MUTATION AND EXPRESSION OF THE P53 TUMOR SUPPRESSOR GENE IN TUMOR SAMPLES AND CANCER CELL-LINES - COMPARISON OF NONISOTOPIC DIRECT DNA-SEQUENCING, IMMUNOBLOTTING AND IMMUNOHISTOCHEMISTRY

Abstract

Mutations in the nuclear phosphoprotein p53 are the most frequent genetic alterations in human solid tumors detected so far. These mutations are clustered in highly conserved domains spanning from exon 4 to 9 of the gene. A very precise method of detecting p53 mutations is to sequence these domains. However, 2 to 3 overlapping PCR-amplifications were needed to span the whole mutation-prone region. We used a very rapid non radioactive solid-phase DNA sequencing method starting from mRNA to sequence the p53 domains in both directions with T7 DNA-polymerase allowing detection of the heterozygous state, where one allele shows the wild-type sequence, the other a mutated one. First we sequenced four colon carcinoma cell lines with known p53 mutations and one T-cell-leukemia cell line with a heterozygous situation to validate our method. Using this method we sequenced the p53 gene (exons four to nine) from 16 primary colon carcinomas. Seven of these 16 (44%) carcinomas showed mutations in the p53 gene resulting in amino acid exchanges. One showed a silent mutation, another one showed two point mutations in the highly conserved domain of the p53 gene. These colorectal carcinomas have been examined for overexpression of the p53 protein using a panel of monoclonal antibodies directed against p53 (PAb1801, PAb240, PAb421, PAb1620) by immunohistochemical analysis and immunoblotting. Furthermore, four colorectal cancer cell lines were examined by indirect immunofluorescence technique with the same mAb PAb1801 as used in histological staining. Analysis of 6 out of 15 (40%) tumor specimens revealed markedly positive p53 nuclear staining patterns using monoclonal antibody PAb1801. These data suggest that there is quite a good correlation between point mutation of the p53 gene and nuclear staining with monoclonal antibody PAb1801 detecting overexpressed p53 protein. Moreover, there is no convincing evidence that wild-type protein can be detected using the monoclonal antibodies PAb 1801 and PAb 1620.