Signal transducers and activators of transcription (STATs) proteins function in transducing signals from the cell membrane to the nucleus and, eventually, activating transcription of genes involved in a wide range of basic cellular functions, including differentiation, proliferation, cell survival, apoptosis and angiogenesis.1, 2 STAT proteins are activated by several soluble factors, including cytokines, growth factors and hormones. When these ligands bind to their specific transmembrane receptor, they usually lead to Janus tyrosine kinases (JAKs)-mediated tyrosine phosphorylation and activation of STATs.3 Upon ligand binding, the intracellular tail of receptor is tyrosine-phosphorylated by activated JAKs and provides docking sites for STATs, which are then tyrosine-phosphorylated by JAKs.4
STAT proteins can be divided into two functional groups.2 One group (STAT2, 4 and 6) is activated by certain cytokines and plays a unique role in the development of T-cells and IFNγ signaling. The other group (STAT1, 3 and 5) is activated by various ligands and is involved in IFN signaling, mammary gland development and embryogenesis. This second group controls cell cycle progression and apoptosis, thus playing an important role in carcinogenesis.2, 5 STAT1 works as a potential tumor suppressor, active in arresting cell growth and promoting apoptosis.4 On the other hand, STAT3 and 5 work as oncoproteins by promoting cell-cycle progression, cellular transformation and preventing apoptosis, after their activation of cyclin D1, c-myc and Bcl-xL expression.4
Inactivation of STAT3 in mice leads to embryonic lethality.6 On the other hand, constitutive activation of STAT3 is required for either enhancing transformation or blocking apoptosis in cancer cell lines and in human cancers, particularly breast cancer cell lines and tissues.7, 8, 9, 10 Using electrophoretic mobility-shift assay, it has also been observed that DNA binding of STAT3 in breast cancer tissues is significantly higher than in normal breast tissues.11, 12 STAT3 is activated through tyrosine phosphorylation at residue 705, in response to growth factors, hormones and cytokines.1 Tyrosine phosphorylation at residue 705 facilitates STAT3 dimerization1 and translocation to the nucleus, where they bind to the specific DNA response elements and enable gene transcription.13 A second phosphorylation event at serine residue 727 modulates the transcriptional activity of STAT3, and is required for maximal transcriptional activity.13, 14, 15, 16, 17 This event is related to the activation of MAPK family members as well as several other signal transduction components, such as Vav, Rac-1 and SEK-1/MKK-4, which are dependent on cell context and the stimulus used.18, 19, 20 It is indicated that serine727 phosphorylation is not required for DNA binding activity of STAT3.15, 16 However, there is also a report showing that leptin-induced ERK activation in J774.2 macrophages corresponded with an increase in both phosphorylation of Ser727 and STAT3 DNA binding activity.21
Although it is undetermined whether serine phosphorylation is an event dependent on tyrosine phosphorylation, phosphorylation at serine residue 727 of STAT3 may contribute to a full STAT3 activity.13, 14, 15, 16, 17 For example, serine phosphorylation of STAT3 is essential for postnatal survival and growth, since knock-in of STAT3SA cDNA, which replaces serine residue 727 with an alanine, into STAT3 knockout mice fails to compensate the phenotype.22 In addition, STAT3β, a truncated form of STAT3, which is without the serine residue 727-containing C-terminus, works as a negative regulator of STAT3-mediated activity in breast cancer cells.23, 24, 25
In this study, immunoblotting approaches were used to analyze the expression patterns of p-ser727-STAT3 in breast infiltrating ductal carcinoma tissues, and the results were further correlated with clinicopathological characteristics. Furthermore, the effects of ER status on the expression of p-ser727-STAT3 in breast cancer cell lines were analyzed.
Material and methods
Tissue samples and cell culture
Sixty-eight newly identified breast infiltrating ductal carcinoma cases diagnosed from 2000 to 2003 were included in this study. Tissue specimens were obtained from patients during surgical treatment at the Department of Surgery, Kaohsiung Medical University Hospital. Specimens were frozen immediately in liquid N2 and stored at −80°C until further use.
Human breast cancer cell line MCF-7 was purchased from American Type Culture Collection. ER-positive human breast cancer cell lines (T47D and ZR75-1) and ER-negative human breast cancer cell lines (MDA-MB-231, SKBR-3 and MDA-MB-468), kindly provided by Dr. M.C. Hung, were grown in DMEM/F12 medium supplemented with 10% fetal bovine serum, 100 U/ml penicillin and 100 μg/ml streptomycin. Tamoxifen, 17β-estradiol and epidermal growth factor were ordered from Sigma.
Frozen breast tissues were ground and dissolved in EBC buffer (50 mM Tris pH 7.6, 120 mM NaCl, 0.5% Nonidet P-40, 1 mM EDTA, 1 mM 2-mercaptoethanol, 50 mM NaF and 1 mM Na3VO4). Detailed procedures for immunoblotting were followed accordingly.26 Rabbit polyclonal anti-p-ser727-STAT3 antibody and mouse monoclonal anti-STAT3 antibody F-2 was ordered from Santa Cruz, and mouse monoclonal anti-p-tyr705-STAT3 antibody was obtained from Cell Signaling. The AlphaImager™ 1220 documentation and analysis system (Alpha Innotech, San Leandro, CA) was used to compare expression levels of p-ser727-STAT3 in breast cancer tissues and the paired noncancer breast tissues, from the same patient, after normalization by respective total STAT3. Notably, no significant difference was observed in the expression levels of total STAT3 between the paired cancer and noncancer breast tissues, after normalization by β-actin. C > N was defined as the ratio of p-ser727-STAT3 to total STAT3 at least 50% higher in cancer tissue than in paired noncancer tissue. Similarly, C < N was defined as the ratio of p-ser727-STAT3 to total STAT3 in noncancer tissue at least 50% higher than in paired cancer tissue. CN was defined as a less than 50% difference in the ratio of p-ser727-STAT3 to total STAT3 between 2 paired tissues.
Immunohistochemical protocol was followed accordingly.27 In brief, the tissue samples were fixed with 10% buffered formalin, then dissected, dehydrated and coated with wax. Samples were sliced to 4 μm, and then either dyed with hematoxylin–eosin or immunostained with the first (mouse monoclonal antiestrogen receptor, antiprogesterone receptor and anti-Her2/Neu antibody from DAKO; rabbit polyclonal anti-p-ser727-STAT3 antibody from Santa Cruz), followed by Universal LAB + kit/HRP (DAKO), or counterstained with hematoxylin. The results were captured by the Nikon E-800M microscope, and then processed by Kodak MGDS330 and Adobe Photoshop 6.0.
All of the statistical analyses were performed using the SPSS 10.0 statistical package for PC (SPSS, Chicago, IL). Grouping results (C > N, CN, C < N) were correlated with the clinicopatholoical parameters, using Student t-test (two-tailed), Pearson's rank correlation coefficients and Spearman's rank correlation coefficients. The binary significance was further examined by Fisher's exact test. Two-tailed p ≤ 0.05 was considered statistically significant.
Small interfering RNA (siRNA) transient transfection
The siRNA kit for knocking down ERα message was obtained from Santa Cruz. The sequences for ERα-specific and nonspecific siRNA used in this study were not released by the manufacturer. MCF-7 cells were grown in 6-well plates and were at 60% confluency when transfected with 100 nM ERα-specific siRNA or control siRNA, by using LipofectAMINE 2000 reagent (Invitrogen). The transfected cells were grown in DMEM supplemented with 10% FBS for 48 hr and then harvested for immunoblotting analysis.
Expression patterns of p-ser727-STAT3 in breast cancer tissues
Clinicopathological characteristics of breast cancer patients included in this study are shown in Table I. Average age of breast cancer patients is 51.4 ± 14.2 years, ranging from 31 to 83 years, and the average tumor size of breast cancer cases is 4.0 ± 1.9 cm, ranging from 1.3 to 10 cm. To investigate the role of STAT3 serine phosphorylation in the tumorigenesis of breast infiltrating ductal carcinoma, immunoblotting analysis was first applied to study the expression pattern of p-ser727-STAT3 in the paired tissue specimens (Fig. 1a). Although the expression of total STAT3 protein was similar between the breast infiltrating ductal carcinoma tissues and their paired noncancer breast tissues, 61.8% (42/68) of cancer tissues had higher p-ser727-STAT3 expression levels than the adjacent noncancer breast tissues, whereas 10.3% (7/68) had similar p-ser727-STAT3 levels and 27.9% (19/68) had lower p-ser727-STAT3 levels (p < 0.001) (Fig. 1a). Further immunohistochemistry study showed an elevated and intense nuclear staining as well as a diffuse ctyoplasmic staining of p-ser727-STAT3 in the cancer lesions, as compared to the matched noncancer breast tissues (Fig. 1b). Also, a weak to absent p-ser272-STAT3 expression was observed in the neighboring stromal cells, except reactive fibroblasts. These results suggest that the elevated levels of p-ser727-STAT3 may play a role in the carcinogenesis of breast infiltrating ductal carcinoma.
Table I. Clinicopathological Characteristics of Breast Cancer Patients
The correlation between p-ser727-STAT3 expression and clinicopathological parameters
When the expression patterns of p-ser727-STAT3 in breast cancer tissues were correlated with clinicopathological parameters, a positive correlation was observed between the p-ser727-STAT3 expression and cancer stage (Table II, p = 0.024) as well as tumor size (Fig. 2, p = 0.014). The tumor size (mean ± SEM) of the high p-ser727-STAT3 (C > N) expression subgroup (n = 42) and the low p-ser727-STAT3 (C ≦ N) expression subgroup (n = 26) were 4.44 ± 1.94 and 3.29 ± 1.48 cm, respectively (Fig. 2). Interestingly, the expression levels of p-ser727-STAT3 in breast cancer tissues were inversely correlated to estrogen receptor (ERα) status (r = −0.391, p = 0.001), but not progesterone receptor (PR) status or Her2/neu status (Table II). The significant association of p-ser727-STAT3 expression with cancer stage, tumor size and ER status suggests that high expression of p-ser727-STAT3 in cancer lesions may be a useful biomarker for a poor prognosis in breast infiltrating ductal carcinoma cases.
Table II. Correlation of P-SER727-STAT3 Expression with Clinicopathological Characteristics.
Analysis of the expression levels of p-ser727-STAT3 in ER-positive and ER-negative breast cancer cell lines
The significant correlation between the expression levels of p-ser727-STAT3 and ERα status in the breast cancer tissues raised the question that whether the expression levels of p-ser727-STAT3 were regulated by ERα. To address this question, different breast cancer cell lines (with and without ER expression) and an immortalized breast epithelial cell line were analyzed for their expression levels of p-ser727-STAT3 by immunoblotting (Fig. 3a). Surprisingly, all the ER-negative breast cancer cell lines we analyzed (MDA-MB-231, SKBR-3 and MDA-MB-468) showed higher p-ser727-STAT3 expression levels than ER-positive cell lines (MCF-7, T47D and ZR75-1) (Fig. 3a). However, the significant correlation between p-ser727-STAT3 and ER in breast cancer cell lines was not observed between p-tyr705-STAT3 and ER (Fig. 3a). A transient transfection of ER-specific siRNA increased p-ser727-STAT3 as well as p-tyr705-STAT3 expression in MCF-7 cells as compared with the mock (no treatment) or nonspecific siRNA treatment, and treatment with ER-specific or control siRNA did not affect total STAT3 expression (Fig. 3a). In addition, tamoxifen, a selective estrogen receptor modulator (SERM), caused a dose-dependent decrease of p-ser727-STAT3 as well as p-tyr705-STAT3 expression, whereas 17β-estradiol (E2) leaded to an increase of p-ser727-STAT3 and p-tyr705-STAT3 expression in ER-positive MCF-7 cells (Fig. 3b). However, no significant effect on the p-ser727-STAT3 or p-tyr705-STAT3 expression was observed in ER-negative MDA-MB-231 cells, upon tamoxifen treatment (Fig. 3c).
In this study, a significant increase of p-ser727-STAT3 expression in breast infiltrating ductal carcinoma tissues, as compared to the matched noncancer breast tissues was observed (Figs. 1a and 1b). Since ser727 phosphorylation in the C-terminal region of STAT3 is required for transactivation by association with p300/CREB-binding protein in HepG2 cells,28 it is theoretically possible that p-ser727 phosphorylation leads to a stronger STAT3 transcriptional activity through its constitutive association with coactivators and then promotes breast cancer oncogenesis. Alternatively, constitutive p-ser727-STAT3 expression in breast cancer tissue may represent a persistent activation of MAPK family members, which are responsible for STAT3 serine727 phosphorylation in primary breast tumor.29, 30, 31
Ligand-activated estrogen receptor is known to induce tyrosine and serine phosphorylation of STAT3 as well as the transcriptional activation of STAT3-regulated promoters.32 In addition, increased STAT3 activation up-regulates VEGF expression and leads to tumor angiogenesis.33, 34 In support of these findings, our study demonstrated the significant correlation between p-ser727-STAT3 expression and negative estrogen receptor (ERα) status, advanced cancer stage as well as increased tumor size in breast infiltrating ductal carcinoma cases (Table II and Fig. 2), suggesting that constitutive p-ser727-STAT3 expression may play a positive role in tumor progression. We also showed that ER-negative breast cancer cell lines had a higher p-ser727-STAT3 expression than ER-positive cell lines (Fig. 3a), as demonstrated in primary breast tumors (Table II). Also, knockdown of ERα expression increased p-ser727-STAT3 expression in ER-positive MCF-7 cells, as aforementioned in primary breast tumors (Table II) (Fig. 3a). It is possible that ERα can modulate, indirectly, serine phosphorylation of STAT3 by a crosstalk with other phosphorylation cascades,35, 36 and thus lead to tumor progression in breast cancer. On the other hand, a direct protein–protein interaction has been observed between STAT molecules and ER37 as well as PR.38, 39 Therefore, alternatively, ERα may inhibit STAT3 serine727 phosphorylation via a direct interaction with STAT3 as reported by a recent study, which demonstrates that ER can repress PRL-induced STAT5 transcriptional activity through its direct interaction with STAT5.40
In this study, we showed that tamoxifen treatment reduced p-ser727-STAT3 and p-tyr705-STAT3 expression in ER-positive MCF-7 cells (Fig. 3b), whereas no significant effect on p-ser727-STAT3 and p-tyr705-STAT3 expression was observed in ER-negative MDA-MB-231 cells upon tamoxifen treatment (Fig. 3c), suggesting that the anticancer activity of tamoxifen against ER-positive breast cancer cases might be through its direct binding and inactivation of ERα and, subsequently, down-regulation of p-tyr705-STAT3 and p-ser727-STAT3 expression. On the other hand, 17β-estradiol treatment increased p-tyr705-STAT3 and p-ser727-STAT3 expression in ER-positive MCF-7 cells, without altering STAT3 expression (Fig. 3b). Therefore, we provide additional information that a functional talk between STAT3 and estrogen receptor may play a role in the development of breast cancer. It should be noted that progesterone treatment of PR-positive T47D breast cancer cells also regulates the tyrosine phosphorylation of STAT3 in a PR-dependent manner,41 and STAT3 expression remains to be unaffected in progesterone-treated PR-transfected MDA-MB-231 cells.42 Additionally, although many studies show that tyrosine and serine phosphorylation of STAT3 are costimulated by certain ligands,13, 20, 21 some studies disagree.15, 19 In the present study, we did not observe any significant, either positive or negative, correlation between p-ser727-STAT3 and p-tyr705-STAT3 expression in primary breast cancer (data not shown). Besides, the expression levels of p-tyr705-STAT3 in cancer tissues were also not correlated to cancer stage, tumor size and ERα status, which were previously correlated to p-ser727-STAT3 expression in our study (data not shown).
In summary, using a cohort of 68 primary breast infiltrating ductal carcinomas, we found a positive correlation between p-ser727-STAT3 level and tumor stage/size. The expression of p-ser727-STAT3 was also negatively correlated with ER status, as demonstrated in breast cancer cell lines and in primary tumors. In addition, a possible crosstalk between STAT3 and estrogen receptor may provide an explanation, at least partially, why tamoxifen is ineffective in the treatment of ER-negative breast cancer cases. Further investigations are required to better understand the detailed mechanisms of STAT3 signaling in breast cancer development.