Memory-specific correlated neuronal activity in higher-order auditory regions of a parrot

Abstract

Male budgerigars (Melopsittacus undulatus) are open-ended learners that can learn to produce new vocalisations as adults. We investigated neuronal activation in male budgerigars using the expression of the protein products of the immediate early genes zenk and c-fos in response to exposure to conspecific contact calls (CCs: that of the mate or an unfamiliar female) in three subregions (CMM, dNCM and vNCM) of the caudomedial pallium, a higher order auditory region. Significant positive correlations of Zenk expression were found between these subregions after exposure to mate CCs. In contrast, exposure to CCs of unfamiliar females produced no such correlations. These results suggest the presence of a CC-specific association among the subregions involved in auditory memory. The caudomedial pallium of the male budgerigar may have functional subdivisions that cooperate in the neuronal representation of auditory memory.

Figure 1

Sonograms of CCs. The female pre-pairing (left column) had two types of CC (A, B), and the male at 5 weeks post-pairing (right column) had five types of CC (C–G). We randomly selected five exemplar sonograms of each CC type and used them for rating the CC Similarity Score (Score). In each of 10 pairwise comparisons, the CC Similarity Score was rated as 1 (no similarity), 2 (modest similarity), or 3 (good similarity). Only type C was rated as good similarity and had a cross-correlation value (Correlation) greater than 0.75. In cross-correlations between two CCs from different CC types, values greater than 0.75 are highly unlikely to occur by chance. Type C was not present in the CC repertoire of male pre-pairing.

Figure 2

Zenk and Fos expressions in response to CCs in the budgerigar brain. (a) Coronal sections cut at level 'a' and 'b' in (b). Rectangles in the sections represent the counting frames. (b) A schematic drawing of a parasagittal plane in the auditory system. The caudal-most part of the LSt was defined as coordinates zero (*). Yellow regions indicate the caudomedial pallium (a, b) and a projection field (HVl/NFl) of the auditory B surrounding the LAM/LAN (b). Caudal and rostral lines shown in blue arrows in (b) indicate anterior and posterior auditory pathways, respectively. †; Reciprocal structural connections from the results of Farabaugh and Wild. Lat, Lateral; Ros, Rostral; D, Dorsal; V, Ventral; C, Caudal; M, Medial. (c) Photomicrographs of coronal sections of the budgerigar brain showing Zenk- and Fos-immunoreactivity. See text for details. Scale bar represents 1 mm in (a, b) and 50 μm in (c). Anatomical Abbreviations: B, Basorostralis; Cb, Cerebellum; CM, Caudal mesopallium; CMM, Caudomedial mesopallium; CoN, Cochlear nucleus; HD, Densocellular part of the hyperpallium; HP, Hippocampus; HVC, A letter-based name, a vocal nucleus in songbirds (no abbreviation); HVl, Lateral ventral hyperstriatum (former name); L1, L2, L3, Subdivisions of Field L complex; LAM, Lateral nucleus of the anterior mesopallium; LAN, Lateral nucleus of the anterior nidopallium; LL, Nucleus of the lateral lemniscus; LMD, Lamina mesopallium dorsale; LMI, Lamina mesopallium intermediate; LMV, Lamina mesopallium ventrale; LPS, Lamina pallio-subpallialis; LSt, Lateral striatum; MLd, Dorsal part of the lateral mesencephalic nucleus; NC, Caudal nidopallium; NCM, Caudomedial nidopallium; dNCM, Dorsal caudomedial nidopallium; vNCM, Ventral caudomedial nidopallium; NIVL, Ventral lateral intermediate nidopallium; NFl, Lateral frontal neostriatum (former name); NLC, Central nucleus of the lateral nidopallium; Ov, Nucleus ovoidalis of the thalamus; pB, Peri-basorostralis; VNC, Caudoventral nidopallium.

Figure 3

Differential patterns of Zenk and Fos expression in response to CCs in the caudomedial pallium. Mean (± SEM, n = 4) number of IEG-IR cells per square millimetre in the CMM, dNCM, vNCM, and hippocampus (HP) in Zenk expression (a) and in the CMM, dNCM, and vNCM in Fos expression (b). Statistical significance is indicated by the asterisks (*P < 0.05; **P < 0.01).

Figure 4

Analysis of correlations in Zenk expression between subregions of the caudomedial pallium after CC stimulation. (a) Scatterplots of number of Zenk-IR cells per square millimetre activated in response to mate CCs (right side of each figure), or unfamiliar CCs, or silence (left side of each figure) are plotted for individual subjects. Shown are correlations in CMM versus dNCM (left), CMM versus vNCM (centre), and dNCM versus vNCM (right). There were three significant linear relationships between subregions with Bayesian analysis, where linear regression lines in red are fitted. (b) The statistics of the data in (a) are given in the table with CIs of Bayesian analysis, correlation coefficients (r) and P values of Pearson's correlation analysis. Statistically significant values are highlighted. (c) Shown are patterns of correlation across subregions of the number of Zenk-IR cells activated in response to mate CC (left), unfamiliar CC (centre), or silence (right). Within a group, the expressions were scaled to the data of the bird with the highest Zenk expression in each of the three subregions, and the CMM, dNCM, and vNCM values for the same subject are plotted as percentages and connected with lines. The values of the first-ranked bird in the CMM within each group are indicated by circles in the figure. The values of the second-, third-, and fourth-ranked bird in the CMM within each group are indicated by squares, stars, and diamonds, respectively. The types of stimuli are color-coded as in (a). Note that only in the case of mate CC (left) was a well-organised and ordered 'triangle' relationship observed within the group. In the other cases, i.e., unfamiliar CC (centre) or silence (right), these triangles overlap irregularly or do not form. Inter-subregion lines drawn in red indicate significant linear relationships between subregions with Bayesian analysis.

Figure 5

Analysis of correlations in Fos expression between subregions of the caudomedial pallium after CC stimulation. Shown are data with the same conventions as in Fig. 4 but pertaining to the Fos-immunopositive cell counts. Unlike the results revealed by Zenk expression (Fig. 4), there was no linear relationship between subregions for the MATE, the UNFAMILIAR or the SILENCE group with Bayesian analysis.