- Email: [email protected]
Ro 15-1788 antagonizes the discriminative stimulus effects of diazepam in rats but not similar effects of pentobarbital
Life Sciecnes, Vol. 31, pp. 2105-2112 Printed in the U,S.A.
Pergamon Press
Ro 15-1788 ANTAGONIZES THE DISCRIMINATIVE STIMULUS EFFECTS OF DIAZEPAM IN RATS BUT NOT SIMILAR EFFECTS OF PENTOBARBITAL Seymore Herling and Harlan E. Snannon National Institute on Drug Abuse Division of Research Addiction Research Center Lexington, KY 40583, U.S.A.
(Received in final form August 2, 1982) Summary The benzodiazepine antagonist properties of Ro 15-1788 were evaluated in rats trained to discriminate between saline and either 1.0 mg/kg of diazepam or i0 mg/kg of pentobarbital in a two-choice discrete-trial shock avoidance procedure. When administered alone, 1.0 mg/kg of diazepam and i0 mg/kg of pentobarbital produced comparable amounts of drug-appropriate responding (> 84%), whether rats were trained to discriminate between diazepam or pentobarbital and saline. Ro 15-1788 (3-32 mg/kg, p.o.), administered I0 min before diazepam or pentobarbital, produced a dose-related blockade of the discriminative effects of diazepam in both groups of rats, but was completely ineffective in blocking the discriminative effects of pentobarbital. The dose-effect curve for the discriminative effects of diazepam was shifted to the right in a parallel fashion 3- and 13-fold by i0 and 32 mg/kg of Ro 15-1788, respectively, indicating that Ro 15-1788 acts as a surmountable, competitive antagonist of diazepam. When administered alone, Ro 15-1788 (32-i00 mg/kg, p.o.) produced primarily saline-appropriate responding, although I00 mg/kg of Ro 15-1788 produced drug-appropriate responding in one out of eight rats. When administered orally 30 min after diazepam, Ro 15-1788 (32 mg/kg) completely reversed within i0 min the discriminative effects of diazepam. The blockade of diazepam's discriminative effects by 32 mg/kg of Ro 15-1788 appeared to last at least as long (approximately 2 nr) as the effects of diazepam alone. Specific receptors in the central nervous system which selectively recognize pharmacologically active benzodiazepines, but not other sedative-hypnotics such as barbiturates, are thought to mediate many of the pharmacologic effects of benzodiazepines (1,2). Recently, an ±midazodiazepine has been synthesized which potently inhibits [3H]-diazepam binding to rat brain, but which lacks characteristic benzodiazepine-like agonist activity (3). This compound, Ro 15-1788 (ethyl-8-fluoro-5-6-d±hydro-5-methyl-6-oxo-4H-imidazol [l,5a] [1,4] benzodiazepine-3-carboxylate), antagonizes diazepam-induced decreases in cerebellar cGMP levels in rats (4) and certain electrophysiological effects of various benzodiazepines in rats and cats (5) without affecting similar changes produced by barbiturates, ethanol, methaqualone or meprobamate. In addition, Ro 15-1788 reverses the antiaversive and anticonvulsant effects of diazepam but not similar effects produced by phenobarbital (3) or
0024-3205/82/192105-08503~00/0
2106
Antagonism of Diazepam by Ro 15-1788
progabide, a GABA agonist (6). Thus selective benzodiazepine antagonist.
Ro
15-1788
Vol. 31, No.
appears
to
be
19, 1982
a
highly
In addition to sharing the neurophysiologic and behavioral effects noted above, benzodiazepines and barbiturates produce similar discriminative stimulus effects (e.g., 7-9). Drugs that snare a wide spectrum of pharmacologic activity are often thought to have similar mechanisms of action (e.g. i0, ll). With the discovery of Ro 15-1788, however, it has been possible to determine that certain of the effects of benzodiazepines (e.g., effects on spinal cord activities; anticonvulsant and anticonflict effects) are mediated by a mechanism (e.g., via an interaction with benzodiazepine receptors) that differs from that mediating similar effects of barbiturates, since only the effects of benzodiazepines are blocked by Ro 15-1788 (3,5). In order to determine whether the discriminative stimulus effects of benzodiazepines and barbiturates might be differentiated on the basis of their susceptibility to antagonism by Ro 15-1788, rats were trained to discriminate Oetween s.c. injections of either diazepam (i.0 mg/kg) or pentobarbital (i0 mg/Mg) and saline and the ability of graded doses of Ro 15-1788 to block the discriminative effects produced by diazepam or pentobarbital was investigated. In addition, dose-effect curves for diazepam were determined in the presence of graded doses of Ro 15-1788 or vehicle in order to determine whether Ro 15-1788 acts as a competitive antagonist of diazepam. Finally, the onset and duration of the antagonist activity of Ro 15-1788 was investigated. Methods Su0~ects. The suojects were Fischer-derived F344 male rats (Harlan Industries, Indianapolis, IN) weighing 200-250 g at the start of discrimination training. Between experimental sessions, the animals were housed three per cage in a large colony room with food and water continuously available. The lights in the colony room were illuminated between 6:00 a.m. and 6:00 p.m. Apparatus. Standard two-lever rat chambers (model llO1-L, Grason-Stadler Co., Inc., Bolton, MA) were modified as described previously (12). A single lever midway along one wall was designated the starting response lever, and the two levers on the opposite wall were designated the choice response levers. A clear Plexiglas partition separated the two choice response levers. A scrambled electric shock was delivered to the grid floor of the test chamber by a constant current shock generator (model 700, Grason-Stadler Co., Inc.). The test chamber was enclosed in a light-proof, sound-attenuating and ventilated enclosure. Schedule contingencies were programmed and data recorded by a SCAT 3002/PDP8 system (BKP Scientific, Berlin, MA). Discrimination training. The procedure used has been described in detail previously (12). Briefly, rats were trained in a two-choice, discrete-trial avoidance procedure to complete the two-response chain of a starting response followed by an appropriate choice response in order to terminate each trial. The start of a trial was signaled by the simultaneous onset of the houselight and white noise. The first starting response on each trial terminated the white noise; if an appropriate choice response was then emitted, the houselight was extinguished and the trial ended. If the inappropriate choice response was emitted, the white noise was presented again and the rat was required to begin the two-response chain again. A trial was defined as being correct only if the first two-response chain completed by the animal was a starting response followed by an appropriate choice response. Beginning 5 sec after the onset of a trial, a scrambled 1.0 mA shock was intermittently presented as 1.0 sec pulses every 3.0 sec for the duration of the trial. The intertrial interval was 45 sec. Sessions ended after 20 trials or 30 min, whichever occurred first.
Vol. 31, No. 19, 1982
Antagonism of Diazepam by Ro 15-1788
2107
Training sessions were conducted 6 days per week. One group of rats was trained to discriminate between diazepam (i.0 mg/kg) and saline; a second group of rats was trained to discriminate between pentobarbital (i0 mg/kg) and saline. Drug or saline was administered s.c. 30 min before the session. Trials were terminated by responses on one of the two choice levers during sessions preceded by an injection of drug or by responses on the other choice lever when sessions were preceded by an injection of saline. Training continued as described previously (12) until a rat completed four consecutive sessions in which at least 90%, i.e., 18 of the 20 trials, were correct for each session. When an animal met this criterion, the next two sessions (one drug and one saline) were conducted as test sessions during which responses on either choice lever terminated a trial. An animal was considered to have acquired the discrimination if at least 90% of the trials during both test sessions were completed on the choice lever appropriate for the pretreatment injection. Tests of stimulus ~eneralization and antagonism. After stimulus control of behavior was established, test sessions, during which responses on either choice lever terminated a trial, were interposed among training sessions. Training sessions continued to be conducted on the first, second, fourth and fifth sessions of each week in order to maintain performance of the discrimination at criterion levels. Saline or drug (diazepam or pentobarbital) was administered prior to training sessions on a double alternation basis (e.g., S, D, D, S). Test sessions were conducted on the third and sixth sessions of each week if the animal completed at least 90% of the trials during the preceding training sessions on the appropriate choice lever; otherwise, additional training sessions were conducted. In both the diazepam- and pentobarbital-trained rats, test sessions were conducted in which the effects of 1.0 mg/kg of diazepam or i0 mg/Kg of pentobarbital were determined in the presence of graded doses of Ro 15-1788 (3-32 mg/kg) or the Ro 15-1788 vehicle. The effects of Ro 15-1788 (32-100 mg/kg) or vehicle administered alone were also determined. In addition, dose-effect curves for diazepam in diazepam-trained rats were determined in the presence of Ro 15-1788 vehicle or graded doses of Ro 15-1788 (3-32 mg/kg). Diazepam and pentobarbital were always administered s.c. 30 min before a test session. Ro 15-1788 was always administered orally by garage 40 min before the session. Time-course experiments. Time-course determinations were obtained in single 4.5 hr sessions by testing the animals during nine consecutive 30 min intervals witn a maximum of 20 trials in each interval. The 20 trials were usually completed within 20 min; during the remainder of each 30 min interval, all lights in the chamber were extinguished. Three separate time-course experiments were conducted. The first 20 trials of each tlme-course determination were immediately preceded by a s.c. injection of saline. The second 20 trials were always preceded immediately by a s.c. injection of 1.0 mg/kg of diazepam. The third set of 20 trials was preceded immediately by an oral intubation of either 10 or 32 mg/kg of Ro 15-1788 or vehicle. Subsequent sets of 20 trials were not preceded by an injection. All time-course determinations were made in the same group of four rats with at least one weeM intervening between each time-course experiment. Druss. Diazepam base was dissolved in saline and the pH of the solution adjusted to 1.5 with I N hydrochloric acid. Sodium pentobarbital was dissolved in saline and the pH adjusted to ii with i N sodium hydroxide. Ro 15-1788 was suspended in distilled water to which two drops of Tween 80 per 10 ml were added. Drugs or vehicle were administered in a volume of 1.0-3.0 ml/kg.
2108
Antagonism of Diazepam by Ro 15-1788
DIAZEPAM -TRAINED w co z 0 13_ w
,,L w _J
Vol. 31, No. 19, 1982
PENTOBARBITAL- TRAINED
I00-
80-
60-
I
n~ C:I I-z
','
40-
20-
u.I
OT
I
I
I
V
3
I0
32
u
I00
Ro 1 5 - 1 7 8 8
3-
i
I
I
V
3
I0
32
I
I00
(mg/kg)
FIG. I Dose-related antagonism by Ro 15-1788 of the discriminative effects produced Oy 1.0 mg/~g of diazepam s.c. (closed circles) but not by i0 mg/kg of pentobarbital s.c. (open circles) in rats trained to discriminate between either diazepam (left panel, n = 4) or pentobarbital (right panel, n = 3) and saline. The effects of Ro 15-1788 administered alone in diazepam-trained (n = 4) or pentobarbital-trained (n = 4) rats are shown by open triangles. Ordinates: % diazepamor pentobarbital-lever responses. Abscissae: dose of Ro 15-1788 (mg/Kg, p.o.). Points at "V" indicate the effects of the vehicle. Vertical lines indicate + I S.E.M. The dashed lines at i0 and 90% indicate tne minimum levels at which the discrimination performance was maintained with saline and diazepam or pentobarbital prior to test sessions. Data analysis. Discrimination data are presented as the percentage of trials, of a maximum of 20, completed on the drug-appropriate choice lever; the remaining trials were completed on the saline-approprlate choice lever (see also 12). Shifts in the dose-response curves were quantitated using biDassay statistics (13). The bioassays were conducted as crossover designs within subjects. Results Oiazepam tantly with diazepam- or above "V").
(i.0 mg/kg) and pentobarbital (i0 mg/kg) administered concomithe Ro 15-1788 vehicle produced comparable percentages of pentobarbital-appropriate responding ( > 84%; Fig. i, circles The Ro 15-1788 vehicle produced only saline-appropriate re-
Vol. 31, No. 19, 1982
Antagonism if Diazepam by Ro 15-1788
t',4
t,/) LI..I Or)
,,>,
60
"
f
DOSE OF Ro 15-1788 • VEHICLE
40
~1
r ~
20
n-
w Q_
2109
o
s mQ/kg
0
I0 mg/kg 32 rng/kg
[]
0
I
I
I
I
I
0.1
0.5
I
5
I0
,
I 32
DIAZEPAM (mg/kg) FIG. 2 Competitive antagonism by Ro 15-1788, administered p.o., of the discriminative stimulus effects of diazepam in rats trained to discriminate between 1.0 mg/kg of diazepam and saline (n = 5)Ordinates: % diazepam-lever responses. Abscissae: dose of diazepam (mg/kg, s.c.). Verticle lines indicate + i S.E.M. sponding (open triangles at "V"). Increasing doses of Ro 15-1788 (3-32 mg/kg, p.o.) resulted in a dose-related decrease in diazepam- or pentobarbital-appropriate responding produced by diazepam (Fig. i, closed circles), but had no effect on drug-appropriate responding produced by pentobarbital (open circles). A dose of 32 mg/kg of Ro 15-1788 was required in order to completely antagonize the effects of 1.0 mg/kg of diazepam. When administered alone, Ro 15-1788 (32 or i00 mg/kg) produced little or no drug-appropriate responding in seven out of eight rats tested, but in one pentobarbital-trained rat, i00 mg/Kg of Ro 15-1788 produced 95% drug-appropriate responding. When lO0 mg/kg of Ro 15-1788 was retested in this rat, this dose of Ro 15-1788 resulted in 90% drug-appropriate responding. Higher doses of the drug were not tested due to limited supply. The average effect of Ro 15-1788, when administered alone, is shown by open triangles in Fig. i. In rats trained to discriminate between saline and diazepam, Ro 15-1788 (3-32 mg/kg) produced dose-related parallel shifts to the right in the doseeffect curve for diazepam (Fig. 2). The diazepam dose-effect curve was relatively unaffected by 3 mg/kg of Ro 15-1788, but was shifted to the right in a parallel manner 3-fold (95% C.L.: 2.00-4.96) and 13-fold (7.05-33.04) by i0 and 32 mg/kg of Bo 15-1788, respectively.
2110
Antagonism of Diazepam by Ro 15-1788
o.. Go la_l n,"
SALINE
Ro 1 5 - 1 7 8 8
IO0-
80la.J > I,I --J
DOSE OF R o l 5 - 1 7 8 8 •
r'~ t-Z Ll.I (-) rr W 0-
VEHICLE
© I0 m g / k g 0 32 m g / k g
60-
I
13W N
19, 1982
1.0 m g / k g DIAZEPAM
if) w o') Z
O
Vol. 31, No.
40-
20-
0I
-30
I
0
I
I0
I
1
i
30 40
I
i
I
60
90
TIME
FIG.
I
120
150
I
180
I
210
(minutes)
3
Time-course of the antagonist activity of Ro 15-1788. Ordinates: % diazepam-lever responses. Abscissae: time (min). Saline was always administered at -30 min. Diazepam s.c. was always administered at O min. Points at O min represent measurements taken between O and I0 min after the injection of diazepam. Points at iO min represent measurements taken between iO and 20 min after the injection of diazepam. Varying doses of Ro 15-1788 (vehicle, i0 mg/kg or 32 mg/kg p.o.) were administered at 30 min (30 min after diazepam). Points at 30 min represent measurements taken O to I0 min after a dose of Ro 15-1788 (i.e., 30 to 40 min after diazepam). Points at 40 min represent measurements taken i0 to 20 min after Ro 15-1788. Points at subsequent times represent measurements taken over approximately 20 min beginning at the time after injection of diazepam indicated along the abscissae. Each point is the average of one observation in each of 4 subjects. Vertical lines indicate + i S.E.M. Time-course determinations. In each time-course determination, all choice responses during the first test interval, which was preceded by an injection of saline, occurred on the saline-appropriate lever (Fig. 3: points above -30 min). During the second test interval, which was preceded by an injection of 1.0 mg/kg of diazepam, the rats completed an average of approximately 40% of the trials on the diazepam lever during the first i0 min after the injection (points above 0 min in Fig. 3). Diazepam-appropriate responding increased to between 70 and 85% of the total responses between i0 and 20 min after the injection of diazepam (points above i0 min in Fig. 3). When Ro 151788 vehicle was administered 30 min after diazepam (Fig. 3: closed circles),
Vol. 31, No.
19, 1982
Antagonism of Diazepam by Ro 15-1788
2111
90-100% diazepam-appropriate responding occurred between 30 and 50 min after diazepam. Subsequently, there was a progressive decline in the percentage of diazepam-appropriate responses emitted, with predominantly saline-appropriate responding occurring 150 min after diazepam administration. In contrast, 32 mg/kg of Ro 15-1788 p.o. (Fig. 3: open circles) produced an almost immediate and sustained blockade of diazepam-appropriate responding that was complete within i0 min of its injection. A lower dose of Ro 15-1788 (I0 mg/kg) also antagonized the effects of diazepam, but considerably less rapidly and completely than the higher dose of tne antagonist. Discussion The imidazodiazepine Ro 15-1788 selectively antagonized the discriminative stimulus effects of diazepam, whereas similar effects of pentobarbital were unaffected by Ro 15-1788. In addition, increasing doses of Ro 15-1788 produced dose-related parallel shifts to the right in the diazepam dose-effect curve, indicating that Ro 15-1788 acts as a surmountable, competitive antagonist of diazepam. Finally, Ro 15-1788 was virtually devoid of diazepam-like agonist activity up to a dose 3-fold higher than that required to antagonize the effects of diazepam. These results replicate and extend previous reports (3,4) that Ro 15-1788 is a selective antagonist of Oenzodiazepines. Drugs that produce similar discriminative stimulus effects are often thought to have similar mechanisms of action. Such appears to be the case for certain narcotics (e.g., 14). Likewise, the similarity in the discriminative stimulus effects of diazepam and pentobarbital suggests that these two drugs also share a common mode of action. However, Ro 15-1788 antagonized the discriminative effects of diazepam but was completely ineffective in blocking similar effects of pentobarbital. Since Ro 15-1788 binds selectively to CNS but not to peripheral benzodiazepine receptors (3,4), the discriminative effects of diazepam, but not those of pentobarbital, appear to be mediated via benzodiazepine receptors in the CNS. The possibility cannot be ruled out, however, that pentobarOital acts on the same neurons as does diazepam but through a mechanism other than benzodiazepine receptors. Ro 15-1788 was considerably less potent as an antagonist of the discriminative effects of diazepam than would be expected from its potency in inhibiting [3H]-benzodiazeplne binding. Ro 15-1788 is approximately 4 times more potent than diazepam in inhibiting [3H]-diazepam binding to homogenates of rat cerebral cortex and approximately half as potent as diazepam in preventing the in vivo binding of [3H]-flunitrazepam in mouse brain (3). In the present study, a dose of Ro 15-1788 (32 mg/kg) approximately 30 times greater than that of diazepam (i.0 mg/kg) was required to completely antagonize the discriminative effects of diazepam. The reasons for this discrepancy are not entirely clear. Although we administered Ro 15-1788 orally, it is doubtful that this route of administration can account entirely for the low potency of the antagonist for the following reasons. Firstly, as an antagonist, Ro 15-1788 administered orally is roughly equipotent to diazepam as an agonist in decreasing cerebellar cGMP levels in rats (4) as well as in preventing the accumulation of [3H]-flunitrazepam in mouse brain (3). Secondly, in the present time-course determinations, Ro 15-1788 was rapidly effective by the oral route yet a nearly 30-fold higher dose than that of diazepam was required. The possibility that Ro 15-1788 administered orally is rapidly degraded to an inactive product cannot, however, be ruled out, since a much lower dose of the drug (0.6 mg/kg) administered i.v. reverses the anticonvulsant effects of diazepam (6). Alternatively, zodiazeplne type.
Ro 15-1788 may be a very weak partial agonist of the benIn the present study, i00 mg/kg of Ro 15-1788 produced
2112
Antagonism of Diazepam by Ro 15-1788
Vol. 31, No. 19, 1982
agonist activity in one out of eight rats. Two previous studies have also indicated that high doses of Ro 15-1788 have weak benzodiazepine-like agonist activity (6,15). Further, the dose-effect curve for precipitating abstinence in the diazepam-dependent rat has a very shallow slope (16). It should be noted, however, that in the initial report of its effects, up to subtoxic doses of Ro 15-1788 failed to produce any of the classic effects of benzodiazepines (3). Thus, it is presently unclear whether Ro 15-1788 is a weak partial agonist of the benzodiazepine type or if the modest agonist activity that has been observed with Ro 15-1788 occurs through mechanisms other than an interaction with benzodiazepine receptors. In summary, the imidazodiazepine Ro 15-1788 selectively antagonizes the discriminative stimulus effects of diazepam but not similar effects of pentobarbital. Further, Ro 15-1788 is rapidly effective by the oral route. The present results are consistent with the interpretation that Ro 15-1788 is a surmountable, competitive benzodiazepine antagonist, but the possibility also exists that it is a very weak partial agonist. Acknowledsements We thank Drs. W. E. Scott and W. Haefely of Hoffmann-LaRoche and Co. Ltd. for providing the diazepam and Ro 15-1788 used in this study and W. Roberts and P. Settlemire for assistance in preparing the manuscript. References i. 2. 3. 4. 5. 6. 7. 8.
9. iO. 11.
12. 13. 14. 15. 16.
H. MOHLER and T. OKADA, Science 198 849-851, (1977). C. BRAESTRUP and R.F. SQUIRES, Eur. J. Pharmacol. 48 263-270 (1978). W. MUNKELER, H. MOHLER, L. PIERI, P. POLC, E.P. BONETTI, R. CUMIN, R. SCHAFFNER and W. HAEFELY, Nature 290 514-516 (1981). H. MOHLER, W.P. BURKARD, H.H. KELLER, J.G. RICHABDS and W. HAEFELY, J. Neurochem. 37 714-722 (1981). P. POLC, J.-P. LAURENT, R. SCHERSCHLICHT and W. HAEFELY, NaunynSchmiedeberg's Arch. Pharmacol. 316 317-325 (1981). K.G. LLOYD, P. BOVIER, C.L. BROEKKAMP and P. WORMS, Eur. J. Pnarmacol. 75 77-78 (1981). F.C. COLPAERT, L.K.C. DESMEDT and P.A.J. JANSSEN, Eur. J. Pnarmacol. 37 i13-123 (1976). S. HERLING, R.J. VALENTINO and G.D. WINGER, Psychopbarmacol. 71 21-28 (1980). O. WINGER and S. HERLING, Psychopharmacol. 76172-176 (1982). W.R. MARTIN, C.G. EADES, W.O. THOMPSON, R.E. HUPPLER and P.E. GILBERT, J. Pbarmacol. Exp. Tner. 197 517-532 (1976). D.R. JASINSKI, H.E. SHANNON, E.J. CONE, D.B. VAUPEL, M.E. RISNER, R.L. McQUINN, T.-P. SU and W.B. PICKWORTH, PCP (Phencyclidine: H i s t o r i c a l and Current Perspectives, E.F. Domino ( e d . ) , pp. 331-400, NPP Books, Ann Arbor, MI (1981). H.E. SHANNON, J. Pharmaeol. Exp. Trier. 216 543-551 (1981). D.J. FINNEY, J. Pharmaool. Exp. Tber. 198 497-506 (1976). S. HERLING and J.H. WOODS, L i f e Sci. 28 1571-1584 (1981). D.J. NUTT, K.P.J. COWENand H.J. LITTLE, Nature 295 436-438 (1982). L.F. McNICHOLAS and W.R. MARTIN, Life Sci., in press (1982).
Pergamon Press
Ro 15-1788 ANTAGONIZES THE DISCRIMINATIVE STIMULUS EFFECTS OF DIAZEPAM IN RATS BUT NOT SIMILAR EFFECTS OF PENTOBARBITAL Seymore Herling and Harlan E. Snannon National Institute on Drug Abuse Division of Research Addiction Research Center Lexington, KY 40583, U.S.A.
(Received in final form August 2, 1982) Summary The benzodiazepine antagonist properties of Ro 15-1788 were evaluated in rats trained to discriminate between saline and either 1.0 mg/kg of diazepam or i0 mg/kg of pentobarbital in a two-choice discrete-trial shock avoidance procedure. When administered alone, 1.0 mg/kg of diazepam and i0 mg/kg of pentobarbital produced comparable amounts of drug-appropriate responding (> 84%), whether rats were trained to discriminate between diazepam or pentobarbital and saline. Ro 15-1788 (3-32 mg/kg, p.o.), administered I0 min before diazepam or pentobarbital, produced a dose-related blockade of the discriminative effects of diazepam in both groups of rats, but was completely ineffective in blocking the discriminative effects of pentobarbital. The dose-effect curve for the discriminative effects of diazepam was shifted to the right in a parallel fashion 3- and 13-fold by i0 and 32 mg/kg of Ro 15-1788, respectively, indicating that Ro 15-1788 acts as a surmountable, competitive antagonist of diazepam. When administered alone, Ro 15-1788 (32-i00 mg/kg, p.o.) produced primarily saline-appropriate responding, although I00 mg/kg of Ro 15-1788 produced drug-appropriate responding in one out of eight rats. When administered orally 30 min after diazepam, Ro 15-1788 (32 mg/kg) completely reversed within i0 min the discriminative effects of diazepam. The blockade of diazepam's discriminative effects by 32 mg/kg of Ro 15-1788 appeared to last at least as long (approximately 2 nr) as the effects of diazepam alone. Specific receptors in the central nervous system which selectively recognize pharmacologically active benzodiazepines, but not other sedative-hypnotics such as barbiturates, are thought to mediate many of the pharmacologic effects of benzodiazepines (1,2). Recently, an ±midazodiazepine has been synthesized which potently inhibits [3H]-diazepam binding to rat brain, but which lacks characteristic benzodiazepine-like agonist activity (3). This compound, Ro 15-1788 (ethyl-8-fluoro-5-6-d±hydro-5-methyl-6-oxo-4H-imidazol [l,5a] [1,4] benzodiazepine-3-carboxylate), antagonizes diazepam-induced decreases in cerebellar cGMP levels in rats (4) and certain electrophysiological effects of various benzodiazepines in rats and cats (5) without affecting similar changes produced by barbiturates, ethanol, methaqualone or meprobamate. In addition, Ro 15-1788 reverses the antiaversive and anticonvulsant effects of diazepam but not similar effects produced by phenobarbital (3) or
0024-3205/82/192105-08503~00/0
2106
Antagonism of Diazepam by Ro 15-1788
progabide, a GABA agonist (6). Thus selective benzodiazepine antagonist.
Ro
15-1788
Vol. 31, No.
appears
to
be
19, 1982
a
highly
In addition to sharing the neurophysiologic and behavioral effects noted above, benzodiazepines and barbiturates produce similar discriminative stimulus effects (e.g., 7-9). Drugs that snare a wide spectrum of pharmacologic activity are often thought to have similar mechanisms of action (e.g. i0, ll). With the discovery of Ro 15-1788, however, it has been possible to determine that certain of the effects of benzodiazepines (e.g., effects on spinal cord activities; anticonvulsant and anticonflict effects) are mediated by a mechanism (e.g., via an interaction with benzodiazepine receptors) that differs from that mediating similar effects of barbiturates, since only the effects of benzodiazepines are blocked by Ro 15-1788 (3,5). In order to determine whether the discriminative stimulus effects of benzodiazepines and barbiturates might be differentiated on the basis of their susceptibility to antagonism by Ro 15-1788, rats were trained to discriminate Oetween s.c. injections of either diazepam (i.0 mg/kg) or pentobarbital (i0 mg/Mg) and saline and the ability of graded doses of Ro 15-1788 to block the discriminative effects produced by diazepam or pentobarbital was investigated. In addition, dose-effect curves for diazepam were determined in the presence of graded doses of Ro 15-1788 or vehicle in order to determine whether Ro 15-1788 acts as a competitive antagonist of diazepam. Finally, the onset and duration of the antagonist activity of Ro 15-1788 was investigated. Methods Su0~ects. The suojects were Fischer-derived F344 male rats (Harlan Industries, Indianapolis, IN) weighing 200-250 g at the start of discrimination training. Between experimental sessions, the animals were housed three per cage in a large colony room with food and water continuously available. The lights in the colony room were illuminated between 6:00 a.m. and 6:00 p.m. Apparatus. Standard two-lever rat chambers (model llO1-L, Grason-Stadler Co., Inc., Bolton, MA) were modified as described previously (12). A single lever midway along one wall was designated the starting response lever, and the two levers on the opposite wall were designated the choice response levers. A clear Plexiglas partition separated the two choice response levers. A scrambled electric shock was delivered to the grid floor of the test chamber by a constant current shock generator (model 700, Grason-Stadler Co., Inc.). The test chamber was enclosed in a light-proof, sound-attenuating and ventilated enclosure. Schedule contingencies were programmed and data recorded by a SCAT 3002/PDP8 system (BKP Scientific, Berlin, MA). Discrimination training. The procedure used has been described in detail previously (12). Briefly, rats were trained in a two-choice, discrete-trial avoidance procedure to complete the two-response chain of a starting response followed by an appropriate choice response in order to terminate each trial. The start of a trial was signaled by the simultaneous onset of the houselight and white noise. The first starting response on each trial terminated the white noise; if an appropriate choice response was then emitted, the houselight was extinguished and the trial ended. If the inappropriate choice response was emitted, the white noise was presented again and the rat was required to begin the two-response chain again. A trial was defined as being correct only if the first two-response chain completed by the animal was a starting response followed by an appropriate choice response. Beginning 5 sec after the onset of a trial, a scrambled 1.0 mA shock was intermittently presented as 1.0 sec pulses every 3.0 sec for the duration of the trial. The intertrial interval was 45 sec. Sessions ended after 20 trials or 30 min, whichever occurred first.
Vol. 31, No. 19, 1982
Antagonism of Diazepam by Ro 15-1788
2107
Training sessions were conducted 6 days per week. One group of rats was trained to discriminate between diazepam (i.0 mg/kg) and saline; a second group of rats was trained to discriminate between pentobarbital (i0 mg/kg) and saline. Drug or saline was administered s.c. 30 min before the session. Trials were terminated by responses on one of the two choice levers during sessions preceded by an injection of drug or by responses on the other choice lever when sessions were preceded by an injection of saline. Training continued as described previously (12) until a rat completed four consecutive sessions in which at least 90%, i.e., 18 of the 20 trials, were correct for each session. When an animal met this criterion, the next two sessions (one drug and one saline) were conducted as test sessions during which responses on either choice lever terminated a trial. An animal was considered to have acquired the discrimination if at least 90% of the trials during both test sessions were completed on the choice lever appropriate for the pretreatment injection. Tests of stimulus ~eneralization and antagonism. After stimulus control of behavior was established, test sessions, during which responses on either choice lever terminated a trial, were interposed among training sessions. Training sessions continued to be conducted on the first, second, fourth and fifth sessions of each week in order to maintain performance of the discrimination at criterion levels. Saline or drug (diazepam or pentobarbital) was administered prior to training sessions on a double alternation basis (e.g., S, D, D, S). Test sessions were conducted on the third and sixth sessions of each week if the animal completed at least 90% of the trials during the preceding training sessions on the appropriate choice lever; otherwise, additional training sessions were conducted. In both the diazepam- and pentobarbital-trained rats, test sessions were conducted in which the effects of 1.0 mg/kg of diazepam or i0 mg/Kg of pentobarbital were determined in the presence of graded doses of Ro 15-1788 (3-32 mg/kg) or the Ro 15-1788 vehicle. The effects of Ro 15-1788 (32-100 mg/kg) or vehicle administered alone were also determined. In addition, dose-effect curves for diazepam in diazepam-trained rats were determined in the presence of Ro 15-1788 vehicle or graded doses of Ro 15-1788 (3-32 mg/kg). Diazepam and pentobarbital were always administered s.c. 30 min before a test session. Ro 15-1788 was always administered orally by garage 40 min before the session. Time-course experiments. Time-course determinations were obtained in single 4.5 hr sessions by testing the animals during nine consecutive 30 min intervals witn a maximum of 20 trials in each interval. The 20 trials were usually completed within 20 min; during the remainder of each 30 min interval, all lights in the chamber were extinguished. Three separate time-course experiments were conducted. The first 20 trials of each tlme-course determination were immediately preceded by a s.c. injection of saline. The second 20 trials were always preceded immediately by a s.c. injection of 1.0 mg/kg of diazepam. The third set of 20 trials was preceded immediately by an oral intubation of either 10 or 32 mg/kg of Ro 15-1788 or vehicle. Subsequent sets of 20 trials were not preceded by an injection. All time-course determinations were made in the same group of four rats with at least one weeM intervening between each time-course experiment. Druss. Diazepam base was dissolved in saline and the pH of the solution adjusted to 1.5 with I N hydrochloric acid. Sodium pentobarbital was dissolved in saline and the pH adjusted to ii with i N sodium hydroxide. Ro 15-1788 was suspended in distilled water to which two drops of Tween 80 per 10 ml were added. Drugs or vehicle were administered in a volume of 1.0-3.0 ml/kg.
2108
Antagonism of Diazepam by Ro 15-1788
DIAZEPAM -TRAINED w co z 0 13_ w
,,L w _J
Vol. 31, No. 19, 1982
PENTOBARBITAL- TRAINED
I00-
80-
60-
I
n~ C:I I-z
','
40-
20-
u.I
OT
I
I
I
V
3
I0
32
u
I00
Ro 1 5 - 1 7 8 8
3-
i
I
I
V
3
I0
32
I
I00
(mg/kg)
FIG. I Dose-related antagonism by Ro 15-1788 of the discriminative effects produced Oy 1.0 mg/~g of diazepam s.c. (closed circles) but not by i0 mg/kg of pentobarbital s.c. (open circles) in rats trained to discriminate between either diazepam (left panel, n = 4) or pentobarbital (right panel, n = 3) and saline. The effects of Ro 15-1788 administered alone in diazepam-trained (n = 4) or pentobarbital-trained (n = 4) rats are shown by open triangles. Ordinates: % diazepamor pentobarbital-lever responses. Abscissae: dose of Ro 15-1788 (mg/Kg, p.o.). Points at "V" indicate the effects of the vehicle. Vertical lines indicate + I S.E.M. The dashed lines at i0 and 90% indicate tne minimum levels at which the discrimination performance was maintained with saline and diazepam or pentobarbital prior to test sessions. Data analysis. Discrimination data are presented as the percentage of trials, of a maximum of 20, completed on the drug-appropriate choice lever; the remaining trials were completed on the saline-approprlate choice lever (see also 12). Shifts in the dose-response curves were quantitated using biDassay statistics (13). The bioassays were conducted as crossover designs within subjects. Results Oiazepam tantly with diazepam- or above "V").
(i.0 mg/kg) and pentobarbital (i0 mg/kg) administered concomithe Ro 15-1788 vehicle produced comparable percentages of pentobarbital-appropriate responding ( > 84%; Fig. i, circles The Ro 15-1788 vehicle produced only saline-appropriate re-
Vol. 31, No. 19, 1982
Antagonism if Diazepam by Ro 15-1788
t',4
t,/) LI..I Or)
,,>,
60
"
f
DOSE OF Ro 15-1788 • VEHICLE
40
~1
r ~
20
n-
w Q_
2109
o
s mQ/kg
0
I0 mg/kg 32 rng/kg
[]
0
I
I
I
I
I
0.1
0.5
I
5
I0
,
I 32
DIAZEPAM (mg/kg) FIG. 2 Competitive antagonism by Ro 15-1788, administered p.o., of the discriminative stimulus effects of diazepam in rats trained to discriminate between 1.0 mg/kg of diazepam and saline (n = 5)Ordinates: % diazepam-lever responses. Abscissae: dose of diazepam (mg/kg, s.c.). Verticle lines indicate + i S.E.M. sponding (open triangles at "V"). Increasing doses of Ro 15-1788 (3-32 mg/kg, p.o.) resulted in a dose-related decrease in diazepam- or pentobarbital-appropriate responding produced by diazepam (Fig. i, closed circles), but had no effect on drug-appropriate responding produced by pentobarbital (open circles). A dose of 32 mg/kg of Ro 15-1788 was required in order to completely antagonize the effects of 1.0 mg/kg of diazepam. When administered alone, Ro 15-1788 (32 or i00 mg/kg) produced little or no drug-appropriate responding in seven out of eight rats tested, but in one pentobarbital-trained rat, i00 mg/Kg of Ro 15-1788 produced 95% drug-appropriate responding. When lO0 mg/kg of Ro 15-1788 was retested in this rat, this dose of Ro 15-1788 resulted in 90% drug-appropriate responding. Higher doses of the drug were not tested due to limited supply. The average effect of Ro 15-1788, when administered alone, is shown by open triangles in Fig. i. In rats trained to discriminate between saline and diazepam, Ro 15-1788 (3-32 mg/kg) produced dose-related parallel shifts to the right in the doseeffect curve for diazepam (Fig. 2). The diazepam dose-effect curve was relatively unaffected by 3 mg/kg of Ro 15-1788, but was shifted to the right in a parallel manner 3-fold (95% C.L.: 2.00-4.96) and 13-fold (7.05-33.04) by i0 and 32 mg/kg of Bo 15-1788, respectively.
2110
Antagonism of Diazepam by Ro 15-1788
o.. Go la_l n,"
SALINE
Ro 1 5 - 1 7 8 8
IO0-
80la.J > I,I --J
DOSE OF R o l 5 - 1 7 8 8 •
r'~ t-Z Ll.I (-) rr W 0-
VEHICLE
© I0 m g / k g 0 32 m g / k g
60-
I
13W N
19, 1982
1.0 m g / k g DIAZEPAM
if) w o') Z
O
Vol. 31, No.
40-
20-
0I
-30
I
0
I
I0
I
1
i
30 40
I
i
I
60
90
TIME
FIG.
I
120
150
I
180
I
210
(minutes)
3
Time-course of the antagonist activity of Ro 15-1788. Ordinates: % diazepam-lever responses. Abscissae: time (min). Saline was always administered at -30 min. Diazepam s.c. was always administered at O min. Points at O min represent measurements taken between O and I0 min after the injection of diazepam. Points at iO min represent measurements taken between iO and 20 min after the injection of diazepam. Varying doses of Ro 15-1788 (vehicle, i0 mg/kg or 32 mg/kg p.o.) were administered at 30 min (30 min after diazepam). Points at 30 min represent measurements taken O to I0 min after a dose of Ro 15-1788 (i.e., 30 to 40 min after diazepam). Points at 40 min represent measurements taken i0 to 20 min after Ro 15-1788. Points at subsequent times represent measurements taken over approximately 20 min beginning at the time after injection of diazepam indicated along the abscissae. Each point is the average of one observation in each of 4 subjects. Vertical lines indicate + i S.E.M. Time-course determinations. In each time-course determination, all choice responses during the first test interval, which was preceded by an injection of saline, occurred on the saline-appropriate lever (Fig. 3: points above -30 min). During the second test interval, which was preceded by an injection of 1.0 mg/kg of diazepam, the rats completed an average of approximately 40% of the trials on the diazepam lever during the first i0 min after the injection (points above 0 min in Fig. 3). Diazepam-appropriate responding increased to between 70 and 85% of the total responses between i0 and 20 min after the injection of diazepam (points above i0 min in Fig. 3). When Ro 151788 vehicle was administered 30 min after diazepam (Fig. 3: closed circles),
Vol. 31, No.
19, 1982
Antagonism of Diazepam by Ro 15-1788
2111
90-100% diazepam-appropriate responding occurred between 30 and 50 min after diazepam. Subsequently, there was a progressive decline in the percentage of diazepam-appropriate responses emitted, with predominantly saline-appropriate responding occurring 150 min after diazepam administration. In contrast, 32 mg/kg of Ro 15-1788 p.o. (Fig. 3: open circles) produced an almost immediate and sustained blockade of diazepam-appropriate responding that was complete within i0 min of its injection. A lower dose of Ro 15-1788 (I0 mg/kg) also antagonized the effects of diazepam, but considerably less rapidly and completely than the higher dose of tne antagonist. Discussion The imidazodiazepine Ro 15-1788 selectively antagonized the discriminative stimulus effects of diazepam, whereas similar effects of pentobarbital were unaffected by Ro 15-1788. In addition, increasing doses of Ro 15-1788 produced dose-related parallel shifts to the right in the diazepam dose-effect curve, indicating that Ro 15-1788 acts as a surmountable, competitive antagonist of diazepam. Finally, Ro 15-1788 was virtually devoid of diazepam-like agonist activity up to a dose 3-fold higher than that required to antagonize the effects of diazepam. These results replicate and extend previous reports (3,4) that Ro 15-1788 is a selective antagonist of Oenzodiazepines. Drugs that produce similar discriminative stimulus effects are often thought to have similar mechanisms of action. Such appears to be the case for certain narcotics (e.g., 14). Likewise, the similarity in the discriminative stimulus effects of diazepam and pentobarbital suggests that these two drugs also share a common mode of action. However, Ro 15-1788 antagonized the discriminative effects of diazepam but was completely ineffective in blocking similar effects of pentobarbital. Since Ro 15-1788 binds selectively to CNS but not to peripheral benzodiazepine receptors (3,4), the discriminative effects of diazepam, but not those of pentobarbital, appear to be mediated via benzodiazepine receptors in the CNS. The possibility cannot be ruled out, however, that pentobarOital acts on the same neurons as does diazepam but through a mechanism other than benzodiazepine receptors. Ro 15-1788 was considerably less potent as an antagonist of the discriminative effects of diazepam than would be expected from its potency in inhibiting [3H]-benzodiazeplne binding. Ro 15-1788 is approximately 4 times more potent than diazepam in inhibiting [3H]-diazepam binding to homogenates of rat cerebral cortex and approximately half as potent as diazepam in preventing the in vivo binding of [3H]-flunitrazepam in mouse brain (3). In the present study, a dose of Ro 15-1788 (32 mg/kg) approximately 30 times greater than that of diazepam (i.0 mg/kg) was required to completely antagonize the discriminative effects of diazepam. The reasons for this discrepancy are not entirely clear. Although we administered Ro 15-1788 orally, it is doubtful that this route of administration can account entirely for the low potency of the antagonist for the following reasons. Firstly, as an antagonist, Ro 15-1788 administered orally is roughly equipotent to diazepam as an agonist in decreasing cerebellar cGMP levels in rats (4) as well as in preventing the accumulation of [3H]-flunitrazepam in mouse brain (3). Secondly, in the present time-course determinations, Ro 15-1788 was rapidly effective by the oral route yet a nearly 30-fold higher dose than that of diazepam was required. The possibility that Ro 15-1788 administered orally is rapidly degraded to an inactive product cannot, however, be ruled out, since a much lower dose of the drug (0.6 mg/kg) administered i.v. reverses the anticonvulsant effects of diazepam (6). Alternatively, zodiazeplne type.
Ro 15-1788 may be a very weak partial agonist of the benIn the present study, i00 mg/kg of Ro 15-1788 produced
2112
Antagonism of Diazepam by Ro 15-1788
Vol. 31, No. 19, 1982
agonist activity in one out of eight rats. Two previous studies have also indicated that high doses of Ro 15-1788 have weak benzodiazepine-like agonist activity (6,15). Further, the dose-effect curve for precipitating abstinence in the diazepam-dependent rat has a very shallow slope (16). It should be noted, however, that in the initial report of its effects, up to subtoxic doses of Ro 15-1788 failed to produce any of the classic effects of benzodiazepines (3). Thus, it is presently unclear whether Ro 15-1788 is a weak partial agonist of the benzodiazepine type or if the modest agonist activity that has been observed with Ro 15-1788 occurs through mechanisms other than an interaction with benzodiazepine receptors. In summary, the imidazodiazepine Ro 15-1788 selectively antagonizes the discriminative stimulus effects of diazepam but not similar effects of pentobarbital. Further, Ro 15-1788 is rapidly effective by the oral route. The present results are consistent with the interpretation that Ro 15-1788 is a surmountable, competitive benzodiazepine antagonist, but the possibility also exists that it is a very weak partial agonist. Acknowledsements We thank Drs. W. E. Scott and W. Haefely of Hoffmann-LaRoche and Co. Ltd. for providing the diazepam and Ro 15-1788 used in this study and W. Roberts and P. Settlemire for assistance in preparing the manuscript. References i. 2. 3. 4. 5. 6. 7. 8.
9. iO. 11.
12. 13. 14. 15. 16.
H. MOHLER and T. OKADA, Science 198 849-851, (1977). C. BRAESTRUP and R.F. SQUIRES, Eur. J. Pharmacol. 48 263-270 (1978). W. MUNKELER, H. MOHLER, L. PIERI, P. POLC, E.P. BONETTI, R. CUMIN, R. SCHAFFNER and W. HAEFELY, Nature 290 514-516 (1981). H. MOHLER, W.P. BURKARD, H.H. KELLER, J.G. RICHABDS and W. HAEFELY, J. Neurochem. 37 714-722 (1981). P. POLC, J.-P. LAURENT, R. SCHERSCHLICHT and W. HAEFELY, NaunynSchmiedeberg's Arch. Pharmacol. 316 317-325 (1981). K.G. LLOYD, P. BOVIER, C.L. BROEKKAMP and P. WORMS, Eur. J. Pnarmacol. 75 77-78 (1981). F.C. COLPAERT, L.K.C. DESMEDT and P.A.J. JANSSEN, Eur. J. Pnarmacol. 37 i13-123 (1976). S. HERLING, R.J. VALENTINO and G.D. WINGER, Psychopbarmacol. 71 21-28 (1980). O. WINGER and S. HERLING, Psychopharmacol. 76172-176 (1982). W.R. MARTIN, C.G. EADES, W.O. THOMPSON, R.E. HUPPLER and P.E. GILBERT, J. Pbarmacol. Exp. Tner. 197 517-532 (1976). D.R. JASINSKI, H.E. SHANNON, E.J. CONE, D.B. VAUPEL, M.E. RISNER, R.L. McQUINN, T.-P. SU and W.B. PICKWORTH, PCP (Phencyclidine: H i s t o r i c a l and Current Perspectives, E.F. Domino ( e d . ) , pp. 331-400, NPP Books, Ann Arbor, MI (1981). H.E. SHANNON, J. Pharmaeol. Exp. Trier. 216 543-551 (1981). D.J. FINNEY, J. Pharmaool. Exp. Tber. 198 497-506 (1976). S. HERLING and J.H. WOODS, L i f e Sci. 28 1571-1584 (1981). D.J. NUTT, K.P.J. COWENand H.J. LITTLE, Nature 295 436-438 (1982). L.F. McNICHOLAS and W.R. MARTIN, Life Sci., in press (1982).