early work in this field was based on the premise that the gonads of young
male birds secrete T, which is then aromatized to E 2 within neurons of song
nuclei to steer their developmental program in the masculine direction. The
demonstration that the bird forebrain has very high levels of aromatase
activity, sufficient to influence circulating estrogen levels elsewhere in the
body, supported this idea (Shen et al. 1992), as did the observation that
estrogen receptors are found in HVc (Walters et al. 1988). However, to date,
no treatment designed to decrease levels of endogenous estrogen (including antiestrogens and castration) has been effective in demasculinizing the
song systems of males (Jacobs et al. 1995; Wade and Arnold 1996; Arnold
1997). At stages during which exogenous estrogen masculinizes the song
system, there are no sex differences in levels of circulating androgens (the
substrate for aromatase), circulating estrogens (mostly derived from brain
aromatization), or levels of aromatase itself. The song nuclei themselves do
not express the aromatase gene (Metzdorf et al. 1999); of the song nuclei,
only HVc expresses the estrogen receptor, and the expression levels are the
same in both sexes (Johnson and Bottjer 1995). If developing zebra finches
are treated in ovo with the aromatase inhibitor fadrozole, genetic females
develop testicular tissue that secretes androgens, but song nuclei are not
masculinized (Wade and Arnold 1996; Wade et al. 1996). Taken together,
the results suggest that sex differences in steroid hormone secretion are not
sufficent to explain masculinization of the song system. Although there is a
great deal of evidence that steroid hormones play some role, the native
mechanisms of sexual differentiation of the vocal system are entirely
obscure (Bottjer and Arnold 1997; Kelley 1997).
Despite these caveats, the observation that exogenous steroids can masculinize song nuclei has been affirmed repeatedly and used as a basis for
understanding how the number of cells, their size, and their connectivity
become sexually differentiated. For example, steroids interact with neurotrophic factors in the differentiation of song nuclei. These interactions can
be demonstrated when the trophic connections between song nuclei are disrupted. In young zebra finches, lesions of MAN lead to the death of neurons
in RA; treatment of RA with the neurotrophin BDNF suppresses neuronal
death (Johnson et al. 1997). BDNF expression is higher in developing males
than in females (Akutagawa and Konishi 1998), and estradiol increases
BDNF expression (Dittrich et al. 1999).
4.2. Frogs
In the African clawed frog, Xenopus laevis, males and females produce sexually distinct vocalizations. The calls of both sexes are made up of clicks
that are repeated in distinctive, sex-specific temporal patterns (Kelley and
Tobias 1999). Reflecting the vocal differences, the vocal organ shows dramatic differences between males and females. The adult male larynx has
more muscle and cartilage cells than the female, and the cells also differ
302
A. Yamaguchi and D.B. Kelley
male birds secrete T, which is then aromatized to E 2 within neurons of song
nuclei to steer their developmental program in the masculine direction. The
demonstration that the bird forebrain has very high levels of aromatase
activity, sufficient to influence circulating estrogen levels elsewhere in the
body, supported this idea (Shen et al. 1992), as did the observation that
estrogen receptors are found in HVc (Walters et al. 1988). However, to date,
no treatment designed to decrease levels of endogenous estrogen (including antiestrogens and castration) has been effective in demasculinizing the
song systems of males (Jacobs et al. 1995; Wade and Arnold 1996; Arnold
1997). At stages during which exogenous estrogen masculinizes the song
system, there are no sex differences in levels of circulating androgens (the
substrate for aromatase), circulating estrogens (mostly derived from brain
aromatization), or levels of aromatase itself. The song nuclei themselves do
not express the aromatase gene (Metzdorf et al. 1999); of the song nuclei,
only HVc expresses the estrogen receptor, and the expression levels are the
same in both sexes (Johnson and Bottjer 1995). If developing zebra finches
are treated in ovo with the aromatase inhibitor fadrozole, genetic females
develop testicular tissue that secretes androgens, but song nuclei are not
masculinized (Wade and Arnold 1996; Wade et al. 1996). Taken together,
the results suggest that sex differences in steroid hormone secretion are not
sufficent to explain masculinization of the song system. Although there is a
great deal of evidence that steroid hormones play some role, the native
mechanisms of sexual differentiation of the vocal system are entirely
obscure (Bottjer and Arnold 1997; Kelley 1997).
Despite these caveats, the observation that exogenous steroids can masculinize song nuclei has been affirmed repeatedly and used as a basis for
understanding how the number of cells, their size, and their connectivity
become sexually differentiated. For example, steroids interact with neurotrophic factors in the differentiation of song nuclei. These interactions can
be demonstrated when the trophic connections between song nuclei are disrupted. In young zebra finches, lesions of MAN lead to the death of neurons
in RA; treatment of RA with the neurotrophin BDNF suppresses neuronal
death (Johnson et al. 1997). BDNF expression is higher in developing males
than in females (Akutagawa and Konishi 1998), and estradiol increases
BDNF expression (Dittrich et al. 1999).
4.2. Frogs
In the African clawed frog, Xenopus laevis, males and females produce sexually distinct vocalizations. The calls of both sexes are made up of clicks
that are repeated in distinctive, sex-specific temporal patterns (Kelley and
Tobias 1999). Reflecting the vocal differences, the vocal organ shows dramatic differences between males and females. The adult male larynx has
more muscle and cartilage cells than the female, and the cells also differ
302
A. Yamaguchi and D.B. Kelley