© 2005 Oxford University Press
ARTICLE |
Hypoxia-inducible Factor 1
and Antiangiogenic Activity of Farnesyltransferase Inhibitor SCH66336 in Human Aerodigestive Tract Cancer
Affiliation of authors: Department of Thoracic/Head and Neck Medical Oncology (J-YH, SHO, FM, EK, WKH, H-OL), Department of Head and Neck Surgery (JNM), The University of Texas M. D. Anderson Cancer Center, Houston, TX
Correspondence to: Ho-Young Lee, PhD, Unit 432, Department of Thoracic/Head and Neck Medical Oncology, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030 (e-mail: hlee{at}mdanderson.org).
 |
ABSTRACT |
|---|
 TopNotes Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Background: The farnesyltransferase inhibitor SCH66336, in combination with other receptor tyrosine kinase inhibitors, inhibits the growth of non–small-cell lung cancer (NSCLC) cells. We examined whether SCH66336 inhibits angiogenesis of aerodigestive tract cancer cells. Methods: Antiangiogenic activities of SCH66336 against NSCLC, head and neck squamous cell carcinoma (HNSCC), and endothelial cells were examined with cell proliferation, capillary tube formation, and chick aorta (under hypoxic, normoxic, insulin-like growth factor I (IGF)–stimulated, and unstimulated conditions); reverse transcription–polymerase chain reaction; and western blot analyses. The specific roles of the ubiquitin-mediated proteasome machinery, mitogen-activated protein kinase (MAPK) and Akt pathways, and heat shock protein 90 (Hsp90) in the SCH66336-mediated degradation of hypoxia-inducible factor 1
(HIF-1) were assessed with ubiquitin inhibitors and adenoviral vectors that express constitutively active MAP kinase kinase (MEK)1, constitutively active Akt, or Hsp90. Results: SCH66336 showed antiangiogenic activities and decreased the expression of vascular endothelial cell growth factor (VEGF) and HIF-1 in hypoxic, IGF-stimulated, and unstimulated aerodigestive tract cancer and endothelial cells. SCH66336 reduced the half-life of the HIF-1 protein, and ubiquitin inhibitors protected the hypoxia- or IGF-stimulated HIF-1 protein from SCH66336-mediated degradation. SCH66336 inhibited the interaction between HIF-1 and Hsp90. The overexpression of Hsp90, but not constitutive Akt or constitutive MEK, restored HIF-1 expression in IGF-stimulated or hypoxic cells but not in unstimulated cells. Conclusions: SCH66336 appears to inhibit angiogenic activities of NSCLC and HNSCC cells by decreasing hypoxia- or IGF-stimulated HIF-1 expression and to inhibit VEGF production by inhibiting the interaction between HIF-1 and Hsp90, resulting in the proteasomal degradation of HIF-1.
|
INTRODUCTION |
|---|
|
TopNotesAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Despite recent therapeutic advances, the survival rate of patients with aerodigestive tract cancer has not improved substantially (1), and most patients with aerodigestive tract cancer die of metastatic disease (2). Generation of blood vessels (i.e., angiogenesis) plays a critical role in malignant solid tumor growth and subsequent metastasis to other organs (3), suggesting that targeting the mechanisms that stimulate tumor angiogenesis should be explored as a therapeutic approach for solid tumors of the aerodigestive tract.
Among various proteins that have been identified as potential targets of antiangiogenesis therapy is vascular endothelial cell growth factor (VEGF). VEGF is expressed by activated endothelial cells and promotes the proliferation, survival, and migration of endothelial cells, and thus it is essential for blood vessel formation (4). In addition, the expression and secretion of VEGF by tumor cells are stimulated by activation of Ras (5) and the phosphatidylinositol 3-kinase (PI3K)–Akt pathway (6,7). The major physiologic stimulus for VEGF expression is hypoxia, which induces transcription of the VEGF gene by hypoxia-inducible factor 1 (HIF-1). HIF-1 is a heterodimer composed of HIF-1 and HIF-1
subunits (1,8). A nuclear localization signal at the carboxyl-terminal end of HIF-1 allows its transport from the cytoplasm to the nucleus, where it forms an active transcription complex by binding to HIF-1. HIF-1 is constitutively expressed, whereas the expression and activity of HIF-1 protein are regulated by O2-dependent and -independent mechanisms [for review, see Harris (9)]. Under normoxic conditions, HIF-1 is subject to O2-dependent prolyl hydroxylation, which triggers binding of von Hippel–Lindau tumor suppressor protein (VHL) and ubiquitin-mediated protein degradation by proteasome (10,11). Under the hypoxic condition, O2 becomes limiting for prolyl hydroxylase activity, HIF-1 ubiquitination is inhibited, and active HIF-1 transcription complexes can be formed. However, the level of HIF-1 also increases via an O2-independent mechanism (6,7). Growth factors, such as epidermal growth factor, heregulin, insulin-like growth factors (IGFs) I and -II, and insulin, induce expression of HIF-1 protein under nonhypoxic conditions (7,12,13). They bind to cognate receptor tyrosine kinases and activate the PI3K or mitogen-activated protein kinase (MAPK) pathway, which in turn increases the rate of HIF-1 protein synthesis. PI3K–Akt and MAPK have also been implicated in the stabilization of HIF-1 induced by oncogenes, hypoxia, and growth factors [for review, see Semenza (14)]. HIF-1 associates with the molecular chaperone heat shock protein 90 (Hsp90); pharmacologic disruption of this association promotes the ubiquitination and proteasome-mediated degradation of HIF-1 in an oxygen- and VHL-independent manner (15), suggesting that inhibitors of the interaction between HIF-1 and Hsp90 could be used to regulate the expression of hypoxia- or IGF-I–induced HIF-1 protein.
We have shown that SCH66336, a nonpeptide tricyclic farnesyltransferase inhibitor (FTI) that inhibits the farnesylation of various components, in combination with other inhibitors of the receptor tyrosine kinase signaling pathway, inhibits the growth of non–small-cell lung cancer (NSCLC) cells (16). In addition, in this study, we found that treatment with SCH66336 alone led to regression of orthotopic tongue tumors of human head and neck squamous cell carcinoma (HNSCC) in mice. Because angiogenesis is an essential step in the transition of a tumor from a small cluster of mutated cells to a large, malignant growth (17), we investigated the activity of SCH66336 on VEGF and HIF-1 expression and the mechanisms of its antiangiogenic action in aerodigestive tract cancers.
|
MATERIALS AND METHODS |
|---|
|
TopNotesAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Cells, Animals, and Materials
UMSCC38 HNSCC cells established originally by Dr. Thomas Carey (University of Michigan, Ann Arbor) were obtained from Dr. Reuben Lotan (18). We purchased the human NSCLC cell line H1299 from American Type Culture Collection (Manassas, VA). These cells 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 (Cambrex BioScience) at 37 °C in a humidified atmosphere of 5% CO2–95% air. HUVECs used in this study were from passages 2–7. Tissue culture reagents and plasticware were from Nunc (Roskilde, Denmark). Transwell chambers were from Corning-Costar (Corning, NY). Amicon Ultra-4 centrifugal filter was obtained from Millipore Co. (Bedford, MA). Cell culture inserts incorporating polyester transwell membranes (6.4-mm diameter with a 8-µm pore size) and 24-well plates were from Costar (Cambridge, MA). We used the following adenoviral (Ad) vectors for experiments in this article: a vector expressing constitutively active MAP kinase kinase (MEK) 1 (in which both serine residues at positions 217 and 221 were changed to a glutamine residue), referred to as Ad-MEK1 (19); a vector expressing constitutively active Akt (MyrAkt), referred to as Ad-HA-MyrAkt (20); a vector expressing hemagglutinin (HA)-tagged Hsp90, referred to as Ad-HA-Hsp90 (21); and an empty vector, referred to as EV. These vectors were amplified as described previously (16). Female nude mice, 6 weeks old, were purchased from Harlan-Sprague Dawley (Indianapolis, IN). SCH66336, i.e., [+]4-(2-[4-(8-chloro-3,0-dibromo-6,11-dihydro-5-benzocyclohepa(1,2-) pyridin-11-yl)-1-piperidinyl]-2-oxoethyl)-1-piperidinecarboxamide, was obtained from Schering-Plough (Kenilworth, NJ). We confirmed that SCH66336 can inhibit farnesylation (22) by assessing the level of unfarnesylated H-Ras in NSCLC cell lines after treatment with SCH66336 (data not shown). FTI-277, another farnesyltransferase inhibitor, was purchased from Calbiochem (San Diego, CA). SCH66336 and FTI-277 were dissolved in dimethyl sulfoxide at various concentrations to establish dose–response relationships. Bovine serum albumin, gelatin, the ubiquitin inhibitors MG132 and ALLN, cycloheximide, and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) were obtained from Sigma-Aldrich (St. Louis, MO). Inhibitors were prepared as 20 mM stock solutions in dimethyl sulfoxide and stored at –20 °C. Synthetic small interfering RNAs (siRNAs) targeting H-ras, VHL, or HIF-1 were purchased from Ambion (Austin, TX).
Hypoxic Treatment
Tissue culture dishes were transferred to a modular incubator chamber (Billups-Rothenberg, Del Mar, CA), the chamber was sealed, and the temperature was set at 37 °C. For hypoxic exposures, cells were placed in an airtight chamber (Biospherix, Redfield, NY), the chamber was flushed with a mixture of 1% O2, 5% CO2, and 94% N2 to maintain O2 concentration at 1% with Pro-Ox model 110 O2 regulators (BioSpherix), and cells were incubated at 37 °C.
Conditioned Medium
To obtain conditioned medium from SCH66336-treated H1299 NSCLC cells, we plated 106 H1299 cells in a 10-cm-diameter plate containing RPMI 1640 medium with 10% fetal bovine serum. After 24 hours, the medium on these cells was replaced with fresh growth medium containing SCH66336 (0 or 5 µM). The plates were then incubated under normoxic or hypoxic conditions for 1 day, and cells were washed with phosphate-buffered saline, and then serum-free medium containing the same concentration of SCH66336 was added. After 2 days of incubation, conditioned medium was removed and centrifuged at 4000g for 20 minutes at 4 °C through an Amicon Ultra-4 centrifugal filter (Millipore) to remove any trace of SCH66336. The molecular mass cutoff of the filters was 5 kDa, and the molecular mass of SCH66336 (Mr = 638.6) is 0.56 kDa. The flowthrough containing excess SCH66336 was discarded, and the retentate was collected. Because SCH66336 itself may have an inhibitory effect on this assay, we confirmed that this approach efficiently removed SCH66336 from the conditioned medium by treating UMSCC38 HNSCC cells with the conditioned medium and measuring farnesylated H-ras levels in the cells by western blot analysis. We found that inhibition of H-ras farnesylation required more than 0.5 µM SCH66336 (data not shown), which was insufficient to induce antiangiogenic effects in H1299 cells. By comparing supernatants from filtration spins with that from control cells treated with known concentrations of SCH66336, we determined that the amount of SCH66336 remaining after two successive filtration spin supernatants was not sufficient to inhibit H-Ras farnesylation when the starting concentration did not exceed 10 µM. The final filter retentate was concentrated 40-fold for use in the angiogenesis assay. This conditioned medium was used for several analyses, including Matrigel plug, chick aortic arch, HUVEC proliferation, and tube formation assays.
Cell Treatments
To assess the effects of SCH66336 on the expression of various proteins and mRNAs by western blot and reverse transcription–polymerase chain reaction (RT–PCR) analyses, we treated 106 cells per 100-mm3-diameter dish (H1299 NSCLC cells, UMSCC38 HNSCC cells, or HUVECs) with SCH66336 (0.5, 1, or 5 µM) in complete medium. For H-ras, HIF-1, VHL, or scrambled (control) siRNA transfection, H1299 cells were plated at a concentration of 105 cells per well in six-well plates. The next day, cells were transfected with 60 nmol of the indicated siRNAs by use of Oligofectamine (Invitrogen, Carlsbad, CA) for 6 hours and then cells were placed in fresh medium with or without 5 µM SCH66336. Scrambled siRNA was used as a negative control. After 2 days of incubation, medium was replaced with complete medium or serum-free medium, and then cells were incubated for another day. Serum-starved cells were stimulated by IGF-I at 100 ng/mL under normoxic (20% O2) or hypoxic (1% O2) conditions for 4 hours before harvest. Total protein extracts were collected for western blot analysis as described elsewhere (16). Briefly, cells were washed in phosphate-buffered saline and lysed in a buffer containing 50 mM Tris–HCl (pH 8.0), 150 mM NaCl, 0.1% sodium dodecyl sulfate (SDS), 1% Nonidet P-40, 1 mM phenylmethylsulfonyl fluoride, aprotinin at 5 µg/mL, and leupeptin at 5 µg/mL. After incubation on ice for 30 minutes and centrifugation at 1500g for 20 minutes at 4 °C, the supernatants were collected, and the protein concentration of each was determined with a BCA protein assay kit (Pierce, Rockford, IL).
To test the effects of SCH66336 on HIF-1 half-life, cycloheximide (100 ng/mL) was added to the medium from SCH66336-treated H1299 cells that had been stimulated with hypoxia or IGF-I for 4 hours, and whole-cell extracts were prepared as described above. In some experiments, ubiquitin inhibitors (10 µM MG132 or 10 µM ALLN) were added to growth medium when cells were stimulated with IGF-I or hypoxia.
To assess the contributions of the MAPK and Akt signaling pathways and Hsp90 expression to the SCH66336-mediated regulation of HIF-1 stability, H1299 cells were infected with Ad-HA-Myr-Akt (25 plaque-forming units [PFU] per cell), Ad-MEK (25 PFU per cell), Ad-HA-Hsp90 (50 PFU per cell), or Ad-EV (the parental virus control, 25–50 PFU per cell). For infection, cells and vectors were incubated for 2 hours in the absence of serum, incubated in growth medium containing 0.5–5 µM SCH66336 for 2 days, and then incubated with complete medium or with serum-free medium for another day. Cells were unstimulated or stimulated by IGF-I or hypoxia for 4 hours as described above.
To test the effects of conditioned medium on proliferation of vascular endothelial cells, 3 x 103 HUVECs (in 100 µL of endothelial basal medium [EBM], Cambrex Bioscience) per well of 96-well culture plates were treated with 10 µg of conditioned medium from the hypoxic, normoxic IGF-stimulated, or unstimulated NSCLC or HNSCC cells. To test the direct effects of SCH66336 on proliferation of vascular endothelial cells, we treated 3 x 103 HUVECs per well with 1 or 5 µM SCH66336 in 96-well culture plates. After 72 hours of incubation under hypoxic or normoxic conditions, cell proliferation was assessed by the MTT assay. For each analysis, six replicate wells were used, and at least three independent experiments were performed.
Chick Aortic Arch Assay
The chick aortic arch assay was as described elsewhere (24). In brief, thoracic aortas were obtained from chick eggs after 13–15 days of incubation. Excess perivascular tissue was removed, and transverse sections (1–2 mm) were cut. The resulting aortic rings were washed in medium 199 (Life Technologies) and embedded in 30 µL of Matrigel in 24-well plates (Costar) with the lumen perpendicular to the base of the well. Each ring was covered with 4 µL of Matrigel, which was allowed to gel, and then 300 µL of ECM and 10 µL of conditioned medium were added to each well. These plates were incubated for 24 hours or 48 hours at 37 °C to allow microvessel sprouting from the adventitial layer of the ring. The plates were photographed under a stereomicroscope (Zeiss, Göttingen, Germany), and average sprouting was measured with the imageJ program (National Institutes of Health, Bethesda, MD). Each condition was tested in six wells. The experiment was repeated three times, each with similar results.
In Vitro Capillary Tube Formation Assay
The capillary tube formation assay was as described elsewhere (23). We evenly dispersed 5 x 104 HUVECs on 250-µL Matrigel surfaces that were depleted of growth factors (Matrigel; Becton Dickinson, Bedford, MA); cells were incubated in 250 µL of EBM containing 30 µL of conditioned medium from H1299 cells cultured under hypoxic or normoxic conditions. To determine the direct effects of SCH66336 on HUVECs, cells were incubated in complete medium containing 5 µM SCH66336 under normoxic or hypoxic conditions. After incubation at 37 °C for 1 day, capillary tube formation was assessed as described elsewhere (23). Morphologic changes in the cells were assessed under a microscope and photographed at x40 magnification. Tube formation was scored; a three-branch-point event was scored as one tube (25). The experiment was repeated three times, each with similar results.
Immunoblot Assays
Whole-cell lysates were prepared in lysis buffer from 106 cells as described elsewhere (16). Equivalent amounts of protein (30–80 µg) were resolved by SDS–polyacrylamide electrophoresis in 7.5%–12% gels (80 V for 20 minutes and 100 V for 1 hour) and transferred by electroblotting overnight at 20 V to a nitrocellulose membrane. After nonspecific binding to the blot was blocked in Tris-buffered saline (TBS) containing 0.05% Tween 20 (TBST) and 5% nonfat powdered milk, the blot was incubated with primary antibody at the appropriate dilution in TBS–5% nonfat milk at 4° C for 16 hours. The membrane was then washed multiple times with TBST and incubated with the appropriate horseradish peroxidase–conjugated secondary antibody for 1 hour at room temperature. The protein–antibody complexes were detected by using the enhanced chemiluminescence kit (Amersham, Arlington Heights, IL), according to the manufacturer's recommended protocol. To identify HIF-1 on these blots, 80 µg of protein was analyzed with a monoclonal antibody against HIF-1 (BD-Transduction Laboratories, Lexington, KY; 1 : 500 dilution), as described elsewhere (24). We used 30 µg of protein for other western blot analyses with mouse monoclonal antibodies against phosphorylated p44/42MAPK (pp44/42 MAPK; Thr202/Tyr204; 1 : 500 dilution) or HIF-1 (H1234; 1 : 1000 dilution) (Novus Biologicals); rabbit polyclonal antibodies against phosphorylated Akt (Ser473; 1 : 1000 dilution), phosphorylated glycogen synthase kinase 3 (GSK-3; 1 : 1000 dilution), GSK-3, and VHL antibodies (1 : 1000 dilution) (Cell Signaling Technology, Beverly, MA); goat polyclonal antibodies against p44/42 MAPK (1 : 1000 dilution), rabbit polyclonal anti-Akt, -HA, -MEK1/2, and --actin (1 : 4000 dilution) antibodies (Santa Cruz Biotechnology, Santa Cruz, CA); rabbit polyclonal anti-HSP90 (1 : 5000 dilution) (Stressgen, Victoria, British Columbia, Canada); rabbit anti–mouse immunoglobulin G (IgG)–horseradish peroxidase conjugate (1 : 2000 dilution; DAKO, Carpinteria, CA); and donkey anti–rabbit IgG–horseradish peroxidase conjugate (1 : 2000 dilution) and rabbit anti–goat IgG–horseradish peroxidase conjugate (1 : 2000 dilution; both from Amersham Pharmacia Biotech, Arlington Heights, IL).
RT–PCR Assay
Total RNA was isolated from cells by using the Trizol reagent (Invitrogen). cDNA was synthesized from 1 µg of total RNA as a template in a 50-µL reaction mixture by using TaqMan reverse transcription reagents, according to the manufacturer's protocol (Applied Biosystems, Foster City, CA). The reaction was incubated at 25 °C for 10 minutes and at 48 °C for 30 minutes and then inactivated at 95 °C for 5 minutes. After inactivation, the cDNA was stored at –20 °C until use. RT–PCR was performed by coamplification of the genes with a reference gene (18S ribosomal RNA) by use of the cDNA template and corresponding gene-specific primer sets. The primer sequences were as follows: (sense) 5'-GGGAGAAAATCAAGTCGTGC-3' and (antisense) 5'-AGCAAGGAGGGCCTCTGATG-3' for HIF-1; (sense) 5'-CCATGAACTTTCTGCTGTCTT-3' and (antisense) 5'-ATCGCATCAGGGGCACACAG-3' for VEGF; (sense) 5'-GGTGAAGGTCGGTGTGAACGGATTT-3' and (antisense) 5'-AATGCCAAAGTTGTCATGGATGACC-3' for GAPDH. To avoid amplification of genomic DNA, we chose the primers from different exons. PCR was carried out in a total volume of 25 µL containing 1 µL of cDNA solution, 0.2 µM of sense primers, and 0.2 µM of antisense primers. The RT–PCR exponential phase was determined after 28–33 cycles to allow quantitative comparisons among the cDNAs developed from identical reactions. The thermocycler conditions used for amplification were 94 °C for 6 minutes (hot start) and then cycles of 94 °C for 45 seconds, 56–60 °C for 45 seconds, and 72 °C for 1 minute. The control PCR was performed for 26 or 27 cycles with 0.5 µL of cDNA solution to allow quantitative comparisons among the cDNAs developed from identical reactions with primers for GAPDH. Amplified products (8 µL) were resolved on 2% agarose gels, stained with ethidium bromide, visualized with a transilluminator, and photographed.
Immunoprecipitation
Whole-cell lysates were prepared in lysis buffer (50 mM Tris, 150 mM NaCl, 5 mM EDTA, 0.5% Nonidet P-40, 5% glycerol, 1 mM phenylmethylsulfonyl fluoride, and protease inhibitor mixture [Roche diagnostics, Mannheim, Germany], pH 7.5), followed immediately by three 15-second periods of vortex mixing. After centrifugation (15 000g for 30 minutes at 4 °C), supernatants were transferred to fresh tubes. Supernatants (1 mg of protein each) were mixed with 1 µg of anti-HIF-1 antibody (Santa Cruz Biotechnology) and incubated at 4 °C overnight. Thereafter, 50 µL of protein A–Sepharose beads (Amersham Pharmacia Biotech AB, Uppsala, Sweden) was added, and the mixture was incubated at 4 °C for 4 hours. Beads were washed three times with the lysis buffer and two times with 1x phosphate-buffered saline, boiled in Laemmli loading buffer, and separated by SDS–polyacrylamide electrophoresis in 8% gels. HIF-1 (BD-Transduction Laboratories, Lexington, KY), Hsp90, and ubiquitin were assessed by immunoblot analysis with corresponding antibodies.
In Vivo Tumor Model and Immunohistochemical Staining
Sublingual injections for implantation of orthotopic tumors were performed as described elsewhere (26). All animal procedures were performed in accordance with a protocol approved by the M. D. Anderson institutional Animal Care and Usage Committee. We injected 2 x 106 UMSCC38 cells into the lateral tongue of anesthetized 6-week-old female nude mice (n = 5 per group). One week after tumor cell injection, when tumors started to develop, drug treatment was started. SCH66336 (40 mg/kg of body weight) or 20% hydroxyl-propyl-betacyclodexatrin control vehicle was given orally twice a day for 3 weeks. Two and 4 weeks after tumor cell injection, tumor size and body weight were measured; thereafter, 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. Four weeks after tumor cell injection, 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 (27). After each mouse was killed, its tongue was removed and divided into two parts.
One part of the tongue was fixed, embedded in paraffin, and sectioned for VEGF and HIF-1 staining. The 5-µm tumor tissue sections were deparaffinized through a series of xylene baths and rehydrated through a series of graded ethanol baths. The sections were then immersed in methanol containing 0.3% hydrogen peroxidase for 20 minutes to block endogenous peroxidase activity and incubated in 2.5% blocking serum to reduce nonspecific binding. Sections were incubated overnight at 4 °C with primary antibody against VEGF (Santa Cruz Biotechnology; 1 : 200 dilution) or HIF-1 (Santa Cruz Biotechnology; 1 : 100 dilution) and then processed for avidin–biotin immunohistochemistry according to the manufacturer's recommendations (Vector Laboratories, Burlingame, CA). Diaminobenzidine was used as a chromogen, and commercial hematoxylin was used for counterstaining. The other part of tongue was frozen for CD31 staining; 20-µm frozen sections of tumor tissues were stained with anti-CD31 antibody (1 : 100 dilution; BD-Pharmingen, San Diego, CA), and then the antibody was detected by Cy3-conjugated secondary antibody as previously described (28).
Statistical Analysis
Data are expressed as the means and 95% confidence intervals (CIs) from triplicate samples, calculated with Microsoft Excel software (version 5.0; Microsoft Corporation, Seattle, WA). The Wilcoxon rank-sum test in the statistical SPSS statistical program (SPSS version 10; SPSS, Chicago, IL) was used to determine the statistical significance of antitumorigenic and antiangiogenic effects of SCH66336. Cell proliferation data were analyzed with a Student's t test. The null hypothesis that there was no difference in the cell proliferation between control and treatment groups was rejected at the probability (P) of less than .05. All statistical tests were two-sided.
|
RESULTS |
|---|
|
TopNotesAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
SCH66336 and Angiogenic Activities of Aerodigestive Tract Cancer Cells
We investigated whether the farnesyltransferase inhibitor SCH66336 had antiangiogenic activity on aerodigestive tract cancer cells, including NSCLC and HNSCC cells. The angiogenic process consists of several steps that include degradation of the basement membrane by endothelial cells, migration to the extravascular space, proliferation, and synthesis of a new basement membrane (29). We first performed the ex vivo chick aortic arch ring assay to determine the effects of SCH66336 on the ability of aerodigestive tract cancer cells to promote growth and migration of endothelial cells in three-dimensional cultures. The cell sprouting was evaluated by scoring angiogenic activity on a scale of 0–4. Conditioned medium from untreated control H1299 cells stimulated statistically significantly more endothelial cell sprouting (mean = 3, 95% CI = 2.2 to 3.8) than EBM (mean = 1, 95% CI = 0.2 to 1.8) (P = .04) (Fig. 1, A). However, conditioned medium from SCH66336-treated cells stimulated less endothelial cell sprouting (mean = 0.5, 95% CI = 0.1 to 0.9) than conditioned medium from untreated control cells (P = .03).
|
The in vitro capillary tube formation assay uses HUVECs to assess endothelial cell morphogenesis into capillaries on Matrigel-coated plates. Because tumor angiogenesis is also stimulated by hypoxia, we used conditioned medium from untreated control H1299 cells incubated under hypoxic (1% O2) or normoxic (20% O2) conditions. Conditioned medium from untreated control cells cultured under hypoxic (mean = 32.8 tubes per field, 95% CI = 28.0 to 37.7; n = 3 samples) or normoxic (mean = 29.1 tubes per field, 95% CI = 24.4 to 33.8; n = 3 samples) conditions stimulated capillary tube formation of HUVECs on Matrigel-coated culture plates statistically significantly more than EBM alone (Fig. 1, B). However, conditioned medium from cells treated with 5 µM SCH66336 under hypoxic (mean = 18.4 tubes per field, 95% CI = 14.8 to 22.0; P<.001) or normoxic (mean = 19.9 tubes per field, 95% CI = 16.3 to 23.4; P = .04) conditions stimulated endothelial cell morphogenesis statistically significantly less than conditioned medium from untreated control cells. These findings indicate that SCH66336 inhibited the angiogenic activities of H1299 cells.
Finally, because tumor angiogenesis is induced by angiogenic growth factors that are secreted from hypoxia- or IGF-stimulated cancer cells and then bind to their corresponding receptors expressed on endothelial cells to stimulate endothelial cell proliferation (30), we investigated whether conditioned medium from SCH66336-treated H1299 cells altered HUVEC proliferation. We found that addition of conditioned medium from hypoxic, normoxic IGF-stimulated, or unstimulated normoxic H1299 cells statistically significantly increased proliferation of HUVECs, whereas addition of conditioned medium from SCH66336-treated cells statistically significantly decreased HUVEC proliferation under hypoxic (73.2% of control growth stimulation, 95% CI = 68.2% to 78.2%; P = .004) or normoxic (67.1% of control growth stimulation, 95% CI = 63.0% to 71.1%; P = .002) conditions (Fig. 1, C). HUVECs incubated with conditioned medium from hypoxic, normoxic IGF–stimulated, or unstimulated normoxic UMSCC38 cells had comparable patterns of decreased cell proliferation (data not shown). Thus, SCH66336 appears to inhibit hypoxia- or growth factor–stimulated and constitutive secretion of angiogenic growth factors from aerodigestive tract cancer cells.
Expression of HIF-1 and VEGF by SCH66336-treated Cells
Because VEGF plays an important role in angiogenesis and because its expression is regulated largely by the transcription factor HIF-1, we investigated whether SCH66336 treatment alters the expression, and thus the secretion, of HIF-1 and VEGF in aerodigestive tract cancer cells. To determine whether HIF-1 is involved in the regulation of VEGF in H1299 cells, we transfected these cells with an siRNA targeting HIF-1, to inhibit the expression of HIF-1, or with a control scrambled siRNA and measured VEGF levels in corresponding conditioned media. The basal level of HIF-1 protein in untransfected H1299 cells was very low but increased markedly after cells were incubated under hypoxic conditions for 4 hours (Fig. 2, A). Levels of HIF-1 and VEGF proteins in H1299 cells transfected with HIF-1 siRNA were lower in normoxic and hypoxic conditions than cells transfected with scrambled siRNA. As expected, HIF-1 protein expression was unchanged in these cells. Thus, the expression of HIF-1 and VEGF appear to be associated in H1299 cells.
|
Because growth factors and oncogenes can increase the expression of HIF-1
(14,31,32), we investigated whether the level of HIF-1 protein in H1299 cells is regulated by Ras or IGF-I. The role of Ras in HIF-1 expression was tested in H1299 cells transfected with siRNAs targeting H-Ras. H-Ras siRNA-transfected H1299 cells had lower levels of H-Ras mRNA and protein under normoxic (Fig. 2, B, upper panels) and hypoxic (data not shown) conditions than control cells transfected with scrambled siRNA. Levels of GAPDH mRNA and -actin protein were unchanged in these cells under these conditions (data not shown). In addition, HIF-1 protein levels were not affected by H-Ras siRNA under hypoxic or normoxic conditions (Fig. 2, B, lower panels). To determine whether HIF-1 expression was affected by IGF-I, we cultured H1299 cells in serum-free medium containing IGF-I (50 ng/mL). As previously reported (14), we found that IGF-I had induced the expression of HIF-1 protein after 4 hours of incubation (Fig. 2, C). Thus, the level of HIF-1 protein appears to be regulated by oxygen concentration and growth factors in H1299 cells.
We then investigated whether SCH66336 treatment could alter the level of HIF-1 protein in H1299 cells under hypoxic, normoxic and IGF-stimulated, or normoxic and unstimulated conditions. Hypoxic conditions or IGF treatment increased the expression of HIF-1 protein in H1299 cells, compared with that in cells under normoxic conditions, and this increased expression was reduced by SCH66336 treatment in a time- and dose-dependent manner (Fig. 2, D). In addition, the level of HIF-1 protein in unstimulated H1299 cells was also reduced by SCH66336 treatment, but the level of HIF-1 protein was not altered by SCH66336 treatment under any condition tested. Another farnesyltransferase inhibitor, FTI-277, also decreased the level of HIF-1 protein in hypoxic, normoxic IGF-stimulated, and normoxic unstimulated H1299 cells (Fig. 2, E). In addition, patterns of decreased HIF-1 protein expression, which were similar to those in H1299 cells, were observed in UMSCC38 HNSCC cells treated with SCH66336 or FTI-277 under hypoxic, normoxic IGF-stimulated, and normoxic unstimulated conditions (Fig. 2, F). Thus, inhibition of HIF-1 protein in aerodigestive tract cancer cells appears to be a generic response to farnesyltransferase inhibitors.
We next tested whether the SCH66336-mediated decreased HIF-1 protein expression was associated with VEGF expression in hypoxic, normoxic IGF-stimulated, and normoxic unstimulated H1299 cells. Both VEGF mRNA expression in the SCH66336-treated cells and VEGF protein secreted in the conditioned medium decreased under all conditions tested, whereas GAPDH mRNA expression was not affected (Fig. 2, G). Thus, suppression of HIF-1 expression by SCH66336 appears to contribute to the inhibition of hypoxia- or IGF-stimulated, or constitutive VEGF production.
Because HIF-1 has recently been shown to have an important role in endothelial-cell survival (33), we investigated whether SCH66336 treatment would alter the expression of HIF-1 in HUVECs. Pretreatment of HUVECs with 5 µM SCH66336 inhibited the expression of HIF-1 protein and VEGF mRNA under hypoxic, normoxic IGF-stimulated, and normoxic unstimulated conditions, similar to the patterns previously observed (Fig. 3, A). We also examined whether SCH66336 treatment would alter HUVEC capillary tube formation and cell proliferation. For the capillary tube formation assay, HUVECs were untreated or treated with 5 µM SCH66336 for 24 hours under hypoxic or normoxic conditions, and 5 x 104 untreated or treated cells were cultured on a Trans-well chamber coated with a thin layer of Matrigel. In control medium, capillary tube formation was observed after 6 hours of incubation and was almost completed at 24 hours (Fig. 3, B); however, in medium containing 5 µM SCH66336 under hypoxic (mean = 16.4% of control capillary tube formation, 95% CI = 9.4% to 23.5%, P<.001) or normoxic (mean = 6.4% of control capillary tube formation, 95% CI = 3.0% to 9.7%, P<.001) conditions, HUVECs were statistically significantly less able to differentiate into capillary tube-like structures. For the cell proliferation assay, HUVECs were treated with 5 µM SCH66336 for 3 days under hypoxic or normoxic conditions, and cell proliferation was assessed. Proliferation of SCH66336-treated cells was statistically significantly lower under hypoxic (1.0 µM SCH66336, 83.2%, 95% CI = 80.3% to 86.1%, P<.001; 5 µM SCH66336, 66.3%, 95% CI = 63.5% to 69.1%, P<.001) and normoxic (1.0 µM SCH66336, 69.6%, 95% CI = 65.8% to 73.4%, P<.001; 5 µM SCH66336, 52.7%, 95% CI = 50.2% to 55.2%, P<.001) conditions than that of untreated cells (Fig. 3, C). These data suggest that SCH66336 acts directly on endothelial cells to inhibit the angiogenic-associated processes of tube formation and cell proliferation.
|
SCH66336 and Angiogenic Activities in HNSCC cells
To determine whether SCH66336 inhibits HIF-1 protein and VEGF expression in vivo and whether inhibition of these proteins affects tumor angiogenesis and tumor growth in vivo, we established HNSCC orthotopic tongue tumors in nude mice and treated the mice with SCH66336. Three weeks of oral treatment with SCH66336 (40 mg/kg) statistically significantly suppressed tongue tumor growth compared with untreated tumors (P = .02). On day 28 of cell injection, the average tumor volume among untreated control mice had increased to 336.4% of the pretreatment volume, whereas that among SCH66336-treated mice was 123.8% of the pretreatment volume (difference in means = 212.6, 95% CI for difference = 150.7% to 274.4%; P = .02) (Fig. 4, A, left panel). In contrast, the average body weight of SCH66336-treated mice was not reduced during the treatment (Fig. 4, A, middle panel), indicating that side effects of SCH66336 are minimal.
|
We next evaluated whether SCH66336 treatment would alter angiogenesis and HIF-1
and VEGF expression in HNSCC tumors in nude mice. SCH66336 treatment statistically significantly decreased tumor vascularization (P = .02), as reflected by microvessel densities that was measured with anti-CD31 staining in tongue tumor tissue sections from control and SCH6636-treated nude mice (Fig. 4, A, right panel, and 4, B). Decreased levels of HIF-1 and VEGF were detected in SCH66336-treated tongue tumor tissue, compared with levels in untreated control mice, which did not change over time. Consequently, SCH66336 appears to target HIF-1 and VEGF expression in aerodigestive tract cancer tissue.
SCH66336 and IGF-I–induced HIF-1 Protein Synthesis
We next investigated the mechanism through which SCH66336 reduces the expression of HIF-1. Because HIF-1 protein expression is very weak in normoxic unstimulated cells, we used hypoxic and normoxic IGF-stimulated H1299 cells for these studies. H1299 cells under hypoxic and normoxic IGF-stimulated conditions were treated with SCH66336, and then total RNA was isolated and subjected to RT–PCR. Untreated cells served as the control group. SCH66336 treatment did not measurably alter HIF-1 mRNA levels in the hypoxic and normoxic IGF-stimulated cells (Fig. 5, A), indicating that SCH66336 affects the level of HIF-1 at the posttranscriptional level, possibly by decreasing the rate of degradation and/or increasing the rate of synthesis (14).
|
To determine which mechanism is used by SCH66336 to mediate a reduction in the level of HIF-1
protein, H1299 cells were incubated with 5 µM SCH66336 for 3 days. These cells were then exposed to hypoxia or IGF-I at 100 ng/mL for 4 hours to induce HIF-1 expression, and cycloheximide was added to a final concentration of 100 µM to block protein synthesis. HIF-1 protein levels were measured before and at various times during cycloheximide treatment. Half-lives of both hypoxia- and IGF-I–induced HIF-1 proteins in the SCH66336-treated cells were statistically significantly deceased compared with those in untreated cells (half-life of hypoxia-induced HIF-1 protein = 180 minutes in untreated H1299 cells and 115 minutes in SCH66336-treated cells, difference = 65 minutes; half-life of IGF-I–induced HIF-1 protein = more than 30 minutes, difference = 150 minutes) (Fig. 5, B). Thus, SCH66336 appears to decrease the stability of HIF-1 protein in H1299 cells.
Because HIF-1 protein is degraded mainly through the ubiquitin–proteasome pathway, we investigated whether SCH66336-induced reduced HIF-1 protein level is dependent on proteasomal degradation by incubating SCH66336-treated H1299 cells with proteasome inhibitors and measuring the level of HIF-1 protein in such cells. Pretreatment for 8 hours with various proteasome inhibitors, including 10 µM MG132 or 10 µM ALLN, prevented the SCH66336-mediated decrease in the level of HIF-1 protein under hypoxic or IGF-I–stimulated conditions (Fig. 5, C). Because the polyubiquitin chain, which serves as a recognition signal for targeting the proteasome (34), is added to most proteins degraded in the proteasome, we examined polyubiquitination of HIF-1 and found that treatment of cells with a proteasome inhibitor, alone or in combination with SCH66336, resulted in the formation of polyubiquitinated, higher-molecular-weight forms of HIF-1 (Fig. 5, D). Thus, SCH66336 appears to interfere with HIF-1 protein accumulation by modifying a proteasome-dependent degradation pathway under hypoxic and normoxic IGF-stimulated conditions.
Role of VHL in SCH66336-mediated Degradation of HIF-1
We next investigated the role of VHL in the mechanism of SCH66336-mediated reduced HIF-1 expression in H1299 cells, because HIF-1 is constitutively stabilized in normoxic tumors and in cell lines that are VHL-null or that express a nonfunctional mutant form of VHL (35) and because VHL-inactivated tumors are highly vascularized and overproduce VEGF (36). We assessed the levels of VHL in hypoxic or normoxic IGF-stimulated or normoxic unstimulated SCH66336-pretreated H1299 cells and found that the expression of VHL protein in these cells was not changed by treatment with 5 µM SCH66336 (Fig. 6, A). We also examined whether inhibition of VHL expression with a VHL siRNA abrogates the effects of SCH66336 on HIF-1 expression by use of H1299 cells that were untransfected or transfected with a VHL or a control scrambled siRNA. We confirmed that VHL expression decreased in VHL siRNA-transfected H1299 cells (data not shown) and then measured HIF-1 protein levels in cells under various conditions. HIF-1 protein expression was higher in normoxic IGF-stimulated or unstimulated VHL siRNA-transfected H1299 cells than in control scrambled siRNA–transfected cells (Fig. 6, B). However, use of VHL siRNA to inhibit VHL expression did not abrogate the effects of SCH66336 on HIF-1 expression in H1299 cells under any condition tested. Thus, SCH66336 appears to reduce the level of HIF-1 by acting through VHL-independent pathways.
|
Role of MAPK and PI3K–Akt Pathways in SCH66336-mediated Degradation of HIF-1
Because PI3K–Akt and p44/42 MAPK pathways have been implicated in synthesis and stabilization of HIF-1 protein [for review, see Semenza (14)] and because SCH66336 can inhibit these pathways, we first investigated whether the PI3K–Akt and p44/42 MAPK pathways are involved in SCH66336-mediated decreased HIF-1 expression in H1299 cells. SCH66336 treatment decreased the levels of phosphorylated Akt (Ser-473) and phosphorylated p44/42 MAPK in hypoxic and normoxic IGF-stimulated H1299 cells, whereas the level of phosphorylated Akt was mildly decreased and the level of phosphorylated p44/42 MAPK was not affected by SCH66336 treatment in normoxic unstimulated H1299 cells (Fig. 6, C). Unphosphorylated Akt and p44/42 MAPK remained unchanged in the cells under every condition examined. Thus, expression of Akt and p44/42 MAPK does not appear to be affected by SCH66336 treatment.
Second, we explored whether the PI3K–Akt and p44/42 MAPK signaling pathways are involved in SCH66336-mediated decreased HIF-1 protein levels. In this experiment, we infected H1299 cells with a control adenoviral vector (EV), an adenovirus expressing HA-MyrAkt (Ad-HA-MyrAkt) to generate constitutively active Akt, or an adenovirus expressing MEK1 (Ad-MEK1) to generate constitutively active MEK1. The induced expression of HA-MyrAkt and MEK1 and the constitutive activity of these proteins were confirmed by western blot analysis with antibodies against HA, MEK1, phosphorylated GSK-3, an Akt downstream effector, and p44/42MAPK, as previously described (19) (Fig. 6, D–G). We next determined whether infection by Ad-HA-MyrAkt or Ad-MEK1 could protect HIF-1 expression from SCH66336-mediated degradation. Basal levels of HIF-1 protein levels were increased by Ad-HA-MyrAkt or Ad-MEK1 infection in hypoxic (Fig. 6, D, F), normoxic IGF-stimulated (Fig. 6, E, G), or normoxic unstimulated (Fig. 6, H) H1299 cells. However, treatment with SCH66336 mediated a decrease in the level of HIF-1 protein in these cells that was not restored when HA-MyrAkt or MEK1 were overexpressed, suggesting that neither the PI3K–Akt pathway nor the MAPK pathway plays a major role in protecting HIF-1 protein from SCH66336-mediated degradation in H1299 cells.
Hsp90 and Destabilization of HIF-1 Induced by SCH66336 Under Normoxic and Hypoxic Conditions
Because the protein chaperone Hsp90 plays a pivotal role in mediating the proper folding and subsequent activation of its many client proteins, including HIF-1 (37), we investigated the role of Hsp90 in HIF-1 protein expression under hypoxic, normoxic IGF-stimulated, or normoxic unstimulated conditions in H1299 cells, by examining whether overexpression of Hsp90 would protect H1299 cells from the SCH66336-induced degradation of HIF-1. We introduced Hsp90 into H1299 cells by infection with an adenoviral vector containing HA-tagged Hsp90 under the control of the cytomegalovirus promoter (i.e., Ad-HA-Hsp90). Induction of Hsp90 protein expression in the cells infected with Ad-HA-Hsp90 was confirmed by the appearance of the HA band on a western blot probed with an anti-HA antibody (Fig. 7, A). Under hypoxic or normoxic IGF-stimulated conditions, cells infected with Ad-HA-Hsp90 were completely rescued from the HIF-1–inhibiting effect of SCH66336, unlike cells infected w