Identification of novel antiangiogenic anticancer activities of deguelin targeting hypoxia-inducible factor-1 alpha



Hypoxia-inducible factor 1 (HIF-1) plays an essential role in tumor angiogenesis and growth by regulating the transcription of several genes in response to hypoxic stress and changes in growth factors. This study was designed to investigate the effects of deguelin on tumor growth and angiogenesis, and the mechanisms underlying the antitumor activities of deguelin. We show here that orally administered deguelin inhibits tumor growth and blocks tumor angiogenesis in mice. Deguelin decreased expression of HIF-1α protein and its target genes, such as VEGF, in a subset of cancer cell lines, including H1299 lung cancer cells, and vascular endothelial cells in normoxic and hypoxic conditions. Overexpression of vascular endothelial growth factor by adenoviral vector infection abolished the antiangiogenic effects of deguelin on H1299 nonsmall cell lung cancer cells. Deguelin inhibited de novo synthesis of HIF-1α protein and reduced the half-life of the synthesized protein. MG132, a proteasome inhibitor, protected the hypoxia- or IGF-induced HIF-1α protein from deguelin-mediated degradation. Our findings suggest that deguelin is a promising antiangiogenic therapeutic agent in cancer targeting HIF-1α. Considering that HIF-1α is overexpressed in a majority of human cancers, deguelin could offer a potent therapeutic agent for cancer. © 2007 Wiley-Liss, Inc.

Despite recent advances in therapeutic modalities, the survival rate of patients with cancer, especially those with aerodigestive tract cancer, has not improved substantially,1 indicating the need for novel anticancer therapy. Tumor progression requires the generation of new blood vessels (i.e., angiogenesis) to provide nutrients and oxygen. Angiogenesis is also essential for the metastasis of primary tumor cells.2 Therefore, small-molecule inhibitors blocking angiogenesis in tumor could provide effective therapeutic opportunities.

One of ideal therapeutic targets for such an approach is hypoxia-inducible factor (HIF). HIF is an α/β heterodimeric transcription factor for the expression of genes, including vascular endothelial growth factor (VEGF), glucose transporters 1 and 3 and most of the glycolytic enzymes (aldolase A and C, enolase 1, hexokinase 1 and 3, lactate dehydrogenase A, phosphofructokinase L and phosphoglycerate kinase 1) and insulin-like growth factor 2. These proteins have important roles in several aspects of tumor biology, such as angiogenesis, glucose/energy metabolism, cellular growth, metastasis and apoptosis.2, 3, 4 Changes in the tumor microenvironment, such as hypoxia and growth factors, loss or inactivation of tumor-suppressor genes such as p53, PTEN, von Hippel-Lindau (VHL) and oncogenic activation, can all induce increases in HIF-1 expression and/or activity.2, 3, 4 Thus, HIF-1 can provide cancer and vascular endothelial cells with an advantage for survival and proliferation, leading to the formation of more aggressive and metastatic vascular tumors.

The transcriptional activity of HIF is regulated mainly by HIF- α subunits, with the best characterized being HIF-1α and HIF-2α. HIF-β is constitutively expressed, whereas the expression and activity of HIF-α protein are regulated by multiple oxygen-dependent and -independent mechanisms.4, 5, 6 Under normoxic conditions, HIF-α is subjected to prolyl hydroxylase-mediated hydroxylation within the oxygen-dependent degradation domain, which triggers binding of the VHL tumor suppressor protein and ubiquitin-mediated protein degradation.2, 3, 5 In contrast, factor inhibiting HIF-mediated hydroxylation of HIF-α within the C-terminal transactivation domain regulates transcriptional activity of HIF.7 The phosphatidylinositol 3-kinase (PI3K)/Akt pathway is involved in HIF-1α protein expression/stability.4, 8 HIF-α expression is also affected by its interaction with heat shock protein 90 (Hsp90), an ATPase-directed molecular chaperone that controls folding and stabilization of a number of client proteins that are typically required for cancer progression.9

While searching for effective anticancer agents, we identified a novel antiangiogenic action of deguelin, a natural product.10 Deguelin has shown potential chemopreventive activities against several types of cancers.11, 12, 13 However, anticancer activities of deguelin have not been defined. The data presented here demonstrate the therapeutic efficacy of deguelin in aerodigestive tract cancer, including nonsmall cell lung cancer (NSCLC) and head and neck squamous cell carcinoma (HNSCC). We found that deguelin inhibits HIF-1α expression in vitro and in vivo at the translational and post-translational levels and suppresses VEGF secretion, resulting in the effective antiangiogenic and therapeutic activities.

Material and methods

Cells, animals and materials

H1299, A549 human NSCLC, PC-3 human prostate cancer, MKN45 human gastric cancer and MCF7 human breast cancer, 786-0 renal carcinoma cell lines were obtained from the American Type Culture Collection (Manassas, VA). UMSCC38 HNSCC cells established originally by Dr. Thomas E. Carey (University of Michigan, Ann Arbor) were obtained from Dr. Reuben Lotan.14 These cancer cell lines were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum and antibiotics. Human umbilical vein endothelial cells (HUVECs; Cambrex BioScience, Walkersville, MD) were maintained in a gelatin-coated dish in endothelial growth medium (EGM, Cambrex BioScience) at 37°C in a humidified atmosphere of 5% CO2/95% air. HUVECs used in this study were taken from passages 2 to 7. Expression vectors containing GFP-tagged wild-type were provided by Dr. Kyu-Won Kim (Seoul National University, Seoul, Korea). HA-tagged wild-type HIF-1α was kindly given by Len Neckers (National Cancer Institute, Rockville, MD). The adenoviral vector expressing human VEGF (Ad-VEGF) and its parental vector (Ad-EV) were amplified as described.15 Tissue culture reagents and plastic ware were purchased from USA Scientific (Ocala, FL). Cell culture inserts incorporating polyester transwell membranes (6.4-mm diameter with an 8-μm pore size) and 24-well plates were obtained from Corning (Acton, MA). The Amicon Ultra-4 centrifugal filter was obtained from Millipore (Billerica, MA). Female nude mice, 6 weeks old, were purchased from Harlan Sprague Dawley (Indianapolis, IN). Deguelin was purchased from Gaia Chemical Corporation (Gaylordsville, CT) and was prepared as 100 μM stock solutions in dimethyl sulfoxide (DMSO) and stored at −20°C. Growth factor reduced Matrigel and heparin were purchased from Trevigen (Gaithersburg, MD) and Sigma–Aldrich (St. Louis, MO), respectively. Basic fibroblast growth factor (bFGF) and IGF-1 were obtained from R&D systems (Minneapolis, MN). Fertilized eggs were supplied from Charles River (Wilmington, MA).

In vivo tumor model and immunohistochemical staining

Subcutaneous injections for implantation of xenograft tumors of H1299, UMSCC38, A549 and MKN45 cells were performed as described elsewhere.16 After the tumor volume reached approximately 40–80 mm3 (day 0), mice were administered deguelin (4 mg/kg) by gavage contained in 100 μl 50% DMSO in corn oil or vehicle twice a day for 15–28 days. Tumor size was measured every 2–4 days. Orthotopic sublingual injections of UMSCC38 cells were described elsewhere.14 We injected 2 × 106 UMSCC38 cells into the lateral tongue of anesthetized 6-week-old female nude mice (n = 5 per group). Three days after tumor cell injection, when tumors started to develop, drug treatment was started. Two weeks after tumor cell injection, the mouse food was replaced by commercially available soft food (transgenic mice dough; Bio-serv, Frenchtown, NJ) that the mice could swallow even when the oral cavity was blocked by tumor. After 3 weeks of deguelin treatment, the mice were humanely killed by exposure to CO2. Tumor growth was assessed by measuring tumor size in two dimensions and calculating tumor volume as described elsewhere.14, 15 Histopathologic evidence of pulmonary toxicity (i.e., edema or inflammation of the bronchial epithelium and alveoli), inflammation and injury in other organs were evaluated by a veterinary pathologist. After the sacrifice of the animals, tumor tissues were collected for immunohistochemical analysis on CD31 (frozen section) and VEGF (paraffin block) and Western blot analysis on HIF-1α and HIF-1β as described.14 All animal procedures were performed in accordance with a protocol approved by the M. D. Anderson Institutional Animal Care and Usage Committee.

Conditioned medium

H1299 cells (3 × 106) in a 10-cm plate were incubated in growth medium with or without 100 nM deguelin. The plates were incubated under normoxic or hypoxic conditions for 24 hr, the cells were washed with phosphate-buffered saline (PBS) and then serum-free medium was added. After 24 hr of incubation, conditioned medium (CM) was removed and centrifuged at 4,000g for 20 min at 4°C through an Amicon Ultra-4 centrifugal filter (Millipore) to remove any trace of deguelin. The molecular mass cut-off of the filters was 5 kDa, and the molecular mass of deguelin (Mr = 400.6) is 0.36 kDa. The flow-through containing excess deguelin was discarded, and the retentate was collected. To test the specific role of VEGF in the inhibitory mechanism of deguelin in NSCLC angiogenesis, H1299 cells were infected with 25 plaque-forming units (pfu)/cell Ad-EV or Ad-VEGF as described previously14 and incubated in growth medium with or without 100 nM deguelin. CM was collected as described above. After filtrations, the final filter retentate was concentrated 40-fold. We determined removal of deguelin from the CM by treating H1299 cells with the CM and measuring HIF-1α expression by Western blot analysis. This CM was used for several analyses, including the chick aortic arch, tube formation and HUVEC proliferation assays.

Chick aortic arch assay and chorioallantoic membrane assay

The chick aortic arch assay was performed as described14 by incubating chick aortic arch ring in EGM containing 5 nmol deguelin or endothelial basal medium (EBM, Cambrex BioScience) containing CM from H1299 cells, which had been incubated with or without deguelin (100 nM). Average sprouting was measured with the Axiovision 4.3 software (Carl Zeiss, Jena, Germany) after the plates were photographed under a stereomicroscope (Carl Zeiss). Each condition was tested in 6 wells. The experiment was repeated 3 times, each with similar results. The chorioallantoic membrane (CAM) assay was conducted by using 4.5-day-old chick embryos as previously described.17 Retinoic acid (1 μg) was used as a positive control for determining antiangiogenic response. This independent experiment was repeated 3 times with more than 20 eggs each time.

Matrigel plug assay

The in vivo mouse Matrigel plug assay was performed as described elsewhere.17 Briefly, 10 nmol of deguelin was mixed with heparin (10 U), bFGF (50 ng/ml), and growth factor reduced Matrigel (600 μl) and then injected subcutaneously into nude mice. After 7 days, Matrigel plugs were excised, and the level of vessel formation was determined. One half of each plug was fixed in formalin, embedded in paraffin and stained for Masson's trichrome. The other half was homogenized and its hemoglobin content was determined by using the Drabkin reagent kit (Sigma–Aldrich). Each treatment group included 5 mice, and 1 representative result from 2 independent experiments with similar results is shown.

Invasion assay

The invasion assay was performed on transwell units coated with Matrigel as described elsewhere.14 Briefly, 1 × 104 HUVECs in 200 μl of EGM containing deguelin (1–100 nM) were added to the upper chamber. After 6 hr of incubation at 37°C under normoxic or hypoxic conditions, invaded cells were fixed and stained with hematoxylin and eosin, and the cells in 4 independent fields were counted under a microscope (Carl Zeiss). The results are expressed as percentages of invaded cells compared with control cultures.

Tube formation and proliferation assays

The tube formation assay was performed as described elsewhere.14 Briefly, HUVECs were seeded onto Matrigel surfaces and grown in complete medium containing various doses of deguelin or CM from H1299 cells, which had been incubated with or without deguelin (100 nM) under normoxic or hypoxic conditions. Morphologic changes of cells were photographed at 40× magnification and scored; a 3 branch-point event was scored as 1 tube. The experiment was repeated 3 times, each time with similar results. The effects of CM on HUVEC proliferation were tested as described elsewhere.14 Briefly, 2 × 103 HUVECs per well of 96-well culture plates were treated with 10 μg of CM from H1299 cells, which had been incubated with or without deguelin (100 nM) under normoxic or hypoxic conditions, and cell proliferation was assessed by the MTT assay after 48 hr of incubation. For each analysis, 6 replicate wells were used, and at least 3 independent experiments were performed.

Cell treatments

To assess the effects of deguelin on protein and mRNA expressions by Western blot and reverse-transcriptase polymerase chain reaction (RT-PCR) analyses, H1299, A549, UMSCC38, MKN45, PC3 and MCF7 cells were treated with deguelin (100 nM) in complete medium for 4 hr (RT-PCR) or16 hr under hypoxic or normoxic conditions. We treated HUVECs with deguelin (1–100 nM) in complete medium for 16 hr under hypoxic or IGF-1 treated normoxic conditions. Total protein extracts were collected for Western blot analysis as described elsewhere.10

Immunoblot assays

Preparation of whole-cell lysates and Western blotting were performed as described elsewhere,14 using a monoclonal antibodies against HIF-1α (1:500 dilution; BD Transduction Laboratories, Lexington, KY) and HIF-1β (1:1,000 dilution; Novus Biologicals, Littleton, CO); rabbit polyclonal antibodies against VEGF (1:1,000 dilution; Santa Cruz, CA) and bFGF (1:1,000 dilution, Santa Cruz); goat polyclonal antibodies against actin (1:4,000 dilution; Santa Cruz); rabbit anti-mouse immunoglobulin G (IgG)-horseradish peroxidase conjugate (1:2,000 dilution; Santa Cruz); donkey anti-rabbit IgG-horseradish peroxidase conjugate (1:2,000 dilution; Santa Cruz) and rabbit anti-goat IgG-horseradish peroxidase conjugate (1:2,000 dilution; Santa Cruz).

RT-PCR assay

Preparation of total RNA from whole-cell lysates and RT-PCR reaction were performed as described elsewhere.14 Amplification products were separated on agarose gels and visualized by ethidium bromide staining under ultraviolet transillumination. The primer sequences were as follows: (sense) 5′-CTCAAAGTCGGA CAGCCTCA-3′ and (antisense) 5′-CCCTGCAGTAGGTTTCT GCT-3′ for HIF-1α; (sense) 5′-CCATGAACTTTCTGCTGTCTT-3′ and (antisense) 5′-ATCGCATCAGGGGCACACAG-3′ for VEGF; (sense) 5′-GACTACCTCATGAAGATC-3′ and (antisense) 5′-GATCCACATCTGCTGGAA-3′ for β-actin.

Luciferase assays

H1299 cells were transfected with various combinations of effector plasmids, including pSV40promoter-EpoHRE-Luc reporter, pCMV-β-gal, pBOS-hHIF-1α and pBOS-hARNT using lipofectamine transfection methods. After transfection, cell lysates were analyzed for luciferase activity using an assay kit (Promega, Madison, WI) and a luminometer (Turner Design, Sunnyvale, CA). Extracts were also assayed for β-galactosidase enzyme activity. Each extract was assayed 3 times, and the relative luciferase activity was normalized to the RLU/β-galactosidase activity.

Metabolic labeling

Metabolic labeling was performed as described elsewhere.18 Briefly, H1299 cells were treated with deguelin (100 nM) in the presence of IGF-1 (100 ng/ml) or CoCl2 (100 nM) for 12 hr and incubated with medium lacking methionine and cysteine for 1 hr. The cells were washed with 1× PBS and labeled with [35S] methionine–cysteine with deguelin for protein synthesis. During labeling, cells were harvested at the indicated time points. For pulse-chase experiment, after 1 hr of pulse-labeling with [35S] methionine–cysteine, cells were washed with 1× PBS and then added to chase medium containing 150 mg/l of methionine and cysteine and 100 nM deguelin. During chasing, cells were harvested at the indicated time points. Equal amount of protein was used for immunoprecipitation employing an antibody against HIF-1α. The immunoprecipitates were washed and separated by SDS-PAGE and analyzed by autofluorography. Expression of β-actin was analyzed as a loading comparison. Laser densitometry was performed to quantify the density of the HIF-1α bands relative to that of cells at time 0.

Zebrafish maintenance and drug treatment

Transgenic zebrafish (Danio rerio) fli:EGFP were purchased from the Zebrafish International Resource Center (ZIRC, OR) and maintained according to the Zebrafish book.19 Adult fish were kept at 28.5°C on a 14 hr light/10 hr dark cycle similar to natural conditions and all embryos were collected from the natural mating as a unit of hours-post fertilization (hpf). Embryos were treated 24 and 32 hpf and deguelin was rinsed 1.5 hr after treatment. At 72 hpf photograph was taken under fluorescent microscope.

Statistical analysis

Data are expressed as the means and standard errors of the mean (SEM) or standard errors (SD) from at least triplicate samples, as calculated with Microsoft Excel software (version 5.0; Microsoft Corporation). Antiangiogenic effects, cell proliferation data and antitumorigenic effect of deguelin on the UMSCC38 orthotopic tumor were analyzed using Student's t-test. All statistical tests were two-sided and p < 0.05 was considered statistically significant for all tests.


Effect of deguelin on tumor growth

To evaluate the therapeutic efficacy of deguelin against aerodigestive tract cancers, we established H1299 xenograft and UMSCC38 orthotopic tongue tumors in nude mice as previously described.14, 20 Fifteen days of daily treatment with deguelin (4 mg/kg), based on maximum tolerable dose of intragastrically administered deguelin in rats,21 significantly inhibited the growth of H1299 xenograft tumors by 42.1% compared to that of control mice (Fig. 1a). We also tested dose dependence of deguelin with higher dose (8 mg/kg); however, there was no benefit in tumor growth inhibition (data not shown). Deguelin treatment for 3 weeks also induced significant decreases in the growth of UMSCC38 orthotopic tongue tumors in nude mice; on day 21, the average tongue tumor volume of deguelin-treated mice was reduced by 40.4% compared to that of control mice (Fig. 1b). After the treatment, control mice showed body weight lost due to the tongue tumors blocking diet consumption (Fig. 1b). Tissue samples taken from several organs (liver, lung, heart, kidneys, spleen, urinary bladder, ovary, stomach, duodenum, pancreas, jejunum, ileum, cecum, colon and rectum) of deguelin-treated mice revealed no pathohistological changes (data not shown), which is consistent with the lack of major toxic effects in A/J mice after 19 weeks of oral administration (twice daily) of deguelin in a previous study.11 These data showed that deguelin is a promising therapeutic agent in aerodigestive tract cancer, including lung and head and neck cancers.

Figure 1.

Deguelin suppresses NSCLC tumor growth. (a) Tumor volume in the vehicle-(Control) and deguelin-treated mice bearing a H1299 xenograft tumor (n = 6). (b) Representative photograph nude mice (n = 3) bearing UMSCC38 orthotopic tongue tumors treated with vehicle (Con) or deguelin. Tumor growth and changes in body weight in the vehicle and the deguelin treatment groups are expressed as the means + SEM. *p < 0.05; **p < 0.01, compared with the control group.

Deguelin inhibits tumor angiogenesis

We investigated the mechanisms that mediate the antitumor activities of deguelin. We have previously demonstrated that deguelin has apoptotic activities in premalignant and malignant human bronchial epithelial cells and NSCLC cells.10, 22 In this study, we investigated whether deguelin has antiangiogenic activity. We examined the effects of deguelin on tumor angiogenesis by measuring microvessel density in the H1299 and UMSCC38 xenograft tumor tissues obtained from vehicle- or deguelin-treated nude mice. Anti-CD31 staining of xenograft tumor tissue samples revealed that administration of deguelin significantly decreased tumor vascularization as reflected by the number of tumor vessels per high-power field (HPF; classical microvessel density) (Fig. 2a).

Figure 2.

Effect of deguelin on angiogenesis in NSCLC and HNSCC. Immunohistochemical analysis of CD31 was performed in H1299 and UMSCC38 xenograft tissues from nude mice that were treated with vehicle (Con) or deguelin. (a) Number of CD31-immunoreactive vessels. (b) Representative CD31 immunostaining in H1299 xenograft tissues. Results are expressed as the mean percentage + SEM. *p < 0.05; ***p < 0.005, compared with the control group.

A representative CD31 staining on H1299 xenograft tissue samples is shown (Fig. 2b). These findings suggested efficient suppression of tumor angiogenesis by the deguelin treatment.

Deguelin functions as a potent inhibitor of angiogenesis

We further investigated antiangiogenic activities of deguelin by the use of a series of ex vivo, in vivo and in vitro antiangiogenic assays. The ex vivo chick aortic arch ring assay indicated that deguelin (5 nmol) significantly decreased the number of endothelial cell sprouts (Fig. 3a). The CAM assay (Fig. 3b) and Matrigel plug assay (Fig. 3c), both of which are established in vivo angiogenesis models, also showed significant antiangiogenic activity of deguelin. Treatment with deguelin (1 nmol/egg) reduced new vessel formation in chick embryos to (46.7 ± 3.33)% (mean ± SEM.) (n = 20) (Fig. 3b). The chick embryos treated with RA as a positive control for angiogenesis inhibitor showed reduced vessel formation to 84.3%. We found no signs of toxicity of deguelin, such as thrombosis, hemorrhage and egg lethality, in the CAM assay. In the Matrigel plug assay, control plugs in which Matrigel was mixed with bFGF (100 ng/ml) showed abundant vessel development inside the plugs when compared with the plugs (−) in which Matrigel was injected with heparin alone, whereas deguelin (10 nmol) strongly inhibited the bFGF-induced angiogenesis (Fig. 3c). Functional vasculature was quantified by measuring hemoglobin contents in the Matrigels; deguelin (1 nmol) yielded significantly lower hemoglobin levels. We further tested the effects of deguelin in blood vessel patterning in vivo using zebrafish fli1:EGFP, where GFP is expressed specifically in the endothelial cells in live fish. Deguelin-treated zebrafish embryos at 24 or 32 hpf stages showed retarded intersegmental vessel formation during vessel development compared to that of untreated zebrafish embryo (Fig. 3d). In vitro treatment of HUVECs with deguelin significantly suppressed their potential for invasion (Fig. 3e) and the stability of tube formation (Fig. 3f) under normoxic (20% O2) and hypoxic (1% O2) conditions.

Figure 3.

Antiangiogenic activities of deguelin. (a) Angiogenic inhibition by deguelin (5 nmol) was photographed under a microscope (left) or evaluated by scoring angiogenic activity (right) in the chick aortic ring assay. (b) Angiogenic inhibition by deguelin (1 nmol) or all-trans retinoic acid (RA) (1 μg) was photographed under a microscope (left) or evaluated quantitatively (right), in the CAM assay. Arrow indicates the region covered with plastic ring. Independent experiments were repeated 3 times; values are means + SEM. (c) Matrigel plug assay. Left: Gross observation. Middle: samples were stained with Masson-trichrome; scale bar, 40 μm. E, endothelial cells; M, Matrigel. Right: Hemoglobin content. -, Matrigel containing heparin alone (10 U/ml); Con, heparin plus bFGF (50 ng/ml); Deguelin, heparin plus bFGF (50 ng/ml) and deguelin (10 nmol). Values are means + SEM. (d) Deguelin was treated to 24 and 32 hpf zebrafish (Danio rerio) fli:EGFP embryos and deguelin was rinsed 1.5 hr after treatment. The blood vessels of zebrafish vessels were visualized under fluorescent microscope at 72 hpf. Arrow indicates intersegmental vessel. (e, f) Invasion (6 h) (e) and tube formation (24 hr) (f) of HUVEC were inhibited by treatment with deguelin (100 nM) under hypoxic and normoxic conditions. Values are means + SEM. *p < 0.05; **p < 0.01; ***p < 0.005, compared with the control group.

Deguelin inhibits angiogenic activities of cancer cells

Because tumor angiogenesis is partly mediated by tumor-secreted angiogenic growth factors that interact with their surface receptors expressed on endothelial cells,23, 24, 25 we next investigated whether deguelin is able to inhibit angiogenic activities of cancer cells. To this end, we analyzed the effects of CM derived from deguelin (100 nM)-pretreated H1299 cells on vascular endothelial cells. Chick aortic arch assay revealed that chick aortas incubated with CM from deguelin-treated H1299 cells showed significantly decreased endothelial cell sprouting when compared with those incubated with CM from control H1299 cells (Con) (Fig. 4a). Moreover, when compared with EBM or CM from deguelin-treated cells, treatment with CM from deguelin-treated H1299 cells in normoxic or hypoxic conditions significantly decreased HUVEC tube formation (Fig. 4b) and proliferation (Fig. 4c).

Figure 4.

Deguelin inhibits angiogenic activities of cancer cells. (a) Endothelial cell sprouting from chick aortic rings, incubated with EBM alone or with CM from untreated (Con) or deguelin-treated (100 nM) H1299 cells, were evaluated as described. (b, c) Angiogenic activity in CM from H1299 cells was tested by tube formation (b) and proliferation of HUVEC (c). Values are means + SEM of triplicate results. *p < 0.05; **p < 0.01; ***p < 0.005, compared with the control group.

Deguelin inhibits VEGF expression in vitro

We explored the mechanisms that mediate the antiangiogenic activities of deguelin. We investigated the effects of deguelin on VEGF and bFGF secretion, which is commonly associated with angiogenesis.24, 25 Three bands of VEGF, which were defined as VEGF isoforms, VEGF189 at 28 kDa, VEGF165 at 22–23 kDa and VEGF121 at 18 kDa,26 respectively in CM from H1299 cells treated with deguelin (100 nM) in normoxic IGF-stimulated or hypoxic conditions were obviously lower than those from untreated cells (Fig. 5a). In contrast, bFGF (18 kDa) secretion was mildly affected by deguelin treatment. To test the specific role of VEGF in the antiangiogenic activities of deguelin, H1299 cells were infected by the adenoviral vector expressing human VEGF (Ad-VEGF) and its parental vector (Ad-EV) as a control and then treated with deguelin. HUVECs were treated with the CM from these cells. We found that the CM from H1299 cells that were infected with Ad-EV prior to the deguelin treatment significantly decreased HUVEC proliferation compared with that from untreated cells (Fig. 5b). However, the antiangiogenic activities of deguelin were obviously abrogated when the H1299 cells were infected with Ad-VEGF. These data suggest that deguelin induce antiangiogenic activities in the H1299 cell by regulating VEGF expression.

Figure 5.

Role of VEGF in deguelin-mediated inhibition of NSCLC angiogenesis. (a) Western blot analysis on VEGF and bFGF protein levels in CM secreted from the deguelin-treated cells. One representative result from two independent experiments is shown. (b) Angiogenic activity in CM from H1299 cells infected with Ad-EV or Ad-VEGF (25 pfu/cell) was tested by proliferation of HUVEC. Values are means + SD. ***p < 0.005 compared with the control group.

Deguelin inhibits HIF-1α expression in vitro and in vivo

Because HIF-1 plays a crucial role in tumor progression by regulating the expression of key apoptotic and angiogenic factors, including VEGF, we investigated the effects of deguelin on HIF-1α expression. Because the HIF-1α protein expression under normoxic conditions was very weak owing to its rapid degradation, we enhanced HIF-1α expression by the treatment with insulin-like growth factor (IGF) under normoxic condition (20% O2) or by exposure to hypoxic condition (1% O2). Treatment with deguelin (100 nM) reduced HIF-1α protein expression in H1299 cells in a time-dependent manner (Fig. 6a). Decreased HIF-1α protein and correlative VEGF mRNA expressions were also observed in deguelin-treated HUVECs (Fig. 6b). We further examined the functional consequences of deguelin-based treatment. As shown in Figure 6c, luciferase activity was significantly induced (4-fold) in response to hypoxia upon transfection of H1299 cells with pSV40 promoter-EpoHRE-Luc,27 whereas there was no induction in the cells transfected with mutated EpoHRE-Luc reporters. Treatment with deguelin yielded a concentration-dependent decrease in the luciferase activity of HIF under both hypoxic and normoxic conditions, indicating that deguelin inhibited the HIF transcriptional activity. We next sought to determine the effect of deguelin on HIF-1α expression in other types of cancer cells, including A549 NSCLC, UMSCC38 HNSCC, MKN45 stomach, PC3 prostate and MCF7 breast cancer cells. All of these cells showed similar decreases in HIF-1α protein expression by the deguelin treatment (Fig. 6d). Thus, inhibition of HIF-1α protein appeared to be a generic response of cancer and endothelial cells to the deguelin treatment. We then evaluated whether deguelin treatment would alter HIF-1α and VEGF expression in vivo. H1299 xenograft tumors obtained from the deguelin-treated mice (Fig. 1a) showed lower levels of HIF-1α expression compared to those from control mice (Fig. 6e). Immunohistochemical studies revealed decreased levels of VEGF staining in H1299 and UMSCC38 tumor tissues obtained from deguelin-treated nude mice when compared with that in control tumors (Fig. 6f). A549 NSCLC and MKN45 stomach xenograft tumors from deguelin-treated nude mice also showed decreases in the levels of VEGF expression compared with those from control mice (Fig. 6f).

Figure 6.

Deguelin inhibits HIF-1α and VEGF expression. (a) Western blot analysis on HIF-1α, actin expression in normoxic (20% O2), IGF-1 (50 ng/ml)-stimulated or hypoxic (1% O2) H1299 cells untreated (−) or treated (+) with deguelin for indicated time. One representative result from two experiments performed is shown. (b) Western blot analysis on HIF-1α and actin protein expression (upper two) and RT-PCR (lower two) analysis on VEGF and β-actin mRNA expression in HUVECs treated with the indicated concentrations of deguelin for 16 hr in the presence or absence of IGF-1 (50 ng/ml) or CoCl2 under normoxic (20% O2) condition or in hypoxic condition (1% O2). (c) H1299 cells were cotransfected with pSV40pro-EpoHRE-Luc (1 μg) or mutated EpoHRE-Luc (1 μg), and pCMV-β-gal (0.1 μg). Transfected cells were incubated under normoxic or hypoxic conditions in the presence of different concentrations of deguelin for 16 hr. Results are expressed as the mean + SD of triplicate results. *p < 0.05; **p < 0.01, compared with the control group. (d) Western blot analysis on HIF-1α and actin protein expression A549, UMSCC38, MKN45, PC3 and MCF7 cells treated with deguelin for 16 hr under hypoxic (1% O2) condition. (e) Western analysis on HIF-1α, HIF-1β and actin using H1299 xenografts from vehicle- (Con) or deguelin-treated animals. (f) Immunohistochemical analysis on VEGF staining on the tissues from H1299, A549, UMSCC38 and MKN45 xenograft tumors in the mice treated with vehicle- (Con) or deguelin for 15–28 days.

Deguelin inhibits HIF-1α expression by inhibiting de novo protein synthesis and by inducing ubiquitin- and proteasome-mediated protein degradation.

To investigate the mechanism by which deguelin regulates HIF-1α expression, we first tested the effect of deguelin on HIF-1α mRNA expression. It has been reported that HIF-1α mRNA expression changed marginally after the deguelin-based treatment (100 nM, for 6 hr),28 indicating posttranscriptional regulation of HIF-1α by deguelin. Therefore, we tested the effects of deguelin on HIF-1α protein synthesis by performing metabolic labeling on deguelin-treated H1299 cells with [35S] methionine–cysteine. As shown in Figure 7a, the rate of 35S-labeled HIF-1α protein synthesis in the cells treated with deguelin in the presence of IGF or CoCl2, hypoxia-mimicking agent, was remarkably lower than that in the control cells incubated with IGF or CoCl2 alone, indicating the inhibitory effects of deguelin on HIF-1α protein synthesis (Fig. 7a). We further investigated the effects of deguelin on the pre-existed HIF-1α protein. To this end, H1299 cells and HIF-1α- and VHL-deficient 786-0 RCC cells29 were transfected with an expression vector containing GFP-conjugated or HA-conjugated HIF-1α. Treatment with deguelin efficiently suppressed the expression of exogenously induced HIF-1α protein in these cells (Fig. 7b), suggesting the presence of mechanisms that mediate destabilization of HIF-1α protein by deguelin. We then determined the effects of deguelin on HIF-1α protein half-life. To this end, IGF-1- or CoCl2-stimulated cells were pulse-labeled with [35S], and then the radioactivity was chased for the indicated periods in the presence or absence of deguelin (Fig. 7c). The half-life of 35S-labeled HIF-1α in deguelin-treated H1299 cells showed a dramatic decrease when compared to the half-life of 35S-labeled HIF-1α in untreated cells (longer than 12 min in untreated H1299 cells and less than 3 min in deguelin-treated cells in IGF-1-induced condition; more than 4.5 hr in untreated H1299 cells and less than 1.5 hr in deguelin-treated cells in CoCl2-induced conditions). Because HIF-1α protein is degraded mainly through the ubiquitin-proteasome pathway, we tested whether blockade of proteasomal activity by a proteasome inhibitor affects the HIF-1α protein level in deguelin-treated cells. Treatment with proteasome inhibitor MG132 (10 μM) resulted in the formation of polyubiquitinated, higher-molecular-weight forms of HIF-1α that were further increased by the deguelin treatment in normoxic IGF-1 or CoCl2-stimulated and hypoxic H1299 cells (Fig. 7d). Moreover, combined treatment with deguelin and MG132 prevented the deguelin-mediated decrease in normoxic IGF- or CoCl2-stimulated or hypoxia-accumulated HIF-1α protein levels in H1299 cells (Fig. 7e). These findings indicate that deguelin can effectively inhibit HIF-1α protein expression by blocking protein synthesis and by inducing the ubiquitin-mediated proteasome degradation pathway.

Figure 7.

Deguelin inhibits HIF-1α expression by inhibiting de novo protein synthesis. (a) Metabolic labeling was performed in normoxic IGF-1 (100 ng/ml)- or normoxic CoCl2-stimulated H1299 cells. The deguelin (100 nM)-treated cells were labeled with [35S] methionine-cystein. Cells were harvested at the indicated time points for the analysis of protein synthesis. Laser densitometry was performed to quantify the density of the HIF-1α bands relative to that of cells treated with vehicle at time 1 hr. Western blot analyses on actin were included. (b) Western blot analysis on HIF-1α and actin protein expression using antibodies against GFP and actin in H1299 cells transfected with a GFP-tagged HIF-1α expression vector (GFP-HIF-1α) and then untreated or treated with deguelin (100 nM) for 6 hr (left). Western blot analysis was performed in 786-0 cells that were transfected with a vector expressing HA-tagged wild-type HIF-1α. These cells were incubated in normoxia (20% O2) or exposed to hypoxia (1% O2) for 6 hr and then untreated or treated deguelin (100 nM) for 6 hr (right). (c) H1299 cells were pulse-labeled and then the radioactivity was chased with unlabeled methionine–cystein in the presence of deguelin (100 nM). Cells were harvested at the indicated periods for the analysis of protein stability. Laser densitometry was performed to quantify the density of the HIF-1α bands relative to that of cells at time 0. One representative result from 2 experiments performed is shown. (d) HIF-1α ubiquitination by deguelin. Whole-cell extracts from H1299 cells that were untreated or treated with deguelin (100 nM) for 1 hr in the absence or presence of MG132 (10 μM) were immunoprecipitated (IP) with an anti-HIF-1α antibody, and immunoprecipitates (IP) were subjected to Western blot (WB) analysis with an antiubiquitin antibody. Western blot analyses on HIF-1α and actin were included. Ub (Ubiquitin). (e) Western blot analysis was performed in normoxic IGF-1 or CoCl2-stimulated and hypoxic H1299 cells incubated with deguelin (100 nM, 6 hr) in the presence or absence of MG132 (10 μM).


In this article, we have demonstrated that deguelin, a natural product, has antiangiogenic and antitumor activities in aerodigestive tract cancers, in NSCLC and HNSCC. Our observations show that deguelin-based treatment decreases de novo protein synthesis and the half-life of HIF-1α protein and production of VEGF in cancer and vascular endothelial cells in normoxic and hypoxic conditions. Blockade of proteasomal activity restores HIF-1α protein expression in the presence of deguelin, indicating an induction of ubiquitin-mediated degradation of HIF-1α protein by the drug. Because HIF-1α-driven expression of VEGF has an important role in the survival of cancer cells as well as vascular endothelial cells,30, 31 deguelin appears to inhibit tumor angiogenesis at least in part by blocking HIF-1α -driven, VEGF-induced autocrine and paracrine loops.

HIF-1 plays essential roles in the progressive growth of solid tumors under an unfavorable environment (e.g. hypoxia) by stimulating the transcription of genes, such as VEGF.2, 3, 4 Based on a major role of VEGF in angiogenesis, antibodies and soluble proteins that block VEGF and its receptors have been investigated as antiangiogenic therapeutic agents for solid tumors.32, 33, 34 HIF-1 also induces the transcription of genes that are involved in crucial aspects of cancer biology, including cell survival, glucose metabolism and invasion.4 In support of the multiple role of HIF-1 in tumor progression, inhibition of HIF-1 activity has marked effects on tumor growth in preclinical studies.35, 36

While searching for effective therapeutic agents that would block HIF-1 activity, we identified novel antiangiogenic and antitumor actions of deguelin. In the current study, we show that orally administered deguelin suppresses the growth of human NSCLC and HNSCC established in nude mice and vessel formation in the tumors. It appears that deguelin can function by various mechanisms. Deguelin reduced HIF-1α expression in multiple cancer cell lines, including NSCLC (H1299), head and neck (UMSCC38), prostate (PC3), stomach (MKN45) and breast (MCF7) cancer cells, as well as vascular endothelial cells. Deguelin has previously shown potential cancer chemopreventive efficacy against the carcinogenesis of lung, skin, colon, and breast in several in vitro and in vivo models.11, 37, 38 In addition, deguelin has been reported to inhibit the growth of different types of cancer cells by inducing apoptosis.38, 39 Deguelin induces apoptosis by inhibiting Akt or nuclear factor-kappaB kinase pathway,11, 40 which are important in cell proliferation and antiapoptosis.41 Deguelin also increases sensitivity of cancer cells to other chemotherapy drugs by downregulation of inhibitor of apoptosis.42 These collective findings indicated that deguelin could be an effective cancer therapeutic agent through its orchestrated actions on apoptosis and angiogenesis in cancer and vascular endothelial cells.

We investigated the mechanism of deguelin-mediated regulation of HIF-1α protein expression. On the basis of the inhibitory effects of deguelin on Akt activation,10, 11 we first paid our attention to the effects of deguelin on HIF-1α protein synthesis. PI3K/Akt activates mTOR (mammalian target for rapamycin), which phosphorylates and inactivates 4E-BP1 to connect the translational regulatory protein eIF-4E and thus to stimulate translation of HIF-1α mRNA into protein.8, 43, 44 Treatment with inhibitors of PI3K/Akt, such as LY294002 and wortmannin, gene transfection with dominant negative PI3K or kinase-dead Akt mutants, or increase in PTEN level have shown to decrease HIF-1α protein expression in certain cancer cell types.8, 45, 46, 47 Indeed, deguelin effectively suppressed de novo synthesis of HIF-1α protein. Surprisingly, the results from our experiments indicate that deguelin also induce HIF-1α protein degradation in addition to its expected effects on HIF-1α mRNA translation. Deguelin appeared to induce destabilization of the HIF-1α protein through oxygen-independent mechanisms because the drug treatment reduced the half-life of the protein both in normoxia and hypoxia. Recent studies have shown that activated Akt contributes to HIF-1α stabilization by provoking the expression of heat shock proteins, including Hsp70 and Hsp90.45 However, deguelin left the expression of the Hsp70 and Hsp90 proteins unaltered (data not shown). Therefore, deguelin appears to affect the stability of HIF-1α through an yet unknown mechanisms other than inactivation of Akt. We are currently investigating the mechanism underlying deguelin-induced HIF-1α protein degradation.

In summary, our findings revealed novel antiangiogenic and antitumor activities of deguelin in aerodigestive cancers, including lung and head and neck cancers. Treatment with deguelin halted HIF-1α protein synthesis and induced ubiquitin and proteasome-mediated protein degradation, resulting in decreased production of VEGF. Deguelin has several features that make it an attractive targeted antineoplastic drug for aerodigestive tract cancer in clinical trials. First, HIF-1α overexpression has been frequently observed in aerodigestive cancers, in which HIF-1α protein overexpression is associated with increased tumor angiogenesis and mortality.2 Disruption of HIF-1α transcriptional activity has shown therapeutic effects in xenograft models of colon and breast cancers.48 Second, deguelin can affect multiple aberrant signaling components in cancer cells by virtue of its effects on the expression of HIF-1α and HIF-1 target genes in addition to its established effects on PI3K/Akt and its downstream signaling mediators. Third, deguelin decreased HIF-1α protein expression at a concentration less than 100nM, which is considerably below the concentration achievable in vivo (about 100nM-1μM) in mice given a single oral dose of deguelin at 4 mg/kg.11 Finally and most importantly, when tested in the animal models, deguelin was well tolerated, causing no organ or systemic toxicity after prolonged administration to achieve the chemopreventive11 and therapeutic activities. In addition to the effects of deguelin on NSCLC and HNSCC cells, deguelin decreased the expression of HIF-1α protein in MKN45 stomach, PC3 prostate, MCF7 breast cancer cells. Furthermore, we and others have demonstrated that deguelin regulates proliferation of a variety of cancer cells.22, 38 All of these results provide an important new rationale for the use of deguelin as an antineoplastic drug in clinical trials.


This work was supported by National Institutes of Health grants RO1 CA100816-01 and R01 CA109520-01 (to Dr. Ho-Young Lee), and American Cancer Society grant RSG-04-082-01-TBE 01 (to Dr. Ho-Young Lee). Dr. W. K. H. is an American Cancer Society clinical research professor.