No other lacZ mutant showed a marked difference
No other lacZ mutant showed a marked difference in mutation rate between fpg and fpg backgrounds (that is, rates were comparable between HS101 and CSH1191, HS102 and CSH1192, etc.). The number of reversion mutants was much higher in HS1194 than in any other fpg strain. The inclusion of the pTRC99a plasmid did not affect the reversion rate (Fig. 1). We, like Palmer et al. , found very few reversion mutations in HS1193 and HS1196 (Table 2).
Table 2 describes the effects of OGG, FPG-1, and FPG-2 on the appearance of revertant colonies after five days on minimal medium containing lactose. Fig. 2A illustrates the results for the case in which OGG and FPG activity is expected. There were significantly fewer revertants in HS1194(pTRC99a::ogg) and HS1194(pTRC99a:fpg-1) than in HS1194(pTRC99a): the presence of either OGG or FPG-1 reduced the number of revertants by approximately 75%. We infer that the plant OGG and FPG were each effective in initiating the base-excision repair process by removing oxidized G bases, thus blocking 8-oxo-G:A FIPI pairing and consequent G→T transversions. In contrast, there was no evidence that FPG-2 had any significant effect in lowering the G→T reversion rate.
Although the numbers of mutants were not large, there was a consistent indication that the presence of OGG and FPG-1 reduced the number of revertants in HS1195 (AT→TA) and that FPG-1 (but not OGG) reduced the number of revertants in HS1192 (GC→AT)(Table 2). For HS1195, the numbers of colonies were sufficient to suggest a strong effect in lowering mutation rate (Fig. 2B). For HS1192, the number of revertant colonies was small and the effect of repair thus difficult to quantify.
To test the suggestion that OGG and FPG-1 blocked reversion in HS1192 and HS1195, we treated plated cells with UV radiation. The treatment reduced the viability of the cells by 95% with no substantial difference in survival among HS1191 strains containing the different plasmids (data not shown). The UV treatment increased the rates of reversion of HS1192(pTRC99a) and HS1195(pTRC99a) by factors of 10,000 and 325, respectively (compare Table 2, Table 3). With this increase, it became clear that both OGG and FPG-1 were effective in reducing reversion in both strains. To check whether the reversion of irradiated cells was equivalent to reversion of unirradiated cells, we extracted DNA from five revertant colonies each of HS1192(pTRC99a) and HS1195(pTRC99a) and determined their lacZ base sequences around the critical base (Fig. 3). In all cases, there was evidence for mutation to a triplet that would transcribe and translate to an active glu in the LacZ protein. In both HS1192 and HS1195, some colonies showed mixed bases at the critical position, indicating either that colonies contained both revertant and non-revertant cells or that cells contained both revertant and non-revertant plasmids, a finding not observed in previous studies involving spontaneous reversions . In one HS1192 revertant, two adjacent bases were mutated.
Cells of HS1191(pTRC99a:fpg-1) consistently showed a higher number of revertants than did HS1191(pTRC99a) or HS1191(pTRC99a:fpg-2) (Table 2, Table 3). The mutations were stable, in that they maintained their lacZ+ phenotype on repeated transfer to new media. However, sequence analysis of PCR fragments from the lacZ genes of five revertants showed the expected AT→CG mutations at position 461 in only three of the five (data not shown). The mutational changes leading to the other reversions are not known.
Discussion The results of these experiments confirm that both OGG and FPG-1 enzymes coded by their corresponding Arabidopsis genes are capable of reducing the frequency of G→T transversions. These results were predicted by earlier reports of their ability in vitro to cleave DNA containing 8-oxo-G (OGG: , ; FPG: , . 8-Oxo-G base-pairs equally well with A and C. To the degree that the A is not removed from an 8-oxo-G:A base-pair by the MutY gene product, adenine-DNA glycosylase, the next round of replication will yield a stable T:A base-pair (and an active β-galactosidase if this base-pair is in the appropriate position in the lacZ gene). OGG and FPG-1 reduce the frequency of G→T transversions by removing the 8-oxo-G and thus preventing the formation of the 8-oxo-G:A base-pair (Fig. 4).