Temozolomide is a DNA-alkylating agent used to treat brain tumors, but resistance to this drug is common. In this study, we provide evidence that efficacious responses to this drug can be heightened significantly by coadministration of an artificial nucleoside (5-nitroindolyl-2′-deoxyriboside, 5-NIdR) that efficiently and selectively inhibits the replication of DNA lesions generated by temozolomide. Conversion of this compound to the corresponding nucleoside triphosphate, 5-nitroindolyl-2′-deoxyriboside triphosphate, in vivo creates a potent inhibitor of several human DNA polymerases that can replicate damaged DNA. Accordingly, 5-NIdR synergized with temozolomide to increase apoptosis of tumor cells. In a murine xenograft model of glioblastoma, whereas temozolomide only delayed tumor growth, its coadministration with 5-NIdR caused complete tumor regression. Exploratory toxicology investigations showed that high doses of 5-NIdR did not produce the side effects commonly seen with conventional nucleoside analogs. Collectively, our results offer a preclinical pharmacologic proof of concept for the coordinate inhibition of translesion DNA synthesis as a strategy to improve chemotherapeutic responses in aggressive brain tumors.

Significance: Combinatorial treatment of glioblastoma with temozolomide and a novel artificial nucleoside that inhibits replication of damaged DNA can safely enhance therapeutic responses. Cancer Res; 78(4); 1083–96. ©2017 AACR.

Glioblastoma multiforme (GBM) is the most aggressive form of all malignant primary brain tumors found in humans. This is also the most common type of brain cancer as over 12,000 children and adults are diagnosed with GBM each year in the United States (1). Current standard of care for GBM is surgical resection followed by focal radiotherapy and chemotherapy. However, even with aggressive treatments, the median survival time for GBM patients is less than 16 months (2). One important chemotherapeutic agent used to treat GBM is temozolomide (TMZ), an orally administered DNA-alkylating agent (3). TMZ is a second-generation imidazotetrazine prodrug that does not require hepatic metabolism for activation but instead undergoes spontaneous conversion to become an active alkylating agent under normal physiologic conditions (4). TMZ displays antitumor activity against a wide variety of malignancies including lymphoma, leukemia, and colon cancer (5). However, its ability to easily cross the blood–brain barrier makes it particularly useful in the treatment of brain tumors (6). This drug produces cytostatic and cytotoxic effects primarily through the nonenzymatic methylation of DNA. TMZ creates a number of DNA lesions including N3-methyladenine, O6-methylguanine, and N7-methylguanine, the most commonly formed DNA adduct (7). In addition, methylation at the N7 position of guanine produces a more toxic DNA lesion, termed an abasic site, which forms by the spontaneous depurination of the methylated base (8). Although each type of DNA lesion stimulates DNA repair pathways to induce apoptosis, resistance to TMZ can unfortunately develop through the inactivation of DNA repair pathways such as DNA mismatch repair (MMR). In fact, defects in MMR coupled with continued treatment with TMZ can generate significantly higher amounts of mutagenesis (9–13). Higher mutation frequencies can occur as unrepaired lesions formed by TMZ can be inappropriately replicated by various DNA polymerases in a process known as translesion DNA synthesis (Fig. 1A). In addition to directly causing drug resistance, the promutagenic nature of translesion DNA synthesis (TLS) activity can also diminish the efficacy of TMZ through mutagenesis of key regulator proteins. Indeed, Johnson and colleagues recently reported that genomic DNAs isolated from recurrent GBM tumors treated with TMZ were highly mutated, containing between 30 and 90 mutations per megabase compared with initial tumors that had significant lower mutation frequencies (0.2 to 4.5 mutations per Mb; ref. 14). These recurrent GBM tumors were drug resistant, and this coincided with the accumulation of acquired somatic mutations in MMR genes as well as through mutations in the retinoblastoma and mTOR (Akt–mTOR) pathways (14). These findings collectively highlight how the efficacy of TMZ can be severely compromised by inappropriate replication of DNA lesions caused by TLS activity. This article describes a new therapeutic strategy to increase the efficacy of DNA-damaging agents such as TMZ that involves inhibiting the inappropriate replication of damaged DNA during TLS.

Figure 1.

Model for the strategy employing artificial nucleosides as a method to combat drug resistance caused by translesion DNA synthesis. A, Generalized model for translesion DNA synthesis. In this model, a DNA polymerase misinserts a nucleotide opposite a DNA lesion and then extends beyond it. The biological consequences of this activity include the onset of drug resistance and an increase in mutagenesis. B, Comparison of the chemical structures of dATP, 5-NITP, and 3-Eth-5-NITP. C, Dose-response curves for TMZ against the human glioblastoma cell lines U87, A172, and SW1088. Cell viability was normalized against 0.1 % DMSO treatment (vehicle control). Each experiment represents an average of three independent determinations performed on different days. In all cases, TMZ displays low potency as exhibited in high LD50 values greater than 100 μmol/L. D, Dose-response curves for 5-NIdR against human glioblastoma cell lines (U87, A172, and SW1088). Each experiment represents an average of three independent determinations performed on different days. In all cases, 5-NIdR displays low potency as exhibited in high LD50 values greater than 100 μg/mL (360 μmol/L). E, Combining sublethal doses of 5-NIdR and TMZ generates a synergistic cytotoxic effect compared with treatment with either TMZ or 5-NIdR alone. The percentage of nonviable cells of 33 ± 2% generated using the combination of sublethal doses of 5-NIdR and TMZ is greater than the additive effects of treatment with TMZ or 5-NIdR alone (percentage of nonviable cells = 22 ± 2%). Each experiment represents an average of three independent determinations performed on different days. **, P > 0.01.

Figure 1.

Model for the strategy employing artificial nucleosides as a method to combat drug resistance caused by translesion DNA synthesis. A, Generalized model for translesion DNA synthesis. In this model, a DNA polymerase misinserts a nucleotide opposite a DNA lesion and then extends beyond it. The biological consequences of this activity include the onset of drug resistance and an increase in mutagenesis. B, Comparison of the chemical structures of dATP, 5-NITP, and 3-Eth-5-NITP. C, Dose-response curves for TMZ against the human glioblastoma cell lines U87, A172, and SW1088. Cell viability was normalized against 0.1 % DMSO treatment (vehicle control). Each experiment represents an average of three independent determinations performed on different days. In all cases, TMZ displays low potency as exhibited in high LD50 values greater than 100 μmol/L. D, Dose-response curves for 5-NIdR against human glioblastoma cell lines (U87, A172, and SW1088). Each experiment represents an average of three independent determinations performed on different days. In all cases, 5-NIdR displays low potency as exhibited in high LD50 values greater than 100 μg/mL (360 μmol/L). E, Combining sublethal doses of 5-NIdR and TMZ generates a synergistic cytotoxic effect compared with treatment with either TMZ or 5-NIdR alone. The percentage of nonviable cells of 33 ± 2% generated using the combination of sublethal doses of 5-NIdR and TMZ is greater than the additive effects of treatment with TMZ or 5-NIdR alone (percentage of nonviable cells = 22 ± 2%). Each experiment represents an average of three independent determinations performed on different days. **, P > 0.01.

Close modal

Reagents

PBS, antibiotic and antifungal agents, amphotericin, propidium iodide (PI), PrestoBlue, DAPI, Alexa Fluor 488, and apoptosis assay kit containing Alexa Fluor 488–labeled Annexin V were from Invitrogen. 5-nitroindolyl-2′-deoxyriboside (5-NIdR), 5-nitroindolyl-2′-deoxyriboside triphosphate (5-NITP), 3-Eth-5-NIdR, and 3-Eth-5-NITP were synthesized and purified as previously described (15, 16). DNAs including that containing an abasic site were obtained from Operon and purified as described (46). TMZ (>98% purity) was purchased from Sigma-Aldrich. Recombinant human DNA polymerases including pol delta, pol epsilon, pol eta, pol iota, pol kappa, pol lambda, and pol mu were obtained from Enzymax, LLC. Each polymerase was judged to be >97% pure as assessed by sodium dodecylsulfate-polyacrylamide denaturing gel electrophoresis. Male and female C57BL/6 mice (∼6 weeks) and female (Crl:NU(NCr)-Foxn1nu) mice were obtained from Charles River Laboratories. All mice were group-housed (2–4 per cage) with unlimited access to food and water, and maintained on a 12-hour light/dark cycle (lights on at 06:00 hours).

Cell-based studies

All experiments used human cancer cell lines including U87, SW1088, and A172 and were obtained from the ATCC. As such, informed consent was not required for these cell lines. Cell lines were routinely authenticated based on morphology and growth characteristics. All cells were expanded and then frozen at low passage (passages 2–5) within 2 weeks after the receipt of the original stocks. All cells used for experiments were between passages 6 and 12. They were tested for mycoplasma after each thaw or every 4 weeks when grown in culture. Mycoplasma infection was detected using the MycoAlert Mycoplasma Detection Kit from Lonza. All adherent cancer cell lines were grown in DMEM (Cellgro) supplemented with 10% FBS (Biowest) and 1.0% penicillin streptomycin (Gibco) at 37°C with 5.0% CO2.

Cell proliferation assays

Cells were plated at a density of approximately 10,000/well in 200 μL media overnight in a 96-well plate. TMZ was added in a dose-dependent manner (1–100 μmol/L). 5-NIdR was added in a dose-dependent manner (1–100 μg/mL). After variable periods of time (24–72 hours), media were removed and replenished with 90 μL fresh medium followed by the addition of 10 μL of PrestoBlue reagent (Invitrogen). Cells were incubated for at least 30 minutes, and the optical density of samples was read at 560 nm. Background dye absorbance was subtracted from each sample. Cell viability was normalized against cells treated with DMSO. IC50 values were obtained using Equation A:

formula

where y is the fraction of viable cells, IC50 is the concentration that inhibits 50% cell growth, and inhibitor is the concentration of compound tested. Each experiment represents an average of three independent determinations performed on different days.

Apoptosis measurements

Cells were plated at an initial density of 200,000 cells/mL. TMZ or 5-NIdR was added in a dose-dependent fashion for 72 hours. TMZ was varied at concentrations ranging from 5 to 100 μmol/L, whereas 5-NIdR was varied from 1 to 100 μg/mL. In some cases, cells were treated with fixed concentrations of 5-NIdR (100 μg/mL) and TMZ (100 μmol/L). At variable time intervals (24–72 hours), media containing TMZ and/or 5-NIdR were removed and replaced with fresh media. Cells were then treated with 0.25% trypsin, harvested by centrifugation, washed in PBS, and resuspended in 100 μL of binding buffer containing 5 μmol/L of Annexin V–Alexa Fluor 488 conjugate. Cells were treated with 1 μg/ mL PI and incubated at room temperature for 15 minutes followed by flow cytometry analysis. Cells were analyzed using either Muse Cell analyzer or Beckman Coulter EPICS-XL with EXPO 32 Data Acquisition software. A total of 15,000 gated events were observed for each sample.

Cell-cycle analyses

U87 cells were plated at an initial density of 200,000/mL. Cells were treated with a fixed concentration of 5-NIdR (100 μg/mL), TMZ (100 μmol/L), or a combination of 100 μg/mL 5-NIdR and 100 μmol/L TMZ. Cell-cycle analysis on fixed, permeabilized cells was performed using PI (Invitrogen). Approximately 106 cells were fixed in 500 μL cold methanol and stored at 4°C overnight. Cells were resuspended in 500 μL PI solution [1 μg/mL PI, 0.1% (v/v) Triton, 0.1% (w/v) NaN3, 1.2% (v/v) 100 μg/mL RNAse A) and incubated for 1 hour at 37°C. Cells were analyzed using either Muse Cell analyzer or Beckman Coulter EPICS-XL with EXPO 32 Data Acquisition software. A total of 15,000 gated events were observed for each sample.

DNA damage response

U87 cells were plated at an initial density of 200,000/mL. Cells were treated with a fixed concentrations of 5-NIdR (100 μg/mL), TMZ (100 μmol/L), or a combination of 100 μg/mL 5-NIdR and 100 μmol/L TMZ. After 72 hours, media containing TMZ and/or 5-NIdR were removed and replaced with fresh media. Cells were then treated with 0.25% trypsin, harvested by centrifugation, washed in PBS, and resuspended in 1X Assay Buffer (50 μL per 100,000 cells). Cells were permeabilized by adding ice-cold 1X permeabilization buffer and incubated on ice for 10 minutes. Cells were again centrifuged at 300 × g for 5 minutes, resuspended in 1X assay buffer, and 5 μL of antibodies against pATM and p-γH2AX were added. Cells were incubated for 30 minutes in the dark at room temperature and then analyzed using a Muse Cell analyzer. A total of 15,000 gated events were observed for each sample.

Kinetic parameters for nucleotide incorporation

Kinetic studies using human DNA polymerases including pol delta, pol epsilon, pol eta, pol iota, pol kappa, pol lambda, and pol mu were performed using an assay buffer consisting of 50 mmol/L TrisOAc, 1 mg/mL BSA, 10 mmol/L DTT, and 5 mmol/L MgCl2 at pH 7.5. All assays were performed at 37°C. kcat, Km, and kcat/Km values for nucleotides were measured as described (15) Briefly, a typical assay was performed by preincubating DNA substrate (250 nmol/L) with a limiting amount of polymerase (10 or 20 nmol/L) in assay buffer and Mg2+. Reactions were initiated by adding variable concentrations of either dATP or 5-NITP (1–500 μmol/L). At variable time intervals, 5 μL aliquots of the reaction were quenched by adding an equal volume of 200 mmol/L EDTA. Polymerization reactions were monitored by analyzing products on 20% sequencing gels as previously described (28). Gel images were obtained with a Packard PhosphorImager using the OptiQuant software supplied by the manufacturer. Product formation was quantified by measuring the ratio of 32P-labeled extended and unextended primer. The ratios of product formation are corrected for substrate in the absence of polymerase (zero point). Corrected ratios are then multiplied by the concentration of primer/template used in each assay to yield total product. Steady-state rates were obtained from the linear portion of the time course and were fit to Equation B:

formula

where m is the slope of the line and the rate of polymerization reaction (nmol/L min−1), b is the y intercept, and t is time. Data for the dependency of rate as a function of nucleotide concentration were fit to the Michaelis–Menten equation (Equation C):

formula

where ν is the rate of product formation (nmol/L min−1), Vmax is the maximal rate of polymerization, Km is the Michaelis constant for dXTP, and [dXTP] is the concentration of nucleotide substrate. The turnover number, kcat, is Vmax divided by the final concentration of polymerase used in the experiment.

Chain-termination capabilities of 5-NITP against high fidelity and specialized DNA polymerases

Assays were performed using pseudo-first order reaction conditions in which a limiting concentration of DNA polymerase (20 nmol/L) was added last to a preincubated solution containing 250 nmol/L DNA containing an abasic site (13/20Sp-mer) or DNA-containing thymine (13/20T-mer) in assay buffer and a fixed concentration of 5-NITP (100 μmol/L) or dATP (100 μmol/L). After 10 minutes, an aliquot of the reaction was quenched with 200 mmol/L EDTA to verify insertion of the artificial or natural nucleotide opposite the abasic site or thymine. At this time, an aliquot of dNTPs (500 μmol/L final concentration of dCTP, dGTP, and dTTP) was added to the remaining reaction mixture to initiate elongation. Aliquots of the elongation reaction were then quenched with 200 mmol/L EDTA at variable times (0–30 minutes) and analyzed by denaturing gel electrophoresis to assess elongation beyond either dATP or 5-NITP.

Monitoring translesion DNA synthesis in cells

“Click” reactions were performed using cells harvested after 2 days of treatment with DMSO, TMZ (100 μmol/L), 3-Eth-5-NIdR (25 μg/mL), or TMZ (100 μmol/L) with 3-Eth-5-NIdR (25 μg/mL). All cells were fixed with cold methanol overnight. Cells were treated with 0.3 mL of saponin-based permeabilization and wash buffer for 45 minutes at 37°C. “Click” reactions were initiated with click-iT reaction cocktail followed by incubation at 37°C for 90 minutes. Cells were washed twice with wash buffer. Cell pellets were dislodged using 0.5 mL solution of 10 μg/mL PI and RNAase A in saponin-based permeabilization buffer. Cells were incubated for 15 minutes with 1 μg/mL DAPI prior to analysis. Images were obtained using an EVOSfl Advanced microscope (×40 magnification).

Animal studies

Protocols for animal use were approved by the Institutional Animal Care and Use Committee at Cleveland State University. All procedures were carried out in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals. U87 cells were injected into the bilateral flanks of female athymic nude mice (nu/nu, 6–8 weeks old), and tumor growth was measured with calipers. Once the tumor reached a specified volume (100 or 500 mm3), mice received 5 consecutive days of i.p. treatment with DMSO/PBS, 40 mg/kg TMZ, 100 mg/kg 5-NIdR, or 40 mg/kg TMZ and 100 mg/kg 5-NIdR. Following treatment, tumor growth was monitored every day to measure tumor volumes calculated using Equation D:

formula

Relative tumor volumes were compared with an unpaired t test and expressed as the mean of the tumor volume ± SEM. The tumor growth delay was calculated as (number of days for treated tumor to double in size from the first day of treatment) – (number of days for the control tumor to double in size from the first day of treatment). A one-way ANOVA compared the relative tumor volume of all treatment groups to determine P values. Blood samples were collected 14 days after treatment via tail nick and analyzed using a Hemavet 950FS.

Statistical analyses

All data showing error bars are presented as mean ± SEM. The significance of difference in the mean value was determined using a two-tailed Student t test, and normal distribution was assumed in all cases. P values <0.05 were considered significant. All calculations were performed using KaliedaGraph software. All cell culture experiments were reproduced at least 3 times independently. For each experiment, the number of samples and replicates are provided in the text or figure legend.

5-NIdR increases the potency of TMZ in glioblastoma cell lines

The most common DNA adduct formed by TMZ is N7-methylguanine, which undergoes spontaneous depurination to produce abasic sites (7). Under normal physiologic conditions, abasic sites frequently form in human cells at an estimated rate of approximately 10,000 lesions generated per cell per day (17). However, this rate significantly increases with exposure to DNA-damaging agents. Because abasic sites lack Watson–Crick coding information, they are classified as noninstructional DNA lesions and are thus highly promutagenic (18). Indeed, many DNA polymerases display unique behavior as they preferentially incorporate dATP opposite the noninstructional DNA lesion (19–24). Furthermore, the formed mispair can be efficiently extended during TLS to cause drug resistance and increase mutagenesis. In an effort to combat drug resistance caused by TLS, we previously developed an artificial nucleotide analog designated 5-NITP (Fig. 1B) that is efficiently and selectively inserted opposite abasic sites (25). Our previous in vitro studies demonstrated that 5-NITP is utilized approximately 1,000-fold more efficiently than dATP during the replication of an abasic site (25, 26). Despite being efficiently inserted opposite this lesion, 5-NITP is refractory to elongation and thus acts as chain terminator to inhibit TLS activity (27). These inhibitory effects predicted that the corresponding artificial nucleoside, 5-NIdR, could increase the potency of TMZ and make the DNA-damaging agent more effective against drug-resistant brain cancers. This hypothesis was tested by quantifying the ability of 5-NIdR to potentiate the cytotoxic effects of TMZ against three human GBM cell lines (U87, SW1088, and A172). We first validated that each cell line is resistant to TMZ as the LD50 value (measured 3 days after treatment) for the DNA-damaging agent is greater than 100 μmol/L in all three cancer cell lines (Fig. 1C). Next, we demonstrated that 5-NIdR also displays low potency as a monotherapeutic agent against each GBM cell line as the LD50 value (measured 3 days after treatment) is greater than 100 μg/mL in all three cancer cell lines (Fig. 1D). Please note that 100 μg/mL 5-NIdR corresponds to a molar concentration of 360 μmol/L. In general, the low potency of 5-NIdR is expected as the corresponding artificial nucleoside triphosphate (5-NITP) is poorly incorporated opposite undamaged DNA (11). However, data provided in Fig. 1E show that combining sublethal concentrations of 5-NIdR (100 μg/mL) with TMZ (100 μmol/L) produces a synergistic increase in cell death compared with the additive effects of either compound used independently. As illustrated, the percentage of nonviable cells of 33.8 ± 3.0% caused by combining 5-NIdR and TMZ is greater than the additive effects (22.5 ± 2.8%) of TMZ (10.2 ± 1.1%) plus 5-NIdR (12.3 ± 2.2%) used individually. Data are provided using U87 cells as this cell line is more resistant to both TMZ and 5-NIdR compared with either the A172 or SW1088 cell line (Fig. 1C and D, respectively). However, similar synergistic cytotoxic effects are also observed with the A172 and SW1088 cell lines (Supplementary Fig S1A and Fig S1B, respectively).

The cellular mechanisms accounting for this synergistic cell killing effect were examined using PI uptake and Annexin V staining to distinguish viable cells from those undergoing early- and late-stage apoptosis or necrosis. Representative data provided in Supplementary Fig. S2A show that U87 cells cotreated with 100 μg/mL 5-NIdR, and 100 μmol/L TMZ displays significantly higher levels of early- and late-stage apoptosis compared with cells treated individually with TMZ or 5-NIdR. This was confirmed through multiple determinations (n = 3), and Table 1 provides a summary of average values obtained from three independent determinations. Close inspection shows that the net overall apoptotic effect of 28.8% observed by combining 5-NIdR with TMZ is approximately 4-fold greater than the additive effects (6.8%) of 5-NIdR and TMZ used individually.

Table 1.

Summary of dual-parameter flow cytometry measuring apoptosis in U87 cells

ConditionViableEarly apoptoticLate apoptoticNecroticTotal apoptotic
DMSO 87.5 ± 3.8% 6.5 ± 1.2% 5.0 ± 1.5% 1.0 ± 0.3% 11.5% (0%) 
100 μmol/L TMZ 87.8 ± 4.1% 6.9 ± 1.4% 4.7 ± 1.1% 0.6 ± 0.1% 11.6% (0.1%) 
100 μg/mL 5-NIdR 80.9 ± 3.3% 9.7 ± 2.5% 8.5 ± 2.2% 0.9 ± 0.4% 18.2% (6.7%) 
Combination 58.6 ± 6.0% 12.0 ± 2.8% 28.3 ± 4.2% 1.1 ± 0.3% 40.3% (28.8%) 
ConditionViableEarly apoptoticLate apoptoticNecroticTotal apoptotic
DMSO 87.5 ± 3.8% 6.5 ± 1.2% 5.0 ± 1.5% 1.0 ± 0.3% 11.5% (0%) 
100 μmol/L TMZ 87.8 ± 4.1% 6.9 ± 1.4% 4.7 ± 1.1% 0.6 ± 0.1% 11.6% (0.1%) 
100 μg/mL 5-NIdR 80.9 ± 3.3% 9.7 ± 2.5% 8.5 ± 2.2% 0.9 ± 0.4% 18.2% (6.7%) 
Combination 58.6 ± 6.0% 12.0 ± 2.8% 28.3 ± 4.2% 1.1 ± 0.3% 40.3% (28.8%) 

NOTE: Values represent an average of three independent determinations performed on different days. Values in parenthesis represent the difference in percent apoptosis of treatment compared with treatment with DMSO (vehicle control).

We next examined the effect of this drug combination on cell-cycle progression using PI staining to measure cellular DNA content. The results of representative experiments are provided in Supplementary Fig. S2B. Table 2 provides a summary of average values obtained from three independent determinations. In all experiments, a baseline for cell-cycle progression was first determined by treating cells with 0.1% DMSO over a 3-day period. The histogram for cells treated with DMSO displays a pattern consistent with an asynchronous cell population as the majority of cells exist at G0–G1 (54.2 ± 3.2%), whereas smaller populations exist at S-phase (11.9 ± 2.5%), G2–M (25.2 ± 3.1%), and sub-G1 (8.7 ± 1.4%). Treatment with 100 μg/mL 5-NIdR for 3 days produces a similar profile (G0–G1 = 46.2 ± 3.9%, S-phase = 13.7 ± 3.2%, G2–M = 25.7 ± 2.1%, and sub-G1 = 14.4 ± 1.9%). These results again indicate that the artificial nucleoside generates a minimal cytotoxic effect in the absence of an exogenous DNA-damaging agent. However, treatment with 100 μmol/L TMZ over the same time period produces striking effects on cell-cycle progression. In particular, there is a significant increase in cells at G2–M (46.7 ± 2.4%) with a concomitant reduction in cells at G0–G1 (29.6 ± 4.1%). These changes occur with little effect on cells undergoing chromosomal DNA synthesis (S-phase = 14.2 ± 1.5%) or cells undergoing apoptosis (sub-G1 = 9.5 ± 2.5%). The inability of TMZ to generate an effect on S-phase cells coupled with the accumulation of cells at G2–M suggests that DNA lesions produced by TMZ do not inhibit chromosomal DNA synthesis. This likely reflects the activity of specialized DNA polymerases such as pol eta and pol iota to easily by-pass the DNA lesions produced by TMZ. Consistent with this hypothesis, cotreating U87 cells with TMZ and 5-NIdR produces an approximately 2-fold increase in the population of cells at S-phase (23.4 ± 1.9%) compared with treatment with TMZ (14.2 ± 1.5%) or 5-NIdR (13.7 ± 3.2%) alone. This increase in S-phase likely results from the ability of the corresponding nucleoside triphosphate, 5-NITP, to inhibit the replication of abasic sites produced by TMZ. Furthermore, cells that accumulate at S-phase appear to undergo apoptosis as there is also a significant increase in sub-G1 DNA (19.2 ± 2.9%) compared with cells treated with TMZ (9.5 ± 2.5%) or 5-NIdR (14.4 ± 1.9%) alone.

Table 2.

Summary of the effects of drug treatment on cell-cycle progression in U87 cells

ConditionG0–G1S-phaseG2–MSub-G1
DMSO 54.2 ± 3.2% 11.9 ± 2.5% 25.2 ± 3.1% 8.7 ± 1.4% (0%) 
100 μmol/L TMZ 29.6 ± 4.1% 14.2 ± 1.5% 46.7 ± 2.4% 9.5 ± 2.5% (0.8%) 
100 μg/mL 5-NIdR 46.2 ± 3.9% 13.7 ± 3.2% 25.7 ± 2.1% 14.4 ± 1.9% (5.7%) 
Combination 33.1 ± 2.5% 23.4 ± 1.9% 24.3 ± 1.5% 19.2 ± 2.9% (10.5%) 
ConditionG0–G1S-phaseG2–MSub-G1
DMSO 54.2 ± 3.2% 11.9 ± 2.5% 25.2 ± 3.1% 8.7 ± 1.4% (0%) 
100 μmol/L TMZ 29.6 ± 4.1% 14.2 ± 1.5% 46.7 ± 2.4% 9.5 ± 2.5% (0.8%) 
100 μg/mL 5-NIdR 46.2 ± 3.9% 13.7 ± 3.2% 25.7 ± 2.1% 14.4 ± 1.9% (5.7%) 
Combination 33.1 ± 2.5% 23.4 ± 1.9% 24.3 ± 1.5% 19.2 ± 2.9% (10.5%) 

NOTE: Values represent an average of three independent determinations performed on different days. Values in parentheses represent the difference in percent sub-G1 DNA measured with various treatments compared with treatment with DMSO (vehicle control).

To further assess the mechanism of cell death, we measured single- and double-strand DNA breaks (DSBs) formed in cells treated with TMZ alone or combined with 5-NIdR. This was approached by quantifying the levels of pATM caused by single-strand DNA breaks (28) and pH2AX that increases after formation of DSBs (29). Supplementary Fig. S2C provides representative flow cytometry data, whereas Table 3 summarizes the average of values obtained from three independent experiments. The collective dataset shows that U87 cells treated with a combination of 100 μmol/L TMZ and 100 μg/mL 5-NIdR have significantly higher levels of pATM (14.8 ± 0.5%) and DSBs (5.2 ± 0.3%) compared with cells treated individually with TMZ (4.2 ± 0.2% and 1.1 ± 0.1%, respectively) or 5-NIdR (4.5 ± 0.2% and 1.1 ± 0.3%, respectively). The increase in pATM levels is indicative of the production of single-strand DNA caused by inhibiting chromosomal DNA synthesis during S-phase. Collectively, these biochemical results suggest that 5-NIdR potentiates the anticancer effects of TMZ by inhibiting the ability of DNA polymerases to replicate DNA lesions produced by the DNA-damaging agent.

Table 3.

Summary of the effects of drug treatment on the DNA damage response in U87 cells

ConditionNegativepATMpH2AXDSBsTotal
DMSO 95.0 ± 1.5% 2.7 ± 0.1% 1.2 ± 0.1% 1.1 ± 0.2% 5.0% (0%) 
100 μmol/L TMZ 91.8 ± 2.1% 4.2 ± 0.2% 2.9 ± 0.2% 1.1 ± 0.1% 8.2% (3.2%) 
100 μg/mL 5-NIdR 91.7 ±1.8% 4.5 ± 0.2% 2.7 ± 0.3% 1.1 ± 0.3% 8.3% (3.3%) 
Combination 79.4 ± 1.1% 14.8 ± 0.5% 0.6 ± 0.1% 5.2 ± 0.3% 20.6% (15.6%) 
ConditionNegativepATMpH2AXDSBsTotal
DMSO 95.0 ± 1.5% 2.7 ± 0.1% 1.2 ± 0.1% 1.1 ± 0.2% 5.0% (0%) 
100 μmol/L TMZ 91.8 ± 2.1% 4.2 ± 0.2% 2.9 ± 0.2% 1.1 ± 0.1% 8.2% (3.2%) 
100 μg/mL 5-NIdR 91.7 ±1.8% 4.5 ± 0.2% 2.7 ± 0.3% 1.1 ± 0.3% 8.3% (3.3%) 
Combination 79.4 ± 1.1% 14.8 ± 0.5% 0.6 ± 0.1% 5.2 ± 0.3% 20.6% (15.6%) 

NOTE: Values represent an average of three independent determinations performed on different days. Values in parentheses represent the difference in the total amount of pATM, pH2AX, and DSBs measured with various treatments compared with treatment with DMSO (vehicle control).

5-NIdR increases the efficacy of TMZ by inhibiting translesion DNA synthesis

To verify that the artificial nucleoside functions as a chain terminator of TLS activity, we measured the ability of several human DNA polymerases to incorporate 5-NITP, the triphosphate form of 5-NIdR, opposite an abasic site. The DNA polymerases used in this study include two high-fidelity DNA polymerases (pol delta and pol epsilon), three specialized human DNA polymerases (pol eta, pol iota, and pol kappa), and two DNA polymerases involved in DNA repair (pol lambda and pol mu). The activity of each DNA polymerase was tested using the DNA substrate illustrated in Fig. 2A that contains an abasic site (Sp) at the 14th position of the template strand. Initial studies compared the efficiency for incorporating the preferred natural substrate, dATP, opposite an abasic site against two artificial analogs, 5-NITP and 3-Eth-5-NITP, a “clickable” nucleoside analog that can visualize TLS activity (15, vide infra). In these in vitro experiments, a fixed concentration of 10 μmol/L nucleotide substrate was added to a preincubated solution containing 250 nmol/L DNA and 20 nmol/L DNA polymerase. Reactions were quenched with EDTA after a time interval of 30 minutes. Representative gel electrophoresis images provided in Fig. 2B show that the high-fidelity DNA polymerases, pol delta and pol epsilon, poorly incorporate dATP opposite the noninstructional lesion and efficiently insert 5-NITP and 3-Eth-5-NITP opposite an abasic site. Similar results are obtained with pol eta and pol iota as both specialized DNA polymerases insert 5-NITP and 3-Eth-5-NITP opposite an abasic site more effectively compared with identical concentrations of dATP. Surprisingly, the specialized DNA polymerase, pol kappa, and the repair DNA polymerases (pol lambda and pol mu) poorly incorporate dATP, 5-NITP, and 3-Eth-5-NITP opposite an abasic site.

Figure 2.

In vitro analyses demonstrate that 5-NITP is an efficient chain terminator of translesion DNA synthesis. A, DNA substrate used in kinetic studies to measure translesion DNA synthesis by various human DNA polymerases. X in the template at position 14 denotes an abasic site (Sp) or thymine (T). B, Denaturing gel electrophoresis images comparing the incorporation of dATP, 5-NITP, and 3-Eth-5-NITP opposite an abasic site by high fidelity (pol delta and pol epsilon), specialized (pol eta, pol iota, and pol kappa), and repair DNA polymerases (pol lambda and pol mu). Assays were performed using a fixed concentration of 100 μmol/L nucleotide substrate. Reactions were quenched at a time interval of 30 minutes. The following legend is used to denote nucleotide substrates used in each reaction: 1 = no dNTP, 2 = 100 μmol/L dATP, 3 = 100 μmol/L 5-NITP, and 4 = 100 μmol/L 3-Eth-5-NITP. In general, high-fidelity DNA polymerases (pol delta and pol epsilon) poorly incorporate dATP but efficiently insert 5-NITP and 3-Eth-5-NITP opposite an abasic site. Pol eta and pol iota also poorly incorporate dATP but efficiently insert 5-NITP and 3-Eth-5-NITP opposite an abasic site. The specialized DNA polymerase, pol kappa, and the repair DNA polymerases (pol lambda and pol mu) poorly incorporate dATP, 5-NITP, and 3-Eth-5-NITP opposite an abasic site. C, Michaelis–Menten plots comparing the insertion of 5-NITP opposite an abasic site catalyzed by pol delta (), pol epsilon (), pol eta (), and pol iota (). In general, high-fidelity DNA polymerases display higher kcat and lower Km values for 5-NITP compared with specialized DNA polymerases. D, Denaturing gel electrophoresis images comparing the incorporation and extension of dATP or 5-NITP beyond an abasic site catalyzed by the high-fidelity DNA polymerase, pol delta, and the specialized DNA polymerase, pol eta. E, Microscopy images showing that 3-Eth-5-NIdR is incorporated into genomic DNA in cells treated with TMZ. Cells treated with a combination of 100 μmol/L TMZ and 25 μg/mL 3-Eth-5-NIdR display significantly higher levels of green fluorescence that colocalizes in the nucleus compared with cells treated with 0.1% DMSO or 25 μg/mL 3-Eth-5-NIdR.

Figure 2.

In vitro analyses demonstrate that 5-NITP is an efficient chain terminator of translesion DNA synthesis. A, DNA substrate used in kinetic studies to measure translesion DNA synthesis by various human DNA polymerases. X in the template at position 14 denotes an abasic site (Sp) or thymine (T). B, Denaturing gel electrophoresis images comparing the incorporation of dATP, 5-NITP, and 3-Eth-5-NITP opposite an abasic site by high fidelity (pol delta and pol epsilon), specialized (pol eta, pol iota, and pol kappa), and repair DNA polymerases (pol lambda and pol mu). Assays were performed using a fixed concentration of 100 μmol/L nucleotide substrate. Reactions were quenched at a time interval of 30 minutes. The following legend is used to denote nucleotide substrates used in each reaction: 1 = no dNTP, 2 = 100 μmol/L dATP, 3 = 100 μmol/L 5-NITP, and 4 = 100 μmol/L 3-Eth-5-NITP. In general, high-fidelity DNA polymerases (pol delta and pol epsilon) poorly incorporate dATP but efficiently insert 5-NITP and 3-Eth-5-NITP opposite an abasic site. Pol eta and pol iota also poorly incorporate dATP but efficiently insert 5-NITP and 3-Eth-5-NITP opposite an abasic site. The specialized DNA polymerase, pol kappa, and the repair DNA polymerases (pol lambda and pol mu) poorly incorporate dATP, 5-NITP, and 3-Eth-5-NITP opposite an abasic site. C, Michaelis–Menten plots comparing the insertion of 5-NITP opposite an abasic site catalyzed by pol delta (), pol epsilon (), pol eta (), and pol iota (). In general, high-fidelity DNA polymerases display higher kcat and lower Km values for 5-NITP compared with specialized DNA polymerases. D, Denaturing gel electrophoresis images comparing the incorporation and extension of dATP or 5-NITP beyond an abasic site catalyzed by the high-fidelity DNA polymerase, pol delta, and the specialized DNA polymerase, pol eta. E, Microscopy images showing that 3-Eth-5-NIdR is incorporated into genomic DNA in cells treated with TMZ. Cells treated with a combination of 100 μmol/L TMZ and 25 μg/mL 3-Eth-5-NIdR display significantly higher levels of green fluorescence that colocalizes in the nucleus compared with cells treated with 0.1% DMSO or 25 μg/mL 3-Eth-5-NIdR.

Close modal

To further quantify the utilization of 5-NITP during TLS, we measured the kinetic parameters (kcat, Km, and kcat/Km) for the artificial nucleotide by the four human DNA polymerases that displayed the highest activity observed through initial screening efforts described above. These include the high-fidelity polymerases, pol delta and pol epsilon, that are involved in chromosomal replication as well as pol eta and pol iota, which are specialized DNA polymerases that can efficiently replicate DNA lesions to produce drug resistance (30, 31). Figure 2C provides Michaelis–Menten plots for the utilization of 5-NITP by all four DNA polymerases during the replication of an abasic site. The kinetic parameters kcat, Km, and kcat/Km for 5-NITP were determined from these plots, and these values are provided in Supplementary Table S1. Inspection of these kinetic parameters shows that both high-fidelity DNA polymerases utilize 5-NITP with an approximately10-fold higher kcat/Km values compared with the specialized DNA polymerase. The kcat/Km value is an important parameter as it represents the overall catalytic efficiency of the polymerase to utilize a nucleotide substrate under physiologic conditions. The higher kcat/Km value displayed by both pol delta and pol epsilon is caused primarily by a lower Km value for 5-NITP (Km ∼2.5 μmol/L) compared with the higher Km values of 10 μmol/L and 120 μmol/L measured with pol eta and pol iota, respectively.

To better gauge the potency of 5-NITP, we attempted to measure kcat, Km, and kcat/Km values for the utilization of the natural substrate, dATP, by these DNA polymerases. These parameters could not be measured with either pol epsilon or pol iota as both DNA polymerases showed weak TLS activity using dATP to replicate an abasic site. Note that we previously reported these kinetic parameters for pol delta (32), whereas the Michaelis–Menten plot for the utilization of dATP by human pol eta is provided as Supplementary Fig. S3. Regardless, kcat, Km, and kcat/Km values for pol delta and pol eta are summarized in Supplementary Table S1. These data reveal that each DNA polymerase poorly utilizes dATP during the replication of an abasic site. In fact, the global dataset shows that both high-fidelity and specialized DNA polymerases utilize 5-NITP with a significantly higher catalytic efficiency compared with dATP. In particular, pol delta displays a kcat/Km for 5-NITP that is approximately 200-fold higher than that measured for dATP. In this case, the higher catalytic efficiency is caused exclusively by a lower Km value for 5-NITP (Km ∼2 μmol/L) compared with dATP (Km ∼560 μmol/L). Similar effects are observed with the other DNA polymerases examined in this study. The significantly higher catalytic efficiency for 5-NITP suggests that the artificial nucleotide would outcompete dATP for insertion opposite abasic sites. This would favor the incorporation of 5-NITP under cellular conditions and subsequently inhibit TLS activity catalyzed by either high-fidelity and specialized DNA polymerases.

To further interrogate this possibility, we next examined the ability of these DNA polymerases to extend beyond an abasic site when paired opposite dAMP or 5-NIMP. In these experiments, mispairs were formed by adding a fixed concentration of nucleotide substrate to a preincubated solution of DNA substrate and polymerase. After 4 half-lives, an aliquot of dNTP (500 μmol/L final concentration) was added to initiate the elongation reaction. Figure 2D provides representative gel electrophoresis data that again demonstrate the ability of pol delta and pol eta to efficiently insert 5-NITP opposite the abasic site. However, both polymerases are unable to extend beyond the artificial nucleotide. Identical results are obtained with pol epsilon and pol iota (Supplementary Fig. S4A and S4B, respectively). Collectively, these data demonstrate that 5-NITP inhibits the ability of both high-fidelity and specialized DNA polymerases to replicate abasic sites formed by exposure to TMZ.

Similar experiments were performed to investigate if 5-NITP inhibits normal DNA synthesis. This was assessed by measuring the ability of the high-fidelity DNA polymerases to extend beyond thymine (T) when paired opposite dAMP or 5-NIMP. Gel images provided in Supplementary Fig. S5A and S5B show that pol delta and pol epsilon, respectively, can efficiently extend DNA when supplied with dATP to incorporate opposite T. In contrast, these polymerases are unable to stably insert 5-NITP opposite T, even after long incubation periods (∼30 minutes). Furthermore, 5-NITP does not inhibit normal DNA synthesis as each polymerase completely extends the primer when supplied with the all four natural dNTPs in the presence of 5-NITP. Collectively, the inability of 5-NITP to be efficiently inserted opposite normal DNA indicates that the artificial nucleotide is a selective substrate for TLS activity (27).

High-field microscopy was also used to visualize the artificial nucleoside in genomic DNA in U87 cells treated with TMZ. These experiments were performed using techniques previously described (32), in which U87 cells were treated with the “clickable” nucleoside analog, 3-Eth-5-NIdR, in the absence and presence of 100 μmol/L TMZ. In this case, a concentration of 25 μg/mL 3-Eth-5-NIdR was used as this represents a sublethal dose of the artificial nucleoside against U87 cells (Supplementary Fig. S6). Please note that 25 μg/mL 3-Eth-5-NIdR corresponds to a molar concentration of 83 μmol/L. At 2 days after treatment, cells were harvested by centrifugation and washed with PBS to remove 3-Eth-5-NIdR and/or 3-Eth-5-NITP not incorporated into DNA. Cells were fixed, permeabilized, and then treated with AlexaFluor488-azide and Cu(I) catalyst to covalently attach the fluorogenic probe to 3-Eth-5-NITP incorporated into DNA. Prior to microscopy analysis, cells were costained with DAPI to identify the nucleus. Microscopy images provided in Fig. 2E show that U87 cells treated with DMSO show insignificant levels of green fluorescence. U87 cells treated with 25 μg/mL 3-Eth-5-NIdR show low levels of green fluorescence, which indicates that 3-Eth-5-NIdR is not efficiently incorporated into genomic DNA in the absence of exogenous DNA damage. However, cells cotreated with 100 μmol/L TMZ and 25 μg/mL 3-Eth-5-NIdR display significantly higher levels of green fluorescence, which colocalizes in the nucleus. The increased amount of 3-Eth-5-NITP incorporated into genomic DNA coincides with an increased production of abasic sites caused by TMZ treatment (32).

Combining 5-NIdR with TMZ causes tumor ablation in mice

The efficacy of combining 5-NIdR with TMZ as a therapeutic regimen was next examined using a xenograft mouse model. In these experiments, U87 cells were injected subcutaneously into the hindflank of athymic mice, and tumor growth was monitored as a function of time. When tumors reached a volume of approximately 100 mm3, mice received injections of 40 mg/kg TMZ, 100 mg/kg 5-NIdR, or 40 mg/kg TMZ combined with 100 mg/kg 5-NIdR for 5 consecutive days. As a negative control, mice were treated with 0.1% DMSO/PBS, which is used as the cosolvent for 5-NIdR and TMZ. Tumor growth was evaluated by visual inspection and measured using calipers as previously described (26). In each treatment group, a minimum of 6 mice were used to define statistical relevance. Figure 3A provides a Kaplan–Meier plot comparing animal survival as a function of time under each treatment. As illustrated, treatment with 100 mg/kg 5-NIdR alone has no significant effect on animal survival as the median time for death (MTD) of 32.5 days is identical to that for treatment with DMSO/PBS (MTD = 33 days). Mice treated with TMZ survive approximately 2-fold longer (MTD = 61 days), and this increase is consistent with the ability of the DNA-damaging agent to delay tumor growth by inducing cell death (33). More importantly, combining 5-NIdR with TMZ has a much more pronounced effect on survival as greater than 60% of mice treated with this combination survive beyond 200 days after treatment. The increase in overall survival is consistent with previously described cell-based experiments demonstrating that 5-NIdR potentiates the cell killing effects of TMZ by inhibiting the replication of damaged DNA. This conclusion is recapitulated by the data provided in Fig. 3B, which show that the combination of 5-NIdR and TMZ causes tumor ablation, whereas treatment with TMZ alone only delays tumor growth. In fact, the time course in tumor growth with TMZ treatment shows an interesting trend as the size of the tumor remains relatively static for approximately 35 days after treatment before showing rapid tumor growth. The ability of these treated tumors to grow rapidly may reflect the onset of drug resistance and/or the generation of mutated cancer cells that have increased growth potential.

Figure 3.

Combining 5-NIdR with TMZ increases animal survival by reducing tumor burden. A, Kaplan–Meier plot comparing animal survival as a function of time in mice treated with 0.1% DMSO/PBS (vehicle control, red line), 100 mg/kg 5-NIdR (black line), 40 mg/kg TMZ (blue line), or 40 mg/kg TMZ combined with 100 mg/kg 5-NIdR (green line). Combining 5-NIdR with TMZ has a much more pronounced effect on survival as >60% of mice cotreated with 40 mg/kg TMZ and 100 mg/kg 5-NIdR survive beyond 250 days after treatment. B, Representative time courses in tumor growth in mice treated with DMSO/PBS (vehicle control, black line), 40 mg/kg TMZ (red line), 100 mg/kg 5-NIdR (blue line), or 40 mg/kg TMZ combined with 100 mg/kg 5-NIdR (green line). Combining 5-NIdR and TMZ causes tumor ablation, whereas treatment with TMZ alone delays tumor growth by 2-fold. C, Representative images demonstrating that treatment with 40 mg/kg TMZ alone (top row) has a minimal effect on tumor growth, whereas cotreatment with 100 mg/kg 5-NIdR and 40 mg/kg TMZ (bottom row) causes complete tumor regression within 33 days after treatment. D, Kaplan–Meier plot comparing survival as a function of time in mice bearing large tumors (∼500 mm3) treated with 40 mg/kg TMZ (red line) versus 40 mg/kg TMZ combined with 100 mg/kg 5-NIdR (black line). E, Treatment with TMZ combined with 5-NIdR results in a bimodal effect on tumor growth. A total of 60% of mice treated with the combination of 5-NIdR and TMZ show complete tumor regression within 30 days after treatment (blue line), whereas 40% of mice receiving this dual treatment show a delay in tumor growth followed by rapid growth in tumor mass (black line).

Figure 3.

Combining 5-NIdR with TMZ increases animal survival by reducing tumor burden. A, Kaplan–Meier plot comparing animal survival as a function of time in mice treated with 0.1% DMSO/PBS (vehicle control, red line), 100 mg/kg 5-NIdR (black line), 40 mg/kg TMZ (blue line), or 40 mg/kg TMZ combined with 100 mg/kg 5-NIdR (green line). Combining 5-NIdR with TMZ has a much more pronounced effect on survival as >60% of mice cotreated with 40 mg/kg TMZ and 100 mg/kg 5-NIdR survive beyond 250 days after treatment. B, Representative time courses in tumor growth in mice treated with DMSO/PBS (vehicle control, black line), 40 mg/kg TMZ (red line), 100 mg/kg 5-NIdR (blue line), or 40 mg/kg TMZ combined with 100 mg/kg 5-NIdR (green line). Combining 5-NIdR and TMZ causes tumor ablation, whereas treatment with TMZ alone delays tumor growth by 2-fold. C, Representative images demonstrating that treatment with 40 mg/kg TMZ alone (top row) has a minimal effect on tumor growth, whereas cotreatment with 100 mg/kg 5-NIdR and 40 mg/kg TMZ (bottom row) causes complete tumor regression within 33 days after treatment. D, Kaplan–Meier plot comparing survival as a function of time in mice bearing large tumors (∼500 mm3) treated with 40 mg/kg TMZ (red line) versus 40 mg/kg TMZ combined with 100 mg/kg 5-NIdR (black line). E, Treatment with TMZ combined with 5-NIdR results in a bimodal effect on tumor growth. A total of 60% of mice treated with the combination of 5-NIdR and TMZ show complete tumor regression within 30 days after treatment (blue line), whereas 40% of mice receiving this dual treatment show a delay in tumor growth followed by rapid growth in tumor mass (black line).

Close modal

We next examined if combining 5-NIdR with TMZ could produce similar beneficial effects against tumors larger than 100 mm3. This is important because patients with brain cancers are typically diagnosed only after the manifestation of severe neurological symptoms (headaches, nausea, vomiting, and seizures, or display visual, speech, coordination and/or cognitive problems) that result from increased intracranial pressure caused by the large mass of the brain tumor (34). Animal studies were performed similarly to those described above except that tumors were allowed to reach a volume of 500 mm3 prior to treatment with 40 mg/kg TMZ alone or combined with 100 mg/kg of 5-NIdR. Representative images provided in Fig. 3C show that treatment with 40 mg/kg TMZ has a minimal effect on tumor growth, whereas cotreatment with 100 mg/kg 5-NIdR and 40 mg/kg TMZ results in complete tumor regression within 33 days after treatment. This result was confirmed using multiple mice (n > 8) and is summarized in the Kaplan–Meier plot provided as Fig. 3D. In this case, the MTD for mice treated with 40 mg/kg TMZ is 45 days, whereas the MTD for mice treated with a combination of TMZ (40 mg/kg) and 5-NIdR (100 mg/kg) is greater than 250 days. In fact, the majority of mice treated with this drug combination show complete tumor regression (Fig. 3E). We note that approximately 40% of mice do not survive 50 days after treatment due to the development of excessively large tumors (>2,000 mm3). The molecular reason accounting for this bimodal response toward the combination of 5-NIdR and TMZ is currently not clear. However, there are several possibilities that may explain this phenomenon. One possibility may reflect problems associated with adequate drug delivery to the tumor. This could be caused by the presence of a “drug sanctuary” and/or a lack of adequate vasculature that prevents efficient delivery of the compounds to the tumor. An alternative mechanism is that the observed bimodal response reflects a subset of cells that have developed resistance to TMZ or 5-NIdR. We are currently attempting to differentiate these possibilities. Despite this intriguing result, it is clear that combining 5-NIdR with TMZ produces a beneficial effect as this combination produces a significantly longer delay in tumor growth compared with treatment with TMZ alone.

5-NIdR is a nontoxic nucleoside analog

Collectively, these in vivo studies demonstrate that combining 5-NIdR with TMZ produces significant beneficial anticancer effects by ablating tumor growth. However, there are concerns regarding the potential safety of this artificial nucleoside that primarily involve the presence of the nitro group. In general, nitro groups are disfavored as pharmacophores due to the potential enzymatic reduction of the -NO2 group to an -NH2 group (35). This reaction is typically catalyzed by liver cytochrome P450s and proceeds via the production of radical intermediates that can damage liver cells (36). Additional safety concerns include the occurrence of side effects including nausea, diarrhea, fatigue, and immunosuppression that are commonly observed with conventional nucleoside analogs such as fludarabine and gemcitabine (38, 38). To evaluate the overall safety of 5-NIdR, we performed a dose-escalation study in which male and female C57BL/6 mice (n = 6 per group) received ascending doses of 5-NIdR ranging from 50 to 500 mg/kg via tail vein injections. All mice survived treatment for 7 days prior to being euthanized to harvest organs for pathologic evaluation. During this time interval, mice did not show any significance decrease (>5%) in body weight (Fig. 4A), indicating that the artificial nucleoside does not produce nausea or diarrhea in mice. In addition, blood was analyzed from mice treated with DMSO/PBS and with 500 mg/kg of 5-NIdR, the highest dose used in this study. The data provided in Fig. 4B show that there are no adverse effects upon administration of 500 mg/kg 5-NIdR on the levels of white blood cells, neutrophils, red blood cell, hemoglobin, and platelets. Thus, acute treatment with a high dose of 5-NIdR does not produce hematologic disorders such as leukopenia, anemia, and/or thrombocytopenia, which can occur with anticancer nucleosides (37, 38).

Figure 4.

Toxicology studies demonstrate that 5-NIdR does not produce toxic side effects. A, Time courses monitoring the change in percent weight in male and female C57BL/6 mice treated with DMSO/PBS (white bars) versus 500 mg/kg 5-NIdR (gray bars). B, Analyses of key hematologic indicators in mice treated with DMSO/PBS versus 500 mg/kg of 5-NIdR. There are no significant decreases in the levels of white blood cells (WBC), neutrophils (NE), red blood cells (RBC), hemoglobin (Hb), and platelets at a dose of 500 mg/kg 5-NIdR. C, Histologic examination of brain, heart, liver, and kidneys isolated from male and female mice treated with PBS/DMSO (vehicle control) versus 500 mg/kg of 5-NIdR. Microscopic imaging reveals no differences in tissues isolated from mice treated with 500 mg/kg of 5-NIdR compared with treatment with DMSO/PBS.

Figure 4.

Toxicology studies demonstrate that 5-NIdR does not produce toxic side effects. A, Time courses monitoring the change in percent weight in male and female C57BL/6 mice treated with DMSO/PBS (white bars) versus 500 mg/kg 5-NIdR (gray bars). B, Analyses of key hematologic indicators in mice treated with DMSO/PBS versus 500 mg/kg of 5-NIdR. There are no significant decreases in the levels of white blood cells (WBC), neutrophils (NE), red blood cells (RBC), hemoglobin (Hb), and platelets at a dose of 500 mg/kg 5-NIdR. C, Histologic examination of brain, heart, liver, and kidneys isolated from male and female mice treated with PBS/DMSO (vehicle control) versus 500 mg/kg of 5-NIdR. Microscopic imaging reveals no differences in tissues isolated from mice treated with 500 mg/kg of 5-NIdR compared with treatment with DMSO/PBS.

Close modal

We also performed a repeat-dosing study in which adult male and female mice (n = 4 per group) received two injections of variable doses of 5-NIdR (50 to 500 mg/kg) given 7 days apart. As before, all mice that received repeat doses of 5-NIdR survived for 7 days after the final treatment prior to being euthanized for pathologic evaluation. Over the time period tested (Δt = 14 days), mice did not show any significant decrease (>10%) in body weight nor did they present with signs of leukopenia, anemia, and/or thrombocytopenia. Histologic examination of major organs including brain, heart, liver, and kidneys was performed in animals treated with PBS/DMSO (vehicle control) and 500 mg/kg of 5-NIdR. Inspection of the microscopy data provided in Fig. 4C reveals no differences in tissues isolated from mice treated with 500 mg/kg of 5-NIdR compared with mice treated with DMSO/PBS. Collectively, the results of these acute and repeat dosing studies indicate that 5-NIdR is a well-tolerated nucleoside analog that does not produce overt toxic side effects typically seen with conventional nucleoside analogs.

DNA-damaging agents such as ionizing radiation, carboplatin, cyclophosphamide, and TMZ are widely used to treat cancer. Unfortunately, these types of therapeutic agents can present with devastating complications such as immunosuppression, drug resistance, and an increased risks of developing secondary cancers such as leukemia (39–41). Some of these complications are caused by the ability of DNA polymerases to inappropriately replicate unrepaired DNA lesions formed by these agents. The results described in this article provide evidence for a new therapeutic strategy to alleviate these complications by using an artificial nucleoside that is efficiently inserted opposite damaged DNA to selectively inhibit TLS activity. Our animal studies demonstrate that coadministration of the artificial nucleoside, 5-NIdR, with TMZ significantly increases the overall therapeutic efficacy of the DNA-damaging agent without causing side effects that may prohibit its clinical utility. Cell-based pharmacologic and in vitro DNA polymerization studies show that the cell-killing effects of 5-NIdR involve the ability of the corresponding nucleoside triphosphate, 5-NITP, to inhibit the ability of DNA polymerases to replicate DNA damage caused by TMZ.

A model describing the most widely accepted mechanism for the coordination of DNA polymerase activity during TLS is provided in Supplementary Fig. S7A (42–44). In this model, high-fidelity DNA polymerases such as pol delta and pol epsilon encounter an unrepaired DNA lesion during chromosomal replication. However, the intrinsic high-fidelity associated with these polymerases prevents the stable incorporation of a nucleotide opposite the DNA lesion. Instead, a specialized DNA polymerase such as pol eta is recruited to the DNA lesion to incorporate a dNTP opposite the damaged DNA. This allows the lesion to be by-passed in a timely manner such that a high-fidelity polymerase can displace the specialized polymerase to continue chromosomal replication. Although TLS activity is essential for cells to survive genomic stress caused by unrepaired DNA lesions, the process of replicating damaged DNA can also produce detrimental effects at the cellular and organismal level. As described above, upregulated TLS activity can cause significant complications in patients receiving chemotherapy. The misreplication of DNA lesions caused by TMZ can directly cause drug resistance, whereas promutagenic TLS activity can increase genetic mutations to create more aggressive cancers. Our in vitro kinetic data provide a strategy to inhibit this process and combat complications caused by TLS activity. In particular, these data show that the high-fidelity DNA polymerases, pol delta and pol epsilon, are far more efficient at incorporating 5-NITP opposite an abasic site compared with inserting the natural substrate, dATP. In fact, it is surprising that both high-fidelity DNA polymerases are approximately 10-fold more efficient at inserting 5-NITP opposite the noninstructional DNA lesion compared with the specialized DNA polymerases, pol eta and pol iota, which are typically associated with drug resistance. The unique combination of facile and preferential utilization of 5-NITP by chromosomal DNA polymerases coupled with the chain-termination capabilities of the artificial nucleotide disrupts the normal coordination of polymerase activity during the by-pass of abasic sites. The model provided in Supplementary Fig. S7B shows that the incorporation of 5-NITP opposite an abasic site inhibits lesion bypass, which subsequently alters the continuity of DNA synthesis during S-phase. Subsequent stalling of the replication fork would produce large stretches of single-strand DNA that leads to the activation of ATM to ultimately induce apoptosis.

The ability of 5-NITP to inhibit TLS activity during S-phase has several important implications for translational applications. First, 5-NITP is almost exclusively inserted opposite damaged DNA, and this high selectivity explains why the artificial nucleoside displays low toxicity in the absence of exogenous DNA-damaging agents. Second, the ability of 5-NIdR to inhibit the replication of abasic sites explains how the nucleoside analog increases the cytotoxic effects of compounds such as TMZ that produce these DNA lesions. In this case, inhibiting TLS activity could provide several beneficial effects in patients undergoing chemotherapy. Sensitizing cancer cells to the effects of a DNA-damaging agent provides a strategy to administer lower drug doses to reduce the risk of potential side effects. This is particularly important with DNA-damaging agents such as TMZ, cisplatin, and cyclophosphamide as they often produce severe and debilitating side effects that reflect their ability to kill normal cells. Perhaps more relevant toward treating neurological cancers such as GBM is whether or not 5-NIdR can potentiate the cell-killing effects of ionizing radiation. We hypothesize that 5-NIdR could increase the efficacy of ionizing radiation as this treatment modality ultimately creates DSBs, which are noninstructional DNA lesions similar to abasic sites. Because 5-NITP is efficiently inserted opposite abasic sites, we expect that the artificial nucleoside will also be utilized by DNA polymerases such as pol lambda, pol mu, and pol theta that replicate DSBs.

Another important aspect of this report is the concept of targeting TLS activity as a rational way to combat drug resistance that may be caused by the upregulation of promutagenic DNA synthesis in cancers (45, 46). Although our in vitro and cell-based studies collectively show that 5-NIdR inhibits TLS activity, our data have not definitively shown that the artificial nucleoside does indeed delay or prevent the onset of drug resistance. Unambiguously determining this capability is complicated for several reasons. In particular, it is relatively easy to identify drug-resistant cells as they survive exposure to high drug concentrations. This positive selection provides a way to assess genomic and/or proteomic changes that give a survival advantage to exposure to DNA-damaging agents. In this report, we have shown that the majority of cells treated with the combination of 5-NIdR and TMZ actually undergo cell death. This presents a negative selection mechanism and subsequently hinders attempts to directly assess possible cellular changes associated with drug resistance. Despite these difficulties, we are currently performing genomic and proteomic analyses on tumor biopsies from mice treated with TMZ alone or combined with 5-NIdR to identify cellular changes that correlate with drug sensitivity or resistance.

Finally, it is important to compare and contrast the potential translational applications of 5-NIdR with fludarabine and gemcitabine, two nucleoside analogs that are widely used in chemotherapy. Although 5-NIdR bears close structural resemblance to fludarabine, their mechanisms of action appear distinctive different. For instance, fludarabine is almost exclusively used as a monotherapuetic agent against hematologic malignancies such as chronic lymphoblastic leukemia and never used in combination with DNA-damaging agents. Indeed, several clinical trials examining the combination of fludarabine with drugs such as chlorambucil and cyclophosphamide were discontinued as patients displayed severe hematologic toxicities with no improvement in overall response compared with fludarabine monotherapy (47, 48). These hematologic toxicities likely reflect a lack of selectivity exhibited by fludarabine, which potently inhibits replicative DNA polymerases such as pol alpha and epsilon during normal DNA synthesis (49). This inhibition places a high burden on normal DNA replication and DNA repair in healthy cells, a feature not observed with 5-NIdR, which does not inhibit normal DNA synthesis.

In contrast to fludarabine, the pyrimidine analog, gemcitabine, displays anticancer activities against both hematologic cancers and solid tumors. In addition, it is frequently combined with platinum drugs such as cisplatin. However, combining gemcitabine with platinum agents does not appear to cause cell death by inhibiting TLS activity. Instead, gemcitabine is primarily used as a sensitizing agent so that lower acute and cumulative doses of platinum-based DNA-damaging agents can be applied. This is important because gemcitabine produces less severe side effects (mild myelosuppression and nausea/vomiting) than platinum drugs (cumulative peripheral neurotoxicity and nephrotoxicity). Combination therapy typically consists of gemcitabine (900 mg/m2) and oxaliplatin (60 mg/m2) given as an intravenous infusion once per week over the course of several weeks (50). By comparison, the dose of 100 mg/kg of 5-NIdR used in our animal studies corresponds to a predicted human dose of 300 mg/m2. This value is substantially lower than the dose of gemcitabine used as a monotherapeutic agent (>1,000 mg/m2) or when combined with platinum drugs. The low predicted dose for 5-NIdR likely reflects a different mechanism of action in which the artificial nucleoside analog functions as a potent inhibitor of TLS activity while gemcitabine inhibits normal DNA synthesis.

A. Berdis is cofounder/CSO of RED5 Pharmaceuticals, LLC. No potential conflicts of interest were disclosed by the other authors.

Conception and design: J.-S. Choi, A. Berdis

Development of methodology: J.-S. Choi, A. Berdis

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J.-S. Choi, A. Berdis

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J.-S. Choi, A. C.S. Kim, A. Berdis

Writing, review, and/or revision of the manuscript: J.-S. Choi, A. Berdis

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C.S. Kim, A. Berdis

Study supervision: A. Berdis

This work was supported by grants to A. Berdis from the Department of Defense (W81XWH-13-1-0238), the Ohio Third Frontier Foundation, and the Berdis Glioblastoma Fund.

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.
Ellis
HP
,
Greenslade
M
,
Powell
B
,
Spiteri
I
,
Sottoriya
A
,
Kurian
KM
. 
Current challenges in glioblastoma: intratumour heterogeneity, residual disease, and models to predict disease recurrence
.
Front Oncol
2015
;
5
:
251
.
2.
Cachia
D
,
Kamiya-Matsuoka
C
,
Mandel
JJ
,
Olar
A
,
Cykowski
MD
,
Armstrong
TS
, et al
Primary and secondary gliosarcomas: clinical, molecular and survival characteristics
.
J Neurooncol
2015
;
125
:
401
10
.
3.
Weller
M
,
Steinbach
JP
,
Wick
W
. 
Temozolomide: a milestone in the pharmacotherapy of brain tumors
.
Future Oncol
2005
;
1
:
747
54
.
4.
Sanderson
BJ
,
Shield
AJ
. 
Mutagenic damage to mammalian cells by therapeutic alkylating agents
.
Mutat Res
1996
;
355
:
41
57
.
5.
Bonmassar
L
,
Marchesi
F
,
Pascale
E
,
Franzese
O
,
Margison
GP
,
Bianchi
A
, et al
Triazene compounds in the treatment of acute myeloid leukemia: a short review and a case report
.
Curr Med Chem
2013
;
20
:
2389
401
.
6.
Agarwala
SS
,
Kirkwood
JM
. 
Temozolomide, a novel alkylating agent with activity in the central nervous system, may improve the treatment of advanced metastatic melanoma
.
Oncologist
2000
;
5
:
144
51
.
7.
Gates
KS
,
Nooner
T
,
Dutta
S
. 
Biologically relevant chemical reactions of N7-alkylguanine residues in DNA
.
Chem Res Toxicol
2004
;
17
:
839
56
.
8.
Erasimus
H
,
Gobin
M
,
Niclou
S
,
Van Dyck
E
. 
DNA repair mechanisms and their clinical impact in glioblastoma
.
Mutat Res Rev Mutat Res
2016
;
769
:
19
35
.
9.
Silber
JR
,
Bobola
MS
,
Blank
A
,
Chamberlain
MC
. 
O(6)-methylguanine-DNA methyltransferase in glioma therapy: promise and problems
.
Biochim Biophys Acta
2012
;
1826
:
71
82
.
10.
Cabrini
G
,
Fabbri
E
,
Lo Nigro
C
,
Dechecchi
MC
,
Gambari
R
. 
Regulation of expression of O6-methylguanine-DNA methyltransferase and the treatment of glioblastoma
.
Int J Oncol
2015
;
47
:
417
28
.
11.
Hunter
C
,
Smith
R
,
Cahill
DP
,
Stephens
P
,
Stevens
C
,
Teague
J
, et al
A hypermutation phenotype and somatic MSH6 mutations in recurrent human malignant gliomas after alkylator chemotherapy
.
Cancer Res
2006
;
66
:
3987
91
.
12.
Cahill
DP
,
Levine
KK
,
Betensky
RA
,
Codd
PJ
,
Romany
CA
,
Reavie
LB
, et al
Loss of the mismatch repair protein MSH6 in human glioblastomas is associated with tumor progression during temozolomide treatment
.
Clin Cancer Res
2007
;
13
:
2038
45
.
13.
Yip
S
,
Miao
J
,
Cahill
DP
,
Iafrate
AJ
,
Aldape
K
,
Nutt
CL
, et al
MSH6 mutations arise in glioblastomas during temozolomide therapy and mediate temozolomide resistance
.
Clin Cancer Res
2009
;
15
:
4622
9
.
14.
Johnson
BE
,
Mazor
T
,
Hong
C
,
Barnes
M
,
Aihara
K
,
McLean
CY
, et al
Mutational analysis reveals the origin and therapy-driven evolution of recurrent glioma
.
Science
2014
;
343
:
189
93
.
15.
Motea
EA
,
Lee
I
,
Berdis
AJ
. 
A non-natural nucleoside with combined therapeutic and diagnostic activities against leukemia
.
ACS Chem Biol
2012
;
7
:
988
98
.
16.
Motea
EA
,
Lee
I
,
Berdis
AJ
. 
Quantifying the energetic contributions of desolvation and π-electron density during translesion DNA synthesis
.
Nucleic Acids Res
2011
;
39
:
1623
37
.
17.
Lindahl
T
. 
Instability and decay of the primary structure of DNA
.
Nature
1993
;
362
:
709
15
.
18.
Taylor
JS
. 
New structural and mechanistic insight into the A-rule and the instructional and non-instructional behavior of DNA photoproducts and other lesions
.
Mutat Res
2002
;
510
:
55
70
.
19.
Berdis
AJ
. 
Dynamics of translesion DNA synthesis catalyzed by the bacteriophage T4 exonuclease-deficient DNA polymerase
.
Biochemistry
2001
;
40
:
7180
91
.
20.
Maor-Shoshani
A
,
Hayashi
K
,
Ohmori
H
,
Livneh
Z
. 
Analysis of translesion replication across an abasic site by DNA polymerase IV of Escherichia coli
.
DNA Repair (Amst)
2003
;
2
:
1227
38
.
21.
Ling
H
,
Boudsocq
F
,
Woodgate
R
,
Yang
W
. 
Snapshots of replication through an abasic lesion; structural basis for base substitutions and frameshifts
.
Mol Cell
2004
;
13
:
751
62
.
22.
Hogg
M
,
Wallace
SS
,
Doublié
S
. 
Crystallographic snapshots of a replicative DNA polymerase encountering an abasic site
.
EMBO J
2004
;
23
:
1483
93
.
23.
Zhong
X
,
Pedersen
LC
,
Kunkel
TA
. 
Characterization of a replicative DNA polymerase mutant with reduced fidelity and increased translesion synthesis capacity
.
Nucleic Acids Res
2008
;
36
:
3892
904
.
24.
Avkin
S
,
Adar
S
,
Blander
G
,
Livneh
Z
. 
Quantitative measurement of translesion replication in human cells: evidence for bypass of abasic sites by a replicative DNA polymerase
.
Proc Natl Acad Sci USA
2002
;
99
:
3764
9
.
25.
Reineks
EZ
,
Berdis
AJ
. 
Evaluating the contribution of base stacking during translesion DNA replication
.
Biochemistry
2004
;
43
:
393
404
.
26.
Golden
J
,
Motea
E
,
Zhang
X
,
Choi
J-S
,
Feng
Y
,
Xu
Y
, et al
Development and characterization of a non-natural nucleoside that displays anticancer activity against solid tumors
.
ACS Chem Biol
2013
;
8
:
2452
65
.
27.
Zhang
X
,
Lee
I
,
Berdis
AJ
. 
A potential chemotherapeutic strategy for the selective inhibition of promutagenic DNA synthesis by nonnatural nucleotides
.
Biochemistry
2005
;
44
:
13111
21
.
28.
Bakkenist
CJ
,
Kastan
MB
. 
DNA damage activates ATM through intermolecular autophosphorylation and dimer dissociation
.
Nature
2003
;
421
:
499
506
.
29.
Burma
S
,
Chen
BP
,
Murphy
M
,
Kurimasa
A
,
Chen
DJ
. 
ATM phosphorylates histone H2AX in response to DNA double-strand breaks
.
J Biol Chem
2001
;
276
:
42462
7
.
30.
Zhao
Y
,
Biertümpfel
C
,
Gregory
MT
,
Hua
YJ
,
Hanaoka
F
,
Yang
W
. 
Structural basis of human DNA polymerase η-mediated chemoresistance to cisplatin
.
Proc Natl Acad Sci U S A
2012
;
109
:
7269
74
.
31.
Ummat
A
,
Rechkoblit
O
,
Jain
R
,
Roy Choudhury
J
,
Johnson
RE
,
Silverstein
TD
, et al
Structural basis for cisplatin DNA damage tolerance by human polymerase η during cancer chemotherapy
.
Nat Struct Mol Biol
2012
;
19
:
628
32
.
32.
Choi
J-S
,
Kim
S
,
Motea
E
,
Berdis
AJ
. 
Inhibiting translesion DNA synthesis as an approach to combat drug resistance to DNA damaging agents
.
Oncotarget
2017
;
8
:
40804
16
.
33.
Liu
L
,
Nakatsuru
Y
,
Gerson
SL
. 
Base excision repair as a therapeutic target in colon cancer
.
Clin Cancer Res
2002
;
8
:
2985
91
.
34.
Schmidt-Hansen
M
,
Berendse
S
,
Hamilton
W
. 
Symptomatic diagnosis of cancer of the brain and central nervous system in primary care: a systematic review
.
Fam Pract
2015
;
32
:
618
23
.
35.
Uetrecht
J
. 
N-oxidation of drugs associated with idiosyncratic drug reactions
.
Drug Metab Rev
2002
;
34
:
651
65
.
36.
Wardman
P
,
Dennis
MF
,
Everett
SA
,
Patel
KB
,
Stratford
MR
,
Tracy
M
. 
Radicals from one-electron reduction of nitro compounds, aromatic N-oxides and quinones: the kinetic basis for hypoxia-selective, bioreductive drugs
.
Biochem Soc Symp
1995
;
61
:
171
94
.
37.
Lukenbill
J
,
Kalaycio
M
. 
Fludarabine: a review of the clear benefits and potential harms
.
Leuk Res
2013
;
37
:
986
94
.
38.
Serdjebi
C
,
Milano
G
,
Ciccolini
J
. 
Role of cytidine deaminase in toxicity and efficacy of nucleosidic analogs
.
Expert Opin Drug Metab Toxicol
2015
;
11
:
665
72
.
39.
Stojanovska
V
,
Sakkal
S
,
Nurgali
K
. 
Platinum-based chemotherapy: gastrointestinal immunomodulation and enteric nervous system toxicity
.
Am J Physiol Gastrointest Liver Physiol
2015
;
308
:
G223
232
.
40.
Ranchoux
B
,
Günther
S
,
Quarck
R
,
Chaumais
MC
,
Dorfmüller
P
,
Antigny
F
, et al
Chemotherapy-induced pulmonary hypertension: role of alkylating agents
.
Am J Pathol
2015
;
185
:
356
71
.
41.
Albertella
MR
,
Lau
A
,
O'Connor
MJ
. 
The overexpression of specialized DNA polymerases in cancer
.
DNA Repair (Amst)
2005
;
4
:
583
93
.
42.
McCulloch
SD
,
Kunkel
TA
. 
The fidelity of DNA synthesis by eukaryotic replicative and translesion synthesis polymerases
.
Cell Res
2008
;
18
:
148
61
.
43.
Sutton
MD
. 
Coordinating DNA polymerase traffic during high and low fidelity synthesis
.
Biochim Biophys Acta
2010
;
1804
:
1167
79
.
44.
Sale
JE
. 
Translesion DNA synthesis and mutagenesis in eukaryotes
.
Cold Spring Harb Perspect Biol
2013
;
5
:
a012708
.
45.
Tomicic
MT
,
Aasland
D
,
Naumann
SC
,
Meise
R
,
Barckhausen
C
,
Kaina
B
, et al
Translesion polymerase η is upregulated by cancer therapeutics and confers anticancer drug resistance
.
Cancer Res
2014
;
74
:
5585
96
.
46.
Albertella
MR
,
Green
CM
,
Lehmann
AR
,
O'Connor
MJ
. 
A role for polymerase eta in the cellular tolerance to cisplatin-induced damage
.
Cancer Res
2015
;
65
:
9799
806
.
47.
Weiss
M
,
Spiess
T
,
Berman
E
,
Kempin
S
. 
Concomitant administration of chlorambucil limits dose intensity of fludarabine in previously treated patients with chronic lymphocytic leukemia
.
Leukemia
1994
;
8
:
1290
3
.
48.
Rai
KR
,
Peterson
BL
,
Appelbaum
FR
,
Kolitz
J
,
Elias
L
,
Shepherd
L
, et al
Fludarabine compared with chlorambucil as primary therapy for chronic lymphocytic leukemia
.
N Engl J Med
2000
;
343
:
1750
57
.
49.
Parker
WB
. 
Enzymology of purine and pyrimidine antimetabolites used in the treatment of cancer
.
Chem Rev
2009
;
109
:
2880
93
.
50.
Liliemark
JO
,
Plunkett
W
,
Dixon
DO
. 
Relationship of 1-beta-D-arabinofuranosylcytosine in plasma to 1-beta-D-arabinofuranosylcytosine 5′-triphosphate levels in leukemic cells during treatment with high-dose 1-beta-D-arabinofuranosylcytosine
.
Cancer Res
1985
;
45
:
5952
7
.