br Cell Reports Resource br Breast Cancer Classification Bas
Cell Reports Resource
Breast Cancer Classification Based on Proteotypes Obtained by SWATH Mass Spectrometry
Pavel Bouchal,1,2,10,* Olga T. Schubert,3,4 Jakub Faktor,2 Lenka Capkova,1 Hana Imrichova,1,5 Karolina Zoufalova,1 Vendula Paralova,1 Roman Hrstka,2 Yansheng Liu,6 Holger Alexander Ebhardt,3,7 Eva Budinska,2,8 Rudolf Nenutil,2 and Ruedi Aebersold3,9,* 1Department of Biochemistry, Faculty of Science, Masaryk University, Brno, Czech Republic 2Regional Centre for Applied Molecular Oncology, Masaryk Memorial Cancer Institute, Brno, Czech Republic 3Department of Biology, Institute of Molecular Systems Biology, Federal Institute of Technology (ETH) Zurich, Zurich, Switzerland 4Department of Human Genetics, University of California, Los Angeles, Los Angeles, CA, USA 5Center for Human Genetics, University of Leuven, Leuven, Belgium 6Department of Pharmacology, Yale Cancer Biology Institute, Yale University School of Medicine, West Haven, CT, USA 7Systems Biology Ireland, University College Dublin, Dublin, Ireland 8Research Centre for Toxic Compounds in the Environment, Faculty of Science, Masaryk University, Brno, Czech Republic 9Faculty of Science, University of Zurich, Zurich, Switzerland 10Lead Contact *Correspondence: [email protected] (P.B.), [email protected] (R.A.)
Accurate classification of breast tumors is vital for patient management decisions and enables more precise cancer treatment. Here, we present a quanti-tative proteotyping approach based on sequential windowed acquisition of all theoretical fragment ion spectra (SWATH) mass spectrometry and establish key proteins for breast tumor classification. The study is based on 96 tissue samples representing five conventional breast cancer subtypes. SWATH proteotype patterns largely recapitulate these sub-types; however, they also reveal varying heterogene-ity within the conventional subtypes, with triple nega-tive tumors being the most heterogeneous. Proteins that contribute most strongly to the proteotype-based classification include INPP4B, CDK1, and ERBB2 and are associated with Emetine receptor (ER) status, tumor grade status, and HER2 status. Although these three key proteins exhibit high levels of correlation with transcript levels (R > 0.67), general correlation did not exceed R = 0.29, indicating the value of protein-level measurements of disease-regulated genes. Overall, this study highlights how cancer tissue proteotyping can lead to more accu-rate patient stratification.
Despite the progress achieved in early cancer diagnosis and therapy, many patients develop fatal disease. This also applies to breast cancer, even though it is one of the best characterized malignant diseases. Breast cancer is currently classified into five intrinsic subtypes, typically using immunohistological markers
(estrogen receptor [ER], progesterone receptor [PR], HER2 gene, and/or ERBB2 protein status), tumor grade, and/or prolif-eration. We will refer to these subtypes as ‘‘conventional sub-types’’; they have been defined as follows: luminal A (ER+, HER2 , low proliferation), luminal B HER2 (ER+, HER2 , high proliferation), luminal B HER2+ (ER+, HER2+, high proliferation), HER2 enriched (ER , HER2+, high proliferation), and triple nega-tive (ER , PR , HER2 , high proliferation; Brouckaert et al., 2013; Lam et al., 2014; Parise and Caggiano, 2014). This classi-fication guides decisions for the adjuvant therapy, which, howev-er, fails in a substantial proportion of cases due to cancer recurrence, therapy resistance, and/or metastasis (Parise and Caggiano, 2014). The development of advanced, generalized disease despite the therapy guided by the tumor classification into the subtypes described above indicates that the current classification scheme may not fully capture the genetic and molecular status of the cancer and that a refined classification system might better predict which patient groups respond best to the range of available therapies.
Nowadays, the search for better tumor classifiers significantly concentrates on the application of omics approaches, which are able to analyze thousands of gene sequences, gene transcripts, or proteins in a single experiment. The biochemical effector molecules in cells are proteins, and their direct measurement is, therefore, in principle preferable over the inference of protein quantities from transcript measurements (expression arrays and RNA sequencing). However, the commonly used proteomic approaches based on mass spectrometry analysis in data-dependent acquisition (DDA) mode are often hampered by limited consistency and quantitative accuracy and are therefore less suitable for application to clinical cohorts of significant size. In contrast, targeted proteomic technologies overcome some of these limitations and provide improved quantification precision and reproducibility (Pernika´rova´ and Bouchal, 2015). Kennedy and colleagues (Kennedy et al., 2014) recently demonstrated the ability of the targeted proteomic technique selected or multiple reaction monitoring (S/MRM) to quantify