Steroid-induced plasticity in the sexually dimorphic vasotocinergic innervation of the avian brain: behavioral implications.

Vasotocin (VT, the antidiuretic hormone of birds) is synthesized by diencephalic magnocellular neurons projecting to the neurohypophysis. In addition, in male quail and in other oscine and non-oscine birds, a sexually dimorphic group of VT-immunoreactive (ir) parvocellular neurons is located in a region homologous to the mammalian nucleus of the stria terminalis, pars medialis (BSTm) and in the medial preoptic nucleus (POM). These cells are not visible in females. VT-ir fibers are present in many diencephalic and extradiencephalic locations. Quantitative morphometric analyses demonstrate that, in quail, these elements are expressed in a sexually dimorphic manner (males>females) in regions involved in the control of different aspects of reproduction: i.e., the POM (copulatory behavior), the lateral septum (secretion of gonadotropin-releasing hormone [GnRH]), the nucleus intercollicularis (control of vocalizations), and the locus coeruleus (the main noradrenergic center of the avian brain). In many of these regions, VT-ir fibers are closely related to aromatase-ir, GnRH-ir, or estrogen receptor-expressing neurons. This dimorphism has an organizational nature: administration of estradiol-benzoate to quail embryos (a treatment that abolishes male sexual behavior) results in a dramatic decrease of the VT-immunoreactivity in all sexually dimorphic regions of the male quail brain. Conversely, the inhibition of estradiol (E2) synthesis during embryonic life (a treatment that stimulates the expression of male copulatory behavior in adult testosterone (T)-treated females) results in a male-like distribution of VT-ir cells and fibers. Castration markedly decreases the immunoreactivity in both the VT-immunopositive elements of the BSTm and the innervation of the SL and POM, whereas T-replacement therapy restores the VT immunoreactivity to a level typical of intact birds. These changes reflect modifications of VT mRNA concentrations (and probably synthesis) as demonstrated by in situ hybridization and they are paralleled by similar changes in male copulatory behavior (absent in castrated male quail, fully expressed in CX+T males). The aromatization of T into estradiol (E2) also controls VT expression and, in parallel limits the activation of male sexual behavior by T. In castrated male quail, the restoration by T of the VT immunoreactivity in POM, BSTm and lateral septum could be fully mimicked by a treatment with E2, but the androgen 5alpha-dihydrotestosterone (DHT) had absolutely no effect on the VT immunoreactivity in these conditions. At the doses used in this study, DHT also did not synergize with E2 to enhance the density of VT immunoreactive structures. Systemic or i.c.v. injections of VT markedly inhibit the expression of all aspects of male sexual behavior. VT, presumably, does not simply represent one step in the biochemical cascade of events that is induced by T in the brain and leads to the expression of male sexual behavior. Androgens and estrogens presumably affect reproductive behavior both directly, by acting on steroid-sensitive neurons in the preoptic area, and indirectly, by modulating peptidergic (specifically vasotocinergic) inputs to this and other areas. The respective contribution of these two types of actions and their interaction deserves further analysis.