Search

Antisense oligonucleotides targeting p53 have been hailed as a potentially new technique for treating patients with cancer, and there have been encouraging reports of good patient tolerance in vivo and of antiproliferative effects in vitro. However, evidence is lacking that these oligonucleotides are acting via an antisense interaction to modulate p53 expression. We examined a phosphorothioate antisense oligonucleotide, directed against exon 10 of the TP53 gene, and a chimaeric phosphorothioate-phosphodiester oligonucleotide directed against the p53 translation initiation codon. Both failed to specifically suppress p53 protein production in a cell-free assay system or to have any effect on mutant p53 expression by human pancreatic cancer cell lines. Antiproliferative effects were apparent, especially with the phosphorothioate antisense oligonucleotide, but this was independent of the p53 status of the cells (mutant, wild-type or absent) and also occurred with the control (sense and randomised) oligonucleotides. The most dramatic antiproliferative effects were seen with the 'control' phosphorothioate oligonucleotides. These findings suggest that the antiproliferative effects of some antisense oligonucleotides may be unrelated to expression of the gene they have been designed to target.

We developed a high-titer amphotropic retroviral vector that expresses mutant (Ala143) human p53 to test directly the response of genetically normal human epithelial cells to p53 mutation. Contrary to our prediction, we found that in pancreatic epithelium (whose tumors display a high frequency of p53 mutation) but not in thyroid (whose tumors show an exceptionally low mutation frequency), expression of mutant p53 induced a dramatic, though self-limiting, proliferative response. This result questions the assumption that p53 mutation is relevant only to the later stages of tumorigenesis.

The tumour suppressor gene p53 has been found to be mutated or inactivated at high frequency in several common human tumours. We have examined a series of exocrine pancreatic carcinomas for over-expression of mutant forms of p53 by immunohistochemistry with a panel of specific antibodies. We found immunodetectable p53 in 13 of 22 (60%) frozen pancreatic cancers and seven of 13 pancreatic cell lines. One of the antibodies, CM1, recognises p53 in formalin-fixed, paraffin-embedded archival material and using this reagent we found immunodetectable p53 in 28 of 124 (23%) pancreatic cancers. We have successfully demonstrated the presence of point mutations by direct sequencing of genomic DNA extracted from archival tissue showing CM1 immunoreactivity. We conclude that p53 activation is an important event in human pancreatic tumorigenesis and that the CM1 antibody can detect a proportion of cases of overexpression of mutant p53 in archival pathological material.