Psychopharmacology (2000) 152:208–215 DOI 10.1007/s002130000507
O R I G I N A L I N V E S T I G AT I O N
Sheryl S. Moy · Darin J. Knapp · Gary E. Duncan George R. Breese
Enhanced ultrasonic vocalization and Fos protein expression following ethanol withdrawal: effects of flumazenil Received: 29 February 2000 / Accepted: 2 May 2000 / Published online: 9 August 2000 © Springer-Verlag 2000
Abstract Rationale: Administration of flumazenil, a benzodiazepine (BZD) antagonist, has therapeutic efficacy against some anxiogenic effects of ethanol withdrawal. This observation has led to the suggestion that anxiety associated with ethanol withdrawal is related to release in brain of an endogenous BZD inverse agonist. Objective: The present studies further tested this hypothesis by assessing the effect of flumazenil on withdrawal-induced changes in a behavioral task and on the expression of the neuronal protein, Fos. Methods: Male Sprague-Dawley rats were withdrawn from a chronic ethanol regimen and tested, with or without flumazenil pretreatment, for either ultrasonic vocalization in response to air puff or for the induction of Fos protein-like immunoreactivity (Fos-LI) in brain. In addition, flumazenil effects on Fos-LI were measured in a group of animals treated with the BZD inverse agonist DMCM (0.75 and 1.0 mg/kg). Results: Flumazenil (5.0 mg/kg) significantly reduced the number of ultrasonic vocalizations observed following withdrawal from chronic ethanol. In contrast, flumazenil (5.0 mg/kg), given either 14 h before withdrawal from chronic ethanol, or during hours 3 and 5 following withdrawal, did not attenuate the effects of withdrawal on Fos-LI. Subsequent testing with DMCM confirmed that a benzodiazepine inverse agonist can induce Fos-LI in most of the same brain regions as observed following ethanol withdrawal, and that this change in Fos protein can be attenuated by pretreatment with flumazenil (5.0 mg/kg). Conclusions: Overall, these results demonstrate that specific behavioral indices of anxiety, but not measures of Fos-LI, support the contribution of an endogenous BZD inverse agonist in the ethanol withdrawal syndrome. Key words Anxiety · Benzodiazepine receptor · DMCM · Ethanol withdrawal · Flumazenil · Ultrasonic vocalization S.S. Moy (✉) · D.J. Knapp · G.E. Duncan · G.R. Breese Skipper Bowles Center for Alcohol Studies, Department of Psychiatry and UNC Neuroscience Center, CB# 7178, UNC School of Medicine, Chapel Hill, NC 27599, USA e-mail:
[email protected] Fax: +1-919-966-5679
Introduction Flumazenil (Ro 15-1788) is most frequently categorized as a benzodiazepine (BZD) antagonist (Polc et al. 1981; Bonetti et al. 1982). A substantial number of studies have demonstrated that flumazenil can attenuate the effects of anxiolytic compounds such as chlordiazepoxide and diazepam (e.g., Bonetti et al. 1981, 1982), and can precipitate a withdrawal syndrome in animals maintained on a chronic diazepam regimen (Martinez et al. 1992; Martin et al. 1993). However, paradoxically, flumazenil can also alleviate the anxiety and other symptoms observed during withdrawal from chronic ethanol use (Gerra et al. 1991). In rat models of the ethanol withdrawal syndrome, flumazenil can reverse anxiogenic withdrawal effects such as inhibition during a social interaction test (File et al. 1989, 1992) and reduced open arm exploration on an elevated plus maze (Moy et al. 1997). In addition, Uzbay and colleagues (1995) found that flumazenil could prevent the agitation and stereotyped behavior induced by withdrawal from long-term ethanol exposure in rats. One possible explanation for these anxiolytic effects is that flumazenil can block the action of an endogenous BZD inverse agonist that is released in brain during withdrawal from chronic ethanol or diazepam, and during other aversive situations (Baldwin and File 1988; File et al. 1989; Maier et al. 1995). Diazepam binding inhibitor, or DBI, is a neuropeptide that has been isolated from brain and has anxiogenic activity when centrally administered in rats (Guidotti et al. 1983). The anxietylike response to DBI can be blocked by flumazenil (Guidotti et al. 1983). Levels of DBI in brain increase during stress (Ferrarese et al. 1991a, 1991b) and following interruption of chronic ethanol consumption (Alho et al. 1987; Katsura et al. 1995, 1998) and chronic diazepam treatment (Miyata et al. 1987), suggesting that the release of endogenous DBI might underlie at least some of the anxiety-like symptoms following withdrawal from chronic dosing regimens. In addition to effects on behavior, ethanol withdrawal can also induce activation of the immediate-early proto-
209
oncogene c-fos in brain (Dave et al. 1990; Morgan et al. 1992; Matsumoto et al. 1993). Increased levels of Fos protein, measured as Fos-like immunoreactivity (Fos-LI), can be detected in specific brain regions of ethanolwithdrawn rats (Matsumoto et al. 1993; Knapp et al. 1998) and have been used to identify activated neural circuitry in brain (Morgan et al. 1987; Sagar et al. 1988). The following studies were designed to test further the possible role of an endogenous BZD inverse agonist in the anxiogenic effects of ethanol withdrawal. First, the ability of flumazenil to attenuate withdrawal-induced changes in a behavioral measure, ultrasonic vocalization in response to an aversive air puff, was determined. Previous work in our laboratory had demonstrated that flumazenil, given 5 min before testing, could prevent the withdrawal-induced reduction in open arm exploration on the elevated plus maze (Moy et al. 1997). While the maze task has been well established as a sensitive measure for changes in GABAergic function (e.g., Pellow and File 1986), there is evidence that startle-induced ultrasonic vocalization is more dependent upon serotonergic mediation. For example, an acute dose of a 5-hydroxytryptamine1A (5-HT1A) receptor agonist, gepirone, but not diazepam, can attenuate the enhanced ultrasonic vocalization seen after withdrawal from repeated treatment with a relatively high dose of diazepam (Vivian et al. 1994). Therefore, in the first experiment, flumazenil was tested in the ultrasound task to assess whether withdrawal-induced changes in vocalization might be based on a similar underlying mechanism as indicated using the plus maze task, or if the two measures might reflect different types of anxiety, driven by different systems in brain. A second experiment tested whether flumazenil would have a significant impact against the Fos-LI observed following ethanol withdrawal. In this case, flumazenil was administered either 14 h before the withdrawal period or twice during the withdrawal period, in order to investigate both short-term and long-term drug effects. Previous work has shown that flumazenil, administered several hours before ethanol withdrawal, can reduce signs of ethanol dependence and tolerance in mice, an effect attributed to the ability of flumazenil to directly modify or “reset” elements of the benzodiazepine/GABA receptor complex (Buck et al. 1991). A final experiment examined, first, whether a BZD inverse agonist, DMCM (methyl 6,7-dimethoxy-4-ethyl-β-carboline-3-carboxylate), would induce Fos-LI in a way comparable to ethanol withdrawal and, second, whether pretreatment with flumazenil would block the induction of this immediate early gene by DMCM.
Materials and methods Subjects Male Sprague-Dawley rats were purchased from Charles River Laboratories (Raleigh, N.C., USA). The animals were experimentally naive at the beginning of each study and were tested only
once, at the age of 2–4 months. Subjects were kept in a temperature-controlled environment with a 12-h light-dark cycle. Rats maintained on lab chow (Purina Rat Chow) were given free access to food and water and were group housed, while rats maintained on liquid diet were given free access to water and were housed individually. All procedures involved in this work were in strict compliance with the policies on animal welfare of the National Institutes of Health and the University of North Carolina (stated in the “Guide for the Care and Use of Laboratory Animals,” Institute of Laboratory Animal Resources, National Research Council, 1996 edition). Drugs Flumazenil (Ro 15-1788, Hoffman-La Roche Pharmaceutical Co., Basel, Switzerland) was ultrasonically dispersed in a vehicle of 0.5% carboxymethylcellulose (CMC). A dose of 5.0 mg/kg, with a concentration of 2.5 mg/ml, was used for each injection. DMCM (Research Biochemicals International, Natick, Mass., USA) was dissolved in sterile water with 1–2 drops 1.0 N HCl. Two doses were used: a low dose (0.75 mg/kg) and a high dose (1.0 mg/kg). Drug concentration was 0.375 mg/ml and 0.5 mg/ml, respectively. Both flumazenil and DMCM were administered by IP injection. Chronic ethanol regimen The chronic administration of ethanol involved a nutritionally balanced liquid diet, as previously described (Frye et al. 1981). At the beginning of the liquid diet procedure for each experiment, rats were given 2 days of control liquid diet to allow adaptation to the novel mixture. On day 3, one group of the subjects was given diet consisting of 5% w/v ethanol, while the remaining subjects continued on the control diet in which the ethanol content was replaced by an equicaloric amount of dextrose. After 2 days at 5%, the ethanol concentration was increased to 6% for 3 days, and then to 7% for the remaining 9 days of the diet regimen. In all, ethanol liquid diet was given for 14 days. Using a paired-feeding approach, the volume of diet given to each control animal was matched to the amount of diet consumed by an ethanol subject, a procedure which allows adequate and comparable weight gain in both ethanol and control groups (Frye et al. 1981). Testing procedures Effect of flumazenil on withdrawal-induced ultrasonic vocalization Subjects were randomly divided into two groups, with one group receiving liquid diet containing ethanol (n=24) and one group receiving control liquid diet (n=16). One subject was removed from each group for failure to consume adequate amounts of the liquid diet. After 14 days of treatment, withdrawal was induced by replacing the ethanol liquid diet with control liquid diet. Six to 8 h following withdrawal, the animals were tested for ultrasonic vocalizations. Five minutes before the test, an injection of flumazenil (5.0 mg/kg) was given to one group of subjects (11 ethanol diet animals and eight control diet subjects), while the CMC vehicle was administered to another set (12 ethanol diet animals and seven control diet subjects). For the ultrasound procedure (Knapp and Pohorecky 1995), each animal was placed in a standard wire mesh-bottomed cage for a 5-min habituation period. Next, an air puff, dispensed at a force of 9 PSI by a picospritzer, was aimed at the animal’s neck and head from approximately 15–20 cm above the rat. After administration of the first air puff, the subject was monitored for ultrasonic vocalization during the subsequent 15 s period by an experimenter equipped with a set of headphones connected to a Mini-2 Bat Detector (Ultrasound Advice, UK). If no vocalizations were detected, another air puff was administered, followed by an-
210 other 15 s listening period. Rats were monitored for the duration of their vocalization period, and a maximum of 15 air puffs were administered for each subject. Measures were taken of number of air puffs to elicit the first vocalization, number of vocalizations during the first three minutes following the first air puff, total number of vocalizations, and duration of vocalization. Flumazenil effects on withdrawal-induced Fos-LI Subjects were randomly divided into two groups, with one group receiving liquid diet containing ethanol and one group receiving control liquid diet. After 14 days of treatment, withdrawal was induced by replacing the ethanol liquid diet with control liquid diet. One group of animals, the pre-WD (pre-withdrawal) group, received a single dose of flumazenil (5.0 mg/kg) 14 h before the ethanol was withdrawn (n=4 control diet rats and six ethanol diet rats). A second group of animals, the post-WD (post-withdrawal) group, received flumazenil (5.0 mg/kg) at hours 3 and 5 following withdrawal (n=4 control diet rats and six ethanol diet rats). A control group of animals was given CMC vehicle injections at each time point (n=4 for both control and ethanol diet rats). CMC vehicle was also administered to the pre-WD group at hours 3 and 5 following withdrawal, and to the post-WD group at 14 h before withdrawal. Thus, each subject in the study received a total of three injections. Eight hours following withdrawal of the ethanol diet, each subject was deeply anesthetized with sodium pentobarbital (80–100 mg/kg) and prepared for the Fos-LI procedure. Flumazenil effects on DMCM-induced Fos-LI To habituate the subjects to handling, the animals were weighed and gently held on three occasions during the week before testing. On the test day, each subject received either flumazenil (5.0 mg/kg) or the CMC vehicle. Five minutes following the first injection, each subject received either the low dose (0.75 mg/kg) or high dose (1.0 mg/kg) of DMCM, or the sterile water vehicle. The study consisted of the following experimental groups (and number of rats): flumazenil/low DMCM dose (n=5), flumazenil/high DMCM dose (n=3), flumazenil/vehicle (n=3), vehicle/low DMCM dose (n=6), vehicle/high DMCM dose (n=3), and vehicle/vehicle (n=4). One subject in the vehicle/high DMCM dose group died from a seizure shortly after the drug was administered. In addition, two other subjects were removed from the study when both were observed to have seizures and, subsequently, had extremely high levels of Fos-LI. These changes resulted in only one subject in the vehicle pretreatment/high DMCM dose group and five animals in the vehicle/low DMCM dose group. Animals were returned to their home cages following each injection and observed for any seizure-like activity. Results from these behavioral observations indicated that, within 10 min of the DMCM injection, two of the three rats in the high dose DMCM group and two of the six animals in the low dose DMCM group had tonic-clonic seizures. Four of the remaining five subjects given DMCM alone showed occasional head and tail jerks. At 140 min after the first injection, each subject was deeply anesthetized with pentobarbital (80–100 mg/kg) and prepared for the Fos-LI procedure. Measurement of Fos-LI Anesthetized animals were perfused transcardially with 0.1 M phosphate-buffered saline (PBS), followed by 4% paraformaldehyde in 0.1 M phosphate-buffer solution. Brains were then placed in paraformaldehyde and kept in cold storage until cut with a vibratome. Sections (50 µm thick) were taken for selected brain regions, and a standard avidin-biotin/horseradish peroxidase method was used for the Fos-LI assay, as previously described (Duncan et al. 1996). Briefly, Fos-LI was detected using an affinity-purified sheep polyclonal antibody (Vector Laboratories, Inc., Burlingame,
Calif., USA) raised to a synthetic peptide of Fos (Cambridge Research Biochemicals, Inc., Wilmington, Del., USA; current source is Biogenesis, Brentwood, N.H., USA), and visualized with 3,3’-diamino-benzidene tetrahydrochloride (DAB; Polysciences, Inc., Warrington, Pa., USA). Cells demonstrating Fos-LI were counted at a total magnification of ×100–200 for selected brain regions. Statistics Data are presented as group means±SEM. Measures taken from each experiment were first analyzed using a two-way analysis of variance (ANOVA), with the variables of group (DMCM/vehicle or ethanol diet/control diet) and drug pretreatment (flumazenil versus vehicle). Significant effects were further tested using oneway ANOVAs. Fisher’s protected least significant difference (PLSD) tests were conducted on group means only when a significant F value was found. Significance was set at P
