Interaction between the Spinal Melanocortin and Opioid Systems in a Rat Model of Neuropathic Pain

Thirty male Wistar rats weighing 250–300 g at the start of the study were used. Animals were housed in groups of two or three in plastic cages on sawdust bedding. They were kept at a 12/12-h light/dark cycle, with food and water available ad libitum . All testing procedures in this study were performed according to the Ethical Guidelines of the International Association for the Study of Pain 24and approved of by the Ethics Committee on Animal Experiments of Utrecht University (Utrecht, The Netherlands).

Animals were anesthetized with a single intramuscular injection of Hypnorm (Janssen Pharmaceutical Ltd., Grove, Oxford) containing 0.315 mg/ml fentanyl citrate and 10 mg/ml fluanisone, at a dose of 1 ml/kg body weight.

Four ligatures were placed around the right sciatic nerve as previously described by Bennett and Xie, 8and the incision was closed with silk sutures. Two weeks after the CCI lesion, rats were anesthetized with a mixture of oxygen-nitrous oxide (1:2) and 2.5% halothane. A lumbar spinal catheter was placed as described by Storkson et al.  25and subcutaneously tunneled to the neck region. After recovery from anesthesia and also at the end of all experiments, proper placement of the catheters was checked by injecting 15 μl of lidocaine, 2%, which gives an immediate and short-lasting motor paralysis of the hind limbs on intraspinal injection. Because of incorrect placement of the spinal catheter, 3 rats were excluded from this study, thus leaving a total of 27 animals. They were allowed to recover before testing was initiated.

For in vivo  administration, MTII (melanotan II or cyclo-[Nle⁴, Asp⁵, D-Phe⁷, Lys¹⁰]α-MSH-(4-10)), SHU9119 (cyclo-[Nle⁴, Asp⁵, D-Nal(2)⁷, Lys¹⁰]α-MSH-(4-10)), naloxone (naloxone-hydrochloride), and morphine (morphine-hydrochloride) were used. MTII was purchased from Bachem Feinchemicalien (Buberdorf, Switzerland); SHU9119 was synthesized using Fmoc solid-phase synthesis as reported elsewhere. 26Naloxone and morphine were obtained from the Utrecht University Medical Center Pharmacy (Utrecht, The Netherlands). Drugs were dissolved in 5 μl saline and injected through the spinal catheter by means of a 25-μl Hamilton syringe, followed by a saline flush (12 μl).

The following 18 different treatments were given: saline, naloxone (0.1, 10, 30, and 100 μg), SHU9119 (0.5, 1, and 1.5 μg), morphine (1, 3, 10, and 30 μg), MTII (0.5 μg), or the following combinations: SHU9119 (1.5 μg) with naloxone (0.1 μg) pretreatment, morphine (1, 3, or 10 μg) with SHU9119 (0.5 μg) pretreatment, or MTII (1.5 μg) and morphine (100 μg) simultaneously. In experiments in which a “pretreatment” was given, the first drug was given 15 min before the second drug, because we previously demonstrated that the peak effect of SHU9119 is at 15 min after injection. 9 

On a typical testing day, all 27 animals were randomly divided into three groups (n = 8 or 9 each, except for simultaneous administration of MTII and morphine, where n = 4). Each group randomly received one treatment, with the experimenter blinded to the allocation. Thereafter, animals were given at least 2 days of rest to minimize any possibility of drug interactions or development of tolerance. 9On the next testing day, all animals were again randomly divided into three groups, and three other treatments were given. This design was continued until all treatments had been given to one group of animals. Thus, in total, 18 groups were tested, and therefore, each animal was used multiple times on consecutive testing days.

Foot withdrawal threshold in response to a mechanical stimulus was determined using a series of von Frey filaments (Stoelting, Wood Dale, IL), ranging from 1.08 to 21.09 g. Animals were placed in a plastic cage with a metal mesh floor, allowing them to move freely. They were allowed to acclimatize to this environment before the experiment. The filaments were presented to the midplantar surface as described by Chaplan et al. , 27and the smallest filament eliciting a foot withdrawal response was considered the threshold stimulus.

Baseline values were determined, and measurements were repeated 60 min after drug or vehicle administration (in case of a pretreatment, measurements were taken 60 min after injection of the second drug).

All data are expressed as mean ± SEM for visualization purposes only. Effects of treatments resulting in an increase or decrease in mechanical allodynia are plotted on a negative or positive axis, respectively. For the mechanical stimulation test, the effect of different drug treatments was quantified as the percentage of maximum possible effect (%MPE), using the following formula:%MPE = 100 × (postdrug value − baseline value)/(cutoff value − baseline value).

Nonparametric tests were performed to analyze the data: Overall group differences were analyzed by a Kruskal-Wallis test followed by a two-tailed Mann-Whitney U test to analyze differences between groups. A probability level of less than 5% was considered significant. To assess the statistical significance of the combination of SHU9119 and morphine, dose-response curves were generated for each compound. The composite line of additivity for the combination was determined and compared to the experimental data using analysis of variance. These results are also represented graphically by isobolographic analysis.

Two weeks after the CCI lesion, mean baseline withdrawal threshold to von Frey stimulation was reduced from 20.3 ± 0.4 g to 5.3 ± 0.6 g, confirming the presence of mechanical allodynia. Baseline withdrawal thresholds were not altered by placement of the spinal catheters and remained stable throughout the testing period.

With increasing doses of naloxone (10, 30, and 100 μg), a significant decrease in withdrawal thresholds was observed, with a %MPE of −67.2 ± 9.3 at the highest dose tested (fig. 1).

SHU9119, 1.5 μg, increased withdrawal thresholds, with a %MPE of 60.0 ± 13.3. When 0.1 μg naloxone, a dose that by itself had no effect on mechanical withdrawal thresholds, was administered 15 min before SHU9119 (1.5 μg), this antiallodynic effect of SHU9119 was largely decreased (%MPE was reduced to 14.8 ± 10.9, which was not significantly different from saline;fig. 2).

Administration of 0.5 μg MTII significantly decreased withdrawal thresholds, with a %MPE of −93.8 ± 6.2, whereas 30 μg morphine increased thresholds, with a %MPE of 89.6 ± 10.4. A combination of approximately three times higher doses of both drugs (1.5 μg MTII and 100 μg morphine) resulted in an intermediate %MPE of 12.2 ± 9.8, which was not significantly different from saline (fig. 3).

Spinal administration of morphine (1, 3, and 10 μg) produced a significant antiallodynic effect, as shown by a dose-dependent increase in withdrawal thresholds to von Frey stimulation (fig. 4), compared to saline treatment. A single dose of SHU9119 (0.5 μg) also increased withdrawal thresholds, with a %MPE of 41.4 ± 8.6. When given 15 min before morphine, 0.5 μg SHU9119 resulted in a significant increase in withdrawal thresholds, with a maximum %MPE of 86.2 ± 6.9 (0.5 μg SHU9119 and 10 μg morphine;fig. 4).

Dose-response curves for both SHU9119 and morphine, and for their combination, were generated. From these curves, isobolograms for different levels of effect were plotted (fig. 5). Although these isobolograms might suggest synergy, the comparison of the experimental data and the calculated composite additive line, as well as statistical analysis of these data (as described in Materials and Methods), reveal additivity (fig. 5, inset ).

In the current study, we were able to demonstrate an interaction between the spinal melanocortin and spinal opioid systems in a rat model of neuropathic pain. In the spinal cord, the MC⁴receptor, 28,29as well as α-MSH, an endogenous ligand for the melanocortin receptors, and its precursor molecule pro-opiomelanocortin 12,29,30are shown to be expressed in the dorsal horn, an area involved in the processing of nociceptive information. Proteolytic cleavage from pro-opiomelanocortin yields not only α-MSH but also the endogenous opioid β-endorphin, which is also present in the dorsal horn. 12Moreover, both the μ- and δ-opioid receptors, for which β-endorphin displays a high affinity, 31are expressed in the same area. 32Thus, both a functional melanocortin and an opioid system appear to be present at the same anatomic site in the spinal cord. Possibly, there is a natural balance between these two systems, each with opposite activities on nociceptive processing (see also Adan and Gispen 33). In neuropathic pain, this balance might be out of equilibrium, by increased activity of the melanocortin system 34and decreased activity of the opioid system. 19–22 

Considering this balance between the spinal melanocortin and opioid systems, we proposed that by blocking the pronociceptive activity of the spinal melanocortin system with SHU9119, 34tonic analgesic effects of β-endorphin, through activation of opioid receptors in the dorsal horn, will predominate (see also Vrinten et al.  35). As shown in figure 1, administration of naloxone increased mechanical allodynia, consistent with blockade of a tonically active endogenous opioid agonist at the spinal level. This confirms the results of previous work in which high doses of naloxone increased pain-related behavior in mononeuropathic and spinally injured rats. 36,37When a low dose of naloxone, which by itself had no effect on allodynia, was administered 15 min before SHU9119, the antiallodynic effect of SHU9119 was almost completely blocked, as can be seen in figure 2.

These data suggest that the melanocortin and opioid systems act at the same common pathway and that the level of activity of one system (i.e. , the opioid system) might modulate the threshold for activation of the other system (i.e. , the melanocortin system), even if the naloxone dose is so low that it does not affect functional responses by itself. Other evidence for the close interactions between the melanocortin and opioid systems has been provided previously. For instance, chronic morphine treatment is shown to induce a down-regulation of MC⁴receptors in several brain areas, 38whereas ablation of pituitary pro-opiomelanocortin neurons induces both hypothalamic pro-opiomelanocortin overexpression and a down-regulation of μ-opioid receptors in several brain areas. 39Our current experiments suggest that an interaction is also present at the spinal level. However, from these data, it is not possible to establish the neuroanatomic origin of this interaction. The melanocortin and opioid systems might be organized in a linear pathway, with the melanocortin system either upstream or downstream of the opioid system. Another possibility is that the two systems are organized in a parallel fashion. The MC⁴receptor and opioid receptor could be located on different neurons, with pathways converging further downstream, or they could be colocalized on the same neuron. The latter has been demonstrated in locus ceruleus-derived cells, in which α-MSH and β-endorphin respectively increase and decrease the level of the second messenger cAMP (via  adenylate cyclase) through melanocortin receptors and δ-opioid receptors that are expressed in the same cell. 40Adenylate cyclase might thus function as an integrator of melanocortin- and opioid-mediated signaling, thereby regulating the output of the cell in vivo . A similar mechanism has been proposed to be involved in the effect of melanocortins on opioid addiction. 16 

To further investigate the nature of the neuroanatomic substrate underlying the interaction between the spinal melanocortin and opioid system, we tested the effect of coadministration of a high dose of the melanocortin receptor agonist MTII and morphine. As seen in figure 3, each of these drugs alone already had an almost maximal effect on allodynia with a three-times-lower dose than we used for the coadministration. When the two systems are organized in a linear fashion, it would be expected that the resultant effect of the combined doses on allodynia would resemble that of activation of the downstream receptor. Thus, when the opioid receptor is downstream of the melanocortin receptor, this combination of drugs would produce antiallodynia, and vice versa , whereas with parallel pathways, the resulting effect is expected to be in-between. The net effect of the combined drugs was indeed between those of the separate drugs (fig. 3), suggesting a parallel organization of the melanocortin and opioid system. Because a dose of naloxone that was in itself inactive blocked the antiallodynic effect of the melanocortin receptor antagonist SHU9119, we conclude that the melanocortin and opioid systems converge on the same pathway, possibly the same neurons, involved in nociceptive processing. In the experiments described here, we focussed solely on functional parameters. An alternative approach to investigate the interaction between the spinal opioid and melanocortin systems, to further test our hypothesis and address the question on which neurons the two systems converge, is to conduct neuroanatomic or electrophysiologic studies.

From a clinical point of view, an integration of the melanocortin and opioid system is interesting because by coadministration of melanocortin receptor antagonists, it might be possible to modulate opioid effectiveness. We tested this hypothesis by administering SHU9119 15 min before morphine. With this treatment paradigm, SHU9119 and morphine had an additive antiallodynic effect, as shown in figure 4. Although this does not imply a modulation of opioid effect by melanocortins, it raises the possibility that combined treatment with MC⁴receptor antagonists and opioids might have a place in the management of neuropathic pain. With concomitant administration of MC⁴receptor antagonists, the total dose of opioid needed to obtain adequate analgesia might be reduced. This might diminish the incidence and severity of opioid side effects, development of tolerance, and possible tolerance-associated hyperalgesia. 41,42Also, by changing the interinjection interval of SHU9119 and opioids, it could be that the combined antiallodynic action is enhanced, as has been demonstrated for the combination of N -methyl-d-aspartate receptor antagonists and morphine. 43 

In conclusion, we were able to demonstrate an interaction between the spinal melanocortin and opioid systems in a rat model of neuropathic pain. The antiallodynic actions of the MC⁴receptor antagonist SHU9119 were largely blocked by a low dose of naloxone, whereas SHU9119 and morphine had an additive antiallodynic effect. It is conceivable that coadministration of MC⁴receptor antagonists and opioids might contribute to a better treatment of human neuropathic pain.