Representation learning in neural networks may be implemented with supervised or unsupervised algorithms, distinguished by the availability of instruction. In the sensory cortex, perceptual learning drives neural plasticity1-13, but it is not known whether this is due to supervised or unsupervised learning. Here we recorded populations of up to 90,000 neurons simultaneously from the primary visual cortex (V1) and higher visual areas (HVAs) while mice learned multiple tasks, as well as during unrewarded exposure to the same stimuli. Similar to previous studies, we found that neural changes in task mice were correlated with their behavioural learning. However, the neural changes were mostly replicated in mice with unrewarded exposure, suggesting that the changes were in fact due to unsupervised learning. The neural plasticity was highest in the medial HVAs and obeyed visual, rather than spatial, learning rules. In task mice only, we found a ramping reward-prediction signal in anterior HVAs, potentially involved in supervised learning. Our neural results predict that unsupervised learning may accelerate subsequent task learning, a prediction that we validated with behavioural experiments.

Preprint: https://www.biorxiv.org/content/early/2024/02/27/2024.02.25.581990

Tumors are complex ecosystems composed of malignant and non-malignant cells embedded in a dynamic extracellular matrix (ECM). In the tumor microenvironment, molecular phenotypes are controlled by cell-cell and ECM interactions in 3D cellular neighborhoods (CNs). While their inhibition can impede tumor progression, routine molecular tumor profiling fails to capture cellular interactions. Single-cell spatial transcriptomics (ST) maps receptor-ligand interactions but usually remains limited to 2D tissue sections and lacks ECM readouts. Here, we integrate 3D ST with ECM imaging in serial sections from one clinical lung carcinoma to systematically quantify molecular states, cell-cell interactions, and ECM remodeling in CN. Our integrative analysis pinpointed known immune escape and tumor invasion mechanisms, revealing several druggable drivers of tumor progression in the patient under study. This proof-of-principle study highlights the potential of in-depth CN profiling in routine clinical samples to inform microenvironment-directed therapies. A record of this paper's transparent peer review process is included in the supplemental information.

Cognitive maps confer animals with flexible intelligence by representing spatial, temporal and abstract relationships that can be used to shape thought, planning and behaviour. Cognitive maps have been observed in the hippocampus1, but their algorithmic form and learning mechanisms remain obscure. Here we used large-scale, longitudinal two-photon calcium imaging to record activity from thousands of neurons in the CA1 region of the hippocampus while mice learned to efficiently collect rewards from two subtly different linear tracks in virtual reality. Throughout learning, both animal behaviour and hippocampal neural activity progressed through multiple stages, gradually revealing improved task representation that mirrored improved behavioural efficiency. The learning process involved progressive decorrelations in initially similar hippocampal neural activity within and across tracks, ultimately resulting in orthogonalized representations resembling a state machine capturing the inherent structure of the task. This decorrelation process was driven by individual neurons acquiring task-state-specific responses (that is, 'state cells'). Although various standard artificial neural networks did not naturally capture these dynamics, the clone-structured causal graph, a hidden Markov model variant, uniquely reproduced both the final orthogonalized states and the learning trajectory seen in animals. The observed cellular and population dynamics constrain the mechanisms underlying cognitive map formation in the hippocampus, pointing to hidden state inference as a fundamental computational principle, with implications for both biological and artificial intelligence.

Genetically encoded voltage indicators (GEVIs) allow optical recording of membrane potential from targeted cells in vivo. However, red GEVIs that are compatible with two-photon microscopy and that can be multiplexed in vivo with green reporters like GCaMP, are currently lacking. To address this gap, we explored diverse rhodopsin proteins as GEVIs and engineered a novel GEVI, 2Photron, based on a rhodopsin from the green algae Klebsormidium nitens. 2Photron, combined with two photon ultrafast local volume excitation (ULoVE), enabled multiplexed readout of spiking and subthreshold voltage simultaneously with GCaMP calcium signals in visual cortical neurons of awake, behaving mice. These recordings revealed the cell-specific relationship of spiking and subthreshold voltage dynamics with GCaMP responses, highlighting the challenges of extracting underlying spike trains from calcium imaging.

The interplay between two major forebrain structures - cortex and subcortical striatum - is critical for flexible, goal-directed action. Traditionally, it has been proposed that striatum is critical for selecting what type of action is initiated while the primary motor cortex is involved in the online control of movement execution. Recent data indicates that striatum may also be critical for specifying movement execution. These alternatives have been difficult to reconcile because when comparing very distinct actions, as in the vast majority of work to date, they make essentially indistinguishable predictions. Here, we develop quantitative models to reveal a somewhat paradoxical insight: only comparing neural activity during similar actions makes strongly distinguishing predictions. We thus developed a novel reach-to-pull task in which mice reliably selected between two similar, but distinct reach targets and pull forces. Simultaneous cortical and subcortical recordings were uniquely consistent with a model in which cortex and striatum jointly specify flexible parameters of action during movement execution.

Foraging animals must use decision-making strategies that dynamically adapt to the changing availability of rewards in the environment. A wide diversity of animals do this by distributing their choices in proportion to the rewards received from each option, Herrnstein’s operant matching law. Theoretical work suggests an elegant mechanistic explanation for this ubiquitous behavior, as operant matching follows automatically from simple synaptic plasticity rules acting within behaviorally relevant neural circuits. However, no past work has mapped operant matching onto plasticity mechanisms in the brain, leaving the biological relevance of the theory unclear. Here we discovered operant matching in Drosophila and showed that it requires synaptic plasticity that acts in the mushroom body and incorporates the expectation of reward. We began by developing a novel behavioral paradigm to measure choices from individual flies as they learn to associate odor cues with probabilistic rewards. We then built a model of the fly mushroom body to explain each fly’s sequential choice behavior using a family of biologically-realistic synaptic plasticity rules. As predicted by past theoretical work, we found that synaptic plasticity rules could explain fly matching behavior by incorporating stimulus expectations, reward expectations, or both. However, by optogenetically bypassing the representation of reward expectation, we abolished matching behavior and showed that the plasticity rule must specifically incorporate reward expectations. Altogether, these results reveal the first synaptic level mechanisms of operant matching and provide compelling evidence for the role of reward expectation signals in the fly brain.

Recent success in training artificial agents and robots derives from a combination of direct learning of behavioural policies and indirect learning through value functions. Policy learning and value learning use distinct algorithms that optimize behavioural performance and reward prediction, respectively. In animals, behavioural learning and the role of mesolimbic dopamine signalling have been extensively evaluated with respect to reward prediction; however, so far there has been little consideration of how direct policy learning might inform our understanding. Here we used a comprehensive dataset of orofacial and body movements to understand how behavioural policies evolved as naive, head-restrained mice learned a trace conditioning paradigm. Individual differences in initial dopaminergic reward responses correlated with the emergence of learned behavioural policy, but not the emergence of putative value encoding for a predictive cue. Likewise, physiologically calibrated manipulations of mesolimbic dopamine produced several effects inconsistent with value learning but predicted by a neural-network-based model that used dopamine signals to set an adaptive rate, not an error signal, for behavioural policy learning. This work provides strong evidence that phasic dopamine activity can regulate direct learning of behavioural policies, expanding the explanatory power of reinforcement learning models for animal learning.

To accurately track self-location, animals need to integrate their movements through space. In amniotes, representations of self-location have been found in regions such as the hippocampus. It is unknown whether more ancient brain regions contain such representations and by which pathways they may drive locomotion. Fish displaced by water currents must prevent uncontrolled drift to potentially dangerous areas. We found that larval zebrafish track such movements and can later swim back to their earlier location. Whole-brain functional imaging revealed the circuit enabling this process of positional homeostasis. Position-encoding brainstem neurons integrate optic flow, then bias future swimming to correct for past displacements by modulating inferior olive and cerebellar activity. Manipulation of position-encoding or olivary neurons abolished positional homeostasis or evoked behavior as if animals had experienced positional shifts. These results reveal a multiregional hindbrain circuit in vertebrates for optic flow integration, memory of self-location, and its neural pathway to behavior.Competing Interest StatementThe authors have declared no competing interest.

Internal representations are thought to support the generation of flexible, long-timescale behavioral patterns in both animals and artificial agents. Here, we present a novel conceptual framework for how Drosophila use their internal representation of head direction to maintain preferred headings in their surroundings, and how they learn to modify these preferences in the presence of selective thermal reinforcement. To develop the framework, we analyzed flies’ behavior in a classical operant visual learning paradigm and found that they use stochastically generated fixations and directed turns to express their heading preferences. Symmetries in the visual scene used in the paradigm allowed us to expose how flies’ probabilistic behavior in this setting is tethered to their head direction representation. We describe how flies’ ability to quickly adapt their behavior to the rules of their environment may rest on a behavioral policy whose parameters are flexible but whose form is genetically encoded in the structure of their circuits. Many of the mechanisms we outline may also be relevant for rapidly adaptive behavior driven by internal representations in other animals, including mammals.

Chemical neurotransmission constitutes one of the fundamental modalities of communication between neurons. Monitoring release of these chemicals has traditionally been difficult to carry out at spatial and temporal scales relevant to neuron function. To understand chemical neurotransmission more fully, we need to improve the spatial and temporal resolutions of measurements for neurotransmitter release. To address this, we engineered a chemi-sensitive, two-dimensional nanofilm that facilitates subcellular visualization of the release and diffusion of the neurochemical dopamine with synaptic resolution, quantal sensitivity, and simultaneously from hundreds of release sites. Using this technology, we were able to monitor the spatiotemporal dynamics of dopamine release in dendritic processes, a poorly understood phenomenon. We found that dopamine release is broadcast from a subset of dendritic processes as hotspots that have a mean spatial spread of ≈3.2 µm (full width at half maximum) and are observed with a mean spatial frequency of 1 hotspot per ≈7.5 µm of dendritic length. Major dendrites of dopamine neurons and fine dendritic processes, as well as dendritic arbors and dendrites with no apparent varicose morphology participated in dopamine release. Remarkably, these release hotspots colocalized with Bassoon, suggesting that Bassoon may contribute to organizing active zones in dendrites, similar to its role in axon terminals.