Figure 1.
Neural substrates for singing and song learning.
A and B: schematics of the song control system and relevant auditory structures. Represented are the posterior vocal-motor pathway (VMP, blue), and the anterior forebrain pathway (AFP, red). HVC is the origin of both pathways and the entry site of inputs from auditory areas (yellow) into the song system. For abbreviations, see table 1. In A, brain topography is preserved, for easier comparison with experimental brain sections. In B, indication of broad brain subdivisions (on the right) facilitates comparison with mammalian brains.
Table 1.
Abbreviations for zebra finch brain structures.
Figure 2.
Simplified schematic of the all-trans retinoic acid (ATRA) signaling pathway.
ATRA is synthesized from retinol (Vitamin A) in a two-step oxidation process which takes place in the cytoplasm. To induce transcription, retinoic acid receptors (RARs) need to be activated by ATRA binding, and both RARs and retinoid X receptors (RXRs) need to bind retinoic acid response elements (RAREs) in the vicinity of target genes. Several CYP26 isoforms are responsible for ATRA degradation. Due to its small size and hydro- and lipophilic properties, ATRA is able to cross cell membranes and thus act in other cells.
Figure 3.
RXRα expression in adult male zebra finch brain.
Drawings on the left depict brain areas and nuclei of serial parasagittal sections. The corresponding sections on the right were processed for in situ hybridization (ISH) for RXRα. For all images, anterior is to the right and dorsal is up; medial to lateral levels are represented from top to bottom. Scale bar = 2 mm (all panels). For abbreviations, see table 1.
Figure 4.
RXRα expression in song control nuclei of adult male zebra finch.
A: Schematic of a parasagittal section at the level of song nuclei HVC, RA, Area X, and lMAN. The dashed rectangles indicate the areas shown in B, D, and G. Photos in B–H are bright field views of parasagittal sections hybridized with digoxigenin-labeled RXRα antisense probe. Anterior is right, dorsal is up in all panels. B, C: Expression in RA. Many cells were strongly labeled; the surrounding arcopallium showed less labeling. D-F: Expression in lMAN and Area X. Expression is comparable to the surrounding areas, but lower in Area X/striatum than in lMAN/nidopallium. G, H: Expression in HVC. The high expression is comparable to the adjacent nidopallium. In C, E, F, and H, black arrows depict strongly labeled cells, gray arrows weakly labeled cells, and white arrows unlabeled cells. Scale bars = 200µm in B, D, G; 20µm in C, E, F, H.
Figure 5.
RXRα expression in higher auditory areas of adult male zebra finch telencephalon.
A: Drawing of a parasagittal section at the level of the auditory areas NCM, Field L2a, and CMM. The dashed rectangle indicates the area shown in B. B–D: Bright-field views of parasagittal sections processed for RXRα in situ hybridization. Anterior is right, dorsal is up in all panels. B: Low power view of higher auditory areas, with strong labeling in NCM and CMM and lower labeling in L2a. C, D: High power views of NCM and CMM, respectively; black arrows depict strongly labeled cells in clusters, white arrows depict unlabeled cells. Scale bars = 200µm in B; 20µm in C, D.
Figure 6.
RXRγ expression in Area X of adult male zebra finch.
A: Drawing of a parasagittal section of adult brain, indicating detail area shown in B; anterior is to the right, dorsal is up. For abbreviations see table 1. B: Detail view of Area X and surrounding area in section processed for RXRγ ISH showing sparse labeled cells in Area X. D: High-magnification view of Area X; black and white arrows depict labeled and unlabeled cells, respectively. Scale bars: 2mm in B, 200µm in C, 50µm in D.
Figure 7.
ATRA mapping in brain sections through a reporter cell assay.
A and B: Schematic overview of ATRA reporter cell assay. A: Reporter cells (they contain a LacZ gene under a retinoic acid response element, or RARE, and express retinoic acid and retinoid X receptors - RAR/RXRs - needed for ATRA induced gene expression) are seeded onto a Petri dish; a freshly cut brain section is placed into the dish and attaches to the cell monolayer. B: ATRA locally generated in the brain slice reaches a reporter cell, binds to RAR/RXR complexes and causes LacZ expression, revealed as blue label by LacZ/X-gal staining. C and D: LacZ expression in reporter cells is specifically induced by ATRA. C (negative control 1): In the absence of ATRA, reporter cells are LacZ-negative and do not turn blue upon LacZ/X-gal staining. D (positive control): Labeling is generated when ATRA is added to the medium. E (negative control 2): A slice co-cultured with an F9 cell line without RARE and LacZ does not generate blue signal upon LacZ/X-gal staining. F: A slice from an adult male bird co-cultured with reporter cells results in blue labeling in regions where ATRA is present. Blue labeling is seen under song nucleus HVC, which expresses zRalDH, but not in cells without overlying tissue (top), or in cells that underlie a part of the tissue that does not contain ATRA (bottom). Photos in C–F were taken through the bottom of the Petri dish. Scale bars: for C, D = 200µm; for E, F = 100µm.
Figure 8.
ATRA-induced reporter expression in song nuclei HVC and RA of adult male zebra finch is consistent with zRalDH expression.
The drawing illustrates regions shown in A (HVC) and B (lMAN), abbreviations in table 1. Left panels show ISH for zRalDH, right panels show detection of ATRA (blue label) in brain sections by reporter assay. HVC but not the surrounds, and lMAN and its surrounds both strongly express zRalDH and induce reporter expression. In all panels, frontal is to the right and dorsal is up.
Figure 9.
Induction of ATRA reporter expression by song nuclei that do not express zRalDH in adult male zebra finch.
Line drawing on top indicates position of song nuclei. Left panels: detail views of zRalDH expression by ISH. Middle panels: Sites of ATRA presence by reporter cell assay. Right panels: summary and examples of transcript distribution for retinoid receptors. Data for RAR expression are from Jeong et al., 2005. Brain diagram on top indicates position of song control nuclei. In all panels, frontal is to the right and dorsal is up. A: Area X; B: RA. In both nuclei, zRalDH is not expressed, but ATRA-induced reporter expression is detected, as well as receptors that may mediate ATRA effects.
Figure 10.
zRalDH protein expression in areas of adult male zebra finch brain that express or lack zRalDH mRNA.
A: Drawing indicates location of images shown in other panels. B: zRalDH mRNA expression by in situ hybridization, level similar to A. C–G: zRalDH protein detection through immunohistochemistry. C: HVC and RA, as well as the fiber tracts extending from HVC to RA (arrows) are labeled. D: lMAN and surrounding nidopallium, as well as Area X, are labeled; labeling in Area X is diffuse and not in somata. Thus, besides HVC and lMAN, which express zRalDH mRNA, protein is present in two song nuclei (X and RA) that lack zRalDH transcript (white arrows in B). E: Detail view of zRalDH protein in RA; labeling is diffuse and not in somata. F: Detail view of zRalDH protein in HVC, somata are labeled. G: High power view of zRalDH in lMAN, somata are labeled. In all panels, frontal is to the right, and dorsal is up. For abbreviations, see table 1.Scale bars for C = 0.5 mm; D–F = 100µm, G = 50µm.
Figure 11.
zRalDH immunoreactivity in HVC and lMAN neurons that project to RA (HVCRA and LMANRA).
Schematic drawing on top illustrates injection of the retrograde tracer cholera toxin subunit B (CTB) and location of retrogradely labeled neurons. A and B: High power views of HVC (A) and lMAN (B) in zRalDH-immunolabeled sections from an adult male zebra finch that received an CTB injection into RA resulting in retrogradely labeled neurons in HVC and lMAN. Left panels show bright field views of zRalDH protein expression, right panels show CTB signal (red) and cell nuclei stained with DAPI (blue) in the same fields. Arrowheads point to zRalDH-immunoreactive cells that are retrogradely labeled. Scale bars: 20µm.
Figure 12.
In the absence of axonal input from HVC, ATRA induced reporter staining is reduced in RA of adult male zebra finches.
Experimental design and time course are indicated by diagrams on the left and on top of photomicrographs. After surgical procedure, birds were allowed to survive for 14 days before reporter assay was performed. A–D: High power views of reporter expression induced in the monolayer by RA; dashed circles indicate position of RA. A and C: control hemispheres with intact HVC-to-RA projections. B and D: experimental hemispheres with HVC lesion (B) or knife-cut fibers (D).
Figure 13.
RXRα expression in Area X, lMAN, and RA in relation to time of day and age.
A: RXRα expression across the day. Bars represent the difference in normalized expression strength between a song nucleus and the surrounding area (“Δ RXRα inside nucleus-outside”, mean ± SEM); a 0 value indicates no expression difference, positive values indicate more expression inside than outside, negative values the opposite. At different times, expression compared to the surrounds varied significantly in lMAN, but not in Area X or RA; asterisk indicates significance by MANCOVA (Area X: F = 0.79, p = 0.471; lMAN: F = 5.7, p = 0.014; RA: F = 0.31, p = 0.736); for details, see Methods. During the day, the RXRα expression inside and outside lMAN was similar, whereas RXRα expression was lower in lMAN than in the surrounding nidopallium in the morning and evening. B: Representative bright field images of sagittal sections showing variation in RXRα expression in lMAN compared to surrounding nidopallium at different times of the day. C: RXRα expression across ages. Plotted is the same measure of expression strength as in A vs. age (post-hatch days), and linear fits; a significant decrease with age was seen in lMAN (MANCOVA, F = 4.72, p = 0.046) but not in Area X (F = 2.13, p = 0.165) or RA (F = 0.54, p = 0.476).
Figure 14.
CYP26B1 expression in adult zebra finch brain.
The drawings on the left depict brain areas of the parasagittal sections on the right. The right panels show sections processed for ISH for CYP26B1. For all images, anterior is to the right and dorsal is up. A and B are medial sections, C a lateral one. CYP26B1 expression is sparse, mostly in CMM and CLM (black arrowheads), and does not overlap with zRalDH (see fig. 14, 15, and fig. 7). CYP26B1 is also expressed in the fronto-lateral mesopallium (empty arrow), in the medial habenula (white arrowhead), in the caudal-dorsal hippocampus (white arrow), and in cerebellar Purkinje cells. For abbreviations, see table 1. Scale bars = 1 mm.
Figure 15.
CYP26B1 and zRalDH exhibit graded expression in higher auditory areas of adult zebra finch.
The drawing in the top left indicates approximate regions shown in A–E; the drawings on the left depict brain areas shown on the right. The middle and right columns show sections processed for ISH for CYP26B1 and zRalDH. For all images, anterior is to the right and dorsal is up; medial to lateral is represented from top to bottom. CYP26B1 is expressed in a dorsoventral gradient-like pattern throughout the caudal, caudo-medial and caudo-lateral mesopallium (CM, CMM and CLM; arrows in middle panels point to region of high expression). This distribution does not overlap with zRalDH mRNA expression, which is absent in the mesopallium; some cells in the rostral part of the caudal nidopallium express zRalDH (arrows in right panels), tapering off caudally. For other abbreviations, see table 1. Scale bar = 0.5 mm.
Figure 16.
Overlapping expression of RARs and RXRs differ across song control nuclei, as well as between most song nuclei and their surroundings.
Expression strength of all receptors based on visual estimates are schematically represented by symbol size (for RARs, data are from Jeong et al. (2005)). For all song nuclei except HVC, the receptor expression profile differs from the surroundings. Thus, each song nucleus has a unique combination of RAR(s)/RXR(s), which may lead to nucleus-specific sets of ATRA effects.