Dendrites transform local electrical activity into intracellular Ca2+ signals that drive plasticity1,2, yet the voltage→Ca2+ mapping during natural behavior remains poorly defined. Here, we measure this transfer function via simultaneous voltage and Ca2+ imaging throughout the dendritic arbors of hippocampal CA2 pyramidal neurons in behaving mice. Dendritic Ca2+ exhibited a hierarchical activation pattern dominated by back-propagating action potentials: simple spikes primarily drove somatic and proximal Ca2+, whereas complex spikes produced larger somatic Ca2+ signals and propagated farther into distal dendrites, sometimes in a branch-selective manner. Dendrite-restricted co-activation of voltage and Ca2+ without concurrent somatic events was rare. A biophysics-inspired model accurately predicted local Ca2+ transients from local voltage waveforms. Our data and model provide a quantitative understanding of when – and why – dendritic Ca2+ signals in CA2 pyramidal cells arise during behavior.

The cytoskeleton is a key mediator of mechanical interactions in cells, but specific contributions of septins remains unclear. Septins preferentially localize with a subset of actin stress fibers positioned under the nucleus, where they are situated between the membrane and stress fibers. Removing the nucleus from the cell results in the loss of these subnuclear septin-decorated stress fibers. Surprisingly, however, their formation can be rescued using a large glass bead in place of the nucleus. Similarly, applying a compressive force to the cell via confinement, whether externally or through internally generated actomyosin forces, results in increased septin accumulation in regions where the nucleus engages the cell cortex. Finally, loss of septin filaments via knockdown of SEPT7 increases the likelihood of nuclear membrane rupture during confinement. Together these data suggest that septins act as a dynamic mechanosensitive protective mechanism to buffer mechanical forces on the nucleus.

Spiking neural networks (SNNs) offer a promising paradigm for modeling brain dynamics and developing neuromorphic intelligence, yet an online learning system capable of training rich spiking dynamics over long horizons with low memory footprints has been missing. Existing online approaches either incur quadratic memory growth, sacrifice biological fidelity through oversimplified models, or lack end-to-end automated tooling. Here, we introduce BrainTrace, a model-agnostic, linear-memory, and automated online learning system for spiking neural networks. BrainTrace standardizes model specification to encompass diverse neuronal and synaptic dynamics; implements a linear-memory online learning rule by exploiting intrinsic properties of spiking dynamics; and provides a compiler that automatically generates optimized online-learning code for arbitrary user-defined models. Across diverse dynamics and tasks, BrainTrace achieves strong learning performance with a low memory footprint and high computational throughput. Critically, these properties enable online fitting of a whole-brain-scale Drosophila SNN that recapitulates region-level functional activity. By reconciling generality, efficiency, and usability, BrainTrace establishes a foundation for spiking network modeling at scale.

The first step to probing any potential interaction between two biomolecules is to determine their spatial association. In other words, if two biomolecules localize similarly within a cell, then it is plausible they could interact. Traditionally, this is quantified through various colocalization metrics. These measures infer this association by estimating the degree to which fluorescent signals from each biomolecule overlap or correlate. However, these metrics are, at best, proxies, and they depend strongly on various experimental choices. Here, we define a new strategy that leverages multispectral imaging and phasor analysis, termed the phasor mixing coefficient (PMC). The PMC measures the precise mixing of fluorescent signals in each pixel. We demonstrate how the PMC captures complex biological subtlety by offering two distinct values, a global measure of overall color mixing and the homogeneity thereof. We additionally show that the PMC exhibits less sensitivity to signal-to-noise ratio, intensity threshold and background signal compared to canonical methods. Moreover, this method provides a means to visualize color mixing at each pixel. We show that the PMC offers users a nuanced and robust metric to quantify biological association.

To successfully forage for food, animals must balance the energetic cost of searching for food sources with the energetic benefit of exploiting those sources. While the Marginal Value Theorem provides one normative account of this balance by specifying that a forager should leave a food patch when its energetic yield falls below the average yield of other patches in the environment, it assumes the presence of other readily reachable patches. In natural settings, however, a forager does not know whether it will encounter additional food patches, and it must balance potential energetic costs and benefits accordingly. Upon first encountering a patch of food, it faces a decision of whether and when to leave the patch in search of better options, and when to return if no better options are found. Here, we explore how a forager should structure its search for new food patches when the existence of those patches is unknown, and when searching for those patches requires energy that can only be harvested from a single known food patch. We identify conditions under which it is more favorable to explore the environment in several successive trips rather than in a single long exploration, and we show how the optimal sequence of trips depends on the forager’s beliefs about the distribution and nutritional content of food patches in the environment. This optimal strategy is well approximated by a local decision that can be implemented by a simple neural circuit architecture. Together, this work highlights how energetic constraints and prior beliefs shape optimal foraging strategies, and how such strategies can be approximated by simple neural networks that implement local decision rules.

Across social species, social touch enhances well-being and reduces pain — two seemingly distinct benefits that enhance survival. Yet where and how the nervous system integrates these functions, and whether a single mechanism could serve both, remains unknown. Here we show that massage triggers oxytocin release, which shapes both pain and touch reward at the earliest stage of central processing — the spinal cord — through a single, state-dependent circuit mechanism. We report that in humans, massage enhances well-being, effects that correlate with endogenous oxytocin release. In mice, gentle touch activates hypothalamic oxytocin neurons that project directly to the spinal dorsal horn. Genetic manipulation of spinal oxytocin circuits alters behavioral responses to both gentle touch and noxious stimuli. Spinal calcium imaging and slice electrophysiology reveal that oxytocin acts on both excitatory and inhibitory spinal neurons to sculpt the relative activity of spinal ascending systems that convey both social touch and pain to the brain. Extending these findings to humans, we show that oxytocin receptors are also expressed on spinal excitatory and inhibitory neurons, and that endogenous oxytocin during massage correlates with altered spinal touch processing. Thus, spinal oxytocin signaling provides an evolutionarily conserved mechanism for the therapeutic benefits of massage.

We present a method, open-source software, and experiments which embed arbitrary deformation vector fields produced by any method (e.g., ANTs or VoxelMorph) in the Large Deformation Diffeomorphic Metric Mapping (LDDMM) framework. This decouples formal diffeomorphic shape analysis from image registration, which has many practical benefits. Shape analysis can be added to study designs without modification to already chosen image registration methods and existing databases of deformation fields can be reanalyzed within the LDDMM framework without repeating image registrations. Pairwise time series studies can be extended to full time series regression with minimal added computing. The diffeomorphic rigor of image registration methods can be compared by embedding deformation fields and comparing projection distances. Finally, the added value of formal diffeomorphic shape analysis can be more fairly evaluated when it is derived from and compared to a baseline set of deformation fields. In brief, the method is a straightforward use of geodesic shooting in diffeomorphisms with a deformation field as the target, rather than an image. This is simpler than the image registration case which leads to a faster implementation that requires fewer user derived parameters.

Messenger RNA (mRNA) transfection enables rapid, transient protein expression without nuclear entry, providing a powerful alternative to DNA or viral delivery in post-mitotic and otherwise difficult-to-transfect cells. Although in vitro transcribed (IVT) mRNAs have revolutionized therapeutic applications, their adoption in experimental biology remains limited by challenges in synthesis, variability across cell types, and concerns about cytotoxicity. Here, we define design principles that maximize IVT mRNA performance across diverse cellular and organismal systems. Through systematic comparison of capping strategies and base modifications, including N1-methyl-pseudouridine, 5-methylcytidine, and 5-methoxyuridine, we identify modifications that enhance translation while minimizing activation of cellular stress responses. Optimized transcripts drive robust protein expression within four hours, persist for up to one week, and support multiplexed expression of structurally and functionally distinct proteins in mammalian cells, including cancer cell lines, iPSC-derived systems, primary cells, and organoids, as well as in vivo in zebrafish embryos and in less genetically tractable models such as Danionella cerebrum and sea urchin embryos. To further expand accessibility for community use, we developed mRNAbow, a platform for generating low-toxicity mRNAs encoding organelle-targeted fluorescent proteins and biosensors for multiplex imaging, with corresponding plasmids made publicly available. Together, these advances establish a generalizable framework for IVT mRNA design and expand experimental access to synthetic mRNA technologies for dissecting cellular architecture and dynamics.

Understanding how neural signals control muscle activity during behavior is a key challenge in motor neuroscience. To this end, recent advances in intramuscular multielectrode arrays have enabled high-quality multichannel recordings of many motor unit action potentials (MUAPs) in freely moving subjects. However, identifying individual MUAP events within multichannel recordings is a significant challenge for existing spike sorting methods, which are typically optimized for identifying action potentials from neurons in the brain. To overcome this challenge, we developed the Enhanced Motor Unit sorter (EMUsort), an extension of Kilosort4 (KS4) that achieves high-performance MUAP spike sorting. We applied EMUsort to high-resolution intramuscular recordings from rat forelimb during locomotion and monkey forelimb during a reaching task. EMUsort improves upon prior methods by addressing key challenges encountered with MUAP datasets, including: 1) long time delays across electrodes due to propagation along muscle fibers, 2) more complex waveform shapes compared to neuronal action potentials, and 3) a high degree of MUAP overlap due to cumulative motor unit recruitment. We compared EMUsort to existing spike sorting methods quantitatively using simulated datasets that closely emulated the rat and monkey datasets we recorded. EMUsort provided median error rate reductions of 67.5% and 49.9% during periods of high motor unit activation for the rat and monkey datasets, respectively. In sum, EMUsort provides a substantial improvement to MUAP spike sorter accuracy, especially during regions of high MUAP overlap, in an easy-to-use software package.

Single-walled carbon nanotubes (SWCNTs) functionalized with single-stranded DNAs can function as near-infrared nanosensors for molecular analytes. However, predicting which analytes elicit strong optical responses for specific nanosensors remains challenging. We developed machine learning (ML) models to predict analyte-induced fluorescence changes in a DNA–SWCNT dopamine nanosensor. Using a data set of 63 small molecules sampling chemical space around dopamine, we encoded analytes with RDKit fingerprints, with or without HOMO and LUMO energies, and applied principal component analysis to identify structural motifs associated with optical response strength. We trained support vector regression and classification models using two strategies: ensembles of 200 models and cross-validation. Regression models achieved mean R2 values of 0.2–0.4, with cross-validation outperforming ensembles, while classifiers reached mean F1 scores of ∼0.8. Cross-validation performed best for predictions on a blind set of 21 molecules. These findings show that ML can capture structure–response patterns in modest data sets and guide in silico DNA–SWCNT nanosensor design.