Nutlin-3 radiosensitizes hypoxic prostate cancer cells independent of p53

Localized prostate cancer is characterized by marked hypoxia and significant heterogeneity in oxygenation, similar to other human tumors (1, 2). Intratumoral hypoxia can be chronic (e.g., diffusion-limited) or acute (e.g., by temporary obstructions or variable blood flow in tumor vessels; ref. 3). The acute and chronic hypoxia found in tumors significantly complicates cancer therapy. Hypoxic cancer cells are considered to be resistant to radiation and chemotherapy and are more prone to develop metastasis (4). Hypoxia can increase radioresistance by up to 3-fold due to a decrease in lethal DNA double-strand breaks (5). Hypoxia can also lead to increased chemoresistance due to limited drug diffusion through hypoxic areas, acquisition of resistance to drug-induced apoptosis, and the decreased production of free radicals required for the cytotoxic action of certain chemotherapy drugs (6). In certain tumors, hypoxia can also induce a metastatic phenotype (4, 7), which decreases cancer-free survival. These effects are thought to occur by hypoxia-induced changes in gene expression and genetic stability including the down-regulation of DNA damage signaling and repair (8).

Hypoxia stabilizes and promotes the activity of hypoxia-inducible factor-1α (HIF-1α). This in turn promotes the transcription of several downstream genes such as vascular endothelial growth factor, carbonic anhydrase IX, and GLUT-1, whose overexpression has been linked to poor prognosis and decreased therapeutic response (8, 9). Additionally, HIF-1α may increase MDM2, which is an important protein in the ATM-mediated response to DNA damage and activation of DNA repair and cell cycle pathways (10, 11).

MDM2 negatively regulates p53 by inhibiting its transcriptional activity and promoting its degradation by functioning as an E3-ubiquitin ligase in the proteasomal degradation pathway (12). The transcription factor p53 exerts a pivotal role in controlling DNA damage sensing, cell cycle progression, and apoptosis in response to various forms of genotoxic and cellular stress (reviewed in ref. 13). MDM2 also has p53-independent functions that require binding of MDM2 to proteins such as p14^ARF (14), HIF-1α (15), NBS1 (16), E2F-1 (17), and ribosomal protein L5 (ref. 18; reviewed in ref. 19). Importantly, MDM2 can be up-regulated by cellular hypoxia and has been shown to increase the metastatic efficiency of a murine fibrosarcoma cell line by increasing resistance to apoptosis (11). Therefore, overexpression or aberrant function of MDM2 can lead to altered DNA damage signaling, hypoxic gene expression, and apoptotic thresholds.

Currently, several molecular signaling inhibitors are being tested both preclinically and clinically for their potential efficiency when combined with ionizing radiation (20). However, few agents are being tested under conditions of hypoxia. This is surprising given that hypoxia is common in tumors and is a major determinant of radioresponse. Nutlins are small molecules that inhibit MDM2 binding to p53 (21). These compounds bind in the p53-binding pocket of MDM2 to displace p53 from the complex and induce p53 stabilization. p53 then activates downstream targets leading to p21^WAF induction, cell cycle arrest, and apoptosis. In studies of chemosensitization, Nutlin-3a, the active enantiomer of Nutlin-3, increased the sensitivity to cisplatin by inhibiting MDM2 binding to E2F-1 (22). Increased sensitivity was linked to the induction of proapoptotic proteins, such as p73α and Noxa. Nutlin-3 also prevents the association between MDM2 and HIF-1α (23), thereby inhibiting vascular endothelial growth factor production and possibly tumor angiogenesis. Finally, a recent study reported that Nutlin-3a can radiosensitize wild-type p53 (WTp53)–expressing lung cancer cells (24). However, whether there are differential radiosensitizing effects of Nutlin-3 on cells that vary in p53 status or levels of oxygenation has not been studied.

We hypothesized that Nutlin-3 inhibition of MDM2-p53 interaction would lead to increased p53 activation under normoxia and hypoxia. Herein, we show novel data indicating that Nutlin-3 radiosensitizes prostate cancer cells under both normoxia and hypoxia independent of p53 status. Nutlin-3 and similar MDM2-p53 inhibitors may therefore be useful adjuncts to radiotherapy to improve clonogenic cell killing in tumor cells that have acquired hypoxia-mediated local or systemic resistance.

Nutlin-3, a racemic mix of Nutlin-3a (active enantiomer) and Nutlin-3b (inactive enantiomer), was purchased from Sigma-Aldrich and dissolved in DMSO at a concentration of 1 mmol/L. Using the stock solution, Nutlin-3 was further diluted in medium for individual experiments. For control experiments, DMSO dissolved in medium at a concentration of 0.5% was used.

Three human malignant prostate cancer cell lines with varying p53 status were selected for study: DU145 (p53 mutated) and PC-3 (p53 null) were purchased from American Type Culture Collection. The 22RV1 cell line (p53 wild-type) was a kind gift of Dr. J.H. Pinthus (McMaster University). These cell lines were grown in α-MEM, Ham's F-12K, and RPMI 1640 supplemented with 10% FCS and 1% l-glutamine, respectively. Approximate doubling times for cell cultures under these conditions were as follows: DU145, 18 h; PC-3, 24 h; and 22RV1, 40 h, as described previously (25).

For all experiments, 1 × 10⁵ to 3 × 10⁵ cells were plated in 10-cm dishes, such that cells were growing exponentially at all time points tested. The cell lines were passaged in 5% CO₂ and air at 37°C for routine culture. For certain experiments, cells were exposed to either acute hypoxic or chronic anoxic gassing. For acute hypoxia, exponentially growing cells were plated as above, placed in a hypoxic environment (0.2% O₂) using an In vivo₂ 400 chamber (Ruskinn Technology) for a minimum of 2 h before irradiation. To induce chronic anoxia, flasks were flushed with a gas concentration of 0% O₂, 5% CO₂, and balanced with N₂ for 24 h before irradiation as described previously (26).

Western blotting was carried out as described previously (26). Primary antibodies were used at the following dilutions: anti-actin rabbit monoclonal antibody (Sigma-Aldrich) 1:50,000, anti-p53 mouse monoclonal antibody clone BP53-12 (Santa Cruz Biotechnology) 1:10,000, anti-p21^WAF mouse monoclonal antibody (Rockland Immunochemicals) 1:10,000, anti-MDM2 mouse monoclonal antibody (Calbiochem) 1:1,000, and anti-HIF-1α mouse monoclonal antibody (BD Translab) 1:750. The washed blot was incubated with secondary antibody conjugated to Alexa 680 (Invitrogen) or IRdye-800 (Rockland Immunochemicals) diluted 1:10,000 in LI-COR blocking buffer (diluted 1:1 in PBS) for 30 min at room temperature followed by five 5-min washes with TBS-Tween 20 at room temperature. Western blots were imaged and analyzed using a LI-COR Odyssey infrared imaging system (LI-COR Biosciences) and densitometry values calculated based on actin loading controls.

Clonogenic survival assays were done as described previously (27). Cells were counted and plated at different densities 15 h before Nutlin-3 exposure. Cells were irradiated at room temperature using a ¹³⁷Cs γ-ray irradiator (Nordion) at a dose rate of 0.9 Gy/min. Following incubation, drug was removed and fresh medium was added. Colonies were stained and counted under a microscope, with 50 cells as the minimum number to define a surviving colony. The best-fit survival curve was generated according to a linear-quadratic survival model and the mean inactivation dose (area under survival curve) was calculated as described previously (28). The sensitizing enhancement ratio (SER) was defined as SER = mean inactivation dose control irradiation / mean inactivation dose Nutlin-3 + irradiation. A SER value greater than 1.0 is indicative of radiosensitization. Survival curves for combined treatment were corrected for cytotoxicity due to Nutlin-3 alone.

For cell cycle analysis, cells were stained with propidium iodide and counted with a FACScan (Becton Dickinson) as described previously (28).

For apoptosis measurements, cells were collected 48 h after starting treatment with the drug and stained with Hoechst 33342 for microscopic evaluation of the presence of apoptotic bodies as described previously (27). A minimum of 200 cells were counted under a microscope for each experimental point.

The mean ± SE (standard error of the mean) was calculated for a minimum of three independent experiments. For survival curves, statistical comparisons were done on mean inactivation dose. For statistical analysis, paired t-test was calculated using Instat (Graphpad Software).

Nutlin-3 induced p53 and p21^WAF expression in a dose-dependent manner in 22RV1 cells (WTp53; Fig. 1A). In contrast, p21^WAF was not induced in the DU145 (mutated p53) or PC-3 (p53-null) cell lines (data not shown). Short-term cell cycle assays showed that, at a dose of 10 μmol/L, Nutlin-3 increased slightly the G₁-phase fraction and decreased S-phase fraction of all three cell lines (Fig. 1B). Nutlin-3 decreased significantly clonogenic survival of 22RV1 cells (IC₅₀, 4.3 μmol/L) when compared when DU145 and PC-3 cells (e.g., IC₅₀, >10 μmol/L; Fig. 1C). Nutlin-3 also induced a significant increase in apoptosis in 22RV1 cells (P < 0.05), whereas no significant increase in apoptosis was observed for p53-deficient DU145 or PC-3 cells (Fig. 1D). We conclude that, as a single agent, Nutlin-3 increases apoptosis and decreases clonogenic survival preferentially in WTp53-expressing cells.

Radiotherapy is usually given as a series of daily fractions of 2 Gy to a total dose of 60 to 80 Gy. To select the optimal combination of Nutlin-3 and irradiation for clonogenic cell killing, initially we incubated cells in Nutlin-3 for 0 to 8 h followed by irradiation at the clinically relevant dose of 2 Gy. In the clonogenic assay, the surviving fraction at 2 Gy (Fig. 2A) decreased significantly when 22RV1 or DU145 cells were treated with Nutlin-3 for 24 h with irradiation at 2 h following the start of incubation. No significant differences were noted between the two Nutlin-3 concentrations tested (e.g., 2 and 5 μmol/L). Therefore, we selected a 24-h Nutlin-3 incubation (5 μmol/L) with irradiation at 2 h following the start of this incubation for all further experiments.

We next compared p53 and p21^WAF expression following DNA damage in the presence of Nutlin-3 among the cell lines. The combination of Nutlin-3 and irradiation induced p53 and p21^WAF to levels greater than either treatment or radiation alone in the WTp53-expressing 22RV1 cells (Fig. 2B). As expected, no induction was noted in DU145 or PC-3 cells (data not shown). We then tested loss of clonogenic survival and apoptosis following treatment with both Nutlin-3 and irradiation under normoxia. Nutlin-3 significantly increased the radiosensitivity of 22RV1 (SER, 1.24 ± 0.1), DU145 (SER, 1.27 ± 0.13), and PC-3 (SER, 1.12 ± 0.08) cells (Fig. 2C) in the clonogenic assay. Nutlin-3 significantly increased apoptosis when combined with radiation compared with drug or radiation alone only for the WTp53-expressing 22RV1 cells (P < 0.05; Fig. 2D).

Hypoxia can increase MDM2 levels and also induce radioresistance. We therefore tested Nutlin-3 effects under acute hypoxia. Acute hypoxia alone did not lead to clonogenic cell kill. Nutlin-3 clonogenic cell kill was similar under conditions of both normoxia and hypoxia in all three cell lines (data not shown). Hypoxia induced HIF-1α stabilization under these gassing conditions (Fig. 3A). HIF-1α levels increased more rapidly (starting at 2 h) in cells treated with Nutlin-3. As expected, Nutlin-3-induced stabilization of p53 increased MDM2 expression as part of a p53-induced feedback loop. Hypoxia alone did not induce MDM2 expression, but we did observe increased MDM2 levels following treatment with hypoxia plus Nutlin-3 (Fig. 3B). Under acute hypoxia, Nutlin-3 alone, irradiation alone, or both treatments combined induced a dose-dependent activation of p53 and p21^WAF in 22RV1 cells (Fig. 3C) but not in DU145 or PC-3 cells (data not shown). Hypoxic radioresistance was observed as early as 2 h and reached a plateau at 5 h of gassing (Fig. 3D). Incubation of cells in Nutlin-3 partially reversed this radioresistance. Under acute hypoxia, Nutlin-3 significantly increased killing of 22RV1 (SER, 1.78 ± 0.19), DU145 (SER, 1.31 ± 0.18), and PC-3 (SER, 1.28 ± 0.09) cells (Fig. 3E). These SER values were increased compared with the SER values under normoxia. We conclude that Nutlin-3 increases radiosensitivity under acute hypoxic conditions and that sensitization may be greater in cells that maintain p53 function.

To ascertain whether Nutlin-3 was a radiosensitizer of anoxic prostate cancer cells, cells were exposed to Nutlin-3 under chronic anoxia for 24 h and then irradiated. Anoxia during 24 h induced HIF-1α to a similar extent with or without Nutlin-3 (Fig. 4A). Nutlin-3 induced p53 to a greater extent under anoxia compared with normoxia in 22RV1 cells (Fig. 4B). Under anoxia, clonogenic cell killing was more marked when Nutlin-3 was combined with radiation compared with radiation alone (Fig. 4C). Nutlin-3 alone induced more apoptosis under anoxia than normoxia for 22RV1 and DU145 cells (Fig. 4D), but Nutlin-3 increased radiation-induced apoptosis only in 22RV1 cells. No significant differences were noted in this cell line between the apoptotic levels following combined treatment under normoxia or anoxia.

The use of molecular inhibitors of signaling pathways of DNA repair and cell cycle checkpoints is increasing in radiation oncology (20). However, little data are available on whether molecular inhibitors can radiosensitize cancer cells under both oxic and hypoxic/anoxic conditions. This is important as hypoxic and anoxic cells can severely reduce the efficacy of cancer treatment and drive a metastatic phenotype (4).

In our study, we observed that Nutlin-3 is a radiosensitizer under both hypoxic and anoxic conditions, further supporting the use of this agent in solid tumors that contain populations of cells with varying levels of oxygen. The SER were rather small, with the exception of the effect seen in p53WT cells under hypoxia. Our data support a previous finding (24) that Nutlin-3 radiosensitizes cancer cell lines under oxia. However, in our study, Nutlin-3 was found to radiosensitize cells independent of p53. This supports another report in which MDM2 inhibition using antisense oligonucleotides led to increased radiosensitization through a p53-independent mechanism (29).

Nutlin-3 may radiosensitize WTp53-expressing cells, such as 22RV1, via multiple factors, including p21^WAF activation and increased apoptosis. However, the mechanism of the radiosensitization of mutated p53–expressing or p53-null cells in our study remains unclear. In treated cells, we did not find a 1:1 correlation between the short-term death endpoint of apoptosis and the longer-term death endpoint of colony formation, suggesting that other forms of cell death (e.g., mitotic catastrophe or terminal growth arrest) may also be factors in Nutlin-associated sensitization.

An important finding of our study is that Nutlin-3 radiosensitized chronically anoxic cells or acutely hypoxic cells to a greater extent than normoxic cells. These radiosensitizing properties of Nutlin-3 were not affected by the oxygen levels (0.2% or 0%) or the duration of hypoxia before radiation (e.g., 2 or 24 h). The molecular mechanisms for the increased radiosensitizing properties of Nutlin-3 under hypoxic or anoxic conditions also remain unclear. Previous studies have shown that Nutlin-3 can alter the binding of MDM2 to both HIF-1α and p53, which could modify cell survival. Consistent with these results, our data show that Nutlin-3 increased the level of HIF-1α at the time of irradiation and that increased apoptosis was observed in WTp53 prostate cells under hypoxic and anoxic gassing conditions.

In conclusion, Nutlin-3 can radiosensitize hypoxic and anoxic prostate cancer cells and should be further evaluated in preclinical studies as a possible radiosensitizer in vivo given that hypoxia can occur in 30% to 90% of prostate tumors (8). The SER of Nutlin-3 under hypoxia is similar to the hypoxia-selective nitroacridine drug nitracrine (30) but lower than tirapazamine (SER between 2.55 and 2.70; refs. 31, 32). Further studies need to be completed with tumor cells in vitro and in vivo that are derived from a range of tumors to determine better the general implications of our data relating to the radiosensitizing properties of Nutlin-3 under hypoxia/anoxia. However, if confirmed for a range of cell lines, using modern radiotherapy techniques to target hypoxic subpopulations in combination with Nutlin-3 could be part of biologically adaptive radiotherapy using boosts (a concept named “dose Painting”; ref. 33). In this manner, the dose to hypoxic/anoxic regions can be selectively increased to enhance radiotherapy clonogenic cell kill. Precise targeting of hypoxic/anoxic subvolumes and the use of drugs that radiosensitize hypoxic cancer cells could improve clonogenic cell kill and maintain the therapeutic ratio. Nutlin-3 could therefore be a useful drug in this instance of combined modality therapy to target radioresistant hypoxic cancer cells within prostate tumors.