Tumor hypoxia modifies the efficacy of conventional anticancer therapy and promotes malignant tumor progression. Human chorionic gonadotropin (hCG) is a glycoprotein secreted during pregnancy that has been used to monitor tumor burden in xenografts engineered to express this marker. We adapted this approach to use urinary β-hCG as a secreted reporter protein for tumor hypoxia. We used a hypoxia-inducible promoter containing five tandem repeats of the hypoxia-response element (HRE) ligated upstream of the β-hCG gene. This construct was stably integrated into two different cancer cell lines, FaDu, a human head and neck squamous cell carcinoma, and RKO, a human colorectal cancer cell line. In vitro studies showed that tumor cells stably transfected with this plasmid construct secrete β-hCG in response to hypoxia or hypoxia-inducible factor 1α (HIF-1α) stabilizing agents. The hypoxia responsiveness of this construct can be blocked by treatment with agents that affect the HIF-1α pathways, including topotecan, 1-benzyl-3-(5′-hydroxymethyl-2′-furyl)indazole (YC-1), and flavopiridol. Immunofluorescent analysis of tumor sections and quantitative assessment with flow cytometry indicate colocalization between β-hCG and 2-(2-nitro-1H-imidazol-1-yl)-N-(2,2,3,3,3-pentafluoropropyl)acetamide (EF5) and β-hCG and pimonidazole, two extrinsic markers for tumor hypoxia. Secretion of β-hCG from xenografts that contain these stable constructs is directly responsive to changes in tumor oxygenation, including exposure of the animals to 10% O2 and tumor bed irradiation. Similarly, urinary β-hCG levels decline after treatment with flavopiridol, an inhibitor of HIF-1 transactivation. This effect was observed only in tumor cells expressing a HRE-regulated reporter gene and not in tumor cells expressing a cytomegalovirus-regulated reporter gene. The 5HRE β-hCG reporter system described here enables serial, noninvasive monitoring of tumor hypoxia in a mouse model by measuring a urinary reporter protein.

Tumor hypoxia is a major determinant of the malignant progression of transformed cells and their response to therapy. It has been shown to promote tumor progression through the selection of tumor cells with diminished apoptotic potential, stimulate proangiogenic gene expression, and increase metastatic potential (1). Clinical studies have shown a strong correlation between pretreatment tumor pO2 and tumor control and survival in patients with head and neck cancers (2, 3) and cervical cancers (46). Clinical and preclinical studies have also indicated that hypoxia increases tumor invasiveness and dissemination in several human solid tumor types (710).

At the molecular level, hypoxia-inducible factor 1 (HIF-1) is an important transcription factor that regulates gene expression under hypoxic conditions (11). Over three dozen HIF-1–regulated genes have been identified to date (12, 13). The protein products of these genes play key roles in angiogenesis, vascular remodeling, glucose metabolism, cell proliferation, and cell survival. Recent studies suggest that HIF-1 is a novel target for anticancer therapy and many research groups are actively developing small molecular inhibitors of HIF-1 (1416). Therefore, a noninvasive and cost-effective system to monitor tumor hypoxia in vivo would aid in determining the efficacy of targeted therapy to HIF-1 or other hypoxia-regulated genes or proteins.

Various methods have been developed for assessing tumor hypoxia in xenograft tumor models and in patients (1721). Most approaches only allow assessment of tumor hypoxia only at a single time point, as it is often necessary to remove the tumor or to sacrifice the animal at the time of hypoxia measurement. Only few methods can be used for serial measurements of tumor hypoxia. These include various oxygen sensing devices and hypoxia imaging techniques (2224). The oxygen sensing devices, such as needle electrodes, can only access superficial tumors, are highly dependent on the technical skill of the user, and often fail to distinguish viable hypoxic cells from necrosis (25, 26). Imaging studies have also been developed to monitor hypoxia-induced stability of reporter proteins, such as luciferase (27), or HIF-1 reporter gene activity, such as positron emission tomography (PET) imaging of the 2′-[18F]flouro-2′deoxy-1β-d-arabionofuranosyl-5-ethyl-uracil (FEAU) product of a HIF-1–driven herpes simplex virus type 1 thymidine kinase/green fluorescent protein fusion (23, 24), but these approaches often require expensive dedicated animal imaging facilities and local imaging expertise for optimal results. Our goal was therefore to develop a noninvasive, rapid, and inexpensive system for dynamic monitoring of tumor hypoxia in living animals.

Human chorionic gonadotropin (hCG) is a glycoprotein normally secreted by syncytiotrophoblasts during pregnancy and has long been used as a sensitive and specific marker for pregnancy. β-hCG is readily measured in both serum and urine and has been used for the diagnosis and monitoring of pregnancy. Shih et al. (28) have previously developed a system using β-hCG engineered to be constitutively secreted from implanted tumor xenografts to monitor tumor growth and treatment response through a simple urine test. We have modified this approach by engineering a hypoxia-inducible promoter containing five tandem repeats of the hypoxia responsive element (HRE; ref. 29) ligated upstream of the β-hCG gene. We showed that tumor cells stably transfected with this construct secreted β-hCG in response to hypoxia and other HIF-1α stabilizing agents, and that the hypoxia responsiveness of this construct was blocked by treatment with agents targeting the HIF-1 pathway, including topotecan, 1-benzyl-3-(5′-hydroxymethyl-2′-furyl)indazole (YC-1), and, less directly, flavopiridol. Tumors growing from these cells also showed partial colocalization between β-hCG staining and EF5, a known marker for tumor hypoxia (30). Urinary secretion of β-hCG in xenografts expressing these stable constructs showed hypoxia-responsiveness to different manipulations of tumor oxygenation. Treatment with flavopiridol, a drug that decreases the stability of inducible mRNAs (31), including HIF-1α targets such as vascular endothelial growth factor (32), blocked β-hCG secretion in vitro and in mice bearing these engineered tumors. We conclude that the 5HRE β-hCG reporter system described hereenables serial, noninvasive monitoring of tumor hypoxia in a mouse model, and that it can potentially be used for in vivo studies to evaluate the efficacy of drugs targeting the HIF-1α pathway.

Materials. The pVSneo-hCG that expresses β-hCG under a constitutive cytomegalovirus (CMV) promoter was purchased from Promega (Madison, WI). Plasmids containing stabilized human HIF-1α cDNA and HIF-1α lacking a transactivation domain were generously provided by Denise Chan and Fiona Kaper (Department of Radiation Oncology, Stanford University, Stanford, CA), respectively. The stabilized HIF-1α vector contains mutations that inhibit HIF-1α hydroxylation (33, 34). The HIF-1α vector (GenBank accession no. NM_001530) that lacks a transactivation domain was generated by removing a cytosine from site 2,768 on the cDNA, thus creating a frame-shift mutation. There is also a point mutation that changes threonine 123 into an alanine. The frame shift upstream the transactivation domain creates a protein that can bind to the hypoxia-response element (HRE) but not produce mRNA from the target gene. The cell lines FaDu (human head and neck squamous carcinoma cells) and RKO (human colorectal carcinoma cells) were purchased from the American Type Culture Collection (Manassas, VA) and cultured in DMEM + 10% fetal bovine serum.

Vector construction. To generate a hypoxia-responsive β-hCG expression vector, the β-hCG gene was amplified by PCR from the pVSneo-hCG vector and engineered to contain the NcoI and NotI restriction sites at the 5′ and 3′ ends, respectively. This product was ligated into the same restriction sites in the 5HRE-hCMVmp d2eGFP vectors (35), replacing d2eGFP with β-hCG. DNA sequencing was used to verify the resulting sequences. The 5HRE-hCMVmp promoter contains five tandem repeats of the HRE (CCACAGTGCATACGTGGGCTCCAACAGGTCCTCTT) placed upstream to a CMV minimal promoter. The 5HRE-hCMVmp β-hCG vector will be referred to as 5HRE β-hCG.

Transfection and selection of clones. FaDu and RKO cells (1 × 106) were transfected with 2 μg of 5HRE β-hCG and CMV β-hCG using the Lipofectamine PLUS reagents (Invitrogen, Carlsbad, CA) according to the protocol of the manufacturer. Stable transfectants were selected in G418-containing media and single clones were isolated. Mitomycin C–pretreated feeder cells were used to promote growth of the single clones. Each clone was assayed for β-hCG expression by electrochemiluminescent immunoassay, described below, and the highest expressing clones were selected for subsequent experiments. Two to three clones were selected for further analysis on the basis of high β-hCG expression. The following stable clones were generated: RKO and FaDu cells that contain the constitutive β-hCG vector pVSneo-hCG (CMV-RKO and CMV-FaDu) and RKO and FaDu cells that contain the 5HRE-β-hCG vector (5HRE-RKO and 5HRE-FaDu).

Human β-chorionic gonadotropin and creatinine assays. β-hCG was measured using the Elecsys 2010 immunoassay (Roche Diagnostics, Indianapolis, IN), which is an electrochemiluminescent immunoassay that recognizes both intact human chorionic gonadotropin and the free β-chain. The overall imprecision was less than 5% coefficient of variation. A 10-fold dilution of mouse urine in Roche Universal Diluent (Roche Diagnostics) was generally done before electrochemiluminescent immunoassay analysis. Urinary β-hCG results were corrected for urine creatinine concentration, which was analyzed on Vitros 950 (Ortho-Clinical Diagnostics, Raritan, NJ).

Cell culture assays for human β-chorionic gonadotropin. Cells were seeded at 3 × 105 per well, allowed to settle overnight, washed with PBS, and exposed to either normoxia, 2% O2 (intermediate hypoxia), or <0.1% O2 (extreme hypoxia) in humidified 37°C chambers for 24 hours. In parallel, cells were treated for 24 hours with 50 μmol/L desferrioxamine (an iron chelator that stabilizes HIF-1α) or 5 mmol/L dimethyloxallyl glycine (a proline hydroxylase inhibitor). A separate set of cells was subjected to 24 hours of <0.1% O2, followed by 0, 24, or 48 hours of reoxygenation, with media replacement every 24 hours. For all in vitro assays, secreted β-hCG was measured by electrochemiluminescent immunoassay as described above and adjusted to the total cell number at the end of the experiment. Cell lysates were generated by lysing cells in 150 mmol/L sodium chloride, 10 mmol/L Tris (pH 7.5), 0.1% SDS, 1% Triton X-100, and 0.5% deoxycholate, and intracellular β-hCG levels were assayed by electrochemiluminescent immunoassay.

Hypoxia-inducible factor 1α transfection. Approximately 1 × 105 5HRE-RKO cells were transiently transfected with 0.4 μg of stabilized HIF-1α, empty vector, or HIF-1α lacking a transactivation domain using Lipofectamine PLUS reagents (Invitrogen). Cotransfection with 0.04 μg β-galactosidase cDNA (Promega) was used to correct for transfection efficiency. Eighteen hours after the start of transfection, the media was replaced and β-hCG was allowed to accumulate in the new media for 24 hours. Cells treated with <0.1% O2 were used as a positive control. Measured β-hCG levels were corrected for transfection efficiency.

Topotecan treatment. 5HRE-RKO and CMV-RKO cells were subjected to 2, 5, and 10 μmol/L topotecan (LKT Laboratories, Inc., St. Paul, MN) under <0.1% O2 for 24 hours. To distinguish anti-HIF-1α effects from nonspecific effects on cell proliferation and viability, we calculated the ratio of β-hCG secretion in 5HRE-RKO and CMV-RKO cells, and arbitrarily assigned the ratio of 5HRE hypoxic treatment to CMV hypoxic treatment without topotecan as 1.

Flavopiridol treatment. Approximately 1.3 × 105 cells were treated with 24 hours of normoxia, hypoxia (<0.1% O2), or hypoxia with 100 nmol/L flavopiridol (Aventis Pharmaceutical, Inc., Bridgewater, NJ). The ratio of secreted β-hCG from 5HRE to CMV cells was calculated. To allow for easy comparison, the ratio of 5HRE to CMV secretion without flavopiridol treatment under hypoxia was arbitrarily set to 1. To confirm that flavopiridol blocks 5HRE reporter protein induction under hypoxia rather than protein secretion, transient luciferase transfections were done with both 5HRE-firefly luciferase plasmids (0.36 μg) and CMV-Renilla luciferase plasmids (0.04 μg) in 1 × 105 FaDu cells. Both luciferase protein species were assayed using the Dual-Luciferase kit (Promega) and the ratio of firefly/Renilla was calculated. The ratio of firefly/Renilla under hypoxia without flavopiridol was also set to 1.

Northern blot. Cells were lysed with TRIzol reagent (Gibco BRL, Gaithersburg, MD) and total RNA was extracted according to the protocol of the manufacturer. Eight micrograms of RNA per lane were denatured with glyoxal and DMSO, subjected to electrophoresis on a 1% agarose/sodium phosphate gel, transferred to a nylon membrane, UV cross-linked, and hybridized with a 32P-labeled β-hCG probe. Radioactivity was exposed to a phosphor screen and visualized on a Storm 860 scanner (Amersham Biosciences, Piscataway, NJ).

Animal studies. Animal protocols were approved by the Institutional Animal Care and Use Committee at Stanford University. Tumor xenografts were formed by injecting cells intradermally into severe combined immunodeficiency (SCID) BALB/c mice (The Jackson Laboratory, Bar Harbor, ME). The mice were maintained under pathogen-free conditions.

Urine collection. Mice were placed on 96-well plates (Nalge Nunc International, Rochester, NY) that line the bottom of a standard mouse cage with food and water. Urine specimens were collected from the wells for 2 to 6 hours and frozen at −20°C until assayed. At each collection, tumor length and width were recorded. Tumor volume was calculated according to the equation (length × width2 × π / 6), where width was the shorter dimension. In the experiments in which mice were treated to 10% oxygen, the height dimension was also recorded for greater accuracy. For these experiments, volume was calculated according to the equation (length × width × height × π / 6). At the time of sacrifice, the tumor was removed and weighed.

Tumor bed irradiation. The lower backs of mice were irradiated with a single fraction of 20 Gy ionizing radiation 11 days before tumor cell implantation, an established procedure that increases hypoxia in implanted tumors (36). Irradiation was delivered with a 250 kVp X-ray unit (250 mA, 0.5 copper filter, 1.54 Gy/min) while the mice were inside a lead-shielded chamber, exposing only the lower backs. Nonirradiated mice injected with the same tumor cells in the same location were used as controls. Urine was collected just before sacrifice, and urinary β-hCG was corrected for creatinine and the mass of the excised tumor. Unirradiated mice were sacrificed 6 weeks after implantation and, due to slow growth in irradiated tumor beds, the irradiated mice were sacrificed 16 weeks after implantation.

Ten percent oxygen breathing. Mice with 5HRE-RKO xenografts were placed in a sealed environment chamber that was infused with 10% oxygen and 90% nitrogen (Praxair, San Carlos, CA) for 72 hours. Control mice with matched tumor volumes were maintained in room air. Urine samples were collected daily during and up to 5 days after 10% oxygen treatment and assayed for β-hCG and creatinine.

Flavopiridol treatment in vivo. Mice were injected with 5 × 106 5HRE-FaDu cells and the tumors were allowed to grow to ∼400 mm3. Mice were injected with 5 mg/kg flavopiridol daily for 10 days or with 10% DMSO in saline for control. Urine was collected every 1 to 2 days starting just before the first injection and assayed for β-hCG and creatinine.

2-(2-Nitro-1H-imidazol-1-yl)-N-(2,2,3,3,3-pentafluoropropyl)acetamide or pimonidazole treatment and immunofluorescence. Ninety minutes before sacrifice, each animal received a 0.1-mL i.p. injection of 10 mmol/L EF5 per 10 g body weight or pimonidazole at a dose of 100 mg/kg. After sacrifice, the tumor was removed, frozen in liquid nitrogen, and stored at −80°C until assayed. For tissue staining, 14 μm sections were made on a cryostat, fixed in 4% paraformaldehyde in PBS (1 hour at 4°C), and washed thrice in ice-cold PBS. Blocking was done overnight at 4°C in PBS with 0.3% Tween 20, 1.5% bovine serum albumin (BSA), 20% skim milk, and 5% heat-inactivated mouse serum (30 minutes, 56°C). After washing, EF5 adducts were stained with the mouse monoclonal ELK3-51-Cy3 antibody at 80 μg/mL (1). All antibodies were diluted in PBS, 0.3% Tween 20, and 1.5% BSA. β-hCG was stained with a mouse anti-β-hCG primary antibody (NeoMarkers, Fremont, CA) at 1:100 dilution for 6 hours at 4°C, followed by washing and staining with a goat anti-mouse Alexa Flour 488 fluorochrome-conjugated secondary antibody (Molecular Probes, Eugene, OR) overnight at 4°C. The sections were visualized with VectaShield with 1.5 μg/mL 4′,6-diamidino-2-phenylindole (DAPI; Molecular Probes) on a fluorescence microscope with filters for propidium iodide (for EF5-Cy3), FITC (for β-hCG-Alexa Flour 488), and UV (DAPI).

Flow cytometric analysis for pimonidazole was done with the appropriate antibodies as previously described (37, 38). Approximately 2 × 106 cells were used for each experiment. All experiments were done in duplicate and repeated at least once. RKO cells cultured under hypoxia (2% and <0.1% O2) were used as positive controls. As a negative control, we omitted either the primary or secondary antibody during the staining process. Fluorescence data were acquired and analyzed as previously described (37) on a Becton Dickinson FacsCalibur Flow cytometer equipped with a 15 mW argon laser with the FL1-H (488/30) channel.

Statistical analysis. Statistical analysis was done using Statview (SAS Institute, Inc., Cary, NC) statistical software. Student's t test was used to compare the β-hCG levels for control and treated mice in the tumor bed irradiation and flavopiridol experiments (39).

In vitro studies. We engineered a HIF-1α responsive vector that selectively up-regulates β-hCG under hypoxia in two separate cell lines: 5HRE-RKO, a colorectal adenocarcinoma cell line, and 5HRE-FaDu, a pharyngeal squamous cell carcinoma cell line. Figure 1 shows very strong induction of β-hCG mRNA (Fig. 1A), intracellular protein (data not shown), and secreted protein (Fig. 1B) under hypoxia. Returning the samples to aerobic conditions (reoxygenation) resulted in a rapid reduction in β-hCG mRNA but a less rapid reduction in the cellular and secreted protein levels. Kinetic analysis indicated that the maximal induction of the secreted protein levels occurred at 24 hours of hypoxic treatment, in which there was a 46-fold increase in β-hCG secretion per 5HRE-RKO cell under hypoxia compared with normoxia (data not shown). A similar response was noted in 5HRE-FaDu cell lines with a 33-fold induction (data not shown). We observed hypoxic induction of β-hCG at both 2% and <0.1% oxygen in 5HRE-RKO cells, with slightly higher induction under more stringent hypoxia (Fig. 1C). Treatment with the hypoxia mimetics desferrioxamine (50 μmol/L) and dimethyloxallyl glycine (5 mmol/L) for 24 hours under normoxia resulted in a similar induction of β-hCG protein secretion as with treatment for 24 hours under hypoxia (Fig. 1C). The CMV-RKO stable cell lines, containing the vector pVSneo-hCG, showed only a small increase in constitutive β-hCG secretion (<1.7-fold compared with >33-fold for the 5HRE-RKO cells) after treatment with either hypoxia-mimetic drugs or hypoxia, presumably related to some cell death and lysis with hypoxia and drug treatment (Fig. 1D). All in vitro experiments were done in triplicate in at least two independent experiments. The mean and SD of the data are shown.

Figure 1.

Validation of 5HRE β-hCG as a new reporter protein for hypoxia and hypoxia mimetics. A, Northern blot showing significant induction of β-hCG mRNA after 24 hours of <0.1% oxygen (H) over normoxia (N), and a rapid reduction of the mRNA after 24 hours (H+24R) and 48 hours of reoxygenation (H+48R). 18S rRNA (18S) was used as a loading control. B, electrochemiluminescent immunoassay studies showing induction of secreted β-hCG protein levels in the conditioned media after 24 hours of hypoxia and a gradual reduction of the protein on reoxygenation. C, electrochemiluminescent immunoassay studies showing induction of β-hCG secreted protein in the conditioned media after 24 hours of exposure to <0.1% oxygen (H), 2% oxygen (H, 2%), and 50 μmol/L desferrioxamine or 5 mmol/L dimethyloxallyl glycine (DMOG). D, electrochemiluminescent immunoassay studies showing only a slight increase (<1.7-fold) in the levels of secreted β-hCG protein in the conditioned media of CMV-RKO cells treated with either hypoxia or 5 mmol/L dimethyloxallyl glycine.

Figure 1.

Validation of 5HRE β-hCG as a new reporter protein for hypoxia and hypoxia mimetics. A, Northern blot showing significant induction of β-hCG mRNA after 24 hours of <0.1% oxygen (H) over normoxia (N), and a rapid reduction of the mRNA after 24 hours (H+24R) and 48 hours of reoxygenation (H+48R). 18S rRNA (18S) was used as a loading control. B, electrochemiluminescent immunoassay studies showing induction of secreted β-hCG protein levels in the conditioned media after 24 hours of hypoxia and a gradual reduction of the protein on reoxygenation. C, electrochemiluminescent immunoassay studies showing induction of β-hCG secreted protein in the conditioned media after 24 hours of exposure to <0.1% oxygen (H), 2% oxygen (H, 2%), and 50 μmol/L desferrioxamine or 5 mmol/L dimethyloxallyl glycine (DMOG). D, electrochemiluminescent immunoassay studies showing only a slight increase (<1.7-fold) in the levels of secreted β-hCG protein in the conditioned media of CMV-RKO cells treated with either hypoxia or 5 mmol/L dimethyloxallyl glycine.

Close modal

To confirm that β-hCG is a reporter gene for HIF-1α activity, we transiently transfected 5HRE-RKO cells with plasmids containing a copy of stabilized HIF-1α, which has point mutations in proline 402 and proline 564 that stabilize HIF-1α levels under oxygenated conditions (34). A vector for HIF-1α lacking a functional transactivation domain was used as a negative control. Cotransfection with a plasmid containing the β-galactosidase cDNA was used to control for transfection efficiency. Overexpression of stabilized HIF-1α resulted in increased β-hCG protein secretion when compared with the vector control or to the HIF-1α lacking a transactivation domain (Fig. 2A), confirming that secreted β-hCG is a reporter protein for HIF-1α activity. In addition, 24-hour treatment with increasing concentrations of topotecan, a drug that has been shown to have an inhibitory effect on HIF-1α activity (40, 41), resulted in a dose-dependent inhibition of β-hCG secretion under hypoxia in 5HRE-RKO cells but minimal effect on β-hCG secretion in CMV-RKO cells. Figure 2B shows the ratio of secreted β-hCG in 5HRE-RKO cells to the secreted β-hCG in CMV-RKO cells. To distinguish anti-HIF-1α effects from nonspecific effects on cell proliferation and viability, we calculated the ratio of β-hCG secretion in 5HRE-RKO and CMV-RKO cells, and arbitrarily assigned the ratio of 5HRE hypoxic treatment to CMV hypoxic treatment without topotecan as 1 with all other results normalized to this number for ease of interpretation. We obtained similar results using 1-benzyl-3-(5′-hydroxymethyl-2′-furyl)indazole (YC-1; A.G. Scientific, San Diego, CA), another compound targeting the HIF-1 pathway (data not shown; ref. 42).

Figure 2.

Confirmation of 5HRE β-hCG as a HIF-1 reporter gene. A, electrochemiluminescent immunoassay studies showing enhanced β-hCG secretion in RKO cell conditioned media after transfection with plasmids containing stabilized HIF-1α cDNA (HIF-1α, stable), but not with plasmids containing the HIF-1α cDNA lacking the transactivation domain (HIF-1α, no TAD). B, electrochemiluminescent immunoassay studies in RKO cells showing decreased β-hCG reporter protein in conditioned media after 24 hours of treatment with increasing concentrations of topotecan, a known HIF-1α inhibitor. Because topotecan selectively suppressed β-hCG production only in the HIF-1–dependent 5HRE-RKO cells but not in the constitutively active CMV-RKO cells, β-hCG concentrations were expressed as a ratio of that produced by 5HRE-RKO cells to that of CMV-RKO cells, and the ratio of 5HRE hypoxic treatment to CMV hypoxic treatment without topotecan was arbitrarily assigned as 1 for ease of comparison. The results from normoxia- and topotecan-treated cells were normalized to this number.

Figure 2.

Confirmation of 5HRE β-hCG as a HIF-1 reporter gene. A, electrochemiluminescent immunoassay studies showing enhanced β-hCG secretion in RKO cell conditioned media after transfection with plasmids containing stabilized HIF-1α cDNA (HIF-1α, stable), but not with plasmids containing the HIF-1α cDNA lacking the transactivation domain (HIF-1α, no TAD). B, electrochemiluminescent immunoassay studies in RKO cells showing decreased β-hCG reporter protein in conditioned media after 24 hours of treatment with increasing concentrations of topotecan, a known HIF-1α inhibitor. Because topotecan selectively suppressed β-hCG production only in the HIF-1–dependent 5HRE-RKO cells but not in the constitutively active CMV-RKO cells, β-hCG concentrations were expressed as a ratio of that produced by 5HRE-RKO cells to that of CMV-RKO cells, and the ratio of 5HRE hypoxic treatment to CMV hypoxic treatment without topotecan was arbitrarily assigned as 1 for ease of comparison. The results from normoxia- and topotecan-treated cells were normalized to this number.

Close modal

In vivo studies. Intradermal implantation of CMV-RKO cells, transfected with the constitutive β-hCG vector pVSneo-hCG, into SCID mice showed that urinary β-hCG correlated with tumor volume, as previously described (ref. 28; Fig. 3A), whereas the hypoxia-responsive tumor cells 5HRE-RKO showed significantly less expression for the same tumor volume (Fig. 3A,, inset). Similar data were obtained for the CMV-FaDu and 5HRE-FaDu cell lines (data not shown). Manipulations of tumor oxygenation showed that increasing hypoxia resulted in elevated urinary β-hCG secretion. Subjecting mice with 5HRE-RKO xenografts to 72 hours of breathing 10% oxygen increased urinary β-hCG up to 5.7-fold over the mice breathing room air (21% oxygen; Fig. 3B). Similarly, irradiation of the tumor implantation site before tumor inoculation, a procedure that increases tumor hypoxia (36), resulted in elevated urinary β-hCG 5.8-fold over unirradiated controls in 5HRE-FaDu–bearing mice (P = 0.02, Mann-Whitney; Fig. 3C).

Figure 3.

Characterization of the 5HRE β-hCG system in transplanted tumors in immunodeficient mice. A, relationship between urinary β-hCG and tumor size. Dashed line, relationship between urinary β-hCG levels (corrected for urine creatinine) and tumor size in CMV-RKO xenografts constitutively producing β-hCG. Solid line and inset, relationship between urinary β-hCG levels (corrected for urine creatinine) and tumor size in HIF-1α-responsive 5HRE-RKO xenografts. Points, average value for 10 urine samples and the corresponding tumor volume measurements; bars, SE. Thirty mice were used for this experiment, 15 for CMV-RKO tumors, and 15 for 5HRE-RKO tumors. B, expression of urinary β-hCG in response to different manipulations of tumor oxygenation. Solid line, rapid increase in urinary β-hCG level when 5HRE-RKO tumor-bearing mice were placed in a 10% O2 chamber (solid arrow) for 72 hours. Urinary β-hCG subsequently returned to near pretreatment level when the mice were removed from the chamber (dashed arrow). Dashed curve, urinary β-hCG levels for a separate control group of mice that were matched for tumor size and maintained at room air (21% O2) for the duration of the experiment. Points, average value of four mice; bars, SE. C, increased urinary β-hCG levels in 5HRE-FaDu xenografts that received tumor bed irradiation (n = 3) compared with unirradiated controls (n = 5). Because tumors growing in irradiated tumor beds grow slower than control tumors, mice with unirradiated tumor beds were sacrificed 6 weeks after injection, whereas those with irradiated tumor beds were sacrificed after 16 weeks. Urine was collected before sacrifice, assayed for β-hCG and creatinine, and these values were corrected for the mass of the excised tumor. Columns, average; bars, SE (P = 0.02, Mann-Whitney).

Figure 3.

Characterization of the 5HRE β-hCG system in transplanted tumors in immunodeficient mice. A, relationship between urinary β-hCG and tumor size. Dashed line, relationship between urinary β-hCG levels (corrected for urine creatinine) and tumor size in CMV-RKO xenografts constitutively producing β-hCG. Solid line and inset, relationship between urinary β-hCG levels (corrected for urine creatinine) and tumor size in HIF-1α-responsive 5HRE-RKO xenografts. Points, average value for 10 urine samples and the corresponding tumor volume measurements; bars, SE. Thirty mice were used for this experiment, 15 for CMV-RKO tumors, and 15 for 5HRE-RKO tumors. B, expression of urinary β-hCG in response to different manipulations of tumor oxygenation. Solid line, rapid increase in urinary β-hCG level when 5HRE-RKO tumor-bearing mice were placed in a 10% O2 chamber (solid arrow) for 72 hours. Urinary β-hCG subsequently returned to near pretreatment level when the mice were removed from the chamber (dashed arrow). Dashed curve, urinary β-hCG levels for a separate control group of mice that were matched for tumor size and maintained at room air (21% O2) for the duration of the experiment. Points, average value of four mice; bars, SE. C, increased urinary β-hCG levels in 5HRE-FaDu xenografts that received tumor bed irradiation (n = 3) compared with unirradiated controls (n = 5). Because tumors growing in irradiated tumor beds grow slower than control tumors, mice with unirradiated tumor beds were sacrificed 6 weeks after injection, whereas those with irradiated tumor beds were sacrificed after 16 weeks. Urine was collected before sacrifice, assayed for β-hCG and creatinine, and these values were corrected for the mass of the excised tumor. Columns, average; bars, SE (P = 0.02, Mann-Whitney).

Close modal

Immunohistochemical staining of EF5, a known extrinsic hypoxia marker (30), and β-hCG expression show colocalization of the two markers, with both staining the perinecrotic areas of the tumor (Fig. 4A). However, the colocalization of the two markers is not 100%, as the oxygen dependence and kinetics of HIF-1 activation and hypoxia marker bindings are not identical (43).

Figure 4.

A, colocalization between β-hCG and EF5 in 5HRE-RKO xenografts. Mice were injected with EF5 1.5 hours before sacrifice and flash-frozen tumors were stained for both β-hCG and EF5. β-hCG, immunofluorescent staining of β-hCG; EF5, immunofluorescent staining of EF5 using the same tissue slide; β-hCG + EF5, merged image of β-hCG and EF5; DAPI, nuclear staining of the same slide, showing no edge or folding artifacts in this tissue section. Bar, 500 μm. B, correlation between cellular pimonidazole staining (assayed by flow cytometry) and urinary β-hCG (assayed by electrochemiluminescent immunoassay) in 5HRE-RKO tumor growing in SCID mice (n = 10; r = 0.66, r2 = 0.40, and P = 0.0037). C, increased pimonidazole staining (assayed by flow cytometry) and urinary β-hCG levels (assayed by electrochemiluminescent immunoassay) in 5HRE-RKO xenografts treated with 10% O2 (n = 10) compared with room air–treated controls (n = 10).

Figure 4.

A, colocalization between β-hCG and EF5 in 5HRE-RKO xenografts. Mice were injected with EF5 1.5 hours before sacrifice and flash-frozen tumors were stained for both β-hCG and EF5. β-hCG, immunofluorescent staining of β-hCG; EF5, immunofluorescent staining of EF5 using the same tissue slide; β-hCG + EF5, merged image of β-hCG and EF5; DAPI, nuclear staining of the same slide, showing no edge or folding artifacts in this tissue section. Bar, 500 μm. B, correlation between cellular pimonidazole staining (assayed by flow cytometry) and urinary β-hCG (assayed by electrochemiluminescent immunoassay) in 5HRE-RKO tumor growing in SCID mice (n = 10; r = 0.66, r2 = 0.40, and P = 0.0037). C, increased pimonidazole staining (assayed by flow cytometry) and urinary β-hCG levels (assayed by electrochemiluminescent immunoassay) in 5HRE-RKO xenografts treated with 10% O2 (n = 10) compared with room air–treated controls (n = 10).

Close modal

We also did quantitative assessment of urinary β-hCG and misonidazole staining in 5HRE-RKO tumors (Fig. 4B) using flow cytometry. There was a significant correlation between urinary β-hCG and pimonidazole staining in the mice evaluated (r2 = 0.40, P = 0.0037; Fig. 4B). A similar increase in both urinary β-hCG levels and pimonidazole staining was observed when the mice were treated with 3 days of 10% O2 (n = 10) compared with room-air controls (n = 10; Fig. 4C).

Flavopiridol studies. Treatment of 5HRE-FaDu cells with 100 nmol/L flavopiridol significantly inhibited β-hCG secretion under hypoxia whereas there was minimal effect on β-hCG secretion of CMV-FaDu or CMV-RKO cells (Fig 5A). Flavopiridol had the same effect on FaDu cells transfected with 5HRE-luciferase and CMV-luciferase (Fig. 5B), illustrating that flavopiridol does not inhibit secretion per se. Mice were matched for pretreatment urinary β-hCG levels before being assigned to treatment with either flavopiridol or DMSO alone. Whereas control mice treated with daily DMSO injection had rising β-hCG levels, those treated with daily flavopiridol injections had stable urinary β-hCG levels and the difference between the 2 groups was ∼2-fold (P = 0.09, n = 17 mice/group; Fig. 5C) starting after day 5. We observed a similar 2-fold difference in a smaller cohort of mice (n = 5 for each treatment group) treated with 5 days of flavopiridol (data not shown). Because the two curves seemed to remain stably separated after day 5, we pooled the urinary β-hCG values for days 6, 8, and 10 together and compared the pooled values for two treatment groups (Fig. 5). There was a significant difference in the mean urinary β-hCG level for the flavopiridol-treated mice compared with that of control mice (0.65 +/− 0.06 for flavopiridol versus 1.28 +/− 0.25 for control; P = 0.01). Immunoblot analysis showed no inhibition of HIF-1α protein accumulation on exposure to 100 nmol/L flavopiridol (data not shown) at either 6 or 24 hours of hypoxia, suggesting that the drug does not directly affect HIF-1α protein expression but rather its transactivating activity or the expression of downstream HIF-1 targets.

Figure 5.

Flavopiridol inhibits β-hCG induction by the 5HRE promoter in vitro and in vivo. A, electrochemiluminescent immunoassay studies showing a decrease in hypoxia-induced β-hCG secretion on treatment with 100 nmol/L flavopiridol in RKO and FaDu cells. Ratio of β-hCG secretion between 5HRE and CMV, showing the effects that are specific to the 5HRE promoter and not the CMV promoter. B, 5HRE-luciferase and CMV-luciferase transient transfection studies of FaDu cells showing a decrease in hypoxia-induced luciferase expression. Ratio of 5HRE expression to CMV expression. C, in vivo 5HRE-FaDu xenograft studies showing a difference in urinary β-hCG levels between flavopiridol-treated mice (5 mg/kg/d for 10 days) and DMSO control mice. Mice were matched for pretreatment urinary β-hCG levels (day −1). Seventeen mice were used per group and urine was collected every other day during treatment. Points and columns, mean; bars, SE. Left, serial urinary β-hCG levels during treatment. P values (flavopiridol versus control) are 0.09 and 0.10 for days 6 and 8 and >0.10 for the remainders. Right, mean and SE for pooled day 6 to 10 urinary β-hCG values for flavopiridol-treated compared with control mice (P = 0.01).

Figure 5.

Flavopiridol inhibits β-hCG induction by the 5HRE promoter in vitro and in vivo. A, electrochemiluminescent immunoassay studies showing a decrease in hypoxia-induced β-hCG secretion on treatment with 100 nmol/L flavopiridol in RKO and FaDu cells. Ratio of β-hCG secretion between 5HRE and CMV, showing the effects that are specific to the 5HRE promoter and not the CMV promoter. B, 5HRE-luciferase and CMV-luciferase transient transfection studies of FaDu cells showing a decrease in hypoxia-induced luciferase expression. Ratio of 5HRE expression to CMV expression. C, in vivo 5HRE-FaDu xenograft studies showing a difference in urinary β-hCG levels between flavopiridol-treated mice (5 mg/kg/d for 10 days) and DMSO control mice. Mice were matched for pretreatment urinary β-hCG levels (day −1). Seventeen mice were used per group and urine was collected every other day during treatment. Points and columns, mean; bars, SE. Left, serial urinary β-hCG levels during treatment. P values (flavopiridol versus control) are 0.09 and 0.10 for days 6 and 8 and >0.10 for the remainders. Right, mean and SE for pooled day 6 to 10 urinary β-hCG values for flavopiridol-treated compared with control mice (P = 0.01).

Close modal

The system described here allows real-time, noninvasive monitoring of tumor hypoxia before and after interventions in live animals. Theoretically, this system should work for both s.c. tumors, as shown here, as well as orthotopically implanted tumors, although further testing will be needed to validate this hypothesis. In comparison with other systems that allow for serial measurement of hypoxia, such as oxygen sensing devices or imaging approaches, the urinary β-hCG system is simple,noninvasive, and does not require expensive imaging equipments.

HIF-1α–mediated gene transcription represents a major hypoxia regulated signaling pathway and is a valid therapeutic target (16, 44). Recent data demonstrating that HIF-1α cooperates with other oncogenic signaling pathways contribute to the anticancer strategy for inhibiting this factor (4548). Current approaches include the use of antisense constructs (49), small molecule inhibitors of the HIF-1 pathway (15, 40, 41), and methods of destabilizing the protein (16). However, one of the biggest challenges for the future development of HIF-1 inhibitors lies in the establishment of surrogate markers that are sensitive enough for in vivo and clinical testing of these agents. Noninvasive imaging techniques using imidazole- and non-imidazole-based agents coupled with PET and single-photon emission computed tomography hold promise but are still at early stage of development and may not have sufficient resolution to provide quantitative information on the efficacy of new HIF-1 inhibitors (22, 50). Similarly, strategies that use optical imaging to assess changes in HIF stability require an imaging facility with the prerequisite cameras. More recently, Serganova et al. (23) described an elegant approach using PET to image FEAU products of an 8HRE-driven herpes simplex virus type 1 thymidine kinase/green fluorescent protein gene fusion. In vitro and in vivo studies showed responsiveness of this construct to changes in concentrations of oxygen and hypoxia mimetic and in induction of acute hypoxia by ischemia-reperfusion injury. However, this approach requires a dedicated animal PET scanner and local imaging expertise. The method described by our group using 5HRE β-hCG reporter gene has the potential for studying novel hypoxia-targeted drugs and HIF-1α inhibitors in both cell cultures and transplanted tumors in a rapid and inexpensive manner.

A pitfall of this approach is its reliance on measurements of tumor volume with calipers for β-hCG normalization. Although tumor volume is commonly used to assess tumor growth and treatment response in xenograft models, they are error-prone, highly subjected to interobserver variability, and do not take into account nonviable or necrotic regions in the tumors that do not produce or secrete β-hCG. A better approach is to normalize the β-hCG levels to the levels of another secreted marker that is either engineered to be constitutively secreted by the tumor cells or spontaneously produced and secreted by the tumor cells themselves. In addition, this second marker has to be efficiently filtered through the renal glomeruli and directly secreted into the urine for ease of noninvasive measurement. One such candidate marker is the κ or λ immunoglobulin light chain. We are in the process of exploring this dual marker concept to overcome the previously stated problems associated with caliper measurements.

Although our xenograft study with 5HRE-FaDu cells and flavopiridol showed a wide variability in the urinary β-hCG measurements, we did observe a progressive increase in the levels of secreted urinary β-hCG in most saline-treated mice, whereas the levels in the flavopiridol-treated mice remained the same or declined in most mice. Starting from day 5 of treatment, the difference in the urinary β-hCG levels between the saline-treated and flavopiridol-treated mice was ∼2-fold. This is consistent with the data reported for other in vivo systems that used HRE-based luciferase reporters for assessing the efficacy of HIF-1 inhibitors. Kung et al. (15) showed a 50% decrease in the luminescence ratio of hypoxia reporter activity (normalized to a constitutive internal control) after two doses of chetomin injection. Only two mice were used per group; therefore, there was no statistical analysis. Rapisarda et al. (40) noted progressive increases in luminescence in the vehicle-treated mice compared with stable or slight decreases in luminescence in the topotecan-treated mice. All values were normalized to the calculated tumor volumes. These mice were treated with low-dose topotecan continually for 10 days, a schedule similar to ours for flavopiridol.

Our in vitro and in vivo data suggest that flavopiridol, when administered at low doses, decreases the levels of HIF-reporter genes. Flavopiridol has been shown to be a broad mRNA synthesis inhibitor, but at lower concentrations selectively inhibits inducible mRNAs (31), such as those regulated by HIF-1α. Similarly, low-dose flavopiridol (100 nmol/L) blocks the 5HRE reporter genes significantly but leaves the CMV reporter genes mostly unchanged. Although immunoblot analysis shows no significant inhibition of HIF-1α protein levels during hypoxia, flavopiridol may be working as a selective inhibitor of mRNA synthesis by blocking induction of the HRE-regulated reporter transcript. This effect is specific to the inducible mRNA species as we observed no effect on the CMV regulated reporter.

In summary, we have developed a cost-effective and noninvasive system for real-time monitoring of tumor hypoxia in vivo. This system is responsive to different manipulations of tumor oxygenation and has the potential to be used for in vivo studies with compounds targeting HIF-1α and HIF-1α–mediated gene expression.

Note: D.W. Nelson and H. Cao contributed equally to this work.

Presented at the 94th Annual Meeting of the AACR, July 11-14, 2003, Washington, DC.

Grant support: U.S. Public Health Service grant CA 67166 (supported all except A.C. Koong, J.D. Faix, and B. Sunar-Reeder), the Damon Runyon-Lilly Clinical Investigator Award (A.C. Koong), and an Aventis Investigator-Initiated Award (Q-T. Le and H. Cao).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Yun Z, Giaccia AJ. Tumor deprivation of oxygen and tumor suppressor gene function.
Methods Mol Biol
2003
;
223
:
485
–504.
2
Rudat V, Vanselow B, Wollensack P, et al. Repeatability and prognostic impact of the pretreatment pO(2) histography in patients with advanced head and neck cancer.
Radiother Oncol
2000
;
57
:
31
–7.
3
Brizel DM, Dodge RK, Clough RW, Dewhirst MW. Oxygenation of head and neck cancer: changes during radiotherapy and impact on treatment outcome.
Radiother Oncol
1999
;
53
:
113
–7.
4
Hockel M, Schlenger K, Hockel S, Vaupel P. Hypoxic cervical cancers with low apoptotic index are highly aggressive.
Cancer Res
1999
;
59
:
4525
–8.
5
Hockel M, Schlenger K, Aral B, et al. Association between tumor hypoxia and malignant progression in advanced cancer of the uterine cervix.
Cancer Res
1996
;
56
:
4509
–15.
6
Fyles AW, Milosevic M, Wong R, et al. Oxygenation predicts radiation response and survival in patients with cervix cancer.
Radiother Oncol
1998
;
48
:
149
–56.
7
Nordsmark M, Overgaard M, Overgaard J. Pretreatment oxygenation predicts radiation response in advanced squamous cell carcinoma of the head and neck.
Radiother Oncol
1996
;
41
:
31
–9.
8
Fyles A, Milosevic M, Hedley D, et al. Tumor hypoxia has independent predictor impact only in patients with node-negative cervix cancer.
J Clin Oncol
2002
;
20
:
680
–7.
9
Cairns RA, Hill RP. Acute hypoxia enhances spontaneous lymph node metastasis in an orthotopic murine model of human cervical carcinoma.
Cancer Res
2004
;
64
:
2054
–61.
10
Brizel DM, Scully SP, Harrelson JM, et al. Tumor oxygenation predicts for the likelihood of distant metastases in human soft tissue sarcoma.
Cancer Res
1996
;
56
:
941
–3.
11
Semenza GL. Targeting HIF-1 for cancer therapy.
Nat Rev Cancer
2003
;
3
:
721
–32.
12
Harris AL. Hypoxia—a key regulatory factor in tumour growth.
Nat Rev Cancer
2002
;
2
:
38
–47.
13
Denko NC, Fontana LA, Hudson KM, et al. Investigating hypoxic tumor physiology through gene expression patterns.
Oncogene
2003
;
22
:
5907
–14.
14
Melillo G. HIF-1: a target for cancer, ischemia and inflammation—too good to be true?
Cell Cycle
2004
;
3
:
154
–5.
15
Kung A, Zabludoff S, France D, et al. Small molecule blockade of transcriptional coatcivation of the hypoxia-inducible factor pathway.
Cancer Cell
2004
;
6
:
33
–43.
16
Giaccia A, Siim BG, Johnson RS. HIF-1 as a target for drug development.
Nat Rev Drug Discov
2003
;
2
:
803
–11.
17
Saitoh J, Sakurai H, Suzuki Y, et al. Correlations between in vivo tumor weight, oxygen pressure, 31P NMR spectroscopy, hypoxic microenvironment marking by β-d-iodinated azomycin galactopyranoside (β-D-IAZGP), and radiation sensitivity.
Int J Radiat Oncol Biol Phys
2002
;
54
:
903
–9.
18
Olsen DR, Rofstad EK. Monitoring of tumor reoxygenation following irradiation by 31P magnetic resonance spectroscopy: an experimental study of human melanoma xenografts.
Radiother Oncol
1999
;
52
:
261
–7.
19
Kavanagh MC, Tsang V, Chow S, et al. A comparison in individual murine tumors of techniques for measuring oxygen levels.
Int J Radiat Oncol Biol Phys
1999
;
44
:
1137
–46.
20
Evans SM, Koch CJ. Prognostic significance of tumor oxygenation in humans.
Cancer Lett
2003
;
195
:
1
–16.
21
Bentzen L, Keiding S, Horsman MR, et al. Assessment of hypoxia in experimental mice tumours by [18F]fluoromisonidazole PET and pO2 electrode measurements. Influence of tumour volume and carbogen breathing.
Acta Oncol
2002
;
41
:
304
–12.
22
Koch CJ, Evans SM. Non-invasive PET and SPECT imaging of tissue hypoxia using isotopically labeled 2-nitroimidazoles.
Adv Exp Med Biol
2003
;
510
:
285
–92.
23
Serganova I, Doubrovin M, Vider J, et al. Molecular imaging of temporal dynamics and spatial heterogeneity of hypoxia-inducible factor-1 signal transduction activity in tumors in living mice.
Cancer Res
2004
;
64
:
6101
–8.
24
Wen B, Burgman P, Zanzonico P, et al. A preclinical model for noninvasive imaging of hypoxia-induced gene expression; comparison with an exogenous marker of tumor hypoxia.
Eur J Nucl Med Mol Imaging
2004
;
31
:
1530
–8.
25
Jenkins WT, Evans SM, Koch CJ. Hypoxia and necrosis in rat 9L glioma and Morris 7777 hepatoma tumors: comparative measurements using EF5 binding and the Eppendorf needle electrode.
Int J Radiat Oncol Biol Phys
2000
;
46
:
1005
–17.
26
Brown JM, Le QT. Tumor hypoxia is important in radiotherapy, but how should we measure it?
Int J Radiat Oncol Biol Phys
2002
;
54
:
1299
–301.
27
Safran M, Kim WY, Kung AL, et al. Mouse reporter strain for noninvasive bioluminescent imaging of cells that have undergone Cre-mediated recombination.
Mol Imaging
2003
;
2
:
297
–302.
28
Shih IM, Torrance C, Sokoll LJ, et al. Assessing tumors in living animals through measurement of urinary β-human chorionic gonadotropin.
Nat Med
2000
;
6
:
711
–4.
29
Shibata T, Giaccia AJ, Brown JM. Development of a hypoxia-responsive vector for tumor-specific gene therapy.
Gene Ther
2000
;
7
:
493
–8.
30
Evans SM, Hahn S, Pook DR, et al. Detection of hypoxia in human squamous cell carcinoma by EF5 binding.
Cancer Res
2000
;
60
:
2018
–24.
31
Lam LTPO, Peng AC, Rosenwald A, et al. Genomic-scale measurement of mRNA turnover and the mechanisms of action of the anti-cancer drug flavopiridol.
Genome Biol
2001
;
2
.
32
Melillo G, Sausville EA, Cloud K, et al. Flavopiridol, a protein kinase inhibitor, down-regulates hypoxic induction of vascular endothelial growth factor expression in human monocytes.
Cancer Res
1999
;
59
:
5433
–7.
33
Masson N, Willam C, Maxwell PH, Pugh CW, Ratcliffe PJ. Independent function of two destruction domains in hypoxia-inducible factor-α chains activated by prolyl hydroxylation.
EMBO J
2001
;
20
:
5197
–206.
34
Chan DA, Sutphin PD, Denko NC, Giaccia AJ. Role of prolyl hydroxylation in oncogenically stabilized hypoxia-inducible factor-1α.
J Biol Chem
2002
;
277
:
40112
–7.
35
Vordermark D, Shibata T, Brown JM. Green fluorescent protein is a suitable reporter of tumor hypoxia despite an oxygen requirement for chromophore formation.
Neoplasia
2001
;
3
:
527
–34.
36
Kim IH, Lemmon MJ, Brown JM. The influence of irradiation of the tumor bed on tumor hypoxia: measurements by radiation response, oxygen electrodes, and nitroimidazole binding.
Radiat Res
1993
;
135
:
411
–7.
37
Vordermark D, Brown J. Evaluation of hypoxia-inducible factor-1α (HIF-1α) as an intrinsic marker of tumor hypoxia in U87 MG human glioblastoma: in vitro and xenograft studies.
Int J Radiat Oncol Biol Phys
2003
;
56
:
1184
–93.
38
Olive P, Durand R, Raleigh J, Luo C, Aquino-Parsons C. Comparison between the comet assay and pimonidazole binding for measuring tumour hypoxia.
Br J Cancer
2000
;
83
:
1525
–31.
39
Glanz SA, Slinker BK. Primer of applied regression analysis of variance. New York: McGraw-Hill, Inc.; 1990. p. 50–109, 512–168.
40
Rapisarda A, Zalek J, Hollingshead M, et al. Schedule-dependent inhibition of hypoxia-inducible factor-1α protein accumulation, angiogenesis, and tumor growth by topotecan in U251-HRE glioblastoma xenografts.
Cancer Res
2004
;
64
:
6845
–8.
41
Rapisarda A, Uranchimeg B, Scudiero DA, et al. Identification of small molecule inhibitors of hypoxia-inducible factor 1 transcriptional activation pathway.
Cancer Res
2002
;
62
:
4316
–24.
42
Yeo EJ, Chun YS, Cho YS, et al. YC-1: a potential anticancer drug targeting hypoxia-inducible factor 1.
J Natl Cancer Inst
2003
;
95
:
516
–25.
43
Bussink JKJ, van der Kogel AJ. Tumor hypoxia at the micro-regional level: clinical relevance and predictive value of exogenous and endogenous hypoxic cell markers.
Radiother Oncol
2003
;
67
:
3
–15.
44
Brown JM, Wilson WR. Exploiting tumour hypoxia in cancer treatment.
Nat Rev Cancer
2004
;
4
:
437
–47.
45
Zundel W, Schindler C, Haas-Kogan D, et al. Loss of PTEN facilitates HIF-1-mediated gene expression.
Genes Dev
2000
;
14
:
391
–6.
46
Zhong H, Chiles K, Feldser D, et al. Modulation of hypoxia-inducible factor 1α expression by the epidermal growth factor/phosphatidylinositol 3-kinase/PTEN/AKT/FRAP pathway in human prostate cancer cells: implications for tumor angiogenesis and therapeutics.
Cancer Res
2000
;
60
:
1541
–5.
47
McMenamin ME, Soung P, Perera S, et al. Loss of PTEN expression in paraffin-embedded primary prostate cancer correlates with high Gleason score and advanced stage.
Cancer Res
1999
;
59
:
4291
–6.
48
Laughner E, Taghavi P, Chiles K, Mahon PC, Semenza GL. HER2 (neu) signaling increases the rate of hypoxia-inducible factor 1α (HIF-1α) synthesis: novel mechanism for HIF-1-mediated vascular endothelial growth factor expression.
Mol Cell Biol
2001
;
21
:
3995
–4004.
49
Sun X, Kanwar JR, Leung E, et al. Gene transfer of antisense hypoxia inducible factor-1 α enhances the therapeutic efficacy of cancer immunotherapy.
Gene Ther
2001
;
8
:
638
–45.
50
Van de Wiele C, Lahorte C, Oyen W, et al. Nuclear medicine imaging to predict response to radiotherapy: a review.
Int J Radiat Oncol Biol Phys
2003
;
55
:
5
–15.