Use this URL to cite or link to this record in EThOS:
Title: Determination of p53 role in cancer biology and therapy through interactome development and analysis
Author: Hussain, M.
Awarding Body: University of Salford
Current Institution: University of Salford
Date of Award: 2016
Availability of Full Text:
Access through EThOS:
Access through Institution:
Cancer is a heterogeneous pathology of cell and tissue type, involving multi- dysregulation of pathways that govern fundamental cellular processes. Chemotherapy efficacy is highly affected by these molecular deviations and the majority of patients are non-responsive. The stress responsive transcription factor, p53 is a powerful tumour suppressor implicated in over 50 % of all human cancers, and the chemo- therapeutic applications of p53 have been well described. However, the vast literature base of p53 and complexity of its interactions makes it a challenging system for integration of this diverse information. Computational methodologies are novel tools for integration of diverse molecular information into a coherent framework. The in silico Boolean PKT206 p53–DNA damage model has previously demonstrated good predictive capability for p53 wildtype and null tumours. Here, we have expanded PKT206 to generate a more clinically robust representation of p53-cancer dynamics. The Boolean PMH260 and PMH302 p53 models were constructed to consider 260 nodes and 980 interactions (PMH260) and 302 nodes with 1398 interactions (PMH302). Processes of angiogenesis, DNA repair and cell cycle arrest were amalgamated into both models with an additional input of hypoxia included into PMH302. For greater representation we further constructed a logical model that was more relevant to a specific cancer and integrated 61 deregulated genes considered as important to mesothelioma (Meso-PMH61 model), and superimposed microarray data generated in our laboratory of mesothelioma cells treated with etoposide and 1 % O2 oxygen. We analysed all models, in silico, using CellNetAnalyzer. In silico knockout analysis of various nodes mimicking in vivo mutations revealed 98 and 514 potential novel predictions (PNPs) for PMH260 and PMH302 respectively, some were validated by laboratory and literature verification. Validation of 4 PNPs were investigated in our laboratory using transient gene and protein knockdown, q-PCR and western blot analysis. Of these, 2 PNPs were in agreement with the models prediction. For further validation we superimposed various human cancer transcriptome in vitro profiles and compared omics results to in silico LSSA data. Greater correct predictions were achieved than the earlier p53 logical models (PKT206 and PMH260) when using the expanded p53 interactomes (PMH302 and Meso-PMH61) resulting in 68.5 - 83 % dependant on model and simulation. We further tested the models capability to predict gene expression changes on a clinical and individual patient basis, and superimposed patient derived in vivo transcriptome tumour profiles with a p53 mutant and wild type status. Correct predictions were again in the majority ranging between 57 - 61.5 % for NSCLC and between 56-63 % for biphasic and epithelioid mesothelioma tumours. Gene expression analysis of individual tumour profiles identified several significantly over-represented pathways and deregulated genes common and unique contributing to these tumours, correctly predicted by the model. In particular, we highlight pathways, MAPK, Erbb and Ca+ as contributing to non-small cell lung cancer dependant on patient. For malignant pleural mesothelioma, we highlight cell cycle and MAPK pathways and the cell cycle genes; CCNB1, CCNB2 along with KIF14, PDGFRB and SULF1. These offer potential for further investigation that could be exploited for greater therapeutic efficacy in sarcomatoid, and CYP24A1, HIPK4 and PEG3 in biphasic p53 (+/+) malignant mesothelioma tumours. Drug profile analysis of deregulated genes identified by the model highlight the need for individualised therapeutic approaches, and we offer putative combined targeted therapeutic suggestions dependant on tumour profile. In summary, we have generated the largest p53 signalling model to date, and have successfully identified overall system attributes when compared to in vitro and in vivo patient derived data. We show the importance of individualised therapies and highlight the enlarged p53 interactome as a promising predictive tool for further investigation into personalised anti-cancer therapies with clinical relevance.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available