Selection-Enhanced Mutagenesis of lac Genes Is Due to Their Coamplification with dinB Encoding an Error-Prone DNA Polymerase

Abstract

To test whether growth limitation induces mutations, Cairns and Foster constructed an Escherichia coli strain whose mutant lac allele provides 1-2% of normal ability to use lactose. This strain cannot grow on lactose, but produces ∼50 Lac⁺ revertant colonies per 10⁸ plated cells over 5 days. About 80% of revertants carry a stable lac⁺ mutation made by the error-prone DinB polymerase, which may be induced during growth limitation; 10% of Lac⁺ revertants are stable but form without DinB; and the remaining 10% grow by amplifying their mutant lac allele and are unstably Lac⁺ Induced DinB mutagenesis has been explained in two ways: (1) upregulation of dinB expression in nongrowing cells ("stress-induced mutagenesis") or (2) selected local overreplication of the lac and dinB⁺ genes on lactose medium (selected amplification) in cells that are not dividing. Transcription of dinB is necessary but not sufficient for mutagenesis. Evidence is presented that DinB enhances reversion only when encoded somewhere on the F'lac plasmid that carries the mutant lac gene. A new model will propose that rare preexisting cells (1 in a 1000) have ∼10 copies of the F'lac plasmid, providing them with enough energy to divide, mate, and overreplicate their F'lac plasmid under selective conditions. In these clones, repeated replication of F'lac in nondividing cells directs opportunities for lac reversion and increases the copy number of the dinB⁺ gene. Amplification of dinB⁺ increases the error rate of replication and increases the number of lac⁺ revertants. Thus, reversion is enhanced in nondividing cells not by stress-induced mutagenesis, but by selected coamplification of the dinB and lac genes, both of which happen to lie on the F'lac plasmid.

Keywords: Escherichia coli; adaptive mutation; copy number variant; dinB; error-prone polymerase; gene amplification; lactose operon; local overreplication; mutagenesis; plasmid.

Figure 1

Deletions and insertions used to move dinB on the F’₁₂₈ plasmid. The two deletions described above (16 and 17 kb) were central to the strain constructions described later. Strain #1 is the original Cairns–Foster strain with dinB⁺ located 16 kb from lac (TR7178). Strain #2 has a 1.5-kb sequence (FRT Cm^R FRT) replacing the 16 kb between dinB and lac. Strain #4, with the 16-kb deletion, has dinB⁺ closest to lac, 180-bp away (TT27001). Strains #1, #2, #4, and #6 all show normal reversion under selection. The demonstrably functional dinB⁺ allele near Cm^R in strain #2 was used as donor for inserting the (Cm^R-FRT-dinB⁺) sequence into new sites. Strain #5 with the 17-kb dinB deletion (TT27002) received the (Cm^R-FRT-dinB⁺) fragment in place of its plasmid yebB gene to produce strain #6 (TT27282). Cm^R, chloramphenicol resistance cassette.

Figure 2

The relevant loci on chromosome and plasmid. This figure depicts the genotype of the standard Cairns–Foster strain (TR7178) and the changes introduced to test the effect of dinB⁺ position on reversion. A heavy line indicates the portion of the F’lac plasmid derived from the original F plasmid; the rest of the plasmid is derived from the chromosome of E. coli (Kofoid et al. 2003). The orientation of the lac and dinB genes is reversed by the way in which the plasmid is excised from the chromosome. Two deletions (16 and 17 kb) share an endpoint near lac on the plasmid, but extend different distances to either leave or remove the dinB⁺ gene. The insert is a PCR fragment from the 16-kb deletion strain, as described in Figure 1. The yebB gene is located at a maximum distance from lac in the tester F’₁₂₈lac plasmid. To make the isogenic set of strains used for reversion, these insertions and deletions were assembled into isogenic combinations by either transduction or conjugation. Cm^R, chloramphenicol resistance cassette.

Figure 3

MMS sensitivity of strains with dinB⁺ at various positions. Strains were plated on LB plates with or without 7.5 mM MMS by spotting 5 µl droplets of cultures that had been diluted 10⁻⁴, 10⁻⁵, or 10⁻⁶ fold. The first column from the left denotes line numbers referenced in the text as well as strain numbers (genotypes described in Table 1). The second and third columns describe dinB genotypes in the chromosome (Chrom.) and the F’₁₂₈ plasmid, using “+” to indicate dinB⁺ and “−” to indicate a dinB deletion. For the “+,” annotations in parentheses indicate where a new dinB allele was inserted, and no annotation means that the dinB gene is in its original position on the chromosome or F’₁₂₈ episome. For the “−,” annotations indicate what was inserted to replace the dinB gene. All strains tested for resistance to MMS were isogenic to E. coli K-12. The fourth and fifth columns show the effects of the MMS treatment on cell viability and growth compared to a control. The last column on the right indicates whether the strain is sensitive (S) or resistant (R) to MMS.

Figure 4

Effects of dinB⁺ gene position on reversion under selection. Isogenic strains were tested for accumulation of revertants under selection. Four isogenic strains have dinB⁺ alleles at various genomic positions. The top two strains above have a dinB⁺ allele on F’lac, and a chromosome with either a dinB⁺ allele (TR7178) or a dinB deletion formed by drug cassette replacement (TT27281). The bottom two strains have a F’lac plasmid whose dinB gene was removed by a 17-kb deletion. Their chromosome has either a functional dinB⁺ allele (TT27002) or a dinB deletion made by drug cassette replacement (TT27010). The lower graph shows the lawn populations, which were assessed by removing agar plugs from random spots on the selection plate. The results are presented for each strain individually and expressed relative to the cell number at the time of plating.

Figure 5

Effects of dinB position on reversion and on fraction of unstable revertants on day 5. (A) Four strains compared with a dinB⁺ allele at various positions in the F’lac plasmid. All strains have a dinB⁺ allele at the normal position in the chromosome (Chrom.). The strain (TR7178) has dinB⁺ located 16 kb from lac on the F’lac plasmid. The second strain (TT27001) has dinB⁺ moved nearer to lac (180-bp away). The third strain (TT27282) lacks the normal 17-kb dinB region but carries a functional dinB⁺ inserted at the opposite side of the F’lac plasmid, within the yebB gene (see Figure 2). The fourth strain (TT27002) has a plasmid with no dinB⁺ allele. The lower graph shows lawn cell population, which was assessed by removing plugs from random spots on the selection plate. The results are presented for each strain individually relative to the cell number at the time of plating. (B) The number of Lac+ revertants at day 5 and the percentage of revertants with an unstable or stable Lac⁺ phenotype for each strain of (A) (TT27001; TR7178; TT27282; and TT27002), and also strain TT27010 with dinB deleted from both the F’lac plasmid and the chromosome. The stability phenotype was assessed by restreaking Lac⁺ colonies at day 5 from the lactose plates onto rich medium with 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside, as described in the Materials and Methods.

Figure 6

Contribution to reversion of ectopic dinB⁺copies. The top three strains all carry an F’lac plasmid with a functional dinB⁺ gene. The first strain (TT27286) also has a dinB⁺ allele at its normal chromosomal location and an additional copy inserted into the chromosomal hisC gene. The second (TT27295) has only the dinB⁺ allele inserted into the hisC gene. The third strain (TT27290) carries the normal dinB⁺ and a Cm⁺ determinant inserted in hisC without a dinB⁺. Lower strains (TT27292, TT27296, and TT27291) all have an F’lac with a dinB deletion. These strains have one, the other, or both of the chromosomal dinB⁺ alleles. The lower graph shows the lawn population, which was assessed by removing plugs from random spots on the selection plate. The results are presented for each strain individually relative to the cell number at day 0. Lack of lawn growth shows that revertant accumulation is not a simple result of population growth.