Biology subjects

Mizuno, T.

Publications and source records attributed to Mizuno, T..

At least 19 recordsLinked to original sources

bioRxiv2026

Once in a blue moon: an unexpected evolutionary route to blue floral pigmentation in orchids

Blue flowers are rare in the hyperdiverse orchid family, and the genetic mechanisms enabling their blue delphinidin-based pigmentation remain poorly understood. Within the Australian orchid subtribe Caladeniinae (~350 spp.), repeated transitions from maroon to blue flowers, followed by reversions to maroon, provide a unique opportunity to investigate blue flower evolution. Using phylotranscriptomics, anthocyanin profiling, molecular evolutionary analyses, and gene expression data from 51 species, we reconstructed the genetic changes associated with these transitions. Despite its near-universal role in delphinidin biosynthesis across flowering plants, unexpectedly, we found no evidence for a functional CYP75A-mediated flavonoid 3'5'-hydroxylase (F3'5'H) pathway. Instead, blue-flowered species possess a duplicated CYP75B lineage gene (CYP75B-D) with apparent 5'-hydroxylase activity, a derived D-type dihydroflavonol reductase (DFR), and a distinct anthocyanin biosynthetic pathway co-expression pattern. Phylogenetic analyses suggest that the D-type DFR pre-dated blue flower evolution, representing a permissive evolutionary shift later coupled with the evolution of CYP75B-D. Notably, blue flower evolution coincided with repeated pollinator transitions and major palaeoenvironmental change during the Middle Miocene. Together, these findings reveal an unexpected route to blue floral pigmentation and suggest that its evolution arose through the interaction of molecular innovation, regulatory evolution, and ecological opportunity.

evolutionary biology↗

bioRxiv2026

Biosensor-enabled single-molecule enzyme activity profiling for function-based molecular counting

Single-molecule enzyme activity analysis enables proteoform-resolved quantification of catalytically active enzyme molecules but remains limited by the need for enzyme-specific fluorogenic substrates, particularly for enzymes involved in central metabolism. Here we establish a biosensor-based framework that integrates genetically encoded fluorescent metabolite biosensors with single-molecule enzyme activity profiling (SEAP), enabling native enzymatic reactions to be monitored without enzyme-specific fluorogenic substrates. By employing different biosensors, the framework was applied to distinct metabolic enzymes, including lactate dehydrogenase and pyruvate kinase. Biosensor-based SEAP enables absolute quantification of active enzyme molecules, transforming conventional activity measurements (U/mL) into molecular abundance (molecules/mL). Applying this approach to single-cell functional proteoform analysis revealed functional heterogeneity in lactate dehydrogenase activity during T-cell activation. In addition, analysis of circulating enzymes showed that much of the plasma lactate dehydrogenase pool could be quantitatively accounted for by physiological erythrocyte turnover. These results establish genetically encoded fluorescent protein-based metabolite biosensors as modular optical transducers for single-molecule functional enzyme analysis across cellular and circulating samples.

biochemistry↗

bioRxiv2026

Early detection of pancreatic ductal adenocarcinoma by single-molecule profiling of pancreatic enzyme activities

Pancreatic ductal adenocarcinoma (PDAC) remains a leading cause of cancer-related mortality, largely due to diagnosis at advanced stages. Early detection through minimally invasive liquid biopsy holds promise for improving patient outcomes. Here, we report a blood-based liquid biopsy platform based on single-molecule enzyme activity profiling (SEAP), which detects proteoform-level alterations in circulating pancreatic enzymes at single-molecule resolution. Using a tissue-centric biomarker discovery strategy, we identified activity signatures of pancreas-specific digestive enzymes associated with PDAC. A combinatorial classifier detected PDAC (stage I-IV) with 95.5% specificity and 75.0% sensitivity (74.2% for stage I-II) across 690 blood samples collected from multiple hospitals and biobanks. Performance was further validated in an independent cohort enriched for early-stage disease, where 54.2% (13/24) of stage IA, 64.3% (9/14) of stage IB, and 21.4% (3/14) of stage 0 lesions were classified as positive. These findings support the clinical potential of SEAP for early detection of PDAC.

biochemistry↗

bioRxiv2026

Single-Molecule Activity Profiling of Glycosidase Proteoforms Using Water-Soluble Fluorogenic Probes

In this study, we report a series of highly water-soluble fluorogenic probes for glycosidases useful for microdevice-based single-molecule enzyme activity profiling (SEAP). The assay was able to detect trace glycosidase activities in blood with high sensitivity and with proteoform resolution. This platform revealed the potential of -mannosidase as sensitive blood-based liquid-biopsy biomarkers for liver injury, highlighting the potential of this approach for activity-based diagnostics.

biochemistry↗

bioRxiv2026

Synthesis based on covalent capture and release (SCCR): a programmable strategy for automated preparation of protease-activatable molecules

Enzyme-activatable chemical tools, including fluorogenic probes and prodrugs, are essential in chemical biology and targeted therapeutics but remain challenging to access in structurally diverse forms because their synthesis is often bespoke and difficult to standardize. Here, we introduce synthesis based on covalent capture and release (SCCR) as a programmable chemical strategy that enables the modular assembly of protease-activatable molecules through specifically designed protecting-group logic. The SCCR framework establishes a standardized capture-elongation- release workflow that decouples molecular diversification from individual synthetic optimization, thereby enabling automated preparation of complex libraries. Using this chemistry, we generated a diverse set of fluorogenic probes and applied them to single-molecule enzyme activity analyses to identify candidate activity-based biomarkers of liver diseases. The generality of the SCCR strategy was further demonstrated by extending the same chemical logic to the preparation of antibody-drug conjugate (ADC) linkers, allowing systematic evaluation of plasma stability and cytotoxic potential. By establishing a programmable capture-release chemistry for the synthesis of enzyme-activatable molecules, this work provides a generalizable chemical foundation for the scalable and automated construction of functional small-molecule tools across biological and translational research.

biochemistry↗

bioRxiv2025

Designing Cell Cell Relation Extraction Models: A Systematic Evaluation of Entity Representation and Pre-training Strategies

Extracting cell-cell relations from biomedical literature is essential for understanding intercellular communication in immunity, inflammation, and tissue biology. However, cell-cell relation extraction has not been established as a standalone biomedical relation extraction task, and no benchmark corpus or systematic evaluation framework currently exists. Fully manual corpus construction is costly and difficult to scale, limiting literature-based analyses of cell-cell communication. Here, we define a sentence-level cell-cell relation extraction task and construct complementary manually annotated corpora under realistic annotation constraints. To enable scalable annotation, rule-based literature mining is used solely as an annotation accelerator to identify candidate sentences, while all relation labels are assigned manually. In addition, an independently annotated PubMed corpus without rule-based filtering is constructed to evaluate robustness on natural sentence distributions. Using these resources, we evaluate representative model configurations involving entity indication strategies, classification architectures, and continued pre-training. Our results show that cell-cell relation extraction remains challenging under realistic conditions. Increasing training data size yields consistent performance gains, and specific combinations of entity-aware architectures and continued pre-training provide modest robustness improvements. Nevertheless, performance on unfiltered PubMed sentences remains in the 70% accuracy range, and error analyses indicate that failures cannot be readily explained by simple surface-level factors. Comparisons with general-purpose large language models further suggest that task complexity, rather than model class, is the primary limiting factor. Together, these findings establish a practical foundation for literature-scale cell-cell relation extraction while clarifying its intrinsic limitations.

bioinformatics↗

bioRxiv2025

The caninized anti-canine PD-1 monoclonal antibody in canine oral malignant melanoma: Efficacy and exploratory biomarker analysis

BackgroundCanine oral malignant melanoma (OMM) is a highly aggressive tumor with few treatment options. Because of its biological similarity to human mucosal melanoma, canine OMM represents a valuable spontaneous model for translational immunotherapy studies. Anti-PD-1 antibody therapy has shown promise in canine OMM; however, predictive biomarkers for treatment response and survival have not been identified. MethodsWe conducted a multicenter, prospective, investigator-initiated clinical trial to evaluate the safety and efficacy of a caninized anti-canine PD-1 monoclonal antibody (ca-4F12-E6) in 150 dogs with advanced OMM. The dogs were administered 3 mg/kg ca-4F12-E6 intravenously every two weeks. Treatment response was assessed using cRECIST v1.0. Biomarker analyses included peripheral blood parameters, cytokine/chemokines, peripheral lymphocyte subsets, microsatellite instability (MSI), immunohistochemistry for immune cell and mismatch repair protein markers, and RNA sequencing of the tumor tissue. Associations with clinical outcomes were determined by logistic regression and Cox proportional hazard models. ResultsThe best overall response rate was 16.7%. Treatment-related adverse events occurred in 40.0% of the dogs, which were primarily grade 1-3. Increased baseline white blood cell, neutrophil count, and C-reactive protein (CRP) levels were significantly associated with poor response, shorter progression-free survival (PFS), and reduced overall survival (OS). MSI-high tumors were associated with significantly prolonged OS compared with MSI-low/microsatellite stable tumors. Transcriptome analysis revealed differentially expressed genes and enriched immune-related pathways in responders, though limited by sample size. Conclusionca-4F12-E6 exhibited durable antitumor activity and an acceptable safety profile in dogs with OMM. Baseline systemic inflammatory markers and MSI status may serve as predictive biomarkers for clinical outcomes. The results support the use of canine OMM as a comparative model for human immuno-oncology and biomarker discovery.

cancer biology↗

bioRxiv2025

Local niche-derived immunosuppressive CXCR2+ cells impair antiviral immunity

Symptom severity after viral infection varies among individuals, yet its mechanism remains poorly understood. We categorized mice into recovering and non-recovering groups based on body weight after infection of the same titer of vesicular stomatitis virus (VSV). We revealed that in the olfactory bulb (OB) where VSV initially expands, non-recovering mice exhibited an anti-inflammatory environment, suggesting its detrimental effect. Importantly, CXCR2+ cells resembling immunosuppressive myeloid-derived suppressor cells (MDSCs) were more abundant in the OB of non-recovering mice than in that of recovering mice after VSV infection. Depleting CXCR2+ MDSC-like cells from the brain increased inflammatory responses and the animals survival after infection. Furthermore, site-specific labeling indicated that a significant fraction of these cells in the OB originate from the skull-bone marrow (skull-BM) as well as circulation. This study reveals the lethal effects of CXCR2+ MDSC-like cells on local immune responses to viral infection, highlighting their therapeutic potential for antiviral defense.

immunology↗

bioRxiv2025

Platform of single-molecule enzyme activity-based liquid biopsy for detection of pancreatic adenocarcinoma at early stages

Liquid biopsy analyzes biomolecules in biofluids to provide critical information for the early detection of diseases and the development of more effective treatment strategies. Among these biomolecules, the functional states of proteins offer particularly valuable insights due to their direct correlation with phenotypic changes. However, current approaches for assessing protein function in liquid biopsies are often limited by low detection sensitivity. Here, we present a liquid biopsy platform that analyzes single-molecule protease/peptidase activity to detect pathological alterations in enzyme function in blood samples. The platform demonstrates potential for identifying patients with early-stage (stage I-II) pancreatic ductal adenocarcinoma (PDAC).

biochemistry↗

bioRxiv2025

Development of single-molecule enzyme activity assay for serine hydrolases using activity-based protein labeling probes

Single-molecule enzyme activity assays have proven their potential in elucidating aberrant protein function at the proteoform level. However, the limited number of targetable enzymes is the major drawback of such assays. Here, we report the development of single-molecule enzyme activity assays utilizing activity-based probes that label active enzymes in an enzyme superfamily-wide manner. A proof-of-principle using fluorophosphonate-based probes was conducted to detect the active form of serine hydrolases such as PSA and granzyme B at the single-molecule level in complex biological systems such as blood. The assay suggested the potential of active granzyme B in blood as a suitable biomarker of liver damage with specific activation of immune cells.

biochemistry↗

bioRxiv2025

Model-informed Deep Q-Networks to Guide Infliximab Dosing in Pediatric Crohn's Disease

Model-informed precision dosing (MIPD) utilizes pharmacokinetic/pharmacodynamic (PK/PD) models to optimize drug therapy. However, conventional MIPD often requires manual simulation and regimen selection, which are time-consuming and demand specialized expertise. Reinforcement learning (RL), in which an agent learns optimal decisions through iterative interactions with an environment, offers a scalable and automated alternative. In this study, we developed a model-informed Deep Q-Network (DQN) to personalize infliximab dosing for patients with Crohns disease. The DQN was trained in a simulation environment incorporating a population PK model, inter-individual variability, and assay error. Virtual patients with randomly sampled covariates were used to explore dosing strategies at infusions 1, 3, and 4. Doses ranged from 1 to 10 mg/kg at infusion 1 and from 1 to 20 mg/kg thereafter, with intervals of 4 to 12 weeks. The reward function prioritized achieving trough concentrations of 18-26 {micro}g/mL before infusion 3 and 5-10 {micro}g/mL before infusions 4 and 5, while penalizing overtreatment and additional infusions. The DQN policy converged after 80,000 episodes, yielding target attainment probabilities (PTAs) of 92.9% and 98.4% at infusions 4 and 5, respectively, in 1,000 virtual patients. High doses (11-20 mg/kg) were selected in only 0.2% of cases. At infusion 4, 66.8% of patients received an 8-week interval, and 57.3% at infusion 5. Retrospective real-world validation showed that patients whose actual doses matched DQN recommendations had trough levels significantly closer to target ranges. These findings support the feasibility of using DQN-based agents to enhance and automate infliximab individualized dosing in pediatric populations.

pharmacology and toxicology↗

bioRxiv2025

Hybrid Population PK-Machine Learning Modeling to Predict Infliximab Pharmacokinetics in Pediatric and Young Adult Patients with Crohn's Disease

Population pharmacokinetic (PK) model-based Bayesian estimation is widely used for dose individualization, particularly when sample availability is limited. However, its predictive accuracy can be compromised by factors such as misspecified prior information, intra-patient variability, and uncertainties in PK variations. In this study, we developed a hybrid approach that combines machine learning (ML) with population PK-based Bayesian methods to improve the prediction of infliximab concentrations in children with Crohns disease. We calculated prediction errors between Bayesian-estimated and observed infliximab concentrations from 292 measurements across 93 patients. Incorporating clinical patient features, we explored various ML algorithms, including linear regression, random forest, support vector regression, neural networks, and XGBoost to correct the Bayesian-based prediction errors. The predictive performance of these ML models was assessed using root mean square error (RMSE) and mean prediction error (MPE) with 5-fold cross-validation. For Bayesian estimation alone, the RMSE and MPE were 4.8 {micro}g/mL and -0.67 {micro}g/mL, respectively. Among the ML algorithms, the XGBoost model demonstrated the best performance, achieving an RMSE of 3.78 {+/-} 0.85 {micro}g/mL and an MPE of -0.03 {+/-} 0.69 {micro}g/mL in 5-fold cross-validation. The ML-corrected Bayesian estimation significantly reduced the absolute prediction error compared to Bayesian estimation alone. This hybrid population PK-ML approach provides a promising framework for improving the predictive performance of Bayesian estimation, with the potential for continuous learning from new clinical data to enhance dose individualization. Key pointsO_LIA new hybrid model combining population pharmacokinetic model-based Bayesian estimation and machine learning significantly improved the accuracy of infliximab concentration predictions in young adult and pediatric patients with Crohns disease. C_LIO_LIThe developed hybrid model can facilitate infliximab individualized dosing by accounting for changes in clinical conditions and patient-specific factors that the conventional Bayesian estimation approach may not address, and can be integrated into precision dosing dashboards, such as RoadMAB, for real-world clinical application. C_LIO_LIThis study indicates that model predictive accuracy can be enhanced by combining the Bayesian method with machine learning, even with a relatively small amount of clinical data. This is particularly encouraging for specific populations, such as pediatric patients, where obtaining rich clinical data is challenging. C_LI

pharmacology and toxicology↗

bioRxiv2025

Low-Background Cancer Imaging With a Bioorthogonal Fluorescence Probe and Engineered Reporter Enzyme Bearing a Targeting Moiety

Combinatorial use of an antibody-reporter enzyme conjugate and a fluorescence probe activated by the enzyme is a powerful strategy for fluorescence-guided cancer surgery. However, conventional probes for typical reporter enzymes are insufficiently bioorthogonal, resulting in high background signals in non-target tissues. We screened a library of HMRef (rhodol derivative)-based fluorescence probes bearing various sugar moieties, and discovered that HMRef-{beta}-D-fucose is bioorthogonal in mammalian systems, but is activated by a metagenomic glycosidase, Td2F2. Directed evolution generated a mutant with a kcat/Km value for HMRef-{beta}-D-fucose of 3.3 x 105/M/sec, 7.3 times higher than that of wild-type Td2F2 and comparable to that of {beta}-galactosidase (LacZ) with a corresponding probe. Theoretical calculation suggested that E296G mutation in Td2F2 causes structural changes that facilitate the probes access to the enzymes active site. In a proof-of-concept study, cancer cells were visualized with a minimal background in the mesentery of a mouse model of peritoneally disseminated human-ovarian-cancer-derived SKOV-3 cells, which endogenously express HER2, by using HMRef-{beta}-D-fucose together with engineered Td2F2 conjugated/fused to a HER2-binding antibody/nanobody.

bioengineering↗

bioRxiv2024

Establishment of an easy-to-construct liver injury mouse model for longitudinal analysis by drinking-water administration of 4,4'-methylenedianiline

Background & AimsLongitudinal animal models are essential for understanding the temporal dynamics of liver injury and recovery. While drinking water-based administration is ideal for sustained exposure in high feasibility, available compounds are limited, with thioacetamide (TAA) being the primary option. Here, we aimed to establish a novel drinking water-induced mouse model of cholestatic liver injury using 4,4-methylenedianiline (MDA), and to characterize its pathological trajectory in comparison to the TAA model. MethodsMice were administered MDA via drinking water for 28 Days. To elucidate the early events that give rise to chronic pathological divergence, we conducted a multi-layered analysis comprising plasma biochemical assays, immune cell profiling by flow cytometry, and hepatic transcriptomics at five time points during the early phase. The MDA model was evaluated against the established TAA model. ResultsMDA administration induced sustained ALT elevation, peribiliary fibrosis, and spatially irregular focal hepatocellular necrosis, distinguishing it from the centrilobular injury observed with TAA. Additionally, the MDA model showed significant elevations in ALP, TBIL, and TCHO, indicating cholestatic liver dysfunction. Early-phase analyses revealed model-specific differences in immunological and molecular responses, including increased CD8 T cell populations and enrichment of fibrinolysis-related gene expression in MDA-DW mice. ConclusionsWe present the MDA-DW model as a novel, longitudinally tractable liver injury model that complements existing systems by capturing alternative spatial, immunological, and transcriptional patterns of injury. This model offers a valuable platform for dissecting the temporal dynamics of liver disease progression in experimental settings. Significance StatementWe developed a cost-effective, non-invasive mouse model of cholestatic liver injury using drinking-water administration of 4,4-methylenedianiline (MDA). This model exhibits periportal-predominant damage, peribiliary fibrosis, and spatially irregular focal hepatocellular necrosis, distinct from conventional centrilobular models. Early-phase multi-omics analysis revealed immunological and transcriptomic differences, including increased CD8 T cells and activation of fibrinolysis-related pathways. The low mortality rate and ease of implementation enable long-term studies and cross-sectional comparisons across time points or interventions. This study provides not only a practical model for investigating chronic liver injury, but also a rich time-series, multi-view dataset, offering a valuable resource for advancing research on liver pathophysiology and toxicological mechanisms.

pharmacology and toxicology↗

bioRxiv2024

GLDADec: marker-gene guided LDA modelling for bulk gene expression deconvolution

Inferring cell type proportions from bulk transcriptome data is crucial in immunology and oncology. Here, we introduce GLDADec (Guided LDA Deconvolution), a bulk deconvolution method that guides topics using cell type-specific marker gene names to estimate topic distributions for each sample. Through benchmarking using blood-derived datasets, we demonstrate its high estimation performance and robustness. Moreover, we apply GLDADec to heterogeneous tissue bulk data and perform comprehensive cell type analysis in a data-driven manner. We show that GLDADec outperforms existing methods in estimation performance and evaluate its biological interpretability by examining enrichment of biological processes for topics. Finally, we apply GLDADec to TCGA tumor samples, enabling subtype stratification and survival analysis based on estimated cell type proportions, thus proving its practical utility in clinical settings. This approach, utilizing marker gene names as partial prior information, can be applied to various scenarios for bulk data deconvolution. GLDADec is available as an open-source Python package at https://github.com/mizuno-group/GLDADec.

bioinformatics↗

bioRxiv2023

Olfaction foraging in visually oriented tropical arboreal ants Oecophylla smaragdina: Implications for insect predation studies using artificial sentinel prey

Predation is the key to understanding trophic interactions. Because of the brief and cryptic nature of predatory behavior, sentinel prey has been widely adopted as an indirect way to identify predators and understand trophic interactions. However, sentinel prey presents only static visual cues, potentially biasing toward visually oriented predators whilst ignoring those that require other sensory cues to forage. Despite this, the effectiveness of sentinel prey has rarely been tested. Here, we focused on the weaver ant, Oecophylla smaragdina, a keystone predator widespread in the Asian and Australian tropics. As this species has large eyes and is known to visually navigate in their arboreal habitats, we hypothesized that they rely on visual cues to forage and that their predatory behavior will be captured by caterpillar-shaped sentinel prey. Ants were collected as colonies, and preference trials on baits were conducted using combinations of olfactory and static visual cues including the caterpillar shape. Surprisingly, O. smaragdina showed little or no preference for baits in the absence of olfactory cues and did not differentiate the shapes of baits regardless of the presence of olfactory cues. Our results indicate that O. smaragdina is likely to make predatory decisions based primarily on olfactory cues, while visual cues might be used for other behaviors. Furthermore, O. smaragdina is likely to be left out by the predation studies using sentinel prey models which is particularly alarming considering the dominant role of this species in the trophic interactions of tropical rainforests. Our study demonstrates that morphological characteristics, arboreal habitats, and visually oriented behavior do not necessarily suggest the use of static visual cues for predatory decisions. We suggest that sentinel prey models should not be used alone when the dominant predators are unlikely to use visual cues to make predatory decisions.

ecology↗

bioRxiv2023

Investigation of normalization procedures for transcriptome profiles of compounds oriented toward practical study design

The transcriptome profile is a representative phenotype-based descriptor of compounds, widely acknowledged for its ability to effectively capture compound effects. However, the presence of batch differences is inevitable. Despite the existence of sophisticated statistical methods, many of them presume a substantial sample size. How should we design a transcriptome analysis to obtain robust compound profiles, particularly in the context of small datasets frequently encountered in practical scenarios? This study addresses this question by investigating the normalization procedures for transcriptome profiles, focusing on the baseline distribution employed in deriving biological responses as profiles. Firstly, we investigated two large GeneChip datasets, comparing the impact of different normalization procedures. Through an evaluation of the similarity between response profiles of biological replicates within each dataset and the similarity between response profiles of the same compound across datasets, we revealed that the baseline distribution defined by all samples within each batch under batch-corrected condition is a good choice for large datasets. Subsequently, we conducted a simulation to explore the influence of the number of control samples on the robustness of response profiles across datasets. The results offer insights into determining the suitable quantity of control samples for diminutive datasets. It is crucial to acknowledge that these conclusions stem from constrained datasets. Nevertheless, we believe that this study enhances our understanding of how to effectively leverage transcriptome profiles of compounds and promotes the accumulation of essential knowledge for the practical application of such profiles.

bioinformatics↗

bioRxiv2023

Barcoding of small extracellular vesicles with CRISPR-gRNA enables comprehensive, subpopulation-specific analysis of their biogenesis/release regulators

Small extracellular vesicles (sEVs) are important intercellular information transmitters in various biological contexts, but their release processes remain poorly understood. Herein, we describe a high-throughput assay platform, CRISPR-assisted individually barcoded sEV-based release regulator (CIBER) screening, for identifying key players in sEV release. CIBER screening employs sEVs barcoded with CRISPR-gRNA through the interaction of gRNA and dead Cas9 fused with an sEV marker. Barcode quantification enables the estimation of the sEV amount released from each cell in a massively parallel manner. Barcoding sEVs with different sEV markers in a CRISPR pooled-screening format allows genome-wide exploration of sEV release regulators in a subpopulation-specific manner, successfully identifying previously unknown sEV release regulators and uncovering the exosomal/ectosomal nature of CD63+/CD9+ sEVs, respectively, as well as the synchronization of CD9+ sEV release with the cell cycle. CIBER should be a valuable tool for detailed studies on the biogenesis, release, and heterogeneity of sEVs.

cell biology↗