European Journal of Pharmacology

Transcription

1 European Journal of Pharmacology 674 (2012) Contents lists available at SciVerse ScienceDirect European Journal of Pharmacology journal homepage: Neuropharmacology and Analgesia Synergistic antinociceptive actions and tolerance development produced by morphine fentanyl coadministration: Correlation with μ-opioid receptor internalization Arturo Silva-Moreno a, Claudia Gonzalez-Espinosa a, Martha León-Olea b, Silvia L. Cruz a,b, a Departamento de Farmacobiología, Cinvestav, Sede Sur. Calzada de los Tenorios # 235, Col. Granjas Coapa, México D.F., 14330, México b Departamento de Neuromorfología Funcional, Dirección de Investigaciones en Neurociencias, Instituto Nacional de Psiquiatría Ramón de la Fuente Muñiz. Calzada México-Xochimilco 101, Col. San Lorenzo Huipulco, Tlalpan, México D.F , México article info abstract Article history: Received 11 June 2011 Received in revised form 16 October 2011 Accepted 27 October 2011 Available online 31 October 2011 Keywords: Fentanyl Morphine μ-opioid receptor internalization Synergism Tolerance Isobologram It has been described that coadministration of opioids with low doses of other analgesics can reduce adverse effects and increase antinociception, but combinations of two μ-opioid receptor agonists have been poorly explored. The objective of this work was threefold: 1) to evaluate the antinociceptive combination of i.c.v. morphine and fentanyl at different doses; 2) to compare the antinociception produced by acute or repeated administration of an effective morphine dose (1 μg) alone, or combined with a low fentanyl dose (1 ng); and 3) to correlate these effects with μ-opioid receptor internalization in periaqueductal gray matter and locus coeruleus. Antinociception was evaluated by the tail-flick test and receptor internalization was analyzed by confocal microscopy in Wistar rats. Drug interactions were examined by administering combinations of opioids in 1:3, 1:1 and 3:1 ratios of their respective ED 50 fractions. For tolerance and internalization studies, animals were i.c.v. injected only once (acute treatment) or twice a day until five administrations were completed. Our results show that morphine and fentanyl have synergistic effects. The combination of 1 ng fentanyl with 1 μg morphine increases the magnitude and duration of antinociception not only after a single injection, but also after five administrations when tolerance develops to morphine alone. Increased and longlasting antinociception correlates positively with increased β-arrestin 2 activity and μ-opioid receptor internalization in periaqueductal gray matter and locus coeruleus. These results suggest that combined administration of morphine and fentanyl increases long-lasting antinociception and β-arrestin 2 signaling contributes to the combination effects Elsevier B.V. All rights reserved. 1. Introduction Morphine and fentanyl are effective μ-opioid receptor analgesics, but their repeated administration can produce tolerance and physical dependence (Trescot et al., 2008; Waldhoer et al., 2004; Walwyn et al., 2010). A strategy to attenuate opioid-induced adverse effects is to combine them with low doses of non-steroidal anti-inflammatory drugs (Goldstein, 2002; Silva-Moreno et al., 2009) or other opioids (Smith, 2008). Controlled studies of combined opioids are scarce and their results, controversial (Mercadante et al., 2004). Dottrens et al. (1992) reported that the analgesic effects of a morphine sufentanyl combination after surgery lasted longer than those of morphine alone. In contrast, Friedman et al. (2008) did not find benefits from a morphine fentanyl combination in patients recovering from surgery. In preclinical studies, synergistic effects have been observed between L-methadone and Corresponding author at: Calzada de los Tenorios # 235, Col. Granjas Coapa, México, 14330, D.F., Mexico. Tel.: ; fax: address: slcruz@cinvestav.mx (S.L. Cruz). other μ-opioid receptor agonists, including morphine (Bolan et al., 2002). Opioid analgesics differ in efficacy (Sirohi et al., 2008), affinity for different μ-opioid receptor isoforms (Pasternak, 2004; Xu et al., 2011), activation of transcriptional factors (Zheng et al., 2010a), and β-arrestin-dependent receptor internalization (von Zastrow, 2010). Arrestin-dependent signals can be pharmacologically dissociated from those dependent on G-protein activation by using biased ligands that induce differential coupling of G-protein coupled receptors favoring one pathway or the other. High internalization capacity has been associated with preferential β-arrestin signaling, whereas low endocytosis is considered an indicator of preferential G-protein dependent activation (Luttrell and Gesty-Palmer, 2010). Fentanyl, a β arrestin-biased ligand (Zheng et al., 2008, 2010b), readily induces receptor endocytosis and recycling (Bohn et al., 2004), whereas morphine, a G protein-biased ligand, produces lower internalization (Beaulieu, 2005; Rajagopal et al., 2010). Acute coadministration of morphine with low doses of 2-Ala-4- mephe-5-gly-enkephalin (DAMGO) or methadone, two μ agonists with high internalization capacity, enhances antinociception and /$ see front matter 2011 Elsevier B.V. All rights reserved. doi: /j.ejphar

2 240 A. Silva-Moreno et al. / European Journal of Pharmacology 674 (2012) receptor endocytosis (Hashimoto et al., 2006; He et al., 2002). Also, coadministration of morphine and DAMGO results in reduced tolerance and enhanced μ-opioid receptor endocytosis (He and Whistler, 2005; Koch et al., 2005). To our knowledge, no systematic studies have been done to correlate the antinociception produced by acute and chronic morphine fentanyl combination, μ-opioid receptor internalization and β-arrestin 2 activation. The aim of this work was threefold: 1) to determine the type of morphine fentanyl (i.c.v.) interaction in the tail-flick test; 2) to evaluate the antinociceptive effects of acute and repeated administration of morphine combined with a low fentanyl dose; and 3) to compare these results with those produced by morphine alone, analyzing changes in μ-opioid receptor endocytosis in periaqueductal gray matter and locus coeruleus. These brain regions were selected due to their pivotal role in pain processing and physical dependence (Millan, 2002). The i.c.v. route was used to avoid possible differences in absorption between drugs and ensure immediate access to the brain. The hypotheses to be tested were that: a) morphine and fentanyl would produce synergistic effects; b) a combination of an effective morphine dose with a low fentanyl dose would induce antinociception and prevent the development of morphine tolerance; and c) μ-opioid receptor internalization would be better observed with the morphine fentanyl combination than with morphine alone. 2. Material and methods 2.1. Animals We used male adult Wistar rats ( g) from our breeding facilities, individually housed with free access to drinking water and food under controlled conditions of temperature (22±2 C) and light (12/12 h light/dark cycle). Our local Ethics Committee on Animal Experimentation (Protocol No ) approved all experimental procedures, which followed regulations established in the Mexican official norm for the use and care of laboratory animals NOM-062- ZOO-1999 and guidelines on ethical standards for investigation of experimental pain in animals (Zimmermann, 1983) Drugs Morphine sulfate and fentanyl citrate were obtained via the Mexican Health Secretariat. Ketamine hydrochloride was bought from Probiomed (Mexico City, Mexico) and xylazine hydrochloride, from Pisa (TulaHidalgo,Mexico).Allanalgesic drugs were dissolved in sterile saline solution and administered i.c.v., in a final volume of 5 μl during 1 min, using an infusion pump (KD Scientific, USA). This volume was kept constant whether opioid drugs were administered individually or in combination Surgical procedure Rats were anesthetized with a mixture of ketamine/xylazine (45/ 12 mg/kg, i.p.) and mounted on a stereotaxic apparatus to place a permanent guide cannula (10 mm long) in the right lateral cerebral ventricle using the following coordinates from bregma: 1.5, midline: 1.4, and 4 from dura (Paxinos and Watson, 1986). The cannula was fixed with two screws and dental cement. Animals underwent surgery on Friday and were allowed to recover during the weekend. To reduce the stress associated with experiments, rats were handled for 15 min twice a day, during 2 additional days in the experimental environment, with the injection cannula inserted into the guide cannula. Nociception studies were performed at least five days after surgery. Once experiments were completed, animals were perfused with paraformaldehyde and brain coronal sections were observed in the microscope to corroborate the injection site Antinociception assessment We used a tail-flick apparatus with a radiant heat source connected to an automatic timer (Ugo Basile, Italy) and determined the increase in tail withdrawal latency for nociception assessment (Nance and Sawynok, 1987). The stimulus intensity was adjusted at the beginning of the study to get a baseline tail-flick latency of 6.0±0.5 s. Each animal was tested three times before the first drug injection to obtain a mean baseline latency value. Rats that showed no flicking within 5.5 to 6.5 s (5 to 10% of the total) were discarded. We established a cutoff time of 15 s to avoid tail skin tissue damage. Tail withdrawal latency was recorded every 15 min during the first hour after drug administration, and every 30 min in the second hour Protocol Ten independent groups of rats (n=6, each) were used to determine the dose response relationships for acute i.c.v. administration of fentanyl (0.001, 0.01, 0.031, 0.1 or 0.2 μg) or morphine (0.31, 0.56, 1.0, 3.1 or 5.6 μg). Each animal was tested only once. Based on these curves, we selected the doses to be used in combination for synergistic analysis, tolerance development and μ-opioid receptor internalization. For chronic treatment, four independent groups of animals were used (n=9, each). Rats received five injections (at 8:00 and 20:00 h until treatment was completed) of saline solution, 1 μg morphine, 1 ng fentanyl or the combination of these opioids at the same doses. In all cases, the final injection volume was 5 μl. Nociception was evaluated only after the morning dose to minimize exposure to noxious stimuli. After completing treatment, 6 animals from each group were used to evaluate withdrawal signs after a challenge naloxone dose (31 μg, i.c.v.) administered 30 min after the last morphine injection, and precipitated abstinence signs were monitored for 30 min. The remaining animals were used for immunohistochemistry analysis Immunohistochemistry To evaluate μ-opioid receptor presence, β-arrestin 2 activation and changes in μ-opioid receptor cell localization, independent groups of rats (n=3, per group) were deeply anesthetized with sodium pentobarbital 15 min after one or five administrations of each treatment and then perfused intracardially with 200 ml of 0.1 M saline phosphate buffer (ph ), followed by 200 ml cold 4% paraformaldehyde in the phosphate buffer. The brain was dissected, removed, and post-fixed during 4 h in the paraformaldehyde solution at 4 C. Finally, the brain was cryoprotected at least 48 h in 30% sucrose buffer solution. Coronal sections (30 μm thick) of periaqueductal gray matter (bregma 6.04 to 6.8 mm) and locus coeruleus (bregma 9.68 to 10.3 mm) were cut with a cryostat (Cryostat 1750, Leitz). Freefloating sections were washed four times with phosphate buffer containing 0.4% Triton X-100 and blocked with 2% donkey serum, 2% bovine serum antigen (Sigma, St. Louis MO, USA) for 60 min. Sections were incubated with goat antiserum (1:50 dilution) raised against C-18 of β-arrestin 2 (Catalog No. SC-6383, Santa Cruz, Santa Cruz, CA, USA) and rabbit antiserum (1:150 dilution) raised against amino acids of the cloned rat μ opioid receptor (No. Catalog PC165L Calbiochem Merck, Darmstadt, Germany). The epitope corresponds to the complete exon 4 and the adjacent three amino acids of exon 3 of MOR-1 C terminal (Abbadie et al., 2000). Once sections were washed three times in phosphate buffer with Triton, they were incubated for two additional hours at 37 C with the secondary antibodies: tetramethyl rhodamine isothiocyanate-labeled donkey anti-rabbit, and fluorescein isothiocyanate-labeled donkey anti-goat (both obtained from Jackson Immunoresearch) diluted at 1:150. After three more washes, sections were mounted on glass slides

3 A. Silva-Moreno et al. / European Journal of Pharmacology 674 (2012) with antifade kit (Invitrogen, Carlsbad, CA, USA) to minimize photobleaching, and cover-slipped. Negative controls, designed to assess non-specific antibody binding, were conducted in parallel. In these controls, sections were treated as described but they were not exposed to primary antibody. Preliminary experiments were also conducted with different dilutions of primary and secondary antibodies, in an effort to optimize visualization of the two proteins of interest and minimize the background. Slides were kept in the dark until analysis. Images of brain slices (0.45 μm optical thickness) were acquired with a Zeiss META 510 laser scanner microscope. Fluorescein isothiocyanate signal (green) was obtained after excitation with an argon (488 nm) laser using an emission window of nm. Tetramethyl rhodamine isothiocyanate signal (red) was excited with a helium-neon laser (543 nm), and emission was detected through a 560 nm long-pass filter in multi-track mode to eliminate cross talk between the channels. The pinhole was 1 airy unit. Images were obtained with a 40 oil-immersion objective (numerical aperture: 1.3). Detector gain and background were adjusted to provide optimal fluorescent range; all images were acquired in the same conditions. The images were stored and analyzed with Zeiss laser scanner microscope 4.0 spl software. For analysis of whole cell immunoreactivity to μ-opioid receptor and to β-arrestin 2, the mean for total specific pixels on the surface of 30 neurons (10 per animal), was obtained with Zeiss 4.0 software and the data were utilized for the graphs shown in Figs. 4 and 5. For μ-opioid receptor internalization data shown in Fig. 6, the intensity of μ-opioid receptor and β-arrestin 2 signal was analyzed along of a line of 10 μm that started on the plasma membrane of each cell and crossed through the intracellular space. Specific signal distribution was obtained from at least five cells in each treatment and a representative cell with its distribution pattern is shown. All cell images were processed with Photoshop CS3 (Adobe Systems) using identical values for contrast and brightness in all cases Data and statistical analysis Antinociception was evaluated by an increase in the latency to tail withdrawal as a function of time. The latencies of the tail-flick responses were converted to the percentage of maximum possible effect (%MPE) for each animal, at each time, according to the following formula: %MPE=[(test latency baseline latency)/(cutoff time baseline latency)] 100%. The cumulative antinociceptive effect during the whole observation period was determined as the area under the curve for the %MPE (0 120 min) and these values were used for dose response curves (Yoshikawa et al., 2007). Results are expressed as the mean±s.e.m. for six animals per group. Comparisons between two experimental groups were done by Student's t test, while comparisons among several drug treatments were done by a one-way analysis of variance followed by Dunnett's test. In order to compare the time-course of the nociceptive effect produced by morphine alone or in combination with fentanyl, we used a two-way analysis of variance for repeated measures (factors: time and treatment) followed by Tukey test. The ED 50 (i.e., the dose that caused 50% of maximum antinociception) and associated 95% confidence intervals were generated from standard non-linear regression analysis of the log dose response curves (Prism 4.0, Graphpad Software, San Diego, CA, USA). Morphine and fentanyl interaction was evaluated by isobolographic analysis (Tallarida, 2001) using dose response curves in which the effects were expressed as the area under the curve of the time course for antinociception (Déciga-Campos et al., 2003; Granados-Soto and Argüelles, 2005; Yoshikawa et al., 2007). This approach was followed because both maximal responses and duration of analgesic effects were important for this study. The ED 50 values of morphine and fentanyl were placed on the x- and y-axes, respectively, and the two points were connected with a line representing the theoretical additive effect of both drugs. According to this analysis, when the experimentally determined data points and their confidence interval lie along this line, the drug effects are additive (no interaction); when the points lie below this line, they are superadditive (synergy), and when they lie above this line, they are antagonistic. We administered fixed-ratio combinations of opioids (i.c.v.) in 1:3, 1:1 and 3:1 ratios of their respective ED 50 fractions and the theoretical additive ED 50 value was then compared with the experimental ED 50 of the combination to determine whether there was a statistically significant difference. The degree of synergism was calculated by the isobolar relation (Tallarida, 2001, 2002): γ=a/a+b/b, where A and B are the doses of drug A (fentanyl alone) and B (morphine alone), respectively, that give the specified effect and (a,b) are the combination doses that produce the same effect. The quantities in the equation were obtained from the dose response curves of drugs A, B, and their combinations. If γ=1, the interaction was additive; if γb1, it was super-additive (synergy); and if γ>1, it was subadditive (antagonism). The program used for Statistics was SigmaStat (version 2.03, Jandel Scientific). 3. Results 3.1. Dose response curves Fig. 1 shows the time course of antinociception for increasing i.c.v. doses of morphine and fentanyl, and their corresponding dose response curves. With few exceptions, peak effects were achieved 15 min after administration of both drugs. In general, high doses of morphine produced peak antinociceptive effects that lasted up to 30 min and declined rapidly thereafter, while fentanyl achieved peak levels at approximately 15 min and they gradually decreased towards control levels. As expected, fentanyl was similarly effective, but more potent than morphine. Thus, morphine produced a maximal antinociceptive effect of 7887± 353 area units and fentanyl of 6878±739 area units. The calculated ED 50 was 1.5 μg (CL ) for morphine and μg (CL ) for fentanyl. Once the type of interaction between fentanyl and morphine was determined, we selected an effective morphine dose (1 μg) to be used in combination with a low dose of fentanyl (1 ng) for internalization studies (Section 3.3). This drug proportion was chosen based on reports from the literature showing that a low dose of fentanyl or methadone (another β-arrestin biased ligand), increased morphine antinociceptive effects (Hashimoto et al., 2006; He and Whistler, 2005) Isobologram The results of the isobolographic analysis are depicted in Fig. 2. The oblique line between the x- and y-axes is the additive line, and the points in the middle are the theoretical additive points calculated from the 1:3, 1:1 and 3:1 combination of separate ED 50 values. Table 1 shows the additive and experimental ED 50 values and γ values. These data indicate a synergistic interaction between morphine and fentanyl because the experimental points lie far below the additive line and the interaction index values are b1 for the three examined fixed-ratio combinations Combined actions of an effective morphine dose and a low dose of fentanyl Nociception Once the type of pharmacological interaction between these analgesics was determined, we studied whether the combination of an effective dose of morphine with a low, almost sub-effective fentanyl dose could also yield synergistic effects. The results are shown in Fig. 3. Fentanyl (1 ng) produced mild antinociception during the first 15 min after administration and no effect thereafter; this pattern

4 242 A. Silva-Moreno et al. / European Journal of Pharmacology 674 (2012) A C B Fig. 1. Time courses (A and B) and dose response curves (C) for the antinociceptive actions of morphine and fentanyl in the tail-flick test. Control animals were injected with saline solution. White symbols in the dose response curve indicate the doses used in chronic combination treatments. Each point represents the mean±s.e.m. of six animals. was similar in animals that received one or five fentanyl administrations (%MPE: 16.2±4.4% and 14.6±4.8%, respectively; n.s., Student's t test). A single morphine dose (1 μg) produced a 65.1±11% increase in the latency to tail-flick that peaked at 15 min, remained stable for 30 min and decreased afterwards. The fifth morphine administration was not effective i.e., at this time, animals were already tolerant. The combination of 1 μg morphine plus 1 ng fentanyl produced a peak effect of 89.1±3% at 45 min and this effect lasted longer than the one produced by morphine alone in acutely- and chronicallytreated animals. Although the efficacy of this morphine fentanyl combination diminished by the fifth administration (peak effect: 72.3± 9.8% at 30 min), it was still significantly higher when compared with control or morphine-treated rats. Morphine fentanyl coadministration was particularly effective after one hour of drug administration, when morphine was no longer effective on its own (panels A and C). Differences between the overall effect of antinociceptive treatments were also evident when the effects were expressed as area under the curve values (panels B and D) Naloxone-precipitated withdrawal Naloxone (31 μg) did not elicit abstinence signs in animals injected five times with a low dose of fentanyl (1 ng), morphine (1 μg) or morphine plus fentanyl (data not shown) Mu opioid receptor internalization and β-arrestin 2 activity In order to correlate antinociceptive effects with changes in μ-opioid receptor localization and density and β-arrestin 2 activity in periaqueductal gray and locus coeruleus, immunofluorescence studies were performed in animals treated once or five times with morphine and fentanyl, alone or in combination. Fig. 4 (upper panels) shows representative images from neurons stained with antibodies that recognize μ opioid receptors (red) and β-arrestin 2 in its active conformation (green) in animals that received one or five doses of each treatment. Active β-arrestin 2 immunoreactivity was barely detectable in control animals and in those that received one morphine dose. A significant increase in β-arrestin 2 activation was seen with the morphine fentanyl combination in comparison with the other groups both in acutelyand chronically-treated animals (Fig. 4, lower panels). No significant changes in total μ-opioid receptor immunoreactivity were observed under any circumstances. Mu opioid receptors were located on the plasma membrane in cells of animals that received one injection of saline solution or morphine, while their distribution was more diffuse when morphine and fentanyl were coadministered (panel A and B, second column). This combination was associated not only with an Table 1 Effect of fentanyl, morphine and their fixed-ratio combinations in the rat tail-flick test. Fig. 2. Isobolograms for the i.c.v. coadministration of fentanyl and morphine in the tailflick test in rats. Three fixed-ratio combinations of fentanyl:morphine were used: 1:3, 1:1 and 3:1 ratios of their respective ED 50 ratios. Points located on the additive line correspond to the theoretical additive ED 50 from each dose response curve considering that the observed effect was the sum of the individual antinociception produced by each component. Points below the line are the experimentally obtained ED 50 values for each combination. Each point represents the mean±s.e.m. of six animals. Fentanyl morphine combinations Experimental ED 50 ± S.E.M (μg, i.c.v.) Theoretical additive ED 50 ± S.E.M (μg, i.c.v.) γ (interaction index) 1: ±0.04 a 1.13 ± ±0.08 1: ±0.05 a 0.77 ± ±0.09 3: ±0.02 a 0.4± ±0.07 a Pb0.05 vs. the respective additive group; Student's t test.

5 A. Silva-Moreno et al. / European Journal of Pharmacology 674 (2012) A B C D Fig. 3. The acute or chronic coadministration of morphine and fentanyl (M+Fen) induces higher antinociception in the tail-flick test than morphine alone. Panels A and C show antinociception observed after a single or repeated drug administration. Panels B and D show the antinociception calculated as area under curve (AUC) from data shown in A and C. Data are presented as the mean ±S.E.M. + Pb0.05, +++ Pb0.001 vs. morphine alone, Tukey test; ***Pb0.05 vs. control, Dunnett's test; ### Pb0.01; Student's t test. increase in β-arrestin 2 immunoreactivity, but also with μ-opioid receptor and β-arrestin 2 colocalization (yellow). Fig. 5 shows the results obtained in locus coeruleus. The same pattern of changes observed in the periaqueductal gray was found in this brain area; i.e., there was an increase in β-arrestin 2 activity and μ-opioid receptor internalization with the morphine fentanyl combination in acutely- and chronically-treated animals. Fig. 6 analyzes the effect of different analgesic treatments on μ-opioid receptor and β-arrestin 2 signal redistribution in periaqueductal gray neurons. The quantitative fluorescence intensity vs. distance profiles was scanned along the straight arrow indicated in confocal images. Control group showed fluorescence principally in the plasma membrane. This pattern was also observed in animals treated with morphine or fentanyl alone. No differences in μ-opioid receptor subcellular localization were observed between acute and repeatedly treated animals. When the morphine-fentanyl combination was administered, fluorescence was not only found in the plasma membrane, but also in the cytoplasm, indicating μ-opioid receptor endocytosis. Similar results were seen in locus coeruleus (data not shown). 4. Discussion Our results show that the combined treatment of morphine with fentanyl results in long-lasting synergistic antinociceptive effects aftersingleorrepeatedadministration, and that these actions are positively correlated with an increase in μ-opioid receptor endocytosis and β-arrestin 2 activity in the periaqueductal gray matter and locus coeruleus. No differences in the time to reach peak effects and duration of antinociception produced by each separate drug were found in the present study probably due to the use of the i.c.v. route which precludes several pharmacokinetic processes. Using this administration route, morphine and fentanyl showed similar efficacy, but different antinociceptive potency. These results reproduce what has been seen in clinical practice and preclinical experiments with other administration routes where fentanyl has proven to be consistently, and significantly (up to a hundred times) more potent than morphine (Jeal and Benfield, 1997; Meert and Vermeirsch, 2005; Romero et al., 2010; Trescot et al., 2008). The antinociception produced by morphine fentanyl combination has been analyzed before, but only after a single injection and under different experimental conditions. Romero and coworkers found additive effects in the hot plate test when effective doses of morphine and fentanyl were coadministered to mice (Romero et al., 2010). Their study differs from ours in the use of different animal species, the response under study paw lick is a supra-spinal response (Langerman et al., 1995) and tail-flick is a spinal reflex, the administration route (s.c. vs. i.c.v.), and the drug proportions used in combination (effective vs. sub-effective doses of fentanyl). Also, Romero's study measured peak responses, whereas we analyzed the cumulative effects during the whole observation period (2 h), which allowed antinociception evaluation at late times at which synergism became more evident (Fig. 3). In another study, Bolan and coworkers reported only additive actions in the tail-flick test in mice treated with a combination of s.c. morphine plus fentanyl. They reached this conclusion using a quantal approach to assess antinociception (i.e., the percentage of animals that responded to a fixed thermal stimulus) and effective

6 244 A. Silva-Moreno et al. / European Journal of Pharmacology 674 (2012) Fig. 4. Coadministration of morphine and fentanyl (M+Fen) increases μ-opioid receptor (red) and β-arrestin 2 (green) colocalization (yellow) in periaqueductal gray. Rats received one (panel A) or five (panel B) i.c.v. administrations of saline (Control), 1 μg morphine (M), 1 ng fentanyl (Fen) or the combination of M+Fen. A representative neuron obtained from animals with each treatment is shown in the upper panels. Lower graphs show the pixels of specific fluorescence on the surface of at least 30 neurons per treatment. Data are presented as the mean±s.e.m. *Pb0.05, ***Pb0.001 vs. control; Dunnett's test. doses of both compounds in combination (Bolan et al., 2002). Variations among results are not surprising because it has been reported that analgesic combination effects can change with the ratio of the two components of the drug pair, the level of effect, the site of injection, and the nociception test (Le Bars et al., 2001; Romero et al., 2010; Tallarida, 2001). Nonetheless, our study shows that synergistic effects between morphine and fentanyl can be observed within a wide dose range providing that long-lasting nociception is being evaluated. Another significant finding was that coadministration of morphine and fentanyl produced analgesia under circumstances where morphine was no longer effective. Moreover, in spite of the enhanced antinociception observed in chronically treated animals we did not observe significant withdrawal signs after a high naloxone challenge in any of the groups studied. This suggests that no significant physical dependence occurred under our experimental conditions; i.e., the synergism found with morphine and fentanyl did not include this adverse effect, at least after five administrations. It is worth noting that although physical dependence is frequently intertwined with tolerance, each can occur independently (Aceto, 1990; Smith and Picker, 1998). In our laboratory we have found tolerance, but not dependence to the repeated administration of a morphine dipyrone combination (Hernández-Delgadillo et al., 2003). With the combination used in the present study we cannot discount the possibility that dependence would occur with more prolonged treatments, but even in that case, the delay in physical dependence development would constitute a potential advantage of using this combination for pain treatment. It has been considered that coadministration of two μ-agonists is an empirical approach without sound theoretical ground because if drugs act similarly, one could expect to see, at best, additive effects (Barrera et al., 2005). Nonetheless, the analgesic synergy found in the present study between morphine and fentanyl could be due to a mixture of mechanisms with subtle, but physiologically relevant differences. Mu opioid agonists differ in their capacity to induce receptor signaling and trafficking. In particular, morphine has been described as

7 A. Silva-Moreno et al. / European Journal of Pharmacology 674 (2012) Fig. 5. Effects of acute and chronic coadministration of morphine and fentanyl (M+Fen) on μ-opioid receptor immunoreactivity (red), β-arrestin 2 activity (green) and μ-opioid receptor and β-arrestin 2 colocalization (yellow) in locus coeruleus. Rats received one (panel A) or five (panel B) i.c.v. administrations of saline (Control), 1 μg morphine (M), 1 ng fentanyl (Fen) or the combination of M+Fen. A representative neuron obtained from animals with each treatment is shown in the upper panels. Lower graphs show the pixels of specific fluorescence on the surface of at least 30 neurons per treatment. Data are presented as the mean±s.e.m. *P b0.05, ***Pb0.001 vs. control; Dunnett's test. a G-protein biased ligand, whereas fentanyl seems to preferentially act in a β-arrestin 2 dependent manner (Rajagopal et al., 2010). Both morphine and fentanyl produce receptor desensitization (Virk and Williams, 2008), but morphine provokes considerably less receptor internalization than fentanyl (Finn and Whistler, 2001; He et al., 2002; Johnson et al., 2005; Koch et al., 2005) and both ligands differ in receptor desensitization. Several authors have shown that morphine's ability to induce μ-opioid receptor internalization can be improved when morphine is combined with another opioid agonist with high internalization capacity. Hashimoto and colleagues studied the antinociception produced by acute i.t. administration of 2.5 μg morphine and 0.5 μg fentanyl alone or in combination, and correlated this effect with μ-opioid receptor internalization in the spinal dorsal horn neurons of rats. At the doses tested, morphine and fentanyl on their own were not effective in the hot plate test, but together they produced significant antinociception. The authors suggested that this was due to an increase in μ-opioid receptor endocytosis (Hashimoto et al., 2006). Our results agree with these data because we also found increased antinociception and enhanced μ-opioid receptor internalization with acute coadministration of morphine and fentanyl; moreover, present results extend these findings to two supraspinal areas (periaqueductal gray and locus coeruleus) that are central in pain regulation (Millan, 2002), tolerance and withdrawal (Scavone and Van Bockstaele, 2009; Williams et al., 2001). The fact that receptor endocytosis is positively correlated with increased antinociception in different anatomical sites suggests that the mechanisms regulating μ-opioid receptor endocytosis are similar at different levels of nociceptive pathways.

8 246 A. Silva-Moreno et al. / European Journal of Pharmacology 674 (2012) Fig. 6. The combined administration of morphine and fentanyl (M+Fen) increases μ-opioid receptor endocytosis in periaqueductal gray sections. Animals were subjected to one or five i.c.v. administrations of saline (C), 1 μg morphine (M), 1 ng fentanyl (Fen) or the combination of M+Fen at the same doses. Intensity of the μ-opioid receptor and β-arrestin 2 signal was analyzed along of a line of 10 μm starting at the plasma membrane of each cell and crossing through the intracellular space. Distribution of specific red (μ-opioid receptor) or green (β-arrestin 2) signal was obtained from the depicted cell and placed in the lower part of each photograph. Images show amplifications of a representative cell analyzed in each treatment. Among the few researchers that have addressed the chronic effects of two μ-opioid receptor agonists in combination, He and Whistler (2005) assessed nociception in the tail-flick test in rats after repeated i.c.v. administration (twice a day for five days) of an effective morphine dose and a low dl-methadone dose, and found reduced tolerance and dependence associated with enhanced endocytosis of μ-opioid receptors. They proposed that a small number of methadone-activated receptors facilitated endocytosis of morphine-activated receptors because they formed dimers/oligomers. Because fentanyl, like methadone, is a biased β-arrestin ligand (Rajagopal et al., 2010), it is reasonable to suppose that fentanylactivated receptors could promote endocytosis of morphinestimulated receptors even after repeated administration. The fact that β-arrestin 2 activation increased with repeated administration of our analgesic combination supports this idea and suggests that β-arrestin 2 dependent signaling pathways remain active for a long time without showing substantial desensitization. Our results also agree with other authors who have found that μ-opioid receptor endocytosis prevents tolerance development in simplified cell culture model systems (Koch et al., 2005) and other experimental preparations (reviewed in von Zastrow, 2010). As previously mentioned, morphine and fentanyl differ not only in signaling preferences, but also in efficacy (Sirohi et al., 2008; Smith, 2008) and affinity for splice variants of the μ-opioid receptor gene (Pasternak, 2004). Our results can be discussed in the light of different recent findings in the field. The antibody used in this work recognizes μ-opioid receptor isoforms with the common C terminal sequence codified by exon 4 of the gene. With the exception of mmor-1, those variants also have the N terminal amino acid sequence codified by exon 11 (Pan et al., 2009; Xu et al., 2011). Interestingly, in exon 11 knockout mice, fentanyl loses its antinociceptive efficacy to a much higher extent than morphine (Pan et al., 2009). This suggests that more than one receptor isoform could be involved in the morphine fentanyl combination effects observed in our study. In conclusion, the analgesic combination of morphine plus fentanyl provides advantages over single drug treatments because synergism occurs for antinociception, but not for physical dependence; moreover, less tolerance develops after repeated administration. The mechanisms behind these actions could involve selective activation of β-arrestin 2 dependent signaling pathways in anatomical structures implicated in pain processing. Due to different profiles in pharmacological actions, this analgesic combination could also be also used as an experimental tool for the study of differential μ-opioid receptor activation and transduction. Acknowledgements This paper includes data from Arturo Silva-Moreno's doctoral dissertation and was supported by scholarship from Conacyt. This work was partially funded by a grant to MLO from the INPRFM (project No. NC ). Authors wish to thank Dr. Vinicio Granados-Soto for valuable comments to the manuscript and Mr. Abraham Contreras and Isaí Méndez for technical assistance. References Abbadie, C., Pan, Y.X., Pasternak, G.W., Differential distribution in rat brain of mu opioid receptor carboxy terminal splice variants MOR-1C-like and MOR-1-like immunoreactivity: evidence for region-specific processing. J. Comp. Neurol. 419, Aceto, M.D., Assessment of physical dependence techniques for the evaluation of abused drugs. In: Adler, M.W., Cowan, A. (Eds.), Testing and Evaluation of Drugs of Abuse. Modern Methods in Pharmacology. Wiley-Liss, Inc., New York, NY, pp Barrera, N.P., Morales, B., Torres, S., Villalón, M., Principles: mechanisms and modeling of synergism in cellular responses. Trends Pharmacol. Sci. 26, Beaulieu, J.M., Morphine-induced μ-opioid receptor internalization: a paradox solved in neurons. J. Neurosci. 25, Bohn, L.M., Dykstra, L.A., Lefkowitz, R.J., Caron, M.G., Barak, L.S., Relative opioid efficacy is determined by the complements of the G protein-coupled receptor desensitization machinery. Mol. Pharmacol. 66,

9 A. Silva-Moreno et al. / European Journal of Pharmacology 674 (2012) Bolan, E.A., Tallarida, R.J., Pasternak, G.W., Synergy between mu opioid ligands: evidence for functional interactions among mu opioid receptor subtypes. J. Pharmacol. Exp. Ther. 303, Déciga-Campos, M., López, U.G., Reval, M.I., López-Muñoz, F.J., Enhancement of antinociception by co-administration of an opioid drug (morphine) and preferential cyclooxygenase-2 inhibitor (rofecoxib) in rats. Eur. J. Pharmacol. 460, Dottrens, M., Rifat, K., Morel, D.R., Comparison of extradural administration of sufentanil, morphine and sufentanil morphine combination after caesarean section. Br. J. Anaesth. 69, Finn, A.K., Whistler, J.L., Endocytosis of the mu opioid receptor reduces tolerance and a cellular hallmark of opiate withdrawal. Neuron 32, Friedman, Z., Katznelson, R., Phillips, S.R., Zanchetta, C., Nistor, O.I., Elsen, L.B., Siddiqui, N., A randomized double-blind comparison of a morphine fentanyl combination vs. morphine alone for patient-controlled analgesia following bowel surgery. Pain Pract. 8, Goldstein, F.J., Adjuncts to opioid therapy. J. Am. Osteopath. Assoc. 102, S15 S21. Granados-Soto, V., Argüelles, C.F., Synergic antinociceptive interaction between tramadol and gabapentin after local, spinal and systemic administration. Pharmacology 74, Hashimoto, T., Saito, Y., Yamada, K., Hara, N., Kirihara, Y., Tsuchiya, M., Enhancement of morphine analgesic effect with induction of mu-opioid receptor endocytosis in rats. Anesthesiology 105, He, L., Whistler, J.L., An opiate cocktail that reduces morphine tolerance and dependence. Curr. Biol. 15, He, L., Fong, J., von Zastrow, M., Whistler, J.L., Regulation of opioid receptor trafficking and morphine tolerance by receptor oligomerization. Cell 108, Hernández-Delgadillo, G.P., López-Muñoz, F.J., Salazar, L.A., Cruz, S.L., Morphine and dipyrone co-administration delays tolerance development and potentiates antinociception. Eur. J. Pharmacol. 469, Jeal, W., Benfield, P., Transdermal fentanyl. A review of its pharmacological properties and therapeutic efficacy in pain control. Drugs 53, Johnson, E.E., Christie, M.J., Connor, M., The role of opioid receptor phosphorylation and trafficking in adaptations to persistent opioid treatment. Neurosignals 14, Koch, T., Widera, A., Bartzsch, K., Schulz, S., Brandenburg, L.O., Wundrack, N., Beyer, A., Grecksch, G., Hollt, V., Receptor endocytosis counteracts the development of opioid tolerance. Mol. Pharmacol. 67, Langerman, L., Zakowski, M.I., Piskoun, B., Grant, G.J., Hot plate versus tail flick: evaluation of acute tolerance to continuous morphine infusion in the rat model. J. Pharmacol. Toxicol. Methods 34, Le Bars, D., Gozariu, M., Cadden, S.W., Critical analysis of animal models of acute pain. II. Ann. Fr. Anesth. Reanim. 20, Luttrell, L.M., Gesty-Palmer, D., Beyond desensitization: physiological relevance of arrestin-dependent signaling. Pharmacol. Rev. 62, Meert, T.F., Vermeirsch, H.A., A preclinical comparison between different opioids: antinociceptive versus adverse effects. Pharmacol. Biochem. Behav. 80, Mercadante, S., Villari, P., Ferrera, P., Casuccio, A., Addition of a second opioid may improve opioid response in cancer pain: preliminary data. Support. Care Cancer 12, Millan, M.J., Descending control of pain. Prog. Neurobiol. 66, Nance, P.W., Sawynok, J., Substance P-induced long-term blockade of spinal adrenergic analgesia: reversal by morphine and naloxone. J. Pharmacol. Exp. Ther. 240, Pan, Y.X., Xu, J., Xu, M., Rossi, G.C., Matulonis, J.E., Pasternak, G.W., Involvement of exon 11-associated variants of the mu opioid receptor MOR-1 in heroin, but not morphine, actions. Proc. Natl. Acad. Sci. 106, Pasternak, G.W., Multiple opiate receptors: déjà vu all over again. Neuropharmacology 47 (Suppl. 1), Paxinos, G., Watson, C., The Rat Brain in Stereotaxic Coordinates. Academic Press, New York, NY. Rajagopal, S., Rajagopal, K., Lefkowitz, R.J., Teaching old receptors new tricks: biasing seven-transmembrane receptors. Nat. Rev. Drug Discov. 9, Romero, A., Miranda, H.F., Puig, M.M., Analysis of the opioid-opioid combinations according to the nociceptive stimulus in mice. Pharmacol. Res. 61, Scavone, J.L., Van Bockstaele, E.J., μ-opioid receptor redistribution in the locus coeruleus upon precipitation of withdrawal in opiate-dependent rats. Anat. Rec. (Hoboken) 292, Silva-Moreno, A., López-Muñoz, F.J., Cruz, S.L., D-propoxyphene and dipyrone coadministration produces greater antinociception and fewer adverse effects than single treatments in rats. Eur. J. Pharmacol. 607, Sirohi, S., Dighe, S.V., Walker, E.A., Yoburn, B.C., The analgesic efficacy of fentanyl: relationship to tolerance and mu-opioid receptor regulation. Pharmacol. Biochem. Behav. 91, Smith, H.S., Combination opioid analgesics. Pain Physician 11, Smith, M.A., Picker, M.J., Tolerance and cross-tolerance to the rate-suppressing effects of opioids in butopharnol-treated rats: influence of maintenance dose and relative efficacy at the mu receptor. Psychopharmacology (Berl) 140, Tallarida, R.J., Drug synergism: its detection and applications. J. Pharmacol. Exp. Ther. 298, Tallarida, R.J., The interaction index: a measure of drug synergism. Pain 98, Trescot, A.M., Datta, S., Lee, M., Hansen, H., Opioid pharmacology. Pain Physician 11, S133 S153. Virk, M.S., Williams, J.T., Agonist-specific regulation of μ-opioid receptor desensitization and recovery from desensitization. Mol. Pharmacol. 2008, von Zastrow, M., Regulation of opioid receptors by endocytic membrane traffic: mechanisms and translational implications. Drug Alcohol Depend. 108, Waldhoer, M., Bartlett, S.E., Whistler, J.L., Opioid receptors. Annu. Rev. Biochem. 73, Walwyn, W.M., Miotto, K.A., Evans, C.J., Opioid pharmaceuticals and addiction: the issues, and research directions seeking solutions. Drug Alcohol Depend. 108, Williams, J.T., Christie, M.J., Manzoni, O., Cellular and synaptic adaptations mediating opioid dependence. Physiol. Rev. 81, Xu, J., Xu, M., Rossi, G.C., Pasternak, G.W., Pan, Y.X., Identification and characterization of seven new exon 11-associated splice variants of the rat mu opioid receptor gene, OPRM1. Mol. Pain 7, 9. Yoshikawa, M., Ito, K., Maeda, M., Akahori, K., Takahashi, S., Jin, X.L., Matsuda, M., Suzuki, T., Oka, T., Kobayashi, H., Hashimoto, A., Activation of supraspinal NMDA receptors by both D-serine alone or in combination with morphine leads to the potentiation of antinociception in tail-flick test of rats. Eur. J. Pharmacol. 565, Zheng, H., Loh, H.H., Law, P.Y., β-arrestin-dependent μ opioid receptor-activated extracellular signal-regulated kinases (ERKs) translocate to nucleus in contrast to G protein-dependent ERK activation. Mol. Pharmacol. 73, Zheng, H., Loh, H.H., Law, P.Y., 2010a. Agonist-selective signaling of G protein-coupled receptor: mechanisms and implications. IUBMB Life 62, Zheng, H., Zeng, Y., Zhang, X., Chu, J., Horace, H., Law, P.Y., 2010b. mu-opioid receptor agonists differentially regulate the expression of mir-190 and NeuroD. Mol. Pharmacol. 77, Zimmermann, M., Ethical guidelines for investigations of experimental pain in conscious animals. Pain 16,