The alpha-fetoprotein knock-out mouse model suggests that parental behavior is sexually differentiated under the influence of prenatal estradiol

Abstract

In rodent species, sexual differentiation of the brain for many reproductive processes depends largely on estradiol. This was recently confirmed again by using the alpha-fetoprotein knockout (AFP-KO) mouse model, which lacks the protective actions of alpha-fetoprotein against maternal estradiol and as a result represents a good model to determine the contribution of prenatal estradiol to the sexual differentiation of the brain and behavior. Female AFP-KO mice were defeminized and masculinized with regard to their neuroendocrine responses as well as sexual behavior. Since parental behavior is also strongly sexually differentiated in mice, we used the AFP-KO mouse model here to ask whether parental responses are differentiated prenatally under the influence of estradiol. It was found that AFP-KO females showed longer latencies to retrieve pups to the nest and also exhibited lower levels of crouching over the pups in the nest in comparison to WT females. In fact, they resembled males (WT and AFP-KO). Other measures of maternal behavior, for example the incidence of infanticide, tended to be higher in AFP-KO females than in WT females but this increase failed to reach statistical significance. The deficits observed in parental behavior of AFP-KO females could not be explained by any changes in olfactory function, novelty recognition or anxiety. Thus our results suggest that prenatal estradiol defeminizes the parental brain in mice.

Figure 1

Relative time spent by male (M) and female (F) mice of both WT and AFP-KO genotypes in activities directed (time spent crouching over pups and licking) or not (time spent out of the nest or self-grooming) towards pups. *p<0.05 compared to F-KO, M-WT, and M-KO.

Figure 2

Proportion of animals attacking pups in each group during the 30 min observation test.

Figure 3

Mean (± SEM) time spent by AFP-KO and WT mice of both sexes investigating deionized water or volatile urinary stimuli. * p < 0.05 post-hoc comparisons between the time spent investigating the third presentation of a particular stimulus and the first presentation of the next stimulus. # p < 0.05 post-hoc comparisons between sexes.

Figure 4

Mean (± SEM) total time spent in the open and closed arms in the elevated plus maze test. ANOVA with sex and genotype as independent factors and the time spent investigating the arms of the maze as repeated factors indicated a general effect of time spent investigating each arm (open/closed) but no interactions between factors.

Figure 5

Novel object recognition test. Time spent investigating the novel object is expressed as a preference index (time investigating the novel object/total time of investigation).