The role of medial amygdala inhibitory neurons in regulating social behavior

Aggression is an evolutionary conserved social behavior that is necessary for an animalÂŽs survival. The neural substrates coordinating aggression are known to exist in the amygdala and the hypothalamus. While the medial amygdala (MeA) and the ventrolateral subdivision of the ventromedial hypothalamus (VMHvl) have been implicated in aggression control and are known to be anatomically inter-connected, it has not yet been described at the circuit-level whether an interaction between these nuclei plays a role in aggression control.Posteriodorsal MeA (MeApd) is a majorly GABAergic nucleus (~70%), however the role of heterogeneous sub-populations of the MeA-GABA neurons in aggression is poorly understood. Since we were interested to identify the inhibitory cell types in the MeApd that control aggression, we first attempted to stimulate all the GABA neurons in the MeApd. Using a fast variant of channelrhodopsin-2E123T,H134R (ChETA) in VGATCre mice, we found that optogenetic stimulation of MeA-GABA neurons in the resident males interrupted intruder-directed aggression, contrary to the previous findings by Hong et al., where stimulation of MeApd-GABA neurons using channelrhodopsin-2H134R (ChR2) evoked aggression. An in-depth investigation of this discrepancy, using two AAV serotypes, revealed that optogenetic stimulation with ChETA suppressed aggression, whereas optogenetic stimulation with ChR2 increased aggression, irrespective of the AAV serotype. Recordings of the membrane potential changes reported larger plateau depolarizations, smaller action potential amplitudes, and larger local inhibition when using ChR2 as compared to ChETA. Thus, in the first part I showed that optogenetic stimulation of the brain areas with heterogenous population of interconnected GABA neurons such as the MeApd, can strongly depend on the properties of optogenetic tools.In the second part, we further explored how the MeApd-GABA neurons control aggression via long-range projections to their target nuclei in hypothalamus. Anterograde tracing revealed synapses from the MeApd-GABA neurons in the ventral premammillary (PMv) and the lateral ventromedial hypothalamic nuclei (VMHvl); both nuclei were previously identified in positive control of aggression (Lin et al., 2011; Stagkourakis et al., 2018). Ex-vivo optogenetic mapping of synaptic outputs from the MeApd-GABA projecting axons revealed evoked IPSCs in glutamatergic neurons in the VMHvl and the PMv, suggesting likely mechanism for inhibiting aggression. In-vivo optogenetic activation of the MeApd-GABA axonal terminals above the VMHvl or the PMv had no effect on aggression and instead led to an increased mounting, likely caused by optic fiber implantation in the hypothalamus. To avoid this artifact, we have combined retrograde AAV with Cre ON/Flip ON approach and specifically stimulated the VMHvl projecting MeApd-GABA neurons. This approach revealed that in aggressive mice, stimulation of this pathway tends to reduce aggression. Finally, we performed in-vivo Ca2+ imaging of the VMHvl projecting MeApd-GABA neurons during the RI test, which showed that these neurons are largely multimodal and get activated during different sub-types of social behaviours including aggression.This study shows that direct inhibitory path from the MeApd-GABA neurons to the VMHvl might suppress aggressive behaviour. Altogether, this work provides a step forward in our understanding of the neural substrates underlying aggression.