N-terminal Extension of N-Methylpurine DNA Glycosylase Is Required for Turnover in Hypoxanthine Excision Reaction*

N-Methylpurine DNA glycosylase (MPG) initiates base excision repair in DNA by removing a wide variety of alkylated, deaminated, and lipid peroxidation-induced purine adducts. In this study we tested the role of N-terminal extension on MPG hypoxanthine (Hx) cleavage activity. Our results showed that MPG lacking N-terminal extension excises hypoxanthine with significantly reduced efficiency, one-third of that exhibited by full-length MPG under similar conditions. Steady-state kinetics showed full-length MPG has higher Vmax and lower Km than NΔ100 MPG. Real time binding experiments by surface plasmon resonance spectroscopy suggested that truncation can substantially increase the equilibrium binding constant of MPG toward Hx, but under single-turnover conditions there is apparently no effect on catalytic chemistry; however, the truncation of the N-terminal tail affected the turnover of the enzyme significantly under multiple turnover conditions. Real time binding experiments by surface plasmon resonance spectroscopy further showed that NΔ100 MPG binds approximately six times more tightly toward its product apurinic/apyrimidinic site than the substrate, whereas full-length MPG similarly binds to both the substrate and the product. We thereby conclude that the N-terminal tail in MPG plays a critical role in overcoming the product inhibition, which is achieved by reducing the differences of MPG binding affinity toward Hx and apurinic/apyrimidinic sites and thus is essential for the Hx cleavage reaction of MPG. The results from this study also affirm the need for reinvestigation of full-length MPG for its enzymatic and structural properties, which are currently available mostly for the truncated protein.

Cellular DNA is continuously damaged by various endogenous or exogenous chemical or physical agents. Multiple DNA repair pathways repair damaged bases and prevent cell death and mutations responsible for genomic instability, cancer, and aging (1)(2)(3).
In all organisms, repair of DNA-containing small adducts, as well as altered and abnormal bases, occurs primarily via the base excision repair pathway, beginning with cleavage of the base by a DNA glycosylase. Mammalian N-methylpurine DNA glycosylase (MPG), 2 a monofunctional glycosylase, is known to excise at least 17 structurally diverse modified purine bases, including toxic and mutagenic alkylated, deaminated, and etheno adducts from both the major and minor grooves of duplex DNA (4 -12).
In our previous study, we showed that MPG is organized into three distinct domains with a protease-hypersensitive ϳ100amino acid region at the N terminus (13). We also found that truncated (N⌬100C⌬18) and full-length enzymes retained similar binding and kinetic properties toward ⑀A (7). Later, several x-ray structures of human truncated MPG in complex with ⑀A or control DNA were published with the notion that the seemingly unstructured (protease-sensitive) N-terminal extension may hinder crystallization of MPG (14 -16). But different studies by us and others showed that the N-terminal extension of MPG could be critical for recognition of substrates such as 3-methylguanine, 7-methylguanine, and 1,N 2 -ethenoguanine (1,N 2 -⑀G) adducts (9,17). Notably, there is a report that truncated and full-length forms of human MPG do not show a significant difference in Hx cleavage activity, although the long incubation period, instability of the full-length MPG, and insufficient kinetic details could have contributed to such results (11). Hx is one of the preferred substrates elucidated so far for MPG or related enzymes that are ubiquitously present in all organisms, including humans (18). Moreover, Hx was shown to be significantly mutagenic (19,20).
In the present study, we have demonstrated through analyzing individual intermediate kinetic steps that the N-terminal tail is crucial for MPG binding to Hx, dissociation from its product AP site, and its overall turnover. However, it does not have an effect on catalysis of Hx-containing DNA. The results from this study also underscore the need for the reinvestigation of full-length MPG for its enzymatic and structural properties, which are currently available in literature primarily for the truncated protein.

MATERIALS AND METHODS
Purification of Recombinant Mouse MPG-Mouse MPG wild type was purified as previously described (7).
Cloning and Purification of N⌬100 MPG-An expression construct encoding N⌬100 mMPG was prepared by ligating a * This work was supported by National Institutes of Health Grants RO1 CA 92306 (to R. R.) and RO1 CA 108641 (to A. U.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  PCR product containing the MPG coding sequence but lacking the first 100 amino acid residues at the NdeI and EcoRI sites of the pRSETB vector. PCR was carried out using a previous mouse MPG (N⌬48) construct as template and the primers (5Ј-CATATGGACCATTCTGGCCGGCTA-3Ј and 5Ј-GAATTC-TTACTTTTGAACAATTAAAAGCCCC-3Ј). The primers allowed the introduction of NdeI and EcoRI sites at the 5Ј and 3Ј ends, respectively. The PCR products were then subcloned in pCR2.1-TOPO cloning vector digested with NdeI and EcoRI and subcloned into expression vector pRSETB at NdeI/EcoRI sites, allowing us to express a nonfusion N⌬100 mMPG protein. The identity of the construct was confirmed by DNA sequencing. N⌬100 mMPG was overexpressed in E. coli BL21(DE3) cells and purified to near electrophoretical homogeneity. The purification was carried out as follows. The conditions for cell growth and induction of protein expression were as described previously (9).
For purification, the cells were harvested from 1 liter of culture in 50 ml of Buffer A (20 mM PIPES, pH 6.1, 10% glycerol, 50 mM NaCl, 0.1% Tween 20, 1 mM DTT). The cells were then lysed as described previously (7). The lysate was clarified by centrifugation at 15000 rpm for 30 min, and the supernatant was applied onto tandemly attached ion exchange columns consisting of Q Sepharose (5 ml) and SP Sepharose (1 ml), which were pre-equilibrated with Buffer A. The columns were washed with 10 column volumes of Buffer A. The Q column was then detached, and the SP column was eluted with a gradient of 0 -100% of Buffer B (Buffer A plus 600 mM NaCl) in Buffer A. The peak fractions containing N⌬100 mMPG, tested by SDS-PAGE, were pooled and diluted three times with Buffer C (20 mM HEPES, pH 7.5, 1 M NaCl, 0.8 M ammonium sulfate, 1 mM DTT) before loading onto a hydrophobic phenyl Sepharose column that was pre-equilibrated with Buffer C. After washing with Buffer C, the protein was eluted using a linear gradient of Buffer D (20 mM HEPES, pH 7.5, 5% glycerol, 1 mM DTT). The peak fractions were pooled, dialyzed against Buffer E (Tris-HCl, pH 7.5, 10% glycerol, 50 mM NaCl, and 1 mM DTT), and stored at Ϫ80°C in aliquots for future use.
Oligonucleotide Substrates Preparation-Hx and ⑀A containing 50-mer oligonucleotide with the sequence 5Ј-TCGAGGA-TCCTGAGCTCGAGTCGACGXTCGCGAATTCTGCGGA-TCCAAGC-3Ј (where X represents Hx) were purchased from Operon Technologies (Alameda, CA) and Gene Link (Hawthorne, NY). The complementary oligonucleotide containing T opposite Hx was synthesized by the Recombinant DNA Laboratory Core Facility at the University of Texas Medical Branch (Galveston, TX). The oligonucleotides were purified on a sequencing gel. The Hx or ⑀A oligonucleotide was labeled at the 5Ј end using T4 polynucleotide kinase and [ 32 P]ATP and annealed to complementary oligonucleotide to prepare 32 Pend-labeled duplex oligonucleotide as described previously (21).
MPG-mediated Excision Activity Assay-The MPG proteins, mouse full-length (15-30 nM) and N⌬100 (15-30 nM) were individually incubated with 5Ј-32 P-labeled Hx-containing duplex oligonucleotide substrates (4 -8 nM) for 10 min at 37°C in an assay buffer (25 mM HEPES-KOH, pH 7.9, 0.5 mM DTT, 10 g/ml nuclease-free bovine serum albumin, 150 mM NaCl and 10% glycerol) in a total volume of 20 l. The reaction was stopped by inactivating the enzyme at 75°C for 5 min. The products containing the AP sites were analyzed as described previously (21).
Steady-state Kinetic Study-The full-length or truncated enzyme (8.5 nM) was incubated with 5Ј-32 P-labeled Hx-containing duplex oligonucleotide (0 -60 nM) substrates for 5 min at 37°C under assay conditions similar to those described above. The reaction products were also analyzed and quantified as described for the activity assay.
DNA Binding Studies Using Surface Plasmon Resonance-A 50-mer duplex oligonucleotide containing an Hx or abasic site (tetrahydrofuran) at the 26 th position from the 5Ј end of one strand was used for measuring enzyme-DNA interactions. Oligonucleotides were biotinylated and immobilized on streptavidin-coated Biacore chips (21). Then we measured the binding parameters of truncated (0 -300 nM) or full-length MPG (0 -25 nM) for Hx and abasic site using a binding buffer (10 mM HEPES-KOH pH 7.6, 150 mM NaCl, and 0.5% surfactant) at 25°C. The MPGs at various concentrations were injected, and the surface plasmon resonance units were measured, with 60-s injections. Following each injection, the chip was regenerated with 1 M NaCl. The binding kinetics for oligonucleotides containing Hx or AP sites were established with a series of MPG concentrations. The Langmuir isotherms (1:1 binding) at various protein concentrations allowed us to calculate the kinetic binding parameters based on on/off rates and protein concentrations.
Single-Turnover (STO) Kinetic Study-The full-length (20 -40 nM) and truncated enzymes (450 -900 nM) were individually incubated with 2 nM of 5Ј-32 P-labeled Hx-containing duplex oligonucleotide substrates at 37°C in an assay buffer (25 mM HEPES-KOH, pH 7.9, 10 g/ml nuclease-free bovine serum albumin, 0.5 mM DTT, 150 mM NaCl, and 10% glycerol) in a total volume of 100 l. Aliquots of 5 l were taken out at different time points (0 -17 min) and heat-inactivated at 80°C in a preheated micro centrifuge tube. The products containing the AP sites were quantitatively cleaved into smaller fragments, followed by resolution on denaturing gels. Radioactivity in the incised oligonucleotide was also quantified as described in the activity assay.
Burst Analysis-The enzymes (10 nM) were individually incubated with 5Ј-32 P-labeled Hx-containing duplex oligonucleotide (84 -112 nM) at 37°C under conditions similar to those described in the STO kinetic study.

Purification of N⌬100 MPG and MPG-mediated Excision
Activity Assay-The wild type and N⌬100 MPGs were purified using the methods described under "Materials and Methods." Both the proteins were 85-88% pure electrophoretically (Fig.  1). Then the activity of the purified full-length and truncated MPG was measured using Hx as a substrate. The full-length MPG had 2-3-fold more activity than the truncated protein at different enzyme:substrate (2-4:1) molar ratios, indicating that the overall decreasing effect on product formation during the Hx reaction can most likely be attributed to the deletion of the Specific Interaction of MPG with Hx OCTOBER 12, 2007 • VOLUME 282 • NUMBER 41 JOURNAL OF BIOLOGICAL CHEMISTRY 30079 N-terminal extension. Thus, it appears that that the N-terminal extension plays a role in the excision activity of MPG toward Hx (Fig. 2).
Steady-state Kinetic Study-To understand the mechanism of reduced activity of truncated MPG, we then compared the steady-state kinetic parameters for both the full-length and truncated MPGs. We measured the full-length as in a previous study (21) but included N⌬100 MPG here and found that N⌬100 MPG has significantly higher K m and lower V max compared with its full-length counterpart (Table 1).
Mechanism Analysis to Understand the Role of N-terminal Tail in Hx-MPG Reaction-To have further insight into this mechanism, we used both surface plasmon resonance spectroscopy and pre-steady-state kinetics. In Fig. 3 we described the basic reaction steps of MPG, which are slightly modified from our published scheme (21). Previously, we hypothesized that the cleaved base will be diffused spontaneously, but that may not be the case. The cleaved base may still remain bound with the enzyme by a specific hydrogen bond (details under "Discussion"). Therefore, the dissociation of both products from MPG may contribute to the overall product dissociation rate. Using surface plasmon resonance, we measured the binding of MPG toward substrate (Hx) and product (abasic site). Pre-steady-state kinetic analysis provides the opportunity to identify the intermediate reaction step(s) that might be affected by the N-terminal tail. We took advantage of the slow reaction rates of MPG and measured the effect of the N-terminal tail on the glycosidic bond cleavage (catalysis) step by STO kinetics and the product dissociation step by multiple turnover reaction conditions.
Hx Binding Studies Using Surface Plasmon Resonance-In search of a mechanism of modulation of MPG activity by N-terminal extension, we examined the MPG-Hx binding using a Biacore-T100 (Biacore, Uppsala, Sweden). Our results showed that the equilibrium binding constant (K D ) is 0.15 nM for the full-length protein, whereas it is 25 nM for the truncated one (Fig. 4). Apparently, the major effect of N-terminal extension on K D is primarily due to the high rate of microscopic association (k on; 4.03 Ϯ 0.4 ϫ 10 8 M Ϫ1 s Ϫ1 ) for full-length MPG, which is close to the diffusion-controlled limit (10 9 M Ϫ1 s Ϫ1 ) (22). Thus, the N-terminal extension plays a significant role in Hx recognition and binding.
STO Kinetics-Prompted by the observation that the N-terminal tail could play a critical role in product formation in Hx-MPG reactions, we tested whether the N-terminal tail affects any of the catalytic intermediate steps other than the binding step. We conducted STO kinetics with full-length MPG proteins to measure the k chem (21,23). The reaction was performed at substrate and enzyme concentrations of 2 and 20 -40 nM for full-length enzyme or 450 -900 nM for truncated enzyme, respectively. The data were analyzed using the first order rate equation, where A 0 represents the amplitude of the exponential phase, and k obs is the observed rate constant associated with the reaction process. Under the STO conditions ([E] Ͼ Ͼ [S]), all of the substrate molecules should remain bound by enzymes. The binding step should not affect the rate of product formation, and hence, under these conditions k obs can be considered as k chem . Two different enzyme substrate ratios for both of the proteins provide similar values of k chem , ensuring the enzymatic reactions following STO conditions. However, the full-length and N⌬100 MPG have similar k chem (0.3 Ϯ 0.03) (Fig. 5), indicating the minimal effect of N-terminal extension on the chemistry step of MPG-Hx reaction.
Burst Analysis-Next, we tried to measure the rate of product release  (k pd ) and active enzyme concentration available for the reaction (21,24,25). The extremely slow turnover rate of MPG during excision of Hx provided the opportunity to perform burst analysis under the reaction conditions of [S] Ͼ Ͼ [E], where the substrate and enzyme concentration were 84 -112 and 10 nM, respectively. The data were fit to the following equation.
Plot of product concentration (P t ) versus time (t) can be analyzed using Equation 2 as before (21,23) to determine the kinetic parameters, A 0 (amplitude of the burst) and k ss (slope of the linear phase; turnover). At full-length MPG and Hx concentrations of 5 and 50 nM, we previously found a k pd value (0.016 Ϯ 0.001 min Ϫ1 ) that was low and apparently rate-limiting in the MPG-mediated multi-step reaction process (21). Here, we found a similar k pd value for full-length MPG, but notably at the same MPG and Hx concentrations the truncated protein had much lower turnover (Fig. 6). In fact, the dissociation was extremely slow, and it was neither practical nor possible to perform curve fitting to assign an accurate value of k pd for the truncated protein. Therefore, it is evident that the N-terminal extension regulates the product dissociation rate or MPG turnover for Hx reactions in a significant fashion in addition to its modulating effect on the substrate binding step.
AP Site Binding Studies Using Surface Plasmon Resonance-Further, we asked whether the N-terminal extension has any direct effect on MPG-AP site interactions to reflect the poor product dissociation rate and turnover of truncated MPG. We directly measured the binding kinetics of full-length and N⌬100 MPGs to AP sites using a Biacore-T100. Our results showed an apparent equilibrium binding constant (K D ) of 0.18 nM for the full-length protein, whereas 4.3 nM was found for the truncated one (Fig. 7). Apparently, the truncated MPG binds AP sites with ϳ6-fold higher affinity than to its substrate Hx, whereas the full-length protein has affinity comparable with that of both AP site and Hx (Figs. 4 and 7). These results indicate that the N-terminal extension plays a critical role in assisting MPG to overcome product inhibition and ensure facile turnover.

DISCUSSION
N-terminal extension is present in all MPGs from higher eukaryotes. This suggests an evolutionarily conserved critical function of the N-terminal tail for MPG activity. However, neither the structure of MPG including the N-terminal tail nor the role of this extension in enzymatic activity is fully elucidated yet.
In the present study, we have demonstrated that the N-terminal tail is a regulatory domain and is essential for MPG turnover. We found in a previous study that the N-terminal tail also had a modulating effect on the function of a bifunctional DNA glycosylase, such as human endonuclease III, but it inhibited the turnover of this enzyme and in turn its activity (24). In fact, the truncation of the N-terminal tail stimulated human endonucleases III activity at the product dissociation step (24), whereas the present study shows that the truncation of the N-terminal tail inhibited MPG activity. Thus, the N-terminal tail apparently regulates the activity of DNA glycosylases, such as a monofunctional enzyme MPG and a bi-functional NTH1, but in a diverse manner.

Specific Interaction of MPG with Hx
We had shown before by systematic deletion analysis of MPG from N and C termini that a minimally sized polypeptide (N⌬100C⌬18) lacking 100 and 18 amino acid residues from the N and C termini, respectively, and wild type enzyme had similar kinetic and binding properties for ⑀A (7). Since then, there have been several reports on the crystallographic structures of similarly truncated protein in complex with ⑀A or control DNA (14 -16). But our present study indicates that the structural information on ⑀A-truncated MPG complex is not sufficient and does not reveal the full scenario for other DNA adducts including Hx. In another previous study with a recombinant chimeric protein containing N-and C-terminal halves of human and mouse MPG, we found that the N-terminal half is critical for the recognition of 3-methylguanine and 7-methylguanine (9). However, both of the methylated substrates were positively charged, and the proteins were partially purified, so it was of interest to evaluate the role of the N-terminal tail for MPG activity toward a neutral substrate, such as Hx. From this study, it is apparent that the mode of action of MPG toward Hx and ⑀A are strikingly different.
The first 70 amino acid residues of hMPG were also indis-pensable for 1,N 2 -⑀G excision reaction (17), because the N-terminally truncated protein was less active toward 1,N 2 -⑀G. However, unlike Hx the 1,N 2 -⑀G binding was not affected by truncation; rather, a possible change in the catalytic pocket or lack of amino acids specifically involved in the catalysis of that substrate was suggested to be the cause (17). However, it is important to note that 1,N 2 -⑀G has yet to be detected in genomic DNA of biological samples (17). In the present study, we have shown by detailed mechanistic analysis of all three major steps in the MPG-Hx reaction that product formation is affected by the N-terminal tail.
Nonetheless, being a broad substrate enzyme, MPG must be flexible for DNA binding to recognize DNA lesions of varied structures. In fact, for similar reasons O'Brien and Ellenberger (18) proposed that a "nonspecific catalytic mechanism" must be met for an enzyme to succeed as a generalist as one of the major criteria, which comes "at the expense of catalytic power." In the future, it would be interesting to study the details of the binding and catalytic mechanisms of full-length MPG toward substrates other than Hx of different structures, including N-3 and N-7 of deoxyguanine adducts generated by nitrogen and sulfur mustards (26). Obviously, the structures of these MPG-cleavable adducts containing large modifications, especially at N-7 of deoxyguanine, are very different from those of Hx and ⑀A.
Another interesting point is that the equilibrium binding constant (K D ) between MPG and Hx is entirely different for the full-length and truncated versions. The K D is apparently ϳ160-fold less for N⌬100 MPG compared with full-length protein, but overall product formation is only ϳ3-fold different. One possible explanation could be that compared with its chemistry and turnover step, the binding of MPG to Hx is extremely fast, as is evident from the k on value of 4.03 Ϯ 0.4 ϫ 10 8 M Ϫ1 s Ϫ1 , which is close to the diffusion-controlled limit (10 9 M Ϫ1 s Ϫ1 ). Thus, the chemistry or turnover should provide the overall rate determining step. The effect of the N-terminal tail on turnover is significant, because there is no apparent turnover of N⌬100 MPG caused by product inhibition, whereas the latter is less significant for full-length MPG. Interestingly, the k off (from surface plasmon resonance studies) of two different forms of MPG for AP site are similar. However, the turnover (k ss ) of the truncated MPG is extremely slow compared with the wild type protein (Fig. 6). This could be explained by the apparent relative affinity of both the proteins toward Hx and AP site DNA (Figs. 4 and 7). Notably, k pd is composed of k off values for both free base and the AP site containing DNA. Moreover, the k ss consists of k pd and the effective rate constant (kЈ) for the process, MPG ϩ Hx-DNA3 MPG⅐AP site DNA (details discussed in Refs. 21,23). Generally, for obtaining the definite value for product FIGURE 5. Effect of N-terminal tail on MPG-Hx reaction under single-turnover conditions. The reaction was performed using different substrate and enzyme concentrations as described under "Materials and Methods." A, effect of N-terminal tail on pre-steady-state kinetic parameters under single-turnover conditions. B, data derived from A were analyzed using the first order rate equation: [P] t ϭ A 0 { 1 Ϫ exp(Ϫk obs t)} as described under "Results." k chem , catalytic constant at the chemistry step. dissociation from burst kinetics, it is assumed that k pd is much slower than the kЈ, and thus the contribution of the latter is ignored, and k ss becomes identical to k pd . But huge product inhibition in the case of N-terminally truncated MPG obscures that assumption, and therefore, an accurate k pd value for the truncated protein could not be assigned. Therefore, the overall effect on product formation is apparently arising from the alternations in the turnover step, and also the N-terminal extension is playing a critical role in that regard. N-terminally truncated hMPG also showed ϳ5-fold higher affinity toward AP sites than Hx (22).
The contribution of free base in the overall product dissociation cannot be ruled out. From the tertiary structures of MPG in complexes with ⑀A-containing DNA, it was suggested that the enzyme selects the substrate bases through a combination of hydrogen bonds and the steric constraints of the active site (18). Several other principles, such as shape and electrostatic properties of the binding pocket that accepts the flipped-out nucleotide, also dictate MPG to select its substrates. As revealed from crystal structures, the backbone amide of His-136 donates a hydrogen bond to the etheno N-6 nitrogen of ⑀A (18). This hydrogen bond can discriminate against adenine, which has a 6-amino group that cannot accept a hydrogen bond (18). Thus, this hydrogen bond may also be important in discrimination and excision of Hx and ⑀A and perhaps in the dissociation of the cleaved base.
Notably, the microscopic dissociation rates of full-length and N⌬100 MPGs from MPG-Hx complex (k off ) are comparable, whereas the k on of these proteins are significantly different. Full-length MPG binds Hx at a ϳ500-fold faster rate than N⌬100. The most probable explanation could be that MPG lacking N-terminal tail must collide with the substrates multiple times before engaging in a productive binding leading to catalysis. This indicates that without the N terminus, MPG possibly changes its conformation, or the N terminus in the full-length enzyme helps guide productive interaction with the modified base. Surprisingly, the differences in k on values for the subsequent product, an abasic site, are only ϳ10-fold for these two proteins. However, the apparent lack of an effect for the N terminus for AP site binding further underscores the importance of the N-terminal tail for substrate binding.
In the future, it would be interesting to study in detail the role of the N-terminal tail in substrate binding and perhaps in scanning and base flipping. Furthermore, the reinvestigation of biochemistry, kinetics, and structural analysis of full-length MPG (so far elucidated for truncated protein) with different substrates is undoubtedly important.