Cooperative thalamocortical circuit mechanism for sensory prediction errors

Nature, Journal Year: 2024, Volume and Issue: 633(8029), P. 398 - 406

Published: Aug. 28, 2024

Abstract The brain functions as a prediction machine, utilizing an internal model of the world to anticipate sensations and outcomes our actions. Discrepancies between expected actual events, referred errors, are leveraged update guide attention towards unexpected events 1–10 . Despite importance prediction-error signals for various neural computations across brain, surprisingly little is known about circuit mechanisms responsible their implementation. Here we describe thalamocortical disinhibitory that required generating sensory in mouse primary visual cortex (V1). We show violating animals’ predictions by stimulus preferentially boosts responses layer 2/3 V1 neurons most selective stimulus. Prediction errors specifically amplify input, rather than representing non-specific surprise or difference how input deviates from animal’s predictions. This amplification implemented cooperative mechanism requiring thalamic pulvinar cortical vasoactive-intestinal-peptide-expressing (VIP) inhibitory interneurons. In response VIP inhibit specific subpopulation somatostatin-expressing interneurons gate excitatory V1, resulting pulvinar-driven stimulus-selective V1. Therefore, prioritizes unpredicted information selectively increasing salience features through synergistic interaction neocortical circuits.

Language: Английский

---

Functional connectomics reveals general wiring rule in mouse visual cortex

Nature, Journal Year: 2025, Volume and Issue: 640(8058), P. 459 - 469

Published: April 9, 2025

Abstract Understanding the relationship between circuit connectivity and function is crucial for uncovering how brain computes. In mouse primary visual cortex, excitatory neurons with similar response properties are more likely to be synaptically connected 1–8 ; however, broader rules remain unknown. Here we leverage millimetre-scale MICrONS dataset analyse synaptic functional of across cortical layers areas. Our results reveal that preferentially within areas—including feedback connections—supporting universality ‘like-to-like’ hierarchy. Using a validated digital twin model, separated neuronal tuning into feature (what respond to) spatial (receptive field location) components. We found only component predicts fine-scale connections beyond what could explained by proximity axons dendrites. also discovered higher-order rule whereby postsynaptic neuron cohorts downstream presynaptic cells show greater similarity than predicted pairwise like-to-like rule. Recurrent neural networks trained on simple classification task develop patterns mirror both rules, magnitudes those in data. Ablation studies these recurrent disrupting impairs performance random connections. These findings suggest principles may have role sensory processing learning, highlighting shared biological artificial systems.

Language: Английский

---

Lateral inhibition in V1 controls neural and perceptual contrast sensitivity

Nature Neuroscience, Journal Year: 2025, Volume and Issue: unknown

Published: March 3, 2025

Language: Английский

---

Functional connectomics reveals general wiring rule in mouse visual cortex

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: March 14, 2023

Understanding the relationship between circuit connectivity and function is crucial for uncovering how brain implements computation. In mouse primary visual cortex (V1), excitatory neurons with similar response properties are more likely to be synaptically connected, but previous studies have been limited within V1, leaving much unknown about broader rules. this study, we leverage millimeter-scale MICrONS dataset analyze synaptic functional of individual across cortical layers areas. Our results reveal that responses preferentially connected both areas — including feedback connections suggesting universality ‘like-to-like’ hierarchy. Using a validated digital twin model, separated neuronal tuning into feature (what respond to) spatial (receptive field location) components. We found only component predicts fine-scale connections, beyond what could explained by physical proximity axons dendrites. also higher-order rule where postsynaptic neuron cohorts downstream presynaptic cells show greater similarity than predicted pairwise like-to-like rule. Notably, recurrent neural networks (RNNs) trained on simple classification task develop patterns mirroring rules, magnitude those in data. Lesion these RNNs disrupting has significantly impact performance compared lesions random connections. These findings suggest principles may play role sensory processing learning, highlighting shared biological artificial systems.

Language: Английский

---

Synapse-type-specific competitive Hebbian learning forms functional recurrent networks

Proceedings of the National Academy of Sciences, Journal Year: 2024, Volume and Issue: 121(25)

Published: June 13, 2024

Cortical networks exhibit complex stimulus–response patterns that are based on specific recurrent interactions between neurons. For example, the balance excitatory and inhibitory currents has been identified as a central component of cortical computations. However, it remains unclear how required synaptic connectivity can emerge in developing circuits where synapses neurons simultaneously plastic. Using theory modeling, we propose wide range response properties arise from single plasticity paradigm acts at all connections—Hebbian learning is stabilized by synapse-type-specific competition for limited supply resources. In plastic circuits, this enables formation decorrelation inhibition-balanced receptive fields. Networks develop an assembly structure with stronger connections similarly tuned normalization orientation-specific center-surround suppression, reflecting stimulus statistics during training. These results demonstrate self-organize into functional suggest essential role competitive development circuits.

Language: Английский

---

Computational functions of precisely balanced neuronal microcircuits in an olfactory memory network

Cell Reports, Journal Year: 2025, Volume and Issue: 44(3), P. 115330 - 115330

Published: Feb. 20, 2025

Language: Английский

---

Multi-scale spiking network model of human cerebral cortex

Cerebral Cortex, Journal Year: 2024, Volume and Issue: 34(10)

Published: Oct. 1, 2024

Abstract Although the structure of cortical networks provides necessary substrate for their neuronal activity, alone does not suffice to understand activity. Leveraging increasing availability human data, we developed a multi-scale, spiking network model cortex investigate relationship between and dynamics. In this model, each area in one hemisphere Desikan–Killiany parcellation is represented by $1\,\mathrm{mm^{2}}$ column with layered structure. The aggregates data across multiple modalities, including electron microscopy, electrophysiology, morphological reconstructions, diffusion tensor imaging, into coherent framework. It predicts activity on all scales from single-neuron area-level functional connectivity. We compared electrophysiological resting-state magnetic resonance imaging (fMRI) data. This comparison reveals that can reproduce aspects both statistics fMRI correlations if inter-areal connections are sufficiently strong. Furthermore, study propagation single-spike perturbation macroscopic fluctuations through network. open-source serves as an integrative platform further refinements future silico studies structure, dynamics, function.

Language: Английский

---

Cortical multi-area model with joint excitatory-inhibitory clusters accounts for spiking statistics, inter-area propagation, and variability dynamics

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Jan. 30, 2024

Abstract The primate brain uses billions of interacting neurons to produce macroscopic dynamics and behavior, but current methods only allow neuroscientists investigate a subset the neural activity. Computational modeling offers an alternative testbed for scientific hypotheses, by allowing full control system. Here, we test hypothesis that local cortical circuits are organized into joint clusters excitatory inhibitory investigating influence this organizational principle on resting-state spiking activity, inter-area propagation, variability dynamics. model represents all vision-related areas in one hemisphere macaque cortex with biologically realistic neuron densities connectivities, expanding previous unclustered Each area is represented square millimeter microcircuit including density synapses, avoiding downscaling artifacts testing at natural scale. We find excitatory-inhibitory clustering normalizes activity statistics terms firing rate distributions inter-spike interval variability. A comparison data from V1, V4, FEF, 7a, DP shows enables especially higher be better captured. In addition, supports signal propagation across both feedforward feedback directions reasonable latencies. Finally, also show localized stimulation clustered quenches agreement experimental observations. conclude likely circuits, supporting statistics,

Language: Английский

---

Dopamine D1 receptor expression in dlPFC inhibitory parvalbumin neurons may contribute to higher visuospatial distractibility in marmosets versus macaques

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: June 16, 2024

Abstract Marmosets and macaques are common non-human primate models of cognition, but evidence suggests that marmosets perform more poorly appear distractible during cognitive tasks. The dorsolateral prefrontal cortex (dlPFC) plays a key role in regulating attention, prior research dopaminergic modulation inhibitory parvalbumin (PV) neurons could contribute to distractibility performance. Thus, we compared the two species using visual fixation task with distractors, performed molecular anatomical analyses dlPFC, linked functional microcircuitry performance computational modeling. We found than macaques, marmoset dlPFC PV contain higher levels dopamine-1 receptor (D1R) transcripts, similar mice, D1R protein. model suggested expression may increase by suppressing microcircuits, e.g., when dopamine is released salient stimuli.

Language: Английский

---

The mechanics of correlated variability in segregated cortical excitatory subnetworks

Proceedings of the National Academy of Sciences, Journal Year: 2024, Volume and Issue: 121(28)

Published: July 3, 2024

Understanding the genesis of shared trial-to-trial variability in neuronal population activity within sensory cortex is critical to uncovering biological basis information processing brain. Shared often a reflection structure cortical connectivity since it likely arises, part, from local circuit inputs. A series experiments segregated networks (excitatory) pyramidal neurons mouse primary visual challenge this view. Specifically, across-network correlations were found be larger than predicted given known weak cross-network connectivity. We aim uncover mechanisms responsible for these enhanced through biologically motivated models. Our central finding that coupling each excitatory subpopulation with specific inhibitory provides most robust network-intrinsic solution shaping correlations. This result argues existence excitatory–inhibitory functional assemblies early areas which mirror not just response properties but also between cells. Furthermore, our findings provide theoretical support recent experimental observations showing inhibition forms structural and subnetworks cells, contrast classical view nonspecific blanket suppression excitation.

Language: Английский

---