Program#/Poster#: |
479.25/FF126 |
Title: |
Changes in behavioral decision making
due to aversive outcomes requires the lateral and basal nuclei of the amygdala
|
Location: |
South Hall A |
Presentation Time: |
Monday, Oct 19, 2009, 1:00 PM - 2:00
PM |
Authors: |
J. P. JOHANSEN, D. M. GIRARD,
J. E. LEDOUX;
Ctr. for Neural Sci., New York Univ., New York, NY |
Abstract: |
While the neural mechanisms mediating
fear learning have been well studied, relatively little known about the
neural mechanisms through which the probability of a specific action is
reduced when that action (by itself or in conjunction with environmental
cues) is associated with an aversive outcome. We used a two choice punishment
procedure in which animals pressing for food on two levers are then punished
(footshock) for pressing on one of the levers resulting in a reduction of
responding on the punished, but not on the non-punished, lever. To begin
to characterize the neural circuits mediating this change in response selection
we examined the contribution of the lateral (LA) and basal (B) nuclei of
the amygdala to this behavior. B neurons project to areas involved in goal
directed behavior such as the ventral striatum and we, and others (Belova
et al, 2007), have recently demonstrated that LA and B neurons encode a
prediction error-like signal for aversive events. These findings suggest
the hypothesis that LA and B neurons are part of an aversive teaching signal
pathway for this form of learning. To test this idea, LA and B neurons were
temporarily inactivated (intra-LA/B muscimol microinfusions) prior to punishment
training and animals were tested for their memory of the training 48 hours
later in a drug and shock free test session. We found that vehicle infused
animals exhibited a decrease in pressing on the punished lever after training,
but that this learning was significantly reduced in animals which had received
pre-training LA/B inactivations (consistent with Killcross et al, 1997).
Another potential prediction of our hypothesis is that LA and B inactivation
prior to overtraining in animals that have previously been well trained
should lead to extinction of punishment responding. To test this possibility,
animals received one day of punishment training and then vehicle or muscimol
was injected into the LA and B prior to a second day of overtraining and
learning was assessed drug free 48 hours later. We found that while vehicle
injected animals displayed a robust preference for the non-punished lever
following overtraining, LA and B inactivated animals increased their responding
on the punished lever such that pressing on both levers was indistinguishable.
This suggests that inactivation of LA and B neurons prior to overtraining
resulted in extinction of punishment learning. Together these results demonstrate
that the LA and B are required for both learning and maintaining action
selection under aversive contingencies and provide preliminary support for
the hypothesis that the LA and B are part of a teaching signal pathway for
the aversive regulation of action selection. |
Support: |
NIH NRSA 1F32MH082505-01A1 to JPJ |
|
R37 MH038774 to JEL |
|
P50 MH058911 to JEL |
|
R01 MH046516 to JEL |
|
|