Expression of cyclooxygenase 2 (COX-2) in breast cancer correlates with poor prognosis, and COX-2 enzyme inhibitors reduce breast cancer incidence in humans. We recently showed that COX-2 overexpression in the mammary gland of transgenic mice induced mammary cancer. Because prostaglandin E2 (PGE2) is the major eicosanoid and because the EP2 subtype of the PGE2 receptor is highly expressed in the mammary tumors, we tested if this G protein–coupled receptor is required for tumorigenesis. We crossed the MMTV-COX-2 transgenic mice with Ep2−/− mice and studied tumor development in bigenic mice. Lack of EP2 receptor strongly suppressed COX-2–induced effects such as precocious development of the mammary gland in virgins and the development of mammary hyperplasia in multiparous female mice. Interestingly, the expression of amphiregulin, a potent mammary epithelial cell growth factor was down regulated in mammary glands of Ep2−/− mice. Total cyclic AMP (cAMP) levels were reduced in Ep2−/− mammary glands suggesting that PGE2 signaling via the EP2 receptor activates the Gs/cAMP/protein kinase A pathway. In mammary tumor cell lines, expression of the EP2 receptor followed by treatment with CAY10399, an EP2-specific agonist, strongly induced amphiregulin mRNA levels in a protein kinase A–dependent manner. These data suggest that PGE2 signaling via the EP2 receptor in mammary epithelial cells regulate mammary gland hyperplasia by the cAMP-dependent induction of amphiregulin. Inhibition of the EP2 pathway in the mammary gland may be a novel approach in the prevention and/or treatment of mammary cancer.

Genetic, pharmacologic, and epidemiologic evidence support the critical role of cyclooxygenase 2 (COX-2) in tumorigenesis (1). Thus, significant efforts have been expended to inhibit the COX-2 enzyme to determine if this indeed provides a benefit in the prevention and/or treatment of various cancers. However, recent withdrawal of the COX-2 inhibitor Rofecoxib due to enhanced cardiovascular adverse effects raised concerns about this pharmacologic approach (2). Because the significance of COX-2 in cancer is not in dispute, alternative mechanisms of blocking COX2 signaling must be sought. One avenue that has not been well studied is the potential role for COX-2 targets that act downstream in specific tumor settings. Therefore, the fundamental question is how COX-2 regulates tumor development, and to this end, various studies have implicated the important role played by prostaglandin E2 (PGE2), a major COX-2 product in various cancers (3). We recently reported that overexpression of human COX-2 in the mammary glands of transgenic mice (MMTV-COX-2 mice) resulted in the development of mammary tumors in female multiparous CD1 mice (4). This effect was inhibited by indomethacin (a nonselective COX inhibitor) as well as by Celecoxib (a specific COX-2 inhibitor) suggesting that the enzyme activity of COX-2 is required for the tumorigenic effect (5).4

4

Unpublished observations.

Analysis of eicosanoids synthesized in the mammary gland indicated that PGE2 is the major product (5). These studies suggest that PGE2 signaling is involved mammary tumorigenesis. However, it is not clear which receptors and signal transduction pathways are involved. In addition, critical downstream molecular events are also not defined. In this study, we focused our efforts on the EP2 subtype of the G protein–coupled receptor for PGE2, which is expressed in the mammary epithelial cells and is induced during the proliferative stage of mammary gland (5). We show that the EP2 receptor is required for mammary epithelial hyperplasia in COX-2 transgenic mice and that it induces the expression of potent epithelial cell growth factor amphiregulin by a cyclic AMP (cAMP)/protein kinase A signaling pathway.

Chemicals. The EP2 agonist CAY10399, a free acid, two-series analogue of butaprost was obtained from Cayman Chemical (Ann Arbor, MI).

Animals.MMTV-COX-2 transgenic mice (4) in FVB/N background were crossed with Ep2−/− mice in the C57BL/6J background (6). The resulting MMTV-COX-2 Ep2+/− mice in the (FVB/N;C57BL/6) background were backcrossed with FVB/N mice three to five times. The resulting MMTV-COX-2 Ep2+/− female mice were crossed with Ep2+/− male mice in same strain yielding the three experimental groups: MMTV-COX-2 Ep2+/+, MMTV-COX-2 Ep2+/−, and MMTV-COX-2 Ep2−/−. Some females are kept as virgins for 3 months and dissected for the precocious development study. All other females underwent three rounds of pregnancies. Pups were removed after 4 days postpartum. After 1 month of weaning, mammary glands were dissected for experiments.

Histology and immunohistochemistry. Mammary glands were isolated and mammary gland 4 was used for whole mount analysis and mammary gland 9 was embedded in paraffin for H&E and immunohistochemistry as described (5). After blocking with goat serum (COX-2 and EP2) or rabbit serum (amphiregulin), anti-COX-2 (Cayman Chemical), anti-EP2 (Cayman Chemical), or anti-amphiregulin (R&D Systems, Minneapolis, MN) was incubated and followed by the Vectastain avidin-biotin complex reagent and 3,3′-diaminobenzidine for color development as described (5).

Cyclic AMP measurement. Mammary glands (100 mg) were homogenized in radioimmunoprecipitation assay buffer (RIPA: 0.1% SDS, 0.5% sodium deoxycholate, 1% NP40, 1 mmol/L sodium orthovanadate, 50 mmol/L β-glycerophosphate, and 1× protease inhibitor cocktail). Samples were centrifuged 14,000 × g for 10 minutes, and 50 μL of supernatant were used for cAMP measurement by a RIA (Amersham Biosciences, Piscataway, NJ). cAMP amount was quantitated after normalization of protein concentration measured by Bradford assay.

RNA purification and reverse transcription-PCR analysis. Total RNA was isolated from mammary gland tissues as well as mammary tumor cells and quantitative reverse transcription-PCR was done as described (7). Primers for amphiregulin and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) are as follows. Amphiregulin sense 5′-GGTCTTAGGCTCAGGCCATTA-3′ and Amphiregulin antisense 5′-CGCTTATGGTGGAAACCTCTC-3′. GAPDH sense 5′-CAACTACATGGTCTACATGTTCCAGTATG-3′ and GAPDH antisense 5′-TGACCCGTTTGGCTCCA-3′.

Western blot analysis. Mammary tumor cells were lysed with RIPA buffer and Western blot analysis was carried out as described (4). Incubation of primary antibodies for amphiregulin (1:500, R&D Systems) or β-actin (1:10,000, Sigma, St. Louis, MO) was followed by incubation of secondary antibodies, respectively, anti-goat or anti-mouse (1:5,000, ICN Amersham Pharmacia) conjugated to horseradish peroxidase and visualized by an enhanced chemiluminescence system (Amersham Pharmacia).

EP2 receptor is required for mammary hyperplasia. We recently reported that MMTV-COX-2 mice in the CD1 background exhibit precocious development of the mammary gland in virgins and hyperplasia of the mammary glands in multiparous mice (4, 5). To determine the role for the EP2 receptor, we crossed the MMTV-COX-2 transgenic mice with the Ep2−/− mice (6) and obtained MMTV-COX-2 Ep2+/− mice. We have recently determined that mammary hyperplasia was readily observed in the FVB/N background but not in the C57/Bl6 background.4 These mice were then backcrossed three to five times into the FBV/N background. Subsequently, such mice were intercrossed to obtain littermates of MMTV-COX-2 Ep2+/+ and MMTV-COX-2 Ep2−/− mice. At 12 weeks of age, mammary glands from virgin MMTV-COX-2 Ep2+/+ female mice indicated precocious development (Fig. 1A). In contrast, MMTV-COX-2 Ep2−/− mice showed normal mammary development. These data suggest that abnormal mammary gland development induced by the COX-2 pathway requires the EP2 receptor.

Figure 1.

Mammary hyperplasia requires the EP2 receptor. A, whole mount analysis. Mammary glands from EP2+/+/nontransgenic, EP2−/−/nontransgene, EP2+/+/MMTV-COX-2, and EP2−/−/MMTV-COX-2 littermates at 12-week virgin females were dissected out and stained with carmine red. Magnification, ∼3×. B, morphology of mammary glands. Left, whole mounts (magnification, ∼3×) and middle, H&E staining (magnification, 10×; bar, 100 μm) of mammary glands from multiparous female mice. Expression of human COX-2 transgene in EP2+/− and EP2−/− mammary glands was confirmed by reverse transcription-PCR analysis (right).

Figure 1.

Mammary hyperplasia requires the EP2 receptor. A, whole mount analysis. Mammary glands from EP2+/+/nontransgenic, EP2−/−/nontransgene, EP2+/+/MMTV-COX-2, and EP2−/−/MMTV-COX-2 littermates at 12-week virgin females were dissected out and stained with carmine red. Magnification, ∼3×. B, morphology of mammary glands. Left, whole mounts (magnification, ∼3×) and middle, H&E staining (magnification, 10×; bar, 100 μm) of mammary glands from multiparous female mice. Expression of human COX-2 transgene in EP2+/− and EP2−/− mammary glands was confirmed by reverse transcription-PCR analysis (right).

Close modal

We next allowed MMTV-COX-2 Ep2+/+, MMTV-COX-2 Ep2+/−, and MMTV-COX-2 Ep2−/− mice to undergo several rounds of pregnancy and lactation and analyzed the mammary gland morphology of the multiparous mice after weaning for 4 weeks. Mammary hyperplasia was observed in 9 of 11 mice that express the EP2 receptor (both Ep2+/− and Ep2+/+). Mammary glands showed abundant ductal branches and alveolar development as shown in the whole mount analysis (Fig. 1B). In tissue sections, epithelial hyperplasia was evident. Lack of EP2 receptor strongly suppressed the development of mammary hyperplasia in multiparous mice (9 of 11 mice). Expression of the Cox-2 transcript was essentially similar in the mammary gland RNA extracts of mice that lack the EP2 receptor compared with those that express it. These data suggest that the EP2 receptor is required for the development of mammary hyperplasia in COX-2 transgenic mice.

Reduced expression of amphiregulin in the EP2 null MMTV-COX-2 mammary glands. Epidermal growth factor receptor (EGFR) ligands are potent regulators of mammary gland development and tumorigenesis (8). We hypothesized that PGE2 signaling via the EP2 receptor might regulate the EGF/EGFR system and thereby induce mammary gland hyperplasia. Therefore, we examined the expression of EGF-like ligands in these tissues (9). Whereas the level of EGF mRNA was relatively similar in both groups, the expression of amphiregulin mRNA was markedly reduced in MMTV-COX-2 Ep2−/− compared with MMTV-COX-2 Ep2+/− mammary glands obtained from multiparous mice (Fig. 2A). Consistently, amphiregulin-expressing cells were markedly reduced in MMTV-COX-2 Ep2−/− mammary glands. In contrast, amphiregulin was strongly expressed in most ductal and alveolar epithelial compartment of MMTV-COX-2 Ep2+/− mammary glands (Fig. 2B). These data suggest that EP2 signaling in the mammary gland regulates amphiregulin expression.

Figure 2.

Reduced expression of amphiregulin (AREG) in the MMTV-COX-2 Ep2−/− mammary glands. A, quantitative reverse transcription-PCR analysis. mRNA expression of amphiregulin and EGF was measured from three independent mammary glands in MMTV-COX-2 Ep2+/− and MMTV-COX-2 Ep2−/− multiparous mammary glands. P of amphiregulin expression (P ≤ 0.0312) was significant by Student's t test. B, immunohistochemistry. Enhanced expression of amphiregulin in mammary epithelial cells in low (5×) and high (40×) magnification in the wild-type mice. Bar, 100 μm.

Figure 2.

Reduced expression of amphiregulin (AREG) in the MMTV-COX-2 Ep2−/− mammary glands. A, quantitative reverse transcription-PCR analysis. mRNA expression of amphiregulin and EGF was measured from three independent mammary glands in MMTV-COX-2 Ep2+/− and MMTV-COX-2 Ep2−/− multiparous mammary glands. P of amphiregulin expression (P ≤ 0.0312) was significant by Student's t test. B, immunohistochemistry. Enhanced expression of amphiregulin in mammary epithelial cells in low (5×) and high (40×) magnification in the wild-type mice. Bar, 100 μm.

Close modal

Regulation of amphiregulin expression through EP2/PKA pathway. To show that PGE2 signaling via the EP2 receptor is directly involved in the induction of amphiregulin, we used the EP2 null mammary myoepithelial cell line (C11 cell line) isolated from tumors of MMTV-COX-2 transgenic mice (10). EP2 receptor was expressed in C11 cells by adenoviral transduction. Treatment with CAY10399, an EP2-specific agonist, induced mRNA expression of amphiregulin in a dose-dependent manner only when cells are expressing the EP2 receptor (Fig. 3A). Previous studies showed that CAY10399 dose-dependently increased cAMP levels in EP2-expressing C11 cells (10), suggesting that cAMP/PKA pathway might mediate the induction of amphiregulin.

Figure 3.

Regulation of amphiregulin via the EP2/cAMP pathway. A and C, after transduction of adenovirus expressing EP2 (adEP2), or vector (adCON) into EP2-null nontumorigenic mammary cells, indicated dose of CAY10399 (A) and vehicle (V), 1 μmol/L of CAY10399 (C1), or 1 μmol/L of CAY10399 with 10 μmol/L of H-89 (C1H, C) were administered for 1 hour and mRNA expression of amphiregulin (AREG) was analyzed. B, cAMP level was measured in the mammary gland extracts from MMTV-COX-2 Ep2+/− and MMTV-COX-2 Ep2−/− mice. Column, mean from three independent animals; bars, ±SD. P ≤ 0.0134. D, after treatment as indicated for 6 hours in vector or EP2 transduced cells, the expression of amphiregulin and β-actin was examined by Western blot analysis.

Figure 3.

Regulation of amphiregulin via the EP2/cAMP pathway. A and C, after transduction of adenovirus expressing EP2 (adEP2), or vector (adCON) into EP2-null nontumorigenic mammary cells, indicated dose of CAY10399 (A) and vehicle (V), 1 μmol/L of CAY10399 (C1), or 1 μmol/L of CAY10399 with 10 μmol/L of H-89 (C1H, C) were administered for 1 hour and mRNA expression of amphiregulin (AREG) was analyzed. B, cAMP level was measured in the mammary gland extracts from MMTV-COX-2 Ep2+/− and MMTV-COX-2 Ep2−/− mice. Column, mean from three independent animals; bars, ±SD. P ≤ 0.0134. D, after treatment as indicated for 6 hours in vector or EP2 transduced cells, the expression of amphiregulin and β-actin was examined by Western blot analysis.

Close modal

To further study the functional significance of the link of EP2, cAMP, and amphiregulin, we measured the level of cAMP in the mammary gland because EP2 is a Gs-coupled receptor (6) and amphiregulin is a cAMP-regulatory gene (11). cAMP was reduced ∼50% in MMTV-COX-2 Ep2−/− mammary glands when compared with MMTV-COX-2 Ep2+/− mammary glands (Fig. 3B).

We next tested if inhibition of PKA signaling affects EP2-dependent induction of amphiregulin. Treatment with H-89, a PKA inhibitor, potently blocked the induction of amphiregulin mRNA by CAY10399 in EP2-expressing cells (Fig. 3C). Similar changes in amphiregulin protein expression were observed (Fig. 3D). These data indicate that EP2/Gs/cAMP/PKA signaling pathway is directly involved in the induction of amphiregulin.

The COX-2 pathway is now recognized important in human cancer development and progression (1). Studies conducted in mouse models indicate that PGE2 signaling by EP and PPAP receptors are important in intestinal polyposis (3, 12). In addition, EP4 receptor may also be involved in promoting intestinal tumorigenesis (13). Specifically, the EP2 receptor was implicated to control polyp angiogenesis and growth (14). Although both EP2 and EP4 receptors are coupled to the Gs signaling pathway, EP4 was shown to also regulate the phosphatidylinositol-3 kinase pathway (15). Recent studies in mammary tumorigenesis also show that COX-2 is critical for breast cancer development (1). PGE2 is a major eicosanoid synthesized by the COX-2 pathway in the mammary gland (5). However, role of specific receptors for PGE2 in mammary tumor development has not been addressed.

We recently showed that overexpression of COX-2 in the mammary gland of CD1 transgenic mice resulted in tumorigenic transformation (4). This effect was only seen in multiparous female mice, was inhibited by indomethacin and Celecoxib, and was seen in CD1 and FVB/N background but not in the C57Bl/6 background (4, 5).4 These data suggest that enzymatic function of COX-2 cooperate with other events to induce tumorigenesis and that genetic modifier genes can dramatically alter the effect. Molecular mechanisms involved are important to be delineated as this information has the potential to be rapidly applied to cancer prevention and/or treatment.

In this study, we focused on the EP2 receptor, a Gs-coupled receptor for PGE2. Deletion of the Ep2 gene in mice led to defects in ovulation, embryonic implantation, and salt-sensitive hypertension (6). However, viability of the Ep2−/− mice are generally uncompromised. EP2 receptor is expressed in mammary epithelial cells, its expression is induced during mammary gland development (during pregnancy), and is highly expressed in mammary tumor epithelial cells (5). Because COX-2 is also expressed in the mammary epithelial cells of the MMTV-COX-2 mice (4), we hypothesized that autocrine signaling of EP2 receptor may be involved in the tumorigenic phenotype. Indeed, the development of mammary hyperplasia, a precursor lesion for invasive mammary tumors, is strongly suppressed in Ep2−/− mice. Similarly, precocious development of mammary gland was also inhibited in Ep2−/− mice. These data suggest that PGE2/EP2 signaling within the mammary epithelial compartment are involved in cellular proliferation and hyperplasia. Interestingly, we did not observe significant changes in microvessel density or vascular endothelial growth factor expression. This is probably due to the fact that EP2 is expressed in the epithelial cell compartment and therefore does not regulate stromal events such as angiogenesis. For example, the EP4 receptor is expressed in the mammary stromal cells (5) and may be involved in the regulation of angiogenic response and tumorigenesis. Indeed, both EP2 and EP4 receptors are important for different aspects of intestinal tumorigenesis (16).

PGE2 signaling intersects with EGF signaling in the context of cancer. For example, PGE2 transactivates the EGF receptors in colon cancer cells in vitro (17). Moreover, PGE2 treatment of colon cancer cells induces the expression of amphiregulin, an EGF family member ligand implicated in mammary cancer development (11). In that study, it was shown that the cAMP/PKA pathway is important in the transcriptional induction of amphiregulin. As COX-2 inhibitors delay cancer development in HER-2/neu (EGFR) model of mammary cancer (18), we investigated the expression patterns of EGF-ligand family members in Ep2−/− and wild-type mice. Our results clearly show that EP2 receptor controls the expression of amphiregulin in the mammary epithelial cells in vivo. Amphiregulin was shown to regulate ductal morphogenesis as a mitogen in mice (9). In addition, retroviral overexpression of amphiregulin into mammary glands induces hyperplasia in vivo (19), suggesting an important role of amphiregulin in premalignant mammary hyperplasia. Our data imply that COX-2–derived PGE2/EP2 signaling may influence the EGF signaling system by up-regulating the expression of amphiregulin.

The novel mechanism elucidated in this study (i.e., EP2/Gs/cAMP/amphiregulin pathway in the induction of mammary hyperplasia) adds to our knowledge of the role of the COX-2 pathway in mammary cancer development. EP2 receptor inhibitors may be useful in the prevention and/or treatment of breast cancer.

Grant support: NIH grants CA95181 (T. Hla), CA77839 (T. Hla), and GM15431 (R.M. Breyer).

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

1
Dannenberg AJ, Subbaramaiah K. Targeting cyclooxygenase-2 in human neoplasia: rationale and promise.
Cancer Cell
2003
;
4
:
431
–6.
2
Couzin J. Clinical trials. Nail-biting time for trials of COX-2 drugs.
Science
2004
;
306
:
1673
–5.
3
Gupta RA, Dubois RN. Colorectal cancer prevention and treatment by inhibition of cyclooxygenase-2.
Nat Rev Cancer
2001
;
1
:
11
–21.
4
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.
5
Chang SH, Liu CH, Conway R, et al. Role of prostaglandin E2-dependent angiogenic switch in cyclooxygenase 2-induced breast cancer progression.
Proc Natl Acad Sci U S A
2004
;
101
:
591
–6.
6
Kennedy CR, Zhang Y, Brandon S, et al. Salt-sensitive hypertension and reduced fertility in mice lacking the prostaglandin EP2 receptor.
Nat Med
1999
;
5
:
217
–20.
7
Sanchez T, Estrada-Hernandez T, Paik JH, et al. Phosphorylation and action of the immunomodulator FTY720 inhibits vascular endothelial cell growth factor-induced vascular permeability.
J Biol Chem
2003
;
278
:
47281
–90.
8
Normanno N, Ciardiello F. EGF-related peptides in the pathophysiology of the mammary gland.
J Mammary Gland Biol Neoplasia
1997
;
2
:
143
–51.
9
Luetteke NC, Qiu TH, Fenton SE, et al. Targeted inactivation of the EGF and amphiregulin genes reveals distinct roles for EGF receptor ligands in mouse mammary gland development.
Development
1999
;
126
:
2739
–50.
10
Chang S, Liu CH, Wu MT, Hla T. Regulation of vascular endothelial growth factor expression in mouse mammary tumor cells by the EP2 subtype of prostaglandin E2 receptor. Prostaglandins Other Lipid Mediat. In press 2005.
11
Shao J, Evers BM, Sheng H. Prostaglandin E2 synergistically enhances receptor tyrosine kinase-dependent signaling system in colon cancer cells.
J Biol Chem
2004
;
279
:
14287
–93.
12
Wang D, Wang H, Shi Q, et al. Prostaglandin E(2) promotes colorectal adenoma growth via transactivation of the nuclear peroxisome proliferator-activated receptor δ.
Cancer Cell
2004
;
6
:
285
–95.
13
Mutoh M, Watanabe K, Kitamura T, et al. Involvement of prostaglandin E receptor subtype EP(4) in colon carcinogenesis.
Cancer Res
2002
;
62
:
28
–32.
14
Seno H, Oshima M, Ishikawa TO, et al. Cyclooxygenase 2- and prostaglandin E(2) receptor EP(2)-dependent angiogenesis in Apc(Δ716) mouse intestinal polyps.
Cancer Res
2002
;
62
:
506
–11.
15
Sheng H, Shao J, Washington MK, DuBois RN. Prostaglandin E2 increases growth and motility of colorectal carcinoma cells.
J Biol Chem
2001
;
276
:
18075
–81.
16
Hull MA, Ko SC, Hawcroft G. Prostaglandin EP receptors: targets for treatment and prevention of colorectal cancer?
Mol Cancer Ther
2004
;
3
:
1031
–9.
17
Pai R, Soreghan B, Szabo IL, Pavelka M, Baatar D, Tarnawski AS. Prostaglandin E2 transactivates EGF receptor: a novel mechanism for promoting colon cancer growth and gastrointestinal hypertrophy.
Nat Med
2002
;
8
:
289
–93.
18
Howe LR, Subbaramaiah K, Patel J, et al. Celecoxib, a selective cyclooxygenase 2 inhibitor, protects against human epidermal growth factor receptor 2 (HER-2)/neu-induced breast cancer.
Cancer Res
2002
;
62
:
5405
–7.
19
Kenney NJ, Smith GH, Rosenberg K, Cutler ML, Dickson RB. Induction of ductal morphogenesis and lobular hyperplasia by amphiregulin in the mouse mammary gland.
Cell Growth Differ
1996
;
7
:
1769
–81.