Biology subjects

van Nimwegen, E.

Publications and source records attributed to van Nimwegen, E..

14 recordsLinked to original sources

bioRxiv2026

Single-cell growth rate fluctuations contribute independently from protein production to gene expression noise and decrease with average growth rate

Phenotypic heterogeneity is a universal feature in microbial life and has profound consequences for the behavior of single cells. However, dissecting and quantifying the sources of phenotypic noise is challenging as it requires precise quantification of the dynamics of gene expression fluctuations in vivo. Here, we analyze time-lapse microscopy data of single E. coli cells carrying fluorescent reporters for constitutive and ribosomal promoters to quantify how fluctuations in protein production, growth rate, and random sampling at division contribute to noise in protein concentrations. We find that growth rate fluctuations are largely uncorrelated with protein production and contribute substantially to noise in protein concentrations. In addition, as nutrients are varied, growth rate fluctuations decrease both in amplitude and duration with the mean growth rate. In contrast, fluctuations in production do not show a systematic dependence on mean growth rate, but exhibit surprisingly variable behaviors, even across different constitutive promoters. Moreover, although production and growth rate fluctuations are only weakly correlated, they exhibit complex cross-correlation patterns with growth rate that are both condition-and promoter-dependent. These observations suggest that, while growth rate fluctuations might be dominated by a single underlying mechanism, there are multiple sources of protein production fluctuations that vary in strength across conditions and affect different promoters to different extents. Finally, the sampling noise of proteins at cell division only contributes substantially to protein concentration noise for low-expressed genes with ten or fewer molecules per cell.

systems biology↗

bioRxiv2025

RealTrace: Uncovering biological dynamics hidden under measurement noise in time-lapse microscopy data

One of the most powerful approaches for identifying the mechanisms underlying complex biological phenomena is not just to measure bulk populations, or even take single-cell snapshots, but to directly track the behaviors of single cells in time. Indeed, in recent years there has been a steep rise in the use of time-lapse microscopy with fluorescent reporters to quantify single-cell time dynamics in many different biological systems. However, there is a major challenge with the analysis of such data. Since biological changes are inherently small on short time scales, much of the true biological dynamics is hidden under unavoidable measurement noise, but we do not know what functional form the time dynamics may take, and smoothing over neighboring time points simply replaces the true dynamics with a time average. Here, we present RealTrace, a Bayesian method that rigorously solves this challenge using only the assumption that, while true biological fluctuations are correlated on short time scales, measurement errors are uncorrelated. We show that this assumption can be formalized in a Bayesian model using maximum entropy process priors and that this model can be accurately solved by recursively approximating the non-linear dynamics over short time intervals. Given raw input cell size and fluorescence measurements, RealTrace calculates the estimated true sizes, total GFP contents, instantaneous growth rates, and volumic GFP production rates across all cells and time, together with rigorous error bars on all these quantities. We demonstrate the broad applicability of RealTrace by analyzing time-lapse microscopy datasets from E. coli bacteria, mouse embryonic stem cells, and entire C. elegans embryos. To exemplify the subtle dynamical features RealTrace can uncover, we perform an in-depth study of E. coli cells carrying fluorescent reporters for constitutive and ribosomal promoters, across multiple growth conditions. We find growth rates of single cells substantially fluctuate in time and, as the average growth rate decreases, these fluctuations systematically increase in amplitude and duration. Both growth rate and gene expression vary systematically across the cell cycle, but while the pattern of growth rate variation is condition-independent, the systematic variation in gene expression across the cell cycle is both condition and promoter-dependent. Finally, under a sudden change in nutrients, single cells exhibit highly consistent transient changes in growth rate and gene expression. While gene expression rates settle into a new steady-state surprisingly rapidly, the growth rate shows a dramatic drop followed by an overshoot that takes several cell cycles to settle down. All these observations give fundamental new insights into the single-cell physiology of bacteria. RealTrace is a highly versatile method that accurately removes measurement noise from any fluorescence time-lapse microscopy data, enabling the identification and quantification of subtle features in the biological dynamics that would otherwise be hidden under measurement noise.

bioinformatics↗

bioRxiv2025

Bonsai: Tree representations for distortion-free visualization and exploratory analysis of single-cell omics data

Single-cell omics methods promise to revolutionize our understanding of gene regulatory processes during cell differentiation, but analysis of such data continues to pose a major challenge. Apart from technical challenges such as the sparsity and heterogeneous noise properties of these data, the crucial problem is that we know little about the potentially very complex high-dimensional structures that the data represent. Consequently, there is an urgent need for exploratory analysis methods that allow rigorous representation and visualization of the structure in the data. However, currently popular methods such as UMAP and t-SNE are unsatisfactory because they are ad hoc, stochastic, uninterpretable, and known to severely distort the structure in the data. Here we show that these challenges can be overcome by representing the data on tree structures and present Bonsai : a novel method that reconstructs the most likely tree relating any set of high-dimensional objects while rigorously accounting for heterogeneous measurement noise. We show that, in contrast to other visualization methods, distances along the Bonsai trees accurately represent true distances between the objects in high-dimensional space across many types of datasets. Moreover, Bonsai automatically regularizes measurement noise, outperforming even methods specifically designed for that purpose on tasks such as nearest-neighbor identification. By analyzing a blood cell dataset, we show that Bonsai trees not only capture known lineage relationships but also provide novel biological insights. For example, Bonsai uncovers that different subsets of NK cells derive from the myeloid and lymphoid lineage, and pinpoints genes that distinguish myeloid-NK from lymphoid-NK cells. Bonsai is free from tunable parameters and scales to datasets of hundreds of thousands of cells. The accompanying tool, Bonsai-scout, provides visualizations of the Bonsai trees and allows for interactive data exploration such as identifying subclades and their markers, visualizing features along the tree, changing the tree layout, and zooming in on substructures. Finally, application to a dataset of football statistics shows the generality of Bonsai in successfully capturing complex structures in high-dimensional data.

bioinformatics↗

bioRxiv2024

Transient transcription factor depletions explain diverse single-cell responses of LexA target promoters to mild DNA damage

In bacteria, the effects of transcription factors (TFs) on the expression of their target genes are highly stochastic at the single-cell level. Not only do TF concentrations fluctuate in time and from cell to cell, in each cell the actual binding and unbinding of TFs to promoters is also stochastic. However, how this noise propagation from TFs to their targets determines the expression fluctuations in target genes of the same TF has so far not been quantitatively characterized. Here we use fluorescence time-lapse microscopy in combination with microfluidics and automated image analysis to quantitatively track the single-cell growth and gene expression dynamics of different target promoters of the LexA SOS response regulator in E. coli under mild DNA damage induced by the antibiotic ciprofloxacin (CIP). We find that, at the single-cell level, LexA target promoters exhibit stochastic bursts in protein production with specific quantitative characteristics. First, the durations of protein production bursts are short, identical for all promoters, and independent of CIP levels. Second, the heights of the production peaks are exponentially distributed for each promoter, and independent of CIP level, but with different average heights for different promoters. Third, the frequency of the production bursts increases with CIP level for all promoters, but differs across promoters. Importantly, we show that these observations are inconsistent with popular equilibrium models of bacterial gene regulation that assume that the dynamics of binding and unbinding of TFs is fast relative to the time-scale of TF concentration fluctuations. Instead, all observations can be quantitatively fit by a relatively simple model in which DNA damage events lead to short transient dips in LexA concentration, and an expression burst occurs in a cell whenever LexA unbinds from the target promoter during such a short dip in LexA concentration. Moreover, because these transient dips in LexA concentration occur on the same time-scale as the stochastic unbinding of LexA from promoters, the response of a target promoter to an individual DNA damage event is stochastic and depends in a complex manner on the strengths of its LexA binding sites and the duration of the event. Consequently, the expression levels of different LexA target genes in a given cell reflect distinct statistics of the frequency and durations of DNA damage events in the recent history of that cell. Our results show that, even in the simple scenario where a single TF is induced, different target promoters can exhibit distinct stochastic responses that depend on the precise kinetics of the induction events.

systems biology↗

bioRxiv2024

E. coli leverages growth arrest to remodel its proteome upon entry into starvation

It is widely believed that due to nutrient limitations in natural environments, bacteria spend most of their life in non-growing states. However, very little is known about how bacteria change their phenotype during starvation and what controls the concentration of different gene products inside the cells. Here we used microfluidics with quantitative fluorescence microscopy to quantitatively monitor growth and gene expression in many independent single-cell E. coli lineages as cells were switched from exponential growth to carbon starvation. In contrast to the hypothesis that stationary phase at the population level may reflect a balance between continuing growth and death in different sub-populations, we found that all cells immediately enter growth arrest, that cells further in their cell cycle subsequently undergo reductive division, and no cell death occurs for more than two days. Second, we observed dramatic time-dependent changes in protein production that are highly homogeneous across single cells. Some promoters shut off protein production immediately, some show a slow exponential decay of production on a 10 h time scale, while others exhibit a transient burst of increased production before decaying exponentially at different rates. Notably, the reduction in protein production 30-60 h into starvation relative to production in exponential phase varies by more than two orders of magnitude across promoters and is highly correlated with production in the first 10 h of starvation. Control experiments show that protein degradation itself also decays expo-nentially and using mathematical modeling we show how the fold-change in a genes protein concentration between exponential phase and late starvation depends on the size of the expression burst at the onset of starvation, the rate of subsequent production decay, and the rate of degradation decay. For many genes, the expression in late starvation is driven by production during the first 10 h. Finally, we establish that this expression program at the onset of starvation is critical for cell viability. In particular, by inhibiting gene expression during different periods of starvation, we show that tolerance to stress later in starvation is determined by gene expression occurring during the first 10 h. Our study provides a foundation for quantitative studies of bacterial starvation by uncovering a gene expression program that fundamentally remodels the proteome during the first 10 h of starvation, is highly homogeneous across single cells, sets the proteome later in starvation, and is crucial for stress tolerance.

microbiology↗

bioRxiv2023

Identifying cell states in single-cell RNA-seq data at statistically maximal resolution

Single-cell RNA sequencing (scRNA-seq) has become a popular experimental method to study variation of gene expression within a population of cells. However, obtaining an accurate picture of the diversity of distinct gene expression states that are present in a given dataset is highly challenging because the sparsity of the scRNA-seq data and its inhomogeneous measurement noise properties. Although a vast number of different methods is applied in the literature for clustering cells into subsets with similar expression profiles, these methods generally lack rigorously specified objectives, involve multiple complex layers of normalization, filtering, feature selection, dimensionalityreduction, employ ad hoc measures of distance or similarity between cells, often ignore the known measurement noise properties of scRNA-seq measurements, and include a large number of tunable parameters. Consequently, it is virtually impossible to assign concrete biophysical meaning to the clusterings that result from these methods. Here we address the following problem: Given raw unique molecule identifier (UMI) counts of an scRNA-seq dataset, partition the cells into subsets such that the gene expression states of the cells in each subset are statistically indistinguishable, and each subset corresponds to a distinct gene expression state. That is, we aim to partition cells so as to maximally reduce the complexity of the dataset without removing any of its meaningful structure. We show that, given the known measurement noise structure of scRNA-seq data, this problem is mathematically well-defined and derive its unique solution from first principles. We have implemented this solution in a tool called CO_SCPLOWELLSTATESC_SCPLOW which operates directly on the raw data and automatically determines the optimal partition and cluster number, with zero tunable parameters. We show that, on synthetic datasets, CO_SCPLOWELLSTATESC_SCPLOW almost perfectly recovers optimal partitions. On real data, CO_SCPLOWELLSTATESC_SCPLOW robustly identifies subtle substructure within groups of cells that are traditionally annotated as a common cell type. Moreover, we show that the diversity of gene expression states that CO_SCPLOWELLSTATESC_SCPLOW identifies systematically depends on the tissue of origin and not on technical features of the experiments such as the total number of cells and total UMI count per cell. In addition to the CO_SCPLOWELLSTATESC_SCPLOW tool we also provide a small toolbox of software to place the identified cellstates into a hierarchical tree of higher-order clusters, to identify the most important marker genes at each branch of this hierarchy, and to visualize these results.

bioinformatics↗

bioRxiv2022

The protooncogene Ski regulates the neuron-glia switch during development of the mammalian cerebral cortex

The brain is the most complex organ in mammals and understanding the origin of this complexity is a major challenge for developmental biologists. Crucial to the size and morphology of the cortex is the timing and transition of neural stem cell (NSC) fate. An interesting candidate for modulating and fine tuning these processes is the transcriptional regulator Ski, a protooncogene expressed in cortical cells. Ski is involved in diverse cellular processes and epigenetic programs, and mice deficient in Ski exhibit complex central nervous system defects that resemble some of the features observed in patients with 1p36 deletion syndrome and Shprintzen-Goldberg syndrome. Here, we took advantage of in vivo transgenic labeling and next-generation sequencing to analyze the gene expression profiles of NSCs, basal progenitor (BP) cells, and newborn neurons (NBNs) from wildtype and Ski-deficient embryos throughout cortical development. We created a unique database that allowed us to identify and compare signaling pathways and transcriptional networks within each progenitor population in the presence and absence of Ski. We find that NSCs are the most affected cell population and uncover that mutant NSCs fail to switch to a gliogenic fate in time. We show that Ski functions in concert with the Bone Morphogenetic Protein (BMP) signaling pathway to alter the cell differentiation fate of NSCs from neurons to glia, which is key to generating adequate numbers of specific cell types during corticogenesis. Thus, by combining genetic tools and bioinformatic analysis, our work not only deepens the knowledge of how Ski functions in the brain, but also provides an immense resource for studying neurodevelopmental disorders.

systems biology↗

bioRxiv2022

Coupling phenotype stability to growth rate overcomes limitations of bet-hedging strategies

Microbes in the wild face highly variable and unpredictable environments, and are naturally selected for their average growth rate across environments. Apart from using sensory-regulatory systems to adapt in a targeted manner to changing environments, microbes employ bet-hedging strategies where cells in an isogenic population switch stochastically between alternative phenotypes. Yet, bet-hedging suffers from a fundamental trade-off: increasing the phenotype switching rate increases the rate at which maladapted cells explore alternative phenotypes, but also increases the rate at which cells switch out of a well-adapted state. Consequently, it is currently believed that bet-hedging strategies are only effective when the number of possible phenotypes is limited and when environments last for sufficiently many generations. However, recent experimental results show that gene expression noise generally decreases with growth rate, suggesting that phenotype switching rates may systematically decrease with growth rate. We here show that such growth rate dependent stability (GRDS) can almost completely overcome the trade-off that limits bet-hedging, allowing for effective adaptation even when environments are diverse and change rapidly. GRDS allows cells to be more explorative when maladapted, and more phenotypically stable when well-adapted. We further show that even a small decrease in switching rates of faster growing phenotypes can substantially increase long-term fitness of bet-hedging strategies. Together, our results suggest that stochastic strategies may play an even bigger role for microbial adaptation than hitherto appreciated.

evolutionary biology↗

bioRxiv2022

Growth rate controls the sensitivity of gene regulatory circuits

Unicellular organisms adapt to their changing environments by gene regulatory switches that sense chemical cues and induce specific target genes when the inducing signal is over a critical threshold. Using mathematical modeling we here show that, because growth rate sets the dilution rate of intracellular molecules, the sensitivity of gene regulatory switches automatically couples to growth rate, in a way that can be exploited by natural selection. We confirm the modeling predictions by experimentally demonstrating that, as nutrient quality is varied, the concentration of inducer required for activating the lac operon in E. coli increases quadratically with the populations growth rate. Our theory further predicts that, when growth rate is instead modulated by translation inhibition, critical inducer levels are invariant, and we experimentally validate this prediction as well. Moreover, we establish that this growth-coupled sensitivity allows bacteria to implement concentration-dependent sugar preferences, in which a new carbon source is used only if its concentration is high enough to improve upon the current growth rate of the cells. Using microfluidics in combination with time-lapse microscopy, we validate experimentally that this strategy governs how mixtures of glucose and lactose are used in E. coli at single-cell level. Overall, growth-coupled sensitivity provides a general mechanism through which cells can mute external signals in beneficial conditions when growth is fast, and become highly sensitive to alternative nutrients or stresses when growth is slow or arrested.

systems biology↗

bioRxiv2022

Frequent co-regulation of splicing and polyadenylation by RNA-binding proteins inferred with MAPP

Maturation of eukaryotic pre-mRNAs via splicing, 3 end cleavage and polyadenylation is modulated across cell types and conditions by a variety of RNA-binding proteins (RBPs). Although over 1500 proteins are associated with RNAs in human cells, their binding motifs, targets and functions still remain to be elucidated, especially in the complex environment of human tissues and in the context of diseases. To overcome the lack of methods for systematic and automated detection of sequence motif-guided changes in pre-mRNA processing based on RNA sequencing (RNA-seq) data we have developed MAPP (Motif Activity on Pre-mRNA Processing). We demonstrate MAPPs functionality by applying it to RNA-seq data from 284 RBP knock-down experiments in the ENCODE project, from which MAPP not only infers position-dependent impact profiles of known regulators, but also reveals RBPs that modulate both the inclusion of cassette exons and the poly(A) site choice. Among these, the Polypyrimidine Tract Binding Protein 1 (PTBP1) has a similar activity in glioblastoma samples. This highlights the ability of MAPP to unveil global regulators of mRNA processing under physiological and pathological conditions.

systems biology↗

bioRxiv2020

Response to comment on 'Initiation of chromosome replication controls both division and replication cycles in E. coli through a double-adder mechanism'

Last year we published an article (Witz et al., 2019) in which we used time-lapse microscopy in combination with microfluidics to measure growth, division and replication in single E. coli cells on the one hand, and developed a new statistical analysis method to calculate the ability of different cell cycle models to capture the correlation structure observed in the data on the other hand. This led us to propose a new model of cell cycle control in E. coli which we called the double-adder model. Recently Le Treut et al. published a comment (Le Treut et al., 2020) on our article which made a number of highly critical claims, including allegations that our own data support a different model than the one we proposed, and that our model cannot reproduce the adder phenotype observed in the data. We here show that all these allegations are false and based on basic analysis errors. Although our focus is on explaining the errors in the analysis of Le Treut et al, we have attempted to make the presentation of interest to a broader scientific audience by discussing the issues in the context of what our current understanding is of the bacterial cell cycle, and to what extent recent data either support or reject various proposed models.

systems biology↗

bioRxiv2020

Subpopulations of sensorless bacteria drive fitness in fluctuating environments

Populations of bacteria often undergo a lag in growth when switching conditions. Because growth lags can be large compared to typical doubling times, variations in growth lag are an important but often overlooked component of bacterial fitness in fluctuating environments. We here explore how growth lag variation is determined for the archetypical switch from glucose to lactose as a carbon source in E. coli. First, we show that single-cell lags are bimodally distributed and controlled by a single-molecule trigger. That is, gene expression noise causes the population before the switch to divide cells with zero pre-existing into subpopulations with zero and nonzero lac operon expression. While ’sensorless’ lac expression at the switch have long lags because they are unable to sense the lactose signal, any nonzero lac operon expression suffices to ensure a short lag. Second, we show that the growth lag at the population level depends crucially on the fraction of sensorless cells, and that this fraction in turn depends sensitively on the growth condition before the switch. Consequently, even small changes in basal expression affecting the fraction of sensorless cells can significantly affect population lags and fitness under switching conditions, and may thus be subject to significant natural selection. Indeed, we show that condition-dependent population lags vary across wild E. coli isolates. Since many sensory genes are naturally low expressed in conditions where their inducer is not present, bimodal responses due to subpopulations of sensorless cells may be a general mechanism inducing phenotypic heterogeneity and controlling population lags in switching environments. This mechanism also illustrates how gene expression noise can turn even simple sensory gene circuits into a bet-hedging module, and underlines the profound role of gene expression noise in regulatory responses.Competing Interest StatementThe authors have declared no competing interest.View Full Text

systems biology↗

bioRxiv2019

Whole genome phylogenies reflect long-tailed distributions of recombination rates in many bacterial species

Although homologous recombination is accepted to be common in bacteria, so far it has been challenging to accurately quantify its impact on genome evolution within bacterial species. We here introduce methods that use the statistics of single-nucleotide polymorphism (SNP) splits in the core genome alignment of a set of strains to show that, for many bacterial species, recombination dominates genome evolution. Each genomic locus has been overwritten so many times by recombination that it is impossible to reconstruct the clonal phylogeny and, instead of a consensus phylogeny, the phylogeny typically changes many thousands of times along the core genome alignment. We also show how SNP splits can be used to quantify the relative rates with which different subsets of strains have recombined in the past. We find that virtually every strain has a unique pattern of frequencies with which its lineages have recombined with those of other strains, and that the relative rates with which different subsets of strains share SNPs follow long-tailed distributions. Our findings show that bacterial populations are neither clonal nor freely recombining, but structured such that recombination rates between different lineages vary along a continuum spanning several orders of magnitude, with a unique pattern of rates for each lineage. Thus, rather than reflecting clonal ancestry, whole genome phylogenies reflect these long-tailed distributions of recombination rates.

evolutionary biology↗

bioRxiv2017

Discovery of global regulators of 3’ untranslated region processing in cancers with KAPAC

3 UTR length is regulated in relation to cellular state. To uncover key regulators of poly(A) site (PAS) use in specific conditions, we have developed PAQR, a method for quantifying PAS use from RNA sequencing data and KAPAC, an approach that infers activities of oligomeric sequence motifs on PAS choice. Application of PAQR and KAPAC to RNA sequencing data from normal and tumor tissue samples uncovered sequence motifs that can explain changes in cleavage and polyadenylation in specific cancers. In particular, our analysis points to Polypyrimidine tract binding protein 1 as a regulator of PAS choice in glioblastoma.

systems biology↗