PROJECT SUMMARY Sequential neural activity is ubiquitous throughout the brain and is thought to underly many essential processes from working memory to complex motor behavior. However, the mechanisms of sequence generation still remain unknown; whether neural sequences are generated locally or emerge from the cooperation between local circuits and long-range inputs is controversial. For example, input from thalamic regions has been shown to be involved in the sequential activity that occurs in cortex during skilled motor acts. The adult male zebra finch provides an ideal system to investigate mechanisms of cortical sequence generation. After learning his courtship song through months of extensive practice as a juvenile, the zebra finch produces a stereotyped song composed of repeated complex vocal elements known as syllables. Proper song production requires the coordinated action of multiple nuclei across the song-production network, including two cortical regions, HVC (proper name) and the robust nucleus of the arcopallium (RA) and the thalamic nucleus uvaeformis (Uva). While lesion studies have shown the importance of Uva to song production, the role of these thalamic inputs in facilitating the sequential activity in HVC and RA is still unclear. In this proposal, I will test three models of sequence generation in the zebra finch song-production network, each of which attributes a different role to the thalamus in shaping cortical sequences. In model 1, thalamic input is required for moment-to-moment progression through the cortical sequence. In model 2, thalamic input is necessary to link discrete cortical sequences encoding individual syllables. In model 3, the thalamus acts to orchestrate progression through a continuous cortical sequence that encodes the entire song. Previous studies have sought to distinguish between these models by assessing the behavioral impact of lesioning Uva. Lesions eliminate singing behavior, thus precluding any opportunity to study the circuit. However, sleep provides the opportunity to probe this circuit, where song-like `replay' fragments have been observed in individual cortical neurons and neuron pairs without subsequent vocalizations. I will first adapt high-density silicon probe technology for the zebra finch and build analytical tools to decode the song-content of sleep replay. Next I will investigate the role of thalamic input in cortical sequence generation by perturbing this input and determining the impact this perturbation has on cortical replay. In this way, I will uncover the role of thalamic input in cortical sequences underlying complex skilled behavior in the zebra finch. These experiments will increase our understanding of how thalamic inputs contribute to cortical dynamics in healthy networks and provide insight into pathological processes that disrupt these dynamics during disease.