- Email: [email protected]
Differential susceptibility of the PAG and RVM to tolerance to the antinociceptive effect of morphine in the rat
Pain 113 (2005) 91–98 www.elsevier.com/locate/pain
Differential susceptibility of the PAG and RVM to tolerance to the antinociceptive effect of morphine in the rat Michael M. Morgan*, Cecilea C. Clayton, Jill S. Boyer-Quick Department of Psychology, Washington State University, 14204 NE Salmon Creek Avenue, Vancouver, WA 98686, USA Received 10 June 2004; received in revised form 8 September 2004; accepted 28 September 2004
Abstract The periaqueductal gray (PAG) and rostral ventromedial medulla (RVM) are part of a nociceptive modulatory system. Microinjection of morphine into either structure produces antinociception. Tolerance develops to ventrolateral PAG mediated antinociception with repeated microinjection of morphine. In contrast, there are no published reports of tolerance to morphine administration into the RVM. Three experiments were conducted to determine whether tolerance develops to morphine microinjections into the RVM. Experiment 1 compared tolerance to the antinociceptive effect of microinjecting morphine (5 mg/0.5 ml) into the PAG and RVM following daily injections for four consecutive days. Experiment 2 assessed tolerance to a range of morphine doses (2.5–20 mg) after injecting morphine into the RVM twice a day for two consecutive days. Experiment 3 followed a similar procedure except twice as many RVM injections were made (8 microinjections in 4 days). The degree to which tolerance developed to the antinociceptive effect of morphine was much greater with microinjections into the PAG compared to the RVM. There was a 64% drop in hot plate latency from the first to the fifth injection of morphine into the PAG, but only a 36% drop in latency following RVM microinjections. Reducing the interdose interval to two injections a day or increasing the total number of injections from 4 to 8 did not enhance the development of tolerance to RVM morphine administration. These data demonstrate that opioid-sensitive neurons in the RVM are relatively resistant to the development of tolerance compared to PAG neurons. q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. Keywords: Opiate
Opiates such as morphine are the most effective treatment for pain. Unfortunately, morphine can be rendered ineffective with repeated administration because of the development of tolerance. Although numerous mechanisms for tolerance have been proposed, a clear understanding of the neural changes responsible for tolerance is unclear. An important step in understanding the mechanisms underlying tolerance is to identify the neural structures mediating tolerance. Opiates produce antinociception, at least in part, by activating a descending modulatory system that includes the periaqueductal gray (PAG) and rostral ventromedial medulla (RVM) (Basbaum and Fields, 1984; Fields et al., * Corresponding author. Tel.: C1 360 546 9726; fax: C1 360 546 9038. E-mail address: [email protected] (M.M. Morgan).
1991, 1995). Microinjection of morphine into either the PAG or RVM produces antinociception (Jacquet and Lajtha, 1974; Jensen and Yaksh, 1986; Morgan et al., 1998; Yaksh et al., 1976), and inactivation of the RVM attenuates the antinociceptive effect of systemically administered morphine (Heinricher et al., 1997; Proudfit, 1980; Young et al., 1984). The important role that the PAG and RVM play in morphine antinociception make these structures potential sites where tolerance to morphine may occur. A number of studies have shown that tolerance develops to repeated microinjections of morphine into the PAG (Jacquet and Lajtha, 1976; Lewis and Gebhart, 1977; Siuciak and Advokat, 1987; Tortorici et al., 1999, 2001). In contrast, there are no published reports of tolerance to the antinociceptive effect of microinjecting morphine into the RVM. Given the prominent role of the RVM in modulating
0304-3959/$20.00 q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2004.09.039
92
M.M. Morgan et al. / Pain 113 (2005) 91–98
nociception, it is important to determine whether this structure contributes to morphine tolerance. Antinociception has been shown to be reduced with repeated electrical stimulation of the RVM (Oliveras et al., 1978). In addition, chronic systemic administration of morphine produces changes in the activity of neurons in and around the RVM (Bederson et al., 1990; Haghparast et al., 1998; Li and Wang, 2001), and inactivation of the RVM attenuates tolerance to chronic systemic morphine administration (Vanderah et al., 2001). Although these studies indicate that RVM neurons contribute to the development of tolerance to morphine, the changes in RVM neurons that occur as tolerance develops could be secondary to changes in other structures such as the ventrolateral PAG (Behbehani, 1981; Tortorici et al., 2001). That is, changes in RVM neurons could correlate with the development of tolerance without causing tolerance. The objective of this paper is to test the hypothesis that tolerance will occur to the antinociceptive effect of direct administration of morphine into the RVM.
1. Materials and methods 1.1. Subjects Male Sprague–Dawley rats (280–380 g; Animal Technologies Inc., Kent, WA) were anesthetised with sodium pentobarbital (Nembutal; 60 mg/kg, i.p.) and stereotaxically implanted with a 23 gauge stainless-steel guide cannula aimed at the PAG (9 mm guide; AP 1.2 mm, ML 0.6 mm; DV K5.0 mm) or RVM (15 mm long; AP K2.2 mm; ML 0.0 mm, DV K8.8 mm from lambda). The guide cannula was held in place with dental cement affixed to two screws in the skull. A removable 31-gauge stainless-steel stylet was inserted into the guide cannula following surgery. The rat was allowed to recover for 1 week prior to testing. Animals were housed individually following surgery with food and water available ad libitum. Lights were maintained on a reverse 12 h light/dark cycle. All injections and testing were conducted during the dark phase when rats are awake and active. Experiments were conducted in accordance with the animal care and use guidelines of the Committee for Research and Ethical Issues of IASP. This research was approved by the Institutional Animal Care and Use Committee at Washington State University. Rats were handled daily following surgery. In addition, each rat was habituated to the injection procedure prior to the first day of testing by inserting an injection cannula through the guide in the absence of drug administration. This procedure was used to reduce non-specific effects resulting from mechanical damage to neurons on the first test day. Microinjections of morphine sulfate (a gift from the National Institute on Drug Abuse) and saline were made into the PAG or RVM through a 31-gauge injection cannula that extended 2 mm beyond the tip of the guide cannula. The injection cannula was connected to a 1 ml syringe (Hamilton Co., Reno, NV) with PE20 tubing filled with sterile water. All microinjections were given in a volume of 0.5 ml over 50 s while the rat was gently restrained. The injection cannula remained in place for an additional 20 s to
minimize backflow of the drug up the cannula tract. Following the injection, the stylet was reinserted into the guide cannula and the rat was returned to its home cage. 1.2. Behavioral assessment Nociception was assessed using the hot plate test (IITC, Woodland Hills, CA). The hot plate test measures the latency to lick a hindpaw when placed on a 51–52 8C metal surface. The rat was removed from the hot plate immediately following a response or after 40 s if no response occurred. Previous research has shown a clear tolerance to the antinociceptive effects of morphine using this test (Tortorici et al., 1999). The number of tests was kept to a minimum (a single test on the first and last trials) in the first experiment to avoid changes in latency associated with repeated testing (Gamble and Milne, 1989). 1.3. Histology Rats were given a lethal injection of pentobarbital (120 mg/kg, i.p.) immediately following testing. The microinjection site was marked by injecting cresyl violet into the same site as the drugs were administered previously. The brain was removed and placed in formalin (10%). At least 3 days later, the brain was sectioned coronally (50 mm) and viewed under a microscope to localize the injection site (Paxinos and Watson, 1986). 1.4. Data analysis Students’ t-test and analysis of variance were used to compare differences in hot plate latency between groups. Post-hoc analysis of within group changes from the first to last trial was conducted by calculating the 95% confidence limits for specific means. 1.5. Experiment 1: comparison of tolerance to morphine injections into the PAG and RVM The objective of this experiment was to compare the magnitude of tolerance to the antinociceptive effect of morphine microinjection into the PAG and RVM. Rats were implanted with a guide cannula aimed at the PAG or RVM. Morphine (5 mg/0.5 ml) was injected into the PAG or RVM once a day for four consecutive days. Saline was injected into the RVM on Days 1–4 as a control. All rats received a microinjection of morphine on Day 5 to determine whether tolerance had developed. Nociception was assessed using the hot plate test 30 min after the injection on Days 1 and 5. Rats were returned to their home cage without nociceptive testing following microinjections on Days 2–4 to minimize potential changes in latency caused by repeated testing (Gamble and Milne, 1989). Given that the PAG and RVM are part of the same nociceptive modulatory system, it is hypothesised that tolerance will develop at the same rate in both structures. 1.6. Experiment 2: analysis of interdose interval on tolerance to RVM morphine microinjections Morphine was injected once a day for 5 days in Experiment 1. The objective of Experiment 2 was to determine whether increasing the exposure of RVM neurons to morphine by shortening the interdose interval to two injections a day for 2
M.M. Morgan et al. / Pain 113 (2005) 91–98
93
days would enhance the development of tolerance to the antinociceptive effect of morphine. Rats received the same number of injections as in Experiment 1 but over 2 days instead of four. Rats were randomly assigned to a morphine or saline pretreatment group. The morphine pretreatment group was injected with morphine (5 mg/0.5 ml) into the RVM twice a day (9:00 AM and 4:00 PM) for two consecutive days. The saline pretreatment group was injected with saline at the same times to control for the effects of repeated microinjections. Nociception was assessed using the hot plate test 30 min after each injection. On Day 3 (Trial 5), these morphine and saline pretreated rats were injected with one of five morphine doses (0, 2.5, 5, 10, and 20 mg/0.5 ml). Given that the procedure described above produces tolerance when morphine is injected into the ventrolateral PAG (Tortorici et al., 1999), we hypothesised that tolerance to RVM morphine administration will develop more rapidly using this procedure than with once daily microinjections. A range of doses was tested to determine whether morphine pretreatment causes a change in potency. 1.7. Experiment 3: effect of doubling the number of injections on tolerance to RVM morphine administration Morphine was administered four times to induce tolerance in Experiments 1 (once daily for 4 days) and 2 (twice daily for 2 days). The present experiment doubled the number of morphine injections by combining these two approaches (two daily injections for 4 days: 9:00 AM and 4:00 PM). Doubling the number of injections would be expected to enhance the development of tolerance. Rats were tested as in Experiment 2 except that rats received eight injections of morphine (5 or 10 mg/0.5 ml) or saline into the RVM prior to the final test. Nociception was assessed only after the morning injections so the number of hot plate tests did not differ from Experiment 2. Both morphine and saline pretreated rats were injected with morphine on Day 5 (Trial 9) to determine whether tolerance had developed. The same dose was used for induction (Trials 1–8) and assessment of tolerance (Trial 9). If RVM neurons are susceptible to morphine tolerance, then tolerance should be greater in rats receiving twice as many morphine injections.
2. Results 2.1. Experiment 1: comparison of tolerance in the PAG and RVM Only rats with microinjections into the ventrolateral PAG or RVM (nucleus raphe magnus and adjacent nucleus reticularis gigantocellularis pars alpha) were included in data analysis (Fig. 1). These data consisted of 10 rats microinjected with morphine into the PAG, 16 with morphine microinjected into the RVM, and eight with saline microinjected into the RVM. Given that susceptibility to morphine tolerance has been shown to be site specific within the PAG (Tortorici et al., 1999), injection sites outside the boundaries of the ventrolateral PAG and RVM were not included in data analysis. Most of these misplaced injections were just outside the border, and microinjection
Fig. 1. Location of microinjection sites in the ventrolateral PAG (top) and RVM (bottom) for Experiments 1–3. Only morphine and saline injections within the shaded regions were included in data analysis. Coronal sections are taken from the Atlas of Paxinos and Watson (1986). The number in the top left corner refers to the distance from the interaurel line. All injections fell within G0.5 mm of these coronal planes.
of morphine into these sites produced antinociception in many cases. Microinjection of morphine produced an increase in hot plate latency following PAG and RVM administration compared to saline treated rats on Trial 1 (Fig. 2). As reported previously (Morgan and Whitney, 2000; Tortorici and Morgan, 2002), the antinociceptive effect of morphine was greater following PAG compared to RVM administration. Microinjection of morphine had the opposite effect on Trial 5. Microinjection of morphine into the RVM produced greater antinociception than microinjection into the PAG (Trial 5 in Fig. 2). There was a mean decrease in hot plate latency from Trial 1–5 following daily morphine administration into the PAG of 20.3G4.3 s. The mean decrease in hot plate latency with repeated microinjection of morphine into the RVM was only 9.9G3.2 s. This decrease in hot plate latency across trials was statistically significant (F(1,24)Z32.27, p!0.05). The degree to which tolerance developed following PAG microinjections of morphine was greater than with RVM microinjections. This difference was evident by a nearly significant group (PAG vs. RVM) by trial interaction (F(1,24)Z3.78, PZ0.06). Repeated microinjections of saline into the RVM did not prevent morphine
94
M.M. Morgan et al. / Pain 113 (2005) 91–98
Fig. 2. Comparison of tolerance to the antinociceptive effect of microinjecting morphine into the PAG and RVM. Microinjection of morphine (5 mg/0.5 ml) into the PAG and RVM produced an increase in hot plate latency compared to rats injected with saline into the RVM on Trial 1. All rats were injected with morphine on Trial 5. Hot plate latency was greatly reduced in morphine-pretreated rats compared to Trial 1. A 64% decrease in hot plate latency from Trial 1–5 was evident with PAG administration. A decrease of only 36% occurred with RVM administration. Repeated saline injections into the RVM did not disrupt the antinociceptive effect of morphine administration on Trial 5. These data demonstrate that the ventrolateral PAG is more susceptible to morphine tolerance than the RVM.
from producing antinociception when administered for the first time on Trial 5. 2.2. Experiment 2: analysis of interdose interval on tolerance to RVM morphine microinjections This experiment assessed whether a decrease in the interdose interval would enhance the development of tolerance to RVM morphine microinjection. Only data from injection sites located in the RVM were included in the statistical analysis (Fig. 1). This included 41 rats pretreated with microinjections of morphine (5 mg/0.5 ml) and 28 pretreated with microinjections of saline. These large groups were subdivided for dose–response analysis on Day 3 (Trial 5). MeanGSEM hot plate latency for saline treated rats was 10.7G0.7 s on Trial 1. None of these rats had a hot plate latency greater than 20 s. In contrast, microinjection of morphine into the RVM produced an increase in hot plate latency of 20 s or greater in 23 of the 41 rats (56%). The finding that only a subset of rats had hot plate latencies greater than 20 s following microinjection of morphine into the RVM is consistent with previous research (Morgan and Whitney, 2000; Tortorici and Morgan, 2002). The number of rats with a hot plate latency greater than 20 s dropped on each subsequent trial (46% on Trial 2; 40% on Trial 3; and 37% on Trial 4). Although this decrease suggests tolerance may occur in a subset of rats, analysis of morphine
antinociception in saline pretreated rats suggests this decrease is caused by behavioral and associative tolerance, not from a direct action of morphine on RVM neurons (see analysis of mean hot plate latency below). Microinjection of morphine into the RVM produced antinociception on each of the four pretreatment trials compared to saline treated controls (Fig. 3). Analysis of variance revealed that the hot plate latency for rats injected with morphine was significantly higher than the latency for saline treated rats across all trials (F(1,65)Z297.87, P!0.05). Moreover, there was no group by trial interaction (F(1,195)Z1.085, n.s.), an effect that would be expected if there had been a reduction in antinociception across trials. Although a small decrease in mean hot plate latencies across trials was evident following both morphine and saline administration on Trials 1–4, statistical support for tolerance to RVM morphine administration did not occur. The decrease in hot plate latency across trials occurred in both the saline and morphine treated rats indicating it was caused by repeated testing (Gamble and Milne, 1989; Milne et al., 1989), not tolerance to morphine administration. Morphine and saline pretreated rats were injected with a range of morphine doses on Trial 5 (0, 2.5, 5, 10, and 20 mg/0.5 ml; NZ5–12 rats/condition). Microinjection of morphine produced a dose dependent increase in hot plate latency in both morphine and saline pretreated rats (F(4,58)Z2.997, P!0.05; Fig. 4). Morphine pretreatment did not alter the potency for antinociception produced by microinjecting morphine into the RVM. In fact, the ED50 for antinociception following morphine microinjection was lower in morphine (5.1 mg) compared to saline (5.9 mg) pretreated rats.
Fig. 3. Effect of repeated microinjections of morphine and saline into the RVM on hot plate latency across trials. Rats were injected with morphine (5 mg/0.5 ml) or saline twice a day for two consecutive days. The elevation in hot plate latency across all four trials in morphine treated rats indicates that RVM neurons are relatively insensitive to the development of tolerance. The slight decrease in hot plate latency across trials is evident in both morphine and saline pretreated groups, and suggests that this decrease is caused by repeated nociceptive testing (Gamble and Milne, 1989).
M.M. Morgan et al. / Pain 113 (2005) 91–98
Fig. 4. Comparison of the antinociceptive effect of microinjecting morphine into the RVM of rats pretreated with morphine or saline. Microinjection of morphine into the RVM produced a dose-related increase in hot plate latency in both morphine and saline pretreated rats. Rats were pretreated with morphine (5 mg/0.5 ml) or saline twice a day for two consecutive days. Both groups were injected with morphine (0, 2.5, 5, 10, or 20 mg/0.5 ml) into the RVM on Day 3. Tolerance should cause a rightward shift in the dose– response curve of morphine-pretreated rats. This did not happen suggesting that RVM neurons are resistant to morphine tolerance.
2.3. Experiment 3: effect of doubling the number of injections on tolerance to RVM morphine administration The relatively small decrease in antinociception with repeated morphine administration in Experiments 1 and 2 suggests that four injections may not be sufficient to induce tolerance. If this is the case, then doubling the number of morphine injections should double the magnitude of tolerance. The present experiment tested this hypothesis by injecting morphine twice a day for 4 days. All rats were injected with morphine 3 days prior to the tolerance induction procedure to determine whether the injection site supported antinociception. Only injection sites that were located in the RVM and produced antinociception to morphine administration on the pretest were included in data analysis. Given the relatively low rate of antinociception following RVM morphine administration, screening rats for antinociception allowed for a more focused analysis of changes in nociception with repeated morphine administration. Microinjection of morphine into the RVM produced a significant increase in hot plate latency in morphine treated rats (NZ11) compared to saline treated controls (NZ7) on Trial 1 (F(1,16)Z84.70; P!0.05). Although there was a slight decrease in hot plate latency across the eight trials, as in Experiment 2 this decrease was not statistically significant (F(3,48)Z1.677, n.s.). Moreover, there was no difference in antinociception between morphine and saline pretreated groups following morphine administration on Trial 9 (t(16)Z0.160, n.s.; Fig. 5). In addition, there was no group by trial interaction as would be expected if tolerance had occurred in the morphine treated group (F(3,48)Z0.919, n.s.). These data indicate that tolerance to
95
Fig. 5. Lack of tolerance following eight microinjections of morphine into the RVM. Rats were injected with morphine (5–10 mg/0.5 ml) or saline twice a day for four consecutive days. The first injection of morphine into the RVM produced antinociception compared to saline treated controls (Trial 1). Microinjection of morphine into the RVM of both morphine and saline pretreated rats produced antinociception on Trial 9. The antinociceptive effect was identical despite that this was the ninth injection of morphine for the morphine-pretreated group and the first injection of morphine for the saline pretreated group.
morphine did not occur despite eight prior microinjections of morphine into the RVM.
3. Discussion The present data suggest that RVM neurons are relatively resistant to the development of tolerance to the antinociceptive effect of morphine. Daily administration of morphine into the PAG produced a 64% drop in antinociception from the first to fifth injection, whereas morphine microinjections into the RVM produced a drop in antinociception of only 36% (Experiment 1). The decrease in antinociception was smaller when morphine was administered with a short interdose interval (two injections a day; Experiment 2) or when the number of injections was increased from four to eight (Experiment 3)—procedural changes that would be expected to enhance the development of tolerance to RVM morphine administration. We previously have shown that a profound tolerance develops with as little as two injections of morphine a day for 2 days when the PAG is the target (Tortorici et al., 1999). The present data show that neurons in the RVM are relatively resistant to the development of tolerance compared to neurons in the ventrolateral PAG. A lack of tolerance could simply indicate that the procedure used to induce tolerance is inadequate. Tolerance may require an increase in the number of morphine injections, an increase in the dose of morphine, an increase in the microinjection volume, a reduction in the time between
96
M.M. Morgan et al. / Pain 113 (2005) 91–98
injections, etc. Although it is impossible to test all possible factors in a single manuscript, the present findings demonstrate that shortening the interdose interval or increasing the number of injections did not enhance the development of tolerance to morphine microinjection into the RVM. The relatively stable antinociception produced by repeated microinjections of morphine into the RVM is surprising. The antinociceptive effect of microinjecting morphine into the PAG has been shown to be diminished by a single prior injection 3 days earlier (Lewis and Gebhart, 1977). Although relatively little tolerance was evident with daily microinjections of morphine into the RVM, this procedure was sufficient to produce tolerance when microinjections were directed at the PAG. The magnitude of tolerance to PAG administration was consistent whether morphine was injected into the ventrolateral PAG once (Experiment 1) or twice a day (Tortorici et al., 1999). Given that the same procedures were used for RVM administration, the present data indicate that RVM neurons are relatively resistant to the development of tolerance to the antinociceptive effects of morphine. The present paper appears to be the first published report examining tolerance to direct administration of morphine into the RVM. A previous study reported tolerance to repeated electrical stimulation of the RVM in the cat (Oliveras et al., 1978). This tolerance occurred with as little as 6 min of stimulation spread over 13 min. Whether this is tolerance or an artifact of electrical stimulation (e.g. neuronal fatigue) is debatable. In contrast, there have been numerous reports of tolerance to morphine microinjection into the PAG (Jacquet and Lajtha, 1976; Lewis and Gebhart, 1977; Siuciak and Advokat, 1987; Tortorici et al., 1999, 2001). Five additional studies show tolerance to antinociception produced by electrical stimulation of the PAG (Gebhart and Toleikis, 1978; Lewis and Gebhart, 1977; Mayer and Hayes, 1975; Millan et al., 1987; Morgan and Liebeskind, 1987). Given the important role of the RVM in modulating nociception, the lack of published reports of tolerance to the antinociceptive effect of morphine administered into this structure is surprising. This lack of data makes the present finding that the RVM is relatively resistant to the development of tolerance to morphine administration especially important. Although the decrease in antinociception across trials in all three experiments looks like tolerance, this decrease in hot plate latency is small. A similar decrease in hot plate latency across trials occurred in saline treated rats suggesting that repeated testing, not morphine administration, caused this change. This decrease in hot plate latency with repeated testing is called behavioral tolerance and can occur independent of morphine administration (Gamble and Milne, 1989). In addition, a reduction in morphine antinociception can be caused by neural damage following repeatedly inserting a microinjection cannula into the brain. This damage appears to have only a minor impact because microinjection of morphine into the RVM produced
antinociception in rats pretreated with multiple saline injections (see Figs. 2, 4, and 5). The most surprising finding is that the greatest decrease in morphine antinociception occurred with the fewest injections. That is, daily injections of morphine into the RVM caused a 36% decrease in hot plate latency in Experiment 1, whereas increasing the number of morphine pretreatments from four to eight resulted in a decrease in hot plate latency from Trial 1–9 of only 18% (Experiment 3). Previous research has shown that associative tolerance—tolerance caused by cues that predict drug administration (Siegel, 1975)—is enhanced with long interdose intervals (Cox and Tiffany, 1997; Johnson et al., 2002; Tiffany and Maude-Griffin, 1988; Tiffany et al., 1992). Thus, much of the decrease in antinociception in Experiment 1 is probably caused by associative tolerance. Given that the neural mechanisms for associative and nonassociative tolerance are distinct (Grisel et al., 1996), a decrease in hot plate latency resulting from associative tolerance is not necessarily caused by a change in the response of RVM neurons to morphine. If RVM neurons change in response to morphine binding, then increasing the number of injections should enhance the development of tolerance. The data presented here show the opposite. Changes in the RVM are known to occur in animals made tolerant to systemically administered morphine (Bederson et al., 1990; Haghparast et al., 1998; Li and Wang, 2001). The present findings showing a lack of tolerance with direct application of morphine into the RVM suggest that these previously reported changes in the RVM are secondary to changes in other parts of the nervous system. Repeated systemic administration of morphine will produce changes throughout the nervous system that correlate with the development of tolerance. Given that the PAG projects to the RVM (Basbaum and Fields, 1984) and PAG neurons are susceptible to tolerance, any change in the activity of PAG output neurons will be evident in the RVM (Behbehani, 1981; Tortorici and Morgan, 2002; Tortorici et al., 2001). The tolerance that develops to repeated microinjection of morphine into the ventrolateral PAG (Tortorici et al., 1999) could be caused by a change anywhere along the descending system such as the RVM or spinal cord. Our finding that tolerance does not develop to morphine administration into the RVM suggests that tolerance to PAG morphine administration is caused by a change within the PAG and not along the descending system from the RVM to spinal cord. Our recent finding that tolerance does not occur with repeated activation of PAG output neurons (Morgan et al., 2003) is consistent with this interpretation. If tolerance is not caused by repeated activation of PAG output neurons, then the mechanism for tolerance must be located in neurons that precede output neurons. Opioid-sensitive GABAergic neurons that inhibit PAG output neurons are a good candidate. In vitro single cell recordings support this hypothesis. Chronic morphine administration causes
M.M. Morgan et al. / Pain 113 (2005) 91–98
changes in the intracellular signalling in opioid-sensitive GABAergic neurons in the PAG (Chieng and Christie, 1996; Ingram et al., 1998). It is surprising that tolerance develops so much more rapidly with morphine microinjections into the ventrolateral PAG (Tortorici et al., 1999) than the RVM. It is well known that opioids produce antinociception by a direct action on RVM neurons (Heinricher et al., 1992, 1994). The present data indicate that the response of these neurons to repeated morphine administration remains consistent. In contrast, opioid-sensitive neurons in the ventrolateral PAG appear to have a dynamic response to repeated morphine administration. This difference between the PAG and RVM suggests two approaches to improve the treatment of pain: develop treatments that disrupt the molecular changes underlying tolerance in the ventrolateral PAG, or develop treatments that preferentially target the tolerance–resistant neurons in the RVM.
Acknowledgements This investigation was supported in part by funds provided for medical and biological research by the State of Washington Initiative Measure No. 171 and by National Institute on Drug Abuse grant DA015498 to M. Morgan.
References Basbaum AI, Fields HL. Endogenous pain control systems: brainstem spinal pathways and endorphin circuitry. Annu Rev Neurosci 1984;7: 309–38. Bederson JB, Fields HL, Barbaro NM. Hyperalgesia during naloxoneprecipitated withdrawal from morphine is associated with increased oncell activity in the rostral ventromedial medulla. Somatosens Mot Res 1990;7:185–203. Behbehani MM. Effect of chronic morphine treatment on the interaction between the periaqueductal grey and the nucleus raphe magnus of the rat. Neuropharmacology 1981;20:581–6. Chieng B, Christie MJ. Local opioid withdrawal in rat single periaqueductal gray neurons in vitro. J Neurosci 1996;16:7128–36. Cox LS, Tiffany ST. Associative and nonassociative tolerance: the effects of dose and interdose interval. Pharmacol Biochem Behav 1997;57: 31–6. Fields HL, Heinricher MM, Mason P. Neurotransmitters in nociceptive modulatory circuits. Annu Rev Neurosci 1991;14:219–45. Fields HL, Malick A, Burstein R. Dorsal horn projection targets of ON and OFF cells in the rostral ventromedial medulla. J Neurophysiol 1995;74: 1742–59. Gamble GD, Milne RJ. Repeated exposure to sham testing procedures reduces reflex withdrawal and hot-plate latencies: attenuation of tonic descending inhibition? Neurosci Lett 1989;96:312–7. Gebhart GF, Toleikis JR. An evaluation of stimulation-produced analgesia in the cat. Exp Neurol 1978;62:570–9. Grisel JE, Watkins LR, Maier SF. Associative and non-associative mechanisms of morphine analgesic tolerance are neurochemically distinct in the rat spinal cord. Psychopharmacology 1996;128:248–55. Haghparast A, Semnanian S, Fathollahi Y. Morphine tolerance and dependence in the nucleus paragigantocellularis: Single unit recording study in vivo. Brain Res 1998;814:71–7.
97
Heinricher MM, Morgan MM, Fields HL. Direct and indirect actions of morphine on medullary neurons that modulate nociception. Neuroscience 1992;48:533–43. Heinricher MM, Morgan MM, Tortorici V, Fields HL. Disinhibition of offcells and antinociception produced by an opioid action within the rostral ventromedial medulla. Neuroscience 1994;63:279–88. Heinricher MM, McGaraughty S, Grandy DK. Circuitry underlying antiopioid actions of orphanin FQ in the rostral ventromedial medulla. J Neurophysiol 1997;78:3351–8. Ingram SL, Vaughan CW, Bagley EE, Connor M, Christie MJ. Enhanced opioid efficacy in opioid dependence is caused by an altered signal transduction pathway. J Neurosci 1998;18:10269–76. Jacquet YF, Lajtha A. Paradoxical effects after microinjection of morphine in the periaqueductal gray matter in the rat. Science 1974;185:1055–7. Jacquet YF, Lajtha A. The periaqueductal gray: site of morphine analgesia and tolerance as shown by 2-way cross tolerance between systemic and intracerebral injections. Brain Res 1976;103:501–13. Jensen TS, Yaksh TL. Comparison of antinociceptive action of morphine in the periaqueductal gray, medial and paramedial medulla in rat. Brain Res 1986;363:99–113. Johnson AS, Lynch WC, Block RA, Schwartz EA. Associative tolerance in the rat: a meta-analytic review. Soc Neurosci Abstr 2002;32:586.14. Lewis VA, Gebhart GF. Morphine-induced and stimulation-produced analgesias at coincident periaqueductal central gray loci: evaluation of analgesic congruence, tolerance, and cross-tolerance. Exp Neurol 1977; 57:934–55. Li AH, Wang HL. G protein-coupled receptor kinase 2 mediates mu-opioid receptor desensitization in GABAergic neurons of the nucleus raphe magnus. J Neurochem 2001;77:435–44. Mayer DJ, Hayes RL. Stimulation-produced analgesia: development of tolerance and cross-tolerance to morphine. Science 1975;188:941–3. Millan MJ, Czlonkowski A, Herz A. An analysis of the tolerance which develops to analgetic electrical stimulation of the midbrain periaqueductal grey in freely moving rats. Brain Res 1987;435:97–111. Milne RJ, Gamble GD, Holford NHG. Behavioural tolerance to morphine analgesia is supraspinally mediated: a quantitative analysis of dose– response relationships. Brain Res 1989;491:316–27. Morgan MM, Liebeskind JC. Site specificity in the development of tolerance to stimulation-produced analgesia from the periaqueductal gray matter of the rat. Brain Res 1987;425:356–9. Morgan MM, Whitney PK. Immobility accompanies the antinociception mediated by the rostral ventromedial medulla of the rat. Brain Res 2000;872:276–81. Morgan MM, Whitney PK, Gold MS. Immobility and flight associated with antinociception produced by activation of the ventral and lateral/dorsal regions of the rat periaqueductal gray. Brain Res 1998;804:159–66. Morgan MM, Clayton CC, Lane DA. Behavioral evidence linking opioidsensitive GABAergic neurons in the ventrolateral periaqueductal gray to morphine tolerance. Neuroscience 2003;118:227–32. Oliveras JL, Hosobuchi Y, Guilbaud G, Besson JM. Analgesic electrical stimulation of the feline nucleus raphe magnus: development of tolerance and its reversal by 5-HTP. Brain Res 1978;146:404–9. Paxinos G, Watson C. The rat brain in stereotaxic coordinates. Sydney: Academic Press; 1986. Proudfit HK. Reversible inactivation of raphe magnus neurons: effects on nociceptive threshold and morphine-induced analgesia. Brain Res 1980; 201:459–64. Siegel S. Evidence from rats that morphine tolerance is a learned response. J Comp Physiol Psychol 1975;89:498–506. Siuciak JA, Advokat C. Tolerance to morphine microinjections in the periaqueductal gray (PAG) induces tolerance to systemic, but not intrathecal morphine. Brain Res 1987;424:311–9. Tiffany ST, Maude-Griffin PM. Tolerance to morphine in the rat: associative and nonassociative effects. Behav Neurosci 1988;102:534–43. Tiffany ST, Drobes DJ, Cepeda-Benito A. Contribution of associative and nonassociative processes to the development of morphine tolerance. Psychopharmacology 1992;109:185–90.
98
M.M. Morgan et al. / Pain 113 (2005) 91–98
Tortorici V, Morgan MM. Comparison of morphine and kainic acid microinjections into identical PAG sites on the activity of RVM neurons. J Neurophysiol 2002;88:1707–15. Tortorici V, Robbins CS, Morgan MM. Tolerance to the antinociceptive effect of morphine microinjections into the ventral, but not lateraldorsal PAG of the rat. Behav Neurosci 1999;113:833–9. Tortorici V, Morgan MM, Vanegas H. Tolerance to repeated microinjection of morphine into the periaqueductal gray is associated with changes in the behavior of off- and on-cells in the rostral ventromedial medulla of rats. Pain 2001;89:237–44.
Vanderah TW, Suenaga NM, Ossipov MH, Malan TPJ, Lai J, Porreca P. Tonic descending facilitation from the rostral ventromedial medulla mediates opioid-induced abnormal pain and antinociceptive tolerance. J Neurosci 2001;21:279–86. Yaksh TL, Yeung JC, Rudy TA. Systematic examination in the rat of brain sites sensitive to the direct application of morphine: observation of differential effects within the periaqueductal gray. Brain Res 1976;114:83–103. Young EG, Watkins LR, Mayer DJ. Comparison of the effects of ventral medullary lesions on systemic and microinjection morphine analgesia. Brain Res 1984;290:119–29.
Differential susceptibility of the PAG and RVM to tolerance to the antinociceptive effect of morphine in the rat Michael M. Morgan*, Cecilea C. Clayton, Jill S. Boyer-Quick Department of Psychology, Washington State University, 14204 NE Salmon Creek Avenue, Vancouver, WA 98686, USA Received 10 June 2004; received in revised form 8 September 2004; accepted 28 September 2004
Abstract The periaqueductal gray (PAG) and rostral ventromedial medulla (RVM) are part of a nociceptive modulatory system. Microinjection of morphine into either structure produces antinociception. Tolerance develops to ventrolateral PAG mediated antinociception with repeated microinjection of morphine. In contrast, there are no published reports of tolerance to morphine administration into the RVM. Three experiments were conducted to determine whether tolerance develops to morphine microinjections into the RVM. Experiment 1 compared tolerance to the antinociceptive effect of microinjecting morphine (5 mg/0.5 ml) into the PAG and RVM following daily injections for four consecutive days. Experiment 2 assessed tolerance to a range of morphine doses (2.5–20 mg) after injecting morphine into the RVM twice a day for two consecutive days. Experiment 3 followed a similar procedure except twice as many RVM injections were made (8 microinjections in 4 days). The degree to which tolerance developed to the antinociceptive effect of morphine was much greater with microinjections into the PAG compared to the RVM. There was a 64% drop in hot plate latency from the first to the fifth injection of morphine into the PAG, but only a 36% drop in latency following RVM microinjections. Reducing the interdose interval to two injections a day or increasing the total number of injections from 4 to 8 did not enhance the development of tolerance to RVM morphine administration. These data demonstrate that opioid-sensitive neurons in the RVM are relatively resistant to the development of tolerance compared to PAG neurons. q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. Keywords: Opiate
Opiates such as morphine are the most effective treatment for pain. Unfortunately, morphine can be rendered ineffective with repeated administration because of the development of tolerance. Although numerous mechanisms for tolerance have been proposed, a clear understanding of the neural changes responsible for tolerance is unclear. An important step in understanding the mechanisms underlying tolerance is to identify the neural structures mediating tolerance. Opiates produce antinociception, at least in part, by activating a descending modulatory system that includes the periaqueductal gray (PAG) and rostral ventromedial medulla (RVM) (Basbaum and Fields, 1984; Fields et al., * Corresponding author. Tel.: C1 360 546 9726; fax: C1 360 546 9038. E-mail address: [email protected] (M.M. Morgan).
1991, 1995). Microinjection of morphine into either the PAG or RVM produces antinociception (Jacquet and Lajtha, 1974; Jensen and Yaksh, 1986; Morgan et al., 1998; Yaksh et al., 1976), and inactivation of the RVM attenuates the antinociceptive effect of systemically administered morphine (Heinricher et al., 1997; Proudfit, 1980; Young et al., 1984). The important role that the PAG and RVM play in morphine antinociception make these structures potential sites where tolerance to morphine may occur. A number of studies have shown that tolerance develops to repeated microinjections of morphine into the PAG (Jacquet and Lajtha, 1976; Lewis and Gebhart, 1977; Siuciak and Advokat, 1987; Tortorici et al., 1999, 2001). In contrast, there are no published reports of tolerance to the antinociceptive effect of microinjecting morphine into the RVM. Given the prominent role of the RVM in modulating
0304-3959/$20.00 q 2004 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2004.09.039
92
M.M. Morgan et al. / Pain 113 (2005) 91–98
nociception, it is important to determine whether this structure contributes to morphine tolerance. Antinociception has been shown to be reduced with repeated electrical stimulation of the RVM (Oliveras et al., 1978). In addition, chronic systemic administration of morphine produces changes in the activity of neurons in and around the RVM (Bederson et al., 1990; Haghparast et al., 1998; Li and Wang, 2001), and inactivation of the RVM attenuates tolerance to chronic systemic morphine administration (Vanderah et al., 2001). Although these studies indicate that RVM neurons contribute to the development of tolerance to morphine, the changes in RVM neurons that occur as tolerance develops could be secondary to changes in other structures such as the ventrolateral PAG (Behbehani, 1981; Tortorici et al., 2001). That is, changes in RVM neurons could correlate with the development of tolerance without causing tolerance. The objective of this paper is to test the hypothesis that tolerance will occur to the antinociceptive effect of direct administration of morphine into the RVM.
1. Materials and methods 1.1. Subjects Male Sprague–Dawley rats (280–380 g; Animal Technologies Inc., Kent, WA) were anesthetised with sodium pentobarbital (Nembutal; 60 mg/kg, i.p.) and stereotaxically implanted with a 23 gauge stainless-steel guide cannula aimed at the PAG (9 mm guide; AP 1.2 mm, ML 0.6 mm; DV K5.0 mm) or RVM (15 mm long; AP K2.2 mm; ML 0.0 mm, DV K8.8 mm from lambda). The guide cannula was held in place with dental cement affixed to two screws in the skull. A removable 31-gauge stainless-steel stylet was inserted into the guide cannula following surgery. The rat was allowed to recover for 1 week prior to testing. Animals were housed individually following surgery with food and water available ad libitum. Lights were maintained on a reverse 12 h light/dark cycle. All injections and testing were conducted during the dark phase when rats are awake and active. Experiments were conducted in accordance with the animal care and use guidelines of the Committee for Research and Ethical Issues of IASP. This research was approved by the Institutional Animal Care and Use Committee at Washington State University. Rats were handled daily following surgery. In addition, each rat was habituated to the injection procedure prior to the first day of testing by inserting an injection cannula through the guide in the absence of drug administration. This procedure was used to reduce non-specific effects resulting from mechanical damage to neurons on the first test day. Microinjections of morphine sulfate (a gift from the National Institute on Drug Abuse) and saline were made into the PAG or RVM through a 31-gauge injection cannula that extended 2 mm beyond the tip of the guide cannula. The injection cannula was connected to a 1 ml syringe (Hamilton Co., Reno, NV) with PE20 tubing filled with sterile water. All microinjections were given in a volume of 0.5 ml over 50 s while the rat was gently restrained. The injection cannula remained in place for an additional 20 s to
minimize backflow of the drug up the cannula tract. Following the injection, the stylet was reinserted into the guide cannula and the rat was returned to its home cage. 1.2. Behavioral assessment Nociception was assessed using the hot plate test (IITC, Woodland Hills, CA). The hot plate test measures the latency to lick a hindpaw when placed on a 51–52 8C metal surface. The rat was removed from the hot plate immediately following a response or after 40 s if no response occurred. Previous research has shown a clear tolerance to the antinociceptive effects of morphine using this test (Tortorici et al., 1999). The number of tests was kept to a minimum (a single test on the first and last trials) in the first experiment to avoid changes in latency associated with repeated testing (Gamble and Milne, 1989). 1.3. Histology Rats were given a lethal injection of pentobarbital (120 mg/kg, i.p.) immediately following testing. The microinjection site was marked by injecting cresyl violet into the same site as the drugs were administered previously. The brain was removed and placed in formalin (10%). At least 3 days later, the brain was sectioned coronally (50 mm) and viewed under a microscope to localize the injection site (Paxinos and Watson, 1986). 1.4. Data analysis Students’ t-test and analysis of variance were used to compare differences in hot plate latency between groups. Post-hoc analysis of within group changes from the first to last trial was conducted by calculating the 95% confidence limits for specific means. 1.5. Experiment 1: comparison of tolerance to morphine injections into the PAG and RVM The objective of this experiment was to compare the magnitude of tolerance to the antinociceptive effect of morphine microinjection into the PAG and RVM. Rats were implanted with a guide cannula aimed at the PAG or RVM. Morphine (5 mg/0.5 ml) was injected into the PAG or RVM once a day for four consecutive days. Saline was injected into the RVM on Days 1–4 as a control. All rats received a microinjection of morphine on Day 5 to determine whether tolerance had developed. Nociception was assessed using the hot plate test 30 min after the injection on Days 1 and 5. Rats were returned to their home cage without nociceptive testing following microinjections on Days 2–4 to minimize potential changes in latency caused by repeated testing (Gamble and Milne, 1989). Given that the PAG and RVM are part of the same nociceptive modulatory system, it is hypothesised that tolerance will develop at the same rate in both structures. 1.6. Experiment 2: analysis of interdose interval on tolerance to RVM morphine microinjections Morphine was injected once a day for 5 days in Experiment 1. The objective of Experiment 2 was to determine whether increasing the exposure of RVM neurons to morphine by shortening the interdose interval to two injections a day for 2
M.M. Morgan et al. / Pain 113 (2005) 91–98
93
days would enhance the development of tolerance to the antinociceptive effect of morphine. Rats received the same number of injections as in Experiment 1 but over 2 days instead of four. Rats were randomly assigned to a morphine or saline pretreatment group. The morphine pretreatment group was injected with morphine (5 mg/0.5 ml) into the RVM twice a day (9:00 AM and 4:00 PM) for two consecutive days. The saline pretreatment group was injected with saline at the same times to control for the effects of repeated microinjections. Nociception was assessed using the hot plate test 30 min after each injection. On Day 3 (Trial 5), these morphine and saline pretreated rats were injected with one of five morphine doses (0, 2.5, 5, 10, and 20 mg/0.5 ml). Given that the procedure described above produces tolerance when morphine is injected into the ventrolateral PAG (Tortorici et al., 1999), we hypothesised that tolerance to RVM morphine administration will develop more rapidly using this procedure than with once daily microinjections. A range of doses was tested to determine whether morphine pretreatment causes a change in potency. 1.7. Experiment 3: effect of doubling the number of injections on tolerance to RVM morphine administration Morphine was administered four times to induce tolerance in Experiments 1 (once daily for 4 days) and 2 (twice daily for 2 days). The present experiment doubled the number of morphine injections by combining these two approaches (two daily injections for 4 days: 9:00 AM and 4:00 PM). Doubling the number of injections would be expected to enhance the development of tolerance. Rats were tested as in Experiment 2 except that rats received eight injections of morphine (5 or 10 mg/0.5 ml) or saline into the RVM prior to the final test. Nociception was assessed only after the morning injections so the number of hot plate tests did not differ from Experiment 2. Both morphine and saline pretreated rats were injected with morphine on Day 5 (Trial 9) to determine whether tolerance had developed. The same dose was used for induction (Trials 1–8) and assessment of tolerance (Trial 9). If RVM neurons are susceptible to morphine tolerance, then tolerance should be greater in rats receiving twice as many morphine injections.
2. Results 2.1. Experiment 1: comparison of tolerance in the PAG and RVM Only rats with microinjections into the ventrolateral PAG or RVM (nucleus raphe magnus and adjacent nucleus reticularis gigantocellularis pars alpha) were included in data analysis (Fig. 1). These data consisted of 10 rats microinjected with morphine into the PAG, 16 with morphine microinjected into the RVM, and eight with saline microinjected into the RVM. Given that susceptibility to morphine tolerance has been shown to be site specific within the PAG (Tortorici et al., 1999), injection sites outside the boundaries of the ventrolateral PAG and RVM were not included in data analysis. Most of these misplaced injections were just outside the border, and microinjection
Fig. 1. Location of microinjection sites in the ventrolateral PAG (top) and RVM (bottom) for Experiments 1–3. Only morphine and saline injections within the shaded regions were included in data analysis. Coronal sections are taken from the Atlas of Paxinos and Watson (1986). The number in the top left corner refers to the distance from the interaurel line. All injections fell within G0.5 mm of these coronal planes.
of morphine into these sites produced antinociception in many cases. Microinjection of morphine produced an increase in hot plate latency following PAG and RVM administration compared to saline treated rats on Trial 1 (Fig. 2). As reported previously (Morgan and Whitney, 2000; Tortorici and Morgan, 2002), the antinociceptive effect of morphine was greater following PAG compared to RVM administration. Microinjection of morphine had the opposite effect on Trial 5. Microinjection of morphine into the RVM produced greater antinociception than microinjection into the PAG (Trial 5 in Fig. 2). There was a mean decrease in hot plate latency from Trial 1–5 following daily morphine administration into the PAG of 20.3G4.3 s. The mean decrease in hot plate latency with repeated microinjection of morphine into the RVM was only 9.9G3.2 s. This decrease in hot plate latency across trials was statistically significant (F(1,24)Z32.27, p!0.05). The degree to which tolerance developed following PAG microinjections of morphine was greater than with RVM microinjections. This difference was evident by a nearly significant group (PAG vs. RVM) by trial interaction (F(1,24)Z3.78, PZ0.06). Repeated microinjections of saline into the RVM did not prevent morphine
94
M.M. Morgan et al. / Pain 113 (2005) 91–98
Fig. 2. Comparison of tolerance to the antinociceptive effect of microinjecting morphine into the PAG and RVM. Microinjection of morphine (5 mg/0.5 ml) into the PAG and RVM produced an increase in hot plate latency compared to rats injected with saline into the RVM on Trial 1. All rats were injected with morphine on Trial 5. Hot plate latency was greatly reduced in morphine-pretreated rats compared to Trial 1. A 64% decrease in hot plate latency from Trial 1–5 was evident with PAG administration. A decrease of only 36% occurred with RVM administration. Repeated saline injections into the RVM did not disrupt the antinociceptive effect of morphine administration on Trial 5. These data demonstrate that the ventrolateral PAG is more susceptible to morphine tolerance than the RVM.
from producing antinociception when administered for the first time on Trial 5. 2.2. Experiment 2: analysis of interdose interval on tolerance to RVM morphine microinjections This experiment assessed whether a decrease in the interdose interval would enhance the development of tolerance to RVM morphine microinjection. Only data from injection sites located in the RVM were included in the statistical analysis (Fig. 1). This included 41 rats pretreated with microinjections of morphine (5 mg/0.5 ml) and 28 pretreated with microinjections of saline. These large groups were subdivided for dose–response analysis on Day 3 (Trial 5). MeanGSEM hot plate latency for saline treated rats was 10.7G0.7 s on Trial 1. None of these rats had a hot plate latency greater than 20 s. In contrast, microinjection of morphine into the RVM produced an increase in hot plate latency of 20 s or greater in 23 of the 41 rats (56%). The finding that only a subset of rats had hot plate latencies greater than 20 s following microinjection of morphine into the RVM is consistent with previous research (Morgan and Whitney, 2000; Tortorici and Morgan, 2002). The number of rats with a hot plate latency greater than 20 s dropped on each subsequent trial (46% on Trial 2; 40% on Trial 3; and 37% on Trial 4). Although this decrease suggests tolerance may occur in a subset of rats, analysis of morphine
antinociception in saline pretreated rats suggests this decrease is caused by behavioral and associative tolerance, not from a direct action of morphine on RVM neurons (see analysis of mean hot plate latency below). Microinjection of morphine into the RVM produced antinociception on each of the four pretreatment trials compared to saline treated controls (Fig. 3). Analysis of variance revealed that the hot plate latency for rats injected with morphine was significantly higher than the latency for saline treated rats across all trials (F(1,65)Z297.87, P!0.05). Moreover, there was no group by trial interaction (F(1,195)Z1.085, n.s.), an effect that would be expected if there had been a reduction in antinociception across trials. Although a small decrease in mean hot plate latencies across trials was evident following both morphine and saline administration on Trials 1–4, statistical support for tolerance to RVM morphine administration did not occur. The decrease in hot plate latency across trials occurred in both the saline and morphine treated rats indicating it was caused by repeated testing (Gamble and Milne, 1989; Milne et al., 1989), not tolerance to morphine administration. Morphine and saline pretreated rats were injected with a range of morphine doses on Trial 5 (0, 2.5, 5, 10, and 20 mg/0.5 ml; NZ5–12 rats/condition). Microinjection of morphine produced a dose dependent increase in hot plate latency in both morphine and saline pretreated rats (F(4,58)Z2.997, P!0.05; Fig. 4). Morphine pretreatment did not alter the potency for antinociception produced by microinjecting morphine into the RVM. In fact, the ED50 for antinociception following morphine microinjection was lower in morphine (5.1 mg) compared to saline (5.9 mg) pretreated rats.
Fig. 3. Effect of repeated microinjections of morphine and saline into the RVM on hot plate latency across trials. Rats were injected with morphine (5 mg/0.5 ml) or saline twice a day for two consecutive days. The elevation in hot plate latency across all four trials in morphine treated rats indicates that RVM neurons are relatively insensitive to the development of tolerance. The slight decrease in hot plate latency across trials is evident in both morphine and saline pretreated groups, and suggests that this decrease is caused by repeated nociceptive testing (Gamble and Milne, 1989).
M.M. Morgan et al. / Pain 113 (2005) 91–98
Fig. 4. Comparison of the antinociceptive effect of microinjecting morphine into the RVM of rats pretreated with morphine or saline. Microinjection of morphine into the RVM produced a dose-related increase in hot plate latency in both morphine and saline pretreated rats. Rats were pretreated with morphine (5 mg/0.5 ml) or saline twice a day for two consecutive days. Both groups were injected with morphine (0, 2.5, 5, 10, or 20 mg/0.5 ml) into the RVM on Day 3. Tolerance should cause a rightward shift in the dose– response curve of morphine-pretreated rats. This did not happen suggesting that RVM neurons are resistant to morphine tolerance.
2.3. Experiment 3: effect of doubling the number of injections on tolerance to RVM morphine administration The relatively small decrease in antinociception with repeated morphine administration in Experiments 1 and 2 suggests that four injections may not be sufficient to induce tolerance. If this is the case, then doubling the number of morphine injections should double the magnitude of tolerance. The present experiment tested this hypothesis by injecting morphine twice a day for 4 days. All rats were injected with morphine 3 days prior to the tolerance induction procedure to determine whether the injection site supported antinociception. Only injection sites that were located in the RVM and produced antinociception to morphine administration on the pretest were included in data analysis. Given the relatively low rate of antinociception following RVM morphine administration, screening rats for antinociception allowed for a more focused analysis of changes in nociception with repeated morphine administration. Microinjection of morphine into the RVM produced a significant increase in hot plate latency in morphine treated rats (NZ11) compared to saline treated controls (NZ7) on Trial 1 (F(1,16)Z84.70; P!0.05). Although there was a slight decrease in hot plate latency across the eight trials, as in Experiment 2 this decrease was not statistically significant (F(3,48)Z1.677, n.s.). Moreover, there was no difference in antinociception between morphine and saline pretreated groups following morphine administration on Trial 9 (t(16)Z0.160, n.s.; Fig. 5). In addition, there was no group by trial interaction as would be expected if tolerance had occurred in the morphine treated group (F(3,48)Z0.919, n.s.). These data indicate that tolerance to
95
Fig. 5. Lack of tolerance following eight microinjections of morphine into the RVM. Rats were injected with morphine (5–10 mg/0.5 ml) or saline twice a day for four consecutive days. The first injection of morphine into the RVM produced antinociception compared to saline treated controls (Trial 1). Microinjection of morphine into the RVM of both morphine and saline pretreated rats produced antinociception on Trial 9. The antinociceptive effect was identical despite that this was the ninth injection of morphine for the morphine-pretreated group and the first injection of morphine for the saline pretreated group.
morphine did not occur despite eight prior microinjections of morphine into the RVM.
3. Discussion The present data suggest that RVM neurons are relatively resistant to the development of tolerance to the antinociceptive effect of morphine. Daily administration of morphine into the PAG produced a 64% drop in antinociception from the first to fifth injection, whereas morphine microinjections into the RVM produced a drop in antinociception of only 36% (Experiment 1). The decrease in antinociception was smaller when morphine was administered with a short interdose interval (two injections a day; Experiment 2) or when the number of injections was increased from four to eight (Experiment 3)—procedural changes that would be expected to enhance the development of tolerance to RVM morphine administration. We previously have shown that a profound tolerance develops with as little as two injections of morphine a day for 2 days when the PAG is the target (Tortorici et al., 1999). The present data show that neurons in the RVM are relatively resistant to the development of tolerance compared to neurons in the ventrolateral PAG. A lack of tolerance could simply indicate that the procedure used to induce tolerance is inadequate. Tolerance may require an increase in the number of morphine injections, an increase in the dose of morphine, an increase in the microinjection volume, a reduction in the time between
96
M.M. Morgan et al. / Pain 113 (2005) 91–98
injections, etc. Although it is impossible to test all possible factors in a single manuscript, the present findings demonstrate that shortening the interdose interval or increasing the number of injections did not enhance the development of tolerance to morphine microinjection into the RVM. The relatively stable antinociception produced by repeated microinjections of morphine into the RVM is surprising. The antinociceptive effect of microinjecting morphine into the PAG has been shown to be diminished by a single prior injection 3 days earlier (Lewis and Gebhart, 1977). Although relatively little tolerance was evident with daily microinjections of morphine into the RVM, this procedure was sufficient to produce tolerance when microinjections were directed at the PAG. The magnitude of tolerance to PAG administration was consistent whether morphine was injected into the ventrolateral PAG once (Experiment 1) or twice a day (Tortorici et al., 1999). Given that the same procedures were used for RVM administration, the present data indicate that RVM neurons are relatively resistant to the development of tolerance to the antinociceptive effects of morphine. The present paper appears to be the first published report examining tolerance to direct administration of morphine into the RVM. A previous study reported tolerance to repeated electrical stimulation of the RVM in the cat (Oliveras et al., 1978). This tolerance occurred with as little as 6 min of stimulation spread over 13 min. Whether this is tolerance or an artifact of electrical stimulation (e.g. neuronal fatigue) is debatable. In contrast, there have been numerous reports of tolerance to morphine microinjection into the PAG (Jacquet and Lajtha, 1976; Lewis and Gebhart, 1977; Siuciak and Advokat, 1987; Tortorici et al., 1999, 2001). Five additional studies show tolerance to antinociception produced by electrical stimulation of the PAG (Gebhart and Toleikis, 1978; Lewis and Gebhart, 1977; Mayer and Hayes, 1975; Millan et al., 1987; Morgan and Liebeskind, 1987). Given the important role of the RVM in modulating nociception, the lack of published reports of tolerance to the antinociceptive effect of morphine administered into this structure is surprising. This lack of data makes the present finding that the RVM is relatively resistant to the development of tolerance to morphine administration especially important. Although the decrease in antinociception across trials in all three experiments looks like tolerance, this decrease in hot plate latency is small. A similar decrease in hot plate latency across trials occurred in saline treated rats suggesting that repeated testing, not morphine administration, caused this change. This decrease in hot plate latency with repeated testing is called behavioral tolerance and can occur independent of morphine administration (Gamble and Milne, 1989). In addition, a reduction in morphine antinociception can be caused by neural damage following repeatedly inserting a microinjection cannula into the brain. This damage appears to have only a minor impact because microinjection of morphine into the RVM produced
antinociception in rats pretreated with multiple saline injections (see Figs. 2, 4, and 5). The most surprising finding is that the greatest decrease in morphine antinociception occurred with the fewest injections. That is, daily injections of morphine into the RVM caused a 36% decrease in hot plate latency in Experiment 1, whereas increasing the number of morphine pretreatments from four to eight resulted in a decrease in hot plate latency from Trial 1–9 of only 18% (Experiment 3). Previous research has shown that associative tolerance—tolerance caused by cues that predict drug administration (Siegel, 1975)—is enhanced with long interdose intervals (Cox and Tiffany, 1997; Johnson et al., 2002; Tiffany and Maude-Griffin, 1988; Tiffany et al., 1992). Thus, much of the decrease in antinociception in Experiment 1 is probably caused by associative tolerance. Given that the neural mechanisms for associative and nonassociative tolerance are distinct (Grisel et al., 1996), a decrease in hot plate latency resulting from associative tolerance is not necessarily caused by a change in the response of RVM neurons to morphine. If RVM neurons change in response to morphine binding, then increasing the number of injections should enhance the development of tolerance. The data presented here show the opposite. Changes in the RVM are known to occur in animals made tolerant to systemically administered morphine (Bederson et al., 1990; Haghparast et al., 1998; Li and Wang, 2001). The present findings showing a lack of tolerance with direct application of morphine into the RVM suggest that these previously reported changes in the RVM are secondary to changes in other parts of the nervous system. Repeated systemic administration of morphine will produce changes throughout the nervous system that correlate with the development of tolerance. Given that the PAG projects to the RVM (Basbaum and Fields, 1984) and PAG neurons are susceptible to tolerance, any change in the activity of PAG output neurons will be evident in the RVM (Behbehani, 1981; Tortorici and Morgan, 2002; Tortorici et al., 2001). The tolerance that develops to repeated microinjection of morphine into the ventrolateral PAG (Tortorici et al., 1999) could be caused by a change anywhere along the descending system such as the RVM or spinal cord. Our finding that tolerance does not develop to morphine administration into the RVM suggests that tolerance to PAG morphine administration is caused by a change within the PAG and not along the descending system from the RVM to spinal cord. Our recent finding that tolerance does not occur with repeated activation of PAG output neurons (Morgan et al., 2003) is consistent with this interpretation. If tolerance is not caused by repeated activation of PAG output neurons, then the mechanism for tolerance must be located in neurons that precede output neurons. Opioid-sensitive GABAergic neurons that inhibit PAG output neurons are a good candidate. In vitro single cell recordings support this hypothesis. Chronic morphine administration causes
M.M. Morgan et al. / Pain 113 (2005) 91–98
changes in the intracellular signalling in opioid-sensitive GABAergic neurons in the PAG (Chieng and Christie, 1996; Ingram et al., 1998). It is surprising that tolerance develops so much more rapidly with morphine microinjections into the ventrolateral PAG (Tortorici et al., 1999) than the RVM. It is well known that opioids produce antinociception by a direct action on RVM neurons (Heinricher et al., 1992, 1994). The present data indicate that the response of these neurons to repeated morphine administration remains consistent. In contrast, opioid-sensitive neurons in the ventrolateral PAG appear to have a dynamic response to repeated morphine administration. This difference between the PAG and RVM suggests two approaches to improve the treatment of pain: develop treatments that disrupt the molecular changes underlying tolerance in the ventrolateral PAG, or develop treatments that preferentially target the tolerance–resistant neurons in the RVM.
Acknowledgements This investigation was supported in part by funds provided for medical and biological research by the State of Washington Initiative Measure No. 171 and by National Institute on Drug Abuse grant DA015498 to M. Morgan.
References Basbaum AI, Fields HL. Endogenous pain control systems: brainstem spinal pathways and endorphin circuitry. Annu Rev Neurosci 1984;7: 309–38. Bederson JB, Fields HL, Barbaro NM. Hyperalgesia during naloxoneprecipitated withdrawal from morphine is associated with increased oncell activity in the rostral ventromedial medulla. Somatosens Mot Res 1990;7:185–203. Behbehani MM. Effect of chronic morphine treatment on the interaction between the periaqueductal grey and the nucleus raphe magnus of the rat. Neuropharmacology 1981;20:581–6. Chieng B, Christie MJ. Local opioid withdrawal in rat single periaqueductal gray neurons in vitro. J Neurosci 1996;16:7128–36. Cox LS, Tiffany ST. Associative and nonassociative tolerance: the effects of dose and interdose interval. Pharmacol Biochem Behav 1997;57: 31–6. Fields HL, Heinricher MM, Mason P. Neurotransmitters in nociceptive modulatory circuits. Annu Rev Neurosci 1991;14:219–45. Fields HL, Malick A, Burstein R. Dorsal horn projection targets of ON and OFF cells in the rostral ventromedial medulla. J Neurophysiol 1995;74: 1742–59. Gamble GD, Milne RJ. Repeated exposure to sham testing procedures reduces reflex withdrawal and hot-plate latencies: attenuation of tonic descending inhibition? Neurosci Lett 1989;96:312–7. Gebhart GF, Toleikis JR. An evaluation of stimulation-produced analgesia in the cat. Exp Neurol 1978;62:570–9. Grisel JE, Watkins LR, Maier SF. Associative and non-associative mechanisms of morphine analgesic tolerance are neurochemically distinct in the rat spinal cord. Psychopharmacology 1996;128:248–55. Haghparast A, Semnanian S, Fathollahi Y. Morphine tolerance and dependence in the nucleus paragigantocellularis: Single unit recording study in vivo. Brain Res 1998;814:71–7.
97
Heinricher MM, Morgan MM, Fields HL. Direct and indirect actions of morphine on medullary neurons that modulate nociception. Neuroscience 1992;48:533–43. Heinricher MM, Morgan MM, Tortorici V, Fields HL. Disinhibition of offcells and antinociception produced by an opioid action within the rostral ventromedial medulla. Neuroscience 1994;63:279–88. Heinricher MM, McGaraughty S, Grandy DK. Circuitry underlying antiopioid actions of orphanin FQ in the rostral ventromedial medulla. J Neurophysiol 1997;78:3351–8. Ingram SL, Vaughan CW, Bagley EE, Connor M, Christie MJ. Enhanced opioid efficacy in opioid dependence is caused by an altered signal transduction pathway. J Neurosci 1998;18:10269–76. Jacquet YF, Lajtha A. Paradoxical effects after microinjection of morphine in the periaqueductal gray matter in the rat. Science 1974;185:1055–7. Jacquet YF, Lajtha A. The periaqueductal gray: site of morphine analgesia and tolerance as shown by 2-way cross tolerance between systemic and intracerebral injections. Brain Res 1976;103:501–13. Jensen TS, Yaksh TL. Comparison of antinociceptive action of morphine in the periaqueductal gray, medial and paramedial medulla in rat. Brain Res 1986;363:99–113. Johnson AS, Lynch WC, Block RA, Schwartz EA. Associative tolerance in the rat: a meta-analytic review. Soc Neurosci Abstr 2002;32:586.14. Lewis VA, Gebhart GF. Morphine-induced and stimulation-produced analgesias at coincident periaqueductal central gray loci: evaluation of analgesic congruence, tolerance, and cross-tolerance. Exp Neurol 1977; 57:934–55. Li AH, Wang HL. G protein-coupled receptor kinase 2 mediates mu-opioid receptor desensitization in GABAergic neurons of the nucleus raphe magnus. J Neurochem 2001;77:435–44. Mayer DJ, Hayes RL. Stimulation-produced analgesia: development of tolerance and cross-tolerance to morphine. Science 1975;188:941–3. Millan MJ, Czlonkowski A, Herz A. An analysis of the tolerance which develops to analgetic electrical stimulation of the midbrain periaqueductal grey in freely moving rats. Brain Res 1987;435:97–111. Milne RJ, Gamble GD, Holford NHG. Behavioural tolerance to morphine analgesia is supraspinally mediated: a quantitative analysis of dose– response relationships. Brain Res 1989;491:316–27. Morgan MM, Liebeskind JC. Site specificity in the development of tolerance to stimulation-produced analgesia from the periaqueductal gray matter of the rat. Brain Res 1987;425:356–9. Morgan MM, Whitney PK. Immobility accompanies the antinociception mediated by the rostral ventromedial medulla of the rat. Brain Res 2000;872:276–81. Morgan MM, Whitney PK, Gold MS. Immobility and flight associated with antinociception produced by activation of the ventral and lateral/dorsal regions of the rat periaqueductal gray. Brain Res 1998;804:159–66. Morgan MM, Clayton CC, Lane DA. Behavioral evidence linking opioidsensitive GABAergic neurons in the ventrolateral periaqueductal gray to morphine tolerance. Neuroscience 2003;118:227–32. Oliveras JL, Hosobuchi Y, Guilbaud G, Besson JM. Analgesic electrical stimulation of the feline nucleus raphe magnus: development of tolerance and its reversal by 5-HTP. Brain Res 1978;146:404–9. Paxinos G, Watson C. The rat brain in stereotaxic coordinates. Sydney: Academic Press; 1986. Proudfit HK. Reversible inactivation of raphe magnus neurons: effects on nociceptive threshold and morphine-induced analgesia. Brain Res 1980; 201:459–64. Siegel S. Evidence from rats that morphine tolerance is a learned response. J Comp Physiol Psychol 1975;89:498–506. Siuciak JA, Advokat C. Tolerance to morphine microinjections in the periaqueductal gray (PAG) induces tolerance to systemic, but not intrathecal morphine. Brain Res 1987;424:311–9. Tiffany ST, Maude-Griffin PM. Tolerance to morphine in the rat: associative and nonassociative effects. Behav Neurosci 1988;102:534–43. Tiffany ST, Drobes DJ, Cepeda-Benito A. Contribution of associative and nonassociative processes to the development of morphine tolerance. Psychopharmacology 1992;109:185–90.
98
M.M. Morgan et al. / Pain 113 (2005) 91–98
Tortorici V, Morgan MM. Comparison of morphine and kainic acid microinjections into identical PAG sites on the activity of RVM neurons. J Neurophysiol 2002;88:1707–15. Tortorici V, Robbins CS, Morgan MM. Tolerance to the antinociceptive effect of morphine microinjections into the ventral, but not lateraldorsal PAG of the rat. Behav Neurosci 1999;113:833–9. Tortorici V, Morgan MM, Vanegas H. Tolerance to repeated microinjection of morphine into the periaqueductal gray is associated with changes in the behavior of off- and on-cells in the rostral ventromedial medulla of rats. Pain 2001;89:237–44.
Vanderah TW, Suenaga NM, Ossipov MH, Malan TPJ, Lai J, Porreca P. Tonic descending facilitation from the rostral ventromedial medulla mediates opioid-induced abnormal pain and antinociceptive tolerance. J Neurosci 2001;21:279–86. Yaksh TL, Yeung JC, Rudy TA. Systematic examination in the rat of brain sites sensitive to the direct application of morphine: observation of differential effects within the periaqueductal gray. Brain Res 1976;114:83–103. Young EG, Watkins LR, Mayer DJ. Comparison of the effects of ventral medullary lesions on systemic and microinjection morphine analgesia. Brain Res 1984;290:119–29.