Cancer Research CTRC-AACR San Antonio Breast Cancer Symposium 09 AM Call for Abstracts
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Vanden Berghe, W.
Right arrow Articles by Haegeman, G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Vanden Berghe, W.
Right arrow Articles by Haegeman, G.
[Cancer Research 66, 4852-4862, May 1, 2006]
© 2006 American Association for Cancer Research

Experimental Therapeutics, Molecular Targets, and Chemical Biology

Attenuation of Mitogen- and Stress-Activated Protein Kinase-1–Driven Nuclear Factor-{kappa}B Gene Expression by Soy Isoflavones Does Not Require Estrogenic Activity

Wim Vanden Berghe1, Nathalie Dijsselbloem1, Linda Vermeulen1, 'Matladi N. Ndlovu1, Elke Boone2 and Guy Haegeman1

1 Laboratory for Eukaryotic Gene Expression and Signal Transduction, Department of Molecular Biology, Ghent University, Gent, Belgium and 2 Laboratorium voor Moleculaire Diagnostiek, H.-Hartziekenhuis, Roeselare, Belgium

Requests for reprints: Wim Vanden Berghe, Laboratory for Eukaryotic Gene Expression and Signal Transduction 11HB, Department of Molecular Biology, Universiteit Gent, K.L. Ledeganckstraat 35, B-9000 Gent, Belgium. Phone: 32-9-264-51-47; Fax: 32-9-264-53-04; E-mail: w.vandenberghe{at}ugent.be.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
We have analyzed in molecular detail how soy isoflavones (genistein, daidzein, and biochanin A) suppress nuclear factor-{kappa}B (NF-{kappa}B)–driven interleukin-6 (IL6) expression. In addition to its physiologic immune function as an acute stress cytokine, sustained elevated expression levels of IL6 promote chronic inflammatory disorders, aging frailty, and tumorigenesis. Our results in estrogen-unresponsive fibroblasts, mitogen- and stress-activated protein kinase (MSK) knockout cells, and estrogen receptor (ER)–deficient breast tumor cells show that phytoestrogenic isoflavones can selectively block nuclear NF-{kappa}B transactivation of specific target genes (in particular IL6), independently of their estrogenic activity. This occurs via attenuation of mitogen-activated protein/extracellular signal-regulated kinase (ERK) kinase (MEK) and ERK activity, which further down-regulates MSK-dependent NF-{kappa}B p65 and histone H3 phosphorylation. As constitutive NF-{kappa}B and MSK activity are hallmarks of aggressive metastatic ER-deficient breast cancer, the MSK signaling pathway may become an attractive target for chemotherapy. (Cancer Res 2006; 66(9): 4852-62)


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
The inflammatory response is a tightly controlled process that is critically important to homeostasis. The pleiotropic cytokine interleukin-6 (IL6) affects inflammatory reactions, hematopoiesis, bone metabolism, reproduction (spermatogenesis and menstrual cycle), and aging frailty. In addition to its role in inflammation, IL6 acts as a paracrine/autocrine growth factor of many tumor cells, which establishes a functional link between both types of affections. IL6-mediated signaling pathways have been implicated in tumor progression, invasion, motility, and chemoresistance in solid and hematopoietic tumors (renal cell carcinoma, breast, lung, colon, ovarian, and gut cancer and multiple myeloma; reviewed in ref. 1). In this respect, serum IL6 levels are considered as a diagnostic and prognostic marker for tumor progression. Underscoring the potential value of targeted anti-IL6 therapy in cancer, anti-IL6 monoclonal antibodies were found to induce apoptosis and regression of xenografted human prostate cancer cells in a nude mouse model. Similarly, inhibition of IL6 trans-signaling inhibits tumor progression in colon cancer.

IL6 is normally expressed at low levels, except during infection, trauma, aging, or other stress conditions. Tumor necrosis factor (TNF)–induced IL6 gene expression is primarily controlled at the transcriptional level by the transcription factor nuclear factor-{kappa}B (NF-{kappa}B) and requires in addition to I{kappa}B kinase (IKK) activity, activation of the mitogen-activated protein kinase (MAPK)/MSK kinase pathway, which phosphorylates NF-{kappa}B p65 and histone H3, to establish a transcription-competent promoter complex (enhanceosome; ref. 2). After menopause or andropause, IL6 levels are increasing with age as a consequence of a rapid decline in circulating estrogen or testosterone hormones. This altered regulation may certainly account for several disease-associated inflammatory pathologies, phenotypical changes of advanced age, and accelerated tumorigenesis (1, 3). With the aging of our population, prevention of these types of complaints and maintenance of the important physiologic inflammatory balance has attained paramount importance. Until recently, conventional hormone replacement therapy (HRT) was thought of as a cornerstone in that process (4, 5). The ground for estrogen supplementation following menopause was based on the clinical observations that elderly women without circulating sex steroids had a higher incidence of osteoporotic fractures, coronary heart disease, hot flashes, and mood fluctuations. However, as conventional HRT has recently been associated with an increased incidence of trombosis and breast and endometrial cancer (4, 5), there is a renewed interest in using dietary natural plant estrogens (phytoestrogens), fueled by observational studies showing a lower incidence of menopausal symptoms, osteoporosis, cardiovascular disease, and breast and endometrial cancers in Asian women who have a diet rich in soy products (6, 7). The isoflavones genistein, daidzein, and biochanin A, which are abundant in soybeans and available as herbal tablets, gained enormous attention as structure-function studies have revealed a stable, strong binding to the estrogen receptor (ER), which raised assumptions ranging from mimicking normal estrogenic actions to competitive inhibitory effects (8). The predominant biological effects of estrogen hormones are mediated through two distinct intracellular receptors, ER{alpha} and ERß (9). The functional interaction or "cross-talk" between the ER and NF-{kappa}B has been suggested to play a key role in estrogen prevention of age-related inflammatory pathologies and tumorigenesis in vivo (reviewed in refs. 1, 10, 11). Of special note, loss of ER function has been associated with constitutive NF-{kappa}B activity and hyperactive MAPK in response to constitutive secretion of cytokines and growth factors, which culminates in aggressive, metastatic, hormone-resistant cancers.

Besides its ER-dependent activities, several other estrogen-independent properties of isoflavones may contribute to their actions as well, as antiosteoporotic activities by isoflavones were also described in ovariectomized rats. Other features observed include binding to other (nuclear) receptors [estrogen-related receptor, peroxisome proliferator-activated receptor (PPAR), aryl hydrocarbon receptor, etc.], antioxidant effects due to their polyphenolic nature, modulation of detoxification and of steroid metabolism, interference with Ca transport, Na+/K+ ATPases, favorable effects on lipid and lipoprotein profiles, inhibition of tyrosine protein kinases, phosphatidylinositol 3-kinase (PI3K)/Akt kinase, topoisomerase II and of cyclic AMP (cAMP)-phosphodiesterase-4 enzymes, and interference with cell cycle transition (reviewed in refs. 1, 12). Today, various activities of isoflavones have been linked to suppression of the NF-{kappa}B signaling pathway (reviewed in refs. 7, 13, 14). The presumable therapeutic strength of phytoestrogens in inflammatory disorders or tumorigenesis could therefore rely on the combination of several features in one natural molecule.

In this study, we have further analyzed the suppressive effects of soybean-derived phytoestrogens (i.e., genistein, daidzein, and biochanin A) on NF-{kappa}B–dependent IL6 gene expression in estrogen-unresponsive fibroblast cells and characterized in more detail the specific signaling pathways affected. Finally, results were also applied to an ER-deficient breast cancer model.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cell culture assays. Mouse fibroblast L929sA cells, primary fibroblasts from wild-type and MSK1–/–/MSK2–/– mice, human embryonic kidney HEK293T cells, MDA-MB231, SKBR3, and MCF7 breast cancer cells were regularly cultured in DMEM supplemented with 5% FCS and 5% newborn calf serum, 100 units/mL penicillin, and 0.1 mg/mL streptomycin. TM4 cells were grown in DMEM/Nut Mix F12 supplemented with 5% horse serum, 2.5% FCS, and 1% penicillin/streptomycin (all reagents from Invitrogen, San Diego, CA).

Cytokines, kits, and inhibitors. Recombinant TNF has been described previously (2). Secreted IL6 levels were determined by IL6 immunoassay kit purchased from R&D Systems, Inc. (Minneapolis, MN). SB203580, PD98059, tyrphostin A23, rolipram, LY294002, and wortmannin were purchased from Alexis (Lausen, Switzerland) and H89 was obtained from Calbiochem-Novabiochem International (San Diego, CA). Trichostatin A, genistein, daidzein, biochanin A, 17ß-estradiol, OH-tamoxifen, BHA, ciglitazone, and N-acetyl-L-cystein were purchased from Sigma (St. Louis, MO). ICI 182780 was obtained from Tocris (Ellisville, MO) and U0126 from Promega (Madison, WI).

Western blotting and antibodies. L929sA cells were grown until subconfluency in six-well plates and were treated as indicated in the figure legends. For phospho-specific Western analysis [p38, extracellular signal-regulated kinase (ERK), MAP/ERK kinase (MEK), MSK], cells were serum starved for at least 24 hours. After induction, total lysates were prepared with SDS-Laemmli buffer, containing 62.5 mmol/L Tris-HCl (pH 6.8), 2% SDS, 10% glycerol, 1 mmol/L DTT, and bromophenol blue. Cell lysates were separated by SDS-PAGE and blotted onto a polyvinylidene difluoride membrane (Schleicher & Schuell Bioscience, Keene, NH). For Western detection of p65 and ER{alpha}, sc-372 (C20) and sc-7202 (H184) were used, respectively (Santa Cruz Biotechnology, Santa Cruz, CA). Phospho-specific polyclonal rabbit antibodies to p38 (T180/Y182), p42/44 ERK (T202/Y204), MAPK, MEK1 (S217/S221), MSK1 (S376), IKK{alpha}/ß (S180/S181), and NF-{kappa}B p65 (S276) were used to detect the respective phosphorylated forms and purchased from Cell Signaling (Beverly, CA). Dimethyl Lys9 H3 antibodies were purchased from Upstate (Lake Placid, NY). Anti-phosphorylated-H3 (p-H3) and anti-phosphoacetyl-H3 antibodies were kindly provided by A. Clayton (15). ERß antibody was a kind gift of M. Warner (16). Anti-actin antibody was obtained from ICN (Irvine, CA).

Promoter analysis and transactivation assays, electrophoretic mobility shift assay. The plasmids p1168hu.IL6P-luc+, p(IL6-{kappa}B)350hu.IL6P-luc+, pPGKßgeobpA, and pCMV-CBP were described previously (17, 18). L929sA cells or HEK293T cells were transiently or stably transfected with the plasmids indicated in the figure legends, by the DEAE dextrane method or the calcium phosphate precipitation procedure, respectively, as described previously (17). Reporter gene assays were carried out essentially as described elsewhere (17). Luciferase activity, expressed in arbitrary light units, was corrected for the protein concentration and transient transfection efficiency by normalization for coexpressed ß-galactosidase levels. The latter were quantified with a chemiluminescent reporter assay Galacto-star kit from TROPIX (San Francisco, CA).

L929sA cells constitutively expressing Gal4-p651-551 were transiently transfected with the reporter gene construct p(GAL4)2-50hu.IL6-luc+ using the DEAE dextrane method (2). The reporter plasmid containing two sites for the yeast transcription factor Gal4 in front of an IL6-TATA box–containing minimal promoter was described previously (17). Electrophoretic mobility shift assay (EMSA) has been done as described elsewhere (17).

Immunoprecipitation-MSK1 kinase assay. L929sA cells were seeded at 1 x 106 per dish and grown until subconfluence. After 48 hours of starvation, cells were treated as indicated in figure legends. Endogenous MSK1 was immunoprecipitated, as described elsewhere (2). Immunoprecipitated MSK1 was incubated with 30 µmol/L p65-tide (CMQLRRPSDRELSE) for 20 minutes at 30°C in the phosphorylation assay buffer [50 mmol/L Tris-HCl (pH 7.5), 0.1 mmol/L EGTA, 0.1% ß-mercaptoethanol, 2.5 µmol/L PKI, 1 µmol/L microcystin, 10 mmol/L Mg(Ac)2, 0.1 mmol/L {gamma}[32P]ATP (100-200 cpm/pmol)]. Incorporation of phosphate into peptides was determined using p81 phosphocellulose paper.

Chromatin immunoprecipitation assay. Chromatin immunoprecipitation analysis and semiquantitative PCR has been previously described (2). The following promoter-specific primers were used: hIL6 sense, 5'-GCGCTAGCCTCAATGACGACCTAAG-3' and hIL6 antisense, 5'-GAGCCTCAGACATCTCCAGTCCTAT-3'; mIL6 sense, 5'-TGACTTCAGCTTTACTCTTGT-3' and mIL6 antisense, 5'-CTGATTGGAAACCTTATTAAG-3'; H4 sense, 5'GACACCGCATGCAAAGAATAGCTG-3' and H4 antisense, 5'-CTTTCCCAAGGCCTTTACCACC-3'. One tenth of the immunoprecipitated DNA and one fiftieth of the input DNA was used for each PCR. Histone cell extraction and acid-urea gel electrophoresis has been described elsewhere (2, 15).

Reverse transcription-PCR and dot blot analysis. Subconfluent L929sA fibroblasts were treated as indicated, and total cellular RNA was isolated with the acid-guanidinium-thiocyanate-phenol chloroform method using the Trizol reagent (Invitrogen). Reverse transcription was done on 5 µg total RNA to prepare cDNA for a conventional reverse transcription-PCR on IL6 (sense, 5'-GGAGTACCATAGCTACCTGG-3' and antisense, 5'-GACCACAGTGAGGAATGTCC-3'; amplicon, 331 bp) and glyceraldehyde-3-phosphate dehydrogenase (sense, 5'-GTCCATGCCATCACTGCCA-3' and antisense, 5'-GTGGGAGTTGCTGTTGAAG-3'; amplicon, 342 bp). PCR conditions applied were as follows: at 94°C for 4 minutes, 29 cycles at 62.5°C for 30 seconds, at 72°C for 30 seconds, and at 94°C for 45 seconds; final annealing for 1 minute; final elongation for 5 minutes. For dot blot analysis, the obtained total RNA was denatured in formaldehyde mix at 60°C and as a dilution series spotted on nylon filters (Hybond membranes, Amersham Pharmacia Biotech, Piscataway, NJ). After fixation by UV cross-linking, the filter was hybridized with a specific 32P-IL6 probe at 42°C for 24 hours.

Superarray analysis. The mouse NF-{kappa}B signaling pathway gene array and human breast cancer and ER signaling gene array (GEA array) kits were obtained from SuperArray, Inc. (Bethesda, MD) and used according to the manufacturer's instructions (2). Quantification and normalization of the obtained hybridization signals was done using Phosphor-Imager and SuperArray software. Comparison of the GEArray and RT-PCR or Northern results revealed good correlation and confirmed signal specificity.

Supplementary array data analysis. We analyzed published raw data sets of 78 breast cancer patients, of which 44 have good prognosis signature and 34 have bad prognosis signature (ref. 19; data sets and details regarding sample selection, preparation, and expression profiling are freely available at http://www.rii.com/publications/2002/vantveer.htm). The accompanying Microsoft Excel spreadsheet ArrayData_less_than_5yr.xls and ArrayData_greater_than_5yr.xls contains pages with actual results of ~24,500 gene measurements. Data sets for various cytokines, NF-{kappa}B signaling players, hormone receptors, and kinases were extracted and log10(ratio), the mean ratio of the intensities of the red and green channels that reflect the extent of induction or repression of a given gene, of the gene of interest of the various patients are represented in scatter column graph plots with indication of the calculated means of good and worse prognosis patient populations. Increasing or decreasing trends are indicated by arrows, whereas unaffected targets are indicated with an equation sign.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Phytoestrogenic isoflavones genistein, daidzein, and biochanin A but not synthetic estrogen 17ß-estradiol inhibit endogenous IL6 gene induction in response to TNF. Because estrogens can elicit stimulatory as well as repressive effects on NF-{kappa}B–dependent gene expression in a cell type– and gene-specific way (11), we measured the effects of various (phyto)estrogens (i.e., genistein, daidzein, and biochanin A versus 17ß-estradiol) on TNF-induced IL6 gene expression in L929sA mouse fibroblasts. As we and others detect at least 100-fold weaker estrogenic activity of isoflavones than 17ß-estradiol to elicit comparable estrogenic activities (refs. 20, 21; data not shown), phytoestrogen concentrations were increased ~100-fold compared with 17ß-estradiol to reach a similar ER hormone efficacy in subsequent experiments. Secreted IL6 protein levels present in the supernatants were quantified by IL6 ELISA. L929sA cells were treated with TNF alone or in combination with different concentrations of soy isoflavones compared with the synthetic estrogen 17ß-estradiol or the reference glucocorticoid hormone dexamethasone (22). In Fig. 1A , strongly elevated levels of IL6 protein are detected after TNF treatment as expected. Interestingly, whereas low micromolar concentrations of the reference hormone dexamethasone potently inhibit TNF-induced IL6 gene expression, and phytoestrogenic isoflavones are weakly repressing, 17ß-estradiol hormone seems completely ineffective (Fig. 1A; data not shown). Clearly, a specific and dose-dependent inhibition of IL6 production can be observed upon cotreatment of the various isoflavones, genistein, daidzein, and biochanin A in the micromolar range (Fig. 1A), whereas high micromolar concentrations of pure dexamethasone or 17ß-estradiol were found to be cytotoxic at these supraphysiologic hormone doses (data not shown).


Figure 1
View larger version (32K):
[in this window]
[in a new window]
  		Figure 1. Regulation of NF-{kappa}B–driven gene expression by soy isoflavones. A, L929sA cells were pretreated for 2 hours with reference hormone compounds dexamethasone (2 µmol/L) or 17ß-estradiol (2 and 20 µmol/L), or various doses of the soy isoflavones genistein, daidzein, biochanin A (200, 100, 50, and 20 µmol/L) followed by 6 hours of treatment with 2,000 IU/mL TNF. Corresponding levels of secreted IL6 protein were quantified by mIL6 ELISA. B, L929sA cells were pretreated for 2 hours with genistein (200 µmol/L) or 17ß-estradiol (2 µmol/L) followed by 6 hours of treatment with 2,000 IU/mL TNF. Total RNA was isolated and semiquantitative RT-PCR was done with mouse-specific IL6 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) primer sets. Alternatively, dose response of genistein (micromolar concentrations as indicated) on TNF-induced IL6 mRNA levels was revealed by dot blot analysis upon mIL6 cDNA probe hybridization to total RNA of various treatments, spotted as a 1:3 dilution series on a Hybond nylon filter. Presence of ER{alpha}/ß mRNA was revealed by RT-PCR on total RNA of L929sA cells with isoform-specific mouse primers, negative RT-PCR control without cDNA was done in the same experiment. C, L929sA fibroblasts were stably transfected with reporter gene plasmids controlled by the natural IL6 promoter (p1168hu.IL6P-luc+), a synthetic NF-{kappa}B–responsive promoter [p(IL6-{kappa}B)350hu.IL6P-luc+] or the constitutive phospho-glycerokinase promoter (pPGKßgeobpA), referred to as IL6, 3xNF-{kappa}B, and PGK, respectively. L929sA transfectants were pretreated for 2 hours with genistein, daidzein, or biochanin A (200 µmol/L) followed by 6 hours of treatment with 2,000 IU/mL TNF upon which lysates were prepared for quantification of luciferase and/or galactosidase reporter gene levels. After normalization as described previously (17), the induction factor is defined as the amount of luciferase produced in treated cells compared with untreated cells (the latter is arbitrarily set to 1). D, L929sA cells, stably transfected with the synthetic NF-{kappa}B–responsive promoter construct p(IL6-{kappa}B)350hu.IL6P-luc+, were pretreated for 2 hours with either genistein (200 µmol/L), 17ß-estradiol (1 µmol/L), OH-tamoxifen (1 µmol/L), ICI 182780 (1 µmol/L), genistein + ICI 182780, ciglitazone (30 µmol/L), tyrphostin A23 (50 µmol/L), rolipram (200 µmol/L), NAC (30 mmol/L), BHA (200 µmol/L), wortmannin (100 nmol/L), or LY294002 (20 µmol/L) followed by 6 hours of treatment with 2,000 IU/mL TNF, upon which lysates were prepared for quantification of luciferase levels. Corresponding luciferase expression levels are represented as bar graphs.

 
Briefly, although synthetic estrogens may fail to specifically transrepress NF-{kappa}B in L929sA fibroblasts in a hormone-dependent way, soy isoflavones can dose-dependently repress IL6 gene expression. These results are confirmed at the mRNA level because IL6 RT-PCR and blot assays reveal a similar dose-dependent (range, 2-200 µmol/L) and specific decrease in IL6 mRNA levels upon exposure to genistein but not in the presence of 17ß-estradiol (Fig. 1B). The lack of estrogen-dependent NF-{kappa}B transrepression is remarkable and will be discussed further throughout the results, as ER{alpha} and ERß mRNA can both be detected in L929sA fibroblasts (Fig. 1B).

Isoflavones genistein, daidzein, and biochanin A inhibit NF-{kappa}B–driven reporter gene expression in response to TNF. To verify whether IL6 gene repression by isoflavones is reflected at the transcriptional level, various promoter reporter gene constructs [i.e., p1168hu.IL6P-luc+ or p(IL6-{kappa}B)350hu.IL6P-luc+, and pPGKßGeobpA], containing, respectively, the natural IL6 promoter, a synthetic promoter with multimerized NF-{kappa}B–responsive elements and the housekeeping promoter phosphoglycerate kinase (PGK), were stably transfected into L929sA cells (or Sertoli TM4 as indicated in figure legends). The resulting stable cell pools were identically treated, and the lysates were assayed for corresponding reporter gene activity (Fig. 1C-D; Supplementary Fig. S1). Enhanced luciferase expression levels were measured in response to TNF, whereas cotreatment with various phytoestrogens consistently decreased IL6 promoter activity, and more specifically, NF-{kappa}B–driven reporter gene activity, thus mimicking endogenous IL6 gene regulation. The promoter specificity of the observed regulatory effects is further shown by the housekeeping promoter PGK, which remained unaffected by the different stimulating agents used. In line with our results obtained in Fig. 1A, 17ß-estradiol does not inhibit NF-{kappa}B–driven gene expression in L929sA fibroblasts, although dexamethasone is strongly repressing NF-{kappa}B at the same hormone concentrations. Interestingly, in Sertoli TM4 cells, significant transrepression can be observed with both estrogen and glucocorticoid hormones, pointing to cell type–dependent regulation of ER transrepression (Supplementary Fig. S1). Upon Western blot analysis of ER{alpha}/ß expression levels in both cell types, it becomes clear that lack of ER-dependent NF-{kappa}B effects in L929sA fibroblasts may originate from the limiting trace amounts of ER{alpha}/ß present compared with the corresponding protein levels observed in TM4 (Supplementary Fig. S1). In contrast, soy isoflavones can potently inhibit NF-{kappa}B-dependent reporter gene activity in both cell types (Fig. 1C; data not shown).

To further evaluate if genistein-dependent NF-{kappa}B repression is a superposition of multiple effects, ranging from hormone-like activities (ER{alpha}/ß and PPAR{gamma}), antioxidant properties, inhibition of tyrosine kinases, of PI3K/Akt kinases, and/or of phosphodiesterases (PDE4; see Introduction), we have measured NF-{kappa}B effects in presence of a panel of reference compounds related to these activities. Inhibitor doses presented in Fig. 1D reflect optimized doses that give specific effects (but leave PGK housekeeping promoter activity unaffected). Upon further exploring hormone-like activities, no significant NF-{kappa}B repression could be measured in the presence of an ER agonist (17ß-estradiol), a SERM with mixed agonist/antagonist properties (OH-tamoxifen), or a PPAR{gamma} agonist (ciglitazone). In addition, an ER antagonist (ICI 182780) was not able to reverse genistein-dependent NF-{kappa}B repression. Furthermore, as wortmannin, LY294002, and rolipram also fail to potently repress NF-{kappa}B at optimal doses tested, PI3K and PDE4 are presumably no major players involved in NF-{kappa}B–driven gene expression in mouse fibroblasts either. In contrast, the tyrosine kinase inhibitor tyrphostin A23 and the antioxidants NAC and BHA are able to potently inhibit NF-{kappa}B–driven reporter gene expression to the same extent as genistein. Whether the antioxidant and tyrosine kinase inhibitor properties of genistein may mimic tyrphostin and antioxidant effects on NF-{kappa}B by a common repression mechanism, will be addressed in further experiments.

Isoflavones inhibit NF-{kappa}B–driven gene expression by attenuation of the ERK-MAPK/MSK1 cascade. As NF-{kappa}B–driven gene expression requires a coordinated interplay of IKK activation and MAPK (MSK)–dependent transactivation mechanisms (2, 23–26), we evaluated to which extent genistein-dependent IL6 gene repression relies on inhibition of either pathway in L929 mouse fibroblasts. The first control level concerns the cytoplasmic regulatory event, in which NF-{kappa}B is released from its physiologic inhibitor I{kappa}B upon its degradation. The second level affects MAPK signaling–dependent enhanceosome dynamics in relation to the surrounding chromatin environment.

In this respect, L929sA cells were treated with TNF alone or in combination with genistein for the indicated time points and cell extracts were subsequently analyzed by Western blot for protein expression levels of I{kappa}B, or the activation status of various kinases (IKK{alpha}/ß, p38, ERK, MEK1, and MSK) by use of phospho-specific antibodies. Furthermore, complementary to I{kappa}B Western analysis, release of NF-{kappa}B followed by nuclear DNA binding is revealed by EMSA with an NF-{kappa}B–specific oligonucleotide probe. From Fig. 2A , it seems that in mouse fibroblasts, isoflavones do not affect NF-{kappa}B activation by the IKK pathway, as the p-IKK activation pattern, and I{kappa}B degradation and resynthesis kinetics upon TNF stimulation are similar in presence or absence of genistein. These results are completely in line with the EMSA data, which reveal induced NF-{kappa}B p50-p65/DNA-binding upon TNF stimulation that remains unaffected in presence of genistein (Fig. 2B). The amount of constitutively binding factor RBP-J{kappa} remained unchanged under the various conditions evaluated. Characterization of the various inducible and constitutive transcription factor binding complexes at the IL6 NF-{kappa}B site has previously been shown by extensive supershift analysis and binding competition experiments (17, 27). Interestingly, in contrast to results obtained in L929sA, we and others found significant suppression of NF-{kappa}B/DNA binding in other cell types pointing to cell specific thresholds for IKK inhibition in response to anti-inflammatory phytochemicals (data not shown).


Figure 2
View larger version (43K):
[in this window]
[in a new window]
  		Figure 2. Effect of soy isoflavones on NF-{kappa}B/DNA binding and MAPK-MSK activation. A, L929sA cells were either or not pretreated for 2 hours with genistein followed by treatment with 2,000 IU/mL TNF for the indicated times. Total cell lysates were analyzed for P-IKK{alpha}/ß and I{kappa}B{alpha} levels by Western blot analysis. B, L929sA cells were either or not pretreated for 2 hours with genistein followed by 30 minutes of treatment with 2,000 IU/mL TNF. Nuclear cell lysates were incubated with a 32P-labeled IL6 {kappa}B site-containing probe. Binding complexes formed were analyzed by EMSA. Loading of equal amounts of protein was verified by comparison with the binding activity of the repressor molecule RBP-J{kappa} (27). C to G, serum-starved L929sA fibroblasts or (D) SKBR3 (ER{alpha}–/ß–) breast cancer cells, were either or not pretreated for 2 hours with 200 µmol/L genistein or biochanin A, 2 µmol/L estradiol, 50 µmol/L A23, or 30 mmol/L NAC followed by treatment with 2,000 IU/mL TNF for the indicated times. Cell lysates were analyzed for P-p38 and P-ERK MAPK, P-MEK1/2, and P-MSK1 by phospho-specific Western blot analysis. As a control for equal protein loading, blots were developed against constitutive p38, ERK kinase levels. E, similarly, P-MSK levels, detected in a prolonged time kinetics experiment of L929sA cells exposed to TNF alone or in combination with genistein, were revealed by Western blot analysis and quantified by Image J software (open source Image J software available at http://rsb.info.nih.gov/ij/). Signal intensities are plotted in function of time and corrected for protein loading by normalisation with the constitutive MSK signal. F, along the same line, P-MSK Western signals detected after 15' TNF treatment, in the absence or presence of different doses of genistein or biochanin A, were again quantified by Image J software and normalized for loaded protein levels. G, similar experiments were done in presence of 2 µmol/L 17ß-estradiol, 30 mmol/L NAC, or 50 µmol/L A23 followed by treatment with 2000 IU/mL TNF for the indicated times in serum-starved L929sA fibroblasts. Cell lysates were again analyzed for P-MSK1 by phospho-specific Western blot analysis. As a control for equal protein loading, blots were developed against p65 (data not shown).

 
However, although genistein effects are absent in cytoplasmic NF-{kappa}B activation and DNA binding in L929sA cells, important effects can be observed on the NF-{kappa}B transactivation pathways. We previously established that the TNF-activated p38 and ERK MAPK pathways contribute to the transactivation capacity of NF-{kappa}B (17), by driving MSK1-mediated phosphorylation of p65 Ser276 (2). We now found that genistein attenuates TNF-induced ERK activation, whereas p38 activation remains unaffected (Fig. 2C). Consequently, MSK1 activation is hampered upon cotreatment of TNF and genistein in a time-dependent fashion (Fig. 2D and E). Upstream of ERK MAPK, we also observe inhibition of MEK1 activation by genistein, which may contribute to the reduction in ERK activity (Fig. 2D). Of particular interest is the observation that the MAPK/MSK pathways are also decreased by isoflavones in ER{alpha}/ß-deficient SKBR3 breast cancer cells (Fig. 2D), which again suggests that attenuation of MSK1 activation by phytoestrogens does not require classic estrogen (receptor) activity (Fig. 2D). This is further corroborated by the observation that in analogy to genistein, p-MSK levels can be reduced with antioxidant compounds or tyrphostins (i.e., NAC and A23; Fig. 2G), in line with the reporter gene results (Fig. 1D). Finally, upon further testing different doses of the most potent isoflavones, significant repression of MSK1 activation could be measured with a minimal dose of 12.5 µmol/L biochanin A and 50 µmol/L genistein (Fig. 2F), suggesting subtle differences in structure requirements to elicit inhibition of MSK signaling pathways, this being in line with the different immunosuppressive potencies of biochanin A > genistein > daidzein isoflavones.

Isoflavones decrease NF-{kappa}B p65 transactivation by interfering with MSK1 kinase activity and histone acetyltransferase/histone deacetylase cofactor activities. Further proof for soy isoflavones in targeting NF-{kappa}B transactivation via the MSK1 pathway comes from an immunoprecipitation MSK kinase assay and the Gal4 one hybrid technique. Effects of TNF and/or genistein on endogenous MSK1 activity can be measured by a kinase assay of MSK1 immunoprecipitates from cell lysates (i.e., of cells treated with TNF alone, or in combination with genistein or 17ß-estradiol). The Gal4 one-hybrid system is a read-out assay for NF-{kappa}B p65 transactivation activity, independent of I{kappa}B. Before, we have established the crucial link between MSK-dependent NF-{kappa}B p65 S276 phosphorylation and TNF-induced pGal4-p651-551 transactivation driving the reporter gene construct p(GAL4)2-50hu.IL6-luc+ (2). Upon comparison of MSK1 kinase activity (Fig. 3A ) and Gal4-p65 transactivation potency (Fig. 3B), a tight correlation in responses can be observed in line with previous findings (2). TNF treatment clearly up-regulates MSK1 kinase activity and p65 transactivation, which can be completely abrogated in presence of the MAPK inhibitor cocktail SB205380 and U0126. As genistein only blocks ERK, but not p38 pathways (Fig. 2C and D), only partial reduction of MSK kinase activity can be observed (Fig. 2D and E), but to a similar extent as treatment with the ERK inhibitor U0126 alone, whereas 17ß-estradiol is completely ineffective (Fig. 3A). These results perfectly mirror the pattern obtained in the Gal4-p65 transactivation results.


Figure 3
View larger version (16K):
[in this window]
[in a new window]
  		Figure 3. Soy isoflavones affect MSK activity, NF-{kappa}B transactivation, and cofactor activity. A, L929sA cells were starved for 48 hours in serum-free medium, and cells were left untreated or were pretreated for 2 hours with the MAPK inhibitors SB203580 (10 µmol/L) and/or U0126 (10 µmol/L), genistein (200 µmol/L), or 17ß-estradiol (2 µmol/L) followed by 30 minutes of treatment with 2,000 IU/mL TNF. Cells were lysed, and endogenous MSK1 was isolated by immunoprecipitation. The corresponding activity of MSK1 was assessed by an in vitro kinase assay on p65-S276-peptide. B, pools of L929sA cells stably expressing Gal4-p65 were transiently transfected with p(Gal4)2-50hu.IL6-luc+. At 48 hours after transfection, cells were left untreated or were pretreated for 2 hours with the MAPK inhibitors SB203580 (10 µmol/L) and/or U0126 (10 µmol/L), genistein (200 µmol/L), or 17ß-estradiol (2 µmol/L) followed by 6 hours of treatment with 2,000 IU/mL TNF. After normalization, the corresponding induction factors are represented as bar graphs. C, HEK293T cells were transiently transfected with a combination of expression plasmids (i.e., 300 ng p1168hu.IL6-luc+, 20 ng pPGKßgeobpA, 20 ng pRcRSVp65, and/or 80 ng CBP expression plasmid). The total amount of DNA was kept constant in all setups by supplementing empty vector DNA. Cells were either or not treated for 16 hours with genistein (200 µmol/l), starting at 32 hours after transfection, and all transfected setups were lysed at time point 48 hours. Corresponding luciferase expression levels in lysates are represented as bar graphs and have been normalized for protein concentrations and transfection efficiency. D, L929sA were stably transfected with a reporter gene plasmid controlled by the natural IL6 promoter (p1168hu.IL6P-luc+). Transfectants were untreated or pretreated for 2 hours with genistein (200 µmol/L) followed by 6 hours of treatment with 2,000 IU/mL TNF, trichostatin A (TSA; 100 nmol/L), or both together. Corresponding normalized induction factors are represented as bar graphs.

 
Another aspect of nuclear regulation of NF-{kappa}B activity relies on its interactions with chromatin-modifying cofactor complexes (23, 28–30). Transcription factors mediate gene activation or repression via recruitment of "tandem" cofactor complexes of CBP/p300, histone acetyltransferases (HAT), and RSK2/MSK1 kinases (31, 32), or histone deacetylases (HDAC) and phosphatases PP1, respectively (33). Previously, we have shown the crucial role of the NF-{kappa}B p65 subunit in engaging CBP/p300 and HAT activity for transcriptional activation of the IL6 promoter (18), which strongly depends on MSK-dependent p65 Ser276 phosphorylation (2). In this respect, we have investigated the ability of CBP to synergize with NF-{kappa}B p65 in absence or presence of genistein by means of transient cofactor transfection assays or reporter gene experiments with stable chromatinized promoters in presence of the HDAC inhibitor trichostatin A. Subsaturating amounts of expression plasmids of p65 and/or CBP were transiently transfected to HEK293T cells together with the IL6 reporter gene construct p1168hu.IL6P-luc+. As is clear from Fig. 3C, NF-{kappa}B–dependent gene expression is strongly increased if CBP is coexpressed with p65 but reversed upon prior addition of genistein. Expression levels of p65 or CBP remained unaffected under the tested conditions as confirmed by Western analysis (data not shown). In a more physiologic approach with trichostatin A on a chromatinized IL6 promoter, trichostatin A augments NF-{kappa}B–driven but not basal gene expression in response to TNF, as expected (ref. 18; Fig. 3D). Interestingly, genistein is able to completely counteract this synergy. These results indicate that genistein is able to block NF-{kappa}B triggered CBP/p300 HAT activity by inhibiting MSK1-dependent phosphorylation of NF-{kappa}B p65 S276 (2). Similar results were obtained with other isoflavones (data not shown).

Isoflavones interfere with MSK1 recruitment and consequent factor phosphorylation and chromatin dynamics at IL6 gene promoters. Considering that genistein attenuates the TNF-activated MSK1 pathway and that MSK1 kinase substrates include transcription factors (i.e., NF-{kappa}B p65 Ser276 and cAMP-responsive element binding protein Ser133) as well as chromatin protein histone H3 Ser10 (2, 34), we have investigated isoflavone effects on NF-{kappa}B or H3 phosphorylation by phospho-specific Western analysis. Interestingly, in serum-starved L929sA fibroblasts treated with genistein, a significant reduction of basal and TNF-induced NF-{kappa}B Ser276 and H3 Ser10 phosphorylation can be observed (Fig. 4A and B ). A similar reduction can be detected with biochanin A and to a lesser extent daidzein but not with 17ß-estradiol (Fig. 4B; data not shown). Total cellular levels of NF-{kappa}B p65 and H3 Lys9 dimethylation remained unaffected under the various conditions tested, pointing to the specificity of the isoflavone effect. To further confirm our results on an IL6 gene–specific basis, we did chromatin immunoprecipitation against various histone H3 modifications and/or MSK1 at the level of the IL6 gene promoter. As chromatin immunoprecipitation grade p-NF-{kappa}B S276 antibodies are currently not available, we were unable to include this setup in our further chromatin immunoprecipitation experiments, although a similar pattern can be expected as with the p-H3 S10 chromatin immunoprecipitations. More particularly, we found increased phospho(acetylated) H3 levels at the endogenous IL6 promoter following TNF stimulation, but these effects are completely lost in the presence of genistein (Fig. 4C). The complete inhibition of histone phospho(acetyl) modifications is remarkable because MSK1 kinase activity could only be partially inhibited by isoflavones (Fig. 3A). However, upon investigating MSK1 recruitment at the IL6 promoter under the same conditions, we did observe a complete loss of TNF-induced MSK1 recruitment at the IL6 promoter, explaining complete disappearance of histone H3 (phospho)acetylation in the presence of genistein, comparable with treatment with the MAPK inhibitor cocktail SB205380 + PD98059 (Fig. 4D). Similar results were obtained with other isoflavones (Supplementary Fig. S2). This suggests that genistein may have cumulative effects on MSK-driven gene expression by interfering with MSK1 activation and kinase activity, as well as with MSK recruitment on chromatinized promoters. As the spatial context in which MAPK/MSK kinases operate in transcription complexes is poorly understood, future studies of subcellular dynamics of MAPK/MSK regulation will be of high interest (35).


Figure 4
View larger version (32K):
[in this window]
[in a new window]
  		Figure 4. Soy isoflavones strongly reduce MSK1-effects at the IL6 gene promoter. A to B, L929sA cells were serum starved and left untreated or pretreated for 2 hours with genistein, daidzein, biochanin A (200 µmol/L), or 17ß-estradiol (2 µmol/L) followed by 30 minutes of treatment with 2,000 IU/mL TNF. p-NF-{kappa}B Ser276 (A) and histone H3 (B) levels are revealed by phospho-specific Western analysis. Equal protein loading is shown by Western detection of corresponding NF-{kappa}B p65 or methylated histone H3 levels. C to D, serum-starved L929sA mouse fibroblasts were treated for 30 minutes with 2,000 IU/mL TNF alone, or following 2 hours of pretreatment with the inhibitors SB203580 + PD98059 (10 µmol/L) or H89 (10 µmol/L) or genistein (200 µmol/L). Chromatin immunoprecipitation analysis was done against phosphorylated (S10) or phosphoacetylated (K9-S10) histone H3 (C) or MSK1 (D). After reversal of cross-linking, coimmunoprecipitated genomic DNA fragments were analyzed by quantitative PCR for 27 cycles with IL6 or H4 promoter-specific primer sets and is revealed by gel electrophoresis. Input reflects the relative amounts of sonicated DNA fragments present before immunoprecipitation and revealed by quantitative PCR with either IL6- or H4-specific primers.

 
Isoflavones selectively inhibit NF-{kappa}B–dependent gene expression. To obtain final proof for involvement of the MSK pathway in isoflavone effects on NF-{kappa}B–driven gene expression, we compared gene expression patterns of TNF-treated wild-type and MSK1/2 double-knockout mouse embryonal fibroblasts (2, 36, 37), in comparison with TNF- and TNF + genistein–treated L929sA fibroblast cells (Fig. 5 ). Upon analysis of various NF-{kappa}B superarray filter hybridizations, a remarkable similarity in gene expression regulation can be observed between IL6 or VCAM mRNA expression, affected by genistein treatment (Fig. 5B and D), or by knocking out MSK1/2 (Fig. 5A and C). In contrast, the target gene NFKB2 that is also strongly induced by TNF seems to be insensitive to genistein treatment and in parallel to genetic disruption of MSK1/2. Finally, 17ß-estradiol treatment has only minor effects on NF-{kappa}B gene expression in fibroblasts in agreement with our results discussed above.


Figure 5
View larger version (45K):
[in this window]
[in a new window]
  		Figure 5. Similarities in NF-{kappa}B–driven gene expression pattern in MSK KO (A and C) and genistein-treated cells (B and D). Serum-starved MSK1–/–/MSK2–/– and wild-type MEF cells were treated with 2,000 IU/mL TNF for 4 hours (A and C). Alternatively, serum-starved L929sA cells were treated for 4 hours with 2,000 IU/mL TNF alone, or following a 2-hour pretreatment with genistein (200 µmol/L) or 17ß-estradiol (2 µmol/L; B and D). Total RNA was isolated and analyzed using NF-{kappa}B GEArray technology according to the manufacturer's instructions. Specific mRNA expression was normalized for loading differences with housekeeping gene signals. Spot intensities of IL6, VCAM1, and NFKB2 are marked in the array figures (A and B), and the corresponding signal intensities quantified by phosphor-imager analysis are represented as bar graphs (C and D).

 
Genistein interferes with MSK1 signaling in ER{alpha}-deficient breast cancer cells. Because soy bean–derived isoflavones have been associated with reduced rates of breast cancer (1) and are recommended as HRT alternative for people with increased breast cancer risk or breast cancer survivors (5, 38), we next investigated whether isoflavones may have similar effects in a breast cancer model. Upon reinvestigation of raw array data sets of breast cancer patients among which 44 with good and 34 with worse prognosis signature (19), a remarkable increase can be observed in MSK1, IL6, and IL-8 but not in NFKB2 gene expression levels in the worse prognosis population. Remarkably, these data coincide with loss of ER{alpha} (Supplementary Fig. S3) and are in line with previous reports that classify IL6 and IL-8 as metastasis-promoting cytokines in advanced breast cancer (1). Similarly, we and others have observed a strong increase in IL6 gene expression in the invasive metastatic breast cancer cells MDA-MB231 (ER{alpha}–/ß+) compared with noninvasive MCF7 (ER{alpha}+/ß+) cells (Fig. 6A ; refs. 39, 40). Besides a very strong TNF-induced NF-{kappa}B activation in MDA-MB231 cells compared with MCF7 cells (Fig. 6B), we do also observe a constitutive MSK phosphorylation, NF-{kappa}B/DNA-binding activity (Fig. 6B and C), and occupation of the IL6 gene promoter by p65 and MSK1 in the basal state in MDA-MB231, as revealed by IL6 promoter-specific chromatin immunoprecipitation analysis (Fig. 6D). Finally, in line with our results obtained in fibroblasts, genistein is able to significantly reduce IL6 mRNA gene expression levels, MSK1 recruitment, and H3 phosphorylation at the IL6 promoter in the metastatic MDA-MB231 cells (Fig. 6E and F). This suggests that soy isoflavones may protect against aggressive breast cancer progression by attenuation of MSK1 activities, irrespective of the cellular ER status.


Figure 6
View larger version (35K):
[in this window]
[in a new window]
  		Figure 6. Genistein affects MSK-dependent IL6 gene expression in ER{alpha}-negative breast cancer cells. A, to compare IL6 gene expression levels in MCF7 and MDA-MB231 cells, both cell types were left untreated or TNF-treated (2,000 IU/mL) for 6 hours. Corresponding levels of secreted IL6 protein were quantified by hIL6 ELISA. B and C, similarly, MCF7 and MDA-MB231 cells were left untreated or TNF-treated (2,000 IU/mL) for 30 minutes, and nuclear lysates were analyzed for NF-{kappa}B/DNA-binding activity (B) and P-MSK levels (C) by means of EMSA or Western analysis, respectively. For the latter, equal loading was detected by an anti-actin antibody. D, In addition, the constitutive occurrence of NF-{kappa}B and MSK1 in untreated MDA-MB231 or MCF7 cells were revealed by chromatin immunoprecipitation analysis against NF-{kappa}B p65 and MSK1 at the endogenous IL6 gene promoter. E, finally, we measured the effects of 200 µmol/L genistein on TNF-stimulated H3 phosphorylation and MSK recruitment on the IL6 gene promoter by ChIP analysis in serum-starved MDA-MB231 cells, treated with 2,000 IU/mL TNF for 30 minutes, either or not after a 2-hour pretreatment with genistein (200 µmol/L). F, to determine genistein effects at IL6 mRNA gene expression in MDA-MB231, cells were TNF treated with 2,000 IU/mL for 4 hours either or not following a 2-hour pretreatment with genistein (200 µmol/L) or 17ß-estradiol (2 µmol/L). Total RNA was isolated and analyzed using GEArray technology according to the manufacturer's instructions. Specific mRNA expression was normalized for loading differences, quantified by phosphor-imager, and represented as bar graphs according to hybridization intensity.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The isoflavones genistein, daidzein, and biochanin A, which are abundant in soybeans and widely available as herbal tablets, are especially popular among postmenopausal women. Only lately, many studies have been undertaken to unveil the mode by which phytoestrogens mediate their NF-{kappa}B–suppressive effects. Our results in estrogen-unresponsive fibroblasts, MSK knockout cells, and ER-deficient breast tumor cells show that phytoestrogenic isoflavones but not 17ß-estradiol selectively block nuclear NF-{kappa}B transactivation of particular NF-{kappa}B target genes, including IL6. This occurs via attenuation of MEK1 and ERK activity, which slows down MSK1-dependent NF-{kappa}B p65 and histone H3 phosphorylation.

Glucocorticoids are able but 17ß-estradiol fails to inhibit NF-{kappa}B–driven IL6 gene expression in L929sA fibroblasts at the same hormone concentrations. Along the same line, glucocorticoids but not 17ß-estradiol are able to stimulate GR- or ER-driven reporter gene activity, respectively, in L929sA cells. This suggests that L929sA mouse fibroblasts may lack functional ER, which is required to mediate NF-{kappa}B crosstalk. Nowadays, constitutive growth factor signaling and kinase cascades, as well as elevated threshold levels of NF-{kappa}B, have all been shown to drastically affect ER functionality by interfering with its ligand sensitivity, localization, and turnover rate (11). As ER{alpha}/ß mRNA transcription can clearly be detected by RT-PCR in L929sA fibroblasts (Fig. 1B), but corresponding protein levels are at the detection limit (Supplementary Fig. S1), we assume that high turnover rates of ER{alpha}/ß in these cells may prevent dose-dependent responses (either transactivation or transrepression) by agonists, antagonists, or mixed SERMS, as can also be observed in other cell types, that result in an estrogen-resistant cell phenotype (5). In this respect, the isoflavone-dependent IL6 gene inhibition in L929sA mouse fibroblasts is most probably independent of classic estrogenic properties. This hypothesis is further strengthened by the observation that the ER antagonist ICI 182780 is unable to modulate NF-{kappa}B–dependent gene expression or reverse the isoflavone effects. Although PPAR{gamma} too has been proposed as a hormone receptor for soy isoflavones (41), we found no evidence for PPAR{gamma} involvement in mediating isoflavone effects because IL6 promoter activity could not be significantly down-regulated by ciglitazone. Interestingly, comparison of other possible properties of isoflavones (12) with small-molecule inhibitors as reference compounds rather suggests potential involvement of the tyrosine kinase inhibitory and/or antioxidant activities by genistein because A23 or NAC were found to inhibit NF-{kappa}B–dependent gene expression to a similar extent as genistein and as such may affect similar regulatory targets.

Further experiments clearly established that isoflavones rather attenuate the NF-{kappa}B transactivation potency than cytoplasmic NF-{kappa}B activation by blocking the ERK/MSK pathway, as shown in estrogen-unresponsive fibroblasts and in ER{alpha}/ß-deficient SKBR3 breast cancer cells. In analogy to genistein, the tyrosine kinase inhibitor A23 and antioxidant NAC were also found to inhibit the ERK and MSK1 pathway. Various receptor tyrosine kinases with an intrinsic, ligand-dependent tyrosine kinase activity stimulate the Ras/Raf-MEK-ERK pathway, which controls fundamental cellular processes, including proliferation, differentiation, survival, and NF-{kappa}B transactivation (42, 43). Furthermore, TNF has also been shown to activate various non-receptor tyrosine kinases (i.e., c-Src, lyn, Pyk2, Eth/Bm, and Syk; refs. 43, 44). In addition, previous studies in L929 cells have also revealed that formation of reactive oxygen intermediates is essential for the IL6 gene–inductive effects of TNF (45), whereas depletion of mitochondrial oxidative metabolism results in inhibition of IL6 gene induction by TNF. Interestingly, the mitochondrial antioxidant system controlling the cellular redox balance is sensitive to growth factor tyrosine kinase activity (46). Altogether, MSK1 may integrate upstream tyrosine kinase and ROS activities via the MAPK signaling cascades, leading towards IL6 gene induction, which itself may thus be sensitive to the inhibitory effects of soy isoflavones at multiple levels.

Although it has been proposed that antioxidants may inhibit TNF-induced NF-{kappa}B activation by lowering the affinity of TNF for its receptor resulting in a general reduction in magnitude of all TNF signaling events (47), this can definitely not be generalized for all kind of antioxidants in every cell type. More particularly, genistein was found to only affect ERK/MSK1 pathways, leaving TNF-induced p38 and IKK activity unaffected. Furthermore, genistein only inhibits particular NF-{kappa}B target genes, such as IL6 and VCAM, whereas other TNF-responsive genes remain unaffected (such as NFKB2). Alternatively, depending on cellular redox changes in the cell, distinct redox forms of TNFRI could be identified with distinct ligand binding, clustering, and signaling ability (48).

At the level of the IL6 promoter enhanceosome, it seems that genistein totally blocks recruitment of the histone H3 kinase MSK1, coinciding with a complete loss of H3 phosphorylation and acetylation, although MSK activity itself is only partially inhibited. This suggests that genistein may have cumulative effects on MAPK/MSK-driven gene expression by interfering with MSK kinase activity as well as with its recruitment on chromatinized promoters. Of particular interest, RSK (and in many cases also MSK) interactions with the acetylase CBP, the phosphatase PP2C, or 14-3-3ß proteins have been found to regulate its subcellular localization and restrict its activities in time and space (31, 49–51). Further experiments will be required to unravel genistein effects on spatiotemporal dynamics of MSK(RSK)-cofactor complexes in relation to selective NF-{kappa}B–dependent gene expression.

Because isoflavone-rich soy diets have been associated with reduced rates of breast cancer (1), we also investigated genistein effects in a breast cancer model. Publically available array data sets from breast cancer patients with good or worse prognosis (19) already illustrate a significant increase in IL6 and IL8 but not NFKB2 gene expression coinciding with higher MSK expression levels and loss of ER{alpha} in breast cancer patients with bad prognosis signature (Supplementary Fig. S3). Along the same line, we observed strongly elevated (but still TNF-inducible) NF-{kappa}B and MSK1 activity coinciding with elevated IL6 gene expression in the aggressive metastatic breast tumor cells MDA-MB231, lacking ER{alpha}, which could still be reversed in presence of genistein. The fact that in breast cancer cells, IL6 gene expression has been associated with mitogenic, angiogenic, metastatic, and invasive responses, whereas phytoestrogens have been found to inhibit these activities may offer interesting therapeutic opportunities for isoflavone compounds. Indeed, as constitutive NF-{kappa}B and MSK activity are hallmarks of aggressive metastatic ER-deficient breast cancer, the MSK pathway may be a relevant therapeutic target for soy phytoestrogens, in case classic hormone therapy fails.

Of special note, genistein plasma concentrations of 3.4 µmol/L in mice or 10 to 25 µmol/L among humans were measured upon oral feeding of genistein (52), which could be considered as a physiologic isoflavone concentration range. Because we observe weak inhibition of endogenous IL6 gene expression in presence of 10 to 50 µmol/L soy isoflavones in L929sA fibroblasts (Fig. 1A and B) and significant inhibition of P-MSK levels with the most potent soy isoflavone biochanin at 12.5 µmol/L, or mild isoflavone genistein around 50 µmol/L, the experimental concentration range applied is within reach of a physiologic frame. In a real life diet, responses to physiologic-dose phytochemicals might reveal synergistic effects, in combination with other classes of functional foods (53). To limit excessive IL6 production to a health-beneficial concentration range, a partial decrease of MSK activity at physiologic isoflavone concentrations may be sufficient to elicit chemopreventive effects, whereas a complete block of IL6 gene expression at elevated isoflavone concentrations may be detrimental for immune homeostasis.

Furthermore, long-term (time frame of various months or years) chemopreventive effects, upon daily exposure to an isoflavone-rich diet, may be difficult to mimic in short-term (timeframe of a few hours) tissue culture experiments at similar concentrations and may require higher doses of individual compounds to reveal molecular black and white effects in the latter case. Further pharmacologic studies are required with respect to bioavailability and metabolism of soy isoflavones to determine health beneficial doses at short-term/long-term periods, which may interfere with pathologic MSK1 activity in various cell types in vivo. Finally, chemopreventive (daily diet) or chemotherapeutic use (medicinal application to hormone-resistant tumors) of isoflavone preparations may require distinct dose ranges to fine-tune MSK activities.

In summary, our results show that soy phytoestrogens in contrast to 17ß-estradiol, can counteract MSK-dependent NF-{kappa}B transactivation on specific NF-{kappa}B target genes in estrogen-unresponsive fibroblasts and ER-defective breast tumor cells, presumably via inhibition of tyrosine kinases or by their antioxidant capacity. Structure/function analysis of different isoflavone metabolites could reveal minimal core structure domains involved in its antioxidant, tyrosine kinase inhibitor, or hormone ligand properties. This may finally allow metabolic engineering of superior isoflavone drugs as "selective MSK pathway modulators" with therapeutic benefit in chronic inflammatory disorders, longevity, and/or cancer biology.


   Acknowledgments
 
Grant support: Interuniversitaire Attractiepolen and the Geconcerteerde Onderzoeksacties.

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.

We thank K. Van Wesemael, J. Claeys, and P. Faes for technical expertise; F. Jacquemotte (Institut Meurice, Brussels, Belgium) for synthesis p65 peptide; Drs. J.A. Gustafsson and M. Warner (Karolinska Institute, Huddinge, Sweden) for ERß immunoreagents; Dr. W. D'Hooghe (University Gent, Belgium) for supplying TM4 cells; Drs. P. Cohen and J.S. Arthur and G. Wiggin (University of Dundee, Dundee, United Kingdom) for giving us the opportunity to use the MSK1–/–MSK2–/– MEFs in our studies; Drs. L. Mahadevan and A. Clayton (Department of Biochemistry, Nuclear Signaling Laboratory, University of Oxford, Oxford, United Kingdom) for H3 chromatin immunoprecipitation expertise and immunoreagents; P. Van Damme for excellent help with chromatin immunoprecipitation assays; and other members of the lab for critical discussions.


   Footnotes
 
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

W. Vanden Berghe and N. Dijsselbloem equally contributed to this work.

L. Vermeulen is a fellow with the Stichting tegen kanker. N. Dijsselbloem is a fellow with the Vlaams Instituut voor de Bevordering van het Wetenschappelijk-Technologisch Onderzoek in de Industrie. L. Vermeulen and W. Vanden Berghe are now postdoctoral fellows with the Fonds voor Wetenschappelijk Onderzoek-Vlaanderen.

Due to space restrictions, we apologize for having replaced various original references by more extensive review articles referring to the original work.

Received 8/19/05. Revised 12/27/05. Accepted 3/ 3/06.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Dijsselbloem N, Vanden Berghe W, De Naeyer A, Haegeman G. Soy isoflavone phyto-pharmaceuticals in interleukin-6 affections: multi-purpose nutraceuticals at the crossroad of hormone replacement, anti-cancer and anti-inflammatory therapy. Biochem Pharmacol 2004;68:1171–85.[CrossRef][Medline]
  2. Vermeulen L, De Wilde G, Damme PV, Vanden Berghe W, Haegeman G. Transcriptional activation of the NF-kappaB p65 subunit by mitogen- and stress-activated protein kinase-1 (MSK1). EMBO J 2003;22:1313–24.[CrossRef][Medline]
  3. Erschler WB, Keller ET. Age-associated increased interleukin-6 gene expression, late-life diseases, and frailty. Annu Rev Med 2000;51:245–70.[CrossRef][Medline]
  4. Speroff L. A clinician's review of the WHI-related literature. Int J Fertil Womens Med 2004;49:252–67.[Medline]
  5. Turgeon JL, McDonnell DP, Martin KA, Wise PM. Hormone therapy: physiological complexity belies therapeutic simplicity. Science 2004;304:1269–73.[Abstract/Free Full Text]
  6. Adlercreutz H, Mazur W. Phyto-oestrogens and Western diseases. Ann Med 1997;29:95–120.[Medline]
  7. Surh YJ. Cancer chemoprevention with dietary phytochemicals. Nat Rev Cancer 2003;3:768–80.[CrossRef][Medline]
  8. Gronemeyer H, Gustafsson JA, Laudet V. Principles for modulation of the nuclear receptor superfamily. Nat Rev Drug Discov 2004;3:950–64.[CrossRef][Medline]
  9. Gustafsson JA. What pharmacologists can learn from recent advances in estrogen signalling. Trends Pharmacol Sci 2003;24:479–85.[CrossRef][Medline]
  10. Karin M, Greten FR. NF-kappaB: linking inflammation and immunity to cancer development and progression. Nat Rev Immunol 2005;5:749–59.[CrossRef][Medline]
  11. Kalaitzidis D, Gilmore TD. Transcription factor cross-talk: the estrogen receptor and NF-kappaB. Trends Endocrinol Metab 2005;16:46–52.[CrossRef][Medline]
  12. Dixon RA. Phytoestrogens. Annu Rev Plant Physiol Plant Mol Biol 2004;55:225–61.[CrossRef][Medline]
  13. Surh YJ, Chun KS, Cha HH, et al. Molecular mechanisms underlying chemopreventive activities of anti-inflammatory phytochemicals: down-regulation of COX-2 and iNOS through suppression of NF-kappa B activation. Mutat Res 2001;480–481:243–68.
  14. Aggarwal BB, Takada Y, Oommen OV. From chemoprevention to chemotherapy: common targets and common goals. Expert Opin Investig Drugs 2004;13:1327–38.[CrossRef][Medline]
  15. Clayton AL, Rose S, Barratt MJ, Mahadevan LC. Phosphoacetylation of histone H3 on c-fos- and c-jun-associated nucleosomes upon gene activation. EMBO J 2000;19:3714–26.[CrossRef][Medline]
  16. Warner M, Wang L, Weihua Z, et al. Analysis of estrogen receptor expression in tissues. Methods Enzymol 2003;364:448–63.[Medline]
  17. Vanden Berghe W, Plaisance S, Boone E, et al. p38 and extracellular signal-regulated kinase mitogen-activated protein kinase pathways are required for nuclear factor-kappaB p65 transactivation mediated by tumor necrosis factor. J Biol Chem 1998;273:3285–90.[Abstract/Free Full Text]
  18. Vanden Berghe W, De Bosscher K, Boone E, Plaisance S, Haegeman G. The nuclear factor-kappaB engages CBP/p300 and histone acetyltransferase activity for transcriptional activation of the interleukin-6 gene promoter. J Biol Chem 1999;274:32091–8.[Abstract/Free Full Text]
  19. van't Veer LJ, Dai H, van de Vijver MJ, et al. Gene expression profiling predicts clinical outcome of breast cancer. Nature 2002;415:530–6.[CrossRef][Medline]
  20. Milligan S, Kalita J, Pocock V, et al. Oestrogenic activity of the hop phyto-oestrogen, 8-prenylnaringenin. Reproduction 2002;123:235–42.[Abstract]
  21. Wober J, Weisswange I, Vollmer G. Stimulation of alkaline phosphatase activity in Ishikawa cells induced by various phytoestrogens and synthetic estrogens. J Steroid Biochem Mol Biol 2002;83:227–33.[CrossRef][Medline]
  22. De Bosscher K, Schmitz ML, Vanden Berghe W, Plaisance S, Fiers W, Haegeman G. Glucocorticoid-mediated repression of nuclear factor-kappaB-dependent transcription involves direct interference with transactivation. Proc Natl Acad Sci U S A 1997;94:13504–9.[Abstract/Free Full Text]
  23. Chen LF, Greene WC. Shaping the nuclear action of NF-kappaB. Nat Rev Mol Cell Biol 2004;5:392–401.[CrossRef][Medline]
  24. Viatour P, Merville MP, Bours V, Chariot A. Phosphorylation of NF-kappaB and IkappaB proteins: implications in cancer and inflammation. Trends Biochem Sci 2005;30:43–52.[CrossRef][Medline]
  25. Schmitz ML, Bacher S, Kracht M. I kappa B-independent control of NF-kappa B activity by modulatory phosphorylations. Trends Biochem Sci 2001;26:186–90.[CrossRef][Medline]
  26. Yamamoto Y, Gaynor RB. IkappaB kinases: key regulators of the NF-kappaB pathway. Trends Biochem Sci 2004;29:72–9.[CrossRef][Medline]
  27. Plaisance S, Vanden Berghe W, Boone E, Fiers W, Haegeman G. Recombination signal sequence binding protein Jkappa is constitutively bound to the NF-kappaB site of the interleukin-6 promoter and acts as a negative regulatory factor. Mol Cell Biol 1997;17:3733–43.[Abstract]
  28. Vanden Berghe W, De Bosscher K, Vermeulen L, De Wilde G, Haegeman G. Induction and repression of NF-kB-driven Inflammatory Genes. Ernst Schering Res Found Workshop 2002;40:233–78.[Medline]
  29. Natoli G, Saccani S, Bosisio D, Marazzi I. Interactions of NF-kappaB with chromatin: the art of being at the right place at the right time. Nat Immunol 2005;6:439–45.[CrossRef][Medline]
  30. Hayden MS, Ghosh S. Signaling to NF-kappaB. Genes Dev 2004;18:2195–224.[Abstract/Free Full Text]
  31. Merienne K, Pannetier S, Harel-Bellan A, Sassone-Corsi P. Mitogen-regulated RSK2-CBP interaction controls their kinase and acetylase activities. Mol Cell Biol 2001;21:7089–96.[Abstract/Free Full Text]
  32. Janknecht R. Regulation of the ER81 transcription factor and its coactivators by mitogen- and stress-activated protein kinase 1 (MSK1). Oncogene 2003;22:746–55.[CrossRef][Medline]
  33. Brush MH, Guardiola A, Connor JH, Yao TP, Shenolikar S. Deacetylase inhibitors disrupt cellular complexes containing protein phosphatases and deacetylases. J Biol Chem 2004;279:7685–91.[Abstract/Free Full Text]
  34. Soloaga A, Thomson S, Wiggin GR, et al. MSK2 and MSK1 mediate the mitogen- and stress-induced phosphorylation of histone H3 and HMG-14. EMBO J 2003;22:2788–97.[CrossRef][Medline]
  35. Edmunds JW, Mahadevan LC. MAP kinases as structural adaptors and enzymatic activators in transcription complexes. J Cell Sci 2004;117:3715–23.[Abstract/Free Full Text]
  36. Wiggin GR, Soloaga A, Foster JM, Murray-Tait V, Cohen P, Arthur JS. MSK1 and MSK2 are required for the mitogen- and stress-induced phosphorylation of CREB and ATF1 in fibroblasts. Mol Cell Biol 2002;22:2871–81.[Abstract/Free Full Text]
  37. Arthur JS, Cohen P. MSK1 is required for CREB phosphorylation in response to mitogens in mouse embryonic stem cells. FEBS Lett 2000;482:44–8.[CrossRef][Medline]
  38. Duffy C, Cyr M. Phytoestrogens: potential benefits and implications for breast cancer survivors. J Womens Health (Larchmt) 2003;12:617–31.[Medline]
  39. Patel NM, Nozaki S, Shortle NH, et al. Paclitaxel sensitivity of breast cancer cells with constitutively active NF-kappaB is enhanced by IkappaBalpha super-repressor and parthenolide. Oncogene 2000;19:4159–69.[CrossRef][Medline]
  40. Nakshatri H, Bhat-Nakshatri P, Martin DA, Goulet RJ, Jr., Sledge GW, Jr. Constitutive activation of NF-kappaB during progression of breast cancer to hormone-independent growth. Mol Cell Biol 1997;17:3629–39.[Abstract]
  41. Dang ZC, Audinot V, Papapoulos SE, Boutin JA, Lowik CW. Peroxisome proliferator-activated receptor gamma (PPARgamma) as a molecular target for the soy phytoestrogen genistein. J Biol Chem 2003;278:962–7.[Abstract/Free Full Text]
  42. Chen G, Goeddel DV. TNF-R1 signaling: a beautiful pathway. Science 2002;296:1634–5.