A Dual-Sensor-Based Screening System for In Vitro Selection of TDP1 Inhibitors

DNA sensors can be used as robust tools for high-throughput drug screening of small molecules with the potential to inhibit specific enzymes. As enzymes work in complex biological pathways, it is important to screen for both desired and undesired inhibitory effects. We here report a screening system utilizing specific sensors for tyrosyl-DNA phosphodiesterase 1 (TDP1) and topoisomerase 1 (TOP1) activity to screen in vitro for drugs inhibiting TDP1 without affecting TOP1. As the main function of TDP1 is repair of TOP1 cleavage-induced DNA damage, inhibition of TOP1 cleavage could thus reduce the biological effect of the TDP1 drugs. We identified three new drug candidates of the 1,5-naphthyridine and 1,2,3,4-tetrahydroquinolinylphosphine sulfide families. All three TDP1 inhibitors had no effect on TOP1 activity and acted synergistically with the TOP1 poison SN-38 to increase the amount of TOP1 cleavage-induced DNA damage. Further, they promoted cell death even with low dose SN-38, thereby establishing two new classes of TDP1 inhibitors with clinical potential. Thus, we here report a dual-sensor screening approach for in vitro selection of TDP1 drugs and three new TDP1 drug candidates that act synergistically with TOP1 poisons.


Introduction
Modern small-molecule-based anticancer treatments, especially within precision medicine, are based on targeting specific cellular mechanisms or enzymatic reactions driving the cancer. With the increased knowledge of the biological mechanisms driving different cancer types, specific anticancer targets are identified, and with the large amount of already synthesized small molecular compounds, the modern approach often is to perform a wide screen of small molecule panels for inhibitory effect of the relevant targets. DNA sensors capable of determining the effect of small molecular compounds in screening-based setups are therefore becoming increasingly important for discovery, characterization, and validation of new anticancer drugs [1][2][3].

TDP1 In Vitro Activity Assay
In vitro testing of the small molecules' capability of TDP1 inhibition was carried out by using a fluorescent TDP1 DNA sensor (5 -ATTO488-AAA GCA GGC TTC AAC GCA ACT Sensors 2021, 21, 4832 3 of 16 GTG AAG ATC GCT TGG GTG CGT TGA AGC CTG CTT T-BHQ1-3 , DNA technology) described in P. Jensen et al. [51] and A.-K. Jakobsen et al. [63]. Then, 20 ng of purified human TDP1 (purified as described in Jensen et al. [51]) was incubated with the potential TDP1 inhibitors in a final concentration of 80 µM for each drug or DMSO equivalent to the DMSO content of the potential TDP1 inhibitors (1.6%). After incubation, a final concentration of 0.5 µM TDP1 DNA sensor was added on ice in a buffer containing 20 mM Tris-HCl, pH 8, 100 mM KCl, 10 mM DTT, 10 mM EDTA, and 0.05% Triton X-100. A final volume of 25 µL reaction was transferred to a black 384-well plate, and fluorescence was measured every 30 s for 1 h at 494 nm excitation and 518 nm emission, in a FlexStation ® 3 Multi-Mode Microplate Reader, at 37 • C. The initial linear slope was used as a relative measure of TDP1 activity. All measurements were performed in triplicate.
To test if the different groups were statistically significantly different from each other, the means of the two groups were compared by using a paired t-test.
The inhibitory effect of the compounds was evaluated by comparing the TDP1 activity at the 13 different concentrations of drug to the no-drug control. The data were analyzed with GraphPad Prism version 8.4.1, using non-linear curve fitting.
The REEAD assay was carried out as previously described in Stougaard et al. [49]. Briefly, SuperEpoxy2 slides were coupled with 10 µM REEAD primer specific for the TOP1 substrate in 300 mM Na 3 PO 4 , pH 8.5, and incubated overnight in a humidity chamber with saturated NaCl. The slides were blocked at 50 • C in blocking buffer (50 mM Tris, 50 mM Tris-HCl, and 32 mM ethanolamine, pH 9), followed by wash in wash buffer 1 (4× SSC and 0.1% SDS).
Rolling-circle amplification was carried out in 1× Phi29 buffer supplemented with 0.2 µg/µL BSA, 1 unit/µL phi29 polymerase, and 1 mM dNTP. The slides were incubated for 1 h at 37 • C and subsequently washed for 1 min at 20 • C in wash buffer 2 (100 mM Tris-HCl, 150 mM NaCl, and 0.3% SDS), followed by 1 min wash in wash buffer 3 (100 mM Tris-HCl, 150 mM NaCl, and 0.05% Tween20). Finally, the slides were dehydrated in 96% in EtOH for 1 min. The generated rolling-circle products were visualized by hybridization of 200 nM fluorescent detection probe (FAM labeled) to the rolling-circle products in hybridization buffer (20% formamide, 2× SSC, and 5% glycerol) for 30 min at 37 • C. The slides were washed in wash buffer 2 and 3, followed by dehydration in EtOH. The slides were mounted with Vectashield without DAPI and then visualized by using 60× objective in a fluorescent microscope (Olympus IX73). Signals detected in an average of 12 microscopic images were counted in ImageJ, and all data were plotted as mean with standard deviation. All measurements were performed in triplicate.

Cell Survival after Co-Treatment of TDP1 Drugs and SN-38
To test for survival of HeLa cells after treatment with TDP1 inhibitors and/or SN-38, HeLa cells were seeded in a quantity of 5000 cells/well in 100 µL culture media in a 96-well culture plate. The following day, the cells were treated with DMSO alone (0.5%), TDP1 inhibitor alone, SN-38 alone, or a combination of TDP1 inhibitor and SN-38. The final concentration of SN-38 was 10 nM, and the final concentration of TDP1 inhibitor was 17.5 µM for NAF-15 and 35 µM for PSTHQ-2 and PSTHQ-13. After 72 h of drug treatment, cell survival was measured by using PrestoBlue Cell Viability Reagent (ThermoFisher). Each well was treated with 10 µL PrestoBlue and incubated at 37 • C and 5% CO 2 for 30 min, before fluorescence was measured at 560 nm excitation and 590 nm emission in a FlexStation ® 3 Multi-Mode Microplate Reader. The mean of fluorescence from 12 wells without cells was subtracted from the results in order to correct for background. Twelve different wells were treated with the same drug conditions.
To test if the different groups were statistically significantly different from each other, the means of the two groups were compared by using a paired t-test.
To test if the different groups were statistically significantly different from each other, the means of the two groups were compared by using a paired t-test.

Structure of Potential TDP1 Inhibitors
We have, in an unpublished study, screened 40 small molecular compounds for inhibitory effect on TDP1 activity to select the best compounds for this study. The 40 compounds were previously developed and investigated for TOP1 inhibitory effect or antileishmanial activity. Based on the structural comparison of the most potent TDP1 inhibitors among the 40 compounds, seven small molecular compounds were selected from a library of compounds developed at Department of Organic Chemistry I, University of Basque Country (UPV/EHU), and resynthesized for investigation in this study. The molecular structures, chemical formula, and molecular weight of the seven compounds are presented in Figure 1. pounds were previously developed and investigated for TOP1 inhibitory effect or antileishmanial activity. Based on the structural comparison of the most potent TDP1 inhibitors among the 40 compounds, seven small molecular compounds were selected from a library of compounds developed at Department of Organic Chemistry I, University of Basque Country (UPV/EHU), and resynthesized for investigation in this study. The molecular structures, chemical formula, and molecular weight of the seven compounds are presented in Figure 1.

In Vitro Inhibition of TDP1 and TOP1 Enzyme Activity with NAF-15, PSTHQ-2, and PSTHQ-13
In order to investigate the inhibitory effect of the compounds NAF-15, NAF-17, NAF-18, NAF-POEt, HNAF-1, PSTHQ-2, and PSTHQ-13 ( Figure 1) on TDP1 enzymatic activity, we used a DNA sensor previously developed by our group and designed to measure the ability of TDP1 to remove 3′-adducts [51]. The DNA sensor is composed of a DNA oligonucleotide that fold into a hairpin structure having a fluorophore at the 5′-end and quencher at the 3′-end ( Figure 2a). TDP1 mediated hydrolysis removes the quencher, thereby enabling light emission from the fluorophore by which TDP1 activity can be measured as fluorescent development over time. In the experimental setup, we measured the

In Vitro Inhibition of TDP1 and TOP1 Enzyme Activity with NAF-15, PSTHQ-2, and PSTHQ-13
In order to investigate the inhibitory effect of the compounds NAF-15, NAF-17, NAF-18, NAF-POEt, HNAF-1, PSTHQ-2, and PSTHQ-13 ( Figure 1) on TDP1 enzymatic activity, we used a DNA sensor previously developed by our group and designed to measure the ability of TDP1 to remove 3 -adducts [51]. The DNA sensor is composed of a DNA oligonucleotide that fold into a hairpin structure having a fluorophore at the 5 -end and quencher at the 3 -end ( Figure 2a). TDP1 mediated hydrolysis removes the quencher, thereby enabling light emission from the fluorophore by which TDP1 activity can be measured as fluorescent development over time. In the experimental setup, we measured the activity of 20 ng of purified TDP1 in terms of fluorescent emission generated in the absence or presence of 80 µM of each of the molecular compounds, as shown in Figure 2b. The relative TDP1 activity after treatment with the small molecular compounds was calculated as the slope of the initial linear phase of florescence development and presented in Figure 2c. As seen from Figure 2c, addition of NAF-15, PSTHQ-2, and PSTHQ-13 resulted in a statistically significant inhibition of TDP1 activity when comparing to the DMSO control. DMSO was used as a control since all compounds were dissolved in DMSO. NAF-15, PSTHQ-2, and PSTHQ-13 reduced the TDP1 activity with 64.1%, 90.4%, and 86.1%, respectively. The IC50 value for TDP1 inhibition was measured for all three compounds by incubating the TDP1 sensor with purified TDP1 and 13 different concentrations of compounds or with DMSO. Based on the results from these experiments, the IC50 values were calculated to be 37.8 µM for NAF-15, 4.28 µM for PSTHQ-2, and 13.1 µM for PSTHQ-13 (Table 1).
DMSO. NAF-15, PSTHQ-2, and PSTHQ-13 reduced the TDP1 activity with 64.1%, 90.4%, and 86.1%, respectively. The IC50 value for TDP1 inhibition was measured for all three compounds by incubating the TDP1 sensor with purified TDP1 and 13 different concentrations of compounds or with DMSO. Based on the results from these experiments, the IC50 values were calculated to be 37.8 µ M for NAF-15, 4.28 µ M for PSTHQ-2, and 13.1 µ M for PSTHQ-13 (Table 1). DMSO control contained 1.6% DMSO, which is the same DMSO amount as in the samples with potential TDP1 inhibitors. All data were normalized to DMSO control. *** Represents a p-value < 0.001 compared to the DMSO control. The four compounds, NAF-17, NAF-18, NAF-POEt, and HNAF, did not produce a significant decrease in TDP1 activity and were not included in any further experiments.
To investigate if the compounds had any effect on TOP1 activity, we tested them by using the REEAD assay illustrated in Figure 3a,b for measuring the TOP1 cleavage-ligation activity, which has previously been described by Stougaard et al. [49,64]. TOP1 activity was measured in the presence of purified TOP1 and NAF-15, PSTHQ-2, PSTHQ-13, The slope of the initial linear phase is a representation of the TDP1 activity assay visualized as a columns chart in c. (c) TDP1 activity for purified TDP1 enzyme after treatment with potential TDP1 inhibitors. All potential TDP1 inhibitors had a final concentration of 80 µM. DMSO control contained 1.6% DMSO, which is the same DMSO amount as in the samples with potential TDP1 inhibitors. All data were normalized to DMSO control. *** Represents a p-value < 0.001 compared to the DMSO control. The four compounds, NAF-17, NAF-18, NAF-POEt, and HNAF, did not produce a significant decrease in TDP1 activity and were not included in any further experiments.
To investigate if the compounds had any effect on TOP1 activity, we tested them by using the REEAD assay illustrated in Figure 3a,b for measuring the TOP1 cleavage-ligation activity, which has previously been described by Stougaard et al. [49,64]. TOP1 activity was measured in the presence of purified TOP1 and NAF-15, PSTHQ-2, PSTHQ-13, CPT, or DMSO. The results are depicted in Figure 3c. As evident from Figure 3c, TOP1 activity was not affected by any of the compounds when comparing to the DMSO control.    PSTHQ-13 (c). Cell survival was measured by using PrestoBlue. All data were normalized to DMSO control. *** Represents a p-value < 0.001 compared to TDP1 inhibitor or SN-38 alone.

Discussion
In this study, we reported a dual-sensor-based in vitro screening system for identification and validation of new TDP1 inhibitors that take into account the biological interplay between TOP1 and TDP1. The sensor system is based on two previously published DNA sensors with a fluorescent readout and measures the small molecular compounds' effect on the activity of TDP1 and TOP1. We demonstrated the use of the in vitro sensor system for screening and validation of new small molecular compounds with the ability to inhibit TDP1 activity without inhibiting TOP1 activity and validated the in vitro results in vivo.
It is of great importance to measure the inhibitory effect of potential TDP1 inhibitors on both TDP1 and TOP1 activity to ensure that the compounds inhibit only TDP1 activity and not TOP1 catalytic activity. The most well-documented application for TDP1 inhibitors is in combination with TOP1 targeting anticancer drugs [28,65,66]. In relation to this, it is important to realize that the TOP1 targeting anticancer drugs are TOP1 poisons stabilizing the transient TOP1-DNA intermediate in the TOP1 catalytic cycle [37]. This will eventually result in cytotoxic double-strand breaks in dividing cells, such as cancer cells [32]. If the potential TDP1 inhibitor blocks the catalytic activity of TOP1, TOP1 poisons cannot exert their effect. Thus, the advantage of using the TDP1 inhibitor to enhance the effect of TOP1 poisons will be lost. Therefore, the concept of testing potential TDP1 inhibitors for effect on TOP1 catalytic activity is crucial. Despite this, most published TDP1 inhibitors are not tested for inhibitory effect on TOP1 activity; however, this can easily be accomplished by using a DNA sensor, as presented in Figure 3.
Initially, 40 small molecular compounds from different studies investigating TOP1 inhibitory effect or antileishmanial activity were screened for their TDP1 inhibitory effect, using the TDP1 sensor (data not shown). Based on structural comparisons of the compounds with the strongest TDP1 inhibition, seven small molecular compounds were selected from a large library of compounds to be used for validation of the dual-sensorbased in vitro screening system. The seven small molecular compounds were, in this study, used to validate our new dual-sensor-based in vitro screening setup. Thus, they PSTHQ-13 (c). Cell survival was measured by using PrestoBlue. All data were normalized to DMSO control. *** Represents a p-value < 0.001 compared to TDP1 inhibitor or SN-38 alone.

Discussion
In this study, we reported a dual-sensor-based in vitro screening system for identification and validation of new TDP1 inhibitors that take into account the biological interplay between TOP1 and TDP1. The sensor system is based on two previously published DNA sensors with a fluorescent readout and measures the small molecular compounds' effect on the activity of TDP1 and TOP1. We demonstrated the use of the in vitro sensor system for screening and validation of new small molecular compounds with the ability to inhibit TDP1 activity without inhibiting TOP1 activity and validated the in vitro results in vivo.
It is of great importance to measure the inhibitory effect of potential TDP1 inhibitors on both TDP1 and TOP1 activity to ensure that the compounds inhibit only TDP1 activity and not TOP1 catalytic activity. The most well-documented application for TDP1 inhibitors is in combination with TOP1 targeting anticancer drugs [28,65,66]. In relation to this, it is important to realize that the TOP1 targeting anticancer drugs are TOP1 poisons stabilizing the transient TOP1-DNA intermediate in the TOP1 catalytic cycle [37]. This will eventually result in cytotoxic double-strand breaks in dividing cells, such as cancer cells [32]. If the potential TDP1 inhibitor blocks the catalytic activity of TOP1, TOP1 poisons cannot exert their effect. Thus, the advantage of using the TDP1 inhibitor to enhance the effect of TOP1 poisons will be lost. Therefore, the concept of testing potential TDP1 inhibitors for effect on TOP1 catalytic activity is crucial. Despite this, most published TDP1 inhibitors are not tested for inhibitory effect on TOP1 activity; however, this can easily be accomplished by using a DNA sensor, as presented in Figure 3.
Initially, 40 small molecular compounds from different studies investigating TOP1 inhibitory effect or antileishmanial activity were screened for their TDP1 inhibitory effect, using the TDP1 sensor (data not shown). Based on structural comparisons of the compounds with the strongest TDP1 inhibition, seven small molecular compounds were selected from a large library of compounds to be used for validation of the dual-sensor-based in vitro screening system. The seven small molecular compounds were, in this study, used to validate our new dual-sensor-based in vitro screening setup. Thus, they were tested in vitro for their ability to inhibit the activities of TDP1 and TOP1. Of these, three compounds, NAF-15, PSTHQ-2, and PSTHQ-13, all showed a high degree of TDP1 inhibition ( Figure 2). No inhibitions of TOP1 were detected (Figure 3), thereby proving the capability of the in vitro screening system for the identification of molecular compounds capable of inhibiting TDP1 without influencing the TOP1 activity. To address the performance of the three compounds, the IC50 values for TDP1 inhibition were investigated by using the TDP1 sensor and calculated to be 37.8 µM for NAF-15, 4.28 µM for PSTHQ-2, and 13.1 µM for PSTHQ-13 (Table 1). This is within the normal range compared to other TDP1-inhibiting compounds [67].
Development of compounds that specifically inhibit TDP1 without also inhibiting TOP1 will, in anticancer treatment, allow for more diverse treatment than dual inhibitors, since TDP1 is important in other repair processes than the repair of TOP1cc. So far, TDP1 has been found to participate in the repair of a wide range of 3 -lesions, such as 3 -deoxyribose phosphate ends, 3 -phosphoglycolate ends, and 3 -abasic sites [21,22,68]. Furthermore, several studies have indicated that TDP1 is involved in resolving doublestranded breaks by participation in the NHEJ repair process [24][25][26]. Additionally, TDP1 has been suggested as a promising target in the treatment of HPV-induced cancers, where it, along with PARP1, has been shown to be essential for the initial amplification of the high-risk HPV genome [69]. Because of the various repair functions of TDP1-and with the potential of more being found-the enzyme could prove to be an extremely interesting target in repair deficient tumors. If a tumor is repair-deficient in other repair processes than the ones involving TDP1, inhibition of TDP1 might be beneficial to achieve synthetic lethality and thereby increase cytotoxicity of tumor cells. Treating repair-deficient tumors with TDP1 inhibitors will most likely keep the toxicity and side effects of anticancer treatment at a minimum, since the non-tumor cells in the body will not be repair-deficient and presumably more resistant to drug treatment. Furthermore, toxicity and side effects are expected to be low with the TDP1 inhibitors developed in this study, as they have no cytotoxicity by themselves, even at high concentrations (Figures 4 and 5). The concept of synthetic lethality of repair enzymes has already been applied clinically with great success in the anticancer treatment of BRCA-deficient ovarian tumors treated with PARP1 inhibitors [70]. Interestingly, a combination of TDP1 knockdown and PARP1 inhibition has been shown to be cytotoxic in rhabdomyosarcoma cells even without BRCA alterations [8]. As the cytotoxic effects were preferential for rhabdomyosarcoma cells over those of control myoblasts, the authors hypothesized that this effect is due to the rhabdomyosarcoma cells having repair deficiencies that have to be compensated for by PARP and TDP1. This, combined with TDP1s involvement in many different repair mechanisms, suggests that TDP1 drugs may be useful in many other settings than combined TDP1/TOP1 inhibition.
Even though TDP1 drugs have many potentials besides combined TDP1/TOP1 inhibition, it is still of great importance that the TDP1 drugs can be used in combination with TOP1 poisons of the CPT family. These are already used clinically but have many side effects and are usually administered in low doses, in combination with other drugs, to reduce side effects [71]. Thus, we hypothesized that, if the TDP1 drug could act synergistically with TOP1 poisons, it could enable the use of low-dose TOP1 poison. Therefore, the three TDP1-inhibiting molecules tested in this study, NAF-15, PSTHQ-2, and PSTHQ-13, were tested in combination with the TOP1 poison, SN-38. All three compounds decreased cell survival significantly when combined with a low dose of SN-38 ( Figure 4). Furthermore, PSTHQ-2 and PSTHQ-13 were found to reduce cell survival significantly, with a concentration down to 100 nM (no lower concentrations were tested) in combination with a low dose of SN-38 ( Figure 5). The same results were observed for NAF-15 with a concentration down to 50 nM (no lower concentrations were tested). This indicates that low doses of all three TDP1 inhibitors can kill cancer cells when combined with a low dose of TOP1 poison. This is very promising for their usability in cancer treatment, where low doses of drugs are preferred to keep the side effects and toxicity of chemotherapeutics at a minimum.
Thus, we here reported, for the first time, a sensor based in vitro screening setup for identification of TDP1 inhibitors capable of taking into account the biological interplay between TDP1 and TOP1. We demonstrated that the sensor system can identify TDP1 inhibitors in vitro that do not affect TOP1 activity, and we confirmed that the identified inhibitors act synergistically with TOP1 poisons in vivo even at low doses. Using the sensor based in vitro screening setup presented in this study, we identified three new TDP1 inhibitors, namely NAF-15, PSTHQ-2, and PSTHQ-13, with promising clinical perspectives.