Inhibition of the medial amygdala disrupts escalated aggression in lactating female mice after repeated exposure to male intruders

Abstract

Virgin female laboratory mice readily express pup care when co-housed with dams and pups. However, pup-sensitized virgins fail to express intruder-directed aggression on a single session of testing. To study whether repeated testing would affect the onset and dynamics of maternal or intruder-directed aggression, we tested dams and their accompanying virgins from postpartum day 4 to 6. Repeated testing led to escalated aggression towards male intruders in dams, but virgins never developed aggression. In dams, inhibition of the medial amygdala using DREADD (designer receptors exclusively activated by designer drugs) vectors carrying the hM4Di receptor blocked the expected increase in maternal aggression on the second testing day. Our data support that the onset of maternal aggression is linked to physiological changes occurring during motherhood, and that medial amygdala, a key centre integrating vomeronasal, olfactory and hormonal information, enables the expression of escalated aggression induced by repeated testing. Future studies selectively targeting specific neuronal populations of the medial amygdala are needed to allow a deeper understanding of the control of experience-dependent aggression increase, a phenomenon leading to the high aggression levels found in violent behaviours.

Fig. 1. Outline of the protocols followed in the study of maternal aggression behaviour.

a In Experiment 1, repeated aggression, females were randomly assigned into two groups: dams (n = 8) and pup-sensitised females (n = 8) (not represented). On postpartum days (PPD) 4–6, both groups of females were exposed to a different intruder male each day. b In Experiment 2, effect of DREADD-inhibition of MeA in aggression, females had stereotaxic injections of DREADD-AAV in the MeA. Then, they were mated and randomly assigned to two groups. After parturition, for behavioural testing, we followed an intragroup design. Group 1 received i.p. injection of CNO in PPD4 and Veh in PPD5, and Group 2 received Veh injection in PPD4 and CNO in PPD5. Then, 30 min after i.p. injections, females were exposed to a different male intruder each day. In parallel, we run two control groups of females, without DREADD injection, and the same drug regime, to discard any effect of CNO treatment (Supplementary information). CNO clozapine N-oxide, Veh vehicle.

Fig. 2. Lactating females displayed increasing levels of maternal aggression across testing days, whereas pup-sensitised females did not develop intruder-directed aggression.

a Total duration of attacks averaged across testing days was significantly higher in dams than in pup-sensitised virgins. a’ In lactating females, repeated testing leads to an increase in the total duration of attacks. a” Attacks were negligible in pup-sensitised females across testing days. b The latency to attack was significantly shorter in lactating females than in pup-sensitised virgin females. b’ In lactating females, the latency to the first attack was significantly shorter on the second day of testing than on the first. b” Since attacks were negligible in pup-sensitised females, latency to attack did not change across testing. c, d Social investigation was significantly higher in pup-sensitised females than in lactating dams, and did not significantly change across testing days, except for a trend towards decreased anogenital investigation in lactating dams (c’). Data are represented as mean ± SEM. **p ≤ 0.01, *p < 0.05, ^#p = 0.058.

Fig. 3. Injection sites in the MeA showing the rostro-caudal extent of the DREADD infections.

Photomicrographs showing the biggest (a, b) and the smallest (c, d) injection included in the analysis. Scale bar a, b: 500 µm and c, d: 100 µm. e, f Schematic reconstruction of coronal brain sections showing the extent of DREADD infections. Different coloured outlines represent the biggest (light pink) and the smallest (green) injection. The brain drawings are modified from ref. . ACo anterior cortical amygdaloid nucleus, BMA anterior basomedial amygdaloid nucleus, BSTIA bed nucleus of the stria terminalis intraamygdaloid division, CeM medial part of the central amygdaloid nucleus, ic internal capsule, MeAa anterior medial amygdaloid nucleus, MeApv posteroventral medial amygdaloid nucleus, opt optic tract (opt), MeApd posterodorsal medial amygdaloid nucleus, SI substantia innominate.

Fig. 4. Inhibition of MeA did not block maternal aggression but prevented its increase during the second day of testing.

In Group 1, injected with CNO in PPD4 and VEH in PPD5, the ANOVA showed significant differences between PPD4 and PPD5, with an increase in the total duration of aggression during the second session of testing (a) and a trend towards a decrease in latency to attack on the second day of testing (c). For Group 2, injected with Veh in PPD4 and CNO in PPD5, the ANOVA did not reveal significant differences between studied days in the total amount of time (b) or the latency to attack (d). Data are represented as mean ± SEM; *p ≤ 0.05, ^#p = 0.051. CNO clozapine N-oxide, Veh vehicle.

Fig. 5. Representation of the social brain network (SBN) nuclei involved in aggression in dam mice related to MeA projections.

Projections in red represent circuits with direct experimental evidence for the corresponding behaviour. AOB accessory olfactory bulb, BST bed nucleus of the stria terminalis, MeA medial amygdaloid nucleus, MOB main olfactory bulb, PAG periaqueductal grey, VMH ventromedial hypothalamic nucleus.