- Research article
- Open Access
Structural insights on the pamoic acid and the 8 kDa domain of DNA polymerase beta complex: Towards the design of higher-affinity inhibitors
© Hazan et al; licensee BioMed Central Ltd. 2008
Received: 12 December 2007
Accepted: 16 April 2008
Published: 16 April 2008
DNA polymerase beta (pol beta), the error-prone DNA polymerase of single-stranded DNA break repair as well as base excision repair pathways, is overexpressed in several tumors and takes part in chemotherapeutic agent resistance, like that of cisplatin, through translesion synthesis. For this reason pol beta has become a therapeutic target. Several inhibitors have been identified, but none of them presents a sufficient affinity and specificity to become a drug. The fragment-based inhibitor design allows an important improvement in affinity of small molecules. The initial and critical step for setting up the fragment-based strategy consists in the identification and structural characterization of the first fragment bound to the target.
We have performed docking studies of pamoic acid, a 9 micromolar pol beta inhibitor, and found that it binds in a single pocket at the surface of the 8 kDa domain of pol beta. However, docking studies provided five possible conformations for pamoic acid in this site. NMR experiments were performed on the complex to select a single conformation among the five retained. Chemical Shift Mapping data confirmed pamoic acid binding site found by docking while NOESY and saturation transfer experiments provided distances between pairs of protons from the pamoic acid and those of the 8 kDa domain that allowed the identification of the correct conformation.
Combining NMR experiments on the complex with docking results allowed us to build a three-dimensional structural model. This model serves as the starting point for further structural studies aimed at improving the affinity of pamoic acid for binding to DNA polymerase beta.
DNA polymerase β (pol beta) is the smallest human DNA polymerase and the first discovered adaptative polymerase. The structure of the full-length protein of 39 kDa has been solved by X ray cristallography . It is divided into two structural subdomains, which have distinct functions. The first is termed the 8 kDa domain and is located at the N-terminal position. It binds to single-stranded and double stranded DNA, recognizes the 5'-phosphate group in gapped DNA and possesses dRP lyase activity . The second, the C-terminal 31 kDa domain, displays the catalytic activity of polymerisation .
Pol beta is the major enzyme of the single-stranded break DNA repair and base excision repair pathways [4–6]. It is able to remove damaged base residues, nucleotides and abasic sites arising from various endogenous and exogenous sources . Thereby, when pol beta gene is deleted from mouse fibroblasts, hypersensitivity to monofunctional alkylation agents, e.g., methylmethanesulfonate, is observed [8, 9]. Furthermore, pol beta is able to bypass DNA lesions which block the replication by the replicative DNA polymerases. In fact, pol beta allows an error-prone translesion replication of some adducts, like those generated by cisplatin [10–13]. A large and readily adaptable binding site and a lack of 3'-5'-exonuclease activity facilitates synthesis through lesions. Hence, pol beta seems to be involved in chemotherapeutic agent resistance, as its overexpression diminishes the efficacy of anticancer drug therapies using cisplatin . Indeed, in breast, colon and prostate tumors, a cisplatin resistance is often observed while pol beta is overexpressed .
Since it causes genetic instability and resistance to anticancer drugs, pol beta is a therapeutic target. Within the last few years, there has been considerable effort to find inhibitors of pol beta with higher affinity and greater specificity [16–18].
Among the approaches that aim at conceiving inhibitors, the fragment-based drug design is a recent one, that has proven to be successful . Briefly, in this approach, two weakly binding ligands or fragments of the target are sought so that they can be covalently bound to obtain a final molecule, whose affinity is roughly the product of individual affinities of the two fragments .
To use this approach, sizeable structural information is required, in particular orientation and conformation of both fragments in their respective sites on the target. This information, together with the distance between fragments, is used to define a linker capable of connecting both fragments without disrupting the global affinity. Nuclear Magnetic Resonance (NMR) spectroscopy is a very powerful structural technique in the fragment-based strategy, as it can detect weakly interacting fragments and provide distances between target and ligand protons and structural information on the interaction. In the context of searching a pol beta inhibitor, we concentrated our efforts on the 8 kDa domain, defined by a protease sensitive hinge region at Lys87. Its structure has been solved by NMR [21, 22]. It is formed by 4 alpha helices, packed as two antiparallel pairs. The helices 1 (from Gly13 to Val 29) and 2 (from Ile33 to Lys48) are linked by a 4 amino-acid loop. A "helix-hairpin-helix" motif links helix 3 (from Ser55 to Lys61) to helix 4 (from Thr67 to Thr79), which is responsible for non-specific DNA binding [23, 24]. The alpha helix from Arg83 to Gln90 formed in the full-length protein is not folded in the 8 kDa domain. The first step of polymerisation process by pol beta involves the 8 kDa domain. Moreover, the alkylation-sensitive phenotype can be rescued by expression of the 8 kDa domain, suggesting that removal of the dRP group is most critical during base excision repair .
Among the large number of pol beta inhibitors that have been reported so far, the pamoic acid binds to the 8 kDa domain with a reasonable affinity . Its solubility in aqueous buffer makes it interesting for further NMR analysis. Moreover, the presence of hydroxyl and carboxyl groups allows simple chemical modifications, aiming at tethering other fragments.
This paper describes structural data obtained on the complex of pamoic acid with the 8 kDa domain. A combination of computational approaches and experimental data obtained from NMR provided the exact orientation of pamoic acid bound to the 8 kDa domain, allowing to propose a three-dimensional model of the complex. We found that pamoic acid binds to a site formed by helix 2 and helix 4, which also corresponds to the single-stranded DNA binding site. The conformation of the bound pamoic acid affords possible binding of a second fragment in its vicinity. The current work is the starting point to apply the fragment-based strategy, that is to say identifying a second fragment that binds in the vicinity of the pamoic acid site.
Results and Discussion
Docking pamoic acid to the 8 kDa domain
We used AutoDock 3.0.5  to dock pamoic acid to the 8 kDa domain of pol beta (PDB code 1DK3). Structures generated by AutoDock have been ranked according to their binding energy with the protein and 100 lowest energy structures were retained for further analysis. With the force field used by AutoDock, the energy values for the best ligands varied from -9.58 kcal/mol to -8.96 kcal/mol.
Two other sites can be defined in the proximity of the pamoic binding site. The first site could be located further beyond Lys35, between helices 1 and 4 and neighboring Leu19, Leu22, Glu26 or Lys72. The other potential site could be related to Lys60, Leu62 and Ala70, further in the direction of Gly64 and Gly66, which belong to pamoic acid site. Since we intend to modify one of the carboxyl groups of pamoic acid to tether a second fragment to reach one of the other two sites, the precise orientation of pamoic acid in the binding site has to be determined.
Chemical Shift Mapping confirms the pamoic acid binding site found by docking
A series of 2D 1H-15N HSQC spectra was monitored upon addition of pamoic acid. As the spectra could be easily superimposed, we concluded that the structure of the 8 kDa domain is not altered upon ligand addition. Chemical shifts of the crosspeaks have been followed during titration (Fig. 3a). The peak positions changed between those characteristic of the free and bound forms, indicating fast exchange on the NMR chemical shift timescale.
Comparison of the two spectra recorded for the protein in its free state and in the presence of a 4-fold molar excess of pamoic acid revealed a small number of amino-acids that are affected upon ligand binding. The chemical shift changes D (cf. NMR chemical shift mapping of Methods section) of the same crosspeak between the free and the ligand bound state were plotted versus the pol beta residue numbers (Fig. 3b). Amide groups of Ala42, Gly64, Gly66 and Ile69 showed a significant D value (≥ 3 × standard deviation), indicating that these amino-acids belong to the binding site of pamoic acid. Previous studies reported that Lys35, Tyr39, Arg83 and Leu85 were also affected by binding . To characterize binding, we used a threshold for D value equal to 3 × standard deviation instead of 1.5 as in the paper . With the multiplier set to 1.5, Lys35, Arg83 and Leu85 were also found to be affected by binding in our experiments. Interestingly, chemical shift mapping at higher protein concentration revealed that Lys35, Tyr39, Arg83 and Leu85 were also affected by binding (data not shown). However, Arg83 and Leu85 are located far from the other affected residues, on an unstructured loop, which is a part of an alpha helix in the entire 39 kD protein. These residues are therefore unlikely to be involved in the binding of pamoic acid to the entire protein.
Furthermore, comparing the results of the in silico docking with those obtained from chemical shift mapping showed a good agreement. All retained docked structures are located in the site revealed by Chemical Shift Mapping experiments, the one which corresponds to the single-stranded DNA binding. Among the five conformations proposed by AutoDock, conformation n°5 could be rejected as it was not close enough to Gly64 and Gly66, which have been shown to be involved in pamoic acid binding. Therefore, at this step, four conformations were retained.
A NOE correlation helps in determining the correct conformation
In order to further discriminate the right position, we looked for intermolecular NOEs. 3D 15N-TOCSY-HSQC (TOtal Correlated SpectroscopY) and 3D 15N-NOESY-HSQC (Nuclear Overhauser Enhancement SpectroscopY) spectra of the free 15N-labeled 8 kDa domain were acquired and 3D 15N-NOESY-HSQC spectrum was recorded in the presence of equimolar pamoic acid concentration.
The atomic contacts were computed from the four retained structures by assuming that the maximal distance between a pair of protein-ligand atoms does not exceed 6 Å. Among these distances we selected the ones corresponding to the amide proton of Ile69 and the two Ha protons of pamoic acid. The 55 best structures that fulfilled this condition were visually inspected and showed that three different conformations were admissible. More than half of the 55 structures corresponded to conformation n°1 depicted on Fig. 2a. The two other conformations represented 30% (conformation n°2, Fig. 2b) and 18% (conformation n°3, Fig. 2c) of the full AutoDock structure set.
Two important protons for the interaction between the 8 kDa domain and pamoic acid
In order to get further structural information on the binding, Saturation Transfer Difference (STD) experiments were carried out. The principle of this technique is to transfer the saturation from the protein to the bound ligand . Protein protons that are shifted outside of the spectral window of low molecular weight compounds are selectively irradiated. Within a short time, the saturation is spread over the entire 8 kDa domain through spin diffusion. The spin diffusion is more efficient when the molecular mobility is restricted, which is the case for large molecules like proteins. The spin diffusion may also occur between a protein and its ligand, assuming they are close enough. The closer the distance from the binding site, the larger the saturation, indicating the ligand's binding epitope .
Searching through the computed atomic contacts, we analyzed the distances between Hd or He and protein atoms and found that two of the three remaining conformations did not feature sufficiently close contacts to give rise to STD signals, namely conformations n°1 and n°3. Moreover, ligand protons that are closer to the protein than Hd and He in these conformations did not give STD signals. By contrast, every structure in conformation n°2 family showed at least one distance below 2,6 Å between these two protons of pamoic acid and one of the protein protons.
STD experiments were used to assess ligand affinity, as the ligand has been added gradually up to around a 50-fold molar excess with respect to the protein concentration. As STD technique requires low concentrations in ligand and protein, it is less affected by interferences due to additional non-specific binding visible in Kd measurements . STD amplification factor was plotted versus the ligand excess. Assuming a stoichiometric complex, fitting the curve of the STD amplification factor of the major proton He provided the Kd value of 13 μ M ± 5 μ M (Fig. 5b) .
Model of the interaction between the pamoic acid and the 8 kDa domain
Under these assumptions, we analysed the contacts between the validated ligand structure and the 8 kDa domain atoms. The aromatic groups of pamoic acid probably form favorable hydrophobic interactions with the main amino-acids of the site, such as Tyr39, Ala42, Gly64 and Gly66. Amide protons of Gly64 and Gly66 are shielded upon pamoic acid addition. This can be correlated with the fact that both residues are affected by magnetic anisotropy of the naphtalene nucleus . Furthermore, numerous lysine residues present in the site can form electrostatic interactions with both carboxyl groups. One of the carboxyl groups is oriented towards His34 and Lys35. It makes close contacts with Ile69 amide proton and electrostatic interaction with the terminal group of the Lys68 sidechain. The other carboxyl group forms hydrogen bonds with the amide proton of Lys68 (distance of 1,67 Å) and with the hydroxyl group of Thr67 (distance of 1,94 Å). Obviously, the two carboxyl groups contribute to pamoic acid affinity for the 8 kDa domain. Using one of them to tether a second fragment is likely to lower the affinity but this may be compensated by the properties of the second fragment. From our data, we have defined two other potential sites close to the pamoic acid binding site (see above). In the proposed model, each of the carboxyl groups is oriented towards one of the other sites. Therefore the possibility of increasing the pamoic acid affinity by using the fragment-based approach could be considered.
Pol beta gets involved in DNA repair pathway and in translesion synthesis, particularly when it is overexpressed in cancer cell lines treated by cisplatin agent. This process leads to a chemotherapeutic drug resistance, which could be prevented by an adjuvant treatment, that is to say a pol beta inhibitor. One of the key benchmarks for a small molecule to become a drug is the affinity for its target. No currently known pol beta inhibitors rise above micromolar affinity, which is insufficient for any pharmacological development. The X family DNA polymerases is the only one to feature the 8 kDa domain . Hence, an inhibitor of this domain is less subject to bind to replicative DNA polymerases. Moreover, inhibition of the 8 kDa of pol lambda and pol mu, both involved in non-homologous end joining of DNA break [38, 39], could improve the radiosensitivity of tumors by preventing cells from repairing radiotherapy-induced DNA damage.
Even if pamoic acid is one of the most known pol beta specific inhibitors, its affinity (low micromolar) has to be improved. Structural insights in the interaction between 8 kDa domain of pol beta and pamoic acid are prerequisites to improve the ligand affinity by applying the fragment-based strategy.
A previous work has reported the binding of pamoic acid to pol beta using chemical shift mapping. Pamoic acid is one of the best known pol beta specific inhibitors. It inhibits the deoxyribose phosphate lyase activity and increases sensibility to MMS . As pol beta has been shown to be a pharmacological target, increasing the affinity of pamoic acid for pol beta could transform pamoic acid into a drug-candidate. In the present paper, we have combined NMR (chemical shift mapping, STD and NOESY data) and computational approaches to generate a detailed 3D model of the complex of the 8 kDa domain of the DNA polymerase with pamoic acid. Validation of the computational model by experimental NMR data gave a unique structure for the complex (Fig. 6). The site occupied by pamoic acid corresponds to the one where single-stranded DNA binds to.
Indeed, the model thus established is the starting point to search for a fragment that could bind in one of the other two sites found in the vicinity of the pamoic acid binding site. The orientation of its bound carboxyl groups towards two distinct potential second sites makes pamoic acid a very interesting candidate for further attempts to increase its affinity for pol beta, using fragment-based approach. Furthermore, as NMR techniques can screen small molecules, they can be used to find a second fragment which binds to the 8 kDa domain in a site, which is close to but disjointed from the pamoic acid binding site . The present structure of the complex opens an avenue for the development of new families of specific pol beta inhibitors by the well-established fragment based approach. Glu26 or Lys72. Another potential second site could be related to Lys60, Leu62 and Ala70, further in the direction of Gly64 and Gly66, which belong to the pamoic acid site. be reported elsewhere.
Protein expression and purification
For NMR experiments, the recombinant 8 kDa-domain of DNA polymerase beta (Met1-Lys87) was produced as His-tag fusion protein in a pET28 plasmid (Novagen) in E. coli BL21(DE3). Bacteria were grown in LB medium at 37°C to an optical density at 600 nm of 0.8 before induction with 1 mM IPTG during 4 hours, to obtain an unlabeled sample. Isotopically 15N-labeled protein was expressed in minimal (M9) medium containing 15NH4Cl. Proteins were purified using a Ni-NTA column (HiTrap, Amersham) eluted by an imidazole gradient from 100 mM to 500 mM in 5 column volumes, after 8 column volumes at imidazole 100 mM. The protein was eluted at imidazole 250 mM. The protein was then concentrated to 0.5–2 mM in 20 mM deuterated Tris-HCl, pH = 6.8, 200 mM NaCl. The His-tag was cleaved using one unit of thrombin per milligram of protein overnight at 4°C (Novagen).
Pamoic acid preparation
Pamoic acid was obtained from Sigma-Aldrich. It was dissolved in high-pH Tris-HCl 20 mM, 200 mM NaCl at a concentration of 10 mM. Then, the pH was brought down to 6.9 by addition of the appropriate volume of HCl. The final solution was stored at 4°C.
NMR spectra were recorded at 293 K on a Bruker DRX600 spectrometer equipped with a cryoprobe. NMR samples were prepared in 20 mM deuterated Tris-HCl, pH = 6.8, 200 mM NaCl with 10% D2O. NMR data were processed using TOPSPIN software (Bruker) and NMRPipe  and analyzed using NMRView . Resonance assignment was performed using previously reported BMRB data [43, 44] and 3D TOCSY-HSQC and 3D NOESY-HSQC data. Spectra were recorded on the free 15N-labeled 8 kDa domain at 500 μ M and on the equimolar complex with 1024, 80 and 128 complex points for 1H, 15N and 1H respectively, and 8 scans. The mixing time for TOCSY and NOESY were 60 ms and 200 ms, respectively.
NMR chemical shift mapping
For NMR titration, 2D 15N-HSQC spectra of the 15N-labeled 8 kDa domain of pol beta at the concentration of 150 μ M were collected at 293 K after incremental addition of pamoic acid (typically 0.25, 0.5, 1, 1.5, 2, 3, 4 molar equivalents of ligand) with 2048 and 256 complex points for 1H and 15N, respectively, and 4 scans. The NMR sample was diluted by no more than 10% by increasing the pamoic acid concentration.
The factor of 5 appearing for Δδ N corresponds to the spectral widths ratio of 15N (25 ppm) and 1H (5 ppm) and is commonly used in the literature .
Saturation Transfer Difference experiments
For STD titration, a sample of 3 μ M of unlabeled protein in deuterated buffer was prepared. Increasing amounts of pamoic acid were added, in molar ligand/protein excess from 1 to 30.
Protein signals were saturated by forty 50 ms gaussian pulses, with interpulse spacing of 1 ms. The total saturation time was 2 s. Water signal elimination was performed using Watergate W5 sequence . Protein signals were eliminated by a 60 ms T1ρ filter with a 5000 Hz excitation. For each ligand concentration, two spectra were recorded with and without protein irradiation. For I0 spectrum, no irradiation of the protein occurred. I std spectrum is obtained by substraction of spectrum with protein irradiation from the I0 spectrum.
[L] represent pamoic acid concentration. K d and A max were searched parameters in fitting using the GOSA program .
The protein-ligand docking was performed using the AutoDock 3.05 software . The input files were prepared with the MGLTools 1.2. A pdbqs file was created, based on PDB: 1DK3 file information. Polar hydrogens were added and charges were calculated. From these data, a grid map of 96 × 126 × 96 points with a distance spacing of 0.442 Angströms representing the three-dimensional protein structure was calculated and used for the docking. The appropriate atom-specific affinity maps were created. The grid map covered the entire protein structure so that the ligand could move freely within the grid around the protein. For the ligand, pamoic acid, flexible docking has been enabled by defining 8 rotatable bonds within the molecule. The search of the parameter space followed the Lamarckian genetic algorithm, with most of the user-defined parameters set to their default values. We modified the population size by setting it to 1000 individuals, as well as the number of surviving individuals, the maximum number of energy evaluations and the maximum number of generations, which were multiplied by 10 with respect to their default values. With the number of generated structures set to 256, a typical run averaged 6 hours on a single Xeon 3 GHz CPU. Run in parallel on a 40-nodes Linux cluster, with different seeds for random number generator, the above setup allowed obtaining 10240 ligand conformations each time the program was launched. The results have been analyzed with an in-house software, allowing the analysis of the intermolecular energy and atomic contacts.
We are grateful to F. Amalric at the IPBS Institute, Toulouse for various support. We wish to acknowledge P. Ramos at the IPBS, Toulouse for NMR technical assistance. The 600 MHz cryoprobe and 700 MHz IPBS spectrometers were acquired within the CPER 2000–2006 involving CNRS, University of Toulouse, Region Midi-Pyrenees and European Structural Funds (FEDER). This work was supported by the French Research Ministry (to CH), Canceropole Grand Sud Ouest (to CC) and by Ligue Nationale Contre le Cancer (Equipe labellisee, to CC and to J.S. Hoffmann). "Genetic Instability and Cancer" and "NMR and membrane-protein interactions" teams are acknowledged for various support and discussions.
- Sawaya MR, Pelletier H, Kumar A, Wilson SH, Kraut J: Crystal structure of rat DNA polymerase beta: evidence for a common polymerase mechanism. Science 1994, 264(5167):1930–1935. 10.1126/science.7516581View ArticleGoogle Scholar
- Prasad R, Beard WA, Chyan JY, Maciejewski MW, Mullen GP, Wilson SH: Functional analysis of the amino-terminal 8-kDa domain of DNA polymerase beta as revealed by site-directed mutagenesis. DNA binding and 5'-deoxyribose phosphate lyase activities. J Biol Chem 1998, 273(18):11121–11126. 10.1074/jbc.273.18.11121View ArticleGoogle Scholar
- Beard WA, Prasad R, Wilson SH: Activities and mechanism of DNA polymerase beta. Methods Enzymol 2006, 408: 91–107. 10.1016/S0076-6879(06)08007-4View ArticleGoogle Scholar
- Wilson SH, Sobol RW, Beard WA, Horton JK, Prasad R, Berg BJV: DNA polymerase beta and mammalian base excision repair. Cold Spring Harb Symp Quant Biol 2000, 65: 143–155. 10.1101/sqb.2000.65.143View ArticleGoogle Scholar
- Sobol RW, Horton JK, Kühn R, Gu H, Singhal RK, Prasad R, Rajewsky K, Wilson SH: Requirement of mammalian DNA polymerase-beta in base-excision repair. Nature 1996, 379(6561):183–186. 10.1038/379183a0View ArticleGoogle Scholar
- Whitehouse CJ, Taylor RM, Thistlethwaite A, Zhang H, Karimi-Busheri F, Lasko DD, Weinfeld M, Caldecott KW: XRCC1 stimulates human polynucleotide kinase activity at damaged DNA termini and accelerates DNA single-strand break repair. Cell 2001, 104: 107–117. 10.1016/S0092-8674(01)00195-7View ArticleGoogle Scholar
- Sobol RW, Wilson SH: Mammalian DNA beta-polymerase in base excision repair of alkylation damage. Prog Nucleic Acid Res Mol Biol 2001, 68: 57–74.View ArticleGoogle Scholar
- Ochs K, Sobol RW, Wilson SH, Kaina B: Cells deficient in DNA polymerase beta are hypersensitive to alkylating agent-induced apoptosis and chromosomal breakage. Cancer Res 1999, 59(7):1544–1551.Google Scholar
- Fortini P, Pascucci B, Belisario F, Dogliotti E: DNA polymerase beta is required for efficient DNA strand break repair induced by methyl methanesulfonate but not by hydrogen peroxide. Nucleic Acids Res 2000, 28(16):3040–3046. 10.1093/nar/28.16.3040View ArticleGoogle Scholar
- Bergoglio V, Canitrot Y, Hogarth L, Minto L, Howell SB, Cazaux C, Hoffmann JS: Enhanced expression and activity of DNA polymerase beta in human ovarian tumor cells: impact on sensitivity towards antitumor agents. Oncogene 2001, 20(43):6181–6187. 10.1038/sj.onc.1204743View ArticleGoogle Scholar
- Vaisman A, Chaney SG: The efficiency and fidelity of translesion synthesis past cisplatin and oxaliplatin GpG adducts by human DNA polymerase beta. J Biol Chem 2000, 275(17):13017–13025. 10.1074/jbc.275.17.13017View ArticleGoogle Scholar
- Hoffmann JS, Pillaire MJ, Garcia-Estefania D, Lapalu S, Villani G: In vitro bypass replication of the cisplatin-d(GpG) lesion by calf thymus DNA polymerase beta and human immunodeficiency virus type I reverse transcriptase is highly mutagenic. J Biol Chem 1996, 271(26):15386–15392. 10.1074/jbc.271.26.15831View ArticleGoogle Scholar
- Hoffmann JS, Pillaire MJ, Maga G, Podust V, Hübscher U, Villani G: DNA polymerase beta bypasses in vitro a single d(GpG)-cisplatin adduct placed on codon 13 of the HRAS gene. Proc Natl Acad Sci USA 1995, 92(12):5356–5360. 10.1073/pnas.92.12.5356View ArticleGoogle Scholar
- Ali-Osman F, Berger MS, Rairkar A, Stein DE: Enhanced repair of a cisplatin-damaged reporter chloramphenicol-O-acetyltransferase gene and altered activities of DNA polymerases alpha and beta, and DNA ligase in cells of a human malignant glioma following in vivo cisplatin therapy. J Cell Biochem 1994, 54: 11–19. 10.1002/jcb.240540103View ArticleGoogle Scholar
- Albertella MR, Lau A, O'Connor MJ: The overexpression of specialized DNA polymerases in cancer. DNA Repair (Amst) 2005, 4(5):583–593. 10.1016/j.dnarep.2005.01.005View ArticleGoogle Scholar
- Mizushina Y, Yoshida S, Matsukage A, Sakaguchi K: The inhibitory action of fatty acids on DNA polymerase beta. Biochim Biophys Acta 1997, 1336(3):509–521.View ArticleGoogle Scholar
- Mizushina Y, Tanaka N, Kitamura A, Tamai K, Ikeda M, Takemura M, Sugawara F, Arai T, Matsukage A, Yoshida S, Sakaguchi K: The inhibitory effect of novel triterpenoid compounds, fomitellic acids, on DNA polymerase beta. Biochem J 1998, 330(Pt 3):1325–1332.View ArticleGoogle Scholar
- Boudsocq F, Benaim P, Canitrot Y, Knibiehler M, Ausseil F, Capp JP, Bieth A, Long C, David B, Shevelev I, Frierich-Heinecken E, Hübscher U, Amalric F, Massiot G, Hoffmann JS, Cazaux C: Modulation of cellular response to cisplatin by a novel inhibitor of DNA polymerase beta. Mol Pharmacol 2005, 67(5):1485–1492. 10.1124/mol.104.001776View ArticleGoogle Scholar
- Hajduk PJ, Greer J: A decade of fragment-based drug design: strategic advances and lessons learned. Nat Rev Drug Discov 2007, 6(3):211–219. 10.1038/nrd2220View ArticleGoogle Scholar
- Shuker SB, Hajduk PJ, Meadows RP, Fesik SW: Discovering high-affinity ligands for proteins: SAR by NMR. Science 1996, 274(5292):1531–1534. 10.1126/science.274.5292.1531View ArticleGoogle Scholar
- Liu D, Prasad R, Wilson SH, DeRose EF, Mullen GP: Three-dimensional solution structure of the N-terminal domain of DNA polymerase beta and mapping of the ssDNA interaction interface. Biochemistry 1996, 35(20):6188–6200. 10.1021/bi952656oView ArticleGoogle Scholar
- Maciejewski MW, Liu D, Prasad R, Wilson SH, Mullen GP: Backbone dynamics and refined solution structure of the N-terminal domain of DNA polymerase beta. Correlation with DNA binding and dRP lyase activity. J Mol Biol 2000, 296: 229–253. 10.1006/jmbi.1999.3455View ArticleGoogle Scholar
- Doherty AJ, Serpell LC, Ponting CP: The helix-hairpin-helix DNA-binding motif: a structural basis for non-sequence-specific recognition of DNA. Nucleic Acids Res 1996, 24(13):2488–2497. 10.1093/nar/24.13.2488View ArticleGoogle Scholar
- Pelletier H, Sawaya MR: Characterization of the metal ion binding helix-hairpin-helix motifs in human DNA polymerase beta by X-ray structural analysis. Biochemistry 1996, 35(39):12778–12787. 10.1021/bi960790iView ArticleGoogle Scholar
- Sobol RW, Prasad R, Evenski A, Baker A, Yang XP, Horton JK, Wilson SH: The lyase activity of the DNA repair protein beta-polymerase protects from DNA-damage-induced cytotoxicity. Nature 2000, 405(6788):807–810. 10.1038/35015598View ArticleGoogle Scholar
- Hu HY, Horton JK, Gryk MR, Prasad R, Naron JM, Sun DA, Hecht SM, Wilson SH, Mullen GP: Identification of small molecule synthetic inhibitors of DNA polymerase beta by NMR chemical shift mapping. J Biol Chem 2004, 279(38):39736–39744. 10.1074/jbc.M402842200View ArticleGoogle Scholar
- Goodsell DS, Morris GM, Olson AJ: Automated docking of flexible ligands: applications of AutoDock. J Mol Recognit 1996, 9: 1–5. Publisher Full Text 10.1002/(SICI)1099-1352(199601)9:1<1::AID-JMR241>3.0.CO;2-6View ArticleGoogle Scholar
- Murakami S, Kamisuki S, ichi Takata K, Kasai N, Kimura S, Mizushina Y, Ohta K, Sugawara F, Sakaguchi K: Site-directed mutational analysis of structural interactions of low molecule compounds binding to the N-terminal 8 kDa domain of DNA polymerase beta. Biochem Biophys Res Commun 2006, 350: 7–16. 10.1016/j.bbrc.2006.08.195View ArticleGoogle Scholar
- Clarkson J, Campbell ID: Studies of protein-ligand interactions by NMR. Biochem Soc Trans 2003, 31(Pt 5):1006–1009.View ArticleGoogle Scholar
- Liu D, DeRose EF, Prasad R, Wilson SH, Mullen GP: Assignments of 1H, 15N, and 13C resonances for the backbone and side chains of the N-terminal domain of DNA polymerase beta. Determination of the secondary structure and tertiary contacts. Biochemistry 1994, 33(32):9537–9545. 10.1021/bi00198a020View ArticleGoogle Scholar
- Oschkinat H, Griesinger C, Kraulis PJ, Sorensen OW, Ernst RR, Gronenborn AM, Clore GM: Three-dimensional NMR spectroscopy of a protein in solution. Nature 1988, 332(6162):374–376. 10.1038/332374a0View ArticleGoogle Scholar
- Clore GM, Gronenborn AM: Structures of larger proteins in solution: three- and four-dimensional heteronuclear NMR spectroscopy. Science 1991, 252(5011):1390–1399. 10.1126/science.2047852View ArticleGoogle Scholar
- Meyer B, Klein J, Mayer M, Meinecke R, Möller H, Neffe A, Schuster O, Wülfken J, Ding Y, Knaie O, Labbe J, Palcic MM, Hindsgaul O, Wagner B, Ernst B: Saturation transfer difference NMR spectroscopy for identifying ligand epitopes and binding specificities. Ernst Schering Res Found Workshop 2004, (44):149–167.Google Scholar
- Mayer M, Meyer B: Group epitope mapping by saturation transfer difference NMR to identify segments of a ligand in direct contact with a protein receptor. J Am Chem Soc 2001, 123(25):6108–6117. 10.1021/ja0100120View ArticleGoogle Scholar
- Fielding L, Rutherford S, Fletcher D: Determination of protein-ligand binding affinity by NMR: observations from serum albumin model systems. Magn Reson Chem 2005, 43(6):463–470. 10.1002/mrc.1574View ArticleGoogle Scholar
- Wannere CS, Schleyer PVR: How do ring currents affect (1)h NMR chemical shifts? Org Lett 2003, 5(5):605–608. 10.1021/ol027327kView ArticleGoogle Scholar
- Ramadan K, Shevelev I, Hübscher U: The DNA-polymerase-X family: controllers of DNA quality? Nat Rev Mol Cell Biol 2004, 5(12):1038–1043. 10.1038/nrm1530View ArticleGoogle Scholar
- Capp JP, Boudsocq F, Bertrand P, Laroche-Clary A, Pourquier P, Lopez BS, Cazaux C, Hoffmann JS, Canitrot Y: The DNA polymerase lambda is required for the repair of non-compatible DNA double strand breaks by NHEJ in mammalian cells. Nucleic Acids Res 2006, 34(10):2998–3007. 10.1093/nar/gkl380View ArticleGoogle Scholar
- Capp JP, Boudsocq F, Besnard AG, Lopez BS, Cazaux C, Hoffmann JS, Canitrot Y: Involvement of DNA polymerase mu in the repair of a specific subset of DNA double-strand breaks in mammalian cells. Nucleic Acids Res 2007, 35(11):3551–3560. 10.1093/nar/gkm243View ArticleGoogle Scholar
- Hajduk PJ, Gerfin T, Boehlen JM, Häberli M, Marek D, Fesik SW: High-throughput nuclear magnetic resonance-based screening. J Med Chem 1999, 42(13):2315–2317. 10.1021/jm9901475View ArticleGoogle Scholar
- Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A: NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR 1995, 6(3):277–293. 10.1007/BF00197809View ArticleGoogle Scholar
- Johnson BA: Using NMRView to visualize and analyze the NMR spectra of macromolecules. Methods Mol Biol 2004, 278: 313–352.Google Scholar
- Ulrich EL, Akutsu H, Doreleijers JF, Harano Y, Ioannidis YE, Lin J, Livny M, Mading S, Maziuk D, Miller Z, Nakatani E, Schulte CF, Tolmie DE, Wenger RK, Yao H, Markley JL: BioMagResBank. Nucleic Acids Res 2008, 36: D402-D408. 10.1093/nar/gkm957View ArticleGoogle Scholar
- BMRB : Biological Magnetic Resonance Data Bank[http://www.bmrb.wisc.edu/]
- Mollica L, Marchis FD, Spitaleri A, Dallacosta C, Pennacchini D, Zamai M, Agresti A, Trisciuoglio L, Musco G, Bianchi ME: Glycyrrhizin binds to high-mobility group box 1 protein and inhibits its cytokine activities. Chem Biol 2007, 14(4):431–441. 10.1016/j.chembiol.2007.03.007View ArticleGoogle Scholar
- Piotto M, Saudek V, Sklenár V: Gradient-tailored excitation for single-quantum NMR spectroscopy of aqueous solutions. J Biomol NMR 1992, 2(6):661–665. 10.1007/BF02192855View ArticleGoogle Scholar
- BioLog: GOSA : Global Optimization by Simulated Annealing.[http://www.bio-log.biz/index.php?page=gosa]
- DeLano W: The PyMOL Molecular Graphics System.2002. [http://www.pymol.org]Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.