Prof. Maurice S Fox

The maintenance of the genetic continuity of any organism requires an efficient process of DNA surveillance and repair. Cellular DNA is subject to damaging events and, during replication, subject to copying errors. If such lesions or mistakes were to persist, mutations would accumulate and compromise cell functions. In higher organisms mutations have been shown to play a role in tumor etiology.

The risks of flaws in the fidelity of transmission of DNA sequence from parent to progeny is prominently reduced by the activity of mutation avoidance systems. These systems play roles in the elimination of replication misincorporated nucleotides and in removal of chemically modified nucleotides. Some of the systems that have been characterized in E. coli include l) the methyl directed mismatch repair system, dedicated to the removal of misincorporated nucleotides in a newly replicated DNA strand, 2) the vsr system, dedicated to the removal of thymine in T:G base pairs, that are probably the consequence of deamination of methylated cytosine in duplex DNA, 3) the GO system, involving mutY, mutT, andfpg that act to avoid the consequences of the 7,8dihydro-8oxoguanine (8-oxodG) product of guanine oxidation either in duplex DNA or in the nucleotide DNA precursor pools. Since 8-oxodG is evident in all cells so far examined, it is reasonable to infer that there must exist enzymatic activities capable of dealing with the mutagenic consequences of incorporation of this base analogue. Indeed, there is biochemical evidence for functional analogs of the GO system components in yeast and mammalian cells. We have developed a screen to detect cDNA from eucaryotic cells capable of complementing bacterial mutants in this system. We have also isolated, through sequence homology to the yeast gene OGG 1, (coding for an 8-oxodG specific glycosylase), a human cDNA that, when expressed in E. coli, is able to complement an fpg mutant. We are now in the process of mapping the chromosomal locus and studying the biochemical function encoded by the cDNA.

Several lines of evidence show that superimposed on these systems are other physiological features such as replication or transcription that can modify observed mutation frequencies. We have shown that the frequency of occurrence of certain sequence changes is influenced by whether the target sequence resides on the bacterial chromosome or on an F' plasmid. Much of the work, investigating the roles of mutation avoidance functions on specific sequence changes, has focused on mutation events in loci residing on F' plasmids. Our studies of the post plating revertants of a lactose frameshift mutation located in the episome, i.e. adaptive mutations, have shown that about 6% of the colonies appearing after prolonged selection display an unstable lactose phenotype when grown under non selective conditions. Molecular analysis of these mutants showed that the region of the episome that includes the lac coding sequence is amplified. Tandem repeats of 15 to over 30kb are present including up to 50 copies. This rearrangement seems linked not only to the unstable lac phenotype but also to the presence of deletions in nearby loci. Since this lac allele is leaky, this is consistent with a model in which at least a fraction of the post plating revertants are formed through the amplification of the allele, the accumulation of multiple copies of the allele allowing growth on lactose and increasing the likelihood of a nucleotide change mutation to a bona fide lac+ sequence.

We have examined the lac+ revertants that have been associated with episome transfer to a scavenger background. When the episome harbors a drug resistance bearing transposon e.g. TnlO, we have found that a large fraction of the late appearing lac+ revertants have lost the transposon marker and that among those products there appears to be a substantial enrichment of revertants with amplification of the lac region. This enrichment is independent of the location of the transposon with respect to lac. These results suggest an array of different events taking place possibly during the single stranded stage of the DNA being transferred to the recipient cell. Due to starvation, DNA replication of the single stranded DNA may be slow, allowing the formation of stem and loop structures in the transposon and, leading to its loss.

Much of the work on the investigation of mutagenesis include measurement of mutation rates. Fluctuation analysis of the frequencies of mutants detected by selection in independently derived cultures is the conventional method used by Luria and Delbruck to demonstrate that mutant cells were present in cultures prior to the imposition of lethal selection. The question of the impact of selection on the detection of mutants is evidently more complicated when the selection is not lethal. We have examined the reversion of defined mutations in the lac gene of E. coli and shown that under selection conditions in which the Lac-bacteria cannot grow, many of the mutant bacteria that first appear as colonies on the selective plate were not present in the culture deposited on the plate. It would appear therefore that even for the earliest detected mutants in the culture, a substantial fraction of the mutations are the result of events occurring after selection has been imposed.