Understanding the Role of Androgen Receptor Signalling in Prostate Cancer: A Lipidomics Approach by MALDI Mass Spectrometry Imaging

Abstract

Prostate cancer (PCa) remains one of the deadliest cancers globally and is primarily driven by aberrant androgen hormone signalling. Clinical blood-based biomarker testing which detects prostate specific antigen (PSA) and the Gleason grading system have been important diagnostic and prognostic tools to aid clinical management. However, a significant research challenge is the ability to stratify men according to their risk of tumour progression without having to overtreat many of them. Recent research has discovered that the androgen receptor (AR) regulates key lipid metabolic genes throughout PCa tumorigenesis. Advances in metabolomics/lipidomics technologies have further emphasised this unique hallmark of PCa with changes across the diversity of the lipid classes. The hypothesis of this thesis is that lipid composition of PCa has prognostic value. Currently, conventional lipidomics, like other omics, is confounded by prostate tumour heterogeneity. This project utilised matrix-assisted laser desorption/ionisation mass spectrometry imaging (MALDI MSI) that has the advantage of in situ spatial information to histologically delineate tissue-type specific lipid profiles in prostate glands. Initial MALDI MSI experiments in 10 patients, achieved by sublimation of MALDI matrix on tissue sections collected at interspersed levels, revealed lipidomic profiles consistent with morphological and pathological features of PCa. These findings were further underpinned by validation using MALDI MS/MS imaging and lipidomics. A subsequent study in an independent cohort employed automated spray coating of matrix that afforded markedly improved lipid ion sensitivity in prostate epithelia and critically, the stroma which is the physiological milieu for the tumour microenvironment.Crucially, it was demonstrated using a combination of clustering and multivariate analysis methods that benign glands are enriched in particular lysophospholipids compared to tumours. However, lipid metabolism appeared to be heterogenous in tumours which did not show sufficiently discriminative lipid species. MALDI MSI did also appear to indicate spatial changes in relative abundance for the common monounsaturated phosphatidylcholine species PC(34:1 ). This project also assessed the uptake of the AR inhibitor, enzalutamide (ENZ), in a well-characterised ex vivo model of human PCa of patient-derived explants (PDE). An LC-MS/MS bioanalytical assay was developed and validated according to US FDA guidelines. Together with targeted drug imaging by MALDI MS/MS, these studies revealed a time-dependent uptake of ENZ and its enhanced localisation to AR-positive epithelial rich areas. The significance of this work is to better understand pharmacodynamic implications of experimental design in tissue culture models. Finally, a more advanced high resolution MALDI Q-TOF system was applied to interrogate the lipid profiles of PDE samples. Explants from five patients' biopsies with high grade prostate tumours were cultured in DMSO or ENZ. Spatial characterisation in dual MALDI polarities was conducted at a lateral resolution of 20 µm that offered well-defined tissue segment lipidomic fingerprints. Negative ion mode imaging uncovered heterogenous abundance of cholesterol sulphate in multifocal tissues. Altered lipid composition in drug treated PDEs possibly because of pharmacological inhibition of AR was seen. This facet of the project has potential to unravel novel lipid biomarkers of treatment response and other metabolic susceptibilities of aggressive PCa and promises exciting investigations in the future.