Somatic Mutations in the p53 Tumor Suppressor Gene

27 May

Somatic mutations in the p53 tumor suppressor gene in human cancer have been seen more com- monly than somatic mutations in any other single gene. About half of all cases of human cancer studied exhibit somatic mutations in the p53 gene (Harris, 1993), including nearly 70% of col- orectal cancers, 50% of lung cancers, and 40% of breast cancers (Culotta and Koshland, 1993). As Harris (1993) has noted, mutations in the p53 tumor suppressor gene are different from those seen in other tumor suppressor genes. Whereas Rb and APC are commonly inactivated by non- sense mutations, causing truncation or instability of the protein, more than 90% of the mutations in the p53 tumor suppressor gene are missense mutations, transitions, or transversions that result in a change of an amino acid within the protein. Such changes may alter the conformation and increase the stability of the protein. This observation has led a number of authors to utilize im- munohistochemistry  of tissue slices with antibodies specific to a mutant p53 protein to obtain evidence for presumed mutation in the p53 gene (Bartek et al., 1991; Zusman, 1995). Guinee et al. (1995) extended these studies, demonstrating  that enhanced immunohistochemical  staining for p53 correlated  predominantly  with the presence  of missense  mutations  in specific  exons (5–8) of the p53 gene. The p53 protein consists of 393 amino acids with functional and evolu- tionarily conserved domains (Harris, 1995). A diagram of the relative frequency of mutations at specific sites in the human p53 tumor suppressor gene is seen in Figure 6.2. The majority of missense mutations occur in the conserved hydrophobic  region, while nonmissense  mutations are distributed throughout the protein. Direct mutagenesis of p53 cDNA by benzo[a]pyrene di-

Figure 6.2 Schematic representation  of the p53 molecule. The protein extending from the amino termi- nus (N) on the left to the carboxy terminus (C) on the right is diagramed with its functional and evolution- arily conserved  domains.  The numbers  indicate  the codons  where  a very high incidence  of mutation (mutational hot spots) occurs. (Adapted from Harris, 1995, with permission of the author and publisher.)

olepoxide and the epoxide of aflatoxin B1 results in G:C-T:A transversions in a number of loca- tions in the conserved region by the diolepoxide and a similar change with the ultimate form of aflatoxin. These mutations were localized to a considerable extent to codon 249 (Puisieux et al.,

1991), but this codon was not a target for the benzo[a]pyrene diolepoxide. Interestingly, in hepa- tocellular carcinomas developing in individual regions of the world where aflatoxin B1 exposure is endemic, a similar transversion in codon 249 was noted (Hsu et al., 1991). A diagram of the proportion and types of various mutations seen in a number of cancer types as well as the distri- bution of p53 mutations in general is given in Figure 6.3. As noted in the figure, there are dis-

tinct, different distributions  of the various types of mutations  in each of the types of cancer depicted. This is particularly notable in colon cancer, which develops in adenomatous polyposis coli in patients exhibiting a germline mutation in APC, compared with sporadic colon cancer where only somatic mutation in this gene is presumably involved. The proportions of the transi- tion mutations G:C to A:T at CpG dinucleotide positions where the C is presumably methylated are quite different in colon and lung cancer. Rideout et al. (1992) have suggested that this indi- cates a more important role for endogenous mutagens in colon cancer than in lung cancer, where presumably exogenous tobacco smoke is the primary carcinogenic factor.

In later chapters (Chapters 9 and 15) the molecular functions of p53 and other tumor sup- pressor genes are considered. However, it is obvious that this gene in particular is the target for a variety of exogenous and endogenous mutagens. It is also clear that many other somatic muta- tions may occur in both oncogenes and tumor suppressor genes, and it is unlikely that any single mutation in any single gene is the ubiquitous lesion that leads directly to cancer.

Random Posts

Comments are closed.