Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Benzodiazepines that positively modulate GABA at the GABA_A receptor complex are the most commonly prescribed drugs for anxiety, insomnia, and convulsions because they are safe and effective (for review, see Woods et al., 1992). However, long-term daily treatment with benzodiazepines can lead to tolerance and, as evidenced by withdrawal signs that emerge upon discontinuation of treatment, dependence. A variety of approaches have been used to study benzodiazepine tolerance and dependence (e.g., observational procedures; Lukas and Griffiths, 1982), including drug discrimination. For example, the benzodiazepine antagonist flumazenil has been established as a discriminative stimulus in rhesus monkeys treated daily with the benzodiazepine diazepam (Gerak and France, 1999; McMahon et al., 2001, 2002). The consequences of chronic benzodiazepine treatment (e.g., tolerance) have been examined by comparing the discriminative stimulus effects of various compounds in diazepam-treated monkeys discriminating flumazenil to the discriminative stimulus effects of the same compounds in untreated monkeys discriminating midazolam (Lelas et al., 1999; McMahon et al., 2001, 2002). Such comparisons might be relevant to benzodiazepine dependence and withdrawal because the flumazenil discriminative stimulus in diazepam-treated monkeys is qualitatively similar to discriminative stimulus effects that emerge when diazepam treatment is temporarily discontinued (Gerak and France, 1999).

In addition to benzodiazepines, some barbiturates and neuroactive steroids positively modulate GABA by acting at nonbenzodiazepine sites on the GABA_A receptor complex (Gee et al., 1988; Turner et al., 1989). The effects of positive modulators acting at different sites have been shown to be qualitatively similar in untreated and diazepam-treated monkeys, i.e., benzodiazepines, barbiturates, and a neuroactive steroid substituted for midazolam and attenuated flumazenil in diazepam-treated monkeys (McMahon et al., 2001). On the other hand, the potency of positive modulators acting at different sites has been shown to be markedly different in untreated compared with diazepam-treated monkeys (McMahon et al., 2001). For example, doses of benzodiazepines larger than doses substituting for midazolam in untreated monkeys were required to attenuate flumazenil in diazepam-treated monkeys. The opposite relationship was revealed for nonbenzodiazepine ligands, i.e., doses of barbiturates and a neuroactive steroid smaller than doses substituting for midazolam attenuated flumazenil in diazepam-treated monkeys. One goal of the present study was to determine whether similar differences in relative potency are evident in untreated and diazepam-treated monkeys for other positive modulators acting at benzodiazepine (e.g., flunitrazepam and abecarnil) and neuroactive steroid sites (e.g., androsterone, dihydroandrosterone, and epipregnanolone).

Previous studies have demonstrated that efficacy in positively modulating GABA is an important determinant of discriminative stimulus effects in untreated and diazepam treated-monkeys, i.e., low-efficacy benzodiazepines (e.g., bretazenil) substituted for flumazenil and not for midazolam (Gerak and France, 1999; Lelas et al., 1999). Thus, another goal of this study was to further evaluate the importance of efficacy as a determinant of discriminative stimulus effects in untreated and diazepam-treated monkeys. Positive modulators chosen for study were reported to have low efficacy at some benzodiazepine receptor subtypes comprising ₁-, ₂-, ₃-, and ₅-subunits (e.g., abecarnil; Smith et al., 2001) or low efficacy at neuroactive steroid sites (e.g., dihydroandrosterone and epipregnanolone; Park-Chung et al., 1999).

Thus far, the features of the midazolam and flumazenil discriminative stimulus in untreated and diazepam-treated monkeys, respectively, have been evaluated with GABA_A modulators. A number of other drugs that do not have as a primary mechanism of action modulation of GABA share certain effects with benzodiazepines and other positive GABA_A modulators. For example, buspirone has anxiolytic effects (Uhlenhuth, 1982), ketamine has dissociative-anesthetic effects (Sadove et al., 1971), and diphenhydramine has sedative-hypnotic effects (Sunshine et al., 1978). Other drugs, such as -hydroxybutyrate (GHB; Mamelak et al., 1977) and valproic acid (Bruni and Wilder, 1979) can influence GABA transmission; however, the mechanism of action that mediates the behavioral effects of these compounds is unclear. Thus, this study examined whether these compounds substitute for or modulate the discriminative stimulus effects of midazolam. In addition, the present study examined whether non-GABAergic drugs that modulate various aspects of benzodiazepine withdrawal (e.g., buspirone, File and Andrews, 1991; valproic acid, Harris et al., 2000) substitute for or modulate the discriminative stimulus effects of flumazenil in diazepam-treated monkeys.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Subjects. Adult rhesus monkeys (Macaca mulatta) discriminating midazolam (four females and one male) or flumazenil (one female and four males) were housed individually on a 14-h light/10-h dark schedule and maintained at 95% free-feeding weight (range 3.8-10.0 kg) with a diet provided in the home cage comprising primate chow (High Protein Monkey Diet; Harlan Teklad, Madison, WI), fresh fruit, peanuts, and water. Monkeys discriminating flumazenil were treated daily with diazepam (5.6 mg/kg p.o.) for at least 1 year before these studies. The animals used in these studies were maintained in accordance with the Institutional Animal Care and Use Committee, The University of Texas Health Science Center at San Antonio, and with the 1996 Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources on Life Sciences, National Research Council, National Academy of Sciences).

Apparatus. Monkeys were trained to discriminate midazolam or flumazenil as described previously (Gerak and France, 1999; Lelas et al., 1999). During experimental sessions, monkeys were seated in chairs (model R001; Primate Products, Miami, FL) that provided neck restraint and placed in ventilated, sound-attenuating chambers equipped with two response levers, lights, and a food cup into which pellets could be delivered from a dispenser. For monkeys discriminating midazolam under a schedule of stimulus-shock termination, their feet were placed in shoes containing brass electrodes through which a brief electric stimulus (3 mA, 250 ms) could be delivered from an A/C generator. An interface (MED Associates, St. Albans, VT) connected the chambers to a computer, which controlled and recorded experimental events.

Midazolam Discrimination Procedure. Experimental sessions consisted of multiple 15-min cycles, each comprising a 10-min time-out period, during which responses had no programmed consequence. A 5-min response period followed, during which red lights were illuminated, thereby signaling the beginning of the response period in which an electric stimulus could be delivered every 15 s. The correct lever was designated by an injection (saline or midazolam) during the first minute of the cycle; designation of correct levers (e.g., left, saline; right, midazolam) varied among monkeys and remained the same for an individual throughout the study. Ten consecutive responses (fixed ratio 10) on the correct lever extinguished the red lights and postponed delivery of the electric stimulus for 30 s. Responding on the incorrect lever reset the response requirement on the correct lever. Response periods ended after 5 min or after the delivery of four electric stimuli, whichever occurred first.

Flumazenil Discrimination Procedure. Diazepam was given 3 h before experimental sessions. Multiple cycle procedures were similar to those described above except that the 5-min response period comprised a fixed ratio 5 schedule of food presentation during which a maximum of 10 food pellets (300-mg banana-flavored pellets; Bio-Serv, Frenchtown, NJ) was available. When the maximum number of food pellets was obtained in less than 5 min, the remainder of the response period was a time-out. Vehicle training comprised administration of vehicle or sham injections during the first minute of each of no more than eight cycles. Flumazenil training sessions comprised administration of flumazenil (0.32 mg/kg for three monkeys or 0.1 mg/kg for two monkeys s.c.) during the first minute of a cycle followed by a vehicle or sham injection during the first minute of a second cycle; flumazenil training cycles could be preceded by one to six vehicle or sham injection cycles. Test sessions were conducted after training sessions in which 80% of the total responses occurred on the lever designated as correct and fewer than five responses occurred on the incorrect lever before the first completion of the response requirement on the correct lever for all cycles. Before each test, these criteria had to be satisfied for both flumazenil and vehicle training sessions.

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-ml syringe. To obtain a dose of 5.6 mg/kg diazepam, a standard concentration of diazepam (1.0 mg/ml) was administered in a volume adjusted to individual body weights. The diazepam mixture was prepared immediately before administration.

Data Analyses. Drug discrimination data are expressed as the percentage of total responses occurring on the drug-appropriate lever averaged among monkeys (± S.E.M.) and plotted as a function of dose. Substitution for the training drug was defined as 80% responding on the drug-appropriate lever. When a test with a given compound was conducted more than once, the determinations were averaged for an individual subject for further analyses. Doses of a compound required to produce 50% drug-appropriate responding (ED₅₀) and the 95% confidence limits (95% CLs) were estimated using linear regression by using more than two appropriate data points, otherwise by interpolation. These values were determined first for individual monkeys and then averaged among all monkeys. Control midazolam or flumazenil ED₅₀ values and 95% CLs were determined periodically throughout the course of these studies, and these values were averaged to obtain an overall mean of the control ED₅₀ values and 95% CLs. ED₅₀ values for midazolam- or flumazenil-appropriate responding after administration of another drug were compared to the overall average of the control ED₅₀ values and 95% CLs. ED₅₀ values were considered to be significantly different from control when the average ED₅₀ value was not within the 95% CL values of the overall average of the control. The magnitude of shift elicited by a given drug was determined from the averaged data. Responding on both levers was divided by the duration of time that both levers were active. Control response rate represents the average of the five saline or vehicle-training sessions immediately preceding a test. Response rate was calculated as a percentage of control rate for individual animals and then averaged among subjects (±S.E.M.) and plotted as a function of dose.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Substitution of Positive GABA_A Modulators for Midazolam in Untreated Monkeys. Cumulative doses of midazolam increased midazolam-lever responding in a dose-related manner with a dose of 0.32 mg/kg occasioning 80% midazolam-lever responding in all monkeys (Fig. 1, top, closed circles). Administration of saline during the first cycle of these tests occasioned predominantly saline-appropriate responding (data not shown). The largest dose (0.32 mg/kg) of midazolam did not substantially modify response rate (Fig. 1, bottom, closed circles). The range of control midazolam ED₅₀ values was 0.11 to 0.17 mg/kg; the overall average of ED₅₀ values and 95% CLs was 0.15 mg/kg (0.08-0.23) (Table 1).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Discriminative-stimulus and rate effects of positive GABAA modulators in rhesus monkeys discriminating midazolam. Abscissae: dose in milligrams per kilogram of body weight. Ordinates: mean (±S.E.M.) percentage of responding on the midazolam lever (% DR, drug responding, top) and mean (±S.E.M.) response rate expressed as percentage of control rate [rate (% of control), bottom]. Data for flunitrazepam and androsterone are redrawn from 30 min in Figs. 2 and 4, respectively; data for abecarnil are redrawn from 90 min in Fig. 3. Data represent average values from at least three monkeys.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Time course of discriminative-stimulus (% DR) and rate (rate) effects of flunitrazepam in monkeys discriminating midazolam. Abscissae: time in minutes. Ordinates: mean (±S.E.M.) percentage of responding on the midazolam lever (% DR, drug responding, top) and mean (±S.E.M.) response rate expressed as percentage of control rate [rate (% of control), bottom]. Data represent average values from at least three monkeys.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Time course of discriminative-stimulus (% DR) and rate (rate) effects of abecarnil in monkeys discriminating midazolam. See Fig. 2 for other details.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Time course of discriminative-stimulus (% DR) and rate (rate) effects of androsterone in monkeys discriminating midazolam. See Fig. 2 for other details.

Attenuation of Flumazenil Discriminative Stimulus in Diazepam-Treated Monkeys with Positive GABA_A Modulators. In monkeys treated daily with diazepam, administration of flumazenil dose dependently increased responding on the flumazenil-appropriate lever with a dose of 0.1 mg/kg occasioning 80% flumazenil-lever responding (Figs. 5-9 and 12, top, closed circles). Administration of the flumazenil vehicle solution during the first cycle of these tests occasioned predominantly vehicle-appropriate responding (Figs. 5-9 and 12, top, closed circles above "V"). The range of control flumazenil ED₅₀ values was 0.009 to 0.050 mg/kg; the overall average ED₅₀ (95% CL) was 0.031 mg/kg (0.013-0.057) flumazenil (Table 2). Before administration of flumazenil in control tests, response rate was slightly decreased during the first cycle; on average, flumazenil did not substantially modify response rate (Figs. 5-9 and 12, bottom, closed circles).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Discriminative-stimulus and rate effects of flumazenil in diazepam-treated monkeys that received flunitrazepam. Abscissae: dose in milligrams per kilogram of body weight; V, vehicle. Ordinates: mean (±S.E.M.) percentage of responding on the drug-appropriate lever (% DR, drug responding, top) and mean response rate expressed as percentage of control (vehicle training days) rate [rate (% of control), bottom]. Flunitrazepam was administered at the start of the session 15 min before the administration of the first dose of flumazenil.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Discriminative-stimulus and rate effects of flumazenil in diazepam-treated monkeys that received abecarnil. Abecarnil was administered 105 min before the start of the session and 120 min before the administration of the first dose of flumazenil. See Fig. 5 for other details.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Discriminative-stimulus and rate effects of flumazenil in diazepam-treated monkeys that received androsterone. Androsterone was administered at the start of the session 15 min before the administration of the first dose of flumazenil. See Fig. 5 for other details.

View larger version (18K): [in this window] [in a new window]   Fig. 8.   Discriminative-stimulus and rate effects of flumazenil in diazepam-treated monkeys that received dihydroandrosterone. Dihydroandrosterone was administered at the start of the session 15 min before the administration of the first dose of flumazenil. See Fig. 5 for other details.

View larger version (16K): [in this window] [in a new window]   Fig. 9.   Discriminative stimulus and rate effects of flumazenil in diazepam-treated monkeys that received epipregnanolone. Epipregnanolone was administered at the start of the session 15 min before the administration of the first dose of flumazenil. See Fig. 5 for other details.

Effects of Other Anxiolytics, Sedatives, and Anticonvulsants in Untreated Monkeys Discriminating Midazolam. Single doses (0.1-1.0 mg/kg) of buspirone administered at the beginning of eight cycles occasioned predominantly saline-lever responding (data not shown), except in one monkey for which buspirone (0.32 mg/kg) substituted for midazolam at a single time point (30 min after injection). The group average response on the midazolam lever was 27% at 30 min after injection of 0.32 mg/kg buspirone (data not shown). In the monkey for which buspirone (0.32 mg/kg) substituted for midazolam, single doses (0.1 and 0.32 mg/kg) of buspirone administered before midazolam dose-response tests enhanced the midazolam discriminative stimulus. In this monkey, each dose (0.1 and 0.32 mg/kg) of buspirone shifted the midazolam discrimination dose-effect curve to the left as evidenced by 3-fold decreases in the ED₅₀ value for midazolam discrimination. However, when data from this monkey were averaged with data from three other monkeys, buspirone (0.1 and 0.32 mg/kg) did not significantly modify the ED₅₀ value for midazolam discrimination (Table 3). The smaller doses (0.1 and 0.32 mg/kg) of buspirone administered before cumulative doses of midazolam did not alter response rate (Table 3, before midazolam); however, buspirone (0.1 and 0.32 mg/kg) in combination with midazolam decreased response rate (Table 3, after midazolam). The largest dose (1.0 mg/kg) of buspirone decreased responding to 36% at 15 min after injection (Table 3, before midazolam).

View this table: [in this window] [in a new window]   TABLE 3 Mean ED50 values and 95% CLs for the midazolam discrimination dose-effect curve alone and after pretreatment with various compounds Rate of responding includes rate after administration of the pretreatment drug alone (before midazolam) and rate after administration of the pretreatment drug followed by the dose of midazolam resulting in a group average of 80% midazolam-lever responding (after midazolam). Data are from at least three monkeys.

View larger version (19K): [in this window] [in a new window]   Fig. 10.   Discriminative stimulus and rate effects of midazolam alone and after pretreatment with diphenhydramine. Diphenhydramine was administered at the start of the session 15 min before the administration of the first dose of midazolam. See Fig. 5 for other details.

View larger version (21K): [in this window] [in a new window]   Fig. 11.   Discriminative stimulus and rate effects of midazolam alone and after pretreatment with valproic acid. Valproic acid was administered at the start of the session 15 min before the administration of the first dose of midazolam. See Fig. 5 for other details.

Effects of Other Anxiolytics, Sedatives, and Anticonvulsants in Diazepam-treated Monkeys Discriminating Flumazenil. A dose of 0.1 mg/kg buspirone did not substitute for flumazenil and did not substantially modify the discriminative stimulus effects of flumazenil (Table 4). A larger dose (0.32 mg/kg) of buspirone alone or in combination with flumazenil suppressed responding in two of four monkeys (Table 4, before flumazenil). Flumazenil partially reversed the rate-decreasing effects produced by the combination of buspirone (0.32 mg/kg) and the daily dose of diazepam in two other monkeys (Table 4, after flumazenil); buspirone (0.32 mg/kg) did not significantly modify the ED₅₀ value for flumazenil discrimination in these two monkeys (Table 4). Similarly, single doses (100.0-1000.0 mg/kg) of GHB administered before flumazenil dose-effect tests did not substitute for flumazenil and did not substantially modify response rate; GHB (100.0-1000.0 mg/kg) did not modify the discriminative stimulus effects of flumazenil (Table 4).

View this table: [in this window] [in a new window]   TABLE 4 Mean ED50 values and 95% CLs for the flumazenil discrimination dose-effect curve alone and after pretreatment with various compounds Rate of responding includes rate after administration of the pretreatment drug alone (before flumazenil) and rate after administration of the pretreatment drug followed by the dose of flumazenil resulting in a group average of 80% midazolam-lever responding (after flumazenil). Data are from at least three monkeys.

View larger version (19K): [in this window] [in a new window]   Fig. 12.   Discriminative stimulus and rate effects of flumazenil in diazepam-treated monkeys that received valproic acid. Valproic acid was administered at the start of the session 15 min before the administration of the first dose of flumazenil. See Fig. 5 for other details.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Positive GABA_A modulators acting at benzodiazepine and neuroactive steroid sites substituted for midazolam in untreated monkeys and attenuated flumazenil in diazepam-treated monkeys. These results suggest that positive modulators acting at benzodiazepine and neuroactive steroid sites had qualitatively similar effects that might include attenuation of diazepam withdrawal. Although positive modulators at different sites had qualitatively similar effects, site of action was an important determinant of potency in attenuating a flumazenil discriminative stimulus in diazepam-treated monkeys, i.e., neuroactive steroids were relatively more potent than benzodiazepines compared with their respective potency in midazolam-discriminating monkeys. Drugs with mechanisms not predominantly involving modulation of GABA did not substitute for midazolam or flumazenil; however, valproic acid enhanced midazolam and attenuated flumazenil. Thus, although these discrimination assays seem to be selective for positive GABA_A modulation and not other mechanisms underlying anxiolytic, sedative or anticonvulsant activity, some drugs (e.g., valproic acid) have in common with positive GABA_A modulators the ability to modulate benzodiazepine dependence and withdrawal.

Compounds that modulate GABA with high efficacy at benzodiazepine receptors have been shown to substitute for midazolam and to attenuate flumazenil in diazepam-treated monkeys (Gerak and France, 1999; Lelas et al., 1999; McMahon et al., 2001, 2002). In the current study, flunitrazepam substituted for midazolam and attenuated flumazenil; both results are consistent with flunitrazepam having high affinity at benzodiazepine receptor subtypes comprising ₁-, ₂-, ₃-, and ₅-subunits and having high efficacy in positively modulating GABA at these subtypes in vitro (Smith et al., 2001). Some compounds that modulate GABA with low efficacy at benzodiazepine receptors have effects in discrimination assays that are qualitatively different from high-efficacy benzodiazepines. For example, the low-efficacy positive modulator bretazenil substitutes for flumazenil (e.g., antagonizes diazepam) and not for midazolam (Gerak and France, 1999; Lelas et al., 1999). Abecarnil binds with high affinity at all benzodiazepine receptor subtypes and positively modulates GABA with high efficacy at receptors comprising ₁-subunits and relatively low efficacy at receptors comprising other -subunits (Smith et al., 2001). Substitution of abecarnil for midazolam suggested that high efficacy at benzodiazepine receptors comprising ₁-subunits was sufficient for producing midazolam-like effects; alternatively, efficacy of abecarnil at receptors comprising ₂-, ₃-, and ₅-subunits might have been sufficient for occasioning midazolam-like effects. Unlike flunitrazepam, abecarnil occasioned 57% flumazenil-lever responding in one diazepam-treated monkey and disrupted responding when combined with flumazenil. The effects of abecarnil in diazepam-treated monkeys were somewhat different from other benzodiazepine receptor ligands and might be related to differential efficacy of abecarnil at benzodiazepine receptor subtypes.

Pregnanolone positively modulates GABA with high efficacy at neuroactive steroid sites in vitro (Park-Chung et al., 1999), substitutes for midazolam, and attenuates flumazenil in diazepam-treated monkeys (McMahon et al., 2001). In the current study, androsterone, a neuroactive steroid with efficacy comparable with that of pregnanolone, substituted for midazolam and attenuated flumazenil. Androsterone (ED₅₀ value of 26.64) was less potent than pregnanolone (ED₅₀ value of 6.40 mg/kg) in substituting for midazolam, a result consistent with androsterone being less potent than pregnanolone in modulating GABA in vitro (Park-Chung et al., 1999). Dihydroandrosterone and epipregnanolone are neuroactive steroids that have been reported to have approximately 5-fold lower efficacy than other steroids (Pignataro and Fiszer de Plazas, 1997; Park-Chung et al., 1999). Dihydroandrosterone and epipregnanolone, up to a dose of 100 mg/kg, substituted for midazolam in two of three monkeys and attenuated flumazenil in diazepam-treated monkeys. Failure of dihydroandrosterone and epipregnanolone to substitute for midazolam could be related to the limited range of doses that could be studied, to limited bioavailability, or to low efficacy.

A previous study reported that potency of positive GABA_A modulators in substituting for midazolam in untreated monkeys did not predict potency in attenuating flumazenil in diazepam-treated monkeys (McMahon et al., 2001). Results of the present study confirm and extend these findings to other modulators acting at different sites on the GABA_A receptor complex. Figure 13 compares the potency of positive modulators in untreated and diazepam-treated monkeys by depicting the magnitude of the rightward shift in the flumazenil dose-effect curve (ordinate) as a function of each dose of each of the modulators, where the dose of the modulator is expressed as a multiple of its ED₅₀ value in substituting for midazolam (abscissa). A dose of 1 (abscissa) represents the midazolam substitution ED₅₀ value for the appropriate positive modulator. Benzodiazepine site ligands (Fig. 13, closed symbols) were relatively less potent in diazepam-treated monkeys, shifting the flumazenil dose-effect curve 2- to 7-fold rightward; it is possible that larger doses of benzodiazepine site ligands would shift the flumazenil dose-effect curve further to the right. In contrast, neuroactive steroids (Fig. 13, open symbols) were relatively more potent in diazepam-treated monkeys compared with untreated monkeys. A dose of pregnenolone one-half the ED₅₀ value in midazolam-discriminating monkeys shifted the flumazenil dose-effect curve more than 20-fold to the right, whereas comparable doses of other neuroactive steroids produced smaller rightward shifts in the flumazenil dose-effect curve. Differences in relative potency among drugs that act at neuroactive steroid sites might be due to differential efficacy at GABA_A receptor subtypes. The greater relative potency of positive modulators acting at neuroactive steroid sites compared with benzodiazepine sites could be related to whether modulators act at flumazenil-sensitive sites on the GABA_A receptor complex. For example, the behavioral effects of pregnanolone are not antagonized by flumazenil (McMahon and France, 2001), whereas flumazenil antagonizes benzodiazepines in a competitive manner (Lelas et al., 2000); thus, benzodiazepines and not steroids act at the same site as flumazenil on the GABA_A receptor complex. Noncompetitive interactions at the GABA_A receptor complex seem to more potently attenuate flumazenil than competitive interactions at benzodiazepine receptors.

View larger version (26K): [in this window] [in a new window]   Fig. 13.   Rightward shift in the flumazenil dose-effect function elicited by positive GABAA modulators expressed as a multiple of their midazolam substitution ED50 value. Abscissa: multiple of the ED50 value of the appropriate positive GABAA modulator in substituting for midazolam. Ordinate: mean (±S.E.M.) rightward shift in the flumazenil dose-effect curve expressed as flumazenil ED50 value after pretreatment with a dose of positive GABAA modulator divided by the corresponding control flumazenil ED50 value. Vertical dashed line represents the ED50 value for midazolam substitution. Data for pregnanolone, diazepam, and midazolam are from McMahon et al. (2001).

Drugs with mechanisms that do not predominantly involve modulation of GABA and that share therapeutic or other effects with benzodiazepines (e.g., buspirone, ketamine, and diphenhydramine) were studied to examine the pharmacological specificity of these discrimination assays and to determine whether these compounds modify the acute and withdrawal-related effects of benzodiazepines. It is not clear to what extent GABA is modulated by other compounds that were studied (e.g., GHB and valproic acid). Buspirone did not substitute for midazolam; however, buspirone occasioned some responding on the midazolam lever (Spealman, 1985). Although buspirone facilitates flunitrazepam binding in vivo (Oakley and Jones, 1983), the discriminative stimulus effects of midazolam were not enhanced by buspirone. Buspirone can also decrease anxiety-like behavior during diazepam withdrawal in rodents (File and Andrews, 1991). However, buspirone did not modify the flumazenil-discriminative stimulus in diazepam-treated monkeys, a result that is consistent with clinical studies showing that buspirone does not attenuate benzodiazepine withdrawal (Schweizer and Rickels, 1986; Lader and Olajide, 1987). Collectively, these results suggest that buspirone and benzodiazepines have qualitatively different effects and that buspirone does not substantially modify the behavioral effects of benzodiazepines. However, because daily administration of buspirone is typically required for anxiolysis in humans (Goldberg, 1984), daily treatment with buspirone might be required to attenuate the flumazenil discriminative stimulus in diazepam-treated monkeys if that attenuation is due to anxiolytic activity.

Valproic acid is used to treat anxiety, seizures, bipolar disorder, and drug abuse (for review, see Davis et al., 2000). Although valproic acid blocks sodium channels and increases GABA levels through inhibition of GABA transaminase, the mechanism responsible for its therapeutic effects is not well understood. Valproic acid, up to a dose of 1000 mg/kg, did not substitute for midazolam or flumazenil; however, valproic acid shifted the midazolam dose-effect curve leftward and the flumazenil dose-effect curve rightward. Valproic acid enhances the anesthetic effects of ethanol and a barbiturate (Hoffman and Habib, 1994), suggesting that valproic acid can modulate effects of some positive GABA_A modulators. The ability of valproic acid to attenuate the flumazenil discriminative stimulus in diazepam-treated monkeys is consistent with some (Harris et al., 2000) and not other (Rickels et al., 1999) results obtained in benzodiazepine-dependent subjects.

The sedative hypnotics and anesthetics GHB and ketamine did not substitute for or modify the discriminative-stimulus effects of midazolam or flumazenil (Table 3; Gerak and France, 1999; Lelas et al., 1999; Woolverton et al., 1999). Diphenhydramine did not substitute for midazolam or flumazenil, a finding that is consistent with previous studies in monkeys (Spealman, 1985; Evans and Johanson, 1989). However, relatively large doses of diphenhydramine substituted for amphetamine in some nonprimate species, produced hyperactivity and convulsions in monkeys (Evans and Johanson, 1989), and shifted the midazolam dose-effect curve to the right in the present study. Antagonism of midazolam by diphenhydramine is likely due to a functional interaction mediated by histaminic or muscarinic and not benzodiazepine receptors (Kubo et al., 1987). Diphenhydramine did not modify the flumazenil discriminative stimulus in diazepam-treated monkeys, suggesting that diphenhydramine does not exacerbate diazepam withdrawal. Differential effects of diphenhydramine in untreated and diazepam-treated monkeys might suggest that antagonism of the acute effects of benzodiazepines does not necessarily confer flumazenil-like effects in diazepam-treated monkeys.

In summary, the present results demonstrate that the discriminative-stimulus effects of midazolam in untreated monkeys and of flumazenil in diazepam-treated monkeys are mediated by GABA_A modulation and not other mechanisms underlying anxiolytic, sedative, or anticonvulsant activity. These results also suggest that valproic acid and diphenhydramine do not act at the GABA_A receptor complex to modify behavioral effects of benzodiazepines resulting from positive GABA_A modulation. Results with compounds acting at benzodiazepine and neuroactive steroid sites demonstrate that positive GABA_A modulation at different sites results in qualitatively similar effects. However, the greater relative potency of neuroactive steroids in diazepam-treated monkeys, compared with benzodiazepines, indicates that site of action is an important determinant of relative potency under conditions of benzodiazepine treatment that result in tolerance and dependence. If the flumazenil discrimination assay represents diazepam withdrawal, these results might be relevant to the future development of neuroactive steroids and other drugs (e.g., valproic acid) for modulating benzodiazepine dependence and withdrawal.