Reactive oxygen species (ROS) oxidize nucleotide triphosphate pools (e.g., 8-oxodGTP), which may kill cells if incorporated into DNA. Whether cancers avoid poisoning from oxidized nucleotides by preventing incorporation via the oxidized purine diphosphatase MTH1 remains under debate. Also, little is known about DNA polymerases incorporating oxidized nucleotides in cells or how oxidized nucleotides in DNA become toxic. Here we show that replacement of one of the main DNA replicases in human cells, DNA polymerase delta (Pol δ), with an error-prone variant allows increased 8-oxodG accumulation into DNA following treatment with TH588, a dual MTH1 inhibitor and microtubule targeting agent. The resulting elevated genomic 8-oxodG correlated with increased cytotoxicity of TH588. Interestingly, no substantial perturbation of replication fork progression was observed, but rather mitotic progression was impaired and mitotic DNA synthesis triggered. Reducing mitotic arrest by reversin treatment prevented accumulation of genomic 8-oxodG and reduced cytotoxicity of TH588, in line with the notion that mitotic arrest is required for ROS buildup and oxidation of the nucleotide pool. Furthermore, delayed mitosis and increased mitotic cell death was observed following TH588 treatment in cells expressing the error-prone but not wild-type Pol δ variant, which is not observed following treatments with antimitotic agents. Collectively, these results link accumulation of genomic oxidized nucleotides with disturbed mitotic progression.

Significance:

These findings uncover a novel link between accumulation of genomic 8-oxodG and perturbed mitotic progression in cancer cells, which can be exploited therapeutically using MTH1 inhibitors.

See related commentary by Alnajjar and Sweasy, p. 3459

Oxygen metabolism is critical for many cellular processes, and it is established that reactive oxygen species (ROS) can be both beneficial and deleterious to cells. Loss of balanced redox homeostasis and/or ROS may have a role in the etiology of many diseases, including cancer, cardiovascular disease, hypertension, inflammatory diseases (e.g., atherosclerosis, rheumatoid arthritis), ischemia-reperfusion injury, diabetes mellitus, neurodegenerative diseases, and ageing (1, 2). ROS causes oxidative DNA damage, either directly in the DNA, which then can be repaired by the 8-oxoguanine glycosylase OGG1, or by oxidizing the nucleobases in the free deoxynucleoside triphosphate (dNTP) pool. The human MutT homologue 1 (MTH1) protein, an oxidized purine diphosphatase, is important to prevent oxidized dNTPs, such as 8-oxodGTP and 2-OHdATP, from being incorporated into DNA (3, 4).

Previously, we and others showed that the normally non-essential MTH1 enzyme is required for survival of cancer cells (5–8), whereas others have ruled out MTH1 as an anticancer target (9–11). We showed that the wild-type MTH1 or bacterial MutT protein can rescue 8-oxodGTP incorporation and toxicity generated following MTH1 siRNA depletion and treatment with the MTH1 inhibitor (MTH1i) TH588 (6, 12), consistent with incorporation of 8-oxodGTP into DNA having toxic effects. MTH1i TH588 and the best-in-class analogue, the clinical candidate TH1579 (karonudib), possess dual pharmacology, inhibiting microtubule polymerization directly in addition to inhibiting MTH1 (10). Also, emerging data indicate MTH1 protein itself is required for mitosis and binds tubulin directly (13), which altogether are complicating the understanding of what effects are generated through mitosis or through incorporation of 8-oxodGTP.

Pol δ is a main DNA polymerase involved in lagging strand synthesis as well as being the main DNA polymerase for repair or break-induced replication (14). Recent data also suggest Pol δ is the main DNA polymerase for leading strand synthesis (15). Unlike trans-lesion synthesis polymerases, Pol δ has less tolerance to replicate across lesions and only accepts very slight base modifications to be able to polymerize DNA with modified nucleotide substrates. Hence, 8-oxodGTP is likely a poor substrate for Pol δ in cells. To investigate the cellular fate of 8-oxodGTP, we employed a recently reported protein-replacement system allowing the replacement of endogenous Pol δ with an error-prone variant (16). Using this system, we report increased incorporation of 8-oxodGTP when cells express the error prone variant are exposed to MTH1i TH588 enabling us to use this model to increase our understanding of the role of genomic 8-oxodG lesions. These experiments are highly relevant for cancer treatment as MTH1i TH1579 (karonudib) is currently evaluated in clinical trials for anticancer efficacy.

Cell culture

Cell lines were cultured at 37°C in 5% CO2. U2OS cells (2014; ATCC) were cultured in DMEM media supplemented with 10% FBS and penicillin–streptomycin (100 U/mL), A2780 Pol δ replacement cells (16) were obtained from Professor Josef Jiricny (2014; University of Zurich) and were cultured in DMEM with 5% low TET FBS and 100 μg/mL G418, 10 μg/mL blasticidin and 1 μg/mL puromycin. Cell culture reagents were from Gibco/Thermo Fisher Scientific. Cell lines were authenticated by the provider and cultivated according to the conditions specified, typically maintained in culture for 1 to 2 months before thawing of a new batch. Mycoplasma screening was performed routinely using the MycoAlert Mycoplasma Detection Kit (Lonza).

Compounds

The following chemical inhibitors were used: CENP-E inhibitor (GSK923295; SelleckChem), vincristine sulfate (Sigma Aldrich), paclitaxel (Sigma Aldrich), reversine (Axon MedChem), RO3306 (SelleckChem), Aphidicolin (Sigma Aldrich). TH588(6) was synthesized in house as reported previously.

Antibodies

The following antibodies were used: β-actin (ab6276; Abcam), ATM pS1981 (sc47739; Santa Cruz Biotechnology), ATR phospho-S428 (ab178407; Abcam), Cdk2 phospho-T14/T15 (Santa Cruz Biotechnology, sc28435-R), Chk1 phospho-S345 (#2341; Cell Signaling Technology), Chk1 (#2360; Cell Signaling Technology), GAPDH (sc-25778; Santa Cruz Biotechnology), H2A.X phospho-S139 (05-636; Millipore), Histone H3 phospho-S10 (H3-pS10; ab14955; Abcam), Histone H3 (#4318; Cell Signaling Technology), MTH1 (NB100-109; Novus Biologicals), p21 (H164; Santa Cruz Biotechnology), cleaved PARP (#9541; Cell Signaling Technology), PLK1 (#5844; Millipore), α-tubulin, DNA Pol δ catalytic subunit (p125; ab10362; Abcam).

Generation of H2B-GFP cell lines

The H2B-GFP vector was constructed by amplifying H2B from the H2B-RFP pENTR1A vector (Addgene, #22525) by PCR and subcloning the product into the pENTR1A-GFP-N2 vector (Addgene, #9364) at the HindIII and BamH1 restriction sites. The H2B-pENTR1A-GFP-N2 vector was verified by sequencing and transferred into the pLenti-CMV-Blast and pLenti-CMV-Hygro vectors using LR clonase (Invitrogen). H2B-RFP in pENTR1A (w507-1) and pENTR1A-GFP-N2 (FR1) were gifts from Eric Campeau and Paul Kaufman (Addgene plasmids, #22525 and #19364). H2B-GFP plasmids were transduced using lentivirus infections.

Western blot analysis

Cells were washed with 1× PBS and scraped in lysis buffer [50 mmol/L Tris-HCl pH 8, 150 mmol/L NaCl, 1 mmol/L EDTA, 1% Triton X-100, 0.1% SDS, 1× protease inhibitor cocktail (Thermo Fisher Scientific), and 1× Halts phosphatase inhibitor cocktail (Thermo Fisher Scientific)]. Samples were incubated on ice for 30 minutes with occasional vortexing before centrifugation to pellet insoluble material. Protein concentration was measured using BCA Assay (Thermo Fisher Scientific) before preparation in Laemmli Sample Buffer (Bio-Rad) containing 355 mmol/L final concentration 2-mercaptoethanol, and samples were denatured at 95°C for 5 minutes. Samples were run on 4% to 12% SDS-PAGE gels (Mini-Protean TGX Precast gel; Bio-Rad) and subsequently transferred to nitrocellulose membranes using the Trans-Blot Turbo instrument (Bio-Rad). Membranes were stained with Ponceau S and blocked in Odyssey TBS Blocking Buffer (Li-Cor) for 1 hour. Antibodies were diluted in Odyssey TBS Blocking Buffer and incubated overnight at 4°C. Membranes were washed with TBS supplemented with Tween-20 (0.1%) before incubation with secondary antibodies (IRDye 800CW and IRDye 680LT) diluted in TBS-Tween before further washes and imaging using the LI-COR Odyssey Fc Imaging system.

Cell viability assay

Cell viability was assessed using the resazurin assay. In the case of Pol δ replacement cells, prior to the cell viability assay, cells were seeded in six-well plates and cultured in medium supplemented with 10 ng/mL doxycycline for 3 days before trypsinization and reseeding; doxycycline supplementation was maintained for the remainder of the assay. For the cell viability assay, typically 1,500 cells per well were seeded in 90 μL media in 96-well plates, and the following day, the drug to be tested was serially diluted in media and 10 μL 10× drug dilution added to the desired well. Following 3 days treatment, viable cells were measured using resazurin (Sigma Aldrich; catalog no. R7017), as described previously (6). Measurements for each well were normalized to control wells (cells with DMSO, 100% viable control; medium with DMSO, 0% viable control) and analyzed using four-parameter logistic model in Prism 8 (GraphPad Software).

DNA fiber technique

Cells were seeded in six-well plates, typically 200,000 to 300,000 per well, to reach 70% confluency on day of experiment. For the time-course analysis in U2OS cells, 5 μmol/L TH588 was added 8, 4, and 1 hours before labeling or directly with labeling, and an 8-hour DMSO control was also included. For experiment with the Pol δ replacement cells, cells cultured in 10 ng/mL doxycycline for 4 days were treated with 1 μmol/L TH588 or mock treated with DMSO for the indicated times. Treatment was followed by pulse-labeling for 20 minutes with 25 μmol/L 5-chloro-2′-deoxyuridine (CldU; Sigma) and then 20 minutes with 250 μmol/L 5-Iodo-2′-deoxyuridine (IdU; Sigma). The inhibitor and mock treatment was present during labeling. Cells were harvested and resuspended in ice-cold PBS. Lysing of cells were performed by incubating a drop of cell suspension together with spreading buffer (200 mM Tris-HCl, pH 7.4, 50 mM EDTA, and 0.5% SDS) on SuperFrost objective glass (Thermo Fisher Scientific). DNA fibers were spread by tilting the glass slides allowing the cell lysate to slowly run down the slide. For immunodetection of CldU and IdU, acid treated fibers (2.5 M HCl for 1 hour) were stained with monoclonal rat anti-BrdUrd [1:500; Clone BU1/75 (ICR1; AbD Serotec)] and monoclonal mouse anti-BrdUrd [1:500; Clone B44, 347580 (BD Biosciences)] respectively, for 1 hour in 37°C. This was followed by incubation with Alexa Fluor 555-conjugated anti-rat and Alexa Fluor 488-conjugated anti-mouse (Molecular Probes). DNA fibers were studied in a LSM780 confocal microscope using a 64× oil objective and lengths of CldU and IdU tracks were measured using the ImageJ software (http://fiji.sc/Fiji), 50 to 100 replication tracks were scored for each condition.

Mitotic DNA replication

U2OS cells were seeded on coverslips one day before. The next day the cells were incubated for 16 hours with the CDK inhibitor RO3306 (9 μmol/L) alone or in combination with either 0.4 μmol/L Aphidicolin (APH) or 5 μmol/L TH588. Afterwards, the cells were washed with prewarmed PBS four to five times and within 5 minutes to release the cells into mitosis and incubated with fresh media containing 10 μmol/L EdU for 30 minutes at 37°C. Fixation and permeabilization was done simultaneously with PTEMF (20 mM PIPES pH 6.8, 10 mM EGTA, 0.2% Triton X-100, 1 mM MgCl2, and 4% formaldehyde) solution for 20 minutes at room temperature. Blocking was done in 3% BSA in PBST 0.5% for 1 hour at room temperature followed by EdU detection with the Click-iT EdU reaction Kit (Thermo Fisher Scientific), following the manufacturer's instructions. The nucleus was visualized with DAPI. For labeling of mitotic cells, anti-H3pS10 primary antibody was used.

High content microscopy

Pol δ replacement cells were first cultured with media supplemented with 10 ng/mL doxycycline for 3 days before seeding at 10,000 cells per well in black, clear-bottomed 96-well plates (BD Falcon). The following day, if labeling S-phase cells, cells were pulse-labeled with 10 μmol/L EdU for 20 minutes prior to fixation in 4% paraformaldehyde in PBS (Santa Cruz Biotechnology) containing 0.5% Triton X-100 for 15 minutes. Wells were washed with PBS before permeabilization in 0.3% Triton X-100. For detection of EdU labeled cells, a click reaction mix was assembled in PBS containing 2.5 mmol/L copper (II) sulphate, 10 mmol/L ascorbic acid, and 2 μmol/L ATTO-488 azide fluorophore (ATTO-TEC), which was added to wells and incubated in the dark for 30 minutes. Wells were washed in PBS, cell nuclei stained with 1 μg/mL DAPI for 10 minutes, and further washed in PBS prior to imaging. Plates were imaged on an ImageXpress high-throughput microscope (Molecular Devices) with a 20× objective. Images were analyzed (nuclei counting, DAPI, and EdU nuclear intensity measurements) with CellProfiler (Broad Institute) and data handled in Excel and plotted in Prism 8.

Modified comet assay

Cells were seed in six-well plates at a density of 150 to 200,000 cells/well and either on the same day (suspension cells) or the day after (adherent cells) treated for 24 hours with compound or DMSO control. Cells were harvested by trypsinization, washed once in PBS, and resuspended in 300 μL PBS. The suspension (100 μL) was mixed with 500 μL 1.2% low melting point agarose at 37°C and the mixture were added to agarose coated slides and a coverslip was added on top. The slides were solidified on ice and lysed O/N at 4°C in Lysis buffer (2.5 M NaCl, 100 mmol/L EDTA, 10 mmol/L Tris, 10% DMSO, 1% Triton X100). Samples were washed three times in enzyme buffer (40 mmol/L HEPES, 0.1 M KCl, 0.5 mmol/L EDTA, 0.2 mg/mL BSA, pH 8.0) and treated with hOGG1 enzyme (2 μg/mL) or buffer alone for 45 minutes at 37°C. hOGG1 was expressed in Escherichia coli and purified as described previously (6). Slides were transferred to alkaline electrophoresis buffer (300 mmol/L NaOH, 10 mmol/L EDTA) for 30 minutes and electrophoresis was performed at 25 V, 300 mA for 30 minutes at 4°C. Samples were washed in 400 mmol/L Tris pH 7.5 for 45 minutes. DNA was stained with SybrGOLD and comets were quantified with Comet Assay IV software in live video mode.

Time-lapse microscopy

Cells were seeded in 96-well plates (BD Falcon plate 353376; 5,000 cells/well). The day after, cells were treated and the time-lapse was initiated after 30 minutes and images of GFP and brightfield channels were acquired every 10 minutes for 24 hours in a Perkin-Elmer ImageXpress instrument. Cells were kept at 37°C in 5% CO2 atmosphere during the entire time-lapse experiment. Movie files were assembled in MetaXpress and ImageJ softwares. For each condition, images were acquired from two different wells. Individual cells were followed manually and scored for defects in mitosis including mitotic slippage/polynucleation (MS/PN), micronuclei formation (G1/MN), mitotic slippage (MS), or cell death during mitosis (DiM). The time in mitosis was defined as the time from the cell rounding up and condensing the chromatin to the end of cytokinesis.

Caspase activation assay

Pol δ replacement cells were cultured in the presence or absence 10 ng/mL doxycycline for 3 days and subsequently reseeded into a 96-well plate (3,000 cells per well). Following overnight incubation, cells were treated with TH588 for 24 hours and caspase activity was subsequently measured using the Caspase-GloR 3/7 Assay Kit (Promega). Kit protocol was followed and after 1 hour of incubation with reagent, luminescence was measured using a Hidex Sense plate reader.

Statistical analysis

Statistical analysis was performed using Prism 8 (GraphPad software).

An experimental system to study 8-oxodGTP incorporation into DNA

Our original model on the role of MTH1 in protecting cancer cells centered upon the misincorporation of oxidized nucleotides, such as 8-oxodGTP, into DNA as the toxic lesion (6). In accordance with this model, we have also demonstrated that 8-oxodGTP is cytotoxic to zebrafish embryos in the absence of MTH1 protein or in the presence of TH588 (17), and that TH588 toxicity in cultured cell lines relies upon ROS (18). More recently, others and ourselves have demonstrated direct action of TH588 on tubulin polymerization (10, 13, 19, 20) and that MTH1 itself has a novel role in mitosis, rendering a more complex mechanism of action of TH588 and MTH1 as an anticancer target (13). With this background it is important to understand if 8-oxodG in DNA is a potentially lethal lesion and how this is incorporated into the genome. To re-examine this model, we employed a protein-replacement system (16), which allows the doxycycline-inducible replacement of endogenous Pol δ, one of the main DNA replicases in cells, with either a wild-type (WT), error-prone (EP, L606G), proofreading-deficient (PD, D402A), or EP and PD double mutant (DM) variant in the ovarian cancer cell line A2780 (Fig. 1A). We were particularly interested in the EP variant as we reasoned this would increase misincorporation of oxidized nucleotides into DNA and thus provide a controlled experimental system to study the consequences of TH588 treatment.

Figure 1.

An experimental system to study 8-oxodGTP incorporation into DNA. A, Schematic detailing Pol δ (p125) replacement system. B, Pol δ replacement cells (WT, DM, EP, and PD double mutant) were cultured in the presence or absence 10 ng/mL doxycycline (Dox) for 4 days and lysates prepared for Western blot analysis with the indicated antibodies. C, Pol δ replacement cells were cultured in 10 ng/mL doxycycline for 7 days and viable cells measured using the resazurin assay. Values were normalized to Pol δ replacement cells cultured in the absence of doxycycline. Averages of three experiments are shown, each performed in sextuplet. Error bars, SD. D, Pol δ replacement cells were cultured in 10 ng/mL doxycycline for 4 days and were labeled with 10 μmol/L EdU for 20 minutes prior to fixation. S-phase cells were defined at >9 EdU foci per nucleus; at least 2,000 cells were analyzed per well, with experiments performed at least with duplicate wells. Average of at least three independent experiments is plotted. Error bars, SD. E and F, Pol δ replacement cells were cultured in 10 ng/mL doxycycline for 4 days before treatment with 1 μmol/L TH588 for 24 hours. Genomic levels of 8-oxo-dG were analyzed using the modified comet assay. Representative images are shown in E. Scale bar, 50 μm. Fold difference in tail moment comparing OGG1-treated to buffer only control is shown in F. At least 200 cells were analyzed in two independent experiments. Horizontal bar, median. Statistical comparison performed using Mann–Whitney tests. ****, P < 0.0001; ns, not significant.

Figure 1.

An experimental system to study 8-oxodGTP incorporation into DNA. A, Schematic detailing Pol δ (p125) replacement system. B, Pol δ replacement cells (WT, DM, EP, and PD double mutant) were cultured in the presence or absence 10 ng/mL doxycycline (Dox) for 4 days and lysates prepared for Western blot analysis with the indicated antibodies. C, Pol δ replacement cells were cultured in 10 ng/mL doxycycline for 7 days and viable cells measured using the resazurin assay. Values were normalized to Pol δ replacement cells cultured in the absence of doxycycline. Averages of three experiments are shown, each performed in sextuplet. Error bars, SD. D, Pol δ replacement cells were cultured in 10 ng/mL doxycycline for 4 days and were labeled with 10 μmol/L EdU for 20 minutes prior to fixation. S-phase cells were defined at >9 EdU foci per nucleus; at least 2,000 cells were analyzed per well, with experiments performed at least with duplicate wells. Average of at least three independent experiments is plotted. Error bars, SD. E and F, Pol δ replacement cells were cultured in 10 ng/mL doxycycline for 4 days before treatment with 1 μmol/L TH588 for 24 hours. Genomic levels of 8-oxo-dG were analyzed using the modified comet assay. Representative images are shown in E. Scale bar, 50 μm. Fold difference in tail moment comparing OGG1-treated to buffer only control is shown in F. At least 200 cells were analyzed in two independent experiments. Horizontal bar, median. Statistical comparison performed using Mann–Whitney tests. ****, P < 0.0001; ns, not significant.

Close modal

To confirm replacement of endogenous Pol δ with each variant, A2780 Pol δ replacement cells were cultured in medium supplemented with doxycycline for 4 days to allow shRNA-mediated depletion of endogenous Pol δ p125 and the simultaneous expression of shRNA-resistant FLAG-tagged Pol δ p125 variants (Fig. 1A), as described previously (16), and cell lysates were analyzed by Western blot analysis. In each case, following doxycycline-induction, successful depletion of endogenous Pol δ p125 was observed whilst expression of the FLAG-tagged variants was clearly visible (Fig. 1B). Importantly, within the timeline of the experiments performed, replacement of endogenous Pol δ with the Pol δ variants did not significantly reduce cell viability (Fig. 1C) and had a minimal effect on the proportion of S-phase cells (Fig. 1D), with the exception of the DM variant, which had a small but significant increase in the proportion of S-phase cells, as reported previously (16).

To determine whether replacement of Pol δ with the EP variant and subsequent TH588 treatment leads to increased genomic 8-oxodG, we employed the modified comet assay in which cells embedded in agarose are exposed to recombinant OGG1 prior to electrophoresis under alkaline conditions. The presence of 8-oxodG in DNA would lead to increased OGG1-induced comet tails in the assay. Consistent with our previous reports (6, 12), TH588 treatment, irrespective of Pol δ status, resulted in increased 8-oxodG in DNA as compared with DMSO-treated controls (Fig. 1E and F). The highest accumulation of genomic 8-oxodG was observed in the TH588-treated cells expressing the Pol δ EP variant (Fig. 1E and F), indicating that this reduced-fidelity variant of this replicase misincorporates 8-oxodG into DNA. Thus, we have an experimental system with which to study the cellular consequences of increased misincorporation of 8-oxodGTP into DNA, on top of the baseline 8-oxodGTP misincorporation by other DNA polymerases.

Increased genomic 8-oxodG in TH588-treated Pol δ EP cells correlates with increased toxicity

The cytotoxic effects of TH588 have been proposed to be independent of targeting MTH1 and independent of 8-oxodGTP (9, 10, 21), whereas we have shown TH588 cytotoxicity is related to ROS and partly rescued by expression of the bacterial 8-oxodGTPase MutT (6, 12, 17). Here, we wanted to use this experimental system to determine the contribution of 8-oxodGTP incorporation to the cytotoxicity of TH588. In accordance with the original proposed mechanism of action of MTH1 inhibition, replacement of endogenous Pol δ with the EP but not WT variant increased TH588-induced cytotoxicity (Fig. 2A). Elevated sensitivity was also observed in the cells expressing the Pol δ DM but not PD variant (Fig. 2A), indicating the increased sensitivity was due to the point mutation reducing Pol δ fidelity, that is, the EP variant. In line with this, elevated DNA damage (γH2Ax) and apoptotic (cleaved-PARP) signaling was observed in the EP and DM variant expressing cells compared with WT control (Fig. 2B and C). Furthermore, using a quantitative luminescence-based assay to measure caspase activity, we could observe elevated apoptosis in Pol δ EP variant cells compared with WT following TH588 treatment (Fig. 2D). Taken together, these data indicate that incorporation of TH588-induced 8-oxodGTP contributes to cell killing.

Figure 2.

An error-prone variant of Pol δ sensitizes cells to TH588. A, Pol δ replacement cells (WT, DM, EP, and PD double mutant) were cultured in the presence or absence 10 ng/mL doxycycline (Dox) for 3 days prior to reseeding, and 24 hours later, treatment with a dose–response of TH588 in media with or without doxycycline. Viable cells were measured following a 3-day treatment. A representative of four independent experiments is shown, performed in sextuplet, with error bars indicating SD. B and C, Pol δ replacement cells were cultured with doxycycline for 4 days and then treated with 1 μmol/L TH588 for the indicated times before collection for Western blot analysis with the indicated antibodies. Representative blot of two independent experiments is shown in B. Quantification of band intensity for cleaved-PARP, relative to β-actin loading control, is plotted in C. Individual values are shown. Error bars, SD from two independent experiments. D, Pol δ replacement cells were cultured with doxycycline for 4 days and then treated with 2 μmol/L TH588 for 24 hours before measurement of caspase activity. Individual values are shown. Error bars, SEM from two independent experiments, each performed in sextuplet. Statistical comparisons performed with a two-way ANOVA. *, P ≤ 0.05; **, P ≤ 0.01; ****, P < 0.0001.

Figure 2.

An error-prone variant of Pol δ sensitizes cells to TH588. A, Pol δ replacement cells (WT, DM, EP, and PD double mutant) were cultured in the presence or absence 10 ng/mL doxycycline (Dox) for 3 days prior to reseeding, and 24 hours later, treatment with a dose–response of TH588 in media with or without doxycycline. Viable cells were measured following a 3-day treatment. A representative of four independent experiments is shown, performed in sextuplet, with error bars indicating SD. B and C, Pol δ replacement cells were cultured with doxycycline for 4 days and then treated with 1 μmol/L TH588 for the indicated times before collection for Western blot analysis with the indicated antibodies. Representative blot of two independent experiments is shown in B. Quantification of band intensity for cleaved-PARP, relative to β-actin loading control, is plotted in C. Individual values are shown. Error bars, SD from two independent experiments. D, Pol δ replacement cells were cultured with doxycycline for 4 days and then treated with 2 μmol/L TH588 for 24 hours before measurement of caspase activity. Individual values are shown. Error bars, SEM from two independent experiments, each performed in sextuplet. Statistical comparisons performed with a two-way ANOVA. *, P ≤ 0.05; **, P ≤ 0.01; ****, P < 0.0001.

Close modal

8-oxodGTP incorporation does not greatly perturb S-phase DNA synthesis

It is easy to envision that misincorporation of 8-oxodGTP would generate replication stress, as MTH1 overexpression reduces Ras-induced DNA lesions and counteracts oncogenic stress (8, 22), and replication stress is associated with oxidative stress in cancer (23) and replication difficulties (24, 25). Therefore, we wanted to investigate if MTH1 inhibition and resulting incorporation of oxidized nucleotides would result in replication stress. To test this we used the well-established DNA fiber assay in cells expressing either the Pol δ WT or EP variant treated with TH588 for 0 to 6 hours prior to the assay. Two hours of TH588 treatment resulted in a small, but significant, decrease in replication fork speeds regardless of whether cells were expressing the WT or EP variant of Pol δ, and this decreased fork speed recovered to DMSO-treated control levels following 6 hours TH588 treatment (Fig. 3A and B). Notably, expression of the Pol δ EP variant did not significantly alter this TH588-induced replication defect compared with WT controls (Fig. 3A and B). In a parallel set of experiments, the replication stress response was also monitored in osteosarcoma U2OS cells treated with TH588, which we have previously reported are sensitive to TH588 treatment and accumulate genomic 8-oxodG (6). Similarly, an initial small, but significant, decrease in replication fork speeds were observed following TH588 treatment, but fork speeds quickly recovered to that of the control following longer treatment times (Supplementary Figs. S1A and S1B). Consistent with no great perturbation of ongoing DNA synthesis, no activation of the intra-S-phase checkpoint kinase Chk1 was observed following TH588 treatment, despite the induction of DNA damage signaling at later timepoints (Supplementary Fig. S1C).

Figure 3.

Incorporation of TH588-induced 8-oxodGTP does not cause sustained replication stress in S phase. A and B, Pol δ replacement cells were cultured with doxycycline (Dox) for 4 days before treatment with 1 μmol/L TH588 for the indicated time and sequential pulse labeling with CldU and IdU prior to preparation and analysis of spread DNA fibers. Representative DNA fiber images are shown in A. Measured fork speeds from two independent experiments are shown in B, with at least 200 fibers measured in total. Horizontal bars, median. Statistical testing was performed using Mann–Whitney tests. ns, not significant; **, P ≤ 0.01. C and D, Alkaline DNA unwinding assay (schematic shown in C) with Pol δ replacement WT or EP cells treated with 10 ng/mL doxycycline for 4 days prior to the indicated treatments. Average from two independent experiments is shown in D. Mean values are plotted. Error bars, SD. Caffeine was used as the postreplication repair of gaps, which are caffeine dependent, hence the detection of gaps can be increased with caffeine.

Figure 3.

Incorporation of TH588-induced 8-oxodGTP does not cause sustained replication stress in S phase. A and B, Pol δ replacement cells were cultured with doxycycline (Dox) for 4 days before treatment with 1 μmol/L TH588 for the indicated time and sequential pulse labeling with CldU and IdU prior to preparation and analysis of spread DNA fibers. Representative DNA fiber images are shown in A. Measured fork speeds from two independent experiments are shown in B, with at least 200 fibers measured in total. Horizontal bars, median. Statistical testing was performed using Mann–Whitney tests. ns, not significant; **, P ≤ 0.01. C and D, Alkaline DNA unwinding assay (schematic shown in C) with Pol δ replacement WT or EP cells treated with 10 ng/mL doxycycline for 4 days prior to the indicated treatments. Average from two independent experiments is shown in D. Mean values are plotted. Error bars, SD. Caffeine was used as the postreplication repair of gaps, which are caffeine dependent, hence the detection of gaps can be increased with caffeine.

Close modal

Upon DNA damage, replication may proceed downstream of a stalled fork by origin-independent re-priming, for example, by PrimPol (26, 27), leaving a gap on the nascent DNA strand that cannot be detected by the DNA fiber assay (28). To test if gaps are generated along the replicating nascent DNA strand, we made use of an in house assay that detects such gaps (Fig. 3C; ref. 29). Briefly, we 3H-thymidine pulse-labelled ongoing forks, and then let these progress from the labeled area for different times. By addition of alkaline solution, DNA unwinding takes place from the single-stranded ends of the fork or gaps, if present. Hence, as the fork moves forward or the gaps are filled, the labeling is removed from the fraction of DNA that becomes single-stranded. We isolated the DNA, ultra-sonicated the sample and separated double-stranded (ds) and single-stranded (ss) DNA on hydroxyl apatite columns and determined the ds/ssDNA ratio to determine the progression of replication forks, as described previously (29). Using this assay we demonstrate that no gaps are generated behind the fork during replication following TH588 treatment, regardless of Pol δ status (Fig. 3D). Taken together, these data indicate that despite incorporation of oxidized nucleotides, and the ultimate cytotoxicity, no substantial, sustained perturbation of ongoing DNA synthesis in S-phase cells is observed following TH588 treatment.

Treatment with TH588 triggers mitotic DNA replication

Emerging data show TH588 and MTH1 siRNA arrest cells in mitosis through interference of tubulin dynamics, and in the case of TH588, primarily due to direct inhibition of tubulin polymerization (10, 13). The explanation for MTH1i or MTH1 siRNA to fail to be toxic in cancer cell lines could be explained by lack of ROS, which is supported by nontoxic MTH1i becoming toxic in the presence of ROS and also that MTH1 knockout zebrafish embryos are sensitive to microinjection of 8-oxodGTP (13). It has been reported that ROS and oxidative stress are elevated during mitosis and that mitotic arrest itself generates ROS (30), with one report linking this to mitophagy (31), and hence MTH1 may have a role in hydrolyzing oxidized purine nucleotides during mitosis. Normally, DNA replication is completed during the S-phase of the cell cycle. However, under conditions of replication stress, such as that generated experimentally by low-dose aphidicolin treatment, it has been demonstrated that under-replicated regions are repaired by Pol δ–mediated mitotic DNA replication at common fragile sites (32). Because replication stress is often found in cancer (24), this may explain how 8-oxodGTP could be specifically incorporated in cancer cells following MTH1 siRNA or TH588 treatment, as reported previously. To test if TH588 would induce mitotic DNA replication, we synchronized U2OS cells in G2 using the Cdk1 inhibitor RO-3306 together with 0.4 μmol/L aphidicolin as a positive control or TH588. Cells were released into mitosis for 30 minutes by washout of the compounds and EdU was added to the media to label newly synthesized DNA. As expected, aphidicolin induced EdU foci in mitotic cells and so did TH588 (Fig. 4A and B), suggesting that mitotic replication is stimulated with this compound.

Figure 4.

TH588 triggers mitotic replication. A and B, U2OS cells were treated with Cdk1 inhibitor RO3306 together with 0.4 μmol/L aphidicolin (APH) or 5 μmol/L TH588 for 16 hours. Compounds were washed away and cells were released for 30 minutes in the presence of 10 μmol/L EdU. EdU was labeled with Alexa Fluor-488-azide, mitotic cells with an H3-pS10 antibody, and DNA with DAPI. Representative images are shown in A. Scale bar, 10 μm. Quantification of cells containing mitotic EdU foci is shown in B. Data shown as mean of >160 cells analyzed per condition from two independent experiments. Error bars, SD. Statistical test performed using Student t test. *, P < 0.05.

Figure 4.

TH588 triggers mitotic replication. A and B, U2OS cells were treated with Cdk1 inhibitor RO3306 together with 0.4 μmol/L aphidicolin (APH) or 5 μmol/L TH588 for 16 hours. Compounds were washed away and cells were released for 30 minutes in the presence of 10 μmol/L EdU. EdU was labeled with Alexa Fluor-488-azide, mitotic cells with an H3-pS10 antibody, and DNA with DAPI. Representative images are shown in A. Scale bar, 10 μm. Quantification of cells containing mitotic EdU foci is shown in B. Data shown as mean of >160 cells analyzed per condition from two independent experiments. Error bars, SD. Statistical test performed using Student t test. *, P < 0.05.

Close modal

Spindle assembly checkpoint-mediated mitotic arrest is required for 8-oxodG accumulation and toxicity induced by TH588

Because replication in S phase is not greatly perturbed by TH588 treatment and instead this agent triggers mitotic DNA replication, we hypothesized that incorporation of 8-oxodGTP caused by TH588 into DNA may be dependent upon mitotic arrest, to generate ROS and provide time for mitotic DNA synthesis. To test this hypothesis, we used an inhibitor to the mitotic kinase mps1 (reversine) that prevents the spindle assembly checkpoint (SAC) activation (33) in combination with TH588 treatment in U2OS cells. TH588 treatment induces a mitotic arrest (Fig. 5A), and here we show that this arrest is dependent upon a functional SAC as reversin cotreatment ablates this mitotic arrest (Fig. 5A). In accordance with our hypothesis, reversin treatment also prevents accumulation of TH588-induced genomic 8-oxodG and cell death (Fig. 5B and C). Thus, these data demonstrate that SAC-dependent mitotic arrest is required for accumulation of genomic 8-oxodG and cell death following treatment with TH588.

Figure 5.

Mitotic arrest is required for accumulation of genomic 8-oxodG following treatment with TH588. A, Quantification of the time in mitosis from the onset of prophase until the end of cytokinesis in U2OS-H2B-GFP cells treated with 10 μmol/L TH588 and 0.5 μmol/L reversine (REV) for 24 hours. Individual cells are shown from a representative experiment. Horizontal line, mean. B, Relative tail moment in the modified comet assay of U2OS cells treated with 10 μmol/L TH588 and 0.5 μmol/L reversine for 24 hours. Data from two independent experiments shown with 400 cells analyzed in total. Horizontal line, median. Statistical testing was performed using the Mann–Whitney test. ****, P < 0.001. C, Viability assay of U2OS cells treated with TH588 with or without 0.5 μmol/L reversine for 72 hours. Mean of three independent experiments is shown. Error bars, SD.

Figure 5.

Mitotic arrest is required for accumulation of genomic 8-oxodG following treatment with TH588. A, Quantification of the time in mitosis from the onset of prophase until the end of cytokinesis in U2OS-H2B-GFP cells treated with 10 μmol/L TH588 and 0.5 μmol/L reversine (REV) for 24 hours. Individual cells are shown from a representative experiment. Horizontal line, mean. B, Relative tail moment in the modified comet assay of U2OS cells treated with 10 μmol/L TH588 and 0.5 μmol/L reversine for 24 hours. Data from two independent experiments shown with 400 cells analyzed in total. Horizontal line, median. Statistical testing was performed using the Mann–Whitney test. ****, P < 0.001. C, Viability assay of U2OS cells treated with TH588 with or without 0.5 μmol/L reversine for 72 hours. Mean of three independent experiments is shown. Error bars, SD.

Close modal

Delayed mitotic progression in Pol δ EP cells treated with TH588, but not by antitubulin poisons

We hypothesized that additional misincorporation of oxidized nucleotides prior to, or during, mitotic entry could contribute to the mitosis-dependent cell death mechanism we observed. We first sought further evidence that oxidized nucleotide incorporation can occur during mitosis. We arrested Pol δ WT and EP cells in G2 phase using the CDK1 inhibitor RO336 and then released these cells into mitosis in the presence of TH588. To ensure these cells did not enter the subsequent S-phase, 5 hours post release, high concentrations of thymidine were added to cell media. Twenty-four hours after release, as TH588-treated cells spend longer in mitosis, cells were harvested and genomic oxidized nucleotides assessed using the modified comet assay. Under these conditions, in which cells have progressed through mitosis but have not begun the subsequent S-phase, we observed elevated genomic 8-oxodG in the Pol δ EP cells (Fig. 6A), which could be consistent with incorporation of oxidized nucleotides during mitosis.

Figure 6.

Expression of an error-prone variant of Pol δ results in a prolonged mitotic arrest and elevated mitotic cell death following TH588 treatment. A, Pol δ replacement cells were cultured with doxycycline for 4 days before being arrested in G2 using RO3366 and were subsequently released into mitosis in the presence of TH588 (2.5 μmol/L). Five hours post-release thymidine (2 mmol/L) was added to the media to prevent S-phase entry. Twenty-four hours after release, cells were collected and genomic oxidized nucleotides assessed using the modified comet assay. Individual values from two independent experiments are shown. Horizontal bar, median. Statistical testing performed with a Kruskal–Wallis test. ****, P > 0.001. B, Quantification of the time in mitosis from onset of prophase until end of cytokinesis. Data are from two experiments. Fifty cells were analyzed in total. Horizontal line, mean. Statistical testing performed with unpaired t test. ****, P > 0.001. C and D, Pol δ replacement-H2B-GFP cells (WT; EP) were treated with 10 ng/mL doxycycline for 4 days before treatment with 2.5 μmol/L TH588 and subsequent initiation of the time-lapse microscopy after 30 minutes. Images were acquired every 10 minutes for 24 hours. Individual cells were followed manually and scored for defects in mitosis including MS/PN, G1/MN, MS, or DiM. Representative of two independent experiments is shown in C. Representative images are shown in D. Scale bar, 10 μm.

Figure 6.

Expression of an error-prone variant of Pol δ results in a prolonged mitotic arrest and elevated mitotic cell death following TH588 treatment. A, Pol δ replacement cells were cultured with doxycycline for 4 days before being arrested in G2 using RO3366 and were subsequently released into mitosis in the presence of TH588 (2.5 μmol/L). Five hours post-release thymidine (2 mmol/L) was added to the media to prevent S-phase entry. Twenty-four hours after release, cells were collected and genomic oxidized nucleotides assessed using the modified comet assay. Individual values from two independent experiments are shown. Horizontal bar, median. Statistical testing performed with a Kruskal–Wallis test. ****, P > 0.001. B, Quantification of the time in mitosis from onset of prophase until end of cytokinesis. Data are from two experiments. Fifty cells were analyzed in total. Horizontal line, mean. Statistical testing performed with unpaired t test. ****, P > 0.001. C and D, Pol δ replacement-H2B-GFP cells (WT; EP) were treated with 10 ng/mL doxycycline for 4 days before treatment with 2.5 μmol/L TH588 and subsequent initiation of the time-lapse microscopy after 30 minutes. Images were acquired every 10 minutes for 24 hours. Individual cells were followed manually and scored for defects in mitosis including MS/PN, G1/MN, MS, or DiM. Representative of two independent experiments is shown in C. Representative images are shown in D. Scale bar, 10 μm.

Close modal

To test whether the elevated genomic oxidized nucleotide observed in the Pol δ EP cells contributes to the mitosis-dependent cell death mechanism, we generated H2B-GFP expressing Pol δ replacement cells, and following TH588 treatment, monitored mitotic progression using live cell microscopy. In doxycycline-induced Pol δ WT cells, low doses of TH588 induced a mitotic arrest, but the majority of cells could proceed into anaphase without apparent problems (Fig. 6BD, b; Supplementary Movie S1). In contrast, the Pol δ EP expressing cells had a prolonged mitotic arrest in the presence of TH588 and underwent mitotic catastrophe, either becoming polynucleated or undergoing mitotic cell death (Fig. 6BD; Supplementary Movie S2). These data suggest that oxidized dNTPs incorporated during mitotic replication can result in mitotic catastrophe.

To offer an alternative explanation, the mitotic defects induced by TH588 in Pol δ EP but not in WT expressing cells could potentially be a consequence of direct induction of ROS by the compounds, and Pol δ EP cells being able to incorporate 8-oxodGTP into DNA. Hence, the effect may be independent of MTH1 inhibition by TH588. To test this, we triggered mitotic arrest and ROS using the tubulin poison paclitaxel and followed mitosis using live cell imaging. Interestingly, we found no difference in mitotic progression between Pol δ EP and WT expressing cells (Fig. 7A and B), showing that a mitotic delay by itself without concomitant inhibition of MTH1 is insufficient to cause a differential response in Pol δ EP and WT expressing cells. To further support that MTH1 inhibition is important for the observed effect, Pol δ EP and WT cells were treated with other mitosis-targeted compounds, vincristine, or an inhibitor to CENP-E, which resulted in no difference in mitotic progression between the two cell types (Fig. 7C). Taken together, these data indicate that cancer cell death caused by TH588 is a combinatorial effect due to perturbation of microtubule dynamics and incorporation of oxidized nucleotides into the genome. Furthermore, misincorporation of oxidized dNTPs may alter mitotic progression.

Figure 7.

Expression of an error-prone variant of Pol δ does not result in prolonged mitotic arrest and elevated mitotic cell death induced by mitotic poisons. A, Quantification of the time in mitosis from the onset of prophase until the end of cytokinesis followed treatment with 2 nmol/L paclitaxel (PTX). Individual cells are shown from a representative experiment. Horizontal line, mean. Statistical testing performed using unpaired t test. B, Time-lapse imaging of Pol δ replacement H2B-GFP cells induced with 10 ng/mL doxycycline (Dox) for 4 days prior to treatment with 2 nmol/L paclitaxel. Images were acquired every 10 minutes for 24 hours. Individual cells were followed manually and scored for defects in mitosis including MS/PN, G1/MN, MS, or DiM. Representative experiment is shown. C, Quantification of the time in mitosis of Pol δ replacement H2B-GFP cells induced with 10 ng/mL doxycycline for 4 days prior to treatment with 1 nmol/L vincristine or 20 nmol/L CENP-Ei. Individual cells are shown from a representative experiment. Horizontal line, mean. Statistical testing performed using unpaired t test. D, Mechanism of action of TH588. We propose MTH1 inhibition during S phase does not cause replication stress and is related to overall low ROS levels also in cancer cells. Prolonged mitotic arrest causes mitophagy and ROS (31), likely above physiological levels (dotted line). We propose TH588 is causing ROS because of mitotic arrest mediated by directly interfering with tubulin polymerization or breaking MTH1-tubulin interactions. Through MTH1 inhibition, TH588 prevents 8-oxodGTP sanitization, resulting in 8-oxodG incorporation into DNA during mitotic replication. We speculate cancer-specific toxicity of TH588 is related to replication stress associated in cancer, likely triggering mitotic replication, that is, repair synthesis of under replicated regions. ns, not significant.

Figure 7.

Expression of an error-prone variant of Pol δ does not result in prolonged mitotic arrest and elevated mitotic cell death induced by mitotic poisons. A, Quantification of the time in mitosis from the onset of prophase until the end of cytokinesis followed treatment with 2 nmol/L paclitaxel (PTX). Individual cells are shown from a representative experiment. Horizontal line, mean. Statistical testing performed using unpaired t test. B, Time-lapse imaging of Pol δ replacement H2B-GFP cells induced with 10 ng/mL doxycycline (Dox) for 4 days prior to treatment with 2 nmol/L paclitaxel. Images were acquired every 10 minutes for 24 hours. Individual cells were followed manually and scored for defects in mitosis including MS/PN, G1/MN, MS, or DiM. Representative experiment is shown. C, Quantification of the time in mitosis of Pol δ replacement H2B-GFP cells induced with 10 ng/mL doxycycline for 4 days prior to treatment with 1 nmol/L vincristine or 20 nmol/L CENP-Ei. Individual cells are shown from a representative experiment. Horizontal line, mean. Statistical testing performed using unpaired t test. D, Mechanism of action of TH588. We propose MTH1 inhibition during S phase does not cause replication stress and is related to overall low ROS levels also in cancer cells. Prolonged mitotic arrest causes mitophagy and ROS (31), likely above physiological levels (dotted line). We propose TH588 is causing ROS because of mitotic arrest mediated by directly interfering with tubulin polymerization or breaking MTH1-tubulin interactions. Through MTH1 inhibition, TH588 prevents 8-oxodGTP sanitization, resulting in 8-oxodG incorporation into DNA during mitotic replication. We speculate cancer-specific toxicity of TH588 is related to replication stress associated in cancer, likely triggering mitotic replication, that is, repair synthesis of under replicated regions. ns, not significant.

Close modal

Here, we aimed to explore if and how oxidized dNTPs, incorporated into DNA, are toxic to cells. We used a previously established system based upon doxycycline-inducible replacement of endogenous Pol δ with an error-prone (L606G) variant (16), which we confirm did not greatly alter cell-cycle progression or overall viability of cells. We demonstrate increased 8-oxodGTP incorporation in these cells, providing an experimental system by which to study elevated incorporation of oxidative nucleotides into DNA and the resultant cellular consequences.

Because Pol δ is a main DNA polymerase one would expect a huge increase in 8-oxodG DNA levels and detrimental effects on replication in Pol δ EP cells exposed to MTH1i TH588, which we show is not the case. Because we observe minor alterations, and no generation of post-replication gaps, we suggest that the overall level of 8-oxodGTP is very low in the S phase of the cell cycle making MTH1 function redundant during this phase in unstressed conditions. However, it has recently been shown that mild replication stress is sufficient to induce chromosome mis-segregation through premature centriole disengagement (34).

ROS is known to accumulate throughout the cell cycle, peaking in mitosis, and accordingly is induced following extended mitotic arrest (30, 31). TH588 causes arrest in mitosis owing to a direct effect on tubulin (10, 13, 19, 20) and likely also through disrupting the binding between MTH1 and tubulin in cells, resulting in reduced microtubule polymerization rates (13). The discussion of the role of MTH1 on tubulin and in mitosis and the role of TH588 binding to tubulin is detailed elsewhere (13) and not the topic of this report. Nonetheless, TH588-induced arrest in mitosis is associated with an increase in ROS, likely resulting in increased 8-oxodGTP levels in mitosis. Here, we report an increase in mitotic DNA replication following treatment with TH588 and also that the incorporation of 8-oxodGTP into DNA is dependent on an extended mitotic arrest, as pharmacologically reducing the time in mitosis reduced the incorporation of 8-oxodGTP as well as toxicity of TH588 (Fig. 5). Here, we find only a slight increase in toxicity of TH588 in Pol δ EP cells as compared with those expressing Pol δ WT (Fig. 2), which suggests that the increase in genomic 8-oxodG is only explaining a marginal part of the toxic effects of TH588. However, it should be emphasized that the lower fidelity DNA polymerases Pol κ and Pol β are abundant in cells and have been implicated in 8-oxodGTP incorporation into DNA (35, 36) and that 8-oxoG is also incorporated in the Pol δ WT cells (Fig. 1E and F). Hence, the slight difference in cytotoxicity is explained by other polymerases being responsible for 8-oxoGTP incorporation also into the control cells.

A highly surprising finding is that Pol δ EP cells show a more pronounced delay in mitosis as compared with Pol δ WT cells, suggesting incorporation of 8-oxodGTP into DNA is increasing the mitotic defect (Fig. 6). Understanding how this work mechanistically will be challenging and should be addressed in future studies. An alternative explanation is that Pol δ EP cells would be prone to generate severe mitotic arrest owing to MTH1 and 8-oxodGTP independent effects. We argue this is not the case, as this mitotic arrest is not a phenomena observed generally following treatments with microtubule poisons (Fig. 7).

Here, we propose a suitable model on the mechanism of action of TH588 (Fig. 7D). TH588 does not lead to substantial obstruction of replication forks in S-phase due to low levels of ROS, which is supported by a recent study demonstrating elevated ROS levels in cancer cells primarily during G2 and particularly mitosis (30), which may also explain why other MTH1i that do not arrest cells in mitosis are nontoxic. ROS are accumulating upon mitotic arrest by TH588, mediated through a direct effect on tubulin and also likely by disrupting MTH1 mitotic function and preventing the tubulin-MTH1 interaction (13). The nucleotide pool is oxidized in the presence of ROS in mitosis generating 8-oxodGTP, which is a substrate for MTH1 that prevents mitotic DNA damage. Upon inhibition of MTH1 8-oxodGTPase activity by TH588, 8-oxodGTP is incorporated into DNA during mitotic replication in cancer cells and enhances the mitotic arrest by a mechanism yet to be established.

S.G. Rudd, H. Gad, U. Warpman Berglund, and T. Helleday all have ownership interest (including patents) in Oxcia AB. H. Gad, U. Warpman Berglund, and T. Helleday are unpaid consultants and have an advisory board relationship at Helleday Foundation. U. Warpman Berglund is a chairman of the board at Oxcia AB. No potential conflicts of interest were disclosed by the other authors.

Conception and design: S.G. Rudd, H. Gad, C.E. Ström, U. Warpman Berglund, T. Helleday

Development of methodology: S.G. Rudd, K. Sanjiv, N. Amaral, C.E. Ström, T. Helleday

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S.G. Rudd, H. Gad, N. Amaral, A. Hagenkort, P. Groth

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S.G. Rudd, H. Gad, N. Amaral, A. Hagenkort, P. Groth, O. Mortusewicz, T. Helleday

Writing, review, and/or revision of the manuscript: S.G. Rudd, H. Gad, N. Amaral A. Hagenkort, O. Mortusewicz, U. Warpman Berglund, T. Helleday

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): T. Helleday

Study supervision: S.G. Rudd, U. Warpman Berglund, T. Helleday

We thank Prof. Jircny (University of Zurich) for sharing Pol δ-replacement cell lines. Financial support was given by EMBO Long-Term Fellowships (ALTF-2014-605 to S.G. Rudd; ALTF-2017-196 to N. Amaral), Swedish Foundation for Strategic Research (RB13-0224 to U. Warpman Berglund), Swedish Research Council (2016-2025 to T. Helleday), Swedish Cancer Society (19-0056-JIA to S.G. Rudd; CAN2015/225 to T. Helleday), Swedish Children's Cancer Foundation (TJ2017-0021 to S.G. Rudd; PR2014-0048 to T. Helleday), the Swedish Pain Relief Foundation (T. Helleday), and The European Research Council (ERC-AdG-695376, TAROX to T. Helleday).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Jackson
SP
,
Bartek
J
. 
The DNA-damage response in human biology and disease
.
Nature
2009
;
461
:
1071
78
.
2.
Zhang
Y
,
Du
Y
,
Le
W
,
Wang
K
,
Kieffer
N
,
Zhang
J
. 
Redox control of the survival of healthy and diseased cells
.
Antioxid Redox Signal
2011
;
15
:
2867
908
.
3.
Maki
H
,
Sekiguchi
M
. 
MutT protein specifically hydrolyses a potent mutagenic substrate for DNA synthesis
.
Nature
1992
;
355
;
273
5
.
4.
Rudd
SG
,
Valerie
NCK
,
Helleday
T
. 
Pathways controlling dNTP pools to maintain genome stability
.
DNA Repair (Amst)
2016
;
44
:
193
204
.
5.
Huber
KV
,
Salah
E
,
Radic
B
,
Gridling
M
,
Elkins
JM
,
Stukalov
A
, et al
Stereospecific targeting of MTH1 by (S)-crizotinib as an anticancer strategy
.
Nature
2014
;
508
;
222
227
.
6.
Gad
H
,
Koolmeister
T
,
Jemth
AS
,
Eshtad
S
,
Jacques
SA
,
Strom
CE
, et al
MTH1 inhibition eradicates cancer by preventing sanitation of the dNTP pool
.
Nature
2014
;
508
;
215
21
.
7.
Patel
A
,
Burton
DG
,
Halvorsen
K
,
Balkan
W
,
Reiner
T
,
Perez-Stable
C
, et al
MutT Homolog 1 (MTH1) maintains multiple KRAS-driven pro-malignant pathways
.
Oncogene
2015
;
34
:
2586
96
.
8.
Burton
DG
,
Rai
P
. 
MTH1 counteracts oncogenic oxidative stress
.
Oncoscience
2015
;
2
:
785
6
.
9.
Kettle
JG
,
Alwan
H
,
Bista
M
,
Breed
J
,
Davies
NL
,
Eckersley
K
, et al
Potent and selective inhibitors of MTH1 probe its role in cancer cell survival
.
J Med Chem
2016
;
59
:
2346
61
.
10.
Kawamura
T
,
Kawatani
M
,
Muroi
M
,
Kondoh
Y
,
Futamura
Y
,
Aono
H
, et al
Proteomic profiling of small-molecule inhibitors reveals dispensability of MTH1 for cancer cell survival
.
Sci Rep
2016
;
6
:
26521
.
11.
Rahm
F
,
Viklund
J
,
Tresaugues
L
,
Ellermann
M
,
Giese
A
,
Ericsson
U
, et al
Creation of a novel class of potent and selective mutt homologue 1 (MTH1) inhibitors using fragment-based screening and structure-based drug design
.
J Med Chem
2018
;
61
:
2533
51
.
12.
Warpman Berglund
U
,
Sanjiv
K
,
Gad
H
,
Kalderen
C
,
Koolmeister
T
,
Pham
T
, et al
Validation and development of MTH1 inhibitors for treatment of cancer
.
Ann Oncol
2016
;
27
:
2275
83
.
13.
Gad
H
,
Mortusewicz
O
,
Rudd
SG
,
Stolz
A
,
Amaral
N
,
Brautigam
L
, et al
MTH1 promotes mitotic progression to avoid oxidative DNA damage in cancer cells
.
bioRxiv
2019
.
14.
Costantino
L
,
Sotiriou
SK
,
Rantala
JK
,
Magin
S
,
Mladenov
E
,
Helleday
T
, et al
Break-induced replication repair of damaged forks induces genomic duplications in human cells Science
2014
;
343
:
88
91
.
15.
Johnson
RE
,
Klassen
R
,
Prakash
L
,
Prakash
S
. 
A Major role of DNA polymerase delta in replication of both the leading and lagging DNA strands
.
Mol Cell
2015
;
59
:
163
75
.
16.
Ghodgaonkar
MM
,
Kehl
P
,
Ventura
I
,
Hu
L
,
Bignami
M
,
Jiricny
J
. 
Phenotypic characterization of missense polymerase-delta mutations using an inducible protein-replacement system
.
Nat Commun
2014
;
5
:
4990
.
17.
Brautigam
L
,
Pudelko
L
,
Jemth
AS
,
Gad
H
,
Narwal
M
,
Gustafsson
R
, et al
Hypoxic signaling and the cellular redox tumor environment determine sensitivity to MTH1 inhibition
.
Cancer Res
2016
;
76
:
2366
75
.
18.
Wang
JY
,
Jin
L
,
Yan
XG
,
Sherwin
S
,
Farrelly
M
,
Zhang
YY
, et al
Reactive oxygen species dictate the apoptotic response of melanoma cells to TH588
.
J Invest Dermatol
2016
;
136
:
2277
86
.
19.
Gul
N
,
Karlsson
J
,
Tangemo
C
,
Linsefors
S
,
Tuyizere
S
,
Perkins
R
, et al
The MTH1 inhibitor TH588 is a microtubule-modulating agent that eliminates cancer cells by activating the mitotic surveillance pathway
.
Sci Rep
2019
;
9
;
14667
.
20.
Patterson
JC
,
Joughin
BA
,
Prota
AE
,
Muhlethaler
T
,
Jonas
OH
,
Whitman
MA
, et al
VISAGE reveals a targetable mitotic spindle vulnerability in cancer cells
.
Cell Syst
2019
;
9
;
74
92 e78
.
21.
Petrocchi
A
,
Leo
E
,
Reyna
NJ
,
Hamilton
MM
,
Shi
X
,
Parker
CA
, et al
Identification of potent and selective MTH1 inhibitors
.
Bioorg Med Chem Lett
2016
;
26
:
1503
7
.
22.
Rai
P
,
Onder
TT
,
Young
JJ
,
McFaline
JL
,
Pang
B
,
Dedon
PC
, et al
Continuous elimination of oxidized nucleotides is necessary to prevent rapid onset of cellular senescence
.
Proc Natl Acad Sci U S A
2009
;
106
:
169
74
.
23.
Bartkova
J
,
Hamerlik
P
,
Stockhausen
MT
,
Ehrmann
J
,
Hlobilkova
A
,
Laursen
H
, et al
Replication stress and oxidative damage contribute to aberrant constitutive activation of DNA damage signalling in human gliomas
.
Oncogene
2010
;
29
:
5095
102
.
24.
Bartkova
J
,
Rezaei
N
,
Liontos
M
,
Karakaidos
P
,
Kletsas
D
,
Issaeva
N
, et al
Oncogene-induced senescence is part of the tumorigenesis barrier imposed by DNA damage checkpoints
.
Nature
2006
;
444
:
633
7
.
25.
Dobbelstein
M
,
Sorensen
CS
. 
Exploiting replicative stress to treat cancer. Nature reviews
.
Drug Discov
2015
;
14
:
405
23
.
26.
Bianchi
J
,
Rudd
S
,
Jozwiakowski
SK
,
Bailey
L
,
Soura
V
,
Taylor
E
, et al
PrimPol bypasses UV photoproducts during eukaryotic chromosomal DNA replication
.
Mol Cell
2013
;
52
:
566
73
.
27.
Mouron
S
,
Rodriguez-Acebes
S
,
Martínez-Jiménez
MI
,
García-Gómez
S
,
Chocrón
S
,
Blaco
L
, et al
Repriming of DNA synthesis at stalled replication forks by human PrimPol
.
Nat Struct Mol Biol
2013
;
20
:
1383
9
.
28.
Elvers
I
,
Johansson
F
,
Groth
P
,
Erixon
K
,
Helleday
T
. 
UV stalled replication forks restart by re-priming in human fibroblasts
.
Nucleic Acids Res
2011
;39:7049–57.
29.
Bryant
HE
,
Petermann
E
,
Schultz
N
,
Jemth
AS
,
Loseva
O
,
Issaeva
N
, et al
PARP is activated at stalled forks to mediate Mre11-dependent replication restart and recombination
.
EMBO J
2009
;
28
:
2601
15
.
30.
Patterson
JC
,
Joughin
BA
,
van de Kooij
B
,
Lim
DC
,
Lauffenburger
DA
,
Yaffe
MB
. 
ROS and oxidative stress are elevated in mitosis during asynchronous cell cycle progression and are exacerbated by mitotic arrest
.
Cell Syst
2019
;
8
:
163
7 e162
.
31.
Domenech
E
,
Maestre
C
,
Esteban-Martinez
L
,
Partida
D
,
Pascual
R
,
Fernandez-Miranda
G
, et al
AMPK and PFKFB3 mediate glycolysis and survival in response to mitophagy during mitotic arrest
.
Nat Cell Biol
2015
;
17
:
1304
16
.
32.
Minocherhomji
S
,
Ying
S
,
Bjerregaard
VA
,
Bursomanno
S
,
Aleliunaite
A
,
Wu
W
, et al
Replication stress activates DNA repair synthesis in mitosis
.
Nature
2015
;
528
:
286
90
.
33.
Chen
S
,
Zhang
Q
,
Wu
X
,
Schultz
PG
,
Ding
S
. 
Dedifferentiation of lineage-committed cells by a small molecule
.
J Am Chem Soc
2004
;
126
:
410
1
.
34.
Wilhelm
T
,
Olziersky
AM
,
Harry
D
,
De Sousa
F
,
Vassal
H
,
Eskat
A
, et al
Mild replication stress causes chromosome mis-segregation via premature centriole disengagement
.
Nat Commun
2019
;
10
;
3585
.
35.
Freudenthal
BD
,
Beard
WA
,
Perera
L
,
Shock
DD
,
Kim
T
,
Schlick
T
, et al
Uncovering the polymerase-induced cytotoxicity of an oxidized nucleotide
.
Nature
2015
;
517
:
635
9
.
36.
Foti
JJ
,
Devadoss
B
,
Winkler
JA
,
Collins
JJ
,
Walker
GC
. 
Oxidation of the guanine nucleotide pool underlies cell death by bactericidal antibiotics
.
Science
2012
;
336
;
315
9
.