Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

It is well known that opioids have many diverse effects on the immune system. Interestingly, many of these effects are indirect and appear to be mediated via the central nervous system, suggesting that alterations in opioid neurotransmission can alter the efficacy of the immune response. Our laboratory and others have demonstrated that central morphine administration decreases mitogen-induced blood lymphocyte proliferation (Hernandez et al., 1993; Mellon and Bayer, 1998) or splenocyte proliferation (Lysle et al., 1996), decreases NK cell activity (Weber and Pert, 1989; Moeniralam et al., 1998), as well as increases macrophage nitric oxide release (Fecho et al., 1994). Studies have also indirectly demonstrated a modulatory role of central opioid receptors in the stimulatory effect of interleukin-1 on serum interleukin-6 (IL-6), which was prevented by naloxone pretreatment (De Simoni et al., 1993). Direct evidence of opioids modulating IL-6 release was obtained by both peripheral and central routes of administration (Bertolucci et al., 1996; Zubelewicz et al., 1999).

Interleukin-6 is a 26-kDa protein that is produced by both immune and nonimmune tissues and thus has many different biological effects. For example, circulating IL-6 leads to the release of acute phase proteins from the liver (for review, see Heinrich et al., 1990), elicits a febrile response, and activates the hypothalamic-pituitary-adrenal axis (Lenczowski et al., 1999). As such, administration of opioids via illicit or therapeutic use could lead to a dysregulation of circulating IL-6 levels. Likewise, during times of stress and adaptation the release of endogenous opioids may alter the homeostasis of an organism to respond to bodily injury or inflammation. The observations of Bertolucci et al. (1996) suggest that opioids may regulate peripheral levels of this cytokine. However, the characteristics and mechanism(s) by which opioids induce this elevation in IL-6 still remain largely uncharacterized.

To gain a more clear understanding of opioid-induced immunoregulation as measured by the elevation of plasma IL-6 levels, studies were designed to characterize the dose, time, and receptor dependence of morphine administration to rats. In a previous study, only a single time point and dose of morphine were used to determine the effect of opioids on sera IL-6 (Bertolucci et al., 1996). Furthermore, because the mechanism of this effect has yet to be determined, studies were completed to elucidate whether the autonomic nervous system (ANS) or activation of the adrenal glands was involved. Previous studies from our laboratory have demonstrated a role of the ANS in the effect of morphine to decrease mitogen-stimulated lymphocyte proliferation (Flores et al., 1996; Mellon and Bayer, 1999; Mellon and Bayer, 2001). However, it remains to be determined whether activation of the autonomic nervous system is required for the effects of morphine on plasma IL-6 levels. To test this possibility, the bisquaternary nicotinic ganglionic receptor antagonist chlorisondamine was used to effectively inhibit peripheral autonomic neurotransmission (Clarke et al., 1994).

Finally, it has been previously determined that the adrenal gland may be a potential source of IL-6. Interleukin-6 mRNA has been observed in the adrenal gland by in situ hybridization (Gonzalez-Hernandez et al., 1994; Gadient et al., 1995). Additionally, the adrenal cortex and medulla have been implicated in the release of IL-6, because primary cultures of adrenal cells produce IL-6 when stimulated with several neurotransmitters, hormones, and cytokines (for review, see Nussdorfer and Mazzocchi, 1998). To determine the potential involvement of the adrenal gland in this response to morphine, adrenalectomized and adrenal demedullated animals were used. The results suggest a unique mechanism for central opioid regulation of a peripheral cytokine, which is distinct from the mechanisms reported for the suppressive effects of opioids on lymphocyte proliferation and NK cell activity.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. Male Sprague-Dawley rats were purchased from Taconic Farms (Germantown, NY), weighing 200 to 225 g upon arrival. Animals were group housed (three per polypropylene cage with microisolator tops) and were provided free access to food (Purina Rat Chow; Purina, St. Louis, MO) and water ad libitum. The animals were housed in a thermoregulated (23 ± 1°C) and light (12-h light/dark)-controlled room for 1 week to allow for acclimation to this environment before experimentation. Animals that underwent surgery were allowed to recover for an additional week before experimentation. All animal studies have been approved by the Georgetown University Animal Care and Use Committee in accordance with the guidelines adopted by National Institutes of Health.

Drugs. Morphine sulfate was generously provided by the National Institute on Drug Abuse (Research Triangle Park, NC). Chlorisondamine chloride (Ecolid) was a gift from Novartis (Summit, NJ). Recombinant mouse IL-6 and rat IL-6 (rrIL-6) were purchased from Genzyme (Cambridge, MA) and Peprotech (Rocky Hill, NJ), respectively. Naltrexone hydrochloride and all other drugs were purchased from Sigma Chemical (St. Louis, MO). All vehicle and drug solutions were made using sterile, pyrogen-free 0.9% saline.

Surgical Procedures. For all surgeries, a proper and sterile surgical environment was maintained. Specifically, all surgical instruments were autoclaved before and soaked in Novalsan solution during surgical procedures. Incision sites for either central cannula placement or bilateral adrenalectomy were prepared by shaving the skin and cleaning with Novalsan solution and ethyl alcohol (70%). The incision was made using either a sterile surgical blade or sterile surgical scissors. After the completion of the surgical procedure, the wound site was either sutured or stapled as described below. To minimize possible infection after the surgical procedure, each animal was injected intramuscularly with 0.1 ml of gentamicin sulfate (40 mg/ml).

Intracerebroventricular Cannulation and Microinjection Procedures. Animals were anesthetized with equithesin (3 ml/kg i.p.). Equithesin was prepared by dissolving chloral hydrate (4.25 g) and magnesium sulfate (2.13 g) in a solution of propylene glycol (26.6 ml), pentobarbitol (18 ml), and deionized water (43 ml). Animals were placed in a stereotaxic apparatus (David Kopf Instruments, Tujunga, CA), and a single 5.0-mm guide cannula was directed into the lateral ventricle using the following coordinates relative to bregma: 0.9 mm anterior-posterior, 1.6 mm mediolateral, and 3.1 mm dorsoventral obtained from Paxinos and Watson (1997). The guide cannula was fixed with dental acrylic and jewelers' screws and fitted with an inert 28-gauge cannula to maintain patency. Animals were allowed to recover for at least 1 week before experimentation.

Adrenalectomy and Adrenal Demedullation. Animals were anesthetized with equithesin. Bilateral incisions were made 1 cm below the rib cage. Underlying muscle was cut with sterile surgical scissors and the adrenal gland was isolated and removed. The musculature was sutured with polyglycolic acid sutures 4-0 (American Cyanamid Co., Wayne, NJ) and the skin wound was stapled. Postoperative care for animals included a bolus subcutaneous injection of sucrose (20%). Adrenalectomized animals were provided saline (0.9%) drinking solution supplemented with corticosterone 21-acetate (25 µg/ml in 0.2% ethanol) to restore basal corticosterone levels (Jacobson et al., 1989; Zhou et al., 1993). Sham-operated animals were provided with drinking water containing 0.2% ethanol. All animals were allowed at least 1-week recovery from surgery before experimentation. To verify the completeness of the adrenalectomy, plasma corticosterone levels were measured. Any plasma corticosterone level that was 75 ng/ml was considered incomplete and that animal was eliminated from data analysis.

Plasma Corticosterone Measurement. Plasmas collected at the time of sacrifice were stored at 20°C and assayed for corticosterone levels using a solid phase ¹²⁵I radioimmunoassay purchased from ICN Pharmaceuticals (Costa Mesa, CA).

Plasma IL-6 Measurement. Interleukin-6 activity was determined using the murine splenocyte hybridoma cell line 7TD1 (American Type Culture Collection, Manassas, VA). The cells were maintained in culture medium (RPMI-1640 with glutamine) supplemented with 10% heat-inactivated fetal calf serum, 50 µM -mercaptoethanol, and 10 units/ml recombinant mouse interleukin-6. IL-6 activity was measured as previously described (Van Snick et al., 1986; LeMay et al., 1990) except that rrIL-6 was used instead of recombinant human IL-6 in the standard curve. Briefly, 7TD1 cells were washed three times in the culture medium described above without IL-6 and centrifuged at 1100 rpm (DuPont Sorvall, Newtown, CT). Experiments were completed using 96-well plates containing 100 µl of control media, serially diluted rat plasma (2- to 4096-fold) or rrIL-6 (0.1-1000 U/ml), and 100-µl aliquots of 2 × 10³ cells/well were cultured for 72 h. During the last 6 h, the cultures were pulsed with 0.5 µCi of [³H]thymidine (PerkinElmer Life Sciences, Boston, MA), harvested onto glass fiber filters by using a 96-well harvester (Brandel Inc., Gaithersburg, MD), and radioactivity was counted on a Betaplate scintillation counter (Wallac, Gaithersburg, MD).

Statistical and Data Analysis. Plasma IL-6 activity and corticosterone data were analyzed using a one-way ANOVA with Newman-Keuls for post hoc comparisons. Values of plasma IL-6 that were more than 2 S.D. from the mean of the treatment group were considered outliers at 95% confidence intervals and omitted from data analysis. For the experiments presented in these studies, plasma IL-6 values for 12 of 245 (5%) animals met this criterion and were omitted. Adrenalectomized animals received exogenous corticosterone therapy and values of plasma corticosterone levels <75 ng/ml were considered complete. Using this criterion, 1 of 24 (4%) adrenalectomized animals was excluded due to elevated corticosterone levels (147 ng/ml). In adrenalectomized animals receiving corticosterone replacement therapy, the range of plasma corticosterone values for saline- and morphine-treated animals was 25 to 69 and 25 to 41 ng/ml, respectively. In adrenal demedullated animals, plasma corticosterone levels were monitored to determine that the remaining cortex was intact. Statistical significance was accepted at a level of p  0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Time Course of Morphine-Induced Plasma IL-6 Activity. To determine the time course of morphine-induced elevation of plasma IL-6, animals received subcutaneous treatments of either saline (1 ml/kg) or morphine (10 mg/kg) and were sacrificed 30, 60, or 120 min after drug treatment. As shown in Fig. 1, saline-injected rats were sacrificed 30 min after treatment and basal IL-6 activity was observed. After 30 min of morphine treatment, plasma IL-6 activity increased 2-fold compared with the saline-injected controls. In additional studies, saline treatment for either 1 or 2 h had similar plasma IL-6 levels (data not shown). Likewise, 60- and 120-min treatment with morphine significantly elevated plasma IL-6 activity, as well. For all subsequent studies, plasma IL-6 activity was observed 1 h after morphine treatment.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Time course of morphine-induced elevation of plasma IL-6 activity. Animals were subcutaneously injected with either saline (1 ml/kg) or morphine (10 mg/kg) and sacrificed 30, 60, or 120 min after drug treatment. Plasma IL-6 activity was measured using the 7TD1 bioassay (see Materials and Methods). Data are combined from two independent experiments and expressed as the mean ± S.E.M. IL-6 (units/milliliter) for groups of 8 to 10 animals. , p < 0.01 or , p < 0.001 compared with saline-injected controls (one-way ANOVA, Newman-Keuls).

Dose-Dependent Effect of Morphine on Plasma IL-6 Activity. To determine whether the elevation of plasma IL-6 levels was dose-dependent, increasing doses of morphine were subcutaneously injected into groups of rats. Animals received either saline (1 ml/kg) or morphine (0.1-30 mg/kg), and plasma IL-6 activity was determined 1 h after drug treatment. Lower doses (0.1 and 1 mg/kg) were not significantly different from saline-injected controls (Fig. 2). Doses of morphine greater than 10 mg/kg led to significant elevations (5- to 14-fold) in IL-6 activity. These effects were dose-dependent with approximately 500, 1100, and 1400% increases in plasma IL-6 with 10, 20, and 30 mg/kg doses, respectively.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Dose-dependent effect of morphine on plasma IL-6 activity. Animals were subcutaneously injected with either saline (1 ml/kg) or morphine (0.1-30 mg/kg). One hour after drug treatment, animals were sacrificed and plasma IL-6 was determined. Data are expressed as the percentage of saline-injected controls (mean ± S.E.M.). The data were obtained from two separate experiments with overlapping doses. In the first experiment, saline and morphine at doses 0.1, 1, 10, and 20 mg/kg were examined. In the second experiment, IL-6 responses to saline and morphine at doses 10, 20, and 30 mg/kg were measured. The data represent groups of six rats for each of the low doses of morphine (0.1 and 1 mg/kg) and groups of 10 to 12 rats for each of the higher doses of morphine (10, 20, and 30 mg/kg). , p < 0.05 compared with saline-treated animals and , p < 0.001 compared with both saline and morphine (0.1, 1, and 10 mg/kg)-treated animals (one-way ANOVA, Newman-Keuls). Nonlinear regression analysis of this morphine dose response on plasma IL-6 activity was presented in the inset. An ED50 value for the morphine-induced IL-6 activity was approximated at 13.8 mg/kg.

Effect of Central Administration of Morphine on Plasma IL-6 Activity. To determine whether the activation of central opioid receptors also increased IL-6 levels, morphine was administered directly into the lateral ventricle. As shown in Fig. 3, intracerebroventricular microinjection of morphine (10 µg/5 µl) led to a significant (3-fold) increase in plasma IL-6 activity compared with the saline-injected (5 µl) control.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Central morphine increased plasma IL-6 activity through opioid receptor activation. One week before experiments, animals were surgically implanted with a guide cannula in the lateral ventricle. On the day of the experiment, animals were pretreated for 30 min with a subcutaneous injection of either saline (S, 1 ml/kg) or naltrexone (N, 10 mg/kg) followed by microinjection into the lateral ventricle of either saline (5 µl) or morphine sulfate (M, 10 µg/5 µl). One hour after the last drug treatment, animals were sacrificed and plasma IL-6 was determined. Data are expressed as the mean units/milliliter ± S.E.M. for groups of six animals and were representative of two independent determinations. , p < 0.05 compared with all groups (one-way ANOVA, Newman-Keuls).

Effect of Naltrexone Pretreatment on Morphine-Induced Plasma IL-6 Activity. To determine whether the elevation of plasma IL-6 by morphine was mediated through activation of opioid receptors, the animals described above were pretreated with the opioid receptor antagonist naltrexone. Animals were pretreated with a subcutaneous injection of either saline (1 ml/kg) or naltrexone (10 mg/kg). Thirty minutes later, saline (5 µl) or morphine (10 µg/5 µl) was microinjected into the lateral ventricle, and animals were sacrificed 1 h after the last drug treatment. Naltrexone pretreatment had no effect on plasma IL-6 activity after saline treatment (Fig. 3). However, naltrexone pretreatment completely blocked the central morphine-induced elevation of plasma IL-6 activity.

Effect of Ganglionic Blockade on Morphine-Induced Plasma IL-6 Activity. Studies were initiated to determine whether activation of the autonomic ganglia after morphine administration may be involved in the elevation of plasma IL-6. Animals were pretreated with chlorisondamine, a quaternary neuronal nicotinic receptor antagonist, which has previously been shown to effectively inhibit peripheral autonomic neurotransmission (Schneider and Moore, 1955; Clarke et al., 1994). As shown in Fig. 4, morphine (10 mg/kg s.c.) treatment significantly stimulated plasma IL-6 activity. No effect on plasma IL-6 activity was observed in animals treated with chlorisondamine (0.5 mg/kg i.p.) alone. However, chlorisondamine pretreatment significantly decreased the morphine-induced elevation of plasma IL-6 activity. Based upon the dose-response study (Fig. 2), a higher dose of morphine was used for subsequent studies. Additional studies using a higher dose of both chlorisondamine (5 mg/kg i.p.) and morphine (20 mg/kg s.c.) also demonstrated that chlorisondamine pretreatment was accompanied by a blockade of the elevation of opioid-induced plasma IL-6 activity (data not shown).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Blockade of morphine-induced effects by chlorisondamine. Animals were pretreated for 30 min with an intraperitoneal injection of either saline (S, 1 ml/kg) or chlorisondamine (C, 0.5 mg/kg) and then were treated with a subcutaneous injection of either saline (S, 1 ml/kg) or morphine sulfate (M, 10 mg/kg). Animals were sacrificed 1 h after the last drug treatment, and plasma IL-6 was determined. Data are combined from two independent experiments and expressed as the mean ± S.E.M. IL-6 (units/milliliter) for groups of 10 to 11 animals. , p < 0.01 compared with all groups (one-way ANOVA, Newman-Keuls).

Effect of Adrenalectomy on Morphine-Induced Plasma IL-6 Activity. To determine the role of adrenal gland activation in elevation of plasma IL-6, animals were either sham-operated (sham) or bilateral adrenalectomized and a basal level of corticosterone (25 µg/ml) was supplied in the drinking water (ADX-CORT). To verify the completeness of adrenalectomy, plasma corticosterone levels were measured. As shown in Fig. 5A, sham animals were injected subcutaneously with either saline (1 ml/kg) or morphine (20 mg/kg). Saline-treated shams had basal levels of corticosterone (100 ng/ml), whereas animals treated with morphine had significantly elevated plasma corticosterone levels (approximately 5-fold). Comparison of corticosterone levels in saline-treated sham and ADX-CORT animals were not significantly different, although basal levels in ADX-CORT animals were approximately 3-fold lower. Unlike sham animals, morphine treatment (20 mg/kg s.c.) did not elevate plasma corticosterone in ADX-CORT animals.

View larger version (11K): [in this window] [in a new window]   Fig. 5.   Blockade of morphine-induced elevation of plasma corticosterone and IL-6 activity by adrenalectomy. One week before experiments, animals were sham-operated or adrenalectomized. Adrenalectomized animals were provided with corticosterone in the drinking fluid as described under Materials and Methods. Animals (5-6/group) were treated with a subcutaneous injection of either saline (1 ml/kg) or morphine (20 mg/kg) and sacrificed 1 h after treatment. Data are representative of two independent determinations. A, plasma corticosterone was measured using a radioimmunoassay, and data are expressed as the mean ± S.E.M. corticosterone (ng/ml). , p < 0.001 compared with all groups (one-way ANOVA, Newman-Keuls). B, plasma IL-6 activity is expressed as the mean ± S.E.M. IL-6 (units/milliliter). , p < 0.001 compared with all groups (one-way ANOVA, Newman-Keuls).

Effect of Adrenal Demedullation on Morphine-Induced Plasma IL-6 Activity. Because adrenalectomy prevented morphine-induced plasma IL-6 release, studies were initiated to elucidate whether the adrenal cortex or medulla was involved in the effect of morphine. Animals were either sham or adrenal demedullated (AD-DM) and were subcutaneously injected with either saline (1 ml/kg) or morphine (20 mg/kg). One hour after treatment, animals were sacrificed and IL-6 activity was determined.

View larger version (11K): [in this window] [in a new window]   Fig. 6.   Morphine elevated plasma IL-6 activity in adrenal demedullated rats. Animals were sham or AD-DM and were treated with either saline (1 ml/kg) or morphine (20 mg/kg). One hour after treatment, animals were sacrificed and plasma corticosterone and IL-6 were determined. The data presented (n = 11-13/group) are combined from two independent experiments. A, plasma corticosterone was measured using a radioimmunoassay. Data are presented as the mean ± S.E.M. corticosterone (ng/ml). , p < 0.05 compared with sham morphine (Mor); , p < 0.01 compared with sham saline (Sal) and AD-DM Sal; and , p < 0.001 compared with sham Sal and AD-DM Sal treatments (ANOVA, Newman-Keuls). B, plasma IL-6 levels were measured and data are presented as the mean ± S.E.M. IL-6 (units/milliliter). , p < 0.05 compared with sham Mor; , p < 0.01 compared with sham; and AD-DM Sal, and , p < 0.001 compared with sham and AD-DM Sal treatments (one-way ANOVA, Newman-Keuls).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The studies described here provide the first characterization of the time and dose dependence of an opioid-mediated elevation of plasma IL-6. The results show that the effect of systemic morphine administration on IL-6 release is both rapid and dose-dependent. Plasma IL-6 is elevated 2-fold within 30 min of systemic morphine administration. In addition to being rapid, IL-6 elevation after morphine treatment is prolonged and maintained for at least 2 h. Because the half-life of plasma IL-6 is very short (t_1/2 = 1.5-5 min) (Castell et al., 1988), the maintenance of a prolonged elevation of circulating IL-6 levels is most likely due to either a continual release of IL-6 or a decrease in the clearance of IL-6 from plasma. Systemic morphine treatment was also shown to elevate plasma IL-6 in a dose-dependent manner, but requires high doses (10-30 mg/kg). These data suggest that morphine acts through a receptor-mediated effect with an ED₅₀ value of 13.8 mg/kg. However, this dose is relatively higher than the dose of morphine required to inhibit lymphocyte proliferative responses (Bayer et al., 1992) or to produce analgesic responses (for review, see Martin, 1984). The high-dose requirements suggest the effect of morphine on plasma IL-6 may have greater impact during the clinical use of opioids to produce anesthesia or in the recreational illicit use of these drugs.

These studies provide additional evidence for a critical role of central opioid receptors in the regulation of peripheral IL-6 levels. First, comparison of the dose of morphine that was required to produce the elevation of plasma IL-6 after different routes of administration suggested involvement of central opioid receptors. Subcutaneous administration of morphine (10 mg/kg) produced a similar elevation of plasma IL-6 as did microinjection of morphine (10 µg/5 µl) into the lateral ventricle. In considering the dose-response study (Fig. 2), if the central dose of morphine (10 µg) would be administered subcutaneously no effect would be predicted, because the effect of morphine on plasma IL-6 required subcutaneous doses approximately 1000-fold greater. Second, the elevation of plasma IL-6 by central microinjection of morphine was completely antagonized by pretreatment with naltrexone, an opiate receptor antagonist, thus demonstrating for the first time that this effect of morphine is opioid receptor-dependent. Our observations are in agreement with previous studies suggesting a role for central opioid receptors in the elevation of serum IL-6 after administration of IL-1 (De Simoni et al., 1993). Additional direct evidence for a role of opioids was the observation that intraperitoneal injection of either morphine (15 mg/kg) or etorphin (10 mg/kg) and central administration of morphine (10 µg) elevated serum IL-6 levels at 2 h after administration (Bertolucci et al., 1996). The results reported here confirm these initial studies and further demonstrate that the central receptor mediating the effect of morphine is naltrexone-sensitive. Taken together, these data suggest that direct stimulation of central opioid receptors by morphine activates opioid pathways that may be involved in altering circulating levels of cytokines, such as IL-6.

The mechanism of opioid-induced IL-6 levels has not been previously examined. In the present studies, it is shown that chlorisondamine pretreatment completely blocked the morphine-induced elevation of plasma IL-6 (Fig. 4). These data are consistent with the fact that central opioid administration has also been shown to modulate the sympathetic nervous system as well as the sympathoadrenal axis. For example, microinjection of morphine into the central nervous system of rats has been shown to increase blood pressure (Conway et al., 1983) and to elevate plasma epinephrine, norepinephrine, and/or dopamine (Van Loon et al., 1981; Conway et al., 1983; Appel et al., 1986). Chlorisondamine is a quaternary neuronal nicotinic receptor antagonist that has been shown to effectively block peripheral ganglionic neurotransmission (Schneider and Moore, 1955; Clarke et al., 1994). High-dose chlorisondamine (10 mg/kg) administration has been reported to access the brain and produce a prolonged blockade of central nervous system nicotinic receptors (Clarke et al., 1994). However, these studies were designed to use a considerably lower dose of chlorisondamine (0.5 mg/kg), which has been previously shown to completely block the effect of epibatidine (a potent nicotinic receptor agonist) on lymphocyte proliferation responses to mitogen (Mellon and Bayer, 1999). In contrast, chlorisondamine (0.5 mg/kg) did not block epibatidine stimulation of the hypothalamic-pituitary-adrenal axis (Mellon and Bayer, 1999), which is mediated by central nicotinic receptor stimulation (Matta et al., 1987). Consistent with a role of the autonomic ganglia to elevate IL-6, pretreatment of animals with chlorisondamine (3 mg/kg) also prevented IL-1-stimulated sera IL-6 levels (Kitamura et al., 1998). Therefore, these data provide the first evidence that stimulation of peripheral ganglionic nicotinic receptors was required for the elevation of plasma IL-6 by morphine.

Because central opioid receptor stimulation also activates the sympathoadrenal axis (Van Loon et al., 1981; Conway et al., 1983; Appel et al., 1986), studies were conducted to determine whether the adrenal glands were involved in this response. The results presented here demonstrated that in adrenalectomized animals, receiving replacement of corticosterone in their drinking water, morphine administration did not elevate IL-6 levels (Fig. 5). Thus, intact adrenal glands were required for this effect of morphine. Additional studies examined the potential of the adrenal cortex or medulla to be involved in this response. In contrast to complete adrenalectomy, surgical removal of the adrenal medulla did not prevent the increase in plasma IL-6 by morphine (Fig. 6). Thus, these data provide the first evidence that morphine-mediated effects on IL-6 are adrenal-dependent and require that the adrenal cortex remain intact.

Although these studies do not directly demonstrate IL-6 is released from the adrenal gland, the adrenal demedullation studies suggest that IL-6 release is dependent on the adrenal cortex. The source of IL-6 in circulation after treatment with morphine is unknown. Interleukin-6 is produced by a variety of cells of both immune and nonimmune origin (for review, see Akira et al., 1993). The results presented here indicate that morphine rapidly induced plasma IL-6 within 30 min and remained elevated for at least 2 h (Fig. 1), required ANS stimulation (Fig. 4), and was adrenal cortex-dependent (Figs. 5 and 6). Based on the adrenalectomy and adrenal demedullation studies, the adrenal cortex may be a potential source of IL-6 release after morphine treatment. The adrenal gland has been demonstrated to be innervated by pre- and postganglionic fibers of the autonomic nervous system (Kesse et al., 1988), and IL-6 mRNA has been detected by in situ hybridization in the adrenal gland (Gadient et al., 1995). Because it is currently unknown which cell type releases IL-6 from the adrenal gland, two potential candidates are the resident macrophages and the adrenal cortical cells. Interestingly, primary cultures of adrenal cortical cells have been stimulated to produce IL-6 by a variety of neurotransmitters, hormones, and cytokines (Nussdorfer and Mazzocchi, 1998). Collectively, these data suggest that elevation of plasma IL-6 by morphine may occur via release by adrenocortical cells.

An alternate possibility is that morphine treatment leads to the release of an adrenal-derived factor that elevates plasma IL-6 levels. Several adrenal cortex-derived factors could be potential candidates for the elevation of plasma IL-6. First, there appears to be a correlation between the elevation of plasma corticosterone and plasma IL-6. In both the adrenalectomy and adrenal demedullation studies, the elevation of IL-6 activity mirrored that of plasma corticosterone. Furthermore, the ED₅₀ value of morphine (7.8 mg/kg) to elevate plasma corticosterone has been previously reported by this laboratory (Mellon and Bayer, 2001), which is approximately 2-fold lower than that required to increase plasma IL-6 (Fig. 2), thus suggesting that corticosterone is increased before the elevation of plasma IL-6. Although these two effects of morphine appear related there is evidence that the elevation of plasma corticosterone is independent of the increase in plasma IL-6. The promoter for the IL-6 gene contains a glucocorticoid response element that negatively regulates IL-6 gene transcription (Ray et al., 1990). Additionally, chlorisondamine pretreatment completely blocked the elevation of plasma IL-6, but did not block the elevation of plasma corticosterone (data not shown). Taken together, the elevation of corticosterone after acute morphine treatment does not appear to be the adrenal cortical factor that increases plasma IL-6. A second potential adrenal-derived factor is the proinflammatory cytokine IL-1, which is localized within the adrenal cortex (Schultzberg et al., 1989) and stimulates the release of IL-6 from adrenal cortical cells (Judd and MacLeod, 1995) as well as many other tissues (for review, see Heinrich et al., 1990). Acute morphine treatment to mice has been shown to increase IL-1 and tumor necrosis factor- levels released from phytohemagglutinin-stimulated splenocyte cultures (Pacifici et al., 2000), yet it is unknown whether this elevation also occurs in the adrenal gland. Collectively, these findings suggest that an adrenal cortex derived factor such as corticosterone may not be involved in this effect of acute morphine treatment, whereas the cytokine IL-1 may potentially be involved in this response.

In summary, acute morphine administration rapidly increases plasma IL-6 levels (within 30 min) largely through activation of central opioid receptors. This effect of morphine is prolonged (at least 2 h) and is mediated through stimulation of the autonomic ganglia. The adrenalectomy studies suggest that the effect of morphine to elevate plasma IL-6 is adrenal-dependent. An intact adrenal cortex appears to be largely required for the morphine-induced effect on plasma IL-6, because adrenalectomy, and not adrenal demedullation, completely prevented the morphine-induced elevation of plasma IL-6. This observation is unique, because few responses to date have implicated the autonomic innervation of the adrenal cortex in mediating effects on the immune system. Whether the mechanism of the increased plasma IL-6 levels is due to the release of IL-6 directly from adrenal cortical cells or indirectly via an adrenal cortex derived factor remains to be determined. These data suggest that opioid administration for clinical uses, such as analgesia and/or anesthesia or abuse via illicit use may potentially alter plasma IL-6 homeostasis.