Solid tumors generally possess hypoxic areas in their central portion because of decreased vascular supply associated with the effects of treatment and the originally increased energy demand of cancer cells, and the hypoxic tissue is one of the serious matters for consideration in the control of malignant tumors. Most tumor cells possess the ability to undergo apoptosis in response to hypoxic conditions. However, tumor cells can adapt to hypoxic conditions by employing a variety of survival tools, which result in the promotion of cancer cell growth and metastasis.1, 2, 3 This adaptation of cancer cells to hypoxia is mainly mediated by the transcription factor hypoxia-inducible factor-1 (HIF-1).4 HIF-1 is a heterodimeric transcription factor consisting of an oxygen-regulated α subunit (HIF-1α) and a stable nuclear factor, HIF-1β/aryl hydrocarbon receptor nuclear translocator (ARNT). Under normoxic conditions, HIF-1α is rapidly degraded by the proteosome after being targeted for ubiquitination. HIF-1α translocates to the nucleus under hypoxic conditions and forms an active complex with HIF-1β; the complex binds to the hypoxia-response element (HRE) in the target genes, which results in the transactivation of these genes.5 Proteins encoded by such genes contribute to the blood supply, energy production, growth/survival, invasion/metastasis and resistance.
It has been frequently reported that hypoxic conditions in tumor tissues are associated with the resistance of tumor cells to radiotherapy and chemotherapy.6, 7 In fact, the association of HIF-1α protein over-expression with cell proliferation and poor prognosis has been observed in many kinds of human cancers.8 In addition, recent data indicate that the elevated level of HIF-1α is closely correlated with radio-resistance and chemo-resistance of tumor cells and that HIF-1α inhibits the induction of apoptosis in tumor cells.9, 10, 11, 12, 13, 14, 15 However, the precise mechanism of the hypoxia-induced resistance of tumor cells is not fully known.
P-glycoprotein (P-gp), a membrane-resident glycoprotein encoded by the multidrug resistance (MDR1) gene, works to decrease the intracellular concentration of many types of chemotherapeutic drugs by acting as a drug efflux pump and plays an important role in drug resistance in cancer therapy.16, 17MDR1 gene expression with subsequent functional P-gp expression is markedly upregulated in a HIF-1-dependent manner in response to hypoxia.18, 19 Hypoxia-elicited MDR1 expression resulting from HIF-1 activation depends at least in part on signaling via activation of JNK.20 Therefore, the inhibition of HIF-1 or JNK may prevent the development of multidrug resistance by the abrogation of P-gp upregulation.
Many chemotherapeutic drugs and γ-rays induce the generation of reactive oxygen species (ROS) in tumor cells. When the level of generated ROS exceeds the ROS-scavenging activity of the tumor cells, the cells are impaired to some degree, depending on the amount of excessive ROS. We previously reported that manganese-superoxide dismutase (Mn-SOD) antisense upregulates the apoptosis of oral squamous cell carcinoma (OSCC) cells induced in vivo by γ-rays by regulating the expression of the Bcl-2 family proteins COX-2 and p21Cip1/WAF1.21 Thus, the activity of scavengers such as SOD, catalase and glutathione peroxidase in tumor cells is closely associated with the susceptibility of the cells to anticancer agents. HIF-1 promotes the transcription of some genes such as heme oxygenase-1 (HO-1)22 and ceruloplasmin (Cp).23 HO-1 is a microsomal enzyme that catalyses the degradation of heme groups to yield biliverdin, iron and carbon monoxide. Biliverdin is subsequently converted to the antioxidant bilirubin by biliverdin reductase. HO-1 has also been identified as an inducer of Mn-SOD.24 Therefore, HO-1 seems to counteract oxidative damage and confer cytoprotection, as suggested by studies of deficient and increased HO-1 expression.25, 26, 27 Although HO-1 is induced by a variety of stress-inducing agents such as heavy metals, hyperthermia and ultraviolet irradiation, which generate ROS and/or diminish glutathione levels, the involvement of HO-1 in chemo- and radio-resistance is not fully understood. Cp is also an antioxidant protein that blocks free oxygen radical-induced proteolysis and DNA damage through sequestration of free copper ions and superoxide anion radical scavengers,28 and high levels of Cp expression have been demonstrated in various cancers such as thyroid carcinoma, melanoma and ovarian cancer.29, 30
In the present study, we attempted to explore how HIF-1α contributes to radio-resistance and chemo-resistance in OSCC cells. We showed here that the enhanced expression of HIF-1α in OSCC cells was associated with resistance to chemotherapeutic drugs and γ-ray-irradiation and that the downregulation of HIF-1α expression by small interfering RNA (siRNA) enhanced the susceptibility of OSCC cells to chemotherapeutic drugs and γ-ray-irradiation through the downregulation of P-gp, HO-1 and Cp expression and the upregulation of ROS.
OSCC cell lines (OSC-2, -4, -5 and -6 cells) established from patients with oral cancer in our laboratory were cultured in Dulbecco's modified Eagle's medium (DMEM, Nissui Pharmaceutical Co., Tokyo, Japan) supplemented with 10% (v/v) fetal bovine serum, 10 mM glutamine, 100 units/ml penicillin and 100 μg/ml streptomycin (Invitrogen Co., Carlsbad, CA) and treated with 100 μM 5-fluorouracil (5-FU), cis-diamminedichloroplatinum (CDDP) (Sigma-Aldrich Inc., St. Louis, MO) or irradiated with 30 Gy of 137Cs. Ascorbic acid (AA), cobalt chloride (CoCl2), cyclosporine A (CsA) and N-acetyl-L-cysteine (NAC) were obtained from Sigma-Aldrich Inc. Zinc protoporphyrin IX (ZnPP) was supplied by Calbiochem (La Jolla, CA). The details of the treatment are shown in the legends of each figure.
Plasmid construct and siRNA
Full-length human HIF-1α cDNA was cloned into the pcDNA3.1/V5-His TOPO expression vector (Invitrogen Co.). The fidelity of the constructed vector was confirmed by sequencing. HIF-1α-siRNA (5′-CUGAUGACCAGCAACUUGAtt-3′) and control scrambled siRNA (5′-AGUUCAACGACCAGUAGUCtt-3′) were synthesized by the Ambion siRNA facility (Austin, TX). Plasmids expressing a sequence corresponding to the MDR1 promoter (−194 to + 158) were constructed using pGL3-Basic (Promega, Madison, WI). Transfection was performed with Trans IT polyamine (Pan Vera, Madison, WI) 24 hr after cell seeding. Stable transfectants were selected by culturing the cells in medium containing geneticin (200–400 μg/ml), and cell lines were developed individually from surviving colonies.
Preparation of total cellular and nuclear extracts
Cells were solubilized with ice-cold lysis buffer containing 1% (v/v) Triton X-100, 50 mM NaCl, 25 mM Hepes (pH 7.4), 1 mM EDTA, 1 mM EGTA, 1 mM phenylmethylsulphonyl fluoride (PMSF), 20 μg/ml aprotinin, 10 μg/ml leupeptin and 1 μg/ml pepstatin. To isolate the total cellular fraction, the lysates were centrifuged at 12,500g for 15 min at 4°C and the clarified supernatants were used as total cellular extracts.
Nuclear extracts were isolated by resuspending cells in buffer A (50 mM Hepes, pH 7.5, 10 mM KCl, 1 mM EDTA, 1 mM dithiothreitol, 0.5 mM PMSF, 20 μg/ml aprotinin, 10 μg/ml leupeptin, 1 μg/ml pepstatin and 1 μg/ml soy bean trypsin inhibitor). Triton X-100 was added at a final concentration of 0.5% (v/v), and nuclei were pelleted by centrifugation for 30 sec at 12,500g. The nuclei were then resuspended in buffer C (same as buffer A except with 400 mM KCl) and centrifuged for 5 min at 12,500g after incubation for 20 min at 4°C. The supernatants containing the nuclear proteins were used for nuclear extracts. Total cellular and nuclear extracts were snap-frozen and stored at −80°C until use. Protein concentrations were determined using a BCA protein assay kit (Pierce Biotechnology, Rockford, IL).
Western blot analysis
Extracted proteins (50 μg/lane) were separated by SDS-polyacrylamide gel electrophoresis and transferred onto an Immobilon-P membrane (Immobilon, Millipore Corporation, Bedford, MA). Blocking was performed in Tris-buffered saline containing 5% (w/v) skim milk powder and 0.1% (v/v) Tween-20. The membranes were probed with the following diluted antibodies (Abs): anti-HIF-1α monoclonal Ab (Transduction Laboratories, Lexington, KY) at 1:500, anti-HO-1 polyclonal Ab (Santa Cruz Biotechnology, Santa Cruz, CA) at 1:200, anti-Cp polyclonal Ab (Santa Cruz Biotechnology) at 1:200, anti-P-gp monoclonal Ab (Calbiochem) at 1:1,000, anti-Mn-SOD monoclonal Ab (Chemicon International, Temecula, CA) at 1:200 and anti-β-actin monoclonal Ab (Santa Cruz Biotechnology) at 1:1,000. Detection was performed with an ECL system (Amersham, Piscataway, NJ).
RNA extraction and real-time quantitative RT-PCR analysis
Total cellular RNA was extracted using an RNeasy total RNA isolation system (QIAGEN, Valentia, CA). The RNA was quantitated by measuring the optical density at 260 nm. The extracted RNA (1 μg) was added to 20 μl of reverse transcription buffer (50 mM Tris-HCl, pH 8.3, 50 mM KCl, 10 mM MgCl2, 1 mM EDTA, 10 μg/ml bovine serum albumin and 1 mM DTT) containing 10 mM dNTP, 50 U RNase inhibitor, 1 μg oligo dT primer and 50 U of avian myeloblastosis virus reverse transcriptase (all from Takara Biomedicals, Kyoto, Japan). This mixture was incubated at 42°C for 40 min and heated at 99°C for 5 min.
PCR was performed using a master mix based on the TaqMan universal PCR master mix and run on the ABI PRISM 7000 sequence detection system (Applied Biosystems, Foster City, CA). Primers for HIF-1α mRNA and the TaqMan probe were obtained from Applied Biosystems. The expression level of the endogenous reference gene was determined using the commercially available human GAPDH TaqMan Predeveloped Assay Reagent (Applied Biosystems).
Cell proliferation assay
Cell proliferation was estimated by the 3-(4,5-dimethylthiazol-2-yl)-2, 5-diphenyl-2H-tetrazolium bromide (MTT) colorimetric assay. Cells (1.5 × 104 cells/well) were cultured in a 96-well microplate for 24 hr after irradiation with 30 Gy γ-rays or in the presence of 100 μM 5-FU or CDDP. After each treatment, the cells in each well were washed with 200 μl of phosphate buffered saline (PBS) and incubated with 5 mg/ml MTT solution (Sigma-Aldrich) at 37°C for 4 hr. The supernatants were then removed and the formazan crystals in each well were solubilized by the addition of 200 μl of dimethyl sulfoxide for 30 min. The colored formazan product was measured using a plate reader at a wavelength of 570 nm. Experiments were repeated 3 times with triplicate samples for each experiment.
After each indicated treatment, cells were trypsinized, washed once with complete medium and stained with propidium iodide and FITC-conjugated annexin V (Sigma-Aldrich), according to the manufacturer's instructions. The cells were then analyzed on a FACScan cytometer using CELLQUEST software (Becton Dickinson, San Jose, CA).
Electrophoretic mobility shift assay (EMSA)
Nuclear extracts were incubated with a 32P-end-labeled oligonucleotide probe encoding the HIF-1 binding site within the promoter region of the MDR1 gene (5′-AGGACAAGCGCCGGGGCGTGGGCTGAGCACAGCCGCTTC-3′), 1 μg of poly(dI)·poly(dC), 50 mM NaCl, 0.2 mM EDTA, 0.5 mM dithiothreitol, 2% (v/v) glycerol and 10 mM Tris-HCl, pH 7.6 for 30 min at 4°C. Proteins were separated by electrophoresis through a native 4% polyacrylamide gel in a running buffer consisting of 67 mM Tris-HCl, pH 7.5, 33 mM sodium acetate and 10 mM EDTA. The gel was dried and exposed to Kodak XOMAT film with intensifying screens for 5–15 hr at −70°C, and the image was analyzed using an image analyzer (BAS2000, Fuji Photo Film Co., Tokyo, Japan). The HIF-1 probe was labeled with [γ-32P] ATP (Amersham), using T4 polynucleotide kinase (Takara Biomedicals) and purified using a G-50 spin column.
Assay for intracellular concentrations of chemotherapeutic drugs
After cells were exposed to 100 μM CDDP and 100 nM [3H]5-FU (Daiichi Pure Chemicals Co., Tokyo, Japan) for 1 hr in the presence of 10 μM CsA, a P-gp inhibitor, the cells were rinsed with warm serum-free medium. At the time points of experiments indicated in the text, the cells were collected. The platinum content in the cell lysates was measured by atomic absorption spectrophotometry using a polarized Zeeman atomic absorption spectrophotometer (Z-9000, Hitachi Co., Ibaraki, Japan). For the [3H]5-FU concentration assay, the harvested cells were placed in vials containing scintillation fluid, and the radioactivity in the cells was counted.
Measurement of intracellular ROS levels and SOD activity
Intracellular ROS levels were measured using a fluorescent dye, dichlorofluorescein diacetate (DCFH-DA), which is converted to DCFH by esterases when it is taken up by cells. The treated cells were washed 3 times in 5 mM Hepes-buffered saline, pH 7.4, incubated in Hepes buffered saline containing 10 μM DCFH-DA for 15 min and analyzed on a FACScan cytometer (Becton Dickinson).
SOD activity was determined using a Superoxide Dismutase Assay Kit (Dojindo Laboratories, Kumamoto, Japan). The assay kit utilizes a tetrazolium salt for detection of superoxide radicals generated by xanthine oxidase hypoxanthine.31
Results are expressed as the mean ± SEM. Differences were compared by Mann–Whitney's U-test and p values of less than 0.05 were considered statistically significant.
Different expression of HIF-1α and susceptibility to γ-rays and chemotherapeutic drugs in OSCC cell lines
Of the four OSCC cell lines (OSC-2, -4, -5 and -6 cells) examined, OSC-2 and -4 cells expressed low levels of HIF-1α mRNA and protein, and OSC-5 and -6 cells expressed high levels under normoxic conditions (Fig. 1a). Treatment with γ-rays, CDDP and 5-FU induced nuclear translocation of HIF-1α protein in all of the examined cell lines (Fig. 1b). The MTT assay revealed that treatment of cells with the anticancer agents suppressed the cell growth (Fig. 1c). The suppressive effects were more remarkable (50–65%) in OSC-2 and -4 cells than in OSC-5 and -6 cells (about 30%). To determine whether the increased cytotoxicity reflected the induction of apoptosis, cells were analyzed by flow cytometry after annexin V/propidium iodide staining. The induction of apoptosis by treatment with irradiation and chemotherapeutic drugs was observed more clearly in OSC-2 and -4 cells than in OSC-5 and -6 cells (Fig. 1d).
Effects of HIF-1α-overexpression and silencing on susceptibility of OSCC cell lines to γ-rays and chemotherapeutic drugs
To explore the precise role of HIF-1α in chemoradiotherapy-induced apoptosis of OSCC cell lines, we transfected HIF-1α expression vector into OSC-2 and -4 cells and HIF-1α siRNA duplexes into OSC-5 and -6 cells. The overexpression of HIF-1α in OSC-2 and -4 cells and the suppression of the expression of HIF-1α in OSC-5 and -6 cells were confirmed by Western blot analysis of whole cell lysates derived from each cell line (Fig. 2a). Since the induction of apoptosis by anticancer agents was strongest and weakest in OSC-2 and -5 cells, respectively (Fig. 1d), we used these 2 cell lines in the following experiments.
When HIF-1α expression vector-transfected OSC-2 cells were treated with γ-rays, CDDP and 5-FU, the inhibitory effects of these anticancer agents were partially diminished compared to those in empty vector-transfected OSC-2 cells (Fig. 2b). In contrast, when HIF-1α-targeting siRNA duplex-transfected OSC-5 cells were treated with γ-rays and the chemotherapeutic drugs, the cell growth was more strongly inhibited as compared to that of scrambled siRNA-transfected OSC-5 cells. The induction of apoptosis by γ-rays and chemotherapeutic drugs was suppressed in the HIF-1α expression vector-transfected OSC-2 cells (Fig. 2c). In contrast, the transfection of HIF-1α-siRNA duplexes into OSC-5 cells promoted the apoptosis induced by these treatments.
Influence of HIF-1α overexpression and silencing on MDR1 promoter activity and P-gp expression in OSCC cell lines
The binding of radio-labeled HRE probe to the HIF-1 complex in nuclear extracts was low in OSC-2 and -4 cells, and high in OSC-5 and -6 cells (Fig. 3a). When OSC-2 and -4 cells were transfected with the HIF-1α expression vector, the binding was increased (Lane 2 and 5). In contrast, the binding of the HIF-1α complex in nuclear extracts to the HRE probe was decreased by the transfection of HIF-1α-siRNA duplexes into OSC-5 and -6 cells as compared with that in extracts from the control cells (Lane 8 and 11). Further, the binding of the HIF-1α complex was induced by the treatment of the cells with CoCl2 in all cell lines examined (Lane 3, 6, 9 and 12) and DNA–protein complexes in nuclear extracts from OSC-6 cells treated with CoCl2 were diminished by the addition of unlabeled probe (Lane 13). The luciferase reporter assay revealed that MDR1 promoter activity was low in OSC-2 and -4 cells and high in OSC-5 and -6 cells and that the activity was increased by HIF-1α over-expression in OSC-2 and OSC-4 cells and decreased by the transfection of HIF-1α-siRNA duplexes into OSC-5 and -6 cells (Fig. 3b). The levels of P-gp protein corresponded with the expression of HIF-1α and the binding of the HIF-1 complex to the HRE of P-gp probe (Fig. 3c).
Influence of HIF-1α overexpression and silencing on uptake and efflux of anticancer drugs in OSCC cell lines
We next examined the cellular uptake and efflux of CDDP and 5-FU in OSC-2 and -5 cells. As shown in Figure 4a, the amount of platinum in the cells at 1 h increased dose dependently with CDDP concentration in both types of cells, and the basal level in OSC-2 cells was higher than that in OSC-5 cells. The intracellular CDDP content in HIF-1α expression vector-transfected OSC-2 cells was lower than that in empty vector-transfected OSC-2 cells, whereas the intracellular CDDP level was slightly increased in HIF-1α-siRNA duplex-transfected OSC-5 cells compared with that in scrambled siRNA-transfected OSC-5 cells. There was no significant difference in the CDDP efflux of scrambled- and HIF-1α-siRNA duplex-transfected OSC-5 cells, although the intracellular platinum level in HIF-1α expression vector-transfected OSC-2 cells decreased more rapidly compared to that in empty vector-transfected OSC-2 cells (Fig. 4b). In accordance with the CDDP uptake, the uptake of 5-FU was suppressed by the transfection of HIF-1α expression vector into OSC-2 cells and enhanced by the transfection of HIF-1α-siRNA duplexes into OSC-5 cells (Fig. 4c). A P-gp inhibitor, CsA, suppressed the inhibitory effect of HIF-1α on the intracellular accumulation of chemotherapeutic drugs and enhanced the upregulatory effect of HIF-1α-siRNA duplexes on the accumulation (Figs. 4a and 4c). Moreover, the overexpression of HIF-1α in OSC-2 cells promoted and the knockdown of HIF-1α in OSC-5 cells suppressed the efflux of 5-FU, but these effects were slight (Fig. 4d).
Influence of HIF-1α overexpression and knockdown on the expression of HO-1 and Cp and ROS generation in OSCC cell lines
The expression levels of HO-1 and Cp in OSCC cell lines were upregulated and downregulated by the transfection of HIF-1α expression vector and HIF-1α-siRNA duplexes, respectively (Fig. 5a). The nuclear translocation of HIF-1α induced by γ-rays, CDDP and 5-FU was abolished by the pretreatment with an antioxidant, NAC, in all cell lines examined (Fig. 5b). Treatment with γ-rays and chemotherapeutic drugs induced the generation of ROS in OSC-2 and -5 cells (Fig. 5c). The level of production of ROS was decreased by transfection of the HIF-1α expression vector into OSC-2 cells (Fig. 5c, left panel). Conversely, the ROS level was increased by treatment of with each of γ-rays, CDDP and 5-FU in the HIF-1α-siRNA duplex-transfected OSC-5 cells (Fig. 5c, right panel). When HIF-1α expression vector-transfected OSC-2 cells and HIF-1α-targeted siRNA duplex-transfected OSC-5 cells were treated with antioxidants such as AA and NAC, anticancer agent-induced ROS production was decreased but not completely inhibited.
Involvement of HO-1 in anticancer agents-induced cytotoxicity and Mn-SOD activity
To explore the involvement of HO-1 in the susceptibility to γ-rays and chemotherapeutic drugs, we treated each type of cells with a specific competitive inhibitor of HO-1, ZnPP. When OSC-2 cells were pretreated with ZnPP, the effect of HIF-1α overexpression was inhibited (Fig. 6a, left panel). In contrast, the pretreatment of HIF-1α-targeted siRNA duplex-transfected OSC-5 with ZnPP had no effect on anticancer agents-induced cell growth inhibition (Fig. 6a, right panel). The Mn-SOD expression corresponded to the expression of HO-1 and the expression levels of Mn-SOD in OSCC cell lines were upregulated and downregulated by the transfection of HIF-1α expression vector and HIF-1α-siRNA duplexes, respectively (Fig. 6b). Furthermore, SOD activity was induced by the transfection of HIF-1α expression vector into OSC-2 cells and inhibited in the HIF-1α-siRNA duplex-transfected OSC-5 cells (Fig. 6c). The upregulation of SOD activity by the transfection of HIF-1α expression vector into OSC-2 cells was abolished by pretreatment with ZnPP.
It is well known that hypoxia of tumors is strongly associated with the progression of malignant phenotypes, poor prognosis and resistance to radiation and anticancer drugs.6, 7 HIF-1α overexpression has been detected in the majority of solid tumors, including breast, colon, ovarian and oral cancers.8 Some reports have shown a correlation between HIF-1α expression levels and resistance to radiation and chemotherapeutic drugs in oropharyngeal, breast and lung carcinomas.12, 13, 14, 15 However, there are some contradictory reports indicating that the susceptibility of tumor cells to radiation and chemotherapeutic drugs is not correlated with the expression level of HIF-1α,32, 33, 34, 35 which suggests that the involvement of HIF-1α in tumor malignant phenotypes varies among different tumor subtypes. On the basis of the functions of HIF-1-targeted gene products, however, it seems that HIF-1 is quite likely to contribute to tumor progression and chemo-radio-resistance. This suspicion is supported by the facts that inhibitors of HIF-1 inhibit tumor growth and enhance the tumor sensitivity to anticancer agents.36, 37 In the present study, however, OSCC cells did not become completely susceptible to chemotherapeutic drugs and γ-rays even though the expression of HIF-1α was completely knocked down by HIF-1α-siRNA duplexes. This result suggests that many factors other than HIF-1α are involved in the chemo-radio-resistance of cancer cells. It was reported that HIF-1α and HIF-2α are expressed at high levels in a variety of human tumors and transactivate overlapping sets of target genes.38
Many approaches have been taken to suppress the expression or the transcriptional activity of HIF-1 using antisense HIF-1α plasmid, small interfering RNA targeted against HIF-1α and chemical compounds such as an HSP90 inhibitor and a thioredoxin inhibitor.39, 40, 41 Although it is clear that these agents decrease the expression and transcriptional activity of HIF-1 in vitro, the precise mechanisms of the suppression and the antitumor effects of these agents remain to be determined.
Some studies have indicated that the effects of anticancer drugs on cancer cells are inhibited by several mechanisms. P-gp, which is one of the major resistance factors against anticancer drugs, seems to be closely associated with HIF-1.18, 19, 20 As is well known, P-gp makes tumor cells highly resistant to a variety of anticancer drugs, and the expression of P-gp is related to the high growth activity of the tumor cells and the poor prognosis of the tumor.16, 17 The fact that P-gp expression is regulated by HIF-1α suggests that HIF-1α expression is involved in chemo-resistance of OSCC cell lines. The present study revealed that the expression levels of P-gp increased in accordance with the HIF-1α expression levels in the 4 OSCC cell lines examined, and the increase of P-gp was intimately associated with the increase of the efflux of CDDP and 5-FU. Conversely, the efflux of these drugs was decreased when HIF-1α was downregulated, which was associated with a decrease of P-gp expression. Multi-drug resistance-related protein (MRP) modulates the toxicities of chemotherapeutic drugs. Although MRP1 was expressed in 4 OSCC cell lines, the overexpression and knockdown of HIF-1α did not affect the expression levels of MRP1, and a MRP inhibitor, probenecid, did not influence the retention/efflux of CDDP and 5-FU in OSCC cell lines (data not shown). These findings appear to indicate that not MRP1 but P-gp is an important molecule, which regulates the susceptibility of OSCC cell lines to chemotherapeutic drugs.
Apoptosis is induced via a variety of signaling pathways that are finally transmitted to caspase-3. In the apoptotic pathways, the Fas/Fas ligand system and mitochondria play important roles. Of the mitochondria-associated apoptosis signals, ROS and Bcl-2 family proteins are crucial. The role of ROS in the mitochondrial signal has been indicated by many investigators. Hydrogen peroxide and ROS-inducing agents such as chemotherapeutic drugs and γ-rays induce apoptosis, while antioxidants, including NAC, vitamin C and vitamin E, as well as ROS scavengers such as Mn-SOD and glutathione peroxidase inhibit apoptosis.42, 43, 44 In contrast, ROS act as signaling molecules mediating the response to growth factors, chemical stress and hormones. Interestingly, many researchers reported that ROS played a crucial role in the regulation of HIF-1α in a variety of cell types.45 However, the role of HIF-1α in the regulation of ROS generation and degradation is poorly understood. In the present study, we found that the nuclear translocation of HIF-1α protein induced by anticancer agents was suppressed by antioxidant. On the contrary, ROS production induced by anticancer agents was decreased by the transfection of the HIF-1α expression vector into OSC-2 cells and that the anticancer agent-induced ROS generation was increased in the HIF-1α-siRNA duplex-transfected OSC-5 cells. In addition, antioxidants such as AA and NAC decreased the susceptibility of OSC cells to the anticancer agents. Previous reports and our present results coordinately support the conclusion that the induction of HIF-1α by the treatment with anticancer agents is dependent on the elevation of ROS levels induced by these compounds and that HIF-1α acts as a suppressor of apoptosis by inhibiting ROS generation in cancer cells.
Of the HIF-1α target gene products, HO-1 and Cp have ROS scavenging activity. HO-1 is a key enzyme for the protection of cells against oxidative stress. Its overexpression in different types of human cancers appears to be related to the fact that HO-1 supports cellular growth and induces resistance against chemotherapy and radiotherapy.46 Our results indicated that the expression levels of HO-1 and HIF-1α were correlated with each other and that the expression levels of HO-1 were negatively correlated with the intracellular ROS levels. Recently, it was reported that HIF-1α induced Mn-SOD and the induced Mn-SOD suppressed HIF-1α upregulation in hypoxic conditions.47 In addition, HO-1 has been identified as an inducer of Mn-SOD.24 To explore the involvement of HIF-1α in Mn-SOD expression, we examined Mn-SOD levels in OSCC cells that were transfected with HIF-1α expression vector or HIF-1α-siRNA. The results showed that the Mn-SOD levels in the OSCC cells were upregulated by the transfection of HIF-1α expression vector and conversely downregulated by the siRNA. In addition, ZnPP, a specific HO-1 inhibitor, inhibited the upregulation of Mn-SOD activity by the vector. These results indicate that HIF-1α upregulates the degradation, decomposition or disintegration of ROS via HO-1 transactivation and Mn-SOD induction.
The details of the antioxidant properties of Cp have not been clarified yet, although the accumulated research indicates that the ferroxidase activity of Cp is the basis of its antioxidant effect.28 In OSCC cell lines, the expression levels of Cp were not correlated with the susceptibility of the cells to anticancer agents. However, the expression levels of Cp were upregulated by the transfection of HIF-1α expression vector into OSC-2 cells. We have no explanation of these contradictory results and it will be necessary to further explore the precise role of Cp in the susceptibility to anticancer drugs and radiation.
The expression of HIF-1α is regulated by a variety of mechanisms, including transcription, translation, posttranslational modification, protein–protein interaction and degradation, and HIF-1α expression may be enhanced by genetic alterations of the oncogenes or tumor suppressor genes regardless of oxygen tension. The present study revealed that chemotherapeutic drugs as well as radiation enhanced the expression and nuclear translocation of HIF-1α. It was recently demonstrated that mitochondrial ROS activate HIF-1α through the activation of p38 mitogen-activated protein kinase-associated signaling.48 Since radiation and chemotherapeutic drugs, including CDDP and 5-FU, induce ROS production in mitochondria, such anticancer agents may activate HIF-1α via mitochondrial ROS and p38. However, the precise mechanism of the activation still remains unknown.
In conclusion, the results obtained in this study revealed that the susceptibility of OSCC cells to chemotherapeutic drugs and γ-rays was negatively correlated with the expression level of HIF-1α and that the treatments with these agents enhanced the expression and nuclear translocation of HIF-1α. This study also showed that HIF-1α activated by the anticancer agents transactivates MDR1, HO-1 and Cp genes, which resulted in decreases of intracellular levels of chemotherapeutic drugs and ROS (Fig. 7). Furthermore, the knock-down of HIF-1α by siRNA enhanced the susceptibility of OSCC cells to the anticancer agents through the downregulation of P-gp, HO-1 and Mn-SOD, which resulted in increases of the intracellular levels of chemotherapeutic drugs and ROS. These results suggest that the enhanced HIF-1α expression is related to the resistance of tumor cells to chemo- and radio-therapy and that HIF-1α would be an effective therapeutic target for cancer treatment.