Biology subjects

Kepecs, A.

Publications and source records attributed to Kepecs, A..

14 recordsLinked to original sources

bioRxiv2026

Parallel processing chains span cytoarchitectures to organize association cortex

Task fMRI1 and electrophysiology2 have revealed distributed, linked cortical patches with shared category preferences (e.g., faces, objects, places)1,3-5, smaller than cytoarchitectonic areas. Resting-state functional connectivity (RSFC) similarly showed that somato-cognitive action network (SCAN) nodes interleave with effectors (foot, hand, mouth), subdividing the precentral gyrus6. Here, using multiple precision functional mapping (PFM) modalities (RSFC, task, lags), we discovered that most of association cortex is organized like face processing and SCAN, with small, discrete patches interconnected into chains. Such patch-chains densely tile prefrontal cortex but are largely absent from primary cortex. Cortico-striatal connectivity is organized such that patches of the same chain connect to the same striatal location. Within chains, infra-slow fMRI signals are ordered in time. RSFC-defined chains align with task fMRI localizers (e.g., visual, motor, pain). Chains are absent at birth and emerge in the first year of life, suggesting their formation is at least partially experience-driven. Cytoarchitectonic areas are subdivided by patches, and patches in the same chain are distributed across different cytoarchitectures. Chains represent parallel ordered processing streams that are separated by information domain and behavioral goals, not cytoarchitectonics. Functional subdivision of architectonics into smaller patches, interlinked to form cross-architecture chains, enable greater parallelization and flexible specialization of processing.

neuroscience↗

bioRxiv2024

Bounded optimality of time investments in rats, mice, and humans

Time is our scarcest resource. Allocating time optimally presents a universal challenge for all organisms because the future benefits of time investments are uncertain. We developed a normative framework for assessing bounded optimality in time allocation, emphasizing the accuracy of future predictions, independent of subjective costs and benefits. In a common decision task across humans, rats, and mice, we varied uncertainty by titrating ambiguous sensory evidence and measured the time each subject was willing to invest post-decision. We observed that all species and subjects invested more time when they were more likely to be correct, which reflected a statistical confidence of uncertain evidence. Time allocation strategy approached the lower bound of optimality, indicating an accurate decision-by-decision assessment of confidence in the likelihood that waiting will pay off - independent of the subjective payoff values and time costs. We demonstrate that an elementary algorithm based on a drift-diffusion process algorithm can implement this optimal time investment strategy. These results illuminate the computational mechanisms governing rational time investment, showing that humans, rats, and mice can maximize payoffs via confidence-guided time allocation. HighlightsO_LIComputational and behavioral framework to assess bounded optimality of investments. C_LIO_LIHumans, rats, and mice invest more time to obtain more likely payoffs, in proportion to statistical confidence. C_LIO_LITime investment was close to optimal model predictions, reflecting bounded optimality of investments under uncertainty. C_LIO_LIBounded-optimal time investment may be an evolutionary ancient adaptive behavioral strategy. C_LI

animal behavior and cognition↗

bioRxiv2024

A Cortical Microcircuit for Region-Specific Credit Assignment in Reinforcement Learning

The distributed architecture of the cortex poses a fundamental challenge for reinforcement learning: how to assign credit specifically to regions that contribute to successful behavior? Cortical neurons can be driven by both global reinforcers, like rewards, and local sensory features, making it difficult to disentangle these influences. To address this, we investigated cortical reinforcement learning by manipulating the reward-predictive sensory modality during learning tasks, while monitoring key regulators of cortical activity--local inhibitory neurons, and cholinergic inputs. We found that VIP interneurons are broadly recruited by reward-predictive cues via a modality-independent cholinergic signal. However, when task demands aligned with local computation, SST interneurons suppressed VIP recruitment through an inhibitory feedback loop. A computational model demonstrates that this cholinergic-VIP-SST interneuron circuit motif enables targeted reinforcement learning and region-specific credit assignment in the cortex. These results offer a neurobiologically-grounded framework for how the cortex uses global reinforcement signals to direct plasticity to task-relevant regions, enabling those regions to adapt and fine-tune their responses.

neuroscience↗

bioRxiv2024

The neural mechanisms of fast versus slow decision-making

Is more haste less speed? Decision time variability has been attributed to speed/accuracy trade-offs(1), internal mental architecture(2), and noisy evidence accumulation(3-5). However, exploring these possibilities is difficult in rodents that consistently behave impulsively. Here, we demonstrate a novel floating-platform system that allows head-fixed mice to voluntarily vary decision times, akin to observed human behavior, in combination with sensitive neuroimaging approaches. We track the activity flow from medial to lateral frontal cortex (MFC to LFC) and record sequences of single-neuron activity. Choice-selective neurons displayed divergent temporal codes between MFC and LFC, with remarkable MFC susceptibility to optogenetic inhibition. These results suggest that LFC acts as an integrative motor threshold, while MFC plays a broader cognitive role in strategy and choice-selection.

neuroscience↗

bioRxiv2024

A disinhibitory basal forebrain to cortex projection supports sustained attention

Sustained attention, as an essential cognitive faculty governing selective sensory processing, exhibits remarkable temporal fluctuations. However, the underlying neural circuits and computational mechanisms driving moment-to-moment attention fluctuations remain elusive. Here we demonstrate that cortex-projecting basal forebrain parvalbumin-expressing inhibitory neurons (BF-PV) mediate sustained attention in mice performing an attention task. BF-PV activity predicts the fluctuations of attentional performance metrics [-] reaction time and accuracy [-] trial-by-trial, and optogenetic activation of these neurons enhances performance. BF-PV neurons also respond to motivationally salient events, such as predictive cues, rewards, punishments, and surprises, which a computational model explains as representing motivational salience for allocating attention over time. Furthermore, we found that BF-PV neurons produce cortical disinhibition by inhibiting cortical PV+ inhibitory neurons, potentially underpinning the observed attentional gain modulation in the cortex. These findings reveal a disinhibitory BF-to-cortex projection that regulates cortical gain based on motivational salience, thereby promoting sustained attention. HIGHLIGHTSO_LIBF-PV activity predicts attentional performance metrics: reaction time and accuracy C_LIO_LIBF-PV responses reflect the computation of motivational salience-guided attention allocation C_LIO_LIOptogenetic activation of BF-PV neurons improves attentional performance C_LIO_LIBF-PV neurons produce cortical disinhibition through topographic projections and mediate gain modulation C_LI

neuroscience↗

bioRxiv2024

Structured behavioral data format: An NWB extension standard for task-based behavioral neuroscience experiments

Understanding brain function necessitates linking neural activity with corresponding behavior. Structured behavioral experiments are crucial for probing the neural computations and dynamics underlying behavior; however, adequately representing their complex data is a significant challenge. Currently, a comprehensive data standard that fully encapsulates task-based experiments, integrating neural activity with the richness of behavioral context, is lacking. We designed a data model, as an extension to the NWB neurophysiology data standard, to represent structured behavioral neuroscience experiments, spanning stimulus delivery, timestamped events and responses, and simultaneous neural recordings. This data format is validated through its application to a variety of experimental designs, showcasing its potential to advance integrative analyses of neural circuits and complex behaviors. This work introduces a comprehensive data standard designed to capture and store a spectrum of behavioral data, encapsulating the multifaceted nature of modern neuroscience experiments.

neuroscience↗

bioRxiv2022

A cortical circuit for orchestrating oromanual food manipulation

The seamless coordination of hands and mouth--whether in humans eating corn on the cob or mice extracting sunflower seeds--represents one of evolutions most sophisticated motor achievements. Whereas spinal and brainstem circuits implement basic forelimb and orofacial actions, whether there is a specialized cortical circuit that assembles these actions to enable skilled oromanual manipulation remains unclear. Here, we discover a cortical area and its cell-type-specific circuitry that govern oromanual food manipulation in mice. An optogenetic screen of cortical areas and projection neuron types identified a rostral forelimb-orofacial area (RFO), wherein activation of pyramidal tract (PTFezf2) and intratelencephalic (ITPlxnD1) neurons induced concerted posture, forelimb and orofacial movements resembling eating. In a freely moving pasta-eating behavior, pharmacological RFO inactivation impaired the sitting posture, hand recruitment, and oromanual coordination in pasta eating. RFO PTFezf2 and ITPlxnD1 activity was closely correlated with oromanual pasta manipulation and hand-assisted biting. Optogenetic inhibition revealed that PTsFezf2 regulate dexterous hand and mouth movements while ITsPlxnD1 play a more prominent role in oromanual coordination. RFO forms the hub of an extensive network, with reciprocal connections to cortical forelimb and orofacial sensorimotor areas, as well as insular and visceral areas. Within this cortical network, RFO PTsFezf2 project unilaterally to multiple subcortical, brainstem and spinal areas associated with forelimb and orofacial control, while ITsPlxnD1 project bilaterally to the entire network and the ventrolateral striatum, and can mediate concurrent forelimb and mouth movement in part through their striatal projection. Together, these findings uncover the cell-type-specific implementation of a cortical circuit that orchestrates oromanual manipulation, essential for skilled feeding.

neuroscience↗

bioRxiv2022

Cortex-wide fast activation of VIP-expressing inhibitory neurons by reward and punishment

Reward and punishment powerfully inform ongoing behaviors and drive learning throughout the brain, including neocortex. Yet it remains elusive how these global signals are represented and impact local cortical computations. Previously we found that in auditory cortex, VIP-expressing interneurons are recruited by reinforcement feedback. Here, we used 3D random-access two-photon microscopy and fiber photometry to monitor VIP neural activity in dozens of cortical areas while mice learned an auditory decision task. We show that reward and punishment evoke a rapid, cortex-wide activation of most VIP interneurons. This global recruitment mode of VIP interneurons showed variations in temporal dynamics in individual neurons and across areas. Neither their weak sensory tuning in visual cortex, nor their arousal state modulation was predictive of reinforcer responses of VIP interneurons. We suggest that VIP-expressing cortical inhibitory neurons transduce global reinforcement signals to provide disinhibitory control over local circuit computations and their plasticity.

neuroscience↗

bioRxiv2021

Apparent sunk cost effect in rational agents

Rational decision makers aim to maximize their gains, but humans and other animals often fail to do so, exhibiting biases and distortions in their choice behavior. In a recent study of economic decisions, humans, mice, and rats have been reported to succumb to the sunk cost fallacy, making decisions based on irrecoverable past investments in detriment of expected future returns. We challenge this interpretation because it is subject to a statistical fallacy, a form of attrition bias, and the observed behavior can be explained without invoking a sunk cost-dependent mechanism. Using a computational model, we illustrate how a rational decision maker with a reward-maximizing decision strategy reproduces the reported behavioral pattern and propose an improved task design to dissociate sunk costs from fluctuations in decision valuation. Similar statistical confounds may be common in analyses of cognitive behaviors, highlighting the need to use causal statistical inference and generative models for interpretation.

animal behavior and cognition↗

bioRxiv2020

Total recall: episodic memory retrieval,choice, and memory confidence in the rat

Episodic memory enables recollection of past experiences to guide future behavior. Humans know which memories to trust (high confidence) and which to doubt (low confidence). How memory retrieval, memory confidence, and memory-guided decisions are related, however, is not understood. Additionally, whether animals can assess confidence in episodic memories to guide behavior is unknown. We developed a spatial episodic memory task in which rats were incentivized to gamble their time: betting more following a correct choice yielded greater reward. Rat behavior reflected memory confidence, with higher temporal bets following correct choices. We applied modern machine learning to identify a memory decision variable, and built a generative model of memories evolving over time that accurately predicted both choices and confidence reports. Our results reveal in rats an ability thought to exist exclusively in primates, and introduce a unified model of memory dynamics, retrieval, choice, and confidence.

neuroscience↗

bioRxiv2020

Basal forebrain-derived acetylcholine encodes valence-free reinforcement prediction error.

Acetylcholine (Ach) is released by the cholinergic basal forebrain (CBF) throughout the cortical mantle and is implicated in behavioral functions ranging from arousal to attention to learning. Yet what signal ACh provides to cortex remains unresolved, hindering our understanding of its functional roles. Here we demonstrate that the CBF signals unsigned reinforcement prediction error, in contrast to dopamine (DA) neurons that encode reward prediction error. We show that both CBF neuronal activity and acetylcholine (ACh) release at cortical targets signal reinforcement delivery, acquire responses to predictive stimuli and show diminished responses to expected outcomes, hallmarks of a prediction error. To compare ACh with DA, we simultaneously monitored the activity of both neuromodulators during a serial reversal learning task. ACh tracked learning as swiftly as DA during acquisition but lagged slightly during extinction, suggesting that these neuromodulators play complementary roles in reinforcement as their patterns of innervation, cellular targets, and signaling mechanisms are themselves complementary. Through retrograde viral tracing we show that the cholinergic and dopaminergic systems engage overlapping upstream circuits, accounting for their coordination during learning. This predictive and valence-free signal explains how ACh can proactively and retroactively improve the processing of behaviorally important stimuli, be they good or bad.

neuroscience↗

bioRxiv2019

Centrioles control the capacity, but not the specificity, of cytotoxic T cell killing

Immunological synapse formation between cytotoxic T lymphocytes (CTLs) and the target cells they aim to destroy is accompanied by reorientation of the CTL centrosome to a position beneath the synaptic membrane. Centrosome polarization is thought to enhance the potency and specificity of killing by driving lytic granule fusion at the synapse and thereby the release of perforin and granzymes toward the target cell. To test this model, we employed a genetic strategy to delete centrioles, the core structural components of the centrosome. Centriole deletion altered microtubule architecture, as expected, but surprisingly had no effect on lytic granule polarization and directional secretion. Nevertheless, CTLs lacking centrioles did display substantially reduced killing potential, which was associated with defects in both lytic granule biogenesis and synaptic actin remodeling. These results reveal an unexpected role for the intact centrosome in controlling the capacity, but not the specificity, of cytotoxic killing.

immunology↗

bioRxiv2019

Distinct synchronization, cortical coupling and behavioural function of basal forebrain cholinergic neuron types

Basal forebrain cholinergic neurons (BFCN) densely innervate the forebrain and modulate synaptic plasticity, cortical processing, brain states and oscillations. However, little is known about the functional diversity of cholinergic neurons and whether distinct types support different functions. To examine this question we recorded BFCN in vivo, to examine their behavioral functions, and in vitro, to study their intrinsic properties. We identified two distinct types of BFCN that markedly differ in their firing modes, synchronization properties and behavioral correlates. Bursting cholinergic neurons (BFCNBURST) fired in zero-lag synchrony with each other, phase-locked to cortical theta activity and fired precisely timed bursts of action potentials after reward and punishment. Regular firing cholinergic neurons (BFCNREG) were found predominantly in the posterior basal forebrain, displayed strong theta rhythmicity (5-10 Hz), fired asynchronously with each other and responded with precise single spikes after behavioral outcomes. In an auditory detection task, synchronization of BFCNBURST neurons to auditory cortex predicted the timing of mouse responses, whereas tone-evoked cortical coupling of BFCNREG predicted correct detections. We propose that cortical activation relevant for behavior is controlled by the balance of two cholinergic cell types, where the precise proportion of the strongly activating BFCNBURST follows an anatomical gradient along the antero-posterior axis of the basal forebrain.

neuroscience↗

bioRxiv2017

Categorical Representations Of Decision-Variables In Orbitofrontal Cortex

The brain creates internal representations of the external world in the form of neural activity, which is structured to support adaptive behavior. In many cortical regions, individual neurons respond to specific features that are matched to the function of each region and statistics of the world. In frontal cortex, however, neurons display baffling complexity, responding to a mixture of sensory, motor and other variables. Here we use an integrated new approach to understanding the architecture of higher-order cortical representations, and use this approach to show that discrete groups of orbitofrontal cortex (OFC) neurons encode distinct decision variables. Using rats engaged in a complex task combining perceptual and value guided decisions, we found that OFC neurons can be grouped into distinct, categorical response types. These categorical representations map directly onto decision-variables of a choice model explaining our behavioral data, such as reward size, decision confidence and integrated value. We propose that, like sensory neurons, frontal neurons form a sparse and over complete population representation aligned to the natural statistics of the world - in this case spanning the space of decision-variables required for optimal behavior.

neuroscience↗