Forebrain delta opioid receptors regulate the response of delta agonist in models of migraine and opioid-induced hyperalgesia

Abstract

Delta opioid receptor (DOR) agonists have been identified as a promising novel therapy for headache disorders. DORs are broadly expressed in several peripheral and central regions important for pain processing and mood regulation; and it is unclear which receptors regulate headache associated symptoms. In a model of chronic migraine-associated pain using the human migraine trigger, nitroglycerin, we observed increased expression of DOR in cortex, hippocampus, and striatum; suggesting a role for these forebrain regions in the regulation of migraine. To test this hypothesis, we used conditional knockout mice with DORs deleted from forebrain GABAergic neurons (Dlx-DOR), and investigated the outcome of this knockout on the effectiveness of the DOR agonist SNC80 in multiple headache models. In DOR loxP controls SNC80 blocked the development of acute and chronic cephalic allodynia in the chronic nitroglycerin model, an effect that was lost in Dlx-DOR mice. In addition, the anti-allodynic effects of SNC80 were lost in a model of opioid induced hyperalgesia/medication overuse headache in Dlx-DOR conditional knockouts. In a model reflecting negative affect associated with migraine, SNC80 was only effective in loxP controls and not Dlx-DOR mice. Similarly, SNC80 was ineffective in the cortical spreading depression model of migraine aura in conditional knockout mice. Taken together, these data indicate that forebrain DORs are necessary for the action of DOR agonists in relieving headache-related symptoms and suggest that forebrain regions may play an important role in migraine modulation.

Figure 1

Chronic intermittent NTG increases DOR-eGFP expression. Mice were treated every other day for 9 days with vehicle of NTG (10 mg kg⁻¹ IP) and tissue was collected on day 10. (A) Representative ×10 images of saline and NTG treated brain regions. Hippocampus, inset—×20 zoom of representative neurons from the CA3 region of the hippocampus. Somatosensory Cortex, Striatum, inset—×20 zoom of representative image. (B) Comparison of number of DOR-eGFP positive cells in different brain regions of saline and NTG treated mice. (C) Average fluorescent intensity of DOR-eGFP positive cells in different brain regions of saline and NTG treated mice. For striatal regions overall fluorescence was quantified. DG Dentate gyrus, CA1 Region 1 of the hippocampus, CA2 Region 2 of the hippocampus, CA3 Region 3 of the hippocampus, SC primary somatosensory cortex, NAC nucleus accumbens core, NAS nucleus accumbens shell, Cpu caudate putamen *p < 0.05, **p < 0.01, ***p < 0.001 compared to saline treated group in the same region. Multiple t-test with Holm-Sidak correction. Specific p values can be found in Supplementary Table 1. n = 5/group.

Figure 2

Expression of δ receptors in Dlx-DOR conditional knockouts. Quantitative RT-PCR was performed on (A) striatum (B) hippocampus (C) trigeminal nucleus caudalis (TNC) and (D) trigeminal ganglia (TG) taken from adult loxP and Dlx-DOR mice. Multiple t-tests with Holm-Sidak correction showed a significant reduction of DOR in the two forebrain regions (*p < 0.05, **p < 0.01). Specific p values can be found in Supplementary Table 1. n = 6/genotype.

Figure 3

Role of central δ receptors in NTG-induced periorbital allodynia. (A) Mice were injected every other day for 9 days with vehicle or NTG (10 mg kg⁻¹ IP) and 1 h 15 min later with vehicle or SNC80 (10 mg kg⁻¹ IP). (B) They were tested for basal and post-treatment cephalic responses on days 1, 5, and 9. Repeated administration of NTG produced (C) basal and (D) acute periorbital allodynia in loxP and Dlx-DOR mice. (C) SNC80 prevented the development of basal hypersensitivity in loxP controls; an effect that was not observed in Dlx-DOR mice. Three-way repeated measures ANOVA revealed significant effects of day (F(2,2) = 62.43, p < 0.001), SNC80 dose (F(1,2) = 23.37, p < 0.001), and genotype (F(1,2) = 35.28, p < 0.001), as well as significant interaction of day X SNC80 dose (F(2,2) = 8.32, p = 0.0006), day X genotype (F(2,2) = 8.46, p = 0.0006), SNC80 dose X genotype (F(1,2) = 28.42, p < 0.001), and day X SNC80 dose X genotype (F(2,2) = 7.70, p = 0.001). (D) SNC80 inhibited the acute allodynic effects of NTG in loxP controls, but failed to do so in Dlx-DOR mice. Three-way repeated measures ANOVA revealed significant effects of SNC80 dose (F(1,2) = 181.4, p < 0.001), genotype (F(1,2) = 129.5, p < 0.001), and a significant SNC80 dose X genotype interaction (F(1,2) = 129.4, p < 0.001). n = 6/group for all panels. ***p < 0.001 compared to all other groups on the same day. n = 6/group.

Figure 4

Role of central δ receptors in opioid-induced hyperalgesia. Mice were treated twice daily for 4 days with vehicle or morphine (20 mg kg⁻¹ days 1–3 and 40 mg kg⁻¹ day 4, SC) and tested for cephalic responses on days 1,3 and 5 (A, grey circles), and peripheral responses on day 8 (A, shadowed circle). Repeated morphine treatment produced periorbital (B) and hind paw (C) hyperalgesia in both loxP and Dlx-DOR mice. (B) In control mice, SNC80 reversed morphine-induced hyperalgesia in the periorbital region but had no effect in Dlx-DOR mice. Three-way ANOVA revealed significant effects of morphine dose (F(1,2) = 46.28, p < 0.001), SNC80 dose (F(1,2) = 9.28, p = 0.004), genotype (F(1,2) = 8.80 p = 0.005), and significant morphine dose X SNC80 dose (F(1,2) = 8.33, p = 0.006), SNC80 dose X genotype (F(1,2) = 4.45, p = 0.041), and morphine dose X SNC80 dose X genotype interactions (F(1,2) = 6.40, p = 0.016). (C) SNC80 inhibited opioid-induced hyperalgesia in the hind paw in loxP but not Dlx-DOR mice. Three-way ANOVA revealed significant effects of morphine dose (F(1,2) = 71.02, p < 0.001), SNC80 dose (F(1,2) = 8.76, p = 0.005), genotype (F(1,2) = 7.90, p = 0.008), and significant SNC80 dose X genotype (F(1,2) = 6.12, p = 0.018), and morphine dose X SNC80 dose X genotype interactions (F(1,2) = 4.59, p = 0.038). n = 6–10/group for all panels. ***p < 0.001 vs loxP control with same drug regimen. ^##p < 0.01 vs. vehicle treated controls of the same genotype ^###p < 0.001 vs. vehicle treated controls of the same genotype. n = 6/group; except loxP-MOR-VEH, cKO-MOR-VEH and cKO-MOR-SNC80 where n = 7/group.

Figure 5

Role of central δ receptors in NTG-induced CPA. (A) Outline of experiment. (B) Conditioning with 10 mg kg⁻¹ NTG produced significant CPA in all genotypes. Conditioning with SNC80 prevented the development of NTG-induced CPA in controls, but SNC80 was ineffective in Dlx-DOR mice. Three-way ANOVA revealed significant main effects of NTG dose (F(1,2) = 13.40, p < 0.001) and SNC80 dose (F(1,2) = 10.40, p < 0.002). There was also a trend for genotype X SNC80 dose effect (F(1,2) = 3.14, p = 0.08) and a significant NTG dose X SNC80 dose X genotype interaction (F(1,2) = 4.14, p = 0.045). n = 8–12/group. *p < 0.001 vs. loxP control with same drug regimen. ^#p < 0.01 vs. vehicle treated controls of the same genotype. n = 12/group, except cKO-NTG-VEH where n = 13/group.

Figure 6

Role of central δ receptors in cortical spreading depression. (A) The location of the thinned skull and placement of the burr hole where KCl injection takes place and LFP recording is done. (B) To panel: Montage demonstrating the change in reflectance that can be seen in a typical CSD event. Lower panels: representative local field potential (LFP) recording that demonstrates the reflectance versus time that is typical for an hour of CSD recording following saline (upper panel) or SNC80 (lower panel) injection (IP). (C) SNC80 reduced the average number of CSD events in loxP controls, but failed to do so in Dlx-DOR animals. Two-way ANOVA revealed significant effects of SNC80 dose (F(1,27) = 6.92, p = 0.014), and a significant genotype X SNC80 dose interaction (F(1,27) = 4.61, p = 0.041). n = 7–8/group, **p < 0.01 compared to saline treated loxP group. n/group: loxP-VEH = 8, loxP-SNC80 = 9, cKO-VEH and cKO-SNC80 = 7/group.