W P Bennett

Cell cycle arrest at the G1 checkpoint allows completion of critical macromolecular events prior to S phase. Regulators of the G1 checkpoint include an inhibitor of cyclin-dependent kinase, p16INK4; two tumor-suppressor proteins, p53 and RB (the product of the retinoblastoma-susceptibility gene); and cyclin D1. Neither p16INK4 nor the RB protein was detected in 28 of 29 tumor cell lines from human lung, esophagus, liver, colon, and pancreas. The presence of p16INK4 protein is inversely correlated with detectable RB or cyclin D1 proteins and is not correlated with p53 mutations. Homozygous deletions of p16INK4 were detected in several cell lines, but intragenic mutations of this gene were unusual in either cell lines or primary tumors. Transfection of the p16INK4 cDNA expression vector into carcinoma cells inhibits their colony-forming efficiency and the p16INK4 expressing cells are selected against with continued passage in vitro. These results are consistent with the hypothesis that p16INK4 is a tumor-suppressor protein and that genetic and epigenetic abnormalities in genes controlling the G1 checkpoint can lead to both escape from senescence and cancer formation.

Preinvasive lesions of squamous cell carcinoma are well defined morphologically and provide a model for multistage carcinogenesis. Since alterations in the p53 tumor suppressor gene occur frequently in invasive esophageal squamous cell carcinoma, we examined a set of preinvasive lesions to investigate the timing of p53 mutation. Surgically resected tissues from nine patients with esophageal squamous cell carcinoma contained precursor lesions which had not yet invaded normal tissues. Immunohistochemistry showed high levels of p53 protein in both preinvasive lesions and invasive carcinomas in six cases; sequence analysis of all invasive tumors identified p53 missense mutations in two cases. Preinvasive lesions from both tumors with mutations plus one wild-type tumor were microdissected and sequenced. In one patient there were different mutations in the invasive carcinoma (codon 282, CGGarg > TGGtrp) and a preinvasive lesion (codon 272, GTGval > T/GTGleu/val). In a second case, an invasive carcinoma had a mutation in codon 175 (CGCarg > CAChis), and adjacent preinvasive lesions contained a wild-type sequence. A carcinoma and preinvasive lesion from the third case contained high levels of protein and a wild-type DNA sequence. Therefore, p53 mutation may precede invasion in esophageal carcinogenesis, and multifocal esophageal neoplasms may arise from independent clones of transformed cells. The timing of p53 protein accumulation is favorable for an intermediate biomarker in multistage esophageal carcinogenesis.

Twenty cell lines from 17 individuals with malignant mesothelioma have been examined for p53 alterations by direct sequencing of genomic DNA, by evaluation of mRNA expression levels, and by immunocytochemical analysis of p53 protein expression in comparison with normal human pleural mesothelial cells. The results of this study show p53 abnormalities in cell lines from 3 individuals. These include 2 point mutations and one null cell line. Interestingly, while both cell lines with point mutations exhibit high levels of p53 protein, normal mesothelial cells as well as 12 of the mesotheliomas evaluated express low but significant levels. In addition, sequencing of K-ras at codons 12, 13, and 61 reveals wild-type sequence in all 20 mesothelioma cell lines. The capacity to induce tumors in athymic nude mice did not correlate with the presence of a p53 mutation or elevated p53 protein levels. These data suggest that neither p53 alteration nor K-ras activation constitutes a critical step in the development of human mesothelioma.