The oxidative DNA glycosylases of Mycobacterium tuberculosis exhibit different substrate preferences from their Escherichia coli counterparts

Abstract

The DNA glycosylases that remove oxidized DNA bases fall into two general families: the Fpg/Nei family and the Nth superfamily. Based on protein sequence alignments, we identified four putative Fpg/Nei family members, as well as a putative Nth protein in Mycobacterium tuberculosis H37Rv. All four Fpg/Nei proteins were successfully overexpressed using a bicistronic vector created in our laboratory. The MtuNth protein was also overexpressed in soluble form. The substrate specificities of the purified enzymes were characterized in vitro with oligodeoxynucleotide substrates containing single lesions. Some were further characterized by gas chromatography/mass spectrometry (GC/MS) analysis of products released from gamma-irradiated DNA. MtuFpg1 has substrate specificity similar to that of EcoFpg. Both EcoFpg and MtuFpg1 are more efficient at removing spiroiminodihydantoin (Sp) than 7,8-dihydro-8-oxoguanine (8-oxoG). However, MtuFpg1 shows a substantially increased opposite base discrimination compared to EcoFpg. MtuFpg2 contains only the C-terminal domain of an Fpg protein and has no detectable DNA binding activity or DNA glycosylase/lyase activity and thus appears to be a pseudogene. MtuNei1 recognizes oxidized pyrimidines on both double-stranded and single-stranded DNA and exhibits uracil DNA glycosylase activity. MtuNth recognizes a variety of oxidized bases, including urea, 5,6-dihydrouracil (DHU), 5-hydroxyuracil (5-OHU), 5-hydroxycytosine (5-OHC) and methylhydantoin (MeHyd). Both MtuNei1 and MtuNth excise thymine glycol (Tg); however, MtuNei1 strongly prefers the (5R) isomers, whereas MtuNth recognizes only the (5S) isomers. MtuNei2 did not demonstrate activity in vitro as a recombinant protein, but like MtuNei1 when expressed in Escherichia coli, it decreased the spontaneous mutation frequency of both the fpg mutY nei triple and nei nth double mutants, suggesting that MtuNei2 is functionally active in vivo recognizing both guanine and cytosine oxidation products. The kinetic parameters of the MtuFpg1, MtuNei1 and MtuNth proteins on selected substrates were also determined and compared to those of their E. coli homologs.

Fig. 1

DNA glycosylase/lyase activity of MtuFpg1 on double-stranded DNA substrates. Labeled substrate, 15 nM, was incubated with either no enzyme (−), 15 nM EcoFpg or increasing amounts of MtuFpg1 (1.5 nM, 15 nM or 150 nM).

Fig. 2

Enzymatic activities of MtuFpg1 (●) and EcoFpg (○) on 8-oxoG:A. Labeled 8-oxoG:A, 15 nM, was incubated with increasing amounts of either MtuFpg1 or EcoFpg at 37 °C for 30 min. The data represent the means of three independent experiments, the uncertainties are standard deviations.

Fig. 3

(A) DNA glycosylase/lyase activities of MtuNth and MtuNei1 on double-stranded DNA substrates. Labeled substrate, 15 nM, was incubated with no enzyme (−), 15 nM EcoNth, EcoNei or NEIL1, or increasing amounts of MtuNth or MtuNei1 (1.5 nM, 15 nM or 150 nM). (B) Uracil DNA glycosylase activity of MtuNei1. Double-stranded DNA substrate containing uracil, 15 nM, was incubated with no enzyme (−), 15 nM EcoFpg, MtuFpg1, EcoNth, MtuNth, EcoNei or NEIL1, or increasing amounts of MtuNei1 (1.5 nM, 15 nM or 150 nM).

Fig. 3

(A) DNA glycosylase/lyase activities of MtuNth and MtuNei1 on double-stranded DNA substrates. Labeled substrate, 15 nM, was incubated with no enzyme (−), 15 nM EcoNth, EcoNei or NEIL1, or increasing amounts of MtuNth or MtuNei1 (1.5 nM, 15 nM or 150 nM). (B) Uracil DNA glycosylase activity of MtuNei1. Double-stranded DNA substrate containing uracil, 15 nM, was incubated with no enzyme (−), 15 nM EcoFpg, MtuFpg1, EcoNth, MtuNth, EcoNei or NEIL1, or increasing amounts of MtuNei1 (1.5 nM, 15 nM or 150 nM).

Fig. 4

DNA glycosylase/lyase activity of MtuNei1 on single-stranded DNA substrates. (A) Labeled substrate, 15 nM, was incubated with no enzyme (−), 15 nM EcoFpg, MtuFpg, EcoNth, MtuNth, EcoNei or NEIL1, or MtuNei1. (B) The fraction cleaved was calculated for the reactions with 15 nM enzyme. The data represent the means of three independent experiments and the error bars indicate the standard deviations.

Fig. 4

DNA glycosylase/lyase activity of MtuNei1 on single-stranded DNA substrates. (A) Labeled substrate, 15 nM, was incubated with no enzyme (−), 15 nM EcoFpg, MtuFpg, EcoNth, MtuNth, EcoNei or NEIL1, or MtuNei1. (B) The fraction cleaved was calculated for the reactions with 15 nM enzyme. The data represent the means of three independent experiments and the error bars indicate the standard deviations.

Fig. 5

Activity of MtuNei1 on 5-OHU-containing (A) DNA bubble structures with 5, 11 or 19 unpaired bases and (B) 3′ forked structure. Labeled substrate, 15 nM, was incubated with 15 nM EcoNei, NEIL1 or MtuNei1. Cleaved fractions were calculated for each reaction. The double-stranded and single-stranded 51-mer substrates containing 5-OHU were used as controls for the reactions with the bubble structures. The double-stranded and single-stranded 35-mer substrates containing 5-OHU were controls for the reactions with the 3′ forked structure. The data represent the means of three independent experiments, the uncertainties are standard deviations.

Fig. 5

Activity of MtuNei1 on 5-OHU-containing (A) DNA bubble structures with 5, 11 or 19 unpaired bases and (B) 3′ forked structure. Labeled substrate, 15 nM, was incubated with 15 nM EcoNei, NEIL1 or MtuNei1. Cleaved fractions were calculated for each reaction. The double-stranded and single-stranded 51-mer substrates containing 5-OHU were used as controls for the reactions with the bubble structures. The double-stranded and single-stranded 35-mer substrates containing 5-OHU were controls for the reactions with the 3′ forked structure. The data represent the means of three independent experiments, the uncertainties are standard deviations.

Fig. 6

Stereospecificity of MtuNth and MtuNei1 on thymine glycol. Double-stranded oligodeoxynucleotide substrates, 15 nM, containing either 5R ((5R,6S), (5R,6R)) or 5S ((5S,6R), (5S,6S)) Tg isomers were incubated with either no enzyme (−) or increasing amounts of MtuNth, EcoNth, MtuNei1 and EcoNei (1.5 nM, 15 nM or 150 nM).

Fig. 7

Spontaneous mutation frequencies to rifampin resistance in E. coli wild type (W.T.), fpg mutY nei triple mutant, nei nth double mutant as well as the mutant strains expressing either EcoNei, MtuNei1 or MtuNei2. Mid-log-phase cultures were plated on LB agar with or without 100 μg/ml of rifampin. Mutation frequencies per 10⁸ cells were calculated. The data shown are the averages of the three or four independent experiments. The error bars indicate the standard deviations. * - significant difference (p < 0.05 by the student’s t test) compared to the mutation frequency of the fpg mutY nei triple mutant or the nei nth double mutant.

Fig. 7

Spontaneous mutation frequencies to rifampin resistance in E. coli wild type (W.T.), fpg mutY nei triple mutant, nei nth double mutant as well as the mutant strains expressing either EcoNei, MtuNei1 or MtuNei2. Mid-log-phase cultures were plated on LB agar with or without 100 μg/ml of rifampin. Mutation frequencies per 10⁸ cells were calculated. The data shown are the averages of the three or four independent experiments. The error bars indicate the standard deviations. * - significant difference (p < 0.05 by the student’s t test) compared to the mutation frequency of the fpg mutY nei triple mutant or the nei nth double mutant.