BackgroundBreast cancer tumors are known to be highly heterogeneous and differences in their metabolic phenotypes, especially at protein level, are less well-understood. Profiling of metabolism-related proteins harbors the potential to establish new patient stratification regimes and biomarkers promoting individualized therapy. In our study, we aimed to examine the relationship between metabolism-associated protein expression profiles and clinicopathological characteristics in a large cohort of breast cancer patients.MethodsBreast cancer specimens from 801 consecutive patients, diagnosed between 2009 and 2011, were investigated using reverse phase protein arrays (RPPA). Patients were treated in accordance with national guidelines in five certified German breast centers. To obtain quantitative expression data, 37 antibodies detecting proteins relevant to cancer metabolism, were applied. Hierarchical cluster analysis and individual target characterization were performed. Clustering results and individual protein expression patterns were associated with clinical data. The Kaplan-Meier method was used to estimate survival functions. Univariate and multivariate Cox regression models were applied to assess the impact of protein expression and other clinicopathological features on survival.ResultsWe identified three metabolic clusters of breast cancer, which do not reflect the receptor-defined subtypes, but are significantly correlated with overall survival (OS, p ≤ 0.03) and recurrence-free survival (RFS, p ≤ 0.01). Furthermore, univariate and multivariate analysis of individual protein expression profiles demonstrated the central role of serine hydroxymethyltransferase 2 (SHMT2) and amino acid transporter ASCT2 (SLC1A5) as independent prognostic factors in breast cancer patients. High SHMT2 protein expression was significantly correlated with poor OS (hazard ratio (HR) = 1.53, 95% confidence interval (CI) = 1.10–2.12, p ≤ 0.01) and RFS (HR = 1.54, 95% CI = 1.16–2.04, p ≤ 0.01). High protein expression of ASCT2 was significantly correlated with poor RFS (HR = 1.31, 95% CI = 1.01–1.71, p ≤ 0.05).ConclusionsOur data confirm the heterogeneity of breast tumors at a functional proteomic level and dissects the relationship between metabolism-related proteins, pathological features and patient survival. These observations highlight the importance of SHMT2 and ASCT2 as valuable individual prognostic markers and potential targets for personalized breast cancer therapy.Trial registrationClinicalTrials.gov, NCT01592825. Registered on 3 May 2012.Electronic supplementary materialThe online version of this article (doi:10.1186/s13058-017-0905-7) contains supplementary material, which is available to authorized users.

SummaryThe transcription factor Meis1 drives myeloid leukemogenesis in the context of Hox gene overexpression but is currently considered undruggable. We therefore investigated whether myeloid progenitor cells transformed by Hoxa9 and Meis1 become addicted to targetable signaling pathways. A comprehensive (phospho)proteomic analysis revealed that Meis1 increased Syk protein expression and activity. Syk upregulation occurs through a Meis1-dependent feedback loop. By dissecting this loop, we show that Syk is a direct target of miR-146a, whose expression is indirectly regulated by Meis1 through the transcription factor PU.1. In the context of Hoxa9 overexpression, Syk signaling induces Meis1, recapitulating several leukemogenic features of Hoxa9/Meis1-driven leukemia. Finally, Syk inhibition disrupts the identified regulatory loop, prolonging survival of mice with Hoxa9/Meis1-driven leukemia.

To evaluate whether tumour-derived microvesicles (T-MV), originating from the plasma membrane, represent suitable cancer biomarkers, we isolated MV from peripheral blood samples of cancer patients with locally advanced and/or metastatic solid tumours (n = 330, including 79 head & neck cancers, 74 lung cancers, 41 breast cancers, 28 colorectal cancers and 108 with other cancer forms) and controls (n = 103). Whole MV preparations were characterised using flow cytometry. While MV carrying the tumour-associated proteins MUC1, EGFR and EpCAM were found to be enhanced in a tumour-subtype-specific way in patients’ blood, expression of the matrix metalloproteinase inducer EMMPRIN was increased independent of tumour type. Higher levels of EMMPRIN+-MV correlated significantly with poor overall survival, whereas the other markers were prognostic only in specific tumour subgroups. By combining all four tumour-associated antigens, cancer patients were separated from healthy controls with an AUC of up to 0.85. Ex vivo, whole MV preparations from cancer patients, in contrast to those of controls, induced a tumour-supporting phenotype in macrophages and increased tumour cell invasion, which was dependent on the highly glycosylated isoform of EMMPRIN. In conclusion, the detection of T-MV in whole blood, even in minor amounts, is feasible with standard techniques, proves functionally relevant and correlates with clinical outcome.

Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Mutations and activation of the PI3K signaling pathway in breast cancer cells have been linked to brain metastases. However, here we describe that in some breast cancer brain metastases samples the protein expression of PI3K signaling components is restricted to the metastatic microenvironment. In contrast to the therapeutic effects of PI3K inhibition on the breast cancer cells, the reaction of the brain microenvironment is less understood. Therefore we aimed to quantify the PI3K pathway activity in breast cancer brain metastasis and investigate the effects of PI3K inhibition on the central nervous system (CNS) microenvironment. First, to systematically quantify the PI3K pathway activity in breast cancer brain metastases, we performed a prospective biomarker study using a reverse phase protein array (RPPA). The majority, namely 30 out of 48 (62.5%) brain metastatic tissues examined, revealed high PI3K signaling activity that was associated with a median overall survival (OS) of 9.41 months, while that of patients, whose brain metastases showed only moderate or low PI3K activity, amounted to only 1.93 and 6.71 months, respectively. Second, we identified PI3K as a master regulator of metastasis‐promoting macrophages/microglia during CNS colonization; and treatment with buparlisib (BKM120), a pan‐PI3K Class I inhibitor with a good blood‐brain‐barrier penetrance, reduced their metastasis‐promoting features. In conclusion, PI3K signaling is active in the majority of breast cancer brain metastases. Since PI3K inhibition does not only affect the metastatic cells but also re‐educates the metastasis‐promoting macrophages/microglia, PI3K inhibition may hold considerable promise in the treatment of brain metastasis and the respective microenvironment.

astrid.wachter@med.uni-goettingen.de.

Analysis of large-scale proteomic data sets requires specialized software tools, tailored toward the requirements of individual approaches. Here we introduce an extension of an open-source software solution for analyzing reverse phase protein array (RPPA) data. The R package RPPanalyzer was designed for data preprocessing followed by basic statistical analyses and proteomic data visualization. In this update, we merged relevant data preprocessing steps into a single user-friendly function and included a new method for background noise correction as well as new methods for noise estimation and averaging of replicates to transform data in such a way that they can be used as input for a new time course plotting function. We demonstrate the robustness of our enhanced RPPanalyzer platform by analyzing longitudinal RPPA data of MET receptor signaling upon stimulation with different hepatocyte growth factor concentrations.

No abstract