Molecular Cancer Research AM No Date
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
Molecular Cancer Research 5, 435-442, May 1, 2007. doi: 10.1158/1541-7786.MCR-07-0010
© 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 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 Google Scholar
Google Scholar
Right arrow Articles by Stasinopoulos, I.
Right arrow Articles by Bhujwalla, Z. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Stasinopoulos, I.
Right arrow Articles by Bhujwalla, Z. M.

Angiogenesis, Metastasis, and the Cellular Microenvironment

Silencing of Cyclooxygenase-2 Inhibits Metastasis and Delays Tumor Onset of Poorly Differentiated Metastatic Breast Cancer Cells

Ioannis Stasinopoulos, David R. O'Brien, Flonne Wildes, Kristine Glunde and Zaver M. Bhujwalla

JHU ICMIC Program, The Russell H. Morgan Department of Radiology and Radiological Science, The Johns Hopkins University School of Medicine, Baltimore, Maryland

Requests for reprints: Zaver M. Bhujwalla, Department of Radiology, The Johns Hopkins University School of Medicine, Room 208C, Traylor Building, 720 Rutland Avenue, Baltimore, MD 21205. Phone: 410-955-9698; Fax: 410-614-1948. E-mail: zaver{at}mri.jhu.edu


   Abstract
Top
 Notes
 Abstract
Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
Cyclooxygenases (COX) are rate-limiting enzymes involved in the conversion of PLA2-mobilized arachidonic acid into prostaglandins and thromboxanes. COX-2 is a key mediator of inflammation during both physiologic and pathologic responses to endogenous stimuli and infectious agents. Its overexpression has been detected in different cancers, including that of the breast. Using RNA interference, we have reduced the expression of COX-2 in the highly malignant breast cancer cell line MDA-MB-231 below detectable levels in response to interleukin-1ß or 12-O-tetradecanoylphorbol-13-acetate treatment. Microarray analysis showed that COX-2 silencing resulted in the loss of mRNA expression of several oncogenic markers, such as matrix metalloproteinase-1, chemokine (C-X-C motif) receptor 4, and interleukin-11, which have been correlated with poor disease outcome, and in the up-regulation of antimetastatic transcripts, such as thrombospondin-1 and Epstein-Barr-Induced 3. Cells lacking COX-2 were less able to invade reconstituted extracellular matrix than parental cells in vitro. Consistent with these changes, loss of COX-2 resulted in the abolition or the significant delay of tumor onset when the cells were injected in the mammary fat pad of severe combined immunodeficient mice. Finally, silencing of COX-2 resulted in the inhibition of metastasis to the lungs of severe combined immunodeficient mice after intravenous injection. These data show that silencing of COX-2 abolishes the metastatic potential of MDA-MB-231 cells in vivo. (Mol Cancer Res 2007;5(5):435–42)


   Introduction
Top
 Notes
 Abstract
 Introduction
Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
Cyclooxygenase (COX)-1 and COX-2 are cytoplasmic enzymes involved in the conversion of PLA2-mobilized arachidonic acid into lipid signal transduction molecules known as prostaglandins and thromboxanes. The major product of the COX-2–catalyzed reaction is prostaglandin E2 (PGE2), an inflammatory mediator participating in many biological processes, including development, pain, immunity, and angiogenesis (1). COX-2 function has been the target of pharmaceutical intervention in a multitude of widespread degenerating conditions, including autoimmune diseases, gastric inflammation, and several different cancers, such as gastric, lung, breast, and colon cancer (2–7). Its expression is induced by proinflammatory cytokines, such as interleukin (IL)-1ß and tumor necrosis factor (TNF)-{alpha}, and its promoter contains a cyclic AMP response element, a nuclear factor-{kappa}B binding site, and two nuclear factors for IL-6 target sequences (8). Extensive expression of COX-2 has been observed in colon cancer (9) and in 37% to 43% of breast tumors (10, 11). Mechanisms responsible for the enhancement of tumor progression by COX-2 are thought to involve the promotion of invasion and angiogenesis (12, 13), processes necessary for tumorigenesis and metastasis. COX-2 overexpression in the mammary glands of transgenic mice resulted in increased tumor formation (14). Studies on the possible role of COX-2 in cancer have predominantly been conducted using COX-2–selective inhibitors. Although the overall data arguing for the involvement of COX-2 in some human cancers are compelling, the possible COX-2–independent functions of COX-2–selective inhibitors have recently come under intense scrutiny (15). To our knowledge, this is the first study reporting that nonpharmacologically mediated loss of COX-2 delays tumor onset and inhibits metastasis formation in a metastatic human breast cancer xenograft.


   Results
Top
 Notes
 Abstract
 Introduction
 Results
Discussion
 Materials and Methods
 Acknowledgements
 References
 
Generation of MDA-MB-231 Cells Lacking COX-2
We used RNA interference to study the effects of loss of COX-2 in the estrogen receptor/progesterone receptor–negative, poorly differentiated, malignant human breast cancer cell line MDA-MB-231. COX-2 protein was induced in response to IL-1ß treatment in a dose-dependent manner in MDA-MB-231 cells with maximum stimulation observed at 10 ng/mL (Fig. 1A ). Transient transfection experiments using COX-2 small interfering RNA (siRNA) showed specific inhibition of COX-2 expression in response to IL-1ß (Fig. 1A, lane 6). Additionally, a second COX-2 siRNA sequence described previously (16) also showed specific COX-2 down-regulation (Fig. 1A, lane 7). In contrast, a glyceraldehyde-3-phosphate dehydrogenase–specific siRNA duplex had no effect on the expression of COX-2 (Fig. 1A, lane 8). Stable clones expressing a short hairpin RNA complementary to COX-2 mRNA were obtained as described previously (17). To assess the extent of the loss of COX-2 expression following IL-1ß stimulation, immunoassays for PGE2 were done on supernatants from individual clones (data not shown). Clone 2, which could not be induced to secrete PGE2, was selected for further experimentation. To avoid artifacts resulting from the use of single clones of transfected cells, we decided to simultaneously use a pool of four individual clones, which could also not be induced to secrete PGE2. Immunoblotting revealed no detectable COX-2 protein in the clone 2 and pooled clone cell lines induced with IL-1ß or 12-O-tetradecanoylphorbol-13-acetate (a known mitogen) for 24 h, in contrast to MDA-MB-231 and empty vector cells, which responded to identical treatment (Fig. 1B). IL-1ß stimulation provided the most stable means to evaluate the ability of cells to express COX-2, which is consistent with the fact that COX-2 expression in vivo is a result of induction by proinflammatory cytokines, such as IL-1ß, TNF-{alpha}, and interferon (IFN)-{gamma}. Although it has been observed previously that MDA-MB-231 cells express high background levels of COX-2 (18), in our experience, background COX-2 expression, as well as associated PGE2 production in MDA-MB-231 and empty vector cells, varied significantly. Real-time PCR did not reveal a specific COX-2 product in the clone 2 and pooled clone cell lines, in contrast with parental cells, which amplified a specific product at 23.5 or 26.4 cycles when 50 or 5 ng total mRNA template was used (data not shown).


Figure 1
View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 1. Generation of MDA-MB-231 stable clones lacking COX-2. A. MDA-MB-231 (231) cells were induced with varying concentrations of IL-1ß for 24 h to ensure that COX-2 was inducible. Two COX-2–specific siRNA molecules (-2, -2nv), one designed by us and the other described previously (16), severely down-regulated COX-2 expression in response to IL-1ß. B. COX-2 short hairpin RNA–transfected clones were obtained using G418 as a selection marker and following standard protocols. One individual clone (c2) and four pooled clones (P) were induced with IL-1ß (10 ng/mL) or 12-O-tetradecanoylphorbol-13-acetate (TPA; 50 nmol/L V, empty vector cells). Membranes were probed with a COX-2–specific antibody and with a glyceraldehyde-3-phosphate dehydrogenase–specific antibody (GAPDH).

 
Loss of COX-2 Does Not Alter Cell Growth Rate In vitro
To measure cell growth, cells (1.5 x 105) were plated on 35-mm dishes and trypsinized and counted at 24-, 48-, and 72-h time intervals. No significant differences in growth rate were seen between cells inducible for COX-2 (MDA-MB-231 and empty vector cells) and cells lacking COX-2 (clone 2 and pooled clones) as shown in Fig. 2 , suggesting that COX-2 does not affect MDA-MB-231 doubling time in culture. Additional experiments showed that loss of COX-2 did not significantly affect cell size (data not shown).


Figure 2
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 2. Cells lacking COX-2 have similar growth characteristics with parental MDA-MB-231 cells. Cells (1.5 x 105) were seeded in 35-mm dishes and cells were trypsinized and counted 24, 48, and 72 h later. MDA-MB-231 ({square}), empty vector cells (Figure 2), clone 2 (Figure 2), and pooled clones ({blacksquare}).

 
Cells Lacking COX-2 Have a Less Oncogenic Transcriptome
PGE2 may promote cell growth by interfering with the ß-catenin and Ras pathways (19, 20). We thus hypothesized that loss of COX-2 may result in alteration of the expression of factors responsible for many cellular functions. To assess differences between the transcriptomes of IL-1ß–challenged COX-2–lacking and parental MDA-MB-231 cells, we did GeneChip expression arrays using the Affymetrix Human Genome U133 Plus 2.0 GeneChip. We sought to identify candidate genes whose expression was suppressed or increased in both the clone 2 and the pooled clone cells relative to parental MDA-MB-231 cells from two separate experiments. Results show that COX-2–lacking cells display altered expression of an array of genes coding for factors regulating gene expression, invasion, and metastasis as well as tumor markers, including matrix metalloproteinase-1 (MMP-1), chemokine (C-X-C motif) receptor 4 (CXCR4), SMA- and MAD-related protein 1 (SMAD1), IL-11, and thrombospondin-1 (THBS-1; Fig. 3A ). Cells not induced with IL-1ß also showed similar differences in transcript levels (Fig. 3B) with the exception of S100P, HAS2, TPM1, ROBO4, and LUM, which were not differentially expressed in pooled clone cells compared with the parental MDA-MB-231 cells to a significant extent. Only transcripts that were modified in both the single clone as well as the pooled clones are shown, as we hypothesized that transcripts responsible for a common phenotypic change would be modified in both the clone 2 and the pooled clone cells. To validate the results of the microarray data, quantitative PCR was done to probe RNA purified from MDA-MB-231– and COX-2–lacking cells clone 2 and pooled clones challenged with IL-1ß (10 ng/mL) and probed for COX-2, CXCR4, JAG1, and SMAD1. No specific COX-2 or CXCR4 product was detected in the COX-2–silenced cells, which also displayed reduced levels of JAG1 (MDA-MB-231 versus clone 2, 3.9-fold; MDA-MB-231 versus pooled clones, 3.3-fold) and SMAD1 (MDA-MB-231 versus clone 2, 5.8-fold; MDA-MB-231 versus pooled clones, 3.2-fold). The values presented represent the mean from two independent experiments. These data validate the observations of the microarray analyses for COX-2, CXCR4, JAG1, and SMAD1.


Figure 3
View larger version (13K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 3. Loss of COX-2 alters the transcriptomes of MDA-MB-231 cells and reduces invasion. A. MDA-MB-231, clone 2, and pooled clone cells were induced with IL-1ß for 24 h and total mRNA was subjected to hybridization on an U133 Plus 2.0 Affymetrix microarray chip. All represented transcripts were regulated in a similar manner in two independent experiments. Down-regulated or up-regulated transcripts are shown with negative or positive values, respectively. White columns and black columns, the differential expression of transcripts in clone 2 and pooled clone cells compared with MDA-MB-231 cells, respectively. PTGS2*, Affymetrix probe ID 1554997_a_at (National Center for Biotechnology Information public ID AY151286), which corresponds to the alternatively spliced COX-2b protein. Expression signals were obtained by Robust Multiarray Analysis (21). The posterior probabilities of the differential expression of genes between the treatments were estimated with an empirical Bayses method using Gamma-Gamma modeling by the bioconductor R package, EBarrays. The criterion of the posterior probability >0.5 was used to produce the differentially expressed gene lists shown. B. RNA from MDA-MB-231 and COX-2–silenced clone 2 and pooled clone cells, which were not challenged with IL-1ß, was purified and subjected to microarray analyses as above. The criterion of the posterior probability >0.5 was used to produce the differentially expressed gene lists shown. C. Cells lacking COX-2 show significantly decreased invasion in vitro. MDA-MB-231 cells, empty vector cells, clone 2 cells, and pooled clone cells (1.5 x 105) were plated onto a Biocoat Tumor Invasion System and allowed to invade the Matrigel for 24 h. The difference in serum concentration between the top and the bottom of the plate was only 2-fold (5-10%) to avoid any deleterious effects to the cells arising from serum starvation. Results are representative of six individual experiments done in triplicates for each cell line. Columns, relative invasion index; bars, SE. Statistical significance was assessed by doing a homoscedastic, two-tailed Student's t test. P = 0.856205, empty vector cells; P = 0.006042, clone 2; P = 0.000026, pooled clones.

 
Loss of COX-2 Reduces Invasion In vitro
MMPs are broad specificity enzymes that cancer cells use to degrade extracellular matrix. Because MMP-1 mRNA levels were significantly reduced in both clone 2 and pooled clone cells in two independent experiments, we examined the invasive potential of the cells lacking COX-2 using Matrigel-based invasion assays. Parental MDA-MB-231 cells, empty vector cells, clone 2 cells, and pooled clone cells (1.5 x 105) were plated on Matrigel-containing Transwell plates and allowed to invade for 24 h. COX-2–lacking cells (clone 2 and pooled clones) showed a significant reduction in invasion compared with MDA-MB-231 and empty vector cells that can be induced to express COX-2 (Fig. 3C). These results agree with a recent publication that suggested that nuclear factor for activated T cells can induce COX-2 expression, which correlated with increased invasiveness (12).

Loss of COX-2 Attenuates Tumor Onset
To assess the effect of loss of COX-2 on tumor onset in vivo, 2 x 106 of MDA-MB-231 cells (n = 10), empty vector cells (n = 9), clone 2 cells (n = 11), and pooled clone cells (n = 10), suspended in 50 µL HBSS, were inoculated in the mammary fat pad of severe combined immunodeficient (SCID) mice in at least two separate experiments. Inoculation of parental MDA-MB-231 and empty vector cells resulted in palpable tumors within 6 weeks after injection (Table 1 ). In contrast, of the 11 mice injected with clone 2 cells, only 1 mouse developed a palpable tumor within the first 10 weeks after injection. Similarly, of the 10 mice injected with pooled clone cells, 4 developed small but palpable tumors within the first 10 weeks after injection (summarized in Table 1). Average tumor volumes reflecting the differences in tumor onset are shown in Fig. 4A . The tumor from the mouse injected with clone 2 cells was excised and cells were isolated and cultured. To ascertain whether the clone 2 tumor-derived cells (c2T1205) had lost the siRNA-mediated COX-2 inhibition, immunoblotting was done. c2T1205 cells regained their ability to express COX-2 in the absence of, or in response to, exogenous IL-1ß to a significant extent (Fig. 4B), suggesting that COX-2 is required for tumor initiation in vivo and that loss of COX-2 is sufficient for abolishing the tumorigenicity of MDA-MB-231 cells in vivo.


View this table:
[in this window]
[in a new window]

  		Table 1. COX-2–Silenced MDA-MB-231 Cells Form Delayed, if Any, Tumors When Injected in the Mammary Fat Pad of SCID Mice

 

Figure 4
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 4. Cells lacking COX-2 have delayed and limited tumor onset in vivo. A. MDA-MB-231 cells ({diamondsuit}), empty vector cells (•), clone 2 cells ({blacktriangleup}), and pooled clone cells ({blacksquare}), 2 x 106/50 µL HBSS, were injected into the mammary fat pad of 10, 9, 11, and 10 SCID mice, respectively. Tumor size was measured before sacrificing the mice. Points, tumor volumes for each cell line; bars, SE. B. Tumor formation correlates with COX-2 expression. MDA-MB-231, clone 2, and cells derived from clone 2 tumor (T1205) were induced with IL-1ß (10 ng/mL) or TPA (50 nM) as described previously.

 
Loss of COX-2 Inhibits Extrapulmonary Metastasis
To examine the effect of loss of COX-2 expression on experimental metastasis from parental MDA-MB-231 cells (n = 5), empty vector cells (n = 5), clone 2 cells (n = 5), and pooled clone cells (n = 3), 2 x 106 cells resuspended in 50 µL HBSS were injected in the tail vein of SCID mice. Mice were sacrificed 5 weeks after injection and their lungs were infused with agarose to prevent collapse. A stark contrast in metastatic nodules was apparent both visually as well as in H&E-stained sections. Mice injected with parental MDA-MB-231 and empty vector cells had multiple visible metastatic nodules in the lungs, in contrast to COX-2–lacking cell lines, clone 2 and pooled clones, for which lungs appeared clear of any metastasis (Fig. 5A ). H&E-stained lung sections showing multiple metastatic lesions from parental and empty vector cells, but none from COX-2–lacking cells, are shown in Fig. 5B. Images of H&E-stained lung sections representing the cross-sectional area from each pair of lungs were analyzed to determine the percentage area covered by metastases using ImageJ (NIH). The percentage areas (mean ± SE) were 4.2 ± 0.7, 24.9 ± 3.7, 0, and 0, for MDA-MB-231 cells (n = 5), empty vector cells (n = 5), clone 2 cells (n = 5), and pooled clone cells (n = 3), respectively. The two-tailed nonparametric Mann-Whitney U test was used to determine if there was a significant difference in the percentage areas between the cell lines using {alpha} of 0.05. A significant difference in the percentage area of metastases was evident between MDA-MB-231 and clone 2 (P = 0.007) or pooled clones (P = 0.032) and between empty vector cells and clone 2 (P = 0.007) or pooled clones (P = 0.032). These data suggest that COX-2 function is critical for metastasis of MDA-MB-231 cells in vivo.


Figure 5
View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 5. Cells lacking COX-2 fail to metastasize to the lungs of SCID mice. A. Representative photographs of whole, agarose-infused lungs from mice injected with MDA-MB-231 cells, empty vector cells (V), clone 2 cells, and pooled clone cells (2 x 106/50 µL HBSS) in the tail vein of 5, 5, 5, and 3 SCID mice. Lungs were excised 5 wks after injection. B. Photographs of lung sections stained with H&E using standard protocols. Representative photographs from sections of lungs obtained from two different mice per cell line taken using a Nikon inverted microscope and a 4x objective lens.

 

   Discussion
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
Materials and Methods
 Acknowledgements
 References
 
Although the association of persistent inflammation and increased cancer risk has long been established, the mechanisms underlying it remain to be delineated. COX-2 is thought to promote tumorigenesis by the actions of lipid hydroperoxide formation and through the effects of its major product, PGE2. COX-2 lipid peroxidation products have been shown to result in the formation of DNA adducts (22), whereas PGE2 has been shown to affect a multitude of pathways promoting invasion and angiogenesis (19, 20), including the induction of hypoxia-inducible transcription factor (23), a potent signaling molecule responsible for angiogenesis. Lewis lung carcinoma–derived tumors growing in COX-2 knockout mice had decreased vascular density and a significant decrease in the production of VEGF, suggesting that the absence of COX-2 from the tumor microenvironment affects tumor vasculature (24).

Deciphering the mechanism by which COX-2 promotes tumorigenesis is complicated by the fact that COX-2–selective inhibitors seem to also inhibit carcinogenic signaling independent of COX-2 at the concentrations used to study cell signaling in vitro (15). Additionally, COX-2–dependent pathways can vary depending on the tissue and the nature of the tissue microenvironment studied. For example, it was reported previously that celecoxib inhibits TNF-{alpha}–induced extracellular signal-regulated kinase phosphorylation in the non–small cell lung carcinoma cell line H1299 (25). However, in our cell lines that lack COX-2 expression, TNF-{alpha} (10 ng/mL) successfully induced mitogen-activated protein/extracellular signal-regulated kinase kinase phosphorylation within 45 min.1

To investigate the effect of COX-2 in breast cancer, we decided to silence COX-2 expression using RNA interference in an effort to circumvent some of the nonspecific and transient effects associated with pharmacologic inhibitors. We have shown that loss of COX-2 in the metastatic human breast xenograft model MDA-MB-231 delayed or inhibited orthotopic tumor onset and inhibited extrapulmonary metastasis in a model of experimental metastasis. Interestingly, COX-2–silenced cells did not show any differences in growth rate in vitro. This is likely a result of the complex interactions required for tumor onset and growth in vivo in contrast to culture conditions. Such complex interactions may include alteration or loss of chemotactic signaling or other modifications of the tissue microenvironment that can only take place in an in vivo model of tumorigenesis. Cells depleted of COX-2 express messages that code for the production of tumor endothelial markers, such as THBS-1, which is thought to contribute to an environment less permissive to the establishment of metastasis (26), tropomyosin-1, which sensitizes breast cancer cells to apoptosis (27), IL-27 (EBI-3), which generates a Th1 immune response (28), and CSR1 (SCARA3), an oxygen scavenger down-regulated in prostate cancer (29). Additionally, MDA-MB-231 cells lacking COX-2 showed reduced mRNA expression of several oncogenic markers, including IL-11, a marker for metastasis of breast cancer to bone (30), the Notch1 receptor ligand JAG1, whose expression is correlated with poor breast cancer prognosis (31), CXCR4, a receptor involved in the regulation of the tumor microenvironment (32), and MMP-1, a secreted enzyme responsible for the degradation of the stroma during breast cancer cell invasion (33).

Our results indicate that COX-2 expression modulates the expression of many extracellular matrix components, including collagen, glycoproteins (e.g., THBS-1), hyaluronan, and proteoglycans (e.g., lumican). It is therefore possible that tumor-derived COX-2 modifies the extracellular matrix enabling tumors to successfully establish metastases. Modification of the extracellular matrix by microenvironment-derived COX-2 could also explain why tissues with active inflammatory processes, such as wounds, are sites of frequently successful metastases. The notion that sites of inflammation are highly permissive to successful metastasis was experimentally proven when Rous sarcoma virus–induced tumor metastasis was shown to be dependent on wound-induced inflammation (34). Further experimentation is required to prove the hypothesis that tumor- and microenvironment-derived COX-2 and the products of the COX reaction contribute significantly to the successful establishment of metastasis. With pharmacologic inhibition of COX-2 in vivo, it is difficult to determine the contribution of COX-2 secreted by stromal cells to tumor growth and metastasis. Here, we have shown that silencing COX-2 selectively in cancer cells, without directly altering COX-2 expression in stromal cells, has a profound effect on tumor onset and metastasis. Another critical difference between the use of pharmacologic agents and our model is that using RNA interference, we have reduced the level of COX-2 beyond the limit of detection. This extent of inhibition may not be easily achieved without associated side effects using pharmacologic agents. Silencing COX-2 expression selectively in tumor cells using cancer cell–targeted delivery of siRNA may minimize the side effects of COX-2 inhibition while achieving effective treatment of the tumor.

Taken together, our data indicate that COX-2 is important for tumorigenesis in the mammary fat pad in vivo. Additionally, COX-2 has a critical role in the establishment of metastasis in the lungs, a primary metastatic target of human breast cancers and associated with a large percentage of breast cancer–associated mortality. Microarray data suggest that loss of COX-2 results in the reduction of the expression of proteins associated with poor breast cancer outcome and the up-regulation of angiostatic and antimetastatic transcripts. These data strongly suggest that at least in some cancers, COX-2 is a key mediator of the signaling interplay between the invading cancer and the tumor microenvironment, an event critical to tumor metastasis.

The effects of the COX-2–selective inhibitor celecoxib are currently studied in many clinical trials alone or in combination with chemotherapy and radiation.

Of particular interest is the clinical trial designed to test the effect of celecoxib in estrogen receptor–negative premenopausal women at high risk of breast cancer (National Cancer Institute identifier: NCT00056082). Our results suggest that COX-2 inhibition may be of particular clinical importance in poorly differentiated breast cancer tumors.


   Materials and Methods
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
Acknowledgements
 References
 
Derivation of a COX-2–Specific siRNA Target Sequence
The target siRNA sequence was determined by the online, freely available siRNA target finder search engine offered by Ambion2 and was synthesized by Dharmacon against the following COX-2 mRNA target sequence: AACAUUCCCUUCCUUCGAAAU.

Stable Expression of the COX-2 Short Hairpin RNA–Containing Plasmid in MDA-MB-231 Cells
The COX-2 short hairpin RNA–coding plasmid was constructed and placed under the control of the U6 promoter as described previously by Glunde et al. (17). Cells were transfected with COX-2–specific short hairpin RNA–bearing plasmid using LipofectAMINE 2000 as suggested by the manufacturer (Invitrogen). Individual clones were selected for G418 resistance using standard protocols. Several individual clones were analyzed for PGE2 production from the supernatant using the PGE2 EIA Kit-Monoclonal as described by the manufacturer (Cayman Chemical; data not shown). Parental MDA-MB-231 cells, empty vector cells (17), clone 2 cells, and cells pooled from four individual clones were selected for further experimentation.

Cell Culture and Immunoblotting
All cells were cultured with RPMI 1640 (Sigma-Aldrich) supplemented with 8.25% fetal bovine serum (Sigma-Aldrich) without penicillin or streptomycin. To establish single clones, cells were selected for G418 resistance (400 µg/mL; Invitrogen) according to standard protocols. Established COX-2–lacking clones were not further selected for G418 resistance; instead, cells were assayed functionally for loss of COX-2. MDA-MB-231, empty vector cells, clone 2 and pooled clone cells, and c2T1205 cells were plated in 100-mm dishes (1.5 x 106) and induced with IL-1ß (10 ng/mL) or 12-O-tetradecanoylphorbol-13-acetate (50 nmol/L) for all experiments involving immunoblotting for COX-2. Twenty-four hours later, cells were lysed using M-PER (Pierce) and a protease inhibitor cocktail (Sigma-Aldrich). Crude cell extracts were quantified using DC protein assay (Bio-Rad). Fifty micrograms of protein were subjected to SDS-PAGE and immunoblotting following standard protocols. Membranes were probed overnight with a COX-2–specific antibody (Oxford Biomedical) and with a glyceraldehyde-3-phosphate dehydrogenase–specific antibody (Santa Cruz Biotechnology), washed, and probed with antigoat horseradish peroxidase–labeled secondary antibodies (Jackson ImmunoResearch) following standard protocols. Substrate (SuperSignal West Pico, Pierce) was added to the probed membranes and films were developed using an automatic film processor (Kodak and Agfa).

Biocoat Tumor Invasion Assay
Cells (1.5 x 105) were plated on the top part of a rehydrated, Matrigel-containing Biocoat Tumor Invasion System (Becton Dickinson) and allowed to invade for 24 h. Cells were labeled using calcein AM (Invitrogen) for 1 h as directed by the manufacturer. The difference in serum concentration between the top and the bottom of the plate was only 2-fold (5-10%) to avoid any deleterious effects to the cells arising from serum starvation. Fluorescence measurements were taken 1 h after labeling using a Victor plate reader set to read the bottom of the plate at 485/535 nm.

Cell Preparation for Injection into the Murine Fat Pad and Tail Vein
Cells were cultured as described above and trypsinized using standard protocols. The cells were then centrifuged and adjusted to a cell density of 2 x 106/50 µL HBSS (Sigma-Aldrich) without Ca2+ or Mg2+. The cells were then injected in the mammary fat pad or the mouse tail vein of anesthetized 5- or 6-week-old female SCID mice (National Cancer Institute) and mice were returned to their cages for further examination.

Mice Work Committee Approval
The authors of this publication state that all mice studies described were approved by the Animal Care and Use Committee and that all requirements for humane animal treatment dictated by the above committee were met.

Microarray Analysis
Cells used in the microarray experiment described in Fig. 3A and B were seeded (1 x 106) in 100-mm dishes and allowed to attach to the plate overnight. Cells were induced with IL-1ß (10 ng/mL) for 24 h. RNA was purified using the RNeasy kit (Qiagen) as described by the manufacturer and the on-column DNase I digestion done as described by the manufacturer. Samples were subjected to agarose formaldehyde electrophoresis to determine the quality of the RNA by comparing the ratios of 18 S and 28 S RNA species. All samples were run in commercial arrays from Affymetrix, using Human Genome U133 Plus 2.0 GeneChip arrays as described in the Affymetrix Web site.3 Probe hybridization and analysis was run as suggested by the manufacturer in the aforementioned Web site. These descriptions include all information currently considered under the Minimum Information About a Microarray Experiment supportive guidelines, with which the JHMI Microarray Core Facility abides in all its procedures. The authors will reposit the Minimum Information About a Microarray Experiment–compliant original data in the Gene Expression Omnibus database immediately following the acceptance of the manuscript.

Quantitative Real-time RT-PCR
Cells were induced with IL-1ß (10 ng/mL) for 24 h and RNA was purified as above. Twenty nanograms of purified RNA were subjected to One-Step Real-time RT-PCR using SYBR Green (Qiagen) in a Bio-Rad iCycler as described by the manufacturer. Primers for COX-2, CXCR4, JAG1, and SMAD1 studies were obtained from Qiagen. Primer sequences for 18 S transcript used as a control: 5'-GGTTGATCCTGCCAGTAGC-3' and 5'-GCGACCAAAGGAACCATAAC-3'. Fold differences were calculated using the 2{Delta}{Delta}Ct method.


   Acknowledgements
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
References
 
We thank Dr. Venu Raman for useful discussions and advice about the characterization of the cell lines, Dr. Arvind Pathak for statistical analyses of the lung data, and Tomoyo Takagi and Gary Cromwell for their assistance with the animal studies.


   Notes
Top
 Notes
Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
Grant support: NIH research grants 2R01CA82337 and P50 CA103175.

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.

Note: Microarray data will be submitted to the Gene Expression Omnibus database pending acceptance of the manuscript.

1 Unpublished observation. Back

2 http://ambion.com/techlib/misc/siRNA_finder.html Back

3 http://www.affymetrix.com Back

Received 1/ 9/07; revised 3/ 1/07; accepted 3/ 9/07.


   References
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 

  1. Smith WL, DeWitt DL, Garavito RM. Cyclooxygenases: structural, cellular, and molecular biology. Annu Rev Biochem 2000;69:145–82.[CrossRef][Medline]
  2. Xu L, Zhang L, Yi Y, Kang HK, Datta SK. Human lupus T cells resist inactivation and escape death by upregulating COX-2. Nat Med 2004;10:411–5.[CrossRef][Medline]
  3. Wallace JL. Recent advances in gastric ulcer therapeutics. Curr Opin Pharmacol 2005;5:573–7.[CrossRef][Medline]
  4. Wang D, Mann JR, DuBois RN. The role of prostaglandins and other eicosanoids in the gastrointestinal tract. Gastroenterology 2005;128:1445–61.[CrossRef][Medline]
  5. Krysan K, Reckamp KL, Sharma S, Dubinett SM. The potential and rationale for COX-2 inhibitors in lung cancer. Anticancer Agents Med Chem 2006;6:209–20.[Medline]
  6. Diaz-Cruz ES, Brueggemeier RW. Interrelationships between cyclooxygenases and aromatase: unraveling the relevance of cyclooxygenase inhibitors in breast cancer. Anticancer Agents Med Chem 2006;6:221–32.[Medline]
  7. Koehne CH, Dubois RN. COX-2 inhibition and colorectal cancer. Semin Oncol 2004;31 Suppl 7:12–21.[Medline]
  8. Chun KS, Surh YJ. Signal transduction pathways regulating cyclooxygenase-2 expression: potential molecular targets for chemoprevention. Biochem Pharmacol 2004;68:1089–100.[CrossRef][Medline]
  9. Eberhart CE, Coffey RJ, Radhika A, Giardiello FM, Ferrenbach S, DuBois RN. Up-regulation of cyclooxygenase 2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology 1994;107:1183–8.[Medline]
  10. Ristimaki A, Sivula A, Lundin J, Reitsma JB, Haglund C. Bosma PJ et al. Prognostic significance of elevated cyclooxygenase-2 expression in breast cancer. Cancer Res 2002;62:632–5.[Abstract/Free Full Text]
  11. Half E, Tang XM, Gwyn K, Sahin A, Wathen K, Sinicrope FA. Cyclooxygenase-2 expression in human breast cancers and adjacent ductal carcinoma in situ. Cancer Res 2002;62:1676–81.[Abstract/Free Full Text]
  12. Yiu GK, Toker A. NFAT induces breast cancer cell invasion by promoting the induction of cyclooxygenase-2. J Biol Chem 2006;281:12210–7.[Abstract/Free Full Text]
  13. Gately S, Li WW. Multiple roles of COX-2 in tumor angiogenesis: a target for antiangiogenic therapy. Semin Oncol 2004;31 Suppl 7:2–11.[Medline]
  14. Liu CH, Chang SH, Narko K, et al. Overexpression of cyclooxygenase-2 is sufficient to induce tumorigenesis in transgenic mice. J Biol Chem 2001;276:18563–9.[Abstract/Free Full Text]
  15. Grosch S, Maier TJ, Schiffmann S, Geisslinger G. Cyclooxygenase-2 (COX-2)-independent anticarcinogenic effects of selective COX-2 inhibitors. J Natl Cancer Inst 2006;98:736–47.[Abstract/Free Full Text]
  16. Denkert C, Furstenberg A, Daniel PT, 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]
  17. Glunde K, Raman V, Mori N, Bhujwalla ZM. RNA interference-mediated choline kinase suppression in breast cancer cells induces differentiation and reduces proliferation. Cancer Res 2005;65:11034–43.[Abstract/Free Full Text]
  18. Basu GD, Pathangey LB, Tinder TL, Gendler SJ, Mukherjee P. Mechanisms underlying the growth inhibitory effects of the cyclo-oxygenase-2 inhibitor celecoxib in human breast cancer cells. Breast Cancer Res 2005;7:R422–35.[CrossRef][Medline]
  19. Castellone MD, Teramoto H, Williams BO, Druey KM, Gutkind JS. Prostaglandin E2 promotes colon cancer cell growth through a Gs-axin-ß-catenin signaling axis. Science 2005;310:1504–10.[Abstract/Free Full Text]
  20. Krysan K, Reckamp KL, Dalwadi H, et al. Prostaglandin E2 activates mitogen-activated protein kinase/Erk pathway signaling and cell proliferation in non-small cell lung cancer cells in an epidermal growth factor receptor-independent manner. Cancer Res 2005;65:6275–81.[Abstract/Free Full Text]
  21. Irizarry RA, Hobbs B, Collin F, et al. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 2003;4:249–64.[Abstract]
  22. Lee SW, Williams MV, DuBois RN, Blair IA. Cyclooxygenase-2-mediated DNA damage. J Biol Chem 2005;31:28337–46.
  23. Liu XH, Kirschenbaum A, Lu M, et al. Prostaglandin E2 induces hypoxia-inducible factor-1{alpha} stabilization and nuclear localization in a human prostate cancer cell line. J Biol Chem 2002;277:50081–6.[Abstract/Free Full Text]
  24. Williams CS, Tsujii M, Reese J, Dey SK, DuBois RN. Host cyclooxygenase-2 modulates carcinoma growth. J Clin Investig 2000;105:1589–94.[Medline]
  25. Shishodia S, Koul D, Aggarwal BB. Cyclooxygenase (COX)-2 inhibitor celecoxib abrogates TNF-induced NF-{kappa}B activation through inhibition of activation of I{kappa}B{alpha} kinase and Akt in human non-small cell lung carcinoma: correlation with suppression of COX-2 synthesis. J Immunol 2004;173:2011–22.[Abstract/Free Full Text]
  26. Guo N, Krutzsch HC, Inman JK, Roberts, DD. Thrombospondin 1 and type I repeat peptides of thrombospondin 1 specifically induce apoptosis of endothelial cells. Cancer Res 1997;57:1735–42.[Abstract/Free Full Text]
  27. Raval G-N, Bharadwaj S, Levine EA, et al. Loss of expression of tropomyosin-1, a novel class II tumor suppressor that induces anoikis, in primary breast tumors. Oncogene 2003;22:6194–203.[CrossRef][Medline]
  28. Chiyo M, Shimozato O, Yu L, et al. Expression of IL-27 in murine carcinoma cells produces antitumor effects and induces protective immunity in inoculated host animals. Int J Cancer 2005;115:437–42.[CrossRef][Medline]
  29. Yu G, Tseng GC, Yu YP, et al. CSR1 suppresses tumor growth and metastasis of prostate cancer. Am J Pathol 2006;168:597–607.[Abstract/Free Full Text]
  30. Kang Y, He W, Tulley S, et al. Breast cancer bone metastasis mediated by the Smad tumor suppressor pathway. Proc Natl Acad Sci U S A 2005;27:13909–14.
  31. Reedijk M, Odorcic S, Chang L, et al. High-level coexpression of JAG1 and NOTCH1 is observed in human breast cancer and is associated with poor overall survival. Cancer Res 2005;65:8530–7.[Abstract/Free Full Text]
  32. Orimo A, Weinberg RA. Stromal fibroblasts in cancer: a novel tumor-promoting cell type. Cell Cycle 2006;5:1597–601.[Medline]
  33. Sato H, Takino T, Miyamori H. Roles of membrane-type matrix metalloproteinase-1 in tumor invasion and metastasis. Cancer Sci 2005;96:212–7.[CrossRef][Medline]
  34. Martins-Green M, Boudreau N, Bissell MJ. Inflammation is responsible for the development of wound-induced tumors in chickens infected with Rous sarcoma virus. Cancer Res 1994;54:4334–41.[Abstract/Free Full Text]




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 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 Google Scholar
Google Scholar
Right arrow Articles by Stasinopoulos, I.
Right arrow Articles by Bhujwalla, Z. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Stasinopoulos, I.
Right arrow Articles by Bhujwalla, Z. M.

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