The pallial basal ganglia pathway modulates the behaviorally driven gene expression of the motor pathway

Keywords: immediate-early gene, motor-driven gene expression, song, zebra finch, ZENK

Animals

We used 85 adult (more than 90 days old) male zebra finches (Taeniopygia guttata) from our breeding colony. In zebra finches only males learn to sing. Females were used as stimuli for directed singing. Animal treatments and procedures were approved by the Duke University Institutional Animal Care and Use Committee.

Anatomy

Because the boundary between MAreaX and LAreaX had not been determined in the prior study (Jarvis et al., 1998), as their sections were processed in the sagittal plane, we defined it here using directed singing-driven gene expression patterns in frontal sections that spanned medial and lateral AreaX. We found a MAreaX–LAreaX boundary that is diagonally positioned and starts at an angle where AreaX dips ventrally to become larger (Fig. 3A, higher power image in Supplementary material, Fig. S1). Using this functionally defined boundary, we calculated that LAreaX is 19.69 ± 0.48 times larger than MAreaX (n = 18 animals singing directed song). Because this boundary is not visible after undirected singing, when we quantified ZENK expression in MAreaX, we did so in an area within MAreaX but more medial to the expected (undirected) or observed (directed) functionally defined boundary.

Surgery

Unilateral ibotenic acid lesions were made to LAreaX, LMAN, LAreaX + LMAN, MAreaX, MMAN and MAreaX + MMAN. We used ibotenic acid, as opposed to electrolytic lesions, to minimize damage to fibers of passage through a vocal nucleus. Ibotenic acid lesions to cell bodies cause degeneration of their axons and terminals to their connected neurons within 48 h (Milner & Veznedaroglu, 1993; Halim & Swerdlow, 2000). Ibotenic acid (Sigma, USA) was dissolved in 1 m NaCl to a concentration of 1% (pH 7.5–8.0). Fresh aliquots were used for each experiment. To lesion LAreaX and LAreaX + LMAN, surgeries were performed with isoflurane anesthesia (1–3%; flow 1 L/min), and ibotenic acid was injected using a single-barrel glass micropipette attached to a Nanoject II injector (Drummond Scientific, USA). The bird was fixed in the stereotaxic apparatus and the location of a vocal nucleus found according to the following coordinates. For LAreaX, two injections of 55.2 nL each were made within 3.9–4.4 mm rostrally, at 1.3 mm laterally and 3.5 mm ventrally, from bregma. For LAreaX + LMAN together, two injections of 55.2 nL were made at the same coordinates and one additional injection of 55.2 nL at 2.2–2.4 mm ventrally. The coordinates were adjusted individually for each bird according to the size and shape of the skull. A different approach was taken to lesion LMAN, MAreaX and MMAN alone, because of their smaller size. Surgeries were performed with ketamine–xylazine anesthesia (15 mg/mL ketamine: 3 mg/mL xylazine; 40 µL/12 mg of bird weight), which enables recording of the spontaneous activity in LMAN. Spontaneous activity of vocal nuclei is different from the surrounding brain areas, allowing us to locate them. To record this activity, one barrel of a double-barrel glass micropipette was filled with 1 m NaCl, attached to a silver wire, and used as an electrode. The other barrel was attached to the Nanoject II injector and filled with 1% ibotenic acid. Once found (coordinates: 4.0–4.8 mm rostrally, 1.6–1.8 mm laterally, 2.0–2.6 mm ventrally), LMAN was either lesioned (three injections of 32.2 nL) or the micropipette was moved medially to MMAN and/or ventrally to MAreaX and these nuclei were lesioned (coordinates: 0.35 mm laterally, 2.0–2.3 mm ventrally for MMAN, 2.7–2.8 mm ventrally for MAreaX; one injection each of 32.2 nL). Left and right sides were alternated. After all surgeries, the bird was kept under a heat lamp overnight and moved the next day to a recording sound isolation box.

Lesion size was determined using Hu⁺ immunostaining (a neuronal marker; Barami et al., 1995) on brain sections, which unequivocally allowed identification of the lesioned area as opposed to Nissl staining, which is not as unequivocal. In addition, vocal nuclei neuronal morphology and density are different from the surrounding areas allowing identification of their boundaries using Hu staining. Non-lesioned parts of vocal nuclei, when present, were also identified by singing-induced ZENK expression. To determine lesion size, the area of the remaining non-lesioned part of a nucleus (0 when completely lesioned) and the area of the intact nucleus in the control hemisphere was calculated in every fourth section throughout a vocal nucleus using the lasso tool, then histogram, pixel number function in Adobe Photoshop software. The measurements for each hemisphere were summed, and the percentage of the non-lesioned part was calculated and subtracted from 100 to generate lesion size as a percentage. The comparison to the control hemisphere allowed us to control for variations in vocal nuclei sizes between animals. The lesion sizes for each bird and lesion type are shown in Table 1. Because medial vs. lateral divisions of AreaX have not been well studied, specificity of the MAreaX lesions is provided in Supplementary Fig. S2, A. The lesions damaged the whole MAreaX or its medial, dorsal or ventral portion. None of the six birds had LAreaX damaged. Two other birds with lesions that crossed the MAreaX/LAreaX boundary (not shown) were excluded from the analyses for clarity of interpretation.

We used Hu labeling and LANT6 and parvalbumin double-labeling to assess the percentage of neurons and type of neurons lesioned within the lesioned area. LANT6⁺/parvalbumin⁻ AreaX neurons project to DLM (Reiner et al., 2004a). We found that within the lesioned area, ibotenic acid eliminated more than 96 ± 2.2% and 86 ± 1.6% of the Hu⁺ neurons in LAreaX and LMAN, respectively (n = 6 animals each). The small percentages of neurons remaining in the lesioned AreaX were not projection neurons (LANT6⁺/parvalbumin⁻, Fig. 2), and thus the lesions removed all output from AreaX. The small percentage of neurons remaining in the lesioned area of LMAN was of neurons small in size (4 ± 0.8 µm diameter vs. 9.2 ± 1.4 µm diameter of large neurons in control LMAN; based on 50 neurons from four birds), and thus are also unlikely projection neurons. Alternatively, the lesions could have reduced the size of residual projection neurons.

Tracer injections

In eight birds we bilaterally targeted MAreaX (n = 16 injections total) with dextran amine Alexa Fluor^R 488 (Molecular Probes, USA), which is a combined anterograde and retrograde tracer. We used the same coordinates as listed above for lesions and made one 32.2-nL injection of a 5% tracer solution in sterile phosphate-buffered saline (PBS). Five days after surgery, singing behavior was obtained, and the injection location relative to vocal nuclei was identified by singing-driven ZENK protein expression. We found that our exact injection locations varied as follows: predominantly in MAreaX and surrounding striatum (n = 8), MAreaX with MMAN along the injection track (n = 4), the medial striatum (n = 2) and the medial nidopallium caudal to MMAN (n = 2).

Singing behavior

Singing behavior was recorded before surgery and the first morning the bird started to sing after surgery. All birds were able to sing after unilateral lesions. The birds were killed on the second–11th day after surgery, but usually on the third–fifth day, after an overnight period of silence in a sound isolation box to prevent examining ZENK induction as a result of unobserved behaviors. The male was either alone in the cage (for undirected singing) or with a female that was placed, while in the dark overnight, in the other half of the cage separated by a cage wall barrier (for directed singing). We found that the barrier encouraged more directed singing. Most birds would start singing within minutes after the lights came on. In the presence of the female, song was scored as directed when the bird was facing the female during singing or undirected when the bird did not face the female during singing. Sometimes an additional female was added midway into the singing session to induce more directed singing. If a bird did not sing 20 or more bouts within a 55-min period, the experiment was repeated on subsequent days until we obtained a sufficient amount of singing. Singing was recorded using Sound Analysis Life 229 (Tchernichovski et al., 2004). The number of song bouts was counted and we refer to them as songs. A song was defined as vocalizations that started usually with several introductory notes, continued with at least one motif, and was surrounded by more than 2 s of silence. We counted song bouts instead of motifs or singing time because in control animals without surgery the number of song bouts showed the highest correlation with the number of neurons that expressed ZENK protein in vocal nuclei (P < 0.0001, r = 0.96 for number of song bouts; P = 0.0004, r = 0.94 for number of motifs; P = 0.0009, r = 0.93 for singing time; correlations for HVC, n = 7 birds), similarly as shown for mRNA (Jarvis et al., 1998). After the 55-min singing session, the bird was injected with a lethal dose of equithesin and then within another 5 min the bird was perfused transcardially with PBS and then with 4% paraformaldehyde in PBS. The brain was dissected, post-fixed for 5 h, and cryoprotected in 20% and then 30% sucrose in PBS at 4 °C. The frozen brain was stored at −80 °C until used for immunocytochemistry. Brains of intact singers were taken under the same behavioral paradigm, and those of silent intact controls were taken after an overnight period followed by at least 1 h of lights on with no singing.

Immunocytochemistry

The brains were cut on a cryostat in 30-µm coronal sections, and the sections were collected in PBS. Every fourth section was used for Hu and ZENK double-labeling. Up to eight animals from different groups were processed together for each immunocytochemistry experiment. Adjacent sections from several brains were used for LANT6 (pallidal neuron and striatal interneuron marker) and parvalbumin (striatal interneuron marker) double-labeling (Reiner et al., 2004a). We used the mouse monoclonal anti-Hu (Molecular Probes), diluted 1 : 500, the rabbit polyclonal anti-egr-1 (ZENK; Santa Cruz Biotech, Santa Cruz, CA, USA), diluted 1 : 200, the rabbit polyclonal anti-LANT6 (provided generously by Dr R. Carraway, University of Massachusetts, Worchester, USA), diluted 1 : 1000, and the mouse monoclonal anti-parvalbumin (Sigma), diluted 1 : 1000. The specificity of these antibodies for birds has been described in previous studies (Reiner & Carraway, 1987; Barami et al., 1995; Mello & Ribeiro, 1998; Reiner et al., 2004a). The free-floating sections were washed 3 × in PBS, and non-specific binding was blocked in 0.1% bovine serum albumin with the addition of 0.3% Triton X-100 in PBS for 1 h. The sections were then incubated overnight (Hu and ZENK) or over 2 nights (LANT6 and parvalbumin) in the mixture of primary antiseras in blocking solution at 4 °C. After three washes in PBS, the sections were incubated for 2 h with Cy3-conjugated donkey anti-mouse IgG and FITC-conjugated donkey anti-rabbit IgG (Jackson Immunoresearch, USA); their concentrations were 1 : 200 and 1 : 50, respectively. For the tracer-injected animals, we used Cy3-conjugated donkey anti-rabbit IgG against the egr-1 antibody. After three washes in PBS, the sections were mounted on silanated slides, washed in deionized water and coverslipped with DAPI to stain cell nuclei (Vectashield, USA).

Gene expression quantification

We measured ZENK protein levels because protein is the functional molecule and we could perform double-label immunohistochemistry to determine the portion of neurons that co-expressed ZENK. Images of immunolabeled vocal nuclei from injected and control hemispheres were taken with a Spot III CCD camera attached to a Leica DMRXA2 microscope under 2.5 × and 10 × objectives using Spot 3.2.4 software (Diagnostic Instruments). The images taken with the 10 × objective were used for counting. The total number of Hu-labeled and ZENK-labeled cells was manually counted in representative 0.06-mm² areas for LAreaX, LMAN and HVC, in 0.02-mm² areas for MAreaX and MMAN, and in the whole nucleus in sections with RA. To count double- and single-labeled neurons, we created a layer in Adobe Photoshop of the ZENK image, then marked the 0.02-mm² or 0.06-mm² area, and counted all the labeled nuclei using the brush tool to mark counted cells; we considered the cell as ZENK⁺ if we could recognize the circle-shaped cell nucleus. Then we copied this layer and superimposed it to the Hu image; all ZENK⁺ cells were Hu⁺. Then we counted the rest of the Hu-labeled neurons in the area. The counts were taken from the center of the vocal nucleus in a given section. Because the cell densities near the rostral and caudal ends of a vocal nucleus can vary, the sections were taken from the middle of a vocal nucleus. For HVC, there has been reported a trend of less directed singing-driven mRNA levels in lateral part in comparison to the medial part (Jarvis et al., 1998). We did not find such a trend in the protein levels in control animals (r = 0.64 vs. 0.74, P = 0.89; multiple regression, n = 6), and thus for all other animals we measured the protein expression from the center of HVC, in frontal sections. For all nuclei, we counted at least two sections containing the vocal nucleus in each hemisphere. The variability in counts between adjacent sections was low; the standard deviations of the mean in HVC and LAreaX in intact birds were 3.5% and 4.4%, and in LMAN-lesioned birds they were 3.5% and 2.8%. The number of ZENK-labeled neurons was divided by the number of Hu-labeled neurons to obtain the percentage of ZENK-expressing neurons (% of ZENK expressing neurons = 100 × # of ZENK⁺ neurons/# of Hu⁺ neurons). Counting was done blind to group.

Song analysis

Song motif duration was measured in 20 undirected motifs before surgery and 20 undirected motifs from the last day of observation, and also at two different comparable time points for intact birds (n = 7 intact birds, 5 birds with LAreaX lesion, 5 birds with LMAN lesion, and 3 birds with LAreaX + LMAN lesion).

Statistics

We present the data of intact and lesioned animals in two complementary analyses: first as regression analysis of gene expression in each vocal nucleus with singing amount; and second as ratio analysis between the same nuclei of each brain hemisphere. The regressions allow for comparisons of each hemisphere relative to control animals but are sensitive to only relatively large differences between groups because they do not control for within-animal variability. The ratios control for within-animal variability, and as such are sensitive to subtle but significant changes in the expression levels between hemispheres of individual animals because each animal serves as its own control. The ratio analysis also normalizes for singing amount. The information lost in the ratio analysis is the direction (increase or decrease) of a change that occurs relative to intact controls.

For the regression analysis, we calculated P- and r-values from simple regressions when comparing expression of a single vocal nucleus and singing amount. We used multiple regressions when comparing undirected and directed singing and when comparing the effect of vocal nuclei lesions on gene expression in other vocal nuclei between lesioned and intact animals. For the ratio analysis, the number of ZENK-expressing neurons in a vocal nucleus of the lesioned side was divided by the number on the intact side, and statistical differences across groups were assessed by anova followed by the Fisher’s protected least significant difference (Fisher’s PLSD) post hoc test. The anova analyses were done separately for each nucleus and social context. The independent variable was lesion type and the dependent variable was percentage of ZENK-expressing neurons in the ipsilateral : contralateral vocal nucleus.

Singing-driven ZENK protein levels and social context

The social context-dependent differences in LAreaX, LMAN and RA are known for ZENK mRNA levels per cell (Jarvis et al., 1998), but not for ZENK protein (but see Castelino & Ball, 2005). Before performing experiments in this study, we first established whether the social context mRNA differences persist at the protein level in control animals without lesions. Consistent with the mRNA regulation (Jarvis et al., 1998), the number of neurons expressing ZENK protein in LAreaX, MAreaX, LMAN, MMAN, RA and HVC increased with the number of undirected songs produced (Fig. 3B, filled circles and solid lines). However, unlike the amount of mRNA/cell, which continues to increase linearly even after 120 songs in 45 min (Jarvis et al., 1998), the number of ZENK protein-expressing neurons began to show saturation with about 80 songs. The saturation was not at 100% of the neurons, but less, with the level of saturation depending on vocal nucleus (Fig. 3B). When a bird directed its singing to a female, the number of neurons expressing ZENK protein in MAreaX, MMAN and HVC also increased, while in LAreaX, LMAN and RA the numbers were low (Fig. 3B, open circles and dashed lines), which is consistent with the amount of mRNA/cell (Jarvis et al., 1998). We could not ascertain for directed singing whether saturation occurs in MAreaX, MMAN and HVC, as it was difficult to obtain birds that would sing more than 80 songs to a female in the time period (55 min) tested. The variations between the mRNA and protein results could be due to post-translational regulation of ZENK (Whitney et al., 2000; Whitney & Johnson, 2005) or differences in the relationships of singing amount and ‘expression/cell’ vs. ‘percentage of expressing neurons’; the latter by definition has to saturate below or at 100% of the neurons.

Furthermore, after undirected singing the ZENK protein expression levels in LAreaX, LMAN and RA were strongly correlated (graphs not shown; r = 0.85–0.94; P = 0.0004–0.007; simple regressions); after directed singing, no significant correlations were found with any of these three nuclei among each other or with other nuclei, due to little or no increases (r = 0.09–0.74, P = 0.06–0.83; simple regressions). The expression levels in MAreaX, MMAN and HVC were strongly correlated with each other in both singing contexts (r = 0.91–0.95, P = 0.0007–0.0014 after undirected singing; r = 0.95–0.98, P = 0.0001–0.0007 after directed singing; simple regressions), and with LAreaX, LMAN and RA only during undirected singing (r = 0.88–0.93, P = 0.0001–0.005; simple regressions). Thus, the social context-dependent differences persist at the protein level, where the lateral part of the vocal pallial basal ganglia pathway (LAreaX and LMAN) together with RA function in a coordinated manner, whereas the medial part (MAreaX and MMAN) together with HVC may function in a coordinated manner.

Lateral part of pallial basal ganglia pathway

Medial part of pallial basal ganglia pathway

Pallial basal ganglia pathway and social contexts

Given that the lateral part of the pallial basal ganglia pathway modulates singing-driven ZENK expression in RA and given that the lateral part is differentially active during undirected and directed singing, this leads to the prediction that the lateral part may modulate RA expression differentially in different social contexts. To test this idea, we lesioned the lateral basal ganglia pathway nuclei and assessed whether social context-dependent differences remained in RA and other nuclei using multiple regression analyses. We found that such lesions either eliminated or reduced the social context-dependent differences.

LAreaX + LMAN lesions and directed vs. undirected singing-driven gene expression

Combined lesions to ipsilateral LAreaX and LMAN also eliminated the social context difference in RA of both ipsilateral and contralateral hemispheres (P = 0.32 and 0.11, respectively; multiple regressions comparison within lesion groups; not shown). The difference between directed and undirected singing in RA was lost because relative to intact birds some LAreaX + LMAN-lesioned birds showed an increase in both contexts, similar to that seen with lesions only to LMAN.

Summarizing, the results above show that LAreaX and LMAN are required for social context-dependent gene expression in opposite directions for different contexts. LAreaX is required for ‘high’ expression in LMAN and RA when producing ‘undirected song’ (Fig. 11A, top middle panel). In a reciprocal manner, LMAN is required for ‘low’ expression in LAreaX when producing ‘directed song’ (Fig. 11A, bottom right panel). Because RA can still have some low expression during directed song after a LMAN lesion or combined LAreaX + LMAN lesion, an additional input must be responsible for the full low expression in RA during directed singing. We have not yet investigated the source of this additional input.

Behavior

There was no statistical difference in singing amount between unilateral lesion and intact birds either for undirected singing (P = 0.52, anova) or for directed singing (P = 0.37, anova). Although we performed unilateral and not bilateral lesions, we carried out a measurement of song tempo that has been previously shown to change after bilateral lesions to LMAN (Williams & Mehta, 1999). Our analysis showed that there are only subtle changes towards slower singing in unilateral LAreaX-lesioned birds and no changes for unilateral LMAN or LAreaX + LMAN-lesioned birds (anova, P = 0.06; Fisher’s PLSD is P = 0.01 for LAreaX lesion, P = 0.16 for LMAN lesion, and P = 0.12 for LAreaX + LMAN lesion; mean percentage change ± SEM was 2.14 ± 0.7% for LAreaX lesion, 0.9 ± 0.9% for LMAN lesion and 1.3 ± 0.5% for LAreaX + LMAN lesion).

Medial and lateral subdivisions of vocal basal ganglia pathway

Traditionally, the role of the vocal pallial basal ganglia pathway is said to process information from HVC and then send output to RA (Fig. 11A; Williams, 1989; Nottebohm, 1993). The instructive signal has been proposed to change RA activity to correct the song produced (Brainard & Doupe, 2000). We argue that this is only part of the story, the lateral part, which has been the inadvertent focus of most studies on this pathway in songbirds. Our results suggest that information from HVC is also processed in the medial part of the vocal pallial basal ganglia pathway, which then sends feedback to HVC (Fig. 11B). In this view, the main role of the vocal pallial basal ganglia pathway may be to process motor pathway input from HVC and use it to modulate activity of both major nuclei, HVC and RA, of the motor pathway. For RA, this modulation could be for learning or maintaining syllable acoustic structure, and for HVC, learning or maintaining syllable sequences (Hahnloser et al., 2002).

Our lesion results indicate that the lateral part of the vocal pallial basal ganglia pathway has stronger unilateral control over RA than the medial part does over HVC. This could be because of either a larger lateral pathway volume or unilateral vs. bilateral connectivity. The lateral part of the pallial basal ganglia pathway and RA has mainly unilateral connections (Fig. 1). In contrast, the medial part and HVC have bilateral feedback connections: DMP to MMAN, and nucleus uvaeformis (Uva) to NIf and HVC (Fig. 1). Thus, when MAreaX is lesioned, ipsilateral MMAN and HVC could still be activated and compensated by input from the contralateral side.

Our results also suggest that there is an interaction between the medial and lateral components. Birds with LAreaX lesions had an effect on the singing-driven gene expression in HVC. This might be due to two reasons. First, the downstream series of connections from LAreaX to DLM to LMAN to RA to dorsal medial nucleus of the midbrain (DM) to Uva to NIf to HVC or from RA to DMP to MMAN to HVC (Fig. 1) could influence HVC (Akutagawa & Konishi, 2005). A second possibility is that LAreaX and MAreaX are interconnected. The lesion to LAreaX influenced the expression in MAreaX, which then via a loop with MMAN could have affected the expression in HVC. This interconnection is supported by the fact that numerous striatal spiny neurons terminate on other striatal neurons in AreaX (Reiner et al., 2004a), and lesions in one part of LAreaX affected ZENK expression throughout other parts of AreaX. However, because MAreaX lesions did not affect ZENK expression in LAreaX, the interconnections may not be reciprocal (and we did not note projections from MAreaX to LAreaX in our tracer injections). Alternatively, LAreaX is much larger than MAreaX and thus LAreaX may have enough redundancy to withstand damage of input from MAreaX.

Opposing roles for striatal vs. pallial nuclei

Our data show that relative to intact animals, LAreaX lesions lead to a lower ZENK expression in downstream nuclei and LMAN lesions lead to the same or higher ZENK expression in downstream nuclei. Given the later bimodal effect, we were still concerned whether we can conclude that LMAN lesions can have an effect of increased ZENK expression in RA. Our concern, though, is assuaged by the facts that first there are clearly much higher levels of visible ZENK expression in RA of the majority of LMAN-lesioned birds at levels we have never seen before in intact birds; second there is a similar result with combined LAreaX + LMAN lesions; third there is a significant difference in the RA expression in LMAN-lesioned vs. LAreaX-lesioned groups; and fourth a large increase in RA ZENK expression has been previously found following LMAN lesions in juvenile birds (Whitney et al., 2000). A parsimonious explanation of these results is that LAreaX enhances while LMAN permits or dampens (possibly by suppression) the expression in their targets, the two working in concert to balance each other. The enhancing effect that LAreaX has on RA expression during undirected singing appears to go through LMAN (and thus also through DLM), where LAreaX influences the effect that LMAN has on RA. That is, in intact animals activation of LAreaX may induce (via DLM) the high activation levels in LMAN, which in turn permits HVC (or other upstream nuclei) to induce high levels in RA up to a certain amount (Fig. 11A, top left panel); a role for HVC activation of RA is supported by electrophysiology studies of Hahnloser et al. (2002) and preliminary gene expression findings of ours (where unilateral HVC lesions prevented singing-driven ZENK activation in ipsilateral RA). When LAreaX is removed and LMAN connections to RA are intact, LMAN may no longer be activated or is altered in a manner that does not permit HVC to induce ZENK expression in RA (Fig. 11A, top middle panel). But, when LMAN is removed, HVC can then induce high levels in RA (and in LAreaX) without the need for permissive signals from LMAN (Fig. 11A, top right panel). A similar activation scenario can be made for directed singing (Fig. 11A, bottom panels), with the exception that other as yet to be determined input to RA still does not permit HVC to fully activate ZENK expression in RA when LMAN is lesioned. This other input to RA might be dopaminergic or noradrenergic innervation from the midbrain (Appeltants et al., 2002). A recent test of noradrenergic involvement for LAreaX showed that lesions to noradrenergic neurons result in abnormally high levels of directed singing-driven ZENK expression in LAreaX, but still relatively low levels in RA and LMAN (Castelino & Ball, 2005), similar to what we find in RA and in LAreaX following directed singing in LMAN-lesioned birds. Thus, like LMAN input, noradrenergic input to LAreaX is required for the low expression levels during directed singing. Consistent with the idea of opposing effects that LMAN and LAreaX (via LMAN) have on downstream nuclei is the prior behavior findings that LAreaX is required for song stereotypy and LMAN is required for song variability (Scharff & Nottebohm, 1991; Kao et al., 2005; Olveczky et al., 2005).

As for the role the medial nuclei MAreaX and MMAN have on the activation of each other and on HVC, the changes following lesions were too small for regression analyses to reveal the direction of change, enhancement or suppression. However, the results of the ratio analyses of MAreaX-lesioned birds with lower ipsilateral than contralateral expression in MMAN and HVC suggest similar effects as its lateral counterpart LAreaX has on LMAN and RA. Further, in parallel with the lateral counterpart, the change in HVC requires MMAN input, MMAN lesions alone do not cause a lowering of expression in HVC, and there is a significant difference in the HVC expression in MMAN-lesioned vs. MAreaX-lesioned groups. To reveal the full effect of medial pathway lesions on HVC expression may require bilateral lesions in comparison with intact controls to prevent possible contralateral compensation.

Proposed synaptic mechanisms

It has been hypothesized that presynaptic neurons regulate postsynaptic IEG expression, where neurotransmitters released by presynaptic neural activity onto postsynaptic receptors induce the motor-driven expression of postsynaptic IEGs during behavior (Lerea, 1997; Sgambato et al., 1999; Jarvis, 2004a). Our results are consistent with this hypothesis, as lesions to the presynaptic nuclei disrupt normal postsynaptic IEG induction. As for RA, there are two main inputs: HVC (Fig. 11A), which releases glutamate onto AMPA and N-methyl-d-aspartate (NMDA) receptors of RA neurons; and LMAN (Fig. 11A), which releases glutamate mainly onto NMDA receptors of the same RA neurons (Mooney & Konishi, 1991; Stark & Perkel, 1999). When bound to glutamate, NMDA receptors activate ZENK transcription (Lerea, 1997). But if LMAN input to RA (and presumably also to LAreaX) is glutamatergic, i.e. considered excitatory (Perkel, 2004), how could LMAN suppress ZENK expression in the connected nuclei during singing? One possible explanation is that LMAN activates in RA and in LAreaX not only the excitatory NMDA receptors but also some inhibitory glutamate receptors. Potential candidates are metabotropic glutamate group II receptors, such as mGluR4, which is rich in both RA and AreaX (Wada et al., 2004). Differential activation of NMDA and mGluR4 receptor types in different social context could then possibly allow high and low ZENK expression during undirected and directed singing, respectively. A second possibility is that LMAN preferentially activates RA inhibitory neurons and these inhibit ZENK activation in RA’s tonically active projection neurons. This feedforward inhibition of neural firing from LMAN to RA inhibitory neurons that synapse onto excitatory neurons has been shown in electrophysiological studies (Spiro et al., 1999). A similar argument can be made for LMAN input to AreaX and the ZENK activation in its striatal neurons. Detailed cellular analysis will be required to find out which hypothesis is correct.

As for the contralateral changes that we found, perhaps some compensation occurs in the contralateral hemisphere, but we do not suppose that the expression changes are simply due to compensation, because we would expect to see an opposite pattern of expression in the contralateral hemisphere. For example, compensation for the high expression in LAreaX after directed singing and unilateral LMAN lesions would be a decreased expression in the contralateral LAreaX and vice versa for the LAreaX-lesioned birds. Instead, in LAreaX-lesioned birds we see bilateral decreases in downstream nuclei, and in LMAN-lesioned birds we can see bilateral increases in downstream nuclei. These contralateral effects presumably occur via communication through the anterior commissure and/or bilateral thalamic and midbrain vocal nuclei connections (Fig. 1). Birds do not have a corpus callosum (fibers connecting the two hemispheres), as mammals. In mammals, although social context and behavior were not evaluated, unilateral blocking of dopamine D1 receptors in the striatum causes bilateral reduction of basal egr-1 (ZENK) levels in the cortex (Steiner & Kitai, 2000), with less reduction on the contralateral side, similar to our findings with striatal lesions in songbirds. Thus, even without a corpus callosum, bilateral control of pallial/cortical gene activation by the basal ganglia pathway could be a general feature of vertebrate brain function.

Possible behavioral consequences

The modulation of the vocal motor pathway by the vocal pallial basal ganglia pathway could potentially lead to immediate and long-term changes in vocal behavior output. The absence of any effect on adult singing following LMAN and LAreaX lesions (Bottjer et al., 1984; Sohrabji et al., 1990; Scharff & Nottebohm, 1991) is being revised in the literature. An immediate effect on adult vocal output was shown in bilateral lesion studies where MMAN is necessary for consistent stereotyped sequencing of syllables (Foster & Bottjer, 2001), LMAN is necessary for generation of variability in syllable structure (Kao et al., 2005; Kao & Brainard, 2006), and LAreaX is necessary for producing repeated syllables without stuttering (Kobayashi et al., 2001). However, the immediate behavioral changes are presumably not the result of ZENK activation, as ZENK is induced after the behavior is produced (Jarvis & Nottebohm, 1997). ZENK can have long-term effects; gene manipulation studies in the mammalian hippocampus have shown a functional role of ZENK on memory reconsolidation, but not on acquisition (Lee et al., 2004). If a similar functional role exists for other neural circuits, then high activation of ZENK throughout the vocal pathways during undirected singing may be useful for reconsolidation of motor networks for the songs produced. In this context, perhaps for RA and HVC, AreaX is required for higher ZENK expression and higher reconsolidation, while MAN is required for lower ZENK expression and thus more plasticity. During directed singing, however, the lateral part of the basal ganglia pathway may be utilized less or inhibited and thus LMAN may not have such control of RA gene expression, but reconsolidation is still needed for HVC. In this context, the behavioral performance to the listening animal may be more crucial than memory reconsolidation.

1. Akutagawa E, Konishi M. Connections of thalamic modulatory centers to the vocal control system of the zebra finch. Proc. Natl Acad. Sci. USA. 2005;102:14086–14091. doi: 10.1073/pnas.0506774102. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Appeltants D, Ball GF, Balthazart J. The origin of catecholaminergic inputs to the song control nucleus RA in canaries. Neuroreport. 2002;13:649–653. doi: 10.1097/00001756-200204160-00023. [DOI] [PubMed] [Google Scholar]
3. Barami K, Iversen K, Furneaux H, Goldman SA. Hu protein as an early marker of neuronal phenotypic differentiation by subependymal zone cells of the adult songbird forebrain. J. Neurobiol. 1995;28:82–101. doi: 10.1002/neu.480280108. [DOI] [PubMed] [Google Scholar]
4. Berretta S, Sachs Z, Graybiel AM. Cortically driven Fos induction in the striatum is amplified by local dopamine D2-class receptor blockade. Eur. J. Neurosci. 1999;11:4309–4319. doi: 10.1046/j.1460-9568.1999.00866.x. [DOI] [PubMed] [Google Scholar]
5. Blandini F, Fancellu R, Orzi F, Conti G, Greco R, Tassorelli C, Nappi G. Selective stimulation of striatal dopamine receptors of the D1- or D2-class causes opposite changes of fos expression in the rat cerebral cortex. Eur. J. Neurosci. 2003;17:763–770. doi: 10.1046/j.1460-9568.2003.02520.x. [DOI] [PubMed] [Google Scholar]
6. Bottjer SW, Halsema KA, Brown SA, Miesner EA. Axonal connections of a forebrain nucleus involved with vocal learning in zebra finches. J. Comp. Neurol. 1989;279:312–326. doi: 10.1002/cne.902790211. [DOI] [PubMed] [Google Scholar]
7. Bottjer SW, Miesner EA, Arnold AP. Forebrain lesions disrupt development but not maintenance of song in passerine birds. Science. 1984;224:901–903. doi: 10.1126/science.6719123. [DOI] [PubMed] [Google Scholar]
8. Brainard MS, Doupe AJ. Auditory feedback in learning and maintenance of vocal behaviour. Nat. Rev. Neurosci. 2000;1:31–40. doi: 10.1038/35036205. [DOI] [PubMed] [Google Scholar]
9. Castelino CB, Ball GF. A role for norepinephrine in the regulation of context-dependent ZENK expression in male zebra finches (Taeniopygia guttata) Eur. J. Neurosci. 2005;21:1962–1972. doi: 10.1111/j.1460-9568.2005.04028.x. [DOI] [PubMed] [Google Scholar]
10. Dudek SM, Fields RD. Somatic action potentials are sufficient for late-phase LTP-related cell signaling. Proc. Natl Acad. Sci. USA. 2002;99:3962–3967. doi: 10.1073/pnas.062510599. [DOI] [PMC free article] [PubMed] [Google Scholar]
11. Farries MA, Perkel D. Pallidum-like elements in the avian ‘striatum’ outside of specialized vocal structures project directly to the thalamus. Soc. Neurosci. Abstr. 2002:680.618. [Google Scholar]
12. Fields RD, Eshete F, Dudek S, Ozsarac N, Stevens B. Regulation of gene expression by action potentials: dependence on complexity in cellular information processing. Novartis Found. Symp. 2001;239:160–172. doi: 10.1002/0470846674.ch13. Discussion, 172–166, 234–140. [DOI] [PubMed] [Google Scholar]
13. Foster EF, Bottjer SW. Lesions of a telencephalic nucleus in male zebra finches: influences on vocal behavior in juveniles and adults. J. Neurobiol. 2001;46:142–165. doi: 10.1002/1097-4695(20010205)46:2<142::aid-neu60>3.0.co;2-r. [DOI] [PubMed] [Google Scholar]
14. Foster EF, Mehta RP, Bottjer SW. Axonal connections of the medial magnocellular nucleus of the anterior neostriatum in zebra finches. J. Comp. Neurol. 1997;382:364–381. doi: 10.1002/cne.903820305. [DOI] [PubMed] [Google Scholar]
15. Hahnloser RHR, Kozhevnikov AA, Fee MS. An ultra-sparse code underlies the generation of neural sequences in a songbird. Nature. 2002;419:65–70. doi: 10.1038/nature00974. [DOI] [PubMed] [Google Scholar]
16. Halim ND, Swerdlow NR. Distributed neurodegenerative changes 2–28 days after ventral hippocampal excitotoxic lesions in rats. Brain Res. 2000;873:60–74. doi: 10.1016/s0006-8993(00)02497-5. [DOI] [PubMed] [Google Scholar]
17. Hessler NA, Doupe AJ. Singing-related neural activity in a dorsal forebrain-basal ganglia circuit of adult zebra finches. J. Neurosci. 1999a;19:10461–10481. doi: 10.1523/JNEUROSCI.19-23-10461.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
18. Hessler NA, Doupe AJ. Social context modulates singing-related neural activity in the songbird forebrain. Nat. Neurosci. 1999b;2:209–211. doi: 10.1038/6306. [DOI] [PubMed] [Google Scholar]
19. Iyengar S, Viswanathan SS, Bottjer SW. Development of topography within song control circuitry of zebra finches during the sensitive period for song learning. J. Neurosci. 1999;19:6037–6057. doi: 10.1523/JNEUROSCI.19-14-06037.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Jarvis ED. Brains and birdsong. In: Marler P, Slabbekoorn H, editors. Nature’s Music: The Science of BirdSong. San Diego: Elsevier; 2004a. pp. 226–271. [Google Scholar]
21. Jarvis ED. Learned birdsong and the neurobiology of human language. Ann. N Y Acad. Sci. 2004b;1016:749–777. doi: 10.1196/annals.1298.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Jarvis ED, Nottebohm F. Motor-driven gene expression. Proc. Natl Acad. Sci. USA. 1997;94:4097–4102. doi: 10.1073/pnas.94.8.4097. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Jarvis ED, Scharff C, Grossman MR, Ramos JA, Nottebohm F. For whom the bird sings: context-dependent gene expression. Neuron. 1998;21:775–788. doi: 10.1016/s0896-6273(00)80594-2. [DOI] [PubMed] [Google Scholar]
24. Johnson F, Sablan MM, Bottjer SW. Topographic organization of a forebrain pathway involved with vocal learning in zebra finches. J. Comp. Neurol. 1995;358:260–278. doi: 10.1002/cne.903580208. [DOI] [PubMed] [Google Scholar]
25. Kao MH, Brainard MS. Lesions of an avian basal ganglia circuit prevent context-dependent changes to song variability. J. Neurophysiol. 2006;96:1441–1455. doi: 10.1152/jn.01138.2005. [DOI] [PubMed] [Google Scholar]
26. Kao MH, Doupe AJ, Brainard MS. Contributions of an avian basal ganglia-forebrain circuit to real-time modulation of song. Nature. 2005;433:638–643. doi: 10.1038/nature03127. [DOI] [PubMed] [Google Scholar]
27. Keefe KA, Gerfen CR. Local infusion of the (+/−)-alpha-amino-3-hydroxy-5-methylisoxazole-4-propionate/kainate receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione does not block D1 dopamine receptor-mediated increases in immediate early gene expression in the dopamine-depleted striatum. Neuroscience. 1999;89:491–504. doi: 10.1016/s0306-4522(98)00327-3. [DOI] [PubMed] [Google Scholar]
28. Kobayashi K, Uno H, Okanoya K. Partial lesions in the anterior forebrain pathway affect song production in adult Bengalese finches. Neuroreport. 2001;12:353–358. doi: 10.1097/00001756-200102120-00034. [DOI] [PubMed] [Google Scholar]
29. Lee JL, Everitt BJ, Thomas KL. Independent cellular processes for hippocampal memory consolidation and reconsolidation. Science. 2004;304:839–843. doi: 10.1126/science.1095760. [DOI] [PubMed] [Google Scholar]
30. Lerea L. Glutamate receptors and gene induction: signalling from receptor to nucleus. Cell. Signal. 1997;9:219–226. doi: 10.1016/s0898-6568(96)00134-9. [DOI] [PubMed] [Google Scholar]
31. Liu WC, Nottebohm F. Variable rate of singing and variable song duration are associated with high immediate early gene expression in two anterior forebrain song nuclei. Proc. Natl Acad. Sci. USA. 2005;102:10724–10729. doi: 10.1073/pnas.0504677102. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Luo M, Ding L, Perkel DJ. An avian basal ganglia pathway essential for vocal learning forms a closed topographic loop. J. Neurosci. 2001;21:6836–6845. doi: 10.1523/JNEUROSCI.21-17-06836.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Mello CV, Ribeiro S. ZENK protein regulation by song in the brain of songbirds. J. Comp. Neurol. 1998;393:426–438. doi: 10.1002/(sici)1096-9861(19980420)393:4<426::aid-cne3>3.0.co;2-2. [DOI] [PubMed] [Google Scholar]
34. Mello CV, Vicario DS, Clayton DF. Song presentation induces gene expression in the songbird forebrain. Proc. Natl Acad. Sci. USA. 1992;89:6818–6822. doi: 10.1073/pnas.89.15.6818. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Milner TA, Veznedaroglu E. Septal efferent axon terminals identified by anterograde degeneration show multiple sites for modulation of neuropeptide Y-containing neurons in the rat dentate gyrus. Synapse. 1993;14:101–112. doi: 10.1002/syn.890140202. [DOI] [PubMed] [Google Scholar]
36. Mooney R, Konishi M. Two distinct inputs to an avian song nucleus activate different glutamate receptor subtypes on individual neurons. Proc. Natl Acad. Sci. USA. 1991;88:4075–4079. doi: 10.1073/pnas.88.10.4075. [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Nordeen KW, Nordeen EJ. Long-term maintenance of song in adult zebra finches is not affected by lesions of a forebrain region involved in song learning. Behav. Neural Biol. 1993;59:79–82. doi: 10.1016/0163-1047(93)91215-9. [DOI] [PubMed] [Google Scholar]
38. Nottebohm F. The search for neural mechanisms that define the sensitive period for song learning in birds. Neth. J. Zool. 1993;43:193–234. [Google Scholar]
39. Nottebohm F, Kelley DB, Paton JA. Connections of vocal control nuclei in the canary telencephalon. J. Comp. Neurol. 1982;207:344–357. doi: 10.1002/cne.902070406. [DOI] [PubMed] [Google Scholar]
40. Nottebohm F, Stokes TM, Leonard CM. Central control of song in the canary, Serinus canarius. J. Comp. Neurol. 1976;165:457–486. doi: 10.1002/cne.901650405. [DOI] [PubMed] [Google Scholar]
41. Olveczky BP, Andalman AS, Fee MS. Vocal experimentation in the juvenile songbird requires a basal ganglia circuit. PLoS Biol. 2005;3:e153. doi: 10.1371/journal.pbio.0030153. [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Parthasarathy HB, Graybiel AM. Cortically driven immediate-early gene expression reflects modular influence of sensorimotor cortex on identified striatal neurons in the squirrel monkey. J. Neurosci. 1997;17:2477–2491. doi: 10.1523/JNEUROSCI.17-07-02477.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Perkel DJ. Origin of the anterior forebrain pathway. Ann. N Y Acad. Sci. 2004;1016:736–748. doi: 10.1196/annals.1298.039. [DOI] [PubMed] [Google Scholar]
44. Perkel DJ, Farries MA, Luo M, Ding L. Electrophysiological analysis of a songbird basal ganglia circuit essential for vocal plasticity. Brain Res. Bull. 2002;57:529–532. doi: 10.1016/s0361-9230(01)00690-6. [DOI] [PubMed] [Google Scholar]
45. Reiner A, Carraway RE. Immunohistochemical and biochemical studies on Lys8-Asn9-neurotensin8-13 (LANT6)-related peptides in the basal ganglia of pigeons, turtles, and hamsters. J. Comp. Neurol. 1987;257:453–476. doi: 10.1002/cne.902570312. [DOI] [PubMed] [Google Scholar]
46. Reiner A, Laverghetta AV, Meade CA, Cuthbertson SL, Bottjer SW. An immunohistochemical and pathway tracing study of the striatopallidal organization of area X in the male zebra finch. J. Comp. Neurol. 2004a;469:239–261. doi: 10.1002/cne.11012. [DOI] [PubMed] [Google Scholar]
47. Reiner A, Perkel DJ, Bruce LL, Butler AB, Csillag A, Kuenzel W, Medina L, Paxinos G, Shimizu T, Striedter G, Wild M, Ball GF, Durand S, Guturkun O, Lee DW, Mello CV, Powers A, White SA, Hough G, Kubikova L, Smulders TV, Wada K, Dugas-Ford J, Husband S, Yamamoto K, Yu J, Siang C, Jarvis ED. Revised nomenclature for avian telencephalon and some related brainstem nuclei. J. Comp. Neurol. 2004b;473:377–414. doi: 10.1002/cne.20118. [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Scharff C, Nottebohm F. A comparative study of the behavioral deficits following lesions of various parts of the zebra finch song system: implications for vocal learning. J. Neurosci. 1991;11:2896–2913. doi: 10.1523/JNEUROSCI.11-09-02896.1991. [DOI] [PMC free article] [PubMed] [Google Scholar]
49. Sgambato V, Maurice N, Besson M, Thierry A, Deniau J. Effect of a functional impairment of corticostriatal transmission on cortically evoked expression of c-Fos and zif 268 in the rat basal ganglia. Neuroscience. 1999;93:1313–1321. doi: 10.1016/s0306-4522(99)00267-5. [DOI] [PubMed] [Google Scholar]
50. Sohrabji F, Nordeen EJ, Nordeen KW. Selective impairment of song learning following lesions of a forebrain nucleus in the juvenile zebra finch. Behav. Neural Biol. 1990;53:51–63. doi: 10.1016/0163-1047(90)90797-a. [DOI] [PubMed] [Google Scholar]
51. Spiro JE, Dalva MB, Mooney R. Long-range inhibition within the zebra finch song nucleus RA can coordinate the firing of multiple projection neurons. J. Neurophysiol. 1999;81:3007–3020. doi: 10.1152/jn.1999.81.6.3007. [DOI] [PubMed] [Google Scholar]
52. Stark L, Perkel D. Two-stage, input-specific synaptic maturation in a nucleus essential for vocal production in the zebra finch. J. Neurosci. 1999;19:9107–9116. doi: 10.1523/JNEUROSCI.19-20-09107.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
53. Steiner H, Kitai ST. Regulation of rat cortex function by D1 dopamine receptors in the striatum. J. Neurosci. 2000;20:5449–5460. doi: 10.1523/JNEUROSCI.20-14-05449.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
54. Striedter GF, Vu ET. Bilateral feedback projections to the forebrain in the premotor network for singing in zebra finches. J. Neurobiol. 1998;34:27–40. [PubMed] [Google Scholar]
55. Tchernichovski O, Lints TJ, Deregnaucourt S, Cimenser A, Mitra PP. Studying the song development process: rationale and methods. Ann. N Y Acad. Sci. 2004;1016:348–363. doi: 10.1196/annals.1298.031. [DOI] [PubMed] [Google Scholar]
56. Vates GE, Nottebohm F. Feedback circuitry within a song-learning pathway. Proc. Natl Acad. Sci. USA. 1995;92:5139–5143. doi: 10.1073/pnas.92.11.5139. [DOI] [PMC free article] [PubMed] [Google Scholar]
57. Vates GE, Vicario DS, Nottebohm F. Reafferent thalamo-‘cortical’ loops in the song system of oscine songbirds. J. Comp. Neurol. 1997;380:275–290. [PubMed] [Google Scholar]
58. Wada K, Sakaguchi H, Jarvis ED, Hagiwara M. Differential expression of glutamate receptors in avian neural pathways for learned vocalization. J. Comp. Neurol. 2004;476:44–64. doi: 10.1002/cne.20201. [DOI] [PMC free article] [PubMed] [Google Scholar]
59. Whitney O, Johnson F. Motor-induced transcription but sensory-regulated translation of ZENK in socially interactive songbirds. J. Neurobiol. 2005;65:251–259. doi: 10.1002/neu.20187. [DOI] [PMC free article] [PubMed] [Google Scholar]
60. Whitney O, Soderstrom K, Johnson F. Post-transcriptional regulation of zenk expression associated with zebra finch vocal development. Brain Res. Mol. Brain Res. 2000;80:279–290. doi: 10.1016/s0169-328x(00)00178-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
61. Wild JM. Descending projections of the songbird nucleus robustus archistriatalis. J. Comp. Neurol. 1993;338:225–241. doi: 10.1002/cne.903380207. [DOI] [PubMed] [Google Scholar]
62. Williams H. Multiple representations and auditory–motor interactions in the avian song system. Ann. N.Y. Acad. Sci. 1989;563:148–164. [Google Scholar]
63. Williams H, Mehta N. Changes in adult zebra finch song require a forebrain nucleus that is not necessary for song production. J. Neurobiol. 1999;39:14–28. [PubMed] [Google Scholar]

Supplementary Materials