Abstract:
Melanoma is the most aggressive form of skin cancer, with a rapidly increasing incidence rate. Malignant melanoma is characterized by mutations in the mitogen-activated kinase (MAPK) pathway, which strongly correlate with poor prognosis of the disease. The kinase BRAF is mutated in ~48% of melanoma cases, resulting predominantly in V600E substitution that leads to constitutive activation of the BRAF kinase and downstream signaling pathways. Over the last decade, several therapeutic treatments for melanoma have been developed with improved efficiency and overall survival rates. Targeted inhibition of the mutated BRAF with selective inhibitors, such as vemurafenib or dabrafenib and immunotherapy with the immune checkpoint antibodies targeting PD-1 and CTLA-4 receptors, results in regression of the disease. However, only a minority of patients can benefit from the current therapies and most of them quickly develop resistance to the treatment. Prognostic biomarkers, resistance mechanisms and mutational profiles of melanoma are mainly studied by genomics and transcriptomics. Although only about 2% of the genome codes for proteins, variants in these region of the genome have a high potential to rewire signal transduction networks. In addition, the majority of targeted cancer therapies do not target the genome, but rather the protein itself. Thus, it is highly important to analyze proteins and their patient-specific alterations in context of personalized medicine. Mass spectrometry-based proteomics can be used to study protein-specific clinical questions and can identify molecular mechanisms of treatment-resistant melanoma. By combining personalized genomics and proteomics, in an approach called proteogenomics, it is possible to derive patient-specific protein sequence databases – that include patient-specific amino acid variants. These in turn can provide deeper and more comprehensive molecular characterization of cellular processes that underlie disease progression. Several mechanisms for acquired resistance and even cross-resistance in melanoma have been detected, but key (phospho)proteins involved in resistance, as well as mutations altering protein modification status are still largely elusive. This thesis develops and applies personalized proteogenomics workflows to study these mechanisms on the level of individual melanoma cells and patient tissues. In the first part of this thesis, a SILAC-based quantitative (phospho)proteomics profiling of vemurafenib-resistant and -sensitive A375 melanoma cells was performed to gain new insights into molecular processes that govern resistance to BRAF inhibitors. Among down-regulated proteins in vemurafenib-resistant cell lines were multiple cytoskeletal proteins including the intermediate filament nestin. Previous studies showed that nestin is expressed in various types of solid tumors and its abundance correlates with malignant phenotype of transformed cells. However, the role of nestin in cancer cells with regard to acquired resistance is still poorly understood. CRISPR/Cas9 knockout of the nestin gene showed that the loss of nestin leads to increased cellular proliferation and colony formation upon treatment with kinase inhibitors. Moreover, nestin depletion is associated with an invasive phenotype and acquired resistance to MEK and BRAF inhibitors. Finally, phosphoproteome analysis revealed that nestin depletion affects integrin and PI3K/AKT/mTOR pathway signaling similar to resistant cells. In this part, proteomic and phosphoproteomic changes have been determined for BRAF inhibitor resistant and sensitive cells. In the second part of the thesis, an individualized proteogenomics approach was applied to two melanoma cell lines, A375 and SkMel28, to analyze non-synonymous mutations and their impact on signal transduction networks in context of acquired resistance to kinase inhibitors. Integration of genomics and proteomics highlighted the distinct mutational landscape of both cell lines and revealed that cancer mutations are accumulating in MAPK and ErbB signaling pathways in resistant cells. Several alternate peptides interfering with the modification status of proteins with a potential to rewire signal transduction pathways were confirmed by high resolution mass spectrometry. Among them was transcription factor RUNX1, previously connected with myeloid leukemia and breast cancer. Validation of a loss of a known phosphorylation site on RUNX1 using SILAC-based protein interaction studies suggested that this mutation has an impact on the interactome of the protein and may alter its transcriptional activity. Taken together, this part of the thesis established the individualized proteogenomic workflow for analysis of mutational profiles of cancer cell lines and tissues. In the third part of the thesis, this individualized proteogenomics approach was applied to four clinical melanoma samples in response to immunotherapy. Integration of the matching genomics and (phospho)proteomics datasets revealed an extensive number of patient-specific variants and disproportional number of shared variants in immune checkpoint inhibitor (ICi)-treated patient samples compared to untreated (naïve) samples. The proteogenomic signatures of human tissues could be recapitulated in patient-derived xenografts, thus allowing phosphoproteomics analysis. MS-measurements confirmed mutation-driven modification changes of several proteins specific to one sample, most of them were previously not reported in melanoma. Statistical analysis revealed differing mechanisms and associated network-attacking mutations in response to immunotherapy, such as PI3K/AKT signaling or GTPase activation in ICi treated samples. The gain of a new phosphorylation site on the GEF protein DOCK1 was further investigated by interactome studies and the results showed that this mutation has an impact on the interactome of DOCK1. The obtained results have demonstrated that the developed individualized proteogenomic workflow can be efficiently applied to human melanoma tissue and patient-derived xenografts in response to immunotherapy. Taken together, this thesis presents a new personalized proteogenomics workflow that can be routinely applied to numerous types of cancer and other diseases involving patient-specific accumulation of mutations in protein-coding genes. Datasets reported in this thesis provide new insights into resistance mechanisms and associated mutations with the potential to rewire signal transduction networks in malignant melanoma. This work can therefore serve as a basis for further improvement of therapeutic treatment of cancer patients.
