- Email: [email protected]
Antinociception produced by mu opioid receptor activation in the amygdala is partly dependent on activation of mu opioid and neurotensin receptors in the ventral periaqueductal gray
Brain Research 865 (2000) 17–26 www.elsevier.com / locate / bres
Research report
Antinociception produced by mu opioid receptor activation in the amygdala is partly dependent on activation of mu opioid and neurotensin receptors in the ventral periaqueductal gray Sheralee A. Tershner a , *, Fred J. Helmstetter b b
a Department of Psychology, Western New England College, Springfield, MA 01119, USA Department of Psychology, University of Wisconsin–Milwaukee, Milwaukee, WI 53201, USA
Accepted 15 February 2000
Abstract Exposure to stressful or fear-inducing environmental stimuli activates descending antinociceptive systems resulting in a decreased pain response to peripheral noxious stimuli. Stimulating mu opioid receptors in the basolateral nucleus of the amygdala (BLA) in anesthetized rats produces antinociception that is similar to environmentally induced antinociception in awake rats. Recent evidence suggests that both forms of antinociception are mediated via projections from the amygdala to the ventral periaqueductal gray (PAG). In the present study, we examined the types of neurochemicals released in the ventral PAG that may be important in the expression of antinociception produced by amygdala stimulation in anesthetized rats. Microinjection of a mu opioid receptor agonist into the BLA resulted in a time dependent increase in tail flick latency that was attenuated by preadministration of a mu opioid receptor or a neurotensin receptor antagonist into the ventral PAG. Microinjection of a delta 2 opioid receptor antagonist or an NMDA receptor antagonist into the ventral PAG was ineffective. These findings suggest that amygdala stimulation produces antinociception that is mediated in part by opioid and neurotensin release within the ventral PAG. 2000 Elsevier Science B.V. All rights reserved. Themes: Sensory systems Topics: Pain modulation: pharmacology Keywords: Amygdala; Antinociception; Neurotensin; Opioid; Periaqueductal gray; Tail flick
1. Introduction Exposure to environmental stressors can inhibit ascending pain transmission [9,27,42]. Some forms of stress induced hypoalgesia are dependent on opioids [7,8,12,18,19] and are mediated via a descending antinociceptive circuit which includes the amygdala [14–16], periaqueductal gray (PAG) [7,19–21,26,36], and rostral ventromedial medulla (RVM) [20,36]. There is accumulating evidence that environmentally induced hypoalgesia is often a result of the activation of opioid receptors within the amygdala [17,33,35]. There is a heterogenous distribution of opioid receptors within the amygdaloid nuclei [2], with mu opioid receptors heavily
*Corresponding author. Tel.: 11-413-796-2193. E-mail address: [email protected] (S.A. Tershner)
localized within the BLA and lateral nuclei, while delta and kappa opioid receptors are only moderately represented. In comparison, there are relatively few opioid receptors of any type located within the central nucleus of the amygdala (CeA) [30]. We have recently demonstrated that activation of mu opioid receptors, not delta or kappa, in the amygdala produces antinociception in anesthetized rats [16]. The amygdala appears to modulate nociceptive responses by activating a descending antinociceptive pathway through the midbrain and medulla. Although both the CeA and BLA send neuronal projections to the ventral PAG, it appears that the primary input from the BLA to the ventral PAG includes a critical synapse within the CeA [22,34]. The ventral PAG in turn projects to nociceptive modulating neurons in the RVM [3]. Microinjection of the mu opioid receptor agonist DAMGO into the BLA produces inhibition of the tail flick (TF) reflex that is
0006-8993 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 00 )02179-X
18
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
blocked by lesions or lidocaine inactivation of the ventral PAG or RVM [21]. Although there is evidence for a serial connection between the amygdala and ventral PAG necessary for antinociception following opioid injection into the amygdala [21], little is known regarding the neurotransmitters involved. The ventral PAG contains opioid sensitive cells that when stimulated can produce antinociception [33,38,46]. There are mu and kappa, but relatively few delta opioid receptors located within the PAG, with the highest concentration of mu opioid receptors located in the ventral region [41]. A population of ventral PAG cells is excited following electrical stimulation of the amygdala, and this excitation is inhibited by the local application of an opioid antagonist [11]. In addition, injections of naltrexone into the ventral PAG in anesthetized rats reduce the antinociception produced by mu opioid receptor activation in the amygdala [43] and mu, but not kappa, opioid receptor antagonists injected into the ventral PAG block conditioned fear induced hypoalgesia in awake rats [7]. There appears to be a critical opioid synapse within the ventral PAG in antinociception produced by amygdala manipulation, however there is little evidence that direct afferents from the CeA to the PAG release opioids [13]. Therefore, it is likely that additional neurotransmitters are involved and released within the PAG by terminals of CeA cells as well as by local interneurons. Neurotensin-immunoreactive cells originating in the lateral regions of the CeA project to the caudal ventrolateral regions of the PAG [13]. Cell bodies containing neurotensin located in the ventral PAG [4,5,37] and neurotensinergic interneurons are believed to contribute to prolonged excitation of PAG neurons during antinociception [4]. Injections of neurotensin into the PAG can produce a dose-dependent inhibition of TF latency that is not reversible with naloxone and is thus believed to be independent of opioids [1,5]. In addition to the neurotensin there are excitatory amino acid (EAA) projections from the CeA to the PAG [25] and a substantial number of EAA immunoreactive terminals that contain glutamate and aspartate [6,29]. Site specific injections of glutamate and NMDA into the PAG produce antinociception in the TF test [23,45]. To examine whether ventral PAG opioid, neurotensin and NMDA receptor activation is necessary for the antinociception produced by amygdala stimulation their respective antagonists were microinjected into the ventral PAG in anesthetized rats prior to the administration of DAMGO into the BLA. We chose to use a reduced preparation that involves direct pharmacological activation of cell groups in this model circuit in anesthetized animals. It is believed that this neural circuit has evolved to subserve defensive and antinociceptive functions in the behaving animal and can be studied in this reduced manner in cases where it would be difficult to perform in behaving animals, e.g. experiments requiring multiple simultaneous intracerebral injections and cell recording. Results found
using this anesthetized preparation has yielded comparable results as those found in subsequent studies involving awake subjects [14,15,20] and is therefore considered a valid model for studying the circuit involved in environmentally induced analgesia.
2. Materials and methods
2.1. Subjects Subjects were 111 male Long–Evans rats (325–450 g) obtained from Harlan Sprague–Dawley (Madison, WI). Animals were housed individually in stainless steel cages with unlimited access to food and water. All experimental manipulations were performed during the light portion of a 14:10-h light / dark cycle.
2.2. Surgical preparation Prior to surgery, all animals were anesthetized with sodium pentobarbital (50 mg / kg / ml; i.p.). A catheter constructed from polyethylene tubing (PE-20) was inserted into the jugular vein at a depth of 1.0 cm. The vessel was secured around the tubing using 3-0 silk suture. Once implanted, the rat received a constant infusion of methohexital (10 mg / ml) delivered through the catheter at an infusion rate of 0.8 ml / h. The subjects were positioned in a David Kopf stereotaxic apparatus. Cannulae implantation consisted of exposing the surface of the skull and lowering 26 Gauge stainless steel guide cannulae into the BLA (bilateral) and ventral PAG (unilateral) using coordinates obtained from a rat brain atlas [32] (AP: 23.0; L: 1 / 25.2; V: 27.9 and AP: 27.8; L: 10.8; V: 26.0, respectively). All guide cannulae were secured to the skull with epoxy. Subjects remained anesthetized and in the stereotaxic apparatus during the entire testing session.
2.3. Apparatus Heart rate was recorded during the test period to measure possible effects of administered substances on cardiovascular activity. This variable also provided an index of the stability of anesthesia throughout the test session. Recording electrodes for measuring heart rate were positioned subcutaneously near the sternum ventrally and between the scapulae dorsally. Signals from the EKG electrodes were amplified (Grass Instruments) and recorded by a microcomputer with an A / D converter. Raw data waveforms were recorded and analyzed with software written with the ASYST data acquisition system. TF latency was measured using a custom made apparatus constructed from a stainless steel box with a 500-W projection bulb mounted inside. There was a 3-mm circular aperture in the top of the box that permitted
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
radiant heat from the bulb to make contact with the ventral side of the rat’s tail. A thermocouple positioned within the TF apparatus recorded stimulus intensity. The heat stimulus was set to elicit baseline TF latencies between 4.5 and 6.5 s resulting in a radiant heat pulse which reached 45.68C in 20 s. This heat setting was held constant between all subjects and throughout the testing period. A cooling fan attached to the apparatus maintained the surface of the apparatus between 28 and 328C. When the rat’s tail deflected from the heat source a photocell shut off a digital timer and deactivated the stimulus. If a TF response did not occur within 20 s, the heat stimulus was terminated to prevent tissue damage.
2.4. Drugs and injection procedure H-D-Phe-Cys-Tyr-D-Trp-Arg-Pen-Thr-NH2 (CTAP) (Multiple Peptide Systems; 0.0005, 0.005, and 0.05 mg / 0.25 ml); and naltriben (NTB) (Research Biochemicals Inc.; 0.12, 1.2, 4.0 mg / 0.25 ml) were used to examine the role of mu and delta opioid receptors, respectively, within the ventral PAG in amygdala induced antinociception [7,38–40]. The partial neurotensin antagonist [D-Trp 11 ] neurotensin (NTA) (Sigma; 4.0 pmol (1.69310 23 pg), 16 pmol (6.78310 23 pg), and 64 pmol (2.71310 22 pg) / 0.25 ml), and an NMDA receptor antagonist DL-2-amino-5phosphono-valerate (APV) (Research Biochemicals Inc.; 0.03, 0.30, and 3.0 mg / 0.25 ml) were injected into the ventral PAG to examine the role of neurotensin and NMDA receptor activation in amygdala antinociception [23,44]. The selective mu opioid receptor agonist D-Ala 2 N-MePhe 4 -Gly-Ol 5 enkephalin (DAMGO) (Sigma; 0.5 mg / 0.25 ml / side) was injected into the BLA to produce antinociception [16]. All drugs were dissolved in isotonic sterile saline. A pooled control group which received saline injections into the ventral PAG and DAMGO into the BLA (SAL-DAMGO) was used for statistical comparisons in the analyses of all receptor antagonists. Subjects assigned to the SAL-DAMGO control group were tested concurrently with the testing of the subjects in the other treatment groups. Drug delivery was performed with an injector constructed from a 33-Gauge stainless steel cannula connected to a piece of PE-20 tubing. The injector was attached to a 1.0-ml Hamilton microsyringe mounted on a Harvard Apparatus infusion pump. All drugs were delivered in a volume of 0.25 ml over a 45-s period.
2.5. Testing procedure Once the rat was prepared for the experiment, a standard protocol was implemented. The test session consisted of 30 TF trials with an intertrial interval of 120 s. Heart rate was recorded during the twenty seconds prior to and the twenty seconds following the onset of the heat stimulus. Four
19
baseline TF trials were recorded before drug administration. Immediately after trial four (time58 min), an opioid receptor antagonist or isotonic sterile saline was infused into the ventral PAG followed by DAMGO injections into the BLA at the end of the fifth trial (time510 min).
2.6. Histology and data analyses After testing rats were overdosed with an intraperitoneal injection of euthanasia solution consisting of ethanol and sodium pentobarbital (100 mg / kg). They were then perfused transcardially with 0.9% saline followed by 10% phosphate buffered formalin. Following decapitation, the heads were left in buffered formalin for 24 h before brain removal. The brains were soaked in sucrose / formalin for an additional 24 h. To verify cannula placements the brains were sliced into coronal sections (40 mm), mounted on slides, and stained using cresyl violet. After reviewing the histology, only animals with injection sites located within the ventral PAG and bilaterally within the BLA, as defined by Paxinos and Watson [32], were included in the analyses. Amygdala injection sites were considered acceptable if both cannulae were located in the region of the BLA or the ventral portions of the lateral nucleus. PAG injection sites were considered acceptable if they were located in the ventral portion of the PAG and lateral to the dorsal raphe nucleus. See Fig. 1A and B for a schemata depicting the areas of acceptance within the amygdala and ventral PAG, respectively. TF and heart rate responses were subjected to within and between subjects ANOVA. TF latency was measured in seconds (0–20 s) and heart rate values were recorded in beats / min and transformed to reflect change from baseline values. Tukey post-hoc tests (P,0.05) were performed for between group comparisons when needed.
3. Results
3.1. Histology After examination of the histology, rats with cannulae placements outside of the regions of acceptance, as described in Materials and methods section, were excluded from the data analysis. There were five rats in each of the CTAP drug conditions, seven subjects in the SALDAMGO control group, five subjects in each of the NTB drug conditions, four subjects in the NTB-SAL group, five subjects in each of the NTA drug conditions, four subjects in the NTA-SAL group, and five subjects in each of the APV drug conditions. Cannulae placements with the BLA and PAG can be seen in Fig. 1A and B, respectively.
3.2. Mu opioid receptor antagonist Microinjections of CTAP into the ventral PAG at-
20
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
Fig. 1. Schemata representation of the areas of acceptance for cannulae placements. All BLA injections were made bilaterally (panel A). PAG injections were made unilaterally and were counterbalanced across subjects for side of injection site (panel B).
tenuated the increase in TF latency produced by DAMGO injections in the BLA (see Fig. 2). An ANOVA on the mean of the four TF latencies prior to drug administration indicated that baseline nociception between the groups was not different, F(4,22)51.41, P.0.05. DAMGO injections into the BLA produced a time dependent inhibition of TF latency in rats that received saline injections into the PAG. The intermediate (0.005 mg) and high (0.05 mg) doses of CTAP injected into the PAG attenuated the increase in TF latency resulting from DAMGO administration in the BLA. These observations were supported by an ANOVA for repeated measures on post injection TF latencies which yielded a main effect for group, F(4,22)514.73, P,0.05, a significant trial effect, F(25,550)514.80, P,0.05, and a significant trial by group interaction, F(100,550)52.99, P,0.05. All doses of CTAP tested delayed the onset of TF inhibition, post-hoc Tukey analyses (P,0.05), but only the intermediate and higher doses continued to decrease the inhibition throughout the session (P,0.05). The rats receiving saline injections into the BLA and the high dose of CTAP into the PAG did not show any change in TF latency demonstrating that this dose of CTAP is not
producing hyperalgesia. Heart rate was not different in any of the groups either before F(4,19)52.017, P.0.05, or after drug administration F(4,19),1.0.
3.3. Delta receptor antagonist None of the doses of NTB, when injected in to the PAG were effective in attenuating antinociception produced by DAMGO administration in the BLA (see Fig. 3). An ANOVA failed to show any significant differences between the groups on TF latency before drug administrations F(4,21)51.070, P.0.05. All groups except for the NTBSAL group showed a time dependent increase in TF latency indicated by a main effect for group F(4,21)5 7.458, P,0.05, trial F(25,525)516.023, P,0.05, and a group by trial interaction F(100,525)52.125, P,0.05. Surprisingly, at 24 min after DAMGO injection the low dose of NTB (0.12 mg) potentiated the increase in TF latency, post-hoc Tukey comparison (P,0.05). Heart rate was not different in any of the groups either before F(4,19),1.0, or after drug administration F(4,19),1.0.
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
21
Fig. 2. Effects of injecting a mu opioid receptor antagonist into the ventral PAG on antinociception produced by DAMGO injections into the BLA. DAMGO administration alone produced a time-dependent increase in TF latency which lasted the duration of the test session. This increase in TF latency was attenuated by pretreatment with CTAP in the ventral PAG. Vertical lines denote a standard error of the mean.
3.4. Neurotensin receptor antagonist An ANOVA on mean baseline TF latencies indicated that the groups were not significantly different before drug administration F(4, 21)52.712, P.0.05. Ventral PAG injections of the NTA attenuated the increase in TF latency produced by activating mu opioid receptors within the BLA (see Fig. 4). TF trials measured after drug administration were subjected to an ANOVA for repeated measures which showed that there was a significant main effect for group F(4,21)54.942, P,0.05, trials F(25,525)512.092, P,0.05, and a drug by trial interaction F(100,525)51.701. The low and intermediate dose of NTA (0.0016 and
0.0067 pg, respectively) injected into the PAG delayed the onset of TF inhibition produced by DAMGO administration in the BLA as measured at time points 10–18 min after DAMGO injection, post-hoc Tukey analyses (P, 0.05). At the end of the session the attenuation of TF inhibition with these two doses of NTA was still evident, P,0.05. TF latencies in the NTA-SAL group did not increase, thus providing evidence that the intermediate dose of NTA is not apposing DAMGO antinociception by producing hyperalgesia. Although the highest dose of NTA did not have an effect on TF latency it did increased HR throughout the test session (see Fig. 5). An ANOVA for repeated measures
Fig. 3. Effects of injecting a delta 2 opioid receptor antagonist into the ventral PAG on antinociception produced by DAMGO injections into the BLA. NTB, at the doses examined in this experiment, did not attenuate the increase in TF latency produced by DAMGO injections into the BLA. Vertical lines denote a standard error of the mean.
22
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
Fig. 4. Effects of injecting a neurotensin receptor antagonist into the ventral PAG on antinociception produced by DAMGO injections into the BLA. DAMGO administration alone produced a time-dependent increase in TF latency which lasted the duration of the test session. This increase in TF latency was attenuated by pretreatment with NTA in the ventral PAG. Vertical lines denote a standard error of the mean.
on post drug trials failed to reveal a significant main effect for group F(4,20)52.724, P.0.05, but did reveal a significant trial effect F(25,500)52.640, P,0.05, and a significant group by trial interaction F(100,500)52.038, P,0.05. Post-hoc Tukey analyses (P,0.05) revealed that 16 min after drug administration HR in subjects that received a high dose of NTA was significantly higher than the SAL-DAMGO controls. None of the other comparisons were significant at this time point. HR differences between
these two groups persisted throughout the duration of the test session. HR measures taken prior to drug administrations were not different between the groups F(4, 20),1.0.
3.5. NMDA receptor antagonist An ANOVA on the baseline TF latencies of the groups indicated that there were significant differences between the groups before drug administration, F(4,22)52.915,
Fig. 5. HR recordings were transformed to reflect change from baseline HR values after drug administration. The high dose of NTA (64 pmol) produced a slight increase in HR that lasted the duration of the test session. Vertical lines denote a standard error of the mean.
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
P50.045. There is no definite explanation for the statistical difference found, however after examining the data it appears that the difference may be due to the extremely small variance in the group that received the intermediate dose of APV. Upon inspection of the data we believe that the finding is a statistical anomaly. None of the doses of APV injected in the PAG attenuated the increase in TF latency produced by DAMGO injections into the BLA (see Fig. 6). TF latencies of the subjects that received the high (3.0 mg) dose of APV into the PAG increased immediately following administration. These observations were supported by an ANOVA for repeated measures on post drug TF trials which yielded a significant main effect F(4,22)5 5.249, P,0.05, trial effect F(25,550)514.740, P,0.05 and a group by trial interaction F(100,550)51.953, P, 0.05. Post-hoc Tukey comparisons (P,0.05) between the groups revealed that the highest dose of APV facilitated the onset of TF inhibition. The lack of an antinociceptive effect of the high dose of APV in the absence of DAMGO suggests that the increase in the onset of TF inhibition does not reflect a summation of a primary analgesic effect of the drug. An ANOVA on mean baseline HR measures indicated that the groups were not different prior to drug administration F(4,20)51.708, P.0.05. Rats that received APV following DAMGO injections into the BLA showed a small tachycardiac response (less than a 25 beat / min increase) which persisted throughout the remainder of the test session indicated by a significant trial by group interaction F(100,500)52.296, P,0.05. (see Fig. 7). The high dose of APV (3.0 mg) did not have an effect on HR
23
when administered alone, suggesting that the ability of APV to increase HR is dependent on the coadministration of APV and DAMGO.
4. Discussion The purpose of these experiments was to determine which neurotransmitters are released within the PAG in response to mu opioid receptor stimulation in the BLA. By injecting selective opioid receptor antagonists into the PAG this experiment demonstrated that mu opioid receptors are involved on the expression of DAMGO induced antinociception from the BLA. This finding in anesthetized rats compliments the results of similar studies in awake rats. Fear-induced antinociception is mediated by opioid release within the PAG [19], and more specifically through the activation of mu opioid receptors within the ventral PAG [7]. In addition, a study by Pavlovic et al. [31] demonstrated that analgesia produced by injecting morphine into the amygdala of awake rats was blocked by CTAP injections into the PAG. Taken together, these results clearly demonstrate the importance of a mu opioid receptor link within the ventral PAG in antinociception produced by the amygdala. The reports on the effectiveness of delta opioid agonists producing antinociception when injected into the PAG have been mixed. Some studies have failed to show delta opioid antinociception [28], while others have found antinociceptive effects within the PAG [24,36]. The present lack of evidence suggesting a role for PAG delta opioid
Fig. 6. Effects of injecting an NMDA receptor antagonist into the ventral PAG on antinociception produced by DAMGO injections into the BLA. An ANOVA revealed that the TF latencies were different between groups before drug administration. However, it is clear that this is merely a statistical anomaly and not a meaningful difference. APV injected into the ventral PAG did not attenuate the increase in TF latency produced by DAMGO injections into the BLA. Vertical lines denote a standard error of the mean.
24
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
Fig. 7. HR recordings were transformed to reflect change from baseline HR values after drug administration. All doses of APV tested produced a slight increase in HR that lasted the duration of the test session. Vertical lines denote a standard error of the mean.
receptors in amygdala antinociception directly contradicts the aforementioned study by Pavlovic et al. [31]. They found that intra-PAG injections of a delta 2 opioid receptor antagonist were effective in blocking antinociception produced by amygdala stimulation. Although there are some significant methodological differences between our studies, the reason for this discrepancy is not clear. Their study used awake and freely moving rats, different tests for pain sensitivity, and a delta 2 opioid receptor antagonist (naltrindole isothiocyanate) that required administration 24 h prior to testing. Any of these factors, or a combination of several, may have contributed to the differences between our findings. Since it is difficult to use naltrindole isothiocyanate in an acute preparation, additional behavioral studies using naltriben, at similar doses and using the TF test, are needed to clarify the role of PAG delta 2 opioid receptors in amygdala antinociception. Finally, the facilitation of TF inhibition by a delta opioid antagonist is difficult to explain since there is little or no evidence for delta opioid antagonists producing analgesia or enhancing mu agonist analgesia Selectively blocking mu receptors within the PAG did not completely block the antinociceptive input from the amygdala suggesting the involvement of multiple pain inhibitory processes. The results from this study suggest that in addition to a critical opioid synapse within the ventral PAG, neurotensin release within the ventral PAG is important for amygdala antinociception. The low and intermediate doses of NTA (0.0016 and 0.0067 pg) delayed the onset of the increase in TF latency following DAMGO administration and continued to attenuate the response throughout the test session. These doses of NTA did not produce changes in HR. The highest dose of NTA (0.0271 pg) was ineffective in attenuating TF inhibition
but did produce an increase in HR that lasted the duration of the test session. The partial agonist–antagonist properties of the NTA makes it very difficult to examine the endogenous release of neurotensin. Doses similar to those used in the present experiment were shown to block antinociception produced by neurotensin administration [44]. Since the NTA is not a true antagonist but a partial agonist, higher doses of NTA (100–300 pmol) injected into the RVM produced TF inhibition and low doses of NTA (3 pmol) decreased TF inhibition produced by neurotensin. Interestingly, Urban and Smith [44] reported that the same dose of NTA (3 pmol) injected into the RVM enhanced the TF inhibition produced by morphine administration into the PAG. It is quite possible that activation of neurotensin receptors may have inhibitory and facilitatory effects on nociceptive processes depending on the location of the receptors. The results of this experiment suggest that NMDA receptors within the PAG are not involved in pain inhibition produced by amygdala stimulation. None of the doses tested were successful in attenuating or blocking the increase in TF latency produced by DAMGO injections into the BLA. TF inhibition appeared to be facilitated in the group receiving the highest dose of APV (3.0 mg). Although, the same dose of APV (3.0 mg) did not produce antinociception when administered alone, a 10-mg dose of APV has been shown to produce analgesia in the formalin test when administered intrathecally [10]. It is possible that 3.0 mg of APV, with out producing antinociception when administered into the ventral PAG, could produce a synergistic enhancement of mu opioid analgesia. It is still not known if opioids and neurotensin are being released by afferent projections or by local interneurons within the PAG. There is little evidence to suggest that
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
projections from the CeA to the PAG contain opioids [13]. There are neurotensin immunoreactive cells located in the CeA projecting to the ventral PAG [13], and neurotensin interneurons within the PAG [4]. In addition, there is a population of neurotensin cells within the PAG that provides direct efferent connections with the RVM [5]. Although the results from these experiments provide information on the neurotransmitters involved in pain modulation within the ventral PAG, the intrinsic circuitry of the PAG as it receives input from the amygdala remains to be described. In summary, antinociception produced by injecting the mu opioid agonist DAMGO into the BLA is partly dependent on an additional mu opioid synapse within the ventral PAG. At least in anesthetized rats, it does not appear that delta opioid or NMDA receptors located in the ventral PAG are involved in amygdala antinociception. This study also suggests that neurotensin release within the ventral PAG plays a role in the antinociception produced by amygdala stimulation.
[9] [10]
[11]
[12]
[13]
[14] [15]
[16]
Acknowledgements This work was supported by a grant from the National Institute on Drug Abuse (DA 09429) and was performed as partial fulfillment for a PhD degree by the primary author. Portions of this data were previously presented at the Society for Neuroscience annual meetings of 1995 and 1996.
References [1] N.R. Al-Rodham, E. Richelson, J.A. Gilbert, D.J. McCormick, K.S. Kanba, M.A. Pfenning, A. Nelson, E.W. Larson, T.L. Yaksh, Structure–antinociceptive activity of neurotensin and some novel analogues in the periaqueductal gray region of the brainstem, Brain Res. 557 (1991) 227–235. [2] S.F. Atweh, M.J. Kuhar, Autoradiographic localization of opiate receptors in rat brain. III. The telencephalon, Brain Res. 134 (1977) 393–405. [3] A.I. Basbaum, H.L. Fields, Endogenous pain control systems: Brainstem spinal pathways and endorphin circuitry, Annu. Rev. Neurosci. 7 (1984) 309–338. [4] M.M. Behbehani, Physiological mechanisms of the analgesic effect of neurotensin, in: P. Kitabgi, C.B. Nemeroff (Eds.), The Neurobiology of Neurotesin, Annals of the New York Academy of Sciences, 1992, pp. 253–265. [5] M.M. Behbehani, A.A. Pert, A mechanism for the analgesic effect of neurotensin as revealed by behavioral and electrophysiological techniques, Brain Res. 324 (1984) 35–42. [6] A.J. Beitz, The nuclei of origin of brain stem enkephalin and substance P projections to the rodent nucleus raphe magnus, Neuroscience 11 (1982) 2753–2768. [7] P.S. Bellgowan, F.J. Helmstetter, The role of mu and kappa opioid receptors within the periaqueductal gray in the expression of conditional hypoalgesia, Brain Res. 791 (1998) 83–89. [8] R.J. Bodnar, C.L. Williams, S.J. Lee, G.W. Pasternak, Role of
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
25
m-opioid receptors in supraspinal opiate analgesia: a microinjection study, Brain Res. 447 (1988) 25–34. R.C. Bolles, M.S. Fanselow, A perceptual-defensive recuperative model of fear and pain, Behav. Brain Sci. 3 (1980) 291–301. T.J. Coderre, I. Van Empel, The utility of excitatory amino acid (EAA) antagonists as analgesic agents. I. Comparison of the antinociceptive activity of various classes of EAA antagonists in mechanical, thermal and chemical nociceptive tests, Pain 59 (1994) 345–352. T.M. Da Costa Gomez, M.M. Behbehani, An electrophysiological characterization of the projection from the central nucleus of the amygdala to the periaqueductal gray of the rat: the role of opioid receptors, Brain Res. 689 (1995) 21–31. M.S. Fanselow, D.J. Calcagnetti, F.J. Helmstetter, The role of mu and kappa opioid receptors in conditional fear-induced analgesia: The antagonistic actions of nor-binaltorphimine and the cyclic somatostatin octapeptide, CTOP, J. Pharmacol. Exp. Ther. 250 (1980) 825–830. T.S. Gray, D.J. Magnuson, Peptide immunoreactive neurons in the amygdala and the bed nucleus of the stria terminalis project to the midbrain central gray in the rat, Peptides 13 (1992) 451–460. F.J. Helmstetter, The amygdala is essential for the expression of conditional hypoalgesia, Behav. Neurosci. 106 (1992) 518–528. F.J. Helmstetter, P.S. Bellgowan, Lesions of the amygdala block conditional hypoalgesia on the tail flick test, Brain Res. 612 (1993) 253–257. F.J. Helmstetter, P.S. Bellgowan, L.H. Poore, Microinfusion of mu, but not delta or kappa opioid antagonists into the basolateral amygdala results in inhibition of the tail flick reflex in pentobarbital anesthetized rats, J. Pharmacol. Exp. Ther. 275 (1995) 381–388. F.J. Helmstetter, P.S. Bellgowan, S.A. Tershner, Inhibition of the tail flick reflex following microinjection of morphine into the amygdala, NeuroReport 4 (1993) 471–474. F.J. Helmstetter, M.S. Fanselow, Strain differences in reversal of conditional analgesia by opioid antagonists, Behav. Neurosci. 101 (1987) 735–737. F.J. Helmstetter, J. Landeira-Fernandez, Conditional hypoalgesia is attenuated by naltrexone applied to the periaqueductal gray, Brain Res. 537 (1990) 88–92. F.J. Helmstetter, S.A. Tershner, Lesions of the periaqueductal gray and rostral ventromedial medulla disrupt antinociception but not cardiovascular aversive conditional responses, J. Neurosci. 14 (1994) 7099–7108. F.J. Helmstetter, S.A. Tershner, L.H. Poore, P.S.F. Bellgowan, Antinociception following opioid stimulation of the basolateral amygdala is expressed through the periaqueductal gray and rostral ventromedial medulla, Brain Res. 779 (1998) 104–118. A. Hopkins, G. Holstege, Amygdaloid projections to the mesencephalon, pons and medulla oblongata in the cat, Exp. Brain Res. 32 (1978) 529–547. T.S. Jensen, T.L. Yaksh, The antinociceptive activity of excitatory amino acids in the rat brainstem: an anatomical and pharmacological analysis, Brain Res. 569 (1992) 255–267. T.S. Jensen, T.L. Yaksh, Comparison of the antinociceptive action of mu and delta opioid receptor ligands in the periaqueductal gray matter, medial and paramedial ventral medulla in the rat as studied by the microinjection technique, Brain Res. 372 (1986) 301–312. P. Kalen, M. Karlson, L. Wiklund, Possible excitatory amino acid afferents to nucleus raphe dorsalis of the rat investigated wheat germ agglutinin and D-[ 3 H]aspartate tracing, Brain Res. 360 (1985) 285– 297. I.B. Kinscheck, L.R. Watkins, D.J. Mayer, Fear is not critical to classically conditioned analgesia: The effects of periaqueductal gray lesions and administration of chlordiazepoxide, Brain Res. 298 (1984) 33–44. A.J. MacLennan, R.L. Jackson, S.F. Maier, Conditioned analgesia in the rat, Bull. Psychonom. Soc. 15 (1980) 387–390.
26
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
[28] M.H. Ossipov, C.J. Kovelowski, H. Wheeler-Aceto, A. Cowan, J.C. Hunter, J. Lai, T.P. Malan, F. Porreca, Opioid antagonists and antisera to endogenous opioids increase the antinociceptive response to formalin: demonstration of an opioid kappa and delta inhibitory tone, J. Pharmacol. Exp. Ther. 277 (1996) 784–788. [29] O. P Ottersen, J. Storm-Mathisen, Glutamate- and GABA-containing neurons in the mouse and rat brain as demonstrated with a new immunocytochemical technique, J. Comp. Neurol. 229 (1984) 374– 392. [30] C.M. Paden, S. Krall, W.C. Lynch, Heterogeneous distribution and upregulation of mu, delta, and kappa opioid receptors in the amygdala, Brain Res. 418 (1987) 349–355. [31] Z.W. Pavlovic, M.L. Cooper, R.J. Bodnar, Opioid antagonists in the periaqueductal gray inhibit morphine and beta-endorphin analgesia elicited from the amygdala of rats, Brain Res. 741 (1996) 13–26. [32] G. Paxinos, C. Watson (Eds.), The Rat Brain in Stereotaxic Coordinates, Vol. 2, Academic Press, New York, 1986. [33] A. Pert, T. Yaksh, Sites of morphine induced analgesia in the primate brain: relation to pain pathways, Brain Res. 180 (1974) 135–140. [34] T.A. Rizvi, M. Ennis, M.M. Behbehani, M.T. Shipley, Connections between the central nucleus of the amygdala and the midbrain periaqueductal gray: Topography and reciprocity, J. Comp. Neurol. 303 (1991) 121–131. [35] R.J. Rodgers, Elevation of aversive thresholds in rats by intraamygdaloid injection of morphine sulfate, Pharmacol. Biochem. Behav. 6 (1977) 385–390. [36] G.C. Rossi, G.W. Pasternak, R.J. Bodnar, Mu and delta opioid synergy between the periaqueductal gray and the rostro-ventral medulla, Brain Res. 665 (1994) 85–93. [37] M.T. Shipley, J.H. McLean, M.M. Behbehani, Heterogenous distribution of neurotensin-like immunoreactive neurones and fibers in the midbrain periaqueductal grey of the rat, J. Neurosci. 7 (1987) 2025–2034.
[38] D.J. Smith, J.M. Perrotti, T. Crisp, M.E. Y Cabral, J.T. Long, J.M. Scalzitti, The mu opiate receptor is responsible for descending pain inhibition originating in the periaqueductal gray region of the rat brain, Eur. J. Pharmacol. 156 (1988) 47–54. [39] D.J. Smith, B. Robertson, P.J. Monroe, D.A. Taylor, J.A. Leedham, J.D. Cabral, Opioid receptors mediating antinociception from Bendorphin and morphine in the periaqueductal gray, Neuropharmacology 31 (1992) 1137–1150. [40] M. Sofuoglu, P.S. Portoghese, A.E. Takemori, Differential antagonism of delta opioid agonists by naltrindole and its benzofuran analog (NTB) in mice: Evidence for delta opioid receptor types, J. Pharmacol. Exp. Ther. 257 (1991) 676–680. [41] A. Tempel, R.S. Zukin, Neuroanatomical patterns of the mu, delta, and kappa opioid receptors of rat brain as determined by quantitative in vitro autoradiography, Proc. Nat. Acad. Sci. USA 84 (1987) 4308–4312. [42] G.W. Terman, Y. Shavit, J.W. Lewis, J.T. Cannon, J.C. Liebeskind, Intrinsic mechanisms of pain inhibition: activation by stress, Science 226 (4680) (1984) 1270–1277. [43] S.A. Tershner, F.J. Helmstetter, Spinal antinociception following stimulation of the amygdala depends on opioid receptors in the ventral periaqueductal gray, Soc. Neurosci. Abstr. 1 (1995) 1. [44] M.O. Urban, D. Smith, Neurotensin in the nucleus raphe magnus in opioid antinociception from the periaqueductal gray, J. Pharmacol. Exp. Ther. 265 (1993) 580–586. [45] G. Urca, R.L. Nahin, J.C. Liebeskind, Glutamate-induced analgesia: Blockade and potentiation by naloxone, Brain Res. 192 (1980) 523–530. [46] T.L. Yaksh, J.C. Yeung, T.A. Rudy, 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. 114 (1976) 83–103.
Research report
Antinociception produced by mu opioid receptor activation in the amygdala is partly dependent on activation of mu opioid and neurotensin receptors in the ventral periaqueductal gray Sheralee A. Tershner a , *, Fred J. Helmstetter b b
a Department of Psychology, Western New England College, Springfield, MA 01119, USA Department of Psychology, University of Wisconsin–Milwaukee, Milwaukee, WI 53201, USA
Accepted 15 February 2000
Abstract Exposure to stressful or fear-inducing environmental stimuli activates descending antinociceptive systems resulting in a decreased pain response to peripheral noxious stimuli. Stimulating mu opioid receptors in the basolateral nucleus of the amygdala (BLA) in anesthetized rats produces antinociception that is similar to environmentally induced antinociception in awake rats. Recent evidence suggests that both forms of antinociception are mediated via projections from the amygdala to the ventral periaqueductal gray (PAG). In the present study, we examined the types of neurochemicals released in the ventral PAG that may be important in the expression of antinociception produced by amygdala stimulation in anesthetized rats. Microinjection of a mu opioid receptor agonist into the BLA resulted in a time dependent increase in tail flick latency that was attenuated by preadministration of a mu opioid receptor or a neurotensin receptor antagonist into the ventral PAG. Microinjection of a delta 2 opioid receptor antagonist or an NMDA receptor antagonist into the ventral PAG was ineffective. These findings suggest that amygdala stimulation produces antinociception that is mediated in part by opioid and neurotensin release within the ventral PAG. 2000 Elsevier Science B.V. All rights reserved. Themes: Sensory systems Topics: Pain modulation: pharmacology Keywords: Amygdala; Antinociception; Neurotensin; Opioid; Periaqueductal gray; Tail flick
1. Introduction Exposure to environmental stressors can inhibit ascending pain transmission [9,27,42]. Some forms of stress induced hypoalgesia are dependent on opioids [7,8,12,18,19] and are mediated via a descending antinociceptive circuit which includes the amygdala [14–16], periaqueductal gray (PAG) [7,19–21,26,36], and rostral ventromedial medulla (RVM) [20,36]. There is accumulating evidence that environmentally induced hypoalgesia is often a result of the activation of opioid receptors within the amygdala [17,33,35]. There is a heterogenous distribution of opioid receptors within the amygdaloid nuclei [2], with mu opioid receptors heavily
*Corresponding author. Tel.: 11-413-796-2193. E-mail address: [email protected] (S.A. Tershner)
localized within the BLA and lateral nuclei, while delta and kappa opioid receptors are only moderately represented. In comparison, there are relatively few opioid receptors of any type located within the central nucleus of the amygdala (CeA) [30]. We have recently demonstrated that activation of mu opioid receptors, not delta or kappa, in the amygdala produces antinociception in anesthetized rats [16]. The amygdala appears to modulate nociceptive responses by activating a descending antinociceptive pathway through the midbrain and medulla. Although both the CeA and BLA send neuronal projections to the ventral PAG, it appears that the primary input from the BLA to the ventral PAG includes a critical synapse within the CeA [22,34]. The ventral PAG in turn projects to nociceptive modulating neurons in the RVM [3]. Microinjection of the mu opioid receptor agonist DAMGO into the BLA produces inhibition of the tail flick (TF) reflex that is
0006-8993 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 00 )02179-X
18
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
blocked by lesions or lidocaine inactivation of the ventral PAG or RVM [21]. Although there is evidence for a serial connection between the amygdala and ventral PAG necessary for antinociception following opioid injection into the amygdala [21], little is known regarding the neurotransmitters involved. The ventral PAG contains opioid sensitive cells that when stimulated can produce antinociception [33,38,46]. There are mu and kappa, but relatively few delta opioid receptors located within the PAG, with the highest concentration of mu opioid receptors located in the ventral region [41]. A population of ventral PAG cells is excited following electrical stimulation of the amygdala, and this excitation is inhibited by the local application of an opioid antagonist [11]. In addition, injections of naltrexone into the ventral PAG in anesthetized rats reduce the antinociception produced by mu opioid receptor activation in the amygdala [43] and mu, but not kappa, opioid receptor antagonists injected into the ventral PAG block conditioned fear induced hypoalgesia in awake rats [7]. There appears to be a critical opioid synapse within the ventral PAG in antinociception produced by amygdala manipulation, however there is little evidence that direct afferents from the CeA to the PAG release opioids [13]. Therefore, it is likely that additional neurotransmitters are involved and released within the PAG by terminals of CeA cells as well as by local interneurons. Neurotensin-immunoreactive cells originating in the lateral regions of the CeA project to the caudal ventrolateral regions of the PAG [13]. Cell bodies containing neurotensin located in the ventral PAG [4,5,37] and neurotensinergic interneurons are believed to contribute to prolonged excitation of PAG neurons during antinociception [4]. Injections of neurotensin into the PAG can produce a dose-dependent inhibition of TF latency that is not reversible with naloxone and is thus believed to be independent of opioids [1,5]. In addition to the neurotensin there are excitatory amino acid (EAA) projections from the CeA to the PAG [25] and a substantial number of EAA immunoreactive terminals that contain glutamate and aspartate [6,29]. Site specific injections of glutamate and NMDA into the PAG produce antinociception in the TF test [23,45]. To examine whether ventral PAG opioid, neurotensin and NMDA receptor activation is necessary for the antinociception produced by amygdala stimulation their respective antagonists were microinjected into the ventral PAG in anesthetized rats prior to the administration of DAMGO into the BLA. We chose to use a reduced preparation that involves direct pharmacological activation of cell groups in this model circuit in anesthetized animals. It is believed that this neural circuit has evolved to subserve defensive and antinociceptive functions in the behaving animal and can be studied in this reduced manner in cases where it would be difficult to perform in behaving animals, e.g. experiments requiring multiple simultaneous intracerebral injections and cell recording. Results found
using this anesthetized preparation has yielded comparable results as those found in subsequent studies involving awake subjects [14,15,20] and is therefore considered a valid model for studying the circuit involved in environmentally induced analgesia.
2. Materials and methods
2.1. Subjects Subjects were 111 male Long–Evans rats (325–450 g) obtained from Harlan Sprague–Dawley (Madison, WI). Animals were housed individually in stainless steel cages with unlimited access to food and water. All experimental manipulations were performed during the light portion of a 14:10-h light / dark cycle.
2.2. Surgical preparation Prior to surgery, all animals were anesthetized with sodium pentobarbital (50 mg / kg / ml; i.p.). A catheter constructed from polyethylene tubing (PE-20) was inserted into the jugular vein at a depth of 1.0 cm. The vessel was secured around the tubing using 3-0 silk suture. Once implanted, the rat received a constant infusion of methohexital (10 mg / ml) delivered through the catheter at an infusion rate of 0.8 ml / h. The subjects were positioned in a David Kopf stereotaxic apparatus. Cannulae implantation consisted of exposing the surface of the skull and lowering 26 Gauge stainless steel guide cannulae into the BLA (bilateral) and ventral PAG (unilateral) using coordinates obtained from a rat brain atlas [32] (AP: 23.0; L: 1 / 25.2; V: 27.9 and AP: 27.8; L: 10.8; V: 26.0, respectively). All guide cannulae were secured to the skull with epoxy. Subjects remained anesthetized and in the stereotaxic apparatus during the entire testing session.
2.3. Apparatus Heart rate was recorded during the test period to measure possible effects of administered substances on cardiovascular activity. This variable also provided an index of the stability of anesthesia throughout the test session. Recording electrodes for measuring heart rate were positioned subcutaneously near the sternum ventrally and between the scapulae dorsally. Signals from the EKG electrodes were amplified (Grass Instruments) and recorded by a microcomputer with an A / D converter. Raw data waveforms were recorded and analyzed with software written with the ASYST data acquisition system. TF latency was measured using a custom made apparatus constructed from a stainless steel box with a 500-W projection bulb mounted inside. There was a 3-mm circular aperture in the top of the box that permitted
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
radiant heat from the bulb to make contact with the ventral side of the rat’s tail. A thermocouple positioned within the TF apparatus recorded stimulus intensity. The heat stimulus was set to elicit baseline TF latencies between 4.5 and 6.5 s resulting in a radiant heat pulse which reached 45.68C in 20 s. This heat setting was held constant between all subjects and throughout the testing period. A cooling fan attached to the apparatus maintained the surface of the apparatus between 28 and 328C. When the rat’s tail deflected from the heat source a photocell shut off a digital timer and deactivated the stimulus. If a TF response did not occur within 20 s, the heat stimulus was terminated to prevent tissue damage.
2.4. Drugs and injection procedure H-D-Phe-Cys-Tyr-D-Trp-Arg-Pen-Thr-NH2 (CTAP) (Multiple Peptide Systems; 0.0005, 0.005, and 0.05 mg / 0.25 ml); and naltriben (NTB) (Research Biochemicals Inc.; 0.12, 1.2, 4.0 mg / 0.25 ml) were used to examine the role of mu and delta opioid receptors, respectively, within the ventral PAG in amygdala induced antinociception [7,38–40]. The partial neurotensin antagonist [D-Trp 11 ] neurotensin (NTA) (Sigma; 4.0 pmol (1.69310 23 pg), 16 pmol (6.78310 23 pg), and 64 pmol (2.71310 22 pg) / 0.25 ml), and an NMDA receptor antagonist DL-2-amino-5phosphono-valerate (APV) (Research Biochemicals Inc.; 0.03, 0.30, and 3.0 mg / 0.25 ml) were injected into the ventral PAG to examine the role of neurotensin and NMDA receptor activation in amygdala antinociception [23,44]. The selective mu opioid receptor agonist D-Ala 2 N-MePhe 4 -Gly-Ol 5 enkephalin (DAMGO) (Sigma; 0.5 mg / 0.25 ml / side) was injected into the BLA to produce antinociception [16]. All drugs were dissolved in isotonic sterile saline. A pooled control group which received saline injections into the ventral PAG and DAMGO into the BLA (SAL-DAMGO) was used for statistical comparisons in the analyses of all receptor antagonists. Subjects assigned to the SAL-DAMGO control group were tested concurrently with the testing of the subjects in the other treatment groups. Drug delivery was performed with an injector constructed from a 33-Gauge stainless steel cannula connected to a piece of PE-20 tubing. The injector was attached to a 1.0-ml Hamilton microsyringe mounted on a Harvard Apparatus infusion pump. All drugs were delivered in a volume of 0.25 ml over a 45-s period.
2.5. Testing procedure Once the rat was prepared for the experiment, a standard protocol was implemented. The test session consisted of 30 TF trials with an intertrial interval of 120 s. Heart rate was recorded during the twenty seconds prior to and the twenty seconds following the onset of the heat stimulus. Four
19
baseline TF trials were recorded before drug administration. Immediately after trial four (time58 min), an opioid receptor antagonist or isotonic sterile saline was infused into the ventral PAG followed by DAMGO injections into the BLA at the end of the fifth trial (time510 min).
2.6. Histology and data analyses After testing rats were overdosed with an intraperitoneal injection of euthanasia solution consisting of ethanol and sodium pentobarbital (100 mg / kg). They were then perfused transcardially with 0.9% saline followed by 10% phosphate buffered formalin. Following decapitation, the heads were left in buffered formalin for 24 h before brain removal. The brains were soaked in sucrose / formalin for an additional 24 h. To verify cannula placements the brains were sliced into coronal sections (40 mm), mounted on slides, and stained using cresyl violet. After reviewing the histology, only animals with injection sites located within the ventral PAG and bilaterally within the BLA, as defined by Paxinos and Watson [32], were included in the analyses. Amygdala injection sites were considered acceptable if both cannulae were located in the region of the BLA or the ventral portions of the lateral nucleus. PAG injection sites were considered acceptable if they were located in the ventral portion of the PAG and lateral to the dorsal raphe nucleus. See Fig. 1A and B for a schemata depicting the areas of acceptance within the amygdala and ventral PAG, respectively. TF and heart rate responses were subjected to within and between subjects ANOVA. TF latency was measured in seconds (0–20 s) and heart rate values were recorded in beats / min and transformed to reflect change from baseline values. Tukey post-hoc tests (P,0.05) were performed for between group comparisons when needed.
3. Results
3.1. Histology After examination of the histology, rats with cannulae placements outside of the regions of acceptance, as described in Materials and methods section, were excluded from the data analysis. There were five rats in each of the CTAP drug conditions, seven subjects in the SALDAMGO control group, five subjects in each of the NTB drug conditions, four subjects in the NTB-SAL group, five subjects in each of the NTA drug conditions, four subjects in the NTA-SAL group, and five subjects in each of the APV drug conditions. Cannulae placements with the BLA and PAG can be seen in Fig. 1A and B, respectively.
3.2. Mu opioid receptor antagonist Microinjections of CTAP into the ventral PAG at-
20
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
Fig. 1. Schemata representation of the areas of acceptance for cannulae placements. All BLA injections were made bilaterally (panel A). PAG injections were made unilaterally and were counterbalanced across subjects for side of injection site (panel B).
tenuated the increase in TF latency produced by DAMGO injections in the BLA (see Fig. 2). An ANOVA on the mean of the four TF latencies prior to drug administration indicated that baseline nociception between the groups was not different, F(4,22)51.41, P.0.05. DAMGO injections into the BLA produced a time dependent inhibition of TF latency in rats that received saline injections into the PAG. The intermediate (0.005 mg) and high (0.05 mg) doses of CTAP injected into the PAG attenuated the increase in TF latency resulting from DAMGO administration in the BLA. These observations were supported by an ANOVA for repeated measures on post injection TF latencies which yielded a main effect for group, F(4,22)514.73, P,0.05, a significant trial effect, F(25,550)514.80, P,0.05, and a significant trial by group interaction, F(100,550)52.99, P,0.05. All doses of CTAP tested delayed the onset of TF inhibition, post-hoc Tukey analyses (P,0.05), but only the intermediate and higher doses continued to decrease the inhibition throughout the session (P,0.05). The rats receiving saline injections into the BLA and the high dose of CTAP into the PAG did not show any change in TF latency demonstrating that this dose of CTAP is not
producing hyperalgesia. Heart rate was not different in any of the groups either before F(4,19)52.017, P.0.05, or after drug administration F(4,19),1.0.
3.3. Delta receptor antagonist None of the doses of NTB, when injected in to the PAG were effective in attenuating antinociception produced by DAMGO administration in the BLA (see Fig. 3). An ANOVA failed to show any significant differences between the groups on TF latency before drug administrations F(4,21)51.070, P.0.05. All groups except for the NTBSAL group showed a time dependent increase in TF latency indicated by a main effect for group F(4,21)5 7.458, P,0.05, trial F(25,525)516.023, P,0.05, and a group by trial interaction F(100,525)52.125, P,0.05. Surprisingly, at 24 min after DAMGO injection the low dose of NTB (0.12 mg) potentiated the increase in TF latency, post-hoc Tukey comparison (P,0.05). Heart rate was not different in any of the groups either before F(4,19),1.0, or after drug administration F(4,19),1.0.
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
21
Fig. 2. Effects of injecting a mu opioid receptor antagonist into the ventral PAG on antinociception produced by DAMGO injections into the BLA. DAMGO administration alone produced a time-dependent increase in TF latency which lasted the duration of the test session. This increase in TF latency was attenuated by pretreatment with CTAP in the ventral PAG. Vertical lines denote a standard error of the mean.
3.4. Neurotensin receptor antagonist An ANOVA on mean baseline TF latencies indicated that the groups were not significantly different before drug administration F(4, 21)52.712, P.0.05. Ventral PAG injections of the NTA attenuated the increase in TF latency produced by activating mu opioid receptors within the BLA (see Fig. 4). TF trials measured after drug administration were subjected to an ANOVA for repeated measures which showed that there was a significant main effect for group F(4,21)54.942, P,0.05, trials F(25,525)512.092, P,0.05, and a drug by trial interaction F(100,525)51.701. The low and intermediate dose of NTA (0.0016 and
0.0067 pg, respectively) injected into the PAG delayed the onset of TF inhibition produced by DAMGO administration in the BLA as measured at time points 10–18 min after DAMGO injection, post-hoc Tukey analyses (P, 0.05). At the end of the session the attenuation of TF inhibition with these two doses of NTA was still evident, P,0.05. TF latencies in the NTA-SAL group did not increase, thus providing evidence that the intermediate dose of NTA is not apposing DAMGO antinociception by producing hyperalgesia. Although the highest dose of NTA did not have an effect on TF latency it did increased HR throughout the test session (see Fig. 5). An ANOVA for repeated measures
Fig. 3. Effects of injecting a delta 2 opioid receptor antagonist into the ventral PAG on antinociception produced by DAMGO injections into the BLA. NTB, at the doses examined in this experiment, did not attenuate the increase in TF latency produced by DAMGO injections into the BLA. Vertical lines denote a standard error of the mean.
22
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
Fig. 4. Effects of injecting a neurotensin receptor antagonist into the ventral PAG on antinociception produced by DAMGO injections into the BLA. DAMGO administration alone produced a time-dependent increase in TF latency which lasted the duration of the test session. This increase in TF latency was attenuated by pretreatment with NTA in the ventral PAG. Vertical lines denote a standard error of the mean.
on post drug trials failed to reveal a significant main effect for group F(4,20)52.724, P.0.05, but did reveal a significant trial effect F(25,500)52.640, P,0.05, and a significant group by trial interaction F(100,500)52.038, P,0.05. Post-hoc Tukey analyses (P,0.05) revealed that 16 min after drug administration HR in subjects that received a high dose of NTA was significantly higher than the SAL-DAMGO controls. None of the other comparisons were significant at this time point. HR differences between
these two groups persisted throughout the duration of the test session. HR measures taken prior to drug administrations were not different between the groups F(4, 20),1.0.
3.5. NMDA receptor antagonist An ANOVA on the baseline TF latencies of the groups indicated that there were significant differences between the groups before drug administration, F(4,22)52.915,
Fig. 5. HR recordings were transformed to reflect change from baseline HR values after drug administration. The high dose of NTA (64 pmol) produced a slight increase in HR that lasted the duration of the test session. Vertical lines denote a standard error of the mean.
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
P50.045. There is no definite explanation for the statistical difference found, however after examining the data it appears that the difference may be due to the extremely small variance in the group that received the intermediate dose of APV. Upon inspection of the data we believe that the finding is a statistical anomaly. None of the doses of APV injected in the PAG attenuated the increase in TF latency produced by DAMGO injections into the BLA (see Fig. 6). TF latencies of the subjects that received the high (3.0 mg) dose of APV into the PAG increased immediately following administration. These observations were supported by an ANOVA for repeated measures on post drug TF trials which yielded a significant main effect F(4,22)5 5.249, P,0.05, trial effect F(25,550)514.740, P,0.05 and a group by trial interaction F(100,550)51.953, P, 0.05. Post-hoc Tukey comparisons (P,0.05) between the groups revealed that the highest dose of APV facilitated the onset of TF inhibition. The lack of an antinociceptive effect of the high dose of APV in the absence of DAMGO suggests that the increase in the onset of TF inhibition does not reflect a summation of a primary analgesic effect of the drug. An ANOVA on mean baseline HR measures indicated that the groups were not different prior to drug administration F(4,20)51.708, P.0.05. Rats that received APV following DAMGO injections into the BLA showed a small tachycardiac response (less than a 25 beat / min increase) which persisted throughout the remainder of the test session indicated by a significant trial by group interaction F(100,500)52.296, P,0.05. (see Fig. 7). The high dose of APV (3.0 mg) did not have an effect on HR
23
when administered alone, suggesting that the ability of APV to increase HR is dependent on the coadministration of APV and DAMGO.
4. Discussion The purpose of these experiments was to determine which neurotransmitters are released within the PAG in response to mu opioid receptor stimulation in the BLA. By injecting selective opioid receptor antagonists into the PAG this experiment demonstrated that mu opioid receptors are involved on the expression of DAMGO induced antinociception from the BLA. This finding in anesthetized rats compliments the results of similar studies in awake rats. Fear-induced antinociception is mediated by opioid release within the PAG [19], and more specifically through the activation of mu opioid receptors within the ventral PAG [7]. In addition, a study by Pavlovic et al. [31] demonstrated that analgesia produced by injecting morphine into the amygdala of awake rats was blocked by CTAP injections into the PAG. Taken together, these results clearly demonstrate the importance of a mu opioid receptor link within the ventral PAG in antinociception produced by the amygdala. The reports on the effectiveness of delta opioid agonists producing antinociception when injected into the PAG have been mixed. Some studies have failed to show delta opioid antinociception [28], while others have found antinociceptive effects within the PAG [24,36]. The present lack of evidence suggesting a role for PAG delta opioid
Fig. 6. Effects of injecting an NMDA receptor antagonist into the ventral PAG on antinociception produced by DAMGO injections into the BLA. An ANOVA revealed that the TF latencies were different between groups before drug administration. However, it is clear that this is merely a statistical anomaly and not a meaningful difference. APV injected into the ventral PAG did not attenuate the increase in TF latency produced by DAMGO injections into the BLA. Vertical lines denote a standard error of the mean.
24
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
Fig. 7. HR recordings were transformed to reflect change from baseline HR values after drug administration. All doses of APV tested produced a slight increase in HR that lasted the duration of the test session. Vertical lines denote a standard error of the mean.
receptors in amygdala antinociception directly contradicts the aforementioned study by Pavlovic et al. [31]. They found that intra-PAG injections of a delta 2 opioid receptor antagonist were effective in blocking antinociception produced by amygdala stimulation. Although there are some significant methodological differences between our studies, the reason for this discrepancy is not clear. Their study used awake and freely moving rats, different tests for pain sensitivity, and a delta 2 opioid receptor antagonist (naltrindole isothiocyanate) that required administration 24 h prior to testing. Any of these factors, or a combination of several, may have contributed to the differences between our findings. Since it is difficult to use naltrindole isothiocyanate in an acute preparation, additional behavioral studies using naltriben, at similar doses and using the TF test, are needed to clarify the role of PAG delta 2 opioid receptors in amygdala antinociception. Finally, the facilitation of TF inhibition by a delta opioid antagonist is difficult to explain since there is little or no evidence for delta opioid antagonists producing analgesia or enhancing mu agonist analgesia Selectively blocking mu receptors within the PAG did not completely block the antinociceptive input from the amygdala suggesting the involvement of multiple pain inhibitory processes. The results from this study suggest that in addition to a critical opioid synapse within the ventral PAG, neurotensin release within the ventral PAG is important for amygdala antinociception. The low and intermediate doses of NTA (0.0016 and 0.0067 pg) delayed the onset of the increase in TF latency following DAMGO administration and continued to attenuate the response throughout the test session. These doses of NTA did not produce changes in HR. The highest dose of NTA (0.0271 pg) was ineffective in attenuating TF inhibition
but did produce an increase in HR that lasted the duration of the test session. The partial agonist–antagonist properties of the NTA makes it very difficult to examine the endogenous release of neurotensin. Doses similar to those used in the present experiment were shown to block antinociception produced by neurotensin administration [44]. Since the NTA is not a true antagonist but a partial agonist, higher doses of NTA (100–300 pmol) injected into the RVM produced TF inhibition and low doses of NTA (3 pmol) decreased TF inhibition produced by neurotensin. Interestingly, Urban and Smith [44] reported that the same dose of NTA (3 pmol) injected into the RVM enhanced the TF inhibition produced by morphine administration into the PAG. It is quite possible that activation of neurotensin receptors may have inhibitory and facilitatory effects on nociceptive processes depending on the location of the receptors. The results of this experiment suggest that NMDA receptors within the PAG are not involved in pain inhibition produced by amygdala stimulation. None of the doses tested were successful in attenuating or blocking the increase in TF latency produced by DAMGO injections into the BLA. TF inhibition appeared to be facilitated in the group receiving the highest dose of APV (3.0 mg). Although, the same dose of APV (3.0 mg) did not produce antinociception when administered alone, a 10-mg dose of APV has been shown to produce analgesia in the formalin test when administered intrathecally [10]. It is possible that 3.0 mg of APV, with out producing antinociception when administered into the ventral PAG, could produce a synergistic enhancement of mu opioid analgesia. It is still not known if opioids and neurotensin are being released by afferent projections or by local interneurons within the PAG. There is little evidence to suggest that
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
projections from the CeA to the PAG contain opioids [13]. There are neurotensin immunoreactive cells located in the CeA projecting to the ventral PAG [13], and neurotensin interneurons within the PAG [4]. In addition, there is a population of neurotensin cells within the PAG that provides direct efferent connections with the RVM [5]. Although the results from these experiments provide information on the neurotransmitters involved in pain modulation within the ventral PAG, the intrinsic circuitry of the PAG as it receives input from the amygdala remains to be described. In summary, antinociception produced by injecting the mu opioid agonist DAMGO into the BLA is partly dependent on an additional mu opioid synapse within the ventral PAG. At least in anesthetized rats, it does not appear that delta opioid or NMDA receptors located in the ventral PAG are involved in amygdala antinociception. This study also suggests that neurotensin release within the ventral PAG plays a role in the antinociception produced by amygdala stimulation.
[9] [10]
[11]
[12]
[13]
[14] [15]
[16]
Acknowledgements This work was supported by a grant from the National Institute on Drug Abuse (DA 09429) and was performed as partial fulfillment for a PhD degree by the primary author. Portions of this data were previously presented at the Society for Neuroscience annual meetings of 1995 and 1996.
References [1] N.R. Al-Rodham, E. Richelson, J.A. Gilbert, D.J. McCormick, K.S. Kanba, M.A. Pfenning, A. Nelson, E.W. Larson, T.L. Yaksh, Structure–antinociceptive activity of neurotensin and some novel analogues in the periaqueductal gray region of the brainstem, Brain Res. 557 (1991) 227–235. [2] S.F. Atweh, M.J. Kuhar, Autoradiographic localization of opiate receptors in rat brain. III. The telencephalon, Brain Res. 134 (1977) 393–405. [3] A.I. Basbaum, H.L. Fields, Endogenous pain control systems: Brainstem spinal pathways and endorphin circuitry, Annu. Rev. Neurosci. 7 (1984) 309–338. [4] M.M. Behbehani, Physiological mechanisms of the analgesic effect of neurotensin, in: P. Kitabgi, C.B. Nemeroff (Eds.), The Neurobiology of Neurotesin, Annals of the New York Academy of Sciences, 1992, pp. 253–265. [5] M.M. Behbehani, A.A. Pert, A mechanism for the analgesic effect of neurotensin as revealed by behavioral and electrophysiological techniques, Brain Res. 324 (1984) 35–42. [6] A.J. Beitz, The nuclei of origin of brain stem enkephalin and substance P projections to the rodent nucleus raphe magnus, Neuroscience 11 (1982) 2753–2768. [7] P.S. Bellgowan, F.J. Helmstetter, The role of mu and kappa opioid receptors within the periaqueductal gray in the expression of conditional hypoalgesia, Brain Res. 791 (1998) 83–89. [8] R.J. Bodnar, C.L. Williams, S.J. Lee, G.W. Pasternak, Role of
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
25
m-opioid receptors in supraspinal opiate analgesia: a microinjection study, Brain Res. 447 (1988) 25–34. R.C. Bolles, M.S. Fanselow, A perceptual-defensive recuperative model of fear and pain, Behav. Brain Sci. 3 (1980) 291–301. T.J. Coderre, I. Van Empel, The utility of excitatory amino acid (EAA) antagonists as analgesic agents. I. Comparison of the antinociceptive activity of various classes of EAA antagonists in mechanical, thermal and chemical nociceptive tests, Pain 59 (1994) 345–352. T.M. Da Costa Gomez, M.M. Behbehani, An electrophysiological characterization of the projection from the central nucleus of the amygdala to the periaqueductal gray of the rat: the role of opioid receptors, Brain Res. 689 (1995) 21–31. M.S. Fanselow, D.J. Calcagnetti, F.J. Helmstetter, The role of mu and kappa opioid receptors in conditional fear-induced analgesia: The antagonistic actions of nor-binaltorphimine and the cyclic somatostatin octapeptide, CTOP, J. Pharmacol. Exp. Ther. 250 (1980) 825–830. T.S. Gray, D.J. Magnuson, Peptide immunoreactive neurons in the amygdala and the bed nucleus of the stria terminalis project to the midbrain central gray in the rat, Peptides 13 (1992) 451–460. F.J. Helmstetter, The amygdala is essential for the expression of conditional hypoalgesia, Behav. Neurosci. 106 (1992) 518–528. F.J. Helmstetter, P.S. Bellgowan, Lesions of the amygdala block conditional hypoalgesia on the tail flick test, Brain Res. 612 (1993) 253–257. F.J. Helmstetter, P.S. Bellgowan, L.H. Poore, Microinfusion of mu, but not delta or kappa opioid antagonists into the basolateral amygdala results in inhibition of the tail flick reflex in pentobarbital anesthetized rats, J. Pharmacol. Exp. Ther. 275 (1995) 381–388. F.J. Helmstetter, P.S. Bellgowan, S.A. Tershner, Inhibition of the tail flick reflex following microinjection of morphine into the amygdala, NeuroReport 4 (1993) 471–474. F.J. Helmstetter, M.S. Fanselow, Strain differences in reversal of conditional analgesia by opioid antagonists, Behav. Neurosci. 101 (1987) 735–737. F.J. Helmstetter, J. Landeira-Fernandez, Conditional hypoalgesia is attenuated by naltrexone applied to the periaqueductal gray, Brain Res. 537 (1990) 88–92. F.J. Helmstetter, S.A. Tershner, Lesions of the periaqueductal gray and rostral ventromedial medulla disrupt antinociception but not cardiovascular aversive conditional responses, J. Neurosci. 14 (1994) 7099–7108. F.J. Helmstetter, S.A. Tershner, L.H. Poore, P.S.F. Bellgowan, Antinociception following opioid stimulation of the basolateral amygdala is expressed through the periaqueductal gray and rostral ventromedial medulla, Brain Res. 779 (1998) 104–118. A. Hopkins, G. Holstege, Amygdaloid projections to the mesencephalon, pons and medulla oblongata in the cat, Exp. Brain Res. 32 (1978) 529–547. T.S. Jensen, T.L. Yaksh, The antinociceptive activity of excitatory amino acids in the rat brainstem: an anatomical and pharmacological analysis, Brain Res. 569 (1992) 255–267. T.S. Jensen, T.L. Yaksh, Comparison of the antinociceptive action of mu and delta opioid receptor ligands in the periaqueductal gray matter, medial and paramedial ventral medulla in the rat as studied by the microinjection technique, Brain Res. 372 (1986) 301–312. P. Kalen, M. Karlson, L. Wiklund, Possible excitatory amino acid afferents to nucleus raphe dorsalis of the rat investigated wheat germ agglutinin and D-[ 3 H]aspartate tracing, Brain Res. 360 (1985) 285– 297. I.B. Kinscheck, L.R. Watkins, D.J. Mayer, Fear is not critical to classically conditioned analgesia: The effects of periaqueductal gray lesions and administration of chlordiazepoxide, Brain Res. 298 (1984) 33–44. A.J. MacLennan, R.L. Jackson, S.F. Maier, Conditioned analgesia in the rat, Bull. Psychonom. Soc. 15 (1980) 387–390.
26
S. A. Tershner, F. J. Helmstetter / Brain Research 865 (2000) 17 – 26
[28] M.H. Ossipov, C.J. Kovelowski, H. Wheeler-Aceto, A. Cowan, J.C. Hunter, J. Lai, T.P. Malan, F. Porreca, Opioid antagonists and antisera to endogenous opioids increase the antinociceptive response to formalin: demonstration of an opioid kappa and delta inhibitory tone, J. Pharmacol. Exp. Ther. 277 (1996) 784–788. [29] O. P Ottersen, J. Storm-Mathisen, Glutamate- and GABA-containing neurons in the mouse and rat brain as demonstrated with a new immunocytochemical technique, J. Comp. Neurol. 229 (1984) 374– 392. [30] C.M. Paden, S. Krall, W.C. Lynch, Heterogeneous distribution and upregulation of mu, delta, and kappa opioid receptors in the amygdala, Brain Res. 418 (1987) 349–355. [31] Z.W. Pavlovic, M.L. Cooper, R.J. Bodnar, Opioid antagonists in the periaqueductal gray inhibit morphine and beta-endorphin analgesia elicited from the amygdala of rats, Brain Res. 741 (1996) 13–26. [32] G. Paxinos, C. Watson (Eds.), The Rat Brain in Stereotaxic Coordinates, Vol. 2, Academic Press, New York, 1986. [33] A. Pert, T. Yaksh, Sites of morphine induced analgesia in the primate brain: relation to pain pathways, Brain Res. 180 (1974) 135–140. [34] T.A. Rizvi, M. Ennis, M.M. Behbehani, M.T. Shipley, Connections between the central nucleus of the amygdala and the midbrain periaqueductal gray: Topography and reciprocity, J. Comp. Neurol. 303 (1991) 121–131. [35] R.J. Rodgers, Elevation of aversive thresholds in rats by intraamygdaloid injection of morphine sulfate, Pharmacol. Biochem. Behav. 6 (1977) 385–390. [36] G.C. Rossi, G.W. Pasternak, R.J. Bodnar, Mu and delta opioid synergy between the periaqueductal gray and the rostro-ventral medulla, Brain Res. 665 (1994) 85–93. [37] M.T. Shipley, J.H. McLean, M.M. Behbehani, Heterogenous distribution of neurotensin-like immunoreactive neurones and fibers in the midbrain periaqueductal grey of the rat, J. Neurosci. 7 (1987) 2025–2034.
[38] D.J. Smith, J.M. Perrotti, T. Crisp, M.E. Y Cabral, J.T. Long, J.M. Scalzitti, The mu opiate receptor is responsible for descending pain inhibition originating in the periaqueductal gray region of the rat brain, Eur. J. Pharmacol. 156 (1988) 47–54. [39] D.J. Smith, B. Robertson, P.J. Monroe, D.A. Taylor, J.A. Leedham, J.D. Cabral, Opioid receptors mediating antinociception from Bendorphin and morphine in the periaqueductal gray, Neuropharmacology 31 (1992) 1137–1150. [40] M. Sofuoglu, P.S. Portoghese, A.E. Takemori, Differential antagonism of delta opioid agonists by naltrindole and its benzofuran analog (NTB) in mice: Evidence for delta opioid receptor types, J. Pharmacol. Exp. Ther. 257 (1991) 676–680. [41] A. Tempel, R.S. Zukin, Neuroanatomical patterns of the mu, delta, and kappa opioid receptors of rat brain as determined by quantitative in vitro autoradiography, Proc. Nat. Acad. Sci. USA 84 (1987) 4308–4312. [42] G.W. Terman, Y. Shavit, J.W. Lewis, J.T. Cannon, J.C. Liebeskind, Intrinsic mechanisms of pain inhibition: activation by stress, Science 226 (4680) (1984) 1270–1277. [43] S.A. Tershner, F.J. Helmstetter, Spinal antinociception following stimulation of the amygdala depends on opioid receptors in the ventral periaqueductal gray, Soc. Neurosci. Abstr. 1 (1995) 1. [44] M.O. Urban, D. Smith, Neurotensin in the nucleus raphe magnus in opioid antinociception from the periaqueductal gray, J. Pharmacol. Exp. Ther. 265 (1993) 580–586. [45] G. Urca, R.L. Nahin, J.C. Liebeskind, Glutamate-induced analgesia: Blockade and potentiation by naloxone, Brain Res. 192 (1980) 523–530. [46] T.L. Yaksh, J.C. Yeung, T.A. Rudy, 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. 114 (1976) 83–103.