Background and aim: More than 75% of the genome transcribe non-coding RNAs. This thesis explored long non-coding RNAs (lncRNAs) and microRNAs (miRNAs), which are important gene regulators that when misregulated promote cancer progression. Most lncRNAs are uncharacterised but some could be novel cancer drivers. The potential of non-coding RNA versus mRNA signatures as biomarkers is relatively under explored. The overall hypothesis of the thesis was non-coding RNAs are important novel regulators of cellular processes and biomarkers in cancer. The specific objectives were to: 1) identify a novel lncRNA and kinase regulating MAPK signalling in lung adenocarcinoma (LUAD); 2) identify novel chromatin associated lncRNAs involved in chemoresistance in small cell lung cancer (SCLC); 3) develop a lncRNA signature that reflects intrinsic radiosensitivity; 4) develop mRNA hypoxia-related signatures and explore the potential of lncRNAs in hypoxia in lung cancer; and 5) derive a miRNA hypoxia signature for muscle invasive bladder cancer (MIBC), comparing its performance with the 24-gene mRNA signature. Methods: 1) RNA-seq co-expression networks were used to identify a novel lncRNA and kinase in the MAPK signalling pathway. The functions of the lncRNA and kinase were explored using knockdown and overexpression approaches respectively. 2) Candidate lncRNAs were identified using baseline circulating-tumour-cell-derived explant (CDX) models. Three lncRNAs identified were studied using siRNA knockdown and by measuring cisplatin sensitivity in SCLC cell lines. 3) Signature generation involved radiosensitivity data of primary cervix tumours (n=46), radiosensitivity data and RNA-seq expression data of cell lines from the Cancer Cell Line Encyclopedia (n=531) as well as Affymetrix exon array data for patients with cervix (n=160) and bladder (n=151) cancer who had radiotherapy. 4) Hypoxia-induced mRNAs and lncRNAs were identified using RNA-seq data for 8 lung cancer cell lines grown in 1% vs 21% O2 and the hypoxia induced mRNA genes were used to derive a LUAD hypoxia-related signature using RNA-seq data from part of the The Cancer Genome Atlas (TCGA) LUAD dataset. 5) Four bladder cancer cell lines grown in 0.2% vs 21% O2 were profiled using miRNA array cards. Hypoxia-induced genes were used to derive a signature using part of the RNA-seq data from the TCGA-BLCA dataset. Results: 1) LINC01267 and MAP3K19 were identified, with LINC01267 potentially regulating the ERK signalling pathway and MAP3K19 shown to activate the ERK and JNK signalling pathways. 2) Three lncRNAs identified using CDX models had no effect on cisplatin sensitivity in SCLC cell lines. 3) A pan-cancer signature was not possible but a ten-lncRNA bladder radiosensitivity signature was derived. The signature had a trend for prognostic significance in patients who had radiotherapy but not in patients treated with surgery. The radiosensitivity index based on mRNA outperformed the bladder lncRNA signature. 4) A 30 gene hypoxia-related LUAD mRNA signature was derived and three lncRNAs (MIR210HG, DARS-AS1, LINC00887) were identified that could be explored further for a role in regulating hypoxia in LUAD. 5) A 14-miRNA hypoxia signature was derived in MIBC, which predicted benefit from having hypoxia-modifying treatment with radiotherapy. The miRNA hypoxia signature had a comparable performance to a 24-gene mRNA signature, but their combined performance was superior to either used alone. Conclusions: Bioinformatic approaches can identify novel lncRNAs provided multiple gene expression datasets are available. Finding lncRNAs associated with chemoresistance in SCLC should explore paired baseline and progression CDX models. It is challenging to develop lncRNA signatures as biomarkers to validate in formalin fixed paraffin embedded (FFPE) tissue. For validation in FFPE tissue, miRNA signatures show potential. Future work should explore combined miRNA-mRNA signatures as biomarkers.
|Date of Award
|31 Dec 2020
- The University of Manchester
|Catharine West (Supervisor) & Claus Jorgensen (Supervisor)