A cerebellar-thalamocortical pathway drives behavioral context-dependent movement initiation

•DN/IPN thalamocortical activity conveys a reliable feedforward motor timing signal•Silencing DN/IPN or recipient regions of thalamus blocks movement initiation•Photostimulation the pathway triggers movement•Thalamocortical activation drives behavioral context-dependent initiation Executing learned behaviors often requires transformation sensory cues into patterns commands that generate appropriately timed actions. The cerebellum and are two key areas involved in shaping cortical output movement, but contribution cerebellar-thalamocortical to voluntary remains poorly understood. Here, we investigated how an auditory “go cue” transforms these changes relate initiation. Population responses dentate/interpositus-recipient reflect time-locked increase immediately prior is temporally uncoupled from go cue, indicative fixed-latency signal. Blocking cerebellar thalamic suppresses initiation, while stimulation movements manner. Our findings show output, via thalamus, shapes necessary for ability actions response hallmark mammalian control. Movement controlled part by cerebellum, as dysfunction leads execution (Bastian Thach, 1995Bastian A.J. Thach W.T. Cerebellar outflow lesions: comparison deficits resulting lesions at levels thalamus.Ann. Neurol. 1995; 38: 881-892Crossref PubMed Scopus (0) Google Scholar; Holmes, 1939Holmes G. man.Brain. 1939; 62: 1-30Crossref (578) Milak et al., 1997Milak M.S. Shimansky Y. Bracha V. Bloedel J.R. Effects inactivating individual nuclei on performance retention operantly conditioned forelimb movement.J. Neurophysiol. 1997; 78: 939-959Crossref (56) 1975Thach Timing dentate nucleus cerebral cortex during prompt volitional movement.Brain Res. 1975; 88: 233-241Crossref (111) Scholar). However, circuit dynamics initiating remain unclear. Two distinct pathways could contribute cerebellar-rubrospinal tract (Asanuma 1983Asanuma C. Jones E.G. Brainstem spinal projections deep monkey, with observations brainstem dorsal column nuclei.Brain 1983; 286: 299-322Crossref Gibson 1985Gibson A.R. Houk J.C. Kohlerman N.J. Magnocellular red different types limb macaque monkey.J. Physiol. 1985; 358: 527-549Crossref (129) Teune 2000Teune T.M. van der Burg J. Moer Voogd Ruigrok T.J. Topography nuclear brain stem rat.Prog. Brain 2000; 124: 141-172Crossref Scholar) (Bostan 2013Bostan A.C. Dum R.P. Strick P.L. networks basal ganglia.Trends Cogn. Sci. 2013; 17: 241-254Abstract Full Text PDF (390) Gornati 2018Gornati S.V. Schäfer C.B. Eelkman Rooda O.H.J. Nigg A.L. De Zeeuw C.I. Hoebeek F.E. Differentiating impact nuclei.Cell Rep. 2018; 23: 2690-2704Abstract (20) Horne Butler, 1995Horne M.K. Butler role cerebello-thalamo-cortical skilled movement.Prog. Neurobiol. 46: 199-213Crossref (70) Kuramoto 2009Kuramoto E. Furuta T. Nakamura K.C. Unzai Hioki H. Kaneko rat: single neuron-tracing study using viral vectors.Cereb. Cortex. 2009; 19: 2065-2077Crossref (152) Nashef 2019Nashef A. Cohen O. Harel R. Israel Z. Prut Reversible Block Outflow Reveals Cortical Circuitry Motor Coordination.Cell 2019; 27: 2608-2619.e4Abstract (10) latter supported neuronal dentate/interpositus (DN/IPN) preceding (Nashef 2018Nashef Shaping Firing Is Correlated Actions.Cell 1275-1285Abstract Strick, 1976Strick Activity ventrolateral neurons arm 1976; 39: 1032-1044Crossref (118) Scholar, 2014Thach Does initiate movement?.Cerebellum. 2014; 13: 139-150Crossref (24) (Anderson Turner, 1991Anderson M.E. Turner R.S. cerebellar-receiving pallidal-receiving behaving 1991; 66: 879-893Crossref 1992Butler Hawkins monkey neurones skilled, ballistic 1992; 445: 25-48Crossref (31) 1996Butler Finkelstein D.I. Harvey M.C. Churchward P.R. Forlano L.M. relationship between kinematic parameters wrist 1996; 736: 146-159Crossref (11) Fortier 1989Fortier P.A. Kalaska J.F. Smith A.M. related whole-arm reaching 1989; 198-211Crossref (158) 1979Harvey R.J. Porter Rawson J.A. Discharges intracerebellar cells monkeys.J. 1979; 297: 559-580Crossref (55) Porter, 1980Horne discharges conscious 1980; 304: 349-372Crossref (40) Kurata, 2005Kurata K. properties location project 2005; 94: 550-566Crossref (32) Macpherson 1980Macpherson J.M. Rasmusson D.D. Murphy J.T. Activities “motor” control primate.J. 44: 11-28Crossref (35) Mushiake 1993Mushiake Preferential guided vision.J. 1993; 70: 2660-2664Crossref Schmied 1979Schmied Benita M. Conde Dormont relation simple reaction time task cat.Exp. 36: 285-300Crossref (26) Donkelaar 1999van P. Stein Passingham R.E. Miall R.C. Neuronal primate visually triggered internally generated movements.J. 1999; 82: 934-945Crossref (79) Scholar), disrupting either region alters sensory-triggered (Meyer-Lohmann 1977Meyer-Lohmann Hore Brooks V.B. participation generation 1977; 40: 1038-1050Crossref (101) Spidalieri 1983Spidalieri Busby L. Lamarre Fast visual, auditory, somesthetic stimuli monkey. II. unilateral lesion discharge precentral time.J. 50: 1359-1379Crossref (43) 2000van Temporary inactivation 83: 2780-2790Crossref Beyond proposed coordinate accuracy ongoing given focal endpoint (dysmetria/hypermetria), reach path curvature grasping (Becker Person, 2019Becker M.I. Person Control Reach Kinematics Endpoint Precision.Neuron. 103: 335-348.e5Abstract 1999Bracha Kolb F.P. Irwin K.B. Inactivation interposed cat: classically withdrawal reflexes, action primitive hypothesis.Exp. 126: 77-92Crossref (36) Cooper 2000Cooper S.E. Martin J.H. Ghez anterior interpositus dynamic multijoint 84: 1988-2000Crossref Ishikawa 2014Ishikawa Tomatsu S. Tsunoda Lee Hoffman D.S. Kakei Releasing Purkinje cell inhibition generates cerebrocerebellum.PLoS ONE. 9: e108774Crossref 2000Martin Hacking Differential effects adaptive control.J. 1886-1899Crossref Mason 1998Mason C.R. Miller L.E. Baker Organization revealed muscimol inactivations.J. 1998; 79: 537-554Crossref (73) Thanawalla 2020Thanawalla Chen A.I. Azim Nuclei Dexterous Limb Movements.Neuroscience. 2020; 450: 168-183Crossref loss anticipatory adjustments unexpected obstacles complex locomotion (Martin In contrast, fastigial results posture, locomotion, planning, minimal (Li 2015Li N. T.W. Guo Z.V. Gerfen Svoboda A planning movement.Nature. 2015; 519: 51-56Crossref Bastian, 2004Thach Bastian Role adaptation gait health disease.Prog. 2004; 143: 353-366Crossref (80) Thus, provide differing contributions control, where likely convey signals order modify (Kurata, rodents, ventral thalamus. primarily targets ventromedial (VM), target anteromedial (AM), anterolateral (VAL) subdivisions (Angaut 1985Angaut Cicirata F. Serapide Topographic organization cerebellothalamic rat. An autoradiographic study.Neuroscience. 15: 389-401Crossref (60) Haroian 1981Haroian Massopust L.C. Young Cerebellothalamic degeneration study.J. Comp. 1981; 197: 217-236Crossref (115) axon terminal fields overlap substantially displaying morphological functional characteristics consistent strong driver inputs, such large synaptic boutons (Aumann Horne, 1996aAumann T.D. ultrastructure synapses cerebello-rubral cerebello-thalamic rat.Neurosci. Lett. 211: 175-178Crossref (14) Aumann 1996bAumann Ramification termination axons rat.J. 376: 420-430Crossref (37) 1994Aumann Projections lateral light electron microscopic double anterograde labelling.J. 1994; 349: 165-181Crossref (74) unitary (Gornati Sawyer 1994Sawyer S.F. S.J. Groves P.M. Tepper Cerebellar-responsive ventroanterior-ventrolateral rats: vivo electrophysiology.Neuroscience. 63: 711-724Crossref 2021Schäfer Gao Temporal cerebello-cortical convergence ventro-lateral thalamus.J. 2021; 599: 2055-2073Crossref input short-latency spiking superficial layers (Hooks 2013Hooks B.M. Mao Gutnisky D.A. Yamawaki Shepherd G.M. pyramidal mouse cortex.J. Neurosci. 33: 748-760Crossref (160) transforming top-down excitation through layer 2/3 (Weiler 2008Weiler Wood Yu Solla S.A. Top-down laminar excitatory network cortex.Nat. 2008; 11: 360-366Crossref (192) direct 5 Sauerbrei 2020Sauerbrei B.A. J.Z. J.D. Mischiati W. Kabra Verma Mensh B. Branson Hantman A.W. pattern dexterous input-driven.Nature. 577: 386-391Crossref (34) Key remaining questions whether plays this dependent input. To address questions, developed cued lever push mice requiring basic stimulus-response behavior reward. This habitual depends antecedent rather than goal value, recruiting feedback reinforcement circuits, including VAL (Balleine, 2019Balleine B.W. meaning behavior: discriminating reflex volition brain.Neuron. 104: 47-62Abstract (21) Graybiel, 2008Graybiel Habits, rituals, evaluative brain.Annu. Rev. 31: 359-387Crossref (989) Using imaging, electrophysiology, gain- loss-of-function experiments, cue DN/IPN-recipient were dominated providing signal cortex. Consistent view, membrane potential 5B projection matched pre-movement activation, suppressing blocked Conversely, photostimulation demonstrate important causal investigate mice. design required execute horizontal (4 mm) after randomized inter-trial interval (4–6 s) 6-kHz cue. Miss trials, partial pushes, spontaneous resulted no reward reset (Figure 1A; Video S1). Mice rapidly (mean = 7.5 days [6.3, 8.6] 95% confidence [CI], N 16 mice; all data unless otherwise stated presented mean [bootstrapped CI]; percentage successful trials [last session], 63.7% [56.0, 71.7]), moderate times (RTs; last session median 0.32 s [0.30, 0.34]) reproducible trajectories (Figures 1B–1E; Even expert mice, observed miss reflecting changing attention satiation within sessions 1E; https://www.cell.com/cms/asset/c3c41015-621d-43e8-823f-368e7e1617d3/mmc2.mp4Loading ... Download .mp4 (6.74 MB) Help files S1. Cued task, Figure 1 Since both DN IPN implicated send glutamatergic Bosch-Bouju 2013Bosch-Bouju Hyland B.I. Parr-Brownlie integration cortical, ganglia information: implications normal parkinsonian conditions.Front. Comput. 7: 163Crossref (102) sought define receives projects caudal area (CFA) dual labeling strategy 1F). dense centered AM nuclei, sparse colocalization posteromedial (VPM). We found which (Gao 2018Gao Davis Thomas Economo M.N. Abrego Li cortico-cerebellar loop planning.Nature. 563: 113-116Crossref (97) 1G S1A–S1E). Although injections targeted DN/IPN, low-level expression was detected some adjacent vestibular do not S1A S1C). Within overlap, ∼76% CFA, CFA-projecting received (vesicular glutamate transporter 2 [VGluT2] positive) (Bosch-Bouju Rovó 2012Rovó Ulbert I. Acsády Drivers 2012; 32: 17894-17908Crossref (62) 2020Schäfer Cerebello-Thalamic Spike Transfer Coding Adaptation.bioRxiv. https://doi.org/10.1101/2020.01.19.911610Crossref 1H S2A–S2D). degree connectivity (MThDN/IPN) being node connecting CFA. explore MThDN/IPN population employed gradient-index (GRIN) lens-mediated two-photon calcium imaging 1I 1J). Lens implantation above did affect overall when compared (control versus GRIN lens-implanted mice: two-sample Kolmogorov-Smirnov test, RT p 0.56, duration 0.22, number pushes 0.35, 23 9 shown). Most displayed push-related (210/248 neurons) (early-onset ΔF/F0, 104/210 neurons; early-onset decrease 32/210 execution/reward period (late-onset 47/210 late-onset 27/210 neurons), absent (11 view [FOVs], 8 mice) 1K–1L). Increased appeared dominant neurons: increased activity, 76.4%; decreased 23.5%) 1L) across extent 1M). If signal, then be described three hypothetical models. First, rising before onset. regard, increasing dictates (model i). Second, rises maintained until additional inputs trigger movement. contributes to, does dictate, ii). Third, reflects continuous sensorimotor slope iii) 2A). distinguish models, grouped short, medium, long RTs aligned trial-averaged ΔF/F0 focusing early 2B). Changes occurred irrespective (median onsets: short RT, −267 ms [−361, −178] CI; medium −276 [−374, −177] −367 [−464, −271] CI, n 104 neurons/9 FOVs, 6 0.46, one-way ANOVA). During RTs, baseline upon presentation, 2C 2D). Response profiles trial (i.e., model i) 2E–2G). Organiza