Biology subjects

Tong, A.

Publications and source records attributed to Tong, A..

16 recordsLinked to original sources

bioRxiv2026

Release of the Staphylococcus aureus autoinducing peptide is enabled by active export of the biosynthetic intermediate

Staphylococcus aureus is a species of Gram-positive bacteria that uses the accessory gene regulator (agr) quorum sensing system to regulate many aspects of cell behavior such as virulence and toxin production. Quorum sensing is activated in dense S. aureus populations by the extracellular accumulation of a signaling molecule called the autoinducing peptide (AIP). While the biosynthetic enzymes that produce AIP have been elucidated in certain subgroups of S. aureus, the mechanism of AIP secretion is unknown. Here, we show that the biosynthetic intermediate known as AgrD (1-32) thiolactone is actively exported into the surrounding environment in certain S. aureus subgroups. Using a high-throughput transposon mutagenesis screen, we identify the integral membrane transporters, AbcA and PmtCD, as being involved in the biosynthesis of certain AIPs. Subsequent genetic complementation and biochemical experiments reveal that these proteins actively transport AgrD (1-32) thiolactone out of the cell. This enables the production of AIP directly on the cell surface following the action of MroQ, an integral membrane protease demonstrated here to possess an active site facing the extracellular milieu. Furthermore, we show that media supplementation with catalytically-active MroQ nanodiscs rescues AIP production in an mroQ knockout strain. This observation suggests the possibility that some proportion of AgrD (1-32) thiolactone exported by one cell may travel to neighboring cell surfaces for further processing into the diffusive AIP molecule. Collectively, our studies shed light on a poorly understood step in the agr quorum sensing pathway with clear implications for the design of novel therapeutic strategies.

biochemistry↗

bioRxiv2026

Reframing enzyme function prediction as conditional generation

Enzymes frequently exhibit promiscuous activity beyond their native roles, providing starting-points for new functions. Finding these promiscuous enzymes, especially for non-native chemical transformations, is challenging but highly valuable, as they promise novel, sustainable solutions for chemistry and biotechnology. However, current machine learning methods are poorly suited to discovering unseen chemistry as they often frame function prediction as closed set classification or a retrieval task. Here, we present Fluxion, a generative deep learning framework that learns enzymatic catalysis by modeling dynamic electron flow trajectories across the enzyme's catalytic residues. By combining both synthetic chemistry and biochemical datasets with protein language model representations, Fluxion generates multi-step electron-flow trajectories analogous to the arrow-pushing representations used to describe enzyme reaction mechanisms. Generation is conditioned on enzyme context, including the enzyme sequence, catalytic residues, substrates, and cofactors. We show that this conditioning allows Fluxion to learn enzyme-dependent regioselectivity across cytochrome P450 enzymes with different sequences shifting the predicted reaction sites for the same substrate. We then demonstrate that Fluxion's embeddings are useful for downstream tasks, such as specificity prediction on two experimental datasets, with and without finetuning. Finally, we show that Fluxion has the potential to transfer synthetic chemical logic to biology; it can generate the observed non-native product from real-world non-native directed evolution screens. Our results establish a proof of concept that generative modeling through mechanistic representations of enzymes can shift enzyme function prediction beyond static database retrieval and closed set classification to function generation. This conceptual framework provides a stepping stone towards an in silico generative method to discover non-native biocatalysts.

biochemistry↗

bioRxiv2026

Nutrient microenvironments reprogram RPE metabolism

Induced pluripotent stem cell-derived retinal pigment epithelium (iPSC RPE) has become a widely used model for studying the mechanisms of age-related macular degeneration (AMD). However, the nutrient composition of currently used RPE culture media is highly variable, posing a major challenge to reproducibility in RPE metabolism and phenotype. We systematically investigate how six distinct nutrient microenvironments shape RPE phenotype, function and metabolism in both iPSC RPE and fetal RPE (fRPE). These included MEM, DMEM-HG/F12 basal media, physiological human plasma-like medium (HPLM) supplemented with FBS or B27, and X-VIVO 10. Although canonical RPE markers were expressed across all conditions, B27 supplementation and X-VIVO 10 increased RPE cell size, hexagonality, and transepithelial resistance. Culture in HPLM+FBS induced accumulation of lipid droplets and sub-RPE deposits, whereas X-VIVO 10 resulted in the formation of large intracellular vacuoles. B27 supplementation enhanced basal respiration, while X-VIVO 10 increased glycolytic capacity. Amino acid consumption was broadly conserved across media types, including complete depletion of proline in all conditions by 48 hours; however, lipid and nucleotide metabolism varied substantially between conditions. Notably, B27 supplementation in specific media types reversed the net direction of several metabolites, with creatine, serine and taurine shifting from consumption to production, while riboflavin and guanine shifted from production to consumption. These findings establish the nutrient environment as a key determinant of RPE phenotype, function and metabolism. Our work provides a valuable resource for media selection and interpretation of cellular and metabolic phenotypes relevant to RPE disease modeling.

cell biology↗

bioRxiv2025

Hidden sampling biases inflate performance in gene regulatory network inference

Accurate reconstruction of gene regulatory networks (GRNs) from single-cell transcriptomic data remains a major methodological challenge. Recent machine learning approaches, particularly graph neural networks and graph autoencoders, have reported improved performance, yet these gains do not consistently translate to realistic biological settings. Here, we show that a key reason for that is the way negative regulatory interactions are sampled for supervised training and evaluation. We find that widely used sampling strategies introduce node-degree biases that allow models to exploit trivial graph-structural cues rather than biological signals. Across multiple benchmarks, simple degree-based heuristics match or exceed state-of-the-art graph neural network models under these biased evaluation protocols. We further introduce a degree-aware sampling approach that eliminates these artifacts and provides more reliable assessments of GRN inference methods. Our results call for standardized, bias-aware benchmarking practices to ensure meaningful progress in supervised GRN inference from single-cell RNA-seq data.

bioinformatics↗

bioRxiv2025

Quantitative analysis of genetic interactions in human cells from genome-wide CRISPR-Cas9 screens

Genetic interaction (GI) networks in model organisms have revealed how combinations of genome variants can impact phenotypes. To advance efforts toward a reference human GI network, we developed the quantitative Genetic Interaction (qGI) score, a method for precise GI measurement from genome-wide CRISPR-Cas9 screens in different query mutants constructed in a single human cell line. We found surprising prevalent systematic variation unrelated to GIs in CRISPR screen data, including both genomically linked effects and functionally coherent covariation. Leveraging [~]40 control screens in wild-type cells and half a billion differential fitness effect measurements, we developed a pipeline for CRISPR screen data processing and normalization to correct these artifacts and measure accurate, quantitative GIs. We also comprehensively characterized GI reproducibility by characterizing 4 - 5 biological replicates for [~]125,000 unique gene pairs. The qGI framework enables systematic identification of human GIs and provides broadly applicable strategies for analyzing context-specific CRISPR screen data.

bioinformatics↗

bioRxiv2025

geneRNIB: a living benchmark for gene regulatory network inference

Gene regulatory networks (GRNs) underpin cellular identity and function, playing a key role in health and disease. GRN inference has received substantial attention, motivating systematic benchmarking. Despite various benchmarking efforts, existing studies remain limited in the number of methods, datasets, and metrics, fail to capture the context-specific nature of regulatory interactions across biological conditions, and are constrained by the absence of a reliable ground truth. Here, we introduce geneRNIB, a comprehensive GRN inference benchmarking framework built on three key principles: continuous integration, context-specific evaluation, and holistic assessment in the absence of a true reference network. geneRNIB enables the seamless incorporation of new algorithms, datasets, and evaluation metrics to reflect ongoing developments. In the current version, we systematically integrated and assessed 12 GRN inference methods, spanning single- and multiomics approaches across 11 datasets including thousands of perturbation scenarios. We introduced complementary metrics specifically designed to assess context-specific inference. Our findings indicate that simple models with fewer assumptions often outperform more complex pipelines across several perturbation-informed and predictive metrics. Notably, gene expression-based algorithms yielded better results than more advanced multimodal approaches. In addition, we identify several potential factors that influence the performance of GRN inference and offer actionable guidelines for the future development of the method. By addressing these critical limitations in existing benchmarks, geneRNIB advances GRN inference research and fosters progress toward personalized medicine.

bioinformatics↗

bioRxiv2024

Beta cells are essential drivers of pancreatic ductal adenocarcinoma development

Pancreatic endocrine-exocrine crosstalk plays a key role in normal physiology and disease. For instance, endocrine islet beta ({beta}) cell secretion of insulin or cholecystokinin (CCK) promotes progression of pancreatic adenocarcinoma (PDAC), an exocrine cell-derived tumor. However, the cellular and molecular mechanisms that govern endocrine-exocrine signaling in tumorigenesis remain incompletely understood. We find that {beta} cell ablation impedes PDAC development in mice, arguing that the endocrine pancreas is critical for exocrine tumorigenesis. Conversely, obesity induces {beta} cell hormone dysregulation, alters CCK-dependent peri-islet exocrine cell transcriptional states, and enhances islet proximal tumor formation. Single-cell RNA-sequencing, in silico latent-space archetypal and trajectory analysis, and genetic lineage tracing in vivo reveal that obesity stimulates postnatal immature {beta} cell expansion and adaptation towards a pro-tumorigenic CCK+ state via JNK/cJun stress-responsive signaling. These results define endocrine-exocrine signaling as a driver of PDAC development and uncover new avenues to target the endocrine pancreas to subvert exocrine tumorigenesis.

cancer biology↗

bioRxiv2023

Learning transcriptional and regulatory dynamics driving cancer cell plasticity using neural ODE-based optimal transport

While single-cell technologies provide snapshots of tumor states, building continuous trajectories and uncovering causative gene regulatory networks remains a significant challenge. We present Cflows, an AI framework that combines neural ODE networks with Granger causality to infer continuous cell state transitions and gene regulatory interactions from static scRNA-seq data. In a new 5-time point dataset capturing tumorsphere development over 30 days, Cflows reconstructs two types of trajectories leading to tumorsphere formation or apoptosis. Trajectory-based cell-of-origin analysis delineated a novel cancer stem cell profile characterized by CD44hiEPCAM+CAV1+, and uncovered a cell cycle-dependent enrichment of tumorsphere-initiating potential in G2/M or S-phase cells. Cflows uncovers ESRRA as a crucial causal driver of the tumor-forming gene regulatory network. Indeed, ESRRA inhibition significantly reduces tumor growth and metastasis in vivo. Cflows offers a powerful framework for uncovering cellular transitions and dynamic regulatory networks from static single-cell data.

cancer biology↗

bioRxiv2022

Cancer-Associated Fibroblasts Regulate Patient-Derived Organoid Drug Responses

Patient-derived organoids (PDOs) can model personalized therapy responses, however current screening technologies cannot reveal drug response mechanisms or study how tumor microenvironment cells alter therapeutic performance. To address this, we developed a highly-multiplexed mass cytometry platform to measure post translational modification (PTM) signaling in >2,500 colorectal cancer (CRC) PDOs and cancer-associated fibroblasts (CAFs) in response to clinical therapies at single-cell resolution. To compare patient- and microenvironment-specific drug responses in thousands of single-cell datasets, we developed Trellis -- a highly-scalable, hierarchical tree-based treatment effect analysis method. Trellis single-cell screening revealed that on-target cell-cycle blockage and DNA-damage drug effects are common, even in chemorefractory PDOs. However, drug-induced apoptosis is patient-specific. We found drug-induced apoptosis does not correlate with genotype or clinical staging but does align with cell-intrinsic PTM signaling in PDOs. CAFs protect chemosensitive PDOs by shifting cancer cells into a slow-cycling cell-state and CAF chemoprotection can be reversed by inhibiting YAP. HighlightsO_LI>2,500 single-cell PTM signaling, DNA-damage, cell-cycle, and apoptosis responses from drug-treated PDOs and CAFs. C_LIO_LITrellis: hierarchical tree-based treatment effect method for single-cell screening analysis. C_LIO_LIPDOs have patient-specific drug responses that align with cell-intrinsic PTM signaling states. C_LIO_LICAFs chemoprotect PDOs by altering PDO cell-state via YAP signaling. C_LI

cancer biology↗

bioRxiv2020

FAM72A antagonizes UNG2 to promote mutagenic uracil repair during antibody maturation

Activation-induced cytidine deaminase (AID) catalyzes the deamination of deoxycytidines within Immunoglobulin (Ig) genes to induce somatic hypermutation (SHM) and class switch recombination (CSR) 1,2. AID-induced deoxyuracils within Ig loci are recognized and processed by subverted base excision and mismatch repair pathways that ensure a mutagenic outcome in B lymphocytes 3-8. However, it is unclear why DNA repair pathways that remove deoxyuracil from DNA are not efficient at faithfully repairing AID-induced lesions. Here, we identified through a genome-wide CRISPR screen that FAM72A, a protein with no ascribed function, is a major determinant for the error-prone processing of deoxyuracil. Fam72a-deficient CH12F3-2 B cells and primary B cells from Fam72a-/- mice have reduced CSR and SHM frequencies. The SHM spectrum in B cells from Fam72a-/- mice is opposite to that observed in Ung2-/- mice 9, suggesting that UNG2 is hyperactive in FAM72A-deficient cells. Indeed, FAM72A binds to UNG2 resulting in reduced UNG2 activity, and significantly reduced protein levels in the G1 phase, coinciding with peak AID activity. This effect leads to increased genome-wide deoxyuracils in B cells. By antagonizing UNG2, FAM72A therefore increases U*G mispairs that engage mutagenic mismatch repair promoting error-prone processing of AID-induced deoxyuracils. This work shows that FAM72A bridges base-excision repair and mismatch repair to modulate antibody maturation.

immunology↗

bioRxiv2020

Multiscale PHATE Exploration of SARS-CoV-2 Data Reveals Multimodal Signatures of Disease

1The biomedical community is producing increasingly high dimensional datasets, integrated from hundreds of patient samples, which current computational techniques struggle to explore. To uncover biological meaning from these complex datasets, we present an approach called Multiscale PHATE, which learns abstracted biological features from data that can be directly predictive of disease. Built on a continuous coarse graining process called diffusion condensation, Multiscale PHATE creates a tree of data granularities that can be cut at coarse levels for high level summarizations of data, as well as at fine levels for detailed representations on subsets. We apply Multiscale PHATE to study the immune response to COVID-19 in 54 million cells from 168 hospitalized patients. Through our analysis of patient samples, we identify CD16hi CD66blo neutrophil and IFN{gamma}+GranzymeB+ Th17 cell responses enriched in patients who die. Further, we show that population groupings Multiscale PHATE discovers can be directly fed into a classifier to predict disease outcome. We also use Multiscale PHATE-derived features to construct two different manifolds of patients, one from abstracted flow cytometry features and another directly on patient clinical features, both associating immune subsets and clinical markers with outcome.

bioinformatics↗

bioRxiv2020

A DNA translocase operates by cycling between planarand lock-washer structures

Ring ATPases that translocate disordered polymers possess lock-washer architectures that they impose on their substrates during transport via a hand-over-hand mechanism. Here, we investigate the operation of ring motors that transport substrates possessing a preexisting helical structure, such as the bacteriophage {phi}29 dsDNA packaging motor. During each cycle, this pentameric motor tracks one helix strand (the tracking strand), and alternates between two segregated phases: a dwell in which it exchanges ADP for ATP and a burst in which it packages a full turn of DNA in four steps. We challenge this motor with DNA-RNA hybrids and dsRNA substrates and find that it adapts the size of its burst to the corresponding shorter helical pitches by keeping three of its power strokes invariant while shortening the fourth. Intermittently, the motor loses grip when the tracking strand is RNA, indicating that it makes load-bearing contacts with the substrate that are optimal with dsDNA. The motor possesses weaker grip when ADP-bound at the end of the burst. To rationalize all these observations, we propose a helical inchworm translocation mechanism in which the motor increasingly adopts a lock-washer structure during the ATP loading dwell and successively regains its planar form with each power stroke during the burst.

biophysics↗

bioRxiv2019

Direct Binding of Phosphatidylglycerol at Specific Sites Modulates Desensitization of a Pentameric Ligand-Gated Ion Channel

Pentameric ligand-gated ion channels (pLGICs) are essential determinants of synaptic transmission, and are modulated by specific lipids including anionic phospholipids. The exact modulatory effect of anionic phospholipids in pLGICs and the mechanism of this effect are not well understood. Using native mass spectrometry, coarse-grained molecular dynamics simulations and functional assays, we show that the anionic phospholipid, 1-palmitoyl-2-oleoyl-phosphatidylglycerol (POPG), preferentially binds to and stabilizes the pLGIC, Erwinia ligand-gated ion channel (ELIC), and decreases ELIC desensitization. Mutations of five arginines located in the interfacial regions of the transmembrane domain (TMD) reduce POPG binding, and a subset of these mutations increase ELIC desensitization. In contrast, the L240A mutant known to decrease ELIC desensitization, increases POPG binding. The results support a mechanism by which POPG stabilizes the open state of ELIC relative to the desensitized state by direct binding at specific sites.

biophysics↗

bioRxiv2019

An Addressable Microfluidics device for Noninvasive Manipulation of Cells and Chemicals within Live Cultures

According to the U.S. Department of Health & Human Services, nearly 115,000 people in the U.S needed a lifesaving organ transplant in 2018, while only [~]10% of them have received it. Yet, almost no artificial FDA-approved products are commercially available today - three decades after the inception of tissue engineering. It is hypothesized here that the major bottlenecks restricting its progress stem from lack of access to the inner pore space of the scaffolds. Specifically, the inability to deliver nutrients to, and clear waste from, the center of the scaffolds limits the size of the products that can be cultured. Likewise, the inability to monitor, and control, the cells after seeding them into the scaffold results in nonviable tissue, with an unacceptable product variability. To resolve these bottlenecks, we present a prototype addressable microfluidics device capable of minimally disruptive fluid and cell manipulations within living cultures. As proof-of-concept, we demonstrate its ability to perform additive manufacturing by seeding cells in spatial patterns (including co-culturing multiple cell types); and subtractive manufacturing by removing surface adherent cells via focused flow of trypsin. Additionally, we show that the device can sample fluids and perform cell "biopsies" (which can be subsequently sent for ex-situ analysis), from any location within its Culture Chamber. Finally, the on-chip plumbing is completely automated using external electronics. This opens the possibility to perform long-term computer-driven tissue engineering experiments, where the cell behavior is modulated in response to the minimally disruptive observations (e.g. fluid sampling and cell biopsies) throughout the entire duration of the cultures. It is expected that the proof-of-concept technology will eventually be scaled up to 3D addressable microfluidic scaffolds, capable of overcoming the limitations bottlenecking the transition of tissue engineering technologies to the clinical setting.

bioengineering↗

bioRxiv2017

Predator-secreted sulfolipids induce fear-like defense responses in C. elegans

Animals respond to predators by altering their behavior and physiological states, but the underlying signaling mechanisms are poorly understood. Using the interactions between Caenorhabditis elegans and its predator, Pristionchus pacificus, we show that neuronal perception by C. elegans of a predator-specific molecular signature induces instantaneous escape behavior and a prolonged reduction in oviposition. Chemical analysis revealed this predator-specific signature to consist of a class of sulfolipids, produced by a biochemical pathway required for developing predacious behavior and specifically induced by starvation. These sulfolipids are detected by four pairs of C. elegans amphid sensory neurons that act redundantly and recruit cyclic nucleotide-gated (CNG) or transient receptor potential (TRP) channels to drive both escape and reduced oviposition. Specific abolishment of predator-evoked C. elegans responses by the anti-anxiety drug sertraline as well as functional homology of the delineated signaling pathways suggests a conserved or convergent strategy for managing predator threats.

neuroscience↗

bioRxiv2017

Evaluation and Design of Genome-wide CRISPR/Cas9 Knockout Screens

The adaptation of CRISPR/Cas9 technology to mammalian cell lines is transforming the study of human functional genomics. Pooled libraries of CRISPR guide RNAs (gRNAs), targeting human protein-coding genes and encoded in viral vectors, have been used to systematically create gene knockouts in a variety of human cancer and immortalized cell lines, in an effort to identify whether these knockouts cause cellular fitness defects. Previous work has shown that CRISPR screens are more sensitive and specific than pooled library shRNA screens in similar assays, but currently there exists significant variability across CRISPR library designs and experimental protocols. In this study, we re-analyze 17 genome-scale knockout screens in human cell lines from three research groups using three different genome-scale gRNA libraries, using the Bayesian Analysis of Gene Essentiality (BAGEL) algorithm to identify essential genes, to refine and expand our previously defined set of human core essential genes, from 360 to 684 genes. We use this expanded set of reference Core Essential Genes (CEG2), plus empirical data from six CRISPR knockout screens, to guide the design of a sequence-optimized gRNA library, the Toronto KnockOut version 3.0 (TKOv3) library. We demonstrate the high effectiveness of the library relative to reference sets of essential and nonessential genes as well as other screens using similar approaches. The optimized TKOv3 library, combined with the CEG2 reference set, provide an efficient, highly optimized platform for performing and assessing gene knockout screens in human cell lines.

systems biology↗