Share Email Print

Proceedings Paper

DeepDyeDrop: an image-based approach to quantify the phenotypic response of cancer cells to therapeutics (Conference Presentation)
Author(s): Marc Hafner; Caitlin E. Mills; Luca Gerosa; Mirra Chung; Mario Niepel; Peter K. Sorger

Paper Abstract

Accurate measurement of the response of cancer cells to therapeutic agents is confounded by many variables. Lysate-based assays such as CellTiter-Glo® (CTG) are commonly used but rely on a surrogate of cell number (ATP in the case of CTG) and provide limited information on the phenotypic response to treatment. To characterize drug response at a single-cell level, we developed the DeepDyeDrop assay, a high throughput minimally disruptive microscopy-based assay. In brief, the assay involves plating cells in 384-well plates, delivering treatments of interest, and staining and fixing the treated plates at an appropriate time point. Nuclei are stained with Hoechst, dead cells with a membrane impermeable LIVE/DEAD (LDR) dye, cells progressing through S phase of the cell cycle with EdU, and cells in mitosis with a conjugated phospho-histone H3 (pH3) antibody. Images are acquired with a Perkin Elmer Operetta microscope, nuclei are segmented based on the Hoechst signal, and the intensity of all signals within the nuclear area are measured. After segmentation, we first identify dead cells based on the Hoechst and LDR signals. The remaining live cells are then assigned to the phases of the cell cycle based on their DNA content, and EdU and pH3 intensities. To quantify the differences among treatments, we relied on normalized growth rate inhibition (GR) metrics, which allow for the comparison of cell lines with variable division rates. We added new theoretical framework to the standard GR metrics to account for differential responses within a treated population of cells to independently score the contributions of the cytostatic and cytotoxic components to the overall response. By conducting the assay on a panel of breast cancer cell lines treated with a library of kinase inhibitors, we observed a wide range of responses that reflect drug mechanisms of action.

Paper Details

Date Published: 14 March 2018
Proc. SPIE 10475, Visualizing and Quantifying Drug Distribution in Tissue II, 104750E (14 March 2018); doi: 10.1117/12.2293204
Show Author Affiliations
Marc Hafner, Harvard Medical School (United States)
Caitlin E. Mills, Harvard Medical School (United States)
Luca Gerosa, Harvard Medical School (United States)
Mirra Chung, Harvard Medical School (United States)
Mario Niepel, Harvard Medical School (United States)
Peter K. Sorger, Harvard Medical School (United States)

Published in SPIE Proceedings Vol. 10475:
Visualizing and Quantifying Drug Distribution in Tissue II
Kin Foong Chan; Conor L. Evans, Editor(s)

© SPIE. Terms of Use
Back to Top