Thank you for your interest in our work. We are using the powerful zebra finch model to study the neurophysiological bases of vocal learning.

Almost 40 years ago, the Template Hypothesis proposed that songbirds learn their songs in two steps: (1) they memorize the song of a song tutor and (2) they match their own song to the memory of the tutor song (the "template") using auditory feedback (Konishi, 1965). Since then, birdsong physiologists have looked for a neural signal that might be consistent with this hypothesis, the template-matching signal. In addition, we have been searching for evidence that this template-matching signal instructs vocal learning through direct interactions with the motor control circuitry.

Our lab has found evidence for a template-matching signal that activates certain neurons in a song motor control brain area (HVC, formerly known as the High Vocal Center). Further, the HVC neurons that respond to auditory input are physiologically very different during a critical period for song learning (the sensorimotor phase) than during adulthood: They have prolonged bursts (Figure 4).

We hypothesize that the prolonged bursts enable temporal overlap of sensory and motor activity (Figure 5). Temporal coincidence is important in many forms of activity-dependent plasticity, such as long-term potentiation. Several synapses lie between the cortical HVC area and the muscles that produce singing. In addition, several synapses lie between the ear and HVC. Because of the large number of synapses, there are substantial pauses between the timing of motor activity that produces singing and the sensory activity that would indicate how well the singing was performed (called ‘auditory feedback’ and essential for vocal learning). Thus, the prolonged bursting may serve to close the temporal gap between motor-related activity and sensory activity and enable activity dependent plasticity.

Major Projects

The span of our research closes the gap between cellular/molecular and systems/behavioral neuroscience. We study the neural bases of song development by recording from awake, behaving juvenile finches in the process of song learning. The Nick lab has recorded neural populations longitudinally over several weeks. This is unparalleled in the field, as no other lab has been able to record neural activity for more than a few days. We have also recorded from single neurons in neuronal ensembles over several hours in behaving animals (with multi-electrode bundles). No other birdsong lab has reported success with this technique. Our lab utilizes rigorous sorting and analysis techniques that enable us to have high confidence that our single unit recordings represent the activity of a single neuron and only the activity of that neuron.

Beginning with HVC, we are systematically examining the development of neural activity in the context of behavior and then identifying the cellular and molecular mechanisms that underlie the changes in activity. We have made several important discoveries that are described below:

1. Development of song system activity. This series of studies utilizes chronic longitudinal neural recordings of song control areas in juveniles as they actively learn their songs. This approach is unusual because it examines activity in the same minimally perturbed animal across development, learning, and wake/sleep states. We have found:

• HVC vocal-associated bursting activity changes with development and correlates with behavior. Prolonged HVC activity outlasts the vocalization during the sensorimotor critical period.
  
• HVC sleep activity increases with development and correlates with behavior. HVC is less active during sleep in juveniles, whose song degrades overnight, than during sleep in adults, whose song behavior is stable overnight. These data suggest that sleep activity in the song system stabilizes song behavior.

Collectively, these data indicate that HVC activity is plastic during the sensorimotor phase and correlates with behavior. Further, they illuminate the function and mechanism of sleep activity and of prolonged bursting during a sensorimotor critical period.

2. Activity of neuronal ensembles. Using multi-electrode ensemble recording, we are illuminating the cellular mechanisms of song learning. The Nick lab has found that:

• A subset of HVC neurons exhibit prolonged bursting during the sensorimotor phase.
  
• The neurons that show prolonged bursting are the only neurons in the juvenile that respond to auditory stimulation.
  
• The neurons that show prolonged bursting appear to be interneurons, based on action potential rate and duration.

3. Perturbation of neural sleep activity and auditory feedback. Most of the experiments described above are necessary descriptive analyses of activity during development. We have also begun to perturb sleep activity and auditory feedback with novel methods that we have designed and implemented. Our lab is testing specific hypotheses that are suggested by the experiments described above.

• Preliminary data indicate that we can eliminate HVC sleep activity. Our novel method does not perturb sleep overall, but only in the song system (according to electroencephalography in other brain locations). Our lab is utilizing this method to selectively perturb song system activity during sleep in an effort to parse out the neuronal and behavioral correlates of sleep activity in HVC.
  
• Preliminary data indicate that perturbation of auditory feedback using masking noise or acute deafening decreases HVC activity during singing. Auditory feedback is essential for song learning, yet no neural evidence of auditory feedback in the song system has ever been reported (Konishi, 2004). We have developed the acute deafening protocol to rigorously test the hypothesis that the decrease that we see in HVC activity is due to changes in auditory feedback and not, for example, in premotor activity that drives singing.

4. Expression of perineuronal nets during song development. Perineuronal nets (PNNs) have a role in the closure of the critical period of ocular dominance plasticity (Hensch, 2005). We are using immunohistochemistry, song and neural recording, and destabilization of PNNs at various stages of song learning to investigate the role of PNNs in the song system during the sensorimotor critical period.

Please contact me with any questions.