Structural insights by molecular dynamics simulations into specificity
http://www.100md.com
《核酸研究医学期刊》
Donner Laboratory, Life Sciences Division, Lawrence Berkeley National Laboratory, University of California, Berkeley, CA 94720, USA
*To whom correspondence should be addressed. Tel: +1 510 495 2537; Fax: +1 510 486 6488; Email: bo_hang@lbl.gov
Correspondence may also be addressed to B. Singer. Tel: +1 510 642 0637; Fax: +1 510 486 6488.
ABSTRACT
The benzetheno exocyclic adduct of the cytosine (C) base (pBQ-C) is a product of reaction between DNA and a stable metabolite of the human carcinogen benzene, p-benzoquinone (pBQ). We reported previously that the pBQ-C-containing duplex is a substrate for the human AP endonuclease (APE1), an enzyme that cleaves an apurinic/apyrimidinic (AP) site from double stranded DNA. In this work, using molecular dynamics simulation (MD), we provided a structural explanation for the recognition of the pBQ-C adduct by APE1. Molecular modeling of the DNA duplex containing pBQ-C revealed significant displacement of this adduct toward the major groove with pronounced kinking of the DNA at the lesion site, which could serve as a structural element recognized by the APE1 enzyme. Using 3 ns MD it was shown that the position of the pBQ-C adduct is stabilized by two hydrogen bonds formed between the adduct and the active site amino acids Asp 189 and Ala 175. The pBQ-C/APE1 complex, generated by MD, has a similar hydrogen bond network between target phosphodiester bond at the pBQ-C site and key amino acids at the active site, as in the crystallographically determined APE1 complexed with an AP site-containing DNA duplex. The position of the adduct at the enzyme active site, together with the hydrogen bond network, suggests a similar reaction mechanism for phosphodiester bond cleavage of oligonucleotide containing pBQ-C as reported for the AP site.
INTRODUCTION
Benzene has been known as an ubiquitous human carcinogen, which causes hematological disorders, such as acute myeloid leukemia (1). A large population is widely exposed to this carcinogen, which can be found in cigarette smoke, gasoline and automobile exhaust, and can be an occupational hazard in some professions (1–4). The carcinogenic activity of benzene appears to be partially related to its highly reactive metabolite, p-benzoquinone (pBQ) (5). This compound is the most potent mutagen of 12 structurally related simple benzoquinones tested in the Salmonella typhimurium based mutation reversion assay (6) and found to be also carcinogenic in rodents (5). Chemical studies showed that pBQ is highly reactive with DNA in vitro and leads to the formation of several two-ring benzetheno exocyclic base derivatives: 3,N4-benzetheno-2'-dC (pBQ-dC) (7), 1,N6-benzetheno-2'-dA (pBQ-dA) (8) and 1,N2-benzetheno-2'-dG (pBQ-dG) (9,10). The formation of these adducts in vivo, if found, would be a clear threat to the genome integrity. It has been shown by this laboratory that the HeLa cells contain a repair activity toward pBQ-C, p-BQ-A and pBQ-G incorporated in defined oligonucleotides (11,12). Later, it was unexpectedly found that this function belongs to the major human apurinic/apyrimidinic (AP) endonuclease (APE1), with the highest nicking and binding activity toward pBQ-C (Fig. 1A) (12–15). The APE1 enzyme has been known to primarily remove AP sites from DNA by hydrolyzing the phosphodiester backbone immediately 5' to the abasic nucleotide (16–18). An AP site is a frequently formed DNA lesion occurring either by spontaneous or damage-induced hydrolysis of the N-glycosyl bond or through the excision activity of a DNA glycosylase during base excision repair (BER) (19–22). Additional substrates for APE1 include strand breaks with 3' phosphate or 3' phosphoglycolate termini, some oxidative DNA lesions and also the RNA strand of RNA–DNA hybrids (23–29). Moreover, APE1 is known to modulate the binding activity of certain transcription factors toward DNA (30) and regulates the redox-state of cysteine residue in the target factor (31,32).
Figure 1. (A) Chemical structure of the pBQ-dC adduct produced by the reaction of the highly reactive benzene metabolite pBQ with deoxycytidine (dC). (B) Sequence of the 25mer DNA oligonucleotide used for the modeling. For the control pBQ-C was replaced by cytosine (C).
The ability of APE1 to cleave two apparently structurally unrelated substrates, an AP site and pBQ-C, raised interest in understanding the structural requirements for recognition and cleavage of the bulky pBQ adduct. It has been shown, using site-directed mutagenesis and enzymatic assays, that the cleavage of the pBQ-C-containing oligomers requires the same catalytic active site as that proposed for the AP site. However, the AP site remains the preferred substrate for APE1 (33). Recent crystallographic studies revealed the 3-D structure of the human APE1 alone as well as its complex with the abasic DNA (34,35). Based on the high-resolution structures of the APE1 bound to the 11 and 15mer DNA substrates and the ternary product complex of APE1 with nicked AP site, Mol et al. proposed a reaction mechanism for the phosphodiester bond cleavage, which required the Mg2+ ion in order to stabilize the transition state and to facilitate the O3' leaving group (34). Based on this mechanism, Asp 210 activates a hydroxyl nucleophile, which then attacks the target 5' phosphate. The role of Asp 210 in the catalytic function of this enzyme was also confirmed by mutation experiments (15). Recently, an independent study solved the crystal structure of the full length human APE1 at two different pHs. It was shown that at pH 7.5 there are two metal ions bound 5 ? apart in the active site (36). The authors proposed a two metal ion-mediated mechanism for the hydrolysis of the DNA backbone, which involves one of the metal ions in coordination of the attacking hydroxyl and the other ion in stabilizing the transition state with the 3' leaving group.
In this work, we investigated the structural aspects of the recognition and cleavage of a pBQ-C-containing oligonucleotide by human APE1 using molecular dynamics (MD) simulations. Recent advances in the molecular modeling procedures, which include full-scale solvation and explicit treatment of electrostatic interactions, allow us to gain a reliable structural data concerning conformation and molecular motion, which is in a reasonable agreement with the experiment (37–40). Based on MD simulations, we recently reported structural evidence for the reduced repair efficiency of 1,N6-ethanoadenine (EA) by human alkylpurine-DNA N-glycosylase, as compared with the structural related 1,N6-ethenoadenine (A) adduct (41). For this study, we employed MD using newly implemented polarizable force field (42–44). The enzyme activity toward pBQ-C is expected to be rather specific since this enzyme does not cleave other structurally related exocyclic adducts such as A, 3,N4-etheno-dC (C), 1,N2-etheno-dG (G) and 1,N6-propano-dG (33). The availability of the high-resolution structure of APE1 bound to the AP-containing DNA (34) enabled us to use it as a starting point for our molecular modeling. The analysis of the structural changes induced by the replacement of the AP site with pBQ-C at the enzyme active site should provide insights into how such a bulky adduct can fit into the active site for further processing by the APE1 protein. The favorable interaction between the adduct and the amino acids at the enzyme active site should be one of the key elements determining the recognition and cleavage of the pBQ-C adduct. However, a number of other structural factors, which could determine the initial binding/recognition of a particular adduct, could be determined by the conformational features of the adduct-containing duplexes. In order to evaluate the effects of the pBQ-C adduct on the local and global structural features of the DNA duplex, simulation of the pBQ-C·G-containing 25mer DNA duplex was also performed, since biochemical studies showed that maximal activity of APE1 was observed for the pBQ-C·G pair and minimal activity for the pBQ-C·C pair (33).
MATERIALS AND METHODS
Parameterization and MD simulation of the pBQ-C-containing 25mer DNA duplex
The pBQ-C adduct was built by addition of two exocyclic rings to the C nucleotide. A set of force field parameters and geometry optimized coordinates for the pBQ-C adduct was developed with ab initio quantum mechanical calculation using Spartan 5.0 suite (Wavefunction, Inc., Irvine, CA). Atom-centered charges were calculated with the RESP module of AMBER 7.0 using the partial charges obtained by Hartree-Fock calculations with 6-311G* basis set. The adduct was placed opposite G in the eighth position into the 25mer duplexes with identical sequences to the ones used in the biochemical studies (Fig. 1B) (33). A total of 200 ps of equilibration and 3 ns of unrestrained MD were carried out on the duplex-containing pBQ-C8·G43 pair and corresponding control (duplex with the C8·G43 base pair). Equilibrations and production runs were performed using explicit solvent with the TIP3P waters and 12 ? Lennard–Jones interaction distance cut-off according to the previously reported molecular modeling procedure (41,45). All calculations were performed with the SANDER module of AMBER 7.0 using a polarizable force field without extra points (no accounting for electron lone pairs) (42,43).
Preparation of the starting structure and MD simulations of the AP-DNA/APE1 and the pBQ-C-DNA/APE1 complexes
The high-resolution X-ray coordinates for the AP-DNA/APE1 complex (PDB code 1DEW ) served as the starting structure in our simulation. Hydrogen atoms were modeled using the xLeap module of AMBER 7.0. The force field parameters for the AP site were generated using a similar procedure as described for the pBQ-C adduct. To generate the starting structure for the pBQ-C-DNA/APE1 complex, the AP site was replaced by the geometry optimized pBQ-C adduct. Two sets of topology and coordinates files for the APE1/DNA complexes were generated using the xLeap module of AMBER 7.0. To overcome a direct overlap between the pBQ-C adduct and amino acids in the active site, and to minimize modeled hydrogens, the system was subjected to 100 steps of steepest descent (SD) followed by 200 steps of conjugate gradient minimization using the Hingerty distance-dependent dielectric function. The Mg2+ was modeled into the active site of the DNA/APE1 complex according to its position derived from the 3 ? resolution ternary product complex of APE1, nicked abasic DNA and a divalent metal ion (34). Twenty-five Na+ ions were placed around the complex at the positions of the minimum electrostatic potential of the system in order to neutralize negative charges. A rectangular box of the TIP3P water molecules was added, providing at least 10 ? of explicit solvent around each DNA/enzyme complex, which yielded 12 997 randomly oriented water molecules. The complete system consisted of approximately 44 285 atoms and has the initial dimensions 87.721, 81.320 and 77.345 ? in the x, y and z directions, respectively. The initial density of the water around the protein was 0.826 g/cm3. The system was first relaxed using several rounds of the short minimization and then the water box was subjected to a series of equilibration MD runs while holding the DNA/APE1 complex fixed as described previously (41,45). The system was then minimized using the five rounds of energy minimization under harmonic constraints, limiting the displacement of the solute atoms and counter-ions from their initial positions, with force constants 25, 20, 15, 10 and 5 kcal/mol ?, respectively, followed by 600 steps of unconstrained energy minimization. The system was then heated to 310 K in 10 ps and finally a production phase of 3 ns was started. The translational and rotational motions of the solute were removed every 5 ps. The Watson–Crick hydrogen bonds of the terminal bases were reinforced by means of 5 kcal/mol ? soft distance constraints in order to avoid possible opening of these base pairs. All calculations were performed using periodic boundary conditions. The electrostatic interactions were calculated with the particle-mesh-Ewald method (46,47) using 1 ? charge grid spacing with B-spline interpolation and a sum tolerance of 10–6 ?. A 12 ? cut-off was applied to Lennard–Jones interactions and SHAKE algorithm was used for all X-H bonds (48) with the 2 fs time step. The final structures representing the conformational family for the DNA/APE1 complexes were generated by averaging the MD trajectories based on root mean square deviation (RMSD) profiles (from 0.5 to 3 ns). The atom coordinates were stored every 1 ps.
Data analysis
The MD trajectories were processed using the analytical modules of AMBER 7.0 and Visual Molecular Dynamics (VMD) program (49). Nucleic acid structural parameters were obtained using CURVES 5.1 (50). All calculations were performed on a Silicon Graphics Origin 200 server interfaced with a dual processor Octane workstation (Silicon Graphics, Inc., Mountain View, CA). Each 3 ns production run for the DNA/APE1 complex required 42 days of the CPU time. The figures were generated using the VMD and Raster3D (51) software.
RESULTS
Conformation of the pBQ-C·G-containing duplex
Conformations of the pBQ-C·G-containing 25mer DNA duplex (pBQ-C·G-DNA) and the corresponding control were determined by 3 ns MD simulations. The conformational stability was evaluated by calculating all-atom RMSD values for the entire structure, lesion base pair with the flanking bases and lesion base pair alone. The evolution of RMSD of the pBQ-C·G-DNA as a function of the simulation time is shown in Figure 2. Based on RMSD values, the pBQ-C·G-DNA duplex reached conformational equilibrium after the first 750 ps and showed a plateau for these values for the rest of the simulation. Thus, the representative structures for this 25mer were generated by averaging MD trajectory from 0.75 to 3 ns. Large fluctuations of all atoms RMSD values are associated with the high flexibility and partial opening of the terminal base pairs (fraying effects). Calculation of the Watson–Crick hydrogen bond distances, including the 5'-GC and 3'-GC base pairs flanking the lesion site, shows complete integrity of the duplex with the 98–100% occupancy rate during the entire simulation. Terminal bases were not included in these calculations due to the fraying effects mentioned above. The conformation of the pBQ-C·G-DNA duplex was compared with the C·G-DNA, which was generated and evaluated using identical procedures. Figure 3 shows 3 and 10 bp motifs for averaged minimized structures of the C·G-DNA (left panel) and pBQ-C·G-DNA (right panel) generated by MD simulations. As expected, conformational differences between the control and modified duplex were localized at the lesion site. Structural perturbations induced by the pBQ-C adduct were analyzed by calculations of the helical parameters, which describe geometry and stacking interaction of the bases in the DNA ladder (52). Figure 4 shows average values for the intra- and inter-base pair parameters for pBQ-C·G-DNA in comparison with the C·G-DNA.
Figure 2. The evolution of the RMSD values from the minimized coordinates as a function of the simulation time of the different structural components for the 25mer DNA duplex-containing pBQ-C opposite G (A) and human AP endonuclease complexed to the pBQ-C-containing DNA duplex (B).
Figure 3. Top view on the 3 bp motifs (-G7X8G9/C44G43C42-, where X = C or pBQ-C) and the 10 bp motifs (-C4T5A6G7X8G9G10G11T12A13/G47A46T45C44G43C42C41C40A39T38-, where X = C or pBQ-C) side view for the C·G- and pBQ-C·G-containing DNA duplexes produced by the 3 ns MD simulations. The pBQ-C·G pair characterized by the displacement of the pBQ-C adduct toward major groove (right panel) as compared with the C·G (left panel). Yellow dashed lines show hydrogen bonding between in the C·G and pBQ-C·G pairs.
Figure 4. (A) Average values for the intra-base pair parameters describing the geometry of base pairing for the 5 bp in the C·G and pBQ-C·G-containing duplexes. (B) Average values for the inter-base pair parameters describing the stacking interactions for the 5 bp steps in the C·G and pBQ-C·G-containing duplexes. The tick marks on the x-axis indicate the base pair step. For example: label T5-A6 corresponds the T5-A46/A6-T45 base pair step.
The most noticeable structural contrast between the canonical Watson–Crick base pair and the pBQ-C·G pair, produced by MD, is the significant displacement of the adduct toward the major groove (Fig. 3, right panel). The magnitude of the displacement can be characterized by several helical parameters including shear (SHR), which measured 1.75 ? for the pBQ-C·G pair (Fig. 4A). Another two inter-base pair parameters, which are considerably affected by deposition of pBQ-C from the normal Watson–Crick pair, are the Stretch (STR) and Opening (OPN) parameters. For the pBQ-C8·G43 base pair the STR value, which shows separation between the bases along the y axis, was 1.53 ? and opening between the bases was measured at 54°. However, there was only a small vertical displacement of the lesion base pair, which is indicated by the slightly higher Rise value (3.9 ?) for the pBQ-C·G-DNA as compared with the rest of the duplex, which remained in the range for the canonical B-DNA (average value 3.3 ?) (Fig. 4B). A bifurcated hydrogen bond was observed between the pBQ-C O7 and NH2/NH of the opposite G (Fig. 3, left panel). The effect of the pBQ-C on the neighboring bases was monitored only up to 3 bp on both sides of the lesion. These structural distortions around the lesion were expressed in the large deviation from unmodified DNA in the Propeller (PRP), Tilt (TLT) and Roll (ROL) parameters, which best described the stacking interactions between the bases in the DNA duplex (Fig. 4A and B). The high magnitude for the Twist (TWS) parameters for the G7·C44/pBQ-C8·G43 base step (45°) and the lower value (11°) for the succeeding pBQ-C8·G43/G9·C42 step indicated partial untwisting of the DNA at the lesion site. The presence of the pBQ-C adduct at the DNA duplex increases local DNA curvature, which was measured for the first 15 bp for the pBQ-C·G-DNA and C·G-DNA duplexes. The curvature was calculated by searching for the optimal curved axis using the algorithm implemented in ‘Curves’ and the contribution from the terminal steps was omitted. The values for the C·G-DNA and pBQ-C·G-DNA were 14.5 and 34.8°, respectively (Fig. 3). The measurement of the sugar conformation for the pBQ-C adduct showed that its puckering falls into the C4'-exo range. The conformation around the glycosidic bond for the adduct remained in the -anti range during the entire simulation. The molecular motion of the pBQ-C adduct during 3 ns MD simulation was in comparable range with cytosine in the C·G-DNA duplex. The RMSD values for the pBQ-C adduct and C in C·G-DNA were 0.80 ± 0.14 and 0.77 ± 0.2 ?, respectively.
pBQ-C-induced conformational rearrangements in the APE1 active site
To study the effect of pBQ-C on the APE1 active site, we used crystal data of human APE1 bound to abasic DNA (AP-DNA/APE1) as a starting point in our modeling. The simple superimposition of the pBQ-C adduct over the AP site in the AP-DNA/APE1 complex revealed multiple steric clashes between pBQ-C and the amino acids at the enzyme active site. To overcome these clashes we performed structural refinement of the pBQ-C-DNA/APE1 complex using 3 ns of unrestrained MD simulation. The conformational rearrangements produced by the MD simulation in order to accommodate the pBQ-C adduct at the enzyme active site were evaluated to provide possible insights into interaction between this adduct and enzyme active site. In addition, these active site interactions in the pBQ-C-DNA/APE1 complex were evaluated against a previously proposed catalytic mechanism for this enzyme (34,53).
To validate our modeling approach, the AP-DNA/APE1 (PDB ID code 1DEW ) complex was first subjected to 3 ns simulations. The analysis of the complex conformation, as well as the position of the AP site in the active site, showed that the average minimized structure produced by our modeling deviates minimally from the crystal coordinates. The largest RMSD fluctuations were observed for the DNA duplex bound to enzyme (2.35 ± 0.4 ?) and associated with the high flexibility of the DNA termini. The average RMSD values for the complex were measured at 2.1 ± 0.2 ? for the all atoms system, 1.85 ± 0.15 ? for the protein alone and 1.03 ± 0.07 ? for the enzyme active site. All stacking and key hydrogen bond interactions unidentified in the crystallographic studies were intact during the MD simulations (34). Figure 5 shows APE1 interaction with the flipped-out AP site produced by molecular modeling. The average minimized structure generated by MD simulation maintained all necessary contacts between the amino acids in the active site and target phosphate atom, in order to orient and polarize the P-O3' bond (Fig. 5). MD also conserved the specific position of the Asp 210, which is involved in activation of an attacking hydroxyl nucleophile, as suggested by one of the proposed mechanisms (34). The orientation of the Asp 210 is coordinated by alignment with Asn 68 and Asn 212 amino acids (Fig. 5). The position of the divalent metal ion (Mg2+), coordinated with the Glu 96, was also unchanged during the simulation (data not shown).
Figure 5. Interaction between APE1 active site and the AP-containing DNA produced by 3 ns MD simulations. Only AP site with the flanking bases (the backbone indicated by a ribbon) and key amino acids are shown from the averaged minimized structure. The generated complex preserved hydrogen bond network (shown by green lines), identified in a crystallographically determined structure of the AP-DNA/APE1 complex. These interactions include: alignment of Asp 210 by Asn 68 and Asn 212; orientation of the target phosphate by hydrogen bonds between O5' and Asn 174, O2P and Asn 212 and His 309 and O1P atom. Arg 177 inserted between the bases flanking the AP site and forms hydrogen bond to the non-target AP site 3' phosphate.
The molecular modeling of the pBQ-C-DNA/APE1 complex started with several rounds of minimization. This initial step was used to reduce steric hindrance in the active site resulting from the replacement of the AP site by the geometry optimized pBQ-C adduct. After initial minimization, the system was subjected to equilibration followed by 3 ns MD simulation. The all atoms RMSD profiles for the pBQ-C-DNA/APE1 complex are shown in Figure 2B. All average RMSD values were <2.7 ?, with the value of 1.39 ± 0.08 ? for the enzyme active site, 1.96 ± 0.14 ? for the enzyme, 2.66 ± 0.47 ? for the bound DNA and 2.41 ± 0.24 ? for the entire complex. The higher average RMSD values for the pBQ-C-DNA/APE1 active site (1.39 ?), as compared with the AP-DNA/APE1 (1.03 ?), indicated that some structural rearrangement was necessary in order to accommodate the pBQ-C adduct. Figure 6 shows the selective interaction between the amino acids in the active site and the DNA phosphodiester backbone generated by MD simulations. Similar to the AP-DNA/APE1 complex, the position of the Asp 210 and Mg2+ atom remained uncharged relatively to the position of the target P-O3' bond. The Asp283/His309 pair was conserved during the simulation and His 309 N2 formed a hydrogen bond with the O1P atom (Fig. 6). However, the position of Asp 212, which formed hydrogen bond with the O2P of the target phosphate in the AP-DNA/APE1 complex, changed in the pBQ-C-DNA/APE1 active site. In this complex, the side chain of Asp 212 makes a direct hydrogen bond with the O5' atom. The O2P of the target phosphate makes a new hydrogen bond with the hydroxyl group of Tyr 171 (yellow lines in Fig. 6). The orientation of the Arg 177 remained unchanged in both complexes and its main chain amide forms a hydrogen bond with the O2P of the P-O5' bond. The position of the pBQ-C adduct in the APE1 active site was stabilized by the two hydrogen bonds. First, one was formed between the main chain amide of Ala 175 and pBQ-C O7 and the second one between the side chain of Asp 189 and hydroxyl group of the adduct (Fig. 7). The exocyclic rings of pBQ-C stack against the side chain of Arg 185, thus providing additional stabilization of the adduct at the enzyme active site (Fig. 7). The hydrophobic side of the pBQ-C sugar packs against the hydrophobic pocket formed by Phe 266, Trp 280 and Leu 282, which was identified previously from the crystallographic studies (34).
Figure 6. APE1 interactions with the target phosphodiester backbone at the pBQ-C site produced by 3 ns MD simulations. Green lines show hydrogen bond interactions, which are also observed in the AP-DNA/APE1 complex. The target phosphate is oriented by the hydrogen bond between the His 309 N2 and O1P, which was also reported for the AP-DNA/APE1 complex. Additionally the target phosphate is oriented by two new hydrogen bonds produced by structural rearrangement at the active site in order to accommodate the pBQ-C adduct (yellow lines between hydroxyl group of Tyr 171 and O2P and side chain of Asn 212 and O5' of the target phosphate). The position of Arg 177 was unchanged in the pBQ-C-DNA/APE1 complex, as compared with the AP-DNA/APE1 complex, and its side chain makes a hydrogen bond to the non-target 3' phosphate. The sugar moiety of the pBQ-C packs against Trp 290, which could provide initial contact with the adduct base thus facilitating the insertion of the pBQ-C into the active site.
Figure 7. Hydrogen bond interactions stabilizing the position of the pBQ-C at the APE1 active site as determined by molecular modeling. The main chain amide of Ala 175 hydrogen bonded to pBQ-C O7 and the side chain of Ala 189 hydrogen bonded to the hydroxyl group of adduct (green lines). The exocyclic ring of the pBQ-C stacks against the side chain of the Arg 185.
DISCUSSION
Removal of modified bases from the DNA ladder is one of the key mechanisms in defence by living cells against mutations. The major human AP endonuclease (APE1) plays an important role in protecting the integrity of the genome by incising the AP sites generated via glycosylase-catalyzed hydrolysis of the N-glycosidic bond in BER or by spontaneous, chemical-induced reactions (54–57). Previously, Hang at el. (13) showed that APE1 has the ability to act on structurally unrelated benzene-derived exocyclic adduct, pBQ-C. This endonuclease directly incises the oligonucleotide 5' to the pBQ-C adduct, leaving the adduct as a dangling base on the 5' side of the 3' fragment, suggesting for the first time, that this enzyme may be involved in nucleotide excision repair (13,14). The biochemical experiments using site-directed mutagenesis showed that the cleavage of pBQ-C-containing oligomer requires the same catalytic active site as that proposed for the AP function (33).
In this work we employed MD simulations to gain structural insights into recognition and cleavage of the pBQ-C adduct by human APE1. MD simulations of the pBQ-C-containing 25mer DNA duplex revealed a specific structural motif for the lesion base pair. Characteristic elements of this motif include partial untwisting of the DNA at the lesion site (20°), pronounced displacement of the adduct toward the major groove (X disp for the pBQ-C is 2 ?) and increased curvature around the lesion site (34°). These structural features could be one of the essential ingredients for the recognition of the pBQ-C adduct by the human APE1 protein. It has been shown that this enzyme does not form stable complexes with the unmodified DNA duplex or fully base paired, nicked DNA, which retains an unkinked conformation with the minimum distortion at the lesion site (58–60). Earlier it has been proposed that the extrahelical conformation of the deoxyribose moiety and increased molecular motion near the abasic site are required for the recognition of the AP site by APE1 (35,61). Our calculations showed similar molecular motion for pBQ-C as for the unmodified cytosine. The ability to form an extrahelical conformation for pBQ-C is likely to be constrained by the formation of the bifurcated hydrogen bond between the O7 of the adduct and the opposite G. However, the role of the displacement of the adduct toward the major groove, as one of the potential recognition signals, is in agreement with the biochemical data showing lower activity of this enzyme toward the pBQ-C·C pair, as compared with the pBQ-C·G pair (33). The bulky pBQ-C adduct, which occupies almost three times van der Waals space as compared with the normal C, could be accommodated more easily opposite the smaller C rather than G. Therefore, the pBQ-C·C base pair has less distortion around the lesion with the smaller displacement of the adduct toward the major groove (data not shown). The less evident changes induced by pBQ-C·C pair in the DNA duplex may make this lesion less recognizable by the enzyme, thus affecting the activity of APE1 toward the pBQ-C·C pair. The local conformational perturbations, observed by MD in the pBQ-C·G-containing 25mer duplexes, are overall in agreement with the thermodynamics studies reported previously for the same oligonucleotide (62). It has been shown that pBQ-C reduces thermal and thermodynamic stabilities of the 25mer DNA duplex, indicating the local conformational destabilization around the lesion site. The overall B-conformation for the duplex was only slightly altered as determined by circular dichroism measurements.
The conformational changes in the adduct-containing DNA duplex play an important role in the initial enzyme recognition. However, the interaction of the adduct with the enzyme active site should be crucial for the catalytic reaction. As shown by MD simulation in this work, pBQ-C can be accommodated at the APE1 active site with certain structural rearrangements. The hydrophobic pocket of the APE1 active site, particularly Trp 280 (Fig. 6), could provide initial contact/stacking with pBQ-C, facilitating the insertion of the adduct into enzyme active site in order to achieve proper alignment of the DNA backbone for the phosphate hydrolysis.
Currently, several catalytic reaction mechanisms have been proposed for the human APE1 enzyme. One of the mechanisms involves His 309 in generation of the hydroxyl nucleophile with the metal ion stabilizing the transition state intermediate and orienting the phosphate group (35). The second mechanism, based on the three co-crystal structures of the human APE1 bound to abasic DNA, proposed that Asp 210 acts as a Lewis base and generates the active site nucleophile, when the His 309 involves in orientation and polarization of the target P-03' bond (34). The metal ion, in this mechanism, stabilizes the leaving group. Based on the recent studies using the crystal structure of the APE1 at pH 7.5, another report proposed two divalent metal ions mechanism, in which one of the metals coordinates the attacking hydroxyl nucleophile and the other metal is involved in neutralizing of the pentacovalent intermediate and/or stabilizing the 3' leaving group (36). At this point we are not be able to specify which mechanism, proposed for the APE1 based on its activity toward AP site, could be utilized for the catalytic reaction with the pBQ-C adduct. However, the analysis of the interaction in the pBQ-C-DNA/APE1 active site showed all necessary contacts for the neutralization and orientation of the negatively charged oxygens of the target phosphodiester bond, which are essential for the catalytic reaction (Fig. 6). The interaction at the active site, such as positioning of the Asp 210 and Asp283/His309 pair remained unchanged relative to the target phosphate in the pBQ-C-DNA/APE1 complex as compared with the AP-DNA/APE1 crystal coordinates. Moreover, the pH dependence of the APE1 activity toward pBQ-C, with highest nicking efficiency between pH 6 and 6.5, supports the role of His 309, which remained protonated at the higher pH due to the formation of the catalytic diad with Asp 283.
Structural rearrangements produced by MD at the active site in order to accommodate the pBQ-C adduct, suggest a possible involvement of the hydroxyl group of Tyr 171 in stabilizing the target phosphate by making the hydrogen bond with the O2P (Fig. 6). The Asn 212 residue makes a contact with the O5' replacing the role of Asn 174, as determined for the AP-DNA/APE1 complex. However, the other side chain amine of Asn 212 remains hydrogen bonded with Asp 210 thus, together with Asn 68, stabilizing the position of Asp 210 in the pBQ-C-DNA/APE1 complex. Based on the structural data from this work, we propose that the position of the adduct at the enzyme active site, which is essential to support a suitable hydrogen bond network for the catalytic activity, could be stabilized by the two hydrogen bonds with Asp 189 and Ala 175 and also by stacking with the side chain of the Arg 185 (Fig. 7). It is tempting to speculate that the incapability of forming these stabilizing hydrogen bonds could be one of the reasons for the lack of activity of APE1 toward the adducts such as 1,N6-etheno-A, 1,N2-etheno-G, 1,N2-propano-G or 8-oxo-G. These adducts do not show correct positioning of the hydrogen bond acceptor, which is needed in order to form a hydrogen bond with the Ala 175. Moreover, the C base, which could form one of two hydrogen bonds (with Ala 175), may not form the correct structural motif in the DNA duplex in order to be recognized by this enzyme. Previous structural studies of the C-containing DNA duplexes, including MD simulations, showed less pronounced displacement of this adduct toward the major groove (SHR 1.38 versus 1.75 ? in the pBQ-C-containing duplex) (63,64).
In this work, based on the conformation of the pBQ-C-containing duplex together with elucidation of the active site interactions in the pBQ-C-DNA/APE1 complex, we provide a structural rationale for the recognition of the carcinogen generated, bulky pBQ-C adduct by APE1. The newly identified interaction between this adduct and amino acids in the APE1 active site, such as Asp 189, Ala 175 and Tyr 171, should aid in determination of the specificity of this enzyme toward the pBQ-C adduct as compared with the AP site. Future structural and biochemical work with the other substrates will be needed in order to expand understanding of the chemical details of the catalytic reaction mechanism and substrate specificity of this multifunctional enzyme.
ACKNOWLEDGEMENTS
This work was supported by NIH Grants CA 72079 (to B.H.) and CA 47723 (to B.S.) and was administered by the Lawrence Berkeley National Laboratory under Department of Energy contract DE-AC03-76SF00098.
REFERENCES
IARC (1987) Monographs on the Evaluation of the Carcinogenic Risks to Humans. International Agency for Research on Cancer, Lyon, France.
Goldstein,B.D. (1977) Benzene toxicity: a critical evaluation: hematotoxicity in humans. J. Toxicol. Environ., 69–105.
Hricko,A. (1994) Rings of controversy around benzene. Environ Health Perspect., 102, 276–281.
Wallace,L. (1996) Environmental exposure to benzene: an update. Environ Health Perspect., 104 (Suppl. 6), 1129–1136.
Huff,J.E., Eustis,S.L. and Haseman,J.K. (1989) Occurrence and relevance of chemically induced benign neoplasms in long-term carcinogenicity studies. Cancer Metastasis Rev., 8, 1–22.
Hakura,A., Mochida,H., Tsutsui,Y. and Yamatsu,K. (1995) Mutagenicity of benzoquinones for Ames Salmonella tester strains. Mutat. Res., 347, 37–43.
Pongracz,K., Kaur,S., Burlingame,A.L. and Bodell,W.J. (1990) Detection of (3'-hydroxy)-3,N4-benzetheno-2'-deoxycytidine-3'-phosphate by 32P-postlabeling of DNA reacted with p-benzoquinone. Carcinogenesis, 11, 1469–1472.
Pongracz,K. and Bodell,W.J. (1991) Detection of 3'-hydroxy-1,N6-benzetheno-2'-deoxyadenosine 3'-phosphate by 32P postlabeling of DNA reacted with p-benzoquinone. Chem. Res. Toxicol., 4, 199–202.
Jowa,L., Witz,G., Snyder,R., Winkle,S. and Kalf,G.F. (1990) Synthesis and characterization of deoxyguanosine-benzoquinone adducts. J. Appl. Toxicol., 10, 47–54.
Jowa,L., Winkle,S., Kalf,G., Witz,G. and Snyder,R. (1986) Deoxyguanosine adducts formed from benzoquinone and hydroquinone. Adv. Exp. Med. Biol., 197, 825–832.
Chenna,A., Hang,B., Rydberg,B., Kim,E., Pongracz,K., Bodell,W.J. and Singer,B. (1995) The benzene metabolite p-benzoquinone forms adducts with DNA bases that are excised by a repair activity from human cells that differs from an ethenoadenine glycosylase. Proc. Natl Acad. Sci. USA, 92, 5890–5894.
Hang,B. and Singer,B. (1998) In Aruoma,O.I. and Halliwell,B. (eds), DNA and Free Radicals: Techniques, Mechanisms and Applications. OICA International, London, 27–53.
Hang,B., Chenna,A., Fraenkel-Conrat,H. and Singer,B. (1996) An unusual mechanism for the major human apurinic/apyrimidinic (AP) endonuclease involving 5' cleavage of DNA containing a benzene-derived exocyclic adduct in the absence of an AP site. Proc. Natl Acad. Sci. USA, 93, 13737–13741.
Hang,B., Chenna,A., Sagi,J. and Singer,B. (1998) Differential cleavage of oligonucleotides containing the benzene-derived adduct, 1,N6-benzetheno-dA, by the major human AP endonuclease HAP1 and Escherichia coli exonuclease III and endonuclease IV. Carcinogenesis, 19, 1339–1343.
Rothwell,D.G., Hang,B., Gorman,M.A., Freemont,P.S., Singer,B. and Hickson,I.D. (2000) Substitution of Asp-210 in HAP1 (APE/Ref-1) eliminates endonuclease activity but stabilises substrate binding. Nucleic Acids Res., 28, 2207–2213.
Demple,B. and Harrison,L. (1994) Repair of oxidative damage to DNA: enzymology and biology. Annu. Rev. Biochem., 63, 915–948.
Barzilay,G., Walker,L.J., Rothwell,D.G. and Hickson,I.D. (1996) Role of the HAP1 protein in repair of oxidative DNA damage and regulation of transcription factors. Br. J. Cancer Suppl., 27, S145–150.
Kane,C.M. and Linn,S. (1981) Purification and characterization of an apurinic/apyrimidinic endonuclease from HeLa cells. J. Biol. Chem., 256, 3405–3414.
Lindahl,T. and Karlstrom,O. (1973) Heat-induced depyrimidination of deoxyribonucleic acid in neutral solution. Biochemistry, 12, 5151–5154.
Lindahl,T. and Nyberg,B. (1972) Rate of depurination of native deoxyribonucleic acid. Biochemistry, 11, 3610–3618.
Lindahl,T. (1993) Instability and decay of the primary structure of DNA. Nature, 362, 709–715.
Loeb,L.A. and Preston,B.D. (1986) Mutagenesis by apurinic/apyrimidinic sites. Annu. Rev. Genet., 20, 201–230.
Barzilay,G., Mol,C.D., Robson,C.N., Walker,L.J., Cunningham,R.P., Tainer,J.A. and Hickson,I.D. (1995) Identification of critical active-site residues in the multifunctional human DNA repair enzyme HAP1. Nat. Struct. Biol., 2, 561–568.
Barzilay,G., Walker,L.J., Robson,C.N. and Hickson,I.D. (1995) Site-directed mutagenesis of the human DNA repair enzyme HAP1: identification of residues important for AP endonuclease and RNase H activity. Nucleic Acids Res., 23, 1544–1550.
Wilson,D.M., 3rd, Takeshita,M., Grollman,A.P. and Demple,B. (1995) Incision activity of human apurinic endonuclease (Ape) at abasic site analogs in DNA. J. Biol. Chem., 270, 16002–16007.
Suh,D., Wilson,D.M., 3rd and Povirk,L.F. (1997) 3'-Phosphodiesterase activity of human apurinic/apyrimidinic endonuclease at DNA double-strand break ends. Nucleic Acids Res., 25, 2495–2500.
Winters,T.A., Weinfeld,M. and Jorgensen,T.J. (1992) Human HeLa cell enzymes that remove phosphoglycolate 3'-end groups from DNA. Nucleic Acids Res., 20, 2573–2580.
Ishchenko,A.A., Sanz,G., Privezentzev,C.V., Maksimenko,A.V. and Saparbaev,M. (2003) Characterisation of new substrate specificities of Escherichia coli and Saccharomyces cerevisiae AP endonucleases. Nucleic Acids Res., 31, 6344–6353.
Gros,L., Ishchenko,A.A., Ide,H., Elder,R.H. and Saparbaev,M.K. (2004) The major human AP endonuclease (Ape1) is involved in the nucleotide incision repair pathway. Nucleic Acids Res., 32, 73–81.
Xanthoudakis,S., Miao,G., Wang,F., Pan,Y.C. and Curran,T. (1992) Redox activation of Fos-Jun DNA binding activity is mediated by a DNA repair enzyme. EMBO J., 11, 3323–3335.
Xanthoudakis,S., Miao,G.G. and Curran,T. (1994) The redox and DNA-repair activities of Ref-1 are encoded by nonoverlapping domains. Proc. Natl Acad. Sci. USA, 91, 23–27.
Walker,L.J., Robson,C.N., Black,E., Gillespie,D. and Hickson,I.D. (1993) Identification of residues in the human DNA repair enzyme HAP1 (Ref-1) that are essential for redox regulation of Jun DNA binding. Mol. Cell. Biol., 13, 5370–5376.
Hang,B., Rothwell,D.G., Sagi,J., Hickson,I.D. and Singer,B. (1997) Evidence for a common active site for cleavage of an AP site and the benzene-derived exocyclic adduct, 3,N4-benzetheno-dC, in the major human AP endonuclease. Biochemistry, 36, 15411–15418.
Mol,C.D., Izumi,T., Mitra,S. and Tainer,J.A. (2000) DNA-bound structures and mutants reveal abasic DNA binding by APE1 and DNA repair coordination . Nature, 403, 451–456.
Gorman,M.A., Morera,S., Rothwell,D.G., de La Fortelle,E., Mol,C.D., Tainer,J.A., Hickson,I.D. and Freemont,P.S. (1997) The crystal structure of the human DNA repair endonuclease HAP1 suggests the recognition of extra-helical deoxyribose at DNA abasic sites. EMBO J., 16, 6548–6558.
Beernink,P.T., Segelke,B.W., Hadi,M.Z., Erzberger,J.P., Wilson,D.M.,3rd and Rupp,B. (2001) Two divalent metal ions in the active site of a new crystal form of human apurinic/apyrimidinic endonuclease, Ape1: implications for the catalytic mechanism. J. Mol. Biol., 307, 1023–1034.
Beveridge,D.L. and McConnell,K.J. (2000) Nucleic acids: theory and computer simulation, Y2K. Curr. Opin. Struct. Biol., 10, 182–196.
Arthanari,H., McConnell,K.J., Beger,R., Young,M.A., Beveridge,D.L. and Bolton,P.H. (2003) Assessment of the molecular dynamics structure of DNA in solution based on calculated and observed NMR NOESY volumes and dihedral angles from scalar coupling constants. Biopolymers, 68, 3–15.
Cheatham,T.E.,3rd and Young,M.A. (2000) Molecular dynamics simulation of nucleic acids: successes, limitations and promise. Biopolymers, 56, 232–256.
Kollman,P.A., Massova,I., Reyes,C., Kuhn,B., Huo,S., Chong,L., Lee,M., Lee,T., Duan,Y., Wang,W. et al. (2000) Calculating structures and free energies of complex molecules: combining molecular mechanics and continuum models. Acc. Chem. Res., 33, 889–897.
Guliaev,A.B., Hang,B. and Singer,B. (2002) Structural insights by molecular dynamics simulations into differential repair efficiency for ethano-A versus etheno-A adducts by the human alkylpurine-DNA N-glycosylase. Nucleic Acids Res., 30, 3778–3787.
Cieplak,P., Caldwell,J. and Kollman,P. (2001) Molecular mechanical models for organic and biological systems going beyond the atom centered two body additive approximation: aqueous solution free energies of methanol and N-methyl acetamide, nucleic acid base and amide hydrogen bonding and chloroform/water pertition coefficients of the nucleic acid bases. J. Computat. Chem., 22, 1048–1057.
Case,D.A., Pearlman,D.A., Caldwell,J.W., Cheatham,T.E., Wang,J., Ross,W.S., Simmerling,C.L., Darden,T.A., Merz,K.M., Stanton,R.V. et al. (2002) AMBER 7. University of California, San Francisco.
Grossfield,A., Ren,P. and Ponder,J.W. (2003) Ion solvation thermodynamics from simulation with a polarizable force field. J. Am. Chem. Soc., 125, 15671–15682.
Guliaev,A.B., Sagi,J. and Singer,B. (2000) Sequence-dependent conformational perturbation in DNA duplexes containing an epsilonA.T mismatch using molecular dynamics simulation. Carcinogenesis, 21, 1727–1736.
Darden,T.A. and Pedersen,L.G. (1993) Molecular modeling: an experimental tool. Environ. Health Perspect., 101, 410–412.
Essman,W.D., Luedtke,R.R., McGonigle,P. and Lucky,I. (1995) Variations in the behavioral responses to apomorphine in different strains of rats. Behav. Pharmacol., 6, 4–15.
Ryckaert,J.P., Ciccotti,G. and Berendsen,H.J.C. (1977) Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alcanes. J. Comput. Phys., 23, 327–341.
Humphrey,W., Dalke,A. and Schulten,K. (1996) VMD: visual molecular dynamics. J. Mol. Graph., 14, 33–38, 27–38.
Lavery,R. and Sklenar,H. (1996) CURVES5.1. Helical analysis of irregular nucleic acids. Laboratory for Theoretical Biochemistry.
Merrit,E.A. and Bacon,D.J. (1997) Raster3D-photorealistic molecular graphics. Methods Enzymol., 277, 505–524.
Neidle,S. (1999) Oxford Handbook of Nucleic Acid Structure. Oxford University Press, New York.
Wilson,D.M.,III and Barsky,D. (2001) The major human abasic endonuclease: formation, consequences and repair of abasic lesions in DNA. Mutat. Res., 485, 283–307.
Demple,B., Herman,T. and Chen,D.S. (1991) Cloning and expression of APE, the cDNA encoding the major human apurinic endonuclease: definition of a family of DNA repair enzymes. Proc. Natl Acad. Sci. USA, 88, 11450–11454.
Nakamura,J., Walker,V.E., Upton,P.B., Chiang,S.Y., Kow,Y.W. and Swenberg,J.A. (1998) Highly sensitive apurinic/apyrimidinic site assay can detect spontaneous and chemically induced depurination under physiological conditions. Cancer Res., 58, 222–225.
Robson,C.N. and Hickson,I.D. (1991) Isolation of cDNA clones encoding a human apurinic/apyrimidinic endonuclease that corrects DNA repair and mutagenesis defects in E. coli xth (exonuclease III) mutants. Nucleic Acids Res., 19, 5519–5523.
Seki,S., Hatsushika,M., Watanabe,S., Akiyama,K., Nagao,K. and Tsutsui,K. (1992) cDNA cloning, sequencing, expression and possible domain structure of human APEX nuclease homologous to Escherichia coli exonuclease III. Biochim. Biophys. Acta, 1131, 287–299.
Erzberger,J.P., Barsky,D., Scharer,O.D., Colvin,M.E. and Wilson,D.M.,3rd. (1998) Elements in abasic site recognition by the major human and Escherichia coli apurinic/apyrimidinic endonucleases. Nucleic Acids Res., 26, 2771–2778.
Strauss,P.R., Beard,W.A., Patterson,T.A. and Wilson,S.H. (1997) Substrate binding by human apurinic/apyrimidinic endonuclease indicates a Briggs-Haldane mechanism. J. Biol. Chem., 272, 1302–1307.
Evans,A.R., Limp-Foster,M. and Kelley,M.R. (2000) Going APE over ref-1. Mutat. Res., 461, 83–108.
Barsky,D., Foloppe,N., Ahmadia,S., Wilson,D.M.,3rd and MacKerell,A.D.,Jr (2000) New insights into the structure of abasic DNA from molecular dynamics simulations. Nucleic Acids Res., 28, 2613–2626.
Sagi,J., Chenna,A., Hang,B. and Singer,B. (1998) A single cyclic p-benzoquinone adduct can destabilize a DNA oligonucleotide duplex. Chem. Res. Toxicol., 11, 329–334.
Cullinan,D., Johnson,F. and de los Santos,C. (2000) Solution structure of an 11-mer duplex containing the 3,N-4-ethenocytosine adduct opposite 2'-deoxycytidine: Implications for the recognition of exocyclic lesions by DNA glycosylases. J. Mol. Biol., 296, 851–861.
Hang,B., Downing,G., Guliaev,A.B. and Singer,B. (2002) Novel activity of Escherichia coli mismatch uracil-DNA glycosylase (Mug) excising 8-(hydroxymethyl)-3,N4-ethenocytosine, a potential product resulting from glycidaldehyde reaction. Biochemistry, 41, 2158–2165.(Anton B. Guliaev, Bo Hang* and B. Singer)
*To whom correspondence should be addressed. Tel: +1 510 495 2537; Fax: +1 510 486 6488; Email: bo_hang@lbl.gov
Correspondence may also be addressed to B. Singer. Tel: +1 510 642 0637; Fax: +1 510 486 6488.
ABSTRACT
The benzetheno exocyclic adduct of the cytosine (C) base (pBQ-C) is a product of reaction between DNA and a stable metabolite of the human carcinogen benzene, p-benzoquinone (pBQ). We reported previously that the pBQ-C-containing duplex is a substrate for the human AP endonuclease (APE1), an enzyme that cleaves an apurinic/apyrimidinic (AP) site from double stranded DNA. In this work, using molecular dynamics simulation (MD), we provided a structural explanation for the recognition of the pBQ-C adduct by APE1. Molecular modeling of the DNA duplex containing pBQ-C revealed significant displacement of this adduct toward the major groove with pronounced kinking of the DNA at the lesion site, which could serve as a structural element recognized by the APE1 enzyme. Using 3 ns MD it was shown that the position of the pBQ-C adduct is stabilized by two hydrogen bonds formed between the adduct and the active site amino acids Asp 189 and Ala 175. The pBQ-C/APE1 complex, generated by MD, has a similar hydrogen bond network between target phosphodiester bond at the pBQ-C site and key amino acids at the active site, as in the crystallographically determined APE1 complexed with an AP site-containing DNA duplex. The position of the adduct at the enzyme active site, together with the hydrogen bond network, suggests a similar reaction mechanism for phosphodiester bond cleavage of oligonucleotide containing pBQ-C as reported for the AP site.
INTRODUCTION
Benzene has been known as an ubiquitous human carcinogen, which causes hematological disorders, such as acute myeloid leukemia (1). A large population is widely exposed to this carcinogen, which can be found in cigarette smoke, gasoline and automobile exhaust, and can be an occupational hazard in some professions (1–4). The carcinogenic activity of benzene appears to be partially related to its highly reactive metabolite, p-benzoquinone (pBQ) (5). This compound is the most potent mutagen of 12 structurally related simple benzoquinones tested in the Salmonella typhimurium based mutation reversion assay (6) and found to be also carcinogenic in rodents (5). Chemical studies showed that pBQ is highly reactive with DNA in vitro and leads to the formation of several two-ring benzetheno exocyclic base derivatives: 3,N4-benzetheno-2'-dC (pBQ-dC) (7), 1,N6-benzetheno-2'-dA (pBQ-dA) (8) and 1,N2-benzetheno-2'-dG (pBQ-dG) (9,10). The formation of these adducts in vivo, if found, would be a clear threat to the genome integrity. It has been shown by this laboratory that the HeLa cells contain a repair activity toward pBQ-C, p-BQ-A and pBQ-G incorporated in defined oligonucleotides (11,12). Later, it was unexpectedly found that this function belongs to the major human apurinic/apyrimidinic (AP) endonuclease (APE1), with the highest nicking and binding activity toward pBQ-C (Fig. 1A) (12–15). The APE1 enzyme has been known to primarily remove AP sites from DNA by hydrolyzing the phosphodiester backbone immediately 5' to the abasic nucleotide (16–18). An AP site is a frequently formed DNA lesion occurring either by spontaneous or damage-induced hydrolysis of the N-glycosyl bond or through the excision activity of a DNA glycosylase during base excision repair (BER) (19–22). Additional substrates for APE1 include strand breaks with 3' phosphate or 3' phosphoglycolate termini, some oxidative DNA lesions and also the RNA strand of RNA–DNA hybrids (23–29). Moreover, APE1 is known to modulate the binding activity of certain transcription factors toward DNA (30) and regulates the redox-state of cysteine residue in the target factor (31,32).
Figure 1. (A) Chemical structure of the pBQ-dC adduct produced by the reaction of the highly reactive benzene metabolite pBQ with deoxycytidine (dC). (B) Sequence of the 25mer DNA oligonucleotide used for the modeling. For the control pBQ-C was replaced by cytosine (C).
The ability of APE1 to cleave two apparently structurally unrelated substrates, an AP site and pBQ-C, raised interest in understanding the structural requirements for recognition and cleavage of the bulky pBQ adduct. It has been shown, using site-directed mutagenesis and enzymatic assays, that the cleavage of the pBQ-C-containing oligomers requires the same catalytic active site as that proposed for the AP site. However, the AP site remains the preferred substrate for APE1 (33). Recent crystallographic studies revealed the 3-D structure of the human APE1 alone as well as its complex with the abasic DNA (34,35). Based on the high-resolution structures of the APE1 bound to the 11 and 15mer DNA substrates and the ternary product complex of APE1 with nicked AP site, Mol et al. proposed a reaction mechanism for the phosphodiester bond cleavage, which required the Mg2+ ion in order to stabilize the transition state and to facilitate the O3' leaving group (34). Based on this mechanism, Asp 210 activates a hydroxyl nucleophile, which then attacks the target 5' phosphate. The role of Asp 210 in the catalytic function of this enzyme was also confirmed by mutation experiments (15). Recently, an independent study solved the crystal structure of the full length human APE1 at two different pHs. It was shown that at pH 7.5 there are two metal ions bound 5 ? apart in the active site (36). The authors proposed a two metal ion-mediated mechanism for the hydrolysis of the DNA backbone, which involves one of the metal ions in coordination of the attacking hydroxyl and the other ion in stabilizing the transition state with the 3' leaving group.
In this work, we investigated the structural aspects of the recognition and cleavage of a pBQ-C-containing oligonucleotide by human APE1 using molecular dynamics (MD) simulations. Recent advances in the molecular modeling procedures, which include full-scale solvation and explicit treatment of electrostatic interactions, allow us to gain a reliable structural data concerning conformation and molecular motion, which is in a reasonable agreement with the experiment (37–40). Based on MD simulations, we recently reported structural evidence for the reduced repair efficiency of 1,N6-ethanoadenine (EA) by human alkylpurine-DNA N-glycosylase, as compared with the structural related 1,N6-ethenoadenine (A) adduct (41). For this study, we employed MD using newly implemented polarizable force field (42–44). The enzyme activity toward pBQ-C is expected to be rather specific since this enzyme does not cleave other structurally related exocyclic adducts such as A, 3,N4-etheno-dC (C), 1,N2-etheno-dG (G) and 1,N6-propano-dG (33). The availability of the high-resolution structure of APE1 bound to the AP-containing DNA (34) enabled us to use it as a starting point for our molecular modeling. The analysis of the structural changes induced by the replacement of the AP site with pBQ-C at the enzyme active site should provide insights into how such a bulky adduct can fit into the active site for further processing by the APE1 protein. The favorable interaction between the adduct and the amino acids at the enzyme active site should be one of the key elements determining the recognition and cleavage of the pBQ-C adduct. However, a number of other structural factors, which could determine the initial binding/recognition of a particular adduct, could be determined by the conformational features of the adduct-containing duplexes. In order to evaluate the effects of the pBQ-C adduct on the local and global structural features of the DNA duplex, simulation of the pBQ-C·G-containing 25mer DNA duplex was also performed, since biochemical studies showed that maximal activity of APE1 was observed for the pBQ-C·G pair and minimal activity for the pBQ-C·C pair (33).
MATERIALS AND METHODS
Parameterization and MD simulation of the pBQ-C-containing 25mer DNA duplex
The pBQ-C adduct was built by addition of two exocyclic rings to the C nucleotide. A set of force field parameters and geometry optimized coordinates for the pBQ-C adduct was developed with ab initio quantum mechanical calculation using Spartan 5.0 suite (Wavefunction, Inc., Irvine, CA). Atom-centered charges were calculated with the RESP module of AMBER 7.0 using the partial charges obtained by Hartree-Fock calculations with 6-311G* basis set. The adduct was placed opposite G in the eighth position into the 25mer duplexes with identical sequences to the ones used in the biochemical studies (Fig. 1B) (33). A total of 200 ps of equilibration and 3 ns of unrestrained MD were carried out on the duplex-containing pBQ-C8·G43 pair and corresponding control (duplex with the C8·G43 base pair). Equilibrations and production runs were performed using explicit solvent with the TIP3P waters and 12 ? Lennard–Jones interaction distance cut-off according to the previously reported molecular modeling procedure (41,45). All calculations were performed with the SANDER module of AMBER 7.0 using a polarizable force field without extra points (no accounting for electron lone pairs) (42,43).
Preparation of the starting structure and MD simulations of the AP-DNA/APE1 and the pBQ-C-DNA/APE1 complexes
The high-resolution X-ray coordinates for the AP-DNA/APE1 complex (PDB code 1DEW ) served as the starting structure in our simulation. Hydrogen atoms were modeled using the xLeap module of AMBER 7.0. The force field parameters for the AP site were generated using a similar procedure as described for the pBQ-C adduct. To generate the starting structure for the pBQ-C-DNA/APE1 complex, the AP site was replaced by the geometry optimized pBQ-C adduct. Two sets of topology and coordinates files for the APE1/DNA complexes were generated using the xLeap module of AMBER 7.0. To overcome a direct overlap between the pBQ-C adduct and amino acids in the active site, and to minimize modeled hydrogens, the system was subjected to 100 steps of steepest descent (SD) followed by 200 steps of conjugate gradient minimization using the Hingerty distance-dependent dielectric function. The Mg2+ was modeled into the active site of the DNA/APE1 complex according to its position derived from the 3 ? resolution ternary product complex of APE1, nicked abasic DNA and a divalent metal ion (34). Twenty-five Na+ ions were placed around the complex at the positions of the minimum electrostatic potential of the system in order to neutralize negative charges. A rectangular box of the TIP3P water molecules was added, providing at least 10 ? of explicit solvent around each DNA/enzyme complex, which yielded 12 997 randomly oriented water molecules. The complete system consisted of approximately 44 285 atoms and has the initial dimensions 87.721, 81.320 and 77.345 ? in the x, y and z directions, respectively. The initial density of the water around the protein was 0.826 g/cm3. The system was first relaxed using several rounds of the short minimization and then the water box was subjected to a series of equilibration MD runs while holding the DNA/APE1 complex fixed as described previously (41,45). The system was then minimized using the five rounds of energy minimization under harmonic constraints, limiting the displacement of the solute atoms and counter-ions from their initial positions, with force constants 25, 20, 15, 10 and 5 kcal/mol ?, respectively, followed by 600 steps of unconstrained energy minimization. The system was then heated to 310 K in 10 ps and finally a production phase of 3 ns was started. The translational and rotational motions of the solute were removed every 5 ps. The Watson–Crick hydrogen bonds of the terminal bases were reinforced by means of 5 kcal/mol ? soft distance constraints in order to avoid possible opening of these base pairs. All calculations were performed using periodic boundary conditions. The electrostatic interactions were calculated with the particle-mesh-Ewald method (46,47) using 1 ? charge grid spacing with B-spline interpolation and a sum tolerance of 10–6 ?. A 12 ? cut-off was applied to Lennard–Jones interactions and SHAKE algorithm was used for all X-H bonds (48) with the 2 fs time step. The final structures representing the conformational family for the DNA/APE1 complexes were generated by averaging the MD trajectories based on root mean square deviation (RMSD) profiles (from 0.5 to 3 ns). The atom coordinates were stored every 1 ps.
Data analysis
The MD trajectories were processed using the analytical modules of AMBER 7.0 and Visual Molecular Dynamics (VMD) program (49). Nucleic acid structural parameters were obtained using CURVES 5.1 (50). All calculations were performed on a Silicon Graphics Origin 200 server interfaced with a dual processor Octane workstation (Silicon Graphics, Inc., Mountain View, CA). Each 3 ns production run for the DNA/APE1 complex required 42 days of the CPU time. The figures were generated using the VMD and Raster3D (51) software.
RESULTS
Conformation of the pBQ-C·G-containing duplex
Conformations of the pBQ-C·G-containing 25mer DNA duplex (pBQ-C·G-DNA) and the corresponding control were determined by 3 ns MD simulations. The conformational stability was evaluated by calculating all-atom RMSD values for the entire structure, lesion base pair with the flanking bases and lesion base pair alone. The evolution of RMSD of the pBQ-C·G-DNA as a function of the simulation time is shown in Figure 2. Based on RMSD values, the pBQ-C·G-DNA duplex reached conformational equilibrium after the first 750 ps and showed a plateau for these values for the rest of the simulation. Thus, the representative structures for this 25mer were generated by averaging MD trajectory from 0.75 to 3 ns. Large fluctuations of all atoms RMSD values are associated with the high flexibility and partial opening of the terminal base pairs (fraying effects). Calculation of the Watson–Crick hydrogen bond distances, including the 5'-GC and 3'-GC base pairs flanking the lesion site, shows complete integrity of the duplex with the 98–100% occupancy rate during the entire simulation. Terminal bases were not included in these calculations due to the fraying effects mentioned above. The conformation of the pBQ-C·G-DNA duplex was compared with the C·G-DNA, which was generated and evaluated using identical procedures. Figure 3 shows 3 and 10 bp motifs for averaged minimized structures of the C·G-DNA (left panel) and pBQ-C·G-DNA (right panel) generated by MD simulations. As expected, conformational differences between the control and modified duplex were localized at the lesion site. Structural perturbations induced by the pBQ-C adduct were analyzed by calculations of the helical parameters, which describe geometry and stacking interaction of the bases in the DNA ladder (52). Figure 4 shows average values for the intra- and inter-base pair parameters for pBQ-C·G-DNA in comparison with the C·G-DNA.
Figure 2. The evolution of the RMSD values from the minimized coordinates as a function of the simulation time of the different structural components for the 25mer DNA duplex-containing pBQ-C opposite G (A) and human AP endonuclease complexed to the pBQ-C-containing DNA duplex (B).
Figure 3. Top view on the 3 bp motifs (-G7X8G9/C44G43C42-, where X = C or pBQ-C) and the 10 bp motifs (-C4T5A6G7X8G9G10G11T12A13/G47A46T45C44G43C42C41C40A39T38-, where X = C or pBQ-C) side view for the C·G- and pBQ-C·G-containing DNA duplexes produced by the 3 ns MD simulations. The pBQ-C·G pair characterized by the displacement of the pBQ-C adduct toward major groove (right panel) as compared with the C·G (left panel). Yellow dashed lines show hydrogen bonding between in the C·G and pBQ-C·G pairs.
Figure 4. (A) Average values for the intra-base pair parameters describing the geometry of base pairing for the 5 bp in the C·G and pBQ-C·G-containing duplexes. (B) Average values for the inter-base pair parameters describing the stacking interactions for the 5 bp steps in the C·G and pBQ-C·G-containing duplexes. The tick marks on the x-axis indicate the base pair step. For example: label T5-A6 corresponds the T5-A46/A6-T45 base pair step.
The most noticeable structural contrast between the canonical Watson–Crick base pair and the pBQ-C·G pair, produced by MD, is the significant displacement of the adduct toward the major groove (Fig. 3, right panel). The magnitude of the displacement can be characterized by several helical parameters including shear (SHR), which measured 1.75 ? for the pBQ-C·G pair (Fig. 4A). Another two inter-base pair parameters, which are considerably affected by deposition of pBQ-C from the normal Watson–Crick pair, are the Stretch (STR) and Opening (OPN) parameters. For the pBQ-C8·G43 base pair the STR value, which shows separation between the bases along the y axis, was 1.53 ? and opening between the bases was measured at 54°. However, there was only a small vertical displacement of the lesion base pair, which is indicated by the slightly higher Rise value (3.9 ?) for the pBQ-C·G-DNA as compared with the rest of the duplex, which remained in the range for the canonical B-DNA (average value 3.3 ?) (Fig. 4B). A bifurcated hydrogen bond was observed between the pBQ-C O7 and NH2/NH of the opposite G (Fig. 3, left panel). The effect of the pBQ-C on the neighboring bases was monitored only up to 3 bp on both sides of the lesion. These structural distortions around the lesion were expressed in the large deviation from unmodified DNA in the Propeller (PRP), Tilt (TLT) and Roll (ROL) parameters, which best described the stacking interactions between the bases in the DNA duplex (Fig. 4A and B). The high magnitude for the Twist (TWS) parameters for the G7·C44/pBQ-C8·G43 base step (45°) and the lower value (11°) for the succeeding pBQ-C8·G43/G9·C42 step indicated partial untwisting of the DNA at the lesion site. The presence of the pBQ-C adduct at the DNA duplex increases local DNA curvature, which was measured for the first 15 bp for the pBQ-C·G-DNA and C·G-DNA duplexes. The curvature was calculated by searching for the optimal curved axis using the algorithm implemented in ‘Curves’ and the contribution from the terminal steps was omitted. The values for the C·G-DNA and pBQ-C·G-DNA were 14.5 and 34.8°, respectively (Fig. 3). The measurement of the sugar conformation for the pBQ-C adduct showed that its puckering falls into the C4'-exo range. The conformation around the glycosidic bond for the adduct remained in the -anti range during the entire simulation. The molecular motion of the pBQ-C adduct during 3 ns MD simulation was in comparable range with cytosine in the C·G-DNA duplex. The RMSD values for the pBQ-C adduct and C in C·G-DNA were 0.80 ± 0.14 and 0.77 ± 0.2 ?, respectively.
pBQ-C-induced conformational rearrangements in the APE1 active site
To study the effect of pBQ-C on the APE1 active site, we used crystal data of human APE1 bound to abasic DNA (AP-DNA/APE1) as a starting point in our modeling. The simple superimposition of the pBQ-C adduct over the AP site in the AP-DNA/APE1 complex revealed multiple steric clashes between pBQ-C and the amino acids at the enzyme active site. To overcome these clashes we performed structural refinement of the pBQ-C-DNA/APE1 complex using 3 ns of unrestrained MD simulation. The conformational rearrangements produced by the MD simulation in order to accommodate the pBQ-C adduct at the enzyme active site were evaluated to provide possible insights into interaction between this adduct and enzyme active site. In addition, these active site interactions in the pBQ-C-DNA/APE1 complex were evaluated against a previously proposed catalytic mechanism for this enzyme (34,53).
To validate our modeling approach, the AP-DNA/APE1 (PDB ID code 1DEW ) complex was first subjected to 3 ns simulations. The analysis of the complex conformation, as well as the position of the AP site in the active site, showed that the average minimized structure produced by our modeling deviates minimally from the crystal coordinates. The largest RMSD fluctuations were observed for the DNA duplex bound to enzyme (2.35 ± 0.4 ?) and associated with the high flexibility of the DNA termini. The average RMSD values for the complex were measured at 2.1 ± 0.2 ? for the all atoms system, 1.85 ± 0.15 ? for the protein alone and 1.03 ± 0.07 ? for the enzyme active site. All stacking and key hydrogen bond interactions unidentified in the crystallographic studies were intact during the MD simulations (34). Figure 5 shows APE1 interaction with the flipped-out AP site produced by molecular modeling. The average minimized structure generated by MD simulation maintained all necessary contacts between the amino acids in the active site and target phosphate atom, in order to orient and polarize the P-O3' bond (Fig. 5). MD also conserved the specific position of the Asp 210, which is involved in activation of an attacking hydroxyl nucleophile, as suggested by one of the proposed mechanisms (34). The orientation of the Asp 210 is coordinated by alignment with Asn 68 and Asn 212 amino acids (Fig. 5). The position of the divalent metal ion (Mg2+), coordinated with the Glu 96, was also unchanged during the simulation (data not shown).
Figure 5. Interaction between APE1 active site and the AP-containing DNA produced by 3 ns MD simulations. Only AP site with the flanking bases (the backbone indicated by a ribbon) and key amino acids are shown from the averaged minimized structure. The generated complex preserved hydrogen bond network (shown by green lines), identified in a crystallographically determined structure of the AP-DNA/APE1 complex. These interactions include: alignment of Asp 210 by Asn 68 and Asn 212; orientation of the target phosphate by hydrogen bonds between O5' and Asn 174, O2P and Asn 212 and His 309 and O1P atom. Arg 177 inserted between the bases flanking the AP site and forms hydrogen bond to the non-target AP site 3' phosphate.
The molecular modeling of the pBQ-C-DNA/APE1 complex started with several rounds of minimization. This initial step was used to reduce steric hindrance in the active site resulting from the replacement of the AP site by the geometry optimized pBQ-C adduct. After initial minimization, the system was subjected to equilibration followed by 3 ns MD simulation. The all atoms RMSD profiles for the pBQ-C-DNA/APE1 complex are shown in Figure 2B. All average RMSD values were <2.7 ?, with the value of 1.39 ± 0.08 ? for the enzyme active site, 1.96 ± 0.14 ? for the enzyme, 2.66 ± 0.47 ? for the bound DNA and 2.41 ± 0.24 ? for the entire complex. The higher average RMSD values for the pBQ-C-DNA/APE1 active site (1.39 ?), as compared with the AP-DNA/APE1 (1.03 ?), indicated that some structural rearrangement was necessary in order to accommodate the pBQ-C adduct. Figure 6 shows the selective interaction between the amino acids in the active site and the DNA phosphodiester backbone generated by MD simulations. Similar to the AP-DNA/APE1 complex, the position of the Asp 210 and Mg2+ atom remained uncharged relatively to the position of the target P-O3' bond. The Asp283/His309 pair was conserved during the simulation and His 309 N2 formed a hydrogen bond with the O1P atom (Fig. 6). However, the position of Asp 212, which formed hydrogen bond with the O2P of the target phosphate in the AP-DNA/APE1 complex, changed in the pBQ-C-DNA/APE1 active site. In this complex, the side chain of Asp 212 makes a direct hydrogen bond with the O5' atom. The O2P of the target phosphate makes a new hydrogen bond with the hydroxyl group of Tyr 171 (yellow lines in Fig. 6). The orientation of the Arg 177 remained unchanged in both complexes and its main chain amide forms a hydrogen bond with the O2P of the P-O5' bond. The position of the pBQ-C adduct in the APE1 active site was stabilized by the two hydrogen bonds. First, one was formed between the main chain amide of Ala 175 and pBQ-C O7 and the second one between the side chain of Asp 189 and hydroxyl group of the adduct (Fig. 7). The exocyclic rings of pBQ-C stack against the side chain of Arg 185, thus providing additional stabilization of the adduct at the enzyme active site (Fig. 7). The hydrophobic side of the pBQ-C sugar packs against the hydrophobic pocket formed by Phe 266, Trp 280 and Leu 282, which was identified previously from the crystallographic studies (34).
Figure 6. APE1 interactions with the target phosphodiester backbone at the pBQ-C site produced by 3 ns MD simulations. Green lines show hydrogen bond interactions, which are also observed in the AP-DNA/APE1 complex. The target phosphate is oriented by the hydrogen bond between the His 309 N2 and O1P, which was also reported for the AP-DNA/APE1 complex. Additionally the target phosphate is oriented by two new hydrogen bonds produced by structural rearrangement at the active site in order to accommodate the pBQ-C adduct (yellow lines between hydroxyl group of Tyr 171 and O2P and side chain of Asn 212 and O5' of the target phosphate). The position of Arg 177 was unchanged in the pBQ-C-DNA/APE1 complex, as compared with the AP-DNA/APE1 complex, and its side chain makes a hydrogen bond to the non-target 3' phosphate. The sugar moiety of the pBQ-C packs against Trp 290, which could provide initial contact with the adduct base thus facilitating the insertion of the pBQ-C into the active site.
Figure 7. Hydrogen bond interactions stabilizing the position of the pBQ-C at the APE1 active site as determined by molecular modeling. The main chain amide of Ala 175 hydrogen bonded to pBQ-C O7 and the side chain of Ala 189 hydrogen bonded to the hydroxyl group of adduct (green lines). The exocyclic ring of the pBQ-C stacks against the side chain of the Arg 185.
DISCUSSION
Removal of modified bases from the DNA ladder is one of the key mechanisms in defence by living cells against mutations. The major human AP endonuclease (APE1) plays an important role in protecting the integrity of the genome by incising the AP sites generated via glycosylase-catalyzed hydrolysis of the N-glycosidic bond in BER or by spontaneous, chemical-induced reactions (54–57). Previously, Hang at el. (13) showed that APE1 has the ability to act on structurally unrelated benzene-derived exocyclic adduct, pBQ-C. This endonuclease directly incises the oligonucleotide 5' to the pBQ-C adduct, leaving the adduct as a dangling base on the 5' side of the 3' fragment, suggesting for the first time, that this enzyme may be involved in nucleotide excision repair (13,14). The biochemical experiments using site-directed mutagenesis showed that the cleavage of pBQ-C-containing oligomer requires the same catalytic active site as that proposed for the AP function (33).
In this work we employed MD simulations to gain structural insights into recognition and cleavage of the pBQ-C adduct by human APE1. MD simulations of the pBQ-C-containing 25mer DNA duplex revealed a specific structural motif for the lesion base pair. Characteristic elements of this motif include partial untwisting of the DNA at the lesion site (20°), pronounced displacement of the adduct toward the major groove (X disp for the pBQ-C is 2 ?) and increased curvature around the lesion site (34°). These structural features could be one of the essential ingredients for the recognition of the pBQ-C adduct by the human APE1 protein. It has been shown that this enzyme does not form stable complexes with the unmodified DNA duplex or fully base paired, nicked DNA, which retains an unkinked conformation with the minimum distortion at the lesion site (58–60). Earlier it has been proposed that the extrahelical conformation of the deoxyribose moiety and increased molecular motion near the abasic site are required for the recognition of the AP site by APE1 (35,61). Our calculations showed similar molecular motion for pBQ-C as for the unmodified cytosine. The ability to form an extrahelical conformation for pBQ-C is likely to be constrained by the formation of the bifurcated hydrogen bond between the O7 of the adduct and the opposite G. However, the role of the displacement of the adduct toward the major groove, as one of the potential recognition signals, is in agreement with the biochemical data showing lower activity of this enzyme toward the pBQ-C·C pair, as compared with the pBQ-C·G pair (33). The bulky pBQ-C adduct, which occupies almost three times van der Waals space as compared with the normal C, could be accommodated more easily opposite the smaller C rather than G. Therefore, the pBQ-C·C base pair has less distortion around the lesion with the smaller displacement of the adduct toward the major groove (data not shown). The less evident changes induced by pBQ-C·C pair in the DNA duplex may make this lesion less recognizable by the enzyme, thus affecting the activity of APE1 toward the pBQ-C·C pair. The local conformational perturbations, observed by MD in the pBQ-C·G-containing 25mer duplexes, are overall in agreement with the thermodynamics studies reported previously for the same oligonucleotide (62). It has been shown that pBQ-C reduces thermal and thermodynamic stabilities of the 25mer DNA duplex, indicating the local conformational destabilization around the lesion site. The overall B-conformation for the duplex was only slightly altered as determined by circular dichroism measurements.
The conformational changes in the adduct-containing DNA duplex play an important role in the initial enzyme recognition. However, the interaction of the adduct with the enzyme active site should be crucial for the catalytic reaction. As shown by MD simulation in this work, pBQ-C can be accommodated at the APE1 active site with certain structural rearrangements. The hydrophobic pocket of the APE1 active site, particularly Trp 280 (Fig. 6), could provide initial contact/stacking with pBQ-C, facilitating the insertion of the adduct into enzyme active site in order to achieve proper alignment of the DNA backbone for the phosphate hydrolysis.
Currently, several catalytic reaction mechanisms have been proposed for the human APE1 enzyme. One of the mechanisms involves His 309 in generation of the hydroxyl nucleophile with the metal ion stabilizing the transition state intermediate and orienting the phosphate group (35). The second mechanism, based on the three co-crystal structures of the human APE1 bound to abasic DNA, proposed that Asp 210 acts as a Lewis base and generates the active site nucleophile, when the His 309 involves in orientation and polarization of the target P-03' bond (34). The metal ion, in this mechanism, stabilizes the leaving group. Based on the recent studies using the crystal structure of the APE1 at pH 7.5, another report proposed two divalent metal ions mechanism, in which one of the metals coordinates the attacking hydroxyl nucleophile and the other metal is involved in neutralizing of the pentacovalent intermediate and/or stabilizing the 3' leaving group (36). At this point we are not be able to specify which mechanism, proposed for the APE1 based on its activity toward AP site, could be utilized for the catalytic reaction with the pBQ-C adduct. However, the analysis of the interaction in the pBQ-C-DNA/APE1 active site showed all necessary contacts for the neutralization and orientation of the negatively charged oxygens of the target phosphodiester bond, which are essential for the catalytic reaction (Fig. 6). The interaction at the active site, such as positioning of the Asp 210 and Asp283/His309 pair remained unchanged relative to the target phosphate in the pBQ-C-DNA/APE1 complex as compared with the AP-DNA/APE1 crystal coordinates. Moreover, the pH dependence of the APE1 activity toward pBQ-C, with highest nicking efficiency between pH 6 and 6.5, supports the role of His 309, which remained protonated at the higher pH due to the formation of the catalytic diad with Asp 283.
Structural rearrangements produced by MD at the active site in order to accommodate the pBQ-C adduct, suggest a possible involvement of the hydroxyl group of Tyr 171 in stabilizing the target phosphate by making the hydrogen bond with the O2P (Fig. 6). The Asn 212 residue makes a contact with the O5' replacing the role of Asn 174, as determined for the AP-DNA/APE1 complex. However, the other side chain amine of Asn 212 remains hydrogen bonded with Asp 210 thus, together with Asn 68, stabilizing the position of Asp 210 in the pBQ-C-DNA/APE1 complex. Based on the structural data from this work, we propose that the position of the adduct at the enzyme active site, which is essential to support a suitable hydrogen bond network for the catalytic activity, could be stabilized by the two hydrogen bonds with Asp 189 and Ala 175 and also by stacking with the side chain of the Arg 185 (Fig. 7). It is tempting to speculate that the incapability of forming these stabilizing hydrogen bonds could be one of the reasons for the lack of activity of APE1 toward the adducts such as 1,N6-etheno-A, 1,N2-etheno-G, 1,N2-propano-G or 8-oxo-G. These adducts do not show correct positioning of the hydrogen bond acceptor, which is needed in order to form a hydrogen bond with the Ala 175. Moreover, the C base, which could form one of two hydrogen bonds (with Ala 175), may not form the correct structural motif in the DNA duplex in order to be recognized by this enzyme. Previous structural studies of the C-containing DNA duplexes, including MD simulations, showed less pronounced displacement of this adduct toward the major groove (SHR 1.38 versus 1.75 ? in the pBQ-C-containing duplex) (63,64).
In this work, based on the conformation of the pBQ-C-containing duplex together with elucidation of the active site interactions in the pBQ-C-DNA/APE1 complex, we provide a structural rationale for the recognition of the carcinogen generated, bulky pBQ-C adduct by APE1. The newly identified interaction between this adduct and amino acids in the APE1 active site, such as Asp 189, Ala 175 and Tyr 171, should aid in determination of the specificity of this enzyme toward the pBQ-C adduct as compared with the AP site. Future structural and biochemical work with the other substrates will be needed in order to expand understanding of the chemical details of the catalytic reaction mechanism and substrate specificity of this multifunctional enzyme.
ACKNOWLEDGEMENTS
This work was supported by NIH Grants CA 72079 (to B.H.) and CA 47723 (to B.S.) and was administered by the Lawrence Berkeley National Laboratory under Department of Energy contract DE-AC03-76SF00098.
REFERENCES
IARC (1987) Monographs on the Evaluation of the Carcinogenic Risks to Humans. International Agency for Research on Cancer, Lyon, France.
Goldstein,B.D. (1977) Benzene toxicity: a critical evaluation: hematotoxicity in humans. J. Toxicol. Environ., 69–105.
Hricko,A. (1994) Rings of controversy around benzene. Environ Health Perspect., 102, 276–281.
Wallace,L. (1996) Environmental exposure to benzene: an update. Environ Health Perspect., 104 (Suppl. 6), 1129–1136.
Huff,J.E., Eustis,S.L. and Haseman,J.K. (1989) Occurrence and relevance of chemically induced benign neoplasms in long-term carcinogenicity studies. Cancer Metastasis Rev., 8, 1–22.
Hakura,A., Mochida,H., Tsutsui,Y. and Yamatsu,K. (1995) Mutagenicity of benzoquinones for Ames Salmonella tester strains. Mutat. Res., 347, 37–43.
Pongracz,K., Kaur,S., Burlingame,A.L. and Bodell,W.J. (1990) Detection of (3'-hydroxy)-3,N4-benzetheno-2'-deoxycytidine-3'-phosphate by 32P-postlabeling of DNA reacted with p-benzoquinone. Carcinogenesis, 11, 1469–1472.
Pongracz,K. and Bodell,W.J. (1991) Detection of 3'-hydroxy-1,N6-benzetheno-2'-deoxyadenosine 3'-phosphate by 32P postlabeling of DNA reacted with p-benzoquinone. Chem. Res. Toxicol., 4, 199–202.
Jowa,L., Witz,G., Snyder,R., Winkle,S. and Kalf,G.F. (1990) Synthesis and characterization of deoxyguanosine-benzoquinone adducts. J. Appl. Toxicol., 10, 47–54.
Jowa,L., Winkle,S., Kalf,G., Witz,G. and Snyder,R. (1986) Deoxyguanosine adducts formed from benzoquinone and hydroquinone. Adv. Exp. Med. Biol., 197, 825–832.
Chenna,A., Hang,B., Rydberg,B., Kim,E., Pongracz,K., Bodell,W.J. and Singer,B. (1995) The benzene metabolite p-benzoquinone forms adducts with DNA bases that are excised by a repair activity from human cells that differs from an ethenoadenine glycosylase. Proc. Natl Acad. Sci. USA, 92, 5890–5894.
Hang,B. and Singer,B. (1998) In Aruoma,O.I. and Halliwell,B. (eds), DNA and Free Radicals: Techniques, Mechanisms and Applications. OICA International, London, 27–53.
Hang,B., Chenna,A., Fraenkel-Conrat,H. and Singer,B. (1996) An unusual mechanism for the major human apurinic/apyrimidinic (AP) endonuclease involving 5' cleavage of DNA containing a benzene-derived exocyclic adduct in the absence of an AP site. Proc. Natl Acad. Sci. USA, 93, 13737–13741.
Hang,B., Chenna,A., Sagi,J. and Singer,B. (1998) Differential cleavage of oligonucleotides containing the benzene-derived adduct, 1,N6-benzetheno-dA, by the major human AP endonuclease HAP1 and Escherichia coli exonuclease III and endonuclease IV. Carcinogenesis, 19, 1339–1343.
Rothwell,D.G., Hang,B., Gorman,M.A., Freemont,P.S., Singer,B. and Hickson,I.D. (2000) Substitution of Asp-210 in HAP1 (APE/Ref-1) eliminates endonuclease activity but stabilises substrate binding. Nucleic Acids Res., 28, 2207–2213.
Demple,B. and Harrison,L. (1994) Repair of oxidative damage to DNA: enzymology and biology. Annu. Rev. Biochem., 63, 915–948.
Barzilay,G., Walker,L.J., Rothwell,D.G. and Hickson,I.D. (1996) Role of the HAP1 protein in repair of oxidative DNA damage and regulation of transcription factors. Br. J. Cancer Suppl., 27, S145–150.
Kane,C.M. and Linn,S. (1981) Purification and characterization of an apurinic/apyrimidinic endonuclease from HeLa cells. J. Biol. Chem., 256, 3405–3414.
Lindahl,T. and Karlstrom,O. (1973) Heat-induced depyrimidination of deoxyribonucleic acid in neutral solution. Biochemistry, 12, 5151–5154.
Lindahl,T. and Nyberg,B. (1972) Rate of depurination of native deoxyribonucleic acid. Biochemistry, 11, 3610–3618.
Lindahl,T. (1993) Instability and decay of the primary structure of DNA. Nature, 362, 709–715.
Loeb,L.A. and Preston,B.D. (1986) Mutagenesis by apurinic/apyrimidinic sites. Annu. Rev. Genet., 20, 201–230.
Barzilay,G., Mol,C.D., Robson,C.N., Walker,L.J., Cunningham,R.P., Tainer,J.A. and Hickson,I.D. (1995) Identification of critical active-site residues in the multifunctional human DNA repair enzyme HAP1. Nat. Struct. Biol., 2, 561–568.
Barzilay,G., Walker,L.J., Robson,C.N. and Hickson,I.D. (1995) Site-directed mutagenesis of the human DNA repair enzyme HAP1: identification of residues important for AP endonuclease and RNase H activity. Nucleic Acids Res., 23, 1544–1550.
Wilson,D.M., 3rd, Takeshita,M., Grollman,A.P. and Demple,B. (1995) Incision activity of human apurinic endonuclease (Ape) at abasic site analogs in DNA. J. Biol. Chem., 270, 16002–16007.
Suh,D., Wilson,D.M., 3rd and Povirk,L.F. (1997) 3'-Phosphodiesterase activity of human apurinic/apyrimidinic endonuclease at DNA double-strand break ends. Nucleic Acids Res., 25, 2495–2500.
Winters,T.A., Weinfeld,M. and Jorgensen,T.J. (1992) Human HeLa cell enzymes that remove phosphoglycolate 3'-end groups from DNA. Nucleic Acids Res., 20, 2573–2580.
Ishchenko,A.A., Sanz,G., Privezentzev,C.V., Maksimenko,A.V. and Saparbaev,M. (2003) Characterisation of new substrate specificities of Escherichia coli and Saccharomyces cerevisiae AP endonucleases. Nucleic Acids Res., 31, 6344–6353.
Gros,L., Ishchenko,A.A., Ide,H., Elder,R.H. and Saparbaev,M.K. (2004) The major human AP endonuclease (Ape1) is involved in the nucleotide incision repair pathway. Nucleic Acids Res., 32, 73–81.
Xanthoudakis,S., Miao,G., Wang,F., Pan,Y.C. and Curran,T. (1992) Redox activation of Fos-Jun DNA binding activity is mediated by a DNA repair enzyme. EMBO J., 11, 3323–3335.
Xanthoudakis,S., Miao,G.G. and Curran,T. (1994) The redox and DNA-repair activities of Ref-1 are encoded by nonoverlapping domains. Proc. Natl Acad. Sci. USA, 91, 23–27.
Walker,L.J., Robson,C.N., Black,E., Gillespie,D. and Hickson,I.D. (1993) Identification of residues in the human DNA repair enzyme HAP1 (Ref-1) that are essential for redox regulation of Jun DNA binding. Mol. Cell. Biol., 13, 5370–5376.
Hang,B., Rothwell,D.G., Sagi,J., Hickson,I.D. and Singer,B. (1997) Evidence for a common active site for cleavage of an AP site and the benzene-derived exocyclic adduct, 3,N4-benzetheno-dC, in the major human AP endonuclease. Biochemistry, 36, 15411–15418.
Mol,C.D., Izumi,T., Mitra,S. and Tainer,J.A. (2000) DNA-bound structures and mutants reveal abasic DNA binding by APE1 and DNA repair coordination . Nature, 403, 451–456.
Gorman,M.A., Morera,S., Rothwell,D.G., de La Fortelle,E., Mol,C.D., Tainer,J.A., Hickson,I.D. and Freemont,P.S. (1997) The crystal structure of the human DNA repair endonuclease HAP1 suggests the recognition of extra-helical deoxyribose at DNA abasic sites. EMBO J., 16, 6548–6558.
Beernink,P.T., Segelke,B.W., Hadi,M.Z., Erzberger,J.P., Wilson,D.M.,3rd and Rupp,B. (2001) Two divalent metal ions in the active site of a new crystal form of human apurinic/apyrimidinic endonuclease, Ape1: implications for the catalytic mechanism. J. Mol. Biol., 307, 1023–1034.
Beveridge,D.L. and McConnell,K.J. (2000) Nucleic acids: theory and computer simulation, Y2K. Curr. Opin. Struct. Biol., 10, 182–196.
Arthanari,H., McConnell,K.J., Beger,R., Young,M.A., Beveridge,D.L. and Bolton,P.H. (2003) Assessment of the molecular dynamics structure of DNA in solution based on calculated and observed NMR NOESY volumes and dihedral angles from scalar coupling constants. Biopolymers, 68, 3–15.
Cheatham,T.E.,3rd and Young,M.A. (2000) Molecular dynamics simulation of nucleic acids: successes, limitations and promise. Biopolymers, 56, 232–256.
Kollman,P.A., Massova,I., Reyes,C., Kuhn,B., Huo,S., Chong,L., Lee,M., Lee,T., Duan,Y., Wang,W. et al. (2000) Calculating structures and free energies of complex molecules: combining molecular mechanics and continuum models. Acc. Chem. Res., 33, 889–897.
Guliaev,A.B., Hang,B. and Singer,B. (2002) Structural insights by molecular dynamics simulations into differential repair efficiency for ethano-A versus etheno-A adducts by the human alkylpurine-DNA N-glycosylase. Nucleic Acids Res., 30, 3778–3787.
Cieplak,P., Caldwell,J. and Kollman,P. (2001) Molecular mechanical models for organic and biological systems going beyond the atom centered two body additive approximation: aqueous solution free energies of methanol and N-methyl acetamide, nucleic acid base and amide hydrogen bonding and chloroform/water pertition coefficients of the nucleic acid bases. J. Computat. Chem., 22, 1048–1057.
Case,D.A., Pearlman,D.A., Caldwell,J.W., Cheatham,T.E., Wang,J., Ross,W.S., Simmerling,C.L., Darden,T.A., Merz,K.M., Stanton,R.V. et al. (2002) AMBER 7. University of California, San Francisco.
Grossfield,A., Ren,P. and Ponder,J.W. (2003) Ion solvation thermodynamics from simulation with a polarizable force field. J. Am. Chem. Soc., 125, 15671–15682.
Guliaev,A.B., Sagi,J. and Singer,B. (2000) Sequence-dependent conformational perturbation in DNA duplexes containing an epsilonA.T mismatch using molecular dynamics simulation. Carcinogenesis, 21, 1727–1736.
Darden,T.A. and Pedersen,L.G. (1993) Molecular modeling: an experimental tool. Environ. Health Perspect., 101, 410–412.
Essman,W.D., Luedtke,R.R., McGonigle,P. and Lucky,I. (1995) Variations in the behavioral responses to apomorphine in different strains of rats. Behav. Pharmacol., 6, 4–15.
Ryckaert,J.P., Ciccotti,G. and Berendsen,H.J.C. (1977) Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alcanes. J. Comput. Phys., 23, 327–341.
Humphrey,W., Dalke,A. and Schulten,K. (1996) VMD: visual molecular dynamics. J. Mol. Graph., 14, 33–38, 27–38.
Lavery,R. and Sklenar,H. (1996) CURVES5.1. Helical analysis of irregular nucleic acids. Laboratory for Theoretical Biochemistry.
Merrit,E.A. and Bacon,D.J. (1997) Raster3D-photorealistic molecular graphics. Methods Enzymol., 277, 505–524.
Neidle,S. (1999) Oxford Handbook of Nucleic Acid Structure. Oxford University Press, New York.
Wilson,D.M.,III and Barsky,D. (2001) The major human abasic endonuclease: formation, consequences and repair of abasic lesions in DNA. Mutat. Res., 485, 283–307.
Demple,B., Herman,T. and Chen,D.S. (1991) Cloning and expression of APE, the cDNA encoding the major human apurinic endonuclease: definition of a family of DNA repair enzymes. Proc. Natl Acad. Sci. USA, 88, 11450–11454.
Nakamura,J., Walker,V.E., Upton,P.B., Chiang,S.Y., Kow,Y.W. and Swenberg,J.A. (1998) Highly sensitive apurinic/apyrimidinic site assay can detect spontaneous and chemically induced depurination under physiological conditions. Cancer Res., 58, 222–225.
Robson,C.N. and Hickson,I.D. (1991) Isolation of cDNA clones encoding a human apurinic/apyrimidinic endonuclease that corrects DNA repair and mutagenesis defects in E. coli xth (exonuclease III) mutants. Nucleic Acids Res., 19, 5519–5523.
Seki,S., Hatsushika,M., Watanabe,S., Akiyama,K., Nagao,K. and Tsutsui,K. (1992) cDNA cloning, sequencing, expression and possible domain structure of human APEX nuclease homologous to Escherichia coli exonuclease III. Biochim. Biophys. Acta, 1131, 287–299.
Erzberger,J.P., Barsky,D., Scharer,O.D., Colvin,M.E. and Wilson,D.M.,3rd. (1998) Elements in abasic site recognition by the major human and Escherichia coli apurinic/apyrimidinic endonucleases. Nucleic Acids Res., 26, 2771–2778.
Strauss,P.R., Beard,W.A., Patterson,T.A. and Wilson,S.H. (1997) Substrate binding by human apurinic/apyrimidinic endonuclease indicates a Briggs-Haldane mechanism. J. Biol. Chem., 272, 1302–1307.
Evans,A.R., Limp-Foster,M. and Kelley,M.R. (2000) Going APE over ref-1. Mutat. Res., 461, 83–108.
Barsky,D., Foloppe,N., Ahmadia,S., Wilson,D.M.,3rd and MacKerell,A.D.,Jr (2000) New insights into the structure of abasic DNA from molecular dynamics simulations. Nucleic Acids Res., 28, 2613–2626.
Sagi,J., Chenna,A., Hang,B. and Singer,B. (1998) A single cyclic p-benzoquinone adduct can destabilize a DNA oligonucleotide duplex. Chem. Res. Toxicol., 11, 329–334.
Cullinan,D., Johnson,F. and de los Santos,C. (2000) Solution structure of an 11-mer duplex containing the 3,N-4-ethenocytosine adduct opposite 2'-deoxycytidine: Implications for the recognition of exocyclic lesions by DNA glycosylases. J. Mol. Biol., 296, 851–861.
Hang,B., Downing,G., Guliaev,A.B. and Singer,B. (2002) Novel activity of Escherichia coli mismatch uracil-DNA glycosylase (Mug) excising 8-(hydroxymethyl)-3,N4-ethenocytosine, a potential product resulting from glycidaldehyde reaction. Biochemistry, 41, 2158–2165.(Anton B. Guliaev, Bo Hang* and B. Singer)