Figure 5.

LHb inactivation before PMA retrieval test biases choice-mediated defensive behavior toward reward-seeking. A, left, Schematic representation of rats training. Rats were trained to associate tone with foot-shocks as well as associating the tone with moving to a safety platform learn to avoid the shock (day 1–10; PMA). Early PMA (day 2) represents classical conditioned-mediated threat learning (red), whereas late PMA (day 10) represents instrumental avoidance-mediated safety learning (green). One day later, rats were infused with SAL or M&B into LHb (gray arrow) before being presented with tones only (day 11, test), which involves the retrieval competition of coexisting threat (red) and safety (green) memories to control choice behavior. Right, Coronal drawings of rat brain atlas showing placements of the injector tips in LHb. B, During early PMA training (day 2), freezing and avoidance levels are similar. In contrast, during late PMA training (day 10), avoidance levels are higher than freezing levels. Thus, avoidance levels dramatically increased with PMA training while freezing levels remain similar (n = 15). C, LHb inactivation before PMA test (gray arrow and shaded area in graph), impaired retrieval of PMA memory, as indicated by lower tone-elicited percent freezing responses (but not suppression ratio) of the experimental group as compared with control group (SAL, n = 7; M&B, n = 8). D_, LHb inactivation before PMA test (gray arrow) biased choice-mediated avoidance toward reward-seeking behavior as indicated by lower tone-elicited percent avoidance (D₁) and time spent in platform (D₂), as well as lower suppression ratio (inset) in the experimental group as compared with the control group (SAL, n = 7; M&B, n = 8). Data are shown as ±SEM in blocks of two trials in B, C, D₁ and in 6-s bins in D₂. LHb, lateral habenula. Inf, infusion. Supp. ratio, suppression ratio; ****p < 0.00001, *p < 0.05.