Daurinol used in this study was chemically synthesized according to recently reported methods (9). Natural daurinol was isolated from Haplophyllum dauricum as previously described (6). Dimethyl sulfoxide (DMSO), etoposide, novobiocin, propidium iodide (PI) and ATP disodium salt were purchased from Sigma (St. Louis, MO, USA). Camptothecin was purchased from TopoGEN, Inc. (Port Orange, FL, USA). Daurinol, etoposide and camptothecin were dissolved in DMSO for cellular treatment. Antibodies against E2F-1 (3742), cyclin A (4656) and Cdk4 (2906) were purchased from Cell Signaling Technology (Danvers, MA, USA). Antibodies against Cdk2 (sc-163), cyclin E (sc-247), cyclin D1 (sc-753) and β-actin (sc-47778) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Secondary anti-rabbit and anti-mouse antibodies were purchased from Santa Cruz Biotechnology.

The crystal structure of human topoisomerase IIα (ATP2ATPase) complexed with AMPPNP (a non-hydrolyzable ATP analog) from the Protein Data Bank (PDB code 1ZXM) was used for the docking simulation. Daurinol was built using the Maestro build panel and minimized using the impact module of Maestro in the Schrödinger suite program. The starting coordinates of the ATP2ATPASE were further modified for the prediction of daurinol binding. The protein structure was minimized using the Protein Preparation Wizard by applying an OPLS force field. For grid generation, the binding site was defined as the centroid of the AMPPNP. Ligand docking into the active site of ATP2ATPASE was performed using the Schrödinger docking program Glide (Schrödinger, Inc., USA). Energy-minimized daurinol was docked into the prepared receptor grid. The best-docked poses were selected based on the lowest Glide scores. The molecular graphics of the inhibitor-binding pocket and refined docking model for daurinol were generated using the PyMol package (http://www.pymol.org).

The inhibitory activity of daurinol against human topoisomerase IIα was measured using the Topoisomerase II Drug Screening kit (TopoGEN, Inc.). The standard reaction mixture (20 μl) contained 250 ng of supercoiled DNA (pHOT1), 0.7 μl of topoisomerase IIα, and different concentrations of ATP (0.5, 1 or 2 mM) and the tested compounds dissolved in DMSO. The final concentration of DMSO was 1%, and the topoisomerase II poison etoposide was used as a positive control. Reactions were initiated by the addition of the supercoiled DNA substrate. The reaction mixture was incubated at 37°C for 30 min. Other procedures were performed as described previously (5).

The inhibitory activity of daurinol against human topoisomerase I was measured using the Topoisomerase I Assay kit (TopoGEN, Inc.). The standard reaction mixture (20 μl) contained 10 mM Tris-HCl (pH 7.9), 1 mM EDTA, 150 mM NaCl, 0.1% bovine serum albumin (BSA), 0.1 mM spermidine, 5% glycerol, 250 ng of supercoiled DNA (pHOT1), 0.7 μl of topoisomerase I, and the test compound dissolved in DMSO. The final concentration of DMSO was 1%, and the topoiso-merase I poison camptothecin was used as a positive control. Reactions were initiated by the addition of the supercoiled DNA substrate. The reaction mixture was incubated at 37°C for 30 min, and 2 μl of 10% sodium dodecyl sulfate was added to stop the reaction. Additional procedures were performed using procedures similar to the topoisomerase II assay, as described previously (5).

The DNA-dependent ATP hydrolysis activity of human topoisomerase IIα was determined by quantifying hydrolyzed inorganic phosphate using the malachite green assay with slight modifications (10–12). Briefly, the standard reaction mixture (20 μl) contained 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 10 mM MgCl₂, 0.5 mM dithiothreitol, 30 μg/ml BSA, 200 ng of supercoiled DNA (pHOT-1), 2.0 μl of human topoisomerase IIα (TopoGEN, Inc.); and 100 μM daurinol or 400 μM novobiocin dissolved in DMSO. The final concentration of DMSO was 1%, and novobiocin was used as a positive control (12,13). The reaction was initiated by adding ATP at a final concentration of 2 mM (Sigma) and was incubated at 37°C for 30 min. The reaction was stopped by the addition of the Lanzetta reagent (80 μl, freshly prepared mixture of 0.035% malachite green-HCl and 4.2% ammonium molybdate in 4 N HCl at a ratio of 3:1 and 0.2% CHAPS), and the color development was read immediately using a Synergy HT Multi-Mode Microplate Reader (BioTek Instruments, Winooski, VT, USA) at 620 nm. The amount of inorganic phosphate was calculated using a phosphate (potassium dihydrogen phosphate, KH₂PO₃) standard curve. Enzyme activity was expressed as μM phosphate produced per min reaction.

SK-OV-3 and NIH:OVCAR-3 human ovarian cancer; and NTERA-2 cl.D1 (NT2-D1), NCCIT human testicular cancer cell lines were obtained from American Type Culture Collection (ATCC; Rockville, MD, USA). The SNU-840 human ovarian cancer; and NCI-H69, NCI-H146, NCI-H187 and NCI-H417 human small-cell lung cancer cell lines were obtained from the Korean Cell Line Bank (Seoul, Korea). These cells were cultured in RPMI-1640 supplemented with 25 mM HEPES, 10% (v/v) heat-inactivated fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin. Cells were maintained at sub-confluence in a 95% air and 5% CO₂ humidified atmosphere at 37°C.

Cancer cells (5×10³ cells/well) were plated in 96-well plates, incubated at 37°C for 24 h, and treated with daurinol for 48 h. Cell viability was determined using the EZ-Cytox cell viability assay kit (Daeil Lab Service, Ltd., Seoul), as previously described (14).

NIH:OVCAR-3 (5×10⁵ cells/well), SNU-840 (3×10⁵ cells/well) and NCCIT (3×10⁵ cells/well) cells were plated in 6-well plates, incubated at 37°C for 24 h, and treated with daurinol or etoposide for 24 and 48 h. The cells were stained with PI, and their DNA contents were analyzed using a FACSCalibur flow cytometer (Becton-Dickinson, San Jose, CA, USA) and Modfit LT V3.0 software (Verity Software House, Topsham, ME, USA) as previously described (5).

SNU-840 cells (5×10⁵) were seeded on 60-mm dishes, incubated for 24 h, and then treated with daurinol for 24 and 48 h. Additional procedures were performed as previously described (15). The relative protein expression was measured by densitometry.

To evaluate the effect of daurinol and etoposide on cell and nuclear size, we performed fluorescence pulse signal analysis of PI-stained SNU-840 cells as previously described (15,16). The treatment and flow cytometric DNA content analysis were performed as described above in the section of cell cycle analysis. FSC-H and FL2-W values, which correlate with cell and nuclear size, respectively, were measured using flow cytometry. The mean values of FSC-H and FL2-W were calculated using the histogram statistics tool from CellQuest Pro software (Becton-Dickinson). To compare the FSC-H and FL2-W distributions of the control and chemical-treated cells, we used the Kolmogorov-Smirnov statistics tools from CellQuest Pro software. We also observed SNU-840 cell morphology after treatment with daurinol or etoposide using an Olympus CK40 phase contrast microscope (Tokyo, Japan).

The values represent the mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism 5 software (La Jolla, CA, USA). A paired two-tailed Student’s t-test was used to compare the IC₅₀ values of daurinol and etoposide in each cell line. One-way analysis of variance (ANOVA) followed by Dunnett’s or Tukey’s multiple comparison test was used to analyze all other data. P<0.05 was considered statistically significant.

Because only a minute amount of natural daurinol is present in the plant H. dauricum, large-scale production of daurinol is one of the key hurdles to its development as a clinical anticancer drug. Thus, we prepared synthetic daurinol using recently described regioselective chemical synthesis methods (9) and tested its anticancer activity in various cancer cells, including HCT116 cells. The anti-proliferative activity (IC₅₀ value) of synthetic daurinol was equivalent to that of natural daurinol in HCT116 cells. In addition, synthetic daurinol also clearly induced cell cycle arrest at S phase in HCT116 cells (data not shown). These results were in accordance with our previous report (5). Based on these data, we used synthetic daurinol as a substitute for natural daurinol in the present study.

We used molecular docking simulations to evaluate the biochemical mechanism underlying the inhibitory activity of daurinol against human topoisomerase IIα. The results indicated the formation of hydrogen-bond interactions between daurinol and the polar side chains in the catalytic site of topoisomerase IIα. As shown in Fig. 1A, the model of daurinol bound to topoisomerase IIα indicated the following important interactions: the hydrogen bond between the benzodioxol functional group of daurinol and the Lys168 side chain of the enzyme and the interactions between the naphthofuranone ring of daurinol and water in topoisomerase IIα (Fig. 1A and C). Most importantly, the binding site of daurinol is identical to the binding site of AMP-PNP, a non-hydrolyzable ATP analog (Fig. 1B). Therefore, we speculated that daurinol inhibits human topoisomerase IIα by targeting its ATPase domain.

To verify the hypothesis from the molecular docking study, we measured human topoisomerase IIα activity in the presence of different concentrations of ATP. We previously reported the inhibitory activity of daurinol against topoisomerase IIα to be in the millimolar range (5). To more accurately evaluate topoisomerase IIα inhibition, we tested micromolar concentrations of daurinol in this study. As shown in Fig. 2, micromolar concentrations of daurinol potently inhibited the catalytic activity of human topoisomerase IIα and prevented the relaxation of the supercoiled DNA substrate. Notably, increasing the ATP concentration resulted in decreased inhibitory activity. When the enzyme reaction was performed in the presence of 0.5 mM ATP, daurinol inhibited the enzyme activity at concentrations of 10–100 μM. However, in the presence of 2 mM ATP, only 100 μM daurinol clearly inhibited the catalytic activity of topoisomerase IIα (Fig. 2). These data are in agreement with the molecular docking predictions.

To further confirm that daurinol inhibits human topoisomerase IIα by targeting the ATP-binding pocket, we measured the inhibitory activity of daurinol against human topoisomerase I using an in vitro biochemical assay. Because topoisomerase I does not contain an ATPase domain and its catalytic activity is ATP-independent (17), we anticipated that daurinol, which targets ATP binding, would not inhibit human topoisomerase I activity. As expected, daurinol did not severely inhibit the catalytic activity of topoisomerase I at the concentrations (up to 100 μM) tested, whereas camptothecin (a topoisomerase I inhibitor) clearly inhibited the catalytic activity of topoisomerase I (Fig. 3).

We also measured the ATP hydrolysis activity of human topoisomerase IIα in the presence or absence of daurinol using the malachite green method to confirm that daurinol inhibits topoisomerase IIα by targeting its ATPase domain. Both daurinol and novobiocin (a topoisomerase II inhibitor that blocks ATP binding) (12,18) significantly inhibited the ATP hydrolysis activity of human topoisomerase IIα (Fig. 4). Taken together, these results demonstrate that micromolar daurinol selectively inhibits human topoisomerase IIα by targeting the ATPase domain.

Next, we measured the anti-proliferative activity of daurinol in various human cancer lines. We previously reported the anticancer activity of daurinol in colorectal cancer cells compared to etoposide. However, etoposide is usually used to treat various cancers such as ovarian, small cell lung, and testicular cancer rather than colorectal cancer. Therefore, to evaluate the overwhelming anticancer effects of daurinol as a novel alternative to etoposide, we measured the anti-proliferative activity in other cancer cell types. We tested a total of 9 different cell lines (3 ovarian cancers, 4 small cell lung cancers and 2 testicular cancers). Daurinol potently inhibited ovarian cancer cell lines, particularly NIH:OVCAR-3 and SNU-840 cells, compared to etoposide. Daurinol also inhibited small-cell lung cancer cells, but the potency of daurinol was similar to or weaker than that of etoposide (Table I). Because the strongest anti-proliferative activity of daurinol was observed in SNU-840 cells compared to etoposide, we selected SNU-840 for further investigation.

We evaluated the effects of daurinol on cell cycle distribution in human ovarian cancer cells using flow cytometric DNA content analysis. Treatment with daurinol for 24 or 48 h apparenly induced cell cycle arrest in S phase. By contrast, etoposide significantly induced cell cycle arrest in the G2/M phase (Fig. 5). Daurinol also induced an accumulation of the cell population in S phase in NCCIT human testicular cancer cells (data not shown).

We also investigated the expression of cell cycle regulatory proteins such as cyclins and cyclin-dependent kinases to elucidate the molecular mechanism underlying the S phase arrest induced by daurinol in SNU-840 cells. Daurinol increased the expression of cyclins E and A (Fig. 6), which associate with Cdk2. The resulting complexes cyclin E/Cdk2 and cyclin A/Cdk2 regulate the initiation and progression of S phase, respectively (19,20). The expression of cyclin E was increased after 24 h of treatment; thus, we speculated that daurinol accelerated the initiation of S phase at 24 h. The expression of cyclin A was increased at both 24 and 48 h, suggesting that daurinol persisted in the enhanced S phase progression. In addition, daurinol increased E2F-1 expression at 24 and 48 h (Fig. 6). E2F-1 is a transcription factor that promotes the transcription of cyclin E and A (21). In summary, the enhanced expression of cyclin E, cyclin A and E2F-1 appear to be primarily responsible for the S phase arrest induced by daurinol.

Finally, we investigated the effect of daurinol and etoposide on the cell and nuclear size of SNU-840 cells using microscopy and flow cytometry (15). Abnormal cell and nuclear enlargement was previously correlated with the side effects of anticancer agents (5). Microscopic observation showed that etoposide induced cell and nuclear enlargement in SNU-840 cells, while daurinol did not (Fig. 7A). To more accurately evaluate this phenomenon, we performed flow cytometric analysis of the fluorescence pulse signal in cells stained with DNA-binding fluorescent dye. The FSC-H values, which indicate the cell size, of etoposide-treated cells were higher than those of control or daurinol-treated cells (Fig. 7B); the mean values of FSC-H were 554.8±10.5, 587.3±19.6, 578.2±28.0, 604.6±44.2 and 743.3±33.9 for the control, 2.5, 5 and 10 μM daurinol and 10 μM etoposide-treated cells, respectively. Similarly, the FL2-W values (pulse width of PI fluorescence), which indicate the nuclear size (16), of etoposide-treated cells were higher than those of the control- or daurinol-treated cells (Fig. 7C); the mean FL2-H values were 197.4±4.5, 218.9±7.9, 237.0±14.4, 224.9±10.3 and 313.0±21.9 for the control, 2.5, 5 and 10 μM daurinol- and 10 μM etoposide-treated cells, respectively. Kolmogorov-Smirnov statistical analysis demonstrated that the FSC-H and FL2-W distributions of etoposide-treated cells were significantly different from those of the control or daurinol-treated cells. There was no significant difference between the control and daurinol-treated cells (Fig. 7D and E). Based on these data, we concluded that daurinol did not induce abnormal the cell or nuclear enlargement of SNU-840 cells, in contrast to etoposide.