Parallel or serial processes in sexual differentiation?

We argue that estrogen feminization of the brain is the result of a series of events initiated by differential androgen exposure. There is no need to postulate a feminizing process parallel to androgen-induced masculinization to explain the findings. Fitch & Denenberg (F&D) argue that the current view of the process of sexual differentiation of mammalian brain and behavior proposed by Phoenix et al. (1959) needs to be amended. In their view, two parallel periods of hormone action are required for complete sexual differentiation: a perinatal period when androgens masculinize and defeminize the male brain and a later prepubertal period when estrogens act to feminize the female brain. They present evidence from their own work on the sexual differentiation of the rat corpus callosum to support their view that estrogenic feminization, which occurs during postnatal days 12– 35 for the corpus callosum, is a much neglected process that is required for complete neural sex differentiation in mammals. Although this is an interesting idea, particularly in light of the fact that estrogenic feminization is the mechanism underlying sexual differentiation of birds and some reptilian species (Adkins-Regan 1988), we believe that there is another, broader context in which to place these data. Our alternate view is that the effects of estrogens on the female brain during this “second period” of hormone action are a result of the differentiation process, rather than a part of it. That is, because males are exposed to testicular androgens during the perinatal period, they become unresponsive (or differentially responsive) to estrogens for the remainder of their lives. Females, in contrast, do not get exposed to much androgen or estrogen during the perinatal critical period, and thus they remain responsive to estrogens. We would argue that this sexually dimorphic response to estrogen is apparent throughout the lifetime of the rat and is not just restricted to the prepubertal period. To make a case that two parallel processes are required for full sexual differentiation, F&D would have had to provide convincing evidence that there is a critical period for estrogen action as there is for androgen action, and demonstrate that endogenous or exogenously administered estrogen can feminize both females and males. The only evidence of a time-limited period for “feminization” of the corpus callosum is that an enlargement of the corpus callosum occurs following ovariectomy (OVX) on postnatal day 12 (PD 12) but not following OVX on PD 78 (Mack et al. 1996a). The enlargement following OVX on PD 12 requires over 43 days to be expressed, yet the corpora callosi of adult rats were examined only 32 days after OVX. It may be that if F&D had waited another 2–4 weeks, a change in corpus callosum size would be detectable. This is particularly important because the feminizing effect of estrogen takes so long to influence the size of the corpus callosum. Moreover, F&D have not tried to replace estradiol following the presumed critical period to determine whether there is indeed a time-limited window for the developmental effect. A second argument against F&D’s interpretation of androgenic and estrogenic influences as parallel processes lies in the lack of effect of estrogen-feminization manipulations on males. While the manipulation of perinatal androgens has similar consequences for both males and females over a wide range of tasks and measures (see sects. 3 and 5.4), only females appear to be affected by manipulations of estrogen during the prepubertal period. Also, the fact that sex differences in callosal size are apparent at 3 days of age (Zimmerberg & Scalzi 1989) indicates that ovarian secretions are not required for the expression of this sexual dimorphism. A more parsimonious way to account for the data is to remember that the responsivity of the female brain (e.g., corpus callosum) to estrogen during the prepubertal period may simply be the result of differentiation that occurs perinatally. This is certainly not a novel idea. Phoenix et al. (1959) proposed that “androgen administered prenatally has an organizing action on the tissues mediating mating behavior in the sense of producing a responsiveness to exogenous hormones which differs from that of normal adult females” (p. 369). For example, it is well known that male rats and female rats that have been exposed to androgens perinatally do not show a positive feedback effect of estrogen (Karsch et al. 1973). We have recently discovered that acute estradiol administration increases spine density of hippocampal dentate gyrus granule cells in aged female rats, but not in male rats (Henderson et al. 1997). As in the case of the corpus callosum, a structural change is induced in female rats by estradiol, but male rats are unresponsive. In F&D’s terminology, these findings indicate that females are “feminized” by estrogen (while males are not) at numerous times during their life span. Though one might argue that the effects reported above are “activational” while the developmental effects of estrogen reported by F&D may be “organizational” (though as indicated above, the data are not yet convincing), it is clear that the distinctions between these presumably different types of hormone action are blurring. Activational effects can be seen early in development (Williams 1986); organizational effects can be seen in adulthood (Arnold & Breedlove 1985). In addition, it is now well accepted that activational and organizational effects of hormones can alter specific neuronal structures and functions. Furthermore, the work of Toran-Allerand (1995) suggests that estrogens may work by interacting with neurotrophins and their receptors or with other locally synthesized growth factors. This might represent a universal mechanism underlying the multiple and varied actions of estrogens at different stages of life. Response/Fitch & Denenberg: Sexual differentiation of the brain BEHAVIORAL AND BRAIN SCIENCES (1998) 21:3 341 We feel that the crucial event for sexual differentiation is the presence or absence of androgens during the perinatal period. As a consequence of this experience, a cascade of other events occurs that further polarizes the behavior and morphological differences between males and females. One such event is probably the sexually dimorphic response to pubertal exposure to estrogens that normally occurs only in females. Another example might be the finding that androgen exposure perinatally leads to an increase in maternal licking, which has been found to further masculinize the male rat pup (Moore 1992). The responsivity of the corpus callosum to estrogen during late development is an important and intriguing finding that can be explained without the need to amend the “organizational theory” with a second parallel process.