Evaluation of multiple protein profiles from treated xenograft tumor models identifies a marker panel for FFPE tissue analysis with reverse phase protein arrays

Background: Molecular profiling of patient tumors today uses high throughput genome and transcriptome technologies, yet proteins and especially their functions, on which drugs can act, are poorly predicted. Proteomic approaches like reverse phase protein arrays (RPPA) have demonstrated in several studies that signaling pathway profiling with up to hundreds of samples and proteins in parallel can add valuable information (i) providing multiple marker patterns from small amounts of tissue, and (ii) providing biological answers: Are pathways active downstream of a driver mutation? Will drug treatment be effective? What are the underlying mechanisms? Can we identify predictive marker proteins? So far. a lot of RPPA data have been generated from fresh frozen (FF) tissues, however formalin fixation and paraffin embedding (FFPE) is the standard method for tissue preservation in the clinic. Here we performed RPPA to measure tumor profiles in treated FF and FFPE xenograft tumors to identify a marker panel suited for the analysis of clinical FFPE tissue samples. Material and Methods: Protein lysates were prepared from matched FF and FFPE tissues specimen of three different xenograft tumor models. Tumor material was taken from randomized mice, untreated (vehicle control) and treated with either anti-Her3 or bispecific anti-IGF1R/EGFR monoclonal antibodies. Straight after collection, the individual tumors from each study animal were divided in half--one half was immediately snap frozen and the remaining part fixed in formalin and paraffin embedded (n = 3 mice per treatment group). Protein lysates were analyzed via RPPA as one cohort in parallel applying direct, sensitive and quantitative fluorescence read-out immunoassays. Results: In total, protein extracts from matched FF and FFPE tumor tissues from 20 xenograft individual tumors (40 samples) were evaluated by RPPA using a panel of 300 markers. The markers covered a broad range of signaling pathways and had been pre-validated up-front in the FF tissue type. Overall a list of top 60 markers was identified that significantly correlated between the FF and FFPE tissue profiles (p < 0.01). No relationship between the cellular location (membrane, cytoplasm or nucleus) of the correlating markers was apparent, and both phosphorylated and non-phosphorylated proteins were represented. Conclusion: The data demonstrate that using appropriate assay reagents, RPPA analyses from FFPE tissue is well feasible and can provide biologi