Introduction

Top Abstract Introduction Methods Results Discussion References

Opioids play an important role in the regulation of the hypothalamo-pituitary axis. Opioid receptors of the three major subtypes (µ,  and ) are found in varying levels throughout the hypothalamus and posterior pituitary (Bondy et al., 1989; Desjardins et al., 1990), and opioid-containing nerve terminals make substantial synaptic contacts in many different hypothalamic nuclei (e.g., ARC, periventricular, paraventricular and supraoptic nuclei) (Goldsmith et al., 1991; Fitzsimmons et al., 1992). Opioids inhibit the activity of A₁₂ dopamine neurons (Loose et al., 1990; Manzanares et al., 1992), which accounts for the stimulatory effect of opioids on prolactin secretion (Kapoor and Willoughby, 1990) from the anterior pituitary. Opioids also regulate the firing patterns of magnocellular neurons in the supraoptic nucleus (Bourque et al., 1993), thereby inhibiting the secretion of oxytocin and vasopressin from the posterior pituitary (Bondy et al., 1989; Russell et al., 1992). At the cellular level, opioids inhibit hypothalamic neurons in one of two ways. In supraoptic neurons -opioids regulate magnocellular firing by attenuating Ca⁺⁺ currents (Bourque et al., 1993), whereas in ARC neurons µ- and -opioids inhibit cell firing by activating an inwardly rectifying K⁺ conductance (Loose and Kelly, 1990).

The inhibitory neurotransmitter GABA also plays an important role in the regulation of the hypothalamo-pituitary axis. GABA_A receptor subunits and the biosynthetic enzyme for GABA, glutamic acid decarboxylase, are expressed throughout the hypothalamus (Tappaz et al., 1977; Inglefield et al., 1994), and GABA is found in the majority of hypothalamic nerve terminals (Decavel and van den Pol, 1992). Activation of GABA receptors inhibits the activity of A₁₂ dopamine neurons (Wagner et al., 1994), thereby stimulating prolactin secretion (Kimura et al., 1993; Mäkinen et al., 1993; Wagner et al., 1994) from the anterior pituitary. At the cellular level, activation of GABA receptors inhibits cell firing by activating a Cl ionophore that facilitates Cl influx (i.e., the GABA_A receptor) (Inglefield et al., 1994) or by activating an inwardly rectifying K⁺ conductance (i.e., the GABA_B receptor) (Bowery, 1989; Loose et al., 1991).

The opioid agonist DAMGO (Goldstein and Naidu, 1989) and the GABA_B receptor agonist baclofen (Bowery, 1989) inhibit -endorphin (Loose et al., 1991) as well as A₁₂ dopamine (Loose et al., 1990; Wagner et al., 1994) neurons within the ARC nucleus. The DAMGO effect is antagonized by naloxone with an equilibrium constant consistent with a blockade of µ-opioid receptors (Kelly et al., 1992). When a maximally effective dose of DAMGO is coadministered with a maximally effective dose of baclofen, no additional inhibitory effect is observed (Loose et al., 1991; Kelly et al., 1992). This suggests that, in ARC neurons, µ-opioid and GABA_B receptors are coupled to a common pool of ligand-gated, inwardly rectifying, K⁺ channels.

The opiate morphine is capable of binding µ-, - and -opioid receptors but binds µ-opioid receptors with approximately 2 orders of magnitude higher affinity, compared with  or  receptors (Goldstein and Naidu, 1989; Kristensen et al., 1995). Many studies have focused on the effects of chronic morphine treatment and the resultant development of tolerance to various opioid-regulated processes, such as analgesia (Roerig, 1995), inhibition of gastrointestinal smooth muscle contractility (Chavkin and Goldstein, 1984) and neuronal excitability in the locus coeruleus (Christie et al., 1987; Rasmussen et al., 1990). However, relatively few studies have examined the effect of chronic morphine on processes regulated by, or on neurons originating in, the hypothalamus. Much of what has been studied has focused on the development of tolerance to opioid receptor-mediated inhibition of oxytocin release (Russell et al., 1992). However, we recently found that cellular tolerance develops in ARC neurons to µ-opioid agonists (Zhang et al., 1996), with the latter most likely being due to an attenuation of the µ-opioid receptor/K⁺ channel coupling. Because µ-opioid and GABA_B receptor systems converge on the same effector, the possibility exists that cross-tolerance to the baclofen response might develop because of chronic morphine treatment.

The purpose of the present study was to test the hypothesis that in ARC neurons, including TH-positive neurons, tolerance develops to the membrane-hyperpolarizing effects of µ-opioid and GABA_B receptor activation as a result of chronic morphine treatment. To this end, intracellular recordings were made under current-clamp conditions in hypothalamic slices prepared from ovariectomized female guinea pigs treated with placebo or morphine pellets for 4 to 7 days. Dose-response relationships for DAMGO and baclofen were determined, to evaluate the potency and efficacy of these two agonists. The results reveal that, in A₁₂ dopamine neurons, cross-tolerance does not develop between µ-opioid and GABA_B receptor agonists, which suggests that convergence upon a common effector may not be a sufficient criterion for the expression of cross-tolerance between receptor systems. Preliminary results of this work have been presented in abstract form (Wagner et al., 1995).

	Methods

Top Abstract Introduction Methods Results Discussion References

Animals. Female Topeka guinea pigs (440-670 g) were obtained from our institutional breeding facility and maintained under conditions of constant temperature (72.4 ± 0.1°F) and light (lights on between 6:30 A.M. and 8:30 P.M.). Animals were housed individually, with food and water provided ad libitum. The surgical and sacrificing procedures described in the present study are in accordance with institutional guidelines based on National Institutes of Health standards.

Drugs and treatments. Pellets containing either placebo or 75 mg of morphine sulfate were obtained from the Research Technology Branch of the National Institute on Drug Abuse (Research Triangle, NC). (±)-Baclofen [(±)--(aminomethyl)-4-chlorobenzenepropanoic acid; Sigma Chemical Co., St. Louis, MO] was dissolved in 0.1 N HCl to a stock concentration of 40 mM. DAMGO (Peninsula Laboratories Inc., Belmont, CA) was dissolved in Milli-Q water to a stock concentration of 1 mM. TTX (Sigma Chemical Co.) was dissolved in Milli-Q water and diluted to the appropriate volume with 0.1% acetic acid (final concentration, 1 mM; pH 4-5). Aliquots of the baclofen, DAMGO and TTX stock solutions were stored at 80°C.

Hypothalamic slice preparation. On the day of experimentation, each animal was decapitated, and its trunk blood was collected for subsequent determination of serum morphine and metabolite levels. The brain was rapidly removed from the skull and rinsed with ice-cold aCSF [124 mM NaCl, 5 mM KCl, 2.6 mM NaH₂PO₄, 10 mM dextrose, 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, 2 mM MgSO₄, 2 mM CaCl₂], and the hypothalamus was immediately dissected. Four coronal slices (450 µm) through the rostro-caudal extent of the ARC were cut with the aid of a vibrotome. The slices were transferred to a multi-well auxiliary chamber containing oxygenated (95% O₂/5% CO₂) aCSF, where they were kept until electrophysiological recording. During experiments involving morphine-treated animals, slices were kept in aCSF containing 1 µM morphine to prevent withdrawal (Zhang et al., 1996).

Electrophysiology. During recording, slices were maintained in a chamber perfused with warmed (35°C) oxygenated aCSF. aCSF and all drugs (diluted with aCSF) were perfused via a peristaltic pump, at a rate of 1.5 ml/min. Microelectrodes were assembled from borosilicate glass pipettes (1.2-mm outer diameter; Dagan, Minneapolis, MN) pulled on a P-87 Flaming Brown puller (Sutter Instrument Co., Novato, CA) and were filled with a 3% biocytin solution in 1.75 M KCl, 0.025 M Tris (pH 7.4). Electrode resistances varied from 100 to 300 M. The V_m of hypothalamic neurons was measured under current-clamp conditions by intracellular recording from the ARC. Potentials were amplified and current was passed through the electrode using an Axoclamp 2A preamplifier (Axon Instruments, Foster City, CA). Current and V_m traces were stored on a digital oscilloscope (Tektronix 2230; Tektronix, Beaverton, OR) and were recorded on a chart recorder (Gould 2200; Gould Inc., Glen Burnie, MD). They also underwent analog/digital conversion with a CyberAmp 320 signal conditioner (for amplification) connected to a DigiData 1200 analog/digital converter (sampling frequency, 62 Hz), with subsequent storage on a computer containing Axotape software (Axon Instruments).

Histology. After recording, slices were placed in 10 µM forskolin for 15 to 30 min to increase the intracellular TH signal and were then fixed with 4% paraformaldehyde in Sorensen's phosphate buffer (pH 7.4) for 90 to 180 min (Ronnekleiv et al., 1990). They were immersed overnight in 30% sucrose dissolved in Sorensen's buffer and were frozen in Tissue-Tek embedding medium (Miles, Inc., Elkhart, IN) the next day. Coronal sections (16 µm) were cut on a cryostat and were mounted on slides coated with poly-L-lysine. These sections were washed with 0.1 M sodium phosphate buffer (pH 7.4) and then processed with streptavidin-fluorescein isothiocyanate as described previously (Ronnekleiv et al., 1990). After localization of the biocytin-filled neuron, the slides containing the appropriate sections were processed with either an affinity-purified polyclonal TH antibody (PEL-Freez, Rogers, AR) at a 1:1000 dilution or a monoclonal TH antibody (Ink-Star, Stillwater, MN) at a 1:3000 dilution, using fluorescence immunohistochemistry (Ronnekleiv et al., 1990).

Determination of serum morphine and metabolite levels. Morphine and its conjugated metabolites M6G and M3G were measured in serum by high-performance liquid chromatography coupled to UV absorbance detection, using a modification of a procedure that was originally described by Murphey and Olsen (1993). Briefly, the aforementioned compounds were extracted from individual 80-µl plasma aliquots using a C₁₈ extraction column (Bond-Elut; Varian, Harbor City, CA). They were then reconstituted in 300 µl of mobile phase consisting of 10 mM phosphate buffer (pH 2.1) containing 1 mM sodium dodecyl sulfate and 16.5% (v/v) acetonitrile, and 100 µl from each sample was injected onto an octadecyl-silica, reverse-phase analytical column (VeloSep, 100 × 3.2 mm, 3-µm sorbents; Applied Biosystems, Foster City, CA), at a flow rate of 1.4 ml/min. After separation, morphine, M6G and M3G signals were evaluated by an absorbance detector (Waters 486; Millipore, Milford, MA) with the lamp wavelength set at 214 nm. Peak areas from each sample were quantified by linear regression using standard curves generated with known quantities of the compounds. The lower limit of sensitivity for morphine, M6G and M3G was 5 ng/80 µl serum.

Statistical analyses. The homogeneity of variance was evaluated using Cochran's C test. Comparison between two groups were performed using either a two-tailed t test or a rank sum (Mann-Whitney) test. Comparisons among two or more groups were performed using either a one-way analysis of variance or a Kruskal-Wallis one-way analysis of variance followed by the Mann-Whitney test. Differences were considered statistically significant if the probability of error was <5% (P < .05).

	Results

Top Abstract Introduction Methods Results Discussion References

The present study included a total of 83 cells obtained from 73 animals. There were nine instances where cells were recorded in two different slices from the same animal and one instance where two cells were recorded from the same slice. There also were five instances in which both DAMGO and baclofen dose-response relationships were generated in the same cell.

Figure 1 illustrates representative dose-response relationships for DAMGO and baclofen in hypothalamic neurons. Both DAMGO and baclofen elicited a dose-dependent membrane hyperpolarization. Examination of the V/I plots in figure 2 reveals that the reversal potential for both DAMGO- and baclofen-induced hyperpolarization was near 95 mV, which closely approximates the calculated Nernst equilibrium potential for K⁺. This was accompanied by a decrease in the R_in in the region between 60 and 80 mV and an even greater decrease in R_in at potentials more negative than the reversal potential (i.e., inward rectification). If maximal concentrations of DAMGO and baclofen were coadministered, the resultant V_max for the two drugs combined was not different from that determined at the end of a dose-response experiment with a single agonist (table 1). Taken together, this confirms and extends our previous findings that in ARC neurons µ-opioid and GABA_B agonists activate an inwardly rectifying K⁺ conductance and that µ-opioid and GABA_B receptor systems share a common pool of these K⁺ channels (Loose and Kelly, 1990; Loose et al., 1991; Kelly et al., 1992).

View larger version (38K): [in this window] [in a new window]   Fig. 1.   A, cumulative dose-response effects of the µ-opioid receptor agonist DAMGO on an ARC neuron from a placebo-treated animal. The resting Vm of this cell was 48 mV. Successively increasing doses of DAMGO (20, 50 and 300 nM), perfused until a new steady-state Vm was reached (4-7 min), hyperpolarized the cell 8, 14 and 22 mV, respectively. The highest dose tested (1 µM) had no additional effect. The break in the recording represents the time at which the second V/I relationship was established (see "Methods"). B, cumulative dose-response effects of the GABAB receptor agonist baclofen on an ARC neuron from a placebo-treated animal. The resting Vm of this cell was 61 mV. Successively increasing doses of baclofen (1, 2, 5 and 20 µM), perfused until a new steady-state Vm was reached (4-7 min), hyperpolarized this cell 8, 13, 17 and 19 mV, respectively. The downward deflections represent voltage deflections used to monitor Rin, which are caused by intermittent negative current pulses (not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   V/I plots obtained just before and immediately after DAMGO (A) or baclofen (B) application. Depolarizing and hyperpolarizing current pulses (0.2 Hz, 1-sec duration) were given, and the resultant Vm response recorded, just before the dose-response experiment () and immediately after application of the maximal concentration of drug (). Both plots cross at approximately 95 mV, which closely approximates the Nernst equilibrium potential for K+. The pronounced decrease in Rin at potentials less than 100 mV indicates that the underlying K+ current is inwardly rectifying.

View this table: [in this window] [in a new window]   TABLE 1 Nonadditivity of the maximal responses produced by DAMGO and baclofen Dose-response relationships for DAMGO or baclofen were generated in ARC neurons in hypothalamic slices prepared from placebo-treated animals. The Vm was returned to its original resting state via the injection of positive current, and maximal concentrations of DAMGO (200-500 nM) and baclofen (30-40 µM) were coadministered. Values represent the mean ± S.E.M. of the Vmax elicited by maximal concentrations of either DAMGO (n = 11) or baclofen (n = 3) alone or in combination (DAMGO + baclofen; n = 14).

As shown in table 2, chronic morphine treatment produced appreciable serum concentrations of morphine and its conjugated metabolites M6G and M3G. This level of morphine was shown previously to elicit the behavioral expression of tolerance (Goldstein and Schulz, 1973), as well as down-regulation of µ-opioid receptors (Zhang et al., 1996). In contrast, serum levels of these compounds were not detectable in placebo-treated controls. Chronic morphine exposure produced a 3.2-fold rightward shift of the dose-response curve for DAMGO (EC₅₀, 42.1 ± 4.0 nM vs. 134.7 ± 13.4 nM) (fig. 3) but not for baclofen (EC₅₀, 2.2 ± 0.7 µM vs. 2.1 ± 0.7 µM) (fig. 4). This increase in the DAMGO EC₅₀ for all ARC neurons tested was nearly identical to that observed in a subset of TH-positive ARC neurons (fig. 5), an example of which is shown in figure 6, A and B. Likewise, the lack of effect of chronic morphine treatment on the baclofen EC₅₀ was reflected in a subset of TH-positive ARC neurons (fig. 7), an example of which is shown in figure 6, C and D. Chronic morphine treatment did not affect the resting V_m, R_in, membrane time constant or firing rate of ARC neurons (table 3). In addition, the opiate affected neither the V_max nor the agonist-induced decrease in R_in for DAMGO or baclofen (table 4). These results show that in ARC neurons, including A₁₂ dopamine neurons, chronic morphine treatment decreases the potency of the selective µ-opioid agonist DAMGO but decreases neither the potency nor the efficacy of the selective GABA_B agonist baclofen.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Composite cumulative dose-response curves for DAMGO with placebo or morphine treatment. Cells from placebo-treated () or morphine-treated () animals were perfused with successively higher concentrations of DAMGO (10, 20, 50, 100, 200, 300 and 1000 nM, 4-7 min/dose; n = 15 for placebo, n = 48 for morphine). Curves were produced from logistic equations fitted by computer to the data points. The mean DAMGO EC50 values were 42.1 ± 4.0 nM for placebo and 134.7 ± 13.4 nM for morphine (P < .05; Mann-Whitney test).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Composite cumulative dose-response curves for baclofen with placebo or morphine treatment. Cells from placebo-treated () or morphine-treated () animals were perfused with successively higher concentrations of baclofen (0.1, 0.3, 1, 2, 3, 5, 10, 20, 30, 40 and 100 µM, 4-7 min/dose; n = 10 for placebo, n = 6 for morphine). Symbols and vertical lines are means ± 2 S.E.M. of hyperpolarizations produced by a given dose, which have been normalized to agonist-induced Vmax. Curves were produced from logistic equations fitted by computer to the data points. The mean baclofen EC50 values were 2.2 ± 0.7 µM for placebo and 2.1 ± 0.7 µM for morphine.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Effect of chronic morphine treatment on DAMGO potency in ARC neurons, including TH-positive ARC neurons. Columns and vertical lines represent means and 1 S.E.M. of the DAMGO EC50 in either placebo-treated (n = 15), morphine-treated (n = 48) or morphine-treated, TH-positive (TH+; n = 4) ARC neurons. *, Values from morphine-treated and morphine-treated, TH-positive ARC neurons that are significantly different (Kruskal-Wallis analysis of variance/rank sum test; P < .05) from placebo-treated controls.

View larger version (81K): [in this window] [in a new window]   Fig. 6.   Examples of double-labeled TH neurons from morphine-treated animals that exhibit tolerance to µ-opioid receptor agonists but not cross-tolerance to GABAB receptor agonists. A, photomicrograph of the biocytin-streptavidin-fluorescein isothiocyanate labeling of a large pyramidal neuron. This neuron exhibited tolerance to the membrane-hyperpolarizing effect of the µ-opioid receptor agonist DAMGO, as manifested by reduced potency (EC50 = 111 nM). B, photomicrograph of the TH immunoreactivity in the soma and fiber of the same cell as in A, as visualized with CY-3. C, photomicrograph of the biocytin-streptavidin-fluorescein isothiocyanate labeling of a fusiform neuron. This neuron was tested with the GABAB receptor agonist baclofen, and the baclofen EC50 was comparable to that seen in placebo-treated controls. D, photomicrograph of the TH immunoreactivity in the soma and proximal fiber of the same cell as in C, as visualized with CY-3. Scale bar, 25 µm (for all photomicrographs).

View larger version (24K): [in this window] [in a new window]   Fig. 7.   Lack of effect of chronic morphine treatment on baclofen potency in ARC neurons, including TH-positive ARC neurons. Columns and vertical lines represent means and 1 S.E.M. of the baclofen EC50 in either placebo-treated (n = 10), morphine-treated (n = 6) or morphine-treated, TH-positive (TH+; n = 3) ARC neurons.

View this table: [in this window] [in a new window]   TABLE 3 Lack of effect of chronic morphine treatment on either the membrane properties or the firing rate of ARC neurons Values represent the mean ± S.E.M. of the resting Vm, Rin,  and firing rate of ARC neurons (n = 19-49) obtained from placebo- and morphine-treated animals.

View this table: [in this window] [in a new window]   TABLE 4 Lack of effect of chronic morphine treatment on either the Vmax or the Rin elicited by DAMGO or baclofen Cells from placebo-treated or morphine-treated animals were perfused with successively higher concentrations of either DAMGO (10, 20, 50, 100, 200, 300 and 1000 nM, 4-7 min/dose; n = 15 for placebo, n = 48 for morphine) or baclofen (0.1, 0.3, 1, 2, 3, 5, 10, 20, 30, 40 and 100 µM, 4-7 min/dose; n = 10 for placebo, n = 6 for morphine). Values from the second column represent the mean ± S.E.M. of the Vmax produced by either DAMGO or baclofen in placebo- and morphine-treated animals. The agonist-induced decrease in Rin (Rin) was calculated by subtracting the ratio of the slopes of the post- and predrug V/I plots between 60 and 80 mV from unity and multiplying the resultant number by 100. Values from the third column represent the mean ± S.E.M. of the Rin obtained from 4 to 18 cells.

	Discussion

Top Abstract Introduction Methods Results Discussion References

The results of the present study demonstrate for the first time that chronic morphine treatment produces tolerance to the direct effects of µ-opioid receptor agonists on central dopamine neurons. This conclusion is based on the observation that chronic morphine exposure decreases the potency of the µ-opioid receptor agonist DAMGO in a subset of TH-positive ARC neurons, recorded in the presence of TTX to block neuronal firing and synaptic input. The A₁₂ dopamine neurons are the principal inhibitory modulators of prolactin secretion and are different from other central dopamine (e.g., nigrostriatal or mesolimbic) neurons in that they are neurosecretory in nature. These results also indicate that in ARC neurons, including A₁₂ dopamine neurons, chronic morphine treatment does not produce cross-tolerance to GABA_B receptor agonists. This observation is based on the observation that chronic morphine exposure decreases neither the potency nor the efficacy of the GABA_B receptor agonist baclofen.

Baclofen and DAMGO act at different receptor sites that converge at the same effector (Loose et al., 1991; Kelly et al., 1992). Chronic exposure to sedative/hypnotic drugs of abuse such as ethanol, barbiturates and benzodiazepines, all of which act at different modulatory sites on the GABA_A receptor/Cl ionophore, has been postulated to produce cross-tolerance (Cox, 1990). However, little cross-tolerance has been observed between agonists that activate µ- and -opioid receptors, which converge at the same effector (Cox, 1990). In the present study, chronic morphine treatment elicited a 3.2-fold reduction in the potency of DAMGO to hyperpolarize ARC neurons and was without effect on the efficacy of the hyperpolarizing response. A comparable reduction in the potency of DAMGO to produce outward K⁺ current as a result of chronic morphine treatment has been observed in locus coeruleus neurons (Christie et al., 1987). Those authors also showed that chronic morphine treatment resulted in a decreased maximal outward current produced by the partial opioid agonist normorphine but not by the full agonist DAMGO. As described in the present study, chronic morphine treatment failed to elicit a reduction in either the potency or the efficacy of baclofen to hyperpolarize ARC neurons. Similarly, the maximal outward current produced by activation of alpha-2 adrenergic receptors, which couple to the same inwardly rectifying K⁺ channel as do µ-opioid receptors in locus coeruleus neurons, was unchanged by chronic morphine treatment (Christie et al., 1987). In addition, chronic morphine exposure in vitro attenuates µ-opioid but not alpha-2 receptor-mediated inhibition of N-type Ca⁺⁺ current in SH-SY5Y cells (Kennedy and Henderson, 1991).

Receptor desensitization as a result of prolonged acute exposure to µ-opioid agonists has been proposed as the first step in the development of tolerance associated with chronic opiate exposure (Harris and Williams, 1991). This phenomenon is manifested by decreased amplitude and/or decay of the agonist response (Harris and Williams, 1991; Kovoor et al., 1995). In oocyte expression systems, heterologous desensitization has been shown between µ-opioid and 5-hydroxytryptamine_1A receptors coupling to the same inwardly rectifying K⁺ channels (Kovoor et al., 1995). In the present study, chronic morphine exposure did not affect the efficacy of DAMGO or baclofen, and there were no signs of desensitization with the range of agonist doses used (10 nM to 1 µM for DAMGO and 0.1-100 µM for baclofen). Furthermore, in rat locus coeruleus, desensitization caused by a 5-min application of 30 µM Met-enkephalin was primarily homologous, with very little heterologous desensitization being observed between µ-opioid and alpha-2 adrenergic agonists (Harris and Williams, 1991). The results from the present study and other studies (Christie et al., 1987; Cox, 1990) indicate that cross-tolerance does not develop between µ-opioid receptors and other G protein-coupled receptors modulating a common effector. This suggests that, in the mammalian central nervous system, homologous desensitization may underlie the development of tolerance and the selective uncoupling of the µ-opioid receptor from its effector as a result of chronic morphine exposure.

The development of tolerance resulting from chronic morphine exposure is associated with increased activity of adenylate cyclase, increased intracellular levels of cAMP and increased activity of cAMP-dependent protein kinase (Rasmussen et al., 1990; Wang et al., 1994), suggesting that protein phosphorylation may be involved in the development of morphine tolerance. Whereas protein kinase inhibitors have been reported to decrease DAMGO binding to µ-opioid receptors in cerebellar membranes (Aloyo, 1995), they also have been shown to block antinociceptive morphine tolerance (Narita et al., 1994). It is possible that, in ARC neurons, chronic morphine treatment enhances protein phosphorylation, thereby uncoupling µ-opioid receptors from their inwardly rectifying K⁺ channels. Indeed, cAMP-dependent protein kinase treatment decreases the functional coupling of purified µ-opioid receptors reconstituted with purified G_i protein, as measured by agonist-stimulated, low-K_m GTPase activity (Harada et al., 1989, 1990). Studies are underway to determine whether protein kinase inhibition prevents the development of tolerance to µ-opioid agonists in these neurons.

Despite the importance of opioid receptor-mediated regulation of the hypothalamus, few studies have examined the effects of chronic morphine exposure on processes regulated by this brain region. Tolerance to the inhibitory effects of morphine on oxytocin release from magnocellular neurons in the supraoptic nucleus, and from parvocellular neurons in the paraventricular nucleus, has been reported (Russell et al., 1992). Acute exposure to µ-opioid agonists hyperpolarizes dopamine neurons in the ARC in vitro (Loose et al., 1990) and increases dopamine turnover in incertohypothalamic neurons in vivo (Tian et al., 1992). On the other hand, chronic morphine treatment does not alter the turnover of dopamine in neurons terminating in the rostral hypothalamus or the mediobasal hypothalamus, neuronal systems thought to be important in regulating sexual behavior (Clark et al., 1988). On the basis of these studies, it appears that tolerance develops to the effects of opioids on hypothalamic dopaminergic neuronal systems because of chronic morphine treatment. Indeed, in the present study we show that ARC A₁₂ dopamine neurons develop tolerance to the hyperpolarizing effects of µ-opioids, manifested by a reduction in µ-opioid receptor agonist potency. This would account for the observed tolerance to the prolactin-releasing effects of morphine with chronic exposure (Deyo et al., 1980). We also showed previously that ARC -endorphin neurons are even more sensitive to the effects of chronic morphine treatment, exhibiting cellular tolerance in the form of prominent reductions in µ-opioid receptor agonist potency and efficacy (Zhang et al., 1996). The ARC contains a heterogeneous mixture of neuronal phenotypes (Zoli et al., 1993), which ultimately affect a diverse array of behavioral processes by regulating the secretion of a multitude of pituitary hormones and by interacting with other brain nuclei. The present study clearly identifies an ARC phenotype that exhibits tolerance but not cross-tolerance due to chronic morphine administration. In conclusion, the results presented in this study reveal that, in ARC neurons, cross-tolerance does not develop between µ-opioid and GABA_B receptor agonists as a result of chronic morphine exposure, suggesting that coupling to a common effector is not a sufficient criterion for establishing cross-tolerance between receptor systems.