Arsenic trioxide prevents radiation-enhanced tumor ... - Nature

Oncogene (2005) 24, 390–398

& 2005 Nature Publishing Group All rights reserved 0950-9232/05 $30.00 www.nature.com/onc

Arsenic trioxide prevents radiation-enhanced tumor invasiveness and inhibits matrix metalloproteinase-9 through downregulation of nuclear factor jB Lin-Hung Wei1, Kuo-Pao Lai1, Chi-An Chen2, Chia-Hsien Cheng1, Yun-Ju Huang1, Chia-Hung Chou3, Min-Liang Kuo3 and Chang-Yao Hsieh*,2 1

Department of Oncology, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan; 2Department of Obstetrics and Gynecology, National Taiwan University Hospital and National Taiwan University College of Medicine, Taipei, Taiwan; 3Laboratory of Molecular and Cellular Toxicology, Institute of Toxicology, National Taiwan University College of Medicine, Taipei, Taiwan

Arsenic trioxide (ATO) has been implicated as a promising anticancer agent by inhibiting cell growth and inducing apoptosis in certain types of cancer cells. This study explored the antimetastasis property of arsenic, drew potential link between arsenic use and radiotherapy, and uncovered the specific mechanisms underlying these remarkable responses. Using gelatin invasion assay and intravasation assay, we report the novel finding that low-dose ATO (1 lM) reduced the intrinsic migration ability of HeLa cells and significantly inhibited radiation-promoted tumor invasive potential of CaSki cells without inducing apoptotic cell death. Using the murine Lewis lung carcinoma model, the control animals and ATO treatment animals (1 mg/kg i.p., twice weekly) displayed similar in vivo growth kinetics, whereas the radiation (30 Gy in one fraction) and concurrent treatment groups showed considerable growth inhibition. Importantly, although concurrent treatment did not enhance the effectiveness of radiation therapy to the primary tumor, further examination of the lungs revealed that all animals succumbed to radiation-accelerated lung metastases could be effectively treated by combination of ATO and radiation. Radiationinduced matrix metalloproteinase-9 (MMP-9) expression was significantly inhibited by ATO using sequential analysis of the expression of MMPs in xenografts. Supporting this observation, ATO directly downregulates radiation-induced MMP-9 mRNA expression by inhibiting nuclear factor jB activity in human cervical cancer cells. In sum, concurrent arsenic– radiation therapy not only achieves local tumor control but also inhibits distant metastasis. Experimental results of this study highlight a novel strategy in cancer treatment. Oncogene (2005) 24, 390–398. doi:10.1038/sj.onc.1208192 Published online 8 November 2004 Keywords: arsenic trioxide; irradiation; matrix metalloproteinase-9; metastasis; nuclear factor kB

*Correspondence: C-Y Hsieh, Department of Obstetrics and Gynecology, National Taiwan University Hospital, No. 7, Chung-Shan South Road, Taipei 100, Taiwan; E-mail: [email protected] Received 8 July 2004; revised 23 August 2004; accepted 6 September 2004; published online 8 November 2004

Introduction The anticancer activity of arsenic derivatives has been known for hundreds of years. However, due to the strong toxicity of oral arsenic on most patients along with the later development of modern radiotherapy and chemotherapy, the therapeutic use of arsenic trioxide (ATO) was gradually abandoned (Kwong and Todd, 1997). During 1990s, the discovery of intravenous infusion of ATO, which achieved a complete remission on a significant proportional number of patients diagnosed with acute promyelocytic leukemia, has led to the reintroduction of arsenic compounds in current therapeutic regimens (Shen et al., 1997; Soignet et al., 1998). Arsenic can influence various cellular effects, which include apoptosis induction, growth inhibition, promotion of differentiation, and angiogenesis inhibition (Miller et al., 2002). The mechanisms of action of ATO, depending on the cell type and applied concentration of arsenic, may be attributed to its effect on cellular signaling, redox status, stress response, and caspases activation (Miller et al., 2002). These multiple actions of arsenic have stimulated numerous clinical trials of ATO in hematologic as well as solid tumors (Murgo, 2001). However, despite arsenic being an active anticancer drug, chronic exposure to arsenic and arsenic derivatives can also have toxic effects (Hughes, 2002). Only at low doses do the therapeutic benefits outweigh the toxicity. The metastatic spread of cancer causes 90% of human cancer deaths, and thus remains the greatest barrier to curing cancer (Sporn, 1996). Invasion and metastasis of solid tumor are exceedingly complicated processes, which involve the escape of cells from the primary tumor, followed by intravasation, survival and transport in the circulation, lodgement, extravasation, and secondary growth in distant sites of the body (Chambers et al., 2002). Several classes of proteins are altered in cells with invasive or metastatic capabilities, including members of the immunoglobulin, calcium-dependent cadherins family, integrins, extracellular proteases, and angiogenic factors (Chambers and Matrisian, 1997;

Arsenic prevents radiation-enhanced cancer metastasis L-H Wei et al

391

Lukashev and Werb, 1998; Christofori and Semb, 1999; Folkman, 2002). Among these factors, tumor-associated matrix metalloproteinases (MMPs) are important components of the metastatic process through their capacity to degrade extracellular matrix proteins. The MMPs comprise a family of zinc-dependent endopeptidases, which are synthesized as inactive zymogens and activated by proteinase cleavage (Sternlicht and Werb, 2001). MMP activity is tightly controlled by endogenous inhibitors, including a2-macroglobulin, tissue inhibitors of metalloproteinases (TIMPs), and the membranebound inhibitor RECK (Baker et al., 2002). The expression and activity of MMPs are increased in almost every type of human cancer, and this is often associated with advanced tumor stage and poor survival (Sienel et al., 2003; Tanioka et al., 2003; Zucker and Vacirca, 2004). Recent studies demonstrated that MMPs also promote survival of cancer cells by liberating insulin growth factor and by cleavage of FAS ligand (Miyamoto et al., 2004; Strand et al., 2004). Furthermore, MMPs promote angiogenesis by increasing the bioavailability of the proangiogenic factors VEGF, FGF2, and TGF-b (Egeblad and Werb, 2002). The growing awareness of these new functions of MMPs in cancer biology provides the rationale for blocking both tumor metastasis and angiogenesis by inhibiting MMPs activity (Overall and Lopez-Otin, 2002). Ionizing radiation is a common conventional treatment modality for various human solid tumors. However, the therapeutic efficacy of radiotherapy alone for treating locally or regionally advanced cancer is often limited by tumor radioresistance or systemic tumor progression. Synergistic interaction with ATO and radiation has previously been reported. A higher dose of ATO results in sustained reduction of tumor blood flow and enhanced tumor response to radiotherapy in a locally advanced murine tumor model (Lew et al., 2002). Pretreatment of human cervical cancer cells with arsenic enhances radiation-induced apoptosis, which was associated with reactive oxygen species generation and the activation of caspase-9 and caspase-3 (Chun et al., 2002). This study serves as a pioneer in demonstrating that a low dose of ATO could reduce tumor invasiveness. In addition, the concurrent arsenic–radiation therapy could not only achieve local tumor control but also inhibit distant metastasis in murine tumor model. Results of this study indicate the effectiveness of a novel treatment strategy in cancer therapy.

Results

Figure 1 (a) Cytotoxicity of cervical cancer cells treated with different concentrations of ATO for 48 h. Cell viability was determined by MTT assay. (b) Effect of ATO on apoptosis induction in CaSki cells. Cells were treated with 0 mM (I), 1.0 mM (II), 2.5 mM (III), or 5.0 mM (IV) ATO for 48 h. After propidium iodide staining, cell cycle analysis was performed on a flow cytometer. The percentage of apoptotic cells in sub-G1 is displayed. (c) Effect of low concentration ATO on intravasation of tumor cells was analysed using semiquantitative RT–PCR analysis. HeLa cells (1 106 cells/well) were grown in culture and pretreated with 1 mM ATO for 24 h. The cells were detached with EDTA, resuspended in PBS, and inoculated on CAMs. At 50 h after inoculation, the lower CAMs were excised. Aliquots (100 ng) of the DNAs were used to amplify human Alu sequences. The intensity of the B220 bp band increased with increased content of human cancer cells. P, positive control from genomic DNA of HeLa cells

Low-dose arsenic trioxide reduces tumor invasiveness via an apoptosis-independent pathway In vitro growth inhibition of ATO has previously been reported in various human hematologic cancers and solid tumors. Likewise, remarkable in vitro inhibition of cell growth (>50%) was induced in all four cervical cancer cell lines at higher concentrations of ATO (>5 mM) (Figure 1a). Unlike higher dosage, treatment

of cervical cancer cells with low concentrations of ATO (p1 mM) did not induce the characteristic morphology of apoptosis, such as nuclear shrinkage with chromatin condensation and fragmentation (data not shown). Figure 1b demonstrates that 1 mM ATO had little effect on apoptotic cell death as compared with control group Oncogene


392

(3.5 vs 2.2%). By contrast, 2.5 or 5 mM ATO induced a strong apoptotic response (68.3 and 82.2%, respectively). Notably, when HeLa cells were exposed to lowdose ATO (1 mM), the intrinsic intravasation property of cancer cells was significantly reduced, as revealed by a semiquantitative PCR-based assay (Figure 1c). Since intravasation is an early step of the multifaceted process leading to metastasis, the experimental results suggest that ATO may reduce tumor invasiveness under noncytotoxic dosage. Arsenic trioxide prevents radiation-promoted cancer cell migration and metastasis Although radiotherapy is a major therapeutic modality for cancer treatment, previous works have implied that radiation may increase the invasive potential of cancer cells (Wild-Bode et al., 2001; Qian et al., 2002). Accordingly, this study examined the feasibility of ATO as a potent adjuvant therapeutic approach for improving the therapeutic efficacy of radiotherapy. Figure 2 shows that 1 mM ATO not only reduced the intrinsic migration ability of CaSki cells but also significantly inhibited radiation-promoted tumor invasive potential (Po0.01) using the invasion assay. Furthermore, this study examined the combinatorial effect of ATO and radiotherapy in reducing tumor metastasis by the Lewis lung carcinoma (LLC) model in mice (Camphausen et al., 2001). LLC cells were injected s.c. into the right thighs of C57BL/6 mice. At 7 days following implantation, the animals were randomized into one of four treatment groups: no treatment, 30 Gy in one fraction, ATO alone (1 mg/kg i.p., twice weekly), or concurrent ATO treatment and radiation. The mean tumor weight of the primary site in each group was monitored by killing five animals per group every 3

Figure 2 Effect of ATO on radiation-promoted migration and the invasiveness of CaSki cells. Invasion assay was performed using Boyden chambers coated with 0.2 mg/ml gelatin. CaSki cells were treated with the indicated dose of ATO for 24 h and then plated in the upper chamber with serum-free medium. Following 16 h of postradiation incubation, the cells in the lower chamber were stained with crystal violet and counted. The results shown here represent the mean7s.d. of three independent experiments performed in duplicate Oncogene

Figure 3 Effect of ATO on radiation-promoted metastatic tumor growth in murine. (a) LLC cells were implanted in the right hind leg of C57BL/6 mice. Radiation was administered at 30 Gy in a single dose to the primary tumor on day 7. ATO was administered at 1 mg/kg 1 day before radiation exposure and twice weekly according to the following protocol. The animals were killed (n ¼ 5 per group) as indicated and the primary tumor was dissected for weighing. Each value is the mean7s.d. obtained from five mice. (b) In another independent experiment, mice were treated with concurrent irradiation with various concentrations of ATO as indicated. The animals were killed after 21 days. The lungs were stained with H&E, and the foci were counted. Each value is the mean7s.d. obtained from 5 to 8 mice

days. Figure 3a demonstrates that the control animals and ATO treatment alone animals have similar in vivo growth kinetics, whereas the other two treatment groups display considerable growth inhibition (Po0.01), indicating the effectiveness of radiation in local tumor control. Although concurrent treatment did not increase the effectiveness of radiation therapy to the primary tumor, further examination of the lungs revealed that the use of radiation to delay primary tumor growth could actually accelerate metastatic growth, and notably the combination of ATO with radiation therapy may control distant metastases (Figure 3b). Repeat experiments were conducted using different dosages of ATO (1 and 2.5 mg/kg i.p., twice weekly), and all animals that


393

succumbed to lung metastases due to radiation therapy were effectively rescued by a combination of ATO and radiation (Figure 3b). Downregulation of radiation-induced MMP-9 expression by arsenic trioxide in Lewis lung cancer cells and tumors MMPs are important in creating and maintaining an environment that initiates and maintains growth of primary and metastatic tumors (Chambers and Matrisian, 1997). Accordingly, this study sequentially analysed the expression of MMP families in the xenograft of mouse LCC models using RT–PCR. Figure 4a reveals that MMP-9, but not MMP-2 or MMP-3, was markedly induced 3 days after irradiation, and intrinsic production of MMP-9 by LLC cells was observed after day 6. Production of both radiation-induced and intrinsically produced MMP-9 by LLC cells was significantly inhibited by arsenic treatment. Interestingly, tissue inhibitor of metalloproteinases 1 (TIMP-1) was also moderately raised in the arsenic treatment group. RT– PCR was employed to examine whether arsenic directly

regulates MMP-9 expression in the LLC cells, and the experimental results showed that a significant induction of MMP-9 by irradiation in LLC cells could be blocked by arsenic treatment, whereas TIMP-1 remained largely unchanged (Figure 4b). These results imply that MMP-9 is responsible, at least in part, for ATO to prevent radiation-induced tumor invasiveness. Arsenic trioxide transcriptionally inhibits radiationinduced MMP-9 expression in human cervical cancer cells This study next explored whether ATO regulates the expression of MMP-9 in human cancer cells. Initially, CaSki cells were serum starved and pretreated with varying concentrations of arsenic for 24 h. RT–PCR was used to assess the expression of various MMP families at 1 and 6 h after irradiation. Of the genes tested, MMP-9 mRNA, but not other factors, was significantly raised 6 h after irradiation and could be inhibited by arsenic dose-dependently (Figure 5a). Consistent with the mRNA level, the elevation of MMP-9 protein induced by irradiation was substantially reduced by arsenic using Western blotting analysis (Figure 5b). Additionally, arsenic treatment inhibited the activities of active form MMP-9 in conditioned media from irradiated CaSki cell culture using gelatin zymography (Figure 5c). This phenomenon suggests that ATO directly downregulated MMP-9 activities elicited by irradiation. Downregulation of nuclear factor kB is essential for arsenic trioxide to inhibit radiation-induced MMP-9 expression

Figure 4 Effect of ATO on radiation-induced MMP-9 expression in LLC cells and tumors. (a) The treatment of primary tumors in xenograft is described in Materials and methods. After 3 and 6 days, total RNA of the primary tumors was extracted and RT– PCR was used for semiquantification. The PCR products (MMP-9, MMP-2, TIMP-1, MMP-3) and b-actin were separated by 2% agarose gel electrophoresis. (b) LLC-LM cells pretreated or not with 1.0 mM ATO for 24 h and following 10 Gy irradiation. Total RNA was extracted and RT–PCR was used for semiquantification. The samples were probed to MMP-9, MMP-2, and TIMP-1 and separated by 2% gel electrophoresis

It is important to identify nuclear factors that are responsible for the inhibition of irradiation-induced MMP-9 expression by arsenic. To achieve this, a 0.7-kb segment at the 50 -flanking region of the human MMP-9 promoter region, containing both conserved binding sites for nuclear factor kB (NF-kB) and activator protein-1 (AP-1), was cloned into a pGL3-Basic vector for reporter gene assay. From Figure 6a, it is clear that MMP-9 reporter gene activity increased by up to three or four times when CaSki cells were exposed to 10 Gy irradiation. The induction of MMP-9 promoter activity by irradiation was significantly inhibited by decoy-kB dose-dependently, compared to when treated with scrambled oligonucleotides (Figure 6a). In contrast, decoy-AP-1 did not significantly influence MMP-9 promoter activity, indicating that NF-kB is specifically involved in inhibiting radiation-induced MMP-9 expression. Consistent with NF-kB/Rel-specific decoy oligonucleotides, NF-kB-specific chemical inhibitors, pyrrolidinedithiocarbamate (PDTC) and BAY 117082, abolished radiation-induced MMP-9 promoter activity (Figure 6b). Importantly, arsenic also markedly reduced radiation-induced MMP-9 promoter activity dose-dependently. NF-kB activation requires the 26S proteasome-mediated degradation of IkBs, thus releasing NF-kB from the complex to translocate to the nucleus and activate genes (Ghosh et al., 1998). Figure 6c shows that on 10 Gy irradiation, cytoplasmic IkBa was Oncogene


394

Figure 5 Effect of ATO on radiation-induced MMP-9 synthesis and activity. (a) CaSki cells were pretreated with ATO (0, 0.5, or 1.0 mM) for 24 h before exposed to 10 Gy irradiation. Total RNA was extracted 1 and 6 h after irradiation, and RT–PCR was used for semiquantification. The PCR products (MMP-9, MMP-2, MMP-3) and b-actin were separated by 2% agarose gel electrophoresis. (b) CaSki cells were pretreated or nor with ATO (0.5 or 1.0 mM) for 24 h and following 10 Gy irradiation. Conditioned media (CM) were collected 24 h after irradiation. A 10 mg protein sample was separated by SDS–PAGE and the level of MMP-9 protein was determined by Western blot analysis. (c) MMP-9 gelatinolytic activities of CM were analysed with gelatin zymography

significantly degraded, causing NF-kB P65 nuclear translocation. Meanwhile, these effects could be effectively reversed by pretreatment with arsenic or PDTC. These experimental results indicate that NF-kB activity is crucial to the blockade of radiation-induced MMP-9 by arsenic.

Discussion Many compounds that modify the radiation response of mammalian cells have been found to increase the lethality of radiation. For example, halogenated pyrimidines weaken the DNA chain by incorporating into the DNA in place of thymidine, thus sensitizing the cells to g-ray damage (McGinn et al., 1996). Moreover, hypoxic cell sensitizers, for example, misonidazole or Oncogene

Figure 6 Role of NF-kB in the inhibition of radiation-induced MMP-9 expression by arsenic. (a) Radiation activates MMP-9 promoter activity in an NF-kB-dependent manner. CaSki cells were transfected with indicated oligonucleotides 30 min before 5 mg MMP-9 promoter luciferase plasmid was introduced. The transient transfected cells were starved of serum for 16 h and then exposed to 10 Gy irradiation. After 2 h, the MMP-9 promoter activities were determined by luciferase. Three independent triplicate experiments were performed. (b) Effect of ATO on MMP-9 promoter luciferase activity. Transient transfected CaSki cells were pretreated with ATO (24 h) or NF-kB chemical inhibitors (1 h) before irradiation. MMP-9 luciferase activity was measured 2 h after irradiation. Three independent triplicate experiments were performed. (c) ATO inhibits radiation-induced NF-kB nuclear translocation. CaSki cells were pretreated with ATO (24 h) or PDTC (1 h) before irradiation. At 2 h after irradiation, 10 mg of the nuclear protein and cytoplasmic fraction were extracted and were subjected to immunoblotting with either rabbit polyclonal to NF-kB or mouse monoclonal to I-kB. a-Tubulin served as an internal control


395

etanidazole, increase the radiosensitivity of hypoxic cells by fixing damage produced by free radicals (Inanami et al., 2002). Taxanes, a widely used chemotherapeutic agent, can act synergistically with radiotherapy by inducing microtubule stabilization, thus arresting cells in the radiosensitive G2/M phase of the cell cycle (Liebmann et al., 1994). Recent works have demonstrated that adjunct use of radiation with monoclonal antibodies or small molecule tyrosine kinase inhibitors targeting EGF receptor produces notable synergistic antitumor effects by blocking MAP kinase-mediated cytoprotective response (Azria et al., 2003). All of these strategies for improving therapeutic efficacy have offered practical gains in clinical radiation therapy; however, considerable room remains for improving the combined treatment strategies. Radiotherapy paradoxically increased tumor invasiveness at doses used clinically as a fractionated irradiation, implying that sublethal doses of irradiation could promote tumor migration and distant metastasis (Qian et al., 2002). The use of radiation to eradicate a primary LLC has been found to cause accelerated growth of lung metastasis (Camphausen et al., 2001). These findings form the basis of combined arsenic treatment with radiotherapy. At low concentrations (1 mM), the experimental results in this study demonstrated the antimetastasis property of arsenic both in vitro and in vivo. When radiation and arsenic were administered together, arsenic could inhibit radiationinduced MMP-9 expression and thus prevent tumor cells escaping from cumulatively lethal dosage of irradiation and reduce the probability of distant lung metastasis. Other studies using higher concentrations of ATO demonstrated that arsenic preferentially shuts down tumor blood flow and enhances tumor response to fractionated radiation in locally advanced murine tumor (Lew et al., 2002). The remarkable findings of this study indicate an effective strategy for combing arsenic and radiotherapy to combat cancer metastasis. The proinvasive effects of sublethal irradiation are associated with enhanced avb3-integrin expression, an altered profile of MMP-2 and -9 expression and activity, altered MT1-MMP and TIMP-2 expression, and altered BCL-2/BAX rheostat (Wild-Bode et al., 2001). Consequently, pharmaceutical inhibitors of MMPs or avb3-integrin expression have successfully prevented radiation-induced tumor cell invasion (Qian et al., 2002; Wick et al., 2002). Based on the presented data, radiation significantly induced MMP-9 expression in human cervical cancer cells and also in LLC cells. MMP-9 is a rate-limiting extracellular protease involved in cell migration across basement membranes (Yu and Stamenkovic, 1999). MMP-9 increases cancer cell proliferation, reduces cancer cell apoptosis, and triggers tumor angiogenesis (Egeblad and Werb, 2002). Also, MMP-9 regulates the bioavailability characteristics of lung-specific chemokines involved in recruiting malignant cells, and thus potentiates pulmonary metastasis formation (van Kempen and Coussens, 2002). In a mouse model, downregulation of MMP-9 by antisense oligonucleotides reduced tumor metastasis

(Kondraganti et al., 2000). In experimental metastasis assays, the number of colonies formed in the lungs of mice is reduced in the MMP-9-null mice as compared with wild-type mice (Itoh et al., 1999). Accordingly, marked inhibition of MMP-9 by ATO clearly contributes to its antimetastatic effect. However, whether other molecules that reduce metastatic burden during radiation, such as angiostatin (Gorski et al., 2003), are regulated by arsenic remains uncertain. Paradoxically, single use of sodium arsenite has been shown to cause a marked increase in MMP-9 in prostate epithelial cells (Achanzar et al., 2002). The potential reasons for the difference of MMP9 response to arsenic exposure might be the different experimental design and arsenic dosage. In the study by Achanzar et al., no discernible increase in MMP-9 activity from RWPE-1 cells was detected until after 29 weeks of continuous 5 mM arsenite exposure. Reporter gene assay demonstrates that radiation directly activates MMP-9 mRNA expression in an NF-kB-dependent manner (Figure 6a). NF-kB is a transcriptional factor that exists in a latent form with the inhibitory protein IkB. Degradation of IkBs enables transient translocation of NF-kB into the nucleus and regulates induction of various genes via decameric kBbinding sites (Verma et al., 1995). NF-kB appears responsive following clinically relevant doses of radiation. The activation of NF-kB mediates the synthesis of inflammatory cytokines (Zhou et al., 2001) and the modulation of radiosensitivity. Several studies suggested that NF-kB promoted radiosensitivity (Yang et al., 2000); however, accumulating evidences demonstrated that radiosensitization can be achieved by inhibiting NF-kB activation in many cancer cell types (Yamagishi et al., 1997; Eichholtz-Wirth and Sagan, 2000; Kato et al., 2000; Didelot et al., 2001; Russo et al., 2001). Consistent with these observations, the inhibition of radiation-mediated NF-kB activation directly suppressed MMP-9 synthesis and activity, which contributes to arsenic-inhibited tumor invasion and metastasis. To date, consensus is still lacking regarding the activation mechanism for radiation regulation of NFkB activity. Depending on the cell lines or dosage applied, radiation appears to evoke distinct signaling events to activate NF-kB. These events include protein tyrosine kinase, PKC, ROS, ATM, DNA-PK, IkB kinase, and the ubiquitin–proteasome pathway (Criswell et al., 2003). Radiation-induced IkB degradation was significantly suppressed by ATO, resulting in reduced NF-kB nuclear translocation (Figure 6c). Supporting evidence from a previous study demonstrated that the IkBa protein is stabilized post-transcriptionally, most likely due to inhibition of the IKK complex at low and high concentrations of arsenic (Mathas et al., 2003). Furthermore, ATO binds to the cysteine residue (Cys179) in the activation loop of the IKK catalytic subunits IKKa/b (Kapahi et al., 2000). Integrating chemotherapeutic agents and radiotherapy has been found to improve overall survival rates for over a dozen human cancers, including lung, breast, Oncogene


396

head and neck, and cervical cancer (Choy et al., 1998; Rose et al., 1999; Isaac et al., 2002; Forastiere et al., 2003). The emerging rationale for combining radiation and chemotherapy includes improving locoregional tumor control and coping with systemic micrometastasis (Choy and Kim, 2003). The analytical results presented here demonstrate that concurrent arsenic–radiation therapy could not only achieve local tumor control but also inhibit distant metastasis. Inhibition of NF-kB activity and MMP-9 synthesis by arsenic uncovers the specific mechanisms underlying these remarkable responses. Moreover, the experiments were performed using concentrations of arsenic that are achievable in patients and use of arsenic generally not leading to myelosuppression. The findings thus demonstrate the feasibility of using arsenic as an adjuvant agent with radiation in cancer therapy.

5 mg/ml DNase-free RNase), and analysed using a Becton Dickinson FACScan and the Cell Quest software (Becton Dickinson Immunocytometry Systems, San Jose, CA, USA). In vitro invasion assays Invasion chambers were prepared by coating with 0.2 mg/ml gelatin (Sigma) to 24-well, 8 mm pore, cell culture inserts and left to set for 1 h at 371C. CaSki cells were treated with the indicated dose of ATO for 24 h and then detached with 0.25% Tris-EDTA. A total of 105 cells were seeded to the upper chamber, suspended in 200 ml DMEM, and 500 ml of serumfree medium was added to the lower chamber. After attaching the cells to the insert, the medium was changed to serum-free medium and exposed to radiation. After 16 h incubation, the cells on the upper surface were removed using a cotton bud, then the remaining migrated cells were fixed in methanol at 201C for 1 h and stained with 0.1% crystal violet for 1 h at room temperature, and finally migrated cells were counted at 200 magnification in 10 different fields on each filter. Experiments were repeated three times.

Materials and methods Cell culture

In vitro experiments

HeLa, CaSki, C33A, SiHa cells, and mouse LLCs were obtained from the American Type Culture Collection (Rockville, MD, USA). Tumor cells were grown in tissue culture dishes as monolayers at 371C in 5% CO2 in DMEM with 10% heat-inactivated fetal bovine serum plus penicillin–streptomycin under sterile tissue culture conditions. Logarithmically growing cells were treated with arsenic compounds as described for each of the respective experiments.

For the intravasation and invasion experiment, HeLa cells were cultured as described above, and incubated with or without 2.5 mM ATO for 24 h before beginning the experiment. Cells were detached from the culture dish using 0.25% TrisEDTA, counted, resuspended (106 cells) in 50 ml of PBS with Ca2 þ and Mg2 þ , and inoculated onto a CAM of a 9-day-old chick embryo, in which an artificial air sac was created (designated ‘upper CAM’). Following 48 h of incubation, the lower half of the CAM (designated ‘lower CAM’) was removed and stored frozen at 801C. The frozen tissue was crushed in lysis buffer with sterile 5 ml pipettes. DNA extraction and human Alu sequence PCR amplification were performed as described previously (Kim et al., 1998). Briefly, genomic DNA of the frozen tissue was isolated from lower CAMs using the DNA Isolation Kit (QIAGEN Systems) as per the specifications of the manufacturer. The specific primer for human Alu sequences comprised Alu-sense 50 ACG CCT GTA ATC CCA GCA CTT 30 and Alu-antisense 50 TCG CCC AGG CTG GAG TGC A 30 , which produced a band of 224 bp. The PCR conditions were as follows: 951C for 10 min, 951C for 30 s, 581C for 45 s, 721C for 45 s, and 721C for 10 min. Finally, the PCR products were electrophoresed on an 1.8% agarose gel at 100 V and visualized using ethidium bromide.

Antibodies and reagents Arsenical stock solutions were from TTY Biopharm (Taipei, Taiwan). To prevent oxidation of ATO, stock solutions were prepared immediately before each experiment. PDTC and BAY 11-7082 were obtained from Calbiochem (San Diego, CA, USA). The antibodies (a-tubulin, anti-MMP-9, anti-p65, and anti-IkB) were purchased from Santa Cruz Biotechnology Inc. (Paso Robles, CA, USA). Irradiation Cells were grown to 70–80% confluence in DMEM and serum starved for 16 h. The cells were irradiated with g-irradiation (10 Gy), using a JL Shepherd Mark I Irradiator [137Cs] with a dose rate of 3.6 Gy/min. Following irradiation, the cells were incubated at 371C for the indicated time. Cell viability assays Cells were plated into 96-well microplates at a density of 5 103 cells/well for cell viability assay. For the assay, the cells were cultured at 371C for the indicated time, 30 ml of MTT solution (5 mg/ml) was added to each well, and then the cells were incubated for 4 h in darkness. The formazan grain was then dissolved in DMSO, and the absorbance at 570 nm was read with an ELISA plate reader. Determination of subdiploid DNA content Cells (106/ml) were stimulated as indicated and harvested by centrifugation. Cell cycle analysis was performed by quantifying the DNA content using propidium iodide staining (PBS containing 0.5% Tween 20, 15 mg/ml propidium iodide, and Oncogene

Animals model and tumor irradiation Male, 4–6-week-old C57BL/6 mice (National Taiwan University Animal Center, Taiwan) were used. Tumor injections containing 2 106 Lewis lung cells were made s.c. into the right hind limb. At 7 days after implantation, the animals were randomized into one of four groups: control, irradiation (IR) only, ATO only, and IR/ATO. For tumor irradiation, mice were immobilized in a customized harness that exposed the right hind leg while shielding the remainder of the body by 3.5 cm of lead. Mice were irradiated in a cobalt 60 (Picker V9) source using single fraction of 30 Gy operating at 1.05 Gy/min. ATO treatment was started 1 day before radiation exposure and repeated twice weekly according to the following protocol (1 mg/kg, i.p.; the plasma concentration of ATO was 0.6– 1 mM). The animals were killed (n ¼ 5 per group) at the indicated time and the primary tumor and lung were dissected and used in the following experiments.


397 Histomorphometric examination The lung dissected was fixed in formaldehyde. Embedded tissue was serially sectioned into 5-mm sections every 200 mm. Paraffin sections were deparaffinized in xylene, rehydrated with graded ethanol, and antigen retrieval was performed by heating at 901C for 10 min in 10 mM sodium citrate (pH 6.0) and stained with hematoxylin–eosin (HE). The number of micrometastasis was quantified by counting lung colonies presented in the total tissue area per lung section using light microscopy (magnification, 40). RT–PCR Reverse transcription of RNA isolated from cells was performed in a final reaction volume of 20 ml containing 5 mg of total RNA in Moloney murine leukemia virus (MMLV) reverse transcriptase buffer (Promega, Madison, WI, USA), which consists of 10 mM dithiothreitol, all four deoxynucleoside 50 -triphosphates (dNTPs; each at 2.5 mM), 1 mg of (dT)12– 18 primer, and 200 U of MMLV reverse transcriptase (Promega). The reaction mixture was incubated at 371C for 2 h, and the reaction was terminated by heating at 701C for 10 min. A 1 ml portion of the reaction mixture was then amplified by PCR with the following pairs of primers: human(h) MMP-2 (581C): sense 50 -CCA CGT GAC AAG CCC ATG GGG CCC C-30 and antisense 50 -GCA GCC TAG CCA GTC GGA TTT GAT G-30 ; hMMP-3 (551C): sense 50 GAA AGT CTG GGA AGA GGT GAC TCC AC-30 and antisense 50 -CAG TGT TGG CTG AGT GAA AGA GAC CC-50 ; hMMP-9 (601C): sense 50 -CAA CAT CAC CTA TTG GAT CC-30 and antisense 50 -CGG GTG TAG AGT CTC TCG CT-30 ; b-actin (601C): sense 50 -CTT CTA CAA TGA GCT GCG TG-30 and antisense 50 -TCA TGA GGT AGT CAG TCA GG-30 ; mouse(m) MMP-2 (561C): sense 50 -GAG TTG GCA GTG CAA TAC CT-30 and antisense 50 -GCC ATC CTT CTC AAA GTT GT-30 ; mMMP-3 (501C): sense 50 TTC TCC AGG ATC TCT GAA GGA GAG G-30 and antisense 50 -GAT GTC CTC GTG GTA CCC ACC AAG T30 ; and mMMP-9 (601C): sense 50 -AGT TTG GTG TCG CGG AGC AC-30 and antisense 50 -TAC ATG AGC GCT TCC GGC AC-30 ; mTIMP-1 (581C): sense 50 -CTG GCA TCC TCT TGT TGC TA-30 and antisense 50 -AGG GAT CTC CAA GTG CAC AA-30 . A total of 30 cycles were performed, with each cycle comprising 1 min at 941C, 1 min at 50–601C depending on each gene, and 1.5 min at 721C with a final extension of 8 min at 721C. All amplifications were conducted within the linear range of the assay, and human b-actin reactions were amplified for 25 cycles. The reaction products were separated on 2% agarose gel, stained with 1 mg/ml ethidium bromide, and visualized using a UVP GDS-7900 Digital imaging system. The results were confirmed by conducting at least three replicate experiments. Gelatin zymography The supernatants of CaSki cells were electrophoresed for analysis in 9% SDS–PAGE gels containing 1 mg/ml gelatin. The gels were washed for 30 min at room temperature in 2.5%

Triton X-100, and then washed several times with ddH2O. The gels were incubated in 50 mM Tris (pH 7.6), 1 mM ZnCl2, 0.15 M NaCl, and 10 mM CaCl2 for 18 h at 371C. Following incubation, the gels were stained with 0.2% Coomassie blue R250. Bands of lysis representing gelatinase activity were then visualized against a dark background. Western blot analysis Protein extraction from CaSki cells was performed as described previously (Wei et al., 2001). A 50 mg protein sample was separated using 12.5% SDS–PAGE, transferred onto polyvinylidene difluoride (PVDF) membrane, and immunoblotted with various antibodies. Bound antibodies were then detected using appropriate peroxidase-coupled secondary antibodies, followed by use of an enhanced chemiluminescent detection system (ECL, Boehringer Mannheim). Reporter gene assay A 0.7-kb segment at the 50 -flanking region of the human MMP-9 promoter region was amplified by PCR using specific primers from the human MMP-9 gene (accession no. D10051): 50 -ACAATCGAGCTCCTGAAGGAAGAGAGTA AAGC (forward/SacI) and 50 -AATCCCAAGCTTATGGTG AGGGCAGAGGTG (reverse/HindIII). The pGL3-Basic vector containing a polyadenylation signal upstream from the luciferase gene was used to construct expression vectors by subcloning PCR-amplified DNA of MMP-9 promoter into the SacI/HindIII site of the pGL3-Basic vector. The PCR products were confirmed based on their size as determined by electrophoresis and DNA sequencing. CaSki cells (5 105) were transiently transfected with 5 mg of MMP-9 promoter plasmid by Transfast Transfection Reagent (Promega). This study used cotransfection with pSV-b-galactosidase, and conducted data normalization following all transient transfections using triplicate cultures, repeating every cotransfection protocol at least three times. The transient transfected cells were starved of serum for 16 h and then treated with different protocols as described in figure legends. Cell lysate luciferase activity was measured using the Luciferase Enzyme Assay System (Promega). Statistical analysis Statistical evaluation of the data was performed with a twotailed Student’s t-test for simple comparison between two values when appropriate. All statistical analyses were performed with SPSS, version 10.0 (SPSS, Chicago, IL, USA). P-values of less than 0.05 were considered statistically significant. Acknowledgements We thank Dr Yung-Ming Jeng (Department of Pathology, National Taiwan University Hospital) for review of histological work presented in this study and Julie Yu for carefully reading the manuscript. This work was supported by grant from National Taiwan University Hospital 92-S016.

References Achanzar WE, Brambila EM, Diwan BA, Webber MM and Waalkes MP. (2002). J. Natl. Cancer Inst., 94, 1888–1891. Azria D, Larbouret C, Robert B, Culine S, Ychou M, Verrelle P, Dubois JB and Pelegrin A. (2003). Bull Cancer, 90 (Spec No.), S202–S212.

Baker AH, Edwards DR and Murphy G. (2002). J. Cell Sci., 115, 3719–3727. Camphausen K, Moses MA, Beecken WD, Khan MK, Folkman J and O’Reilly MS. (2001). Cancer Res., 61, 2207–2211. Oncogene


398 Chambers AF and Matrisian LM. (1997). J. Natl. Cancer Inst., 89, 1260–1270. Chambers AF, Groom AC and MacDonald IC. (2002). Nat. Rev. Cancer, 2, 563–572. Choy H, Akerley W, Safran H, Graziano S, Chung C, Williams T, Cole B and Kennedy T. (1998). J. Clin. Oncol., 16, 3316–3322. Choy H and Kim DW. (2003). Semin. Oncol., 30, 3–10. Christofori G and Semb H. (1999). Trends Biochem. Sci., 24, 73–76. Chun YJ, Park IC, Park MJ, Woo SH, Hong SI, Chung HY, Kim TH, Lee YS, Rhee CH and Lee SJ. (2002). FEBS Lett., 519, 195–200. Criswell T, Leskov K, Miyamoto S, Luo G and Boothman DA. (2003). Oncogene, 22, 5813–5827. Didelot C, Barberi-Heyob M, Bianchi A, Becuwe P, Mirjolet JF, Dauca M and Merlin JL. (2001). Int. J. Radiat. Oncol. Biol. Phys., 51, 1354–1360. Egeblad M and Werb Z. (2002). Nat. Rev. Cancer, 2, 161–174. Eichholtz-Wirth H and Sagan D. (2000). Apoptosis, 5, 255–263. Folkman J. (2002). Semin. Oncol., 29, 15–18. Forastiere AA, Goepfert H, Maor M, Pajak TF, Weber R, Morrison W, Glisson B, Trotti A, Ridge JA, Chao C, Peters G, Lee DJ, Leaf A, Ensley J and Cooper J. (2003). N. Engl. J. Med., 349, 2091–2098. Ghosh S, May MJ and Kopp EB. (1998). Annu. Rev. Immunol., 16, 225–260. Gorski DH, Mauceri HJ, Salloum RM, Halpern A, Seetharam S and Weichselbaum RR. (2003). Cancer Res., 63, 308–311. Hughes MF. (2002). Toxicol. Lett., 133, 1–16. Inanami O, Sugihara K, Okui T, Hayashi M, Tsujitani M and Kuwabara M. (2002). Int. J. Radiat. Biol., 78, 267–274. Isaac N, Panzarella T, Lau A, Mayers C, Kirkbride P, Tannock IF and Vallis KA. (2002). Cancer, 95, 696–703. Itoh T, Tanioka M, Matsuda H, Nishimoto H, Yoshioka T, Suzuki R and Uehira M. (1999). Clin. Exp. Metast., 17, 177– 181. Kapahi P, Takahashi T, Natoli G, Adams SR, Chen Y, Tsien RY and Karin M. (2000). J. Biol. Chem., 275, 36062–36066. Kato T, Duffey DC, Ondrey FG, Dong G, Chen Z, Cook JA, Mitchell JB and Van Waes C. (2000). Head Neck, 22, 748–759. Kim J, Yu W, Kovalski K and Ossowski L. (1998). Cell, 94, 353–362. Kondraganti S, Mohanam S, Chintala SK, Kin Y, Jasti SL, Nirmala C, Lakka SS, Adachi Y, Kyritsis AP, Ali-Osman F, Sawaya R, Fuller GN and Rao JS. (2000). Cancer Res., 60, 6851–6855. Kwong YL and Todd D. (1997). Blood, 89, 3487–3488. Lew YS, Kolozsvary A, Brown SL and Kim JH. (2002). Cancer Res., 62, 4202–4205. Liebmann J, Cook JA, Fisher J, Teague D and Mitchell JB. (1994). J. Natl. Cancer Inst., 86, 441–446. Lukashev ME and Werb Z. (1998). Trends Cell Biol., 8, 437–441. Mathas S, Lietz A, Janz M, Hinz M, Jundt F, Scheidereit C, Bommert K and Dorken B. (2003). Blood, 102, 1028–1034.

Oncogene

McGinn CJ, Shewach DS and Lawrence TS. (1996). J. Natl. Cancer Inst., 88, 1193–1203. Miller Jr WH, Schipper HM, Lee JS, Singer J and Waxman S. (2002). Cancer Res., 62, 3893–3903. Miyamoto S, Yano K, Sugimoto S, Ishii G, Hasebe T, Endoh Y, Kodama K, Goya M, Chiba T and Ochiai A. (2004). Cancer Res., 64, 665–671. Murgo AJ. (2001). Oncologist, 6 (Suppl 2), 22–28. Overall CM and Lopez-Otin C. (2002). Nat. Rev. Cancer, 2, 657–672. Qian LW, Mizumoto K, Urashima T, Nagai E, Maehara N, Sato N, Nakajima M and Tanaka M. (2002). Clin. Cancer Res., 8, 1223–1227. Rose PG, Bundy BN, Watkins EB, Thigpen JT, Deppe G, Maiman MA, Clarke-Pearson DL and Insalaco S. (1999). N. Engl. J. Med., 340, 1144–1153. Russo SM, Tepper JE, Baldwin Jr AS, Liu R, Adams J, Elliott P and Cusack Jr JC. (2001). Int. J. Radiat. Oncol. Biol. Phys., 50, 183–193. Shen ZX, Chen GQ, Ni JH, Li XS, Xiong SM, Qiu QY, Zhu J, Tang W, Sun GL, Yang KQ, Chen Y, Zhou L, Fang ZW, Wang YT, Ma J, Zhang P, Zhang TD, Chen SJ, Chen Z and Wang ZY. (1997). Blood, 89, 3354–3360. Sienel W, Hellers J, Morresi-Hauf A, Lichtinghagen R, Mutschler W, Jochum M, Klein C, Passlick B and Pantel K. (2003). Int. J. Cancer, 103, 647–651. Soignet SL, Maslak P, Wang ZG, Jhanwar S, Calleja E, Dardashti LJ, Corso D, DeBlasio A, Gabrilove J, Scheinberg DA, Pandolfi PP and Warrell Jr RP. (1998). N. Engl. J. Med., 339, 1341–1348. Sporn MB. (1996). Lancet, 347, 1377–1381. Sternlicht MD and Werb Z. (2001). Annu. Rev. Cell Dev. Biol., 17, 463–516. Strand S, Vollmer P, van den Abeelen L, Gottfried D, Alla V, Heid H, Kuball J, Theobald M, Galle PR and Strand D. (2004). Oncogene, 23, 3732–3736. Tanioka Y, Yoshida T, Yagawa T, Saiki Y, Takeo S, Harada T, Okazawa T, Yanai H and Okita K. (2003). Br. J. Cancer, 89, 2116–2121. van Kempen LC and Coussens LM. (2002). Cancer Cell, 2, 251–252. Verma IM, Stevenson JK, Schwarz EM, Van Antwerp D and Miyamoto S. (1995). Genes Dev., 9, 2723–2735. Wei LH, Kuo ML, Chen CA, Chou CH, Cheng WF, Chang MC, Su JL and Hsieh CY. (2001). Oncogene, 20, 5799–5809. Wick W, Wick A, Schulz JB, Dichgans J, Rodemann HP and Weller M. (2002). Cancer Res., 62, 1915–1919. Wild-Bode C, Weller M, Rimner A, Dichgans J and Wick W. (2001). Cancer Res., 61, 2744–2750. Yamagishi N, Miyakoshi J and Takebe H. (1997). Int. J. Radiat. Biol., 72, 157–162. Yang CR, Wilson-Van Patten C, Planchon SM, WuerzbergerDavis SM, Davis TW, Cuthill S, Miyamoto S and Boothman DA. (2000). FASEB J., 14, 379–390. Yu Q and Stamenkovic I. (1999). Genes Dev., 13, 35–48. Zhou D, Yu T, Chen G, Brown SA, Yu Z, Mattson MP and Thompson JS. (2001). Int. J. Radiat. Biol., 77, 763–772. Zucker S and Vacirca J. (2004). Cancer Metast. Rev., 23, 101–117.