Quantitation of the adaptive response to alkylating agents

Abstract

ESCHERICHIA COLI can acquire resistance to the lethal and mutagenic effects of alkylating agents such as N-methyl-N′-nitro-nitrosoguanidine (MNNG)¹. We have called this the adaptive response and have shown that it is distinct from previously described forms of DNA repair, including the inducible error-prone pathway^2,3. The induction of a similar response has recently been demonstrated independently in the liver cells of animals fed dimethylnitrosamine (DMN)⁴. The phenomenon is therefore apparently widespread and may conceivably protect us from mutagenesis and carcinogenesis induced by low levels of nitrosamines. The main mutagenic lesion produced by agents like MNNG and DMN is probably O⁶-methylguanine (O⁶-MeG)⁵, which can be read as adenine during DNA replication⁶ and transcription⁷; certainly, most of the mutations produced by MNNG are GC to AT transitions⁸. Therefore, it was not surprising to find that adapted bacteria contain less O⁶-MeG after a challenge dose of ³H-MNNG than do non-adapted bacteria⁹. However, the response seemed to have two distinct components, which we can call the fast and the slow reactions. Immediately after a 2-min exposure to ³H-MNNG the adapted bacteria contain less O⁶-MeG, as if some fast reaction were protecting them from accumulating alkylation at the O⁶ positions; there is, however, a limit to the capacity of this part of the response, and at higher concentrations of ³H-MNNG (or after longer periods of exposure) the adapted bacteria start to accumulate O⁶-MeG (and mutations) at the same rate as the non-adapted bacteria. This suggests that the molecules which carry out the fast reaction are used up during their interaction with O⁶-MeG (that is, each can function only once). During subsequent growth in the absence of MNNG, both adapted and non-adapted bacteria lose their O⁶-MeG, but the process is completed much sooner in the adapted bacteria; thus, there is apparently a slow component to the adaptive response. This separation of the response into two components⁹ was supported by the observation that the fast reaction occurs at 0 °C whereas the slow reaction does not (P. F. Schendel and P.R., unpublished results). However, further experiments reported here indicate that these two reactions are different manifestations of a single, rather unusual form of repair.