Systemic desensitization through TRPA1 channels by capsazepine and mustard oil - a novel strategy against inflammation and pain

Abstract

We demonstrate a novel dual strategy against inflammation and pain through body-wide desensitization of nociceptors via TRPA1. Attenuation of experimental colitis by capsazepine (CPZ) has long been attributed to its antagonistic action on TRPV1 and associated inhibition of neurogenic inflammation. In contrast, we found that CPZ exerts its anti-inflammatory effects via profound desensitization of TRPA1. Micromolar CPZ induced calcium influx in isolated dorsal root ganglion (DRG) neurons from wild-type (WT) but not TRPA1-deficient mice. CPZ-induced calcium transients in human TRPA1-expressing HEK293t cells were blocked by the selective TRPA1 antagonists HC 030031 and A967079 and involved three cysteine residues in the N-terminal domain. Intriguingly, both colonic enemas and drinking water with CPZ led to profound systemic hypoalgesia in WT and TRPV1(-/-) but not TRPA1(-/-) mice. These findings may guide the development of a novel class of disease-modifying drugs with anti-inflammatory and anti-nociceptive effects.

Figure 1. Capsazepine (CPZ) enemas attenuate murine DSS colitis independently of TRPV1.

(A) Representative colonoscopy photographs during the course of DSS (5%) colitis at days 2, 4 and 6 (d2, d4, d6) (each group n = 8). Comparison of WT and TRPV1-deficient mice without CPZ (531 μM) enemas with WT and TRPV1^−/− mice that were treated twice daily with CPZ enemas from day 0. (B) Endoscopic score (Mann Whitney U-test) (C) course of body weight analysis of variance (ANOVA) followed by Fisher’s LSD test, (D) histological examination (HE staining) and (E) histological score (Mann Whitney U-test) in the different groups. All *P < 0.05, **P < 0.01.

Figure 2. Capsazepine (CPZ) activates heterologously expressed human TRPA1.

(A) CPZ (10 μM) evokes hTRPA1-mediated currents in transfected HEK293t cells. Note the complete inhibition of currents by the selective TRPA1 antagonist HC030031 (HC, 10 μM). (B) Representative ramp currents evoked in a hTRPA1-transfected HEK293 cell by 400 ms voltage ramps from −100 to +100 mV applied every 4 s. Each symbol represents the current amplitude at −80 mV (squares) and +80 mV (circles). The CPZ-induced current was strongly inhibited by HC. (C) Examples of individual ramp currents corresponding to the filled symbols and numbers in B. (D) CPZ (50 μM, 10 s) evokes large calcium transients in hTRPA1-HEK293 cells (black trace, n = 128). HC (20 μM; red trace, n = 209) and A-967079 (10 μM; blue trace, n = 140) completely inhibited the response to CPZ. Data represent means (straight lines) ± SEMs (dotted lines). (E) The activation of hTRPA1 by CPZ (10 μM) is concentration-dependent. Black trace: hTRPA1-HEK293 cells (n = 52) were stimulated by increasing concentrations of CPZ for 20 s at 3 min-intervals. Gray trace: untransfected HEK293 cells (n = 101) were subjected to the same CPZ concentrations. (F) Activation of hTRPA1 by CPZ (1 μM) involves three critical cysteines in the N-terminus of the channel. Black trace: response of n = 123 HEK293 cells expressing WT hTRPA1 to successive applications of CPZ (1 μM), carvacrol (100 μM) and allyl isothiocyanate (AITC, 50 μM). Gray trace: response of n = 165 HEK293 cells expressing the mutant hTRPA1-3C to the same sequence of stimuli. Note the substantial reduction in the response of the hTRPA1-3C mutant to CPZ and AITC, but not to carvacrol. (G) The difference in CPZ sensitivity between genotypes was abolished at 100 μM CPZ. (H) Activation of hTRPA1 by CPZ can be prevented by the scavenger N-acetyl cysteine (NAC). Black trace: in control experiments hTRPA1-HEK293 cells were challenged with CPZ (50 μM; 60 s; n = 74). Red trace: in a separate experiment cells were first exposed for 30 s to a combination of CPZ (50 μM) and NAC (15 mM), immediately followed by CPZ alone for another 30 s (n = 83).

Figure 3. Capsazepine (CPZ) activates murine TRPA1 in dorsal root ganglion neurons.

(A) CPZ (50 μM) activates a subpopulation of AITC (100 μM)-sensitive mouse dorsal root ganglion (DRG) neurons. Individual responses from three different AITC- and capsaicin (CAP, 1 μM)-sensitive DRG neurons that were activated by CPZ. CPZ was applied for 20 s, AITC for 30 s and CAP for 10 s at intervals of 4 min allowing recovery. KCl (60 mM) was applied at the end of the experiment to to ensure viability of cultured neurons. (B) Averaged response of n = 135 AITC-sensitive neurons to three different concentrations of CPZ (25, 50, 100 μM, 20 s each). Note the concentration-dependence of the amplitude of Ca²⁺ transients. Straight traces represent mean and dotted traces represent SEMs. (C) Concentration-dependent increase of CPZ-induced [Ca²⁺]_i in AITC-sensitive DRG neurons normalized to a depolarizing stimulus with KCl (60 mM). The EC₅₀ of ∼30 μM was calculated by fitting to the Dose-Response function. Data are representative of two sets of experiments and show means ± SEM; for CPZ 1, 10, 25 μM: n = 169, for CPZ 25, 50, 100 μM: n = 138. (D) Illustrating populations of CPZ-, AITC-, and CAP-sensitive DRG neurons and their overlap. In a total of 906 imaged neurons (selected by their response to KCl), 200 (22%) were activated by CPZ (50 μM), 301 (33%) by AITC (100 μM) and 323 (36%) by CAP (1 μM) (detailed analysis of fractions, see SI Figure Legend 3). (E) CPZ (100 μM, 20 s) activates AITC (100 μM, 20 s)-sensitive DRG neurons from TRPV1-deficient mice. Individual responses from different AITC-sensitive neurons that were activated by CPZ. About 90% (509/572 cells) of the TRPV1-deficient DRG neurons that were AITC-sensitive were responsive to CPZ. CAP (1 μM, 10 s) did not induce Ca²⁺ transients in TRPV1^−/− neurons. (F) Lack of CPZ (100 μM)-induced Ca²⁺ influx in TRPA1-deficient DRG neurons. Individual responses from different CAP (1 μM)-sensitive DRG neurons (out of 448 neurons tested) that were neither activated by CPZ nor AITC (10 μM).

Figure 4. Capsazepine enemas desensitize local and distant pain responses.

(A) Acute writhing reactions in response to CPZ (531 μM) enemas during the first 5 min after application. Repeated CPZ enemas (twice daily) led to a progressive decrease of writhing responses. A dramatic step reduction was observed after 5 enemas in WT mice. By contrast, TRPA1^−/− mice treated with CPZ (531 μM) or WT mice treated with vehicle (PBS) enemas displayed only few writhing that occurred within the first 30 s (each n = 6). (B) Visceromotor responses (VMRs) to CPZ enemas. CPZ enemas twice daily led to a continuous decrease of VMRs in WT mice, similar to the course of writhing reactions. VMRs in response to vehicle were not significantly different from those to CPZ enemas in TRPA1^−/− (both n = 6). (A,B) WT CPZ group compared to vehicle group. (C) Repeated CPZ enemas attenuate eye-wipe behavior to AITC (100 μM). Both WT and TRPV1^−/− mice treated with CPZ enemas showed a profound reduction of eye-wipe counts compared with vehicle (both n = 6). Controls were WT mice without enemas receiving vehicle (PBS) drops into the eye. (D) CPZ enemas attenuate eye-wipe behavior to CAP (1 mM). WT mice treated with vehicle and TRPA1^−/− mice treated with CPZ enemas twice daily for 7 d showed numerous eye-wipe reactions to CAP instillation into the eye, tested 12 h after the last CPZ enema (both n = 6). WT mice treated with CPZ enemas showed profound reduction of eye wipe counts. (E) Thermal and (F) mechanical withdrawal thresholds of both hindpaws during the course of oral CPZ (531 μM) or AITC (500 μM) medication via drinking water. (E) CPZ and AITC over 10 d compared to the vehicle-treated group induced a progressive increase during 3–5 d in withdrawal latencies to radiant heat stimulation, which normalized within 2 weeks of wash-out (each n = 6). (F) Mechanical thresholds to stimulation with an electrodynamic von Frey filament showed no systematic trend under both compounds (each n = 6). All repeated measures ANOVA and Dunnett’s post-test, except in (C,D) Mann Whitney U-test.

Figure 5. Local and systemic desensitization of peptidergic sensory nerves by capsazepine (CPZ) indicated by altered release of calcitonin gene-related peptide (CGRP).

(A) Acute CGRP release from isolated colon preparations of WT mice induced by CPZ (100 μM); subsequent mustard oil (AITC, 100 μM) exposures fail to induce CGRP release, while unspecific depolarization by KCl (60 mM) is effective as normal. Data are means + SEMs (n = 8). (B–D) CPZ (100 μM)- and AITC (100 μM)-induced colonic CGRP release was abolished in mice pretreated with twice daily CPZ enemas for 7 d until the day before the release experiment, whereas CAP (1 μM)-induced colonic CGRP-release was strongly reduced but not abolished in these mice compared with controls. (**P < 0.01, ***P < 0.001, Mann Whitney U-test, each n = 6). (E) Similarly, AITC (100 μM)-induced CGRP release was abolished in isolated skin preparations from the hindpaws of mice pretreated with CPZ enemas. (F) In contrast, CAP (1 μM)-induced CGRP release was strongly reduced from the skin of these mice compared with controls. (**P < 0.01, ***P < 0.001, Mann Whitney U-test, each n = 6). Note that all KCl (60 mM) responses following desensitized CPZ and AITC responses were normal (B,C,E), whereas the KCl responses of the incompletely desensitized CAP-stimulated, neuron population (D,F) were as much reduced as in all control experiments, suggesting CGRP store depletion which is prevented by effective desensitization of the CPZ/AITC sensitive neuron subpopulation.