Intact MutY and its catalytic domain differentially contact with A/8-oxoG-containing DNA

Abstract

Escherichia coli MutY is an adenine and a weak guanine DNA glycosylase active on DNA substrates containing A/G, A/8-oxoG, A/C or G/8-oxoG mismatches. A truncated form of MutY (M25, residues 1-226) retains catalytic activity; however, the C-terminal domain of MutY is required for specific binding to the 8-oxoG and is critical for mutation avoidance of oxidative damage. Using alkylation interference experiments, the determinants of the truncated and intact MutY were compared on A/8-oxoG-containing DNA. Several purines within the proximity of mismatched A/8-oxoG show differential contact by the truncated and intact MutY. Most importantly, methylation at the N7 position of the mismatched 8-oxoG and the N3 position of mismatched A interfere with intact MutY but not with M25 binding. The electrostatic contacts of MutY and M25 with the A/8-oxoG-containing DNA substrates are drastically different as shown by ethylation interference experiments. Five consecutive phosphate groups surrounding the 8-oxoG (one on the 3' side and four on the 5' side) interact with MutY but not with M25. The activities of the truncated and intact MutY are modulated differently by two minor groove-binding drugs, distamycin A and Hoechst 33258. Both distamycin A and Hoechst 33258 can inhibit, to a similar extent, the binding and glycosylase activities of MutY and M25 on A/G mismatch. However, binding and glycosylase activities on A/8-oxoG mismatch of intact MutY are inhibited to a lesser degree than those of M25. Overall, these results suggest that the C-terminal domain of MutY specifies additional contact sites on A/GO-containing DNA that are not found in MutY-A/G and M25-A/8-oxoG interactions.

Figure 1

A representative gel image by PhosphorImager of the alkylation interference of MutY and M25. The 40mer heteroduplex DNA containing an A/GO mismatch was 5′ end-labeled on the GO strand, partially methylated with dimethylsulfate (lanes 1–5) or ethylated with N-nitrosoethylurea (lanes 6–10), purified and bound to MutY(D138N) and M25(D138N). After binding, bound (B) and free (F) DNA were separated by 4% native polyacrylamide gel electrophoresis, eluted and chemically cleaved (32). Control samples were DNA treated by the same procedures without adding protein. Cleavage products were then analyzed on a 14% DNA sequencing gel and exposed to a PhosphorImager screen. Arrows mark the position of GO. Alkylation at the underlined nucleotides and backbone phosphates showed significant interference effect with MutY binding.

Figure 2

Quantitation of methylation interference effects. The 40mer heteroduplex DNA containing an A/GO mismatch (boxed) was 5′ end-labeled on the A-strand (top) or GO-strand (bottom). Numbers between the base pairs mark the positions of the nucleotides from the 5′ end of the A-strand. Interference effects on MutY and M25 are indicated by white and black bars, respectively. Results shown are from PhosphorImager quantitative analyses of gel images similar to Figure 1 from more than three experiments. The extent of interference is expressed as the ratio of the intensity of each band from the enzyme-free DNA to that of the corresponding band in the enzyme-bound samples. The ratios of free:bound at G₃₄ on the A-strand and G₈ on the GO-strand are normalized to 1.0. Error bars are ±1 standard deviation. No significant interference was observed for those regions of the 40 base pair DNA, not shown.

Figure 3

Quantitation of ethylation interference effects. The phosphate groups at the 5′ and 3′ sides of mismatched bases are labeled with positive and negative numbers, respectively, and the phosphates immediately 5′ to the mismatched bases are labeled as position 0. Interference effects on MutY and M25 are indicated by white and black bars, respectively. Interferences were quantitated and shown as described in Figure 2. The experiments were performed in triplicate. The ratio of free:bound of the phosphate at position 9 on the GO-containing strand and that of the phosphate at position –10 on the A-containing strand are normalized to 1.0. Error bars are ±1 standard deviation. The phosphate groups upstream of the seventh phosphodiester bond 5′ to the mismatched A may not significantly interfere with MutY binding due to different degrees of cleavage of enzyme-free and enzyme-bound DNA fractions.

Figure 4

Structures of the DNA minor groove binding drugs distamycin A and Hoechst 33258 used in this study.

Figure 5

Distamycin A and Hoechst 33258 inhibit the binding activities of MutY and M25 with A/G- and A/GO-containing DNA. Binding activities of MutY are marked by squares and solid lines, and those of M25 are marked by triangles and dotted lines. Relative MutY and M25 binding activities with A/G- (A and C) and A/GO-containing (B and D) DNA to that of no drug added samples were plotted as a function of drug concentrations. Data were from PhosphorImager quantitative analyses of gel images over three experiments.

Figure 6

Distamycin A and Hoechst 33258 inhibit the glycosylase activities of MutY and M25 with A/G- and A/GO-containing DNA. Squares and solid lines mark glycosylase activities of MutY, and that of M25 are marked by triangles and dotted lines. Relative MutY and M25 glycosylase activities with A/G- (A and C) and A/GO-containing (B and D) to that of no drug added samples were plotted as a function of drug concentrations. Data were from PhosphorImager quantitative analyses of gel images over three experiments.

Figure 7

Summary of the alkylation interference on specific complexes between MutY and M25 with DNA containing an A/G or A/GO mismatch. (A–C) MutY with A/G-containing DNA, MutY with A/GO-containing DNA and M25 with A/GO-containing DNA, respectively. The data in (A) are derived from Lu et al. (17). The methylation interference sites on purines are colored in blue for G_23, G₂₄ and G₂₅ in the major groove and colored in green for mismatched A and GO as well as A₁₈, A₁₉, A₂₀ and A₂₁ that are believed to be in the minor groove. Methylated purines with strong and weak interference effects on protein binding are in dark and light colors, respectively. Ethylated phosphate groups show strong and weak interference effects on protein binding are shaded in red and purple, respectively. The phosphate groups at the 5′ and 3′ sides of mismatched bases are labeled with positive and negative numbers, respectively, and the phosphates immediately 5′ to the mismatched bases are labeled as position 0. The sequence positions are labeled between base pairs with respect to the A-strand from the 5′ end of the 40mer DNA. GO is marked as O. The DNA shown is bent with the mismatched A flipped-out from the DNA helix according to the DNA structure in AlkA–DNA complex (34). Arrows mark the phosphates that interact with the HhH motif of MutY.

Figure 8

Spatial distributions of the contacts of MutY and M25 on the distorted DNA based on the coordinates of DNA from an AlkA–DNA co-crystal structure (34). The DNA shown here contains a 1-azaribose abasic base and has a different sequence than the one used in these experiments. (C and D) are viewed at 180° from (A and B) in which the minor groove is on the left and major groove on the right. The color codes are similar to those in Figure 7. The N3 group of mismatched A and the N7 group of mismatched GO (green blocks) are shown in (A) but are at the back of (C). The N7 groups of G₂₃ and G₂₄ (blue cylinders) as well as the N3 groups of A₂₀ and A₂₁ (green blocks) in the minor groove can be viewed in (C). Ethylated phosphate groups show strong, weak and no interference effects on protein binding are represented as red, purple and yellow spheres in different sizes, respectively. The 5′ and 3′ ends of the A-strand, as well as the phosphates that interact with the HhH motif of MutY, are marked.