Development of topography within song control circuitry of zebra finches during the sensitive period for song learning

Songbird mesostriatal dopamine pathways are spatially segregated before the onset of vocal learning

Diverse dopamine (DA) pathways send distinct reinforcement signals to different striatal regions. In adult songbirds, a DA pathway from the ventral tegmental area (VTA) to Area X, the striatal nucleus of the song system, carries singing-related performance error signals important for learning. Meanwhile, a parallel DA pathway to a medial striatal area (MST) arises from a distinct group of neighboring DA neurons that lack connectivity to song circuits and do not encode song error. To test if the structural and functional segregation of these two pathways depends on singing experience, we carried out anatomical studies early in development before the onset of song learning. We find that distinct VTA neurons project to either Area X or MST in juvenile birds before the onset of substantial vocal practice. Quantitative comparisons of early juveniles (30-35 days post hatch), late juveniles (60-65 dph), and adult (>90 dph) brains revealed an outsized expansion of Area X-projecting neurons relative to MST-projecting neurons in VTA over development. These results show that a mesostriatal DA system dedicated to social communication can exist and be spatially segregated before the onset of vocal practice and associated sensorimotor experience.

Dopaminergic error signals retune to social feedback during courtship

Hunger, thirst, loneliness and ambition determine the reward value of food, water, social interaction and performance outcome1. Dopamine neurons respond to rewards meeting these diverse needs2-8, but it remains unclear how behaviour and dopamine signals change as priorities change with new opportunities in the environment. One possibility is that dopamine signals for distinct drives are routed to distinct dopamine pathways9,10. Another possibility is that dopamine signals in a given pathway are dynamically tuned to rewards set by the current priority. Here we used electrophysiology and fibre photometry to test how dopamine signals associated with quenching thirst, singing a good song and courting a mate change as male zebra finches (Taeniopygia guttata) were provided with opportunities to retrieve water, evaluate song performance or court a female. When alone, water reward signals were observed in two mesostriatal pathways but singing-related performance error signals were routed to Area X, a striatal nucleus specialized for singing. When courting a female, water seeking was reduced and dopamine responses to both water and song performance outcomes diminished. Instead, dopamine signals in Area X were driven by female calls timed with the courtship song. Thus the dopamine system handled coexisting drives by routing vocal performance and social feedback signals to a striatal area for communication and by flexibly re-tuning to rewards set by the prioritized drive.

Responses to Song Playback Differ in Sleeping versus Anesthetized Songbirds

Vocal learning in songbirds is mediated by a highly localized system of interconnected forebrain regions, including recurrent loops that traverse the cortex, basal ganglia, and thalamus. This brain-behavior system provides a powerful model for elucidating mechanisms of vocal learning, with implications for learning speech in human infants, as well as for advancing our understanding of skill learning in general. A long history of experiments in this area has tested neural responses to playback of different song stimuli in anesthetized birds at different stages of vocal development. These studies have demonstrated selectivity for different song types that provide neural signatures of learning. In contrast to the ease of obtaining responses to song playback in anesthetized birds, song-evoked responses in awake birds are greatly reduced or absent, indicating that behavioral state is an important determinant of neural responsivity. Song-evoked responses can be elicited during sleep as well as anesthesia, and the selectivity of responses to song playback in adult birds is highly similar between anesthetized and sleeping states, encouraging the idea that anesthesia and sleep are similar. In contrast to that idea, we report evidence that cortical responses to song playback in juvenile zebra finches (Taeniopygia guttata) differ greatly between sleep and urethane anesthesia. This finding indicates that behavioral states differ in sleep versus anesthesia and raises questions about relationships between developmental changes in sleep activity, selectivity for different song types, and the neural substrate for vocal learning.

Developmentally regulated pathways for motor skill learning in songbirds

Vocal learning in songbirds is mediated by cortico-basal ganglia circuits that govern diverse functions during different stages of development. We investigated developmental changes in axonal projections to and from motor cortical regions that underlie learned vocal behavior in juvenile zebra finches (Taeniopygia guttata). Neurons in LMAN-core project to RA, a motor cortical region that drives vocal output; these RA-projecting neurons send a transient collateral projection to AId, a region adjacent to RA, during early vocal development. Both RA and AId project to a region of dorsal thalamus (DLM), which forms a feedback pathway to cortico-basal ganglia circuitry. These projections provide pathways conveying efference copy and a means by which information about vocal motor output could be reintegrated into cortico-basal ganglia circuitry, potentially aiding in the refinement of juvenile vocalizations during learning. We used tract-tracing techniques to label the projections of LMAN-core to AId and of RA to DLM in juvenile songbirds. The volume and density of terminal label in the LMAN-core→AId projection declined substantially during early stages of sensorimotor learning. In contrast, the RA→DLM projection showed no developmental change. The retraction of LMAN-core→AId axon collaterals indicates a loss of efference copy to AId and suggests that projections that are present only during early stages of sensorimotor learning mediate unique, temporally restricted processes of goal-directed learning. Conversely, the persistence of the RA→DLM projection may serve to convey motor information forward to the thalamus to facilitate song production during both learning and maintenance of vocalizations.

Computational benefits of structural plasticity, illustrated in songbirds

The plasticity of nervous systems allows animals to quickly adapt to a changing environment. In particular, the structural plasticity of brain networks is often critical to the development of the central nervous system and the acquisition of complex behaviors. As an example, structural plasticity is central to the development of song-related brain circuits and may be critical for song acquisition in juvenile songbirds. Here, we review current evidences for structural plasticity and their significance from a computational point of view. We start by reviewing evidence for structural plasticity across species and categorizing them along the spatial axes as well as the along the time course during development. We introduce the vocal learning circuitry in zebra finches, as a useful example of structural plasticity, and use this specific case to explore the possible contributions of structural plasticity to computational models. Finally, we discuss current modeling studies incorporating structural plasticity and unexplored questions which are raised by such models.

The expression of delta opioid receptor mRNA in adult male zebra finches (Taenopygia guttata)

The endogenous opioid system is evolutionarily conserved across reptiles, birds and mammals and is known to modulate varied brain functions such as learning, memory, cognition and reward. To date, most of the behavioral and anatomical studies in songbirds have mainly focused on μ-opioid receptors (ORs). Expression patterns of δ-ORs in zebra finches, a well-studied species of songbird have not yet been reported, possibly due to the high sequence similarity amongst different opioid receptors. In the present study, a specific riboprobe against the δ-OR mRNA was used to perform fluorescence in situ hybridization (FISH) on sections from the male zebra finch brain. We found that δ-OR mRNA was expressed in different parts of the pallium, basal ganglia, cerebellum and the hippocampus. Amongst the song control and auditory nuclei, HVC (abbreviation used as a formal name) and NIf (nucleus interfacialis nidopallii) strongly express δ-OR mRNA and stand out from the surrounding nidopallium. Whereas the expression of δ-OR mRNA is moderate in LMAN (lateral magnocellular nucleus of the anterior nidopallium), it is low in the MSt (medial striatum), Area X, DLM (dorsolateral nucleus of the medial thalamus), RA (robust nucleus of the arcopallium) of the song control circuit and Field L, Ov (nucleus ovoidalis) and MLd (nucleus mesencephalicus lateralis, pars dorsalis) of the auditory pathway. Our results suggest that δ-ORs may be involved in modulating singing, song learning as well as spatial learning in zebra finches.

Song learning and plasticity in songbirds

Birdsong provides a fascinating system to study both behavioral and neural plasticity. Oscine songbirds learn to sing, exhibiting behavioral plasticity both during and after the song-learning process. As a bird learns, its song progresses from a plastic and highly variable vocalization into a more stereotyped, crystallized song. However, even after crystallization, song plasticity can occur: some species' songs become more stereotyped over time, whereas other species can incorporate new song elements. Alongside the changes in song, songbirds' brains are also plastic. Both song and neural connections change with the seasons in many species, and new neurons can be added to the song system throughout life. In this review, we highlight important research on behavioral and neural plasticity at multiple timescales, from song development in juveniles to lifelong modifications of learned song.

Molecular specializations of deep cortical layer analogs in songbirds

How the evolution of complex behavioral traits is associated with the emergence of novel brain pathways is largely unknown. Songbirds, like humans, learn vocalizations via tutor imitation and possess a specialized brain circuitry to support this behavior. In a comprehensive in situ hybridization effort, we show that the zebra finch vocal robust nucleus of the arcopallium (RA) shares numerous markers (e.g. SNCA, PVALB) with the adjacent dorsal intermediate arcopallium (AId), an avian analog of mammalian deep cortical layers with involvement in motor function. We also identify markers truly unique to RA and thus likely linked to modulation of vocal motor function (e.g. KCNC1, GABRE), including a subset of the known shared markers between RA and human laryngeal motor cortex (e.g. SLIT1, RTN4R, LINGO1, PLXNC1). The data provide novel insights into molecular features unique to vocal learning circuits, and lend support for the motor theory for vocal learning origin.

The expression of DARPP-32 in adult male zebra finches (Taenopygia guttata)

Although the catecholaminergic circuitry in the zebra finch brain has been well studied, there is little information regarding the postsynaptic targets of dopamine. To answer this question, we looked at overall patterns of immunoreactivity for DARPP-32 (a dopamine and cAMP-regulated phosphoprotein, present mostly in dopaminoceptive neurons) in adult male zebra finches. Our results demonstrated that as in mammals and other avian species, DARPP-32 expression was highest in both medial and lateral striatum. Interestingly, a specific pattern of immunoreactivity was observed in the song control system, with 'core' song control regions, that is, LMANcore (lateral magnocellular nucleus of the anterior nidopallium), RA (nucleus robustus arcopallialis) and HVC being less immunoreactive for DARPP-32 than 'shell' areas such as LMANshell, RAcup, AId (intermediate arcopallium) and HVCshelf. Our results suggest that whereas dopamine may modulate the shell pathways at various levels of the AFP, dopaminergic modulation of the core pathway occurs mainly through Area X, a basal ganglia nucleus. Further, secondary sensory cortices including the perientopallial belt, Fields L1 and L3 had higher DARPP-32-immunoreactivity than primary sensory cortical areas such as the pallial basolateral nucleus, entopallium proper and Field L2, corresponding to somatosensory, visual and auditory systems, respectively. We also found DARPP-32-rich axon terminals surrounding dopaminergic neurons in the ventral tegmental area-substantia nigra complex which in turn project to the striatum, suggesting that there may be a reciprocal modulation between these regions. Overall, DARPP-32 expression appears to be higher in areas involved in integrating sensory information, which further supports the role of this protein as a molecular integrator of different signal processing pathways.

Molecular architecture of the zebra finch arcopallium

The arcopallium, a key avian forebrain region, receives inputs from numerous brain areas and is a major source of descending sensory and motor projections. While there is evidence of arcopallial subdivisions, the internal organization or the arcopallium is not well understood. The arcopallium is also considered the avian homologue of mammalian deep cortical layers and/or amygdalar subdivisions, but one-to-one correspondences are controversial. Here we present a molecular characterization of the arcopallium in the zebra finch, a passerine songbird species and a major model organism for vocal learning studies. Based on in situ hybridization for arcopallial-expressed transcripts (AQP1, C1QL3, CBLN2, CNTN4, CYP19A1, ESR1/2, FEZF2, MGP, NECAB2, PCP4, PVALB, SCN3B, SCUBE1, ZBTB20, and others) in comparison with cytoarchitectonic features, we have defined 20 distinct regions that can be grouped into six major domains (anterior, posterior, dorsal, ventral, medial, and intermediate arcopallium, respectively; AA, AP, AD, AV, AM, and AI). The data also help to establish the arcopallium as primarily pallial, support a unique topography of the arcopallium in passerines, highlight similarities between the vocal robust nucleus of the arcopallium (RA) and AI, and provide insights into the similarities and differences of cortical and amygdalar regions between birds and mammals. We also propose the use of AMV (instead of nucleus taenia/TnA), AMD, AD, and AI as initial steps toward a universal arcopallial nomenclature. Besides clarifying the internal organization of the arcopallium, the data provide a coherent basis for further functional and comparative studies of this complex avian brain region.

The expression of tyrosine hydroxylase and DARPP-32 in the house crow (Corvus splendens) brain

Birds of the family Corvidae which includes diverse species such as crows, rooks, ravens, magpies, jays, and jackdaws are known for their amazing abilities at problem-solving. Since the catecholaminergic system, especially the neurotransmitter dopamine, plays a role in cognition, we decided to study the distribution of tyrosine hydroxylase (TH), the rate-limiting enzyme in the synthesis of catecholamines in the brain of house crows (Corvus splendens). We also studied the expression of DARPP-32 (dopamine and cAMP-regulated phosphoprotein), which is expressed in dopaminoceptive neurons. Our results demonstrated that as in other avian species, the expression of both TH and DARPP-32 was highest in the house crow striatum. The caudolateral nidopallium (NCL, the avian analogue of the mammalian prefrontal cortex) could be differentiated from the surrounding pallial regions based on a larger number of TH-positive "baskets" of fibers around neurons in this region and greater intensity of DARPP-32 staining in the neuropil in this region. House crows also possessed distinct nuclei in their brains which corresponded to song control regions in other songbirds. Whereas immunoreactivity for TH was higher in the vocal control region Area X compared to the surrounding MSt (medial striatum) in house crows, staining in RA and HVC was not as prominent. Furthermore, the arcopallial song control regions RA (nucleus robustus arcopallialis) and AId (intermediate arcopallium) were strikingly negative for DARPP-32 staining, in contrast to the surrounding arcopallium. Patterns of immunoreactivity for TH and DARPP-32 in "limbic" areas such as the hippocampus, septum, and extended amygdala have also been described.

Differential developmental changes in cortical representations of auditory-vocal stimuli in songbirds

Procedural skill learning requires iterative comparisons between feedback of self-generated motor output and a goal sensorimotor pattern. In juvenile songbirds, neural representations of both self-generated behaviors (each bird's own immature song) and the goal motor pattern (each bird's adult tutor song) are essential for vocal learning, yet little is known about how these behaviorally relevant stimuli are encoded. We made extracellular recordings during song playback in anesthetized juvenile and adult zebra finches ( Taeniopygia guttata) in adjacent cortical regions RA (robust nucleus of the arcopallium), AId (dorsal intermediate arcopallium), and RA cup, each of which is well situated to integrate auditory-vocal information: RA is a motor cortical region that drives vocal output, AId is an adjoining cortical region whose projections converge with basal ganglia loops for song learning in the dorsal thalamus, and RA cup surrounds RA and receives inputs from primary and secondary auditory cortex. We found strong developmental differences in neural selectivity within RA, but not in AId or RA cup. Juvenile RA neurons were broadly responsive to multiple songs but preferred juvenile over adult vocal sounds; in addition, spiking responses lacked consistent temporal patterning. By adulthood, RA neurons responded most strongly to each bird's own song with precisely timed spiking activity. In contrast, we observed a complete lack of song responsivity in both juvenile and adult AId, even though this region receives song-responsive inputs. A surprisingly large proportion of sites in RA cup of both juveniles and adults did not respond to song playback, and responsive sites showed little evidence of song selectivity. NEW & NOTEWORTHY Motor skill learning entails changes in selectivity for behaviorally relevant stimuli across cortical regions, yet the neural representation of these stimuli remains understudied. We investigated how information important for vocal learning in zebra finches is represented in regions analogous to infragranular layers of motor and auditory cortices during vs. after the developmentally regulated learning period. The results provide insight into how neurons in higher level stages of cortical processing represent stimuli important for motor skill learning.

Neural activity in cortico-basal ganglia circuits of juvenile songbirds encodes performance during goal-directed learning

Cortico-basal ganglia circuits are thought to mediate goal-directed learning by a process of outcome evaluation to gradually select appropriate motor actions. We investigated spiking activity in core and shell subregions of the cortical nucleus LMAN during development as juvenile zebra finches are actively engaged in evaluating feedback of self-generated behavior in relation to their memorized tutor song (the goal). Spiking patterns of single neurons in both core and shell subregions during singing correlated with acoustic similarity to tutor syllables, suggesting a process of outcome evaluation. Both core and shell neurons encoded tutor similarity via either increases or decreases in firing rate, although only shell neurons showed a significant association at the population level. Tutor similarity predicted firing rates most strongly during early stages of learning, and shell but not core neurons showed decreases in response variability across development, suggesting that the activity of shell neurons reflects the progression of learning.

Cortical inter-hemispheric circuits for multimodal vocal learning in songbirds

Vocal learning in songbirds and humans is strongly influenced by social interactions based on sensory inputs from several modalities. Songbird vocal learning is mediated by cortico-basal ganglia circuits that include the SHELL region of lateral magnocellular nucleus of the anterior nidopallium (LMAN), but little is known concerning neural pathways that could integrate multimodal sensory information with SHELL circuitry. In addition, cortical pathways that mediate the precise coordination between hemispheres required for song production have been little studied. In order to identify candidate mechanisms for multimodal sensory integration and bilateral coordination for vocal learning in zebra finches, we investigated the anatomical organization of two regions that receive input from SHELL: the dorsal caudolateral nidopallium (dNCL_SHELL ) and a region within the ventral arcopallium (Av). Anterograde and retrograde tracing experiments revealed a topographically organized inter-hemispheric circuit: SHELL and dNCL_SHELL , as well as adjacent nidopallial areas, send axonal projections to ipsilateral Av; Av in turn projects to contralateral SHELL, dNCL_SHELL , and regions of nidopallium adjacent to each. Av on each side also projects directly to contralateral Av. dNCL_SHELL and Av each integrate inputs from ipsilateral SHELL with inputs from sensory regions in surrounding nidopallium, suggesting that they function to integrate multimodal sensory information with song-related responses within LMAN-SHELL during vocal learning. Av projections share this integrated information from the ipsilateral hemisphere with contralateral sensory and song-learning regions. Our results suggest that the inter-hemispheric pathway through Av may function to integrate multimodal sensory feedback with vocal-learning circuitry and coordinate bilateral vocal behavior.

A neural circuit mechanism for regulating vocal variability during song learning in zebra finches

Motor skill learning is characterized by improved performance and reduced motor variability. The neural mechanisms that couple skill level and variability, however, are not known. The zebra finch, a songbird, presents a unique opportunity to address this question because production of learned song and induction of vocal variability are instantiated in distinct circuits that converge on a motor cortex analogue controlling vocal output. To probe the interplay between learning and variability, we made intracellular recordings from neurons in this area, characterizing how their inputs from the functionally distinct pathways change throughout song development. We found that inputs that drive stereotyped song-patterns are strengthened and pruned, while inputs that induce variability remain unchanged. A simple network model showed that strengthening and pruning of action-specific connections reduces the sensitivity of motor control circuits to variable input and neural ‘noise’. This identifies a simple and general mechanism for learning-related regulation of motor variability.

DOI:http://dx.doi.org/10.7554/eLife.03697.001

‘Practice makes perfect’ captures the essence of how we learn new skills. When learning to play a musical instrument, for example, it often takes hours of practice before we can play a single piece of music properly for the first time. And as we get better, the variability in our performance—which is an advantage during the early stages of learning—becomes less. Likewise, songbirds need lots of practice in order to master the intricate songs they need to sing to attract mates.

Studies in songbirds show that the neural circuits in the brain that are responsible for producing song and for generating vocal variability both converge on a motor control region called the robust nucleus of the arcopallium (or RA for short). However, the details of how learning a song leads to reduced variability in vocal performance are poorly understood.

Now Garst-Orozco et al. have investigated the relationship between learning and variability by studying brain slices of zebra finches. Their experiments reveal that the inputs received by RA neurons from a higher-order brain region that controls song change with practice, with some inputs becoming stronger and others being eliminated as the birds' singing ability improves. However, inputs received by RA neurons from the circuit that generates vocal variability do not change despite the song becoming increasingly precise.

Using a computer simulation, Garst-Orozco et al. show that the sensitivity of RA neurons to variable or ‘noisy’ input is reduced when inputs from the brain region that controls song are adaptively strengthened and eliminated. This ensures that when the notes and syllables that make up the bird's song have finally been learned, they will be uttered with high fidelity and precision. Intriguingly, motor skill learning in mammals have been associated with neural connectivity changes very similar to those described by Garst-Orozco et al., suggesting that insights from songbirds may lead to a better understanding of how ‘practice makes perfect’ also works in humans.

DOI:http://dx.doi.org/10.7554/eLife.03697.002

Calbindin expression in developing striatum of zebra finches and its relation to the formation of area X

A sexually dimorphic network of brain regions controls learning and production of song in zebra finches. How this specialized song system evolved is unknown. To start addressing this question, we focused on developmental differences between the sexes, using the expression of the calcium-binding protein calbindin (CB) during embryonic to adult stages to map out the early development of Area X, a male-specific striatal structure. We related this pattern to the expression of three transcription factors, Pax6 and Islet1 to delineate the striatal radial domains, and Nkx2.1 as a marker for cells of pallidal origin. An incipient Area X-CB+ domain became discernable at embryonic day 13 in the Islet1-ventral striatal field. This region contained many Nkx2.1-expressing cells with a morphology characteristic of migrating cells. Eight days after hatching (PHD) CB staining clearly delineated Area X. Another CB+ structure formed around PHD5 at the subpallial/pallial boundary. We call it the CB+striatal capsule (CB-StC) and discuss its relation with the previously described striatal capsule in vertebrates. The CB cell population in both Area X and CB-StC, but not in the surrounding striatum, colocalized with the striatal medium spiny neurons (MSNs) marker, D1-receptor associated signaling protein dopamine-and-cAMP-regulated phosphoprotein of 32 kDa, DARPP32. In females, CB-positive cells were also present in the rostral striatum but did not coalesce into an Area X-like structure. We discuss possible reasons for CB expression in MSNs in songbirds and mammals, but not described in chicken striatum.

Morphology of axonal projections from the high vocal center to vocal motor cortex in songbirds

Only birds that learn complex vocalizations have telencephalic brain regions that control vocal learning and production, including HVC (high vocal center), a cortical nucleus that encodes vocal motor output in adult songbirds. HVC projects to RA (robust nucleus of the arcopallium), a nucleus in motor cortex that in turn projects topographically onto hindbrain neurons innervating vocal muscles. Individual neurons projecting from HVC to RA (HVC(RA) ) fire sparsely to drive RA activity during song production. To advance understanding of how individual HVC neurons encode production of learned vocalizations, we reconstructed single HVC axons innervating RA in adult male zebra finches. Individual HVC(RA) axons were not topographically organized within RA: 1) axon arbors of HVC cell bodies located near each other sent branches to different subregions of RA, and 2) branches of single HVC axons terminated in different locations within RA. HVC(RA) axons also had a simple, sparse morphology, suggesting that a single HVC neuron activates a limited population of postsynaptic RA neurons. These morphological data are consistent with previous work showing that single HVC(RA) neurons burst sparsely for a brief period of time during the production of a song, indicating that ensembles of HVC(RA) neurons fire simultaneously to drive small temporal segments of song behavior. We also examined the morphology of axons projecting from HVC to RA cup, a region surrounding RA that receives input from auditory cortex. Axons projecting to RA cup also sent some branches into RA, suggesting direct integration between the sensory and motor circuits for song control.

Development of auditory-vocal perceptual skills in songbirds

Songbirds are one of the few groups of animals that learn the sounds used for vocal communication during development. Like humans, songbirds memorize vocal sounds based on auditory experience with vocalizations of adult “tutors”, and then use auditory feedback of self-produced vocalizations to gradually match their motor output to the memory of tutor sounds. In humans, investigations of early vocal learning have focused mainly on perceptual skills of infants, whereas studies of songbirds have focused on measures of vocal production. In order to fully exploit songbirds as a model for human speech, understand the neural basis of learned vocal behavior, and investigate links between vocal perception and production, studies of songbirds must examine both behavioral measures of perception and neural measures of discrimination during development. Here we used behavioral and electrophysiological assays of the ability of songbirds to distinguish vocal calls of varying frequencies at different stages of vocal learning. The results show that neural tuning in auditory cortex mirrors behavioral improvements in the ability to make perceptual distinctions of vocal calls as birds are engaged in vocal learning. Thus, separate measures of neural discrimination and behavioral perception yielded highly similar trends during the course of vocal development. The timing of this improvement in the ability to distinguish vocal sounds correlates with our previous work showing substantial refinement of axonal connectivity in cortico-basal ganglia pathways necessary for vocal learning.

Song tutoring in presinging zebra finch juveniles biases a small population of higher-order song-selective neurons toward the tutor song

We explored physiological changes correlated with song tutoring by recording the responses of caudal nidopallium neurons of zebra finches aged P21-P24 (days post hatching) to a broad spectrum of natural and synthetic stimuli. Those birds raised with their fathers tended to show behavioral evidence of song memorization but not of singing; thus auditory responses were not confounded by the birds' own vocalizations. In study 1, 37 of 158 neurons (23%) in 17 of 22 tutored and untutored birds were selective for only 1 of 10 stimuli comprising broadband signals, early juvenile songs and calls, female calls, and adult songs. Approximately 30% of the selective neurons (12/37 neurons in 9 birds) were selective for adult conspecific songs. All these were found in the song system nuclei HVC and paraHVC. Of 122 neurons (17 birds) in tutored birds, all of the conspecific song-selective neurons (8 neurons in 6 birds) were selective for the adult tutor song; none was selective for unfamiliar song. In study 2 with a different sampling strategy, we found that 11 of 12 song-selective neurons in 6 of 7 birds preferred the tutor song; none preferred unfamiliar or familiar conspecific songs. Most of these neurons were found in caudal lateral nidopallium (NCL) below HVC. Thus by the time a bird begins to sing, there are small numbers of tutor song-selective neurons distributed in several forebrain regions. We hypothesize that a small population of higher-order auditory neurons is innately selective for complex features of behaviorally relevant stimuli and these responses are modified by specific perceptual/social experience during development.

Auditory experience refines cortico-basal ganglia inputs to motor cortex via remapping of single axons during vocal learning in zebra finches

Experience-dependent changes in neural connectivity underlie developmental learning and result in life-long changes in behavior. In songbirds axons from the cortical region LMAN(core) (core region of lateral magnocellular nucleus of anterior nidopallium) convey the output of a basal ganglia circuit necessary for song learning to vocal motor cortex [robust nucleus of the arcopallium (RA)]. This axonal projection undergoes remodeling during the sensitive period for learning to achieve topographic organization. To examine how auditory experience instructs the development of connectivity in this pathway, we compared the morphology of individual LMAN(core)→RA axon arbors in normal juvenile songbirds to those raised in white noise. The spatial extent of axon arbors decreased during the first week of vocal learning, even in the absence of normal auditory experience. During the second week of vocal learning axon arbors of normal birds showed a loss of branches and varicosities; in contrast, experience-deprived birds showed no reduction in branches or varicosities and maintained some arbors in the wrong topographic location. Thus both experience-independent and experience-dependent processes are necessary to establish topographic organization in juvenile birds, which may allow birds to modify their vocal output in a directed manner and match their vocalizations to a tutor song. Many LMAN(core) axons of juvenile birds, but not adults, extended branches into dorsal arcopallium (Ad), a region adjacent to RA that is part of a parallel basal ganglia pathway also necessary for vocal learning. This transient projection provides a point of integration between the two basal ganglia pathways, suggesting that these branches convey corollary discharge signals as birds are actively engaged in learning.

Neural distribution of vasotocin receptor mRNA in two species of songbird

The neurohypophyseal hormones vasopressin and oxytocin are produced and released within the mammalian brain, where they act via multiple receptor subtypes. The neural distributions of these receptors, for example, V1a and oxytocin receptors, have been well described in many mammals. In birds, the distribution of binding sites for the homologous neuropeptides, vasotocin (VT) and mesotocin, has been studied in several species by using synthetic radioligands designed to bind to mammalian receptors. Such binding studies, however, may not reveal the specific distributions of each receptor subtype. To identify and map the receptors likely to bind VT and mesotocin, we generated partial cDNA sequences for four VT receptor subtypes, VT1, VT2 (V1b), VT3 (oxytocin-like), and VT4 (V1a), in white-throated sparrow (Zonotrichia albicollis) and zebra finch (Taeniopygia guttata). These genes shared high sequence identity with the homologous avian and mammalian neurohypophyseal peptide receptors, and we found evidence for VT1, VT3, and VT4 receptor mRNA expression throughout the brains of both species. As has been described in rodents, there was striking interspecific and intraspecific variation in the densities and distribution of these receptors. For example, whereas the VT1 receptor mRNA was more widespread in zebra finch brain, the VT3 (oxytocin-like) receptor mRNA was more prevalent in the sparrow brain. Although VT2 (V1b) receptor mRNA was abundant in the pituitary, it was not found in the brain. Because of their association with brain regions implicated in social behavior, the VT1, VT3, and VT4 receptors are all likely candidates for mediating the behavioral effects of VT.

A hypothesis for basal ganglia-dependent reinforcement learning in the songbird

Most of our motor skills are not innately programmed, but are learned by a combination of motor exploration and performance evaluation, suggesting that they proceed through a reinforcement learning (RL) mechanism. Songbirds have emerged as a model system to study how a complex behavioral sequence can be learned through an RL-like strategy. Interestingly, like motor sequence learning in mammals, song learning in birds requires a basal ganglia (BG)-thalamocortical loop, suggesting common neural mechanisms. Here, we outline a specific working hypothesis for how BG-forebrain circuits could utilize an internally computed reinforcement signal to direct song learning. Our model includes a number of general concepts borrowed from the mammalian BG literature, including a dopaminergic reward prediction error and dopamine-mediated plasticity at corticostriatal synapses. We also invoke a number of conceptual advances arising from recent observations in the songbird. Specifically, there is evidence for a specialized cortical circuit that adds trial-to-trial variability to stereotyped cortical motor programs, and a role for the BG in "biasing" this variability to improve behavioral performance. This BG-dependent "premotor bias" may in turn guide plasticity in downstream cortical synapses to consolidate recently learned song changes. Given the similarity between mammalian and songbird BG-thalamocortical circuits, our model for the role of the BG in this process may have broader relevance to mammalian BG function.

Mechanisms and time course of vocal learning and consolidation in the adult songbird

In songbirds, the basal ganglia outflow nucleus LMAN is a cortical analog that is required for several forms of song plasticity and learning. Moreover, in adults, inactivating LMAN can reverse the initial expression of learning driven via aversive reinforcement. In the present study, we investigated how LMAN contributes to both reinforcement-driven learning and a self-driven recovery process in adult Bengalese finches. We first drove changes in the fundamental frequency of targeted song syllables and compared the effects of inactivating LMAN with the effects of interfering with N-methyl-d-aspartate (NMDA) receptor-dependent transmission from LMAN to one of its principal targets, the song premotor nucleus RA. Inactivating LMAN and blocking NMDA receptors in RA caused indistinguishable reversions in the expression of learning, indicating that LMAN contributes to learning through NMDA receptor-mediated glutamatergic transmission to RA. We next assessed how LMAN's role evolves over time by maintaining learned changes to song while periodically inactivating LMAN. The expression of learning consolidated to become LMAN independent over multiple days, indicating that this form of consolidation is not completed over one night, as previously suggested, and instead may occur gradually during singing. Subsequent cessation of reinforcement was followed by a gradual self-driven recovery of original song structure, indicating that consolidation does not correspond with the lasting retention of changes to song. Finally, for self-driven recovery, as for reinforcement-driven learning, LMAN was required for the expression of initial, but not later, changes to song. Our results indicate that NMDA receptor-dependent transmission from LMAN to RA plays an essential role in the initial expression of two distinct forms of vocal learning and that this role gradually wanes over a multiday process of consolidation. The results support an emerging view that cortical-basal ganglia circuits can direct the initial expression of learning via top-down influences on primary motor circuitry.

Twitter evolution: converging mechanisms in birdsong and human speech

Vocal imitation in human infants and in some orders of birds relies on auditory-guided motor learning during a sensitive period of development. It proceeds from 'babbling' (in humans) and 'subsong' (in birds) through distinct phases towards the full-fledged communication system. Language development and birdsong learning have parallels at the behavioural, neural and genetic levels. Different orders of birds have evolved networks of brain regions for song learning and production that have a surprisingly similar gross anatomy, with analogies to human cortical regions and basal ganglia. Comparisons between different songbird species and humans point towards both general and species-specific principles of vocal learning and have identified common neural and molecular substrates, including the forkhead box P2 (FOXP2) gene.

A bird's eye view of neural circuit formation

Neural circuits underlying complex learned behaviors, such as speech in humans, develop under genetic constraints and in response to environmental influences. Little is known about the rules and mechanisms through which such circuits form. We argue that songbirds, with their discrete and well studied neural pathways underlying a complex and naturally learned behavior, provide a powerful model for addressing these questions. We briefly review current knowledge of how the song circuit develops during learning and discuss new possibilities for advancing the field given recent technological advances.

Language-related Cntnap2 gene is differentially expressed in sexually dimorphic song nuclei essential for vocal learning in songbirds

Multiple studies, involving distinct clinical populations, implicate contactin associated protein-like 2 (CNTNAP2) in aspects of language development and performance. While CNTNAP2 is broadly distributed in developing rodent brain, it shows a striking gradient of frontal cortical enrichment in developing human brain, consistent with a role in patterning circuits that subserve higher cognition and language. To test the hypothesis that CNTNAP2 may be important for learned vocal communication in additional species, we employed in situ hybridization to characterize transcript distribution in the zebra finch, an experimentally tractable songbird for which the neural substrate of this behavior is well established. Consistent with an important role in learned vocalization, Cntnap2 was enriched or diminished in key song control nuclei relative to adjacent brain tissue. Importantly, this punctuated expression was observed in males, but not females, in accord with the sexual dimorphism of neural circuitry and vocal learning in this species. Ongoing functional work will provide important insights into the relationship between Cntnap2 and vocal communication in songbirds and thereby clarify mechanisms at play in disorders of human cognition and language.

Conjunction of vocal production and perception regulates expression of the immediate early gene ZENK in a novel cortical region of songbirds

The cortical nucleus LMAN (lateral magnocellular nucleus of the anterior nidopallium) provides the output of a basal ganglia pathway that is necessary for acquisition of learned vocal behavior during development in songbirds. LMAN is composed of two subregions, a core and a surrounding shell, that give rise to independent pathways that traverse the forebrain in parallel. The LMAN(shell) pathway forms a recurrent loop that includes a cortical region, the dorsal region of the caudolateral nidopallium (dNCL), hitherto unknown to be involved with learned vocal behavior. Here we show that vocal production strongly induces the IEG product ZENK in dNCL of zebra finches. Hearing tutor song while singing is more effective at inducing expression in dNCL of juvenile birds during the auditory-motor integration stage of vocal learning than is hearing conspecific song. In contrast, hearing conspecific song is relatively more effective at inducing expression in adult birds, regardless of whether they are producing song. Furthermore, ZENK+ neurons in dNCL include projection neurons that are part of the LMAN(shell) recurrent loop and a high proportion of dNCL projection neurons express ZENK in singing juvenile birds that hear tutor song. Thus juvenile birds that are actively refining their vocal pattern to imitate a tutor song show high levels of ZENK induction in dNCL neurons when they are singing while hearing the song of their tutor and low levels when they hear a novel conspecific. This pattern indicates that dNCL is a novel brain region involved with vocal learning and that its function is developmentally regulated.

Social context rapidly modulates the influence of auditory feedback on avian vocal motor control

Sensory feedback is important for the learning and control of a variety of behaviors. Vocal motor production in songbirds is a powerful model system to study sensory influences on behavior because the learning, maintenance, and control of song are critically dependent on auditory feedback. Based on previous behavioral and neural experiments, it has been hypothesized that songs produced in isolation [undirected (UD) song] represent a form of vocal practice, whereas songs produced to females during courtship interactions [female-directed (FD) song] represent a form of vocal performance. According to this "practice versus performance" framework, auditory feedback should be more influential when birds engage in vocal practice than when they engage in vocal performance. To directly test this hypothesis, we used a computerized system to perturb auditory feedback at precise locations during the songs of Bengalese finches and compared the degree to which feedback perturbations caused song interruptions as well as changes to the sequencing and timing of syllables between interleaved renditions of UD and FD song. We found that feedback perturbation caused fewer song interruptions and smaller changes to syllable timing during FD song than during UD song. These data show that changes in the social context in which song is produced rapidly modulate the influence of auditory feedback on song control in a manner consistent with the practice versus performance framework. More generally, they indicate that, for song, as for other motor skills including human speech, the influence of sensory feedback on activity within vocal premotor circuitry can be dynamically modulated.

Molecular mapping of movement-associated areas in the avian brain: a motor theory for vocal learning origin

Vocal learning is a critical behavioral substrate for spoken human language. It is a rare trait found in three distantly related groups of birds-songbirds, hummingbirds, and parrots. These avian groups have remarkably similar systems of cerebral vocal nuclei for the control of learned vocalizations that are not found in their more closely related vocal non-learning relatives. These findings led to the hypothesis that brain pathways for vocal learning in different groups evolved independently from a common ancestor but under pre-existing constraints. Here, we suggest one constraint, a pre-existing system for movement control. Using behavioral molecular mapping, we discovered that in songbirds, parrots, and hummingbirds, all cerebral vocal learning nuclei are adjacent to discrete brain areas active during limb and body movements. Similar to the relationships between vocal nuclei activation and singing, activation in the adjacent areas correlated with the amount of movement performed and was independent of auditory and visual input. These same movement-associated brain areas were also present in female songbirds that do not learn vocalizations and have atrophied cerebral vocal nuclei, and in ring doves that are vocal non-learners and do not have cerebral vocal nuclei. A compilation of previous neural tracing experiments in songbirds suggests that the movement-associated areas are connected in a network that is in parallel with the adjacent vocal learning system. This study is the first global mapping that we are aware for movement-associated areas of the avian cerebrum and it indicates that brain systems that control vocal learning in distantly related birds are directly adjacent to brain systems involved in movement control. Based upon these findings, we propose a motor theory for the origin of vocal learning, this being that the brain areas specialized for vocal learning in vocal learners evolved as a specialization of a pre-existing motor pathway that controls movement.

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

The discrete neural network for songbird vocal communication provides an effective system to study neural mechanisms of learned motor behaviors in vertebrates. This system consists of two pathways--a vocal motor pathway used to produce learned vocalizations and a vocal pallial basal ganglia loop used to learn and modify the vocalizations. However, it is not clear how the loop exerts control over the motor pathway. To study the mechanism, we used expression of the neural activity-induced gene ZENK (or egr-1), which shows singing-regulated expression in a social context-dependent manner: high levels in both pathways when singing undirected and low levels in the lateral part of the loop and in the robust nucleus of the arcopallium (RA) of the motor pathway when singing directed to another animal. Here, we show that there are two parallel interactive parts within the pallial basal ganglia loop, lateral and medial, which modulate singing-driven ZENK expression of the motor pathway nuclei RA and HVC, respectively. Within the loop, the striatal and pallial nuclei appear to have opposing roles; the striatal vocal nucleus lateral AreaX is required for high ZENK expression in its downstream nuclei, particularly during undirected singing, while the pallial vocal lateral magnocellular nucleus of the anterior nidopallium is required for lower expression, particularly during directed singing. These results suggest a dynamic molecular interaction between the basal ganglia pathway and the motor pathway during production of a learned motor behavior.

Development of selectivity for natural sounds in the songbird auditory forebrain

In adult songbirds, auditory neurons in the primary auditory forebrain region of field L and a secondary auditory forebrain region of caudal mesopallium (CM) are highly responsive to natural sounds, such as conspecific song. Because these nuclei are involved in sensory representations of songs, we investigated how their function changes during development. We recorded neural responses to conspecific and tutor song and acoustically matched synthetic sounds in field L and lateral CM (CLM) of urethane-anesthetized juvenile male zebra finches of approximately 35 days of age. At this age, juvenile songbirds are memorizing the songs of their adult tutors but do not yet sing mature song. They are also starting to recognize songs of individual conspecifics. Compared with adult auditory forebrain neurons, juvenile neurons in field L were on average less responsive to auditory stimuli and exhibited less selectivity for natural sounds compared with the synthetic sounds. This developmental effect was more pronounced in the secondary subregions of L1 and L3 than in the primary thalamo-recipient subregion L2 of field L. CLM showed adultlike selectivity for natural sounds. Also, we did not find any evidence of memory for the tutor song in either field L or CLM. We note that the neural development of selective responses to conspecific song in the secondary subregions of field L is correlated with the emergence of individual song preference around 35 days of age. Therefore we suggest that the emergence of natural sound selectivity in field L could be important for the behavioral development of song recognition.

Enriched expression and developmental regulation of the middle-weight neurofilament (NF-M) gene in song control nuclei of the zebra finch

Songbirds evolved a complex set of dimorphic telencephalic nuclei that are essential for the learning and production of song. These nuclei, which together make up the oscine song control system, present several neurochemical properties that distinguish them from the rest of the telencephalon. Here we show that the expression of the gene encoding the middle-weight neurofilament (NF-M), an important component of the neuronal cytoskeleton and a useful tool for studying the cytarchitectonic organization of mammalian cortical areas, is highly enriched in large neurons within pallial song control nuclei (nucleus HVC, robustus nucleus of the arcopallium, and lateral magnocellular nucleus of the nidopallium) of male zebra finches (Taeniopygia guttata). We also show that this transcript is highly expressed in large neurons in the medulla, pons, midbrain, and thalamus. Moreover, we demonstrate that NF-M expression in song control nuclei changes during postembryonic development, peaking during an early phase of the song-learning period that coincides with the maturation of the song system. We did not observe changes in NF-M expression in auditory areas or in song control nuclei in the contexts of hearing song or singing, although these contexts result in marked induction of the transcription factor ZENK. This observation suggests that NF-M might not be under the regulatory control of ZENK in auditory areas or in song control nuclei. Overall, our data indicate that NF-M is a neurochemical marker for pallial song control nuclei and provide suggestive evidence of an involvement of NF-M in the development and/or maturation of the oscine song control system.

Silent Synapses in a Thalamo-Cortical Circuit Necessary for Song Learning in Zebra Finches

Developmental changes in synaptic properties may act to limit neural and behavioral plasticity associated with sensitive periods. This study characterized synaptic maturation in a glutamatergic thalamo-cortical pathway that is necessary for vocal learning in songbirds. Lesions of the projection from medial dorsolateral nucleus of the thalamus (DLM) to the cortical nucleus lateral magnocellular nucleus of the anterior nidopallium (LMAN) greatly disrupt song behavior in juvenile birds during early stages of vocal learning. However, such lesions lose the ability to disrupt vocal behavior in normal birds at 60–70 days of age, around the time that selective auditory tuning for each bird’s own song (BOS) emerges in LMAN neurons. This pattern has suggested that LMAN is involved in processing song-related information and evaluating the degree to which vocal motor output matches the tutor song to be learned. Analysis of reversed excitatory postsynaptic currents at DLM→LMAN synapses in in vitro slice preparations revealed a pronounced N-methyl-d-aspartate receptor (NMDAR)-mediated component in both juvenile and adult cells with no developmental decrease in the relative contribution of NMDARs to synaptic transmission. However, the synaptic failure rate at DLM→LMAN synapses in juvenile males during the sensitive period for song learning was significantly lower at depolarized potentials than at hyperpolarized potentials. In contrast, the failure rate at DLM→LMAN synapses did not differ at hyper- versus depolarized holding potentials in adult males that had completed the acquisition of a stereotyped song. This pattern indicates that juvenile cells have a higher incidence of silent (NMDAR-only) synapses, which are postsynaptically silent at hyperpolarized potentials due to the voltage-dependent gating of NMDARs. Thus the decreased involvement of the LMAN pathway in vocal behavior is mirrored by a decline in the incidence of silent synapses but not by changes in the relative number of NMDA and AMPA receptors at DLM→LMAN synapses. These findings suggest that a developmental decrease in silent synapses within LMAN may represent a neural correlate of behavioral plasticity during song learning.

Evidence for ?direct? and ?indirect? pathways through the song system basal ganglia

Song learning in oscine birds relies on a circuit known as the "anterior forebrain pathway," which includes a specialized region of the avian basal ganglia. This region, area X, is embedded within a telencephalic structure considered homologous to the striatum, the input structure of the mammalian basal ganglia. Area X has many features in common with the mammalian striatum, yet has distinctive traits, including largely aspiny projection neurons that directly innervate the thalamus and a cell type that physiologically resembles neurons recorded in the mammalian globus pallidus. We have proposed that area X is a mixture of striatum and globus pallidus and has the same functional organization as circuits in the mammalian basal ganglia. Using electrophysiological and anatomical approaches, we found that area X contains a functional analog of the "direct" striatopallidothalamic pathway of mammals: axons of the striatal spiny neurons make close contacts on the somata and dendrites of pallidal cells. A subset of pallidal neurons project directly to the thalamus. Surprisingly, we found evidence that many pallidal cells may not project to the thalamus, but rather participate in a functional analog of the mammalian "indirect" pathway, which may oppose the effects of the direct pathway. Our results deepen our understanding of how information flows through area X and provide more support for the notion that song learning in oscines employs physiological mechanisms similar to basal ganglia-dependent forms of motor learning in mammals.

Differential expression of glutamate receptors in avian neural pathways for learned vocalization

Learned vocalization, the substrate for human language, is a rare trait. It is found in three distantly related groups of birds-parrots, hummingbirds, and songbirds. These three groups contain cerebral vocal nuclei for learned vocalization not found in their more closely related vocal nonlearning relatives. Here, we cloned 21 receptor subunits/subtypes of all four glutamate receptor families (AMPA, kainate, NMDA, and metabotropic) and examined their expression in vocal nuclei of songbirds. We also examined expression of a subset of these receptors in vocal nuclei of hummingbirds and parrots, as well as in the brains of dove species as examples of close vocal nonlearning relatives. Among the 21 subunits/subtypes, 19 showed higher and/or lower prominent differential expression in songbird vocal nuclei relative to the surrounding brain subdivisions in which the vocal nuclei are located. This included relatively lower levels of all four AMPA subunits in lMAN, strikingly higher levels of the kainite subunit GluR5 in the robust nucleus of the arcopallium (RA), higher and lower levels respectively of the NMDA subunits NR2A and NR2B in most vocal nuclei and lower levels of the metabotropic group I subtypes (mGluR1 and -5) in most vocal nuclei and the group II subtype (mGluR2), showing a unique expression pattern of very low levels in RA and very high levels in HVC. The splice variants of AMPA subunits showed further differential expression in vocal nuclei. Some of the receptor subunits/subtypes also showed differential expression in hummingbird and parrot vocal nuclei. The magnitude of differential expression in vocal nuclei of all three vocal learners was unique compared with the smaller magnitude of differences found for nonvocal areas of vocal learners and vocal nonlearners. Our results suggest that evolution of vocal learning was accompanied by differential expression of a conserved gene family for synaptic transmission and plasticity in vocal nuclei. They also suggest that neural activity and signal transduction in vocal nuclei of vocal learners will be different relative to the surrounding brain areas.

Songbirds and the revised avian brain nomenclature

It has become increasingly clear that the standard nomenclature for many telencephalic and related brainstem structures of the avian brain is based on flawed once-held assumptions of homology to mammalian brain structures, greatly hindering functional comparisons between avian and mammalian brains. This has become especially problematic for those researchers studying the neurobiology of birdsong, the largest single group within the avian neuroscience community. To deal with the many communication problems this has caused among researchers specializing in different vertebrate classes, the Avian Brain Nomenclature Forum, held at Duke University from July 18-20, 2002, set out to develop a new terminology for the avian telencephalon and some allied brainstem cell groups. In one major step, the erroneous conception that the avian telencephalon consists mainly of a hypertrophied basal ganglia has been purged from the telencephalic terminology, and the actual parts of the basal ganglia and its brainstem afferent cell groups have been given new names to reflect their now-evident homologies. The telencephalic regions that were incorrectly named to reflect presumed homology to mammalian basal ganglia have been renamed as parts of the pallium. The prefixes used for the new names for the pallial subdivisions have retained most established abbreviations, in an effort to maintain continuity with the pre-existing nomenclature. Here we present a brief synopsis of the inaccuracies in the old nomenclature, a summary of the nomenclature changes, and details of changes for specific songbird vocal and auditory nuclei. We believe this new terminology will promote more accurate understanding of the broader neurobiological implications of song control mechanisms and facilitate the productive exchange of information between researchers studying avian and mammalian systems.

Origin of the anterior forebrain pathway

The brain nuclei and pathways comprising the song system of oscine songbirds bear many similarities with circuits in other bird species and in mammals. This suggests that the song system evolved as a specialization of pre-existing circuits and may retain fundamental properties in common with those of other taxa. Here we review evidence for these similarities, including electrophysiological, morphological, and neurochemical data for identifying specific cell types. In addition, we discuss connectional data, addressing similarities in axonal projections among nuclei across taxa. We focus primarily on the anterior forebrain pathway, a circuit essential for song learning and vocal plasticity, because the evidence is strongest that this circuit is homologous to mammalian circuits. These fundamental similarities highlight the importance of comparative approaches; for example, understanding the role the anterior forebrain pathway plays in song plasticity may shed light on general principles of basal ganglia function. In addition, understanding specializations of such circuits in songbirds may illuminate specific innovations critical for vocal learning.

Developmental regulation of basal ganglia circuitry during the sensitive period for vocal learning in songbirds

A hallmark of sensitive periods of development is an enhanced capacity for learning, such that experience exerts a profound effect on the brain resulting in the establishment of behaviors and underlying neural circuitry that can last a lifetime. Songbirds, like humans, have a sensitive period for vocal learning: they acquire the sounds used for vocal communication during a restricted period of development. In principle, any organism that undertakes vocal learning is faced with the same challenge: to form some representation of target vocal sounds based on auditory experience, and then to translate that auditory target into a motor program that reproduces the sound. Both birds and humans achieve this translation by using auditory (and other) feedback resulting from incipient vocalizations ("babbling" in humans, "subsong" in birds) to adjust motor commands until vocal output produces a good copy of the target sounds. Similarities between vocal learning in birds and humans suggest that many aspects of the learning process have evolved to meet demands imposed by vocal communication. Thus songbirds provide a valuable animal model in which to study the physiological basis of learned vocal communication and the nature of sensitive periods in general. In this article, I describe aspects of both behavioral and neural frameworks that currently inform our thinking about mechanisms underlying vocal learning and behavior in songbirds, and highlight ideas that may need re-examination.

Contributions of the Anterior Forebrain Pathway to Vocal Plasticity

The anterior forebrain pathway (AFP) is a basal ganglia-dorsal forebrain circuit that is prominent specifically in birds that learn to sing. This circuit is interconnected with the song motor pathway, is active during song production, and contains neurons that are selective for the sound of the bird's own song, suggesting an important role for the AFP in vocal behavior. However, interruption of the AFP by lesions in adult birds has little overt effect on the production of learned song. In contrast, lesions in juvenile birds prevent the normal progression of song learning. Moreover, lesions in adults, while not disrupting production, can prevent experience-dependent plasticity of song. Such data implicate the AFP specifically in song learning and vocal plasticity. This chapter reviews some of the experimental evidence supporting a role for the AFP in these processes and discusses potential instructive and permissive functions of the AFP in vocal plasticity.

Revised nomenclature for avian telencephalon and some related brainstem nuclei

The standard nomenclature that has been used for many telencephalic and related brainstem structures in birds is based on flawed assumptions of homology to mammals. In particular, the outdated terminology implies that most of the avian telencephalon is a hypertrophied basal ganglia, when it is now clear that most of the avian telencephalon is neurochemically, hodologically, and functionally comparable to the mammalian neocortex, claustrum, and pallial amygdala (all of which derive from the pallial sector of the developing telencephalon). Recognizing that this promotes misunderstanding of the functional organization of avian brains and their evolutionary relationship to mammalian brains, avian brain specialists began discussions to rectify this problem, culminating in the Avian Brain Nomenclature Forum held at Duke University in July 2002, which approved a new terminology for avian telencephalon and some allied brainstem cell groups. Details of this new terminology are presented here, as is a rationale for each name change and evidence for any homologies implied by the new names. Revisions for the brainstem focused on vocal control, catecholaminergic, cholinergic, and basal ganglia-related nuclei. For example, the Forum recognized that the hypoglossal nucleus had been incorrectly identified as the nucleus intermedius in the Karten and Hodos (1967) pigeon brain atlas, and what was identified as the hypoglossal nucleus in that atlas should instead be called the supraspinal nucleus. The locus ceruleus of this and other avian atlases was noted to consist of a caudal noradrenergic part homologous to the mammalian locus coeruleus and a rostral region corresponding to the mammalian A8 dopaminergic cell group. The midbrain dopaminergic cell group in birds known as the nucleus tegmenti pedunculopontinus pars compacta was recognized as homologous to the mammalian substantia nigra pars compacta and was renamed accordingly; a group of gamma-aminobutyric acid (GABA)ergic neurons at the lateral edge of this region was identified as homologous to the mammalian substantia nigra pars reticulata and was also renamed accordingly. A field of cholinergic neurons in the rostral avian hindbrain was named the nucleus pedunculopontinus tegmenti, whereas the anterior nucleus of the ansa lenticularis in the avian diencephalon was renamed the subthalamic nucleus, both for their evident mammalian homologues. For the basal (i.e., subpallial) telencephalon, the actual parts of the basal ganglia were given names reflecting their now evident homologues. For example, the lobus parolfactorius and paleostriatum augmentatum were acknowledged to make up the dorsal subdivision of the striatal part of the basal ganglia and were renamed as the medial and lateral striatum. The paleostriatum primitivum was recognized as homologous to the mammalian globus pallidus and renamed as such. Additionally, the rostroventral part of what was called the lobus parolfactorius was acknowledged as comparable to the mammalian nucleus accumbens, which, together with the olfactory tubercle, was noted to be part of the ventral striatum in birds. A ventral pallidum, a basal cholinergic cell group, and medial and lateral bed nuclei of the stria terminalis were also recognized. The dorsal (i.e., pallial) telencephalic regions that had been erroneously named to reflect presumed homology to striatal parts of mammalian basal ganglia were renamed as part of the pallium, using prefixes that retain most established abbreviations, to maintain continuity with the outdated nomenclature. We concluded, however, that one-to-one (i.e., discrete) homologies with mammals are still uncertain for most of the telencephalic pallium in birds and thus the new pallial terminology is largely devoid of assumptions of one-to-one homologies with mammals. The sectors of the hyperstriatum composing the Wulst (i.e., the hyperstriatum accessorium intermedium, and dorsale), the hyperstriatum ventrale, the neostriatum, and the archistriatum have been renamed (respectively) the hyperpallium (hypertrophied pallium), the mesopallium (middle pallium), the nidopallium (nest pallium), and the arcopallium (arched pallium). The posterior part of the archistriatum has been renamed the posterior pallial amygdala, the nucleus taeniae recognized as part of the avian amygdala, and a region inferior to the posterior paleostriatum primitivum included as a subpallial part of the avian amygdala. The names of some of the laminae and fiber tracts were also changed to reflect current understanding of the location of pallial and subpallial sectors of the avian telencephalon. Notably, the lamina medularis dorsalis has been renamed the pallial-subpallial lamina. We urge all to use this new terminology, because we believe it will promote better communication among neuroscientists. Further information is available at http://avianbrain.org

An immunohistochemical and pathway tracing study of the striatopallidal organization of area X in the male zebra finch

Area X is a nucleus within songbird basal ganglia that is part of the anterior forebrain song learning circuit. It receives cortical song-related input and projects to the dorsolateral medial nucleus of thalamus (DLM). We carried out single- and double-labeled immunohistochemical and pathway tracing studies in male zebra finch to characterize the cellular organization and circuitry of area X. We found that 5.4% of area X neuronal perikarya are relatively large, possess aspiny dendrites, and are rich in the pallidal neuron/striatal interneuron marker Lys8-Asn9-neurotensin8-13 (LANT6). Many of these perikarya were found to project to the DLM, and their traits suggest that they are pallidal. Area X also contained several neuron types characteristic of the striatum, including interneurons co-containing LANT6 and the striatal interneuron marker parvalbumin (2% of area X neurons), interneurons containing parvalbumin but not LANT6 (4.8%), cholinergic interneurons (1.4%), and neurons containing the striatal spiny projection neuron marker dopamine- and adenosine 3',5'-monophosphate-regulated phosphoprotein (DARPP-32) (30%). Area X was rich in substance P (SP)-containing terminals, and many ended on area X neurons projecting to the DLM with the woolly fiber morphology characteristic of striatopallidal terminals. Although SP+ perikarya were not detected in area X, prior studies suggest it is likely that SP-synthesizing neurons are present and the source of the SP+ input to area X neurons projecting to the DLM. Area X was poor in enkephalinergic fibers and perikarya. The present data support the premise that area X contains both striatal and pallidal neurons, with the striatal neurons likely to include SP+ neurons that project to the pallidal neurons.

Metabolic and neural activity in the song system nucleus robustus archistriatalis: effect of age and gender

The sexually dimorphic robust archistriatal nucleus (RA) represents the telencephalic output of the bird song system. Here, we document sex-dependent changes in both the metabolic and neuronal activity of RA during the sensory and sensorimotor phases of song learning. From posthatching day (PHD) 20-63 in males but not females, RA and its input nucleus HVc showed sharp increases in cytochrome oxidase (CO) activity relative to surrounding archistriatum and the underlying shelf, respectively. In urethane-anesthetized birds, during the same period, the spontaneous activity of male RA neurons underwent dramatic changes in firing rate, distribution of interspike intervals, and bursting frequency, compared with other archistriatal cells. At PHD 20-21, RA neurons had extremely slow, irregular firing rates in birds of both sexes. In males, from PHD 30-36, RA neurons increased their firing rates and spiking activity became more regular, and at approximately PHD 38, strong bursts followed by inhibition (which in awake animals is associated with singing) began to be observed. Dual recordings from RA and HVc revealed synchronous bursting, with RA spikes lagging approximately 10 msec behind HVc. We conclude that changes in relative CO activity correlate with changes in spontaneous firing rates within RA and that patterns of RA spontaneous activity exhibit gradual change as birds enter early song and then again for plastic song. The emergence of strong burst patterns in RA occurs later in life than does input from HVc as established by tracer studies or based on observed HVc bursting in young animals.

What songbirds teach us about learning

Bird fanciers have known for centuries that songbirds learn their songs. This learning has striking parallels to speech acquisition: like humans, birds must hear the sounds of adults during a sensitive period, and must hear their own voice while learning to vocalize. With the discovery and investigation of discrete brain structures required for singing, songbirds are now providing insights into neural mechanisms of learning. Aided by a wealth of behavioural observations and species diversity, studies in songbirds are addressing such basic issues in neuroscience as perceptual and sensorimotor learning, developmental regulation of plasticity, and the control and function of adult neurogenesis.

Development of individual axon arbors in a thalamocortical circuit necessary for song learning in zebra finches

Individual axon arbors within developing neural circuits are remodeled during restricted sensitive periods, leading to the emergence of precise patterns of connectivity and specialized adaptive behaviors. In male zebra finches, the circuit connecting the medial dorsolateral nucleus of the thalamus (DLM) and its cortical target, the lateral magnocellular nucleus of the anterior neostriatum (lMAN), is crucial for the acquisition of a normal vocal pattern during the sensitive period for song learning. The shell subregion of lMAN as well as the entire terminal field of DLM axons within lMAN undergo a striking increase in overall volume during early stages of vocal learning followed by an equally substantial decrease by adulthood, by which time birds have acquired stable song patterns. Because the total number of DLM neurons remains stable throughout this period, the dramatic changes within the overall DLM-->lMAN circuit are presumably attributable to dynamic rearrangements at the level of individual DLM axon arbors over the course of vocal learning. To study such rearrangements directly, we reconstructed individual DLM axon arbors in three dimensions at different stages during vocal learning. Unlike axon arbors in other model systems, in which the number of branches increases during development, DLM arbors are unusual in that they have the greatest number of branches at the onset of vocal learning and undergo large-scale retraction during the sensitive period for song learning. Decreases in the degree of overlap between DLM arbors apparently contribute to the increased overall volume of the DLM-->lMAN circuit during vocal learning. These developmental changes in DLM axon arbors occur at the height of the sensitive period for vocal learning, and hence may represent either a morphological correlate of song learning or a necessary prerequisite for acquisition of song.

The role of auditory experience in the formation of neural circuits underlying vocal learning in zebra finches

The initial establishment of topographic mapping within developing neural circuits is thought to be shaped by innate mechanisms and is primarily independent of experience. Additional refinement within topographic maps leads to precise matching between presynaptic and postsynaptic neurons and is thought to depend on experiential factors during specific sensitive periods in the animal's development. In male zebra finches, axonal projections of the cortical lateral magnocellular nucleus of the anterior neostriatum (lMAN) are critically important for vocal learning. Overall patterns of topographic organization in the majority of these circuits are adult-like throughout the sensitive period for vocal learning and remain stable despite large-scale functional and morphological changes. However, topographic organization within the projection from the core subregion of lMAN (lMAN(core)) to the motor cortical robust nucleus of the archistriatum (RA) is lacking at the onset of song development and emerges during the early stages of vocal learning. To study the effects of song-related experience on patterns of axonal connectivity within different song-control circuits, we disrupted song learning by deafening juvenile zebra finches or exposing them to loud white noise throughout the sensitive period for song learning. Depriving juvenile birds of normal auditory experience delayed the emergence of topographic specificity within the lMAN(core)-->RA circuit relative to age-matched controls, whereas topographic organization within all other projections to and from lMAN was not affected. The projection from lMAN(core) to RA therefore provides an unusual example of experience-dependent modification of large-scale patterns of brain circuitry, in the sense that auditory deprivation influences the development of overall topographic organization in this pathway.

Birdsong: models and mechanisms

Recent studies have provided important information concerning the neural signals that subserve vocal learning in songbirds: advanced signal processing techniques are beginning to clarify the behavioral trajectories followed by developing birds; single-unit physiology in behaving animals is providing important clues about sensory and motor representations during learning; in vitro whole-cell recordings are revealing patterns of synaptic communication; and experimental alterations in song behavior have advanced our understanding of specific structure-function relationships. The construction of theoretical and computational models will be crucial in integrating such disparate experimental results.