Intradental mechano-nociceptors serve as sentinels that prevent tooth damage

Abstract

The trigeminal sensory neurons that innervate the tooth's vital interior-intradental neurons-are expected to drive severe pain, yet their contribution to healthy tooth sensation has not been explored. Here, we uncover a role for myelinated high-threshold mechano-nociceptors (intradental HTMRs) in tooth protection using in vivo Ca²⁺ imaging, opto-/chemogenetics, and the AI-driven behavioral analysis tool LabGym. Intradental HTMRs innervate the inner dentin through overlapping receptive fields and respond as the external structures of the tooth are damaged in the absence of either PIEZO2 or Na_v1.8. Whereas chemogenetic activation of intradental HTMRs results in a pain phenotype marked by facial and postural changes, their transient optogenetic activation triggers a rapid, jaw-opening reflex via contraction of the digastric muscle. Our work indicates that intradental HTMRs not only trigger pain but also protect the teeth by initiating a reflexive movement of the jaws when the teeth experience damage during chewing.

Keywords: CP: Neuroscience; craniofacial biology; dentin; interoception; mechanoreceptors; nociception; pain; sensorimotor reflex; somatosensation; tooth pulp biology; trigeminal.

Figure 1.. Functional imaging of intradental neurons reveals a large diameter population of putative myelinated HTMRs

(A) Schematic in vivo recording of the TG. Intradental neurons were identified using low-voltage pulsing (4 V, 0.6 Hz, 200 ms) of individual mandibular molars. Data in (A)–(F) were obtained from adult Ai95(RCL-GCaMP6f)-D (Ai95D) mice injected postnatally (P0-P3) with AAV9-Cre. (B–D) Electrical stimulation identified intradental neurons innervating a single molar. (B) Imaging field showing merged single frames of Ca²⁺ response (ΔF). Neurons from stimulation of either molar 1 (m1, red) or molar 2 (m2, cyan). Scale bar, 100 μm. (C) Example traces of GCaMP6f from m1 (red) and m2 (cyan) intradental neurons. Numbers correspond to (B). Vertical scale bars, 20 ΔF. Horizonal scale bar, 5 s. (D) Example heatmap from single TG showing responses of 18 intradental neurons. Scale bar, 20 s. Traces in (C) are neurons from (D). (E and F) (E) Bar graph showing mean neurons per trial or (F) mean diameter of neurons responding to ≥2 voltage pulses (=intradental neurons). n = 162 neurons from 10 mice. Plotted individual data points represent neurons in separate imaging trials. Error bars indicate the SEM. (G) Alignment of calcium responses with multiplexed whole-mount in situ hybridization (ISH) enables molecular classification of intradental neurons. Data were obtained from Ai95(RCL-GCaMP6f)-D (Ai95D) mice injected postnatally (P0-P3) with AAV9-hSyn1-mCherry-2A-iCre. mCherry expression was used to align fluorescence responses from Ca²⁺ imaging with ISH images. (Left) Multiplexed whole-mount ISH staining following alignment to GCaMP6f fluorescence. Probes: s100 calcium-binding protein B (S100b, blue), α-Nav1.8 (Scn10a, green), calcitonin gene-related peptide (CGRP; Calca, red), Mas-related G protein-coupled receptor member D (Mrgprd, white). (+), positive staining. Scale bar, 50 μm. (Right) Vertical bar (parts of a whole) indicates number and percentage of intradental neurons expressing selected genes. n = 5 mice, comprising 47 intradental neurons. (H and I) Representative image of retrograde-labeled intradental neurons (CTB-AF647) and ISH for Chrna7. (H) Left panel: CTB labeling; middle panel: Chrna7 expression; right panel: merge. Scale bar, 100 μm. Yellow arrowheads, intradental neurons co-positive for CTB-AF647 and Chrna7; white arrowheads, cells only positive for CTB. (I) Pie chart demonstrating co-expression. n = 4 mice, comprising 633 retrograde-labeled intradental neurons. See also Figure S1 and Video S1.

Figure 2.. Intradental neurons are HTMRs that encode direct force after enamel removal and innervate the dental pulp and inner dentin

(A–C) Intradental neurons do not respond to direct forces applied to the intact tooth. (A) Example trace (top) and associated heatmap (bottom). Stimuli used are indicated above the traces. Vertical scale bar, 10 ΔF. Horizontal scale bars, 20 s (top, trace) or 40 s (bottom, heatmap). (B) Bar graph showing proportion of intradental neurons that respond to direct force. Plotted individual data points represent proportion of force responsive/intradental neurons per trial. Bar shows the mean and error bars indicate the SEM. n = 3 mice, comprising 46 cells. (C) Bar graph showing proportion of intradental neurons responding to cold. Plotted individual data points represent the proportion of cold responsive/intradental neurons/trial. Bar shows the mean and error bars indicate the SEM. n = 3 mice, comprising 46 cells. Data for (A)–(C) were obtained from Ai95(RCL-GCaMP6f)-D (Ai95D) mice injected postnatally (P0) with AAV9-Cre. (D–F) Intradental neurons respond to direct forces following structural damage of the tooth. (D) Combined heatmap showing response profiles of 54 intradental neurons from 3 mice before (left) and following pulp exposure (right). All neurons represent intradental neurons as determined by electrical stimulation. De novo responders were not noted outside of the intradental neurons. Scale bar, 5 s. (E) Example heatmap showing the response of 23 intradental neurons to direct force applied to exposed dentin. Scale bar, 10 s. (F) Bar graph showing proportion of direct force response/intradental neurons. Plotted individual data points represent the proportion of force responsive/intradental neurons per trial. Bar shows the mean and error bars indicate the SEM. n = 6 mice for intact enamel, n = 6 mice for pulp exposed, n = 3 mice for dentin exposed. p < 0.001, p = 0.002, p > 0.05 using one-way ANOVA with Tukey’s correction as indicated with the graph. Data for (D)–(F) obtained from Ai95(RCL-GCaMP6f)-D (Ai95D) mice injected postnatally (P0-P3) with AAV9-Cre. (G) Representative image of immunostained neuronal terminals merged with bright-field image of dentin tubules in demineralized molars from S100b-Cre mice receiving TG injection of AAV9-flex-tdTomato for sparse labeling. n = 5 mice; Scale bar, 20 μm. (H) Representative image of immunostained neuronal terminals merged with bright-field image of dentin tubules in demineralized molars from Scn10a-Cre mice injected using a Cre-dependent AAV to induce sparse labeling of terminals (see STAR Methods). Enamel, a fully mineralized structure, has been removed completely by demineralization. This experiment was repeated in n = 4 mice with additional littermate and no primary controls. Scale bar, 20 μm. See also Figure S2.

Figure 3.. Intradental HTMRs detect enamel damage, independent of Piezo2 or Scn10a (Na_v1.8)

(A–E) Intradental HTMRs respond to cutting and frictional damage of tooth. (A) Example traces showing intradental neuron response to enamel cutting and friction applied to individual molars. Vertical scale bars, 20 ΔF. Horizontal scale bar, 20 s. (B) Example heatmap containing traces from (A). Scale bar, 40 s. (C) Bar graph showing proportion of friction and cutting responsive intradental neurons. Plotted individual data points represent proportion for each trial. Bar shows the mean and error bars indicate the SEM. n = 7 mice, comprising 205 total cells, p < 0.01, unpaired t test. (D) Line graph depicting magnitude of Ca²⁺ responses of intradental neurons to friction vs. cutting. Individual points represent area under the curve (AUC) during stimulation for each cell. n = 23 cells, p < 0.0001, paired t test. (E) Vertical bar (parts of a whole) indicates number and percentage of cutting responsive Scn10a+ neurons (81.4%, 22/27 neurons, n = 4 mice). Results in (E) from primary data also used in Figure 1G. (F and G) Example heatmap showing intradental HTMR responses to cutting persist in (F) Piezo2-cKO mice or (G) Scn10a-KO mice. Scale bars, 20 s. (H) Bar graph depicting the proportion of intradental HTMRs that respond to enamel cutting in wild-type, Piezo2-cKO, or Scn10a-KO mice. Plotted individual data points represent calculated proportion of enamel cutting responsive/intradental neurons. Bar shows the mean, and error bars indicate the SEM. n = 6 or 7 adult mice/genotype. n.s. p > 0.05 using one-way ANOVA with Tukey’s correction. See also Figure S3.

Figure 4.. Chemogenetic activation of intradental HTMRs elicits a marked pain phenotype

(A) Schematic demonstrating the experimental approach. The excitatory Gq receptor hM3Dq (CAG-LSL-Gq-DREADD) was driven by Scn10a-Cre and CNO was directly applied to the tooth. (B) Representative examples showing video frame (left) and corresponding motion pattern images from LabGym (right) to illustrate behavioral categories. Example from the front (top) or side (bottom) for each behavior. Overlaid color curves represent successive time points indicating behavior dynamics. (C–F) Raster plots showing the behavior categorizations over time for Scn10a-Cre; CAG-LSL-Gq-Dreadd (top) and wild-type controls (bottom) during (C) baseline, (D) 0.1 mg/kg CNO i.p., (E) 0.01 mg/kg CNO applied to unilateral m1 and m2 mandibular molars, and (F) 0.01 mg/kg CNO i.p. Color bars indicate behavioral categories, with color intensity reflecting the probability of each behavior. Time (x axis, s) and with each row (y axis) a single animal. n = 8 animals/genotype. (G) Bar graph depicting pain duration corresponding to (C)–(F). Plotted individual data points represent the cumulative pain duration for individual animals. Bar shows the mean, and error bars indicate the SEM (see Table S1 for one-way ANOVA with Bonferroni correction). See also Figure S4, Table S1, Videos S1, and S2.

Figure 5.. Optogenetic activation of intradental HTMRs induces digastric muscle activity and initiates a jaw-opening reflex

(A–C) Optogenetic activation of intradental neurons induces activity in the anterior digastric muscle EMG. (A) Schematic of selective activation of intradental neurons while recording EMG. Channelrhodopsin-2 was selectively expressed in right mandibular intradental neurons innervating m1 and m2 via tooth injection of AAV6/2-hEF1a-iCre into Ai32(RCL-ChR2(H134R)/EYFP) mice. (B) Example traces demonstrating EMG activity elicited by optogenetic activation of intradental neurons (15 ms, 473 nm), n = 5 mice for all conditions. Vertical scale bar, 1 mV. Horizontal scale bar, 15 ms. (C) Graph depicting the amplitude of activity of digastric muscle EMG in response to blue light stimulation (15 ms, 473 nm). Plotted individual data points represent the peak-to-peak amplitude of the elicited EMG trace. Error bar indicates SD. n = 5 mice per condition. p < 0.0001 using unpaired t test. (D–G) Optogenetic activation of intradental neuron terminals in the intact tooth results in mandibular deflection (i.e., jaw opening). (D) Schematic of approach. Channelrhodopsin-2 expression (Ai32) was driven by Scn10a-Cre and an optical fiber (500 μm) delivered blue light to target tissue. (E) Traces of average measured mandibular deflection in response to single blue light pulse (470 nm, 1 s) directed to the molar tooth. Black trace represents the average (n = 4 mice). Multicolored traces represent average measurements (n = 10 traces/mouse). Vertical scale bar, 100 μm. Horizontal scale bar, 100 ms. Inset shows brief timescale of trace from (E). Vertical scale bar, 150 μm. Horizontal scale bar, 10 ms. (F) Average trace of measured mandibular deflection in response to trains of blue or green light pulses (470 or 545 nm, 10 pulses, 0.5 Hz, 1 s). Vertical scale bar, 200 μm. Horizontal scale bar, 10 s. n = 4 mice. (G) Line graph depicting average mandibular deflection in response to sequential blue light pulse train stimulation of molar. Average deflection from each mouse is plotted as multicolored individual points and represents mean peak amplitude and error bars show SEM of mandibular deflection during stimulation. n = 10 measurements/mouse/train related to traces (F). Black points are average of n = 4 mice. n.s. p > 0.05 using unpaired t test. See also Figure S5 and Video S4.