We asked the question: what are the changes in tyrosine phosphorylation of proteins upon EGF stimulation and tyrosine kinase inhibitor (TKI) inhibition in human lung adenocarcinoma cell lines expressing the mutant EGFRs? We used a 1st generation EGFR TKI, erlotinib, a 2nd generation EGFR TKI, afatinib, an irreversible EGFR and ERBB2 inhibitor, and two 3rd generation EGFR TKIs, osimertinib and rociletinib which are irreversible inhibitors. Various large-scale experiments were performed using stable isotope labeling with amino acids in cell culture (SILAC) and mass spectrometry. Lung adenocarcinoma cell lines used in these experiments for the 1st and 2nd generation EGFR TKIs were H3255 and 11-18 (L858R mutation), H1975 (L858R/T790M mutation), PC9 (E746-A750 Del EGFR). This part of the project is complete. We have validated the changes in phosphorylation of a subset of these proteins by immunoprecipitation, western blot experiments and tried to understand the significance of phosphorylation of these proteins. We also performed siRNA-mediated knockdown of proteins identified as phosphorylation targets of mutant EGFRs in lung adenocarcinoma cells harboring mutant EGFRs or mutant KRAS. It is interesting to note that several of the proteins whose tyrosine phosphorylation was inhibited by TKIs in mutant EGFR-expressing cells were also required for survival of EGFR mutant expressing cells but not KRAS mutant expressing cells. We followed two such targets: DAPP1 and SCAMP3. We have showed that both DAPP1 and SCAMP3 interact with mutant EGFRs. We have made several phosphorylation site specific mutants of these proteins and currently conducting biological assays to study the significance of altered phosphorylation of these proteins and the overall role played by these proteins in mutant EGFR-driven lung tumorigenesis. DAPPI1 (dual adapter for phosphotyrosine or 3-phosphoinositides) has not been implicated in EGFR signaling. However tyrosine phosphorylation at Y139 of DAPPI1 is stimulated upon EGF stimulation and inhibited 5-10 fold upon treatment with TKIs, erlotinib and afatinib, suggesting DAPPI1 may be an integral member of the EGFR signaling pathway. We have shown by biochemical experiments that DAPP1 interacts with EGFR. We have also shown that Y139 is a major phosphorylation site of DAPP1. We have planned experiments to assay whether EGFR is the kinase phosphorylating DAPP1, since there is some evidence that SRC may be the kinase involved. SCAMP3 (secretory career membrane protein) has been shown to interact with EGFR. We have demonstrated that mutant EGFRs phosphorylate specific sites in SCAMP3 more than WT EGFR. SCAMP3 is involved in receptor recycling. We are currently studying the role of SCAMP3 in mutant EGFR trafficking. We have also demonstrated that SCAMP3 is a tumor suppressor in EGFR mutant lung adenocarcinoma cell lines.A manuscript on the role of SCAMP3 and the phosphorylation targets downstream of mutant EGFRs is being submitted. Identification of Ser/Thr phosphorylation sites on downstream targets of mutant EGFRs and quantitation of phosphorylation changes upon TKI inhibition resistant to EGFR-directed TKIs This part of the project is complete. We have completed a series of experiments to identify Ser/Thr phosphorylation sites by SILAC labeling of adenocarcinoma cells and mass spectrometry. We used various fractionation techniques after in-solution trypsin digestion, such as strong cation exchange (SCX) or basic reverse phase. Around 30 fractions were then subjected to TiO2 enrichment for phosphopeptide isolation followed by reverse phase liquid chromatography and tandem mass spectrometry using an orbitrap elite mass spectrometer. H3255 lung adenocarcinoma cells harboring the L858R mutation and H1975 cells harboring L858R/T790M mutations were used in a triple-SILAC experiment. A total of around 8000 phosphosites were identified from H3255 cells and around 6000 sites identified in H1975 cells. We demonstrateddifferent patterns of dynamic phosphorylation changes upon EGF stimulation and TKI inhibition in these cells. We have analyzed this data set using several bioinformatic tools, including IPA and Ariadne pathway studio. Various canonical pathways such as p70S6, IRS, ERK/MAPK, mTOR, PKA, JAK/STAT were enriched in our dataset. We have collaborated with Dr. Andrea Califano in Columbia University to use their unique bioinformatics tools to interrogate the data to generate new hypothesis that can then be validated experimentally. Two manuscripts have been published in the journals Proteomics (2015) and Molecular and Cellular Proteomics (2017). We are currently performing quantitative mass spectrometry experiments to identify and quantify the proteome and phosphoproteome of lung adenocarcinoma cells sensitive and resistant to the 3rd generation EGFR TKIs, osimertinib and rociletinib. We have generated several human lung adenocarcinoma cell lines resistant to these TKIs by step-wise TKI treatment. We have finished the mass spectrometry analyses of the osimertinib and rociletinib-sensitive and resistant cells. We have performed SILAC mass spectrometry to quantitate the global proteome and phosphoproteome. Currently we are doing the same for osimertinib resistant cells. In parallel we have also generated in vivo resistance to osimertinib in mutant EGFR genetically engineered mouse (GEM) models. Osimertinib-resistant mouse tumors will be compared with sensitive tumors to identify mechanisms of osimertinib resistance. We have undertaken an integrated proteo-genomics approach to identifying the mechanisms of resistance to the 3rd generation EGFR TKIs, osimertinib and rociletinib. We have completed whole exome sequencing of cell lines and xenografts generated in our laboratory that are resistant to osimertinib. We are currently analyzing this data. In addition, we have completed global proteome quantitation and phosphoproteome quantitation of a subset of lung adenocarcinoma cell lines generated in our laboratory. We have identified specific kinases that are potentially more active in the osimertinib resistant cells compared to the sensitive cells- these include PI3K, AKT, mTOR, p38 MAPK, SGK1, among others. Further bioinformatic analysis of our mass spectrometry data identified dactolisib, a PI3K/mTOR inhibitor, and VX970, an ATR inhibitor. We validated the efficacy of these drugs in osimertinib and rociletinib resistant cells in vitro and in vivo in xenografts. One manuscript describing these findings with the 3rd generation EGFR TKis is being submitted.
