Excision of uracil from DNA by hSMUG1 includes strand incision and processing

Abstract

Uracil arises in DNA by hydrolytic deamination of cytosine (C) and by erroneous incorporation of deoxyuridine monophosphate opposite adenine, where the former event is devastating by generation of C → thymine transitions. The base excision repair (BER) pathway replaces uracil by the correct base. In human cells two uracil-DNA glycosylases (UDGs) initiate BER by excising uracil from DNA; one is hSMUG1 (human single-strand-selective mono-functional UDG). We report that repair initiation by hSMUG1 involves strand incision at the uracil site resulting in a 3'-α,β-unsaturated aldehyde designated uracil-DNA incision product (UIP), and a 5'-phosphate. UIP is removed from the 3'-end by human apurinic/apyrimidinic (AP) endonuclease 1 preparing for single-nucleotide insertion. hSMUG1 also incises DNA or processes UIP to a 3'-phosphate designated uracil-DNA processing product (UPP). UIP and UPP were indirectly identified and quantified by polyacrylamide gel electrophoresis and chemically characterised by matrix-assisted laser desorption/ionisation time-of-flight mass-spectrometric analysis of DNA from enzyme reactions using 18O- or 16O-water. The formation of UIP accords with an elimination (E2) reaction where deprotonation of C2' occurs via the formation of a C1' enolate intermediate. A three-phase kinetic model explains rapid uracil excision in phase 1, slow unspecific enzyme adsorption/desorption to DNA in phase 2 and enzyme-dependent AP site incision in phase 3.

Figure 1.

Indication of hSMUG1 incision at uracil in DNA. (A) DNA substrate and conventional base excision assay. (B, C) Protein dependence of U-DNA incision (red) and uracil excision (blue). hSMUG1 was incubated with U-DNA (substrate 1, 0.5 pmol) in 20 mM Tris–HCl, pH 8.0, 1 mM DTT, 1 mM EDTA, 70 mM KCl at 37°C for 10 min. Each value in C represents the average (±SD) of three independent measurements. ‘U-DNA incision (total)’ corresponds to the values obtained from measuring the strength of the bands on the gel in B (lanes 4–7); the ‘U-DNA incision (enzymatic)’ values are calculated by subtracting the amount of AP site incision caused by the 5-min heat treatment at 95°C (as presented in Figure 2D) from the ‘U-DNA incision (total)’ values, where the number of AP sites formed by hSMUG1 equals the number of uracils excised as measured in parallel in B (lanes 8–10). Abbreviation: nt, nucleotides; UIP, U-DNA incision product; UPP, U-DNA processing product.

Figure 2.

Thermolysis of AP-DNA at high temperature efficiently forms UIP as opposed to UPP. (A) DNA substrate (see below) and assay. (B) Time dependence for cleavage of AP-DNA at 95°C. AP-DNA derived from substrate 1 (0.5 pmol) was treated with loading solution used in conventional denaturing PAGE [containing 80% (v/v) formamide]. UIP forms efficiently, while a smaller amount of UPP/δ-product appears at the longest incubation times. (C) Time dependence for cleavage of AP-DNA at different temperatures. AP-DNA derived from substrate 1 was used at 37°C (1 pmol) and 95°C (see B), while that used at 75°C (1 pmol) was derived from substrate 2 (see Materials and Methods). Each value represents the average (± SD) of 6–15 (95°C; red), 2–6 (75°C; orange) or 5–6 (37°C; dark grey) independent measurements. At 37°C, PAGE was performed on a 15% (w/v) gel containing 3% (v/v) formamide, and identical experiments with AP-DNA dissolved in pure water also showed no significant DNA cleavage (data not shown). UPP (green) was only formed at 95°C. (D) Time dependence for AP-DNA cleavage in different solutions at 95°C. Treatment in loading solution (red; described in B), water (blue) or TE buffer (violet) showed that the initial cleavage of AP-DNA is virtually identical in the different aqueous solutions. To separate incised DNA from un-incised DNA the reaction products were subjected to denaturing (red) or non-denaturing (blue; violet) PAGE. Each value represents the average (± SD) of 4–17 independent measurements, where the slopes of the graphs for the initial DNA incision, i.e. the first three data points (6–17 independent measurements; red, y = 3.95x + 0.769, R = 0.999; blue, y = 3.92x + 9.29, R = 0.998; violet, y = 3.65x + 27.673, R = 0.999) yield the non-enzymatic incision per min. This amounted to 3.95% of the AP sites incised per min, resulting in a background of 19.8% non-enzymatic hydrolysis (as calculated from the red graph; for the 5 min formamide/heat treatment) for the experiment described in Figure 1B and C. The amount of background incision was subtracted giving the value for enzymatic U-DNA incision for all experiments using 5 min heat treatment at 95°C (Figure 1C). (E) Time dependence for AP-DNA cleavage in different solutions at 75°C. AP-DNA (substrate 2, 1 pmol) was exposed to loading solution (red) or water (blue). Each value represents the average (±SD) of 6 (at 2–20 min) or 2–3 (at 30 min) independent measurements. To separate incised DNA from un-incised DNA the reaction products were subjected to denaturing (red) or non-denaturing (blue) PAGE. The initial slopes of the graphs (red, y = 0.722x + 3.65, R = 0.986; blue, y = 0.755x + 7.68, R = 0.977) yield the non-enzymatic incision per min. This amounted to 0.722% of the AP sites incised per min in the formamide solution. Abbreviation: δ, β/δ-elimination product.

Figure 3.

Indirect identification of UIP and UPP by electrophoretic mobility using conventional denaturing conditions. (A, B) Time dependence of UIP (red) and UPP (green) formation by hSMUG1. hSMUG1 (0.3 pmol) was incubated with substrate 1[³²P] (0.12 pmol) in 20 mM Tris-HCl, pH 8.0, 1 mM DTT, 1 mM EDTA, 70 mM KCl at 37°C. To define the different 3′-end products, substrate was incubated with either EcNth (8.7 pmol), EcNfo (0.16 pmol), EcFpg (17 pmol) or hOGG1 (13 pmol) together with EcUng (0.78 pmol) for 10 min. Incised was separated from un-incised DNA by denaturing PAGE. Each value in B represents the average (±SD) of three independent measurements.

Figure 4.

hSMUG1 incises at uracil in DNA. (A) DNA substrate and assay. (B, C) Protein dependence of U-DNA incision (red) and uracil excision (blue). hSMUG1 was incubated with U-DNA (substrate 1, 1 pmol) at 37°C for 10 min. Each value in C represents the average (±SD) of 3–6 independent measurements. Incision product was separated from un-incised DNA by PAGE at 115 V for 1.5 h using a 20% (w/v) gel with 3% (v/v) formamide. (D) hSMUG1(25–270) was incubated with U-DNA (1 pmol of substrate 1; see A) at 37°C for 20 min. Incision product was separated from un-incised DNA by PAGE at 120 V for 2 h using a 20% (w/v) gel with 3% (v/v) formamide. (E) Protein dependence of U-DNA incision/processing (red) and uracil excision (blue). Each value represents the average (± SD) of 4–5 independent measurements as described in D. (F) U-DNA incision by hSMUG1 in different buffers. U-DNA (1 pmol of substrate 1) was incubated with 1 pmol of hSMUG1(25–270) or without enzyme as control in reaction buffer (HEPES), or in 45 mM sodium cacodylate with the same pH and additions as for reaction buffer (see Materials and Methods), at 37°C for 10 min (final volume, 20 μl). Incision product was separated from un-incised DNA by PAGE as described in E. Each value represents the average (±SD) of three independent measurements.

Figure 5.

hSMUG1 incises at uracil in ssDNA. (A, B) Protein dependence of U-DNA incision (red) and uracil excision (blue). hSMUG1 was incubated with ssU-DNA (1 pmol; the labelled strand of substrate 1) at 37°C for 10 min. Each value in B represents the average of 2 independent measurements. Incision product was separated from un-incised DNA by PAGE at 100 V for 50 min using a 12% (w/v) gel with 3% (v/v) formamide.

Figure 6.

Trapping experiments for Schiff base intermediate. Left panel, EcFpg (17 pmol) alone as a negative control, and together with EcUng (3 pmol) as a positive control, EcUng as well as hUNG (5 pmol) alone as negative controls, and hSMUG1 (0.3 pmol) alone, were incubated with substrate 2 (1 pmol) and 50 mM NaBH₄ in reaction buffer at 37°C for 1 h (final volume, 10 μl). Right panel, EcFpg (10 pmol) alone as a negative control, and together with EcUng (10 pmol) as a positive control, EcUng as well as hUNG (10 pmol) alone as negative controls, and hSMUG1 (10 pmol) alone, were incubated with substrate 2 (1 pmol) and 50 mM NaBH₄ in reaction buffer at 37°C for 1 h (final volume, 10 μl). In each case (A and B), trapped was separated from un-trapped substrate by denaturing PAGE [10% (w/v)] at 200 V for 1 h. The experiments were performed in triplicate showing the same result.

Figure 7.

Indirect identification of UIP by electrophoretic mobility without exposure of DNA to high temperature. U-DNA (substrate 1, 1 pmol) was incubated with hSMUG1 (0.3 pmol) at 37°C for 30 min; either alone or together with EcFpg (4 pmol) as indicated. To define the different 3′-end products, substrate was incubated with hUNG (1 pmol) together with either EcFpg (4 pmol), hAPE1 (0.45 pmol) or EcNth (1 pmol), as indicated, under the same conditions. Incubations were also performed with either substrate 1 (dsDNA; lane 2) or the labelled strand of substrate 1 (ssDNA; lane 1) alone, showing that the upper substrate band is ssDNA and the lower band dsDNA. Incision product was separated from un-incised DNA by PAGE at 300 V for 5 h using a 20% (w/v) gel with 7 M urea.

Figure 8.

Chemical identification of UIP and UPP and working model for reaction mechanism causing DNA incision. (A) Proposed E2 elimination reaction for the formation of UIP and chemical identification of UIP and UPP by MALDI-TOF-MS (see Supplementary Data, Figure S3 for MALDI-TOF-MS controls). hSMUG1 amino acid residue(s) suggested being involved in catalysis are coloured green; their hydrogen bonds with catalytic water and substrate are shown by red dotted lines. Proposed electronic and proton transfers involved in the formation of UIP are indicated by blue arrows. In the case of UPP, no reaction mechanism is proposed, and it is still unclear whether it is formed directly as a result of incision or by processing of UIP as depicted here. (B) Confirmation of the chemical nature of UIP. The observed post-enzymatic addition of water (left) or ammonia (middle and right) can be explained by the presence of a conjugated double bond, while the efficient exchange of an oxygen atom when the sample was transferred between ¹⁸O- and ¹⁶O-water can be explained by the presence of an aldehyde group. The MALDI-TOF-MS signals of the different chemical structures are shown in the upper and lower panels in A, and in the lower panel in B.

Figure 9.

hSMUG1 kinetics. (A) Three-phase kinetic model. Phase 1 is shown in blue, phase 2 in violet and phase 3 in red. The uracil excision step is rapid compared to the slow DNA incision step. (B) U-DNA incision rate Vⁱⁿ and (C) uracil excision rate V^ex (see A) as a function of enzyme concentration [E]₀ at an initial U-DNA concentration [S]₀ of 50 nM, where the corresponding time-dependent data in the range [E]₀ = 0.05–0.25 nM (red line) is presented in (D) showing that at higher initial enzyme (E) concentration the model predicts that the formation of the incision product P1 has linear time-dependent kinetics, and in (E), showing that the excision kinetics for U (blue line) are fast and correlate with the removal of substrate DNA (S; black line), respectively. Incubation was performed for 20 min as described in Figure 4B. The Vⁱⁿ in the blue area changes as a result of increased unspecific binding of enzyme to DNA. In the yellow area, the unspecific binding is saturated and the Vⁱⁿ follows Michaelis–Menten (MM) kinetics. Each value represents the average (±SD) of 3–6 independent measurements.

Figure 10.

Proposed steps in the human BER pathway after SMUG1 has targeted uracil in DNA. After uracil has been removed by the DNA glycosylase activity of SMUG1 (step 1; blue), the latter is either replaced by APE1 (dark red) which incises the AP site (step 2a), or SMUG1 itself incises the AP site (step 2b; red) leaving behind a 3′-α,β-unsaturated aldehyde (UIP) which can be removed by APE1 (step 3b). Further processing of UIP (or maybe an alternative type of incision of the AP site; green broken arrows) results in a 3′-phosphate (UPP) which is a substrate for PNKP (orange). The cleaned one nucleotide gap in DNA is now ready for insertion of the correct dCMP (step 4) by the repair DNA polymerase β (Pol β; dark blue), which also exhibits the dRP lyase activity which removes the 5′-dRP remnant (step 3a) after APE1 incision. BER is concluded by nick-sealing (step 5) by DNA ligase III (LIG3; purple). The residues removed are indicated in dark red; those resulting from replacement in dark blue; dR, deoxyribose.