Biology subjects

Pennartz, C.

Publications and source records attributed to Pennartz, C..

12 recordsLinked to original sources

bioRxiv2026

Selective binocular integration in mouse visual cortex: binocular but not monocular neurons respond distinctly to congruent stimuli

Even though binocular integration is a foremost example of integrative computations in the brain, it remains unknown how the visual cortex mediates it. Here, we investigated neuronal responses to binocular stimuli in layers 2/3 of mouse primary visual cortex (V1) and lateromedial (LM) area with in vivo two-photon imaging. We presented mice with three types of binocular stimuli: congruent gratings, orthogonal gratings and a flashing Mondrian pattern in binocular conflict with a grating. We looked for congruency-specific responses to the binocular stimuli, but found that most neurons responded homogeneously across stimulus types. Responses of cells with a strong monocular preference for the contralateral eye responded similarly to monocular and binocular stimulation, while responses of cells with a preference for ipsilateral stimuli showed a strong but nonspecific suppression to all binocular stimuli. Only cells without a strong monocular preference (i.e. binocular cells) exhibited linear integration of monocular responses to incongruent stimuli and binocular facilitation to congruent gratings. Responses to ipsilateral components of a binocular display were found to be poor predictors of binocular responses regardless of eye preference of individual cells. We conclude that most neuronal responses to binocular stimuli in mouse V1 & LM are uninformative about the congruency of the monocular components, with the exception of binocular cells which show either linear or supralinear integration.

neuroscience↗

bioRxiv2026

Baseline activity of V1 interneurons connects pupil-linked arousal to engaged behavioral state

AO_SCPLOWBSTRACTC_SCPLOWHumans and rodents alternate between discrete, persistent behavioral states during perceptual decision-making. For example, trials can be clustered into states of engaged, disengaged, and biased decision-making strategies. In mice, the probability of being in an engaged state exhibits an inverted-U relationship with baseline pupil-linked arousal, consistent with the Yerkes-Dodson Law. We replicated this relationship in mice (N=11; audio-visual change detection task) and showed that it generalizes to humans (N=69; auditory detection task). We then examined its neural basis, using simultaneous pupillometry and electrophysiological recordings in mouse V1. A nonlinear mediation model revealed that baseline activity of putative fast-spiking interneurons, but not putative pyramidal neurons, made a significant statistical contribution to the inverted-U shaped relationship between pupil-linked arousal and engaged-state probability. These findings are consistent with the idea that arousal may shape behavioral states by modulating baseline inhibitory activity within a primary sensory region.

neuroscience↗

bioRxiv2025

Continuous Flash Suppression responses in mouse visual cortex: stimulus laterality and anesthesia effects

We investigated whether binocularly conflicting stimuli adapted from primate binocular rivalry studies could induce binocular response suppression in mouse visual cortex. We presented binocularly conflicting stimuli adapted from the paradigm of continuous flash suppression to awake and anesthetized mice and examined neuronal responses in visual areas V1 and LM. Neurons often showed a preference for the left or right eye where a monocular grating was presented, and their responses were modulated by presenting a conflicting flashing Mondrian stimulus to the other eye. The direction of this modulation depended on whether neurons preferentially responded to contralateral or ipsilateral gratings. Responses of cells with a preference for ipsilateral gratings were suppressed during binocular conflict, while contralateral-preferring cells were less suppressed or even enhanced by binocular conflict. Binocular conflict modulation continued to occur under anesthesia, but response recovery after intermittent binocular conflict only occurred in awake mice. Responses to Mondrians were modulated by binocular conflict in a similar fashion as those to grating stimuli, which suggests that mouse binocular conflict processing exhibits a distinctive dependency on stimulus laterality over other stimulus features. Finally, a canonical binocular rivalry model could be successfully fitted to our data but lacked sufficient competitive inhibition and adaptation strength to create oscillatory activity under binocular conflict.

neuroscience↗

bioRxiv2025

Highly Correlated Activity across Higher-Order Thalamic Nuclei in Awake and Anesthetized States

The thalamus is an egg-shaped structure deep inside the brain and is of central importance for our cognitive functions. The thalamus can be divided into first- and higher-order (HO) nuclei. Unlike first-order thalamic regions which are known to primarily relay sensory signals to the neocortex, the functions of HO thalamic nuclei remain far less clear. Although previous studies indicate HO thalamic nuclei influence cortical processing and consciousness, most studies examined single thalamic nuclei in isolation. To investigate the relation between distinct HO thalamic nuclei, we recorded local field potentials (LFP) and multi-unit activity (MUA) across multiple thalamic nuclei simultaneously in awake and anesthetized mice with sensory stimulation. The effect of voluntary locomotion was also studied in the awake state. Surprisingly, we found that both LFP and MUA were strongly correlated across HO nuclei in both states, which cannot be explained by volume conduction alone but suggests the existence of shared synaptic inputs. Furthermore, in the awake state locomotion and sensory stimuli activated many HO nuclei, including those that have not classically been regarded as sensory or motor. These findings challenge the classical view of the thalamus being the aggregate of isolated, independent nuclei and suggest that the thalamus serves as an ideal bottleneck whose coordination resulted in effective orchestration of the entire brain.

neuroscience↗

bioRxiv2024

Efficient laminar-distributed interactions and orientation selectivity in the mouse V1 cortical column

The emergence of orientation selectivity in the visual cortex is a well-known phenomenon in neuroscience, but the details of such emergence and the role of different cortical layers and cell types, particularly in rodents which lack a topographical organization of orientation-selectivity (OS) properties, are less clear. To tackle this question, we use an existing biologically detailed model of the mouse V1 cortical column, which is constrained by existing connectivity data across cortical layers and between pyramidal, PV, SST and VIP cell types. Using this model as a basis, we implemented activity-dependent structural plasticity induced by stimulation with orientated drifting gratings, leading to a good match of tuning properties of pyramidal cells with experimentally observed OS laminar distribution, their evoked firing rate and tuning width. We then employed a mean-field model to uncover the role of co-tuned subnetworks in laminar signal propagation and explain the effects of intra- and inter-laminar coupling distributions. Our plasticity-induced modified model and mean-field model were able to explain both the excitatory enhancement through co-tuned subnetworks and inter-laminar disynaptic inhibition. Overall, our work highlights the importance of the clustering of neural selectivity features for effective excitatory transmission in cortical circuits.

neuroscience↗

bioRxiv2024

Cortical networks with multiple interneuron types generate oscillatory patterns during predictive coding

Predictive coding (PC) proposes that our brains work as an inference machine, generating an internal model of the world and minimizing predictions errors (i.e., differences between external sensory evidence and internal prediction signals). Theoretical models of PC often rely on high-level approaches, and therefore implementations detailing which neurons or pathways are used to compute prediction errors or adapt the internal representations, as well as their level of agreement with biological circuitry, are currently missing. Here we propose a computational model of PC, which integrates a neuroanatomically informed hierarchy of cortical areas with a precise laminar organization and cell-type-specific connectivity between pyramidal, PV, SST and VIP cells. Our model efficiently performs PC, even in the presence of external and internal noise, by forming latent representations of naturalistic visual input (MNIST, fashion-MNIST and grayscale CIFAR-10) via Hebbian learning and using them to predict sensory input by minimizing prediction errors. The model assumes that both positive and negative prediction errors are computed by stereotypical pyramidal-PV-SST-VIP circuits with the same structure but different incoming input. During sensory inference, neural oscillatory activity emerges in the system due to interactions between representation and prediction error microcircuits, with optogenetics-inspired inactivation protocols revealing a differentiated role of PV, SST and VIP cell types in such dynamics. Finally, our model shows anomalous responses to deviant stimuli within series of same-image presentations, in agreement with experimental results on mismatch negativity and oddball paradigms. We argue that our model constitutes an important step to better understand the circuits mediating PC in real cortical networks. Author summaryPredictive coding (PC) suggests that the brain constantly generates expectations about the world and updates these expectations based on incoming sensory input. While this theory is widely accepted, we still lack detailed models that show how specific neurons and brain circuits might carry out these processes. Here, we present a computational model which addresses this gap by including biologically plausible brain circuitry with specific types of neurons (pyramidal, PV, SST, and VIP cells) and their connections. It efficiently learns to form internal representations of visual information and uses them to predict sensory input, adjusting its predictions when errors occur. We found that particular types of neurons play different roles in these processes, and that neural oscillations emerge during the training and inference processes. Our model also replicates neural patterns observed in experiments where unexpected stimuli appear. By integrating anatomical and functional details, our work brings us closer to understanding how the brain uses predictive coding at the circuit level.

neuroscience↗

bioRxiv2024

Hippocampal place cell sequences during a visual discrimination task: recapitulation of paths near the chosen reward site and independence from perirhinal activity

Compressed hippocampal place-cell sequences have been associated with memory storage, retrieval and planning, but it remains unclear how they align with activity in the parahippocampal cortex. In a visuospatial discrimination task, we found a wide repertoire of hippocampal place cell sequences, which recapitulated paths across the task environment. Place cell sequences generated at reward sites predominantly reiterated trajectories near the chosen maze side, whereas trajectories associated with the side chosen in the previous trial were underrepresented. We hypothesized that neurons in the perirhinal cortex, which during the task display broad firing fields correlated with the animals location, might reactivate in concert with hippocampal sequences. However, we found no evidence of significant perirhinal engagement during virtual trajectories, indicating that these hippocampal memory-related operations can occur independently of the perirhinal cortex.

neuroscience↗

bioRxiv2024

Transient DREADD manipulation of the dorsal Dentate Gyrus in rats impairs disambiguation of similar place-outcome associations

The dentate gyrus subfield of the hippocampus is thought to be critically involved in the disambiguation of similar episodic experiences and places in a context-dependent manner. However, most empirical evidence has come from lesion and gene knock-out studies in rodents, in which the dentate gyrus function is permanently perturbed and compensation of affected functions via other areas within the memory circuit could take place. The acute and causal role of the dentate gyrus herein remains therefore elusive. The present study aimed to investigate the acute role of the dorsal dentate gyrus in disambiguation learning using reversible inhibitory DREADDs. Rats were trained on a location discrimination task and learnt to discriminate between a rewarded and unrewarded location with either small (similar condition) or large (dissimilar condition) separation. Reward contingencies switched after a reversal rule, allowing us to track the temporal engagement of the dentate gyrus during the task. Bilateral but not unilateral DREADD modulation of the dentate gyrus impaired the initial acquisition learning of place-reward associations, but performance rapidly recovered to control levels within the same session. Modelling of the behavioural patterns revealed that reward learning and reward sensitivity were not associated with the DREADD-dependent impairment during acquisition learning, suggesting that either the ability to encode place-reward associations, or the fine-grained coding of place were instead affected. Our study thus provides novel evidence that the dorsal dentate gyrus is acutely and bilaterally engaged during the initial acquisition learning of ambiguous place-reward associations, although the exact neural mechanisms supporting this function still need to be fully understood.

neuroscience↗

bioRxiv2023

Learning to segment self-generated from externally caused optic flow through sensorimotor mismatch circuits

Efficient sensory detection requires the capacity to ignore task-irrelevant information, for example when optic flow patterns created by egomotion need to be disentangled from object perception. To investigate how this is achieved in the visual system, predictive coding with sensorimotor mismatch detection is an attractive starting point. Indeed, experimental evidence for sensorimotor mismatch signals in early visual areas exists, but it is not understood how they are integrated into cortical networks that perform input segmentation and categorization. Our model advances a biologically plausible solution by extending predictive coding models with the ability to distinguish self-generated from externally caused optic flow. We first show that a simple three neuron circuit produces experience-dependent sensorimotor mismatch responses, in agreement with calcium imaging data from mice. This microcircuit is then integrated into a neural network with two generative streams. The motor-to-visual stream consists of parallel microcircuits between motor and visual areas and learns to spatially predict optic flow resulting from self-motion. The second stream bidirectionally connects a motion-selective higher visual area (mHVA) to V1, assigning a crucial role to the abundant feedback connections: the maintenance of a generative model of externally caused optic flow. In the model, area mHVA learns to segment moving objects from the background, and facilitates object categorization. Based on shared neurocomputational principles across species, the model also maps onto primate vision. Our work extends the Hebbian predictive coding to sensorimotor settings, in which the agent actively moves - and learns to predict the consequences of its own movements. Significance statementThis research addresses a fundamental challenge in sensory perception: how the brain distinguishes between self-generated and externally caused visual motion. Using a computational model inspired by predictive coding and sensorimotor mismatch detection, the study proposes a biologically plausible solution. The model incorporates a neural microcircuit that generates sensorimotor mismatch responses, aligning with experimental data from mice. This microcircuit is integrated into a neural network with two streams: one predicting self-motion-induced optic flow and another maintaining a generative model for externally caused optic flow. The research advances our understanding of how the brain segments visual input into object and background, shedding light on the neural mechanisms underlying perception and categorization not only in rodents, but also in primates.

neuroscience↗

bioRxiv2023

The neural and computational architecture of feedback dynamics in mouse cortex during stimulus report

Conscious reportability of visual input is associated with a bimodal neural response in primary visual cortex (V1): an early-latency response coupled to stimulus features and a late-latency response coupled to stimulus report or detection. This late wave of activity, central to major theories of consciousness, is thought to be driven by prefrontal cortex (PFC), responsible for "igniting" it. Here we analyzed two electrophysiological studies in mice performing different stimulus detection tasks, and characterize neural activity profiles in three key cortical regions: V1, posterior parietal cortex (PPC) and PFC. We then developed a minimal network model, constrained by known connectivity between these regions, reproducing the spatio-temporal propagation of visual-and report-related activity. Remarkably, while PFC was indeed necessary to generate report-related activity in V1, this occurred only through the mediation of PPC. PPC, and not PFC, had the final veto in enabling the report-related late wave of V1 activity.

neuroscience↗

bioRxiv2023

Neural correlates of object identity and reward outcome in the corticohippocampal hierarchy: double dissociation between perirhinal and secondary visual cortex

Neural circuits support behavioral adaptations by integrating sensory and motor information with reward and error-driven learning signals, but it remains poorly understood how these signals are distributed across different levels of the corticohippocampal hierarchy. We trained rats on a multisensory object-recognition task and compared visual and tactile responses of simultaneously recorded neuronal ensembles in somatosensory cortex, secondary visual cortex, perirhinal cortex and hippocampus. The sensory regions primarily represented unisensory information, while hippocampus was modulated by both vision and touch. Surprisingly, secondary visual cortex but not perirhinal neurons coded object-specific information, whereas perirhinal but not visual cortical neurons signaled trial outcome. A majority of outcome-related perirhinal cells responded to a negative outcome (reward omission), whereas a minority of other cells coded positive outcome (reward delivery). Our results support a distributed neural coding of multisensory variables in the cortico-hippocampal hierarchy, with a double dissociation between higher visual cortex and perirhinal cortex in coding of object identity versus feedback on trial outcome.

animal behavior and cognition↗

bioRxiv2023

Predictive coding with spiking neurons and feedforward gist signalling

Predictive coding (PC) is an influential theory in neuroscience, which suggests the existence of a cortical architecture that is constantly generating and updating predictive representations of sensory inputs. Owing to its hierarchical and generative nature, PC has inspired many computational models of perception in the literature. However, the biological plausibility of existing models has not been sufficiently explored due to their use of artificial neural network features such as a non-linear, continuous, and clock-driven function approximator as basic unit of computation. Therefore, we have developed a spiking neural network for predictive coding (SNN-PC), in which neurons communicate using event-driven and asynchronous spikes. While adopting the hierarchical structure and Hebbian learning algorithms from previous PC neural network models, SNN-PC introduces two novel features: 1) a fast feedforward sweep from the input to higher areas, which generates a spatially reduced and abstract representation of input (i.e., a neural code for the gist of a scene) and provides a neurobiological alternative to an arbitrary choice of priors; and 2) a separation of positive and negative error-computing neurons, which counters the biological implausibility of a bi-directional error neuron with a very high basal firing rate. After training with the MNIST handwritten digit dataset, SNN-PC developed hierarchical internal representations and was able to reconstruct samples it had not seen during training. SNN-PC suggests biologically plausible mechanisms by which the brain may perform perceptual inference and learning in an unsupervised manner. In addition, it may be used in neuromorphic applications that can utilize its energy-efficient, event-driven, local learning, and parallel information processing nature. Author summaryHow does the brain seamlessly perceive the world, in the midst of chaotic sensory barrage? Rather than passively relaying information that sensory organs pick up from the external world along the cortical hierarchy for a series of feature extractions, it actively gathers statistical regularities from sensory inputs to track causal relationships between physical properties of external objects and the body. In other words, the brains perceptual apparatus is constantly trying to make sense of the incoming streams of sensory input and represent the subjects current situation by building and maintaining internal models of the world and body. While this constructivist theme in understanding perception has been pervasive across multiple disciplines from philosophy to psychology to computer science, a comprehensive theory of brain function called predictive coding aims at unifying neural implementations of perception. In this study, we present a biologically plausible neural network for predictive coding that uses spiking neurons, Hebbian learning, and a feedforward visual pathway to perform perceptual inference and learning on images. Not only does the model show that predictive coding is well behaved under the biological constraint of spiking neurons, but it also provides deep learning and neuromorphic communities with novel paradigms of learning and computational architectures inspired by the natures most intelligent system, the brain.

neuroscience↗