Collybolide is a novel biased agonist of κ-opioid receptors with potent antipruritic activity

Collybolides Exhibit Structural Similarity to Sal A.

Colly and 9-epi-Colly share the common furyl-δ-lactone motif (i.e., a δ-lactone with its Cδ carbon substituted by the furyl group) present in Sal A (Fig. 1). A bicyclic core with rings A (cyclohexane) and B (δ-lactone) is a common feature of these terpene compounds. Colly has a trans-ring A/B junction, as in Sal A, whereas 9-epi-Colly has a cis junction. Besides the δ-lactone, the central ring A is substituted by an additional ring, a highly oxygenated cyclohexane with two esters and a ketone in Sal A, and a γ-lactone in the Collys. In contrast to Sal A, Collys have a bulky C-6 benzoyl ester in ring B (Fig. 1). In this study, we report the absolute configurations of both sesquiterpenes using single crystal X-ray diffraction analysis in the absence of any chemical transformation (SI Materials and Methods) and establish the epimeric nature of Colly and 9-epi-Colly at the level of C-9 (Fig. 1). Interestingly, the absolute configuration of the asymmetric Cδ carbon (furyl-bearing carbon) in the δ-lactone (Fig. 1) is inverse in Sal A and Collys. The presence of the furyl-δ-lactone motif in Collys, as well as their novel stereochemical features led us to examine whether they could function as κOR selective agonists.

Collybolides Exhibit High Affinity Binding to Human κOR.

Ligand binding assays using whole cells or membrane preparations expressing human κOR (hκOR) and [³H]Naloxone, a nonselective opioid receptor antagonist, as the radiolabeled ligand show that Colly exhibits a dose-dependent displacement profile with a K_i ∼10⁻¹⁰ M in membrane preparations (Fig. 2 and Table S1) and ∼10⁻⁷ M in whole cells (Fig. S1). Sal A, the synthetic κOR agonist U69,593, and the peptidic κOR agonist dynorphin A8 (Dyn A8) displace [³H]Naloxone binding with affinities in the nanomolar range (K_i ∼ 1–7 × 10⁻⁹ M) in membrane preparations (Table S1), and 10⁻⁸–10⁻⁹ M in whole cells (Table S2). Interestingly, epimerization at position 9 (9-epi-Colly) decreases the ability of the compound to displace [³H]Naloxone binding to hκOR (Fig. 2, Table S1, and Fig. S1). The displacement profile of receptor-bound [³H]Naloxone by Colly and 9-epi-Colly is not probe specific or dependent on the assay buffer conditions used, because a similar profile is seen with [³H]Diprenorphine or [³H]U69,593 or when using different assay buffer conditions (Fig. 2, Table S1, and Fig. S1). Together these results indicate that Collys bind to hκOR.

Collybolides Exhibit Selectivity to hκOR.

We examined the selectivity of Collys for hκOR by comparing their ability to displace [³H]Naloxone binding to hμOR or hδOR and to signal through these receptors. We find that neither Colly nor 9-epi-Colly displace [³H]Naloxone binding to hμOR or hδOR (Fig. 3). Also, Collys do not elicit changes in [³⁵S]GTPγS binding or ERK1/2 phosphorylation in cells expressing hμOR or hδOR (Fig. S2) nor do they block DAMGO (μOR agonist) or deltorphin II (δOR)-mediated increases in phospho ERK1/2 (Fig. S2). These results indicate that Collys exhibit selectivity toward hκOR. This selectivity is supported by the ability of Colly to displace [³H]Naloxone binding to membranes from WT mice but not from mice lacking κOR (Fig. 3 and Fig. S3).

Collybolide Is a High Affinity Agonist of hκOR.

Next we examined signaling by Colly using [³⁵S]GTPγS binding assays. Colly dose-dependently increases [³⁵S]GTPγS binding with an EC₅₀ of ∼10⁻⁹ M (Fig. 4A), a potency similar to that of 9-epi-Colly, Sal A, U69,593, and Dyn A8 (Fig. 4A and Table 1). Interestingly, 9-epi-Colly exhibits substantial lower efficacy suggesting partial agonism (Fig. 4A). The increase in [³⁵S]GTPγS binding mediated by both Collys is blocked by a κOR antagonist, nor-BNI (Fig. 4B). Moreover, Colly dose-dependently increases [³⁵S]GTPγS binding to WT but not κOR KO membranes (Fig. 4C), supporting Collys as κOR selective agonists.

Because κOR activation leads to inhibition of adenylyl cyclase activity and consequently decreases in cAMP levels (22), we examined effects of Colly on adenylyl cyclase activity. We find that Colly dose-dependently decreases adenylyl cyclase activity with a nanomolar potency comparable to that of Sal A (Fig. 4D and Table 1) but markedly lower than that of 9-epi-Colly that exhibits a potency in the micromolar range (Table 1). Together these results indicate that Collys exhibit agonistic activity at hκOR.

Collybolide Induces Robust Endocytosis of hκOR.

Because treatment with agonists can induce rapid and robust internalization of hκOR (23), we examined the ability of Colly to induce κOR endocytosis. A time course analysis reveals rapid and robust endocytosis (Fig. S4), and the dose–response analysis reveals an EC₅₀ for Colly ∼5 × 10⁻¹⁰ M that is comparable to that of Sal A and much greater than that of 9-epi-Colly (Fig. 4E and Table 1). These results show that Colly, like other high potency hκOR agonists, induces rapid and robust receptor internalization.

Collybolide Exhibits Biased Agonism at hκOR.

Next, we examined the ability of Collys to induce ERK1/2 phosphorylation. We find that Collys exhibit peak ERK1/2 phosphorylation at ∼3–5 min (Fig. S5). Interestingly, Colly exhibits an inverted “U-shape” profile with maximal stimulation at ∼10⁻⁹ M; at higher concentrations, Colly-mediated signaling is rapidly desensitizing both in HEK or CHO cells expressing hκOR (Fig. 4F and Fig. S5). In contrast, Sal A (like 9-epi-Colly and U69,593) exhibits a sigmoidal dose–response curve with maximal ERK1/2 phosphorylation at ∼10⁻⁷–10⁻⁶ M (Fig. 4F and Table 1). To ascertain whether the inverted U-shaped profile for ERK1/2 phosphorylation by Colly is also seen with other signaling pathways, we examined the profile for phosphorylation of Akt at S473 and at T308. We find that Colly exhibits a sigmoidal dose (and not an inverted U-shaped) response curve for Akt both at S473 and T308 (Fig. 4 G and H). Interestingly, Colly appears to be more efficacious than Sal A at phosphorylating Akt at both sites (Table 1). These observations suggest that Colly exhibits differences in signaling from Sal A at hκOR.

Collybolide Exhibits κOR Agonist Activity in Mice.

Next, we focused on Colly and evaluated its behavioral effects in mice. We find that the antinociceptive activity of Colly is similar to that of Sal A with peak activity at 20 min following drug administration (Fig. 5A). We did not find significant differences between Colly and vehicle-treated controls in the forced swim test, although Colly-treated mice tended to spend less time being immobile (Fig. 5B). In the elevated plus maze test, Colly and Sal A did not exhibit significant differences, although Colly-treated animals showed a higher tendency to spend less time in the open arms (i.e., tend to exhibit a higher level of anxiety), and this was blocked in animals treated with nor-BNI (Fig. 5C). The effect of Colly in the conditioned place preference/aversion assay shows that single daily injections for 4 d leads to significant (*P < 0.05) conditioned place aversion (Fig. 5D); this has been shown for κOR agonists including Sal A (3, 24). In contrast to these assays where Colly behaves in a manner similar to Sal A, Colly differs remarkably from Sal A when examined for its ability to block pruritus (non–histamine-mediated itch). Colly (2 mg/kg) significantly attenuates chloroquine-mediated scratching behavior; this is not seen with the same dose of Sal A (Fig. 5E). In addition, Colly exhibits dose-dependent inhibition of pruritus with a high potency (IC₅₀ = 0.08 mg/kg), and this effect is desensitized at higher doses (Fig. 5E). Also, Colly-mediated attenuation of pruritus can be reversed by the κOR antagonist nor-BNI (Fig. 5F). We also find that Colly treatment has no effect on animal movement in an open field test (Fig. 5G). These results indicate that Colly is a novel, high-potency hκOR agonist with behavioral properties distinct from that of Sal A that taken with its biased agonistic activity makes it an attractive candidate for further studies exploring the molecular pharmacology of κOR.

Natural Compounds, Materials, and Analytical Data.

Sal A was obtained from In Solve Scientific; the analytical report (HPLC, MS, UV, and ¹H NMR) with 95.9% purity (HPLC) was in agreement with the reported data for C₂₃H₂₈O₈, i.e., 432.47 g/mol (40): (i) ESI(+) MS: m/z = 455.25 (exptl) for 455.17 (theor) C₂₃H₂₈O₈Na⁺; (ii) UV absorption at 210 nm (max) for furan; (iii) a comparison of the ¹H NMR spectrum (400 MHz; CDCl₃; TMS internal reference) with the graphical ¹H NMR spectrum (800 MHz, CDCl₃) of Sal A (with revised assignments) in the PhD thesis of Munro (41) was carried out in full conformity with: 1.11 ppm (s,3H): CH3-19; 1.45 ppm (s, 3H): CH₃-20; 2.16 ppm (s, 3H): CH₃-22; 3.72 ppm (s, 3H): CH₃-23; 5.14 ppm (1H): C2-H; 5.52 ppm (1H): C12-H; 6.37 ppm (1H): furyl-H14; ∼7.4 ppm (2H): furyl-H15+H16. The two sesquiterpenes of fungal origin extracted from C. maculata (for a review on the phylogeny of C. maculata, see ref. 42), collybolide and 9-epi-collybolide, were part of the collection of two of the coauthors (A.C. and J.P.) after their discovery (21). The structures of collybolide and 9-epi-collybolide were reassessed by X-ray crystallography using monocrystals of collybolide and 9-epi-collybolide. Absolute configurations were determined using the Hooft method, which uses the Bijvoet-pair intensities measured in reciprocal space from anomalous dispersion of the atoms collected (43–45). ¹H and ¹³C NMR spectra for both the natural compounds were found to be in agreement as reported in the literature (21, 46); 10 mM stock solutions of Sal A, collybolide, and 9-epi-collybolide in DMSO were used for all pharmacological assays. U69,593 (34), the arylacetamide κOR agonist [(5α,7α,8β)-(+)-N-methyl-N-[7-(1-pyrrolidinyl)-1-oxaspiro[4,5]dec-8-yl]benzenacetamide], Dyn A8, H-Tyr-Gly-Gly-Phe-Leu-Arg-Arg-Ile-OH, and protease inhibitor mixture were obtained from Sigma. The radioligands, [³H]Diprenorphine, [³H]Naloxone, and [³H]U69,593, with specific radioactivities of 42.3, 61.1, and 37.4 Ci/mmol, respectively, and [³⁵S]GTPγS with a specific radioactivity of 1,250 Ci/mmol were from Perkin-Elmer. κOR KO mice brains were a gift from Charles Chavkin (University of Washington, Seattle).

Cell Culture.

HEK-293 cells were grown in DMEM containing 10% (vol/vol) FBS and penicillin-streptomycin. CHO cells were grown in F12 media containing 10% FBS and penicillin-streptomycin. Cells were transfected with human μOR (hμOR), δOR (hδOR), or hκOR cDNAs using lipofectamine as per manufacturer’s protocol (Invitrogen). Forty-eight hours after transfection cells were used to prepare membranes for receptor binding and signaling assays as described below. Saturation binding assays using [³H]Naloxone (0–10 nM) in 50 mM Tris⋅Cl buffer, pH 7.8, containing 1 mM EGTA, 5 mM MgCl₂, and protease inhibitor mixture in the absence or presence of 10 μM U69,593, deltorphin II, or DAMGO show that the cells express similar amounts of hμOR, hδOR, or hκOR (874 ± 240 fmol/mg protein for hμOR, 837 ± 171 for hδOR, and 816 ± 253 for hκOR).

Membrane Preparation.

Membranes were prepared from HEK-293 cells expressing either hκOR, hδOR, hμOR, or from cerebral cortex or whole brains of WT or κOR KO mice as described previously (35).

Receptor Displacement Binding Assays.

Assays were carried out in glass tubes treated with 0.1% (vol/vol) aquaSil-siliconizing fluid in double distilled water (DDW) for 10 min. Tubes were rinsed once in DDW and air dried. Competition binding assays were carried out as described previously (35, 36). Briefly, membranes (200 μg) from HEK cells expressing either hμOR, hδOR, or hκOR (200 μg/tube) were incubated with [³H]Diprenorphine, [³H]Naloxone, or [³H]U69,593 (3 nM final concentration) in the presence of different concentrations (10⁻¹³–10⁻⁵ M) of κOR ligands for 2 h at 30 °C in a buffer comprising of 50 mM Tris⋅Cl buffer, pH 7.8, containing 1 mM EGTA, 5 mM MgCl₂, and protease inhibitor mixture. Other buffers used in binding assays included (i) 50 mM Tris⋅Cl buffer, pH 7.8, containing protease inhibitor mixture; (ii) 50 mM Tris⋅Cl, pH 7.4, containing 100 mM NaCl, 5 mM MgCl₂, 0.2 mM EGTA, and protease inhibitor mixture; and (iii) 50 mM Tris⋅Cl buffer, pH 7.8, containing 1 mM EGTA, 5 mM MgCl₂, 1 μM GTPγS, and protease inhibitor mixture. Nonspecific binding was determined using [³H]Naloxone (3 nM final concentration) and 10 μM cold naloxone. Bound radioactivity was collected by filtration under reduced pressure using GF/B filters presoaked in 50 mM Tris⋅Cl, pH 7.4, containing 0.1 mg/mL BSA and 0.2% (wt/vol) polyethylenimine. Values obtained with 10⁻¹³ M cold ligand were taken as 100%.

In a separate set of experiments, cerebral cortex membranes or whole brain membranes (100 μg) from WT or κOR KO mice were incubated with [³H]Naloxone (3 nM) and specific bound counts determined in the presence of 10 μM of either collybolide, Sal A, U69,593, or naloxone as described above.

[³⁵S]GTPγS Binding Assays.

Assays were carried out in glass tubes treated as described above for receptor binding assays. Membranes (20 μg) from HEK-293 cells expressing hκOR, hμOR, or hδOR were subjected to a [³⁵S]GTPγS binding assay in 50 mM Tris⋅Cl buffer, pH 7.8, containing 1 mM EGTA, 5 mM MgCl₂, and protease inhibitor mixture in the absence or presence of either collybolide, 9-epi-collybolide, Sal A, U69,593, or Dyn A8 (10⁻¹³–10⁻⁵ M) as described previously (35, 36). Nonspecific binding was determined in the presence of 10 μM cold GTPγS. Values obtained with 10⁻¹³ M ligand were taken as 100%. In a separate set of experiments, membranes were preincubated with nor-BNI (10 μM) for 30 min before carrying out [³⁵S]GTPγS binding as described above. In another set of experiments, membranes (20 μg) from the cerebral cortex of WT mice or from κOR KO mice were subjected to a [³⁵S]GTPγS binding assay with collybolide (10⁻¹³–10⁻⁵ M) as described above.

Measurement of Adenylyl Cyclase Activity.

Membranes from HEK-293 cells expressing hκOR (2 μg/well) were incubated with either collybolide, 9-epi-collybolide, Sal A, U69,593, or Dyn A8 (10⁻¹³–10⁻⁵ M) for 30 min at 37 °C in 50 mM Hepes buffer, pH 7.4, containing 20 μM forskolin, 10 mM MgCl₂, 100 mM NaCl, 200 μM ATP, 10 μM GTP, and 100 μM IBMX. cAMP levels were determined using the HitHunter cAMP MEM chemiluminescence detection kit from DiscoveRx. Values obtained at 10⁻¹³ M ligand were taken as 100%.

Measurement of ERK1/2 Phosphorylation.

CHO or HEK-293 cells expressing hκOR (2 × 10⁵ cells per well) were seeded into a 24-well polylysine-coated plate. Cells were grown for 16 h in growth media without FBS. Cells were treated with either vehicle, collybolide, 9-epi-collybolide, Sal A, or U69,593 (10⁻¹³–10⁻⁶ M) for 3 min at 37 °C. For time course studies, cells were treated with either vehicle, collybolide, or 9-epi-collybolide (100 nM) for different time periods (30 s to 30 min). ERK1/2 phosphorylation was measured as described (36, 37) using antibodies to phospho ERK1/2 (1:1,000; cat. no. 4370S; Cell Signaling Technology) and to total ERK1/2 (1:1,000; cat. no. 4696S; Cell Signaling Technology) as primary antibodies. Anti-rabbit IRDye 800 (1:10,000; cat. no. 926-32211; Li-Cor) and anti-mouse IRDye 680 (1:10,000; cat. no. 926-68070; Li-Cor) were used as secondary antibodies.

In a separate set of experiments, hμOR or hδOR cells were treated with collybolide, 9-epi-collybolide (10⁻¹³–10⁻⁶ M), with 1 μM DAMGO ± 10 μM collybolide/9-epi-collybolide or with 1 μM deltorphin II ± 10 μM collybolide/9-epi-collybolide as described above.

Measurement of Akt Phosphorylation.

CHO or HEK-293 cells expressing hκOR (2 × 10⁵ cells per well) were seeded into a 24-well polylysine-coated plate. Cells were grown for 16 h in growth media without FBS. Cells were treated with either vehicle, collybolide, 9-epi-collybolide, Sal A, or U69,593 (10⁻¹³–10⁻⁶ M) for 3 min at 37 °C. Akt phosphorylation at S473 or T308 was measured by Western blotting as described (36, 37) using rabbit anti-phospho Akt S473 (1:1,000; cat no. 4060S; Cell Signaling Technology), rabbit anti-phospho Akt T308 (1:1,000; cat no. 2965S; Cell Signaling Technology), and mouse anti-tubulin antibodies (1:50,000; cat. no. T5168; Sigma-Aldrich) as primary antibodies. Anti-rabbit IRDye 800 (1:10,000; cat. no. 926–32211; Li-Cor) and anti-mouse IRDye 680 (1:10,000; cat. no. 926-68070; Li-Cor) were used as secondary antibodies.

Receptor Internalization.

HEK-293 cells expressing hκOR (2 × 10⁵ cells per well) were seeded into a 24-well polylysine-coated plate. The following day, the plate was kept on ice, and cells were incubated at 4 °C for 1 h with 1:1,000 anti-HA antibody in media to label cell surface κOR. Cells were washed three times and then treated without or with 100 nM collybolide, Sal A, U69,593, and Dyn A8 for different time intervals (0–120 min) at 37 °C to facilitate receptor internalization. At the end of the incubation period, cells were briefly fixed (3 min) with 4% (wt/vol) paraformaldehyde followed by three washes (5 min each) with PBS. Receptors present at the cell surface were determined using 1:1,000 dilution (in PBS containing 1% BSA) of anti-rabbit IgG coupled to HRP (Vector Laboratories) as described previously (47, 48). The percent internalized receptors was calculated by taking total cell surface receptors before agonist treatment for each individual experiment as 100% and subtracting percent surface receptors following agonist treatment. In a separate set of experiments, the effects of different doses (0–1 μM) of agonists for 30 min on hκOR-mediated internalization was examined.

Animal Studies.

Male C57BL/6J mice, 10–12 wk of age (Jackson Laboratories), were used in these studies. Mice were housed in groups of three to five mice per cage in a humidity- and temperature-controlled room with a 12-h light/dark cycle (lights on: 07:00–19:00 hours) and given free access to food and water throughout the experiment. All behavioral procedures were conducted according to ethical guidelines/regulations approved by the Icahn School of Medicine Animal Care and Use Committee. Animal care and all experimental procedures were in accordance with guidelines from the National Institutes of Health.

Tail Flick Assay.

Mice (7–10/group) were injected i.p. with either vehicle [6% DMSO (vol/vol) in saline], Sal A, or collybolide (2 mg/kg), and antinociception was measured using the tail-flick assay (38). The intensity of the heat source was set to 10 (this results in a basal tail-flick latency of 5–7 s for most animals) and the tail-flick latency was recorded at the indicated time period (0–60 min) after vehicle or drug administration. Cutoff latency was set at 20 s to minimize tissue damage.

CPP/Aversion.

The CPP chambers were housed within a sound attenuating room with red light and contained three distinct sensory environments. The conditioning chambers were differentiated on the basis of wall color and flooring, with the gray conditioning chamber containing a steel mesh floor, the black and white striped conditioning chamber containing a floor made of large steel mesh, and the center white chamber containing a floor made of steel rods. Guillotine doors connect each side compartment to the center compartment. On day 1, mice were placed in the white center compartment and allowed to freely roam the entire apparatus for 20 min. The amount of time spent in each compartment (gray or striped) was recorded to assess the unconditioned preference and then animals were returned to their home cages. Any mouse showing a strong unconditioned preference (>70%) to either chamber was excluded from the experiment. Animals (7–10/group) were administered i.p. with vehicle in one chamber and 8 h later with vehicle/collybolide (2 mg/kg, i.p.) in the other chamber for 45 min. These injections were repeated for 4 d using an unbiased, counterbalanced design in which half the animals had gray and half had striped chambers as the drug-paired side. On the test day (day 6), animals were placed in the central white chamber and allowed to freely roam the entire apparatus. Time spent in each test compartment was recorded separately for each animal for 20 min. The change of preference was calculated as the difference (in seconds) between the time spent in the drug-paired compartment on the testing day and the time spent in this compartment in the preconditioning session.

Forced Swim Test.

Mice (7–10/group) were injected i.p. with either vehicle [6% DMSO (vol/vol) in saline] or collybolide (2 mg/kg), individually placed inside a glass beaker (height 44 cm, diameter 22 cm) filled with water at a depth of 30 cm, at 25 ± 2 °C and behavior recorded for 6 min. At the end of the test mice were removed, dried, and returned to their home cages. Mice were scored blind for the treatment conditions for time spent immobile.

Elevated Plus Maze.

Mice (7–10 per group) were injected i.p. with either vehicle [6% DMSO (vol/vol) in saline], Sal A, or collybolide (2 mg/kg) placed onto the center of the elevated plus maze and the number of open- and closed-arm entries and the time spent in open arms recorded. Percent time spent in open arms was used as a measure of anxiety. A separate group of animals was administered with nor-BNI (10 mg/kg, i.p.) 20 min before collybolide administration.

Open Field.

Mice were injected with vehicle [6% DMSO (vol/vol) in saline], Colly (2 mg/kg, i.p.), Sal A (2 mg/kg, i.p.), or nor-BNI (10 mg/kg i.p.) + Colly (2 mg/kg, i.p.) 20 min before open field test. Distance traveled (in centimeters) over 10 min was quantified using Ethovision XT tracking system (Noldus).

Chloroquine-Mediated Itch Behavior.

Mice were habituated to an acrylic chamber for 1 h before testing. Ten minutes before inducing itch mice were injected i.p. with vehicle [0.9% saline +2% DMSO (vol/vol)], collybolide, or Sal A (2, 20, 200, and 2,000 µg/kg) over the course of 4 d starting with the lowest dose on day 1. Each mouse was assigned to one group consistently throughout the experiment. To induce itch mice were injected s.c. with chloroquine-phosphate (20 mg/kg) into the base of the neck and were placed into the recording chamber. For experiments involving nor-BNI, on day 1 mice were habituated to the chamber for 1 h and 10 min before inducing itch mice were treated with vehicle to obtain a baseline amount of chloroquine-induced itch. On the next day, this procedure was repeated with mice receiving 200 µg/kg collybolide 10 min before inducing itch. The same cohort of mice were injected i.p. with nor-BNI (10 mg/kg) and 48 h later mice received 200 µg/kg collybolide 10 min before inducing itch. Because nor-BNI’s effects are stable for weeks following injection (49), these mice were assessed for itch 72 h later following a vehicle injection 10 min before inducing itch to examine whether i.p. nor-BNI has an impact on itch. All mice were videotaped for 15 min. The total number of scratches in 5-min intervals were scored by experimenters blinded to the drug treatments. A scratch was defined as hind paw movements directed to the site of injection (30, 39).

Statistical Analysis.

Data were analyzed using Prism 6.0 (Graph Pad) and statistical analysis carried out using t test or one-way ANOVA.

Natural Compounds, Materials, and Analytical Data.

Sal A was from In Solve Scientific, Colly and 9-epi-Colly were from the collection of two of the coauthors (A.C. and J.P.) after their discovery (21) (analytical details in SI Materials and Methods); 10 mM stock solutions of Sal A, Colly, and 9-epi-Colly in DMSO were used in all assays. U69,593 (34), Dyn A8, and protease inhibitor mixture were from Sigma. [³H]Diprenorphine, [³H]Naloxone, [³H]U69,593, and [³⁵S]GTPγS with specific radioactivities of 42.3, 61.1, 37.4, and 1250 Ci/mmol, respectively, were from Perkin-Elmer. κOR KO mice brains were a gift from Charles Chavkin (University of Washington, Seattle).

Cell Culture.

HEK-293 or CHO cells were grown in DMEM or F12 media, respectively, that contained 10% (vol/vol) FBS and penicillin-streptomycin. Cells were transfected with hμOR, hδOR, or hκOR cDNAs using lipofectamine as per the manufacturer’s protocol (Invitrogen). Saturation binding assays using [³H]Naloxone (0–10 nM) show that the cells express similar amounts of receptors (874 ± 240 fmol/mg protein for hμOR, 837 ± 171 for hδOR, and 816 ± 253 for hκOR).

Membrane Preparation.

Membranes were prepared from HEK-293 cells expressing hκOR, hδOR, and hμOR, from cerebral cortex or whole brains of WT or κOR KO mice as described previously (35).

Receptor Displacement Binding Assays.

Glass tubes treated for 10 min with 0.1% (vol/vol) aquaSil-siliconizing fluid in double distilled water were used. Membranes from HEK-cells expressing hμOR, hδOR, or hκOR (200 μg/tube), cerebral cortex, or whole brains (100 μg/tube) of WT or κOR KO mice were incubated with [³H]Diprenorphine, [³H]Naloxone, or [³H]U69,593 (3 nM final concentration) in the presence of different concentrations (10⁻¹³–10⁻⁵ M) of κOR ligands for 2 h at 30 °C as described previously (35, 36). Details of buffers used are given in SI Materials and Methods.

[³⁵S]GTPγS Binding Assays.

Assays were carried out in glass tubes treated as described above. Membranes (20 μg) were subjected to a [³⁵S]GTPγS binding assay with κOR ligands (10⁻¹³–10⁻⁵ M) as described previously (35, 36). nor-BNI (10 μM) was added 30 min before the assay.

Measurement of Adenylyl Cyclase Activity.

Membranes (2 μg) were treated with 20 μM forskolin in the absence or presence of κOR ligands (10⁻¹³–10⁻⁵ M) for 30 min at 37 °C. cAMP levels were determined using the HitHunter cAMP MEM chemiluminescence detection kit from DiscoveRx.

Measurement of ERK1/2 and Akt Phosphorylation.

CHO or HEK-293 cells expressing hκOR, hμOR, or hδOR (2 × 10⁵ cells per well) were treated with vehicle or κOR ligands (10⁻¹³–10⁻⁶ M) for 3 min at 37 °C. ERK1/2 phosphorylation and Akt phosphorylation at S473 or T308 were measured by Western blot as previously described (36, 37).

Receptor Internalization.

HEK-293 cells expressing hκOR (2 × 10⁵ cells per well) were incubated at 4 °C for 1 h with 1:1,000 anti-HA antibody to label cell surface κOR. Cells were treated without or with 100 nM of κOR ligands for different time intervals (0–120 min) or with different doses (0–1 μM) of κOR ligands for 30 min at 37 °C. Cells were fixed and cell surface receptors determined by ELISA using 1:1,000 dilution (in PBS containing 1% BSA) of anti-rabbit IgG coupled to HRP (Vector Laboratories).

Animal Studies.

Male C57BL/6J mice, 10–12 wk of age (Jackson Laboratories) were housed in groups of three to five per cage in a humidity- and temperature-controlled room with a 12-h light/dark cycle (lights on 07:00–19:00 hours) and given free access to food and water throughout the experiment. Behavioral procedures were according to ethical guidelines/regulations approved by the Icahn School of Medicine Animal Care and Use Committee. Animal care and all experimental procedures were in accordance with guidelines from the National Institutes of Health.

Tail Flick Assay.

Mice (7–10/group) were injected i.p. with vehicle [6% (vol/vol) DMSO in saline], Sal A, or Colly (2 mg/kg), and antinociception was measured using the tail-flick assay (38).

Conditioned Place Preference/Aversion.

Animals (7–10/group) were administered i.p. with vehicle in one chamber and 8 h later with vehicle/Colly (2 mg/kg, i.p.) in the other chamber. These injections were repeated for 4 d using an unbiased, counterbalanced design. On the test day (day 6), animals were placed in the central white chamber, and place preference/aversion was measured for 20 min. Change in preference (seconds) = time spent in drug-paired compartment on test day − time spent in this compartment in the preconditioning session.

Forced Swim Test.

Mice (7–10/group) were injected i.p. with vehicle or Colly (2 mg/kg) and placed inside a glass beaker (height 44 cm, diameter 22 cm) filled with water at a depth of 30 cm, at 25 ± 2 °C, and behavior was recorded for 6 min. Mice were scored blind for the treatment conditions for time spent immobile.

Elevated Plus Maze.

Mice (7–10/group) were injected i.p. with vehicle, Sal A, or Colly (2 mg/kg) and placed onto the center of the elevated plus maze, and the number of open- and closed-arm entries and time spent in open arms were recorded. Percentage time spent in open arms was used as a measure of anxiety. A separate set of animals was given nor-BNI (10 mg/kg, i.p.) 20 min before Colly administration.

Open Field.

Mice were injected with i.p. with vehicle, Colly (2 mg/kg), Sal A (2 mg/kg), or nor-BNI (10 mg/kg) + Colly (2 mg/kg) 20 min before the open field test. Distance traveled (cm) over 10 min was quantified using Ethovision XT tracking system (Noldus).

Chloroquine-Mediated Itch Behavior.

Mice (7–10/group) were injected i.p. with vehicle (0.9% saline + 2% DMSO), Colly, or Sal A (2–2,000 µg/kg) 10 min before inducing itch. To induce itch, mice were injected s.c. with chloroquine-phosphate (20 mg/kg) into the base of the neck, placed into the recording chamber, and videotaped for 15 min. Total number of scratches in 5-min intervals were scored by experimenters blinded to the drug treatments. A scratch was defined as hind paw movements directed to the site of injection (30, 39). Another set of animals was evaluated for the effect of nor-BNI on collybolide-mediated itch attenuation.

Statistical Analysis.

Data were analyzed using Prism 6.0 (Graph Pad), and statistical analysis was carried out using t test or one-way ANOVA.