A cell-level quality control workflow for high-throughput image analysis.

BACKGROUND: Image-based high throughput (HT) screening provides a rich source of information on dynamic cellular response to external perturbations. The large quantity of data generated necessitates computer-aided quality control (QC) methodologies to flag imaging and staining artifacts. Existing image- or patch-level QC methods require separate thresholds to be simultaneously tuned for each image quality metric used, and also struggle to distinguish between artifacts and valid cellular phenotypes. As a result, extensive time and effort must be spent on per-assay QC feature thresholding, and valid images and phenotypes may be discarded while image- and cell-level artifacts go undetected. RESULTS: We present a novel cell-level QC workflow built on machine learning approaches for classifying artifacts from HT image data. First, a phenotype sampler based on unlabeled clustering collects a comprehensive subset of cellular phenotypes, requiring only the inspection of a handful of images per phenotype for validity. A set of one-class support vector machines are then trained on each biologically valid image phenotype, and used to classify individual objects in each image as valid cells or artifacts. We apply this workflow to two real-world large-scale HT image datasets and observe that the ratio of artifact to total object area (ARcell) provides a single robust assessment of image quality regardless of the underlying causes of quality issues. Gating on this single intuitive metric, partially contaminated images can be salvaged and highly contaminated images can be excluded before image-level phenotype summary, enabling a more reliable characterization of cellular response dynamics. CONCLUSIONS: Our cell-level QC workflow enables identification of artificial cells created not only by staining or imaging artifacts but also by the limitations of image segmentation algorithms. The single readout ARcell that summaries the ratio of artifacts contained in each image can be used to reliably rank images by quality and more accurately determine QC cutoff thresholds. Machine learning-based cellular phenotype clustering and sampling reduces the amount of manual work required for training example collection. Our QC workflow automatically handles assay-specific phenotypic variations and generalizes to different HT image assays.

A high-throughput screen to identify inhibitors of ATP homeostasis in non-replicating Mycobacterium tuberculosis.

Growing evidence suggests that the presence of a subpopulation of hypoxic non-replicating, phenotypically drug-tolerant mycobacteria is responsible for the prolonged duration of tuberculosis treatment. The discovery of new antitubercular agents active against this subpopulation may help in developing new strategies to shorten the time of tuberculosis therapy. Recently, the maintenance of a low level of bacterial respiration was shown to be a point of metabolic vulnerability in Mycobacterium tuberculosis. Here, we describe the development of a hypoxic model to identify compounds targeting mycobacterial respiratory functions and ATP homeostasis in whole mycobacteria. The model was adapted to 1,536-well plate format and successfully used to screen over 600,000 compounds. Approximately 800 compounds were confirmed to reduce intracellular ATP levels in a dose-dependent manner in Mycobacterium bovis BCG. One hundred and forty non-cytotoxic compounds with activity against hypoxic non-replicating M. tuberculosis were further validated. The resulting collection of compounds that disrupt ATP homeostasis in M. tuberculosis represents a valuable resource to decipher the biology of persistent mycobacteria.

Cellular Ser/Thr-kinase assays using generic peptide substrates.

High-throughput cellular profiling has successfully stimulated early drug discovery pipelines by facilitating targeted as well as opportunistic lead finding, hit annotation and SAR analysis. While automation-friendly universal assay formats exist to address most established drug target classes like GPCRs, NHRs, ion channels or Tyr-kinases, no such cellular assay technology is currently enabling an equally broad and rapid interrogation of the Ser/Thr-kinase space. Here we present the foundation of an emerging cellular Ser/Thr-kinase platform that involves a) coexpression of targeted kinases with promiscuous peptide substrates and b) quantification of intracellular substrate phosphorylation by homogeneous TR-FRET. Proof-of-concept data is provided for cellular AKT, B-RAF and CamK2delta assays. Importantly, comparable activity profiles were found for well characterized B-Raf inhibitors in TR-FRET assays relying on either promiscuous peptide substrates or a MEK1(WT) protein substrate respectively. Moreover, IC(50)-values correlated strongly between cellular TR-FRET assays and a gold standard Ba/F3 proliferation assay for B-Raf activity. Finally, we expanded our initial assay panel by screening a kinase-focused cDNA library and identified starting points for >20 cellular Ser/Thr-kinase assays.

1H-filtered correlation experiments for assignment and determination of coupling constants in backbone labelled proteins.

The implementation of [13Calpha,13C',15N,2Halpha] labelled amino acids into proteins allows the acquisition of high resolution triple resonance experiments. We present for the first time resonance assignments facilitated by this new labelling strategy. The absence of 1JCalpha,Cbeta couplings enables us to measure 1JCalpha,C' scalar and 1DCalpha,C' residual dipolar coupling constants using modified HNCA experiments which do not suffer from sensitivity losses characteristic for 13C constant time experiments.