Disengagement of Motor Cortex during Long-Term Learning Tracks the Performance Level of Learned Movements

Motor learning refines thalamic influence on motor cortex

The primary motor cortex (M1) is central for the learning and execution of dexterous motor skills1-3, and its superficial layer (layers 2 and 3; hereafter, L2/3) is a key locus of learning-related plasticity1,4-6. It remains unknown how motor learning shapes the way in which upstream regions activate M1 circuits to execute learned movements. Here, using longitudinal axonal imaging of the main inputs to M1 L2/3 in mice, we show that the motor thalamus is the key input source that encodes learned movements in experts (animals trained for two weeks). We then use optogenetics to identify the subset of M1 L2/3 neurons that are strongly driven by thalamic inputs before and after learning. We find that the thalamic influence on M1 changes with learning, such that the motor thalamus preferentially activates the M1 neurons that encode learned movements in experts. Inactivation of the thalamic inputs to M1 in experts impairs learned movements. Our study shows that motor learning reshapes the thalamic influence on M1 to enable the reliable execution of learned movements.

An Open-Source Joystick Platform for Investigating Forelimb Motor Control, Auditory-Motor Integration, and Value-Based Decision-Making in Head-Fixed Mice

Investigation of neural processes underlying motor control requires behavioral readouts that capture the richness of actions, including both categorical (choice-based) information and motor execution (kinematics). We present an open-source platform for behavioral training of head-fixed mice that combines a stationary or retractable forelimb-based joystick, sound-presentation system, capacitive lick sensor, and water reward dispenser. The setup allows for the creation of multiple behavioral paradigms, two of which are highlighted here: a two-alternative forced-choice auditory-motor discrimination paradigm and a two-armed bandit value-based decision-making task. In the auditory-motor paradigm, mice learn to report high- or low-frequency tones by pushing or pulling the joystick. In the value-based paradigm, mice learn to push or pull the joystick based on the history of rewarded trials. In addition to reporting categorical choices, this setup provides a rich dataset of motor parameters that reflect components of the underlying learning and decision processes in both of these tasks. These kinematic parameters (including joystick speed and displacement, Fréchet similarity of trajectories, tortuosity, angular standard deviation, and movement vigor) provide key additional insights into the motor execution of choices that are not as readily assessed in other paradigms. The system's flexibility of task design, joystick readout, and ease of construction represent an advance compared with currently available manipulandum tasks in mice. We provide detailed schematics for constructing the setup and protocols for behavioral training using both paradigms, with the hope that this open-source resource is readily adopted by neuroscientists interested in mechanisms of sensorimotor integration, motor control, and choice behavior.

An Open-Source Joystick Platform for Investigating Forelimb Motor Control, Auditory-Motor Integration, and Value-Based Decision-Making in Head-Fixed Mice

Investigation of neural processes underlying motor control requires behavioral readouts that capture the richness of actions, including both categorical (choice-based) information and motor execution (kinematics). We present an open-source platform for behavioral training of head-fixed mice that combines a stationary or retractable forelimb-based joystick, sound-presentation system, capacitive lick sensor, and water reward dispenser. The setup allows for the creation of multiple behavioral paradigms, two of which are highlighted here: a two-alternative forced-choice auditory-motor discrimination paradigm, and a two-armed bandit value-based decision-making task. In the auditory-motor paradigm, mice learn to report high or low frequency tones by pushing or pulling the joystick. In the value-based paradigm, mice learn to push or pull the joystick based on the history of rewarded trials. In addition to reporting categorical choices, this setup provides a rich dataset of motor parameters that reflect components of the underlying learning and decision processes in both of these tasks. These kinematic parameters (including joystick speed and displacement, Fréchet similarity of trajectories, tortuosity, angular standard deviation, and movement vigor) provide key additional insights into the motor execution of choices that are not as readily assessed in other paradigms. The system's flexibility of task design, joystick readout, and ease of construction represent an advance compared to currently available manipulandum tasks in mice. We provide detailed schematics for constructing the setup and protocols for behavioral training using both paradigms, with the hope that this open-source resource is readily adopted by neuroscientists interested in mechanisms of sensorimotor integration, motor control, and choice behavior.

Significance Statement

Behavioral paradigms for experiments in head-restrained mice are important for investigating the relationship between neural activity and behavior. However, behavioral setups are often constrained by high cost, design complexity, and implementation challenges. Here, we present an open-source platform for behavioral training of head-fixed mice using a joystick manipulandum. The setup allows for the creation of multiple behavioral paradigms, including an auditory-motor discrimination paradigm, and a value-based decision-making task. We include detailed instructions for construction and implementation of the entire open-source behavioral platform.

Developmental timecourse of aptitude for motor skill learning in mouse

Learning motor skills requires plasticity in the primary motor cortex (M1). But the capacity for cortical circuit plasticity varies over developmental age in sensory cortex. This study assesses the normal developmental trajectory of motor learning to assess how aptitude might vary with age. We trained mice of both sexes to run on a custom accelerating rotarod at ages from postnatal day (P) 20 to P120, tracking paw position and quantifying time to fall and changes in gait pattern. While animals of all ages were able to perform better after five training sessions, performance improved most rapidly on the first training day for mice between ages P30-60, suggesting an age with heightened plasticity. Learning this task required M1, because pharmacological inactivation of M1 prevented improvement in task performance. Paw position and gait patterns changed with learning, though differently between age groups. Successful mice learned to shift their gait from hopping to walking. Notably, this shift in gait happened earlier in the trial for forelimbs in comparison to hindlimbs. Thus, motor plasticity might more readily occur in forelimbs. Changes in gait and other kinematic parameters are an additional learning metric beyond time to fall, offering insight into how mice improve performance. Overall, these results suggest mouse motor learning has a developmental trajectory.

Dynamic lateralization in contralateral-projecting corticospinal neurons during motor learning

Understanding the adaptability of the motor cortex in response to bilateral motor tasks is crucial for advancing our knowledge of neural plasticity and motor learning. Here we aim to investigate the dynamic lateralization of contralateral-projecting corticospinal neurons (cpCSNs) during such tasks. Utilizing in vivo two-photon calcium imaging, we observe cpCSNs in mice performing a "left-right" lever-press task. Our findings reveal heterogeneous populational dynamics in cpCSNs: a marked decrease in activity during ipsilateral motor learning, in contrast to maintained activity during contralateral motor learning. Notably, individual cpCSNs show dynamic shifts in engagement with ipsilateral and contralateral movements, displaying an evolving pattern of activation over successive days. It suggests that cpCSNs exhibit adaptive changes in activation patterns in response to ipsilateral and contralateral movements, highlighting a flexible reorganization during motor learning This reconfiguration underscores the dynamic nature of cortical lateralization in motor learning and offers insights for neuromotor rehabilitation.

Motor cortical inactivation impairs corrective submovements in mice performing a hold-still center-out reach task

Holding still and aiming reaches to spatial targets may depend on distinct neural circuits. Using automated homecage training and a sensitive joystick, we trained freely moving mice to contact a joystick, hold their forelimb still, and then reach to rewarded target locations. Mice learned the task by initiating forelimb sequences with clearly resolved submillimeter-scale micromovements followed by millimeter-scale reaches to learned spatial targets. Hundreds of thousands of trajectories were decomposed into millions of kinematic submovements, while photoinhibition was used to test roles of motor cortical areas. Inactivation of both caudal and rostral forelimb areas preserved the ability to produce aimed reaches, but reduced reach speed. Inactivation specifically of contralateral caudal forelimb area (CFA) additionally impaired the ability to aim corrective submovements to remembered locations following target undershoots. Our findings show that motor cortical inactivations reduce the gain of forelimb movements but that inactivation specifically of contralateral CFA impairs corrective movements important for reaching a target location.NEW & NOTEWORTHY To test the role of different cortical areas in holding still and reaching to targets, this study combined home-cage training with optogenetic silencing as mice engaged in a learned center-out-reach task. Inactivation specifically of contralateral caudal forelimb area (CFA) impaired corrective movements necessary to reach spatial targets to earn reward.

Goal-directed learning is multidimensional and accompanied by diverse and widespread changes in neocortical signaling

New tasks are often learned in stages with each stage reflecting a different learning challenge. Accordingly, each learning stage is likely mediated by distinct neuronal processes. And yet, most rodent studies of the neuronal correlates of goal-directed learning focus on individual outcome measures and individual brain regions. Here, we longitudinally studied mice from naïve to expert performance in a head-fixed, operant conditioning whisker discrimination task. In addition to tracking the primary behavioral outcome of stimulus discrimination, we tracked and compared an array of object-based and temporal-based behavioral measures. These behavioral analyses identify multiple, partially overlapping learning stages in this task, consistent with initial response implementation, early stimulus-response generalization, and late response inhibition. To begin to understand the neuronal foundations of these learning processes, we performed widefield Ca2+ imaging of dorsal neocortex throughout learning and correlated behavioral measures with neuronal activity. We found distinct and widespread correlations between neocortical activation patterns and various behavioral measures. For example, improvements in sensory discrimination correlated with target stimulus evoked activations of response-related cortices along with distractor stimulus evoked global cortical suppression. Our study reveals multidimensional learning for a simple goal-directed learning task and generates hypotheses for the neuronal modulations underlying these various learning processes.

Selective consolidation of learning and memory via recall-gated plasticity

In a variety of species and behavioral contexts, learning and memory formation recruits two neural systems, with initial plasticity in one system being consolidated into the other over time. Moreover, consolidation is known to be selective; that is, some experiences are more likely to be consolidated into long-term memory than others. Here, we propose and analyze a model that captures common computational principles underlying such phenomena. The key component of this model is a mechanism by which a long-term learning and memory system prioritizes the storage of synaptic changes that are consistent with prior updates to the short-term system. This mechanism, which we refer to as recall-gated consolidation, has the effect of shielding long-term memory from spurious synaptic changes, enabling it to focus on reliable signals in the environment. We describe neural circuit implementations of this model for different types of learning problems, including supervised learning, reinforcement learning, and autoassociative memory storage. These implementations involve synaptic plasticity rules modulated by factors such as prediction accuracy, decision confidence, or familiarity. We then develop an analytical theory of the learning and memory performance of the model, in comparison to alternatives relying only on synapse-local consolidation mechanisms. We find that recall-gated consolidation provides significant advantages, substantially amplifying the signal-to-noise ratio with which memories can be stored in noisy environments. We show that recall-gated consolidation gives rise to a number of phenomena that are present in behavioral learning paradigms, including spaced learning effects, task-dependent rates of consolidation, and differing neural representations in short- and long-term pathways.

Rethinking dopamine-guided action sequence learning

As opposed to those requiring a single action for reward acquisition, tasks necessitating action sequences demand that animals learn action elements and their sequential order and sustain the behaviour until the sequence is completed. With repeated learning, animals not only exhibit precise execution of these sequences but also demonstrate enhanced smoothness and efficiency. Previous research has demonstrated that midbrain dopamine and its major projection target, the striatum, play crucial roles in these processes. Recent studies have shown that dopamine from the substantia nigra pars compacta (SNc) and the ventral tegmental area (VTA) serve distinct functions in action sequence learning. The distinct contributions of dopamine also depend on the striatal subregions, namely the ventral, dorsomedial and dorsolateral striatum. Here, we have reviewed recent findings on the role of striatal dopamine in action sequence learning, with a focus on recent rodent studies.

Integrating across behaviors and timescales to understand the neural control of movement

The nervous system evolved to enable navigation throughout the environment in the pursuit of resources. Evolutionarily newer structures allowed increasingly complex adaptations but necessarily added redundancy. A dominant view of movement neuroscientists is that there is a one-to-one mapping between brain region and function. However, recent experimental data is hard to reconcile with the most conservative interpretation of this framework, suggesting a degree of functional redundancy during the performance of well-learned, constrained behaviors. This apparent redundancy likely stems from the bidirectional interactions between the various cortical and subcortical structures involved in motor control. We posit that these bidirectional connections enable flexible interactions across structures that change depending upon behavioral demands, such as during acquisition, execution or adaptation of a skill. Observing the system across both multiple actions and behavioral timescales can help isolate the functional contributions of individual structures, leading to an integrated understanding of the neural control of movement.

Local and global predictors of synapse elimination during motor learning

During learning, synaptic connections between excitatory neurons in the brain display considerable dynamism, with new connections being added and old connections eliminated. Synapse elimination offers an opportunity to understand the features of synapses that the brain deems dispensable. However, with limited observations of synaptic activity and plasticity in vivo, the features of synapses subjected to elimination remain poorly understood. Here, we examined the functional basis of synapse elimination in the apical dendrites of L2/3 neurons in the primary motor cortex throughout motor learning. We found no evidence that synapse elimination is facilitated by a lack of activity or other local forms of plasticity. Instead, eliminated synapses display asynchronous activity with nearby synapses, suggesting that functional synaptic clustering is a critical component of synapse survival. In addition, eliminated synapses show delayed activity timing with respect to postsynaptic output. Thus, synaptic inputs that fail to be co-active with their neighboring synapses or are mistimed with neuronal output are targeted for elimination.

Determination of Neuronal Activity and Its Meaning for the Processes of Learning and Memory

Despite the years of studies in the field of systems neuroscience, the functions of neural circuits and behavior-related systems are still not entirely understood. The systems description of brain activity has recently been associated with cognitive concepts, e.g., a cognitive map, reconstructed via place-cell activity analysis and the like, and a cognitive schema, modeled in consolidation research. The issue we find of importance is that cognitive concepts reconstructed in neuroscience research are mainly formulated in terms of the environment. In this chapter, we present the idea of an element of individual experience that serves as a model of behavioral interaction with the environment, rather than a model of the environment itself. This intangible difference entails the need for substantial revision of several well-known phenomena, including long-term potentiation. The principal questions we address are: how do elements of experience appear and change during learning and performance; and how do the links between these elements create the whole structure of individual experience? We argue that learning and memory research need a clear distinction between processes that provide the emergence of new elements of experience (functional systems) and processes underlying the retrieval and/or changes in the existing experience. We propose to view the activity of a neuron as an "action" directed to the future adaptive "microresult," essential for meeting its metabolic needs. This anticipatory neuronal activity is coordinated with the activity of many other cells of the body in the organism-wide functional system, ensuring the achievement of an adaptive "macroresult" at the behavioral level. From this perspective, the mechanisms of learning are considered as the formation of functional systems, and memory is considered as a dynamic structure constituted by systems formed at different stages of individual development.

Cortico-spinal modularity in the parieto-frontal system: A new perspective on action control

Classical neurophysiology suggests that the motor cortex (MI) has a unique role in action control. In contrast, this review presents evidence for multiple parieto-frontal spinal command modules that can bypass MI. Five observations support this modular perspective: (i) the statistics of cortical connectivity demonstrate functionally-related clusters of cortical areas, defining functional modules in the premotor, cingulate, and parietal cortices; (ii) different corticospinal pathways originate from the above areas, each with a distinct range of conduction velocities; (iii) the activation time of each module varies depending on task, and different modules can be activated simultaneously; (iv) a modular architecture with direct motor output is faster and less metabolically expensive than an architecture that relies on MI, given the slow connections between MI and other cortical areas; (v) lesions of the areas composing parieto-frontal modules have different effects from lesions of MI. Here we provide examples of six cortico-spinal modules and functions they subserve: module 1) arm reaching, tool use and object construction; module 2) spatial navigation and locomotion; module 3) grasping and observation of hand and mouth actions; module 4) action initiation, motor sequences, time encoding; module 5) conditional motor association and learning, action plan switching and action inhibition; module 6) planning defensive actions. These modules can serve as a library of tools to be recombined when faced with novel tasks, and MI might serve as a recombinatory hub. In conclusion, the availability of locally-stored information and multiple outflow paths supports the physiological plausibility of the proposed modular perspective.

Emergence of cortical network motifs for short-term memory during learning

Learning of adaptive behaviors requires the refinement of coordinated activity across multiple brain regions. However, how neural communications develop during learning remains poorly understood. Here, using two-photon calcium imaging, we simultaneously recorded the activity of layer 2/3 excitatory neurons in eight regions of the mouse dorsal cortex during learning of a delayed-response task. Across learning, while global functional connectivity became sparser, there emerged a subnetwork comprising of neurons in the anterior lateral motor cortex (ALM) and posterior parietal cortex (PPC). Neurons in this subnetwork shared a similar choice code during action preparation and formed recurrent functional connectivity across learning. Suppression of PPC activity disrupted choice selectivity in ALM and impaired task performance. Recurrent neural networks reconstructed from ALM activity revealed that PPC-ALM interactions rendered choice-related attractor dynamics more stable. Thus, learning constructs cortical network motifs by recruiting specific inter-areal communication channels to promote efficient and robust sensorimotor transformation.

Cortical circuit dynamics underlying motor skill learning: from rodents to humans

Motor learning is crucial for the survival of many animals. Acquiring a new motor skill involves complex alterations in both local neural circuits in many brain regions and long-range connections between them. Such changes can be observed anatomically and functionally. The primary motor cortex (M1) integrates information from diverse brain regions and plays a pivotal role in the acquisition and refinement of new motor skills. In this review, we discuss how motor learning affects the M1 at synaptic, cellular, and circuit levels. Wherever applicable, we attempt to relate and compare findings in humans, non-human primates, and rodents. Understanding the underlying principles shared by different species will deepen our understanding of the neurobiological and computational basis of motor learning.

Task-specific modulation of corticospinal neuron activity during motor learning in mice

Motor skill learning relies on the plasticity of the primary motor cortex as task acquisition drives cortical motor network remodeling. Large-scale cortical remodeling of evoked motor outputs occurs during the learning of corticospinal-dependent prehension behavior, but not simple, non-dexterous tasks. Here we determine the response of corticospinal neurons to two distinct motor training paradigms and assess the role of corticospinal neurons in the execution of a task requiring precise modulation of forelimb movement and one that does not. In vivo calcium imaging in mice revealed temporal coding of corticospinal activity coincident with the development of precise prehension movements, but not more simplistic movement patterns. Transection of the corticospinal tract and optogenetic regulation of corticospinal activity show the necessity for patterned corticospinal network activity in the execution of precise movements but not simplistic ones. Our findings reveal a critical role for corticospinal network modulation in the learning and execution of precise motor movements.

Delay eyeblink conditioning performance and brain-wide c-Fos expression in male and female mice

Delay eyeblink conditioning has been extensively used to study associative learning and the cerebellar circuits underlying this task have been largely identified. However, there is a little knowledge on how factors such as strain, sex and innate behaviour influence performance during this type of learning. In this study, we used male and female mice of C57BL/6J (B6) and B6CBAF1 strains to investigate the effect of sex, strain and locomotion in delay eyeblink conditioning. We performed a short and a long delay eyeblink conditioning paradigm and used a c-Fos immunostaining approach to explore the involvement of different brain areas in this task. We found that both B6 and B6CBAF1 females reach higher learning scores compared to males in the initial stages of learning. This sex-dependent difference was no longer present as the learning progressed. Moreover, we found a strong positive correlation between learning scores and voluntary locomotion irrespective of the training duration. c-Fos immunostainings after the short paradigm showed positive correlations between c-Fos expression and learning scores in the cerebellar cortex and brainstem, as well as previously unreported areas. By contrast, after the long paradigm, c-Fos expression was only significantly elevated in the brainstem. Taken together, we show that differences in voluntary locomotion and activity across brain areas correlate with performance in delay eyeblink conditioning across strains and sexes.

Dorsolateral Striatum is a Bottleneck for Responding to Task-Relevant Stimuli in a Learned Whisker Detection Task in Mice

A learned sensory-motor behavior engages multiple brain regions, including the neocortex and the basal ganglia. How a target stimulus is detected by these regions and converted to a motor response remains poorly understood. Here, we performed electrophysiological recordings and pharmacological inactivations of whisker motor cortex and dorsolateral striatum to determine the representations within, and functions of, each region during performance in a selective whisker detection task in male and female mice. From the recording experiments, we observed robust, lateralized sensory responses in both structures. We also observed bilateral choice probability and preresponse activity in both structures, with these features emerging earlier in whisker motor cortex than dorsolateral striatum. These findings establish both whisker motor cortex and dorsolateral striatum as potential contributors to the sensory-to-motor (sensorimotor) transformation. We performed pharmacological inactivation studies to determine the necessity of these brain regions for this task. We found that suppressing the dorsolateral striatum severely disrupts responding to task-relevant stimuli, without disrupting the ability to respond, whereas suppressing whisker motor cortex resulted in more subtle changes in sensory detection and response criterion. Together these data support the dorsolateral striatum as an essential node in the sensorimotor transformation of this whisker detection task.SIGNIFICANCE STATEMENT Selecting an item in a grocery store, hailing a cab - these daily practices require us to transform sensory stimuli into motor responses. Many decades of previous research have studied goal-directed sensory-to-motor transformations within various brain structures, including the neocortex and the basal ganglia. Yet, our understanding of how these regions coordinate to perform sensory-to-motor transformations is limited because these brain structures are often studied by different researchers and through different behavioral tasks. Here, we record and perturb specific regions of the neocortex and the basal ganglia and compare their contributions during performance of a goal-directed somatosensory detection task. We find notable differences in the activities and functions of these regions, which suggests specific contributions to the sensory-to-motor transformation process.

Corticospinal neurons from motor and somatosensory cortices exhibit different temporal activity dynamics during motor learning

The ability to learn motor skills implicates an improvement in accuracy, speed and consistency of movements. Motor control is related to movement execution and involves corticospinal neurons (CSp), which are broadly distributed in layer 5B of the motor and somatosensory cortices. CSp neurons innervate the spinal cord and are functionally diverse. However, whether CSp activity differs between different cortical areas throughout motor learning has been poorly explored. Given the importance and interaction between primary motor (M1) and somatosensory (S1) cortices related to movement, we examined the functional roles of CSp neurons in both areas. We induced the expression of GCaMP7s calcium indicator to perform photometric calcium recordings from layer 5B CSp neurons simultaneously in M1 and S1 cortices and track their activity while adult mice learned and performed a cued lever-press task. We found that during early learning sessions, the population calcium activity of CSp neurons in both cortices during movement did not change significantly. In late learning sessions the peak amplitude and duration of calcium activity CSp neurons increased in both, M1 and S1 cortices. However, S1 and M1 CSp neurons display a different temporal dynamic during movements that occurred when animals learned the task; both M1 and S1 CSp neurons activate before movement initiation, however, M1 CSp neurons continue active during movement performance, reinforcing the idea of the diversity of the CSp system and suggesting that CSp neuron activity in M1 and S1 cortices throughout motor learning have different functional roles for sensorimotor integration.

Manipulation-specific cortical activity as mice handle food

Food handling offers unique yet largely unexplored opportunities to investigate how cortical activity relates to forelimb movements in a natural, ethologically essential, and kinematically rich form of manual dexterity. To determine these relationships, we recorded high-speed (1,000 fps) video and multi-channel electrophysiological cortical spiking activity while mice handled food. The high temporal resolution of the video allowed us to decompose active manipulation ("oromanual") events into characteristic submovements, enabling event-aligned analysis of cortical activity. Activity in forelimb M1 was strongly modulated during food handling, generally higher during oromanual events and lower during holding intervals. Optogenetic silencing and stimulation of forelimb M1 neurons partially affected food-handling movements, exerting suppressive and activating effects, respectively. We also extended the analysis to forelimb S1 and lateral M1, finding broadly similar oromanual-related activity across all three areas. However, each area's activity displayed a distinct timing and phasic/tonic temporal profile, which was further analyzed by non-negative matrix factorization and demonstrated to be attributable to area-specific composition of activity classes. Current or future forelimb position could be accurately predicted from activity in all three regions, indicating that the cortical activity in these areas contains high information content about forelimb movements during food handling. These results thus establish that cortical activity during food handling is manipulation specific, distributed, and broadly similar across multiple sensorimotor areas while also exhibiting area- and submovement-specific relationships with the fast kinematic hallmarks of this natural form of complex free-object-handling manual dexterity.

Learning-induced changes in the neural circuits underlying motor sequence execution

As the old adage goes: practice makes perfect. Yet, the neural mechanisms by which rote repetition transforms a halting behavior into a fluid, effortless, and "automatic" action are not well understood. Here we consider the possibility that well-practiced motor sequences, which initially rely on higher-level decision-making circuits, become wholly specified in lower-level control circuits. We review studies informing this idea, discuss the constraints on such shift in control, and suggest approaches to pinpoint circuit-level changes associated with motor sequence learning.

Movement-specific signaling is differentially distributed across motor cortex layer 5 projection neuron classes

Motor cortex generates descending output necessary for executing a wide range of limb movements. Although movement-related activity has been described throughout motor cortex, the spatiotemporal organization of movement-specific signaling in deep layers remains largely unknown. Here we record layer 5B population dynamics in the caudal forelimb area of motor cortex while mice perform a forelimb push/pull task and find that most neurons show movement-invariant responses, with a minority displaying movement specificity. Using cell-type-specific imaging, we identify that invariant responses dominate pyramidal tract (PT) neuron activity, with a small subpopulation representing movement type, whereas a larger proportion of intratelencephalic (IT) neurons display movement-type-specific signaling. The proportion of IT neurons decoding movement-type peaks prior to movement initiation, whereas for PT neurons, this occurs during movement execution. Our data suggest that layer 5B population dynamics largely reflect movement-invariant signaling, with information related to movement-type being routed through relatively small, distributed subpopulations of projection neurons.