Dissecting the brain's fear system reveals the hypothalamus is critical for responding in subordinate conspecific intruders

Abstract

Effective defense against natural threats in the environment is essential for the survival of individual animals. Thus, instinctive behavioral responses accompanied by fear have evolved to protect individuals from predators and from opponents of the same species (dominant conspecifics). While it has been suggested that all perceived environmental threats trigger the same set of innately determined defensive responses, we tested the alternate hypothesis that different stimuli may evoke differentiable behaviors supported by distinct neural circuitry. The results of behavioral, neuronal immediate early gene activation, lesion, and neuroanatomical experiments indicate that the hypothalamus is necessary for full expression of defensive behavioral responses in a subordinate conspecific, that lesions of the dorsal premammillary nucleus drastically reduce behavioral measures of fear in these animals, and that essentially separate hypothalamic circuitry supports defensive responses to a predator or a dominant conspecific. It is now clear that differentiable neural circuitry underlies defensive responses to fear conditioning associated with painful stimuli, predators, and dominant conspecifics and that the hypothalamus is an essential component of the circuitry for the latter two stimuli.

Fig. 1.

Photomicrographs of transverse Fos-stained sections of the dorsal premammillary nucleus from animals exposed to a predator (A) or to a dominant conspecific (B). Abbreviations: fx, fornix; PMDdm and -vl, dorsal premammillary nucleus, dorsomedial and ventrolateral parts. (Scale bars, 200 μm.)

Fig. 2.

Histograms showing for the control (n = 6), cat-exposed (n = 5), intruder (n = 7), and resident (n = 7) groups the counts of PMD Fos-immunoreactive cells (total count and specific counts for the dorsomedial and ventrolateral parts of the nucleus). Data are expressed as mean ± SEM. *, differs significantly from control group, P < 0.0002; #, differs significantly from intruder group, P < 0.0002; ○, differs significantly from intruder and cat-exposed groups, P < 0.0002; **, differs significantly from the ventrolateral part of the PMD of the same group, P < 0.0002.

Fig. 3.

Behavioral measurements for the sham (saline, n = 6), NMDA-control (n = 5), and NMDA-PMD lesioned (n = 9) intruders, during a 5-min observation period after the resident conspecific's initial attack. For conspecific intruders, the total amount of time spent in exploratory and social (approach, contact with the resident, sniffing, and anogenital sniffing) behaviors and in defensive behaviors was evaluated. Defensive behaviors were further separated into passive defense (including freezing and the typical sustained on-the-back position after the resident leaves the intruder alone) and active defense (corresponding to the upright position while trying to push the resident away, boxing, and fleeing from the resident). Data are expressed as mean ± SEM. *, a statistically significant difference between the marked group and both saline and NMDA-control groups for each behavioral category (P < 0.0002 in all statistically significant pairwise comparisons).

Fig. 4.

(A and C) Darkfield photomicrographs of transverse sections to show the periaqueductal gray terminal fields after a PHAL injection in the PMD region mostly responsive to cat exposure (ventrolateral part, A) and in the PMD region mostly responsive to a dominant conspecific (dorsomedial part, C). (B and D) Brightfield photomicrographs of transverse Fos-stained PAG sections from animals exposed to a predator (B) or to a dominant conspecific (D). Abbreviations: III, oculomotor nucleus; PAGdl, -dm, and -l, periaqueductal gray, dorsolateral, dorsomedial, and lateral divisions; SOM, supraoculomotor region. (Scale bars, 200 μm.)

Fig. 5.

A schematic diagram showing the putative brain systems involved in processing predatory and conspecific threats and in organizing predator and conspecific defense. Abbreviations: AHN, anterior hypothalamic nucleus; BMAp, basomedial amygdalar nucleus, posterior part; LA, lateral amygdalar nucleus; MEAad, -pd, and -pv, medial amygdalar nucleus, anterodorsal, posterodorsal, and posteroventral parts; MPO, medial preoptic area; PAGdl, -dm, and -l, periaqueductal gray, dorsolateral, dorsomedial, and lateral parts; PMDdm and -vl, dorsal premammillary nucleus, dorsomedial and ventrolateral parts; PMV, ventral premammillary nucleus; VMHdm and -vl, ventromedial nucleus, dorsomedial and ventrolateral parts.