Stochastic gene expression in a single cell

Highly expressed cell wall genes contribute to robustness of sepal size

Reproducibility in organ size and shape is a fascinating trait of living organisms. The mechanisms underlying such robustness remain, however, to be elucidated. Taking the sepal of Arabidopsis as a model, we investigated whether variability of gene expression plays a role in variation of organ size and shape. Previous work from our team identified cell-wall related genes as being enriched among the genes whose expression is highly variable. We then hypothesized that the variation of measured morphological parameters in cell-wall related single knockout mutants could be correlated with the variation in gene expression of the corresponding gene (the knocked-out gene) in wild-type plants. We analyzed sepal size and shape from 16 cell-wall mutants and found that sepal size variability correlates positively, not with gene expression variation, but with mean gene expression of the corresponding gene in wild type. These findings support a contribution of cell-wall related genes to the robustness of sepal size.

Exploring Single-Probe Single-Cell Mass Spectrometry: Current Trends and Future Directions

The Single-probe single-cell mass spectrometry (SCMS) is an innovative analytical technique designed for metabolomic profiling, offering a miniaturized, multifunctional device capable of direct coupling to mass spectrometers. It is an ambient technique leveraging microscale sampling and nanoelectrospray ionization (nanoESI), enabling the analysis of cells in their native environments without the need for extensive sample preparation. Due to its miniaturized design and versatility, this device allows for applications in diverse research areas, including single-cell metabolomics, quantification of target molecules in single cell, MS imaging (MSI) of tissue sections, and investigation of extracellular molecules in live single spheroids. This review explores recent advancements in Single-probe-based techniques and their applications, emphasizing their potential utility in advancing MS methodologies in microscale bioanalysis.

Circadian Control of Protein Synthesis

Daily rhythms in the rate and specificity of protein synthesis occur in most mammalian cells through an interaction between cell-autonomous circadian regulation and daily cycles of systemic cues. However, the overall protein content of a typical cell changes little over 24 h. For most proteins, translation appears to be coordinated with protein degradation, producing phases of proteomic renewal that maximize energy efficiency while broadly maintaining proteostasis across the solar cycle. We propose that a major function of this temporal compartmentalization-and of circadian rhythmicity in general-is to optimize the energy efficiency of protein synthesis and associated processes such as complex assembly. We further propose that much of this temporal compartmentalization is achieved at the level of translational initiation, such that the translational machinery alternates between distinct translational mechanisms, each using a distinct toolkit of phosphoproteins to preferentially recognize and translate different classes of mRNA.

Cellular variability as a driver for bioprocess innovation and optimization

Cellular heterogeneity plays a crucial role in biotechnological processes, significantly influencing metabolic activity, product yield, and process consistency. This review explores the different dimensions of cellular heterogeneity, focusing on its manifestation at both single-cell and population levels. The study examines how factors such as asymmetric cell division, age, and environmental conditions contribute to functional diversity within cell populations, with an emphasis on microorganisms like yeast. Age-related cellular heterogeneity, in particular, is highlighted for its impact on metabolic pathways, mitochondrial function, and secondary metabolite production, which directly affect bioprocess outcomes. Furthermore, the review discusses advanced techniques for detecting and managing heterogeneity, including surface marker-based approaches, which utilize proteins, polysaccharides, and lipids, and label-free methods that leverage cellular volume and physical properties for separation. Understanding and controlling cellular heterogeneity is essential for optimizing industrial bioprocesses, improving yield, and ensuring product quality. The review also underscores the potential of emerging biotechnological tools, such as real-time single-cell analysis and microfluidic devices, in enhancing separation techniques and managing cellular diversity for better process efficiency and robustness.

Single-cell analyses reveal increased gene expression variability in human neurodevelopmental conditions

Interindividual variation in phenotypic penetrance and severity is found in many neurodevelopmental conditions, although the underlying mechanisms remain largely unresolved. Within individuals, homogeneous cell types (i.e., genetically identical and in similar environments) can differ in molecule abundance. Here, we investigate the hypothesis that neurodevelopmental conditions can drive increased variability in gene expression, not just differential gene expression. Leveraging independent single-cell and single-nucleus RNA sequencing datasets derived from human brain-relevant cell and tissue types, we identify a significant increase in gene expression variability driven by the autosomal aneuploidy trisomy 21 (T21) as well as autism-associated chromodomain helicase DNA binding protein 8 (CHD8) haploinsufficiency. Our analyses are consistent with a global and, in part, stochastic increase in variability, which is uncoupled from changes in transcript abundance. Highly variable genes tend to be cell-type specific with modest enrichment for repressive H3K27me3, while least variable genes are more likely to be constrained and associated with active histone marks. Our results indicate that human neurodevelopmental conditions can drive increased gene expression variability in brain cell types, with the potential to contribute to diverse phenotypic outcomes. These findings also provide a scaffold for understanding variability in disease, essential for deeper insights into genotype-phenotype relationships.

Invasion and metastasis in cancer: molecular insights and therapeutic targets

The progression of malignant tumors leads to the development of secondary tumors in various organs, including bones, the brain, liver, and lungs. This metastatic process severely impacts the prognosis of patients, significantly affecting their quality of life and survival rates. Research efforts have consistently focused on the intricate mechanisms underlying this process and the corresponding clinical management strategies. Consequently, a comprehensive understanding of the biological foundations of tumor metastasis, identification of pivotal signaling pathways, and systematic evaluation of existing and emerging therapeutic strategies are paramount to enhancing the overall diagnostic and treatment capabilities for metastatic tumors. However, current research is primarily focused on metastasis within specific cancer types, leaving significant gaps in our understanding of the complex metastatic cascade, organ-specific tropism mechanisms, and the development of targeted treatments. In this study, we examine the sequential processes of tumor metastasis, elucidate the underlying mechanisms driving organ-tropic metastasis, and systematically analyze therapeutic strategies for metastatic tumors, including those tailored to specific organ involvement. Subsequently, we synthesize the most recent advances in emerging therapeutic technologies for tumor metastasis and analyze the challenges and opportunities encountered in clinical research pertaining to bone metastasis. Our objective is to offer insights that can inform future research and clinical practice in this crucial field.

Inference of a three-gene network underpinning epidermal stem cell development in Caenorhabditis elegans

Gene regulatory networks are crucial in cellular decision-making, making the inference of their architecture essential for understanding organismal development. The gene network of Caenorhabditis elegans epidermal stem cells, known as seam cells, remains undefined. Here, we integrate experimental data, mathematical modeling, and statistical inference to investigate this network, focusing on three core transcription factors (TFs), namely ELT-1, EGL-18, and CEH-16. We use single-molecule FISH to quantify TF mRNA levels in single seam cells of wild-type and mutant backgrounds across four early larval stages. Using Modular Response Analysis, we predict TF interactions and uncover a repressive interaction between CEH-16 and egl-18 consistent across time points. We validate its significance at the L1 stage with ordinary differential equations and Bayesian modeling, making testable predictions for a double mutant. Our findings reveal TF regulatory relationships in seam cells and demonstrate a flexible mathematical framework for inferring gene regulatory networks from gene expression data.

Concentration buffering and noise reduction in non-equilibrium phase-separating systems

Biomolecular condensates have been proposed to buffer intracellular concentrations and reduce noise. However, concentrations need not be buffered in multicomponent systems, leading to a non-constant saturation concentration (csat) when individual components are varied. Simplified equilibrium considerations suggest that noise reduction might be closely related to concentration buffering and that a fixed saturation concentration is required for noise reduction to be effective. Here, we present a theoretical analysis to demonstrate that these suggestions do not apply to mesoscopic fluctuating systems. We show that concentration buffering and noise reduction are distinct concepts, which cannot be used interchangeably. We further demonstrate that concentration buffering and a constant csat are neither necessary nor sufficient for noise reduction to be effective. Clarity about these concepts is important for studying the role of condensates in controlling cellular noise and for the interpretation of concentration relationships in cells. A record of this paper's transparent peer review process is included in the supplemental information.

Transcriptional Heterogeneity and Differential Response of Rod Photoreceptor Pathway Uncovered by Single-Cell RNA Sequencing of the Aging Mouse Retina

Visual function deteriorates throughout the natural course of aging. Age-related structural and functional adaptations are observed in the retina, the light-sensitive neuronal tissue of the eye where visual perception begins. Molecular mechanisms underlying retinal aging are still poorly understood, highlighting the need to identify biomarkers for better prognosis and alleviation of aging-associated vision impairment. Here, we investigate dynamics of transcriptional dysregulation in the retina and identify affected pathways within distinct retinal cell types. Using an optimized protocol for single-cell RNA sequencing of mouse retinas at 3, 12, 18, and 24 months, we detect a progressive increase in the number of differentially expressed genes across all retinal cell types. The extent and direction of expression changes varies, with photoreceptor, bipolar, and Müller cells showing the maximum number of differentially expressed genes at all age groups. Furthermore, our analyses uncover transcriptionally distinct, heterogeneous subpopulations of rod photoreceptors and bipolar cells, distributed across distinct areas of the retina. Our findings provide a plausible molecular explanation for enhanced susceptibility of rod cells to aging and correlate with the observed loss of scotopic sensitivity in elderly individuals.

Contextual computation by competitive protein dimerization networks

Many biological signaling pathways employ proteins that competitively dimerize in diverse combinations. These dimerization networks can perform biochemical computations in which the concentrations of monomer inputs determine the concentrations of dimer outputs. Despite their prevalence, little is known about the range of input-output computations that dimerization networks can perform and how it depends on network size and connectivity. Using a systematic computational approach, we demonstrate that even small dimerization networks of 3-6 monomers are expressive, performing diverse multi-input computations. Further, dimerization networks are versatile, performing different computations when their protein components are expressed at different levels, such as in different cell types. Remarkably, individual networks with random interaction affinities, when large enough, can perform nearly all potential one-input network computations merely by tuning their monomer expression levels. Thus, even the simple process of competitive dimerization provides a powerful architecture for multi-input, cell-type-specific signal processing.

Priestia megaterium cells are primed for surviving lethal doses of antibiotics and chemical stress

Antibiotic resistant infections kill millions worldwide yearly. However, a key factor in recurrent infections is antibiotic persisters. Persisters are not inherently antibiotic-resistant but can withstand antibiotic exposure by entering a non-dividing state. This tolerance often results in prolonged antibiotic usage, increasing the likelihood of developing resistant strains. Here, we show the existence of "primed cells" in the Gram-positive bacterium Priestia megaterium, formerly known as Bacillus megaterium. These cells are pre-adapted to become persisters prior to lethal antibiotic stress. Remarkably, this prepared state is passed down through multiple generations via epigenetic memory, enhancing survival against antibiotics and other chemical stress. Previously, two distinct types of persisters were proposed: Type I and Type II, formed during stationary and log phases, respectively. However, our findings reveal that primed cells contribute to an increase in persisters during transition and stationary phases, with no evidence supporting distinct phenotypes between Type I and Type II persisters.

Attractor Landscape Analysis Reveals a Reversion Switch in the Transition of Colorectal Tumorigenesis

A cell fate change such as tumorigenesis incurs critical transition. It remains a longstanding challenge whether the underlying mechanism can be unraveled and a molecular switch that can reverse such transition is found. Here a systems framework, REVERT, is presented with which can reconstruct the core molecular regulatory network model and a reversion switch based on single-cell transcriptome data over the transition process is identified. The usefulness of REVERT is demonstrated by applying it to single-cell transcriptome of patient-derived matched organoids of colon cancer and normal colon. REVERT is a generic framework that can be applied to investigate various cell fate transition phenomena.

Functional constraints of wtf killer meiotic drivers

Killer meiotic drivers are selfish DNA loci that sabotage the gametes that do not inherit them from a driver+/driver- heterozygote. These drivers often employ toxic proteins that target essential cellular functions to cause the destruction of driver- gametes. Identifying the mechanisms of drivers can expand our understanding of infertility and reveal novel insights about the cellular functions targeted by drivers. In this work, we explore the molecular mechanisms underlying the wtf family of killer meiotic drivers found in fission yeasts. Each wtf killer acts using a toxic Wtfpoison protein that can be neutralized by a corresponding Wtfantidote protein. The wtf genes are rapidly evolving and extremely diverse. Here we found that self-assembly of Wtfpoison proteins is broadly conserved and associated with toxicity across the gene family, despite minimal amino acid conservation. In addition, we found the toxicity of Wtfpoison assemblies can be modulated by protein tags designed to increase or decrease the extent of the Wtfpoison assembly, implicating assembly size in toxicity. We also identified a conserved, critical role for the specific co-assembly of the Wtfpoison and Wtfantidote proteins in promoting effective neutralization of Wtfpoison toxicity. Finally, we engineered wtf alleles that encode toxic Wtfpoison proteins that are not effectively neutralized by their corresponding Wtfantidote proteins. The possibility of such self-destructive alleles reveals functional constraints on wtf evolution and suggests similar alleles could be cryptic contributors to infertility in fission yeast populations. As rapidly evolving killer meiotic drivers are widespread in eukaryotes, analogous self-killing drive alleles could contribute to sporadic infertility in many lineages.

Long-term imaging of individual ribosomes reveals ribosome cooperativity in mRNA translation

The genetic information stored in mRNAs is decoded by ribosomes during mRNA translation. mRNAs are typically translated by multiple ribosomes simultaneously, but it is unclear whether and how the activity of different ribosomes on an mRNA is coordinated. Here, we develop an imaging approach based on stopless-ORF circular RNAs (socRNAs) to monitor translation of individual ribosomes in either monosomes or polysomes with very high resolution. Using experiments and simulations, we find that translating ribosomes frequently undergo transient collisions. However, unlike persistent collisions, such transient collisions escape detection by cellular quality control pathways. Rather, transient ribosome collisions promote productive translation by reducing ribosome pausing on problematic sequences, a process we term ribosome cooperativity. Ribosome cooperativity also reduces recycling of ribosomes by quality control pathways, thus enhancing processive translation. Together, our single-ribosome imaging approach reveals that ribosomes cooperate during translation to ensure fast and efficient translation.

A differentiable Gillespie algorithm for simulating chemical kinetics, parameter estimation, and designing synthetic biological circuits

The Gillespie algorithm is commonly used to simulate and analyze complex chemical reaction networks. Here, we leverage recent breakthroughs in deep learning to develop a fully differentiable variant of the Gillespie algorithm. The differentiable Gillespie algorithm (DGA) approximates discontinuous operations in the exact Gillespie algorithm using smooth functions, allowing for the calculation of gradients using backpropagation. The DGA can be used to quickly and accurately learn kinetic parameters using gradient descent and design biochemical networks with desired properties. As an illustration, we apply the DGA to study stochastic models of gene promoters. We show that the DGA can be used to: (i) successfully learn kinetic parameters from experimental measurements of mRNA expression levels from two distinct E. coli promoters and (ii) design nonequilibrium promoter architectures with desired input-output relationships. These examples illustrate the utility of the DGA for analyzing stochastic chemical kinetics, including a wide variety of problems of interest to synthetic and systems biology.

Gene function revealed at the moment of sitochastic gene silencing

Gene expression is a dynamic and stochastic process characterized by transcriptional bursting followed by periods of silence. Single-cell RNA sequencing (scRNA-seq) is a powerful tool to measure transcriptional bursting and silencing at the individual cell level. In this study, we introduce the single-cell Stochastic Gene Silencing (scSGS) method, which leverages the natural variability in single-cell gene expression to decipher gene function. For a target gene g under investigation, scSGS classifies cells into transcriptionally active (g + ) and silenced (g-) samples. It then compares these cell samples to identify differentially expressed genes, referred to as SGS-responsive genes, which are used to infer the function of the target gene g. Analysis of real data demonstrates that scSGS can reveal regulatory relationships up- and downstream of target genes, circumventing the survivorship bias that often affects gene knockout and perturbation studies. scSGS thus offers an efficient approach for gene function prediction, with significant potential to reduce the use of genetically modified animals in gene function research.

The impact of phenotypic heterogeneity on fungal pathogenicity and drug resistance

Phenotypic heterogeneity in genetically clonal populations facilitates cellular adaptation to adverse environmental conditions while enabling a return to the basal physiological state. It also plays a crucial role in pathogenicity and the acquisition of drug resistance in unicellular organisms and cancer cells, yet the exact contributing factors remain elusive. In this review, we outline the current state of understanding concerning the contribution of phenotypic heterogeneity to fungal pathogenesis and antifungal drug resistance.

Deterministic patterns in single-cell transcriptomic data

We report the existence of deterministic patterns in statistical plots of single-cell transcriptomic data. We develop a theory showing that the patterns are neither artifacts introduced by the measurement process nor due to underlying biological mechanisms. Rather they naturally emerge from finite sample size effects. The theory precisely predicts the patterns in data from multiplexed error-robust fluorescence in situ hybridization and five different types of single-cell sequencing platforms.

Nonlinear regulatory dynamics of bacterial restriction-modification systems modulates horizontal gene transfer susceptibility

Type II restriction-modification (R-M) systems play a pivotal role in bacterial defense against invading DNA, influencing the spread of pathogenic traits. These systems often involve coordinated expression of a regulatory protein (C) with restriction (R) enzymes, employing complex feedback loops for regulation. Recent studies highlight the crucial balance between R and M enzymes in controlling horizontal gene transfer (HGT). This manuscript introduces a mathematical model reflecting R-M system dynamics, informed by biophysical evidence, to minimize reliance on arbitrary parameters. Our analysis clarifies the observed variations in M-to-R ratios, emphasizing the regulatory role of the C protein. We analytically derived a stability diagram for C-regulated R-M systems, offering a more straightforward analysis method over traditional numerical approaches. Our findings reveal conditions leading to both monostability and bistability, linking changes in the M-to-R ratio to factors like cell division timing and plasmid replication rates. These variations may link adjusting defense against phage infection, or the acquisition of new genes such as antibiotic resistance determinants, to changing physiological conditions. We also performed stochastic simulations to show that system regulation may significantly increase M-to-R ratio variability, providing an additional mechanism to generate heterogeneity in bacterial population.

Functional constraints of wtf killer meiotic drivers

Killer meiotic drivers are selfish DNA loci that sabotage the gametes that do not inherit them from a driver+/driver- heterozygote. These drivers often employ toxic proteins that target essential cellular functions to cause the destruction of driver- gametes. Identifying the mechanisms of drivers can expand our understanding of infertility and reveal novel insights about the cellular functions targeted by drivers. In this work, we explore the molecular mechanisms underlying the wtf family of killer meiotic drivers found in fission yeasts. Each wtf killer acts using a toxic Wtfpoison protein that can be neutralized by a corresponding Wtfantidote protein. The wtf genes are rapidly evolving and extremely diverse. Here we found that self-assembly of Wtfpoison proteins is broadly conserved and associated with toxicity across the gene family, despite minimal amino acid conservation. In addition, we found the toxicity of Wtfpoison assemblies can be modulated by protein tags designed to increase or decrease the extent of the Wtfpoison assembly, implicating assembly size in toxicity. We also identified a conserved, critical role for the specific co-assembly of the Wtfpoison and Wtfantidote proteins in promoting effective neutralization of Wtfpoison toxicity. Finally, we engineered wtf alleles that encode toxic Wtfpoison proteins that are not effectively neutralized by their corresponding Wtfantidote proteins. The possibility of such self-destructive alleles reveals functional constraints on wtf evolution and suggests similar alleles could be cryptic contributors to infertility in fission yeast populations. As rapidly evolving killer meiotic drivers are widespread in eukaryotes, analogous self-killing drive alleles could contribute to sporadic infertility in many lineages.

Relative Distribution of DnaA and DNA in Escherichia coli Cells as a Factor of Their Phenotypic Variability

Phenotypic variability in isogenic bacterial populations is a remarkable feature that helps them cope with external stresses, yet it is incompletely understood. This variability can stem from gene expression noise and/or the unequal partitioning of low-copy-number freely diffusing proteins during cell division. Some high-copy-number components are transiently associated with almost immobile large assemblies (hyperstructures) and may be unequally distributed, contributing to bacterial phenotypic variability. We focus on the nucleoid hyperstructure containing numerous DNA-associated proteins, including the replication initiator DnaA. Previously, we found an increasing asynchrony in the nucleoid segregation dynamics in growing E. coli cell lineages and suggested that variable replication initiation timing may be the main cause of this phenomenon. Here, we support this hypothesis revealing that DnaA/DNA variability represents a key factor leading to the enhanced asynchrony in E. coli. We followed the intra- and intercellular distribution of fluorescently tagged DnaA and histone-like HU chromosomally encoded under their native promoters. The diffusion rate of DnaA is low, corresponding to a diffusion-binding mode of mobility, but still one order faster than that of HU. The intracellular distribution of DnaA concentration is homogeneous in contrast to the significant asymmetry in the distribution of HU to the cell halves, leading to the unequal DNA content of nucleoids and DnaA/DNA ratios in future daughter compartments. Accordingly, the intercellular variabilities in HU concentration (CV = 26%) and DnaA/DNA ratio (CV = 18%) are high. The variable DnaA/DNA may cause a variable replication initiation time (initiation noise). Asynchronous initiation at different replication origins may, in turn, be the mechanism leading to the observed asymmetric intracellular DNA distribution. Our findings indicate that the feature determining the variability of the initiation time in E. coli is the DnaA/DNA ratio, rather than each of them separately. We provide a likely mechanism for the 'loss of segregation synchrony' phenomenon.

Population-level amplification of gene regulation by programmable gene transfer

Engineering cells to sense and respond to environmental cues often focuses on maximizing gene regulation at the single-cell level. Inspired by population-level control mechanisms like the immune response, we demonstrate dynamic control and amplification of gene regulation in bacterial populations using programmable plasmid-mediated gene transfer. By regulating plasmid loss rate, transfer rate and fitness effects via Cas9 endonuclease, F conjugation machinery and antibiotic selection, we modulate the fraction of plasmid-carrying cells, serving as an amplification factor for single-cell-level regulation. This approach expands the dynamic range of gene expression and allows orthogonal control across populations. Our platform offers a versatile strategy for dynamically regulating gene expression in engineered microbial communities.

Multimodal screen identifies noise-regulatory proteins

Gene-expression noise can influence cell-fate choices across pathology and physiology. However, a crucial question persists: do regulatory proteins or pathways exist that control noise independently of mean expression levels? Our integrative approach, combining single-cell RNA sequencing with proteomics and regulator enrichment analysis, identifies 32 putative noise regulators. SON, a nuclear speckle-associated protein, alters transcriptional noise without changing mean expression levels. Furthermore, SON's noise control can propagate to the protein level. Long-read and total RNA sequencing shows that SON's noise control does not significantly change isoform usage or splicing efficiency. Moreover, SON depletion reduces state switching in pluripotent mouse embryonic stem cells and impacts their fate choice during differentiation. Collectively, we demonstrate a class of proteins that control noise orthogonally to mean expression levels. This work serves as a proof of concept that can identify other functional noise regulators throughout development and disease progression.

Single-cell data reveal heterogeneity of investment in ribosomes across a bacterial population

Ribosomes are responsible for the synthesis of proteins, the major component of cellular biomass. Classical experiments have established a linear relationship between the fraction of resources invested in ribosomal proteins and the rate of balanced growth of a microbial population. Very little is known, however, about how the investment in ribosomes varies over individual cells in a population. We therefore extended the study of ribosomal resource allocation from populations to single cells, using a combination of time-lapse fluorescence microscopy and statistical inference. We found a large variability of ribosome concentrations and growth rates in conditions of balanced growth of the model bacterium Escherichia coli in a given medium, which cannot be accounted for by the population-level growth law. A large variability in the allocation of resources to ribosomes was also found during the transition of the bacteria from a poor to a rich growth medium. While some cells immediately adapt their ribosome synthesis rate to the new environment, others do so only gradually. Our results thus reveal a range of strategies for investing resources in the molecular machines at the heart of cellular self-replication. This raises the fundamental question whether the observed variability is an intrinsic consequence of the stochastic nature of the underlying biochemical processes or whether it improves the fitness of Escherichia coli in its natural environment.

Asynchronous abundance fluctuations can drive giant genotype frequency fluctuations

Large stochastic population abundance fluctuations are ubiquitous across the tree of life, impacting the predictability and outcomes of population dynamics. It is generally thought that abundance fluctuations with a Taylor's law exponent of two do not strongly impact evolution. However, we argue that such abundance fluctuations can lead to substantial genotype frequency fluctuations if different genotypes in a population experience these fluctuations asynchronously. By serially diluting mixtures of two closely related Escherichia coli strains, we show that such asynchrony can occur, leading to giant frequency fluctuations that far exceed expectations from genetic drift. We develop an effective model explaining that the abundance fluctuations arise from correlated offspring numbers between individuals, and the large frequency fluctuations result from (even slight) decoupling in offspring number correlations between genotypes. The model quantitatively predicts the observed abundance and frequency fluctuation scaling. Initially close trajectories diverge exponentially, suggesting that chaotic dynamics may underpin the excess frequency fluctuations. Our findings suggest that decoupling noise is also present in mixed-genotype Saccharomyces cerevisiae populations. Theoretical analyses demonstrate that decoupling noise can strongly influence evolutionary outcomes, in a manner distinct from genetic drift. Given the generic nature of these frequency fluctuations, we expect them to be widespread across biological populations.

Exact distributions of threshold crossing times of proteins under post-transcriptional regulation by small RNAs

The timings of several cellular events like cell lysis, cell division, or pore formation in endosomes are regulated by the time taken for the relevant proteins to cross a threshold in number or concentration. Since protein synthesis is stochastic, the threshold crossing time is a first passage problem. The exact distributions of these first passage processes have been obtained recently for unregulated and autoregulated genes. Many proteins are however regulated by post-transcriptional regulation, controlled by small noncoding RNAs (sRNAs). Certain mathematical models of gene expression with post-transcriptional sRNA regulation have been recently exactly mapped to models without sRNA regulation. Utilizing this mapping and the exact distributions, we calculate exact results on fluctuations (full distribution, all cumulants, and characteristic times) of protein threshold crossing times in the presence of sRNA regulation. We derive two interesting predictions from these exact results. We show that the size of the fluctuation of the threshold crossing times have a nonmonotonic U-shaped behavior as a function of the rates of binding and unbinding of the sRNA-mRNA complex. Thus there are optimal parameters that minimize noise. Furthermore, the fluctuations in models with sRNA regulation may be higher or lower compared to the model without regulation, depending on the mean protein burst size.

Network for knowledge Organization (NEKO): An AI knowledge mining workflow for synthetic biology research

Large language models (LLMs) can complete general scientific question-and-answer, yet they are constrained by their pretraining cut-off dates and lack the ability to provide specific, cited scientific knowledge. Here, we introduce Network for Knowledge Organization (NEKO), a workflow that uses LLM Qwen to extract knowledge through scientific literature text mining. When user inputs a keyword of interest, NEKO can generate knowledge graphs to link bioinformation entities and produce comprehensive summaries from PubMed search. NEKO significantly enhance LLM ability and has immediate applications in daily academic tasks such as education of young scientists, literature review, paper writing, experiment planning/troubleshooting, and new ideas/hypothesis generation. We exemplified this workflow's applicability through several case studies on yeast fermentation and cyanobacterial biorefinery. NEKO's output is more informative, specific, and actionable than GPT-4's zero-shot Q&A. NEKO offers flexible, lightweight local deployment options. NEKO democratizes artificial intelligence (AI) tools, making scientific foundation model more accessible to researchers without excessive computational power.

Disentangling variational bias: the roles of development, mutation, and selection

The extraordinary diversity and adaptive fit of organisms to their environment depends fundamentally on the availability of variation. While most population genetic frameworks assume that random mutations produce isotropic phenotypic variation, the distribution of variation available to natural selection is more restricted, as the distribution of phenotypic variation is affected by a range of factors in developmental systems. Here, we revisit the concept of developmental bias - the observation that the generation of phenotypic variation is biased due to the structure, character, composition, or dynamics of the developmental system - and argue that a more rigorous investigation into the role of developmental bias in the genotype-to-phenotype map will produce fundamental insights into evolutionary processes, with potentially important consequences on the relation between micro- and macro-evolution. We discuss the hierarchical relationships between different types of variational biases, including mutation bias and developmental bias, and their roles in shaping the realized phenotypic space. Furthermore, we highlight the challenges in studying variational bias and propose potential approaches to identify developmental bias using modern tools.

Deep learning-based models for preimplantation mouse and human embryos based on single-cell RNA sequencing

The rapid growth of single-cell transcriptomic technology has produced an increasing number of datasets for both embryonic development and in vitro pluripotent stem cell-derived models. This avalanche of data surrounding pluripotency and the process of lineage specification has meant it has become increasingly difficult to define specific cell types or states in vivo, and compare these with in vitro differentiation. Here we utilize a set of deep learning tools to integrate and classify multiple datasets. This allows the definition of both mouse and human embryo cell types, lineages and states, thereby maximizing the information one can garner from these precious experimental resources. Our approaches are built on recent initiatives for large-scale human organ atlases, but here we focus on material that is difficult to obtain and process, spanning early mouse and human development. Using publicly available data for these stages, we test different deep learning approaches and develop a model to classify cell types in an unbiased fashion at the same time as defining the set of genes used by the model to identify lineages, cell types and states. We used our models trained on in vivo development to classify pluripotent stem cell models for both mouse and human development, showcasing the importance of this resource as a dynamic reference for early embryogenesis.

The highly rugged yet navigable regulatory landscape of the bacterial transcription factor TetR

Transcription factor binding sites (TFBSs) are important sources of evolutionary innovations. Understanding how evolution navigates the sequence space of such sites can be achieved by mapping TFBS adaptive landscapes. In such a landscape, an individual location corresponds to a TFBS bound by a transcription factor. The elevation at that location corresponds to the strength of transcriptional regulation conveyed by the sequence. Here, we develop an in vivo massively parallel reporter assay to map the landscape of bacterial TFBSs. We apply this assay to the TetR repressor, for which few TFBSs are known. We quantify the strength of transcriptional repression for 17,765 TFBSs and show that the resulting landscape is highly rugged, with 2092 peaks. Only a few peaks convey stronger repression than the wild type. Non-additive (epistatic) interactions between mutations are frequent. Despite these hallmarks of ruggedness, most high peaks are evolutionarily accessible. They have large basins of attraction and are reached by around 20% of populations evolving on the landscape. Which high peak is reached during evolution is unpredictable and contingent on the mutational path taken. This in-depth analysis of a prokaryotic gene regulator reveals a landscape that is navigable but much more rugged than the landscapes of eukaryotic regulators.

A simple MiMIC-based approach for tagging endogenous genes to visualise live transcription in Drosophila

Live imaging of transcription in the Drosophila embryo using the MS2 or PP7 systems is transforming our understanding of transcriptional regulation. However, insertion of MS2/PP7 stem-loops into endogenous genes requires laborious CRISPR genome editing. Here, we exploit the previously described Minos-mediated integration cassette (MiMIC) transposon system in Drosophila to establish a method for simply and rapidly inserting MS2/PP7 cassettes into any of the thousands of genes carrying a MiMIC insertion. In addition to generating a variety of stem-loop donor fly stocks, we have made new stocks expressing the complementary coat proteins fused to different fluorescent proteins. We show the utility of this MiMIC-based approach by MS2/PP7 tagging of endogenous genes and the long non-coding RNA roX1, then imaging their transcription in living embryos. We also present live transcription data from larval brains, the wing disc and ovary, thereby extending the tissues that can be studied using the MS2/PP7 system. Overall, this first high-throughput method for tagging mRNAs in Drosophila will facilitate the study of transcription dynamics of thousands of endogenous genes in a range of Drosophila tissues.

A synthetic protein-level neural network in mammalian cells

Artificial neural networks provide a powerful paradigm for nonbiological information processing. To understand whether similar principles could enable computation within living cells, we combined de novo-designed protein heterodimers and engineered viral proteases to implement a synthetic protein circuit that performs winner-take-all neural network classification. This "perceptein" circuit combines weighted input summation through reversible binding interactions with self-activation and mutual inhibition through irreversible proteolytic cleavage. These interactions collectively generate a large repertoire of distinct protein species stemming from up to eight coexpressed starting protein species. The complete system achieves multi-output signal classification with tunable decision boundaries in mammalian cells and can be used to conditionally control cell death. These results demonstrate how engineered protein-based networks can enable programmable signal classification in living cells.

Emergence of temporal noise hierarchy in co-regulated genes of multi-output feed-forward loop

Natural variations in gene expression, called noise, are fundamental to biological systems. The expression noise can be beneficial or detrimental to cellular functions. While the impact of noise on individual genes is well-established, our understanding of how noise behaves when multiple genes are co-expressed by shared regulatory elements within transcription networks remains elusive. This lack of understanding extends to how the architecture and regulatory features of these networks influence noise. To address this gap, we study the multi-output feed-forward loop motif. The motif is prevalent in bacteria and yeast and influences co-expression of multiple genes by shared transcription factors (TFs). Focusing on a two-output variant of the motif, the present study explores the interplay between its architecture, co-expression (symmetric and asymmetric) patterns of the two genes, and the associated noise dynamics. We employ a stochastic modeling approach to investigate how the binding affinities of the TFs influence symmetric and asymmetric expression patterns and the resulting noise dynamics in the co-expressed genes. This knowledge could guide the development of strategies for manipulating gene expression patterns through targeted modulation of TF binding affinities.

Dynamic heterogeneity in an E. coli stress response regulon mediates gene activation and antimicrobial peptide tolerance

The bacterial stress response is an intricately regulated system that plays a critical role in cellular resistance to drug treatment. The complexity of this response is further complicated by cell-to-cell heterogeneity in the expression of bacterial stress response genes. These genes are often organized into networks comprising one or more transcriptional regulators that control expression of a suite of downstream genes. While the expression heterogeneity of many of these upstream regulators has been characterized, the way in which this variability affects the larger downstream stress response remains hard to predict, prompting two key questions. First, how does heterogeneity and expression noise in stress response regulators propagate to the diverse downstream genes in their regulons. Second, when expression levels vary, how do multiple downstream genes act together to protect cells from stress. To address these questions, we focus on the transcription factor PhoP, a critical virulence regulator which coordinates pathogenicity in several gram-negative species. We use optogenetic stimulation to precisely control PhoP expression levels and examine how variations in PhoP affect the downstream activation of genes in the PhoP regulon. We find that these downstream genes exhibit differences both in mean expression level and sensitivity to increasing levels of PhoP. These response functions can also vary between individual cells, increasing heterogeneity in the population. We tie these variations to cell survival when bacteria are exposed to a clinically-relevant antimicrobial peptide, showing that high expression of the PhoP-regulon gene pmrD provides a protective effect against Polymyxin B. Overall, we demonstrate that even subtle heterogeneity in expression of a stress response regulator can have clear consequences for enabling bacteria to survive stress.

Deciphering regulatory architectures of bacterial promoters from synthetic expression patterns

For the vast majority of genes in sequenced genomes, there is limited understanding of how they are regulated. Without such knowledge, it is not possible to perform a quantitative theory-experiment dialogue on how such genes give rise to physiological and evolutionary adaptation. One category of high-throughput experiments used to understand the sequence-phenotype relationship of the transcriptome is massively parallel reporter assays (MPRAs). However, to improve the versatility and scalability of MPRAs, we need a "theory of the experiment" to help us better understand the impact of various biological and experimental parameters on the interpretation of experimental data. These parameters include binding site copy number, where a large number of specific binding sites may titrate away transcription factors, as well as the presence of overlapping binding sites, which may affect analysis of the degree of mutual dependence between mutations in the regulatory region and expression levels. To that end, in this paper we create tens of thousands of synthetic gene expression outputs for bacterial promoters using both equilibrium and out-of-equilibrium models. These models make it possible to imitate the summary statistics (information footprints and expression shift matrices) used to characterize the output of MPRAs and thus to infer the underlying regulatory architecture. Specifically, we use a more refined implementation of the so-called thermodynamic models in which the binding energies of each sequence variant are derived from energy matrices. Our simulations reveal important effects of the parameters on MPRA data and we demonstrate our ability to optimize MPRA experimental designs with the goal of generating thermodynamic models of the transcriptome with base-pair specificity. Further, this approach makes it possible to carefully examine the mapping between mutations in binding sites and their corresponding expression profiles, a tool useful not only for developing a theory of transcription, but also for exploring regulatory evolution.

Polygenicity in a box: Copy number variants, neural circuit development, and neurodevelopmental disorders

Clinically defined neurodevelopmental disorders (cd-NDDs), including Autistic Spectrum Disorder (ASD) and Schizophrenia (Scz), are primarily polygenic: Multiple risk genes distributed across the genome, in potentially infinite combinations, account for variable pathology. Polygenicity raises a fundamental question: Can "core" cd-NDD pathogenic mechanisms be identified given this genomic complexity? With the right models and analytic targets, a distinct class of polygenic mutations-Copy Number Variants (CNVs): contiguous gene deletions or duplications associated with cd-NDD risk-provide a singular opportunity to define cd-NDD pathology. CNVs orthologous to those that confer cd-NDD risk have been engineered in animals as well as human stem cells. Using these tools, one can determine how altered function of multiple genes cause serial stumbles over cell biological steps typically taken to build optimal "polygenic" neural circuits. Thus, cd-NDD pathology may be a consequence of polygenic deviations-stumbles-that exceed limits of adaptive variation for key developmental steps.

Machine learning based analysis of single-cell data reveals evidence of subject-specific single-cell gene expression profiles in acute myeloid leukaemia patients and healthy controls

Acute Myeloid Leukaemia (AML) is characterized by uncontrolled growth of immature myeloid cells, disrupting normal blood production. Treatment typically involves chemotherapy, targeted therapy, and stem cell transplantation but many patients develop chemoresistance, leading to poor outcomes due to the disease's high heterogeneity. In this study, we used publicly available single-cell RNA sequencing data and machine learning to classify AML patients and healthy, monocytes, dendritic and progenitor cells population. We found that gene expression profiles of AML patients and healthy controls can be classified at the individual level with high accuracy (>70 %) when using progenitor cells, suggesting the existence of subject-specific single cell transcriptomics profiles. The analysis also revealed molecular determinants of patient heterogeneity (e.g. TPSD1, CT45A1, and GABRA4) which could support new strategies for patient stratification and personalized treatment in leukaemia.

Stress-Induced Constraint on Expression Noise of Essential Genes in E. coli

Gene expression is an inherently noisy process that is constrained by natural selection. Yet the condition dependence of constraint on expression noise remains unclear. Here, we address this problem by studying constraint on expression noise of E. coli genes in eight diverse growth conditions. In particular, we use variation in expression noise as an analog for constraint, examining its relationships to expression level and to the number of regulatory inputs from transcription factors across and within conditions. We show that variation in expression noise is negatively associated with expression level, implicating constraint to minimize expression noise of highly expressed genes. However, this relationship is condition dependent, with the strongest constraint observed when E. coli are grown in the presence of glycerol or ciprofloxacin, which result in carbon or antibiotic stress, respectively. In contrast, we do not observe evidence of constraint on expression noise of highly regulated genes, suggesting that highly expressed and highly regulated genes represent distinct classes of genes. Indeed, we find that essential genes are often highly expressed but not highly regulated, with elevated expression noise in glycerol and ciprofloxacin conditions. Thus, our findings support the hypothesis that selective constraint on expression noise is condition dependent in E. coli, illustrating how it may play a critical role in ensuring expression stability of essential genes in unstable environments.

Getting it right: suppression and leveraging of noise in robust decision-making

Noise is a ubiquitous feature for all organisms growing in nature. Noise (defined here as stochastic variation) in the availability of nutrients, water and light profoundly impacts their growth and development. Not only is noise present as an external factor but cellular processes themselves are noisy. Therefore, it is remarkable that organisms can display robust control of growth and development despite noise. To survive, various mechanisms to suppress noise have evolved. However, it is also becoming apparent that noise is not just a nuisance that organisms must suppress but can be beneficial as low noise can facilitate the response of an organism to a sub-threshold input signal in a stochastic resonance mechanism. This review discusses mechanisms capable of noise suppression or noise leveraging that might play a significant role in robust temporal regulation of an organism's response to their noisy environment.

Molecular Spies in Action: Genetically Encoded Fluorescent Biosensors Light up Cellular Signals

Cellular function is controlled through intricate networks of signals, which lead to the myriad pathways governing cell fate. Fluorescent biosensors have enabled the study of these signaling pathways in living systems across temporal and spatial scales. Over the years there has been an explosion in the number of fluorescent biosensors, as they have become available for numerous targets, utilized across spectral space, and suited for various imaging techniques. To guide users through this extensive biosensor landscape, we discuss critical aspects of fluorescent proteins for consideration in biosensor development, smart tagging strategies, and the historical and recent biosensors of various types, grouped by target, and with a focus on the design and recent applications of these sensors in living systems.

Plausible, robust biological oscillations through allelic buffering

Biological oscillators can specify time- and dose-dependent functions via dedicated control of their oscillatory dynamics. However, how biological oscillators, which recurrently activate noisy biochemical processes, achieve robust oscillations remains unclear. Here, we characterize the long-term oscillations of p53 and its negative feedback regulator Mdm2 in single cells after DNA damage. Whereas p53 oscillates regularly, Mdm2 from a single MDM2 allele exhibits random unresponsiveness to ∼9% of p53 pulses. Using allelic-specific imaging of MDM2 activity, we show that MDM2 alleles buffer each other to maintain p53 pulse amplitude. Removal of MDM2 allelic buffering cripples the robustness of p53 amplitude, thereby elevating p21 levels and cell-cycle arrest. In silico simulations support that allelic buffering enhances the robustness of biological oscillators and broadens their plausible biochemical space. Our findings show how allelic buffering ensures robust p53 oscillations, highlighting the potential importance of allelic buffering for the emergence of robust biological oscillators during evolution. A record of this paper's transparent peer review process is included in the supplemental information.

Harnessing microbial heterogeneity for improved biosynthesis fueled by synthetic biology

Metabolic engineering-driven microbial cell factories have made great progress in the efficient bioproduction of biochemical and recombinant proteins. However, the low efficiency and robustness of microbial cell factories limit their industrial applications. Harnessing microbial heterogeneity contributes to solving this. In this review, the origins of microbial heterogeneity and its effects on biosynthesis are first summarized. Synthetic biology-driven tools and strategies that can be used to improve biosynthesis by increasing and reducing microbial heterogeneity are then systematically summarized. Next, novel single-cell technologies available for unraveling microbial heterogeneity and facilitating heterogeneity regulation are discussed. Furthermore, a combined workflow of increasing genetic heterogeneity in the strain-building step to help in screening highly productive strains - reducing heterogeneity in the production process to obtain highly robust strains (IHP-RHR) facilitated by single-cell technologies was proposed to obtain highly productive and robust strains by harnessing microbial heterogeneity. Finally, the prospects and future challenges are discussed.

Abstraction-based segmental simulation of reaction networks using adaptive memoization

BACKGROUND

 Stochastic models are commonly employed in the system and synthetic biology to study the effects of stochastic fluctuations emanating from reactions involving species with low copy-numbers. Many important models feature complex dynamics, involving a state-space explosion, stiffness, and multimodality, that complicate the quantitative analysis needed to understand their stochastic behavior. Direct numerical analysis of such models is typically not feasible and generating many simulation runs that adequately approximate the model's dynamics may take a prohibitively long time.

RESULTS

 We propose a new memoization technique that leverages a population-based abstraction and combines previously generated parts of simulations, called segments, to generate new simulations more efficiently while preserving the original system's dynamics and its diversity. Our algorithm adapts online to identify the most important abstract states and thus utilizes the available memory efficiently.

CONCLUSION

 We demonstrate that in combination with a novel fully automatic and adaptive hybrid simulation scheme, we can speed up the generation of trajectories significantly and correctly predict the transient behavior of complex stochastic systems.

Energy Landscape Reveals the Underlying Mechanism of Cancer-Adipose Conversion in Gene Network Models

Cancer is a systemic heterogeneous disease involving complex molecular networks. Tumor formation involves an epithelial-mesenchymal transition (EMT), which promotes both metastasis and plasticity of cancer cells. Recent experiments have proposed that cancer cells can be transformed into adipocytes via a combination of drugs. However, the underlying mechanisms for how these drugs work, from a molecular network perspective, remain elusive. To reveal the mechanism of cancer-adipose conversion (CAC), this study adopts a systems biology approach by combing mathematical modeling and molecular experiments, based on underlying molecular regulatory networks. Four types of attractors are identified, corresponding to epithelial (E), mesenchymal (M), adipose (A) and partial/intermediate EMT (P) cell states on the CAC landscape. Landscape and transition path results illustrate that intermediate states play critical roles in the cancer to adipose transition. Through a landscape control approach, two new therapeutic strategies for drug combinations are identified, that promote CAC. These predictions are verified by molecular experiments in different cell lines. The combined computational and experimental approach provides a powerful tool to explore molecular mechanisms for cell fate transitions in cancer networks. The results reveal underlying mechanisms of intermediate cell states that govern the CAC, and identified new potential drug combinations to induce cancer adipogenesis.

Drug tolerance and persistence in bacteria, fungi and cancer cells: Role of non-genetic heterogeneity

A common feature of bacterial, fungal and cancer cell populations upon treatment is the presence of tolerant and persistent cells able to survive, and sometimes grow, even in the presence of usually inhibitory or lethal drug concentrations, driven by non-genetic differences among individual cells in a population. Here we review and compare data obtained on drug survival in bacteria, fungi and cancer cells to unravel common characteristics and cellular pathways, and to point their singularities. This comparative work also allows to cross-fertilize ideas across fields. We particularly focus on the role of gene expression variability in the emergence of cell-cell non-genetic heterogeneity because it represents a possible common basic molecular process at the origin of most persistence phenomena and could be monitored and tuned to help improve therapeutic interventions.

Phenotypic Heterogeneity in Pathogens

Pathogen diversity within an infected organism has traditionally been explored through the lens of genetic heterogeneity. Hallmark studies have characterized how genetic diversity within pathogen subpopulations contributes to treatment escape and infectious disease progression. However, recent studies have begun to reveal the mechanisms by which phenotypic heterogeneity is established within genetically identical populations of invading pathogens. Furthermore, exciting new work highlights how these phenotypically heterogeneous subpopulations contribute to a pathogen population better equipped to handle the complex and fluctuating environment of a host organism. In this review, we focus on how bacterial pathogens, including Staphylococcus aureus, Salmonella typhimurium, Pseudomonas aeruginosa, and Mycobacterium tuberculosis, establish and maintain phenotypic heterogeneity, and we explore recent work demonstrating causative links between this heterogeneity and infection outcome.

Evaluating the potential of vancomycin-modified magnetic beads as a tool for sample preparation in diagnostic assays

Vancomycin-functionalized micro- or nanoparticles are frequently used for isolation and enrichment of bacteria from various samples. Theoretically, only Gram-positive organisms should adhere to the functionalized surfaces as vancomycin is an antibiotic targeting a peptidoglycan precursor in the cell wall, which in Gram-negative bacteria is shielded by the outer cell membrane. In the literature, however, it is often reported that Gram-negative bacteria also bind efficiently to the vancomycin-modified particles. The goal of our study was to identify the underlying cause for these different findings. For each species several strains, including patient isolates, were investigated, and effects such as day-to-day reproducibility, particle type, and the antimicrobial effect of vancomycin-coupled beads were explored. Overall, we found that there is a strong preference for binding Gram-positive organisms, but the specific yield is heavily influenced by the strain and experimental conditions. For Staphylococcus aureus average yields of approximately 100% were obtained. Respectively, yields of 44% for Staphylococcus cohnii, 22% for Staphylococcus warneri, 17% for Enterococcus faecalis and 5% for vancomycin-sensitive Enterococcus faecium were found. Yields for Gram-negative species (Escherichia coli, Klebsiella pneumoniae, Acinetobacter baumannii) and vancomycin-resistant Enterococcus faecium were below 3%. Our results indicate that the interaction between vancomycin and the D-alanine-D-alanine terminus of the peptidoglycan precursor in the bacterial cell wall is the dominant force responsible for the adherence of the bacteria to the particle surface. It needs to be considered though, that other factors, such as the specific molecules presented on the bacterial surface, as well as the pH, and the ion concentrations in the surrounding medium will also play a role, as these can lead to attractive or repulsive electrostatic forces. Last but not least, when using colony forming unit-based quantification for determining the yields, the influence of cell cluster formation and different sensitivities towards the antimicrobial effect of the vancomycin beads between species and strains needs to be considered.

Preexisting cell state rather than stochastic noise confers high or low infection susceptibility of human lung epithelial cells to adenovirus

Viruses display large variability across all stages of their life cycle, including entry, gene expression, replication, assembly, and egress. We previously reported that the immediate early adenovirus (AdV) E1A transcripts accumulate in human lung epithelial A549 cancer cells with high variability, mostly independent of the number of incoming viral genomes, but somewhat correlated to the cell cycle state at the time of inoculation. Here, we leveraged the classical Luria-Delbrück fluctuation analysis to address whether infection variability primarily arises from the cell state or stochastic noise. The E1A expression was measured by the expression of green fluorescent protein (GFP) from the endogenous E1A promoter in AdV-C5_E1A-FS2A-GFP and found to be highly correlated with the viral plaque formation, indicating reliability of the reporter virus. As an ensemble, randomly picked clonal A549 cell isolates displayed significantly higher coefficients of variation in the E1A expression than technical noise, indicating a phenotypic variability larger than noise. The underlying cell state determining infection variability was maintained for at least 9 weeks of cell cultivation. Our results indicate that preexisting cell states tune adenovirus infection in favor of the cell or the virus. These findings have implications for antiviral strategies and gene therapy applications.IMPORTANCEViral infections are known for their variability. Underlying mechanisms are still incompletely understood but have been associated with particular cell states, for example, the eukaryotic cell division cycle in DNA virus infections. A cell state is the collective of biochemical, morphological, and contextual features owing to particular conditions or at random. It affects how intrinsic or extrinsic cues trigger a response, such as cell division or anti-viral state. Here, we provide evidence that cell states with a built-in memory confer high or low susceptibility of clonal human epithelial cells to adenovirus infection. Results are reminiscent of the Luria-Delbrück fluctuation test with bacteriophage infections back in 1943, which demonstrated that mutations, in the absence of selective pressure prior to infection, cause infection resistance rather than being a consequence of infection. Our findings of dynamic cell states conferring adenovirus infection susceptibility uncover new challenges for the prediction and treatment of viral infections.

A method for in silico exploration of potential glioblastoma multiforme attractors using single-cell RNA sequencing

We presented a method to find potential cancer attractors using single-cell RNA sequencing (scRNA-seq) data. We tested our method in a Glioblastoma Multiforme (GBM) dataset, an aggressive brain tumor presenting high heterogeneity. Using the cancer attractor concept, we argued that the GBM's underlying dynamics could partially explain the observed heterogeneity, with the dataset covering a representative region around the attractor. Exploratory data analysis revealed promising GBM's cellular clusters within a 3-dimensional marker space. We approximated the clusters' centroid as stable states and each cluster covariance matrix as defining confidence regions. To investigate the presence of attractors inside the confidence regions, we constructed a GBM gene regulatory network, defined a model for the dynamics, and prepared a framework for parameter estimation. An exploration of hyperparameter space allowed us to sample time series intending to simulate myriad variations of the tumor microenvironment. We obtained different densities of stable states across gene expression space and parameters displaying multistability across different clusters. Although we used our methodological approach in studying GBM, we would like to highlight its generality to other types of cancer. Therefore, this report contributes to an advance in the simulation of cancer dynamics and opens avenues to investigate potential therapeutic targets.