Abstract
σ1 Receptors have been implicated in the modulation of opioid analgesia. In the current study, we examined the role of σ1 systems in the periaqueductal gray (PAG), the rostroventral medulla (RVM), and the locus coeruleus (LC) of the rat, regions previously shown to be sensitive to morphine. Morphine was a potent analgesic in all three regions. Coadministration of the σ1 agonist (+)-pentazocine diminished the analgesic actions of morphine in all three regions, although the PAG was far less sensitive than the other two regions. Blockade of the σ1 receptors with haloperidol in the RVM markedly enhanced the analgesic actions of coadministered morphine, implying a tonic activity of the σ1 system in this region. This effect was mimicked by down-regulation of RVM σ1 receptors using an antisense approach. However, no tonic σ1 activity was observed in either the LC or the PAG. The RVM also was important in modulating analgesia elicited from morphine microinjected into the PAG. The analgesic actions of morphine given into the PAG could be attenuated by (+)-pentazocine placed into the RVM, whereas haloperidol in the RVM enhanced PAG morphine analgesia. These studies illustrate the pharmacological importance of σ1 receptors in the brainstem modulation of opioid analgesia.
σ Receptors are unique proteins of approximately 28 kDa that have been cloned from guinea pigs (Hanner et al., 1996), humans (Kekuda et al., 1996), mice (Pan et al., 1998), and rats (Seth et al., 1998; Mei and Pasternak, 2001) with distinct pharmacological characteristics (Martin et al., 1976; Bowen, 2000; Matsumoto, 2007). They have been implicated in a wide range of actions. They have been associated with potassium channels (Aydar et al., 2002) and aspects of cell proliferation and cancer (Bowen, 2000; Aydar et al., 2004; Casellas et al., 2004). Among their actions, σ receptors comprise a tonically active antiopioid system (Chien and Pasternak, 1993, 1994, 1995a,b; Pasternak, 1994; King et al., 1997; Mei and Pasternak, 2002). The σ1 antagonist haloperidol greatly potentiated systemic opioid analgesia. Although (–)-pentazocine is an effective opioid with activity at both μ and κ receptors, (+)-pentazocine, in contrast, has no opioid activity, but it is a potent σ1 agonist. (+)-Pentazocine reduced systemic opioid analgesia for μ, δ, κ1, κ3, and orphanin FQ/nociceptin ligands. This enhanced activity of opioid actions with haloperidol implied that the σ1 system was tonically active. Furthermore, the differences in sensitivity to opioids among several strains of mice could be eliminated by blocking σ1 actions with haloperidol, raising the possibility that these sensitivity differences might reflect varying levels of tonic activity of the σ1 system.
The cloning of σ1 receptors facilitated their study at the molecular level. Down-regulation of σ1 receptors in either the mouse or rat using antisense techniques had effects similar to those of the antagonist haloperidol. Thus, the evidence implicating the ability of σ1 receptors to modulate opioid analgesia is strong.
A number of brainstem structures have shown potent morphine analgesic activities (Pert and Yaksh, 1974; Bodnar et al., 1988; Rossi et al., 1993, 1994a). These studies revealed complex synergistic interactions among three morphine-sensitive sites, the periaqueductal gray (PAG), the locus coeruleus (LC), and the rostroventral medulla (RVM). Autoradiographic studies indicate that σ1 receptors are present within the brainstem, including these morphine sensitive sites (Walker et al., 1992). Furthermore, the σ1 receptor antagonist haloperidol and agonist (+)-pentazocine, influence supraspinal, but not spinal, morphine analgesia (J. F. Mei and G. W. Pasternak, unpublished data). This supraspinal modulation of opioid analgesia by σ1 receptors raises questions regarding the regional localization of these interactions. We now report the mapping of rat σ1 receptor modulation of morphine analgesia in brainstem nuclei.
Materials and Methods
Materials. Na[125I] was purchased from PerkinElmer Life and Analytical Sciences (Boston, MA). [125I](+)-Pentazocine labeling was performed using the chloramine T/sodium metabisulfate method (Letchworth et al., 2000). (+)-Pentazocine and morphine sulfate were gifts from the Research Technology Branch of National Institute on Drug Abuse (Bethesda, MD). Halothane was purchased from Halocarbon Laboratories (Hackensack, NJ). All other chemicals were purchased from Sigma-Aldrich (St. Louis, MO). All the drugs used in the in vivo studies were dissolved in saline. Glass fiber filters (number 32) were purchased from Whatman Schleicher and Schuell (Keene, NH). Formula 963 liquid scintillant was purchased from PerkinElmer Life and Analytical Sciences.
Animals and Analgesic Testing. Male albino Sprague-Dawley rats (175–250 g) were purchased from Charles River Laboratories, Inc. (Wilmington, MA), and they were maintained on a 12-h light/dark cycle with rodent chow and water available ad libitum. Rats were housed in groups of two in clear plastic cages at ambient room temperatures between 21 and 25°C until they were tested. All animal studies were approved by the Institutional Animal Care and Use Committee of the Memorial Sloan-Kettering Cancer Center (New York, NY), and they adhered to National Institutes of Health guidelines (Institute of Laboratory Animal Resources (1996).
Analgesia was determined using the radiant heat tail-flick technique as reported previously (Chien and Pasternak, 1995b). The thermal stimulus was positioned 8 cm above the dorsum and 3 to 9 cm proximal to the tip of the tail of a lightly restrained animal. The mean of two latency readings was taken for each animal at each indicated time. The baseline latencies ranged between 2 and 3 s, and maximal latency of 10 s was used to minimize the tissue damage. Studies used groups of four to seven rats. Saline injections did not appreciably alter the latencies over time compared with the baseline value, as shown in Fig. 2A. The percent maximal possible effect (%MPE) was calculated as the (observed latency – baseline latency)/(maximal latency – baseline latency).
Cannulations. Rats were anesthetized with chlorpromazine HCl (3 mg/kg i.p.) given 20 min before ketamine HCl (100 mg/kg i.m.) injection. The animal was set on the Kopf stereotaxic instrument (David Kopf Instruments, Tujunga, CA), and a stainless steel guide cannula (26-gauge; Plastic Products, Roanoke, VA) was implanted stereotaxically in the specified location. Cannulae coordinates were the same as in prior studies (Rossi et al., 1993). Cannulae were secured to the skull with three anchor screws and dental acrylic. The cannulae were then capped with a dummy cannulae (Plastic Products). Each animal was housed in a single cage, and it was allowed 1 week to recover from surgery before behavioral testing began. All cannulated rats were tested with morphine to confirm cannula placement at least 5 days before the experiment: PAG, 2.5 μg; and RVM and LC, 5 μg. Animals displaying a poor response were assumed to have a faulty cannulae placement, and they were not used in further experiments (Bodnar et al., 1988, 1991).
After behavioral testing, cannulae placement was confirmed anatomically. Rat brains were fixed in 10% buffered formalin overnight and transferred into 30% sucrose until section cutting. The brains were cut coronally and stained with cresyl violet, and placement was confirmed by light microscopy. Animals whose placements were inaccurate were not included in the result (Bodnar et al., 1988, 1991; Rossi et al., 1993).
Antisense and in Vivo Assay. Antisense oligodeoxynucleotides were designed using the rat σ1 receptor cDNA sequences with the Align and Gene Runner programs (Scientific & Educational Software, Cary, NC), and they were purchased from Midland Certified Reagent (Midland, TX). The oligodeoxynucleotides were repurified using sodium acetate precipitation, and they were dissolved in saline to make a final concentration of 5 μg/μl.
The antisense oligodeoxynucleotide sequence (RS357AN) corresponding to the cloned rat σ1 receptor (rs2-2) was 5′-CCAGCCGCCCGCGTTCAC-3′, and the mismatch control differed by switching six bases, 5′-CCACGCGCCGCCGTTACC-3′.
Groups of mice were treated with antisense oligodeoxynucleotides (10 μg in 2 μl of saline per injection) or vehicle (saline) on days 1, 2, and 4, and they were tested for analgesia on day 5 with morphine sulfate (μ). Previous studies in rats and mice have shown that antisense treatment can down-regulate opioid or σ1 receptors at both the mRNA and protein levels (Rossi et al., 1994b; Mei and Pasternak, 2001).
Data Analysis. Rat analgesia data were compared by both the area under the time-action curve and at peak effect using one-way analysis of variance followed by a post hoc Tukey's analysis. ED50 values were calculated by Pharm/PCS software (Springer-Verlag, New York, NY) using nonlinear regression analysis. Results are presented as means ± S.E.M. of triplicate experiments unless specified otherwise.
Results
Modulation of Supraspinal Morphine Analgesia by a σ1 Receptor Agonist. Microinjection of morphine into any one of a number of sites within the rat brain elicits a profound naloxone-sensitive analgesia (Pert and Yaksh, 1974; Bodnar et al., 1988). We chose three sites sensitive to morphine analgesia to explore the interaction between supraspinal opioid and σ1 systems: PAG, LC, and RVM.
Microinjection of morphine into each of the three regions produced a dose-dependent analgesia, with peak effects of 30 min in the PAG and the RVM and 15 min in the LC, and durations of action were between 60 and 90 min (Fig. 1). These results are consistent with our earlier study (Rossi et al., 1993).
We next explored whether σ1 systems modulated morphine action in these regions. The σ1 agonist (+)-pentazocine significantly blocked morphine analgesia in the PAG, RVM, and LC (Fig. 2, A–C) in a dose-dependent manner (Fig. 2D), but the sensitivity of the regions varied markedly. The RVM was the most sensitive to (+)-pentazocine, with an ID50 of 2.6 ng. The LC also was quite sensitive, with an ID50 of 17.4 ng. In contrast, the doses of (+)-pentazocine needed to reduce morphine analgesia in the PAG were up to 1000-fold greater, with an ID50 of 4090 ng (Fig. 2D).
In both the PAG site and RVM sites, the effects of (+)-pentazocine could be overcome by increasing the morphine dose (Fig. 3; Table 1). In the PAG, (+)-pentazocine significantly shifted the morphine dose-response curve more than 4-fold, increasing the ED50 from 1.5 to 6.5 μg. We observed a similar significant shift in the morphine dose-response curve in the RVM, with the morphine ED50 significantly increasing from 2.8 to 12 μg. In both regions, morphine retained the ability to fully overcome the inhibitory actions of (+)-pentazocine.
Modulation of PAG Analgesia through RVM σ1 Receptor Systems. Opioid administration into the PAG activates a circuit between the PAG and the RVM, with release of endogenous opioids (Osborne et al., 1996; Fields, 2000). We therefore examined potential interactions between the PAG and RVM. Administration of (+)-pentazocine into the RVM attenuated the analgesic activity of morphine microinjected into the PAG (Fig. 4A), consistent with the prior demonstration of the circuit. (+)-Pentazocine was equally effective in attenuating the actions of morphine given into either the PAG or RVM (Fig. 4B). The ability of RVM (+)-pentazocine to block PAG morphine analgesia is further supported by the sensitivity of PAG morphine analgesia to RVM (+)-pentazocine at only 50 ng, a dose that was inactive against morphine when both were given directly into the PAG.
This ability of (+)-pentazocine administered into the RVM to block PAG morphine analgesia raised the question of whether the high doses of (+)-pentazocine needed to lower morphine analgesia when both drugs were given into the PAG might be due to diffusion of (+)-pentazocine from the PAG to the RVM. However, this seems to be unlikely for several reasons. Autoradiographically, we failed to see evidence of [125I](–)-pentazocine diffusing in the RVM following its injection into either the LC or PAG after 45 min. However, the sensitivity of autoradiography is limited, and it might not be sufficient to detect (+)-pentazocine doses necessary for activity in the PAG and RVM. Therefore, we examined the ability of the σ1 antagonist haloperidol to reverse (+)-pentazocine actions. If the activity of PAG (+)-pentazocine was due to its diffusion into the RVM, haloperidol administered into the RVM should reverse the actions of (+)-pentazocine and increase morphine analgesia. Haloperidol did not reverse (+)-pentazocine (Fig. 5), making it unlikely that the actions of (+)-pentazocine in the PAG result from diffusion into the RVM. However, it leaves open the question of why this region is so insensitive, particularly because it has readily demonstrable levels of σ binding sites autoradiographically (Gundlach et al., 1986).
Modulation of Supraspinal Morphine Analgesia by a σ1 Receptor Antagonist. Haloperidol enhances opioid analgesia, presumably by blocking a tonically active σ1 system (Chien and Pasternak, 1993, 1994, 1995b; King et al., 1997; Mei and Pasternak, 2002). To assess the possibility of tonic σ1 activity within individual brainstem regions, we examined the effects of the σ1 antagonist haloperidol on morphine analgesia in each region. At the maximal doses, we could test due to solubility constraints (5 μg), haloperidol had little effect on morphine analgesia when both were microinjected into either the PAG or the LC (Fig. 6). However, haloperidol coadministered with morphine into the RVM markedly enhanced the analgesic response. Full dose-response curves revealed that haloperidol shifted the ED50 of morphine in the RVM by approximately 2-fold (Fig. 7; Table 1). These findings confirm tonic σ1 activity in the RVM but not in the PAG and LC sites. Haloperidol can interact with both σ1 and dopamine D2 receptors. To ensure that the actions of haloperidol were mediated through σ1 sites, we also tested (–)-sulpiride, a D2 receptor-selective antagonist. Administering (–)-sulpiride along with morphine into the RVM failed to enhance the analgesic actions of morphine in the RVM, supporting the suggestion that haloperidol was acting through blockade of σ1 receptors (J. F. Mei and G. W. Pasternak, unpublished observations), results similar to those in mice (Chien and Pasternak, 1994).
Effects of Down-Regulation of σ1 Receptors on Morphine Analgesia in the RVM. Antisense approaches have proven valuable in selectively down-regulating opioid receptors and σ receptors within the central nervous system, with a selectivity greater than that of most antagonists (Standifer et al., 1994; Pasternak and Pan, 2000). To confirm the role of σ1 receptors in the potentiation of morphine analgesia, we used an antisense approach similar to that used to down-regulate μ opioid receptors (Rossi et al., 1994b). In earlier work, our group demonstrated the ability of antisense treatment to down-regulate σ1 receptors and to potentiate opioid analgesia in both mice and rats (Pan et al., 1998; Mei and Pasternak, 2001, 2002). In this study, we administered the antisense directly into RVM, and we tested morphine given into the same region. Antisense treatment enhanced morphine analgesia in the RVM [Fig. 8), as shown both by the peak effect in the time-action curve (p < 0.001) and by the area under the curves (p < 0.001)]. To ensure the specificity of the effect, we used a mismatch antisense probe in which the order of three sets of adjacent bases was switched, keeping the total base composition the same. The mismatch was without effect, confirming the selectivity of the antisense effect.
σ1 Actions in the RVM can modulate the analgesic actions of morphine in the PAG. We therefore examined of the effects of antisense treatment in the RVM on the analgesic activity of morphine administered into the PAG (Fig. 8). Morphine in the PAG at a dose that failed to significantly elevate tail-flick latencies above baseline levels elicited a profound analgesic response following antisense treatment in the RVM. Again, the inactivity of the mismatch antisense oligodeoxynucleotide confirmed the selectivity of the response. These antisense studies confirmed the role of σ1 receptors in the modulation of morphine analgesia and the ability of σ1 receptors within the RVM to modulate the analgesic actions of PAG morphine.
Discussion
Within the brain stem region, microinjection studies have identified three morphine-sensitive regions, including the PAG, LC, and RVM (Jacquet and Lajtha, 1973; Pert and Yaksh, 1974; Bodnar et al., 1988, 1991; Rossi et al., 1993). Although each can induce analgesia alone, they also interact with each other, as shown by analgesic synergy between morphine administered in the PAG and the RVM (Rossi et al., 1994a). This pharmacological interaction is consistent with detailed studies demonstrating the importance of the RVM in mediating the actions of opioids in the PAG (Osborne et al., 1996; Christie et al., 2000; Fields, 2000).
The σ receptor system reportedly plays an important role in antiopioid, antipsychotic, neuroprotective, and antidepressant activities (Bowen, 2000; Hayashi and Su, 2005). (+)-Pentazocine, a selective σ1 receptor agonist, blocks opioid analgesia, whereas the σ1 antagonist haloperidol potentiates opioid analgesia (Chien and Pasternak, 1993, 1994, 1995a,b), results supported by antisense approaches in mice (Mei and Pasternak, 2001, 2002). The object of the current study was to explore the role of σ1 receptor systems within a series of brainstem nuclei of the rat known to be important in opioid analgesia.
The ability of the σ1 agonist (+)-pentazocine to diminish opioid analgesia in all three brainstem regions clearly demonstrates the presence and potential importance of the σ1 system in the modulation of opioid mechanisms. However, there were significant differences among the sites. (+)-Pentazocine was quite potent in both the LC and the RVM but not the PAG, despite the high levels of σ1 binding as assessed autoradiographically (Gundlach et al., 1986). Indeed, the doses necessary to influence morphine analgesia in the PAG were several orders of magnitude greater than the other two regions. Initially, we assumed that this might reflect the need for the drug to diffuse from the PAG to the RVM, particularly because (+)-pentazocine at far lower doses administered directly into the RVM attenuated morphine analgesia following morphine injection into the PAG. However, several lines of evidence suggest that this is not likely. First, directly looking at the diffusion of [125I]pentazocine failed to show any appreciable levels in the RVM after administering the compound into the PAG. In addition, the σ1 antagonist haloperidol given into the RVM did not reverse the actions of (+)-pentazocine administered into the PAG, which would have been expected if the (+)-pentazocine were acting in the RVM. Yet, it is still not clear why (+)-pentazocine is so weak in the PAG. It may simply reflect the limitations of the σ1 system in this region in modulating analgesia, but at these high doses, (+)-pentazocine may be acting through alternative, non-σ mechanisms.
Although morphine analgesia in both the LC and RVM was lowered by low doses of (+)-pentazocine, implying a highly sensitive σ1 system, only the RVM seemed to have tonic σ1 activity based upon the ability of the σ1 antagonist haloperidol to enhance morphine actions. Earlier work in mice (Chien and Pasternak, 1993, 1994, 1995a; King et al., 1997; Mei and Pasternak, 2002) and in rats (Chien and Pasternak, 1995b; Mei and Pasternak, 2001) showed that haloperidol enhanced opioid analgesia. Interestingly, varying levels of tonic σ1 receptor activity seemed to be responsible for many of the differences in sensitivities of some strains to opioids.
Although haloperidol is a potent σ1 antagonist, it also blocks dopamine D2 receptors quite effectively, making the interpretation of some of the studies difficult. Unlike haloperidol, the selective dopamine D2 receptor antagonist (–)-sulpiride does not interact with σ1 receptors. In our prior studies in CD-1 mice, (–)-sulpiride did not reproduce the actions of haloperidol. However, in other mouse strains and models, dopamine D2 receptors can influence opioid action (Calcutt et al., 1971; Kunihara et al., 1993; Kamei and Saitoh, 1996), a finding that was supported in a more recent study looking at opioid actions in a dopamine D2 receptor knockout mouse (King et al., 2001). Although there may be situations in which D2 receptors modulate opioid actions, σ1 systems also are important. This was clearly shown by the ability of haloperidol to potentiate and (+)pentazocine to block opioid analgesia in the D2 knockout mice. Because these mice have no D2 receptors, the actions must be mediated through σ1 sites.
To further verify the role of σ1 receptors in the RVM, we also examined the effects of antisense. Antisense approaches can effectively down-regulate receptors and influence opioid behavior (Standifer et al., 1994). Down-regulation of σ1 receptors in the RVM by antisense, but not by mismatch controls, potentiated morphine analgesia in much the same way as haloperidol. This confirms the role of σ1 receptors in the RVM in morphine analgesia.
The interactions between the PAG and RVM were particularly interesting. A large literature has established the importance of the PAG-to-RVM pathway in opioid analgesia (Osborne et al., 1996; Fields, 2000). This is further supported by the synergistic opioid interactions in microinjection studies (Rossi et al., 1993). The current studies support these PAG/RVM interactions. (+)-Pentazocine administered into the RVM attenuated morphine analgesia from the PAG, whereas haloperidol in the RVM enhanced it. Finally, antisense treatment targeting σ1 receptors in the RVM potentiated morphine analgesia from the PAG, much like the actions of haloperidol. Thus, the σ1 system in the RVM also modulates opioid actions from the PAG.
In conclusion the current study supports a role for σ1 receptor actions in the brainstem. Its activity within the three regions differed widely. Whereas both the LC and RVM were sensitive to (+)-pentazocine, the PAG was not, despite the demonstration of σ1 receptors in this site autoradiographically (Walker et al., 1992). The RVM was particularly interesting in that it was the only regions with evidence for tonic σ1 activity, and it could modulate the analgesia from morphine given into the PAG.
Acknowledgments
We thank Dr. Michael King for advice and comments on the manuscript and Drs. Grace C. Rossi and Richard J. Bodnar for advice and assistance.
Footnotes
-
This work was supported from a Senior Scientist Award and National Institute on Drug Abuse Grants DA07241, DA02615, and DA00220 (to G.W.P.) and National Cancer Institute Core Grant CA08748 (to Memorial Sloan-Kettering Cancer Center).
-
doi:10.1124/jpet.107.121137.
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
ABBREVIATIONS: PAG, periaqueductal gray; LC, locus coeruleus; RVM, rostroventral medulla; %MPE, percent maximal possible effect.
- Received February 9, 2007.
- Accepted May 31, 2007.
- The American Society for Pharmacology and Experimental Therapeutics