Neuropathic pain attenuates opioid facilitation of rewarding electrical stimulation of limbic dopaminergic pathways originating from the ventral tegmental area. Whether neuropathic pain alters opioid effects of other brain-reward systems is unknown.
Control and spinal nerve-ligated (SNL) rats had electrodes implanted into the paraventricular nucleus (PVN) of the hypothalamus or medial forebrain bundle. Control and SNL rats were trained to lever-press for intracranial self-stimulation (ICSS), and modulation by morphine or cocaine was assessed.
Control and SNL rats lever-pressed for stimulation of the PVN and medial forebrain bundle. Morphine produced greater reductions in the frequency at which rats emitted 50% of maximal responding for PVN ICSS (maximal effect 24.67 ± 4.60 [mean ± SEM] and 24.11 ± 5.96 in SNL [n = 6] and control [n = 8] rats, respectively, compared with medial forebrain bundle ICSS (12.38 ± 6.77 [n = 8] and 12.69 ± 1.55 [n = 7]). In contrast, cocaine was less efficacious in potentiating PVN ICSS (maximal effect 11.76 ± 2.86 and 12.38 ± 4.01 in SNL [n = 12] and control [n = 8] rats, respectively) compared with medial forebrain bundle ICSS (30.58 ± 3.40 [n = 9] and 27.55 ± 4.51 [n = 7]).
PVN ICSS is facilitated to a greater extent by morphine than cocaine, and the effects of each drug on this behavior are unaltered after spinal nerve ligation. These effects contrast those observed with direct stimulation of limbic dopamine pathways, suggesting that the PVN may have a greater role in the reinforcing effects of opioids than classic limbic regions, particularly in the presence of chronic pain.
What We Already Know about This Topic
Drugs that produce addiction facilitate self-stimulation from various brain regions
Facilitation of brain self-stimulation by opioids is attenuated in a chronic neuropathic pain model
What This Article Tells Us That Is New
Stimulation of the paraventricular nucleus was enhanced more by morphine than by cocaine and morphine remained enhanced in a chronic neuropathic pain model
Understanding the role of the paraventricular nucleus and the reinforcement of opioids could improve our understanding of addiction in chronic pain
INTRACRANIAL self-stimulation (ICSS) is an operant paradigm pairing lever presses with electrical stimulation of discrete brain pathways.1Rats will lever-press for stimulation of brain regions within the mesolimbic dopamine system, including the ventral tegmental area (VTA) and medial forebrain bundle (MFB).2Drugs of abuse such as opioids and cocaine, whose reinforcing effects are mediated through this circuitry, potentiate the reinforcing effects of VTA and MFB ICSS in rats.2Although neuropathic pain does not alter VTA ICSS, morphine, heroin, and commonly abused prescription opioids are less effective in facilitating VTA ICSS in spinal nerve-ligated (SNL) rats compared with control animals.3,4It is not known if this effect of SNL is limited to facilitation of ICSS in the VTA.
The hypothalamic paraventricular nucleus (PVN) may play an important role in the reinforcing effects of drugs of abuse through oxytocin release in the limbic system.5Oxytocin release is stimulated in response to various environmental stimuli and is involved in a host of behavioral effects including maternal and social bonding, anxiety, and pain.6A lesser understood effect of oxytocin is its ability to modulate behavioral effects of drugs of abuse.7These effects likely arise from release of oxytocin from the PVN into limbic and forebrain regions, where oxytocin may interact with dopamine neurotransmission to modulate aspects of drug addiction.5For instance, PVN oxytocin fibers project to the VTA,8which expresses oxytocin receptor messenger RNA,9and intra-VTA oxytocin stimulates dopamine release in the nucleus accumbens that is reversed by pretreatment with an intra-VTA oxytocin receptor antagonist.8Intracerebroventricular administration of oxytocin has been shown to inhibit analgesic tolerance to heroin10and morphine.11In addition, intravenous self-administration of heroin is reduced after intraaccumbens or intrahippocampal oxytocin administration in rats,12although it is unclear if this reflects changes in tolerance or in the rewarding properties of heroin. Oxytocin administered intracerebroventricularly has also been shown to inhibit some of the acute behavioral effects of cocaine, including stereotypy.13It is therefore clear that oxytocin has the potential to modulate effects of drugs of abuse, although it is unclear to what extent the reverse is true.
It is hypothesized that electrical stimulation of the PVN will produce reinforcing effects that may be mediated to some extent through indirect stimulation of mesolimbic dopamine via oxytocin release. It is therefore expected that drugs of abuse, which stimulate mesolimbic dopaminergic neurons, would subsequently enhance the reinforcing effects of PVN ICSS. The current study sought to address this issue by assessing the effects of morphine and cocaine on the reinforcing effects of PVN ICSS. Given the close proximity of the PVN to the MFB, the modulatory effects of morphine and cocaine on the reinforcing effects of MFB ICSS were also assessed. It is also unclear to what extent nerve injury, which diminishes the potentiating effects of opioids for VTA ICSS,3,4will alter the ability of morphine and cocaine to facilitate PVN or MFB ICSS. Therefore, studies were performed in both normal and nerve-injured rats.
Materials and Methods
Subjects were 39 male Fisher 344 rats (12 SNL and 11 control rats for PVN experiments; 9 SNL and 7 control rats for MFB experiments) weighing between 275–325 g at the beginning of the experiment (Harlan Laboratories, Raleigh, NC). Rats were group-housed in a temperature- and humidity-controlled room that was maintained on a reverse light-dark cycle (dark 5:00 PM–5:00 AM); this room was adjacent to the room in which behavioral experiments were performed. Food and water were continuously available except during behavioral experiments. All procedures were conducted according to guidelines of the Committee for Research and Ethical Issues of the International Association for the Study of Pain and were approved by the Animal Care and Use Committee of Wake Forest University School of Medicine, Winston-Salem, North Carolina.
Electrodes were permanently implanted in the rats as previously described.3Briefly, rats were anesthetized with pentobarbital (50 mg/kg, intraperitoneal) and atropine methyl nitrate (10 mg/kg, intraperitoneal) and received penicillin G procaine (75,000 U, intramuscular) to prevent infection. Rats were secured in a stereotaxic frame and platinum bipolar stimulating electrodes (Plastics One, Roanoke, VA) were aimed at the left PVN or left MFB (2.5 mm and 3.0 mm posterior to bregma, 0.3 mm and 1.7 mm lateral from the midline, and 8.0 mm and 8.5 mm below the skull, for PVN and MFB, respectively), which were secured by three stainless-steel screws embedded in dental acrylic on the skull surface.
Spinal Nerve Ligation.
After electrode implantation, 21 of the 39 rats were subjected to SNL as previously described.14Briefly, an incision was made through the skin and muscle of the lower back, and the left transverse process of the fifth lumbar vertebra was removed using bone microrongeurs. The fifth lumbar nerve was exteriorized and ligated with 4–0 silk suture. The sixth lumbar nerve was exteriorized from below the iliac bone at the sciatic notch and similarly ligated. Afterward, muscle layers were sutured with 4–0 chromic gut suture, the skin with 4–0 nylon suture, and exterior wounds dressed with antibiotic powder (Polysporin; Pfizer Healthcare, Morris Plains, NJ).
Paw Withdrawal Threshold
Mechanical allodynia was assessed using von Frey filaments (Touch Test Sensory Evaluators; Stoelting, Wood Dale, IL)15for all animals using Dixon nonparametric statistics.16Paw withdrawal thresholds (PWTs) were assessed 14 days after surgery to verify development of allodynia (PWT less than 4.0 g). Experimenter-delivered electrical stimulation (30 0.5-s stimulations per minute) of the PVN or MFB was delivered using parameters similar to those during self-stimulation sessions. Specific testing procedures were used to prevent motor abnormalities during testing. Baseline PWTs were determined before stimulation. With the frequency set to 50 Hz, the current was gradually increased (10-μA increments) up to that used during self-stimulation sessions (unique to each rat; some rats exhibited motor effects and lower intensities than those used during self-stimulation were used). PWTs were assessed to various frequencies in ascending order (50, 79, 112, 158 Hz for the PVN; 50, 78, 118, 156 Hz for the MFB). The frequency of stimulation was gradually increased (10-Hz increments until the next scheduled test frequency) after each PWT assessment. Testing lasted between 8–12 min per rat.
Morphine sulfate was purchased as a 15 mg/ml sterile solution (Baxter Healthcare; Deerfield, IL). Cocaine hydrochloride was obtained from the Drug Supply Program of the National Institute on Drug Abuse (Rockville, MD), dissolved in 0.9% (wt/vol) saline, and sterilized by filtration through a 0.22-μm nitrocellulose filter. All drugs were diluted using 0.9% (wt/vol) saline, pH 7.4.
Operant chambers housed within sound- and light-attenuating enclosures equipped with a houselight and ventilation fan were used (Med Associates Inc., St. Albans, VT). These chambers have a lever 5 cm above a grid bar floor, a stimulus lamp 2 cm above the lever, and a tone generator. An ICSS stimulator controlled by computer software (Med Associates Inc.) that controlled all stimulation parameters and data collection was located outside of the enclosure. A two-channel swivel commutator (Model SLC2C, Plastics One) positioned above the operant chamber connected the electrodes to the ICSS stimulator via 25 cm cables (Plastics One).
After at least 14 days recovery from surgery, rats were trained to lever-press for electrical stimulation of the PVN or MFB as previously described.3Illumination of the stimulus light indicated stimulus availability, and each lever press generated a 0.5-s train of rectangular alternating cathodal and anodal pulses (0.1-ms pulse durations). During stimulation the stimulus light shut off, the houselight turned on, and a tone sounded. Responding during stimulation resulted in no further stimulation and was not recorded.
During training sessions the frequency was held constant (150 Hz) and the intensity adjusted to maintain consistent responding. After initial training, frequency response curves were generated. For PVN ICSS, 120-min sessions consisted of five 18-min components. Each component consisted of eight 135-s trials. Each trial began with a 5-s timeout followed by a 10-s priming period in which rats received five noncontingent stimulations, and concluded with a 120-s response period. Current intensity remained constant (unique to each animal) and 8 frequencies (158–71 Hz, 0.05 log increments; PVN) corresponding to each trial were made available in descending order. A 30-min timeout between the third and fourth components permitted drug injections. MFB ICSS was similarly performed with the following differences: 90-min sessions consisted of 6 10-min components, which consisted of 10 60-s trials. Each trial began with a 5-s timeout, followed by a 5-s priming period, and concluded with a 50-s response period; 10 frequencies (156–45 Hz, 0.06 log increments) corresponded to each trial. At the beginning of the timeout rats received 1 ml/kg intraperitoneal injections of saline (0.9% wt/vol), morphine (0.3–6 mg/kg), or cocaine (5–10 mg/kg). For data analysis, the two components preceding drug injection and the two (PVN) or three (MFB) components following drug injection were averaged and compared using Prism software (sigmoidal-dose response, variable slope; GraphPad, La Jolla, CA). Test sessions were separated by at least 1 day. Test sessions were performed 1–4 months after initial surgery.
Rats were sacrificed by carbon dioxide asphyxiation. Brains were rapidly removed, frozen in isopentane (−35°C), and stored at −80°C. Coronal sections (25 μm) around the electrode tract were obtained using a cryostat to confirm electrode placement within the PVN and MFB (fig. 1).
The EF50 (frequency at which rats emitted 50% of maximal responding) and maximum response rate for PVN or MFB ICSS was calculated using Prism software (sigmoidal-dose response, variable slope; GraphPad). The effect of drug treatment on PVN and MFB ICSS was analyzed using a two-way analysis of variance with drug dose and experimental group (PVN control, PVN SNL, MFB control, MFB SNL) serving as the independent variables and ΔEF50 (EF50 before injection – EF50 after injection) or maximal response rates serving as the dependent measures. Post hoc analyses were made using Tukey honestly significant difference t test for multiple pairwise comparisons. A two-tailed P value of 0.05 or less was considered statistically significant. All statistical analyses were performed using JMP software (version 5.0.1a, SAS Institute Inc., Cary, NC).
Effects of SNL on ICSS of the PVN or MFB
The intensities used for ICSS were not significantly different between SNL and control groups for either the PVN [F(1,22) = 0.2, P = 0.6] or the MFB [F(1,15) = 0.05, P = 0.8]. The maximum response rates also did not differ between control and SNL rats for ICSS in the PVN [F(1,22) = 0.2, P = 0.6] or in the MFB [F(1,15) = 3.7, P = 0.07] (fig. 2). The EF50 values did not differ between SNL and control rats for ICSS in the MFB [F(1,15) = 0.2, P = 0.7]; however, the EF50 value for ICSS in the PVN was slightly but significantly greater in SNL rats (114.7 ± 1.8) compared with control subjects (109.1 ± 1.8) [F(1,22) = 4.6, P = 0.04].
Comparison of ICSS Parameters in the PVN and MFB
The intensities used for ICSS did not differ (SNL and control rats combined) between the PVN and MFB [F(1,38) = 2.1, P = 0.16]. The maximum response rate maintained by ICSS in the PVN (23.8 ± 1.1 resp/min) was approximately half that maintained in the MFB (45.1 ± 1.3 resp/min) [F(1,38) = 160, P < 0.0001] (fig. 2). The EF50 was also significantly greater for ICSS in the PVN (112.0 ± 1.6) compared to the MFB (96.0 ± 2.0) [F(1,38) = 39, P < 0.0001].
Effects of Morphine on PVN and MFB ICSS in Control and SNL Rats
There was a significant main effect of morphine dose [F(4,160) = 23.4. P < 0.00001] and group [F(3,160) = 5.6, P = 0.0012] on the effects of EF50 for ICSS in the PVN and MFB in SNL and control rats (fig. 3). Post hoc analysis found that the effect of morphine differed between the MFB SNL group and the other three groups. Morphine produced significant leftward shifts in the frequency-rate curve for PVN ICSS, decreasing the EF50 at doses of 1, 3, and 6 mg/kg compared with saline in both control and SNL rats (P ⩽ 0.05). Morphine also produced a leftward shift in the frequency-rate curve for the MFB control group, decreasing the EF50 at doses of 3 and 6 mg/kg. Morphine did not alter the EF50 at any dose in the MFB SNL group however (P ⩽ 0.05).
The effect of morphine on the maximum response rate was also dependent on the dose [F(4,160) = 8.1, P < 0.0001] and group [F(3,160) = 11.1, P < 0.0001], with 3 mg/kg decreasing the maximum response rate in both the PVN control and PVN ICSS groups (P ⩽ 0.05) and no other significant effects being found.
Effects of Cocaine on PVN and MFB ICSS in Control and SNL Rats
Cocaine produced significant effects on both PVN and MFB ICSS in control and SNL rats. The effect of cocaine on the EF50 values for ICSS was dependent upon both dose [F(2,105) = 46.2, P < 0.0001] and group [F(3,105) = 11.3, P < 0.0001], with a significant dose by group interaction [F(6,105) = 3.5, P = 0.004]. Only the highest dose tested of 10 mg/kg cocaine produced a significant effect on PVN ICSS compared with saline and did so in both the control and SNL groups (P ⩽ 0.05). However doses of 5 and 10 mg/kg significantly shifted the frequency-rate curves to the left, thereby reducing the EF50s for MFB ICSS in both the control and SNL groups compared with saline treatment (P ⩽ 0.05) (fig. 4). The effect of cocaine on the EF50s for ICSS was significantly different between the PVN and MFB groups, but not between the control and SNL groups within each brain region (P ⩽ 0.05).
Cocaine increased the maximum response rate values for ICSS at both doses of 5 and 10 mg/kg [F(2,105) = 15.4, P < 0.0001] and there was not a significant effect of ICSS group on this effect [F(3,105) = 1.6, P = 0.2].
Effects of Electrical Stimulation of the PVN and MFB on PWT in SNL Rats
Electrical stimulation of the PVN produced a frequency-dependent antiallodynic effect [F(4,39) = 24.9, P < 0.0001], with frequencies of 112 and 158 Hz increasing PWT compared with baseline values (2.7 ± 0.1g) (fig. 5). In contrast with the PVN, electrical stimulation of the MFB failed to produce a significant effect on PWT over a range of 50–156 Hz [F(4,39)-1.2, P = 0.3], with PWT not being significantly different from baseline values (2.8 ± 0.3 g) at any frequency (intensity = 189.4 ± 12.1 μA) (fig. 5).
The current study sought to determine whether the reinforcing effects of electrical stimulation of the PVN are modulated by drugs of abuse in a manner previously reported for the VTA, and if the presence of neuropathic pain altered these effects.3,4Pharmacology studies revealed that morphine and cocaine potentiated the reinforcing effects of PVN and MFB ICSS, although important differences in drug efficacy for each brain region were observed. Morphine was more efficacious in potentiating the reinforcing effects of PVN compared with MFB ICSS. In contrast, cocaine had decreased potency and efficacy in potentiating PVN compared with MFB ICSS. Studies in SNL rats revealed that nerve injury diminished morphine's ability to facilitate MFB ICSS, similar to previous results with VTA ICSS,3but did not diminish morphine's ability to facilitate PVN ICSS. Cocaine's effects were unchanged by SNL for either PVN or MFB ICSS, similar to previous findings with VTA ICSS.3These studies suggest that the reinforcing effects of PVN ICSS may be modulated to some extent through indirect stimulation of mesolimbic dopamine pathways; however, it is likely that other systems also are involved because PVN ICSS is influenced to a greater extent by opioids compared with cocaine, whereas the reverse is true with MFB or VTA ICSS.3
It is hypothesized that PVN stimulation is reinforcing to some extent through indirect stimulation of mesolimbic dopamine neurons of the VTA, a pathway implicated in the reinforcing aspects of rewarding stimuli.17PVN oxytocin fibers project directly to the VTA and terminate in close proximity to VTA dopamine cell bodies,8providing a neuroanatomic basis for oxytocin interactions with mesolimbic dopamine circuitry. Therefore, PVN stimulation is expected to induce oxytocin release in the VTA. Exogenous administration of oxytocin in the VTA induces dopamine release in the nucleus accumbens,8and thus PVN stimulation-induced oxytocin release in this region would similarly be expected to stimulate dopamine neurotransmission. Under these conditions, it would be expected that drugs that stimulate this circuitry would facilitate PVN ICSS, and the fact that morphine and cocaine facilitated PVN ICSS is in agreement for a role of mesolimbic dopamine mediating the reinforcing effects of PVN ICSS.
Opioids and dopamine agonists also potentiate the effects of MFB ICSS.2Given the close proximity of the PVN to MFB, one possibility is that unintended stimulation of the MFB is responsible for the reinforcing effects of PVN ICSS, and moreover may explain why morphine and cocaine facilitated PVN ICSS. The potency and efficacy of cocaine was diminished in rats responding for PVN compared with MFB ICSS in both control and SNL rats. If PVN ICSS were mediated by unintended stimulation of the MFB, then these results would indicate that weak stimulation of the MFB from a distal electrode diminishes the potency and efficacy of a drug to facilitate ICSS. However, such an interpretation is unlikely because, in the same rats, morphine produced greater facilitation of PVN compared with MFB ICSS. One of the most striking of the differences observed between PVN and MFB ICSS was the relative inability of cocaine to facilitate PVN ICSS compared with the MFB. For MFB ICSS the highest dose of cocaine assessed produced roughly a 2.5-fold greater facilitation compared with morphine, similar to previous findings using VTA ICSS,3and consistent with microdialysis data indicating that cocaine stimulates dopamine release in the nucleus accumbens to a greater extent than morphine.18In contrast, this same dose of cocaine was only half as efficacious as morphine in facilitating PVN ICSS in control and SNL rats. This suggests a greater role for opioid modulation of PVN ICSS compared to MFB ICSS, and further supports the notion that the neuronal mechanisms that support PVN ICSS are fundamentally different from those involved in MFB ICSS.
In the current study all manipulations on ICSS were performed in both control and SNL rats. The only effect of nerve injury on altering drug modulation of PVN and MFB ICSS was to decrease the potency of morphine for potentiation of MFB ICSS. This finding is similar to previous work using VTA ICSS,3and may relate to impairment of μ opioid receptor function within the VTA.19The working hypothesis is that opioids are less effective in stimulating the mesolimbic dopaminergic system in the presence of neuropathic pain,19and therefore are less effective in facilitating either VTA or MFB ICSS in nerve-injured rats. The fact that morphine was equally effective at facilitating PVN ICSS in SNL and control rats is yet another key difference between PVN and MFB ICSS, and indicates that PVN ICSS may not solely be mediated by indirect stimulation of mesolimbic dopaminergic systems, because one would expect morphine to be less efficacious in SNL rats compared with control subjects if this were true. It is possible that morphine facilitates PVN ICSS by stimulating oxytocin release and/or interacting with oxytocin neurotransmission throughout the limbic system, as well as through its own direct actions in stimulating dopamine neurotransmission.20Such additive effects could also explain why morphine produces greater facilitation of PVN compared with MFB ICSS, and may mask the suppressive effects of SNL on opioid activity within the VTA. Similarly, it is possible that the reduced efficacy of cocaine compared with morphine in potentiating PVN ICSS may be due to cocaine exerting direct effects on limbic dopamine neurotransmission, but having little effect on facilitating oxytocin release and/or interacting at other sites that contribute to the reinforcing effects of PVN ICSS. Increased efficacy of morphine compared with cocaine in facilitating PVN ICSS suggests that opioid interactions with the PVN may play a prominent and relatively unknown role in the reinforcing and abuse related effects of opioids, during both pain and nonpain states.
A major limitation of the ICSS methodology is the lack of selectivity of electrical stimulation, particularly when studying brain regions in close proximity to each other. To combat this, pharmacologic studies were performed in groups with electrodes placed in the PVN and MFB, to differentiate the modulatory effects of morphine and cocaine. Because morphine and cocaine facilitate PVN and MFB ICSS, albeit to different degrees, it is impossible to rule out the existence of unintended stimulation reaching the opposing brain region. It should be noted, however, that the differential effects of morphine and cocaine in facilitating PVN and MFB ICSS were obtained in both control and SNL rats; this increases confidence in these findings because in all cases critical pharmacologic differences were found in two groups of animals for each brain region.
Another limitation of the current work pertains to regional selectivity of the drugs studied. Because morphine and cocaine were given systemically, it is unclear to what degree drug effects are attributable to specific brain regions. This is particularly important when trying to understand the mechanism by which morphine facilitates PVN ICSS, because according to current thinking PVN ICSS could be modulated at the level of the PVN, VTA, or in other limbic and frontal brain regions that contain opioid receptors. Future studies addressing these questions would benefit from the use of site-specific intracranial injections of opioid and oxytocin agonists and antagonists.
It was interesting that PVN but not MFB stimulation was effective in reversing mechanical allodynia in SNL rats. This indicates not only that stimulation of limbic dopamine does not alter spinally mediated hypersensitivity, but that stimulation of the MFB is unlikely causing stimulation of the PVN; it is unclear if the reverse is true, however. It is worth noting that compared with previous findings,21,22higher frequencies of PVN stimulation were necessary to reverse hypersensitivity following nerve injury. Initial attempts to provide short, constant stimulation followed by assessment of allodynia proved difficult to do in freely moving rats. Therefore, in the current study PVN stimulation was delivered for minutes, and each 0.5-s train of stimulation was followed by a 1.5-s period of no stimulation. This was done because it allowed animals to habituate to the stimulation, reduced motor effects, allowed testing at higher frequencies, and modeled rates of self-stimulation during ICSS sessions. It is quite possible that the repeated use of on-and-off stimulation may have led to the higher frequency requirements in reversing hypersensitivity. Nonetheless, this protocol was reliable in reversing hypersensitivity during PVN stimulation, and reflects yet another difference between stimulation of the PVN and MFB in the current work.
In conclusion, electrical stimulation of the PVN produces reinforcing effects in control and SNL rats. It is hypothesized that oxytocin release into limbic regions involved in reward (e.g. , VTA) likely mediates to some degree the reinforcing effects of PVN ICSS. Morphine was more effective than cocaine in potentiating PVN ICSS, whereas the reverse is true in the MFB. The reinforcing effects of PVN ICSS are influenced by opioids to a greater extent than cocaine and these effects are unaltered by SNL, suggesting that the PVN may be a brain region intricately involved in the reinforcing effects and abuse liability of opioids under normal as well as chronic pain states.