All procedures were carried out in accordance with the UK Animals (Scientific Procedures) Act (1986) and its associated guidelines. A total of 26 group-housed male Sprague–Dawley rats (Envigo, UK), aged ~ 2 months at the beginning of training, were used. Animals were maintained on a 12-h light/dark cycle (lights on 07.00). During testing, rats were food-restricted to maintain ~ 85–90% of their free-feeding weight. Water was available ad libitum in their home cages. For the NAcC infusion studies, animals were implanted with bilateral guide cannulae (Plastics One) 1.5 mm above the target site of the NAcC (AP: + 1.4 mm, ML: ± 1.7 mm, DV: –6.0 mm from skull surface) under isoflurane anaesthesia and secured with dental acrylic.
Testing was carried out in operant chambers (30.5 × 24.1 × 29.2 cm; Med Associates) (Fig. 2a). Each chamber was housed within a sound-attenuating cabinet ventilated with a fan, which provided a constant background noise of ~ 59 dB. Each chamber contained two retractable levers, 9.5 cm of either side of a central nose poke, which was fitted with an infrared beam signalling when animals entered the receptacle. The wall opposite contained a food magazine into which 45-mg sucrose pellets (Sandown Scientific, UK) could be dispensed. Each chamber was also fitted with a house-light and a speaker for delivering auditory stimuli.
The task design is shown in Figs. 1 and 2a, b. The task required animals to use an auditory cue to guide whether to initiate a response and press either the left or right lever (Go left/right) or to withhold responding (No-Go) to gain either a small or large reward. A trial was initiated when the rat voluntarily entered and stayed in a central nose poke for 0.3–0.7 s. This triggered the presentation of one of four auditory cues, which signalled (a) the action requirement (Go left/right or No-Go) and (b) the reward for a correctly performed trial (small or large, respectively, 1 or 2 sucrose pellets) (Fig. 2b). Go trials required animals to make two presses on the correct lever within 5 s of cue onset (Figs. 1a and 2b). On No-Go trials, animals were required to remain in the nose poke for the No-Go hold period (Fig. 1b). While No-Go trials always posed the same requirement regardless of the promised reward, the left/right mapping on Go trials was consistently associated with a specific reward size (small/large), with the side and reward associations counterbalanced across the cohort. This allowed us to independently manipulate action requirement and reward expectation, and additionally, to assess how this influenced the direction of motor responses on Go trials. Successful trials caused reward to be delivered to a food magazine on the opposite wall of the chamber.
To teach rats to respond or withhold actions to go and no-go cues, animals were trained as described by (Syed et al. 2016). Briefly, animals were first habituated to the operant chamber and learned to retrieve pellet rewards from the food magazine tray. Rats then commenced training initially exclusively with the No-Go trial type. Over several sessions, they were gradually trained to make and hold a response in the central nose poke, such that on No-Go trials, they were eventually able to withhold responding during a 0.3- to 0.7-s pre-cue nose poke hold period and a subsequent No-Go hold period (Cohort 1, systemic SB242084: 1.7–1.9 s; Cohort 2, local SB242084, d-amphetamine and systemic SB242084 replication: 1.5–1.7 s) to gain a reward (1 or 2 sucrose pellets, respectively, for “No-Go small” or “No-Go large” trials). The cue was either a tone, buzz, white noise, or clicker, counterbalanced across animals (each ~ 74 dB). A premature head exit caused the house light to be illuminated as the animal exited the nose poke for the duration of a 5-s time-out; after which, the house light turned off, and a standard 5-s inter-trial interval (ITI) commenced.
Once 60% of No-Go Small and Large trials were performed correctly, rats were next trained exclusively on Go trials. Here, after a 0.3 to 0.7-s central nose poke, one of the two remaining auditory cues would sound, one requiring a left lever press and the other a right lever press (side counterbalanced across animals). Correct responding on a particular lever was associated throughout testing with either a single sucrose pellet (“Go small”) or two sucrose pellets (“Go large”). An incorrect lever press would result in the house light switching on for a 5s time-out period, followed by a 5-s ITI. During training, an error-correction procedure was used so that the next trial after an error would always be of the same cue/trial type with the wrong lever withdrawn. Once a criterion of 60% successful Go responses was reached, interleaved No-Go and Go trials were introduced, each with a 25% probability (other than correction trials).
Once an average 60% success rate on all four trial types in a session was achieved, the rats moved onto the full version of the task. Here, error correction trials were removed. Further, the number of necessary lever presses on Go trials was increased to two. This ensured that the interval between cue onset to reward delivery was similar between Go and No-Go trials. Reward delivery was delayed for 1 s after the successful completion of a trial. Similarly, the error signal (the house light being illuminated) was also delayed for 1 s following an erroneous response. Throughout training and testing, a session ended after the animals had either earned 100 rewards or had spent 60 min in a session, whichever came first (although rats always met the former criterion during test sessions in the current study). Typically, rats took 1–1.5 months to train to reach stable baseline performance, and each rat in the current study had performed the full task at least 10 times before undergoing drug testing sessions. In between every drug testing session, rats underwent at least one training session without drug manipulations to reestablish baseline performance, where error correction trials were reintroduced, but all other parameters were kept constant.
Drugs were administered in a within-subjects regular Latin square design. The local infusion sessions were performed following the full completion of an experiment to assess the effect of intra-NAc infusions of D1 agonist or antagonist drugs on task performance (data to be reported separately) and after returning to stable baseline levels of performance. This meant that, prior to the start of local infusion experiments reported here, rats had received 6–7 NAcC infusion (plus one mock infusion session where no substance was injected). Animals always performed at least one behavioural session without injection or infusion in between each drug manipulation to reestablish baseline performance and rule out lasting effects of drugs.
SB242084 (6-chloro-2,3-dihydro-5-methyl-N-[6-[(2-methyl-3-pyridinyl)oxy]-3-pyridinyl]-1H-indole-1-carboxyamide dihydrochloride, Tocris) was dissolved in 25 mM citric acid in 8% w/v cyclodextrin in distilled water, and the pH adjusted to 6–7 using 5 M NaOH. Doses for systemic administration match those used in previous studies to have demonstrated increases in impulsive responding (Winstanley et al. 2004b, a; Fletcher et al. 2007; Silveira et al. 2020) and instrumental vigour (Higgins et al. 2003). Stock solutions of the vehicle (25 mM citric acid, 8% w/v cyclodextrin in distilled water), SB242084 0.1 mg/ml (referred to in the results’ section as “low” dose) and SB242084 0.5 mg/ml (“high” dose), were prepared and aliquoted before being frozen at − 20 °C. Concentrations of both drugs were calculated as the salt. On each experimental day, one aliquot of each drug was defrosted. Systemic injections of drug or vehicle, containing 25 mM citric acid and 8% w/v cyclodextrin in distilled water, were given intraperitoneally in a volume of 1 ml/kg. Injections were given 20 min prior to testing.
SB242084 was made up as described above. Doses were chosen based on a prior report showing a consistent and dose-dependent increase in impulsive responding on the 5CSRTT (Robinson et al. 2008). Stock solutions (vehicle, SB242084 0.2 μg/μl, 1.0 μg/μl) were prepared and aliquoted before being frozen at − 20 °C. Concentrations of both drugs were calculated as the salt. On each experimental day, one aliquot of each drug was defrosted. D-amphetamine (( +)-α-methylphenethylamine hemisulfate, Sigma-Aldrich) was dissolved in 0.9% NaCl solution to reach a concentration of 10 μg/μl). This was aliquoted and frozen at − 20 °C. On each experimental day, one aliquot of the drug was defrosted.
Rats were anaesthetised using inhaled isoflurane (4% in O2 for induction and 1.5% for maintenance) and given buprenorphine (Vetergesic, 0.03 mg/kg) and meloxicam (Metacam, 2 mg/kg) at the start of surgery. Once animals were secured in a stereotaxic frame and their scalp shaved and cleaned with dilute Hibiscrub and 70% alcohol, a local anaesthetic (bupivacaine, 2 mg/kg) was administered to the area. The skull was then exposed and craniotomies were made for implantation of bilateral guide cannulae (Plastics One, UK), consisting of an 8-mm plastic pedestal holding together two 26 gauge metal cannulae with a centre-to-centre distance of 3.4 mm and a length of 7.5 mm. The cannulae were implanted 1.5 mm above the target site of the NAcC, at AP + 1.4 mm, ML ± 1.7 mm, DV − 6.0 mm from the surface of the skull and relative to bregma (Franklin and Paxinos 2007). Four anchoring screws (Precision Technology Supplies) were also implanted and dental acrylic was applied to secure the cannulae to the skull. At the end of the surgery, dummy cannulae of the same length as the guide cannulae were inserted to ensure patency and a dust cap was secured onto the pedestal to secure the dummy. Following surgery, animals were again administered buprenorphine (0.03 mg/kg) and meloxicam (2 mg/kg) and were thereafter given meloxicam for up to 3 days post-surgery.
Local infusion procedure
Animals were first habituated to the manipulation of their implants by being lightly restrained and having the dummy cannulae removed and a fresh set reinserted. Retraining of the animals commenced approximately 2 weeks after surgery. All animals returned back to criterion performance (≥ 60% success rate on each trial type) before drug infusions began. During infusions, the rats were gently restrained whilst the dummy cannulae were removed and the 33 gauge bilateral infusion cannulae, at a length of 9 mm, were inserted into the NAcC. A total of 0.5 μl of vehicle or drug solution was injected per hemisphere at a rate of 0.25 μl/min. The infusion cannulae were left in place for 2 min after the cessation of the infusion to allow diffusion of solution from the cannulae. Next, the infusion cannulae were removed, the dummy cannulae and dust cap replaced, and the rats were returned to their home cage for 10 min to reduce the possible effects of injection stress on performance. Testing then commenced.
At the end of data collection, animals were deeply anaesthetised with sodium pentobarbitone (200 mg/kg, i.p. injection). They were then transcardially perfused with 0.9% saline followed by a 10% formalin solution (vol/vol). The brains were kept in 10% formalin solution until being sectioned. Brains were sectioned into 60-μm-thick coronal sections using a vibratome (Leica). The sections were then stained with cresyl violet (Sigma-Aldrich) before being mounted in DePeX mounting medium onto 1.5% gelatin-coated slides and enclosed with coverslips.
Latency and performance measures of interest are summarized in Table 1.
Measures of performance
The primary measure of performance was Go and No-Go trial success, measured as the percentage of correctly performed small or large reward Go or No-Go trials, respectively. On Go trials, we considered two types of errors: (i) no response on either lever within 5 s (“response omission” Go) or (ii) incorrect lever response (“wrong lever” Go). No-Go trials were considered to be erroneous if rats left the nose poke prematurely before the end of the holding period while the cue was still playing. We reasoned that premature responses could result from either a failure to inhibit fast cue-driven responses or from a failure to wait for the appropriate time period before initiating a response. We anticipated that the former process would lead to premature responses clustered near cue presentation, while the latter would manifest in No-Go failures nearer the end of the holding period. In support of this distinction, we had previously observed that stimulation of dopamine receptors in this task selectively increased cue-elicited No-Go errors (Grima et al. 2021). Therefore, instances of premature action initiation were quantified within the first and second halves of the No-Go hold period (“early” and “late” epochs, < 0.95 s and ≥ 0.95 s or < 0.85 s and ≥ 0.85 s for the 1st and 2nd cohort, respectively).
Additionally, there were three other behaviours that animals could exhibit that reflected possible changes in performance. First, after exiting the nose poke prematurely on an unsuccessful No-Go trial, animals sometimes pressed a lever. If this occurred during a period equivalent to the minimum No-Go hold interval, these instances were labelled as “invalid lever press” trials and quantified. Second, the number of presses on the correct lever preceding food magazine entry was quantified in Go trials, as rats sometimes continued to press beyond the required number to gain reward. Third, occasionally, rats exited the nose poke prematurely during the pre-cue period and thereby failed to initiate a new trial. These instances were labelled “aborted trials”.
Measures of latency
Trial windows of interest are summarised in Fig. 1. Action initiation latency was measured as the time from cue onset to exit of the nose poke in both Go and No-Go trials. Following on from the planned analyses of the patterns of No-Go errors, we conducted additional exploratory analyses on successful No-Go trials where the nose poke exits were classified as falling within the pre-reward (between cue offset and reward delivery later) or within the post-reward interval. We analysed observations within the first 0.8 s of each of these intervals (instead of the full 1 s), as the jitter of the No-Go hold duration sometimes resulted in the remaining 0.2 s falling outside of the time interval of interest. Similar to our conceptualisation of No-Go errors, we reasoned that responses clustered just after No-Go cue offset in the pre-reward period or just after reward delivery in the post-reward interval could be considered cue-driven responses. Travel time in Go trials was measured as the time from nose poke head exit to 1st press on the appropriate lever. Inter-press latency was measured as the time interval between the 1st and the 2nd press on the appropriate lever. Magazine latency was defined as the time interval between completion of a successful No-Go response or the Go response and entry to the food magazine. Changes in within-trial response latencies were used as proxies for changes in animals’ instrumental vigour. Reengagement latency was defined as the time taken to reenter the nose poke either after leaving the food magazine in successful trials or 1 s following the registration of an error on unsuccessful trials.
In order to determine whether any of the effects in the local SB242084 study might have been masked by tissue damage around the tips of the cannulae, we also carried out intra-NAcC infusions of d-amphetamine in this cohort, which has previously been shown to increase behavioural activation and premature responding (Cole and Robbins 1987; West et al. 1998). This meant that, as well as analysing the effect of a hyperdopaminergic state in NAcC on task performance and latencies, we could also use two behavioural measures that were predicted to be strongly influenced by this drug in the majority of animals – No-Go accuracy and Go trial action initiation latency – as a potential marker for the continued viability of the brain tissue around the cannulae. Specifically, if the effect of d-amphetamine on these measures within a subject did not exceed the cohort mean d-amphetamine-induced change by ≥ 50% in at least one of these behavioural outcomes, that animal was categorized as potentially having functionally significant tissue damage around the cannulae tips.
All data were analysed using MATLAB (Mathworks), SPSS (IBM), and R (The R Foundation). Performance and time measures were mainly analysed using repeated measures ANOVAs with drug dose and reward size as the within-subject factors unless specified otherwise. As we have described the effects of reward on task performance in a detail in a separate manuscript (Grima et al. 2021), the primary focus here was on the effects of and interaction with the drug. Any significant main effects of reward are, however, reported in the figure legends. Behavioural measures not reported in the main text are documented in Supplementary Tables 1–3 (1, systemic SB242084; 2, intra-NAcC SB242084; 3, intra-NAcC d-amphetamine). In the d-amphetamine infusion experiment, we used sessions where saline was locally infused into NAcC from a parallel dopamine receptor pharmacology experiment in the same animals as vehicle data for within-subjects analysis. In the systemic SB242084 replication experiment, an additional within-subject factor “training experience” was specified, due to the inclusion of 2 separate replications of systemic SB242084 administration, one prior to cannulation surgery and one after local infusion experiments had been completed. Percentages of nose poke exits in the early and late parts of the No-Go hold period were compared using the chi-squared test of contingency on the group level, as the low number of such observations resulted in skewed residual distributions in ANOVAs. Whenever there was a significant main effect of a drug or an interaction with a task variable of interest, test results were reported in the main text or figure legends and post hoc comparisons across levels of that effect were carried out. The influence of outliers on the ANOVA results was minimized by excluding any subject on the basis of absolute standardised residual values bigger than 3 from that analysis. A p-value less than 0.05 was considered significant.