Biology subjects

Yoshimura, S.

Publications and source records attributed to Yoshimura, S..

11 recordsLinked to original sources

bioRxiv2026

Robust Cooperative Learning for Multiview Regression via Soft-IPOD

Multiview data analysis has become increasingly important for integrating heterogeneous information from multiple data sources. Cooperative learning provides a flexible framework for multiview integration by bridging early and late fusion strategies. However, its use of squared loss makes it sensitive to outliers. In this study, we propose robust cooperative learning based on the soft-IPOD framework. The proposed method introduces an observation-specific outlier absorption vector into the cooperative learning objective, thereby reducing the influence of outlying observations while preserving information sharing across views. The resulting optimization problem can be formulated as a weighted Lasso-type problem using an augmented design matrix and can be naturally extended to more than two data views. Numerical simulations and a real-data application to labor onset prediction demonstrate the predictive performance of the proposed method in the presence of outliers.

bioinformatics↗

bioRxiv2026

Cerebral microthrombi promote focal Cav-1-dependent blood-brain barrier impairment, triggering neuroinflammation and neuronal damage after traumatic brain injury

Traumatic brain injury (TBI) is frequently accompanied by blood-brain barrier (BBB) dysfunction, yet the microvascular events that initiate barrier failure and secondary neural injury remain poorly understood. Using highly sensitive fluorescent nanoscale tracers, correlative light and electron microscopy, single-cell transcriptomics, and genetic manipulation of Caveolin-1 (Cav-1), we show that BBB leakage after TBI in mice occurs focally at sites of cerebral microthrombus formation. Endothelial cells adjacent to microthrombi remain structurally intact, but exhibit a distinct transcriptional program enriched in both thrombosis- and transcytosis-related genes, including Cav-1. Microthrombi promote Cav-1-dependent transcytosis and size-selective extravasation of blood-borne proteins, exposing the surrounding parenchyma to circulating factors and inducing focal microglia activation and neuronal damage in vivo. In human induced pluripotent stem cell-derived neurons, these proteins directly impaired neuronal structure and network function. Genetic deletion of Cav-1 markedly reduces microthrombus-associated BBB leakage, immune-cell diapedesis, microglial activation and neuronal loss, whereas endothelial re-expression of Cav-1 using AAVs restores leakage. Human TBI data reveal early dynamic coagulation dysregulation associated with adverse clinical outcome, focal cerebral vascular injury and endothelial transcriptional signatures consistent with reduced barrier integrity and increased caveolae-associated transport. Together, these findings identify cerebral microthrombi as focal sites of Cav-1-dependent BBB dysfunction and establish pathological endothelial transcytosis as a mechanistic link between post-traumatic microvascular thrombosis, blood protein extravasation and secondary neuronal injury. Targeting Cav-1-dependent transcytosis may therefore provide a strategy to limit pathological BBB permeability and secondary injury after TBI.

neuroscience↗

bioRxiv2026

Glucocorticoid receptor activation reorganizes Wnt/LEF1 regulatory circuitry associated with therapeutic response in B-cell acute lymphoblastic leukemia

Glucocorticoids (GCs) are foundational to B-cell acute lymphoblastic leukemia (B-ALL) therapy, yet how glucocorticoid receptor (GR) activation interacts with lineage regulatory circuitry to shape therapeutic response remains incompletely understood. Using LEF1 to interrogate canonical Wnt regulatory circuitry, we show that GR activation reorganizes GR+LEF1 elements into recruitment, retention, and dissociation states. LEF1 dissociates from Wnt-responsive regulatory elements while being recruited to GRE-rich GR-associated elements that progressively acquire chromatin accessibility, H3K27ac, BRD4 and BRG1. RUNX1 and AP-1 show class-dependent redistribution across LEF1-defined states, extending organization beyond LEF1. GR+LEF1 reorganization is observed in patient-derived B-ALL. GR+LEF1 elements lose accessibility in GC-resistant B-ALL and during diagnosis-to-relapse evolution, with recruitment elements enriched for remodeling. Perturbation of resistance-associated GR+LEF1 elements at BIM and BMF alters GC-induced apoptotic signaling. Together, these findings identify Wnt/LEF1 circuitry as an interface through which therapeutic GR activation reorganizes the B-ALL regulatory landscape into states associated with glucocorticoid response and resistance.

cancer biology↗

bioRxiv2026

Cooperative Learning with Penalized Linear Mixed-Effects Models for High-Dimensional Clustered Multiview Data

In biomedical research, multiple types of high-dimensional data, such as genomic, transcriptomic, proteomic, and metabolomic data, are increasingly collected from the same subjects. Integrating these multiple data views can improve prediction by exploiting shared or complementary information across the views. Cooperative learning provides an agreement-based framework for multiview supervised learning by encouraging predictions obtained from individual views to be similar. However, the original framework assumes independent observations and therefore does not account for clustered structures, such as repeated measurements obtained from the same subject. To address this limitation, we propose Cooperative Learning with a penalized Linear Mixed Model (CL-pLMM) for high-dimensional multiview data with a clustered structure. CL-pLMM replaces the ordinary prediction loss in cooperative learning with a covariance-weighted loss that accounts for within-cluster dependence, while retaining the agreement penalty between views and a Lasso penalty for variable selection. We further show that its objective function can be represented as a penalized linear mixed-effects model applied to augmented data, allowing existing estimation procedures to be used. The performance of CL-pLMM is evaluated through simulation studies under various signal and dependence settings and an application to longitudinal proteomic and metabolomic data for predicting the time to spontaneous labor.

bioinformatics↗

bioRxiv2025

The Genomic Architecture of Human DNA Replication Origins

The exact sites of DNA replication origins in human and other metazoans remain elusive. Examining human whole-genome sequencing data of 2,616 specimens, we observed conjoining reads at 2,025,756 non-random genomic positions, likely arising from nascent DNA and thus defining replication origins. These origins exhibited a 16 bp motif and periodic occurrence at the intervals of 10.5 bp and 200 bp. Genome-wide replication activity is related to the expression of DNA replication-related genes. Across the genome, DNA replication initiation is more active in early replicating regions, correlated with transcription activity in cis, enriched for de novo mutations and trait-associated polymorphisms. Our high-resolution mapping of human DNA replication origins points to molecular features that govern where and when replication begins in the genome.

cell biology↗

bioRxiv2025

LCK-targeting molecular glues overcome resistance to inhibitor-based therapy in T-cell acute lymphoblastic leukemia

Drug resistance is a major challenge in cancer therapy, especially in the context of kinase inhibitors. While targeted protein degradation (TPD) was a distinct mode of action compared to inhibition-based therapeutic targeting, the potential value of TPD in drug-resistant cancer remains unclear. Here, we report the discovery of cereblon-recruiting molecular glue degraders (MGDs) targeting LCK, an oncogenic kinase in T-cell acute lymphoblastic leukemia (T-ALL). By high-throughput screening and medicinal chemistry optimization, we developed a series of MGDs that induced CRBN-dependent degradation of LCK as well as potent cytotoxicity in T-ALL in vitro. Structure-activity relationship analysis and ternary complex modeling revealed a non-canonical degron at the LCK-CRBN interface involving the G-loop, whose mutation disrupts this interaction. Unlike inhibitors and inhibitor-based PROTACs, these MGDs engage LCK in regions distal to the ATP binding site and thus their activities in T-ALL are not affected by gate-keeper LCK mutations that drive resistance to inhibitor-based therapeutics. Taken together, our data underscore the potential of LCK-targeting MGDs as a strategy to overcome kinase inhibitor resistance in T-ALL, highlighting a potentially generalizable strategy in cancer therapy.

cancer biology↗

bioRxiv2024

Development of a novel in silico classification model to assess reactive metabolite formation in the cysteine trapping assay and investigation of important substructures

Predicting whether a compound can cause drug-induced liver injury (DILI) is difficult due to the complexity of its mechanism. The production of reactive metabolites is one of the major causes of DILI, particularly idiosyncratic DILI. The cysteine trapping assay is one of the methods to detect reactive metabolites which bind to microsomes covalently. However, it is cumbersome to use 35S isotope-labeled cysteine for this assay. Therefore, we constructed an in silico classification model for predicting a positive/negative outcome in the cysteine trapping assay to accelerate the drug discovery process. In this study, we collected 475 compounds (436 in-house compounds and 39 publicly available drugs). Using a Message Passing Neural Network (MPNN) and Random Forest (RF) with extended connectivity fingerprint (ECFP) 4, we built machine learning models to predict the covalent binding risk of compounds. The 5-fold cross-validation (CV) and hold-out test were evaluated in random- and time-split trials. Additionally, we investigated the substructures that contributed to positive results in the cysteine trapping assay through the framework of the MPNN model. In the random-split dataset, the AUC-ROC of MPNN and RF were 0.698 and 0.811 in the 5-fold CV, and 0.742 and 0.819 in the hold-out test, respectively. In the time-split dataset, AUC-ROC of MPNN and RF were 0.729 and 0.617 in the 5-fold CV, and 0.625 and 0.559 in the hold-out test, restrictively. This result suggests that the MPNN model has a higher predictivity than RF in the time-split dataset. Hence, we conclude that the in silico MPNN classification model for the cysteine trapping assay have better predictive power. Furthermore, most of the substructures that contributed positively to the cysteine trapping assay were consistent with previous reports such as propranolol, verapamil, and imipramine. This is a new machine learning model that can determine the outcome of the cysteine trapping assay, namely accurately predicting the covalent binding risk as the one of factors of idiosyncratic DILI. We believe that this can contribute to mitigating DILI risk for reactive metabolites at the early stages of drug discovery.

bioinformatics↗

bioRxiv2024

Minimizing variability in the filament middle cerebral artery occlusion model in C57BL/6 mice by surgical optimization - the PURE-MCAo Model

BACKGROUNDIn the intraluminal filament middle cerebral artery occlusion (fMCAo) model, there is considerable variability in infarct volumes, especially in C57BL/6 mice, which often lack the P1 segment of the posterior cerebral artery (PCA) and therefore develop not only MCA but also PCA area infarcts after fMCAo. Another factor contributing to infarct volume variability is collateral flow to the MCA area. The aim of this study was to establish an optimal surgical method to reduce the infarct volume variability in C57BL/6 mice. METHODSC57BL/6 mice were subjected to 60 min of fMCAo with cerebral blood flow monitored by laser Doppler fluxmetry. The influence of the common carotid artery (CCA) ligation, filament morphology, and the pterygopalatine artery (PPA) ligation on lesion volume and neurological severity score 24 hours after reperfusion were assessed. RESULTThe use of filaments with appropriate length of coating and ligation of the PPA while maintaining perfusion of the CCA prevented the development of infarcts in the PCA area, resulted in pure MCA infarcts (68.3{+/-}14.5mm3, 26.1{+/-}3.6% of the hemisphere with Swansons correction) and reduced the variability of infarct volumes by more than half to 13.9% of the standard deviation divided by mean. CONCLUSIONSUsing improved surgical methods with suitable filaments to induce MCA occlusion in mice, we were able to produce PCA area-unaffected reproducible infarcts exclusively in the MCA area with reduced variability (PURE-MCAo). Our results may thus help to increase the reproducibility of the fMCAo model and reduce the number of animals required in preclinical stroke research.

neuroscience↗

bioRxiv2023

ATF7IP2/MCAF2 directs H3K9 methylation and meiotic gene regulation in the male germline

H3K9 tri-methylation (H3K9me3) plays emerging roles in gene regulation, beyond its accumulation on pericentric constitutive heterochromatin. It remains a mystery why and how H3K9me3 undergoes dynamic regulation in male meiosis. Here, we identify a novel, critical regulator of H3K9 methylation and spermatogenic heterochromatin organization: the germline-specific protein ATF7IP2 (MCAF2). We show that, in male meiosis, ATF7IP2 amasses on autosomal and X pericentric heterochromatin, spreads through the entirety of the sex chromosomes, and accumulates on thousands of autosomal promoters and retrotransposon loci. On the sex chromosomes, which undergo meiotic sex chromosome inactivation (MSCI), the DNA damage response pathway recruits ATF7IP2 to X pericentric heterochromatin, where it facilitates the recruitment of SETDB1, a histone methyltransferase that catalyzes H3K9me3. In the absence of ATF7IP2, male germ cells are arrested in meiotic prophase I. Analyses of ATF7IP2-deficient meiosis reveal the proteins essential roles in the maintenance of MSCI, suppression of retrotransposons, and global upregulation of autosomal genes. We propose that ATF7IP2 is a downstream effector of the DDR pathway in meiosis that coordinates the organization of heterochromatin and gene regulation through the spatial regulation of SETDB1-mediated H3K9me3 deposition.

developmental biology↗

bioRxiv2023

Atypical heat shock transcription factor HSF5 is critical for male meiotic prophase under non-stress conditions

Meiotic prophase progression is differently regulated in males and females. In males, pachytene transition during meiotic prophase is accompanied by robust alteration in gene expression. However, how gene expression is regulated differently to ensure meiotic prophase completion in males remains elusive. Herein, we identified HSF5 as a male germ cell-specific heat shock transcription factor (HSF) for meiotic prophase progression. Genetic analyses and single-cell RNA-sequencing demonstrated that HSF5 is essential for progression beyond the pachytene stage under non-stress conditions rather than heat stress. Chromatin binding analysis in vivo and DNA-binding assays in vitro suggested that HSF5 binds to promoters in a subset of genes associated with chromatin organization. HSF5 recognizes a DNA motif different from typical heat shock elements recognized by other canonical HSFs. This study suggests that HSF5 is an atypical HSF that enforces the gene expression program for pachytene transition during meiotic prophase in males.

developmental biology↗

bioRxiv2021

The rice OsERF101 transcription factor regulates the NLR Xa1-mediated perception of TAL effectors and Xa1-mediated immunity

O_LIPlant nucleotide-binding leucine-rich repeat receptors (NLRs) initiate immune responses and the hypersensitive response by recognizing pathogen effectors. Xa1 encodes an NLR with an N-terminal BED domain, and recognizes transcription activator-like (TAL) effectors of Xanthomonas oryzae pv. oryzae (Xoo). The molecular mechanisms controlling the recognition of TAL effectors by Xa1 and the subsequent induction of immunity remain poorly understood. C_LIO_LIXa1 interacts in the nucleus with two TAL effectors via the BED domain. We identified the AP2/ERF-type transcription factor OsERF101/OsRAP2.6 as an interactor with Xa1, and found that it also interacts with the TAL effectors. Overexpression of OsERF101 exhibited an enhanced resistance to an incompatible Xoo strain only in the presence of Xa1, indicating that OsERF101 functions as a positive regulator of Xa1-mediated immunity. Unexpectedly, oserf101 mutants also showed enhanced Xa1-dependent resistance, but in a different manner from the overexpressing plants. This result revealed an additional Xa1-mediated immune pathway that is negatively regulated by OsERF101. Furthermore, OsERF101 directly interacted with the TAL effectors. C_LIO_LIOur results show that OsERF101 regulates the recognition of TAL effectors and the Xa1-mediated activation of the immune response. These data provide new insights into the molecular mechanism of NLR-mediated immunity in plants. C_LI

plant biology↗