bioRxiv Subject Collection: Neuroscience's Journal

Shared texture-like representations, not global form, underlie deep neural network alignment with human visual processing
Deep neural networks (DNNs) are a leading computational framework for understanding neural visual processing. A standard approach for evaluating their similarity to brain function uses DNN activations to predict human neural responses to the same images, yet which visual properties drive this alignment remains unclear. Here, we show that texture-like representations, operationalized as global summaries of local image statistics, largely underlie this alignment. We recorded electroencephalography (EEG) from 57 participants viewing three image types: natural scenes, texture-synthesized versions that preserve global summaries of local statistics while disrupting global form, and isolated objects without backgrounds. Representational-similarity analysis showed the strongest DNN-EEG alignment when both systems processed texture-synthesized images. When using features from one image condition to predict EEG responses to another, we showed that features from texture-synthesized images generalized to natural scenes. Crucially, we observed a dissociation between DNN-EEG alignment and decodable object category information: alignment increased for texture-synthesized images even when object information was reduced. Together, our findings identify global summaries of local image statistics as a common currency linking DNNs and human visual processing, clarifying that global form features are not required for high DNN-EEG alignment. Our findings highlight the shared importance of local image statistics in artificial and biological visual systems.

The human brain mechanisms of afterimages: From networks to cortical layers
Afterimages are common visual illusions that have long attracted scientific interest, yet their neural mechanisms remain little understood. We used high-spatial-resolution fMRI to investigate human whole brain and cortical layer activity in primary visual cortex (V1) linked with afterimages and perceptually matching animated images: stimuli designed to imitate the appearance of afterimages. Both afterimages and perceptually matched images engaged overlapping, widespread brain activity, particularly in visual sensory regions that follow the contralateral circuitry of the primary visual pathway. However, afterimages elicited weaker fMRI signals across many subcortical and cortical areas compared to images, except in salience network regions, where activity was enhanced for afterimages. Cortical layer-specific analyses in V1 revealed afterimages selectively engaged deep cortical layers, while images activated middle and superficial layers. In addition, we found that baseline eye measures and fMRI signals in arousal and visual networks differed depending on whether afterimages were perceived or not perceived. We argue that these results challenge the prevailing framing on the neurophysiological origins of afterimages as arising from either retinal or central neural processes. As with image perception, typical afterimages emerge by the interaction between retinal activity and central neural processing.

Effects of combined prenatal exposure to air pollution and maternal stress on social behavior and oxytocin and vasopressin systems in male and female mice
Prenatal exposures to air pollution and maternal psychosocial stress are each associated with increased risk of neurodevelopmental disorders, including autism spectrum disorder (ASD) and epidemiological work suggests that concurrent exposure to these risk factors may be particularly harmful. This is important given that the same populations often bear the brunt of both toxicant and psychosocial stress burdens. Social impairments are a defining symptom in ASD. Previous work modeling combined prenatal exposure to diesel exhaust particles (DEP) and maternal stress (MS) in rodents has found male-biased social deficits in offspring, as well as changes to neuroimmune processes and the gut microbiome. However, the precise neural circuits on which these exposures converge to impact social behavior is unclear. Oxytocin (OXT) and vasopressin (AVP) are neuropeptides critical to the regulation of social behavior across species, signaling primarily at the oxytocin receptor (Oxtr) and vasopressin V1a receptor (V1aR) in the brain. Here, we hypothesized that OXT and/or AVP expression would be reduced in the brain following DEP/MS exposure. Following prenatal exposure to DEP/MS or the vehicle/control condition (VEH/CON), we measured maternal and offspring outcomes during the perinatal period, social and anxiety-like behavior during adolescence, OXT and AVP cell/fiber density and Oxtr and Avpr1a mRNA expression in early adulthood in several brain regions in both males and females. We observed a decrease in interaction time in DEP/MS males as compared to VEH/CON in the sociability assay and a decrease in social novelty preference in DEP/MS females as compared to VEH/CON. No effects of sex or treatment were observed on OXT or AVP cell number or fiber density in the hypothalamic regions assessed. However, numerous sex differences were observed in Oxtr and Avpr1a mRNA. Moreover, Avpr1a mRNA was significantly increased following DEP/MS exposure in the nucleus accumbens in both sexes and tended to increase in the dorsal hippocampus. Conversely, Avpr1a mRNA tended to decrease in the amygdala in both sexes following DEP/MS exposure. Together, these findings suggest that DEP/MS exposure has a stronger impact on female social behavior than previously observed. Moreover, while DEP/MS exposure does not appear to impact OXT or AVP expression in the brain, V1aR expression is modulated by DEP/MS exposure in several brain regions.

Responses to taste and eating in the parabrachial nucleus of the pons in awake, unrestrained rats
Background/Objectives. The parabrachial nucleus of the pons (PbN) is a hub in the central pathway for taste in non-primate mammals. Recent evidence has identified a role for the PbN in regulating ingestion; however, little is known about how the PbN responds to solid food. Here, we recorded PbN responses to liquid taste stimuli over days/weeks, and tested whether these responses are good predictors of responses to solid foods consumed in a naturalistic way. Methods. Rats were prepared for one-photon calcium imaging by surgical implantation of a GRIN lens. PbN activity was imaged during multiple sessions spanning up to 112 days while animals licked various tastants in an experimental chamber (Lick phase). In some sessions, following the Lick phase, rats were presented with Granny Smith apples, milk chocolate and/or salted peanuts (Food phase). In one session, chocolate or peanut odor was presented during the Lick phase along with the tastants. Results. PbN cells responded to more than one taste quality with tastant-specific spatiotemporal patterns of response. While response profiles of individual PbN cells shifted over days/weeks, across-unit patterns differentiating taste quality remained relatively stable. Population responses to liquid tastants and solid foods were segregated according to the different motor patterns required for ingestion. Responses to liquid tastants were not good predictors of responses to solid foods. Importantly, PbN cells also responded to food-related odorants. Conclusions. Collectively, these results challenge the traditional view of the PbN as a simple relay for taste and instead position it as an integrative hub involved in processing gustatory, olfactory, and somatosensory information. Moreover, the findings emphasize the importance of population coding in maintaining perceptual stability.

Structural Compression and Entorhinal Vulnerability: Linking Tentorial Adjacency to Tau Burden and Dementia Progression
Alzheimer's disease (AD) is a growing public health crisis. The disease is defined neuropathologically by accumulation of amyloid-{beta} plaques and neurofibrillary tangles (NFTs) composed of abnormal tau protein in the brain. Early neurofibrillary degeneration in the entorhinal cortex (EC) is a hallmark of AD and a critical initiating event in the hierarchical pathoanatomical progression. However, the factors triggering initial tau deposition in the EC remain unclear. We propose a novel biomechanical cascade hypothesis, positing that the unique anatomical inferomedial positioning of the EC, including proximity to the tentorial incisura (TI) and other skull base structures, renders it susceptible to very mild yet persistent age-related mechanical stress, analogous to the effects of repetitive mild traumatic brain injury, triggering tau pathology. To test this hypothesis, we developed a method to quantify Entorhinal-Tentorial (EC-TI) adjacency and applied it to multimodal imaging data from the Alzheimer's Disease Neuroimaging Initiative (ADNI; n=47). Based on this neuroanatomical contact coefficient (NCC), participants were heuristically stratified into high (n=24) and low (n=23) adjacency groups. When controlling for other risk factors, tau PET signal in the EC predicted conversion from mild cognitive impairment to AD only in the high-adjacency group (LLR p=0.009, tau PET in EC p=0.036). These findings identify EC-TI adjacency as a novel and anatomically grounded biomarker of AD progression risk. More broadly, they suggest a previously unrecognized biomechanical contribution to the initiation of tau pathology in aging and sporadic AD, opening new avenues for early detection, risk stratification, and mechanistically targeted prevention strategies.

Development of the Olfactory and Vomeronasal Systems in the Fossorial Water Vole (Arvicola scherman). I. The Late Prenatal Stages
stages of life, yet most of what is known about the prenatal development of the olfactory and vomeronasal systems comes from laboratory rodents. These models, while invaluable, may not fully represent the developmental trajectories of wild species living under natural ecological pressures. Here we investigated the fetal development of the nasal chemosensory systems in the fossorial water vole (Arvicola scherman), a free-living arvicoline rodent with a highly subterranean lifestyle. We analyzed fetuses at embryonic days E17 and E21 (term) using classical histology, immunohistochemistry (markers: Gi2, Go, G{gamma}8, CB, CR, PGP 9.5, GAP-43, {beta}-tubulin, MAP2), and lectin histochemistry (UEA, LEA, SBA, STA, DBA). This combined approach enabled us to assess structural maturation, neuronal differentiation, and the temporal dynamics of glycoconjugate expression in the vomeronasal organ (VNO), olfactory epithelium (OE), and the main (MOB) and accessory olfactory bulbs (AOB). By E21, the MOB displayed a six-layered adult-like organization with well-defined glomeruli and interneuronal populations, whereas the AOB showed delayed morphological maturation but already exhibited selective molecular signatures in its nerve and superficial layers. Prenatally, the VNO underwent conspicuous structural differentiation, including stratification of the sensory epithelium, robust axonal fasciculation, and early development of vomeronasal glands. Immunohistochemical analysis revealed early expression of G-protein subunits and calcium-binding proteins, indicating premature pathway specification and interneuronal circuit formation. Lectin labeling provided additional insights: SBA emerged as a highly selective marker of the vomeronasal pathway; UEA highlighted early compartmentalization of vomeronasal projections; LEA showed a conserved, pan-chemosensory binding pattern across systems; and DBA, despite its lower specificity, revealed late-onset reactivity in postmitotic neurons. Together, these findings demonstrate that A. scherman exhibits a remarkably accelerated prenatal maturation of its chemosensory systems compared with laboratory rodents. This early functional readiness likely reflects adaptive pressures of a fossorial lifestyle, emphasizing the importance of incorporating wild species into developmental neurobiology to refine our understanding of mammalian chemosensory evolution.

Evaluating BOLD functional MRI biophysical simulation approaches: impact of vascular geometry, magnetic field calculations, and water diffusion models
Biophysical simulations have guided the development of blood oxygenation level-dependent (BOLD) functional MRI (fMRI) acquisitions and signal models that relate the BOLD signal to the underlying physiology, such as calibrated BOLD and vascular fingerprinting. Numerous simulation techniques have been developed, however, few of them have been directly compared, thus limiting the assessment of the accuracy and interchangeability of these methods as well as the accuracy of the quantitative techniques derived from them. In this work, we compared the accuracy and computational demands of eight previously published simulation approaches that adopt different geometries (ranging from infinite cylinders to synthetic vascular anatomical networks (VANs)), field offset calculations (analytical and Fourier-based), and water diffusion implementations (Monte Carlo and convolution-based), all of which are available in an open-source Python toolkit, BOLDs{omega}imsuite. The reference simulation approach for comparison used three-dimensional infinite cylinders, analytical field offsets, and Monte Carlo diffusion. When compared with the reference approach, most of the simulations, including two- and three-dimensional geometries, were in excellent agreement when assuming the intravascular signal contribution was small. Two commonly employed simulation approaches were notably biased; both used two-dimensional geometries with overly simplified vasculature or field offset calculations. In general, the simulated intravascular signal was the least consistent across approaches, thus potentially resulting in larger errors when the intravascular signal contribution is large. Lastly, the VAN results were in good agreement with the reference but they diverged slightly, yet systematically, from each other at smaller radii ([lsim] 3 m), primarily driven by intravascular signal differences. We conclude, therefore, that the reference approach is an attractive option for exploratory simulations in the many cases where anatomical and hemodynamic realism is not needed, balancing ease of implementation, accessibility, versatility, computational efficiency, accuracy of results, and interpretability. These findings help pave the way for a broader adoption of forward modelling of the BOLD signal and more reliable interpretations of biophysical simulations aiming to develop quantitative models of the BOLD signal.

High affinity cross-context cellular assays reveal novel protein-protein interactions of peripheral myelin protein of 22 kDa
Peripheral Myelin Protein 22 (PMP22) is a tetraspan membrane protein whose altered dosage causes the most common hereditary neuropathy, Charcot-Marie-Tooth disease type 1A (CMT1A). Despite its clinical significance, the physiological functions of PMP22 and the mechanism behind its tightly controlled gene dosage sensitivity remain unknown since over 30 years, in part due to limited knowledge of its protein-protein interactions (PPIs). In fact, integral membrane proteins such as PMP22 are significantly underrepresented in known cellular interactomes, likely due to limited suitability or technical challenges specific to these hydrophobic molecules in the major PPI discovery approaches. Here, we applied a rigorously optimized co-immunoprecipitation and mass spectrometry workflow using the mild detergent DDM and the high affinity ALFA-tag/anti-ALFA nanobody interaction to identify physiologically PMP22-associated proteins. In a cross-context approach, we ran our standardized pipeline across multiple cell types including HEK293T, MDCKII epithelial cells, the Schwann cell line MSC80, and primary rat Schwann cells. We confirm known interactors, and functional annotation analysis revealed cell-type specific enrichment patterns, with adhesion-related PPIs predominating in MDCKII cells (e.g., CD47, CLDN1, ATP1B1) and myelin-associated proteins enriched in Schwann cells. Importantly, we identified novel PPI candidates that may be highly relevant for PMP22 function including enzymes of the de novo sphingolipid biosynthesis pathway.

Functional imaging of hippocampal layers using VASO and BOLD on the Next Generation (NexGen) 7T Scanner
Spatial accuracy and venous biases are a central concern in mesoscale fMRI, with subcortical brain regions facing additional challenges due to lower contrast-to-noise ratio (CNR), high physiological noise, and complicated vasculature. Here, we optimized CBV VASO on the NexGen 7T scanner for layer-specific investigations of the hippocampus. The presence of venous biases in VASO and BOLD (from the same acquisition) was then compared by using an established autobiographical memory task. While VASO and BOLD based activation patterns converged at macroscale, layer-specific differences emerged in the hippocampal subiculum, consistent with venous bias in the inner layers of the subiculum which can be explained by the unique two-sided venous drainage. Further, both VASO and BOLD showed sensitivity to short blocks (elaboration > construction), revealing an anterior-posterior distinction consistent with stronger involvement of the posterior hippocampus. Hippocampal cortical connectivity revealed brain circuitry between subcortical and cortical regions. Thus, hippocampal fMRI allows mapping layer function with high accuracy, made possible by sequence timing optimization on the high performance NexGen 7T scanner. The improved MR imaging has been developed to enable precision mapping of subcortical brain gray matter. By capturing changes of neural information flow within and across the microcircuitry of the hippocampus, it can provide deeper insights into a number of neuropsychological phenomena and the early changes occurring in Alzheimer's disease (AD) and mild cognitive impairment (MCI).

Glial plasticity and metabolic stability after knockdown of astrocytic Cx43 in the dorsal vagal complex
Obesity affects well over 890 million people worldwide and causes millions of deaths each year due to metabolic complications, making it a major public health challenge. It results from a chronic imbalance between caloric intake and energy expenditure. This balance is regulated by the central nervous system, primarily by the hypothalamus and the dorsal vagal complex (DVC). The latter integrates metabolic signals from energy stores and gastrointestinal tract and coordinates autonomic responses. While historically overshadowed by a focus on neurons, the role of glial cells in regulating energy balance is now well established. Connexin 43 (Cx43) is a well-known protein expressed by astrocytes, playing a key role in glial and neuroglial communication. We previously demonstrated that pharmacological inhibition of Cx43 hemichannels (HCs) in the hypothalamus and the DVC led to anorexia in mice, accompanied by neuronal activation in both regions. To further investigate the role of astrocytic Cx43 within the DVC, where its expression is remarkably high, we developed a mouse model in which Cx43 expression is specifically reduced in DVC astrocytes using an RNA interference approach. The metabolic profile of these animals was assessed under standard feeding conditions as well as under a high-fat, high-sugar diet. Although reduced Cx43 expression led to modified glial (astrocytes and microglia) morphology and phenotype within the DVC, our analyses did not reveal significant changes in the animal's metabolic phenotype. This result is surprising, given the high expression of Cx43 in the DVC and the anorexigenic effect observed with pharmacological HC inhibition. These observations raise crucial questions, about potential compensatory mechanisms involving other cell types and/or alternative neuron-glia communication pathways, as well as the DVC's ability to maintain robust control of energy homeostasis.

Update on the reproduction and interpretation of DIANA fMRI
Three years ago, our group reported direct imaging of neuronal activity (DIANA) with high spatiotemporal resolution, but its reproducibility and signal origin remain controversial. Here, we report the results of our reproduction experiments of DIANA fMRI performed to date using forelimb electrical stimulation at various magnetic field strengths in anesthetized mice, along with the characteristics of DIANA signal, called the pseudo-steady state (PSS). Theoretical analysis and Bloch simulations demonstrated that the spatial location and temporal phase of PSS oscillations are primarily determined by frequency-offset, and that their spatiotemporal superposition can generate peak signals in specific regions at specific timing that closely resemble DIANA signals. These findings suggest that if PSS oscillations are the primary source of DIANA signals, it may be premature to interpret them as neuronal responses to sensory stimulation. Further studies are needed to clarify the relationship between DIANA signals and brain activation.

Inhibited oligodendrogenesis, but not repeated mild traumatic brain injury, impairs attention in adult mice
Attention problems are among the most common long-lasting cognitive symptoms of mild traumatic brain injury (mTBI) and as attention is fundamental to many aspects of cognition, the effects of attentional impairment can be broad. The brain's white matter is particularly vulnerable to damage during mTBI. Damage to oligodendrocytes and myelin contributes to cognitive deficits following injury and myelin plasticity is a potential mechanism for functional recovery. The aim of this work was to assess attentional impairment following mTBI in mice and evaluate the role of newly generated oligodendrocytes in recovery. This study used the Myrf conditional knockout mouse model, in which the Myrf gene, required for oligodendrocyte precursor cell (OPC) differentiation into mature myelinating oligodendrocytes, is deleted from OPCs following tamoxifen injection, thereby halting oligodendrogenesis. Mice were trained on the 5-Choice Serial Reaction Time (5-CSRT) task before receiving tamoxifen followed by three mTBI or sham procedures. Attention was probed on the 5-CSRT with decreasing stimulus durations at four time points following injury out to 12 weeks. While no attentional impairment was observed in mice with mTBI, OPC-MyrfKO mice showed lower choice accuracy ({beta} = -2.85%, p = 0.03) and more omitted trials ({beta} = 4.09%, p = 0.02) across injury groups, time points and stimulus durations, suggesting that active oligodendrogenesis is required for sustained attention. These results suggest that this mouse model of mTBI was not severe enough to impact attention as measured by the 5-CSRT task, however myelin plasticity in adulthood may contribute to attention and complex task performance.

Output-Contingent Working Memory and Decision-Making in Economic Choices
In daily life, economic decisions often unfold sequentially. The brain is thought to compute the subjective value of each option, and comparisons can occur in different reference frames, for example, based on the commodity or presentation order. While primate prefrontal recordings have identified various reference frames for economic choice, it remains unclear how distinct neural mechanisms support them even in similar tasks. To address this, we trained recurrent neural networks (RNNs) on two sequential economic decision-making tasks differing only in output contingencies. Analysis of RNN activity, combined with latent connectivity inference, revealed distinct regimes: commodity-based choices with attractor dynamics and order-based choices with rotational dynamics. Moreover, value and choice representations in the order-based tasks aligned with neural data from a novel experiment where reference frames were not explicitly constrained. Our results suggest that different reference frames emerge depending on task demands and engage distinct working memory and decision-making mechanisms.

Fear, anxiety, and the extended amygdala- Absence of evidence for strict functional segregation
Since the time of Freud, the distinction between fear and anxiety has been a hallmark of influential models of emotion and emotional illness, including the Diagnostic and Statistical Manual of Mental Disorders (DSM) and Research Domain Criteria (RDoC) framework. Fear and anxiety disorders are a leading cause of human misery and morbidity. Existing treatments are inconsistently effective, underscoring the importance of developing accurate models of the underlying neurobiology. Although there is agreement that the extended amygdala (EA) plays a central role in orchestrating responses to threat, the respective contributions of its two major subdivisions- the central nucleus of the amygdala (Ce) and bed nucleus of the stria terminalis (BST)- remain contentious. To help adjudicate this debate, we performed a harmonized mega-analysis of fMRI data acquired from 295 adults as they completed a well-established threat-anticipation paradigm. Contrary to popular double-dissociation models, results demonstrated that the Ce responds to temporally uncertain threat and the BST responds to certain threat. In direct comparisons, the two regions showed statistically indistinguishable responses, with strong Bayesian evidence of regional equivalence. These observations underscore the need to reformulate conceptual models that posit a strict segregation of temporally certain and uncertain threat processing in the EA.

Connectome-Based Predictive Modeling of Concurrent and Prospective Substance Use in Adolescence
Understanding the neural mechanisms of adolescent substance use is a critical public health issue, with direct implications for bolstering prevention and treatment strategies. Yet this effort is challenging because substance use is multi-faceted, commonly used brain network features are not optimized to capture both local and global aspects of intrinsic connectivity, and because the facets themselves sensitive to developmental shifts. In this study, we operationalized adolescent substance use along three dimensions--intent, access, and family-developmental history--and trained predictive models of each facet using resting-state connectivity. Trait impulsivity, a known risk factor, was also examined. Using Baseline and 2 Year Follow-Up data from the ABCD Bids Community Collection (ABCC), we found that prediction was more successful at follow-up than baseline. At baseline, predictive accuracy was modest and intent to use substances was the most accurately predicted facet. Prediction accuracies at follow-up were much higher, with access and family-developmental history being better predicted, signaling a developmental shift in the brain-behavior mapping of substance use vulnerabilities. These findings suggest that the neurobiological correlates of substance are dynamic across adolescence, possibly reflecting changing phenotype. More broadly, these results underscore the importance of modeling distinct substance use facets and accounting for developmental timing to understand risk trajectories, while contributing to a growing literature that shows early-developing individual differences are predictive of later outcomes.

Multiomic profiling reveals aberrant immunomodulatory signature in β-propeller protein-associated neurodegeneration patient iPSC-derived microglia
Microglia are the primary immune cells of the central nervous system and play a crucial role in maintaining brain homeostasis. In common neurodegenerative diseases such as Alzheimer's disease and Parkinson's disease (PD), early and sustained microglial activation has been shown to precede neuronal loss, with elevated levels of microglia-derived inflammatory mediators detected in affected brain regions. In contrast, little is known about the role of microglia in rare neurodegenerative disorders. One such disorder is {beta}-propeller protein-associated neurodegeneration (BPAN), a common subtype of neurodegeneration with brain iron accumulation (NBIA). BPAN shares pathological features with PD, including iron accumulation and selective loss of dopaminergic neurons in the substantia nigra, and is caused by mutations in the WD repeat domain 45 (WDR45) gene encoding an autophagy protein also called WIPI4. However, the pathological role of mutant WDR45 in BPAN and the possible contribution of microglia remain unresolved. We generated the first BPAN patient microglia model system using induced pluripotent stem cells (iPSCs) to identify immune-related alterations and immunomodulatory signaling changes in a disease-relevant context. Integrated transcriptomic and proteomic profiling of iPSC-derived microglia from BPAN patients revealed a consistent shift from a homeostatic to a reactive, disease-associated state. Transcriptomic analysis showed disruption of core microglial pathways, including immune activation, stress response, and autophagy, consistent with a chronic pro-inflammatory phenotype. Complementary secretome analysis identified impaired lysosomal function and increased antigen presentation pathways, further supporting persistent microglial activation. Together this suggests that dysfunctional microglial states may contribute to BPAN pathogenesis. Our findings lay the groundwork for advancing immunomodulatory research in BPAN and may open new avenues for therapeutic development targeting microglial dysfunction.

Personalized real-time inference of momentary excitability from human EEG
The efficacy of transcranial magnetic stimulation (TMS) is often limited by non-adaptive protocols that disregard instantaneous brain states, potentially constraining therapeutic outcomes. Current EEG-guided approaches are hindered by their reliance on motor-evoked potentials (MEPs), which confound cortical and spinal excitability and restrict applications to the motor cortex, and a dependence on static biomarkers that cannot adapt to changing neurophysiological patterns. We introduce PRIME (Personalized Real-time Inference of Momentary Excitability), a deep learning framework that predicts cortical excitability, quantified by TMS-evoked potential (TEP) amplitude, from raw EEG signals. By targeting cortical excitability directly, PRIME enables brain state-dependent stimulation across any cortical region. PRIME incorporates transfer learning and continual adaptation to automatically identify personalized biomarkers, allowing stimulation timing to be adapted across individuals and sessions. PRIME successfully predicts cortical excitability with minimal latency, providing a computational foundation for next-generation, personalized closed-loop TMS interventions.

Neuromedin U Receptor 2 orchestrates sleep continuity and circadian entrainment
Neuromedin S signaling has recently been identified as a modulator of both sleep and circadian rhythms in the central nervous systems of mammals. Nevertheless, the involvement of its preferred receptor, NMUR2, in sleep and circadian regulation remains elusive. Here, employing its only selective antagonist R-PSOP, we demonstrate that NMUR2 function in the mouse brain is essential for physiological sleep architecture under baseline conditions, as well as for responses to homeostatic and circadian challenges. Using a combination of pharmacological, neuronal tagging, brain clearing and chemogenetic tools, we identified and functionally validated a group of PVH neurons as one of the targets mediating the effect of R-PSOP on sleep. These findings establish NMUR2 as a potential therapeutic target for the integral treatment of sleep and circadian neurological disorders.

IRF8 deficiency causes anxiety-like behavior in a sex-dependent manner
Anxiety disorder is a serious psychiatric disease that affects women twice more than men and disrupts patients' daily lives. It is often comorbid with major depression and other mental diseases. Various underlying mechanisms have been proposed, such as neurotransmitters and neuroanatomical disruptions, and more recently, oxidative stress; however, much remained unclear, including the role of glial cells. Here, we investigated the role of IRF8 in anxiety disorders in the mouse model. IRF8 is a transcription factor expressed primarily in microglia in the brain. A battery of behavioral tests revealed that female IRF8 knockout (IRF8KO) mice show increased anxiety relative to male IRF8KO and wild-type mice. Female IRF8KO mice also exhibited a higher tendency for obsessive-compulsive disorder. However, these behavioral abnormalities were not observed when IRF8 was deleted postnatally, indicating that it acts during the fetal stage to control anxiety. Transcriptome analysis revealed that IRF8 deficiency leads to redox dysregulation. Further, 2',7'-dichlorofluorescin diacetate (DCFDA) staining for microglia demonstrated that female IRF8KO microglia produce higher levels of reactive oxygen species (ROS) compared to WT and male IRF8KO counterparts. Detailed RNA-seq analysis, however, did not reveal specific genes that cause high ROS production in female cells. In sum, this work demonstrates that IRF8 in microglia plays a major role in controlling anxiety in a sex dependent manner.

Early endothelial activation at the blood-nerve barrier defines a hallmark of ALS
Vascular defects are common in Amyotrophic Lateral Sclerosis (ALS). The prevailing view is that breakdown of the blood-brain and blood-spinal cord barriers contributes to neurodegeneration. Here, we reveal that selective and early vulnerability of peripheral nerve endothelium --manifested as endothelial cell activation and interlinked blood-nerve barrier dysfunction-- constitutes a core feature of ALS pathogenesis, arising before vascular alterations in the central nervous system (CNS) and motor neuron pathology. Vascular changes in ALS patients have been largely studied in postmortem samples, limiting insight into their onset, causes, and pathogenic role. Surveying diagnostic motor nerve samples from ALS patients across clinical stages, we observed endothelial damage that preceded axonal loss and demyelination, marking vascular dysfunction as an early disease event. Similar ultrastructural abnormalities were detected in pre-symptomatic ALS mouse models (SOD1 and TARDBP mutants). Notably, endothelial cells became dysfunctional even when not carrying mutant TDP-43, indicating they respond to non-cell autonomous disease signals. Transcriptional, histological, and functional analyses revealed that these alterations were largely confined to peripheral nerves, while spinal cord vessels exhibited delayed and more focal changes. Single-cell sequencing identified ALS-susceptible endothelial cell subsets prone to a pro-inflammatory phenotype and impaired blood-nerve barrier function, increasing permeability via the transcellular route. These changes coincided with reactivity of nerve-resident macrophages and neutrophil infiltration. Neutrophil depletion attenuated endothelial activation and barrier leakiness, mitigating axonopathy in ALS mice. Our work unmasks the greater susceptibility of the peripheral nerve vasculature in ALS relative to the CNS. The early activation of peripheral nerve endothelium, combined with its potential reversibility, identifies a therapeutic window and suggests strategies for targeting the vascular-immune axis to protect the motor system.

Zolpidem restores sleep and slows Alzheimer's progression in a mouse model
INTRODUCTION: Deficits in Non-Rapid Eye Movement (NREM) sleep facilitate Alzheimer's disease (AD) progression. Enhancing GABAergic signaling can restore sleep. Unbiased computational analysis identified zolpidem as high-affinity GABA receptor modulator facilitating chloride transport that could slow AD. METHODS: Zolpidem's effects on sleep and Alzheimer's progression were evaluated in young APP/PS1 mice. Sleep was monitored with EEG/EMG telemetry. Widefield imaging with voltage-sensitive dyes was used to track sleep-dependent brain rhythms. Multiphoton microscopy allowed assessments of amyloid plaque load and basal neuronal calcium levels. Behavioral assays were used to measure memory and cognitive function. RESULTS: Zolpidem restored NREM sleep and rescued sleep-dependent brain rhythm, slow oscillation. Zolpidem administration reduced cortical amyloid plaque burden, mitigated neuronal calcium overload, and enhanced sleep-dependent memory consolidation without adverse effects on locomotion. DISCUSSION: Zolpidem effectively slowed Alzheimer's progression in young APP/PS1 mice. This supports zolpidem's therapeutic promise as an intervention strategy at early stages of AD.

Satiation is associated with OGT-dependent regulation of excitatory synapses.
Satiation is essential for energy homeostasis and is dysregulated in metabolic disorders like obesity and eating disorders such as anorexia nervosa. While satiation engages a large neural network across brain regions, how the communication within this network depends on metabolic fluctuations is unclear. This study shows that nutrient access can affect neuron-to-neuron communication in this network by regulating excitatory synaptic plasticity through O-GlcNAc transferase (OGT) in CaMKII satiation neurons in the paraventricular nucleus (PVN). Using cell-specific knockout mice and electrophysiological recordings, we demonstrate that OGT deletion in PVNCaMKII neurons increases input resistance and neuronal excitability while preserving basic membrane electrical properties. Strikingly, feeding triggered a robust 3.8-fold increase in the excitatory synaptic input in wild-type neurons, whereas OGT-knockout neurons failed to exhibit this feeding-induced synaptic activation, instead displayed a paradoxical trend towards increases in synaptic activity during hungry conditions. Furthermore, OGT deletion destabilized glucose-dependent synaptic responses, with knockout neurons displaying maladaptive depression of excitatory transmission in conditions where stability is normally preserved. These findings establish OGT as a nutrient-sensitive modulator of synaptic plasticity that ensures appropriate satiation signalling by coupling metabolic state to synaptic plasticity.

Cryo-EM Structures of Higher Order Gephyrin OligomersReveal Principles of Inhibitory Postsynaptic Scaffold Organization
Gephyrin is the principal scaffolding protein of inhibitory postsynaptic densities, clustering glycine and GABAA receptors via multivalent interactions. It features structured N and C terminal domains connected by an intrinsically disordered linker. Although the structural and functional properties of its terminal domains are well characterized, the mechanism by which full-length gephyrin organizes into higher-order complexes remains unresolved. Here, we combine biochemical reconstitution, cryo-electron microscopy, and mutational analyses to elucidate the structural logic of gephyrin oligomerization. We demonstrate that gephyrin adopts a stable dimeric assembly which constitutes the basic unit for both linear and oblique tetramers as well as linear hexameric arrangements. High resolution structures reveal a critical segment of the flexible linker that adopts two distinct conformations, one of which occludes the receptor-binding site. This segment harbors key phosphorylation sites, providing a mechanistic link between structural conformation and regulatory control. Our findings redefine the architecture of inhibitory synapses and reconcile gephyrin oligomerization models with published in-situ post-synaptic densities characterized by cryo-electron tomography.

Replicable, Transdiagnostic Behavioral and Neural Correlates of Sensory Over-responsivity
Sensory over-responsivity (SOR), characterized by strong negative reactions to typically innocuous stimuli, is considered a symptom of autism spectrum disorder. However, SOR also affects 15-20% of children overall, including a majority of children with common psychiatric conditions. Despite its prevalence, the clinical specificity and neurobiological bases of SOR remain poorly understood. Our study aims to determine the specific clinical significance of SOR across diverse child samples and establish whether SOR is associated with distinct patterns of functional connectivity (FC). We analyzed data from 15,728 children (ages 6-17.9 years) across five datasets: three community samples, including the Adolescent Brain Cognitive Development [ABCD] and Healthy Brain Network [HBN] studies, and two autism-enriched samples. Bivariate and multivariate models examined associations between SOR and symptoms of anxiety, attention-deficit/hyperactivity disorder, depression, conduct disorder, and oppositional defiant disorder, as well as autistic traits. Analysis of resting-state functional MRI (fMRI) data from the ABCD study tested brain-wide and circuit-specific FC correlates of mild SOR, replication of effects in an independent ABCD subsample, as well as extension to severe SOR in ABCD and to SOR in the HBN sample. Multivariate analyses revealed that SOR is associated with a remarkably consistent transdiagnostic profile: greater levels of both autistic traits and anxiety symptoms and, in community samples, lower levels of conduct disorder symptoms. Across samples, SOR is not reliably associated with symptoms of any other analyzed psychiatric conditions. Resting state fMRI analyses reveal that SOR is associated with both brain-wide and circuit-specific functional connectivity (FC) patterns in networks related to tactile processing (somatomotor-hippocampal) and error detection (cingulo opercular-ventral diencephalon) that replicate in independent subsamples. Our results demonstrate that SOR may constitute a transdiagnostic latent trait with both specific clinical risk and protection, and with replicable neural correlates that implicate specific cortico-subcortical circuits. These findings advance our understanding of the neurobiology and clinical relevance of SOR. They may also inform clinical practice and future research aimed at understanding and supporting individuals with sensory challenges.

Comparative analysis of RAN translation from CAG repeats within the Huntingtin coding sequence
Repeat associated non-AUG (RAN) translation is a non-canonical initiation event that occurs in the absence of a start codon in repeat expansion disorders, generating aggregation-prone proteins which may contribute to pathogenicity. The mechanism by which repeats trigger RAN translation is not completely understood, with most prior work focused on how repeats might elicit initiation when placed within a 5 leader in reporter plasmids. However, RAN translation is also reported to generate out-of-frame proteins in Huntington disease (HD), where a CAG repeat expansion in exon 1 of the Huntingtin gene (HTT) resides within the coding sequence of the gene. To explore this process, we generated a series of RAN translation-specific reporter constructs that include the 5 leader and first exon of HTT and compared their translation to CGG and GGGGCC repeats. CAG repeats support RAN translation in both the alanine (GCA) and glutamine (CAG) frames in both the presence and absence of upstream start codons or near-AUG cognate codons. HTT RAN translation in the alanine frame is comparable in efficiency to polyalanine RAN translation from CGG repeats and exhibits cap-dependence and selective enhancement by activation of the integrated stress response. Importantly, this translation was readily detectable from in vitro transcribed RNAs transfected into neurons or cellular lysates, suggesting that plasmid based aberrant splicing into CAG repeats does not explain the observed phenomena. CAG repeats in the context of HTT exon 1 elicit neuronal toxicity in the absence of any AUG initiation codons. Taken together, these data suggest that RAN translation shares key mechanistic parameters across different repeat sequences and surrounding RNA contexts with implications for therapy development.

Direct Evidence for Dendritic Spine Compensation and Regeneration in Alzheimer Disease Models
Dendritic spine loss in Alzheimer disease (AD) strongly correlates with cognitive decline, whereas spine preservation is associated to cognitive resilience. Yet, whether and how neurons compensate for spine loss in AD remains largely unknown. Using a chromophore-assisted light inactivation approach (CALI), we developed a tool to selectively eliminate dendritic spines to model this key feature of AD. Using in vivo and in vitro two-photon imaging, we discovered that the artificial elimination of spines triggers a two-stage compensatory response: rapid enlargement of remaining spines followed by delayed spine regeneration. Remarkably, similar structural plasticity was observed across multiple Amyloid beta-driven models of synapse loss, including the APP/PS1 mouse and following intracortical delivery of oligomeric Amyloid beta. Mechanistically, compensatory spine enlargement required NMDA receptor activation and de novo protein synthesis. These findings suggest that neurons retain an intrinsic capacity to reverse early synaptic loss in AD, potentially contributing to cognitive resilience

Multiplexing behavioral signals in sensory representations
Activity of sensory neurons is influenced not only by external stimuli but also by the animal's behavioral state. It is well documented that behavior influences the general properties of neural activity, such as response gain. However, it is not known whether it could affect the sensory tuning of individual neurons in a more refined way and what the functional benefit of such nuanced modulation might be. Here, we investigate this in the mouse visual cortex, where sensory response gain varies with locomotion speed. First, using numerical simulations, we demonstrate that gain modulation can multiplex behavioral information in sensory populations, without compromising the accuracy of sensory coding. To implement such multiplexing, behavioral signals should modify the sensory tuning in individual neurons. Second, we analyze neural activity in the mouse visual cortex made available by the Allen Brain Observatory. Our analysis indicates that, in agreement with the theory, locomotion-induced gain modulation can modify the tuning of sensory neurons to direction of visual motion. In that way, the visual cortex could instantiate an accurate, joint representation of sensory and movement-related signals and support computations that simultaneously require both types of information.

Sex differences in the developing human cortex intersect with genetic risk of neurodevelopmental disorders
Autism is highly heritable and diagnosed more frequently in males than females. To identify neurodevelopmental processes that might present sex-biased vulnerability, we generated transcriptomic and epigenomic profiles of cell types present in the prenatally developing human cerebral cortex of 27 males and 21 females. By intersecting sex-biased molecular signatures and genes with de novo mutations in male and female autistic probands, we reveal two points of vulnerability contributing to the sex-biased penetrance in neurodevelopmental disorders (NDDs). First, we show that NDD risk genes are biased towards higher expression in females, identifying the NDD gene MEF2C as a critical transcription factor for female-biased expression. Second, we identify a significant contribution of X chromosome genes to NDD pathobiology. We construct a gene regulatory map of X-linked risk genes to enable functional studies of genetic variants that likely disrupt gene expression in the developing brains of autistic males. Together, these results point towards an outsized contribution of the X-chromosome to both the origin of sex differences in the developing human cortex and NDD vulnerability. We propose a model where female-biased vulnerability is driven by coding variation within genes while male-biased vulnerability is driven by noncoding variation in regulatory elements that affect gene expression.

Emergence of Balanced Cortical Activity via Calcium-Regulated Synaptic Homeostasis
Cortical circuits must stabilize activity while retaining the variability and flexibility essential for computation. This raises a fundamental question: how can excitatory (E) and inhibitory (I) synapses co-adapt through homeostatic plasticity without disrupting network function or relying on fine-tuned parameters? We propose a solution grounded in the multidimensional nature of intracellular calcium signaling, which independently regulates protein synthesis at E and I synapses. By analytically characterizing calcium dynamics driven by spike-train statistics, we show that calcium's mean encodes firing rate, while its variance reflects spike-time irregularity, two complementary features critical for stable yet flexible spiking. Leveraging this dual signal, we construct a closed-loop model in which inhibitory synapses are regulated by calcium's mean and excitatory synapses by its variance through independent pathways. This mechanism preserves irregular spiking and stabilizes firing rates across diverse inputs. Strikingly, it also yields the empirically observed weakening of synaptic strengths with the number of inputs K as 1{surd}K, leading to the spontaneous emergence of balanced excitatory-inhibitory dynamics. These results uncover a calcium-driven regulatory principle linking intracellular signaling to the origin of balanced activity in cortical networks.

Transient Early Postnatal Neuronal Hyperexcitation Results in Lasting Social Preference Deficits
Perturbations during critical periods of neurodevelopment are implicated in the etiology of autism spectrum disorder (ASD), a condition marked by considerable heterogeneity and prevalent comorbidities. One such comorbidity is epilepsy, with approximately 30% of children with ASD experiencing seizures and a similar proportion of children with epilepsy displaying ASD-like symptoms. Both conditions have been associated with disruptions in the brain's excitation-to-inhibition (E/I) balance. Although early-life seizures in rodents have been linked to social impairments, direct causal evidence connecting E/I imbalance, interneuron development, and social behavior remains limited. To address this gap, we induced transient, brain-wide hyperexcitation in neonatal mice using pentylenetetrazol (PTZ), a GABA_A receptor antagonist. We administered both convulsive and subconvulsive doses and assessed long-term effects on social behavior, cortical E/I balance, and parvalbumin (PV) interneuron development. PTZ-treated groups displayed impaired social preference as measured in the 3-chamber test and increased PV interneuron density within the medial prefrontal cortex. These findings highlight a critical developmental window during which E/I imbalance leads to social deficits characteristic of ASD and epilepsy. They also reveal dose-dependent neurobiological changes, underscoring the importance of early-life neural activity in shaping social circuitry.

Flexible navigation with neuromodulated cognitive maps
Animals develop specialized cognitive maps during navigation, constructing environmental representations that facilitate efficient exploration and goal-directed planning. The hippocampal CA1 region is implicated as the primary neural substrate for cognitive mapping, housing spatially tuned cells that adapt based on behavioral patterns and internal states. Computational approaches to modeling these biological systems have employed various methodologies. Although labeled graphs with local spatial information and deep neural networks have provided computational frameworks for spatial navigation, significant limitations persist in modeling one-shot adaptive mapping. We introduce a biologically inspired place cell architecture that develops cognitive maps during exploration of novel environments. Our model implements a simulated agent for reward-driven navigation that forms spatial representations online. The architecture incorporates behaviorally relevant information through neuromodulatory signals that respond to environmental boundaries and reward locations. Learning combines rapid Hebbian plasticity, lateral competition, and targeted modulation of place cells. Analysis of the capabilities of the model on a variety of environments demonstrates our approachs efficiency, achieving in one shot what traditional RL models require thousands of epochs to learn. The simulation results show that the agent successfully explores and navigates to the target locations in various environments, showing adaptability when the reward positions change. Analysis of neuromodulated place cells reveals dynamic changes in neuronal density and tuning field size after behaviorally significant events. These findings align with experimental observations of reward effects on hippocampal spatial cells while providing computational support for the efficacy of biologically inspired approaches to cognitive mapping.

Grid cells encode reward distance during path integration in cue-rich environments
The medial entorhinal cortex supports both path integration and landmark anchoring, but how these computations interact during goal-directed navigation is unclear. We show that grid cells dissociate from landmarks and instead encode reward distance when mice perform a path integration task on a cue-rich treadmill. Grid cell population activity reset at rewards and shifted coherently across trials, consistent with continuous attractor dynamics realigned by rewards. Furthermore, grid cells exhibited reduced spatial scales, broadened theta frequency distributions, and altered temporal coordination. These phenomena were captured by a theta interference model incorporating cell competition and two sets of theta oscillating inputs whose frequencies shifted apart. Switching to cue-based navigation stabilized the firing fields and partially restored grid scale, theta frequencies and temporal structure. These results demonstrate that MEC circuits flexibly reset to encode goal-directed trajectories, and suggest that continuous attractor and interference mechanisms normally cooperate but can decouple under path integration demands.

A developmental transition from early neural hyperexcitability in zebrafish
Early brain development is characterized by plastic phases that play a crucial role in the establishment of functional connectivity. However, this plasticity also renders the developing brain highly sensitive to perturbations. The dynamics of the transition of the immature brain from a hyperexcitable state to a relatively stable network remains poorly understood. Using a combination of brain-wide activity mapping, behavioural and molecular analysis in response to pentylenetetrazole (PTZ), a GABAA receptor antagonist, we assessed neural hyperexcitability at different developmental stages in larval zebrafish and observed a distinct age-dependent trajectory. Specifically, we observed that early stages (4-6 dpf) exhibited high susceptibility to external perturbations, whereas by 9 dpf, both neural activity and behavioral responses were markedly attenuated. Further, upon chronic insult, rather than a cumulative increase in excitability, early life hyperexcitation events rendered the larvae resistant to subsequent induction. Put together, our work maps the temporal transition of the vertebrate brain from a state of hyperexcitability to a phase of relative resistance and has potential implications on understanding the progression of neurological conditions in the pediatric population.

Time-adaptive modulation of evidence evaluation in rat posterior parietal cortex
A crucial component of successful decision making is determining the optimal timescale over which to evaluate evidence. For example, when detecting transient changes in the environment, it is best to focus evaluation on the current evidence as opposed to older evidence. However, it is unclear how this adjustment in timescale is achieved in the brain in terms of how the neurons that process evidence adjust their dynamics. To address this question, we used Neuropixel probes to record spiking activity from neurons in the posterior parietal cortex (PPC) of rats performing a free-response auditory change detection task in which subjects evaluate sensory evidence over short timescales to determine when a change occurs in a noisy sensory stream. Consistent with longer timescale temporal integration tasks, we found that PPC neurons modulated their activity by the strength of evidence leading to decisions, were selective for the rats choices, and had opposing populations of neurons that were positively versus negatively modulated by evidence. However, in contrast to temporal integration tasks, responses of neurons to individual pulses of evidence were transient, such that the effect of the evidence on activity tapered off over a timescale corresponding to the subjects behavioral timescale of evidence evaluation. Intriguingly, PPC also exhibited "gain changes" in the influence of evidence as a function of decision time that were consistent with changes in behavioral urgency. In addition, reversible inactivation revealed an important role for PPC in this auditory change detection, such that PPC inactivation altered choice behavior and the timescale over which rats evaluated evidence. Together, our results suggest important contributions of PPC to free-response decisions that involve adjusting timescales of evidence evaluation.

Epigenomic regulation of human oligodendrocyte myelination properties, relation to age and lineage
Multiple sclerosis (MS) is characterized by immune-mediated injury to myelin and oligodendrocytes (OLs). Repair depends on the ability of OL lineage cells to form new myelin and ensheathe axons. We previously showed that late progenitors (O4+A2B5+ cells) and mature human OLs exhibit age-related differences in ensheathment capacity and vulnerability to injury. Here, we test the hypothesis that differences in chromatin accessibility and specific histone marks may underlie transcriptional differences linked to these functional responses. Confocal imaging of cultured cells revealed higher levels of the transcriptionally permissive histone marks H3K27ac and H4K8ac in pediatric than adult derived late progenitors and mature OLs. Levels were higher in adult-derived progenitors versus mature cells from the same individuals. The majority of pathways and genes related to myelination and immune interactions were downregulated in adult cell samples when compared to pediatric samples. Analysis of publicly available datasets indicated that the chromatin accessibility for genes within these categories was more restricted in adult than pediatric OLs. There was less chromatin accessibility and lower H3K27ac chromatin occupancy also in more differentiated OL compared to progenitors. The levels of the transcriptionally repressive H3K27me3 histone mark in mature OLs were enriched in genomic regions encoding for transcriptional inhibitors of myelination and related signaling pathways, as compared to early progenitors. Restrictions in chromatin accessibility were more pronounced in human cells than in mouse cells. These results link the myelination capacity and immune-mediated injury susceptibility of human OLs to their epigenomic state, raising the issue of how epigenetic modulation could influence disease progression.

Hypothalamic recurrent inhibition regulates functional states of stress effector neurons.
Stress triggers rapid and reversible shifts in vital physiological functions from homeostatic operation to emergency response. However, the neural mechanisms regulating such functional stress states remain poorly understood. Here we identify a novel recurrent inhibitory circuit governing functional states of key stress regulatory neurons: corticotropin-releasing hormone (CRH) neurons in the hypothalamic paraventricular nucleus (PVN). Microendoscopic calcium imaging in freely behaving mice revealed synchronized low-activity state at baselines and a reversible high-activity state during mild stress. Ensemble analysis indicated increased dimensionality of network dynamics during high-activity state. Computational modeling of calcium ensemble data, together with independent modeling of single-unit CRHPVN neurons spiking dynamics, converged to show that recurrent inhibition is a key circuit motif for stress-induced functional state transitions. Guided by model predictions, chemogenetic manipulations of PVN-projecting GABAergic neurons revealed their roles in constraining CRHPVN neurons to low-activity state at baselines via a prolonged feedback inhibition. Unexpectedly, slow CRHergic excitation was dispensable for driving this prolonged feedback, whereas glutamatergic transmission predominated at CRHPVN[->]PVN[->]GABA excitatory synapses. Incorporating these findings, we refined our computational model to include fast excitation and slow inhibition, yielding new predictions for circuit operation. Together, our results establish recurrent inhibition as a fundamental circuit motif controlling CRHPVN neurons functional states and highlight the value of iterative experiment-model integration in advancing understanding of neural circuits functions.

Age-Related Changes in Curiosity: The Influence of Locus Coeruleus on Information-Seeking Behavior
Curiosity enhances learning and memory and has been linked to the locus coeruleus (LC), which undergoes age-related decline. To examine how aging affects curiosity and information-seeking, we developed the Photographic Art Storytelling Task. Participants (sixty-eight young and sixty-five older adults) viewed photographs, rated their curiosity, and later read stories associated with selected images. The stories were deliberately constructed to be either interesting or boring, functioning as rewards that elicited prediction errors. Participants reappraised their curiosity, allowing us to separate intrinsic motivation from story-driven reward influences. Associations between performance and both pupil diameter and MRI measures of LC integrity supported a role of LC in curiosity regulation. Aging was associated with greater reliance on novelty-driven (initial) curiosity and a shift away from prediction error-related information-seeking. Both age groups showed curiosity-driven memory enhancement, with older adults exhibiting youth-like effects, suggesting a role for intrinsic motivation in preserving cognitive function in aging.

Claustrum-cortical reciprocal connections orchestrate allostatic responses following stress
Anxiety, whilst often viewed as a disorder, is an evolutionarily conserved mechanism that facilitates threat detection and survival. However, when stress regulation becomes maladaptive, this adaptive response can shift into pathology. Here, we identify the claustrum (CLA) as a key hub for allostasis following stress, integrating Gad2 (GABAergic)-vGluT1 (glutamatergic) microcircuits. We report that acute social defeat stress activated the CLA and induced hypervigilance and anxiety-like behaviors. Multimodal analyses revealed transcriptional plasticity in CLA neurons, and fiber photometry revealed anticipatory activation of Gad2 neurons and reactive activation of vGluT1+ neurons. We further delineated a reciprocal GABAergic-glutamatergic circuit between the CLA and the anterior cingulate cortex (ACC) that orchestrates allostasis following stress via opposing mechanisms: (1) glutamatergic CLA-ACC projections that amplify threat responses, and (2) two distinct GABAergic inhibitory pathways- intrinsic CLA Gad2+ activity and top-down ACC-CLA Gad2+ modulation. Chronic stress drives persistent hyperactivation of CLA Gad2+ neurons, suppressing CLA glutamatergic activity and leading to depression-like behaviors. Our results identify a dynamic CLA circuit that gates stress responses via CLA Gad2+ neurons acting as a brake under acute stress. Chronic stress amplifies this inhibition, thereby disturbing circuit balance and driving behavioral despair affective pathology.

Axon guidance deficits in a human sensory neuron model of Fabry disease
Fabry disease (FD) is a rare genetic galactosidase alpha (GLA) gene associated lysosomal disorder caused by alpha-galactosidase A (AGAL) deficiency, leading to sphingolipid (globotriaosylceramide, Gb3) accumulation in multiple tissues. Burning pain due to small fiber neuropathy is an early symptom with great impact on health-related quality of life. The pathophysiological role of Gb3 accumulations in sensory neurons of the dorsal root ganglia is incompletely understood. We have differentiated induced pluripotent stem cells of an isogenic GLA knockout line (p.S364del, hemizygous) and its healthy control into sensory neurons to model FD in vitro. We have compared both lines on transcriptional and proteomic level and investigated the effects of AGAL enzyme supplementation. FD sensory neurons showed dysregulation of disease-related pathways, including axon guidance at both RNA and protein level and microfluidic assays revealed shorter neurite length. While AGAL did not restore the transcriptomic state, it reduced Gb3 accumulation and lowered protein ephrin 5A and glycoprotein M6A level. These findings highlight axon guidance alterations in an isogenic human FD sensory model, with potential implications for early central and peripheral innervation in small fiber neuropathy.

Representational magnitude as a geometric signature ofimage and word memorability
What makes some stimuli more memorable than others? While memory varies across individuals, research shows that some items are intrinsically more memorable, a property quantifiable as memorability. Recently, a new representational signature for image memorability was identified: the magnitude of the population response in convolutional neural networks (CNNs) correlated with image memorability. However, it is unclear if this geometric principle was confined to the visual domain or whether it represents a broader computational phenomenon observable in other stimuli domains as well. Here we show that this representational magnitude effect not only replicates for images in an independent dataset but also generalizes to an entirely different cognitive domain and neural network architecture: lexical memorability and word embeddings. Across three large-scale lexical datasets, we found that the L2 norm (vector magnitude) of word embeddings reliably predicted recognition memorability, independent of word frequency, valence, or word length. This consistency suggests the effect reflects a general property of distributed representations, where representational magnitude may capture how strongly a stimulus projects onto dominant, conceptual/semantically meaningful features in a networks embedding space. At the same time, this effect does not appear to generalize to all domains, as our analysis of a novel voice memorability dataset showed no such relationship between representational magnitude and memorability. Together, these findings indicate that the geometry of distributed representations offers a useful lens for understanding memorability, suggesting an items representational magnitude reflects its projection onto dominant dimensions of representation, in both artificial and biological systems.

Intrathecal administration of palmitoyl-lysophosphatidylethanolamine reduces secondary injury and improves locomotor recovery following spinal cord injury
Spinal cord injury (SCI) triggers secondary pathophysiological cascades, including glutamate excitotoxicity, that result in neuronal loss and impair functional recovery. We have previously shown that lysophosphatidylethanolamine (LPE), a lysophospholipid, promotes neurite outgrowth and protects against glutamate excitotoxicity in cultured cortical neurons. However, whether these effects extend to spinal cord neurons and occur in vivo has remained unclear. In this study, we compared the effects of different LPE species: myristoyl-LPE (14:0 LPE), palmitoyl-LPE (16:0 LPE), stearoyl-LPE (18:0 LPE), and oleoyl-LPE (18:1 LPE) in cultured spinal commissural neurons, and evaluated their effects in vivo using a mouse model of SCI. In cultured neurons, all LPE species promoted neurite outgrowth. Although several species demonstrated a tendency toward neuroprotection, only 16:0 LPE exhibited a statistically significant protective effect against glutamate-induced excitotoxic cell death. Intrathecal administration of 16:0 LPE after SCI reduced TUNEL-positive cells in the acute phase and attenuated lesion expansion at 8 weeks post-injury. Moreover, 5-HT fluorescence intensity was increased in 16:0 LPE-treated mice, suggesting enhanced serotonergic innervation. Furthermore, administration of 16:0 LPE after SCI significantly improved hind-limb motor performance compared with vehicle controls, as assessed by the Basso Mouse Scale. Collectively, these findings suggest that intrathecal administration of 16:0 LPE reduces secondary injury and promotes functional recovery following SCI. Our findings highlight its potential as a therapeutic candidate for SCI.

A neural network with episodic memory learns causal relationships between narrative events
Humans reflect on past memories to make sense of an ongoing event. Past work has shown that people retrieve causally related past events during comprehension, but the exact process by which this causal inference occurs remains elusive. Here, we employed a recurrent neural network augmented with an episodic memory buffer to examine how memories are retrieved and integrated based on causal relationships between events. The model was trained to predict upcoming scenes as it watched a television episode. At every time step, the model transformed the current scene into two distinct representations--"value" representing memory content and "key" representing memory address--both of which were stored as episodic memory. The model learned to retrieve selective past values by applying self-attention over stored keys, and it integrated these retrieved values with the current scene representation to predict an upcoming scene. By separating representations used for encoding from those used for retrieval, the model learned to retrieve memories in ways that go beyond simple pattern similarity. In turn, the model represented causally related events with similar patterns beyond perceptual or semantic similarities, suggesting that it organized event representations based on latent causal structure. Memories retrieved by the model were similar to those retrieved by human participants who watched the same television episode. The model also exhibited hippocampus-like pattern separation and pattern completion, and its representational structure aligned more closely with human fMRI data than a comparison model without an episodic memory buffer. These findings suggest that the model captures the way humans represent events and retrieve memories based on causal relationships. Together, this work proposes a key-value episodic memory system as a candidate computational mechanism for how humans retrieve causally related memories to comprehend naturalistic events.

Automated Quantification of Stereotypical Motor Movements in Autism Using Persistent Homology
Stereotypical motor movements (SMM) are a core diagnostic feature of autism that remain difficult to quantify efficiently and validly across individuals and developmental stages. The current paper presents a novel pipeline that leverages Topological Data Analysis to quantify and characterize recurrent movement patterns. Specifically, we use persistent homology to construct low-dimensional, interpretable feature vectors that capture geometric properties associated with autistic SMM by extracting periodic structure from time series derived from pose estimation landmarks in video data and accelerometer signals from wearable sensors. We demonstrate that these features, combined with simple classifiers, enable accurate automated quantification of autistic SMM. Visualization of the learned feature space reveals that extracted features generalize across individuals and are not dominated by person-specific SMM. Our results highlight the potential of using mathematically principled features to support more scalable, interpretable, and person-agnostic characterization of autistic SMM in naturalistic settings.

Multi-focal ultrasound neuromodulation to the dorsal anterior cingulate cortex disrupts behavioural and neural pain processing
Transcranial ultrasound stimulation (TUS) is a promising non-invasive technique for modulating deep brain regions involved in pain. TUS applied to the dorsal anterior cingulate cortex (dACC), a region implicated in chronic pain and established target for deep brain stimulation, has shown potential for reducing pain. This study aimed to investigate the neural mechanisms underlying TUS effects on pain in healthy participants using neuroimaging. Thirty-two participants underwent two double-blind, randomised TUS-fMRI sessions (active or sham). A tonic cold stimulus was applied during multifocal dACC-TUS and during fMRI and magnetic resonance spectroscopy (MRS) blocks. While no significant main effect of TUS on pain intensity was observed, active TUS showed a significantly greater reduction in pain ratings between 28- and 55-minutes post-stimulation, suggesting a delayed analgesic effect. Active TUS also disrupted the typical relationship between stimulus temperature and reported pain intensity, indicating altered sensory encoding. There was increased functional connectivity between the dACC and the supplementary motor area, pre-motor cortex, mid-ACC and the supramarginal gyrus, along with decreased coupling with the periaqueductal grey (PAG), and altered salience network connectivity. Overall, these findings suggest TUS to the dACC has multidimensional effects across behavioural and neural aspects of pain processing, supporting its potential therapeutic value.

CARACAS, a novel automated tool for Cardiac Artifact Removal in Absence of CArdiac Signal
EEG recordings can contain cardiac related artifacts. Independent Component Analysis (ICA) followed by removal of cardiac Independent Components (ICs) is a powerful and widely used strategy for artifact correction. Most existing methods for automatic labeling of cardiac ICs require simultaneously recorded ECG (e.g., to compute correlation with IC time course, such as CORR algorithm). However, ECG is not always available. To address this limitation, we developed CARACAS (Cardiac Artifact Removal in Absence of CArdiac Signal), a novel tool that identifies cardiac ICs using only the IC time courses. Because cardiac ICs exhibit temporal profiles highly similar to ECG signals, we applied one of our existing function designed to detect cardiac events (PQRST) in ECG signals to each IC time course. Analysis of the detected events enabled the differentiation of cardiac ICs from non-cardiac ICs, where unrelated signal variations are incorrectly identified as cardiac events. Using the 375 EEG-ECG recordings of the open-source dataset OpenNeuro ds003690, we compared the performances of three algorithms: CARACAS, IClabel (a generic IC classifier which does not require ECG), and CORR algorithm. A total of 21,375 ICs were manually and automatically classified. CARACAS achieved high performance (sensitivity = 0.923, specificity = 0.991), substantially outperforming ICLabel (sensitivity = 0.135, specificity = 0.999) and approaching the performance of CORR (sensitivity = 0.970, specificity = 0.998). We present a reliable ECG-free algorithm for cardiac IC detection in EEG. CARACAS provides a practical solution when ECG is unavailable, and is implemented in SASICA toolbox (MATLAB).

Spontaneous oscillatory activity in episodic timing: an EEG replication study and its limitations
Episodic timing refers to the one-shot, automatic encoding of temporal information in the brain, in the absence of attention to time. A previous magnetoencephalography (MEG) study showed that the relative burst time of spontaneous alpha oscillations () during quiet wakefulness was a selective predictor of retrospective duration estimation. This observation was interpreted as embodying the "ticks" of an internal contextual clock. Herein, we replicate and extend these findings using electroencephalography (EEG), assess robustness to time-on-task effects, and test the generalizability in virtual reality (VR) environments. In three EEG experiments, 147 participants underwent 4-minute eyes-open resting-state recordings followed by an unexpected retrospective duration estimation task. Experiment 1 tested participants before any tasks, Experiment 2 after 90 minutes of timing tasks, and Experiment 3 in VR environments of different sizes. We successfully replicated the original MEG findings in Experiment 1 but did not in Experiment 2. We explain the lack of replication through time-on-task effects (changes in power and topography) and contextual changes yielding a cognitive strategy based on temporal expectation (supported by a fast passage-of-time). In Experiment 3, we did not find the expected duration underestimation in VR, and did not replicate the correlation between bursts and retrospective time estimates. Overall, EEG captures the burst marker of episodic timing, its reliability depends critically on experimental context. Our findings highlight the importance of controlling experimental context when using bursts as a neural marker of episodic timing.

Memorability of Images Positively Influences Associative Memory Retention
Enhancing memory function is essential for daily living and cognitive health, particularly amid population aging and cognitive decline. However, current methods for memory enhancement often require specialized interventions or effortful practice. Here, we present evidence that the intrinsic memorability of images can serve as a simple, scalable, and involuntary strategy for improving associative memory. We studied its effects on associative memory in adults of distinct ages, memory retention in cognitively impaired older adults, and vocabulary learning in foreign language learners. Our results show that highly memorable images significantly enhance the recall of associated words, especially when the cue image and target word are semantically unrelated. Notably, these effects persist for at least one week and are robust across age groups and cognitive status. In a foreign language vocabulary task, pairing words with memorable images led to either improved recall accuracy or reduced learning time. These findings highlight that, in addition to being an intrinsic memory-enhancing property, image memorability also serves as a general facilitator of associative memory, via a method that is easy to adopt, requires minimal conscious effort, and can benefit the general public, particularly learners and cognitive impairment patients in educational and clinical settings.

Canonical Representational Mapping for Cognitive Neuroscience
Understanding neural representations is central to cognitive neuroscience, yet isolating meaningful patterns from noisy or correlated data remains challenging. Canonical Representational Mapping (CRM) is a novel multivariate analysis method to identify neural patterns aligned with specific cognitive hypotheses. CRM maximizes correlations between multivariate datasets -- similar to Canonical Correlation Analysis -- while controlling for confounding sources of variance, such as shared noise or irrelevant task conditions. We validate CRM with simulations and apply it across diverse neurophysiological datasets: In one application, we use CRM to map large language model activations onto intracranial electroencephalography recordings during story listening, controlling for contextual autocorrelations. In another example, we factorize overlapping representations in functional Magnetic Resonance Imaging into distinct context and episode components. Finally, we uncover frequency-coupled representations shared between hippocampus and medial prefrontal cortex in navigating rodents. Our results introduce CRM as a powerful tool to isolate representations across modalities and species.

Modeling Human Visuomotor Adaptation with a DisturbanceObserver Framework
Visuomotor adaptation studies have revealed a range of behaviors, including nonlinear saturation effects, instruction-dependent differences in performance, and distinct explicit and implicit components of adaptation. However, existing computational models have struggled to account for the full breadth of these findings. In this paper, we aim to address this gap by (1) designing a set of experiments to investigate the role of nonlinear saturation under different instructional contexts, and (2) introducing an abstract discrete-time model that captures plausible neural computations underlying the observed behaviors. The model draws on recent advances in control theory and is informed by prior work on oculomotor adaptation and cerebellar function, particularly within the floccular complex.

In Humans, fMRI Reveals That Striosome-like and Matrix-like Striatal Voxels are Engaged in Different Phases of Movement
Introduction: The striatum is organized into two neurochemically and anatomically distinct compartments, the striosome and matrix, that play specialized roles in motor and cognitive functions. While extensive animal research has elucidated compartment-specific contributions to reward, learning and motor control, direct evidence for compartment specialization in humans is lacking. Methods: We defined human striatal voxels as striosome-like or matrix-like based on biases in structural (diffusion) connectivity. Then we investigated functional activation patterns in those compartment-like voxels using task-based functional MRI (tfMRI) during pre-movement cue and five motor conditions (left/right hand, left/right foot, and tongue movements). Results: Functional activation was strikingly segregated: striosome-like voxels were preferentially engaged during the cue phase, while matrix-like voxels dominated activation during motor execution, especially for tongue and foot movement. Motor tasks elicited robust bilateral activation, with contralateral activation dominating during limb movements. Activation was more lateralized in matrix-like than in striosome-like voxels. Both striosome-like and matrix-like voxels exhibited strong activation at the onset of task execution (e.g., within the first few seconds post-cue). However, activation in matrix-like voxels declined modestly over the course of the movement phase, while striosomal activation dropped sharply at task termination, suggesting a role in behavioral transitions. These findings are consistent with the role of the striosome in anticipatory evaluation and dopaminergic modulation, and matrix specialization for executing automatized routines. Conclusions: This study provides the first task-based fMRI evidence of temporally and functionally distinct striatal compartment dynamics in humans, offering novel insights into striatal microcircuitry in motivated behavior and the planning and execution of movements.

Early postpartum development of pup urine preference in mothers
The transition to motherhood involves profound physiological and neural changes, including adaptations in the sensory systems that support infant care. While the olfactory system plays a critical role in guiding maternal behaviors such as pup retrieval and nesting, how olfactory processing itself is reshaped during motherhood remains poorly understood. Here, we show that firsttime mothers develop a selective preference for pup urine following parturition and early postpartum care, a preference not observed for other social or neutral odors. Using odor preference assays combined with liquid and gas chromatography mass spectrometry, we identify specific volatile compounds in pup urine that may contribute to this maternal attraction. Disruption of olfactory input or restriction of contact chemosensation abolished the preference, indicating that both volatile and nonvolatile sensory modalities contribute, likely through combined input from the main olfactory epithelium (MOE) and vomeronasal organ (VNO). Notably, this preference is absent in late-pregnant females, in mothers separated from pups at birth, and in virgins co-housed with pups or exposed to pup urine- highlighting that pup urine preference depends on the convergence of internal hormonal signals and external chemosensory cues. These findings reveal a previously unrecognized specificity in maternal olfactory behavior and provide insight into how motherhood modulates the sense of smell to support offspring recognition and care.

Next generation neural mass model with dopamine modulation mediated by D1-type receptors
Neuromodulation is a complex process in which chemical substances modulate brain activity, allowing its rich repertoire of behaviors. Among these substances, dopamine has a preponderant role, being involved in several mechanisms. Moreover, dysfunctions in the dopamine connections has been observed in pathology, such as Parkinson's disease and schizophrenia. To investigate the mechanism of neuromodulation, we expand a previously proposed mean-field formalism, that describes the average activity of a neural population, by adding the effect of dopamine modulation. This mean-field reduction allows for a direct comparison with the underlying neural network to test its ability to qualitatively reproduce population behavior. The resulting mathematical framework is able to capture network activity in distinct dynamical regimes and transitions between them. Thus, this approach provides a reliable foundation for the development of personalized medicine tools to study how the effect of dopamine modulation on single brain region affects whole brain behavior.