Influence of testosterone metabolites on song-control system neuroplasticity during photostimulation in adult European starlings (Sturnus vulgaris)

Abstract

The song-control system is a network of discrete nuclei in the songbird brain that controls the production and learning of birdsong and exhibits some of the best-studied neuroplasticity found in the adult brain. Photoperiodic growth of the song-control system during the breeding season is driven, at least in part, by the gonadal steroid testosterone. When acting on neural tissue, however, testosterone can be metabolized into 5α-dihydrotestosterone (DHT) or 17β-estradiol (E2), which activate different hormonal signaling pathways. By treating adult starlings with both testosterone metabolites and metabolite antagonists, we attempted to isolate the effects of androgen and estrogen treatment on neuroplasticity during photostimulation in male and female European starlings (Sturnus vulgaris). Photostimulation resulted in a large HVC volume typical of the breeding season in all treatments independent of hormone treatment. E2 had additional effects on HVC growth by reducing neuron density and enhancing early survival of new neurons recruited to HVC in females but did not significantly affect HVC volume. Conversely, DHT reduced the migration of new neurons, assessed by the expression of doublecortin, to HVC. DHT also increased syrinx mass and maintained RA (robust nucleus of the arcopallium) cytoarchitecture in the presence of aromatase inhibitors. In addition, we document the first evidence of sex-specific neuroplastic responses of the song-control system to androgens and estrogens. These findings suggest that the contributions of DHT and E2 signaling in songbird neuroplasticity may be regulated by photoperiod and that future studies should account for species and sex differences in the brain.

Figure 1. Effects of testosterone metabolites on anatomy of HVC.

A) Photomicrograph of HVC in NeuN-stained (NeuN+) tissue. Ventral border indicated by black arrows. Scale bar = 500 µm. B) Volumes of HVC across treatments. C) NeuN+ cell count in HVC. D) Density of NeuN+ cells in HVC. All bars represent mean ± SEM. * <0.05.

Figure 2. Effects of testosterone metabolites on anatomy of RA.

A) Photomicrograph of RA in NeuN-stained (NeuN+) tissue. Nuclear borders indicated by black arrows. Scale bar = 500 µm. B) Volumes of RA across treatments. C) NeuN+ cell count in RA. D) Density of NeuN+ cells in RA. E) Average NeuN+ cell soma size in RA. All bars represent mean ± SEM. * <0.05. RA, robust nucleus of the arcopallium. In C), asterisks without connecting lines mark significant differences found using post-hoc analysis.

Figure 3. Effects of testosterone metabolites on anatomy of area X.

A) Photomicrograph of Area X in NeuN-stained tissue. Nuclear borders indicated by black arrows. Scale bar = 500 µm. B) Volumes of Area X across treatments. All bars represent mean ± SEM. * <0.05.

Figure 4. Effects of testosterone metabolites on doublecortin-stained (DCX+) cell recruitment to HVC.

A) Photomicrograph of DCX+ fusiform cell. Scale bar = 50 µm. B) Number of DCX+ fusiform cells sampled in HVC across treatments. C) Photomicrograph of DCX+ spherical cell. Scale bar = 50 µm. B) Number of DCX+ spherical cells sampled in HVC across treatments. In all figures, bars represent estimated marginal mean ± SEM using the same measurement in the adjacent nidopallium as a covariate. * <0.05. In B), asterisks without connecting lines mark significant differences found using post-hoc analysis.

Figure 5. Neuroplastic differences between castrated male starlings and DHT- E2- male starlings.

A) NeuN-stained cell count in. Bars represent mean ± SEM. B) DCX+ spherical cells sampled in HVC. Bars represent estimated marginal mean ± SEM using DCX+ spherical cells sampled in the adjacent nidopallium as a covariate. * <0.05. DCX+, doublecortin-stained.

Figure 6. HVC volumes in male free-living European starlings compared to male HVC volume in this study.

Bars represent mean ± SEM. Data for Belgian European starlings in nonbreeding (September) and breeding (April) conditions from . Data for photostimulated birds from mean male HVC volume in this study regardless of hormonal manipulation.