Calcium imaging in canary (serinus canaria) HVC reveals latent states supporting behavioral sequencing with long range history dependence Yarden Cohen (yardenc@bu.edu) Department of Biology, Boston University, 24 Cummington Mall Boston, MA 02215 USA Jun Shen (junshen@bu.edu) Department of Psychological and Brain Sciences, Boston University, 610 Commonwealth Avenue Boston, MA 02215 USA Dawit Semu (dawits@bu.edu) Department of Biology, Boston University, 24 Cummington Mall Boston, MA 02215 USA Timothy M. Otchy (totchy@bu.edu) Department of Biology, Boston University, 24 Cummington Mall Boston, MA 02215 USA Timothy J. Gardner (timothyg@ bu.edu) Department of Biology, Boston University, 24 Cummington Mall Boston, MA 02215 USA Abstract: History dependent behavior is a key readout of neural processing. In skills, like speech or dance, motor sequences follow syntactic rules in which transitions between motor elements rely on past actions. Canary songs are defined by syllable repeats, called phrases, whose syntax exhibits long range order. The phrase sequence neural underpinnings must either rely on fixed action patterns or maintain historic context to influence ongoing transitions. To discriminate such mechanisms, we recorded Ca 2+ signals from the premotor nucleus HVC in freely behaving canaries. We find that song history is reflected in identified ROIs up to 4 phrases apart, spanning up to 3 seconds and 40 syllables and that some ROIs exhibit mixed history selectivity. Moreover, we find that signals, reflecting sequence history information are more frequent during phrase transitions that are history dependent compared to history insensitive ones. These findings suggest that the network dynamics reflects historic context relevant to flexible transitions. Additionally, we find ROIs whose signals last several seconds and span 3-4 phrases. These signals are rarely modulated by syllable or phrase boundaries and initiate mostly during stereotyped sequences, suggesting distinct network dynamics during stereotyped and variable behavior. Keywords: canary motor syntax; HVC imaging; hidden states Background Natural behavior provides an uncontrived readout of neural computations. Analyses of neural recordings, accompanying behavior, often require considering additional variables. These factors can be intrinsic, such as electrical and neuromodulator states affecting motion-related spiking (Schiemann et al., 2015), or extrinsic, such as stimuli that drive behaviorally relevant neural representations (Panzeri, Harvey, Piasini, Latham, & Fellin, 2017). Here we pursue the idea that the behavior’s dependence on its own history offers valuable constraints on the neural processing. These constraints can be mechanical or statistical limitations, recently used in relating motor plans to the previous movement and current limb position to reveal its premotor representation (Glaser, Perich, Ramkumar, Miller, & Kording, 2018). Alternatively, in behavior that can be reliably divided into basic elements (c.f. Wiltschko et al., 2015), the syntax of motor sequences provides a powerful tool for studying the neural basis of long range history- dependent behavior. To pursue this approach, we take advantage of a uniquely rich and inherently segmented behavior - the song repertoire of the domestic canary (serinus canaria). All songbirds have a learned and naturally recurring behavior, song, whose temporal structure is largely governed by the premotor nucleus HVC (Hahnloser, Kozhevnikov, & Fee, 2002; Long & Fee, 2008; Nottebohm, Stokes, & Leonard, 1976; Wang, Herbst, Keller, & Hahnloser, 2008). Canary song repertoires are defined by units of repeated syllables, known as phrases (Figure 1a), whose syntax is controlled separately from the syllables’ identity (Gardner, Naef, & Nottebohm, 2005) and which exhibit long range order (Markowitz, Ivie, Kligler, & Gardner, 2013). Specifically, certain phrase transitions have history dependence, originating 2-3 phrases upstream (Figure 1b). Accordingly, to support the generation of phrase sequences, the premotor neural activity must either rely on a complex memorized repertoire, receive directives from other brain regions, or maintain historic context to influence ongoing transitions. To investigate such mechanisms, we recorded and analyzed [Ca 2+ ] dynamics from canary HVC projection neurons. We discover two novel types of neural correlates –