| Nucleic Acids Research | Pages |
Evidence that MutY is a monofunctional glycosylase capable of forming a covalent Schiff base intermediate with substrate DNA
Introduction
Materials and Methods
Materials
MutY purification
Substrate preparation
Quantitative glycosylase/lyase assays
Qualitative assays for reduced enzyme-DNA intermediates
Results
Quantitative glycosylase/lyase assays
Qualitative borohydride reduction experiments
Discussion
A hypothesis for the active site of MutY
Acknowledgements
References
Note Added in Proof
Evidence that MutY is a monofunctional glycosylase capable of forming a covalent Schiff base intermediate with substrate DNA
ABSTRACT
INTRODUCTION
Intricate DNA repair pathways exist in all organisms to preserve the integrity of the genome (1). A major DNA repair pathway, base-excision repair (BER), utilizes DNA glycosylases for the recognition and removal of inappropriate bases by cleavage of the C1[prime]-N glycosyl bond. A wide variety of bases are removed by BER enzymes, including products of direct chemical modification, as well as normal DNA bases in mispair contexts. Currently, most BER enzymes belong to one of two subgroups: monofunctional (simple) glycosylases, or bifunctional glycosylase/lyases, which effect DNA strand scission at a rate equivalent to that of the glycosylase step (2). Figure Figure 1. (A) Proposed catalytic mechanism for AP site generation by monofunctional glycosylase activity (3). Nucleophilic attack at the C1[prime] carbon by an activated water molecule is facilitated by protonation of the leaving base (1.). The resulting AP site tautomerises between the ring-closed and ring-open forms (2.). [beta]- (3.) and [delta]-eliminations (4.) take place to effect DNA phosphodiester strand cleavage upon base treatment of the AP site. (B) Proposed catalytic mechanism for bifunctional glycosylase/AP-lyases (3). A side chain amine serves as the attacking nucleophile at C1[prime] (1.). The resulting Schiff base undergoes open- and closed-form tautomerisation (2.), with the open ring form (3.) facilitating proton abstraction from the 2[prime] position (4.), effecting [beta]-elimination (5.). Hydrolysis cleaves the covalent enzyme/DNA intermediate (6.). Treatment with sodium borohydride reduces the Schiff base intermediate to generate a stable covalent link between the enzyme and DNA substrate (7.). The adenine glycosylase MutY from Escherichia coli represents a subclass of BER enzymes which participate in the repair of oxidative DNA damage, and contain an [Fe4S4]2+ center (4,5). Its most important substrate is the 7,8-dihydro-8-oxo-2[prime]-deoxyguanosine (OG):A mispair, where the base removed is the undamaged adenine (6). The remaining apurinic/apyrimidinic (AP) site is then subject to further enzymatic repair activity, resulting in complete removal of the AP site deoxyribose ring from the phosphodiester backbone. The remaining nucleotide gap is filled in by DNA polymerase, followed by ligase activity to yield an OG:C mispair (7,8). The OG:C mispair is a substrate for the OG glycosylase MutM (FPG), leading to eventual restoration of the original G:C base pair (6,9). Whether or not MutY plays a direct role in removal of the AP site deoxyribose ring remains an issue of dispute. MutY also displays adenine glycosylase activity toward a variety of other substrate mispairs, notably G:A and C:A mispairs (10). Since its discovery, MutY has been classified as both a monofunctional and a bifunctional enzyme. Au et al. detected monofunctional adenine glycosylase activity for MutY in 1989 (11). In 1992, however, the DNA strand nicking behavior at the mispaired 2[prime]-deoxyadenosine observed in assays with MutY by Tsai-Wu et al. was demonstrated to be intrinsic to the enzyme(i.e. not arising by virtue of endonuclease contamination in the purified sample) (10). When the two activities from samples of each purification step were quantified by densitometry, the specific activity of each increased proportionately as the purification progressed. Additionally, the glycosylase activity for each step was shown to be slightly higher than strand scission activity. In 1995, Lu et al. reported strand nicking activity in MutY toward G:A, C:A, OG:A and G:N mispairs (where N = nebularine) (12); this activity was not quantified with respect to AP site generation. In the same year, Sun et al. labeled MutY as a monofunctional glycosylase, in part by virtue of its inability to form a covalent complex with G:A mispair-containing DNA in the presence of NaBH4 (13). Substrate specificity studies with MutY in 1996 by Bulychev et al. indicated an absence of AP lyase activity in their MutY preparations (14). In a recent report on the p26 catalytic domain of MutY by Manuel and Lloyd (15), the authors write that AP lyase activity had a 1:1 rate correspondence to the glycosylase step in all nicking assays, since piperidine treatment after enzyme incubation did not increase the formation of cleaved product relative to reactions not exposed to base. This was true in the cases of both the cloned p26 domain, and for the intact enzyme in the presence of OG:A substrate mispairs (15). In the present work, this controversy is addressed directly through two different assay types. The first involves the qualitative monitoring for a Schiff base intermediate, which has become an emergent characteristic of enzymes catalyzing DNA strand scission at a rate parallel to glycosylase chemistry (16). Figure In addition, a quantitative comparison is conducted between MutY and two other DNA glycosylases, both of which are very well characterized in terms of the activity they exhibit on their respective substrates. One of these is uracil DNA glycosylase (UDG), a commercially available monofunctional glycosylase, the substrate for which is 2[prime]-deoxyuracil-containing DNA, single-stranded or in a duplex (17). The other is FPG (also known as MutM), a DNA repair enzyme known to elicit strand scission at a rate equal to that of glycosylase catalysis directed toward the OG base of OG:C mispair substrates in duplex DNA (18,19). Assays for both glycosylase and lyase activities were conducted on all three enzymes in parallel, with the respective mispair substrates centrally located within 30mer deoxyoligonucleotide duplexes of identical flanking sequence. Substrate/product ratios as a function of enzyme reaction time, with or without subsequent base treatment, were determined by quantitative PhosphorImager analysis of the polyacrylamide gels upon which 32P-5[prime] end-labeled DNA strands of differing molecular weights were resolved. The enzyme reactions carried out in the presence of sodium borohydride produced no detectable covalent protein-DNA adducts in UDG incubation with G:U mispairs, as would be expected for an enzyme harbouring monofunctional glycosylase activity. Such trapped species were seen in the reaction of FPG with the OG:C mispair substrate, as well as that of MutY with OG:A and G:A mispairs. Interestingly, the activity assay results show FPG to be the only enzyme of the three to elicit DNA strand scission at a rate parallel to glycosylase chemistry. The conversion to cleaved DNA product in the presence of FPG took place regardless of base treatment. This is in contrast to the activity assays performed with UDG and MutY, where both enzymes exhibited inefficient DNA strand scission without subsequent base treatment of reaction aliquots. The chemistry elicited toward OG:A and G:A mispair-containing substrates by MutY is therefore that of a monofunctional glycosylase. The results of this work have led to a detailed hypothesis regarding the catalytic mechanism of MutY, based upon active site residues suspected to be involved in this enzyme's chemistry.
MATERIALS AND METHODS
Materials
UDG was purchased from New England Biolabs. FPG was a kind gift from Dr J. Laval, Institut Gustave-Roussy, France. All substrate 2[prime]-deoxyribooligonucleotides were synthesized by standard phosphoramidite chemistry on an Applied Biosystems automated oligonucleotide synthesizer as per the manufacturer's protocol. The OG phosphoramidite was purchased from Glen Research. The 5[prime]-end-labeling was performed with T4 polynucleotide kinase purchased from New England Biolabs in the presence of [[gamma]-32P]ATP from Amersham Life Sciences. Labeled oligonucleotides were purified using the Nensorb purification system (DuPont-NEN). All other buffers and reagents used were purchased from Fisher or USB. Gel imaging and quantitation were performed using a Molecular Dynamics Storm 840 PhosphorImager. Additional qualitative gel images were produced via autoradiographic exposure to Hyperfilm-MP film from Amersham.
MutY purification
MutY was isolated and purified as described previously (20). The percent active enzyme out of total purified MutY protein for the preparation used in the experiments reported herein was determined to be at 60%, using standard active site titration methods modified specifically for MutY (21). The final stock enzyme concentration was determined to be 22 µM based upon UV-vis spectroscopic data at 410 nm ([epsis] ~ 17 000 M-1cm-1). The concentrations of MutY reported throughout are those from UV-visible measurements and are not corrected for active enzyme concentration.
Substrate preparation
Single-stranded DNA 30mers were synthesized of 5[prime]-CGATCATGGAGCCAC X AGCTCCCGTTACAG-3[prime] and 3[prime]-GCTAGTACCTCGGTG Y TCGAGGGCAATGTC-5[prime], where X = dG or dOG, and Y = dU, dC or dA. Mispair substrates were formed by adding two molar equivalents of the complement sequence to the 32P-5[prime] end-labeled strand in buffer containing 20 mM Tris-HCl (pH 7.6), 10 mM EDTA and 150 mM NaCl. Mixtures were then heated to 90°C for 5 min and allowed to cool slowly to 4°C overnight. Additional MutY assays employed an 18 bp DNA duplex of 5[prime]-TCATGGGTC OG TCGGTATA-3[prime] and 3[prime]-AGTACCCAG A AGCCATAT-5[prime]. Strand labeling and annealing procedures were identical to those used for the generation of the above 30mer duplexes.
Quantitative glycosylase/lyase assays
For assays of MutY, OG:A mispair-containing DNA duplexes with the A-containing strand 32P-labeled were employed. Analogously, OG:C mispair duplexes were used in FPG assays, with the OG-containing strand labeled. UDG assays were performed with G:U mispair duplexes, where the U-containing strand was 32P-labeled. In each case, the total reaction volume was 80 µl with final DNA concentrations of ~10 nM. It should be noted that the actual DNA concentrations may be lower due to loss during the post-labeling purification steps. Final enzyme concentrations were kept at 800 nM. Again, it is worth noting that this concentration of UDG is highly approximated, owing to its small volume commercial source. Buffer conditions were 20 mM sodium phosphate (pH 7.6) or 20 mM Tris-HCl (pH 7.6) and 10 mM EDTA. In the case of experiments in Tris-HCl, the assay was conducted both with and without 100 mM NaCl. Reactions were performed at 37°C, and, at time intervals of 0.5, 1, 2, 4, 8, 16, 32 and 64 min, two separate 4 µl aliquots were taken from the reaction. One aliquot was immediately quenched at -78°C; the other was exposed to NaOH or piperidine with a final concentration of 0.1 M in base, and heated to 90°C for 5 min before quenching at -78°C. Five microliters of denaturing loading dye (80% formamide, 0.025% xylene cylanol, 0.025% bromophenol blue in TBE buffer) were subsequently added to all aliquots. Following heat denaturation for 2 min, the aliquots were loaded onto 15% polyacrylamide gels containing 8 M urea to separate products from unreacted DNA species.
Qualitative assays for reduced enzyme-DNA intermediates
As with the quantitative assays above, each enzyme was incubated with its respective substrate mispair. Final buffer component concentrations in these reactions were 25 mM sodium phosphate (pH 6.8), 1 mM EDTA and 0.1 mg/ml BSA. Final substrate DNA duplex concentration was 10 nM; final enzyme concentration was 300 nM. Seven NaBH4 concentrations were employed: 0, 30, 60, 90, 120, 150 and 180 mM. Total reaction volumes were 10 µl each. NaCl was added to each reaction such that the Na+ ion concentration was kept constant at 180 mM. All reactions were performed at 37°C for 30 min. An equal volume (10 µl) of SDS-PAGE loading buffer [125 mM Tris-HCl (pH 8.0), 5% SDS, 25% glycerol, 0.025% bromophenol blue] was added to each reaction, and the resulting 20 µl volume was heated to 90°C for 10 min prior to loading onto an 8% polyacrylamide SDS gel for resolution of free DNA from covalent DNA-enzyme complexes.
RESULTS
Quantitative glycosylase/lyase assays
Quantitative assays to determine the extents of glycosylase activity versus dual glycosylase/lyase activity were performed by monitoring strand scission of the appropriately 32P-end-labeled duplex with or without added base in the presence of MutY, FPG or UDG. The cleaved strand was separated from uncleaved strand by denaturing PAGE and visualized by storage phosphor autoradiography. Quantitation of the percent of cleaved strand in the presence and absence of base provides an indication of each enzyme's ability to catalyze a [beta]-lyase reaction associated with their respective glycosylase reactions.
The quantitative glycosylase/lyase activity assays for the three different enzymes illustrate the distinct difference between the chemistry of simple glycosylases and glycosylases harbouring concomitant strand scission activity (Fig.
Figure 2. (A) Autoradiogram of denaturing polyacrylamide gel of FPG in reaction with the OG:C mispair substrate. S, substrate 30mer DNA; P, 15mer cleavage product. Lane 1, 30mer DNA substrate alone; Lane 2, substrate DNA treated with 0.1 M NaOH; lanes 3-18, substrate DNA in the presence of enzyme for the times indicated. The odd-numbered lanes were non-base treated (-); the even-numbered lanes were treated with 0.1 M NaOH (+). (B) Autoradiograph of UDG in the presence of G:U mispair substrate, and (C) MutY in the presence of OG:A mispair DNA. Lane assignments for (B) and (C) are as described in (A). Images were generated by direct scanning of the autoradiographs. (D-F) Plots of DNA product conversion as a function of time for the three enzymes. Information was obtained by PhosphorImager quantitative analysis of the gel images described in (A-C). Lyase activity was quantified from the non-base treated lanes, while glycosylase activity was determined from lanes treated with NaOH. Band intensity on the gels depicted in Figure Identical experiments with MutY and OG:A mispair-containing 30mer duplexes were carried out in sodium phosphate buffer instead of Tris-HCl at the same pH. No quantifiable difference in conversion to product was detected between the two buffer environments, demonstrating that MutY's activity is independent of buffer type (data not shown). Since a majority of the other published MutY experiments monitoring simple glycosylase activity employ piperidine as the base for effecting DNA strand cleavage, activity assays were repeated, replacing 0.1 M NaOH with 0.1 M piperidine for base treatment. No detectable difference was observed between experiments differing by the type of base used (data not shown). Additionally, there was no variance seen between different preparations of MutY from our laboratory. However, it should be noted that all MutY preparations used were purified in similar fashion (20), and differ only in the final stock enzyme concentration. Many published MutY activity assays also report the use of between 50 and 100 mM NaCl in the reaction buffer. To address the possibility that salt concentration may have an influence upon substrate conversion and the extent of lyase activity under our reported conditions, reactions were also carried out with a final NaCl concentration of 100 mM. However, no detectable difference was seen between these experiments and those carried out without the presence of sodium chloride (data not shown). Figure 3. Storage phosphor autoradiogram of denaturing polyacrylamide gel image of MutY reaction with OG:A mispair substrates in the 18mer duplex flanking sequence. Substrate and product bands are indicated (S,P). Enzyme:substrate molar ratios are indicated above. Lane A, DNA alone; lane B, DNA treated with 0.1 M NaOH. Lanes 1-4, incubation of MutY followed by 0.1 M NaOH treatment for 1, 10, 30 and 60 min, respectively; lanes 5-8, analogous to lanes 1-4, but without the base treatment. Image taken directly from PhosphorImager, 0-25424 counts, with linear adjustment. The activity of MutY toward OG:A mispairs in a different flanking sequence environment and duplex length, as well as different enzyme:DNA molar ratios was also investigated. Molar ratios of 1:1, 10:1 and 100:1 MutY:substrate DNA (OG:A-containing 18mer duplex) were allowed to react in the Tris-HCl buffer environment. As evident in Figure Figure 4. Storage phosphor autoradiogram of SDS-PAGE analysis of enzyme-substrate reactions in the presence of varying sodium borohydride concentrations. Lanes 1, substrate duplex alone; lanes 2, substrate DNA with enzyme; lanes 3-6, substrate DNA with enzyme and 0, 60, 120 and 180 nM [BH4-], respectively; C, covalent adduct; S, substrate bands; P, product bands. Enzymes and substrate mispairs used are indicated. Image taken directly from PhosphorImager, 0-5000 counts, with linear adjustment. The detection of reduced Schiff base intermediates is an emergent hallmark of enzymes harbouring dual function glycosylase/lyase activity (16). Sodium borohydride reduction of such intermediates to form stable covalent enzyme-DNA complexes is shown in Figure Additional borohydride trapping experiments with MutY were performed to compare the efficiency of adduct formation between G:A and OG:A mispair substrates. Figure Figure 5. Storage phosphor autoradiogram of SDS-PAGE experiment with MutY and 30mer sequences. (A) OG:A mispairs; (B) G:A mispairs. Enzyme:DNA molar ratio was 30 for each lane. Lanes 1, DNA alone; lanes 2, DNA + MutY; lanes 3-9, DNA + MutY with increasing amounts of NaBH4: 0, 30, 60, 90, 120, 150 and 180 mM, respectively; C, covalent adduct; F, free DNA (both substrate and product bands). Image taken directly from PhosphorImager, 0-25400 counts, with linear adjustment. Figure 6. Storage phosphor autoradiogram of time-dependent borohydride reduction experiment with MutY. Sodium borohydride concentration was kept constant at 80 mM. Final enzyme concentration was 500 nM; substrate concentration was 10 nM. OG:A mispairs (A) and G:A mispairs (B) in the 30 bp duplex context were allowed to react for 1, 2, 5 and 10 min before quenching at -78°C, represented in lanes 2-5, respectively. Lanes 1, control (no enzyme). Image taken directly from PhosphorImager, 0-16999 counts, with linear adjustment.
Qualitative borohydride reduction experiments
DISCUSSION
When considering the current proposed catalytic mechanisms for both mono- and bifunctional DNA glycosylases, there is a clear discrepancy between the results of the quantitative assays, and the qualitative reduction with NaBH4. As our results have shown, MutY behaves quantitatively in a manner very similar to UDG; both enzymes generate base-labile AP sites in duplex DNA at a rate which far outpaces any strand scission activity intrinsic to either enzyme. Borohydride trapping is seen with FPG and its substrate, as would be expected for an enzyme of known bifunctional activity, proceeding via a Schiff base intermediate (2,16). However, there is clearly formation of a stable covalent complex between MutY and OG:A mispair-containing duplexes in NaBH4 reducing environments, as is also the case with G:A mispair substrates, though less facile.
It is possible that the conflict observed in the data between our quantitative and qualitative experiments is at the root of discrepancies between data obtained in other laboratories working with MutY. The results of the work presented here indicate that MutY bears the activity of a monofunctional glycosylase, but is nevertheless capable of forming stable covalent enzyme-DNA complexes with substrate mispairs in the presence of NaBH4. It is therefore desirable to summarize the detailed reaction conditions reported in experiments published with MutY. Table 1 outlines the experimental conditions used in previously published MutY borohydride trapping and activity assays with G:A and OG:A substrates, the results of which have led the respective authors to classify MutY as either a monofunctional or a bifunctional enzyme.
Table 1.
| Source | MutY classificationa | Reaction conditions | Substrate DNA | Activity assay results | Borohydride reduction resultsb |
| 6,9 | MONO | 20 mM Tris-HCl pH 7.6 | 23mer | Full conversion for both G:A and OG:A | N/A |
| 10 mM EDTA | G:A | ||||
| 0.05 mg/ml BSA | OG:A | ||||
| 37°C, 30[prime] | MutY:DNA ratio = 58 | ||||
| 0.1 M NaOH treatment | |||||
| 13 | MONO | 25 mM sodium phosphate | 50mer | Full conversion with base treatment | +100 mM BH4- |
| pH 6.8 | G:A | Incomplete conversion without base | MutY:DNA ratios = 2.2, 11.2, 56.4 | ||
| 1 mM EDTA | MutY:DNA ratios = | No trapping detected | |||
| 0.1 mg/ml BSA | 0.04, 0.2, 4.4, 22 | ||||
| 100 mM NaCl or KCl | |||||
| 37°C, 30[prime] | |||||
| with/without piperidine | |||||
| 14 | MONO | 62.5 mM Tris-HCl pH 7.5 | 23mer | Incomplete conversion | N/A |
| 6.25 mM EDTA | G:A | (sub-stoichiometric MutY) | |||
| 125 mM KCl | OG:A | ||||
| 25°C, 90[prime]-G:A | MutY:DNA ratio = 0.4 | ||||
| 25°C, 20[prime]-OG:A | |||||
| piperidine treatment | |||||
| 20 | MONO | 20 mM Tris-HCl pH 7.5 | 30mer | Full conversion | N/A |
| 10 mM EDTA | G:A | ||||
| 0.1 mg/ml BSA | OG:A | ||||
| 37°C, 5[prime], 15[prime], 30[prime], 60[prime] | MutY:DNA ratio = 30 | ||||
| 0.1 M NaOH | |||||
| This work | MONO | 20 mM Tris-HCl pH 7.6 | 30mer | Full (>90%) conversion at reaction times above 1[prime], only with base treatment. | [NaBH4] = 30, 60, 90, |
| or | G:A | 120, 150, 180 mM | |||
| 20 mM sodium phosphate | OG:A | 25 mM NaHPO4 | |||
| pH 7.6 | MutY:DNA ratio = 48 | <10% conversion without base, up to 64[prime] | pH 7.6 | ||
| 10 mM EDTA | (based on active site titration) | 1 mM EDTA | |||
| 0 or 100 mM NaCl | 0.1 mg/ml BSA | ||||
| 37°C, 0.5, 1[prime], 2[prime], 4[prime], 8[prime], 16[prime], 32[prime], 64[prime] | utY:DNA ratio = 18 | ||||
| 0.1 M NaOH | Trapping more efficiency: | ||||
| or | OG:A = 60% | ||||
| 0.1 M piperidine | G:A = 12% | ||||
| 10 | BI | 20 mM Tris-HCl pH 7.6 | 40mer | Near full conversion only in base-treated lanes | N/A |
| 1 mM EDTA | G:A | ||||
| 1 mM DTT | MutY:DNA ratio = ? | ||||
| 80 mM NaCl | |||||
| 2.9% glycerol | |||||
| 30°C, 30[prime] | |||||
| with/without piperidine | |||||
| 12 | BI | 20 mM Tris-HCl pH 7.6 | 20mer | Incomplete conversion | N/A |
| 1 mM EDTA | G:A | ||||
| 1 mM DTT | OG:A | ||||
| 80 mM NaCl | MutY:DNA ratio = 933 | ||||
| 2.9% glycerol | |||||
| 37°C, 30[prime] | |||||
| No base treatment | |||||
| 15 | BI | 25 mM sodium phosphate | 30mer | 60% conversion with G:A | +100 mM BH4- |
| pH 6.8 | G:A | 25% conversion with OG:A | MutY:DNA ratio = 100 | ||
| 1 mM EDTA | OG:A | (regardless of base treatment) | Positive for covalent complexes | ||
| 0.1 mg/ml BSA | MutY:DNA ratio = 10 | ||||
| 50 mM NaCl | |||||
| 37°C, 30[prime] | |||||
| with/without piperidine | |||||
| 23 | BI | 20 mM Tris-HCl pH 7.6 | 20mer | Incomplete conversion | +100 mM BH4- |
| 1 mM EDTA | G:A | No significant increase in conversion with base treatment | MutY:DNA ratios = | ||
| 1 mM DTT | MutY:DNA | 0.3, 0.5, 1, 2.5, 10, 40 | |||
| 80 mM NaCl | ratios = 0.3, 0.5, | Trapping more efficient with G:A than with OG:A at ratios of 10 and 40 | |||
| 2.9% glycerol | 1, 2.5, 10, 40 | ||||
| 37°C, 30[prime] | |||||
| with/without piperidine | |||||
| 24 | BI | 20 mM Tris-HCl pH 7.6 | 20mer | Incomplete conversion | +100 mM BH4- |
| 1 mM EDTA | G:A | MutY:DNA ratios = 2.5, 10, 40, 160 | |||
| 1 mM DTT | OG:A | Trapping more efficient with OG:A than with G:A at all ratios | |||
| 80 mM NaCl | MutY:DNA | ||||
| 2.9% glycerol | ratios = 2.5, 5, 10, 20, 40 | ||||
| 37°C, 30[prime] | |||||
| NO base treatment | |||||
| 32 | BI | 25 mM sodium phosphate | G:A 50mer | Full conversion for OG:A at ratios of 5 and 10 | N/A |
| pH 6.8 | OG:A 21mer | ||||
| 1 mM EDTA | (not in same sequence context) | Incomplete conversion for G:A for all MutY:DNA ratios | |||
| 0.1 mg/ml BSA | |||||
| 50 mM NaCl | MutY:DNA ratios = 1, 5, 10 | ||||
| 37°C, 30[prime] | |||||
| with/without piperidine |
bReaction conditions for the borohydride reduction assays are the same as for the activity assays unless otherwise stated.
In terms of published experiments using sodium borohydride to assay for covalent MutY-substrate DNA complexes, the only one reporting an absence of trapped species was work done in 1995 by Sun et al. (13). The authors supported their categorization of MutY as a monofunctional glycosylase by virtue of its inability to generate covalent enzyme-DNA adducts with G:A mispair substrates in the presence of 100 mM NaBH4. Later, Manuel and Lloyd reported borohydride trapping for both G:A and OG:A mispairs (15). Other puzzling results of Lu et al. (23) compared with the data reported by Gogos et al. (24) indicate opposite trends in borohydride trapping of G:A and OG:A mispairs. Lu et al. found that at higher MutY:DNA ratios, borohydride trapping was more efficient with G:A mispairs than with OG:A (23). In contrast, Gogos et al. found that reductive trapping was more efficient for OG:A than G:A mispairs (24). In our work, we consistently observe more efficient trapping with OG:A than with G:A substrates.
The literature results on glycosylase assays and borohydride trapping of MutY compiled in Table 1 suggest that substrate preparation, reaction buffer type, the use of bovine serum albumin (BSA) and dithiothreitol (DTT), do not have a uniform effect upon the extent of substrate conversion to product by MutY, nor its behaviour as a monofunctional or bifunctional enzyme. Thus, there seems to be no single component of the published reaction conditions which uniformly governs the efficiency of enzyme-DNA adduct formation in reducing environments, nor the extent of glycosylase and glycosylase/lyase reaction in the MutY activity assays with G:A and OG:A substrates. Therefore, inconsistent results overall may be attributed to a combination of variables which may be difficult to control.
One variable is the percent active enzyme of any given MutY purification. Most of the published works cited herein do not report quantitation of the active enzyme component in final purified samples of MutY. Using active site titration experiments, we have found some variability in the percent active enzyme from different MutY preparations in our laboratory. Furthermore, kinetic experiments with MutY on G:A and OG:A substrates has shown that MutY is extremely inefficient at effecting multiple turnover of its substrate due to its high affinity for the product of its glycosylase activity (21). The inability to dissociate from the product is particularly severe for OG:A substrates, and in this case the degree of conversion of substrate to product is directly proportional to the amount of active MutY enzyme. It is therefore possible that protein preparations having very low percent active enzyme concentration could lead to incomplete substrate conversion when used in reactions where enzyme is thought to be in molar excess relative to substrate, even in those reactions treated with base. This could be one explanation for work reporting no significant increase in product formation upon base treatment, relative to reactions not treated with base. Indeed, in the reported studies where base-independent DNA strand cleavage has been observed with MutY, there has been incomplete conversion of substrate to product, even in molar excess of enzyme. Thus, under these conditions the small amount of conversion may be very close to the `background' strand-scission reaction that we have observed with both MutY and UDG (~10%). If MutY were a bifunctional glycosylase/AP-lyase with high activity, one would expect under conditions of excess enzyme to witness the near quantitative base-independent substrate conversion as seen with FPG reactions with its substrate.
Figure 7. Possible active site chemistry for MutY. (A) Initial nucleophilic attack at the C1[prime] carbon is by an activated water molecule, facilitated by protonation of the leaving adenine base (1.). The resulting AP site undergoes a second nucleophilic attack at C1[prime] by a side chain amine to form the Schiff base species (2.), the ring-opened tautomer of which (3.) can be reduced to form a trapped enzyme-DNA complex (4.). (B) Initial nucleophilic attack at the C1[prime] carbon is by a side chain amine (1.). The Schiff base tautomerises to the ring-opened form (2.) and undergoes another nucleophilic attack at C1[prime] by an activated water molecule to generate an AP site (3.). As in (A), the Schiff base can be reduced by sodium borohydride to form a covalent enzyme-DNA adduct (4.). Another caveat which may be contributing to conflicts in published MutY data is the issue of sequence context effects. Data from our laboratory have shown that the activity of MutY toward G:A mispairs is highly sensitive to sequence context (B.D.Meads, M.J.Tuttle and S.S.David, unpublished results). Therefore, there may very well be unforeseen sequence effects upon substrate conversion by MutY, which are playing a large role in the different reported experimental outcomes by various groups. The thermodynamic stability of duplexes containing abasic sites (25,26) and the structural properties of abasic sites (27) have been shown to be sensitive to the sequence context and this may affect the lability of abasic sites. Thus, in some sequence contexts, the background strand scission reaction may be more prominent. A related issue may also be one of substrate duplex annealing. MutY is highly specific for its substrate base pairs within duplex DNA. A significant amount of labeled strand inefficiently annealed to its complement strand could also explain low substrate conversions in the presence of excess MutY. The inconsistent results in different laboratories may also be attributed to the different protocols used for the isolation and purification of MutY. It is possible that certain purification steps preserve the strand scission activity of MutY while others reduce or abolish it. This may be expected for an enzyme that contains a cofactor which may be lost upon purification; however, this seems unlikely with the known mechanistic properties of the AP-lyase reaction. Other possible explanations for the different results may lie with the intrinsic lability of abasic sites such that reaction conditions and handling of the duplex may result in different amounts of observed strand cleavage. These variables underscore the usefulness of direct comparisons with well-behaved monofunctional and bifunctional enzymes in difficult cases like MutY. Previous experiments using OG:F and G:F mispairs in 30 bp duplexes yielded valuable insight into the dissociation constants for substrate-like mispairs (where F = 2[prime]-deoxyformycin, a non-cleavable substrate analogue to 2[prime]-deoxyadenosine) (20). Indeed, the Kd determined for the OG:F mispair (0.8 ± 0.4 nM) is significantly lower than that of the G:F mispair (26 ± 6 nM). This difference in Kd as a function of G or OG in mispair with the formycin analogue is mirrored in the values determined for the products of OG:A and G:A substrate turnover. Based upon these results, and those of hydroxyl radical footprinting assays conducted in the same work, it is clear that MutY has a higher affinity for OG:A mispairs than G:A mispairs. Furthermore, MutY exhibits significant affinity for the products of its glycosylase activity. The dissociation constant studies are consistent with the results of the present work. It seems clear that the active site of MutY and the substrate/product of the OG:A mispair stay in close proximity for longer time periods than identical DNA duplexes harbouring a G:A mispair. Therefore, our hypothesis suggests that an amine functionality (perhaps a lysine residue) in the active site of MutY lies close enough to the substrate mispaired adenosine or the product AP site to form a transient imine species. In one model, the original nucleophilic attack at C1[prime] could be performed by an activated water molecule to generate an AP site. The C1[prime] of the AP site could then be subject to a second nucleophilic attack by the amine functional group of the enzyme to yield the trappable Schiff base species (Fig. Figure 8. Sequence alignment of members of a BER superfamily. Active site helix-hairpin-helix (HhH) and Gly-Pro-Asp (GPD) motifs conserved among the BER superfamily DNA glycosylases. Six representative enzymes are illustrated. Alignment was made based upon previously published alignments indicating the existence of a BER superfamily (28,30). Indeed, a similar idea for the MutY active site has recently been suggested (28). In such a scenario where a lysine [epsis]-NH2 group experiences occasional proximity to either the substrate for glycosylase action or the AP site, there could be an increase in strand scission events (presumably via either [beta]- or [delta]-elimination) taking place separately from AP site generation. This would also explain the apparent imine species reduction in the presence of sodium borohydride. This hypothesis agrees well with the fact that OG:A mispair-containing DNA was found to form covalent complexes with MutY more efficiently than G:A mispairs. Similarly, mutagenesis research with DNA polymerase [beta] illustrated a variability in borohydride trapping efficiency (29). Mutation of the active site lysine 72 to alanine did not eliminate Schiff base formation; covalent complexes with the K72A mutant were reported to be ~30% relative to those formed with the wild-type protein. It was suggested that other neighbouring lysine residues can participate as sites for imine species formation. Further credence is lent to the emergent picture of MutY's active site pocket by virtue of sequence alignments with other members of the BER superfamily. Endonuclease III homologues from three other species, and a UV endonuclease, are all dual function glycosylase/AP lyases which harbour a lysine mapping to the E.coli endonuclease III Lys 120 position in sequence alignments (30). Furthermore, studies with a human OG glycosylase (hOGG1) have identified Lys 249 as the residue responsible for imine intermediate generation during catalysis, and indeed, Lys 249 is the residue mapping to Lys 120 of endonuclease III in sequence alignments (Fig. The results of the work reported herein indicate that MutY may indeed have the mechanistic ability to effect DNA strand scission via possible AP-lyase activity, albeit at a rate dramatically slower than that of its glycosylase activity. Therefore, it seems most appropriate to classify MutY as a monofunctional glycosylase, based upon the catalytic activity it directs toward its substrates. Under the conditions used (effectively 16-fold molar excess MutY) near 100% conversion to product was seen in under 1 min incubation time for OG:A substrates, and ~30 min for G:A substrates, upon subsequent treatment with 0.1 M NaOH. Without base treatment, generation of cleaved DNA products reached a plateau at ~10% conversion, similar to the case of UDG. Indeed, both UDG and FPG have provided excellent control experiments for our assays with MutY. Since for MutY, cleavage of the C1[prime]-N glycosyl bond occurs at a much higher rate than DNA strand scission at the resultant AP site, it is our conclusion that any lyase activity inherent to MutY is not mechanistically paired to the activity of base removal.
A hypothesis for the active site of MutY
ACKNOWLEDGEMENTS
This work was supported by the National Institutes of Health Grant CA67985 and the University of Utah. S.S.D. is an A.P.Sloan Research Fellow (1998-2000).
REFERENCES
Since submission of our manuscript, a report has appeared (Zharkov,D.O. and Grollman,A.P., Biochemistry, 37, 12384-12394) in which formation of a trappable Schiff base intermediate with an OG:A substrate was observed. Further, these authors identify Lys142 as the catalytic residue involved in Schiff base formation.NOTE ADDED IN PROOF
This article has been cited by other articles:
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 3 Nov 1998
Copyright©Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
J. O. Blaisdell and S. S. Wallace
Rapid determination of the active fraction of DNA repair glycosylases: a novel fluorescence assay for trapped intermediates
Nucleic Acids Res.,
March 12, 2007;
35(5):
1601 - 1611.
[Abstract]
[Full Text]
[PDF]
T. Watanabe, J. O. Blaisdell, S. S. Wallace, and J. P. Bond
Engineering Functional Changes in Escherichia coli Endonuclease III Based on Phylogenetic and Structural Analyses
J. Biol. Chem.,
October 7, 2005;
280(40):
34378 - 34384.
[Abstract]
[Full Text]
[PDF]
R. C. Manuel, K. Hitomi, A. S. Arvai, P. G. House, A. J. Kurtz, M. L. Dodson, A. K. McCullough, J. A. Tainer, and R. S. Lloyd
Reaction Intermediates in the Catalytic Mechanism of Escherichia coli MutY DNA Glycosylase
J. Biol. Chem.,
November 5, 2004;
279(45):
46930 - 46939.
[Abstract]
[Full Text]
[PDF]
X. Li and A-L. Lu
Molecular Cloning and Functional Analysis of the MutY Homolog of Deinococcus radiodurans
J. Bacteriol.,
November 1, 2001;
183(21):
6151 - 6158.
[Abstract]
[Full Text]
[PDF]
H. Yang, W. M. Clendenin, D. Wong, B. Demple, M. M. Slupska, J.-H. Chiang, and J. H. Miller
Enhanced activity of adenine-DNA glycosylase (Myh) by apurinic/apyrimidinic endonuclease (Ape1) in mammalian base excision repair of an A/GO mismatch
Nucleic Acids Res.,
February 1, 2001;
29(3):
743 - 752.
[Abstract]
[Full Text]
[PDF]
K. Shinmura, S. Yamaguchi, T. Saitoh, M. Takeuchi-Sasaki, S.-R. Kim, T. Nohmi, and J. Yokota
Adenine excisional repair function of MYH protein on the adenine:8-hydroxyguanine base pair in double-stranded DNA
Nucleic Acids Res.,
December 15, 2000;
28(24):
4912 - 4918.
[Abstract]
[Full Text]
[PDF]
O. M. Sidorkina and J. Laval
Role of the N-terminal Proline Residue in the Catalytic Activities of the Escherichia coli Fpg Protein
J. Biol. Chem.,
March 31, 2000;
275(14):
9924 - 9929.
[Abstract]
[Full Text]
[PDF]
X. Li, P. M. Wright, and A-L. Lu
The C-terminal Domain of MutY Glycosylase Determines the 7,8-Dihydro-8-oxo-guanine Specificity and Is Crucial for Mutation Avoidance
J. Biol. Chem.,
March 17, 2000;
275(12):
8448 - 8455.
[Abstract]
[Full Text]
[PDF]
P. M. Wright, J. Yu, J. Cillo, and A-L. Lu
The Active Site of the Escherichia coli MutY DNA Adenine Glycosylase
J. Biol. Chem.,
October 8, 1999;
274(41):
29011 - 29018.
[Abstract]
[Full Text]
[PDF]
M. M. Slupska, W. M. Luther, J.-H. Chiang, H. Yang, and J. H. Miller
Functional Expression of hMYH, a Human Homolog of the Escherichia coli MutY Protein
J. Bacteriol.,
October 1, 1999;
181(19):
6210 - 6213.
[Abstract]
[Full Text]
D. O. Zharkov, T. A. Rosenquist, S. E. Gerchman, and A. P. Grollman
Substrate Specificity and Reaction Mechanism of Murine 8-Oxoguanine-DNA Glycosylase
J. Biol. Chem.,
September 8, 2000;
275(37):
28607 - 28617.
[Abstract]
[Full Text]
[PDF]
E. Nicolas, J. M. Beggs, and T. F. Taraschi
Gelonin Is an Unusual DNA Glycosylase That Removes Adenine from Single-stranded DNA, Normal Base Pairs and Mismatches
J. Biol. Chem.,
September 29, 2000;
275(40):
31399 - 31406.
[Abstract]
[Full Text]
[PDF]
This Article 
Abstract
Print PDF (466K)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (41)
Request Permissions
Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Articles by Williams, S. D.
Articles by David, S. S.
Search for Related Content
PubMed
PubMed Citation
Articles by Williams, S. D.