A genome-wide analysis of promoter-mediated phenotypic noise in Escherichia coli

Genetically encoded transcriptional plasticity underlies stress adaptation in Mycobacterium tuberculosis

Transcriptional regulation is a critical adaptive mechanism that allows bacteria to respond to changing environments, yet the concept of transcriptional plasticity (TP) - the variability of gene expression in response to environmental changes - remains largely unexplored. In this study, we investigate the genome-wide TP profiles of Mycobacterium tuberculosis (Mtb) genes by analyzing 894 RNA sequencing samples derived from 73 different environmental conditions. Our data reveal that Mtb genes exhibit significant TP variation that correlates with gene function and gene essentiality. We also find that critical genetic features, such as gene length, GC content, and operon size independently impose constraints on TP, beyond trans-regulation. By extending our analysis to include two other Mycobacterium species -- M. smegmatis and M. abscessus -- we demonstrate a striking conservation of the TP landscape. This study provides a comprehensive understanding of the TP exhibited by mycobacteria genes, shedding light on this significant, yet understudied, genetic feature encoded in bacterial genomes.

Phenotypic Patterning through Copy Number Adaptation to Environmental Gradients

We recently described a paradigm for engineering bacterial adaptation using plasmids coupled to the same origin of replication. In this study, we use plasmid coupling to generate spatially separated and phenotypically distinct populations in response to heterogeneous environments. Using a custom microfluidic device, we continuously tracked engineered populations along induced gradients, enabling an in-depth analysis of the spatiotemporal dynamics of plasmid coupling. Our observations reveal a pronounced phenotypic separation within 4 h exposure to an opposing gradient of AHL and arabinose. Additionally, by modulating the burden strength balance between coupled plasmids, we demonstrate the inherent limitations and tunability of this system. Intriguingly, phenotypic separation persists for an extended time, hinting at a biophysical spatial retention mechanism reminiscent of natural speciation processes. Complementing our experimental data, mathematical models provide invaluable insights into the underlying mechanisms and guide optimization of plasmid coupling for prospective applications of environmental copy number adaptation engineering across separated domains.

Deep-learning-assisted Sort-Seq enables high-throughput profiling of gene expression characteristics with high precision

Owing to the nondeterministic and nonlinear nature of gene expression, the steady-state intracellular protein abundance of a clonal population forms a distribution. The characteristics of this distribution, including expression strength and noise, are closely related to cellular behavior. However, quantitative description of these characteristics has so far relied on arrayed methods, which are time-consuming and labor-intensive. To address this issue, we propose a deep-learning-assisted Sort-Seq approach (dSort-Seq) in this work, enabling high-throughput profiling of expression properties with high precision. We demonstrated the validity of dSort-Seq for large-scale assaying of the dose-response relationships of biosensors. In addition, we comprehensively investigated the contribution of transcription and translation to noise production in Escherichia coli, from which we found that the expression noise is strongly coupled with the mean expression level. We also found that the transcriptional interference caused by overlapping RpoD-binding sites contributes to noise production, which suggested the existence of a simple and feasible noise control strategy in E. coli.

Fitness cost associated with cell phenotypic switching drives population diversification dynamics and controllability

Isogenic cell populations can cope with stress conditions by switching to alternative phenotypes. Even if it can lead to increased fitness in a natural context, this feature is typically unwanted for a range of applications (e.g., bioproduction, synthetic biology, and biomedicine) where it tends to make cellular response unpredictable. However, little is known about the diversification profiles that can be adopted by a cell population. Here, we characterize the diversification dynamics for various systems (bacteria and yeast) and for different phenotypes (utilization of alternative carbon sources, general stress response and more complex development patterns). Our results suggest that the diversification dynamics and the fitness cost associated with cell switching are coupled. To quantify the contribution of the switching cost on population dynamics, we design a stochastic model that let us reproduce the dynamics observed experimentally and identify three diversification regimes, i.e., constrained (at low switching cost), dispersed (at medium and high switching cost), and bursty (for very high switching cost). Furthermore, we use a cell-machine interface called Segregostat to demonstrate that different levels of control can be applied to these diversification regimes, enabling applications involving more precise cellular responses.

Genetically encoded transcriptional plasticity underlies stress adaptation in Mycobacterium tuberculosis

Transcriptional regulation is a critical adaptive mechanism that allows bacteria to respond to changing environments, yet the concept of transcriptional plasticity (TP) remains largely unexplored. In this study, we investigate the genome-wide TP profiles of Mycobacterium tuberculosis (Mtb) genes by analyzing 894 RNA sequencing samples derived from 73 different environmental conditions. Our data reveal that Mtb genes exhibit significant TP variation that correlates with gene function and gene essentiality. We also found that critical genetic features, such as gene length, GC content, and operon size independently impose constraints on TP, beyond trans-regulation. By extending our analysis to include two other Mycobacterium species -- M. smegmatis and M. abscessus -- we demonstrate a striking conservation of the TP landscape. This study provides a comprehensive understanding of the TP exhibited by mycobacteria genes, shedding light on this significant, yet understudied, genetic feature encoded in bacterial genomes.

Genetically encoded transcriptional plasticity underlies stress adaptation in Mycobacterium tuberculosis

Transcriptional regulation is a critical adaptive mechanism that allows bacteria to respond to changing environments, yet the concept of transcriptional plasticity (TP) remains largely unexplored. In this study, we investigate the genome-wide TP profiles of Mycobacterium tuberculosis (Mtb) genes by analyzing 894 RNA sequencing samples derived from 73 different environmental conditions. Our data reveal that Mtb genes exhibit significant TP variation that correlates with gene function and gene essentiality. We also found that critical genetic features, such as gene length, GC content, and operon size independently impose constraints on TP, beyond trans-regulation. By extending our analysis to include two other Mycobacterium species -- M. smegmatis and M. abscessus -- we demonstrate a striking conservation of the TP landscape. This study provides a comprehensive understanding of the TP exhibited by mycobacteria genes, shedding light on this significant, yet understudied, genetic feature encoded in bacterial genomes.

Phenotypic heterogeneity in the bacterial oxidative stress response is driven by cell-cell interactions

Genetically identical bacterial cells commonly display different phenotypes. This phenotypic heterogeneity is well known for stress responses, where it is often explained as bet hedging against unpredictable environmental threats. Here, we explore phenotypic heterogeneity in a major stress response of Escherichia coli and find it has a fundamentally different basis. We characterize the response of cells exposed to hydrogen peroxide (H2O2) stress in a microfluidic device under constant growth conditions. A machine-learning model reveals that phenotypic heterogeneity arises from a precise and rapid feedback between each cell and its immediate environment. Moreover, we find that the heterogeneity rests upon cell-cell interaction, whereby cells shield each other from H2O2 via their individual stress responses. Our work shows how phenotypic heterogeneity in bacterial stress responses can emerge from short-range cell-cell interactions and result in a collective phenotype that protects a large proportion of the population.

Genome-wide inference reveals that feedback regulations constrain promoter-dependent transcriptional burst kinetics

Gene expression in mammalian cells is highly variable and episodic, resulting in a series of discontinuous bursts of mRNAs. A challenge is to understand how static promoter architecture and dynamic feedback regulations dictate bursting on a genome-wide scale. Although single-cell RNA sequencing (scRNA-seq) provides an opportunity to address this challenge, effective analytical methods are scarce. We developed an interpretable and scalable inference framework, which combined experimental data with a mechanistic model to infer transcriptional burst kinetics (sizes and frequencies) and feedback regulations. Applying this framework to scRNA-seq data generated from embryonic mouse fibroblast cells, we found Simpson's paradoxes, i.e. genome-wide burst kinetics exhibit different characteristics in two cases without and with distinguishing feedback regulations. We also showed that feedbacks differently modulate burst frequencies and sizes and conceal the effects of transcription start site distributions on burst kinetics. Notably, only in the presence of positive feedback, TATA genes are expressed with high burst frequencies and enhancer-promoter interactions mainly modulate burst frequencies. The developed inference method provided a flexible and efficient way to investigate transcriptional burst kinetics and the obtained results would be helpful for understanding cell development and fate decision.

Genome-Wide Analysis of Gene Expression Noise Brought About by Transcriptional Regulation in Pseudomonas aeruginosa

The part of expression noise that is brought about by transcriptional regulation (represented here as NTR) is an important criterion for estimating the regulatory mode of a gene. However, characterization of NTR is an under-explored area, and there is little knowledge regarding the genome-wide NTR in the model pathogen Pseudomonas aeruginosa. Here, with a library of dual-color transcriptional reporters, we estimated the NTR for over 90% of the promoters in P. aeruginosa. Most promoters exhibit low NTR, while 42 and 115 promoters with high NTR were screened out in the exponential and the stationary growth phases, respectively. Specifically, a rearrangement of NTR was found in promoters involved in amino acid metabolism when bacteria enter the exponential phase. In addition, during the stationary phase, high NTR was found in a wide range of iron-related promoters involving siderophore synthesis and heme uptake, ExsA-regulated promoters involving bacterial virulence, and FleQ-regulated promoters involving biofilm development. We also found a large-scale negative dependence of transcriptional regulation between high-NTR promoters belonging to different functional categories. Our findings offer a global view of transcriptional heterogeneity in P. aeruginosa. IMPORTANCE The phenotypic diversity of Pseudomonas aeruginosa is frequently observed in research, suggesting that bacteria adopt strategies such as bet-hedging to survive ever-changing environments. Gene expression noise (GEN) is the major source of phenotypic diversity. Large GEN from transcriptional regulation (represented as NTR) represent an evolutionary necessity to maintain the copy number diversity of certain proteins in the population. Here, we provide a system-wide view of NTR in P. aeruginosa under nutrient-rich and stressed conditions. High NTR was found in genes involved in flagella biosynthesis and amino acid metabolism under both conditions. Specially, iron acquisition genes exhibited high NTR in the stressed condition, suggesting a great diversity of iron physiology in P. aeruginosa. We further revealed a global negative dependence of transcriptional regulation between those high-NTR genes under the stressed condition, suggesting a mutually exclusive relationship between different bacterial survival strategies.

Collective decision-making in Pseudomonas aeruginosa involves transient segregation of quorum-sensing activities across cells

A hallmark of bacterial sociality is that groups can coordinate cooperative actions through a cell-to-cell communication process called quorum sensing (QS). QS regulates key bacterial phenotypes such as virulence in infections and digestion of extracellular compounds in the environment. Although QS responses are typically studied as group-level phenotypes, it is unclear whether individuals coordinate their actions at the single-cell level or whether group phenotypes simply reflect the sum of their noisy members. Here, we studied the behavior of Pseudomonas aeruginosa individuals by tracking their temporal commitments to the two intertwined Las and Rhl-QS systems, from low to high population density. Using chromosomally integrated fluorescent gene reporters, we found that QS gene expression (signal, receptor, and cooperative exoproduct) was noisy with heterogeneity peaking during the build-up phase of QS. Moreover, we observed the formation of discrete subgroups of cells that transiently segregate into two gene expression states: low Las-receptor expressers that instantly activate exoproduct production and high Las-receptor expressers with delayed exoproduct production. Later, gene expression activities converged with all cells fully committing to QS. We developed general mathematical models to show that gene expression segregation can mechanistically be spurred by molecular resource limitations during the initiation phase of regulatory cascades such as QS. Moreover, our models indicate that gene expression segregation across cells can operate as a built-in brake enabling a temporary bet-hedging strategy in unpredictable environments. Altogether, our work reveals that studying the behavior of bacterial individuals is key to understanding emergent collective actions at the group level.

Transcription factor binding process is the primary driver of noise in gene expression

Noise in expression of individual genes gives rise to variations in activity of cellular pathways and generates heterogeneity in cellular phenotypes. Phenotypic heterogeneity has important implications for antibiotic persistence, mutation penetrance, cancer growth and therapy resistance. Specific molecular features such as the presence of the TATA box sequence and the promoter nucleosome occupancy have been associated with noise. However, the relative importance of these features in noise regulation is unclear and how well these features can predict noise has not yet been assessed. Here through an integrated statistical model of gene expression noise in yeast we found that the number of regulating transcription factors (TFs) of a gene was a key predictor of noise, whereas presence of the TATA box and the promoter nucleosome occupancy had poor predictive power. With an increase in the number of regulatory TFs, there was a rise in the number of cooperatively binding TFs. In addition, an increased number of regulatory TFs meant more overlaps in TF binding sites, resulting in competition between TFs for binding to the same region of the promoter. Through modeling of TF binding to promoter and application of stochastic simulations, we demonstrated that competition and cooperation among TFs could increase noise. Thus, our work uncovers a process of noise regulation that arises out of the dynamics of gene regulation and is not dependent on any specific transcription factor or specific promoter sequence.

Low protein expression enhances phenotypic evolvability by intensifying selection on folding stability

Protein abundance affects the evolution of protein genotypes, but we do not know how it affects the evolution of protein phenotypes. Here we investigate the role of protein abundance in the evolvability of green fluorescent protein (GFP) towards the novel phenotype of cyan fluorescence. We evolve GFP in E. coli through multiple cycles of mutation and selection and show that low GFP expression facilitates the evolution of cyan fluorescence. A computational model whose predictions we test experimentally helps explain why: lowly expressed proteins are under stronger selection for proper folding, which facilitates their evolvability on short evolutionary time scales. The reason is that high fluorescence can be achieved by either few proteins that fold well or by many proteins that fold less well. In other words, we observe a synergy between a protein's scarcity and its stability. Because many proteins meet the essential requirements for this scarcity-stability synergy, it may be a widespread mechanism by which low expression helps proteins evolve new phenotypes and functions.

Un-biased housekeeping gene panel selection for high-validity gene expression analysis

Differential gene expression normalised to a single housekeeping (HK) is used to identify disease mechanisms and therapeutic targets. HK gene selection is often arbitrary, potentially introducing systematic error and discordant results. Here we examine these risks in a disease model of brain hypoxia. We first identified the eight most frequently used HK genes through a systematic review. However, we observe that in both ex-vivo and in vivo, their expression levels varied considerably between conditions. When applying these genes to normalise expression levels of the validated stroke target gene, inducible Nox4, we obtained opposing results. As an alternative tool for unbiased HK gene selection, software tools exist but are limited to individual datasets lacking genome-wide search capability and user-friendly interfaces. We, therefore, developed the HouseKeepR algorithm to rapidly analyse multiple gene expression datasets in a disease-specific manner and rank HK gene candidates according to stability in an unbiased manner. Using a panel of de novo top-ranked HK genes for brain hypoxia, but not single genes, Nox4 induction was consistently reproduced. Thus, differential gene expression analysis is best normalised against a HK gene panel selected in an unbiased manner. HouseKeepR is the first user-friendly, bias-free, and broadly applicable tool to automatically propose suitable HK genes in a tissue- and disease-dependent manner.

Gene regulation in Escherichia coli is commonly selected for both high plasticity and low noise

Bacteria often respond to dynamically changing environments by regulating gene expression. Despite this regulation being critically important for growth and survival, little is known about how selection shapes gene regulation in natural populations. To better understand the role natural selection plays in shaping bacterial gene regulation, here we compare differences in the regulatory behaviour of naturally segregating promoter variants from Escherichia coli (which have been subject to natural selection) to randomly mutated promoter variants (which have never been exposed to natural selection). We quantify gene expression phenotypes (expression level, plasticity and noise) for hundreds of promoter variants across multiple environments and show that segregating promoter variants are enriched for mutations with minimal effects on expression level. In many promoters, we infer that there is strong selection to maintain high levels of plasticity, and direct selection to decrease or increase cell-to-cell variability in expression. Taken together, these results expand our knowledge of how gene regulation is affected by natural selection and highlight the power of comparing naturally segregating polymorphisms to de novo random mutations to quantify the action of selection.

Fast bacterial growth reduces antibiotic accumulation and efficacy

Phenotypic variations between individual microbial cells play a key role in the resistance of microbial pathogens to pharmacotherapies. Nevertheless, little is known about cell individuality in antibiotic accumulation. Here, we hypothesise that phenotypic diversification can be driven by fundamental cell-to-cell differences in drug transport rates. To test this hypothesis, we employed microfluidics-based single-cell microscopy, libraries of fluorescent antibiotic probes and mathematical modelling. This approach allowed us to rapidly identify phenotypic variants that avoid antibiotic accumulation within populations of Escherichia coli, Pseudomonas aeruginosa, Burkholderia cenocepacia, and Staphylococcus aureus. Crucially, we found that fast growing phenotypic variants avoid macrolide accumulation and survive treatment without genetic mutations. These findings are in contrast with the current consensus that cellular dormancy and slow metabolism underlie bacterial survival to antibiotics. Our results also show that fast growing variants display significantly higher expression of ribosomal promoters before drug treatment compared to slow growing variants. Drug-free active ribosomes facilitate essential cellular processes in these fast-growing variants, including efflux that can reduce macrolide accumulation. We used this new knowledge to eradicate variants that displayed low antibiotic accumulation through the chemical manipulation of their outer membrane inspiring new avenues to overcome current antibiotic treatment failures.

Division of labor and collective functionality in Escherichia coli under acid stress

The acid stress response is an important factor influencing the transmission of intestinal microbes such as the enterobacterium Escherichia coli. E. coli activates three inducible acid resistance systems - the glutamate decarboxylase, arginine decarboxylase, and lysine decarboxylase systems to counteract acid stress. Each system relies on the activity of a proton-consuming reaction catalyzed by a specific amino acid decarboxylase and a corresponding antiporter. Activation of these three systems is tightly regulated by a sophisticated interplay of membrane-integrated and soluble regulators. Using a fluorescent triple reporter strain, we quantitatively illuminated the cellular individuality during activation of each of the three acid resistance (AR) systems under consecutively increasing acid stress. Our studies highlight the advantages of E. coli in possessing three AR systems that enable division of labor in the population, which ensures survival over a wide range of low pH values.

Dynamic gene expression and growth underlie cell-to-cell heterogeneity in Escherichia coli stress response

Individual bacteria that share identical genomes and growth environments can display substantial cell-to-cell differences in expression of stress-response genes and single-cell growth rates. This phenotypic heterogeneity can impact the survival of single cells facing sudden stress. However, the windows of time that cells spend in vulnerable or tolerant states are often unknown. We quantify the temporal expression of a suite of stress-response reporters, while simultaneously monitoring growth. We observe pulsatile expression across genes with a range of stress-response functions, finding that single-cell growth rates are often anticorrelated with reporter levels. These dynamic phenotypic differences have a concrete link to function, in which individual cells undergoing a pulse of elevated expression and slow growth are predisposed to survive antibiotic exposure.

Cell-to-cell heterogeneity in gene expression and growth can have critical functional consequences, such as determining whether individual bacteria survive or die following stress. Although phenotypic variability is well documented, the dynamics that underlie it are often unknown. This information is important because dramatically different outcomes can arise from gradual versus rapid changes in expression and growth. Using single-cell time-lapse microscopy, we measured the temporal expression of a suite of stress-response reporters in Escherichia coli, while simultaneously monitoring growth rate. In conditions without stress, we found several examples of pulsatile expression. Single-cell growth rates were often anticorrelated with reporter levels, with changes in growth preceding changes in expression. These dynamics have functional consequences, which we demonstrate by measuring survival after challenging cells with the antibiotic ciprofloxacin. Our results suggest that fluctuations in both gene expression and growth dynamics in stress-response networks have direct consequences on survival.

Advances in automated real-time flow cytometry for monitoring of bioreactor processes

Flow cytometry and its technological possibilities have greatly advanced in the past decade as analysis tool for single cell properties and population distributions of different cell types in bioreactors. Along the way, some solutions for automated real-time flow cytometry (ART-FCM) were developed for monitoring of bioreactor processes without operator interference over extended periods with variable sampling frequency. However, there is still great potential for ART-FCM to evolve and possibly become a standard application in bioprocess monitoring and process control. This review first addresses different components of an ART-FCM, including the sampling device, the sample-processing unit, the unit for sample delivery to the flow cytometer and the settings for measurement of pre-processed samples. Also, available algorithms are presented for automated data analysis of multi-parameter fluorescence datasets derived from ART-FCM experiments. Furthermore, challenges are discussed for integration of fluorescence-activated cell sorting into an ART-FCM setup for isolation and separation of interesting subpopulations that can be further characterized by for instance omics-methods. As the application of ART-FCM is especially of interest for bioreactor process monitoring, including investigation of population heterogeneity and automated process control, a summary of already existing setups for these purposes is given. Additionally, the general future potential of ART-FCM is addressed.

Application of an Escherichia coli triple reporter strain for at-line monitoring of single-cell physiology during L-phenylalanine production

Biotechnological production processes are sustainable approaches for the production of biobased components such as amino acids for food and feed industry. Scale-up from ideal lab-scale bioreactors to large-scale processes is often accompanied by loss in productivity. This may be related to population heterogeneities of cells originating from isogenic cultures that arise due to dynamic non-ideal conditions in the bioreactor. To better understand this phenomenon, deeper insights into single-cell physiologies in bioprocesses are mandatory before scale-up. Here, a triple reporter strain (3RP) was developed by chromosomally integrating the fluorescent proteins mEmerald, CyOFP1, and mTagBFP2 into the L-phenylalanine producing Escherichia coli strain FUS4 (pF81_kan) to allow monitoring of growth, oxygen availability, and general stress response of the single cells. Functionality of the 3RP was confirmed in well-mixed lab-scale fed-batch processes with glycerol as carbon source in comparison to the strain without fluorescent proteins, leading to no difference in process performance. Fluorescence levels could successfully reflect the course of related process state variables, revealed population heterogeneities during the transition between different process phases and potentially subpopulations that exhibit superior process performance. Furthermore, indications were found for noise in gene expression as regulation strategy against environmental perturbation.

The Role of Integration Host Factor in Escherichia coli Persister Formation

Persisters represent a small subpopulation of cells that are tolerant of killing by antibiotics and are implicated in the recalcitrance of chronic infections to antibiotic therapy. One general theme has emerged regarding persisters formed by different bacterial species, namely, a state of relative dormancy characterized by diminished activity of antibiotic targets. Within this framework, a number of studies have linked persister formation to stochastic decreases in energy-generating components, leading to low ATP and target activity. In this study, we screen knockouts in the main global regulators of Escherichia coli for their effect on persisters. A knockout in integration host factor (IHF) had elevated ATP and a diminished level of persisters. This was accompanied by an overexpression of isocitrate dehydrogenase (Icd) and a downregulation of isocitrate lyase (AceA), two genes located at the bifurcation between the tricarboxylic acid (TCA) cycle and the glyoxylate bypass. Using a translational ihfA-mVenus fusion, we sort out rare bright cells, and this subpopulation is enriched in persisters. Our results suggest that noise in the expression of ihf produces rare cells with low Icd/high AceA, diverting substrates into the glyoxylate bypass, which decreases ATP, leading to antibiotic-tolerant persisters. We further examine noise in a simple model, the lac operon, and show that a knockout of the lacI repressor increases expression of the operon and decreases persister formation. Our results suggest that noise quenching by overexpression serves as a general approach to determine the nature of persister genes in a variety of bacterial species and conditions. IMPORTANCE Persisters are phenotypic variants that survive exposure to antibiotics through temporary dormancy. Mutants with increased levels of persisters have been identified in clinical isolates, and evidence suggests these cells contribute to chronic infections and antibiotic treatment failure. Understanding the underlying mechanism of persister formation and tolerance is important for developing therapeutic approaches to treat chronic infections. In this study, we examine a global regulator, IHF, that plays a role in persister formation. We find that noise in expression of IHF contributes to persister formation, likely by regulating the switch between the TCA cycle that efficiently produces energy and the glyoxylate bypass. We extend this study to a simple model lac operon and show that when grown on lactose as the sole carbon source, noise in its expression influences ATP levels and determines persister formation. This noise is quenched by overexpression of the lac operon, providing a simple approach to test the involvement of a gene in persister formation.

High-Throughput Time-Lapse Fluorescence Microscopy Screening for Heterogeneously Expressed Genes in Bacillus subtilis

Elucidating phenotypic heterogeneity in clonal bacterial populations is important for both the fundamental understanding of bacterial behavior and the synthetic engineering of bacteria in biotechnology. In this study, we present and validate a high-throughput and high-resolution time-lapse fluorescence microscopy-based strategy to easily and systematically screen for heterogeneously expressed genes in the Bacillus subtilis model bacterium. This screen allows detection of expression patterns at high spatial and temporal resolution, which often escape detection by other approaches, and can readily be extrapolated to other bacteria. A proof-of-concept screening in B. subtilis revealed both recognized and yet unrecognized heterogeneously expressed genes, thereby validating the approach.

IMPORTANCE Differential gene expression among isogenic siblings often leads to phenotypic heterogeneity and the emergence of complex social behavior and functional capacities within clonal bacterial populations. Despite the importance of such features for both the fundamental understanding and synthetic engineering of bacterial behavior, approaches to systematically map such population heterogeneity are scarce. In this context, we have elaborated a new time-lapse fluorescence microscopy-based strategy to easily and systematically screen for such heterogeneously expressed genes in bacteria with high resolution and throughput. A proof-of-concept screening in the Bacillus subtilis model bacterium revealed both recognized and yet unrecognized heterogeneously expressed genes, thereby validating our approach.

OUP accepted manuscript

Operons are a hallmark of the genomic and regulatory architecture of prokaryotes. However, the mechanism by which two genes placed far apart gradually come close and form operons remains to be elucidated. Here, we propose a new model of the origin of operons: Mobile genetic elements called insertion sequences can facilitate the formation of operons by consecutive insertion–deletion–excision reactions. This mechanism barely leaves traces of insertion sequences and thus difficult to detect in nature. In this study, as a proof-of-concept, we reproducibly demonstrated operon formation in the laboratory. The insertion sequence IS3 and the insertion sequence excision enhancer are genes found in a broad range of bacterial species. We introduced these genes into insertion sequence-less Escherichia coli and found that, supporting our hypothesis, the activity of the two genes altered the expression of genes surrounding IS3, closed a 2.7 kb gap between a pair of genes, and formed new operons. This study shows how insertion sequences can facilitate the rapid formation of operons through locally increasing the structural mutation rates and highlights how coevolution with mobile elements may shape the organization of prokaryotic genomes and gene regulation.

Genome-wide gene expression noise in Escherichia coli is condition-dependent and determined by propagation of noise through the regulatory network

Although it is well appreciated that gene expression is inherently noisy and that transcriptional noise is encoded in a promoter’s sequence, little is known about the extent to which noise levels of individual promoters vary across growth conditions. Using flow cytometry, we here quantify transcriptional noise in Escherichia coli genome-wide across 8 growth conditions and find that noise levels systematically decrease with growth rate, with a condition-dependent lower bound on noise. Whereas constitutive promoters consistently exhibit low noise in all conditions, regulated promoters are both more noisy on average and more variable in noise across conditions. Moreover, individual promoters show highly distinct variation in noise across conditions. We show that a simple model of noise propagation from regulators to their targets can explain a significant fraction of the variation in relative noise levels and identifies TFs that most contribute to both condition-specific and condition-independent noise propagation. In addition, analysis of the genome-wide correlation structure of various gene properties shows that gene regulation, expression noise, and noise plasticity are all positively correlated genome-wide and vary independently of variations in absolute expression, codon bias, and evolutionary rate. Together, our results show that while absolute expression noise tends to decrease with growth rate, relative noise levels of genes are highly condition-dependent and determined by the propagation of noise through the gene regulatory network.

Genome-wide flow cytometry measurements reveal that gene expression noise in bacteria is highly condition-dependent; while absolute noise levels of all genes decrease with growth-rate, theoretical modeling shows that the relative noise levels of different genes are determined by the propagation of noise through the gene regulatory network (GRN). Thus GRN structure controls not only mean expression but also noise levels.

Cis-Regulatory Logic Produces Gene-Expression Noise Describing Phenotypic Heterogeneity in Bacteria

Gene transcriptional process is random. It occurs in bursts and follows single-molecular kinetics. Intermittent bursts are measured based on their frequency and size. They influence temporal fluctuations in the abundance of total mRNA and proteins by generating distinct transcriptional variations referred to as “noise”. Noisy expression induces uncertainty because the association between transcriptional variation and the extent of gene expression fluctuation is ambiguous. The promoter architecture and remote interference of different cis-regulatory elements are the crucial determinants of noise, which is reflected in phenotypic heterogeneity. An alternative perspective considers that cellular parameters dictating genome-wide transcriptional kinetics follow a universal pattern. Research on noise and systematic perturbations of promoter sequences reinforces that both gene-specific and genome-wide regulation occur across species ranging from bacteria and yeast to animal cells. Thus, deciphering gene-expression noise is essential across different genomics applications. Amidst the mounting conflict, it is imperative to reconsider the scope, progression, and rational construction of diversified viewpoints underlying the origin of the noise. Here, we have established an indication connecting noise, gene expression variations, and bacterial phenotypic variability. This review will enhance the understanding of gene-expression noise in various scientific contexts and applications.

Individual bacteria in structured environments rely on phenotypic resistance to phage

Bacteriophages represent an avenue to overcome the current antibiotic resistance crisis, but evolution of genetic resistance to phages remains a concern. In vitro, bacteria evolve genetic resistance, preventing phage adsorption or degrading phage DNA. In natural environments, evolved resistance is lower possibly because the spatial heterogeneity within biofilms, microcolonies, or wall populations favours phenotypic survival to lytic phages. However, it is also possible that the persistence of genetically sensitive bacteria is due to less efficient phage amplification in natural environments, the existence of refuges where bacteria can hide, and a reduced spread of resistant genotypes. Here, we monitor the interactions between individual planktonic bacteria in isolation in ephemeral refuges and bacteriophage by tracking the survival of individual cells. We find that in these transient spatial refuges, phenotypic resistance due to reduced expression of the phage receptor is a key determinant of bacterial survival. This survival strategy is in contrast with the emergence of genetic resistance in the absence of ephemeral refuges in well-mixed environments. Predictions generated via a mathematical modelling framework to track bacterial response to phages reveal that the presence of spatial refuges leads to fundamentally different population dynamics that should be considered in order to predict and manipulate the evolutionary and ecological dynamics of bacteria–phage interactions in naturally structured environments.

Bacteriophages represent a promising avenue to overcome the current antibiotic resistance crisis, but evolution of phage resistance remains a concern. This study shows that in the presence of spatial refuges, genetic resistance to phage is less of a problem than commonly assumed, but the persistence of genetically susceptible bacteria suggests that eradicating bacterial pathogens from structured environments may require combined phage-antibiotic therapies.

Cell Growth Model with Stochastic Gene Expression Helps Understand the Growth Advantage of Metabolic Exchange and Auxotrophy

During cooperative growth, microbes often experience higher fitness by sharing resources via metabolite exchange. How competitive species evolve to cooperate is, however, not known. Moreover, existing models (based on optimization of steady-state resources or fluxes) are often unable to explain the growth advantage for the cooperating species, even for simple reciprocally cross-feeding auxotrophic pairs. We present here an abstract model of cell growth that considers the stochastic burst-like gene expression of biosynthetic pathways of limiting biomass precursor metabolites and directly connect the amount of metabolite produced to cell growth and division, using a "metabolic sizer/adder" rule. Our model recapitulates Monod's law and yields the experimentally observed right-skewed long-tailed distribution of cell doubling times. The model further predicts the growth effect of secretion and uptake of metabolites by linking it to changes in the internal metabolite levels. The model also explains why auxotrophs may grow faster when supplied with the metabolite they cannot produce and why two reciprocally cross-feeding auxotrophs can grow faster than prototrophs. Overall, our framework allows us to predict the growth effect of metabolic interactions in independent microbes and microbial communities, setting up the stage to study the evolution of these interactions. IMPORTANCE Cooperative behaviors are highly prevalent in the wild, but their evolution is not understood. Metabolic flux models can demonstrate the viability of metabolic exchange as cooperative interactions, but steady-state growth models cannot explain why cooperators grow faster. We present a stochastic model that connects growth to the cell's internal metabolite levels and quantifies the growth effect of metabolite exchange and auxotrophy. We show that a reduction in gene expression noise can explain why cells that import metabolites or become auxotrophs can grow faster and why reciprocal cross-feeding of metabolites between complementary auxotrophs allows them to grow faster. Furthermore, our framework can simulate the growth of interacting cells, which will enable us to understand the possible trajectories of the evolution of cooperation in silico.

Pareto optimality between growth-rate and lag-time couples metabolic noise to phenotypic heterogeneity in Escherichia coli

Despite mounting evidence that in clonal bacterial populations, phenotypic variability originates from stochasticity in gene expression, little is known about noise-shaping evolutionary forces and how expression noise translates to phenotypic differences. Here we developed a high-throughput assay that uses a redox-sensitive dye to couple growth of thousands of bacterial colonies to their respiratory activity and show that in Escherichia coli, noisy regulation of lower glycolysis and citric acid cycle is responsible for large variations in respiratory metabolism. We found that these variations are Pareto optimal to maximization of growth rate and minimization of lag time, two objectives competing between fermentative and respiratory metabolism. Metabolome-based analysis revealed the role of respiratory metabolism in preventing the accumulation of toxic intermediates of branched chain amino acid biosynthesis, thereby supporting early onset of cell growth after carbon starvation. We propose that optimal metabolic tradeoffs play a key role in shaping and preserving phenotypic heterogeneity and adaptation to fluctuating environments.

Genome-Wide Identification and Expression Analysis of SOS Response Genes in Salmonella enterica Serovar Typhimurium

A bioinformatic search for LexA boxes, combined with transcriptomic detection of loci responsive to DNA damage, identified 48 members of the SOS regulon in the genome of Salmonella enterica serovar Typhimurium. Single cell analysis using fluorescent fusions revealed that heterogeneous expression is a common trait of SOS response genes, with formation of SOS^OFF and SOS^ON subpopulations. Phenotypic cell variants formed in the absence of external DNA damage show gene expression patterns that are mainly determined by the position and the heterology index of the LexA box. SOS induction upon DNA damage produces SOS^OFF and SOS^ON subpopulations that contain live and dead cells. The nature and concentration of the DNA damaging agent and the time of exposure are major factors that influence the population structure upon SOS induction. An analogy can thus be drawn between the SOS response and other bacterial stress responses that produce phenotypic cell variants.

Interplay between genetic, epigenetic, and gene expression variability: Considering complexity in evolvability

Genetic variability, epigenetic variability, and gene expression variability (noise) are generally considered independently in their relationship with phenotypic variation. However, they appear to be intrinsically interconnected and influence it in combination. The study of the interplay between genetic and epigenetic variability has the longest history. This article rather considers the introduction of gene expression variability in its relationships with the two others and reviews for the first time experimental evidences over the four relationships connected to gene expression noise. They show how introducing this third source of variability complicates the way of thinking evolvability and the emergence of biological novelty. Finally, cancer cells are proposed to be an ideal model to decipher the dynamic interplay between genetic, epigenetic, and gene expression variability when one of them is either experimentally increased or therapeutically targeted. This interplay is also discussed in an evolutionary perspective in the context of cancer cell drug resistance.

Microbial metabolic noise

From the time a cell was first placed under the microscope, it became apparent that identifying two clonal cells that "look" identical is extremely challenging. Since then, cell-to-cell differences in shape, size, and protein content have been carefully examined, informing us of the ultimate limits that hinder two cells from occupying an identical phenotypic state. Here, we present recent experimental and computational evidence that similar limits emerge also in cellular metabolism. These limits pertain to stochastic metabolic dynamics and, thus, cell-to-cell metabolic variability, including the resulting adapting benefits. We review these phenomena with a focus on microbial metabolism and conclude with a brief outlook on the potential relationship between metabolic noise and adaptive evolution. This article is categorized under: Metabolic Diseases > Computational Models Metabolic Diseases > Biomedical Engineering.

Using fluorescence flow cytometry data for single-cell gene expression analysis in bacteria

Fluorescence flow cytometry is increasingly being used to quantify single-cell expression distributions in bacteria in high-throughput. However, there has been no systematic investigation into the best practices for quantitative analysis of such data, what systematic biases exist, and what accuracy and sensitivity can be obtained. We investigate these issues by measuring the same E. coli strains carrying fluorescent reporters using both flow cytometry and microscopic setups and systematically comparing the resulting single-cell expression distributions. Using these results, we develop methods for rigorous quantitative inference of single-cell expression distributions from fluorescence flow cytometry data. First, we present a Bayesian mixture model to separate debris from viable cells using all scattering signals. Second, we show that cytometry measurements of fluorescence are substantially affected by autofluorescence and shot noise, which can be mistaken for intrinsic noise in gene expression, and present methods to correct for these using calibration measurements. Finally, we show that because forward- and side-scatter signals scale non-linearly with cell size, and are also affected by a substantial shot noise component that cannot be easily calibrated unless independent measurements of cell size are available, it is not possible to accurately estimate the variability in the sizes of individual cells using flow cytometry measurements alone. To aid other researchers with quantitative analysis of flow cytometry expression data in bacteria, we distribute E-Flow, an open-source R package that implements our methods for filtering debris and for estimating true biological expression means and variances from the fluorescence signal. The package is available at https://github.com/vanNimwegenLab/E-Flow.

Functional roles of microbial cell-to-cell heterogeneity and emerging technologies for analysis and control

Clonal cell populations often display significant cell-to-cell phenotypic heterogeneity, even when maintained under constant external conditions. This variability can result from the inherently stochastic nature of transcription and translation processes, which leads to varying numbers of transcripts and proteins per cell. Here, we showcase studies that reveal links between stochastic cellular events and biological functions in isogenic microbial populations. Then, we highlight emerging tools from engineering, computation, and synthetic and molecular biology that enable precise measurement, control, and analysis of gene expression noise in microorganisms. The capabilities offered by this sophisticated toolbox will shape future directions in the field and generate insight into the behavior of living systems at the single-cell level.

Extrinsic noise prevents the independent tuning of gene expression noise and protein mean abundance in bacteria

It is generally accepted that prokaryotes can tune gene expression noise independently of protein mean abundance by varying the relative levels of transcription and translation. Here, we address this question quantitatively, using a custom-made library of 40 Bacillus subtilis strains expressing a fluorescent protein under the control of different transcription and translation control elements. We quantify noise and mean protein abundance by fluorescence microscopy and show that for most of the natural transcription range of B. subtilis, expression noise is equally sensitive to variations in the transcription or translation rate because of the prevalence of extrinsic noise. In agreement, analysis of whole-genome transcriptomic and proteomic datasets suggests that noise optimization through transcription and translation tuning during evolution may only occur in a regime of weak transcription. Therefore, independent control of mean abundance and noise can rarely be achieved, which has strong implications for both genome evolution and biological engineering.

Contributions to the 'noise floor' in gene expression in a population of dividing cells

Experiments with cells reveal the existence of a lower bound for protein noise, the noise floor, in highly expressed genes. Its origins are still debated. We propose a minimal model of gene expression in a proliferating bacterial cell population. The model predicts the existence of a noise floor and it semi-quantitatively reproduces the curved shape of the experimental noise vs. mean protein concentration plots. When the cell volume increases in a different manner than does the mean protein copy number, the noise floor level is determined by the cell population’s age structure and by the dependence of the mean protein concentration on cell age. Additionally, the noise floor level may depend on a biological limit for the mean number of bursts in the cell cycle. In that case, the noise floor level depends on the burst size distribution width but it is insensitive to the mean burst size. Our model quantifies the contributions of each of these mechanisms to gene expression noise.

Bacterial metabolic heterogeneity: origins and applications in engineering and infectious disease

Bacteria within an isoclonal population display significant heterogeneity in metabolism, even under tightly controlled environmental conditions. Metabolic heterogeneity enables influential functions not possible or measurable at the ensemble scale. Several molecular and cellular mechanisms are likely to give rise to metabolic heterogeneity including molecular noise in metabolic enzyme expression, positive feedback loops, and asymmetric partitioning of cellular components during cell division. Dissection of the mechanistic origins of metabolic heterogeneity has been enabled by recent developments in single-cell analytical tools. Finally, we provide a discussion of recent studies examining the importance of metabolic heterogeneity in applied settings such as infectious disease and metabolic engineering.

Large-scale analysis of human gene expression variability associates highly variable drug targets with lower drug effectiveness and safety

Motivation

The effectiveness of drugs tends to vary between patients. One of the well-known reasons for this phenomenon is genetic polymorphisms in drug target genes among patients. Here, we propose that differences in expression levels of drug target genes across individuals can also contribute to this phenomenon.

Results

To explore this hypothesis, we analyzed the expression variability of protein-coding genes, and particularly drug target genes, across individuals. For this, we developed a novel variability measure, termed local coefficient of variation (LCV), which ranks the expression variability of each gene relative to genes with similar expression levels. Unlike commonly used methods, LCV neutralizes expression levels biases without imposing any distribution over the variation and is robust to data incompleteness. Application of LCV to RNA-sequencing profiles of 19 human tissues and to target genes of 1076 approved drugs revealed that drug target genes were significantly more variable than protein-coding genes. Analysis of 113 drugs with available effectiveness scores showed that drugs targeting highly variable genes tended to be less effective in the population. Furthermore, comparison of approved drugs to drugs that were withdrawn from the market showed that withdrawn drugs targeted significantly more variable genes than approved drugs. Last, upon analyzing gender differences we found that the variability of drug target genes was similar between men and women. Altogether, our results suggest that expression variability of drug target genes could contribute to the variable responsiveness and effectiveness of drugs, and is worth considering during drug treatment and development.

Availability and implementation

LCV is available as a python script in GitHub (https://github.com/eyalsim/LCV).

Supplementary information

Supplementary data are available at Bioinformatics online.

Molecular design of a signaling system influences noise in protein abundance under acid stress in different γ-Proteobacteria

Bacteria have evolved different signaling systems to sense and adapt to acid stress. One of these systems, the CadABC-system, responds to a combination of low pH and lysine availability. In Escherichia coli, the two signals are sensed by the pH sensor and transcription activator CadC and the co-sensor LysP, a lysine-specific transporter. Activated CadC promotes the transcription of the cadBA operon, which codes for the lysine decarboxylase CadA and the lysine/cadaverine antiporter CadB. The copy number of CadC is controlled translationally. Using a bioinformatics approach, we identified the presence of CadC with ribosomal stalling motifs together with LysP in species of the Enterobacteriaceae family. In contrast, we identified CadC without stalling motifs in species of the Vibrionaceae family, but the LysP co-sensor was not identified. Therefore, we compared the output of the Cad system in single cells of the distantly related organisms E. coli and V. campbellii using fluorescently-tagged CadB as the reporter. We observed a heterogeneous output in E. coli, and all the V. campbellii cells produced CadB. The copy number of the pH sensor CadC in E. coli was extremely low (≤4 molecules per cell), but it was 10-fold higher in V. campbellii An increase in the CadC copy number in E. coli correlated with a decrease in heterogeneous behavior. This study demonstrated how small changes in the design of a signaling system allow a homogeneous output and, thus, adaptation of Vibrio species that rely on the CadABC-system as the only acid resistance system.Importance Acid resistance is an important property of bacteria, such as Escherichia coli, to survive acidic environments like the human gastrointestinal tract. E. coli possess both passive and inducible acid resistance systems to counteract acidic environments. Thus, E. coli evolved sophisticated signaling systems to sense and appropriately respond to environmental acidic stress by regulating the activity of its three inducible acid resistance systems. One of these systems is the Cad system that is only induced under moderate acidic stress in a lysine-rich environment by the pH-responsive transcriptional regulator CadC. The significance of our research is in identifying the molecular design of the Cad systems in different Proteobacteria and their target expression noise at single cell level during acid stress conditions.

Metabolic Heterogeneity and Cross-Feeding in Bacterial Multicellular Systems

Cells in assemblages differentiate and perform distinct roles. Though many pathways of differentiation are understood at the molecular level in multicellular eukaryotes, the elucidation of similar processes in bacterial assemblages is recent and ongoing. Here, we discuss examples of bacterial differentiation, focusing on cases in which distinct metabolisms coexist and those that exhibit cross-feeding, with one subpopulation producing substrates that are metabolized by a second subpopulation. We describe several studies of single-species systems, then segue to studies of multispecies metabolic heterogeneity and cross-feeding in the clinical setting. Many of the studies described exemplify the application of new techniques and modeling approaches that provide insights into metabolic interactions relevant for bacterial growth outside the laboratory.

microRNA-mediated noise processing in cells: A fight or a game?

In the past decades, microRNAs (miRNA) have much attracted the attention of researchers at the interface between life and theoretical sciences for their involvement in post-transcriptional regulation and related diseases. Thanks to the always more sophisticated experimental techniques, the role of miRNAs as "noise processing units" has been further elucidated and two main ways of miRNA noise-control have emerged by combinations of theoretical and experimental studies. While on one side miRNAs were thought to buffer gene expression noise, it has recently been suggested that miRNAs could also increase the cell-to-cell variability of their targets. In this Mini Review, we focus on the role of miRNAs in molecular noise processing and on the advantages as well as current limitations of theoretical modelling.

Development and characterization of Escherichia coli triple reporter strains for investigation of population heterogeneity in bioprocesses

Background

Today there is an increasing demand for high yielding robust and cost efficient biotechnological production processes. Although cells in these processes originate from isogenic cultures, heterogeneity induced by intrinsic and extrinsic influences is omnipresent. To increase understanding of this mechanistically poorly understood phenomenon, advanced tools that provide insights into single cell physiology are needed.

Results

Two Escherichia coli triple reporter strains have been designed based on the industrially relevant production host E. coli BL21(DE3) and a modified version thereof, E. coli T7E2. The strains carry three different fluorescence proteins chromosomally integrated. Single cell growth is followed with EmeraldGFP (EmGFP)-expression together with the ribosomal promoter rrnB. General stress response of single cells is monitored by expression of sigma factor rpoS with mStrawberry, whereas expression of the nar-operon together with TagRFP657 gives information about oxygen limitation of single cells. First, the strains were characterized in batch operated stirred-tank bioreactors in comparison to wildtype E. coli BL21(DE3). Afterwards, applicability of the triple reporter strains for investigation of population heterogeneity in bioprocesses was demonstrated in continuous processes in stirred-tank bioreactors at different growth rates and in response to glucose and oxygen perturbation simulating gradients on industrial scale. Population and single cell level physiology was monitored evaluating general physiology and flow cytometry analysis of fluorescence distributions of the triple reporter strains. Although both triple reporter strains reflected physiological changes that were expected based on the expression characteristics of the marker proteins, the triple reporter strain based on E. coli T7E2 showed higher sensitivity in response to environmental changes. For both strains, noise in gene expression was observed during transition from phases of non-growth to growth. Apparently, under some process conditions, e.g. the stationary phase in batch cultures, the fluorescence response of EmGFP and mStrawberry is preserved, whereas TagRFP657 showed a distinct response.

Conclusions

Single cell growth, general stress response and oxygen limitation of single cells could be followed using the two triple reporter strains developed in this study. They represent valuable tools to study population heterogeneity in bioprocesses significantly increasing the level of information compared to the use of single reporter strains.

Gene expression noise in a complex artificial toxin expression system

Gene expression is an intrinsically stochastic process. Fluctuations in transcription and translation lead to cell-to-cell variations in mRNA and protein levels affecting cellular function and cell fate. Here, using fluorescence time-lapse microscopy, we quantify noise dynamics in an artificial operon in Escherichia coli, which is based on the native operon of ColicinE2, a toxin. In the natural system, toxin expression is controlled by a complex regulatory network; upon induction of the bacterial SOS response, ColicinE2 is produced (cea gene) and released (cel gene) by cell lysis. Using this ColicinE2-based operon, we demonstrate that upon induction of the SOS response noise of cells expressing the operon is significantly lower for the (mainly) transcriptionally regulated gene cea compared to the additionally post-transcriptionally regulated gene cel. Likewise, we find that mutations affecting the transcriptional regulation by the repressor LexA do not significantly alter the population noise, whereas specific mutations to post-transcriptionally regulating units, strongly influence noise levels of both genes. Furthermore, our data indicate that global factors, such as the plasmid copy number of the operon encoding plasmid, affect gene expression noise of the entire operon. Taken together, our results provide insights on how noise in a native toxin-producing operon is controlled and underline the importance of post-transcriptional regulation for noise control in this system.

Stochastic Gene Expression Influences the Selection of Antibiotic Resistance Mutations

Bacteria can resist antibiotics by expressing enzymes that remove or deactivate drug molecules. Here, we study the effects of gene expression stochasticity on efflux and enzymatic resistance. We construct an agent-based model that stochastically simulates multiple biochemical processes in the cell and we observe the growth and survival dynamics of the cell population. Resistance-enhancing mutations are introduced by varying parameters that control the enzyme expression or efficacy. We find that stochastic gene expression can cause complex dynamics in terms of survival and extinction for these mutants. Regulatory mutations, which augment the frequency and duration of resistance gene transcription, can provide limited resistance by increasing mean expression. Structural mutations, which modify the enzyme or efflux efficacy, provide most resistance by improving the binding affinity of the resistance protein to the antibiotic; increasing the enzyme's catalytic rate alone may contribute to resistance if drug binding is not rate limiting. Overall, we identify conditions where regulatory mutations are selected over structural mutations, and vice versa. Our findings show that stochastic gene expression is a key factor underlying efflux and enzymatic resistances and should be taken into consideration in future antibiotic research.

Human gene expression variability and its dependence on methylation and aging

BACKGROUND

Phenotypic variability of human populations is partly the result of gene polymorphism and differential gene expression. As such, understanding the molecular basis for diversity requires identifying genes with both high and low population expression variance and identifying the mechanisms underlying their expression control. Key issues remain unanswered with respect to expression variability in human populations. The role of gene methylation as well as the contribution that age, sex and tissue-specific factors have on expression variability are not well understood.

RESULTS

Here we used a novel method that accounts for sampling error to classify human genes based on their expression variability in normal human breast and brain tissues. We find that high expression variability is almost exclusively unimodal, indicating that variance is not the result of segregation into distinct expression states. Genes with high expression variability differ markedly between tissues and we find that genes with high population expression variability are likely to have age-, but not sex-dependent expression. Lastly, we find that methylation likely has a key role in controlling expression variability insofar as genes with low expression variability are likely to be non-methylated.

CONCLUSIONS

We conclude that gene expression variability in the human population is likely to be important in tissue development and identity, methylation, and in natural biological aging. The expression variability of a gene is an important functional characteristic of the gene itself and the classification of a gene as one with Hyper-Variability or Hypo-Variability in a human population or in a specific tissue should be useful in the identification of important genes that functionally regulate development or disease.

Quantitative Flow Cytometry to Understand Population Heterogeneity in Response to Changes in Substrate Availability in Escherichia coli and Saccharomyces cerevisiae Chemostats

Microbial cells in bioprocesses are usually described with averaged parameters. But in fact, single cells within populations vary greatly in characteristics such as stress resistance, especially in response to carbon source gradients. Our aim was to introduce tools to quantify population heterogeneity in bioprocesses using a combination of reporter strains, flow cytometry, and easily comprehensible parameters. We calculated mean, mode, peak width, and coefficient of variance to describe distribution characteristics and temporal shifts in fluorescence intensity. The skewness and the slope of cumulative distribution function plots illustrated differences in distribution shape. These parameters are person-independent and precise. We demonstrated this by quantifying growth-related population heterogeneity of Saccharomyces cerevisiae and Escherichia coli reporter strains in steady-state of aerobic glucose-limited chemostat cultures at different dilution rates and in response to glucose pulses. Generally, slow-growing cells showed stronger responses to glucose excess than fast-growing cells. Cell robustness, measured as membrane integrity after exposure to freeze-thaw treatment, of fast-growing cells was strongly affected in subpopulations of low membrane robustness. Glucose pulses protected subpopulations of fast-growing but not slower-growing yeast cells against membrane damage. Our parameters could successfully describe population heterogeneity, thereby revealing physiological characteristics that might have been overlooked during traditional averaged analysis.

Metabolic activity affects the response of single cells to a nutrient switch in structured populations

Microbes live in ever-changing environments where they need to adapt their metabolism to different nutrient conditions. Many studies have characterized the response of genetically identical cells to nutrient switches in homogeneous cultures; however, in nature, microbes often live in spatially structured groups such as biofilms where cells can create metabolic gradients by consuming and releasing nutrients. Consequently, cells experience different local microenvironments and vary in their phenotype. How does this phenotypic variation affect the ability of cells to cope with nutrient switches? Here, we address this question by growing dense populations of Escherichia coli in microfluidic chambers and studying a switch from glucose to acetate at the single-cell level. Before the switch, cells vary in their metabolic activity: some grow on glucose, while others cross-feed on acetate. After the switch, only few cells can resume growth after a period of lag. The probability to resume growth depends on a cells' phenotype prior to the switch: it is highest for cells cross-feeding on acetate, while it depends in a non-monotonic way on the growth rate for cells growing on glucose. Our results suggest that the strong phenotypic variation in spatially structured populations might enhance their ability to cope with fluctuating environments.

Plasmids for Independently Tunable, Low-Noise Expression of Two Genes

Microbiologists often express foreign proteins in bacteria in order study them or to use bacteria as a microbial factory. Usually, this requires controlling the number of foreign proteins expressed in each cell, but for many common protein expression systems, it is difficult to “tune” protein expression without large cell-to-cell variation in expression levels (called “noise” in protein expression). This work describes two protein expression systems that can be combined in the same cell, with tunable expression levels and very low protein expression noise. One new system was used to detect single mRNA molecules by fluorescence microscopy, and the two systems were shown to be independent of each other. These protein expression systems may be useful in any experiment or biotechnology application that can be improved with low protein expression noise.

Some microbiology experiments and biotechnology applications can be improved if it is possible to tune the expression of two different genes at the same time with cell-to-cell variation at or below the level of genes constitutively expressed from the chromosome (the “extrinsic noise limit”). This was recently achieved for a single gene by exploiting negative autoregulation by the tetracycline repressor (TetR) and bicistronic gene expression to reduce gene expression noise. We report new plasmids that use the same principles to achieve simultaneous, low-noise expression for two genes in Escherichia coli. The TetR system was moved to a compatible plasmid backbone, and a system based on the lac repressor (LacI) was found to also exhibit gene expression noise below the extrinsic noise limit. We characterized gene expression mean and noise across the range of induction levels for these plasmids, applied the LacI system to tune expression for single-molecule mRNA detection under two different growth conditions, and showed that two plasmids can be cotransformed to independently tune expression of two different genes.

IMPORTANCE Microbiologists often express foreign proteins in bacteria in order study them or to use bacteria as a microbial factory. Usually, this requires controlling the number of foreign proteins expressed in each cell, but for many common protein expression systems, it is difficult to “tune” protein expression without large cell-to-cell variation in expression levels (called “noise” in protein expression). This work describes two protein expression systems that can be combined in the same cell, with tunable expression levels and very low protein expression noise. One new system was used to detect single mRNA molecules by fluorescence microscopy, and the two systems were shown to be independent of each other. These protein expression systems may be useful in any experiment or biotechnology application that can be improved with low protein expression noise.

Gene Expression Noise Produces Cell-to-Cell Heterogeneity in Eukaryotic Homologous Recombination Rate

Variation in gene expression among genetically identical individual cells (called gene expression noise) directly contributes to phenotypic diversity. Whether such variation can impact genome stability and lead to variation in genotype remains poorly explored. We addressed this question by investigating whether noise in the expression of genes affecting homologous recombination (HR) activity either directly (RAD52) or indirectly (RAD27) confers cell-to-cell heterogeneity in HR rate in Saccharomyces cerevisiae. Using cell sorting to isolate subpopulations with various expression levels, we show that spontaneous HR rate is highly heterogeneous from cell-to-cell in clonal populations depending on the cellular amount of proteins affecting HR activity. Phleomycin-induced HR is even more heterogeneous, showing that RAD27 expression variation strongly affects the rate of recombination from cell-to-cell. Strong variations in HR rate between subpopulations are not correlated to strong changes in cell cycle stage. Moreover, this heterogeneity occurs even when simultaneously sorting cells at equal expression level of another gene involved in DNA damage response (BMH1) that is upregulated by DNA damage, showing that the initiating DNA damage is not responsible for the observed heterogeneity in HR rate. Thus gene expression noise seems mainly responsible for this phenomenon. Finally, HR rate non-linearly scales with Rad27 levels showing that total amount of HR cannot be explained solely by the time- or population-averaged Rad27 expression. Altogether, our data reveal interplay between heterogeneity at the gene expression and genetic levels in the production of phenotypic diversity with evolutionary consequences from microbial to cancer cell populations.

Effects of mRNA Degradation and Site-Specific Transcriptional Pausing on Protein Expression Noise

Genetically identical cells exhibit diverse phenotypes even when experiencing the same environment. This phenomenon in part originates from cell-to-cell variability (noise) in protein expression. Although various kinetic schemes of stochastic transcription initiation are known to affect gene expression noise, how posttranscription initiation events contribute to noise at the protein level remains incompletely understood. To address this question, we developed a stochastic simulation-based model of bacterial gene expression that integrates well-known dependencies between transcription initiation, transcription elongation dynamics, mRNA degradation, and translation. We identified realistic conditions under which mRNA lifetime and transcriptional pauses modulate the protein expression noise initially introduced by the promoter architecture. For instance, we found that the short lifetime of bacterial mRNAs facilitates the production of protein bursts. Conversely, RNA polymerase (RNAP) pausing at specific sites during transcription elongation can attenuate protein bursts by fluidizing the RNAP traffic to the point of erasing the effect of a bursty promoter. Pause-prone sites, if located close to the promoter, can also affect noise indirectly by reducing both transcription and translation initiation due to RNAP and ribosome congestion. Our findings highlight how the interplay between transcription initiation, transcription elongation, translation, and mRNA degradation shapes the distribution in protein numbers. They also have implications for our understanding of gene evolution and suggest combinatorial strategies for modulating phenotypic variability by genetic engineering.

Respiratory Heterogeneity Shapes Biofilm Formation and Host Colonization in Uropathogenic Escherichia coli

Biofilms are multicellular bacterial communities encased in a self-secreted extracellular matrix comprised of polysaccharides, proteinaceous fibers, and DNA. Organization of these components lends spatial organization to the biofilm community such that biofilm residents can benefit from the production of common goods while being protected from exogenous insults. Spatial organization is driven by the presence of chemical gradients, such as oxygen. Here we show that two quinol oxidases found in Escherichia coli and other bacteria organize along the biofilm oxygen gradient and that this spatially coordinated expression controls architectural integrity. Cytochrome bd, a high-affinity quinol oxidase required for aerobic respiration under hypoxic conditions, is the most abundantly expressed respiratory complex in the biofilm community. Depletion of the cytochrome bd-expressing subpopulation compromises biofilm complexity by reducing the abundance of secreted extracellular matrix as well as increasing cellular sensitivity to exogenous stresses. Interrogation of the distribution of quinol oxidases in the planktonic state revealed that ∼15% of the population expresses cytochrome bd at atmospheric oxygen concentration, and this population dominates during acute urinary tract infection. These data point toward a bet-hedging mechanism in which heterogeneous expression of respiratory complexes ensures respiratory plasticity of E. coli across diverse host niches.IMPORTANCE Biofilms are multicellular bacterial communities encased in a self-secreted extracellular matrix comprised of polysaccharides, proteinaceous fibers, and DNA. Organization of these components lends spatial organization in the biofilm community. Here we demonstrate that oxygen gradients in uropathogenic Escherichia coli (UPEC) biofilms lead to spatially distinct expression programs for quinol oxidases-components of the terminal electron transport chain. Our studies reveal that the cytochrome bd-expressing subpopulation is critical for biofilm development and matrix production. In addition, we show that quinol oxidases are heterogeneously expressed in planktonic populations and that this respiratory heterogeneity provides a fitness advantage during infection. These studies define the contributions of quinol oxidases to biofilm physiology and suggest the presence of respiratory bet-hedging behavior in UPEC.