Transcription-induced mutations: increase in C to T mutations in the nontranscribed strand during transcription in Escherichia coli

Abstract

Cytosines in single-stranded DNA deaminate to uracils at 140 times the rate for cytosines in double-stranded DNA. If resulting uracils are not replaced with cytosine, C to T mutations occur. These facts suggest that cellular processes such as transcription that create single-stranded DNA should promote C to T mutations. We tested this hypothesis with the Escherichia coli tac promoter and found that induction of transcription causes approximately 4-fold increase in the frequency of C to U or 5-methylcytosine to T deaminations in the nontranscribed strand. Excess mutations caused by C to U deaminations were reduced, but not eliminated, by uracil-DNA glycosylase. Similarly, mutations caused by 5-methylcytosine to T deaminations were only partially reduced by the very short-patch repair process in E.coli. These effects are unlikely to be caused by differential repair of the two strands, and our results suggest that all actively transcribed genes in E. coli should acquire more C to T mutations in the nontranscribed strand.

Figure 1

Predicted transcription patterns of pIQS11 and pIQS12. The expected dominant transcription patterns of pIQS11 (A) and pIQS12 (B) in the presence or absence of IPTG are shown. The figure is not drawn to scale, and exaggerates the size of the transcription bubble. Positions and directions of the tac and the kan promoters are shown. The strands of DNA are distinguished as being coding and noncoding, and the 5′ to 3′ directionality of the kan gene is indicated by an arrow below the DNA. The HindIII sites used to clone the kanS-D94 allele in pKK223-3 are marked. The sequence of the Dcm site at codon 94 of the gene is shown and the mutation that would restore Kan^R phenotype is indicated above it. The asterisk above the second cytosine within the Dcm site indicates of methylation of this base when Dcm is present in cells.

Figure 2

Amounts of transcripts produced from the two strands of pIQS11 and pIQS12. Amounts of RNAs calculated from analysis of PhosphorImager intensities of RNAs hybridized with two different probes are shown. The intensities are proportional to amounts of RNAs from the two strands. Because no effort was made to calculate the absolute RNA amounts from these numbers, the units for the x axis are marked as “arbitrary units.” Amounts of RNAs from the two strands are represented in separate histograms.

Figure 3

Effect of UDG-mediated repair on revertant frequency. Mean reversion frequencies in cultures from three independent colonies of GM31 or GM31ung containing pIQS11 are shown. The procedure used for this experiment was similar to that used for generating data presented in Table 1 and Fig. 3. Thus, cultures grown with and without IPTG originated from same bacterial colonies. The data have been regrouped and presented in a fashion that allows comparison of cultures with the same state of the tac promoter induction, but which were in different genetic backgrounds.

Figure 4

Effect of cytosine methylation on revertant frequency. Mean reversion frequencies in cultures from three independent colonies of GM31ung containing pIQS11 and pACYC184 or pDCM72 are shown. Experiments were carried out and the data processed in a manner similar to that described in legend to Fig. 3.

Figure 5

Effect of VSP repair on revertant frequency. Mean reversion frequencies in cultures from three independent colonies of GM31ung containing pIQS11 and pDCM72 or pDCM33 are shown. Experiments were carried out and the data processed in a manner similar to that described in legend to Fig. 3.