LeDoux Lab 2008 SfN Abstracts
 
Program#/Poster#: 591.23/UU10
Title: Performance of well-trained sidman active avoidance in rats is amygdala-independent and constrained by activity in the central nucleus of the amygdala
Location: Washington Convention Center: Hall A-C
Presentation Time: Tuesday, Nov 18, 2008, 10:00 AM -11:00 AM
Authors: *G. LAZARO-MUNOZ, J. E. LEDOUX, C. K. CAIN;
Ctr. Neural Sci., New York Univ., New York City, NY
Abstract: Active avoidance (AA) protocols are ideal for examining the transition from ‘passive’ Pavlovian responding to ‘active’ instrumental responding in dangerous situations. A typical AA experiment involves teaching a rat to emit a shuttle response in order to prevent delivery of an aversive unconditioned stimulus (US; e.g. footshock). Environmental cues (conditioned stimuli, CSs) predictive of the US elicit Pavlovian reactions such as freezing early in AA training. However, as training progresses, these cues are believed to motivate and reinforce instrumental responding, and Pavlovian reactions indicative of fear cease to be expressed.
Formation and storage of Pavlovian CS-US associations is dependent on the lateral nucleus of the amgydala (LA). Previous work in our laboratory suggests that different outputs of the LA mediate ‘passive’ vs. ‘active’ responses to aversive cues. The central nucleus (CE) is critical for Pavlovian responding whereas the basal nucleus (B) appears to be critical for acquisition of instrumental responding. Gross manipulations of the amygdala in rabbits suggest that it is required for AA learning but not for performance following overtraining. We examined the contribution of LA, CE and B to performance of Sidman AA, with an emphasis on comparing Pavlovian responding (freezing) and instrumental responding (shuttling). Freezing is the dominant defensive behavior for rats; thus, in order to transition from fear reactions to actions we hypothesize that the CE must be suppressed; whereas once avoidance is well-learned, the response becomes an LA-, CE- and B-independent habit. Rats were given 15 daily Sidman sessions in which shuttling once every 30s prevented the delivery of footshocks. Following this overtraining, rats received bilateral electrolytic or sham lesions of LA, CE or B. After recovery, rats received 3 Sidman sessions. Sidman performance was unaffected by lesions of LA, CE or B in good performers after overtraining. After 10 Sidman sessions a subset of rats (~20%) failed to show acquisition of the AA response. We hypothesized that bilateral CE lesions to these poor performers would block CE-dependent freezing and allow for AA acquisition during retraining. Surprisingly, CE-lesioned poor performers showed asymptotic AA performance in the first post-lesion training session, suggesting that AA had been acquired but could not be performed pre-lesion due to excessive freezing driven by CE activity. Together, these results indicate that performance of well-learned Sidman AA is independent of LA, CE or B, and that CE-dependent Pavlovian responding needs to be suppressed to allow for AA performance.
Support: F32 MH077458
R37 MH038774
P50 MH058911
R01 MH046516
K05 MH067048