Abstract

The medial prefrontal cortex (mPFC) and nucleus accumbens (NAc) are 2 structures within a larger corticolimbic network mediating goal-directed actions, especially when the procurement of different goals is sensitive to impulsive tendencies. The present study investigated the role of these structures in goal-directed action for differential reward by training rats to respond for sucrose reward at a nosepoke operandum such that longer duration nosepokes (up to 2 s) resulted in correspondingly larger volumes of reward. After 16 weeks of training, neurotoxic lesions of either the mPFC or the NAc-core were performed, followed by reassessment of sustained response behavior. Lesions of mPFC increased choice impulsivity by shifting responding away from large rewards toward rewards of smaller sizes. The total volume of reward earned remained unchanged, thereby dissociating the lesion effects on response parameters from overall motivation for reward. In contrast, NAc-core lesions decreased the total amount of responding and total volume of reward earned without altering choice impulsivity across differing nosepoke durations and reward sizes. These results suggest that the mPFC mediates the ability to maintain behavioral responding over longer durations for larger magnitude rewards, while the NAc-core mediates the initiation of responding, perhaps by affecting motivational drive, independent of reward magnitude.

Introduction

Each day we make countless decisions in which we choose among differing actions; as well, we must decide whether to maintain an action once engaged. These decisions are based largely on their prospective outcomes, making them central to adaptive behavior and survival. Poor or impulsive decision-making tendencies are thought to underlie a wide variety of disorders, most notably obsessive–compulsive disorder, eating disorders, and drug addictions (Dawe et al. 2004; Liao et al. 2008; Smari et al. 2008). Initial investigations into the neural bases of decision making, or neuroeconomics, using animal models of goal-directed behavior began more than 2 decades ago and were originally focused on the role of dopamine (DA) input into the nucleus accumbens (NAc) on effort-based decision-making tasks (Cousins and Salamone 1994; Cousins et al. 1996; Salamone et al. 1991; Salamone et al. 1994). These early studies have spawned more recent investigations within many different brain sites across both primate and rodent species that are just now revealing the neural network mediating decision-making processes (Cardinal 2006; Floresco, St Onge, et al. 2008; Glimicher and Rustichini 2007; Opris and Bruce 2005; Walton et al. 2006). The majority of these studies employ situations in which decisions among instrumental actions are principally controlled by the active representation of the desired goal or expected outcome. Response selections under these conditions are therefore mediated in an excitatory or inhibitory manner based on the incentive value of the expected outcome relative to baseline motivation (e.g., Balleine and Dickinson 1998; Robbins and Everitt 1996). The incentive value created by the outcome is also hypothesized to be modulated by the cost/effort and/or delay required to obtain the desired reward (Cardinal 2006; Walton et al. 2006). Impulsive tendencies are therefore important influences on goal-directed actions and must be directly inhibited during decision-making processes in order to procure the largest magnitude of reward relative to the corresponding effort required for goal attainment.

The ability to inhibit responses toward less desirable easily obtainable goals in favor of more desirable goals, that may also require more effort, is mediated by a wide array of interconnected corticolimbic structures. Experimental evidence suggests that 2 structures within this corticolimbic network, the prefrontal cortex (PFC) and NAc, are critical for decision-making processes involved in the procurement of desirable goals, especially when differential delays, effort, or probabilities need to be balanced against the incentive value of comparative goal choices (Balleine 2005; Cardinal et al. 2004; King et al. 2003; Lee et al. 2007; Rudebeck et al. 2006; Salamone et al. 2007). In particular, PFC function appears to be important for the maintenance of information regarding appropriate responses or rules based on association with reward-related stimuli or with the reward itself (Joel et al. 1997; Kesner and Raggozino 2003;,Miller and Cohen 2001; Montojo and Courtney 2008). The PFC is also thought to mediate the active inhibition of inappropriate responses at the same time as facilitating stimulus-appropriate responding for successful goal completion (Dalley et al. 2008; Knight et al. 1999; Sakagami et al. 2006). Several subregions have been identified within the PFC, such as the infralimbic cortex, anterior cingulate cortex, and the orbitofrontal cortex, that play important roles during effort-based (Hauber and Sommer 2009; Rudebeck et al. 2006; Schweimer and Hauber 2006; Walton et al. 2002, 2003) and delay-based (Rudebeck et al. 2006; Roesch et al. 2006; Walton et al. 2003) decision-making tasks in which competing responses deliver differential magnitudes of reward. The NAc is also a critical component of the neural circuitry underlying goal-directed actions, especially when effort- or delay-based decisions are required between actions that result in variable levels of reward magnitude (Cardinal et al. 2001; Cousins et al. 1996; Hauber and Sommer 2009; Mingote et al. 2005; Salamone et al. 2003).

The majority of studies that have identified the PFC and NAc as key structures mediating effort- or delay-based decisions have used behavioral models that require decisions between 2 competing response choices that are explicitly associated with rewards of distinct value, usually one small or nonpreferred and the other large or preferred. Here, we sought to expand upon such behavioral designs by increasing the number of choices presented to the subject and by doing so have increased the ability to assess choice impulsivity in a more detailed fashion. In particular, rather than presenting an initial choice between 2 options (i.e., left lever vs. right lever), in the present study, subjects had only a single response operandum at which nosepokes of various lengths provided 1 of 6 different reward magnitudes (“sustained response task”). By maintaining the duration of the nosepoke response, which inherently involves more cognitive effort via the inhibition of the natural tendency to withdraw, subjects earned greater reward; therefore, subjects decided among gradations of sustained responding for gradations of reward. As with other decision-making tasks, subjects must make a decision to initiate responding; however, we posit that our task requires an additional process of “online” decision making as the subject decides to maintain its response or to withdraw from the nosepoke operandum. In addition, our task employed free-running trials rather than the fixed trials commonly used in the majority of previously published decision-making studies. Importantly, under free-running trials, subjects have the opportunity to vary their responding to compensate for any alterations in the total amount of reward earned caused by experimental manipulations, thereby allowing examination of variations in the pattern of responding separately from alterations in the total volume of reward earned.

The current study sought to examine the role of the medial prefrontal cortex (mPFC) and the NAc in decision making using the sustained response task in which the animal can actively titrate its own level of sustained responding across a range of reward magnitudes on free-running trials. The results presented here identify dissociable roles for the NAc-core and mPFC in initiating and maintaining, respectively, sustained responding for differential reward magnitude.

Materials and Methods

Subjects

Forty-eight male Long-Evans rats weighing 150–175 g and 2 months of age at the start of the experiment were used as subjects. All rats were individually housed in Plexiglas cages on ventilated racks (Biozone) in a 12-h light/dark cycle (lights on at 7 AM) climate-controlled vivarium. The rats had ad lib access to food and water during the first 2 weeks of acclimation to the colony and daily handling, 5 min/day/rat. The rats weighed approximately 320–430 g at the start of initial training when access to homecage water was restricted to 1–2 h after each daily behavioral session, as described below. All procedures were conducted in accordance with the Guide for the Care and Use of Laboratory Animals (National Research Council 1996) and were approved in advance by the Gallo Center Institutional Animal Care and Use Committee.

Behavioral Apparatus

Rats were trained to nosepoke for sucrose reward in standard operant chambers (12.0″ L × 12.5″ W × 11.5″ H) that were housed in sound-attenuating melamine cubicles (26.0″ W × 22.0″ H × 14.0″ D) (Med Associates Inc.). The chambers were equipped with a recessed nosepoke operandum on the left panel that contained 3 colored cue lights (red, green, and yellow) and an infrared nose entry detector. The right panel had 2 cue lights (2.8 W) with opaque lens that were positioned 10 cm above the stainless steel grid floor, which provided chamber illumination during the behavioral sessions. The liquid reward port was also positioned on the right panel and contained a stainless steel dish equipped with stainless steel tubing that was connected to 60-mL syringes via polyethylene tubing. Sucrose was delivered by the activation of a syringe pump.

Behavioral Training

Rats were initially trained during an overnight session to respond at the nosepoke operandum illuminated with 3 colored cue lights on an FR-1 schedule for 0.1 mL of 20% sucrose (w/v). Following the overnight session, rats were trained for 2 more days on this FR-1 schedule in 45-min sessions. Rats were then required to sustain nosepoke responses across 3 minimum “fixed durations” (200, 400, or 600 ms) to train the rats to associate longer durations of individual nosepoke responses with increasing volumes of 20% sucrose reward (50, 150, and 300 μL, respectively). Trials of each individual “fixed duration” were pseudorandomly presented throughout the session in roughly equal numbers. Nosepokes that were sustained throughout the fixed duration were followed upon snout withdrawal by a 500-ms tone that signaled a successfully completed nosepoke duration and pending reward delivery upon withdrawal from the nosepoke and subsequent head entry into the reward port. The nosepoke cue lights were deilluminated in concordance with the fixed durations, providing an additional aid in determining the completion of the minimum duration: the red cue light turned off at 200 ms, the green cue light turned off at 400 ms, and the yellow cue light turned off at 600 ms. The intertrial interval following successfully completed fixed nosepoke durations was 10 s. If the rat failed to sustain the nosepoke response until the end of the fixed duration, a 10-s time-out period was initiated and the lighting inside the chamber was turned off. After 1 week of training on the fixed duration task, the concentration of sucrose was decreased to 10% for the remainder of the study. Rats continued training for 10% sucrose for a minimum of 10 sessions or until they successfully completed 50% of the trials at the longest fixed duration (600 ms) for 3 consecutive days. After reaching this criterion, the fixed durations were increased to 400, 600, and 800 ms and the corresponding differential reward magnitudes were increased in contrast (50, 100, and 400 μL). Rats were trained for an additional 20 sessions or until they successfully completed 50% of the trials at the longest duration (800 ms) for 3 consecutive days. Rats were trained on the fixed duration versions of the task for an average of 16.46 ± 0.83 sessions for the 200-, 400-, and 600-ms durations and 16.75 ± 1.61 sessions for the 400-, 600-, and 800-ms durations.

The final “Free Choice” phase of training allowed the rats to freely respond with nosepokes of any duration and at any rate during a 45-min session. The delivery of differential magnitudes of sucrose reward remained proportional to the individual duration of the sustained response across 6 duration ranges (see Fig. 1). The nosepoke cues deilluminated in a similar manner to fixed duration training but occurred at 400, 800, and 1200 ms after response initiation. Completion of nosepoke responses was no longer followed by a tone at this phase of training since all nosepokes produced reward delivery. Upon withdrawal from the nosepoke operandum, all remaining cue lights were deilluminated for a minimum of 2.33 s (fixed 1.5-s delay plus minimum reward delivery time) up to a maximum of 7.31 s (fixed 1.5-s delay plus maximum reward delivery time) depending on the corresponding magnitude of reward earned by the duration of the completed response. Additional nosepoke responses occurring during this deilluminated period did not result in additional reward delivery. Rats were trained on the Free Choice program for 17–20 weeks. After the second week of Free Choice nosepoke responding, the rats were returned to ad lib access to water for the remainder of the experiment. The last 4 weeks of Free Choice behavior were used as baseline performance to equate the pattern of responding for the rats selected for the mPFC and NAc sham and lesion groups in Experiments 1 and 2, respectively.

Figure 1.

Sustained response task. (A) Schematic representation of the nosepoke operandum and the cue lights inside the port relative to the duration of the sustained response. If the infrared beam for the nosepoke entry detector remained interrupted, then the port cues were deilluminated according to the following durations: red cue (upper right) off after 400 ms, green cue (upper left) off after 800 ms, and yellow cue (lower) off after 1200 ms. (B) Relationship between sustained response duration and corresponding sucrose volume delivered into the reward port on the opposite chamber wall.

Figure 1.

Sustained response task. (A) Schematic representation of the nosepoke operandum and the cue lights inside the port relative to the duration of the sustained response. If the infrared beam for the nosepoke entry detector remained interrupted, then the port cues were deilluminated according to the following durations: red cue (upper right) off after 400 ms, green cue (upper left) off after 800 ms, and yellow cue (lower) off after 1200 ms. (B) Relationship between sustained response duration and corresponding sucrose volume delivered into the reward port on the opposite chamber wall.

Surgical Procedures

All rats weighed approximately 450–630 g and were 8–9 months of age at the time of surgery. The subjects were anesthetized with isoflurane inhalation (5% induction and 1–3% maintenance) before placement into a stereotaxic apparatus (David Kopf). The scalp was incised along the midline and the underlying fascia was carefully scraped to the sides of the skull surface. The skull was leveled using the dorsal coordinates of bregma relative to lambda. In Experiment 1, burr holes were drilled above the mPFC at stereotaxic coordinates anterior to posterior (A/P) +3.4, medial/lateral (M/L) ±1.0, and A/P +2.7, M/L ±1.2 mm relative to bregma (Paxinos and Watson 1998). A 10-μL Hamilton syringe fitted with a 28-g beveled needle was mounted in a microinjection unit that was attached to the stereotaxic arm. The syringe was then slowly lowered to a depth of −3.8 mm from the dura surface. Either ibotenic acid (10 μg/μL) or artificial cerebrospinal fluid (aCSF) was injected at a rate of 0.1 μL/min for 2 min and a total volume of 0.2 μL in the anterior site, while 0.15 μL was injected at 2 different depths, −4.1 and −2.9 mm, in the posterior site. In Experiment 2, burr holes were drilled above the NAc-core at stereotaxic coordinates A/P +2.2, M/L ±1.7, and A/P +1.5, M/L ±1.8 mm relative to bregma. A similar 10-μL Hamilton syringe was lowered to a depth of −7.0 mm from the dura surface and either ibotenic acid (10 μg/μL) or aCSF was injected at a rate of 0.1 μL/min over 3 min at both sites. For both experiments, the needle was left in place for an additional 5 min after each injection before withdrawal. The incision was sutured and treated with lidocaine and triple antibiotic ointments. All rats were given 1 week of recovery time with ad lib food and water before resuming behavioral testing.

Postsurgical Testing

All rats were retested on the Free Choice program for 4 weeks to determine the pattern of sustained responding for differential magnitudes of sucrose reward following mPFC and NAc-core lesions, followed by 1 week of extinction testing in which no sucrose reward was delivered.

Statistical Analysis

All behavioral measures in both experiments were analyzed using mixed-design repeated measures analysis of variance (ANOVA) using pre- and postsurgical testing, and where appropriate, the number of weeks and/or nosepoke duration ranges, were used as within-subject measures. The dependent behavioral measures included total number of reinforced nosepoke responses, duration of reinforced nosepoke responses, total sucrose consumed, reward efficiency (total volume/number of reinforced nosepokes), number of nosepoke responses across the 6 duration ranges, and percentage of total nosepoke responses within each duration range. Planned comparisons were conducted using the Student–Newman–Keuls test to analyze any significant main effects with multiple factors or significant interaction terms. In addition, a percent change from presurgery baseline measure [((postsurgical value/presurgical value) − 1) × 100] was calculated for the overall behavioral measures in order to directly compare the postsurgical performance between the sham and lesion groups using one-way ANOVAs. The percent change measure was calculated for each individual subject first, given that every animal responded for an individualized amount of sucrose reward, and then, the individual values were averaged to generate the group mean value. Simple linear regression was conducted on the behavioral measures of sustained responding to determine their correlation to the total volume of reward earned.

For both experiments, separate within-subjects repeated measures ANOVAs for the sham and lesion groups were conducted on all behavioral measures over the course of the second to fourth postsurgical week of Free Choice behavior. This analysis determined whether significant changes occurred in the behavioral performance across postsurgical weeks of testing. An average postsurgical measure was used for each subject in subsequent analyses if postsurgical performance did not significantly vary across weeks. The statistical tests were computed using SuperANOVA (Abacus Concepts) and SigmaStat (SPSS Inc.) software.

Results

Baseline Sustained Response Performance

Rats were trained on the Free Choice version of the task in which the duration of individual sustained nosepokes was directly proportional to the volume of sucrose reward delivered. The acquisition of Free Choice performance over the course of 16 weeks of training was marked by a sharp transition to a high percentage of brief responses for the smallest reward (0–400 ms and 50 μL) coupled with sequential increases in the percentage of responses of the longest duration that produced the largest magnitude of reward (>2000 ms and 350 μL; see Fig. 2A). The shift in the pattern of responding across durations also produced significant changes in the percentage of total reward volume earned across response durations. Figure 2B illustrates that the sequential shift across weeks toward responses >2000 ms in duration for the largest individual reward produced considerable increases in the percentage of total reward volume earned by that response. The reward efficiency measure increased across all selected weeks of training, and a significant main effect of Week (F3,123 = 18.88, P < 0.001) was identified. Planned comparisons determined that reward efficiency significantly improved at each successive week selected during Free Choice training (all Ps < 0.05; see Fig. 2C).

Figure 2.

Acquisition of sustained responding for differential magnitudes of reward. (A) Acquisition of sustained response performance during Free Choice training as measured by the percentage of total nosepokes emitted within each duration range for increasing magnitudes of reward and across 4 selected time points. The percentage of nosepokes emitted at both response duration/reward extremes increased over successive weeks. (B) Acquisition of sustained response performance during Free Choice training as measured by the percentage of total sucrose volume earned within each duration range and across 4 selected time points. The longest duration nosepokes, which resulted in the delivery of the largest magnitude rewards, increased sequentially over the course of training and accounted for the largest percentage of reward volume earned. (C) Acquisition of sustained response performance during Free Choice training as measured by reward efficiency (total volume/total number of nosepokes). The reward efficiency or volume of reward earned per nosepoke increased sequentially over the 4 selected weeks of Free Choice training. *P < 0.05 relative to the selected prior week.

Figure 2.

Acquisition of sustained responding for differential magnitudes of reward. (A) Acquisition of sustained response performance during Free Choice training as measured by the percentage of total nosepokes emitted within each duration range for increasing magnitudes of reward and across 4 selected time points. The percentage of nosepokes emitted at both response duration/reward extremes increased over successive weeks. (B) Acquisition of sustained response performance during Free Choice training as measured by the percentage of total sucrose volume earned within each duration range and across 4 selected time points. The longest duration nosepokes, which resulted in the delivery of the largest magnitude rewards, increased sequentially over the course of training and accounted for the largest percentage of reward volume earned. (C) Acquisition of sustained response performance during Free Choice training as measured by reward efficiency (total volume/total number of nosepokes). The reward efficiency or volume of reward earned per nosepoke increased sequentially over the 4 selected weeks of Free Choice training. *P < 0.05 relative to the selected prior week.

Linear regression analysis revealed that the transition to performance marked by an increased percentage of longer duration responses was significantly correlated in a positive manner with the total volume of sucrose reward earned. During the first week of acquisition, only nosepokes of 800–1200 ms in duration correlated with the total volume of reward earned (r = 0.511, P < 0.001); however, that pattern shifted, and analysis of performance during Week 16 revealed that only those reinforced nosepokes with durations greater than 1200 ms significantly correlated with total sucrose volume (rs = 0.763, 0.852, and 0.761 for the 3 longest duration ranges, all Ps < 0.001). These results highlight the substantial impact that the observed shift to longer duration responses for larger magnitudes of sucrose had on the total volume of reward earned.

Sustained response performance within each Free Choice session also exhibited a distinctive pattern of choice impulsivity. Approximately 45% of all responses made and 43% of all sucrose reward earned occurred within the first 15-min interval, and both measures decreased successively across the remainder of the session (see Table 1). This initial behavioral interval was also marked by the lowest reward efficiency score, which increased significantly over the middle and last intervals. The decrease in choice impulsivity across the session is highlighted by examination of the briefest (0–400 ms) and longest (>2000 ms) duration responses. The highest number of brief responses occurred during the initial interval when the percentage of longest responses was at its lowest point, and during the middle and last intervals, brief responding decreased sequentially while the percentage of longest responses showed significant increases. The percentage of total volume earned by the briefest response remained relatively constant throughout the entire session, whereas the percentage of total volume earned by the longest response increased sharply in the middle and last intervals. Hence, higher choice impulsivity was evidenced early in the session when motivation for sucrose reward was high, followed by increased levels of sustained responding in the middle and end of the behavioral session illustrating decreased impulsivity and more efficient response patterns.

Table 1

Presurgery baseline performance

 Session interval
 
Response type 0–45 min 0–15 min 15–30 min 30–45 min 
All durations 
    Number of responses 71.0 ± 4.1 30.8 ± 1.5 22.2 ± 1.418.0 ± 1.6** 
    Percentage of total responding  45.5 ± 1.5 31.2 ± 0.823.3 ± 1.0** 
    Sucrose volume (mL) 8.71 ± 0.51 3.55 ± 0.21 2.93 ± 0.202.23 ± 0.19** 
    Percentage of total volume  42.6 ± 1.6 33.2 ± 0.924.2 ± 1.1** 
    Reward efficiency 0.128 ± 0.005 0.116 ± 0.005 0.138 ± 0.0060.130 ± 0.006
0–400 ms nosepoke 
    Number of responses 27.5 ± 2.2 11.8 ± 0.7 8.5 ± 0.87.3 ± 0.8** 
    Percentage of total responding 37.7 ± 1.5 37.9 ± 1.5 36.3 ± 1.6 37.1 ± 1.9 
    Percentage of total volume 17.5 ± 1.4 19.4 ± 1.4 16.8 ± 1.418.8 ± 1.7 
    Duration (ms) 140 ± 3 142 ± 4 138 ± 4 143 ± 5 
>2000 ms nosepoke 
    Number of responses 9.4 ± 1.1 3.0 ± 0.4 3.7 ± 0.4 2.7 ± 0.4*** 
    Percentage of total responding 14.7 ± 1.6 10.2 ± 1.4 18.7 ± 2.117.0 ± 2.0
    Percentage of total volume 33.6 ± 3.1 24.4 ± 2.7 38.0 ± 3.534.3 ± 3.4** 
    Duration (ms) 2977 ± 100 2650 ± 64 3090 ± 1173254 ± 139
 Session interval
 
Response type 0–45 min 0–15 min 15–30 min 30–45 min 
All durations 
    Number of responses 71.0 ± 4.1 30.8 ± 1.5 22.2 ± 1.418.0 ± 1.6** 
    Percentage of total responding  45.5 ± 1.5 31.2 ± 0.823.3 ± 1.0** 
    Sucrose volume (mL) 8.71 ± 0.51 3.55 ± 0.21 2.93 ± 0.202.23 ± 0.19** 
    Percentage of total volume  42.6 ± 1.6 33.2 ± 0.924.2 ± 1.1** 
    Reward efficiency 0.128 ± 0.005 0.116 ± 0.005 0.138 ± 0.0060.130 ± 0.006
0–400 ms nosepoke 
    Number of responses 27.5 ± 2.2 11.8 ± 0.7 8.5 ± 0.87.3 ± 0.8** 
    Percentage of total responding 37.7 ± 1.5 37.9 ± 1.5 36.3 ± 1.6 37.1 ± 1.9 
    Percentage of total volume 17.5 ± 1.4 19.4 ± 1.4 16.8 ± 1.418.8 ± 1.7 
    Duration (ms) 140 ± 3 142 ± 4 138 ± 4 143 ± 5 
>2000 ms nosepoke 
    Number of responses 9.4 ± 1.1 3.0 ± 0.4 3.7 ± 0.4 2.7 ± 0.4*** 
    Percentage of total responding 14.7 ± 1.6 10.2 ± 1.4 18.7 ± 2.117.0 ± 2.0
    Percentage of total volume 33.6 ± 3.1 24.4 ± 2.7 38.0 ± 3.534.3 ± 3.4** 
    Duration (ms) 2977 ± 100 2650 ± 64 3090 ± 1173254 ± 139

Note: Values are mean ± standard error of mean. The italicized and bolded values are statistically significant. *P < 0.05 compared with 0–15 min interval; **P < 0.05 compared with both 0–15 min and 15–30 min intervals; ***P < 0.05 compared with 15–30 min interval; n = 48.

The probability of emitting responses of different durations also illustrates the same pattern of choice impulsivity over the course of the behavioral session (see Supplementary Table 1). Responses of the briefest duration (0–400 ms) for the smallest reward size dominated behavior throughout the session, and examination of the second most frequent response following any initial response, independent of duration, also verified the tendency for higher choice impulsivity in the first 15-min interval of the session. The second most probable response shifted from the second briefest in duration (400–800 ms) in the first interval to the longest duration (>2000 ms) in the middle and last intervals. This initial impulsive choice pattern to responding is clearly demonstrated in the individual examples of cumulative responding presented in Figure 3. The behavioral performances are dominated by a high rate of responding (responses/min) and choice impulsivity in the first third of the session, followed by a decreased rate of responding and increase in longest duration responding for larger rewards in the latter two-thirds of the session.

Figure 3.

Cumulative response pattern during Free Choice. The examples shown are from 2 mPFC study rats taken from the same presurgical baseline testing session. Both traces illustrate sustained response performance dominated by a higher rate of responding (responses/min) and impulsive choice within the first 15-min interval. The pattern of responding in both traces show a decreased response rate and impulsive choice along with an increased delay between responses in the latter two-thirds of the testing session. The percentage of responding for maximal reward (>2000 ms) was roughly equivalent in both traces; however, the lower trace shows a higher percentage of impulsive responding (0–400 ms) for minimal reward (50 μL) relative to the upper trace, contributing to the smaller total volume of sucrose earned. The reward efficiency measure (total reward volume/total responses) also identifies the effect of a higher percentage of choice impulsivity in the lower trace relative to the upper trace (0.165 vs. 0.208). Each inset identifies the total volume of sucrose reward earned along with the percentage of responding at the briefest and longest duration for the smallest and maximal individual rewards, respectively.

Figure 3.

Cumulative response pattern during Free Choice. The examples shown are from 2 mPFC study rats taken from the same presurgical baseline testing session. Both traces illustrate sustained response performance dominated by a higher rate of responding (responses/min) and impulsive choice within the first 15-min interval. The pattern of responding in both traces show a decreased response rate and impulsive choice along with an increased delay between responses in the latter two-thirds of the testing session. The percentage of responding for maximal reward (>2000 ms) was roughly equivalent in both traces; however, the lower trace shows a higher percentage of impulsive responding (0–400 ms) for minimal reward (50 μL) relative to the upper trace, contributing to the smaller total volume of sucrose earned. The reward efficiency measure (total reward volume/total responses) also identifies the effect of a higher percentage of choice impulsivity in the lower trace relative to the upper trace (0.165 vs. 0.208). Each inset identifies the total volume of sucrose reward earned along with the percentage of responding at the briefest and longest duration for the smallest and maximal individual rewards, respectively.

Experiment 1: Effects of mPFC Lesions on Responding for Differential Reward

Histology of mPFC Sham and Lesion Groups

After 16 weeks of Free Choice training, 22 rats were assigned to mPFC sham and lesion groups in an attempt to form balanced groups based on their presurgical performance across the behavioral measures listed in Table 2. Histological analysis of the brains from the mPFC sham group showed no evidence of general damage or cell loss within any sectors of mPFC. In contrast, examination of brain sections from rats in the lesion group identified an area of gliosis and tissue loss that primarily encompassed the prelimbic and infralimbic sectors extending from +4.2 to +1.7 mm anterior to bregma. In all the lesioned rats, minor damage was seen dorsal to the prelimbic cortex border that included small portions of secondary motor cortex in anterior sections and small portions of the anterior cingulate cortex and precentral region in more posterior sections. Minor damage in a subset of rats was also evident ventral to the infralimbic border that included small portions of the medial and orbital cortex in anterior sections and small portions of the dorsal peduncular cortex in more posterior sections. Figure 4A illustrates representative hemisection photomicrographs of the neuronal damage and tissue loss within the mPFC for a lesioned rat and the same region for a rat from the sham group. Figure 4B depicts the minimal and maximal damage produced within the mPFC-lesioned rats.

Table 2

mPFC lesion effects on sustained responding

 mPFC sham
 
mPFC lesion
 
Behavioral measure Presurgery Postsurgery Percent change Presurgery Postsurgery Percent change 
Reinforced nosepoke number 73.1 ± 9.3 76.5 ± 10.5 3.6 ± 4.7% 62.9 ± 9.7 69.6 ± 10.9 13.6 ± 9.2% 
Nosepoke duration (ms) 1010 ± 93 1002 ± 123 −3.2 ± 4.9% 1049 ± 141 828 ± 138−20.1 ± 7.0% 
Sucrose volume (mL) 8.61 ± 0.60 8.89 ± 0.93 2.6 ± 6.1% 7.56 ± 1.27 6.75 ± 0.89 −0.1 ± 8.6% 
Reward efficiency 0.128 ± 0.009 0.128 ± 0.006 −1.1 ± 4.0% 0.128 ± 0.014 0.108 ± 0.008−14.2 ± 5.7% 
Number of nosepokes 
    0–400 ms 30.5 ± 5.1 35.1 ± 6.7 13.8 ± 7.5% 25.0 ± 5.3 33.0 ± 7.6 33.4 ± 12.7% 
    >2000 ms 8.5 ± 1.4 9.5 ± 2.2 −0.8 ± 14.0% 8.3 ± 2.2 4.9 ± 1.7 −46.6 ± 14.0%** 
Percentage of nosepokes 
    0–400 ms 39.6 ± 2.0 43.3 ± 2.7 10.3 ± 5.9% 38.3 ± 4.6 43.5 ± 5.618.7 ± 8.8% 
    >2000 ms 14.8 ± 2.9 16.0 ± 3.4 −4.2 ± 13.2% 15.4 ± 3.6 9.5 ± 3.6−50.5 ± 12.1%** 
 mPFC sham
 
mPFC lesion
 
Behavioral measure Presurgery Postsurgery Percent change Presurgery Postsurgery Percent change 
Reinforced nosepoke number 73.1 ± 9.3 76.5 ± 10.5 3.6 ± 4.7% 62.9 ± 9.7 69.6 ± 10.9 13.6 ± 9.2% 
Nosepoke duration (ms) 1010 ± 93 1002 ± 123 −3.2 ± 4.9% 1049 ± 141 828 ± 138−20.1 ± 7.0% 
Sucrose volume (mL) 8.61 ± 0.60 8.89 ± 0.93 2.6 ± 6.1% 7.56 ± 1.27 6.75 ± 0.89 −0.1 ± 8.6% 
Reward efficiency 0.128 ± 0.009 0.128 ± 0.006 −1.1 ± 4.0% 0.128 ± 0.014 0.108 ± 0.008−14.2 ± 5.7% 
Number of nosepokes 
    0–400 ms 30.5 ± 5.1 35.1 ± 6.7 13.8 ± 7.5% 25.0 ± 5.3 33.0 ± 7.6 33.4 ± 12.7% 
    >2000 ms 8.5 ± 1.4 9.5 ± 2.2 −0.8 ± 14.0% 8.3 ± 2.2 4.9 ± 1.7 −46.6 ± 14.0%** 
Percentage of nosepokes 
    0–400 ms 39.6 ± 2.0 43.3 ± 2.7 10.3 ± 5.9% 38.3 ± 4.6 43.5 ± 5.618.7 ± 8.8% 
    >2000 ms 14.8 ± 2.9 16.0 ± 3.4 −4.2 ± 13.2% 15.4 ± 3.6 9.5 ± 3.6−50.5 ± 12.1%** 

Note: Values are mean ± standard error of mean. The italicized and bolded values are statistically significant. *P < 0.05 compared with mPFC lesion, presurgery; **P < 0.05 compared with mPFC sham, percent change; sham, n = 11 and lesion, n = 11.

Figure 4.

mPFC lesion histology. (A) Representative photomicrographs of nissl stained coronal hemisections from mPFC sham and lesion brains. The left photomicrograph was taken from a sham-operated brain and illustrates the absence of any substantial neuronal or tissue loss throughout the medial wall of PFC. In contrast, the photomicrograph on the right was taken from a neurotoxic lesioned brain and illustrates the average degree of neuronal and tissue loss that encompassed primarily the prelimbic and infralimbic sectors of the mPFC while sparing the adjacent dorsal and ventral portions in the medial wall. (B) Coronal schematics from Paxinos and Watson (1998) depicting the smallest (black filled) and largest (gray filled) lesion area within the medial wall of PFC throughout the A/P gradient. The smallest and largest lesion depicted reflects the estimation of lesion size at each A/P level and does not necessarily reflect the lesion size from the same brain across all levels. Numbers to the bottom right of each schematic are the A/P coordinates relative to bregma. AC, anterior cingulate; DP, dorsal peduncular; IL, infralimbic; PC, precentral; PrL, Prelimbic.

Figure 4.

mPFC lesion histology. (A) Representative photomicrographs of nissl stained coronal hemisections from mPFC sham and lesion brains. The left photomicrograph was taken from a sham-operated brain and illustrates the absence of any substantial neuronal or tissue loss throughout the medial wall of PFC. In contrast, the photomicrograph on the right was taken from a neurotoxic lesioned brain and illustrates the average degree of neuronal and tissue loss that encompassed primarily the prelimbic and infralimbic sectors of the mPFC while sparing the adjacent dorsal and ventral portions in the medial wall. (B) Coronal schematics from Paxinos and Watson (1998) depicting the smallest (black filled) and largest (gray filled) lesion area within the medial wall of PFC throughout the A/P gradient. The smallest and largest lesion depicted reflects the estimation of lesion size at each A/P level and does not necessarily reflect the lesion size from the same brain across all levels. Numbers to the bottom right of each schematic are the A/P coordinates relative to bregma. AC, anterior cingulate; DP, dorsal peduncular; IL, infralimbic; PC, precentral; PrL, Prelimbic.

Presurgical Sustained Response Behavior

Following histological confirmation of the mPFC sham (n = 11) and lesion (n = 11) groups, analyses were conducted on the behavioral measures used to equate the 2 groups prior to surgery to verify their relative equivalence. One-way ANOVAs revealed no main effects of “Group” (sham vs. lesion) for the number of reinforced nosepokes (F1,20 = 0.58, P = 0.45), overall nosepoke duration (F1,20 = 0.06, P = 0.82), total sucrose volume (F1,20 = 0.54, P = 0.47), reward efficiency (F1,20 = 0.0003, P = 0.99), the number and percentage of the shortest duration nosepokes (0–400 ms) (F1,20 = 0.57, P = 0.46, and F1,20 = 0.14, P = 0.71, respectively), or the number and percentage of the longest duration nosepokes (>2000 ms) (F1,20 = 0.01, P = 0.92, and F1,20 = 0.01, P = 0.91, respectively) (see Table 2). The equivalent behavioral performance during the last 4 weeks prior to surgery established that any postsurgical differences between the groups would be due to neuronal loss within the mPFC and not to any a priori behavioral differences.

mPFC Lesion Effects on Sustained Response Behavior

Neurotoxic lesions of mPFC produced selective deficits in sustained responding for differential magnitudes of reward, and Table 2 summarizes the performance of both groups on the principal measures of sustained responding from pre- to postsurgical testing. Lesions of mPFC produced a significant decrease in the mean duration of nosepoke responses and reward efficiency without altering the total number of responses emitted or the total volume of sucrose earned. Two-way repeated measures ANOVAs determined that the total number of reinforced nosepokes and the total volume of sucrose earned were not significantly altered by mPFC lesions (no main effects of Group, F1,20 = 0.52, P = 0.48, and F1,20 = 1.58, P = 0.22, respectively) from pre- to postsurgical testing (no main effects of Surgery, F1,20 = 1.07, P = 0.31, and F1,20 = 0.31, P = 0.58, respectively) or between the mPFC sham and lesioned groups from pre- to postsurgical testing (no Group × Surgery interactions, F1,20 = 0.01, P = 0.93, and F1,20 = 1.35, P = 0.26, respectively). In contrast, the mean duration of all reinforced nosepokes and reward efficiency did not differ between the lesion and sham groups (no main effects of Group) but did significantly differ by Surgery (F1,20 = 7.08, P < 0.05, and F1,20 = 4.12, P < 0.05, respectively). These analyses also identified significant Group × Surgery interactions for nosepoke duration (F1,20 = 6.13, P < 0.05) and reward efficiency (F1,20 = 4.10, P < 0.05). Additional analyses conducted separately within each group determined that the performance of the mPFC sham group did not differ from pre- to postsurgical testing; however, mPFC lesions caused a significant postsurgical decrease in both overall nosepoke duration (F1,10 = 9.50, P < 0.05) and reward efficiency (F1,10 = 5.51, P < 0.05).

The decreases in overall nosepoke duration and reward efficiency caused by lesions to mPFC, in the absence of a significant change in the number of nosepokes emitted or total volume of reward earned, suggest that the underlying pattern of responding across the range of durations was selectively altered in the lesioned rats. Thus, performance across the 6 duration ranges was specifically examined before and after surgery and between sham and lesion groups. Three-way repeated measures ANOVA conducted on the number of reinforced nosepokes did not identify a main effect of Group (F1,20 = 0.36, P = 0.55), Surgery (F1,20 = 3.36, P = 0.08), or Group × Surgery interaction (F1,20 = 0.46, P = 0.50). Figure 5A,C displays the varying amount of responding across the range of durations for differential magnitudes of sucrose (main effect of Duration, F5,100 = 24.19, P < 0.0001) that also varied from pre- to postsurgical testing (Surgery × Duration interaction, F5,100 = 5.93, P < 0.0001). The 3-way Group × Surgery × Duration interaction was not statistically significant (F5,100 = 1.68, P = 0.15), suggesting that the pre- to postsurgical change in the number of responses across durations did not significantly vary between the sham and lesion groups; however, direct comparison between the groups using the percent change from presurgery baseline measure revealed that the number of nosepokes >2000 ms emitted by the mPFC lesion group decreased by approximately 50%, which was significantly greater than the percent change from presurgery baseline within the mPFC sham group (P < 0.05; see Table 2).

Figure 5.

mPFC lesion effects on sustained responding for differential magnitudes of reward. (A) The total number of reinforced nosepokes across increasing response durations that produced increasing magnitudes of sucrose reward by the mPFC sham group during pre- and postsurgical testing. (B) The percentage of total nosepokes emitted across increasing response durations by the mPFC sham group during pre- and postsurgical testing. The number or percentage of total nosepokes across durations did not significantly shift from pre- to postsurgical testing in the mPFC sham group. (C) The total number of reinforced nosepokes across increasing response durations that produced increasing magnitudes of sucrose reward by the mPFC lesion group during pre- and postsurgical testing. (D) The percentage of total nosepokes emitted across increasing response durations that corresponded to increasing volumes of sucrose reward by the mPFC lesion group during pre- and postsurgical testing. In contrast to the sham group, the mPFC lesion group showed a selective increase in the percentage of total nosepokes of the briefest duration for the smallest reward size while simultaneously showing a decreased percentage of nosepokes of the longest duration for the maximal reward size (*P < 0.05 relative to presurgical values).

Figure 5.

mPFC lesion effects on sustained responding for differential magnitudes of reward. (A) The total number of reinforced nosepokes across increasing response durations that produced increasing magnitudes of sucrose reward by the mPFC sham group during pre- and postsurgical testing. (B) The percentage of total nosepokes emitted across increasing response durations by the mPFC sham group during pre- and postsurgical testing. The number or percentage of total nosepokes across durations did not significantly shift from pre- to postsurgical testing in the mPFC sham group. (C) The total number of reinforced nosepokes across increasing response durations that produced increasing magnitudes of sucrose reward by the mPFC lesion group during pre- and postsurgical testing. (D) The percentage of total nosepokes emitted across increasing response durations that corresponded to increasing volumes of sucrose reward by the mPFC lesion group during pre- and postsurgical testing. In contrast to the sham group, the mPFC lesion group showed a selective increase in the percentage of total nosepokes of the briefest duration for the smallest reward size while simultaneously showing a decreased percentage of nosepokes of the longest duration for the maximal reward size (*P < 0.05 relative to presurgical values).

Similar analysis conducted between the sham and lesion groups on the percentage of nosepokes emitted across response durations revealed additional lesion-induced changes in the pattern of sustained responding. Three-way repeated measures ANOVA on the percentage of total nosepokes identified a main effect of Duration (F5,100 = 42.06, P < 0.001) that also significantly varied from pre- to postsurgical testing (Surgery × Duration interaction, F5,100 = 3.93, P < 0.005) and between sham and lesioned groups from pre- to postsurgery (Group × Surgery × Duration interaction, F5,100 = 2.38, P < 0.05). Additional within-subjects analyses were conducted separately within each Group to determine the source of the 3-way interaction. These analyses revealed that the percentage of nosepokes emitted within each duration range showed a trend toward a significant difference from pre- to postsurgical testing within the mPFC sham group (F5,50 = 2.37, P = 0.053; see Fig. 5B); however, there was a significant Surgery × Duration interaction for the mPFC lesion group (F5,50 = 3.47, P < 0.01; see Fig. 5D). Planned comparisons revealed that lesions of mPFC caused selective and inverse alterations in responding at the 2 extremes of the duration range, with an increased percentage of responding for the smallest sucrose reward (16% increase, P < 0.01) combined with a decreased percentage of responding for the largest sucrose reward (38% decrease, P < 0.01) when compared with their presurgery baseline values (see Supplementary Fig. 1 for individual examples). In addition, direct comparison of postsurgical performance between the groups using the percent change from presurgery baseline measure revealed that the percentage of nosepokes >2000 ms emitted by the mPFC lesion group decreased by approximately 50%, which was significantly greater than the percent change from presurgery baseline within the mPFC sham group (P < 0.05; see Table 2). This bidirectional lesion effect on responding at the duration extremes, for the correspondingly smallest and largest magnitudes of reward, was universally present throughout the lesion group and independent from the individual pattern of baseline responding (see Supplementary Fig. 2).

The mPFC lesion-induced decrease in sustained responding for maximal reward was also independent of the general pattern of responding throughout the course of the 45-min Free Choice session. Lesioned animals displayed the same pattern of responding over the course of the 45-min session (see Supplementary Table 3) relative to presurgical baseline (see Table 1) and sham animals (see Supplementary Table 2). Performance during both pre- and postsurgical testing was dominated by brief responding and higher choice impulsivity in the first 15-min interval, followed by an increased percentage and mean duration of maximal duration responses in the latter two-thirds of the testing session.

mPFC Lesion Effects on Extinction of Sustained Response Behavior

Extinction testing was conducted over 5 days to examine the effects of mPFC lesions on sustained responding in the absence of reward. The absence of sucrose delivery rapidly decreased the total number of nosepokes as well as the number and percentage of nosepokes across the 6 duration ranges. Despite an equivalent total number of nosepokes emitted during the last week of Free Choice behavior, the mPFC lesion group made significantly fewer nosepokes during extinction relative to the sham group. Repeated measures ANOVA on the total number of nosepokes across the 5 days of extinction identified main effects of Group (F1,20 = 6.31, P < 0.05) and Day (F4,80 = 18.23, P < 0.001) and a Group × Day interaction (F4,80 = 2.91, P < 0.05). As illustrated in Figure 6A, planned comparisons revealed that the mPFC lesion group emitted significantly fewer total responses in the absence of reward during the first 3 days and the fifth day of extinction testing compared with the sham group (all Ps < 0.05). The enhanced extinction effect seen in the mPFC lesion group was not due to aberant extinction responding in the sham group since both groups exhibited significantly reduced responding in Day 1 of extinction testing relative to the last week of reinforced responding (P < 0.05).

Figure 6.

mPFC lesion effects on extinction responding. (A) Total nosepoke number plotted across the last postsurgical week of reinforced training and the first 5 days of extinction testing. Despite both the sham and lesion groups displaying an equivalent total number of nosepokes during reinforced responding, the mPFC lesion group displayed more rapid acquisition of extinction and decreased responding across the first 5 days of extinction testing. (B) Number of 0–400 ms duration nosepokes averaged over the last postsurgical week of reinforced training and over the course of the first 5 days of extinction testing. Similar to the total number of nosepokes, the mPFC lesion group exhibited more rapid acquisition of extinction with significantly fewer nosepokes of the briefest duration emitted on the first day of extinction training (*P < 0.05 relative to postsurgery Week 4 values; **P < 0.05 relative to both postsurgery Week 4 values and sham group; #P < 0.05 relative to sham group only).

Figure 6.

mPFC lesion effects on extinction responding. (A) Total nosepoke number plotted across the last postsurgical week of reinforced training and the first 5 days of extinction testing. Despite both the sham and lesion groups displaying an equivalent total number of nosepokes during reinforced responding, the mPFC lesion group displayed more rapid acquisition of extinction and decreased responding across the first 5 days of extinction testing. (B) Number of 0–400 ms duration nosepokes averaged over the last postsurgical week of reinforced training and over the course of the first 5 days of extinction testing. Similar to the total number of nosepokes, the mPFC lesion group exhibited more rapid acquisition of extinction with significantly fewer nosepokes of the briefest duration emitted on the first day of extinction training (*P < 0.05 relative to postsurgery Week 4 values; **P < 0.05 relative to both postsurgery Week 4 values and sham group; #P < 0.05 relative to sham group only).

Reinforced responding in the mPFC lesion group was characterized by a shift in the percentage of responding across nosepoke durations; consequently, extinction responding was further analyzed by evaluating the number of nosepokes across the range of durations. Repeated measures ANOVA revealed that responding significantly varied across nosepoke durations and extinction days between the sham and lesion groups (Group × Duration × Day interaction, F20,400 = 3.11, P < 0.001). This interaction was largely due to the difference in responding at the briefest duration (0–400 ms). Additional analysis focused solely on this duration identified main effects of Group (F1,20 = 5.70, P < 0.05) and Day (F4,80 = 26.98, P < 0.001), as well as a Group × Day interaction (F4,80 = 4.83, P < 0.005). As illustrated in Figure 6B, the mPFC lesion group made significantly fewer 0–400 ms nosepokes on the first day of extinction (P < 0.05), and this decreased responding approached significance on Days 2, 3, and 5 (P = 0.07) relative to the sham group. Interestingly, this was the same preferred response duration in mPFC-lesioned rats that was significantly elevated in percentage during reinforced responding.

Experiment 2: Effects of NAc-core Lesions on Responding for Differential Reward

Histology of NAc-core Sham and Lesion Groups

After 16 weeks of Free Choice training, 26 rats were assigned to NAc-core sham and lesion groups in an attempt to form balanced groups based on their presurgical performance (see Table 3). Histological analysis of the brains from the sham group showed no evidence of neuronal or tissue loss throughout the A/P extent of the core region of the NAc. In contrast, examination of brain sections from lesioned rats identified an area of gliosis and neuronal loss localized to the core region surrounding the anterior commissure. The excitotoxic damage extended from +2.7 to +0.48 mm anterior to bregma. Figure 7A illustrates representative hemisection photomicrographs of neuronal damage within the core region of the NAc from a lesioned rat and the same intact region from a sham-operated rat. Figure 7B depicts the minimal and maximal damage evidenced by the NAc-core-lesioned group. Six rats from the lesion group had either unilateral damage or less than 50% damage to the NAc-core (unilateral or bilateral) and were therefore excluded from the analysis.

Table 3

NAc-core lesion effects on sustained responding

 NAc-core sham
 
NAc-core lesion
 
Behavioral measure Presurgery Postsurgery Percent change Presurgery Postsurgery Percent change 
Reinforced nosepoke number 67.3 ± 8.1 65.4 ± 10.2 −7.0 ± 5.9% 69.4 ± 6.9 39.7 ± 7.0−43.8 ± 6.7%** 
Nosepoke duration (ms) 957 ± 118 883 ± 102 −5.2 ± 5.5% 1026 ± 97 982 ± 118 −5.4 ± 3.6% 
Sucrose volume (mL) 7.93 ± 0.89 7.63 ± 1.24 −8.8 ± 8.4% 8.90 ± 1.27 4.80 ± 0.89−47.3 ± 6.9%** 
Reward efficiency 0.122 ± 0.012 0.117 ± 0.007 −3.1 ± 3.7% 0.128 ± 0.010 0.118 ± 0.009 −7.1 ± 1.4% 
Number of nosepokes 
    0–400 ms 26.7 ± 4.7 24.6 ± 4.9 −12.2 ± 4.4% 25.6 ± 3.4 14.8 ± 2.8−42.6 ± 6.4%** 
    >2000 ms 7.7 ± 2.1 6.2 ± 1.9 −17.6 ± 11.1% 10.0 ± 3.0 5.2 ± 1.7−53.4 ± 7.1%** 
Percentage of nosepokes 
    0–400 ms 37.7 ± 3.1 36.4 ± 3.4 −1.5 ± 4.3% 36.3 ± 2.8 38.0 ± 3.1 4.9 ± 4.4% 
    >2000 ms 12.9 ± 4.0 10.0 ± 3.4 −27.4 ± 11.0% 14.1 ± 3.2 12.5 ± 3.0 −18.8 ± 10.2% 
 NAc-core sham
 
NAc-core lesion
 
Behavioral measure Presurgery Postsurgery Percent change Presurgery Postsurgery Percent change 
Reinforced nosepoke number 67.3 ± 8.1 65.4 ± 10.2 −7.0 ± 5.9% 69.4 ± 6.9 39.7 ± 7.0−43.8 ± 6.7%** 
Nosepoke duration (ms) 957 ± 118 883 ± 102 −5.2 ± 5.5% 1026 ± 97 982 ± 118 −5.4 ± 3.6% 
Sucrose volume (mL) 7.93 ± 0.89 7.63 ± 1.24 −8.8 ± 8.4% 8.90 ± 1.27 4.80 ± 0.89−47.3 ± 6.9%** 
Reward efficiency 0.122 ± 0.012 0.117 ± 0.007 −3.1 ± 3.7% 0.128 ± 0.010 0.118 ± 0.009 −7.1 ± 1.4% 
Number of nosepokes 
    0–400 ms 26.7 ± 4.7 24.6 ± 4.9 −12.2 ± 4.4% 25.6 ± 3.4 14.8 ± 2.8−42.6 ± 6.4%** 
    >2000 ms 7.7 ± 2.1 6.2 ± 1.9 −17.6 ± 11.1% 10.0 ± 3.0 5.2 ± 1.7−53.4 ± 7.1%** 
Percentage of nosepokes 
    0–400 ms 37.7 ± 3.1 36.4 ± 3.4 −1.5 ± 4.3% 36.3 ± 2.8 38.0 ± 3.1 4.9 ± 4.4% 
    >2000 ms 12.9 ± 4.0 10.0 ± 3.4 −27.4 ± 11.0% 14.1 ± 3.2 12.5 ± 3.0 −18.8 ± 10.2% 

Note: Values are mean ± standard error of mean. The italicized and bolded values are statistically significant. *P < 0.05 compared with NAc-core lesion, presurgery; **P < 0.05 compared with NAc-core sham, percent change; sham, n = 10 and lesion, n = 10.

Figure 7.

NAc-core lesion histology. (A) Representative photomicrographs of nissl stained coronal hemisections from NAc-core sham and lesion brains. The left photomicrograph was taken from a sham-operated brain and illustrates the absence of any substantial neuronal or tissue loss throughout the core region of the NAc. In contrast, the photomicrograph on the right was taken from a neurotoxic lesioned brain and illustrates the average degree of neuronal loss that was restricted primarily to the core region of the NAc. (B) Coronal schematics from Paxinos and Watson (1998) depicting the smallest (black filled) and largest (gray filled) lesion area within the NAc and throughout the A/P gradient. The smallest and largest lesion depicted reflects the estimation of lesion size at each level and does not necessarily reflect the lesion size from the same brain across all levels. Numbers to the bottom right of each schematic are the A/P coordinates relative to bregma. AC, anterior commissure; NAc-s, NAc-shell.

Figure 7.

NAc-core lesion histology. (A) Representative photomicrographs of nissl stained coronal hemisections from NAc-core sham and lesion brains. The left photomicrograph was taken from a sham-operated brain and illustrates the absence of any substantial neuronal or tissue loss throughout the core region of the NAc. In contrast, the photomicrograph on the right was taken from a neurotoxic lesioned brain and illustrates the average degree of neuronal loss that was restricted primarily to the core region of the NAc. (B) Coronal schematics from Paxinos and Watson (1998) depicting the smallest (black filled) and largest (gray filled) lesion area within the NAc and throughout the A/P gradient. The smallest and largest lesion depicted reflects the estimation of lesion size at each level and does not necessarily reflect the lesion size from the same brain across all levels. Numbers to the bottom right of each schematic are the A/P coordinates relative to bregma. AC, anterior commissure; NAc-s, NAc-shell.

Presurgical Sustained Response Behavior

Following the histological confirmation of the sham (n = 10) and lesion (n = 10) groups, analyses were conducted on the behavioral measures initially used to equate the 2 groups prior to surgery. The relative equivalence of the groups was confirmed by one-way ANOVAs that did not identify any main effects of Group for the number of reinforced nosepokes (F1,18 = 0.04, P = 0.84), overall nosepoke duration (F1,18 = 0.20, P = 0.66), total sucrose volume (F1,18 = 0.39, P = 0.54), reward efficiency (F1,18 = 0.16, P = 0.70), the number and percentage of the shortest duration nosepokes (0–400 ms) (F1,18 = 0.04, P = 0.85, and F1,18 = 0.003, P = 0.96, respectively), or the number and percentage of the longest duration nosepokes (>2000 ms) (F1,18 = 0.40, P = 0.54, and F1,18 = 0.05, P = 0.82, respectively) (see Table 3). Hence, any postsurgical differences between the sham and lesion groups would be due to neuronal loss within the NAc-core and not to any a priori differences in behavioral performance.

NAc-core Lesion Effects on Sustained Response Behavior

Neurotoxic lesions of the NAc-core produced selective deficits in sustained responding for differential magnitudes of reward. As summarized in Table 3, NAc-core lesions produced an overall decrease in the amount of responding and reward. The number of reinforced nosepokes, as well as the total volume of sucrose earned, significantly decreased from pre- to postsurgical testing (main effect of Surgery for reinforced nosepoke number, F1,18 = 24.54, P < 0.001, and reward volume, F1,18 = 20.55, P < 0.001), and, more importantly, there was a significant Group × Surgery interaction for reinforced nosepoke number (F1,18 = 19.11, P < 0.001) and sucrose volume (F1,18 = 15.40, P < 0.005). Additional within-group repeated analyses determined that nosepoke number and sucrose volume did not vary from pre- to postsurgical testing in the NAc-core sham group; however, a significant decrease in the number of reinforced responses (F1,9 = 30.26, P < 0.001) and total volume of sucrose earned (F1,9 = 28.69, P < 0.001) was seen in the NAc-core lesion group during postsurgical performance. These effects are paralleled by direct comparison between the sham and lesion groups using the percent change from presurgery baseline measures (see Table 3).

Despite the significant decreases in overall behavioral responding and earned rewards caused by NAc-core lesions, the mean duration of reinforced nosepokes did not significantly differ by Group (F1,18 = 0.31, P = 0.58) or Surgery (F1,18 = 2.51, P = 0.13), and there was no significant Group × Surgery interaction (F1,18 = 0.17, P = 0.69). These results suggest that the decrease in responding caused by lesions of the NAc-core was not selectively displayed within any specific response duration; for verification, additional analyses incorporating the range of response durations were conducted (see Fig. 8). A 3-way repeated measures ANOVA on the number of reinforced nosepokes revealed a main effect of Surgery (F1,18 = 24.53, P < 0.0001) and Duration (F5,90 = 25.27, P < 0.0001) along with a Group × Surgery interaction (F1,18 = 19.11, P < 0.0005) but no main effect of Group (F1,18 = 1.13, P = 0.30). The Group × Surgery × Duration interaction approached but did not reach significance (F5,90 = 2.20, P = 0.06), indicating that the significant decrease in responding by the NAc-core lesion group during postsurgical testing was equally distributed across all nosepoke durations (see Fig. 8C).

Figure 8.

NAc-core lesion effects on sustained responding for differential reward magnitudes. (A) The total number of reinforced nosepokes across increasing response durations that corresponded to increasing volumes of sucrose reward for the NAc sham and lesion groups during pre- and postsurgical testing. The NAc sham group did not demonstrate any significant alterations in nosepoke number across the range of nosepoke durations from pre- to postsurgical testing. (B) The percentage of total nosepokes emitted across increasing response durations for the NAc sham group during pre- and postsurgical testing, which remained stable from pre- to postsurgical testing. (C) The total number of reinforced nosepokes across increasing response durations for the NAc-core lesion group. In contrast to the sham group, the NAc-core lesion group exhibited a nonselective decrease in the number of nosepokes across all nosepoke durations. (D) The percentage of total nosepokes emitted across increasing response durations for the NAc sham group during pre- and postsurgical testing with the one exception of a postsurgical increase in the percentage of total nosepokes with durations of 400–800 ms in the NAc lesion group (*P < 0.05 relative to presurgical values).

Figure 8.

NAc-core lesion effects on sustained responding for differential reward magnitudes. (A) The total number of reinforced nosepokes across increasing response durations that corresponded to increasing volumes of sucrose reward for the NAc sham and lesion groups during pre- and postsurgical testing. The NAc sham group did not demonstrate any significant alterations in nosepoke number across the range of nosepoke durations from pre- to postsurgical testing. (B) The percentage of total nosepokes emitted across increasing response durations for the NAc sham group during pre- and postsurgical testing, which remained stable from pre- to postsurgical testing. (C) The total number of reinforced nosepokes across increasing response durations for the NAc-core lesion group. In contrast to the sham group, the NAc-core lesion group exhibited a nonselective decrease in the number of nosepokes across all nosepoke durations. (D) The percentage of total nosepokes emitted across increasing response durations for the NAc sham group during pre- and postsurgical testing with the one exception of a postsurgical increase in the percentage of total nosepokes with durations of 400–800 ms in the NAc lesion group (*P < 0.05 relative to presurgical values).

Lesions of the NAc-core unequivocally decreased responding for all magnitudes of reward; however, the severely attenuated level of overall responding precipitated a small but significant shift in the percentage of emitted nosepokes distributed across the range of durations. A 3-way repeated measures ANOVA on the percentage of total nosepokes during pre- and postsurgical testing, and across the range of durations, identified a significant main effect of Duration (F5,90 = 36.18, P < 0.001), as well as Surgery × Duration (F5,90 = 3.65, P < 0.005) and Group × Surgery × Duration interactions (F5,90 = 2.60, P < 0.05). Additional within-group repeated analyses determined that the NAc-core sham group exhibited a main effect of Duration (F5,45 = 15.45, P < 0.001) but no main effect of Surgery (F1,9 = 1.00, P = 0.34) or Surgery × Duration interaction (F5,45 = 1.94, P = 0.11) (see Fig. 8B). In contrast, analysis of the NAc-core lesion group identified a main effect of Duration (F5,45 = 21.76, P < 0.001) and a Surgery × Duration interaction (F5,45 = 3.98, P < 0.005); this interaction was due to a postsurgical increase only in the percentage of 400–800 ms duration responses (P < 0.005), while the percentage of nosepokes at all other durations did not significantly differ from presurgical testing (see Fig. 8D). In total, these results suggest that lesions of the NAc-core produce a larger overall effect on the number of nosepokes than on the pattern of responding across the range of nosepoke durations and corresponding magnitudes of reward.

NAc-core Lesion Effects on Extinction of Sustained Response Behavior

Extinction training was conducted to examine the effects of NAc-core lesions on sustained responding in the absence of reward. The removal of reward rapidly decreased the total number and duration of responses, as well as the number and percentage of nosepokes across the 6 response durations. Despite the decreased total number of nosepokes during the last week of reinforced responding, the NAc-core lesion group did not significantly differ from the sham group in the amount of extinction responding (see Fig. 9). Repeated measures ANOVA on the total number of responses across the 5 days of extinction testing identified a significant main effect of Day (F4,72 = 16.03, P < 0.001), but the decreased level of responding during extinction was equivalently expressed in both sham and lesion groups (no main effect of Group, F1,18 = 0.43, P = 0.52, and no Group × Day interaction, F4,72 = 0.42, P = 0.80). These results indicate that NAc-core lesions do not alter extinction learning in this procedure and that lesioned subjects are capable of emitting numbers of responses comparable with sham subjects. In addition, these extinction results suggest that the reduced level of reinforced responding exhibited by the lesioned subjects was still mediated by the presence of reward versus its absence.

Figure 9.

NAc core lesion effects on extinction responding. Total nosepoke number plotted across the last postsurgical week of reinforced training and the first 5 days of extinction testing. Despite the lesion group emitting significantly fewer responses during reinforced behavior, the total number of nosepokes emitted during the first day of unreinforced responding by the NAc core lesion group did not significantly differ from sham performance (*P < 0.05 relative to postsurgery Week 4 values; **P < 0.05 relative to sham group).

Figure 9.

NAc core lesion effects on extinction responding. Total nosepoke number plotted across the last postsurgical week of reinforced training and the first 5 days of extinction testing. Despite the lesion group emitting significantly fewer responses during reinforced behavior, the total number of nosepokes emitted during the first day of unreinforced responding by the NAc core lesion group did not significantly differ from sham performance (*P < 0.05 relative to postsurgery Week 4 values; **P < 0.05 relative to sham group).

Discussion

The current study utilized a novel decision-making task that incorporates differential levels of sustained responding for the receipt of differential magnitudes of reward. The contributions of the mPFC and NAc-core to decision-making processes over a range of sustained response durations that assess choice impulsivity have not been previously investigated. The present findings identify dissociable roles for the NAc-core and mPFC on initiating and maintaining sustained responding for differential magnitudes of reward, respectively. NAc-core-lesioned subjects made fewer responses overall but showed little alteration in the ability to sustain individual nosepoke responses or in choice impulsivity, while mPFC-lesioned subjects showed no change in their ability to emit responses but did so with increased choice impulsivity that was most prominently displayed as a decrease in the number and percentage of responses of the longest duration for receipt of the largest reward. The behavioral contributions of these interconnected structures were also dissociated under extinction testing conditions with mPFC-lesioned rats showing enhanced extinction, whereas NAc-core-lesioned rats exhibited normal response extinction despite the significant reduction in responding when rewarded.

Assessment of Decision Making and Choice Impulsivity Using the Sustained Response Task

The majority of studies conducted on the neural basis of decision making, either effort or delay based, make use of tasks that involve an initial decision between 2 competing responses, which by necessity includes a spatial component, and often rely upon the presentation of fixed trials that limit the total amount of responding (Floresco and Ghods-Sharifi 2007; Floresco, Tse, and Ghods-Sharifi 2008; Hauber and Sommer 2009; Salamone et al. 2007; Schweimer and Hauber 2006; Walton et al. 2006). We designed the current task to probe decision making in a different manner: animals have one response operandum, thereby reducing possible effects of spatial response bias, and performance is free-running in that animals choose when to initiate their behavior and can thereby control the total number of trials performed and total amount of reward earned in a given session. In addition, we did not food- or water-restrict subjects to allow the inherent rewarding properties of the sucrose solution to drive behavior. The primary distinction between the current task and other decision-making tasks is that animals must maintain their behavioral response over progressively greater lengths of time to attain progressively greater volumes of sucrose, and this may engage a continual online decision-making process of whether to continue the response or to withdraw. We suggest that this response maintenance, although not equivalent with other traditional tasks measuring larger expenditures of physical effort (i.e., climbing barriers or high-ratio lever pressing), primarily involves cognitive effort as it requires an active inhibition of the impulsive tendency to withdraw and an active maintenance of the nosepoke response. Notably, after withdrawal from the operandum, the delay to reward receipt is the same for all magnitudes of reward and independent of the preceding length of the sustained response; however, further study is required to determine if this task also depends on an evaluation of delay to reward as evidenced in other studies (Floresco, Tse, and Ghods-Sharifi 2008).

Interestingly, performance in the sustained response task was dominated by responses at the extremes of the performance range, which roughly correspond to the no effort/large effort and small/large reward options of most other decision-making tasks. The preferred nosepoke response was of the briefest duration (0–400 ms) for the smallest magnitude of reward, while the overwhelming majority of total reward volume was earned from responses of the longest duration (>2000 ms) (see Table 1). Despite this dichotomy however, and unlike many other decision-making tasks, the use of a range of response durations and corresponding reward magnitudes afforded voluntary responding at choices other than the extremes. Table 1 shows that responses outside the duration extremes made up approximately 50% of total responding and generated 50% of the total reward during baseline performance. In addition, responding at each of the 3 longest duration ranges, not just >2000 ms, was positively associated with earning larger total volumes of sucrose reward. These results suggest that when decision-making tasks provide the opportunity to work for rewards at other than the minimal and maximal extremes, overall performance on these responses can significantly contribute to the total amount of rewarded behavior.

mPFC Mediates Sustained Responding for the Receipt of Maximal Reward

Results from the current study examining the effects of lesions of mPFC, primarily affecting the prelimbic and infralimbic cortices, demonstrate that these cortical regions are important neural structures underlying choice impulsivity and the ability to sustain responding in order to obtain maximal reward. In their absence, more impulsive responding for less than maximal reward per trial ensued; however, the lesion-induced shift in the pattern of responding produced a “functional” level of choice impulsivity in which lesioned animals showed a small but nonsignificant decrease in the amount of sucrose reward earned relative to presurgical performance. Specifically, lesions of the medial wall of PFC did not significantly alter the total amount of responding or volume of sucrose reward earned; however, mPFC lesions shifted behavior away from longer duration responses that produced the largest magnitudes of reward and required more impulsive inhibition, demonstrating increased choice impulsivity for smaller individual rewards. This small but significant shift toward briefer sustained responses in mPFC-lesioned rats served to functionally compensate (helped to prevent a significant decrease in the overall level of reward intake) for the shift away from the longest duration responses that provided the largest individual rewards.

Functional impulsivity has been previously defined in terms of responding with little forethought and evaluation when the situation was optimal for such behavior (Evenden 1999); here, we use the term in reference to the maintenance of reward intake following alteration in the normal pattern of behavioral responding. This type of compensation was revealed by our use of free-running trials that allowed lesioned animals the opportunity to titrate their individual level of reward intake relative to their motivational baseline. Consequently, we have demonstrated that the increase in choice impulsivity induced by mPFC disruptions can remain functional within the larger context of goal-directed behavior. This finding indicates that the motivational influences, or desire for reward, underlying this goal-directed behavior were not impaired—in agreement with other studies of mPFC lesions (Walton et al. 2002, 2003)—even as the mPFC-lesioned animals no longer responded as efficiently in the decision-making task relative to presurgical performance. Notably, the finding that mPFC-lesioned animals responded significantly less at the shortest duration (0–400 ms) during extinction suggests that responding at the shortest duration under rewarded conditions was not merely impulsive action execution. However, future studies utilizing this task will need to incorporate a premature response measure (i.e., no reward for nosepokes less than 400 ms), which will allow for the measurement of response impulsivity, and control procedures for equating the temporal interval differences between brief and sustained responding (i.e., temporal discounting effects) in order to more effectively assess the effects of mPFC lesions in balancing response costs, reward benefits, and impulsive tendencies in mediating decision-making behavior.

Estimation of timing and other timing-related processes used in temporal order and interval schedule paradigms have been shown to be related to several brain regions, including the PFC, across a wide array of species (Hannesson et al. 2004; Thorpe et al. 2002; and for review, see Buhusi and Meck 2005; Rubia and Smith 2004). Other studies specifically focused on the PFC have found attenuated acquisition of timing processes (Dietrich and Allen 1998), general declines in timing abilities during performance that were interpreted as a general slowing of interval estimation (Dietrich et al. 1997; Thorpe et al. 2002), or no effects on the timing of individual stimuli but impaired timing when multiple stimuli were presented (Olton et al. 1988). Interval estimation is the timing process we hypothesize is mediating sustained responding for increasing magnitudes of reward in the current task. The present PFC lesion-induced shifts in performance were localized to longer duration responses that fell within ranges not externally cued by the nosepoke port lights and were therefore principally guided by internal time estimations. Lesion-induced slowing of time estimation would be hypothesized to increase the percentage of longer duration responses, which is directly opposite of the current findings and suggests that impairments in timing cannot fully account for the diminished manner in which lesioned animals sustained their responding for the largest magnitude rewards. In support of this contention, disruption of mPFC function by temporary inactivation appears not to involve disruption of time estimation per se but rather may reflect deficits in waiting to initiate a response (Narayanan et al. 2006). In addition, a recent study by Cheng et al. (2006) concluded that DA input to the dorsal striatum and not to the PFC, based on differential alteration of DA by cocaine and ketamine, is the critical component in interval timing estimation. Taken together, prior research suggests that the mPFC lesion induced shift away from longer duration responses for maximal reward and toward shorter durations for smaller rewards was not likely not caused by a gross impairment in timing the length of each individual response.

Another possibility could be that lesions of mPFC abolished the learned association between the duration of the response and the corresponding magnitude of reward. This seems unlikely, however, since recent evidence has suggested that mPFC (the prelimbic sector in particular) is not essential for the expression of action–outcome associations based on findings that inactivation did not alter sensitivity to outcome devaluation (Tran-Tu-Yen et al. 2009). The current finding that responding extinguished rapidly also suggests that mPFC-lesioned rats were able to process outcome information. Responding for maximal reward with nosepokes >2000 ms in duration persisted in the lesion group, albeit in a diminished amount (50% decrease), also suggests that the learned associations inherent to the sustained response task were not completely abolished. However, additional studies utilizing reversal testing conditions and outcome devaluation will need to be conducted to provide more direct evidence confirming the locus of the mPFC lesion-induced effect on sustained responding for differential reward magnitudes.

Previous research has also found no effect of mPFC lesions on short-term memory for varying reward magnitudes (DeCoteau et al. 1997) or on preferences for larger magnitude rewards (Schweimer and Hauber 2005). It also seems unlikely that the results can be explained by a working memory deficit, such that after initiating a nosepoke response, the lesioned animals were unable to stay on task to complete the response for the desired reward magnitude. Not only is there evidence that mPFC lesions may not produce working memory deficits at these short-time intervals (Gisquet-Verrier and Delatour 2006; Izaki et al. 2008), but these possibilities seem unlikely, given that the ability to sustain nosepokes >2000 ms persisted in the lesioned animals, although emitted in a significantly diminished manner. Nor can the mPFC lesion-induced deficits in sustained responding be attributed simply to elevated hyperactivity, given that total responding was not elevated, maximal duration responses continued to be emitted with equivalent mean duration (see Supplementary Tables 2 and 3), and responding during extinction was lower than in controls. The general pattern of responding across 15-min intervals within the Free Choice session emitted by lesioned animals was similar to their own presurgical baseline and that of the sham-lesioned group despite the alterations in the briefest and longest duration responses (see Supplementary Tables 25), again suggesting that the mPFC lesion-induced deficits in sustained responding were not caused by general alterations in responding or interval timing but reflect enhanced choice impulsivity.

The PFC has also been implicated in decision-making processes that involve resolving the influences of uncertainty stemming from the probability of expected reward and the risk or cost of the action required to procure the desired reward (for review, see Rushworth and Behrens 2008). In the current task, the outcome, in terms of probability and magnitude of reward following individual actions or responses of given durations, has a fixed relationship that is experienced over many, many instances; therefore, the degree of uncertainty over the presence or absence of reward and its corresponding magnitude is likely to be quite low or absent altogether. Despite this fixed relationship between action and outcome in the current task, the mPFC lesion-induced shift in sustained responding could have been the result of increased uncertainty of exactly what magnitude of reward was being earned on any given trial. Estimations of interval timing are obviously interwoven in this aspect of outcome uncertainty, but as stated earlier, mPFC-lesioned rats continued to respond at maximal durations (in a diminished manner), suggesting that interval timing estimation was not completely abolished. Reversal testing, however, will need to be conducted in order to more directly assess the effects of mPFC lesions on alterations in action–outcome uncertainty and whether lesioned animals can effectively process changes in certainty about the relationship between actions of a given length and outcomes varying in magnitude of reward.

The effects of mPFC lesions produced here are in general agreement with alterations in response choice seen in traditional “effort-based” decision-making tasks, whether caused by similar mPFC lesions or smaller lesions focused selectively on the anterior cingulate cortex (Schweimer and Hauber 2005; Walton et al. 2002, 2003;). Walton and colleagues found that mPFC lesions induced a shift in choice behavior away from responses that involved high effort for large rewards toward responses with low effort for small rewards using the T-maze task with climbing barriers. Effort and incentive motivation manipulations within these studies also verified that the shift away from high-effort responses could be reversed when the effort or size of reward was equated for both responses, suggesting that incentive motivation, the ability to maintain the representation of the large reward, and basic motor abilities were not affected by the lesion. In contrast, however, the T-maze studies were unable to identify any possible functional aspect in the shift to more impulsive choice behavior following PFC damage because this task uses a fixed 10-trial sequence preventing the animals from responding for the lesser reward on a greater number of trials, thereby preventing them from maintaining total reward intake. Additional studies are required to determine whether the lesion effects identified here are the combined result of damage to multiple subregions within the medial wall of PFC or primarily due to damage localized within individual subregions, such as the most rostral portions of the anterior cingulate cortex, which have been shown to produce shifts in impulsive choice responding away from responses requiring larger effort for larger rewards (Walton et al. 2003; Schweimer and Hauber 2005).

NAc-core Mediates Response Initiation Relative to Baseline Motivation

In direct contrast to the effects of mPFC lesions, lesions to the NAc-core did not produce impulsive choice responding, that is, they did not shift responding from long-duration responses for larger rewards to short-duration responses for smaller rewards. Lesioned animals were just as willing on average to maintain responding for the largest rewards as they were to respond briefly for small rewards; however, they responded significantly fewer times overall compared with their presurgical baseline and relative to shams, reducing the total volume of reward obtained per session. This overall decrease in behavioral output suggests that the NAc-core is an important neural substrate in the circuit translating baseline motivation into goal-directed actions, in agreement with previous findings (Balleine and Killcross 1994; Cardinal 2006; Mogenson et al. 1980; Nicola et al. 2004; Robbins and Everitt 1996). And although this is far from a novel suggestion, especially when considering the multitude of studies on mesolimbic DA input to the NAc that have supported this hypothesis (for review, see Berridge 2007; Robbins and Everitt 2007; Salamone 2007), the current behavioral paradigm provides a clear dissociation between effects on the rate of responding and choice impulsivity, thereby allowing a clear indication of the effect of selective damage to neurons within the NAc-core subregion.

Numerous studies have examined the effects of NAc DA depletion in effort-based tasks (T-maze with climbing barriers and escalating fixed-ratio schedules of reinforcement), finding decreases in response vigor or rate of responding, and the willingness to expend larger amounts of effort for the procurement of large rewards (Cousins et al. 1996; Mingote et al. 2005; Salamone et al. 2003). When the effort is made comparable across the choices for differential reward, NAc DA-depleted animals choose the larger reward, suggesting that the effects of DA depletion are specific to the effort-based aspects in the allocation of food-related responses. Likewise, excitotoxic lesions of the NAc-core have recently been shown to impair effort-based responding in the T-maze cost–benefit task (Hauber and Sommer 2009). In contrast to our study, Hauber and Sommer (2009) identified an effort-based locus to the deficit, whereas we found a general decrement in response output after excitotoxic NAc-core lesions. This difference could be due to the different levels of physical and cognitive effort required between the behavioral tasks used in these studies. Another possible difference lies in the distinct types of decision-making processes: The T-maze choice task requires the response decision to be made before initiating the response, whereas in the present task, the decision to sustain responding and inhibit withdrawal to receive larger reward can be made after response initiation by maintaining the nosepoke response. It is also possible that larger amounts of physical effort and/or lever-pressing behavior may be more susceptible to NAc-core manipulations than nosepoke behaviors, a dissociation that has been previously noted for DA antagonists (Ettenberg et al. 1981; Mekarski 1988).

NAc-core lesions have also been reported to shift responding away from a lever that provides larger delayed rewards and toward a lever that provides smaller more immediate rewards in a delay-discounting decision-making task (Cardinal et al. 2001), leading these authors to conclude that damage to the NAc-core region resulted in choice impulsivity, given that the effort required for either of the 2 rewards was equivalent. The current results contrast with these findings in that no alteration in choice impulsivity as defined by a shift to smaller more immediately available rewards was seen following lesions of the NAc-core. Interestingly, the shortest delay tested in the temporal discounting task was 10 s, whereas the maximum response duration required in the present study was just over 2 s (albeit the mean response duration was nearer to 3 s; see Table 1), and there was no explicit delay to reward upon response completion. It remains to be tested if a delay-dependent deficit would emerge after NAc-core lesions in the present task if it were explicitly designed such that longer response durations would significantly increase the delay to reward delivery. The present results, however, indicate that NAc-core lesions do not affect choice impulsivity in a behavioral paradigm based on free-running trials and Free Choice sustained responding across a range of reward magnitude options. Rather, these results reveal a distinction between lesion-induced decreases in baseline motivation/response initiation for reward and alterations in choice impulsivity.

The absence of a lesion-induced effect on extinction performance indicates that the core region of the NAc is not critically involved in the acquisition of extinction learning under the current testing conditions. The intact extinction behavior following NAc-core lesions also suggests that the animals were not impaired in utilizing the action–outcome association, supporting the contention that the decreases in overall performance during rewarded behavior were more connected to the loss of incentive motivation than to impulsiveness or impaired action–outcome processing.

Are the Dissociable Roles of the mPFC and NAc-core in Responding for Differential Magnitudes of Reward Cooperative?

The NAc has long been thought to be the limbic–motor interface in which converging inputs from several limbic-associated structures access motor output pathways in the conversion of motivation into voluntary behavioral action (Mogenson et al. 1980). The mPFC provides direct input onto medium spiny neurons within the NAc, with a dense innervation of the core subregion stemming from prelimbic, infralimbic, and cingulate sectors of mPFC (Brog et al. 1993; Groenewegen et al. 1999; Vertes 2004). This direct connectivity suggests that these 2 structures may normally function in a cooperative manner to support responding of maximal effort for maximal reward, while other inputs carrying information relevant to incentive motivation may additionally be integrated within the NAc-core to drive the total amount of responding for reward consumption. Notably, Hauber and Sommer (2009) have shown that the anterior cingulate and the NAc-core are required for normal performance in the T-maze cost–benefit task since asymmetrical unilateral excitotoxic lesions of the 2 regions impaired the bias toward the high-effort/large reward response. In addition, Hauber and Sommer (2009) carefully determined that the detrimental effects on effort-based decisions were not caused by alterations in reward memory, preference, discriminative cue responding, or action–outcome processing during reversal learning.

Several other structures within the larger corticolimbic network have direct connections with the mPFC and/or converge within the NAc-core and may also participate in behavioral responding in the sustained response task. These limbic areas include the orbitofrontal cortex, with its noted role in expected outcomes and delay-based decisions (Rudebeck et al. 2006; Schoenbaum et al. 1998, 2009), and the amygdala, which is important for establishing the learned significance of stimuli and for associating and accessing outcome representations (Balleine and Killcross 2006; Cardinal 2006; Tye et al. 2008, 2010; Winstanley et al. 2004), as well as mediating effort-based decisions via projections to mPFC (Floresco and Ghods-Sharifi 2007). It has yet to be determined what role if any that these other forebrain structures play in mediating decision-making abilities within the current sustained response task.

Summary

The results from the current study using a novel goal-directed task utilizing a single response, free-running trials and a range of duration/reward magnitude ratios have established dissociable roles for the NAc-core and mPFC for initiating and maintaining responding for the receipt of maximal reward, respectively, as well as responding during extinction. The absence of a significant decrease in total reward volume earned by mPFC-lesioned animals indicates that the increase in choice impulsivity caused by mPFC damage proved functional and was sensitive to alterations in expected outcomes since the briefest duration responding rapidly decreased in the absence of reward during extinction testing. In direct contrast, NAc-core-lesioned animals responded less frequently and earned less reward volume overall but did not demonstrate any increase in impulsive choice responding, nor did the diminution of reinforced responding reflect the complete failure of the goal to motivate responding since further decreases were evidenced in the absence of reward. The cooperative or independent nature of these dissociable roles in supporting goal-directed responding for differential magnitudes of reward has yet to be directly tested. In addition, future studies utilizing an explicit response impulsivity measure combined with motivational manipulations will help to further elucidate the distinct roles that the mPFC and NAc-core play in cognitive effort-based decisions when the incentive value of the reward is incremented or decremented, as is known to occur in many compulsive disorders and addictive behaviors.

Funding

State of California for medical research on alcohol and substance abuse through the University of California, San Francisco.

Supplementary Material

Supplementary material can be found at: http://www.cercor.oxfordjournals.org/

The authors wish to acknowledge and thank William Schairer for expert technical assistance with data program analysis and Dr Gemma Guillazo-Blanch for professional assistance with surgery and behavioral training. Conflict of Interest: None declared.

References

Balleine
B
Killcross
S
Effects of ibotenic acid lesions of the nucleus accumbens on instrumental action
Behav Brain Res
 , 
1994
, vol. 
65
 (pg. 
181
-
193
)
Balleine
BW
Neural basis of food-seeking: affect, arousal and reward in corticostriatolimbic circuits
Physiol Behav
 , 
2005
, vol. 
86
 (pg. 
717
-
730
)
Balleine
BW
Dickinson
A
Goal-directed instrumental action: contingency and incentive learning and their cortical substrates
Neuropharmacology
 , 
1998
, vol. 
37
 (pg. 
407
-
419
)
Balleine
BW
Killcross
S
Parallel incentive processing: an integrated view of amygdala function
Trends Neurosci
 , 
2006
, vol. 
29
 (pg. 
272
-
279
)
Berridge
KC
The debate over dopamine’s role in reward: the case for incentive salience
Psychopharmacology
 , 
2007
, vol. 
191
 (pg. 
391
-
431
)
Brog
JS
Salyapongse
A
Deutch
AY
Zahm
DS
The patterns of afferent innervation of the core and shell in the “accumbens” part of the rat ventral striatum: immunohistochemical detection of retrogradely transported fluoro-gold
J Comp Neurol
 , 
1993
, vol. 
338
 (pg. 
255
-
278
)
Buhusi
CV
Meck
WH
What makes us tick? Functional and neural mechanisms of interval timing
Nat Rev Neurosci
 , 
2005
, vol. 
6
 (pg. 
755
-
765
)
Cardinal
RN
Neural systems implicated in delayed and probabilistic reinforcement
Neural Netw
 , 
2006
, vol. 
19
 (pg. 
1277
-
1301
)
Cardinal
RN
Pennicott
DR
Sugathapala
CL
Robbins
TW
Everitt
BJ
Impulsive choice induced in the rats by lesions of the nucleus accumbens core
Science
 , 
2001
, vol. 
292
 (pg. 
2499
-
2501
)
Cardinal
RN
Winstanley
CA
Robbins
TW
Everitt
BJ
Limbic corticostriatal systems and delayed reinforcement
Ann N Y Acad Sci
 , 
2004
, vol. 
1021
 (pg. 
33
-
50
)
Cheng
RK
MacDonald
CJ
Meck
WH
Differential effects of cocaine and ketamine on time estimation: implications for neurobiological models of interval timing
Pharmacol Biochem Behav
 , 
2006
, vol. 
85
 (pg. 
114
-
122
)
Cousins
MS
Atherton
A
Turner
L
Salamone
JD
Nucleus accumbens dopamine depletions alter relative response allocation in a T-maze cost/benefit task
Behav Brain Res
 , 
1996
, vol. 
74
 (pg. 
189
-
197
)
Cousins
MS
Salamone
JD
Nucleus-accumbens dopamine depletions in rats affect relative response allocation in a novel cost/benefit procedure
Pharmacol Biochem Behav
 , 
1994
, vol. 
49
 (pg. 
85
-
91
)
Dalley
JW
Mar
AC
Economidou
D
Robbins
TW
Neurobehavioral mechanisms of impulsivity: fronto-striatal systems and functional neurochemistry
Pharmacol Biochem Behav
 , 
2008
, vol. 
90
 (pg. 
250
-
260
)
Dawe
S
Gullo
MJ
Loxton
NJ
Reward drive and rash impulsiveness as dimensions of impulsivity: implications for substance misuse
Addictive Behav
 , 
2004
, vol. 
29
 (pg. 
1389
-
1405
)
DeCoteau
WE
Kesner
RP
Williams
JP
Short-term memory for food reward magnitude: the role of the prefrontal cortex
Behav Brain Res
 , 
1997
, vol. 
88
 (pg. 
239
-
249
)
Dietrich
A
Allen
JD
Functional dissociation of the prefrontal cortex and the hippocampus in timing behavior
Behav Neurosci
 , 
1998
, vol. 
112
 (pg. 
1043
-
1047
)
Dietrich
A
Frederick
DL
Allen
JD
The effects of total and subtotal prefrontal cortex lesions on the timing ability of the rat
Psychobio
 , 
1997
, vol. 
25
 (pg. 
191
-
201
)
Ettenberg
A
Koob
GF
Bloom
FE
Response artifact in the measurement of neuroleptic-induced anhedonia
Science
 , 
1981
, vol. 
213
 (pg. 
357
-
359
)
Evenden
JL
Varieties of impulsivity
Psychopharmacology
 , 
1999
, vol. 
146
 (pg. 
348
-
361
)
Floresco
SB
Ghods-Sharifi
S
Amygdala-prefrontal circuitry regulates effort-based decision making
Cereb Cortex
 , 
2007
, vol. 
17
 (pg. 
251
-
260
)
Floresco
SB
St Onge
JR
Ghods-Sharifi
S
Winstanley
CA
Cortico-limibic-striatal circuits subserving different forms of cost-benefit decision making
Cogn Affect Behav Neurosci
 , 
2008
, vol. 
8
 (pg. 
375
-
389
)
Floresco
SB
Tse
MT
Ghods-Sharifi
S
Dopaminergic and glutamatergic regulation of effort-and delay-based decision making
Neuropsychopharmacology
 , 
2008
, vol. 
33
 (pg. 
1966
-
1979
)
Gisquet-Verrier
P
Delatour
B
The role of the rat prelimbic/infralimbic cortex in working memory: not involved in the short-term maintenance but in monitoring and processing functions
Neuroscience
 , 
2006
, vol. 
141
 (pg. 
585
-
596
)
Glimicher
PW
Rustichini
A
Neuroeconomics: the consilience of brain and decision
Science
 , 
2007
, vol. 
306
 (pg. 
447
-
452
)
Groenewegen
HJ
Wright
CI
Beijer
AV
Voorn
P
Convergence and segregation of ventral striatal inputs and outputs
Ann N Y Acad Sci
 , 
1999
, vol. 
877
 (pg. 
49
-
63
)
Hannesson
DK
Vacca
G
Howland
JG
Phillips
AG
Medial prefrontal cortex is involved in spatial temporal order memory but not spatial recognition memory in tests relying on spontaneous exploration in rats
Behav Brain Res
 , 
2004
, vol. 
153
 (pg. 
273
-
285
)
Hauber
W
Sommer
S
Prefrontostriatal circuitry regulates effort-related decision making
Cereb Cortex
 , 
2009
, vol. 
19
 (pg. 
2240
-
2247
)
Izaki
Y
Takita
M
Akema
T
Specific role of the posterior dorsal hippocampus-prefrontal cortex in short-term working memory
Eur J Neurosci
 , 
2008
, vol. 
27
 (pg. 
3029
-
3034
)
Joel
D
Weiner
I
Feldon
J
Electrolytic lesions of the medial prefrontal cortex in rats disrupt performance on an analog of the Wisconsin Card Sorting Test, but do not disrupt latent inhibition: implications for animal models of schizophrenia
Behav Brain Res
 , 
1997
, vol. 
85
 (pg. 
187
-
201
)
Kesner
RP
Ragozzino
ME
The role of the prefrontal cortex in object-place learning: a test of the attribute specificity model
Behav Brain Res
 , 
2003
, vol. 
146
 (pg. 
159
-
165
)
King
JA
Tenney
J
Rossi
V
Colamussi
L
Burdick
S
Neural substrates underlying impulsivity
Ann N Y Acad Sci
 , 
2003
, vol. 
1008
 (pg. 
160
-
169
)
Knight
RT
Staines
WR
Swick
D
Chao
LL
Prefrontal cortex regulates inhibition and excitation in distributed neural networks
Acta Psychol
 , 
1999
, vol. 
101
 (pg. 
159
-
178
)
Lee
D
Rushworth
MFS
Walton
ME
Watanabe
M
Sakagami
M
Functional specialization of the primate frontal cortex during decision making
J Neurosci
 , 
2007
, vol. 
27
 (pg. 
8170
-
8173
)
Liao
PC
Uher
R
Lawrence
N
Treasure
J
Schmidt
U
Campbell
IC
Collier
DA
Tchanturia
K
An examination of decision making in bulimia nervosa
J Clin Exp Neuropsychol
 , 
2008
, vol. 
11
 (pg. 
1
-
7
)
Mekarski
JE
Main effects of current and pimozide on prepared and learned self-stimulation behaviors are on performance not reward
Pharmacol Biochem Behav
 , 
1988
, vol. 
31
 (pg. 
845
-
853
)
Miller
EK
Cohen
JD
An integrative theory of prefrontal cortex function
Annu Rev Neurosci
 , 
2001
, vol. 
24
 (pg. 
167
-
202
)
Mingote
S
Weber
SM
Ishiwari
K
Correa
M
Salamone
JD
Ratio and time requirements on operant schedules: effort-related effects of nucleus accumbens dopamine depletions
Eur J Neurosci
 , 
2005
, vol. 
21
 (pg. 
1749
-
1757
)
Mogenson
GJ
Jones
DL
Yim
CY
From motivation to action: functional interface between the limbic and the motor system
Prog Neurobiol
 , 
1980
, vol. 
14
 (pg. 
69
-
97
)
Montojo
CA
Courtney
SM
Differential neural activation for updating rule versus stimulus information in working memory
Neuron
 , 
2008
, vol. 
59
 (pg. 
173
-
182
)
Narayanan
NS
Horst
NK
Laubach
M
Reversible inactivations of rat medial prefrontal cortex impair the ability to wait for a stimulus
Neuroscience
 , 
2006
, vol. 
139
 (pg. 
865
-
876
)
National Research Council.
Guide for the care and use of laboratory animals
 , 
1996
Washington (DC)
National Academy Press
Nicola
SM
Yun
IA
Wakabayashi
KT
Fields
HL
Cue-evoked firing of nucleus accumbens neurons encodes motivational significance during a discriminative stimulus task
J Neurophysiol
 , 
2004
, vol. 
16
 (pg. 
5812
-
5829
)
Olton
DS
Wenk
GL
Church
RM
Meck
WH
Attention and the frontal cortex as examined by simultaneous temporal processing
Neuropsychologia
 , 
1988
, vol. 
26
 (pg. 
307
-
318
)
Opris
I
Bruce
CJ
Neural circuitry of judgement and decision mechanisms
Brain Res Rev
 , 
2005
, vol. 
48
 (pg. 
509
-
526
)
Paxinos
G
Watson
CW
The rat brain in stereotaxic coordinates
 , 
1998
4th ed
New York
Academic Press Inc
Robbins
TW
Everitt
BJ
Neurobehavioural mechanisms of reward and motivation
Curr Opin Neurobiol
 , 
1996
, vol. 
6
 (pg. 
228
-
236
)
Robbins
TW
Everitt
BJ
A role for mesencephalic dopamine in activation: commentary on Berridge, 2006
Psychopharmacology
 , 
2007
, vol. 
191
 (pg. 
433
-
437
)
Roesch
MR
Taylor
AR
Schoenbaum
G
Encoding of time-discounted rewards in orbitofrontal cortex is independent of value representation
Neuron
 , 
2006
, vol. 
51
 (pg. 
509
-
520
)
Rubia
K
Smith
A
The neural correlates of cognitive time management: a review
Acta Neurobiol Exp
 , 
2004
, vol. 
64
 (pg. 
329
-
340
)
Rudebeck
PH
Walton
ME
Smyth
AN
Bannerman
DM
Rushworth
MF
Separate neural pathways process different decision costs
Nat Neurosci
 , 
2006
, vol. 
9
 (pg. 
1161
-
1168
)
Rushworth
MF
Behrens
TE
Choice, uncertainty and value in prefrontal and cingulate cortex
Nat Neurosci
 , 
2008
, vol. 
11
 (pg. 
389
-
397
)
Sakagami
M
Pan
X
Uttl
B
Behavioral inhibition and prefrontal cortex in decision-making
Neural Netw
 , 
2006
, vol. 
19
 (pg. 
1255
-
1265
)
Salamone
JD
Motivational views of reinforcement: implications for understanding the behavioral functions of nucleus accumbens dopamine
Behav Brain Res
 , 
2007
, vol. 
137
 (pg. 
3
-
25
)
Salamone
JD
Correa
M
Farrar
A
Mingote
SM
Effort-related functions of nucleus accumbens dopamine and associated forebrain circuits
Psychopharmacology
 , 
2007
, vol. 
191
 (pg. 
461
-
482
)
Salamone
JD
Correa
M
Mingote
S
Weber
SM
Nucleus accumbens dopamine and the regulation of effort in food-seeking behavior: implications for studies of natural motivation, psychiatry, and drug abuse
J Pharmacol Exp Ther
 , 
2003
, vol. 
305
 (pg. 
1
-
8
)
Salamone
JD
Cousins
MS
Bucher
S
Anhedonia or anergia? Effects of haloperidol and nucleus accumbens dopamine depletion on instrumental response selection in a T-maze cost/benefit procedure
Behav Brain Res
 , 
1994
, vol. 
65
 (pg. 
221
-
229
)
Salamone
JD
Steinpreis
RE
McCullough
LD
Smith
P
Grebel
D
Mahan
K
Haloperidol and nucleus accumbens dopamine depletion suppress lever pressing for food but increase free food consumption in a novel food choice procedure
Psychopharmacology
 , 
1991
, vol. 
104
 (pg. 
515
-
521
)
Schoenbaum
G
Chiba
AA
Gallagher
M
Orbitofrontal cortex and basolateral amygdala encode expected outcomes during learning
Nat Neurosci
 , 
1998
, vol. 
1
 (pg. 
155
-
159
)
Schoenbaum
G
Roesch
MR
Stalnaker
TA
Takahashi
YK
A new perspective on the role of the orbitofrontal cortex in adaptive behaviour
Nat Rev Neurosci
 , 
2009
, vol. 
10
 (pg. 
885
-
892
)
Schweimer
J
Hauber
W
Involvement of the rat anterior cingulate cortex in control of instrumental responses guided by reward expectancy
Learn Mem
 , 
2005
, vol. 
12
 (pg. 
334
-
342
)
Schweimer
J
Hauber
W
Dopamine D1 receptors in the anterior cingulated cortex regulate effort-based decision making
Learn Mem
 , 
2006
, vol. 
13
 (pg. 
777
-
782
)
Smari
J
Bouranel
G
Thornora Eidsdottir
S
Responsibility and impulsivity and their interaction in relation to obsessive-compulsive symptoms
J Behav Ther Exp Psychiatry
 , 
2008
, vol. 
39
 (pg. 
228
-
233
)
Thorpe
CM
Floresco
SB
Carr
JAR
Wilkie
DM
Alterations in time-place learning induced by lesions to the rat medial prefrontal cortex
Behav Processes
 , 
2002
, vol. 
59
 (pg. 
87
-
100
)
Tran-Tu-Yen
DAS
Marchand
AR
Pape
JR
Di Scala
G
Coutureau
E
Transient role of the rat prelimbic cortex in goal-directed behaviour
Eur J Neurosci
 , 
2009
, vol. 
30
 (pg. 
464
-
471
)
Tye
KM
Cone
JJ
Schairer
WW
Janak
PH
Amygdala neural encoding of the absence of reward during extinction
J Neurosci
 , 
2010
, vol. 
30
 (pg. 
116
-
125
)
Tye
KM
Stuber
GD
de Ridder
B
Bonci
A
Janak
PH
Rapid strengthening of thalamo-amygdala synapses mediates cue-reward learning
Nature
 , 
2008
, vol. 
453
 (pg. 
1253
-
1257
)
Vertes
RP
Differential projections of the infralimbic and prelimbic cortex in the rat
Synapse
 , 
2004
, vol. 
51
 (pg. 
32
-
58
)
Walton
ME
Bannerman
DM
Alterescu
K
Rushworth
MFS
Functional specialization within medial frontal cortex of the anterior cingulate for evaluating effort-related decisions
J Neurosci
 , 
2003
, vol. 
23
 (pg. 
6475
-
6479
)
Walton
ME
Bannerman
DM
Rushworh
MFS
The role of rat medial frontal cortex in effort-based decision making
J Neurosci
 , 
2002
, vol. 
22
 (pg. 
10996
-
11003
)
Walton
ME
Kennerley
SW
Bannerman
DM
Phillips
PEM
Rushworth
MFS
Weighing up the benefits of work: behavioral and neural analyses of effort-related decision making
Neural Netw
 , 
2006
, vol. 
19
 (pg. 
1302
-
1314
)
Winstanley
CA
Theobald
DEH
Cardinal
RN
Robbins
TW
Contrasting roles of basolateral amygdala and orbitofrontal cortex in impulsive choice
J Neurosci
 , 
2004
, vol. 
24
 (pg. 
4718
-
4722
)