A Plant Phytotoxin, Solanapyrone A, Is an Inhibitor of DNA Polymerase β and λ

Solanapyrone A, a phytotoxin and enzyme inhibitor isolated from a fungus (SUT 01B1-2) selectively inhibits the activities of mammalian DNA polymerase β and λ (pol β and λ) in vitro. The IC50 values of the compound were 30 μm for pol β and 37 μm for pol λ. Because pol β and λ are in a family and their three-dimensional structures are thought to be highly similar to each other, we used pol β to analyze the biochemical relationship with solanapyrone A. On pol β, solanapyrone A antagonistically competed with both the DNA template and the nucleotide substrate. BIAcore analysis demonstrated that solanapyrone A bound selectively to the N-terminal 8-kDa domain of pol β. This domain is known to bind single-stranded DNA, provide 5′-phosphate recognition of gapped DNA, and cleave the sugar-phosphate bond 3′ to an intact apurinic/apyrimidinic (AP) site (i.e. AP lyase activity) including 5′-deoxyribose phosphate lyase activity. Solanapyrone A inhibited the single-stranded DNA-binding activity but did not influence the activities of the 5′-phosphate recognition in gapped DNA structures and the AP lyase. Based on these results, the inhibitory mechanism of solanapyrone A is discussed.

lymerases and to clarify their biological and in vivo functions. In this study, we found an interesting fungus-produced compound that selectively inhibits the activities of pol ␤ and pol . Interestingly, the compound was found to be a plant phytotoxin, solanapyrone A, isolated from the causal fungus of early blight disease in potato. To our knowledge, such specific natural compounds have not been reported with the exception of the pol ␤-inhibitor, prunasin, which we reported previously (13). The compound was a stronger pol ␤-inhibitor than prunasin, and no pol -inhibitors have been reported.
Pol ␤ is the lowest molecular mass (39 kDa) DNA polymerase lacking such intrinsic accessory activities as 3Ј-or 5Ј-exonuclease, endonuclease, dNMP turnover, or pyrophosphorolysis (16). Pol ␤ consists of an independently folded N-terminal 8-kDa domain and C-terminal 31-kDa domain (17,18). The N-terminal domain was originally characterized as a single-stranded DNA (ssDNA)-binding domain. It was then found to possess binding specificity for the 5Ј-phosphate in gapped DNA (19 -21) and a helix-hairpin-helix motif found in several other DNA repair enzymes (22,23). Recently, Matsumoto and Kim (24) demonstrated that pol ␤ catalyzes the removal of dRP from the AP endonuclease-incised AP site via ␤-elimination as opposed to hydrolysis, and that this dRP lyase activity resides in the 8-kDa domain of pol ␤. According to recent studies (7,25), moreover, pol ␤ is known to have another family polymerase, pol . Pol ␤ has been present in all tissues examined and is generally expressed at a low level as are a number of other so-called constitutive "house-keeping" enzymes. Although the in vivo function of the family enzyme, pol , is unclear as yet, pol appears to work in a similar manner to pol ␤ (7). As to why a plant phytotoxin is a pol ␤-family-specific inhibitor, we presently are analyzing the structure and function of pol ␤ and pol using the inhibitor from two different view-points to understand the precise role of each of the polymerases in vivo and to develop a drug design strategy for cancer chemotherapy agents. We previously reported the three-dimensional structure of pol ␤ and the N-terminal 8-kDa domain (the DNA template-binding domain) with or without long chain fatty acids to further clarify the structure and function of pol ␤ (26). Because at least pol ␤ is an essential enzyme for nucleotide excision repair (1,2), the plant phytotoxin may lead to blockage of the DNA repair systems of rescue cancer cells under clinical radiation-therapy or chemotherapy. The plant phytotoxin is not only a useful molecular tool for analyzing the polymerases, it might also be considered as a potentially useful cancer chemotherapy agent.
Solanapyrone A, found as a novel inhibitor of pol ␤ and pol in this report, is a plant phytotoxin isolated together with other toxic metabolites from Alternaria solani, the causal fungus of early potato blight (27,28). The biochemical mode of action is still unknown, although there is a report on the detection of the enzymatic activity, isolation, and characterization of an enzyme catalyzing a Diels-Alder reaction (29). This compound was found to bind directly to the 8-kDa domain of pol ␤, but not to the 31-kDa domain, and in the three functional activities of the 8-kDa domain (i.e. ssDNA binding, 5Ј-phosphate recognition in gapped DNA, and dRP lyase activity using intact AP site containing DNA substrate) to inhibit only the ssDNA-binding activity. Based on these results, the inhibitory action of solanapyrone A and its relation to the enzyme structure of the 8-kDa domain of pol ␤ is discussed under "Discussion."
L-F1 and L-R1 primers contain EcoRI and XhoI sites (underlined), respectively. PCR was performed by the standard procedure of expand high fidelity (PharMingen). The template used was human placenta cDNA library (CLONTECH). The DNA sequences of the PCR products were determined by the dideoxy termination method. The PCR products were digested with EcoRI and XhoI restriction enzymes, and the DNA fragments obtained were subcloned into the bacteriophage T7 promoter-based expression vector pET28a(ϩ) (Novargen) between the EcoRI and XhoI sites. Recombinant His-pol was overexpressed and purified according to the recommended procedures by Novargen. Pol I (␣-like) and II (␤-like) from a higher plant, cauliflower influorescence, were purified according to the methods outlined by Sakaguchi et al. (34). The Klenow fragment of pol I and human immunodeficiency virus type-1 (HIV-1) reverse transcriptase were purchased from Worthington Biochemical Corp. (Freehold, NJ). Calf thymus terminal deoxynucleotidyltransferase, T7 RNA polymerase, and bovine pancreas deoxyribonuclease I were purchased from Stratagene Cloning Systems (La Jolla, CA). Taq DNA polymerase, T4 DNA polymerase, and T4 polynucleotide kinase were purchased from Takara (Tokyo, Japan).
DNA Polymerase Assays-The reaction mixtures for pol ␣, pol ␤, plant, and prokaryotic DNA polymerases were described previously (9,10), and the reaction mixtures for pol ␦ and ⑀ were described by Ogawa et al. (35). The reaction mixture for pol was the same as that for pol ␤. The substrates of the DNA polymerases used were poly(dA)/ oligo(dT) [12][13][14][15][16][17][18] and dTTP as template-primer DNA and nucleotide substrate, respectively. The substrates of HIV-1 reverse transcriptase used were poly(rA)/oligo(dT) [12][13][14][15][16][17][18] and dTTP as template-primer and nucleotide substrate, respectively. The substrates of terminal deoxynucleotidyltranscriptase used were oligo(dT) [12][13][14][15][16][17][18] (3Ј-OH) and dTTP as template-primer and nucleotide substrate, respectively. Solanapyrone A was dissolved in dimethyl sulfoxide (Me 2 SO) at various concentrations and sonicated for 30 s. 4 l of the sonicated samples was mixed with 16 l of each enzyme (final 0.05 units) in 50 mM Tris-HCl, pH7.5, containing 1 mM dithiothreitol, 50% glycerol, and 0.1 mM EDTA, and kept at 0°C for 10 min. These inhibitor-enzyme mixtures (8 l) were added to 16 l of each of the standard enzyme reaction mixtures, and incubation was carried out at 37°C for 60 min with the exception of Taq DNA polymerase, which was incubated at 74°C for 60 min. The activity without the inhibitor was considered to be 100%, and the remaining activities at each concentration of inhibitor were determined as percentages of this value. One unit of each DNA polymerase activity was defined as the amount of enzyme that catalyzed the incorporation of 1 nmol dTTP into the synthetic template-primers (i.e. poly(dA)/ oligo(dT) 12-18 ϭ 2/1) in 60 min at 37°C under the normal reaction conditions for each enzyme (9,10).
Other Enzyme Assays-Activities of calf DNA primase of pol ␣, T7 RNA polymerase, T4 polynucleotide kinase, and bovine deoxyribonuclease I were measured in each of the standard assays according to the specifications by the manufacturer as described by Koizumi et al. (36), Nakayama and Saneyoshi (37), Soltis and Uhlenbeck (38), and Lu and Sakaguchi (39), respectively. Telomerase activity was determined using the PCR-based telomeric repeat amplification protocol as described previously (40) with some modifications (41).
Surface Plasmon Resonance-Binding analyses of the 39-kDa intact and 8-kDa domain fragment of pol ␤ and solanapyrone A were performed using a Biosensor BIAcore instrument (BIACORE ® X, BIAcore, Sweden). CM5 research grade sensor chips (BIAcore, Sweden) were used. All buffers were filtered before use. Pol ␤ and 8-kDa domain (243 and 50 g/ml, respectively, 30 l each, i.e. 1.87 nmol each) in coupling buffer (10 M sodium acetate, pH 5.0) were injected over a CM5 sensor chip at 20 l/min to capture the protein to the carboxymethyl dextran matrix of the chip by NHS/EDC coupling reaction (60 l of mix) as described previously (42). Unreacted N-hydroxysuccinimide ester groups were inactivated using 1 M ethanolamine-HCl, pH 8.0. This reaction immobilized approximately 5000 response units of the protein.
The binding analysis of solanapyrone A was performed in running buffer including solanapyrone A (5 mM potassium phosphate buffer, pH 7.0, and 10% Me 2 SO) at a flow rate of 20 l/min at 25°C. Kinetic parameters were determined using the software BIA evaluation 3.1.
Gel Mobility Shift Assay-The gel mobility shift assay was carried out as described by Casas-Finet et al. (19). The binding mixture (a final volume of 20 l) contained 20 mM Tris-HCl, pH 7.5, 40 mM KCl, 50 g/ml bovine serum albumin, 10% Me 2 SO, 2 mM EDTA, M13 plasmid DNA (2.2 nmol, nucleotide, single-stranded and singly primed), and 0.15 nmol 80-kDa domain of pol ␤. Various concentrations of solanapyrone A were added to the binding mixture followed by incubation at 25°C for 10 min. Samples were run on a 1.0% agarose gel in 0.1 M Tris acetate buffer, pH 8.3, containing 5 mM EDTA at 50 V for 2 h.
5Ј and 3Ј-End Labeling-The 5Ј-end of the dephosphorylated primer ( Fig. 6A, primer 1, 17-mer) was labeled with T4 polynucleotide kinase using [␥-32 P]ATP as described previously (43). The 3Ј-end of the 29 base oligodeoxyribonucleotide containing an AP site at position 10 ( Fig. 7, A and B) was labeled on its 3Ј-end with terminal deoxynucleotidyltransferase using [␣-32 P]ddATP. These oligonucleotides were annealed to their complementary strands by heating the solution at 90°C for 3 min followed by slow cooling to 25°C. 32 P-Labeled duplex oligodeoxynucleotide was separated from unincorporated [␥-32 P]ATP or [␣-32 P]ddATP using a MicroSpin G-25 column (Amersham Biosciences, Inc.) according to the protocol suggested by the manufacturer.
UV Cross-linking to Gapped DNA-The 8-kDa domain of pol ␤ (1 M) was mixed with the 5-nucleotide-gapped DNA template-primer (0.5 M) (21) and various concentrations of solanapyrone A in a reaction mixture containing 20 mM Tris-HCl, pH 8.0, 20 mM NaCl, 1 mM EDTA, 5 mM MgCl 2 , and 10% Me 2 SO and incubated at room temperature for 15 min. The samples were spotted onto Parafilm and irradiated at 254 nm for 4 min using a UV-stratalinker (Stratagene Cloning Systems). The photochemical cross-linked 8-kDa protein-DNA complexes were separated by 15% SDS-PAGE and visualized by autoradiography.
AP Lyase Assay-AP lyase activity was determined in a reaction mixture (20 l) that contained 50 mM Hepes, pH 7.4, 2 mM dithiothreitol, 10% Me 2 SO, 200 nM 8-kDa domain of pol ␤, and various concentrations of solanapyrone A. The mixture was preincubated, and then 32 P-labeled double-stranded oligodeoxynucleotide containing an AP site at position 10 was added. After incubation for 15 min at 37°C, the reaction was terminated, and the product was stabilized by the addition of 2 M NaBH 4 to a final concentration of 340 mM and then incubated for 30 min at 0°C. The stabilized DNA product was recovered by ethanol precipitation in the presence of 0.1 g/ml tRNA and resuspended in 10 l of gel loading buffer (95% formamide, 20 mM EDTA, 0.02% bromphenol blue, and 0.02% xylene cyanol). After incubation at 75°C for 2 min, the reaction products were separated by electrophoresis in a 20% polyacrylamide gel containing 8 M urea in 89 mM Tris-HCl, pH 8.8, 89 mM boric acid, and 2 mM EDTA and visualized by autoradiography.
Solanapyrone A Docking Modeling-The molecular docking of solanapyrone A and the 8-kDa domain of pol ␤ was done using a fixed docking procedure in the Affinity program within the Insight II modeling software (Molecular Simulations Inc., San Diego, CA, USA, 1999). Solanapyrone A was placed to orient the aldehyde moiety to Lys65 of pol ␤. The calculations used a CVFF force-field in the Discovery program and a Monte Carlo strategy in the Affinity programs (44). Each energyminimized final docking position of solanapyrone A was evaluated using the interactive score function in the Ludi module. Ludi score includes contribution of the loss of translational and rotational entropy of the fragment, number and quality of hydrogen bonds, and contributions from ionic and lipophilic interactions to the binding energy.

Production and Isolation of Solanapyrone A-We screened
for DNA polymerase inhibitors and found a natural compound that selectively inhibited mammalian DNA polymerase ␤ activity from a fungus (SUT 01B1-2) collected from fields in the vicinity of Noda city in Chiba prefecture, Japan. The compound was extracted with CH 2 Cl 2 from the broth of the fungus, and then purified by silica gel column and Sephadex LH-20 column chromatography. Negative FABHR (Fast Atom Bombardment High Resolution) mass and [ 1 H]NMR, [ 13 C]NMR, and DEPT (Distortionless Enhancement by Polarization Transfer) NMR spectroscopic analyses indicated that the inhibitor was found to be an agent known as solanapyrone A previously reported as a phytotoxin isolated together with other toxic metabolites from A. solani, the causal fungus of early potato blight (28). The chemical structure of solanapyrone A is shown in Fig. 1. The fact that a plant phytotoxin is an inhibitor of DNA polymerase is of interest. Fig. 2 shows the inhibition dose-response curves of solanapyrone A against mammalian DNA polymerases. Solanapyrone A was effective at inhibiting rat DNA polymerase ␤ and human DNA polymerase (pol or pol ␤2), but not replicative DNA polymerases such as calf DNA polymerase ␦ and human DNA polymerase ⑀, and only slightly in calf DNA polymerase ␣ (Fig. 2). The inhibition of pol ␤ by solanapyrone A was dose-dependent with a 50% inhibition observed at a dose of 30 M and almost a complete inhibition at 80 M (Fig. 2). Because dideoxy thymidine triphosphate (ddTTP), a potent inhibitor of mammalian pol ␤, shows complete inhibition at 40 M (45), the effect of solanapyrone A on this enzyme was almost as strong as that of ddTTP. The IC 50 values of solanapyrone A on calf pol ␣, calf pol ␦, and human pol ⑀ were over 100 M (Fig. 2). The inhibitory effect of solanapyrone A on higher plant cauliflower pol II (␤-like) was stronger than that of pol I (␣-like), and the IC 50 values were 41 and Ͼ100 M, respectively (Table I). Solanapyrone A had no inhibitory effect on prokaryotic DNA polymerases, such as the Klenow fragment of E. coli pol I, Taq DNA polymerase, T4 DNA polymerase, and other DNA-metabolic enzymes such as calf terminal deoxynucleotidyltransferase, human telomerase, HIV-1 reverse transcriptase, T7 RNA polymerase, T4 polynucleotide kinase, and bovine deoxyribonuclease I ( Table I). The IC 50 values in Table I did not change when the template-primer was activated DNA.

Effects of Solanapyrone A on the Activities of Mammalian DNA Polymerases and Other Enzymes-
Interestingly, in the inhibition spectrum, solanapyrone A inhibited the activity of pol , which has been recently identified as a new family member of pol ␤. The inhibitory effect of pol was the same as that of pol ␤, and the IC 50 value was 37 M. Pol contains a BRCT motif in its N-terminal region, its C-terminal region exhibits a 33% sequence identity to a corresponding region of human pol ␤. The truncated pol , which deleted the BRCT motif in its N-terminal region, has DNA polymerase activity and inhibited the activity of pol ␤ at the same concentration of solanapyrone A (data not shown). Solanapyrone A thus appeared to be a selective inhibitor of eukaryotic pol ␤ and . The remainder of this report is devoted to an analysis of the action of solanapyrone A using pol ␤, because  pol ␤ and are in the pol X family as described above, and their three-dimensional structures are thought to be highly similar to each other (7,25).
Because solanapyrone A was found as a phytotoxin as described above, the inhibition of pol ␤ and its family enzyme, pol , by the compound strongly suggests a relationship between the lack of DNA repair and the activity of the phytotoxin.
Mode of DNA Polymerase ␤ Inhibition by Solanapyrone A-Next, to elucidate the mechanism of inhibition of solanapyrone A on pol ␤, the extent of inhibition as a function of DNA template-primer or dNTP substrate concentrations was studied (Fig. 3). In kinetic analysis, poly(dA)/oligo(dT) [12][13][14][15][16][17][18] and dTTP were used as the template-primer DNA and dNTP substrate, respectively. Double reciprocal plots of the results showed that the solanapyrone A-induced inhibition of pol ␤ activity was competitive with both the DNA template and the dNTP substrate (Fig. 3, A and B). In the case of the DNA template, the apparent maximum velocity (V max ) was unchanged at 111 pmol/h, whereas 147, 294, and 588% increases in the Michaelis-Menten constant (K m ) were observed in the presence of 20, 40, and 60 M solanapyrone A, respectively (Fig.  3A). The V max for the dNTP substrate was 62.5 pmol/h, and the K m for the dNTP substrate decreased from 3.05 to 15 pmol/h in the presence of 60 M solanapyrone A (Fig. 3B). The inhibition constant (K i ) values obtained from Dixon plots were found to be 6.8 and 22 M for the DNA template and substrate dTTP, respectively (Fig. 3, C and D). The inhibition by solanapyrone A against the DNA template was 3-fold more effective than that against the dTTP substrate. When activated DNA and four deoxynucleoside triphosphates (i.e. dNTP substrate) were used as the template-primer DNA and dNTP substrate, respectively, the inhibition of pol ␤ by solanapyrone A was competitive with both the DNA template and the dNTP substrate.
The inhibition of pol ␤ by ddTTP was non-competitive with activated DNA as a DNA template-primer and competitive with respect to the dNTP substrate (45). ddTTP inhibited the pol ␤ activity by competing with TTP (45). In contrast, the inhibition of pol ␤ by solanapyrone A was competitive with the DNA template (Fig. 3A), suggesting that solanapyrone A directly binds to the DNA template-binding site of pol ␤. The ssDNA-binding site occurs in the N-terminal 8-kDa domain of pol ␤ (17, 18). We further studied the interaction between solanapyrone A and the 8-kDa domain of pol ␤.
Binding between Solanapyrone A and DNA Polymerase ␤ or Its 8-kDa Domain Fragment-The rat pol ␤ used in this study has been extensively studied, including its amino acid sequence and its secondary and tertiary structures (46 -51). The enzyme can be divided into two domain fragments using controlled proteolysis, an 8-kDa N-terminal domain fragment and a 31-kDa C-terminal domain fragment (17,18).
To confirm the kinetic parameters precisely, the parameters for the binding of solanapyrone A were determined using 39-kDa pol ␤ and the 8-kDa domain immobilized to the sensor chip in a BIAcore. Five different concentrations of solanapyrone A (20, 40, 60, 80, and 100 M) were used for the binding analysis. Both the enzyme and the 8-kDa domain (1.87 nmol each) were conjugated to the CM5 sensor chip, and then solanapyrone A was added to the conjugated proteins. Solanapyrone A bound to pol ␤ and dissociated from the protein (Fig. 4A). Solanapyrone A also bound to the 8-kDa domain at almost the same rate as pol ␤ (Fig. 4B). The dissociation constants (K d ) of binding of solanapyrone A to pol ␤ and the 8-kDa domain were determined to be 33.6 and 1.03 M, respectively, from the data in Fig. 4. This result suggested that the binding of solanapyrone A to the 8-kDa domain is 30-fold tighter than the binding of solanapyrone A to pol ␤. Therefore, solanapyrone A must interact with the 8-kDa domain directly.
Analysis of the Binding between Solanapyrone A and the N-terminal 8-kDa Domain of DNA Polymerase ␤-We investigated the interaction between the 8-kDa domain of pol ␤ and solanapyrone A more precisely. The DNA-binding activity of the 8-kDa domain was analyzed by a gel mobility shift assay. Fig. 5 shows the gel mobility shift assay of the M13 ssDNA to the 8-kDa domain-binding complex. In the binding assay, M13 ssDNA at 2.2 nmol (nucleotide) was added to 0.15 nmol of the 8-kDa domain (Fig. 5, lanes 2-6). The 8-kDa domain bound to M13 ssDNA was shifted in the gel (Fig. 5, lane 6). The molecular ratios of solanapyrone A and the 8-kDa domain are shown as the inhibitor to enzyme ratio (I/E) in Fig. 5. When the I/E ratio was 1 or more, solanapyrone A interfered with the complex formation between M13 ssDNA and the 8-kDa domain of pol ␤ (Fig. 5, lanes 2-4). At a ratio of 0.1 (Fig. 5, lane 5), the inhibition of the DNA-binding activity of the 8-kDa domain almost disappeared, suggesting that one molecule of solanapyrone A competes with one molecule of M13 DNA and subsequently interferes with the binding of ssDNA to the 8-kDa domain of pol ␤. Kinetic analysis indicated that solanapyrone A acted by competing with the DNA template on pol ␤, and thus solanapyrone A directly binds to the ssDNA-binding site of the 8-kDa domain of pol ␤ (Fig. 3A). However, the interference of the shift in gel mobility by ddTTP did not occur (data not shown), indicating that the action modes of ddTTP and solanapyrone A on pol ␤ differed from each other. The ddTTP data were in agreement with those of a previous study (45), indicating that the inhibition mode of pol ␤ by ddTTP was noncompetitive with the DNA template-primer.
Inhibition of Solanapyrone A on the 5Ј-Phosphate Recognition in Gapped DNA-Cross-linking of pol ␤ to gapped DNA is dependent on a 5Ј-phosphate moiety in the gap. This DNA gap binding of pol ␤ was directed by the 8-kDa domain (21). To investigate the inhibition of solanapyrone A on the 5Ј-phosphate recognition in gapped DNA, a synthetic gapped DNA substrate was formed by annealing two 17-residue oligonucleotides (designated as Primer 1 and Primer 2) to a 39-residue template 1 creating a 5-nucleotide gap between the 3Ј-hydroxyl of Primer 1 and the 5Ј-phosphate or 5Ј-hydroxyl of Primer 2 (see Fig. 6A and "Experimental Procedures"). This DNA substrate was incubated with the 8-kDa domain of pol ␤, and then the complex was photochemically cross-linked with UV light (21). To score the covalently cross-linked complexes, the 5Ј-end of the Primer 1 oligonucleotide was separated by SDS-PAGE, and the gel was analyzed by autoradiography (Fig. 6B). The results showed that the cross-linking among the 8-kDa domain, Template 1, and Primer 1 was strongly influenced by the phosphate group on the 5Ј-end of Primer 2 (Fig. 6B, lanes 7 and 8). The molecular ratios of solanapyrone A and the 8-kDa domain are shown as the inhibitor to enzyme ratio (I/E) in Fig. 6B. Even an excess amount of solanapyrone A (at an I/E ratio of more than 100) could not interfere with the 5Ј-phosphate recognition of the 8-kDa domain (Fig. 6B, lanes 3 and 4).
Effect of Solanapyrone A for Cleavage of AP Site Containing Double-stranded DNA by the N-Terminal 8-kDa Domain of DNA Polymerase ␤-Pol ␤ and its 8-kDa domain could cleave double-stranded DNA containing an AP site (52), whereas the pol ␤-catalyzed reaction is via ␤-elimination incising the AP site on the 3Ј-side of the sugar. Whether solanapyrone A bound on the 8-kDa domain could inhibit not only the DNA-binding activity but also the incise activity is of interest. To study the inhibitory activity of this reaction by solanapyrone A, we used a 20 base pair oligonucleotide duplex DNA that contained an AP site at position 10 (Fig. 7A). The AP site containing the DNA strand was 3Ј-end-labeled with [ 32 P]ddAMP and annealed to its complementary DNA strand (Fig. 7B). The AP site containing DNA was incubated with the 8-kDa domain of pol ␤. The 8-kDa domain incised DNA product bears a 3Ј-dRP moiety and a 5Ј-phosphate (Fig. 7B). To resolve the cleaved labeled DNA products bearing a 5Ј-phosphate, the products were stabilized by NaBH 4 reduction at the end of each reaction period, and the incised DNA products were separated by gel electrophoresis in a 20% denaturing polyacrylamide gel (Fig. 7C). In the AP lyase activity, solanapyrone A was added. The molecular ratio of solanapyrone A and the 8-kDa domain are shown as the inhibitor and enzyme ratio (I/E) in Fig. 7C. The substrate only as a negative control is shown in lane 1, and the control using alkali-hydrolyzed substrate is shown in lane 6. The I/E ratios for the 8-kDa domain shown in lanes 2, 3, 4, and 5 were 100, 10, 1, and 0, respectively. Even an excess amount of solanapyrone A did not interfere with the AP lyase activity. Solanapyrone A must inhibit only the DNA-binding activity of the 8-kDa domain by competing with the DNA template. DISCUSSION As described in the Introduction, we have been screening for new DNA polymerase inhibitors for use in analyzing the structure and function of mammalian DNA polymerases to understand their precise roles in vivo and to develop drug design strategies for the development of cancer chemotherapy agents. These inhibitors are not only molecular tools for analyzing the polymerases but should also be considered as potentially useful cancer chemotherapy agents. Subsequently, we found a potent inhibitor selective to pol ␤ and its family polymerase, pol , which was a potentially useful agent for these purposes, and the inhibitor was one of the phytotoxins known as solanapyrone A.
Also as described in the Introduction, solanapyrone A is a plant phytotoxin isolated together with other toxic metabolites from A. solani, the causal fungus of early potato blight (27,28). Solanapyrone A could inhibit the activities of mammalian pol ␤ and in the range of 50 -100 M. We should emphasize that solanapyrone A suppressed the activities of pol ␤ in vitro to the same extent as dideoxyTTP, a well known potent pol ␤ inhibitor (45). Because of the polymerase species-specificity effects, solanapyrone A could be useful as a pol ␤ family-specific inhibitor in studies to determine the precise roles of pol ␤ and . The IC 50 value on higher plant cauliflower pol II (␤-like) was 41 M, suggesting that solanapyrone A works as a toxic metabolite by inhibiting the activities of ␤-family DNA polymerases in the plant. In mammals, pol ␤, which is widely known to have roles in the short-patch base excision repair pathway (24, 53, 54), plays an essential role in neural development (55). Before this finding, Northern blot analysis indicated that the transcripts of pol ␤ were abundantly expressed in the testis, thymus, and brain in rats (56). The reason why the testis and the thymus require the pol ␤ activity has almost been determined. Both organs need DNA recombination processes specific to meiosis and the immunoglobulin production system (57,58). Although the functions of the plant alternative of pol ␤ remain unknown, the phytotoxin from the causal fungus of early potato blight appears to be an inhibitor of the plant alternative of pol ␤.
Pol ␤ is the smallest known DNA polymerase in animal cells with a molecular mass of 39 kDa, and its structure is highly conserved among mammals (1, 2). This protein has a modular two-domain structure with apparent flexibility within a protease-sensitive region that separates the two domains; 1) an N-terminal domain fragment (8 kDa), which retains the binding affinity for ssDNA, 5Ј-phosphate recognition in gapped DNA, and dRP lyase activity using an intact AP site containing DNA substrate, and 2) a C-terminal domain fragment (31 kDa) with DNA polymerase activity (17,18). Solanapyrone A could bind to the ssDNA-binding site of the 8-kDa domain, but it was not effective at 5Ј-phosphate recognition or AP lyase activity using the intact AP site-containing DNA substrate. These results strongly suggest that solanapyrone A by binding to the 8-kDa domain inhibited only the binding affinity for ssDNA on the domain by competing with the DNA template-primer. NMR, x-ray crystal structures, and site-directed mutagenesis of this domain have suggested several residues that may interact with ssDNA or play a role in the dRP lyase interaction (47, 59 -61). As a result of the site-directed mutagenesis of the 8-kDa domain reported by Prasad et al. (61), Lys-35, Lys-60, and Lys-68 were impaired in the ssDNA-binding activity. Lys-35 was involved in 5Ј-phosphate recognition and, including Lys-72, was significantly reduced in dRP lyase activity. The dRP lyase active site Lys-72 was observed to be part of a lysine-rich pocket consisting of Lys-35, 68, 72, 84, and 87 on the surface of the 8-kDa domain that may also bind to ssDNA (62). It was suggested that solanapyrone A bound to the residues of Lys-60 of the 8-kDa domain and inhibited the ssDNA-binding activity by competing with the DNA template.
The crystal and solution structures of the 8-kDa domain of pol ␤ have been determined (47,48,59,60). According to their results, the 8-kDa domain (residues 1-87) is formed from four ␣-helices packed as two antiparallel pairs (Fig. 8A). The pairs of ␣-helices cross one another at 50°giving them a V-like shape. The 8-kDa domain contains a motif termed the "Helix-hairpin-  Helix." We created a definitive binding set of Lys-60 residue in the ssDNA-binding pocket between helix-3 and helix-4 to move as a 2.5-Å shell around the manually docked ligand during the energy minimization. The number of final docking positions was set to 4, although only one promising position was identified finally (Fig. 8). As shown in Fig. 8A, solanapyrone A could be mapped to one face of the 8-kDa domain. In pol ␤, Lys-60 is a hydrophilic amino acid. The two ketone groups (-C ϭ O) of solanapyrone A may, therefore, show a preference for binding to the hydrophilic residue of Lys-60, and on the other side, the benzene groups may be absorbed to the hydrophobic amino acids in both helix-3 and helix-4 (Fig. 8A). In the docking simulation, the binding energy between NH 3 ϩ of Lys-60 and the ketone groups in solanapyrone A was Ϫ28.230 kcal/mol by hydrogen bond, and the binding force consisted of coulomb force (Ϫ27.212 kcal/mol) and van der Waals forces (Ϫ1.018 kcal/mol) ( Table II). The distances between the two ketone groups of solanapyrone A and the NH 3 ϩ residue of Lys-60 were 2.01 and 2.41 Å (Fig. 8B). The binding energy between the benzene backbone of solanapyrone A and the hydrophobic amino acids (i.e. Ile-53, Gly-56, Ala-59, Ala-70, and Ile-73) was Ϫ7.682 kcal/mol (Table II). The Connolly surface of helix-3 and the three-dimensional position of solanapyrone A is indicated in Fig. 8C. The lysine-rich pocket consisting of Lys-35, 68, 72, 84and 87 had no interaction with solanapyrone A (data not shown). On the 8-kDa domain, solanapyrone A was smoothly intercalated into the pocket between helix-3 and helix-4, and the residues around the Lys-60 site appear to be important for solanapyrone A binding.
As described above, the purpose of this study was to screen a useful agent for analyzing the in vivo functions of DNA repairrelated enzymes in pol ␤-rich tissues, because this agent was found to potently influence mammalian pol ␤ activity, and we could finally report the properties of solanapyrone A with regard to its effect on pol ␤ and its family enzyme pol . The amino acid sequence and the three-dimensional structure of pol ␤ have been studied extensively (46,51), but the in vivo functions are still mostly unknown. For example, pol ␤ is quite highly expressed in the meiotic tissues, the thymus, and the brain in mammals (56), but no reasonable explanation with direct evidence for this localization has yet been proposed. Moreover, the biochemical properties and in vivo functions of pol identified recently are not known. The Northern blot analysis indicated that the expression of pol was almost the same as that of pol ␤ (7). This compound will be useful for analyzing the function of pol ␤ and pol in tissues such as the brain in which replication polymerases do not function. ddTTP is well known as a pol ␤ inhibitor (45). However, ddTTP cannot be used for in vivo experiments, because it cannot penetrate the cells. Other compounds reported to inhibit the pol ␤ activity include long chain fatty acids (9, 10, 26), a bile acid such as lithocholic acid (35,63), terpenoids (12, 64 -67), nucleotide analogs (68,69), flavonoids (15,70,71), sulfate-or sialic acidcontaining glycolipids (11,72,73), and phospholipids (74,75), but these agents inhibited activities of both pol ␣ and pol ␤. As a result, we studied the antibiotic in more detail that was first reported 10 years ago and report here our findings on its unique properties (i.e. a selective inhibitor of pol ␤ and ). Based on the data provided here, solanapyrone A could be used as a molecular probe for the study on pol ␤ and pol . We are at present trying to synthesize solanapyrone A chemically.