bioRxiv Subject Collection: Neuroscience's Journal

Decoding multi-limb movements from low temporal resolution calcium imaging using deep learning
Two-photon imaging has been a critical tool for dissecting brain circuits and understanding brain function. However, relating slow two-photon calcium imaging data to fast behaviors has been challenging due to relatively low imaging sampling rates, thus limiting potential applications to neural prostheses. Here, we show that a recurrent encoder-decoder network with an output length longer than the input length can accurately decode limb trajectories of a running mouse from two-photon calcium imaging data. The encoder-decoder model could accurately decode information about all four limbs (contralateral and ipsilateral front and hind limbs) from calcium imaging data recorded in a single cortical hemisphere. Furthermore, neurons that were important for decoding were found to be well-tuned to both ipsilateral and contralateral limb movements, showing that artificial neural networks can be used to understand the function of the brain by identifying sub-networks of neurons that correlate with behaviors of interest.

Alterations of adult prefrontal circuits induced by early postnatal fluoxetine treatment mediated by 5-HT7 receptors.
The prefrontal cortex (PFC) plays a key role in high-level cognitive functions and emotional behaviors, and PFC alterations correlate with different brain disorders including major depression and anxiety. In mice, the first two postnatal weeks represent a critical period of high sensitivity to environmental changes. In this temporal window, serotonin (5- HT) levels regulate the wiring of PFC cortical neurons. Early life insults and postnatal exposure to the selective serotonin reuptake inhibitor fluoxetine (FLX) affect PFC development leading to depressive and anxiety-like phenotypes in adult mice. However, the mechanisms responsible for these dysfunctions remain obscure. We found that postnatal FLX exposure (PNFLX) results in reduced overall firing, and high-frequency bursting of putative pyramidal neurons (PNs) of deep layers of the medial PFC (mPFC) of adult mice in vivo. Ex-vivo, patch-clamp recordings revealed that PNFLX abolished high-frequency firing in a distinct subpopulation of deep-layer mPFC PNs, which transiently express the serotonin transporter SERT. SERT+ and SERT- PNs exhibit distinct morpho-functional properties. Genetic deletion of 5-HT7Rs prevented the PNFLX-induced reduction of PN firing in vivo and pharmacological 5-HT7R blockade precluded altered firing of SERT+ PNs in vitro. This indicates a pivotal role of this 5-HTR subtype in mediating 5-HT-dependent maturation of PFC circuits that are susceptible to early-life insults. Overall, our results suggest potential novel neurobiological mechanisms, underlying detrimental neurodevelopmental consequences induced by early-life alterations of 5-HT levels.

Remote memory engrams are controlled by encoding-specific tau phosphorylation
The engram represents the physical trace that encodes a specific memory and enables its recall. Functional failure of the engram is linked to the progressive memory decline in Alzheimer's disease. However, it is unknown whether the microtubule-associated protein tau, a central factor in Alzheimer's, has a direct function in the engram. Here, we demonstrate that tau and encoding-associated tau phosphorylation are critical for robust remote memory engrams. Tau is required specifically during memory formation for remote, yet not proximal recall in memory paradigms in mice. Controlled expression of tau exclusively during memory entrainment is necessary and sufficient to restore remote memory deficits in tau knockout mice. Tau is phosphorylated at specific sites during encoding. Gene editing to ablate site-specific phosphorylation at threonine-205 (T205) lowers precision of engram cell recruitment and precludes efficient remote recall. Vector-based engineering of engram cells reveals that T205 phosphorylation of tau is required to engrain memory for recall at remote timepoints. Notably, in the absence of tau, memory is recalled from latency by direct optogenetic activation of engram cells at distal time points but not when natural cues are used, revealing an association-specific gatekeeper function of tau during encoding. Our work delineates a physiologic role of site-specific tau phosphorylation at the inception of episodic memory to support an enduring engram and enable efficient remote recall. Thus, encoding-associated phosphorylation of tau is proximal to the elusive substrate of remote memory and may connect to the basis of amnesia in Alzheimer's disease.

Association between recovering from tempo perturbations and reading measurements.
A strong correlation between auditory temporal processing and reading proficiency has been consistently observed across clinical and nonclinical populations, spanning various age groups and languages. Specifically, rhythm sensitivity in both the music and speech domains has been considered to be fundamental for accurately tracking the hierarchical acoustic components in speech, playing a central role in the development of reading skills. However, the empirical validation of this hypothesis has primarily utilized stimuli with an isochronous underlying beat structure, which is limited in its ability to capture the nonlinearity inherent in the perception of timing within the speech and music domains. In our current study, we introduced perturbation stimuli and demonstrated a relationship between sensorimotor synchronization performance and reading measurements in the neurotypical adult population. Specifically, the current study highlighted that sensorimotor synchronization during the post-perturbation time window yields notably superior predictive value across a wide array of reading measurements when compared to the pre-perturbation time window, which, in contrast, did not predict reading measurements. Furthermore, our novel curve fitting analysis effectively captured the nonlinear aspects of participants' sensorimotor synchronization performance when recovering from tempo perturbation, providing further insight into their auditory temporal processing abilities when responding to timing changes in auditory signals, a phenomenon commonly encountered in both speech and music contexts.

Developmental remodeling repurposes larval neurons for sexual behaviors in adult Drosophila
Most larval neurons in Drosophila are repurposed during metamorphosis for functions in adult life, but their contribution to the neural circuits for sexually dimorphic behaviors is unknown. Here, we identify two interneurons in the nerve cord of adult Drosophila females that control ovipositor extrusion, a courtship rejection behavior performed by mated females. We show that these two neurons are present in the nerve cord of larvae as mature, sexually monomorphic interneurons. During pupal development, they acquire the expression of the sexual differentiation gene, doublesex, undergo doublesex-dependent programmed cell death in males, and are remodeled in females for functions in female mating behavior. Our results demonstrate that the neural circuits for courtship in Drosophila are built in part using neurons that are sexually reprogrammed from former sex-shared activities in larval life.

A fast and flexible approximation of power-law adaptation for auditory computational models
Power-law adaptation is a form of neural adaptation that has been shown to provide a better description of auditory-nerve adaptation dynamics as compared to simpler exponential-adaptation processes. However, the computational costs associated with power-law adaptation are high and, problematically, grow superlinearly with the number of samples in the simulation. This cost limits the applicability of power-law adaptation in simulations of responses to relatively long stimuli, such as speech, or in simulations for which high sampling rates are needed. Here, we present a simple approximation to power-law adaptation based on a parallel set of exponential-adaptation processes with different time constants, demonstrate that the approximation improves on an existing approximation provided in the literature, and provide updates to a popular phenomenological model of the auditory periphery that implements the new approximation.

Negative feedback control of hunger circuits by the taste of food
The rewarding taste of food is critical for motivating animals to eat, but whether taste has a parallel function in promoting meal termination is not well understood. Here we show that hunger-promoting AgRP neurons are rapidly inhibited during each bout of ingestion by a signal linked to the taste of food. Blocking these transient dips in activity via closed-loop optogenetic stimulation increases food intake by selectively delaying the onset of satiety. We show that upstream leptin receptor-expressing neurons in the dorsomedial hypothalamus (DMHLepR) are tuned to respond to sweet or fatty tastes and exhibit time-locked activation during feeding that is the mirror image of downstream AgRP cells. These findings reveal an unexpected role for taste in the negative feedback control of ingestion. They also reveal a mechanism by which AgRP neurons, which are the primary cells that drive hunger, are able to influence the moment-by-moment dynamics of food consumption.

Ivermectin increases striatal cholinergic activity to facilitate dopamine terminal function
Ivermectin (IVM) is a commonly prescribed antiparasitic treatment with pharmacological effects on invertebrate glutamate ion channels resulting in paralysis and death of invertebrates. However, it can also act as a modulator of some vertebrate ion channels and has shown promise in facilitating L-DOPA treatment in preclinical models of Parkinson's disease. The pharmacological effects of IVM on dopamine terminal function were tested, focusing on the role of two of IVM's potential targets: purinergic P2X4 and nicotinic acetylcholine receptors. Ivermectin enhanced electrochemical detection of dorsal striatum dopamine release. Although striatal P2X4 receptors were observed, IVM effects on dopamine release were not blocked by P2X4 receptor inactivation. In contrast, IVM attenuated nicotine effects on dopamine release, and antagonizing nicotinic receptors prevented IVM effects on dopamine release. IVM also enhanced striatal cholinergic interneuron firing. L-DOPA enhances dopamine release by increasing vesicular content. L-DOPA and IVM co-application further enhanced release but resulted in a reduction in the ratio between high and low frequency stimulations, suggesting that IVM is enhancing release largely through changes in terminal excitability and not vesicular content. Thus, IVM is increasing striatal dopamine release through enhanced cholinergic activity on dopamine terminals.

Reduction of low-frequency oscillations in cerebral circulation correlates with pupil dilation during cognition: an fNIRS study
Numerous studies have utilized functional near-infrared spectroscopy (fNIRS) to investigate brain hemodynamics during diverse cognitive tasks. However, although these studies have consistently reported increased oxygenation levels in the prefrontal cortex, they have not explored their effects on the amplitude of brain low-frequency oscillations (LFO). Additionally, other reports have shown that pupil dilation occurs during cognitive tasks, which may indicate a correlation between LFO amplitude and pupil dilation. However, no study has demonstrated such a correlation. Our research has two aims: firstly, to analyze the impact of cognitive tasks on the amplitude of LFO using oxyhemoglobin (HbO2) fNIRS signals, and secondly, to assess the relationship between the amplitude of these LFO and pupil diameters during such cognitive tasks. We found that during arithmetic tasks, the brain LFOs recorded on the prefrontal cortex were temporarily reduced while pupil diameter increased. These findings offer new insights into the physiological functions of reactivity of LFO in cerebral circulation. Additionally, combining fNIRS signals to track LFO on the prefrontal cortex with pupil measurement shows the feasibility of developing efficient hybrid brain-computer interfaces of LFO-pupil detection capable of predicting the initiation and ending of cognitive processes. Significant statement: Inspired by the Berger effect, which demonstrates alpha wave reactivity with eye-opening, our study investigated reactivity in other physiological oscillations related to brain hemodynamics during cognition. Our findings revealed a novel phenomenon: a temporary reduction in the amplitude of low-frequency oscillations (LFOs) of oxyhemoglobin (HbO2) in the brain's prefrontal cortex, concurrent with increased pupil diameter during arithmetic tasks. This discovery expands our knowledge of LFO in cerebral circulation and creates new avenues for studying fNIRS-LFO reactivity during cognitive tasks in healthy individuals and patients.

Profile analysis in listeners with sensorineural hearing loss: behavioral data and computational models
Many sounds contain spectral modulations at multiple scales, but much is still unknown about how such spectral features are represented in the auditory system. One behavioral task that provides insight into this question is profile analysis. In a typical profile-analysis task, listeners are asked to discriminate between a complex tone with equal-amplitude components and a complex tone with a single incremented component. Because listeners can perform profile analysis even when the overall sound level of the stimuli is randomized from interval to interval, this task is thought to be a useful index of relative processing of spectral shape, rather than just sensitivity to absolute level changes. Here, we measured profile analysis across the frequency range in a group of listeners that varied widely in their hearing status. We then modeled the resulting behavioral data by decoding responses to the stimuli from computational models of the auditory nerve and inferior colliculus. We found that both hearing loss at the target frequency and increases in the target frequency were associated with poorer profile-analysis thresholds, and that these results could both be explained as the result of corresponding changes in sensitivity of temporal modulation-sensitive cells at the level of the inferior colliculus. These results suggest that key features of profile-analysis may reflect the limits of central neural tuning to temporal modulations.

Noradrenergic signaling controls Alzheimer's disease pathology via activation of microglial β2 adrenergic receptors
In Alzheimer disease (AD) pathophysiology, plaque and tangle accumulation trigger an inflammatory response that mounts positive feed-back loops between inflammation and protein aggregation, aggravating neurite damage and neuronal death. One of the earliest brain regions to undergo neurodegeneration is the locus coeruleus (LC), the predominant site of norepinephrine (NE) production in the central nervous system (CNS). In animal models of AD, dampening the impact of noradrenergic signaling pathways, either through administration of beta blockers or pharmacological ablation of the LC, heightened neuroinflammation through increased levels of pro-inflammatory mediators. Since microglia are the resident immune cells of the CNS, it is reasonable to postulate that they are responsible for translating the loss of NE tone into exacerbated disease pathology. Recent findings from our lab demonstrated that noradrenergic signaling inhibits microglia dynamics via {beta}2 adrenergic receptors ({beta}2ARs), suggesting a potential anti-inflammatory role for microglial {beta}2AR signaling. Thus, we hypothesize that microglial {beta}2 adrenergic signaling is progressively impaired during AD progression, which leads to the chronic immune vigilant state of microglia that worsens disease pathology. First, we characterized changes in microglial {beta}2AR signaling as a function of amyloid pathology. We found that LC neurons and their projections degenerate early and progressively in the 5xFAD mouse model of AD; accompanied by mild decrease in the levels of norepinephrine and its metabolite normetanephrine. Interestingly, while 5xFAD microglia, especially plaque-associated microglia, significant downregulated {beta}2AR gene expression early in amyloid pathology, they did not lose their responsiveness to {beta}2AR stimulation. Most importantly, we demonstrated that specific microglial {beta}2AR deletion worsened disease pathology while chronic {beta}2AR stimulation resulted in attenuation of amyloid pathology and associated neuritic damage, suggesting microglial {beta}2AR might be used as potential therapeutic target to modify AD pathology.

Activity-dependent capture of neuropeptide vesicles prepares clock neuron synapses for daily release
Drosophila brain sLNv clock neurons release the neuropeptide PDF to control circadian rhythms. Strikingly, PDF content in sLNv terminals is rhythmic with a peak in the morning. Peak content drops because of activity-dependent release from dense-core vesicles (DCVs), but the mechanism for the daily increase in presynaptic PDF in the hours prior to release is unknown. Although transport from the soma was proposed to drive the daily increase in presynaptic PDF, live imaging in sLNv neurons shows that anterograde axonal DCV transport is constant throughout the day. Instead, capture of circulating DCVs, indicated by decreased retrograde axonal transport, rhythmically boosts presynaptic neuropeptide content. Genetic manipulations demonstrate that the late night increase in capture requires electrical activity but is independent of daily morphological changes. These results suggest that each day, during the hours of ongoing electrical activity, a toggle switches from inducing vesicle capture to triggering exocytosis, thereby maximizing daily rhythmic bursts of synaptic neuropeptide release by clock neurons.

Transcriptomic correlates of state modulation in forebrain interneurons: A cross-species analysis
GABAergic inhibitory interneurons comprise many subtypes that differ in their molecular, anatomical and functional properties. In mouse visual cortex, they also differ in their modulation with an animal's behavioural state, and this state modulation can be predicted from the first principal component (PC) of the gene expression matrix (Bugeon et al., 2022). Here, we ask whether this link between transcriptome and state-dependent processing generalises across species. To this end, we analysed seven single-cell and single-nucleus RNA sequencing datasets from mouse, human, songbird, and turtle forebrains. Despite homology at the level of cell types, we found clear differences between transcriptomic PCs, with greater dissimilarities between evolutionarily distant species. These dissimilarities arise from two factors: divergence in gene expression within homologous cell types and divergence in cell type abundance. We also compare the expression of cholinergic receptors, which are thought to causally link transcriptome and state modulation. Several cholinergic receptors predictive of state modulation in mouse interneurons are differentially expressed between species. Circuit modelling and mathematical analyses delineate the conditions under which these expression differences could translate into functional differences.

Dynamic changes in somatosensory and cerebellar activity mediate temporal recalibration of self-touch
An organism's ability to accurately anticipate the sensations caused by its own actions is crucial for a wide range of behavioral, perceptual, and cognitive functions. Notably, the sensorimotor expectations produced when touching one's own body attenuate such sensations, making them feel weaker and less ticklish and rendering them easily distinguishable from potentially harmful touches of external origin. How the brain learns and keeps these action-related sensory expectations updated is unclear. We employed psychophysics and functional magnetic resonance imaging to pinpoint the behavioral and neural substrates of dynamic recalibration of expected temporal delays in self-touch. Psychophysical results revealed that self-touches were less attenuated after systematic exposure to delayed self-generated touches, while responses in the contralateral somatosensory cortex that normally distinguish between delayed and nondelayed self-generated touches became indistinguishable. During the exposure, the ipsilateral anterior cerebellum showed increased activity, supporting its proposed role in recalibrating sensorimotor predictions. Moreover, responses in the cingulate areas gradually increased, suggesting that as delay adaptation progresses, the nondelayed self-touches trigger activity related to cognitive conflict. Together, our results show that sensorimotor predictions in the simplest act of touching one's own body are upheld by a sophisticated and flexible neural mechanism that maintains them accurate in time.

Long-term abstinence reduces optimal energy required for transitions from the exogenous networks to the frontoparietal network in methamphetamine addicts
Methamphetamine (MA) addiction is a chronic neurotoxic brain disorder that places a significant burden on public health, with a high relapse risk. Long-term abstinence can significantly reduce craving, yet the potential alterations caused by long-term abstinence still remain unclear. In this study, a total of 62 MA users who underwent longitudinal follow-up during their period of long-term abstinence (duration of long-term abstinence: mean: 347.52, std: 99.25 days) were enrolled. For the first time, we employed a promising framework known as network control theory to explore the impact of long-term abstinence on MA addicts. Our observations indicated that long-term abstinence led to a decline in the control energy required for transitions from the visual network, and somatomotor network to the frontoparietal network, suggesting a reduced barrier for triggering executive control response to the drug cues. Furthermore, we identified the orbitofrontal cortex and the dorsolateral prefrontal cortex as crucial regions involved in facilitating these transitions. Notably, we discovered significant associations between the influence of long-term abstinence on brain regions and the spatial distribution of key biological factors, such as DAT, 5HTT, D2, MOR, VAChT, and NET. Overall, our findings not only provide a novel perspective on understanding the impact of long-term abstinence in MA addicts but also link this process to biological factors.

Small RNAs in plasma extracellular vesicles define biomarkers of premanifest changes in Huntington's disease
Despite the advances in the understanding of Huntington's disease (HD), there is the need for molecular biomarkers to categorize mutation-carriers during the preclinical stage of the disease preceding the functional decline. Small RNAs (sRNAs) are a promising source of biomarkers since their expression levels are highly sensitive to pathobiological processes. Here, using an optimized method for plasma extracellular vesicles (EVs) purification and an exhaustive analysis pipeline of sRNA sequencing data, we show that EV-sRNAs are early downregulated in mutation-carriers, and that this deregulation is associated with premanifest cognitive performance. Seven candidate sRNAs (tRF-Glu-CTC, tRF-Gly-GCC, miR-451a, miR-21-5p, miR-26a-5p, miR-27a-3p, and let7a-5p) were validated in additional subjects, showing a significant diagnostic accuracy at premanifest stages. Of these, miR-21-5p was significantly decreased over time in a longitudinal study; and miR-21-5p and miR-26a-5p levels correlated with cognitive changes in the premanifest cohort. In summary, the present results suggest that deregulated plasma EV-sRNAs define an early biosignature in mutation carriers with specific species sensing the progression and cognitive changes occurring at the premanifest stage.

Atrophy-driven functional network collapse in neurodegenerative disease
Cognitive deficits in Alzheimer's disease (AD) and frontotemporal dementia (FTD) result from atrophy and altered functional connectivity. However, it is unclear how atrophy and functional connectivity disruptions relate across dementia subtypes and stages. We addressed this question using structural and functional MRI from 221 patients with AD (n=82), behavioral variant FTD (n=41), corticobasal syndrome (n=27), nonfluent (n=34) and semantic (n=37) variant primary progressive aphasia, and 100 cognitively normal individuals. Using partial least squares regression, we identified three principal structure-function components. The first component showed cumulative atrophy correlating with primary cortical hypo-connectivity and subcortical/fronto-parietal association cortical hyper-connectivity. The second and third components linked focal syndrome-specific atrophy patterns to peri-lesional hypo-connectivity and distal hyper-connectivity. Structural and functional component scores collectively predicted global and domain-specific cognitive deficits. Anatomically, functional connectivity decreases and increases reflected alterations in specific brain activity gradients. Eigenmode analysis identified temporal phase and amplitude disruptions as a potential explanation for atrophy-driven functional connectivity changes.

Dynamic Causal Modelling Highlights the Importance of Decreased Self-Inhibition of the Sensorimotor Cortex in Motor Fatigability
Motor fatigability emerges when challenging motor tasks must be maintained over an extended period of time. It is a frequently observed phenomenon in everyday life which affects patients as well as healthy individuals. Motor fatigability can be measured using simple tasks like finger tapping at maximum speed for 30s. This typically results in a rapid decrease of tapping frequency, a phenomenon called motor slowing. In a previous study (Bachinger et al. 2019), we showed that motor slowing goes hand in hand with a gradual increase of activation in the primary sensorimotor cortex (SM1), supplementary motor area (SMA), and dorsal premotor cortex (PMd). Previous electrophysiological measurements further suggested that the increase in SM1 activity might reflect a breakdown of inhibition and, particularly, a breakdown of surround inhibition which might have led to heightened coactivation of antagonistic muscles. It is unclear what drives the activity increase in SM1 caused by motor slowing and whether motor fatigability affects the dynamic interactions between SM1 and upstream motor areas like SMA and PMd. Here, we performed dynamic causal modelling to answer this question. Our main findings revealed that motor slowing was associated with a significant reduction in SM1 self-inhibition which is in line with previous electrophysiological results. Additionally, the model revealed a significant decrease in the driving input to premotor areas suggesting that structures other than cortical motor areas might cause motor fatigability.

Inducing representational change in the hippocampus through real-time neurofeedback
When you perceive or remember one thing, other related things come to mind. This competition has consequences for how these items are later perceived, attended, or remembered. Such behavioral consequences result from changes in how much the neural representations of the items overlap, especially in the hippocampus. These changes can reflect increased (integration) or decreased (differentiation) overlap; previous studies have posited that the amount of coactivation between competing representations in cortex determines which will occur: high coactivation leads to hippocampal integration, medium coactivation leads to differentiation, and low coactivation is inert. However, those studies used indirect proxies for coactivation, by manipulating stimulus similarity or task demands. Here we induce coactivation of competing memories in visual cortex more directly using closed-loop neurofeedback from real-time fMRI. While viewing one object, participants were rewarded for implicitly activating the representation of another object as strongly as possible. Across multiple real-time fMRI training sessions, they succeeded in using the neurofeedback to induce coactivation. Compared with untrained objects, this coactivation led to behavioral and neural integration: The trained objects became harder for participants to discriminate in a categorical perception task and harder to decode from patterns of fMRI activity in the hippocampus.

A dynamic attractor network model of memory formation, reinforcement and forgetting
Empirical evidence shows that memories that are frequently revisited are easy to recall, and that familiar items involve larger hippocampal representations than less familiar ones. In line with these observations, here we develop a modelling approach to provide a mechanistic hypothesis of how hippocampal neural assemblies evolve differently, depending on the frequency of presentation of the stimuli. For this, we added an online Hebbian learning rule, background firing activity, neural adaptation and heterosynaptic plasticity to a rate attractor network model, thus creating dynamic memory representations that can persist, increase or fade according to the frequency of presentation of the corresponding memory patterns. Specifically, we show that a dynamic interplay between Hebbian learning and background firing activity can explain the relationship between the memory assembly sizes and their frequency of stimulation. Frequently stimulated assemblies increase their size independently from each other (i.e. creating orthogonal representations that do not share neurons, thus avoiding interference). Importantly, connections between neurons of assemblies that are not further stimulated become labile so that these neurons can be recruited by other assemblies, providing a neuronal mechanism of forgetting.

Behavioural and electrophysiological correlates of the cross-modal enhancement for unaware visual events
According to many reports, cross-modal interactions can lead to enhancement of visual perception, even when visual events appear below awareness. Yet, the mechanism underlying this cross-modal enhancement is still unclear. The present study addressed whether cross-modal integration based on bottom-up processing can break through the threshold of awareness. We used a binocular rivalry protocol, and measured ERP responses and perceptual switches time-locked to flashes, sounds or flash-sound co-occurrences. In behavior, perceptual switches happened the earliest when subthreshold flashes co-occurred with sounds. Yet, this cross-modal facilitation never surpassed the benchmark indicated by the probability summation, thus suggesting independence rather than integration of sensory signals. Likewise, the ERPs to audiovisual events did not differ from the summed unimodal ERPs, also suggesting that the cross-modal behavioural benefit for unaware visual events can be explained by the independent contribution of unisensory signals and suggest no need for a multisensory integration mechanism. Hence, even though cross-modal benefits appeared behaviourally, we suggest that this cross-modal facilitation might origin from well-known bottom-up attentional capture processes, contributed by each individual sensory stimulus.

Brain network dynamics predict moments of surprise across contexts
We experience surprise when reality conflicts with our expectations. When we encounter such expectation violations in psychological tasks and daily life, are we experiencing completely different forms of surprise? Or is surprise a fundamental psychological process with shared neural bases across contexts? To address this question, we identified a brain network model, the surprise edge-fluctuation-based predictive model (EFPM), whose regional interaction dynamics measured with functional magnetic resonance imaging (fMRI) predicted surprise in an adaptive learning task. The same model generalized to predict surprise as a separate group of individuals watched suspenseful basketball games. The surprise EFPM also uniquely predicts surprise, capturing expectation violations better than models built from other brain networks, fMRI measures, and behavioral metrics. These results suggest that shared neurocognitive processes underlie surprise across contexts and that distinct experiences can be translated into the common space of brain dynamics.

Humans progressively feel agency over events triggered before their actions
AI has become increasingly efficient in anticipating our behavior. Will this impact, in the near future, how much we feel control over events generated with AI-assistance? In everyday life, our sense of agency over events occurring at various delays after our actions has adapted to accommodate these delays. Here we investigate whether our sense of agency can also adapt to a highly unusual situation, in which a consequence precedes an action. We used an online game where players tried to beat the computer at finding and clicking on a target to trigger an animation, while in fact an algorithm triggered the animation before the players' click. The animation was not randomly controlled by the algorithm, but rather based on the history of the players' past movements and on the beginning of their current movement. We used modeling and machine learning decoding approaches to capture how players compute their reported sense of agency over the animation. We found evidence that, in less than an hour, players implicitly learned, despite the unusual timing, that they were controlling the animation and adapted their sense of agency accordingly. Such findings may help us to anticipate how humans will integrate AI-assistance to guide their behavior.

Continuous estimation of reaching space in superficial layers of the motor cortex
Motor cortex plays a key role in controlling voluntary arm movements towards spatial targets. The cortical representation of spatial information has been extensively studied and was found to range from combinations of muscle synergies to cognitive maps of locations in space. How such abstract representations of target space evolve during a behavior, how they integrate with other behavioral features and what role they play in movement control is less clear. Here we addressed these questions by recording the activity of layer 2/3 (L2/3) neurons in the motor cortex using two-photon calcium imaging in head-restrained mice, while they reached for water droplets presented at different spatial locations around their snout. Our results reveal that a majority (>80%) of L2/3 neurons with task-related activity are target-space selective and their activity is contingent on a single target position in an ego-centric reference frame. This spatial framework is preferentially organized along three cardinal directions (Center, Left and Right). Surprisingly, the coding of target space is not limited to the activity during movement planning or execution, but is also predominant during preceding and subsequent phases of the task, and even persists beyond water consumption. More importantly, target specificity is independent of the movement kinematics and is immediately updated when the target is moved to a new position. Our findings suggest that, rather than descending motor commands, the ensemble of L2/3 neurons in the motor cortex conjointly encode internal (behavioral) and external (spatial) aspects of the task, playing a role in higher-order representations related to estimation processes of the ongoing actions.

Network-level encoding of local neurotransmitters in cortical astrocytes
Astrocytes--the most abundant non-neuronal cell type in the mammalian brain--are crucial circuit components that respond to and modulate neuronal activity via calcium (Ca2+) signaling. Astrocyte Ca2+ activity is highly heterogeneous and occurs across multiple spatiotemporal scales: from fast, subcellular activity to slow, synchronized activity that travels across connected astrocyte networks. Furthermore, astrocyte network activity has been shown to influence a wide range of processes. While astrocyte network activity has important implications for neuronal circuit function, the inputs that drive astrocyte network dynamics remain unclear. Here we used ex vivo and in vivo two-photon Ca2+ imaging of astrocytes while mimicking neuronal neurotransmitter inputs at multiple spatiotemporal scales. We find that brief, subcellular inputs of GABA and glutamate lead to widespread, long-lasting astrocyte Ca2+ responses beyond an individual stimulated cell. Further, we find that a key subset of Ca2+ activity--propagative events--differentiates astrocyte network responses to these two major neurotransmitters, and gates responses to future inputs. Together, our results demonstrate that local, transient neurotransmitter inputs are encoded by broad cortical astrocyte networks over the course of minutes, contributing to accumulating evidence across multiple model organisms that significant astrocyte-neuron communication occurs across slow, network-level spatiotemporal scales. We anticipate that this study will be a starting point for future studies investigating the link between specific astrocyte Ca2+ activity and specific astrocyte functional outputs, which could build a consistent framework for astrocytic modulation of neuronal activity.

Short-Term Meditation Training Alters Brain Activity and Sympathetic Responses at Rest, but not during the meditation
Numerous studies have shown that meditation has a number of positive effects on the physical and psychological well-being of practitioners. As a result, meditation has become widely practiced not only as a religious practice but also as a self-regulation technique to achieve specific measurable goals. This raises the question of how quickly physiological changes can be noticed in individuals for whom meditation is not the main focus of their lives but rather a wellbeing keeping technique. Another question is whether it is possible to observe changes occurring directly during meditation and use bio- or neuro-feedback to enhance such meditation training and achieve tangible results. In our study, the experimental group of individuals with no previous meditation experience underwent eight weeks of training in Taoist meditation (2 sessions lasting 1 hour each week), under the guidance of a certified instructor. Participants in the control group attended offline group meetings during the same period, where they listened to audio books. All participants performed meditation testing before and after the intervention, following audio instructions. During the meditation practice, participants' EEG, photoplethysmogram, respiratory rate, and skin conductance were recorded. The meditation training, but not the control group activity, resulted in topically organized changes of the resting state brain activity and heart rate variability. Specifically, we observed an increase in EEG power in multiple frequency bands (delta, theta, alpha, beta) and changes in the heart rate variability indicators associated with sympathetic system activation. However, no significant changes were observed when we compared the physiological indicators during the actual meditation process performed prior and post the 8-week training. We interpret these changes as signs of increased alertness and possibly accelerated resting metabolic rate. Importantly, these changes were observed after only 16 hours of meditation training performed during the 8-week period of time. The absence of difference in the band-specific power profiles between the experimental and control groups during the process of meditation conceptually complicates the development of assistive devices aimed at guiding the novice meditators during the actual meditation. Our results suggest that the focus in creating such digital assistants should rather be shifted towards monitoring neurophysiological activity during the time intervals outside of the actual meditation. The apparent changes occur not only in the EEG derived parameters but are also detectable based on the markers of autonomous nervous system activity that can be readily registered with a range of wearable gadgets which renders hope for a rapid translation of our results into practical applications.

Translating from mice to humans: using preclinical blood-based biomarkers for the prognosis and treatment of traumatic brain injury
Rodent models are important research tools for studying the pathophysiology of traumatic brain injury (TBI) and developing potential new therapeutic interventions for this devastating neurological disorder. However, the failure rate for the translation of drugs from animal testing to human treatments for TBI is 100%, perhaps due, in part, to distinct timescales of pathophysiological processes in rodents versus humans that impedes translational advancements. Incorporating clinically relevant biomarkers in preclinical studies may provide an opportunity to calibrate preclinical models to human TBI biomechanics and pathophysiology. To support this important translational goal, we conducted a systematic literature review of preclinical TBI studies in rodents measuring blood levels of clinically used NfL, t-Tau, p-Tau, UCH-L1, or GFAP, published in PubMed/MEDLINE up to June 13th, 2023. We focused on blood biomarker temporal trajectories and their predictive and pharmacodynamic value and discuss our findings in the context of the latest clinical TBI biomarker data. Out of 369 original studies identified through the literature search, 71 met the inclusion criteria, with a median quality score on the CAMARADES checklist of 5 (interquartile range 4-7). NfL was measured in 17 preclinical studies, GFAP in 41, t-Tau in 17, p-Tau in 7, and UCH-L1 in 19 preclinical studies. Data in rodent models show that all blood biomarkers exhibited injury severity-dependent elevations, with GFAP and UCH-L1 peaking within hours after TBI, NfL peaking within days after TBI and remaining elevated up to 6 months post-injury, whereas t-Tau and p-Tau levels were gradually increased many weeks after TBI. Blood NfL levels emerges as a prognostic indicator of white matter loss after TBI, while both NfL and GFAP hold promise for pharmacodynamic studies of neuroprotective treatments. Therefore, blood-based preclinical biomarker trajectories could serve as important anchor points that may advance translational research in the TBI field. However, further investigation into biomarker levels in the subacute and chronic phases will be needed to more clearly define pathophysiological mechanisms and identify new therapeutic targets for TBI.

An inducible genetic tool for tracking and manipulating specific microglial states in development and disease
Recent single-cell RNA sequencing studies have revealed distinct microglial states in development and disease. These include proliferative region-associated microglia (PAM) in developing white matter and disease-associated microglia (DAM) prevalent in various neurodegenerative conditions. PAM and DAM share a similar core gene signature and other functional properties. However, the extent of the dynamism and plasticity of these microglial states, as well as their functional significance, remains elusive, partly due to the lack of specific tools. Here, we report the generation of an inducible Cre driver line, Clec7a-CreERT2, designed to target PAM and DAM in the brain parenchyma. Utilizing this tool, we profile labeled cells during development and in several disease models, uncovering convergence and context-dependent differences in PAM/DAM gene expression. Through long-term tracking, we demonstrate surprising levels of plasticity in these microglial states. Lastly, we specifically depleted DAM in cuprizone-induced demyelination, revealing their roles in disease progression and recovery.

Astrocyte coverage of excitatory synapses correlates to measures of synapse structure and function in primary visual cortex
Most excitatory synapses in the mammalian brain are contacted by astrocytes, forming the tripartite synapse. This interface is thought to be critical for glutamate turnover and structural or functional dynamics of synapses. While the degree of synaptic contact of astrocytes is known to vary across brain regions and animal species, the implications of this variability remain unknown. Furthermore, precisely how astrocyte coverage of synapses relates to in vivo functional properties of individual dendritic spines has yet to be investigated. Here, we characterized perisynaptic astrocyte processes (PAPs) contacting synapses of pyramidal neurons of the ferret visual cortex and, using correlative light and electron microscopy, examined their relationship to synaptic strength and to sensory-evoked Ca2+ activity. Nearly all synapses were contacted by PAPs, and most were contacted along the axon-spine interface (ASI). Structurally, we found that the degree of PAP coverage scaled with synapse size and complexity. Functionally, we found that PAP coverage scaled with the selectivity of Ca2+ responses of individual synapses to visual stimuli and, at least for the largest synapses, scaled with the reliability of visual stimuli to evoke postsynaptic Ca2+ events. Our study shows astrocyte coverage is highly correlated with structural properties of excitatory synapses in the visual cortex and implicates astrocytes as a contributor to reliable sensory activation.

Mild blast TBI raises gamma connectivity, EEG power, and reduces GABA interneuron density
At least one traumatic brain injury (TBI) will be experienced by approximately 50-60 million of the world's population in their lifetime and is the biggest cause of death and disability in those under 40. Mild traumatic brain injury (mTBI) can induce subtle changes but have long-lasting effects that may be difficult to detect through conventional neurological assessment, including standard clinical imaging techniques. These changes can lead to an increased risk of future neurodegeneration and emphasises the need to use more sensitive diagnostic tools such as EEG in order to identify injury and opportunities for therapeutic intervention. In this study, we investigated electrophysiological and histopathological changes in a rat model of mild blast-induced TBI. We used a 32-channel EEG electrode array to detect global and local changes in neural activity and functional connectivity in acute (3 to 4-hours) as well as chronic phases (1 and 3-months) post-injury. GABAergic inhibitory interneurons, crucial for maintaining an excitatory/inhibitory balance, were quantified using immunohistochemistry. Mild blast-induced TBI had minimal effects on resting power and connectivity at the acute timepoint but resulted in resting-state global power increases at all frequencies as well as a relative power increase in slow-wave frequencies in the chronic phase post-injury. Functional connectivity increases in the gamma frequency along with increases in power in the chronic phase pointed towards an alteration in the excitatory/inhibitory balance. Indeed, electrophysiological changes were associated with reduced density of GABAergic interneurons at 7 days, 1 month, and 3 months post-injury, with a decrease in somatostatin-positive cell density in the 5th layer of all cortical regions of interest, and a parvalbumin decrease in the 5th layer of the primary auditory cortex. In contrast, the total number of neurons, measured by NeuN did not change significantly, thus demonstrating a biased impact on inhibitory interneuron populations. Our work demonstrates that the techniques and metrics of injury assessment employed in this study are sensitive enough to reflect the subtle changes present in mTBI and therefore hold potential clinical relevance. By using non-invasive EEG assessments and histopathology, we were able to reveal direct correlates and potential sources of the abnormalities caused by mild blast-induced TBI.

Cortical Network Disruption is Minimal in Early Stages of Psychosis
Background and Hypothesis: Chronic schizophrenia is associated with white matter disruption and topological reorganization of cortical connectivity but the trajectory of these changes over the disease course are poorly understood. Current white matter studies in first-episode psychosis (FEP) patients using diffusion magnetic resonance imaging (dMRI) suggest such disruption may be detectable at the onset of psychosis, but specific results vary widely and few reports have contextualized their findings with direct comparison to chronic patients. Here, we test the hypothesis that structural changes are not a significant feature of early psychosis. Study Design: Diffusion and T1-weighted 7T MR scans were obtained from N=113 (61 FEP patients, 37 controls, 15 chronic patients) recruited from an established cohort in London, Ontario. Voxel- and network-based analyses were used to detect changes in diffusion microstructural parameters. Graph theory metrics were used to probe changes in the cortical network hierarchy and to assess the vulnerability of hub regions to disruption. Experiments were replicated with N=167 (111 patients, 56 controls) from the Human Connectome Project - Early Psychosis (HCP-EP) dataset. Study Results: Widespread microstructural changes were found in chronic patients, but changes in FEP patients were minimal. Unlike chronic patients, no appreciable topological changes in the cortical network were observed in FEP patients. These results were replicated in the early psychosis patients of the HCP-EP datasets, which were indistinguishable from controls on nearly all metrics. Conclusions: The white matter structural changes observed in chronic schizophrenia are not a prominent feature in the early stages of this illness.

A Multi-Scale Study of Thalamic State-Dependent Responsiveness
The thalamus is the brain's central relay station, orchestrating sensory processing and cognitive functions. However, how thalamic function depends on internal and external states, is not well understood. A comprehensive understanding would necessitate the integration of single cell dynamics with their collective behaviour at population level. For this we propose a biologically realistic mean-field model of the thalamus, describing the dynamics of thalamocortical relay neurons (TC) and thalamic reticular neurons (RE). With this we perform a multi-scale study of thalamic responsiveness and its dependence on cell, synaptic, and brain states. Based on existing single-cell experiments we show that: The transition between tonic and burst states of TC cells acts as a linear-nonlinear switch in the thalamic response to stimuli. In the awake state, sensory stimuli generate a linear thalamic response while cortical input to the thalamus generates a nonlinear response. And that stimulus response and information transfer are controlled by cortical input and synaptic noise, which both suppress responsiveness. In addition, synaptic noise acts as an equalizer between thalamic response at different states and diffuses state transitions between tonic-bursting and awake-sleep states. Finally, the model replicates spindles within a sleep-like state, reducing its responsiveness. This study results in new insight on the brain-state dependent function of the thalamus. In addition, the development of a novel thalamic mean-field model provides a new a tool for incorporating detailed thalamic dynamics in large scale brain simulations. This will help to bridge the gap between computational neuroscience, behavioral experiments, and the study of brain diseases.

Ablation of Mitochondrial RCC1-L Induces Nigral Dopaminergic Neurodegeneration and Parkinsonian-like Motor Symptoms
Mitochondrial dysfunction has been linked to both idiopathic and familial forms of Parkinsons disease (PD). We have previously identified RCC1-like (RCC1L) as a protein of the inner mitochondrial membrane important to mitochondrial fusion. Herein, to test whether deficits in RCC1L mitochondrial function might be involved in PD pathology, we have selectively ablated the Rcc1l gene in the dopaminergic (DA) neurons of mice. A PD-like phenotype resulted that includes progressive movement abnormalities, paralleled by progressive degeneration of the nigrostriatal tract. Experimental and control groups were examined at 2, 3-4, and 5-6 months of age. Animals were tested in the open field task to quantify anxiety, exploratory drive, locomotion, and immobility; and in the cylinder test to quantify rearing behavior. Beginning at 3-4 months, both female and male Rcc1l knockout mice show rigid muscles and resting tremor, kyphosis and a growth deficit compared with heterozygous or wild type littermate controls. Rcc1l knockout mice begin showing locomotor impairments at 3-4 months, which progress until 5-6 months of age, at which age the Rcc1l knockout mice die. The progressive motor impairments were associated with progressive and significantly reduced tyrosine hydroxylase immunoreactivity in the substantia nigra pars compacta (SNc), and dramatic loss of nigral DA projections in the striatum. Dystrophic spherical mitochondria are apparent in the soma of SNc neurons in Rcc1l knockout mice as early as 1.5-2.5 months of age and become progressively more pronounced until 5-6 months. Together, the results reveal the RCC1L protein to be essential to in vivo mitochondrial function in DA neurons. Further characterization of this mouse model will determine whether it represents a new model for in vivo study of PD, and the role of RCC1L as a risk factor that might increase PD occurrence and severity in humans.

Intermediate CA1 is Required for Object-in-Place Recognition Memory in Mice
Many behaviors that are essential for survival, such as retrieving food, finding shelter and locating predator cues, rely on forming effective associations between the identity and location of spatial elements. This identity-location association is commonly assessed in rodents using spontaneous object-in-place (OiP) recognition memory tasks. OiP recognition memory deficits are seen in autism spectrum disorder, schizophrenia, and are used to detect early onset of Alzheimer's disease. These deficits are replicated in animal models of neurodevelopmental, neurodegenerative and chromosomal disorders. Mouse models have been widely adopted in behavioral and systems neuroscience research for their ease of genetic manipulations, and yet very few studies have successfully assessed OiP recognition memory or its neural correlates in mice. To address this limitation, we first established that adult C57/129J and C57BL/6J male and female mice are able to successfully perform the two-object, but not the four-object version of the spontaneous OiP recognition task, with retention intervals of five minutes and one hour. Next, using chemogenetic inhibition, we found that two-object OiP requires the activity of the intermediate CA1 (iCA1) subregion of the hippocampus, but not the medial prefrontal cortex or iCA1-medial prefrontal cortex connections. Our data identify hippocampal subregion specialization in the successful assessment of OiP recognition memory in mice, expanding our understanding of the neural basis of spatial memory processing.