Molecular Biology of Urological Cancer Group, Cancer Research UK Cancer Medicine Research Division

Molecular Genetics (M Knowles)
Our genetic studies of bladder cancer aim to elucidate the molecular pathogenesis of tumour development and to understand how the divergent clinical behaviour of these tumours may be determined by genotype.

Deletions of chromosome 9 are found in more than 50% of bladder tumours of all grades and stages. Deletions commonly involve loss of an entire parental homologue leading to the suggestion that a gene or genes on both arms of the chromosome are involved in tumour development. Several candidate tumour suppressor genes have been identified on each chromosome arm and we continue to explore the role of these individual genes in urothelial transformation.

At 9p21, the complex CDKN2A/ARF locus shows homozygous deletion in many cases. The individual contribution of the two genes encoded, p16 and p14ARF, is not known and we have developed real time PCR assays to allow more precise deletion mapping of 9p21. Recent results show that some melanomas have deletions affecting only the p14ARF and not the p16 transcript. To date we have only found large homozygous deletions including both coding regions in bladder tumours and cell lines.

Studies of the 9q tumour suppressor genes DBCCR1 and TSC1 have continued. DBCCR1 induces G1 arrest in NIH 3T3 cells and marked inhibition of proliferation in human bladder tumour cells. This effect does not appear to be mediated via known G1 checkpoint control proteins suggesting an indirect mechanism. Recent experiments provide some clues to the function of this gene. Transiently transfected cells show rounding that is not indicative of apoptosis and cDNA microarray results indicate altered expression of genes related to the cytoskeleton. We have found that DBCCR1 shares a protein domain with perforin, a protein involved in pore formation during complement mediated cell lysis and several other proteins including astrotactins 1 and 2 which are implicated in neuronal cell migration. A series of truncation constructs of DBCCR1 are being used to examine potential functional domains in more detail.

Re-expression of the Tuberous Sclerosis tumour suppressor gene TSC1 in bladder tumour cells containing only a mutated version of the gene induces a profound effect on cell phenotype, including inhibition of proliferation, reduction in cell saturation density and inhibition of tumorigenicity. Preliminary observations of changes in the distribution of the focal adhesion associated protein paxillin and of tuberin the TSC2 gene product which is known to bind to the TSC1 protein, suggest that some phenotypic effects may be mediated via changes in the cytoskeleton or in cell adhesion. This raises the possibility of the involvement of common pathways in the function of the two 9q candidate genes DBCCR1 and TSC1.

Several other genetic alterations in bladder cancer are being studied. These include novel small regions of LOH on chromosome 15 found in ~30% of tumours and deletions of 8p which are found in >50% of tumours of high grade and stage. We have defined a common breakpoint region on proximal 8p that is shared with breast tumours and several candidate genes have been identified within the refined critical region of deletion. In bladder tumour cell lines, MFISH has identified translocations involving 8p and chromosomes 4, 17 and 21.

We have found constitutive activation of the fibroblast growth factor receptor FGFR3 via point mutation in ~40% of bladder tumours. This represents the most common genetic alteration to an oncogene in bladder cancer. Mutations have not been identified at high frequency in any other tumour type examined to date. Normal urothelial cells express only one of two known alternative transcripts of FGFR3 that binds FGF1. Recent studies of bladder tumour cell lines show that not only is this so-called 'epithelial' transcript expressed but also the alternative transcript commonly expressed by non-epithelial cells. The latter shows wider ligand binding specificity including binding of FGF2, known to be present in the urine of bladder cancer patients. This may represent an alternative mechanism of activation of FGFR3 signalling in these tumours.


Applied Urothelial Biology (M Knowles, P Selby)
The goal of the team is to translate basic urothelial research into novel therapies for patients with bladder cancer. One aim is to create targeted gene therapies for the treatment of transitional cell carcinoma of the bladder. Although the bladder is an ideal target organ for virus-mediated gene therapy, there is little information relating to gene delivery to the urothelium either in vitro or in vivo. Using a lacZ-containing replication-defective adenovirus, we have demonstrated efficient, reproducible transduction of both normal and malignant urothelial cells from clinical specimens, both in primary explant and in organotypic cultures. Both organ cultures of normal urothelium and clinical tumour specimens show successful adenoviral transduction of superficial layers of urothelium, but minimal penetration to deeper layers. Immunohistochemistry on intact urothelium, using an antibody against the human cell surface adenoviral receptor suggest that a physical barrier, rather than receptor status, is more likely to be the limiting factor. Experiments to improve viral penetration to deeper cell layers are under way.

The transcriptional control elements of bladder-specific genes provide a possible mechanism to efficiently target gene therapy for bladder cancer. The uroplakins (UPIa, UPIb, UPII and UPIII) are four proteins that make up the asymmetric unit membrane of terminally differentiated urothelial cells and their expression is restricted to urothelium. We have characterised the promoters of both UPIa and UPIb using a series of deletion constructs linked to the luciferase reporter gene. We have identified a 2.5 kb tissue-specific region of the UPIa promoter and a 400 bp UPIb promoter region which induces high level expression of the reporter gene. The UP1a and UPIb promoters are being incorporated into novel lacZ-expressing adenoviruses to assess their activity in both in vitro and in vivo mouse models. Future work will use such models to assess the possible clinical benefits of novel recombinant adenoviruses linking these promoters to therapeutic genes.

Selectively-replicating adenoviruses targeting defects in the p53 and pRb tumour suppressor pathways provide a possible means for exploiting the known molecular pathogenesis of transitional cell carcinoma of the bladder. Experiments using normal human urothelial cells, a panel of human bladder tumour cell lines with defined p53 and pRb status, and organotypic cultures are under way to assess tumour-selective replication of such viruses in individual cells and in intact epithelium.

We continue to develop clinical trial protocols using these selectively-replicating adenoviruses in the treatment of high-risk superficial bladder cancer. The UK Gene Therapy Advisory Committee have recently approved a phase I clinical trial with the selectively-replicating adenovirus Onyx-015 in patients with high-risk superficial or muscle-invasive bladder cancer. This trial will be linked to laboratory research assessing the relationship of selective viral replication to the molecular phenotype of biopsy specimens and clinical outcome.


Lymphothelial Interactions (LK Trejdosiewicz)
Although the majority of human adult tumours are carcinomas of epithelial cell origin, the functional and molecular basis for interactions between epithelial cells and the immune system is not well understood. Nevertheless, an understanding of the mechanisms governing lymphoepithelial interactions has clear implications for local immunity during malignant transformation and tumour progression.

As local immune-mediated interactions frequently take place in the context of inflammation and wound repair, we have become interested in how the histioenvironment can modify the responses and interactions of epithelial cells and T lymphocytes to juxtacrine and paracrine signalling. Transforming growth factor (TGF)β is central to wound repair processes and, in order to mimic the early events of wound-healing, we investigated the effects of TGFβ on mitogen-stimulated T cells. Whereas TGFβ added at the initiation of mitogenesis did not significantly alter T cell activation, proliferation, CD45 isoform switching or activation-induced cell death, TGFβ added >72 h post-activation enhanced the cumulative increase of apoptotic T cells. IL-2 was found to abolish the pro-apoptotic effects of TGFβ on post-activated T cells, thus demonstrating that TGFβ increases the cytokine-dependence of T cells for survival. This novel observation suggests that once the initiating insult has been removed, locally-produced TGFβ helps to eliminate T cells and hence down-regulates the immune-driven inflammatory response. Many carcinomas produce active TGFβ, which may reflect another anti-lymphocyte weapon in the tumour cell arsenal.

In order to further our understanding of tissue microenvironments in the context of local immunity and wound-healing responses, we have been developing human organ culture models, concentrating on the liver, a frequent site of carcinoma metastasis, and the urinary bladder. In the urothelial system, in collaboration with Jenny Southgate (York) we have shown that the stromal/matrix factors implicated in the wound-healing response induce expression of genes which have been previously implicated in malignant progression, suggesting that malignant cells abrogate the normal epithelial wound response during invasive and metastatic behaviour.

Activated T cells express or secrete several members of the Tumour Necrosis Factor (TNF) family, although the effects of such ligands on normal and malignant urothelial cells have not been studied extensively. Whereas it is well known that soluble versus cell-surface presentation of Fas Ligand can have different effects, we have now shown that the consequences of CD40 ligation are also context-dependent. Ligation of CD40 in bladder tumour cells with soluble trimeric ligand can result in growth inhibition, but not apoptosis. By contrast, CD40L presented by transfected fibroblasts leads to cell death. This pro-apoptotic effect can be triggered by agonistic CD40 antibody, but only if presented via CD32 (Fc receptor)-transfected fibroblasts (in collaboration with John Gordon, Birmingham).

Susceptibility to CD40-mediated apoptosis appears to be a feature of malignantly-transformed bladder cells (ICRF Scientific Report 2000) and we have shown that susceptibility to CD40-mediated cell death can be conferred to normal cells following disablement of p53, a key genetic event in the development of invasive bladder carcinomas. We have used the HPV16 E6 gene to disable p53 function in normal urothelial cells and, in collaboration with Jenny Southgate (York), have demonstrated that this leads to an inability to survive genotoxic damage, whereas normal urothelial cells show tolerance to genotoxic insult. This finding is consistent with clinical observations that bladder tumours with p53 loss have an increased sensitivity to chemotherapy. By contrast, non-apoptotic 'replicative' cell death of proliferating normal urothelial cells in response to non-genotoxic agents, such as long chain polyunsaturated fatty acids, is independent of p53 function.


For a list of refereed research papers, see Publications (in navigation on left).