- Email: [email protected]
Differential development of tolerance to the depressant effects of benzodiazepine and non-benzodiazepine agonists at the omega (BZ) modulatory sites of GABAA receptors
Neurophormaco~ogy Vol. 31, No. I, pp. 693-700, 1992 Printed in Great Britain. AI1 rights reserved
002%3908/92 $5.00 + 0.00
Copyright 6 1992 Pergamon Press Ltd
DIFFERE~IAL DEVELOPME~ OF TOLERANCE TO THE DEPRESSANT EFFECTS OF BENZODIAZEPINE AND NON-BENZODIAZEPINE AGONISTS AT THE OMEGA (BZ) MODULATORY SITES OF GABA, RECEPTORS D. J. SANGER and B. ZIVKOVIC Synthelabo Recherche (L.E.R.S), 31 Ave P.V. Couturier, 92220 Bagneux, France (Accepted 31 January 1992)
Sunmmry-In a previous study, it was found that both the benzodiazepine hypnotic, midazolam, and the imid~opy~dine hypnotic, zolpidem, which has selective affinity for a sub-population of omega (benzodiazepine, BZ) modulatory sites of GABA,, receptors, produced similar decreases in rates of food-reinforced lever pressing in rats. However, during 10 days of repeated administration, marked tolerance developed to the depressant effect of midazolam but little tolerance developed with zolpidem. It was found in the present study that, with a within-subject design similar to that used previously, tolerance developed to the response rate-decreasing activity of the benzodiazepine, triazolam and the cyclopyrrolone, zopiclone but not to that of the triazolopyridazine, CL 218,872. In another experiment, using a between-groups design, tolerance developed to the effect of midazolam, even if the injections were not associated with daily test sessions, providing no evidence for a dNg~nvironment interaction. The lack of tolerance to zolpidem was confirmed in two experiments. Them was little indication of tolerance to the depressant effect of zolpidem, even after 19 days administration of daily doses, up to 30 mg/kg, a dose 10 times greater than that which completely suppressed responding. These results showed that the extent to which tolerance develops to the effects of drugs with affinity for omega (BZ) modulatory sites can show wide variations which may be related to differences in mechanisms of action. Key wor&--tolerance,
benzodiazepines, zolpidem, topiclone, CL 218,872.
It has been well established that tolerance can develop during repeated administration of benzodiazepines. However, the degree of tolerance can depend on a variety of pharmacological, behavioural and environmental factors (File, 1985). In general, tolerance to the behavioural depressant effects of these drugs, which presumably represents their muscle relaxant and sedative effects, develops rapidly, as shown in studies of locomotor activity and instrumental responding (e.g. File, 1981; Margules and Stein, 1968; Griffiths and Goudie, 1987). In recent years a number of compounds have been described which, while being chemically distinct from the benzodiazepines, act at similar receptors. These compounds belong to several chemical series including /I-cat&lines, imidazopy~~nes, cyclopyrrolones and pyrazoloquinolines (Gardner, 1989). Although the pharmacological profiles of these agents have many features in common with benzodiazepines, significant differences have also been described in some cases. It is therefore important to investigate whether tolerance will develop to the pharmacological effects of these newer compounds. Zolpidem is an imidazopyridine with selectivity for the y-aminobutyric acid (GABA*) subtype of recep tor, endowed with the w, (BZ,) modulatory sites (Langer, Arbilla, Scatton, Niddam and Dubois,
1988). In extensive pharmacological studies, zolpidem has been found to have hypnotic activity of rapid onset and short duration and its ~havioural and neuropharmacological activities differ in certain respects from those of the benzodiazepines. Specifically, zolpidem shows hypnotic activity at doses smaller than those which induce anticonvulsant, anxiolyticlike and muscle relaxant effects, it has not been found to increase food and water intake and it produces only limited cross generalization with benzodiazepines in drug-discrimination studies (Depoortere, Zivkovic, Lloyd, Sanger, Perrault, Langer and Bartholini, 1986; Perrault, Morel, Sanger and Zivkovic, 1990a; Sanger, Perrault, Morel, Joly and Zivkovic, 1987). In a study of repeated a~inistration of zolpidem to rats trained to lever-press for food reward, it was found that no tolerance occurred to the zolpideminduced depression of responding during 10 daily injections. In contrast, very clear tolerance developed to the similar effect of midazolam (Sanger and Zivkovic, 1987). The purpose of the present series of experiments was to investigate the generality of this finding. A similar technique was used to study the effects of repeated admiuistration of not only midazolam and zolpidem but also the benzodiazepine, triazolam, the cyclopyrrolone, 693
694
D. J. SANGER and B. ZIVKOVIC
CL 218,872 zopiclone and triazolopyridazine, (3-methyl-6[3-(trifluoromethyl)phenyl]-1,2,4-triazolo[4,3#]pyridazine. The drug CL 218,872 is of particular interest because it, like zolpidem, is selective for the w, subtype of benzodiazepine receptors (Squires, Benson, Braestrup, Coupet, Klepner, Myers and Beer, 1979; Lippa, Coupe& Greenblat, Klepner and Beer, 1979) and has also been reported not to give rise to tolerance (McElroy, Fleming and Feldman, 1985). A second purpose of these studies was to investigate whether the tolerance, previously observed with midazolam (Sanger and Zivkovic, 1987) depended on the fact that the animals were repeatedly tested in the drugged state. Such behaviourally-augmented or learned tolerance has been found with benzodiazepines, including midazolam and both instrumental conditioning (Herberg and Montgomery, 1987) and classical conditioning (King, Bouton and Musty, 1987) processes have been implicated. However, Griffiths and Goudie (1986, 1987) found no evidence for a role for conditioning processes in tolerance to midazolam, METHODS
Experiment 1. Repeated administration of triazolam and CL 218,872 Animals. Subjects were 18 male Wistar rats (Charles River, France). They weighed 18&200g when obtained from the supplier and were allowed to grow during the experiment so that they weighed 400-500g during the period of administration of drugs. Rats were individually housed under standard laboratory conditions (lights on 07.0~19.00), with water freely available in the home cages. At the start of training the rats were deprived of food for 2 days, They were subsequently given a standard quantity of chow each evening and over the weekends, which allowed them to gain weight but maintained motivation for operant responding. This procedure is used routinely in the authors’ laboratory and maintains stable rates of lever pressing over many months. Apparatus. All experiments were carried out in standard operant test chambers (Campden Instruments). Reinforcement for lever pressing was provided by 45 mg food pellets (Bio-Serv). Procedure. After deprivation of food and habituation to the test chambers, the rats were trained to press the lever to the right of the food dispenser to obtain the 45 mg food pellets during daily 15 min sessions. At the beginning of training every response produced a food pellet. The schedule requirement was then gradually increased over approx 10 sessions until 10 lever presses were necessary for each pellet (fixedratio 10: FR 10). The rats were maintained on this FR 10 schedule, until the response rates became stable and during which time they were adapted to receiving intraperitoneal injections of saline, 20 min
before the sessions. Throughout the experiment, daily sessions remained at 15 min duration and rats were tested for 5 days each week. After training on the FR 10 schedule, the rats were divided into 2 groups to receive administration of triazolam and CL 218,872. A dose-response curve was established with 3 doses of each drug; injection of drug being given on a Tuesday and the following Friday and Tuesday. The doses were: triazolam 0.1, 0.3 and 1.0 mg/kg and CL 218,872 3.0, 10 and 30 mg/kg. The largest doses (i.e. triazolam 1.0 mg/kg and CL 218,872 30 mg/kg) were then administered daily for a further 9 days. The rats continued to be tested after each injection, except during the weekend (i.e. days 5 and 6) when they received injections of drug but were not tested. The period of repeated injections finished on Thursday and the rats were injected with saline on the following Friday and Monday. The dose-response curves for triazolam and CL 218,872 were then reestablished with injections being given on Tuesdays and Fridays. Drugs. Triazolam and CL 218,872 were prepared in physiological saline. The drugs were synthesized at the Chemistry Department, Synthelabo Recherche and all doses are expressed in terms of the bases. Injection volume was 2 ml/kg. Data analysis. Values for control response rates were obtained from the days preceding days of injection of drug. The results were analysed separately for the 2 groups of rats, using analysis of variance (ANOVA) and Duncan’s test. Experiment 2. Repeated a~inistration
of zopicione
AnimaZ~. Subjects were 6 male Wistar rats, maintained under conditions similar to those described above. Procedure. The procedure was similar to that of Experiment 1, except for several modifications. Rats were trained to lever press for food pellets on an FR 10 schedule during daily 5 min sessions and adapted to receiving subcutaneous injections of physiological saline, 30 min before the sessions. After a stable baseline had been established, the rats received injections of zopiclone at doses of 1.0, 3.0, 10 and 30 mg/kg (s.c.), 30 min before the sessions. The injections of drug were given at intervals of several days, the first on a Friday and the others on the following Tuesday, Friday and Tuesday. The rats were then injected with saline on the Wednesday after the last acute injection and a period of 9 further days of injection of the largest dose began on the next day. The rats were injected but not tested over the weekend. The period of repeated injection of zopiclone finished on a Friday and the rats were injected with saline on the following Monday. The dose-response curve was reestablished over the following 3 weeks, injections of drug continuing to be given on Tuesdays and Fridays.
Tolerance to benzodiazepine agonists Drugs. Zopiclone was prepared in saline containing 2 drops of Tween 80, to give an injection volume of 2 ml/kg. Experiment 3. Non-contingent repeutedadministration of midazolam and zolpidem Animals. Subjects were 24 male Wistar rats, maintained under conditions similar to those described above. Fro~edure. Rats were trained to lever press for food reward on an FR 10 schedule during daily 15 min sessions. They were also adapted to receiving subcutaneous injections of physiological saline, 15 min before the sessions and at the end of each afternoon (approx 6 hr after testing). When stable baselines of responding were established, the rats were divided into 4 groups of 6. One group was then given 10 daily injections of midazolam (3.0 mg/kg, s.c.) at the end of each afternoon. The second group received similar treatment with a dose of zolpidem (1.0 mg/kg, s.c.) and the other 2 groups received injections of saline. The period of repeated injection of drug began on a Monday and continued over a weekend, when the rats did not receive behavioural testing. After chronic administration of midazolam, zolpidem or saline, the rats were tested but not injected with drugs for 1 week. Dose-response curves were then established to midazolam in the midazolam-treated rats and one group of saline-treated rats and to zolpidem in the zolpidem-treated and the other saline group. The doses of midazolam were 0.3, 1.0, 3.0 and lOmg/kg (s.c.) and those of zolpidem 0.1, 0.3, 1.0 and 3.0 mg/kg (s.c.), injected 15 min before the sessions. Drugs were injected on Tuesdays and Fridays, with saline being administered on the intervening days. Drugs. Zolpidem and midazolam were dissolved in physiological saline, to give injection volumes of 2 ml/kg. Experiment 4. Chronic large doses of zolpidem Animals. Subjects were 6 male Wistar rats, maintained under conditions similar to those described above. Procedure. Rats were trained to lever press on an FR 10 schedule of food reward during daily 15 min sessions. They were also adapted to intraperitoneal injections of physiological saline. Because a cumulative dose procedure was to be used for ev~uating the effects of zolpidem, the animals were also given 4 injections of saline each one 10min before a Smin test session. The second, third and fourth injections were given immediately after the first, second and third tests. Thus, the entire procedure required 60 min from the first injection to the end of the fourth test. In the 10 min periods between tests, the rats were left in their home cages. Two days after this adaptation session a dose-response curve to zolpidem was established, using this cumulative dose procedure. The rats were injected with a dose of 0.1 mg/kg and 10 min later given a 5 min test on the FR 10 schedule.
695
Immediately after this test, they received a second injection, this time at the dose of 0.2 mg/kg, to give a total dose of 0.3 mg/kg. A third (0.7 mg/kg) and fourth (2.0 m&kg) injection was then given, following the same procedure, so that the total cumulative dose received was 3.0 mg/kg. The first dose-response investigation with zolpidem was carried out on a Friday. On the Saturday and Sunday, the rats were given 3.0 mg/kg zolpidem but were not tested. They were given 10 mg/kg at the end of the afternoon for the next 3 days and the dose was then increased to 30 mg/kg daily. After 13 daily administrations of this dose, the dose-response curve was reestablished on the immediately following day (day 20), using the procedure described above, with doses of 0.3, 0.7, 2.0 and 7.0mg/kg. Drugs. Zolpidem was dissolved in physiological saline to give an injection volume of 2ml/kg. RESULTS Experiment 1
The effects of acute admi~stration of triazolam on the rates of FR 10 lever pressing are shown in Fig. 1. The figure shows that while the baseline rate of responding had not changed, the dose-response curve for triazolam was shifted to the right and flattened, following the period of repeated administration. A two-way ANOVA, using the 0.1, 0.3 and 1.0 mg/kg values showed a significant effect of dose (F = 12.1, df = 2,48, P < O.OOl),a significant effect of condition (pre- or post-chronic) (F = 7.2, df = 1,48, P < 0.01) and a significant interaction (F = 6.2, df = 2,48, P < 0.01). Individual comparisons with control values showed that doses of 0.3 and 1.Omg/kg of triazolam decreased response rates at the beginning of the experiment, whereas the effect of the 0.3 mgfkg dose was no longer statistically significant after chronic administration. There were also statistically significant differences between the response rates after the 0.3 and 1.0 mg/kg doses, before and after chronic treatment. The effects of CL 218,872, before and after chronic administration of 30 mg/kg are also shown in Fig. 1. It is clear that the two dose-response curves were almost identical. No tolerance had developed to the depressant effect of this compound. The difference in the development of tolerance with triazolam and CL 218,872 is also apparent in Fig. 2. This figure shows the rates of FR 10 responding during the period of 8 days when the animals were tested after administration of triazolam at 1.Omg/kg or CL 218,872 at 30 mg/kg. The two drugs produced similar decreases in response rates, on the first administration and this effect decreased to a much greater extent during the course of chronic administration of triazolam, than with CL 218,872. One-way ANOVA, using the response rates on the 8 days with triazolam showed a statistically significant effect of days (F = 3.8, df = 7,56, P c 0.01) but a similar analysis
D. J. SANGERand B. ZIWCO~IC
696
Tl?lAZOlAM
CL 218.872
I 0
n&kg
I
I
I
C
3.0
10
I 30
(i.p.)
Fig. 1. Effects of triazolam (left panel) and CL 218,872 (right panel) on rates of FR 10 responding in rats, before and after repeated administration. Triazolam was given at the dose of 1.Omg/kg,‘day for 10 days and CL 218,872 at the dose of 30mg/kg/day for the same period. Control values were taken from the sessions immediately preceding the session with drug. The values shown are means f SEM.
applied to CL 218,872 did not show a significant effect of repeated treatment (F = 1.6, df = 7,56), although Fig. 2 indicates that there was a trend for the effect of this compound to decrease. Experiment
2
The effects on zopiclone on rates of FR 10 responding, before and after the period of repeated administration of 30 mg/kg, are shown in Fig. 3. When first
0
100 -
administered, doses of zopiclone of 3.0, 10 and 30 mg/kg, produced very marked decreases in responding. However, when the dose-response curve was reestablished after the chronic treatment, it was necessary to increase the dose to lOOmg/kg before a decrease in responding occurred. Analysis of variance, using the response rates after the 1.0,3.0, 10
cL218.872 3om@6
i 75-
I
50 -
25 -
‘1
0-L C
1
” 2
3
” 4
5
6
’
I
7
8
SESSIONS Fig. 2. Rates of FR 10 responding in rats during the period of repeated administration of triazolam or CL 218,872. The rats were injected daily with the drug for 10 days and tested on 8 of these 10 days. Control values are from the session immediately preceding the start of repeated administration of drug. The values shown are means f SEM.
sdcmchNmk L
C
t
I
1.0
s.0
1
10
I
so
i
loo
Zopielonc me/kg (r.c.1
Fig. 3. Effects of zopiclone on rates of FR 10 responding in rats before and after repeated administration of a dose of 30 mg/kg/day for 10 days. Control values were taken from sessions immediately preceding the sessions with drug.
Tolerance to benzodiazcpine agonists 75
figure shows response rates during the 8 successive sessions, before which the 30 mg/kg dose was administered. Although this dose greatly reduced response rates after the first few administrations, response rates were not subsequently below control values. An ANOVA applied to the data from the 8 days with drug showed a statistically significant effect (F = 5.8, df = 8,45, P < 0.001).
50 d a i
Experiment 3
It 4
697
25
0
I
C
’
12
’
’
’
’
’
’
I
3
4
5
6
7
8
Fig. 4. Rates of FR 10 responding in rats during the period of repeated administration of zopiclone. The rats were injected with zopiclone (30mg/kg) for 10 days and were tested on 8 of these 10 days. Control values are from the session immediately preceding the first injection of the 30 mg/kg dose. and 30 mg/kg doses, showed a statistically significant effect of dose (F = 6.9, df= 3,40, P < O.OOl), a significant effect of condition (F = 14.5, df = 1,40, P < 0.001) and a significant interaction (F = 5.9, df = 3,40, P < 0.01). The very marked tolerance produced by zopiclone is also illustrated by the data presented in Fig. 4. This
This experiment investigated the effects of midazolam and zolpidem on response rates after chronic administration of saline or drug. Unlike the other experiments in this series, it involved a betweengroups rather than a within-animals design and chronic administration was not associated with behavioural testing. The results are shown in Fig. 5. Midazolam produced a dose-related decrease in responding but the dose-response curve was shifted to the right and somewhat flattened in the rats which had received previous repeated treatment with the drug. Analysis of variance applied to these data, showed a statistically significant effect of the dose of midazolam (F = 30.9, df= 3,40, P < 0.001) and a significant effect of pretreatment condition (F = 18.8, df = 1,40, P < 0.001). The interaction term did not reach an acceptable level of statistical significance, however. In contrast to the results obtained with midazolam, there was no difference between the effects of zolpidem in the 2 groups of rats (Fig. 5). Thus, pretreatment with a daily dose of 1.Omg/kg of zolpidem for 10 days, did not produce any tolerance to the
MOAZOlAM
ZOIPIDEM
mglkg
(w.)
Fig. 5. Effects of midazolam (left panel) and zolpidem (right panel) on rates of FR 10 responding in rats after repeated administration of saline or drug. Rats were given 10 daily injections of saline, midazolam (3.0mg/kg) or zolpidem (l.Omg/kg) and dose-response curves for midazolam and zolpidem were established during a 2-week period which began 1 week after the end of the period of chronic administration of drug. Control values were taken from the sessions immediately preceding drug sessions.
D. J. SANGER and B. ZWKOVIC
698
tration. Analysis of variance applied to the results obtained with the 3 doses given twice (0.3, 1.0, 3.0 mg/kg), showed a statistically significant effect of the dose of zolpidem (F = 48.5, df = 2,30, P < 0.~1) but no significant effect of condition (i.e. before vs after chronic treatment) and the interaction term also was not statistically significant. DISCUSSION
8
3
50-
H 4
25
-
OC
0.1
0.3
1.0
3.0
10
Cumulative zolpidem dose mglkg (i.p.)
Fig. 6. Effects of zolpidem on rates of FR 10 responding in rats at the beginning and at the end of a period of chronic adminis~tion of drug. Each dose-response curve was established on a single day, using a cumulative dose method, this was done on day 1 and on day 20. The period of chronic
treatment with zolpidem consisted of 3 days of 3 mg/kg (including day l), 3 days of lOmg/kg and 13 days of 30 mg/kg. With the exception of days 1 and 20, injections of zolpidem were given at the end of each afternoon. Control response rates were taken from the days immediately preceding the days when the dose-response curves were established.
decrease in rates of operant responding produced by this drug. The ANOVA produced a statistically signiticant effect of the dose of zolpidem (F = 78.5, df = 3,40, P c 0.001) but neither the effects of pretreatment condition nor the interaction term was significant. Experiment 4 In this ex~~ment a dose-response curve was established for cumulative doses of zolpidem in a group of rats on day 1 and day 20, after 19 days of chronic administration of relatively large doses of zolpidem (3-30 mg/kg/day). The chronic injection procedure did not produce any disruption of the behavioural baseline. Figure 6 shows the effects of zolpidem on day 1 and day 20 of chronic administration. It is clear that the drug produced dose-related decreases in response rates on both occasions. There appeared to be very little change in the sensitivity of the animals to the drug produced by the period of chronic adminis-
Many studies have reported that tolerance develops to some of the behavioural effects of benzodiazepines and that the behavioural depressant actions of these drugs decline rapidly with repeated administration. Very little research, however, has investigated whether newer, non-benz~iazepine ligands for similar binding sites also give rise to tolerance. In a previous experiment it was found that, while tolerance developed rapidly to the decrease in rates of operant responding produced by midazolam, there was relatively little tolerance to the similar effect of zolpidem (Sanger and Zivkovic, 1987). In addition, there was little evidence for cross-tolerance between the two drugs. It has also been reported that less tolerance occurs to the anticonvulsant and locomotor depressant actions of zolpidem in mice, than is observed with benzodiazepines (Cox, Boyland, Gent and Feely, 1988; Perrault, Morel, Sanger and Zivkovic, 199Ob). The lack of tolerance with repeated administration of zoipidem was confirmed in the present set of experiments. In a between-groups design experiment (Experiment 3), in which marked tolerance developed after IO days of administration of midazolam, there was no tolerance to the depressant effect of zolpidem. The doses of midazolam and zolpidem, given during the period of repeated treatment (3.0 mg/kg midazolam, 1.Omg/kg zolpidem) were, however, equivalent in that they produced similar decreases in responding in animals which had not received chronic treatment with drugs. Repeated treatment with zolpidem also produced no tolerance to the depressant effect of the drug in another experiment (Experiment 4) where larger doses of zolpidem (up to 30 mg/kg/day) were given for a longer period (19 days). The lack of tolerance to the pharmacological effects of zolpidem appears, therefore, not to be restricted to the particular conditions of the previous study (Sanger and Zivkovic, 1987). In most of the present experiments, dose-response curves were reestablished after chronic treatment with drug had ceased. It is possible therefore that this method might unde~stimate the degree of tolerance if recovery had occurred. However, this does not account for the lack of tolerance with zolpidem as, in Experiment 4, the dose-response curve for zolpidem was reestablished on the day immediately after the end of chronic treatment and no significant tolerance was observed.
Tolerance to henzodiazepine agonists In contrast to the results obtained with zolpidem, clear tolerance occurred with midazolam, as previously described (Griffiths and Goudie, 1987; Sanger and Zivkovic, 1987). The observation that tolerance occurred in animals which were never tested in the drugged state, during the period of repeated administration (Experiment 3) is also consistent with the findings of Griffiths and Goudie (1987) and thus provides no evidence that the tolerance observed was learned or behaviourally-contingent tolerance (Corfield-Summer and Stolerman, 1978; Wolgin, 1989). Other experiments, however, have provided evidence for the importance of associative factors in the development of tolerance to the effects of benzodiazepines, including midazolam (King et al., 1987; Herberg and Montgomery, 1987; File, 1982; Greeley and Cappell, 1985). It should be noted that, as pointed out by Goudie (1986) it is possible that tolerance to the depressant effects of benzodiazepines could occur through the development of behavioural strategies, which are not task or situation specific. Thus, animals repeatedly given drugs in their home cages might acquire compensatory responses which would generalize to test situations, such as those involving operant responding. Indeed, several aspects of the present results seem consistent with the possibility that tolerance was not purely pharmacological. Tolerance to the effects of triazolam and zopiclone developed quite rapidly (Figs 2 and 4) and the tolerance observed previously with midazolam was maintained for up to 6 weeks after repeated treatment had ceased (Sanger and Zivkovic, 1987). The present results and those of other studies (Griffiths and Goudie, 1987), also show that tolerance to benzodiazepines involves not only a shift to the right but also a flattening of the dose-response curves, which might be more easily explained in terms of behavioural rather than pharmacological tolerance. Tolerance, similar to that observed with midazolam, was found in the present study with triazolam and zopiclone but not with CL 218,872, although there was a non-significant trend for the effect of CL 218,872 to decrease during repeated administration. Triazolam, like midazolam, is a benzodiazepine with a relatively short duration of action. Zopiclone, a cyclopyrrolone, is a hypnotic drug which binds to benzodiazepine sites and has a pharmacological profile, similar to that of the benzodiazepines (Julou, Blanchard and Dreyfus, 1985). Few data are available on the behavioural effects of zopiclone after repeated administration. However, Yamamoto and colleagues (Yamamoto, Kumasaka and Ueki, 1989) described the results of a drug-discrimination study in rats, in which the training dose of zopiclone initially produced an approx 30% decrease in response rates. Tolerance appeared to have occurred to this effect after 4 or 5 administrations of drugs. This result is thus similar to the present findings with zopiclone but in contrast to the results of drug-discrimination
699
experiments with zolpidem, where response rates after administration of drug continued to be below control rates over many months @anger and Zivkovic, 1986). Other experiments, involving repeated administration of zopiclone to mice, found no tolerance to its effects on pentylenetetrazole-induced convulsions and footshock-inducing fighting in mice (Julou, Bardone, Blanchard, Garrett and Stutzman, 1983). However, in these studies little tolerance development with chlordiazepoxide, tested under similar conditions. The present finding, that little or no tolerance occurred to the decrease in responding produced by CL 218,872 after administration for 10 days, is consistent with previous results reported by McElroy et al. (1985). These workers found that tolerance had developed to the decrease in locomotor activity of rats produced by a dose of chlordiazepoxide after 6 days of once-daily injection of the same dose. However, there was no tolerance to the decrease in locomotion produced by a dose of CL 218,872, tested under similar conditions. Both CL 218,872 and zolpidem show selectivity for the GABA, receptor, carrying w, modulatory sites (Langer and Arbilla, 1988; Sieghart and Schlerka, 1991). Recent research in molecular biology has indicated that this subtype of receptor may correspond to a subpopulation of GABA, receptors, containing the IX, subunit (Pritchett and Seeburg, 1990). It is thus tempting to speculate that the lack of tolerance, observed with CL 218,872 and zolpidem, may be related to their receptor selectivity. Why this should be so is unclear at present. However, selective compounds might be expected to interact with a smaller proportion of GABAA receptors in the central nervous system. Indeed, it has been shown that, at doses producing equivalent decreases in exploratory behaviour in mice, the number of cortical w receptors occupied was smaller for zolpidem and CL 218,872 than for flunitrazepam (Benavides, Peny, Dubois, Perrault, Morel, Zivkovic and Scatton, 1988). It is possible, therefore, that at pharmacologically active doses, the proportion of receptors occupied by zolpidem and CL 218,872 is insufficient to bring into play the compensatory physiological mechanisms, presumably involved in the development of tolerance. Whatever the mechanisms underlying the present results, these findings demonstrate that the extent to which tolerance develops to behavioural depressant effects, differs among different ligands for the GABA,-linked o receptors. Development of tolerance may depend not only on pharmacological (i.e. dose) and pharmcokinetic (i.e. duration of action) factors but also on the manner in which different drugs interact with GABA, receptors. skilled technical assistance of Michele LePichon and Nathalie Toupin is gratefully acknowledged. Acknowledgements-The
700
D. J. SANGER and B. ZIVKOVIC REFERENCES
Benavides J., Peny B., Dubois A., Perrault G., Morel F., Zivkovic B. and Scatton B. (1988) In vivo interaction of zolpidem with central benzodiazepine (BZD) binding sites (as labelled by 3H-Ro 15-1788) in the mouse brain. Preferential affinity of zolpidem for the o, (BZD,) subtype. .I. Pharmac. exp. Ther. 245: 1033-1041. Corfield-Summer P. K. and Stolerman I. P. (1978) Behavioral tolerance. In: Contemporary Research in Behavioral Pharmacology (Blackman D. E. and Sanger D. J., Eds), pp. 391448. Plenum Press, New York. Cox A., Boyland P., Gent J. P. and Freely M. (1988) An investigation of anti-convulsant tolerance to zolpidem in mice. Br. J. Pharmac. 95: 667P. Depoortere H., Zivkovic B., Lloyd K. G., Sanger D. J., Perrault G.. Lancer S. Z. and Bartholini G. (1986) Zolpidem a no& non-benzodiazepine hypnotic. I. Neuropharmacological and behavioral effects. J. Pharmat. exp. Ther. 231: 649-658.
File S. E. (1981) Rapid development of tolerance to the sedative effects of lorazepam and triazolam in rats. Psychopharmacology 73: 240-245.
File S. E. (1982) Development and retention of tolerance to the sedative effects of chlordiazepoxide: role of apparatus cues. Eur. J. Pharmac. 81: 637643. File S. E. (1985) Tolerance to the behavioral actions of benzodiazepines. Neurosci. Biobehav. Rev. 9: 113-121. Gardner C. R. (1989) Internretation of the behavioral effects of benzodiazepink receptor ligands. Drugs Fufure 14: 51-67. Goudie A. J. (1986) Behavioural influences on benzodiazepine tolerance: When do they occur and what do they mean? Psychopharmacology 9% 546547. Greelev J. and Cannel1 H. (1985) Associative control of tolerance to the* sedative. and’ hypothermic effects of chlordiazepoxide. Psychopharmacology 86: 487-493. Griffiths J. W. and Goudie A. J. (1986) Analysis of the role of drug predictive environmental stimuli in tolerance of the hypothermic effects of the benzodiazepine midazolam. Psychopharmacology 901 513-521.
Griffiths J. W. and Goudie A. J. (1987) Analysis of the role of behavioural factors in the development of tolerance to the benzodiazepine midazolam. Neuropharmacology 26: 201-209.
Herberg L. J. and Montgomery A. M. J. (1987) Learnt tolerance to sedative effects of chlordiazepoxide on selfstimulation performance, but no tolerance to facilitatory effects after 80 days. Psychopharmacology 93: 214-217. Julou L., Blanchard J. C. and Dreyfus J. F. (1985) Pharmacological and clinical studies of cyclopyrrolones: zopiclone and suriclone. Pharmac. Biochem. Behav. 23: 653-659.
Julou L., Bardone M. C., Blanchard J. C., Garrett C. and Stutzman J. M. (1983) Pharmacological studies on zopiclone. Pharmacology 27: Suppl. 2, 46-58. King D. A., Bouton M. E. and Musty R. E. (1987) Associative control of tolerance to the sedative effects of a short-acting benzodiazepine. Behav. Neurosci. 101: 104-l 14. Langer S. Z. and Arbilla S. (1988) Limitations of the benzodiazepine receptor nomenclature: a proposal for a
pharmacological classification as omega receptor subtypes. Fundam. clin. Pharmac. 2: 159-170. Langer S. Z., Arbilla S., Scatton B., Niddam R. and Dubois A. (1988) Receptors involved in the mechanism of action of zolpidem. In: Imidazopyridines in Sleep Disorders (Sauvanet J. P., Langer S. Z. and Morselli P. L., Eds), pp. 55-70. Raven Press, New York. Linna A. S.. Couwt J.. Greenblatt E. N., Klenner C. A. &id Beer B. (19f9) A synthetic non-benzodiazepine ligand for benzodiazepine receptors: a probe for investigating neuronal substrates of anxiety. Pharmac. Biochem. Behav. 11: 99-106. Margules D. L. and Stein L. (1968) Increase of antianxiety activity and tolerance to behavioral depression during chronic administration of oxazepam. Psychopharmacologia 13: 74-80. McElroy J. F., Fleming R. L. and Feldman R. S. (1985) A comparison between chlordiazepoxide and CL 218,872a synthetic non benzodiazepine ligand for benzodiazepine receptors on spontaneous locomotor activity in rats. Psyihopharmacology 85: 224226.
Perrault G., Morel E., Sanger D. J. and Zivkovic B. (1990a) Pharmacological profiles of four new hypnotics with different chemical structures: quazepami- brotizolam, zopiclone and zolpidem. Eur. J. Pharmac. 187: 487-494. Perrault G., Morel E., Sanger D. J. and Zivkovic B. (1990b) Differential development of tolerance to anticonvulsant and sedative effects of zolpidem, an imidazopyridine hypnotic, and benzodiazepine hypnotics. Fundam. clin. Pharmac. 4: 574.
Pritchett D. B. and Seeburg P. H. (1990) y-Aminobutyric acidA receptor a,-subunit creates novel type II benzodiazepine receptor pharmacology. J. Neurochem. 54: 1802-1804.
Sanger D. J. and Zivkovic B. (1986) The discriminative stimulus properties of zolpidem, a novel imidazopyridine hypnotic. Psychopharmacology 89: 317-322. Sanger D. J. and Zivkovic B. (1987) Investigation of the development of tolerance to the actions of zolpidem and midazolam. Neuropharmacoloav 26: 1513-1518. Sanger D. J., Perrauh G., Morel-E., Joly D. and Zivkovic B. (1987) The behavioral profile of zolpidem, a novel hypnotic drug of imidazopyridine structure. Physiol. Behav. 41: 235-240.
Sieghart W. and Schlerka W. (1991) Potency of several type I-benzodiazepine receptor ligands for inhibition of [)H]flunitrazepam binding in different rat brain tissues. Eur. J. Pharmac. 1W: 103-107. Squires R. F., Benson D. I., Braestrup C., Coupet J., Klepner C. A., Myers V. and Beer B. (1979) Some properties of brain specific benzodiazepine receptors: new evidence for multiple receptors. Pharmac. Biochem. Behav. 10: 825-830. Wolgin D. L. (1989) The role of instrumental learning in behavioral tolerance to drugs. In: Psychoactive Drugs. Tolerance and Sensitization (Goudie A. J. and EmmettOglesby M. W., Eds), pp. 17-l 14. Humana Press, Clifton, New Jersey. Yamamoto T., Kumasaka Y. and Ueki S. (1989) Behavioral analvsis of zoniclone on the basis of their discriminative stimulus properties in the rat. Jap. J. Pharmac. 51: 337-345.
002%3908/92 $5.00 + 0.00
Copyright 6 1992 Pergamon Press Ltd
DIFFERE~IAL DEVELOPME~ OF TOLERANCE TO THE DEPRESSANT EFFECTS OF BENZODIAZEPINE AND NON-BENZODIAZEPINE AGONISTS AT THE OMEGA (BZ) MODULATORY SITES OF GABA, RECEPTORS D. J. SANGER and B. ZIVKOVIC Synthelabo Recherche (L.E.R.S), 31 Ave P.V. Couturier, 92220 Bagneux, France (Accepted 31 January 1992)
Sunmmry-In a previous study, it was found that both the benzodiazepine hypnotic, midazolam, and the imid~opy~dine hypnotic, zolpidem, which has selective affinity for a sub-population of omega (benzodiazepine, BZ) modulatory sites of GABA,, receptors, produced similar decreases in rates of food-reinforced lever pressing in rats. However, during 10 days of repeated administration, marked tolerance developed to the depressant effect of midazolam but little tolerance developed with zolpidem. It was found in the present study that, with a within-subject design similar to that used previously, tolerance developed to the response rate-decreasing activity of the benzodiazepine, triazolam and the cyclopyrrolone, zopiclone but not to that of the triazolopyridazine, CL 218,872. In another experiment, using a between-groups design, tolerance developed to the effect of midazolam, even if the injections were not associated with daily test sessions, providing no evidence for a dNg~nvironment interaction. The lack of tolerance to zolpidem was confirmed in two experiments. Them was little indication of tolerance to the depressant effect of zolpidem, even after 19 days administration of daily doses, up to 30 mg/kg, a dose 10 times greater than that which completely suppressed responding. These results showed that the extent to which tolerance develops to the effects of drugs with affinity for omega (BZ) modulatory sites can show wide variations which may be related to differences in mechanisms of action. Key wor&--tolerance,
benzodiazepines, zolpidem, topiclone, CL 218,872.
It has been well established that tolerance can develop during repeated administration of benzodiazepines. However, the degree of tolerance can depend on a variety of pharmacological, behavioural and environmental factors (File, 1985). In general, tolerance to the behavioural depressant effects of these drugs, which presumably represents their muscle relaxant and sedative effects, develops rapidly, as shown in studies of locomotor activity and instrumental responding (e.g. File, 1981; Margules and Stein, 1968; Griffiths and Goudie, 1987). In recent years a number of compounds have been described which, while being chemically distinct from the benzodiazepines, act at similar receptors. These compounds belong to several chemical series including /I-cat&lines, imidazopy~~nes, cyclopyrrolones and pyrazoloquinolines (Gardner, 1989). Although the pharmacological profiles of these agents have many features in common with benzodiazepines, significant differences have also been described in some cases. It is therefore important to investigate whether tolerance will develop to the pharmacological effects of these newer compounds. Zolpidem is an imidazopyridine with selectivity for the y-aminobutyric acid (GABA*) subtype of recep tor, endowed with the w, (BZ,) modulatory sites (Langer, Arbilla, Scatton, Niddam and Dubois,
1988). In extensive pharmacological studies, zolpidem has been found to have hypnotic activity of rapid onset and short duration and its ~havioural and neuropharmacological activities differ in certain respects from those of the benzodiazepines. Specifically, zolpidem shows hypnotic activity at doses smaller than those which induce anticonvulsant, anxiolyticlike and muscle relaxant effects, it has not been found to increase food and water intake and it produces only limited cross generalization with benzodiazepines in drug-discrimination studies (Depoortere, Zivkovic, Lloyd, Sanger, Perrault, Langer and Bartholini, 1986; Perrault, Morel, Sanger and Zivkovic, 1990a; Sanger, Perrault, Morel, Joly and Zivkovic, 1987). In a study of repeated a~inistration of zolpidem to rats trained to lever-press for food reward, it was found that no tolerance occurred to the zolpideminduced depression of responding during 10 daily injections. In contrast, very clear tolerance developed to the similar effect of midazolam (Sanger and Zivkovic, 1987). The purpose of the present series of experiments was to investigate the generality of this finding. A similar technique was used to study the effects of repeated admiuistration of not only midazolam and zolpidem but also the benzodiazepine, triazolam, the cyclopyrrolone, 693
694
D. J. SANGER and B. ZIVKOVIC
CL 218,872 zopiclone and triazolopyridazine, (3-methyl-6[3-(trifluoromethyl)phenyl]-1,2,4-triazolo[4,3#]pyridazine. The drug CL 218,872 is of particular interest because it, like zolpidem, is selective for the w, subtype of benzodiazepine receptors (Squires, Benson, Braestrup, Coupet, Klepner, Myers and Beer, 1979; Lippa, Coupe& Greenblat, Klepner and Beer, 1979) and has also been reported not to give rise to tolerance (McElroy, Fleming and Feldman, 1985). A second purpose of these studies was to investigate whether the tolerance, previously observed with midazolam (Sanger and Zivkovic, 1987) depended on the fact that the animals were repeatedly tested in the drugged state. Such behaviourally-augmented or learned tolerance has been found with benzodiazepines, including midazolam and both instrumental conditioning (Herberg and Montgomery, 1987) and classical conditioning (King, Bouton and Musty, 1987) processes have been implicated. However, Griffiths and Goudie (1986, 1987) found no evidence for a role for conditioning processes in tolerance to midazolam, METHODS
Experiment 1. Repeated administration of triazolam and CL 218,872 Animals. Subjects were 18 male Wistar rats (Charles River, France). They weighed 18&200g when obtained from the supplier and were allowed to grow during the experiment so that they weighed 400-500g during the period of administration of drugs. Rats were individually housed under standard laboratory conditions (lights on 07.0~19.00), with water freely available in the home cages. At the start of training the rats were deprived of food for 2 days, They were subsequently given a standard quantity of chow each evening and over the weekends, which allowed them to gain weight but maintained motivation for operant responding. This procedure is used routinely in the authors’ laboratory and maintains stable rates of lever pressing over many months. Apparatus. All experiments were carried out in standard operant test chambers (Campden Instruments). Reinforcement for lever pressing was provided by 45 mg food pellets (Bio-Serv). Procedure. After deprivation of food and habituation to the test chambers, the rats were trained to press the lever to the right of the food dispenser to obtain the 45 mg food pellets during daily 15 min sessions. At the beginning of training every response produced a food pellet. The schedule requirement was then gradually increased over approx 10 sessions until 10 lever presses were necessary for each pellet (fixedratio 10: FR 10). The rats were maintained on this FR 10 schedule, until the response rates became stable and during which time they were adapted to receiving intraperitoneal injections of saline, 20 min
before the sessions. Throughout the experiment, daily sessions remained at 15 min duration and rats were tested for 5 days each week. After training on the FR 10 schedule, the rats were divided into 2 groups to receive administration of triazolam and CL 218,872. A dose-response curve was established with 3 doses of each drug; injection of drug being given on a Tuesday and the following Friday and Tuesday. The doses were: triazolam 0.1, 0.3 and 1.0 mg/kg and CL 218,872 3.0, 10 and 30 mg/kg. The largest doses (i.e. triazolam 1.0 mg/kg and CL 218,872 30 mg/kg) were then administered daily for a further 9 days. The rats continued to be tested after each injection, except during the weekend (i.e. days 5 and 6) when they received injections of drug but were not tested. The period of repeated injections finished on Thursday and the rats were injected with saline on the following Friday and Monday. The dose-response curves for triazolam and CL 218,872 were then reestablished with injections being given on Tuesdays and Fridays. Drugs. Triazolam and CL 218,872 were prepared in physiological saline. The drugs were synthesized at the Chemistry Department, Synthelabo Recherche and all doses are expressed in terms of the bases. Injection volume was 2 ml/kg. Data analysis. Values for control response rates were obtained from the days preceding days of injection of drug. The results were analysed separately for the 2 groups of rats, using analysis of variance (ANOVA) and Duncan’s test. Experiment 2. Repeated a~inistration
of zopicione
AnimaZ~. Subjects were 6 male Wistar rats, maintained under conditions similar to those described above. Procedure. The procedure was similar to that of Experiment 1, except for several modifications. Rats were trained to lever press for food pellets on an FR 10 schedule during daily 5 min sessions and adapted to receiving subcutaneous injections of physiological saline, 30 min before the sessions. After a stable baseline had been established, the rats received injections of zopiclone at doses of 1.0, 3.0, 10 and 30 mg/kg (s.c.), 30 min before the sessions. The injections of drug were given at intervals of several days, the first on a Friday and the others on the following Tuesday, Friday and Tuesday. The rats were then injected with saline on the Wednesday after the last acute injection and a period of 9 further days of injection of the largest dose began on the next day. The rats were injected but not tested over the weekend. The period of repeated injection of zopiclone finished on a Friday and the rats were injected with saline on the following Monday. The dose-response curve was reestablished over the following 3 weeks, injections of drug continuing to be given on Tuesdays and Fridays.
Tolerance to benzodiazepine agonists Drugs. Zopiclone was prepared in saline containing 2 drops of Tween 80, to give an injection volume of 2 ml/kg. Experiment 3. Non-contingent repeutedadministration of midazolam and zolpidem Animals. Subjects were 24 male Wistar rats, maintained under conditions similar to those described above. Fro~edure. Rats were trained to lever press for food reward on an FR 10 schedule during daily 15 min sessions. They were also adapted to receiving subcutaneous injections of physiological saline, 15 min before the sessions and at the end of each afternoon (approx 6 hr after testing). When stable baselines of responding were established, the rats were divided into 4 groups of 6. One group was then given 10 daily injections of midazolam (3.0 mg/kg, s.c.) at the end of each afternoon. The second group received similar treatment with a dose of zolpidem (1.0 mg/kg, s.c.) and the other 2 groups received injections of saline. The period of repeated injection of drug began on a Monday and continued over a weekend, when the rats did not receive behavioural testing. After chronic administration of midazolam, zolpidem or saline, the rats were tested but not injected with drugs for 1 week. Dose-response curves were then established to midazolam in the midazolam-treated rats and one group of saline-treated rats and to zolpidem in the zolpidem-treated and the other saline group. The doses of midazolam were 0.3, 1.0, 3.0 and lOmg/kg (s.c.) and those of zolpidem 0.1, 0.3, 1.0 and 3.0 mg/kg (s.c.), injected 15 min before the sessions. Drugs were injected on Tuesdays and Fridays, with saline being administered on the intervening days. Drugs. Zolpidem and midazolam were dissolved in physiological saline, to give injection volumes of 2 ml/kg. Experiment 4. Chronic large doses of zolpidem Animals. Subjects were 6 male Wistar rats, maintained under conditions similar to those described above. Procedure. Rats were trained to lever press on an FR 10 schedule of food reward during daily 15 min sessions. They were also adapted to intraperitoneal injections of physiological saline. Because a cumulative dose procedure was to be used for ev~uating the effects of zolpidem, the animals were also given 4 injections of saline each one 10min before a Smin test session. The second, third and fourth injections were given immediately after the first, second and third tests. Thus, the entire procedure required 60 min from the first injection to the end of the fourth test. In the 10 min periods between tests, the rats were left in their home cages. Two days after this adaptation session a dose-response curve to zolpidem was established, using this cumulative dose procedure. The rats were injected with a dose of 0.1 mg/kg and 10 min later given a 5 min test on the FR 10 schedule.
695
Immediately after this test, they received a second injection, this time at the dose of 0.2 mg/kg, to give a total dose of 0.3 mg/kg. A third (0.7 mg/kg) and fourth (2.0 m&kg) injection was then given, following the same procedure, so that the total cumulative dose received was 3.0 mg/kg. The first dose-response investigation with zolpidem was carried out on a Friday. On the Saturday and Sunday, the rats were given 3.0 mg/kg zolpidem but were not tested. They were given 10 mg/kg at the end of the afternoon for the next 3 days and the dose was then increased to 30 mg/kg daily. After 13 daily administrations of this dose, the dose-response curve was reestablished on the immediately following day (day 20), using the procedure described above, with doses of 0.3, 0.7, 2.0 and 7.0mg/kg. Drugs. Zolpidem was dissolved in physiological saline to give an injection volume of 2ml/kg. RESULTS Experiment 1
The effects of acute admi~stration of triazolam on the rates of FR 10 lever pressing are shown in Fig. 1. The figure shows that while the baseline rate of responding had not changed, the dose-response curve for triazolam was shifted to the right and flattened, following the period of repeated administration. A two-way ANOVA, using the 0.1, 0.3 and 1.0 mg/kg values showed a significant effect of dose (F = 12.1, df = 2,48, P < O.OOl),a significant effect of condition (pre- or post-chronic) (F = 7.2, df = 1,48, P < 0.01) and a significant interaction (F = 6.2, df = 2,48, P < 0.01). Individual comparisons with control values showed that doses of 0.3 and 1.Omg/kg of triazolam decreased response rates at the beginning of the experiment, whereas the effect of the 0.3 mgfkg dose was no longer statistically significant after chronic administration. There were also statistically significant differences between the response rates after the 0.3 and 1.0 mg/kg doses, before and after chronic treatment. The effects of CL 218,872, before and after chronic administration of 30 mg/kg are also shown in Fig. 1. It is clear that the two dose-response curves were almost identical. No tolerance had developed to the depressant effect of this compound. The difference in the development of tolerance with triazolam and CL 218,872 is also apparent in Fig. 2. This figure shows the rates of FR 10 responding during the period of 8 days when the animals were tested after administration of triazolam at 1.Omg/kg or CL 218,872 at 30 mg/kg. The two drugs produced similar decreases in response rates, on the first administration and this effect decreased to a much greater extent during the course of chronic administration of triazolam, than with CL 218,872. One-way ANOVA, using the response rates on the 8 days with triazolam showed a statistically significant effect of days (F = 3.8, df = 7,56, P c 0.01) but a similar analysis
D. J. SANGERand B. ZIWCO~IC
696
Tl?lAZOlAM
CL 218.872
I 0
n&kg
I
I
I
C
3.0
10
I 30
(i.p.)
Fig. 1. Effects of triazolam (left panel) and CL 218,872 (right panel) on rates of FR 10 responding in rats, before and after repeated administration. Triazolam was given at the dose of 1.Omg/kg,‘day for 10 days and CL 218,872 at the dose of 30mg/kg/day for the same period. Control values were taken from the sessions immediately preceding the session with drug. The values shown are means f SEM.
applied to CL 218,872 did not show a significant effect of repeated treatment (F = 1.6, df = 7,56), although Fig. 2 indicates that there was a trend for the effect of this compound to decrease. Experiment
2
The effects on zopiclone on rates of FR 10 responding, before and after the period of repeated administration of 30 mg/kg, are shown in Fig. 3. When first
0
100 -
administered, doses of zopiclone of 3.0, 10 and 30 mg/kg, produced very marked decreases in responding. However, when the dose-response curve was reestablished after the chronic treatment, it was necessary to increase the dose to lOOmg/kg before a decrease in responding occurred. Analysis of variance, using the response rates after the 1.0,3.0, 10
cL218.872 3om@6
i 75-
I
50 -
25 -
‘1
0-L C
1
” 2
3
” 4
5
6
’
I
7
8
SESSIONS Fig. 2. Rates of FR 10 responding in rats during the period of repeated administration of triazolam or CL 218,872. The rats were injected daily with the drug for 10 days and tested on 8 of these 10 days. Control values are from the session immediately preceding the start of repeated administration of drug. The values shown are means f SEM.
sdcmchNmk L
C
t
I
1.0
s.0
1
10
I
so
i
loo
Zopielonc me/kg (r.c.1
Fig. 3. Effects of zopiclone on rates of FR 10 responding in rats before and after repeated administration of a dose of 30 mg/kg/day for 10 days. Control values were taken from sessions immediately preceding the sessions with drug.
Tolerance to benzodiazcpine agonists 75
figure shows response rates during the 8 successive sessions, before which the 30 mg/kg dose was administered. Although this dose greatly reduced response rates after the first few administrations, response rates were not subsequently below control values. An ANOVA applied to the data from the 8 days with drug showed a statistically significant effect (F = 5.8, df = 8,45, P < 0.001).
50 d a i
Experiment 3
It 4
697
25
0
I
C
’
12
’
’
’
’
’
’
I
3
4
5
6
7
8
Fig. 4. Rates of FR 10 responding in rats during the period of repeated administration of zopiclone. The rats were injected with zopiclone (30mg/kg) for 10 days and were tested on 8 of these 10 days. Control values are from the session immediately preceding the first injection of the 30 mg/kg dose. and 30 mg/kg doses, showed a statistically significant effect of dose (F = 6.9, df= 3,40, P < O.OOl), a significant effect of condition (F = 14.5, df = 1,40, P < 0.001) and a significant interaction (F = 5.9, df = 3,40, P < 0.01). The very marked tolerance produced by zopiclone is also illustrated by the data presented in Fig. 4. This
This experiment investigated the effects of midazolam and zolpidem on response rates after chronic administration of saline or drug. Unlike the other experiments in this series, it involved a betweengroups rather than a within-animals design and chronic administration was not associated with behavioural testing. The results are shown in Fig. 5. Midazolam produced a dose-related decrease in responding but the dose-response curve was shifted to the right and somewhat flattened in the rats which had received previous repeated treatment with the drug. Analysis of variance applied to these data, showed a statistically significant effect of the dose of midazolam (F = 30.9, df= 3,40, P < 0.001) and a significant effect of pretreatment condition (F = 18.8, df = 1,40, P < 0.001). The interaction term did not reach an acceptable level of statistical significance, however. In contrast to the results obtained with midazolam, there was no difference between the effects of zolpidem in the 2 groups of rats (Fig. 5). Thus, pretreatment with a daily dose of 1.Omg/kg of zolpidem for 10 days, did not produce any tolerance to the
MOAZOlAM
ZOIPIDEM
mglkg
(w.)
Fig. 5. Effects of midazolam (left panel) and zolpidem (right panel) on rates of FR 10 responding in rats after repeated administration of saline or drug. Rats were given 10 daily injections of saline, midazolam (3.0mg/kg) or zolpidem (l.Omg/kg) and dose-response curves for midazolam and zolpidem were established during a 2-week period which began 1 week after the end of the period of chronic administration of drug. Control values were taken from the sessions immediately preceding drug sessions.
D. J. SANGER and B. ZWKOVIC
698
tration. Analysis of variance applied to the results obtained with the 3 doses given twice (0.3, 1.0, 3.0 mg/kg), showed a statistically significant effect of the dose of zolpidem (F = 48.5, df = 2,30, P < 0.~1) but no significant effect of condition (i.e. before vs after chronic treatment) and the interaction term also was not statistically significant. DISCUSSION
8
3
50-
H 4
25
-
OC
0.1
0.3
1.0
3.0
10
Cumulative zolpidem dose mglkg (i.p.)
Fig. 6. Effects of zolpidem on rates of FR 10 responding in rats at the beginning and at the end of a period of chronic adminis~tion of drug. Each dose-response curve was established on a single day, using a cumulative dose method, this was done on day 1 and on day 20. The period of chronic
treatment with zolpidem consisted of 3 days of 3 mg/kg (including day l), 3 days of lOmg/kg and 13 days of 30 mg/kg. With the exception of days 1 and 20, injections of zolpidem were given at the end of each afternoon. Control response rates were taken from the days immediately preceding the days when the dose-response curves were established.
decrease in rates of operant responding produced by this drug. The ANOVA produced a statistically signiticant effect of the dose of zolpidem (F = 78.5, df = 3,40, P c 0.001) but neither the effects of pretreatment condition nor the interaction term was significant. Experiment 4 In this ex~~ment a dose-response curve was established for cumulative doses of zolpidem in a group of rats on day 1 and day 20, after 19 days of chronic administration of relatively large doses of zolpidem (3-30 mg/kg/day). The chronic injection procedure did not produce any disruption of the behavioural baseline. Figure 6 shows the effects of zolpidem on day 1 and day 20 of chronic administration. It is clear that the drug produced dose-related decreases in response rates on both occasions. There appeared to be very little change in the sensitivity of the animals to the drug produced by the period of chronic adminis-
Many studies have reported that tolerance develops to some of the behavioural effects of benzodiazepines and that the behavioural depressant actions of these drugs decline rapidly with repeated administration. Very little research, however, has investigated whether newer, non-benz~iazepine ligands for similar binding sites also give rise to tolerance. In a previous experiment it was found that, while tolerance developed rapidly to the decrease in rates of operant responding produced by midazolam, there was relatively little tolerance to the similar effect of zolpidem (Sanger and Zivkovic, 1987). In addition, there was little evidence for cross-tolerance between the two drugs. It has also been reported that less tolerance occurs to the anticonvulsant and locomotor depressant actions of zolpidem in mice, than is observed with benzodiazepines (Cox, Boyland, Gent and Feely, 1988; Perrault, Morel, Sanger and Zivkovic, 199Ob). The lack of tolerance with repeated administration of zoipidem was confirmed in the present set of experiments. In a between-groups design experiment (Experiment 3), in which marked tolerance developed after IO days of administration of midazolam, there was no tolerance to the depressant effect of zolpidem. The doses of midazolam and zolpidem, given during the period of repeated treatment (3.0 mg/kg midazolam, 1.Omg/kg zolpidem) were, however, equivalent in that they produced similar decreases in responding in animals which had not received chronic treatment with drugs. Repeated treatment with zolpidem also produced no tolerance to the depressant effect of the drug in another experiment (Experiment 4) where larger doses of zolpidem (up to 30 mg/kg/day) were given for a longer period (19 days). The lack of tolerance to the pharmacological effects of zolpidem appears, therefore, not to be restricted to the particular conditions of the previous study (Sanger and Zivkovic, 1987). In most of the present experiments, dose-response curves were reestablished after chronic treatment with drug had ceased. It is possible therefore that this method might unde~stimate the degree of tolerance if recovery had occurred. However, this does not account for the lack of tolerance with zolpidem as, in Experiment 4, the dose-response curve for zolpidem was reestablished on the day immediately after the end of chronic treatment and no significant tolerance was observed.
Tolerance to henzodiazepine agonists In contrast to the results obtained with zolpidem, clear tolerance occurred with midazolam, as previously described (Griffiths and Goudie, 1987; Sanger and Zivkovic, 1987). The observation that tolerance occurred in animals which were never tested in the drugged state, during the period of repeated administration (Experiment 3) is also consistent with the findings of Griffiths and Goudie (1987) and thus provides no evidence that the tolerance observed was learned or behaviourally-contingent tolerance (Corfield-Summer and Stolerman, 1978; Wolgin, 1989). Other experiments, however, have provided evidence for the importance of associative factors in the development of tolerance to the effects of benzodiazepines, including midazolam (King et al., 1987; Herberg and Montgomery, 1987; File, 1982; Greeley and Cappell, 1985). It should be noted that, as pointed out by Goudie (1986) it is possible that tolerance to the depressant effects of benzodiazepines could occur through the development of behavioural strategies, which are not task or situation specific. Thus, animals repeatedly given drugs in their home cages might acquire compensatory responses which would generalize to test situations, such as those involving operant responding. Indeed, several aspects of the present results seem consistent with the possibility that tolerance was not purely pharmacological. Tolerance to the effects of triazolam and zopiclone developed quite rapidly (Figs 2 and 4) and the tolerance observed previously with midazolam was maintained for up to 6 weeks after repeated treatment had ceased (Sanger and Zivkovic, 1987). The present results and those of other studies (Griffiths and Goudie, 1987), also show that tolerance to benzodiazepines involves not only a shift to the right but also a flattening of the dose-response curves, which might be more easily explained in terms of behavioural rather than pharmacological tolerance. Tolerance, similar to that observed with midazolam, was found in the present study with triazolam and zopiclone but not with CL 218,872, although there was a non-significant trend for the effect of CL 218,872 to decrease during repeated administration. Triazolam, like midazolam, is a benzodiazepine with a relatively short duration of action. Zopiclone, a cyclopyrrolone, is a hypnotic drug which binds to benzodiazepine sites and has a pharmacological profile, similar to that of the benzodiazepines (Julou, Blanchard and Dreyfus, 1985). Few data are available on the behavioural effects of zopiclone after repeated administration. However, Yamamoto and colleagues (Yamamoto, Kumasaka and Ueki, 1989) described the results of a drug-discrimination study in rats, in which the training dose of zopiclone initially produced an approx 30% decrease in response rates. Tolerance appeared to have occurred to this effect after 4 or 5 administrations of drugs. This result is thus similar to the present findings with zopiclone but in contrast to the results of drug-discrimination
699
experiments with zolpidem, where response rates after administration of drug continued to be below control rates over many months @anger and Zivkovic, 1986). Other experiments, involving repeated administration of zopiclone to mice, found no tolerance to its effects on pentylenetetrazole-induced convulsions and footshock-inducing fighting in mice (Julou, Bardone, Blanchard, Garrett and Stutzman, 1983). However, in these studies little tolerance development with chlordiazepoxide, tested under similar conditions. The present finding, that little or no tolerance occurred to the decrease in responding produced by CL 218,872 after administration for 10 days, is consistent with previous results reported by McElroy et al. (1985). These workers found that tolerance had developed to the decrease in locomotor activity of rats produced by a dose of chlordiazepoxide after 6 days of once-daily injection of the same dose. However, there was no tolerance to the decrease in locomotion produced by a dose of CL 218,872, tested under similar conditions. Both CL 218,872 and zolpidem show selectivity for the GABA, receptor, carrying w, modulatory sites (Langer and Arbilla, 1988; Sieghart and Schlerka, 1991). Recent research in molecular biology has indicated that this subtype of receptor may correspond to a subpopulation of GABA, receptors, containing the IX, subunit (Pritchett and Seeburg, 1990). It is thus tempting to speculate that the lack of tolerance, observed with CL 218,872 and zolpidem, may be related to their receptor selectivity. Why this should be so is unclear at present. However, selective compounds might be expected to interact with a smaller proportion of GABAA receptors in the central nervous system. Indeed, it has been shown that, at doses producing equivalent decreases in exploratory behaviour in mice, the number of cortical w receptors occupied was smaller for zolpidem and CL 218,872 than for flunitrazepam (Benavides, Peny, Dubois, Perrault, Morel, Zivkovic and Scatton, 1988). It is possible, therefore, that at pharmacologically active doses, the proportion of receptors occupied by zolpidem and CL 218,872 is insufficient to bring into play the compensatory physiological mechanisms, presumably involved in the development of tolerance. Whatever the mechanisms underlying the present results, these findings demonstrate that the extent to which tolerance develops to behavioural depressant effects, differs among different ligands for the GABA,-linked o receptors. Development of tolerance may depend not only on pharmacological (i.e. dose) and pharmcokinetic (i.e. duration of action) factors but also on the manner in which different drugs interact with GABA, receptors. skilled technical assistance of Michele LePichon and Nathalie Toupin is gratefully acknowledged. Acknowledgements-The
700
D. J. SANGER and B. ZIVKOVIC REFERENCES
Benavides J., Peny B., Dubois A., Perrault G., Morel F., Zivkovic B. and Scatton B. (1988) In vivo interaction of zolpidem with central benzodiazepine (BZD) binding sites (as labelled by 3H-Ro 15-1788) in the mouse brain. Preferential affinity of zolpidem for the o, (BZD,) subtype. .I. Pharmac. exp. Ther. 245: 1033-1041. Corfield-Summer P. K. and Stolerman I. P. (1978) Behavioral tolerance. In: Contemporary Research in Behavioral Pharmacology (Blackman D. E. and Sanger D. J., Eds), pp. 391448. Plenum Press, New York. Cox A., Boyland P., Gent J. P. and Freely M. (1988) An investigation of anti-convulsant tolerance to zolpidem in mice. Br. J. Pharmac. 95: 667P. Depoortere H., Zivkovic B., Lloyd K. G., Sanger D. J., Perrault G.. Lancer S. Z. and Bartholini G. (1986) Zolpidem a no& non-benzodiazepine hypnotic. I. Neuropharmacological and behavioral effects. J. Pharmat. exp. Ther. 231: 649-658.
File S. E. (1981) Rapid development of tolerance to the sedative effects of lorazepam and triazolam in rats. Psychopharmacology 73: 240-245.
File S. E. (1982) Development and retention of tolerance to the sedative effects of chlordiazepoxide: role of apparatus cues. Eur. J. Pharmac. 81: 637643. File S. E. (1985) Tolerance to the behavioral actions of benzodiazepines. Neurosci. Biobehav. Rev. 9: 113-121. Gardner C. R. (1989) Internretation of the behavioral effects of benzodiazepink receptor ligands. Drugs Fufure 14: 51-67. Goudie A. J. (1986) Behavioural influences on benzodiazepine tolerance: When do they occur and what do they mean? Psychopharmacology 9% 546547. Greelev J. and Cannel1 H. (1985) Associative control of tolerance to the* sedative. and’ hypothermic effects of chlordiazepoxide. Psychopharmacology 86: 487-493. Griffiths J. W. and Goudie A. J. (1986) Analysis of the role of drug predictive environmental stimuli in tolerance of the hypothermic effects of the benzodiazepine midazolam. Psychopharmacology 901 513-521.
Griffiths J. W. and Goudie A. J. (1987) Analysis of the role of behavioural factors in the development of tolerance to the benzodiazepine midazolam. Neuropharmacology 26: 201-209.
Herberg L. J. and Montgomery A. M. J. (1987) Learnt tolerance to sedative effects of chlordiazepoxide on selfstimulation performance, but no tolerance to facilitatory effects after 80 days. Psychopharmacology 93: 214-217. Julou L., Blanchard J. C. and Dreyfus J. F. (1985) Pharmacological and clinical studies of cyclopyrrolones: zopiclone and suriclone. Pharmac. Biochem. Behav. 23: 653-659.
Julou L., Bardone M. C., Blanchard J. C., Garrett C. and Stutzman J. M. (1983) Pharmacological studies on zopiclone. Pharmacology 27: Suppl. 2, 46-58. King D. A., Bouton M. E. and Musty R. E. (1987) Associative control of tolerance to the sedative effects of a short-acting benzodiazepine. Behav. Neurosci. 101: 104-l 14. Langer S. Z. and Arbilla S. (1988) Limitations of the benzodiazepine receptor nomenclature: a proposal for a
pharmacological classification as omega receptor subtypes. Fundam. clin. Pharmac. 2: 159-170. Langer S. Z., Arbilla S., Scatton B., Niddam R. and Dubois A. (1988) Receptors involved in the mechanism of action of zolpidem. In: Imidazopyridines in Sleep Disorders (Sauvanet J. P., Langer S. Z. and Morselli P. L., Eds), pp. 55-70. Raven Press, New York. Linna A. S.. Couwt J.. Greenblatt E. N., Klenner C. A. &id Beer B. (19f9) A synthetic non-benzodiazepine ligand for benzodiazepine receptors: a probe for investigating neuronal substrates of anxiety. Pharmac. Biochem. Behav. 11: 99-106. Margules D. L. and Stein L. (1968) Increase of antianxiety activity and tolerance to behavioral depression during chronic administration of oxazepam. Psychopharmacologia 13: 74-80. McElroy J. F., Fleming R. L. and Feldman R. S. (1985) A comparison between chlordiazepoxide and CL 218,872a synthetic non benzodiazepine ligand for benzodiazepine receptors on spontaneous locomotor activity in rats. Psyihopharmacology 85: 224226.
Perrault G., Morel E., Sanger D. J. and Zivkovic B. (1990a) Pharmacological profiles of four new hypnotics with different chemical structures: quazepami- brotizolam, zopiclone and zolpidem. Eur. J. Pharmac. 187: 487-494. Perrault G., Morel E., Sanger D. J. and Zivkovic B. (1990b) Differential development of tolerance to anticonvulsant and sedative effects of zolpidem, an imidazopyridine hypnotic, and benzodiazepine hypnotics. Fundam. clin. Pharmac. 4: 574.
Pritchett D. B. and Seeburg P. H. (1990) y-Aminobutyric acidA receptor a,-subunit creates novel type II benzodiazepine receptor pharmacology. J. Neurochem. 54: 1802-1804.
Sanger D. J. and Zivkovic B. (1986) The discriminative stimulus properties of zolpidem, a novel imidazopyridine hypnotic. Psychopharmacology 89: 317-322. Sanger D. J. and Zivkovic B. (1987) Investigation of the development of tolerance to the actions of zolpidem and midazolam. Neuropharmacoloav 26: 1513-1518. Sanger D. J., Perrauh G., Morel-E., Joly D. and Zivkovic B. (1987) The behavioral profile of zolpidem, a novel hypnotic drug of imidazopyridine structure. Physiol. Behav. 41: 235-240.
Sieghart W. and Schlerka W. (1991) Potency of several type I-benzodiazepine receptor ligands for inhibition of [)H]flunitrazepam binding in different rat brain tissues. Eur. J. Pharmac. 1W: 103-107. Squires R. F., Benson D. I., Braestrup C., Coupet J., Klepner C. A., Myers V. and Beer B. (1979) Some properties of brain specific benzodiazepine receptors: new evidence for multiple receptors. Pharmac. Biochem. Behav. 10: 825-830. Wolgin D. L. (1989) The role of instrumental learning in behavioral tolerance to drugs. In: Psychoactive Drugs. Tolerance and Sensitization (Goudie A. J. and EmmettOglesby M. W., Eds), pp. 17-l 14. Humana Press, Clifton, New Jersey. Yamamoto T., Kumasaka Y. and Ueki S. (1989) Behavioral analvsis of zoniclone on the basis of their discriminative stimulus properties in the rat. Jap. J. Pharmac. 51: 337-345.