Biology subjects

Zhang, S.

Publications and source records attributed to Zhang, S..

At least 217 records · Page 12Linked to original sources

bioRxiv2025

A stem cell knockout village reveals lineage rewiring and a non-canonical islet cell fate in monogenic diabetes

Genetics studies have identified a core set of regulators essential for pancreatic {beta} cell development, many of which are mutated in monogenic diabetes. However, how these mutations alter developmental trajectories to produce pathological cell states remains elusive. Here we introduce a knockout village framework that enables longitudinal scRNA-seq profiling of 79 human pluripotent stem cell mutant lines targeting 30 developmental regulators, including 15 diabetes genes, across five islet differentiation stages. We show that loss of lineage regulators impairs {beta} cell formation in a stage-specific manner and rewires developmental trajectories towards competing lineages. Notably, several monogenic diabetes gene mutations drive a shift from {beta} cells to enterochromaffin (EC)-like cells, a recently recognized non-canonical islet cell fate. These EC-like cells exhibit incomplete activation of hormone regulation programs, along with elevated neuron signatures. Leveraging the diversity of cell fate outcomes across mutants, we predicted and experimentally validated ISL1 as a key downstream effector of PDX1 and PAX6 that safeguards {beta} cell identity against an EC-like fate. Together, our findings reveal cell fate rewiring as a widespread, previously underappreciated pathological mechanism in monogenic diabetes and establish a scalable platform for uncovering developmental vulnerabilities in human genetic disorders.

developmental biology↗

bioRxiv2025

nAPOGEE: A machine-learning platform for clinically actionable pathogenicity assessment of all mitochondrial noncoding variants

Mitochondrial noncoding variants, particularly those in tRNA and rRNA genes, pose significant challenges for clinical interpretation due to heteroplasmy, broad phenotypic heterogeneity where symptoms can overlap with other conditions, and the limited availability of well-established genotype-phenotype correlations. Despite their central role in mitochondrial translation, these variants have remained largely unexplored by the existing variant-effect predictors. Here, we present nAPOGEE, a novel machine-learning framework specifically designed to assess the pathogenicity of all possible single-nucleotide variants in human mitochondrial noncoding RNAs. nAPOGEE integrates two specialized predictors: tAPOGEE, which outperforms existing tools for tRNAs, and rAPOGEE, the first dedicated classifier for mitochondrial rRNA variants. Using curated training datasets, phylogenetic conservation metrics, secondary structure modeling, RNA-specific embeddings, and thermodynamic features, nAPOGEE provides biologically interpretable predictions and posterior probabilities aligned with the ACMG/AMP guidelines. Applied to both curated variant sets and population-scale data, nAPOGEE revealed consistent spatial correlation of the predicted pathogenicity, reflecting underlying structural and evolutionary constraints. This study addresses a longstanding gap in mitochondrial genomics and offers a clinically applicable tool for variant prioritization, reclassification, and research into mitochondrial disease mechanisms.

bioinformatics↗

bioRxiv2025

Mortality Risk-Stratified Septic Serum Depresses Contractility and Mitochondrial Function in Human Induced Pluripotent Stem Cell-Derived Cardiomyocytes

BackgroundSepsis-associated myocardial dysfunction (SAMD) is common in children with septic shock, is independently associated with mortality, and has no disease-modifying treatments. Differences in murine cardiomyocyte biology and repeated failures to translate discoveries into novel therapies for septic shock underscore a key translational need for human-relevant disease modeling. We sought to investigate human induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) exposed to mortality risk-stratified septic serum as a model of SAMD. MethodsSerum from children with septic shock (n=120) was stratified by Pediatric Sepsis Biomarker Risk Model (PERSEVERE) II mortality probability as low, intermediate, or high risk. We conducted aptamer-based proteomic analysis of septic serum to determine differentially expressed proteins in children with high compared to low mortality risk. We treated iPSC-CMs with risk-stratified septic serum and defined contractile and mitochondrial functional and transcriptomic responses. ResultsWe found 612 differentially expressed proteins in children with high mortality probability, most prominently interleukins (IL)-6 and -8. High-risk septic serum reversibly depressed iPSC-CM contractility as measured by percent shortening, while low-risk septic serum had no impact. Further, high-risk septic serum depressed basal mitochondrial respiration, maximum uncoupled respiration, and coupled oxidative phosphorylation in iPSC-CMs relative to low-risk serum. We identified distinct patterns of gene expression due to risk-stratified serum with 5,293 differentially expressed genes, including upregulation of acute phase reactants and apolipoproteins and downregulation of chemokines, as well as transcriptional changes reflective of chronic IL-6 and IL-8 signaling. ConclusionsSeptic serum from children with high mortality risk exhibited distinct proteomic signatures, notably enriched for IL-6 and IL-8. Human iPSC-CMs differentially responded to risk-stratified septic serum, recapitulating phenotypic features of SAMD including reversible contractility depression and mitochondrial dysfunction with high-risk septic serum. These findings establish mortality risk-stratified septic serum exposure of iPSC-CMs as a human-relevant translational platform to interrogate mechanisms of myocardial dysfunction in septic shock.

immunology↗

bioRxiv2025

Batch Effects Remain a Fundamental Barrier to Universal Embeddings in Single-Cell Foundation Models

Constructing a cell universe requires integrating heterogeneous single-cell RNA-seq datasets, but is hindered by diverse batch effects. Single-cell foundation models (scFMs), inspired by large language models, aim to learn universal cellular embeddings from large-scale single-cell data. However, unlike language, single-cell data are sparse, noisy, and strongly affected by batch effects that limit cross-dataset transferability. Our systematic evaluation across diverse batch scenarios reveals that current scFMs fail to effectively remove batch effects, with batch signals persisting in pretrained embeddings. Post-hoc batch-centering partially improves alignment, highlighting the need for future scFMs to integrate explicit batch-effect correction mechanisms to achieve true universal cellular embeddings.

bioinformatics↗

bioRxiv2025

6mA demethylase Mi-NMAD-1 modulates nematode development and parasitism through the transcription factor Mi-nhr-118

N6-methyladenine (6mA) DNA methylation has emerged as an important epigenetic mark in eukaryotes, but its biological role in plant-parasitic nematodes remains largely unexplored. Here, we demonstrate that 6mA methylation is essential for the embryonic development and parasitic ability of Meloidogyne incognita. Silencing of the 6mA demethylase genes minmad-1 significantly reduced egg hatching, revealing that 6mA-mediated regulation is required for nematode development. Transcriptome analyses showed that minmad-1 knockdown downregulated genes involved in embryogenesis, RNA biosynthesis, and cell differentiation, suggesting a critical role of 6mA in developmental gene regulation. Among the transcription factors affected by minmad-1 silencing, Mi-nhr-118, a nuclear hormone receptor orthologous to C. elegans nhr-118, was identified as a key regulator of embryonic development. Functional assays confirmed that RNAi or host-induced gene silencing (HIGS) of Mi-nhr-118 led to developmental arrest, reduced gall and egg mass formation, and impaired infectivity across multiple Meloidogyne species. ChIP-seq analysis revealed that Mi-nhr-118 directly binds to promoters of genes associated with cellular regulation and developmental signaling. Several downstream targets, including transcription factors (Mi_23621.1, Mi_39647.1) and genes implicated in metabolic and signaling processes (Mi_03067.1, Mi_08398.1), were shown to be essential for both embryogenesis and parasitism. Functional enrichment analyses of Mi-nhr-118 target genes indicated suppression of immune-related signaling, carbohydrate recognition, and glycoprotein biosynthetic pathways following gene silencing. Collectively, these findings establish Mi-nhr-118 as a central regulator linking 6mA-mediated epigenetic control to development and parasitism in M. incognita.

molecular biology↗

bioRxiv2025

STUltra: scalable and accurate integration for subcellular-level spatial omics data

High-resolution spatial transcriptomics (ST) is generating massive datasets involving multiple tissue sections, continuous developmental stages, or various disease conditions. Yet, their scale, biological heterogeneity, and substantial batch effects hinder data integration for further analysis, particularly the temporal interpretation of disease-associated spatial domains within tissue organization landscapes. Here, we present STUltra, a scalable hypergraph framework that integrates multi-sample ST data for precise domain detection and genome-wide association study (GWAS)-based disease trait mapping. From ST datasets, STUltra first constructs interval-sampled integrative hypergraphs, in which hyperedges capture tissue neighborhoods within the slices as well as shared biological contexts across the slices. It then combines a robust graph autoencoder with contrastive learning to learn batch-corrected, spatially informed embeddings for identifying spatial domains across the tissue sections. These embeddings are then coupled with GWAS statistics to map trait-associated signals onto the integrated tissue sub-structure landscapes spanning different sections, developmental stages, and disease conditions. STUltra substantially outperforms competing methods, especially being much faster, on diverse integration benchmarks, enabling efficient processing of datasets containing millions of spots. As a case study on a mouse embryo dataset, STUltra recovered continuous developmental programs and mapped congenital heart disease signals beyond cardiac regions to broader programs including vascular and mesenchymal domains. On a mouse Alzheimer's disease dataset, STUltra prioritized localized AD-associated genetic signals to microglia and a disease-expanded C1QA/HEXB-high astrocyte state. Together, STUltra provides a scalable framework to bridge the gap between spatiotemporal tissue structures and complex disease traits, facilitating discoveries from previously unrecognized molecular and cellular architectures across diverse contexts.

bioinformatics↗

bioRxiv2025

Distributed fMRI patterns coupled to low-frequency cardiorespiratory dynamics provide markers of aging

How aging affects brain-body connections can be investigated through changes in the coupling between functional magnetic resonance imaging (fMRI) signals and bodily autonomic processes across the adult lifespan. Recent studies using univariate approaches have identified age-related changes in the association between fMRI signals from multiple individual brain regions and low-frequency respiratory and cardiac activity. Here, we investigate if whole-brain spatial fMRI patterns associated with low-frequency physiological processes (heart rate and respiratory volume fluctuations) present generalizable changes with age. Data from human participants of both sexes are included in the analysis. We find that chronological age can be predicted statistically beyond chance from patterns of low-frequency fMRI-physiology coupling, even after accounting for individual differences in physiological signal characteristics and brain anatomy. Notably, brain areas implicated in central autonomic regulation, including nodes within salience and ventral attention networks (e.g., insula and middle cingulate cortex), are amongst the strongest contributors to age prediction. Further, we observe that after removing physiological effects from fMRI data, the residual blood oxygen level-dependent (BOLD) signal variability is still a reliable indicator of age. Together, these findings underscore the close integration between brain and body physiology, and highlight this interaction as a potential biomarker of the aging process. Significance StatementThe association between brain activity and respiratory or cardiac activity is often dismissed as "noise" in functional magnetic resonance imaging (fMRI) studies. However, emerging evidence suggests that coupling between fMRI and peripheral physiological signals can provide valuable insight into the brain-body connection. In this study, we show that whole brain patterns of coupling between fMRI signals and low-frequency respiratory and cardiac processes can reliably predict age across the adult lifespan. Brain regions involved in autonomic regulation, such as insula and cingulate cortex, were among the most informative predictors of age. These findings suggest that fMRI-physiology coupling may capture aging-related changes in brain vascular health and autonomic function and may have broader relevance for tracking disease-related disruptions in brain-body interaction.

neuroscience↗

bioRxiv2025

Single-cell atlas reveals alveolar macrophage-specific Oncostatin-M drives pulmonary fibrosis via fibroblast activation in PCV2d infection

A hallmark of pulmonary fibrosis is the aberrant activation of lung fibroblasts into pathological fibroblasts that produce excessive extracellular matrix(1). Clarifying the cell-specific mechanisms by which viruses induce this process is crucial for developing effective strategies to intervene in disease progression. Here, we demonstrate that alveolar macrophage subsets specifically express tumor suppressor M (Osm) after infection by PCV2d infection and related intervention experiments in mice, combined with single-cell transcriptome sequencing (scRNA-seq). Through validation using histopathological analyses and an in vitro macrophage-fibroblast co-culture model, these alveolar macrophage-derived oncostatin M (Osm) were confirmed to specifically drive fibroblast activation. Vaccine intervention experiments further confirmed that targeted inhibition of oncostatin M (Osm) expression in alveolar macrophages significantly attenuated fibroblast activation and extracellular matrix deposition, thereby mitigating pulmonary fibrosis progression. Therefore, subgroup-specific secretion of oncostatin M (Osm) by alveolar macrophages represents a key driver mediating PCV2d infection-induced pulmonary fibrosis, and targeting this pathway offers a potential therapeutic strategy for viral-induced pulmonary fibrosis.

genomics↗

bioRxiv2025

geneSCOPE: gene Spatial Co-Occurrence of Pairwise Expression

Spatial transcriptomics captures neighborhood-dependent gene expression, but existing workflows do not always fully account for measurement scale and often treat space implicitly. We present geneSCOPE, a framework that integrates ecology-inspired statistics with network analysis. Molecules are binned on a grid whose width is chosen near the mode of the per-gene unit-invariant knee (UIK) distribution derived from Morisitas I{delta} -width curves. Pairwise adjacency-weighted spatial association is quantified with Lees L. We then assemble a spatial gene network and identify gene modules by consensus clustering. To identify cell-cell interactions between different cell types, high Lees L and low Pearsons r is examined. Applied to human colorectal cancer (three Xenium sections) and a lymph node, geneSCOPE recovered spatial gene modules that map to microanatomy such as invasive margins, luminal epithelium, fibroblast-rich territories and germinal-center subdomains, and highlights intercellular neighborhood patterns at tumor-stroma interfaces between LGR5-marked stem-like tumor programs and C3-centered fibroblast/complement-associated niches. geneSCOPE thus provides a scalable, interpretable analytical foundation that generalizes across spatial omics and supports a wide range of applications.

bioinformatics↗

bioRxiv2025

Mapping the Glycogen Synthase 1 Interactome in Brown Adipocytes

Brown adipose tissue (BAT) is distinguished by its ability to dissipate energy through thermogenesis, a process with considerable therapeutic potential for obesity, type 2 diabetes, and related metabolic disorders. Although glycogen represents a relatively minor energy reserve in adipocytes compared with liver and muscle, emerging evidence shows that glycogen turnover increases sharply during metabolic transitions and is required for full thermogenic activation in brown adipocytes. The rate-limiting enzyme glycogen synthase 1 (GYS1) catalyzes glycogen chain elongation, and adipose-specific Gys1 deletion impairs adaptive thermogenesis, highlighting the importance of GYS1 in coordinating glucose and lipid metabolism; however, the mechanisms regulating GYS1 in BAT remain poorly defined. To address this gap, we used an unbiased proximity-labeling strategy in brown adipocytes by fusing GYS1 to the biotin ligase TurboID and identifying biotinylated proteins through mass spectrometry, revealing 425 putative interactors. These included canonical glycogen-handling proteins such as glycogenin and glycogen phosphorylase, validating the approach, while gene ontology analysis uncovered broader enrichment in cytoskeletal remodeling, vesicle trafficking, and transcriptional regulation, suggesting previously unrecognized roles for GYS1 beyond glycogen synthesis. Notably, we identified several novel candidate interactors, including CAST, PDAP1, and CCDC102A. Together, these findings define the most comprehensive GYS1 interactome reported to date and provide new insights into the molecular mechanisms linking glycogen metabolism to thermogenic function in BAT.

biochemistry↗

bioRxiv2025

DPCMHC: efficient prediction of MHC-peptide binding affinity by deep learning based on dual-padding convolution

Predicting peptide binding affinities to major histocompatibility complex class I (MHCI) is a crucial challenge in immunological bioinformatics and is essential for identifying neoantigens in personalized cancer vaccines. Current deep learning methods often struggle, especially with 10-mer and 11-mer peptides. To address this, we developed DPCMHC, an advanced deep learning model featuring an embedding module, a dual-padding convolutional module, a BiLSTM module, and an output module. This design enhances the models understanding of amino acid sequences and their lengths. DPCMHC effectively captures the information on the beginning and end of amino acids, as well as the diverse sizes of adjacent amino acids. Using concatenation, the model extracts continuous sequence information. We rigorously evaluated DPCMHC on three benchmark datasets, demonstrating its superior or comparable performance to existing state-of-the-art methods. Our validation results confirm that DPCMHC is a robust and efficient tool for predicting MHC-peptide binding affinities.

bioinformatics↗

bioRxiv2025

Dorsomedial hypothalamus cold-sensitive neurons drive shivering thermogenesis

Hypothermia, a decline in core body temperature below 35{degrees}C in homeotherms, elicits rapid, asynchronous involuntary shivering thermogenesis; however, the hypothalamic neurons responsible for initiating this response remain uncharacterized. Here, retrograde transsynaptic tracing combined with RNA-seq identified a distinct population of VGLUT2-positive neurons enriched with Calbindin D28k (CALB1) within the dorsomedial hypothalamic nucleus (DMHCalb1/Vglut2). Convergent evidence from chemogenetic, optogenetic, fiber photometry calcium imaging, and electromyography (EMG) experiments demonstrated that these neurons selectively mediate hypothermia-triggered shivering thermogenesis. Bidirectional perturbation of DMHCalb1/Vglut2 activity modulated shivering intensity, while CALB1-negative VGLUT2-expressing DMH neurons were responsible for non-shivering thermogenesis (NST). Genetic ablation of afferent inputs from LSI/POA/VLPAG projecting to DMHCalb1/Vglut2, as well as ex vivo pharmacological synaptic blockade, failed to eliminate shivering responses during hypothermia, indicating that their activation is independent of canonical cold-responsive circuits. Hypothermia significantly suppressed transient A-type K (IA) currents, without affecting delayed rectifier K (IK) currents, accompanied by a pronounced reduction in the IA inactivation time constant relative to normothermic 36 {degrees}C. Collectively, hypothermia elevates DMHCalb1/Vglut2 excitability through IA-dependent reshaping of action potential dynamics. Shortened action potential repolarization and afterhyperpolarization duration cooperatively boost neuronal firing, constituting an intrinsic mechanism for DMHCalb1/Vglut2 neuron-driven shivering initiation.

neuroscience↗

bioRxiv2025

Mineralized Tissue-Targeting Expression System for Local Control of Gene Expression

Targeted control of gene expression in mineralized tissue would enable the use of nucleic acids to modulate the local microenvironment at diseased sites, ultimately promoting bone regeneration. Piperazine-linked bisphosphonate ionizable lipids provide a facile approach to targeting the transfection of mineralized tissue with lipid nanoparticles (LNPs). Here, we develop a Mineralized Tissue-Targeting Expression System (MiTEX) using bisphosphonate LNPs to locally target mineralized tissues by adsorption to mineral surfaces and bone graft materials. MiTEX demonstrated a significant increase in the adsorption of RNA onto hydroxyapatite substrates, which retained the ability to transfect bone mesenchymal cells via the adsorbed layer of mRNA LNPs. Bone graft scaffolds functionalized by adsorbed Cre mRNA-LNP were implanted to genetically label newly formed bone tissues in vivo. The surface affinity and adsorption of bisphosphonate lipids provided a local reservoir in mineralized tissues, sustaining the in vivo delivery of MiTEX. Furthermore, the targeted delivery of RNA therapeutics was demonstrated using STAT3 siRNA to modulate gene expression and proinflammatory cytokine release in ex vivo periodontal tissues. The design of this new RNA-functionalized delivery platform will promote the development of precision nucleic acid therapeutics for local anti-inflammatory treatments and bone regeneration at mineralized tissue interfaces.

bioengineering↗

bioRxiv2025

General features of cognition gene polymorphism patterns in marine fish

A growing body of evidence suggests that the cognitive abilities of fish are far more advanced than traditionally believed. They are not mere "automata" driven by instinct, but complex organisms capable of learning, remembering, solving simple problems, engaging in social interactions, and even deception. This situation has prompted us to explore genes related to cognitive functions in the fish genome, examining the pattern status of a set of SNVs (single nucleotide variations) in their cognitive genes relative to the human genome sequence. This study investigates the locus status of 1,824 SNVs from 33 human cognitive genes across the whole genomes of 90 fish species. Based on this, preliminary findings indicate that fish exhibit approximately four patterns of cognitive gene polymorphism. One of these patterns includes at least two coelacanth/lungfish samples, which directly point to the evolutionary direction of the ancient human sd1 sample (an ancient human hair sample from Sudan, Africa).

molecular biology↗

bioRxiv2025

PRSNet-2: End-to-end genotype-to-phenotype prediction via hierarchical graph neural networks

The emergence of large-scale biobanks has opened unprecedented opportunities for the development of data-driven approaches, especially deep learning-based methods, for genotype-to-phenotype (G2P) prediction. However, designing an end-to-end framework capable of directly leveraging extremely high-dimensional genotypic data while simultaneously mitigating overfitting and ensuring robust prediction remains a significant challenge. In this study, we introduce PRSNet-2, an interpretable end-to-end deep learning framework designed to predict complex phenotypes directly from large-scale genotypic data. PRSNet-2 proposes a novel hierarchical graph neural network (GNN) architecture, which first employs a multi-kernel aggregator to map high-dimensional genotypic features to gene-level representations. Next, it models gene-gene interactions through message-passing operations and uses an attention-based readout module to generate interpretable phenotypic predictions. We further introduce significance-guided regularization strategies to boost models generalizability based on prior genetic associations. Extensive empirical evaluations across multiple complex traits and diseases demonstrate that PRSNet-2 consistently outperforms a variety of baseline methods and exhibits superior capabilities in overcoming overfitting, even when modeling around half a million single nucleotide polymorphisms (SNPs). Moreover, PRSNet-2 is shown to be easily extendable to integrate multiple genome-wide association study (GWAS) datasets into a single model, thereby enhancing predictive performance for both single-phenotype and multi-phenotype prediction tasks. The inherent interpretability of PRSNet-2 further facilitates the identification of disease-relevant genes, functional gene modules, and potential therapeutic targets. In summary, PRSNet-2 offers a powerful, versatile tool for both genetic risk stratification and the discovery of biological insights.

genetics↗

bioRxiv2025

Sex and experience dependent regulation of synaptic protein turnover

Synaptic transmission can be carefully tuned through plasticity mechanisms that regulate synaptic strength, structure, and number. In vivo measurements demonstrate remarkable spine dynamics, with subsets of synapses persisting for months. This correlates with the extreme longevity of certain memories, which can persist for an organisms lifetime. The molecular basis supporting the long-term stability of specific synapses and the long-term durability of memories remains unknown. At the protein level, most proteins persist for a relatively short amount of time before they are degraded and replaced with new molecules. However, recent work has identified a population of proteins, including those present at the synapse, that are exceptionally long-lived. It has been speculated that long-lived proteins (LLPs) could contribute to long-term synapse stability, function, and memory. Here, we used stable isotope labeling in mammals (SILAM) to first identify LLPs in the post synaptic density (PSD) of the hippocampus and subsequently determine if protein turnover rates varied by sex or following learning. We identified novel synaptic LLPs and found that both sex and experience can regulate synaptic protein turnover rates. We identified sex-dependent changes in protein turnover rates in autism spectrum disorder (ASD) risk genes, including increased stability of Gabrg2, a GABA-A receptor subunit, in male mice. Furthermore, we observed stabilization of a subset of PSD proteins, such as Shank3, following contextual fear conditioning. We propose that sex and experience dependent changes in protein turnover rates could help explain sex-differences in psychiatric risk and aid our understanding of the molecular mechanisms that support learning and memory.

neuroscience↗

bioRxiv2025

Metabolomic analysis of Chinese chestnut (Castanea mollissima Bl.) leaves in response to Oligonychus ununguis feeding stress

Piercing-sucking insects such as the chestnut red mite (CRM; Oligonychus ununguis) damage leaves by extracting mesophyll cell nutrients, thereby reducing photosynthesis and nut yield in Chinese chestnut (Castanea mollissima Blume). However, the metabolic basis of the chestnut response to mite stress remains elusive. To address this, leaf metabolomics was conducted on four varieties with different resistance levels. Resistance grades were determined by the leaf chlorosis index, and three comparison groups were established. Secondary metabolome profiling revealed that resistant varieties showed greater metabolic changes under infestation than susceptible ones. In total, 713 total metabolites and 275 metabolites with significantly differential accumulation among germplasms were identified across the comparisons. KEGG enrichment highlighted 246 metabolites and 75 differential metabolites involved in flavonoid, flavonol, and phenylpropanoid biosynthesis. Notably, 27 phenylpropanoids and flavonoids accumulated significantly in resistant varieties but were largely absent or unchanged in susceptible ones. These findings suggest that specialized phenylpropanoids and flavonoids, along with their associated pathways, play key roles in CRM resistance, providing a foundation for breeding mite-resistant chestnut varieties.

plant biology↗

bioRxiv2025

An inherent T cell-activating mRNA delivery platform for in vivo CAR T generation

The clinical success of CAR T cell therapy has underscored the need for scalable, non-invasive strategies for in vivo T cell engineering. Although mRNA delivery offers a promising alternative, current lipid-based vectors lack intrinsic T cell tropism and often rely on antibody conjugation, complicating manufacturing and raising safety concerns. Here we present an inherent T cell-activating polymer-lipid nanoparticle (ERTLNPs) that enable ligand-free, efficient mRNA transfection and activation of T cells in vivo. ERTLNPs, composed of p-toluenesulfonyl arginine (RT)-functionalized polyethylenimine (denoted as PEI-RT) and helper lipids, preferentially accumulate in the spleen following systemic administration. Without exogenous stimulation, ERTLNPs intrinsically activate T cells, triggering robust mRNA expression and proliferation. Mechanistically, ERTLNPs engage the PI3K/AKT/mTOR signaling axis to reprogram T cell metabolism, promoting expansion while restraining exhaustion. Systemic delivery of mRNA encoding fibroblast activation protein chimeric antigen receptor (mFAP CAR) via ERTLNPs leads to in situ generation of functional CAR T cells, which efficiently eliminate pathological fibroblasts in models of cancer and fibrosis, with minimal off-target effects. This ligand-free, metabolically reprogramming mRNA delivery platform provides a clinically translatable approach for in vivo CAR T cell generation and broadens the landscape of non-viral immunotherapeutic engineering.

bioengineering↗