AISTROSIGHT: La pharmacologie des neurones et des astrocytes à l’aide des sciences du numérique

Neural computational power is determined by neuroenergetics, but how and which energy substrates are allocated to various forms of memory engram is unclear. To solve this question, we asked whether neuronal fueling by glucose or lactate scales differently upon increasing neural computation and cognitive loads. Here, using electrophysiology, two-photon imaging, cognitive tasks, and mathematical modeling, we show that both glucose and lactate are involved in engram formation, with lactate supporting long-term synaptic plasticity evoked by high-stimulation load activity patterns and high attentional load in cognitive tasks and glucose being sufficient for less demanding neural computation and learning tasks. Indeed, we show that lactate is mandatory for demanding neural computation, such as theta-burst stimulation, while glucose is sufficient for lighter forms of activity-dependent long-term potentiation (LTP), such as spike timing-dependent plasticity (STDP). We find that subtle variations of spike number or frequency in STDP are sufficient to shift the on-demand fueling from glucose to lactate. Finally, we demonstrate that lactate is necessary for a cognitive task requiring high attentional load, such as the object-in-place task, and for the corresponding in vivo hippocampal LTP expression but is not needed for a less demanding task, such as a simple novel object recognition. Overall, these results demonstrate that glucose and lactate metabolism are differentially engaged in neuronal fueling depending on the complexity of the activity-dependent plasticity and behavior.

Tensor decomposition has recently been gaining attention in the machine learning community for the analysis of individual traces, such as Electronic Health Records. However, this task becomes significantly more difficult when the data follows complex temporal patterns. This paper introduces the notion of a temporal phenotype as an arrangement of features over time and it proposes SWoTTeD ( S liding W ind o w for T emporal Te nsor D ecomposition), a novel method to discover hidden temporal patterns. SWoTTeD integrates several constraints and regularizations to enhance the interpretability of the extracted phenotypes. We validate our proposal using both synthetic and real-world datasets, and we present an original usecase using data from the Greater Paris University Hospital. The results show that SWoTTeD achieves at least as accurate reconstruction as recent state-of-the-art tensor decomposition models, and extracts temporal phenotypes that are meaningful for clinicians.

ABSTRACT Astrocytes are now widely accepted as key regulators of brain function and behavior. Calcium (Ca 2+ ) signals in perisynaptic astrocytic processes (PAPs) enable astrocytes to fine‐tune neurotransmission at tripartite synapses. As most PAPs are below the diffraction limit, their content in Ca 2+ stores and the contribution of the latter to astrocytic Ca 2+ activity is unclear. Here, we reconstruct hippocampal tripartite synapses in 3D from a high‐resolution electron microscopy (EM) dataset and find that 75% of PAPs contain some endoplasmic reticulum (ER), a major calcium store in astrocytes. The ER in PAPs displays strikingly diverse shapes and intracellular spatial distributions. To investigate the causal relationship between each of these geometrical properties and the spatiotemporal characteristics of Ca 2+ signals, we implemented an algorithm that generates 3D PAP meshes by altering the distribution of the ER independently from ER and cell shape. Reaction–diffusion simulations in these meshes reveal that astrocyte activity is governed by a complex interplay between the location of Ca 2+ channels, ER surface–volume ratio, and spatial distribution. In particular, our results suggest that ER‐PM contact sites can act as local signal amplifiers if equipped with IP 3 R clusters but attenuate PAP Ca 2+ activity in the absence of clustering. This study sheds new light on the ultrastructural basis of the diverse astrocytic Ca 2+ microdomain signals and on the mechanisms that regulate neuron‐astrocyte signal transmission at tripartite synapses.

Homeoproteins (HPs) and cell-penetrating peptides (CPPs) enter cells by endocytosis and direct membrane crossing (translocation). However, unlike endocytosis, translocation remains globally unknown. Here, we developed an electrophysiological approach to assess the internalization of CPPs (Tat, R 9 , penetratin and R 6 W 3 ) and the HPs Otx2 and En2 though single-cell unitary transient currents in mammalian cells. At resting membrane potential, CPPs or HPs lead to submillisecond transient pores, faster than any endocytosis event, which reveal the rapid passage of the peptide across the membrane i.e., by translocation. We evidenced that expression of specific membrane glycosaminoglycans is mandatory for translocation-induced transient pores. Associated transient currents are supralinearly enhanced by hyperpolarization and poorly affected by depolarization. Moreover, a CPP-conjugated bioactive cargo similarly translocates into cytosol. Finally, we show similar HPs-evoked transient pores in brain cortical pyramidal cells, showing the physiological relevance of translocation, with crucial biotechnological and therapeutical consequences for cell delivery purposes.

Clustering based on belief functions has been gaining increasing attention in the machine learning community due to its ability to effectively represent uncertainty and/or imprecision. However, none of the existing algorithms can be applied to complex data, such as mixed data (numerical and categorical) or non-tabular data like time series. Indeed, these types of data are, in general, not represented in a Euclidean space and the aforementioned algorithms make use of the properties of such spaces, in particular for the construction of barycenters. In this paper, we reformulate the Evidential C-Means (ECM) problem for clustering complex data. We propose a new algorithm, Soft-ECM, which consistently positions the centroids of imprecise clusters requiring only a semi-metric. Our experiments show that Soft-ECM present results comparable to conventional fuzzy clustering approaches on numerical data, and we demonstrate its ability to handle mixed data and its benefits when combining fuzzy clustering with semi-metrics such as DTW for time series data.

Abstract Static or frozen disorder, characterised by spatial heterogeneities, influences diverse complex systems, encompassing many-body systems, equilibrium and nonequilibrium states of matter, intricate network topologies, biological systems, and wave-matter interactions. While static disorder has been thoroughly examined, delving into evolving disorder brings increased intricacy to the issue. An example of this complexity is the observation of stochastic acceleration of electromagnetic waves in evolving media, where noisy fluctuations in the propagation medium transfer effective momentum to the wave. Here, we investigate elastic wave propagation in a one-dimensional heterogeneous medium with diagonal disorder. We examine two types of complex elastic materials: one with static disorder, where mass density randomly varies in space, and the other with evolving disorder, featuring random variations in both space and time. Our results indicate that evolving disorder enhances the propagation speed of Gaussian pulses compared to static disorder. Additionally, we demonstrate that the enhanced speed effect also occurs when the medium evolves chaotically rather than randomly over time. The latter establishes that evolving randomness is not a unique prerequisite for observing the enhanced transport of wavefronts, introducing the concept of chaotic speed enhancement of waves in complex media.

National audience

One-shot learning-the ability to form memories after a single, brief salient event-is essential for behavioral flexibility in a dynamic world. However, how one-shot learning unfolds in the brain and whether it relies on distinct plasticity mechanisms remains unknown. Here we show that a nonclassical plasticity mechanism requiring only a few stimulations-endocannabinoid-mediated long-term potentiation (eCB-LTP)-underlies one-shot learning in the striatum. To do so, we developed a one-shot behavioral paradigm-the sticky tape avoidance test-in which mice learn to avoid a piece of sticky tape after a spontaneous single and brief contact. Brief, but not prolonged, contacts drive striatal potentiation in vivo and engage coordinated cortical-striatal activity patterns consistent with eCB-LTP induction, as corroborated by ex vivo electrophysiology and computational modeling. Finally, genetic and pharmacological disruptions of eCB-LTP impaired one-shot learning. These results highlight the essential role of nonclassical plasticity mechanisms in supporting memory formation after a single experience.

International audience

In this paper, we considered an age-structured jump model that arises as a description of continuous-time random walks with infinite mean waiting time between jumps. We proved that, under a suitable rescaling, this equation converges in the long-time large-scale limit to a time-fractional subdiffusion equation.

Low-dimensional descriptions of large systems of coupled oscillators and spiking neurons rely heavily on the Lorentzian Ansatz. We show that its privileged role is geometric rather than heuristic: for the transport induced by Riccati dynamics, the Cauchy-Lorentz family indeed emerges as the unique connected two-dimensional family of continuous probability densities that is invariant under the induced projective transport. The key step of the demonstration is to reformulate the dynamics on the circle, where the problem reduces to the uniqueness of the rotation-invariant probability measure. Under stereographic projection, this yields the standard Cauchy law and, under the full projective action, the Lorentzian family. This result gives a unified geometric foundation for the Ott-Antonsen [Chaos 18, 037113 (2008)] and Montbri{ó}-Paz{ó}-Roxin [Phys. Rev. X 5, 021028 (2015)] reductions, explains the failure of Gaussian closures, and identifies the structural condition underlying exact two-parameter reductions.

Low-dimensional descriptions of large systems of coupled oscillators and spiking neurons rely heavily on the Lorentzian Ansatz. We show that its privileged role is geometric rather than heuristic: for the transport induced by Riccati dynamics, the Cauchy-Lorentz family indeed emerges as the unique connected two-dimensional family of continuous probability densities that is invariant under the induced projective transport. The key step of the demonstration is to reformulate the dynamics on the circle, where the problem reduces to the uniqueness of the rotation-invariant probability measure. Under stereographic projection, this yields the standard Cauchy law and, under the full projective action, the Lorentzian family. This result gives a unified geometric foundation for the Ott-Antonsen [Chaos 18, 037113 (2008)] and Montbri{ó}-Paz{ó}-Roxin [Phys. Rev. X 5, 021028 (2015)] reductions, explains the failure of Gaussian closures, and identifies the structural condition underlying exact two-parameter reductions.

The Inositol 1,4,5-trisphosphate receptor channel (IP 3 R) is an important calcium channel involved in calcium-induced calcium release, playing a prominent role in intracellular calcium signaling. However, accurately characterizing its gating behavior remains a challenge, particularly due to the temporal resolution of patch clamp techniques that is not large enough to detect all short-lived events. This limitation can significantly bias the inference of kinetic models describing the receptor activity. To address this issue, we focused on the quantitative analysis of IP 3 R gating behavior using patch clamp data, with particular attention to missed events. We modeled IP 3 R channel gating using Hierarchical Markov chains and used a Bayesian approach that integrates missed event correction directly into the likelihood function, enabling more accurate parameter inference and model evaluation. We show that accounting for missed events deeply clarifies the multi-modal model that emerges from model selection. In this new model, the Park and Drive modes both consist of the same 3-state Markov model, with mode-dependent kinetic parameters: the Drive mode stabilizes the closed state directly connected to the open one, whereas the Park mode stabilizes the other closed state, that is not connected to the open one. Intermediate Ca 2+ concentrations are found to strongly depress the Drive to Park transition rate, so that the IP 3 R channel undergoes frequent transitions to the Park mode only for __ 50 nM or micromolar Ca 2+ concentrations. Overall, our approach provides a refined perspective on IP 3 R channel modeling and highlights the critical importance of accounting for missed events upon model selection based on single-channel recordings.

The Inositol 1,4,5-trisphosphate receptor channel (IP 3 R) is an important calcium channel involved in calcium-induced calcium release, playing a prominent role in intracellular calcium signaling. However, accurately characterizing its gating behavior remains a challenge, particularly due to the temporal resolution of patch clamp techniques that is not large enough to detect all short-lived events. This limitation can significantly bias the inference of kinetic models describing the receptor activity. To address this issue, we focused on the quantitative analysis of IP 3 R gating behavior using patch clamp data, with particular attention to missed events. We modeled IP 3 R channel gating using Hierarchical Markov chains and used a Bayesian approach that integrates missed event correction directly into the likelihood function, enabling more accurate parameter inference and model evaluation. We show that accounting for missed events deeply clarifies the multi-modal model that emerges from model selection. In this new model, the Park and Drive modes both consist of the same 3-state Markov model, with mode-dependent kinetic parameters: the Drive mode stabilizes the closed state directly connected to the open one, whereas the Park mode stabilizes the other closed state, that is not connected to the open one. Intermediate Ca 2+ concentrations are found to strongly depress the Drive to Park transition rate, so that the IP 3 R channel undergoes frequent transitions to the Park mode only for __ 50 nM or micromolar Ca 2+ concentrations. Overall, our approach provides a refined perspective on IP 3 R channel modeling and highlights the critical importance of accounting for missed events upon model selection based on single-channel recordings.

In this paper, we consider an age-structured jump model that arises as a description of continuous time random walks with infinite mean waiting time between jumps. We prove that under a suitable rescaling, this equation converges in the long time large scale limit to a time fractional subdiffusion equation.

This special issue is a selection of the best papers of the French-speaking conference on Knowledge Extraction and Management (EGC) 2025. EGC is the reference conference for the French community in Knowledge Extraction and Management since its inception in 2001. EGC 2025 will focus on fundamental research in data management and knowledge extraction (a.k.a. KDD) approaches that tackles the challenges of empowering scientific research with digitalized data and knowledge, those of the interface between KDD and other scientific disciplines, and the ethical and societal issues of their use.

International audience

National audience

Abstract We present a novel three-state gene expression model designed to elucidate the underlying mechanisms of mRNA transcription and its regulation. Our model incorporates gene regulatory processes by explicitly including a transcription factor-bound state, thereby capturing the dynamic interplay between transcription activation and chromatin dynamics. We fit the model to paired single-cell ATAC-seq and single-cell RNA-seq data, as these data give us simultaneous information on a gene’s transcriptional state and its accompanying chromatin state. Working at the pseudo-bulk level, we extract biologically meaningful high-level descriptors from homogeneous cell (sub)populations, such as the mean and variance of gene expression as well as the fraction of accessible chromatin. Crucial to the computational feasibility of our approach, these descriptors can be analytically related to our model parameters. Despite the increased complexity needed to capture regulatory processes in our model, it remains sufficiently parsimonious to infer parameters reliably from experimental data. Each parameter has a clear biological interpretation, reflecting properties such as burst frequency, chromatin opening and closing dynamics, and basal or regulated expression. Fitting the model to a large collection of genes allows us to analyze the parameters and distinguish so-called gene expression strategies. The model parameters reveal a small number of distinct expression strategies among gene clusters, providing data-driven novel insight into context-dependent regulation of gene expression.

Clustering based on belief functions has been gaining increasing attention in the machine learning community due to its ability to effectively represent uncertainty and/or imprecision. However, none of the existing algorithms can be applied to complex data, such as mixed data (numerical and categorical) or non-tabular data like time series. Indeed, these types of data are, in general, not represented in a Euclidean space and the aforementioned algorithms make use of the properties of such spaces, in particular for the construction of barycenters. In this paper, we reformulate the Evidential C-Means (ECM) problem for clustering complex data. We propose a new algorithm, Soft-ECM, which consistently positions the centroids of imprecise clusters requiring only a semi-metric. Our experiments show that Soft-ECM present results comparable to conventional fuzzy clustering approaches on numerical data, and we demonstrate its ability to handle mixed data and its benefits when combining fuzzy clustering with semi-metrics such as DTW for time series data.