Cyclooxygenase-2 predicts adverse effects of tamoxifen: a possible mechanism of role for nuclear HER2 in breast cancer patients

Transcription

1 Cyclooxygenase-2 predicts adverse effects of tamoxifen: a possible mechanism of role for nuclear HER2 in breast cancer patients Mary F Dillon 1,2, Anthony T Stafford 1, Gabrielle Kelly 3, Aisling M Redmond 1, Marie McIlroy 1, Thomas B Crotty 2, E McDermott 2, Arnold D Hill 1 and Leonie S Young 1 1 Endocrine Oncology Research Group, Department of Surgery, Royal College of Surgeons in Ireland, St Stephens Green, Dublin 2, Ireland 2 UCD Conway Institute, School of Medicine and Medical Science, St Vincent s University Hospital and University College Dublin, Dublin, Ireland 3 School of Mathematical Sciences, Statistics and Actuarial Science, University College Dublin, Dublin, Ireland (Correspondence should be addressed to L S Young; lyoung@rcsi.ie) Abstract Cyclooxygenase-2 (COX-2) is associated with breast tumour progression. Clinical and molecular studies implicate human epidermal growth factor receptor 2 (HER2) in the regulation of COX-2 expression. Recent reports raise the possibility that HER2 could mediate these effects through direct transcriptional mechanisms. The relationship between HER2 and COX-2 was investigated in a cohort of breast cancer patients with or without endocrine treatment. A tissue microarray comprising tumours from 560 patients with 10-year follow-up was analysed for HER2, ERK1/2, polyoma enhancer activator 3 (PEA3) and COX-2 expression. Subcellular localisation of HER2 was assessed by immunofluorescence and confocal microscopy. Expression of markers examined was analysed in relation to classic clinicopathological parameters and disease-free survival in the presence and absence of tamoxifen. COX-2 expression associated with both membranous and nuclear expression of HER2 (PZ and P! respectively). No association was detected between COX-2 and either ERK1/2 or PEA3 (PZ0.7 and PZ0.3 respectively). None of the markers were found to be independently prognostic. Membrane HER2, nuclear HER2 and COX-2, however, were all found to predict poor disease-free survival in patients on endocrine treatment (PZ0.0017, PZ and PZ respectively). Moreover, patients who were positive for COX-2 predicted adverse effects of tamoxifen (PZ0.0427). These clinical ex vivo data are consistent with molecular observations that HER2 can regulate COX-2 expression through direct transcriptional mechanisms. COX-2 expression correlates with disease progression on endocrine treatment. This study supports a role for COX-2 as a predictor of adverse effects of tamoxifen in breast cancer patients. Endocrine-Related Cancer (2008) Introduction There is substantial evidence to suggest that cyclooxygenase-2 (COX-2), an enzyme that catalyses the formation of prostaglandins, is important in breast carcinogenesis. Expression of COX-2 is associated with parameters of disease progression, including size, positive lymph node metastasis and human epidermal growth factor receptor 2 (HER2)-positive status (Howe et al. 2001b, Ristimaki et al. 2002, Nassar et al. 2007). Furthermore, overexpression studies in mice suggest that COX-2 is important in the genesis of mammary tumours (Liu et al. 2001). COX-2-derived prostaglandins drive tumour growth through a number of mechanisms including, cell proliferation (Sheng et al. 2001), angiogenesis (Tsujii et al. 1998) and local invasion (Dohadwala et al. 2001). The tyrosine kinase receptor HER2 is amplified in 20 30% of human cancers and overexpression has been associated with poor patient prognosis (Ross & Fletcher 1998). Molecular and translational studies Endocrine-Related Cancer (2008) /08/ q 2008 Society for Endocrinology Printed in Great Britain Downloaded from Bioscientifica.com DOI: /ERC at 09/10/ :53:06PM Online version via

2 M F Dillon et al.: HER2 and COX-2 in endocrine-resistant breast cancer provide a substantial link between HER2 signalling and COX-2 expression in cancer (Ristimaki et al. 2002, Simeone et al. 2004). Significant associations between COX-2 and HER2 expression has been reported in human breast cancer (Hwang et al. 1998). The HER2 receptor can employ several second messenger signalling mechanisms, including the mitogen-activated protein kinase (MAPK) pathway and the phosphatidylinositol 3-kinase/protein kinase B (AKT) system (Moasser 2007). The Ets proteins are a family of MAPK-dependent transcription factors. The Ets member polyoma enhancer activator 3 (PEA3) is thought to play an essential role in HER2-mediated oncogenesis (Benz et al. 1997, Shepherd et al. 2001). We have recently reported strong associations between co-expression of PEA3 and HER2 and reduced disease-free survival in breast cancer patients (Myers et al. 2006). Several molecular studies suggest that PEA3 can transcriptionally regulate COX-2 (Howe et al. 2001a, Subbaramaiah et al. 2002). In HER2-transformed human mammary epithelial cells, HER2 can signal to PEA3 via the MAP kinase pathway to regulate COX-2 expression (Subbaramaiah et al. 2002). However, in a clinical study focusing on stage I breast cancer patients, no association was observed between COX-2 expression and implicated signalling pathways, including MAP kinase and AKT (Schmitz et al.2006). Although the function of HER2 as a transmembrane receptor has been well described recently, several studies provide evidence of a nuclear function as a transcriptional regulator (Xie & Hung 1994, Vadlamundi et al. 1999, Wang et al. 2004). Nuclear HER2 has been found to associate with multiple genomic targets, including the COX-2 gene promoter. HER2 can form a complex at specific nucleotide sequences within the COX-2 promoter to regulate gene transcription (Wang et al. 2004). These observations raise the possibility that in breast cancer cells HER2 can mediate its effects, at least in part, through direct transcriptional regulation of COX-2. In order to elucidate the relationship between HER2 and COX-2 in endocrine-treated breast cancer, we examined the expression of membrane HER2, nuclear HER2, the signalling mediator ERK1/2, the transcription factor PEA3 and COX-2 in a large cohort of patients. Materials and methods Patient information and construction of tissue microarray (TMA) Following ethical approval, breast tumour samples were obtained from archival cases at St Vincent s 746 University Hospital, Dublin, Ireland, over the period from 1987 to Patients received either no endocrine treatment (nz200; ER positive 68%) or tamoxifen (nz360; ER positive 74%), 20 mg/day for 5 years, which was discontinued only in those patients who suffered a relapse while on endocrine therapy. Excluded from the analysis were patients who did not have breast surgery, those who had neoadjuvant therapy or those whose tissue specimens were irretrievable. Data on the patients included pathological characteristics (tumour size, grade, lymph node status, oestrogen receptor status) as well as treatment with radiotherapy, chemotherapy or tamoxifen. Detailed 10-year follow-up data (median 7.72 years) was collected from the patients to determine diseasefree and overall survival. Archival tissue was attained on the patient population for the purposes of TMA construction. A haematoxylin eosin slide was used on all specimens to mark the site of carcinoma. Four 0.6 mm punch biopsies were taken from each specimen, and transplanted into a recipient block. The recipient block was cut into 5 mm sections, mounted on Superfrost Plus slides (BDH, Poole, UK) and baked in an oven for 1 h at 60 8C. Immunohistochemistry Sections were dewaxed, rehydrated and washed with PBS. Endogenous peroxidase activity was blocked by immersion in 3% hydrogen peroxidase (BDH) in distilled water for 20 min in the dark. Antigen retrieval was performed by immersing the sections in 0.01 M sodium citrate buffer (Sigma Aldrich) and microwaving on high power for 7 min and low power for 15 min. The sections were blocked in the appropriate serum (Sigma Aldrich) for 60 min. All the primary antibodies were diluted in PBS. The sections were incubated with primary antibodies as follows: mouse anti-human ERa (5 mg/ml) (cat no. sc-8002; Santa Cruz Biotechnology, Santa Cruz, CA, USA); rabbit anti-human phospho-p44/42 MAPK (Thr202/Tyr204) (10 mg/ml) (cat no. 4376L; Cell Signalling Tech, Danvers, MA, USA), mouse anti-human PEA3 (50 mg/ml) (cat no. sc-113; Santa Cruz); mouse anti-human COX-2 (8 mg/ml) (cat no ; Cayman, Ann Arbor, MI, USA). The primary antibodies were incubated for 60 min at room temperature, except for anti-cox-2 antibody, which was incubated for 120 min, or anti-pea3 antibody, overnight at 4 8C. The sections are subsequently incubated with the corresponding biotin-labelled secondary (1:2000 in PBS) (Vector Laboratories, Burlingame, CA, USA) for

3 30 min, followed by peroxidase-labelled avidin biotin complex (Vector Laboratories) for 30 min. The sections were developed in 3,3 0 -diaminobenzidine tetrahydrochloride for 3 7 min and counterstained with haematoxylin for 2 min. The sections were then passed through increasing concentrations of industrial methylated spirits (70, 90 and 100%) for 5 min in each container and then xylene for 10 min. Immunostained slides were scored for phospho- ERK1/2, PEA3 and COX-2 using the Allred scoring system (Harvey et al. 1999). Independent observers, without knowledge of prognostic factors, scored slides. Assessment of HER2 status HER2 status was evaluated using the DAKO (Glostrup, Denmark) HercepTest immunocytochemical assay. Scoring was assessed in accordance with the manufacturer s instructions. A score was assigned according to the intensity and pattern of cell membrane staining: 0toC1, no staining or staining in!10% of cells; C2, weak-to-moderate staining in O10% of cells and C3, strong staining in O10% of cells. In tumour samples scoring C2 with the Hercep Test, the samples were selected out and a TMA constructed consisting of these tumours. HER2 status was confirmed by fluorescent in situ hybridisation using the PathVysion kit probe to detect amplification of the HER2 gene (spectrum orange-labelled HER2 and spectrum green-labelled a satellite centromeric region for chromosome 17) (Vysis Inc., Downers Grove, IL, USA), according to the manufacturer s instructions. Criteria for gene amplification were tight clusters of HER2 signals in multiple cells with at least two times more HER2 signal than centromeric 17. Immunofluorescent microscopy Breast cancer sections were prepared as above and incubated in goat serum for 60 min. Rabbit anti-human HER2 (10 mg/ml in 10% human serum, cat no. 2165; Cell Signalling Tech) was placed on each slide for 90 min. The sections were incubated with the corresponding secondary fluorochrome-conjugated antibody (1:100) (Sigma Aldrich) for 60 min. The sections were mounted using fluorescent mounting media (DAKO). The slides were examined under a confocal fluorescent microscope. Negative controls were performed by matched IgG (Rabbit IgG, cat no. X0903; DAKO). A score was assigned according to the intensity and pattern of nuclear staining: 0 to C1, no staining or staining in!10% of cells; C2, weakto-moderate staining in O10% of cells and C3, strong Figure 1 (A) Immunohistochemical localisation of HER2 (100!), with fluorescent in situ hybridisation of HER2 gene amplification (200!) (inset) and matched IgG control. (B) Immunofluorescent localisation of HER2 to the membrane (HER2m) (100!), (400!) and counterstained with DAPI (400!) (insets). Immunofluorescent localisation of HER2 to the membrane and nucleus (HER2m and HER2n) (100!), (400!) and counterstained with DAPI (400!) (insets). (C) Immunohistochemical localisation of phospho-erk1/2 (100!), (400!) (insets) with matched IgG control. (D) Immunohistochemical localisation of PEA3 (100!), (400!) (insets) with matched IgG control. (E) Immunohistochemical localisation COX-2 (100!), (400!)(insets) with matched IgG control

4 M F Dillon et al.: HER2 and COX-2 in endocrine-resistant breast cancer staining in O10% of cells. A score of R2 was considered positive. Statistical analysis Statistical analysis was carried out using Fisher s exact test for categorical variables to compare two proportions. Kaplan Meier estimates of survival functions were computed and the Wilcoxon test was used to compare survival curves. In addition, the Wilcoxon rank-sum test was used to compare two medians. Two-sided P values!0.05 were considered to be statistically significant. Results Localisation and expression of HER2, phospho-erk1/2, PEA3 and COX-2 in human breast cancer tissue The tyrosine kinase receptor HER2 was found to be expressed predominantly at the cell membrane (Fig. 1A). On immunohistochemical analysis, there was also notable nuclear staining in a large number of HER2-positive breast cancer patients. Immunofluorescence and confocal microscopy confirmed nuclear localisation of HER2 (Fig. 1B). Phospho-ERK1/2 was expressed in both the nucleus and the cytoplasm of the tumour epithelial cells, whereas COX-2 was found to be expressed primarily in the cytoplasm. PEA3 was found to be predominantly nuclear, with some scant staining observed in the cytoplasm (Fig. 1C E). Membrane and nuclear HER2 were expressed in 19.9 and 12.3% respectively of the total breast cancer patient population. The MAP kinase signalling mediator, phospho-erk1/2, was expressed in 30.3% of patients. PEA3 and COX-2 were detected in 53.1 and 47.8% of patients respectively (Table 1). Consistent with the theory that HER2 can signal to PEA3 to transcriptionally regulate COX-2 expression, there were significant associations between membrane HER2 and PEA3 (PZ0.0043) and membrane HER2 and COX-2 (PZ0.0033). However, no association was detected between either membrane HER2 and ERK1/2 (PZ0.51) or between PEA3 and COX-2 (PZ0.3) (Table 1). To elucidate the role of HER2 as a direct transcriptional regulator of COX-2 in breast cancer patients, we examined the relationship between nuclear HER2 and COX-2. Tumours that expressed HER2 in the nucleus of the cancer epithelial cells were significantly associated with expression of COX-2 (P! ) (Table 1). These data support a role for HER2 as a transcription factor important in the regulation of COX-2. An inverse association between ER expression and expression of Table 1 Associations of markers, HER2 membrane (HER2m), HER2 nuclear (HER2n), phospho-erk1/2, PEA3 and cyclooxygenase-2 (COX-2) expression with each other using Fisher s exact test PEA3 COX-2 Total (%) ER HER2m HER2n ERK1/2 ERCve !0.001 ERKve HER2mCve ! HER2mKve HER2nCve ! HER2nKve ERK1/2Cve ERK1/2Kve PEA3Cve PEA3 Kve COX-2Cve COX-2Kve

5 Table 2 Comparisons of HER2 membrane (HER2m), HER2 nuclear (HER2n), phospho-erk1/2, PEA3 and cyclooxygenase-2 (COX-2) expression with clinicopathological parameters using Fisher s exact test Total n HER2m HER2n P ERK1/2 PEA3 COX-2 % Positive patients % Axilla Positive % Negative % Size O2.5 cm % !2.5 cm % Grade O grade III % ! grade III % COX-2 and both membrane and nuclear HER2 was observed (PZ0.0013, PZ and PZ respectively, Table 1). Associations between expression of HER2, phospho-erk1/2, PEA3 and COX-2, and clinical variables in breast cancer patients Associations between the qualitative expression of membrane and nuclear HER2, phospho-erk1/2, PEA3 and COX-2 and clinicopathological parameters were examined (Table 2). Nuclear expression of HER2 and phospho-erk1/2 associated with tumour size (PZ and PZ0.006), whereas HER2 both at the membrane and in the nucleus was found to associate with tumour grade (PZ0.03 and PZ respectively). Tumours that expressed COX-2 were associated with lymph node positivity (PZ0.0448). HER2 and COX-2 expression and endocrine response The ability of HER2, phospho-erk1/2, PEA3 and COX-2 to act as prognostic factors or to predict response to endocrine therapy was examined. In patients who received no adjuvant treatment following primary surgery, there was no significant association between any of the markers and disease-free survival detected (Table 3; Fig. 2A (i iii)). In patients who received tamoxifen following initial surgery, expression of membrane HER2, nuclear HER2 and COX-2 was each associated with reduced disease-free survival (PZ0.0017, PZ and PZ respectively) (Fig. 2A (iv vi)). Furthermore, in the total patient population, co-expression of COX-2 and nuclear HER-2 significantly increased the rate of recurrence, compared with patients who expressed COX-2, but not nuclear HER-2 (PZ0.0289) (Fig. 2B). However, in patients on tamoxifen treatment, co-expression of COX-2 with any of the markers examined did not impact on disease-free survival compared with COX-2 alone. Association between PEA3 and poor disease-free survival on endocrine treatment approached significance (PZ0.0672), whereas no association between expression of phospho-erk1/2 and response to treatment was observed (Table 3). As expression of membrane HER2, nuclear HER2 and COX-2 all associated with reduced disease-free survival in patients who received adjuvant tamoxifen, we examined the impact of treatment on patients who were positive for each of these markers. Although patients who expressed either membrane HER2 or nuclear HER2 had a reduced disease-free survival when treated with tamoxifen compared with patients who received no treatment according to Kaplan Meier estimates of survival, these data did not reach statistical significance (PZ0.331 and PZ0.143 respectively) (Fig. 3A and B). However, in patients who were positive for COX-2, those who were treated with tamoxifen, did significantly worse compared with patients who received no adjuvant endocrine treatment (PZ0.0427) (Fig. 3C). Table 3 Correlation of HER2 membrane (HER2m), HER2 nuclear (HER2n), phospho-erk1/2, PEA3 and cyclooxygenase-2 (COX-2) expression with disease-free survival using Wilcoxon test Time to recurrence no endocrine treatment P value Time to recurrence patients treated with tamoxifen P value HER2m HER2n ERK1/ PEA COX

6 M F Dillon et al.: HER2 and COX-2 in endocrine-resistant breast cancer Figure 2 Kaplan Meier estimates of disease-free survival (DFS). (A) DFS of breast cancer patients who received no endocrine therapy according to HER2 membrane (i), HER2 nuclear (ii) and COX-2 (iii) expression and corresponding patient populations treated with tamoxifen (iv vi). (B) DFS according to HER2 nuclear expression in COX-2-positive breast tumour patients. Discussion COX-2 has been shown to participate in tumour development and progression. Its actions are mediated, at least in part, by increased angiogenesis and CD44- induced invasion (Tsujii et al. 1998, Hiscox et al. 2006). In this clinical ex vivo investigation, we examined the contribution of COX-2 to tumour progression and resistance to endocrine treatment and investigated the role of HER2 in regulating COX-2 expression. Clinical and molecular studies provide a strong link between the tyrosine kinase receptor HER2 signalling and regulation of COX-2 expression. In this study, we found strong associations between expression of the membrane receptor and COX-2 in breast tumour tissue. 750 Evidence from in vitro work suggests that HER2 induces COX-2 expression via activation of ERK1/2, JNK and p38 (Subbaramaiah et al. 2002). Transcriptional upregulation of the COX-2 gene by HER2 expression has been found in cultured cancer cells (Howe et al. 2001b, Subbaramaiah et al. 2002). In HER2-transformed human mammary epithelial cells, HER2 has been shown to regulate COX-2 expression via MAP kinase-induced activation of the transcriptional regulator PEA3 (Subbaramaiah et al. 2002). Clinical studies, however, in a cohort of stage I breast cancers failed to detect a relationship between COX-2 expression and activated AKT, ERK1/2, p38 and HER2 (Schmitz et al. 2006). In this ex vivo study, we examined the association between HER2 and

7 Figure 3 Kaplan Meier estimates of disease-free survival (DFS). DFS of breast cancer patients with and without tamoxifen treatment according to (A) HER2 membrane, (B) HER2 nuclear and (C) COX-2 expression. COX-2 regulation in a large breast cancer patient population. Although we found strong associations between membrane HER2 and the transcription factor PEA3, and its putative target COX-2, no association between HER2 and its second messenger signalling mediator phospho-erk1/2 was observed. Furthermore, there was no relationship detected between PEA3 and COX-2. These data suggest that in the clinical setting HER2-induced PEA3 regulation of COX-2 may not be important. However, although no association was observed immunohistochemically, this may not reflect mechanisms operating at a subcellular level that could have some clinical implications. Although the function of HER proteins as transmembrane growth factor receptors has been well documented, a nuclear function as a transcription factor is rapidly being established. Recent molecular studies provide evidence that HER2 can directly regulate COX-2 expression through transcriptional induction (Vadlamundi et al. 1999, Wang et al. 2004). The significance of this nuclear role for HER2 in terms of its ability to directly regulate COX-2 in a clinical setting and to promote tumour progression has not been investigated. In line with our findings for membrane HER2, no association was detected between nuclear HER2 and phospho-erk1/2. Very strong associations were observed, however, between nuclear HER2 and 751

8 M F Dillon et al.: HER2 and COX-2 in endocrine-resistant breast cancer COX-2 expression. The association was substantially greater than that detected between membrane HER2 and COX-2. These observations support a function for HER2 as a transcription factor important in the regulation of COX-2 in breast cancer patients. Several clinical studies have reported a role for COX-2 in tumour progression and reduced disease-free survival (Ristimaki et al. 2002, Denkert et al. 2003, Wülfing et al. 2003). Many of these investigations, however, did not take into account the influence of adjuvant endocrine therapies. Despite the lack of clinical data relating COX-2 expression to response to hormonal therapy, several studies provide indirect evidence of a role for the cyclooxygenase in endocrine resistance. Prostaglandins, in particular PGE2, enhance stromal aromatase production, suggesting crosstalk between the cyclooxygenase and steroid pathways (Zhao et al. 1996, Brueggemeier et al. 1999). Here, we stratified our patient population into those who received no adjuvant endocrine treatment and those who received tamoxifen for a 5-year period. The ability of HER2, phospho-erk1/2, PEA3 and COX-2 to act as either prognostic factors or to predict response to endocrine treatment was assessed. None of the markers examined were found to be independently prognostic. Membrane HER2, nuclear HER2 and COX-2, however, were all found to predict poor disease-free survival in patients on endocrine treatment. Although strong associations between COX-2 and nuclear HER2 in our total patient population suggest that nuclear HER2 may have a role to play in the production of COX-2 in breast cancer patients, in endocrine-treated patients, co-expression of COX-2 with nuclear HER2 did not impact on disease-free survival compared with COX-2 alone. These data suggest that expression of COX-2 in the endocrine setting may be regulated by alternative mechanisms. In determining the consequence of treatment with endocrine therapy in patients who express membrane HER2, nuclear HER2 and COX-2, we found that COX-2 positivity can predict an adverse effect of tamoxifen. In summary, data presented in this ex vivo study supports a role for HER2 in the direct transcriptional regulation of COX-2 in human breast cancer. COX-2 positivity in our patient population predicts adverse effects of hormonal therapy. Declaration of interest The authors declare that there is no conflict of interest that would prejudice the impartiality of this work. 752 Funding This research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector. Acknowledgements The authors wish to thank Prof. Niall O Higgins, St Vincent s University Hospital, Dublin, for his invaluable clinical expertise. Irish Foundation for Breast Diseases and Breast Cancer Ireland. References Benz CC, O Hagan RC, Richter B, Scott GK, Chang CH, Xiong X, Chew K, Ljung BM, Edgerton S, Thor A et al HER2/Neu and the Ets transcription activator PEA3 are coordinately upregulated in human breast cancer. Oncogene Brueggemeier RW, Quinn AL, Parrett ML, Joarder FS, Harris RE & Robertson FM 1999 Correlation of aromatase and cyclooygenase gene expression in human breast cancer specimens. Cancer Letters Denkert C, Winzer KJ, Müller BM, Weichert W, Pest S, Köbel M, Kristiansen G, Reles A, Siegert A, Guski H et al Elevated expression of cyclooxygenase-2 is a negative prognostic factor for disease free survival and overall survival in patients with breast carcinoma. Cancer Dohadwala M, Luo J, Zhu L, Lin Y, Dougherty GJ, Sharma S, Huang M, Pold M, Batra RK & Dubinett SM 2001 Non-small cell lung cancer cyclooxygenase-2-dependent invasion is mediated by CD44. Journal of Biological Chemistry Harvey JM, Clark GM, Osborne CK & Allred DC 1999 Estrogen receptor status by immunohistochemistry is superior to the ligand-binding assay for predicting response to adjuvant endocrine therapy in breast cancer. Journal of Clinical Oncology Hiscox S, Jiang WG, Obermeier K, Taylor K, Morgan L, Burmi R, Barrow D & Nicholson RI 2006 Tamoxifen resistance in MCF7 cells promotes EMT-like behaviour and involves modulation of beta-catenin phosphorylation. International Journal of Cancer Howe LR, Howard CC, Subbaramaiah K, Hassell JA, Dannenberg AJ & Brown AMC 2001a PEA3 is up-regulated in response to Wnt1 and activates the expression of cyclooxygenase-2. Journal of Biological Chemistry Howe LR, Subbaramaiah K, Brown AM & Dannenberg AJ 2001b Cyclooxygenase-2: a target for the prevention and treatment of breast cancer. Endocrine-Related Cancer

9 Hwang D, Scollard D, Byrne J & Levine E 1998 Expression of cyclooxygenase -1 and cyclooxygenase-2 in human breast cancer. Journal of the National Cancer Institute Liu CH, Chang SH, Narko K et al Overexpression of cyclooxygenase-2 is sufficient to induce tumrigenesis in transgenic mice. Journal of Biological Chemistry Moasser MM 2007 The oncogene HER2: its signalling and transforming functions and its role in human cancer pathogenesis. Oncogene Myers E, Hill AD, Kelly G, McDermott EW, O Higgins NJ & Young LS 2006 positive role for PEA3 in HER2- mediated breast tumour progression. British Journal of Cancer Nassar A, Radhakrishnan A, Cabrero IA, Cotsonis G & Cohen C 2007 COX-2 expression in invasive breast cancer: correlation with prognostic parameters and outcome. Applied Immunohistochemistry and Molecular Morphology Ristimaki A, Sivula A, Lundin J, Lundin M, Salminen T, Haglund C, Joensuu H & Isola J 2002 Prognostic significance of elevated cyclooxygenase-2 expression in breast cancer. Cancer Research Ross JS & Fletcher JA 1998 The HER-2/neu oncogene in breast cancer: prognostic factor, predictive factor, and target for therapy. Oncologist Schmitz KJ, Callies R, Wohlschlaeger J, Kimming R, Otterbach F, Bohr J, Lee SH, Takeda A, Schmid RW & Baba HA 2006 Overexpression of cyclooxygenase-2 is an independent predictor of unfavourable outcome in nodenegative breast cancer, but is not associated with protein kinase B (Akt) and mitogen-activated protein kinase (ERK1/2, p38) activation or with HER2/neu signalling pathways. Journal of Clinical Pathology Sheng H, Shao J, Washington MK & DuBois RN 2001 Prostaglandin E2 increases growth and motility of colorectal carcinoma cells. Journal of Biological Chemistry Shepherd TG, Kockeritz L, Szrajber MR, Muller WJ & Hassell JA 2001 The pea3 subfamily ets genes are required for HER2/Neu-mediated mammary oncogenesis. Current Biology Simeone AM, Li YJ, Broemeling LD, Johnson MM, Tuna M & Tari AM 2004 Cyclooxygenase-2 is essential for HER2/neu to suppress N- (4-hydroxyphenyl) retinamide apoptotic effects in breast cancer cells. Cancer Research Subbaramaiah K, Norton L, Gerald W & Dannenberg AJ 2002 Cyclooxygenase-2 is overexpressed in HER-2/neupositive breast cancer. Journal of Biological Chemistry Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M & DuBois RN 1998 Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell Vadlamundi R, Mandal M, Adam L, Steinbach G, Mendelsohn J & Kumar R 1999 Regulation of cyclooxygenase-2 pathway by HER2 receptor. Oncogene Wang SC, Lein HC, Xia W, Chen IF, Lo HW, Wang Z, Ali-Seyed M, Lee DF, Bartholomeusz G, Ou-Yang F et al Binding at and transcription activation of COX-2 promoter by nuclear tyrosine kinase receptor ErbB-2. Cancer Cell Wülfing P, Diallo R, Müller C, Wülfing C, Poremba C, Heinecke A, Rody A, Greb RR, Böcker W & Kiesel L 2003 Analysis of cyclooxygenase-2 expression in human breast cancer: high throughput tissue microarray analysis. Journal of Cancer Research and Clinical Oncology Xie Y & Hung MC 1994 Nuclear localisation of p185neu tyrosine kinase and its association with transcriptional activation. Biochemical and Biophysical Research Communications Zhao Y, Agarwal VR, Mendelson CR & Simpson ER 1996 Estrogen biosynthesis proximal to a breast tumour is stimulated by PGE2 via cyclic AMP, leading to activation of promoter II of the CYP19 armoatase gene. Endocrinology