|Home | About | Journals | Submit | Contact Us | Français|
Administration of μ-opioid receptor (MOR) agonists is known to produce adaptive changes within noradrenergic neurons of the rat locus coeruleus (LC). Alterations in the subcellular distribution of MOR have been shown to occur in the LC in response to full agonists and endogenous peptides; however there is considerable debate in the literature whether trafficking of MOR occurs following chronic exposure to the partial-agonist morphine. In the present study, we examined adaptations in MOR following chronic opioid exposure using immunofluorescence and electron microscopy (EM), with receptor internalization as a functional endpoint. MOR trafficking in LC neurons, identified by tyrosine hydroxylase immunoreactivity, was characterized in rats that were morphine dependent in response to an antagonist, naltrexone, at a dose known to precipitate withdrawal. Following chronic morphine exposure, a subtle redistribution of MOR immunoreactivity from the membrane to the cytosol was detected within dendrites of LC neurons. Interestingly, an acute injection of naltrexone in rats exposed to chronic morphine produced a robust internalization of MOR, while administration of naltrexone failed to do so in naïve animals. The findings provide anatomical evidence for modified regulation of MOR trafficking following chronic morphine treatment in brain noradrenergic neurons, and suggest adaptations in the MOR signaling pathways that regulate internalization as a consequence of chronic treatment and precipitation of withdrawal. Mechanisms underlying this effect might include differential MOR regulation in the LC, or downstream effects of withdrawal-induced enkephalin release from afferents to the LC.
G-protein coupled receptors (GPCRs) comprise a superfamily of transmembrane receptors and are involved in virtually all physiological processes. These receptors are extremely important drug targets and effective sites for modulation of cell signaling. Opioid receptors, which are the site of action for many analgesic compounds, are members of this GPCR superfamily. Opioid GPCRs typically activate a variety of downstream pathways through the GTP-binding Gi/Go regulatory proteins, resulting in adenylyl cyclase inhibition, decreased cAMP levels, and modulation of Ca2+ and K+ conductance (Childers et al., 1992; Dhawan et al., 1996; Yu et al., 1997; Borgland et al., 2003). Widely used analgesic agents such as morphine and endogenous opioids such as enkephalin and endorphin act selectively through the μ-opioid receptor (MOR) (Raynor et al., 1994; Raynor et al., 1995). Chronic activation of MOR leads to the development of tolerance and dependence, and opioids with a high abuse liability have been shown to function largely through this receptor (Raynor et al., 1994; Connor et al., 2004; Contet et al., 2004; Johnson et al., 2005). Side effects and potential for dependence represent a major concern in the use of these compounds as therapeutic agents. For these reasons, the adaptations of MOR to chronic agonist exposure are central to the understanding of opiate tolerance, addiction, and withdrawal.
MOR undergoes many of the regulatory processes common to GPCRs. One phenomenon, in particular, is that of agonist-induced endocytosis, which has been extensively studied in MOR in the dorsal root ganglia (Walwyn et al., 2006), spinal cord (Narita et al., 2006), and locus coeruleus (Van Bockstaele and Commons, 2001). Receptor endocytosis and desensitization is thought to play a modulatory role in opiate tolerance and withdrawal (Whistler et al., 1999; von Zastrow et al., 2003; von Zastrow, 2004; Arttamangkul et al., 2006). In a ligand-dependent manner, receptor internalization and trafficking affect signal transduction (Gintzler and Chakrabarti, 2000; Chakrabarti et al., 2001; Alvarez et al., 2002; Koch et al., 2005; Han et al., 2006). Although full agonists are believed to cause agonist-induced endocytosis of MOR throughout the CNS, considerable debate surrounds the trafficking of MOR in response to morphine treatment. In the past, research has shown that acute exposure to morphine, a partial agonist, causes desensitization of MOR but not endocytosis in vitro and in vivo (Keith et al., 1996; Sternini et al., 1996; Keith et al., 1998; Alvarez et al., 2002). However, a more recent report demonstrates the redistribution of MOR in striatal neurons in response to acute morphine (Haberstock-Debic, et al., 2005), and in dendrites of catecholaminergic neurons in response to chronic morphine (Drake et al., 2005; Haberstock-Debic et al., 2005).
In the locus coeruleus (LC), an area containing a high-density of MORs, chronic administration of opioid agonists are known to produce adaptive changes (Nestler, 1993; Van Bockstaele et al., 2001; Xu et al., 2004; Gintzler and Chakrabarti, 2006). The LC is a homogeneous region of noradrenergic neurons that serves as the primary source of norepinephrine to the rest of the brain, regulating attention, arousal, and pain (Waterhouse et al., 1983; Nakazato, 1987; Aston-Jones et al., 1996; Aston-Jones et al., 1999; Berridge and Waterhouse, 2003). In the LC, MOR responds to ligand activation by desensitization and/or internalization in a manner unique to the specific ligand on board. Full agonists such as etorphine or endogenous opioid peptides cause MOR internalization following receptor stimulation, while acute exposure to partial agonists such as morphine are generally not thought to cause internalization (Van Bockstaele and Commons, 2001; Alvarez et al., 2002). However, recent studies focusing on acute and chronic morphine-induced MOR trafficking again calls into question the response of MOR to morphine in the LC (Drake et al., 2005; Haberstock-Debic et al., 2005). Particularly, it is not well understood how the localization of MOR is changed by chronic opioid exposure and upon initiation of withdrawal.
The current studies are focused on changes in MOR trafficking patterns following chronic exposure to morphine and in response to the rapid precipitation of withdrawal. Our studies using immunofluorescence and immunogold electron microscopy suggest that naltrexone, an opioid antagonist, may cause MOR internalization in the rat LC following chronic morphine administration. Additionally, chronic morphine appears to cause a very subtle, though consistent shift of MOR from the plasmalemma to the cytoplasm of dendritic compartments. These observations provide critical anatomical evidence of the trafficking patterns within noradrenergic neurons that may shape the development of tolerance and withdrawal, and help to explain how MOR trafficking is influenced by chronic opioid exposure and precipitation of withdrawal.
Adult male Sprague Dawley rats were used in this study (250–300g, n=15; Harlan, Indianapolis, IN). Animals were housed 3 to a cage and kept on a 12-hour light-dark cycle, with free access to food and water. All procedures were performed in accordance with NIH and Institutional Animal Care and Use Committee guidelines. Efforts were taken to minimize postoperative pain and discomfort, and to limit the number of animals used.
Rats were subcutaneously implanted with two slow-release morphine pellets for a 5-day period (75g x2, National Institute on Drug Abuse). This period of time has been shown to be sufficient to yield physical dependence to morphine (Koob et al., 1992). Matched controls received placebo pellets. At the end of the 5-day exposure period, rats were given one of four types of an acute drug treatment.
Five different treatment groups (n=3 per group) were analyzed in this study. One group of morphine-treated rats received a subcutaneous (s.c) injection of the MOR antagonist naltrexone [morphine pellet-acute naltrexone injection, s.c.] at a dose sufficient to produce withdrawal (100 mg/kg) (Shaw-Lutchman et al., 2002, (Rasmussen et al., 1990). A second group that had received morphine treatment was administered a saline vehicle injection [morphine pellet-acute saline injection, s.c.] at the end of the 5-day period. The third group of animals that had received chronic morphine treatment was given an acute injection of etorphine (0.05 mg/kg) [morphine pellet-acute etorphine injection s.c.]. Etorphine given to naïve rats was previously shown to cause prominent MOR internalization in the LC (Van Bockstaele and Commons, 2001). One group of matched controls received placebo pellets, followed by an injection of saline vehicle [placebo pellet-acute saline injection, s.c.]. A second group of control rats received placebo pellets, followed by an acute injection of naltrexone (100 mg/kg) [placebo pellet-acute naltrexone injection, s.c.].
Thirty minutes following the acute injection of naltrexone, etorphine or vehicle, rats were anesthetized and sacrificed for tissue preparation. For this, rats were deeply anesthetized with sodium pentobarbital and perfused transcardially through the ascending aorta with 75ml with 3.8% acrolein + 2% paraformaldehyde in 0.1M phosphate buffer (PB, pH 7.4), followed by 400 ml of 2% paraformaldehyde in 0.1M PB. Animals from all treatment groups were perfused on the same day using the same batch of perfusion fixative. Immediately following perfusion/fixation, brains were removed and returned to the 2% paraformaldehyde + acrolein fixative for 30 minutes. Brains were cut in 40μm coronal sections through the locus coeruleus using a Vibratome, at −10.04 from bregma, 1.10mm medial-lateral, and 7.4 from the top of the skull based on the rat brain atlas of Paxinos and Watson (Paxinos, 1986).
Immunohistochemical labeling of MOR and tyrosine hydroxylase (TH) for electron microscopy was performed as previously described by Van Bockstaele, et al. (Van Bockstaele et al., 1996). Staining was done in parallel for control and experimental animals, using the same reagents. Briefly, tissues were incubated in 1% sodium borohydride in 0.1M phosphate buffer (PB) to remove reactive aldehydes, and then blocked in 0.5% BSA in 0.1M Tris-buffered saline (TBS). Tissue sections were incubated overnight in primary antibody in 0.1% BSA in 0.1M TBS. Single-labeled tissues were incubated at a concentration of 1:10,000 in rabbit anti-MOR antiserum directed to amino acids 384–398 of the rat MOR (Immunostar, Hudson, WI) (for specificity and characterization, see (Arvidsson et al., 1995) (Cheng et al., 1996). Dual-labeled tissues were incubated in a cocktail of rabbit anti-MOR (Immunostar, 1:10,000) plus mouse anti-tyrosine hydroxylase (Immunostar, 1:1000) (for specificity and characterization, see (Van Bockstaele and Pickel, 1993).
Detection of MOR and TH for immunofluorescence utilized secondary antibodies that were FITC conjugated donkey-anti mouse IgG for TH detection, and TRITC conjugated goat-anti rabbit IgG for MOR detection at 1:200 dilutions (Jackson Immunoresearch, West Grove, PA). Tissue sections underwent serial dehydration, were mounted on slides, and coverslipped using DPX (Aldrich, St. Louis, MO). Slides were then viewed using a fluorescent confocal microscope, and images were prepared using the Zeiss LSM Image Browser and Adobe Photoshop.
The secondary antibodies used for electron microscopy were biotinylated anti-mouse for TH labeling at 1:400 (Vector Laboratories, Burlingame, CA), and 1nm gold particle conjugated goat-anti rabbit IgG at 1:50 for MOR (GE/Amersham, Piscataway, NJ). Immunoperoxidase labeling of TH was performed using an avidin-biotin complex solution (Vector Laboratories). The peroxidase reaction product was then visualized using 0.02% DAB plus 10 μl of 30% H2O2 in 0.1M TBS. Electron dense labeling of MOR was detected via silver intensification of immunogold MOR particles using a silver enhancement kit (GE/Amersham). Tissues were prepared for visualization under the electron microscope as previously described, with osmification, serial dehydration, flat-embedding, and tissue sectioning at 74 nm on an ultramicrotome (Van Bockstaele and Commons, 2001). Control sections for each experiment were run in parallel in which one of the primary antisera was omitted, with the rest of the protocol identical, and no immunoreactivity was observed. Sections were collected on copper mesh grids and examined using an electron microscope (Morgani, Fei Company, Hillsboro, OR). Digital images were viewed and captured using the AMT advantage HR HR-B CCD camera system (Advance Microscopy Techniques Corp., Danvers, MA). Immunofluorescence and electron micrograph images were prepared using Adobe Photoshop to adjust the brightness and contrast.
To confirm dual-labeling of MOR and TH within LC sections for immunofluorescence studies, Z-stack analysis was performed. For better clarity of receptor distribution in the dense LC, all quantification of the subcellular distribution of MOR immunoreactivity took place in ultra-thin sections using electron microscopy. Receptor quantification and analysis of internalization at the EM level as previously reported (Van Bockstaele and Commons, 2001; Reyes et al., 2007). Data analysis was performed by an experimenter blinded to the treatment condition to prevent any bias. Several measures were taken to prevent inclusion of background labeling in the quantification. First, the extent of background labeling was assessed by examining blood vessels and myelin (which should lack MOR labeling), and these structures showed minimal immunogold labeling. Additionally, to account for potential background, only dendrites with at least two or more gold particles were considered immunoreactive and used for quantification. Cellular elements were isolated and classified based on Peters et al. (Peters, 1991), and post-synaptic effects of drug treatment were assessed through quantification of MOR immunoreactivity on dendrites in the LC.
For each treatment group, data from a series of 3 animals were analyzed, and dendritic profiles were sampled from at least 10 copper grids of ultrathin tissue sections per animal. Localization of the LC was verified by the presence of a high density of TH immunoreactivity. Each grid was systematically scanned and all immunoreactive profiles with intact morphology were captured digitally. A total of 125 dendritic profiles per animal in each treatment group were included in the data analysis. Dendrites with a maximal cross-sectional diameter of greater than 0.5μm and less than 6μm fit the criteria for analysis. To account for possible differences in the cellular dynamics between small and large dendrites, dendritic profiles were classified in to groups of small transverse (0.5μm – 2.0μm) and large transverse (2.01μm – 6.0μm). Extremely small, longitudinal and irregularly shaped profiles were excluded due to possibly higher perimeter/surface ratios and risk of biasing the silver grain counts towards the membrane. Any profiles containing large, irregularly shaped silver grains of more than 0.25μm were excluded from the analysis.
To calculate the distribution ratio, a ratio of cytoplasmic to total immunogold particles per dendritic profile was computed, and an average of all ratios for each animal was taken. One-way ANOVA was used to determine if there were within-group differences in the distribution ratio among animals receiving identical drug treatment protocols. There were no significant differences between animals within any of the treatment groups (see results for values). Since no differences were detected between animals within the same drug treatment group, data from these animals were pooled. In order to address the possibility of MOR shifting within the dendritic cytoplasm, the amount of TH-labeled dendrites that lack immunoreactivity for MOR was compared across treatment groups by calculating the percentages of dendrites singly-labeled for TH versus the percentage dually-labeled for TH and MOR. Immunogold particles were tallied as plasmalemmal if they were clearly in contact with an intact plasma membrane. Because the MOR antibody used in this study targets the C-terminus of the receptor which is oriented towards the cytoplasmic space, only plasmalemmal MOR immunogold particles oriented towards the cytoplasmic compartment of dendrites were included in the quantification of a profile’s distribution ratio. Immunogold MOR was considered cytoplasmic only if there was a clearly discernable gap between the particle and membrane.
Using the mean distribution ratio and standard deviation for each animal, one-way analysis of variance (ANOVA) was used to confirm that there were no significant differences between animals in the same drug treatment group. Once this was confirmed, an average distribution ratio from the three animals in each treatment group was calculated. ANOVA was used for within group comparisons to determine the presence of statistically significant shifts in the distribution of MOR immunoreactivity. Tukey’s post-hoc tests were used for between group comparisons.
Five different treatment groups (n=3 per group) were analyzed in this study: 1) placebo pellet-acute saline injection, 2) morphine pellet-acute saline injection, 3) placebo pellet-acute naltrexone injection, 4) morphine pellet-acute naltrexone injection, and 5) morphine pellet-acute etorphine injection. These groups were compared to previous findings in a group of naïve rats given acute etorphine, where it was shown to cause MOR internalization (Van Bockstaele and Commons, 2001). There were no signs of irritation or infection following pellet implantation, and mobility was unaffected. Shortly after administration of naltrexone to morphine-dependent animals, withdrawal signs became evident. Behavior observed during the 30-minute withdrawal period included wet dog shakes, panting, vocalization, and defensive burying. Chronic morphine animals given an acute dose of etorphine displayed hypolocomotor effects and appeared sedated, behaving similar to naïve animals given acute etorphine in a previous study. No notable behavioral changes occurred upon acute administration of naltrexone to placebo-treated animals, or in chronic morphine animals treated with acute saline. Placebo-acute saline group behavior was also unremarkable.
Dual label immunofluorescence in naïve rats reveals a high density of MOR immunoreactivity in the LC region, with a considerable degree of colocalization between labeling for MOR (red) and TH (green) (Fig. 1). The presence of TH/MOR labeling coexistence confirms these MOR-containing neurons to be positive for the rate-limiting enzyme for norepinephrine synthesis, TH, and therefore specific to the noradrenergic LC. Analysis of fluorescence labeling of MOR using confocal microscopy revealed a greater distribution of granular, punctate immunoreactivity in naltrexone-treated animals (Fig. 1, Column D) as compared to placebo-saline treated controls (Fig. 1, Column A) and chronic morphine-saline treated animals (Fig. 1, Column B). This punctate pattern of MOR immunoreactivity is comparable to that of chronic morphine-acute etorphine treated animals (Fig. 1, Column C). This increase in punctuate MOR immunoreactivity may be indicative of receptor redistribution and localization of receptors to endocytic vesicles, as was observed in previous studies (Van Bockstaele and Commons, 2001). These qualitative findings were therefore analyzed semi-quantitatively using electron microscopy to more closely examine potential changes in MOR immunoreactivity. In a higher magnification view of LC neurons of chronic morphine-acute naltrexone treated rats, bright spots of immunoreactivity are clustered in the cytoplasm, and may represent endocytic vesicles containing MOR labeling (Fig. 2, A and B).
In agreement with the qualitative findings from the immunofluorescence microscopy data, electron microscopy results verified prominent MOR labeling in LC perikarya and dendrites, with occasional labeling in axon terminals and axons. Spurious background was assessed by analyzing the presence of gold-silver labeling in myelin sheaths (a structure that should not exhibit immunoreactivity for MOR) in 50 randomly selected micrographs (for each treatment group) taken at 7000x magnification. Within each set of 50 micrographs, a range of 2–5 single immunogold-silver particles were identified with the myelin sheath indicating minimal spurious background labeling. However, to account for potential non-specific background labeling, the criterion for dendritic immunoreactivity was set at a minimum of 2 immunogold particles per profile. MOR immunogold was detected in dendritic profiles dual-labeled for TH with immunoperoxidase in all treatment groups, confirming localization to the LC region (Fig. 3, E–H). There were no significant differences in the localization of MOR immunoreactivity between animals within each treatment group. The mean internalization percentages (reported as means ± SEM, p-values) were as follows: morphine pellet-acute naltrexone injection group (71% ± 1.4%, p= 0.48), morphine pellet-acute saline injection group (37% ± 1%, p= 0.74), placebo pellet-acute saline injection group (30% ± 0.7%, p= 0.80), and placebo pellet-acute naltrexone injection group (29% ± 1.2%, p= 0.53). Since no differences were detected between animals within the same drug treatment group, data from these animals were pooled.
Importantly, there was minimal variability in the overall density of MOR1 immunogold between treatment groups. For each treatment group, the sum of total immunogold particles from all dendrites analyzed was computed. Each group included 375 dendritic profiles, and groups were composed of comparable ratios of small and large dendrites. The mean immunogold particle sum for the four treatment groups was 1206 ± 101 grains, which equates to a deviation of ≤ 8% above or below the mean. We therefore conclude that the observed changes in MOR immunoreactivity reported were not likely due to changes in MOR protein expression. In all treatment groups, the majority of TH-labeled dendrites were dually-labeled for MOR as well. The percentage of total TH-positive dendrites lacking MOR labeling was 19%, 22%, 18% and 17% in the morphine pellet-acute naltrexone injection group, morphine pellet-acute saline injection group, placebo pellet-acute saline injection group, and placebo pellet-acute naltrexone injection group, respectively, suggesting a small amount of variability in this measure between experimental groups.
A total of 375 dendritic profiles per treatment group (n=3) were quantified for each of four treatment groups. Because the dynamics of receptor trafficking may vary depending upon the location within the cell, potential differences in MOR distribution between proximal (large) and distal (small) dendrites were examined. The profiles from each treatment group were analyzed and divided into groups based upon size. Two-tailed t-tests revealed that no statistically significant differences (p ≥ 0.05) existed in the ratio of cytoplasmic to total immunogold particles between proximal and distal dendrites in any of the treatment groups. The mean cytoplasmic percentages (reported as means ± variance) were as follows: morphine pellet-acute naltrexone injection proximal group (77 ± 6%) versus distal group (78 ± 9%), with p= 0.95, morphine pellet-acute saline injection proximal group (33 ± 11%) versus distal group (37 ± 5%), with p= 0.56,, and placebo pellet-acute naltrexone injection proximal group (34 ± 12%) versus distal group (34 ± 12%), with p= 0.96. In the placebo pellet-acute saline injection group, there was a trend towards a difference in the distribution of MOR immunogold between the proximal and distal proximal dendrite groups, with proximal (41 ± 10%) versus distal group (28 ± 8%), with p= 0.06. Although this did not reach the threshold for statistical significance, there may be a tendency toward a proximal-distal difference in receptor distribution in the placebo-acute saline control group. However, in all groups where pharmacological manipulation took place, there was no difference between proximal and distal dendrites detected. Since no statistically significant differences were seen between proximal and distal distribution means, the data from categories were pooled. The patterns of receptor distribution appear to be similar in dendrites both proximal and distal to the soma, and when receptor internalization was observed, the phenomenon was present regardless of profile size. The trafficking patterns observed in this study appear to indicate an overall effect of treatment that spanned the length of the dendrites in LC neurons.
EM findings corroborate the evidence of MOR internalization in response to chronic morphine-acute naltrexone treatment seen in confocal micrographs. Distribution of MOR immunoreactivity was examined in LC sections singly labeled for MOR (Fig. 3, A–D), and in tissue sections with dual labeling for MOR and TH (Fig. 3, E–H). Under basal conditions, MOR immunogold preferentially associates with the plasma membrane of LC dendrites. A majority of the gold-silver MOR labeling in placebo-acute saline, placebo-acute naltrexone, and morphine-acute saline groups was associated with the plasma membrane (Fig. 3, A–C/E–G). However, a statistically significant shift of MOR immunogold particles from the plasmalemma to the cytoplasm did occur when comparing chronic morphine-acute saline animals to the placebo-acute saline group, with 37% versus 30% internalization of MOR, respectively (p< 0.05, Fig. 4, Table 1). A considerable shift of MOR particles from plasma membrane to cytoplasm occurred when withdrawal was induced via naltrexone in chronic morphine treated rats (Fig. 3, D/H). This withdrawal-induced shift brought MOR immunogold to 71% cytoplasmic, as compared to the basal values of 30% in placebo pellet-acute saline controls and 29% in placebo pellet-acute naltrexone controls (p< 0.001, Fig. 4, Table 1). Notably, this shift was specific to the induction of morphine withdrawal by naltrexone and not an effect due solely to acute naltrexone, because no significant differences mean cytoplasmic percentage (p= 0.84) existed between placebo pellet-acute saline (30% ± 0.7%) and placebo pellet-acute naltrexone animals (29% ± 1.5%).
Based on previous observations of receptor trafficking in the laboratory, redistribution of receptor immunoreactivity into punctate foci is often suggestive of a shift or redistribution of the receptor, as was seen with MOR internalization following etorphine administration to naive animals (Van Bockstaele and Commons, 2001). In this study, similar qualitative observations were made in LC neurons of rats chronically exposed to morphine and given naltrexone. The results support a unique phenomenon of receptor redistribution when withdrawal was induced in a noradrenergic system chronically exposed to morphine. This phenomenon did not appear, however, when naltrexone was given to naïve animals. In order to gain subcellular resolution of the distribution of MOR labeling in vivo, studies were performed at the ultrastructural level using electron microscopy. Semi-quantitative analysis of receptor internalization by electron microscopy supported the immunofluorescence data, and uncovered the more subtle phenomenon of MOR immunogold redistribution in response to chronic morphine administration. Qualitative MOR immunofluorescence findings show a MOR redistribution within the chronic morphine-acute etorphine treatment paradigm that will need to be elucidated further using a more quantitative electron microscopic approach and analysis. However, alterations in the pattern of MOR immunoreactivity (Fig. 1) suggest that chronic morphine treatment may not desensitize MOR to the effects of etorphine. Altogether, this study provides anatomical evidence of withdrawal-induced trafficking of MOR using various types of microscopic analysis, and offers further clarification to the debate surrounding chronic morphine-induced endocytosis of MOR.
MOR trafficking in the central nervous system is widely known as a region- and ligand-dependent process. While full agonists such as etorphine or opioid peptides efficiently activate the receptor and cause MOR endocytosis (Van Bockstaele and Commons, 2001), studies of MOR trafficking following morphine treatment have provided conflicting results. Until recently, many in the field have maintained the notion that acute morphine, while able to activate MOR and result in tolerance, lacks or has limited ability to cause endocytosis of the receptor, especially in comparison to typically high efficacy agonists like etorphine or DAMGO (Keith et al., 1996; Yu et al., 1997; Zaki et al., 2000; Van Bockstaele and Commons, 2001). In the past few years, several reports of morphine-induced trafficking of MOR have surfaced, with both acute and chronic administrations (Haberstock-Debic et al., 2003; Haberstock-Debic et al., 2005). In the current study, we have uncovered a modest, but consistent pattern of endocytosis of MOR labeling in response to chronic morphine exposure. It is possible that the release of endogenous opioids such as enkephalin in response to morphine administration could be contributing to the pattern of internalization seen following a chronic regimen by binding resensitized MOR (Dang and Williams, 2004). The present findings reveal a statistically significant degree of chronic morphine-induced internalization in the LC; this change represents a shift of 7% of the total receptor population from the plasmalemma to the cytoplasmic compartment (30% to 37%) in placebo controls versus chronic morphine. This 7% total shift equates to a loss of approximately 1/5 of the MOR that was present on the membrane. Although significant by ANOVA, further studies are required to determine whether this is a biologically relevant shift, and able to impact global signal transduction.
Unlike agonist-induced endocytosis, antagonist-induced endocytosis is a much rarer and less-understood process (Roth and Willins, 1999). Various reports have however demonstrated cases of antagonist-induced GPCR internalization, suggesting that activation of the receptor is not always necessary for the endocytic event. Antagonist-induced internalization has been documented in several other GPCRs: cholecystokinin receptor in the pituitary (Roettger et al., 1997), serotonin subtype 2A receptors in the prefrontal cortex (Willins et al., 1999), and the endothelin subtype A receptor in smooth muscle (Bhowmick et al., 1998). Similar to our paradoxical finding that both etorphine and naltrexone induced internalization, agonist and antagonist-induced internalization of GPCR A1 adenosine receptors has been demonstrated (Navarro et al., 1999). However, few studies have previously explored the probability of MOR internalization following an acute, high dose antagonist administration given to precipitate morphine withdrawal. When examining the present findings, it is necessary to consider how the pharmacological properties of naltrexone may contribute to this phenomenon of altered signaling pathways and endocytosis upon induction of withdrawal. In morphine naïve animals, evidence suggests that naltrexone functions as a neutral antagonist, but that following chronic agonist exposure, a switch to inverse agonism occurs, whereby naltrexone may suppress the constitutive, basal activity of MOR (Liu and Prather, 2001; Wang et al., 2001; Wang et al., 2007). Although inverse antagonism is often associated with receptor up-regulation and increased membrane presence, in select cases, it has been shown that inverse agonism can result in down-regulation and internalization (Breit et al., 2006; Dupre et al., 2007).
Notably, we cannot be certain whether the redistribution of MOR labeling found in these studies is the direct result of antagonist-induced internalization of MOR in the LC, or the result of system-wide changes surrounding the induction of withdrawal following chronic adaptation of MOR to morphine. It is not yet known whether other MOR antagonists will result in a similar pattern of receptor redistribution, or if these effects are specific to naltrexone-induced withdrawal. Additionally, it will be useful to determine if low-dose administration of this drug functions differently from high-dose naltrexone given to precipitate withdrawal. Along these lines, it is extremely important to note that naltrexone does not induce MOR internalization in placebo-treated animals; rather, this only occurred in morphine-dependent animals given the drug at a dose (Shaw-Lutchman et al., 2002) known to precipitate withdrawal. Moreover, although our analysis focused on the LC noradrenergic system, this is not out of disregard to other regions of the brain and how they may be impacted by adaptations of MOR to chronic morphine.
It is possible that glutamate and endogenous opioid co-release from afferents to the LC may underlie antagonist precipitated/withdrawal-linked endocytosis of MOR. Previous studies in the laboratory have shown that glutamate and enkephalin coexist in axon terminals that originate from the primary excitatory afferent to the LC, the nucleus paragigantocellularis (PGi) (Van Bockstaele et al., 2000; Barr and Van Bockstaele, 2005). . However, rapid ENK release and high levels in the synapse would likely be required to achieve MOR internalization in the presence of a dose of 100 mg/kg naltrexone. Excitatory input from the PGi plays a role in LC hyperactivation observed during opiate withdrawal. This nucleus is particularly sensitive to opiate withdrawal, demonstrating heightened activity in the form of increased transmission of excitatory amino acids, augmented c-Fos expression, greater firing rates, and altered cAMP-mediated signal transduction (Rasmussen and Aghajanian, 1989; Stornetta et al., 1993; Aghajanian et al., 1994; Nestler et al., 1994; Shaw-Lutchman et al., 2002; Han et al., 2006). Glutamate receptor antagonists modulate symptoms of morphine withdrawal and withdrawal-induced LC activation (Rasmussen and Vandergriff, 2003; Rasmussen et al., 2004), suggesting a potential role for enhanced glutamatergic signaling in the current findings. We outline several hypothetical mechanisms for MOR internalization during antagonist-induced opiate withdrawal in Figure 5. MOR internalization during naltrexone-induced withdrawal may be a response to a number of factors: changes in protein conformation and signaling in LC MORs following antagonist binding (Fig. 5, Panel B), endogenous opioid and glutamate release (Fig. 5, Panel C), or a combination thereof (Fig. 5, Panels B + C). The overall pattern of MOR trafficking observed during naltrexone precipitated withdrawal may be influenced by a variety of effects that result from the combined presence of naltrexone, glutamate and enkephalin in the synapse.
Future studies that examine MOR trafficking during precipitated withdrawal using alternative methods and brain systems will help to elucidate whether the observations in this manuscript are unique to the LC, or a more global phenomenon of MOR antagonism. Two opposing schools of thought surround the relationship between MOR endocytosis and the development of opiate tolerance (Koch et al., 2005; Walwyn et al., 2006), and it will also be important to elucidate how the present findings of receptor internalization contribute to the withdrawal behavioral phenotype. Further studies are needed to clarify the driving force behind opiate tolerance, but perhaps more importantly, it may be essential to understand the consequences of MOR endocytosis and the fate of the receptor once internalized. By exploring this aspect of MOR endocytosis, it may be possible to uncover new ways to regulate GPCR trafficking for the potential prevention or treatment of withdrawal symptoms. Overall, the present findings suggest that naltrexone-precipitated withdrawal and subsequent MOR internalization may be an informative new model system that could be used to gain understanding of the alternative signaling pathways, endogenous opioids, and other desensitizing regulatory elements that underlie opiate addiction and withdrawal.
This work was supported by NIH grants DA 09082, 15123, 15395, 023755, and EY 014798. The authors would like to acknowledge and thank Kathryn Commons, PhD for her critical review of this manuscript. We would also like to acknowledge Alex Libowitz for his assistance with data analysis.
Elisabeth Van Bockstaele; Grant number: DA 09082, 15123, & 15395
Jillian Scavone; Grant Number: DA 023755