Noradrenaline and Dopamine Efflux in the Prefrontal Cortex in Relation to Appetitive Classical Conditioning

Introduction

The mesocorticolimbic dopamine (DA) system plays a critical role in motivational processes (Robbins and Everitt, 1996; Kalivas and Nakamura, 1999). DA also exerts an important modulatory influence on cognitive functions typically attributed to the prefrontal cortex (PFC), such as working memory (Arnsten, 1997). The tonic release provides an activation of DA receptors in the PFC that may be related to working memory in an inverted U-shaped manner (Goldman-Rakic et al., 2000). In addition, phasic activation may favor the processing of stimuli that are of direct importance to the animal, given its present needs and preferences.

Evidence from in vivo microdialysis studies indicates that appetitive emotional stimuli can induce such a phasic activation of prefrontal DA release (for review, see Feenstra, 2000). This applies not only to direct, unconditioned stimuli (USs) but also to conditioned stimuli (CSs) that predict reward presentation through classical (pavlovian) conditioning. Consistent with the hypothesis that DA plays a role in appetitively motivated behavior, Bassareo and Di Chiara (1997) and Ahn and Phillips (1999) reported conditioned DA efflux in the PFC.

Although noradrenaline (NA), like DA, may have similar importance as a modulator of prefrontal functions (Arnsten, 1997), the role of noradrenergic innervation of the PFC in the processing of reward-related information has essentially been neglected. We found one direct indication of its activation by stimuli related to natural rewards (Sara and Segal, 1991), while only recently, Drouin et al. (2002) and Ventura et al. (2003) uncovered a role for NA in the motivational effects of addictive drugs. Addressing this subject, we designed experiments to directly compare the activation of prefrontal NA and DA efflux by classical conditioning. Rats were trained in a fully controlled environment (Skinner box) to associate a neutral auditory stimulus with the receipt of an unconditioned stimulus. Applying this approach, we reported that presentation of aversive conditioned stimuli increased both prefrontal NA and DA in vivo efflux (Feenstra et al., 2001), indicating a role for both catecholamines in coding the significance value of stimuli. Therefore, we tested the hypothesis that presentation of appetitive conditioned stimuli would also lead to activation of both catecholamine afferents to the PFC.

Materials and Methods

All experiments were approved by the Animal Experimentation Committee of the Royal Netherlands Academy of Arts and Sciences and were conducted in agreement with Dutch laws (Wet op de Dierproeven, 1996) and European regulations (Guideline 86/609/EEC).

Subjects. Male Wistar rats (Harlan CPB, Zeist, The Netherlands) weighing 100-125 gm at arrival were socially housed in groups of four under a reversed light/dark cycle (dimmed red light from 7:00 A.M. to 7:00 P.M.). They were kept for at least 3 weeks with food and water available ad libitum and were handled frequently. This schedule was maintained for the animals in experiment 1 up to the day of surgery. For the animals of experiment 2, food was taken away in the afternoon 5 d before surgery and was subsequently restricted to 16 gm per rat per day. The weights of these rats were reduced to ∼90% of free-feeding weights.

Rats were anesthetized with chloralhydrate (3.5% solution in saline; 400 mg/kg, i.p.) with additional local anesthesia via lidocaine (6% solution in saline). Under aseptic conditions, two microdialysis probes with 4 mm exposed length of Hospal AN69 dialysis membrane were placed bilaterally in the medial PFC at an angle of 12° (coordinates in millimeters from bregma: anterior, +3.0; lateral, ±1.8; ventral, -5.5), as described previously (Feenstra et al., 1998).

After recovery from anesthesia, each rat was placed in its individual Perspex cage (25 × 25 × 32 cm) with ad libitum access to food and water. On the following day, food was again restricted in experiment 2, and behavioral tests were started. Weights at the time of testing were between 250 and 350 gm.

Behavioral apparatus. Training and testing were conducted in dialysis-compatible Skinner boxes measuring 29.2 × 24.1 × 21 cm (MED Associates, St. Albans, VT), mounted within sound- and light-attenuating chambers, and dimly illuminated by a light oriented toward the ceiling. All Skinner boxes were equipped with a food dispenser from which pellets (Noyes formula P; 45 mg; Sandown Scientific, Hampton, UK) could be delivered and a wall-mounted loudspeaker through which an auditory stimulus (white noise, 25-30 dB measured at 10 cm distance) could be presented. Because of the extended stay in the boxes, water bottles were present. Behavioral responses before and during the CS presentation were recorded by the following: (1) an infrared photo beam that could detect each rat's head entry into the food dispenser (nose poke) and (2) an infrared movement detector mounted on the ceiling that recorded the general locomotor activity throughout the box. The recorded measures included the number of nose pokes, latency until the first nose poke after the start of CS presentation, and general locomotor activity.

Neurochemical apparatus. On the third day after surgery (microdialysis day), implanted probes were connected in series (Feenstra et al., 2002) using flexible polyetheretherketone (inner diameter, 0.13 mm) tubing to a Tsumura (TCS 2-23) quartz-lined dual-channel swivel (Pronexus, Skärholmen, Sweden), and Ringer's solution (in mm: 145 NaCl, 2.7 KCl, 1.2 CaCl₂, 1.0 MgCl₂) was perfused at a constant rate of 3 μl/min. Dialysis samples were automatically injected every 16 min into the HPLC system, as described by Feenstra et al. (1998).

Experimental procedures. On the 2 consecutive days after surgery, the rats were trained in Skinner boxes. On the third day, training was combined with on-line microdialysis measurements. Three different groups of rats were tested: paired, unpaired, and controls.

The paired group was trained to associate a neutral stimulus (white noise, CS) with pellet delivery (appetitive US). On the first day, rats were placed in the Skinner boxes with five pellets available in the food dispenser. They were allowed to explore the box for 90 min. After this adaptation period, three normal acquisition sessions (CS plus US) were given. Each session consisted of six presentations of a 10 sec noise, separated by random intervals ranging between 100 and 200 sec. In the final second of the noise, two (experiment 1) or four (experiment 2) pellets were delivered. Session duration was 16 min, and intersession intervals were variable, averaging 60 min. The animals were returned to their home cages 1 hr after the last session. On the following day, the same training schedule was used but without the adaptation period. On the third day, NA and DA efflux were measured on-line throughout the day. A CS plus US session was given when the efflux had stabilized. Two hours later, a CS-alone (extinction) session was given, in which the animals were subjected to six unrewarded CS presentations. Behavioral scores are now indicative of the strength of the associative learning process.

The unpaired group was subjected to the same schedule of sessions during the three testing days, with one exception: in each session, the CS was presented explicitly unpaired with the US (stimuli separated by a random interval of 50-100 sec). The rats of the control group had similar sessions on all 3 d, consisting only of CS presentations.

Histology. Coronal (20 μm) cryostat sections were stained with thionine and examined for the precise location of the microdialysis probes. Data from rats with incorrectly placed probes were not reported.

Data analysis. For nose poke and locomotor activity data, differences were calculated between the values measured during the 10 sec of CS presentation and the 10 sec immediately before that. Latencies were transformed into ¹⁰log values.

Statistical analyses of these data were first conducted for the full experiment (days 1-3) and later separately for day 3. ANOVA with repeated measures [using the Huynh-Feldt ϵ for correction (Norušis, 1990)] had session as the within-subject factor and group as the between-subject factor. When indicated, between-subject effects were analyzed using Student-Newman-Keuls (SNK) tests. ANOVA with repeated measures for separate groups was followed, when indicated, by simple contrast analysis to compare the data of the first training session with those of successive sessions. The level of significance was p < 0.05 in all circumstances.

Changes in neurochemical concentrations were expressed as percentages, with the mean of the first three samples preceding the stimulus presentation set at 100% as basal level.

The statistical analyses of the microdialysis data were conducted separately for the final CS plus US session and for the CS-alone session. ANOVA with repeated measures was applied with sample as the within-subject factor and group as the between-subject factor. When indicated, between-subject effects were analyzed using SNK tests. ANOVA with repeated measures for separate groups was followed, when indicated, by simple contrast analysis to compare the final basal sample with the subsequent samples.

Results

Discussion

For the first time, prefrontal NA and DA responses in reward-related associative learning were directly compared. The results were surprising, because they revealed that presentation of an explicit, auditory stimulus that was associated previously with reward presentation only increased NA efflux in the PFC of the rats. Although DA is often regarded as the monoamine that is most strongly implicated in motivational processes (Robbins and Everitt, 1996; Kalivas and Nakamura, 1999), no conditioned increases in DA efflux were found.