Traumatic Stress Induces Prolonged Aggression Increase through Synaptic Potentiation in the Medial Amygdala Circuits

Abstract

Traumatic stress can lead to heightened aggression which may be a symptom of psychiatric diseases such as PTSD and intermittent explosive disorder. The medial amygdala (MeA) is an evolutionarily conserved subnucleus of the amygdala that regulates attack behavior and behavioral responses to stressors. The precise contribution of the MeA in traumatic stress-induced aggression, however, requires further elucidation. In this study, we used foot shock to induce traumatic stress in mice and examine the mechanisms of prolonged aggression increase associated with it. Foot shock causes a prolonged increase in aggression that lasts at least one week. In vivo electrophysiological recordings revealed that foot shock induces potentiation of synapses formed between the MeA and the ventromedial hypothalamus (VmH) and bed nucleus of the stria terminalis (BNST). This synaptic potentiation lasts at least one week. Induction of synaptic depotentiation with low-frequency photostimulation (LFPS) immediately after foot shock suppresses the prolonged aggression increase without affecting non-aggressive social behavior, anxiety-like and depression-like behaviors, or fear learning. These results show that potentiation of the MeA-VmH and MeA-BNST circuits is essential for traumatic stress to cause a prolonged increase in aggression. These circuits may be potential targets for the development of therapeutic strategies to treat the aggression symptom associated with psychiatric diseases.

Keywords: PTSD; aggression; medial amygdala; synaptic plasticity; traumatic stress.

Figure 1.

Traumatic stress enhances aggression, fear learning, and anxiety-like and depression-like behaviors while preserving non-aggressive social interactions. A, Experimental schedule for B–F. Ten-week-old mice were individually housed for three weeks before foot shock. Control animals were placed into Context A for the same amount of time but with no foot shocks applied. Different cohorts were used for B, C–F, H–L. B, Analysis of freezing behavior in Context B. C–F, Analysis of attack behavior 7 d after foot shock. G, Experimental schedule for H–L. H, Time in the center of the arena in the open field test. I, Distance traveled during the open field test. J, Ratio of time spent interacting with the cup containing a mouse to the time interacting with the empty cup (SI score) during the social interaction test. K, Time spent in the light compartment during the light/dark box test. L, Time spent immobile during the forced swim test. Data are presented as mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001. Detailed statistics found in Table 2.

Figure 2.

Traumatic stress induces LTP of MeA-VmH and MeA-BNST synapses. A, Experimental schedule for in vivo electrophysiology. Individually housed mice were injected with ChR2 virus into the MeA and six weeks later were implanted with optrodes into the VmH, BNST, or LS. LFPs evoked by stimulating MeA axons at the VmH, BNST, or LS were recorded for 30 min. B, Representative image of ChR2 expression at the MeA. Scale bar: 200 μm. C, D, G, H, K, L, Illustrations for the sites of viral injection and optrode placement (C, G, K) and representative images of ChR2 expression in MeA axons at the VmH (D), BNST (H), or LS (L). Scale bars: 200 μm (D), 200 μm (H), and 100 μm (L). E, I, M, Representative traces of optically evoked responses at the VmH (E), BNST (I), or LS (M) recorded before, 1 d, and 7 d after foot shock or without foot shock. F, J, N, Average slopes of 90 light-evoked fEPSPs recorded at the VmH (F), BNST (J), or LS (N) before (pre), 1 d, and 7 d after foot shock or without foot shock (normalized to preshock). Data are presented as mean ± SEM; *p < 0.05, **p < 0.01. See Extended Data Figure 2-1 for more details. Detailed statistics found in Table 2.

Figure 3.

LFPS immediately after traumatic stress suppresses attack behavior. A, Experimental protocol for B–F. Mice were injected with ChR2 or GFP into the MeA and implanted with optical fibers above. After foot shock, LFPS was delivered immediately or 1 d after. Different mice were tested in B, C–F, H–L. B, Analysis of freezing behavior in Context B. C–F, Analysis of attack behavior 7 d after foot shock. G, Experimental protocol for H–L. H, Time in the center of the arena in the open field test. I, Distance traveled during the open field test. J, Ratio of time spent interacting with the cup containing a mouse to the time interacting with the empty cup (SI score) during the social interaction test. K, Time spent in the light compartment during the light/dark box test. L, Time spent immobile during the forced swim test. Data are presented as mean ± SEM; *p < 0.05, **p < 0.01. Detailed statistics found in Table 2.

Figure 4.

LFPS at the MeA-VmH and MeA-BNST synapses abolishes foot shock-induced synaptic potentiation. A, Experimental schedule for in vivo electrophysiology. LFPS was delivered to the VmH and BNST immediately after foot shock. B, E, Illustrations for the sites of viral injection at the MeA and optrode placement at the VmH (B) or BNST (E). C, F, Representative traces of optically evoked responses at the VmH (C) or BNST (F) recorded before, 1 d, and 7 d after foot shock. D, G, Normalized slopes of light-evoked fEPSPs recorded at the VmH (D) or BNST (G) before and after foot shock. Data are presented as mean ± SEM. See Extended Data Figure 4-1 for more details. Detailed statistics found in Table 2.

Figure 5.

LFPS of the MeA-VmH and MeA-BNST synapses suppresses foot shock-induced aggression increase. A, Experimental protocol. LFPS was delivered to the VmH or BNST immediately after foot shock. B–I, Analysis of attack behavior 7 d after foot shock. Data are presented as mean ± SEM; *p < 0.05, **p < 0.01. Detailed statistics found in Table 2.