Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The GABA_A receptor chloride ionophore complex contains different recognition sites at which benzodiazepines (BZs), barbiturates, and neuroactive steroids can modulate GABA-mediated chloride flux (for review, see Mehta and Ticku, 1999). Drugs facilitating GABA-mediated chloride flux (e.g., positive GABA_A modulators or agonists) at BZ and barbiturate sites are used in medicine to elicit sedation, anxiolysis, muscle relaxation, and anticonvulsant effects (for review, see Woods et al., 1992 ); neuroactive steroid site positive GABA_A modulators are being considered for clinical use (Gasior et al., 1999). Drugs inhibiting GABA-mediated chloride flux (e.g., negative GABA_A modulators or inverse agonists) elicit anxiety-like behaviors and convulsions. Drugs binding to the complex without altering GABA-mediated chloride flux (e.g., neutral GABA_A modulators or antagonists) presumably do not directly elicit behavioral effects by acting at the receptor complex.

Chronic treatment with BZs such as diazepam can lead to tolerance, perhaps via allosteric uncoupling of BZ sites from GABA_A receptors or GABA_A receptor-mediated chloride channels (Hu and Ticku, 1994). Chronic BZ treatment produces other changes in vitro such as allosteric uncoupling of BZ sites from barbiturate or neuroactive steroid sites (i.e., homologous uncoupling) and uncoupling of barbiturate sites from GABA_A receptors (i.e., heterologous uncoupling; Hu and Ticku, 1994; Friedman et al., 1996). It is not clear whether heterologous uncoupling confers cross-tolerance from BZ to barbiturate or neuroactive steroid site positive GABA_A modulators. Another consequence of chronic BZ treatment is dependence as indexed by withdrawal signs emerging upon discontinuation of treatment (for review, see Woods et al., 1992). One consequence of BZ dependence is an increased sensitivity to BZ site neutral modulators such as flumazenil, an effect that is most likely due to the precipitation of withdrawal (Lukas and Griffiths, 1982; Takada et al., 1989; Sannerud et al., 1991; Gerak and France, 1997). Negative modulators acting at BZ sites also antagonize the behavioral effects of BZ site positive modulators (e.g., Wettstein, 1989; Wettstein et al., 1993; Lelas et al., 2000; McMahon and France, 2001). However, it is not clear whether negative modulators precipitate withdrawal in the same way that neutral modulators precipitate withdrawal (e.g., Ongini et al., 1985; Martin et al., 1995). Moreover, it is not clear how sensitivity to a negative modulator would change during chronic BZ treatment because in BZ-dependent animals, the behavioral effects of neutral modulators are qualitatively similar to the effects of negative modulators in untreated animals.

The present study examined whether chronic diazepam treatment confers cross-tolerance to positive modulators acting at different sites on the GABA_A receptor complex. This study also examined whether chronic diazepam treatment increases sensitivity to BZ site negative GABA_A modulators in a manner similar to that observed for flumazenil. Rhesus monkeys responding under a multiple fixed ratio (FR/FR) schedule of food presentation and stimulus-shock termination (SST) received various GABA_A modulators before, during, and after a period of daily diazepam treatment (5.6 mg/kg/day, p.o.). This multiple schedule was chosen because it is sensitive to the withdrawal-precipitating effects of flumazenil and because it has been used to reveal potentially important differences among GABA_A modulators (Ator, 1979; Gerak and France, 1997). BZ site negative modulators that vary in efficacy were studied to test whether efficacy differences correlate with changes in sensitivity that occur during chronic diazepam treatment, as has been shown for BZ site positive modulators (Bronson, 1993; Cohen and Sanger, 1994). The negative modulators were the low efficacy BZ Ro 15-4513, the intermediate efficacy -carboline -CCE, and the higher efficacy -carbolines -CCM and DMCM (Braestrup et al., 1982; Sieghart et al., 1987). The barbiturate pentobarbital and the neuroactive steroid pregnanolone were chosen for study because they are prototypic positive modulators acting at their respective sites.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects. One adult female and three adult male rhesus monkeys (Macaca mulatta) were housed individually under a 14:10-h light/dark schedule, maintained at 95% free-feeding weight (range 5.5-8.5 kg) with primate chow (High Protein Monkey Diet; Harlan Teklad, Madison, WI), fresh fruit, and peanuts, and provided water in the home cage. Monkeys had received BZ ligands acutely in a previous study (McMahon and France, 2001). The animals were maintained in accordance with the Institutional Animal Care and Use Committee, The University of Texas Health Science Center at San Antonio, and Guidelines of the Committee on Care and Use of Laboratory Animal Resources, National Research Council (Department of Health, Education and Welfare, publication no. (National Institutes of Health) 85-23, revised 1996).

Apparatus. During experimental sessions, monkeys were seated in chairs (model R001; Primate Products, Miami, FL) that provided restraint at the neck and placed in ventilated sound-attenuating chambers equipped with a response lever, lights, and a food cup to which pellets could be delivered from a dispenser. Feet were placed in shoes containing brass electrodes to which a brief (250 ms) electric stimulus could be delivered through an a.c. generator. An interface (MedAssociates, St. Albans, VT) connected the chambers to a computer that controlled and recorded experimental events.

Procedure. Monkeys responded under a multiple (FR10/FR10) schedule of food presentation and SST. Experimental sessions consisted of two to eight discrete 15-min cycles comprising a 10-min time-out period, during which lights were extinguished and responses had no programmed consequence, followed by the food presentation and SST response periods. Beginning of a 2-min food component was signaled by illumination of a green light above the response lever; 10 lever responses (FR10) resulted in the delivery of a 300-mg banana-flavored pellet (Bio-Serv, Frenchtown, NJ). The green light was extinguished after 2 min or the delivery of 10 food pellets, and for the latter case the remainder of the 2-min food component was a time-out period. The food component was followed by a 0.9-min time-out after which illumination of a red light signaled the scheduled delivery of an electric stimulus every 10 s. Ten lever responses (FR10) extinguished the red light, prevented the electric stimulus, and initiated a 20-s time-out period, after which the red light was again illuminated. A cycle ended after 5 min (including a 2-min food component, a 0.9-min time-out, and a 2.1-min SST component) or after four electric stimulus presentations, whichever occurred first.

Training and Testing. For training sessions, saline or "sham" injections were given during the first minute of the 10-min time-out period (e.g., first minute of the cycle) with the number of cycles varying nonsystematically across days. Training was conducted until stable rates of responding were established during both components, defined as 10 consecutive days with response rates for both components within ±20% of the mean rate for those days. Test sessions were conducted twice weekly so long as response rates for both components of the training session that immediately preceded a test were within ±20% of the mean rate for the 10 previous training sessions; otherwise, testing was postponed until this criterion was satisfied. During the first cycle of a test session, 0.3 ml of the appropriate vehicle was administered, and on subsequent cycles, increasing doses of a test compound were administered so that the cumulative dose increased by 0.25 or 0.5 log unit per cycle. The number of cycles for a test was determined by the number of cycles required to complete the dose-effect curve (i.e., from an ineffective dose to a dose that resulted in fewer than 10 responses during a food component or in four stimulus presentations during an SST component).

Drugs. The vehicle for oral administration of diazepam was fruit punch combined with suspending agent K (Bio-Serv) in a concentration of 1 g of suspending agent per liter of fruit punch. Tablets containing 10 mg of diazepam (Zenith Laboratories, Inc., Northvale, NJ) were dissolved in vehicle, mixed in a blender, and administered using a 12-gauge drinking needle attached to a 60-cc syringe. To obtain a dose of 5.6 mg/kg diazepam, a standard concentration of diazepam was given in a volume adjusted to individual body weights. The diazepam mixture was prepared immediately before administration.

Data Analyses. For individual monkeys, rates of responding for each session were calculated separately for each component of the multiple schedule by averaging rates of responding for all cycles within a training session. Control response rate was defined as the mean rate of the 10 training sessions immediately preceding the first day of daily diazepam treatment. Rates of responding during a test cycle were expressed as a percentage of the control response rate. Because some drugs did not decrease responding to below 75% of control in all monkeys, group averaged data were used to calculate the dose of a compound and 95% confidence limit (CL) required to decrease mean response rate to 75% of control (ED₇₅) under each component of the multiple schedule (Tallarida and Murray, 1987). ED₇₅ values from test sessions were considered to be significantly different from control when they were not within the control 95% CL. A drug that did not decrease responding to below 75% of control before or after chronic diazepam treatment was considered to have produced a significant effect if an ED₇₅ could be calculated for the same doses of that drug during chronic diazepam treatment. Daily response rate for the 72 days of chronic diazepam treatment and the 6 days after discontinuation of chronic treatment comprises the average of five saline cycles or the value from the first cycle before administration of a test compound during chronic treatment.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of GABA_A Modulators before Chronic Diazepam Treatment. Control response rates (±S.E.M.) for individual monkeys were 1.46 ± 0.04, 1.37 ± 0.06, 1.77 ± 0.11, and 1.76 ± 0.10 responses/s under the schedule of food presentation, and 2.26 ± 0.10, 2.87 ± 0.06, 2.05 ± 0.10, and 2.13 ± 0.10 responses/s under the schedule of SST.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Effects of triazolam on responding maintained under a multiple schedule of food presentation (left panel) or stimulus-shock termination (right panel) before (), during (), and after () chronic diazepam treatment. Abscissae, cumulative dose in mg/kg body weight. Points above V represent the average response rate after administration of vehicle. Ordinates, response rate expressed as a percentage of control rate ± S.E.M. for three monkeys.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Effects of pentobarbital on responding maintained under a multiple schedule of food presentation (left panel) or stimulus-shock termination (right panel) before () and during () chronic diazepam treatment. Open squares () represent effects of pentobarbital determined prior to administration of the daily dose (5.6 mg/kg, p.o.) of diazepam. Data represent mean responding for four monkeys. See Fig. 1 for other details.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effects of pregnanolone on responding maintained under a multiple schedule of food presentation (left panel) or stimulus-shock termination (right panel) before () and during () chronic diazepam treatment. Open squares () represent effects of pregnanolone determined prior to administration of daily dose (5.6 mg/kg, p.o.) of diazepam. Data represent mean responding for four monkeys. See Fig. 1 for other details.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Effects of flumazenil on the responding maintained under a multiple schedule of food presentation (left panel) or stimulus-shock termination (right panel) before, during, and after chronic diazepam treatment. Data represent mean responding for four monkeys up to a dose of 3.2 mg/kg, and two monkeys for larger doses. See Fig. 1 for other details. , before diazepam treatment; , after diazepam treatment; , during day 26; and , during day 72.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Effects of Ro 15-4513 on the responding maintained under a multiple schedule of food presentation (left panel) or stimulus-shock termination (right panel) before (), during (), and after () chronic diazepam treatment. Data represent mean responding for three monkeys. See Fig. 1 for other details.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Effects of -CCE on the responding maintained under a multiple schedule of food presentation (left panel) or stimulus-shock termination (right panel) before (), during (), and after () chronic diazepam treatment. Filled triangles () represent effects of -CCE determined after acute administration of diazepam (5.6 mg/kg, p.o.). Data represent mean responding for four monkeys. See Fig. 1 for other details.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Effects of -CCM on the responding maintained under a multiple schedule of food presentation (left panel) or stimulus-shock termination (right panel) before (), during (), and after () chronic diazepam treatment. Filled triangles () represent effects of -CCM determined after acute administration of diazepam (5.6 mg/kg, p.o.). Data represent mean responding for three monkeys. See Fig. 1 for other details.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Effects of DMCM on the responding maintained under a multiple schedule of food presentation (left panel) or stimulus-shock termination (right panel) before (), during (), and after () chronic diazepam treatment. Filled triangles () represent effects of DMCM determined after acute administration of diazepam (5.6 mg/kg, p.o.). Data represent mean responding for four monkeys. See Fig. 1 for other details.

Effects of GABA_A Modulators during Chronic Diazepam Treatment. Across the period of daily diazepam treatment, rate was unaffected or increased slightly in the food component and decreased in the SST component when determined 3 h after diazepam administration (Fig. 9). Rate-decreasing effects in the SST component were maximal on day 5 and evident throughout the 72 days of chronic treatment. Similar effects were obtained when test sessions were conducted before administration of the daily dose of diazepam on days 55, 57, 60, 69, and 70 of chronic treatment (data not shown).

View larger version (32K): [in this window] [in a new window]   Fig. 9.   The responding maintained under a multiple schedule of food presentation (top panel) or stimulus-shock termination (bottom panel) during daily diazepam treatment and after discontinuation of diazepam treatment. Abscissae, days of chronic diazepam treatment or days after discontinuation of chronic treatment; ordinates, response rate expressed as a percentage of control rate ± S.E.M. for four monkeys.

Effects of GABA_A Modulators after Discontinuation of Chronic Diazepam Treatment. After discontinuation of chronic diazepam treatment, responding in the food component was decreased on days 2 and 3 for one monkey, whereas responding in the SST component was decreased on days 2, 3, and 4 in the same monkey and for another monkey. However, the mean response rate for all four monkeys was only slightly decreased on these days (Fig. 9). Sensitivity to pentobarbital and pregnanolone (food component only) did not differ from sensitivity observed before chronic diazepam administration (Table 1). Sensitivity to pregnanolone in the SST component was significantly decreased as evidenced by a 1.5-fold shift to the right in the dose-effect curve compared with sensitivity observed before chronic diazepam treatment (Table 1). Sensitivity to triazolam, flumazenil, -CCM, and DMCM returned to values obtained before chronic diazepam administration (Figs. 1, 4, 7, and 8, respectively; Tables 1-3). The ED₇₅ from the Ro 15-4513 dose-effect curve in the food component was significantly smaller than that determined before chronic diazepam treatment, although remaining significantly larger than the ED₇₅ determined during chronic diazepam treatment. The ED₇₅ values for the Ro 15-4513 dose-effect curves in the SST component determined before and after discontinuation of chronic diazepam treatment were not different from each other (Fig. 5; Table 3). The ED₇₅ value from the -CCE dose-effect curve in the food component was significantly smaller (i.e., leftward shift) than the ED₇₅ from the -CCE dose-effect curve determined before chronic diazepam treatment (Fig. 6; Table 3).

Acute Interactions between Diazepam and Negative Modulators. Acute diazepam administration shifted the -CCE dose-effect curve in the SST and not the food component significantly to the right (Fig. 6; Table 4). Acute diazepam administration shifted the -CCM dose-effect curves in each schedule component significantly to the right (Fig. 7; Table 4). Acute diazepam administration shifted the DMCM dose-effect curve in the food component significantly to the right and did not alter the effects of DMCM in the SST component (Fig. 8; Table 4).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study examined the effects of chronic diazepam treatment on sensitivity to various different GABA_A modulators. Chronic BZ treatment can uncouple barbiturate sites from GABA_A-mediated chloride channels (Hu and Ticku, 1994), suggesting that chronic diazepam treatment might confer cross-tolerance to positive modulators acting at non-BZ sites on the GABA_A receptor complex (e.g., pentobarbital and pregnanolone). Chronic BZ treatment can also increase sensitivity to the neutral modulator flumazenil, an effect that might be due to the precipitation of withdrawal (Lukas and Griffiths, 1982; Takada et al., 1989; Sannerud et al., 1991; Gerak and France, 1997). Changes in sensitivity to flumazenil during chronic diazepam treatment were compared with changes in sensitivity to various negative modulators acting at BZ sites. BZ site negative modulators that vary in efficacy were studied to test whether efficacy differences correlate with changes in sensitivity that occur during chronic diazepam treatment, as has been shown for positive modulators acting at BZ sites (Bronson, 1993; Cohen and Sanger, 1994).

Positive GABA_A modulators decreased the responding under the multiple schedule with the following order of potency: triazolam > pregnanolone > pentobarbital. This result is consistent with a number of studies demonstrating that various positive GABA_A modulators decrease responding maintained under a variety of operant schedules (e.g., Ator 1979; Wettstein, 1989; Paronis and Bergman 1999; Vanover et al., 1999). Each positive modulator markedly decreased responding with similar potency in the two schedule components. Thus, responding maintained by the two reinforcers (food versus SST) appeared to be similarly affected by these positive GABA_A modulators. In contrast to positive modulators, the neutral modulator flumazenil did not alter the responding up to a dose much larger than the dose antagonizing diazepam and other BZs (e.g., Lelas et al., 2000). This result is consistent with the notion that flumazenil does not substantially modify GABA-mediated chloride flux at BZ sites on the receptor complex (e.g., Smith et al., 2001).

Unlike positive modulators, the potency of some negative modulators (e.g., -CCM and DMCM) to decrease responding was greater in the food component than in the SST component. The qualitatively different effects of positive and negative GABA_A modulators under the multiple schedule are consistent with previous reports on the schedule dependence of some behavioral effects of positive and negative GABA_A modulators (Ator, 1979; Corda et al., 1983; Paronis and Bergman 1999). Rate-decreasing effects among BZ site negative modulators appeared to be related to their apparent efficacy in vitro. For instance, Ro 15-4513 decreased responding to 80% of control at a dose (10 mg/kg) larger than the dose antagonizing diazepam and other BZs (Gerak et al., 1998; Lelas et al., 2000). Low efficacy (Mehta and Ticku, 1989) might be responsible for the failure of Ro 15-4513 to significantly modify schedule-controlled responding in nondependent animals in this and previous studies (Suzdak et al., 1986; for exception, see Britton et al., 1988). In contrast, relatively high efficacy might account for the greater potency of -CCE, -CCM, and DMCM to decrease the responding because rate-decreasing effects occurred at the same doses of -CCE and -CCM that antagonized BZs in previous studies (Lelas et al., 2000; McMahon and France, 2001). The greater potency of -CCM compared with -CCE and DMCM corroborates the results of other studies (Corda et al., 1983; Petersen, 1983; Gerak et al., 1998).

Daily diazepam treatment appeared to increase responding in the food component and to decrease responding in the SST component throughout the course of chronic diazepam treatment. These effects on responding were small compared with the effects produced by triazolam, pentobarbital, and pregnanolone at the largest doses studied. As shown previously, sensitivity to flumazenil increased during chronic diazepam treatment. Increased sensitivity to flumazenil during BZ treatment is thought to be due to the precipitation of withdrawal and has been used as a measure of BZ dependence (Lukas and Griffiths, 1982; Sannerud et al., 1991; Gerak and France, 1997). Rate-decreasing effects of flumazenil were evident after the second daily dose of diazepam (see also Spealman, 1985), suggesting that diazepam dependence begins to develop early in chronic treatment. The effects of flumazenil in the SST component were relatively stable throughout the course of chronic diazepam treatment, whereas sensitivity to flumazenil in the food component increased during the course of treatment. In contrast to previous studies indicating that tolerance develops to flumazenil-precipitated withdrawal during chronic diazepam treatment (Lamb and Griffiths, 1985), sensitivity to flumazenil was relatively stable over time in this and a previous study that used a comparable procedure (Gerak and France, 1998). Thus, to the extent that rate-decreasing effects of flumazenil during chronic BZ treatment represent BZ withdrawal and, therefore, indirectly BZ dependence, these results suggest that BZ dependence is relatively stable over a period of several weeks.

Chronic diazepam treatment elicited cross-tolerance to triazolam, a result consistent with other studies showing tolerance to the effects of BZs on schedule-controlled behavior (Takada et al.; 1989; Sannerud et al., 1993). This cross-tolerance could be related to uncoupling of BZ receptors from GABA_A receptors or GABA_A-mediated chloride channels (Hu and Ticku, 1994). Chronic diazepam treatment did not confer cross-tolerance to pentobarbital or pregnanolone, thereby corroborating the results of previous studies (e.g., Cesare and McKearney, 1980; Reddy and Rogawski, 2000). Collectively, these results suggest that heterologous uncoupling of sites on the GABA_A receptor complex induced by chronic BZ treatment might not lead to behavioral cross-tolerance among drugs acting at non-BZ sites (Hu and Ticku, 1994). Alternatively, the chronic diazepam treatment in the present study might not have been sufficient to promote heterologous uncoupling. Additional study will be required to determine whether allosteric uncoupling of sites on the GABA_A receptor complex is responsible for changes in behavioral sensitivity among drugs acting at these sites.

Chronic diazepam treatment differentially modified sensitivity to negative modulators in a manner that appeared to be related to efficacy. Like flumazenil, sensitivity to Ro 15-4513 was enhanced during chronic diazepam administration, an effect most likely due to antagonism of chronic diazepam at BZ receptors (i.e., precipitation of diazepam withdrawal). In contrast, sensitivity to DMCM decreased during chronic diazepam treatment, whereas sensitivity to -CCE and -CCM was unchanged. A previous study in rodents also reported that chronic treatment with the BZ chlordiazepoxide did not alter sensitivity to -CCE while increasing sensitivity to flumazenil (Takada et al., 1989). These results suggest that chronic diazepam treatment can decrease sensitivity to some negative modulators presumably because the effects of these compounds under these conditions are due to negative modulatory actions at the GABA_A receptor complex. The positive modulatory effects of diazepam, therefore, might attenuate the negative modulatory effects of some compounds. Striking differences among neutral and negative modulators in diazepam-treated monkeys clearly suggest that precipitation of withdrawal does not contribute identically to the behavioral actions of these compounds.

Previous studies have shown that BZ site positive modulators can antagonize negative modulators; however, chronic diazepam did not attenuate the rate-decreasing effects of -CCE and -CCM. This could reflect tolerance to the ability of diazepam to antagonize high efficacy negative modulators, as shown previously for the ability of the BZ lorazepam to antagonize DMCM (Petersen and Jensen, 1987). In support of this notion, acutely administered diazepam antagonized -CCE and -CCM, and produced even greater antagonism of DMCM than observed during chronic treatment. Tolerance to diazepam antagonism of negative modulators could be due to an increased potency of negative modulators in displacing positive modulators from BZ receptors in BZ-tolerant animals (Allan et al., 1992). It is not likely that chronic diazepam treatment conferred cross-tolerance to DMCM since previous studies in rodents have demonstrated increased sensitivity to negative modulators after discontinuation of chronic BZ treatment (Little et al., 1987). Similarly, sensitivity to Ro 15-4513 and -CCE was increased after discontinuation of chronic diazepam treatment in monkeys. Enhanced sensitivity after discontinuation of treatment was selective for these negative modulators insofar as sensitivity to triazolam, pentobarbital, pregnanolone, flumazenil, -CCM, and DMCM was similar to that determined before chronic diazepam treatment.

In summary, the present study demonstrates that daily treatment with the BZ diazepam confers cross-tolerance to a positive modulator acting at BZ sites and not to positive modulators acting at other sites on the GABA_A receptor complex (e.g., a barbiturate or neuroactive steroid). Moreover, changes in sensitivity to ligands that can antagonize diazepam under other conditions, such as neutral and negative modulators acting at BZ sites, appear to be modified in a manner that is correlated with the efficacy of these ligands. It is not clear to what extent the precipitation of withdrawal per se contributes to the rate-decreasing effects of negative modulators in diazepam-treated monkeys. This question might be resolved in monkeys with other procedures (e.g., drug discrimination) that have been successfully applied to the study of diazepam withdrawal.