Abstract 1221: Triple-negative breast cancer (TNBC) phosphoproteomics

Abstract

Abstract Background: We hypothesized that the biphasic relapse pattern of TNBC could be explained by a limited number of activation patterns of signaling nodes. In addition, we sought to determine whether the hyperactive signaling nodes, distinguishing the cases with favorable vs adverse outcome, could be potential targets. Methods: Training set of 34 frozen tumor samples divided in two sets, (A) 13 patients, relapsed in <4 years; (B) 21 patients relapse-free >12 years, (mimicking the percentage and relapse patterns of unselected TNBC, but paired for T, N, G and Ki67). TNBC cell lines: 7 indolent (no metastases in 60 weeks) and 3 aggressive, develop metastases and kill recipient mice in <4 months. Shotgun phosphoproteomics and TiO2-IMAC phosphopeptide enrichment coupled with mass spectrometry runs in a Orbitrap Elite Mass Spectrometer was performed. Spectra were processed with MaxQuant software. Differentially expressed phosphopeptides were obtained by applying linear models R limma package. Differential kinase activation driving the profiles segregating cured vs. relapsing cases was done using linear sequence motif analysis. The hyperactivated kinases were validated in an independent set of 113 consecutive TNBC cases with 12+ years of follow-up spotted in TMAs by using an in-house algorithm for immunohistochemistry coupled with computer-aided quantitation using an Ariol scanning, we took the kinases in the upper quartile (high activity). Survival analysis was performed with KM curves and log rank test; and Cox proportionate hazards model was used for multivariate models. Results: 11405 phosphopeptides were identified and quantified in the training set. Supervised clustering of relapsed vs. cured cases showed that 161 and 541 peptides were significantly up-regulated in the A and B groups, respectively (FDR<0.15). After kinase-to-kinase co-linearity was ruled-out , gathering the high activity (upper quartile) of six kinases (a combined variable, herein Var1) showed statistically significant association with relapse, being these: PRKCE, pERK, c-KIT, CDK6, pP70S6K and pPNKP. Cox proportional hazards model of any of the six probes high (var1) vs rest: 9.9 vs. not reached years (P<0.001). Patients that had any of the 6 kinases high have 47% of chance to relapse (only 2 out of 42 relapsed patients have 0/6 active kinases) vs patients with Var1 negative, 7% of chance (29 patients out of 72 have 0/6 active kinases) we also observed constants patterns of activations in the different sets expressions of kinases. We considered the kinases at Var1 as a potentially targets and we developed a pharmacological in vitro assay, testing pairs inhibitors on 10 TNBC cell lines; 99.3% of the combinations were supra-additive. Conclusion: High throughput p-proteomics allows a parsimonious segregation of early TNBC cases, easily detecting the cases with long-term cure vs the remaining while identifying potential therapeutic solutions for the patients falling in the adverse prognostic subgroups. Citation Format: Sara Fernandez Gaitero, Ivana Zagorac, Jose Francisco Lopez-Acosta, Gonzalo Gomez-Lopez, David González Pisano, Javier Muñoz Peralta, Luis Manso, Soledad Alonso, Renske Penning, Maarten Altelaar, Albert JR Heck, Miguel Quintela-Fandino. Triple-negative breast cancer (TNBC) phosphoproteomics [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1221. doi:10.1158/1538-7445.AM2017-1221