Molecular Markers and Hypoxia in Renal Cell Carcinoma

Sammanfattning: Renal cell carcinomas (RCCs) are a group of tumours that arise from the nephron within the kidney. They are characterised by an indolent growth pattern and do not display overt clinical symptoms in patients with early to locally advanced tumours. Furthermore, the global average age of a RCC patient at diagnosis is 75 and they usually present with co-morbidities that may render invasive surgical/biopsy approaches risky. The most prevalent form of RCC is the clear cell RCC (ccRCC) subtype that displays pseudohypoxic activation of the two transcription factors HIF- 1α and HIF-2α as the result of a non-functional pVHL protein. Although the general influence of these two transcription factors has been deciphered, the full extent of the tumour-promoting activities of HIF-2α via its target genes have not been elucidated. Given these limitations in the clinical management of RCC and understanding of its biology, we set out to address these issues through the papers included in this thesis.In our first paper we aimed to decipher HIF-2α specific target genes operating in normal renal proximal tubule epithelial cells as well as ccRCC tumour cells. We pharmacologically emulated the loss of functional pVHL in renal proximal tubule cells whilst inhibiting HIF-2α transcriptional activity. Subsequent RNA-sequencing revealed potentially HIF-2α specific genes of which we selected SEMA5B, where its protein expression pattern matched our RNA-sequencing findings. We verified the HIF-2α regulatory specificity to SEMA5B in ccRCC cell lines, with other lines of evidence definitively demonstrating that HIF-2α but not HIF-1α specifically regulates the expression of SEMA5B in renal proximal tubule and ccRCC cells. Therefore, SEMA5B may have important role(s) in the context of ccRCC tumour vascularity and microenvironment.With papers two and three, we sought to address issues in RCC prognostication and detection. In paper two, we analysed transcription factor network and regulon activity in RCC subtypes using publicly available datasets. Using this analysis, we identified NFIA, a transcription factor that had similar regulon activity to HNF4A, a well characterised transcription factor in RCC. Based on this data, we examined the relationship between RNA expression of our selected transcription factor NFIA and TCGA-based RCC patient clinicopathological factors such as grade, stage andcancer-specific survival. We assessed the protein expression of NFIA and HNF4A in a large tissue-microarray consisting of ccRCC and papillary RCC (pRCC) tumours and found that NFIA expression can independently predict CSS in ccRCC patients. Molecular markers for the prognostication of RCC patients are not currently used in the clinic and we hope that our work can contribute towards their eventual implementation.In paper three we developed a workflow to enrich, detect and subtype RCC tumour cells from whole blood. Given the poor performance of EpCAM based circulating tumour cell enrichment methods in RCC, we utilised a size-based isolation platform. We demonstrated that RCC tumour cells are suitable for this approach and showed that our methodology can isolate down to one spiked-in tumour cell from whole blood. Furthermore, through differential gene expression analyses between the three RCC subtypes, we identified transcriptomic markers that can be used to detect pRCC and ccRCC tumour cells from whole blood. This paper lays out a large extent of the methodology and fine-tuning required to isolate and detect RCC tumour cells from whole blood and may provide an additional way to monitor RCC patients in adjuvant and/or neoadjuvant settings.