Compared with other vertebrates, the brain of adult teleost fish exhibits two unique features: it exhibits unusually high neurogenic activity and strongly expresses aromatase, a key enzyme that converts aromatizable androgens into estrogens. Until now, these two features, high neurogenic and aromatase activities, have never been related to each other. Recently, it was shown that aromatase is expressed in radial glial cells of the forebrain and not in neurons. Here, we further document that Aromatase B is never detected in cells expressing the markers of postmitotic neurons, Hu and acetylated tubulin. By using a combination of bromodeoxyuridine (BrdU) treatment and immunohistochemical techniques, we demonstrate for the first time to our knowledge that aromatase-positive radial cells actively divide to generate newborn cells in many forebrain regions. Such newborn cells can further divide, as shown by BrdU-proliferating cell nuclear antigen double staining. We also demonstrate that, over time, newborn cells move away from the ventricles, most likely by migrating along the radial processes. Finally, by using antisera to Hu and acetylated tubulin, we further document that some of the newborn cells derived from radial glia differentiate into neurons. These data provide new evidence for the mechanism of neurogenesis in the brain of adult fish. In addition, given that estrogens are well-known neurotrophic and neuroprotective factors affecting proliferation, apoptosis, migration, and differentiation, the expression of aromatase in the neural stem cells of the adult strongly demonstrates that the fish brain is an outstanding model for studying the effects of estrogens on adult neurogenesis and brain repair.

Heightened concern over endocrine-disrupting chemicals is driven by the hypothesis that they could reduce reproductive success and affect wildlife populations, but there is little evidence for this expectation. The pharmaceutical ethynylestradiol (EE2) is a potent endocrine modulator and is present in the aquatic environment at biologically active concentrations. To investigate impacts on reproductive success and mechanisms of disruption, we exposed breeding populations (n = 12) of zebrafish (Danio rerio) over multiple generations to environmentally relevant concentrations of EE2. Life-long exposure to 5 ng/L EE2 in the F1 generation caused a 56% reduction in fecundity and complete population failure with no fertilization. Conversely, the same level of exposure for up to 40 days in mature adults in the parental F0 generation had no impact on reproductive success. Infertility in the F1 generation after life-long exposure to 5 ng/L EE2 was due to disturbed sexual differentiation, with males having no functional testes and either undifferentiated or inter-sex gonads. These F1 males also showed a reduced vitellogenic response when compared with F0 males, indicating an acclimation to EE2 exposure. Depuration studies found only a partial recovery in reproductive capacity after 5 months. Significantly, even though the F1 males lacked functional testes, they showed male-pattern reproductive behavior, inducing the spawning act and competing with healthy males to disrupt fertilization. Endocrine disruption is therefore likely to affect breeding dynamics and reproductive success in group-spawning fish. Our findings raise major concerns about the population-level impacts for wildlife of long-term exposure to low concentrations of estrogenic endocrine disruptors.

Unlike that of mammals, the brain of teleost fish exhibits an intense aromatase activity due to the strong expression of one of two aromatase genes (aromatase A or cyp19a1a and aromatase B or cyp19a1b) that arose from a gene duplication event. In situ hybridization, immunohistochemistry and expression of GFP (green fluorescent protein) in transgenic tg(cyp19a1b-GFP) fish demonstrate that aromatase B is only expressed in radial glial cells (RGC) of adult fish. These cells persist throughout life and act as progenitors in the brain of both developing and adult fish. Although aromatase B-positive radial glial cells are most abundant in the preoptic area and the hypothalamus, they are observed throughout the entire central nervous system and spinal cord. In agreement with the fact that brain aromatase activity is correlated to sex steroid levels, the high expression of cyp19a1b is due to an auto-regulatory loop through which estrogens and aromatizable androgens up-regulate aromatase expression. This mechanism involves estrogen receptor binding on an estrogen response element located on the cyp19a1b promoter. Cell specificity is achieved by a mandatory cooperation between estrogen receptors and unidentified glial factors. Given the emerging roles of estrogens in neurogenesis, the unique feature of the adult fish brain suggests that, in addition to classical functions on brain sexual differentiation and sexual behaviour, aromatase expression in radial glial cells could be part of the mechanisms authorizing the maintenance of a high proliferative activity in the brain of fish.

Compared with adult mammals, the brain of teleost fish is characterized by an extremely high capacity to aromatize androgens into estrogens, and this metabolic activity results from the expression of a specific brain aromatase (AroB) generated by the cyp19b gene. In this study, we first generated antibodies to zebrafish AroB and used them to map AroB-positive structures in the brain of adult zebrafish. We show that AroB is exclusively expressed in radial glial cells, mainly in the olfactory bulbs, telencephalon, preoptic area, and hypothalamus. Second, we investigated in vivo and in vitro the mechanisms involved in the estradiol (E2) regulation of the cyp19b gene. By means of whole-mount hybridization and immunohistochemistry on zebrafish embryos and larvae, we confirmed the E2-dependent upregulation of the cyp19b gene, and we show that E2 triggers AroB expression in radial glial cells mainly in the preoptic area and mediobasal hypothalamus of 48 hpf (hours post fertilization) and 108 hpf larvae. In addition, an in vitro analysis of 0.5 kb of the promoter region of the cyp19b gene demonstrated that this E2-dependent regulation involves a direct transcriptional action of estrogen receptors requiring estrogen-responsive elements. However, the data obtained on different cell lines demonstrate that a glial cell context is necessary for full E2 induction. The correlation between our in vivo and in vitro data suggests that the E2-dependent upregulation of AroB is favored by a glial cell context.

No abstract