Individualised Proteogenomics Applied to Analysis of Amino Acid Variants in Malignant Melanoma

Analysis of patient-specific nucleotide variants is a cornerstone of personalised medicine. Although only 2% of the genomic sequence is protein-coding, mutations occurring in these regions have the potential to influence protein structure and may have severe impact on disease aetiology. Of special importance are variants that affect modifiable amino acid residues, as protein modifications involved in signal transduction networks cannot be analysed by genomics. Proteogenomics enables analysis of proteomes in context of patient- or tissue-specific non-synonymous nucleotide variants. Here, we developed a proteogenomics workflow and applied it to study resistance to serine/threonine-protein kinase B-raf (BRAF) inhibitor (BRAFi) vemurafenib in malignant melanoma cell line A375. This approach resulted in high identification and quantification of non-synonymous nucleotide variants and (phospho)proteins. We integrated multi-omic datasets to reconstruct the perturbed signalling networks associated with BRAFi resistance and to predict drug therapies with the potential to disrupt BRAFi resistance mechanism in A375 cells. Notably, we showed that aurora kinase A (AURKA) inhibition is effective and specific against BRAFi resistant A375 cells. Furthermore, we investigated nucleotide variants that interfere with protein post-translational modification (PTM) status and potentially influence cell signalling. Mass spectrometry (MS) measurements confirmed variant-driven PTM changes in 12 proteins; among them was the runt-related transcription factor 1 (RUNX1) displaying a variant on a known phosphorylation site S(Ph)276L. We confirmed the loss of phosphorylation site by MS and demonstrated the impact of this variant on RUNX1 interactome.