Cell-free biosynthesis combined with deep learning accelerates de novo-development of antimicrobial peptides.

Bioactive peptides are key molecules in health and medicine. Deep learning holds a big promise for the discovery and design of bioactive peptides. Yet, suitable experimental approaches are required to validate candidates in high throughput and at low cost. Here, we established a cell-free protein synthesis (CFPS) pipeline for the rapid and inexpensive production of antimicrobial peptides (AMPs) directly from DNA templates. To validate our platform, we used deep learning to design thousands of AMPs de novo. Using computational methods, we prioritized 500 candidates that we produced and screened with our CFPS pipeline. We identified 30 functional AMPs, which we characterized further through molecular dynamics simulations, antimicrobial activity and toxicity. Notably, six de novo-AMPs feature broad-spectrum activity against multidrug-resistant pathogens and do not develop bacterial resistance. Our work demonstrates the potential of CFPS for high throughput and low-cost production and testing of bioactive peptides within less than 24 h.

Nanoscale organization of the MHC I peptide-loading complex in human dendritic cells.

Dendritic cells (DCs) translate local innate immune responses into long-lasting adaptive immunity by priming antigen-specific T cells. Accordingly, there is an ample interest in exploiting DCs for therapeutic purposes, e.g., in personalized immunotherapies. Despite recent advances in elucidating molecular pathways of antigen processing, in DCs the exact spatial organization of the underlying processes is largely unknown. Here, we unraveled the nanoscale organization of the transporter associated with antigen processing (TAP)-dependent peptide-loading machinery in human monocyte-derived DCs (moDC). We detected an unexpected accumulation of MHC I peptide-loading complexes (PLCs) and TAP-dependent peptide compartmentalization in protrusions of activated DCs. Using single-molecule localization microscopy we revealed that PLCs display homogeneously sized assemblies, independent of the DC activation status or cellular localization. Our data indicate that moDCs show augmentation of subcellular PLC density during DC maturation. We observed a twofold density increase in the cell body, while an even fourfold accumulation was detected in the tips of the protrusions at the mature DC stage in comparison to immature DCs. In these tip regions, PLC assemblies are found along highly compressed tubular ER networks. These findings provide novel insights into nanoscale organization of the antigen presentation machinery, and open new perspectives on the T cell stimulatory capacity of DCs.

DeepBacs for multi-task bacterial image analysis using open-source deep learning approaches.

This work demonstrates and guides how to use a range of state-of-the-art artificial neural-networks to analyse bacterial microscopy images using the recently developed ZeroCostDL4Mic platform. We generated a database of image datasets used to train networks for various image analysis tasks and present strategies for data acquisition and curation, as well as model training. We showcase different deep learning (DL) approaches for segmenting bright field and fluorescence images of different bacterial species, use object detection to classify different growth stages in time-lapse imaging data, and carry out DL-assisted phenotypic profiling of antibiotic-treated cells. To also demonstrate the ability of DL to enhance low-phototoxicity live-cell microscopy, we showcase how image denoising can allow researchers to attain high-fidelity data in faster and longer imaging. Finally, artificial labelling of cell membranes and predictions of super-resolution images allow for accurate mapping of cell shape and intracellular targets. Our purposefully-built database of training and testing data aids in novice users' training, enabling them to quickly explore how to analyse their data through DL. We hope this lays a fertile ground for the efficient application of DL in microbiology and fosters the creation of tools for bacterial cell biology and antibiotic research.

Democratising deep learning for microscopy with ZeroCostDL4Mic.

Deep Learning (DL) methods are powerful analytical tools for microscopy and can outperform conventional image processing pipelines. Despite the enthusiasm and innovations fuelled by DL technology, the need to access powerful and compatible resources to train DL networks leads to an accessibility barrier that novice users often find difficult to overcome. Here, we present ZeroCostDL4Mic, an entry-level platform simplifying DL access by leveraging the free, cloud-based computational resources of Google Colab. ZeroCostDL4Mic allows researchers with no coding expertise to train and apply key DL networks to perform tasks including segmentation (using U-Net and StarDist), object detection (using YOLOv2), denoising (using CARE and Noise2Void), super-resolution microscopy (using Deep-STORM), and image-to-image translation (using Label-free prediction - fnet, pix2pix and CycleGAN). Importantly, we provide suitable quantitative tools for each network to evaluate model performance, allowing model optimisation. We demonstrate the application of the platform to study multiple biological processes.

Microbial Cationic Peptides as a Natural Defense Mechanism against Insect Antimicrobial Peptides.

Bacteria produce a plethora of specialized metabolites (SM), with the ecological function of most of them not known. A major group of SM are peptides derived from nonribosomal peptide synthetases (NRPS). In entomopathogenic bacteria of the genus Xenorhabdus, PAX (peptide-antimicrobial-Xenorhabdus) were described as NRPS-derived lipopeptides, which show antimicrobial activities against bacteria and fungi. We analyzed the production of PAX in Xenorhabdus doucetiae and found the majority bound to the cells. We derivatized PAX with fluorophores and show binding to cells when added externally using super-resolution microscopy. Externally added PAX in X. doucetiae and E. coli as well as inducible PAX production in X. doucetiae showed a protective effect against various antimicrobial peptides (AMPs) from insects, where they are used as a defense mechanism against pathogens. Because AMPs are often positively charged, our results suggest a PAX-induced repulsive force due to positive charge at the bacterial cell wall.

Multi-Color, Bleaching-Resistant Super-Resolution Optical Fluctuation Imaging with Oligonucleotide-Based Exchangeable Fluorophores.

Super-resolution optical fluctuation imaging (SOFI) is a super-resolution microscopy technique that overcomes the diffraction limit by analyzing intensity fluctuations of statistically independent emitters in a time series of images. The final images are background-free and show confocality and enhanced spatial resolution (super-resolution). Fluorophore photobleaching, however, is a key limitation for recording long time series of images that will allow for the calculation of higher order SOFI results with correspondingly increased resolution. Here, we demonstrate that photobleaching can be circumvented by using fluorophore labels that reversibly and transiently bind to a target, and which are being replenished from a buffer which serves as a reservoir. Using fluorophore-labeled short DNA oligonucleotides, we labeled cellular structures with target-specific antibodies that contain complementary DNA sequences and record the fluctuation events caused by transient emitter binding. We show that this concept bypasses extensive photobleaching and facilitates two-color imaging of cellular structures with SOFI.

Protein-Specific, Multicolor and 3D STED Imaging in Cells with DNA-Labeled Antibodies.

Photobleaching is a major challenge in fluorescence microscopy, in particular if high excitation light intensities are used. Signal-to-noise and spatial resolution may be compromised, which limits the amount of information that can be extracted from an image. Photobleaching can be bypassed by using exchangeable labels, which transiently bind to and dissociate from a target, thereby replenishing the destroyed labels with intact ones from a reservoir. Here, we demonstrate confocal and STED microscopy with short, fluorophore-labeled oligonucleotides that transiently bind to complementary oligonucleotides attached to protein-specific antibodies. The constant exchange of fluorophore labels in DNA-based STED imaging bypasses photobleaching that occurs with covalent labels. We show that this concept is suitable for targeted, two-color STED imaging of whole cells.

Whole-Cell, 3D, and Multicolor STED Imaging with Exchangeable Fluorophores.

We demonstrate stimulated emission depletion (STED) microscopy of whole bacterial and eukaryotic cells using fluorogenic labels that reversibly bind to their target structure. A constant exchange of labels guarantees the removal of photobleached fluorophores and their replacement by intact fluorophores, thereby circumventing bleaching-related limitations of STED super-resolution imaging. We achieve a constant labeling density and demonstrate a fluorescence signal for long and theoretically unlimited acquisition times. Using this concept, we demonstrate whole-cell, 3D, multicolor, and live-cell STED microscopy.

A toolbox for multiplexed super-resolution imaging of the E. coli nucleoid and membrane using novel PAINT labels.

Maintenance of the bacterial homeostasis initially emanates from interactions between proteins and the bacterial nucleoid. Investigating their spatial correlation requires high spatial resolution, especially in tiny, highly confined and crowded bacterial cells. Here, we present super-resolution microscopy using a palette of fluorescent labels that bind transiently to either the membrane or the nucleoid of fixed E. coli cells. The presented labels are easily applicable, versatile and allow long-term single-molecule super-resolution imaging independent of photobleaching. The different spectral properties allow for multiplexed imaging in combination with other localisation-based super-resolution imaging techniques. As examples for applications, we demonstrate correlated super-resolution imaging of the bacterial nucleoid with the position of genetic loci, of nascent DNA in correlation to the entire nucleoid, and of the nucleoid of metabolically arrested cells. We furthermore show that DNA- and membrane-targeting labels can be combined with photoactivatable fluorescent proteins and visualise the nano-scale distribution of RNA polymerase relative to the nucleoid in drug-treated E. coli cells.

The Pearling Transition Provides Evidence of Force-Driven Endosomal Tubulation during Salmonella Infection.

Bacterial pathogens exploit eukaryotic pathways for their own end. Upon ingestion, Salmonella enterica serovar Typhimurium passes through the stomach and then catalyzes its uptake across the intestinal epithelium. It survives and replicates in an acidic vacuole through the action of virulence factors secreted by a type three secretion system located on Salmonella pathogenicity island 2 (SPI-2). Two secreted effectors, SifA and SseJ, are sufficient for endosomal tubule formation, which modifies the vacuole and enables Salmonella to replicate within it. Two-color, superresolution imaging of the secreted virulence factor SseJ and tubulin revealed that SseJ formed clusters of conserved size at regular, periodic intervals in the host cytoplasm. Analysis of SseJ clustering indicated the presence of a pearling effect, which is a force-driven, osmotically sensitive process. The pearling transition is an instability driven by membranes under tension; it is induced by hypotonic or hypertonic buffer exchange and leads to the formation of beadlike structures of similar size and regular spacing. Reducing the osmolality of the fixation conditions using glutaraldehyde enabled visualization of continuous and intact tubules. Correlation analysis revealed that SseJ was colocalized with the motor protein kinesin. Tubulation of the endoplasmic reticulum is driven by microtubule motors, and in the present work, we describe how Salmonella has coopted the microtubule motor kinesin to drive the force-dependent process of endosomal tubulation. Thus, endosomal tubule formation is a force-driven process catalyzed by Salmonella virulence factors secreted into the host cytoplasm during infection.IMPORTANCE This study represents the first example of using two-color, superresolution imaging to analyze the secretion of Salmonella virulence factors as they are secreted from the SPI-2 type three secretion system. Previous studies imaged effectors that were overexpressed in the host cytoplasm. The present work reveals an unusual force-driven process, the pearling transition, which indicates that Salmonella-induced filaments are under force through the interactions of effector molecules with the motor protein kinesin. This work provides a caution by highlighting how fixation conditions can influence the images observed.

Sequential Super-Resolution Imaging of Bacterial Regulatory Proteins: The Nucleoid and the Cell Membrane in Single, Fixed E. coli Cells.

Despite their small size and the lack of compartmentalization, bacteria exhibit a striking degree of cellular organization, both in time and space. During the last decade, a group of new microscopy techniques emerged, termed super-resolution microscopy or nanoscopy, which facilitate visualizing the organization of proteins in bacteria at the nanoscale. Single-molecule localization microscopy (SMLM) is especially well suited to reveal a wide range of new information regarding protein organization, interaction, and dynamics in single bacterial cells. Recent developments in click chemistry facilitate the visualization of bacterial chromatin with a resolution of ~20 nm, providing valuable information about the ultrastructure of bacterial nucleoids, especially at short generation times. In this chapter, we describe a simple-to-realize protocol that allows determining precise structural information of bacterial nucleoids in fixed cells, using direct stochastic optical reconstruction microscopy (dSTORM). In combination with quantitative photoactivated localization microscopy (PALM), the spatial relationship of proteins with the bacterial chromosome can be studied. The position of a protein of interest with respect to the nucleoids and the cell cylinder can be visualized by super-resolving the membrane using point accumulation for imaging in nanoscale topography (PAINT). The combination of the different SMLM techniques in a sequential workflow maximizes the information that can be extracted from single cells, while maintaining optimal imaging conditions for each technique.

Nanoscopy of bacterial cells immobilized by holographic optical tweezers.

Imaging non-adherent cells by super-resolution far-field fluorescence microscopy is currently not possible because of their rapid movement while in suspension. Holographic optical tweezers (HOTs) enable the ability to freely control the number and position of optical traps, thus facilitating the unrestricted manipulation of cells in a volume around the focal plane. Here we show that immobilizing non-adherent cells by optical tweezers is sufficient to achieve optical resolution well below the diffraction limit using localization microscopy. Individual cells can be oriented arbitrarily but preferably either horizontally or vertically relative to the microscope's image plane, enabling access to sample sections that are impossible to achieve with conventional sample preparation and immobilization. This opens up new opportunities to super-resolve the nanoscale organization of chromosomal DNA in individual bacterial cells.

BACE-1 is expressed in the blood-brain barrier endothelium and is upregulated in a murine model of Alzheimer's disease.

Endothelial cells of the blood-brain barrier form a structural and functional barrier maintaining brain homeostasis via paracellular tight junctions and specific transporters such as P-glycoprotein. The blood-brain barrier is responsible for negligible bioavailability of many neuroprotective drugs. In Alzheimer's disease, current treatment approaches include inhibitors of BACE-1 (ß-site of amyloid precursor protein cleaving enzyme), a proteinase generating neurotoxic ß-amyloid. It is known that BACE-1 is highly expressed in endosomes and membranes of neurons and glia. We now provide evidence that BACE-1 is expressed in blood-brain barrier endothelial cells of human, mouse, and bovine origin. We further show its predominant membrane localization by 3D-dSTORM super-resolution microscopy, and by biochemical fractionation that further shows an abluminal distribution of BACE-1 in brain microvessels. We confirm its functionality in processing APP in primary mouse brain endothelial cells. In an Alzheimer's disease mouse model we show that BACE-1 is upregulated at the blood-brain barrier compared to healthy controls. We therefore suggest a critical role for BACE-1 at the blood-brain barrier in ß-amyloid generation and in vascular aspects of Alzheimer's disease, particularly in the development of cerebral amyloid angiopathy.

Single cell super-resolution imaging of E. coli OmpR during environmental stress.

Two-component signaling systems are a major strategy employed by bacteria, and to some extent, yeast and plants, to respond to environmental stress. The EnvZ/OmpR system in E. coli responds to osmotic and acid stress and is responsible for regulating the protein composition of the outer membrane. EnvZ is a histidine kinase located in the inner membrane. Upon activation, it is autophosphorylated by ATP and subsequently, it activates OmpR. Phosphorylated OmpR binds with high affinity to the regulatory regions of the ompF and ompC porin genes to regulate their transcription. We set out to visualize these two-components in single bacterial cells during different environmental stress conditions and to examine the subsequent modifications to the bacterial nucleoid as a result. We created a chromosomally-encoded, active, fluorescent OmpR-PAmCherry fusion protein and compared its expression levels with RNA polymerase. Quantitative western blotting had indicated that these two proteins were expressed at similar levels. From our images, it is evident that OmpR is significantly less abundant compared to RNA polymerase. In cross-sectional axial images, we observed OmpR molecules closely juxtaposed near the inner membrane during acidic and hyposomotic growth. In acidic conditions, the chromosome was compacted. Surprisingly, under acidic conditions, we also observed evidence of a spatial correlation between the DNA and the inner membrane, suggesting a mechanical link through an active DNA-OmpR-EnvZ complex. This work represents the first direct visualization of a response regulator with respect to the bacterial chromosome.

Correlative super-resolution imaging of RNA polymerase distribution and dynamics, bacterial membrane and chromosomal structure in Escherichia coli.

We demonstrate correlative super-resolution PALM, PAINT and dSTORM imaging of RNA polymerase, membrane and chromosomal DNA in fixed E. coli. This protocol allows the combination of precise structural information of the nucleoid (dSTORM) with quantitative super-resolution imaging (PALM) of interacting proteins. The spatial distribution and organization of RNA polymerase and DNA are visualized in bacterial cells grown at doubling times of 25 or 44 min. We observe that RNA polymerase is concentrated at the edge of the highly structured nucleoid during fast growth, whereas it is found more evenly distributed during medium-fast growth. In both conditions, the nucleoid shows densely packed areas which appear to be inaccessible to RNA polymerase. This finding is confirmed by live-cell tracking of RNA polymerase and subsequent imaging of the respective nucleoids using a protocol for fast fixation on-the-slide.

Super-resolution imaging of Escherichia coli nucleoids reveals highly structured and asymmetric segregation during fast growth.

Bacterial replication and chromosome segregation are highly organized both in space and in time. However, spatial analysis is hampered by the resolution limit of conventional fluorescence microscopy. In this study, we incubate rapidly-growing Escherichia coli with 5-ethynyl-2'-deoxyuridine (EdU), label the resulting EdU-DNA with photoswitchable fluorophores, and image incorporated molecules with an average experimental precision of 13 nm. During the segregation process, nucleoids develop highly-defined and cell-cycle dependent hetero-structures, which contain discrete DNA fibers with diameters far below the diffraction limit. Strikingly, these structures appear temporally shifted between sister chromosomes, an asymmetry which accumulates for ongoing replication rounds. Moreover, nucleoid positioning and expansion along the bacterial length axis fit into an elongation-mediated segregation model in fast growing E. coli cultures. This is supported by close proximity of the nucleoids to the bacterial plasma membrane, the nature of the observed hetero-structures and recently found interactions of membrane-associated proteins with DNA.

Increasing the brightness of cyanine fluorophores for single-molecule and superresolution imaging.

In spite of their relatively low fluorescence quantum yield, cyanine dyes such as Cy3, Cy5, or Cy7 are widely used in single-molecule fluorescence applications due to their high extinction coefficients and excellent photon yields. We show that the fluorescence quantum yield and lifetime of red-emitting cyanine dyes can be substantially increased in heavy water (D2 O) compared with water (H2 O). We find that the magnitude of the quantum yield increase in D2 O scales with the emission wavelength, reaching a particularly high value of 2.6-fold for the most red-emitting dye investigated, Cy7. We further demonstrate a higher photon yield in single-molecule superresolution experiments in D2 O compared to H2 O, which leads to an improved localization precision and hence better spatial resolution. This finding is especially beneficial for biological applications of fluorescence microscopy, which are typically carried out in aqueous media and which greatly profit from the red spectral range due to reduced cellular auto-fluorescence.

Click chemistry facilitates direct labelling and super-resolution imaging of nucleic acids and proteins†Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra01027bClick here for additional data file.

We demonstrate high-density labelling of cellular DNA and RNA using click chemistry and perform confocal and super-resolution microscopy. We visualize the crescent and ring-like structure of densely packed RNA in nucleoli. We further demonstrate click chemistry with unnatural amino acids for super-resolution imaging of outer-membrane proteins of E. coli.