Biology subjects

Yoshida, N.

Publications and source records attributed to Yoshida, N..

At least 19 recordsLinked to original sources

bioRxiv2026

A chromatin relay from AIRE to ETS transcription factors sustains peripheral antigen expression in the thymic mimetic cells to ensure central tolerance.

Medullary thymic epithelial cells (mTECs) establish central tolerance by expressing a diverse repertoire of peripheral tissue-specific antigens (TSAs). This diversity is regulated not only by the transcriptional regulator AIRE, but also by lineage-defining factors in tissue-mimetic, AIRE-negative descendants of Aire mTECs (post-Aire mimetic TECs). However, whether and how prior AIRE activity contributes to TSA expression in post-AIRE mimetic TECs remains unclear. Here, we identify the ETS transcription factors, EHF and ELF3, as key regulators that sustain AIRE-primed gene expression in these cells. EHF and ELF3 preferentially bind distal genomic regions which are rendered accessible in advance by AIRE. Combined loss of EHF and ELF3 disrupts expression of AIRE-regulated genes expressed in mimetic TECs, leading to tissue-selective autoimmunity. Our findings reveal a relay mechanism in which AIRE primes chromatin for ETS factors to maintain TSA expression in post-AIRE mimetic TECs, thereby safeguarding self-tolerance.

immunology↗

bioRxiv2026

Competition-induced metabolic switching governs the interaction network in plant microbiome assembly

Plants are colonized by characteristic microbiomes that help maintain plant health and function. However, the mechanisms that make these communities assemble in a consistent, repeatable manner remain poorly understood. Here, we show how microbial metabolic strategies for host-derived substrates drive plant microbiome assembly, using a six-member synthetic community that captures the taxonomic diversity of natural duckweed microbiome. Contrary to expectations of metabolic niche partitioning, five of six strains adopt similar metabolic states when colonizing the host alone, consistent with the preferential use of a shared substrate, acetaldehyde. In contrast, during competition with specific community members, these strains consistently switch toward alternative substrates, including sugars and aromatics. Strikingly, this metabolic switching accompanies all negative interspecies interactions observed in the community, indicating that a single class of metabolic response dominates the inhibitory interaction network organizing community structure. By comparison, although metabolite cross-feeding is widespread, its contribution to facilitative interactions depends on the background of resource competition. Together, these findings highlight competition-induced metabolic switching as a primary driver of microbial interaction networks, and provide a cross-scale account of how microbial metabolic strategies underpin the reproducible assembly of plant microbiomes.

microbiology↗

bioRxiv2025

Mitochondria-derived PEP licenses glycerolipid synthesis via SLC25A35

Mitochondria provide a variety of metabolites, in addition to ATP, to meet cell-specific needs. One such metabolite is phosphoenolpyruvate (PEP), which contains a higher-energy phosphate bond than ATP and has diverse biological functions. However, how mitochondria-generated PEP is delivered to the cytosol and fulfills cell-specific requirements remains elusive. Here, we show that SLC25A35 regulates mitochondrial PEP efflux and glyceroneogenesis in lipogenic cells that utilize the pyruvate-to-PEP bypass. Reconstitution and structural studies demonstrated PEP transport by SLC25A35 in a pH gradient-dependent manner. Loss of SLC25A35 in adipocytes impaired the conversion of mitochondrial PEP into glycerol-3-phosphate, thereby reducing glycerolipid synthesis. Significantly, hepatic inhibition of SLC25A35 in obese mice alleviated steatosis and improved systemic glucose homeostasis. Together, these results suggest that mitochondria facilitate glycerolipid synthesis by providing PEP via SLC25A35, offering lipogenic mitochondria as a target to limit glycerolipid synthesis, a pivotal step in the pathogenesis of hepatic steatosis and Type 2 diabetes.

cell biology↗

bioRxiv2025

Sequence-encoded conformational biases shape self-assembly modes of intrinsically disordered proteins

Self-assembly of intrinsically disordered proteins (IDPs) underlies cellular functions and disease pathogenesis. This process is mediated by two intermolecular interaction modes: transient point-to-point contacts described by the sticker-and-spacer framework, and persistent surface-to-surface contacts proposed in the cross-{beta} hypothesis. We investigated the molecular basis of these modes in the context of conformational biases, defined as sequence-encoded structural preferences of local segments. Using a five-residue model, we generated lag-series IDPs from the TIA-1 prion-like domain by systematically modulating conformational biases while preserving amino acid composition. The lag-series IDPs demonstrated distinct condensate properties and varying capacities for amyloid fibril formation. Their structural analyses revealed that strongly biased regions preferentially adopt extended structures, including {beta}-strands, and the spacing between these regions influences metastable {beta}-sheet formation. Our findings demonstrate that local conformational biases shape interaction modes of IDPs, thereby linking sequence to condensate properties and amyloid fibril formation. Significance StatementProteins fold into three-dimensional structures defined by their amino acid sequences. In contrast, intrinsically disordered proteins (IDPs) lack stable structures, yet their sequences encode unique self-assembly behaviors, including phase separation and amyloid fibril formation. Can such behaviors be explained within structure-based frameworks? Using a five-residue model, we showed that IDP sequences encode local structural preferences, termed conformational biases, within short segments. These biases determine whether segments engage in transient point-to-point interactions driving phase separation or persistent surface-to-surface interactions leading to amyloid fibril formation. By bridging sequence, structure, and interaction modes, our work provides a comprehensive mechanism for self-assembly and conceptual tools for understanding IDP-related biological functions and disease mechanisms.

biophysics↗

bioRxiv2025

Chromatin organizer ASCL1 governs gene programs in thymic epithelial cells, defining immunological self

Immunological self-tolerance depends on medullary thymic epithelial cells (mTECs), which express a broad repertoire of self-antigens to support negative selection of autoreactive T cells and the development of regulatory T cells. Although the autoimmune regulator AIRE is essential for this process, additional factors are required to establish the full mTEC gene expression program. Because thymoma is frequently associated with autoimmunity, implicating defective thymic tolerance, we performed single-cell RNA sequencing of TECs from thymoma patients and identified the transcription factor ASCL1 as selectively downregulated in tumor mTECs. Deletion of Ascl1 in mouse TECs resulted in spontaneous autoimmunity without inducing thymic tumorigenesis. Transcriptomic and chromatin accessibility analyses revealed that ASCL1 influences mTEC gene expression programs and is associated with corresponding changes in chromatin accessibility. Genetic interaction analyses further suggested that ASCL1 activity in immature mTECs modulates the AIRE dependency of gene expression in mature mTECs. Together, these findings identify ASCL1 as a key regulator of mTEC function and central tolerance, providing insight into mechanisms that safeguard immune homeostasis and whose disruption may contribute to autoimmune disease.

immunology↗

bioRxiv2025

Breast cancer identity is defined by specialized enhancer sets via lysine deacetylation

Breast cancer subtypes are defined by distinct transcriptional programs, yet the epigenetic mechanisms underlying subtype-specific gene regulation remain unclear. Enhancers, key regulators of gene expression and cell identity, are well positioned to define breast cancer subtypes. Here, we identify a previously unrecognized class of enhancers, termed hypoacetylation-defined (HD) enhancers, that regulate cancer-related genes in a luminal breast cancer cell line. HD enhancers are defined by RNA polymerase II dissociation upon lysine deacetylase inhibition, and bidirectional eRNA transcription. They are distinct from super-enhancers, require a specific Mediator subunit for gene-specific transcription, and form extensive chromatin interactions suggestive of a hub-like architecture. Analyses of clinical datasets further identified a subset of HD enhancers, termed HD cluster 1 enhancers, which classify patients into breast cancer subtypes and are associated with expression quantitative trait loci linked to subtype-specific gene expression. This study identifies the lysine deacetylation-regulated cell identity enhancers, which are potential therapeutic targets.

cancer biology↗

bioRxiv2024

Synthesising robot behaviour through reinforcement learning for homeostasis

Homeostasis is a fundamental property for the survival of animals. Computational reinforcement learning provides a theoretically sound framework for learning autonomous agents. However, the definition of a unified motivational signal (i.e., reward) for integrated survival behaviours has been largely underexplored. Here, we present a novel neuroscience-inspired algorithm for synthesising robot survival behaviour without the need for complicated reward design and external feedback. Our agent, the Embodied Neural Homeostat, was trained solely with feedback generated by its internal physical state and optimised its behaviour to stabilise these internal states: homeostasis. To demonstrate the effectiveness of our concept, we trained the agent in a simulated mechano-thermal environment and tested it in a real robot. We observed the synthesis of integrated behaviours, including walking, navigating to food, resting to cool down the motors, and shivering to warm up the motors, through the joint optimisation for thermal and energy homeostasis. The Embodied Neural Homeostat successfully achieved homeostasis-based integrated behaviour synthesis, which has not previously been accomplished at the motor control level. This demonstrates that homeostasis can be a motivating principle for integrated behaviour generation in robots and can also elucidate the behavioural principles of living organisms.

neuroscience↗

bioRxiv2022

TLR7/8 stress response drives histiocytosis in SLC29A3 disorders

SLC29A3, also known as ENT3, is a lysosomal transmembrane protein that transports nucleosides from the lysosomes to the cytoplasm1. Loss-of-function mutations in SLC29A3 cause lysosomal nucleoside storage and histiocytosis: phagocyte accumulation in multiple organs2,3. However, little is known about the mechanism through which lysosomal nucleoside storage drives histiocytosis. Herein, histiocytosis in Slc29a3-/- mice was demonstrated to depend on TLR7, which senses a combination of nucleosides and oligoribonucleotides4,5. TLR7 responded to lysosomal nucleoside storage and enhanced proliferation of Ly6Chi CX3CR1low immature monocytes and their maturation into Ly6Clow phagocytes in Slc29a3-/- mice. Because accumulated nucleosides primarily originated from cell corpse phagocytosis, TLR7 in immature monocytes recognized nucleoside storage as lysosomal stress and increased phagocyte numbers. This non-inflammatory compensatory response is referred to as the TLR7 stress response where Syk, GSK3{beta}, {beta}-catenin, and mTORC1 serve as downstream signalling molecules. In SLC29A3 disorders, histiocytosis accompanies inflammation6,7. Nucleoside storage failed to induce pro-inflammatory cytokine production in Slc29a3-/- mice, but enhanced ssRNA-dependent pro-inflammatory cytokine production in Ly6Chi classical monocytes and peripheral macrophages, not proliferating immature monocytes. Patient-derived monocytes harbouring G208R SLC29A3 mutation showed higher survival and proliferation in the presence of M-CSF and produced larger amounts of IL-6 upon ssRNA stimulation than did those derived from healthy subjects. A TLR8 antagonist inhibited the survival/proliferation of patient-derived macrophages. These results demonstrated that TLR7/8 responses to lysosomal nucleoside stress drive SLC29A3 disorders.

immunology↗

bioRxiv2022

Binding of cardiolipin to the KcsA channel at the membrane outer leaflet allosterically opens the inner gate

Membrane proteins embedded in the membrane undergo changes in their actions under the influence of membrane lipids. Here, we present a novel type of lipid action on the potassium channel KcsA from Streptomyces lividans. Cardiolipin is present in various cellular membranes, including the host membrane of KcsA. Although the M0 domain, a nontransmembrane helix, is known to sense anionic lipids in the inner leaflet, we found that divalent anionic cardiolipin in the outer leaflet of the membrane interacts with positively charged residues, Arg64 and Arg89, on the extracellular side of the transmembrane domain. This binding propagates its action across the membrane toward the intracellular region of KcsA, thus, opening the inner gate. Such a long-range allosteric effect has not been found for channel-lipid interactions.

biophysics↗

bioRxiv2021

NKG2D signaling regulates IL-17A-producing γδT cells to promote cancer progression

{gamma}{delta}T cells are unconventional T cells particularly abundant in mucosal tissues that play an important role in tissue surveillance and homeostasis. {gamma}{delta}T cell activation is mediated by the T cell receptor composed of {gamma} and {delta} chains, as well as activating receptors for stress-induced ligands, such as NKG2D. Contrary to the well-established anti-tumor function of {gamma}{delta}T cells, recent studies have shown that {gamma}{delta}T cells can promote tumor development in certain contexts. However, the mechanisms leading to this diseasepromoting role remain poorly understood. Here, we show that mice lacking {gamma}{delta}T cells survive longer in a mouse model of intestinal cancer, further supporting their pro-tumoral role. In a surprising conceptual twist, we found that these pro-tumor {gamma}{delta}T cells are regulated by NKG2D signaling, a receptor normally associated with cancer cell killing. Germline deletion of Klrk1, the gene encoding NKG2D, reduced the frequency of {gamma}{delta}T cells in the tumor microenvironment and delayed tumor progression. We further show that blocking NKG2D reduces the capability of {gamma}{delta}T cells to produce IL-17A in the pre-metastatic lung and that co-culture of lung T cells with NKG2D ligand-expressing tumor cells specifically increases the frequency of {gamma}{delta}T cells. Together, these data support the hypothesis that in a tumor microenvironment where NKG2D ligands are constitutively expressed, {gamma}{delta}T cells accumulate in an NKG2D-dependent manner and drive tumor progression by secreting pro-inflammatory cytokines, such as IL-17A.

immunology↗

bioRxiv2021

Balanced release of neuropeptide FF and gonadotropin-releasing hormone 3 modulates male sexual behavior

Animals properly perform sexual behaviors by using multiple sensory cues. However, neural mechanisms integrating multiple sensory cues and regulating motivation for sexual behaviors remain unclear. Here, we focused on peptidergic neurons, terminal nerve gonadotropin-releasing hormone (TN-GnRH) neurons, which receive inputs from various sensory systems and co-express neuropeptide FF (NPFF) in addition to GnRH. Our behavioral analyses using knockout medaka of GnRH (gnrh3) and/or NPFF (npff) demonstrated that some sexual behavioral repertories were delayed, not disrupted, in gnrh3-/- and npff-/- males, while the double knockout showed normal behaviors. We also found anatomical evidence to show that both neuropeptides modulate the sexual behavior-controlling brain areas. Furthermore, we demonstrated that NPFF activates neurons in the preoptic area via indirect pathway, which is considered to induce the increase in the motivation for male sexual behaviors. Considering these results, we propose a novel mechanism by which balanced release of co-existing peptides is important for the neuromodulatory function of TN-GnRH neurons in the control of behavioral motivation. Our results may go a long way toward understanding the functional significance of peptidergic neuromodulation in response to external environments.

neuroscience↗

bioRxiv2021

Epileptic Spike Detection by Recurrent Neural Networks with Self-Attention Mechanism

ObjectiveBecause of the lack of highly skilled experts, automated technologies that support electroencephalogram (EEG)-based in epilepsy diagnosis are advancing. Deep convolutional neural network-based models have been used successfully for detecting epileptic spikes, one of the biomarkers, from EEG. However, a sizeable number of supervised EEG records are required for training. ApproachThis study introduces the Satelight model, which uses the self-attention (SA) mechanism. The model was trained using a clinical EEG dataset labeled by five specialists, including 16,008 epileptic spikes and 15,478 artifacts from 50 children. The SA mechanism is expected to reduce the number of parameters and efficiently extract features from a small amount of EEG data. To validate the effectiveness, we compared various spike detection approaches with the clinical EEG data. Main resultsThe experimental results showed that the proposed method detected epileptic spikes more effectively than other models (accuracy = 0.876 and false positive rate = 0.133). SignificanceThe proposed model had only one-tenth the number of parameters as the other effective model, despite having such a high detection performance. Further exploration of the hidden parameters revealed that the model automatically attended to the EEGs characteristic waveform locations of interest.

bioengineering↗

bioRxiv2020

Epileptic Spike Detection by Using a Linear-Phase Convolutional Neural Network

To cope with the lack of highly skilled professionals, machine learning with proper signal processing is key for establishing automated diagnostic-aid technologies with which to conduct epileptic electroencephalogram (EEG) testing. In particular, frequency filtering with the appropriate passbands is essential for enhancing the biomarkers--such as epileptic spike waves--that are noted in the EEG. This paper introduces a novel class of neural networks (NNs) that have a bank of linear-phase finite impulse response filters at the first layer as a preprocessor that can behave as bandpass filters that extract biomarkers without destroying waveforms because of a linear-phase condition. Besides, the parameters of the filters are also data-driven. The proposed NNs were trained with a large amount of clinical EEG data, including 15,833 epileptic spike waveforms recorded from 50 patients, and their labels were annotated by specialists. In the experiments, we compared three scenarios for the first layer: no preprocessing, discrete wavelet transform, and the proposed data-driven filters. The experimental results show that the trained data-driven filter bank with supervised learning behaves like multiple bandpass filters. In particular, the trained filter passed a frequency band of approximately 10-30 Hz. Moreover, the proposed method detected epileptic spikes, with the area under the receiver operating characteristic curve of 0.967 in the mean of 50 intersubject validations.

neuroscience↗

bioRxiv2020

Mouse fertilization triggers a conserved transcription program in one-cell embryos

Following fertilization, the new embryo reprograms parental genomes to begin transcription (embryonic genome activation, EGA). EGA is indispensable for development, but its dynamics, profile or when it initiates in vertebrates are unknown. We here characterize the onset of transcription in mouse one-cell embryos. Precise embryo staging eliminated noise to reveal a cascading program of de novo transcription initiating within six hours of fertilization. This immediate EGA (iEGA) utilized canonical promoters, produced spliced transcripts, was distinctive and predominantly driven by the maternal genome. Expression represented pathways not only associated with embryo development but with cancer. In human one-cell embryos, hundreds of genes were up-regulated days earlier than thought, with conservation to mouse iEGA. These findings provide a functional basis for epigenetic analysis in early-stage embryos and illuminate networks governing totipotency and other cell-fate transitions. One Sentence SummaryFertilization instates transcription in mouse and human one-cell embryos far sooner than thought and is programmed.

developmental biology↗

bioRxiv2019

Nucleosides drive histiocytosis in SLC29A3 disorders by activating TLR7

A lysosomal transmembrane protein SLC29A3 transports nucleosides from lysosomes to the cytoplasm. Loss-of-function mutations of the SLC29A3 gene cause lysosomal nucleoside storage in monocyte/macrophages, leading to their accumulation called histiocytosis in humans and mice. Little is known, however, about a mechanism behind nucleoside-dependent histiocytosis. TLR7, an innate immune sensors for single stranded RNA, bind and respond to nucleosides. We here show that they drive nucleoside-mediated histiocytosis. Patrolling monocyte/macrophages accumulate in the spleen of Slc29a3-/- mice but not Slc29a3-/- Tlr7-/- mice. Accumulated patrolling monocyte/macrophages stored nucleosides derived from cell corpse. TLR7 was recruited to phagosomes and activated as evidenced by TLR7-dependent phagosomal maturation. TLR7 induced hyper-responsiveness to M-CSF in Slc29a3-/- monocyte/macrophages. These results suggest that TLR7 drives histiocytosis in SLC29A3 disorders. One Sentence SummarySLC29A3 disorders are caused by activation of TLR7 with accumulated nucleosides in lysosomes.

immunology↗

bioRxiv2019

In vivo control of Toxoplasma gondii by zebrafish macrophages

Toxoplasma gondii is an obligate intracellular parasite capable of invading any nucleated cell. Three main clonal lineages (type I, II, III) exist and murine models have driven the understanding of general and strain-specific immune mechanisms underlying Toxoplasma infection. However, murine models are limited for studying parasite-leukocyte interactions in vivo, and discrepancies exist between cellular immune responses observed in mouse versus human cells. Here, we develop a zebrafish infection model to study the innate immune response to Toxoplasma in vivo. By infecting the zebrafish hindbrain ventricle, and using high-resolution microscopy techniques coupled with computer vision driven automated image analysis, we reveal that Toxoplasma invades and replicates inside a parasitophorous vacuole to which type I and III parasites recruit host cell mitochondria. We show that type II and III strains maintain a higher infectious burden than type I strains. To understand how parasites are being cleared in vivo, we analyzed Toxoplasma-macrophage interactions using time-lapse and correlative light and electron microscopy. Strikingly, macrophages are recruited to the infection site and play a key role in Toxoplasma control. These results highlight in vivo control of Toxoplasma by macrophages, and illuminate the possibility to exploit zebrafish for discoveries within the field of parasite immunity.

microbiology↗

bioRxiv2019

Mimicry embedding for advanced neuralnetwork training of 3D biomedical micrographs

The use of deep neural networks (DNNs) for analysis of complex biomedical images shows great promise but is hampered by a lack of large verified datasets for rapid network evolution. Here we present a novel \"mimicry embedding\" strategy for rapid application of neural network architecture-based analysis of biomedical imaging datasets. Embedding of a novel biological dataset, such that it mimics a verified dataset, enables efficient deep learning and seamless architecture switching. We apply this strategy across various microbiological phenotypes; from super-resolved viruses to in vivo parasitic infections. We demonstrate that mimicry embedding enables efficient and accurate analysis of three-dimensional microscopy datasets. The results suggest that transfer learning from pre-trained network data may be a powerful general strategy for analysis of heterogeneous biomedical imaging datasets.

bioinformatics↗

bioRxiv2019

De novo transcriptome assembly of the African bullfrog Pyxicephalus adspersus for molecular analysis of aestivation

The molecular mechanisms of aestivation, a state of dormancy that occurs under dry conditions at ordinary temperature, are not yet clarified. Here, we report the first de novo transcriptome assembly of the African bullfrog Pyxicephalus adspersus, which aestivates for 6-10 months during the hot and dry season. Polyadenylated RNA from tissues was sequenced to 75,320,390 paired-end reads, and the de novo assembly generated 101,682 transcripts. Of these, 100,093 transcripts had open reading frames encoding more than 25 amino acids. BLASTx analysis against the Uniprot Xenopus tropicalis protein database revealed 64,963 transcripts having little homology with an E value higher than 1E-5 and 8,147 transcripts having no homology, indicating that the African bullfrog has many novel genes that are absent in X. tropicalis. The other 28,570 transcripts had homology with an E value lower than 1E-5 for which molecular functions were estimated by gene ontology (GO) analysis and found to contain the aestivation-related genes conserved among other aestivating organisms, including the African lungfish. This study is the first to identify a comprehensive set of genes expressed in the African bullfrog, thus providing basic information for molecular level analysis of aestivation.

molecular biology↗