Classical conditioning paradigms are widely used to study emotional mechanisms of the brain 12345. Fear conditioning is a form of Pavlovian classical conditioning. It is produced experimentally by arranging the occurrence of an insignificant or emotionally neutral sensory stimulus, the conditioning stimulus (CS), with an aversive stimulus, the unconditioned stimulus (US). The CS, by virtue of its relationship with the US, acquires aversive properties and comes to obtainresponses characteristically elicited by threatening stimuli. Thus, a tone that has been previously been paired with foot-shock elicits fear or defense responses. Because these responses are not elicited by the CS before the temporal pairing of the CS with the US, they can be referred to as learned or conditioned emotional responses. Indices of fear conditioning include the measurement of the activity of motor responses (i.e. induction of freezing behavior) and/or a variety of autonomic responses (changes in arterial pressure, heart rate, skin conductance, etc.). Fear conditioning, also known as defense conditioning, is often used as an experimental model of anxiety 67.

Emotional responses to aversive stimuli are usually associated with limbic system components such as the hypothalamus, amygdala, and the cingulate cortex. Lesions of either the hypothalamus or the amygdala result in attenuation of blood pressure, heart rate, or behavioral immobility ("freezing") during the conditioned emotional response 58910. Cingulate cortex lesions result in attenuation of cardiovascular conditioned responses 111.

It is now well established that the amygdala in particular is a critical link in the pathway through which sensory stimuli come to acquire fear evoking properties 58121314. Direct projections from the central nucleus of the amygdala to the lateral hypothalamus mediate the activation of the sympathetic autonomic nervous system that is seen during fear and anxiety 10. Projections to the dorsal motor nucleus of the vagus or parabrachial nucleus 15 are involved in several autonomic measures of fear and anxiety. Projections to the ventral tegmental area 16 may mediate stress-induced increases in dopamine metabolites in the prefrontal cortex. Release of norepinephrine or 5-HT onto motor neurons, via amygdaloid activation of the locus coeruleus or raphe' neurons, could lead to enhanced motor performances. The hippocampus is reciprocally connected with the amygdala 17 and interactions between the amygdala and hippocampus appear to play an important role in some cognitive functions 1819. Thus, the highly correlated set of behaviors seen during fear may result from activation of the amygdala, which then projects to specific target areas that are critical for specific symptoms of fear.

The identity of the transmitters released into these target sites by amygdaloid neurons has been studied 20 and it is estimated that 25% of the neurons in central amygdala contains corticotropin-releasing factor, somatostatin and neurotensin. High density of benzodiazepine 21 and opiate receptors 22 are also found in the nuclei of the amygdala. Local infusion of the opiate agonist levorphanol into the central amygdala has anxiolytic effects in the social interaction test 2324. Local infusion of benzodiazepines into the amygdala has anxiolytic effects in the operant conflict test 25262728.

Cholecystokinin (CCK) is found in high concentrations in cortical and limbic structures of the temporal lobe, including the amygdala and the hippocampus of the rat 2930 and the monkey 31. Evidence has been gathered supporting a role for CCK in the neurobiology of anxiety 5323334. CCK-induced excitation of rat hippocampus units is antagonized by benzodiazepine receptor agonists 35, suggesting that CCK might have anxiogenic properties 3637. Moreover, injections of CCK-8 into the central nucleus of the amygdala enhance arousal and fear in rats 3839. CCKB receptors antagonists show anxiolytic effects in several standard anxiety tests (the light/dark compartment test, the elevated plus-maze test, the social interaction test) 28324041. Interestingly, delivery of electric shock to freely moving rats induces increase of CCK concentrations in limbic regions and in the prefrontal cortex 42. Administration of CCK-4 to healthy patients induces panic attacks 43 and produces respiratory changes in panic disorder and control patients 44.

A test well suited for the study of fear and anxiety will be applied 47. This test allows determining fear responses in a fast and reliable way.

Briefly, rats are placed individually in a rodent sound attenuating conditioning chamber (Coulbourn Instruments). The conditioned stimuli (CS) are an 800 Hz tone, amplified to 80 dB, and presented for 20 sec through a speaker located in the front panel of the chamber. The unconditioned stimulus (US) is a brief (0.5 mA, 500 msec) distributed delivery of direct current produced by a grid floor shocker.

On day 0, the animals are placed in the conditioning chamber for 20 min without stimulus presentation and are then returned to their home cages. On days 1 and 2 conditioning trials (consisting of 2 trials/day during which the US was presented during the last 500 msec of the 20 sec CS) are giving. Extinction trials (2 presentations of the CS alone) begin on day 3 and continue for 3 additional days. Freezing, used as the index of conditioned fear, is assessed by viewing the animals through a peephole in the sound-attenuating chamber and using stopwatches to measure freezing time. Freezing is defined as the absence of all movement except for respiratory-related movements 47. It is a very distinctive behavior that is easy to recognize and rarely occurs in non-fear situations 4849.

The micro-electrodes utilized here were prepared using a 12 µm-diameter carbon fiber (Carbone Lorraine, Lyon, France) and set as described earlier 45. They are cut under the microscope to leave 50-200 µm active area at the tip (Figure 1). This preparation gives the electrodes an impedance of 0.5-1.0 MOhms, allowing electrophysiological single cell activity discrimination 50.

For in vitro and in vivo DPV measurements the carbon fiber electrode is set as described earlier 4551. Briefly, it is electrically treated firstly with a voltage from zero to 3 Volts, 70 Hz, 10 s, then with continuous potentials (+1.5 Volts, 5 s and −0.9 Volts, 5 s), so as to permit the measurement of three oxidation signals related to ascorbic acid, dopamine and serotonin metabolites, respectively, as well as that of a further oxidation peak when the differential pulse voltammetric (DPV) recordings were made in the same solution with the addition of the amino acids or neuropeptides such as CCK and when the DPV scan rate used was 10mV/s from -250 to +950mV at a step size of 50mV (see Figure 2). The DPV measurements of oxidation potentials and of oxidation currents are done automatically every 5min by the Autolab polarograph (potentiostat/galvanostat Ecochemie, The Netherlands) linked to an IBM pc computer equipped with a general purpose electrochemical system software (GPES) package.

The unit wave signals are fed into an amplifier (Fintronics). The signals are also led to an Axon Instruments system for acquisition and subsequently off-line analysis. Single unit discrimination is accomplished through a window discriminator (FHC Instruments) connected to an oscilloscope (Textronics)

For freely moving animal preparations a chronic stage will be used. The animals are placed in a stereotaxic frame under pentobarbital anaesthesia (40mg/kg). An opening is drilled through the skull above the CNS structure of interest (e.g. amygdala, hippocampus, etc.). The microsensor is then located into the area selected as described earlier 51 and fixed into a head stage mounted with dental acrylic on the skull of the animal. After a recovery of 5-7 days the animals will be ready for behavioral tests.

Previous experiments using such microsensors for concomitant in vivo electrophysiological and voltammetric measurements demonstrated that there is no reciprocal influence upon the respective signals recorded in vivo. And that that no damage is done to the single unit recordings by the current used for voltammetry 46. Here, Figure 2 shows single cell discrimination and voltammetry recording in a Central Amygdala preparation. DP Voltammograms lasting 60sec each are carried out every 5 min. In the interval time of 5 min electrophysiological measurements are performed. It appears evident that electrophysiological monitoring of unit activity is not altered after voltammetry recording and this during 1 hour of control analysis. This is in accord with the above mentioned studies that have been performed in other brain areas, therefore further supporting this methodological approach. DPVoltammograms show three signals at approximately +50mV; +300mV and +650mV, respectively. They correspond to the oxidation of endogenous cathechol (peak 1), indole (peak 3) and peptidergic (peak 5) compounds, respectively. In particular peak 5 recorded in amygdala may corresponds to the oxidation of endogenous CCK 5152.

In the experiments performed under urethane anaesthesia, CCK and CCK antagonist will be administered iv (in the lateral tail vein) or icv (in a lateral ventricle) when the drugs do not readily pass the blood-brain barrier. In freely behaving animals experiments a 23 gauge guide canula will be implanted in a lateral ventricle and secured to skull screws with dental cement. Drugs will be then injected before the behavioral test through a 30 gauge canula extending out of the previously implanted guide canula by 1mm.