Biology subjects
Publications and source records attributed to Cheng, X..
Withdrawal StatementThe authors have withdrawn this manuscript because [the results have been updated with new data and revised analyses]. Therefore, the authors do not wish this work to be cited as reference for the project. If you have any questions, please contact the corresponding author.
The O-GlcNAc transferase (OGT) is an essential enzyme that mediates protein O-GlcNAcylation, a unique form of posttranslational modification of many nuclear and cytosolic proteins. Recent studies observed increased OGT and O-GlcNAcylation levels in a broad range of human cancer tissues compared to adjacent normal tissues, indicating a universal effect of OGT in promoting tumorigenesis. Here, we show that OGT is essential for tumor growth in immunocompetent hosts by repressing the cyclic GMP-AMP synthase (cGAS)-dependent DNA sensing pathway. We found that deletion of OGT (Ogt-/-) caused a marked reduction in tumor growth in both syngeneic tumor models and a genetic colorectal cancer (CRC) model induced by mutation of the Apc gene (Apcmin). Pharmacological inhibition or genetic deletion of OGT induced a robust genomic instability (GIN), leading to cGAS-dependent production of the type I interferon (IFN-I) and IFN-stimulated genes (ISGs). As a result, deletion of Cgas or Sting from Ogt-/- cancer cells restored tumor growth, and this correlated with impaired CD8+ T cell-mediated antitumor immunity. Mechanistically, we found that OGT-dependent cleavage of host cell factor C1 (HCF-1) is required for the avoidance of GIN and IFN-I production in tumors. In summary, our results identify OGT-mediated genomic stability and activate cGAS-STING pathway as an important tumor cell-intrinsic mechanism to repress antitumor immunity.
CDCA7, encoding a protein with a C-terminal cysteine-rich domain (CRD), is mutated in immunodeficiency, centromeric instability and facial anomalies (ICF) syndrome, a disease related to hypomethylation of juxtacentromeric satellite DNA. How CDCA7 directs DNA methylation to juxtacentromeric regions is unknown. Here, we show that the CDCA7 CRD adopts a unique zinc-binding structure that recognizes a CpG dyad in a non-B DNA formed by two sequence motifs. CDCA7, but not ICF mutants, preferentially binds the non-B DNA with strand-specific CpG hemi-methylation. The unmethylated sequence motif is highly enriched at centromeres of human chromosomes, whereas the methylated motif is distributed throughout the genome. At S phase, CDCA7, but not ICF mutants, is concentrated in constitutive heterochromatin foci, and the formation of such foci can be inhibited by exogenous hemi-methylated non-B DNA bound by the CRD. Binding of the non-B DNA formed in juxtacentromeric regions during DNA replication provides a mechanism by which CDCA7 controls the specificity of DNA methylation.
Identifying accurate cell markers in single-cell RNA-seq data is crucial for understanding cellular diversity and function. Localized Marker Detector (LMD) is a novel tool to identify "localized genes" - genes exclusively expressed in groups of highly similar cells - thereby characterizing cellular diversity in a multi-resolution and fine-grained manner. LMD constructs a cell-cell affinity graph, diffuses the gene expression value across the cell graph, and assigns a score to each gene based on its diffusion dynamics. LMDs candidate markers can be grouped into functional gene modules, which accurately reflect cell types, subtypes, and other sources of variation such as cell cycle status. We apply LMD to mouse bone marrow and hair follicle dermal condensate datasets, where LMD facilitates cross-sample comparisons, identifying shared and sample-specific gene signatures and novel cell populations without requiring batch effect correction or integration methods. Furthermore, we assessed the performance of LMD across nine single-cell RNA sequencing datasets, compared it with six other methods aimed at achieving similar objectives, and found that LMD outperforms the other methods evaluated.
Single-cell sequencing has revolutionized the scale and resolution of molecular profiling of tissues and organs. Here, we present an integrated multimodal reference atlas of the most accessible portion of the mammalian central nervous system, the retina. We compiled around 2.4 million cells from 55 donors, including 1.4 million unpublished data points, to create a comprehensive human retina cell atlas (HRCA) of transcriptome and chromatin accessibility, unveiling over 110 types. Engaging the retina community, we annotated each cluster, refined the Cell Ontology for the retina, identified distinct marker genes, and characterized cis-regulatory elements and gene regulatory networks (GRNs) for these cell types. Our analysis uncovered intriguing differences in transcriptome, chromatin, and GRNs across cell types. In addition, we modeled changes in gene expression and chromatin openness across gender and age. This integrated atlas also enabled the fine-mapping of GWAS and eQTL variants. Accessible through interactive browsers, this multimodal cross-donor and cross-lab HRCA, can facilitate a better understanding of retinal function and pathology.
Understanding diverse responses of individual cells to the same perturbation is central to many biological and biomedical problems. Current methods, however, do not precisely quantify the strength of perturbation responses and, more importantly, reveal new biological insights from heterogeneity in responses. Here we introduce the perturbation-response score (PS), based on constrained quadratic optimization, to quantify diverse perturbation responses at a single-cell level. Applied to single-cell transcriptomes of large-scale genetic perturbation datasets (e.g., Perturb-seq), PS outperforms existing methods for quantifying partial gene perturbation responses. In addition, PS presents two major advances. First, PS enables large-scale, single-cell-resolution dosage analysis of perturbation, without the need to titrate perturbation strength. By analyzing the dose-response patterns of over 2,000 essential genes in Perturb-seq, we identify two distinct patterns, depending on whether a moderate reduction in their expression induces strong downstream expression alterations. Second, PS identifies intrinsic and extrinsic biological determinants of perturbation responses. We demonstrate the application of PS in contexts such as T cell stimulation, latent HIV-1 expression, and pancreatic cell differentiation. Notably, PS unveiled a previously unrecognized, cell-type-specific role of coiled-coil domain containing 6 (CCDC6) in guiding liver and pancreatic lineage decisions, where CCDC6 knockouts drive the endoderm cell differentiation towards liver lineage, rather than pancreatic lineage. The PS approach provides an innovative method for dose-to-function analysis and will enable new biological discoveries from single-cell perturbation datasets. One sentence summaryWe present a method to quantify diverse perturbation responses and discover novel biological insights in single-cell perturbation datasets.
Almost all Glioblastoma (GBM) are either intrinsically resistant to the chemotherapeutical drug temozolomide (TMZ) or acquire therapy-induced mutations that cause chemoresistance and recurrence. The genome maintenance mechanisms responsible for GBM chemoresistance and hypermutation are unknown. We show that the E3 ubiquitin ligase RAD18 (a proximal regulator of TLS) is activated in a Mismatch repair (MMR)-dependent manner in TMZ-treated GBM cells, promoting post-replicative gap-filling and survival. An unbiased CRISPR screen provides a new aerial map of RAD18-interacting DNA damage response (DDR) pathways deployed by GBM to tolerate TMZ genotoxicity. Analysis of mutation signatures from TMZ-treated GBM reveals a role for RAD18 in error-free bypass of O6mG (the most toxic TMZ-induced lesion), and error-prone bypass of other TMZ-induced lesions. Our analyses of recurrent GBM patient samples establishes a correlation between low RAD18 expression and hypermutation. Taken together we define novel molecular underpinnings for the hallmark tumorigenic phenotypes of TMZ-treated GBM.
Whole-genome time-series allele frequency data are becoming more prevalent as ancient DNA (aDNA) sequences and data from evolve-and-resequence (E&R) experiments are generated at a rapid pace. Such data presents unprecedented opportunities to elucidate the dynamics of genetic variation under selection. However, despite many methods to infer parameters of selection models from allele frequency trajectories available in the literature, few provide user-friendly implementations for large-scale empirical applications. Here, we present diplo-locus, an open-source Python package that provides functionality to simulate and perform inference from time-series data under the Wright-Fisher diffusion with general diploid selection. The package includes Python modules as well as command-line tools and is available at: https://github.com/steinrue/diplo_locus.
Bulk deconvolution with single-cell/nucleus RNA-seq data is critical for understanding heterogeneity in complex biological samples, yet the technological discrepancy across sequencing platforms limits deconvolution accuracy. To address this, we introduce an experimental design to match inter-platform biological signals, hence revealing the technological discrepancy, and then develop a deconvolution framework called DeMixSC using the better-matched, i.e., benchmark, data. Built upon a novel weighted nonnegative least-squares framework, DeMixSC identifies and adjusts genes with high technological discrepancy and aligns the benchmark data with large patient cohorts of matched-tissue-type for large-scale deconvolution. Our results using a benchmark dataset of healthy retinas suggest much-improved deconvolution accuracy. Further analysis of a cohort of 453 patients with age-related macular degeneration supports the broad applicability of DeMixSC. Our findings reveal the impact of technological discrepancy on deconvolution performance and underscore the importance of a well-matched dataset to resolve this challenge. The developed DeMixSC framework is generally applicable for deconvolving large cohorts of disease tissues, and potentially cancer.
For viruses in the family Potyviridae (potyvirids), three virus-encoded proteins (P3N-PIPO, CI and CP) and several host components are known to coordinately regulate viral cell-to-cell movement. Here, we found that HCPro2 encoded by areca palm necrotic ring spot virus is involved in the intercellular movement, which could be functionally complemented by its counterpart HCPro from a potyvirus. The affinity purification and mass spectrum analysis identified several viral factors (including CI and CP) and a variety of host proteins that physically associate with HCPro2. We demonstrated that HCPro2 interacts with either CI or CP in planta, and the three form plasmodesmata (PD)-localized interactive complex in viral infection. Further, we screened HCPro2-associating host proteins, and identified a common host protein RbCS that mediates the interactions of HCPro2-CI, HCPro2-CP and CI-CP among the complex. Knockdown of NbRbCS simultaneously impairs the interactions of HCPro2-CI, HCPro2-CP and CI-CP, and significantly attenuates the intercellular movement and systemic infection for ANRSV and other three tested potyvirids. This study highlights that a nucleus-encoded chloroplast-targeted protein is hijacked by potyvirids as the scaffold protein to mediate the assembly of viral intercellular movement complex to promote viral infection.
AbstractUnderstanding how organisms adapt to unpredictable future environments is a fundamental goal in biology, which becomes even more urgent in an era of rapid climate change. One evolutionary adaptation to randomly fluctuating environments is bet hedging, a strategy that successfully facilitates reproduction and population persistence and has been widely reported from microbes to humans. Empirical evidence for its presence in microalga, one of Earths most important primary producers and carbon sinks, is lacking. Here, we report a bet-hedging strategy in the unicellular microalga Haematococcus pluvialis. In a series of experiments, we show that an isogenic H. pluvialis population reversibly diversifies into hetero-phenotypic mobile and non-mobile subunits, independent of environmental conditions. Mobile cells grow faster but are more susceptible to external stressors, while non-mobile cells hardly grow but are more stress-resistant. This is attributed to dramatic shifts from growth-promoting activities (cell division, photosynthesis) to resilience-promoting cellular metabolic processes, including cell enlargement and aggregation, and accumulation of antioxidant and energy-storaging compounds. Our results provide experimental evidence for bet hedging in microalga, which has implications for their potential to adapt to current and predicted future conditions, and thus for the conservation of ecosystem functions.
Neuroendocrine carcinoma of the cervix (NECC) is a rare subtype of cervical cancer with rapid metastasis. Lack of treatment regimens for metastatic NECC highlights the need for preclinical models and personalized treatment. Here, we established two patient-derived NECC organoid lines from lung metastases from the same patient, one of which was drug-free and the other exhibited tolerance to carboplatin plus etoposide (EP). The two organoids recapitulated the histopathology and genomic spectrum of the original tumors. Whole-genome sequencing revealed genomic structural variations (SVs), including HPV16 infection, NEUROD2, and ERBB2 amplification. The sensitivity phenotype of organoids to ERBB2 inhibitors coincides with the genomic features of ERBB2 amplification. In addition, the patient after relapse responded well to the paclitaxel plus carboplatin (TC) regimen as recommended by organoid drug screening, underscoring the utility of organoids as preclinical models. We first provide two NECC organoids of lung metastases with comprehensive molecular characteristics as valuable resources for this rare cancer and give a typical application for personalized treatment based on organoids.
The study of transcription factors that determine specialised neuronal functions has provided invaluable insights into the physiology of the nervous system. Peripheral chemoreceptors are neurone-like electro-physiologically excitable cells that link the oxygen content of arterial blood to the neuronal control of breathing. In the adult, this oxygen chemosensitivity is exemplified by the Type I cells of the carotid body and recent work has revealed one isoform of the transcription factor HIF, HIF-2, to have a non-redundant role in the development and function of that organ. Here we show that the activation of HIF-2, including isolated overexpression alone, is sufficient to induce oxygen chemosensitivity in the otherwise unresponsive adult adrenal medulla. This phenotypic change in the adrenal medulla was associated with retention of extra-adrenal paraganglioma-like tissues that resemble the foetal organ of Zuckerkandl and also manifest oxygen chemosensitivity. Acquisition of chemosensitivity was associated with changes in the adrenal medullary expression of classes of genes that are ordinarily characteristic of the carotid body, including G-protein regulators and atypical subunits of mitochondrial cytochrome oxidase. Overall, the findings suggest that, at least in certain tissues, HIF-2 acts as a phenotypic driver for cells that display oxygen chemosensitivity, providing a route to mechanistic understanding.
In addition to amide hydrogen bonds and the hydrophobic effect, interactions involving {pi}-bonded sp2 atoms of amides, aromatics and other groups occur in protein self-assembly processes including folding, oligomerization and condensate formation. These interactions also occur in aqueous solutions of amide and aromatic compounds, where they can be quantified. Previous analysis of thermodynamic coefficients quantifying net-favorable interactions of amide compounds with other amides and aromatics revealed that interactions of amide sp2O with amide sp2N unified atoms (presumably C=O{middle dot}{middle dot}{middle dot}H-N hydrogen bonds) and amide/aromatic sp2C (lone pair-{pi}, n-{pi}*) are particularly favorable. Sp3C-sp3C (hydrophobic), sp3C-sp2C (hydrophobic, CH-{pi}), sp2C-sp2C (hydrophobic, {pi}-{pi}) and sp3C-sp2N interactions are favorable, sp2C-sp2N interactions are neutral, while sp2O-sp2O and sp2N-sp2N self-interactions and sp2O-sp3C interactions are unfavorable. Here, from determinations of favorable effects of fourteen amides on naphthalene solubility at 10, 25 and 45 {degrees}C, we dissect amide-aromatic interaction free energies into enthalpic and entropic contributions and find these vary systematically with amide composition. Analysis of these results yields enthalpic and entropic contributions to intrinsic strengths of interactions of amide sp2O, sp2N, sp2C and sp3C unified atoms with aromatic sp2C atoms. For each interaction, enthalpic and entropic contributions have the same sign and are much larger in magnitude than the interaction free energy itself. The amide sp2O-aromatic sp2C interaction is enthalpy-driven and entropically unfavorable, consistent with direct chemical interaction (e.g. lone pair-{pi}) while amide sp3C- and sp2C-aromatic sp2C interactions are entropy-driven and enthalpically unfavorable, consistent with hydrophobic effects. These findings are relevant for interactions involving {pi}-bonded sp2 atoms in protein processes. Table of Contents Graphic O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=109 SRC="FIGDIR/small/548600v2_ufig1.gif" ALT="Figure 1"> View larger version (36K): org.highwire.dtl.DTLVardef@1683bb4org.highwire.dtl.DTLVardef@27db75org.highwire.dtl.DTLVardef@b915ccorg.highwire.dtl.DTLVardef@1461b04_HPS_FORMAT_FIGEXP M_FIG C_FIG
Homeostatic plasticity stabilizes firing rates of neurons, but the pressure to restore low activity rates can significantly alter synaptic and cellular properties. Most previous studies of homeostatic readjustment to complete activity silencing in rodent forebrain have examined changes after two days of deprivation, but it is known that longer periods of deprivation can produce adverse effects. To better understand the mechanisms underlying these effects and to address how presynaptic as well as postsynaptic compartments change during homeostatic plasticity, we subjected mouse cortical slice cultures to a more severe five-day deprivation paradigm. We developed and validated a computational framework to measure the number and morphology of presynaptic and postsynaptic compartments from super resolution light microscopy images of dense cortical tissue. Using these tools, combined with electrophysiological miniature excitatory postsynaptic current measurements, and synaptic imaging at the electron microscopy level, we assessed the functional and morphological results of prolonged deprivation. Excitatory synapses were strengthened both presynaptically and postsynaptically. Surprisingly, we also observed a decrement in the density of excitatory synapses, both as measured from colocalized staining of pre- and postsynaptic proteins in tissue, and from the number of dendritic spines. Overall, our results suggest that cortical networks deprived of activity progressively move towards a smaller population of stronger synapses. SIGNIFICANCE STATEMENTBlocking activity in neocortical slice cultures produced coordinated pre and postsynaptic changes at excitatory synapses. Functional and structural assays suggest that deprivation results in fewer excitatory synapses, but each is strengthened both pre- and postsynaptically. This may contribute to the emergence of epileptiform activity.
Protein language models have shown remarkable success in learning biological information from protein sequences. However, most existing models are limited by either autoencoding or autoregressive pre-training objectives, which makes them struggle to handle protein understanding and generation tasks concurrently. We propose a unified protein language model, xTrimoPGLM, to address these two types of tasks simultaneously through an innovative pre-training framework. Our key technical contribution is an exploration of the compatibility and the potential for joint optimization of the two types of objectives, which has led to a strategy for training xTrimoPGLM at an unprecedented scale of 100 billion parameters and 1 trillion training tokens. Our extensive experiments reveal that 1) xTrimoPGLM significantly outperforms other advanced baselines in 18 protein understanding benchmarks across four categories. The model also facilitates an atomic-resolution view of protein structures, leading to an advanced 3D structural prediction model that surpasses existing language model-based tools. 2) xTrimoPGLM not only can generate de novo protein sequences following the principles of natural ones, but also can perform programmable generation after supervised fine-tuning (SFT) on curated sequences. These results highlight the substantial capability and versatility of xTrimoPGLM in understanding and generating protein sequences, contributing to the evolving landscape of foundation models in protein science. Trained weight for the xTrimoPGLM model, and downstream datasets are available at https://huggingface.co/proteinglm.
The recent outbreak of Corona Virus Disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has been a severe threat to the global public health and economy, however, effective drugs to treat COVID-19 are still lacking. Here, we employ a deep learning-based drug repositioning strategy to systematically screen potential anti-SARS-CoV-2 drug candidates that target the cell entry mechanism of SARS-CoV-2 virus from 2,635 FDA-approved drugs and 1,062 active ingredients from Traditional Chinese Medicine herbs. In silico molecular docking analysis validates the interactions between the top compounds and host receptors or viral spike proteins. Using a SARS-CoV-2 pseudovirus system, we further identify several drug candidates including Fostamatinib, Linagliptin, Lysergol and Sophoridine that can effectively block the cell entry of SARS-CoV-2 variants into human lung cells even at a nanomolar scale. These efforts not only illuminate the feasibility of applying deep learning-based drug repositioning for antiviral agents by targeting a specified mechanism, but also provide a valuable resource of promising drug candidates or lead compounds to treat COVID-19.
Large-scale pretrained models have become foundation models leading to breakthroughs in natural language processing and related fields. Developing foundation models in life science for deciphering the "languages" of cells and facilitating biomedical research is promising yet challenging. We developed a large-scale pretrained model scFoundation with 100M parameters for this purpose. scFoundation was trained on over 50 million human single-cell transcriptomics data, which contain high-throughput observations on the complex molecular features in all known types of cells. scFoundation is currently the largest model in terms of the size of trainable parameters, dimensionality of genes and the number of cells used in the pre-training. Experiments showed that scFoundation can serve as a foundation model for single-cell transcriptomics and achieve state-of-the-art performances in a diverse array of downstream tasks, such as gene expression enhancement, tissue drug response prediction, single-cell drug response classification, and single-cell perturbation prediction.