Dr. Mertins was the Staff scientist and group leader at Broad Institute of MIT and Harvard in 2009-2017. He has been the Head of Proteomics Platform at Max Delbrück Center (MDC) and Berlin Institute of Health (BIH) since 2017.
Proteogenomic identification of drug response markers in patient-derived colon cancer xenografts
Merve Keziban Aktas1,2, Tamara Kanashova1,2, Hans Werner1,2, Shaunt Fereshetian3, Karl Clauser3, DR Mani3, Michael Gillette3, Karsten Krug3, Marlen Keil4, Mathias Gehrmann5, Johannes Haybaeck6, Ulrich Keilholz7, Jens Hoffmann4, Wolfgang Walther4, Ulrike Stein1, Marie-Laure Yaspo8, Hans Lehrach8, David Henderson5, Steven Carr3, Philipp Mertins1,2
1 Max Delbrück Center for Molecular Medicine, Berlin, Germany
2 Berlin Institute of Health, Berlin, Germany
3 Broad Institute of MIT and Harvard, Cambridge, USA
4 EPO-Berlin-Buch GmbH, Berlin, Germany
5 Bayer AG, Germany
6 Universitätsklinikum, Magdeburg, Germany
7 Charité Comprehensive Cancer Center, Berlin, Germany
8 Max Planck Institute for Molecular Genetics, Berlin, Germany
Cancer is caused by genetic alterations, however knowledge about the mutation status of tumors alone often does not help to predict the outcome of therapeutic interventions. Proteogenomics is a new discipline that connects the information derived of genomic and proteomic analyses to directly map the consequences of genetic alterations to the functional proteome. Patient-derived xenograft (PDX) tumors are excellent models to study the determinants of drug response and resistance mechanisms at the level of the genome, transcriptome or proteome. We evaluated the additional value of proteome and phosphoproteome profiling vs. genomics alone in a set of >60 colorectal cancer PDX tumors that have been previously characterized for their response to 14 commonly used drugs, using whole exome sequencing and RNA sequencing (Schütte et al., Nat Commun. 2017). For the proteome/phosphoproteome profiling we used methods established at the Broad Institute as part of the Clinical Proteomic Tumor Analysis Consortium (CPTAC) that allow a comprehensive analysis of the protein status of most expressed genes (Mertins et al., Nat Protocols 2018). At both genomic and proteomic levels, the largest number of markers for drug response could be observed for anti-EGFR therapies, such as the antibody cetuximab or the kinase inhibitor afatinib. Cetuximab is only administered in patients with no mutations in KRAS, NRAS, and BRAF in the clinic. In PDX cohorts a transcriptomic classifier has been shown to be superior to predict cetuximab drug response compared to mutation calling alone, however only few EGFR pathway genes have been identified in these predictive signatures. Here we show that deepscale phosphoproteomic profiling, identified markers of cetuximab response within the EGFR pathway and extended our knowledge of drug response markers in colorectal cancer beyond that provided by genomics methods alone.
