Evidence suggests that N-methyl-d-aspartate (NMDA) glutamate receptors may be involved in the plasticity that arises from long-term administration of morphine (Herman et al. 1995; Inturrisi 1997; Mao 1999; Trujillo 1995, 2000, 2002). Initial evidence supporting this idea was provided by Trujillo and Akil (1991) who reported that the NMDA receptor antagonist MK-801 inhibited the development of tolerance to the antinociceptive effect of morphine and morphine physical dependence without affecting acute morphine antinociception. After this discovery, numerous laboratories replicated these findings, demonstrating that a variety of NMDA receptor antagonists have the ability to inhibit the development of opiate tolerance and dependence (for review, see Bisaga and Popik 2000; Herman et al. 1995; Inturrisi 1997; Mao 1999; Mayer et al. 1999; Trujillo 2000, 2002; and Trujillo and Akil 1995).

Wolf and Jeziorski (1993) provided the first evidence that NMDA receptor antagonists inhibit sensitization to the locomotor stimulant effect of morphine in a manner similar to their effect on tolerance and dependence, a finding that has been supported by others (Jeziorski et al. 1994; Kosten and Bombace 2000). Although these results suggest a role for NMDA receptors in opiate sensitization, there is controversy regarding this idea (Carlezon et al. 2000; Iijima et al. 1996; Jeziorski et al. 1994; Ranaldi et al. 2000; Tzschentke and Schmidt 1996, 1998). Some research has failed to replicate an ability of NMDA receptor antagonists to inhibit the development of sensitization, whereas other research has found inhibition but only at high doses of NMDA receptor antagonists that produce adverse side effects (Iijima et al. 1996; Tzschentke and Schmidt 1996, 1998; Tzschentke and Schmidt 2000; Vanderschuren et al. 1997). An additional controversy is over the selection of NMDA receptor antagonists used to explore the role of NMDA receptors in sensitization to the stimulant effect of morphine. Most studies on this topic have used the potent and selective antagonist MK-801 as the drug of choice. However, MK-801 has been criticized as problematic because of its behavioral effects alone and in combination with morphine (Carlezon et al. 2000; Ranaldi et al. 2000; Tzschentke and Schmidt 1998; Tzschentke and Schmidt 2000). There is, therefore, a need to further explore the potential role of NMDA receptors in opiate sensitization using alternative NMDA receptor antagonists.

Another issue of importance is whether or not NMDA receptor antagonists have the ability to affect tolerance or sensitization to different opiates. Although these antagonists can inhibit the development of tolerance resulting from chronic morphine administration, inhibition of tolerance resulting from other opiates that act at mu opioid receptors has not been consistently reported. For example, Bilsky et al. (1996) reported that NMDA receptor antagonists do not inhibit the development of antinociceptive tolerance to mu opioid agonists other than morphine. On the other hand, Mao et al. (1998) and Allen and Dykstra (2000a) have reported that NMDA receptor antagonists can indeed inhibit tolerance to different mu opioids. Because of these conflicting findings and a relative lack of research, there is a need to further explore the effects of NMDA receptor antagonists on the development of tolerance and sensitization to opiates other than morphine.

The purpose of the current studies is to address some of the controversies regarding the role of NMDA receptors in tolerance and sensitization to opiates. To date, the majority of studies that have investigated sensitization to the locomotor stimulant effect of opiates have used MK-801 as the NMDA receptor antagonist and morphine as the opioid. The present studies replicated earlier research with MK-801 but extended this work to other NMDA receptor antagonists, including memantine and LY235959. Memantine is a clinically available noncompetitive antagonist, whereas LY235959 is a competitive antagonist, and each produces a behavioral profile distinct from MK-801 (Bubser et al. 1992; Danysz et al. 1994; Danysz and Parsons 1998; Geter-Douglass and Witkin 1997, 1999; Parsons et al. 1999). In addition to examining different NMDA receptor antagonists, the present experiments utilized three different opiates: morphine, methadone, and buprenorphine. Each of these opiates is used clinically; however, they differ in affinity, selectivity, and efficacy for mu opioid receptors. The present studies are the first to examine the ability of memantine or LY235959 to affect sensitization to the stimulant effects of opiates and the first to examine the effects of NMDA receptor antagonists on tolerance and sensitization to methadone or buprenorphine.

The general working hypothesis of this research is that NMDA receptors are involved in opiate-induced neural and behavioral plasticity. Given this hypothesis, it is predicted that low selective doses of the NMDA receptor antagonists will inhibit the development of tolerance and sensitization to morphine, methadone, and buprenorphine. If similar effects are obtained across the different combinations of NMDA receptor antagonists and opioids, this will support the idea that NMDA receptors are involved in both tolerance to the antinociceptive effects and sensitization to the stimulant effects of opiates.

Six different experiments were performed in three sets (Table 1). The first set of experiments examined the effects of MK-801 or memantine on sensitization to the stimulant effect of morphine. These experiments replicated the ability of MK-801 to inhibit morphine sensitization and extended these findings to memantine. The second set of experiments examined the effects of MK-801 or LY235959 on both tolerance and sensitization to morphine in the same experimental animals. Similar findings across these two independent behaviors in the same experimental animals have the opportunity to strengthen the idea that the effects are caused by the inhibition of plasticity rather than a pharmacological interaction unique to a particular behavior. The third set of experiments examined the effects of MK-801 on both tolerance and sensitization to methadone and buprenorphine, thereby, extending the research to different mu opioids. Together, the findings support the idea that NMDA receptors are involved in opiate tolerance and sensitization.

Table 1 Summary of experiments

Materials and methods

Animals and groups

One hundred thirty-one experimentally naive adult male Sprague–Dawley rats (200–225 g at time of purchase; Harlan, San Diego, CA, USA) were used in these experiments. The animals were housed in plastic cages, three rats per cage on a 12-h light/dark cycle, with food and water available ad libitum, and allowed at least 1 week of acclimation before use in the experiments. Experimental protocols were approved by the California State University San Marcos Institutional Animal Care and Use Committee and are in compliance with the National Institutes of Health Guide for Care and Use of Laboratory Animals.

Instruments

An IITC model 33 tail-flick apparatus was used to measure antinociception. Methods were similar to those used previously (Redwine and Trujillo 2003; Trujillo and Akil 1991, 1994). Rats were gently restrained by hand for tail flick tests. Three tail flicks in a row, 10–20 s apart, were performed for each animal, with the average used for data analysis. The heat was focused between 2 and 8 cm from the tip of the tail, with the focus of the heat moved on each trial to avoid irritation caused by repeated testing at the same site. The amount of time an animal took to move its tail from the heat was determined. A heat setting that yielded a 2–4 s baseline response was used. A cutoff latency of 8 s was used to prevent tissue damage.

Locomotor stimulation was automatically assessed using a Kinder Scientific open-field motor monitor system. This system consists of eight 16″ × 15″ × 16″ plexiglas enclosures with two 16″ × 16″ photocell arrays surrounding each chamber. The first array is placed 5 cm above the floor and is used to measure horizontal activity; the second is 12.5 cm above the floor and measures vertical (rearing) activity. An automated personal computer collects data on photocell beam interruptions from both arrays and calculates distance traveled, time at rest, number of rears, ambulations (interruptions of successive photocell beams resulting from displacement of the animal horizontally, typically associated with forward movement), and fine movements (interruptions of single photocell beams, typically associated with short repetitive movements).

Drugs

Three NMDA receptor antagonists were used in these studies. The antagonists included the non-competitive antagonists MK-801 (0.1 mg/kg) and memantine (3.0 mg/kg) and the competitive antagonist LY235959 (1.0 mg/kg). Doses chosen for these studies were based on previous research demonstrating that they effectively block NMDA receptors but produce minimal behavioral side-effects (Allen and Dykstra 1999; Bilsky et al. 1996; Bubser et al. 1992; Carter 1994; Danysz et al. 1994; Popik and Kozela 1999; Popik et al. 2000; Redwine and Trujillo 2003; Trujillo et al. 2001a; Trujillo et al. 2000; Trujillo et al. 2001b). Memantine hydrochloride and MK-801 hydrogen maleate were purchased from Sigma, and LY235959 was purchased from Tocris. The opiates used were morphine (5.0 mg/kg), methadone (3.0 mg/kg), and buprenorphine (0.3 or 1.0 mg/kg). Morphine sulfate and buprenorphine hydrochloride were generous gifts from the National Institute on Drug Abuse Drug Supply Program, whereas methadone hydrochloride was purchased from Sigma. All drugs in this study were dissolved in 0.9% saline, except for buprenorphine, which was dissolved in 1% Tween-80; drugs were injected at a volume of 1.0 ml/kg

Protocol

Groups were assigned an initial injection of saline or an NMDA antagonist followed by a second injection of saline or an opiate. The experimental design for each study resulted in four treatment groups for each of six experiments (Table 2). Six animals were included in each group, except where otherwise noted. The comparison of interest is between the opiate group and the experimental group. The antagonist control group is also significant as a measure of any effects the antagonists alone had on the selected behavior.

Table 2 Treatment groups for experimental design

Three sets of experiments were performed. The first (experiments 1.1 and 1.2) examined the effects of MK-801 and memantine on sensitization to the locomotor stimulant effects of morphine. The second set of studies (experiments 2.1 and 2.2) expanded on these findings, examining the effects of MK-801 and LY235959 on both tolerance to the antinociceptive effect and sensitization to the stimulant effect of morphine in the same experimental animals. Finally, to explore other opiates, the third set of studies (experiments 3.1 and 3.2) examined the effects of MK-801 on both tolerance to the antinociceptive effect and sensitization to the stimulant effect of methadone and buprenorphine (see Table 1).

The experimental approach was similar to that used in earlier research (Peterson et al. 2002; Swadley-Lewellen and Trujillo 1998; Trujillo 1995, 2000). For all experiments, animals received, in their home environment, once daily injections of saline or an NMDA receptor antagonist followed 30 min later by an injection of saline or an opiate. On the opiate challenge day, for testing of sensitization, the animals were habituated to the locomotor testing room for 30 min. In the testing room, white noise was used to mask outside sounds, and overhead incandescent lights were dimmed. After habituation to the testing room, the animals were placed in the locomotor apparatus for 30 min, then removed from the enclosure and injected with an opiate in the absence of pretreatment. Upon injection, the animals were immediately placed back in the enclosure, and locomotor activity was recorded for an additional 3.5 h. This approach was used to minimize the opportunity for associative factors to contribute to sensitization. Because repeated injections of the opiate occurred in the home environment, whereas sensitization was assessed in the testing room under unique conditions, association of drug effects with environmental cues was not likely a key factor in the results.

To examine the development of tolerance to the antinociceptive effects of the opiates, baseline tail-flick latencies were taken on the two consecutive days preceding the first day of treatment. Beginning on day 1 of treatment, tail-flick scores were determined 60 min after the second injection (vehicle or opiate) on odd numbered days. The tests during the course of chronic injections occurred in the home environment under the influence of the NMDA receptor antagonist. To further assess tolerance, on the final day of the experiment, during the test for sensitization, the animals were removed briefly 60 min after the opiate injection to obtain a tail-flick score (the time out of the enclosure for tail-flick testing was 2 min or less). This final tail-flick test occurred in the absence of the NMDA receptor antagonist. In this manner (testing for tolerance in both the presence and the absence of the NMDA receptor antagonist), we were able to determine if the effects of the NMDA receptor antagonists resulted from state-dependence (see Trujillo 2000 for further discussion).

The experimental treatment schedule for morphine experiments consisted of 10 days of injections, followed by an opiate challenge on day 11. Development of sensitization to the stimulant effects of methadone and buprenorphine, however, required a 16-day treatment schedule, with sensitization tested on day 17. For methadone, tolerance and sensitization were examined in the same experimental animals using this 17-day protocol. However, because of the different doses and time courses required for buprenorphine tolerance and sensitization, different experimental groups were used to assess tolerance and sensitization. Therefore, 1.0 mg/kg was used to assess the development of tolerance to the antinociceptive effect of buprenorphine over 9 days of treatment (similar to morphine), whereas 0.3 mg/kg was used to assess the development of sensitization to the stimulant effect over 17 days of treatment.

Data analysis

For analysis of tail-flick latencies, raw scores in seconds were determined and used as a measure of pain responsiveness. Effects over days were analyzed by two-way repeated measures analysis of variance (ANOVA; treatment group × day), followed by Fisher’s protected least significant difference (PLSD) post hoc analysis for comparison between individual groups. On test day, a one-way factorial ANOVA was used with Fisher’s PLSD for post hoc comparisons. For locomotor stimulation, activity was collected at 10-min intervals for a total of 210 min post-injection. Raw photo beam breaks (basic movements) were analyzed using a two-way repeated measures ANOVA (treatment group × time) to determine the overall effect of treatment, followed by Fisher’s PLSD post hoc analysis for individual group comparisons. In the present experiments, sensitization for morphine appeared most prominent during the first 60 min after injection, whereas sensitization for methadone and buprenorphine appeared most prominent during the first 120 min post-injection. Therefore, secondary statistical analyses focused on these respective time frames, with a one-way factorial ANOVA examining group differences in the total activity scores during the first 60 min (morphine) or the first 120 min (methadone and buprenorphine) post-injection.

Results

  1. 1.

    Effects of MK-801 and memantine on sensitization to morphine

Experiment 1.1. Effects of the NMDA receptor antagonist MK-801 on sensitization to morphine-induced locomotor stimulation

Animals that received morphine for the first time on day 11 (Sal/Sal and MK/Sal groups) showed a modest level of activity immediately after the morphine injection and a slowly emerging increase in activity that was evident late in the session. In contrast, animals that received morphine throughout the study (Sal/Mor group) showed a potent increase in activity that was evident within 10 min after injection, consistent with the development of sensitization. The group that received MK-801 and morphine (MK/Mor group) showed effects very similar to those that received morphine for the first time (Fig. 1a). Statistical analysis confirmed differences among the groups, showing a significant effect of treatment [F(3,14) = 8.24, p < 0.01] and time [F(23,322) = 21.86, p < 0.01], as well as an interaction [F(69,322) = 3.42, p < 0.01]. Post hoc analysis of the total activity during the first 60 min post-injection revealed that the Sal/Mor group differed from all others, including the MK/Mor group (p < 0.01, Fig. 1b); the MK/Mor group did not differ from the control groups.

Fig. 1
figure 1

MK-801 and memantine each inhibit sensitization to morphine-induced locomotor stimulation. Animals were injected once daily for 10 days with saline, MK-801 (0.1 mg/kg), or memantine (3.0 mg/kg) followed 30 minutes later by saline or morphine (5.0 mg/kg). On day 11, animals were challenged with morphine (5.0 mg/kg), and the time course of activity was assessed. a Time course of response to morphine in the four treatment groups on day 11 in experiment 1.1 (MK-801). b Total activity during the first 60 min post-injection on day 11 in experiment 1.1 (MK-801). c Time course of response to morphine in the four treatment groups on day 11 in experiment 1.2 (memantine). d Total activity during the first 60 min post-injection on day 11 in experiment 1.2 (memantine). Note that, in each study, the Sal/Mor (S/M) group showed greater activity than all other groups (sensitization), whereas the MK/Mor and the Mem/Mor (M/M) groups did not differ from the control groups (sensitization did not develop in these groups). **p < 0.05, Sal/Mor group significantly different from all other treatment groups. N = 6 animals/group (except for Sal/Sal and MK/Sal groups in experiment 1.1; N = 3/group). Arrows identify the point at which animals were briefly removed from the enclosure for tail-flick testing. The burst in activity after the test was likely an interaction between the drug effects and the handling of the animals

Experiment 1.2: Effects of the NMDA receptor antagonist memantine on sensitization to morphine-induced locomotor stimulation

A pattern of results similar to experiment 1.1 was observed. Both the Sal/Sal group and the Mem/Sal group showed modest levels of activity immediately after the morphine injection, with a slowly emerging increase in activity late in the session. The Mem/Mor group showed effects similar to the groups that received morphine for the first time. As in experiment 1.1, the Sal/Mor group showed an increase in activity when compared to all other groups. Statistical analysis confirmed differences among the groups, with a significant effect of treatment [F(3,20) = 4.15, p < 0.05] and time [F(23,460) = 38.14, p < 0.01], as well as a treatment × time interaction [F(69,460) = 3.02, p < 0.01; Fig. 1c]. Analysis of the total activity during the first 60 min post-injection revealed that the Sal/Mor group was significantly more active than all other groups (p < 0.05); however, the Mem/Mor group did not differ from the control groups (Fig. 1d). Although the Mem/Sal group appeared less active than the Sal/Sal control group, this did not achieve statistical significance.

  1. 2.

    Effects of MK-801 and LY235959 on tolerance and sensitization to morphine in the same experimental animals

Experiment 2.1: Effects of the NMDA receptor antagonist MK-801 on tolerance to morphine-induced antinociception and sensitization to morphine-induced locomotor stimulation

Groups receiving Sal/Sal and MK/Sal showed constant tail-flick latencies at approximately 3–4 s throughout the 10 days of injections. Both groups that received morphine showed maximal antinociceptive effects early in treatment, with the Sal/Mor group showing decreasing tail-flick latencies, reflective of tolerance, throughout the course of treatment. The MK/Mor group was resistant to this effect, showing greater tail-flick latencies throughout the course of treatment (Fig. 2a). Statistical analysis of tail-flick scores across the 9 days revealed a significant effect of treatment [F(3,20) = 110.16, p < 0.01], days [F(4,80) = 17.33, p < 0.01], and a treatment × days interaction [F(12,80) = 10.49, p < 0.01]. On day 11, when animals were challenged with morphine a maximal antinociceptive effect was seen in the two control groups that received morphine for the first time (Sal/Sal and MK/Sal), as well as the MK/Mor group. Tail-flick latencies for the MK/Mor group significantly differed from the Sal/Mor group (p < 0.01, Fig. 2b).

Fig. 2
figure 2

MK-801 inhibits morphine tolerance and sensitization in the same experimental animals. Animals were injected once daily for 10 days with saline or MK-801 (0.1 mg/kg) followed 30 min later by saline or morphine (5.0 mg/kg). Tail-flick latencies were assessed 60 min after morphine treatment on odd-numbered days. On day 11, animals were challenged with morphine (5.0 mg/kg) and the time course of activity was assessed. On this day, animals were briefly removed from locomotor testing to assess tail-flick latencies 60 min after morphine treatment (see arrow). a Tail-flick latencies during the first 9 days of treatment. b Tail-flick latencies on day 11, 60 min after morphine challenge. c Time course of the locomotor response to morphine in the four treatment groups on day 11. d Total activity during the first 60 min post-injection on day 11. Note that the MK/Mor (M/M) group showed greater analgesia relative to the Sal/Mor (S/M) group throughout testing, reflecting an inhibition of the development of tolerance. In addition, as in experiment 1, the Sal/Mor (S/M) group showed greater activity than all other groups (sensitization), whereas the MK/Mor (M/M) group did not differ from the control groups (sensitization did not develop in this group). *MK/Mor group significantly different from Sal/Mor group; **Sal/Mor group significantly different (p < 0.05) from all other treatment groups. N = 6 animals per group. The arrow identifies the point at which animals were briefly removed from the enclosure for tail-flick testing. The burst in activity after the test was likely an interaction between the drug effects and the handling of the animals

The pattern of locomotor activity on day 11 was similar to that seen in experiments 1.1 and 1.2. As above, an increase in activity was found in the Sal/Mor group when compared to the other treatment groups (Fig. 2c). Significant effects were found for time [F(23,460) = 42.31, p < 0.01] and the interaction of treatment × time [F(69,460) = 3.70, p < 0.01]. Analysis of the first 60 min post-injection revealed that the Sal/Mor group was significantly more active than all other treatment groups (p < 0.01), whereas the MK/Mor group did not differ from the control groups (Fig. 2d).

Experiment 2.2: Effects of the NMDA receptor antagonist LY235959 on tolerance to morphine-induced antinociception and sensitization to morphine-induced locomotor stimulation

Groups receiving Sal/Sal and LY/Sal showed constant tail-flick scores of approximately 3–4 s throughout the 10 days of injections. Both groups that received morphine showed maximal antinociceptive effects early in treatment, with the Sal/Mor group showing decreasing tail-flick latencies (reflective of tolerance) throughout the course of the treatment. The LY/Mor group was resistant to this effect, showing higher tail-flick latencies throughout the 9 days of testing (Fig. 3a). This was confirmed by statistical analysis, which revealed a significant effect of treatment [F(3,20) = 199.80, p < 0.01), days [F(4,80) = 15.57, p < 0.01) and an interaction for treatment × days [F(12,80) = 5.95, p < 0.01]. On day 11, when all animals were challenged with morphine, maximal tail-flick latencies were evident in the groups that received morphine for the first time (Sal/Sal and LY/Sal), as well as the LY/Mor group. The Sal/Mor group showed lower tail-flick latencies and differed significantly from each of the other groups (p < 0.01, Fig. 3b).

Fig. 3
figure 3

LY235959 inhibits morphine tolerance and sensitization in the same experimental animals. Animals were injected once daily for 10 days with saline or LY235959 (1.0 mg/kg) followed 30 min later by saline or morphine (5.0 mg/kg). Tail-flick latencies were assessed 60 min after morphine treatment on odd-numbered days. On day 11, animals were challenged with morphine (5.0 mg/kg), and the time course of activity was assessed. On this day, animals were briefly removed from locomotor testing to assess tail-flick latencies 60 min after morphine treatment (see arrow). a Tail-flick latencies during the first 9 days of treatment. b Tail-flick latencies on day 11, 60 min after morphine challenge. c Time course of the locomotor response to morphine in the four treatment groups on day 11. d Total activity during the first 60 min post-injection on day 11. Note that the LY/Mor (L/M) group showed greater analgesia relative to the Sal/Mor (S/M) group throughout testing, reflecting an inhibition of the development of tolerance. In addition, the Sal/Mor (S/M) group showed greater activity than all other groups (sensitization), whereas the LY/Mor (L/M) group did not differ from the control groups (sensitization did not develop in this group). *LY/Mor group significantly different from Sal/Mor group; **Sal/Mor group significantly different (p < 0.05) from all other treatment groups. N = 6 animals/group. The arrow identifies the point at which animals were briefly removed from the enclosure for tail-flick testing. The burst in activity after the test was likely an interaction between the drug effects and the handling of the animals

On day 11, an increased stimulant effect of morphine was seen in the Sal/Mor group relative to all other groups, reflective of sensitization (Fig. 3c). Two-way repeated measures ANOVA revealed a significant effect for treatment [F(3,20) = 6.72, p < 0.01) and time [F(23,460) = 31.83, p < 0.01], as well as the interaction of treatment × time [F(69,460) = 3.93, p < 0.01). Analysis of the first 60 min post-injection revealed that the Sal/Mor group was significantly more active than all other treatment groups (p < 0.01), whereas the LY/Mor group did not differ from the control groups (Fig. 3d).

  1. 3.

    Effects of MK-801 on tolerance and sensitization to methadone and buprenorphine

Experiment 3.1: Effects of the NMDA receptor antagonist MK-801 on tolerance to methadone-induced antinociception and sensitization to methadone-induced locomotor stimulation

Groups receiving Sal/Sal and MK/Sal showed constant tail-flick scores at approximately 3–4 s throughout the 15 days of testing. Both groups that received methadone showed maximal antinociceptive effects on day 1 of treatment; however, the Sal/Meth group showed decreasing tail-flick latencies during the course of the treatment, eventually returning to baseline latencies by day 15, reflecting the development of tolerance. The MK/Meth group was resistant to these changes, showing higher tail-flick latencies throughout testing (Fig. 4a). A significant effect for tail-flick responses was found for treatment [F(3,19) = 39.67, p < 0.01), days [F(9,171) = 35.09, p < 0.01), and the interaction of treatment × days [F(27,171) = 8.81, p < 0.01). On day 17, when all animals were injected with methadone, tail-flick scores revealed maximal latencies in the control groups receiving methadone for the first time, as well as the MK/Meth group. Tolerance to these antinociceptive effects was evident in the Sal/Meth group when compared to the MK/Meth group (p < 0.01, Fig. 4b).

Fig. 4
figure 4

MK-801 inhibits methadone tolerance and sensitization in the same experimental animals. Animals were injected once daily for 16 days with saline or MK-801 (0.1 mg/kg) followed 30 min later by saline or methadone (3.0 mg/kg). Tail-flick latencies were assessed 60 min after methadone treatment on odd-numbered days. On day 17, animals were challenged with methadone (3.0 mg/kg) and the time course of activity was assessed. On this day, animals were briefly removed from locomotor testing to assess tail-flick latencies 60 min after methadone treatment (see arrow). a Tail-flick latencies during the first 15 days of treatment. b Tail-flick latencies on day 17, 60 min after methadone challenge. c Time course of the locomotor response to methadone in the four treatment groups on day 17. d Total activity during the first 120 min post-injection on day 17. Note that the MK/Meth (M/M) group showed greater analgesia relative to the Sal/Meth (S/M) group throughout testing, reflecting an inhibition of the development of tolerance. In addition, the Sal/Meth (S/M) group showed greater activity than all other groups (sensitization), whereas the MK/Meth (M/M) group did not differ from the control groups (sensitization did not develop in this group). *MK/Meth group significantly different from Sal/Meth group; **Sal/Meth group significantly different (p < 0.05) from all other treatment groups. N = 6 animals/group (except for Sal/Meth; N = 5). The arrow identifies the point at which animals were briefly removed from the enclosure for tail-flick testing. The burst in activity after the test was likely an interaction between the drug effects and the handling of the animals

During the locomotor test on day 17, a delayed stimulant effect was evident in all groups. However, the Sal/Meth group showed a much earlier stimulation in activity, and the greatest activity overall, reflective of sensitization (Fig. 4c). Two-way repeated measures ANOVA revealed a significant effect for treatment [F(3,16) = 14.26, p < 0.01] and time [F(23,368) = 35.44, p < 0.01], as well as a treatment × time interaction [F(69,368) = 2.75, p < 0.01]. Examination of the first 120 min post-injection revealed that the Sal/Meth group had significantly greater activity than all other treatment groups (p < 0.01), whereas the MK/Meth group did not differ from the control groups (Fig. 4d).

Experiment 3.2: Effects of the NMDA receptor antagonist MK-801 on tolerance to buprenorphine-induced antinociception and sensitization to buprenorphine-induced locomotor stimulation

Both groups receiving buprenorphine showed maximal antinociceptive effects on day 1 of injections. Across the 9 days of treatment, the group receiving Sal/Bup showed decreasing tail-flick scores, reflecting the development of tolerance, whereas the MK/Bup showed higher tail-flick scores, illustrating inhibition of tolerance (Fig. 5a). Statistical analysis of the tail-flick responses revealed a significant effect of the treatment [F(3,20) = 53.46, p < 0.01], days [F(4,80) = 13.28, p < 0.01] and an interaction of treatment × days [F(12,80) = 2.76, p < 0.01). After buprenorphine challenge on day 11, the Sal/Sal, MK/Sal, and MK/Bup groups showed maximal antinociceptive effects, whereas the Sal/Bup group showed lower tail-flick latencies. Statistical analysis of the day 11 tail-flick latencies revealed that the Sal/Bup group differed significantly from all other groups (p < .01, Fig. 5b).

Fig. 5
figure 5

MK-801 inhibits buprenorphine tolerance and sensitization. To assess buprenorphine tolerance, animals were injected once daily for 10 days with saline or MK-801 (0.1 mg/kg) followed 30 min later by 1% Tween 80 vehicle (Twn) or buprenorphine (1.0 mg/kg). Tail-flick latencies were assessed 60 minute after buprenorphine treatment on odd-numbered days. On day 11, animals were challenged with buprenorphine (1.0 mg/kg), and the tail-flick latencies were assessed 60 min after treatment. To assess buprenorphine sensitization, animals were injected once daily for 16 days with saline or MK-801 (0.1 mg/kg) followed 30 min later by 1% Tween 80 vehicle (Twn) or buprenorphine (0.3 mg/kg). On day 17, animals were challenged with buprenorphine (0.3 mg/kg), and the time course of activity was assessed. Animals were briefly removed from locomotor testing 60 min after buprenorphine treatment (see arrow). a Tail-flick latencies during the first 9 days of treatment. b Tail-flick latencies on day 11 after buprenorphine challenge. c Time course of response to buprenorphine in the four treatment groups on day 17. d Total activity during the first 120 minutes post-injection on day 17. Note that the MK/Bup (M/B) group showed greater analgesia relative to the Sal/Bup (S/B) group throughout testing, reflecting an inhibition of the development of tolerance. In addition, the Sal/Bup (S/B) group showed greater activity than all other groups (sensitization), whereas the MK/Bup (M/B) group did not differ from the control groups (sensitization did not develop in this group). *MK/Bup group significantly different from Sal/Bup group; **Sal/Bup group significantly different (p < 0.05) from all other treatment groups. N = 6 animals/group (except for Sal/Sal and MK/Sal groups; N = 3 per group). The arrow identifies the point at which animals were briefly removed from the enclosure for tail-flick testing. The burst in activity after the test was likely an interaction between the drug effects and the handling of the animals

For the locomotor analysis on day 17, the Sal/Bup group showed the greatest activity of all treatment groups (Fig. 5c). Two-way repeated measures ANOVA revealed a significant effect for treatment [F(3,14) = 14.26, p < 0.01] and time [F(23,322) = 35.44, p < 0.01], as well as a treatment × time interaction [F(69,322) = 2.75, p < 0.01]. Analysis of the first 120 min post-injection revealed that the Sal/Bup group was significantly more active than all other groups (p < 0.01), whereas the MK/Bup group did not differ from the control groups (Fig. 5d).

Discussion

The purpose of these studies was to explore the hypothesis that NMDA receptors are involved in opiate-induced plasticity, including tolerance and sensitization. The first two experiments demonstrated that the non-competitive NMDA receptor antagonists MK-801 and memantine inhibit the development of sensitization to the locomotor stimulant effect of morphine. The second set of studies replicated and extended these findings, showing that the non-competitive NMDA receptor antagonist MK-801 and the competitive antagonist LY235959 inhibit the development of tolerance to the antinociceptive effect and sensitization to the locomotor stimulant effect of morphine in the same experimental animals. The final two experiments examined other opiates, showing that MK-801 inhibits tolerance to the antinociceptive effect and sensitization to the stimulant effect of both methadone and buprenorphine. Together, the findings are consistent with the idea that NMDA receptors are involved in opiate tolerance and sensitization (Allen and Dykstra 1999; Allen and Dykstra 2000b; Belozertseva and Bespalov 1998; Bisaga and Popik 2000; Herman et al. 1995; Inturrisi 1997; Mao 1999; Mayer and Mao 1999; Trujillo 2000, 2002; Trujillo and Akil 1991, 1995).

The idea that NMDA receptor antagonists inhibit the development of tolerance to the antinociceptive effect of morphine is widely accepted (for review, see Bisaga and Popik 2000; Herman et al. 1995; Inturrisi 1997; Mao 1999; Mayer et al. 1999; Trujillo 2000, 2002; and Trujillo and Akil 1995). However, there has been some debate over the ability of these antagonists to inhibit the development of tolerance to opiates other than morphine (Allen and Dykstra 2000a; Bilsky et al. 1996; Mao et al. 1998). In the present experiments, MK-801 inhibited the development of tolerance to the antinociceptive effect of three different opiates that act via mu opioid receptors: morphine, methadone, and buprenorphine, suggesting that NMDA receptor antagonists can indeed broadly inhibit opiate tolerance. Of importance to this controversy, Allen and Dykstra (2000a, b) have suggested that inhibition of opioid tolerance by NMDA receptor antagonists depends on the magnitude of tolerance. Although we did not specifically assess the magnitude of tolerance in the current studies, it is notable that tolerance to all three opioids was inhibited by NMDA receptor antagonists. The present results, together with previous findings, demonstrate that NMDA receptors are involved in tolerance to a variety of opiates and suggest that combinations of NMDA receptor antagonists with opiates may prolong their clinical effectiveness.

The ability of NMDA receptor antagonists to inhibit sensitization to opiate-induced locomotor stimulation has been reported in a few studies (Iijima et al. 1996; Jeziorski et al. 1994; Kosten and Bombace 2000; Vanderschuren et al. 1997; Wolf and Jeziorski 1993); however, these findings have generated controversy. First, some attempts to replicate these studies have failed (Atalla and Kuschinsky 2006; Ranaldi et al. 2000). Second, in some studies in which replication was found, it was only seen at doses that produced confounding behavioral effects or lethality (Iijima et al. 1996; Vanderschuren et al. 1997). Finally, use of the potent NMDA receptor antagonist MK-801 has been criticized because of the behavioral side effects of this drug, including locomotor stimulation and ataxia (Carlezon et al. 2000; Ranaldi et al. 2000; Tzschentke and Schmidt 1996, 1998; Tzschentke and Schmidt 2000; Vanderschuren and Kalivas 2000; Vanderschuren et al. 1997; see Trujillo 2000 for further discussion). The present experiments attempted to overcome these concerns through a variety of strategies: First, low, selective doses of NMDA receptor antagonists were used. These doses produced no observable toxicity and few behavioral side effects in the present study. Second, three different NMDA receptor antagonists were used, the high affinity non-competitive antagonist, MK-801; the lower affinity non-competitive antagonist (and clinically available drug), memantine; and the competitive antagonist, LY235959. The two latter drugs produce different behavioral profiles and fewer behavioral side effects than MK-801. For example, memantine produces locomotor stimulation at relatively high doses but no ataxia, whereas LY235959 produces locomotor depression at higher doses rather than stimulation (and therefore, produces a very different behavioral profile than either MK-801 or memantine; Parsons et al. 1999; Peterson 2003; Trujillo et al. 2000). Moreover, at the respective doses utilized, each of these drugs produces a unique acute interaction with morphine in locomotor behavior—MK-801 enhances, LY235959 inhibits, and memantine has no effect on morphine-induced locomotor stimulation (Trujillo, Mendez and Warmoth, unpublished results). As all three antagonists inhibited the development of sensitization, this effect cannot be tied to a specific locomotor side effect. A third strategy that we used was to examine three different opiates, including morphine, methadone, and buprenorphine. As MK-801 inhibited the development of sensitization to all three, the results support the idea that NMDA receptors are broadly involved in sensitization to the stimulant effect of opiates.

We used an additional strategy that increases confidence regarding the conclusions of the present experiments and that is to compare the development of tolerance and sensitization in the same experimental animals. As noted above, it is widely accepted by the scientific community that NMDA receptors are involved in the development of tolerance to the antinociceptive effect of morphine. However, there has been less confidence over the role of these receptors in the development of sensitization to the stimulant effect of opiates. The fact that NMDA receptor antagonists inhibited tolerance and sensitization in a similar manner in the same experimental animals suggests that similar NMDA-dependent neural processes are involved in each of these phenomena. Although antinociceptive tolerance is mediated largely at the level of the spinal cord (see Gutstein and Trujillo 1993) and locomotor sensitization likely involves changes in the mesocorticolimbic system (see Vanderschuren and Kalivas 2000), similar NMDA receptor-dependent synaptic processes appear to be responsible for these distinct forms of plasticity (see Mao 1999; see Trujillo 2002). This should not be surprising, as NMDA receptors have been found to be involved in several different types of behavioral and neural plasticity (see Lodge et al. 2002). A potential alternative explanation for these data is that NMDA receptors are involved in the development of tolerance but that non-specific behavioral disruption accounts for the ability of NMDA receptor antagonists to inhibit sensitization. However, the more parsimonious explanation is that NMDA receptors are involved in both phenomena in a similar manner.

It is presently unclear why some studies have achieved inhibition of morphine sensitization by low doses of NMDA receptor antagonists while others have not. According to Li and Wolf (1999) and Wolf (1998), context-dependent sensitization to psychomotor stimulants (typically obtained by repeated pairings of drugs with the testing environment) is more resistant to NMDA receptor blockade than context-independent sensitization. The current studies used a context-independent approach to examine sensitization, administering repeated injections of the opiates in the vivarium and testing in a separate experimental room. Further research will help to clarify the role of context dependence and other factors in the ability of NMDA receptor antagonists to inhibit sensitization to opiates.

Regarding alternative explanations, it has been suggested that “state dependency” may account for the ability of NMDA receptor antagonists to inhibit the development of tolerance and sensitization (Tzschentke and Schmidt 1998; Tzschentke and Schmidt 2000; see also Trujillo 2000; Wolf 1998). A common experimental design for these studies is to administer the NMDA receptor antagonist with the opiate during chronic administration but then to test for tolerance and sensitization in the absence of the NMDA receptor antagonist. This leads to the possibility that the tolerance and sensitization are dependent on the physiological “state” provided by the NMDA receptor antagonist; according to this hypothesis (Carlezon et al. 1995; Tzschentke and Schmidt 1998; Tzschentke and Schmidt 2000), tolerance and sensitization are not seen during the test because it occurs in a different “state” (in the absence of the NMDA antagonist). In the present studies, the development of tolerance to the antinociceptive effect of the opiates was tested in both the presence and the absence of the NMDA receptor antagonists. Inhibition of tolerance was seen in both conditions, demonstrating that the effects are not caused by state dependency. This supports and extends previous research demonstrating that the inhibition of tolerance to the antinociceptive effects of opiates by NMDA receptor antagonists is not caused by state dependency and instead more likely caused by inhibition of the neural plasticity responsible for tolerance (Ben-Eliyahu et al. 1992; Gutstein and Trujillo 1993; Marek et al. 1991; Tiseo and Inturrisi 1993; Trujillo and Akil 1991; see Trujillo 2000 for review). In the present experiments, because sensitization was inhibited in the same manner as tolerance, in the same experimental animals, the most parsimonious explanation is that inhibition of sensitization was similarly the result of inhibition of plasticity.

An additional argument against state dependency in the present studies is that three different NMDA receptor antagonists with distinct behavioral effects and interoceptive cues (Bubser et al. 1992; Danysz et al. 1994; Danysz and Parsons 1998; Geter-Douglass and Witkin 1997, 1999; Parsons et al. 1999) produced a similar inhibition of tolerance and sensitization. This indicates that the inhibition of these phenomena does not depend on a unique pharmacological state but, instead, occurs whenever NMDA receptors are blocked (see Trujillo 2000 for further discussion). Finally, it is important to note that nonpharmacological approaches, including antisense and transgenic knockdown of NMDA receptors, support the idea that NMDA receptor activation is important for opiate-induced plasticity and provide strong evidence that a unique pharmacological state is not required for inhibition of these phenomena (Miyamoto et al. 2004; Shimoyama et al. 2005). Together, the results argue against a key role for state dependency in the inhibition of tolerance and sensitization. Li and Wolf (1999) and Wolf (1998) have similarly argued that state dependency cannot explain the ability of NMDA receptor antagonists to inhibit the development of sensitization to psychomotor stimulants.

Overall, the results of these experiments support a role for NMDA receptors in opiate-induced neural and behavioral plasticity, including tolerance and sensitization. Interest in the relationship between excitatory amino acids and drug-induced neuronal changes has received significant attention over the last 15 years. This has led to clinical studies on the potential benefits of co-administration of NMDA antagonists with opiates for the treatment of pain and addiction, with promising results (Bell et al. 2005; Bisaga and Popik 2000; Subramaniam et al. 2004). Further research should help to refine our understanding of the role of NMDA receptors in the development of opiate-induced neural and behavioral plasticity, and the potential benefits of glutamatergic drugs in the treatment of pain and addiction.