Glutamate receptor antagonist infusions into the basolateral and medial amygdala reveal differential contributions to olfactory vs. context fear conditioning and expression

Abstract

The basolateral amygdala's involvement in fear acquisition and expression to visual and auditory stimuli is well known. The involvement of the basolateral and other amygdala areas in fear acquisition and expression to stimuli of other modalities is less certain. We evaluated the contribution of the basolateral and medial amygdala to olfactory and to context fear and fear conditioning by infusing into these areas the NMDA receptor antagonist AP5, the AMPA/kainate receptor antagonist NBQX, or vehicle prior to either odor-shock pairings or fear-potentiated startle testing. Pre-training AP5 infusions into the basolateral amygdala disrupted fear conditioning to the odor but not the context conditioned stimulus (CS). Pre-test NBQX infusions disrupted fear-potentiated startle to the odor but not context CS. Neither compound blocked fear conditioning when infused into the medial amygdala prior to training, but pre-test NBQX infusions did block fear-potentiated startle. The results confirm and extend recent findings suggesting a role for the basolateral amygdala in olfactory fear and fear conditioning, reveal an unexpected dissociation of the basolateral amygdala's involvement in discrete cue versus context fear and fear conditioning, and implicate for the first time the medial amygdala in fear-potentiated startle.

Figure 1.

Cannula tip placements transcribed onto atlas plates adapted from Paxinos and Watson (1997). Placements for rats included in the primary analyses of Experiments 1A, 1B, and 2 (basolateral amygdala) are indicated by filled circles, filled triangles (point up), and filled triangles (point down), respectively. Placements for rats included in the primary analyses of Experiments 3, 4A, and 4B (medial amygdala) are indicated by open circles, open triangles (point up), and open triangles (point down), respectively. The distance from bregma is indicated to the left; nuclei and major fiber bundles within the plane of section are identified to the right. BL, basolateral amygdaloid nucleus; BM, basomedial amygdaloid nucleus; Ce, central amygdaloid nucleus; Co, cortical amygdaloid nucleus; ic, internal capsule; L, lateral amygdaloid nucleus; Me, medial amygdaloid nucleus; ot, optic tract; st, stria terminalis.

Figure 2.

Effect on fear conditioning and fear-potentiated startle of glutamate receptor antagonist infusions into the basolateral amygdala. The NMDA receptor antagonist AP5 disrupted fear conditioning to an olfactory CS when infused prior to training (leftmost bars) but did not disrupt fear-potentiated startle when infused prior to testing (middle bars). The AMPA/kainate receptor antagonist NBQX did disrupt fear-potentiated startle when infused prior to testing (rightmost bars). Baseline startle amplitude (i.e., from noise-alone test trials), from which percent potentiation scores are derived, is indicated immediately below each bar. ^*P < 0.05 vs. ACSF control infusions.

Figure 3.

Effect on fear conditioning and fear-potentiated startle of glutamate receptor antagonist infusions into the medial nucleus of the amygdala. Neither AP5 nor NBQX disrupted fear conditioning to an olfactory CS when infused prior to training (leftmost bars). When infused prior to testing, NBQX disrupted fear-potentiated startle to an olfactory CS (middle bars) and to a visual CS (rightmost bars), and significantly reduced startle amplitude on noise-alone trials (indicated immediately below each bar). ^*P < 0.05 vs. ACSF control infusions.

Figure 4.

Effect of glutamate receptor antagonist infusions into the basolateral and medial nucleus of the amygdala on pre- to post-conditioning changes in baseline startle amplitude. Pre-test infusions of NBQX into the medial nucleus of the amygdala disrupted pre- to post-conditioning increases of startle amplitude on noise-alone trials (leftmost bars). In contrast, pre-test NBQX infusions into the basolateral amygdala did not disrupt these increases, nor did pre-training infusions of AP5 (rightmost bars). Baseline startle amplitude from the final pre-conditioning acclimation session, from which percent potentiation scores are derived, is indicated immediately below each bar.

Figure 5.

Pre-test infusions of NBQX into the medial nucleus of the amygdala did not influence startle amplitude in untrained rats.

Figure 6.

Pre-test NBQX infusions disrupt fear-potentiated startle to an olfactory CS via actions within the medial nucleus of the amygdala. For all NBQX-infused rats from Experiment 4, percent potentiation scores were significantly lower for rats that received bilateral infusions into the medial nucleus of the amygdala (MeA) compared to rats that received one infusion into the medial nucleus and a second infusion elsewhere (left). For these same animals, the distance of the infusion sites from the medial nucleus of the amygdala was significantly correlated with the magnitude of fear-potentiated startle to the olfactory CS (right). ^*P < 0.05 vs. ACSF control infusions.

Figure 7.

Previously unpublished findings from our laboratory suggest that the effect of pre-test NBQX infusions into the medial nucleus of the amygdala is not due to drug actions within the central nucleus of the amygdala. Whereas unilateral infusions of NBQX into the central nucleus potently disrupt fear-potentiated startle (C.-J. Shi and M. Davis, unpubl.), unilateral infusions into the medial nucleus of the amygdala do not (Z. Zhao and M. Davis, unpubl.). ^*P < 0.05 vs. ACSF control infusions.