Histological evaluation of putative biomarkers in prostate cancer

Prostate cancer is a significant cause of illness and death in males. Marked disease heterogeneity is associated with prostate cancer. Current detection strategies do not detect the disease at an early stage and cannot distinguish aggressive versus non aggressive prostate cancer leading to over-treatment of the disease and associated morbidity. Thus, prostate cancer is a disease that would benefit from the discovery of novel biomarkers for both early detection and treatment selection. In October 2003, the Prostate Cancer Research Consortium (PCRC), a multidisciplinary, trans-institutional collaboration composed of researchers from University College Dublin's Conway Institute, Trinity College Dublin's Institute for Molecular Medicine, the Royal College of Surgeons in Ireland and Dublin City University, was established. Using a comprehensive approach which involves utilizing genomic, transcriptomic and proteomic platforms and using a highthroughput biomarker validation approach, the consortium's overall aim is to identify novel biomarkers of prostate cancer. This PhD project aims to evaluate the potential prostate cancer biomarkers in tissue which are identified in the discovery projects within the PCRC. Following previous proteomic analysis within the PCRC group, Zinc-a-2- glycoprotein (ZAG), Proteasome subunit P type 6 (PSMB-6), Vitamin D binding protein (VDBP) and Kininogen-1 (KNG-I) were found to be upregulated in the serum of prostate cancer patients. The tissue expression profile of these proteins was investigated to determine if ZAG, PSMB-6, VDBP and KNG-1 were also upregulated in prostatic tumour epithelial cells in the tissue of prostate cancer patients. ZAG expression in epithelial cells of the prostate was inversely associated with Gleason grade (BPH>G3>G4/G5). PSMB-6 was not expressed in either tumour or benign epithelium. However, strong PSMB-6 expression was noted in stromal and inflammatory cells. KNG-I and VDBP were not found to be expressed in prostate tissue blood. These results indicate ZAG as a possible predictive marker of Gleason grade. The inverse association between grade and tissue expression with a rising serum protein level is similar to that seen with PSA. In addition the results for all proteins (ZAG, PSMB-6, KNG-1 and VDBP) highlight the challenges in trying to associate the protein levels in serum with tissue expression. Following methylation analysis of Wingless signalling (Wnt) molecules and their antagonists using Quantitative Methylation-Specific PCR (QMSP), secreted frizzled related protein 2 (SFRP-2) was found to be highly methylated in prostate cancer. The prostatic tissue expression profile of SFRP-2 was investigated to determine whether there was a correlation between SFRP-2 methylation and SFRP-2 expression in prostate cancer. Strong to moderate SFRP-2 expression was observed in benign prostatic hyperplasia (BPH) epithelial cells and negative to weak SFRP-2 expression observed in tumour epithelium particularly Gleason grade 3 and 4. However, in Gleason grade 5 carcinoma there was a 40:60 split in the immunoexpression of SFRP-2, where 40% displayed strong to moderate SFRP-2 expression and 60% displayed negative SFRP-2 expression in epithelial cells. A morphological difference was noted in the Gleason grade 5 tumours that had strong to moderate expression of SFRP-2 (Type A) and the Gleason grade 5 tumours that had no SFRP-2 expression (Type B). It was also noted that biochemical recurrence occurred after 5 years in patients that had strong to moderate SFRP-2 expression in Gleason grade 5 tumours and had "Type A morphology. These preliminary results propose SFRP-2 as a possible marker of histologically benign glands and a possible subgroup of Gleason grade 5 tumours that may predict prognosis and biochemical recurrence. The use of well-characterised antibodies is vital for clinical diagnostics and protein studies. One of the major challenges facing researchers and clinicians in the area of cancer research is the lack of high quality, well characterised antibodies to novel proteins. Successful antibody generation depends on the use and availability of high quality antigen. However, like antibodies to novel proteins in cancer research, there is also a lack of high quality, well characterised, commercially available antigens to these proteins. Thus, generating antibodies to these targets is not straightforward and often requires heterologous production of recombinant proteins. SFRP-2 is a novel marker in prostate cancer and paucity of quality antigen and antibody is an impediment to research on this marker. An antigen grade recombinant protein was designed, produced and used as an immunogen in an avian immune model to determine whether a recombinant antibody against SFRP-2 can be generated based on the polyclonal serum response of the chicken. The first strategy taken to produce the SFRP-2 antigen involved using a prokaryotic (E. coli) expression system. Both a pGS21 -a vector containing a glutathione-s-transferase (GST) tag and a PET-28b(+) vector modified to contain a heart fatty acid binding protein (hFABP) tag were transformed into E. coli cells for soluble SFRP-2 recombinant fusion protein expression which could be subsequently used as an immunogen in an avian immune model. However, both fusion proteins were insoluble in the form of bacterial fusion proteins trapped within the bacterial cell periplasm. The second strategy taken to produce a soluble SFRP-2 antigen involved using a "cell-free" translation expression system. "Cell-free" translation is an in vitro expression system that does not require the use of a host cell to express the target protein. Soluble expression of SFRP-2 was achieved. However, the yield of the soluble protein was too low for its use as either an immunogen or in molecular applications. The challenges and difficulties faced in this study may reflect the lack of high quality, well characterised, commercially available antibodies to novel cancer targets such as SFRP-2.