Increased error frequency of DNA polymerases from senescent human fibroblasts

Abstract

The error catastrophe theory of ageing predicts that the fidelity of DNA polymerase should be reduced in extracts from senescent cells. This prediction has been experimentally verified with MRC-5 human diploid fibroblasts. Using “cytoplasmic” DNA polymerase α the old/young ratio of error frequencies was 3.4 with the poly[d(A-T)]/dGTP/Mg 2+ system, 1.9 with poly[d(A-T)]/dGTP/Mn 2+ and 2.0 with poly[d(I-C)]/dTTP/Mg 2+. With DNA polymerase γ the old/young ratio was 3.8. The fidelity of DNA polymerase was examined at seven points during the life span of MRC-5 fibroblasts and the increase in error frequency with cell age was found to be statistically highly significant ( P < 0.001). By means of control experiments, artifactual explanations of the results can be eliminated. The close correlation between homologous and non-homologous DNA synthesis was demonstrated by following the time-course of the reaction, varying the enzyme concentration and by other means. The non-homologous dNMP incorporation was sensitive to DNase, but not RNase or protease treatment. No terminal transferase activity could be detected. The discrimination of the DNA polymerase from young and old extracts was constant over a wide range of homologous and non-homologous dNTP concentrations. Degradation of the products of the misincorporation assay to 3′-dNMPs revealed that a guanine-thymine base mispairing was the main method of non-homologous synthesis in young and old cells while degradation to 5′-dNMPs revealed that dGTP was not contaminated by other homologous dNTPs. By use of DNA polymerase with a lower error frequency than MRC-5 polymerase, the contamination of poly[d(A-T)] by cytosine could be ruled out. Mixing experiments could not detect a diffusible agent in young or old extracts which was capable of modifying the error frequency of a DNA polymerase from another extract. Since the DNA polymerase extract was not pure, an alternative explanation is possible for these results. Nevertheless, these control experiments do strongly suggest that the misincorporation assay measures the frequency of DNA polymerase-directed errors which are present as single base substitutions.