bioRxiv Subject Collection: Neuroscience's Journal

MCWs (MiCroWire sorter): A new framework for automated and reliable spike sorting in human intracerebral recordings.
Efficient and accurate spike sorting is critical for isolating single neurons from extracellular recordings to distinguish neural activity of interest. However, while the electrodes and acquisition systems for non-human electrophysiology have been enhanced over the past decades to enable higher-yield single-neuron detections, those advances have not been translated into human electrophysiology. Single-wire electrodes are still ubiquitously used, and although acquisition systems have augmented their signal-to-noise ratio over the last 15 years, we are still limited by their low electrode count. Moreover, unlike non-human recordings, human recordings often take place in hospitals where different noise sources and subject breaks can compromise the recording quality during experimental sessions. To bridge this gap, this work presents an automatic, open-source spike sorting pipeline that leverages contemporary computational capabilities and is tailored to single-neuron recordings from humans acquired via microwires. The pipeline is implemented in both MATLAB and Python, ensuring accessibility and compatibility across computational environments. Its modular and comprehensive structure supports customization and even opportunities for new developments as per the requirements of the user and the application. One feature is a data-driven automatic module to remove narrow-band interference, besides electrical line noise, which can be an essential tool while recording in clinical settings, particularly for online processing implementations. Following spike detection, the pipeline implements an artifact rejection module that separates waveforms that are unlikely to be associated with actual spikes. Additionally, we introduce a configurable feature-extraction, clustering, and benchmarking framework that not only allows flexibility in employing user-defined or conventional algorithms, such as wavelet transform with superparamagnetic clustering, but can also evaluate multi-method agreement among the different sorters. The pipeline also utilizes established and novel quality metrics to support semiautomatic curation of isolated clusters. Furthermore, we can integrate the customized pipeline with experimental tasks by removing task-unrelated waveforms (e.g., during a break in a task), and prevent over-clustering with the aid of metrics for comparing response profiles. Thus, the presented pipeline addresses the three-pronged objectives of algorithm-adaptability, rigorous validation, and human single-neuron recording optimization to support clinical and cognitive neuroscience applications.

Active Dissociation of Intracortical Spiking and High Gamma Activity
Cortical high gamma activity (HGA) is used in many scientific investigations, yet its biophysical source is a matter of debate. Two leading hypotheses are that HGA predominantly represents summed postsynaptic potentials or, more commonly, predominantly represents summed local spikes. If the latter were true, the nearest neurons to an electrode should contribute most to HGA recorded on that electrode. We trained subjects to decouple spiking from HGA on a single electrode using a brain-machine interface. Their ability to decouple them indicated that HGA is not primarily generated by summed local spiking. Instead, HGA correlated with neuronal population co-firing of neurons that were widely distributed across millimeters. The neuronal spikes that contributed more to this co-firing also contributed more to, and preceded, spike-triggered HGA. These results suggest that HGA arises predominantly from summed postsynaptic potentials triggered by synchronous co-firing of widely distributed neurons.

Brain-wide connectivity patterns of feedforward and feedback cortico-cortical neurons in the mouse secondary visual cortex
Feedforward and feedback cortico-cortical neurons are distinct yet spatially intermingled subtypes distributed across cortical layers, playing specialized roles in sensory and cognitive processing. However, whether their presynaptic inputs differ to support these functions remains unknown. Using projection- and layer-specific monosynaptic rabies tracing, we mapped brain-wide long-distance inputs to multiple feedforward and feedback neuron types in VISl (also known as LM), the mouse secondary visual cortex. Overall, long-distance input patterns for these feedforward and feedback neurons were largely similar, as all received the majority of their inputs from VISp, the primary visual cortex, along with substantial inputs from various other cortical and visual thalamic regions. Despite their similarities, these feedforward and feedback types differed in the proportion of long-distance cortical inputs originating from specific visual, retrosplenial, and auditory cortices. These findings reveal the input connectivity patterns of cortico-cortical neurons based on feedforward and feedback projections, providing an anatomical framework for future studies on their functions and circuit integration.

Sparsening and decorrelation of granule cell activity in the dentate gyrus by noradrenaline
The dentate gyrus in the hippocampus makes important contributions to the acquisition of episodic memories by transforming synaptic inputs from the entorhinal cortex into sparse and decorrelated activity patterns of its principal neurons, the granule cells. However, the underlying mechanism remains unclear. Using a combination of electrophysiological and optical recordings, together with optogenetic and pharmacological manipulations, we demonstrate that the release of noradrenaline plays a key role in this specialization via an enhancement of feedforward inhibition generated by cholecystokinin-expressing interneurons. By imposing coincidence detection with milliseconds temporal resolution onto granule cells, this enhancement of feedforward inhibition makes granule cell activity sparser and their firing patterns decorrelated. Since decorrelation contributes to efficient memory storage during auto-associative learning, these findings reveal a circuit mechanism by which an arousal signal facilitates memory formation in the hippocampus.

Dysregulated nuclear Lamin B1 in DYT1 dystonia thickens the nuclear lamina and disrupts 14-3-3 proteins
Childhood-onset DYT1 dystonia is caused by a heterozygous {Delta}E mutation in the TOR1A gene, which encodes a membrane-embedded AAA+ (ATPase Associated with diverse cellular Activities) ATPase. However, the mechanism by which {Delta}E induces dystonia remains poorly understood. Previously, using patient-derived neurons, we identified dysregulation of nuclear Lamin B1, at both expression levels and subcellular distribution, as a key contributor to DYT1 pathology. In the present study, we utilized DYT1 patient fibroblast cells and induced human neurons to investigate the molecular basis and consequences of Lamin B1 dysregulation. We found that elevated nuclear Lamin B1 thickens the nuclear lamina and deforms the nucleus, impairing nucleocytoplasmic transport. Proteomic analysis of human iPSC-derived neurons revealed that mislocalized Lamin B1 disrupts essential signaling pathways involved in neuronal function. Notably, 14-3-3 proteins, abundant brain molecular chaperones critical for neuronal development and homeostasis, were the most strongly associated with mislocalized Lamin B1. Functional studies showed that downregulation of 14-3-3 proteins impairs neurodevelopment in healthy neurons, while their upregulation rescues DYT1 neuronal defects by reducing Lamin B1 mislocalization. These findings elucidate a mechanistic link between nuclear deformation and cellular dysfunction in DYT1 dystonia and highlight Lamin B1 and 14-3-3 proteins as potential therapeutic targets.

Uncertainty in Deep Learning for EEG under Dataset Shifts
Objective: As artificial intelligence (AI) is increasingly integrated into medical diagnostics, it is essential that predictive models provide not only accurate outputs but also reliable estimates of uncertainty. In clinical applications, where decisions have significant consequences, understanding the confidence behind each prediction is as critical as the prediction itself. Uncertainty modelling plays a key role in improving trust, guiding decision-making, and identifying unreliable outputs, particularly under dataset shift or in out-of-distribution settings. The primary aim of uncertainty metrics is to align model confidence closely with actual predictive performance, ensuring confidence estimates dynamically adjust to reflect increasing errors or decreasing reliability of predictions. This study investigates how different ensemble learning strategies affect both performance and uncertainty estimation in a clinically relevant task: classifying Normal, Mild Cognitive Impairment, and Dementia from electroencephalography (EEG) data. Approach: We evaluated the performance and uncertainty of ensemble methods and Monte Carlo dropout on a large EEG dataset. Models are assessed in three settings: (1) in-distribution performance on a held-out test set, (2) generalisation to three out-of-distribution datasets, and (3) performance under gradual, EEG-specific dataset shifts simulating noise, drift, and frequency perturbation. Main results: Ensembles consisting of multiple independently trained models, such as deep ensembles, consistently achieved higher performance in both the in-distribution test set and the out-of-distribution datasets. These models also produced more informative and responsive uncertainty estimates under various types of EEG dataset shift. Significance: These results highlight the benefits of ensemble diversity and independent training to build robust and uncertainty-aware EEG classification models. The findings are particularly relevant for clinical applications, where reliability under distribution shift and transparent uncertainty are essential for safe deployment.

The somatosensory barrel cortex controls the spindlethalamocortical oscillation by frequency locking
The sleep spindle is a characteristic oscillation typically observed in NREM sleep and anesthesia. It is generated by a closed-loop thalamocortical circuit that is allegedly contributing to thalamocortical gating, sensory processing and memory consolidation. Yet, the circuit intricacy in terms of electrophysiological neuronal properties and connectivity has so far contributed to hinder a clear understanding of its regulation and function. In this study, we experimentally demonstrate that, when driven by the somatosensory cortex, the spindle circuit behaves as a macroscopic single-frequency self-sustained oscillator. We frequency-modulated cortical inputs to the thalamocortical spindle circuit by periodic microstimulation of the barrel cortex in the anesthetized rat. Cortical spindles exhibited synchronization by frequency locking and not resonance, displaying a characteristic Arnold tongue, a hallmark of the self-sustained oscillator. With a rate model of the barrel cortex-thalamus circuit reproducing the oscillator behavior we show that frequency-locking can govern synchronization under whisking.

Distinct brain-wide neural dynamics predict social approach behavior
Social behavior is essential for animal survival and adaptation, requiring the integration of sensory cues to guide interactions with conspecifics. A key component of social behavior is approach, where animals actively move toward social partners to maintain group cohesion, establish affiliations, and coordinate actions. While a continuous stream of social information is encoded across sensory modalities, it remains unclear whether a distinct neural process underlies social approach. Here, we developed a novel assay in which a head-fixed, tail-free zebrafish interacts with a freely swimming conspecific, enabling precise quantification of social behavior alongside whole-brain functional imaging at cellular resolution. We demonstrate that zebrafish approach behavior is jointly shaped by spatial and temporal information from conspecifics rather than by these features acting independently. Social approach behavior is preceded by distinct brain-wide neural activity patterns emerging seconds before movement onset, characterized by increased activity in a small subset of forebrain neurons and decreased activity in midbrain and hindbrain neuronal populations. These activity patterns reliably predict upcoming approach movements from each of these regions separately. Moreover, the extent to which neural activity distinguishes approach from non-approach movements predicts individual differences in social behavior, directly linking neural dynamics to behavioral variability. Together, our findings reveal a neural mechanism underlying social approach behavior, highlighting how a distributed yet functionally coordinated network facilitates social interaction.

Methyl gallate Attenuates Post-Stroke Emotional and Cognitive Symptoms by Promoting Hippocampal Neurogenesis via PI3K/GSK3 and AMPK Signaling
Restoring hippocampal neurogenesis is an effective strategy for post-stroke recovery. Methyl gallate (MET) exhibits neuroprotective properties. However, the effect of MET in improving brain functional recovery in the post-stroke depression (PSD) model and its underlying mechanism remains unknown. Single-cell data analysis showed that the cell types and molecular characteristics of PSD are similar to those of primary depression but exhibit weaker synaptic plasticity and stronger inflammatory signals. In addition, molecular docking studies revealed that MET exhibits a significant binding capacity with AMPK/GSK3, suggesting that MET mediates the neuroprotective effects of both. In this study, we created a post-stroke depression (PSD) model by performing physical restraint after ischemia and tested the treatment effects of MET. We observed that MET significantly attenuated PSD-induced depressive-/anxiety- behaviors associated with a reduction of stress hormone corticosterone and ACTH levels. Morris water maze and recognition task results indicate that MET can also alleviate cognitive impairments in the PSD model. In the hippocampus of the PSD model, MET improved the proliferation and differentiation of neural stem/progenitor cells (NSPCs). MET treatment significantly enhanced the activity of AMPK and decreased the activity of GSK3{beta}. Furthermore, in primary neural progenitors under hypoxia, both the PI3K inhibitor LY294002 and the AMPK inhibitor compound C blocked the effects of MET to promote neural development. Animal experiments also confirmed that LY294002/compound C treatment could reduce the effects of MET in antidepressant behaviours. Taken together, our results indicate that PI3K, as well as AMPK-mediated adult neurogenesis, is restored by MET to improve brain functions in the PSD model.

Integration of steady-state diffusion MRI with Neural Posterior Estimation (NPE) for post-mortem investigations
Post-mortem diffusion MRI plays a key role in investigative pipelines to characterise tissue microstructure, with long scan times facilitating the acquisition of datasets with improved spatial/angular resolution and reduced artefacts versus in vivo. Diffusion-weighted steady-state free precession (DW-SSFP) has emerged as a powerful technique for post-mortem imaging, achieving high SNR-efficiency and strong diffusion weighting in the challenging imaging environment of fixed tissue. However, the sophisticated signal forming mechanisms of DW-SSFP limit the integration of advanced microstructural models (e.g. incorporating time-dependence; Monte-Carlo simulations) with parameter estimation routines. Here, I investigate the integration of DW-SSFP with neural posterior estimation (NPE), a parameter inference technique leveraging concepts from Bayesian statistics and machine learning to directly estimate P({theta} | S) (i.e. the posterior distribution of parameters {theta} given signal S). A key challenge is that diffusion attenuation in DW-SSFP is dependent on tissue relaxation properties (T1/T2) and transmit inhomogeneity (B1), which must be incorporated into the NPE network for accurate modelling. By using NPE to estimate P({theta} | S,T1,T2,B1) (i.e. conditioning on S and known T1/T2/B1), using a Tensor representation, I demonstrate that NPE achieves accurate parameter estimation even in the presence of non-Gaussian (Rician) noise in low-SNR regimes. Comparisons with conventional non-linear least-squares (NLLS) using both synthetic and experimental DW-SSFP data (whole human post-mortem brain) give excellent agreement, with NPE providing 1000s of posterior samples in a matched evaluation time. Taken together, findings provide a framework to integrate advanced microstructural models with DW-SSFP, and an intuitive approach to incorporate conditional dependencies with NPE.

Motor Cortex Modulates Ipsilateral Limb Movement Through a Direct Cortico-Cerebellar Circuit
Motor cortex is traditionally associated with control of contralateral limb movements via corticospinal and cortico-ponto-cerebellar pathways. However, the contribution of ipsilateral motor cortical outputs on motor control remains unclear. Here, we identify and characterize a distinct population of cortico-cerebellar (C-C) neurons in the motor cortex that form monosynaptic projections to the ipsilateral cerebellar nuclei. The C-C neurons receive preferential local motor cortical inputs and exhibit projection patterns distinct from cortico-pontine projecting neurons. Using in vivo imaging and optogenetic perturbations, we show that these neurons are active during locomotion and transitions of volitional movements. Disruption of the C-C projection severely affects the locomotion and balancing. Interestingly, the C-C pathway is selectively involved in the initiation and coordination of ipsilateral forelimb movements, without affecting contralateral movement kinematics. These findings shed light on a non-canonical cortico-cerebellar pathway that supports ipsilateral motor control, complementing the traditional control mechanisms of the cerebral cortex over the contralateral motor domains.

Proteomic and in silico dissection of MetaAggregates in amyotrophic lateral sclerosis brains
RNA-binding proteins (RBPs), key translation regulators, are thought to be involved in the pathogenesis of amyotrophic lateral sclerosis (ALS). The pathological entities associated with ALS are known as ''MetaAggregates'': heterogeneous coaggregates composed of amyloids, RBPs, and RNA G-quadruplexes (rG4s). In this study, to explore the molecular constituents of ALS-associated MetaAggregates, we developed a proteomic approach using a psoralen-conjugated RBP and crosslinked it with a biotinylated rG4 to enable the isolation of MetaAggregates from ALS brain extracts. Single-cell RNA-seq using in vitro ALS models identified ELAVL4 as a cytoplasmic RBP and revealed the enrichment of an IGFBP2-derived rG4 structure in ALS-specific neurons. Mass spectrometry and amyloidogenicity-based principal component analysis revealed 79 candidate proteins with roles in RNA processing, metabolism, trafficking, and stress responses. Docking simulations highlighted a subset of proteins with potential pro-aggregation characteristics, diverse cytosolic associations and functional links to RNA processing relevant to ALS. Through proteomic and in silico dissection of ALS-associated MetaAggregates, the findings of this study establish a conceptual framework for the exploration of unrecognized amyloidogenic drivers of neurodegeneration.

The cryo-EM-delineated mechanism underlying mimicry of CXCR4 agonism enables widespread stem cell neuroprotection in a mouse model of ALS
G-protein coupled receptors (GPCRs) are transmembrane proteins that mediate a range of signaling functions and, therefore, offer targets for a number of therapeutic interventions. Chemokine receptor CXCR4, a GPCR, plays versatile roles in normal and abnormal physiological processes. Synthetic CXCR4 antagonists have been extensively studied and approved for the clinical treatment of cancer and other diseases. We recently elucidated the structural mechanisms underlying CXCR4 antagonism using cryogenic electron microscopy (cryo-EM). CXCR4 agonism by synthetic molecules is an unanticipated therapeutic intervention we recently unveiled. The structural mechanisms underlying those actions remain poorly understood yet could help elucidate a new class of drugs. Here we demonstrate a synthetic dual-moiety strategy that combines simplified agonistic and antagonistic moieties taken from natural agonistic and antagonistic chemokines, respectively, to design de novo peptide mimics of biological function of natural CXCR4 agonist SDF-1. Two peptides so generated, SDV1a and SDVX1 were shown to mimic the action of SDF-1 in activating CXCR4 signaling pathways and cell migration. The structural mechanism of these peptides in the mimicry of CXCR4 agonism was illustrated by cryo-EM structures of CXCR4 bound and activated by the peptides in the presence of G protein, revealing common interactions with the receptor by these peptides in comparison with SDF-1 that explain their close mimicry and conformational changes leading to CXCR4 signal activation. The therapeutic benefit of one of these peptides, SDV1a, was demonstrated in the SOD1G93A mouse model of the spinal motor neuron degenerative disease, amyotrophic lateral sclerosis (ALS) wherein the success of neuroprotective actions of transplanted human neural stem cells (hNSCs) is directly correlated with the expanse of diseased neuroaxis traversed by the donor cells; SDV1a enabled broader neuroprotective coverage while also permitting a much less invasive route of cell administration for extending life. Taken together, these results provide insights into the structural determinants of therapeutic CXCR4 agonism which may allow the design of adjunctive drugs that improve cell-based treatments of central nervous system (CNS) diseases.

Dissecting the contribution of recent reward versus recent performance history on cognitive effort allocation
An extensive body of literature has shown that humans tend to avoid expending cognitive effort, just like for physical effort or financial resources. How then, do we decide whether to put this effort in? Decision-making not only involves choosing our actions, but also the meta-decision of how much cognitive effort to invest in making this choice, weighing the costs of cognitive effort against potential rewards. Popular recent theories, grounded in the field of reinforcement learning, suggest that this cost-benefit trade-off can be informed by the opportunity costs of effort investment, which the brain may approximate by the estimated average reward rate per unit time. It follows from intuition that in a low reward environment, investing cognitive resources in the task at hand will less likely lead to missed opportunities. Recent studies provided support for this idea, showing that people exert more cognitive effort when reward rate is low. Here, we replicate one of the key previous findings but provide an important nuance to this result. Cognitive effort allocation was better explained by participants' recent performance history (i.e. accuracy rate) than average reward rate. In combination with the observation that participants were insensitive to the reward currently at stake, this invites a reinterpretation of these previous findings and suggests the need for further studies to assess whether environmental richness may indeed serve as a heuristic to modulate cognitive effort allocation.

Surprising effects of stimulus repetition on neuronal firing rates and gamma-band synchronization in awake macaque V1
Stimulus repetition is abundant, because the environment is redundant and/or because it is redundantly sampled. This offers an opportunity to optimize the processing of repeated stimuli. Indeed, stimulus repetition leads to classically described neuronal response decreases, and to more recently described neuronal gamma synchronization increases (sometimes preceded by decreases for a few trials). Here, we used a full-screen colored background (FSCB) and a flashed black bar, while recording multi-unit activity (MUA) and local field potentials (LFP) from area V1 of an awake macaque monkey. We found that the FSCB repetition induced neuronal response increases (sometimes preceded by decreases for a few trials) and gamma synchronization decreases (preceded by increases for a few trials). These effects are largely opposite to the dominant previous findings. Intriguingly, these surprising effects largely reversed when we isolated the responses to the flashed black bar. We discuss these findings, considering differences to previous studies with regards to the subject of the study, the stimuli and the task. We notice that in studies reporting classical results for gamma, sometimes in combination with firing rates, the stimuli were typically (partly) predictive of the reward. Here, we found non-classical results for the FSCB that was not reward predictive, and classical results for the black bar that was reward predictive. Whether this has revealed a general effect of reward predictive versus non-predictive stimuli will require further investigation with stimuli and task designs tailored specifically for this question.

Alpha and theta activity during reward anticipation are modulated by implicit expectations about sequential risk
Reward anticipation potentially guides sequential decision making, yet its underlying neural dynamics remain unclear. In this study, we investigated how risk levels and uncertainty modulate oscillatory brain activity during reward anticipation. EEG was recorded while participants (N = 44) performed a modified version of the Balloon Analogue Risk Task where balloon burst probabilities changed throughout the task, promoting uncertainty. We analyzed induced spectral power time-locked to three risk levels: early no-risk pumps, final successful pumps (preceding a cash out), and unsuccessful pumps (preceding a balloon burst). Time-frequency decomposition using Morlet wavelets revealed a parieto-occipital alpha power increase following early no-risk pumps, interpreted as disengagement from deliberative processing when anticipating sure rewards. In contrast, centroparietal alpha and frontocentral theta power decreased most prominently following final successful pumps, suggesting heightened attentional and expectancy-related processes in response to high reward potential. The results indicate that alpha and theta dynamics are sensitive to risk levels and reward expectations. These findings provide novel insights into the oscillatory mechanisms of reward anticipation in uncertain decision environments.

Conditions for replay of neuronal assemblies
From cortical synfire chains to hippocampal replay, the idea that neural populations can be activated sequentially with precise spike timing is thought to be essential for several brain functions. It has been shown that neuronal sequences with weak feedforward connectivity can be replayed due to amplification via intra-assembly recurrent connections. However, this phenomenon was thought to depend on inhibitory feedback, but its mechanisms were still unclear. Here, we arrive at a minimal spiking model that shows that feedback inhibition is not needed for this amplification to occur. We then introduce a population model of membrane-potential distributions that explains the spiking network behavior, and we analytically describe how different connectivity structures determine replay speed, with weaker feedforward connectivity generating slower pulses that can be sustained by recurrent connections. These pulses can only propagate if neuronal leak currents are slow enough with respect to the pulse speed. Together, our simulations and analytical results predict the conditions for replay of neuronal assemblies.

A novel, evolutionarily conserved inhibitory circuit selectively regulates dentate gyrus mossy cell function.
The mammalian dentate gyrus contributes to mnemonic function by parsing similar events and places. The disparate activity patterns of mossy cells and granule cells is believed to enable this function yet the mechanisms that drive this circuit dynamic remain elusive. We identified a novel inhibitory interneuron subtype, characterized by VGluT3 expression, with overwhelming target selectivity for mossy cells while also revealing that CCK, PV, SOM and VIP interneurons preferentially innervate granule cells. Leveraging pharmacology and novel enhancer viruses, we find that this target-specific inhibitory innervation pattern is evolutionarily conserved in non-human primates and humans. In addition, in vivo chemogenetic manipulation of VGluT3+ interneurons selectively alters the activity and functional properties of mossy cells. These findings establish that mossy cells and granule cells have unique, evolutionarily conserved inhibitory innervation patterns and suggest selective inhibitory circuits may be necessary to maintain DG circuit dynamics and enable pattern separation across species.

REVS: A New Open-Source Platform for High-Resolution Analysis of Rodent Wheel Running Behavior
Background Rodent wheel running is widely used in neuroscience and preclinical research to assess locomotor function, recovery post-trauma or disease, circadian rhythms, and exercise physiology. However, most existing wheel-running systems offer limited metrics, lack flexibility in hardware, or require costly proprietary software, reducing their usefulness for detailed behavioral phenotyping, especially in models of injury or rehabilitation. New method We developed REVS (Revolution Evaluation and Visualization Software), a low-cost, open-source hardware and software platform for analyzing and visualizing rodent wheel running behavior. REVS captures wheel revolutions using Hall effect sensors and computes 13 day-level behavioral metrics along with detailed bout-level data. Users can interactively explore high-resolution temporal features and export data in Open Data Commons (ODC)-compatible formats. REVS supports customizable wheel types, facilitating use in animals with motor and/or sensory impairments. Results We validated REVS using a mouse model of partial spinal cord injury, where fine motor control is compromised. REVS detected impairments in 10 of 13 behavioral metrics post-injury, with varied recovery trajectories across measures. Principal component analysis revealed that recovery was closely linked to bout quality and intensity, rather than timing. Comparison with existing methods Unlike commercial and open-source systems, REVS offers more detailed metrics, customizable wheel compatibility, seamless blending with common vivarium hardware, integrated data visualizations, and ODC-compatible data export. It also supports flexible analysis across individuals and groups. Conclusions REVS provides a powerful, scalable tool for granular behavioral phenotyping in rodent studies, enhancing reproducibility and revealing insights into subtle locomotor changes associated with injury, recovery, and intervention.

Deep mapping of the endomembrane system of cerebellar Purkinje neurons
Neuronal function relies on the precise spatial organization of intracellular membrane-bounded organelles involved in anabolism and Ca2+ sequestration, such as the Golgi apparatus, mitochondria and the endoplasmic reticulum (ER), along with structures involved in catabolism, such as lysosomes. Despite their known roles in energy supply, calcium homeostasis, and proteostasis, our understanding of how the anabolism-linked organelles are structurally arranged within neurons remains incomplete. Due to the tremendous complexity in the morphologies and fine structural features and interwoven nature of these intracellular organelles, particularly the ER, our understanding of their structural organization is limited, particularly, with regard to quantitative assessments of their sites of interaction and accurate measures of their volumetric proportions inside of a single large neuron. To approach this challenge, we used serial block-face scanning electron microscopy (SBEM) to generate large-scale 3D EM volumes and electron tomography on high-pressure frozen tissue of the rodent cerebellum, including the largest cells in the vertebrate brain, the cerebellar Purkinje neuron as well as the most abundant cell type in the vertebrate brain, the much smaller cerebellar granule neuron. We reconstructed the neuronal ultrastructure of these different cell types, focusing on the ER, mitochondria and membrane contact sites, to then characterize intracellular motifs and organization principles in detail, providing a first full map to quantitatively describe a neuronal endoarchitectome. At the gross level organization, we found that the intracellular composite of organelles are cell type specific features, with specific differences between Purkinje neurons and Granule cells. At the level of fine structure, we mapped ultrastructural domains within Purkinje neurons where ER and mitochondria associate directly. In addition to cell type specific differences, we observed significant subcellular regional variation, particularly within the axon initial segment (AIS) of Purkinje neurons, where we identified ultrastructural domains with sharply contrasting distributions of ER and mitochondria. These findings suggest a finely tuned spatial organization of organelles that may underpin the distinct functional demands along the axon. We expect that our subcellular map, along with the methods developed to obtain these maps, will facilitate future studies in health, aging and disease to characterize defined features, by developing a framework for quantitative analysis of the neuronal ultrastructure.

Light stimulation system for measuring pupillary light responses based on Maxwellian view optical system using a retina presentation type viewfinder
Purpose: To evaluate the feasibility of applying a retinal projection viewfinder based on the Maxwellian view (MV) optical system for measuring post-illumination pupil response (PIPR) by comparing its performance with a typical LED-based optical system. Methods: Twenty-two healthy participants underwent pupillometry using both the MV-based viewfinder and a typical LED-based system. Monochromatic red and blue light stimuli were presented for durations of 1 and 10 seconds. Pupil responses, including maximum constriction, PIPR amplitude after 6 seconds from the light offset, area under the curve (AUC) values of PIPR, and sustained slopes, were analyzed using a linear mixed-effects model to assess the differences between the two systems. Results: The MV-based viewfinder significantly enhanced net PIPR amplitude (p < 0.05) and sustained slope (p < 0.01) during 10-second light stimulation compared to the LED system, demonstrating its capability to effectively measure ipRGC-driven responses. In contrast, no significant differences were observed in the net AUC values. These results highlight that the MV-based viewfinder enables effective PIPR measurements by delivering constant and controlled light stimulation directly to the retina, minimizing the effects of dynamic pupil constriction during light stimulation. Conclusions: The MV-based viewfinder showed feasibility as an effective method for measuring PIPR without requiring pharmacological dilation and technical knowledge to build the MV system. This innovative approach has significant potential for clinical and research applications in pupillometry. Translational Relevance: Our methodology provides practical solutions for dilation-free effective measurement of PIPR, accelerating its translation from experimental tool to routine clinical diagnostic.

Domain-General Brain Networks Support Language Development
Understanding the neural basis of verbal intelligence across development requires disentangling the contributions of domain-general and language-selective brain systems. Although language is often considered a domain-specific function, complex language tasks also engage domain-general networks, such as the Default-Mode (DM) and Multiple-Demand (MD) systems. Yet how these systems contribute to the maturation of verbal competence remains poorly understood. Here, we examined this question using gray matter volume measures in an accelerated longitudinal dataset of children and adolescents from Beijing (N = 170), using the Verbal Comprehension Index (VCI) from the Wechsler Intelligence Scale as a benchmark for verbal abilities. We observed that individual differences in VCI were more strongly associated with structural maturation of domain general networks (DM and MD) than with the language-selective network, and that these effects varied with age. Targeted validation in an independent cohort from Chongqing (N = 150) confirmed significant contributions of domain-general networks in adolescence (13-15 years), highlighting the robustness of these developmental effects. These findings suggest that domain-general cortical systems play a critical and previously underappreciated role in the emergence of verbal intelligence during adolescence, with implications for understanding how large-scale brain networks support the development of abstract verbal reasoning.

AAV-based temporal APOE4-to-APOE2 replacement reveals rebound adaptation and RAB24-mediated Aβ and cholesterol dysregulation
APOE-targeted gene therapy offers a promising strategy for modifying Alzheimer's disease (AD) risk, yet the temporal dynamics and context-dependent effects of APOE isoform modulation remain poorly defined. Here, we developed a rapid AAV-based platform enabling inducible in vivo replacement of APOE4 with APOE2. In 5xFAD mice, sustained APOE4 expression exacerbated cognitive decline, A{beta} deposition (parenchymal and vascular), and glial activation, whereas long-term APOE2 expression--with concurrent APOE4 silencing--significantly reversed these pathological features and rescued cognitive function. In contrast, short-term APOE2 replacement conferred no benefit and unexpectedly worsened behavioral and pathological outcomes. Transcriptomic profiling revealed that APOE4-associated gene signatures were broadly reversed by long-term APOE2 expression, but paradoxically aggravated by short-term replacement. Among these, RAB24--a regulator of autophagic trafficking--was upregulated by APOE4 and short-term APOE2 but suppressed by long-term APOE2. RAB24 elevation impaired A{beta} clearance and cholesterol homeostasis via lysosomal retention in primary astrocytes and neurons. Together, these findings uncover a rebound-adaptation mechanism that shapes APOE2 therapeutic outcomes, identify RAB24 as a modifiable node in A{beta} and cholesterol metabolism, and establish a temporally controlled gene therapy platform to inform the design of future APOE-targeted interventions in AD.

Genetic inactivation of the Translin/Trax RNase activity alters small RNAs including miRNAs, disrupts gene expression and impairs distinct forms of hippocampal synaptic plasticity and memory
Neurons utilize RNA interference in the reversible translational repression of synaptically localized mRNAs, enabling rapid translation in response to synaptic activity. Two evolutionarily conserved proteins, Translin and Trax, form an RNase complex which processes miRNAs, tRNAs and siRNAs. To determine the specific role of the RNase activity of this complex in brain function, we employed a mouse line harboring a point mutation in Trax (E126A) that renders the Translin/Trax RNase inactive. At the molecular level, we found alterations in the levels of multiple small RNAs including miRNAs, tsRNAs and substantial downregulation of gene expression at the mRNA level in the hippocampus of TraxE126A mice. At the synaptic level, TraxE126A mice exhibit deficits in specific forms of long-term hippocampal synaptic plasticity. At the behavioral level, TraxE126A mice display impaired long-term spatial memory and altered open-field and acoustic-startle behavior. These studies reveal the functional role of Translin/Trax RNase in the mammalian brain.

Memorization of novel patterns in working memory in a model based on dendritic bistability
Working memory can hold many types of information and is crucial for cognition. Commonly, models of working memory maintain information such as hues or words by forming memory attractors through structured connectivities. However, real-world information can be novel, making it infeasible to use pre-trained attractors. In addition, most models-with or without attractors-have focused on maintaining binary categories instead of continuous information in each neuron. In the present study, we investigate how the brain might maintain working memory representations of arbitrary novel patterns with graded values. We propose an unstructured network model in which each neuron has multiple bistable dendrites. Each dendrite effectively implements fast Hebbian plasticity due to dendritic dynamics and dendrite-soma interactions. This network can maintain novel graded patterns under various perturbations without fine tuning of parameters. Through analytical characterization of network dynamics during the encoding and memory periods, we identify different conditions that yield either perfect memories or several types of memory errors. We also demonstrate memory robustness under various conditions and resilience to temporal inhibitory perturbations. Thus, this architecture provides robust and analytically tractable storage of novel graded patterns in working memory.

Synaptic properties of layer 6 auditory corticothalamic inputs in normal hearing and noise-induced hearing loss
Layer 6 corticothalamic neurons (CTs) provide strong feedback input that is crucial to perception and cognition in normal and pathological states; however, the synaptic properties of this input remain largely unknown, especially in pathology. Here, we examined the synaptic properties of CT axon terminals in the medial geniculate body (MGB), the auditory thalamus, in normal hearing mice and in a mouse model of noise-induced hearing loss. In normal hearing mice, we found that the synaptic strength of CTs to the core-type ventral subdivision of the auditory thalamus (MGv), which mainly conveys rapid sensory information, is stronger than the synaptic strength of CTs to the matrix-type dorsal subdivision of the auditory thalamus (MGd), which likely conveys higher-order internal state information. This is due to increased functional release sites (n) in CT[->]MGv compared to CT[->]MGd synapses. After noise trauma, we observed enhanced short-term facilitation in CT[->]MGd but not CT[->]MGv synapses. Our findings reveal a previously unknown mechanism of short-term synaptic plasticity after noise-induced hearing loss via which CTs enhance the throughput of matrix-type thalamus, likely to improve perceptual recovery via higher-order contextual modulation.

Somatosensory Realignment Following Single and Dual Force Field Adaptation
Evidence that adaptive motor learning coincides with a realignment of somatosensory perception has led to hypotheses that a shared mechanism underlies both processes, predicting similar properties. However, studies of somatosensory realignment with visuomotor adaptation have shown mixed support, possibly due to a confounding coactivation of sensory prediction errors and multisensory integration. While the former is thought to drive adaptation, both processes may contribute to somatosensory realignment. Here, we examined somatosensory realignment following force field adaptation, which is not confounded by multisensory integration. Across two experiments, we tested whether somatosensory realignment mimics three properties of adaptation in this paradigm. Our first experiment examined the specificity of somatosensory realignment to the perceptions of movement or static position and the generalization to reach directions adjacent to the one performed during the adaptation task. The results showed that force field adaptation coincided with a selective realignment of somatosensory perception of movement in the direction of the perturbing force that did not correlate with the magnitude of adaptation or generalize beyond the reach direction of the adaptation task. In a second experiment, we tested whether context-dependent dual adaptation to opposing force field perturbations coincides with a context-dependent dual realignment of somatosensory perception. The results showed no evidence of context-dependent somatosensory realignment after dual adaptation. Our results indicate that somatosensory realignment does not show the same properties as force field adaptation; however, it displays some coherence with the nature of the perturbation. Overall, our data suggest that somatosensory realignment and adaptation likely stem from distinct mechanisms.

Norepinephrine acts through radial astrocytes in the developing optic tectum to enhance threat detection and escape behavior
The ability to switch behavioral states is essential for animals to adapt and survive. We investigated how norepinephrine (NE) activation of radial astrocytes alters visual processing in the optic tectum (OT) of developing Xenopus laevis. NE activates calcium transients in radial astrocytes through 1-adrenergic receptors. NE and radial astrocyte activation shifted OT response selectivity to preferentially respond to looming stimuli, associated with predation threat. NE-mediated astrocytic release of ATP/adenosine reduced excitatory transmission by retinal ganglion cell axons, without affecting inhibitory transmission in the OT. Blockade of adenosine receptors prevented both decreased neurotransmission and the selectivity shift. Chemogenetic activation of tectal radial astrocytes reproduced NE's effects and enhanced behavioral detection of looming stimuli in freely swimming animals. NE signaling via radial astrocytes enhances network signal-to-noise for detecting threatening stimuli, with important implications for sensory processing and behavior.

Characterization of astrocyte density in the Pitt-Hopkins Syndrome mouse model of ASD
Transcription factor 4 (TCF4) is a proneural basic helix-loop-helix transcription factor that plays a critical role in brain development and is associated with a variety of psychiatric disorders including autism spectrum disorder (ASD), major depressive disorder, and schizophrenia. Autosomal dominant mutations in TCF4 result in a profound neurodevelopmental disorder called Pitt-Hopkins Syndrome (PTHS). Germline TCF4 loss-of-function (LOF) studies using human and mouse models have identified dysregulation in neural cell proliferation, genesis, and specification which lead to disruption in neuronal, astroglial and oligodendroglial lineages. In this study, we focused on the role of TCF4 in the genesis of the astrocyte lineage, specifically in the context of modeling PTHS. We show that germline heterozygous mutations in Tcf4 had no effect on the expression of astrocyte marker genes in primary astrocyte cultures and whole brain lysates. Immunohistochemical (IHC) analysis of pan- and subclass-specific astrocyte markers showed Tcf4 mutation had no effect on the proportions of astrocytes in the dorsal cortex and corpus callosum. Lastly, we tracked ventrally-derived astrocytes using an Nkx2.1 reporter mouse and observed that germline Tcf4 LOF did not result in misallocation of ventrally-derived astrocytes into the dorsal cortex, a phenotype previously observed when both Tcf4 alleles were conditionally deleted in the Nkx2.1 lineage. These data indicate that germline heterozygous TCF4 LOF, which models PTHS, does not appear to affect the astrocyte lineage at the cell population level.

Retinal waves reveal axial biases in modular patterns of cortical activity that predict future orientation preferences
Spontaneous activity before sensory onset is thought to guide the formation of functional neural circuits. In visual cortex, spontaneous activity prior to experience exhibits modular patterns that resemble future visually evoked orientation selective responses. However, the factors at this early stage that build the initial network interactions that support the orientation tuning of modular responses at eye opening remain unknown. Here we provide the first evidence that retinal waves could play an important role by shaping the modular biases in patterns of intracortical connectivity and lay the foundation for orientation selective responses. We demonstrate that slow propagating waves in developing cortex that are dependent on retinal activity recruit specific modular patterns during their movement across the cortical surface, resulting in strikingly elongated patterns of modular coactivity that predict visually-evoked orientation responses at eye opening. Thus axial biases in spontaneous patterns of co-activity are present before the onset of visual experience and could serve as the seed for the developmental emergence of modular orientation representations.

Loss of FMRP in microglia promotes degeneration of parvalbumin neurons and audiogenic seizures via progranulin insufficiency
Fragile X syndrome (FXS) results from loss of FMR1-encoded FMRP and is associated with reduced density of parvalbumin (PV) neurons; however, the mechanism underlying this abnormality remains unknown. Here we report that microglial FMRP regulates PV neuron density through lysosomal function. Mice with Fmr1 deletion in microglia exhibited audiogenic seizures (AGS) and decreased PV neuron density in the cortex and AGS-associated inferior colliculus (IC). FMRP increased the expression of lysosomal genes in microglia, including the progranulin-encoding Grn gene. Its loss in microglia led to impaired lysosomal function and increased apoptosis in microglia and PV neurons. Furthermore, PV neuron density in the IC was reduced similarly in male Grn+/-, Fmr1-/y, and Grn+/-;Fmr1-/y mice, and AAV8-mediated overexpression of progranulin rescued AGS and PV neuron loss in Fmr1-/y mice. This indicates that progranulin insufficiency is a determinant for PV neuron loss in FXS and elevating progranulin is a therapeutic strategy for FXS.

Dynamic Modulation of Beta-Band Oscillations in the LGN and Their Role in Visual Processing
Neuronal oscillations are a ubiquitous feature of thalamocortical networks and can be dynamically modulated across processing states, enabling thalamocortical communication to flexibly adapt to varying environmental and behavioral demands. The lateral geniculate nucleus (LGN), like all thalamic nuclei, engages in reciprocal synaptic interactions with the cortex, relaying retinal information to and receiving feedback input from primary visual cortex (V1). While retinal excitation is the primary driver of LGN activity, retinal synapses represent a minority of the total synaptic input onto LGN neurons, allowing for both retinogeniculate and geniculocortical signals to be influenced by nonretinal sources. To gain a holistic view of network processing in the geniculocortical pathway, we performed simultaneous extracellular recordings from the LGN and V1 of behaving macaque monkeys, measuring local field potentials (LFPs) and spiking activity. These recordings revealed prominent beta-band oscillations coherent between the LGN and V1 that influenced spike timing in the LGN and were statistically consistent with a feedforward process from the LGN to V1. These thalamocortical oscillations were suppressed by visual stimulation, spatial attention, and behavioral arousal, strongly suggesting that these oscillations are not a feature of active visual processing. Instead, they appear analogous to occipital lobe, alpha oscillations recorded in humans and may represent a signature of signal suppression that occurs during periods of low engagement or active distractor suppression.

The Functional Epididymal Amyloid Cystatin-Related Epididymal Spermatogenic (CRES) is a Component of the Mammalian Brain Extracellular Matrix
CRES is the defining member of a reproductive subgroup of family 2 cystatins of cysteine protease inhibitors. We previously showed that CRES and other subgroup members are part of a highly plastic amyloid-containing extracellular matrix (ECM) with host defense functions in the mouse epididymal lumen. Based on parallels between the epididymis and the brain, we hypothesized that CRES and CRES amyloids might also function within the brain including the ECM. Here we show that CRES is produced by hippocampal neurons and astrocytes in the male and female mouse and human brain. Further, approximately 50% of hippocampal astrocytes from aged mice, like the aged human donor samples, had significantly reduced levels of CRES compared to younger mice, suggesting an age-related decline in CRES could contribute to altered brain function. Immunofluorescence experiments showed CRES colocalized with the ECM markers phosphacan and wisteria floribunda agglutinin indicating that CRES is part of the ECM. CRES monomer and high molecular weight SDS-resistant forms were found in insoluble fractions of the hippocampus, cortex, cerebellum, and midbrain and bound to the protein aggregation disease (PAD) ligand, which preferentially binds amyloids but not protein monomers, suggesting a population of CRES exists in the brain as an amyloid structure. Collectively, our studies demonstrate that CRES/CRES amyloid is present in the mammalian brain and may contribute to ECM structure and function.

Spatially resolved mapping of monoacylglycerol lipase activity in the brain
Visualizing signaling systems in the brain with high spatial resolution is critical to understand brain function and to develop therapeutics. Especially enzymes are often regulated on the post-translational level, resulting in a disconnect between protein levels and activity. Conventional antibody-based methods have limitations, including potential cross reactivity and the inability of antibodies to discriminate between active and inactive enzyme states. Monoacylglycerol lipase (MAGL), an enzyme degrading the neuroprotective endocannabinoid 2-arachidonoylglycerol, is the target of inhibitors currently in clinical trials for the treatment of several neurological disorders. To support translational and (pre)clinical studies and fully realize the therapeutic opportunities of MAGL inhibitors, it is essential to map the spatial distribution of MAGL activity throughout the brain in both health and disease. Here, we introduce selective fluorescent activity-based probes for MAGL enabling direct visualization of its enzymatic activity in lysates, cultured cells and tissue sections. We show that oxidative stress, which inactivates MAGL through the oxidation of regulatory cysteines, reduces probe labeling , thereby validating the probes activity-dependence. Extending this approach, we developed an activity-based histology protocol to visualize MAGL activity in fresh-frozen mouse and human brain tissues. This approach revealed robust MAGL activity in astrocytes and presynaptic terminals within the mouse hippocampus, and further allows detection of MAGL activity in the human cerebral cortex. Collectively, these findings establish selective activity-based probes as powerful tools mapping MAGL activity with high spatial resolution across mammalian brain tissue.

Perineuronal Net and Inhibitory Synapse Remodeling on Striatal Fast-spiking Interneurons by Chronic Alcohol Exposure
Alcohol use disorder is characterized by persistent drinking in the face of negative consequences. Such inflexible drinking requires dorsolateral striatum fast-spiking interneurons, which comprise roughly 1% of all striatal neurons. How chronic ethanol exposure affects fast-spiking interneuron physiology is poorly understood. We discover in mice that chronic ethanol exposure induced a dramatic loss of GABAergic, but not glutamatergic, synapses onto dorsolateral striatum fast-spiking interneuron somata and proximal dendrites where perineuronal nets, a subdivision of the extracellular matrix, are enriched. We found that chronic ethanol exposure degraded these perineuronal nets and that enzymatically degrading perineuronal nets similarly reduced GABAergic transmission onto dorsolateral striatum fast-spiking interneurons. Modeling the effect of alcohol, we find that silencing extrinsic GABAergic projections to the dorsolateral striatum increased voluntary ethanol consumption. Taken together, these data suggest chronic alcohol exposure remodels perineuronal nets and inhibitory synapses on fast-spiking interneurons to facilitate alcohol drinking.