Advertisement
Human Cancer Biology

Activation of Stat3 in Primary Tumors from High-Risk Breast Cancer Patients Is Associated with Elevated Levels of Activated Src and Survivin Expression

Nills Diaz, Susan Minton, Charles Cox, Tammy Bowman, Tanya Gritsko, Roy Garcia, Ibrahim Eweis, Marek Wloch, Sandy Livingston, Ed Seijo, Alan Cantor, Ji-Hyun Lee, Craig A. Beam, Daniel Sullivan, Richard Jove and Carlos A. Muro-Cacho
DOI: 10.1158/1078-0432.CCR-04-1749 Published January 2006

Abstract

Purpose: Constitutive activation of signal transducer and activator of transcription 3 (Stat3) protein has been observed in a wide variety of tumors, including breast cancer, and contributes to oncogenesis at least in part by prevention of apoptosis. In a study of 45 patients with high-risk breast cancer enrolled in a phase II neoadjuvant chemotherapy trial with docetaxel and doxorubicin, we evaluated the levels of Stat3 activation and potentially associated molecular biomarkers in invasive breast carcinoma compared with matched nonneoplastic tissues.

Experimental Design: Using immunohistochemistry and image analysis, we quantified the levels of phospho-Stat3 (pY-Stat3), phospho-Src (pY-Src), epidermal growth factor receptor, HER2/neu, Ki-67, estrogen receptor, Bcl-2, Bcl-xL, Survivin, and apoptosis in formalin-fixed, paraffin-embedded sections from invasive carcinomas and their paired nonneoplastic parenchyma. The levels of molecular biomarkers in nonneoplastic and tumor tissues were analyzed as continuous variables for statistically significant correlations.

Results: Levels of activated pY-Stat3 and pY-Src measured by immunohistochemistry were significantly higher in invasive carcinoma than in nonneoplastic tissue (P < 0.001). In tumors, elevated levels of pY-Stat3 correlated with those of pY-Src and Survivin. Levels of pY-Stat3 were higher in partial pathologic responders than in complete pathologic responders. In partial pathologic responders, pY-Stat3 levels correlated with Survivin expression.

Conclusions: Our findings suggest important roles for elevated activities of Stat3 and Src, as well as Survivin expression, in malignant progression of breast cancer. Furthermore, elevated Stat3 activity correlates inversely with complete pathologic response. These findings suggest that specific Stat3 or Src inhibitors could offer clinical benefits to patients with breast cancer.

  • Signal transduction
  • neoadjuvant chemotherapy
  • molecular biomarkers

Identification of molecular biomarkers in breast cancer that could help in a more accurate assessment of individual disease risk and prognosis, and in the development of target-specific therapies, is of significant interest. Abnormalities in signal transduction pathways and their biological functions are recognized as being responsible for a growing number of pathologic conditions, including neoplasia. In normal cells, transmission of signals from the cell surface to the nucleus depends on tightly regulated signal transduction pathways (1, 2). The signal transducer and activator of transcription (STAT) proteins are a family of transcription factors originally identified as key components of IFN signaling pathways and subsequently found to mediate responses to a variety of cytokines and growth factors (3–5). Stimulation of cell surface receptors with their cognate ligands results in activation of receptor-intrinsic tyrosine kinase activity and/or activation of receptor-associated tyrosine kinases, such as the Janus kinases or Src family kinases. Activation of these tyrosine kinases, in turn, induces the formation of tyrosine-phosphorylated STAT dimers that translocate to the nucleus and bind to specific STAT DNA response elements, thereby directly regulating gene expression (3–5). Normal STAT activation is a transient and tightly regulated process. However, persistent activation of one STAT family member, Stat3, and to a lesser extent Stat5, has been observed with unexpectedly high frequency in a variety of tumors, including myeloma, leukemia, lymphoma, melanoma, and carcinomas from prostate, ovary, and head and neck (for reviews, see refs. 6–12). Numerous studies have revealed that Stat3 contributes to oncogenesis by several mechanisms, including inhibition of apoptosis, enhancement of cell proliferation, induction of angiogenesis, and suppression of immune responses (12).

In the present study, we assessed the levels of activated, tyrosine-phosphorylated Stat3 (pY-Stat3), and other proteins of potential interest to oncogenic signaling in 45 patients with stage III breast carcinoma. We also investigated changes in the expression of these molecular biomarkers in relation to pathologic response in a phase II neoadjuvant chemotherapy trial of sequential doxorubicin and docetaxel treatment. Consistent with a role for activated Stat3 in breast cancer progression, elevated Stat3 activity was detected in tumors compared with matched nonneoplastic tissues. Levels of activated Stat3 in patients who had a complete pathologic response were significantly lower than those of patients who had a partial pathologic response, indicating an inverse correlation of constitutive Stat3 activity with chemotherapy response. In patients who had partial pathologic responses, elevated levels of activated Stat3 correlated with the expression of Survivin, a member of the inhibitors of apoptosis protein family (13–16). These findings are consistent with a role of constitutively active Stat3 in mediating breast tumor cell survival and response to therapy, in part by induction of Survivin (see Gritsko et al., in this issue). Furthermore, levels of activated Src kinase (pY-Src) were significantly higher in tumors compared with matched nontumor tissues, consistent with a role of activated Src kinase in breast cancer progression. Together, our findings suggest that breast cancer patients may benefit from therapies targeting Src kinase and Stat3 signaling, either as standalone therapy or in combination with conventional chemotherapy.

Materials and Methods

Clinical trial and biomarkers. Forty-five women with stage III breast carcinoma were enrolled in a 3-year clinical trial of neoadjuvant dose-dense chemotherapy with sequential doxorubicin (80 mg/m2) followed by docetaxel (100 mg/m2) i.v. every 2 weeks for three cycles each.6 After neoadjuvant chemotherapy, all participants underwent definitive surgery with either lumpectomy or mastectomy and axillary lymph node dissection. All tumors were at least 5 cm in size, 85% were ductal, 10% were lobular, and 5% had ductal and lobular features. We quantified levels of activated tyrosine-phosphorylated Stat3 (pY-Stat3) and nine other proteins were selected for their reported relevance to STAT activation or breast oncogenesis: pY-Src (17, 18), HER2/neu (19), estrogen receptor (ER; ref. 20), progesterone receptor (PR; ref. 21), Ki-67 (22), Bcl-2 (23, 24), Bcl-xL (24), epidermal growth factor receptor (EGFR) (25), and Survivin (16). Apoptosis in tissues was measured by the terminal deoxynucleotidyl transferase–mediated nick end labeling (TUNEL) assay (26). In an attempt to identify expression patterns that correlate with response to treatment, the analysis was done in tissues obtained before and after therapy. Before treatment, tissue was obtained by incisional biopsies of tumor and nontumoral parenchyma, either from a distant ipsilateral quadrant or from the contralateral breast (majority of cases). After chemotherapy, tissue was obtained at the time of definitive surgery. The clinical protocol was approved by the Institutional Review Board and written informed consent was obtained from every patient before participation in this study.

Tissue collection. Our previous studies (18) had shown that to preserve the activated phosphorylation state of signal transduction proteins, tissues have to be snap frozen in liquid nitrogen within 15 minutes from the moment of interruption of blood supply to the specimen. All tissues in this study were snap frozen or fixed in 10% neutral-buffered formalin within 15 minutes to minimize antigen loss and optimize immunohistochemical detection. The presence of normal tissue or tumor was confirmed in mirror image sections of the respective samples by examination of frozen sections immediately following collection. After chemotherapy, tumor was only available from those patients with a partial pathologic response. All data obtained were entered into a web-based database for statistical correlation of clinical, pathologic, and molecular data. Patient data were deidentified and maintained confidential by assigning each case a research specimen number.

Heterogeneity of signal across tissue sections. An important aspect of the initial experimentation process was to determine the degree of staining variability across consecutive tissue sections. Twenty consecutive sections were prepared without discarding intervening tissue and used to perform imunohistochemistry for a signaling protein not related to the project (transforming growth factor receptor type II). Quantitative image analysis revealed minimal variation in expression intensity for the first 12 consecutive sections (data not shown). Expression levels on the next set of 12 sections showed statistically significant differences when compared with the first set. Therefore, for the purposes of our study, a maximum of 11 consecutive sections were used.

Immunohistochemistry and TUNEL procedures. Consecutive 3 μm sections were prepared without discarding intervening tissue (see above). The first section was stained with H&E and the rest of the sections were used for immunohistochemistry and TUNEL assays. For all antigens, except pY-Stat3 (see procedure for pY-Stat3 detection below), the following procedure was used. Formalin-fixed, paraffin-embedded tissue sections were dried at 37°C overnight. Sections were deparaffinized by an initial warming to 60°C, followed by two xylene changes 10 minutes each, two series of 30 dips in absolute alcohol, 30 dips in 95% alcohol, and 20 dips in deionized water. Antigen retrieval or enzyme digestion procedures were done as described by the supplier of each antibody. Slides were placed for 5 minutes in TBS/Tween and processed on a DAKO Autostainer using the Dako LSAB+ peroxidase detection kit (DAKO, Carpinteria, CA). Endogenous peroxidase was blocked with 3% aqueous hydrogen peroxide followed by two 20 dips in deionized water. The anti-EGFR monoclonal antibody clone 111.6 (Signet Pathology Systems, Dedham, MA) was applied at 1:100 for 30 minutes following proteinase K digestion (25 μg/mL in TBS/Tween) for 17 minutes. The rest of the antibodies were applied for 30 minutes after microwave antigen retrieval with 0.1 mol/L citrate buffer (pH 6.0; Emerson 1,100 W microwave, high to boiling, then 20 minutes on power level 5) as follows: Bcl-2 (1:40; DAKO), Bcl-xL (1:50; Santa Cruz Biotechnology, Santa Cruz, CA), pY-Src (1:100; Cell Signaling Technology, Beverly, MA), Ki-67 (1:50; Immunotech, Norcross, GA), c-ErbB-2 (1:40; HER2/neu; Signet Pathology Systems), ER and PR (1:40; BioGenex, San Ramon, CA), and Survivin (1:100; Cell Signaling Technology). The chromogen 3,3′-diaminobenzidine was used for all proteins except for Survivin, which was detected using Nova-Red (Vector Laboratories, Burlingame, CA). Survivin expression was evaluated only in samples obtained before treatment because analysis of this antigen was added at a later time in this study based on microarray analyses (see Gritsko et al., in this issue). Counterstain was done with modified Mayer's hematoxylin. Slides were dehydrated through graded alcohol, cleared with xylene, and mounted with resinous mounting medium. Apoptosis was detected by the TUNEL assay using the Intergen Apopta G Peroxidase In situ Apoptosis detection kit as indicated by the supplier.

pY-Stat3 immunohistochemistry. After deparaffinizing, a two-stage pretreatment procedure was done as follows. First, antigen retrieval was done in a pressure cooker by placing a total of 600 mL deionized water in three containers, one containing the slides in citrate buffer (pH 6.0) and the other two containing only deionized water. The microwave oven was set on high to pressurize for 12 minutes and then at power level 4 for 10 minutes. Slides were then cooled at room temperature for 30 minutes. This was followed by limited enzymatic digestion at 37°C for 5 minutes with 0.025% trypsin in 5 mmol/L Tris-Cl (pH 7.6) with 0.05% calcium chloride. At the end of the digestion, slides were rinsed with deionized water, placed in TBS/Tween for 5 minutes, drained, and framed with an ImmunoEdge pen. Hydrogen peroxide 3% was applied for 10 minutes and 3% bovine serum albumin/PBS for 10 minutes. Sections were incubated with anti-phospho-Stat3 antibody (rabbit polyclonal P-Stat3, Cell Signaling) at 1:400 in a humid chamber at 4°C overnight and returned to the autostainer for detection and substrate development using the Dako LSAB+ detection system and 3,3′-diaminobenzidine as chromogen. Counterstaining was done for 30 seconds with modified Mayer's hematoxylin. Sections were allowed to sit in tap water for 10 to 15 minutes and dehydrated before mounting with resinous mounting medium.

Image analysis and quantification. The Optimas 6.5 (Media Cybernetics, Silver Springs, MD) software was used to quantify protein expression. Regions of interest were identified on the H&E-stained slide and the same areas were marked on the consecutive sections used for each of the biomarkers and TUNEL assay. Digital images of these areas were obtained using identical magnifications (×400) and camera settings with a Leica DM microscope (Leica Microscopes, Bannockburn, IL) with neutral density 6 and 12 filters, coupled to a SPOT Digital Camera System (Diagnostic Instruments, Sterling Heights, MI) and SPOT software set as AutoGain, RGB filter color, no binning, full chip area, and adjustment factor set to 1. Before photography, Koehler epi-illumination was done. Optimal light conditions for each objective were stored as software settings to be replicated in each measurement session. Image acquisition was done after 1 hour of microscope lamp warm up. White balance was done on an area of the slide with no tissue, and background values were subtracted using a negative control slide. Images of the selected areas were stored as TIFF images. A macro was specifically set in the software to automate the process and transfer mathematical calculations to a Microsoft Excel spreadsheet. These were converted to SAS data sets for statistical analysis. Quantitative image analysis was done on all pretreatment samples and in those posttreatment samples for which tumor tissue was still identifiable (partial pathologic responders).

HER2/neu clinical testing. HER2/neu status was also assessed in all tumors at an independent laboratory using a Food and Drug Administration–approved immunohistochemistry procedure. Results of this test are reported negative when intensity scores are 0 or 1+ and positive when the intensity score is 3+. Cases with intermediate intensity score of 2+ are further evaluated by fluorescence in situ hybridization. Using this method, 11 cases (24.4%) were positive, 31 (68.9%) were negative, and 3 (6.7%) remained undetermined. These percentages are consistent with previous reports (27). In the group of tumors that were positive by the Food and Drug Administration–approved test, the average HER2/neu index calculated by computerized image analysis in this study was 50 and the SD 14.5. In the group of tumors that were negative by the Food and Drug Administration–approved test, the average HER2/neu index by computerized image analysis was 35.5 and the SD was 30. The difference was not statistically significant but suggests a similar trend of HER2/neu detection with both methods.

Electrophoretic mobility shift assay. Details of this procedure are described in by Gritsko et al., in this issue. Briefly, a 32P-radiolabeled oligonucleotide probe containing a high-affinity cis-inducible element (hSIE) derived from the c-fos promoter that binds Stat3 proteins was incubated with nuclear extract. Protein-DNA complexes were resolved by nondenaturing PAGE and detected by autoradiography. Quantification of Stat3 DNA-binding activity was determined relative to internal control standards, which were serial dilutions of nuclear extract prepared from the MDA-MB-468 breast cancer cell line [frozen aliquots of the exact same extract were used in all electrophoretic mobility shift assays (EMSA) for consistent comparisons across tissue specimens and experiments] and measured by PhosphorImager (Molecular Dynamics, Sunnyvale, CA). EMSA was done in triplicate, independent analyses for all matched tumor and nontumor specimens in this trial, and Stat3 DNA-binding activities were compared with the levels of phospho-Stat3 as measured by immunohistochemistry of mirror image tissue samples.

Statistical considerations. Before statistical evaluation, pathologic response was classified as either complete pathologic response or partial pathologic response based on the size of residual tumor after treatment (complete pathologic response if 0 cm, partial pathologic response if >0 cm). Biomarker values were analyzed by descriptive statistics, including mean, SD, and range (minimal-maximal), values in all patients as a group and in the complete pathologic response and partial pathologic response groups separately both in nonneoplastic tissue and in tumor tissue. Comparison of pretreatment samples between the complete pathologic response and partial pathologic response groups was examined using the Student's t test if the data followed a normal distribution and the Wilcoxon Mann-Whitney test if the normality assumption was not met. In addition, pretreatment to posttreatment changes were evaluated using the paired t test in the partial pathologic response group. Correlations between pretreatment values were assessed by the Spearman's rank correlation coefficient. No adjustments were made for multiple testing owing to the exploratory nature of this study. All tests were two-sided and declared significant at the 5% level.

Results

Immunohistochemistry and quantitative image analysis

Figure 1 shows examples of immunohistochemical staining in serial sections of nontumor (A, C, and E) and matched tumor tissues (B, D, and F) from the same patient using antibodies to HER2/neu (A and B), pY-Src (C and D), and pY-Stat3 (E and F). For quantification by digital image analysis, regions of interest were selected to include only epithelial components (nonneoplastic or carcinoma) and to eliminate mesenchymal areas. Each of the markers was evaluated in its corresponding cellular compartment (nucleus, cytoplasm, or membrane), as shown in Fig. 2. In the case of predominantly nuclear markers like pY-Stat3 (A, B, and C), thresholds were set to discriminate between the brown color of 3,3′-diaminobenzidine (or the red color of Nova-Red for Survivin) and the blue color of hematoxylin in negative nuclei. For membranous markers like HER2/neu (D, E, and F), the 3,3′-diaminobenzidine signal was measured as a percentage of the total cellular area after subtraction from the region of interest of both nuclear and cytoplasmic components. For predominantly cytoplasmic markers like pY-Src (G, H, and I), the brown 3,3′-diaminobenzidine signal was quantified in the total cellular area of the selected region of interest after subtraction of the blue nuclear area. In all three situations, positive signals were reported as an index reflecting the optical intensity of the marker in relationship to the total optical intensity of the region of interest (average of three measurements).

Fig. 1.
Fig. 1.

Tissue expression of HER2/neu, pY-Src, and pY-Stat3 as detected by immunohistochemistry. Representative immunohistochemistry staining for HER2/neu (A and B), pY-Src (C and D), and pY-Stat3 (E and F) was done on consecutive sections of nonneoplastic tissue (A, C, and E) and paired carcinoma (B, D, and F) from the same patient. No expression of HER2/neu (A) and pY-Src (C) are detected in nonneoplastic tissue, whereas luminal cells show expression of pY-Stat3 (E). In contrast, elevated expression of membranous HER2/neu (B), cytoplasmic pY-Src (D), and nuclear pY-Stat3 (F) is present in the carcinoma.

Fig. 2.
Fig. 2.

Computerized image analysis of pY-Stat3, HER2/neu, and pY-Src staining levels in invasive carcinoma. Representative image analysis from the same patient shown in Fig. 1. For pY-Stat3: A, immunohistochemistry showing predominantly nuclear pY-Stat3 staining. B, the entire nuclear area (false-color green). C, nuclear pY-Stat3 signal (black) quantified by digital image analysis. For HER2/neu: D, immunohistochemistry showing predominantly membranous HER2/neu signal. E, the nuclei are subtracted from the image. F, cytoplasmic signal is extracted from the image and membranous signal (dark brown) is quantified by digital image analysis. For pY-Src: G, immunohistochemistry showing predominantly cytoplasmic pY-Src staining. H, the region of interest corresponding to invasive carcinoma is isolated from the image. I, nuclear component (green) subtracted from the region of interest for quantification of cytoplasmic pY-Src by digital image analysis.

Table 1 summarizes the descriptive statistics of the immunohistochemical quantitative analysis in nonneoplastic and tumor tissues by pathologic response for pretreatment and posttreatment. Of the 45 patients included in the clinical trial, 12 (27%) showed a complete pathologic response and 33 (73%) showed a partial pathologic response.6 Statistical analysis of the immunohistochemical data was done in all 45 patients as a group and in the complete pathologic response and partial pathologic response groups separately, as described below.

View this table:
Table 1.

Descriptive statistics of molecular biomarker values as measured by immunohistochemistry in nonneoplastic tissues and tumors

Molecular biomarker levels in tumor versus nonneoplastic tissues

Table 2 shows that in tissues obtained before chemotherapy treatment, levels of pY-Stat3, pY-Src, and Ki-67 were significantly higher in tumors than in nonneoplastic tissues, both when all tumors were analyzed as a group (ALL) without regard to pathologic response (P ≤ 0.001) and when the analysis was done in the complete pathologic response and partial pathologic response groups separately. In contrast, pretreatment levels of ER in the complete pathologic response group, and Bcl-2 in both complete pathologic response and partial pathologic response groups, were significantly higher in nonneoplastic tissues than in tumors (compare Tables 1 and 2). Similar relationships were found in tissues after chemotherapy treatment, with higher levels of pY-Stat3, pY-Src, and Ki-67 in tumors and higher levels of Bcl-2 in nonneoplastic tissues.

View this table:
Table 2.

Statistical analysis (P values) of differences in biomarker values between tumors and paired nonneoplastic tissues

Correlation between biomarker levels in tumors

No significant correlations were found among biomarkers in nonneoplastic tissues, either in the pretreatment or in the posttreatment groups (data not shown). In tumors, however, the following significant statistical correlations among molecular biomarkers were observed (Table 3; data not shown).

View this table:
Table 3.

Significant correlations between biomarker values in tumors

Pretreatment. When pretreatment tumors were analyzed as a group, without regard to pathologic response, ER correlated with EGFR; PR with Bcl-2 and Bcl-xL; pY-Src with Bcl-2, Bcl-xL, EGFR, and Ki-67; and pY-Stat3 with HER2/neu and Survivin. In the complete pathologic response group, Bcl-2 correlated with Bcl-xL and PR, and pY-Stat3 with HER2/neu and pY-Src. In the partial pathologic response group, EGFR correlated with ER and TUNEL; pY-Src with Bcl-2, Bcl-xL, and EGFR; and pY-Stat3 with ER and Survivin. When pretreatment tumor values in the complete pathologic response group were compared with those of the partial pathologic response, only pY-Stat3 was found to have a statistically significant correlation with response to therapy (P = 0.028). Levels of pY-Stat3 were lower in tumors of patients who showed complete pathologic response, suggesting that higher levels of activated Stat3 made tumors less responsive to the treatment. Taking all of the data into consideration, the most statistically significant correlation with clinical relevance among the pretreatment biomarkers was between pY-Stat3 and Survivin levels in the partial pathologic response group (P = 0.001; see Gritsko et al., in this issue).

Posttreatment. In the posttreatment partial pathologic response group, a direct correlation between HER2/neu and Ki-67 and an inverse correlation between EGFR and pY-Src were observed.

Pretreatment versus posttreatment (partial pathologic responders only). Levels of Ki-67 were higher in pretreatment samples than in posttreatment samples (P < 0.02), and levels of Bcl-2 were higher in posttreatment samples than in pretreatment samples (P < 0.02).

Stat3 activation determined by EMSA

Nuclear extract preparations from snap frozen, mirror image samples of the same tissues used for immunohistochemistry were analyzed by EMSA for Stat3 DNA-binding activity. EMSA results show constitutive activation of Stat3 homodimers in the majority of the breast tumors tested (Fig. 3). These results are consistent with our previous findings (18). However, we did not observe a direct correlation between the values for activated Stat3 DNA-binding levels obtained by EMSA and those for immunohistochemistry of pY-Stat3 levels in the same matched patient pairs (data not shown). This may be due, in large part, to the presence of nontumor stromal cells, which contain reduced levels of activated Stat3 (e.g., Fig. 1; Table 1), in the nuclear extract preparations of whole tumor specimens that would affect overall DNA-binding activity. Thus, we consider immunohistochemistry to more accurately reflect levels of activated Stat3 in tumor cells specifically. Furthermore, activated Stat1:Stat3 heterodimers and Stat1 homodimers were also detected by EMSA in some cases (top). It is not clear from this analysis whether the activated Stat1 arises predominantly from tumor or nontumor cells (stromal and/or immune cells) present in the tissue specimens. Importantly, both immunohistochemistry and EMSA showed a similar overall trend of higher levels of activated Stat3 in tumors compared with nonneoplastic tissues (bottom).

Fig. 3.
Fig. 3.

STAT DNA-binding activity in pretreatment tissue specimens. DNA-binding activity was measured in whole tissue specimens by EMSA. Top, nuclear extracts were prepared from frozen biopsies of paired tumor (T) and matched nonneoplastic (N) tissues from nine representative patients. The radiolabeled hSIE probe that binds Stat3 and Stat1 proteins specifically was incubated with nuclear extracts and DNA complexes were resolved by gel electrophoresis. DNA-binding activity in tumors was determined relative to serial dilutions of nuclear extract prepared from the MDA-MB-468 breast cancer cell line as an internal reference standard (not shown). The positions of Stat3 homodimers, Stat1:Stat3 heterodimers, and Stat1 homodimers in the gel are indicated. Bottom, quantification of DNA-binding activities in all of the patient specimens shows overall higher Stat3 DNA-binding activity in tumors (striped columns) than in nonneoplastic tissues (dotted columns). Columns, percent Stat3 activity relative to that of the internal reference standard, which had a higher level of constitutive Stat3 activity than any of the tissue specimens.

Discussion

We present evidence for elevated activities of Stat3 and Src in tumor specimens compared with matched nonneoplastic tissues from the same patients with invasive carcinoma. In particular, our findings with 45 patients are the most compelling to date showing that Stat3 and Src signaling are significantly activated (P < 0.001) in breast tumors compared with normal mammary tissues. These observations are consistent with previous reports that suggested important roles for activated Stat3 and Src in human breast cancer cell lines and in limited numbers of patient tumor specimens (18, 28–33). Moreover, our studies reveal a significant correlation of elevated Stat3 activity with partial pathologic response to chemotherapy, which may reflect the antiapoptotic properties of Stat3 that are also favorable for promoting the tumorigenic phenotype.

An earlier report suggests that nuclear expression of pY-Stat3 is associated with a better prognosis for overall survival in node-negative breast cancer (34). Although this apparent discrepancy with our findings is not well understood at the present time, we note the critical influence of processing time following tissue resection to preserve pY-Stat3 levels. Our experience indicates that prolonged delay over 15 minutes in fixing or freezing tissue specimens results in reduced phosphorylation levels of STATs (18), probably due to tyrosine phosphatase activity. In this context, the use of retrospective archived specimens in the earlier study (34) may in part explain some of the discrepancy with our study because the archived specimens might not have been processed within 15 minutes. For the present study, we used only prospectively collected specimens for which a tissue processing time of <15 minutes was documented. Furthermore, the earlier study (34) correlated pY-Stat3 levels with prognosis, whereas our study focused on response to chemotherapy. The present study is consistent with a previous report that elevated Stat1 activity in breast cancer is a prognostic indicator of longer and relapse-free survival (35), reflecting the known function of Stat1 in inducing growth arrest and apoptosis in response to physiologic signals, such as IFN-γ (36). In particular, evidence indicates that Stat1 functions antagonistically to Stat3, which has intrinsic oncogenic potential, whereas Stat1 has opposing tumor suppressor–like properties (37, 38). Our findings reported here show that elevated levels of nuclear pY-Stat3 are associated with invasive breast cancer.

Analysis of biomarker expression in pretreatment samples revealed numerous significant correlations (Tables 2 and 3) of potential relevance to breast oncogenesis. As might be expected, the cell proliferation marker Ki-67 was significantly higher in tumor compared with normal tissues and correlated well with Src activation. Paradoxically, however, levels of Bcl-2 were higher in normal tissues compared with tumors, consistent with other studies (39). In addition, levels of pY-Stat3, HER2/neu, and pY-Src are correlated with each other in the group of complete pathologic responders. These observations are consistent with the known link between Src and Stat3 activation in fibroblasts and breast carcinoma cells (18, 40). Furthermore, Src activation has been implicated in HER2/neu-mediated migratory potential of breast cancer cell lines (17) and in the maintenance of the transformed phenotype of cells overexpressing HER2/neu (41). Other studies indicate activation of Stat3 signaling by the HER2/neu pathway (42, 43). Thus, our present study and those of previous reports suggest cooperation between HER2/neu, Src, and Stat3 signaling pathways in breast oncogenesis although the relevance of this interaction to chemotherapy response is unclear. Our studies did not reveal a good correlation between EGFR and constitutively active pY-Stat3, despite the fact that EGF is a potent physiologic inducer of Stat3 signaling (44). This finding may in part reflect induction of constitutive Stat3 activation in breast carcinoma cells by other tyrosine kinases independent of EGFR, including Src (18, 45) and Janus kinase family tyrosine kinases (18, 46). Further study is required to better understand the potential relevance of all the correlations noted among the various molecular biomarkers, some of which are consistent with expectations, whereas others are surprising in the context of our current knowledge.

A clinically important correlation was observed between pY-Stat3 and Survivin, an antiapoptotic protein of the inhibitors of apoptosis protein family (13–16). The findings revealed a strong correlation between pY-Stat3 and Survivin levels in those patients who were partial pathologic responders to chemotherapy.6 Further support for a link between Stat3 and Survivin comes from studies showing that Survivin is a direct downstream target of Stat3 (see Gritsko et al., in this issue). Together, these findings suggest that elevated levels of Stat3 activity and consequently Survivin expression in breast tumor cells are key factors in promoting the survival of tumor cells and determining their response to chemotherapy. Our conclusion is consistent with reports that inhibition of persistently active Stat3 signaling enhance sensitivity to chemotherapy in B-cell non-Hodgkin's lymphoma and myeloma cells (,47, 48).

The observation that elevated Stat3 activity is associated with breast oncogenesis and correlates inversely with complete pathologic response to chemotherapy suggests that the Stat3 signaling pathway is a potential therapeutic target in breast cancer. Furthermore, inhibitors of Src tyrosine kinase activity block constitutively active Stat3 signaling and induce apoptosis in breast cancer cells (18, 49). Our present study provides support for the development of small molecule inhibitors of Stat3 as well as those of Src kinase for cancer therapeutics (3, 12, 50). This conclusion is further strengthened by our finding that inhibition of Stat3 and downstream Survivin expression induces apoptosis in human breast cancer cell lines harboring constitutively active Stat3 (see Gritsko et al., in this issue). Small molecule inhibitors of Stat3 and Src could be used in combination with conventional chemotherapy or with other signal transduction blockers for increased therapeutic benefits in breast cancer patients.

Acknowledgments

We thank members of the Tissue Procurement, Translational Research, Analytical Microscopy, and Biostatistics Cores at the Moffitt Cancer Center and Research Institute, and the Frank and Carol Morsani Endowment in Molecular Oncology.

Footnotes

  • ↵6 S. Minton et al., submitted for publication.

  • Grant support: NIH grants CA82533 and CA55652; Angela Musette Russo Foundation; and Tissue Procurement, Translational Research, Analytic Microscopy, and Biostatistics Core Facilities of Moffitt Cancer Center.

  • 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: N. Diaz and S. Minton contributed equally to this work.

    • Accepted September 20, 2005.
    • Received August 27, 2004.
    • Revision received June 30, 2005.

References

  1. ↵
    Bishop JM. The molecular genetics of cancer. Science 1987;235:305–11.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    Hahn WC, Weinberg RA. Modelling the molecular circuitry of cancer. Nat Rev Cancer 2002;2:331–41.
    OpenUrlCrossRefPubMed
  3. ↵
    Darnell JE. Transcription factors as targets for cancer therapy. Nat Rev Cancer 2002;2:740–9.
    OpenUrlCrossRefPubMed
  4. Bromberg J, Darnell JE. The role of STATs in transcriptional control and their impact on cellular function. Oncogene 2000;19:2468–73.
    OpenUrlCrossRefPubMed
  5. ↵
    Levy DE, Darnell JE. STATs: transcriptional control and biologic impact. Nat Rev Mol Cell Biol 2002;3:651–62.
    OpenUrlCrossRefPubMed
  6. ↵
    Bowman T, Garcia R, Turkson J, Jove R. STATs in oncogenesis. Oncogene 2000;9:2474–88.
  7. Song JI, Grandis JR. STAT signaling in head and neck cancer. Oncogene 2000;19:2489–95.
    OpenUrlCrossRefPubMed
  8. Turkson J, Jove R. STAT proteins: novel molecular targets for cancer drug discovery. Oncogene 2000;19:6613–26.
    OpenUrlCrossRefPubMed
  9. Lin TS, Mahajan S, Frank DA. STAT signaling in the pathogenesis and treatment of leukemias. Oncogene 2000;19:2496–504.
    OpenUrlCrossRefPubMed
  10. Bromberg J. Signal transducers and activators of transcription as regulators of growth, apoptosis and breast development. Breast Cancer Res 2000;2:86–90.
    OpenUrlCrossRefPubMed
  11. Buettner R, Mora LB, Jove R. Activated STAT signaling in human tumors provides novel molecular targets for therapeutic intervention. Clin Cancer Res 2002;8:945–54.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    Yu H, Jove R. The STATs of cancer—new molecular targets come of age. Nat Rev Cancer 2004;4:97–105.
    OpenUrlCrossRefPubMed
  13. ↵
    Ambrosini G, Adida C, Altieri DC. A novel anti-apoptosis gene, survivin, expressed in cancer and lymphoma. Nat Med 1997;3:917–21.
    OpenUrlCrossRefPubMed
  14. Li F, Altieri DC. Transcriptional analysis of human Survivin gene expression. Biochem J 1999;344:305–11.
  15. Altieri DC. Validating Survivin as a cancer therapeutic target. Nat Rev Cancer 2003;3:46–54.
    OpenUrlCrossRefPubMed
  16. ↵
    Altieri DC. Survivin and apoptosis control. Adv Cancer Res 2003;88:31–52.
    OpenUrlCrossRefPubMed
  17. ↵
    Belsches-Jablonski AP, Biscardi JS, Peavy DR, Tice DA, Romney DA, Parsons SJ. Src family kinases and HER2 interactions in human breast cancer cell growth and survival. Oncogene 2001;20:1465–75.
    OpenUrlCrossRefPubMed
  18. ↵
    Garcia R, Bowman TL, Niu G, et al. Constitutive activation of Stat3 by the Src and Janus kinase tyrosine kinases participates in growth regulation of human breast carcinoma cells. Oncogene 2001;20:2499–513.
    OpenUrlCrossRefPubMed
  19. ↵
    Pegram MD, Pauletti G, Slamon DJ. HER2/neu as a predictive marker of response to breast cancer therapy. Breast Cancer Res Treat 1998;52:65–77.
    OpenUrlCrossRefPubMed
  20. ↵
    Gao B, Shen X, Kunos G, et al. Cross-talk between signal transducer and activator of transcription 3 and estrogen receptor signaling. FEBS Lett 2000;486:143–8.
    OpenUrlCrossRefPubMed
  21. ↵
    Boonyaratanakornkit V, Scott MP, Ribon V, et al. Progesterone receptor contains a proline-rich motif that directly interacts with SH3 domains and activates c-Src family of tyrosine kinases. Mol Cell 2001;8:269–80.
    OpenUrlCrossRefPubMed
  22. ↵
    Archer CD, Parton M, Smith IE, et al. Early changes in apoptosis and proliferation following primary chemotherapy for breast cancer. Br J Cancer 2003;89:1035–41.
    OpenUrlCrossRefPubMed
  23. ↵
    Real PJ, Sierra A, De Juan A, Segovia JC, Lopez-Vega JM, Fernandez-Luna JL. Resistance to chemotherapy via Stat3-dependent overexpression of Bcl-2 in metastatic breast cancer cells. Oncogene 2002;21:7611–8.
    OpenUrlCrossRefPubMed
  24. ↵
    Sjostrom J, Blomqvist C, von Boguslawski K, et al. The predictive value of bcl-2, bax, bcl-xL, bag-1, fas, and fasL for chemotherapy response in advanced breast cancer. Clin Cancer Res 2002;8:811–6.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    Prenzel N, Zwick E, Leserer M, Ullrich A. Tyrosine kinase signaling in breast cancer. Epidermal growth factor receptor: convergence point for signal integration and diversification. Breast Cancer Res 2000;2:184–90.
    OpenUrlCrossRefPubMed
  26. ↵
    Li X, Traganos F, Melamed MR, Darzynkiewicz Z. Single step procedure for labeling DNA strand breaks with fluorescein- or BODIPY-conjugated deoxynucleotides: detection of apoptosis and bromodeoxyuridine incorporation. Cytometry 1995;20:172–80.
    OpenUrlCrossRefPubMed
  27. ↵
    Bilous M, Ades C, Armes J, et al. Predicting the HER2 status of breast cancer from basic histopathology data: an analysis of 1500 breast cancers as part of the HER2000 International Study. Breast 2003;12:92–8.
    OpenUrlCrossRefPubMed
  28. ↵
    Rosen N, Bolen JB, Schwartz AM, Cohen P, DeSeau V, Israel MA. Analysis of pp60c-src protein kinase activity in human tumor cell lines and tissues. J Biol Chem 1986;261:13754–9.
    OpenUrlAbstract/FREE Full Text
  29. Watson CJ, Miller WR. Elevated levels of members of the STAT family of transcription factors in breast carcinoma nuclear extracts. Br J Cancer 1995;71:840–4.
    OpenUrlCrossRefPubMed
  30. Garcia R, Yu CL, Hudnall A, et al. Constitutive activation of Stat3 in fibroblasts transformed by diverse oncoproteins and in breast carcinoma cells. Cell Growth Differ 1997;8:1267–76.
    OpenUrlAbstract
  31. Sartor CI, Dziubinski ML, Yu CL, Jove R, Ethier SP. Role of epidermal growth factor receptor and STAT-3 activation in autonomous proliferation of SUM-102PT human breast cancer cells. Cancer Res 1997;57:978–87.
    OpenUrlAbstract/FREE Full Text
  32. Biscardi JS, Ishizawar RC, Silvan CM, Parsons SJ. Tyrosine kinase signaling in breast cancer: epidermal growth factor receptor and c-Src interactions in breast cancer. Breast Cancer Res 2000;2:203–10.
    OpenUrlCrossRefPubMed
  33. ↵
    Hung W, Elliott B. Co-operative effect of c-Src tyrosine kinase and Stat3 in activation of hepatocyte growth factor expression in mammary carcinoma cells. J Biol Chem 2001;276:12395–403.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    Dolled-Filhart M, Camp R, Kowalski D, Smith B, Rimm D. Tissue microarray analysis of signal transducers and activators of transcription 3 (Stat3) and phospho-Stat3 (Tyr705) in node-negative breast cancer shows nuclear localization is associated with a better prognosis. Clin Cancer Res 2003;9:594–600.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Widschwendter A, Tonko-Geymayer S, Welt T, Daxenbichler G, Marth C, Doppler W. Prognostic significance of signal transducer and activator of transcription 1 activation in breast cancer. Clin Cancer Res 2002;8:3065–74.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    Shen Y, Devgan G, Darnell JE, Bromberg JF. Constitutively activated Stat3 protects fibroblasts from serum withdrawal and UV-induced apoptosis and antagonizes the proapoptotic effects of activated Stat1. Proc Natl Acad Sci U S A 2001;98:1543–8.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    Kaplan DH, Shankaran V, Dighe AS, et al. Demonstration of an interferon γ-dependent tumor surveillance system in immunocompetent mice. Proc Natl Acad Sci U S A 1998;95:7556–61.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    Bromberg JF, Wrzeszczynska MH, Devgan G, et al. Stat3 as an oncogene. Cell 1999;98:295–303.
    OpenUrlCrossRefPubMed
  39. ↵
    Siziopikou KP, Prioleau JE, Harris JR, Schnitt SJ. Bcl-2 expression in the spectrum of preinvasive breast lesions. Cancer 1996;77:499–506.
    OpenUrlCrossRefPubMed
  40. ↵
    Yu CL, Meyer DJ, Campbell GS, et al. Enhanced DNA-binding activity of a Stat3-related protein in cells transformed by the Src oncoprotein. Science 1995;269:81–3.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    Karni R, Jove R, Levitzki A. Inhibition of pp60c-Src reduces Bcl-XL expression and reverses the transformed phenotype of cells overexpressing EGF and HER-2 receptors. Oncogene 1999;18:4654–62.
    OpenUrlCrossRefPubMed
  42. ↵
    Fernandes A, Hamburger AW, Gerwin BI. ErbB-2 kinase is required for constitutive stat3 activation in malignant human lung epithelial cells. Int J Cancer 1999;83:564–70.
    OpenUrlCrossRefPubMed
  43. ↵
    DeArmond D, Brattain M, Jessup J, et al. Autocrine-mediated ErbB-2 kinase activation of STAT3 is required for growth factor independence of pancreatic cancer cell lines. Oncogene 2003;22:7781–95.
    OpenUrlCrossRefPubMed
  44. ↵
    Ruff-Jamison S, Zhong Z, Wen Z, Chen K, Darnell JE, Cohen S. Epidermal growth factor and lipopolysaccharide activate Stat3 transcription factor in mouse liver. J Biol Chem 1994;269:21933–5.
    OpenUrlAbstract/FREE Full Text
  45. ↵
    Kloth MT, Laughlin KK, Biscardi JS, Boerner JL, Parsons SJ, Silva CM. STAT5b, a mediator of synergism between c-Src and the epidermal growth factor receptor. J Biol Chem 2003;278:1671–9.
    OpenUrlAbstract/FREE Full Text
  46. ↵
    Ren Z, Schaefer TS. ErbB-2 activates Stat3α in a Src and JAK2-dependent manner. J Biol Chem 2002;277:38486–93.
    OpenUrlAbstract/FREE Full Text
  47. ↵
    Alas S, Bonavida B. Rituximab inactivates signal transducer and activation of transcription 3 (STAT3) activity in B-non-Hodgkin's lymphoma through inhibition of the interleukin 10 autocrine/paracrine loop and results in down-regulation of Bc1–2 and sensitization to cytotoxic drugs. Cancer Res 2001;61:5137–44.
    OpenUrlAbstract/FREE Full Text
  48. ↵
    Alas S, Bonavida B. Inhibition of constitutive STAT3 activity sensitizes resistant non Hodgkin's lymphoma and multiple myeloma to chemotherapeutic drug-mediated apoptosis. Clin Cancer Res 2003;9:316–26.
    OpenUrlAbstract/FREE Full Text
  49. ↵
    Moasser MM, Srethapakdi M, Sachar KS, Kraker AJ, Rosen N. Inhibition of Src kinases by a selective tyrosine kinase inhibitor causes mitotic arrest. Cancer Res 1999;59:6145–52.
    OpenUrlAbstract/FREE Full Text
  50. ↵
    Turkson J. STAT proteins as novel targets for cancer drug discovery. Expert Opin Ther Targets 2004;8:409–22.
    OpenUrlCrossRefPubMed
Back to top
Clinical Cancer Research: 12 (1)
January 2006
Volume 12, Issue 1

Sign up for alerts

View this article with LENS

Open full page PDF
Article Alerts
Email Article
Citation Tools
Activation of Stat3 in Primary Tumors from High-Risk Breast Cancer Patients Is Associated with Elevated Levels of Activated Src and Survivin Expression
Nills Diaz, Susan Minton, Charles Cox, Tammy Bowman, Tanya Gritsko, Roy Garcia, Ibrahim Eweis, Marek Wloch, Sandy Livingston, Ed Seijo, Alan Cantor, Ji-Hyun Lee, Craig A. Beam, Daniel Sullivan, Richard Jove and Carlos A. Muro-Cacho
Clin Cancer Res January 1 2006 (12) (1) 20-28; DOI: 10.1158/1078-0432.CCR-04-1749
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Advertisement