Nucleus accumbens lesions decrease sensitivity to rapid changes in the delay to reinforcement

Abstract

Both humans and non-humans discount the value of rewards that are delayed or uncertain, and individuals that discount delayed rewards at a relatively high rate are considered impulsive. To investigate the neural mechanisms that mediate delay discounting, the present study examined the effects of excitotoxic lesions of the nucleus accumbens (NAC) on discounting of reward value by delay and probability. Rats were trained on delay (n=24) or probability discounting (n=24) tasks. Following training, excitotoxic lesions of the NAC were made by intracranial injections of 0.5 microl 0.15 M quinolinic acid (n=12) or vehicle (n=12) aimed at the NAC (AP +1.6, ML +/-1.5, DV -7.1). NAC lesions did not alter performance in animals tested with a constant delay (4s) or probability (0.4) of reinforcement. However, when tested with between session changes in the delay (0, 1, 2, 4, and 8s) of reinforcement, the lesioned rats had flatter discount curves than the sham group, indicating that they were less sensitive to frequent changes in the delay to reward. In contrast, the NAC lesions did not affect discounting of probabilistic rewards. NAC lesions impaired the ability to adapt to frequent between session changes in the delay to reward but did not increase or decrease discounting when the delay was held constant across sessions. NAC lesions may disrupt the ability of the animals to predict the timing of delayed rewards when the delay to reward is changed frequently.

Fig. 1

Histological reconstruction of the range of lesion damage. Grey area represents the rat with the largest lesion damage. Black area represents the rat with the smallest lesion damage. Left column—eight rats tested on delay discounting and included in the analysis. Right column—nine rats tested on probability discounting and included in the analysis.

Fig. 2

Indifference points for rats on delay discounting (upper panel) and probability discounting (lower panel) averaged across high and low start conditions for 10 days prior to surgery (5 days high start condition, 5 days low start) and the indifference points of the final 10 days of the 2 weeks of postoperative testing averaged across high and low start conditions (5 days high start, 5 days low start) for rats tested on delay discounting with a 4 second delay to the standard rein-forcer. There was a significant increase in indifference points in both sham and lesioned animals on delay discounting following surgery. ^*p ≤ 0.05, **p ≤ 0.01.

Fig. 3

Determination of indifference points for rats tested on delay discounting before surgery (left columns) and post surgery (right columns) in shams (upper row) and lesioned rats (lower row). The dashed line indicates the point beyond which the median amount available on the adjusting alternative was used to determine indifference points. The same indifference points were reached whether the amount on the adjusting alternative is started at 25 µl or 150 µl both before and after surgery in both sham and lesioned rats.

Fig. 4

Determination of indifference points for rats tested on probability discounting before surgery (left columns) and postoperatively (right columns) in shams (upper row) and lesioned rats (lower row). The dashed line indicates the point beyond which the median amount available on the adjusting alternative was used to determine indifference points. The same indifference points were reached whether the amount on the adjusting alternative is started at 25 or 150 µl both before and after surgery in both sham and lesioned rats.

Fig. 5

Median indifference point for rats tested with 0, 1, 2, 4, and 8 s delays to the standard reinforcer collapsed across high and low start conditions taken from the last month of the delay curve phase. Lesioned rats discounted less than shams. ^**p≤ 0.01.

Fig. 6

Median indifference points for rats tested with 1.0, 0.7, 0.4, 2.0, and 0.1. Probabilities of receiving the standard reinforcer, transformed as odds-against winning and collapsed across high and low start conditions taken from the last month of the probability curve phase.

Fig. 7

Parametric challenges. (A) Effects of challenging rats trained on delay discounting with a 4 s delay to the standard reinforcer with delays of 2 and 8 s on indifference points. Lesioned rats were significantly less responsive to the delay challenge than shams. (B) Effects of challenging rats trained on probability discounting with a 40% probability of receiving the standard reinforcer with probabilities of 20% and 70% on indifference points. (C) Effects of challenging rats on delay discounting with 75 and 300 µl delayed water rewards. Lesioned rats were significantly less responsive to the magnitude challenge than shams. (D) Effects of challenging rats on probability discounting with 75 and 300 µl probabilistic water rewards. (E) Effects of challenging rats on delay discounting with ITIs of 5 and 60 s. Indifference points significantly decreased as ITI increased. (F) Effects of challenging rats on probability discounting with ITIs of 5 and 60 s. Indifference points significantly increased as ITI increased. ^**p ≤ 0.01.