Memory for reward location is enhanced even though acetylcholine efflux within the amygdala is impaired in rats with damage to the diencephalon produced by thiamine deficiency

Abstract

A rodent model of diencephalic amnesia produced by thiamine deficiency (pyrithiamine-induced thiamine deficiency [PTD]) was implemented to assess both changes in behavior and acetylcholine (ACh) efflux in the amygdala across four training sessions of a delayed alternation task. Two versions of the delayed alternation task were used. In one version, when a correct alternation was made a unique reward was paired with each spatial location ([left arm-chocolate milk] or [right arm-rat chow]). This paradigm is called the differential outcomes procedure (DOP). In the second version of the task, correct delayed alternation resulted in the same rewards but randomized across location (Nondifferential Outcomes Procedure [NOP]). The PTD rats were impaired on the first session of delayed alternation testing. However, both control and PTD rats using the DOP performed significantly better on delayed alternation than rats trained with the NOP.This effect was driven primarily by the PTD rats in the DOP condition outperforming all other groups on sessions 2-4. Although ACh efflux in the amygdala increased during delayed alternation testing in all groups, the NOP-trained rats had a greater rise in training-related ACh release in the post-training period. This suggests that increased amygdalar cholinergic activation is more critical for processing spatial information than episodic reward information. These data correspond with the idea that cholinergic activation of the amygdala promotes processing in other neural systems.

Figure 1

Panel A. Photomicrograph (4X) of the midline thalamus of a PTD rat displaying the prototypic PTD-induced lesion (arrows). Panel B. Photomicrograph (4X) of the midline thalamus of a PF rat. Panel C. An example of an acceptable probe placement in the amygdala. The placement of the 2-mm probe membrane is represented by the arrows. Although the probes were directed at the lateral amygdala, dialysis most likely included acetylcholine drawn from much of the amygdala. Abbreviations: BLA= basolateral amygdaloid nucleus, anterior; BLP= basolateral amygdaloid nucleus, posterior; CeL=central amygdaloid nucleus, lateral; CM= central medial thalamic nucleus; LHb= lateral habenular nucleus; MDL=mediodorsal thalamic nucleus, lateral; PC= paracentral thalamic nucleus; Po= posterior thalamic nucleus;

Figure 2

Panel A. Differences in accuracy performance on two variations of a delayed alternation task as a function of training across four sessions. PTD rats are represented by the red circles, whereas the PF rats are represented by the black squares. Rats trained with the DOP have solid lines and those trained with the NOP have dashed lines. Panel B. The PTD rats, regardless of reward parameters, where impaired on the first session of delayed alternation, relative to PF control rats (p<.01). Panel C. There was a main effect of training procedure (p<.01). Specifically, PTD rats (red circles) trained with the DOP had higher choice accuracy than those trained in the NOP condition (p<.05). Although there was a trend for the same phenomenon to occur in the PF rats (black squares), the effect did not reach significance.

Figure 3

Panel A. Increases in release of acetylcholine (ACh) in the amygdala during (M1, M2, M3) and after (A1, A2, A3) training on T-maze version of delayed alternation. Note that training resulted in increased release of ACh efflux in both PF and PTD rats (p<.01). However, the duration of release was reduced in PTD rats in the acute post maze period (p<.05). Panel B. Regardless of brain damage (PF [black square] vs. PTD [red circle]), when spatial location was not correlated with unique reward (NOP condition) there was higher ACh release in the amygdala during training (trend p=.09) and into the post training period (p<.05) than when spatial location was paired with unique rewards (DOP condition).