Biology subjects

Feng, D.

Publications and source records attributed to Feng, D..

At least 19 recordsLinked to original sources

bioRxiv2026

The Onset of Breathing at Birth Switches Off Regenerative Capacity

Many mammalian tissues, including the central nervous system (CNS), lose regenerative capacity in adulthood, but when and how this transition is initiated remains unclear. By monitoring mTOR activity, an evolutionarily conserved readout of cellular growth and intrinsic regenerative potential, we found that mTOR signaling declines within minutes after birth in murine CNS neurons and in other organs such as the heart and skeletal muscles. Focusing on retinal ganglion cells (RGCs), we show that this abrupt decline is driven primarily by CO2 exhalation following the onset of breathing after birth. Using real-time in vivo imaging of RGCs during the earliest postnatal period, we further discover that breathing-induced CO2 exhalation rapidly causes intracellular alkalinization, activates an AMPK-family kinase, SNF-related kinase (SNRK), which in turn suppresses mTOR signaling. Importantly, this process remains reversible in adulthood: CO2 inhalation or Snrk deletion restores mTOR activity in RGCs and enables robust regeneration of their axons after injury. Similar CO2 inhalation-induced mTOR reactivation occurs across multiple cell types and organs, suggesting a general mechanism by which breathing-associated CO2/pH signaling shapes the regenerative capacity across life and pointing to a potential CO2-based strategy for promoting tissue regeneration.

neuroscience↗

bioRxiv2026

FMR1 Perturbations Disrupt Sensory Corticostriatal Circuits and Mitochondrial Energetics in Fragile X Syndrome

Fragile X syndrome (FXS) features sensory, social, and cognitive dysfunction, but how fragile X messenger ribonucleoprotein 1 (FMR1) deficiency disrupts sensory corticostriatal circuitry and whether distinct FMR1 perturbations converge on shared cellular mechanisms remain incompletely defined. We integrated diffusion magnetic resonance imaging (MRI) in individuals with FXS and Fmr1-knockout (KO) mice with projection-defined calcium recording, bidirectional optogenetic manipulation, primary somatosensory cortex (S1)-targeted expression of three patient-associated FMR1 missense variants, regional transcriptomics, and optical manipulation of mitochondrial energetic state. Individuals with FXS exhibited disrupted sensory corticostriatal structural connectivity, with mouse analyses further localizing abnormalities to S1-striatal pathways. Fmr1 deficiency impaired recruitment of the S1-dorsomedial striatum (DMS) pathway during social interaction, and projection-specific manipulation demonstrated that S1-DMS dysfunction contributes to social and cognitive deficits. Three patient-associated FMR1 variants, I234T, E329G, and R586Q, converged on social behavioral deficits, which were improved by S1-DMS activation. Transcriptomic analyses revealed convergent oxidative phosphorylation (OXPHOS)-related mitochondrial remodeling in S1 and DMS across Fmr1 deficiency and variant models. Systemic treatment with the mitochondria-targeted antioxidant MitoQ attenuated novel-location-recognition, social-interaction, and social-novelty deficits in Fmr1-KO mice. Moreover, mt-EcGAPR-based Mitochondrial OptoEnergetic Activation (Mito-OptoEA) enhanced mitochondrial energetic state in S1 and improved novel-location recognition and sociability in both Fmr1-KO and I234T models. Together, these findings define a sensory corticostriatal circuit-metabolic framework for FXS in which S1-DMS circuit dysfunction and convergent mitochondrial energetic abnormalities contribute to FMR1-related social and cognitive phenotypes.

neuroscience↗

bioRxiv2026

3D-MAESTRO: A scalable, modular, portable pipeline for automated processing of large-scale volumetric brain microscopy data

Light microscopy is routinely used to explore the cellular and molecular structure of tissues, but the scale and complexity of data remains a bottleneck for discovery. We introduce 3D-MAESTRO (3D-Microscopy Automation and Execution with Scalable Tools, Rendering, and Orchestration): an automated image processing workflow for large-scale microscopy data, built for scalable execution across cloud and local computing environments. 3D-MAESTRO orchestrates denoising, stitching of image tiles, atlas registration, and segmentation. Its modular architecture permits the integration and benchmarking of new packages, ensuring that performance evolves as more efficient or accurate algorithms emerge. We introduce a new 3D image template for automated registration of mouse brains cleared with aqueous reagents and an efficient method for detection of fluorescent cells. We apply 3D-MAESTRO to lightsheet images of whole mouse brains in the context of diverse anatomical and functional experiments, illustrating high-throughput and reproducible mapping of microscopic structures across the brain.

neuroscience↗

bioRxiv2026

OmniTCR: a foundation model unifying T cell receptor recognition prediction and conditional sequence generation

T cell receptor (TCR) recognition prediction and receptor generation are traditionally modelled separately, leaving vast TCR sequence collections disconnected from smaller TCR-peptide-MHC datasets. Here we present OmniTCR, a 113-million-parameter autoregressive foundation model pretrained on 328 million formatted human immune-sequence records. Sequence-type tokens and complementary component orders enable joint learning from individual TCR chains and partial or complete TCR-pMHC associations. On unseen epitopes, OmniTCR achieved AUPRCs of 0.7009 for peptide-TCR{beta}; recognition and 0.8235 for TCR-pMHC interaction prediction, exceeding the strongest evaluated comparators by 0.3396 and 0.3451, respectively. It distinguishes cancer from healthy repertoires across 11 independent pan-cancer cohorts (mean AUROC, 0.9436). The model achieved the highest sequence recovery on internal and external generation benchmarks. Structural modelling supported the plausibility of selected pMHC-conditioned CDR3{beta}; candidates. OmniTCR bridges heterogeneous immune sequence data, providing a foundation for computational immunology and receptor design.

bioinformatics↗

bioRxiv2026

Charting the small-molecule universe from mass spectra with neuro-symbolic AI

Mass spectrometry (MS) has revealed millions of small organic molecules across organisms, yet most remain uncharacterized, limiting progress in biology and medicine. Despite computational advances, MS workflows rely heavily on expert input and reference libraries that cover only a fraction of known chemical space. Here, we introduce AIMe (AI Molecule Explorer), a multi-agent neuro-symbolic AI framework that transforms the interpretation of unknown spectra into an omics-scale exploration across the known structural space, providing chemically interpretable annotations. At its core, AIMe combines chemical reasoning with structure- informed learning to predict MS2 spectra by modeling fragmentation as a sequence of actions, outperforming existing methods. AIMe dynamically constructs fragmentation pathways by assigning likelihoods to individual fragmentation actions, linking spectral peaks to explicit fragment molecular formulas and structures. At scale, AIMe predicted MS2 spectra for over 100 million small organic molecules in PubChem and organized them into MS2KOSMOS, a substructure-informed community resource comprising over 800 million predicted spectra that expands the searchable small-molecule universe by roughly three orders of magnitude relative to experimental libraries. Analogous to sequence homology-based searches in genomics and proteomics, AIMe maps unknown spectra to molecular neighborhoods in MS2KOSMOS. Exact- formula indexing enables ranked retrieval of candidates and related structures, with peak-level structural and fragmentation-pathway annotations. Applied to mouse microbiota-dependent metabolites, AIMe enabled putative annotation of knowns and guided structure elucidation of unknowns, revealing previously unreported types of microbiota-dependent polyamines that also occur in humans. At repository scale, AIMe enabled putative annotation of roughly a third of 7 million spectral clusters representing most of the unknowns in the GNPS database. By extending MS2 annotation beyond curated-library matching to interpretable search across the known small-molecule universe, AIMe accelerates discovery and large-scale exploration of small molecules across biomedicine, agriculture, and ecology.

biochemistry↗

bioRxiv2025

Distinct Programs Drive Organ Development and Regeneration

A prevailing paradigm in stem cell and developmental biology posits that regeneration is a recapitulation of embryonic programs, yet the molecular logic underlying these processes remains unclear. Here, using the mammary gland as a model, we identified a novel intermediate cell state essential for basal-to-luminal differentiation during regeneration. The transcription factor Pou2f3 is a master regulator of this state, critical for regeneration but dispensable for organ physiology. Mechanistically, we found that POU2F3 functions as a transcriptional repressor that drives lineage conversion by restricting chromatin accessibility of basal genes rather than by directly activating luminal programs. Furthermore, we uncovered a POU2F3-TNF feedback loop that maintains epithelial homeostasis, where TNF signaling suppresses Pou2f3 to inhibit differentiation, and loss of luminal cells relieves this repression to trigger regeneration. Strikingly, cell fate is governed by signaling pathways essential for cell-cell communication during development, but by Pou2f3-dependent chromatin remodeling for rapid response and injury repair. We validate the regeneration-specific requirement for Pou2f3 in mammary gland and further demonstrate analogous regeneration defects in prostate and pancreas. These findings challenge the long-held assumption that regeneration recapitulates development and establish a framework for understanding context-specific mechanisms of tissue plasticity, with implications for regenerative medicine and epithelial biology.

cell biology↗

bioRxiv2025

Efficient and reproducible pipelines for spike sorting large-scale electrophysiology data

The scale of in vivo electrophysiology has expanded in recent years, with simultaneous recordings across thousands of electrodes now becoming routine. These advances have enabled a wide range of discoveries, but they also impose substantial computational demands. Spike sorting, the procedure that extracts spikes from extracellular voltage measurements, remains a major bottleneck: a dataset collected in a few hours can take days to spike sort on a single machine, and the field lacks rigorous validation of the many spike sorting algorithms and preprocessing steps that are in use. Advancing the speed and accuracy of spike sorting is essential to fully realize the potential of large-scale electrophysiology. Here, we present an end-to-end spike sorting pipeline that leverages parallelization to scale to large datasets. The same workflow can run reproducibly on individual workstations, high-performance computing clusters, or cloud environments, with computing resources tailored to each processing step to reduce costs and execution times. In addition, we introduce a benchmarking pipeline, also optimized for parallel processing, that enables systematic comparison of multiple sorting pipelines. Using this framework, we show that Kilosort4, a widely used spike sorting algorithm, outperforms Kilosort2.5 (Pachitariu et al. 2024). We also show that 7x lossy compression, which substantially reduces the cost of data storage, has minimal impact on spike sorting performance. Together, these pipelines address the urgent need for scalable and transparent spike sorting of electrophysiology data, preparing the field for the coming flood of multi-thousand-channel experiments.

neuroscience↗

bioRxiv2025

Enabling regiospecific di-halogenation in one-pot reactions using an engineered single-component flavin-dependent tryptophan halogenase

Installation of halogen atoms to organic compounds through flavin-dependent halogenases (FDHs) catalyzed reactions has emerged as an attractive strategy in organic biochemistry and synthetic biology. A cyanobacterial FDH termed AetF that contains FDH and flavin reductase in the same polypeptide chain could become a useful tool enzyme as the need of including a reductase is omitted. Notably, AetF exploits a unique substrate-interaction network to consecutively brominate tryptophan (Trp) on C5 and then C7 to generate di-brominated Trp. In this study, structure-based engineering is used to eliminate the C7 halogenation capacity of AetF. The resulting variant AetF-AIF that contains three residue substitutions mainly catalyzes bromination or iodination on C5 of Trp. We also show that combining AetF-AIF in tandem with the wild type AetF enables the generation of heterogeneously halogenated Trp in a one-pot reaction. The final products, 5-Br-7-I-Trp or 5-I-7-Br-Trp depending on the addition order of halides, account for a high ratio in the reaction mixture without additional purification step. These results display the potentials of single-component Trp-FDHs in a wide application of the halogenation reactions. Table of Content Graphic O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=105 SRC="FIGDIR/small/687318v1_ufig1.gif" ALT="Figure 1"> View larger version (28K): org.highwire.dtl.DTLVardef@101bd34org.highwire.dtl.DTLVardef@12ca797org.highwire.dtl.DTLVardef@aee280org.highwire.dtl.DTLVardef@1984907_HPS_FORMAT_FIGEXP M_FIG A single-component flavin-dependent halogenase that catalyzes di-bromination of tryptophan was engineered as a mono-halogenase and coupled in a one-pot reaction for the synthesis of heterogeneously halogenated tryptophan. C_FIG

molecular biology↗

bioRxiv2025

In-depth proteomic profiling of the extracellular matrix of pancreatic ductal adenocarcinomas identifies signatures correlating with lymphocyte infiltration

The extracellular matrix (ECM) is a complex assembly of proteins surrounding cells. It is a critical component of the tumor microenvironment that plays an active role in tumor progression and modulation of tumor response to treatment. Pancreatic ductal adenocarcinoma (PDAC) is a cancer type characterized by one of the worst prognoses, as it is often diagnosed at an advanced stage. It is also characterized by a very dense ECM, which hinders efficient drug delivery. In addition, PDACs are considered "cold" tumors as they fail to elicit a strong immune response, challenging the use of immunotherapy for PDAC cancer patients. Yet, the interplay between the ECM and immune cells within the PDAC tumor microenvironment remains poorly understood. Here, we employed ECM-focused proteomics to profile the ECM compositions of PDAC mouse models characterized by different levels of CD8+ T-cell infiltration. We found that CD8lo, or "cold" tumors, and CD8hi, or "hot" tumors, exhibited different ECM compositions. Interrogation of publicly available single-cell RNA-sequencing datasets of human PDACs further revealed that the ECM proteins distinguishing hot and cold PDACs are secreted by multiple stromal cell populations, including cancer-associated fibroblasts, stellate cells, and macrophages. Last, we found that the expression of a subset of the genes encoding ECM proteins characteristic of the CD8lo phenotype correlated with CD8+ T-cell infiltration in human PDAC samples and patient survival. This study paves the way for the development of ECM-modulating interventions to enhance immune cell infiltration and responsiveness to immunotherapy. SignificanceWe report the identification of ECM protein signatures correlating with the level of CD8+ lymphocyte infiltration in murine models of pancreatic ductal adenocarcinomas and human samples. This work paves the way for the development of ECM-modulating therapeutic strategies to enhance lymphocyte infiltration and, hence, the efficacy of immunotherapies.

cancer biology↗

bioRxiv2025

Modular Inter-brain Synchrony Network Associated with Social Difficulty in Autism Spectrum Disorder: a Graph Neural Network-Driven Hyperscanning Study

Understanding social difficulties in Autism Spectrum Disorder (ASD) remains challenging due to its neurobiological heterogeneity and the limited ecological validity of conventional neuroimaging methods in capturing dynamic social interactions. Hyperscanning analysis based on functional near-infrared spectroscopy (fNIRS), which measures inter-brain synchrony (IBS) during dyadic interaction, offers a novel avenue to address these challenges. However, prior studies on ASD have reported inconsistent findings, primarily focusing on intra-regional synchronization while overlooking cross-regional network dynamics. To bridge this gap, we proposed an interpretable graph neural network (GNN) model to systematically identify ASD-specific IBS modular network between child-caregiver dyads during naturalistic cooperative puzzle-solving and free-talking tasks. We identified distinctive key IBS sub-networks for the cooperative puzzle-solving task and free-talking task, with the frontal eye field (FEF) of caregivers, the dorsal lateral prefrontal cortex (DLPFC) and the motor region of children highlighted. Furthermore, the key IBS sub-networks were found to be able to predict multiple domains of the core ASD symptoms. By integrating hyperscanning with GNN-driven analysis, this work uncovers task-dependent inter-brain neural mechanisms underlying social difficulties in ASD. These findings advance the field by proposing a data-driven framework to identify IBS biomarkers tied to clinical profiles, paving the way for personalized interventions that integrate computational neuroscience with clinical practice.

neuroscience↗

bioRxiv2025

A Long-Term Drosophila Body Mass Measurement System

We propose a novel system for long-term monitoring of Drosophila body mass. In this approach, a cantilever beam is placed inside a Drosophila culture vial and undergoes random vibrations induced by Drosophila landings. An infrared camera connected to a microcomputer records these vibrations. Image processing techniques then extract the beams vibration signals from the video recordings. Applying the Euler-Bernoulli beam theory, we calculate the Drosophila body mass. As a demonstration, we used this system to measure body mass variations in wild-type Drosophila over 14 days. Summary statementThis study presents a system that enables long-term monitoring of fruit fly body mass, offering new opportunities to study relationships between genes, circadian rhythms, phenotype, and long-term mass variation.

animal behavior and cognition↗

bioRxiv2025

HortaCloud: An Open and Collaborative Platform for Whole Brain Neuronal Reconstructions

HortaCloud is a cloud-based, open-source platform designed to facilitate the collaborative reconstruction of long-range projection neurons from whole-brain light microscopy data. By providing virtual environments directly within the cloud, it eliminates the need for costly and time-consuming data downloads, allowing researchers to work efficiently with terabyte-scale volumetric datasets. Standardization of computational resources in the cloud make deployment easier and more predictable. The pay-as-you-go cloud model reduces adoption barriers by eliminating upfront investments in expensive hardware. Finally, HortaClouds decentralized architecture enables global collaboration between researchers and between institutions.

neuroscience↗

bioRxiv2025

Anti- and Pro-tumor Functions of Distinct Regulatory T Cell Subsets in Colorectal Cancer

Regulatory T (Treg) cells contribute to solid organ cancer progression, except in colorectal cancer (CRC) despite being abundantly present. Here, we demonstrate that two distinct tumoral IL-10 and IL-10- Treg cell subsets exert opposing functions by counteracting and promoting CRC tumor growth, respectively. The tumor restraining activity of IL-10 Treg cells was mediated by their suppression of effector CD4 T cell production of IL-17, which directly stimulates CRC tumor cell proliferation. Consistently, IL-10- Treg cells were more abundant in both mouse and human CRC tumors than in tumor-adjacent normal tissues, whereas IL-10+ Treg cells exhibited the opposite distribution. Furthermore, relative abundance of IL-10 and IL-10- Treg cells correlated with better and worse disease prognoses in human CRC, respectively. This functional dichotomy between Treg cell subsets provides a rationale for therapeutic strategies to selectively target pro-tumoral Treg cells while preserving their anti-tumoral counterparts across barrier tissue cancers that harbor both subsets.

immunology↗

bioRxiv2025

Sequencing-based Spatial Transcriptomics with scRNA-seq Sensitivity

The advent of spatial transcriptomics has dramatically expanded our ability to study the vast network of cell-cell interactions at the molecular level in tissue. Among current methods, sequencing-based approaches have great potential in discovering because of its unbiased capture. In the last couple of years, the spatial resolution for the capture addresses has been significantly improved from 100 m to <1 m, well below the size of a mammalian cell. However, the capture efficiency has always been a pain point, [~]one order of magnitude lower than that of single cell RNA sequencing (scRNA-seq). The low capture efficiency limits the depth and breadth of its applications in the study of complex biological systems and diseases. Here, we introduce Salus Spatial transcriptomic system (Salus-STS), which provides [~]1 m capture resolution and a capture efficiency [~]1 order of magnitude higher than other current methods. Analysis with sub-cellular resolution becomes practical for sequencing-based spatial transcriptomics.

molecular biology↗

bioRxiv2025

Peripheral alcohol metabolism dictates ethanol consumption and drinking microstructure in mice

BackgroundEthanol metabolism is intimately linked with the physiological and behavioral aspects of ethanol consumption. Ethanol is mainly oxidized by alcohol dehydrogenase (ADH) to acetaldehyde and further to acetate via aldehyde dehydrogenases (ALDHs). Understanding how ethanol and its metabolites work together to initiate and drive continued ethanol consumption is crucial for identifying interventions for alcohol use disorder (AUD). Therefore, the goal of our study was to determine how ADH1, which is mainly peripherally-expressed and metabolizes >90% of ingested ethanol, modulates ethanol metabolite distribution and downstream behaviors. MethodsEthanol consumption in drinking-in-the-dark (DID) and two-bottle choice (2BC) drinking paradigms, ethanol metabolite concentrations, and lickometry were assessed after ADH1 inhibition and/or in Adh1-knockout (Adh1 KO) mice. ResultsWe found that Adh1 KO mice of both sexes exhibited decreased ethanol consumption and preference compared to wild-type (WT) mice in DID and 2BC. ADH1 inhibitor fomepizole (4-MP) also significantly decreased normal and sweetened ethanol consumption in DID studies. Measurement of ethanol and its metabolites revealed that ethanol was increased at 1h but not 15 min, peripheral acetaldehyde was slightly decreased at both time points, and ethanol-induced increases in acetate were abolished after ethanol administration in Adh1 KO mice compared to controls. Similarly, ethanol accumulation as a function of consumption was 2-fold higher in Adh1 KO or 4-MP treated mice compared to controls. We then used lickometry to determine how this perturbation in ethanol metabolism affects drinking microstructure. Adh1 KO mice consume most of their ethanol in the first 30 min like WT mice but display altered temporal shifts in drinking behaviors and do not form normal bout structures, resulting in lower ethanol consumption. ConclusionsOur study demonstrates that ADH1-mediated ethanol metabolism is a key determinant of ethanol consumption, highlighting a fundamental knowledge gap around how ethanol and its metabolites drive ethanol consumption.

neuroscience↗

bioRxiv2024

TCPGdb: A comprehensive T cell perturbation genomics database for identification of critical T cell regulators

CAR-T therapies utilizing T cells engineered with chimeric antigen receptors (CARs) have revolutionized the treatment of hematologic and immune-related malignancies. However, the anti-tumor capability of engineered CAR-T cells are often not persistent, which greatly dampen its clinical efficacy. To address this limitation, massive parallel genetic screens were widely used to identified novel targets and regulators that enhance T cell anti-tumor capability and persistence in tumor microenvironment. We hypothesized that by combining the pooled screen data from multiple independent genetic screens we could provide a systematic, comprehensive, and robust analysis of the effect of gene perturbation on T-cell based immunotherapies. After collecting data from previously published T cell screens, including CRISPR-based and ORF-based screens, through Gene Expression Omnibus (GEO), we reprocessed the gene hits summary and conducting a pathway enrichment analysis. A T cell screen perturbation score (TPS) was employed to quantifies the impact of the gene on T cell function. Additionally, gene expression data (both bulk RNA level and single-cell RNA level) from autoimmune disease and T cell-derived cancers were analyzed to pick up gene perturbations that potentially augment T cell proliferation. We integrated all data and analysis on 27 T-cells screens into our state-of-the-art T cell perturbation genomics database (TCPGdb), which is made easily accessible through our webserver and allows users to interactively explore the impact of query genes on T cell function based on prior screen data and our TPS scoring module. TCPGdb is publicly accessible at http://tcpgdb.sidichenlab.org/.

bioinformatics↗

bioRxiv2024

Basal EFNA3 Facilitates Luminal-Driving Mammary Epithelial Migration and Cancer Metastasis by Promoting OXPHOS

Collective migration is essential for tissue development, repair, and cancer metastasis. However, the mechanisms by which different cell types coordinate to facilitate cohesive movement remain poorly understood. Here, we report distinct migratory behaviors in luminal and basal cells during collective migration of the mammary epithelium. We show that luminal cells are leaders of directional migration. However, basal cells also play a central role in promoting collective migration by producing EFNA3, which targets luminal cells and enhances oxidative phosphorylation and energy production. This promotion of epithelial migration and branching morphogenesis is confirmed in vivo. Importantly, EFNA3 is a prognostic marker for breast cancer metastasis, with high expression correlating with increased metastasis. Conversely, inhibition of EFNA3 significantly slows metastasis. These findings reveal a novel mechanism of epithelial migration that is critical for organogenesis and identify EFNA3 as a potential therapeutic target for cancer metastasis.

developmental biology↗

bioRxiv2024

Informing biologically relevant signal from spatial transcriptomic data

Visium is a spatial sequencing technology that utilizes messenger RNA (mRNA) to spatially map gene expression within tissues. Despite its potential, research utilizing deconvolution tools and exploring microenvironment dynamics remains challenging. We address this gap by benchmarking deconvolution tools across diverse biological contexts, identifying optimal methodologies. Subsequently, we introduce a novel pipeline integrating advanced deconvolution techniques and novel tools for reproducible tissue microenvironment analysis. Through this approach, we uncover intricate immune aggregate biology, highlighting the power of our methodology in unraveling complex biological phenomena.

bioinformatics↗