In the last ten years, extensive progress was made in the area of precision medicine due to genome analysis routines. Nevertheless, it became evident that genetic analyses are not sufficient to understand the complexity of certain diseases, but also the translation of genetic information into proteins, the interaction of proteins, and the metabolism are determining factors. Understanding the interplay of these disease-relevant cellular components will support the future progress of precision medicine.
Mass spectrometry (MS) enables the analysis of a variety of biomolecules. This technique offers a deep insight in the molecular composition of patients under healthy and diseased conditions, delivering information on the patient’s, e.g. proteome, metabolome, glycome. This is why, the establishment of a MS platform became a strategic priority of four major Berlin institutions: Charité – Universitätsmedizin Berlin, Max Delbrück Center for Molecular Medicine (MDC), Berlin Institute of Health (BIH), and Humboldt University Berlin.
To achieve the goal of implementing mass spectrometric analyses in clinical routines and decisions, scientists, bioinformaticians, clinicians as well as industrial partners teamed up to establish a MS platform. Berlin has an ideal basis to reach the aim with excellent local scientists, recently established core facilities for proteomics (BIH and MDC) and metabolomics (BIH and MDC), focusing on systems medicine, and the integration of the Institute for Pathology as well as Labor Berlin.
The repertoire of mass spectrometric techniques comprises:
- Shotgun proteomics/phospho-proteomics
- Quantitative Interaction-proteomics
- Ultra high-throughput quantitative proteomics
- Targeted proteomics
- GC- and LC-/MS-based metabolomics
- Imaging MS
- Imaging Mass Cytometry
In addition to technical advancements of the MS technology, the MSTARS consortium aims at answering two key questions:
- Which biomarkers predict treatment response?
- Which mechanisms underlie the resistance towards targeted therapy?
With the chosen approach, MSTARS has from the beginning on a direct translational potential and tackles the two key questions. Initially, MS-based analyses are performed on retrospective sample material of head and neck squamous cell carcinoma (HNSCC) patients with the goal to identify marker for therapy resistance and molecular signatures. Typical characteristics of HNSCC are alterations in the EGFR, MAPK, or PI3K/mTOR signaling pathways that originate in a variety of mutations. Targeted inhibition of EGFR or PI3K/mTOR mediated signaling is highly efficient in some patients, but without effect in others. To differentiate these patient groups, MS derived data can be applied that subsequently guide therapeutic decisions. In the end, diagnostic tests with high clinical relevance are going to be initialized.
Two complementary approaches, subdivided in discovery, computational and validation phase, are followed to integrate MS-based technologies in the clinics. The mechanistic approach serves to acquire MS data to achieve a mechanistic understanding of disease processes. Clinical samples as well as experiments with preclinical models are performed to analyze the impact of perturbations (e.g. drug treatment, mutations) on diverse omics-level. For most of the samples genome and transcriptome data are available for integration and mechanistic modelling. These datasets give new insights and increase the knowledge on therapy success and treatment resistance mechanisms. The signature-based approach aims for the prediction of treatment results by the systematic measurement of high sample numbers in big patient cohorts. With the help of machine learning, diagnostic characteristics and signatures, that reflect perturbations, are derived from the datasets.
In the MSTARS project two principles of therapy resistance are addressed: On the one hand direct resistance, which is described by an inefficient targeting of a signaling pathway by a drug. On the other hand adaptive resistance, which describes an efficient targeting, that is compensated by the reaction of alternative signaling pathways that circumvent the drug’s effect. Proteins and metabolites mediate cellular functions and are therefore, closely connected to therapy responses and treatment resistance. This is why, we assume that the MS-analysis will result in the identification of two different biomarkers: 1) Diagnostic baseline-biomarkers that indicate therapy success and can be used for patient stratification, and 2) mechanistic biomarkers for the two principles of therapy resistance that can guide treatment options and drug development.
The availability of a large number of patient-derived models and well characterized clinical cohorts defined HNSCC as the central use case that connects all work packages. To demonstrate the usability of MS analyses in the clinics and to implement a sustainable pipeline, which is applicable for cancer and other disease entities, the disease entities are expanded progressively. In the second half of the first funding period samples before and after treatment of acute myeloid leukemia, ovarian cancer, neuroblastoma as well as myocarditis patients are measured by high-throughput analyses. During the second funding period, the pool is completed by samples of patients with prostate cancer or chronic liver disease.
The MSTARS project is divided in six work packages with the focus on proteomics, metabolomics, imaging MS, data management, data modelling and a clinical work package. Two young investigator’s groups complete the project structure.