Abstract 3141: High-content live-cell imaging analysis of therapeutic monoclonal antibody binding and potency across 240 genomically defined cancer cell lines

• Published in: Cancer research (Chicago, Ill.) Ill.), 2013-04, Vol.73 (8_Supplement), p.3141-3141
• Main Authors: Crane, Jonathan M., Alison, Angione, Snead, Katie, Bernards, Karen, Nelson, Brian, Waikins, Kate, McKinley, Keith, Nguyen, Phuong, Warrior, Usha, Shah, O. Jameel
• Format: Article
• Language: English
• ISSN: 0008-5472; 1538-7445
• DOI: 10.1158/1538-7445.AM2013-3141

OncoPanelTM is a collection of 240 genomically and histologically diverse cancer cell lines that comprise an unparalleled platform for multiplexed high-content imaging analysis of potential therapeutic targets in oncology. High-content imaging is ideally suited to detect binding, localization, and trafficking of therapeutic antibodies, adding an important level of resolution when interpreting drug response data. Moreover, understanding relative antigen expression is important both to define the patient population most likely to benefit from therapy as well as to define the antigen expression threshold needed to elicit biologic response. Here we describe the high-throughput parallelization of quantitative, comparative therapeutic antibody binding measurements and drug response profiling across the OncoPanel using high-content imaging. High-quality binding data for monoclonal antibodies was achieved by analysis of referenced, background-subtracted fluorescence intensities on a per cell basis. Binding studies were performed on living cells to best recapitulate natural antigen-antibody interactions, and results were quantified and expressed as potency (EC50 values), total molecules bound per cell, and spatially analyzed to yield a reliable estimate of antigen density. Observed binding intensities were internally referenced against calibrated fluorescent antibody-conjugated microspheres to yield a direct correlation between fluorescence intensity and number of antibodies bound. Cellular areas were independently measured by a non-specific fluorescent counterstain of the plasma membrane, and cell count determined by nuclear staining with DAPI. The therapeutic antibodies Vectibix, Erbitux, and Herceptin were profiled against the OncoPanel, providing a rank-ordering of cell surface expression of EGFR and ErbB2/Her2 across 240 cancer cell lines, and defining thresholds of antibody binding needed to inhibit cell proliferation. Compared to mRNA expression by microarray, quantitative antibody binding was a superior predictor of biologic response. Moreover, we demonstrate that genomic analysis of the response to therapeutic antibodies is significantly improved when only antigen-positive cells are considered. Using the power of binding and genomics data gained from the screening of OncoPanel, targeted studies may be optimized to determine the most relevant cancer subtypes and conditions under which the activity of any monoclonal antibody is likely to be most effective. C