Profiling of experience-regulated proteins in the songbird auditory forebrain using quantitative proteomics

Abstract

Auditory and perceptual processing of songs are required for a number of behaviors in songbirds such as vocal learning, territorial defense, mate selection and individual recognition. These neural processes are accompanied by increased expression of a few transcription factors, particularly in the caudomedial nidopallium (NCM), an auditory forebrain area believed to play a key role in auditory learning and song discrimination. However, these molecular changes are presumably part of a larger, yet uncharacterized, protein regulatory network. In order to gain further insight into this network, we performed two-dimensional differential in-gel expression (2D-DIGE) experiments, extensive protein quantification analyses, and tandem mass spectrometry in the NCM of adult songbirds hearing novel songs. A subset of proteins was selected for immunocytochemistry in NCM sections to confirm the 2D-DIGE findings and to provide additional quantitative and anatomical information. Using these methodologies, we found that stimulation of freely behaving birds with conspecific songs did not significantly impact the NCM proteome 5 min after stimulus onset. However, following 1 and 3 h of stimulation, a significant number of proteins were consistently regulated in NCM. These proteins spanned a range of functional categories that included metabolic enzymes, cytoskeletal molecules, and proteins involved in neurotransmitter secretion and calcium binding. Our findings suggest that auditory processing of vocal communication signals in freely behaving songbirds triggers a cascade of protein regulatory events that are dynamically regulated through activity-dependent changes in calcium levels.

FIG. 1

(A) Schematic representation of a parasagittal section through a zebra finch brain detailing the connectivity of the main stations of the ascending auditory pathway. The focus of the present study, NCM, receives input from the thalamo-recipient layer Field L2, and has reciprocal connectivity with the caudal mesopallium (CM). (B) Experimental approach. All animals were isolated overnight in sound-proof boxes and randomly included into one of four experimental groups. Control animals were killed after isolation without stimulation. The remaining animals were stimulated with playbacks of conspecific songs for either 5 min, 1 h or 3 h. (C) Representative brain sections were obtained with a vibratome for dissection of NCM. The red dashed lines indicate the approximate locations where sections were taken to isolate NCM. Note the darker colour of the thalamo-recipient field L2 is in these preparations, which significantly facilitates the isolation of NCM. (D) Representation of the 2D-DIGE experiments conducted in the present work. Gels were run for individual comparisons across experimental groups (control vs. 5 min; control vs. 1 h and control vs. 3 h). Each experiment was repeated three times with different animals. Intra-gel (DIA), inter-gel (BVA), as well as across-repetition (EDA) comparisons were conducted in a quantitative manner in order to identify statistically significant differences across these variables (see Methods for further details). Anatomical abbreviations not mentioned above or in the text: CN, cochlear nuclei; DM, dorsal medial mesencephalic nucleus; E, entopallium; H, hyperpalium; Hp, hippocampus; LL, lateral lemniscal nuclei; M, mesopallium; MLd, dorsal lateral mesencephalic nucleus; N, nidopallium; Ov, ovoidalis; St, striatum; v, ventricle; SO, superior olive.

FIG. 2

Principal Component Analysis (PCA) of ~2000 protein spots per condition reveals that the NCM proteome is significantly different across experimental conditions. Independent gels for each of the conditions are not significantly different from each other. The circle represents the 95% confidence interval. Values that cluster in a quadrant are significantly different than values falling within other quadrants. See Methods for description of internal standard (IS).

FIG. 3

Top: representative 2D-DIGE gels illustrating fractionated proteins from NCM in comparisons for the 1-h (control vs. 1 h; left) and 3-h (control vs. 3 h; right) groups. Control samples were labeled with the fluorophore Cy3 (green) while experimental samples were labeled with Cy5 (red). Internal control samples were labeled with Cy2 (blue, not shown; see Methods for details). Bottom: Comassie blue-stained gels illustrating differentially regulated spots (arrows) in the 1-h (left) and 3-h (right) conditions, as revealed by quantitative and statistical analyses with DeCyder software. All differentially regulated spots underwent protein fingerprinting by mass spectrometry (see Methods for details).

FIG. 4

Pie charts illustrating the cellular location (A) and molecular function (B) of the proteins identified in the present study (from Table 2). Data are expressed as percentage of the total number of differentially expressed proteins (in parentheses). Numbers adjacent to the parentheses indicate the number of proteins identified in that category. Categories were determined based upon known gene ontologies. No protein was grouped into more than one category for A and B. We note, however, that synapsin II is involved in both calcium binding and neurotransmitter secretion; we placed it in the latter category as its main function.

FIG. 5

Western blots illustrating the specificity of antibodies used for ICC experiments. These blots revealed the specific detection of SYN2, PKM2, 14-3-3 and CALB2 proteins, from left to right, in the zebra finch brain. The single bands were within the predicted molecular weight for the proteins in question with each antibody, indicating that these antibodies specifically recognize these four proteins in the zebra finch NCM. Blots were conducted with NCM tissue obtained from a bird that was killed after 1 h of song playbacks.

FIG. 6

ICC directed against selected proteins identified in the 2D-DIGE screening. Shown are representative photomicrographs of NCM sections of animals from each experimental group (columns). All photomicrographs depicted here for each individual antibody were reacted as a single immunocytochemical batch. (A) egr-1/ZENK protein is expressed in cell nuclei. (B) Synapsin II is expressed in neuropil, as suggested by the punctate staining, which is consistent with presynaptic terminals. (C and D) Pyruvate kinase is expressed in mitochondria; sections from both caudal and rostral parts of NCM are shown. (E) 14-3-3 and (F) Calbindin 2 are both expressed in the cytoplasm. Calbindin 2-positive cells were found sparsely in NCM and, thus, the images shown are at lower magnification; the insets show high-power images of representative cells for each of the conditions. Scale bars: A, C–E, µm; B, 25 µm; F, 200 µm.

FIG. 7

Quantification of immunocytochemistry data. Due to its typical terminal labeling, SYN2 immunoreactivity was quantified with standard optical density measurements (A). Immunolabeled neurons were counted as profiles for PKM2 (B), 14-3-3 (C) and CALB2 (D). With the exception of 14-3-3, significant regulation was found for all proteins. Asterisks depict significant differences (P < 0.05, anova) across groups.

FIG. 8

Schematic representation of known interacting biochemical pathways of proteins detected in this study mapped onto previously known pathways. Proteins identified to be regulated by auditory stimulation in the current and previous studies are colour-coded in red, and identified without and with asterisks, respectively. Most of the interactions of the molecular signaling pathways shown here have been determined in non-songbird species, including the ZENK binding site in the synapsin II promoter. For a detailed discussion of potential mechanisms see Discussion section and Pinaud (2005). Figure modified from Pinaud (2005).