Isoeugenol is a selective potentiator of camptothecin cytotoxicity in vertebrate cells lacking TDP1

Camptothecin (CPT), a topoisomerase I (TOP1) inhibitor, exhibits anti-tumor activity against a wide range of tumors. Redundancy of TOP1-mediated repair mechanisms is a major challenge facing the efficiency of TOP1-targetting therapies. This study aims to uncover new TOP1 targeting approaches utilising a selection of natural compounds in the presence or absence of tyrosyl DNA phosphodiesterase I (TDP1); a key TOP1-mediated protein-linked DNA break (PDB) repair enzyme. We identify, isoeugenol, a phenolic ether found in plant essential oils, as a potentiator of CPT cytotoxicity in Tdp1 deficient but not proficient cells. Consistent with our cellular data, isoeugenol did not inhibit Tdp1 enzymatic activity in vitro nor it sensitized cells to the PARP1 inhibitor olaparib. However, biochemical analyses suggest that isoeugenol inhibits TDP2 catalytic activity; a pathway that can compensate for the absence of TDP1. Consistent with this, isoeugenol exacerbated etoposide-induced cytotoxicity, which generates TOP2-mediated PDBs for which TDP2 is required for processing. Together, these findings identify isoeugenol as a potential lead compound for developing TDP2 inhibitors and encourage structure-activity relationship studies to shed more light on its utility in drug discovery programs.

Scientific RepoRts | 6:26626 | DOI: 10.1038/srep26626 natural compounds in combination with CPT in the presence or absence of TDP1, with the goal of unraveling novel approaches to target TOP1.
We assembled a library of natural derived chemicals that are isolated from either plant or animal origin. Natural products provide a vast resource for drug discovery research. For example, in 1971 and as a part of a National Cancer Institute program, paclitaxel was identified as the active constituent from a crude extract of the bark of Taxus brevifolia 9 . Paclitaxel is the first taxane to enter clinical trials as a chemotherapeutic agent and is used against ovarian, breast, and lung cancers 10 . Ginkgo biloba extract 761 (EGb761) is another natural compound extracted from Ginkgo biloba leaves 11 that was proposed to have anti-angiogenic and antioxidant activities.
Our screen yielded a hit compound that appears to be a promising lead as a tyrosyl DNA phosphodiesterase II (TDP2) inhibitor. Like TDP1, TDP2 is a tyrosyl DNA phosphodiesterase but that catalyzes the cleavage of 5′tyrosyl bonds that are present between topoisomerase II (TOP2) and the DNA 3 . It is known that both enzymes while structurally different, can serve as back up for each other in circumstances where one is lacking, but with much less efficiency than with their preferred substrate 12 .

Results
Establishment of sub-lethal concentrations of the library compounds. A selection of 41 diverse natural products most of which occur in foods or dietary supplements were subjected to a series of cellular and biochemical experiments to identify potential hits to improve TOP1 targeting therapy (Fig. 1A). The large therapeutic index reported for these chemicals is encouraging in terms of developing future and safe drugs for treatment of cancer. The library is comprised of molecules belonging to different natural products classes viz. alkaloids, flavonoids, saponins, terpenes, and phenylpropanoids. To determine sub-lethal doses, the cytotoxic effect of each compound was examined at six different concentrations (0.01, 0.1, 1, 5, 10, and 50 μ M) in Tdp1− /− DT40 cells (Supplementary Fig. 1). The viability assay used in this study utilizes a blue dye, resazurin, that when encountered with viable cells, is reduced to the highly fluorescent red dye resorufin 13 . The cytotoxicity, therefore, is measured as fluorescence intensity quantified by a plate reader. The selected sub-lethal doses (Table 1) were then used for the primary screen in combination with a single lethal and three sub-lethal doses of CPT (4, 1, 0.5, 0.25 nM) in the same cellular model ( Supplementary Figs 2 and 3). The combination of any of the compounds 4,  Table 1. Results are presented on a semi-log scale and represent the average of 3 biological replicates ± s.e.m. Asterisks denote statistical difference (p < 0.001, 4 nM; p < 0.05, 1 nM) using t-test. Examination of the TOP1-mediated cytotoxic effect in the presence or absence of human TDP1. Seven compounds including prunin, isoquercetin, 2,4,4′ trihydroxy chalcone, isoeugenol, xanthohumol, harmine, and thymol (4, 8, 10, 14, 21, 24, 33, resp.) that showed the highest synergistically lethal effect with CPT in Tdp1− /− cells were selected for further screening by comparing their effect on chicken DT40 Tdp1− /− and an isogeneic control stably expressing human TDP1 (hTDP1) ( Fig. 1 and Supplementary Fig. 4). The intensity of fluorescence was normalized to a DMSO control (same volume as that of the highest concentration of combined drugs) and the results were obtained from three biological replicates. All tested compounds, except compound 14, gave the same effect on both cell lines (Tdp1− /− and Tdp1− /− complemented with hTDP1). Isoeugenol annotated as compound 14 revealed the highest, reproducible cytotoxic effect (p < 0.001, 4 nM; p < 0.05, 1 nM) when combined with CPT on Tdp1− /− cells but not on cells complemented with human TDP1, wherein the viability dropped from 65.5% (CPT alone) to 34.6% (CPT plus isoeugenol) at 4 nM CPT, and a reduction of ~18.2% at 1 nM CPT ( Fig. 1G and Supplementary Fig. 4F). These results indicate that isoeugenol (4-propenyl-2-methoxyphenol), at 1 nM, does not exhibit a cytotoxic effect as a single agent, but rather holds a synergistic effect when combined with CPT in a TDP1-dependent manner. This was not an epiphenomenon of DT40 cells since isoeugenol also sensitized TDP1 depleted human MCF-7 cells to CPT, although the effect was much less apparent than DT40 cells ( Supplementary Fig. 5).
Isoeugenol does not act through the PARP1/TDP1 axis. This data suggest that isoeugenol sensitizes cells to CPT only if TDP1 is absent, and that it may suppress TDP1 parallel pathways for repairing TOP1-induced PDBs. To further ascertain that the observed effects are independent of TDP1, we next examined the effect of isoeugenol on TDP1 catalytic activity using an in vitro biochemical assay employing a 13-mer oligonucleotide Cy5.5-labelled substrate containing a 3′ -phosphotyrosine modification. TDP1, in the reaction mixture, cleaves tyrosine from the substrate allowing it to move faster and further through the gel, giving a slightly shifted band. As a positive control, total cell lysate from wild-type DT40 cells was used as a source of TDP1 while Tdp1− /− cell lysate was used as a negative control. Different concentrations of isoeugenol (200 μ M, 500 μ M, and 1000 μ M) were incubated with the 3′ -phosphotyrosine substrate along with the cell lysates and reaction products were fractionated by denaturing PAGE and analysed by fluorescence imaging. The addition of isoeugenol to the reaction did not inhibit the conversion of 3′ -PY substrate to 3′ -P (indicated by arrows) even at the highest concentration (1000 μ M), confirming that isoeugenol does not inhibit TDP1 activity ( Fig. 2A,B). We reasoned that if isoeugenol potentiates CPT cytotoxicity independently of TDP1 one would predict that co-addition of PARP1 inhibitors would be epistatic to isoeugenol, since PARP1 and TDP1 are working together in the same pathway for TOP-PDB repair. To test this hypothesis, we compared the viability of Tdp1− /− cells and controls following CPT treatment in presence and absence of isoeugenol and the FDA approved PARP inhibitor olaparib, either separately or in combination. Addition of olaparib to TDP1 proficient cells led to a marked increase in CPT cytotoxicity (Fig. 2C, p < 0.01). Importantly, co-addition of isoeugenol and olaparib did not further sensitise cells to CPT than addition of olaparib alone (Fig. 2C). In contrast to Tdp1 proficient cells, addition of olaparib to Tdp1− /− cells did not result in additional sensitisation, confirming that TDP1 and PARP1 operate in the same pathway ( Fig. 2D). Notably, whilst addition of isoeugenol alone sensitized Tdp1− /− cells to CPT (p < 0.05), co-addition of isoeugenol did not. Together, we conclude from these findings that isoeugenol potentiates CPT cytotoxicity independently of the TDP1/PARP1 axis.

Isoeugenol does not increase single-strand break (SSB) accumulation in CPT-treated cells. It
is known that TDP1 deficient cells accumulate higher levels of DNA single-strand breaks (SSBs) in the presence of TOP1 inhibitors than controls [14][15][16][17] . We therefore examined if isoeugenol would increase the level of DNA strand breaks observed in Tdp1− /− DT40 cells, using alkaline single-cell gel electrophoresis (comet assays). As expected, the lack of TDP1 caused a CPT-dependent increase in SSB accumulation (P < 0.05); yet, isoeugenol did not potentiate such effect neither alone nor in combination with CPT (Fig. 3).

Isoeugenol inhibits TDP2 catalytic activity in vitro and potentiates the cytotoxic effect of etoposide in vivo.
We previously reported that TDP2 could protect from TOP1-induced damage in the absence of TDP1 18 . Consequently, we examined whether or not isoeugenol would inhibit TDP2 activity, potentially providing a mechanistic insight explaining our data. Whole cell lysate from wild-type DT40 was employed as the source for TDP2. Incubation of 5′ -phosphotyrosine substrate with increasing concentrations of isoeugenol revealed a dose-dependent reduction in the conversion of 5′ -PY to 5′ P, which is specific for TDP2 activity (Fig. 4A). Whilst incubation with extract alone led to ~50% processing, co-addition of 200 μ M isoeugenol suppressed TDP2 activity resulting in ~34% processing (Fig. 4B). Increasing the concentration of isoeugenol to 500 μ M and 1000 μ M showed a corresponding increase in the inhibition of TDP2 activity reaching as little as 6% or no conversion, respectively. The putative TDP2 inhibitory effect of isoeugenol was not due to the vehicle in which it was dissolved (DMSO) since we kept the concentration of DMSO constant in all control and test conditions. These observations suggest that isoeugenol inhibits chicken TDP2 catalytic activity at high doses. To test if this is an epiphenomenon for chicken DT40 or is also true in human cells, we repeated the experiments using HeLa cell lysates (Fig. 4C). Consistent with its TDP2 inhibitory activity, isoeugenol also inhibited human TDP2 but less efficiently than chicken TDP2 with only 20% inhibition at the highest isoeugenol dose examined (1000 μ M). If isoeugenol inhibits TDP2 activity we reasoned that it would sensitize human cells to TOP2 poisons, which specifically require TDP2 to liberate stalled TOP2 from DNA termini. Incubation with etoposide led to a dose dependent decline in viability of both HeLa and MCF7 cells and the co-addition of isoeugenol resulted in further sensitisation (Fig. 4E,F). At 50 μ M etoposide, the viability of HeLa and MCF-7 cells dropped, upon the addition of isoeugenol, from 94.8% to 64.1% and from 91.1% to 72.1%, (P < 0.001, P < 0.05; resp.). At a higher dose of 100 μ M etoposide, viability of HeLa cells decreased from 75.5% to 49.6%, and from 91.2% to 56.9% for MCF-7 cells (P < 0.01). Importantly, addition of low micromolar doses of isoeugenol (1 μ M) sensitized HeLa cells to all concentrations of etoposide tested, as measured by clonogenic survival assays (Fig. 4G). Together, these findings identify isoeugenol as a promising lead compound with potential TDP2 inhibitory activity.

Discussion
Drug combination therapy is a promising strategy used in treating complex diseases such as cancer, cardiovascular diseases, and infectious diseases 19 . Synergistic drug combinations are recognized as effective and therapeutically specific 20 . Overcoming toxicity, side effects linked to high doses of single drugs, and drug resistance can be achieved by synergistic combinations of two or more agents. Consequently, this study aimed to achieve a synergistic therapeutic effect and minimize dose through combination between the TOP1 inhibitor CPT and natural compounds. Screening of 41 compounds from natural origin belonging to diverse classes of natural products viz. alkaloids, phenylpropanoids, saponins and terpenes in combination with CPT revealed a synergistic effect between the phenolic ether isoeugenol and CPT against Tdp1− /− DT40 cells. Such a synergistic effect encourages further research and may lead to a new anti-cancer strategy that target a specific class of tumors, such as those that develop resistance to camptothecins, for example colorectal cancers (CRCs). Our screen identified isoeugenol, a phenylpropanoid, as a compound that potentiates CPT cytotoxic effect on cells that lack TDP1. Several studies have shown that phenylpropanoids exhibit different biological activities that include analgesic, anti-inflammatory, and anti-tumor activities [21][22][23] . Isoeugenol (4-propenyl-2-methoxyphenol) is abundantly present in several plant essential oils and is regularly used in spices, perfumes, and detergents, thus is unlikely to exert toxic effects on humans 24 . It is a structural isomer of eugenol found in clove and cinnamon oil and is mostly recognized medicinally for its local anesthetic effect 25 . Notably, previous studies have shown that eugenol exhibits a topoisomerase II (TOP2) inhibition activity 26,27 , however, to the best of our knowledge, there are no studies conducted on isoeugenol. Eugenol was found to display genotoxic activity via TOP2 inhibition, halting cells in the replication phase, causing S-phase arrest and apoptosis. In addition, eugenol upregulates numerous enzymes involved in the base excision repair pathway and E2F family members in addition to its potential synergistic effect with gemcitabine and fluorouracil on HeLa cells 26 .
In an attempt to examine the mechanism through which isoeugenol potentiates CPT cytotoxic effect on TDP1-deficient cells, we have conducted a series of cellular and biochemical experiments to test specific PDB repair pathways. First, the PARP1-TDP1 pathway was assessed. PARP1 directs TDP1 towards the break induced by CPT, where TDP1 cleaves the bond between TOP1 and the DNA 3′ -terminus. This is followed by modification in the 3′ -and 5′ -termini via polynucleotide kinase phosphatase (PNKP), and finally ligation by DNA Ligase III (Lig3α ) 28 . Our findings suggest that isoeugenol potentiates CPT cytotoxicity independently of TDP1 for multiple reasons. First, only TDP1 deficient, but not proficient, cells were responsive to isoeugenol. Second, biochemical analyses failed to detect inhibitory effect of isoeugenol on established TDP1 substrates. Third, isoeugenol did not further sensitise CPT-treated cells to the PARP1 inhibitor olaparib. It is possible that isoeugenol targets a parallel nucleolytic pathway or another tyrosyl DNA phosphodiesterase that has been implicated for TOP1-mediated PDB repair, particularly in absence of TDP1. Our data favors the latter possibility since isoeugenol displayed detectable TDP2 inhibitory activity in vitro when incubated with cell lysates from Chicken DT40 or HeLa cells.
These observations were further supported by a significant potentiation of cytotoxicity inflicted by the TOP2 poison etoposide, which specifically generates TOP2-mediated PDBs for which TDP2 is required for processing. TDP2′ s discovery as a 5′ -tyrosyl DNA phosphodiesterase in 2009 29 has since impelled scientists to search for an inhibitor. It has been proposed that drug resistance to TOP2 inhibitors, such as etoposide, may stem from an acquired or intrinsic overexpression in TDP2, which as discussed earlier guards against the abortive activity of TOP2 during replication and transcription. In 2013, Oglivie et al. have managed to uncover two classes of compounds, toxoflavins and deazaflavins, as selective TDP2 inhibitors following a high-throughput screening 30 . More recently, isoquinoline-1,3-diones was also found to selectively inhibit TDP2 at a low micromolar concentration 31 . Isoeugenol exhibits a dose dependent inhibitory activity on TDP2, but not TDP1, albeit at higher concentrations. Notably, vanillin (4-hydroxy-3-methoxybenzaldehyde) a structural analogue for isoeugenol ( Supplementary Fig. 5) that only differs in an aldehyde group instead of a prop-1-ene moiety displayed no activity in our assays, suggesting that the presence of an alkenyl group is crucial for biological activity. In planta, there exists several other isoeugenol analogues particularly found in spices and medicinal herbs such as eugenol anethole, estragole, safrole, myristicin, and methyl isoeugenol, all of which have yet to be assessed and might provide more potent drug candidates. Finally, We noted that the effect of isoeugenol on Tdp1− /− DT40 cells was more prominent compared to TDP1 depleted human cells. This may reflect residual TDP1 in human cells, due to siRNA knockdown versus genetic disruption in DT40 cells, which is sufficient to mask the requirement for alterative pathways (i.e. TDP2). Alternatively, it could reflect the differential inhibition of chicken and human TDP2 by isoeugenol. In support of the latter possibility, incubation of 500 μ M isoeugenol with DT40 cell extract nearly reached full inhibition with ~100% substrate remaining (Fig. 4A,B) versus ~80% for HeLa cell extract (Fig. 4C,D). These observations point at cross species differences, which may be illuminating in future studies aiming at improving the potency of isoeugenol as a candidate for drug discovery.
In summary, we identify isoeugenol as a potentiator of CPT cytotoxicity in a TDP1 dependent manner and suggest that it acts by inhibiting TDP2 activity. Isoeugenol therefore may act as a potential lead compound for developing TDP2 inhibitors. Future structure-activity relationship studies might be warranted to improve potency and shed light on its utility in drug discovery programs.