A critical role for cortical amygdala circuitry in shaping social encounters

Abstract

Aggression is an evolutionarily conserved behavior that controls social hierarchies and protects valuable resources like mates, food, and territory. In mice, aggressive behaviour can be broken down into an appetitive phase, which involves approach and investigation, and a consummatory phase, which involves biting, kicking, and wrestling. By performing an unsupervised weighted correlation network analysis on whole-brain c-Fos expression, we identified a cluster of brain regions including hypothalamic and amygdalar sub-regions and olfactory cortical regions highly co-activated in male, but not female aggressors (AGG). The posterolateral cortical amygdala (COApl), an extended olfactory structure, was found to be a hub region based on the number and strength of correlations with other regions in the cluster. Our data further show that estrogen receptor 1 (ESR1)-expressing cells in the COApl exhibit increased activity during attack behaviour, and during bouts of investigation which precede an attack, in male mice only. Chemogenetic or optogenetic inhibition of COApl ESR1 cells in AGG males reduces aggression and increases pro-social investigation without affecting social reward/reinforcement behavior. We further confirmed that COApl ESR1 projections to the ventrolateral portion of the ventromedial hypothalamus and central amygdala are necessary for these behaviours. Collectively, these data suggest that in aggressive males, COApl ESR1 cells respond specifically to social stimuli, thereby enhancing their salience and promoting attack behaviour.

Figure 1. The COApl is a key node in a differentially connected module in aggressive males.

a, Experimental timeline. b, Aggression duration in male and female AGGs were significantly higher than in NONs (main effect of phenotype, ${\mathrm{F}}_{\left(\mathrm{1,45}\right)}=88.31$, $p<0.0001$). c, NONs of both sexes exhibited more social investigation than AGGs (main effect of phenotype, ${\mathrm{F}}_{\left(\mathrm{1,45}\right)}=7.686$, $p=0.0081$, all post-hoc comparisons were $p>0.05$) n = 13 (Male AGG), 14 (Female AGG), 11 Male NON), 11 (Female NON). d, Whole-brain c-Fos network analysis in AGG and NON mice. Each module is arbitrarily assigned a unique colour label in the left and top bars of the topological overlap matrix (TOM). Darker colours indicate stronger c-Fos co-expression between brain regions. e, Network plot of the pink module. Module preservation analysis indicated that this module was the least preserved in terms of module connectivity in the NON network relative to the AGG network. f, Male AGGs exhibited higher pink module expression than female AGGs, ${\mathrm{t}}_{\left(25\right)}=4.612$, $p=0.0001$) g, Regions in the pink module which differed in intramodular connectivity between AGG and NON networks. Regions are ordered from highest to lowest average connectivity in the AGG network. Each point represents the weighted connectivity between the listed region and another region within the pink module. All $\mathrm{q}$ values were less than 0.05. h-j, c-Fos expression in the COAp. Both iDISCO+ (h) and FISH (j) revealed a significant phenotype × sex interaction (iDISCO+: ${\mathrm{F}}_{\left(\mathrm{1,45}\right)}=29.16$, $p<0.0001$; FISH: ${\mathrm{F}}_{\left(\mathrm{1,14}\right)}=28.93$, $p<0.0001$) with male AGGs exhibiting higher c-Fos than male NONs (iDISCO: $p<0.0001$. FISH: $p=0.0002$, Tukey’s post-hoc) and female AGGs (iDISCO+: $p<0.0001$. FISH: $p<0.0001$, Tukey’s post-hoc). Male AGGs also exhibited a higher number of c-Fos⁺ cells expressing Esr1 (phenotype x sex interaction: ${\mathrm{F}}_{\left(\mathrm{1,12}\right)}=10.56$, $p=0.007$) than male NONs ($p=0.0067$, Tukey’s post-hoc) and female AGGs ($p=0.0032$, Tukey’s post-hoc) n = 6 (Male AGG), 4 (Female AGG), 4 Male NON), 6 (Female NON) * p < 0.05, ** p < 0.01, *** p < 0.01, **** p < 0.001) See regions table for abbreviations.

Figure 2. In vivo activity of COApl^Esr1 cells during distinct bouts of investigation in male AGGs

a, Experimental timeline. b, Peri-event plot of COApl^Esr1 activity two seconds before and two seconds after the investigation of an odour; c, Quantification of activity before and after the event. There was a main effect of cue (${\mathrm{F}}_{\left(\mathrm{2,16}\right)}=5.407$, $\mathrm{p}=0.0161$) and a cue x time interaction (${\mathrm{F}}_{\left(\mathrm{2,16}\right)}=11.91$, $\mathrm{p}=0.0007$). Post-hoc comparisons revealed an increase in COApl^Esr1 activity only during the investigation of soiled bedding ($\mathrm{p}<0.0001$) d, f, h, peri-event plot of COApl^Esr1 activity four seconds before and four seconds after performing the behaviour of interest; e, g, i, quantification of activity before and after the event. When combining all bouts of investigation on days in which an attack did not occur compared to bouts of investigation on a day in which an attack occurred, there was a significant increase in activity during investigation on attack day (main effect of trial: ${\mathrm{F}}_{\left(\mathrm{1,5}\right)}=45.56$, $p=0.0010$; trial × time interaction: ${\mathrm{F}}_{\left(\mathrm{1,5}\right)}=10.31$, $p=0.0237$, d & e) compared to investigation bouts during a no-attack day ($p=0.0093$, Tukey’s post-hoc test, d & e). When examining investigation bouts during attack days, it was found that there was a significant increase in COApl^Esr1 activity during investigation bouts which preceded an attack (behaviour × time interaction: ${\mathrm{F}}_{\left(\mathrm{1,5}\right)}=25.17$, $p=0.0040$, f & g) compared to those which occurred in isolation ($p=0.0010$, Tukey’s post-hoc, f & g). When examining attack bouts, we found a significant difference in COApl^Esr1 activity before the onset of attack when mice were engaged in investigation prior to the attack (behaviour × time interaction: ${\mathrm{F}}_{\left(\mathrm{1,6}\right)}=5.668$, $p=0.0547$) compared to when they were not investigating ($p=0.0213$, Tukey’s post-hoc, h & i). Once the mice initiated the attack, we did not see any significant difference in COApl^Esr1 activity whether or not it was preceded by a bout of investigation ($p=0.8233$, Tukey’s post-hoc, h & i) $\mathrm{n}=7$, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.001.

Figure 3. Manipulation of COApl^Esr1 cells alters consummatory drive in male AGGs.

a,Experimental timeline. b, Virus expression in COApl. c-d, Inhibition of COApl^Esr1 cells significantly decreased aggression (c, virus × drug interaction: ${\mathrm{F}}_{\left(\mathrm{2,27}\right)}$, $p=0.0007$; Sidak’s post-hoc $p=0.0133$) and increased investigation (d, virus × drug interaction: ${\mathrm{F}}_{\left(\mathrm{2,27}\right)}=8.343$, $p=0.0015$; Sidak’s post-hoc, $p=0.0035$). Excitation of COApl^Esr1 cells significantly increased aggression (Sidak’s post-hoc, $\mathrm{p}=0.014$) without affecting investigation (Sidak’s post hoc, $p=0.2174$) n = 10 (HM4di), n = 10 (HM3dq), n = 10 (mCherry). e-f, Inhibition of COApl^Esr1 cells did not affect the latency to find hidden food (e, virus × drug interaction: ${\mathrm{F}}_{\left(\mathrm{1,16}\right)}=1.192$, $\mathrm{p}=0.2911$) or affect sex discrimination (f, virus × drug interaction: ${\mathrm{F}}_{\left(\mathrm{1,16}\right)}=.5122$, $p=0.4845$) n = 9 (HM4di), n = 9 (mCherry). g-i, In mice injected with hM4Di, we observed a significant negative correlation between the change in aggression and the change in investigation (g, $\mathrm{r}=-0.7668$, ${\mathrm{R}}^2=0.5880$, $p=0.0014$), but not in mice injected with hM3Dq (h, $\mathrm{r}=-0.5413$, ${\mathrm{R}}^2=0.2930$, $p=0.1061$) or mCherry (i, $\mathrm{r}=0.1559$, ${\mathrm{R}}^2=0.2431$, $p=0.5945$). j, Aggression self-administration experiments, both groups increased the number of rewarded trials with attacks (j, main effect of session: ${\mathrm{F}}_{\left(\mathrm{15,135}\right)}=5.412$, $p<0.0001$) equally. k, Inhibition of COApl^Esr1 cells had no effect on lever pressing (k, virus × drug interaction: ${\mathrm{F}}_{\left(\mathrm{1,18}\right)}=0.3147$, $p=0.5817$). l, However, inhibition of COApl^Esr1 led to a significant reduction in the percentage of rewarded trials which led to an attack (l, virus × drug interaction: ${\mathrm{F}}_{\left(\mathrm{1,18}\right)}=13.60$, $p=0.0022$; Sidak’s post hoc: $p<0.0001$) n = 11 (HM4di), n = 9 (mCherry). * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.001.

Figure 4. Manipulation of COApl^Esr1 cells alters connectivity in the pink module.

a,Experimental timeline. b-c, Attack duration (b, ${\mathrm{t}}_{\left(38\right)}=3.917$, $p=0.0007$) and investigation duration (c, ${\mathrm{t}}_{\left(38\right)}=4.723$, $p<0.0001$) change after COApl^Esr1 cells inhibition. d-e mCherry network demonstrated higher pink module expression (${\mathrm{t}}_{\left(38\right)}=4.442$, $p<0.0001$) when the two networks were combined, and better preserved the connectivity and density of the pink module than the hM4Di network (${\mathrm{t}}_{\left(9\right)}=3.983$, $p=0.0032$) when separately analyzed. Each pair of points connected by a line are a single preservation metric from each network. $\mathrm{n}=19$. (HM4di), 21 (mCherry) f, g, Whole brain c-Fos network in mice injected with mCherry (MC) or hM4Di (H4) and network plot of the pink module from the WT AGG network. Regions from the mCherry and hM4Di networks were plotted using the colour label from the original network. h, Of the 28 regions in the pink module, inhibition of COApl^Esr1 cells significant decreased c-Fos expression in 25 (89%) regions. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.001.

Figure 5. Optogenetic inhibition of regions downstream of COApl^Esr1 cells alters social behaviour in male AGGs

. a, Schematic experimental timeline. b, Regions which receive monosynaptic input from COApl^Esr1 cells as revealed by HSV-1 trans-synaptic anterograde tracing. c-d, Images of c-Fos⁺ cells in the VMH (c) and CEA (d) from mice injected with hM4Di or mCherry. e, Schematic behavioural setup. f, Inhibition of terminals in the VMHvl reduces attack duration in mice injected with NpHR (virus x laser interaction: $\mathrm{F}\left(\mathrm{1,14}\right)=7.893$, $p=0.0139$, Sidak’s post-hoc: $p=0.0048$) and increased investigation duration (virus x laser interaction: $\mathrm{F}\left(\mathrm{1,14}\right)=4.920$, $p=0.0436$, Sidak’s post-hoc: $p=0.0402$). g, Inhibition of terminals in the CEA reduces attack duration in mice injected with NpHR (virus × laser interaction: $\mathrm{F}\left(\mathrm{1,17}\right)=3.654$, $p=0.0729$, Sidak’s post-hoc: $p=0.0239$) but has no effect on investigation duration (virus × laser interaction: $\mathrm{F}\left(\mathrm{1,17}\right)=\mathrm{0,7045}$, $p=0.4192$, Sidak’s post-hoc: $p=0.2945)$ $\mathrm{n}=9$ (VMH NpHR), 7 (VMH YFP). 11 (CEA NpHR), 8 (CEA YFP). * p < 0.05, ** p < 0.01, *** p < 0.001.