Cell Reports Methods

In general, animal behavior can be described as the neuronal-driven sequence of reoccurring postures through time. Current technologies enable offline pose estimation with high spatio-temporal resolution, however to understand complex behaviors, it is necessary to correlate the behavior with neuronal activity in real-time. Here we present DeepLabStream, a highly versatile, closed-loop solution for freely moving mice that can autonomously conduct behavioral experiments ranging from behavior-based learning tasks to posture-dependent optogenetic stimulation. DeepLabStream has a temporal resolution in the millisecond range, can operate with multiple devices and can be easily tailored to a wide range of species and experimental designs. We employ DeepLabStream to autonomously run a second-order olfactory conditioning task for freely moving mice and to deliver optogenetic stimuli based on mouse head-direction.

Differentiation of human pluripotent stem cells (hPSC) into distinct neuronal populations holds substantial potential for disease modeling in vitro, toward both elucidation of pathobiological mechanisms and screening of potential therapeutic agents. For successful differentiation of hPSCs into subtype-specific neurons using in vitro protocols, detailed understanding of the transcriptional networks and their dynamic programs regulating endogenous cell fate decisions is critical. One major roadblock is the heterochronic nature of neurodevelopment, during which distinct cells and cell types in the brain and during in vitro differentiation mature and acquire their fates in an unsynchronized manner, hindering pooled transcriptional comparisons. One potential approach is to "translate" chronologic time into linear developmental and maturational time. Attempts to partially achieve this using simple binary promotor-driven fluorescent proteins (FPs) to pool similar cells have not been able to achieve this goal, due to asynchrony of promotor onset in individual cells. Toward solving this, we generated and tested a range of knock-in hPSC lines that express five distinct dual FP timer systems or single time-resolved fluorescent timer (FT) molecules, either in 293T cells or in human hPSCs driving expression from the endogenous paired box 6 (PAX6) promoter of cerebral cortex progenitors. While each of these dual FP or FT systems faithfully reported chronologic time when expressed from a strong inducible promoter in 293T cells, none of the tested FP/FT constructs followed the same fluorescence kinetics in developing human neural progenitor cells, and were unsuccessful in identification and isolation of distinct, developmentally synchronized cortical progenitor populations based on ratiometric fluorescence. This work highlights unique and often surprising expression kinetics and regulation in specific cell types differentiating from hPSCs.

BackgroundLong-term home-cage monitoring is essential to quantify spontaneous locomotor and social behaviors in group-housed mice, but analysis of high-density RFID tracking data remains a barrier to reproducibility. New methodsWe developed IntelliProfiler 2.0, a fully R-based pipeline tailored to the eeeHive 2D floor-mounted RFID array. The workflow performs data import from text logs, preprocessing, coordinate reconstruction, missing-value handling, feature extraction, statistical testing, and visualization in a single environment. Behavioral metrics include travel distance, close contact ratio (CCR), and a newly implemented inter-individual distance metric. ResultsIn four-day recordings of group-housed C57BL/6J mice (8 males and 8 females), IntelliProfiler 2.0 captured circadian phase-dependent locomotion and proximity patterns and reproduced sex-dependent differences consistent with prior analyses while incorporating updated hardware specifications. Radar-chart summaries enabled intuitive comparison of multidimensional behavioral profiles and inter-individual variability across light/dark phases. Comparison with existing methodsCompared with IntelliProfiler 1.0 and multi-tool workflows, IntelliProfiler 2.0 consolidates analysis into a single, script-based R pipeline, reducing operational complexity and improving reproducibility. The updated implementation supports recent manufacturer-driven changes, including antenna renumbering and multi-USB data export. ConclusionsIntelliProfiler 2.0 provides a reproducible, extensible framework for high-throughput behavioral phenotyping of group-housed mice and is scalable across hardware configurations, including simplified single-board recordings. HighlightsO_LIEnd-to-end R pipeline for eeeHive 2D floor-based RFID tracking analysis C_LIO_LIStandardized setup with comprehensive manuals and protocols C_LIO_LIInter-individual distance metric to quantify group spatial structure C_LIO_LICircadian- and sex-dependent behavioral profiling in group-housed mice C_LIO_LIRadar-charts summarize multidimensional behavioral profiles and variability C_LI

Abnormal response to sensory stimuli characterizes multiple neuropsychiatric conditions. However, not many tools are currently available to assess somatosensory abnormalities in rodent models of brain disorders, limiting the possibilities to study cellular and molecular correlates of this phenotypic trait. To this goal, previous studies relied on the whisker nuisance test (WNt), in which freely moving mice are constantly stimulated on their whiskers using a wood stick for a set time. The whisker-guided response is then deconstructed in behavioral categories indicative of either anxiety or curiosity. Thus far, WNt was shown to be a valuable tool to investigate sensory-driven abnormalities in mouse models of autism spectrum disorders (ASD), demonstrating a solid translational validity. Nevertheless, assessment of behavioral response in the WNt is currently limited by the lack of an objective quantification method. To overcome this limitation, we developed WNt3R (Whisker Nuisance Test in 3D for Rodents), a MATLAB toolbox that uses the output of the open-source software DeepLabCut to determine discrete body postures associated with multiple ethologically-relevant behaviors. Our results show that behavioral modules identified using WNt3R reliably decode mouse body postures, outperforming the manual user, thus offering a novel and unbiased approach to study altered somatosensory function in rodents.

Visualizing neuronal anatomy often requires labor-intensive immunohistochemistry on fixed and dissected brains. To facilitate rapid anatomical staining in live brains, we used genetically targeted membrane tethers that covalently link fluorescent dyes for in vivo neuronal labeling. We generated a series of extracellularly trafficked small molecule tethering proteins, HaloTag-CD41 and SNAPf-CD4, which directly label transgene expressing cells with commercially available ligand substituted fluorescent dyes. We created stable transgenic Drosophila reporter lines which express extracellular HaloTag-CD4 and SNAPf-CD4 with LexA and Gal4 drivers. Expressing these enzymes in live Drosophila brains, we labeled the expression patterns of various Gal4 driver lines recapitulating histological staining in live brain tissue. Pan-neural expression of SNAPf-CD4 enabled registration of live brains to an existing template for anatomical comparisons. We predict that these extracellular platforms will not only become a valuable complement to existing anatomical methods but will also prove useful for future genetic targeting of other small molecule probes, drugs, and actuators.

We describe here an efficient method for gene editing in mouse T cells, based on well-established, high-efficiency retroviral transduction protocols. Our platform allows analysis of mutant phenotypes in primary murine T cells in vitro and in vivo. This approach uses a single retroviral vector to simultaneously knockout an endogenous gene and ectopically express its mutant version. This knockout/re-expression vector can be used as the only plasmid to transduce Cas9-expressing T cells, or used together with a Cas9 retroviral vector to transduce T cells from any mouse strain. We validated the system for analysis of murine T cells by targeting key molecules in proximal T cell signaling, i.e. CD3{gamma} and Zap70. We obtain high knockout and re-expression efficiencies in both Cas9-expressing and non-Cas9 T cells. Knockout efficiencies can be further improved by gRNA multiplexing. Endogenous proteins compete with their ectopically expressed mutants or tagged versions for cellular location, protein interactions and cellular functions. Here, we quantified the incorporation of CD3{gamma}-GFP into surface T cell receptor (TCR) complexes. Our data shows that the knockout and re-expression platform improves integration of CD3{gamma}-GFP into the TCR. Therefore, eliminating competition between endogenous and ectopic proteins benefits analyses of protein assemblies and signaling pathways in primary T cells. Furthermore, we validated advantages of our system for mutant analysis using wild-type and mutant Zap70s. Zap70 mutants deficient in TCR binding or kinase activity show their phenotypes only in the absence of endogenous protein, further validating our knockout/re-expression approach. Most importantly, this system can be used to generate gene-edited primary T cells for in vivo studies, such as the quantification of anti-tumor responses. Our knockout and re-expression platform provides a useful gene editing tool for primary T cells in fundamental research and immunotherapy development.

Methylation profiling of circulating cell-free DNA (cfDNA) is of great interest as a liquid biopsy assay for the detection and monitoring of cancer and other pathologies. Here we describe circulating cell-free DNA reduced representation bisulfite sequencing (cf-RRBS), enabling the use of highly effective RRBS on fragmented plasma cfDNA. This method enriches the CpG-rich RRBS target regions by enzymatic degradation of all off-target DNA rather than by targeted capture, in contrast to previous methods. Critical steps are fully enzymatic in a single-tube, making it rapid, cost-effective, robust, and easily implemented on a liquid-handling station for high-throughput sample preparation. We benchmark cf-RRBS results to those obtained by previous more complex methods and exemplify its use for accurate non-invasive subtyping of lung cancer, a frequent onco-pathology task. cf-RRBS enables any molecular pathology lab to tap into the cfDNA methylome, only making use of off-the-shelf reagents and open-source data analysis tools. One Sentence SummaryNovel methodology enables facile and cost-effective methylation profiling of fragmented plasma DNA, allowing for routine liquid biopsy clinical diagnostics, exemplified here for differential diagnosis of lung cancer subtypes.

Accurate quantification of animal behavior is crucial for advancing neuroscience and for defining reliable physiological markers. We introduce Megabouts (megabouts.ai), a software package standardizing zebrafish larvae locomotion analysis across experimental setups. Its flexibility, achieved with a Transformer neural network, allows the classification of actions regardless of tracking methods or frame rates. We demonstrate Megabouts ability to quantify sensorimotor transformations and enhance sensitivity to drug-induced phenotypes through high-throughput, high-resolution behavioral analysis.

The cellular composition of tumors is now recognized as an essential phenotype, with implications to diagnostic, progression and therapy efficacy. A tool for accurate profiling of the tumor microenvironment is lacking, as single-cell methods and computational approaches are not applicable or suffer from low accuracy. Here we present EpiSort, a novel strategy based on targeted bisulfite sequencing, which allows the accurate enumeration of 23 cell types and may be applicable to cancer studies.

Analysis of cytometry data predominantly relies on clustering and dimensionality reduction approaches for computational tractability. This is particularly relevant for modern spectral flow cytometers, which can simultaneously measure an increasingly large number of antibody marker channels. While dimensionality reduction provides for more efficient data processing, this comes at the expense of data loss that may miss subtle patterns among rare cell types that may be critical for disease detection. Maintaining analysis at full dimensions presents opportunities to preserve resolution and provides complete downstream data interpretability. However, a significant obstacle to analysis of cytometry data without dimensionality reduction is the significant batch effects observed in this data between experiment days, operators and equipment types. Here we show a new strategy to denoise batch variation from both flow- and mass-cytometry datasets using an autoencoder neural network architecture. We generated a benchmark flow cytometry dataset in mice to compare this approach to current toolsets and find our approach shows superior preservation of biological signals, whilst also performing batch correction equal to current best methodology. Our batch alignment approach works to such an extent that it becomes practically possible to project batch aligned data into multidimensional space to generate a novel representation of cellular phenotype for downstream model building. This hyperdimensional approach maintains original data resolution without requiring dimensionality reduction, and thus any resultant cell populations that differentiate phenotypes remain fully interpretable. We show with two large clinical datasets that our batch-alignment approach coupled with the multi-dimensional representation successfully detects meaningful patterns in cases where the original analysis methods struggled. This new framework removes some of the inherent technical limitations encountered in the integration of large, multi-batch cytometry datasets and provides a framework for machine learning model building from this modality. An implementation of this framework and an associated web application accompanies this manuscript at http://voxelcoder.cloud. Significance StatementThis work addresses a critical bottleneck in cytometry analysis by introducing a neural network-based approach that effectively removes technical batch variation while preserving biological signals. The novel hyperdimensional representation maintains full data resolution without dimensionality reduction, enabling more sensitive detection of disease-associated cellular signatures than current methods. Importantly, this framework enables reliable batch normalization of fresh samples processed at different times and locations, overcoming the practical constraints of clinical sample collection where simultaneous processing is often impossible. The enhanced sensitivity for identifying pathogenic cellular patterns has immediate implications for biomarker discovery and personalized medicine applications.

Accurate identification of immune cell subsets in single-cell (sc)RNA-seq data is critical for understanding immune responses in autoimmune diseases, infections, and cancer. One caveat of scRNA-seq is the inability to properly assign rare immune cell subsets due to gene dropout events. To circumvent this caveat, we here developed Optimized Detection and Inference of Names in scRNA-seq data (scODIN). scODIN uses an informed holistic two-step approach combining expert knowledge with machine learning to rapidly assign cell type identities to large scRNA-seq dataset. First, scODIN uses key lineage-defining markers to identify a set of core cell types. Second, scODIN compensates for dropout events by integrating a k-nearest neighbors (kNN) algorithm. We additionally programmed scODIN to detect dual and transitional phenotypes, which are usually overlooked in conventional analyses. Consequently, scODIN may enhance our understanding of immune cell heterogeneity and provides comprehensive insights into immune regulation, with broad implications for immunology and personalized medicine.

Tissue-clearing and light-sheet microscopy have transformed volumetric imaging of intact organs, yet limited mechanistic understanding of dehydration-based clearing continues to constrain rational protocol design and broader applicability. Here, we define the cardinal chemical and physical principles underlying dehydration-based tissue-clearing and establish a new pipeline for large-volume imaging. To maximize imaging performance, we developed the mesoSPIM-ultra, an upgraded mesoSPIM platform with a temperature-controlled sample chamber, a large field-of-view (FoV) camera and specialized optics to achieve long-working-distance, high-resolution imaging of cleared samples. We applied this approach to investigate the projectome of Chx10+ neurons, a cell population with complex axonal morphologies along the entire mouse spinal-cord and brain, and implicated in ipsilateral orienting behaviors. By combining behavioral analysis with post-hoc single-neuron reconstructions, we revealed previously inaccessible branching architectures and long-range projections extending from the brainstem to the spinal cord. Together, our work establishes a mechanistic foundation for tissue-clearing and scalable imaging.

Microglia play critical roles in brain health and disease by adopting a spectrum of dynamic activation states. However, capturing this continuous heterogeneity in a scalable way remains a major challenge. To address this, we developed an imaging and data analysis framework to map the activation landscape of human iPSC-derived microglia (iMG) at single-cell resolution. We combined high-content imaging using two complementary strategies: a hypothesis-driven immunofluorescence (IF) panel targeting key activation markers (NF-{kappa}B, ASC, CD45) and a discovery-oriented, pan-morphological Cell Painting (CP) assay. Diverse phenotypes were captured through handcrafted and representation learning-based features. To classify cells, we applied Gaussian Mixture Models (GMMs) to image-derived features, enabling soft probabilistic assignments that capture transitional states between phenotypes. Compared to graph-based methods like the Leiden algorithm, GMMs provided comparable classification performance while offering a more nuanced and biologically interpretable view of microglial heterogeneity. We demonstrate that deep learning features from the targeted IF panel are most powerful, achieving high classification accuracy and strong correlation with biological states such as functional NLRP3 inflammasome activation. Our model system provides a robust and scalable platform for quantifying microglial heterogeneity, offering a new tool to identify novel disease-associated states and compounds that precisely modulate microglial phenotypes for therapeutic discovery.

Studying the spatiotemporal dynamics of cells in living organisms is a current frontier in bioimaging. Intravital Microscopy (IVM) provides direct, long-term observation of cell behavior in living animals, from tissue to sub-cellular resolution. Hence, IVM has become crucial for studying complex biological processes in motion and across scales, such as the immune response to pathogens and cancer. However, IVM data are typically kept in private repositories inaccessible to the scientific community, hampering large-scale analysis that aggregates data from multiple laboratories. To solve this issue, we introduce Immunemap, an atlas of immune cell motility based on an Open Data platform that provides access to over 58000 single-cell tracks and 1049000 cell-centroid annotations from 360 videos in murine models. Leveraging Immunemap and unsupervised learning, we systematically analyzed cell trajectories, identifying four main patterns of cell migration in immune cells. Two patterns correspond to behaviors previously characterized: directed movement and arresting. However, we identified two other patterns, characterized by low directionality and twisted paths, often considered random migration. We show that the newly defined patterns can be subdivided into two distinct types: within small areas, suggesting a focused patrolling around one or a few cells, and over larger areas, indicative of a more extended tissue patrolling. Furthermore, we show that the percentage of cells displaying these motility patterns changes in response to immune stimuli. Altogether, Immunemap embraces the FAIR principles, promoting data reuse to extract novel insights from immune cell dynamics through an image-based systems biology approach.

Single-cell RNA sequencing (scRNA-seq) can reveal accurate and sensitive RNA abundance in a single sample, but robust integration of multiple samples remains challenging. Large-scale scRNA-seq data generated by different workflows or laboratories can contain batch-specific systemic variation. Such variation challenges data integration by confounding sample-specific biology with undesirable batch-specific systemic effects. Therefore, there is a need for guidance in selecting computational and experimental approaches to minimize batch-specific impacts on data interpretation and a need to empirically evaluate the sources of systemic variation in a given dataset. To uncover the contributions of experimental variables to systemic variation, we intentionally perturb four potential sources of batch-effect in five human peripheral blood samples. We investigate sequencing replicate, sequencing depth, sample replicate, and the effects of pooling libraries for concurrent sequencing. To quantify the downstream effects of these variables on data interpretation, we introduced a new scoring metric, the Cell Misclassification Statistic (CMS), which identifies losses to cell type fidelity that occur when merging datasets of different batches. CMS reveals an undesirable overcorrection by popular batch-effect correction and data integration methods. We show that optimizing gene expression matrix normalization and merging can reduce the need for batch-effect correction and minimize the risk of overcorrecting true biological differences between samples.

Tissues such as the liver lobule, kidney nephron, and intestinal gland exhibit intricate patterns of zonated gene expression corresponding to distinct cell types and functions. To quantitatively understand zonation, it would be important to measure cellular or genetic features as a function of position along a zonal axis. While it is possible to manually count, characterize, and locate features in relation to the zonal axis, it is very difficult to do this for more than a few hundred instances. We addressed this challenge by developing a deep-learning-based quantification method called the "Tissue Positioning System" (TPS), which can automatically analyze zonation in the liver lobule as a model system. By using algorithms that identified vessels, classified vessels, and segmented zones based on the relative position along the portal vein to central vein axis, TPS was able to spatially quantify gene expression in mice with zone specific reporters. TPS could discern expression differences between zonal reporter strains, ages, and disease states. TPS could also reveal the zonal distribution of cells previously thought to be randomly distributed. The design principles of TPS could be generalized to other tissues to explore the biology of zonation. The software is available at https://github.com/yunguan-wang/Tissue_positioning_system.

The diverse functions of the enteric nervous system (ENS) arise from communication between molecularly distinct neuronal populations organized into complete circuits. While recent single-cell transcriptomic studies have resolved the molecular identity of enteric neuron classes, methods for defining their synaptic connectivity remain limited. Here, we describe the implementation of mWmC, an anterograde monosynaptic tracer based on a fusion of wheat germ agglutinin (WGA) and mCherry, as a non-toxic, single-component viral tool for mapping neuronal circuits within and beyond the ENS. Following adeno-associated virus (AAV)-mediated expression in enteric neurons, mWmC was efficiently expressed and transmitted selectively to postsynaptic neurons, with no detectable transfer to enteric glia, interstitial cells of Cajal, blood vessels or other mesenchymal cell types. The method also identified postsynaptic neurons in the celiac-superior mesenteric ganglia following tracing of intestinofugal enteric neurons, demonstrating its utility for mapping inter-organ circuits. As proof of principle, we applied mWmC to two genetically defined myenteric interneuron populations and identified preferential postsynaptic targets, revealing selective connectivity with distinct enteric neuron classes. Time-course experiments showed that transsynaptic labeling occurred between 4 and 10 days and reached a plateau thereafter, consistent with monosynaptic transfer. Finally, we developed a dual-reporter version of the system that simultaneously distinguishes input and target neurons within the same tissue. Together, mWmC provides a robust approach for defining circuit architecture in the ENS, linking molecular cell atlases with neuronal connectivity and paving the way for deeper insights into the circuit mechanisms underlying gut physiology. Graphical AbstractSchematics illustrating the implementation and applications of the anterograde monosynaptic tracer mWmC for mapping enteric neuronal circuits. Parts of schematics are generated with Biorender. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=36 SRC="FIGDIR/small/741165v2_ufig1.gif" ALT="Figure 1"> View larger version (12K): org.highwire.dtl.DTLVardef@19635bborg.highwire.dtl.DTLVardef@a18ffborg.highwire.dtl.DTLVardef@f3d011org.highwire.dtl.DTLVardef@e1359c_HPS_FORMAT_FIGEXP M_FIG C_FIG

Recent advances with deep neural networks have shown the feasibility of acquiring brightfield images with transmitted light and applying in-silico labeling to predict fluorescent images. We have developed a novel in-silico labeling method based on a generative adversarial network and outperforms the state-of-the-art Unet method in generating realistic fluorescent images and quantitatively recapitulating real staining signals, as demonstrated in a complex co-culture myelination assay. Furthermore, we have performed the assay in live mode with multiple kinetic points, applied in-silico labeling to predict fluorescent images from brightfield and quantified the kinetic phenotypic changes. Thus, the proposed approach provides a potential tool to study the kinetics of cellular phenotypic changes with brightfield imaging.

Fluorescence microscopy is a key method in the life sciences. State of the art -omics methods combine fluorescence microscopy with complex protocols to visualize tens to thousands of features in each of millions of pixels across samples. These -omics methods require precise control of temperature, reagent application, and image acquisition parameters during iterative chemistry and imaging cycles conducted over the course of days or weeks. Automated execution of such methods enables robust and reproducible data generation. However, few commercial solutions exist for temperature controlled, fluidics coupled fluorescence imaging, and implementation of bespoke instrumentation requires specialized engineering expertise. Here we present PySeq2500, an open source Python code base and flow cell design that converts the Illumina HiSeq 2500 instrument into an open platform for programmable applications. Customizable PySeq2500 protocols enable experimental designs involving simultaneous 4-channel image acquisition, temperature control, reagent exchange, stable positioning, and sample integrity over extended experiments. To demonstrate accessible automation of complex, multi-day workflows, we use the PySeq2500 system for unattended execution of iterative indirect immunofluorescence imaging (4i). Our automated 4i method uses off-the-shelf antibodies over multiple cycles of staining, imaging, and antibody elution to build highly multiplexed maps of cell types and pathological features in mouse and postmortem human spinal cord sections. As demonstrated here, PySeq2500 enables non-specialists to develop and implement state of the art fluidics coupled imaging methods in a widely available benchtop system.

In the past years, machine-learning-based approaches to track animal poses with high spatial and temporal resolution have become available, but toolkits to extract, integrate, and analyze coordinate datasets in a user-friendly manner have lagged behind. Here, we introduce Fear-Mouse Tracker (FMT), a simple and open-source MATLAB-based pipeline to extract and quantitatively analyze DeepLabCut-derived coordinates of mice presented with threatening stimuli that commonly trigger innate defensive responses. This framework allows for unbiased quantitative estimations of stretch-attend posture (SAP) observed during risk assessment behaviors, as well as for measurements of the timing and extent of freezing and escape responses that follow the presentation of threatening stimuli such as a predator odor, or sweeping and looming stimuli resembling predator approaches. FMT is specially designed for users not very experienced in using programming languages, thus making it more accessible to a broader audience.