bioRxiv Subject Collection: Neuroscience's Journal

Unveiling the role of miRNAs from MSC-EVs in neuroinflammation and behavioral impairments induced by chronic alcohol consumption
Extracellular vesicles derived from mesenchymal stromal cells (MSC-EVs) have emerged as a promising form of regenerative and immunomodulatory therapy; indeed, micro (mi)RNAs contained within MSC-EVs modulate target gene expression and impact disease-associated pathways. Chronic alcohol consumption results in neuroinflammation, brain damage, and impaired cognition; in this study, we asked whether the repeated intravenous administration of MSC-EVs could ameliorate neuroinflammation and behavioral impairment induced by chronic alcohol consumption in mice. MSC-EVs diminished the increased binding of a micro-positron emission tomography tracer (18F-FDG) when analyzing whole-brain 3D images and brain coronal sections of ethanol-treated mice. MSC-EV administration protected against ethanol-induced proinflammatory gene upregulation, cognitive dysfunction, and addictive-like behavior. miRNA sequencing data from MSC-EVs helped to reveal the elevated expression of EV-derived miR-483-5p and miR-140-3p in the brains of ethanol-treated mice following MSC-EV administration. In addition, MSC-EVs modulated the expression of pro-inflammatory-related miRNA target genes (e.g., Socs3, Tnf, Mtor, Atf6) in the brains of ethanol-treated mice. These results suggest that MSC-EVs could function as a neuroprotective therapy to ameliorate the neuroinflammation, cognitive dysfunction, and addictive-like behavior associated with chronic alcohol consumption.

Deep-learning and analytical models give distinct results for the brain structure-function relationship in health and in psychosis
Quantifying the intricate relationship between brain structure and function is of extreme importance in neuroscience. In this work, we present a comprehensive framework for mapping brain structural connectivity measured via diffusion-MRI to resting-state functional connectivity measured via magnetoencephalography, utilizing a deep-learning model based on a Graph Multi-Head Attention AutoEncoder. We compare the results to those from an analytical model that utilizes shortest-path- length and search-information communication mechanisms. The deep-learning model performed well at predicting healthy participant functional connectivity at individual-participant level, in particular in the alpha and beta frequency bands (mean correlation coefficient over 0.8), and better than the analytical model. The two models identified distinct differences in the structure-function relationship in people with psychosis compared to healthy participants, which were highly statistically significant (p < 2x10-4 for the deep-learning model, p < 3x10-3 in the delta band for the analytical model). Our results imply that human brain structural connectivity and electrophysiological functional connectivity are tightly coupled. They also show that simple analytical algorithms are very good models for communication between brain areas in people with psychosis in the delta and theta bands, while more sophisticated models are required for the alpha and beta bands.

HBN-EEG: The FAIR implementation of the Healthy Brain Network (HBN) electroencephalography dataset
The Child Mind Institute (CMI) Healthy Brain Network (HBN) project has recorded phenotypic, behavioral, and neuroimaging data from ~5,000 children and young adults between the ages of 5 and 21. Here, we present "analysis-ready" data from its high-density (128-channel) electroencephalographic (EEG) recording sessions formatted as Brain Imaging Data Structure (BIDS) datasets (HBN-EEG) with behavioral and task-condition events annotated using Hierarchical Event Descriptors (HED), making it analysis-ready for many purposes, without 'forensic' search for unreported details. We also ensured individual data files data and event integrity and marked inconsistencies. HBN-EEG sessions include six tasks, three with no participant behavioral input and three including button press responses following task instructions. Openly available participant information includes age, gender, and four psychopathology dimensions (internalizing, externalizing, attention, and p-factor) derived from bifactor model of questionnaire data. Currently, HBN-EEG data from more than 2,600 participants is freely available on NEMAR (nemar.org) and OpenNeuro in the form of nine dataset releases, with further dataset releases to follow.

Exploring the trade-off between deep-learning and explainable models for brain-machine interfaces
People with brain or spinal cord-related paralysis often need to rely on others for basic tasks, limiting their independence. A potential solution is brain-machine interfaces (BMIs), which could allow them to voluntarily control external devices (e.g., robotic arm) by decoding brain activity to movement commands. In the past decade, deep-learning decoders have achieved state-of-the-art results in most BMI applications, ranging from speech production to finger control. However, the 'black-box' nature of deep-learning decoders could lead to unexpected behaviors, resulting in major safety concerns in real-world physical control scenarios. In these applications, explainable but lower-performing decoders, such as the Kalman filter (KF), remain the norm. In this study, we designed a BMI decoder based on KalmanNet, an extension of the KF that augments its operation with recurrent neural networks to compute the Kalman gain. This results in a varying "trust" that shifts between inputs and dynamics. We used this algorithm to predict finger movements from the brain activity of two monkeys. We compared KalmanNet results offline (pre-recorded data, n=13 days) and online (real-time predictions, n=5 days) with a simple KF and two recent deep-learning algorithms: tcFNN (non-ReFIT version) and LSTM. KalmanNet achieved comparable or better results than other deep learning models in offline and online modes, relying on the dynamical model for stopping while depending more on neural inputs for initiating movements. We further validated this mechanism by implementing a heteroscedastic KF that used the same strategy, and it also approached state-of-the-art performance while remaining in the explainable domain of standard KFs. However, we also see two downsides to KalmanNet. KalmanNet shares the limited generalization ability of existing deep-learning decoders, and its usage of the KF as an inductive bias limits its performance in the presence of unseen noise distributions. Despite this trade-off, our analysis successfully integrates traditional controls and modern deep-learning approaches to motivate high-performing yet still explainable BMI designs.

A triple serine motif in the intracellular domains of sortilin-related receptors SorCS1-3 regulates neurotrophic activity
The Vps10p-domain receptors SorCS1-3 have been repeatedly associated with the development of neurological and psychiatric disorders. They have emerged as key regulators of synaptic activity and neurotrophic signaling, but the underlying molecular mechanism remains poorly understood. Here we report that the SorCS1-3 intracellular domains (ICDs) contain a conserved triple serine motif that potentially functions as a signaling switch to induce neurotrophic signaling in hippocampal neurons. We demonstrate that phosphorylation mimicking mutations of the SorCS1-3 triple serine motifs display neurotrophic activity independently of both their extracellular domains (ECDs) and BDNF, and that the substitution of serines to alanines renders neurons less responsive to BDNF. Hence, we develop triple serine motif-based cell-penetrating peptides that modulate downstream signaling kinases of the BDNF pathway, ultimately activating the transcription factor CREB. Taken together, we provide the first mechanistic insights into SorCS1-3 mediated neurotrophic signaling and use this knowledge to develop pharmacologically active modulators.

Dynamic Attentional Control Through Mixed Prefrontal Cortex Resources
Spatial attention can be either involuntarily drawn to salient, unexpected events (exogenous attention) or voluntarily directed (endogenous attention). Evidence suggests a link of both forms of attention to the (oculo)motor system. We investigated the role of the prefrontal cortex (PFC) in this relationship using a task that spatially dissociated attention from motor planning in two monkeys. PFC units flexibly encoded either the attended location or motor target, with some showing overlapping coding. Endogenous attention promotes an early engagement of these shared resources, producing two key effects: (1) it prevents interference between motor and attention systems, and (2) it regulates the extent of the exogenous shifts for attentional capture, with stronger shifts occurring when endogenous attention is weak. Overall, we show that there are separate attention and motor systems, each of which can flexibly draw on some 'overlapping' resources based on environmental and internal demands.

Out of the single-neuron straitjacket: neurons within assemblies change selectivity and their reconfiguration underlies dynamic coding
We investigated cell assemblies in the frontal cortex of macaques during two discrimination tasks. Focusing on the period of goal-action transformation, we extracted spikes fired during assembly activation from the full neural activity and showed that the contribution of a neuron to assembly coding, when it coordinates with other assembly neurons, differs from its coding in isolation. Neurons with their flexible participation to multiple assemblies contributed to the encoding of new information not encoded by the neurons alone. Even non-discriminative neurons acquired selectivity as part of the collective activity of the assemblies. Thus, neurons in their assemblies process distinct information for various purposes as a chess simul master, playing on multiple chessboards. The reconfiguration of the participation of the neurons into different assemblies in the goal-action transformation process translated into a dynamic form of coding, whereas minimal reconfiguration was associated with the static goal coding of the memory period.

Limiting hearing loss in transgenic mouse models
Transgenic mice provide unprecedented access to manipulate and visualize neural circuits, however, those on a C57BL/6 background develop progressive hearing loss, significantly confounding systems-level and behavioral analysis. While outbreeding can limit hearing loss, it introduces strain variability and complicates the generation of complex genotypes. Here, we propose an approach to preserve hearing by crossing transgenic mice with congenic B6.CAST-Cdh23Ahl+ mice, which maintain low-threshold hearing into adulthood. Widefield and two-photon imaging of the auditory cortex revealed that 2.5-month-old C57BL/6 mice exhibit elevated thresholds to high frequency tones and widespread cortical reorganization, with most neurons responding best to lower frequencies. In contrast, Ahl+ C57BL/6 mice exhibited robust neural responses across a range of frequencies and sound levels (4-64 kHz, 30-90 dB SPL) and retained low thresholds into adulthood. Our approach offers a cost-effective solution for generating complex genotypes and facilitates more interpretable systems neuroscience research by eliminating confounding effects from hearing loss.

Examining Relationships between Functional and Structural Brain Network Architecture, Age, and Attention Skills in Early Childhood
Early childhood is a critical period showing experience-dependent changes in brain structure and function. The complex link between the structural connectivity (SC) and functional connectivity (FC) of the brain is of particular interest, however, its relationship with both age and attention in early childhood is not well understood. In this study, children between the ages of 4 and 7, and at a one-year follow-up visit, underwent neuroimaging (diffusion-weighted and passive-viewing functional magnetic resonance imaging) and assessments for selective, sustained, and executive attention. We examined regional graph theory metrics and SC-FC coupling of the structural and functional networks. Partial least squares (PLS) was used to investigate longitudinal brain measure changes and cross-sectional associations with age and attention. We observed longitudinal changes in functional graph theory metrics and age-related decreases in SC modularity. Region-wise graph theory analyses revealed variable brain-behaviour relationships across the brain, highlighting regions where structural topology is linked to age and attentional performance. Furthermore, we identified SC as a dominant predictor of age when compared to FC and SC-FC coupling. The findings emphasise how early childhood is a dynamic period where cognitive functioning is intricately and predominantly linked to structural network features.

Guided Visual Search is associated with a feature-based priority map in early visual cortex
Visual search models have long emphasised that task-relevant items must be prioritised for optimal performance. While it is known that search efficiency also benefits from active distractor inhibition, the underlying neuronal mechanisms are debated. Here, we used MEG in combination with Rapid Invisible Frequency Tagging (RIFT) to probe how neuronal excitability in early visual cortex is modulated during feature-guided visual search. Participants were instructed to indicate the presence or absence of a letter "T" presented amongst 16 and 32 "L"s. In the guided search condition, participants were informed about the colour of the "T" and could infer the colour of the irrelevant distractors. In the unguided search condition, the target colour was unknown. We found that guided search was associated with enhanced RIFT responses to the target colour, and decreased responses to the distractor colour compared to unguided search. These results conceptually replicated using both a conventional coherence approach, as well as with a General Linear Model approach based on a single-trial measure of the RIFT response. The present findings expand on previous reports based on electrophysiology and fMRI in humans and non-human primates by revealing that feature-guidance in visual search affects neuronal excitability as early as primary visual cortex.

Semantic Dimensions Support the Cortical Representation of Object Memorability
Recent work in vision sciences contends that objects carry an intrinsic property, memorability, that describes the likelihood that an object can be successfully encoded and later retrieved from memory. It has been shown that object memorability is supported by semantic information, but the neural correlates involved are largely unexplored. The present study explores these premises and asks whether neural correlates of object memorability can be accounted for by semantic dimensions. To investigate these questions, we combine three datasets: (1) feature norms for a database of ~1000 natural object images, (2) normative conceptual and perceptual memory data for those objects, and (3) neuroimaging data from an fMRI study collected using a subset (n=360) of those objects. We found that object-wise memorability elicits consistent brain activation across participants in key mnemonic regions (e.g., hippocampus and rhinal cortex), and that a substantial portion of the variance in this brain activity is mediated by the semantic factors describing these images. Regions with the strongest mediation effects are associated with sensory, motor, and visual processes, suggesting that the relationship between memorability and semantics may align with a sensory-functional account of the representation of concepts in the brain.

Interpersonal Synchronization in Mother-Child Dyads: Neural and Motor Coupling as a Mechanism for Motor Learning and Development in Preschoolers
Interpersonal movement synchrony (IMS) and brain-to-brain coupling play a crucial role in social behavior across species. In humans, IMS is often studied in structured tasks that require specific body movements, while spontaneous, unstructured movements have received less attention. In this study, we investigated both structured and spontaneous motor coordination in mother-child dyads. We recorded upper-body kinematics and dual-EEG from mothers and their preschool children during motor tasks and spontaneous face-to-face interactions. Our findings show that mother-child dyads synchronize their movements and neural activity, particularly in gamma band oscillations. This motor and neural synchrony evolves across task repetitions, with a strong correlation between motor and neural measures. Further, we observed that only motor synchronization was significantly related to the child's motor development stage, as assessed by the Movement Assessment Battery for Children. These results suggest that gamma band brain-to-brain coupling reflects joint motor coordination and mutual adaptation shaped by structured tasks and spontaneous interpersonal interactions.

Reliable sensory processing of superficial cortical interneurons is modulated by behavioral state
The GABAergic interneurons within cortical layer 1 (L1) integrate sensory and top-down inputs to modulate network activity and induce the plasticity underlying learning. However, little is known about how sensory inputs drive L1 interneuron activity. We used two-photon calcium imaging to measure the sound-evoked responses of two L1 interneuron populations expressing VIP (vasoactive intestinal peptide) or NDNF (neuron-derived neurotrophic factor) in mouse auditory cortex. We find that L1 interneurons respond to both simple and complex sounds with high trial-to-trial variability. However, these interneurons respond reliably to just a narrow range of stimuli, reflecting selectivity to spectrotemporal sound features. This response reliability is modulated by behavioral state and predicted by the activity of neighboring interneurons. Our data reveal that L1 interneurons exhibit sensory tuning and identify the modulation of response reliability as a potential mechanism by which L1 relays state-dependent cues to shape sensory representations.

Cell-type specific projection patterns promote balanced activity in cortical microcircuits
Brain structure provides the stage on which activity unfolds. Models linking connectivity to dynamics have relied on probabilistic estimates of connectivity derived from paired electrophysiological recordings or single-neuron morphologies obtained by light microscopy (LM) studies. Only recently have electron microscopy (EM) data sets been processed and made available for volumes of cortex on the cubic millimeter scale, thereby exposing the actual connectivity of neurons. Here, we construct a population-based, layer-resolved connectivity map from EM data, taking into account the spatial scale of local cortical connectivity. We compare the obtained connectivity with a map based on an established LM data set. Simulating spiking neural networks constrained by the derived microcircuit architectures shows that both models allow for biologically plausible ongoing activity when synaptic currents caused by neurons outside the network model are adjusted for every population independently. However, differentially varying the external current onto excitatory and inhibitory populations uncovers that only the EM-based model robustly shows plausible dynamics. Our work confirms the long-standing hypothesis that a preference of excitatory neurons for inhibitory targets, not present in the LM-based model, promotes balanced activity in the cortical microcircuit.

Differential effects of conventional transcranial direct current stimulation (tDCS) and high-definition transcranial direct current stimulation (HD-tDCS) of the cerebellum on offset analgesia
Background : Endogenous analgesic systems in the brain modulate pain perception. Offset analgesia (OA) describes the large decrease in perceived pain in response to a minor decrease in applied painful thermal stimulus. Here non-invasive brain stimulation (NIBS) of the cerebellum is used to probe OA. Methods : An OA protocol individualized to heat pain threshold (HPT) was applied via TSA-II (Medoc, Israel). Heat was ramped to HPT, with a transient temperature increase of HPT+1{degrees}C. NIBS interventions were applied prior to OA in 46 participants within a sham controlled repeated measures design. Cathodal cerebellar transcranial direct current stimulation (tDCS) and high-definition (4X1) transcranial direct current stimulation (HD-tDCS) were applied in separate experimental sessions to examine whether diffuse (tDCS) or focal (HD-tDCS) stimulation differentially modulates OA. Results : OA induced hypoanalgesia was robust, with 90% of responses showing a substantial drop in perceived pain ({delta}VAS) following the 1{degrees}C fall in temperature, with an average VAS decrease of 38 in response to the 1{degrees}C fall in temperature. Cathodal cerebellar HD-tDCS enhances the analgesic impact of OA on four OA parameters (OA latency, VAS minimum, VAS mean and VAS 2nd max) relative to pre-stimulation. Conventional tDCS modulates two OA metrics relative to pre-stimulation (OA duration, VAS 2nd max) with an increase in OA duration following sham tDCS. Conclusion : There is a differential influence of conventional and high-definition cerebellar NIBS on OA. This is suggestive of cerebellar modulation of OA and highlights the importance of electrode montage in delineating the influence of the cerebellum in pain processing.

Unraveling Integration-Segregation Imbalances in Schizophrenia Through Topological High-Order Functional Connectivity
Background: The occurrence of brain disorders correlates with detectable dysfunctions in the specialization of brain connectomics. While extensive research has explored this relationship, there remains a lack of studies specifically examining the statistical correlation between psychotic brain networks using high-order networks, considering the limitations of low-order networks. Moreover, these dysfunctions are believed to be linked to information imbalances in brain functions. However, our understanding of how these imbalances give rise to specific psychotic symptoms remains limited. Methods: This study aims to address this gap by investigating variations at the topological high-order level of the system with regard to specialization in both healthy individuals and those diagnosed with schizophrenia. Employing graph-theoretic brain network analysis, we systematically examine resting-state functional MRI data to delineate system-level distinctions in the connectivity patterns of brain networks. Results: The findings indicate that topological high-order functional connectomics highlight differences in the connectome between healthy controls and schizophrenia, demonstrating increased cingulo-opercular task control and salience interactions, while the interaction between subcortical and default mode networks, dorsal attention and sensory/somatomotor mouth decreases in schizophrenia. Furthermore, we observed a reduction in the segregation of brain systems in healthy controls compared to individuals with schizophrenia, which means the balance between segregation and integration of brain networks is disrupted in schizophrenia, suggesting that restoring this balance may aid in the treatment of the disorder. Additionally, the increased segregation and decreased integration of brain systems in schizophrenia patients compared to healthy controls may serve as a novel indicator for early schizophrenia diagnosis. Conclusions: We discovered that topological high-order functional connectivity highlights brain network interactions compared to low-order functional connectivity. Furthermore, we observed alterations in specific brain regions associated with schizophrenia, as well as changes in brain network information integration and segregation in individuals with schizophrenia.

Decision Confidence and Outcome Variability Optimally Regulate Separate Aspects of Hyperparameter Setting
Reinforcement learning models describe how agents learn about the world (value learning), and how they interact with their environment based on the learned information (decision policy). As in any optimization problem, it is important to set the process hyperparameters, a process which also is thought to be learned (meta-learning). Here, we test a key prediction of meta-learning frameworks, namely that there exist one or more meta-signals that govern hyperparameter setting. Specifically, we test whether decision confidence, in a context of varying outcome variability, informs hyperparameter setting. Participants performed a 2-armed bandit task with confidence ratings. Model comparison shows that confidence and outcome variability are differentially involved in hyperparameter setting. A high level of confidence in the previous choice decreased hyperparameter setting of decision noise on the current trial: when a trial was made with low confidence, the choice on the next trial tended to be more explorative (i.e. high decision noise). Outcome variability influenced another hyperparameter, the learning rate for positive prediction errors (thus affecting value learning). Both strategies are rational approaches that maximize earnings at different temporal loci: the modulation by confidence causes more frequent exploration early after a change point, the modulation by outcome variability is advantageous late after a change point. Finally, we show that (reported) confidence in value-based choices reflects the action value of the chosen option (irrespective of the unchosen value). In sum, decision confidence and outcome variability reflect distinct signals that optimally guide the setting of hyperparameters in decision policy and value learning, respectively.

Single and paired TMS pulses engage spatially distinct corticomotor representations in human pericentral cortex
Single-pulse transcranial magnetic stimulation (TMS) of the primary motor hand area (M1-HAND) can assess corticomotor function in humans by evoking motor evoked potentials (MEP). Paired-pulse TMS at peri-threshold intensity elicits short-latency intracortical facilitation (SICF) with early peaks at inter-pulse intervals of 1.0-1.8ms (SICF1) and 2.4-3ms (SICF2). The similarity between the periodicity of SICF and indirect (I-)waves in the corticospinal volleys evoked by single-pulse TMS suggests that SICF originates from I-wave generating circuits. This study aimed to explore the mechanisms of MEP generation by mapping the corticomotor representations of single-pulse and paired-pulse TMS targeting SICF1 and SICF2 peaks in 14 participants (7 female). MEPs were recorded from two hand muscles and the spatial properties of each corticomotor map were analyzed. For both hand muscles, we found a consistent posterior shift of the center-of-gravity (CoG) for SICF maps compared to single-pulse maps, with a larger shift for SICF1. CoG displacement in the SICF1 map correlated with individual SICF1 latencies. Further, ADM maps consistently peaked more medially than FDI maps and paired-pulse TMS resulted in larger corticomotor maps than single-pulse TMS. This is the first study to show that circuits responsible for SICF have a more posterior representation in the precentral crown than those generating MEPs via single-pulse TMS. These findings indicate that paired-pulse TMS probing SICF1, SICF2, and single-pulse TMS engage overlapping but spatially distinct cortical circuits, adding further insights into the intricate organization of the human motor hand area.