Hippocampal regulation of context-dependent neuronal activity in the lateral amygdala

Abstract

Pavlovian fear conditioning is a robust and enduring form of emotional learning that provides an ideal model system for studying contextual regulation of memory retrieval. After extinction the expression of fear conditional responses (CRs) is context-specific: A conditional stimulus (CS) elicits greater conditional responding outside compared with inside the extinction context. Dorsal hippocampal inactivation with muscimol attenuates context-specific CR expression. We have previously shown that CS-elicited spike firing in the lateral nucleus of the amygdala is context-specific after extinction. The present study examines whether dorsal hippocampal inactivation with muscimol disrupts context-specific firing in the lateral amygdala. We conditioned rats to two separate auditory CSs and then extinguished each CS in separate and distinct contexts. Thereafter, single-unit activity and conditional freezing were tested to one CS in both extinction contexts after saline or muscimol infusion into the dorsal hippocampus. After saline infusion, rats froze more to the CS when it was presented outside of its extinction context, but froze equally in both contexts after muscimol infusion. In parallel with the behavior, lateral nucleus neurons exhibited context-dependent firing to extinguished CSs, and hippocampal inactivation disrupted this activity pattern. These data reveal a novel role for the hippocampus in regulating the context-specific firing of lateral amygdala neurons after fear memory extinction.

Figure 1.

(A) Bilateral cannulae placements in dorsal hippocampus (left) and unilateral electrode placements in the lateral nucleus of the amygdala (right). Coronal sections were based on Swanson (1992). (B) Mean (±SEM) percentage of freezing for all rats before conditioning (PRE; average of 3-min period before first conditioning trial) and after the first (EXT-Early) and last (EXT-Late) five trials extinction. There were no differences between rats that were to be treated with either saline (SAL) or muscimol (MUS) during subsequent retention tests. (C) Mean (±SEM) percentage of freezing during the retention tests in rats treated with either saline (SAL) or muscimol (MUS). Rats in each group were tested to a single CS in both the context that the CS was extinguished (consistent or CON) and another context (inconsistent or INCON).

Figure 2.

Mean (± SEM) spontaneous (pre-CS) firing rate among lateral nucleus neurons after saline (SAL) or muscimol (MUS) infusion into the dorsal hippocampus during the consistent (CON) and inconsistent (INCON) retention tests.

Figure 3.

(A) Representative single-unit waveforms in a lateral nucleus neuron recorded during consistent (left) and inconsistent (right) test session in a rat treated with saline. (B,C) Panels show peristimulus time histograms (summed over 10 CS trials) and associated spike rasters for representative single units recorded in saline- (B, for the neuron in A) or muscimol-treated (C) rats. Each rat was tested to a single CS both inside the extinction context (consistent or CON) and outside the extinction context (inconsistent or INCON).

Figure 4.

Normalized neuronal activity (“population averages”) for all CS-responsive neurons recorded in the lateral amygdala of saline- (A) or muscimol-treated (B) rats. Spike firing was summed across the 10 CS trials in each retention test, and post-CS activity (binned in 50-msec intervals) was normalized to the pre-CS baseline (a total of 500 msec before CS onset). Standard scores (z-scores) were averaged across all units in the CON (open bars) and INCON (filled bars) tests. (Shaded rectangles) The 2-sec CS.

Figure 5.

(A) Mean (±SEM) CS-evoked firing (z-scores) in lateral amygdala neurons 50–100 msec after CS onset after saline (SAL) or muscimol (MUS) infusion into the dorsal hippocampus during the consistent (CON) and inconsistent (INCON) retention tests. (B) Scatterplot of normalized spike firing (50–100 msec after CS onset) for each neuron recorded under SAL (solid circles) or MUS (open circles) in the consistent and inconsistent test sessions. Regression lines indicate the deviation from a slope of 1 (dashed line) under the different drug conditions.