Investigation of the role of βarrestin2 in kappa opioid receptor modulation in a mouse model of pruritus

Abstract

The kappa opioid receptor (KOR) is involved in mediating pruritus; agonists targeting this receptor have been used to treat chronic intractable itch. Conversely, antagonists induce an itch response at the site of injection. As a G protein-coupled receptor (GPCR), the KOR has potential for signaling via G proteins and βarrestins, however, it is not clear which of these pathways are involved in the KOR modulation of itch. In this study asked whether the actions of KOR in pruritus involve βarrestins by using βarrestin2 knockout (βarr2-KO) mice as well as a recently described biased KOR agonist that biases receptor signaling toward G protein pathways over βarrestin2 recruitment. We find that the KOR antagonists nor-binaltorphimine (NorBNI) and 5'-guanidinonaltrindole (5'GNTI) induce acute pruritus in C57BL/6J mice, with reduced effects in KOR-KO mice. βArr2-KO mice display less of a response to KOR antagonist-induced itch compared to wild types, however no genotype differences are observed from chloroquine phosphate (CP)-induced itch, suggesting that the antagonists may utilize a KOR-βarrestin2 dependent mechanism. The KOR agonist U50,488H was equally effective in both WT and βarr2-KO mice in suppressing CP-induced itch. Furthermore, the G protein biased agonist, Isoquinolinone 2.1 was as effective as U50,488H in suppressing the itch response induced by KOR antagonist NorBNI or CP in C57BL/6J mice. Together these data suggest that the antipruritic effects of KOR agonists may not require βarrestins.

Keywords: Biased ligand; KOR antagonist; Kappa opioid receptor; Mouse models of itch; Pruritus; U50,488H.

Figure 1

NorBNI and 5′GNTI induce scratching behaviors in C57BL/6J mice with some response still present in KOR-KO mice. (A) NorBNI (1.0-10 mg/kg, s.c._neck) and (B) 5′GNTI (0.1-1.0 mg/kg, s.c._neck) induce scratching behaviors over a 1 hour period compared to vehicle (two-way ANOVA for interaction of dose and time (p < 0.0001) for both; Vehicle n = 10; all other groups n = 5-6). (C) The sum of the total number of scratching bouts over the hour session is presented for each dose. (D) KOR-KO mice display scratching behaviors in response to NorBNI (10 mg/kg, s.c._neck) and 5′GNTI (0.1 mg/kg, s.c._neck) compared to vehicle (two-way ANOVA for interaction of treatment and time (p < 0.0001) for both; n = 3 all groups). (E) The sum of the scratching bouts over the hour session for KOR-KO reveals a significant effect of the KOR antagonists in the absence of KOR (t-test comparing treatment to vehicle, **p < 0.01; n = 3 all groups). (F) For visualization comparison, the fold stimulation over vehicle comparison of the antagonist effects presented in C and F are shown. Data are presented as mean ± SEM.

Figure 2

βarr2-KO mice are less responsive to NorBNI and 5′GNTI. (A) WT and (B) βarr2-KO mice display a dose-dependent scratching response to NorBNI (s.c._neck) compared to vehicle (WT: two-way ANOVA for dose, p < 0.0001; n = 6-8; βarr2-KO: two-way ANOVA for dose, p < 0.0001; n = 6-8) as well as to 5′GNTI (C) WT and (D) barr2-KO mice display a dose-dependent response to 5′GNTI (s.c._neck) compared to vehicle (WT: two-way ANOVA for dose, p < 0.0001; n = 7-8; βarr2-KO: two-way ANOVA for dose, p < 0.0001; n = 7-8). (E) Comparison of the sum of dose effects over the hour test period, WT mice display a significantly greater response to NorBNI than barr2-KO mice (two-way ANOVA for genotype (p = 0.0015) and dose (p < 0.0001), *p<0.05 Bonferroni post hoc analysis; n = 6-8). (F) Comparison of the sum of dose effects over the hour test period, WT mice display a significantly greater response to 5′GNTI than βarr2-KO mice (two-way ANOVA for genotype (p = 0.0006) and dose (p < 0.0001), *p < 0.05 Bonferroni post hoc analysis; n = 7-8). Data presented as mean ± SEM.

Figure 3

WT and βarr2-KO mice display equally robust scratching behaviors in response to administration of the general pruritic agent, chloroquine phosphate (CP). (A, B) CP (20, 40 mg/kg, s.c_neck) induces itch in WT and βarr2-KO mice (two-way ANOVA for dose, p < 0.0001 for both WT and βarr2-KO genotypes; WT, Vehicle n = 7, CP n = 5 for both doses; βarr2-KO, Veh n = 8, CP n = 5 for both doses). Two-way ANOVA comparison between genotypes, within each dose including vehicle, did not reveal a genotype effect when analyzed with the 5 minute binning data (p > 0.05). (C) Comparison of the sum of the response over the hour testing session at each dose revealed a significant effect of dose with no difference between the genotypes (two-way ANOVA for genotype and dose (p < 0.0001), ***p < 0.001, ****p < 0.0001 Bonferroni post hoc analysis). Data are presented as mean ± SEM.

Figure 4

A balanced kappa agonist suppresses scratching behaviors induced by NorBNI and chloroquine phosphate in C57BL/6J mice. (A, B) U50,488H (0.1 - 3.0 mg/kg, s.c.) administered 10 minutes prior to NorBNI (5 mg/kg, s.c._neck) suppresses the antagonist-induced scratching response. For the vehicle response, vehicle (s.c.) was administered 10 min prior to a challenge with vehicle (s.c._neck). (A) Dose effect over the time course compared to vehicle: two-way ANOVA for dose (p < 0.0001); Vehicle, n = 11; all other groups, n = 5-7. (B) Sum of the response over the hour test session: one-way ANOVA vs. Veh + NorBNI (p < 0.0001), ***p < 0.001, ****p < 0.0001 Bonferroni post hoc analysis. (C, D) U50,488H (1.0 and 3.0 mg/kg, s.c.) administered 10 minutes prior to CP (40 mg/kg, s.c._neck) suppresses the scratching response. For the vehicle response, vehicle (s.c.) was administered 10 min prior to a challenge with vehicle (s.c._neck). (C) Dose effect over the time course compared to vehicle: two-way ANOVA for dose (p < 0.0001); Vehicle, n = 11; all other groups, n = 6-7. (D) Sum of the response over the hour test session: one-way ANOVA vs. Veh + CP (p < 0.0001), ****p < 0.0001 Bonferroni post hoc analysis. (E, F) Twenty-four hour pretreatment with kappa antagonist NorBNI blocks the antipruritic effects of U50,488H on CP-induced itch. (E) Effect of Veh (i.p.) + U50,488H (1 mg/kg, s.c.) + CP (40 mg/kg, s.c._neck) vs. NorBNI (10 mg/kg, i.p.) + U50,488H (1 mg/kg, s.c.) + CP (40 mg/kg, s.c._neck): two-way ANOVA (p < 0.0001). There were no differences detected between vehicle pretreatment and NorBNI pretreatment groups (p > 0.05). (F) The comparison of the sum of the responses over the hour test session reveals suppression of CP-induced itch by systemic U50,488H injection and reversal of U50,488H effects by the 24 hour pretreatment with NorBNI: one-way ANOVA, ****p < 0.0001 Bonferroni post hoc analysis compared to vehicle pretreatment. Data are presented as the mean ± SEM.

Figure 5

The KOR agonist U50,488H suppresses scratching behaviors induced by chloroquine phosphate independent of the presence of βarr-2. (A) U50,488H (1.0 and 3.0 mg/kg, s.c.) administered 10 minutes prior to CP (40 mg/kg, s.c._neck) suppresses the scratching response in both genotypes. For the vehicle response, vehicle (s.c.) was administered 10 min prior to a challenge with vehicle (s.c._neck). Compared to vehicle: (A) WT: two-way ANOVA for dose (p < 0.0001); (B) βarr2-KO: two-way ANOVA for dose p < 0.0001); all groups, both genotypes n = 5. Two-way ANOVA comparison between genotypes, within each dose, did not reveal a genotype effect when analyzed with the 5 minute binning data (p > 0.05). (C) Comparison of the sum of the response over the hour testing session at each dose revealed a significant effect of dose, but no difference between the genotypes (two-way ANOVA for genotype (p > 0.05), and dose (p < 0.0001), ****p < 0.0001 Bonferroni post hoc analysis; all groups, both genotypes n = 5). Data are presented as mean ± SEM.

Figure 6

A KOR agonist that is biased towards G protein signaling over βarrestin2 recruitment suppresses chloroquine phosphate-induced itch. (A). Isoquinolinone 2.1 (Iso2.1) is a highly selective KOR agonist; ^†affinity measures for radioligand binding are from Zhou et al., 2013. (B) Iso2.1 (1.0 and 3.0 mg/kg, s.c.) administered 10 minutes prior to CP (40 mg/kg, s.c._neck) suppresses the scratching response compared to vehicle: two-way ANOVA for dose (p < 0.0001); n = 5 - 6. (C) Significance is also revealed by the comparison of the sum of the response over the hour test session: one-way ANOVA, ***p < 0.001, ****p < 0.0001 Bonferroni post hoc analysis; n = 5-6). (D, E), Twenty-four hour pretreatment with KOR antagonist NorBNI (10 mg/kg, i.p.) blocks the antipruritic effects of Iso2.1 (1 mg/kg, i.p.) on CP-induced itch. (D) Effect of Veh + Iso2.1 + CP vs. NorBNI + Iso2.1 + CP: two-way ANOVA (p < 0.0001). There were no differences detected between vehicle pretreatment and NorBNI pretreatment groups (p > 0.05). (E) The comparison of the sum of the responses over the hour test session reveals suppression of CP-induced itch by systemic Iso2.1 injection and reversal of Iso2.1 effects by the 24 hour pretreatment with systemic NorBNI: one-way ANOVA, NorBNI + Veh + CP vs. Veh + Iso2.1 + CP, **p < 0.01; NorBNI + Iso2.1 + CP vs. Veh + Iso2.1 + CP, ^#p < 0.05, student's t-test. Data are presented as the mean ± SEM.

Figure 7

A G protein biased KOR agonist remains potent in the absence of βarr-2. (A) Iso2.1 (1 and 3 mg/kg, s.c.) administered 10 minutes prior to CP (40 mg/kg, s.c._neck) suppresses the scratching response in both genotypes. Compared to vehicle: (B) WT: two-way ANOVA for dose (p < 0.0001); βarr2-KO: (p < 0.0001); all groups, both genotypes n = 5. Two-way ANOVA comparison between genotypes, within each dose, did not reveal a genotype effect when analyzed with the 5 minute binning data (p > 0.05). (C) Comparison of the sum of the response over the hour testing session at each dose revealed a significant effect of dose, but no difference between the genotypes: two-way ANOVA for genotype (p > 0.05), and dose (p < 0.0001), ****p < 0.0001 Bonferroni post hoc analysis; all groups, both genotypes n = 5. Data are presented as mean ± SEM.

Figure 8

A G protein biased KOR agonist does suppress spontaneous locomotor activity. While U50,488H (3 mg/kg, s.c.) significantly suppressed activity (two way ANOVA vs. vehicle, p = 0.0003), Iso2.1 (3 mg/kg, s.c.) had no significant effect on spontaneous activity when compared to vehicle (two-way ANOVA vs. vehicle, p>0.05). Twenty-four hour pretreatment with NorBNI (10 mg/kg, i.p.) prevented the decreased locomotor activity associated with U50,488H (two-way ANOVA vs. 3 mg/kg U50488H, p < 0.0001). Vehicle, n = 9; Iso2.1, n = 6; U50,488H, n = 7; NorBNI + U50,488H, n = 6. Data presented as mean ± SEM.