Intelligent Image-Activated Cell Sorting.

A fundamental challenge of biology is to understand the vast heterogeneity of cells, particularly how cellular composition, structure, and morphology are linked to cellular physiology. Unfortunately, conventional technologies are limited in uncovering these relations. We present a machine-intelligence technology based on a radically different architecture that realizes real-time image-based intelligent cell sorting at an unprecedented rate. This technology, which we refer to as intelligent image-activated cell sorting, integrates high-throughput cell microscopy, focusing, and sorting on a hybrid software-hardware data-management infrastructure, enabling real-time automated operation for data acquisition, data processing, decision-making, and actuation. We use it to demonstrate real-time sorting of microalgal and blood cells based on intracellular protein localization and cell-cell interaction from large heterogeneous populations for studying photosynthesis and atherothrombosis, respectively. The technology is highly versatile and expected to enable machine-based scientific discovery in biological, pharmaceutical, and medical sciences.

Multi-foci parallelised RESOLFT nanoscopy in an extended field-of-view.

Live-cell imaging of biological structures at high resolution poses challenges in the microscope throughput regarding area and speed. For this reason, different parallelisation strategies have been implemented in coordinate- and stochastic-targeted switching super-resolution microscopy techniques. In this line, the molecular nanoscale live imaging with sectioning ability (MoNaLISA), based on reversible saturable optical fluorescence transitions (RESOLFT), offers 45 - 65 nm $45 - 65\;{\rm{nm}}$ resolution of large fields of view in a few seconds. In MoNaLISA, engineered light patterns strategically confine the fluorescence to sub-diffracted volumes in a large area and provide optical sectioning, thus enabling volumetric imaging at high speeds. The optical setup presented in this paper extends the degree of parallelisation of the MoNaLISA microscope by more than four times, reaching a field-of-view of ( 100 - 130 µ m ) 2 ${( {100 - 130\;{\rm{\mu m}}} )^2}$ . We set up the periodicity and the optical scheme of the illumination patterns to be power-efficient and homogeneous. In a single recording, this new configuration enables super-resolution imaging of an extended population of the post-synaptic density protein Homer1c in living hippocampal neurons.

Machine Learning Methods for Endocrine Disrupting Potential Identification Based on Single-Cell Data.

Humans are continuously exposed to a variety of toxicants and chemicals which is exacerbated during and after environmental catastrophes such as floods, earthquakes, and hurricanes. The hazardous chemical mixtures generated during these events threaten the health and safety of humans and other living organisms. This necessitates the development of rapid decision-making tools to facilitate mitigating the adverse effects of exposure on the key modulators of the endocrine system, such as the estrogen receptor alpha (ERα), for example. The mechanistic stages of the estrogenic transcriptional activity can be measured with high content/high throughput microscopy-based biosensor assays at the single-cell level, which generates millions of object-based minable data points. By combining computational modeling and experimental analysis, we built a highly accurate data-driven classification framework to assess the endocrine disrupting potential of environmental compounds. The effects of these compounds on the ERα pathway are predicted as being receptor agonists or antagonists using the principal component analysis (PCA) projections of high throughput, high content image analysis descriptors. The framework also combines rigorous preprocessing steps and nonlinear machine learning algorithms, such as the Support Vector Machines and Random Forest classifiers, to develop highly accurate mathematical representations of the separation between ERα agonists and antagonists. The results show that Support Vector Machines classify the unseen chemicals correctly with more than 96% accuracy using the proposed framework, where the preprocessing and the PCA steps play a key role in suppressing experimental noise and unraveling hidden patterns in the dataset.

Cells infected with human papilloma pseudovirus display nuclear reorganization and heterogenous infection kinetics.

Human papillomaviruses (HPV) are small, non-enveloped DNA viruses, which upon chronic infection can provoke cervical and head-and-neck cancers. Although the infectious life cycle of HPV has been studied and a vaccine is available for the most prevalent cancer-causing HPV types, there are no antiviral agents to treat infected patients. Hence, there is a need for novel therapeutic entry points and a means to identify them. In this work, we have used high-content microscopy to quantitatively investigate the early phase of HPV infection. Human cervical cancer cells and immortalized keratinocytes were exposed to pseudoviruses (PsV) of the widespread HPV type 16, in which the viral genome was replaced by a pseudogenome encoding a fluorescent reporter protein. Using the fluorescent signal as readout, we measured differences in infection between cell lines, which directly correlated with host cell proliferation rate. Parallel multiparametric analysis of nuclear organization revealed that HPV PsV infection alters nuclear organization and inflates promyelocytic leukemia protein body content, positioning these events at the early stage of HPV infection, upstream of viral replication. Time-resolved analysis revealed a marked heterogeneity in infection kinetics even between two daughter cells, which we attribute to differences in viral load. Consistent with the requirement for mitotic nuclear envelope breakdown, pharmacological inhibition of the cell cycle dramatically blunted infection efficiency. Thus, by systematic image-based single cell analysis, we revealed phenotypic alterations that accompany HPV PsV infection in individual cells, and which may be relevant for therapeutic drug screens.

MAPS: machine-assisted phenotype scoring enables rapid functional assessment of genetic variants by high-content microscopy.

BACKGROUND: Genetic testing is widely used in evaluating a patient's predisposition to hereditary diseases. In the case of cancer, when a functionally impactful mutation (i.e. genetic variant) is identified in a disease-relevant gene, the patient is at elevated risk of developing a lesion in their lifetime. Unfortunately, as the rate and coverage of genetic testing has accelerated, our ability to assess the functional status of new variants has fallen behind. Therefore, there is an urgent need for more practical, streamlined and cost-effective methods for classifying variants. RESULTS: To directly address this issue, we designed a new approach that uses alterations in protein subcellular localization as a key indicator of loss of function. Thus, new variants can be rapidly functionalized using high-content microscopy (HCM). To facilitate the analysis of the large amounts of imaging data, we developed a new software toolkit, named MAPS for machine-assisted phenotype scoring, that utilizes deep learning to extract and classify cell-level features. MAPS helps users leverage cloud-based deep learning services that are easy to train and deploy to fit their specific experimental conditions. Model training is code-free and can be done with limited training images. Thus, MAPS allows cell biologists to easily incorporate deep learning into their image analysis pipeline. We demonstrated an effective variant functionalization workflow that integrates HCM and MAPS to assess missense variants of PTEN, a tumor suppressor that is frequently mutated in hereditary and somatic cancers. CONCLUSIONS: This paper presents a new way to rapidly assess variant function using cloud deep learning. Since most tumor suppressors have well-defined subcellular localizations, our approach could be widely applied to functionalize variants of uncertain significance and help improve the utility of genetic testing.

Multi-labelled proteins recognition for high-throughput microscopy images using deep convolutional neural networks.

BACKGROUND: Proteins are of extremely vital importance in the human body, and no movement or activity can be performed without proteins. Currently, microscopy imaging technologies developed rapidly are employed to observe proteins in various cells and tissues. In addition, due to the complex and crowded cellular environments as well as various types and sizes of proteins, a considerable number of protein images are generated every day and cannot be classified manually. Therefore, an automatic and accurate method should be designed to properly solve and analyse protein images with mixed patterns. RESULTS: In this paper, we first propose a novel customized architecture with adaptive concatenate pooling and "buffering" layers in the classifier part, which could make the networks more adaptive to training and testing datasets, and develop a novel hard sampler at the end of our network to effectively mine the samples from small classes. Furthermore, a new loss is presented to handle the label imbalance based on the effectiveness of samples. In addition, in our method, several novel and effective optimization strategies are adopted to solve the difficult training-time optimization problem and further increase the accuracy by post-processing. CONCLUSION: Our methods outperformed the SOTA method of multi-labelled protein classification on the HPA dataset, GapNet-PL, by above 2% in the F1 score. Therefore, experimental results based on the test set split from the Human Protein Atlas dataset show that our methods have good performance in automatically classifying multi-class and multi-labelled high-throughput microscopy protein images.

Cargo selectivity of yeast sorting nexins.

Sorting nexins are PX domain-containing proteins that bind phospholipids and often act in membrane trafficking where they help to select cargo. However, the functions and cargo specificities of many sorting nexins are unknown. Here, a high-throughput imaging screen was used to identify new sorting nexin cargo in the yeast Saccharomyces cerevisiae. Deletions of 9 different sorting nexins were screened for mislocalization of a set of green fluorescent protein (GFP)-tagged membrane proteins found at the plasma membrane, Golgi or endosomes. This identified 27 proteins that require 1 or more sorting nexins for their correct localization, 23 of which represent novel sorting nexin cargo. Nine hits whose sorting was dependent on Snx4, the sorting nexin-containing retromer complex, or both retromer and Snx3, were examined in detail to search for potential sorting motifs. We identified cytosolic domains of Ear1, Ymd8 and Ymr010w that conferred retromer-dependent sorting on a chimeric reporter and identified conserved residues required for this sorting in a functional assay. This work defined a consensus sequence for retromer and Snx3-dependent sorting.

Quantitative comparison of taxa and taxon concepts in the diatom genus Fragilariopsis: a case study on using slide scanning, multiexpert image annotation, and image analysis in taxonomy1.

Semiautomated methods for microscopic image acquisition, image analysis, and taxonomic identification have repeatedly received attention in diatom analysis. Less well studied is the question whether and how such methods might prove useful for clarifying the delimitation of species that are difficult to separate for human taxonomists. To try to answer this question, three very similar Fragilariopsis species endemic to the Southern Ocean were targeted in this study: F. obliquecostata, F. ritscheri, and F. sublinearis. A set of 501 extended focus depth specimen images were obtained using a standardized, semiautomated microscopic procedure. Twelve diatomists independently identified these specimen images in order to reconcile taxonomic opinions and agree upon a taxonomic gold standard. Using image analyses, we then extracted morphometric features representing taxonomic characters of the target taxa. The discriminating ability of individual morphometric features was tested visually and statistically, and multivariate classification experiments were performed to test the agreement of the quantitatively defined taxa assignments with expert consensus opinion. Beyond an updated differential diagnosis of the studied taxa, our study also shows that automated imaging and image analysis procedures for diatoms are coming close to reaching a broad applicability for routine use.

Large-scale image-based screening and profiling of cellular phenotypes.

Cellular phenotypes are observable characteristics of cells resulting from the interactions of intrinsic and extrinsic chemical or biochemical factors. Image-based phenotypic screens under large numbers of basal or perturbed conditions can be used to study the influences of these factors on cellular phenotypes. Hundreds to thousands of phenotypic descriptors can also be quantified from the images of cells under each of these experimental conditions. Therefore, huge amounts of data can be generated, and the analysis of these data has become a major bottleneck in large-scale phenotypic screens. Here, we review current experimental and computational methods for large-scale image-based phenotypic screens. Our focus is on phenotypic profiling, a computational procedure for constructing quantitative and compact representations of cellular phenotypes based on the images collected in these screens. © 2016 International Society for Advancement of Cytometry.

Three-dimensional imaging flow cytometry through light-sheet fluorescence microscopy.

Flow cytometry is the tool of choice for high-speed acquisition and analysis of large cell populations, with the tradeoff of lacking intracellular spatial information. Although in the last decades flow cytometry systems that can actually acquire two-dimensional spatial information were developed, some of the limitations remained though, namely constrains related to sample size and lack of depth or dynamic information. The combination of fluidics and light-sheet illumination has the potential to address these limitations. By having cells travelling with the flowing sheath one can, in a controlled fashion, force them at constant speed through the light-sheet enabling the synchronized acquisition of several optical sections, that is, three-dimensional imaging. This approach has already been used for imaging cellular spheroids, plankton, and zebra-fish embryos. In this review, we discuss the known solutions and standing challenges of performing three-dimensional high-throughput imaging of multicellular biological models using fluidics, while retaining cell and organelle-level resolution. © 2017 International Society for Advancement of Cytometry.

Multiplexed imaging of intracellular protein networks.

Cellular functions emerge from the collective action of a large number of different proteins. Understanding how these protein networks operate requires monitoring their components in intact cells. Due to intercellular and intracellular molecular variability, it is important to monitor simultaneously multiple components at high spatiotemporal resolution. However, inherent trade-offs narrow the boundaries of achievable multiplexed imaging. Pushing these boundaries is essential for a better understanding of cellular processes. Here the motivations, challenges and approaches for multiplexed imaging of intracellular protein networks are discussed. © 2016 International Society for Advancement of Cytometry.

Maximizing the quantitative accuracy and reproducibility of Förster resonance energy transfer measurement for screening by high throughput widefield microscopy.

Förster resonance energy transfer (FRET) between fluorescent proteins (FPs) provides insights into the proximities and orientations of FPs as surrogates of the biochemical interactions and structures of the factors to which the FPs are genetically fused. As powerful as FRET methods are, technical issues have impeded their broad adoption in the biologic sciences. One hurdle to accurate and reproducible FRET microscopy measurement stems from variable fluorescence backgrounds both within a field and between different fields. Those variations introduce errors into the precise quantification of fluorescence levels on which the quantitative accuracy of FRET measurement is highly dependent. This measurement error is particularly problematic for screening campaigns since minimal well-to-well variation is necessary to faithfully identify wells with altered values. High content screening depends also upon maximizing the numbers of cells imaged, which is best achieved by low magnification high throughput microscopy. But, low magnification introduces flat-field correction issues that degrade the accuracy of background correction to cause poor reproducibility in FRET measurement. For live cell imaging, fluorescence of cell culture media in the fluorescence collection channels for the FPs commonly used for FRET analysis is a high source of background error. These signal-to-noise problems are compounded by the desire to express proteins at biologically meaningful levels that may only be marginally above the strong fluorescence background. Here, techniques are presented that correct for background fluctuations. Accurate calculation of FRET is realized even from images in which a non-flat background is 10-fold higher than the signal.

A Novel Nanoprobe for Multimodal Imaging Is Effectively Incorporated into Human Melanoma Metastatic Cell Lines.

To facilitate efficient drug delivery to tumor tissue, several nanomaterials have been designed, with combined diagnostic and therapeutic properties. In this work, we carried out fundamental in vitro and in vivo experiments to assess the labeling efficacy of our novel theranostic nanoprobe, consisting of glycogen conjugated with a red fluorescent probe and gadolinium. Microscopy and resazurin viability assays were used to study cell labeling and cell viability in human metastatic melanoma cell lines. Fluorescence lifetime correlation spectroscopy (FLCS) was done to investigate nanoprobe stability. Magnetic resonance imaging (MRI) was performed to study T1 relaxivity in vitro, and contrast enhancement in a subcutaneous in vivo tumor model. Efficient cell labeling was demonstrated, while cell viability, cell migration, and cell growth was not affected. FLCS showed that the nanoprobe did not degrade in blood plasma. MRI demonstrated that down to 750 cells/µL of labeled cells in agar phantoms could be detected. In vivo MRI showed that contrast enhancement in tumors was comparable between Omniscan contrast agent and the nanoprobe. In conclusion, we demonstrate for the first time that a non-toxic glycogen-based nanoprobe may effectively visualize tumor cells and tissue, and, in future experiments, we will investigate its therapeutic potential by conjugating therapeutic compounds to the nanoprobe.

Pipeline for illumination correction of images for high-throughput microscopy.

The presence of systematic noise in images in high-throughput microscopy experiments can significantly impact the accuracy of downstream results. Among the most common sources of systematic noise is non-homogeneous illumination across the image field. This often adds an unacceptable level of noise, obscures true quantitative differences and precludes biological experiments that rely on accurate fluorescence intensity measurements. In this paper, we seek to quantify the improvement in the quality of high-content screen readouts due to software-based illumination correction. We present a straightforward illumination correction pipeline that has been used by our group across many experiments. We test the pipeline on real-world high-throughput image sets and evaluate the performance of the pipeline at two levels: (a) Z'-factor to evaluate the effect of the image correction on a univariate readout, representative of a typical high-content screen, and (b) classification accuracy on phenotypic signatures derived from the images, representative of an experiment involving more complex data mining. We find that applying the proposed post-hoc correction method improves performance in both experiments, even when illumination correction has already been applied using software associated with the instrument. To facilitate the ready application and future development of illumination correction methods, we have made our complete test data sets as well as open-source image analysis pipelines publicly available. This software-based solution has the potential to improve outcomes for a wide-variety of image-based HTS experiments.

Anti-Aspergillus activities of olorofim at sub-MIC levels during early-stage growth.

Olorofim, the first member of the novel class of antifungal drugs, the orotomides, shows promising anti-Aspergillus activity and is currently in phase III clinical development. Using high-throughput microscopy, we monitored olorofim's antifungal potential at sub-minimum inhibitory concentration (MIC) levels with a focus on early-stage growth. Unlike voriconazole, olorofim showed significant growth inhibitory activities against three main pathogenic Aspergillus species, Aspergillus fumigatus, Aspergillus flavus, and Aspergillus niger, at concentrations >100,000-fold below its MIC. IMPORTANCE: Among antifungal compounds in clinical development for systemic disease, the orotomide olorofim is one of only two that target a completely new mechanism of action. Olorofim is highly potent against pathogenic Aspergillus species including cryptic species that frequently show increased resistance to current agents. In this study, our primary focus was on evaluating in detail the inhibitory activity of voriconazole and olorofim against different pathogenic Aspergillus species employing high-throughput microscopy. Compared to standardized, less-sensitive visual assessment-based methods, microscopy-assisted growth monitoring allowed us to detect sub-MIC drug concentration ranges with significant inhibitory activity at early-stage growth. This revealed that olorofim exerts growth inhibition at concentrations that are several magnitudes below those of voriconazole.

Multi-omic analysis reveals genes and proteins integral to bioactivity of Echinochrome A isolated from the waste stream of the sea urchin industry in Aotearoa New Zealand.

Evechinus chloroticus (commonly known as kina) is a sea urchin species endemic to New Zealand. Its roe is a culinary delicacy to the indigenous Maori and a globally exported food product. Echinochrome A (Ech A) is a bioactive compound isolated from the waste product of kina shells and spines; however, the molecular mechanisms of Ech A bioactivity are not well understood, partly due to Ech A never being studied using unbiased genome-wide analysis. To explore the high-value pharmaceutical potential of kina food waste, we obtained unbiased functional genomic and proteomic profiles of yeast cells treated with Echinochrome A. Abundance was measured for 4100 proteins every 30 min for four hours using fluorescent microscopy, resulting in the identification of 92 proteins with significant alterations in protein abundance caused by Ech A treatment that were over-represented with specific changes in DNA replication, repair and RNA binding after 30 min, followed by specific changes in the metabolism of metal ions (specifically iron and copper) from 60-240 min. Further analysis indicated that Ech A chelated iron, and that iron supplementation negated the growth inhibition caused by Ech A. Via a growth-based genome-wide analysis of 4800 gene deletion strains, 20 gene deletion strains were sensitive to Ech A in an iron-dependent manner. These genes were over-represented in the cellular response to oxidative stress, suggesting that Ech A suppressed growth inhibition caused by oxidative stress. Unexpectedly, genes integral to cardiolipin and inositol phosphate biosynthesis were required for Ech A bioactivity. Overall, these results identify genes, proteins, and cellular processes mediating the bioactivity of Ech A. Moreover, we demonstrate unbiased genomic and proteomic methodology that will be useful for characterizing bioactive compounds in food and food waste.

A novel ERß high throughput microscopy platform for testing endocrine disrupting chemicals.

In this study we present an inducible biosensor model for the Estrogen Receptor Beta (ERß), GFP-ERß:PRL-HeLa, a single-cell-based high throughput (HT) in vitro assay that allows direct visualization and measurement of GFP-tagged ERß binding to ER-specific DNA response elements (EREs), ERß-induced chromatin remodeling, and monitor transcriptional alterations via mRNA fluorescence in situ hybridization for a prolactin (PRL)-dsRED2 reporter gene. The model was used to accurately (Z' = 0.58-0.8) differentiate ERß-selective ligands from ERα ligands when treated with a panel of selective agonists and antagonists. Next, we tested an Environmental Protection Agency (EPA)-provided set of 45 estrogenic reference chemicals with known ERα in vivo activity and identified several that activated ERß as well, with varying sensitivity, including a subset that is completely novel. We then used an orthogonal ERE-containing transgenic zebrafish (ZF) model to cross validate ERß and ERα selective activities at the organism level. Using this environmentally relevant ZF assay, some compounds were confirmed to have ERß activity, validating the GFP-ERß:PRL-HeLa assay as a screening tool for potential ERß active endocrine disruptors (EDCs). These data demonstrate the value of sensitive multiplex mechanistic data gathered by the GFP-ERß:PRL-HeLa assay coupled with an orthogonal zebrafish model to rapidly identify environmentally relevant ERß EDCs and improve upon currently available screening tools for this understudied nuclear receptor.

High throughput microscopy and single cell phenotypic image-based analysis in toxicology and drug discovery.

Measuring single cell responses to the universe of chemicals (drugs, natural products, environmental toxicants etc.) is of paramount importance to human health as phenotypic variability in sensing stimuli is a hallmark of biology that is considered during high throughput screening. One of the ways to approach this problem is via high throughput, microscopy-based assays coupled with multi-dimensional single cell analysis methods. Here, we will summarize some of the efforts in this vast and growing field, focusing on phenotypic screens (e.g., Cell Painting), single cell analytics and quality control, with particular attention to environmental toxicology and drug screening. We will discuss advantages and limitations of high throughput assays with various end points and levels of complexity.

High-throughput "read-on-ski" automated imaging and label-free detection system for toxicity screening of compounds using personalised human kidney organoids.

Organoid models are used to study kidney physiology, such as the assessment of nephrotoxicity and underlying disease processes. Personalized human pluripotent stem cell-derived kidney organoids are ideal models for compound toxicity studies, but there is a need to accelerate basic and translational research in the field. Here, we developed an automated continuous imaging setup with the "read-on-ski" law of control to maximize temporal resolution with minimum culture plate vibration. High-accuracy performance was achieved: organoid screening and imaging were performed at a spatial resolution of 1.1 µm for the entire multi-well plate under 3 min. We used the in-house developed multi-well spinning device and cisplatin-induced nephrotoxicity model to evaluate the toxicity in kidney organoids using this system. The acquired images were processed via machine learning-based classification and segmentation algorithms, and the toxicity in kidney organoids was determined with 95% accuracy. The results obtained by the automated "read-on-ski" imaging device, combined with label-free and non-invasive algorithms for detection, were verified using conventional biological procedures. Taking advantage of the close-to-in vivo-kidney organoid model, this new development opens the door for further application of scaled-up screening using organoids in basic research and drug discovery.

Automated counting of airborne asbestos fibers by a high-throughput microscopy (HTM) method.

Inhalation of airborne asbestos causes serious health problems such as lung cancer and malignant mesothelioma. The phase-contrast microscopy (PCM) method has been widely used for estimating airborne asbestos concentrations because it does not require complicated processes or high-priced equipment. However, the PCM method is time-consuming and laborious as it is manually performed off-site by an expert. We have developed a high-throughput microscopy (HTM) method that can detect fibers distinguishable from other spherical particles in a sample slide by image processing both automatically and quantitatively. A set of parameters for processing and analysis of asbestos fiber images was adjusted for standard asbestos samples with known concentrations. We analyzed sample slides containing airborne asbestos fibers collected at 11 different workplaces following PCM and HTM methods, and found a reasonably good agreement in the asbestos concentration. Image acquisition synchronized with the movement of the robotic sample stages followed by an automated batch processing of a stack of sample images enabled us to count asbestos fibers with greatly reduced time and labors. HTM should be a potential alternative to conventional PCM, moving a step closer to realization of on-site monitoring of asbestos fibers in air.