European Journal of Neuroscience, Vol. 17, pp. 971±980, 2003
ß Federation of European Neuroscience Societies
Regulation of RGS proteins by chronic morphine in rat locus coeruleus Stephen J. Gold,1 Ming-Hu Han,1 Amy E. Herman,1 Yan G. Ni,2 Cindy M. Pudiak,1 George K. Aghajanian,4 Rong-Jian Liu,4 Bryan W. Potts,1 Susanne M. Mumby2 and Eric J. Nestler1,3 1
Department of Psychiatry, Department of Pharmacology and 3 Centre for Basic Neuroscience, The University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd, Dallas TX 75390 4 Department of Psychiatry, Yale University School of Medicine, New Haven, CT 06508, USA 2
Keywords: m-opioid receptor, opiate tolerance, RGS2, RGS4, ubiquitination, withdrawal
Abstract The present study explored a possible role for RGS (regulators of G protein signalling) proteins in the long term actions of morphine in the locus coeruleus (LC), a brainstem region implicated in opiate physical dependence and withdrawal. Morphine in¯uences LC neurons through activation of m-opioid receptors, which, being Gi/o-linked, would be expected to be modulated by RGS proteins. We focused on several RGS subtypes that are known to be expressed in this brain region. Levels of mRNAs encoding RGS2, -3, -4, -5, -7, 8 and -11 are unchanged following chronic morphine, but RGS2 and 4 mRNA levels are increased 2±3-fold 6 h following precipitation of opiate withdrawal. The increases in RGS2 and 4 mRNA peak after 6 h of withdrawal and return to control levels by 24 h. Immunoblot analysis of RGS4 revealed a striking divergence between mRNA and protein responses in LC: protein levels are elevated twofold following chronic morphine and decrease to control values by 6 h of withdrawal. In contrast, levels of RGS7 and 11 proteins, the only other subtypes for which antibodies are available, were not altered by these treatments. Intracellular application of wild-type RGS4, but not a GTPase accelerating-de®cient mutant of RGS4, into LC neurons diminished electrophysiological responses to morphine. The observed subtype- and time-speci®c regulation of RGS4 protein and mRNA, and the diminished morphine-induced currents in the presence of elevated RGS4 protein levels, indicate that morphine induction of RGS4 could contribute to aspects of opiate tolerance and dependence displayed by LC neurons.
Introduction The locus coeruleus (LC), the principal noradrenergic nucleus in the central nervous system, has served as a useful model system in which to study chronic actions of opiates in the nervous system. The LC, which regulates alertness and vigilance under normal conditions, comprises a relatively homogeneous neuronal population that densely expresses the m-opioid receptor. Electrophysiological and pharmacological studies have implicated the LC as one mediator of opiate physical dependence and the expression of opiate withdrawal (reviewed in Koob et al., 1992; Nestler & Aghajanian, 1997; Williams et al., 2001). Indeed, following chronic morphine treatment, LC neurons show tolerance to morphine as well as cellular correlates of dependence and withdrawal (Aghajanian, 1978; Christie et al., 1987; Kogan et al., 1992). At the biochemical level in the LC, chronic exposure to morphine increases levels of the G protein subunits Gai/o, adenylyl cyclase types 1 and 8, protein kinase A (PKA), and the transcription factor CREB (Duman et al., 1988; Nestler & Tallman, 1988; Nestler et al., 1989; Matsuoka et al., 1994; Widnell et al., 1994; Lane-Ladd et al., 1997; Ivanov & Aston-Jones, 2001; Shaw-Lutchman et al., 2002). Upregulation of the cAMP pathway (namely, increased expression of adenylyl Correspondence: Dr Stephen J. Gold, as above. E-mail: [email protected]
Received 22 September 2002, revised 12 December 2002, accepted 3 January 2003 doi:10.1046/j.1460-9568.2003.02529.x
cyclase, PKA, and CREB) has been related directly to the dependence and withdrawal exhibited by these neurons and to the resulting behavioural withdrawal syndrome (Kogan et al., 1992; Lane-Ladd et al., 1997; Punch et al., 1997). In contrast, the functional signi®cance of changes in Gai/o levels has remained obscure. Interestingly, even though Gai/o levels are increased in the LC following chronic morphine administration, both basal and m-opioid receptor-stimulated G-protein activity are reduced, with no concomitant change in the percent of opiate stimulation of G-protein activity (Selley et al., 1997). This suggests that m-opioid receptors might not themselves be desensitized by chronic morphine in the LC, but rather that signalling might be decreased downstream of the receptor. As these data were generated, genetic studies in nonmammalian model organisms (Dohlman et al., 1996; Koelle & Horvitz, 1996) and mammalian cells (De Vries et al., 1995) identi®ed the RGS (regulators of G protein signalling) family of proteins as negative modulators of G protein function. This activity is mediated via the RGS proteins' GTPase-activating properties (GAP; Berman & Gilman, 1998). More recent studies have suggested that RGS proteins also in¯uence G protein signalling through more complex interactions (Ross & Wilkie, 2000). The negative regulation of G proteins by RGS proteins makes the RGS family a good candidate for mediating the decreased m-opioid receptor signalling in the LC (and perhaps elsewhere) after chronic morphine exposure. Indeed, RGS proteins can modulate m-opioid
972 S. J. Gold et al. receptor signalling in both heterologous expression systems (Chuang et al., 1998; Potenza et al., 1999; Rahman et al., 1999) and in vivo (GarzoÂn et al., 2001). Here, we studied the regulation of RGS proteins in the LC by chronic morphine and naltrexone-precipitated withdrawal. The results suggest novel mechanisms by which opiateinduced adaptations in RGS proteins could contribute to aspects of opiate tolerance and dependence.
Materials and methods Animal treatments All experiments were approved by the Institutional Animal Care and Research Advisory Committee (IACRAC) of The University of Texas Southwestern Medical Centre at Dallas. Adult male Sprague±Dawley rats (175±300 g; Charles River, Kingston, NY, USA) were used in all the studies described and were allowed to habituate to the vivarium for at least 6 days before the beginning of treatments. Except where indicated all reagents were purchased from Sigma (St. Louis, MO, USA). For the in situ hybridization-based survey of RGS subtype mRNA levels following sham surgery (sham), chronic morphine (mor), naltrexone-precipitated withdrawal (wdwl), and naltrexone alone (nalt), six male Sprague±Dawley rats (200 g, Charles River) per group were lightly anaesthetized with iso¯uorane and received either sham surgery or subcutaneous implantation of morphine pellets (75 mg, National Institute on Drug Abuse, Bethesda, MD, USA) for ®ve consecutive days between 15.00 h and 17.00 h. On day 6, rats received subcutaneous injections of either saline or naltrexone hydrochloride (100 mg/kg), were killed by decapitation 6 h after the injection, and their brains removed from the skull, blocked, rapidly frozen on dry ice and processed later for in situ hybridization. For Western blot analysis of RGS4 following chronic morphine or naltrexone-precipitated withdrawal, rats were killed as described, except that the LC was dissected rapidly (see Western blot analysis below and ®gure legends for further details). To examine the time course of changes in RGS mRNA levels in LC following naltrexoneprecipitated withdrawal, rats (n 5 per time point) received either daily sham surgery or morphine pellet implantation for ®ve consecutive days. The last two pellets were ensheathed in ®ne nylon netting tied off with suture, to facilitate their later removal, as described (Rasmussen et al., 1990). Thirty minutes after the lights were turned on, on day 6, rats were lightly anaesthetized and received either sham surgery or surgical removal of the last two nylon-ensheathed morphine pellets. Two hours later, rats received either saline or naltrexone (100 mg/kg, s.c.). Rats were killed at 2, 6, 24 or 48 h after naltrexone injection. Animals killed at 24 and 48 h received a second and third (48 h group only) naltrexone injection (100 mg/kg, s.c.) at 6 and 24 h after the initial injection to maintain maximal withdrawal through the duration of the experiment (Rasmussen et al., 1990). In situ hybridization Slide-mounted, fresh-frozen in situ hybridization was used to study regulation of RGS mRNA levels by morphine treatment and withdrawal (see Seroogy & Herman, 1997). For dual-labelling analysis of RGS4 mRNA and tyrosine hydroxylase (TH)-like immunoreactiivty in LC neurons, free-¯oating in situ hybridization for RGS4 mRNA was followed by free-¯oating immunohistochemistry for TH using a published procedure (Seroogy & Gall, 1993). For fresh-frozen ISH, rats were decapitated, and the brains removed from the skull, rapidly frozen on dry ice, and stored at 80 8C. Coronal, fresh frozen sections were cut at 14 mm in a cryostat, thaw-mounted onto Superfrost Plus (Fisher Scienti®c, Pittsburgh, PA, USA) glass slides, and stored at 80 8C until use. Ten series of sections representing planes 9.3±10.3
(LC) and 7.04±8.3 mm (occipital cortex/dorsal raphe) posterior to bregma were cut such that each series had a representation every 140 mm (Paxinos & Watson, 1986). For free-¯oating in situ hybridization, rats were anaesthetized with chloral hydrate, exsanguinated via transcardial perfusion with 0.9% saline followed by 250 mL 4% paraformaldehyde in 0.1 M sodium phosphate buffer (PPB). Brains were post®xed overnight in PPB, cryoprotected overnight in PPB containing 20% sucrose, and four series of 30 mm coronal sections were cut on a freezing sledge microtome into PPB-representing planes 9.3±10.3 mm posterior to bregma (Paxinos & Watson, 1986); sections were stored at 4 8C until use 48 h later. Pretreatment for the freshfrozen and free-¯oating protocols were identical (G...