Daphna Sagher

The role of DNA polymerase in base substitution mutagenesis on non-instructional templates

In vitro DNA synthesis on phi X174 or M13 templates with non-instructional lesions such as UV dimers or AP (apurinic/apyrimidinic) sites terminates one base before the site of the lesion when synthesis is catalyzed by T4 DNA polymerase or E. coli polymerase I. E. Coli polymerase I also produces termination bands at the site of AP lesions. Substitution of Mn2+ for Mg2+ and increasing the concentration of dNTPs results in elongation of the newly synthesized strand opposite the site of the lesion and beyond. Purine deoxynucleoside triphosphates are utilized for insertion opposite lesions to a greater extent than are pyrimidine deoxynucleoside triphosphates. Deoxy ATP is used almost exclusively for elongation opposite AP sites with pol I-Klenow fragment in the presence of Mg2+. We suppose that these results illustrate the previously observed greater affinity of polymerases under template-free conditions for purine nucleotides. We also suppose that the results can be used to account for mutagenic base selection on noninstructional DNA templates. If purines are preferentially selected by polymerases, then treatments which inactivate pyrimidines will lead to an excess of transitions whereas inactivation of purines will produce more transversions. Data in the literature support this hypothesis.

Stabilization of the intermediate in frameshift mutation.

A mismatch repair, proofreading deficient mutant of Escherichia coli lost a C from a C8 run at a rate 10 times higher than the loss of A from an A8 sequence in the same double mutant. This greater frameshift instability of a homopolymeric run of Cs may be due to stabilization of a stacked intermediate. Gain of a (CA) unit in a similarly constructed (CA)15 sequence occurred at a rate about 1/3 that previously reported for a (CA)14 construct losing a (CA) repeat unit.

Mutagenic Consequences of the Alteration of DNA by Chemicals and Radiation

The induction of mutations in vivo is a process that involves the interaction of exogenous agents, the Biologically Reactive Intermediates (BRIs), with particular nucleotides in the DNA. The finding of “hot spots” (see Hsia et al., 1989) and of “mutator” strains of organisms (Modrich, 1987) indicates that superimposed on the primary interaction of BRIs and nucleotides is an effect of DNA sequence (Burns et al., 1987) and of the proteins involved in replication and in the monitoring of the DNA. Many of the primary interactions of BRIs with DNA result in alterations which block DNA synthesis, at least in vitro. It appears to be a truism that mutation, at least point mutation as a result of damage induced by an agent which inhibits DNA synthesis, requires that the DNA synthetic machinery should bypass the damage by some mechanism. An understanding of the phenomena of mutation therefore requires knowledge of the relationships between the altered DNA bases, the arrest of DNA synthesis, and the location of the damage within the DNA sequence. For example, one might assume that the sites most subject to modification by BRIs are those at which mutation occurs most readily. In fact, it appears that this simplest of hypotheses is not inevitably so, and that other factors may intervene (Brash et al., 1987).

Methyl Transferase Activity in Secondary Leukemia

During a typical course of treatment for Hodgkin’s disease or non-Hodgkin’s lymphoma with MOPP, individuals receive approximately 1400 mg/m2 of procarbazine over a period of 14 days (DeVita et al., 1970). Procarbazine is metabolized and one of its products is a methylating ion which produces O-alkylation products including O6-methylguanine (Wiestler et al., 1984). Most organisms have a protein which removes 06-methylguanine in a stoichiometric manner (Samson and Cairns, 1977; Pegg, 1983). In bacteria, this protein is adaptive so that exposure to methylating agents increases the amount of O6-alkylguanine DNA alkyltransferase (AGT) over a hundred-fold in less than a cell generation (Lindahl and Sedgwick, 1988). It is still not clear whether the formation of protein responds adaptively to methylation damage in mammals. In rats there does seem to be a nonspecific response of AGT production to damage (Pegg, 1983). If there is such an effect in humans, the effect is not rapid. As with other repair processes, the AGT content of peripheral blood lymphocytes (PBLs) from individuals varies (Waldstein et al., 1982; Sagher et al., 1988). If this variation were idiosyncratic and persisted, then individual response to procarbazine might be an etiologic factor in the development of secondary cancer.

Heterogeneity in the O6 Alkylguanine DNA Alkyltransferase (AGT) Activity of Human Peripheral Blood Lymphocytes (PBL’s)

The nature of the mechanisms controlling the protein content of mammalian cells and tissues is a central problem of modern biology. Recent studies on the structure of genes and their control elements in eukaryotes indicate the early paradigms based on the lac operon of E. coli (complex as they are) to be much too simple (Watson, et al., 1987). It is likely that numerous factors are involved in the determination of the level of activity of any protein, particularly proteins which are developmentally regulated so that their activity differs in different tissues. The repair protein, O6-alkylguanine DNA alkyltransferase (AGT) is an example of a developmentally regulated protein since it has widely different activities in brain, liver and other tissues (Pegg, 1983). In addition, it has been reported that the peripheral blood lymphocytes (PBL’s) from different individuals may vary widely in their AGT activity (Waldstein, et al., 1982). In this paper, we report that average AGT activity is characteristic of individuals and of their cells but that the PBL’s constitute a heterogeneous population of cells of differing inherent AGT activity. Furthermore, we show that cells may vary in their response to the mutagen N-methyl-N′-nitro-N-nitrosoguanidine by a mechanism which need not involve the AGT protein.

The Response of Human Cells to In Vivo Methylation Damage

Individuals with Hodgkin’s disease (HD) may be treated by a chemotherapeutic regimen which includes procarbazine, a compound metabolized to a methylating species. A small group of all HD patients eventually develop acute nonlymphocytic leukemia (ANLL) some years after therapy (t-ANLL). We therefore initiated a study of the O6− methylguanine DNA methyltransferase (MT) activity in normal controls and in patients with HD, ANLL de novo or t-ANLL. As part of the study we also prepared lymphoblastoid lines by Epstein Barr virus (EBV) transformation of peripheral blood lymphocytes (PBL’s). We express MT activity per gg of DNA. The activity in normal individuals varies from about 2.2 to 14.7 fmol/µg of DNA with a mean ± SE of 7.2 ± 0.35. HD patients before treatment have MT values of 5.6 ± 0.53. The MT value of HD patients receiving procarbazine was 4.3 ± 0.52; t-ANLL patients before treatment gave 4.2 ± 0.63; six recently diagnosed ANLL patients had MT values of 7.8 ± 1.72, and 13 ANLL patients in remission had MT activities of 9.3 ± 2.3. The differences in MT value between normal subjects and HD patients, between t-ANLL and normal subjects, and between ANLL de novo and t-ANLL groups are significant. There is significant variation in the MT values of normal PBL’s sampled at different times. MT activity diminishes slightly with age in both normal and HD groups, but this does not account for the observed differences between normal and HD groups. We find a significant correlation between the MT activity of PBL’s and of the lines derived from them, particularly when normal or untreated individuals are used as the source of the lines. There is variation in the the MT values of successive lines from the same individual, but it is within a factor of about two in these experiments. Although there is significant variation in repeated samples of PBL’s from the same individual, patients with Hodgkin’s disease have significantly lower MT activity than do normal controls. In addition, the HD group includes individuals with almost no MT activity. The observation that individuals with t-ANLL have lower MT activity than either controls or ANLL patients de novo and that HD patients on procarbazine have reduced levels as compared to HD patients before treatment would, if confirmed, lend credence to the hypothesis that MT levels play a role in the etiology of secondary malignancy. There is a clear relationship between the MT level in lymphocytes and in the lines derived from them. We interpret these results to mean that the MT level of a cell is a characteristic which survives the events of EBV transformation.