Construction of sigG mutant strains. (a) A total of 691 bp internal to sigG was deleted by allelic recombination and replaced with a 1.6-kb hygromycin cassette. The schematic shows the locations of the regions replaced within sigG (dotted lines) and the DNA binding regions 2.4 (R2) and 4.2 (R4) (black boxes). The positions of the probes (white boxes below lines) and XmnI restriction sites used in Southern blot analysis are indicated, along with the expected fragment sizes detected in the wild-type and mutant strains (double-headed arrows). (b and c) Genomic DNA was extracted from potential ΔsigG mutant colonies (lanes 1 to 4) and Southern blot analysis of the genomic DNA performed alongside wild-type DNA analysis (lanes WT) after digestion was performed with XmnI and either a radiolabeled probe (b) or a horseradish peroxidase-labeled probe (c) used to detect potential double crossovers. (b) ΔsigG1 mutant construction; colonies 1, 2, and 4 showed the correct genotype. (c) ΔsigG2 mutant construction; all 4 colonies showed the correct genotype.

SigG does not effect expression of LexA. (a) The upstream region of lexA, showing the LexA start codon (italics), the transcriptional start site (dashed box), the SigA promoter motif (bold, solid underline), the SigG promoter motif (bold, dotted underline), and the SOS box (shaded). (b) β-Galactosidase activity of a lacZ transcriptional fusion to the wild-type lexA promoter region (pEMD8) or of that seen with a mutation in the sigA consensus promoter (pEJ554). Activity of the wild-type lexA promoter was increased 3- to 4-fold following induction by 0.02 μg ml⁻¹ mitomycin C for 24 h in both the wild-type (WT) and ΔsigG strains of M. tuberculosis, with no significant difference between the two strains in induced expression levels (t test; P > 0.05). Mutation of the potential SigA promoter resulted in a dramatic reduction in activity of the lexA promoter and no induction by mitomycin C. Data represent means + standard deviations of the results determined with at least three different biological replicates.

Induction of sigG in response to different DNA-damaging agents. qRT-PCR was used to determine expression of sigG in response to exposure to the following DNA-damaging agents under the indicated conditions: mitomycin C (MC), 0.2 μg ml⁻¹ for 5 h at 37°C; bleomycin (BM), 1.5 μg ml⁻¹ for 5 h at 37°C; ofloxacin (OX), 7.5 μg ml⁻¹ for 5 h at 37°C; and UV irradiation, 23.5 J m⁻² followed by 5 h of recovery at 37°C. Each data set represents the means + standard deviations of the results determined with three biological replicates. Alternatively, the response to cell surface stress was determined by exposure to SDS (0.05% for 5 h at room temperature) or ethambutol (Eth) (15 μg ml⁻¹ for 24 h at 37°C); for these determinations, each data set represents the means + standard deviations of the results determined with two biological replicates. Expression levels were determined and normalized to rrs expression, and induction ratios were calculated relative to untreated control results. The dotted line indicates an induction ratio of 1 (signifying lack of induction). sigG expression was induced under DNA-damaging conditions in both wild-type and ΔrecA strains of M. tuberculosis but was not induced by the cell surface stresses caused by the presence of SDS or ethambutol.

Expression of selected DNA repair genes in ΔsigG compared to wild-type M. tuberculosis strains in response to mitomycin C induction. Expression levels were determined by qRT-PCR after treatment with 0.2 μg ml⁻¹ mitomycin C for 24 h and normalized to rrs expression, and induction ratios were calculated relative to untreated control results. The dotted line indicates an induction ratio of 1 (signifying no induction). Data represent means + standard deviations of the results obtained with three biological replicates. There were no significant differences in the levels of induction of any of these genes in the two strains.

Susceptibility of the sigG mutant to UV irradiation. CFUs for wild-type and mutant strains were determined after exposure to different doses of UV and compared to untreated control results. The UV-sensitive ΔuvrB strain is shown as a control. Data represent means + standard deviations of the results obtained with three different biological replicates.