The Novel Isocoumarin 2-(8-Hydroxy-6-methoxy-1-oxo-1H -2-benzopyran-3-yl) Propionic Acid (NM-3) Induces Lethality of Human Carcinoma Cells by Generation of Reactive Oxygen Species¹

Cytogenin (8-hydroxy-3-hydroxymethyl-6-methoxyisocoumarin) is a novel microbial product isolated from a culture filtrate of Streptoverticillium eurocidicum. Screening in mouse models demonstrated that cytogenin has antitumor activity (1). In vivo studies also demonstrated that cytogenin suppresses angiogenesis induced by Sarcoma-180 tumor cells (2). As a result of these findings, synthetic cytogenin derivatives were assayed for antiangiogenic activity in the mouse dorsal air sac model. Among the derivatives, NM-3³ was selected for oral bioavailability, little if any toxicity, and dose-dependent antiangiogenic effects (3). NM-3 exhibits in vivo activity against Lewis lung carcinoma and human tumor xenografts (4). Moreover, NM-3 has been found to potentiate the antitumor effects of radiotherapy without an increase in toxicity (4). Other work has shown that NM-3 is cytotoxic to HUVECs and inhibits HUVEC migration (4). These findings have indicated that the in vivo activity of NM-3 is related to effects on the tumor vasculature. Studies have not to determined whether NM-3 has direct effects on tumor cells.

The present studies have assessed the effects of NM-3 on human carcinoma cells. The results demonstrate that NM-3 treatment is associated with ROS generation, growth arrest, and the induction of apoptosis.

Cell Culture. Human breast (MCF-7 and ZR-75-1), ovarian (PA-1), and cervical (HeLa) cancer cells were obtained from the American Type Culture Collection (Manassas, VA). MCF-7 and HeLa cells were grown in DMEM (high glucose; Mediatech, Cellgro) supplemented with 10% FCS and 2 mml-glutamine. ZR-75-1 cells were cultured in RPMI 1640 (Mediatech, Cellgro) supplemented with 10% FCS and 2 mml-glutamine. PA-1 cells were cultured in MEM (Life Technologies, Inc.) supplemented with 10% FCS and 2 mml-glutamine. Cells were treated with NM-3 (Mercian, Tokyo, Japan) and 30 mm NAC (Sigma Chemical Co.).

Cell Growth and Clonogenic Survival Assays. Cells in logarithmic growth phase were treated with NM-3, released by trypsinization, pelleted, and counted in a hemocytometer. For clonogenic survival assays, cells were treated with NM-3 for 48 h, washed, and maintained in complete medium for 10–14 days. Cells were then stained with Giemsa, and colonies >50 cells were scored as positive.

Measurement of ROS Production. Cells were incubated with 10 μm DCF-DA (Sigma Chemical Co.) for 15 min at 37°C to assess ROS-mediated oxidation of DCF-DA to the fluorescent compound DCF (5). Fluorescence of oxidized DCF was measured at an excitation wavelength of 480 nm and an emission wavelength of 525 nm using a flow cytometer (Becton Dickinson, Lincoln Park, NJ).

Cell Cycle Analysis. Cells were fixed in 80% ice-cold ethanol/PBS for 30 min, rinsed, and resuspended in PBS containing 30 μg/ml RNase A for 1 h at 37°C. The cells were collected, stained with propidium iodide for 30 min, and analyzed by flow cytometry.

Immunoblot Analysis. Cells were lysed in ice-cold lysis buffer [20 mm Tris-HCl (pH 8.0), 150 mm NaCl, 1% Triton X-100, 1 mm phenylmethylsulfonyl fluoride, 1 nm DTT, 10 μg/ml leupeptin, and 10 μg/ml aprotinin] for 30 min. Lysates were cleared by centrifugation at 14,000 × g for 20 min at 4°C. Protein concentration was determined by the Bradford assay (Bio-Rad, Richmond, CA). Proteins were separated by SDS-PAGE, transferred to nitrocellulose, and probed with anti-p53 (Santa Cruz Biotechnology), anti-p21 (Santa Cruz Biotechnology), or anti-caspase-3 (Transduction Labs, Lexington, KY). The antigen-antibody complexes were visualized by enhanced chemiluminescence (ECL; Amersham Life Science).

Cdk2 Kinase Assays. Cell lysates were subjected to immunoprecipitation with anti-Cdk2 antibody (Santa Cruz Biotechnology). Immune precipitates were collected with protein A-Sepharose beads equilibrated in lysis buffer. The beads were washed and resuspended in kinase buffer [50 mm Tris-HCl (pH 7.5), 10 mm MgCl₂] containing 20 nm ATP, 5 μCi of [γ-³²P]ATP, and 2.5 μg of histone H1 for 20 min at 37°C. The reaction products were analyzed by SDS-PAGE and autoradiography.

Quantitation of ATP. Levels of ATP were measured using an ATP Determination kit (Molecular Probes).

NM-3 Induces Cytotoxicity of Human Carcinoma Cells. To assess the effects of NM-3 on cell proliferation and survival, MCF-7 breast carcinoma cells were treated with 100–400 μg/ml NM-3 for 24 and 48 h. The results demonstrate a dose- and time-dependent inhibition of cell growth (Fig. 1A). Similar findings were obtained with ZR-75-1 breast cancer cells (Fig. 1B). To determine whether NM-3 affects clonogenic survival, cells were treated for 48 h and then cultured for 10–14 days. Assessment of colonies demonstrated that NM-3 decreases clonogenicity of MCF-7 cells with an IC₅₀ of 180 μg/ml (Fig. 2A). Similar results were found for ZR-75-1 cells (IC₅₀, 180 μg/ml; Fig. 2B). To extend these findings to other types of carcinomas, we assessed the effects of NM-3 on the survival of PA-1 ovarian carcinoma and HeLa cervical carcinoma cells. The results show that, as found for breast carcinoma cells, NM-3 inhibits the clonogenic survival of PA-1 (IC₅₀, 60 μg/ml) and HeLa (IC₅₀, 70 μg/ml) cells (Fig. 2, C and D). Short exposures (1–6 h) to 400 μg/ml NM-3 had no apparent effect on cell survival (data not shown).

NM-3 Induces ROS Formation. Recent studies have demonstrated that NM-3 potentiates the effects of ionizing radiation (4). Because irradiation of cells is associated with the generation of ROS, we asked whether NM-3 treatment affects intracellular levels of reactive oxygen intermediates. After incubating with DCF-DA, cells were exposed to NM-3 and ROS-mediated oxidation of the fluorochrome was assayed by flow cytometry. Compared with control MCF-7 cells, treatment with NM-3 for 30 min was associated with detectable increases in ROS (Fig. 3A). Moreover, the increases in ROS levels persisted through 24–48 h of NM-3 exposure (data not shown). Similar studies in HeLa and PA-1 cells demonstrated more pronounced effects of NM-3 on ROS production (Fig. 3, B and C). As controls, MCF-7 and HeLa cells were incubated with NAC, a precursor of glutathione and scavenger of ROS (6, 7). NAC pretreatment of MCF-7 and PA-1 cells was associated with abrogation of NM-3-induced ROS production (Fig. 3D). For HeLa cells, NAC pretreatment attenuated NM-3-induced generation of ROS (Fig. 3D). These results indicate that the response of cells to NM-3 includes the generation of ROS.

NM-3 Activates the p53 Tumor Suppressor. p53 is activated by aberrant growth signals and diverse types of stress that include ROS-induced damage (8, 9). To determine whether NM-3 treatment is associated with induction of p53, we analyzed p53 expression in lysates of NM-3-treated MCF-7 cells. The results demonstrate that p53 levels are increased in the response to NM-3 (Fig. 4). Similar results were obtained after treatment of ZR-75-1 and PA-1 cells (Fig. 4). Moreover, NM-3 concentrations that resulted in activation of p53 were comparable with those associated with growth arrest and loss of clonogenic survival. By contrast, NM-3 had no detectable effect on p53 levels in HeLa cells (data not shown), which are mutant for this tumor suppressor (10). These findings indicate that the cellular response to NM-3 includes, in part, p53 activation.

NM-3 Induces p21, Down-Regulation of Cdk2, and G₁ Cell Cycle Arrest. p53 activates expression of the p21 Cdk inhibitor (11, 12). To determine whether NM-3 induces p21 levels, lysates of NM-3-treated MCF-7 cells were analyzed by immunoblotting with anti-p21. The results demonstrate that MCF-7 cells respond to NM-3 with dose-dependent increases in p21 levels (Fig. 5A). ZR-75-1 and PA-1 cells also responded to NM-3 with induction of p21 (Fig. 5A). To extend these findings, studies were performed to assess the effects of NM-3 on Cdk2 kinase activity. Cdk2 immune complexes were prepared from lysates of NM-3-treated MCF-7 cells and assayed for phosphorylation of histone H1. In concert with the induction of p21, the results demonstrate that NM-3 induces a dose-dependent inhibition of Cdk2 activity (Fig. 5B). Similar results were obtained in NM-3-treated ZR-75-1 cells (Fig. 5B). These findings demonstrate that NM-3 induces p21 expression and the down-regulation of Cdk2 activity.

To determine the effects of NM-3 on cell cycle distribution, MCF-7 cells treated with different concentrations of this agent were stained with propidium iodide and analyzed by flow cytometry. NM-3 treatment was associated with a dose-dependent accumulation of MCF-7 cells in the G₁ phase and a decrease in the S-phase population (Fig. 6A). ZR-75-1 cells also responded to NM-3 with a dose-dependent arrest of growth at G₁-S-phase (Fig. 6B). Significantly, there was no detectable induction of MCF-7 or ZR-75-1 cells with sub-G₁ DNA after treatment with NM-3 for 120 h (data not shown). Whereas these findings indicate that MCF-7 and ZR-75-1 cells do not undergo apoptosis in response to NM-3 and necrosis can be distinguished from apoptosis by depletion of ATP (13, 14), we measured ATP levels in these cells. The results demonstrate that NM-3 (400 μg/ml) decreases ATP levels in MCF-7 cells by 56.6 + 3.0% (mean + SE of three replicates) and in ZR-75-1 cells by 61.8 + 5.0% (mean + SE of three replicates) at 48 h. These results collectively indicate that NM-3 activates a p21-dependent inhibition of Cdk2, arrest of cells at the G₁-S interphase, and a necrotic cell death response.

NM-3 Induces Cell Type-dependent Apoptosis. By contrast to the induction of G₁-S arrest in NM-3-treated MCF-7 and ZR-75-1 cells, analysis of PA-1 cells exposed to this agent demonstrated the induction of sub-G₁ DNA (Fig. 7A). Similar results were obtained with HeLa cells (Fig. 7A). To confirm the induction of apoptosis, lysates from NM-3-treated HeLa cells were analyzed for the activation of caspase-3. The results demonstrate cleavage of pro-caspase-3 in response to NM-3 (Fig. 7B). Moreover, in concert with the finding that ZR-75-1 cells respond to NM-3 with arrest in G₁ and not accumulation of sub-G₁ DNA, there was no detectable cleavage of pro-caspase-3 after NM-3 treatment of these cells (data not shown). To assess involvement of NM-3-induced ROS production, HeLa cells were pretreated with NAC and then assayed for cleavage of pro-caspase-3. The results demonstrate that NAC blocks NM-3-induced cleavage of pro-caspase-3 (Fig. 7D). Moreover, NAC treatment of HeLa cells blocked induction of sub-G₁ DNA by NM-3 (Fig. 7C). These findings demonstrate that NM-3 induces apoptosis by a ROS-dependent mechanism.

Certain derivatives of isocoumarin have been identified as potent irreversible inhibitors of blood coagulation serine proteases (15, 16). The isocoumarin cytogenin, however, has no apparent effect on serine proteases in vitro.⁴ Also, the cytogenin derivative, NM-3, has had no effect on blood coagulation in animal models (3) and in early clinical trials.⁴ The available evidence has supported a role for NM-3 as an inhibitor of angiogenesis. NM-3 decreases the survival of HUVEC (IC₅₀, 1 μg/ml) by a nonapoptotic mechanism (4). In addition, oral administration of NM-3 to mice (0.3–10 mg/kg/day) produced a dose-dependent suppression of angiogenesis induced by tumor cells (3). At an oral dose of 10 mg/kg in mice, plasma levels of NM-3 exceed 100 μg/ml, a concentration ∼10²-fold higher than the IC₅₀ of endothelial cells. Oral doses of 1 g/kg have been well tolerated in mice.⁴ Thus, NM-3 plasma levels of >500 μg/ml are achievable, at least in animal models, without acute toxicity. Dose escalation in the ongoing NM-3 Phase I trials has not as yet determined whether similar concentrations can be achieved in humans without normal tissue toxicity.

The present studies demonstrate that NM-3 induces the generation of ROS and inhibits the growth of human cancer cells. In contrast to HUVECs that exhibit an IC₅₀ of 1 μg/ml (4), the IC₅₀s of carcinoma cells ranged from 60 to 180 μg/ml. The results indicate that treatment of certain carcinoma cells with NM-3 results in death responses that are nonapoptotic. Human MCF-7 and ZR-75-1 breast cancer cells responded to NM-3 with growth arrest at G₁-S-phase and loss of clonogenic survival. As found for HUVECs treated with NM-3 (4), the breast cancer cells showed no evidence of apoptosis in response to this agent. By contrast, the PA-1 ovarian carcinoma and HeLa cervical carcinoma cells were more sensitive to NM-3 in clonogenic survival studies and responded with the induction of apoptosis. These findings suggest that, although cytotoxic to diverse human carcinoma cells, the mechanism of cell death to NM-3 appears to be cell type dependent.

The effects observed with NM-3 can be explained, at least in part, by activation of the p53 tumor suppressor. Rapid induction of p53 activity occurs by mechanisms involving posttranslational stabilization in the response of cells to various types of stress (8, 9). The present results demonstrate that NM-3 treatment is associated with dose-dependent increases in p53 levels. Moreover, in certain human carcinoma cells, activation of p53 was associated with induction p21 expression, an event in part regulated by p53-mediated transactivation. In concert with these findings, cells that respond to NM-3 with increases in p21 exhibited downregulation of Cdk2 and arrest in G₁-S-phase. This response to NM-3 was also associated with loss of clonogenic survival without a detectable induction of apoptosis. Alternatively, other human carcinoma cells responded to NM-3 with activation of p53 and the induction of apoptosis. In clonogenic survival studies, PA-1 and HeLa cells that exhibit an apoptotic response to NM-3 were more sensitive to the lethal effects of this agent.

The precise mechanism of action of NM-3 has been unknown. Although the present work demonstrates that NM-3 induces ROS formation and activates p53-dependent signaling, this agent may function in the regulation of other genes. Notably, the present results also demonstrate that NM-3 induces apoptosis of the p53-mutant HeLa cells. These findings indicate that NM-3-induced apoptosis is not dependent on p53-mediated signaling. Taken together with the observations that NM-3 induces endothelial cell death (3, 4), the demonstration that NM-3 has direct effects on human carcinoma cells indicates that this agent, although more potent against endothelial cells, could be effective in targeting both the tumor and its vasculature.

We appreciate the technical assistance of Kamal Chauhan.