| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
Department of Obstetrics and Gynecology [A. W., G. D., C. M.], and Institute of Medical Chemistry and Biochemistry [A. W., S. T-G., T. W., W. D.], University of Innsbruck, A-6020 Innsbruck, Austria
 |
ABSTRACT |
|---|
 Top ABSTRACT IntroductionPatients and MethodsResultsDiscussionREFERENCES |
|---|
Experimental Design: In this study we have separately determined activation of STAT1, STAT3, and STAT5 by measuring their DNA binding activity and tyrosine phosphorylation in breast cancer tissue samples. The predictive value of STAT activation on relapse-free and overall survival among women who received treatment for primary breast cancer was evaluated in a retrospective study.
Results: Survival analysis demonstrated that patients with high STAT1 activation have substantially longer overall and relapse-free survival, irrespective of whether STAT1 activation was determined by its DNA binding activity (P = 0.003 and 0.010, respectively) or by its tyrosine phosphorylation (P = 0.046 and 0.011, respectively). In accordance, Cox proportional hazard regression analysis revealed an enhanced hazard of death (hazard ratio, 3.77; P = 0.018) and relapse of disease (hazard ratio, 6.55; P = 0.013) for the group of women with low STAT1 activation. After adjusting for known prognostic variables (lymph node status, stage of disease, estrogen receptor status, and cathepsin D), STAT1 activation remained an independent prognostic value. Activation of STAT3 and STAT5 DNA binding did not significantly correlate with prognosis.
Conclusion: Our study reveals a favorable and independent prognostic significance of STAT1 activation in mammary carcinoma, and is in accordance with the documented role of STAT1 in growth arrest, and in pro-apoptotic signaling pathways.
|
Introduction |
|---|
|
TopABSTRACTIntroductionPatients and MethodsResultsDiscussionREFERENCES |
|---|
|
Patients and Methods |
|---|
|
TopABSTRACTIntroductionPatients and MethodsResultsDiscussionREFERENCES |
|---|
. The surgically resected specimens used for this study were obtained from patients with breast carcinoma who underwent surgical treatment at the Department of Obstetrics and Gynecology, University Hospital Innsbruck, during the period from 1989 to 1994. Only patients from whom unfixed frozen tumor tissue was available were included in the study. Our 73 patients represent 25% of the breast cancer patients treated at our Department at that time. The surgical procedure included breast-conserving lumpectomy followed by local radiotherapy (40 patients) or modified radical mastectomy (33 patients). Sixty-seven patients (92%) underwent axillary LN dissection. LN biopsies were histologically negative for cancer in 19 (28%) of these 67 patients, whereas 48 (72%) had histological evidence of metastatic spread to axillary LNs (Table 1)
. Of the 73 patients, 38 (52%) were treated with adjuvant systemic chemotherapy, and 39 (53%) were treated with tamoxifen therapy. For 72 patients, follow-up data were available. The median follow-up period of the patient cohort was 6.8 years. Patients were followed up after primary treatment, at our department, in intervals increasing from 3 months to 1 year until death or end of the study, respectively. Routine examinations included systemic review, tumor marker testing, clinical examination, chest X-ray, and liver ultrasound to evaluate the outcome of the disease, which was classified as disease free/relapse/death by the WHO criteria of clinical response. During follow-up, there were 25 recurrences (34%) and 32 deaths (44%). Recurrence was confirmed by biopsy and/or relevant diagnostic procedures. The actual cause of death was not available for the patients; therefore, the recording of survival was based on death from all causes. Biopsy specimens were obtained immediately after resection of the breast or lumpectomy. Specimens were brought to the pathologist, and a part of the specimen containing tumor tissue was snap-frozen in liquid nitrogen. Tumor content of samples was more than 90% as checked by histological analysis of frozen sections from this area. Before storage at -80°C, the samples were ground to powder under liquid nitrogen and used for analysis of STAT activation and other biochemical parameters (Table 2)
.
|
|
EMSA.
Double-stranded oligonucleotides labeled with [
-32P]ATP (>6000 Ci/mmol) containing specific binding sites for either STAT5 (LHRR of the rat ß-casein gene promoter; Ref. 17
) or STAT1 and STAT3 (SIE; Ref. 17
) were used. The sequence of the upper strand of the oligonucleotides was: LHRR, 5'-TGTGGACTTCTTGGAATTAAGGGACTTTTG-3'; and SIE, 5'-GTGCATTTCCCGTAAATCTTGTCTACAATTC-3'. Binding reactions were performed with 8 µg of protein from whole cell extracts. In supershift experiments, extracts were preincubated with the indicated antibodies for 30 min at 0°C before the addition of the labeled oligonucleotide. Complexes were analyzed on a 4% polyacrylamide gel as described previously (17)
. Whole cell extracts from COS-7 cells expressing activated recombinant STAT1, STAT3, and STAT5 were prepared as described previously (18)
and used as controls for verifying the specificity of DNA probes and antibodies for the individual STAT factors. After electrophoresis, gels were fixed in methanol/acetic acid, dried, and exposed to a phosphoscreen for 4–24 h. The radioactivity was analyzed with a PhosphorImager (Molecular Dynamics, Amersham Pharmacia Biotech United Kingdom Limited, Buckinghamshire, England) and the formation of STAT-specific complex quantified with the ImageQuant software by determining the fraction of oligonucleotides forming STAT-specific complexes. To allow normalization of the results of different experiments, every EMSA included the same two control samples with activated STAT5 or STAT1.
Western Blotting.
Immunoblotting experiments were performed with whole cell extracts. The analysis was restricted to the 53 tumors of those patients, of which enough material was left over after the determination of DNA binding activity by EMSA. Ten µg of protein per lane were applied on 8–16% Tris-glycine-SDS precast polyacrylamide gels (Gradipore Ltd, French Forest, Australia). The separated proteins were transferred to poly(vinylidene) difluoride membranes and probed with antibodies as indicated in the figure legends. The following antibodies were used: rabbit polyclonal anti-phospho-STAT1 (Tyr701); rabbit polyclonal anti-STAT1, (CT); and rabbit polyclonal anti-phopho-STAT3 (Tyr705), mouse monoclonal anti-phospho-STAT3 (Tyr705), and mouse monoclonal anti-phospho-STAT5A/B (Tyr694/Tyr699; all from Upstate Biotechnology, Lake Placid, NY); and rabbit polyclonal anti-
-tubulin (H-300) from Santa Cruz Biotechnology, Santa Cruz, CA. For immunodetection, the enhanced chemiluminescence protocol of Amersham (Pharmacia Biotech United Kingdom Limited) was used. Autoradiographies of blots were scanned by laser densitometry (PDSI; Molecular Dynamics), and the relative abundance of the band specific for each protein was quantified with the ImageQuant software. Nine samples were investigated in parallel. Normalization of different experiments was performed by direct comparison of one representative extract of each experiment in a separate experiment.
Expression Analysis of Other Biochemical Markers in Tumor Tissue.
ER and PR content in the cytosol was determined by a standard ligand binding assay using estradiol or ORG-2058, respectively, as ligand, essentially as described previously (19)
. Tissue samples with 
10 fmol/mg protein receptor content were classified as receptor positive. EGF-R protein and ErbB-2 content was determined by commercially available ELISAs (Oncogene Science, Cambridge, MA). For cathepsin D measurement the immunoradiometric assay of CIS bio international (Gif Sur Yvette, France) was used. The cathepsin D protein levels are given as the ratio of cathepsin D value divided by the total protein content of the samples. PRL-R mRNA content was measured in samples by quantitative reverse transcriptase-PCR as described previously (20)
. Briefly, total cytoplasmic RNA was extracted from tissue powder and 400 ng of RNA was reverse transcribed with hexanucleotide primers and AMV- reverse transcriptase. The abundance of PRL-R transcripts relative to GAPDH transcripts was assessed by semi-quantitative PCR. Fluorescence-labeled primer pairs used for PCR were: 5'-ACTTACATAGTTCAGCCAGACC-3' and 5'-TGAATGAAGGTCGCTGGACTCC-3' for amplification of a 310-bp fragment specific for the extracellular domain of the PRL-R; 5'-TGCACCACCAACTGCTTAGCA-3' and 5'-GAAGTCAGAGGAGACCACCT-3' for amplification of a 400-bp GAPDH fragment. The PCR reaction products were analyzed by 7% PAGE, and quantification of the yield was determined with the GENESCAN software of Perkin-Elmer (Foster City, CA). The amount of cDNA used for PCR was adjusted to ensure mRNA concentration-dependent yield of PCR products.
Statistical Analysis.
All of the statistical analyses were performed using the SPSS 10 statistical package for Macintosh (SPSS, Inc., Chicago, IL). For determination of the association between STAT1 DNA binding and STAT1 tyrosine phosphorylation, Kendall’s tau and Spearman’s rho correlation coefficients and their two-tailed significances were determined. The effect of tumor and patient characteristics on OS and RFS was assessed by the Cox univariant proportional hazard regression model. OS was taken as time from the initial tumor resection to death, and RFS as the time from tumor resection to date of recurrence. Patients were categorized into two groups according to the value of the investigated characteristic as indicated. The groups with low values of the investigated biomarker contained all of the patients with biomarker levels lower than the median value. The relative risk of death or relapse of one group in comparison with the other was calculated as HR. The group of patients with low STAT1 activation was also investigated in a Cox multivariant analysis by adjusting the regression model for LN status, stage of disease, ER status, and cathepsin D expression. The Kaplan-Meier survival function was used to generate survival curves for the patient groups with either low- or high-STAT1 activation, and differences between curves were tested by the log-rank test. Differences in the distribution of STAT1 activation and other patient or tumor characteristics among patient subgroups were analyzed using Fisher’s exact test. A two-sided P of 
0.05 was considered statistically significant.
|
Results |
|---|
|
TopABSTRACTIntroductionPatients and MethodsResultsDiscussionREFERENCES |
|---|
, Lanes 1–10). The STAT molecules were activated by erythropoietin and a chimeric erythropoietin-gp130 receptor (Fig. 1
, Lanes 2–4 and 6–10) as described previously (18)
, and the identity of the complexes were verified by the formation of supershifts with antibodies specific for STAT1 and STAT3 (Fig. 1
, Lanes 3 and 9, respectively). Complexes of STAT5 and the SIE probe were not observed (Fig. 1
, Lane 10). In extracts of mammary carcinoma samples, DNA binding activity of STAT1 and STAT3 was determined by the detection of the formation of STAT1 and STAT3 homodimeric complexes, which had the same mobility as the recombinant STATs. In Fig. 1
, the results are shown for three tumor samples, which contain different amounts of activated STAT1 and STAT3 (Fig. 1
Lanes 11, 14, and 17). The identity of the STAT complexes was verified by supershift experiments with antibodies specific for either STAT1 (Lanes 12, 15, and 18) or STAT3 (Lanes 13, 16, and 19). The STAT3 and STAT1 heterodimeric complex migrated in between the STAT1 and STAT3 homodimeric complexes. The formation of such a complex was verified by its reactivity with both the STAT1- and the STAT3-specific antibody (tumor 1684, Fig. 1
; compare Lane 14 with Lanes 15 and 16). For determination of the individual STAT1 or STAT3 DNA binding activity by PhosphorImager analysis in tumor samples, only the homodimeric complexes of STAT1 or STAT3 were used for quantification. Analysis in 68 tumor samples revealed a frequent activation of both STAT1 and STAT3. The results for 27 tumors are shown in Fig. 2
, top panel. Interestingly, not only the strength of STAT1 and STAT3 activation but also the ratio of STAT1:STAT3 activation varied largely between different tumor carcinoma samples, as shown previously for a smaller number of tumors (14)
.
|
|
. Only in 1 of the 63 tumors (specimen 618), was strong STAT5 DNA binding activity observable. The complex reacted with a STAT5A-specific antibody (not shown). However, in the majority of mammary tumors, STAT5 DNA binding activity was either low or below the limit of the sensitivity of our assay. Thus, in mammary carcinoma, the predominately activated STAT forms are STAT1 and STAT3.
Tyrosine Phosphorylation of STAT1.
As another independent means to determining STAT1 activation in mammary carcinoma, we measured the amount of STAT1 phosphorylation at tyrosine 701 by Western blotting experiments with a STAT1 phosphotyrosine-specific antibody. This phosphorylation is required for STAT1 dimerization and DNA binding. The antibody used was highly selective for the tyrosine phosphorylated form as verified with extracts of COS-7 cells expressing activated or nonactivated STAT1 (Fig. 3
, last two lanes). As shown for 17 representative samples of 53 analyzed tumors (Fig. 3
, first row), there was a high variation in the amount of phosphorylated STAT1 detected in the tumor samples. This variation was not the consequence of different expression levels of the STAT1 protein, as revealed by reprobing the same blots after stripping with an antibody recognizing STAT1 independent of its tyrosine phosphorylation status (Fig. 3
, second row). In this analysis, samples with low or no STAT1 tyrosine phosphorylation and samples with tyrosine phosphorylated STAT1 had similar levels of the STAT1 protein. Thus, the different levels of tyrosine phosphorylated STAT1 seen in tumor tissues appear to be the consequence of tumor-specific differences in the activation of STAT1. To assess the correspondence between STAT1 tyrosine phosphorylation and DNA binding activity of STAT1, the ratio of tyrosine phosphorylated STAT1 and total STAT1 in different tumor samples was estimated by densitometric analysis of the Western blots and compared with DNA binding activity of STAT1 quantified by analysis of the EMSAs with a PhosphorImager. The data obtained with both methods were found to be highly correlated with a significance level P = 0.004 for the Spearman’s rank correlation and of P = 0.005 for the Kendall’s tau b correlation. When data were normalized to total STAT1 expressed in the samples, the significance level was further increased in the correlations (P <0.001 in both tests). The correlations obtained for each tumor investigated are depicted in the scatter diagram of Fig. 4
. There, data were grouped into for quadrants according to the median values for activation of STAT1 measured either by EMSA or immunoblotting. This grouping was used in the survival analysis below to categorize the tumors into classes with either high or low STAT1 activation. The strong correlation between the tyrosine phosphorylation and DNA binding was expected, because tyrosine phosphorylation is a prerequisite for DNA binding. A few tumor samples (in the upper left quadrant of Fig. 4
) exhibited low DNA binding activity despite high tyrosine phosphorylation levels. One possible explanation for this is the expression the protein inhibitor of activated STAT PIAS1, which inhibits the DNA binding of the tyrosine phosphorylated STAT1 (24)
, or other functionally similar acting proteins in these tumors.
|
|
-tubulin (Fig. 3
, bottom row). Comparison of the expression levels of -tubulin with STAT1 proteins revealed that the majority of samples did not have a high variation in the cell content; e.g., only 2 of 17 tumors in Fig. 3
(tumors 1471 and 1857) had consistent lower levels of STAT1 and -tubulin, which is indicative of low cell content.
Survival Analysis.
The primary tumors investigated for STAT activation were derived from a patient cohort (Table 1)
for which follow-up information on outcome of disease was available with a median time period of 6.8 years. This allowed us to test the predictive value of the activation status of an individual STAT factor on OS and RFS. Patients were categorized into two groups, according to the relative activation of the STAT factor investigated. In a univariant analysis of HR according to the Cox proportional hazard regression model, the group of patients with low STAT1 activation had a significant enhanced risk of death and recurrence of disease (Table 3)
. This result was obtained, irrespective of whether DNA binding activity or tyrosine phosphorylation was used as a parameter to estimate STAT1 activation. The HR of the group of patients with either low DNA binding or tyrosine phosphorylation (lower left quadrant of Fig. 4
) as compared with the group with high DNA binding and/or tyrosine phosphorylation in their primary tumors (Fig. 4
, other three quadrants) was especially high. There, the risk of death was elevated 3.77-fold and the risk of recurrence 6.66-fold for the group with low STAT1 activity (Table 3)
. In accordance with the results of the Cox analysis, Kaplan-Meier survival curves demonstrated that patients with high STAT1 activation live longer and relapse less frequently, irrespectively of whether STAT1 activation was determined by STAT1 DNA binding or tyrosine phosphorylation (Fig. 5)
.
|
|
, which indicated that the activation level of these factors does not influence the progression of disease or has a small influence not detected because of the limited size of our patient cohort. However the size of the cohort was big enough to obtain significant effects of known predictive factors such as LN status, PR status, and cathepsin D levels on the outcome of disease (Table 3)
.
STAT1 Activation in Relationship to Other Tumor Characteristics.
To assess the dependence of the predictive value of STAT1 activation on other known predictive factors, a multivariant Cox proportional hazard analysis was performed, in which the risk for death or recurrence of disease of the patient group with low STAT1 activation was adjusted to LN status, stage, ER status, and cathepsin D expression. STAT1 activation remained a significant predictor for risk of death and for recurrence of disease (Table 4)
. This indicates that STAT1 activation can serve as an independent predictor of outcome of disease. In combination with the LN status, STAT1 activation status was a valuable predictor of outcome of disease for the group of LN-positive patients. There, from the 7 patients who are LN positive and exhibit high STAT1 activation in their primary tumors, only 1 patient relapsed, whereas from the 25 LN positive patients with low STAT1 activation, 14 patients relapsed.
|
. The only parameter that correlated with STAT1 activation was the LN status; the majority of tumors derived from LN-negative patients had a high STAT1 activation. This was expected because the group of LN patients had a much better prognosis and thus likely contained a higher proportion of tumors with the good prognosis indicator, high STAT1 activation.
|
|
Discussion |
|---|
|
TopABSTRACTIntroductionPatients and MethodsResultsDiscussionREFERENCES |
|---|
Although the molecular basis for the link between STAT1 activation and good prognosis remains unknown, one intriguing possibility is a direct protective role of STAT1 activation for tumor progression. In fact, several reports have documented an antiproliferative and proapoptotic role of STAT1. In cell lines, STAT1 was required for the induction of the cell cycle inhibitor p21WAF1 (10
, 25 , 26)
, and cells with a hypermethylation of the STAT1 site in p21 promoter failed to express the cell cycle inhibitor p21 (27)
. Furthermore, STAT1 was required for induction of the genes encoding caspase 1 and 11 (8
, 11)
and cooperated with the breast cancer susceptibility gene BRCA1 in the induction of IFN--dependent tumor surveillance systems (28)
. Mice with a targeted disruption of the STAT1 gene have provided additional evidence for the protective role of STAT1 against tumor formation: There, a higher frequency and more rapid formation of sarcomas was found and attributed to a deficient IFN--dependent tumor surveillance system (29)
or an increased susceptibility to the formation of carcinogen-induced thymic tumors together with enhanced survival and proliferation of lymphocytes (11)
. STAT1 possibly also acts, similarly as described for STAT3 (12)
, as a proapoptotic factor during normal mammary gland involution. This notion is supported by the observation of a STAT1 activation subsequent to STAT3 during involution and by the precocious activation of STAT1 in mice with a targeted deletion of STAT3 in the mammary gland (12)
.
What is the mechanism of STAT1 activation in mammary tumors? In our tumor population, STAT1 activation and expression of the STAT1 protein was not linked, which indicated that the different activation levels are caused by differences in signaling pathways involved in either activation or repression of STAT1. The elucidation of the molecular nature of these signaling pathway remains a subject for additional investigations. In other studies, IFN- has been described as the principal activator of STAT1 (30)
. However, genetic evidence (11)
indicates additional IFN--independent modes of STAT1 activation, and biochemical analysis of tumor tissue and tumor cell lines has demonstrated STAT1 activation by different hormones and growth factors such as all-trans-retinoic acid (25)
, FGF (9)
, PRL (31)
, and EGF (8
, 32)
. Only from 3 of 30 esophageal squamous cell carcinoma patients was it possible to derive cell lines, in which EGF was able to induce STAT1. Interestingly, these three patients exhibited a dramatically better prognosis (32)
. Activation of STAT1 in tumors could be also the consequence of relief of repression, e.g., by inactivation of the SOCS1, a repressor of STAT1 activation; mice with a targeted deletion of SOCS1 exhibit constitutive STAT1 activation and enhanced IFN- signaling (33)
.
In accordance with previous reports demonstrating activated STAT3 in tumors and tumor cell lines (14 , 15 , 34) , STAT3 activation was found in a high percentage of the tumors investigated in our study. It has been suggested that STAT3, induced via src and JAK tyrosine kinases, contributes to the growth and survival of breast cancer cell lines and potentially contributes also in vivo to the establishment of the primary tumor (14) . Alternatively, it acts similarly to STAT1 by activation of proapoptotic pathways. The latter possibility is supported by the proapoptotic function of STAT3 during mammary gland involution in mice (12) . In our study, STAT3 was an insignificant predictor of prognosis. A more accurate estimate of the role of STAT3 activation in primary mammary carcinoma would require the investigation of a higher number of tumors.
Strong STAT5 activation was seen only in 1 of 63 tumors investigated, whereas in the other tumors, only a weak or no activation was detectable. Thus, constitutive activation of STAT5 appears to be of minor importance in established primary mammary tumors. The low activation of STAT5 might be the result of the previously described inhibitory effect of NF-
B on STAT5 activation (35)
and the high frequency of NF-B activation in mammary tumors (36)
.
The activation of STAT1 was studied in tumor tissue by biochemical techniques allowing the determination of DNA binding activity and tyrosine phosphorylation status in the tumor specimens. Because these techniques do not allow an assessment of the cells that contain activated STAT1, the development of immunohistochemical techniques would be highly desirable, e.g., by procedures allowing selective staining of tyrosine-phosphorylated STAT1 in paraffin sections. In light of the predictive value of STAT1 activation for prognosis, the development of such a technique would be highly desirable.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by the Jubiläeumsfonds of the Austrian National Bank, Project 4487, and by the Austrian Science Fund, Project F209. 
2 Present address: Department of Pathology, Yale University School of Medicine, New Haven, CT 06520-8023.
3 To whom requests for reprints should be addressed, at Institute of Medical Chemistry and Biochemistry, University of Innsbruck, Fritz-Pregl-Strasse 3, A-6020 Innsbruck, Austria. Phone: 43-512-507-3512; Fax: 43-512-507-2638; E-mail: Wolfgang.Doppler{at}uibk.ac.at
4 The abbreviations used are: STAT, signal transducer(s) and activator(s) of transcription; EGF, epidermal growth factor; EGF-R, EGF receptor; EMSA, electromobility shift assay; ER, estrogen receptor; GAPDH, glucose-phosphate dehydrogenase; HR, hazard ratio; LN, lymph node; LHRR, lactogenic hormone response element; OS, overall survival; PR, progesterone receptor; PRL-R, prolactin receptor; RFS, relapse-free survival; SIE, sis-inducible element; SOCS, suppressor(s) of cytokine signaling.
Received 4/ 8/02; revised 6/12/02; accepted 7/ 2/02.
|
REFERENCES |
|---|
|
TopABSTRACTIntroductionPatients and MethodsResultsDiscussionREFERENCES |
|---|
-mediated growth inhibition of human breast cancer cells. Cell Growth Differ., 11: 335-342, 2000.. J. Immunol., 164: 1286-1292, 2000./signal transducers and activators of transcription (STAT) pathway by hypermethylation at a STAT-binding site in the p21WAF1 promoter region. Cancer Res., 60: 3290-3298, 2000. target genes. Proc. Natl. Acad. Sci. USA, 97: 5208-5213, 2000.-dependent tumor surveillance system in immunocompetent mice. Proc. Natl. Acad. Sci. USA, 95: 7556-7561, 1998.B p50/p65 is regulated in the developing mammary gland and inhibits STAT5-mediated ß-casein gene expression. FASEB J., 14: 1159-1170, 2000.B subunits in human breast cancer: potential roles for NF-B2/p52 and Bcl-3. Oncogene, 19: 1123-1131, 2000.[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  D. Ilkovitch, M. E. Handel-Fernandez, L. M. Herbert, and D. M. Lopez Antitumor Effects of Mucin 1/sec Involves the Modulation of Urokinase-Type Plasminogen Activator and Signal Transducer and Activator of Transcription 1 Expression in Tumor Cells Cancer Res., April 1, 2008; 68(7): 2427 - 2435. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. A. Lynch, J. Etchin, T. E. Battle, and D. A. Frank A Small-Molecule Enhancer of Signal Transducer and Activator of Transcription 1 Transcriptional Activity Accentuates the Antiproliferative Effects of IFN-{gamma} in Human Cancer Cells Cancer Res., February 1, 2007; 67(3): 1254 - 1261. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. N. Torrero, X. Xia, W. Henk, S. Yu, and S. Li Stat1 deficiency in the host enhances interleukin-12-mediated tumor regression. Cancer Res., April 15, 2006; 66(8): 4461 - 4467. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Xi, K. F. Dyer, M. Kimak, Q. Zhang, W. E. Gooding, J. R. Chaillet, R. L. Chai, R. E. Ferrell, B. Zamboni, J. Hunt, et al. Decreased STAT1 Expression by Promoter Methylation in Squamous Cell Carcinogenesis J Natl Cancer Inst, February 1, 2006; 98(3): 181 - 189. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Diaz, S. Minton, C. Cox, T. Bowman, T. Gritsko, R. Garcia, I. Eweis, M. Wloch, S. Livingston, E. Seijo, et al. Activation of Stat3 in Primary Tumors from High-Risk Breast Cancer Patients Is Associated with Elevated Levels of Activated Src and Survivin Expression Clin. Cancer Res., January 1, 2006; 12(1): 20 - 28. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Thomas, C. E. Finnegan, K. M.-A. Rogers, J. W. Purcell, A. Trimble, P. G. Johnston, and M. P. Boland STAT1: A Modulator of Chemotherapy-induced Apoptosis Cancer Res., November 15, 2004; 64(22): 8357 - 8364. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. V. Clevenger Roles and Regulation of Stat Family Transcription Factors in Human Breast Cancer Am. J. Pathol., November 1, 2004; 165(5): 1449 - 1460. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. N. Dechow, L. Pedranzini, A. Leitch, K. Leslie, W. L. Gerald, I. Linkov, and J. F. Bromberg Requirement of matrix metalloproteinase-9 for the transformation of human mammary epithelial cells by Stat3-C PNAS, July 20, 2004; 101(29): 10602 - 10607. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. T. Nevalainen, J. Xie, J. Torhorst, L. Bubendorf, P. Haas, J. Kononen, G. Sauter, and H. Rui Signal Transducer and Activator of Transcription-5 Activation and Breast Cancer Prognosis J. Clin. Oncol., June 1, 2004; 22(11): 2053 - 2060. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention |
