Spinal Glucocorticoid Receptors Contribute to the Development of Morphine Tolerance in Rats

Adult male Sprague-Dawley rats (Charles River Laboratories, Wilmington, MA) weighing 300–350 g were used. Animals were housed in cages with water and food pellets available ad libitum . The animal room was artificially illuminated from 7:00 to 19:00 h. The general experimental protocol was approved through our Institutional Animal Care and Use Committee (Boston, MA).

An intrathecal catheter (PE-10) was implanted in each rat according to the previously published method.25Animals that exhibited neurologic deficits such as paralysis after the intrathecal catheter implantation were excluded from the experiments.9Drugs were delivered via  an intrathecal catheter in a total volume of 10 μl followed by a saline flush. Each drug was injected separately via  the intrathecal catheter during an injection session. The following drugs were purchased from Sigma (St. Louis, MO): mifepristone (RU38486), morphine, spironolactone, and dexamethasone. Morphine was dissolved in normal saline, and other drugs were dissolved in a 10% ethanol solution (vehicle).

For those experiments using GR oligodeoxynucleotides, the sequences for antisense, sense, and mixed-base oligodeoxynucleotides were chosen based on a previous study that has verified the down-regulation of GRs after the administration with the antisense but not sense or mixed-base oligodeoxynucleotide.26Therefore, the sequences overlapping the respective initiation codon (Gene Bank No. M14053) were used for targeting GR mRNA: GR antisense (TGG AGT CCA TTG GCA AAT), GR sense (ATT TGC CAA TGG ACT CCA), and the mixed-base (TGA AGT TCA GTG TCA ACT), as described in a previous study.26Sequencing was performed by MWG-Biotech (High Point, NC), and the product stability was ensured at the time of delivery. Oligodeoxynucleotides were dissolved in 0.9% saline and injected intrathecally (5 μmol each) twice daily. After the last behavioral testing, spinal cord tissue samples were taken to examine changes in the GR protein concentration.

Tolerance to the antinociceptive effect of morphine was induced using an intrathecal treatment regimen of 10 μg morphine given twice daily for 6 consecutive days. Differences in morphine antinociception among treatment groups were assessed using the tail-flick test by testing at 30 min after a probe dose of 10 μg morphine. The dose–response effect of morphine was examined by generating cumulative dose–response curves to reduce the total number of rats used, as described in a previous study.27The increment log doses (0.2 log unit) of morphine were given to the same rats, and the morphine antinociceptive effects were examined at 20 min after each morphine injection. This process was repeated until either no additional antinociceptive effect was demonstrated or the cutoff time (10 s) of the tail-flick test was reached in response to a next higher dose, indicative of the maximal antinociceptive effect of morphine under the experimental condition.27The routine tail-flick test was used with baseline latencies of 4–5 s and a cutoff time of 10 s. At least two trials were made for each rat, with an intertrial interval of 1 min and with changes of the tail position receiving radiant heat stimulation at each trial.

The percent of maximal possible antinociceptive effect (%MPAE) was determined by comparing the tail-flick latency before (baseline) and after a drug injection using the equation: %MPAE =[(After Drug Injection − Baseline)/(10 − Baseline)]× 100% (the constant 10 refers to the cutoff time). The data were analyzed by using two-way analysis of variance with post hoc  Newman–Keuls tests. For the dose–response data analysis, AD⁵⁰values and 95% confidence intervals were generated using a computerized regression model.

For Western blot, rats were rapidly (< 1 min) killed through decapitation after being anesthetized with pentobarbital (100 mg/kg intraperitoneal). Lumbar spinal cord segments (L3–L5) were removed through a lumbar laminectomy. The spinal segments were divided into dorsal and ventral horns and homogenized in sodium dodecyl sulfate sample buffer containing a cocktail of proteinase inhibitors (Sigma: 2 mm AEBSF, 1 mm EDTA, 130 μm bestatin, 14 μm E-64, 1 μm leupeptin, 0.3 μm aprotinin). The lumbar segments were harvested because the intrathecal drug delivery was aimed at this site. Protein samples were separated on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (4–15% gradient gel; Bio-Rad, Hercules, CA) and transferred to polyvinyl difluoride membranes (Millipore, Bedford, MA). The membranes were blocked with 5% milk and incubated overnight at 4°C with a primary antibody (GR: 1:1,000; rabbit polyclonal from Santa Cruz, Santa Cruz, CA) and 1 h at room temperature with horseradish peroxidase–conjugated secondary antibody (1:7,000; Amersham Biosciences, Arlington Heights, IL). The blots were visualized in existing chemiluminescent solution (NEN, Boston, MA) for 1 min and exposed to hyperfilms (Amersham) for 1–10 min. The blots were then incubated in a stripping buffer (67.5 mm Tris, pH 6.8, 2% sodium dodecyl sulfate, and 0.7% β-mercaptoethanol) for 30 min at 50°C and reprobed with a polyclonal rabbit anti–β-actin antibody (1:20,000; Alpha Diagnostic International, San Antonio, TX) as a loading control. The Western analysis was made in triplicates. The band density was measured with a computer-assisted imaging analysis system and normalized against the corresponding loading control. Differences were compared using analysis of variance followed by post hoc  Newman–Keuls tests.

The role of GRs in the development of morphine tolerance was examined in three groups of rats each receiving the GR antagonist RU38486 (0.5 or 1 μg intrathecal) or a vehicle with morphine (10 μg intrathecal) for 6 days (n = 5 or 6). The RU38486 doses were chosen based on the literature28and our pilot experiments. Antinociceptive tolerance developed in the morphine plus vehicle group, whereas RU38486 significantly attenuated the development of morphine tolerance when examined on days 5 and 7 (P < 0.05; fig. 1A). Consistently, the rightward shift of the morphine dose–response curve on day 7 in the morphine-plus-vehicle group was prevented in the morphine-plus-RU38486 groups (fig. 1Band table 1). In contrast, intrathecal coadministration of 10 μg morphine with the mineralocorticoid receptor antagonist spironolactone (3 μg) for 6 days did not prevent the development of morphine tolerance (P > 0.05, n = 5; figs. 1A and B), nor did spironolactone alone change baseline nociceptive threshold, indicating a selective role of GRs in this process. The selected spironolactone dose has been shown to be effective for blocking the mineralocorticoid receptor.22 

Two additional groups of rats (n = 5) each received RU38486 (1 μg intrathecal) or saline alone for 6 days. Such treatment did not change the baseline tail-flick latency, nor did the RU38486 treatment change the motor activity, indicating a specific effect of RU38486. The effect of RU38486 on morphine tolerance was unlikely due to a direct effect on morphine antinociception, because a single dose of RU38486 (1 μg) did not reverse morphine tolerance on day 7 in those rats made tolerant to morphine (P > 0.05, n = 5; fig. 1C), nor did it change the antinociception induced by a single morphine (10 μg intrathecal) treatment in a separate group of rats (%MPAE, morphine alone: 91 ± 2.3%; morphine plus RU38486: 89 ± 3.1%). Therefore, the data indicate that the GR antagonist RU38486 effectively attenuated the development of morphine tolerance.

In addition to using RU38486, we examined the functional role of GRs in morphine tolerance by interfering with the GR expression within the spinal cord dorsal horn. Although there are reports on mice with genetically altered GRs, they are either lethal or unavailable.29–31As an alternative, an antisense oligodeoxynucleotide against GRs was given as described previously.26In this experiment, a total of four groups of rats were used, and each group (n = 5–7) received 10 μg morphine in combination with 5 μmol GR antisense oligodeoxynucleotide, sense oligodeoxynucleotide, mixed-base oligodeoxynucleotide, or vehicle. The oligodeoxynucleotide treatment (twice daily, intrathecal) was first given alone for 7 days before the morphine regimen. These rats then received the intrathecal morphine treatment along with an oligodeoxynucleotide for 6 days. The spinal cord dorsal horn samples were collected from these groups at the end of behavioral testing.

The intrathecal morphine regimen induced a substantial up-regulation of GRs within the spinal cord dorsal horn when examined on day 7, whereas the same morphine treatment did not up-regulate GRs in the morphine–plus–antisense oligodeoxynucleotide group, as compared with each of the groups receiving morphine plus vehicle, a sense oligodeoxynucleotide, or mixed-based oligodeoxynucleotide (P < 0.05; fig. 2A). No motor changes such as paralysis were observed after oligodeoxynucleotide treatment. Of note is that the RU38486 treatment that prevented the development of morphine tolerance (see below) also reduced the GR up-regulation induced by morphine administration (P < 0.05, n = 5; fig. 2B), whereas RU38486 alone did not change the GR expression as compared with the vehicle group (P > 0.05; fig. 2B). Consistent with the Western blot results, the development of morphine tolerance was substantially diminished in the morphine–plus–GR antisense group as compared to each of the remaining groups (P < 0.05, n = 6; fig. 3). These results indicate that morphine tolerance was significantly attenuated when the morphine-induced GR up-regulation within the spinal cord dorsal horn was diminished by the antisense oligodeoxynucleotide treatment.

Coadministration of morphine (10 μg intrathecal) and the GR agonist dexamethasone (4 μg intrathecal),23,32given twice daily for 4 days, significantly shortened the onset of morphine tolerance such that antinociceptive tolerance to 10 μg morphine developed on day 5 of the morphine cycle (%MPAE: 46.5 ± 7.1%) as compared with the morphine alone group (%MPAE: 79.2 ± 2.3%) (P < 0.05, n = 6; fig. 4). Dexamethasone alone did not affect the baseline nociceptive threshold.

The current findings demonstrate that (1) the development of morphine antinociceptive tolerance was substantially attenuated by the GR antagonist RU38486, (2) a single treatment with RU38486 did not affect morphine antinociception, (3) a GR antisense oligodeoxynucleotide but not a sense or mixed-base oligodeoxynucleotide also attenuated the development of morphine tolerance as well as the up-regulation of GRs within the spinal cord dorsal horn, and (4) the GR agonist dexamethasone significantly shortened the onset of morphine tolerance. Collectively, these data indicate that central GRs may play a significant role in the development of morphine tolerance in rats.

Several technical considerations are worthwhile mentioning. First, although RU38486 (mifepristone, RU486) has been extensively used as a conventional GR antagonist and is the only commercially available GR antagonist, it does have the antiprogesterone effect.28The antiprogesterone effect is unlikely to be involved in this study because we only included male rats and the GR knockdown experiment using an antisense oligodeoxynucleotide produced a result comparable to that from using RU38486. Second, mineralocorticoid receptors are unlikely to play a significant role in this process because the mineralocorticoid receptor antagonist spironolactone within the dose range effective for blocking the mineralocorticoid receptor22did not influence the development of morphine tolerance. Third, although the current data indicate a spinal locus of GR actions because morphine and RU38486 or the GR antisense oligodeoxynucleotide were delivered using an intrathecal treatment regimen, it does not exclude the possibility that a similar GR-mediated mechanism of morphine tolerance may be involved in other central nervous system regions.

Although GRs have been shown to act through nongenomic mechanisms,33,34the role of GRs in morphine tolerance probably is not due to a direct interaction between GRs and μ-opioid receptors, because a single injection with the GR antagonist RU38486 did not reverse the established morphine tolerance and the same RU38486 treatment did not affect the morphine antinociceptive effect in nontolerant rats. At the cellular level, GRs are activated through the formation of a GR homodimer after the dissociation from a cytosolic complex consisting of heat shock proteins.35A GR homodimer binds to specific nuclear DNA–responsive elements to activate gene transcription and translation for a variety of cellular elements.35Given that the development of morphine tolerance was attenuated in this model after either repeated administration of RU38496 or prevention of GR up-regulation with an antisense oligodeoxynucleotide, it is plausible that GRs may have a modulatory role in one or more intracellular elements contributory to the cellular mechanisms of morphine tolerance.1–5,36 

It may be considered that an interaction between GRs and NMDA receptors might play a role in this process because NMDA receptors have been known to contribute to the cellular mechanism of neural plasticity related to learning and memory as well as opioid tolerance.6–10,37–39For example, GRs have a direct modulatory effect on the NMDA receptor function40and potentiate NMDA-induced responses in dopamine-sensitive neurons in the ventral tegmental area.41In addition, a recent study suggested the involvement of proinflammatory cytokines in opioid tolerance,42raising the possibility that the glucocorticoid effects may be mediated in part though the actions of spinal cytokines. Future studies would examine these possibilities.

The relation among stress, corticosteroids, and morphine analgesia has been suggested in several previous studies.43,44For example, Takahashi et al.  43have demonstrated that the antinociceptive effect of morphine was potentiated by adrenalectomy in mice, suggesting that endogenous corticosteroids might have a negative effect on morphine antinociception. However, in the same study, adrenalectomy seemed to be ineffective in affecting the development of morphine tolerance in mice, although that study did not specifically examine the effect of GR activation. The current data support a potential role of endogenous glucocorticoids in the mechanisms of opioid tolerance via  activation of neuronal GRs, suggesting that those factors (e.g. , stress, emotional disturbance, Addison or Cushing disease) capable of changing the endogenous corticosteroid level could be linked to the cellular mechanism of opioid tolerance through neuronal GRs. Our results also suggest that a GR antagonist such as RU38486 could be a useful pharmacologic tool, alone or in combination with other agents (e.g. , an NMDA receptor antagonist), to prevent the development of opioid tolerance, an issue of clinical relevance.45,46