Schwann cells modulate nociception in neurofibromatosis 1

Abstract

Pain of unknown etiology is frequent in individuals with the tumor predisposition syndrome neurofibromatosis 1 (NF1), even when tumors are absent. Nerve Schwann cells (SCs) were recently shown to play roles in nociceptive processing, and we find that chemogenetic activation of SCs is sufficient to induce afferent and behavioral mechanical hypersensitivity in wild-type mice. In mouse models, animals showed afferent and behavioral hypersensitivity when SCs, but not neurons, lacked Nf1. Importantly, hypersensitivity corresponded with SC-specific upregulation of mRNA encoding glial cell line-derived neurotrophic factor (GDNF), independently of the presence of tumors. Neuropathic pain-like behaviors in the NF1 mice were inhibited by either chemogenetic silencing of SC calcium or by systemic delivery of GDNF-targeting antibodies. Together, these findings suggest that alterations in SCs directly modulate mechanical pain and suggest cell-specific treatment strategies to ameliorate pain in individuals with NF1.

Keywords: Calcium; Mouse models; Neuroscience; Pain.

Figure 1. SC-specific knockout of Nf1 leads to mechanical hypersensitivity.

(A) Nf1 haploinsufficient mice (Nf1^+/–) do not show mechanical hypersensitivity using the Randall-Selitto (R-S) assay (n = 12 control; n = 13 mutant). (B) Deletion of Nf1 in sensory neurons (PirtCre Nf1^+/fl) does not cause any mechanical hypersensitivity using R-S (PirtCre Nf1^+/fl; n = 17 control; n = 17 mutant). (C) Mice with SC-specific Nf1 deletion (DhhCre Nf1^fl/fl) show a trend toward reduced mechanical withdrawal thresholds (n = 20 control; n = 12 mutant; P < 0.077 vs. time-matched littermate controls; 2-way repeated measures [RM] ANOVA, Tukey’s post hoc test; mean ± SEM). (D) Using MCA assay, Nf1^+/– mice prefer to spend more time during the assay exposed to an aversive light stimulus compared with a noxious mechanical stimulus. (*P < 0.05 vs. controls, 1-way ANOVA with Tukey’s post hoc; mean ± SEM). (E) PirtCre Nf1^+/fl mice did not show any significant difference in time spent in either light or dark chambers. (F) DhhCre Nf1^fl/fl mice display increased mechanical avoidance even with smaller spikes present versus littermate controls. (*P < 0.05, ^#P < 0.05 vs. controls 1-way ANOVA with Tukey’s post hoc; mean ± SEM.)

Figure 2. Sensitization of high-threshold mechanoreceptors and polymodal C-fibers in DhhCre Nf1^fl/fl mice as assessed with ex vivo recording.

(A) Representative image of the ex vivo electrophysiological recording preparation. (B) Firing pattern of high-threshold mechanoreceptors (HTMRs) or A-fibers and polymodal C-fibers (CPMs) in WT C57BL/6 (C57) controls and Nf1^fl/fl and DhhCre Nf1^fl/fl mice at 4–5 months of age. (C) HTMRs from DhhCre Nf1^fl/fl mice showed a significant reduction in mechanical thresholds compared with control HTMRs (WT C57, n = 9; Nf1^fl/fl, n = 9, mutant, n = 7; *P < 0.05 vs. Nf1^fl/fl. **P < 0.01, vs. Nf1^fl/fl, 1-way ANOVA with Tukey’s post hoc; mean ± SEM; total no. of cells, WT C57, n = 45; Nf1^fl/fl, n = 57, mutant, n = 37). (D) Firing rates of HTMRs showed the increased firing to mechanical stimuli in DhhCre Nf1^fl/fl mice when compared with controls (*P < 0.05 vs. Nf1^fl/fl 1-way ANOVA with Tukey’s post hoc; mean ± SEM). (E) HTMRs showed no change in heat thresholds. (F) CPMs in DhhCre Nf1^fl/fl mice also showed reduced mechanical thresholds compared with controls (WT C57, n = 14; Nf1^fl/fl, n = 14, mutant, n = 8; *P < 0.05 vs. Nf1^fl/fl, 1-way ANOVA with Tukey’s post hoc; mean ± SEM). (G) DhhCre Nf1^fl/fl mice also showed the increased firing rate of CPMs (*P < 0.05 vs. Nf1^fl/fl, 1-way ANOVA with Tukey’s post hoc; mean ± SEM). (H) CPMs in DhhCre Nf1^fl/fl mice showed no change in heat thresholds.

Figure 3. Chemogenetic activation of SCs induces peripheral hypersensitivity.

(A) DhhCre hM3Dq mice expressing DREADD reporter (top left) and mCitrine in SCs (yellow) surrounding putative NeuN⁺ neurons (purple) in DRGs (top right). SC cultures from DhhCre hM3Dq mice treated with CNO (40 μM) display enhanced calcium fluorescence (Fluo-4) compared with untreated cultures (bottom left and right, and bar graph, scale bar = 100 μm) (****P < 0.0001 vs. no treatment, t = 14.44, df = 148; t test; mean ± SEM). (B) Treatment of DhhCre hM3Dq mice (n = 5) for 7 days with CNO (2 mg/kg, i.p. once/d) induces mechanical hypersensitivity compared with CNO-treated controls (n = 8) by R-S (*P < 0.05 vs. hM3Dq after CNO, 2-way ANOVA with Tukey’s post hoc; mean ± SEM). (C) Similar results are also seen using the MCA assay (*P < 0.05 vs. hM3Dq after CNO, 1-way ANOVA with Tukey’s post hoc; mean ± SEM). (D) Ex vivo recording of saphenous afferents indicates reduced mechanical thresholds in CPM fibers in CNO-treated DhhCre hM3Dq mice (n = 12 CPMs) compared with controls (n = 12 CPMs) (*P < 0.05 vs. hM3Dq, 1-way ANOVA with Tukey’s post hoc; mean ± SEM). (E) Enhanced firing over increasing forces (stimulus encoding) observed in control CPMs was not found in the DhhCre hM3Dq CPMs. ^P = 0.0038, DhhCre hM3Dq vs. hM3Dq. (F) Area under the curve for firing rates for CPMs. *P < 0.05, 1-way (A, C, D, and F) or 2-way RM (B and E) ANOVA with Tukey’s post hoc as appropriate; mean ± SEM. (G) Example firing patterns of CPM neurons from DhhCre hM3Dq and hM3Dq mice after 7days of CNO injections.

Figure 4. Chemogenetic inhibition of SCs suppresses mechanical hypersensitivity in DhhCre Nf1^fl/fl hM4Di mice.

(A) In isolated SCs from sciatic nerves of DhhCre Nf1^fl/fl hM4Di mice, no significant changes in calcium are detected in SCs treated with vehicle (DMSO) (top panel, left). Calcium release is increased upon addition of ATP (100 μM with vehicle) (bottom panel, left). No significant changes in calcium are detected in SCs treated with compound 21 (C21) alone. Inhibition of ATP-induced calcium is observed, however, with C21 in SCs isolated from DhhCre Nf1^fl/fl hM4Di mice (bottom panel, right) (scale bar = 100 μm). White arrow indicates cells displaying green fluorescence, and the red arrows mark the absence of fluorescence within cells in the respective images (bottom, left and right). (B) Quantification of fluorescence intensity from SCs depicting changes in calcium release from conditions outlined in A (****P < 0.0001 ATP with vehicle vs. ATP with C21 only, and ^^^^P < 0.0001 C21 vs. ATP with C21, 2-way ANOVA with HSD post hoc; mean ± SEM). (C) DhhCre Nf1^fl/fl hM4Di mice display increased mechanical avoidance even with smaller spikes present vs. littermate controls (n = 16 control, n = 7 mutant, *P < 0.05 vs. controls 2-way ANOVA, Tukey’s post hoc; mean ± SEM), before C21 injection, but after 7 days of C21 injection (i.p.), mechanical avoidance is reduced to control levels.

Figure 5. GDNF signaling from SC to neurons is enhanced in DhhCre Nf1^fl/fl DRGs.

(A) Analysis of cell-cell signaling in DhhCre Nf1^fl/fl DRGs indicates that GDNF signals from nonmyelinating SC and SC precursors to neurons containing its co-receptor (GFRa1). (B) Four neuronal subtypes (TH⁺, peptidergic 2 [PEP2], nonpeptidergic 3 [NP3], and NP6) display unique signaling of GDNF with SC precursors (SCPs) and nonmyelinating SC in 7-month tumor compared with 2-month control/pretumor or 7-month control. L, ligand; R, receptor.

Figure 6. GDNF is elevated in the DRG of DhhCre Nf1^fl/fl when compared with control mice and regulates behavioral hypersensitivity.

(A) Representative images of DRGs stained with different markers including S100β (green), GDNF (purple), and DAPI (blue). GDNF is mainly expressed in glial cells (arrows). Scale bar, 100 μm. (B) Quantifying the fluorescence intensity from each image shows elevated GDNF in DRGs of DhhCre Nf1^fl/fl mice compared with controls (*P < 0.05 vs. control, 1-way ANOVA with Tukey’s post hoc; mean ± SEM). (C) DhhCre Nf1^fl/fl mice normally display mechanical hypersensitivity in the MCA assay at 4 months; however, 24 hours after being injected (i.v.) with GDNF-targeting antibody, mechanical avoidance is reduced to control levels (n = 19 control, n = 8 mutant). (*P < 0.05 vs. controls, 2-way ANOVA, with Tukey’s post hoc; mean ± SEM.)