A Multipronged Unbiased Strategy Guides the Development of an anti-EGFR/EPHA2 Bispecific Antibody for Combination Cancer Therapy

Purpose: Accumulating analyses of pro-oncogenic molecular mechanisms triggered a rapid development of targeted cancer therapies. Although many of these treatments produce impressive initial responses, eventual resistance onset is practically unavoidable. One of the main approaches for preventing this refractory condition relies on the implementation of combination therapies. This includes dual-specificity reagents that affect both of their targets with a high level of selectivity. Unfortunately, selection of target combinations for these treatments is often confounded by limitations in our understanding of tumor biology. Here, we describe and validate a multipronged unbiased strategy for predicting optimal co-targets for bispecific therapeutics. Experimental design: Our strategy integrates ex-vivo genome-wide loss of function screening, BioID interactome profiling and gene expression analysis of patient data to identify the best fit co-targets. Final validation of selected target combinations is done in tumorsphere cultures and xenograft models. Results: Integration of our experimental approaches unambiguously pointed towards EGFR and EPHA2 tyrosine kinase receptors as molecules of choice for co-targeting in multiple tumor types. Following this lead, we generated a human bispecific anti-EGFR/EPHA2 antibody that, as predicted, very effectively suppresses tumor growth, compared to its prototype anti-EGFR therapeutic antibody, cetuximab. Conclusion: Our work not only presents a new bispecific antibody with a high potential for being developed into clinically-relevant biologics, but more importantly, successfully validates a novel unbiased strategy for selecting biologically optimal target combinations. This is of a significant translational relevance, as such multifaceted unbiased approaches are likely to augment the development of effective combination therapies for cancer treatment. Samples were collected at different time points after treating cells with IgG or anti-EphA2 and probes were prepared to hybridize to microarray chips to identify anti-EphA2 resistant genes.