Biology subjects

Liu, X.

Publications and source records attributed to Liu, X..

At least 559 records · Page 31Linked to original sources

bioRxiv2025

Comparative molecular landscapes of immature neurons in the mammalian dentate gyrus across species reveal special features in humans

Immature dentate granule cells (imGCs) arising from adult hippocampal neurogenesis contribute to plasticity, learning and memory, but their evolutionary changes across species and specialized features in humans remain poorly understood. Here we performed machine learning-augmented analysis of published single-cell RNA-sequencing datasets and identified macaque imGCs with transcriptome-wide immature neuronal characteristics. Our cross-species comparisons among humans, monkeys, pigs, and mice showed few shared (such as DPYSL5), but mostly species-specific gene expression in imGCs that converged onto common biological processes regulating neuronal development. We further identified human-specific transcriptomic features of imGCs and demonstrated functional roles of human imGC-enriched expression of a family of proton-transporting vacuolar-type ATPase subtypes in development of imGCs derived from human pluripotent stem cells. Our study reveals divergent gene expression patterns but convergent biological processes in the molecular characteristics of imGCs across species, highlighting the importance of conducting independent molecular and functional analyses for adult neurogenesis in different species.

neuroscience↗

bioRxiv2025

A new structural paradigm for outer membrane protein biogenesis in the Bacteroidota

In Gram-negative bacteria the outer membrane (OM) is the first line of defence against antimicrobial agents and immunological attacks. A key part of OM biogenesis is insertion of OM proteins (OMPs) by the {beta}-barrel-assembly machinery (BAM). Here we report the cryo-electron microscopy (cryoEM) structure of a BAM complex isolated from Flavobacterium johnsoniae, a member of the phylum Bacteroidota that includes key human commensals and major anaerobic pathogens. This BAM complex is radically different from the canonical Escherichia coli system and includes an extensive extracellular canopy that overhangs the substrate folding site and a subunit that inserts into the BAM pore. We find that two of the novel subunits involved in forming the extracellular canopy are essential for BAM function. For one of these subunits isolation of a suppressor mutation allows the separation of their essential and non-essential functions. The need for a highly remodelled and enhanced BAM complex reflects the unusually complex membrane proteins found in the OM of the Bacteroidota.

microbiology↗

bioRxiv2025

CRISPR-based shuttle cloning of 1,397 human genes into UAS vectors

The construction of genome-wide DNA libraries from publicly available resources is essential for leveraging functional genomics to investigate complex biological systems. However, all existing high-throughput cloning methods for transferring DNA fragments between vectors require PCR amplification of the DNA fragments, rendering the construction of genome-wide DNA libraries labor-intensive and time-consuming. By introducing a concept of CRISPRshuttle cassette, we herein present a method named CRISPR-based shuttle cloning (CRISPRshuttle cloning). This method enables the high-throughput transfer of numerous DNA fragments from original plasmids with identical backbones to a different vector background without the need for PCR amplification of the DNA fragments. The procedure comprises two-step test tube reactions followed by bacterial transformation. Using CRISPRshuttle we successfully generated a library of GAL4/UAS-based UAS-ORF plasmids covering 1,397 human genes conserved in Drosophila. This library may serve as a valuable resource for gain-of-function screening in cultured cells and for the creation of a transgenic UAS-ORF library in Drosophila.

molecular biology↗

bioRxiv2025

cfDecon: Accurate and Interpretable methylation-based cell type deconvolution for cell-free DNA

Cell-free DNA (cfDNA) analysis has emerged as a powerful tool for noninvasive diagnostics and monitoring of aberrant methylation. However, computational deconvolution of cfDNA remains challenging due to the complexity of methylation data, the diverse cellular compositions in cfDNA mixtures, and the limited interpretability of current methods. We present cfDecon, an efficient deep-learning framework for cfDNA deconvolution at the resolution of individual reads. cfDecon employs a multichannel autoencoder core module and an iterative refinement process to estimate cell-type proportions and generate condition-aware cell-type-specific methylation profiles. We rigorously evaluate cfDecon through comprehensive simulation experiments, including scenarios with normal cellular compositions, rare cell types, and unknown cell types, where cfDecon consistently outperforms previous state-of-the-art methods in deconvolution. Using strict data separation and careful control for potential data leakage, cfDecon shows scalability and applicability to large-scale studies by raising Lins Concordance Correlation Coefficient (CCC) over 33% on an atlas-level reference dataset. We further validate cfDecons utility in disease diagnosis on two clinical datasets, Amyotrophic Lateral Sclerosis (ALS) and Hepatocellular Carcinoma (HCC). cfDecon raises the performance of disease detection from 0.53 to 0.79 compared to existing methods on the ALS dataset and 0.55 to 0.77 on the HCC dataset. The capacity of cfDecon to predict condition-specific DNA methylation patterns across cell types offers a valuable framework for studying differentially methylated CpGs. Furthermore, integrating cfDecon with enrichment analysis facilitates the exploration of functional variations in CpGs across different conditions. cfDecon represents a significant advancement in cfDNA analysis, offering improved accuracy, interpretability, and potential for personalized medicine applications. The codebase is available at https://github.com/Susanxuan/cfDecon.

bioinformatics↗

bioRxiv2025

Learning universal knowledge graph embedding for predicting biomedical pairwise interactions

Predicting biomedical interactions is crucial for understanding various biological processes and drug discovery. Graph neural networks (GNNs) are promising in identifying novel interactions when extensive labeled data are available. However, labeling biomedical interactions is often time-consuming and labor-intensive, resulting in low-data scenarios. Furthermore, distribution shifts between training and test data in real-world applications pose a challenge to the generalizability of GNN models. Recent studies suggest that pre-training GNN models with self-supervised learning on unlabeled data can enhance their performance in predicting biomedical interactions. Here, we propose LukePi, a novel self-supervised pre-training frame-work that pre-trains GNN models on biomedical knowledge graphs (BKGs). LukePi is trained with two self-supervised tasks: topology-based node degree classification and semantics-based edge recovery. The former is to predict the degree of a node from its topological context and the latter is to infer both type and existence of a candidate edge by learning semantic information in the BKG. By integrating the two complementary tasks, LukePi effectively captures the rich information from the BKG, thereby enhancing the quality of node representations. We evaluate the performance of LukePi on two critical link prediction tasks: predicting synthetic lethality and drug-target interactions, using four benchmark datasets. In both distribution-shift and low-data scenarios, LukePi significantly outperforms 15 baseline models, demonstrating the power of the graph pre-training strategy when labeled data are sparse.

bioinformatics↗

bioRxiv2025

ML-MAGES: A machine learning framework for multivariate genetic association analyses with genes and effect size shrinkage

A fundamental goal of genetics is to identify which and how genetic variants are associated with a trait, often using the regression results from genome-wide association (GWA) studies. Important methodological challenges are accounting for inflation in GWA effect estimates as well as investigating more than one trait simultaneously. We leverage machine learning approaches for these two challenges, developing a computationally efficient method called ML-MAGES. First, we shrink the inflation in GWA effect sizes caused by non-independence among variants using neural networks. We then cluster variant associations among multiple traits via variational inference. We compare the performance of shrinkage via neural networks to regularized regression and fine-mapping, two approaches used for addressing inflated effects but dealing with variants in focal regions of different sizes. Our neural network shrinkage outperforms both methods in approximating the true effect sizes in simulated data. Our infinite mixture clustering approach offers a flexible, data-driven way to distinguish different types of associations--trait-specific, shared across traits, or spurious--among multiple traits based on their regularized effects. Clustering applied to our neural network shrinkage results also produces consistently higher precision and recall for distinguishing gene-level associations in simulations. We demonstrate the application of ML-MAGES on association analyses of two quantitative traits and two binary traits in the UK Biobank (genetic and phenotypic data from 500,000 residents of the UK). Our identified associated genes from single-trait enrichment tests overlap with those having known relevant biological processes to the traits. Besides trait-specific associations, ML-MAGES identifies several variants with shared multi-trait associations, suggesting putative shared genetic architecture.

genetics↗

bioRxiv2025

TAC-C uncovers open chromatin interaction in crops and SPL-mediated photosynthesis regulation

Cis-regulatory elements (CREs) direct precise gene expression for development and environmental response, yet their spatial organization in crops is largely unknown. We introduce Transposase-Accessible Chromosome Conformation Capture (TAC-C), a method integrating ATAC-seq and Hi-C to capture fine-scale chromatin interactions in four major crops: rice, sorghum, maize, and wheat. TAC-C reveals that chromatin interaction frequency aligns with genome size and gene expression, exhibiting distinct loop structures between C3 and C4 plants, particularly in C4-specific enzymes coding genes. Integrating chromatin structure with population genetics data highlights that chromatin loops connect distal regulatory elements to phenotypic variation. Asymmetrical open chromatin interactions among subgenomes, driven by transposon insertions and sequence variations, contribute to biased homoeolog expression. Furthermore, TaSPL7/15 regulate photosynthesis-related genes through chromatin interactions, with enhanced photosynthetic efficiency and starch content in Taspl7&15 mutant. TAC-C provides new insights into the spatial organization of regulatory elements in crops, especially for SPL-mediated photosynthesis regulation in wheat. TeaserTAC-C Reveals Chromatin Interactions and SPL-Mediated Photosynthesis Regulation in Crops.

plant biology↗

bioRxiv2025

Proteoform identification using multiplexed top-down mass spectra

Top-down mass spectrometry (TDMS) is the method of choice for analyzing intact proteoforms, as well as their post-translational modifications and sequence variations. In TDMS experiments, multiple proteoforms are often co-fragmented in tandem mass spectrometry (MS/MS) analysis, resulting in multiplexed TD-MS/MS spectra. Since multiplexed TD-MS/MS spectra are more complex than common spectra generated from single proteoforms, these spectra pose a significant challenge for proteoform identification and quantification. Here we present TopMPI, a new computational tool specifically designed for the identification of multiplexed TD-MS/MS spectra. Experimental results demonstrate that TopMPI significantly increases proteoform identifications and reduces identification errors in multiplexed TD-MS/MS spectral analysis compared to existing tools.

bioinformatics↗

bioRxiv2025

Mechanism of forced-copy-choice RNA recombination by enteroviral RNA-dependent RNA polymerases

Forced-copy-choice recombination occurs at the end of a template, differing from copy-choice recombination, which happens at internal positions. This mechanism may produce full-length genomes from fragments created by host antiviral responses. Previous studies from our laboratory demonstrated that poliovirus (PV) RNA-dependent RNA polymerase (RdRp) switches to an "acceptor" template in vitro when initiated on a heteropolymeric RNA-primed "donor" template. Surprisingly, recombinants showed template switching from the 3-end of the donor template. We have developed a primed-template system to study PV RdRp-catalyzed forced-copy-choice RNA recombination. PV RdRp adds a single, non-templated nucleotide to the 3-end of a blunt-ended, double-stranded RNA product, forming a "plus-one" intermediate essential for template switching. Non-templated addition of CMP was favored over AMP and GMP (80:20:1); UMP addition was negligible. A single basepair between the plus-one intermediate and the 3-end of the acceptor template was necessary and sufficient for template switching, which could occur without RdRp dissociation. Formation of the plus-one intermediate was rate limiting for template switching. PV RdRp also utilized synthetic, preformed intermediates, including those with UMP 3-overhangs. Reactions showed up to five consecutive template-switching events, consistent with a repair function for this form of recombination. PV RdRp may exclude UMP during forced-copy-choice RNA recombination to preclude creation of nonsense mutations during RNA fragment assembly. Several other picornaviral RdRps were evaluated, and all were capable of RNA fragment assembly to some extent. Lastly, we propose a structure-based hypothesis for the PV RdRp-plus-one intermediate complex based on an elongating PV RdRp structure.

biochemistry↗

bioRxiv2025

Myelination across cortical hierarchies and depths in humans and macaques

Myelination is fundamental to brain function, enabling rapid neural communication and supporting neuroplasticity throughout the lifespan. While hierarchical patterns of myelin maturation across the cortical surface are well-documented in humans, it remains unclear which features reflect evolutionarily conserved developmental processes versus human-characteristic adaptations. Moreover, the laminar development of myelin across the primate cortical surface, which shapes hierarchies and supports functions ranging from sensory integration to network communication, has been largely unexplored. Using neuroimaging to measure the T1-weighted/T2-weighted ratio in tissue contrast as a proxy for myelin content, we systematically compared depth-dependent trajectories of myelination across the cortical surface in humans and macaques. We identified a conserved "inside-out" pattern, with deeper layers exhibiting steeper increases in myelination and earlier plateaus than superficial layers. This depth-dependent organization followed a hierarchical gradient across the cortical surface, progressing from early maturation in sensorimotor regions to prolonged development in association areas. Humans exhibited a markedly extended timeline of myelination across both cortical regions and depths compared to macaques, allowing for prolonged postnatal plasticity across the entire cortical hierarchy -- from sensory and motor processing to higher-order association networks. This extended potential for plasticity may facilitate the shaping of cortical circuits through postnatal experience in ways that support human-characteristic perceptual and cognitive capabilities.

developmental biology↗

bioRxiv2025

Histone Deacetylase Inhibitor Largazole Deactivates A Subset of Superenchancers and Causes Mitotic Chromosome Mis-alignment by Suppressing SP1 and BRD4

Histone deacetylase inhibitors have been investigated as potential therapeutic agents for cancer and other diseases. HDIs are known to promote histone acetylation, resulting in an open chromatin conformation and generally increased gene expression. In previous work, we reported that a subset of genes, particularly those regulated by superenhancers, can be suppressed by the HDAC inhibitor largazole. To elucidate the molecular mechanisms underlying gene repression by largazole, we conducted transposase-accessible chromatin sequencing, ChIP-seq, and RNA-seq studies. Our findings revealed that while largazole treatment generally enhances chromatin accessibility, it selectively decreases the accessibility of a subset of superenhancer regions. These genomic regions, showing the most significant changes in the presence of largazole, were enriched with transcription factor binding motifs for SP1, BRD4, CTCF, and YY1. ChIP-seq analysis confirmed reduced binding of BRD4 and SP1 at their respective sites on chromatin, particularly at superenhancers regulating genes such as ID1, c-Myc and MCMs. Largazole exerts its effects by inhibiting DNA replication, RNA processing, and cell cycle progression, partially through the suppression of SP1 expression. Depletion of SP1 by shRNA mimics several key biological effects of largazole and increases cellular sensitivity to the drug. Specific to cell cycle regulation, we demonstrated that largazole disrupts G/M transition by interfering with chromosome alignment during metaphase, a phenotype also observed with SP1 depletion. Our results suggest that largazole exerts its growth-inhibitory effect by suppressing BRD4 and SP1 at super-enhancers, leading to cytostatic responses and mitotic dysfunction.

cancer biology↗

bioRxiv2025

mRNA lipid nanoparticle-incorporated nanofiber-hydrogel composite generates a local immunostimulatory niche for cancer immunotherapy

Hydrogel materials have emerged as versatile platforms for various biomedical applications. Notably, the engineered nanofiber-hydrogel composite (NHC) has proven effective in mimicking the soft tissue extracellular matrix, facilitating substantial recruitment of host immune cells and the formation of a local immunostimulatory microenvironment. Leveraging this feature, here we report an mRNA lipid nanoparticle (LNP)-incorporated NHC microgel matrix, termed LiNx, by incorporating LNPs loaded with mRNA encoding tumour antigens. Harnessing the potent transfection efficiency of LNPs in antigen-presenting cells (APCs), LiNx demonstrates remarkable immune cell recruitment, antigen expression and presentation, and cellular interaction. These attributes collectively create an immunostimulating milieu and yield a potent immune response achievable with a single dose, comparable to the conventional three-dose LNP immunization regimen. Further investigations reveal that the LiNx not only generates heightened Th1 and Th2 responses but also elicits a distinctive Type 17 T helper cell-mediated response pivotal for bolstering antitumour efficacy. Our findings elucidate the mechanism underlying LiNxs role in potentiating antigen-specific immune responses, presenting a new strategy for cancer immunotherapy.

bioengineering↗

bioRxiv2025

Zmap: an intelligent region-allocation method to map single-cell into spatial data

Integration of single-cell and spatial transcriptome represents as a fundamental strategy to enhance spatial data quality. However, existing methods for mapping single-cell data to spatial coordinates struggle with large-scale datasets comprising millions of cells. Here, we introduce Zmap, an intelligent region-allocation method inspired by the region-of-interest (ROI) concept from image processing. By using gradient descent, Zmap allocates cells to structured spatial regions that matching the most significant biological information, optimizing the integration of data and improving speed and accuracy. Zmap excels in integrating data even in the presence of various sequencing artifacts, such as cell segmentation errors and imbalanced cell-type representations. Zmap outperforms state-of-the-art methods by 10 to 1000 times in speed, and it is the only approach capable of integrating datasets containing millions of cells in a single run. As a result, Zmap uncovers originally hidden gene expression patterns in the brain section, offers new insights into organogenesis and tumor microenvironments, all with exceptional efficiency and accuracy.

bioinformatics↗

bioRxiv2025

Optimized antibody with W32I mutation enhances antiviral efficacy against influenza

The influenza virus has caused a global pandemic with significant morbidity and mortality, highlighting the need to optimize antibodies for improved antiviral efficacy. The 3E1 antibody effectively neutralizes influenza subtypes H1 and H5 by inhibiting acid-induced conformational changes of hemagglutinin (HA). This study aimed to optimize the antibodys bioactivity by modifying amino acid residues, resulting in single-point mutants (3E1-L [W32I], 3E1-H [F103I]) and a double mutant (3E1-H+L [F103I, W32I]). The binding affinity, neutralizing activity, and antiviral mechanisms of the mutants were evaluated. Notably, the 3E1-L mutant showed significantly enhanced antiviral activity against H1N1 and H3N2 compared to wild-type 3E1, inhibiting both viral entry and release. The prophylactic and therapeutic efficacy of the 3E1-L mutant was validated. Molecular dynamics simulations of the 3E1-L/HA complex showed that the W32I mutation reduces steric hindrance between tryptophan at position 32 and the complementarity-determining region (CDR) L1 loop of HA. In conclusion, the W32I substitution enhances the antiviral activity of wild-type 3E1, making the optimization of 3E1-L a promising strategy for developing more effective influenza therapies.

immunology↗

bioRxiv2025

CDP-DAG synthases regulate plant growth and broad-spectrum disease resistance

Phosphatidic acid (PA) functions as a cell membrane component and signaling molecule in plants. PA metabolism has multiple routes, in one of which PA is converted into cytidine diphosphate diacylglycerol (CDP-DAG) by CDP-DAG synthases (CDSs). CDS genes are highly conserved in plants. Here, we found that knock-down of the CDS gene enhanced the resistance of Arabidopsis thaliana to multiple pathogens, with a growth penalty. When Arabidopsis leaves were treated with chitin or flg22, reactive oxygen species (ROS) production in cds mutants was significantly higher than that in the wild-type (WT). Similarly, phosphorylation of mitogen-activated protein kinases (MAPKs) in the cds1cds2 double mutant was significantly increased compared to the WT. By integrating lipidomics, transcriptomics, and metabolomics data, PA accumulation was observed in mutants cds1cds2, activating the jasmonic acid (JA) and salicylic acid (SA) signaling pathway, and increasing transcript levels of plant defense-related genes. Significant accumulation of the downstream metabolites including serotonin and 5-methoxyindole was also found, which plays important roles in plant immunity. In conclusion, our study indicated the role of CDSs in broad-spectrum disease resistance in Arabidopsis and that CDSs are involved in plant metabolic regulation.

plant biology↗

bioRxiv2025

SpaIM: Single-cell Spatial Transcriptomics Imputation via Style Transfer

Spatial transcriptomics (ST) technologies have revolutionized our understanding of cellular ecosystems. However, these technologies face challenges such as sparse gene signals and limited gene detection capacities, which hinder their ability to fully capture comprehensive spatial gene expression profiles. To address these limitations, we propose leveraging single-cell RNA sequencing (scRNA-seq), which provides comprehensive gene expression data but lacks spatial context, to enrich ST profiles. Herein, we introduce SpaIM, an innovative style transfer learning model that utilizes scRNA-seq information to predict unmeasured gene expressions in ST data, thereby improving gene coverage and expressions. SpaIM segregates scRNA-seq and ST data into data-agnostic contents and data-specific styles, with the contents capture the commonalities between the two data types, while the styles highlight their unique differences. By integrating the strengths of scRNA-seq and ST, SpaIM overcomes data sparsity and limited gene coverage issues, making significant advancements over 12 existing methods. This improvement is demonstrated across 53 diverse ST datasets, spanning sequencing- and imaging-based spatial technologies in various tissue types. Additionally, SpaIM enhances downstream analyses, including the detection of ligand-receptor interactions, spatial domain characterization, and identification of differentially expressed genes. Released as open-source software, SpaIM increases accessibility for spatial transcriptomics analysis. In summary, SpaIM represents a pioneering approach to enrich spatial transcriptomics using scRNA-seq data, enabling precise gene expression imputation and advancing the field of spatial transcriptomics research.

bioinformatics↗

bioRxiv2025

TransBrain: A computational framework for translating brain-wide phenotypes between humans and mice

Despite remarkable advances in whole-brain imaging technologies, the lack of quantitative approaches to bridge rodent preclinical and human studies remains a critical challenge. Here we present TransBrain, a computational framework enabling bidirectional translation of brain-wide phenotypes between humans and mice. TransBrain improves human-mouse homology mapping accuracy through: (1) a novel detached region-specific deep neural network trained on integrated multi-modal human transcriptomics to improve cortical correspondence (89.5% improvement over the original transcriptome), which revealed two evolutionarily conserved gradients explaining >50% of cortical organizational variance, and (2) random walk-based graph representation learning to construct a unified cross-species latent space incorporating anatomical hierarchies and structural connectivity. We demonstrated TransBrains utility through three cross-species applications: quantitative assessment of resting-state brain organizational features, inferring human cognitive functions from mouse optogenetic circuits, and translating molecular insights from mouse models to individual-level mechanisms in autism. TransBrain enables quantitative cross-species comparison and mechanistic investigation of both normal and pathological brain functions.

neuroscience↗

bioRxiv2025

Identification and Engineering of UDP-rhamnosyltransferase from Trillium tschonoskii for Heterologous Biosynthesis of Polyphyllin II in Engineered Yeast

AbstractPolyphyllins, a class of isospirostan-type steroidal saponins, exhibit potent cytotoxicity against a variety range of cancer cells. Although extensive research efforts have been made, the complete biosynthetic pathway of these compounds remains unclear. In order to elucidate these pathways, different tissues from Trillium tschonoskii were collected for sequencing that yielded 173,382 high-quality unigene sequences, among which 353 were annotated as glycosyltransferases. Then, a novel rhamnosyltransferase gene, UGT738A3, was characterized, which catalyzes the conversion of triglycoside polyphyllin III and pennogenin 3-O-beta-chacotrioside into tetraglycoside polyphyllin II and polyphyllin VII. The key residues that affect the catalytic activity of UGT738A3 were identified through site-directed mutation that the mutant A158T/P101L exhibited improved catalytic activity towards polyphyllin III and pennogenin 3-O-beta-chacotrioside by 2.5-fold and 6.5-fold respectively. We thus reconstructed the biosynthesis pathway of polyphyllin II in yeast by introducing the UGT93M3(catalyzing the formation of polyphyllin VI) and UGT738A3A158T/P101L3. This study not only elucidates the pivotal role of UGT738A3 in catalyzing the formation of tetraglycoside, but also provides highly efficient enzymatic components essential for the heterologous biosynthesis of polyphyllin saponins.

synthetic biology↗