Biology subjects

Liu, P.

Publications and source records attributed to Liu, P..

At least 91 records · Page 5Linked to original sources

bioRxiv2025

Phosphatidylcholine Metabolism Controls Alveolar Progenitor Renewal and Pulmonary Fibrosis

Idiopathic pulmonary fibrosis (IPF) is a progressive, fatal lung disease marked by alveolar type 2 (AT2) stem cell dysfunction and excessive matrix deposition, with no effective treatments. Recent advances have recognized that AT2 cells act as stem cells, in addition to their role in the production of pulmonary surfactants in the distal alveolar space. We and others have reported a failure of AT2 regeneration and a loss of AT2 cells in IPF. We recently further reported that there is a defect in lipid metabolism in IPF AT2 cells and we discovered a selective loss of lysophosphatidylcholine acyltransferase 1 (LPCAT1) in AT2 cells from IPF, as well as in AT2 cells from bleomycin-injured mice. Pharmacological and genetic experiments confirm that LPCAT1 is required for AT2 cell renewal in 3D organoid assays. AT2 cell-specific Lpcat1 deletion resulted in reduced AT2 renewal, spontaneous lung fibrosis, and heightened susceptibility to bleomycin-induced fibrosis in mice in vivo. Expression-based high-content drug screening with an LPCAT1 knock-in cell line identified several drug families that upregulated LPCAT1 expression. We further confirmed that anti-malarial artesunate and PLA2 inhibitor ONO-RS-082 increased LPCAT1 mRNA expression, promoted AT2 renewal, and attenuated bleomycin-induced lung fibrosis in mice in vivo. Our findings establish LPCAT1 as a critical regulator of AT2 renewal and lipid metabolism in IPF, suggesting that reactivation of LPCAT1 could offer a novel therapeutic strategy for restoring alveolar progenitor function and mitigating lung fibrosis.

cell biology↗

bioRxiv2025

De Novo Design, Directed Evolution and Computational Study of Heme-Binding Helical Bundle Protein Catalysts for Biocatalytic Enantioselective Ge-H Insertion

De novo designed proteins offer a malleable platform for the development of stereoselective transformations not found in biochemistry. Here, we report the de novo design and directed evolution of helical bundle protein catalysts for enantioselective germylation through Ge-H insertion, a transformation not previously achieved by enzymatic catalysis. Comparative computational analysis revealed that, relative to Si-H insertion, the Ge-H insertion reaction proceeds through an earlier and more flexible transition state, introducing distinct challenges for stereocontrol. Using a fully de novo designed truncated four-helix bundle scaffold as the starting point, directed evolution afforded a quadruple mutant that catalyzes Ge-H insertion with high efficiency, enantioselectivity, and broad substrate scope. Molecular dynamics simulations indicated that beneficial mutations introduced from directed evolution enhanced active-site preorganization and modulated local back-bone flexibility, contributing to improved transition-state complementarity with fine-tuned binding pocket size and more stable cofactor positioning regulated by hydrogen bonding interactions. These findings showcase the excellent potential for de novo proteins to achieve stereoselective transformations previously unknown to biocatalysts and underscore the importance of active-site remodeling of de novo protein scaffolds via directed evolution in achieving selective catalysis involving flexible transition states. Table of Contents artwork O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=87 SRC="FIGDIR/small/679279v1_ufig1.gif" ALT="Figure 1"> View larger version (26K): org.highwire.dtl.DTLVardef@11d5beaorg.highwire.dtl.DTLVardef@7d6559org.highwire.dtl.DTLVardef@88d0fborg.highwire.dtl.DTLVardef@501945_HPS_FORMAT_FIGEXP M_FIG C_FIG

biochemistry↗

bioRxiv2025

Programming Microbe-Computer Interaction for Customized Life-Nonlife and Cross-Life Communication

This extension of human-computer interaction (HCI) paradigm to the non-intelligent microbe-computer interaction (MCI) is attractive for building generalized life-nonlife communication. Main challenges lie in how to construct biohybrid interface within individual MCI unit and thereafter program cross-life communication based on these MCI units. Here, we propose the design and construction of opto-interfaced MCI unit composed of engineered microbes and computer terminals, for enabling quantifiable transduction between microbial signals and digital signals. These units were able to serve as biohybrid sensors, actuators and transceivers, supporting biomonitoring, bio-actuation and microbe-centric bidirectional operations. Based on these functional units, we established programmable cross-life communication architectures capable of unicast, multicast, and logic-gate-driven communication with temporal dynamic control. The architectures were validated in biomedical scenarios including tumor hypoxia biomarking, epidermal heavy metal detoxification, and colitis intervention, demonstrating robust performance in closed-loop diagnostic-therapeutic applications. Additionally, we integrated MCI with an unmanned surface vessel for environmental monitoring, highlighting its cross-domain potential. Collectively, we provide a scalable strategy for building customized life-nonlife and cross-life communication, which is expected as an important complement to internet of things (IoT).

synthetic biology↗

bioRxiv2025

Colonization of vineyards by non-Saccharomyces yeast species without evolution of copper and sulfite resistance

Humans have generated ecological and environmental disturbances, such as vineyards, across the globe. Disturbed environments create widespread and repeated selective pressures that can drive colonization and local adaptation in microbial species. We investigated the distribution of fermentative yeast species in vineyards compared to nearby arboreal habitats and measured their resistance to two commonly used vineyard antimicrobials, copper and sulfite. We analyzed 4,101 strains, representing 70 species, collected from grapevine- and oak-associated substrates at 17 vineyard and 20 non-vineyard sites in the USA and Slovenia. Species frequency varied with geography and substrate, but the majority of species commonly present in vineyards were also found in non-vineyard arboreal environments, representing a potential source for vineyard colonization and exploitation of sugar from grapes. Species varied in both copper and sulfite resistance, but only Saccharomyces cerevisiae showed elevated resistance in vineyard compared to non-vineyard samples. Our results indicate that S. cerevisiae has uniquely taken advantage of vineyard environments through adaptations that appear either unnecessary or inaccessible to other yeast species present in vineyards.

evolutionary biology↗

bioRxiv2025

Dietary methionine restriction primes T cell metabolism for activation and tumor inhibition and enhances the efficacy of immune checkpoint blockade

The proliferation of many cancer cells is methionine dependent and dietary methionine restriction (MR) has shown anti-tumor effects in a wide variety of immunodeficiency preclinical models. Yet, whether MR exerts an anti-tumor effect in the presence of an immune-competent background remains inconclusive. Accumulating evidence has shown an essential role of methionine in immune cell differentiation and function. Thus, competition for methionine between tumor cells and immune cells in the tumor microenvironment may drive tumor growth and tumor response to therapy. Here, we aim to define the impact of MR on tumor growth and associated immunity. We first assessed the effect of MR in a series of immunocompetent mouse models of melanoma, colorectal cancer, breast cancer, and lung. MR led to a broad tumor inhibition effect across these models and such tumor inhibition was not sex-or genetic background-dependent but appears to be fully or partially immune-dependent. Through flow cytometry analysis, we found a consistent increase in intratumoral activated CD8+ T cells across different tumor models and depletion of CD8+ T cells partially or completely reversed MR-induced tumor inhibition in a model dependent manner. Interestingly in young healthy non-tumor-bearing mice, MR increased spleen CD3+ and CD8+ T cell populations. Metabolomics and RNAseq analysis of spleen-derived CD8+ T cells revealed significant increase in purine metabolism and amino acid metabolism and that are in line with the metabolic feature of activated T cells. Furthermore, MR improved the efficacy of anti-PD1 immune checkpoint blockade. Together, MR primes T cell metabolism for its anti-tumor effect and improves the efficacy of anti-PD1 checkpoint blockade.

cell biology↗

bioRxiv2025

Convergent utilization of APN receptor by hedgehog merbecoviruses

Dipeptidyl peptidase-4 (DPP4) and angiotensin-converting enzyme 2 (ACE2) are well-established receptors for merbecoviruses, yet the receptor usage of merbecoviruses of European and Asian hedgehogs (EriCoVs) remains unknown. Here, by testing hedgehog orthologs of known coronavirus receptors, we identify hedgehog aminopeptidase N (APN) as a functional receptor for EriCoVs. Analysis of 94 APN orthologs indicates that EriCoVs have a limited host range, primarily utilizing hedgehog APN and, to a lesser extent, APN from certain felids, shaped by specific determinants at the virus-receptor interface. Cryo-EM reveals an APN-binding mode distinct from those used by alpha- and deltacoronaviruses. Functional assays indicate that hedgehog transmembrane serine protease 2 (TMPRSS2) enhances spike activation and promotes pseudovirus entry. Neutralizing antibodies targeting RBD and APN were developed and could effectively block EriCoV pseudovirus entry and propagation. These findings reveal an unexpected convergent evolution of APN utilization among merbecovirus, establishing a foundation for risk assessment and countermeasure development. Graphic abstractConvergent utilization of hedgehog APN by EriCoVs O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=115 SRC="FIGDIR/small/676546v1_ufig1.gif" ALT="Figure 1"> View larger version (48K): org.highwire.dtl.DTLVardef@1ea334aorg.highwire.dtl.DTLVardef@cd53bforg.highwire.dtl.DTLVardef@352102org.highwire.dtl.DTLVardef@a1eed_HPS_FORMAT_FIGEXP M_FIG C_FIG

molecular biology↗

bioRxiv2025

Selective and Controllable Trapping of Single Proteins in Nanopores using Reversible Covalent Bonds

Analysis of individual proteins using nanopores makes it possible to determine their size and shape in a label-free approach, within minutes, and from L sample volumes. Short residence times of proteins in the nanopore, high electrical current noise, and bandwidth limitations of the recording electronics during resistive pulse recordings, however, limit the accuracy of size and shape analysis of individual proteins. The work presented here introduces a polymer surface coating of solid-state nanopores to minimize non-specific interactions of proteins with the nanopore wall while functionalizing it covalently with phenylboronic acid (PBA) groups. These PBA groups make it possible to trap selectively glycated proteins by taking advantage of the formation of reversible covalent bonds between PBA and vicinal diol groups of glycated amino acid residues on proteins. Dwell time analysis revealed two populations of resistive pulses: short pulses with dwell times td below 0.4 ms from free translocation of proteins and resistive pulses that we term "long events" that last from 0.4 ms to 2 s and result from intended transient covalent bonds between glycated proteins and PBA groups in the nanopore lumen. The choice of applied potential differences during nanopore recordings or the pH value of the recording buffer makes it possible to control and extend the most probable trapping time of proteins in the nanopore within one to two orders of magnitude. This approach provides the highest accuracy for the determination of protein volume and shape achieved to date with solid-state nanopores and reveals that a trapping time of 1 to 20 ms is ideal to achieve reliable volume and shape analysis while retaining high throughput of the analysis. This approach, hence, extends the residence time of natively glycated proteins or of proteins that are intentionally glycated by straightforward incubation in a glucose solution, thereby providing selectivity and improving the accuracy of nanopore-based characterization of single proteins.

biophysics↗

bioRxiv2025

Rewiring of chromatin regulation underlies the evolution of brown algal multicellularity

Chromatin structure plays a central role in regulating transcription, genome stability, and epigenetic inheritance in eukaryotes. Much of our understanding of chromatin architecture and histone post-translational modifications (hPTMs) comes from a narrow set of animal and plant models, but emerging data from non-model lineages are challenging canonical views of how chromatin functions across the tree of life. Brown algae are complex multicellular eukaryotes that provide a unique perspective on chromatin evolution given their independent origin of complex multicellularity. Here, we compile the chromatin toolkit of brown algae and show that canonical silencing systems involving DNA cytosine methylation and PRC2-mediated H3K27 methylation were lost early in their evolution. By generating hPTM profiles from diverse brown algal clades, we resolve the nature and regulatory roles of chromatin states in this lineage and show how H3K79 methylation emerged and diversified as a repressive system. We further uncover sex-specific reconfigurations in species with varying degrees of sexual dimorphism and reconstruct the ancestral regulatory landscape that likely preceded the emergence of brown algae. Together, our findings illuminate the dynamic evolution of chromatin regulation in a distinct multicellular lineage and challenge assumptions about the universality of chromatin-based mechanisms across eukaryotes.

evolutionary biology↗

bioRxiv2025

Macropinocytosis-Mediated Recyclable LYTACs (McR-TACs) for Membrane and Extracellular Protein Degradation

The degradation of cell membrane and extracellular proteins with lysosome-targeting chimeras (LYTACs) is hampered by non-recyclable and lysosome-shuttling receptor-dependent mechanisms, limiting the wide application of this emerging technology. Here, we developed macropinocytosis-mediated recyclable LYTACs (McR-TACs) that could durably degrade cell membrane and extracellular proteins in a receptor-independent manner. This McR-TAC platform demonstrated the feasibility of harnessing endogenous transport pathways for developing next-generation recyclable protein degraders with broad applicability.

biochemistry↗

bioRxiv2025

STARNet enables spatially resolved inference of gene regulatory networks from spatial multi-omics data

Biological tissues are composed of distinct microenvironments that spatially orchestrate gene expression and cell identity. However, the regulatory principles governing domain-specific cellular functions remain poorly understood due to the lack of effective methods for mapping gene regulatory networks (GRNs) in situ. To address this gap, we introduce STARNet, a representation learning approach that leverages heterogeneous hypergraph modeling of spatial transcriptomic and epigenomic data to resolve tissue-domain-specific regulatory interactions. By integrating graph neural networks with contrastive learning in a self-supervised framework, STARNet learns unified embeddings that preserve both multi-modal molecular features and anatomical spatial context, enabling accurate and domain-resolved GRN reconstruction within complex tissues. Benchmarking on both simulated and real datasets demonstrates that STARNet achieves state-of-the-art performance. We further demonstrate its broad applicability across diverse biological contexts, including neural development, genetic disease risk, and drug-induced developmental toxicity. In the mouse brain, it delineates region-specific regulatory networks and reconstructs spatiotemporal programs underlying neural stem cell differentiation. In human genetics, it provides a mechanistic link between genotypes and phenotypes by showing how genome-wide association study (GWAS) variants for complex diseases perturb hippocampus-specific GRNs. In developmental toxicology, STARNet reveals that drug-induced disruptions of GRNs in defined embryonic regions underlie tissue-specific vulnerability. Collectively, STARNet offers a powerful and versatile framework for resolving the spatial regulatory logic of complex tissues, providing multi-angle insights into tissue patterning, development, and disease mechanisms.

bioinformatics↗

bioRxiv2025

Lipid Droplets of Human Carotid Atherosclerotic Plaques: Cholesteryl Arachidonate Core and Cytoskeletal Fibrotic Cover

Lipid droplets (LDs) within atherosclerotic plaques are central to cardiovascular disease, yet their molecular composition remains largely unknown. More than 20 years experiences of developing LD purification method allowed us recently to successfully purify LDs from atherosclerotic plaques of patients undergoing carotid artery surgery, including three types, soft (lipid), mixed, and hard (calcified). In addition, the subpopulations of large LDs and small LDs were purified and studied, respectively. We report the first comprehensive proteo-lipidomic analysis of these LDs. Unexpectedly, about 19.3% of the plaque LD proteins belonged to cytoskeleton, while the content of lipid metabolism enzymes on plaque LDs was low. Our results show that the compositions of lipids and proteins were similar between two population LDs. Compared to soft (lipid) plaque LDs, LDs from mixed and hard (calcified) plaques had fewer metabolic enzymes and more cytoskeletal proteins. Lipidomics of the LD core uncovered an unexpected enrichment of polyunsaturated fatty acid cholesteryl esters (PUFA-CEs), which constituted 88.3% of all cholesterol esters. Notably, esters of the pro-inflammatory precursor arachidonic acid (AA) were among the most abundant species (16.5%). In addition, plaque LD were uniquely enriched in phospholipids including phosphatidic acid (PA), phosphatidylethanolamine plasmagens (PE-P), sphingomyelin (SM), and phosphatidylserine (PS). Our findings redefine the plaque LD as a dual-function "bad" organelle: a metabolically inert container for cholesterol ester, coated with cytoskeletons, and a massive reservoir for inflammatory precursors.

cell biology↗

bioRxiv2025

Retinal pigment epithelium-derived PD-L1 reprograms microglial cells and protects against retinal degeneration in mouse models of experimental AMD and genetic retinitis pigmentosa

Retinal degeneration is a leading cause of blindness and a significant health burden during aging worldwide. Because of its complex pathogenic mechanisms, however, effective interventions applicable to multiple retinal degenerations remain limited. Excessive microglia-mediated inflammation is a common pathological feature of different retinal degenerative diseases, and inhibiting these overactive inflammatory responses by reprogramming microglia may provide a therapeutic strategy with broad-spectrum potential. Here, we uncover a novel retinal pigment epithelium (RPE) homeostatic signal, PD-L1, that drives the reprogramming of activated microglial toward a retinal-protective phenotype. Using multiple analyses, including gene expression profiling bioinformatics, adeno-associated virus-mediated RPE-specific gene overexpression, pharmacological, and genetic manipulations, we show that RPE cells upregulate the expression of PD-L1 to modulate microglia for protection against retinal degeneration in an experimental dry AMD mouse model. Bioinformatics enrichment analysis of the high-expression gene signatures of subretinal microglia in AMD patients, combined with pharmacological blocking of NF-{kappa}B activity or using Nlrp3 knockout mice, show that PD-L1/PD-1 signaling suppresses NF-{kappa}B activation in microglia, thereby inhibiting NLRP3 inflammasome priming and attenuating microglia-mediated excessive inflammation. Furthermore, the overexpression of PD-L1 by the AAV8 virus exerts similar microglial reprogramming and retinal protective effects to alleviate retinal degeneration in a preclinical model of retinitis pigmentosa. These findings reveal that RPE-derived PD-L1 suppresses inflammatory reprogramming of subretinal microglia and mitigates retinal degeneration, highlighting a promising proof of concept for PD-L1 as a potential therapeutic target for the broad treatment of retinal degenerative diseases.

genetics↗

bioRxiv2025

MNP33 is a novel mitochondrial protein that reprograms cellular bioenergetics and lipid metabolism

Genomic regions previously annotated as non-coding are now recognized sources of functional microproteins, termed the "dark proteome", though their identification and characterization remain challenging. Addressing this, we developed a novel proteomic strategy combining lipid droplet (LD) enrichment with mass spectrometry against the database of sORF-derived proteins. This approach yielded the first reported LD-resident microprotein, LDANP1, and subsequently revealed other candidates from the same screen, such as MNP33, were localized to mitochondria. Here, we provide the first in-depth functional characterization of MNP33, a novel 28-amino acid protein encoded by an sORF within the Cdhr4 gene locus. We confirm MNP33 localizes to the inner mitochondrial membrane and demonstrate its interaction with the adenine nucleotide translocase 2 (ANT2). Functionally, MNP33 expression remodels mitochondrial bioenergetics, increasing basal respiration and proton leak while paradoxically elevating membrane potential, partly through modulating ANT2 activity linked to enhanced glycolysis. Furthermore, MNP33 shifts cellular lipid metabolism, favoring cholesteryl ester storage over triacylglycerols via stabilization of ACAT1, and is associated with closer ER-mitochondria contacts. MNP33 also induces ROS production and autophagy while inhibiting cell proliferation without increasing apoptosis. This study establishes MNP33 as a novel mitochondrial regulator emerging from the dark proteome, providing a mechanistic link between non-coding genetic elements, ANT2 function, mitochondrial bioenergetics, cellular metabolism, and stress responses. Our work highlights functional importances of the dark proteins and validates organelle-focused proteomics for discovering key cellular regulators.

cell biology↗

bioRxiv2025

Polycomb Repressive-Deubiquitinase Complex Safeguards Oocyte Epigenome and Female Fertility by Restraining Polycomb Activity

Mouse oocytes exhibit a unique chromatin landscape characterized by broad H3K27ac and H3K27me3 domains, demarcating euchromatin and facultative heterochromatin, respectively. However, the mechanisms underlying this non-canonical landscape remain elusive. Here we report BAP1, a core component of the Polycomb Repressive-Deubiquitinase (PR-DUB) complex, as a key negative regulator of Polycomb activity during oogenesis. BAP1 restricts pervasive H2AK119ub1 accumulation in oocytes and protects oocyte-specific broad H3K27ac, particularly within gene-poor regions, from ectopic H3K27me3 deposition. While PR-DUB has been linked to gene repression, in oocytes BAP1 primarily promotes transcription and contributes minimally to Polycomb-mediated silencing. BAP1-dependent transcriptional activation is essential for oocyte developmental competence and female fertility. BAP1 loss disrupts the maternal-to-zygotic transition and impairs embryonic enhancer activation, ultimately compromising preimplantation development. Notably, while H3K27ac patterns are reset after fertilization, the aberrant H3K27me3 landscape established in BAP1-deficient oocytes persists in early embryos. Together, these findings reveal a critical role for PR-DUB in safeguarding the oocyte epigenome by protecting euchromatin from ectopic Polycomb activity, rather than enforcing transcriptional repression.

developmental biology↗

bioRxiv2025

Lipid Droplet Proteome Reveals that Associated ATGL and Anchored Mitochondria Lead to Higher Skeletal Muscle Insulin Sensitivity in Endurance Athletes than Type 2 Diabetes Mellitus

As the key organelle governed lipid homeostasis, lipid droplets (LDs) in skeletal muscle plays a critical role in regulating systemic insulin sensitivity. In type 2 diabetes mellitus (T2DM) patients intramyocellular lipid (IMCL) content negatively associates with insulin sensitivity, while endurance athletes exhibit IMCL levels were positively correlated to insulin sensitivity, which is known as the athletes paradox. To solve this paradox, LDs were isolated from skeletal muscle samples of seven athletes and nine T2DM patients, and the first proteome dataset of human skeletal muscle LD was established. Comparative proteomic analysis revealed 733 upregulated proteins in athlete LDs, primarily enriched in mitochondria, and 755 downregulated proteins, mainly associated with cytoskeletal components, relative to T2DM patients. Integrated analysis with tissue proteomics further indicated enhanced energy metabolism activity of LD-anchored mitochondria (LDAM) in athletes. Notably, adipose triglyceride lipase (ATGL) was specifically upregulated by 2.97-fold on LDs in athletes, while its overall tissue expression remained unchanged. Moreover, deletion of ATGL in myoblasts significantly reduced insulin-stimulated AKT phosphorylation. Additionally, skeletal muscle LDs from db/db and ob/ob mice exhibited reduced levels of mitochondrial proteins and ATGL. Together, our findings reveal a LD-based solution for the athletes paradox, and identify ATGL as a key regulator of skeletal muscle insulin sensitivity. The results also suggest that the "bad LDs" in the skeletal muscle of T2DM patients are characterized by higher levels of cytoskeletal proteins, reduced LDAM, and lower ATGL compared to the "good LDs" observed in athletes.

cell biology↗

bioRxiv2025

Pluripotent stem cell-derived siTNK cells attack tumors via a synthetic CD8-CD28 TCR complex

Tumor-associated antigen-specific T cell receptor (TCR)-engineered T cells offer a promising strategy for cancer therapy. Natural killer (NK) cells exhibit broad anti-tumor activity with low side effects but lack the capacity to recognize intracellular antigens. Here, we found that the human pluripotent stem cell (hPSC)-derived iNK cells, unlike tissue-isolated NK cells, expressed all four CD3 subunits at the transcriptome level. We introduced a synthetic gene-expressing complex (SCOTR), encoding a tumor antigen- specific TCR, CD8 coreceptor, and CD28 costimulatory molecule, into hPSCs to generate SCOTR-hPSCs. The SCOTR-hPSCs gave rise to abundant synthetic TCR complex-expressing iNK (siTNK) cells via an organoid induction method. These siTNK cells showed HLA-dependent, antigen-specific cytotoxicity against tumor cells and significantly suppressed tumor growth in tumor xenograft animal models, while also preserving universal non-specific tumor-killing activity. Collectively, siTNK cells, as hPSC-derived hybrid cells with dual features of adaptive T and inherent NK cells, offer an artificial cell source for human immunotherapies.

immunology↗

bioRxiv2025

Copper-Containing ROS-Scavenging Nanozyme Paradoxically Drives Alpha-Synucleinopathy by Amplifying Nitrosative Stress

Reactive oxygen and nitrogen species (RONS) are implicated in neurodegeneration, but their specific pathogenic roles remain unclear. Here, we developed a pair of iridium-based nanozymes with opposing functionalities to dissect these pathways. We show that a copper-tuned iridium nanozyme (Ir{square}Cu), despite being a superior ROS scavenger, paradoxically and dramatically exacerbated -synuclein (Syn) pathology in vivo. This pathology was causally linked to its ability to amplify RNS, as pharmacological inhibition of nitric oxide synthase (NOS) with L-NAME completely abrogated the pathology and reversed a human Parkinsons disease (PD)-like transcriptomic signature. In contrast, a copper-free, broad-spectrum RONS-scavenging iridium (Ir) nanozyme demonstrated substantial therapeutic efficacy across diverse brain-first, body-first, and Alzheimers disease with Lewy body co-pathology models. Our findings uncover the importance of the RNS pathway in driving -synucleinopathies and establish a critical design principle for nanomedicine, mandating caution in the use of redox-active copper for neuroprotective applications.

pathology↗

bioRxiv2025

MassNet: billion-scale AI-friendly mass spectral corpus enables robust de novo peptide sequencing

Breakthroughs in artificial intelligence (AI) for natural language processing and computer vision have been largely driven by high-quality, large-scale datasets such as OpenWebText and ImageNet. Inspired by this, we present MassNet, a foundational resource for proteomics designed to accelerate deep learning applications. MassNet is the largest known corpus of data-dependent acquisition (DDA) mass spectrometry (MS) data, derived from ~30 TB of raw files and comprising 1.54 billion MS/MS spectra, resulting in 558 million peptide-spectrum matches (PSMs) across 35 species, including animals, plants, and microbes. Within the human subset, MassNet includes more than 1.7 million precursors and 19,966 proteins, covering 98% of annotated human proteins. To enable efficient AI training, we developed the Mass Spectrometry Data Tensor (MSDT), a structured format based on Parquet that enables standardized, high-performance batch access and seamless integration with GPU and TPU platforms for distributed training. We further extended MassNet to support de novo peptide sequencing, which infers peptide sequences directly from MS/MS spectra without reference databases, and is critical for discovering novel proteins, characterizing non-model organisms, and identifying post-translational modifications (PTMs). We introduce XuanjiNovo, a non-autoregressive Transformer model that leverages a curriculum learning strategy to enhance training stability. By dynamically adjusting learning difficulty based on model performance, XuanjiNovo achieves smooth convergence on complex, multi-distributional data without manual hyperparameter tuning. Trained on 100 million PSMs from the MassNet, it consistently outperforms state-of-the-art methods across diverse benchmarking tasks. Peptide recall exceeds 0.8 on the Bacteroides thetaiotaomicron and Zea mays datasets. On human data acquired using the Orbitrap Astral platform, XuanjiNovo achieves achieves 38.8% to 144.3% improvement over existing models. MassNet represents the first large-scale, standardized foundational dataset in proteomics, marking a critical milestone in the integration of artificial intelligence into proteomics research.

bioinformatics↗