bioRxiv Subject Collection: Neuroscience's Journal

PINK1 knockout rats show premotor cognitive deficits measured through a complex maze
Cognitive decline in Parkinsons disease (PD) emerges up to 10 years before clinical recognition. Neurobiological mechanisms underlying premotor cognitive impairment in PD can potentially be examined in the PINK1 / rat, which exhibits a protracted motor onset. To enhance translation to human PD cognitive assessments, we tested a modified multiple T-maze, which measures cognitive flexibility similarly to the Trail-Making Test in humans. Like human PD outcomes, PINK1-/- rats made more errors and took longer to complete the maze than wild types. Thus, we have identified a potential tool for assessing cross-species translation of cognitive functioning in an established PD animal model.

Reduced recruitment of inhibitory control regions in very young children with ADHD during a modified Kiddie Continuous Performance Task: a fMRI study.
Attention-Deficit/Hyperactivity Disorder (ADHD) symptom profiles are known to undergo changes throughout development, rendering the neurobiological assessment of ADHD challenging across different developmental stages. Particularly in young children (ages 4 to 7 years), measuring inhibitory control network activity in the brain has been a formidable task due to the lack of child-friendly functional Magnetic Resonance Imaging (fMRI) paradigms. This study aims to address these difficulties by focusing on measuring inhibitory control in very young children within the MRI environment. A total of 56 children diagnosed with ADHD and 78 typically developing (TD) 4-7-year-old children were examined using a modified version of the Kiddie-Continuous Performance Test (K-CPT) during BOLD fMRI to assess inhibitory control. We concurrently evaluated their performance on the established and standardized K-CPT outside the MRI scanner. Our findings suggest that the modified K-CPT effectively elicited robust and expected brain activity related to inhibitory control in both groups. Comparisons between the two groups revealed subtle differences in brain activity, primarily observed in regions associated with inhibitory control, such as the inferior frontal gyrus, anterior insula, dorsal striatum, medial pre-supplementary motor area (pre-SMA), and cingulate cortex. Notably, increased activity in the right anterior insula was associated with improved response time (RT) and reduced RT variability on the K-CPT administered outside the MRI environment, although this did not survive statistical correction for multiple comparisons. In conclusion, our study successfully overcame the challenges of measuring inhibitory control in very young children within the MRI environment by utilizing a modified K-CPT during BOLD fMRI. These findings shed light on the neurobiological correlates of inhibitory control in ADHD and TD children, provide valuable insights for understanding ADHD across development, and potentially inform ADHD diagnosis and intervention strategies. The research also highlights remaining challenges with task fMRI in very young clinical samples.

Emergence of neocortical function in heterotopic neurons
Brains come in various sizes and shapes, yet how neuronal position constrains the type of circuits that they can form remains largely unknown. The spatial layout of anatomical structures with corresponding functions varies widely across species. Also, during evolution, anatomical structures have duplicated and then diverged to generate new circuits and functions. Thus, it is critical to understand how the position of neurons constrains their integration into circuits and, ultimately, their function. To address this question, we studied Eml1 knockout mice in which subsets of neocortical neurons form a new structure below the neocortex termed heterotopia (Ht). We examined how this new location affects the molecular identity, topography, input-output circuit connectivity, electrophysiology, and functional properties of these neurons. Our results reveal a striking conservation of the cellular features and circuit properties of Ht neurons, despite their abnormal location and misorientation. Supporting this observation, these neurons were able to functionally substitute for overlying necortical neurons in a behaviorally relevant task when the latter were optogenetically silenced. Hence, specific neuronal identities and associated function can be reproduced in altered anatomical settings, revealing a remarkable level of self-organization and adaptability of neocortical circuits.

A Cell Model for the Detection of a Planar Surface Using Motion Stereo in Area MST of the Visual Cortex
I propose a series of modeled cells for detecting three plane parameters (i.e. a time-to-contact to a plane, its orientation, and its shortest distance) with motion stereo. This series is composed of lateral geniculate nucleus cells, nondirectionally selective simple cells, directionally selective (DS) simple cells, DS complex cells, motion-detection cells, and planar-surface-detection cells. These cell types perform the time delay, Hough transform, spatio-temporal correlation, accumulation, inverse Hough transform, and a combinations of the cross-ratio and polar transforms (or a small-circle transform) to detect these plane parameters, respectively. Each connection of the neural network connecting this series of cell types is modeled mathematically one by one using these transforms. The selective responses of SDCs to planar parameters are consistent with those of physiological experiments.

Gut Analysis Toolbox: Automating quantitative analysis of enteric neurons
The enteric nervous system (ENS) plays an important role in coordinating gut function. The ENS consists of an extensive network of neurons and glial cells within the wall of the gastrointestinal tract. Alterations in neuronal distribution, function, and type are strongly associated with enteric neuropathies and gastrointestinal (GI) dysfunction and can serve as biomarkers for disease. However, current methods for assessing neuronal counts and distribution suffer from undersampling. This is partly due to challenges associated with imaging and analyzing large tissue areas, and operator bias due to manual analysis. Here, we present the Gut Analysis Toolbox (GAT), an image analysis tool designed for characterization of enteric neurons and their neurochemical coding using 2D images of GI wholemount preparations. GAT is developed for the Fiji distribution of ImageJ. It has a user-friendly interface and offers rapid and accurate cell segmentation. Custom deep learning (DL) based cell segmentation models were developed using StarDist. GAT also includes a ganglion segmentation model which was developed using deepImageJ. In addition, GAT allows importing of segmentation generated by other software. DL models have been trained using ZeroCostDL4Mic on diverse datasets sourced from different laboratories. This captures the variability associated with differences in animal species, image acquisition parameters, and sample preparation across research groups. We demonstrate the robustness of the cell segmentation DL models by comparing them against the state-of-the-art cell segmentation software, Cellpose. To quantify neuronal distribution GAT applies proximal neighbor-based spatial analysis. We demonstrate how the proximal neighbor analysis can reveal differences in cellular distribution across gut regions using a published dataset. In summary, GAT provides an easy-to-use toolbox to streamline routine image analysis tasks in ENS research. GAT enhances throughput allowing unbiased analysis of larger tissue areas, multiple neuronal markers and numerous samples rapidly.

The type of inhibition provided by thalamic interneurons alters the input selectivity of thalamocortical neurons.
A fundamental problem in neuroscience is how neurons select for their many inputs. A common assumption is that a neuron's selectivity is largely explained by differences in excitatory synaptic input weightings. Here we describe another solution to this important problem. We show that within the first order visual thalamus, the type of inhibition provided by thalamic interneurons has the potential to alter the input selectivity of thalamocortical neurons. To do this, we developed conductance injection protocols to compare how different types of synchronous and asynchronous GABA release influence thalamocortical excitability in response to realistic patterns of retinal ganglion cell input. We show that the asynchronous GABA release associated with tonic inhibition is particularly efficient at maintaining information content, ensuring that thalamocortical neurons can distinguish between their inputs. We propose a model where alterations in GABA release properties results in rapid changes in input selectivity without requiring structural changes in the network.

Single and Multi-Site Cortical Stimulation Related to Human Sensorimotor Function
Somatosensory feedback is crucial for precise control of our body and thereby affects various sensorimotor-related brain areas for movement control. Electrical stimulation on the primary somatosensory cortex (S1) elicits various artificial somatosensations. However, replicating the spatiotemporal dynamics of somatosensory feedback and fine control of elicited somatosensation are still challenging. Furthermore, how and where the somatosensory feedback interacts with neural activity for sensorimotor processing is unclear. Here, we replicate the spatiotemporal dynamics of somatosensory feedback and control the quality of elicited somatosensation using multi-site direct cortical stimulation (DCS). We also investigate how and where the neural feedback activity interacts with neural activity for motor processing by stimulating the downstream areas of the S1. We found that multi-site DCS on the S1 elicits different sensations simultaneously. Using the artificial feedback, blindfolded patients could efficiently perform a DCS-guided reach-and-grasp task successfully. Interestingly, we also found that multi-site DCS close to each other elicits different qualities of somatosensation in the same body part. Additionally, we found that DCS on the ventral premotor area (vPM) can affect hand grasping with eliciting artificial sensation of the hand. Throughout this study, we showed that semi-invasive, macro-level, and multi-site DCS can precisely elicit/modulate somatosensations in human. We suggest that activation of multiple cortical areas elicits simultaneous and independent somatosensations and that interplay among the stimulated sites can change the somatosensation quality. Finally, the results of vPM stimulation indicate that vPM has a critical role in function-specific sensorimotor interactions, such as hand grasping.

Sensory choices as logistic classification
Logistic classification is a simple way to make choices based on a set of factors: give each factor a weight, sum them, and use the result to set the log odds of a random draw. This operation is known to describe human and animal choices based on value (economic decisions). There is increasing evidence that it also describes choices based on sensory inputs (perceptual decisions). Here I briefly review this evidence and I fit data from multiple studies in multiple species to show that logistic classification can describe a variety of choices. These include sensory choices influenced by stimuli of other modalities (multisensory integration) or by non-sensory factors such as value and recent history. Logistic classification is the optimal strategy if factors are independent of each other, and a useful heuristic in other conditions. Using it to describe sensory choices is useful to characterize brain function and the effects of brain inactivations.

Energy Optimization Induces Predictive-coding Properties in a Multicompartment Spiking Neural Network Model
Predictive coding is a prominent theoretical framework for understanding the hierarchical sensory processing in the brain, yet how it could be implemented in networks of cortical neurons is still unclear. While most existing works have taken a hand-wiring approach to creating microcircuits that match experimental results, recent work in applying an optimisation approach to rate-based artificial neural networks revealed that cortical connectivity might result from self-organisation given some fundamental computational principle, such as energy efficiency. As no corresponding approach has studied this in more plausible networks of spiking neurons, we here investigate whether predictive coding properties in a multi-compartment spiking neural network can emerge from energy optimisation. We find that a model trained with an energy objective in addition to a task-relevant objective is able to reconstruct internal representations given top-down expectation signals alone. Additionally, neurons in the energy-optimised model also show differential responses to expected versus unexpected stimuli, qualitatively similar to experimental evidence for predictive coding. These findings indicate that predictive-coding-like behaviour might be an emergent property of energy optimisation, providing a new perspective on how predictive coding could be achieved in the cortex.

Spatially targeted inhibitory rhythms differentially affect neuronal integration
Pyramidal neurons form dense recurrently connected networks with multiple types of inhibitory interneurons. A major differentiator between interneuron subtypes is whether they synapse onto perisomatic or dendritic regions. They can also engender local inhibitory rhythms, beta (12-35 Hz) and gamma (40-80 Hz). The interaction between the rhythmicity of inhibition and its spatial targeting on the neuron may determine how it regulates neuronal integration. Thus, we sought to understand how rhythmic perisomatic and distal dendritic inhibition impacted integration in a layer 5 pyramidal neuron model with elaborate dendrites and Na+, NMDA, and Ca2+ dendritic spikes. We found that inhibition regulated the coupling between dendritic spikes and action potentials in a location and rhythm-dependent manner. Perisomatic inhibition principally regulated action potential generation, while distal dendritic inhibition regulated the incidence of dendritic spikes and their temporal coupling with action potentials. Perisomatic inhibition was most effective when provided at gamma frequencies, while distal dendritic inhibition functioned best at beta. Moreover, beta modulated responsiveness to apical inputs in a phase-dependent manner, while gamma did so for basal inputs. These results may provide a functional interpretation for the reported association of soma-targeting parvalbumin positive interneurons with gamma, and dendrite-targeting somatostatin interneurons with beta.

Physiological JNK3 Concentrations Are Higher in Motor-related and Disease-implicated Brain Regions of C57BL6/J Mice
The c-Jun N-terminal kinase 3 (JNK3) is a stress-responsive protein kinase primarily expressed in the central nervous system (CNS). JNK3 exhibits nuanced neurological activities, such as roles in behavior, circadian rhythms, and neurotransmission, but JNK3 is also implicated in cell death and neurodegeneration. Despite the critical role of JNK3 in neurophysiology and pathology, its localization in the brain is not fully understood due to a paucity of tools to distinguish JNK3 from other isoforms. While previous functional and histological studies suggest locales for JNK3 in the CNS, a comprehensive and higher resolution of JNK3 distribution and abundance remained elusive. Here, we sought to define the anatomical and cellular distribution of JNK3 in adult mouse brains. Data reveal the highest levels of JNK3 and pJNK3 were found in the cortex and the hippocampus. JNK3 possessed neuron-type selectivity as JNK3 was present in GABAergic, cholinergic, and dopaminergic neurons, but was not detectable in VGLUT-1-positive glutamatergic neurons and astrocytes in vivo. Intriguingly, higher JNK3 signals were found in motor neurons and relevant nuclei in the cortex, basal ganglia, brainstem, and spinal cord. While JNK3 was primarily observed in the cytosol of neurons in the cortex and the hippocampus, JNK3 appeared commonly within the nucleus in the brainstem. These distinctions suggest the potential for significant differences between JNK3 actions in distinct brain regions and cell types. Our results provide a significant improvement over previous reports of JNK3 spatial organization in the adult CNS and support continued investigation of JNK3 actions in neurophysiology and pathophysiology.

Altered physiological, affective, and functional connectivity responses to acute stress in patients with alcohol use disorder
Background There is evidence that the processing of acute stress is altered in alcohol use disorder (AUD), but little is known about how this is manifested simultaneously across different stress parameters and which neural processes are involved. The present study examined physiological and affective responses to stress and functional connectivity in AUD. Methods Salivary cortisol samples, pulse rate, and affect ratings were collected on two days from 34 individuals with moderate or severe AUD and 34 controls. On one day, stress was induced, and on the other day, a non-stressful control task was performed. Following the intervention, participants underwent fMRI to assess functional connectivity, focusing on cortical and subcortical seed regions previously reported to be involved in AUD and/or stress. Results For pulse rate and cortisol, stress responses were blunted in AUD, whereas negative affect was increased. Furthermore, stress-related changes in pulse rate, cortisol, and affect were only correlated in healthy controls. Neuroimaging analyses revealed stress-related group differences in functional connectivity, involving the connectivity of striatal seeds with the posterior DMN, cerebellum and midcingulate cortex, and of the posterior DMN seed with the striatum and thalamus. Conclusions The results suggest a dissociation between subjective experienced distress and the physiological stress response in AUD as well as stress-related alterations in functional connectivity. These findings highlight the complex interplay between chronic alcohol use and acute stress regulation, offering valuable considerations for the development of therapeutic strategies.

DeepResBat: deep residual batch harmonization accounting for covariate distribution differences
Pooling MRI data from multiple datasets requires harmonization to reduce undesired inter- site variabilities, while preserving effects of biological variables (or covariates). The popular harmonization approach ComBat uses a mixed effect regression framework that explicitly accounts for covariate distribution differences across datasets. There is also significant interest in developing harmonization approaches based on deep neural networks (DNNs), such as conditional variational autoencoder (cVAE). However, current DNN approaches do not explicitly account for covariate distribution differences across datasets. Here, we provide mathematical results, suggesting that not accounting for covariates can lead to suboptimal harmonization outcomes. We propose two DNN-based harmonization approaches that explicitly account for covariate distribution differences across datasets: covariate VAE (coVAE) and DeepResBat. The coVAE approach is a natural extension of cVAE by concatenating covariates and site information with site- and covariate-invariant latent representations. DeepResBat adopts a residual framework inspired by ComBat. DeepResBat first removes the effects of covariates with nonlinear regression trees, followed by eliminating site differences with cVAE. Finally, covariate effects are added back to the harmonized residuals. Using three datasets from three different continents with a total of 2787 participants and 10085 anatomical T1 scans, we find that DeepResBat and coVAE outperformed ComBat, CovBat and cVAE in terms of removing dataset differences, while enhancing biological effects of interest. However, coVAE hallucinates spurious associations between anatomical MRI and covariates even when no association exists. Therefore, future studies proposing DNN-based harmonization approaches should be aware of this false positive pitfall. Overall, our results suggest that DeepResBat is an effective deep learning alternative to ComBat.

Psilocybin biphasically modulates cortical and behavioral activity in mice
Psilocybin is a serotonergic psychedelic believed to have therapeutic potential for neuropsychiatric conditions. Despite well-documented prevalence of perceptual alterations, hallucinations, and synesthesia associated with psychedelic experiences, little is known about how psilocybin affects sensory cortex or alters the activity of neurons in awake animals. To investigate, we conducted 2-photon imaging experiments in auditory cortex of awake mice and video analysis of mouse behavior, both at baseline and during psilocybin treatment. We found biphasic effects of psilocybin on behavioral and cortical activity. A 2 mg/kg dose of psilocybin initially increased behavioral activity and neural responses to sound. 30 minutes post-dose, mice became behaviorally hypoactive and cortical responses to sound decreased, while neural response variance and noise correlations increased. In contrast, neuronal selectivity for auditory stimuli remained stable during psilocybin treatment. Our results suggest that psilocybin modulates the role of intrinsic versus stimulus-driven activity in sensory cortex, while preserving fundamental sensory processing.

Intranasally Administered EVs from hiPSC-derived NSCs Alter the Transcriptomic Profile of Activated Microglia and Conserve Brain Function in an Alzheimer's Model
Antiinflammatory extracellular vesicles (EVs) derived from human induced pluripotent stem cell (hiPSC)-derived neural stem cells (NSCs) hold promise as a disease-modifying biologic for Alzheimer's disease (AD). This study directly addressed this issue by examining the effects of intranasal administrations of hiPSC-NSC-EVs to 3-month-old 5xFAD mice. The EVs were internalized by all microglia, which led to reduced expression of multiple genes associated with disease-associated microglia, inflammasome, and interferon-1 signaling. Furthermore, the effects of hiPSC-NSC-EVs persisted for two months post-treatment in the hippocampus, evident from reduced microglial clusters, inflammasome complexes, and expression of proteins and/or genes linked to the activation of inflammasomes, p38/mitogen-activated protein kinase, and interferon-1 signaling. The amyloid-beta (A{beta}) plaques, A{beta}-42, and phosphorylated-tau concentrations were also diminished, leading to better cognitive and mood function in 5xFAD mice. Thus, early intervention with hiPSC-NSC-EVs in AD may help maintain better brain function by restraining the progression of adverse neuroinflammatory signaling cascades.

ER and SOCE Ca2+ signals are not required for directed cell migration in human microglia
The central nervous system (CNS) is constantly surveilled by microglia, highly motile and dynamic cells deputed to act as the first line of immune defense in the brain and spinal cord. Alterations in the homeostasis of the CNS are detected by microglia that respond by migrating toward the affected area. Understanding the mechanisms controlling directed cell migration of microglia is crucial to dissect their responses to neuroinflammation and injury. We used a combination of pharmacological and genetic approaches to explore the involvement of calcium (Ca2+) signaling in the directed migration of induced pluripotent stem cell (iPSC)-derived microglia challenged with a purinergic stimulus. This approach mimics cues originating from injury of the CNS. Unexpectedly, simultaneous imaging of microglia migration and intracellular Ca2+ changes revealed that this phenomenon does not require Ca2+ signals generated from the endoplasmic reticulum (ER) and store-operated Ca2+ entry (SOCE) pathways. Instead, we find evidence that human microglial chemotaxis to purinergic signals is mediated by cyclic AMP in a Ca2+-independent manner. These results challenge prevailing notions, with important implications in neurological conditions characterized by perturbation in Ca2+ homeostasis.

Differential Short-Term Facilitation Of Synaptic Inputs And Spike Transmission At The Retinocollicular Synapse In Vivo
Short-term plasticity (STP) is important for understanding how neuronal circuits can perform different computations. The STP of a neuron pair can be measured directly using paired whole-cell recordings. Besides, the cross-correlation between the presynaptic and postsynaptic neuronal firing is usually used as a proxy for estimating the synaptic properties. However, the relationships between the synaptic inputs and the spiking properties of the postsynaptic neurons during the STP in vivo still remain unclear. Here, we characterized the STP of both synaptic input, measured by the postsynaptic field potential (PFP), and spike transmission at the retinocollicular pathway of mice. We found that the STP of the retinocollicular pathway is mainly facilitating, where the second presynaptic spike induces a larger PFP and higher postsynaptic firing rate than the first presynaptic spike. The facilitation in the postsynaptic firing rate is generally larger than the PFP facilitation. Interestingly, the last postsynaptic spike timing also has a large facilitating effect on the postsynaptic spiking upon receiving a presynaptic input spike. However, the PFP does not depend on the last postsynaptic spike timing, suggesting that there is an input-independent component of spike transmission in STP. Overall, our results indicate that the STP of the retinocollicular pathway is likely a two-stage process, where the spiking plasticity of the postsynaptic neuron could be independent of its inputs.

Differential and temporally dynamic involvement of primate amygdala nuclei in face animacy and reward information processing
Decision-making is influenced by both expected reward and social factors, such as who offered the outcomes. Thus, although a reward might originally be independent from social factors, the two elements are closely related. However, whether and how they are processed separately or conjointly remains unclear. Here, we show that neurons in distinct sub-nuclei of the amygdala encode expected reward and face animacy, which is a vital aspect of face perception. Although these encoding processes are distinct, they rely on partially shared neuronal circuits with characteristic temporal dynamics. Two male macaque monkeys made saccades under different social and reward contexts, created by presenting facial images with independent attributes: animacy (a monkey or cartoon face) and associated reward (large or small). The stimulus image was presented twice per trial: during the initial stimulus encoding (S1) and before saccades were made (S2). A longer gaze duration for eye region of the monkey versus cartoon images indicated more robust social engagement for realistic faces. During S1, a similar number of lateral nucleus neurons encoded either animacy only with a monkey-image preference, reward only with a large-reward preference, or both. Conversely, neurons in the basal and central nuclei primarily encoded reward, preferring large- versus small-reward associated face images. The reward-dependent modulation was continuous after S1, but was more conspicuous during S1 in the basal nucleus and during both S1 and S2 in the central nucleus. This anatomically- and temporally-specific encoding in the amygdala may underlie the computation and integration of face animacy and reward information.

It's all in the timing: Delayed feedback in autism may weaken predictive mechanisms during contour integration
Humans rely on predictive mechanisms during visual processing to efficiently resolve incomplete or ambiguous sensory signals. While initial low-level sensory data are conveyed by feedforward connections, feedback connections are believed to shape sensory processing through conveyance of statistical predictions based on prior exposure to stimulus configurations. Individuals with autism spectrum disorder (ASD) show biases in stimulus processing toward parts rather than wholes, suggesting their sensory processing may be less shaped by statistical predictions acquired through prior exposure to global stimulus properties. Investigations of illusory contour (IC) processing in neurotypical (NT) adults have established a well-tested marker of contour integration characterized by a robust modulation of the visually evoked potential (VEP), the IC-effect, that occurs over lateral occipital scalp during the timeframe of the N1 component. Converging evidence strongly supports the notion that this IC-effect indexes a signal with significant feedback contributions. Using high-density VEPs, we compared the IC-effect in 6-17 year-old children with ASD (n=32) or NT development (n=53). Both groups of children generated an IC-effect that was equivalent in amplitude. However, the IC-effect notably onset 21ms later in ASD, even though timing of initial VEP afference was identical across groups. This suggests that feedforward information predominated during perceptual processing for 15% longer in ASD compared to NT children. This delay in the feedback dependent IC-effect, in the context of known developmental differences between feedforward and feedback fibers, suggests a potential pathophysiological mechanism of visual processing in ASD, whereby ongoing stimulus processing is less shaped by statistical prediction mechanisms.

Cholinergic neuromodulation of prefrontal attractor dynamics controls performance in spatial working memory
The behavioral and neural effects of the endogenous release of acetylcholine following stimulation of the Nucleus Basalis of Meynert (NB) have been recently examined (Qi et al. 2021). Counterintuitively, NB stimulation enhanced behavioral performance while broadening neural tuning in the prefrontal cortex (PFC). The mechanism by which a weaker mnemonic neural code could lead to better performance remains unclear. Here, we show that increased neural excitability in a simple continuous bump attractor model can induce broader neural tuning and decrease bump diffusion, provided neural rates are saturated. Increased memory precision in the model overrides memory accuracy, improving overall task performance. Moreover, we show that bump attractor dynamics can account for the nonuniform impact of neuromodulation on distractibility, depending on distractor distance from the target. Finally, we delve into the conditions under which bump attractor tuning and diffusion balance in biologically plausible heterogeneous network models. In these discrete bump attractor networks, we show that reducing spatial correlations or enhancing excitatory transmission can improve memory precision. Altogether, we provide a mechanistic understanding of how cholinergic neuromodulation controls spatial working memory through perturbed attractor dynamics in PFC.

Pharmacological manipulations of the dorsomedial and dorsolateral striatum during fear extinction have opposing effects on fear renewal
Systemic manipulations that enhance dopamine (DA) transmission around the time of fear extinction can strengthen fear extinction and reduce conditioned fear relapse. Prior studies investigating the brain regions where DA augments fear extinction focus on targets of mesolimbic and mesocortical DA systems originating in the ventral tegmental area, given the role of these DA neurons in prediction error. The dorsal striatum (DS), a primary target of the nigrostriatal DA system originating in the substantia nigra (SN), is implicated in behaviors beyond its canonical role in movement, such as reward and punishment, goal-directed action, and stimulus-response associations, but whether DS DA contributes to fear extinction is unknown. We have observed that chemogenetic stimulation of SN DA neurons during fear extinction prevents the return of fear in contexts different from the extinction context, a form of relapse called renewal. This effect of SN DA stimulation is mimicked by a DA D1 receptor (D1R) agonist injected into the DS, thus implicating DS DA in fear extinction. Different DS subregions subserve unique functions of the DS, but it is unclear where in the DS D1R agonist acts during fear extinction to reduce renewal. Furthermore, although fear extinction increases neural activity in DS subregions, whether neural activity in DS subregions is causally involved in fear extinction is unknown. To explore the role of DS subregions in fear extinction, adult, male Long-Evans rats received microinjections of either the D1R agonist SKF38393 or a cocktail consisting of GABAA/GABAB receptor agonists muscimol/baclofen selectively into either dorsomedial (DMS) or dorsolateral (DLS) DS subregions immediately prior to fear extinction, and extinction retention and renewal were subsequently assessed drug-free. While increasing D1R signaling in the DMS during fear extinction did not impact fear extinction retention or renewal, DMS inactivation reduced later renewal. In contrast, DLS inactivation had no effect on fear extinction retention or renewal but increasing D1R signaling in the DLS during extinction reduced fear renewal. These data suggest that DMS and DLS activity during fear extinction can have opposing effects on later fear renewal, with the DMS promoting renewal and the DLS opposing renewal. Mechanisms through which the DS could influence the contextual gating of fear extinction are discussed.

Deciphering the Rhythmic Symphony of Speech: A Neural Framework for Robust and Time-Invariant Speech Comprehension
Unraveling the mysteries of how humans effortlessly grasp speech amidst diverse environmental challenges has long intrigued researchers in systems and cognitive neuroscience. This study delves into the neural intricacies underpinning robust speech comprehension, giving a computational mechanistic proof for the hypothesis proposing a pivotal role for rhythmic, predictive top-down contextualization facilitated by the delta rhythm in achieving time-invariant speech processing. We propose a Brain-Rhythm-Based Inference (BRyBI) model that integrates three key rhythmic processes - theta-gamma interactions for parsing phoneme sequences, dynamic delta rhythm for inferred prosodic-phrase context, and resilient speech representations. Demonstrating mechanistic proof-of-principle, BRyBI replicates human behavioral experiments, showcasing its ability to handle pitch variations, time-warped speech, interruptions, and silences in non-comprehensible contexts. Intriguingly, the model aligns with human experiments, revealing optimal silence time scales in the theta- and delta-frequency ranges. Comparative analysis with deep neural network language models highlights distinctive performance patterns, emphasizing the unique capabilities of our rhythmic framework. In essence, our study sheds light on the neural underpinnings of speech processing, emphasizing the role of rhythmic brain mechanisms in structured temporal signal processing - an insight that challenges prevailing artificial intelligence paradigms and hints at potential advancements in compact and robust computing architectures.

Transfer functions for burst firing probability in a model neocortical pyramidal cell
Neocortical layer 5 thick-tufted pyramidal cells are prone to exhibiting burst firing on receipt of coincident basal and apical dendritic inputs. These inputs carry different information, with basal inputs coming from feedforward sensory pathways and apical inputs coming from diverse sources that provide context in the cortical hierarchy. We explore the information processing possibilities of this burst firing using computer simulations of a noisy compartmental cell model. Simulated data on stochastic burst firing due to brief, simultaneously injected basal and apical currents allows estimation of burst firing probability for different stimulus current amplitudes. Information-theory-based partial information decomposition (PID) is used to quantify the contributions of the apical and basal input streams to the information in the cell output bursting probability. Different operating regimes are apparent, depending on the relative strengths of the input streams, with output burst probability carrying more or less information that is uniquely contributed by either the basal or apical input, or shared and synergistic information due to the combined streams. We derive and fit transfer functions for these different regimes that describe burst probability over the different ranges of basal and apical input amplitudes. The operating regimes can be classified into distinct modes of information processing, depending on the contribution of apical input to output bursting: apical cooperation, in which both basal and apical inputs are required to generate a burst; apical amplification, in which basal input alone can generate a burst but the burst probability is modulated by apical input; apical drive, in which apical input alone can produce a burst; and apical integration, in which strong apical or basal inputs alone, as well as their combination, can generate bursting. In particular, PID and the transfer function clarify that the apical amplification mode has the features required for contextually-modulated information processing.

Modular Phoneme Processing in Human Superior Temporal Gyrus
Modular organization is fundamental to cortical processing, but its presence is human association cortex is unknown. We characterized phoneme processing with 128-1024 channel micro-arrays at 50-200m pitch on superior temporal gyrus of 7 patients. High gamma responses were highly correlated within [~]1.7mm diameter modules, sharply delineated from adjacent modules with distinct time-courses and phoneme-selectivity. We suggest that receptive language cortex may be organized in discrete processing modules.

Efficient 3D cone trajectory design for improved combined angiographic and perfusion imaging using arterial spin labeling
PurposeTo improve the spatial resolution and repeatability of a non-contrast MRI technique for simultaneous time resolved 3D angiography and perfusion imaging by developing an efficient 3D cone trajectory design.

MethodsA novel parameterized 3D cone trajectory design incorporating the 3D Golden Angle was integrated into 4D combined angiography and perfusion using radial imaging and arterial spin labeling (CAPRIA) to achieve higher spatial resolution and sampling efficiency for both dynamic angiography and perfusion imaging with flexible spatiotemporal resolution. Numerical simulations and physical phantom scanning were used to optimize the cone design. Eight healthy volunteers were scanned to compare the original radial trajectory in 4D CAPRIA with our newly designed cone trajectory. A locally low rank reconstruction method was used to leverage the complementary k-space sampling across time.

ResultsThe improved sampling in the periphery of k-space obtained with the optimized 3D cone trajectory resulted in improved spatial resolution compared with the radial trajectory in phantom scans. Improved vessel sharpness and perfusion visualization were also achieved in vivo. Less dephasing was observed in the angiograms due to the short echo time of our cone trajectory and the improved k-space sampling efficiency also resulted in higher repeatability compared to the original radial approach.

ConclusionThe proposed 3D cone trajectory combined with 3D Golden Angle ordering resulted in improved spatial resolution and image quality for both angiography and perfusion imaging and could potentially benefit other applications that require an efficient sampling scheme with flexible spatial and temporal resolution.

A general theoretical framework unifying the adaptive, transient and sustained properties of ON and OFF auditory neural responses
Sounds are temporal stimuli decomposed into numerous elementary components by the auditory nervous system. For instance, a temporal to spectro-temporal transformation modelling the frequency decomposition performed by the cochlea is a widely adopted first processing step in today's computational models of auditory neural responses. Similarly, increments and decrements in sound intensity (i.e., of the raw waveform itself or of its spectral bands) constitute critical features of the neural code, with high behavioural significance. However, despite the growing attention of the scientific community on auditory OFF responses, their relationship with transient ON, sustained responses and adaptation remains unclear. In this context, we propose a new general model, based on a pair of linear filters, named "AdapTrans" that captures both sustained and transient ON and OFF responses into a unifying and easy to expand framework. We demonstrate that filtering audio cochleagrams with AdapTrans permits to accurately render known properties of neural responses measured in different mammal species such as the dependence of OFF responses on the stimulus fall time and on the preceding sound duration. Furthermore, by integrating our framework into gold standard and state-of-the-art machine learning models that predict neural responses from audio stimuli, following a supervised training on a large compilation of electrophysiology datasets (ready-to-deploy PyTorch models and pre-processed datasets shared publicly), we show that AdapTrans systematically improves the prediction accuracy of estimated responses within different cortical areas of the rat and ferret auditory brain. Together, these results motivate the use of our framework for computational and systems neuroscientists willing to increase the plausibility and performances of their models of audition.

Auditory Forward Masking Explained by a Subcortical Model with Efferent Control of Cochlear Gain
Previous physiological and psychophysical studies have explored whether feedback to the cochlea from the efferent system influences forward masking. The present work proposes that the limited growth-of-masking (GOM) observed in auditory-nerve (AN) fibers may have been misunderstood; namely, that this limitation may be due to the influence of anesthesia on the efferent system. Building on the premise that the unanesthetized AN may exhibit GOM similar to more central nuclei, the present computational modeling study demonstrates that feedback from the medial olivocochlear (MOC) efferents may account for GOM observed physiologically in onset-type neurons in both the cochlear nucleus and inferior colliculus (IC), as well as in human psychophysics. Additionally, the computational model of MOC efferents used here generates a decrease in masking with longer masker-signal delays similar to that observed in IC physiology and in psychophysical studies. An advantage of this explanation over alternative physiological explanations (e.g., that forward masking requires inhibition from the superior paraolivary nucleus) is that this theory can explain forward masking observed in the brainstem, early in the ascending pathway. For explaining psychoacoustic results, one strength of this model is that it can account for the lack of elevation in thresholds observed when masker level is randomly varied from interval-to-interval, a result that is difficult to explain using the conventional temporal-window model of psychophysical forward masking.

Novel FKBP12 ligand promotes functional improvement in SOD1-G93A ALS mice
Amyotrophic lateral sclerosis (ALS) is a devastating neurodegenerative disease with limited treatment options. ALS pathogenesis involves intricate processes within motor neurons (MNs), characterized by dysregulated Ca2+ influx and buffering in early ALS-affected MNs. This study proposes the modulation of ryanodine receptors (RyRs), key mediators of intracellular Ca2+, as a therapeutic target. A novel class of novel FKBP12 ligands that show activity as cytosolic calcium modulators through stabilizing RyR channel activity, were tested in the SOD1-G93A mouse model of ALS. Different outcomes were used to assess treatment efficacy including electrophysiology, histopathology, neuromuscular function, and survival. Among the novel FKBP12 ligands, MP-010 was chosen for its central nervous system availability. Chronic administration of MP-010 to SOD1-G93A mice produced a dose-dependent preservation of motor nerve conduction, with the 61 mg/kg dose significantly delaying the onset of motor impairment. This was accompanied by improved motor coordination, increased innervated endplates, and significant preservation of MNs in the spinal cord of treated mice. Notably, MP-010 treatment significantly extended lifespan by an average of 10 days compared to vehicle. In conclusion, FKBP12 ligands, particularly MP-010, exhibit promising neuroprotective effects in ALS, highlighting their potential as novel therapeutic agents. Further investigations into the molecular mechanisms and clinical translatability of these compounds are needed for their application in ALS treatment.

PYK2 in the dorsal striatum of Huntington disease R6/2 mouse model
Huntington disease (HD) is a devastating disease due to autosomal dominant mutation in the HTT gene. Its pathophysiology involves multiple molecular alterations including transcriptional defects. We previously showed that in HD patients and mouse model, the protein levels of the non-receptor tyrosine kinase PYK2 were decreased in the hippocampus and that viral expression of PYK2 improved the hippocampal phenotype. Here, we investigated the possible contribution of PYK2 in the striatum, a major brain region altered in HD. PYK2 mRNA levels were decreased in the striatum and hippocampus of R6/2 mice, a severe HD model. PYK2 protein levels were also decreased in the dorsal striatum of R6/2 mice and in the putamen of human patients. PYK2 knockout by itself did not result in motor symptoms observed in HD mouse models. Yet, we examined whether PYK2 deficiency participated in the R6/2 mice phenotype by expressing PYK2 in the dorsal striatum using AAV vectors. With an AAV1/Camk2a promoter, we did not observe significant improvement of body weight, clasping, motor activity and coordination (rotarod) alterations observed in R6/2 mice. With an AAV9/SYN1 promoter we found an improvement of body weight loss and a tendency to better rotarod performance. DARPP-32 immunofluorescence was increased following AAV9/SYN1-PYK2 injection compared to AAV9/SYN1-GFP, suggesting a possible partial beneficial effect on striatal projection neurons. We conclude that PYK2 mRNA and protein levels are decreased in the striatum as in hippocampus of HD patients and mouse models. However, in contrast to hippocampus, striatal viral expression of PYK2 has only a slight effect on the R6/2 model striatal motor phenotype.

The Calcitron: A Simple Neuron Model That Implements Many Learning Rules via the Calcium Control Hypothesis
Theoretical neuroscientists and machine learning researchers have proposed a variety of learning rules for linear neuron models to enable artificial neural networks to accomplish supervised and unsupervised learning tasks. It has not been clear, however, how these theoretically-derived rules relate to biological mechanisms of plasticity that exist in the brain, or how the brain might mechanistically implement different learning rules in different contexts and brain regions. Here, we show that the calcium control hypothesis, which relates plastic synaptic changes in the brain to calcium concentration [Ca2+] in dendritic spines, can reproduce a wide variety of learning rules, including some novel rules. We propose a simple, perceptron-like neuron model that has four sources of [Ca2+]: local (following the activation of an excitatory synapse and confined to that synapse), heterosynaptic (due to activity of adjacent synapses), postsynaptic spike-dependent, and supervisor-dependent. By specifying the plasticity thresholds and amount of calcium derived from each source, it is possible to implement Hebbian and anti-Hebbian rules, one-shot learning, perceptron learning, as well as a variety of novel learning rules.

Dynamic connectivity profiles characteristic of conscious states are associated with enhanced conscious processing of external stimuli
One of the goals of the neuroscience of consciousness is to identify neural markers capable of distinguishing brain dynamics in awake, healthy individuals from unconscious conditions. This problem also has a clinical diagnostic interest in disorders of consciousness. Recent research has shown that brain connectivity patterns characterized by long-range interactions and anticorrelations are associated with conscious states and diminish with loss of consciousness in human and non-human primates. However, the precise contribution of these patterns to conscious processing and subjective experience formation remains unclear. In this study, we investigated the functional role of these brain patterns in shaping conscious content by examining their influence on participants' ability to process external information during wakefulness. Participants underwent fMRI recordings during an auditory detection task. Phase coherence-based functional connectivity and k-means clustering confirmed that the ongoing dynamics were underpinned by brain patterns consistent with those identified in previous research, including the "high pattern" characteristic of conscious states. We found that the detection of auditory stimuli at threshold was specifically improved when the connectivity pattern at the time of presentation corresponded to this high-pattern. In return, the occurrence of the high-pattern increased after detection, indicating that participants were more likely to transition to a high-pattern following stimulus detection. Our findings suggest that ongoing brain dynamics and conscious perception mutually influence each other and that certain brain configurations are more favorable for conscious processing of external stimuli. In the future, targeting these moments of favorable patterns in patients with disorders of consciousness may help us identify windows of greater receptivity to the external world, paving the way for developing individualized patient care protocols.

Neurophysiological Correlates of Phase-Specific Enhancement of Motor Memory Consolidation via Slow-Wave Closed-Loop Targeted Memory Reactivation
Memory consolidation can be enhanced during sleep using targeted memory reactivation (TMR) and closed-loop (CL) acoustic stimulation on the up-phase of slow oscillations (SOs). Here, we tested whether applying TMR at specific phases of the SOs (up vs. down vs. no reactivation) could influence the behavioral and neural correlates of motor memory consolidation in healthy young adults. Results showed that up- (as compared to down-) state cueing resulted in greater performance improvement. Sleep electrophysiological data indicated that up-stimulated SOs exhibited higher amplitude and greater peak-nested sigma power. Task-related functional magnetic resonance images revealed that up-state cueing strengthened activity in - and segregation of - striato-motor and hippocampal networks; and that these modulations were related to the beneficial effect of TMR on sleep features and performance. Overall, these findings highlight the potential of CL-TMR to induce phase-specific modulations of motor performance, sleep oscillations and brain responses during motor memory consolidation.

Heat Shock Proteins Function as Signaling Molecules to Mediate Neuron-Glia Communication During Aging
The nervous system is primarily composed of neurons and glia, and the communication between them plays profound roles in regulating the development and function of the brain. Neuron-glia signal transduction is known to be mediated by secreted or juxtacrine signals through ligand-receptor interactions on the cell membrane. Here, we report a novel mechanism for neuron-glia signal transduction, wherein neurons transmit proteins to glia through extracellular vesicles, activating glial signaling pathways. We find that in the amphid sensory organ of Caenorhabditis elegans, different sensory neurons exhibit varying aging rates. This discrepancy in aging is governed by the crosstalk between neurons and glia. We demonstrate that early-aged neurons can transmit heat shock proteins (HSP) to glia via extracellular vesicles. These neuronal HSPs activate the IRE1-XBP1 pathway, further increasing their expression in glia, forming a positive feedback loop. Ultimately, the activation of the IRE1-XBP-1 pathway leads to the transcriptional regulation of chondroitin synthases to protect glia-embedded neurons from aging-associated functional decline. Therefore, our studies unveil a novel mechanism for neuron-glia communication in the nervous system and provide new insights into our understanding of brain aging.

Thermodynamic analog of integrate-and-fire neuronal networks by maximum entropy modelling
Relying on maximum entropy arguments, certain aspects of time-averaged experimental neuronal data have been recently described using Ising-like models, allowing the study of neuronal networks under an analogous thermodynamical framework. Here, we apply for the first time the Maximum Entropy method to an Integrate-and-fire (IF) model that can be tuned at criticality, offering a controlled setting for a systematic study of criticality and finite-size effects in spontaneous neuronal activity, as opposed to experiments. We show that generalized Ising models that accurately predict the average local activities and correlation functions between neurons of the IF model networks in the critical state exhibit a spin glass phase for low temperatures, having mostly negative intrinsic fields and a bimodal distribution of interaction constants that tends to become unimodal for larger networks. Results appear to be affected by sample-to-sample connectivity variations and subsampling. Furthermore, we also found that networks with higher percentage of inhibitory neurons lead to Ising-like systems with reduced thermal fluctuations. Finally, considering only neuronal pairs associated with the largest correlation functions allows the study of larger system sizes.

Δ9-tetrahydrocannabinol (THC) Increases the Rewarding Value of Oxycodone During Self-Administration in Rats
Background: Cannabis may reduce the nonmedical use of prescription opioids. Causality of polydrug use is difficult to establish from epidemiological data, and thus controlled laboratory models can test whether cannabinoid co-use with opioids can modulate opioid intake. Methods: Male and female rats were trained to intravenously self-administer (IVSA) oxycodone (0.15 mg/kg/infusion) during 6 h sessions. Separate groups were injected with the vehicle or with THC (5 mg/kg, i.p.; N=10) 30 minutes before sessions for the first three weeks. Treatments were swapped in the fourth week. One male group was trained in the intracranial self-stimulation (ICSS) procedure and assessed for brain reward thresholds prior to each IVSA session. Results: THC treated animals self-administered less oxycodone during acquisition, with a larger differential expressed in the female group. Tolerance to the THC effect developed over the initial weeks, and increasing the dose of THC (10 mg/kg, i.p.) prolonged the suppressing effect on IVSA. While ICSS thresholds increased with sequential IVSA sessions, no differences between THC- and Vehicle-treated groups were observed. Oxycodone IVSA was increased following the first 60 h abstinence interval in THC-treated, but not vehicle-treated, rats. Acute injection of THC, when all animals had been THC abstinent for several weeks, increased breakpoints in a Progressive Ratio procedure. Conclusion: These data support the interpretation that THC enhances the reinforcing efficacy of a given dose of oxycodone and may therefore increase the addiction liability.

miRNA family miR-29 inhibits PINK1-PRKN dependent mitophagy via ATG9A
Loss-of-function mutations in the genes encoding PINK1 and PRKN result in early-onset Parkinson disease (EOPD). Together the encoded enzymes direct a neuroprotective pathway that ensures the elimination of damaged mitochondria via autophagy. We performed a genome-wide high content imaging miRNA screen for inhibitors of the PINK1-PRKN pathway and identified all three members of the miRNA family 29 (miR-29). Using RNAseq we identified target genes and found that siRNA against ATG9A phenocopied the effects of miR-29 and inhibited the initiation of PINK1-PRKN mitophagy. Furthermore, we discovered two rare, potentially deleterious, missense variants (p.R631W and p.S828L) in our EOPD cohort and tested them experimentally in cells. While expression of wild-type ATG9A was able to rescue the effects of miR-29a, the EOPD-associated variants behaved like loss-of-function mutations. Together, our study validates miR-29 and its target gene ATG9A as novel regulators of mitophagy initiation. It further serves as proof-of-concept of finding novel, potentially disease-causing EOPD-linked variants specifically in mitophagy regulating genes. The nomination of genetic variants and biological pathways is important for the stratification and treatment of patients that suffer from devastating diseases, such as EOPD

Protective Effects of Lipoxin A4 and B4 Signaling on the Inner Retina in a Mouse Model of Experimental Glaucoma
Glaucoma is a common neurodegenerative disease characterized by progressive degeneration of retinal ganglion cells (RGCs) and the retinal nerve fiber layer (RNFL), resulting in a gradual decline of vision. A recent study by our groups indicated that the levels of lipoxins A4 (LXA4) and B4 (LXB4) in the retina and optic nerve decrease following acute injury, and that restoring their function is neuroprotective. Lipoxins are members of the specialized pro-resolving mediator (SPM) family and play key roles to mitigate and resolve chronic inflammation and tissue damage. Yet, knowledge about lipoxin neuroprotective activity remains limited. Here we investigate the in vivo efficacy of exogenous LXA4 and LXB4 administration on the inner retina in a mouse model of chronic experimental glaucoma. To investigate the contribution of LXA4 signaling we used transgenic knockout (KO) mice lacking the two mouse LXA4 receptors (Fpr2/Fpr3-/-). Functional and structural changes of inner retinal neurons were assessed longitudinally using electroretinogram (ERG) and optical coherence tomography (OCT). At the end of the experiment, retinal samples were harvested for immunohistological assessment. While both lipoxins generated protective trends, only LXB4 treatment was significant, and consistently more efficacious than LXA4 in all endpoints. Both lipoxins also appeared to dramatically reduce Muller glial reactivity following injury. In comparison, Fpr2/Fpr3 deletion significantly worsened inner retinal injury and function, consistent with an essential protective role for endogenous LXA4. Together, these results support further exploration of lipoxin signaling as a treatment for glaucomatous neurodegeneration.

Effects of High Fat Diet on Metabolic Health Vary by Age of Menopause Onset
Menopause accelerates metabolic dysfunction, including (pre-)diabetes, obesity and visceral adiposity. However, the effects of endocrine vs. chronological aging in this progression are poorly understood. We hypothesize that menopause, especially in the context of middle-age, will exacerbate the metabolic effects of a high fat diet. Using young-adult and middle-aged C57BL/6J female mice, we modeled diet-induce obesity via chronic administration of high fat (HF) diet vs. control diet. We modeled peri-menopause/menopause via injections of 4-vinylcyclohexene diepoxide, which accelerates ovarian failure vs. vehicle. We performed glucose tolerance tests 2.5 and 7 months after diet onset, during the peri-menopausal and menopausal phases, respectively. Peri-menopause increased the severity of glucose intolerance and weight gain in middle-aged, HF-fed mice. Menopause increased weight gain in all mice regardless of age and diet, while chronological aging drove changes in adipose tissue distribution towards more visceral vs. subcutaneous adiposity. These data are in line with clinical data showing that post-menopausal women are more susceptible to metabolic dysfunction and suggest that greater chorological age exacerbates the effects of endocrine aging (menopause). This work highlights the importance of considering both chronological and endocrine aging in studies of metabolic health.

Claustrum projections to the anterior cingulate modulate nociceptive and pain-associated behaviour
The anterior cingulate cortex (ACC) processes nociceptive information and pain unpleasant-ness. The claustrum is a subcortical region that provides robust feed-forward inhibition onto the ACC, suggesting this circuit could play a role in modulating pathological states such as pain. However, the function of this circuit in the context of acute and chronic inflammatory pain is un-clear. Here, we show that claustrocingulate neurons exhibit a bimodal pattern of activation follow-ing acute pain but are suppressed during states of chronic inflammatory pain. Molecular lesion and chemogenetic suppression of claustrocingulate neurons increased acute nociception but interfered with pain learning. Activation of this pathway rescued mechanical allodynia associated with chronic pain. Together, these results suggest claustrocingulate neurons are a critical com-ponent of the pain neuromatrix, and dysregulation of this connection may contribute to chronic pain.

Self-Supervised Transformer Model Training for a Sleep-EEG Foundation Model
The American Academy of Sleep Medicine (AASM) recognizes five sleep/wake states (Wake, N1, N2, N3, REM), yet this classification schema provides only a high-level summary of sleep and likely overlooks important neurological or health information. New, data-driven approaches are needed to more deeply probe the information content of sleep signals. Here we present a self-supervised approach that learns the structure embedded in large quantities of neurophysiological sleep data. This masked transformer training procedure is inspired by high performing self-supervised methods developed for speech transcription. We show that self-supervised pre-training matches or outperforms supervised sleep stage classification, especially when labeled data or compute-power is limited. Perhaps more importantly, we also show that our pretrained model is flexible and can be fine-tuned to perform well on new tasks including distinguishing individuals and quantifying "brain age" (a potential health biomarker). This suggests that modern methods can automatically learn information that is potentially overlooked by the 5-class sleep staging schema, laying the groundwork for new schemas and further data-driven exploration of sleep.

Counter-balancing X-linked Mecp2 hypofunction by hyperfunction ameliorates disease features in a model of Rett syndrome: implications for genetic therapies
Treating monogenic neurodevelopmental disorders remains challenging and mostly symptomatic. X-linked disorders affecting women such as the postnatal neurodevelopmental disorder Rett syndrome caused by mutations in the gene MECP2 have additional challenges due to dosage sensitivity and to cellular mosaicism caused by random X-chromosome inactivation. An approach to augment MECP2 expression from wild-type cells in RTT may be feasible and simpler than gene replacement but has never been tested due to known toxicity of MECP2 over-expression, as evidenced by the distinct neurological condition known as MECP2 Duplication Syndrome. Here, using genetic techniques, we find that counter-balancing Mecp2-null cells in female Mecp2-null/+ mice by a complementary population of cells harboring an X-linked transgene associated with 3X normal levels of MECP2 leads to normalization of multiple whole animal phenotypic outcomes without noticeable toxicity. In addition, in vivo LFP recordings demonstrate that counter-balancing Mecp2 loss-of-function improves select within-region and between-region abnormalities. By comparing the counter-balance approach with an approach based on cell autonomous restoration of MeCP2 using an autosomal transgene expressing 2X normal levels of MECP2 in all cells (mimicking gene replacement), we identify neurobehavioral and electrographic features best suited for preclinical biomarkers of a therapeutic response to cell autonomous versus non-cell autonomous correction. Notably, these proof-of-concept findings demonstrate how non-cell autonomous suppression of MeCP2 deficiency by boosting overall wild-type MeCP2 levels may be a viable disease-modifying therapy for RTT, with potential implications for genetic-based therapies of monogenic X-linked disorders.

Laying the Foundation: Modern Transformers for Gold-Standard Sleep Analysis
Accurate sleep assessment is critical to the practice of sleep medicine and sleep research. Excitingly, the recent availability of large quantities of publicly available sleep data present opportunities for data-driven discovery efforts. Transformers are flexible neural network architectures that not only excel at classification tasks, but also can enable data-driven discovery through un- or self-supervised learning, which requires no human annotations to the input data. The first step for data-driven discovery in sleep data with transformers is to train a model on a supervised learning task (here, scoring sleep) to form the foundation for unsupervised extensions; it is not yet clear which model optimizations are ideal for this. To address this gap and lay the groundwork for data-driven discovery, we explored optimizations of a transformer-based model that learned the canonical 5-class sleep stage classification task. We examined different configurations of model size, input data size (sleep sequence length), input channels, and the use of multiple or single nights of sleep during training. We found that most configurations met or exceeded clinical criteria (accuracy/agreement above approximately 80--85%). Curiously, they achieve this when operating over as little as 10 minutes of input data at a time, which would preclude the use of valuable sleep cycle history that a human expert typically uses. Compared to most recurrent neural network analogues, the attention-based transformer was better able to distinguish REM sleep from neighboring classes using EEG alone. Finally, the full-size transformer model did not significantly outperform smaller versions, indicating that lightweight versions are sufficient for this task. In summary, we found that small transformer models can match and even outclass other state-of-the-art automated sleep scoring techniques, while also providing the basis for future data-driven discovery efforts using large sleep data sets, with or without human annotations.

Retrieval of human aversive memories involves reactivation of gamma activity patterns in the hippocampus that originate in the amygdala during encoding
Emotional memories require coordinated activity of the amygdala and hippocampus. Human intracranial recordings have shown that formation of aversive memories involves an amygdala theta-hippocampal gamma phase code. Yet, the mechanisms engaged during translation of aversive experiences into memories and subsequent retrieval remain unclear. Directly recording from human amygdala and hippocampus, here we show that hippocampal gamma activity increases for correctly remembered aversive scenes, while exerting unidirectional oscillatory influence within the theta/beta frequency range on the amygdala for previously seen aversive scenes. Crucially, patterns of amygdala high amplitude gamma activity at encoding are reactivated in the hippocampus, but not amygdala, during both aversive encoding and retrieval. Trial-specific hippocampal gamma patterns showing highest representational similarity with amygdala activity at encoding are replayed in the hippocampus during aversive retrieval. This reactivation process occurs against a background of gamma activity that is otherwise decorrelated between encoding and retrieval. Thus, retrieval of aversive memories is hippocampal-centered, with hippocampal activity patterns apparently entrained by the amygdala during encoding.

CXCR3-expressing myeloid cells recruited to the hypothalamus protect against diet-induced body mass gain and metabolic dysfunction
Microgliosis is an important component of diet-induced hypothalamic inflammation in obesity. A few hours after the introduction of a high-fat diet, the mediobasal hypothalamus resident microglia undergo morphological and functional changes toward an inflammatory phenotype. If the consumption of large amounts of dietary fats persists for long periods, bone marrow-derived myeloid cells are recruited and integrated into a new landscape of hypothalamic microglia. However, it is currently unknown what are the transcriptional signatures and specific functions exerted by either resident or recruited subsets of hypothalamic microglia. Here, the elucidation of the transcriptional signatures revealed that resident microglia undergo only minor changes in response to dietary fats; however, under the consumption of a high-fat diet, there are major transcriptional differences between resident and recruited microglia with a major impact on chemotaxis. In addition, in recruited microglia, there are major transcriptional differences between females and males with an important impact on transcripts involved in neurodegeneration and thermogenesis. The chemokine receptor CXCR3 emerged as one of the components of chemotaxis with the greatest difference between recruited and resident microglia, and thus, was elected for further intervention. The hypothalamic immunoneutralization of CXCL10, one of the ligands for CXCR3, resulted in increased body mass gain and reduced energy expenditure, particularly in females. Furthermore, the chemical inhibition of CXCR3 resulted in a much greater change in phenotype with increased body mass gain, reduced energy expenditure, increased blood leptin, glucose intolerance, and reduced insulin. Thus, this study has elucidated the transcriptional differences between resident and recruited hypothalamic microglia in diet-induced obesity, identifying chemokines as a relevant subset of genes undergoing regulation. In addition, we showed that a subset of recruited microglia expressing CXCR3 has a protective, rather than a detrimental role in the metabolic outcomes promoted by the consumption of a high-fat diet, thus, establishing a new concept in obesity-associated hypothalamic inflammation.