bioRxiv Subject Collection: Neuroscience's Journal

The Neuroprotective Effects of Ocimum gratissimum-Supplemented Diet on Scopolamine-Induced Memory Impairment in Mice Model of Alzheimer's Disease
Background: Cognitive decline is a hallmark of Alzheimer's disease, a progressive neurodegenerative illness primarily caused by the buildup of amyloid plaque, which is brought on by oxidative stress and neuroinflammation. Currently, therapeutic agents are focused on addressing clinical symptoms with associated side effects. This study aims to explore the neuroprotective potential of Ocimum gratissimum, a plant known for its richness in bioactive compounds. By investigating its effects on cognitive health, this study addresses a significant gap in the literature regarding dietary interventions for Alzheimer's disease. Methods: Thirty-six animals were divided into six groups of six mice each: the control group received distilled water intraperitoneally; the scopolamine group received only scopolamine (1 mg/kg i.p.); the three test groups were fed 5%, 10%, and 20% Ocimum gratissimum-supplemented diets while also receiving scopolamine (1 mg/kg i.p.); and the positive control group received Donepezil (5 mg/kg) followed by the injection of scopolamine (1 mg/kg i.p.). Donepezil was administered orally 30 minutes before the scopolamine injection. All treatments were administered daily for 14 consecutive days. Results: The supplemented diet groups showed significantly improved spatial memory and navigation compared to the scopolamine-only group. Biochemical analyses revealed that O. gratissimum mitigated scopolamine-induced oxidative stress and neuroinflammation, with marked improvements in antioxidant enzyme levels, reduced lipid peroxidation, and modulation of pro-inflammatory cytokines. Conclusion: Dietary intervention using Ocimum gratissimum leaf was able to improve spatial memory and protect against memory impairment, suggesting its potential as a neuroprotective agent against Alzheimer's disease-like pathology.

A proteomic signature of vascular dysfunction linked to tauopathy and degeneration in the aging brain
Small vessel disease (SVD) impacts healthy aging of organs across the body, yet its contributions to adverse brain aging remain poorly defined. Here we show thromboinflammation, a core feature of SVD, as a driver of adverse brain aging. We identify cerebrospinal fluid fibrinogen as a marker of brain thromboinflammation and screen neurovascular biosignatures mediating its impact on synaptic vulnerability along the full spectrum of brain aging from cognitively typical, amyloid-negative to cognitively impaired, amyloid-positive older adults. We identified 53 proteins mediating the effect of fibrinogen on synaptic markers in 1,655 donors from three independent cohorts. Single-cell transcriptomic mapping revealed mediator enrichment in neurovascular unit cells. Pathway analysis demonstrated dysregulation of angiogenesis, fibrosis, and immune signaling. Vascular and microglial-enriched biosignatures associated with compromised white matter integrity. These findings implicate thromboinflammation as an early, amyloid-independent pathway to neurodegeneration and tauopathy, establishing vascular health as fundamental to preserving brain healthspan.

Frontal and Parietal Contributions to Proprioception and Motor Skill Learning
Motor skill learning is the process of developing new movements with practice until they can be performed automatically. This necessitates the interaction of high-level cognitive processes with low-level sensorimotor mechanisms. Skill learning involves not only changes in the motor system but also proprioception (position sense). Proprioceptive deficits increase variability and decrease accuracy of movement. The somatosensory cortex, where low-level proprioception is processed, is known to play a role in motor skill learning, but the involvement of higher-level proprioceptive regions is unclear. Dorsolateral prefrontal cortex (DLPFC) has been linked to the high-level early stages of motor learning, and indirectly to proprioception. Supramarginal gyrus (SMG), an interface area between motor and sensory cortices, has been linked to higher-order proprioceptive processing. In this study, we asked how activity in DLPFC and SMG influences motor skill learning. Participants learned an upper limb motor skill designed to be spatially complex and dependent on proprioception: tracing a two-dimensional maze as accurately as possible within the desired speed range, using a Kinarm Endpoint robotic manipulandum. Proprioceptive acuity (sensitivity and bias) was assessed before and after continuous theta burst transcranial magnetic stimulation (cTBS) was applied to inhibit activity in DLPFC, SMG, or Sham. To measure motor skill, movement accuracy and variability were examined at the trained speed as well as at a faster (more difficult) and a slower (easier) speed. Skill was assessed before and after cTBS, after 40 training trials (early learning) and after another 80 trials (late learning). All three groups showed improvements in movement accuracy and variance, indicating they learned the maze tracing skill. However, the Sham group improved movement variability at the faster speed significantly more than the DLPFC or SMG groups did. This suggests that DLPFC and SMG are important for the more challenging aspects of motor skill learning, consistent with their association with higher-level proprioceptive function.

Microglia coordinate activity-dependent protein synthesis in neurons through metabolic coupling
De novo protein synthesis is required for long-lasting synaptic plasticity and memory, but it comes with a great metabolic cost. In the mammalian brain, it remains unclear which cell types and biological mechanisms are critical for sensing and responding to increased metabolic demand. Here we demonstrate that microglia, resident macrophages of the brain, coordinate metabolic coupling between endothelial cells, astrocytes, and neurons to fuel protein synthesis in active neurons. Increasing metabolic demand via a motor task stimulates microglia to secrete the hypoxia-responsive protein CYR61, increasing glucose transporter expression in brain vasculature. Depleting microglia reduces training-induced metabolic fluxes and neuronal protein synthesis, which can be reproduced by blocking CYR61 signaling. Thus, we define a neuroimmune metabolic circuit required for on-demand protein synthesis in mouse motor cortex.

Extracellular space diffusion modelling identifies distinct functional advantages of archetypical glutamatergic and GABAergic synapse geometries
The brain extracellular space (ECS) is a convoluted compartment of nano- and microscale interconnected ducts. A key step in signaling between neural cells is diffusion of signaling molecules through the ECS, yet, signaling is generally considered solely from the stance of cells and their properties. Where ECS diffusion is addressed, this is commonly done using volume-averaging techniques blind to individual signaling events and ECS geometry. We hypothesized that ECS geometry can shape local diffusion and thereby tune signaling arising from point-sources. To access the scale of individual transmitter release events and synapse geometries, we developed a computational diffusion model, DifFlux, based on super-resolved images of hippocampal ECS in live mouse brain slices and combined this with single molecule Monte Carlo diffusion simulations. Our approach allows us to simulate diffusion of molecules of our choosing in true live ECS geometries. We asked how the ECS shapes local diffusion in dense neuropil and along larger cellular processes in CA1 stratum radiatum. We observed local diffusional anisotropy and directionality imposed by ECS geometry. Further, we identified distinct functional advantages of dendritic spine and somatodendritic synapse ECS geometries, shedding light on the longstanding conundrum of why glutamatergic and GABAergic synapses are so conspicuously morphologically different. Our modelling broadly identifies ECS structure as a direct modulator of extrasynaptic signaling that can operate in parallel to conventional regulation mechanisms.

Natural Scene Coding Consistency in Genetically-Defined Cell Populations
Understanding how genetically-defined cell populations encode visual information remains a fundamental challenge in systems neuroscience. While extensive research has characterized individual cell responses to simple stimuli such as static gratings, the population-level coding principles that govern naturalistic visual processing across cell types remain largely unexplored. We analyzed population responses from 43,018 neurons across 12 genetically-defined cell types in 243 mice from the Allen Brain Observatory, comparing representational geometry between natural scenes and static gratings. We found that inhibitory cell populations (VIP, SST, PV) cluster distinctly in representational space almost independent of anatomical location when responding to natural scenes but not static gratings, suggesting preserved cell-type specific computational functions specific for natural scenes. To assess coding capacity of a population of neurons, we developed Inter-Individual Representational Similarity (IIRS), which measures consistency of neural representations across different individuals in response to an ensemble of stimuli. All inhibitory populations showed significantly higher IIRS for natural scenes compared to static gratings, indicating consistent encoding of naturalistic visual features across individuals comparable to excitatory populations (Cux2, Rorb, Rbp4). Parallel analysis of neural networks trained on natural images (ImageNet) with different random initializations revealed similar patterns: models showed higher cross-initialization similarity for naturalistic stimuli compared to static gratings, suggesting that cross-individual consistency emerges when experimental stimuli engage the features that neural circuits are adapted to extract. These findings establish IIRS as a metric for identifying coding capacity in cell populations and reveal that inhibitory cell populations encode consistent aspects of natural scenes across individuals, indicating these circuits may have evolved specialized tuning for naturalistic visual environments.

Global organisation of structural covariance networks derived from parcellated cortical surface area in atypical populations.
Higher-order features of brain organisation are powerful measures for understanding the relationship between brain and experience. In particular, the global arrangement of structural features of the cortex provides insight into neurodevelopmental processes that underlie individual differences in perception and cognition. Structural covariance networks (SCNs), which capture regional coordination of brain morphometry, are an efficient method to derive global properties of the cortex. However, their interpretation relies on an array of methodological choices that are often inconsistent between studies. Using a hierarchically-clustered version of the Human Connectome Project (HCP) atlas, we constructed SCNs of regional cortical surface area for groups with four different conditions -- synaesthesia, autism, early psychosis, and anxiety or depression -- and compared global network properties with those of age- and sex-matched controls. SCNs for synaesthesia and autism showed globally stronger connectivity, with specific increases at moderate cortical distances, as well as lower network complexity. Conversely, the SCN for early psychosis showed a globally lower connectivity and a greater complexity, while depression and anxiety showed few differences compared to controls. The results for autism and depression were replicated across two datasets each. These findings support the notion that synaesthesia and autism share neurodevelopmental mechanisms, while psychosis may involve a diverging process. This study is also an important proof of principle for analysing diverse populations under one methodological framework.

Acetylation of Axonal G3BP1 through ELP3 Accelerates Axon Regeneration
Nerve injury triggers localized translation of axonal mRNAs to respond to injury and nerve regeneration. The core stress granule protein G3BP1 sequesters axonal mRNAs in granules before and after axotomy. G3BP1 granule disassembly can be regulated by post-translational modifications, including phosphorylation of S149 phosphorylation and acetylation of human K376 (mouse K374). Axonal G3BP1 undergoes phosphorylation after axotomy, but acetylation of G3BP1 in axons was unknown. Here we show that rodent G3BP1 undergoes K374 acetylation after axotomy is ELP3-dependent, which enhances axonal protein synthesis, accelerates nerve regeneration, and supports functional recovery. ELP3-depleted neurons exhibit reduced axon growth and increased axonal G3BP1 granules. The proximal axons degenerate rapidly despite maintaining soma connectivity, an effect prevented by expression of acetylmimetic G3BP1.Together, these findings identify G3BP1 acetylation via ELP3 as a critical regulator of both axonal regeneration and neuronal resilience, revealing a post-translational mechanism that links stress granule regulation to neuronal repair and protection.

Aerobic Exercise Intensity: A Dose-Response Effect on Motor Adaptation and Learning
Acute aerobic exercise (AEX) can enhance motor learning. While AEX intensity likely plays a key role, there is mixed evidence for AEX-enhanced motor skill acquisition and learning across a spectrum of exercise intensities. This may stem, in part, from inconsistent AEX parameters (i.e., intensity, structure, and duration) employed within and across studies. Additionally, evidence suggests that AEX can enhance a specific form of motor learning, namely motor adaptation. Moderate- and high-intensity AEX can increase motor adaptation, but evidence remains limited and inconsistent. Hence, the impact of AEX intensity on motor adaptation remains unclear. Here, we investigated the influence of AEX intensity on motor adaptation, while controlling for AEX structure and duration. Eighty young adults were assigned to four cycling AEX/Rest groups (n=20/group): 20 min of light (LIIT), moderate (MIIT), or high (HIIT) intensity interval training, or Rest (control). AEX consisted of four 3-min cycling intervals (LIIT, 35% heart rate reserve [HRR]; MIIT, 55%HRR; HIIT, 80%HRR) and 2-min active recovery (25%HRR). Participants practiced a visuomotor rotation task immediately after AEX/Rest (adaptation) and at a no-AEX 24 h retention test (motor learning). We found that: (1) all AEX intensities enhanced motor learning compared to Rest, and (2) HIIT enhanced motor adaptation and learning to the greatest extent, followed by MIIT then LIIT. This is the first study to demonstrate a dose-response effect of AEX intensity on motor adaptation and learning. Our results highlight the importance of considering intensity when prescribing AEX in sports and clinical contexts to promote motor learning.

Dissociable Roles of Primary Motor and Supplementary Motor Cortex in Shaping the Neural Drive to Muscle
The primary motor cortex (M1) and supplementary motor area (SMA) are critical for motor execution and planning, yet their distinct causal contributions to modulating the neural drive to muscles remain incompletely understood. To dissociate their roles, we applied inhibitory transcranial magnetic stimulation (TMS) over M1, SMA, or a sham condition in 72 healthy participants and characterized the activity of single motor units from high-density EMG recorded during a sustained isometric contraction. Our results revealed a clear functional divergence. M1 inhibition produced a direct failure of motor output, causing a rapid force decline compared to sham, which was strongly correlated with a reduction in motor unit firing rates. Conversely, SMA inhibition did not impair net force. Instead, it altered the fundamental structure of the motor command, compelling a compensatory strategy characterized by a reliance on smaller-amplitude motor units with lower firing rates and a marked degradation of the low-frequency (delta-band) coherence that organizes stable output. These results provide direct causal evidence that M1 directly dictates the magnitude of motor output via population firing rates, while SMA orchestrates the composition and temporal structure of the active motor unit pool to generate an efficient and stable command.

Stimulation with ECoG electrodes modulates cortical activity and sensory processing in the awake mouse brain
Electrical stimulation has been widely used to probe neural network properties and treat dysfunction. Electrocorticography (ECoG) electrodes, long used for activity monitoring, can also stimulate the brain in a minimally invasive and chronic manner. However, how cortical surface electrical stimulation impacts cortical network activity remains poorly understood. Using in vivo calcium imaging in the awake mouse brain with chronically implanted ECoG electrodes, we measured how electrical stimulation modulates the activity of visual cortical neurons, including during concurrent visual stimulation. We found that cortical surface electrical stimulation initially activates L2/3 neurons followed by prolonged inhibition lasting seconds after stimulation. Electrical stimulation suppresses the activity of neurons at their preferred grating orientation but enhances their responses to non-preferred visual stimuli, thereby reducing sensory feature selectivity. By measuring how electrical stimulation modulates the activity of inhibitory neuron subtypes including PV, SST, and NDNF interneurons, we propose a circuit model in which L1 NDNF interneurons are strongly activated by cortical electrical stimulation and, in turn, inhibit L2/3 excitatory neurons and PV interneurons through volume transmission of GABA.

Connectivity biases generate a learning hierarchy in the Drosophila mushroom body
Learning and memory centers must balance maximizing coding capacity with prioritizing biologically relevant information. Expansion layers, a circuit motif common to many learning and memory centers, including the insect mushroom body, transform dense sensory representations into sparse, distributed ones, and theoretical models propose that random connectivity within these layers maximizes coding capacity by generating highly discriminable responses. Yet this solution creates a fundamental problem: purely random connectivity treats all stimuli equally, without prioritizing survival-relevant cues over neutral ones. Here, we show that the Drosophila melanogaster mushroom body resolves this capacity-selectivity trade-off through systematic biases in projection neuron-Kenyon cell connectivity. Although connectivity is random at the single-cell level, some projection neuron types connect up to 15-fold more frequently than others. These biases translate directly into function: Kenyon cell responses scales with projection neuron connectivity, and the breadth of odor-evoked responses predicts learning performance. Odors activating more than 20% of Kenyon cells drive robust associative memories, whereas those activating fewer than 10% are poorly learned. VL1 projection neurons are a notable exception: despite their weak connectivity, they elicit broad Kenyon cell activity but fail to support learning, revealing a circuit-level gate on learning. These results show that the mushroom body embeds a learning hierarchy in its connectivity architecture, prioritizing ethologically relevant odors while preserving coding capacity for diverse associations.

A sleep promoting role of phototransduction in Drosophila melanogaster
Daily sleep-wake cycle is a conserved behaviour defined by locomotion quiescence and enhanced responsive threshold to sensory stimuli. Both circadian rhythm and a homeostatic process determine the daily sleep profile, which is also regulated by environmental light, a major sensory input to regulate circadian rhythm and alertness. With decades of investigation in Drosophila, the cellular and circuital mechanism underlying light-mediated circadian synchronization are well-established, yet the direct relationship between light/visual input and sleep remains unclear. To address this knowledge gap, we have started an investigatory survey of sleep behaviour using classic mutant lines to manipulate phototransduction and downstream neural transmission. We observed consistent day sleep fragmentation in flies with mutations in multiple phototransduction components. We also found hyperpolarised Drosophila photoreceptor resulted in shorter day sleep. We found a severe reduction in locomotor speed in several visual mutants during normal waking time preventing assessment of their sleep-linked immobility. In summary, we provide a rigorous quantification of several phototransduction genes and reveal the key role of visual input to promote sleep.

Inhibitory-stabilization is sufficient for history-dependent computation in a randomly connected attractor network
For effective information processing, the response to a sensory stimulus should depend on both the incoming stimulus and the history of prior stimuli. Existing models of neural circuits based on multiple attractor states produced with strong self-excitation can exhibit these properties, but they do not stabilize at biologically realistic firing rates. We demonstrate how a randomly connected inhibition-stabilized attractor network can preserve the computational abilities of recurrent excitatory networks, while stabilizing at arbitrarily low firing rates. Not only does excitatory-inhibitory balance stabilize network activity, inhibitory-stabilization also plays a functional role in history-dependent computation: transient oscillations made possible by inhibitory feedback are sufficient for state-dependent responses to stimulation. Such networks may underlie many cognitive tasks, suggesting a functional role for inhibition-stabilized dynamics in cortical computation.

A hyperglycosylated form of Kv1.2 upregulated in LGI1 knockout mice
Kv1 voltage-gated potassium channels determine key functional neuronal properties. Their activity is modulated by subunit composition and post-translational modifications such as phosphorylation and glycosylation. Using an antibody directed against a phosphotyrosine (Y458) located in the C-terminal tail of Kv1.2, we identified yet unreported high molecular weight forms of Kv1.2 among them, a phosphorylated and heavily glycosylated 100 kDa form. Owing to the significant downregulation of Kv 1.2 in LGI1-dependent autosomal dominant lateral temporal lobe epilepsy, we investigated, in total brain and the hippocampal formation of both WT and Lgi1-/- mice, the distribution of phosphoY458 Kv1.2 and we compared their respective proteomic interactomes with those of Kv1.2. In addition to major differences between the interactomes of pY458 Kv1.2 and Kv 1.2 in WT and Lgi1-/-, we found a major reshaping of pY458Kv1.2 molecular neighbourhood between WT and Lgi1-/- as well as a significant upregulation of the glycosylated form in Lgi1-/- .

DISTINCT NEURAL SIGNATURES OF HIPPOCAMPAL POPULATION DYNAMICS DURING LOCOMOTION-IN-PLACE
Hippocampal CA1 neurons modulate their activity with movement variables such as time, distance, and speed, yet it remains unclear how these representations reorganize across behavioral states, from externally driven to self-paced movement and immobility. Here, we investigated how sensory events that initiate or terminate locomotion, structure CA1 population codes and how these codes reorganize across sensory-driven locomotion, spontaneous locomotion, and forced immobility. Using two-photon calcium imaging in head-fixed Thy1-GCaMP mice (n = 5) performing the air-induced running task on a non-motorized conveyor belt, we examined neuronal firing-rate modulation across a series of behavioral configurations designed to probe distinct forms of locomotion-in-place. In the No-Brake (locomotion-permitted) condition, the belt rotated freely, allowing animals to execute full cyclic limb movements in response to air stimulation. In the Brake (immobility) condition, the belt was fixed, restricting movement to partial or attempted locomotor motions. Firing-rate modulation with respect to time, distance, and speed was quantified using linear (Pearson correlation) and nonlinear (mutual information) metrics under permutation testing in the natural reference domains. Behaviorally, air stimuli produced faster, sustained running during air-on and more variable, self-paced movement during air-off. Neurally, a larger fraction of CA1 cells was active and significantly modulated during air-off. Within the modulated set, singularly tuned cells (time, distance, or speed) predominated over mixed-tuned cells, and speed-modulated cells peaked earlier after stimulus onset or offset than time- or distance-modulated cells. Under Brake, CA1 activity was predominantly singularly tuned to time or movement-in-place, with stronger movement modulation and engagement post-stimulation. Despite substantial single-cell turnover across configurations and phases, population-level analyses revealed a coherent, air-phase-locked organization and distinct movement-related populations across Brake and No-Brake conditions. These results indicate a state-dependent reweighting of sensorimotor features implemented atop a conserved ensemble scaffold in CA1.

Crosstalk between the Methyl-Cytosine Dioxygenase TET3 and the Methyl-CpG-binding protein MECP2 Controls Neuronal Maturation
Active DNA demethylation depends on Ten-Eleven-Translocation (TET) enzymes, which oxidize 5-methylcytosine (mC) to 5-hydroxymethylcytosine (hmC) and further derivatives. Mutations in TET3, encoding the predominant neuronal isoform, lead to Beck-Fahrner syndrome, a neurodevelopmental disorder. Using human iPSC-derived neurons, we show that TET3 is dispensable for neuronal specification but critical for subsequent maturation. Differentiating TET3-deficient neurons exhibit delayed transcriptional and proteomic transitions, altered synaptic signatures, and impaired network activity, indicating delayed functional maturation. Mechanistically, we identified an interaction between TET3 and the mC/hmC-binding protein MECP2, pathogenic variants of which cause Rett syndrome. MECP2 negatively regulates TET3 activity, as demonstrated in functional assays and by inverse hmC patterns in MECP2- and TET3-deficient neurons. Despite this, MECP2- and TET3-deficient neurons exhibit highly similar phenotypes later in differentiation. Our findings uncover a functional interplay between TET3 and MECP2 that coordinates DNA methylation and chromatin dynamics during neuronal maturation, suggesting a shared pathogenic mechanism in Beck-Fahrner and Rett syndromes.

Temporal dynamics of noradrenaline release at fine spatial scales during motor learning
Noradrenaline (NA) released from the locus coeruleus (LC) has been known to play pivotal roles in arousal, sensory processing, decision-making, and learning through global release across the entire neocortex. Recent studies have demonstrated heterogeneous and modular NA release in distinct brain regions and highlighted its spatiotemporal dynamics across the neocortex. However, the spatiotemporal specificity of NA release at fine scales within a single brain region remains unclear, and it has not been reported whether the release patterns evolve functionally throughout prolonged learning processes such as motor skill acquisition. Here, by employing in vivo two-photon imaging with various genetically encoded NA sensors, we reveal that behavior-induced NA release in the primary motor cortex (M1) is spatially heterogeneous at the scale of local microcircuitry. Over the course of learning, the release pattern is locally refined, achieving consistent spatial precision within M1. Intriguingly, pharmacological manipulations that disrupt the spatial specificity also alter local neurons activity and representations. Furthermore, LC-NA axonal calcium imaging uncovered two distinct temporal activity patterns, in which non-behavior-related rapid axonal activity (sub-second duration events) profoundly affect the temporally precise behavior-induced persistent axonal activity (seconds duration events). Closed-loop optogenetic manipulations that bi-directionally modulate non-behavior-related rapid events directly impacted the learning process. Together, our results provide novel insights into the temporal dynamics of NA release at fine spatial scales within one brain region and underscore the critical role of local NA specificity in regulating circuit plasticity during motor skill acquisition.

Sustained dynamics and modality-specific network reconfigurations define crossmodal prediction error
Predictive processing theories posit that the brain continuously generates expectations about incoming sensory inputs, updating them through prediction errors. While extensively studied within single modalities, it remains unclear how prediction errors unfold when expectations and violations occur across different senses. Using a local-global hierarchical oddball paradigm combined with high-density EEG in 47 participants, we contrasted unimodal and crossmodal prediction errors across auditory, visual, and somatosensory domains. We found that crossmodal violations elicit temporally sustained cortical responses which diverge from the transient, localised dynamics observed for unimodal prediction errors. Temporal decoding revealed that crossmodal effects maintain shared prolonged neural representations, suggestive of supramodal integration across all levels of cortical processing. Computational modelling further demonstrated that crossmodal prediction errors reorganize effective connectivity within and between sensory hierarchies, engaging distinct early cortical pathways depending on the sensory combination. Our findings refine hierarchical predictive coding for crossmodal transitions by demonstrating that, unlike unimodal prediction errors, crossmodal prediction errors recruit dedicated prolonged supramodal representations and flexibly adapted modality-specific networks.

From optimality to reality: decision-making in basal ganglia output
Value-based decision-making requires translating abstract representations of alternative values into concrete decisions. Although this process is widespread, the neural mechanisms underlying this transformation remain unclear. Here, we address this question by examining neural dynamics in the substantia nigra pars reticulata (SNpr) of two monkeys performing a decision-making task. We found a dynamic shift in SNpr activity during the decision process. Before choice, SNpr activity encoded the optimal selection regardless of the eventual action taken, representing the action and value of the optimal choice even when the non-optimal target was ultimately selected. During choice, the SNpr shifted to code the actual selection made by the monkey. This pattern was reinforced by trials with suboptimal choices, revealing a clear transition in SNpr coding from representing the abstract value of the optimal selection to encoding the chosen action, thus suggesting a neural mechanism that transforms value representations into concrete decisions.

Brain insulin signaling restores deficits in striatal dopamine release in overweight male mice with preexisting low D2-receptor expression
Obesity is characterized by insulin resistance, motivational impairments, and, in some cases, reduced availability of dopamine D2 receptors in the brain. However, whether the low D2 receptor levels represent a predisposing factor or a consequence of obesity, and how these processes are mechanistically linked, remains unclear. Here, we directly tested this causal relationship by selectively reducing D2 receptor density in striatal neurons. Male, but not female, mice with a low density of striatal D2 receptors consumed more food, gained more weight, and developed metabolic features of peripheral insulin resistance despite being maintained on standard chow. Motivational deficits preceded weight gain, manifesting as delayed circadian locomotor onset, reduced physical activity, and diminished effort to obtain food. In the brain, male mice with low D2 receptor density showed reduced dopamine release capacity and age-dependent alterations in brain insulin sensitivity. Prior to weight gain, brain insulin responses were blunted compared to those of controls, in which insulin potentiates dopamine release and enhances striatal acetylcholine signaling. Once overweight, however, these mice exhibited brain insulin hypersensitivity, with insulin strongly restoring dopamine release capacity. Together, these findings demonstrate that low striatal D2 receptor density predisposes male mice to an obesity-like phenotype through early dopaminergic dysfunction that precedes weight gain and is later compensated by insulin hypersensitivity in the brain.

Tegmental atrophy in isolated REM sleep behaviour disorder: Ex vivo-informed in vivo imaging
Isolated rapid eye movement (REM) sleep behaviour disorder (iRBD) is an early-stage synucleinopathy characterized by brainstem pathology. In rodents, the pontine tegmentum contains an REM sleep centre, the sublaterodorsal nucleus (SLD), which expresses corticotropin-releasing hormone binding protein (CRHBP). While the involvement of brainstem pathophysiology is thus implicated in iRBD, its solid evidence remains scarce in humans due to the difficulty in identifying small brainstem nuclei with conventional MRI technology alone. Here, we aimed to detect tegmental atrophy in iRBD with voxel-based morphometry (VBM) analysis combined with a novel human brainstem atlas. Structural MRIs from 98 patients with iRBD and 114 controls were analysed to investigate grey matter volume (GMV) using VBM. Our unique approach involved detailed assessments of the VBM results, guided by a high-resolution MRI-based atlas of the human brainstem. This brainstem atlas was founded on ex vivo MRI of 10 postmortem human specimens. We validated it with CRHBP immunostaining, which aided in identifying putative REM sleep-regulating nuclei in humans. We applied this brainstem atlas to identify atrophy in specific brainstem regions in iRBD and correlate their volumes with clinical measures, including autonomic functions. VBM revealed a focal cluster of grey matter atrophy in the dorsal pontine tegmentum of iRBD patients, including the laterodorsal tegmental nucleus, ventral part (LDTgV) and the pedunculopontine tegmental nucleus (PTg). Our atlas-based analysis confirmed the LDTgV as the site of most conspicuous atrophy, revealing a significant volume reduction in iRBD patients compared to controls with a moderate effect size (Cohen's d = 0.46, Bonferroni-corrected p = 0.019). Furthermore, greater atrophy in the LDTgV and the PTg was associated with more severe autonomic dysfunction as measured by Scales for Outcomes in Parkinson's Disease-Autonomic dysfunction (SCOPA-AUT) scores (partial r = -0.237, p = 0.019 and partial r = -0.236, p = 0.019, respectively). Histological analysis confirmed that the LDTgV is selectively enriched with CRHBP-positive neurons, a putative marker for REM sleep-on neurons. We provided novel evidence for the involvement of LDTgV, the putative human homolog of the murine SLD, in iRBD. The present findings advance our understanding of the neuroanatomical basis of iRBD and will contribute to the development of early biomarkers for -synucleinopathies.

Effect of Large Language Models on P300 Speller Performance with Cross-Subject Training
Amyotrophic lateral sclerosis (ALS), a progressive neuromuscular degenerative disease, rapidly impairs communication within years of onset. This loss of communication necessitates assistive technologies to restore interaction and independence. One such technology, the P300 speller brain-computer interface (BCI), translates EEG signals into text by tracking a subjects neural responses to highlighted characters on a screen. A central challenge in P300-based research is enhancing performance to enable faster and more efficient user interaction. In this context, this study addresses key limitations, particularly in the training of multi-subject classifiers, and integrating advanced language models to optimize stimuli presentation and word prediction, thereby improving communication efficiency. Specifically, we introduce three key innovations:- Advanced multi-subject classifier training- Integrating and evaluating impact of numerous large language models (LLMs) on speller performance- Determining P300 LLM performance bounds using an ideal LLM with perfect prediction We conduct extensive simulations using randomly sampled EEG data. Our results demonstrate substantial speed improvements in typing passages that include rare and out-of-vocabulary (OOV) words. The magnitude of improvement depends on the type of language model used. More specifically, character-level models provide typing speed improvements of approximately 10%, while open-source LLMs such as Llama, Mistral and GPT2 achieve around 40% improvement through efficient word prediction. Additionally, we construct an ideal LLM to establish theoretical performance limits and show that many modern LLMs achieve performance levels within 10% of it. Further, we show that these LLM-driven speed improvements generalize across classifiers, including those designed to reduce subject-specific training.

Behavioural and neuronal insights into multisensory combination of unpracticed cues.
Effective decision-making requires integrating multiple information sources, weighted by their reliability and context. While classic studies show near-optimal cue combination with well-learned signals and extensive feedback, everyday choices often rely on unfamiliar or cross-modal cues without such training. We examined cue combination under these conditions using an online perceptual estimation task in large and diverse participant cohorts. Participants combined unpracticed cues, including visual motion direction, spatial visual information, and auditory location, with minimal feedback and occasional cue conflict. Integration strategies varied with age and self-reported ADHD or Autism. Visual cues were combined near-optimally, whereas audio-visual combinations exhibited winner-take-all behaviour, typically but not always favouring the more reliable cue. To test the generality of these findings, we used electrical microstimulation in non-human primates, targeting unimodal or cross-modal association areas. Stimulation of visual cortex was integrated with sensory motion cues, while stimulation of prefrontal cortex promoted winner-take-all choices. These findings suggest universal circuit-level distinctions between within- and across-modality integration, with deviations potentially diagnostic for neuropsychiatric conditions.

A brain-wide, trial- and time-dependent deterministic drive for self-initiated action decisions
Deciding when to act in the absence of external cues is essential for exploration, learning and survival. Yet how the brain makes such decisions remains controversial, with current models favoring either deterministic or stochastic underpinnings. We performed large-scale, single-unit recordings across eight brain areas in mice self-initiating voluntary actions. We found that action timing is predictable from neural activity, up to several seconds before action initiation, across all eight brain areas. This predictive activity reflects a deterministic drive whose initial value and rate vary trial-by-trial, and whose rate accelerates within trials. While noise contributes to timing variability, the deterministic drive is sufficient to trigger action. Thus, self-initiated timing decisions arise from a variable, brain-wide deterministic drive, challenging models where noise is the primary driver.