Structure-based strategies for synthesis, lead optimization and biological evaluation of N-substituted anthra[1,2-c][1,2,5]thiadiazole-6,11-dione derivatives as potential multi-target anticancer agents

2.1 General experimental procedures

Melting points were determined by a melting point apparatus (Büchi® 545). All reactions were monitored by thin-layer chromatography (TLC; Silica Gel 60 F₂₅₄). ¹H nuclear magnetic resonance (NMR) spectra were recorded using GEMINI-300 MHz (Varian®) and AM-500 MHz (Bruker®) instruments. Chemical shift (δ) values were in ppm ranges relative to tetramethylsilane (TMS) as an internal standard. Fourier-transformed infrared (IR) spectra (KBr) were determined with a Perkin-Elmer 983G spectrometer. Mass spectra were obtained on Finnigan MAT 95 XL high-resolution mass spectroscopy (HRMS) and Finnigan/Thermo Quest MAT HRMS. Reagents and solvents were purchased from Merck and Aldrich and were used without further purification. Typical experiments illustrating the general procedures for the preparation of the anthraquinone derivatives are described below.

2.2 General procedure A: Preparation of compounds (3 ∼ 23)

A mixture of 1 (300 mg, 1 mmole) in 20 ml of ethylene glycol and 3 mmole of serial organoamine was heated to 160 °C for 30 min. After its complete conversion (as determined by TLC), the solution was cooled to 80 °C and diluted with hot water. Precipitates were filtered off, washed with water, recrystallized from ethanol, and dried to produce 3 ∼ 6. The other amino derivatives, 7 ∼ 23, were similarly obtained.

2.3 General procedure B: Preparation of compounds (3 ∼ 23)

A mixture of 2 (1 mmol) in 20 ml of dry dimethylformamide (DMF) and copper acetate was heated to 75 °C, and then serial organoamine compounds (2 mmol) in the same solvent were added. Reaction mixtures were refluxed for 2 h at 75 °C. Then, the crude products were purified by ethanol to afford the desired compounds.

2.4 Initial in vitro cytotoxicity screening of compounds

In an initial attempt to investigate the cytotoxic activities of the newly synthesized analogues 3 ∼ 23, the 50% inhibitory concentration (IC₅₀) values of each compound were measured by SRB and MTT assays against two prostate cancer cell lines (PC-3 and DU-145) obtained from the American Type Culture Collection (ATCC). Both cell lines were maintained in RPMI-1640 culture medium containing 5% heat-inactivated fetal calf serum in an atmosphere of 5% CO₂ in a humidified incubator. The tetrazolium reagent (MTT; 3-(4, 5-di-methylthiazol)-2,5-diphenyltetrazolium bromide, USB) was designed to yield a colored formazan complex upon metabolic reduction by viable cells (Denizot and Lang, 1986; Mosmann, 1983). Approximately 2 × 10³ cells were plated into each well of a 96-well plate and incubated in an 5% CO₂ atmosphere at 37 °C for 24 h. To assess the in vitro cytotoxicity, each compound was dissolved in DMSO, prepared immediately before the experiments, and diluted into complete medium before addition to cell cultures. Test compounds and the vehicle control (0.1% DMSO) were then added to the culture medium at various designated concentrations. After 72 h, an amount of 100 μL of an MTT solution was added to each well, and samples were incubated at 37 °C for 4 h. A 100 μL solution of DMSO was added to each well and incubated at 37 °C for another 3 h. The absorbance at 570 nm was measured using an enzyme-linked immunosorbent assay (ELISA) reader.

2.5 One-dose and Five-dose assays of the NCI-60 human tumor cell lines Screen program

The growth inhibition capabilities of our selected compounds were tested against a panel of 60 human cancer cell lines under the NCI Drug Screen Program. Compounds were tested initially at a single high dose (10 μM) concentration, then those who showed significant growth inhibition capabilities and fulfilled the pre-determined threshold inhibition criteria of the NCI proceeded to the Five-dose assay. Details of the methodology are described in our previous publications (Ali et al., 2016b; Huang et al., 2012, 2009, 2010; Lee et al., 2013a, 2012a, 2012b, 2015).

2.6 Selectivities of the compounds tested by the Five-dose NCI-60 assays towards different types of cancer

The panel of NCI-60 cell lines used to evaluate our compounds by the Five-dose assays includes nine tumor subpanels; leukemia, melanoma, lung, colon, kidneys, ovaries, breasts, prostate, and central nervous system (Stinson et al., 1992). In order to investigate the selectivity of our compounds towards each cancer type, we calculated their selectivity ratios by dividing the average GI₅₀values of each particular subpanel by the average GI₅₀value of all the tested cell lines (mean graph midpoint (MID)) (Ali et al., 2016b).

2.7 Identification of the drugs with similar profile to our compounds using COMPARE analysis

Results from the NCI five-dose cytotoxicity experiments of our compounds were used as “seed” in the COMPARE algorithms to correlate to those in the NCI databases using the Pearson’s correlation coefficient calculations according to our previous publication (Ali et al., 2016b).

2.8 Apoptotic morphological changes in prostate cancer cells

DU-145 cells were plated at a density of 5 × 10⁵ cells per well in a 10-cm dish and then treated with various concentrations of compound 11 for 48 h. Cells were directly examined and photographed under a phase-contrast microscope (Cheng et al., 2012).

2.9 Western blot assay

Treated cells were collected and washed with PBS. After centrifugation, cells were lysed in lysis buffer. The lysates were incubated on ice for 30 min and centrifuged at 12 × 10⁴ g for 15 min. Supernatants were collected, and protein concentrations were determined using the Bradford assay. Protein samples were electrophoresed on 12% sodium dodecylsulfate (SDS)-polyacrylamide gels and transferred to a polyvinylidene difluoride membrane (Ali et al., 2016a; Ali et al., 2020). Immunoreactivity was detected using a Western blot chemiluminescence regent system (Huang et al., 2005) β-actin was used as a loading control.

2.10 Flow cytometry

DU-145 cells (10⁵ cells/well) were treated with various concentrations of compound 11 (0.25, 0.5, or 1.5 µM) or with DMSO for 48 h. Cells were washed with phosphate-buffered saline (PBS) and removed using trypsin-EDTA (Invitrogen, Carlsbad, CA) (Ali et al., 2020). Cells were then washed with PBS, fixed, and permeabilized in 3 ml of 70% ice-cold ethanol cooled at −20 °C. After an overnight incubation, cells were washed again in PBS then stained for 30 min at room temperature with 50 μg/mL propidium iodide (PI), and 50 μg/mL DNase-free ribonuclease A in PBS for 30 min at room temperature. Cells were analyzed using a FACScan flow cytometer (Riccardi and Nicoletti, 2006).

2.11 Molecular modelling

We used the crystal structure of the DNA topoisomerase I (TOP1) from the RCSB protein databank; code: 1K4T (Staker et al., 2002), and that of the DNA topoisomerase III (TOP3) from the RCSB protein databank; code: 4CGY (Bocquet et al., 2014). We established hypothetical binding models of our test compounds to the obtained X-ray crystal structures using molecular docking and calculated the docking score for each compound.

3.1 Chemistry

According to our previous series of studies (Chang et al., 2011; 2013a, 2013b; Huang et al., 2009; Lee et al., 2015; 2013b), the structural complexity of NSC745885, arising from the presence of a tetraheterocyclic ring restricted most structure–activity relationship (SAR) modifications to derivatization of the parent lead structure rather than de novo chemical synthesis. We planned our strategy of synthesizing N-substituted anthra[1,2-c][1,2,5]thiadiazole-6,11-diones as outlined in Scheme 1. Compound 1 was prepared in good yield by the reaction of 1,2-diaminoanthraquinone with thionyl chloride and trimethylamine. In an attempt to prepare the final products (3 ∼ 23) in the highest yields, two different methods of synthesis were utilized as follows: (i) an aniline series was obtained by treating 1 with copper acetate as a catalyst in dimethylformamide (DMF) for several hours to yield the corresponding N-substituted side chains; and (ii) 1 was reacted by nucleophilic substitution of chlorine atoms to obtain 2 and then with a series of nitrogen derivatives in ethylene glycol to produce diverse N-substituted anthra[1,2-c][1,2,5]thiadiazole-6,11-dione derivatives (3 ∼ 23).

3.2 Initial in vitro cytotoxicity screening of compounds and their structure activity relationship (SAR) observations

To initially screen the active members of our library of compounds, we evaluated their cytotoxicities against two androgen-independent human prostate cancer cell lines namely PC-3 and DU-145 (Table 1). Generally, most of the tested compounds demonstrated potent cytotoxic activities in both cell lines. From our structure activity relationship (SAR) observations, we found that compounds with the alkyl amine-side chain substitutions (alkyl nitrogen series) exhibited greater antitumor activities (IC₅₀ values of 0.40 to > 15 µM) than those with the aminated benzene group substitutions (phenyl nitrogen series). In addition, results of the alkyl nitrogen series indicated that the length of the carbon side chain of the anthra[1,2-c][1,2,5]thiadiazole-6,11-dione scaffold played an important role in enhancing the cytotoxic activities of our synthesized compounds. For the phenyl nitrogen and benzyl nitrogen series, increasing the length between the core scaffold and the benzene side chain by adding a CH₃ group (22 and 23) may have increased the inhibitory activity against the prostate cancer cell lines. Compounds 3 (NSC763958) and 11 (NSC763967) were the most active derivatives with IC₅₀ values of 0.21 and 0.40 µM against PC-3 cells and 0.16 and 0.70 µM against DU-145 cells, respectively. We further investigated the effect of compound 11 on the viability of DU-145 cells and found that it exerts potent cytotoxic effect on cells in a time-dependent and dose-dependent manner (Fig. 2 (a - b)).

Close

Fig. 2

Time- and concentration-dependent cytotoxic effects of compound 11 on the human prostate cancer cell line DU-145 determined by the MTT assay after exposure of cells to the IC₅₀ dose of compound 11 for 12, 24, 36, 48, and 72 h (a) or to 0.0625, 0.125, 0.25, 0.75, 3.75, 15 μM of compound 11 for 72 h (b).

Export to PPT

3.3 Evaluating the cytotoxic activities of our compounds by the One-dose and Five-dose assays of the NCI-60 human tumor cell lines Screen program

As shown in Table 2, compounds 3 (NSC763958), 5 (NSC763965), 6 (NSC763954), 10 (NSC764639), 11 (NSC763967), 14 (NSC764640), 15 (NSC763966), 17 (NSC757967), and 18 (NSC763952) were selected for evaluation by the One-dose assays of the NCI. Results for each compound were reported as the percentage of growth of cells treated with a concentration of 10 μM of the compound compared to the percentage of growth of the untreated control cells. Among the tested compounds, 11 was the most potent one which reduced the growth of all tested cell lines by a mean of −49.52%. The most inhibited cell lines were NCI-H552, COLO 205, LOX IMVI, M14, SK-MEL-5, SN12C, DU145, and BT-549 which showed percentages of growth of −91.81%, −99.38%, −90.91%, −96.08%, −98.79%, −95.78%, −95.45%, and −99.43, respectively, after treatment with compound 11. Compound 10 achieved inhibitory effects against the SK-MEL-5 melanoma cell line and reduced its growth percentage to −57.85%. Compound 14 was effective against the melanoma cell lines MDA-MB-435, SK-MEL-5, and UACC-62 and reduced their growth percentages to −67.52%, −81.47%, and −44.11%, respectively.

Tested compounds that met the predetermined growth inhibition criteria of the NCI were selected to proceed to the Five-dose assays of the NCI-60 Human Tumor Cell Lines Screen program. Compounds 3 (NSC763958), 6 (NSC763954), 10 (NSC764639), 11 (NSC763967), 14 (NSC764640), and 17 (NSC757967) were selected for the Five-dose assays. Dose response curves were plotted for every compound using the percentages of growth of every tested cell line after the treatment with different concentrations of every compound (Fig. 3), and the 50% growth inhibitory concentration (GI₅₀), total growth inhibition (TGI), and 50% lethal concentration (LC₅₀) were calculated for every tested compound, and their values are presented in Table 3. Interestingly, compound 11 exhibited remarkable potent cytotoxic activity against most of the tested cell lines of the nine different cancer subpanels with GI₅₀ values in the range of 0.145 ∼ 2.54 μM with a sub-micromolar GI₅₀ mean value of 0.63 μM. Compound 11 showed high activity against leukemia (GI₅₀ range is 0.21 ∼ 0.43 μM), non-small-cell lung cancer (GI₅₀ range is 0.24 ∼ 1.53 μM), colon cancer (GI₅₀ range is 0.20 ∼ 1.62 μM), central nervous system (CNS) cancer (GI₅₀ range is 0.20 ∼ 1.48 μM), melanoma (GI₅₀ range is 0.15 ∼ 1.63 μM), ovarian cancer (GI₅₀ range is 0.19 ∼ 1.29 μM), renal cancer (GI₅₀ range is 0.16 ∼ 1.46 μM), prostate cancer (GI₅₀ range is 0.17 ∼ 0.25 μM), and breast cancer (GI₅₀ range is 0.25 ∼ 1.41 μM). Such results indicate the exceptional high cytotoxicity of compound 11 toward the NCI-60 human tumor cell lines, extend our initial screening observations, and emphasize the potential of this compound as a future anticancer drug. Accordingly, we aimed to characterize further and understand more the biological activities of compound 11 in this study. Along the same line, we found that compound 3 exhibited potent cytotoxic capabilities with GI₅₀ values in the range of 0.0668 ∼ 40.6 μM with a mean of 2.23 μM. The leukemia subpanel was quite sensitive to compound 3, while the non-small cell lung cancer, colon, CNS, melanoma, ovarian, renal, prostate, and breast panels were less sensitive. For compound 6, the GI₅₀ values were in the range of 0.625 μM (non-small cell lung cancer: HOP-92) to > 100 μM. Compound 10 showed GI₅₀ values in the range of 1.19 μM (melanoma: SK-MEL-5) to 94.9 μM (melanoma: SK-MEL-28). Compound 14 showed GI₅₀ values in the range of 0.627 μM (melanoma: SK-MEL-5) to 9.00 μM (melanoma: SK-MEL-28). Compound 17 showed GI₅₀ values in the range of 0.373 μM (melanoma: LOXIMVI) to >100 μM (non-small cell lung cancer: EKVX and NCI-H322M; ovarian cancer: SK-OV-3; and renal carcinoma: TK-10).

Close

Fig. 3

Dose response curves of the NCI-60 human cancer cell lines after exposure to compound 11 (RV59, NSC763967) obtained from the Five-dose assays.

Export to PPT

3.4 Selectivities of compounds tested by the Five-dose assays towards different types of cancer

We aimed to determine the selectivities of our compounds towards different types of cancer of the NCI-60 cell lines panel by calculating the selectivity ratios of the tested compounds using the obtained GI₅₀ values (Ali et al., 2016b). Every tested compound was identified as “highly selective” to the cell line panel if its ratio is more than 6, identified as “moderately selective” if its ratio is between 3 and 6, and identified as “non-selective” if its ratio is less than 3. According to Table 4, compounds 3, 6, 10, 11, 14, and 17 exhibited broad spectrum growth inhibiting activities towards the tested cancer subpanels, except the following observations. Compounds 3, and 6 were highly selective for cell lines of the Leukemia subpanel with selectivity ratios of 7.96, and 7.88, respectively. Similarly, compound 6 was highly selective for cell lines of the Prostate Cancer subpanel with a selectivity ratio of 6.06. Compound 11 showed moderate selectivity towards cell lines of the Prostate Cancer subpanel with a selectivity ratio of 3.00. Compound 17 showed moderate selectivity towards cell lines of the Ovarian Cancer subpanel with a selectivity ratio of 4.05.

3.5 Cytotoxicity of compound 11 towards normal human cells

Due to the exceptional potency of compound 11 and its promising role as an effective anticancer drug, we aimed to evaluate its cytotoxicity against three normal human cell lines and compare its results to those of the standard chemotherapeutic drug doxorubicin (DXR) to get insight into its safety profile. We observed that the survival of normal human urothelial cells (SV-HUC-1) was about 80% of the untreated population after 72 h of exposure to a concentration of 15 µM of compound 11, whereas their survival was about 42% for the same exposure time and concentration of DXR. For the normal human prostatic stromal myofibroblasts cells (WPMY-1), the survival percentages were about 50% and 19% of the untreated cells, after an exposure time of 72 h to a concentration of 15 µM of either compound 11 or DXR, respectively. Similarly, the survival percentages of normal human prostate epithelial cells (RWPE-1) were about 52% and 30% of the untreated cells, after exposure to 15 µM of either compound 11 or DXR, respectively, for 72 h. Collectively, these data show that compound 11 is a potent inhibitor of cancer cell growth while being less toxic to normal human cells compared to the standard chemotherapeutic drug doxorubicin.

3.6 Discovering drugs with similar profile to compound 11 using COMPARE analysis

Drugs exhibiting similar activity profiles in the NCI-60 cell lines experiments usually possess similar mechanisms of action and modes of drug resistance. A data mining strategy using the “COMPARE analysis” method was proved to be an effective and facile way to predict similar drugs to the test compound and the underlying molecular targets (Ali et al., 2016b). We used the fingerprint of GI₅₀ values of compound 11 in the Five-dose assay of NCI-60 cell lines as a “seed” to correlate to drugs and compounds in the “Marketed Drugs”, “Investigational Drugs”, “Mechanistic Set”, “Standard Agents”, “Synthetic Compounds”, “BEC Referral Set”, and “Diversity Set” sets of the NCI database. Selected correlated drugs and compounds of importance to our study are displayed in Table 5. Similarities between our test compound and correlated drugs are presented as “Pearson’s Correlation Coefficient” values, where a value of 1 denotes perfect positive similarity (correlation), a value of zero denotes no similarity (correlation), and a value of −1 denotes negative similarity (correlation). We observed that compound 11 is positively correlated to Bafilomycin antibiotic, Nifuroxazide, Lapachol, Tryptanthrin, Austrocortirubin, Methotrexate, Aminopterin, Combretastatin A-4, Estramustine, Tipifarnib, Raltitrexed, Cyclopentenyl cytosine, Thioguanine, Cytosine arabinoside. These findings suggest further the multi-target nature of compound 11, however, the top similar drugs in our list exert their cytotoxic actions by hampering the cell cycle and causing apoptosis to cancer cells, suggesting strongly that compound 11 possess the same effects on the cell cycle and apoptosis of cancer cells. In order to verify these findings, we aimed to characterize further compound 11 by performing further experiments to discover the mechanisms of its underlying potent activity as follows in this manuscript.

3.7 Morphological changes and apoptosis caused by compound 11 to prostate cancer cells

To investigate whether compound 11 has any effect on the apoptosis of cells, we studied the cellular morphological changes due to the treatment by compound 11 on the prostate cancer cells DU-145. As shown in Fig. 4, there were significant morphological changes to DU-145 cells typical to apoptosis such as rounding and shrinkage of cells, membrane blebbing, presence of apoptotic bodies, and cell detachment from the culture wells. The observed morphological changes were dependent on the treatment concentration of compound 11, confirming the apoptotic capabilities of the compound.

Close

Fig. 4

Apoptotic morphological changes of DU-145 cells induced by compound 11 after exposure to concentrations of either 0.25, 1, or 2.5 µM of the compound for 48 h.

Export to PPT

3.8 Detection of the change in expression of the apoptosis-related proteins after treatment with compound 11

To gain insight into the observed apoptotic activity of compound 11, we investigated the change in expression of the apoptosis-related proteins cyclin D1 (important regulator of cell proliferation and cell cycle; inhibitors of cyclin D1 induce apoptosis) and COX-2 (promotes cancer cell growth and carcinogenesis; inhibitors of COX-2 induce apoptosis) by Western blot analysis. We found that the expression levels of both cyclin D1 and COX-2 were inhibited in a concentration-dependent manner after exposure to compound 11 (Fig. 5 (a-b)). Inhibition of cyclin D1 by compound 11 suggests that it affects the cell cycle of cancer cells, accordingly, we aimed to investigate the effect of compound 11 on the cell cycle of DU-145 cells as presented in the following section. Inhibition of COX-2 by compound 11 may result in apoptosis that occurs through COX-2 independent pathways (Sobolewski et al., 2010), highlighting the multi-target properties of our compound.

Close

Fig. 5

Western blot of whole-cell extracts of DU-145 cells after treatment with different concentrations of compound 11 and their corresponding relative expression quantification. (a) Compound 11 inhibited expression of cyclin D1 in a concentration-dependent manner. (b) Compound 11 inhibited expression of COX-2 in a concentration-dependent manner.

Export to PPT

3.9 Effects of compound 11 on the cell cycle of prostate cancer cells

In order to investigate whether compound 11 exhibits its antiproliferative effect against the prostate cancer cells through induction of apoptosis, we performed flow cytometric experiments to analyze the cell cycle of treated DU-145 cells. As shown in Fig. 6, the effects of compound 11 on the cell cycle were concentration dependent. We observed an increase in the accumulation of cells in the G₀/G₁ phase of the cell cycle due to exposure to compound 11 with a significant reduction of cells in the S and G₂/M phases. Also, compound 11 produced a high population of apoptotic cells (86.04%). Such findings support the previous ones regarding the mechanisms of cytotoxicity of compound 11. An other study involving 4-amine substituted derivatives of anthra[2,1-c][1,2,5]thiadiazole-6,11-dione showed that such compounds may modulate the cell cycle of cancer cells (Dong et al., 2017), supporting our current findings.

Close

Fig. 6

Effects of different concentrations of compound 11 on the cell cycle distribution of DU-145 cells determined by propidium iodide staining for DNA followed by flow cytometry.

Export to PPT

3.10 Molecular modeling studies for topoisomerase inhibition by our compounds

Drugs with similar structure to our compounds, such as topotecan, camptothecin, and doxorubicin proved to be potent inhibitors to different types of topoisomerase by interfering with the topoisomerase-DNA complexes, leading to DNA damage and death of cell (Bali et al., 2018; Foglesong et al., 1992; Li et al., 2017; Marinello et al., 2018). It is worth noting that other 4-thio/sulfonyl and 4-amine substituted derivatives of anthra[2,1-c][1,2,5]thiadiazole-6,11-dione showed strong topoisomerase I inhibition capabilities (Dong et al., 2017), suggesting that our compounds may have similar activities.

To investigate whether our series of compounds are potential topoisomerase inhibitors, we carried out molecular modeling experiments to study the interaction between our compounds and either of topoisomerase I (TOP1) or topoisomerase III (TOP3). We selected nine of our compounds with the lowest IC₅₀ values against the PC-3 prostate cancer line (Table 1) for the molecular modeling experiments. We started with testing the binding of our compounds to the topotecan pocket of the crystal structure of human TOP1 (hTOP1)-DNA complex (Fig. 7 (a)). The docking poses of the nine compounds (Fig. 7 (b)) show that they bind to the DNA by intercalating into the G11:C12 and T10:A13 DNA base pairs, by the same mechanism as topotecan. The binding scores of our compounds (Fig. 7 (c)) revealed that they may bind to hTOP1 with high affinity. The highest binding score was 13.41 for compound 11 followed by compounds 3, 5, 9, 16, 14, 4, 10, and 1 with binding scores of 13.10, 13.01, 13.00, 12.76, 12.73, 12.32, 11.86, and 10.50, respectively. In general, the binding affinities were going along the same line with the observed cytotoxic activities and IC₅₀ values (with few exceptions), suggesting strongly that the cytotoxic activities are related to binding of compounds to hTOP1.

Close

Fig. 7

(a) Binding pocket of the human topoisomerase I (hTOP1) (PDB code 1K4T) defined by the ligand topotecan. The pocket was shown as molecular surface in green. (b) Binding poses of our compounds (marked by a black arrow) within DNA. (c) The binding scores of our compounds (1, 3, 4, 5, 9, 10, 11, 14, 16) to the hTOP1.

Export to PPT

For the TOP3, we used the crystal structure of the hTOP3-RMI1 complex where a RMI1 loop (purple color) is inserted into the interface of the DNA binding site (Fig. 8 (a)). The catalytic tyrosine residue (Tyr362) and the metal ion Mg(II) are shown inside a drawn vertical box. Since no DNA is bound with the complex, the TOP3 is in its closed conformation, accordingly, we suggested that the binding of the TOP3 inhibitor to be at the RMI1 loop binding site at the hTOP3-RMI1 interface as highlighted in Fig. 8 (a). Three potential binding pockets for the inhibitor were identified at the RMI1-hTOP3 interface. Pocket 1 as shown in Fig. 8 (b), was defined by the RMI1 loop with the loop still bound. Pocket 2 as shown in Fig. 8 (c) was at the interface between TOP3 and RMI1 and was identified while the RMI1 loop was being removed, accordingly, ligands of pocket 2 should be competitive inhibitors of the RMI1 loop. Pocket 2 could be further classified as two pockets; 2n and 2c, as the interfaces of the N-terminal and C-terminal portions of the tail loops. Therefore, the three proposed pockets of hTOP3 inhibitors can be summarized as: 1) Pocket 1 – enclosed by the RMI1 loop and TOP3, stabilizing RMI1 binding; 2) Pocket 2n – defined by the N-terminal portion of the RMI1 loop, destabilizing RMI1 binding; 3) Pocket 2c – defined by the C-terminal portion of the RMI1 loop, destabilizing RMI1 binding. Docking scores (Fig. 8 (d)) showed that topotecan is a better binder for all three pockets (although with weaker binding affinity compared to TOP1) compared to compound 11. Compound 11 exhibited the highest binding affinity towards pocket 1 of hTOP3 with a docking score of 7.97 (its binding pose is presented in Fig. 8 (e)). Compound 11 fitted with the RMI1 loop very well and bound to RMI1 with two hydrogen bonds through Lys 105 and Thr111. Moreover, the hydrophilic binding surface includes some hydrophobic residues, Leu115, Pro98, Gln97 and Gln104, which may play a role in the binding at the interface. The binding of topotecan to pocket 2n (docking score = 7.70) is shown in Fig. 8 (f). Pocket 2n has many charged residues and hydrophobic clusters. Topotecan bound by hydrogen bonding with Arg255 and Arg338 at each end of the topotecan molecule. The binding of topotecan to pocket 2c (docking score = 7.44) is shown in Fig. 8 (g). Pocket 2c has both hydrophilic and hydrophobic features. topotecan bound to the pocket by hydrogen bonding with His288, Arg289 and Ser132 at one end and Glu265 at the other. Moreover, the hydrophobic residues Val93, Pro131, Leu210, Leu290, and Phe291 stabilized binding of the ligand. In summary, three different inhibition mechanisms are proposed in Fig. 8 (h). The ligand of Pocket 1 is stabilizing the RMI1 loop. The ligands of Pocket 2n and 2c inhibit the TOP3 activity by interfering with the RMI1 binding. Also, our compounds may have high binding affinity for TOP1 but not for TOP3, suggesting that they are TOP1 specific inhibitors.

Close

Fig. 8

(a) Human topoisomerase III (hTOP3) (PDB code 4CGY) (green) in complex with RMI1 loop (purple). Tyr362 and Mg(II) are shown as stick and ball rendering. The proposed inhibitor binding pocket is also labeled. (b) The identified pocket 1 of hTOP3 presented in green molecular surface. (c) The identified pockets 2, 2n, and 2c of hTOP3 presented in green molecular surfaces. (d) The binding scores of topotecan and compound 11 to pockets 1, 2n, and 2c of hTOP3. (e) Docking pose of compound 11 in pocket 1 of hTOP3 (Score 7.97). (f) Docking pose of topotecan in pocket 2n of hTOP3 (Score 7.70). (g) Docking pose of topotecan in pocket 2c of hTOP3 (Score 7.44). (h) Docking poses of ligands in pockets 1, 2n and 2c of hTOP3 with consequent inhibition effects.

Export to PPT