Repetitive stress in mice causes migraine-like behaviors and calcitonin gene-related peptide-dependent hyperalgesic priming to a migraine trigger

Abstract

Migraine is one of the most disabling disorders worldwide but the underlying mechanisms are poorly understood. Stress is consistently reported as a common trigger of migraine attacks. Here, we show that repeated stress in mice causes migraine-like behaviors that are responsive to a migraine therapeutic. Adult female and male mice were exposed to 2 hours of restraint stress for 3 consecutive days, after which they demonstrated facial mechanical hypersensitivity and facial grimace responses that were resolved by 14 days after stress. Hypersensitivity or grimace was not observed in either control animals or those stressed for only 1 day. After return to baseline, the nitric oxide donor sodium nitroprusside (SNP; 0.1 mg/kg) elicited mechanical hypersensitivity in stressed but not in control animals, demonstrating the presence of hyperalgesic priming. This suggests the presence of a migraine-like state, because nitric oxide donors are reliable triggers of attacks in migraine patients but not controls. The stress paradigm also caused priming responses to dural pH 7.0 treatment. The presence of this primed state after stress is not permanent because it was no longer present at 35 days after stress. Finally, mice received either the calcitonin gene-related peptide monoclonal antibody ALD405 (10 mg/kg) 24 hours before SNP or a coinjection of sumatriptan (0.6 mg/kg). ALD405, but not sumatriptan, blocked the facial hypersensitivity due to SNP. This stress paradigm in mice and the subsequent primed state caused by stress allow further preclinical investigation of mechanisms contributing to migraine, particularly those caused by common triggers of attacks.

Figure 1.. The repeated stress paradigm primes male and female mice to subthreshold doses of a NO donor.

Facial withdrawal thresholds were measured in male (A) and female (C) mice following repeated restraint stress. Upon returning to baseline thresholds at 14 days following the final day of stress, stressed mice were administered either 0.1 mg/kg SNP (n = 8 males, n = 8 females) or vehicle (n = 10 males, n = 6 females). All control mice were given SNP (n = 8 males, n = 7 females). (†) denotes statistical significance between stressed mice that received SNP and control mice; (*) denotes statistical significance between stressed mice that received vehicle and control mice. (§) denotes significance between stressed mice that received SNP and stressed mice that received vehicle. In a separate cohort of mice, grimace responses (B, D) were measured for stressed (n = 15) and control (n = 13) mice following acute stress and administration of 0.1 mg/kg SNP. Two-way RM ANOVA followed by Bonferroni multiple comparison analysis indicated significant differences between stressed and control mice following SNP in both males and females. Data are represented as means ± SEM. ^†,§p<0.05, ^††,§§p<0.01, ^†††,***p<0.001, ^{††††,}****p< 0.0001. See Table 2 for additional results of analysis.

Figure 2.. Single-restraint stress does not produce facial allodynia or priming to SNP.

(A) Facial withdrawal thresholds of male mice following a single session of stress and administration of either SNP(n=10) or vehicle (n=7) after 5 days of testing or naïve mice (n=9). (B) Facial withdrawal thresholds of female mice following a single session of stress and administration of either SNP (n=7) or vehicle (n=6) after 5 days of testing or naïve mice (n=6). Two-way RM with Bonferroni multiple comparison analysis indicates there was no statistical difference among control mice and all cohorts that were stressed in both the acute phase and the priming phase (mean ±SEM). See table 2 for additional statistical analysis.

Figure 3.. Repeated stress induces transient priming to SNP in male and female mice.

Facial withdrawal thresholds were measured in (A) male and (B) female mice following acute stress (n = 8 male, 8 female). Following the final day of stress, animals were tested out to 35 days and received either 0.1 mg/kg SNP (n = 8 males, n = 9 females) or vehicle (n = 9 males, n = 8 females). All controls received SNP. (†) denotes statistical significance between stressed mice will receive SNP and control mice. (*) denotes statistical significance between stressed mice that will receive vehicle and control mice. (§) denotes statistical significance between stressed mice that received vehicle and stressed miced that received SNP. Two-way RM ANOVA with Bonferroni multiple comparison analysis indicated a significant difference between stressed and control mice in the acute phase, but revealed no significant differences following administration of SNP. Data are represented as means ± SEM. *p<0.05, ^{††††,}****p<0.0001. See Table 2 for additional results of analysis.

Figure 4.. Effects of CGRP monoclonal antibodies in repeated stress induced priming to SNP in mice.

Facial withdrawal thresholds were measured in male (A) and female (B) mice following acute stress (n = 10 males, n = 7 females). 15 days following the final day of stress, animals received either the anti-CGRP antibody ALD405 (n = 10 males, n = 8 females) or an isotype control (n = 9 males, n = 9 females). 24 hours following administration of antibody, all animals received 0.1 mg/kg SNP. In the acute phase, (†) denotes statistical significance between control mice and stressed mice that will receive ALD405 and SNP. (*) denotes statistical significance between control mice and stressed mice that will receive isotype control IgG. In the priming phase, (§) denotes statistical significance between stressed mice that received ALD405 and those that received the isotype control prior to SNP. Two-way ANOVA followed by Bonferroni multiple comparison analysis revealed significant differences in the priming phase between female stressed mice that received ALD405 and those that received the isotype. Statistical significance between these groups was observed in males at 48 hours following SNP. Data are represented as means ± SEM. ^§p<0.05,^§§p < 0.01, ^§§§p<0.001, ^{††††,}****p<0.0001. See Table 2 for additional results of analysis.

Figure 5.. Sumatriptan does not block SNP responses in stress-primed mice.

Facial withdrawal thresholds of female mice either naïve (n=8) or following acute stress and administration of either SNP (0.1 mg/kg; n=8), suma (0.6 mg/kg; n=7), or a coadministration of suma and SNP (n=7). In the acute phase, (*) denotes significance between control mice and stressed mice that will receive SNP; (^) denotes significance between control mice and stressed mice that will receive suma; (†) denotes significance between control mice and stressed mice that will receive suma and SNP. For 24hr-7DPS in the acute phase, the statistical significance is the same for each time point. Two-way RM with Bonferroni multiple comparison analysis indicates a statistical difference among control mice and all cohorts that were stressed. In the priming phase there was no statistical difference detected between stressed mice that received SNP and those that received suma and SNP (mean ±SEM, ^^p<0.01, ***p<0.001, ****^,^^^^^{, ††††}p<0.0001). See table 2 for additional statistical analysis.

Figure 6.. Repeated stress primes female mice to decreased dural pH.

(A) Facial withdrawal thresholds of female mice following acute stress and priming to dural pH 7.0 (n=12) or 7.4 (n=12) and control mice given pH 7.0 (n=12) and pH 7.4 (n=11). (B) Grimace scores of mice that were first stressed and then administered either pH 7.0 (n=8) or pH 7.4 (n=8) or mice that were naïve and then administered either pH 7.0 (n=8) or pH 7.4 (n=8). For the acute phase, (*) denotes significance between stressed and control mice that will receive dural pH 7.0 ; (^) denotes significance between stressed and control mice that will receive dural pH 7.4. For the priming phase, (†) denotes significance between stressed mice that received pH 7.0 and stressed mice that received pH 7.4. Dural pH stimuli are given at the time point indicated by the arrows in A; prior to dural pH injections, animals were only exposed to either stress or control. Two-way ANOVA RM with Bonferroni multiple comparison analysis indicate a significant difference between stressed and control mice and stressed mice that received pH 7.0 compared to stressed mice that received pH 7.4 (mean ±SEM, *p<0.05, ^††p<0.01, ***^, ^^^p<0.001, ****^,^^^^^{,††††}p<0.0001). See table 2 for additional statistical analysis.