Mechanisms of action of midazolam on expression of contextual fear in rats

L. Pain*,1,3, A. Launoy1, N. Fouquet2 and P. Oberling2

1 GRERCA, Hôpitaux Universitaires de Strasbourg and INSERM U405, Faculté de Médecine, Strasbourg, France. 2 Institut de Physiologie de la Faculté de Médecine and INSERM U405, Faculté de Médecine, Strasbourg, France 3 Present address: Department of Psychology, University of Vermont, Burlington, VT 05405, USA*Corresponding author: U405 Inserm, Faculté de Médecine, 11 rue Humann, F-67085 Strasbourg Cedex, France

Accepted for publication: May 14, 2002


    Abstract
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
Background. Midazolam may suppress conditioned fear after an aversive event by disrupting the memory trace formed during conditioning, by altering the emotional part of the aversive event, or by the combination of both effects. The purpose of the present study was to determine whether affective-related processes contribute to the amnesic-like effects of midazolam on aversive events.

Methods. The effects of acute administration of low doses of midazolam (0.37–3 mg kg–1) on fear conditioning (association between a neutral context and an aversive stimulus) and on innate anxiety in fearful surroundings were examined in rats. The effect of midazolam on the deleterious consequences of pre-exposure to the context (a non-aversive event) for subsequent fear conditioning was then compared with its effect on fear conditioning. The role of midazolam as an affective context was assessed by performing the testing phase under midazolam. Possible locomotor impairment or long-term effects of midazolam were controlled in additional experiments.

Results. Midazolam reduced both contextual fear conditioning and spontaneous fear. The deleterious effect of midazolam on pre-exposure to the context was of the same magnitude as its effect on the acquisition phase of fear conditioning. The effects of midazolam on both pre-exposure to the context and fear conditioning were unchanged when rats received a second injection of midazolam before the retention phase.

Conclusions. Low doses of midazolam that do not impair locomotion suppress conditioned fear to the context by acting on memory processes rather than on affective or anxiolytic processes.

Br J Anaesth 2002; 89: 614–21

Keywords: hypnotics, benzodiazepine, midazolam; memory; model, rat; psychological responses


    Introduction
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
Some forms of learning and memory may be preserved despite the decrease in cortical awareness that results from the use of sedative agents.14 In clinical practice, midazolam is largely used to alleviate anxiety and to prevent memory (implicit or explicit) of frightening experiences in patients. Benzodiazepines are known to impair the recall of information, especially when administered shortly before the occurrence of the event for which information is to be encoded.3 57 However, the mechanism by which benzodiazepines act on the recall of events that have an aversive component remains unclear. Interactions between affective and mnemonic processes of benzodiazepines cannot be excluded when dealing with the recall of aversive events.811 In this respect, midazolam has been shown to display not only amnesic effects1215 but also potent affective properties, such as anxiolysis16 and a pleasant affective state.17 Midazolam might therefore interact with the emotional component of an aversive event because of its affective properties. More data are thus necessary on the mechanisms by which midazolam acts on the learning and memory for aversive events.

During sedation for medical and/or surgical procedures, the patient experiences passively aversive information that needs to be discarded. This clinical situation could be responsible for contextual fear conditioning, a particular form of automatic associative learning15 that is relatively impenetrable to cognitive control.18 19 Contextual fear conditioning is defined as experiencing an aversive event in a specific context so as to result in conditioned fear to this context. It is a fundamental way in which humans and animals learn to associate context with biologically significant aversive events. This particular form of learning now appears to be deeply involved in the development of anxiety disorder after a stressful experience.20 Fear conditioning may lead to an emotional memory trace in the absence of conscious thought or reflection, the conditioned production of fear relying on unconscious processing mediated by subcortical neural circuitry centred on the amygdala.19 21

The aim of the present study was to examine the mechanisms of action of midazolam on Pavlovian (classical) contextual conditioned fear. We determined the extent to which low doses of midazolam abolished the contextual conditioned fear and the extent to which any reduction in fear conditioning resulted from anxiolytic-like attenuation of the emotional component of the aversive event, impairment of memory processes or a combination of both.


    Material and methods
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
All procedures were conducted in accordance with National Council Directive 87848 (October 19, 1989, Ministère de l’Agriculture et de la Forêt, Service Vétérinaire de la Santé et de la Protection animales) and European Communities Council Directive 86/609/EEC (November 24, 1987).

Animals
Three hundred and twenty-six naive male Long–Evans rats (CERJ, LeGenest, France), weighing 300–330 g at the beginning of the experiments, were used. All animals were housed two per cage in the colony room [mean temperature 22 (SD 2)°C], maintained on a 14-h light, 10-h dark cycle (light on at 07:00) and given food and water ad libitum. All animals were handled three times a week before each experiment.

Drugs
Midazolam (Roche, Basel, Switzerland) was dissolved in saline (0.9% sodium chloride) immediately before intraperitoneal injection at different doses (0, 0.37, 0.75, 1.5 and 3.0 mg kg–1) in a volume of 2 ml kg–1 body weight.

Passive avoidance procedures
The effect of midazolam on anxiety and learning was studied using a passive avoidance apparatus that consisted of a cage with two adjacent compartments, one bright and large and the other dark and small. The bright compartment had white Plexiglas walls, measured 46 cm long, 42 cm wide and 44 cm high and was illuminated by a 25 W bulb fixed on the wall opposite the entrance to the dark box. This latter compartment had black Plexiglas walls and roof and measured 30 cm long, 15 cm wide and 15 cm high. The two compartments were separated by a guillotine door (10 cm wide and 12 cm high). The floor was made of stainless steel bars (diameter 0.6 cm) 1.8 cm apart. Scrambled electric foot-shocks could be delivered through the floor of the dark compartment.

The general procedure consisted of two trials performed 24 h apart in the two-compartment apparatus, except when otherwise specified. Rats were placed in the large, bright compartment (fearful environment), from which they could escape into the adjacent small, dark compartment (neutral environment). Latency to enter the small dark compartment was used as the measure of spontaneous fear for the bright compartment. Latency to enter the dark compartment in rats previously given shocks in this compartment was used as the measure of learning and memory for this aversive event.22

Assessment of drug-induced locomotor impairment
The effect of midazolam on forced and exploratory locomotion was studying using a rotarod apparatus and an ‘activity cage’ respectively.23 Briefly, the rotarod apparatus consisted of an elevated (40 cm high) rotating (2 r.p.m.) bar (40 cm long, 6 cm in diameter), divided into three 11 cm long sections by four perpendicular disks (diameter 50 cm). The activity cage (45x30x30 cm) had an infrared detector (IPR124; Talco, Paris, France) placed behind a Fresnel lens and located in the roof of the cage, with which it was possible to monitor animal movement in the different sections of the cage. The signal was fed into a computer that totalled all horizontal movements (i.e. the total number of crossings from one section to the other) during periods of 5 min.

Study design
We studied the acute effects of low doses of midazolam on contextual fear conditioning, as assessed by Pavlovian (classical) fear conditioning (experiment 1) and on spontaneous escape from the innate fearful surrounding (experiment 2). [In a Pavlovian (classical) fear conditioning procedure, an initially neutral stimulus (the so-called conditioned stimulus, CS), either a discrete signal such as a tone or a light, or the experimental surrounding itself, is paired with an aversive stimulus such as a brief electrical foot-shock (the so-called unconditioned stimulus, US). When exposed later on to the CS, normal animals exhibit conditioned fear reactions, such as block of ongoing behaviour, freezing and modification of escape or avoidance latency.] In our procedure of contextual fear conditioning, the animal passively experienced mild electric foot-shocks [unconditioned stimulus (US)] in a distinctive neutral context [conditioned stimulus (CS)]. The mnemonic and/or anxiolytic mechanisms of midazolam with regard to fear conditioning were then examined by comparing the effect of midazolam on the aversively loaded event (the CS–US association) and on a non-aversive experience (pre- exposure to the CS before fear conditioning) (experiment 3). Initial exposure to the CS (with no aversive stimulus) has a deleterious effect on the subsequent CS–US association.24 This CS-pre-exposure effect (otherwise called the latent inhibition phenomenon) is thought to depend upon information about the CS being retained at the time of pre-exposure to it.25 Finally, we tested the possibility that midazolam could act as an affective contextual stimulus during both pre-exposure to the CS and fear conditioning (experiment 4). Additional experiments (experiments 5 and 6) were conducted in order to control for locomotor impairment and/or possible long-lasting residual actions of midazolam.

Experiment 1. Effect of midazolam on fear conditioning
On day 1 (conditioning phase), rats were enclosed for 5 min in the small, dark compartment, where they were given two mild inescapable foot-shocks (effective intensity 0.1 mA, for 5 s, 20 ms on/140 ms off) delivered 120 and 210 s after the beginning of the trial. On day 2 (testing phase), rats were placed in the large, bright compartment and latency to enter the small dark compartment was recorded (cut-off time 900 s). Rats were given a single injection of midazolam 0.37, 0.75, 1.50 or 3.00 mg kg–1 or saline 15 min before phase 1 (10 rats/group).

Experiment 2. Effect of midazolam on spontaneous fear
This was conducted as Experiment 1 except that rats were not subjected to foot-shocks during day 1 and a single administration of midazolam 0.37, 0.75, 1.50 or 3.00 mg kg–1 or saline took place 15 min before the testing phase on day 2 (10 rats/group).

Experiment 3. Comparison of the effect of midazolam on aversive and non-aversive events
Rats of two groups were initially pre-exposed to the dark compartment, where they were enclosed for 30 min; no shock was delivered during this period. All rats were injected with midazolam 0 or 3 mg kg–1 15 min before this CS pre-exposure period (non-aversive event). The animals were then subjected to the conditioning phase of the fear conditioning procedure as in experiment 1, without additional treatment. Rats in two further groups were not pre-exposed to the dark compartment and were injected with midazolam 0 or 3 mg kg–1 15 min before the conditioning phase (aversive event) as in experiment 1. All animals in the four groups were subjected to the testing phase of the fear conditioning procedure, as in experiment 1, without additional treatment (12 rats/group).

Experiment 4. Contextual effect of midazolam
Rats were injected with saline or midazolam 3 mg kg–1 15 min before the scheduled experimental phase, as follows: On day 1, rats of two non pre-exposed groups were given midazolam (group 1) or saline (group 2) in their home cage. Rats of group 3 were given midazolam and pre-exposed for 30 min to the dark compartment; no shock was delivered during this period. Twenty-four hours later, the three groups of rats were then subjected to the conditioning and testing phases of the fear conditioning procedure, conducted as in experiment 1. All rats were injected with either saline (groups 1 and 3) or midazolam (group 2) before the conditioning phase. All animals in the three groups were given midazolam before the testing phase (14 rats/group).

Experiment 5. Delayed effect of midazolam
This experiment was conducted in order to control for a possible residual effect of midazolam (at the time of the acquisition and/or retention phases) 24 h after the injection. If observed, such an effect would hinder the interpretation of experiment 3 because rats from the pre-exposed group were given midazolam 24 h before their non-pre-exposed counterparts. All animals were subjected to the procedure described for experiment 1. All rats were injected with saline or midazolam, but at different times: 24 h before, 15 min before or 15 min after the conditioning phase (12 rats/group).

In each of the above experiments (1–5), the latency to enter (four paws) the dark compartment during the testing phase was recorded to the nearest second by an experimenter unaware of the treatment conditions.

Experiments 6A and B. Control experiments
These experiments were intended to control for the possibility that midazolam-induced locomotor impairment might have interfered with the results of the present study. In experiment 6A, 50 rats were trained in a forced locomotor activity task, using a rotarod apparatus. Rats were trained during two daily sessions until each of them reached the criterion of stability on the cylinder for at least 60 s. On the third day, rats were given midazolam 0.37, 0.75, 1.50 or 3.00 mg kg–1 or saline 15 min before testing. The numbers of animals staying on the cylinder at least during the first 30 s and the number of animals falling during the 60 s test session were recorded.

In experiment 6B, 50 rats were given midazolam 0.37, 0.75, 1.50 or 3.00 mg kg–1 or saline 15 min before being placed in the activity cage. Locomotor activity was recorded during a 5 min session test (10 rats/group).

Statistical analysis
For experiments 1–5, data were transformed (log10) to better approximate the normal distribution, which is necessary for the use of parametric statistics. The data were subjected to one-way analysis of variance (ANOVA) followed by post hoc tests when needed.


    Results
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
Experiment 1. Effect of midazolam on fear conditioning
Figure 1 shows the latency to enter the dark compartment in rats previously subjected to two shocks in the dark compartment according to the dose of midazolam they received before the conditioning phase. When placed in the large bright compartment, previously non-shocked rats usually enter the small dark compartment after about 10 s (at least 6 s). We observed increased latency (about 200 s) in control rats (saline) previously subjected to two shocks in the dark compartment, which was indicative of the fear-conditioned response. In rats given midazolam, the latency to enter the dark compartment decreased linearly to about 20 s as the dose of midazolam increased (test of linearity, r=–0.547; P<0.001; y=–0.302x+1.893). One-way ANOVA revealed an overall significant dose effect [F(4,45)=5.30; P<0.05]. Post hoc pairwise comparisons indicated that the doses of 1.5 and 3.0 mg kg–1 differed significantly from saline [Dunnett’s t(4,45)=3.33, t(4,45)=3.85; all P<0.05].



View larger version (15K):
[in this window]
[in a new window]
 
Fig 1 Midazolam-induced impairment of fear conditioning. Latency (log10) to enter the dark compartment (mean and SEM) during the testing phase in rats given midazolam 0–3 mg kg–1 15 min before the conditioning phase, during which they received two mild electric foot-shocks in the dark compartment. *P<0.05 vs control (Dunnett’s t test after ANOVA).

 
Experiment 2. Effect of midazolam on spontaneous fear
Figure 2 shows the latency to enter the dark compartment in previously non-shocked rats according to the dose of midazolam they received before being placed in the large bright compartment. Control rats given saline escaped from the large bright compartment after about 10 s, confirming their spontaneous aversion to this compartment. One-way ANOVA revealed a significant main effect [F(4, 45)=6.97; P<0.05]. Post hoc pairwise comparisons indicated that only the 3 mg kg–1 dose of midazolam increased escape latency significantly [Dunnett’s t(4,45)=4.15; P<0.05].



View larger version (12K):
[in this window]
[in a new window]
 
Fig 2 Midazolam-induced anxiolysis. Latency (log10) to enter the dark compartment (mean and SEM) during the testing phase in rats given midazolam 0–3 mg kg–1. No shocks were delivered in the dark compartment. *P<0.05 vs control (Dunnett’s t test after ANOVA).

 
Experiment 3. Comparison of the effect of midazolam on aversive and non-aversive events
In control rats not pre-exposed to the dark compartment before receiving two shocks in this compartment, the latency to enter this compartment during the retention phase was on average 400 s (Fig. 3). This value was reduced to about 60 s in rats given midazolam 3 mg kg–1 before the conditioning phase (aversive event), thus replicating the results obtained in experiment 1. Saline-injected rats given a 30 min pre-exposure to the dark compartment (non-aversive event) 24 h before the conditioning phase of the fear-conditioning procedure escaped from the large compartment within 100 s; this latency was increased to 480 s in rats given midazolam 3 mg kg–1 before pre-exposure. A 2x2 ANOVA (pre-exposure, dose) indicated a significant interaction [F(1, 44)=29.81; P<0.0001] but no main effects [both F(1,44)<1]. Post hoc pairwise comparisons using the Bonferroni test indicated that non-pre-exposed rats given saline differed from those given midazolam and from pre-exposed rats given saline. The performance of pre-exposed rats given midazolam differed from that of their saline counterparts and from those of non-pre-exposed rats given midazolam [all t(44)>3.37; all P<0.05].



View larger version (14K):
[in this window]
[in a new window]
 
Fig 3 Comparison of the effects of midazolam on aversive (fear conditioning) and non-aversive (pre-exposure to CS) events. Latency (log10) to enter the dark compartment (mean and SEM) in (left) non-pre-exposed rats injected with saline (white column) or midazolam (black column) 15 min before the conditioning phase and (right) rats pre-exposed to the dark compartment 15 min after saline (white column) or midazolam (black column). On day 1, non-pre-exposed rats stayed in their home cage, whereas pre-exposed rats were placed for 30 min in the dark compartment without shocks. The four groups were subjected to the conditioning and testing phases on days 2 and 3 respectively. *P<0.05 (Bonferroni test after ANOVA).

 
Experiment 4. Contextual effect of midazolam
In group 1 animals not pre-exposed to the dark compartment before the conditioning phase, the latency to enter this compartment during the retention phase (preceded by midazolam injection) was on average 300 s (Fig. 4). This value was reduced to about 30 s in rats of group 2 given midazolam before the conditioning phase, thus replicating the results obtained in experiments 1 and 3. In group 3 animals given midazolam before 30 min of pre-exposure to the dark compartment (without shock) before the fear-conditioning procedure, the latency to enter this compartment during the retention phase (preceded by midazolam injection) was on average 300 s. This latency was in the range of the values obtained in rats not pre-exposed. This replicated the results obtained in experiment 3. One-way ANOVA (group) revealed a significant main effect [F(2, 38)=24.22; P<0.0001]. Post hoc pairwise comparisons using the Bonferroni test indicated that only group 2 animals given midazolam 15 min before the conditioning phase exhibited a significant reduction in latency to move out of the bright compartment compared with groups 1 and 3 [t(38)=6.05 and 6.71 respectively; all P<0.05].



View larger version (21K):
[in this window]
[in a new window]
 
Fig 4 Midazolam did not induce state-dependent retrieval. Latency (log10) to enter the dark compartment (mean and SEM) during the testing phase of the fear-conditioning procedure in three groups of rats given midazolam 15 min before testing. The inset summarizes the experimental schedule of drug administration. S=saline; M3=midazolam 3 mg kg–1 (see Material and methods section). *P<0.05 (Bonferroni test after ANOVA).

 
Experiment 5. Delayed effect of midazolam
Whether the injection was given before or after the conditioning phase, rats given saline exhibited a similar latency to enter the dark compartment during the testing phase (Fig. 5). These values were in the range of those obtained in control rats in experiments 1 and 3. A 3x2 ANOVA (injection time, treatment) indicated a significant injection time effect [F(2, 66)=14.34; P<0.0001], no treatment effect [F(1, 66)=3.33] and a significant interaction [F(2, 66)=8.04; P<0.001]. Post hoc pairwise comparisons using the Bonferroni test indicated that only rats given midazolam 3 mg kg–1 15 min before the conditioning phase exhibited a significant reduction in latency to move out of the bright compartment compared with their associated controls [t(66)=4.20; P<0.05].



View larger version (18K):
[in this window]
[in a new window]
 
Fig 5 Temporal contiguity for midazolam effect on fear conditioning. Latency (log10) to enter the dark compartment (mean and SEM) during the testing phase for groups injected with saline (white columns) or midazolam 3 mg kg–1 (black columns) 24 h before, 15 min before or 15 min after the conditioning phase. *P<0.05 (Bonferroni test after ANOVA).

 
Experiments 6A and B. Control experiments
Rats given midazolam 0–3 mg kg–1 exhibited no alteration of performance on the rotarod test. All the animals stayed on the bar for more than 30 s; during the 60 s test, the numbers of falls were 0, 1, 0, 1 and 1 for the five groups respectively ({chi}2=2.09; not significant) (experiment 6A). The mean (SEM) activity scores were 229 (10), 220 (9), 244 (7), 226 (12) and 227 (11) units for rats receiving saline and midazolam 0.37, 0.75, 1.5 and 3 mg kg–1 respectively. One-way ANOVA (dose) showed no overall effect [F(4, 45)=0.85] (experiment 6B).


    Discussion
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
The present study provides evidence that midazolam impairs conditioned contextual fear conditioning by acting on pure mnemonic rather than emotional processes or a combination of the two. Such an effect was observed at low doses of midazolam (1.5 and 3.0 mg kg–1), at which spontaneous or forced motor activities were unaffected, the hypnotic dose being above 30 mg kg–1 intraperitoneally in our experimental conditions.17

Indeed, rats subjected to inescapable electric foot-shocks (unconditioned stimulus, US) in the dark side (contextual conditioned stimulus, CS) of a two-compartment box exhibited a clear increase in latency to enter this compartment on a subsequent session, an effect that reflects the acquisition of the CS–US association. When administered before the shock session, midazolam dose-dependently counteracted this effect (experiment 1). Such a disruption may rely on drug-induced anterograde amnesia, but could also have resulted from an anxiolytic effect by the time of acquisition. Low doses of midazolam increased the latency to escape from the bright compartment, which is indicative of an anxiolytic effect in rodents (experiment 2). This increase in latency cannot be attributed to any locomotor impairment, because spontaneous or forced motor activities were unaffected in the same range of doses (experiments 6A and B). Moreover, conditioned fear was altered (i.e. the latency to enter the dark compartment during the retention test was shortened) when midazolam was administered 15 min before the shock session but not when the injection took place 24 h before or 15 min after the conditioning phase (experiment 4). Therefore, residual activity of midazolam at the time of the retention test cannot account for the reduction in the fear reaction observed in experiment 1. However, one could argue that midazolam reduced the perception of the aversive unconditioned stimulus (shocks) by an antinociceptive effect. There is clear evidence in the literature that midazolam is totally devoid of any analgesic effect.26 27 Similarly, preliminary data obtained in our laboratory showed no effect of midazolam 3 mg kg–1 on the threshold of the shock intensity to elicit a jump in rats compared with saline administration.

Because learning an aversive effect occurs in an emotional frame, the reduction in conditioned fear of the context by midazolam may result from changes in the emotional/affective aspects of the situation at the learning stage, the mnemonic processes or the combination of both phenomena.8 10 Midazolam could have disrupted the establishment of the CS–US association through an anxiolytic-related reduction in the emotional part of this aversive event. If midazolam disrupts the CS–US association through mnemonic processes, it should exert this effect whether or not the event that is to be conditioned has an aversive component. The CS pre-exposure procedure (see Material and methods) was used to control for such a possibility (experiment 3). Thirty minutes of pre-exposure to the dark box (CS alone) before the CS–US association (conditioning phase) resulted in a reduction in the latency to enter this compartment during the retention phase. Pre-exposure to the stimulus that was to be conditioned had a deleterious effect on the subsequent acquisition and/or expression of the CS–US association. In rats given midazolam before CS pre-exposure, latency to enter the dark compartment during the retention phase of the fear conditioning procedure was lengthened to a value similar to that observed in control non-pre-exposed rats. Pre-exposure to the dark compartment was a non-aversive experience compared with the conditioning phase, during which rats received uncontrollable shocks. The administration of midazolam resulted in total suppression of the effect of CS pre-exposure on fear conditioning. This shows that the midazolam-induced alteration in fear conditioning was more likely to be accounted for by disruption of memory formation rather than by an anxiolytic-like reduction in the emotional component of the CS–US (shocks) association or by reduction in the perception of the aversive unconditioned stimulus. Another benzodiazepine, chloradiazepoxide, has been shown to suppress the CS pre-exposure effect on fear conditioning in rats, but such an effect was observed only at high sedative doses.28 29

However, it could be argued that midazolam acted as a context alone during both the pre-exposure and the conditioning phase.30 Because the retention phase was performed with no additional treatment in experiments 1 and 3, a drug-to-no-drug state change could account for the observed effects of midazolam on both the CS pre-exposure and fear conditioning. It has been reported regularly that drugs such as benzodiazepines might induce a state-dependent retrieval phenomenon that is seen only when the retention test is performed with drug.3133 Moreover, because midazolam displays affective properties (pleasant affective state),17 one could argue that midazolam acted as a pleasant contextual stimulus during the pre-exposure phase, changing a non-aversive experience (pre-exposure) to a pleasant experience. Such a shift of affective context between the pre-exposure and testing phases has been shown to impair the effect of pre-exposure on fear conditioning.34 In experiment 4, we found that the midazolam dose of 3 mg kg–1 totally suppressed the deleterious effect of CS pre-exposure on subsequent fear conditioning, even when the testing phase was performed in a ‘midazolam state’. This replicated the results obtained in experiment 3. At the same time, midazolam disrupted the fear conditioning in rats given midazolam before both the acquisition and the retention phase, as observed in experiments 1 and 3. These data show clearly that a shift of context or a state-dependent retrieval phenomenon cannot be responsible for the observed effects of midazolam on the CS pre-exposure effect and on fear conditioning.

One of the principal goals of using midazolam in clinical practice is to produce amnesia for untoward events.3537 In humans, it has been demonstrated that midazolam induces anterograde amnesia, depending on the dose.12 15 3840 However, the work mentioned above studied the memory (implicit or explicit) for non-emotionally loaded information (words, pictures) or the explicit recall of untoward events (pain, fear) in patients in whom a state of excessively deep sedation was often produced to obtain some-amnesic like effect. Midazolam-induced amnesia, independently of a reduction in awareness, has been described in particular forms of learning: memory (implicit or explicit) for non-emotionally loaded information,15 38 41 procedural learning in humans42 and operant learning in animals (in which the animal learns what occurred in the environment in response to its behaviour).26 43 These forms of learning do not reflect exactly the mechanism of learning in patients passively experiencing an aversive event. It was for this reason that we examined the effects of low doses of midazolam on Pavlovian fear conditioning in animals passively experiencing an aversive stimulus. Our findings in rats suggest that midazolam might suppress the subsequent conditioned fear resulting from an aversive event occurring under conscious sedation (when verbal contact is maintained in humans).

In conclusion, low doses of midazolam (less than 10% of the hypnotic dose) were able to prevent the expression of conditioned fear after an aversive event in rats. The main mechanism by which midazolam acts on such contextual fear conditioning relies on its memory effect. The anxiolytic effect of midazolam in a fearful context plays a negligible role compared with its effect on memory processes. The mnemonic processes (acquisition, storage or retrieval) involved in the disruption of contextual fear learning by midazolam are currently under investigation.


    Acknowledgements
 
We thank Lilianne Manning, John M. Pearce and Marie-Hélène Thiébot for their helpful comments during the preparation of this manuscript. Financial support for this research was provided by SFAR (contrat de recherche 1997), MENRT-INSERM 1999/2000 ‘Soutien aux Sciences du Vivant’ au sein de l’Institut Fédératif de Recherches Neurosciences 37 and Université Louis Pasteur ‘Appel à projets Exceptionnels 2000’. This work was presented in part at the IVth symposium on Memory and Awareness in Anaesthesia, London, July 1998.


    References
 Top
 Abstract
 Introduction
 Material and methods
 Results
 Discussion
 References
 
1 Tinnin L. Conscious forgetting and subconscious remembering of pain. J Clin Ethics 1994; 5: 151–152[ISI][Medline]

2 Andrade J. Learning during anaesthesia: a review. Br J Psychol 1995; 86: 479–506[ISI][Medline]

3 Ghoneim MM, Block RI. Learning and memory during general anesthesia: an update. Anesthesiology 1997; 87: 387–410[ISI][Medline]

4 Osterman JE, van der Kolb BA. Awareness during anesthesia and posttraumatic stress disorder. Gen Hosp Psychiatry 1998; 20: 274–81[ISI][Medline]

5 Ghoneim MM, Melwaldt SP. Benzodiazepines and human memory: review. Anesthesiology 1990; 72: 926–38[ISI][Medline]

6 Lister RG. The amnesic action of benzodiazepines in man. Neurosci Biobehav Rev 1985; 9: 87–94[ISI][Medline]

7 Thiébot MH. Some evidence for amnesic-like effects of benzodiazepines in animals. Neurosci Biobehav Rev 1985; 9: 95–100[ISI][Medline]

8 Borde N, Krazem A, Jaffard R, Beracochea DJ. Memory deficits following diazepam administration in mice: evidence for a time-dependent retrieval impairment. Psychobiology 1997; 25: 202–9[ISI]

9 Curran HV. Benzodiazepines, memory and mood: a review. Psychopharmacology 1991; 105: 1–8[ISI][Medline]

10 Blaney PH. Affect and memory: a review. Psychol Bull 1986; 99: 229–46[ISI][Medline]

11 Eich E. Searching for mood dependent memory. Psychol Sci 1995; 6: 67–75[ISI]

12 Veselis B, Reinsel R, Alagesan R, Heino R, Bedford R. The EEG as a monitor of midazolam amnesia: changes in power and topography as a function of amnesic state. Anesthesiology 1991; 74: 866–74[ISI][Medline]

13 Polster MR, McCarthy RA, O’Sullivan G, Gray PA, Park GR. Midazolam-induced amnesia: implications for the implicit/explicit memory distinction. Brain Cogn 1993; 22: 244–65[ISI][Medline]

14 Thomas-Anterion C, Koening O, Navez M, Laurent B. Midazolam effects on implicit and explicit memory processes in healthy subjects. Psychopharmacology 1999; 145: 139–43[ISI][Medline]

15 Hirshman E, Passanante A, Arndt J. The effect of midazolam on the modality-match effect in implicit memory. Brain Res Cogn Brain Res 1999; 7: 473–9[ISI][Medline]

16 Salonen M, Onaivi ES, Maze M. Dexmedetomidine synergism with midazolam in the elevated plus-maze test in rats. Psychopharmacology 1992; 108: 229–34[ISI][Medline]

17 Pain L, Oberling P, Sandner G, Di Scala G. Effect of midazolam on propofol-induced pleasant affective state as assessed by place conditioning paradigm in rats. Anesthesiology 1997; 87: 935–44[ISI][Medline]

18 Pearce JM, Bouton ME. Theories of associative learning in animals. Annu Rev Psychol 2001; 52: 111–39[ISI][Medline]

19 Ohman A, Mineka S. Fears, phobias, and preparedness: toward an evolved module of fear and fear learning. Psychol Rev 2001; 3: 483–522

20 Bouton ME, Mineka S, Barlow DH. A modern learning theory perspective on the etiology of panic disorder. Psychol Rev 2001; 108: 4–32[ISI][Medline]

21 Davis M. The role of the amygdala in conditioned fear. In: Aggleton JP, ed. The Amygdala: Neurobiological Aspects of Emotion, Memory and Mental Dysfunction. New York: Wiley-Liss 1992; 255–305

22 Mora PD, Fouquet N, Oberling P, Gobaille S, Graeff FG, Sandner G. A neurotoxic lesion of serotoninergic neurons using 5,7-dihydroxytryptamine does not disrupt latent inhibition in paradigms sensitive to low doses of amphetamine. Behav Brain Res 1999; 100: 167–75[ISI][Medline]

23 Pain L, Oberling P, Launoy A, Di Scala G. Effect of non sedative doses of propofol on an innate anxiogenic situation in rats. Anesthesiology 1999; 90: 191–6[ISI][Medline]

24 Grahame NJ, Barnet RC, Gunther LM, Miller RR. Latent inhibition as a performance deficit resulting from CS–context associations. Anim Learn Behav 1994; 22: 395–408[ISI]

25 Oberling P, Gosselin O, Miller RR. Latent inhibition in animals as a model of acute schizophrenia: a reanalysis. In: Haug M, Whalen RE, eds. Animal Models of Human Emotion and Cognition. Washington DC: American Psychological Association, 1999; 87–102

26 Harris JA, Westbrook RF. Midazolam impairs the acquisition of conditioned analgesia if rats are tested with an acute but not a chronic noxious stimulus. Brain Res Bull 1996; 39: 227–33[ISI][Medline]

27 Zacny JP, Coalson D, Young C et al. A dose–response study of the effects of intravenous midazolam on cold pressor-induced pain. Anesth Analg 1995; 80: 521–5[Abstract]

28 Feldon J, Weiner I. Abolition of the acquisition but not the expression of latent inhibition by chlordiazepoxide in rats. Pharmacol Biochem Behav 1989; 32: 123–7[ISI][Medline]

29 Lacroix L, Spinelli S, Broersen LM, Feldon J. Blockade of latent inhibition following pharmacological increase or decrease of GABAA transmission. Pharmacol Biochem Behav 2000; 4: 893–901

30 Harris JA, Westbrook RF. Contextual control over the expression of fear in rats conditioned under a benzodiazepine. Psychopharmacology 2001; 156: 92–7[ISI][Medline]

31 Colpaert FC. Amnesic trace locked into the benzodiazepine state of memory. Psychopharmacology 1990; 102: 28–36[ISI][Medline]

32 Bouton ME, Kenney FA, Rosengard C. State-dependent fear extinction with two benzodiazepine tranquilizers. Behav Neurosci 1990; 104: 44–55[ISI][Medline]

33 File SE, Goodall EM, Mabbutt PS, Harris A, Skelly AM. State-dependent retrieval and midazolam. Human Psychopharmacol 1993; 8: 243–51[ISI]

34 Killcross S, Balleine B. Role of primary motivation in stimulus preexposure effects. J Exp Psychol Anim Behav Process 1996; 22: 32–42[ISI][Medline]

35 Nadin G, Coulthard P. Memory and midazolam conscious sedation. Br Dent J 1997; 183: 399–407[ISI][Medline]

36 Wagner BK, O’Hara DA, Hammond JS. Drugs for amnesia in the ICU. Am J Crit Care 1997; 6: 192–201[Medline]

37 Henderson A, Dipplesman J, Miller J. Failure of intravenous low dose to influence memory recall in drug paralysed post-operative patients sedated with papaveretum. Aust Crit Care 1994; 7: 22–4[Medline]

38 Glass PS, Bloom M, Kearse L, Rosow C, Sebel P, Manberg P. Bispectral analysis measures sedation and memory effects of propofol, midazolam, isoflurane, and alfentanil in healthy volunteers. Anesthesiology 1997; 86: 836–47[ISI][Medline]

39 Thompson JM, Neave N, Moss MC, Scholey AB, Wesnes K, Girdler NM. Cognitive properties of sedation agents: comparison of the effects of nitrous oxide and midazolam on memory and mood. Br Dent J 1999; 187: 557–62[ISI][Medline]

40 Ghoneim MM, Block RI, Ping ST, el-Zahaby HM, Hinrichs JV. The interactions of midazolam and flumazenil on human memory and cognition. Anesthesiology 1993; 79: 1183–92[ISI][Medline]

41 Veselis RA, Reinsel RA, Feshenko VA, Wronski M. The comparative amnestic effects of midazolam, propofol, thiopental, and fentanyl at equisedative concentrations. Anesthesiology 1997; 87: 734–6[ISI][Medline]

42 Rammsayer TH, Rodewald S, Groh D. Dopamine-antagonistic, anticholinergic, and GABAergic effects on declarative and procedural memory functions. Brain Res Cogn Brain Res 2000; 9: 61–71[ISI][Medline]

43 Salinas JA, Dickinson-Anson H, McGaugh JL. Midazolam administered to rats induces anterograde amnesia for changes in reward magnitude. Behav Neurosci 1994; 108: 1059–64[ISI][Medline]