A noisy linear map underlies oscillations in cell size and gene expression in bacteria.

During bacterial growth, a cell approximately doubles in size before division, after which it splits into two daughter cells. This process is subjected to the inherent perturbations of cellular noise and thus requires regulation for cell-size homeostasis. The mechanisms underlying the control and dynamics of cell size remain poorly understood owing to the difficulty in sizing individual bacteria over long periods of time in a high-throughput manner. Here we measure and analyse long-term, single-cell growth and division across different Escherichia coli strains and growth conditions. We show that a subset of cells in a population exhibit transient oscillations in cell size with periods that stretch across several (more than ten) generations. Our analysis reveals that a simple law governing cell-size control-a noisy linear map-explains the origins of these cell-size oscillations across all strains. This noisy linear map implements a negative feedback on cell-size control: a cell with a larger initial size tends to divide earlier, whereas one with a smaller initial size tends to divide later. Combining simulations of cell growth and division with experimental data, we demonstrate that this noisy linear map generates transient oscillations, not just in cell size, but also in constitutive gene expression. Our work provides new insights into the dynamics of bacterial cell-size regulation with implications for the physiological processes involved.

Deep Learning Automates the Quantitative Analysis of Individual Cells in Live-Cell Imaging Experiments.

Live-cell imaging has opened an exciting window into the role cellular heterogeneity plays in dynamic, living systems. A major critical challenge for this class of experiments is the problem of image segmentation, or determining which parts of a microscope image correspond to which individual cells. Current approaches require many hours of manual curation and depend on approaches that are difficult to share between labs. They are also unable to robustly segment the cytoplasms of mammalian cells. Here, we show that deep convolutional neural networks, a supervised machine learning method, can solve this challenge for multiple cell types across the domains of life. We demonstrate that this approach can robustly segment fluorescent images of cell nuclei as well as phase images of the cytoplasms of individual bacterial and mammalian cells from phase contrast images without the need for a fluorescent cytoplasmic marker. These networks also enable the simultaneous segmentation and identification of different mammalian cell types grown in co-culture. A quantitative comparison with prior methods demonstrates that convolutional neural networks have improved accuracy and lead to a significant reduction in curation time. We relay our experience in designing and optimizing deep convolutional neural networks for this task and outline several design rules that we found led to robust performance. We conclude that deep convolutional neural networks are an accurate method that require less curation time, are generalizable to a multiplicity of cell types, from bacteria to mammalian cells, and expand live-cell imaging capabilities to include multi-cell type systems.

Optimality and robustness in quorum sensing (QS)-mediated regulation of a costly public good enzyme.

Bacteria secrete a variety of public good exoproducts into their environment. These exoproducts are typically produced under the control of quorum sensing (QS), a signaling mechanism by which bacteria sense and respond to changes in their density. QS seems to provide an advantageous strategy to regulate these costly but beneficial exoproducts: it delays production until sufficiently high cell density, when the overall benefit of exoproducts outweighs cost of their production. This notion raises several fundamental questions about QS as a general control strategy adopted by bacteria. How much delay is advantageous? Under what conditions does QS-mediated regulation become advantageous? How does this advantage depend on the kinetic properties of QS? How robust is a given QS system to the stochastic events that occur over bacterial lifecycles? To quantitatively address these questions, we engineered a gene circuit in Escherichia coli to control the synthesis and secretion of a costly but beneficial exoenzyme. We show that exoenzyme production is overall advantageous only if initiated at a sufficiently high density. This property sets the potential advantage for QS-mediated regulation when the initial density is low and the growth cycle is sufficiently long compared with the exoenzyme response time. This advantage of QS-mediated regulation is robust to varying initial cell densities and growth durations, and it is particularly striking when bacteria face uncertainty, such as from stochastic dispersal during their lifecycle. We show, however, that, for QS to be optimal, its kinetic properties must be appropriately tuned; this property has implications for antibacterial treatments that target QS.

Programming stress-induced altruistic death in engineered bacteria.

Programmed death is often associated with a bacterial stress response. This behavior appears paradoxical, as it offers no benefit to the individual. This paradox can be explained if the death is 'altruistic': the killing of some cells can benefit the survivors through release of 'public goods'. However, the conditions where bacterial programmed death becomes advantageous have not been unambiguously demonstrated experimentally. Here, we determined such conditions by engineering tunable, stress-induced altruistic death in the bacterium Escherichia coli. Using a mathematical model, we predicted the existence of an optimal programmed death rate that maximizes population growth under stress. We further predicted that altruistic death could generate the 'Eagle effect', a counter-intuitive phenomenon where bacteria appear to grow better when treated with higher antibiotic concentrations. In support of these modeling insights, we experimentally demonstrated both the optimality in programmed death rate and the Eagle effect using our engineered system. Our findings fill a critical conceptual gap in the analysis of the evolution of bacterial programmed death, and have implications for a design of antibiotic treatment.

Computation of steady-state probability distributions in stochastic models of cellular networks.

Cellular processes are "noisy". In each cell, concentrations of molecules are subject to random fluctuations due to the small numbers of these molecules and to environmental perturbations. While noise varies with time, it is often measured at steady state, for example by flow cytometry. When interrogating aspects of a cellular network by such steady-state measurements of network components, a key need is to develop efficient methods to simulate and compute these distributions. We describe innovations in stochastic modeling coupled with approaches to this computational challenge: first, an approach to modeling intrinsic noise via solution of the chemical master equation, and second, a convolution technique to account for contributions of extrinsic noise. We show how these techniques can be combined in a streamlined procedure for evaluation of different sources of variability in a biochemical network. Evaluation and illustrations are given in analysis of two well-characterized synthetic gene circuits, as well as a signaling network underlying the mammalian cell cycle entry.

Engineered Fluorescent E. coli Lysogens Allow Live-Cell Imaging of Functional Prophage Induction Triggered inside Macrophages.

Half of the bacteria in the human gut microbiome are lysogens containing integrated prophages, which may activate in stressful immune environments. Although lysogens are likely to be phagocytosed by macrophages, whether prophage activation occurs or influences the outcome of bacterial infection remains unexplored. To study the dynamics of bacteria-phage interactions in living cells-in particular, the macrophage-triggered induction and lysis of dormant prophages in the phagosome-we adopted a tripartite system where murine macrophages engulf E. coli, which are lysogenic with an engineered bacteriophage λ, containing a fluorescent lysis reporter. Pre-induced prophages are capable of lysing the host bacterium and propagating infection to neighboring bacteria in the same phagosome. A non-canonical pathway, mediated by PhoP, is involved with the native λ phage induction inside phagocytosed E. coli. These findings suggest two possible mechanisms by which induced prophages may function to aid the bactericidal activity of macrophages.

Noise reduction by diffusional dissipation in a minimal quorum sensing motif.

Cellular interactions are subject to random fluctuations (noise) in quantities of interacting molecules. Noise presents a major challenge for the robust function of natural and engineered cellular networks. Past studies have analyzed how noise is regulated at the intracellular level. Cell-cell communication, however, may provide a complementary strategy to achieve robust gene expression by enabling the coupling of a cell with its environment and other cells. To gain insight into this issue, we have examined noise regulation by quorum sensing (QS), a mechanism by which many bacteria communicate through production and sensing of small diffusible signals. Using a stochastic model, we analyze a minimal QS motif in Gram-negative bacteria. Our analysis shows that diffusion of the QS signal, together with fast turnover of its transcriptional regulator, attenuates low-frequency components of extrinsic noise. We term this unique mechanism "diffusional dissipation" to emphasize the importance of fast signal turnover (or dissipation) by diffusion. We further show that this noise attenuation is a property of a more generic regulatory motif, of which QS is an implementation. Our results suggest that, in a QS system, an unstable transcriptional regulator may be favored for regulating expression of costly proteins that generate public goods.

Combining Comprehensive Analysis of Off-Site Lambda Phage Integration with a CRISPR-Based Means of Characterizing Downstream Physiology.

During its lysogenic life cycle, the phage genome is integrated into the host chromosome by site-specific recombination. In this report, we analyze lambda phage integration into noncanonical sites using next-generation sequencing and show that it generates significant genetic diversity by targeting over 300 unique sites in the host Escherichia coli genome. Moreover, these integration events can have important phenotypic consequences for the host, including changes in cell motility and increased antibiotic resistance. Importantly, the new technologies that we developed to enable this study-sequencing secondary sites using next-generation sequencing and then selecting relevant lysogens using clustered regularly interspaced short palindromic repeat (CRISPR)/Cas9-based selection-are broadly applicable to other phage-bacterium systems.IMPORTANCE Bacteriophages play an important role in bacterial evolution through lysogeny, where the phage genome is integrated into the host chromosome. While phage integration generally occurs at a specific site in the host chromosome, it is also known to occur at other, so-called secondary sites. In this study, we developed a new experimental technology to comprehensively study secondary integration sites and discovered that phage can integrate into over 300 unique sites in the host genome, resulting in significant genetic diversity in bacteria. We further developed an assay to examine the phenotypic consequence of such diverse integration events and found that phage integration can cause changes in evolutionarily relevant traits such as bacterial motility and increases in antibiotic resistance. Importantly, our method is readily applicable to other phage-bacterium systems.

Long-term growth data of Escherichia coli at a single-cell level.

Long-term, single-cell measurement of bacterial growth is extremely valuable information, particularly in the study of homeostatic aspects such as cell-size and growth rate control. Such measurement has recently become possible due to the development of microfluidic technology. Here we present data from single-cell measurements of Escherichia coli growth over 70 generations obtained for three different growth conditions. The data were recorded every minute, and contain time course data of cell length and fluorescent intensity of constitutively expressed yellow fluorescent protein.

Generic metric to quantify quorum sensing activation dynamics.

Quorum sensing (QS) enables bacteria to sense and respond to changes in their population density. It plays a critical role in controlling different biological functions, including bioluminescence and bacterial virulence. It has also been widely adapted to program robust dynamics in one or multiple cellular populations. While QS systems across bacteria all appear to function similarly-as density-dependent control systems-there is tremendous diversity among these systems in terms of signaling components and network architectures. This diversity hampers efforts to quantify the general control properties of QS. For a specific QS module, it remains unclear how to most effectively characterize its regulatory properties in a manner that allows quantitative predictions of the activation dynamics of the target gene. Using simple kinetic models, here we show that the dominant temporal dynamics of QS-controlled target activation can be captured by a generic metric, 'sensing potential', defined at a single time point. We validate these predictions using synthetic QS circuits in Escherichia coli. Our work provides a computational framework and experimental methodology to characterize diverse natural QS systems and provides a concise yet quantitative criterion for selecting or optimizing a QS system for synthetic biology applications.

Linear population allocation by bistable switches in response to transient stimulation.

Many cellular decision processes, including proliferation, differentiation, and phenotypic switching, are controlled by bistable signaling networks. In response to transient or intermediate input signals, these networks allocate a population fraction to each of two distinct states (e.g. OFF and ON). While extensive studies have been carried out to analyze various bistable networks, they are primarily focused on responses of bistable networks to sustained input signals. In this work, we investigate the response characteristics of bistable networks to transient signals, using both theoretical analysis and numerical simulation. We find that bistable systems exhibit a common property: for input signals with short durations, the fraction of switching cells increases linearly with the signal duration, allowing the population to integrate transient signals to tune its response. We propose that this allocation algorithm can be an optimal response strategy for certain cellular decisions in which excessive switching results in lower population fitness.

Programmed cell death in bacteria and implications for antibiotic therapy.

It is now well appreciated that programmed cell death (PCD) plays critical roles in the life cycle of diverse bacterial species. It is an apparently paradoxical behavior as it does not benefit the cells undergoing PCD. However, growing evidence suggests that PCD can be 'altruistic': the dead cells may directly or indirectly benefit survivors through generation of public goods. This property provides a potential explanation on how PCD can evolve as an extreme form of cooperation, although many questions remain to be addressed. From another perspective, as PCD plays a critical role in bacterial pathogenesis, it has been proposed as a potential target for new antibacterial therapy. To this end, understanding the population and evolutionary dynamics resulting from PCD and public goods production may be a key to the success of designing effective antibiotic treatment.

Engineering microbial systems to explore ecological and evolutionary dynamics.

A major goal of biological research is to provide a mechanistic understanding of diverse biological processes. To this end, synthetic biology offers a powerful approach, whereby biological questions can be addressed in a well-defined framework. By constructing simple gene circuits, such studies have generated new insights into the design principles of gene regulatory networks. Recently, this strategy has been applied to analyze ecological and evolutionary questions, where population-level interactions are critical. Here, we highlight recent development of such systems and discuss how they were used to address problems in ecology and evolutionary biology. As illustrated by these examples, synthetic ecosystems provide a unique platform to study ecological and evolutionary phenomena that are challenging to study in their natural contexts.

Oscillations by minimal bacterial suicide circuits reveal hidden facets of host-circuit physiology.

Synthetic biology seeks to enable programmed control of cellular behavior though engineered biological systems. These systems typically consist of synthetic circuits that function inside, and interact with, complex host cells possessing pre-existing metabolic and regulatory networks. Nevertheless, while designing systems, a simple well-defined interface between the synthetic gene circuit and the host is frequently assumed. We describe the generation of robust but unexpected oscillations in the densities of bacterium Escherichia coli populations by simple synthetic suicide circuits containing quorum components and a lysis gene. Contrary to design expectations, oscillations required neither the quorum sensing genes (luxR and luxI) nor known regulatory elements in the P(luxI) promoter. Instead, oscillations were likely due to density-dependent plasmid amplification that established a population-level negative feedback. A mathematical model based on this mechanism captures the key characteristics of oscillations, and model predictions regarding perturbations to plasmid amplification were experimentally validated. Our results underscore the importance of plasmid copy number and potential impact of "hidden interactions" on the behavior of engineered gene circuits - a major challenge for standardizing biological parts. As synthetic biology grows as a discipline, increasing value may be derived from tools that enable the assessment of parts in their final context.

Decoding biological principles using gene circuits.

A major flavor of synthetic biology is the creation of artificial gene circuits to perform user-defined tasks. One aspect of this area is to realize ever-increasingly more complicated circuit behavior. Such efforts have led to the identification and evaluation of design strategies that enable robust control of dynamics in single cells and in cell populations. On the other hand, there is increasing emphasis on using artificial systems programmed by simple circuits to explore fundamental biological questions of broad significance.

Engineering multicellular systems by cell-cell communication.

Synthetic biology encompasses the design of new biological parts and systems as well as the modulation of existing biological networks to generate novel functions. In recent years, increasing emphasis has been placed on the engineering of population-level behaviors using cell-cell communication. From the engineering perspective, cell-cell communication serves as a versatile regulatory module that enables coordination among cells in and between populations and facilitates the generation of reliable dynamics. In addition to exploring biological 'design principles' via the construction of increasingly complex dynamics, communication-based synthetic systems can be used as well-defined model systems to study ecological and social interactions such as competition, cooperation, and predation. Here we discuss the dynamic properties of cell-cell communication modules, how they can be engineered for synthetic circuit design, and applications of these systems.