Molecular and Functional Identification of Tumour Suppressor Genes Involved in Mouse Skin Carcinogenesis

Mouse skin tumourigenesis occurs through a series of discrete steps associated with specific genetic alterations, from an initiated cell, to a benign papilloma, and subsequently to a malignant squamous carcinoma. A proportion of these malignant tumours undergo a dramatic change in cell phenotype, which is accompanied by substantial alterations in the expression of several markers of epithelial differentiation. These spindle cell carcinomas have been well characterised biologically, but the genetic events which are responsible for the transition to these invasive tumours are not well understood. The work presented in this thesis was designed to address the genetic basis of the spindle transition by cytogenetic, molecular genetic and functional approaches. Amplification of the mutant H-ra5 allele and / or loss of the normal W-ras allele are consistently found in spindle cell carcinomas, which also frequently show imbalances of chromosome 7. The mechanistic basis for these ras / chromosome 7 changes was investigated by Fluorescence in situ Hybridisation using chromosome paints. In the squamous carcinoma cell line, B9, increases in mutant W-ras were caused by whole chromosome duplication, while in the clonally-related spindle cell lines, A5 and D3, a further increase in expression was achieved by localised amplification of the mutant W-ras gene on double minute chromosomes. To establish the nature of the gene(s) lost at the squamous-spindle transition, somatic cell fusions were carried out between the clonally-related B9 and A5 cells. The hybrids were epithelial in morphology and expressed characteristic epithelial proteins, such as E- cadherin and the keratins. In addition, they were found to be suppressed in their ability to form tumours following injection into nude mice. These experiments demonstrate that spindle cells arise by a mechanism involving loss of a tumour suppressor gene. To identify putative tumour suppressor loci, we developed an approach using hybrid cells generated between a keratinocyte cell line, C5N, and the spindle carcinoma cell line, carB. Tumourigenicity was initially suppressed in these hybrids, but the tumours which did arise, after a long latency, were poorly differentiated squamous carcinomas or undifferentiated spindle carcinomas. Allelotype analysis of the tumours enabled us to identify regions on mouse chromosomes 4 and 7, which harbour putative tumour suppressor genes involved in mouse skin tumourigenesis. In order to determine the function and relevance of each of these loci in the acquisition of the spindle phenotype, the syntenic regions from the human genome were introduced by microcell-mediated monochromosome transfer into spindle carcinoma cell lines. This strategy, involving a single human chromosome on a mouse background, facilitates finer mapping of the loci in revertant clones. Using this approach, we have identified a locus on chromosome 15 corresponding to the locus identified on mouse chromosome 7, which may cause growth inhibition of A5 cells. Introduction of human chromosome 9 into A5 and carB cells caused a partial reversion to the squamous phenotype and tumour suppression. The loss of the pi6 tumour suppressor gene located on human chromosome 9p21, is know n to be associated with the loss of differentiation which occurs in the transition to the spindle phenotype. Therefore the effect of human chromosome 9 on spindle cells may be partly explained by the consequences of introducing pi6. However, several lines of evidence point to the existence of a second locus on human chromosome 9 which may be important in the conversion to spindle carcinomas. Future work will focus on the identification of this gene(s) and the role it plays in this last stage of mouse skin tumourigenesis