Schwann cell endosome CGRP signals elicit periorbital mechanical allodynia in mice

Abstract

Efficacy of monoclonal antibodies against calcitonin gene-related peptide (CGRP) or its receptor (calcitonin receptor-like receptor/receptor activity modifying protein-1, CLR/RAMP1) implicates peripherally-released CGRP in migraine pain. However, the site and mechanism of CGRP-evoked peripheral pain remain unclear. By cell-selective RAMP1 gene deletion, we reveal that CGRP released from mouse cutaneous trigeminal fibers targets CLR/RAMP1 on surrounding Schwann cells to evoke periorbital mechanical allodynia. CLR/RAMP1 activation in human and mouse Schwann cells generates long-lasting signals from endosomes that evoke cAMP-dependent formation of NO. NO, by gating Schwann cell transient receptor potential ankyrin 1 (TRPA1), releases ROS, which in a feed-forward manner sustain allodynia via nociceptor TRPA1. When encapsulated into nanoparticles that release cargo in acidified endosomes, a CLR/RAMP1 antagonist provides superior inhibition of CGRP signaling and allodynia in mice. Our data suggest that the CGRP-mediated neuronal/Schwann cell pathway mediates allodynia associated with neurogenic inflammation, contributing to the algesic action of CGRP in mice.

Fig. 1. Schwann cell RAMP1 mediates PMA evoked by CGRP.

a Representative real-time PCR plot and cumulative data for GAPDH, S100, CLR and RAMP1 mRNA in HSCs (n = 3 independent experiments) and MSCs from trigeminal or sciatic nerve (MSCs from sciatic nerve n = 3 independent experiments; MSCs from trigeminal nerve, n = 4 independent experiments). b Representative images of DAPI and immunoreactive S100, RAMP1 and CLR in human and mouse cutaneous nerve bundles (scale, 10 µm human, 50 μm mouse) (n = 3 subjects). c PMA induced by CGRP (1.5 nmol) or vehicle in male and female Plp1-Cre^ERT+/Ramp1^fl/fl and Control mice treated with periorbital 4-OHT (n = 8 mice per group). d Representative images and colocalization value (Rcoloc) of S100 and RAMP1 in periorbital nerve and sciatic nerve trunks from Plp1-Cre^ERT+;Ramp1^fl/fl and Control mice (scale, 20 μm) (n = 4 replicates). TN (trigeminal nerve), SN (sciatic nerve). e PMA induced by intraperitoneal (i.p.) CGRP (0.1 mg/kg) or vehicle in male and female C57BL/6 J mice (n = 8 mice per group). f PMA and (g) paw mechanical allodynia induced by intraperitoneal (i.p.) CGRP (0.1 mg/kg) or vehicle in male Plp1-Cre^ERT+/Ramp1^fl/fl and Control mice (n = 8 mice per group) treated with periorbital 4-OHT. Mean±SEM. *P < 0.05, **P < 0.01, ***P < 0.001 vs. Veh, Control-Veh, and TN-Control, ^§§§P < 0.001. vs. Control-CGRP. 2-way (c, e, f, g) or 1-way (d) ANOVA, Bonferroni correction. Source data are provided as a Source Data file.

Fig. 2. Capsaicin induces PMA via CGRP and CLR/RAMP1 in Schwann cells.

a Acute nociception and (b) PMA after periorbital injection of capsaicin (CPS) or vehicle in C57BL/6 J mice. c Acute nociception and (d) PMA after CPS (50 pmol) or vehicle in Trpv1^+/+ and Trpv1^−/− mice. e Acute nociception and (f) PMA after CPS (50 pmol) or vehicle in C57BL/6 J mice pretreated with capsazepine (CPZ, 100 pmol) or vehicle. g, h PMA after periorbital SP (3.5 nmol), CPS (50 pmol) or vehicle in C57BL/6 J mice pretreated (0.5 h) with L-733,060 (20 nmol) or vehicle. i, j PMA after CPS (50 pmol) or vehicle in C57BL/6 J mice pretreated (0.5 h) with olcegepant (1 nmol) or CGRP8-37 (10 nmol) or vehicle. k PMA after CPS (50 pmol) or vehicle in Plp1-Cre^ERT+/Ramp1^fl/fl or Control mice treated with periorbital 4-OHT or vehicle. l PMA after periorbital CGRP (1.5 nmol) or vehicle and (m) acute nociceptive response and PMA after periorbital CPS (50 pmol) or vehicle in Adv-Cre^+/Ramp1^fl/fl or Control mice. n PMA induced by intraperitoneal (i.p.) CGRP (0.1 mg/kg) or vehicle in Adv-Cre^+/Ramp1^fl/fl or Control mice. Mean±SEM., n = 8 mice per group. **P < 0.01, ***P < 0.001 vs. Veh/Veh, Trpv1^+/+-Veh and Control-Veh^{; §}P < 0.05, ^§§P < 0.01, ^§§§P < 0.001 vs. Trpv1^+/+-CPS, CPS/Veh, SP/Veh, Control-CPS, Control-CGRP. 1-way (a, c, e and m left panel) or 2-way (b, d, f–l, m right panel and n) ANOVA, Bonferroni correction. Source data are provided as a Source Data file.

Fig. 3. GTN induces PMA via CGRP released from periorbital trigeminal terminals and, CLR/RAMP1 in Schwann cells.

a PMA and (b) paw mechanical allodynia induced by intraperitoneal (i.p.) GTN (10 mg/kg) or vehicle in Adv-Cre^+/Ramp1^fl/fl or Control mice post-treated (1.5 hs after GTN) with olcegepant (1 nmol) or vehicle. (c) PMA and (d) paw mechanical allodynia induced by GTN (10 mg/kg, i.p.) or vehicle in Plp1-Cre^ERT+/Ramp1^fl/fl or Control (treated with periorbital 4-OHT or vehicle) post-treated (1.5 hs after GTN) with olcegepant (1 nmol) or vehicle. Mean±SEM., n = 8 mice per group. ***P < 0.001 vs. Control-Veh/Veh; ^§§§P < 0.001 vs. Control-GTN/Veh. 2-way ANOVA, Bonferroni correction. Source data are provided as a Source Data file.

Fig. 4. Functional CLR/RAMP1 is expressed by HSCs and undergoes clathrin- and dynamin-mediated endocytosis, which underlies nociception.

a, b Effects of graded concentrations of CGRP on cAMP formation (n = 5 independent experiments). c, d Effects of graded concentrations of olcegepant on CGRP (100 nM)-evoked cAMP formation (n = 4 independent experiments). e Pharmacological targets. f Representative images of HSCs expressing Rab5a-GFP at 30 min after incubation with TAMRA-CGRP (100 nM). Arrows denote colocalization of TAMRA-CGRP and Rab5a-GFP. Arrowheads denote retention of a weak TAMRA-CGRP signal at the plasma membrane. Cells were preincubated with vehicle, Dyngo-4a (Dy4), Pitstop 2 (PS2), inactive analogs (PS2 and Dy4 inact) (all 30 µM) or sucrose (0.45 M) (n = 4 independent experiments). Scale, 10 µm. g Quantification of localization of TAMRA-CGRP in endosomes (data represent n = 949 veh 0 min, n = 1209 veh 30 min, n = 1111 Dy4 30 min, n = 1016 Dy4 inact 30 min, n = 1700 PS2 30 min, n = 714 PS2 inact 30 min, n = 1896 sucrose) and (h) quantification of the number of TAMRA-CGRP+ve endosomes (data represent n = 5 veh 0 min, n = 7 veh 30 min, n = 7 Dy4 30 min, n = 5 Dy4 inact 30 min, n = 7 PS2 30 min, n = 5 PS2 inact 30 min, n = 5 sucrose). i, j PMA induced by periorbital CGRP (1.5 nmol) or vehicle in C57BL/6 J male mice pretreated (0.5 h) with PS2, Dy4, PS2 or Dy4 inact (all 500 pmol) (n = 8 mice per group). k, l PMA induced by periorbital capsaicin (CPS, 50 pmol) or vehicle in C57BL/6 J male mice pretreated (0.5 h) with PS2, Dy4, PS2 or Dy4 inact (all 500 pmol) (n = 8 mice per group). Mean±SEM. ***P < 0.001 vs. Veh 0 min, and Veh/Veh; ^§§P < 0.01, ^§§§P < 0.001 vs. Veh 30 min, PS2 30 min, Dy4 30 min, CGRP/PS2 inact, CGRP/Dy4 inact, CPS/PS2 inact, CPS/Dy4 inact. 1-way (g, h) or 2-way (i–l) ANOVA, Bonferroni correction. Source data are provided as a Source Data file.

Fig. 5. CGRP leads to Gα protein activation and βARR2 recruitment at the plasma membrane and in early endosomes in HEK-hCLR/RAMP1 cells and HSCs-hCLR/RAMP1.

Endosomal signaling generates sustained formation of cAMP in HSCs. (a–d) CGRP (100 nM) increased EbBRET between Rluc8-mGα_s, Rluc8-mGα_sq, Rluc8-mGα_si, and Rluc2-βARR2 with RGFP-CAAX (a, b) and tdRGFP-Rab5a (c, d) in HEK-hCLR/RAMP1 cells. a, c time course. b, d area under curve (AUC) (a, b n = 10 mGα_s, n = 8 mGα_sq, n = 7 mGα_si, n = 7 βARR2, n = 9 veh; c, d n = 8 mGα_s, n = 8 mGα_sq, n = 8 mGα_si, n = 6 βARR2, n = 9 veh). e–h CGRP (100 nM) increased EbBRET between Rluc8-mGα_s, Rluc8-mGα_sq, Rluc8-mGα_si, and Rluc2-βARR2 with RGFP-CAAX (e, f) and tdRGFP-Rab5a (g, h) in HSC-hCLR/RAMP1 cells. e, g time course. f, h AUC (e, f n = 7 mGα_s, n = 9 mGα_sq, n = 5 mGα_si, n = 6 βARR2, n = 9 veh; g, h n = 7 mGα_s, n = 7 mGα_sq, n = 7 mGα_si, n = 5 βARR2, n = 7 veh). (i) Hypertonic sucrose (0.45 M) inhibited CGRP (100 nM)-stimulated EbBRET between hCLR-Rluc8 and tdRGFP-Rab5a in HEK-hCLR/RAMP1 cells (n = 5 independent experiments). (j, k) Hypertonic sucrose (0.45 M) inhibited CGRP (100 nM)-stimulated EbBRET between Rluc8-mGα_s, Rluc8-mGα_sq, Rluc8-mGα_si and Rluc2-βARR2 with tdRGFP-Rab5a in HEK-hCLR/RAMP1 cells (j, k n = 10 independent experiments). (l, m) Sucrose (0.45 M) inhibited CGRP (100 nM)-stimulated EbBRET between Rluc8-mGα_s, Rluc8-mGα_sq, Rluc8-mGα_si, and Rluc2-βARR2 with tdRGFP-Rab5a in HSC-hCLR/RAMP1 cells. (n = 8 independent experiments). (n, o) Sucrose (0.45 M) inhibited CGRP (100 nM)-stimulated formation of cAMP in HSCs. n time course. o, AUC (n = 5 independent experiments). Mean±SEM.. *P < 0.05, **P < 0.01, ***P < 0.001, vs. Veh. 1-way ANOVA, Dunnett’s correction (b, d, f, h) or parametric unpaired t test (i–m, o). Source data are provided as a Source Data file.

Fig. 6. Endogenous and exogenous CGRP induces PMA via NO production.

a Pharmacological targets. b–e PMA after periorbital injection of CGRP (1.5 nmol), capsaicin (CPS, 50 pmol) or vehicle in C57BL/6 J mice pretreated (0.5 h) with L-NAME (1 μmol) and cPTIO (200 nmol) or vehicle (n = 8 mice per group). f, g Real-time PCR plot and cumulative data for GAPDH, S100, NOS1, NOS2 and NOS3 in primary HSCs and MSCs (n = 3 independent experiments). h In-cell p-NOS3^s1177 ELISA in HSCs and IMS32 cells before (time 0, n = 11 independent experiments in both HSCs and IMS32) or after CGRP (time 5, n = 18 and n = 31, time 10, n = 16 and n = 21, time 15, n = 15 and n = 25, time 30, n = 12 and n = 29, time 60, n = 9 and n = 22 independent experiments in HSCs and IMS32, respectively). cAMP assay in (i) HSCs, IMS32 and (j) MSCs from Plp1-Cre^ERT+/Ramp1^fl/fl or Control mice treated with intraperitoneal 4-OHT, k nitric oxide assay in HSCs and IMS32 cells treated with CGRP or vehicle (n = 6 independent experiments). Some cells were treated with olcegepant (1 µM), CGRP8-37 (1 µM), SQ22536 (100 µM), L-NAME (100 µM) or vehicle (n = 6 independent experiments). Mean±SEM. (−) represents the combination of different vehicles. *P < 0.05, **P < 0.01, ***P < 0.001 vs. Veh/Veh, time 0 (min); ^§§§P < 0.001 vs. CGRP/Veh, CGRP 1 µM and 10 µM. 1-way (h–k) or 2-way (b–e) ANOVA, Bonferroni correction. Source data are provided as a Source Data file.

Fig. 7. CGRP induces ROS release via Schwann cell TRPA1 activation.

a Representative images of localization of immunoreactive DAPI, S100, RAMP1 and TRPA1 in human abdominal and mouse periorbital cutaneous nerve bundles (Scale: 10 µm human, 50 μm mouse, inset 10 µm) (n = 3 subjects). b Pharmacological targets. c, d PMA after periorbital injection of CGRP (1.5 nmol) or vehicle in (c) Trpa1^+/+ and Trpa1^−/− mice and in (d) Adv-Cre⁺/Trpa1^fl/fl or Control mice (n = 8 mice per group). e, f Ca²⁺ response in HSCs and IMS32 cells exposed to CGRP (0.001 µM, HSCs n = 284; 0.01 µM, HSCs n = 291, IMS32 n = 62; 0.1 µM, HSCs n = 297, IMS32 n = 53; 1 µM, HSCs n = 362, IMS32 n = 57; 10 µM, IMS32 n = 53 cells) in the presence of olcegepant (100 nM, HSCs n = 291, IMS32 n = 44 cells), CGRP8-37 (100 nM, HSCs n = 359, IMS32 n = 55 cells), SQ22536 (100 μM, HSCs n = 225, IMS32 n = 47 cells), H89 (1 µM, HSCs n = 292, IMS32 n = 58 cells), L-NAME (10 μM HSCs n = 285, IMS32 n = 38), Ca²⁺-free medium (HSCs n = 290, IMS32 n = 28 cells), PBN (50 μM, HSCs n = 309, IMS32 n = 24 cells), ML171 (1 μM, HSCs n = 320, IMS32 n = 49 cells), A967079 (A96, 50 μM, HSCs n = 276 IMS32 n = 533 cells) or vehicle (HSCs n = 297, IMS32 n = 41 cells). g Nitric oxide release in HSCs and IMS32 cells exposed to CGRP (1 or 10 µM) in the presence of A96 (50 μM) or vehicle (n = 6 independent experiments). (h–j) PMA after CGRP (1.5 nmol) or vehicle in C57BL/6 J male mice pre-treated (0.5 h) with h A96 (300 nmol), i PBN (670 nmol) or (j) ML171 (250 nmol) or vehicle. k PMA after CGRP (1.5 nmol) or vehicle in Plp1-Cre^ERT+/Trpa1^fl/fl or Control mice treated with 4-OHT (h–k n = 8 mice per group). Mean±SEM. (−) represents the combination of different vehicles. *P < 0.05, **P < 0.01, ***P < 0.001 vs. Trpa1^+/+-Veh, Control-Veh, Veh and Veh/Veh; ^§§P < 0.01, ^§§§P < 0.001 vs. Trpa1^+/+-CGRP, Control-CGRP, CGRP 1 µM and 1 µM, CGRP/Veh. 2-way (c, d, h-k) or 1-way (e, f) ANOVA, Bonferroni correction. Source data are provided as a Source Data file.

Fig. 8. DIPMA-MK-3207 nanoparticles target endosomal CLR/RAMP1 signaling and provide superior relief from CGRP-evoked PMA.

a pH-responsive DIPMA-MK-3207. b Transmission electron micrograph image of DIPMA-MK-3207 (Scale: 0.1 µm), from two different nanoparticle preparations (3 images captured per sample). c Physicochemical properties of DIPMA-MK-3207 and DIPMA-Ø. d Uptake of DIPMA-Cy5 into HSCs expressing EEA1-GFP. Cells were preincubated with DIPMA-Cy5 (40–60 ng/ml) for 30 min and were then incubated with TAMRA-CGRP (100 nM) for 30 min. Arrows denote accumulation of TAMRA-CGRP in early endosomes containing DIPMA-Cy5. Representative images from n = 5 independent experiments (Scale: 10 µm). e–g Effects of DIPMA-MK-3207, MK-3207, DIPMA-Ø or vehicle on CGRP- (100 nM) stimulated cAMP formation in HEK-rCLR/RAMP1 cells. e Time course and f, g integrated response (AUC) before (1st phase) and after (2nd phase) washing to remove extracellular CGRP (n = 6 independent experiments). h Concentration-response curves of the inhibition by DIPMA-MK-3207 or free MK-3207 on the Ca²⁺ response to CGRP in HSCs (DIPMA-MK-3207: −9M, n = 145; −8M, n = 361; −7M, n = 213: −6.5 M, n = 150; or free MK-3207: −8M, n = 83; −6M, n = 106; −5M, n = 87; −4M, n = 127: −3M, n = 127 cells). i PMA, expressed as AUC, after periorbital injection of CGRP (1.5 nmol), capsaicin (CPS, 50 pmol) or vehicles in C57BL/6 J male mice pre-treated (0.5 h) with DIPMA-MK-3207, MK-3207 (0.1, 0.3, 1 pmol), DIPMA-Ø or vehicle (n = 8 mice per group). Mean±SEM. ***P < 0.001 vs. DIPMA-Ø/Veh, ^###P < 0.001 vs. MK-3207 0.3 pmol and MK-3207 1 pmol. f, g, i 1-way ANOVA, Bonferroni correction. Source data are provided as a Source Data file.

Fig. 9. Schematic representation of the pathway that signal prolonged cutaneous allodynia elicited by CGRP released and associated with neurogenic inflammation.

The pro-migraine neuropeptide, CGRP, released from trigeminal cutaneous afferents, activates CLR/RAMP1 on Schwann cells. CLR/RAMP1 traffics to endosomes, where sustained G protein signaling increases cAMP and stimulates PKA that results in nitric oxide synthase activation. The ensuing release of nitric oxide targets the oxidant-sensitive channel, TRPA1, in Schwann cells, which elicits persistent ROS generation. ROS triggers TRPA1 on adjacent C- (1) or Aδ-fiber (2) afferents resulting in periorbital allodynia, a hallmark of migraine pain. The inset shows several unmyelinated axons invaginated into a Schwann cell forming a Remak bundle.