Abstract

Two strains of male mice have bred over fortygenerations, starting with the work of RobertCairns and his colleagues, one strain with ahigh level of intra-species aggression, theother a low level of aggression. Thehigh-aggression mice tend to establishdominance hierarchies and particularly fight inthe presence of female mice. Thelow-aggression mice tend, in groups of theirown, to have a high degree of low-intensity,peaceful social contact, and to be more timidin initiating action than the high-aggressionmice. Biochemical differences have beenobserved between the two strains, and confirmedby the present data: the high-aggression micehave greater dopamine concentrations (in thecaudate nucleus and nucleus accumbens), lowerlevels of the stress hormone corticosterone,and higher levels of testosterone than thelow-aggression mice. The current experimentswere designed to answer questions about theflexibility of adaptive behaviors:specifically, what is the effect of early dailymaternal separation on adult stress response ineach strain? What are the behavioral andhormonal mechanisms by which, as has beenobserved, low-aggression mice achieve adominant status when brought into situationswhere they compete for territory withhigh-aggression mice? Finally, what are thesocial and neurochemical mechanisms by whichhigh-aggression mice can develop low-aggressionbehavior if brought out of isolation and intogroups?Maternal separation was found to lead todecreases in stress levels, as measured bycorticosterone, in the low-aggression but notthe high-aggression, mice – presumably becauseof the surplus of maternal care the pupsreceive on returning to the nest. When alow-aggression mouse became dominant and ahigh-aggression mouse became submissive, theirusual pattern of corticosterone andtestosterone levels was found to be reversed. The change to low-aggression behavior inhigh-aggression mice switching from anisolation condition to a group condition, wasmediated by a decrease in D 1 dopaminereceptor densities.These results, like the ones on which theybuild, argue for substantial developmentalinfluences in expressions of the genesinfluencing aggressive or cooperative behavior. In this approach to evolution, epigenesis istreated not as a set of traits and behaviorspredetermined by the genome, but as a set ofprobabilistic tendencies toward certain traitsand behaviors.