Pleiotropic Control by Testosterone of a Learned Vocal Behavior and Its Underlying Neuroplasticity(1,2,3)

Abstract

Steroid hormones coordinate multiple aspects of behavior and physiology. The same hormone often regulates different aspects of a single behavior and its underlying neuroplasticity. This pleiotropic regulation of behavior and physiology is not well understood. Here, we investigated the orchestration by testosterone (T) of birdsong and its neural substrate, the song control system. Male canaries were castrated and received stereotaxic implants filled with T in select brain areas. Implanting T solely in the medial preoptic nucleus (POM) increased the motivation to sing, but did not enhance aspects of song quality such as acoustic structure and stereotypy. In birds implanted with T solely in HVC (proper name), a key sensorimotor region of the song control system, little or no song was observed, similar to castrates that received no T implants of any sort. However, implanting T in HVC and POM simultaneously rescued all measures of song quality. Song amplitude, though, was still lower than what was observed in birds receiving peripheral T treatment. T in POM enhanced HVC volume bilaterally, likely due to activity-dependent changes resulting from an enhanced song rate. T directly in HVC, without increasing song rate, enhanced HVC volume on the ipsilateral side only. T in HVC enhanced the incorporation and recruitment of new neurons into this nucleus, while singing activity can independently influence the incorporation of new neurons into HVC. These results have broad implications for how steroid hormones integrate across different brain regions to coordinate complex social behaviors.

Keywords: medial preoptic area; neurogenesis; singing behavior; social behavior; songbirds.

Figure 1.

A, B, Representative photomicrographs illustrating the typical implant localizations for POM (A) and HVC (B). Cannulae tips seen here are adjacent to the nucleus of interest to minimize damage to the nucleus. Arrows outline the borders of each nucleus. A Nissl stain was used to visualize the nuclei. Scale bar, 200 μm. Hp, Hippocampus.

Figure 2.

Semi-schematic presentation of the implant locations and their content. A–H, Successive locations in or around POM (A–D) or HVC (E–H), presented in each case in a rostral-to-caudal order. T-filled implants that contacted their target are represented by black circles, T-filled implants that failed to reach their target are represented by open circles. Empty implants are represented by squares and empty implants in the PER group (birds with a Silastic subcutaneous T-filled implants) are represented by diamonds. Since each bird received two brain implants (T-filled or empty), all symbols are associated with a number that allows reconstructing each pair of implants. In A, B, F, and G, implants near POM or HVC have been slightly spread apart to preserve visibility. AVT, Area ventralis of Tsai; CA, commissura anterior; CO, chiasma opticum; FPL, fasciculus prosencephalis lateralis; HA, hyperpallium apicale; Hp, hippocampus; LAD, lamina arcopalllialis dorsalis; LPS, lamina pallio-subpallialis; LV, lateral ventricle; NIII, nervus oculomotorius; OM, tractus occipitomesencephalicus; OV, nucleus ovoidalis; PVN, nucleus paraventricularis; Rt, nucleus rotundus; Tn, nucleus taeniae of the amygdala; Tu, tuberal hypothalamus.

Figure 3.

Effects of T implants on POM, HVC, and both on POM volume in the hemisphere implanted with T [ipsilateral (ipsi)] and in the opposite hemisphere [contralateral (contra)]. Bars represent the mean ± SEM. The same letter above individual groups indicates the absence of statistical differences, and different letters above bars indicate a significant difference. The asterisks refer to differences between ipsilateral and contralateral sides. PER-T, n = 7; HVC-POM T, n = 7; POM-T, n = 8; HVC-T, n = 5; HVC-POM NO T, n = 4. Differences were considered significant at p ≤ 0.05.

Figure 4.

A–C, Effects of treatments on serum concentrations of testosterone (A) and on two peripheral measures of T action, the cloacal protuberance (CP) width (B) and the syrinx mass (C). Bars represent the mean ± SEM. The same letter above individual groups indicates the absence of statistical differences, and different letters above bars indicate a significant difference. PER-T, n = 7; HVC-POM T, n = 7; POM-T, n = 8; HVC-T, n = 5; HVC-POM NO T, n = 4. Differences were considered significant at p ≤ 0.05.

Figure 5.

Effects of treatments on various acoustic and stereotypy measures of song. Bars represent the mean ± SEM. The same letter above individual groups indicates the absence of statistical differences, and different letters above bars indicate a significant difference. A, An example of a male canary song. B, F, The effects on measures of song acoustic stereotypy (lower CV means higher stereotypy). G, H, The effects on measures of song complexity (acoustic variance correlates positively with song complexity; see Materials and Methods). I, The effects on song energy (a measure of song amplitude). C–E, Histograms for individual birds from PER-T, HVC-POM T, and POM-T groups, respectively, to demonstrate differences in bandwidth stereotypy (the higher the CV, the lower the stereotypy). PER-T, n = 7; HVC-POM T, n = 7; POM-T, n = 8; HVC-T, n = 5; HVC-POM NO T, n = 4. Differences were considered significant at p ≤ 0.05.

Figure 6.

A–C, Effects of T implants on the volume of the song control nuclei HVC (A), RA (B), and Area X (C) in the hemisphere implanted with T [ipsilateral (ipsi)] and in the opposite side [contralateral (contra)]. Bars represent the mean ± SEM. The same letter above individual groups indicates the absence of statistical differences, and different letters above bars indicate a significant difference. The asterisks refer to differences between ipsilateral and contralateral sides. PER-T, n = 7; HVC-POM T, n = 7; POM-T, n = 8; HVC-T, n = 5; HVC-POM NO T, n = 4. Differences were considered significant at p ≤ 0.05.

Figure 7.

Relationship between song rate (number/minute), and the total volumes of HVC, RA, and Area X. Individual data points from the different experimental groups are coded by different colors. The equation for the regression line, the Pearson’s correlation coefficient r, and the associated p value are shown in the bottom right portion of each correlation graph. PER-T, n = 7; HVC-POM T, n = 7; POM-T, n = 8; HVC-T, n = 5; HVC-POM NO T, n = 4.

Figure 8.

Effects of treatments on the number of DCX-IR cells in HVC. A, B, The numbers of fusiform (A) and round (B) cells in the hemispheres contralateral (contra) and ipsilateral (ipsi) to the implant site, Bars represent the mean ± SEM. The asterisk refers to a significant difference between ipsilateral and contralateral sides. #A difference between the ipsilateral and contralateral side, but at p = 0.08. The same letter above individual groups indicates the absence of statistical differences, and different letters above bars indicate a significant difference. PER-T, n = 7; HVC-POM T, n = 6 (one bird had damage within HVC that precluded DCX quantification; see Materials and Methods); POM-T, n = 8; HVC-T, n = 5; HVC-POM NO T, n = 4. Differences were considered significant at p ≤ 0.05.

Figure 9.

A, B, Relationship between song rate (number/minute) and the numbers of DCX-IR cells in HVC calculated separately for the two types of DCX-IR cells, fusiform (A) and round (B), across hemispheres (A, B) and on the contralateral (Contra; C, D) side. Individual data points from the different experimental groups are coded by different colors. Correlations were calculated using Pearson’s r. The equation for the regression line, the Pearson’s correlation coefficient r, and the associated p value are shown in the bottom right portion of each graph. PER-T, n = 7; HVC-POM T, n = 6 (one bird had damage within HVC that precluded DCX quantification; see Materials and Methods); POM-T, n = 8; HVC-T, n = 5; HVC-POM NO T, n = 4.