Reactive Neurogenesis in Response to Naturally Occurring Apoptosis in an Adult Brain

Apoptosis in HVC occurs rapidly after transition from breeding to nonbreeding conditions. A, Number of Hu positive neurons in one hemisphere of HVC across the time course of regression. HVC neuron number drops by 7 d following transition to nonbreeding conditions. B, Number of apoptotic cells in HVC on one side of the brain over the time course of regression. Number of AC-3-positive cells peaks at 2 d following transition from breeding to nonbreeding conditions. Letters above bars indicate significant differences among groups (one-way ANOVA, Tukey post hoc pairwise comparisons). All data plotted as mean ± SEM. C, Representative image of AC-3 immunohistochemistry in fresh frozen tissue counter-stained for Nissl (cytoplasm stains light purple, nuclei stain dark purple) illustrating apoptotic cells in HVC. The arrow indicates an AC-3-positive cell (brown). Note the condensed chromatin (a distinctive feature of apoptotic cells; shown in dark purple) contained within the AC-3-positive cell.

Proliferation in the vVZ changes rapidly during regression of the song control circuits. A, An image of Hu staining in the brain illustrating the method for counting BrdU-positive cells in the VZ. Arrowheads demonstrate the medial and lateral-most extents within which BrdU-positive cells were counted on the ventral side of the VZ. The star indicates the location of the image shown in B. HP, Hippocampus; TEL, telencephalon. B, A representative image of BrdU immunohistochemistry showing proliferative cells and their progeny in the VZ. BrdU labeling in the nucleus is stained dark purple whereas Hu labeling in the cytoplasm is stained brown. The arrow indicates BrdU-positive cells (i.e., neural progenitor cells or daughter cells) labeled at the time of BrdU injections. The arrowheads point to BrdU-positive cells on the dorsal side of the VZ (ventral to HP) that were not included in counts of BrdU-positive cells in the vVZ due to their inability to migrate across the ventricle to HVC. C, Number of BrdU-positive cells in the vVZ of one hemisphere over the experimental time course. Proliferation in the vVZ peaked at 4 d following transition from breeding to nonbreeding conditions. Letters above bars indicate significant differences among groups. All data plotted as mean ± SEM.

The relationship between proliferation in the vVZ and the number of apoptotic cells in HVC during rapid regression of the song control system. A, The lag in time of peak vVZ proliferation compared with HVC cell death. The levels of cell death in HVC and proliferation in the vVZ are plotted relative to the maximum value of each measure to facilitate comparison of the time course of both measures. Proliferation in the vVZ peaks at 4 d following transition from breeding to nonbreeding conditions and 2–3 d following the peak of HVC cell death. B, The correlation between proliferation in the vVZ and the number of apoptotic cells in HVC during rapid regression of the song control system. All data pooled is indicated by the dashed line (R² = 0.0464, p = 0.2140, n = 30 birds), pooled baseline condition data (i.e., 28 d of LD+T and SD; n = 9 birds) by the dark gray diamonds (individual data points) and dark gray solid line, and pooled active regression data (i.e., 1, 2, 4, 7, and 14 d of SD; n = 21 birds) by the light gray diamonds and solid line. See Table 3 for correlations between vVZ proliferation and AC-3 cell number within individual time points.

Cell death in HVC is functionally linked to stem-cell proliferation in the vVZ. A, Number of apoptotic cells in HVC of birds infused with vehicle or caspase inhibitor cocktail infusion near HVC for 3 d into the seasonally induced regression of HVC. The number of AC-3-positive cells was significantly lower in birds infused with the caspase inhibitor cocktail, confirming that the infusion prevented HVC cell death. B, Number of BrdU-positive cells in the vVZ of birds infused with vehicle (n = 5) or caspase inhibitor cocktail (n = 6) near HVC after 3 d of SD. Proliferation in the vVZ was significantly reduced in birds receiving the caspase inhibitor cocktail compared with those receiving vehicle following transition from breeding to nonbreeding conditions. Asterisk above bars indicate significant differences among groups. All data plotted as mean ± SEM.

The effect of age on neuronal turnover in male white-crowned sparrows following HVC regression. A, HVC cell death negatively correlates with sparrow age at stable baseline breeding and nonbreeding conditions (i.e., LD+T and SD 28 d). Solid line is the regression line from pooled LD+T and SD 28 d birds, whereas gray dots represent the value for each bird (n = 11). B, Proliferation in the vVZ negatively correlates with sparrow age in stable baseline conditions (n = 11). C, HVC volume does not change with age. Maximal HVC growth (i.e., LD+T volume; n = 4) correlation with age is represented by the dotted line and complete HVC regression (i.e., 28 d of SD volume; n = 7) correlation with age is indicated by the solid line. For all scatter plots individual birds' values and age are represented by one dot: light gray for LD+T and dark gray for 28 d of SD.

Song degrades as Gambel's white-crowned sparrows are transferred from breeding to nonbreeding conditions. A, Spectrograms of songs produced by three different individuals before and after the birds were transferred to nonbreeding conditions. The manner in which song degraded varied between birds. Bird 1 song illustrates degradation of the spectral properties within individual syllables. Birds 2 and 3 continued to sing individual syllables similar in spectral properties to those sung under breeding condition, but sang fewer syllables over time. Bird 3 stopped singing before 14 d of SD, as did 40% of the birds remaining in the experiment through at least 14 d of SD. B, Song completeness across the time course of regression plotted as percentage of songs that contained each of the syllable types produced under breeding conditions. Song completeness decreased significantly by 4 d into nonbreeding conditions. C, Cross-correlation, a measure of similarity, of song from 21 and 23 d of LD+T and days 1–14 of SD compared with 21 d of LD+T song. Song similarity decreased significantly by 7 d of SD relative to 23 d of LD+T. Letters above bars indicate significant differences among groups. All data plotted as mean ± SEM.