highlight the top 2–3 up-regulated and down-regulated DEGs, respectively. (B) Stage-resolved expression dynamics of DEGs across the time course. The size of each dot is proportional to the percentage of cells expressing DEGs. Purple and green dots indicate DEGs showing higher and lower expression, respectively, in plucked, as compared to control. (C) Time-course GO biological processes reveal enriched functional profiling at each developmental stage. The size of each dot is proportional to the number of genes contributing to each enriched term, and the color scale indicates statistical significance.
Figure 5—figure supplement 1
Compositional analysis of cichlid dental cell types across timepoints using single-nucleus RNA sequencing.
(A) UMAP Cacoa case–control compositional loadings (x-axis) for the cell types (y-axis) each timepoint. Positive loadings correspond to over-representation in the plucked cells and negative loadings mean over-representation in control cells. The uncertainty of the loading coefficients obtained by resampling with 1000 bootstrapping was represented using boxplot, where the boxes are IQRs split by the median (middle line) and the whiskers represent minimum and maximum loading coefficients. The red horizontal line separates cell types passing significance threshold (p-value <0.05 after Benjamini–Hochberg correction). (B, C) Gene-weighted The scCODA v.0.1.9 was used for compositional analysis of the epithelial subpopulation (B) and mesenchymal subpopulation (C). The false discovery rate (FDR) value was set to 0.4 to be able to detect subtle yet biologically relevant changes, as described by the authors in their documentation. In all boxplots, the central line denotes the median, boxes represent the IQR, and whiskers show the distribution except for outliers. Outliers are all points outside 1.5× the IQR. In all barplots, a value of zero means that no statistically credible effect was detected. For a value other than zero, a credible change was detected. A positive sign indicates increased abundance in plucked cells, and a negative sign denotes decreased abundance relative to control cells.
Figure 5—figure supplement 2
CytoTRACE-predicted cell state in epithelial and mesenchymal subpopulations.
Box-and-whisker plots show the predicted ordering based on CytoTRACE scores for individual cells across (A) epithelial and (B) mesenchymal subpopulations by condition. CytoTRACE scores range from 0 to 1, where higher scores indicate greater developmental potential (less differentiated states) and lower scores indicate more mature, differentiated states.
Figure 6
Cell–cell communication during the time course of accelerated tooth replacement.
(A) Heatmap of differential interaction strength among cell populations between plucked and control shows the outgoing and incoming signaling change associated with each cell group. Red (positive values) represents increased signaling in plucked and blue (negative values) represents decreased signaling in plucked, between cell groups, as compared to control. The color scale indicates the magnitude of differential interaction strength.
Figure 7
Signaling pathways and ligand–receptor interactions during accelerated tooth replacement.
(A) Signaling pathways between each sender–receiver cell type pair were ranked by differences in communication probability between conditions. Pathways detected exclusively in plucked are shown in red, whereas pathways detected exclusively in control are shown in blue. Pathways shown in black were detected in both conditions but exhibited significant differences in communication probability between conditions. (B) Up-regulated ligand–receptor interaction network in plucked compared to control. Signaling sources (senders) and targets (receivers) are shown on the bottom and top, respectively. Colored segments indicate their cell identity. Segment size is proportional to the total outgoing or incoming interaction strength associated with each ligand–receptor pair in the corresponding cell sub-population.
Figure 8
Summary of accelerated tooth replacement following plucking.
(A) Plucking accelerates tooth replacement, only on the plucked side, with a~3–4× increase in replacement rate that is conserved across cichlid species with widely divergent dentitions. (B) Conceptual timeline of biological processes activated following tooth removal that accompany accelerated tooth replacement. Early immune activation is associated with increased axonogenesis and vasculogenesis, followed by elevated odontogenesis and bone formation. (C) Schematic representation of prominent intercellular communication events and signaling pathways associated with accelerated tooth replacement after plucking. Red-labeled signaling pathways represent plucked-only signaling events identified in Figure 7A, corresponding to pathways shown as full red bars. Black-labeled signaling pathways represent interactions detected in both plucked and control conditions in Figure 7A but with significantly different interaction probabilities between conditions.
Author response image 1
Single-nucleus RNA-seq data demonstrate consistency across biological replicates in dental cell-state representation and differential gene expression.
(A) UMAP embedding of dental nuclei. Each point represents a single nucleus, colored by test subject. (B) Representative differentially expressed genes show consistent expression differences between plucked and control samples across biological replicates. Paired boxplots of average gene expression for representative differentially expressed genes from multiple cell populations at Days 0, 1, 3, and 7. Each point represents one biological replicate (test subject), with paired plucked and control samples connected by dashed lines. The y-axis shows average gene expression, and the x-axis indicates the experimental condition. These representative examples illustrate the consistent direction of differential expression across biological replicates, supporting the reproducibility of the single-nucleus RNA-seq dataset.
Additional files
-
Supplementary file 1
Summary of cell type annotations, cluster composition, marker genes, and functional enrichment analyses.
Summary of primary and secondary clusters, including the number of nuclei, cell type annotations, and top marker genes for each cluster (Sheet 1); number of nuclei assigned to each cluster for each subject (Sheet 2); cell type-specific marker genes identified using the MAST hurdle model with a 5% false discovery rate (FDR) adjustment (Sheet 3); and Gene Ontology (GO) functional enrichment analysis of cluster-specific marker genes using a one-sided hypergeometric test with a 5% FDR adjustment (Sheet 4).
- https://cdn.elifesciences.org/articles/110584/elife-110584-supp1-v1.xlsx
-
Supplementary file 2
Temporal gene expression dynamics and plucking-associated differential expression across cell types.
Results of linear mixed-effects regression modeling of temporal gene expression dynamics within each cell type (Sheet 1); and plucking-associated differentially expressed genes identified by comparison of plucked and control samples at day 0 (Sheet 2), day 1 (Sheet 3), day 3 (Sheet 4), and day 7 (Sheet 5).
- https://cdn.elifesciences.org/articles/110584/elife-110584-supp2-v1.xlsx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/110584/elife-110584-mdarchecklist1-v1.pdf
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Cellular basis of accelerated whole-tooth regeneration
eLife 15:RP110584.
https://doi.org/10.7554/eLife.110584.3
