Quantification of Giant Unilamellar Vesicle Fusion Products by High-Throughput Image Analysis.

Artificial cells are based on dynamic compartmentalized systems. Thus, remodeling of membrane-bound systems, such as giant unilamellar vesicles, is finding applications beyond biological studies, to engineer cell-mimicking structures. Giant unilamellar vesicle fusion is rapidly becoming an essential experimental step as artificial cells gain prominence in synthetic biology. Several techniques have been developed to accomplish this step, with varying efficiency and selectivity. To date, characterization of vesicle fusion has relied on small samples of giant vesicles, examined either manually or by fluorometric assays on suspensions of small and large unilamellar vesicles. Automation of the detection and characterization of fusion products is now necessary for the screening and optimization of these fusion protocols. To this end, we implemented a fusion assay based on fluorophore colocalization on the membranes and in the lumen of vesicles. Fluorescence colocalization was evaluated within single compartments by image segmentation with minimal user input, allowing the application of the technique to high-throughput screenings. After detection, statistical information on vesicle fluorescence and morphological properties can be summarized and visualized, assessing lipid and content transfer for each object by the correlation coefficient of different fluorescence channels. Using this tool, we report and characterize the unexpected fusogenic activity of sodium chloride on phosphatidylcholine giant vesicles. Lipid transfer in most of the vesicles could be detected after 20 h of incubation, while content exchange only occurred with additional stimuli in around 8% of vesicles.

Automated cell isolation from photodegradable hydrogel based on fluorescence image analysis.

We report an automated cell-isolation system based on fluorescence image analysis of cell aggregates cultured in a photodegradable hydrogel. The system incorporates cell culture in a humidified atmosphere with controlled CO2 concentration and temperature, image acquisition and analysis, micropatterned light exposure, and cell collection by pipetting. Cell aggregates were cultured on hydrogels, and target cells were selected by phase contrast and fluorescence image analysis. After degradation of the hydrogel by exposure to micropatterned UV light, cell aggregates were transferred to a collection vessel by robotic pipetting. We assessed the system for hydrogel degradation, recovery of target cells, and contamination by off-target cells. We demonstrated two practical applications of our method: (i) in cell aggregates from MCF-7-RFP strains in which 18.8% of cells produced red fluorescent protein (RFP), we successfully obtained 14 proliferative fluorescence-positive cell aggregates from 31-wells, and all of the isolated strains produced a higher proportion of RFP production than the original populations; (ii) after fluorescent immunostaining of human epidermal growth factor receptor 2 (HER2) in cancer cells, we successfully isolated HER2-positive cells from a mixed population of HER2-positive and -negative cells, and gene sequence analysis confirmed that the isolated cells mainly contained the target cells.

Myelin detection in fluorescence microscopy images using machine learning.

BACKGROUND: The myelin sheath produced by glial cells insulates the axons, and supports the function of the nervous system. Myelin sheath degeneration causes neurodegenerative disorders, such as multiple sclerosis (MS). There are no therapies for MS that promote remyelination. Drug discovery frequently involves screening thousands of compounds. However, this is not feasible for remyelination drugs, since myelin quantification is a manual labor-intensive endeavor. Therefore, the development of assistive software for expedited myelin detection is instrumental for MS drug discovery by enabling high-content image-based drug screens. NEW METHOD: In this study, we developed a machine learning based expedited myelin detection approach in fluorescence microscopy images. Multi-channel three-dimensional microscopy images of a mouse stem cell-based myelination assay were labeled by experts. A spectro-spatial feature extraction method was introduced to represent local dependencies of voxels both in spatial and spectral domains. Feature extraction yielded two data set of over forty-seven thousand annotated images in total. RESULTS: Myelin detection performances of 23 different supervised machine learning techniques including a customized-convolutional neural network (CNN), were assessed using various train/test split ratios of the data sets. The highest accuracy values of 98.84±0.09% and 98.46±0.11% were achieved by Boosted Trees and customized-CNN, respectively. COMPARISON WITH EXISTING METHODS: Our approach can detect myelin in a common experimental setup. Myelin extending in any orientation in 3 dimensions is segmented from 3 channel z-stack fluorescence images. CONCLUSIONS: Our results suggest that the proposed expedited myelin detection approach is a feasible and robust method for remyelination drug screening.

Protein Crystallization Segmentation and Classification Using Subordinate Color Channel in Fluorescence Microscopy Images.

The accuracy of detecting protein crystals for fluorescence microscopy images is very critical for high throughput and automated systems. Although the trace fluorescent labeling method could highlight protein crystals, reflection and emission from the fluorescence dye is not always due to crystal regions. Therefore, the analysis of the peak wavelength in the emission spectra of a fluorophore may not always yield effective results. In this paper, we show that using the subordinate color intensity corresponding to longer wavelengths than the peak wavelength of the emission spectra could improve the accuracy of protein crystal detection. Hence, we have built a segmentation method based on the percentile intensity of the subordinate color for trace fluorescently labeled (TFL'd) protein crystallization trial images. Compared to using the dominant color channel, our segmentation method on subordinate color channel was able to reduce the misclassification rate of likely-leads or crystals as non-crystals by the percentage of from 9.71% to 2.02% depending on the classifier. Similarly, the accuracy of classifiers were increased by the percentage of from 1.77% to 5.53%. Our method reached around 94% accuracy while keeping misclassification of likely-leads and crystals as non-crystals below 1%. Moreover, to evaluate the generalizability of our method, we have conducted new wet lab experiments on two proteins, Concanavalin A (Con A) and Ab inorganic pyrophosphate (AbIPPase), and the misclassification rate was below 1%. Our experiments show that using the subordinate channel may be more helpful for TFL'd protein trial image classification.

Background-free upconversion-encoded microspheres for mycotoxin detection based on a rapid visualization method.

Methods for detecting mycotoxins are very important because of the great health hazards of mycotoxins. However, there is a high background and low signal-to-noise ratio in real-time sensing, and therefore it is difficult to meet the fast, accurate, and convenient requirements for control of food quality. Here we constructed a quantitative fluorescence image analysis based on multicolor upconversion nanocrystal (UCN)-encoded microspheres for detection of ochratoxin A and zearalenone. The background-free encoding image signal of UCN-doped microspheres was captured by fluorescence microscopy under near-infrared excitation, whereas the detection image signal of phycoerythrin-labeled secondary antibodies conjugated to the microspheres was captured under blue light excitation. We custom-wrote an algorithm to analyze the two images for the same sample in 10 s, and only the gray value in the red channel of the secondary probe confirmed the quantity. The results showed that this novel detection platform performed feasible and reliable fluorescence image measurements by this method. Additionally, the limit of detection of was 0.34721 ng/mL for ochratoxin A and 0.41162 ng/mL for zearalenone. We envision that this UCN encoding strategy will be usefully applied for fast, accurate, and convenient testing of multiple food contaminants to ensure the safety of the food.

Subgroup Characteristics of Marine Methane-Oxidizing ANME-2 Archaea and Their Syntrophic Partners as Revealed by Integrated Multimodal Analytical Microscopy.

Phylogenetically diverse environmental ANME archaea and sulfate-reducing bacteria cooperatively catalyze the anaerobic oxidation of methane oxidation (AOM) in multicelled consortia within methane seep environments. To better understand these cells and their symbiotic associations, we applied a suite of electron microscopy approaches, including correlative fluorescence in situ hybridization-electron microscopy (FISH-EM), transmission electron microscopy (TEM), and serial block face scanning electron microscopy (SBEM) three-dimensional (3D) reconstructions. FISH-EM of methane seep-derived consortia revealed phylogenetic variability in terms of cell morphology, ultrastructure, and storage granules. Representatives of the ANME-2b clade, but not other ANME-2 groups, contained polyphosphate-like granules, while some bacteria associated with ANME-2a/2c contained two distinct phases of iron mineral chains resembling magnetosomes. 3D segmentation of two ANME-2 consortium types revealed cellular volumes of ANME and their symbiotic partners that were larger than previous estimates based on light microscopy. Polyphosphate-like granule-containing ANME (tentatively termed ANME-2b) were larger than both ANME with no granules and partner bacteria. This cell type was observed with up to 4 granules per cell, and the volume of the cell was larger in proportion to the number of granules inside it, but the percentage of the cell occupied by these granules did not vary with granule number. These results illuminate distinctions between ANME-2 archaeal lineages and partnering bacterial populations that are apparently unified in their ability to perform anaerobic methane oxidation.IMPORTANCE Methane oxidation in anaerobic environments can be accomplished by a number of archaeal groups, some of which live in syntrophic relationships with bacteria in structured consortia. Little is known of the distinguishing characteristics of these groups. Here, we applied imaging approaches to better understand the properties of these cells. We found unexpected morphological, structural, and volume variability of these uncultured groups by correlating fluorescence labeling of cells with electron microscopy observables.

Phosphatidylserine Lateral Organization Influences the Interaction of Influenza Virus Matrix Protein 1 with Lipid Membranes.

Influenza A virus matrix protein 1 (M1) is an essential component involved in the structural stability of the virus and in the budding of new virions from infected cells. A deeper understanding of the molecular basis of virion formation and the budding process is required in order to devise new therapeutic approaches. We performed a detailed investigation of the interaction between M1 and phosphatidylserine (PS) (i.e., its main binding target at the plasma membrane [PM]), as well as the distribution of PS itself, both in model membranes and in living cells. To this end, we used a combination of techniques, including Förster resonance energy transfer (FRET), confocal microscopy imaging, raster image correlation spectroscopy, and number and brightness (N&B) analysis. Our results show that PS can cluster in segregated regions in the plane of the lipid bilayer, both in model bilayers constituted of PS and phosphatidylcholine and in living cells. The viral protein M1 interacts specifically with PS-enriched domains, and such interaction in turn affects its oligomerization process. Furthermore, M1 can stabilize PS domains, as observed in model membranes. For living cells, the presence of PS clusters is suggested by N&B experiments monitoring the clustering of the PS sensor lactadherin. Also, colocalization between M1 and a fluorescent PS probe suggest that, in infected cells, the matrix protein can specifically bind to the regions of PM in which PS is clustered. Taken together, our observations provide novel evidence regarding the role of PS-rich domains in tuning M1-lipid and M1-M1 interactions at the PM of infected cells.IMPORTANCE Influenza virus particles assemble at the plasma membranes (PM) of infected cells. This process is orchestrated by the matrix protein M1, which interacts with membrane lipids while binding to the other proteins and genetic material of the virus. Despite its importance, the initial step in virus assembly (i.e., M1-lipid interaction) is still not well understood. In this work, we show that phosphatidylserine can form lipid domains in physical models of the inner leaflet of the PM. Furthermore, the spatial organization of PS in the plane of the bilayer modulates M1-M1 interactions. Finally, we show that PS domains appear to be present in the PM of living cells and that M1 seems to display a high affinity for them.

Neurite analyzer: An original Fiji plugin for quantification of neuritogenesis in two-dimensional images.

BACKGROUND: In life sciences, there is a growing need for new informatics tools designed to provide automated solutions in order to analyze big amounts of images obtained from high-throughput imaging systems. Among the most widely used assays in neurotoxicity, endocrinology and brain diseases, the neurite outgrowth assay is popular. NEW METHOD: Cell-to-cell quantification of the main morphological features of neurite outgrowth assays remains very challenging. Here, we provide a new pipeline developed on Fiji software for analysis of series of two-dimensional images. It allows the automated analysis of most of these features. RESULTS: We tested the accuracy and usefulness of the software by confirming the effects of estradiol and hypoxia on in vitro neuronal differentiation, previously published by different authors with manual analysis methods. With this new method, we highlighted original interesting data. COMPARISON WITH EXISTING METHOD(S): The innovation brought by this plugin lies in the fact that it can process multiple images at the same time, in order to obtain: the number of nuclei, the number of neurites, the length of neurites, the number of neurites junctions, the number of neurites branches, the length of each branch, the position of the branch in the image, the angle of each branch, but also the area of each cell and the number of neurites per cell. CONCLUSIONS: This plugin is easy to use, highly sensitive, and allows the experimenter to acquire ready-to-use data coming from a vast amount of images.