Biology subjects

Chen, C.

Publications and source records attributed to Chen, C..

At least 145 records · Page 8Linked to original sources

bioRxiv2025

Thalamic Interictal Epileptic and Non-Epileptic Events during NREM Sleep in Patients with Focal Epilepsy: a Stereo-EEG Study

BackgroundThalamic recordings are increasingly incorporated into stereo-electroencephalography (SEEG) evaluations of drug-resistant focal epilepsy to guide neuromodulation targeting. Human thalamic electrophysiology, however, is poorly defined, limiting the distinction between pathological and physiological activity. Here, we characterized interictal epileptic and non-epileptic events during non-rapid eye movement (NREM) sleep across multiple thalamic nuclei and examined their relationship to seizure outcomes. MethodsWe analyzed NREM sleep SEEG recordings from 64 patients with drug-resistant focal epilepsy. Electrodes sampled four thalamic nuclei: centromedian (CM), pulvinar (Pu), ventral lateral (VL), and ventral posterolateral (VPL). Patients were classified into three outcome groups: favorable, unfavorable, and surgically non-remediable. Rates of thalamic spikes, high-frequency oscillations (HFOs), spike-fast activity, and sleep spindles were analyzed and compared across nuclei and outcomes. FindingsRecordings of the thalamus revealed both pathological and physiological interictal events. Interictal epileptic events were infrequent. Only [~]0.2% of seizure-onset zone spikes propagated to the thalamus. Thalamic spike-fast activity was indicative of unfavorable surgical outcomes (CM: p = 0.047, d = 0.46) or surgically non-remediable epilepsy (VL: p = 0.002, d = 0.84). In contrast, thalamic sleep spindles were ubiquitous but reduced in surgically non-remediable patients (CM: p = 0.031, d = -0.58; VL: p = 0.005, d = -0.79). Finally, unique thalamic SEEG patterns were identified, including spikes concomitant with spindles, isolated spikes, and physiological fast ripples. InterpretationThis study provides a comprehensive characterization of thalamic interictal events during NREM sleep, enriching our understanding of thalamic pathophysiology and highlighting the value of thalamic recordings in presurgical evaluation.

neuroscience↗

bioRxiv2025

AI recognizes convergent somatic hypermutation signatures to allow the discovery of variant-resilient broadly neutralizing antibodies

Elucidating the mechanisms by which broadly neutralizing antibodies (bnAbs) develop to confer durable immunity in humans is pivotal for the rational design of next-generation vaccines and therapeutics. VH3-53/3-66-encoded public antibodies are widely elicited in the population after the COVID-19 pandemic. Here, we isolated 15 VH3-53/3-66-encoded bnAbs from elite neutralizers who experienced sequential SARS-CoV-2 Omicron breakthrough infections. These bnAbs exhibit exceptional potency and breadth against circulating variants and, as members of a long-lived public antibody lineage, likely contribute to durable humoral protection. Genetic analysis of VH3-53/3-66-encoded bnAbs reveals a convergent somatic hypermutation pattern at seven positions in and around the CDR loops that emerges with repeated viral exposures and distinguishes them from non-bnAbs. Grafting these combined mutations onto clinically escaped antibodies broadens their breadth and restores neutralization activity. Structural analysis shows that these convergent mutations cooperatively remodel CDRs to bind and tolerate virus receptor-binding domain mutations. Using an AI-based antibody language model trained on data linking somatic hypermutations to binding affinity and neutralization, the model is able to identify mutational patterns predictive of breadth, enabling the discovery of a rare bnAb with protective activity against the latest variants from early-pandemic antibody repertoires. This demonstrates the feasibility of establishing an AI-empowered pipeline to identify mutation-tolerant bnAbs from early-pandemic antibody repertoires to fight fast-evolving newly introduced viruses.

immunology↗

bioRxiv2025

Selective Inactivation of Astrocytic Monoacylglycerol Lipase for Alzheimer Disease Therapy

Alzheimers disease (AD) is the leading cause of dementia in the elderly, yet effective therapies remain lacking. Here, we identify monoacylglycerol lipase (MAGL), the principal enzyme that degrades the endocannabinoid 2-arachidonoylglycerol (2-AG) in the brain, specifically in astrocytes as a promising therapeutic target for AD. APP transgenic (TG) mice with astrocyte-specific MAGL deletion (TG-aKO), but not neuron-specific deletion (TG-nKO), exhibit markedly reduced A{beta} pathology, neurodegeneration, and neuroinflammation, along with preserved synaptic structure and function and improved cognition. These benefits are recapitulated in TG mice by AAV-mediated silencing of astrocytic MAGL, whereas MAGL overexpression exacerbates neuropathology and accelerates synaptic and cognitive decline. Transcriptomic analyses reveal that dysregulated synapse-, inflammation-, and AD-related gene expression profiles are reversed in TG-aKO mice. MAGL expression is elevated in astrocytes from both AD patients and TG mice. Remarkably, a single intracerebroventricular (ICV) injection of AAV-MAGL-shRNA administered at either presymptomatic or postsymptomatic stages prevents or reverses neuropathology and synaptic and cognitive impairments in TG mice, providing preclinical evidence that astrocytic MAGL silencing represents an effective therapeutic strategy for AD.

neuroscience↗

bioRxiv2025

Early blood-spinal cord barrier stabilization preserves peri-lesional tissue and improves recovery after SCI

Vascular disruption is an early and critical event in spinal cord injury (SCI), but its contribution remains poorly understood. Using novel in vivo two-photon dual-dye imaging, we found simultaneous blood-spinal cord barrier (BSCB) leakage and venous dilation in both the injury epicenter and adjacent transitional segment after cervical SCI in rats. Notably, vascular permeability in the transitional zone preceded axonal and neuronal loss, revealing a therapeutic window. Systemic delivery of ferulic acid-glycol chitosan (FA-GC) nanoparticles, a membrane-sealant, rapidly stabilized the compromised vasculature, reduced neuronal loss in the transitional region, and improved forelimb muscle strength. These findings identified acute vascular leakage beyond the injury epicenter as a driver of secondary pathology and highlight early vascular stabilization as promising therapeutic strategy.

pathology↗

bioRxiv2025

IgG-Bridging-Seeded Synergistic Aggregation of SARS-CoV-2 Spikes Underlies Potent Neutralization by A Low-Affinity Antibody

Mechanistic studies of viral neutralization typically prioritize high-affinity antibodies, relegating low-affinity binders to the sidelines. We report P5-1C8, a Class 1 SARS-CoV-2 antibody that exemplifies this underexplored "low-affinity yet high-potency" phenotype, retaining strong neutralization of Omicron JN.1 despite markedly weakened trimer binding (KD = 225 nM; IC50 = 0.06 nM). Structural and biophysical analyses reveal that P5-1C8 engages WT and BA.1 spikes through canonical intra-spike bivalency, but with JN.1 it induces aggregation. Using virion-like nanoparticles displaying multiple spikes, we show that IgG remains bound with no detectable dissociation and triggers pronounced aggregation. Coarse-grained molecular dynamics delineate the stepwise pathway in which weak IgG-spike contacts seed aggregation via transient inter-spike bridging. Together, these findings establish the first mechanistic framework demonstrating how weak-binding antibodies can nonetheless achieve potent neutralization through higher-order aggregation, thereby expanding the conceptual landscape of antibody function and opening new directions for antibody evaluation and design. Graphical AbstractLow-affinity antibodies are frequently disregarded in discovery pipelines. This work reports P5-1C8, a Class 1 SARS-CoV-2 antibody with weak trimer binding (KD-to-IC50 > 3,700-fold) yet potent neutralization of Omicron JN.1. Structural, biophysical, functional and coarse-grained simulations collectively demonstrate that transient inter-spike IgG bridging seeds higher-order aggregation, which in turn drives neutralization and provides a mechanistic framework. O_FIG O_LINKSMALLFIG WIDTH=199 HEIGHT=200 SRC="FIGDIR/small/691454v1_ufig1.gif" ALT="Figure 1000"> View larger version (79K): org.highwire.dtl.DTLVardef@1d4f132org.highwire.dtl.DTLVardef@1274400org.highwire.dtl.DTLVardef@e186a6org.highwire.dtl.DTLVardef@4f1870_HPS_FORMAT_FIGEXP M_FIG C_FIG

cell biology↗

bioRxiv2025

Antibacterial Activity of the Natural Product Elaiophylin against Streptococcus suis and its Biofilm Eradication Effect

The multidrug resistance (MDR) problem in Streptococcus suis(S. suis) is becoming increasingly severe, necessitating the development of novel antibacterial agents and strategies. In this study, seven Elaiophylin derivatives were isolated from Streptomyces sp. WS-30248, and the antibacterial activity and mechanism of action of the principal compound, Elaiophylin, against S. suis were systematically evaluated for the first time. Through comprehensive approaches including in vitro efficacy assays, biofilm inhibition and eradication tests, bacterial membrane integrity analysis, reactive oxygen species (ROS) level detection, and a mouse infection model, Elaiophylin was found to exhibit significant antibacterial activity against multiple clinically MDR S. suis strains. Its minimum inhibitory concentration (MIC) was as low as 0.5 g/mL, and complete bactericidal activity was achieved within 24 hours at 4xMIC. The compound effectively inhibited and eradicated bacterial biofilms, directly killing embedded cells within the biofilm matrix. Mechanistic studies revealed that Elaiophylin functions through multiple synergistic pathways, including disruption of bacterial membrane integrity and induction of massive ROS accumulation, thereby interfering with the proton motive force (PMF), depleting intracellular ATP, and blocking energy metabolism. In the mouse infection model, the Elaiophylin-treated group showed a significantly increased survival rate of 60% and effectively reduced bacterial loads in tissues. This study demonstrates that Elaiophylin is a highly promising natural candidate drug with multi-target synergistic effects, offering a new strategy to combat MDR S. suis and biofilm-associated infections.

microbiology↗

bioRxiv2025

Engineering an Anaerobic Microenvironment to Empower Hydrogenase-Catalyzed Hydrogen Therapy for Diabetic Wound Healing

The inherent oxygen sensitivity of hydrogenases has long restricted their applications to anaerobic systems. Here, we present a hybrid peptide-nanocluster hydrogel that establishes a self-sustaining anaerobic microenvironment to enable hydrogenase-catalyzed hydrogen therapy under aerobic conditions. This system integrates a custom-designed Fmoc-KYF peptide hydrogel with integrated silver nanoclusters (AgNCs), which host an engineered [NiFe]-hydrogenase (3S1C-Hyd-2). Mechanistic studies combining molecular dynamics simulations and spectroscopic analysis revealed that hydrophobic pockets within the peptide network trap oxygen molecules, while photoexcited AgNCs rapidly deplete invasive O2. Our design ensures stable hydrogen evolution, effectively overcoming the oxygen sensitivity that has long restricted hydrogenase applications in biomedical contexts. In vitro, the in-situ-generated hydrogen effectively scavenges reactive oxygen species, suppresses pro-inflammatory cytokine production. In a diabetic mouse model, this light-activated system markedly accelerates wound closure (87% by day 11), reduces oxidative stress, promotes macrophage polarization toward a reparative phenotype, enhances angiogenesis, and upregulates tissue repair pathways, as confirmed by transcriptomic profiling. Our work establishes a new bioengineering paradigm for constructing anaerobic microenvironments through synergistic material-enzyme-light interactions, offering a versatile platform for hydrogenase-enabled therapeutic interventions in oxidative stress-related diseases.

biochemistry↗

bioRxiv2025

UHRF1 Overexpression Generates Distinct Senescent States with Different Tp53 Dependencies

SummarySenescence is a pleiotropic phenotype that alternatively suppresses or promotes cancer. The tumor suppressive roles are linked to clearance of damaged cells by the immune system, while cells with tumor promoting functions resist apoptosis, evade immune clearance and either persist and support the tumor microenvironment, or escape and proliferate. What generates these diverse populations is unclear. We investigated this in preneoplastic zebrafish livers where the epigenetic regulator, UHRF1, is overexpressed in hepatocytes. Double strand breaks, DNA methylation repatterning, retrotransposon expression, cell cycle withdrawal and activation of Atm and Tp53 dependent senescence were early responses to UHRF1 overexpression. This evolved to generate diverse populations of senescent cells, some which expressed immune and senescence signatures plus anti-apoptotic markers, and others co-expressed proliferative genes. The fate of these populations was dictated by UHRF1 levels and Tp53, as Tp53 loss enabled proliferation of cells with reduced UHRF1 expression but not in cells expressing high UHRF1. The senolytic Navitoclax targeted only a subset of senescent cells. Thus, the diversity of senescent cells driven by epigenetic changes can generate divergent outcomes. Short summaryUHRF1 overexpression in zebrafish hepatocytes induces DNA damage leading to Atm-tp53-dependent senescence with UHRF1 levels dictating whether hepatocytes remain senescent, escape, or undergo senolytic elimination.

cancer biology↗

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

MiRformer: a dual-transformer-encoder framework for predicting microRNA-mRNA interactions from paired sequences

MotivationMicroRNAs (miRNAs) regulate gene expression by binding to target messenger RNAs (mRNAs), inducing translational repression or mRNA degradation. Accurate prediction of miRNA-mRNA interactions and precise localization of binding and cleavage sites are critical for understanding post-transcriptional regulation and for enabling RNA therapeutics. Existing computational methods often rely on handcrafted or indirect features, scale poorly to kilobase-long mRNA sequences, or provide limited interpretability. ResultsWe present MiRformer, a transformer-based framework that jointly predicts miRNA-mRNA interactions and localizes miRNA binding and cleavage sites directly from raw sequence pairs. MiRformer employs a dual-transformer encoder architecture for miRNA and mRNA sequences and incorporates a sliding-window attention mechanism to efficiently model kilobase-long mRNA contexts while preserving nucleotide-level resolution. Across multiple benchmrks, MiRformer achieves state-of-the-art performance on interaction prediction, binding-site localization, and cleavage-site identification from experimental Human Degradome-seq data. Beyond accuracy, MiRformer provides strong interpretability: attention patterns consistently highlight miRNA seed regions within 500-nt mRNA windows, revealing clear and biologically meaningful interaction signals. When applied to jointly infer binding and cleavage sites across 13k miRNA-mRNA pairs, predicted sites frequently co-localize, supporting a miRNA-mediated degradation mechanism. AvailabilityPython code and datasets are publicly available at https://github.com/li-lab-mcgill/miRformer. ContactPlease contact corresponding author at yueli@cs.mcgill.ca

bioinformatics↗

bioRxiv2025

Scalable prediction of symmetric protein complex structures

All life relies on proteins to function, yet accurately modeling protein structures that exceed {approx} 10, 000 amino acids or have higher-order geometries remains difficult. Existing solutions are limited to specific scenarios, require considerable computational resources, or are otherwise unscalable. Consequently, many large, disease-relevant protein complexes in the human proteome, as well as nearly all viruses and numerous other classes, are impractical to model with high fidelity for drug development. To modulate these protein complexes and viruses, structural information is eminently valuable, and often essential. In the last two years, machine learning based-tools that can generate binders to a given target structure with high hit rates have emerged. Combined with high-throughput screening, these technologies can far outpace traditional drug discovery. However, they cannot function well without accurate models of their target structures. Thus, to unlock the full power of AI-driven drug discovery, a scalable method must be developed to predict large protein complex structures. To overcome this bottleneck, we introduce Plica-1, a physics-based method to rapidly and accurately predict the structure of arbitrarily large, symmetric protein complexes. Validated across 4 major symmetry classes (icosahedral, tetrahedral, octahedral, and cyclic), the method consistently achieves near-experimental levels of accuracy, i.e., RMSD < 5[A]. In test cases, the method runs in < 5 minutes on consumer hardware, 103-105 times faster than the closest comparable software. The largest structure currently built, at {approx}40,000 amino acids, is > 8 times the limit of existing machine learning methods. The results demonstrate that protein complexes can be modeled at significantly improved speeds and scales, making Plica-1 a promising tool for protein engineering and drug development.

bioengineering↗

bioRxiv2025

APOE4-Aβ synergy drives brain network dysfunction and neuronal lysosomal-ER proteostasis dysregulation in preclinical Alzheimer's disease

Amyloid-{beta} (A{beta}) and APOE4 represent two of the strongest pathological and genetic risk factors for Alzheimers disease (AD), but how these co-pathogens interact during preclinical stages remains undefined. We addressed this question by developing a humanized knock-in model expressing physiological, endogenously regulated human A{beta} and APOE4. Aged AppNLF:APOE4 mice displayed incipient amyloidosis with subtle memory-related changes, consistent with preclinical AD. We found largely distinct, non-overlapping APOE4- and A{beta}-driven functional synaptic, sleep, and behavioral alterations. However, at the transcriptomic level, APOE4xA{beta} had a pronounced detrimental interaction in neuronal populations, whereas glial populations were primarily affected by either genotype. We found APOE4xA{beta} molecular interactions in neuronal populations, including excitatory and inhibitory cells, converged on a core lysosomal-ER proteostasis axis. We propose that APOE4xA{beta} interaction produces an early neuronal pathogenic signature, involving the lysosomal-ER proteostasis axis, preceding functional decline and driving disease progression. APOE4xA{beta}-KI models provide a physiologically relevant platform to study early pathogenesis. HighlightsO_LIEarly synergistic APOE4xA{beta} interaction emerges predominantly at the transcriptomic level in neurons, but not in glial cells. C_LIO_LIAPOE4 and A{beta} drive largely non-overlapping physiological changes in preclinical stages of disease, but converge at the level of network hyperexcitability. C_LIO_LIAPOE4xA{beta} neuronal synergy converges on a conserved lysosomal-ER proteostasis axis. C_LIO_LIHumanized APOE4xA{beta} KI mice provide a physiologically relevant model to dissect early AD pathogenesis in preclinical stages C_LI

neuroscience↗

bioRxiv2025

An EEG-based Multi-Scale Hybrid Attention and Squeeze Network for objective taste assessment

Taste perception is central to flavor experiences. Electroencephalography (EEG) signals carry rich information about taste perception. Because these neural data are independent of verbal reports and bypassing conscious filtering, EEG holds promise as a neural-sensing platform for objective taste representation. However, taste-evoked EEG signals exhibit complex spatiotemporal and spectral dynamics that require advanced computational approaches to decode effectively. This study developed an EEG-based Multi-Scale Hybrid Attention and Squeeze Network (MHASNet) for taste evaluation. EEG signals evoked by distinct taste stimuli were recorded, a dedicated taste-EEG dataset was compiled, and a novel deep learning architecture, MHASNet, was designed to classify these signals. MHASNet synergistically integrates multi-scale convolutions to capture temporal dynamics across different time scales, dual-attention mechanisms to localize discriminative brain regions and electrode positions, and a squeeze-and-excitation module to optimize frequency-band contributions-collectively enabling precise extraction of taste-specific neural signatures. Results showed that the proposed model achieved superior performance across five taste categories (sweet, sour, salty, bitter, and tasteless), with 94.33% accuracy, 91.37% F1-score, 91.89% precision, and 92.07% recall. These results surpass benchmark models while maintaining millisecond-level inference latency suitable for real-time applications. By complementing subjective evaluations and instrumental analyses, the model offers an objective, neurophysiology-based solution for taste evaluation.

neuroscience↗

bioRxiv2025

GNL3 SUMOylation is essential for DNA double-strand break repair by homologous recombination

DNA double-strand break (DSB) repair via homologous recombination (HR) is critical for maintaining genomic integrity and requires proper DNA end resection to generate single-stranded DNA (ssDNA) overhangs. However, the mechanisms governing this critical step in cells remains poorly understood. Here, we report that GNL3, a nucleolar GTP-binding protein, plays a key role in HR repair via regulating DNA end resection dependently on its SUMOylation. Ectopic expression of wild-type, but not the SUMO-defective K196R mutant, GNL3 completely abolished DNA damage response induced by knockdown of endogenous GNL3. GNL3 interacts with the BLM-DNA2 helicase-nuclease complex and is critical for DNA end resection and subsequent loading of RPA and RAD51. This interaction requires SUMOylation and SUMO-interacting motifs (SIMs) in both proteins. We further demonstrate that USP36, a nucleolar deubiquitinating enzyme, functions as a novel SUMO ligase for GNL3, while the SUMO protease SENP3 deSUMOylates GNL3. Notably, several breast cancer-derived GNL3 variants that disrupt its SUMOylation or SIM fail to interact with the BLM-DNA2 complex. Knockdown of GNL3 sensitizes HR-proficient breast cancer cells to etoposide or Olaparib treatment. Together, our results reveal that GNL3 SUMOylation is crucial for HR repair and suggest that targeting GNL3 SUMOylation may induce HR deficiency, thereby sensitizing breast cancers to DNA damage-inducing agents.

cell biology↗

bioRxiv2025

Spatiotemporal control of PIWI compartmentalization by mitochondrial scaffolds defines pachytene piRNA pathway organization

Pachytene piRNAs are the least understood class of piRNAs in the mammalian male germ line. During meiosis, their biogenesis occurs near mitochondrial outer membrane in germ granules known as intermitochondrial cement (IMC). However, how mitochondrial factors regulate the trafficking of PIWI proteins into and out of the IMC remain poorly understood. Here we show that the cytoplasmic PIWI proteins MILI and MIWI are recruited for pachytene piRNA biogenesis via distinct mitochondrial membrane proteins. Loss of the mitochondrial scaffold protein ASZ1 during meiosis in mice disrupts multiple downstream biogenesis steps, leading to misregulation of MILI and MIWI, failure of IMC formation, and a near-complete loss of mature pachytene piRNAs. Strikingly, despite the drastic depletion of pachytene piRNAs, LINE1 transposon silencing remains unaffected. We identify three classes of pachytene piRNA pathway components that coordinate piRNA production and compartmentalization. Our findings reveal that chromatoid body precursors serve as a central hub for the accumulation of pachytene PIWI-piRNA complexes, thus establishing a connection between IMC-based biogenesis and downstream piRNA function.

developmental biology↗

bioRxiv2025

Sex differences in thyroid aging and their implications in thyroid disorders: insights from gene regulatory networks

Thyroid disorders are more prevalent in females than males, with sex differences in disease risk varying by age, e.g. although risk of anaplastic thyroid carcinoma (ATC) increases with age for both sexes, females are diagnosed at younger ages than males. In contrast, Hashimotos thyroiditis (HT) is more common in ages 30-50, with females having higher risk. These patterns suggest that in thyroid, gene regulatory processes evolve with age in a sex-specific manner. To characterize these sex differences in thyroid aging, we estimated and analyzed individual-specific gene regulatory networks using BONOBO and PANDA algorithms. We found that regulatory patterns of pathways involved in cell proliferation, immune response, and metabolic processes vary nonlinearly with age, with sex-specific differences in direction and rate. Females show two inflection points around ages 40 and 60, when regulatory patterns show significant change for most pathways, consistent with changes in targeting by estrogen and androgen receptors. Investigating gene regulatory patterns in HT and ATC, we observed that disease-relevant immune and metabolic pathways changed with age in the same direction as they change in ATC. In contrast, in HT, disease-related regulatory patterns are in the opposite direction of aging. In age groups where HT is most diagnosed, regulatory patterns in females more closely resembled disease states than males, highlighting sex-biased regulations role in elevated disease risk in females. Our findings demonstrate immune response and metabolic processes are regulated in age- and sex-specific manner, potentially driving variation in disease risk, emphasizing the need for tailored screening and prevention strategies. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=113 SRC="FIGDIR/small/684890v3_ufig1.gif" ALT="Figure 1"> View larger version (39K): org.highwire.dtl.DTLVardef@cd93aeorg.highwire.dtl.DTLVardef@9a913aorg.highwire.dtl.DTLVardef@1b4a862org.highwire.dtl.DTLVardef@19adca8_HPS_FORMAT_FIGEXP M_FIG C_FIG

systems biology↗

bioRxiv2025

KSMoFinder - Knowledge graph embedding of proteins and motifs for predicting kinases of human phosphosites

MotivationProtein kinases regulate cellular signaling pathways through a cascade of phosphorylation activity, selectively targeting specific residues on substrate proteins (phosphosites). Determining the characteristics of kinases that phosphorylate specific substrates have been extensively studied. Most tools utilize amino acid sequence motifs around phosphosites but do not consider the biological characteristics of substrate proteins. ResultsWe present KSMoFinder, a kinase-substrate-motif prediction model that learns factors beyond motif similarities by integrating the biological contexts of proteins. We learn the semantics in a knowledge graph containing contextual relationships of proteins, kinase-specific motifs and motif composition, and represent the proteins and motifs as embedded vectors. Using the representations as features, we train a supervised deep learning classifier to identify kinase-phosphosite relationships. We use ground truth kinase-substrate-motif dataset from iPTMnet and PhosphositePlus and evaluate the prediction performance of KSMoFinder. Pairwise comparative assessments with prior kinase-substrate prediction tools demonstrate the superior performance of KSMoFinder. KSMoFinder trained using our knowledge graph embeddings surpasses the prediction performances using embeddings of popular protein language models such as ProtT5, ESM2 and ESM3 with a ROC-AUC of 0.851 and PR-AUC of 0.839 on a testing dataset with equal number of positives and negatives. Unlike most existing tools, KSMoFinder can be utilized to predict at the motif and at the substrate protein level. Availability and implementationAll code to reproduce the results are available at https://github.com/manju-anandakrishnan/KSMoFinder. All data and KSMoFinder predictions are deposited at https://doi.org/10.5281/zenodo.15730847.

bioinformatics↗

bioRxiv2025

Tumor-intrinsic MHC-II activation in pancreatic ductal adenocarcinoma enhances immune response and treatment efficacy

Pancreatic ductal adenocarcinoma (PDAC) is characterized by an immunosuppressive tumor microenvironment (TME) and poor prognosis. While major histocompatibility complex class II (MHC-II) expression is traditionally associated with professional antigen-presenting cells, its role in PDAC malignant cells remains underexplored. Herein, we utilized single-cell RNA sequencing (scRNA-seq), spatial transcriptomics, bulk RNA sequencing, multiplex immunohistochemistry (mIHC) and ex vivo studies in culture with both human and murine models to investigate the prognostic relevance of MHC-II expression in malignant PDAC cells. Elevated MHC-II expression in malignant cells was strongly associated with increased infiltration of CD4+ T and CD8+ T cells in human PDAC, and pronounced co-localization with plasma cells, indicative of an antigen-activated immune microenvironment. In the KPC mouse model of PDAC, pharmacologic induction of MHC-II expression by cobimetinib treatment in malignant epithelial cells significantly enhanced the therapeutic response to immune checkpoint blockade (ICB). These findings highlight the role of malignant cell- intrinsic MHC-II expression in promoting antigen presentation and fostering an anti-tumor immune microenvironment. Our results position MHC-II as a promising prognostic biomarker and therapeutic target in PDAC, paving the way for novel immunomodulatory strategies. Summary of highlightsSingle-cell and spatial transcriptomic analyses reveal that elevated MHC-II expression in malignant PDAC cells correlates with increased infiltration of CD4 and CD8 T cells. Stimulating MHC-II expression in tumors effectively enhances immunotherapeutic responses to ICB in the PDAC KPC mouse model, including PDAC tumors previously resistant to therapeutic interventions. MHC-II serves as a prognostic biomarker and a promising target for immunotherapy in PDAC.

cancer biology↗