Dramatic improvements in cancer therapies have been hampered the heterogeneous nature of cancer. Molecular therapeutics targets abnormalities in signaling pathways critical in tumor development and progression. In this regard, it is important to define qualified cellular targets for cancer diagnoses and prognoses, as well as accurately predict and monitor responses to therapies. The reverse phase protein microarray (RPPA) is a recently developed antibody-based quantitative assay that analyzes nanoliter amounts of sample for potentially hundreds of proteins at their expression levels and protein modifications, such as phosphorylation or cleavage. RPPAs allow concordant interrogation of multiple signaling molecules at their functional status. We utilized the RPPAs to profile signaling networks in human breast cancer cell lines and breast tumor tissue. Here, we report that results from RPPAs demonstrate remarkable concordance with Immunoblotting. However, the RPPA approach is preferable in mapping signaling pathways due to specificity, quantification, amount of protein required, costs and labor intensity. We have validated over 100 different monospecific antibodies as useful for this approach. The assay is sensitive to analyze small changes in signaling events. We have demonstrated that linear signals can be obtained from < 15 cells suggesting that sufficient protein is available from needle biopsy or microdissection to allow screening with multiple antibodies. The approach is reproducible with coefficients of determination (r2) of 0.99 for spotting reliability and r2 of 0.76-0.83 for replicated assays of the same samples. The CV across experiments and arrays is 7-20%. The assay is unique in its quantitative nature, readily detected phospho-AKT recombinant protein at 36 femtograms and phospho-AKT peptide at 3.6 femtograms. The use of recombinant protein/phosphopeptides enables absolute quantification of intracellular signaling levels. We have validated signaling events by RPPA in a series of breast cancer cell lines with defined genetic lesions as well as in 80 breast cancer tissues. We classified tumors by RPPAs in 3 subtypes with different sets of genomic and proteomic aberrations. We demonstrated high levels of PI3K pathway activity in ER/HER2- breast cancer cell lines. The PI3K activation is associated with overexpression of EGFR, loss of PTEN, and activating mutations of PIK3CA, suggesting that the PI3K pathway is a rational target in ER/HER2- breast cancers. We characterized the applicability of RPPA to analyze cellular responses to PI3K pathway inhibitors in breast cancer cell lines. Hierarchical clustering exhibits characteristic “fingerprints” of cellular responses and effects of inhibitors. Thus, linking robust pathway mapping to molecular therapeutics should provide an expanding repertoire of validated biomarkers and targeted therapeutics for clinical evaluation.

[Proc Amer Assoc Cancer Res, Volume 47, 2006]