Abstract
Background/Aim: Arsenite is a radiosensitizer of glioma cells both in vitro and in vivo; however, the underlying mechanism of action is unclear. Radiosensitizers specific for p53-deficient tumors are a promising adjunct to radiotherapy because, unlike normal cells, many tumor cells lack p53. Previously, we demonstrated that arsenite sensitizes the p53-deficient glioma cell line U87MG-E6 to X-rays. Materials and Methods: Using flowcytometry, we expand these findings to p53-proficient U87MG cells exposed to heavy ion beams, including carbon and iron ions. Results: Arsenite sensitized U87MG-E6, but not U87MG, cells to heavy ion beams and X-rays. Cell cycle analysis indicated that sensitization of U87MG-E6 was related to an increase in the percentage of cells in the late S/G2/M phases after combined treatment with arsenite, especially when carbon ion beams were used. Induction of γH2AX was significant in U87MG-E6, but not in U87MG, cells after irradiation with carbon ion beams plus arsenite. Conclusion: Arsenite sensitizes cells by increasing the percentage of cells in the late S/G2/M phases after irradiation, possibly via inhibition of DNA repair in the context of p53 deficiency. The findings provide information that may be useful for the development of advanced radiotherapy protocols.
Radiotherapy, especially with heavy ion beams, post-chemotherapy is a widely accepted treatment for unresectable and recurrent glioma. Although glioblastoma (glioma) is resistant to radiotherapy, cisplatin treatment can induce radiosensitization in a variety of mammalian cell lines (1). However, cisplatin penetrates the intact blood-brain barrier poorly, making it difficult to obtain therapeutically effective concentrations in the brain (2, 3). Thus, cisplatin is not suitable for radiosensitization of glioma.
Arsenite is a potential candidate radiosensitizer, but to the best of our knowledge, few studies have been conducted. Many reports suggest that arsenite alone has anticancer effects; indeed, it has been used to treat patients with acute promyelocytic leukemia (APL) (4). After oral administration, the concentration of arsenic in the cerebrospinal fluid of APL patients is approximately half that in the plasma, suggesting that arsenite can penetrate the blood-brain barrier effectively (5). A recent report showed that arsenite induces cell death and inhibits growth of a wide variety of hematologic and solid tumors (6). In addition, arsenite is a radiosensitizer tested in a clinical trial (7).
With respect to its combination with radiation, arsenite radiosensitizes gliomas in a xenograft model based on rat 9L, human SNB75, and human U87MG cells (8-10), although the underlying mechanism remains unclear. A relationship between the functions of p53 and radiosensitization has been reported in some cases. For example, an Hsp90 inhibitor induces apoptosis synergistically with radiation in human oral squamous cell carcinoma cells harboring wild-type p53 (11). In addition, an antagonist of MDM-p53 interactions increases radiation sensitivity, and induces senescence synergistically with radiation, in human prostate cancer cells accumulating p53, which indicates activation of p53 (12). Another study showed that a histone deacetylase inhibitor radiosensitized human glioblastoma cells lacking wild-type p53 (13). Because approximately half of tumors lack p53 or harbor aberrant p53 pathways (14), evaluation of radiosensitization in p53-deficient cells is important for radiotherapy. Heavy ion radiotherapy irradiates the tumor and normal tissues with high and low linear energy transfer (LET) radiation, respectively; therefore, it is more effective than conventional radiotherapies. A combination of heavy ion radiotherapy plus promising radiosensitizers should improve its efficacy even further.
In the present study, we evaluated the mechanism by which arsenite radiosensitizes glioma cells, focusing on p53 status. Previously, we showed that arsenite-induced radiosensitization of p53-deficient cells involves inhibition of a DNA repair protein, BRCA2, via production of reactive oxygen species (ROS) (15). Here, we used p53-normal and -deficient cells to assess the relationship between p53 function and radiosensitization by arsenite, as well as its dependency on LET.
Materials and Methods
Reagents. The medium, serum, and antibiotics used for cell culture were purchased from Sigma (St. Louis, MO, USA). Sodium arsenite (analytical grade) was from Wako Chemicals (Tokyo, Japan). The mouse monoclonal anti-γH2AX (Ser 139) antibody was from Upstate (New York, NY, USA), and Alexa Fluor 488-conjugated goat antimouse IgG was from Molecular Probe (Junction City, OR, USA).
Cell culture and irradiation. The U87MG human malignant glioblastoma cell line (ATCC HTB-14, Manassas, CA, USA), U87MG-neo cells transfected with vector alone (control), and U87MG-E6 cells harboring inactivated p53 via E6 (encoded by human papillomavirus)-mediated degradation of p53 (48) were supplied by Dr. Paul S. Mischel (University of California Los Angeles School of Medicine) (49). The human colon cancer p53−/− HCT116 (p53 knockout) was kindly provided by Dr. Bert Vogelstein (Johns Hopkins University) (50). Cells were maintained in Dulbecco’s modified essential medium supplemented with 10% fetal bovine serum (FBS) and antibiotics. All cells were incubated in a 5% CO2 incubator at 37°C. Cells were irradiated with X-rays produced by an X-ray generator (Pantak, New Haven County, CT, USA) set at 200 kV and 20 mA, with a 0.5 mm copper and 0.5 mm aluminum filters, at a dose rate of 0.9 Gy/min, or with heavy ion beams generated by the Heavy Ion Medical Accelerator in Chiba (HIMAC) at the National Institute of Radiological Sciences (NIRS), National Institutes for Quantum Science and Technology. At 1 h before irradiation, 1.25 μM arsenite was added to the medium. At 4 days after irradiation, the medium was replaced with medium lacking arsenite.
Colony formation assays. Cell proliferation was measured in a colony formation assay. Cells were trypsinized, diluted, counted, and seeded into dishes at various densities. At 1 h before irradiation, arsenite was added to the medium. At 4 days after irradiation, the medium was replaced with medium lacking arsenite. After 2 weeks, colonies were stained with crystal violet, and colonies containing more than 50 cells were counted. The concentration of arsenite was approximately the 10% inhibitory concentration (data not shown).
Flow cytometry analysis. Cells were trypsinized, collected, washed in phosphate-buffered saline (PBS), and fixed in 2% paraformaldehyde and 70% (v/v) ethanol. The cells were then treated with 0.5% Triton X-100 and washed twice in PBS containing 5% FBS before incubation with the primary antibody. After washing once in PBS once, the cells were incubated with secondary antibodies, washed in PBS once, and incubated in PBS with or without 5 ng/μl propidium iodide (Sigma Chemical Co.) and 50 ng/μl RNase A (Sigma Chemical Co.). Finally, cells were analyzed (10,000 events/run) using a Becton Dickinson FACSCalibur cytometer, and data were analyzed using CELLQuest software (San Jose, CA, USA).
Statistical analysis. Data are expressed as the mean±SD. p<0.05 was considered statistically significant (Student’s t-test analyzed using ystat2013 based on Microsoft Excel software).
Results
Arsenite-induced radiosensitization of p53-deficient cells. First, we performed colony formation assays using two human p53-proficient glioma cell lines, U87MG and U87MG-neo; the latter is a subclone of U87MG selected using the neomycin resistance gene on a plasmid vector. A p53-deficient glioma cell line, U87MG-E6, was also used. As in our previous report, cells were treated with 1.25 μM arsenite for 1 h prior to X-irradiation (15). Each survival curve was normalized to account for drug cytotoxicity. U87MG-E6 cells were radiosensitized when exposed to a combination of arsenite plus X-rays. In contrast to this p53-deficient cell line, p53-proficient U87MG (data not shown) and U87MG-neo cells were not radiosensitized (Figure 1). In U87MG-E6 cells, radiosensitization was also observed when arsenite was combined with heavy ion beams; again, no radiosensitization was observed in p53-proficient U87MG-neo cells (Figure 1). The radiosensitivity enhancement ratios [the ratio of D10 (i.e., mean dose that reduces the survival to 10%)] in the presence of arsenite were 1.14, 1.14, 1.29, and 1.54 for U87MG-E6 irradiated with X-rays, carbon ion beams with a LET of ~13 keV/μm, carbon ion beams with a LET of ~70 keV/μm, and iron ion beams with a LET of ~200 keV/μm, respectively (Figure 1). The D10 values for nontreated cells are shown in Table I.
Surviving fraction of p53-proficient (U87MG-neo) and p53-deficient (U87MG-E6) glioma cells. Cells were pretreated for 1 h with 1.25 μM arsenite and then exposed to various doses of X-rays and heavy ion beams. Cell survival was measured in a colony formation assay. Each data point represents the mean value (n≥3); bars, SD.
D10 values of p53-proficient U87MG-neo cells and p53-deficient U87MG-E6 cells after irradiation at low and high linear energy transfer.
DNA repair deficiency sensitizes cells to high LET radiation more weakly than X-rays (16). Our previous report indicated that inhibition of DNA repair by arsenite plays a role in radiosensitization of p53-deficient cells (15). Therefore, one may assume that, as is the case for DNA repair-deficient cell lines, radiosensitization by arsenite via inhibition of DNA repair ability should be weaker when the LET is high. It is therefore surprising that radiosensitization by arsenite was similar among treatments covering a wide range of LET.
We also observed that p53−/− HCT116 cells, but not p53+/+ HCT116 cells, were radiosensitized by 10 μM arsenite, which is an amount that reduces cell survival to about 10% (as 1.25 μM arsenite does in U87MG-neo and U87MG-E6 cells) when X-irradiation was combined. The radiosensitivity enhancement ratio for p53−/− HCT116 cells in the presence of arsenite was 1.13 at D10 (data not shown).
Synergistic effects of arsenite plus radiation on cell cycle distribution. Previous colony formation assays suggested that carbon ion beams have a relative biological effectiveness (RBE) of ~2 when using X-rays as a reference (17, 18). Therefore, we analyzed differences in cell cycle distribution after irradiation of U87MG-neo and U87MG-E6 cells with 5-Gy carbon ion beams and 10-Gy X-rays (i.e., biologically equivalent doses). The results revealed that combined treatment with arsenite and carbon ion beams had no significant synergistic effect at 6, 12, and 24 h after irradiation (data not shown). Interestingly, the percentage of U87MG-E6 cells in the late S/G2/M phases, and the percentage of cells showing hyperploidy (over-4N), increased at 36 h after irradiation with carbon ion beams, whereas arsenite itself had no effect; however, arsenite plus carbon ions led to a significant increase in these percentages, which was not the case with X-rays. The percentage of U87MG-E6 cells in the G1 and early S phases decreased significantly at 36 h postirradiation with carbon ion beams. Again, arsenite itself had no effect, but arsenite plus carbon ion beams led to a further and significant reduction in these percentages, which was also the case with X-rays (Figure 2). Furthermore, we observed no significant synergistic effect on the percentage of cells in the sub-G1 fraction (data not shown).
Relationship between arsenite-induced radiosensitization and the increase in the percentage of U87MG-E6 cells in the late S/G2/M phase at 36 h post-treatment with arsenite plus radiation. (A) Flow cytometry results showing the cell cycle distribution of U87MG-neo and U87MG-E6 cells irradiated with X-rays in the presence/absence of arsenite; cells were gated on the M1 (Sub-G1), M2 (G1/early S phase), M3 (late S/G2/M phase), and M4 (>4N: hyperploidy) regions. The arrow shows a change in the increase in the percentage of cells in the late S/G2/M phase. (B) Percentage of cells in the G1/early S phase and in the late S/G2/M phases, and the percentage of cells showing hyperploidy. Data are expressed as the mean±SD (n≥3). *p<0.05 (Student’s t-test).
Arsenite increases the percentage of γH2AX-positive cells within the irradiated U87MG-E6 population. The above-mentioned changes in the cell cycle induced by arsenite plus radiation suggest that the underlying mechanism is related to the control of cell cycle progression via DNA repair (19, 20). Therefore, we performed an immunostaining assay to detect γH2AX foci and analyzed the positive rate by flow cytometry (Figure 3A). Carbon ion irradiation at 5 Gy increased the percentage of γH2AX-positive U87MG-neo cells at 24 h post-treatment (p=0.053, Figure 3B). The percentage of γH2AX-positive U87MG-neo and U87MG-E6 cells post-treatment with arsenite plus radiation was larger than that after treatment with arsenite alone (Figure 3B). Because some nontreated cells were positive for γH2AX, we subtracted the spontaneous rate from the values for the individual treated groups, and then plotted the net increase (Figure 3C). Carbon ion irradiation at 5 Gy resulted in a significant increase in the percentage of γH2AX-positive U87MG-E6 cells, but not U87MG-neo cells (Figure 3C), when compared with arsenite alone; arsenite alone did not increase the percentage of γH2AX-positive cells of either type in comparison with the control (Figure 3B). Notably, arsenite plus carbon ions induced expression of γH2AX in U87MG-E6 cells to a level significantly above that theoretically predicted by a simple additive response (Figure 3C).
Induction of γH2AX foci in U87MG-neo and U87MG-E6 cells at 24 h post-treatment with arsenite and/or irradiation. (A) Flow cytometry gating of γH2AX-positive cells was set at the cross-point of nonirradiated control (broken line) and the combination of arsenite plus irradiation (solid line). A typical result for U87MG-neo cells is shown. (B) Summary of flow cytometry analyses of the percentage of γH2AX-positive U87MG-neo and U87MG-E6 cells irradiated in the presence/absence of arsenite. The vertical axis shows the percentage of positive cells. (C) Summary of flow cytometry analyses of the percentage of γH2AX-positive U87MG-neo and U87MG-E6 cells irradiated in the presence/absence of arsenite. The vertical axis shows the percentage of positive cells, values after subtraction of the value of the nontreated control group. Data are expressed as the mean±SD (n≥3). *p<0.05, **p<0.01 (Student’s t-test).
Discussion
Previously, we showed that arsenite sensitizes p53-deficient U87MG-E6 glioma cells to X-rays. Here, we show that arsenite increases radiosensitization of U87MG-E6 cells (but not in p53-proficient U87MG cells) to heavy ion beams in addition to X-rays.
Reportedly, the radiosensitivity of p53-proficient and - deficient cells is similar at an LET of >100 keV/μm (21). In the present study, we found that p53-proficient cells were more resistant to X-rays and carbon ion beams (LET ~70 keV/μm) than p53-deficient cells, but their resistance to iron ion beams (LET ~200 keV/μm) was the same as that of deficient cells (Table I), which is in agreement with our previous findings. X-irradiation induces cellular senescence in p53-proficient cells by inducing expression of p21 (22, 23), and in p53-deficient cells via p21-independent mechanisms (24-28); this difference might account for their distinct radiosensitivity to relatively low LET radiation (29, 30). However, the mechanism underlying the p53-independence of radiosensitivity at very high LET is less clear. Studies suggest that it might be related to p53-independent induction of apoptosis at high LET (21, 31-33).
Although high LET radiation in itself has higher cell-killing activity than low LET radiation, arsenite sensitized p53-deficient cells more efficiently to high LET radiation than to low LET radiation; indeed, we observed radiosensitivity enhancement ratios of 1.14, 1.29, and 1.54 for X-rays, carbon ion beams (LET ~70 keV/μm), and iron ion beams (LET ~200 keV/μm), respectively (Figure 1). Previously, we showed that arsenite sensitizes cells to X-rays by inhibiting BRCA2-mediated DNA repair via homologous recombination (HR) (15). Although nonhomologous end joining (NHEJ) dominates HR with respect to repairing DNA damage induced by high LET radiation, the contribution of HR is still substantial (17, 34, 35). Therefore, we suppose that arsenite-mediated inhibition of HR underlies radiosensitization to high LET radiation.
We observed an arsenite-associated increase in the percentage of p53-deficient cells in the late S/G2/M phases, as well as in the percentage of cells showing hyperploidy, which is suggestive of mitotic catastrophe related to p53-independent DNA repair pathways (36-38); these increases were more prominent in cells irradiated with carbon ions than in cells exposed to X-rays. Thus, cell cycle regulation might be involved in radiosensitization by arsenite, especially at high LET. In the absence of cell death, a decrease in the percentage of cells in the G1/early S phases should, in effect, increase the percentage of cells in the late S/G2/M phases. However, this was not the case for U87MG-E6 cells (Figure 2), indicating that radiosensitization by arsenite is related to cell death. We presume that the change at 36 h postirradiation is related to increases in the percentage of γH2AX-positive cells at 24 h after irradiation. We found a supra-additive increase in γH2AX-positive p53-deficient (U87MG-E6), but not in p53-proficient (U87MG-neo), cells at 24 h (Figure 3C), which is consistent with a previous report (39). Because the role of p53 in HR, which functions at the late S/G2/M phases, is more prominent than that of NHEJ (40), our results might reflect dysfunction of HR due to p53 deficiency. Moreover, arsenite alone did not change the proportion of γH2AX-positive U87MG-neo and U87MG-E6 cells significantly (Figure 3), indicating that radiosensitization by arsenite was due to inhibition of DNA repair rather than induction of DNA damage. Thus, accumulation of DNA damage in U87MG-E6 cells at 24 h postirradiation (induced by arsenite-mediated inhibition of DNA repair) appears to increase the proportion of cells in the late S/G2/M phases at 36 h postirradiation and, consequently, radiosensitization by arsenite.
Because HR is regulated by p53 (40), loss of HR through p53 deficiency might only radiosensitize p53-deficient cells in the presence of arsenite. Collectively, our observations support the scheme shown in Figure 4. The results of the colony formation assay using other cell lines (i.e., p53−/− HCT116 and p53+/+ HCT116 cell lines) revealed arsenite-induced radiosensitization only in the p53-deficient line (data not shown), supporting the general role of p53 status in arsenite-induced radiosensitization.
Proposed model mechanism underlying radiosensitization of p53-deficient glioma cells by arsenite. In our previous study (15), we found that arsenite induced radiosensitization by generating reactive oxygen species (ROS), which inhibited BRCA2-mediated repair of DNA damage (broken lines). The present study (solid lines) shows that radiation increases DNA double strand breaks (DSB) and γH2AX foci (which are suppressed by p53) and increases the percentage of cells in the late S/G2/M phases (which is suppressed by p53). Therefore, deficiency of p53 increases radiosensitivity.
Phase I and II trials have been conducted to assess combined use of arsenite trioxide plus radiotherapy for brain tumors (41-43). The present study reveals for the first time that arsenite sensitizes p53-deficient tumor cells to X-rays and heavy ion beams over a wide LET range (~13–200 keV/μm). Arsenic trioxide is a more potent generator of ROS and cellular toxicity than sodium arsenite at the same concentration (44); therefore, arsenic trioxide is expected to induce radiosensitization at lower concentrations than sodium arsenite.
Many tumors are p53-deficient (14) and are therefore expected to be radiosensitized by arsenite; thus, prior determination of tumor p53 status, as well as combined use of p53 inhibitors, should increase the efficacy of radiotherapy with X-rays and heavy ion beams, including multi-ion beams (45), in the presence of arsenite. This might hold true for other metal-related elements, such as gadolinium, for which the underlying mechanism of radiosensitization differs from that of arsenite (46, 47). Further studies are needed to clarify the role of p53 in arsenite-induced radiosensitization.
Acknowledgements
The Authors thank Dr. Paul S. Mischel for supplying the U87MG-neo and U87M-E6 cell lines, and Dr. Dong Yu and Dr. Emiko Sekine-Suzuki for the valuable help during preparation of the manuscript.
Footnotes
Authors’ Contributions
YN conceived the idea of the study, performed all experiments, and contributed to data analysis and interpretation. TI and TN contributed to the interpretation of the results. KD substantially contributed to the manuscript drafting. SK supervised the conduct of this study. All Authors reviewed and revised the manuscript draft and approved the final version of the manuscript to be published.
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
- Received March 8, 2023.
- Revision received April 6, 2023.
- Accepted April 7, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.










