Role of TMEM100 in mechanically insensitive nociceptor un-silencing

Abstract

Mechanically silent nociceptors are sensory afferents that are insensitive to noxious mechanical stimuli under normal conditions but become sensitized to such stimuli during inflammation. Using RNA-sequencing and quantitative RT-PCR we demonstrate that inflammation upregulates the expression of the transmembrane protein TMEM100 in silent nociceptors and electrophysiology revealed that over-expression of TMEM100 is required and sufficient to un-silence silent nociceptors in mice. Moreover, we show that mice lacking TMEM100 do not develop secondary mechanical hypersensitivity-i.e., pain hypersensitivity that spreads beyond the site of inflammation-during knee joint inflammation and that AAV-mediated overexpression of TMEM100 in articular afferents in the absence of inflammation is sufficient to induce mechanical hypersensitivity in remote skin regions without causing knee joint pain. Thus, our work identifies TMEM100 as a key regulator of silent nociceptor un-silencing and reveals a physiological role for this hitherto enigmatic afferent subclass in triggering spatially remote secondary mechanical hypersensitivity during inflammation.

Fig. 1. NGF treatment selectively upregulates TMEM100 in silent nociceptors.

a Cartoon depicting the RNAseq screen workflow. b Volcano plot showing the mean fold-change of expression upon 24 h NGF treatment (log₂FC) vs. the log P-value determined by paired-end RNAseq (n = 3 biologically independent samples, 20 cells per sample, from three different mice; two-sided Student’s T-test). c Comparison of the mean ± SEM expression levels (counts per million, CPM) of peptidergic nociceptor subclass markers and d mechanically-gated ion channels and PIEZO2 modulators, determined by RNAseq in CHRNA3-EGFP⁺ neurons cultured with and without NGF, using two-sided Students t-test (***, P = 4.59843E-05; N-numbers are the same as in b). e Comparison of the mean ± SEM expression levels of TMEM100 (normalized to GAPDH expression) determined by qPCR in the indicated nociceptor subclasses, cultured without (CTL) and with NGF. To enable identification of peptidergic C-fiber nociceptors, non-peptidergic C-fiber nociceptors and Aδ-fiber nociceptors for sample collection, cultures were prepared from Tg(Npy2r-cre)SM19Gsat/Mmucd x B6;129S-Gt(ROSA)26Sortm32(CAG-COP4*H134R/EYFP)Hze/J (Npy2r^Cre;ChR2-EYFP) mice, in which Aδ-fiber nociceptors express EYFP and were additionally labeled with Alex-Fluor-568 conjugated Isolectin B4 (IB4), which selectively binds to non-peptidergic C-fiber nociceptors. Numbers of samples (20 cells each) per subpopulation are indicated in brackets above the bars and individual values are shown as dots (two-sided Mann-Whitney test: P-values provided above bars). f Image showing cultured DRG neurons from CHRNA3-EGFP mice transfected with TMEM100-dsRed. g Cartoon depicting the mechano-clamp configuration of the patch-clamp technique (left, originally published in Verkest et al. 10.1038/s41467-022-28974-6 Nat Commun), example traces of mechanically-evoked currents in CHRNA3-EGFP⁺ control cells (middle, top) and in TMEM100-dsRed-transfected CHRNA3-EGFP⁺ cells (middle, bottom) as well as bar graph showing the proportion of cells responding to mechanical stimulation. Proportions were compared with two-sided Fisher’s exact test (P = 0.023). h Mean ± SEM peak amplitudes of mechanically-evoked currents are shown as a function of membrane displacement for control (white circles) and TMEM100 transfected cells (red squares). Current amplitudes were compared using multiple two-sided Mann-Whitney tests (P-values are provided above the symbols). N-numbers differ from g, because some recordings crashed before maximal mechanical stimulation. Source data are provided as a Source Data file.

Fig. 2. Intraarticular CFA injection induces knee joint pain and secondary mechanical hypersensitivity in the hind paw.

a Representative image of a knee joint section immunostained for CGRP and EGFP. Note, only EGFP⁺ fibers that co-express CGRP are sensory afferents. CGRP^–/EGFP⁺ fibers are sympathetic efferents (see Prato et al., 2017). b Close-up of the region marked by the white rectangle in a, emphasizing co-expression of EGFP and CGRP. c Quantification of silent afferent density (EGFP⁺/CGRP⁺ fibers) in anatomically defined regions. FP (Hoffa’s fat pad), LM (lateral meniscus), MM (medial meniscus), LJC (lateral joint capsule), MJC (medial joint capsule), CL (cruciate ligament). Bars represent means ± SEM from three different mice. For each mouse, at least three photomicrographs per anatomical region were analyzed and averaged. Means from each mouse are shown as black dots. d Cartoon depicting the experimental approach (left), photograph of an inflamed knee (middle) and time course (days post-injection, dpi) of saline- (n = 5 mice) and CFA-induced (n = 6 mice) changes in knee size (right). Data are presented as means ± SEM. e Freeze frames of Catwalk XT movies from saline (top) and CFA (bottom) treated mice. Note, CFA-treated mice only put little weight on the left hind paw (small foot print). f Comparison of the foot print area ratio (left/right hindpaw; LH/RH) and the leg swing time ratio (LH/RH) measured before (solid bars) and three day after (3 dpi, hatched bars) saline (gray) and CFA (orange) injection. Two-sided paired Student’s t-test (saline N = 15, CFA N = 16; print area CFA, P = 2.6 × 10⁻⁷; swing speed CFA P = 8 × 10⁻⁹). g Comparison of mechanical paw withdrawal thresholds before (solid bars) and three day after (3 dpi, hatched bars) saline (gray) and CFA (blue) injection. Paired two-sided Student’s t-test (saline N = 15, CFA N = 16; CFA, P = 5.18 × 10⁻¹²). h Comparison of thermal paw withdrawal latencies before (solid bars) and three day after (3 dpi, hatched bars) saline (gray) and CFA (blue) injection. Paired two-sided Student’s t-test (saline N = 15, CFA N = 16; CFA, P = 1.8 × 10⁻⁹). Source data are provided as a Source Data file.

Fig. 3. CFA-induced knee joint inflammation induces mechanosensitivity and potentiates TRPA1 activity in CHRNA3-EGFP⁺ afferents.

a Cartoon depicting retrograde labeling of knee joint afferents (left), example image of FB⁺ neurons in DRG section (middle) and quantification of the mean ± SEM numbers of FB⁺ neurons in L3 and L4 DRGs from 6 mice. Dots represent individual values for each DRG. b Example image of a DRG culture from a CHRNA3-EGFP mouse after intraarticular FB (left). Stacked bar graph (right) shows the proportions of IB4⁺ cells (non-peptidergic nociceptors, red), MIAs (green), peptidergic nociceptors (IB4^–, <30 µm, blue) and group-II articular afferents (IB4^–, > 30 µm, blue hatched). c Quantification of CFA-induced changes in TMEM100 expression in FB-labeled MIAs and peptidergic nociceptors. Expression levels in ipsi- and contralateral DRGs (20 cells per sample), at the indicated timepoints, were compared by qPCR (ΔΔCt method). Bars represent means ± SEM and data from individual mice are shown as black dots (N = 5 mice for 3 and 21 dpi in CHRNA3⁺/FB⁺ cells and N = 4 for IB4^–/FB⁺ cells). d Example traces of mechanically-evoked currents. e, f Comparison of the mean ± SEM peak amplitudes of mechanotransduction currents evoked by increasing membrane displacements of e MIAs and f peptidergic nociceptors from saline (open symbols) and CFA-treated (solid symbols) mice, using two-sided Mann-Whitney test (P-values are provided next to the symbols in e and f). g Comparison of the mean ± SEM inactivation time constants of the mechanically-evoked currents using two-sided T-test. P-values and N-numbers of independent experiments (numbers in brackets) are shown above the bars. h Comparison of the proportions of CHRNA3⁺/FB⁺ (green) and IB4^–/FB⁺ (blue) DRG neurons from wildtype mice that exhibit Ca²⁺ transients (visualized with Calbryte-590, see Supplementary Fig. 3a, b) in response to the TRPA1 agonist AITC (10 µM) 3 dpi of saline (solid fill) or CFA (cross-hatched bars) using two-sided Fishers exact test (P-values above bars). Numbers above bars indicate the number of responders and number of tested cells. Panel d drawing originally published in Verkest et al. 10.1038/s41467-022-28974-6 Nat Commun Source data are provided as a Source Data file.

Fig. 4. TMEM100 knock-out mice develop normal inflammatory knee joint pain but no long-lasting secondary mechanical hypersensitivity.

a Comparison of the time courses of changes in stand time (left), foot print area (middle) and leg swing speed (right) of saline injected WT mice (white circles), CFA injected WT mice (black circles) and CFA injected TMEM100KO mice (orange squares). b Cartoon depicting the experimental approach for measuring secondary mechanical and thermal hypersensitivity in the ipsilateral hindpaw (left), time courses of changes in mechanical paw withdrawal thresholds of saline injected WT mice (white circles), CFA injected WT mice (black circles) and CFA injected TMEM100KO mice (blue squares) (middle) and responsiveness of WT mice (white circles), CFA injected WT mice (black circles) and CFA injected TMEM100KO mice (blue squares) three days post saline/CFA injection (3 dpi) to all tested von Frey filaments (right). Response rates were compared using 2-way ANOVA. P-values of multiple comparisons are as follows: WT-saline vs. WT-CFA, P_0.07g = 7.5E-08, P_0.16g = 2.8E-08, P_0.4g = 8.7E-12, P_0.6g = 7.5E-08, P_1g = 2.6E-06, P_1.4g = 1.2E-01; WT-saline vs. TMEM100KO-CFA, P_0.07g = 5.9E-01, P_0.16g = 3.3E-01, P_0.4g = 1.5E-04, P_0.6g = 7.8E-04, P_1g = 6.8E-05, P_1.4g = 1.2E-01; WT-CFA vs. TMEM100KO-CFA, P_0.07g = 5.2E-08, P_0.16g = 3.6E-08, P_0.4g = 3.9E-10, P_0.6g = 9.9E-05, P_1g = 6.1E-02, P_1.4g = not determined). c Time courses of changes in thermal paw withdrawal latencies of saline injected WT mice (white circles), CFA injected WT mice (black circles) and CFA injected TMEM100KO mice (green squares). a–c Symbols represent means ± SEM. Unless otherwise stated, ratios at different time points were compared using mixed model ANOVA and P-values of multiple comparisons are indicated above the white circles. Top, WT-saline vs. WT-CFA; middle, WT-saline vs. TMEM100KO-CFA; bottom, WT-CFA vs. TMEM100KO-CFA (*P < 0.05; **P < 0.01; ***P < 0.001; ns, P > 0.05, N-numbers are provided in the graph legends). Exact P-values and detailed statistical information is provided together with the source data as a Source Data file.

Fig. 5. CFA-induced knee joint inflammation fails to sensitize CHRNA3-EGFP⁺ neurons to mechanical stimuli in TMEM100 knock-out mice.

a Cartoon depicting the mechano-clamp configuration of the patch clamp technique (left, originally published in Verkest et al. 10.1038/s41467-022-28974-6 Nat Commun) and example traces of mechanically-evoked currents in the indicated cell types and conditions (right). Note, in contrast to WT mice, CHRNA3-EGFP⁺ neurons from TMEM100KO mice do not acquire mechanosensitivity during CFA-induced inflammation. b, c mean ± SEM peak amplitudes of mechanically-evoked currents are shown as a function of membrane displacement for b MIAs and c peptidergic nociceptors from saline (open symbols) and CFA-treated (solid symbols) mice. Current amplitudes were compared using two-sided Mann-Whitney test (P-values for b from left to right: ns = 0.99, ns = 0.45, ns = 0.21, ns = 0.57, ns = 0.33, ns = 0.21, ns = 0.22, ns = 0.12, ns = 0.55; P-values for c from left to right: ns = 0.99, ns = 0.99, ns = 0.25, ns = 0.4, ns = 0.73, ns = 0.99, ns = 0.47, ns = 0.5, ns = 0.92). d Comparison of the mean ± SEM inactivation time constants of mechanically evoked using two-sided unpaired t-tests (P-values are provided above bars and individual values are shown as dots. e Comparison of the proportions of CHRNA3⁺/FB⁺ (green bars) and IB4^–/FB⁺ (blue bars) DRG neurons from TMEM100KO mice that exhibit Ca²⁺ transients (visualized with Calbryte-590, also see Supplementary Fig. 3c, d) in response to the TRPA1 agonist AITC (10 µM) 3 days post injection (3 dpi) of saline (solid fill) or CFA (cross-hatched bar) using two-sided Fishers exact test (P-values are indicated above the bars. Numbers in brackets above the bars indicate the number of responders and number of tested cells. Source data are provided as a Source Data file.

Fig. 6. Sensitization of cutaneous C-fiber nociceptors contributes to secondary mechanical hypersensitivity.

a Cartoon depicting the experimental approach. b Scatter dot plots of the conduction velocities [m/s] of the examined afferent fibers from the indicated mice. Bars represent means ± SEM. c Comparison of the proportions of C-fiber nociceptors that respond to mechanical stimulation with the indicated von Frey filaments. The proportions were calculated from all C-fibers (same N-numbers as in b) recorded from 3 different mice per group and were compared pairwise using the two-sided Chi-square test. P-values (ns, P > 0.05; *P < 0.05; **P < 0.01, ***P < 0.001) are provided next to the symbols in the graph and refer to WT-saline vs. WT-CFA (top, black), WT-saline vs. TMEM100KO-CFA (middle, blue) and WT-CFA vs. TMEM100KO (bottom, black). Exact P-values are provided together with additional statistical information in the source data file. d Comparison of the firing rates evoked by a series of ramp-and-hold stimuli with increasing amplitudes that exerted the indicated force to the receptive fields. Symbols represent the mean ± SEM numbers of action potentials, which were compared using multiple two-sided Mann-Whitney tests. (WT-saline vs. WT-CFA: *P < 0.05; **P < 0.01; WT-CFA vs. TMEM100KO-CFA: #P < 0.05). Exact P-values are provided together with additional statistical information in the source data file. e Example traces of mechanically-evoked action potentials recorded from single nerve fibers from the tibial nerve of WT mice 3 days post intraarticular saline (top), WT mice 3 dpi CFA (middle) and TMEM100KO mice 3 dpi CFA (bottom). Source data are provided as a Source Data file.

Fig. 7. Overexpression of TMEM100 in articular afferents induces secondary mechanical hypersensitivity in the hind paw but no knee joint pain.

a Cartoon depicting the experimental approach (left), image of a DRG section from a mouse that had received intraarticular AAV-PHP.S-TMEM100-Ires-dsRed (middle) and quantification of the mean ± SEM number of dsRed⁺ neurons (right) in L3 and L4 DRGs from two mice. b dsRed fluorescence in the tibial nerve distal to the knee (right) and the saphenous (left) nerve proximal to the knee which contains the medial articular nerve. c Comparison of the mean ± SEM stand time (left) and leg swing speed ratios (right) of WT mice at the indicated timepoints after i.a. injection of AAV-PHP.S-dsRed control virus (white circles) and AAV-PHP-S-TMEM100-Ires-dsRed (orange circles). Ratios were compared using multiple two-sided Mann-Whitney tests (P-values and N-numbers are provided next to the symbols and in the graph legend). d Comparison of the mean ± SEM mechanical paw withdrawal thresholds (left) and the response rates at 14 dpi (right, % paw withdrawals in response to five successive stimulations with the indicated von Frey filaments) of the same mice as in c, using two-sided Mann-Whitney test (P-values are provided next to the symbols; number of tested animals is the same in both panels and is indicated in the graph legend). e Example traces of mechanically-evoked action potentials recorded from cutaneous C-fiber nociceptors in the tibial nerve from control mice (top, AAV-PHP.S-dsRed) and from mice that overexpress TMEM100 in articular afferents (bottom, AAV-PHP.S-TMEM100-Ires-dsRed). f Comparison of the firing rates evoked by ramp-and-hold stimuli that exerted the indicated force to the receptive fields. Symbols represent means ± SEM numbers of action potentials, which were compared using multiple two-sided Mann-Whitney tests (exact P-values are provided next to the symbols). g comparison of the proportions of C-fiber nociceptors that respond to mechanical stimulation with the indicated von Frey filaments. The proportions were compared pairwise using two-sided Chi-square test (exact P-values are provided next to the symbols). h cartoon depicting the proposed mechanism underlying the induction of secondary hypersensitivity. Source data are provided as a Source Data file.