Effects of quinolinic acid-induced lesions of the nucleus accumbens core on inter-temporal choice: a quantitative analysis

Abstract

Rationale: There is evidence that lesions of the nucleus accumbens core (AcbC) promote preference for smaller earlier reinforcers over larger delayed reinforcers in inter-temporal choice paradigms. It is not known whether this reflects an effect of the lesion on the rate of delay discounting, on sensitivity to reinforcer magnitude, or both.

Aim: We examined the effect of AcbC lesions on inter-temporal choice using a quantitative method that allows effects on delay discounting to be distinguished from effects on sensitivity to reinforcer size.

Materials and methods: Sixteen rats received bilateral quinolinic acid-induced lesions of the AcbC; 14 received sham lesions. They were trained under a discrete-trials progressive delay schedule to press two levers (A and B) for a sucrose solution. Responses on A delivered 50 microl of the solution after a delay d(A); responses on B delivered 100 microl after d(B). d(B) increased across blocks of trials, while d(A) was manipulated across phases of the experiment. Indifference delay d(B(50)) (value of d(B) corresponding to 50% choice of B) was estimated in each phase, and linear indifference functions (d(B(50)) vs d(A)) derived.

Results: d(B(50)) increased linearly with d(A) (r(2) > 0.95 in each group). The intercept of the indifference function was lower in the lesioned than the sham-lesioned group; slope did not differ between groups. The lesioned rats had extensive neuronal loss in the AcbC.

Conclusions: The results confirm that lesions of the AcbC promote preference for smaller, earlier reinforcers and suggest that this reflects an effect of the lesion on the rate of delay discounting.

Figure 1

Group mean data from the sham-lesioned group (upper panels) and AcbC-lesioned group (lower panels). Left-hand panels show preference functions (percent responding on lever B, %B, vs. delay to the larger of the two reinforcers after a response on B (d_B, s). Each set of points shows data collected from one phase of the experiment, in which the delay to the smaller reinforcer (d_A) was set at the value indicated (see inset). The horizontal reference line denotes indifference (%B=50). The intersection between each preference function and the indifference level denotes the indifference delay (d_B(50)) for that phase. Right-hand panels show transformations of the preference functions with d_B (s) on a logarithmic scale, and fitted logistic psychophysical functions, for each value of d_A. Note the leftward displacement and flatter slopes of the functions derived for the AcbC-lesioned group compared to those derived for the sham-lesioned group.

Figure 2

Linear indifference functions obtained for the sham-lesioned (empty circles) and AcbC-lesioned (filled circles) groups. Ordinate: indifference delay to the larger reinforcer (d_B(50), s); abscissa: imposed delay to the smaller reinforcer (d_A, s). Points show group mean data; vertical bars indicate SEMs; lines are best-fit linear functions. Equations for the fitted functions: sham-lesioned rats, d_B(50)=3.58d_A+6.39 (r²>0.97); AcbC-lesioned rats: d_B(50)=2.88d_A+0.45 (r²>0.95).

Figure 3

Parameters of the linear indifference functions from individual rats in the sham-lesioned (open columns) and AcbC-lesioned (filled columns) groups; columns show group mean values, vertical bars indicate SEMs. Left-hand panel: slope of linear function; right-hand panel: intercept of linear function. Significant difference between groups: * P<0.05.

Figure 4

Slope parameter (left-hand panel) and Weber fraction (right-hand panel) derived from psychophysical analysis of preference functions (see right-hand panels in Figure 1). See text for derivation of the Weber fraction. Columns show group mean data; vertical bars indicate SEMs. Significant difference between groups: * P<0.01.

Figure 5

Overall percentage preference for the larger reinforcer (%B) when delays to both levers (A and B) were set to 0 s in all blocks of trials. B>A denotes phase 7, where the reinforcer sizes were the same as in the previous 6 phases (B = 100 μl, A = 50 μl); B<A denotes phase 8 where the reinforcer sizes where switched. Columns show group mean data; vertical bars indicate SEMs.

Figure 6

Sample photomicrographs showing coronal sections of the brains of a sham-lesioned rat (panels a nand c) and a AcbC-lesioned rat (panels b and d). Left-hand panels: sections stained with cresyl violet; right-hand panels: sections stained for NeuN. LV, lateral ventricle; CPu, caudate-putamen; AcbC, nucleus accumbens core; AcbSh, nucleus accumbens shell; aca, anterior commisure. Note ventricular dilatation, neuronal loss and atrophy of the AcbC in the lesioned brain.

Figure 7

Diagram of approximate area of destruction of the AcbC in the lesioned group. Drawings were made from the microscopic sections, superimposed on the relevant pages of Paxinos and Watson’s (1998) stereotaxic atlas. Locations of the sections in the AP plane (mm anterior to bregma) are indicated on the left. The black area represents the smallest and the shaded area the largest lesion.

Figure 8

Theoretical preference and indifference functions, illustrating the implications of three alternative versions of the logistic psychometric preference function for the linear indifference function (see Discussion for details). A: Preference functions (%B vs. d_B, in semi-logarithmic coordinates) for a hypothetical sham-lesioned group; each curve corresponds to a different value of d_A; the curves were derived from equation 3. B, C and D: Corresponding preference functions for a hypothetical AcbC-lesioned group based on three models. In model 1 (panel B), the curves were derived from equation 3, under the assumption that the location parameter, d_B(50), was reduced in the lesioned group due to an increase in the rate of delay discounting. In models 2a and 2b (panels C and D), the curves were derived from equation 4 and 5, under the assumption that the reduction of d_B(50) in the lesioned group was brought about by downward displacement of the preference function (in equation 4, downward displacement is generated by a subtractive parameter b [b > 0]; in equation 5, downward displacement is generated by a multiplicative paramater max [max < 100]). Note that the curves in B, C and D are flatter than those in A, reflecting a lower value of ε in hypothetical AcbC-lesioned group, as seen in the present experiment. E, F and G Linear indifference functions corresponding to the preference functions shown in B, C and D. Note that downward displacement of the preference functions results in a reduction of the slope of the indifference function (F and G), in contrast to the parallel shift of the indifference function that is implied by a higher rate of delay discounting (E); the parallel shift is consonant with the empirical finding shown in Figure 2, indicating that model 1 (increased rate of delay discounting) provides a better account of the findings than models 2a and 2b (downward displacement).