High-content, high-throughput screening for the identification of cytotoxic compounds based on cell morphology and cell proliferation markers

• Published in: PloS one 2014-02, Vol.9 (2), p.e88338
• Main Authors: Martin, Heather L, Adams, Matthew, Higgins, Julie, Bond, Jacquelyn, Morrison, Ewan E, Bell, Sandra M, Warriner, Stuart, Nelson, Adam, Tomlinson, Darren C
• Format: Article
• Language: English
• Subjects: Algorithms; Anthracyclines; Apoptosis; Assaying; Biocompatibility; Biological Products - adverse effects; Biological Products - toxicity; Biology; Cell cycle; Cell growth; Cell Line; Cell morphology; Cell proliferation; Cell Proliferation - drug effects; Cell Shape - drug effects; Cell size; Chemistry; Cytology; Cytotoxicity; Doxorubicin; Drug discovery; Drug Evaluation, Preclinical - methods; High-throughput screening; High-Throughput Screening Assays - methods; Humans; Identification; Image analysis; Image processing; Image processing equipment; In vivo methods and tests; Medical screening; Medicine; Molecular biology; Natural products; Neurosciences; Nocodazole; Paclitaxel; Screening; Small Molecule Libraries - adverse effects; Small Molecule Libraries - toxicity; Software; Taxol; Toxicity
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0088338

Toxicity is a major cause of failure in drug discovery and development, and whilst robust toxicological testing occurs, efficiency could be improved if compounds with cytotoxic characteristics were identified during primary compound screening. The use of high-content imaging in primary screening is becoming more widespread, and by utilising phenotypic approaches it should be possible to incorporate cytotoxicity counter-screens into primary screens. Here we present a novel phenotypic assay that can be used as a counter-screen to identify compounds with adverse cellular effects. This assay has been developed using U2OS cells, the PerkinElmer Operetta high-content/high-throughput imaging system and Columbus image analysis software. In Columbus, algorithms were devised to identify changes in nuclear morphology, cell shape and proliferation using DAPI, TOTO-3 and phosphohistone H3 staining, respectively. The algorithms were developed and tested on cells treated with doxorubicin, taxol and nocodazole. The assay was then used to screen a novel, chemical library, rich in natural product-like molecules of over 300 compounds, 13.6% of which were identified as having adverse cellular effects. This assay provides a relatively cheap and rapid approach for identifying compounds with adverse cellular effects during screening assays, potentially reducing compound rejection due to toxicity in subsequent in vitro and in vivo assays.