*2.2. Biology*

A previously designed [64], real-time, hexadecameric oligonucleotide biosensor with 5(6)-carboxyfluorescein (FAM) at the 50 end and fluorophore quencher BHQ1 (Black Hole Quencher-1) at the 30 -end was used to determine the inhibitory properties of the new compounds.

The results of the Tdp1 assay for the arylcoumarin derivatives are shown in Figure 2 and Supplementary Table S1. All arylcoumarin derivatives containing a geraniol residue (**3aa–3da**) showed high inhibitory activity, with IC<sup>50</sup> values in the submicromolar range; compound **3ac** with a bromine atom in the aromatic ring was slightly less active. Among the derivatives of nopol **3ab–3db**, only the fluorine derivative **3bb** showed a markedly lower activity; the remaining compounds were comparable in activity with the derivatives of geraniol. Since it was necessary to use column chromatography to obtain nopol-arylcoumarin hybrids, which complicates and increases the cost of the synthesis process, geraniol-containing inhibitors are more promising for further studies.

Almost all derivatives of (−)- and (+)-myrtenols (**3ac–3dc** and **3ad–3dd**, respectively) showed similar inhibitory activity with IC<sup>50</sup> values in the 0.4-1.0 µM range, except compound **3cc**. Interestingly, compounds **10a, c, d** containing a benzyl substituent instead of monoterpenoid fragments were significantly less active than most of their monoterpenoid-containing analogues, with an IC<sup>50</sup> in the micromolar range.

**Figure 2.** The Tdp1 inhibitory activities of compounds **3aa**–**3dd** and **10a**, **c**, **d**. Furamidine (Fur) was used as a positive control.

An analysis of the cytotoxicity of the synthesized compounds was performed on cell lines of human breast adenocarcinoma MCF-7 and human cervical cancer HeLa. It turned out that cytotoxicity is absent or insignificant in the entire range of studied concentrations (up to 100 µM) for all the tested compounds, which makes it possible to use them as tumor sensitizers for currently-used antitumor drugs without introducing additional toxic burden (Figure 3).

**Figure 3.** The effect of compounds **3** on the survival of cells of the lines MCF-7 (**a**) and HeLa (**b**).

Since most monoterpene-arylcoumarin hybrids showed comparable inhibitory activity against Tdp1 (~0.5 µM) and no or limited cytotoxicity, we selected a candidate for subsequent studies based on the following considerations. Derivatives of nopol were the most complex compounds to synthesize and purify. Therefore, they were excluded from further consideration. Since we previously obtained contradictory results in in vivo experiments with the myrtenol derivative **2** (unpublished data), in this work, we decided to focus on geraniol derivatives for the in vivo studies. Among the three derivatives of geraniol that showed similar inhibitory activity (Figure 2), we selected compound **3ba** containing a fluorine atom in the para position of the aryl substituent, which can contribute to greater metabolic stability of the inhibitor [65]. In addition to the activity assay that we reported above, we wanted to confirm the interactions between compound **3ba** and Tdp1 before progressing to in vivo studies. Using an intrinsic tryptophan fluorescence quenching assay that we previously applied to study the binding interactions of Tdp1 and its inhibitors [43,45,48,49,52], we evaluated the binding of compound **3ba** to recombinant Tdp1. Clear quenching of the Tdp1 intrinsic fluorescence was observed upon the addition of **3ba** (Supplementary Figure S1). Titration experiments were then performed to determine the dissociation constant (*K*D) of compound **3ba** with Tdp1. A *K*<sup>D</sup> value of 63.0 ± 11 µM was obtained, indicating that it is a reasonable binder to the enzyme. This confirmed that the inhibition efficacy of compound **3ba** was due to binding to the enzyme, and gave us confidence to progress with this compound towards in vivo studies.

A study of the influence of **3ba** on the antitumor effect of tpc (topotecan) was performed using a murine Krebs-2 carcinoma model. An ascitic tumor model combines the advantages of in vitro and in vivo approaches in studying the cytotoxic effect of compounds, since ascitic cells grow in the context of the organism (in vivo), and the intraperitoneal administration of drug ensures its direct contact with tumor cells (in vitro). The experiments were performed using female C57BL/6 mice, which were injected intraperitoneally with 2 <sup>×</sup> <sup>10</sup><sup>5</sup> ascitic cells on day zero. The mice were divided into six groups of ten animals each. Control group 1 did not receive treatment; group 2 received tpc in a single dose of 0.5 mg/kg of body weight intraperitoneally after 2 days; group 3 received tpc as described above, and **3ba** at a dose of 80 mg/kg intraperitoneally; group 4 received tpc and **3ba** 40 mg/kg; group 5 received tpc and **3ba** 20 mg/kg; and finally, group 6 received **3ba** 80 mg/kg only.

The combined use of **3ba** at a maximum concentration of 80 mg/kg with tpc (group 3) led to a significant decrease in the weight of the ascitic tumor compared to the use of only tpc (Figure 4). The **3ba** dose of 40 mg/kg also caused a decrease in ascites weight, although the difference between groups 2 and 4 was not significant. The dose of **3ba** 20 mg/kg co-administered with tpc, as well as the use of **3ba** (80 mg/kg) in the absence of tpc, did not affect tumor growth.

Int. J. Mol. Sci. 2019, 20, x FOR PEER REVIEW 7 of 20

Int. J. Mol. Sci. 2019, 20, x FOR PEER REVIEW 7 of 20

Figure 4. Box plot of the 3ba influence on the antitumor effect of tpc against Krebs-2 carcinoma with intraperitoneal administration. P1–2 = 0.002; P1–3 = 0.00013; P2–3 = 0.04. The differences between group 2 and groups 4–6 are not significant. **Figure 4.** Box plot of the **3ba** influence on the antitumor effect of tpc against Krebs-2 carcinoma with intraperitoneal administration. P1–2 = 0.002; P1–3 = 0.00013; P2–3 = 0.04. The differences between group 2 and groups 4–6 are not significant. Figure 4. Box plot of the 3ba influence on the antitumor effect of tpc against Krebs-2 carcinoma with intraperitoneal administration. P1–2 = 0.002; P1–3 = 0.00013; P2–3 = 0.04. The differences between group 2 and groups 4–6 are not significant.

The number of tumor cells in ascites (Figure 5) in the control group (1750 million per mouse) and in group 2, which received only tpc (950 mln), was significantly different, i.e., by half, p = 0.005. The size of ascites in group 3 (tpc + 80 mg/kg 3ba) was very small; we managed to extract ascitic fluid to determine the number of cells in only one mouse; the number of tumor cells in this mouse was 250 mln. Differences in other groups are not significant. The number of tumor cells in ascites (Figure 5) in the control group (1750 million per mouse) and in group 2, which received only tpc (950 mln), was significantly different, i.e., by half, p = 0.005. The size of ascites in group 3 (tpc + 80 mg/kg **3ba**) was very small; we managed to extract ascitic fluid to determine the number of cells in only one mouse; the number of tumor cells in this mouse was 250 mln. Differences in other groups are not significant. The number of tumor cells in ascites (Figure 5) in the control group (1750 million per mouse) and in group 2, which received only tpc (950 mln), was significantly different, i.e., by half, p = 0.005. The size of ascites in group 3 (tpc + 80 mg/kg 3ba) was very small; we managed to extract ascitic fluid to determine the number of cells in only one mouse; the number of tumor cells in this mouse was 250 mln. Differences in other groups are not significant.

Figure 5. A box plot of the 3ba influence on the number of tumor cells in ascites. Figure 5. A box plot of the 3ba influence on the number of tumor cells in ascites. **Figure 5.** A box plot of the **3ba** influence on the number of tumor cells in ascites.

We then examined the effect of 3ba in combination with tpc on the lifespan of mice. C57BL/6 mice were intraperitoneally inoculated with the ascites variant of Lewis carcinoma. Group 1: control without treatment; group 2: tpc 0.5 mg/kg intraperitoneally; group 3: tpc and 3ba intraperitoneally We then examined the effect of 3ba in combination with tpc on the lifespan of mice. C57BL/6 mice were intraperitoneally inoculated with the ascites variant of Lewis carcinoma. Group 1: control We then examined the effect of **3ba** in combination with tpc on the lifespan of mice. C57BL/6 mice were intraperitoneally inoculated with the ascites variant of Lewis carcinoma. Group 1: control

without treatment; group 2: tpc 0.5 mg/kg intraperitoneally; group 3: tpc and 3ba intraperitoneally

120 mg/kg; and group 4: 3ba 120 mg/kg only. The results are given in Figure 6.

120 mg/kg; and group 4: 3ba 120 mg/kg only. The results are given in Figure 6.

without treatment; group 2: tpc 0.5 mg/kg intraperitoneally; group 3: tpc and **3ba** intraperitoneally 120 mg Int. J. Mol. Sci. /kg; and group 4: 2019, 20, x FOR PEER REVIEW 8 of 20 **3ba** 120 mg/kg only. The results are given in Figure 6.

Figure 6. The influence of 3ba in combination with tpc on the lifespan of mice. The numbers above the boxes indicate the average lifespan in the group. **Figure 6.** The influence of **3ba** in combination with tpc on the lifespan of mice. The numbers above the boxes indicate the average lifespan in the group.

When using a combination of tpc and 3ba, a significant increase in lifespan was noted by 26% (p = 0.0065) compared with mice receiving only tpc, and by 42% compared with the control group (p = 0.0002). Monotherapy with tpc or 3ba at selected doses prolonged the life of experimental animals unreliably, i.e., by 13–19%, р > 0.05. When using a combination of tpc and **3ba**, a significant increase in lifespan was noted by 26% (p = 0.0065) compared with mice receiving only tpc, and by 42% compared with the control group (p = 0.0002). Monotherapy with tpc or **3ba** at selected doses prolonged the life of experimental animals unreliably, i.e., by 13–19%, > 0.05.

#### 2.3. In Silico *2.3. In Silico*

Figure 7.

#### 2.3.1. Molecular Modeling 2.3.1. Molecular Modeling

The 19 compounds were docked into the binding site of Tdp1 (PDB ID: 6DIE, resolution 1.78 Å) [66] with three water molecules (HOH 814, 821 and 1078). It has been shown that keeping these crystalline water molecules improves the prediction quality of the docking scaffold [45]. The modeling shows that all the ligands have a plausible binding mode and good scores with the four scoring functions used, i.e., Astex Statistical Potential (ASP) [67], improved Piecewise Linear Potential (ChemPLP) [68], ChemScore (CS) [69,70] and GoldScore (GS) [71]; the results are given in Table S2, Supplementary Information. Considering 3ba, one of the most active compounds, the coumarin moiety occupies the hydrophilic binding region, which contains amino acids such as threonine and glutamic acid, whilst the alkene side chain occupies the hydrophobic region formed by isoleucine, leucine, and phenylalanine. The carbonyl on the benzopyrone group forms hydrogen bonds with the amine side chain groups of Lys495 and Asn516. The predicted binding mode of 3ba is shown in The 19 compounds were docked into the binding site of Tdp1 (PDB ID: 6DIE, resolution 1.78 Å) [66] with three water molecules (HOH 814, 821 and 1078). It has been shown that keeping these crystalline water molecules improves the prediction quality of the docking scaffold [45]. The modeling shows that all the ligands have a plausible binding mode and good scores with the four scoring functions used, i.e., Astex Statistical Potential (ASP) [67], improved Piecewise Linear Potential (ChemPLP) [68], ChemScore (CS) [69,70] and GoldScore (GS) [71]; the results are given in Table S2, Supplementary Information. Considering **3ba**, one of the most active compounds, the coumarin moiety occupies the hydrophilic binding region, which contains amino acids such as threonine and glutamic acid, whilst the alkene side chain occupies the hydrophobic region formed by isoleucine, leucine, and phenylalanine. The carbonyl on the benzopyrone group forms hydrogen bonds with the amine side chain groups of Lys495 and Asn516. The predicted binding mode of **3ba** is shown in Figure 7.

**Figure 7.** The docked configuration of **3ba** in the binding site of Tdp1 as predicted using the ChemPLP scoring function. (**a**) The protein surface is rendered. The ligand occupies the binding pocket. Blue depicts a hydrophilic region with a partial positive charge on the surface; brown depicts hydrophobic region with a partial negative charge and grey shows neutral areas. (**b**) Hydrogen bonds are shown as green lines between the ligand and residues Lys495 and Asn516. The water molecules also form hydrogen bonds with Ser514 and Lys459.
