The Role of Harvey-ras in Mouse Skin Tumorigenesis

This thesis outlines two alternative approaches to investigating the role of H-ras in mouse skin tumour development. The first, the in vitro approach utilises a number of unique cell lines representative of a late stage of tumorigenesis, the conversion from a well differentiated squamous carcinoma to a undifferentiated spindle form. Through manipulation of the gene dosage ratio of mutant:normal H-ras in these cell lines, by plasmid transfection and virus infection, it was hoped to provide some clues as to the function of H-ras in this late stage transition event. The second approach involves the technique of gene targeting. The aim of this approach was to create a mouse hemizygous or completely deficient in H-ras and to examine the effects in the context of both chemical carcinogenesis studies and normal mouse development. Initial observations of the squamous/spindle transition event showed that the cell lines representative of both phenotypes were characteristically very different. The squamous cell line B9 produced a well differentiated tumour when injected into nude mice whereas the spindle cell lines A5 and D3 gave rise to aggressive, disorganised tumours with a much reduced latency. Genetic analysis of the cell lines showed that the most dramatic difference between the squamous and spindle derivatives was in the copy number and expression levels of H-ras. Both spindle cell lines A5 and D3 showed an approximate 5-10 fold increase in mutant H-ras, best explained by a genetic amplification mechanism. Experimentally overexpressing mutant H-ras in the squamous cells, and vice versa the levels of normal H-ras in the spindle variants, was unable to mimic the conversion event. However it was found that several A5 and D3 transfectant clones showed a dramatic reduction in tumorigenicity. This correlated not with the levels of the introduced normal H-ras but the absolute levels of endogenous mutant H-ras. Spontaneous loss of mutant copies of H-ras resulted in clones that morphologically resembled spindle cells yet produced tumours with a vastly increased latency period. It seems therefore that in this system mutant H-ras is the critical determinant of tumorigenicity, and relates perhaps to the growth of the tumour cell in vivo, possibly providing the driving force for the conversion. However, alteration of the H-ras gene dosage ratio was not in itself sufficient for the transition of a squamous to a spindle cell carcinoma, implying that another locus as yet unknown must play a role in controlling the epithelial phenotype. Finally, the gene targeting approach, provides a means to examining the earlier stages of mouse skin tumorigenesis, in particular the proposed role of normal H-ras in suppressing the development of papillomas from an initiated cell. For instance, if the consistent duplication of the chromosome carrying the mutant H-ras gene, seen in the development of papillomas, is related to overcoming a suppressive effect of normal H-ras, then carcinogen treatment of a mouse deficient in H-ras may result in a mouse more susceptible to tumorigenesis. Experiments with the gene targeting technique are on going and as yet no targeted ES cell clone has been isolated, but hope remains that this approach will provide new insights into the functions of H-ras both in early tumorigenesis and normal development