Cancer Research Annual Meeting 2010 Protein Translation and Cancer
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
Cancer Research 67, 3801, April 15, 2007. doi: 10.1158/0008-5472.CAN-06-3780
© 2007 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bock, J. M.
Right arrow Articles by Trask, D. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bock, J. M.
Right arrow Articles by Trask, D. K.

Experimental Therapeutics, Molecular Targets, and Chemical Biology

Celecoxib Toxicity Is Cell Cycle Phase Specific

Jonathan M. Bock1, Sarita G. Menon2, Lori L. Sinclair1, Nichole S. Bedford1, Prabhat C. Goswami2, Frederick E. Domann2 and Douglas K. Trask1,2

Departments of 1 Otolaryngology-Head and Neck Surgery and 2 Radiation Oncology, University of Iowa Hospital and Clinics, Iowa City, Iowa

Requests for reprints: Douglas K. Trask, Department of Otolaryngology-Head and Neck Surgery, University of Iowa Hospitals & Clinics, 200 Hawkins Drive, Iowa City, IA 52242. Phone: 319-356-1616; Fax: 319-356-4547; E-mail: douglas-trask{at}uiowa.edu.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Celecoxib inhibits proliferation and induces apoptosis in human tumors, but the molecular mechanisms for these processes are poorly understood. In this study, we evaluated the ability of celecoxib to induce toxicity in head and neck squamous cell carcinomas (HNSCC) and explored the relationships between celecoxib-induced cell cycle inhibition and toxicity in HNSCC. Celecoxib inhibited the proliferation of UM-SCC-1 and UM-SCC-17B cells both in vitro and in vivo, accompanied by G1 phase cell cycle arrest and apoptosis. Celecoxib induced p21waf1/cip1 at the transcriptional level independent of wild-type p53 function, leading to decreased expression of cyclin D1 and hypophosphorylation of Rb, with subsequent marked downstream decreases in nuclear E2F-1 protein expression and E2F transactivating activity by luciferase reporter assay. Cell cycle phase–specific cytometric sorting showed that celecoxib induced clonogenic toxicity preferentially to cells within the S phase greater than G1 and G2 phases. Levels of p21waf1/cip1 and cyclin D1 protein were reduced in the S phase compared with the G1 and G2 phases, suggesting a possible protective role for p21waf1/cip1 expression in celecoxib toxicity. In conclusion, we show that celecoxib has marked antiproliferative activity against head and neck cancer cells through transcriptional induction of p21waf1/cip1 and G1 phase accumulation leading to S phase–specific clonogenic toxicity. We additionally show that a profound inhibition of nuclear E2F function provides a possible mechanism for this S phase–specific toxicity. [Cancer Res 2007;67(8):3801–8]


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Over 27,000 new cases of squamous cell carcinoma of the head and neck (HNSCC) will be diagnosed in the United States this year, and surgery, chemotherapy, and radiation therapy remain the hallmarks of modern HNSCC treatment (1). Despite dramatic improvements in surgical technique, radiation delivery, and chemotherapy regimens, overall survival rates for this disease have remained nearly constant for the last 40 years. Any therapeutic modalities that would increase curative outcomes without additional toxicity would be of great benefit.

Cyclooxygenase-2 (COX-2) inhibition has emerged as a novel molecular target for the prevention and treatment of head and neck cancers with this goal in mind (2). COX-2 is the key inducible enzyme in the conversion of arachidonic acid into prostaglandins (3) and seems to be elevated in a number of human cancers (4–8), including HNSCC (9). Celecoxib, a selective COX-2 inhibitor, is currently Food and Drug Administration–approved for chemoprevention in familial adenomatous polyposis based on demonstrable inhibition of colon polyp formation, further solidifying the use of this agent in cancer therapy (10). Recent large-cohort clinical trials have shown the significant activity that celecoxib has in vivo against colorectal adenomas (11). These same studies have also shown a risk of cardiac toxicity with long-term administration of celecoxib at high doses, and authors of these studies suggest that the significant antineoplastic benefits of celecoxib need to be balanced against these toxicities in human cancer treatment protocols (12, 13). It seems likely that long-term use of nonsteroidal anti-inflammatory drugs (NSAID) for chemoprevention may not be possible due to these toxicity issues, but short-term combination therapies with standard agents may significantly augment standard cancer treatments. A clearer understanding of the mechanism of the toxicity of celecoxib will allow for improved clinical outcomes while limiting treatment morbidity.

NSAIDs have been shown to markedly alter cell cycle kinetics in a number of human tumors, and cell cycle alteration seems to be a major contributing factor in celecoxib-mediated cancer toxicity. Celecoxib usually inhibits cell cycle progression through the G1 phase, and this inhibition of proliferation seems to be independent of COX-2 enzyme inhibition in breast cancer models (14). Ovarian tumors are similarly blocked in G1 following treatment with NS-398 (a COX-2–specific inhibitor), and this effect also seems to be independent of COX-2 inhibition (15). This implies that the actual mechanism underlying the antiproliferative activity of NSAIDs may be independent of COX-2 inhibition. The cellular factors involved in this cell cycle inhibition and how these changes affect the subsequent toxicity of celecoxib and other NSAIDs are not clearly known. Celecoxib alters the levels of several important G1 cell cycle checkpoint proteins including p21waf1/cip1 (p21) and cyclin D1, and this may mediate the activity of this drug (14, 16). NSAIDs have additionally been shown to induce peroxisome proliferator–activated receptor (PPAR) transcription factor family members (17), and PPAR-{gamma} antisense oligonucleotides have been shown to abrogate the antiproliferative effects of NSAID compounds. These data propose a possible mechanism for the antineoplastic activity of NSAIDs apart from COX-2 inhibition through cell cycle inhibition by a transcriptional activation system. Unfortunately, few studies have confirmed these mechanistic patterns in HNSCC, and the connections between these pathways in any cell system remain unclear at best.

In this paper, we show that celecoxib has strong antiproliferative and toxic activity against HNSCC cells, shown by decreased in vitro and in vivo cellular growth, G1 cell cycle phase accumulation, and apoptosis. We show that celecoxib induces expression of both PPAR-{gamma} and p21 protein independent of wild-type p53 activity, and that the induction of p21 occurs at the transcriptional level. We further show that these G1 checkpoint protein alterations lead to a marked decrease in E2F activity and subsequent cell cycle phase–specific toxicity in the S phase, correlating with p21 expression changes across the cell cycle phases. To date, no one has shown cell cycle phase–specific toxicity of any NSAID agent, and this describes a connection between the cytostatic and toxic effects of celecoxib treatment. This cell cycle phase–specific toxicity also points toward a potential avenue for therapeutic synergism by combining celecoxib with treatment modalities that are active in the S and G2 phases, including radiation therapy and chemotherapy.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cells and reagents. All drugs and biochemicals were supplied by Sigma-Aldrich Fine Chemical Co. (St. Louis, MO), unless otherwise specified. Celecoxib was purchased from LKT Laboratories (St. Paul, MN). Celecoxib was prepared as a 0.1 mol/L stock solution in DMSO. Tissue culture reagents were supplied by Invitrogen Corporation/Life Technologies Life Sciences (Carlsbad, CA). The UM-SCC-1 and UM-SCC-17B oral cancer cell lines were generously provided by Dr. Thomas Carey (University of Michigan, Ann Arbor, MI). All cell lines were maintained as monolayer cultures in DMEM supplemented with 10% fetal bovine serum, 1% penicillin-streptomycin, 1% L-glutamine, and 1% non–essential amino acids.

Proliferation assays. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays were done according to standard published protocols (18). Cells were incubated with drug and harvested after 96 h of drug exposure, followed by exposure to the MTT reagent (500 µg/mL) and data analysis. Relative MTT absorbance was measured at 570 nm on a Syva Autotrak EIA Autoreader (Bio-Tek Instruments, Inc., Winooski, VT). Growth curve assays were done according to standard parameters. For xenograft assays, 6- to 8-week-old athymic nu/nu female mice (National Cancer Institute, Bethesda, MD) received bilateral flank injections of 1 x 106 HNSCC cells. For growth analysis, mice were immediately started on dietary celecoxib chow (1,500 ppm in irradiated NIH 31 chow, Harlan Teklad, Madison, WI) or control chow (NIH 31, Harlan Teklad). For gavage experiments, mice received daily oral gavage with a solution of 1,500 mg/kg/day celecoxib in 0.5% methylcellulose or control methylcellulose gavage for 14 days. Tumor volume was calculated using the following equation: tumor volume = {pi}/6 x (large diameter) x (small diameter)2. All animal experiments were done according to University of Iowa Animal Care Facility protocols.

Flow cytometry assays. For cell cycle phase–sorting experiments, cells were incubated with 5 µmol/L Hoechst 33342 (Calbiochem, San Diego, CA) for 1 h at 37°C and filtered through 70-µm mesh. Cells were sorted on a Becton Dickinson (Mountain View, CA) FACS DiVa using 100 mW at 488 nm from a Coherent I90C-4 argon ion laser for generation of forward- and side-scatter signals and 100 mW of multi-line UV from a Coherent 302C krypton laser for Hoechst excitation. Hoechst emissions were detected after passing through a 510SP dichroic filter and 424/44 band-pass filter. Amorphous gates were used in bivariate plots of forward-scatter area (FSC-A) versus side-scatter area (SSC-A), forward-scatter area (FSC-A) versus forward-scatter width (FSC-W), and Hoechst area versus Hoechst width to eliminate dead cells, debris, and aggregates. Following sorting, cells were plated at clonogenic density (500–2,500 cells) into 60-mm tissue culture plates in the absence of drug and allowed to grow for 2 weeks. Additional cells were prepared for SDS-PAGE immunoblotting as described below. Cells were fixed with 70% ethanol for 2 min and then stained with crystal violet (0.5% in H2O). Colonies containing more than 50 cells were scored, and survival fractions were calculated. Active caspase-3 assays were done according to manufacturer's instructions (PE-Conjugated Monoclonal Active Caspase-3 Antibody Apoptosis Kit; BD PharMingen, San Diego, CA). Propidium iodide (PI)–staining for sub-G1 DNA accumulation was done according to published protocols (19). Bromodeoxyuridine (BrdUrd) pulse flow cytometric assays for cell cycle progression were done according to previously published guidelines (20).

Immunoblotting. Protein lysates were generated in reducing lysis buffer and boiled for 7 min. Nuclear extracts were prepared according to standard parameters. Protein content was measured by Bradford assay. A total of 20 µg of protein was loaded on 12% agarose gels (except 7.5% for Rb gels) according to standard SDS-PAGE protocol. Proteins were transferred to a nitrocellulose membrane for 1 h according to standard protocol. Equal protein loading was confirmed via Ponceau S staining of nitrocellulose. Blots were probed for 1 h with primary antibody diluted in 5% nonfat dry milk in Tris-buffered saline with 0.1% Tween 20 (TBS-T), followed by 3 x 10-min washes. Primary antibodies included COX-1 (1:200 dilution, Cayman Chemical Company, Ann Arbor, MI), COX-2 (1:250, Cayman Chemical Company), p21waf1/cip1 (1:250, BD PharMingen), p53 (1:500, BD PharMingen), Rb (1:250, BD PharMingen), cyclin D1 (1:333, BD PharMingen), E2F-1 (1:250, Santa Cruz Biotechnology, Santa Cruz, CA), PPAR-{gamma} (1:250, sc-7273, Santa Cruz Biotechnology) actin (1:2,000, Sigma) and {alpha}-tubulin (1:4,000, Oncogene Research Products, San Diego, CA). Blots were then probed with appropriate peroxidase-conjugated secondary antibody (1:5,000–10,000, Jackson Immunoresearch Laboratories, West Grove, PA) in 5% nonfat dry milk in TBS-T for 1 h at room temperature. Blots were washed 5 x 8 min in TBS-T and exposed to enhanced chemiluminescence reagents (Amersham Biosciences, Buckinghamshire, England) followed by radiographic exposure and development. Relative band intensities were determined using CCD camera-based densitometry of bands (Alpha Imager 2200, Alpha Innotech Corporation, San Leandro, CA). Band intensities were first normalized to {alpha}-tubulin in each sample, and results were calculated relative to untreated controls.

Quantitative analysis of p21waf1/cip1 mRNA levels by real-time quantitative reverse transcription-PCR. PCR assays used TaqMan Universal RT-PCR Master Mix (PE Applied Biosystems, Foster City, CA), and reactions were done and analyzed using an ABI Prism 7700 sequence detector equipped with a 96-well thermal cycler (PE Applied Biosystems). RNA was isolated by TRIzol and chloroform extraction (Invitrogen Corporation). About 100 ng of RNA was loaded per PCR reaction. RNA was reverse transcribed (30 min at 48°C, denaturation at 95°C for 10 min), followed by 40 cycles of PCR (92°C for 30 s and 60°C for 30 s). Fluorogenic probes were labeled at 5' and 3' ends with a reporter [6-carboxy-fluorescein (FAM)] and a quencher dye [6-carboxy-tetramethylrhodamine (TAMRA)]. Primers for PCR amplification of p21 were 5'-CTGGAGACTCTCAGGGTCGAA-3' (forward), 5'-GGCGTTTGGAGTGGTAGAAATCT-3'(reverse), and 5'-FAM-ACGGCGGCAGACCAGCATGA-TAMRA-3' (probe; ref. 21). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probes and primers were purchased from Applied Biosystems.

p21 mRNA turnover assay. UM-SCC-1 cells were plated to 50% confluence in 60-mm3 tissue culture plates and allowed to adhere for 24 h. Cells were incubated with celecoxib (100 µmol/L) or DMSO control (0.1%) for 14 h. Actinomycin D was prepared as a 1-mg/mL solution in sterile water, and added to culture plates at 1 µg/mL concentration. RNA was harvested over 12 h using TRIzol and chloroform extraction technique (Invitrogen Corporation). p21 RNA levels were then evaluated using real-time quantitative real-time reverse transcription-PCR (RT-PCR) techniques as described above.

E2F promoter luciferase reporter assays. Dr. Mark May (University of Michigan, Ann Arbor, MI) graciously provided the 3x E2F luciferase reporter construct. pRL-TK Renilla luciferase reporter plasmid was purchased from Promega (Madison, WI). UM-SCC-1 cells were cultured in 24-well plates and pretreated with a range of celecoxib doses for 6 h. A 4:1 ratio of E2F reporter plasmid to pRL-TK vector was transfected into UM-SCC-1 cells using Effectene transfection reagent (Qiagen, Chatsworth, CA) according to manufacturer's instructions. E2F promoter activity was assayed using the Dual Luciferase Reporter kit from Promega according to manufacturer's instructions. Luciferase reactions were done on Luminoskan Ascent Luminometer (Thermo Electron Corporation, Helsinki, Finland). Luciferase activity was reported as a ratio of E2F activity relative to untreated control UM-SCC-1 cells in triplicate experiments.

Data analysis. All experiments were repeated at least in triplicate, and representative data from a single experiment are presented in this manuscript as means, SD, and SE. Statistical comparisons between experimental groups included the use of a two-tailed Student's t test and ANOVA. Data points marked with an asterisk indicate P < 0.05.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Celecoxib inhibits HNSCC viability. UM-SCC-1 and UM-SCC-17B cells were treated with a dose range of celecoxib (0–100 µmol/L) for 96 h and assayed for viability using a standard MTT assay (Fig. 1A ). Celecoxib induced a strong dose-dependent decrease in MTT absorbance in UM-SCC-1 and UM-SCC-17B cells, with an IC50 of 39.9 ± 1.1 and 35.2 ± 2.3 µmol/L, respectively (averaged from triplicate experiments). We also evaluated the ability of celecoxib to inhibit HNSCC proliferation by standard growth curve assays (Fig. 1B). UM-SCC-1 cells were incubated with a range of doses of celecoxib, and samples were harvested 3, 5, and 7 days after drug addition. Celecoxib administration produced a dose-dependent inhibition of cellular proliferation across a wide dose range. These data show that celecoxib treatment produces strong in vitro inhibition of HNSCC viability and proliferation.


Figure 1
View larger version (9K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Celecoxib inhibits proliferation of HNSCC in vitro and in vivo. A, MTT assay. UM-SCC-1 and UM-SCC-17B cells were treated with a dose range of celecoxib and incubated for 120 h, followed by viability assay using standard MTT protocols. B, growth curve. A total of 5 x 105 UM-SCC-1 cells were plated and incubated with a dose range of celecoxib, and proliferation was analyzed at 3, 5, and 7 d. Control cells were incubated with 0.05% DMSO only. C, mouse xenograft growth assay. On day 0, 10 athymic nu/nu female mice were injected with bilateral flank xenograft tumors of 1 x 106 UM-SCC-1 cells. Mice were started immediately on either control NIH 31 chow or NIH 31 chow formulated with 1,500 ppm of celecoxib. Mouse tumor volumes were determined using caliper measurements in two dimensions. Bars, SD of experimental replicates done at least in triplicate. *, P < 0.05.

 
In vivo celecoxib antiproliferative activity was assayed using a mouse xenograft assay (Fig. 1C). Bilateral UM-SCC-1 xenograft tumors were grown on the flanks of athymic nu/nu mice, and treatment mice consumed celecoxib chow (1,500 ppm). Celecoxib administration began on the day of tumor injection, and all mice were allowed continuous access to either drug or control chow throughout the experimental period. Celecoxib strongly inhibited in vivo tumor xenograft growth, inducing nearly a 5-fold decrease in overall tumor size (82%). This decrease in tumor size was statistically significant after 14 days of treatment and remained throughout the duration of the experimental period.

Celecoxib induces apoptosis in HNSCC. The antiproliferative effect of celecoxib can occur through the induction of apoptosis or cell cycle arrest. Therefore, we assessed the ability of celecoxib to induce apoptosis in our HNSCC model by evaluating the levels of active caspase-3, a central and irreversible component of both the cytosolic and mitochondrial apoptosis cascades, following treatment with celecoxib. UM-SCC-1 cells were treated for 36 h with a dose range of celecoxib, fixed in paraformaldehyde, and incubated with a phycoerythrin (PE)-conjugated antibody against active caspase-3. Dose-dependent increases in active caspase-3 levels were seen between 100 and 200 µmol/L celecoxib (Fig. 2A ). We also assayed for apoptosis by analyzing sub-G1 DNA formation. Cells were treated with a similar dose range of celecoxib, and dose-dependent accumulation of sub-G1 DNA was observed again between doses of 100 and 200 µmol/L celecoxib (Fig. 2B).


Figure 2
View larger version (6K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Celecoxib induces apoptosis in HNSCC. A, caspase-3 activation assay. UM-SCC-1 cells were incubated with a dose range of celecoxib for 36 h, fixed in paraformaldehyde, and incubated with PE-conjugated monoclonal antibody against the activated form of caspase-3. Cells were analyzed for PE intensity by flow cytometry. B, sub-G1 DNA fragmentation assay. UM-SCC-1 cells were treated with a dose range of celecoxib for 24 h, fixed in ethanol, and stained with propidium iodide, followed by FACScan analysis for sub-G1 DNA content. Points, means of a representative experiment repeated in triplicate; bars, SD.

 
Celecoxib inhibits progression through the G1-S transition via induction of p21. The effects of celecoxib on cell cycle progression in HNSCC were determined using BrdUrd pulse-chase experiments, allowing for clear visualization of G1, S, and G2 populations as previously described (20). UM-SCC-1 cells were treated with 0 to 100 µmol/L celecoxib for 24 h before ethanol fixation and staining. Celecoxib treatment caused a dose-dependent inhibition of cell cycle progression through the G1-S transition, leading to >77% G1 phase accumulation after 24 h of treatment with 100 µmol/L celecoxib [Fig. 3A (histograms) and B (graphic data )]. A concomitant decrease in both S phase and G2 phase distributions was noted as G1 populations increased.


Figure 3
View larger version (41K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Celecoxib inhibits progression into S phase by induction of p21. UM-SCC-1 cells were treated with indicated doses of celecoxib for 24 h, incubated with BrdUrd for 30 min at 37°C, and then fixed in 90% ethanol. Secondary conjugation with FITC-labeled antibody and propidium iodide staining allowed for identification of G1, S, and G2 populations. A, histogram data from a representative experiment. B, graphical representation of data from experiment in (A); bars, SD from triplicate experimental replicates. *, P < 0.05. C, protein expression changes in HNSCC cells following treatment with celecoxib. UM-SCC-1 cells were treated with a dose range of celecoxib for 24 h, followed by protein extraction and separation by standard SDS-PAGE technique. Proteins were transferred to nitrocellulose membranes and probed with antibodies to important G1 cell cycle checkpoint proteins. Blots were stripped and reprobed for {alpha}-tubulin to confirm even protein loading. Data are from a representative experiment of at least three repetitions. D, tumor protein expression changes following celecoxib treatment. Four athymic nu/nu female mice received s.c. flank injections with 1 x 106 UM-SCC-17B cells, and tumors were allowed to form for 7 d. Mice were subsequently gavaged with control 0.5% methylcellulose solution (lanes 1 and 2) or celecoxib suspension (1,500 mg/kg daily in 0.5% methylcellulose, lanes 3 and 4) for 2 wks. Following euthanasia, tumors were dissected, and proteins were extracted and separated via standard SDS-PAGE protocols, followed by Western blot analysis as above.

 
Further studies were done to evaluate the specific cellular proteins involved in these observed G1 phase cell cycle rearrangements. Celecoxib administration induced COX-2 expression at high doses, but COX-1 levels did not change appreciably following celecoxib administration. This increase in COX-2 expression was not associated with increased prostaglandin E2 (PGE2) levels by enzyme-linked PGE2 immunoassay (data not shown). Celecoxib induced p21 in a dose-dependent fashion, increasing to more than 16-fold of baseline expression by 50 µmol/L (Fig. 3C). This increase in p21 protein expression levels was seen within 6 h of initiation of treatment (data not shown). UM-SCC-1 cells have previously been reported to have wild-type p53 activity (22). Wild-type p53 levels were expressed at very low levels and did not change appreciably following celecoxib treatment. Induction of p21 was accompanied by a decrease in cyclin D1 expression and a dose-dependent increase in the predominance of hypophosphorylated Rb, consistent with the previously observed G1 phase accumulation (Fig. 3A). PPAR-{gamma} may also be involved in the activity of NSAIDs (23, 24), and we therefore evaluated the expression of this transcription factor following celecoxib treatment. PPAR-{gamma} protein levels increased in a dose-dependent fashion by more than 40-fold at the highest doses of celecoxib and were increased by even low doses of celecoxib (1–5 µmol/L). These data show that celecoxib strongly induces p21 and PPAR-{gamma} without significant changes in p53 expression, leading to decreased cyclin D1 expression and a shift to the hypophosphorylated state of Rb.

We also evaluated the effects of celecoxib therapy on G1 checkpoint protein expression in a head and neck tumor xenograft model (Fig. 3D). Four athymic nude mice were injected with bilateral xenograft tumors of UM-SCC-17B cells, and tumors were allowed to grow for 1 week to an average volume of ~100 mm3. Mice were then gavaged with 1,500 mg/kg/day of celecoxib or control methylcellulose solution for 14 days. Tumors were dissected, and protein was extracted, followed by Western blot analysis of G1 checkpoint protein expression, as well as PPAR-{gamma}. As seen in the previous in vitro studies, p21 and PPAR-{gamma} were induced in the tumor xenografts compared with controls (1.5- and 1.25-fold, respectively), with a corresponding 4-fold decrease in cyclin D1 expression. UM-SCC-17B cells also express wild-type p53 (25), and again, p53 levels did not seem to change significantly following celecoxib gavage. This implies that p21 induction was also independent of p53 function in this in vivo head and neck xenograft model. Overall, this pattern of in vivo G1 checkpoint protein expression recapitulates the previous in vitro data and suggests that p21 induction leads to downstream inhibition of cyclin D1 and subsequent hypophosphorylation of Rb.

Celecoxib induces p21 expression by transcriptional activation. Celecoxib strongly induced p21 expression in our model system of HNSCC, and this seemed to be independent of wild-type p53 activity. To determine the mechanism of this increase in p21 protein levels, we evaluated the expression of p21 mRNA following celecoxib administration using quantitative real-time RT-PCR analysis. Subconfluent UM-SCC-1 cells were treated with 100 µmol/L celecoxib or 0.1% DMSO for 14 h. After drug exposure, actinomycin D (1 µg/mL) was added to the culture media to inhibit de novo RNA transcription, and samples were collected over the following 12 h (Fig. 4 ). RNA levels were standardized relative to DMSO 0.1% 0-h control point. Celecoxib treatment induced a 5.3-fold increase in p21 mRNA expression after 14 h of treatment. After addition of actinomycin D, both control and celecoxib treatments showed a marked decrease in p21 mRNA over the following 12 h, with a plateau phase between 4 and 8 h. Final levels of p21 mRNA 12 h after addition of actinomycin D were ~10% of 0-h levels for both groups. The overall rate of decay for both groups was nearly identical, indicating that the mechanism for the observed increase in p21 mRNA levels was due to transcriptional induction and not to decreased RNA degradation.


Figure 4
View larger version (12K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Celecoxib induces transcriptional activation of p21. UM-SCC-1 cells were treated with 100 µmol/L celecoxib for 14 h, and actinomycin D was added to all plates to arrest transcription. RNA was extracted at 4, 8, and 12 h, and relative p21 mRNA levels were determined by real-time quantitative RT-PCR. Data are from a representative experiment repeated in triplicate. Bars, SD.

 
Celecoxib treatment induces profound inhibition of E2F activity. We showed above that p21 induction induced a predominance of the hypophosphorylated form of Rb following celecoxib treatment. In this state, E2F should be highly bound to Rb in the cytoplasm of the cell, thereby inactivating E2F activity. We therefore evaluated the transcriptional transactivating activity of the E2F promoter using a luciferase reporter assay (Fig. 5A ). UM-SCC-1 cells were pretreated with celecoxib for 6 h before transfection of a synthetic 3x E2F-driven luciferase reporter plasmid and then incubated with the reporter construct overnight along with the drug. Celecoxib treatment produced a dose-dependent decrease in E2F transactivating activity that roughly paralleled the amount of hypophosphorylated Rb present on Western blots (Fig. 3C). At maximal dosing of 100 µmol/L celecoxib, E2F activity decreased by more than 50-fold compared with DMSO-treated controls.


Figure 5
View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Celecoxib inhibits E2F expression and activity. UM-SCC-1 cells were pretreated with indicated dose range of celecoxib for 6 h, followed by transfection with a 4:1 ratio of a 3x E2F promoter firefly luciferase reporter vector and control Renilla luciferase vector. Celecoxib was added to the media for an additional 14 h following luciferase construct transfections. Relative luciferase activity was determined by dual luciferase reporter assay. A, relative E2F transactivating activity for whole-cell UM-SCC-1 protein extracts following celecoxib treatment. Data are from a representative experiment; bars, SD from triplicate experimental replicates. *, P < 0.05. B, celecoxib inhibits nuclear E2F-1 protein expression. UM-SCC-1 and UM-SCC-17B cells were treated with either DMSO vehicle control (0.1%) or 100 µmol/L celecoxib for 24 h, and nuclear proteins were extracted as described in Materials and Methods. 20 µg of nuclear protein were separated by standard SDS-PAGE technique, transferred to nitrocellulose, and probed for E2F-1 expression. Equal protein loading between lanes was confirmed by Ponceau S staining of the nitrocellulose membrane. Data are from a representative experiment repeated at least in triplicate.

 
We further evaluated the nuclear expression of E2F-1 following treatment with celecoxib. UM-SCC-1 and UM-SCC-17B cells were treated with 0.1% DMSO or 100 µmol/L celecoxib for 24 h, followed by nuclear protein extraction and Western blotting for E2F-1 (Fig. 5B, inset). Baseline levels of nuclear E2F-1 expression were similar for both cell lines. Celecoxib produced a significant decrease (4–5-fold) in the amount of E2F-1 protein in the nucleus in both cell lines. Thus, celecoxib treatment reduced overall levels of E2F-1 protein within the nucleus and produced a profound decrease in E2F transactivating activity.

Celecoxib specifically kills S phase cells, which correlates with phase-specific alterations in p21 induction. Celecoxib induced apoptosis at doses where we also observed strong inhibition of E2F function and cell cycle rearrangement. To further evaluate the effect of these observed cell cycle rearrangements on celecoxib-induced cytotoxicity, we treated UM-SCC-1 cells with 0.1% DMSO and 50 or 100 µmol/L celecoxib for 24 h and then did vital cellular staining with 5 µmol/L Hoechst 33342, followed by cell cycle phase–specific cytometric sorting. This technique allowed for identification and collection of pure cell populations from the G1, S, and G2 phases (see Fig. 6D for representative sorting histograms). Cells from each phase were then plated at clonogenic density in the absence of drug and evaluated for growth potential by standard clonogenic survival assay techniques. Figure 6A shows representative clonogenic survival plates seeded at similar density and treated for 24 h with 100 µmol/L celecoxib before sorting and plating, whereas Fig. 6B shows overall experimental data presented in graphic form for celecoxib 50 and 100 µmol/L dosing compared with control plating efficiency. Treatment with 50 µmol/L celecoxib for 24 h before sorting did not cause significant changes in plating efficiency across all cell cycle phases, although a general trend was noted toward decreased plating efficiency in the S phase at this dose (Fig. 6B). In contrast, 24-h treatment with 100 µmol/L celecoxib produced significant changes in plating efficiency across all cell cycle phases, with the most profound toxicity in the S phase (Fig. 6A and B). G1 phase clonogenic survival was 43.2 ± 12.3% of control survival efficiency, a statistically significant decrease. G2 phase cells had even lower plating efficiency, with 24.1 ± 2.0% of control efficiency. In contrast, cells from the S phase had a plating efficiency of only 3.64 ± 2.6% of control, more than a 27-fold decrease in plating efficiency compared with control.


Figure 6
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6. Celecoxib induces specific clonogenic toxicity within the S and G2 phases. UM-SCC-1 cells were treated with DMSO control (0.1%) or celecoxib (50 and 100 µmol/L) for 24 h, followed by 30 min of incubation with Hoechst 33342. Cells were trypsinized and sorted by FACS DiVa for cell cycle distribution, allowing for identification and collection of pure G1, S, and G2 populations. These cells were plated at clonogenic density and allowed to incubate at 37°C for 2 wks, followed by ethanol fixation, crystal violet staining, and colony counting. Colonies with a minimum of 50 cells were counted, and all experiments were done in triplicate. A, representative clonogenic survival plates demonstrating decreased plating efficiency in the S phase. After treatment with 100 µmol/L celecoxib for 24 h and cell cycle sorting, UM-SCC-1 cells were seeded at equivalent density (2000 cells/60 mm3 plate) and were cultured for clonogenic survival. Representative clonogenic survival plates from G1, S, and G2 populations are presented. Note decreased colony count in S phase relative to G1 and G2. B, overall relative plating efficiencies following treatment for 24 h with 50 or 100 µmol/L celecoxib versus 0.1% DMSO control. Data show results from a representative experiment standardized to control plating efficiency (DMSO 0.1%) within each group. Bars, SD from triplicate experimental replicates. *, P < 0.05. C, Western blot analysis of G1 checkpoint control protein expression in pure G1, S, and G2 populations. Pure cell cycle phase populations were isolated as above, followed by protein extraction and separation by standard SDS-PAGE technique and probing for p21 and cyclin D1 expression. Blots were stripped and reprobed for {alpha}-tubulin to confirm even protein loading. Data shown are indicative of a representative experiment repeated at least in triplicate. D, representative histograms from cell cycle phase–specific sorting experiment demonstrating standard gating parameters for G1, S, and G2 populations that were subsequently used for experimental purposes. Gating parameters were drawn to isolate the purest possible population of cells within each cell cycle phase.

 
We further hypothesized that changes in p21 expression may affect this celecoxib-induced phase-specific toxicity due to effects of p21 expression on cell cycle transition out of the G1 phase. Using standard SDS-PAGE techniques on protein extracts from pure sorted cell cycle populations, we observed faint expression of p21 in the G1 and G2 phase of control cells (0.1% DMSO treated, Fig. 6C). Strong induction of p21 was seen in the G1 and G2 phase populations after 24 h of treatment with 50 and 100 µmol/L celecoxib, 6- and 8.5-fold, respectively, compared with controls. Expression of p21 in the S phase was barely detectable at both doses. Further blotting for cyclin D1 showed a general trend toward decreased cyclin D1 expression as doses of celecoxib decreased, especially within the S phase. These data show that p21 and cyclin D1 are relatively underexpressed in the S phase of celecoxib-treated cells and suggest that p21 and cyclin D1 expression may perhaps be protective against celecoxib-induced toxicity in the G1 and G2 cell cycle phases.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
We show for the first time in this study that celecoxib toxicity is targeted specifically to the S phase of the cell cycle. This is of considerable significance because celecoxib is currently being used in clinical trials with agents most active in the S and G2 phases, such as chemotherapeutic agents (docetaxel, ref. 26; cisplatin/5-fluorouracil, ref. 27) and radiation (28, 29). This demonstration of marked toxicity specifically limited to cells within the S phase following celecoxib administration provides a clear basis for additive or synergistic toxicity with agents that function specifically in the S and G2 phases. We also show that this specific toxicity is likely mediated at least in part through profound downstream inhibition of E2F-1 activity within the nucleus because celecoxib induces derangements of G1 checkpoint proteins and downstream sequestration of E2F by hypophosphorylated Rb. Without active nuclear E2F-1 function, crucial S phase activities such as DNA repair are impaired, leading to eventual cell death (30). Indeed, celecoxib has been previously shown to inhibit DNA repair enzymes such as Ku70 through the inhibition of nuclear factor {kappa}B, leading to radiosensitization (31). Of note, decreased expression of Ku70 is seen at doses of celecoxib where we routinely observe strong E2F inhibition, suggesting that this E2F inhibition may be involved in downstream inhibition of intracellular DNA repair activity. Further work is needed to clarify these connections between toxicity and E2F activity and how these relate to the observed G1 arrest following celecoxib treatment.

The mechanism for celecoxib-induced specific S phase toxicity may also be related to the relative inability of S phase cells to manage reactive oxygen stress. NSAID compounds are able to markedly increase the formation of intracellular reactive oxygen species (ROS; ref. 32). Celecoxib has previously been shown to increase the formation of ROS in osteosarcoma cells (33). G1 and G2 populations are able to perform silent repair of oxidative DNA lesions through the base excision repair pathway, whereas the S phase remains highly sensitive to perturbations of peroxide and other ROS (34). It is possible that the generation of ROS by celecoxib leads to pronounced toxicity to S phase cells through reactive oxygen toxicity and DNA damage, while leaving G1 and G2 phases relatively protected. Additional studies are needed to clarify the exact relationship between the E2F inhibition that we observe and S phase progression following celecoxib treatment and how this may specifically relate to DNA replication and repair.

We also show for the first time that levels of p21 and cyclin D1 change markedly across the cell cycle phases following treatment with celecoxib, correlating with observed levels of toxicity. In sorted cell cycle populations, p21 was strongly induced in the G1 phase and, to a lesser extent, the G2 phase, and this expression roughly correlates with clonogenic survival within these cell cycle phases. Cells that are overexpressing p21 will arrest in the G1 phase and are then inhibited from further cell cycle progression. We routinely observe significant G1 phase accumulation only at doses of celecoxib sufficient to induce p21 and inhibit E2F function, suggesting that p21 function is important in this G1 blockade. Our data show that it is transit through S phase that is most toxic to celecoxib-treated cells; therefore, p21 induction would offer a protective effect by prohibiting cells from progressing into S phase. One previous study has also shown a similar protective effect of p21 expression in NSAID toxicity. Jung et al. (35) showed that murine thymocytes deficient in p21 were more susceptible to apoptosis following sulindac administration, another NSAID compound. This protective component of p21 expression has also been observed in chemotherapeutic toxicity studies, where cells expressing p21 are inhibited from progressing through S phase and developing polyploidy (36). Conversely, it is also likely that overexpression of p21 is at least partly responsible for the inhibition of E2F function that we observe, which may contribute to S phase toxicity (31). In this way, p21 induction may have a dual role in celecoxib toxicity, protecting cells from toxicity by inducing G1 arrest, but inhibiting S phase E2F activity and thereby causing toxicity to cells progressing through S phase.

In conclusion, we show that celecoxib treatment produces marked toxicity in HNSCC cells through transcriptional induction of p21 and E2F-1 inhibition. This toxicity is specifically targeted to S phase cells, leading to a direct mechanism for synergistic toxicity with many common anticancer agents such as chemotherapeutics and radiation. This provides a rationale for the continued use of celecoxib in clinical cancer trials and dictates a basis for developing further combination therapies that exploit this facet of celecoxib-induced toxicity.


   Acknowledgments
 
Grant support: NIH KO8 CA104361 (D.K. Trask), NIH P01 CA66081 (F.E. Domann), and NIH R01 CA111365 (P.C. Goswami).

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.

We thank Justin Fishbaugh and Gene Hess of the University of Iowa Flow Cytometry facility for able experimental assistance and Kellie Bodeker-Goranson for assistance with manuscript preparation.

Received 10/13/06. Revised 1/ 2/07. Accepted 2/12/07.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Jemal A, Murray T, Samuels A, et al. Cancer statistics, 2003. CA Cancer J Clin 2003;53:5–26.[Abstract/Free Full Text]
  2. Lin DT, Subbaramaiah K, Shah JP, et al. Cyclooxygenase-2: a novel molecular target for the prevention and treatment of head and neck cancer. Head Neck 2002;24:792–9.[CrossRef][Medline]
  3. Dubois RN, Abramson SB, Crofford L, et al. Cyclooxygenase in biology and disease. FASEB J 1998;12:1063–73.[Abstract/Free Full Text]
  4. Aoki T, Nagakawa Y, Tsuchida A, et al. Expression of cyclooxygenase-2 and vascular endothelial growth factor in pancreatic tumors. Oncol Rep 2002;9:761–5.[Medline]
  5. Fujiwaki R, Iida K, Kanasaki H, et al. Cyclooxygenase-2 expression in endometrial cancer: correlation with microvessel count and expression of vascular endothelial growth factor and thymidine phosphorylase. Hum Pathol 2002;33:213–9.[CrossRef][Medline]
  6. Gupta RA, Dubois RN. Colorectal cancer prevention and treatment by inhibition of cyclooxygenase-2. Nat Rev Cancer 2001;1:11–21.[CrossRef][Medline]
  7. Kulkarni S, Rader JS, Zhang F, et al. Cyclooxygenase-2 is overexpressed in human cervical cancer. Clin Cancer Res 2001;7:429–34.[Abstract/Free Full Text]
  8. Kirkpatrick K, Ogunkolade W, Elkak A, et al. The mRNA expression of cyclo-oxygenase-2 (COX-2) and vascular endothelial growth factor (VEGF) in human breast cancer. Curr Med Res Opin 2002;18:237–41.[CrossRef][Medline]
  9. Chan G, Boyle JO, Yang EK, et al. Cyclooxygenase-2 expression is up-regulated in squamous cell carcinoma of the head and neck. Cancer Res 1999;59:991–4.[Abstract/Free Full Text]
  10. Steinbach G, Lynch PM, Phillips RK, et al. The effect of celecoxib, a cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med 2000;342:1946–52.[Abstract/Free Full Text]
  11. Arber N, Eagle CJ, Spicak J, et al. Celecoxib for the prevention of colorectal adenomatous polyps. N Engl J Med 2006;355:885–95.[Abstract/Free Full Text]
  12. Psaty BM, Potter JD. Risks and benefits of celecoxib to prevent recurrent adenomas. N Engl J Med 2006;355:950–2.[Free Full Text]
  13. Solomon SD, McMurray JJ, Pfeffer MA, et al. Cardiovascular risk associated with celecoxib in a clinical trial for colorectal adenoma prevention. N Engl J Med 2005;352:1071–80.[Abstract/Free Full Text]
  14. Grosch S, Tegeder I, Niederberger E, et al. COX-2 independent induction of cell cycle arrest and apoptosis in colon cancer cells by the selective COX-2 inhibitor celecoxib. FASEB J 2001;15:2742–4.[Free Full Text]
  15. Denkert C, Fursternberg A, Daniel P, et al. Induction of G0/G1 cell cycle arrest in ovarian carcinoma cells by the anti-inflammatory drug NS-398, but not by COX-2 specific RNA interference. Oncogene 2003;22:8653–61.[CrossRef][Medline]
  16. Patel MI, Subbaramaiah K, Du B, et al. Celecoxib inhibits prostate cancer growth: evidence of a cyclooxygenase-2–independent mechanism. Clin Cancer Res 2005;11:1999–2007.[Abstract/Free Full Text]
  17. Jaradat MS, Wongsud B, Phornchirasilp S, et al. Activation of peroxisome proliferator-activated receptor isoforms and inhibition of prostaglandin H(2) synthases by ibuprofen, naproxen, and indomethacin. Biochem Pharmacol 2001;62:1587–95.[CrossRef][Medline]
  18. Carmichael J, DeGraff WG, Gazdar AF, et al. Evaluation of a tetrazolium-based semiautomated colorimetric assay: assessment of chemosensitivity testing. Cancer Res 1987;47:936–42.[Abstract/Free Full Text]
  19. Bock JM, Pickart MA, Pink JJ, et al. Modulation of tumor cell proliferation and apoptosis by polyamine depletion in cells of head and neck squamous cell carcinomas. Radiat Res 1999;152:604–10.[Medline]
  20. Menon SG, Sarsour EH, Spitz DR, et al. Redox regulation of the G1 to S phase transition in the mouse embryo fibroblast cell cycle. Cancer Res 2003;63:2109–17.[Abstract/Free Full Text]
  21. Specht K, Richter T, Muller U, et al. Quantitative gene expression analysis in microdissected archival formalin-fixed and paraffin-embedded tumor tissue. Am J Pathol 2001;158:419–29.[Abstract/Free Full Text]
  22. Bradford CR, Zhu S, Ogawa H, et al. P53 mutation correlates with cisplatin sensitivity in head and neck squamous cell carcinoma lines. Head Neck 2003;25:654–61.[CrossRef][Medline]
  23. Wick M, Hurteau G, Dessev C, et al. Peroxisome proliferator–activated receptor-{gamma} is a target of nonsteroidal anti-inflammatory drugs mediating cyclooxygenase-independent inhibition of lung cancer cell growth. Mol Pharmacol 2002;62:1207–14.[Abstract/Free Full Text]
  24. Nikitakis NG, Hebert C, Lopes MA, et al. PPAR{gamma}-mediated antineoplastic effect of NSAID sulindac on human oral squamous carcinoma cells. Int J Cancer 2002;98:817–23.[CrossRef][Medline]
  25. Hauser U, Balz V, Carey TE, et al. Reliable detection of p53 aberrations in squamous cell carcinomas of the head and neck requires transcript analysis of the entire coding region. Head Neck 2002;24:868–73.[CrossRef][Medline]
  26. Nugent FW, Mertens WC, Graziano S, et al. Docetaxel and cyclooxygenase-2 inhibition with celecoxib for advanced non–small cell lung cancer progressing after platinum-based chemotherapy: a multicenter phase II trial. Lung Cancer 2005;48:267–73.[CrossRef][Medline]
  27. Govindan R, McLeod H, Mantravadi P, et al. Cisplatin, fluorouracil, celecoxib, and RT in resectable esophageal cancer: preliminary results. Oncology (Huntingt) 2004;18:18–21.
  28. Liao Z, Komaki R, Milas L, et al. A phase I clinical trial of thoracic radiotherapy and concurrent celecoxib for patients with unfavorable performance status inoperable/unresectable non–small cell lung cancer. Clin Cancer Res 2005;11:3342–8.[Abstract/Free Full Text]
  29. Cerchietti LC, Bonomi MR, Navigante AH, et al. Phase I/II study of selective cyclooxygenase-2 inhibitor celecoxib as a radiation sensitizer in patients with unresectable brain metastases. J Neurooncol 2005;71:73–81.[CrossRef][Medline]
  30. Cam H, Dynlacht BD. Emerging roles for E2F: beyond the G1/S transition and DNA replication. Cancer Cell 2003;3:311–6.[CrossRef][Medline]
  31. Raju U, Ariga H, Dittmann K, et al. Inhibition of DNA repair as a mechanism of enhanced radioresponse of head and neck carcinoma cells by a selective cyclooxygenase-2 inhibitor, celecoxib. Int J Radiat Oncol Biol Phys 2005;63:520–8.[CrossRef][Medline]
  32. Inoue A, Muranaka S, Fujita H, et al. Molecular mechanism of diclofenac-induced apoptosis of promyelocytic leukemia: dependency on reactive oxygen species, Akt, Bid, cytochrome and caspase pathway. Free Radic Biol Med 2004;37:1290–9. Erratum in: Free Radic Biol Med 2005;38:149.[CrossRef][Medline]
  33. Takahashi Y, Hayashi I, Tominari Y, et al. Sulindac sulfide is a noncompetitive {gamma}-secretase inhibitor that preferentially reduces abeta 42 generation. J Biol Chem 2003;278:18664–70.[Abstract/Free Full Text]
  34. Leroy C, Mann C, Marsolier M-C. Silent repair accounts for cell cycle specificity in the signaling of oxidative DNA lesions. EMBO J 2001;20:2896–906.[CrossRef][Medline]
  35. Jung B, Barbier V, Brickner H, et al. Mechanisms of sulindac-induced apoptosis and cell cycle arrest. Cancer Lett 2005;219:15–25.[CrossRef][Medline]
  36. Bulavin DV TN, Aksenov ND, Pospelov VA, Pospelova TV. Deregulation of p53/p21Cip1/Waf1 pathway contributes to polyploidy and apoptosis of E1A+cHa-ras transformed cells after {gamma}-irradiation. Oncogene 1999;18:5611–9.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Molecular Cancer TherapeuticsHome page
D. P. Kunte, R. K. Wali, J. L. Koetsier, and H. K. Roy
Antiproliferative effect of sulindac in colonic neoplasia prevention: role of COOH-terminal Src kinase
Mol. Cancer Ther., July 1, 2008; 7(7): 1797 - 1806.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bock, J. M.
Right arrow Articles by Trask, D. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bock, J. M.
Right arrow Articles by Trask, D. K.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online