The Effects of Morphine, Nicotine and Epibatidine on Lymphocyte Activity and Hypothalamic-Pituitary-Adrenal Axis Responses

Materials and Methods

Animals.

Pathogen-free male Sprague-Dawley rats (200–225 g upon receipt) were obtained from Taconic Laboratories (Germantown, NY). Animals were group-housed three per cage with microisolator tops and were provided food (Purina rat chow, Ralston Purina Co., St. Louis, MO) and water ad libitum. The light cycle was regulated automatically (lights on at 6 AM, off at 6 PM) and temperature was maintained at 23 ± 1°C. All animals were allowed to acclimate to this environment for 1 week before use in an experiment or surgical implantation of a guide cannula into the ventricular system. All experiments were conducted between 7 and 11 AM.

Drugs.

Morphine sulfate was generously provided by the National Institute on Drug Abuse (Research Triangle Park, NC). Epibatidine dihydrochloride was purchased from Research Biochemicals International (Natick, MA). Nicotine hydrogen tartrate and naltrexone hydrochloride were purchased from Sigma Chemical Co. (St. Louis, MO). Chlorisondamine chloride (Ecolid) was kindly provided by Ciba-Geigy Corporation (Summit, NJ). All drugs were dissolved in sterile isotonic saline, which also served as the control treatment in these studies. The injection volume for all systemic studies was 1 ml/kg and the route of administration was as indicated in the figure legends. The doses of all drugs reported in these studies are based on the molecular weight of the salt and not that of the freebase; however, the freebase doses of nicotine are also provided in parentheses following that of the salt.

Cannulation and Microinjection Procedures.

Animals were anaesthetized 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 consisting of deionized water (43 ml), propylene glycol (26.6 ml), ethyl alcohol (12.45 ml), and pentobarbital (18 ml, 50 mg/ml). Surgical implantation of a guide cannula into either the right lateral ventricle (final coordinates relative to bregma: anterior-posterior = −3.6; medial-lateral = −4.6, dorsal-ventral = −7.5) or the dorsal third ventricle (final coordinates relative to bregma: anterior-posterior = −4.2; medial-lateral = 0; dorsal-ventral = −3.4) was completed based on the atlas of Paxinos and Watson (1997). After surgery, a stylet was inserted into the guide cannula to maintain sterility and patency. All animals were given gentamicin sulfate (4 mg, s.c.) and allowed to recover for approximately 1 week before experimentation.

Microinjection of drugs was accomplished in freely moving awake animals as described previously (Hernandez et al., 1993). Briefly, test animals were removed from the home cage, the protective stylet was removed, and the internal cannula, which extended 1 mm beyond the end of the guide cannula, was inserted. The animal was placed into a separate cage for the injection procedure. The injection volume for all centrally administered drugs was 2 μl, administered over 45 s using a Sage Infusion pump set to dispense 2.7 μl/min. The internal cannula remained in place for an additional 75 s to ensure drug dispersal into the ventricle. Flow of the drug solution was monitored via tracking a small air bubble created in the polyethylene tubing that separated the drug solution and distilled water. The animal was returned to the home cage for the duration of the treatment period. Nociception was measured as described 30 min after either morphine administration or 3 min after either nicotine or epibatidine treatment. Animals were sacrificed either 1 or 2 h later, as indicated in the figure legends.

Whole-Blood Lymphocyte Proliferation Assays.

Animals were sacrificed via decapitation and trunk blood was collected in 50-ml polypropylene tubes containing 0.2 ml of heparin (1000 U/ml) and immediately placed on ice until processed. Whole blood was diluted 1:5 with cold RPMI 1640 cell culture media (Gibco, Grand Island, NY) containing 1% fetal bovine serum and gentamicin (20 μg/ml). Triplicate samples of blood (0.1 ml) were plated into 96-well microtiter plates containing 0.1 ml/well of the T cell-specific mitogen concanavalin A (ConA; 0, 2, 4, and 6 μg/well) and incubated for 72 h at 37°C with 8% CO₂. Cells were pulsed with 0.5 μCi of [methyl-³H]thymidine (6.7 Ci/mmol; New England Nuclear, Boston, MA) in a 20 μl volume and incubated for an additional 24 h. Samples were lysed with distilled water and harvested onto glass fiber filters using a 96-well cell harvester (Brandel, Gaithersburg, MD). The amount of labeled DNA was determined via liquid scintillation spectrophotometry (Beta Plate; L.K.B. Pharmacia).

Spleen and Thymus Lymphocyte Proliferation Assays.

After decapitation, spleen and thymus were removed via sterile forceps and each placed in cold RPMI 1640 cell culture media (10 ml) containing 1% fetal bovine serum and gentamicin (20 μg/ml). The tissues were teased apart with sterile forceps, and the resulting cell suspension was washed twice in cold RPMI 1640 media and adjusted to a concentration of 2 × 10⁶ cells/ml for spleen and 5 × 10⁶ cells/ml for thymus. Cell preparations (100 μl) were then cultured with increasing concentrations of ConA (0, 0.0625, 0.125, 0.25, and/or 0.5 mg/well) as described for the blood proliferation assay to provide both optimal and suboptimal concentrations of mitogen.

Antinociception.

Antinociception was measured by the radiant heat tail-flick method (D’Amour and Smith, 1941), as described previously (Hernandez et al., 1993). All animals were acclimated to handling and the tail-flick device for 2 to 3 days before experimental manipulations. Light intensity was controlled to provide a predrug latency between 2 and 3 s. A cut-off of 8 s was used to prevent damage to tail tissue. Data were expressed as the percent maximum possible effect (% MPE) as defined below:

Plasma Corticosterone Assay.

Plasma samples were maintained at −20°C until assayed via solid phase ¹²⁵I radioimmunoassay kits purchased from ICN Biochemicals, Inc. (Costa Mesa, CA). Because baseline levels of plasma corticosterone were found to vary depending on the nature of the experimental manipulations involved in the assays, matched treatment controls were included in all studies.

Histological Analysis.

Brains of centrally treated animals were removed postmortem and stored in 10% formaldehyde followed by 20% sucrose solution. All brains from drug-treated animals were cut into 40-μm sections on a freezing microtome (HistoSTAT Cryostat Microtome, model 975c; Reichert, Buffalo, NY) and placed on gelatin-coated slides for microscopic verification of cannula placement by comparison to the atlas of Paxinos and Watson (1997).

Statistical Analysis.

Only animals whose cannula were determined to be accurately placed in the ventricle were used for data analysis. For proliferation assays, the median of triplicate samples was determined for each concentration of mitogen and the resulting dose-response curves generated. To compare the ConA dose-response curves, nonlinear regression analysis was completed to generate the best-fit curve using GraphPad Prism software (San Diego, CA). For blood proliferation data, regression analysis indicated that the curves best fit a sigmoidal dose-response curve with the following equation:Y = E_min + (E_max − E_min)/(1 + 10^LogEC50 − X), where Y is the proliferation response measured in cpm,E_min is the calculated minimum proliferation response (i.e., absence of mitogen), E_max is the calculated maximal proliferation response, EC₅₀ is the concentration of mitogen producing 50% of the maximal response, andX is the logarithm of concentration. From the resulting equations the values for maximal response (E_max) and EC₅₀ were obtained and compared. Significant differences inE_max were interpreted as alterations in the magnitude of the response of lymphocytes to ConA treatment, whereas alterations in EC₅₀ values were interpreted as alterations in the sensitivity of the lymphocytes to ConA treatment. For all analyses, Student’s t test was used to compare results from experiments involving only two treatment groups, whereas a one-way analysis of variance (ANOVA) with Newman-Keuls post hoc analysis was conducted for comparison of three or more groups. When no significant difference between basal proliferation in the control groups occurred, the results obtained in individual experiments were combined directly and proliferation data were expressed as the mean cpm ± S.E.M. For all parameters, any value greater than 2 SDs from the mean of the treatment group was omitted.

Results

Comparison of the Effects of Morphine with Nicotinic Agonists.

The antinociception, lymphocyte suppression, and activation of the HPA axis induced by morphine were compared to those of the prototypical ganglionic agonist nicotine and the potent nicotinic agonist epibatidine. The rationale for the dose of each drug was based on the results of previously published studies in our and other investigators’ laboratories (Tripathi et al., 1982; Bayer et al., 1992; Caggiula et al., 1992; Qian et al., 1993; McAllister et al., 1994). A representative study comparing the antinociceptive effects of these compounds as measured by tail-flick latency is shown in Fig.1. Both nicotine (2.85 mg/kg, s.c. = 1.0 mg/kg freebase) and epibatidine (5 μg/kg, s.c.) treatment produced approximately 60% of the maximum possible effect 3 min after administration, whereas the response to morphine (10 mg/kg, s.c.) was not significant at this time point. However, when tested 30 min after drug treatment, the antinociceptive effect of morphine was maximal (Fig. 1), whereas by this time, nicotine and epibatidine were only marginally effective (data not shown).

In addition to their antinociceptive effects, treatment with morphine, nicotine or epibatidine significantly decreased the magnitude (E_max) of the peripheral blood lymphocyte proliferation response to the T cell mitogen ConA when compared with responses of saline-treated control animals (Fig.2). No significant alteration in the sensitivity (EC₅₀) of lymphocytes to ConA was noted after any drug treatment. In contrast to the effect of these agents on peripheral blood lymphocytes, no significant effects were detected in lymphocytes isolated from either the spleen or the thymus under these conditions (data not shown).

Since activation of the HPA axis has been reported to produce altered immune parameters (Parrillo and Fauci, 1979), plasma levels of corticosterone were also measured in these studies. Figure 3represents the combined results from two studies where animals were sacrificed at either 1 or 2 h after drug administration. As depicted, all drug treatments produced significantly elevated plasma levels of corticosterone compared with saline treatment when measured at 1 h. However, only morphine treatment significantly continued to sustain elevated corticosterone levels for 2 h (Fig.3).

Dose Dependence of Nicotine-Induced Alterations in Peripheral Blood Lymphocyte Proliferation.

Unlike epibatidine or morphine, the dose of nicotine used in these studies (2.85 mg/kg, s.c. = 1.0 mg/kg freebase) produced some seizure activity. Because this could complicate the interpretation of the results, experiments were carried out to determine whether the effects of nicotine on lymphocyte responses could be observed at lower doses. Figure 4represents the combined results from two studies testing nicotine at four different doses. ANOVA indicated that nicotine at a dose of either 2.0 mg/kg (0.7 mg/kg freebase) or 2.85 mg/kg (1.0 mg/kg freebase) significantly decreased the magnitude of blood lymphocyte proliferation responses compared with saline-treated controls and doses of nicotine at either 0.5 mg/kg (0.175 mg/kg freebase) or 1 mg/kg (0.35 mg/kg freebase). Although not significant, a dose of 1 mg/kg (0.35 mg/kg freebase) slightly decreased peripheral blood lymphocyte responses, whereas the lowest dose of nicotine (0.5 mg/kg, s.c. = 0.175 mg/kg freebase) slightly elevated the response. The sensitivity of lymphocytes to ConA as measured by EC₅₀ values was not significantly altered. In these studies, nicotine at doses of 2 mg/kg (0.7 mg/kg freebase) or 2.85 mg/kg, s.c. (1.0 mg/kg freebase) also significantly increased latency to tail-flick, whereas lower doses were without any detectable effect (data not shown).

The Role of Central Nicotinic Receptor Activation on Peripheral Lymphocyte Activity.

As nicotine and epibatidine readily enter the CNS, the role of central nicotinic receptors in the effect of these drugs on peripheral blood lymphocyte proliferation was examined by administering these compounds directly into the ventricular system of freely moving conscious rats. Figure 5shows the effects of administration of nicotine at doses of 30 and 240 nmol into the third ventricle. Neither of these doses of nicotine significantly altered either the magnitude (E_max) or sensitivity (EC₅₀) of blood lymphocytes to ConA treatment. To maximize the central tissues potentially exposed to injected drug, the higher dose of nicotine (240 nmol) was administered into the lateral ventricle. As shown in the inset in Fig. 5, nicotine treatment was still without effect on blood lymphocyte proliferation. No significant alterations in spleen lymphocyte proliferation were noted after microinjection into either the third or the lateral ventricle under these conditions (data not shown).

To evaluate whether epibatidine may have centrally mediated effects on lymphocyte activity, the drug was also administered directly into the lateral ventricle of the rat. The doses of epibatidine used (5, 50 and 500 ng) were based on our previous experience with central administration of morphine (Hernandez et al., 1993), which detected significant effects on blood lymphocyte proliferation and antinociceptive responses after microinjection of doses 1000-fold lower than those used systemically. After the central administration of epibatidine, the highest dose tested (500 ng) produced a slight but insignificant decrease in the magnitude of the peripheral blood lymphocyte proliferation response (Fig.6).

Nicotinic Antagonist Studies.

To further examine whether the effects of epibatidine were mediated predominantly by peripheral nicotinic receptors, chlorisondamine (0.5 mg/kg, i.p.), the quaternary nicotinic antagonist, was administered 30 min before epibatidine (5 μg/kg, s.c.) treatment. Animals were sacrificed 1 h later, at a time when epibatidine significantly elevates corticosterone levels (Fig. 3). The rationale for the dose of chlorisondamine was based on a review of the published literature. Although higher doses of chlorisondamine have been reported in the literature (Irwin et al., 1988; Britton and Indyk, 1989; Nikolarakis et al., 1989; Saito et al., 1991), we chose a lower dose to decrease the potential for this quaternary drug to gain access to the CNS (Clarke, 1984). As previously shown at 2 h (Fig. 2), epibatidine treatment also significantly suppressed the magnitude of the blood lymphocyte proliferation response within 1 h (Fig. 7). Furthermore, pretreatment with chlorisondamine completely antagonized the suppressive effects of systemic epibatidine on the blood lymphocyte proliferation response.

To examine the possibility that this dose of chlorisondamine may be blocking central nicotinic receptors, plasma levels of corticosterone were measured. As shown in Fig. 8, epibatidine significantly elevated corticosterone 1 h after treatment; however, pretreatment with chlorisondamine did not antagonize this effect. Thus, in contrast to the effects of chlorisondamine on blood lymphocyte proliferation, chlorisondamine did not antagonize the effects of epibatidine on plasma levels of corticosterone (Fig. 8).

Naltrexone Antagonist Studies.

To determine whether the effects of epibatidine were due to interaction with opioid receptors, animals were pretreated with the opioid receptor antagonist naltrexone (10 mg/kg, s.c.) 30 min before epibatidine treatment (5 μg/kg, s.c.). Animals were sacrificed 1 h after epibatidine treatment. As shown in Fig. 9, epibatidine treatment significantly suppressed the blood lymphocyte proliferation response. In contrast to the effects of the nicotinic antagonist chlorisondamine, the opioid antagonist naltrexone did not antagonize the effect of epibatidine. Likewise, epibatidine treatment significantly elevated plasma levels of corticosterone in an opioid-independent manner (Fig.10).

Additional experiments were carried out under identical experimental conditions to confirm that the effects of morphine were sensitive to antagonism by naltrexone. Morphine treatment (10 mg/kg, s.c.) significantly decreased blood lymphocyte proliferation compared with all other treatment groups (Fig. 11). However, in contrast to the lack of an effect of naltrexone in epibatidine-treated animals, naltrexone (10 mg/kg, s.c.) pretreatment completely antagonized the effect of morphine on the blood lymphocyte proliferation response (Fig. 11). Likewise, naltrexone antagonized the effect of morphine on the HPA axis (Fig.12).

Discussion

The close similarity of the reported findings with nicotine on immune function (Caggiula et al., 1992; McAllister et al., 1994) with the results of the studies conducted in our laboratory with morphine led us to further explore the relationship between nicotinic and opioid-induced alterations in immune function. The results (Figs.1-3) with acute systemic morphine, nicotine and epibatidine treatment presented here demonstrated that each of these compounds produce: 1) antinociception, 2) decreased magnitude of peripheral blood lymphocyte proliferation responses to mitogen without altering the sensitivity of the lymphocytes, 3) no alteration of either splenic or thymic proliferation responses, and 4) an elevation of circulating corticosterone levels. Collectively, these results indicate that the effects of systemic morphine are largely mimicked by both nicotine and epibatidine treatment.

Although considerable evidence supports the involvement of a central site of action for systemic morphine on blood lymphocyte responses (Hernandez et al., 1993; Mellon and Bayer, 1998), the site(s) of action for nicotine-induced immunomodulatory effects has not been well characterized. Based on several observations in this report, the CNS does not appear to be the predominant site of action of nicotine- and epibatidine-induced alterations in peripheral blood lymphocyte responses. First, central administration of either nicotine or epibatidine failed to significantly alter blood lymphocyte proliferation (Figs. 5 and 6). Although the highest dose of epibatidine tested did produced a slight trend toward a decrease in blood lymphocyte proliferation, this dose was only 10-fold lower than the dose used for systemic administration. Second, chlorisondamine, a quaternary nicotinic antagonist, completely antagonized the effects of systemic epibatidine on blood lymphocyte proliferation responses. High doses of chlorisondamine (10 mg/kg, s.c.) have been reported to produce prolonged blockade of central nicotinic receptors (Clarke, 1984; Clarke et al., 1994), suggesting that some of the drug gained access to the central compartment under these conditions. However, the lower doses of chlorisondamine used in the present study (0.5 mg/kg, i.p.) had no effect on the stimulation of the HPA axis by epibatidine (Fig. 8). Because nicotinic receptor-mediated activation of the HPA axis appears to be due to central nicotinic receptors (Cam et al., 1979; Matta et al., 1987), these data suggest that it is unlikely that chlorisondamine gained access to the CNS. The conclusion that this dose of chlorisondamine does not gain access to the CNS is further supported by studies examining nicotine-induced antinociception (Caggiula et al., 1995). These observations also suggest that systemic administration of epibatidine decreases peripheral blood lymphocyte responses by stimulating peripheral rather than central nicotinic receptors.

Similar to the findings reported here, Caggiula et al. (1992) observed that systemic nicotine administration to the Sprague-Dawley rat decreased peripheral blood lymphocyte proliferation responses in a dose-dependent manner. Additional studies by this group also demonstrated that acute systemic nicotine treatment in the Sprague-Dawley rat had no effect on splenic lymphocyte proliferation responses to ConA, suggesting that the blood lymphocytes were specifically targeted by nicotine (McAllister et al., 1994). However, in contrast to these observations, in the Lewis rat, Fecho et al. (1993) found that systemic administration of low doses of 1,1-dimethyl-4-phenylpiperazinium (DMPP), a quaternary nicotinic agonist, increased lymphocyte proliferation responses of peripheral blood lymphocytes at low doses of DMPP, an effect which appeared to return to baseline with higher doses. DMPP treatment suppressed the proliferation response of splenic lymphocytes, splenic natural killer cell cytolytic activity, interleukin-2 production, interferon-γ production, and total white blood cell counts (Fecho et al., 1993). Explanations for the reported differential sensitivity of lymphocytes from blood and spleen compartments after nicotine and epibatidine compared with those with DMPP are not clear. Strain differences may account, in part, for some of these observations. For example, Fecho et al. (1993) used the Lewis rat model, whereas Caggiula et al. (1992),McAllister et al. (1994), and our studies used the Sprague-Dawley rat. Differences in autonomic tone have been suggested between these two animals strains (Dimitrova et al., 1995). Regardless of these compartmental differences, these studies collectively suggest that, like morphine, acute systemic nicotinic agonists produce alterations in peripheral lymphocyte activity. However, the studies presented in this report suggest that, unlike acute morphine, nicotinic agonists alter lymphocyte activity via peripheral nicotinic receptors rather than central pathways.

In conjunction with our previous report (Flores LR et al., 1996), these studies collectively provide strong evidence that opioid receptors are located upstream from nicotinic receptors in a neuroimmunomodulatory pathway. This conclusion is supported by the following observations: 1) acute morphine appears to act only at central opioid receptors (Hernandez et al., 1993; Mellon and Bayer, 1998b); 2) nicotinic agonists appear to act predominantly at peripheral nicotinic receptors; 3) an opioid antagonist blocks the effect of morphine without altering those of the nicotinic agonist epibatidine; and 4) the peripherally acting nicotinic antagonist chlorisondamine antagonized both the effects of morphine (Flores LR et al., 1996) and epibatidine (Fig. 7). Therefore, these pathways appear to overlap at the level of the peripheral nicotinic receptor, presumably at the autonomic ganglia. In addition, as both opioid and nicotinic receptor activation stimulate corticosterone release via central mechanisms, the neuronal pathways mediating activation of the HPA axis may overlap. The results of the studies with naltrexone pretreatment of epibatidine, however, suggest that an opioid receptor is not downstream from the nicotinic receptor in the pathway whereby epibatidine stimulates corticosterone release.

The results reported here demonstrating the effects of chlorisondamine pretreatment on epibatidine-induced responses also provide evidence that the suppressive effect of epibatidine on lymphocyte proliferation is not mediated by the increase in corticosterone levels. This conclusion is based on the observation that chlorisondamine pretreatment antagonized epibatidine-induced suppression of blood lymphocyte proliferation but did not antagonize the centrally mediated increase in plasma corticosterone. These results support the conclusion of Caggiula et al. (1992), who reported that the suppressive effect of systemic nicotine on blood lymphocyte proliferation were largely independent of plasma corticosterone. The observation that the effect of epibatidine is antagonized by chlorisondamine even in the presence of elevated corticosterone supports the more global observation that transient elevation of glucocorticoids are not necessarily immunosuppressive (Keller et al., 1983; Keller et al., 1988; Bayer et al., 1990; Flores et al., 1990).

Overall, the results are consistent with the conclusion that activation of the autonomic ganglia appear to lead to decreased peripheral blood lymphocyte proliferation. However, the possibility for a direct effect of nicotinic agonists on lymphocyte proliferation response has not been eliminated by these studies. Although there are a few published reports which provide evidence for nicotinic receptors on lymphocytes (Richman and Arnason, 1979; Maslinski et al., 1980; Konno et al., 1986; Menard and Rola-Pleszczynski, 1987; Toyabe et al., 1997), the pharmacology and functional significance of these receptors are not entirely clear.

In summary, these studies provide the first comparison of the immune effects of nicotine with the potent nicotinic agonist epibatidine and suggest that nicotine and epibatidine treatment appear to mimic the effects of morphine on peripheral lymphocyte responses. These treatments appear to decrease the overall magnitude of the response to mitogen treatment without clearly altering the sensitivity of the cells to this potent stimuli. Although the physiological significance of this altered cellular responsiveness is not known, this profile resembles that which would be produced in the presence of a noncompetitive antagonist to immunological stimuli. Given this potential immunomodulatory role, activation of opioid or nicotinic receptors may lead to a functional, although attenuated, immune response. Furthermore, these studies provide evidence that nicotinic agonists produce their immunomodulatory effects on blood lymphocyte proliferation by acting predominantly at peripheral nicotinic receptors. In addition, nicotinic receptor mediated immunomodulation, like that produced by opioids, appear to be independent of their ability to activate the HPA axis. Collectively, the studies discussed here suggest that peripheral nicotinic receptors appear to be located downstream of central opioid receptors in an neuroimmunomodulatory pathway.