Biology subjects

Fang, M.

Publications and source records attributed to Fang, M..

At least 19 recordsLinked to original sources

bioRxiv2026

A KaiA Conformational Switch Links Day-Night Transitions to Clock Robustness

Biological clocks require mechanisms that enable transition between temporal phases and buffer against environmental fluctuations. In the cyanobacterial KaiABC-based circadian oscillator, KaiA has been regarded as a daytime activator that is passively sequestered at night. Here we define KaiA as a conformationally regulated activator and an active catalyst of the night-to-dawn transition. We identify a single linker helix, 5, that acts as a structural switch, mediating a domain-driven "tug-of-war" that sets the open-closed equilibrium in KaiA. Structure-guided mutations that shift this balance in either direction abolish stable rhythmicity both in vitro and in cells. Also, we show that substoichiometric KaiA dismantles nighttime KaiB-KaiC complexes through interactions with the KaiC C-terminal tails and likely initiates an autocatalytic feedback loop that resets the phosphorylation cycle at dawn. Furthermore, we present two independent kinetic models that offer distinct, deterministic explanations for the clock's low-temperature amplitude decline and failure at 20 {degrees} C.

biochemistry↗

bioRxiv2026

Machine learning classifies subcellular Golgi morphology from imaging flow cytometry during inflammatory activation: an exploratory study

Inflammatory activation remodels the Golgi apparatus to meet increased secretory demand, yet whether these structural changes carry computationally accessible information about immune cell state has not been explored. Here we show that machine learning applied to imaging flow cytometry images of the Golgi can distinguish lipopolysaccharide (LPS) treated from control cells. We trained models on Golgi images from human THP-1 monocytes before and 16 h after LPS stimulation in vitro, and from circulating myeloid cells of mice 24 h after intratracheal LPS or saline in vivo. In each dataset, the best-performing model was a deep learning model trained directly on images rather than a classifier using predefined morphometric features. Cross-domain performance varied with the model, transfer direction and use of labelled target data: several deep learning models trained on mouse images classified THP-1 cells above chance without changing model parameters. Grad-CAM highlighted regions within or adjacent to the GM130 fluorescence pattern. These findings indicate that GM130 images carry information associated with LPS treatment that supports classification within and across the two datasets. To our knowledge, this is the first application of supervised deep learning to imaging flow cytometry images of a subcellular organelle for inflammatory phenotyping, establishing organelle level image analysis as a potential complement to molecular assays for immune cell classification.

bioinformatics↗

bioRxiv2026

Morphological Hallmarks of White Matter Astrocytes Across Development

While the mechanisms underlying protoplasmic, or gray matter (GM), astrocyte morphogenesis have been extensively studied, far less is known about the development of fibrous, or white matter (WM), astrocytes. In particular, no study has comprehensively characterized the morphological development of WM astrocytes. To address this gap, we used an intracortical adeno-associated virus injection technique to sparsely label individual astrocytes in the mouse corpus callosum for morphological analysis across postnatal development. We found that astrocytes in the corpus callosum follow a distinct trajectory of morphological development from their GM cousins. They establish their branching complexity in the first two postnatal weeks and significantly increase in volume into adulthood. We further show that tiling is conserved in WM astrocytes and that a subset of mature astrocytes in the corpus callosum lack endfeet. These findings provide the first developmental framework for the morphogenesis of callosal astrocytes and establish a foundation for future studies investigating the molecular mechanisms regulating the development of WM astrocytes.

neuroscience↗

bioRxiv2026

A superfolding sfPepper enables superresolution and single-molecule RNA imaging in live cells

Fluorescent RNAs (FRs) have emerged as powerful tools for the visualization of RNA in live cells. However, the poor folding of FRs is a major limiting factor in many advanced RNA imaging applications. Here, we describe a novel strategy to significantly increase the folding efficiency of FRs. By introducing a stabilizing stem and performing targeted mutagenesis at key junction regions, we develop a superfolder variant of the Pepper aptamer, termed sfPepper, that exhibits a striking twofold increase in folding efficiency compared with Pepper under physiological conditions. sfPepper exhibits much improved thermostability and reduced ion dependence and, most importantly, is substantially brighter than Pepper in live cells. sfPepper enables robust imaging of diverse noncoding RNAs and messenger RNAs with increased signal-to-background ratios and stimulated emission depletion (STED) superresolution imaging of CUG trinucleotide repeat-containing toxic RNAs. Remarkably, only four sfPepper repeats facilitate sensitive single-molecule mRNA tracking, revealing dynamic heterogeneity among ER-associated mRNA molecules. Together, this study establishes an efficient strategy for improving FR folding and offers a powerful tool for fluorescently labelling RNAs in live cells.

molecular biology↗

bioRxiv2026

De novo design of ligand binding and sensing with a physics based generative approach

The de novo design of ligand-binding proteins has tremendous potential to revolutionize biosensor technology, yet converting these designs into functional sensors remains a major challenge due to the need for ligand-induced conformational changes or modulation of protein-protein interactions. Here, we introduce a physics-based generative approach for the de novo creation of proteins that bind small molecules and metal ions. Our method achieves customizable ligand-binding pocket formation in parallel with simulated protein folding, allowing for precise architectural control of the protein-ligand complex and facilitating the development of biosensors based on either ligand-triggered protein reassociation via split-protein reassembly or ligand-induced protein folding. We demonstrate the versatility of our computational method through successful designs targeting five small molecules, including the very small neurotransmitters serotonin and dopamine, and two metal ions. Biophysical characterization confirmed correct ligand binding, and crystal structures closely matched computational models. We demonstrated the biosensor engineering potential of these designs by constructing serotonin and dopamine sensors using a split protein strategy and explored several approaches to enhance sensor activity. Additionally, we developed a zinc sensor through a zinc-induced protein folding mechanism. Overall, our physics-based generative approach provides a robust framework for the de novo design of ligand-binding proteins, opening new avenues for the development of ligand-responsive biosensors.

bioinformatics↗

bioRxiv2026

Integrating Machine-learning and Ultra-high-throughput Screening for Enzyme spaces exploration

AbstractThe systematic navigation of biocatalyst space is constrained by elusive structure-activity rules and a lack of evolutionary history. Here, we present IMUSE, a strategy integrating machine learning with ultra-high-throughput screening. By screening millions of droplet-encapsulated de novo enzymes, we generated massive synthetic sequence-structure datasets to train models that capture their complex fitness landscapes and biophysical principles. These models effectively guide functional exploration across both sequence and novel structure spaces. IMUSE identified synergistic triple mutations yielding [~]5-fold activity improvements and discovered active second-generation designs with novel catalytic pockets, boosting the experimental success rate >4.9-fold ([~]30%). This work demonstrates how synthetic fitness landscapes bridge the data gap in de novo enzyme space, transforming stochastic search into deterministic navigation to unlock highly proficient biocatalysts beyond natural boundaries.

biochemistry↗

bioRxiv2026

CSF glucosylsphingosine is a central readout of GCase impairment across genetic and sporadic Parkinson’s disease

Parkinsons disease (PD) risk converges on lysosomal biology, including reduced activity of glucocerebrosidase (GCase), encoded by GBA1. Glucosylsphingosine (GlcSph), a toxic deacylated glycosphingolipid and established target engagement biomarker in Gaucher disease, is difficult to quantify in cerebrospinal fluid (CSF) because of low abundance and isomeric interference. We optimized and qualified targeted LC-MS/MS assays for GlcSph in CSF and plasma and measured GlcSph and GCase activity (4-MU assay) in participants from the Parkinsons Progression Markers Initiative (PPMI), including GBA1 and LRRK2 mutation carriers with and without PD, sporadic PD, and healthy controls. Group level analyses showed higher levels of CSF and plasma GlcSph in GBA1 heterozygous variant carriers independent of disease status, [~]120% (p<0.0001) and [~]61% (p<0.0001), respectively, with CSF elevations reflecting allelic dosage and, to a lesser degree, variant severity. Notably, elevated CSF GlcSph was observed not only in carriers of pathogenic GBA1 mutations but also in individuals harboring common GBA1 risk variants. CSF GlcSph was also elevated by [~]30% in sporadic PD (p=0.002) and by [~]40% LRRK2 mutation carriers (p=0.0003), indicating shared central GCase pathway perturbation across PD subtypes. CSF and plasma GlcSph levels showed poor concordance across groups with the exception of the GBA-PD group, supporting the hypothesis that central and peripheral GCase pathway dysfunction arise through distinct biological mechanisms. Together, these findings establish CSF GlcSph as a sensitive biomarker of central GCase pathway impairment and may support its use as a pharmacodynamic and target engagement biomarker for therapeutics targeting GCase in PD. More broadly, CSF GlcSph may enable identification of biologically defined GCase-pathway dysfunction beyond genetically defined GBA1-associated PD, with potential implications for patient stratification in future disease-modifying trials.

Molecular Biology↗

bioRxiv2026

HalluDesign-NA: Extending HalluDesign for De Novo Nucleic Acid Design

AlphaFold3 has revolutionized the prediction of biomolecular structures and interactions, including atomic-level modeling of nucleic acids. However, the de novo design of structured and functional nucleic acids remains a significant challenge. Here, we extend our HalluDesign framework to nucleic acid design by integrating NA-MPNN for nucleic acid sequence optimization and design. This new framework, HalluDesign-NA, enables iterative sequence-structure co-optimization, facilitating the de novo design of nucleic acids. Computational benchmarking across ssDNA, ssRNA, and aptamer design tasks demonstrates consistent improvements in confidence scores (pLDDT, ipTM), supporting the feasibility of de novo nucleic acid design under various constraints, such as sequence length, symmetry, and protein structure context. We anticipate that HalluDesign-NA will accelerate the de novo design of functional nucleic acids for applications in biotechnology and medicine. The source code for HalluDesign-NA is available at https://github.com/MinchaoFang/HalluDesign_NA.

bioinformatics↗

bioRxiv2026

Dual histone methylation reader ZCWPW2 links histone methylation to initiation of meiotic recombination

Meiotic homologous recombination initiates with the formation of programmed DNA double-strand breaks (DSBs) by complexes comprising SPO11 and accessory proteins at discrete sites called recombination hotspots. In mammals, PRDM9-dependent H3K4me3 and H3K36me3 define recombination hotspots, but how these epigenetic characteristics determine the physiological DSB formation remains unknown. Here we show that dual histone methylation reader ZCWPW2 can recognize H3K4me3 and H3K36me3 marks in testis. The binding activity of ZCWPW2 to histone methylation is dependent on PRDM9 function. Moreover, we find the epigenetic writer-reader axis PRDM9-ZCWPW2 is essential for DSB formation at meiotic recombination hotspots, which may be partly explained by the finding that ZCWPW2 can physically interact with HORMAD1, IHO1, MEI4, and REC114. Finally, the absence of ZCWPW2 leads to disrupted chromosomal synapsis and recombination, thereby obstructing meiotic progression. Taken together, our findings provide new insights into how histone modifications and their associated regulatory proteins collectively regulate meiotic homologous recombination initiation.

cell biology↗

bioRxiv2026

The resource-rational dynamics of evidence accumulation

Evidence accumulation is a fundamental aspect of human decision-making. However, how the precise temporal structure of evidence shapes the accumulation process has not been systematically studied. As a result, current understanding of evidence accumulation remains largely limited to its time-averaged behavior. We tested human subjects in a visual estimation task in which they inferred the angular position of an unknown source from a noisy stimulus sequence. Introducing systematic temporal perturbations, i.e., breaks of different durations and at different positions in the otherwise regular evidence sequence, revealed that subjects actively compensated for the memory loss endured during the break by dynamically enhancing evidence integration and memory maintenance immediately after the break. We derived a new time-continuous Bayesian updating model that is dynamically constrained by optimal performance-effort trade-offs. With two free parameters determining the overall resource-efficiencies of encoding and memory maintenance, the model accurately predicts the rich dependencies of subjects accumulation behavior on the evidence schedule, including subjects individual tendencies to emphasize either early (primacy) or late (recency) samples in the evidence sequence. Our results suggest that evidence accumulation is a non-stationary, dynamically controlled process that optimally balances the information gained from incoming evidence against the cognitive effort required to acquire and maintain it. The proposed model is general and should apply broadly across many task domains.

animal behavior and cognition↗

bioRxiv2025

Highly Efficient Multiplexed Genome Engineering and Clone Selection to Enable Next Generation Induced Pluripotent Stem Cell (iPSC)-based Cell Therapies

Human induced pluripotent stem cells (iPSCs) have revolutionized regenerative medicine and cellular therapies. To improve the functionality and safety of iPSC-based therapies, genome engineering has been employed to disrupt gene expression and introduce therapeutic transgenes. However, current genome editing methods face significant challenges, including labor-intensive procedures, low efficiency, and safety concerns. Here, we report a novel and efficient approach for multiplex genome engineering in iPSCs using Integrase-mediated Programmable Genomic Integration (I-PGI) technology. I-PGI combines CRISPR-mediated genome editing and site-specific integrases, enabling precise gene insertion without DNA double-strand breaks. By optimizing I-PGI components and protocols, we achieved multiple gene knockouts and knock-ins of up to five transgenes in a single process. Additionally, we developed a streamlined workflow for enrichment, deposition, and screening of single-cell clones with the desired genome edits. This approach significantly accelerates and improves the precision of multiplex iPSC engineering, paving the way for next-generation iPSC-based therapies.

genomics↗

bioRxiv2025

Joint Biophysical Modeling of Paired Single-Cell RNA and Protein Measurements

Surface protein measurements can supplement gene expression information from single-cell RNA sequencing to provide a more complete assessment of cell identity and function. Recently developed multiomic assays facilitate such measurements, and can, in principle, be utilized to understand the dynamics of transcription and translation. We develop a framework for biophysical modeling of transcription jointly with translation from single-cell data, along with a suitable technical noise model for sequencing data. We demonstrate its efficacy using simulations, and illustrate how it can be useful in practice with 10x multiomic data. Our proof-of-principle highlights the potential for jointly modeling transcription and translation as data quality and measurement accuracy improves.

biophysics↗

bioRxiv2025

Skull Bone Marrow Drainage and Its Associations with Inflammation, Sleep Quality, and Cognitive Performance

BackgroundRecent studies using human craniotomy samples and murine models have revealed channels that connect the inner skull cortex to the skull marrow cavities. While cerebrospinal fluid (CSF) tracer access to skull bone marrow has been described in CSF circulation disorders, evidence on regional patterns and clinical correlates in broader clinical cohorts is limited. MethodsThree-dimensional imaging was performed at baseline, 4.5, 15, and 39 hours post-gadolinium administration in 87 patients with neurological conditions. Signal changes in skull bone marrow near the superior sagittal sinus, lateral fissure, and cisterna magna were recorded. CSF drainage function was characterized as the percentage changes in the signal unit ratio within the skull bone marrow from baseline to time points post administration of gadolinium. ResultsThe tracer drained from CSF into the skull bone marrow, highlighting the drainage variations nearby different regions. Reduced drainage was associated with female sex, hypertension, diabetes, and elevated neutrophil levels (p < 0.05). Drainage near the superior sagittal sinus was inversely associated with sleep quality and positively associated with cognitive scores, and partially accounted for the association of sleep quality with cognitive function (indirect effect {Delta}{beta} = -0.092; bootstrap 95% CI -0.195 to -0.004; proportion mediated {approx} 38%). ConclusionThese results support the presence of skull bone marrow adjacent routes for CSF tracer movement and suggest that skull bone marrow tracer dynamics relate to systemic inflammation, sleep quality, and cognitive performance in this clinical cohort.

neuroscience↗

bioRxiv2025

HalluDesign: Protein Optimization and de novo Design via Iterative Structure Hallucination and Sequence design

Deep learning has revolutionized biomolecular modeling, enabling the prediction of diverse structures with atomic accuracy. However, leveraging the atomic-level precision of the structure prediction model for de novo design remains challenging. Here, we present HalluDesign, a general all-atom framework for protein optimization and de novo design, which iteratively update protein structure and sequence. HalluDesign harnesses the inherent hallucination capabilities of AlphaFold3-style structure prediction models and enables fine-tune free, forward-pass only sequence-structure co-optimization. Structure conditioning at different noise level in the structure prediction stage allows precise control over the sampling space, facilitating tasks from local and global protein optimization to de novo design. We demonstrate the versatility of this framework by optimizing suboptimal structures, rescuing previously unsuccessful designs, designing new biomolecular interactions and generating new protein structures from scratch. Experimental characterization of a diverse set of proteins--including protein binders for small molecules, a metal ion and proteins; antibody design of phosphorylation-specific peptide; and monomeric proteins--revealed high design success rates and excellent structural accuracy. Together, our comprehensive computational and experimental results highlight the broad utility of this framework. We anticipate that HalluDesign will further unlock the modeling and design potential of AlphaFold3-like models, enabling the robust creation of complex biomolecules for a wide range of biotechnological applications.

bioinformatics↗

bioRxiv2025

Engineering Antibody-Drug Conjugates targeting an Adhesion GPCR, CD97

Adhesion G protein-coupled receptors (aGPCRs) are key cell-adhesion molecules involved in many cellular functions and contribute to human diseases, including cancer. aGPCRs are characterized by large extracellular regions that could serve as readily accessible antigens. However, the potential of aGPCRs as targets for biologic therapeutics has not been extensively explored. CD97, also known as ADGRE5, is an aGPCR that is upregulated in various cancer types, including acute myeloid leukemia (AML) and glioblastoma (GBM), and their respective cancer stem cells. Here, we developed antibody-drug conjugates (ADCs) targeting CD97 and assessed their efficacy against AML and GBM cells. We generated a panel of synthetic human antibodies targeting distinct epitopes of CD97, from which we identified an antibody that was efficiently internalized. This antibody binds to all isoforms of human CD97 but not to its close homolog, EMR2. Structure determination by single-particle cryo-electron microscopy revealed that this antibody targets the CD97 GPCR autoproteolysis-inducing (GAIN) domain, whose presence is conserved in aGPCRs, through an unconventional binding mode where it extensively utilizes the light chain framework for antigen recognition. Screening of conjugation methods and payloads resulted in a stable ADC that effectively killed AML and GBM cell lines, as well as patient-derived GBM stem cells, with minimal cytotoxicity against peripheral blood mononuclear cells from healthy donors. Our study demonstrates the therapeutic potential of targeting CD97, as well as the aGPCR GAIN domain in general, and broadens our mechanistic understanding of antibody-antigen interactions.

biochemistry↗

bioRxiv2025

A small molecule inhibitor of NVL suppresses tumor growth by blocking ribosome biogenesis

A longstanding hypothesis, stemming from the enlarged nucleoli typical of cancer cells, posits ribosome production as a selective cancer liability. Certain genotoxic chemotherapies work partly by disrupting ribosome biogenesis, highlighting the need for selective inhibitors of this pathway. Using forward genetics, we identified mutations in the essential 60S ribosomal subunit assembly factor NVL that confer resistance to MM17, a dibenzothiazepinone with anticancer activity. Cryo-EM reconstructions of the NVL hexameric assembly reveal two MM17 docking sites adjacent to resistance mutations. NVL inhibition by MM17 arrests 60S biogenesis in the nucleolus and induces cell cycle arrest or apoptosis through both MDM2/p53-dependent and p53-independent pathways, without causing DNA damage. A bioavailable analog, MM927, suppresses tumor growth in mouse models of leukemia and colorectal cancer without observable toxicity. These findings establish NVL inhibitors as a promising new class of targeted therapeutics and validate ribosome biogenesis as a cancer-specific vulnerability.

biochemistry↗

bioRxiv2025

Longitudinal Imaging of Experimental Intracerebral Hemorrhage Pathology with Iodine-enhanced Micro-CT

BackgroundIodine-enhanced micro-computed tomography (Micro-CT) enables high-resolution three-dimensional imaging of brain architecture. This study aimed to characterize both acute and chronic pathological changes following intracerebral hemorrhage (ICH) using iodine-enhanced micro-CT. MethodExperimental ICH was induced in 8- to 10-week-old C57BL/6 mice (n = 76) via stereotaxic injection of either 0.075 U collagenase IV or 30 l autologous blood. Iodine-enhanced micro-CT imaging was performed at 4 hours, 1, 3, and 7 days after intracerebral hemorrhage post-ICH to evaluate hematoma formation and erythrolysis. Chronic alterations, including ventriculomegaly and ipsilateral lesion, were assessed at 28 days post-ICH. In parallel, MRI was conducted at 1, 7, and 28 days following autologous blood injection, followed by micro-CT, to facilitate cross-modality quantitative analysis. Lesion volumes were compared between imaging modalities over time. ResultsMicro-CT enabled quantification of hematoma volume and erythrolysis in ICH models. Hematomas extended along perivascular pathways toward the cerebral surface in both collagenase- and autologous blood-induced ICH models. At 28 days post-ICH, micro-CT detected ventriculomegaly and hypodense lesions without concurrent expansion of the choroid plexus. Lesion volume measurements derived from micro-CT correlated with those from MRI, enabling quantitatively assessment of iron deposition and hematoma size alterations after ICH. ConclusionIodine-enhanced micro-CT provides a robust and high-resolution imaging platform for evaluating hematoma evolution, hemolysis, ventricular enlargement, and chronic brain lesions in experimental ICH. When integrated with MRI, these multimodal imaging approaches enhance the characterization of both hematoma volume and iron deposition following ICH.

neuroscience↗

bioRxiv2025

Extrinsic biological stochasticity and technical noise normalization of single-cell RNA sequencing data

The technical noise introduced during single-cell RNA sequencing (scRNA-seq) has led to the use of size factor normalization as a first step prior to data analysis. However, this scaling approach inherently affects extrinsic (between cell) variability of gene expression, which stems from both biological and technical factors. Based on previous models on biological and technical extrinsic noise, we propose a general extrinsic noise model for scRNA-seq to provide a theoretical basis for size factor normalization, thus providing a framework for estimating both biological and technical components of extrinsic noise. We highlight the relationship between normalized gene expression covariance, extrinsic noise, and overdispersion, showing that extrinsic noise explains the baseline overdispersion commonly observed in scRNA-seq data. We validated the technical model by testing the relationship on data from pooled RNA. Interestingly, our model accurately describes mature mRNA counts but not nascent mRNA counts, suggesting the need for an alternative technical model for data derived from nascent transcripts. Using single-cell RNA-seq data, we characterize both biological and technical extrinsic noise and cell size factors estimated using Poisson-like genes. Overall, our model helps clarify common misconceptions and provides insight into the role of extrinsic noise and size factor normalization in scRNA-seq data.

bioinformatics↗