Genetic events in colorectal tumorigenesis
Cripps, Kathryn Jane
More is known about the genes involved in colorectal tumorigenesis than for any other human cancer. Mutations have been identified in many genes, including K-ras,APC, M C C , D C C and p53. However, whilst much is known about these events, there are many questions that remain unanswered. Three specific questions involving the p53, M C C and A P C genes were addressed in this thesis.Firstly, the relationship between p53 mutation and stabilisation of p53 protein was assessed. It had been previously assumed that stabilised p53 protein product was indicative of point mutation within the p53 gene. More recently, however, situations have been described in which this relationship is not absolute. Accordingly, the incidence of stabilised product detected by immunocytochemistry (ICC) using p53 specific antibodies, was compared in a series of 47 colorectal carcinomas with that of mutation, detected by single strand conformational polymorphism analysis (SSCP) and sequencing of p53 exons 5-8. The results suggest a high correlation between the detection or absence of p53 protein stabilisation and the presence or absence of mutations within the gene (35/47:74%). However, specific incidences are highlighted in which concordance is not absolute. In particular, mutations in exon 6 ofp 5 3 do not appear to result in stabilised protein and 2 tumours with a high degree of staining contained no apparent mutation within the entire coding region of the gene. Eight tumours within the series with positive ICC staining have also been shown to have no mutations within exons 4-9, a region where the majority of mutations of p53 have previously been shown to occur. Possible explanations for the observed lack of an absolute concordance are discussed.Secondly, mutations in M C C were studied in a subset of colorectal carcinomas. Although the M C C gene, on chromosome 5q21, was identified prior to A P C, and found to be mutated in some colorectal cancers (hence Mutated in Colorectal Cancer), its role in colorectal carcinogenesis is still uncertain. Only five sporadic colorectal tumours with M C C mutations have been published and no germline mutations of the gene have been identified in Familial Polyposis Coli (FPC) patients. A previous study from the Edinburgh Cancer Research Campaign Group had shown that whilst 5q21 allelic losses in a series of twenty colorectal carcinomas invariably involved M C C , no mutations of the remaining allele could be detected. It remained possible however, that M C C might be significant in other forms of colon cancer. Mutations were therefore sought in the coding region of M C C by SSCP in five colorectal tumours from non-polyposis patients who had presented with colon cancer at a young age (less than 48 years). Tumours had been maintained as xenografts in SCID mice and were therefore considered to be free from human contaminating (normal) stromal elements. No mutations were found in any of the young cases, although all previously published tumours with M C C mutations were detectable by the same analysis. These results therefore add to the evidence that inactivation of M C C by mutation is not important in colorectal carcinogenesis at least at the level detectable by this study.Finally, essential preliminary steps were taken in establishing an experimental system in which to study the role of A P C mutation in colorectal carcinogenesis. APC , which is close to M C C on 5q21, is mutated in at least 60% of sporadic colorectal tumours, and germline transmission of A P C mutations are responsible for FPC. Mutations in APC. are believed to be an early event in colorectal tumour progression and are detectable at equal frequency in benign and malignant tumours. It is of interest to establish whether such 'early' events are still important in determining the phenotype of tumours at later stages ot progression, as this would influence gene therapy strategies in colorectal cancer. Accordingly, gene targeting constructs were designed to execute a "Hit and run" strategy that would convert a known truncating mutation in the A P C gene in the colorectal cancer cell line SW480 to wild type. The SW480 cell line was characterised by a variety of techniques and shown to contain three mutated copies of the A P C gene, no normal copies and to be tumorigenic in SCID mice. In addition to a putative correcting construct (which remains to be confirmed and tested), a similar construct was made that would introduce a novel mutation 21 codons from the corrected one. This latter construct has provided a control for the efficiency of targeting the A P C gene in this cell line.