P53 modeling as a route to mesothelioma patients stratification and novel therapeutic identification

Abstract

Background: Malignant pleural mesothelioma (MPM) is an orphan disease that is difficult to treat using traditional chemotherapy, an approach which has been effective in other types of cancer. Most chemotherapeutics cause DNA damage leading to cell death. Recent discoveries have highlighted a potential role for the p53 tumor suppressor in this disease. Given the pivotal role of p53 in the DNA damage response, here we investigated the predictive power of the p53 interactome model for MPM patients' stratification. Methods: We used bioinformatics approaches including omics type analysis of data from MPM cells and from MPM patients in order to predict which pathways are crucial for patients' survival. Analysis of the PKT206 model of the p53 network was validated by microarrays from the Mero-14 MPM cell line and RNA-seq data from 71 MPM patients, whilst statistical analysis was used to identify the deregulated pathways and predict therapeutic schemes by linking the affected pathway with the patients' clinical state. Results: In silico simulations demonstrated successful predictions ranging from 52 to 85% depending on the drug, algorithm or sample used for validation. Clinical outcomes of individual patients stratified in three groups and simulation comparisons identified 30 genes that correlated with survival. In patients carrying wild-type p53 either treated or not treated with chemotherapy, FEN1 and MMP2 exhibited the highest inverse correlation, whereas in untreated patients bearing mutated p53, SIAH1 negatively correlated with survival. Numerous repositioned and experimental drugs targeting FEN1 and MMP2 were identified and selected drugs tested. Epinephrine and myricetin, which target FEN1, have shown cytotoxic effect on Mero-14 cells whereas marimastat and batimastat, which target MMP2 demonstrated a modest but significant inhibitory effect on MPM cell migration. Finally, 8 genes displayed correlation with disease stage, which may have diagnostic implications. Conclusions: Clinical decisions related to MPM personalized therapy based on individual patients' genetic profile and previous chemotherapeutic treatment could be reached using computational tools and the predictions reported in this study upon further testing in animal models.