Two clearly opposing views exist on the function of alpha-fetoprotein (AFP), a fetal plasma protein that binds estrogens with high affinity, in the sexual differentiation of the rodent brain. AFP has been proposed to either prevent the entry of estrogens or to actively transport estrogens into the developing female brain. The availability of Afp mutant mice (Afp(-/-)) now finally allows us to resolve this longstanding controversy concerning the role of AFP in brain sexual differentiation, and thus to determine whether prenatal estrogens contribute to the development of the female brain. Here we show that the brain and behavior of female Afp(-/-) mice were masculinized and defeminized. However, when estrogen production was blocked by embryonic treatment with the aromatase inhibitor 1,4,6-androstatriene-3,17-dione, the feminine phenotype of these mice was rescued. These results clearly demonstrate that prenatal estrogens masculinize and defeminize the brain and that AFP protects the female brain from these effects of estrogens.

The alpha-fetoprotein gene (Afp) is a member of a multigenic family that comprises the related genes encoding albumin, alphaalbumin, and vitamin D binding protein. The biological role of this major embryonic serum protein is unknown although numerous speculations have been made. We have used gene targeting to show that AFP is not required for embryonic development. AFP null embryos develop normally, and individually transplanted homozygous embryos can develop in an AFP-deficient microenvironment. Whereas mutant homozygous adult males are viable and fertile, AFP null females are infertile. Our analyses of these mice indicate that the defect is caused by a dysfunction of the hypothalamic͞ pituitary system, leading to anovulation.A lpha-fetoprotein (AFP) is a serum glycoprotein produced at high levels during fetal life by the liver and the visceral endoderm of the yolk sac and at lower levels by the developing gastrointestinal tract (1, 2). The protein expressed by the embryo is transferred to the maternal blood circulation, and abnormal levels of embryonic AFP in the maternal serum are indicative of spina bifida or Down's syndrome in the fetus (3, 4). The synthesis of AFP decreases dramatically after birth, and only trace amounts are detected in the adult (2). AFP expression has been shown to be regulated by transcriptional mechanisms involving a relatively large promoter (P1) and three distant enhancers (for reviews see refs. 5 and 6). More recently it has been shown that the first intron of the Afp gene contains an enhancer and an alternative promoter called P2 (7). The genes of the albumin family are linked in the mammalian genomes, and this conserved organization has been proposed to be important for the developmental expression switch of the genes of the family after birth (8). Several hypotheses have been proposed for the physiological function of AFP (for reviews see refs. 6, 9, and 10). Because AFP is synthesized during the cell cycle G 1 and S phases, it has been hypothesized that it affects cell growth (11, 12). The observation that AFP is able to bind estrogen led to the suggestion that AFP plays a role in sexual differentiation of the brain by protecting the fetus from the effects of circulating estrogen (13). In addition to binding estrogen, AFP, like albumin, is able to bind other steroids as well as endogenous and exogenous substances such as fatty acids, bilirubin, and various pharmaceutical agents, suggesting that AFP may play a general transportation role (for review see ref. 14). Moreover, because cellular internalization of the protein has been reported, AFP could also interact with cytoplasmic chaperone proteins that normally escort nuclear receptors or transcription cofactors through the cytoplasm toward organelle interfaces (15, 16). AFP has also been proposed to be one protein that protects the embryo against the maternal immune system, on the basis of the observation that addition of purified AFP into the culture of splenic or lymph node mononuclear cells exerts a suppressive effect on ant...

Sex steroid hormones act on developing neural circuits regulating the hypothalamic-pituitary-gonadal axis and are involved in hormone-sensitive behaviours. These hormones act mainly via nuclear receptors, such as oestrogen receptor (ER)-α and androgen receptor (AR). By using immunohistochemistry, we analysed the expression level of ERα and AR throughout perinatal life [at embryonic (E) day 19 and postnatal (P) days 5, 15 and 25] and in adulthood in several hypothalamic nuclei controlling reproduction in both wild-type and aromatase knockout (ArKO) (i.e. which cannot convert testosterone into oestradiol) mice to determine whether there are sex differences in hypothalamic ERα and AR expression and, if so, whether these are established by the action of oestradiol. As early as E19, ERα immunoreactivity (-IR) was observed at same expression levels in both sexes in the anteroventral periventricular nucleus (AVPv), the medial preoptic area (MPOA), the bed nucleus of the stria terminalis (BnST), the ventrolateral part of the ventromedial hypothalamic nucleus and the arcuate nucleus (ARC). Sex differences (female > male) in ERα-IR were observed not only during the prepubertal period in the BnST (P5 to P25) and the MPOA (P15), but also in adulthood in these two brain regions. Sex differences in AR-IR (male > female) were observed at P5 in the AVPv and ARC, and at P25 in the MPOA and ARC, as well as in adulthood in all hypothalamic regions analysed. In adulthood, gonadectomy and hormonal treatment (oestradiol or dihydrotestosterone) also strongly modulated ERα-IR and AR, respectively. Taken together, sex differences in ERα-IR and AR-IR were observed in all hypothalamic regions analysed, although they most likely do not reflect the action of oestradiol because ArKO mice of both sexes showed expression levels very similar to wild-type mice throughout perinatal development.

Sex differences in gonadal function are driven by either cyclical (females) or tonic (males) hypothalamic GnRH1 release and, subsequently, gonadotrophin (LH and FSH) secretion from the pituitary. This sex difference seems to depend on the perinatal actions of gonadal hormones on the hypothalamus. We used alpha-fetoprotein (AFP) knockout mice (Afp(-/-)) to study the mechanisms by which estrogens affect the sexual differentiation of the GnRH1 system. Afp(-/-) mice lack the protective actions of AFP against estrogens circulating during embryonic development, leading to infertility probably due to a hypothalamic dysfunction. Therefore, we first determined whether Afp(-/-) females are capable of showing a steroid-induced preovulatory LH surge by FOS/GnRH1 immunohistochemistry and RIA of plasma LH levels. Because the KISS1/GPR54 system is a key upstream regulator of the GnRH1 system as well as being sexually dimorphic, we also analyzed whether Kisspeptin-10 neurons were activated in Afp(-/-) mice after treatment with estradiol and progesterone. We found that the GnRH1 and Kisspeptin-10 neuronal systems are defeminized in Afp(-/-) females because they did not show either steroid-induced LH surges or significant FOS/GnRH1 double labeling. Furthermore, Kisspeptin-10 immunoreactivity and neural activation, measured by the number of double-labeled FOS/Kisspeptin-10 cells, were lower in Afp(-/-) females, suggesting a down-regulation of GnRH1 function. Thus, the sex difference in the ability to show preovulatory LH surges depends on the prenatal actions of estrogens in the male hypothalamus and, thus, is lost in Afp(-/-) females because they lack AFP to protect them against the defeminizing effects of estrogens during prenatal development.