Time determines the neural circuit underlying associative fear learning

Abstract

Ultimately associative learning is a function of the temporal features and relationships between experienced stimuli. Nevertheless how time affects the neural circuit underlying this form of learning remains largely unknown. To address this issue, we used single-trial auditory trace fear conditioning and varied the length of the interval between tone and foot-shock. Through temporary inactivation of the amygdala, medial prefrontal-cortex (mPFC), and dorsal-hippocampus in rats, we tested the hypothesis that different temporal intervals between the tone and the shock influence the neuronal structures necessary for learning. With this study we provide the first experimental evidence showing that temporarily inactivating the amygdala before training impairs auditory fear learning when there is a temporal gap between the tone and the shock. Moreover, imposing a short interval (5 s) between the two stimuli also relies on the mPFC, while learning the association across a longer interval (40 s) becomes additionally dependent on a third structure, the dorsal-hippocampus. Thus, our results suggest that increasing the interval length between tone and shock leads to the involvement of an increasing number of brain areas in order for the association between the two stimuli to be acquired normally. These findings demonstrate that the temporal relationship between events is a key factor in determining the neuronal mechanisms underlying associative fear learning.

Keywords: amygdala; hippocampus; mPFC; muscimol; single-trial; trace fear conditioning.

Figure 1

Experimental design. (A) Diagram showing experimental protocol. Animals were pre-exposed to Context A and B for 15 min the day before conditioning. Context A and context B were use as training and testing environments in a counterbalanced fashion. (B) Photomicrographs showing example coronal sections of the brain of rats infused with fluorescently labeled muscimol targeted to BLA, mPFC, and dorsal-hippocampus.

Figure 2

Recruitment of amygdala, medial prefrontal-cortex but not dorsal-hippocampus in short tFC. (A) Bar graph showing percent time spent freezing during the tone CS, averaged across the three tones of the test session and across animals (average ± SEM) and during the first pre-CS period, in vehicle and muscimol infusion groups. (B) Cannula placement for both vehicle and muscimol infused rats in each structure: amygdala (BLA) [amyg-mus (n = 9); amy-veh (n = 8)], medial prefrontal-cortex (mPFC) [mPFC-mus (n = 7); mPFC-veh (n = 9)], and dorsal-hippocampus (dHipp) [dHipp-mus (n = 7); dHipp-veh (n = 6)]. *Denotes P < 0.017.

Figure 3

Recruitment of amygdala, medial prefrontal-cortex, and dorsal-hippocampus in long tFC. (A) Bar graph showing percent time spent freezing during the tone CS, averaged across the three tones of the test session and across animals (average ± SEM) and during the first pre-CS period, in vehicle and muscimol infusion groups. (B) Cannula placement for both vehicle and muscimol infused rats in each structure: amygdala (BLA) [amy-veh (n = 13); amy-mus (n = 14)]; medial prefrontal-cortex (mPFC) mPFC-veh (n = 9); mPFC-mus (n = 9); and dorsal-hippocampus (dHipp) [dHipp-veh (n = 9); dHipp-mus (n = 7)]. *Denotes P < 0.017.

Figure 4

Single-trial dFC is independent of mPFC activity. (A) Bar graph showing percent time spent freezing during the tone CS, averaged across the three tones of the test session and across animals (average ± SEM) and during the first pre-CS period, in vehicle and muscimol infusion groups. (B) Cannula placement for both vehicle and muscimol infused rats in medial prefrontal-cortex (mPFC) [mPFC-mus (n = 12); mPFC-veh (n = 8)].