Bacterial persisters are a stochastically formed subpopulation of low-energy cells.

Persisters represent a small subpopulation of non- or slow-growing bacterial cells that are tolerant to killing by antibiotics. Despite their prominent role in the recalcitrance of chronic infections to antibiotic therapy, the mechanism of their formation has remained elusive. We show that sorted cells of Escherichia coli with low levels of energy-generating enzymes are better able to survive antibiotic killing. Using microfluidics time-lapse microscopy and a fluorescent reporter for in vivo ATP measurements, we find that a subpopulation of cells with a low level of ATP survives killing by ampicillin. We propose that these low ATP cells are formed stochastically as a result of fluctuations in the abundance of energy-generating components. These findings point to a general "low energy" mechanism of persister formation.

Synthetic Micrographs of Bacteria (SyMBac) allows accurate segmentation of bacterial cells using deep neural networks.

BACKGROUND: Deep-learning-based image segmentation models are required for accurate processing of high-throughput timelapse imaging data of bacterial cells. However, the performance of any such model strictly depends on the quality and quantity of training data, which is difficult to generate for bacterial cell images. Here, we present a novel method of bacterial image segmentation using machine learning models trained with Synthetic Micrographs of Bacteria (SyMBac). RESULTS: We have developed SyMBac, a tool that allows for rapid, automatic creation of arbitrary amounts of training data, combining detailed models of cell growth, physical interactions, and microscope optics to create synthetic images which closely resemble real micrographs, and is capable of training accurate image segmentation models. The major advantages of our approach are as follows: (1) synthetic training data can be generated virtually instantly and on demand; (2) these synthetic images are accompanied by perfect ground truth positions of cells, meaning no data curation is required; (3) different biological conditions, imaging platforms, and imaging modalities can be rapidly simulated, meaning any change in one's experimental setup no longer requires the laborious process of manually generating new training data for each change. Deep-learning models trained with SyMBac data are capable of analysing data from various imaging platforms and are robust to drastic changes in cell size and morphology. Our benchmarking results demonstrate that models trained on SyMBac data generate more accurate cell identifications and precise cell masks than those trained on human-annotated data, because the model learns the true position of the cell irrespective of imaging artefacts. We illustrate the approach by analysing the growth and size regulation of bacterial cells during entry and exit from dormancy, which revealed novel insights about the physiological dynamics of cells under various growth conditions. CONCLUSIONS: The SyMBac approach will help to adapt and improve the performance of deep-learning-based image segmentation models for accurate processing of high-throughput timelapse image data.

Time-dependent effects of transcription- and translation-halting drugs on the spatial distributions of the Escherichia coli chromosome and ribosomes.

Previously observed effects of rifampicin and chloramphenicol indicate that transcription and translation activity strongly affect the coarse spatial organization of the bacterial cytoplasm. Single-cell, time-resolved, quantitative imaging of chromosome and ribosome spatial distributions and ribosome diffusion in live Escherichia coli provides insight into the underlying mechanisms. Monte Carlo simulations of model DNA-ribosome mixtures support a novel nucleoid-ribosome mixing hypothesis. In normal conditions, 70S-polysomes and the chromosomal DNA segregate, while 30S and 50S ribosomal subunits are able to penetrate the nucleoids. Growth conditions and drug treatments determine the partitioning of ribosomes into 70S-polysomes versus free 30S and 50S subunits. Entropic and excluded volume effects then dictate the resulting chromosome and ribosome spatial distributions. Direct observation of radial contraction of the nucleoids 0-5 min after treatment with either transcription- or translation-halting drugs supports the hypothesis that simultaneous transcription, translation, and insertion of proteins into the membrane ('transertion') exerts an expanding force on the chromosomal DNA. Breaking of the DNA-RNA polymerase-mRNA-ribosome-membrane chain in either of two ways causes similar nucleoid contraction on a similar timescale. We suggest that chromosomal expansion due to transertion enables co-transcriptional translation throughout the nucleoids.

Nonperturbative imaging of nucleoid morphology in live bacterial cells during an antimicrobial peptide attack.

Studies of time-dependent drug and environmental effects on single, live bacterial cells would benefit significantly from a permeable, nonperturbative, long-lived fluorescent stain specific to the nucleoids (chromosomal DNA). The ideal stain would not affect cell growth rate or nucleoid morphology and dynamics, even during laser illumination for hundreds of camera frames. In this study, time-dependent, single-cell fluorescence imaging with laser excitation and a sensitive electron-multiplying charge-coupled-device (EMCCD) camera critically tested the utility of "dead-cell stains" (SYTOX orange and SYTOX green) and "live-cell stains" (DRAQ5 and SYTO 61) and also 4',6-diamidino-2-phenylindole (DAPI). Surprisingly, the dead-cell stains were nearly ideal for imaging live Escherichia coli, while the live-cell stains and DAPI caused nucleoid expansion and, in some cases, cell permeabilization and the halting of growth. SYTOX orange performed well for both the Gram-negative E. coli and the Gram-positive Bacillus subtilis. In an initial application, we used two-color fluorescence imaging to show that the antimicrobial peptide cecropin A destroyed nucleoid-ribosome segregation over 20 min after permeabilization of the E. coli cytoplasmic membrane, reminiscent of the long-term effects of the drug rifampin. In contrast, the human cathelicidin LL-37, while similar to cecropin A in structure, length, charge, and the ability to permeabilize bacterial membranes, had no observable effect on nucleoid-ribosome segregation. Possible underlying causes are suggested.

Structural and functional analyses of nematode-derived antimicrobial peptides support the occurrence of direct mechanisms of worm-microbiota interactions.

The complex relationships between gastrointestinal (GI) nematodes and the host gut microbiota have been implicated in key aspects of helminth disease and infection outcomes. Nevertheless, the direct and indirect mechanisms governing these interactions are, thus far, largely unknown. In this proof-of-concept study, we demonstrate that the excretory-secretory products (ESPs) and extracellular vesicles (EVs) of key GI nematodes contain peptides that, when recombinantly expressed, exert antimicrobial activity in vitro against Bacillus subtilis. In particular, using time-lapse microfluidics microscopy, we demonstrate that exposure of B. subtilis to a recombinant saposin-domain containing peptide from the 'brown stomach worm', Teladorsagia circumcincta, and a metridin-like ShK toxin from the 'barber's pole worm', Haemonchus contortus, results in cell lysis and significantly reduced growth rates. Data from this study support the hypothesis that GI nematodes may modulate the composition of the vertebrate gut microbiota directly via the secretion of antimicrobial peptides, and pave the way for future investigations aimed at deciphering the impact of such changes on the pathophysiology of GI helminth infection and disease.

Partitioning of RNA polymerase activity in live Escherichia coli from analysis of single-molecule diffusive trajectories.

Superresolution fluorescence microscopy is used to locate single copies of RNA polymerase (RNAP) in live Escherichia coli and track their diffusive motion. On a timescale of 0.1-1 s, most copies separate remarkably cleanly into two diffusive states. The "slow" RNAPs, which move indistinguishably from DNA loci, are assigned to specifically bound copies (with fractional population ftrxn) that are initiating transcription, elongating, pausing, or awaiting termination. The "mixed-state" RNAP copies, with effective diffusion constant Dmixed = 0.21 µm(2) s(-1), are assigned as a rapidly exchanging mixture of nonspecifically bound copies (fns) and copies undergoing free, three-dimensional diffusion within the nucleoids (ffree). Longer trajectories of 7-s duration reveal transitions between the slow and mixed states, corroborating the assignments. Short trajectories of 20-ms duration enable direct observation of the freely diffusing RNAP copies, yielding Dfree = 0.7 µm(2) s(-1). Analysis of single-particle trajectories provides quantitative estimates of the partitioning of RNAP into different states of activity: ftrxn = 0.54 ± 0.07, fns = 0.28 ± 0.05, ffree = 0.12 ± 0.03, and fnb = 0.06 ± 0.05 (fraction unable to bind to DNA on a 1-s timescale). These fractions disagree with earlier estimates.

Localized permeabilization of E. coli membranes by the antimicrobial peptide Cecropin A.

Fluorescence microscopy enables detailed observation of the effects of the antimicrobial peptide Cecropin A on the outer membrane (OM) and cytoplasmic membrane (CM) of single E. coli cells with subsecond time resolution. Fluorescence from periplasmic GFP decays and cell growth halts when the OM is permeabilized. Fluorescence from the DNA stain Sytox Green rises when the CM is permeabilized and the stain enters the cytoplasm. The initial membrane disruptions are localized and stable. Septating cells are attacked earlier than nonseptating cells, and curved membrane surfaces are attacked in preference to cylindrical surfaces. Below a threshold bulk Cecropin A concentration, permeabilization is not observed over 30 min. Above this threshold, we observe a lag time of several minutes between Cecropin A addition and OM permeabilization and ∼30 s between OM and CM permeabilization. The long lag times and the existence of a threshold concentration for permeabilization suggest a nucleation mechanism. However, the roughly linear dependence of mean lag time on bulk peptide concentration is not easily reconciled with a nucleation step involving simultaneous insertion of multiple peptides into the bilayer. Monte Carlo simulations suggest that within seconds, the OM permeability becomes comparable to that of a pore of 100 nm diameter or of numerous small pores distributed over a similarly large area.

Divin: a small molecule inhibitor of bacterial divisome assembly.

Bacterial cell division involves the dynamic assembly of division proteins and coordinated constriction of the cell envelope. A wide range of factors regulates cell division--including growth and environmental stresses--and the targeting of the division machinery has been a widely discussed approach for antimicrobial therapies. This paper introduces divin, a small molecule inhibitor of bacterial cell division that may facilitate mechanistic studies of this process. Divin disrupts the assembly of late division proteins, reduces peptidoglycan remodeling at the division site, and blocks compartmentalization of the cytoplasm. In contrast to other division inhibitors, divin does not interact with the tubulin homologue FtsZ, affect chromosome segregation, or activate regulatory mechanisms that inhibit cell division indirectly. Our studies of bacterial cell division using divin as a probe suggest that dividing bacteria proceed through several morphological stages of the cell envelope, and FtsZ is required but not sufficient to compartmentalize the cytoplasmic membrane at the division site. Divin is only moderately toxic to mammalian cells at concentrations that inhibit the growth of clinical pathogens. These characteristics make divin a useful probe for studying bacterial cell division and a starting point for the development of new classes of therapeutic agents.

Superresolution imaging of ribosomes and RNA polymerase in live Escherichia coli cells.

Quantitative spatial distributions of ribosomes (S2-YFP) and RNA polymerase (RNAP; ß'-yGFP) in live Escherichia coli are measured by superresolution fluorescence microscopy. In moderate growth conditions, nucleoid-ribosome segregation is strong, and RNAP localizes to the nucleoid lobes. The mean copy numbers per cell are 4600 RNAPs and 55,000 ribosomes. Only 10-15% of the ribosomes lie within the densest part of the nucleoid lobes, and at most 4% of the RNAPs lie in the two ribosome-rich endcaps. The predominant observed diffusion coefficient of ribosomes is D(ribo) = 0.04 µm(2) s(-1), attributed to free mRNA being translated by one or more 70S ribosomes. We find no clear evidence of subdiffusion, as would arise from tethering of ribosomes to the DNA. The degree of DNA-ribosome segregation strongly suggests that in E. coli most translation occurs on free mRNA transcripts that have diffused into the ribosome-rich regions. Both RNAP and ribosome radial distributions extend to the cytoplasmic membrane, consistent with the transertion hypothesis. However, few if any RNAP copies lie near the membrane of the endcaps. This suggests that if transertion occurs, it exerts a direct radially expanding force on the nucleoid, but not a direct axially expanding force.

The context matrix: Navigating biological complexity for advanced biodesign.

Synthetic biology offers many solutions in healthcare, production, sensing and agriculture. However, the ability to rationally engineer synthetic biosystems with predictable and robust functionality remains a challenge. A major reason is the complex interplay between the synthetic genetic construct, its host, and the environment. Each of these contexts contains a number of input factors which together can create unpredictable behaviours in the engineered biosystem. It has become apparent that for the accurate assessment of these contextual effects a more holistic approach to design and characterisation is required. In this perspective article, we present the context matrix, a conceptual framework to categorise and explore these contexts and their net effect on the designed synthetic biosystem. We propose the use and community-development of the context matrix as an aid for experimental design that simplifies navigation through the complex design space in synthetic biology.

Subdiffraction-limit study of Kaede diffusion and spatial distribution in live Escherichia coli.

Photoactivation localization microscopy (PALM) is used to study the spatial distribution and diffusion of single copies of the protein Kaede in the cytoplasm of live Escherichia coli under moderate growth conditions (67 min doubling time). The spatial distribution of Kaede is uniform within the cytoplasm. The cytoplasmic radius of 380 ± 30 nm varies little from cell to cell. Single-particle tracking using 4 ms exposure times reveals negatively curved plots of mean-square displacement versus time. A detailed comparison with Monte Carlo simulations in a spherocylindrical volume shows that the curvature can be quantitatively understood in terms of free diffusion within a confining volume. The mean diffusion coefficient across cells is = 7.3 ± 1.1 µm(2)·s(-1), consistent with a homotetrameric form of Kaede. The distribution of squared displacements along the long axis for individual Kaede molecules is consistent with homogeneous diffusion. However, for longer cells, a spatial map of one-step estimates of the diffusion coefficient along x suggests that diffusion is ∼20-40% faster within nucleoids than in the ribosome-rich region lying between nucleoid lobes at the cell mid-plane. Fluorescence recovery after photobleaching yielded = 8.3 ± 1.6 µm(2)·s(-1), in agreement with the single-particle tracking results.

Challenges of analysing stochastic gene expression in bacteria using single-cell time-lapse experiments.

Stochastic gene expression causes phenotypic heterogeneity in a population of genetically identical bacterial cells. Such non-genetic heterogeneity can have important consequences for the population fitness, and therefore cells implement regulation strategies to either suppress or exploit such heterogeneity to adapt to their circumstances. By employing time-lapse microscopy of single cells, the fluctuation dynamics of gene expression may be analysed, and their regulatory mechanisms thus deciphered. However, a careful consideration of the experimental design and data-analysis is needed to produce useful data for deriving meaningful insights from them. In the present paper, the individual steps and challenges involved in a time-lapse experiment are discussed, and a rigorous framework for designing, performing, and extracting single-cell gene expression dynamics data from such experiments is outlined.

Tracking bacterial lineages in complex and dynamic environments with applications for growth control and persistence.

As bacteria transition from exponential to stationary phase, they change substantially in size, morphology, growth and expression profiles. These responses also vary between individual cells, but it has proved difficult to track cell lineages along the growth curve to determine the progression of events or correlations between how individual cells enter and exit dormancy. Here, we developed a platform for tracking more than 105 parallel cell lineages in dense and changing cultures, independently validating that the imaged cells closely track batch populations. Initial applications show that for both Escherichia coli and Bacillus subtilis, growth changes from an 'adder' mode in exponential phase to mixed 'adder-timers' entering stationary phase, and then a near-perfect 'sizer' upon exit-creating broadly distributed cell sizes in stationary phase but rapidly returning to narrowly distributed sizes upon exit. Furthermore, cells that undergo more divisions when entering stationary phase suffer reduced survival after long periods of dormancy but are the only cells observed that persist following antibiotic treatment.

Stochastic antagonism between two proteins governs a bacterial cell fate switch.

Cell fate decision circuits must be variable enough for genetically identical cells to adopt a multitude of fates, yet ensure that these states are distinct, stably maintained, and coordinated with neighboring cells. A long-standing view is that this is achieved by regulatory networks involving self-stabilizing feedback loops that convert small differences into long-lived cell types. We combined regulatory mutants and in vivo reconstitution with theory for stochastic processes to show that the marquee features of a cell fate switch in Bacillus subtilis-discrete states, multigenerational inheritance, and timing of commitments-can instead be explained by simple stochastic competition between two constitutively produced proteins that form an inactive complex. Such antagonistic interactions are commonplace in cells and could provide powerful mechanisms for cell fate determination more broadly.

Quantification of very low-abundant proteins in bacteria using the HaloTag and epi-fluorescence microscopy.

Cell biology is increasingly dependent on quantitative methods resulting in the need for microscopic labelling technologies that are highly sensitive and specific. Whilst the use of fluorescent proteins has led to major advances, they also suffer from their relatively low brightness and photo-stability, making the detection of very low abundance proteins using fluorescent protein-based methods challenging. Here, we characterize the use of the self-labelling protein tag called HaloTag, in conjunction with an organic fluorescent dye, to label and accurately count endogenous proteins present in very low numbers (<7) in individual Escherichia coli cells. This procedure can be used to detect single molecules in fixed cells with conventional epifluorescence illumination and a standard microscope. We show that the detection efficiency of proteins labelled with the HaloTag is ≥80%, which is on par or better than previous techniques. Therefore, this method offers a simple and attractive alternative to current procedures to detect low abundance molecules.

Bacterial variability in the mammalian gut captured by a single-cell synthetic oscillator.

Synthetic gene oscillators have the potential to control timed functions and periodic gene expression in engineered cells. Such oscillators have been refined in bacteria in vitro, however, these systems have lacked the robustness and precision necessary for applications in complex in vivo environments, such as the mammalian gut. Here, we demonstrate the implementation of a synthetic oscillator capable of keeping robust time in the mouse gut over periods of days. The oscillations provide a marker of bacterial growth at a single-cell level enabling quantification of bacterial dynamics in response to inflammation and underlying variations in the gut microbiota. Our work directly detects increased bacterial growth heterogeneity during disease and differences between spatial niches in the gut, demonstrating the deployment of a precise engineered genetic oscillator in real-life settings.

Single-cell microscopy of suspension cultures using a microfluidics-assisted cell screening platform.

Studies that rely on fluorescence imaging of nonadherent cells that are cultured in suspension, such as Escherichia coli, are often hampered by trade-offs that must be made between data throughput and imaging resolution. We developed a platform for microfluidics-assisted cell screening (MACS) that overcomes this trade-off by temporarily immobilizing suspension cells within a microfluidics chip. This enables high-throughput and automated single-cell microscopy for a wide range of cell types and sizes. As cells can be rapidly sampled directly from a suspension culture, MACS bypasses the need for sample preparation, and therefore allows measurements without perturbing the native cell physiology. The setup can also be integrated with complex growth chambers, and can be used to enrich or sort the imaged cells. Furthermore, MACS facilitates the visualization of individual cytoplasmic fluorescent proteins (FPs) in E. coli, allowing low-abundance proteins to be counted using standard total internal reflection fluorescence (TIRF) microscopy. Finally, MACS can be used to impart mechanical pressure for assessing the structural integrity of individual cells and their response to mechanical perturbations, or to make cells take up chemicals that otherwise would not pass through the membrane. This protocol describes the assembly of electronic control circuitry, the construction of liquid-handling components and the creation of the MACS microfluidics chip. The operation of MACS is described, and automation software is provided to integrate MACS control with image acquisition. Finally, we provide instructions for extending MACS using an external growth chamber (1 d) and for how to sort rare cells of interest.

Mechanical slowing-down of cytoplasmic diffusion allows in vivo counting of proteins in individual cells.

Many key regulatory proteins in bacteria are present in too low numbers to be detected with conventional methods, which poses a particular challenge for single-cell analyses because such proteins can contribute greatly to phenotypic heterogeneity. Here we develop a microfluidics-based platform that enables single-molecule counting of low-abundance proteins by mechanically slowing-down their diffusion within the cytoplasm of live Escherichia coli (E. coli) cells. Our technique also allows for automated microscopy at high throughput with minimal perturbation to native physiology, as well as viable enrichment/retrieval. We illustrate the method by analysing the control of the master regulator of the E. coli stress response, RpoS, by its adapter protein, SprE (RssB). Quantification of SprE numbers shows that though SprE is necessary for RpoS degradation, it is expressed at levels as low as 3-4 molecules per average cell cycle, and fluctuations in SprE are approximately Poisson distributed during exponential phase with no sign of bursting.