bioRxiv Subject Collection: Neuroscience's Journal

Spatial isoform sequencing at sub-micrometer single-cell resolution reveals novel patterns of spatial isoform variability in brain cell types
Spatial long-read technologies are becoming more common but lack nanometer- and therefore often single-cell resolution. This leaves the question unanswered whether spatially variable isoforms represent spatial variability within one cell type or differences in cell-type abundance between different regions. Here, we develop Spl-ISO-Seq2 with 220nm spot size and 500nm resolution, and the accompanying software packages Spl-IsoQuant-2 and Spl-IsoFind and apply it to the adult mouse brain. We compare spatial variability within a fixed cell type by examining (a) differential isoform abundance between known brain regions and (b) spatial isoform patterns that do not align with predefined regions. The former reveals larger numbers of spatial isoform differences, e.g. Rps24 in oligodendrocytes. For the previously appreciated gene with spatially-variable isoforms Snap25, we can now show that this variability exists in excitatory neurons. However, the latter approach reveals patterns that the former cannot conceptually model, e.g., Tnnc1 in excitatory neurons. Taken together, our experimental and analytical methods enrich spatial transcriptomics with a so-far elusive isoform view of spatial variation for individual cell types.

Stimulation modulates cell assemblies linked with gene networks in the human temporal cortex ex vivo
Deep brain stimulation of the temporal cortex can enhance learning and memory in the face of cognitive impairment. Despite the potential of such therapies, the neural and genetic mechanisms underlying the effect of stimulation on human brain circuits are not understood. To explicate direct mechanisms of neural modulation elicited by brain stimulation, we developed an ex vivo approach utilizing microelectrode array stimulation and recording of resected temporal cortex from neurosurgical patients. We find that stimulation preferentially increases firing rates in pyramidal cells compared to interneurons and also strengthens cell assemblies. Using single cell multiomics, we link the observed physiological changes to cell type-specific gene expression patterns. We detail gene regulatory networks that indicate preferential involvement of specific excitatory neuron subtypes and the response of non-neurons. We conclude that the overall impact of stimulation on the human temporal cortex is activation of specific excitatory neurons and enhanced cell assembly activity, and that these changes are supported by gene networks involving immediate early, synaptic, and ion channel genes. Our findings establish a foundation to identify targetable cell type-specific genetic signatures that may be harnessed for therapeutic benefit in future neuromodulation strategies.

A scalable, all-optical method for mapping synaptic connectivity with cell-type specificity
Single-cell transcriptomics has uncovered the enormous heterogeneity of cell types that compose each region of the mammalian brain, but describing how such diverse types connect to form functional circuits has remained challenging. Current methods for measuring the probability and strength of cell-type specific connectivity motifs principally rely on low-throughput whole-cell recording approaches. The recent development of optical tools for perturbing and observing neural circuit activity, now notably including genetically encoded voltage indicators, presents an exciting opportunity to employ optical methods to greatly increase the throughput with which circuit connectivity can be mapped physiologically. At the same time, advances in spatial transcriptomics now enable the identification of cell types in situ based on their unique gene expression signatures. Here, we demonstrate how long-range synaptic connectivity can be assayed optically with high sensitivity, high throughput, and cell-type specificity. We apply this approach in the motor cortex to examine cell-type-specific synaptic innervation patterns of long-range thalamic and contralateral input onto more than 1000 motor cortical neurons. We find that even cell types occupying the same cortical lamina receive vastly different levels of synaptic input, a finding which was previously not possible to uncover using lower-throughput approaches that can only describe the connectivity of broad cell types.

Multiple autism genes influence GABA neuron remodeling via distinct developmental trajectories
Variation in over 100 genes are now associated with increased risk for autism and related neurodevelopmental condition, but how this variation results in distinct and overlapping behavioral changes is still not well understood. Recent efforts have focused on screening many autism genes at once for functional and phenotypic convergence, and identified subsets that are crucial for many early steps of neurodevelopment. Few studies have screened later steps of neurodevelopment, circuit function, circuit plasticity, or behaviors. We screened twenty conserved autism-associated genes for impact on experience-dependent neuron remodeling in C. elegans. Loss of unc-44/ANK2, set-4/KMT5B, daf-18/PTEN, gap-2/SYNGAP1, and chd-1/CHD8 increased, while CACNA2D3/unc-36 decreased, neurite outgrowth of the GABAergic DVB neuron in adults. Although daf-18/PTEN, set-4/KMD5B, and unc-44/ANK2 had convergent phenotypes, they arise from distinct temporal trajectories with differential impact on DVB pre-synaptic morphology. Screening for the DVB regulated spicule protraction behavior identified multiple autism genes involved, but only unc-44/ANK2 and CACNA2D3/unc-36 were shared between screens. Application of a metric geometry computational framework (CAJAL) to the DVB morphology dataset identified 5 additional genes that impact DVB morphology, including unc-2/CACNA1A and unc-10/RIMS1, which also significantly impacted behavior. This work defines new regulators and molecular mechanisms of experience-dependent neuron remodeling and circuit plasticity, and further links these processes with conserved autism genes. It also demonstrates the utility of using intact, behavior generating circuits in C. elegans, to screen for novel roles for conserved autism genes.

Melanocortin-responsive Kiss1 neurons of the arcuate nucleus drive energy expenditure through glutamatergic signaling to the dorsomedial hypothalamus.
Energy expenditure (EE) is essential for metabolic homeostasis, yet its central regulation remains poorly understood. Here, we identify arcuate Kiss1 neurons as key regulators of EE in male mice. Ablation of these neurons induced obesity, while their chemogenetic activation increased brown adipose tissue (BAT) thermogenesis without affecting food intake. This action is mediated by glutamatergic projections from Kiss1ARC neurons to CART/Lepr-expressing neurons in the dorsomedial hypothalamus, which activate the raphe pallidus-BAT pathway. CRISPR-mediated deletion of the vesicular glutamate transporter 2 (Vglut2) from Kiss1ARC neurons replicated the obesogenic effect. Furthermore, deletion of the melanocortin 4 receptor (MC4R) from Kiss1 neurons resulted in obesity, reduced energy expenditure and impaired thermogenesis. Optogenetic stimulation of pro-opiomelanocortin (POMC) fibers evoked inward currents in Kiss1 neurons, that were attenuated by MC4R antagonism. Our findings reveal a previously unrecognized neural circuit that mediates melanocortin action on energy expenditure, offering new insights into central mechanisms of metabolic control.

Misinterpreting electrophysiology in human cognitiveneuroscience
An axiomatic view in contemporary neuroscience is that EEG components such as eventrelated brain potentials (ERPs) and oscillations are directly interpretable as manifestations of biological processes that support sensory, motor, and cognitive constructs of interest. This premise justifies and propels research programs in laboratories worldwide, but with a substantial social and economic cost, warranted by the potential for basic-science discovery and the resulting bench-to-bedside transfer for health and disease. But a different premise would be more fruitful. This article proposes that EEG components in psychophysiological experiments relate to cognition indirectly through their more direct relationship with oculomotor action. The common experimental design that includes a baseline ocular fixation period preceding stimulus presentation provides an excellent template with which to develop the present proposal. Electrophysiological and eye-tracking evidence (3 published and 3 new data sets: 6 experiments, Ntotal = 204, in the context of face and affective picture viewing, reading, listening, rest, and microsleep) demonstrates how and why common conclusions, and reliance on them in clinical practice/treatment efficacy and drug development studies, are at best premature. Results indicate that the oculomotor system plays a mediating role between such EEG phenomena and cognition. Present evidence supports a complementary view of how EEG can shape the development of a broader thought horizon in psychophysiological theory and practice.

Gamma - Theta - Spike Coupling Coordinates Sequential Memory in Human MTL
It has been suggested that hierarchical synchronization of theta and gamma oscillations coordinates neural activity during sequence memory. Yet, the role of gamma oscillations and their interaction with theta and single - unit activity (SUA) has not been directly examined in humans. We analysed simultaneous micro wire recordings of SUA (N = 1417 units) and local field potentials (N = 917 channels) from the medial temporal lobe (MTL) of epilepsy patients performing a visual multi-item sequence memory task. During encoding, both spiking activity and gamma power contained item - specific information and were temporally coupled. During memory maintenance, stimulus-specific gamma and spiking were structured in brief burst periods during which both signals were tightly synchronized and preferentially aligned to similar theta phases predictive of sequential stimulus position. These findings demonstrate that theta - gamma - spike interactions support a phase - based multiplexed code for sequential memories in the human MTL.

Anterior cingulate folding pattern is altered in autism spectrum disorder
Neuroimaging research has identified focal differences in the cerebral cortex of individuals with autism spectrum disorder (ASD), particularly in the cortical folds (sulci) within higher-level association cortices. The present study investigated the sulcal patterning and morphology of the anterior cingulate cortex (ACC) in individuals with ASD compared to neurotypical (NT) individuals for the first time. We used neuroimaging data from 50 NT and 50 ASD participants. All participants were under 20 years old and male. The two groups were age-matched. Using established criteria and cortical reconstructions generated from each participants T1-weighted magnetic resonance imaging scans with FreeSurfer, we identified the defining sulcal feature of ACC, the variably present paracingulate sulcus (PCGS): its presence in the left and right hemispheres, and asymmetry in PCGS presence between hemispheres. Finally, multiple quantitative morphological features (length, depth, and cortical thickness mean and standard deviation) were extracted from the PCGS using FreeSurfer tools. Analyses revealed that NT participants were more likely to have asymmetrical PCGS patterns than ASD participants (controlling for age and scanner site). However, none of the quantitative morphological features differed between groups. These findings suggest the presence of a variation in the prenatal neurodevelopment of ACC in young males with ASD; however, further research is necessary to uncover the role of this observed difference in the pathogenesis of ASD. The present study also adds to the growing literature implicating variations in PCGS patterning as a trait marker across multiple disorders.

Lay SummaryThis study found that young males with autism spectrum disorder (ASD) show less hemispheric asymmetry in the presence of a notoriously variable brain structure (paracingulate sulcus (PCGS)) compared to neurotypical individuals. Considering that this feature of the PCGS develops before birth, the reduced asymmetry may indicate focal differences in brain development in ASD. These findings further enhance our understanding of the neurodevelopmental characteristics of ASD and highlight growing findings indicating that the PCGS may be a useful transdiagnostic marker for various psychiatric conditions.

Neural correlates of evidence accumulation in social-affective decision-making under perceptual ambiguity
Evidence accumulation models have been successfully applied to decision-making in sensory and cognitive domains; however, it remains unclear how this process is regulated when perceptual ambiguity arises from social-affective content. Here, we integrate computational modeling with multimodal neuroscience to characterize how perceptual ambiguity in emotion judgment shapes decision dynamics. Participants viewed perceptually ambiguous stimuli - morphed images of two categories, such as happy and fearful facial expression - and made binary categorization decisions. Using drift diffusion modeling (DDM), we first demonstrate that drift rate, a key index of evidence accumulation, decreases as perceptual ambiguity increases. Scalp electroencephalography (EEG) data reveal that the magnitude of the late positive potential (LPP) tracks the speed of evidence accumulation in both emotional and non-emotional stimulus categories, but only when the ambiguous dimension is relevant to the categorization decision. Similar to LPP magnitude, single-unit recordings from the dorsomedial prefrontal cortex (dmPFC) and amygdala show that neuronal firing rates in both regions also encode drift rate during the emotion categorization task. Moreover, fMRI-based functional connectivity reveals that the strength of connectivity between the amygdala and dmPFC correlates with individual differences in drift rate. To establish the causal role of the dmPFC, we applied anodal transcranial direct current stimulation (tDCS) targeting the dmPFC in patients with schizophrenia and found that stimulation enhanced evidence accumulation speed in emotion categorization under perceptual ambiguity. These findings identify a distributed corticolimbic circuit that dynamically modulates evidence accumulation during social-affective decision-making under perceptual ambiguity. Our results bridge social-affective and perceptual neuroscience, offering a translational framework for understanding emotion recognition and decision-making impairments.

NERVE GROWTH FACTOR IS SUFFICIENT TO CAUSE MULTIPLE OSTEOARTHRITIS-RELEVANT PATHOLOGICAL FEATURES IN NAIVE MURINE KNEE JOINTS
BACKGROUNDNerve growth factor (NGF), a key mediator of pain and inflammation, is increased in joints with osteoarthritis (OA). Neutralizing NGF with monoclonal antibodies has shown analgesic effects in painful knee OA, but clinical development was stopped due to side effects in the joints. Knowledge about the biological effects of long-term exposure of joint tissues to NGF is limited. Therefore, we aimed to explore the effects of repeated intra-articular (IA) injections of NGF into the knee joints of healthy mice on pain and sensitization, as well as joint innervation and structure.

METHODSWe conducted five experiments in male C57BL/6 mice. In Experiment 1, NGF (50ng or 500ng) or vehicle was injected IA into the knee of naive wildtype (WT) mice, twice a week for 4 weeks. We assessed knee swelling, knee hyperalgesia and histopathology. In Experiment 2, mice were injected with 500ng NGF or vehicle, twice a week for 4 weeks and microCT of the knee was performed. In Experiment 3, NaV1.8-tdTomato reporter mice were injected with 500ng NGF or vehicle, twice a week for 4 weeks, and joint innervation was assessed. In Experiment 4, WT mice received 500ng NGF or vehicle twice a week for 4 weeks and were used for single cell RNA sequencing (scRNAseq) of the synovium. In Experiment 5, L3-L5 DRGs of mice that received 3 IA injections of 500ng NGF or vehicle twice a week were used for bulk RNA sequencing.

RESULTSRepeated bi-weekly IA injections of NGF caused knee hyperalgesia in naive mice. NGF caused dose-dependent knee swelling, synovial pathology, increased bone mineral density and trabecular bone thickness in the medial subchondral bone, growth of pre-osteophytes in the medial compartment, but no cartilage degeneration. NGF injection caused sprouting of NaV1.8+ neurons in the medial but not the lateral synovium. ScRNAseq of the synovium revealed upregulated genes related to neuronal sprouting, synovial fibrosis and ossification, confirming histopathological findings. Bulk RNA seq of DRG showed upregulated pathways related to axonal growth.

CONCLUSIONSIn healthy mouse knees, NGF induced mechanical sensitization, synovitis, neoinnervation in the medial synovium, subchondral bone changes and pre-osteophyte growth in the medial compartment, thus capturing many pathological changes observed in OA, except cartilage damage.

Dynamic feedforward and feedback influences on left ventral occipitotemporal cortex: evidence from word and pseudoword reading
Left ventral occipitotemporal cortex (vOT) is crucial in reading, yet its functional role remains debated. Competing theories paint it as either a prelexical feedforward hub or a bidirectional interface between sensory and higher-order linguistic systems. To address the debate, we investigated the temporal and spectral dynamics of information flow involving left vOT during visual word and pseudoword reading using magnetoencephalography (MEG). The pseudowords varied in the degree to which they orthographically resembled real words. By combining two directed connectivity metrics, i.e., phase slope index (PSI) and Granger causality (GC), we converged on a hybrid model of left vOT function that reconciles the competing perspectives. Feedforward connectivity from low-level visual areas to vOT emerged at around 100 ms post stimulus similarly across all conditions, spanning a wide frequency range. Subsequently, feedforward orthographic information flowed from left vOT to higherorder areas, especially left superior temporal cortex (ST), at the low gamma band. This flow strength was modulated by word-likeness, being stronger for real words and word-like pseudowords than complete pseudowords. Conversely, feedback flow from left ST to vOT was observed in the low beta band for pseudowords, and occurred later for word-like than complete pseudowords. This indicates that greater processing demands modulate the direction of information flow, necessitating top-down linguistic constraints to facilitate reading. Our findings clarify the functional role of left vOT and explain when and why its connectivity may show as feedforward or bidirectional depending on time and task.

Learning, sleep replay and consolidation of contextual fear memories: A neural network model
Contextual fear conditioning is an experimental framework widely used to investigate how aversive experiences affect the valence an animal associates with an environment. While the initial formation of associative context-fear memories is well studied -- dependent on plasticity in hippocampus and amygdala -- the neural mechanisms underlying their subsequent consolidation remain less understood. Recent evidence suggests that the recall of contextual fear memories shifts from hippocampal-amygdalar to amygdalo-cortical networks as they age. This transition is thought to rely on sleep. In particular, neural replay during hippocampal sharp-wave ripple events seems crucial, though open questions regarding the involved neural interactions remain. Here, we propose a biologically informed neural network model of context-fear learning. It expands the scope of previous models through the addition of a sleep phase. Hippocampal representations of context, formed during wakefulness, are replayed in conjunction with cortical and amygdalar activity patterns to establish long-term encodings of learned fear associations. Additionally, valence-coding synapses within the amygdala undergo overnight adjustments consistent with the synaptic homeostasis hypothesis of sleep. The model reproduces experimentally observed phenomena, including context-dependent fear renewal and time-dependent increases in fear generalisation. Few neural network models have addressed fear memory consolidation and to our knowledge, ours is the first to incorporate a neural mechanism enabling it. Our framework yields testable predictions about how disruptions in synaptic homeostasis may lead to pathological fear sensitization and generalisation, thus potentially bridging computational models of fear learning and mechanisms underlying anxiety symptoms in disorders such as PTSD.

Mapping the projectional architecture of the mouse midbrain dopaminergic system using cell type-specific barcoding
Brain-wide neural circuits are formed by the diverse axonal branching patterns of many individual neurons. Here we introduce POINTseq (projections of interest by sequencing), a high-throughput and user-friendly barcoded connectomics method that uses cell type specific barcoding and sequencing to rapidly map single-cell projections of a cell type of interest for thousands of neurons per animal. POINTseq leverages pseudotyping of Sindbis virus and a specific alphavirus-cellular receptor pair to make Sindbis infections cell type specific. It thus integrates MAPseq-style high-throughput barcoded projection mapping with the established viral-genetic neural circuit analysis toolbox. We validated POINTseq by mapping genetically and projection-defined cell populations in the mouse motor cortex. We then applied POINTseq to midbrain dopaminergic neurons and reconstructed the brain-wide single-cell projections of 3,813 dopaminergic neurons in ventral tegmental area (VTA) and substantia nigra pars compacta (SNc). We define over 30 connectomic cell types, vastly exceeding the known diversity of dopaminergic cell types, and identify stereotyped projection motifs that may mediate parallel dopamine signaling. This data constitutes the anatomical substrate on which the diverse functions of dopamine in the brain are built.

HIGHLIGHTSO_LIWe develop POINTseq, which uses pseudotyped Sindbis virus and cell type-specific expression of a viral receptor for cell type-specific barcoding.
C_LIO_LIPOINTseq enables massively multiplexed single-cell projection mapping of cell types of interest.
C_LIO_LIWe map the brain-wide projections of 3,813 individual VTA and SNc dopaminergic neurons.
C_LIO_LIVTA and SNc dopaminergic neurons form over 30 connectomic cell types.
C_LIO_LIProjections organize into stereotyped motifs that may mediate parallel dopamine signalling.
C_LI

Neuromechanical Simulation with NEURON and MuJoCo
In computational neuroscience, simulation platforms generally do not have adequate tools to model the brain, body and environment simultaneously. We demonstrate a method for simulating neuromechanical models using a novel combination of widely used software platforms: NEURON and MuJoCo. Different neural models are used to control a realistic musculoskeletal model in both open-loop and closed-loop configurations. Three models are presented: (1) an open-loop model using simple spiking neurons from the NEURON model library; (2) an open-loop model using realistic, spiking motoneurons; and (3) a closed-loop central pattern generator with feedback from the physics engine.

Scanning for Representation: A Scoping Review of Racial and Ethnic Diversity in MRI Studies of the Maternal Brain
A growing number of magnetic resonance imaging (MRI) studies are examining brain changes across pregnancy and early motherhood, gaining fundamental insight into the neural adaptations of motherhood, with critical clinical and policy implications for supporting mother, child, and family unit. As the field takes off, now is the time to take stock of the current literature and neuroscience practices, to ensure that the field is based on studies that are robust, representative, and transparent. Here, we conducted a scoping review to understand the racial and ethnic diversity of participants reported in MRI studies of the maternal brain, guided by the Joanna Briggs Institute methodology. Our findings highlight three key issues in the 185 identified studies of the maternal brain using MRI: (1) the widespread underreporting of participant racial and ethnic data, with only 38.38% of studies reporting race and/or ethnicity demographics; (2) the overrepresentation of white participants, with 46.83% of the samples that report race and/or ethnicity identifying as white/Caucasian; and (3) the disproportionate geographical locations of studies, with 68.65% of studies from North America or Europe and Central Asia. These findings raise concerns about the generalizability of existing research beyond WEIRD (western, educated, industrialized, rich and democratic) populations, and underscore the urgent need for concerted structural change in neuroscience research practices. While identifying a lack of diversity is only the first step, this scoping review serves as a call to action for greater representation in future research, for our own research group as well as others.

Comparative Systematic Analysis of Gray Matter BiophysicalModels on a Public Dataset
Biophysical models of diffusion tailored to characterize gray matter (GM) microstructure are gaining traction in the neuroimaging community. NEXI, SMEX, SANDI, and SANDIX represent recent efforts to incorporate different microstructural features,such as soma contributions and inter-compartment exchange, into the diffusion MRI (dMRI) signal. In this work, we present a comparative evaluation of these four gray matter models on a single, publicly available in vivo human dataset, the Connectome Diffusion Microstructure Dataset (CDMD), acquired with two diffusion times. Using the open-source Gray Matter Swiss Knife toolbox, we estimate cortical microstructure metrics in 26 healthy subjects and evaluate goodness of fit, anatomical patterns and consistency with previous studies. CDMD data yielded GM parameter estimates consistent with values reported in previous studies. This retrospective cross-model analysis establishes the feasibility of estimating exchange models from only two diffusion times and highlights trade-offs in biological specificity, model complexity, and fitting robustness, critical considerations when choosing a model for future clinical and research applications

Asymmetric increase in episodic and procedural memory interference in older adults
In younger adults, newly formed procedural memories are weakened by the subsequent formation of episodic memories (E[->]P interference) and vice versa (P[->]E interference; "cross-memory interference"). Older adults experience significant decline in episodic memory but maintain relatively intact procedural memory. This asymmetric decline in memory may also cause an asymmetric change in cross-memory interference compared to younger adults. For example, older adults may experience a significant increase in one type of cross-memory interference while leaving the other unchanged. Additionally, decline in episodic memory may cause E[->]P interference to either increase or decrease depending on how the episodic and procedural memory systems interact. However, no study to our knowledge has compared cross-memory interference between younger and older adults. We investigated cross-memory interference in younger and older adults by measuring E[->]P (Exp. 1) and P[->]E (Exp. 2) interference in 40 younger (18-40 years old) and 40 older ([≥] 55 years old) adults. Compared to younger adults, the results show that older adults experience significantly stronger E[->]P interference while P[->]E interference was statistically indistinguishable between groups. These results confirm that older adults experience an asymmetric increase in cross-memory interference and suggest that the increase in E[->]P interference is related to the asymmetric decline in episodic memory relative to procedural memory.

Early transcriptional signatures of MeCP2 positive and negative cells in Rett syndrome
Rett syndrome (RTT) is an X-linked neurological disorder caused by MECP2 mutations. Like other X-linked disorders, RTT patients have sex-specific differences in clinical presentation due to distinct cellular environments, where females have ~50% of cells expressing either a mutant or wild-type copy of MECP2 (mosaic) and males have 100% of cells expressing a mutant MECP2 (non-mosaic). Typical RTT females have a short window of normal early development until ~6-18 months, followed by regression and progressive decline, whereas neonatal encephalopathy is more likely in RTT males. How these sex-specific differences in cellular context contribute molecularly to RTT pathogenesis, particularly in the presymptomatic stages of RTT females, remains poorly understood. Here, we profiled the hippocampal transcriptomes of female (Mecp2+/-) and male (Mecp2-/y) RTT mice at early timepoints using both bulk and single-nucleus RNA-seq, including sorted MeCP2 positive (MeCP2+) and MeCP2 negative (MeCP2-) neurons in female mice. We identified a core disease signature consisting of 12 genes consistently dysregulated only in MeCP2- cells across RTT models. Moreover, we uncovered non-cell-autonomous effects exclusively in female MeCP2+ excitatory neurons, but not inhibitory neurons, suggesting excitatory circuits are more vulnerable early in the mosaic RTT environment. The single-nuclei data also revealed that a previously underappreciated MeCP2- interneuron subtype had the most transcriptional dysregulation in both male and female RTT hippocampi. Together, these data highlight the different effects of MeCP2 loss on excitatory and inhibitory circuits between the mosaic and non-mosaic environment that appear early in RTT pathogenesis.