Steroid hormones act transsynaptically within the forebrain to regulate neuronal phenotype and song stereotypy

Abstract

Steroid sex hormones induce dramatic seasonal changes in reproductive related behaviors and their underlying neural substrates in seasonally breeding vertebrates. For example, in adult white-crowned sparrows, increased Spring photoperiod raises circulating testosterone, causing morphological and electrophysiological changes in song-control nuclei, which modify song behavior for the breeding season. We investigated how photoperiod and steroid hormones induce these changes in morphology, electrophysiology, and behavior. Neurons in a song premotor nucleus, the robust nucleus of the arcopallium (RA), show increased intrinsic spontaneous firing rate and soma size when birds are in breeding condition. Using combinations of systemic and unilateral local intracerebral hormonal manipulations, we show that long-day photoperiod accelerates the effects of systemic testosterone on RA neurons via the estradiol-synthesizing enzyme aromatase (CYP19A1); these changes require inputs from the afferent song control nucleus HVC (used as a proper name) and steroid receptor activation within HVC; local coactivation of androgen and estrogen receptors (ARs and ERs, respectively) within HVC, but not RA, is sufficient to cause neuronal changes in RA; activation of ARs in RA is also permissive. Using bilateral local intracerebral hormone-receptor blockade, we found that ARs and ERs in the song-control nucleus HVC mediate systemic testosterone-induced changes in song stereotypy but not rate. This novel transsynaptic effect of gonadal steroids on activity and morphology of RA neurons is part of a concerted change in key premotor nuclei, enabling stereotyped song.

Figure 1.

A, Simplified schematic of the avian song control system showing the distribution of steroid receptors. HVC and RA are in the main descending motor circuit. RA projects to the brainstem motoneurons that control the muscles involved with respiration and the sound production organ, the syrinx. HVC, X, DLM, and LMAN comprise the anterior forebrain pathway (AFP), which is necessary for song learning. DLM, Medial portion of the dorsolateral nucleus of the anterior thalamus; LMAN, lateral magnocellular nucleus of anterior nidopallium; RA, robust nucleus of the arcopallium; X, Area X, a subdivision of the medial striatum. B, Top, Representative extracellular recordings. An isolated unit from an LD photoperiod and systemic T-treated bird is shown on the left. An isolated unit from an SD photoperiod-treated bird is shown on the right. Bottom, Representative Nissl-stained RA cells from an LD photoperiod and systemic T-treated animal (on left) and a SD photoperiod-treated animal (on right). Scale bar, 5 μm. C, Representative HVC lesion. Arrows indicate the border of remaining HVC. The plane is parasagittal, and anterior is to the right. Scale bar, 500 μm. D, Representative cannula track near HVC. Arrows denote the border of HVC. The plane is coronal, and lateral is to the right. Scale bar, 500 μm.

Figure 2.

Photoperiod modulates the time course of a T-induced increase in firing rate and neuronal growth in RA. A1–A3, Response to systemic T under LD photoperiod. B1–B3, Response to systemic T under SD photoperiod. Different letters indicate significant differences between treatment groups. A1, B1, Spontaneous firing rate (mean ± SEM). A2, B2, Cumulative frequency distributions of firing rates (see Results for statistical significance). A3, B3, Soma area (mean ± SEM).

Figure 3.

Faster increase in spontaneous firing rate and growth of RA neurons in LD than SD photoperiod is attributable to increased availability of E₂. Different treatment groups were exposed to either an SD or LD photoperiod and received systemic implants of T, E₂, an inhibitor of estrogen synthesis (ATD), a combination of implants, or no steroid implant. The letters above the bars indicate significant differences between treatment groups. Numbers at base indicate units and animals. A, Average spontaneous firing rates of each treatment group (mean ± SEM). B, Cumulative frequency distributions (see Results for statistical significance). C, Soma area (mean ± SEM). The SE of the SD+T+E₂ group is obscured by the mean.

Figure 4.

AR and ER activation in HVC is necessary and sufficient for increases in RA firing rate and soma area. A₁–A₃, Unilateral lesion of HVC with systemic exposure to LD and T for 3 weeks. B₁–B₃, The ER antagonist faslodex and the AR antagonist flutamide unilaterally infused near HVC for 3 weeks, with systemic exposure to LD and T. C₁–C₃, DHT+E₂ unilaterally infused near HVC for 3 weeks with SD photoperiod. A₁, B₁, C₁, Firing rate (horizontal line indicates mean) in the lesioned or infused hemisphere compared with the noninfused hemisphere. A₂, B₂, C₂, Comparison of firing rate cumulative frequency distributions between the two hemispheres (see Results for statistical significance). A₃, B₃, C₃, Comparison of soma areas between the two hemispheres.

Figure 5.

AR activation in RA is permissive but not sufficient for increases in RA firing rate and soma area. A₁–A₃, DHT+E₂ unilaterally infused near RA for 3 weeks under SD. B₁–B₃, Flutamide, an AR antagonist, infused near RA for 3 weeks with LD photoperiod and systemic T. A₁, B₁, Firing rate (horizontal line indicates mean) in the infused hemisphere compared with the noninfused hemisphere. A₂, B₂, The cumulative frequency distribution in the two hemispheres (see Results for statistical significance). A₃, B₃, Comparison of soma areas between the two hemispheres.

Figure 6.

Androgen and estrogen receptors in HVC are not necessary for LD photoperiod plus systemic T-induced increases in song rate. Song rate does not differ between LD photoperiod plus systemic T-treated birds that received bilateral infusions near HVC of vehicle or a combination of AR and ER antagonists (p > 0.05).

Figure 7.

Androgen and estrogen receptors in HVC are necessary for LD photoperiod plus systemic T-induced changes in song stereotypy. A, Representative examples of stereotyped white-crowned sparrow song. All three songs were sequentially recorded on the same day. The average CV (higher CV indicates lower stereotypy; CV = “SD/mean” × 100) of the three songs is 3.77. B, Representative examples of less stereotyped song. All three songs were sequentially recorded on the same day, and their average CV is 7.22. C, Comparison of mean day 20 CVs between AR plus ER antagonist-treated birds and vehicle-treated birds. Each point represents the mean CV of one of the 16 song characteristics measured across seven birds. The dotted line is the unity line; all points above the line show less stereotypy in birds treated with antagonists compared with vehicle. Song was significantly more stereotyped on day 20 in birds exposed to vehicle compared with birds exposed to AR plus ER antagonists (p = 0.02). D, Comparison of mean CVs between days 13 and 20 in both vehicle and AR plus ER antagonist-treated birds. Song stereotypy significantly increased between days 13 and 20 in birds treated with vehicle (p < 0.0003), whereas stereotypy between days 13 and 20 in AR plus ER antagonist-treated birds did not significantly differ (p > 0.05). E, Song stereotypy on day 20 instead analyzed by calculating a mean CV for each bird. The horizontal line indicates the mean. F, Comparison of mean CVs for each bird between days 13 and 20. Song stereotypy significantly increased between days 13 and 20 in birds treated with vehicle (p < 0.0007); stereotypy in AR plus ER antagonist-treated birds did not differ (p > 0.05).

Figure 8.

Working model demonstrating that direct and indirect coactivation of ARs and ERs regulates neuronal activity in RA. Growth in RA requires both androgens and estrogens. Plasma T concentration and the enzyme aromatase that converts T to its estrogenic metabolite E₂ are regulated by photoperiod. Estradiol acts on HVC. T and/or its androgenic metabolite DHT act on both HVC and RA. Steroid receptor binding in HVC is necessary and sufficient to create a trophic signal that is transported by HVC afferents to RA. Steroid receptor binding in RA is permissive for morphological and electrophysiological change to occur in response to the trophic signal released by HVC.