Inhibition of DNA Replication by Tirapazamine¹

TPZ³ (SR 4233; WIN 59075; 3-amino-1,2,4-benzotriazine-1,4-dioxide; Tirazone) is a chemotherapeutic agent that is activated at low oxygen levels to become a cytotoxic agent (1, 2). It is currently showing considerable promise in Phase II and III clinical trials in combination with radiation and cisplatin (3, 4). The basis for the tumor selectivity of TPZ is that most solid tumors have a much lower average oxygen tension than normal tissues. These hypoxic tumors are often refractory to radiotherapy and are probably also refractory to chemotherapy (5, 6, 7, 8). However, because TPZ is selectively toxic to hypoxic cells, it can exploit low tumor oxygenation for therapeutic advantage (9).

Intracellular reductases convert TPZ to a cytotoxic radical that, under hypoxic conditions, produces base damage, DNA single-strand breaks, DNA double-strand breaks, and chromosome aberrations (2, 10, 11, 12). In aerobic cells, the cytotoxicity of the TPZ radical is greatly reduced because oxygen causes back-oxidation of the radical to the nontoxic parent compound. Many enzymes (including cytochrome P450 and NADPH cytochrome P450 reductase) can reduce TPZ to its cytotoxic radical (11, 13, 14, 15, 16, 17, 18, 19). However, although a large majority of the TPZ is metabolized in the cytoplasm, nuclear metabolism of TPZ accounts for essentially all of the TPZ-induced DNA damage (20).

The efficient production of DNA double-strand breaks by TPZ, as well as their inefficient repair compared with those produced by ionizing radiation (10), suggested that TPZ may be producing radicals in a highly localized area (2). The nuclear matrix, a proteinaceous scaffold composed of lamins A, lamins B, and topoisomerase IIα, provides a structure to organize DNA in the nucleus. Matrix attachment regions are the regions of DNA that attach to the nuclear matrix (21, 22). Whereas specific nuclear reductases associated with the nuclear matrix that metabolize TPZ have yet to be identified, we have proposed that these reductases could produce highly localized damage at the matrix attachment regions (23).

The possibility that the enzymes responsible for the cytotoxicity of TPZ associate with the nuclear matrix led us to examine whether the TPZ radical modifies or disrupts cellular functions that occur at or near the nuclear matrix. One critical cellular function that occurs at the nuclear matrix is DNA replication (24, 25, 26, 27, 28). Thus, we hypothesized that because TPZ radicals may be highly concentrated in this region, TPZ might have a disproportionately large effect on DNA replication compared with other DNA-damaging agents at comparable levels of cytotoxicity. In agreement with this hypothesis, we show that TPZ dramatically inhibits DNA synthesis in HCT116 colon cancer cells and HeLa cells. This inhibition of DNA synthesis by TPZ (75–80% reduction after 1 h at 20 μm) is of greater magnitude and duration than that caused by more toxic doses of ionizing radiation, suggesting that the TPZ radicals may directly affect components involved in replication. In support of this model, we show using a SV40-based in vitro DNA replication assay that extracts from TPZ-treated cells have a reduced ability to support in vitro DNA replication (20–50% of control after 1 h at 20 μm). The reduced DNA replication activity in these extracts could be traced to a reduction in the levels of RPA, a heterotrimeric protein essential for DNA replication but also involved in DNA repair and recombination (reviewed in Ref. 29). Addition of rRPA to the extracts of TPZ-treated cells restores replication activity to control levels. Using indirect immunofluorescence, we demonstrate that the p34 subunit of RPA (RPA2) forms small subnuclear foci after cells are treated with TPZ under hypoxic conditions. These results reveal that exposure of cells to TPZ under hypoxic conditions greatly inhibits DNA synthesis and suggest that relocation of RPA or damage to RPA may contribute to this effect.

TPZ was obtained from Sanofi-Winthrop (Malvern, PA). [¹⁴C]Thymidine and [³H]thymidine were obtained from Amersham (Cleveland, OH). All other reagents were purchased from Sigma Chemical Co. (St. Louis, MO) unless otherwise indicated.

HeLa cells were grown and maintained in monolayer culture in Joklik-modified MEM supplemented with 5% iron-supplemented calf serum. HCT116 (p21^+/+) human colon carcinoma cells and HCT116 (p21^−/−) cells were graciously provided by Dr. Bert Vogelstein (Johns Hopkins Oncology Center, Baltimore, MD) and grown and maintained in monolayer culture in McCoy’s 5A modified medium supplemented with 10% fetal bovine serum. For all experiments, both HeLa cells and HCT116 cells were used in the logarithmic phase of growth.

HCT116 cells and HeLa cells were treated with TPZ at the concentrations and times indicated. For hypoxia treatment, cells were grown in notched glass dishes, which were loaded into prewarmed aluminum jigs. The jigs were evacuated five times to 0.1 atmosphere with N₂ + 5% C0₂ reintroduction and constant shaking to achieve hypoxia (less than 200 ppm O₂) as described previously (30). For aerobic treatment, jigs were evacuated and refilled five times with 95% air + 5% CO₂. After evacuation, the jigs were placed in a 37°C incubator for 1 h. All TPZ exposures were for 1 h unless otherwise indicated.

For shorter exposures to TPZ (15–45 min), spinner flasks were used to achieve hypoxia. The control spinner flask contained 10.0 ml of control media, the treatment spinner flask contained 10 ml of media supplemented with 26.0 μm TPZ, and the cell spinner flask contained 10.0 ml of cells at a concentration of 4.3 × 10⁶ cells/ml. Control spinner and treatment spinner flasks were allowed to become hypoxic for 1.0 h by circulating N₂ + 5% CO₂ gas through the flasks. Cells from the cell spinner flask were then added to both the control spinner flask and the treatment spinner flask to bring the final concentration of TPZ to 20 μm. At the appropriate time, 1.0 ml of cells was removed from the control spinner flask and the treatment spinner flask. Cells were centrifuged and washed twice with PBS. After washing, the cells were immediately pulse-labeled with [³H]thymidine as indicated below.

Irradiations were performed with ¹³⁷Cesium γ-rays from a J. L. Shepherd and Associates Mark I Model 25 Irradiator (San Fernando, CA) operating at 369.1 cGy/min. Cells were placed on ice during treatment with irradiation.

Before treatment with TPZ or ionizing radiation, cells were continuously labeled for 24 h with 1 μCi/ml [¹⁴C]thymidine. Immediately before treatment, the [¹⁴C]thymidine-containing media was removed, and the cells were rinsed with PBS. After treatment with TPZ or with irradiation, the cells were incubated at 37°C for 30 min. After recovery, the cells were pulse-labeled with 100 μCi/ml [³H]thymidine for 30 min. The radioactive media were removed after pulse-labeling, and the cells were washed twice with PBS.

HCT116 or HeLa cells were lysed after pulse-labeling in radioimmunoprecipitation assay buffer [50 mm Tris-HCl (pH 7.5), 150 mm sodium chloride, 1% Triton X-100, 1% sodium deoxycholate, and 0.1% SDS] at 4°C. The lysate was precipitated on ice for 30 min with 10% trichloroacetic acid + 2% Na PP_i. The lysate was then resuspended to a final volume of 2.0 ml. Aliquots (400 μl) of each sample were applied in triplicate to prewetted 24.0 mm Whatman glass filters (Maidstone, Kent, United Kingdom) placed on a filtration system (Millipore, Bedford, MA). The samples were then washed three times with 5.0 ml of 10% trichloroacetic acid + 2% Na PP_i and twice with 5.0 ml of 95% ethanol. Filters were allowed to air dry and then transferred to scintillation tubes. A total of 10.0 ml of Ecolume were added to each sample. The samples were analyzed using a Beckman LS6000IC liquid scintillation counter (Fullerton, CA). DNA synthesis was calculated as the ratio of the [³H]thymidine counts:[¹⁴C]thymidine counts. The percentage of replication activity was calculated by comparison with identically treated controls not exposed to TPZ or to ionizing radiation.

Colony-forming assays were performed as described previously (31). Briefly, HCT116 p21^+/+ cells treated with varying doses of TPZ under hypoxia for 1 h or varying doses of irradiation were collected by trypsinization, resuspended in ice-cold media, and plated in 60- or 100-mm Petri dishes. After 2 weeks, the media were removed from the dishes, and the cells were stained with crystal violet. Individual colonies of more than 50 cells were counted, and the survival of treated groups was obtained by comparing the number of colonies in the treated groups versus the number of colonies in the control group.

HeLa cells were grown in 1-liter spinner flasks, filled to 250 ml, to a final concentration of ∼10⁶ cells/ml. Cells were made hypoxic by a 1-h exposure to a N₂/CO₂ mixture and treated for 1 h with TPZ. Hypoxia was monitored using an O₂ electrode (kindly provided by Dr. Cameron Koch, Department of Radiation Oncology, University of Pennsylvania School of Medicine, Philadelphia, PA). After treatment, cells were collected, an aliquot was taken to measure DNA replication in vivo, and the remaining cells were processed for cytoplasmic extract preparation using standard protocols (32, 33, 34, 35). In vitro DNA replication activity was measured using an SV40-based in vitro DNA replication assay. In this assay, replication of plasmids carrying SV40 origin of DNA (ori⁺ DNA) replication is accomplished in vitro with cytoplasmic extracts, with TAg as the only noncellular protein. Reaction mixtures (25 μl) contained the following: 40 mm HEPES; 8 mm MgCl₂; 0.5 mm DTT; 3 mm ATP; 200 μm each of CTP, GTP, and UTP; 100 μm each of dATP, dGTP, and dTTP; 25 μm [α-³²P]dCTP (1000 cpm/pmol); 40 mm creatine phosphate; 2.5 mg of creatine phosphokinase; 0.3 μg of superhelical plasmid DNA; 100–150 μg of cytoplasmic extracts; and 0.5 μg of TAg. The reaction mixture was incubated at 37°C for 1 h. The extent of DNA synthesis was evaluated by measuring incorporation into acid-insoluble material. DNA replication products were analyzed by gel electrophoresis followed by radioautography.

Standard procedures were followed for Western blots using nuclear or cytoplasmic extracts. We used whole cell lysates to assess RPA2 levels and RPA2 phosphorylation. Whole cell lysates from HeLa cells were obtained by lysing cells with radioimmunoprecipitation assay buffer supplemented with protease inhibitor mixture for 30 min at 4°C. Lysates were analyzed with SDS-PAGE using the Novex NuPAGE Electrophoresis System (San Diego, CA). Membranes were blocked with 10 mm Tris (pH 7.5), 150 mm sodium chloride, and 0.1% Tween 20 with 5% milk and 0.5% BSA and blotted with RPA2 mouse monoclonal antibody (1:400; Neomarkers, Fremont, CA). Blots were then stained with secondary antibody conjugated to alkaline phosphatase and visualized using the Enhanced Chemifluorescence (ECF) system (Amersham Life Science, Little Chalfont, United Kingdom). Data were analyzed using a STORM Optical Scanner (Molecular Dynamics, Sunnyvale, CA).

To determine cellular localization of RPA, HeLa cells were plated onto glass chamber microscope slides at a density of 1 × 10⁴ cells/slide. Slides were treated with media containing 0 or 50 μm TPZ and then exposed to hypoxia using a Bacton anaerobic/environmental chamber (Sheldon Manufacturing, Inc., Cornelius, OR). After 1 h of treatment with TPZ under hypoxic conditions, slides were removed from the chamber, and treatment media were removed. Slides were washed twice with PBS and then fixed with 2 ml of −20°C methanol. To insure fixation, slides were incubated in methanol for 15 min at −20°C. After fixation, the methanol was removed, and slides were washed twice with PBS and then blocked with PBS + 3% BSA for 30 min at 25°C. Fixed HeLa cells were probed with RPA2 mouse monoclonal antibody (1:250) for 1 h at 25°C and washed with PBS + 3% BSA to remove unbound primary antibody. Bound antibodies were detected with mouse antihuman IgG conjugated to fluorescein. The slides were stained with the DNA stain 4′,6-diamidino-2-phenylindole dihydrochloride hydrate containing Vectashield (Vector, Burlingame, CA) and visualized using a Nikon optiphot fluorescence microscope (Melville, NY).

We measured DNA synthesis as a function of dose and time after TPZ treatment of HCT116 (p21^+/+) and (p21^−/−) human colon carcinoma cells under both aerobic (Fig. 1,A) and hypoxic (Fig. 1,B) conditions. After a 1-h treatment under aerobic conditions and for concentrations of up to 50 μm (Fig. 1,A), TPZ produced only a small inhibition of DNA synthesis. However, when cells were treated with TPZ under hypoxic conditions, TPZ inhibited DNA synthesis in a dose-dependent manner in both p21^+/+ and p21^−/− cells (Fig. 1,B). After a 1-h exposure to 20 μm TPZ under hypoxic conditions, the rate of DNA synthesis decreased by approximately 70% (Fig. 1 B). HTC116 cells show no signs of toxicity to 20 μm TPZ for up to 24 h after treatment, as judged by trypan blue exclusion and cell detachment (data not shown). Therefore, the inhibition of DNA synthesis caused by TPZ is not the result of the cells being metabolically dead. We obtained very similar results with HeLa cells (data not shown).

We next examined the recovery of DNA synthesis inhibition in cells treated with TPZ under hypoxic conditions. HCT116 (p21^−/−) cells were treated with TPZ (20 μm, 1 h) and allowed to recover at 37°C. DNA synthesis remained at 15–30% of controls for up to 24 h (Fig. 1,C). The lack of recovery of DNA synthesis may be due in part to a reduction in the fraction of S-phase cells as a result of their accumulation in G₂-M phase. To determine the kinetics of inhibition of DNA synthesis, we treated HeLa cells with 20 μm TPZ under hypoxic conditions for a period of 15–60 min. After only 15 min of treatment, the rate of DNA synthesis was approximately 50% of the control and continued to decrease with increasing time of exposure to TPZ (Fig. 2). The rapidity of the decrease seen argues against a G₁-S checkpoint here with HeLa cells and in Fig. 1 C with HCT116 cells, particularly because the HCT116 cells had both alleles of p21 deleted and would therefore not be expected to delay in G₁ (36, 37, 38, 39).

Both TPZ and ionizing irradiation introduce DNA single-strand breaks and DNA double-strand breaks and produce equal numbers of lethal chromosome breaks at equal levels of cell kill (10). Ionizing irradiation also produces a dose-dependent inhibition of DNA synthesis (reviewed in Ref. 40). To test our hypothesis that TPZ, unlike ionizing radiation, produces nonrandom damage in DNA possibly concentrated at the nuclear matrix, we compared the inhibition of DNA synthesis caused by TPZ under hypoxic conditions with that caused by ionizing irradiation under hypoxic conditions. Doses of 15–60 Gy of irradiation had a modest effect on DNA synthesis in comparison with 20 μm TPZ under hypoxic conditions, which inhibited the rate of DNA synthesis by approximately 80% (Fig. 3 A).

Fig. 3, B and C, shows the clonogenic survival of HCT116 p21^+/+ cells after exposure to varying doses of TPZ (under hypoxia) or ionizing radiation (under euoxia). Comparison of these results with those of Fig. 2 demonstrate that TPZ, under hypoxic conditions, inhibits DNA synthesis to a much greater extent than doses of ionizing radiation that produce greater cell kill (even allowing for the 3-fold reduction in equivalent dose for the hypoxic irradiation; Fig. 3, A–C). These observations suggest that in addition to introducing DNA damage, TPZ may also interfere directly with the replication machinery. We examined this possibility using an in vitro assay of DNA replication.

To evaluate the effects of TPZ on DNA replication at the biochemical level, we treated HeLa cells under hypoxic conditions with varying concentrations of TPZ and then prepared cytoplasmic extracts. The ability of these extracts to support in vitro DNA replication was evaluated using plasmids containing the SV40 origin of DNA replication and recombinant TAg as the only noncellular protein (35). Although extracts prepared from untreated cells could actively support in vitro DNA replication, those of cells treated with 20 or 50 μm TPZ under hypoxic conditions showed a 50% reduction in this activity (Fig. 4). This observation suggests that extracts of TPZ-treated cells are deficient in a factor or activity that is critical for DNA replication. Because RPA is the rate-limiting factor in this type of reaction (29), we evaluated DNA replication after the addition of rRPA in reactions assembled with extracts prepared from TPZ-treated cells. The results shown in Fig. 4 indicate that rRPA restored the DNA replication activity of extracts from TPZ-treated cells to control levels (Fig. 4). Thus, exposure of cells to TPZ under hypoxic conditions leads to a partial loss of RPA activity from the cytoplasmic component of the extract.

Other DNA-damaging agents, such as camptothecin and high doses of ionizing irradiation, also decrease DNA replication as measured in extracts of treated cells by the SV40-based in vitro DNA replication assay. One component in this inhibition is a reduction of the available pool of RPA in the cytoplasmic fraction of the extract (32, 33, 34). We therefore tested this possibility in extracts prepared from TPZ-treated cells. Western blots of cytoplasmic extracts of TPZ-treated HeLa cells show a decrease in both the p70 (RPA1) and p34 (RPA2) subunits of RPA when compared with untreated cells (Fig. 5 A). This reduction is not directly obvious in nuclear extracts of TPZ-treated HeLa cells, suggesting that the overall levels of RPA in the cells probably remain unchanged.

After exposure of cells to UV irradiation or X-rays, the RPA2 subunit is found in a partly phosphorylated form (29). After treatment of HeLa cells with 50 μm TPZ, there is a shift in the electrophoretic mobility of a portion of RPA2 present in whole cell extracts (Fig. 5,B) that suggests phosphorylation. Moreover, the phosphorylation of RPA2 persists for up to 28 h after a 1-h exposure to 50 μm TPZ (data not shown). We note that there is no reduction in the total levels of RPA2 after exposure to TPZ (Fig. 5,B). Localization of RPA2 was further examined with indirect immunofluorescence. HeLa cells that were plated on glass chamber slides were exposed to hypoxia alone or 50 μm TPZ under hypoxic conditions. Under hypoxic conditions only, RPA2 localized uniformly throughout the nucleus (Fig. 5,C). After treatment with 50 μm TPZ, RPA2 localizes to small subnuclear foci as detected with indirect immunofluorescence (Fig. 5 C). In conclusion, these results suggest that TPZ under hypoxic conditions either traps RPA in the nucleus or enhances the recruitment of RPA into the nucleus, possibly to sites of TPZ-induced damage.

TPZ, a promising chemotherapeutic agent, takes advantage of the hypoxic environment of solid tumors. At low oxygen levels, TPZ radicals formed by 1-electron reduction produce extensive DNA damage, which leads to cytotoxicity (10). We have proposed that although TPZ metabolism occurs throughout the cell, it is the nuclear metabolism that is responsible for the DNA damage and cytotoxicity (9, 20). Moreover there is evidence that this metabolism is associated, at least in part, with the nuclear matrix (23). Because DNA replication occurs at sites of DNA attachment to the nuclear matrix (24, 25, 26, 27, 28), we speculated that TPZ might disrupt DNA replication to a greater degree than other agents that produce DNA damage more randomly. The results presented here confirm that TPZ produces a profound inhibition of DNA replication in both HCT116 and HeLa cells. Other hypoxia-activated cytotoxic drugs such as metronidazole, misonidazole, and nitrofurazone have been tested for their effect on DNA synthesis (41, 42). Variable effects were seen, which, in some cases, were not hypoxia specific. It is possible that some of their similarities with TPZ to inhibit DNA synthesis could be the result of activation by the same nuclear matrix-activating enzymes that are responsible for the reduction of TPZ.

Ionizing irradiation also inhibits DNA replication, but the extent of inhibition at equitoxic or more toxic doses is much less than that produced by TPZ (Fig. 3, A–C). The extent to which ionizing radiation inhibits DNA replication is cell line dependent (40), and with the HCT116 cells used in our studies, there was only a small level of inhibition at doses <15 Gy under hypoxic conditions (Fig. 3 A). Because DNA damage is believed to be directly or indirectly responsible for the bulk of the inhibition of DNA replication (40), it is relevant to ask why ionizing radiation and TPZ inhibit DNA replication to such different efficiencies. We hypothesize that this difference arises because damage induced by ionizing radiation is randomly distributed throughout the DNA, whereas TPZ-induced damage may be localized preferentially at the nuclear matrix. This could result in a strong inhibition of the DNA replication activity. Because the number of chromosome breaks produced by radiation and by TPZ are comparable at equitoxic doses (10), our data suggest that many more DNA replication-inhibiting lesions are produced per lethal event after exposure to TPZ than after exposure to ionizing radiation. A possible reason for this would be direct damage to the replication machinery.

As a direct test of this possibility, we investigated DNA replication in vitro using a SV40-based in vitro DNA replication assay. The results obtained using this system demonstrated a decrease in the ability of extracts prepared from TPZ-treated cells to support SV40 DNA replication (Fig. 4), suggesting that a component required for DNA replication was lost or damaged. However, we do not believe that such damage to the replication machinery can account for all of the inhibition because at doses of TPZ that gave 80% inhibition in vivo, we observed only 50% inhibition in vitro.

When rRPA was added to extracts of TPZ-treated cells, SV40 DNA synthesis was restored to control levels (Fig. 4). This implies that RPA may be affected by TPZ treatment. RPA, a component of the origin recognition complex, is critical for eukaryotic DNA replication and is present in cells as a heterotrimeric complex of 70, 34, and 14 kDa referred as RPA1, RPA2, and RPA3 (p14 subunit of RPA), respectively. Other cellular processes, including DNA repair and recombination, have also been shown to require RPA (29). Western blots of TPZ-treated cell extracts revealed that there was a reduction in RPA1 and RPA2 subunits of RPA in the cytoplasmic extracts (Fig. 5,A). On the other hand, nuclear extracts of TPZ-treated cells contained RPA1 and RPA2 at levels similar to those of extracts from untreated cells (Fig. 5 A). This is consistent with the hypothesis that the localized action of TPZ at the nuclear matrix traps or damages RPA and prevents it from leaching into the cytoplasmic fraction during preparation of the extract. However, similar effects may also be induced by either redirection of the activities of RPA or posttranslational modifications.

Whereas RPA is critical for DNA replication, it also participates in DNA repair (29). It is therefore possible that on DNA damage induction by TPZ, RPA redistributes to sites of DNA repair, thereby abandoning sites of DNA replication. We believe that this could be due to RPA’s known association with hRad51, a protein critical for repair of DNA damage by homologous recombination that we have recently found to be the major pathway of repair of TPZ damage.⁴ In support of this, we show that small subnuclear foci of RPA2 are seen in TPZ-treated cells (Fig. 5,C). Whatever the cause, this redistribution of RPA activity would likely reduce the cytoplasmic component of RPA or further prevent its leaching into the cytoplasmic fraction during extract preparation. Indeed, the Western blots in Fig. 5,A indicate a reduction in both RPA1 and RPA2 in cytoplasmic extracts prepared from TPZ-treated cells. Moreover, on treatment with TPZ, HeLa cells phosphorylate RPA2 (Fig. 5 B). Such phosphorylation of RPA2 after treatment with DNA-damaging agents has been linked to the role of RPA in DNA repair (29), and recently, it was also found that this phosphorylation of RPA2 led to a partial dissociation of the heterotrimeric complex (43). This posttranslational modification may therefore also contribute to the altered distribution of RPA in the cytoplasmic fraction in extracts of TPZ-treated cells.

In summary, our results show that TPZ, a hypoxia-selective cytotoxin, inhibits DNA replication to a greater extent than ionizing irradiation. This inhibition is independent of p21^Waf1/Cip1 and may result, at least in part, from direct damage to or intracellular redistribution of RPA, a protein critical for DNA replication.