bioRxiv Subject Collection: Neuroscience's Journal

Structural Brain Correlates of Speech Disfluency in Early Childhood: A Dimensional Analysis in a Non-Clinical Cohort
Most neuroimaging studies of speech disfluency have compared individuals who stutter with fluent controls. However, treating speech disfluency as a continuous, dimensional trait offers new insights into the neural basis of fluency during early childhood. This study aimed to investigate whether naturally occurring variation in speech disfluency is associated with grey matter structure in a non-clinical, population-based sample of 5-year-old children. The study included 120 participants (65M, 55F) from the FinnBrain Birth Cohort study. Speech disfluency was evaluated as a continuous measure from audiovisual speech samples, with transcription and analysis conducted using the SALT software. Ambrose & Yairi (1999) classification system was used to categorize speech disfluencies into stuttering-like (SLD) and other disfluency types. T1-weighted images obtained through magnetic resonance imaging were analyzed using voxel-based morphometry (VBM) with the CAT12 toolbox and complemented by surface-based morphometry with FreeSurfer. Whole-brain statistical analysis was employed to examine the association between grey matter metrics and speech disfluency. We found that VBM-derived proportional grey matter volume in the left middle frontal gyrus, left posterior cerebellum, and right superior frontal gyrus was positively associated with speech disfluency, specifically SLD, in children (p <.001; p=.002; p <.001, FDR corrected). No significant associations were found for cortical thickness or surface area. Additionally, no notable sex differences were observed. Our findings suggest that speech disfluency in early childhood is linked to localized structural differences in regions supporting motor planning and cognitive control, without broader changes in cortical thickness or surface area. Importantly, similar brain regions have been implicated in studies comparing children who stutter to those who do not, suggesting that normal variation in disfluency captures meaningful neurobiological differences even in non-clinical populations. This supports the value of treating speech disfluency as a spectrum and underscores the importance of longitudinal, multimodal research to clarify how these structural features evolve and influence later fluency outcomes.

Temperature-Induced Bifurcations and Timescale Warping in the Hodgkin-Huxley Model: Dynamical Systems Insights into Temperature Modulation of Neuronal Excitability and Action Potential Morphology
Temperature fluctuations can have detrimental effects on the firing pattern and electrical activity of biological neurons, eliciting diverse responses depending on the neuronal cell types and the underlying ion channels exhibited. Using the classical Hodgkin-Huxley (HH) model, we performed a comprehensive dynamical systems analysis to determine how temperature fluctuations alter neuronal excitability, spike morphology, and bifurcation structure. We first relied on experimentally-derived temperature coefficients, or Q10 values, associated with gating kinetics and conductances, and examined codimension-1 and codimension-2 bifurcations across a range of temperatures and standard HH parameters governing the intrinsic properties (firing frequency, spike amplitude, spike width, afterhyperpolarization (AHP), time-to-peak AHP, etc...) of the model HH neuron. Our analysis revealed that increasing temperature accelerates gating dynamics, leading to narrower and higher-frequency spikes but reduced amplitudes, and ultimately to a loss of sustained firing via temperature-induced depolarization block. We identified generalized Hopf (Bautin) bifurcations as critical boundaries beyond which the system becomes strictly monostable. Extending the model to independently scale sodium activation, sodium inactivation, and potassium activation kinetics showed that excitability is particularly sensitive to potassium gating dynamics. Our findings provide a quantitative framework for understanding temperature modulations of neuronal activity, highlighting how temperature reshapes the excitability landscape, unveiling the intricate interplays between the activation/inactivation kinetics of ion channels, and identifying key parameters governing temperature robustness in neuronal models.

Global search metaheuristics for neural mass model calibration
Neural mass models (NMMs) are often used to help understand the circuitry that underpins observed brain dynamics in basic and clinical research. A key step is to fuse models with data so that model parameter values can be inferred for a given data set, a process called model fitting or model calibration. This can shed light on putative physiological mechanisms underlying the observed signals. Calibration is notoriously challenging in biology since models are often non-identifiable, high-dimensional, and nonlinear. Established methods such as dynamic causal modelling (DCM) circumvent some of these issues, for example, by incorporating prior information and employing fast local search methods in the space of feasible parameter values ("parameter space"). However, it is pertinent to better understand the potential limitations of these methods so that we can increase our confidence in the use of models to interpret brain activity, and to develop new approaches as required. Here we use tools from dynamical systems theory to illustrate some of the complexities of model calibration in an archetypal NMM. We use this information to motivate the use of calibration methods that work across large regions of parameter space, rather than focusing on informative priors or localised search methods. We subsequently evaluate the performance of approximate Bayesian computation (ABC) and evolutionary search metaheuristics (ESMs) for mapping feasible sets of parameters for which an NMM can recreate electroencephalographic recordings during an eyes-closed resting state. Our results demonstrate the superiority of ESMs in terms of computational efficiency and accuracy. Furthermore, we elucidate potential reasons why ESMs are able to perform better than ABC, i.e. that they are less susceptible to biases induced by the complexity of underlying cost landscapes. These results highlight the importance of incorporating ESMs in future efforts to model brain dynamics.

High-Throughput Screening and Initial SAR Studies Identify a Novel Sub-micromolar Potent Human cGAS Inhibitor
Cyclic GMP-AMP synthase (cGAS) has emerged as a promising therapeutic target of several human diseases, including Alzheimer's disease (AD) and other neurodegenerative disorders. As a cytosolic DNA sensor, cGAS generates an innate immune response to promote neuroinflammation by producing an endogenous agonist of the stimulator of interferon genes (STING), 2'3'-cyclic GMP-AMP (cGAMP), which activates the cGAS-STING pathway. We have performed a high-throughput screening of a chemical library containing over 300K small molecules at the Fisher Drug Discovery Resource Center (DDRC), Rockefeller University (RU), to identify multiple hit inhibitors of human (h)-cGAS. We used a modified Kinase Glo(R) Luminescent Kinase assay, which was earlier developed at RU and later used by multiple groups, including ours, to perform primary screening of the library using h-cGAS. The hit candidates bearing novel scaffolds are structurally diverse and exhibited in vitro activity in the low micromolar range. RU-0610270 or compound (cpd) 1, a sulfonamide derivative, is one of the most potent hits (IC<50>=1.88 M), selected for hit expansion and structure-activity relationship (SAR) analysis. We synthesized new analogs of cpd 1 and evaluated them in vitro against h-cGAS to identify cpd 6 (IC<50>=0.66 M) as the most potent hit analog. We further profiled cpd 6 and found that it modestly inhibited cGAMP levels by 29% at 30 M in THP1 cells without detectable toxicity, and by 76% at 100 M, albeit with a moderate decrease (~20%) in cell viability. These results highlight a novel chemical series with promising in vitro activity, providing a starting point for the development of selective and potent human cGAS inhibitors for clinical use.

Streamline Density Normalization: A Robust Approach to Mitigate Bundle Variability in Multi-Site Diffusion MRI
Tractometry enables quantitative analysis of tissue microstructure is sensitive to variability introduced during tractography and bundle segmentation. Differences in processing parameters and bundle geometry can lead to inconsistent streamline reconstructions and sampling, ultimately affecting the reproducibility of tractometry analysis. In this study, we introduce Streamline Density Normalization (SDNorm), a supervised two-step method designed to reduce variability in bundle reconstructions. SDNorm first computes streamline weights using linear regression to match a subject's bundle to a template streamline density map, then iteratively prunes streamlines to achieve a target density using a novel metric called effective Streamline Point Density (eSPD). We evaluate SDNorm across multiple bundles and acquisition protocols in dMRI data from a subset of subjects from Alzheimer's Disease Neuroimaging Initiative and demonstrate that it can significantly reduce variability in streamline density, improve consistency in along-tract microstructure profiles, and provide useful metrics for automated bundle quality control. These results suggest that SDNorm can help enhance the reproducibility and robustness of bundle reconstruction across heterogeneous image acquisition protocols and tractography settings, making it well-suited for large-scale and multi-site neuroimaging studies.

Context-dependent structurally informed effective connectivity under psilocybin
The extent to which anatomical connectivity constrains pharmacologically altered brain dynamics remains poorly understood. Here, we combined psilocybin administration with a structurally informed effective-connectivity model to examine how structural connectivity shapes directed inter-regional influences across experiential contexts. Using dynamic causal modeling embedded in a hierarchical empirical Bayes framework, we analyzed fMRI data acquired from a hippocampo--thalamo--cortical network during rest, guided meditation, music listening and movie viewing. Across contexts, psilocybin reorganized directed interactions while preserving structure-based scaling. Effects converged on efferents (outgoing influences) from the left hippocampus--a hub interfacing mnemonic and associative systems with the default-mode network and thalamus. Notably, the left-hippocampus-to-thalamus pathway showed a sign-reversed association with mystical-experience scores (downregulation during guided meditation and upregulation during music listening). In model-based leave-one-out cross-validation, left-hippocampal efferents predicted individual differences in mystical-experience intensity. A minimal model-free benchmark (hippocampal signal variability) also showed modest associations with mystical experience. Together, these findings link context-specific, structurally informed effective connectivity to individual differences in the acute psychedelic experience, providing a mechanistic bridge between anatomy, neurodynamics, and phenomenology.

The claudin-like molecule CLC-3 regulates neuromuscular function in Caenorhabditis elegans by modulating cholinergic signalling
Cell adhesion molecules (CAMs play important roles in neurons, contributing to nervous system development, synapse formation, and activity-dependent plasticity. Claudins, the cell-adhesion molecules known for their roles at tight junctions in epithelial and endothelial cells, remain underexplored in neurons, particularly in vertebrates. In contrast, emerging studies in Caenorhabditis elegans have begun to reveal neuronal functions of claudin-like proteins. However, a systematic analysis of their neuronal expression has not been performed. We conducted a transcriptional reporter screen of all claudin-like genes in C. elegans and identified several candidates with previously unreported neuronal expression, highlighting a broader role of this family in the nervous system. One candidate, clc-3, showed robust expression in head, tail, and ventral cord neurons, with no detectable expression in non-neuronal tissues. Functional analyses of clc-3 mutants revealed increased body-bend amplitude and elevated evoked postsynaptic currents at the cholinergic neuromuscular synapses. Imaging and molecular interaction studies demonstrated that CLC-3 localises to the presynaptic membranes in cholinergic neurons, where it interacts with the actin-binding protein NAB-1 and regulates cholinergic signalling. This presynaptic role of CLC-3 likely contributes to the regulation of sinusoidal movement in C. elegans. Our findings identify CLC-3 as a neuronally expressed claudin that regulates motor system output by influencing synaptic vesicle organization and illustrate how changes in synaptic organization are coupled to whole-animal behaviour.

Artificial Intelligence-driven Whole-brain Cell Mapping with Highly Multiplexed In Situ Hybridization
Recent advances in three-dimensional single-cell-resolution imaging have begun to link organ-wide and cellular level research in development and disease research. Harnessing the power of whole-mount cell staining and tissue-clearing, it became possible to quantify the cell populations throughout an intact organ. While powerful, whole-organ imaging remains limited by the inability to stain a broad range of molecular markers simultaneously and by the lack of an analytical scheme to precisely quantify the cell population. Here, we present a highly multiplexed whole-mount staining technique, utilizing the repeated application of fluorescent in situ hybridization. This technique, termed mFISH3D, was designed by extensively dissecting the chemical basis of hybridization reactions in fixed tissue. mFISH3D enabled the visualization of 10 types of mRNAs in an intact mouse brain and has been demonstrated in various biological specimens including the human brain. To achieve unprecedented levels of accuracy in spatial cell mapping, we developed artificial intelligence (AI)-driven workflow using self-supervised learning, significantly reducing the need for extensive manual annotations. The integration of mFISH3D with our AI solution sets a standard for high-dimensional tissue analysis, provides a new systematic framework for analyzing complex cellular ecosystems and enables comprehensive investigation of selective cellular vulnerabilities in diseases.

A long-lived Human Neurovascular PENTA Culture Model Reveals Incomplete Vascular Repair and Glial-Mediated Signaling After Traumatic Brain Injury.
Traumatic brain injury (TBI) frequently leads to chronic neurovascular dysfunction, yet mechanistic insights into human-specific responses have been limited by the absence of long-term, multicellular in vitro models. Here, we report a five-cell-type human neurovascular culture system, comprising endothelial cells, astrocytes, pericytes, microglia, and neurons, engineered within a 3D scaffold to study injury-induced remodeling over multiple weeks. This PENTA-culture platform recapitulates hallmark features of the neurovascular unit and enables dissection of cell-specific contributions to vascular repair and degeneration. Upon mechanical trauma, cultures exhibit a biphasic response marked by acute endothelial disintegration, mitochondrial stress, and glial activation, followed by a delayed and incomplete repair. Confocal and proteomic analyses reveal persistent disruptions in tight junction organization, elevated TDP-43 and APP expression, and altered angiogenic and immunomodulatory signaling involving Tie2 and JAK/STAT pathways. Compared to simpler culture systems, the inclusion of microglia and neurons enhances post-injury cytokine resolution and junctional recovery, underscoring the importance of neuroimmune crosstalk. This system offers a mechanistically rich, human-relevant model for studying chronic neurovascular dysfunction and therapeutic revascularization.

Closed-Loop Connectivity Best Supports Angular Tuning and Sleep Dynamics in a Biophysical Thalamocortical Circuit Model
Despite recent advancements in mapping thalamic and cortical projections, the specific organization of intrathalamic and corticothalamic connectivity remains elusive. Current experimental approaches cannot definitively determine whether these connections are arranged in reciprocal (closed-) or non-reciprocal (open-loop) circuits. We developed a biophysically detailed multi-compartmental model of the mouse whisker pathway, built on anatomical and physiological data. We showed that closed-loop intrathalamic projections between the thalamocortical (TC) relay neurons in the ventral posteromedial nucleus and the inhibitory neurons in the thalamic reticular nucleus (TRN) best reproduce thalamic spiking and local field potential responses across awake and sleep states. Increasing the percentage of closed-loop projections regulates the angular tuning in the awake state, while also supporting spindle oscillations during sleep. We also showed that direct activation of closed-loop corticothalamic feedback (CT[->]TC and CT[->]TRN) simulating TC inputs sharpens the angular tuning in the thalamus. These results contribute to resolving a long-standing question regarding the organization of intrathalamic projections, offering mechanistic insights into how thalamo-cortical circuits balance precise sensory tuning with robust oscillatory rhythms across behavioral states.

Spectral patterns of MEG oscillatory coupling emerge from meta-stable dynamics with small coupling delays
Functional connectivity (FC) is a fundamental mechanism of communication in the brain, connecting distinct oscillating neuronal populations. Oscillatory networks exhibit heterogeneity across frequencies and coupling modes whose origins are not well understood, but have been suggested to involve a complex interplay of critical-like dynamics and structure-function coupling. We here utilized structural connectivity (SC) to tune a whole-brain computational model of delay-coupled damped oscillators near a Hopf bifurcation to match oscillations and FC as observed in resting-state magnetoencephalography (MEG) data. We assessed two forms of oscillation-based FC from empirical and model data, namely phase synchronization (PS) and amplitude coupling (AC). We found that both oscillations and FC best matched with empirical observations in a meta-stable regime which was characterized by small delays, realistic oscillation lifetimes, and FC with intermediate strength and high variability. How well MEG FC patterns were matched by the model varied between frequency bands and best fits were observed for high-alpha and beta band networks. These fits could partially, but not fully, be explained by correlations with SC, implicating that both structure-function coupling and critical-like metastable dynamics underlie empirical FC, and their contributions vary between different frequency bands.

Connectome-guided personalization of optimal tDCS intervention selection in Alzheimers disease
Transcranial direct current stimulation (tDCS) is being investigated as a clinical intervention in Alzheimers disease (AD) with the goal of reducing its neurophysiological effects (i.e. oscillatory slowing and loss of functional connectivity). However, progress is hampered by variable outcomes across studies, likely related to both methodological and individual differences. We recently described a virtual brain network simulation method for optimizing tDCS interventions and now propose a novel methodology for further personalizing this approach. While the general model of the AD brain we used for our previous study was based on an average human connectome and brain anatomy, we now created new personalized models for 10 biomarker-confirmed AD patients. We used individual, amplitude envelope correlation (AEC)-based connectivity matrices extracted from magnetoencephalography (MEG) scans and implemented individual structural MRI data for current flow modeling of the tDCS effects. We then assessed a set of previously established stimulation strategies based on their ability to restore relevant neurophysiological outcome parameters in each personalized model, while undergoing AD damage. Personalized tDCS strategies were able to delay neurophysiological deterioration, but in dissimilar ways compared to our previous results. While the general model favored posterior anodal stimulation targeting the precuneus region, the personalized models favored frontal anodal stimulation targeting the dorsolateral prefrontal cortex (dlPFC) region in 90% of the cases. This may be explained by higher connectivity levels of frontal regions in the personalized connectivity matrices, as anodal stimulation of highly connected regions produced more beneficial effects. In this methodological study we propose several ways to improve personalized computational tDCS stimulation prediction modeling. We conclude that connectome-guided personalization of tDCS effects lead to different strategies with potentially better intervention outcomes. For external validation of this model-guided tDCS approach, personalized and general model predictions are currently being tested and compared in a clinical tDCS-MEG trial in AD patients.

Identifiability-Guided Assessment of Digital Twins in Alzheimer's Disease Clinical Research and Care
Digital twins -- personalized, data-driven computational models -- are emerging as a powerful paradigm for representing and predicting disease trajectories at the individual level. These models have the potential to support diagnosis, monitor disease evolution, and evaluate therapeutic interventions in virtual settings in the context of clinical trials and patient care. Rigorous model assessment is thus critical for its implementation, but medical data are often sparse, noisy, and vary significantly across individuals, making it challenging to determine whether a digital twin optimized on such data is valid. In such settings, identifiability analysis becomes essential for evaluating whether model parameters can be reliably estimated and interpreted. To address this, we investigate how identifiability can support the clinical application of a computational causal digital twin model for Alzheimer's Disease (AD), where data sparsity and variability are particularly pronounced. Our results show that the magnitude and distribution of biomarker data influence the parameter practical identifiability, and that constraints on the model structure and parameters can significantly affect identifiability. We also observe differences in identifiability across diagnostic groups, with several parameters showing significantly different values between individuals with AD, mild cognitive impairment (MCI), and cognitively normal (CN) subjects. Uncertainty quantification for identifiable parameters and their corresponding model trajectories provides visual insight into variability in disease progression and reveals mild trends related to biomarker data spread. This study represents a first step toward incorporating identifiability techniques into clinical digital twin frameworks, using a data-driven, interpretable example based on a previously published AD model.

Direct Generation of Images from EEG using Schrödinger Bridge
Real-world data is often noisy, making it challenging to extract true signals. Non-invasively recorded neural activities are among the most difficult data, yet its precise signal reconstruction is highly anticipated by communities developing non-invasive brain-machine interfaces. Several noise sources contribute to this challenge, including unrelated neuronal activity, non-brain bioelectricity, attenuation by the skull and scalp, and environmental noises. Additionally, the accumulation of noise varies significantly across subjects and recording sessions, resulting in widely diverging distributions of degraded observations. In this study, we propose modeling the noise accumulation process as a Schrodinger bridge and decoding the true signal by reversing this process. Compared to conventional guided Diffusion approach, our Schrodinger bridge approach effectively models diverse noise processes within a single framework, exhibiting greater robustness to inter-subject variability. Also, our approach doesn't require pre-aligning brain and image representations, which is an additional compute cost in the conventional approach.

Frequency bands EEG Biomarkers for Dementia using Graph Neural Networks
We introduce a simple and interpretable model for classification of electroencephalography (EEG) signals. Our focus essentially is on using deep learning to study how connectivity patterns that are integrated to classify the EEG signals and highlight the important discriminative features used by the model in predictions. In this study, we utilize the connectivity features across different frequency bands in multi edge Graph Neural Networks (GNN) and showed that edge features are complimentary. We use a simple GNN model to predict Frontotemporal Dementia (FTD) in EEG. Our model is capable of achieving average accuracy of approximately 76% using Leave-One-Subject-Out-subject for FTD predictions which are better than the baselines and comparable to State of the arts models. In this article, we study the importance of the connectivity edges, nodes and frequency bands in the prediction of the model, focusing in explainable AI methods through saliency maps to interpret the model both locally and globally. The Saliency maps highlight the importance of Occipital and anterior temporal regions in the prediction of FTD. Furthermore, our results highlight the importance of Alpha and Theta bands in the prediction of FTD. Our observations align with previous research done using classical statistical methods. We argue that there are complimentary information in each each connectivity feature and frequency band brain networks. The impacts of each connectivity metrics on the prediction of the model are quantified to highlight the complimentary information in each connectivity measure.

Thinging Through Modelling. Active Inference Meets Material Engagement
In this simulation study, we adopt the comprehensive neurocomputational approach of Active Inference (AIF) to illustrate some key concepts of Material Engagement Theory (MET) [1]. MET posits that craftwork does not require, or rely on, rich internal 'pre-planning', i.e., complex and highly detailed representations that occur mainly in the maker's head. Instead, the maker engages materiality through 'thinging', where the human agent (the maker) is guided by and leverages the materiality of the artefact (such as a spinning of clay or a chunk of marble). MET assigns a crucial co-participatory role to materiality, attributing agency to it. We investigate MET's claims through the widely adopted theory of AIF [2]. Our first aim is to simulate the plausibility of the (creative) thinging, adopting a simple modelling scenario. Then, we also discuss its applicability to other, more complex cases, which seem to require greater levels of 'pre-thinking' (e.g., planning, imagining and conceptualising outcomes). We highlight how these cases, too, align with the general principle of thinging. With our AIF neurocomputational understanding, we explain that even in these situations, the predictive brains involved in the creative process attempt to minimise the complexity of their internal model. The upshot of this is that, always and everywhere, our human minds engage the materiality to make the most of the characteristic dynamics of the world surrounding us - things and processes alike.

REM Sleep Misfires: Intruding Delta Waves Forecast Tau, Amyloid, and Forgetting in Aging
Rapid Eye Movement (REM) sleep degrades with age, and more severely in Alzheimer's disease (AD). REM sleep comprises about twenty percent of adult sleep, alternates between phasic and tonic periods, and includes delta waves (1-4Hz) in two forms: fast sawtooth waves and slower, NREM-like waves, whose expression dynamically varies across REM periods. Yet, the functional relevance of these REM sleep delta waves remains unknown. Here, using two independent cohorts, we show that aging is associated with a shift from fast sawtooth to slow NREM-like delta waves, particularly during phasic REM sleep, a period typically marked by high cortical activation. Beyond chronological age, this shift is associated with amyloid-beta and tau burden, suggesting that AD pathology disrupts REM-specific oscillatory patterns. Furthermore, this shift in REM oscillations is linked to impaired overnight memory consolidation, independent of NREM sleep quality. Moreover, variation in ApoE alleles, a major genetic risk factor for AD, was independently associated with a reduction in fast sawtooth wave density, thereby linking a genetic predisposition for AD to these specific REM microstructural changes. These findings identify a novel signature of memory decline in aging and implicate REM sleep as a distinct vulnerable substrate through which AD pathology may impair brain function.

Inner speech and the neurobiology of psychosis
Aberrations of inner speech have been linked to psychotic symptoms such as thought insertion and auditory verbal hallucinations. These symptoms may reflect failures of prediction and source monitoring. Normally, efference copies of speech motor commands are sent to auditory cortices and suppressed, helping distinguish self-generated from external input. If suppression malfunctions, predicted auditory input may become perceptually salient. Further, if self-monitoring or error detection-related regions are also impaired (e.g., anterior cingulate cortex, ACC), inner speech may be misattributed as external. We tested this proposal using neuroimaging meta-analyses, examining how the brain systems in overt and inner speech production in neurotypical participants overlap with findings from psychosis-spectrum participants performing a range of tasks. They showed increased activity in motor-related regions associated with inner speech (e.g., ventral premotor cortices) and decreased grey matter in bilateral auditory cortices and ACC, in regions specific to overt speech. Coactivation-based network analyses revealed that these ventral premotor and auditory regions form distinct, inversely coupled audiomotor networks. Classification suggests the ventral premotor network supports 'higher-level' language processing, while the audiomotor network supports 'lower-level' speech and self-referential processing. Overall, results accord with the proposal that psychotic symptoms like auditory verbal hallucinations derive from phenotypic hyperactivation in inner speech-related regions that yield affectively salient efference copy signals that are insufficiently suppressed and monitored as self-produced. In line with a hierarchical predictive-processing account, disruption of a distributed recurrent system distorts self-awareness and conscious experience.

Developmental relationships between the human alpha rhythm and intrinsic neural timescales are dependent on neural hierarchy
Maturation of human brain structure has been well-studied, but developmental changes to brain physiology are not as well understood. One consistent finding is that the peak alpha rhythm frequency (PAF) increases throughout childhood. Another is that resting-state functional connectivity shifts from sensorimotor regions in children to association regions in adolescents, a reorganization along a hierarchy called the sensorimotor-to-association (S-A) axis. In mature brains, the S-A axis has been parcellated physiologically using the duration of persistent neural activity, known as the intrinsic neural timescale (INT), which increases along the hierarchy. Here we studied the development of PAF and INT in a cohort of epilepsy patients 3 to 33 years of age undergoing intracranial electrocorticographic (ECoG) monitoring. Given the well-known developmental trajectory of PAF, and the ability to delineate hierarchy using INT, we hypothesized that changes to PAF and INT would correlate across development, but that their relationship may be influenced by hierarchy. Consequently, we predicted that age-dependent PAF increases would accompany INT decreases, and we tested whether their relationship varied between sensorimotor and association regions. We found that PAF increased significantly with age in both sensorimotor and association regions, while age-dependent INT decreases were only significant in association regions. Supporting this finding, we found a strong negative relationship between PAF and INT that was specific to association regions. Together, our results suggest that developmental divisions across the S-A axis manifest in the relationships between neurophysiological measures, providing further evidence that asynchronous development along the S-A axis depends on maturation of brain function.

The rhythmic bimodal sensory stimulation in synchronous manner entrains the network oscillation in basolateral amygdala
The state of neural oscillation is important for various brain functions. In the basolateral nucleus of amygdala (BA), the oscillation frequency is accelerated in retrieval of conditioned fear memory. The amygdala receives sensory inputs from associated cortex and thalamus. Therefore, we tried to apply the bimodal sensory stimulation at slow frequency (5 Hz, functional frequency in behavioral context) for the entrainment of the BA oscillation. Young adult rats (P24-30) were stimulated by LED illuminator and acoustic speaker at 1 or 5 Hz for 1 hour. Immediately after the stimulus was finished, BA slices were prepared and whole-cell recording was applied to projection neuron. The slow (0.5-2 Hz) rhythmic IPSCs obtained from the pyramidal neuron was accelerated at ~4 Hz by synchronous opto-acoustic stimulation at 5 Hz. However, the frequency of the neuron at the later recording did not change in the same slice, suggesting that this induced entrainment is transient and reversible phenomenon. As a result, the power distribution was shifted from 0.1-2 to 2-6 Hz by synchronous bimodal 5 Hz stimulation. The regularity of the interval between IPSCs, quantified by rhythm index and the concentration of power around peak frequency in the power spectrum, was not changed by rhythmic sensory stimulation. These results suggest that synchronous bimodal sensory stimuli control the neuronal oscillation frequency by applying with rhythmicity.

Predictive encoding of auditory sequences in the human prefrontal cortex
Humans extract regularities from the environment to form expectations that guide perception and optimize behavior. Although the prefrontal cortex (PFC) is central to this process, the relative contributions of orbitofrontal (OFC) and lateral PFC (LPFC) remain unclear. Here, we show that the brain tracks sound regularities in an auditory deviance detection task to predict when a target deviant will occur. Intracranial EEG in epilepsy patients reveals prefrontal engagement, with earlier expectancy-related modulation in OFC and later modulation in LPFC. Connectivity analyses indicate bidirectional but asymmetrical expectancy-related information exchange between the two areas with a first lead by OFC, consistent with its role in initiating predictive encoding. Converging causal evidence shows that OFC lesions abolish sensitivity to expectancy, whereas LPFC lesions yield only modest effects not significantly different from controls. Together, these results provide electrophysiological and causal evidence for distinct, temporally organized contributions of prefrontal subregions to predictive processing.

Individual differences shape conceptual representation in the brain
Each person experiences the world through a unique conceptual lens, shaped by personal experiences, natural variations, or disease. These individual differences have remained largely inaccessible to cognitive neuroscience and clinical neurology, limiting the development of precision medicine approaches to cognitive disorders. To overcome this limitation, here we develop a new statistical framework to measure and interpret individual differences in functional brain representations. We apply this framework to characterize how different individuals represent the same concepts. Twenty-four participants listened to narrative stories while their brain activity was measured with functional MRI (fMRI). Encoding models were used to recover how hundreds of concepts were represented in each person's brain. Despite listening to identical stories, participants showed systematic individual differences in conceptual representations. These differences reveal person-specific biases in how concepts are represented in the brain. Individual variability was highest in regions that represent social information. Because these regions are thought to integrate sensory information with personal beliefs and experiences, the observed individual differences may reflect cognitive traits unique to each person. Our work reveals that individual differences are a systematic, measurable principle of conceptual representations in the human brain. By enabling researchers to measure and interpret differences in person-specific functional brain representations, our work establishes a new paradigm for precision neuroscience. This paradigm provides a rigorous foundation for developing fMRI applications in precision medicine to diagnose and monitor cognitive disorders.

Neuron morphological physicality and variability define the non-random structure of connectivity
Connectivity in neuronal networks is characterized by high complexity that is required for the correct function of the circuitry. Our attempts to capture it has thus far been limited to the addition of individual features of complexity to stochastic models, leading to limited insights into its origin. Based on the idea that the morphologies of neurons underlies the network structure, we developed an intuitive explanation for the mechanisms leading to structured, non-random connectivity. While a class of neurons on average innervates its entire surroundings, each individual one can only cover a small part of the space. That part is different for each neuron, but in a way that is not completely random, as it must be spatially continuous due to the physicality of neurites. We tested predictions from our hypothesis successfully in biophysically-detailed models and an electron-microscopic (EM) reconstruction of cortical connectivity. We distilled it into a simple stochastic algorithm that generates networks, which accurately match the EM reconstruction in basic network statistics as well as functionally relevant metrics of complexity. Our work may improve the understanding of the impact of neuron malformations and enable the study of the functional role of non-random network structure in simplified models.

Voltage Imaging of CA1 Pyramidal Cells and SST+ Interneurons Reveals Stability and Plasticity Mechanisms of Spatial Firing
Hippocampal place cells (PCs) are important for spatial coding and episodic memory. PCs' representations are modulated upon transitioning between environments (global remapping) but also change with repeated exposure to familiar spaces (representational drift). To gain insights into the mechanistic basis for this unique balance between circuit plasticity and stability, we used voltage imaging to longitudinally record the subthreshold and spiking activity of pyramidal neurons (PNs) and somatostatin-positive (SST) interneurons in CA1 during virtual navigation. A fraction of cells from both populations showed spatial representations, but many SSTs were speed-tuned or fired uniformly across space. Intracellular recordings revealed increased theta power and asymmetric ramp-like depolarization in PN place fields, while SSTs exhibited symmetric depolarization with no theta increase. Longitudinal recordings across weeks demonstrated representational drifts in both populations, although SSTs exhibited remarkably stable firing and subthreshold properties. Transition to a novel environment induced remapping in both populations, accompanied by increase in SST activity and reduction in PNs. These results provide new insights into how hippocampal circuits balance representational stability with experience-dependent plasticity.

Connectivity of serotonin neurons reveals a constrained inhibitory subnetwork within the olfactory system.
Inhibitory local interneurons (LNs) play an essential role in sensory processing by refining stimulus representations via a diverse collection of mechanisms. The morphological and physiological traits of individual LN types, as well as their connectivity within sensory networks, enable each LN type to support different computations such as lateral inhibition or gain control and are therefore ideal targets for modulatory neurons to have widespread impacts on network activity. In this study, we combined detailed connectivity analyses, serotonin receptor expression, neurophysiology, and computational modeling to demonstrate the functional impact of serotonin on a constrained LN network in the olfactory system of Drosophila. This subnetwork is composed of three LN types and we describe each of their distinctive morphology, connectivity, biophysical properties and odor response properties. We demonstrate that each LN type expresses different combinations of serotonin receptors and that serotonin differentially impacts the excitability of each LN type. Finally, by applying these serotonin induced changes in excitability to a computational model that simulates the impact of inhibition exerted by each LN-type, we predict a role for serotonin in adjusting the dynamic range of antennal lobe output neurons and in noise reduction in odor representations. Thus, a single modulatory system can differentially impact LN types that subserve distinct roles within the olfactory system.

Adaptive generalization and efficient learning under uncertainty
People often use recognizable features to infer the value of novel consumables. This "generalization" strategy is known to be beneficial in stable environments, such that individuals can use previously learned rules and values in efficiently exploring new situations. However, it remains unclear whether and how individuals adjust their generalization strategy in volatile environments where previously learned information becomes obsolete. We hypothesized that individuals adaptively use generalization by continuously updating their beliefs about the credibility of the feature-based reward generalization model at each state. Our data showed that participants used generalization more when the novel environment remained consistent with the previously learned monotonic association between feature and reward, suggesting efficient utilization of prior knowledge. Against other accounts, we found that individuals incorporated an arbitration mechanism between feature-based value generalization and model-based learning based on volatility tracking. Notably, our suggested model captured differential impacts of generalization dependent on the context-volatility, such that individuals who were biased the most toward generalization showed the lowest learning errors when the value of stimuli are generalized along the recognizable feature, but showed the highest errors in a volatile environment. This work provides novel insights into the adaptive usage of generalization, orchestrating two distinctive learning mechanisms through monitoring their credibility, and highlights the potential adverse effects of overgeneralization in volatile contexts.

Evolution of Cajal-Retzius Cells in Vertebrates from an Ancient Class of Tp73+ Neurons
In the developing cerebral cortex, Cajal Retzius (CR) cells are early-born neurons that orchestrate the development of mammalian-specific cortical features. However, this cell type has not been conclusively identified in non-mammalian species. Here we studied neurons expressing Tp73, a transcription factor specifically expressed in most mammalian CR cells. Comparisons of chicken, salamander, zebrafish, and little skate data indicate that Tp73-expressing neurons have conserved spatial distribution and transcriptomic signatures in vertebrates. Among the conserved Tp73-expressing cell types we find CR cells andTp73+ external tufted cells (ETCs) in the olfactory bulb of jawed vertebrates. ETCs and CR cells share the expression of most "canonical" CR cell transcription factors, such as Tp73, Lhx1, Lhx5, Ebf3, and Nr2f2, indicating that they are sister cell types. Our findings suggest that CR and ETCs evolved in stem vertebrates from cells involved in olfactory processing, with CR cells progressively acquiring new specialized roles in developmental signaling.

The Microglial Trem2 R47H Alzheimer's Disease Risk Variant Impairs Early Hippocampal Synaptic Remodeling
The rare R47H variant of the microglial TREM2 gene increases Alzheimers disease (AD) risk, but its effects on early hippocampal circuitry remain unclear. We examined basal synaptic transmission and synaptic density in 3-week-old Trem2 R47H knockin mice using in vitro whole-cell patch clamp recordings and immunohistochemistry. R47H mice displayed increased synaptic density in CA1 and CA3 regions, consistent with a trend towards increased spontaneous excitatory current frequency. Evoked synaptic currents, miniature EPSCs and spontaneous inhibitory current frequency were unaltered. Microglia shape synaptic circuits in early development by removing inactive synapses. Thus, inhibition of this role due to the Trem2 R47H mutation decreases the microglial reshaping of hippocampal connectivity during early postnatal development, without overtly altering basal synaptic activity. With ongoing synaptic plasticity, such early structural changes may predispose neural circuits to later dysfunction, particularly in the context of AD pathology, highlighting the importance of microglial TREM2 in developmental synaptic refinement.

The Role of the Basal Ganglia in the Human Cognitive Architecture: A Dynamic Causal Modeling Comparison Across Tasks and Individuals
Researchers agree input from the basal ganglia (BG) to the prefrontal cortex (PFC) plays an important role in cognition, but they disagree on its computational properties. Theoretical models characterize the majority of BG input as either direct (directly transmitting information to the PFC), or modulatory (indirectly influencing PFC activity through the gating of signals from other cortical areas). To determine the computational nature of these BG-PFC inputs in cognition, we tested three alternative connectivity configurations (Direct, Modulatory, and Mixed) within a large-scale cognitive architecture. This architecture, the Common Model of Cognition, has been independently validated using fMRI data from the Human Connectome Project (HCP). The Direct model reflected the standard CMC configuration, featuring a bidirectional, direct connection between the BG and PFC modules. The Modulatory model removed this direct link, instead incorporating a unidirectional connection from the PFC to the BG and two modulatory pathways from the BG to the PFC that passed through other cortical modules. The Mixed model included both direct and modulatory connections. Using fMRI data from 200 HCP participants performing six cognitive tasks and one resting-state session, we applied Dynamic Causal Modeling (DCM) to estimate and compare the influence of these different BG-PFC connectivity patterns. Here, we show that for each of the six cognitive tasks and resting state, the Mixed model consistently outperformed the Direct and Modulatory models. The next best model depended on the specific cognitive task, suggesting the ability for the BG to flexibly adapt to various task demands. Taken together, the current data provide evidence for a likely set of core computations that the BG uses to differentially regulate cortical activity.

A common representational code for event and object concepts in the brain
Events and objects are two fundamental ways in which humans conceptualize their experience of the world. Despite the significance of this distinction for human cognition, it remains unclear whether the neural representations of object and event concepts are categorically distinct or, instead, can be explained in terms of a shared representational code. We investigated this question by analyzing fMRI data acquired from human participants (males and females) while they rated their familiarity with the meanings of individual words (all nouns) denoting object and event concepts. Multivoxel pattern analyses indicated that both categories of lexical concepts are represented in overlapping fashion throughout the association cortex, even in the areas that showed the strongest selectivity for one or the other type in univariate contrasts. Crucially, in these areas, a feature-based model trained on neural responses to individual event concepts successfully decoded object concepts from their corresponding activation patterns (and vice versa), showing that these two categories share a common representational code. This code was effectively modeled by a set of experiential feature ratings, which also accounted for the mean activation differences between these two categories. These results indicate that neuroanatomical dissociations between events and objects emerge from quantitative differences in the cortical distribution of more fundamental features of experience. Characterizing this representational code is an important step in the development of theory-driven brain-computer interface technologies capable of decoding conceptual content directly from brain activity.

Adjudicating competing theories of semantic representation with 7T-fMRI and multivariate decoding
Theories of semantic representation in the brain are numerous and contradictory and are supported by an equally contradictory literature of multivariate neuroimaging studies. Equipped with two pioneering tools - (1) a 7T-fMRI acquisition sequence capable of capturing signal in crucial ventral temporal regions, and (2) four innovative decoding methods that are designed to probe specific hypotheses about the nature of the semantic code - we resolve this confusion. Evidence best supports the hypothesis that the ventral anterior temporal lobe represents complex, multidimensional semantic information via an interdependent, dynamic code similar to that observed in neural networks. Similar representations - highly variable across individuals - were discovered in posterior temporal and occipitotemporal cortex. In summary, this work adjudicates competing theories of semantic cognition, thereby reconciling the discrepancy in results between multivariate imaging and other sources of evidence - neuropsyschology, noninvasive brain stimulation, intracranial electrophysiology, and neural network modelling.

Age-related cerebello-thalamo-cortical white matter degradation and executive function performance across the lifespan
The cerebellum supports higher-order cognition, such as working memory and executive function (EF) both directly and through connection with prefrontal areas via cortical loops. Thus, age-related degradation to white matter connectivity comprising cerebello-thalamo-cortical (CTC) loops may underlie age-related differences in EF. In 190 healthy adults (aged 20-94 years) we collected diffusion tensor imaging scans and multiple tests of working memory and EF. Deterministic tractography was used to generate CTC tracts from which white matter metrics (mean, radial, axial diffusivities) were extracted. General linear model results indicated that reduced white matter integrity (i.e., higher diffusivity) was associated with significantly poorer EF performance in an age-dependent fashion. Higher mean, radial, and axial diffusivities in fronto-cerebellar white matter was associated with lower EF scores in older, but not younger, adults. These findings suggest CTC white matter connectivity is important for executive function performance and lend mechanistic evidence to the role of the cerebellum in age-related differences in higher-order cognitive operations.