TRPM8 contributes to sex dimorphism by promoting recovery of normal sensitivity in a mouse model of chronic migraine

Abstract

TRPA1 and TRPM8 are transient receptor potential channels expressed in trigeminal neurons that are related to pathophysiology in migraine models. Here we use a mouse model of nitroglycerine-induced chronic migraine that displays a sexually dimorphic phenotype, characterized by mechanical hypersensitivity that develops in males and females, and is persistent up to day 20 in female mice, but disappears by day 18 in male mice. TRPA1 is required for development of hypersensitivity in males and females, whereas TRPM8 contributes to the faster recovery from hypersensitivity in males. TRPM8-mediated antinociception effects required the presence of endogenous testosterone in males. Administration of exogenous testosterone to females and orchidectomized males led to recovery from hypersensitivity. Calcium imaging and electrophysiological recordings in in vitro systems confirmed testosterone activity on murine and human TRPM8, independent of androgen receptor expression. Our findings suggest a protective function of TRPM8 in shortening the time frame of hypersensitivity in a mouse model of migraine.

Fig. 1. Chronic nitroglycerin administration induces generalized long-lasting mechanical hypersensitivity in male and female mice.

a Model of chronic migraine induced by five intraperitoneal injections of 10 mg/kg nitroglycerin or its vehicle administered every other day (days 0, 2, 4, 6 and 8). Hind paw mechanical sensitivity was assessed with von Frey filaments 2 h after nitroglycerin administration (blue lines), before nitroglycerin injections and after the end of the treatments on days 10, 12, 14, 16, 18 and 20 (black lines). Tissue samples were obtained at the end. b Left panels. Nitroglycerin (red) induces similar acute hypersensitivity in males and females, whereas the vehicle (black, 5% dextrose, 0.105% propylene glycol) did not induce nociceptive sensitization. b Right panels. Chronic administration of nitroglycerin induces a long-lasting hypersensitivity that is persistent exclusively in females, whereas female or male mice treated with vehicle show normal sensitivity. Data are mean ± S.E.M. **p < 0.01, ***p < 0.001 vs. vehicle. ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 vs. baseline. Two-way RM ANOVA n = 6 mice per each vehicle group, n = 7 mice per each NTG group. NTG nitroglycerin, Veh Vehicle. Source data provided in the Source Data file, statistical results in Supplementary Table 1.

Fig. 2. TRPA1-related activity is essential for mechanical hypersensitivity after nitroglycerin and TRPA1 expression and function increase in trigeminal ganglia of chronically exposed males and females.

a TRPA1 deletion prevents acute and chronic mechanical hypersensitivity after nitroglycerin in male and female mice. b Trigeminal ganglia of male and female mice show increased TRPA1 mRNA expression after chronic nitroglycerin exposure. c Representative calcium traces in response to 70 µM AITC normalized to 40 mM KCl in trigeminal cells of mice chronically exposed to vehicle or nitroglycerin. d AITC responses increase in trigeminal cultures from nitroglycerin-treated animals regardless of the sex. Average response size of AITC-sensitive neurons vs. KCl responses. e Wild-type trigeminal neurons responding to 100 µM nitroglycerin also respond to 70 µM AITC. TRPA1 knockout mice lack these responses. f 15% of trigeminal neurons respond to AITC and NTG in wild-type mice vs. 1–2% in TRPA1 knockouts. g TRPA1-transfected HEK293 cells and IMR90 cells constitutively expressing hTRPA1 show calcium responses after 10, 50 and 100 µM nitroglycerin. h Trigeminal culture exposed to vehicle, nitroglycerin or the exocytosis inhibitor DD04107. Cells labeled with pan-neuronal marker antiMAP2 (green), antiαCGRP (red) and nuclei marked with DAPI (blue). i Nitroglycerin decreases intracellular αCGRP immunoreactivity and DD04107 prevents this. Mean mechanical thresholds (a), expression (b), calcium response (d, f) or red fluorescence (i) ± S.E.M. Datapoints without error bars represent values of individual animals, except for i where dots represent quantified images. a ^p < 0.05, ^^p < 0.01, ^^^p < 0.001 vs. vehicle. ^#p < 0.05, ##p < 0.01 vs. baseline, two-way RM ANOVA, n = 6 mice per condition except for WT female group which was of five mice; b ***p < 0.001 two-way ANOVA, n = 6 samples from six different mice per condition, obtained after one behavioral experiment; d ***p < 0.001 two-way ANOVA, n = 5 samples from five different mice per condition, obtained in three independent experiments; f *p < 0.05 two-sided Mann–Whitney U n = 5 samples from five different mice per condition, obtained in three independent experiments; i ***p < 0.001, one-way ANOVA n = 26 images from samples of five mice per condition, obtained in five independent experiments. NTG nitroglycerin, Veh vehicle. Source data are provided in Source Data file, statistical results in Supplementary Table 2.

Fig. 3. TRPM8 allows recovery of normal sensitivity in male mice exposed to nitroglycerin or formalin.

a Wild-type and TRPM8 knockout males display similar acute mechanical hypersensitivity after nitroglycerin (left panel). Deletion of this receptor prevents recovery of baseline mechanical sensitivity after chronic nitroglycerin treatment in males (right). Acute and long-lasting hypersensitivity is similar in wild-type and knockout females. b Trigeminal TRPM8 mRNA expression is similar regardless of sex or nitroglycerin/vehicle treatment. c Calcium responses of trigeminal cultures from males and females chronically exposed to nitroglycerin or vehicle after exposure to TRPM8 agonist WS12 (500 nM). d Average size of calcium transients after WS12 is similar in neurons from males and females. Response sizes vs. respective KCl responses. e Female mice chronically exposed to nitroglycerin show increased mechanical sensitivity when compared to vehicle-treated mice. This sensitivity remained unchanged after 10 mg/kg WS12 i.p. f Nocifensive behavior in the acute phase of the formalin test (5 min) is significantly alleviated in females receiving 6 nmol WS12 i.pl. g TRPM8 antagonist AMTB (i.p.) unmasks latent mechanical pain sensitization in mice chronically exposed to nitroglycerin that already recovered their basal sensitivity. h 10 mg/kg AMTB administered 1 h after formalin injections induces significant reinstatement of licking behavior in males. Mean mechanical thresholds (a, e, g), expression (b), response size (d), nocifensive behavior (f, h) ± S.E.M. Datapoints without error bars represent values of individual animals. a ##p < 0.01 vs. baseline. ^^p < 0.01 vs. vehicle two-way RM ANOVA n = 5 mice per condition. b Two-way ANOVA n = 6 samples from six different mice per condition, obtained after one behavioral experiment. d Two-way ANOVA n = 5 samples from five different mice per condition, obtained after three independent experiments. e ***p < 0.001 nitroglycerin vs. vehicle, three-way ANOVA n = 6(Veh + Veh), n = 5(NTG + Veh), n = 6(Veh + WS12), n = 7(NTG + Veh WS12) mice. f *p < 0.05 vs. vehicle. Multiple two-sided t-test n = 4(WS12) or 7(Vehicle) mice. g +p < 0.05 vs. vehicle. Two-sided Friedman test n = 6 mice. h *p < 0.05 vs. vehicle. Multiple two-sided t-test n = 6 mice per condition. ORX orchidectomy, NTG nitroglycerin, Veh vehicle. Source data are provided in Source Data file, statistical results in Supplementary Table 3.

Fig. 4. Testosterone stimulates neuronal TRPM8 to promote recovery of normal mechanosensitivity after chronic nitroglycerin treatment.

a Orchidectomized and sham mice display similar acute hypersensitivity after nitroglycerin (left panel), however only orchiectomized mice remain sensitized on day 21 (right). b Left, all orchidectomized mice show similar acute hypersensitivity after a first nitroglycerin dose, whereas after the last dose, orchidectomized TRPM8KO mice develop stronger sensitization. Right, mice receiving testosterone supplementation show inhibition of the long-lasting hypersensitivity induced after chronic nitroglycerin. This attenuation allows recovery of baseline sensitivity in wild-type mice, but not in TRPM8KO mice. c Testosterone-supplemented wild-type mice that recovered normosensitivity reinstate sensitization after AMTB on day 21. This response is absent in TRPM8 knockouts d Sensitized wild-type females recover basal sensitivity after acute testosterone treatment on day 21. e Cultured trigeminal neurons showing calcium transients after 10 pM testosterone also respond to 500 nM WS12 whereas TRPM8 knockouts are unresponsive. Right, 7% of neurons from wild-type trigeminal cultures show calcium transients after 10 pM testosterone and 500 nM WS12. This is abolished in TRPM8 knockouts. f 10 pM testosterone elicits calcium transients in HEK293 cells when heterologously expressing rat or human TRPM8. Mean mechanical thresholds (a–d), response size (e, f) ± S.E.M. Datapoints without error bars represent values of individual animals. a ***p < 0.001 vs. vehicle, ^$$p < 0.01 vs. sham, three-way RM ANOVA. n = 6 (Sham-Vehicle, Sham-NTG), n = 5 (ORX-Vehicle), n = 7 (ORX-NTG) mice. b ^^p < 0.01, ^^^p < 0.001 vs. TRPM8KO; ^&p < 0.05, ^&&p < 0.01 vs. testosterone vehicle; ##p < 0.01, ^###p < 0.001 vs. baseline, three-way RM ANOVA. n = 5 (WT-Veh), n = 6 (WT-Testos and TRPM8KO-Veh), n = 7 (TRPM8KO-Testos) mice. c ^^^p < 0.001 vs. TRPM8KO; ^&p < 0.05, vs. testosterone vehicle; ^##p < 0.01 vs. basal, three-way RM ANOVA n = 5 (WT Veh), n = 6 (WT Testos and TRPM8KO Veh), n = 7 (TRPM8KO Testos) mice. d ^^^p < 0.001 vs. TRPM8KO, ^#p < 0.05 vs. basal, two-way ANOVA n = 7–9 mice per condition. e *p < 0.05 two-sided Mann–Whitney U n = 4 samples from four different mice per condition obtained in two independent experiments. f ***p < 0.001 vs. control, two-sided Mann–Whitney U n = 64–100 cells of two independent cultures per group. ORX orchidectomy, NTG nitroglycerin, Veh vehicle, Testos testosterone, TRPM8KO TRPM8 knockout. Source data are provided in Source Data file, statistical results in Supplementary Table 4.

Fig. 5. Testosterone increases outward current of heterologously expressed human TRPM8, independently of androgen receptor expression.

a Current density (pA/pF) vs. voltage relationship evidences increase in outward rectifying current 10 pM testosterone in HEK293 cells expressing human TRPM8 (hTRPM8). 10 µM AMTB inhibits testosterone effect (n = 5 cells from two independent cultures. Each cell received a testosterone pulse followed by a pulse of AMTB with testosterone). b 10 pM testosterone increases basal TRPM8 current at +80 mV (n = 7 cells from two independent cultures; p < 0.017 vs. Basal current) and 10 µM AMTB reverses its effect (n = 5 cells from two cultures, p < 0.009 vs. 10 pM testosterone). c hTRPM8 HEK 293 cells (hTRPM8 HEK) transfected with small interfering RNA (siRNA) against androgen receptor (AR) (n = 7 from two cultures) and control hTRPM8 HEK transfected with scrambled siRNA (n = 6 from two cultures) show similar testosterone-induced outward rectifying current. This current is absent in wild-type HEK293 cells (WT HEK, n = 6 from 1 culture) lacking TRPM8. d At +80 mV, testosterone-evoked current is similar in anti-AR siRNA or scrambled siRNA-transfected hTRPM8 HEK cells, and this current is absent in wild-type HEK (p < 0.046 vs. anti-AR siRNA, p < 0.005 vs. scrambled siRNA). e Western blot of AR (110 kDa) and β-tubulin loading control (50 kDa) shows disrupted AR expression in anti-AR siRNA-transfected hTRPM8 HEK cells. WT HEK and scrambled siRNA-transfected hTRPM8 HEK cells display similar AR expression. f 15 nM AMTB (IC50 for WS12 inhibition, Supplementary Fig. 5) shifts to the right the dose-response curve for testosterone-induced currents. EC50 for testosterone (7pM, 95% CI 6–8) increases in presence of AMTB (34 pM, 95% CI 16–61, n = 9 cells from three independent cultures for both dose-response curves). “n” indicates number of registered cells. b, d Data were subjected to Shapiro–Wilk normality tests, then Kruskal–Wallis followed by Mann–Whitney U tests were applied. **p < 0.01, *p < 0.05. Data expressed as mean ± SEM, dots on bar charts represent individual cells. EC50s were estimated fitting data to a sigmoidal dose-response curve with constraints (Top = 100, Bottom = 0, GraphPad Prism Software). e is a representative western blot of two independent transfections. Source data are provided in Source Data file, statistical results in Supplementary Table 5.