Biology subjects

Lu, Z.

Publications and source records attributed to Lu, Z..

At least 73 records · Page 4Linked to original sources

bioRxiv2025

Mnemonic maps of visual space in human prefrontal cortex

Neural theories of how the prefrontal cortex (PFC) supports working memory rely on evidence from decades of pioneering macaque research. In some respects, efforts to translate these animal models of working memory in human PFC using neuroimaging have largely failed. One possible explanation, before concluding key non-homologies between the species, is that previous neuroimaging studies used resolutions too coarse to be sensitive to intermixed distributions of neurons tuned to memorized features. To resolve this concern, we scanned human PFC at 900 micron resolution using 7T fMRI. We found that population activity in retinotopically-organized superior precentral sulcus-rather than the predicted midlateral PFC-persists during memory, encodes fine-grained information about memorized items, predicts behavioral errors in memory, and forms a stable subspace with a topological organization yoked to visual space. These results have important implications for both functional homologies between the species and theories of how the PFC supports working memory.

neuroscience↗

bioRxiv2025

Computer Vision Methods for Spatial Transcriptomics: A Survey

Spatial transcriptomics (ST) enables the simultaneous measurement of gene expression and spatial localization within tissue sections, providing unprecedented opportunities to dissect tissue architecture and functional organization. As a relatively new omics technology, bioinformatics has driven much of the innovation in ST. However, within these frameworks, "spatial" information is often reduced to locations and relationships between molecular profiles, without fully leveraging the wealth of submicron morphological detail and histological knowledge available. Advances in computer vision-based artificial intelligence (AI) are opening exciting new avenues beyond conventional bioinformatics approaches by modeling complex histological patterns and linking morphology to molecular states. More excitingly, they bring fresh perspectives to potentially address key limitations of ST, including its high cost, limited clinical applicability, and reliance on twodimensional (2D) analysis of inherently three-dimensional (3D) tissues. For instance, models that predict ST directly from histology images enable "virtual sequencing," drastically reducing costs while integrating morphological insights from pathology with molecular biomarkers, thus accelerating clinical translation. Moreover, computer vision techniques can reconstruct pixel-aligned 3D tissue models, overcoming the technical barriers of 2D acquisition and advancing 3D spatial omics analytics. In this paper, we present the first systematic survey of computer vision AI models for ST analytics, categorizing approaches across architectures, learning paradigms, tasks, and datasets, and tracing their technological evolution. We highlight key challenges and future directions, offering a panoramic perspective on vision-driven ST and its potential to transform both basic research and clinical practice. The curated collection of vision-driven spatial transcriptomics papers is available at https://github.com/hrlblab/computer_vision_spatial_omics.

bioinformatics↗

bioRxiv2025

Sexual dimorphism in the complete connectome of the Drosophila male central nervous system

Sex differences in behaviour exist across all animals, typically under strong genetic regulation. In Drosophila, fruitless/doublesex transcription factors can identify dimorphic neurons but their organisation into functional circuits remains unclear. We present the connectome of the entire Drosophila male central nervous system. This contains 166,691 neurons spanning the brain and nerve cord, fully proofread and annotated including fruitless/doublesex expression and 11,691 types. We provide the first comprehensive comparison between male and female brain connectomes to synaptic resolution, finding 7,205 isomorphic, 114 dimorphic, 262 male-specific and 69 female-specific types. This resource enables analysis of full sensory-to-motor circuits underlying complex behaviours and the impact of dimorphic elements. Sex-specific/dimorphic neurons are concentrated in higher brain centres while the sensory and motor periphery are largely isomorphic. Within higher centres, male-specific connections are organised into hotspots defined by male-specific neurons or arbours. Numerous circuit switches reroute sensory information to form antagonistic circuits controlling opposing behaviours.

neuroscience↗

bioRxiv2025

Quantitative Prevalence, Chemical Speciation, and ECM-Embedded Networks of Biogenic Silicon in Animal Tissues Revealed by Refined Analytical Methods

Silicon, abundant in Earths biosphere, plays crucial roles in many organisms. However, its biological functions in animals remain inadequately characterized, primarily constrained by methodological limitations in quantifying, identifying, and visualizing biogenic silicon in tissues. Here, we refined the molybdenum blue colorimetric method (MB) for more accurate silicon quantification, developed integrated mass spectrometry techniques to delineate chemical natures of biogenic silicon, and optimized micro-X-ray fluorescence (micro-XRF) imaging for spatial mapping in animals. Employing these methodologies, we demonstrated that silicon abundance in animals substantially exceeds prior estimates--surpassing essential elements like iron. We also revealed the chemical architectures of diverse silicon compounds and visualized tissue-specific distributions in mice and humans, indicating its predominant localization within extracellular matrix (ECM) and suggesting the formation of silicified networks. These findings establish silicon as a structurally and quantitatively significant element with profound biological implications, warranting its recognition as an essential factor in human physiology and pathology.

biochemistry↗

bioRxiv2025

Divergent Strategies of Mycorrhiza-Mediated Drought Adaptation in Poplar

Mycorrhizal symbiosis shapes plant growth and stress resilience. Here, we compared physiological and molecular responses of poplars colonized by Paxillus involutus (Pi) or Cenococcum geophilum (Cg) under control conditions, drought stress, and recovery. Both fungal species primed distinct local (root) and systemic (leaf) defenses compared to non-inoculated (Ni) plants. Cg-colonized poplars exhibited constitutively elevated transcripts of heat shock proteins (HSPs), galactinol synthase, and aquaporins in roots and leaves, irrespective of drought. Pi colonization enhanced growth and nitrogen-use-efficiency along with transcriptional increases of TOR/RAPTOR complex. Under severe soil moisture decline, Pi and Ni poplars showed reduced water potential, photosynthesis, growth, and leaf shedding, whereas Cg-colonized plants maintained water status, sustained photosynthesis, and retained foliage. These results reveal two contrasting mycorrhiza-mediated drought strategies in poplar: Pi fosters stress acclimation via drought-induced leaf abscission, enabling rapid recovery; Cg suppresses growth even without stress, conferring constitutive tolerance. Ectomycorrhizal species thus occupy different positions on the growth- defense trade-off spectrum. Such species-specific effects have important ecological and applied implications, enabling targeted use of EM fungi in forestry and agriculture depending on whether maximizing productivity or enhancing stress resilience is the primary goal.

plant biology↗

bioRxiv2025

Mining genes underlying perennial-to-annual transitions in Oryza with comparative genomics

Perennial rice has shown great potential for alleviating multidimensional sustainability issues that are prevalent in current rice agriculture, prompting widespread interest in its breeding. Currently, one major hurdle for these breeding efforts is our scarce knowledge about the evolutionary mechanisms of the perennial-to-annual transition in Oryza. Here, by performing comparative genomics of three annual-perennial pairs of Oryza species (Oryza rufipogon/O. nivara, O. glumaepatula/O. glaberrima, and O. meyeriana/O. brachyantha), we identified 91 gene families that are present in all the perennial species but are missing in any of the annual species. Annotations and functional prediction of these gene families demonstrated that they may be associated with a battery of physiological components of perenniality, such as phenology, energy allocation, and source-sink dynamics. Furthermore, comparative co-expression analysis revealed that many of these families may also play important regulatory roles for orchestrating diverse processes that differ between perennial and annual Oryza species, including reproduction, root development, and response to environmental stimuli. These results highlight the potential complexity of the emergence of annuality in the Oryza genus and offer promising candidates for future functional studies.

plant biology↗

bioRxiv2025

A Neural Circuit for Modular Gating of Organ Somatosensation

End-organs such as the bladder rely on a delicate balance between internal urgency and voluntary restraint. However, the specific spinal circuits coordinating these functions remain poorly defined. Here, we identify a genetically defined population of lumbosacral spinal interneurons marked by Trpc4 expression that gate bladder sensory input and regulate micturition reflexes. Single-cell transcriptomics and in situ physiology reveal molecularly distinct Trpc4 subtypes. Functional manipulation reveals that Trpc4 neurons are essential for coordinating bladder-sphincter activity and gating visceral pain. Their ablation leads to bladder hypersensitivity and voiding dysfunction, while targeted activation reverses these maladaptive states. Circuit tracing reveals convergence of primary afferent and descending brainstem inputs onto Trpc4 neurons. These findings establish Trpc4 interneurons are critical for bladder sensory-motor integration and extend classical spinal gating models to encompass visceral pain and organ reflex control.

neuroscience↗

bioRxiv2025

Covalent Inhibition of the Human Papillomavirus Type 16 E6 Protein Restores p53 and Suppresses HPV-Driven Tumorigenesis

High-risk human papillomavirus (HPV) infections are the etiology of approximately 5% of all cancers worldwide, including cervical, anal, and oropharyngeal malignancies. HPV E6 is a multifunctional oncoprotein that drives tumorigenesis and is best known for bridging the ubiquitin ligase E6AP (UBE3A) and p53 into a complex that leads to proteasome mediated destruction of p53. We developed small molecule inhibitors that covalently bind to cysteine-51 (Cys-51) in HPV16 E6. In HPV16-positive cancer cells, these compounds increase p53 protein levels and activate p53-dependent transcriptional programs associated with apoptosis and senescence, resulting in reduced tumor cell viability. In vivo, E6 inhibition suppresses the growth of human HPV16 expressing cervical and oropharyngeal tumor cell lines in mice. The strategy of targeting a viral oncoprotein with a covalent inhibitor demonstrates a genotype-specific therapeutic strategy for HPV-associated cancers and premalignant infections, addressing a significant unmet need in current treatment options.

cancer biology↗

bioRxiv2025

Embodied Intelligence Unlocks Autonomous Microscopy

Advanced microscopy is a cornerstone of modern science, yet its potential is constrained by a reliance on manual operation, which struggles with the complexity and reproducibility of long-term experiments protocols. While scripting offers a degree of automation, it lacks the generalizability to adapt to new samples or dynamic biological events. A fundamental challenge is the absence of an intelligent system that can interpret high-level scientific intent and ground it in the physical actions of the microscope. To bridge this gap, we introduce an Embodied Intelligent Microscope System (EIMS) with hierarchical reasoning structure that reimagines the microscope autonomous control. This system leverages the advanced reasoning of large models to interpret complex user commands and decompose them into actionable steps. To solve the critical grounding problem, we constrain the models output to the feasible action space, effectively serving as the models "hands and eyes" in the physical world. We demonstrate that our system achieves zero-shot generalization on complex, multi-step protocols and successfully automates challenging biological missions requiring expert-level judgment, such as capturing the sparse spatiotemporal events of cell mitosis and locating scatteredly distributed organoids. This work establishes a new paradigm for scientific instrumentation--fusing high-level intent understanding with grounded, dynamic execution--and provides a generalizable framework for deploying embodied intelligence to accelerate autonomous scientific discovery.

bioengineering↗

bioRxiv2025

Global and local origins of trial-to-trial spike count variability in visual cortex

Sensory neuron spiking responses vary across repeated presentations of the same stimuli, but whether this trial-to-trial variability represents noise versus unidentified signals remains unresolved. Some of the variability can be attributed to correlations between neural activity and arousal, locomotion, and other overt movements. We hypothesized that correlations with global activity factors, i.e., patterns of neural activity observable in other brain regions, may explain additional variability in spike count responses of visual cortical neurons. To test this, we used Neuropixels 2.0 probes to record neural activity in mouse primary visual cortex (V1) while subjects passively viewed images. We recorded videos of behavior alongside neural activity from other brain regions, either spiking activity of neural populations in anterior cingulate area (ACA) or widefield calcium signals from across the dorsal cortex. We then used a model based on reduced rank regression to partition the explainable variability of visual cortical responses by source. Some of the trial-to-trial variability in V1 spike counts was attributable to locally shared patterns of activity uncorrelated with either behavior or global activity patterns. Locally shared activity patterns explained trial-to-trial variability that was in excess of Poisson spike generation. Of the parts of variability attributable to non-local sources, global cortical activity predicted significantly more V1 spike count variability than behavioral factors. Additionally, behavioral factors explained little variability uniquely and comprised a geometric subspace of the globally predictable V1 activity. Finally, optogenetically perturbing ACA directly impacted V1 activity, and ACA activity patterns predicted V1 spike count variability even on trials without overt behaviors. Our data indicate that globally shared factors from other cortical areas contribute substantially to shared spike count variability in V1, with only a minority of shared variability confined to local V1 circuits.

neuroscience↗

bioRxiv2025

Profibrotic monocyte-derived alveolar macrophages as a biomarker and therapeutic target in systemic sclerosis-associated interstitial lung disease

Interstitial lung disease (ILD) is present in over 60% of patients with systemic sclerosis (SSc) and is the leading cause of SSc-related deaths. Profibrotic monocyte-derived alveolar macrophages (MoAM) play a causal role in the pathogenesis of pulmonary fibrosis in animal models where their persistence in the niche requires signaling through Colony Stimulating Factor 1 Receptor (CSF1R). We hypothesized that the presence and proportion of MoAM in bronchoalveolar lavage (BAL) fluid from patients with SSc-ILD may be a biomarker of ILD severity. To test this hypothesis, we analyzed BAL fluid from 9 prospectively enrolled patients with SSc-ILD and 13 healthy controls using flow cytometry and single-cell RNA sequencing. Patients with SSc-ILD had more MoAM and interstitial macrophages in BAL fluid than healthy controls, and their abundance was associated with lung fibrosis severity. We identified changes in the MoAM transcriptome as a function of treatment with mycophenolate, an effective therapy for SSc-ILD. In SSc-ILD lung explants, spatial transcriptomics identified an expanded population of interstitial macrophages spilling into the alveolar space. Our findings suggest that the proportion of profibrotic MoAM and interstitial macrophages in BAL fluid may serve as a biomarker of SSc-ILD and credential them as possible targets for therapy.

immunology↗

bioRxiv2025

Spatiotemporal lineage mapping of tumor immune escape with eTRACER

The plasticity-first hypothesis posits that phenotypic plasticity emerges first in novel environments, followed by (epi)genetic changes that stabilize adaptive phenotypes. Whether tumor evolution follows this model, and how non-genetic mechanisms sustain long-term therapeutic resistance, remain unclear. Here, we develop eTRACER, an endogenously targeted, high-resolution and spatially-resolved CRISPR-Cas9 lineage-tracing system, to study tumor adaptation during longitudinal CD8+ T cell-mediated immunotherapy. Integrated clonal and single-cell transcriptomic analyses show that tumor immune evasion is governed by clonal state composition rather than clonal identity, arguing against selection from pre-existing genetic heterogeneity. Spatiotemporal lineage tracing reveals highly reversible state transitions prior to therapy, followed by consolidated epithelial-mesenchymal transition (EMT) during therapy. Single-cell multiomics identifies AP-1 as a key regulator of EMT stabilization and its inhibition enhances tumor sensitivity to immune killing. Together, these findings show that tumor plasticity precedes stable immune evasion, revealing an AP-1-centered mechanism underlying plasticity-first tumor evolution.

cancer biology↗

bioRxiv2025

Unfolding spatiotemporal representations of 3D visual perception in the human brain

Although visual input is initially recorded in two dimensions on our retinas, we perceive and interact with the world in three dimensions. Achieving 3D perception requires the brain to integrate 2D spatial representations with multiple depth cues, such as binocular disparity. However, most studies typically examine 2D and depth information in isolation, leaving the integrated nature of 3D spatial encoding largely underexplored. In this study, we collected a densely sampled multimodal neuroimaging dataset from 10 participants (8 with EEG and fMRI; 2 with fMRI only) across multiple sessions while they viewed stereoscopic 3D stimuli through red-green anaglyph glasses. Participants first completed a behavioral session including depth judgement tasks and a novel cube adjustment task to quantify and calibrate individual depth perception in units of binocular disparity. Then during two EEG and two fMRI sessions, participants passively viewed stimuli presented at 64 systematically sampled 3D locations, yielding over 66,000 trials in total across ten participants. Combining this multimodal dataset with computational methods via representational similarity analysis, we examined how 2D, depth-related, 3D feature-level, and geometric distance representations unfold across time and brain space. We found that the human brain represents 3D visual space not only by encoding position-in-depth as an additional dimension alongside 2D location, but also by constructing richer forms of 3D spatial structure. Specifically, 2D spatial features were represented earliest and most broadly, depth-related representations were weaker and more spatially restricted, and 3D feature representations were sparse and heterogeneous but detectable at the individual feature level. Critically, geometric distance analyses revealed that neural coding extended beyond separable feature dimensions, showing robust integrated 2D geometric representations and more selective evidence for integrated 3D geometric representations. These findings suggest that human 3D spatial perception is supported by a progression from dominant 2D coding to depth-related and 3D representations, with additional evidence for geometric structure in 3D space, contributing to a more comprehensive understanding of the spatiotemporal organization of neural representations that support 3D perception. Additionally, our novel large dataset will be made openly available to support future research on 3D perception and spatial cognition.

neuroscience↗

bioRxiv2025

Target-Stabilized Base Editors Enable Robust High-Fidelity RNA Editing

RNA base editing using engineered deaminases expands target compatibility, offering a powerful tool to correct mutations at the RNA level. However, widespread off-target effects, primarily arising from dissociated free deaminases, remain a significant challenge. Here, we devised the RECODE (RNA editing with conditionally stable and enhanced ADAR1 deaminase variants) system, which employs designer degron-tagged ADAR1 deaminase (ADAR1d) with guide RNA (gRNA)-regulated stability. By promoting degradation of gRNA-unbound ADAR1d, RECODE markedly reduced transcriptome-wide edits while maintaining high on-target efficacy. Engineering gRNA for target RNA-induced conformational switching restricts ADAR1d stabilization, further enhancing editing precision. With structure-guided rational engineering of ADAR1d, RECODE achieved editing efficiencies of 90.3% and 58.8% for an Amyotrophic Lateral Sclerosis-relevant FUS mutation in vitro and in vivo, respectively. These findings establish RECODE as a highly stringent and efficient RNA editing technology and underscore a general principle for enhancing the specificity of RNA-guided protein effectors.

molecular biology↗

bioRxiv2025

Deep Evolutionary Fitness Inference for Variant Nomination from Directed Evolution

Iterative screening techniques, such as directed evolution, enable high-throughput affinity maturation to optimize binders to molecular interfaces. However, the decision problem of selecting variants from rich, evolved populations to enter low-throughput follow-up methods remains a significant bottleneck. Here, we present evolutionary fitness inference (EVFI) and DeepEVFI, two machine learning methods that model directed evolution from time-series sequencing data, and infer fitness, a variants ability to enrich under selection pressure. Our methods flexibly handle mutation mechanisms and starting populations that may be partially unknown - settings relevant to drug discovery - and achieve strong performance on a diverse set of experimental data. We conducted two experimental directed evolution campaigns, using antibodies and macrocyclic peptides libraries to identify and optimize binders to therapeutically relevant targets. EVFI and DeepEVFI identified tighter binders that were missed by human experts using conventional frequency-based approaches, including "rising stars" with low frequency. Beyond initial hit discovery, EVFI and Deep-EVFI enables labeling large-scale sequence-fitness datasets and identifying variants of initial binders with diverse properties.

bioinformatics↗

bioRxiv2025

Integrated analysis of crosslink-ligation data for the detection of RNA 2D/3D structures and interactions in vivo.

ii.SUMMARY/ABSTRACTThe latest advancements and implementation of crosslinking- and proximity-ligation based methods have transformed the study of RNA structure and interactions in living cells. Despite this progress, sequencing data must be meticulously processed to elucidate information on RNA. This presents distinct challenges, as noted in the Psoralen Analysis of RNA Interactions and Structures (PARIS) and Spatial 2'-Hydroxyl Acylation Reversible Crosslinking (SHARC) studies. In this chapter, we detail the process used to analyze RNA sequencing data derived from crosslink-ligation based experiments. We detail the strategies used to pre-process raw data, map to a reference genome, classify reads, and assemble duplex groups. Further, we show an example of analyzing structural data from the ribosomal RNA.

bioinformatics↗

bioRxiv2025

Enzymatically active exudates from Alteromonas facilitate Prochlorococcus survival in stationary phase

The cyanobacterium Prochlorococcus has a conspicuously reduced genome causing it to require help from co-existing organisms for survival under a variety of stressful conditions. In this work we demonstrated that the heterotrophic bacterium Alteromonas macleodii EZ55 facilitated the survival of Prochlorococcus MIT9312 batch co-cultures as they entered stationary phase. We further showed that exudates from both Alteromonas and Prochlorococcus were responsible for this effect. Unidentified toxic exudates of Prochlorococcus lowered the carrying capacity of Pro99 medium for axenic Prochlorococcus cells, whereas heat-labile high-molecular weight exudates of Alteromonas both removed the effect of Prochlorococcus exudates and extended the lifespan of axenic Prochlorococcus cultures. Alteromonas exudates contained a wide variety of proteins and demonstrated enzymatic activities. Some of these proteins and activities may have been packaged within extracellular membrane vesicles, which we identified within Alteromonas exudates and found capable of physically associating with Prochlorococcus cells. Many of the functionalities observed in Alteromonas exudates (e.g., increasing phosphate availability, degrading hydrogen peroxide) were consistent with leaky Black Queen processes, which are defined as services provided by one organism that benefit the entire community and favor the evolution of interdependencies in microbial communities. Therefore, we discuss the potential ramifications of such processes being packaged into vesicles as opposed to freely diffusing through the extracellular milieu.

microbiology↗

bioRxiv2025

Enhancing Biomedical Optical Volumetric Imaging via Self-Supervised Orthogonal Learning

Optical volumetric imaging grapples with inherent noise problems arising from photon budget constraints, light scattering, and space-bandwidth product bottlenecks, all of which degrade structural fidelity and become even more pronounced than in planar imaging. While deep learning presents potential for denoising, supervised methods are hindered by the impracticality of paired dataset, and existing self-supervised approaches fail to fully exploit the intrinsic volumetric structural redundancy inherent to optical imaging. Here, we present a self-supervised Volumetric biomedicAL Imaging Denoiser (VALID) that leverages intrinsic three-dimensional spatial coherence for highly efficient volumetric denoising through a self-supervised orthogonal learning framework. VALID demonstrates robust denoising performance across diverse imaging modalities, including two- and three-photon microscopy, light-field microscopy, and optical coherence tomography, substantially enhancing structural fidelity in deep-tissue, multimodal, and dynamic imaging scenarios. By combining computational efficiency with zero-shot adaptability, VALID establishes a transformative approach to volumetric image enhancement with structure-aware precision.

neuroscience↗