The central goal of the proposed project is to develop a conceptual framework for understanding the function of basal forebrain cholinergic neurons in motor sequence learning. These neurons supply the motor cortex with acetylcholine, a neuromodulator indispensable for motor learning and associated synaptic plasticity. Despite the profound deficits in motor learning observed following selective cholinergic ablation, surprisingly little is known about the activity of cholinergic neurons during sequenced motor behavior. Much of what we know about this system comes from recordings in animals performing simple stimulus-response tasks to obtain external rewards like food or water. Most human behaviors, like speech or playing an instrument, are not learned in pursuit of external rewards, but instead are learned by matching performance to internal goals. The role of basal forebrain cholinergic neurons in this form of learning is not known. Like humans, songbirds learn to vocalize through an iterant process of trial and error ? matching vocal performance to an auditory memory of their tutor?s song without an explicit need for external reinforcement. Importantly, songbirds have a cholinergic projection from the basal forebrain to primary motor cortical circuits required for vocal learning and production. The function of this projection is unknown. In the present proposal, I investigate the role of cholinergic inputs to cortical song motor circuits during vocal learning. I hypothesize that cholinergic neurons transmit an expected performance uncertainty signal that enables plasticity in cortical motor circuits. In principle, expected performance uncertainty could arise as a result of repeated errors during specific parts of a motor sequence. I hypothesize that this uncertainty signal is necessary for motor sequence learning.
In Aim 1, I propose to test the requirement of cholinergic signaling for vocal learning in songbirds. I will use reverse microdialysis in vocal motor cortex to chronically block acetylcholine receptors in young birds as they learn to sing.
In Aim 2, I propose to identify the synaptic inputs onto cholinergic neurons projecting to the song motor cortex, which are presently unknown. In order to construct an expected performance uncertainty signal, two pieces of information are required ? song timing and performance error. We will determine the neural substrates for these signals using established anatomical tracing techniques and electrophysiological mapping in anesthetized birds.
Aim 3 will directly test for neural signature of expected performance uncertainty in basal forebrain neurons projecting to the song motor cortex during singing. I will record from antidromically-identified, motor cortex-projecting basal forebrain neurons while experimentally controlling performance uncertainty at specific times in the song with distorted auditory feedback. Distorted parts of the vocal sequence are expected to have higher outcome uncertainty compared to undistorted parts. I hypothesize that expected uncertainty signals would be expressed as temporally precise increases in neural activity immediately preceding the distorted time in the song and provide preliminary results in support of this hypothesis. Combined together, the proposed research aims to establish precise computational roles for basal forebrain cholinergic neurons during motor sequence learning and is part of a larger endeavor to understand the function of different neuromodulatory systems in learning.
The basal forebrain cholinergic system is important for sensory, cognitive, and motor functions. Despite the involvement of this system in Alzheimer?s and Parkinson?s disease, schizophrenia, amnesia, age-related cognitive decline, and dystonia, little is known about the actual signals transmitted by cholinergic neurons. This proposal aims to leverage advantages of the songbird model system to elucidate the function of basal forebrain cholinergic during natural sensorimotor learning.
| Gadagkar, Vikram; Puzerey, Pavel A; Chen, Ruidong et al. (2016) Dopamine neurons encode performance error in singing birds. Science 354:1278-1282 |
| Puzerey, Pavel A; Galán, Roberto F (2014) On how correlations between excitatory and inhibitory synaptic inputs maximize the information rate of neuronal firing. Front Comput Neurosci 8:59 |
| Pucilowska, Joanna; Puzerey, Pavel A; Karlo, J Colleen et al. (2012) Disrupted ERK signaling during cortical development leads to abnormal progenitor proliferation, neuronal and network excitability and behavior, modeling human neuro-cardio-facial-cutaneous and related syndromes. J Neurosci 32:8663-77 |
