Biology subjects

Xu, M.

Publications and source records attributed to Xu, M..

At least 73 records · Page 4Linked to original sources

bioRxiv2025

A ubiquitous Streptomyces biosynthetic megacluster encodes an arsenal of synergistic biotin-targeting antibiotics

The rise of multidrug-resistant pathogens underscores the urgent need for antibiotics that act through new targets and mechanisms. Biotin metabolism, essential in most bacteria, remains underexploited therapeutically. Here, we uncover a highly conserved, co-located biosynthetic megacluster in Streptomyces, a striking "cluster of clusters", that encodes four distinct natural product families: acidomycin, stravidins, dapamycins, and -methyl-KAPA, and is flanked by genes that encode streptavidin, a high-affinity biotin-binding protein. Remarkably, all molecules target different steps in bacterial biotin metabolism, revealing a multi-pronged natural strategy for biotin starvation. This arrangement of four functionally convergent biosynthetic gene clusters at a single genomic locus is without precedent. Even more surprisingly, we find that this anti-biotin megacluster is widespread across Streptomyces bacteria, suggesting a deeply conserved evolutionary solution to microbial competition. Mechanistically, the compounds inhibit biotin biosynthesis through enzyme blockade, prodrug activation, covalent cofactor mimicry, and biotin sequestration via co-expressed streptavidin. Stravidin S2 and -methyl-KAPA are effective in a murine model of multidrug-resistant E. coli infection. These findings expose a coordinated biosynthetic logic in microbial secondary metabolites and point to higher-order biosynthetic architectures as promising reservoirs of antibiotic innovation.

microbiology↗

bioRxiv2025

Evaluation and Aggregation of Active Module Identification Algorithms

1High-throughput sequencing methods have generated vast amounts of genetic data for candidate gene studies. However, the complexity of the disease genetic structure often results in a large number of candidate genes and poses a significant challenge for these studies. To explore the multi-gene interactions and elucidate the genetic mechanism, candidate genes are often analyzed through Gene-Gene interaction (GGI) networks. These networks can become very large, necessitating efficient methods to reduce their complexity. Active Module Identification (AMI) is a common method to analyze GGI networks by identifying enriched subnetworks representing relevant biological processes. Multiple AMI algorithms have been developed for biological datasets, and a comparative analysis of their behaviors across a variety of datasets is crucial to their application. In this study, we introduce a framework to compare and aggregate the modules produced by multiple AMI algorithms. We first used a modified Empirical Pipeline to validate the output of four AMI algorithms - PAPER, DOMINO, FDRnet, and HotNet2 - and find that no single algorithm performs well across the different datasets. Using the Earth Movers Distance to measure pairwise module similarity, we find that the outputs of different algorithms are structurally distinct, suggesting that each captures different aspects of the underlying biology. These findings suggest that a comprehensive analysis requires the aggregation of outputs from multiple algorithms. We propose two methods to this end: a spectral clustering approach for module aggregation, and an algorithm that combines modules with similar network structures called Greedy Conductance-based Merging (GCM). The merging algorithm not only allows researchers to obtain a set of cohesive modules from multiple algorithms, it also has the potential of identifying "hidden" genes that are not present in the original input data from the network. Overall, our results advance our understanding of AMI algorithms and how they should be applied. Tools and workflows developed in this study will facilitate researchers working with GGI and AMI algorithms to enhance their analyses. Our code is freely available at https://github.com/LiuJ0/AMI-Benchmark/.

bioinformatics↗

bioRxiv2025

scRGP: Prediction of Single-cell Genetic Perturbation Transcriptional Responses based on Rank in Multiple Scenarios

Single-cell perturbation sequencing technologies (e.g., Perturb-seq, CROP-seq), which integrate CRISPR-based gene editing with single-cell transcriptome profiling, have revolutionized the analysis of transcriptomic changes induced by genetic perturbations at single-cell resolution. These technologies serve as a powerful tool for identifying key genes that inhibit tumor growth or reverse cancer cell phenotypes. However, they face two major challenges: data explosion with high experimental costs, and data complexity characterized by high dimensionality, noise, sparsity, and heterogeneity. To address these challenges, we developed the single-cell Rank-based Genetic Perturbation predictor (scRGP), the first deep learning framework leveraging gene expression rank-order information for this task. scRGP demonstrates superior performance in terms of robustness, cross-cell-line perturbation prediction, and high-throughput screening. Specifically, scRGP achieves an approximately 10-16 percentage points improvement in Pearson correlation coefficient (PCC) over state-of-the-art methods (e.g., GEARS and scFoundation) for single- and double-gene perturbation predictions, while also extending prediction capability to triple-gene perturbations. Furthermore, it outperforms these methods by approximately 5-9 percentage points in cross-cell-line predictions. These advancements promise to shift the paradigm of single-cell perturbation studies from experiment-driven to computation-driven approaches, providing new support for functional genomics and precision medicine.

bioinformatics↗

bioRxiv2025

RobPicker: A Meta Learning Framework for Robust Identification of Macromolecules in Cryo-Electron Tomograms

Accurate particle picking of macromolecules of all kinds, shapes, and sizes in cryogenic electron tomograms (cryo-ET) is critical for understanding the molecular architecture of biological systems in their native state. Current deep learning methods have shown potential in identifying macromolecules from tomograms, but they are vulnerable to issues of noise in the training dataset obtained through human or automatic labeling and imbalanced distribution of macromolecule species. To address these limitations, we developed RobPicker, a meta-learning framework that effectively mitigates these issues by automatically learning deep neural networks to correct label errors and give greater emphasis to underrepresented macromolecule species. In evaluations across diverse cryo-ET datasets with noisy labels and imbalanced species distributions, RobPicker substantially outperforms state-of-the-art methods, particularly in identifying small and rare macromolecules. The efficiency and robustness of RobPicker can also be used for rapid fine-tuning of the tilt-series alignment, leading to improved tomogram reconstruction and enabling high-resolution cellular structural biology analysis.

bioinformatics↗

bioRxiv2025

stk32a links sleep homeostasis to suppression of sensory and motor systems

Sleep is regulated by a homeostatic process and associated with an increased arousal threshold, but the genetic and neuronal mechanisms that implement these essential features of sleep remain poorly understood. To address these fundamental questions, we performed a zebrafish genetic screen informed by human genome-wide association studies. We found that mutation of serine/threonine kinase 32a (stk32a) results in increased sleep and impaired sleep homeostasis in both zebrafish and mice, and that stk32a acts downstream of neurotensin signaling and the serotonergic raphe in zebrafish. stk32a mutation reduces phosphorylation of neurofilament proteins, which are co-expressed with stk32a in neurons that regulate motor activity and in lateral line hair cells that detect environmental stimuli, and ablating these cells phenocopies stk32a mutation. Neurotensin signaling inhibits specific sensory and motor populations, and blocks stimulus-evoked responses of neurons that relay sensory information from hair cells to the brain. Our work thus shows that stk32a is an evolutionarily conserved sleep regulator that links neuropeptidergic and neuromodulatory systems to homeostatic sleep drive and changes in arousal threshold, which are implemented through suppression of specific sensory and motor systems.

neuroscience↗

bioRxiv2025

Senolytic-Resistant Senescent Cells Have a Distinct SASP Profile and Functional Impact: The Path to Developing Senosensitizers

The senescent cell (SC) fate is linked to aging, multiple disorders and diseases, and physical dysfunction. Senolytics, agents that selectively eliminate 30-70% of SCs, act by transiently disabling the senescent cell anti-apoptotic pathways (SCAPs), which defend those SCs that are pro-apoptotic and pro-inflammatory from their own senescence-associated secretory phenotype (SASP). Consistent with this, a JAK/STAT inhibitor, Ruxolitinib, which attenuates the pro-inflammatory SASP of senescent human preadipocytes, caused them to become "senolytic-resistant". Administering senolytics to obese mice selectively decreased abundance of the subset of SCs that is pro-inflammatory. In cell cultures, the 30-70% of human senescent preadipocytes or human umbilical vein endothelial cells (HUVECs) that are senolytic-resistant (to Dasatinib or Quercetin, respectively) had increased p16INK4a, p21CIP1, senescence-associated {beta}-galactosidase (SA{beta}gal), {gamma}H2AX, and proliferative arrest similarly to the total SC population (comprising senolytic-sensitive plus -resistant SCs). However, the SASP of senolytic-resistant SCs entailed less pro-inflammatory/ apoptotic factor production, induced less inflammation in non-senescent cells, and was equivalent or richer in growth/ fibrotic factors. Senolytic-resistant SCs released less mitochondrial DNA (mtDNA) and more highly expressed the anti-inflammatory immune evasion signal, glycoprotein non-melanoma-B (GPNMB). Transplanting senolytic-resistant SCs intraperitoneally into younger mice caused less physical dysfunction than transplanting the total SC population. Because Ruxolitinib attenuates SC release of pro-apoptotic SASP factors, while pathogen-associated molecular pattern factors (PAMPs) can amplify the release of these factors rapidly (acting as "senosensitizers"), senolytic-resistant and senolytic-sensitive SCs appear to be interconvertible.

cell biology↗

bioRxiv2025

Actin-modulated nuclear shape controls adaptive reprogramming in cancer-associated fibroblasts

Cancer-associated fibroblasts (CAFs) exhibit striking plasticity, enabling them to adapt to external cues such as therapeutic and mechanical stress in the tumor microenvironment (TME). Here, we uncover a shared mechanotransduction pathway through which both BRAF inhibition and matrix stiffness converge in nuclear remodeling to drive CAF activation. Mechanistically, BRAF inhibitors (BRAFis) accelerate RAS-dependent RAF homo and heterodimerization and ERK signaling, leading to GSK-3{beta} inactivation and Rho kinase (ROCK) activation. In parallel, stiff substrates engage integrin receptors to directly activate ROCK signaling. Activation of ROCK induces actin stress fiber assembly, generating mechanical forces that deform the nucleus. In both contexts, nuclear reshaping promotes {beta}-catenin translocation and actin polymerization through a feedback loop that continuously enforces CAF activation and promotes melanoma progression. Notably, pharmacological inhibition of ROCK activity blocks both BRAFi- and stiffness-induced nuclear remodeling and {beta}-catenin accumulation, identifying the ROCK-cytoskeleton-nucleus axis as a critical regulator of CAF adaptation and functionality. Collectively, these findings reveal a mechanically tuned nuclear signaling hub that integrates chemical and physical cues to promote stromal adaptation and suggest ROCK inhibition as a strategy to counteract therapy-induced fibroblast reprogramming and improve therapy response.

cancer biology↗

bioRxiv2025

Myocyte Enhancer Factor 2A Orchestrates Vascular Redox Homeostasis via Direct Transcriptional Activation of SIRT1

Myocyte enhancer factor 2A (MEF2A), a transcription factor implicated in coronary artery disease, remains unexplored in vascular redox regulation. To address this gap and overcome limitations of current antioxidant therapies, we investigated MEF2As role in oxidative defense using human umbilical vein endothelial cells (HUVECs) and murine models. Adenoviral vectors encoding MEF2A-specific shRNA or mRNA were employed to silence or overexpress MEF2A in HUVECs. For in vivo validation, endothelial-targeted MEF2A knockdown was achieved via AAV1-shRNA delivery in mice fed a high-fat diet. Systemic redox status was assessed by measuring reactive oxygen species (ROS), glutathione homeostasis (GSH/GSSG ratio), NADH/NAD+ balance, mitochondrial membrane potential ({Delta}{Psi}m), and 8-hydroxy-2'-deoxyguanosine (8-OHdG). Mechanistic insights were derived from immunofluorescence, qPCR, western blotting, and dual-luciferase reporter assays. MEF2A silencing induced redox imbalance, characterized by elevated ROS, reduced GSH/GSSG ratio, and {Delta}{Psi}m collapse. Conversely, MEF2A overexpression synergized with SIRT1 to restore glutathione pools, maintain NAD+ homeostasis, and suppress ROS under oxidative stress. Chromatin immunoprecipitation confirmed direct MEF2A binding to two cis-elements in the SIRT1 promoter, driving transcriptional activation. In vivo, MEF2A-deficient mice exhibited amplified vascular oxidative damage, including elevated DNA damage marker (8-OHdG) and ROS levels. Downregulation of SIRT1/PGC-1 in MEF2A silenced cells was verified in vivo. Our findings establish MEF2A as a master regulator of endothelial redox defense via the SIRT1-PGC-1 axis, offering a mechanistic foundation for targeting oxidative cardiovascular disorders. This work suggests pharmacological MEF2A activation as a novel strategy for precision antioxidant therapy in vascular medicine.

molecular biology↗

bioRxiv2025

Structural basis of DNA N6-adenine methylation in eukaryotes

As a transcription-associated epigenetic mark abundantly present in many unicellular eukaryotes, DNA N6-adenine methylation (6mA) is maintained by the AMT1 complex. However, the structure-function relationship of the AMT1 complex remains to be fully elucidated. Here, we employ AlphaFold3 (AF3) for de novo structural modeling of the AMT1 complex, complemented by molecular dynamics simulations. Functional relevance of key structural features, especially those of the AMT1 ternary complex with the double-stranded DNA (dsDNA) substrate and the cofactor, are validated by extensive mutagenesis, both in vivo and in vitro. Our analysis reveals the structural basis for DNA substrate recognition, base flipping, and catalysis in this prototypical eukaryotic DNA 6mA-methyltransferase. It also allows us to delineate the reaction pathway for processive methylation involving translocation of the AMT1 complex along the dsDNA. As the active site is conserved across the MT-A70 family of eukaryotic N6-adenine methyltransferases, the structural insight will facilitate rational design of specific inhibitors.

molecular biology↗

bioRxiv2025

Unsupervised Multi-scale Segmentation of Cellular Cryo-electron Tomograms with Stable Diffusion Foundation Model

We introduce an unsupervised approach for segmenting multiscale subcellular objects in 3D volumetric cryo-electron tomography (cryo-ET) images, addressing key challenges such as large data volumes, low signal-to-noise ratios, and the heterogeneity of subcellular shapes and sizes. The method requires users to select a small number of slabs from a few representative tomograms in the dataset. It leverages features extracted from all layers of a Stable Diffusion foundation model, followed by a novel heuristic-based feature aggregation strategy. Segmentation masks are generated using adaptive thresholding, refined with CellPose to split composite regions, and then utilized as pseudo-ground truth for training deep learning models. We validated our pipeline on publicly available cryo-ET datasets of S. Pombe and C. Eleganscell sections, demonstrating performance that closely approximates expert human annotations. This fully automated, data-driven framework enables the mining of multi-scale subcellular patterns, paving the way for accelerated biological discoveries from large-scale cellular cryo-ET datasets.

bioinformatics↗

bioRxiv2025

5-azacytosine induces cytotoxicity via 5-methylcytosine depletion on chromatin-associated RNA in leukemia

5-azacytidine (5-azaC) is a DNA hypomethylating agent clinically used to improve outcomes in myeloid malignancies. However, 5-azaC treatment causes gene dysregulation inconsistent with DNA hypomethylation changes, suggesting alternative mechanisms of action by 5-azaC. As a ribonucleoside analogue, 5-azaC is more readily incorporated into nascent RNA. Here, we demonstrate that RNA 5-methylcytosine (m5C) depletion by 5-azaC treatment, particularly at early time points, is sufficient to induce leukemia cell death. In contrast to its DNA demethylation function, the RNA-dependent effect of 5-azaC causes transcriptional repression, disrupting genes involved in cell cycle regulation and DNA repair. Mechanistically, 5-azaC impairs two specific m5C-mediated transcriptional regulatory pathways. First, depletion of m5C in chromatin-associated RNA (caRNA) disrupts the MBD6-mediated H2AK119ub deubiquitination. In parallel, this also impairs SRSF2 recruitment and the downstream H3K27ac deposition by p300. Indeed, loss of the caRNA methyltransferase NSUN2 caused prolonged cell cycle, defective DNA repair, and shifted hematopoietic lineage commitment toward erythropoiesis, mirroring the effects of 5-azaC treatment. Furthermore, we performed a leukemia cell line screen and identified that TET2 and IKZF1 depletion can sensitize 5-azaC treatment, consistent with the observed RNA-dependent cytotoxicity of 5-azaC in leukemic cells. In summary, our findings highlight the transcription repression by 5-azaC through depleting caRNA m5C, providing additional insights into the mechanism of action for 5-azaC, the prediction of its efficacy, and future directions for therapy developments based on 5-azaC. HIGHLIGHTO_LIRNA-dependent effects of 5-azaC are sufficient to drive leukemia cell cytotoxicity through transcriptional repression. C_LIO_LI5-azaC-induced caRNA m5C depletion impairs MBD6 binding and H2AK119ub deubiquitination. C_LIO_LI5-azaC-induced caRNA m5C depletion disrupts SRSF2 chromatin-binding, impeding p300 recruitment and H3K27ac deposition. C_LIO_LITET2 or IKZF1 depletion synergizes leukemia sensitivity to 5-azaC C_LI

molecular biology↗

bioRxiv2025

A Novel 'Three-in-one' Mucosal Vaccine Elicits Broad Protection against Three Distinct Clusters of ACE2-using Sarbecoviruseses

With the persistent risk of coronavirus epidemics, the development of universal vaccines is urgently needed and challenging to achieve. In this study, we aimed to design a multivalent immunogen to provide broad protection against diverse ACE2-using sarbecoviruses, which pose a significant threat to global health. Our analysis revealed that the receptor-binding domains (RBDs) of ACE2-using sarbecoviruses segregate into three antigenically distinct clusters, including one cluster that encompasses viruses related to SARS-CoV-2. Based on these findings, we engineered a three-in-one immunogen, designated as 3Rs-NC, which incorporates representative RBDs from all three clusters. 3Rs-NC preserved the natural epitope configuration of each monomeric RBD component, and efficiently elicited high levels of neutralizing antibodies against representative viruses and closely related sarbecoviruses. When administered intranasally with flagellin-based mucosal adjuvant KFD, 3Rs-NC induced robust and durable RBD-specific serum IgG and mucosal IgA responses in mice. Furthermore, KFD-adjuvanted 3Rs-NC conferred sustained protection in both the upper and lower respiratory tracts against SARS-CoV-2 Omicron BA.1 and SARS-like coronavirus WIV1. Additionally, 3Rs-NC immunization protected mice from lethal challenge of SARS-like coronavirus rRsSHC014S, with more efficient protection observed in female mice than male mice. This needle-free three-in-one vaccine represents a promising candidate for a universal mucosal vaccine against ACE2-using sarbecoviruses and could serve as a foundational component of three-in-one vaccine series to form a comprehensive coronavirus vaccine package. HighlightO_LInovel three-in-one immunogen 3Rs-NC induced a high level of neutralizing antibodies against three distinct ACE2-using Sabecovirus clusters. C_LIO_LIadministration of 3Rs-NC plus KFD adjuvant (3Rs-NC+KFDi.n) generated potent and long-lasting RBD-specific serum IgG and mucosal IgA. C_LIO_LI3Rs-NC+KFDi.n provided long-term protection in both upper- and lower-respiratory tracts in mice. C_LIO_LI3Rs-NC+KFDi.n could serve as a foundational component for next-generation coronavirus vaccine strategies. C_LI

immunology↗

bioRxiv2025

KRT6A and KRT17 mark distinct stem cell populations in the adult palpebral conjunctiva and Meibomian gland

PurposeTo investigate whether KRT6A or KRT17 label self-renewing stem cells (SCs) in adult Meibomian gland (MG) and palpebral conjunctiva (PC) homeostasis and explore the mechanisms that regulate their expression in these tissues. MethodsKRT6A and KRT17 expression in adult mouse MG and PC was examined by single nucleus RNA sequencing and immunofluorescence (IF). Bromodeoxyuridine pulse-chase assays were conducted to assess the turnover rate of PC progenitor cells. Lineage tracing experiments were performed using Krt6a-CreERT2and Krt17-CreERT2 mice carrying Rosa26RnTnG or Rosa26RmTmG reporter. As Hedgehog (Hh) signaling, the histone deacetylase HDAC3, and the transcription factor KLF4, regulate KRT6A and KRT17 in other contexts, IF was conducted to assess the in vivo effects of overexpression of the Hh pathway activator GLI2{Delta}N, and inducible epithelial deletion of Hdac3 or Klf4, on KRT6A and KRT17 expression in the MG and PC. ResultsKRT6A and KRT17 are primarily expressed in the MG central duct and ductules. KRT6A also shows robust expression in PC. MG and PC epithelial progenitor cells turn over within 14 days. Lineage tracing indicated that Krt17 labels self-renewing SCs in the MG whereas Krt6a labels SCs in the PC. GLI2{Delta}N overexpression induced ectopic KRT17 expression in MG acini and PC but did not affect KRT6A expression in either MG or PC. Hdac3 deficiency caused expanded expression of KRT6A and KRT17 in MG acini and ectopic KRT17 expression in PC. Klf4 deletion resulted in ectopic KRT17 expression in PC but did not influence KRT6A expression in MG or PC. ConclusionsKrt6a- and Krt17-expressing cells contribute to adult PC and MG homeostasis, respectively. KRT17 expression is enhanced by GLI2{Delta}N, and suppressed by HDAC3 and KLF4, whereas KRT6A expression is controlled only by HDAC3.

cell biology↗

bioRxiv2025

Quantum Cognition Machine Learning for Forecasting Chromosomal Instability

The accurate prediction of chromosomal instability from the morphology of circulating tumor cells (CTCs) enables real-time detection of CTCs with high metastatic potential in the context of liquid biopsy diagnostics. However, it presents a significant challenge due to the high dimensionality and complexity of single-cell digital pathology data. Here, we introduce the application of Quantum Cognition Machine Learning (QCML), a quantum-inspired computational framework, to estimate morphology-predicted chromosomal instability in CTCs from patients with metastatic breast cancer. QCML leverages quantum mechanical principles to represent data as state vectors in a Hilbert space, enabling context-aware feature modeling, dimensionality reduction, and enhanced generalization without requiring curated feature selection. QCML outperforms conventional machine learning methods when tested on out of sample verification CTCs, achieving higher accuracy in identifying predicted large-scale state transitions (pLST) status from CTC-derived morphology features. These preliminary findings support the application of QCML as a novel machine learning tool with superior performance in high-dimensional, low-sample-size biomedical contexts. QCML enables the simulation of cognition-like learning for the identification of biologically meaningful prediction of chromosomal instability from CTC morphology, offering a novel tool for CTC classification in liquid biopsy.

cancer biology↗

bioRxiv2025

MedSAGE: Bridging Generative AI and Medicinal Chemistry for Structure-Based Design of Small Molecule Drugs

While generative AI is transforming the de novo design of proteins, its effectiveness for structure-based design of small molecules remains limited. Current methods, including diffusion models, often produce small molecules with difficult-to-synthesize structures, poor medicinal chemistry properties, and limited target selectivity. To address these limitations, we introduce MedSAGE, a novel generative AI framework that adapts diffusion models specifically for de novo small-molecule design. Rather than using atoms or strings, we develop a novel representation for generating molecules using fragments relevant for medicinal chemistry. Chemical and geometric information characterizing these fragments is embedded in a smooth and interpretable latent space. We also develop an algorithm to optimize the connectivity between generated fragments while preserving chemical validity and synthesizability. In a benchmark of multiple methods across 25 therapeutically relevant protein targets, MedSAGE achieved state-of-the-art performance, producing synthesizable, drug-like molecules with predicted affinity and selectivity closely matching known drugs and drug candidates. Compared to large-scale virtual screening, MedSAGE produced molecules with high predicted affinity over 100 times more efficiently. Our results demonstrate that MedSAGE is already practically useful and paves the way for next-generation tools in structure-guided drug design.

biophysics↗

bioRxiv2025

Prevalent cross-cell type QTL trans-regulatory genetic effects impacting innate lymphoid cells in the small intestine

Tissue function in homeostasis and disease arise from coordinated interactions between different cell types and can vary between individuals in a population, in part due to the impact of genetic variants. In barrier tissues such as the gut, innate lymphoid cells (ILCs) play critical roles in maintaining tissue homeostasis and immunity, but the genetic basis and associated regulatory circuitry affecting tissue ILCs remain largely unknown. Here, we systematically mapped the regulatory functions of genetic variants across 273,370 gut ILCs profiled by single cell RNA-seq from 274 diversity outbred (DO) mice. Computational analysis identified quantitative trait loci (QTL) impacting ILCs at three levels: cell subtype-specific mRNA expression (local and distal), cell subset proportions, and ILC gene expression programs (in trans), of which we experimentally validated a pivotal role for the transcription factor Rbpj in regulating ILC3s. All three classes of ILC QTL display polygenic inheritance signatures and broadly overlap with QTL affecting peripheral cytokine levels that we measured in another 261 DO mice. Strikingly, nearly half of trans-QTL affecting ILC traits did not overlap (within a 50-kb window) genes expressed in ILCs, and such loci were enriched for genes expressed by non-ILCs especially enteric neurons and intestinal glia. This suggests that such loci may act instead across cell type boundaries and allowed us to recover causal trans-cell type regulatory circuits in the tissue, including loci encoding multiple neuron-expressed peptides, epithelial-cell expressed Sox9, and phagocyte-expressed Ccl17. Finally, human orthologs of genes in QTL impacting gene expression and cell proportions are enriched for autoimmune disease risk heritability, suggesting their relevance to human disease. Our findings highlight how the impact of genetic variants may propagate to maintain tissue homeostasis and show a path to understand causality in tissue biology using quantitative genetics.

genetics↗

bioRxiv2025

Isoform switch of CD47 provokes macrophage-mediated pyroptosis in ovarian cancer

Cancer cells expressing CD47 escape macrophage phagocytosis by binding to the SIRP ligand expressed on macrophages. CD47 targeted therapy offers a promising approach to cancer treatment. Here we report that two major isoforms of CD47 differ in ovarian cancer and normal tissues. The truncated isoform lacking exon 9 and exon 10 (CD47-S) was exclusively expressed in normal tissues, whereas the full-length isoform (CD47-L) was predominantly expressed in ovarian cancer tissues. Interestingly, CD47-S was unable to locate at the cell surface to bind with SIRP as CD47-L did, and thereby inactivated "dont-eat-me" signal. Mechanistic investigations revealed that splicing factor HNRNPA1 promoted splicing switch from CD47-L to CD47-S through exon 9 and exon 10 skipping. We further developed antisense oligonucleotides (ASOs) that effectively switched CD47-L to CD47-S. Importantly, ASOs treatment evoked macrophage-mediated antitumor immune response, thereby triggered pyroptosis of ovarian cancer cells. Moreover, CD47-targeting ASOs significantly reduced tumor growth in patient-derived xenograft. Together, ASO-mediated isoform switch of CD47 has emerged as a promising strategy to improve immune responses against tumors.

cancer biology↗

bioRxiv2025

Object categorization based on abstract concepts by monkeys, humans, and vision models

Humans categorize objects at multiple levels of abstraction--animate versus inanimate, big versus small, and many other attributes. Despite its apparent challenge, the advent of deep neural networks (DNNs) has demonstrated that complex visual processing alone can support such classification without language or human-specific knowledge. This raises a natural question: to what extent can non-human primates, without language, perform such categorization? Although basic object-recognition behavior in monkeys such as similarity judgment has been extensively studied, their ability to classify objects across diverse rules remains poorly characterized. Here, we developed a task paradigm that enabled us to train monkeys on a large battery of binary classification tasks using natural object images, spanning more than 10 rules, such as animate versus inanimate, natural versus man-made objects, and mammalian versus non-mammalian animals. Monkeys acquired each rule in a few days, generalized the learned rules to new images, and exhibited error patterns consistent with human judgments. At the same time, their classification performance correlated more strongly with that of visual DNNs trained without language input, whereas human performance was better explained by language-informed DNNs. These results provide an important benchmark for the capacity of biological neural networks to perform image classification without language and human-specific knowledge.

neuroscience↗