John P. Broida

Infanticide: Accounting for genetic variation in mice

Infanticide, the killing of young, is one of a number of sexually-dimorphic traits in mice that is dependent upon androgen stimulation during perinatal life and during adulthood. Genotype also influences infanticide in that males of some strains of mice (C57BL/6J) exhibit high levels of this behavior while males of other strains (DBA/2J) seldom kill young. The experiments conducted here show that strain differences in pup killing behavior exhibited by males are not related to postweaning social factors nor are they due to differences in perinatal, pubertal, or adult levels of circulating hormones. These results, in combination with those previously reported, suggest that strain differences in the tendency of mice to kill young may instead depend upon the interaction of genotypic features such as prenatal hormone titers and/or sensitivity to these hormones, as well as on extra organismic factors such as intrauterine position. A model for understanding the manner in which genes and hormones may interact to influence infanticide and other hormone dependent sexually-dimorphic behaviors in mice is presented.

Strain-typical patterns of pregnancy-induced nestbuilding in mice: Maternal and experiential influences

Pregnant C57BL/6J mice incorporate less material into maternal nests and build fewer fully enclosed nests than do pregnant DBA/2J mice. These strain differences are not ameliorated by additional reproductive experience since multiparous animals also exhibit a similar pattern. Reciprocally-crossed hybrid females exhibit DBA-like levels of pregnancy-induced nestbuilding and cross-fostered C57BL and DBA females retain the phenotype of their strain. Experiential and maternal environmental factors apparently are not responsible for strain differences in pregnancy-induced nestbuilding. Differences in ovarian function and/or central neural tissue sensitivity to ovarian hormones may modulate strain differences in pregnancy-induced nestbuilding.

Genotypic influences on infanticide in mice: Environmental, situational and experiential determinants☆

Approximately 20-30% of adult, 60-70 day old, dBA/2J male mice exhibit infanticide toward standard stimulus 1-3 day old Rockland-Swiss Albino mouse pups, while 50-80% of adult C57BL/6J males display the behavior. Reciprocally-crossed hybrid males exhibit DBA-like levels of low pup-killing behavior and cross-fostered DBA and C57BL males retain their strain-typical phenotype. Also DBA males continue to exhibit low levels of infanticide and C57bL males display high levels of the behavior regardless of the length of pup exposure, or the sex, age, or strain of the stimulus newborn. Thus, strain differences in pup-killing behavior are not related to differences in the prenatal or postnatal maternal environment, or to situational or experiential factors. Inherent biological differences may be responsible for the strain differences in infanticidal behavior.

Infanticide exhibited by female mice: Genetic, developmental and hormonal influences

Approximately 25-40% of 25-45 day old C57BL/6J females killed young (1-3 day old Rockland-Swiss (R-S) albino mouse pups) while similarly aged DBA/2J females were parental or ignored neonates. Beyond 45 days of age C57BL and DBA females seldom killed young. When ovariectomized at weaning and tested for infanticide at 65 days of age, DBA females rarely killed neonates while 40% of C57BL females exhibited the behavior. In contrast to DBA females, significantly more C57BL females killed young in response to the adult administration of testosterone propionate (TP) and estradiol benzoate (EB), but not dihydrotestosterone propionate (DHTP). It is tentatively proposed that strain differences in spontaneous and steroid aroused infanticide in female mice may be related to differences in the prenatal hormone environment.

Mice: progesterone and the regulation of strain differences in pregnancy-induced nest building.

Pregnant DBA/2J females built significantly larger and more completely enclosed nests than did pregnant C57BL/6J mice. This strain difference was restricted to the last half of gestation and was not observed during either the virgin state or lactation. Genotype-based differences in pregnancy-induced nest building were not related to circulating levels of progesterone (P), core temperature, or body weight. Exposure to supplemented P during pregnancy elevated nest building exhibited by pregnant C57BL females but did not induce DBA-like levels of the behavior. Also, virgin DBA females built larger nests in response to P than did C57BL females. These findings suggest that differences in the sensitivity of central neural tissue to steroid hormones may account for genotypically determined variation in patterns of pregnancy-induced nest building.

Postpartum aggression in C57BL/6J and DBA/2J mice: Experiential and environmental influences

Parturient DBA/2J mice exhibit high levels of maternal aggression when confronted by an adult male Rockland-Swiss (R-S) intruder, while postpartum C57BL/6J females are passive. This strain difference is not altered by additional reproductive experience since multiparous animals continue to exhibit patterns of aggressive behavior consistent with their genotype. Reciprocally crossed hybrid females exhibit DBA-like patterns of aggressive behavior and cross-fostered females behave true to their genotype. Thus, experiential and environmental factors are not responsible for strain differences in postpartum aggression. Genotype determined differences in sensitivity to suckling-induced biochemical changes may modulate individual variation in the display of maternal aggression.

Intermale aggression and infanticide in aged C57BL/6J male mice: Behavioral deficits are not related to serum testosterone (T) levels and are not recovered by supplemental T

Healthy aged adult (24-26 months of age) and young adult (2-4 months of age) c57BL/6J male mice were assessed for intermale aggression, pup-killing behavior (infanticide), and circulating levels of testosterone (T). When compared to young adult male mice, aged adult males were highly variable in the exhibition of both androgen-dependent behaviors. Significant numbers of aged males exhibited deficits in aggression and pup-killing while other animals were as behaviorally active as their young male counterparts. Assessment of serum T showed that aging did not produce a reduction in levels of the steroid and individual variability in androgen-dependent behavior of aged males was not related to plasma levels of the hormone. When aged non-aggressive and non-killer males were exposed to supplemental T by way of subcutaneously implanted silastic capsules, circulating levels of the steroid were elevated but T-dependent behavior was not recovered. These findings, in combination with those previously reported for copulatory behavior, indicate that the deficits observed in the androgen-dependent behavior of aged male mice cannot be attributed to a breakdown in the production of testicular androgens. While neural refractoriness to T may account in part for deficits in androgen-dependent behavior of aged males, the variability that is observed in the reproductive behaviors of aged male rodents ultimately may be related to other sources of variation such as the perinatal environment.

Acute endocrine correlates of attack by lactating females in male mice: Effects on plasma prolactin, luteinizing hormone and corticosterone levels

Immediately following defeat inflicted by lactating Rockland-Swiss (R-S) albino mice, adult R-S male mice exhibited significant reductions in circulating prolactin (PRL) and luteinizing hormone (LH), but not corticosterone (CORT). These results suggest that acute neuroendocrine responses to intersex competition may be as dramatic as those previously reported for intermale encounters.