These results localize the anatomical substrate of offense in the hypothalamus more precisely than previous studies that have used electrical stimulation. Whereas the present results with chemical stimulation are localized only in the anterior hypothalamus, previous results with electrical stimulation have produced offense behavior at a series of levels ranging from the anterior hypothalamus to the level of the ventromedial nucleus (11, 15, 21, 26).
Because chemical stimulation affects only the cell bodies of neurons, while electrical stimulation affects the axons as well, one may propose that the results from studies with electrical stimulation may have come, in large part, from stimulation of descending axons originating in the anterior hypothalamus. Such axons are known to course through the lateral hypothalamus in the regions where electrical stimulation has produced offense (6).
Presumably the present results mean that neurons in the anterior hypothalamus controlling offense are under tonic inhibition mediated by the neurotransmitter gamma-aminobutyric acid (GABA), and picrotoxin produces offense by blocking the action of GABA.
While the present results suggest that neurons mediating offense are under tonic GABA inhibition, they leave open the question as to what are the excitatory neurotransmitters that activate these neurons, as well as what transmitter is released by their axons. Further research is needed to address these questions, as well as the source of the presumed GABA inhibition. At the moment we are assessing lesions of the anterior hypothalamus to see if they abolish offense that is elicited naturally.
If we are correct that our procedure removed a tonic inhibition rather than affecting an excitatory neurotransmitter, this may explain why bite-and-kick attack occurred in only 8 of the 47 injection sites of the anterior hypothalamus. One may suppose that despite removal of inhibition, there was not sufficient background stimulation of the neurons in the inactive sites to elicit offense. This is supported by the finding that on-top posture, which may reflect lower intensity offense that fails to be consummated with attack, occurred in eight other sites in this region.
We are not aware of any other studies which have obtained offense by chemical stimulation, although defensive attack has been obtained by cholinergic chemical stimulation of the brain (2).
Our failure to obtain offense by picrotoxin injections in the ventromedial tegmentum calls into question the speculation that it might be the site of the offense motivational mechanism (3). And it corroborates unpublished findings in our laboratory that electrical stimulation of this region does not produce offense, but only circling locomotion instead. The role of the ventromedial tegmentum in offense remains unknown.
The localization of mounting behavior in this study is similar to that of studies that have used electrical stimulation (18, 19, 24, 25). There are fewer comparable studies, however, using chemical stimulation to produce mounting behavior. Fisher (10) reported stimulus-bound mounting by microinjections of testosterone in the lateral preoptic area in 6 of 130 animals tested. Other studies have used bicuculline and picrotoxin injections in the medial preoptic area to speed up ejaculation when males were tested with receptive females (8, 9). Bicuculline, like picrotoxin, antagonizes the action of GABA. The present study differs from others, however, in that we obtained mounting by females as well as by males and mounting of same sex as well as opposite sex opponents.
We are at a loss to explain the circumscribed anatomical localization and time course of mounting by the opponent in response to picrotoxin injections. At first we thought it might be due to pheromone secretion by the test animal which was communicated to the opponent by self-grooming (17). However, in assessing the data, we found no correlation between the amount of self-grooming by the test animal and the occurrence of mounting by the opponent.
Locomotion has been obtained by picrotoxin injections in the basal forebrain (20, 23), and the present paper extends those effects further posterior. As shown in Fig. 2, persistent locomotion and circling was obtained throughout the hypothalamus.
Other behaviors elicited by the picrotoxin, including social and self-grooming, digging, and leaping, have been reported in studies using electrical stimulation of the hypothalamus. For example Lammers et al. (14-16) claimed that self-grooming was localized to the medial hypothalamus, while social grooming, locomotion, and digging were localized laterally. We did not find such specificity of localization in our results from picrotoxin injections.