Development of Automated Microscopy‐Assisted High‐Content Multiparametric Assays for Cell Cycle Staging and Foci Quantitation

Abstract

The investigation of cell cycle stage-dependent processes in a population of cells is often performed using flow cytometry. While this approach is high-throughput, it is relatively low in resolution and unable to measure phenotypic changes or processes occurring in subcellular compartments. We integrated automated microscopy with newly developed informatics workflow that enabled the quantitation of multiple fluorescent markers from specific subnuclear regions throughout a population of cells. Telomeres protect chromosome termini and prevent cellular aging. Cancer cells lengthen telomeres by synthesizing new TTAGGG repeats by the enzyme telomerase, while others activate recombination-dependent alternative lengthening of telomeres (ALT). A key feature of the ALT pathway is the specific clustering of promyelocytic leukemia (PML) nuclear bodies at telomeres. These ALT-associated PML bodies (APBs) common in tumors of mesenchymal origin have gained in diagnostic use in the past decade. Here we applied recent improvements in automated microscopy and developed novel informatics workflows for quantitation of multiple fluorescent markers from specific subnuclear regions at the single cell level. Key to this workflow are customized machine learning algorithms within HCS Studio™ Cell Analysis which automatically identify and segment cells into defined regions of interest based on fluorescent markers, measure marker intensities and compute marker colocalizations in specific segmented regions. These multiparametric cellular assays assess cell cycle dynamics as well as the interactome of APBs, are amenable to adherent cells and histological sections, and are adaptable for use with additional markers. In the future we anticipate exploiting these algorithms for a wide range of research questions related to telomere biology with potential to facilitate clinical development of ALT detection assays to benefit patients with these often-poor prognosis tumors.

Genome-Wide CRISPR Screen Identifies RACK1 as a Critical Host Factor for Flavivirus Replication

“…Pearson’s correlation coefficient (PCC) calculates the relationship between intensities in two images by linear regression and the slope of the fitted line provides the rate of association of two fluorophores. Its value can range from +1 to -1, with 1 standing for complete positive correlation and -1 for negative correlation, with zero indicating no correlation ( 52 , 53 ). To perform immunoprecipitation, cell lysates were harvested from cells using a lysis buffer containing 1% Triton X-100, 50 mM Tris pH 8, 150 mM NaCl and 1x Protease Inhibitor cocktail (Sigma) and centrifuged at 21,000 × g to remove cell debris.…”

Real‐time visualization and quantitation of cell death and cell cycle progression in 2D and 3D cultures utilizing genetically encoded probes

“…Many in vitro 3D models were described to understand the basic biology of the tumorigenic process and drug discovery applications. [1][2][3][4][5] Current methods of 3D models include tumor explant models, scaffold-based 3D models, microfluidicsbased models, multicellular tumor spheroids with or without matrices. 3 Despite the remarkable success in growing cancer cells as spheres and organoids, drug discovery using these models often require extensive use of chemical probes and preprocessing for drug efficacy testing that compromises their utility.…”

“…In recent years, cancer drug discovery has witnessed a dramatic transition from monolayer‐based models to 3D models and multicellular spheroids after the realization that monolayers fail to recapitulate the in vivo physiology of a growing tumor and drug response. Many in vitro 3D models were described to understand the basic biology of the tumorigenic process and drug discovery applications 1–5 . Current methods of 3D models include tumor explant models, scaffold‐based 3D models, microfluidics‐based models, multicellular tumor spheroids with or without matrices 3 .…”

Advances in Quantitative Bioimage Analysis for Proteins and Cells

“…The quantification of cell properties is usually carried out by flow cytometry. The original article by Frölich et al addresses the issue of flow cytometry being relatively low in resolution and unable to measure phenotypic changes or processes occurring in subcellular compartments. The authors integrate automated microscopy with a bioimage analysis pipeline to quantify multiple fluorescent markers from specific subnuclear regions throughout a population of cells.…”