Molecular Cancer Research Genome w/Abstract AM No Date
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
Molecular Cancer Research 5, 833-845, August 1, 2007. doi: 10.1158/1541-7786.MCR-07-0054
© 2007 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
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 Google Scholar
Google Scholar
Right arrow Articles by Matteucci, E.
Right arrow Articles by Desiderio, M. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Matteucci, E.
Right arrow Articles by Desiderio, M. A.

Signaling and Regulation

c-Src/Histone Deacetylase 3 Interaction Is Crucial for Hepatocyte Growth Factor–Dependent Decrease of CXCR4 Expression in Highly Invasive Breast Tumor Cells

Emanuela Matteucci, Elisa Ridolfi, Paola Maroni, Paola Bendinelli and Maria Alfonsina Desiderio

Institute of General Pathology, University of Milan, via Luigi Mangiagalli, Milan, Italy

Requests for reprints: Maria Alfonsina Desiderio, Institute of General Pathology, School of Medicine, University of Milan, via Luigi Mangiagalli, 31-20133 Milan, Italy. Phone: 39-0250315334; Fax: 39-0250315338. E-mail: a.desiderio{at}unimi.it


   Abstract
Top
 Notes
 Abstract
Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
Hepatocyte growth factor (HGF), a cytokine of tumor microenvironment, exerts opposite effects on CXCR4 expression in MCF-7 (low invasive) and MDA-MB231 (highly invasive) breast carcinoma cells, and here, we show that completely different molecular mechanisms downstream of c-Src activation were involved. As experimental models, we used cells transfected with two CXCR4 promoter constructs and treated with HGF or cotransfected with c-Src wild-type (Srcwt) expression vector; phospho–c-Src formation was enhanced in both cell lines. In MCF-7 cells, consistent with activations of CXCR4Luc constructs after HGF treatment and Srcwt expression, Ets1 and nuclear factor-{kappa}B (NF-{kappa}B) transcription factors were activated. In contrast, in MDA-MB231 cells, CXCR4Luc construct, Ets1 and NF-{kappa}B activities decreased. The divergence point seemed to be downstream of HGF/c-Src and consisted in the interaction between c-Src and the substrate histone deacetylase 3 (HDAC3). Only in MDA-MB231 cells, HDAC3 level was enhanced in membranes and nuclei 30 min after HGF and colocalized/coimmunoprecipitated with phospho–c-Src and phosphotyrosine. Thus, the CXCR4 induction by HGF in MCF-7 cells required NF-{kappa}B and Ets1 activations, downstream of phosphoinositide-3-kinase/Akt, whereas in HGF-treated MDA-MB231 cells, HDAC3 activation via c-Src probably caused a reduction of transcription factor activities, such as that of NF-{kappa}B. These results indicate possible roles of HGF in invasive growth of breast carcinomas. By enhancing CXCR4 in low invasive tumor cells, HGF probably favors their homing to secondary sites, whereas by suppressing CXCR4 in highly invasive cells, HGF might participate to retain them in the metastatic sites. (Mol Cancer Res 2007;5(8):833–45)


   Introduction
Top
 Notes
 Abstract
 Introduction
Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
Spatial and temporal control of the signals emanating from tyrosine kinase receptor Met depends on the diverse repertoire of recruited proximal signaling proteins (1) and is modulated by the partner proteins integrins and E-cadherins, differently expressed on the surface of normal and neoplastic cells in relation to their aggressiveness (2-4). Our data show that in breast neoplastic cells (MCF7) but not in normal epithelial cells (MCF-10), Met and E-cadherins coimmunoprecipitate and undergo coendocytosis after hepatocyte growth factor (HGF) treatment (3). The multifunctional cytokine HGF is the ligand for Met, and an increased autocrine HGF-Met signaling is a critical downstream function of c-Src–Stat3 pathway in mammalian tumorigenesis (5). The high levels of HGF and Met expression are considered as a possible indicator of earlier recurrence and shortened survival in breast cancer patients (6, 7). Receptor tyrosine kinases Met and ErbB-2 are associated with morphogenic and functional differentiation of normal mammary gland epithelium and play an important role in malignant transformation (8). A strong correlation has been reported between Met overexpression and high risk of disease progression because Met node-positive breast cancer identifies patients with poor clinical outcome independent of Her2/neu (9).

We have shown that HGF induces the plasminogen activation system in liver tumors and enhances the expression of CXCR4 receptor for the CXCL12 chemokine in the low invasive breast carcinoma cells MCF-7 (10, 11). In the highly invasive MDA-MB231 breast carcinoma cells, however, HGF decreases CXCR4 expression, and the underlying mechanism(s) of these opposite effects is still unknown (11). CXCR4 is the principal chemokine receptor expressed on breast carcinoma cells with possible but undefined roles in various steps of metastatic cell diffusion and homing to secondary sites (12, 13).

The aim of the present paper was to investigate how c-Src signaling downstream of HGF/Met might be involved in the decrease of CXCR4 expression, evaluating the possible interaction and activation of histone deacetylase 3 (HDAC3). To this purpose, we compared the molecular patterns in MDA-MB231 (highly invasive) and MCF-7 (low invasive) tumor cells after HGF treatment that possibly activates c-Src phosphorylation or c-Src wild-type (Srcwt) transfection that would permit c-Src expression and phosphorylation in these cells.

c-Src is a well-characterized cytoplasmic and membrane-associated tyrosine kinase that transduces mitogenic signals from a variety of growth factor receptors like Met (14, 15). In addition, c-Src also associates with integrins as well as other membrane proteins possibly including E-cadherins (16-18). These interactions ultimately result in increased transcription and/or activity of proteins involved in cell growth, proliferation, motility, and invasion, and c-Src has been frequently associated with tumor progression (19). HDAC3 is a member of class I HDACs, possibly functioning as a negative regulator of gene expression (20). By associating with c-Src at plasma membrane level, HDAC3 seems to be activated and to affect protein transactivation (21). Moreover, c-Src tyrosine kinase directly or indirectly triggers multiple phosphorylation signaling cascades, such as mitogen-activated protein kinases (MAPK) and phosphoinositide-3-kinase (PI3K)/Akt pathways (22, 23), that may be blocked by the recruitment of Dok-R (24). We focused our studies on these signaling pathways in Srcwt-transfected cells, and on the DNA binding of nuclear factor-{kappa}B (NF-{kappa}B) and Ets1, the transcription factors possibly important for CXCR4 expression in the two studied models (25).

We show that different molecular mechanisms are involved in the opposite regulation of CXCR4 expression by HGF/c-Src in MCF-7 (low invasive) and MDA-MB231 (highly invasive) breast carcinoma cells. In fact, c-Src-HDAC3 interaction and activation occurred only in MDA-MB231 cells and were possibly involved in the decrease of NF-{kappa}B and Ets1 activities responsible for CXCR4 suppression. On the contrary, PI3K/Akt downstream of c-Src might be implicated in transcription factor (NF-{kappa}B and Ets1) activation and CXCR4 expression enhancement in MCF-7 cells.


   Results
Top
 Notes
 Abstract
 Introduction
 Results
Discussion
 Materials and Methods
 Acknowledgements
 References
 
Effect of HGF on CXCR4 Promoter Activity and Expression
In MCF-7 and MDA-MB231 cells, we comparatively studied the activity of the gene reporters driven by the CXCR4 promoter (–2632/+86)Luc and by a short sequence pCXCR4(–600/+19)Luc, in the presence or the absence of HGF. The short promoter sequence spanned upstream the TATA box and contained multiple (six) Ets1 and one NF-{kappa}B consensus sites, but was devoid of the five hypoxia-inducible factor-1 (HIF-1) binding sites (HRE; ref. 25).

The luciferase activity of the two gene reporters was higher in MDA-MB231 cells than in MCF-7 cells both under starvation and 10% fetal bovine serum (FBS) culture conditions, taking into account that the transfection efficiency was the same in the two cell lines (Fig. 1A ). Consistently, CXCR4 mRNA level in control MDA-MB231 cells was 6-fold that of MCF-7 cells, the latter being, however, increased by the treatment with 5-azacytidine, a known DNA methyltransferase inhibitor (ref. 26; Fig. 1B). However, DNA methylation blockade did not influence CXCR4 transcription by HGF (data not shown). MDA-MB231 cells had a higher level of HGF-receptor Met relative to MCF-7 cells, whereas E-cadherins were absent (ref. 4; Fig. 1B). Vinculin was chosen for loading control in this Western blot as well as in the following blots because {alpha}-tubulin and ß-actin are not expressed at the same levels in MCF-7 and MDA-MB231 cells as reported in our previous papers and in the literature (4, 11, 25). Thus, under our experimental conditions, vinculin seems to be the most suitable loading control to be used because it is equally expressed in MCF-7 and MDA-MB231 cells.


Figure 1
View larger version (44K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 1. HGF differently regulates CXCR4 promoter activity and expression in breast cancer cells. A. MCF-7 and MDA-MB231 cells were transiently + transfected with CXCR4 promoters (–2632/+86)Luc and (–600/+19)Luc. The histograms indicate the fold increases of luciferase activity observed for control (C) and starved (St) cells, relative to MCF-7 control value. Columns, mean of three independent experiments done in triplicate; bars, SE. *, P < 0.05; **, P < 0.005 versus MCF-7 control value. B. Top, Northern blot analysis of total RNA from cells treated with 5-azacytidine (Aza) and hybridized with labeled probes for CXCR4 or 18 S (used for normalization). Numbers at the bottom, fold variations relative to MCF-7 control value. Bottom, Western blot analysis of Met and E-cadherins (E-cad). Vinculin was used for normalization. All the experiments have been repeated thrice with similar results. C. Cells transfected as described above were treated with HGF for 24 h. The histograms indicate the fold changes of luciferase activity relative to MCF-7 control value. Columns, mean of three independent experiments done in triplicate; bars, SE. *, P < 0.05; **, P < 0.005 versus respective pCXCR4(–2632/+86)Luc st value; {Delta}, P < 0.05; {Delta}{Delta}, P < 0.005 versus respective pCXCR4(–600/+19)Luc st value. Northern (D) and Western (E) blot analyses of CXCR4 in HGF-treated cells. Numbers at the bottom, fold variations relative to the respective starvation (st) values. 18S and vinculin were used for normalization. The experiments have been repeated thrice with similar results. F. CXCR4Luc-cells, cotransfected with the dominant negative for c-Src ({Delta}Src) and/or treated with LY 294002 (LY), were exposed to HGF. Columns, mean of three independent experiments done in triplicate; bars, SE. *, P < 0.05; **, P < 0.005 versus respective st value; {Delta}{Delta}, P < 0.005 versus HGF-treated cells. G. Northern blot analysis of CXCR4 in the cells treated with LY 294002 (LY) and exposed or not to HGF for 16 h. 18S was used for normalization. The data are representative of experiments repeated thrice.

 
As reported in Fig. 1C, 24-h HGF treatment increased (2–3-fold) the luciferase activity of both CXCR4 constructs in MCF-7 cells, but decreased (35–50%) gene reporter activities in MDA-MB231 cells. These results were consistent with the patterns of CXCR4 mRNA and protein levels after HGF treatment (Fig. 1D and E). In HGF-treated MCF-7 cells, CXCR4 mRNA level increased at 4 h (4.4-fold), maintaining almost at the same values until 16 h. Protein level slightly increased starting from 6 h, doubled between 12 and 18 h, and was 5-fold higher than starved cells at 24 h. The differences in the time courses were probably consistent with the time needed for protein synthesis, even if we cannot exclude that post-transcriptional mechanisms might be involved. In MDA-MB231 cells, CXCR4 mRNA levels decreased (–50% to –65%) between 4 and 16 h, with a concomitant and progressive protein level reduction to 0.2-fold relative to starved cells at 24 h (Fig. 1D and E).

In MCF-7 cells, the stimulatory effect of HGF on pCXCR4(–2632/+86)Luc activity was completely prevented by c-Src dominant negative ({Delta}Src) expression as well as by LY 294002 treatment alone or in combination with {Delta}Src (Fig. 1F). LY 294002 is a well-known inhibitor of PI3K activity (11). In MDA-MB231 cells, basal pCXCR4(–2632/+86)Luc activity was diminished by HGF alone and in the presence of {Delta}Src by 35% and 55%, respectively. Therefore, the two stimuli showed additive effects. In these cells, LY 294002 exposure prevented (of about 60%) the HGF-dependent decrease of pCXCR4(–2632/+86)Luc activity. Thus, the inhibition (–60%) observed after exposure to the combination HGF/LY 294002 with {Delta}Src seemed to depend on the c-Src blockade. These data indicate critical roles for c-Src in the opposite patterns of CXCR4 transactivation after HGF in MCF-7 and MDA-MB231 cells, and these molecular mechanisms were further investigated in the present paper.

The effects of LY 294002 on endogenous CXCR4 gene were consistent with those obtained with the promoter construct. In fact, LY 294002 reduced (–90%) CXCR4 mRNA level in 16 h HGF-treated MCF-7 cells, whereas it largely prevented the CXCR4 mRNA reduction in 16 h HGF-treated MDA-MB231 cells (Fig. 1G). We chose to study the 16-h HGF treatment of MCF-7 cells in agreement with previous findings showing that LY 294002 exerts the maximum inhibitory effect at this time (11). Thus, PI3K seemed to be involved in the opposite patterns of CXCR4 in the two cell lines treated with HGF.

c-Src Involvement in CXCR4 Transactivation in HGF-Treated Cells
To determine the effect of HGF-Met couple on c-Src activation in the two breast carcinoma cell lines, differing for E-cadherin expression, first, we evaluated by confocal analysis the patterns of phospho–c-Src (green), Met (magenta), and E-cadherins (red; Fig. 2A ). Because we wanted to examine concomitantly the three molecules and study their interaction, we needed to use three different fluorescent signals, indicated above, plus the blue fluorescent 4',6-diamidino-2-phenylindole (DAPI) stain for the nuclei. In MCF7 cells, 60 min HGF treatment increased phospho–c-Src. Both Met and E-cadherins seemed to be internalized, partially overlapping with phospho–c-Src signal, as shown in merge I and II, respectively. In MDA-MB231 cells, phospho–c-Src was well represented in starved cells, and its level increased 60 min after HGF principally assuming a cell membrane localization in a punctuate pattern. Concomitantly, Met assumed a perinuclear localization. In both the cell lines, HGF showed a scatter effect, dissociating the MCF-7 colonies and enhancing MDA-MB231 cell movement. The phospho–c-Src signal increased also 30 min after HGF treatment in MCF-7 and MDA-MB231 cells, but to simplify the figure, the data were omitted (data not shown). After Srcwt expression-vector transfection, we obtained results similar but more pronounced with respect to those observed with HGF (Fig. 2B). In fact, the phospho–c-Src signal increased throughout the MCF-7 cells and at membrane level in MDA-MB231 cells.


Figure 2
View larger version (41K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 2. c-Src phosphorylation in HGF-treated (A) or Srcwt expression vector-transfected (B) cells. Before HGF treatment, the cells were starved (0.1% FBS) for 18 h. Srcwt transfection was done in 10% FBS. Treated cells on coverslips were incubated with anti-Met (magenta) and anti–E-cadherin (red) antibodies, followed by the secondary antibodies, and then with anti-phospho–c-Src–488–conjugated antibody (green). The nuclei were stained with DAPI (blue). Merge I, merged image of phospho–c-Src/Met/DAPI; merge II, merge image of phospho–c-Src/E-cadherin. Confocal laser scanner analysis was done (magnification, 63x). The images shown are representative of experiments done in triplicate.

 
In further experiments, to evaluate the possible involvement of HGF/c-Src in CXCR4 transactivation, pCXCR4(–2632/+86)Luc– and pCXCR4(–600/+19)Luc–transfected cells were exposed to HGF and/or cotransfected with Srcwt expression vector (Fig. 3A ). In MCF-7 cells, Srcwt expression had a stronger stimulatory effect than HGF treatment on luciferase activity. This was particularly evident with pCXCR4(–600/+19)Luc, which was activated 20-fold by Srcwt transfection showing an additive effect with HGF. Exposure to HGF and Srcwt increased pCXCR4(–2632/+86)Luc by 3- and 8-fold, respectively, and the effects did not seem to be additive. In MDA-MB231 cells, however, HGF and Srcwt decreased luciferase activity of the two gene reporters (50–60% diminutions), and the two stimuli given together seemed to be slightly less effective than separately applied, suggesting a possible interference on CXCR4 transactivation. In MCF-7 cells cotransfected with Srcwt, LY 294002 completely reduced pCXCR4(–2632/+86)Luc activation, whereas in MDA-MB231 cells, the inhibitor prevented almost completely the pCXCR4(–2632/+86)Luc activity diminution due to Srcwt (data not shown).


Figure 3
View larger version (32K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 3. Effects of Srcwt expression on cell signaling. Cells transfected with Srcwt were used as follows. A. CXCR4Luc-cells, cotransfected with Srcwt, were exposed or not to HGF. Columns, mean of three independent experiments done in triplicate; bars, SE. *, P < 0.05; **, P < 0.005; and ***, P < 0.001 versus respective st value. B. Western blot analysis of total proteins and immunoblot with the antibodies for c-Src (Src), phospho–c-Src (pSrc), Akt, phospho-Akt (pAkt), ERK1/2, phosphoERK1/2 (pERK1/2), or vinculin (used for normalization). Numbers at the bottom, fold variations relative to MCF-7 control value. The experiments have been repeated thrice with similar results. The histograms show pSrc/Src, pAkt/Akt, and pERK1/2/Erk1/2 ratios. *, P < 0.05 versus respective control values. C. Confocal laser scanner analysis of cells transfected with Srcwt on coverslips and incubated with anti-phospho–c-Src–488–conjugated (green) and anti-phosphotyrosine-647–Cys–conjugated (magenta) antibodies. The images shown are representative of experiments repeated thrice.

 
Based on the similarity of the results obtained with HGF and Srcwt, both in confocal analyses and in CXCR4-Luc transfection experiments, we thought to use the Srcwt transfection model to further investigate the molecular mechanism(s) responsible for the variations of CXCR4 transactivation. To validate this model, first, we did Western blots in Srcwt-transfected MCF-7 and MDA-MB231 cells (Fig. 3B). c-Src protein and the phosphorylated form, revealed with anti-phospho–c-Src(Tyr416) antibody, strongly increased in Srcwt-transfected cells. The phospho–c-Src (pSrc)/Src ratio almost doubled in Srcwt-transfected MDA-MB231 cells. The endogenous c-Src protein was phosphorylated according to the literature (27, 28). By confocal analysis, we observed that in Srcwt-transfected cells, tyrosine phosphorylation (magenta) increased, and this fluorescent signal largely overlapped with that of phospho–c-Src (green), as shown by the merge, indicating a possible activity of phospho–c-Src at cell membrane level (Fig. 3C).

To clarify the signaling pathways downstream of c-Src, we evaluated also Akt and extracellular signal-regulated kinase 1/2 (ERK1/2) phosphorylation in Srcwt-transfected cells (Fig. 3B). In MCF-7 cells, Srcwt expression slightly increased Akt protein and enhanced its phosphorylation, triplicating the ratio phosphoAkt (pAkt)/Akt. In MDA-MB231 cells, Srcwt transfection remarkably reduced Akt protein (–86%) that was phosphorylated, and the pAkt/Akt ratio was 4-fold higher in Srcwt-transfected cells with respect to controls (C). It is worth noting that in these two carcinoma breast cell lines, Akt migrates at ~60 kDa as a tightly spaced triplet (29). ERK1/2 and their phosphorylation were unaffected by the same experimental conditions. These data suggested a possible role of pAkt in the opposite effects of Srcwt on CXCR4 transactivation.

HDAC3 Associated with Phospho–c-Src in HGF-Treated MDA-MB231 Cells
To investigate the divergence point of the molecular mechanisms responsible for the opposite CXCR4 expression in MCF-7 and MDA-MB231 cell lines, we evaluated the possible relationship of HDAC3 with c-Src (Figs. 4 and 5 ). The experiments were done in HGF-treated cells to deepen the knowledge about the influence of this microenvironmental cytokine on tumor-invasive phenotype (30, 31). Phospho–c-Src (green), HDAC3 (red), and phosphotyrosine (magenta) fluorescent signals were evaluated in the two HGF-treated carcinoma cells by confocal analysis, and we show the most significant results obtained in MDA-MB231 cells (Fig. 4). The images were taken at the middle level (A) and at the top (B) of the preparations, the top images showing principally cell nuclei. Interestingly, 30 min after HGF treatment, HDAC3 signal increased at cell membrane and nuclear level. HDAC3 largely colocalized with phospho–c-Src and phosphotyrosine at cell membrane, as shown in merge I and II, respectively. HDAC3 signal seemed to persist in nuclei between 30 and 60 min after HGF. In MCF-7 cells, a substantial activation of HDAC3 could not be observed, and therefore, the confocal analysis images were not relevant (data not shown).


Figure 4
View larger version (58K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 4. HGF-induced phospho–c-Src/HDAC3 association and tyrosine phosphorylation. Confocal laser scanner analysis of MDA-MB231 cells: images taken at the middle (A) and at the top (B) levels of the preparations. The top level shows principally cell nuclei. HGF-treated cells on coverslips were incubated with anti-HADAC3 (red), followed by the secondary antibody, and then with anti-phospho–c-Src–488–conjugated (green) and anti–phosphotyrosine-647–Cys–conjugated (magenta) antibodies. The nuclei were stained with DAPI (blue). Merge I, (A) merged image of phospho–c-Src/HDAC3/DAPI; and (B) merged image of phospho–c-Src/HDAC3; merge II, (A) and (B) merged image of phosphotyrosine/HDAC3. The images were taken as reported above (magnification, 63x) and are representative of experiments repeated thrice.

 

Figure 5
View larger version (51K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 5. Phospho–c-Src immnunoprecipitates active HDAC3. Total protein extracts were (A) immunoprecipitated with anti-phospho–c-Src ({alpha}p-Src) antibody, and immunoblots were done with anti-HDAC3, anti-phospho–c-Src (pSrc), and anti–c-Src (Src) antibodies. Numbers at the bottom, fold variations relative to the value set as 1 for each immunoblot, or (B) immunoprecipitated with anti-HDAC3 ({alpha}HDAC3) antibody, and immunoblots were done with anti-phosphotyrosine (pTyr) and anti-HDAC3. The experiments have been repeated thrice with similar results.

 
To further confirm the association between HDAC3 and phospho–c-Src in HGF-treated MDA-MB231 cells, but not in MCF-7 cells, coimmunoprecipitation experiments were done with total cell extracts (Fig. 5A). Immunoprecipitation was done with anti-phospho–c-Src antibody ({alpha}p-Src), and immunoblots with anti-HDAC3, anti-phospho–c-Src (pSrc) and anti–c-Src (Src) antibodies. In MDA-MB231 cells, HDAC3 became associated with phospho–c-Src 30 to 60 min after HGF treatment (2.2- to 2.7-fold increase versus starved cells). The phospho–c-Src/HDAC3 coimmunoprecipitation seemed related to the doubling of c-Src phosphorylation and was in agreement with the confocal analysis data shown in Fig. 4. c-Src level was substantially unchanged. In 60-min HGF-treated MCF-7 cells, HDAC3 association with phospho–c-Src slightly increased (from 0.6- to 1.0-fold) without relevant changes in c-Src phosphorylation (Fig. 5A). HDAC3, immunoprecipitated by anti-HDAC3 antibody ({alpha}HDAC3), was phosphorylated at tyrosine only in MDA-MB231 cells between 30 and 60 min after HGF treatment (Fig. 5B). Based on these data, HDAC3 activated by c-Src in HGF-treated MDA-MB231 cells might have a role in gene expression regulation. The soundness of our results, excluding the interference of immunoglobulin G (IgG) heavy chains in the detection of the specific proteins in the coimmunoprecipitates, was based on the wide separation of the bands: in fact, we did a big gel and a long-lasting run. Moreover, the variation of phosphotyrosine levels occurred only in HGF-treated MDA-MB231 cells, and HDAC3 levels were higher in MDA-MB231 than in MCF-7 cells.

Ets1 and NF-{kappa}B Activity Involvement in CXCR4 Transactivation in Response to HGF Treatment
Finally, we studied the activity patterns of Ets1 and NF-{kappa}B, two transcription factors important for the HGF-dependent enhancement of CXCR4 expression in MFC-7 cells (25). Here, the study was extended to MDA-MB231 cells to evaluate the role of Ets1 and NF-{kappa}B in the opposite responses of CXCR4 transactivation to HGF. To this purpose, breast carcinoma cells were cotransfected with CXCR4 gene reporters, and the NF-{kappa}B super-repressor (ssNF{kappa}B) or the dominant negatives of ARNT ({Delta}ARNT) and of Ets1 ({Delta}Ets1; Fig. 6A ). ssNF{kappa}B is a highly specific and effective NF-{kappa}B inhibitor with mutations in the I{kappa}B{alpha}-conserved serine residues typically targeted for phosphorylation after cellular stimulation (32). {Delta}ARNT codes for the mutant form of the HIF-1ß subunit that lacks the basic domain and therefore is still capable of heterodimerizing with the {alpha} subunit, but cannot bind the DNA, and {Delta}Ets1 is a dominant negative for Ets1, bearing only the DNA-binding site (25). In MCF-7 cells, the activation of both CXCR4 reporter constructs in response to HGF was completely prevented by ssNF{kappa}B alone or in combination with {Delta}Ets1. In MDA-MB231 cells exposed to HGF, the activities of the two CXCR4 constructs were reduced by 40% to 50%. {Delta}Ets1 alone or in combination with ssNF{kappa}B further decreased pCXCR4(–2632/+86)Luc activity under the value due to HGF. However, ssNF{kappa}B prevented the pCXCR4(–600/+19)Luc activity diminution after HGF and surprisingly doubled luciferase activity above the starved cell value. {Delta}Ets1 alone was as effective as HGF in diminishing pCXCR4(–600/+19)Luc activity and prevented the strong activation due to ssNF{kappa}B. {Delta}ARNT was ineffective in both the cell lines.


Figure 6
View larger version (35K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 6. Involvement of Ets1 and NF-{kappa}B in CXCR4 expression in HGF-treated breast cancer cells. Cells transfected with CXCR4 construct were cotransfected with the dominant negatives for ARNT ({Delta}ARNT) and ETS1 ({Delta}Ets1) or the ssNF{kappa}B, and exposed to HGF. The histograms indicate the fold changes of luciferase activity relative to the respective st value. Columns, mean of three independent experiments done in triplicate; bars, SE. *, P < 0.05; **, P < 0.005 versus respective st value; {Delta}{Delta}, P < 0.005 versus HGF-treated cells. B and C. EMSA of nuclear extract from starved (st) and HGF-treated cells. We evaluated the bindings of Ets1 and NF-{kappa}B (p50/p65 complex) to the labeled oligonucleotides, corresponding to specific consensus sequences present in CXCR4 promoter. ETSA and ETSB oligonucleotides contain two and one consensus sites for Ets1, respectively. Supergelshifts were done with specific antibodies (Ab) for Ets1, p50, or p65, using nuclear extracts as indicated in the figure. ss, supershift; comp, specific competition with 50-fold excess unlabeled oligonucleotide. Numbers at the bottom, fold variations relative to the respective st values. The experiments have been repeated thrice with similar results.

 
We further investigated how Ets1 and NF-{kappa}B transcription factors influenced CXCR4 expression after HGF in the two breast carcinoma cells. The bindings to consensus sequences present in CXCR4 promoter were evaluated using nuclear extracts from HGF-treated cells. ETSA and ETSB oligonucleotides contain a couple and a single consensus sites for Ets1 (Fig. 6B). In MCF-7 cells, the Ets1 binding to ETSA and ETSB, similar under starvation, increased (3–4-fold) after HGF treatment, peaking at 2 h with ETSA and at 8 h with ETSB. In MDA-MB231 cells, the Ets1 binding to ETSA and ETSB promoter sequences was elevated under starvation and decreased progressively after HGF treatment, reaching 80% diminutions at 8 h. The Ets1 protein bindings to ETSA and ETSB in MDA-MB231 starved cells were 2.5- and 5-fold higher than in MCF-7 cells. The specificity of the Ets1/DNA complex was examined by competition experiments using 50-fold molar excess of unlabeled oligonucleotides (comp), which almost completely suppressed the complex formation. Moreover, supergelshift analyses were done with nuclear extracts from MCF-7 and MDA-MB231 cells treated with HGF. The preincubation with the anti-Ets1 antibody disrupted the labeled DNA probe-Ets1 protein complex, causing a shift (ss). We show the results of the supergelshifts obtained in HGF-treated MCF-7 cells at 2 h (for ETSA) and at 8 h (for ETSB) and in MDA-MB231 starved cells, i.e., the samples showing the highest DNA bindings.

Figure 6C shows the electrophoretic mobility shift assay (EMSA) done with nuclear extracts from HGF-treated cells and the response element "p65" in the CXCR4 promoter (about 70 bp upstream the TATA box), which is the NF-{kappa}B consensus sequence most probably important for CXCR4 transcription regulation in response to cytokines (33). In MCF-7 cells, HGF enhanced the NF-{kappa}B binding activity of 3-fold at 2 h, reaching 6-fold at 4 h and decreasing thereafter. In MDA-MB231 cells, HGF remarkably reduced (–60%) NF-{kappa}B binding at 2 h, remaining at low levels (–50% to –40%) for all the observation period. It is worth noting that under starvation, NF-{kappa}B DNA binding was about 3-fold higher in MDA-MB231 than in MCF-7 cells. The DNA probe-protein complex disruption by the antibody against p50 causing a shift (ss) and a blocked shift with anti-p65 antibody showed that p50 and p65 were the NF-{kappa}B proteins in the complex. The use of 50-fold molar excess of the unlabeled oligonucleotide (comp), which almost completely suppressed the p50/p65 complex binding to DNA, showed the specificity of the complex. DNA-binding activity of the constitutively expressed ubiquitous Octamer-1 transcription factor was unaffected by the treatments (data not shown). Thus, Ets1 and NF-{kappa}B were involved in CXCR4 transactivation, and the activities were oppositely regulated by HGF.

Effect of Srcwt Transfection on Ets1 and NF-{kappa}B DNA Binding Activities
Then, we studied the role played by c-Src in the regulation of Ets1 and NF-{kappa}B bindings to CXCR4 promoter. Figure 7A shows that Srcwt transfection in MCF-7 cells enhanced Ets1 DNA binding to ETSA (18-fold) and ETSB (about 6-fold) and was not affected by LY 294002 treatment. In MDA-MB231 cells, Srcwt expression vector transfection prevented Ets1 binding to the two CXCR4 promoter consensus sequences (–50% to –70%), and also in these cells, LY 294002 was almost ineffective. As regards NF-{kappa}B transactivating activity (Fig. 7B), NF-{kappa}B multimer-driven construct was activated (2-fold) by Srcwt expression in MCF-7 cells, whereas it was reduced (by about 80%) in MDA-MB231 cells. Basal activity of the NF-{kappa}B multimer gene reporter was 24-fold higher in MDA-MB231 cells than in MCF-7 cells. In both the cell lines transfected with Srcwt expression vector, the exposure to LY 294002 did not modify luciferase activity of NF-{kappa}B multimer gene reporter. Consistently, in MCF-7 cells, NF-{kappa}B (p50/p65 complex) DNA binding increased (16-fold) after Srcwt transfection and was reduced (–80%) in MDA-MB231 cells. LY 294002 affected (–35%) NF-{kappa}B DNA binding activity in MCF-7 cells expressing Srcwt (Fig. 7C). Also, under control conditions (10% FBS), Ets1 and NF-{kappa}B DNA bindings were remarkably higher in MDA-MB231 than in MCF-7 cells. DNA-binding activity of the constitutively expressed ubiquitous Octamer-1 transcription factor was unaffected by the treatments (data not shown).


Figure 7
View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 7. Effect of Srcwt expression vector on Ets1 and NF-{kappa}B activities. A and C. Nuclear extracts from Srcwt-transfected cells were used for EMSA analyses of Ets1 and NF-{kappa}B as reported above. Numbers at the bottom, fold variations relative to the respective control (C) value. The EMSA shown are representative of experiments repeated thrice. B. Srcwt-transfected cells were cotransfected with the gene reporter driven by NF-{kappa}B multimer in the presence or the absence of LY 294002 (LY). The histograms indicate the fold changes of luciferase activity relative to MCF-7 control value. Columns, mean of three independent experiments done in triplicate; bars, SE. *, P < 0.05; ***, P < 0.001 versus MCF-7 control value; {Delta}{Delta}{Delta}, P < 0.001 versus MDA-MB231 control value.

 

   Discussion
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
Materials and Methods
 Acknowledgements
 References
 
HGF, a cytokine produced in a tumor microenvironment, may affect various aspects of the invasive phenotype of carcinoma cells (31, 34, 35). We observed that HGF increased the CXCR4 promoter activity as well as the mRNA and protein levels in estrogen-responsive MCF-7 breast carcinoma cells. On the contrary, in MDA-MB231 highly invasive, estrogen-resistant breast carcinoma cells, HGF reduced CXCR4 transactivation, decreasing mRNA and protein levels.

The progression of mammary adenocarcinomas has been associated with deregulation of several signaling molecules and epigenetic mechanisms (36, 37). As shown for urokinase-type plasminogen activator, a member of plasminogen activation system important for tumor invasiveness (26), DNA methylation blockade seemed to induce CXCR4 in MCF-7 cells, increasing the mRNA level to reach the elevated value of MDA-MB231 cells. Because methylation did not seem to influence CXCR4 transcription by HGF, the molecular basis of the opposite CXCR4 expression responses to HGF in MCF-7 and MDA-MB231 cells has been addressed in the present research.

In particular, the aim of this paper was to verify the hypothesis that c-Src tyrosine kinase might interact at the membrane level with the substrate HDAC3 (21), thus regulating CXCR4 gene transcription specifically in MDA-MB231 cells. c-Src is a multifunctional protein, and the elevated expression or activity of c-Src has been linked to the cancer metastatic process (38, 39). The direct interaction with Met after HGF binding confers to c-Src a functional role in cell proliferation, transformation, and motility (15).

At the first part of the studies, we examined the activation of c-Src kinase and its phosphorylation state (Tyr416) in MCF-7 and MDA-MB231 breast carcinoma cells treated with HGF or transfected with Srcwt expression vector. The two models might be assimilated because both were characterized by an enhancement of c-Src phosphorylation.

After HGF exposure, phospho–c-Src was partially associated with E-cadherins inside the MCF-7 cells, a finding similar to that observed after Srcwt transfection and consistent with the data reported for active c-Src(Src 531) in colon cancer (40). The significance and the mechanisms of E-cadherin phosphorylation are still partially known and might be related to their ubiquitination/degradation and recycling responsible for the remodeling of adhesive contacts as well as cell signaling (18). After HGF treatment or Srcwt transfection of MDA-MB231 cells lacking E-cadherins, phospho–c-Src localized at the cell membrane, and Met was internalized, assuming a perinuclear disposition. The perinuclear compartment includes the late endosomes/lysosomes, i.e., the major route for Met degradation after proteasome delivery, as well as the Golgi, corresponding to a recycling compartment or containing newly synthesized Met (41). A functional significance for internalized Met might be suggested in the regulation of signal transduction pathways and nuclear gene expression, as reported in MCF-7 cells by us and other authors under different experimental conditions (3, 42).

In the second part, we focused our studies to understand how the regulatory signaling pathways diverge downstream of HGF-Met/c-Src in MCF-7 and MDA-MB231 cells to explain the opposite CXCR4 expression after HGF exposure as well as after Srcwt transfection. The observed wide overlapping of phospho–c-Src with phosphotyrosine fluorescent signals in Srcwt-transfected cells, analyzed by confocal laser scanning, might also indicate the active role played by exogenously expressed c-Src.

Because c-Src regulates members of the MAPK family and PI3K activity, we studied the phosphorylation of Akt (a PI3K substrate) and of ERK1/2 in Srcwt-transfected cells. Akt protein and phosphorylation augmented (3-fold increase of the pAkt/Akt ratio) in MCF-7 cells, whereas in MDA-MB231 cells, Akt protein decreased but was phosphorylated. A ubiquitin-proteasome pathway might be involved in Akt degradation (43), and the remaining Akt protein was probably phosphorylated via PI3K involved in Srcwt inhibitory effect on CXCR4 gene reporter activity. Due to the fact that ERK1/2 protein and phosphorylation were unchanged after Srcwt transfection, we suppose that pAkt might have a role in the downstream steps of CXCR4 expression regulation in the two cell lines.

The most surprising data, explaining the divergence in the regulatory mechanisms, were observed by confocal analysis of HGF-treated MDA-MB231 cells. At 30 min, HDAC3 was enhanced at the cell membrane level and seemed to colocalize with phospho–c-Src. Thus, HDAC3 was probably phosphorylated at tyrosine, as indicated by the noticeable overlapping of the two fluorescent signals. The tyrosine-phosphorylated HDAC3 was concomitantly present in the nuclei of MDA-MB231 cells, where HDAC3 might exert its regulatory function on chromatin structure and gene expression (20, 44). Further coimmunoprecipitation studies deepened and confirmed these results. In fact, phospho–c-Src coimmunoprecipitated HDAC3 in HGF-treated MDA-MB231 cells, and the HDAC3 immunoprecipitated was phosphorylated at tyrosine probably becoming activated.

Our results lead to hypothesize a principal role of activated HDAC3 in the reduction of CXCR4 expression in HGF-treated MDA-MB231 cells. HDACs negatively regulate gene expression, and pAkt might influence HDAC3 stabilization (44). HDACs as corepressors affect transcription factor activities, including NF-{kappa}B, with mechanisms only partially known (20, 45, 46). NF-{kappa}B DNA binding, very elevated in MDA-MB231 cells, was diminished by HGF exposure as well as by Srcwt transfection. Unexpectedly, ssNF{kappa}B caused a superactivation of pCXCR4(–600/+19)Luc but not of pCXCR4(–2632/+86)Luc in MDA-MB231 cells. We cannot exclude an activation of IKK{alpha}-dependent phosphorylation-induced processing of p100 to p52 that may regulate p65 DNA binding, possibly becoming insensitive to ssNF{kappa}B under particular experimental conditions (45, 47). Also, Ets1 DNA binding was reduced in MDA-MB231 cells treated with HGF or transfected with Srcwt, but we do not know whether the underlying mechanisms were similar to those involved in the regulation of NF-{kappa}B activity. In MDA-MB231 cells, a reduced NF-{kappa}B/Ets1 interaction on CXCR4 promoter might occur in contrast to MCF-7 cells (25). The differences in the signaling pathways involved in CXCR4 expression after HGF treatment are reported in Fig. 8 .


Figure 8
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		FIGURE 8. Different signaling pathways involved in HGF-dependent opposite regulation of CXCR4 expression in breast cancer cells.

 
In MCF-7 cells, Ets1 and NF-{kappa}B cooperate to induce CXCR4 after HGF treatment through the ERK1/2 and PI3K signaling pathways. The blockade of ERK1/2 and PI3K transduction pathways prevents CXCR4 induction due to HGF (11, 25). Consistently, we report here that LY294002 suppressed HGF-dependent CXCR4 transactivation, causing a reduction also of the endogenous CXCR4 transcription.

The possible discrepancies observed for Akt might depend on culture media conditions because Akt and its phosphorylation are affected by serum (48). In fact, Akt activity is low under starvation also in the presence of HGF that triggers Met/PI3K/c-Jun-NH2-kinase 1 (JNK1) and NF-{kappa}B downstream rather than PI3K/Akt (25, 49, 50). However, c-Src is known to inhibit PTEN and then to activate Akt (51), suggesting the involvement of this mechanism in CXCR4 induction after Srcwt expression.

Finally, HIF-1 was not involved in CXCR4 transactivation after HGF treatment based on the data that the activity of CXCR4 gene reporter (–600/+19)Luc, lacking HRE sites, was higher than that of pCXCR4(–2632/+86)Luc, which was unaffected by ARNT dominant negative. It is worth noting that in MDA-MB231 cells, HIF-1 is not functional because HIF-1ß/ARNT subunit is mutated (52).

In summary, HGF may be considered a cytokine that, when overproduced by supporting elements of tumor stroma (31, 34, 35), influences the invasive phenotype of breast carcinoma cells. The homing of the less invasive cells to secondary growth sites might be favored by CXCR4 enhancement, whereas the highly invasive MDA-MB231 cells are probably retained in the metastatic sites because of the reduction of CXCR4 receptor level. HGF/Met couple is considered to be an ideal target for cancer therapy even in the absence of genetic alterations (30). Thus, HGF treatment might be used in MDA-MB231 cells to reduce NF-{kappa}B activity as an effective therapy for relatively resistant cells.


   Materials and Methods
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
Acknowledgements
 References
 
Materials
RPMI 1640, FBS, and 5-azacytidine were from Sigma Chemical Co. Recombinant human HGF and anti-CXCR4 monoclonal antibody (MAB-172) were from R&D Systems. [{alpha}-32P]dCTP (3,000 Ci/mmol), [{gamma}-32P]ATP (3,000 Ci/mmol), Hybond C-extra nylon filters, Hybond ECL nitrocellulose membranes, and nick translation kit were from Amersham Biosciences Europe GmbH. Anti–phospho-Akt(Ser473), anti-HDAC3, and anti-phosphotyrosine(pTyr100)–Alexa Fluor647 conjugated were from Cell Signaling. Anti-phosphotyrosine (AG10), anti–c-Src, and anti-phospho–c-Src(Tyr416), conjugated or not with Alexa Fluor488, were from Upstate Biotechnology. Anti-Met (C12), anti-Ets1(C20), anti-Akt1/2 (H136), anti-Erk1/2 (K-23), anti-phospho-Erk1/2 (E-4), and anti-vinculin antibodies were purchased from Santa Cruz Biotechnology. The monoclonal anti–E-cadherin (C20820) antibody was from Transduction Laboratories. LY 294002 was from Calbiochem. AlexaFluor568 and 647 secondary antibodies were from Molecular Probes. pGL2 enhancer, pGEM-T Easy, pRL-TK, T4 ligase, calf intestinal alkaline phosphatase, Pfu DNA polymerase, Klenow polymerase, and dual luciferase reporter assay system were from Promega. ß-Galactosidase staining kit was from Mirus.

Cell Cultures
Human breast carcinoma cells MCF-7 and MDA-MB231 were from the European Cell Cultures Collection. All cells, routinely maintained in RPMI 1640 containing 10% FBS, were starved (0.1% FBS) for 18 and 24 h before HGF treatment (200 ng/mL) to prevent the interference of growth factors present in the serum (11).

Promoter Constructs and DNA Expression Vectors
We used KpnI/XhoI to subclone the original gene reporter pCXCR4(–2632/+86)Luc (kindly provided as pGL2 basic construct by Dr. A.J. Caruz, Universidad de Jaen, Madrid, Spain) in the pGL2-enhancer vector, previously cut with SmaI/KpnI. The pCXCR4(–600/+19)Luc fragment was amplified by PCR with Pfu DNA polymerase using pCXCR4(–2632/+86)Luc as template, and the following sense and antisense primers: 5'-TCCCGGGCTTCTGAAAGTATCTCCTAATTATCTG-3' and 3'-CAAACAACCGACGCCGTCGTCCATGGT-5' (Roche). PCR was done in 50 µL of 1x PCR buffer containing 2 mmol/L MgCl2, 1 mmol/L deoxynucleotide triphosphates, 2 units of Pfu DNA polymerase, 10 µmol/L DNA primers, and 200 ng pCXCR4(–2632/+86)Luc. The samples were heated at 95°C for 120 s, then at 95°C for 45 s, 68°C for 45 s, 72°C for 120 s for 35 cycles, and then at 74°C for 5 min. The PCR fragment was cloned in pGEM-T Easy and cut with SmaI and NotI, blunted with Klenow polymerase, and cloned in pGL2-enhancer, previously cut with SmaI (24).

The reporter plasmid NF-{kappa}BLuc containing three NF-{kappa}B consensus sequences was kindly provided by Dr. M. Hung (Anderson Cancer Center). The c-Src expression vector for the wild type (Srcwt) was from Dr. S. Parson (University of Virginia, Charlottesville, VA). The expression vector RSVI{kappa}B{alpha}MSS, coding for ssNF{kappa}B, was kindly provided by Dr. N.D. Perkins (University of Dundee, Dundee, United Kingdom). The dominant negative forms of ARNT (pcDNA3ARNTdelta_b, {Delta}ARNT), Ets1 ({Delta}Ets1), and c-Src (SrcK295M, {Delta}Src) were, respectively, from Dr. M. Schwarz (University of Tübingen, Germany), Dr. J. Ghysdael (Centre National de la Recherche Scientifique UMR, Orsay, France), and Dr. W.C. Horne (Yale University, New Haven, CT).

Cell Transfections and Reporter Assay
For gene reporter assays, the cells were seeded in 24-multiwell plates and were transfected with Fugene 6 (Roche Applied Science) according to the recommendations of the manufacturer. At 70% to 80% of confluence, the cells were incubated with a DNA/Fugene mixture containing 200 ng pCXCR4Luc or NF-{kappa}BLuc gene reporter per well and 40 ng of pRL-TK (Renilla luciferase) for normalization in the presence or absence of 1 µg (Srcwt, {Delta}ARNT, or {Delta}Ets1) or 500 ng ssNF{kappa}B expression vector DNA per well (24). After refreshment of culture medium with RPMI containing 10% FBS, the cells were collected 24 h later. When 24 h of HGF treatment was done, the cells underwent previously overnight starvation. Firefly and Renilla luciferase activities were measured with the dual luciferase assay system (Promega). The firefly/Renilla luciferase activity ratios were calculated by the software and used to evaluate the changes in luciferase activity of CXCR4 constructs after Srcwt expression vector transfection or HGF treatment compared with respective controls, i.e., 10% FBS-cultured or -starved cells. Transfection efficiency was 20% to 25% for MCF-7 and MDA-MB231 cells, evaluated in ß-galactosidase–stained cells.

For nuclear protein extraction, the cells seeded in T75 flasks and transfected with Srcwt (2.5 µg) at subconfluence were allowed 24 h before collection.

The treatment with 20 µmol/L of LY 294002, dissolved in DMSO at the nontoxic final concentration of 0.02%, was done 30 min before HGF treatment and in the last 8 h of the Srcwt transfection period (11).

Northern Blot Assay
Cells (4 x 106) were seeded in T25 flasks, were allowed to attach, and were treated at subconfluence with HGF. When present, 20 µmol/L LY 294002 was added to the cultures 30 min before HGF. Northern blots of total RNA (30 µg) were done. The filters were sequentially hybridized with the labeled 1.2-kb fragment (XbaI-EcoRI) from CXCR4 cDNA (kindly provided by TNC Wells, Serono Pharmaceutical Research Institute, Geneva, Switzerland) and with the labeled probe for 18S rRNA as a loading standard. The relative amounts of CXCR4 mRNAs were calculated after densitometric evaluation of the Northern blot, and controls were assigned an arbitrary value of 1 (11).

Coimmunoprecipitation and Western Blot Assays
For coimmunoprecipitation experiments, two T75 flasks containing 30 x 106 cells treated or untreated with HGF were pooled. The cells were lysed by homogenization to prepare total extracts containing proteins from nuclei, membranes, and cytosol, using lysis buffer [50 mmol/L Tris-HCl (pH 7.4), 1% Nonidet-P40, 1 mmol/L EDTA, 150 mmol/L NaCl, 0.25% deoxycolate, 1 mmol/L Na3VO4] supplemented with 1 µg/mL proteinase inhibitors aprotinin, leupeptin, and pepstatin and 1 mmol/L NaF. Samples corresponding to 1 mg of protein were immnunoprecipitated overnight at 4°C with 6.5 µg of anti-phospho–c-Src or of anti-HDAC3 antibody. After incubation for 2 h at 4°C with 60 µL (50% v/v) of protein G-Sepharose bead slurry, the immunoprecipitates were managed and resolved by 10% SDS-PAGE as reported before (3). We did a big gel (16 x 18 cm), and a long-lasting run to separate the bands corresponding to the specific proteins from IgG heavy chains. The blot, done with anti-phospho–c-Src antibody immunoprecipitate, was probed with anti-HDAC3 (1:1,000), anti-phospho–c-Src (2 µg/mL), and anti–c-Src (2 µg/mL) antibodies. The blot, done with anti-HDAC3 antibody immunoprecipitate, was probed with anti-phosphotyrosine (3 µg/mL) and anti-HDAC3 (1:1,000) antibodies.

For Western blot analysis, the cells (10 x 106) in T25 flasks were transfected with Srcwt expression vector (2.5 µg). Total proteins were extracted with lysis buffer without homogenization, and then samples (corresponding to 100 µg of protein) were used. Immunoblots were done with anti-phospho–c-Src (2 µg/mL), anti–c-Src (2 µg/mL), anti-phospho-Erk1/2 (1:200), anti-Erk1/2 (1:200), anti-phospho-Akt (1:1,000), and anti-Akt1/2 (1 µg/mL) antibodies. To confirm equal loading, the membranes were immunoblotted with anti-vinculin antibody (1:5,000).

After incubation with the appropriate secondary antibodies, the signals were detected using enhanced chemiluminescence kit (ECL or ECL-plus; Amersham Biosciences). The relative amounts of proteins were calculated after densitometric evaluation of the Western blots, and controls were assigned an arbitrary value of 1 (11).

Electrophoretic Mobility Shift Assay
The EMSA method was used to analyze the DNA-binding activities of Ets1, NF-{kappa}B, and Octamer-1 transcription factors in nuclear extracts, prepared from 30 x 106 cells treated with HGF or transfected with Srcwt expression vector (24). Single-strand oligonucleotides with consensus binding sites for the transcription factors were labeled with T4 polynucleotide kinase (Amersham) using [{gamma}-32P]ATP, annealed to the complementary strand, and purified by polyacrylamide gel electrophoresis. After spectrophotometric measurement, nuclear proteins (10 µg) were incubated for 20 min at 25°C in the binding reaction mixtures, containing 0.5 ng 32P-labeled double-stranded sequences. For competition experiments, 50-fold molar excess of each specific unlabeled double-stranded sequence was added to the binding mixtures. DNA-protein complexes were resolved by electrophoresis on 5% native polyacrylamide gel at 4°C in 0.5 and 1x Tris-borate EDTA [1x TBE: 90 mmol/L Tris borate containing 2 mmol/L EDTA (pH 8)] for the study of Ets1 and NF-{kappa}B DNA binding, respectively. For supergelshift assay, nuclear extracts were preincubated with 1 µg anti-p50 or anti-p65 antibody for 60 min on ice or with 1 µg anti-Ets1 antibody for 15 min at room temperature, followed by incubation with the labeled oligonucleotide and electrophoresis as described above (24). The oligonucleotides containing the Ets1 consensus sequence(s) were 5'-CGCGGGGGAATGGGCGGTTGGAAGCCTGGC-3' (ETSA, containing two sites) and 5'-CCTCCGAAGGAAAGGATCTT-3' (ETSB, containing one site), present in the CXCR4 promoter and spanning at –397 to –412 and at –478 to –481 bp from the transcription start site. The oligonucleotide containing the {kappa}B consensus sequence for NF-{kappa}B was 5'-TCCCCTGGGCTTCCCAAGCC-3' (p65) present in the promoter of CXCR4 at –96 to –111 bp. The Octamer-1 binding site sequence was 5'-TGCGAATGCAAATCACTAGAA-3'. The oligonucleotides were synthesized by Roche. Densitometric evaluation of the transcription factor DNA binding was done and used to calculate the fold variations.

Confocal Laser Scanning
MCF-7 and MDA-MB231 cells (4 x 104) were seeded on sterile coverslips, previously placed in 24-multiwell plates, and were allowed to attach (3). The cells were transfected at 70% to 80% of confluence with Srcwt expression vector or were starved and treated with HGF for 30 or 60 min. Then, all the cells were fixed with 4% paraformaldehyde solution and were permeabilized with 0.2% Triton X-100 before the reaction with the primary antibodies. (a) The cells were allowed to react with anti-Met (1:50) and anti–E-cadherin (1:5000) antibodies for 1 h at room temperature. Then, the reaction with the secondary goat anti-rabbit AlexaFluor647 and anti-mouse AlexaFluor568 antibodies was done. After washing, the same cells were subsequently exposed to anti-phospho–c-Src(Tyr416) antibody AlexaFluor488-conjugated (2 µg/mL) for 1 h at room temperature. (b) In other experiments, the cells were incubated at room temperature for 1 h with anti-HDAC3 antibody (1:50). Then, the secondary goat anti-rabbit AlexaFluor568-conjugated antibody was used. After washing, the cells were exposed to anti-phospho–c-Src(Tyr416) antibody AlexaFluor488-conjugated (2 µg/mL) and to anti-phosphotyrosine–AlexaFluor647–Cys–conjugated antibody (1:10) for 1 h. The antibodies anti-Met, anti–E-cadherin, and anti-HDAC3 were diluted in PBS containing 0.5% bovine serum albumin, whereas all the other antibodies were in PBS. To evaluate phospho–c-Src levels, the cells were pretreated with 1 mmol/L pervanadate for 1 min (40). For negative control staining, only the secondary antibodies were incubated with the cells. No positive staining was observed with secondary antibodies alone. Nuclear staining was done with DAPI (1:2,000). The coverslips were mounted with a glycerol-based solution and observed under Leica TCS SP2-A0BS confocal microscope. Images were taken at 260-nm intervals (63x magnification) and analyzed with the appropriate software (3).

Statistical Analysis
Densitometric and luciferase activity values were analyzed by ANOVA, with P < 0.05 considered significant. Differences from controls were evaluated on original experimental data, and then we assigned to controls the arbitrary value of 1.


   Acknowledgements
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
References
 
We thank Drs. Mario Faretta and Pietro Dransidico of Istituto Oncologia Molecolare (IFOM, Milan, Italy) for confocal laser scanner analysis.


   Notes
Top
 Notes
Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 
Grant support: Ministero Istruzione Università Ricerca and Ministero della Salute, Italy.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: E. Matteucci and E. Ridolfi contributed equally to the paper.

Received 1/31/07; revised 3/26/07; accepted 5/ 1/07.


   References
Top
 Notes
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 Acknowledgements
 References
 

  1. Boccaccio C, Comoglio PM. Invasive growth: a Met-driven genetic programme for cancer and stem cells. Nat Rev Cancer 2006;6:637–45.[CrossRef][Medline]
  2. Trusolino L, Cavassa S, Angelini P, et al. HGF/scatter factor selectively promotes cell invasion by increasing integrin avidity. FASEB J 2000;14:1629–40.[Abstract/Free Full Text]
  3. Matteucci E, Ridolfi E, Desiderio MA. Hepatocyte growth factor differently influences Met–Ecadherin phosphorylation and downstream signalling pathway in two models of breast cells. Cell Mol Life Sci 2006;63:2016–26.[Medline]
  4. Ostapkowicz A, Inai K, Smith L, Kreda S, Spychala J. Lipid rafts remodelling in estrogen receptor-negative breast cancer is reversed by histone deacetylase inhibitor. Mol Cancer Ther 2006;5:238–45.[Abstract/Free Full Text]
  5. Wojcik EJ, Sharifpoor S, Miller NA, et al. A novel activating function of c-Src and Stat3 on HGF transcription in mammary carcinoma cells. Oncogene 2006;25:2773–84.[CrossRef][Medline]
  6. Sheen-Chen SM, Liu YW, Eng HL, Chou F-F. Serum levels of hepatocyte growth factor in patients with breast cancer. Cancer Epidemiol Biomarkers Prev 2005;14:715–7.[Abstract/Free Full Text]
  7. Elliott BE, Hung WL, Boag AH, Tuck AB. The role of hepatocyte growth factor (scatter factor) in epithelial-mesenchymal transition and breast cancer. Can J Physiol Pharmacol 2002;80:91–102.[CrossRef][Medline]
  8. Khoury H, Dankort DL, Sadekova S, Naujokas MA, Muller W, Park M. Distinct tyrosine autophosphorylation sites mediate induction of epithelial mesenchymal like transition by an activated ErbB-2/Neu receptor. Oncogene 2001;20:788–99.[CrossRef][Medline]
  9. Lengyel E, Prechtel D, Resau JH, et al. c-Met overexpression in node-positive breast cancer identifies patients with poor clinical outcome independent of Her2/neu. Int J Cancer 2004;113:678–82.
  10. Tacchini L, Matteucci E, De Ponti C, Desiderio MA. Hepatocyte growth factor signalling regulates transactivation of genes belonging to the plasminogen activation system via hypoxia inducible factor-1. Exp Cell Res 2003;290:391–401.[CrossRef][Medline]
  11. Matteucci E, Locati M, Desiderio MA. Hepatocyte growth factor enhances CXCR4 expression favouring breast cancer cell invasiveness. Exp Cell Res 2005;310:176–85.[CrossRef][Medline]
  12. Balkwill F. Chemokine biology in cancer. Semin Immunol 2003;15:49–55.[CrossRef][Medline]
  13. Balkwill F. Cancer and chemokine network. Nat Rev Cancer 2004;4:540–50.[CrossRef][Medline]
  14. Ishizawar R, Parsons SJ. c-Src and cooperating partners in human cancer. Cancer Cell 2004;6:209–14.[CrossRef][Medline]
  15. Bertotti A, Comoglio PM. Tyrosine kinase signal specificity. Trends Biochem Sci 2003;28:527–33.[CrossRef][Medline]
  16. Moro L, Dolce L, Cabodi S, et al. Integrin-induced epidermal growth factor (EGF) receptor activation requires c-Src and p130Cas and leads to phosphorylation of specific EGF receptor tyrosines. J Biol Chem 2002;277:9405–14.[Abstract/Free Full Text]
  17. Owens DW, McLean GW, Wyke AW, et al. The catalytic activity of Src family kinases is required to disrupt cadherin-dependent cell-cell contacts. Mol Biol Cell 2000;11:51–64.[Abstract/Free Full Text]
  18. Fujita Y, Krause G, Scheffner M, et al. Hakai, a c-Cbl–like protein, ubiquitinates and induces endocytosis of the E-cadherin complex. Nat Cell Biol 2002;4:222–31.[CrossRef][Medline]
  19. Summy JM, Gallick GE. Src family kinases in tumor progression and metastasis. Cancer Metastasis Rev 2003;22:337–58.[CrossRef][Medline]
  20. Gao Z, Chiao P, Zhang X, et al. Coactivators and corepressors of NF-{kappa}B in I{kappa}B{alpha} gene promoter. J Biol Chem 2005;280:21091–8.[Abstract/Free Full Text]
  21. Longworth MS, Laimins LA. Histone deacetylase 3 localizes to the plasma membrane and is substrate of Src. Oncogene 2006;25:4495–500.[Medline]
  22. Penuel E, Martin GS. Transformation by v-Src: Ras-MAPK and PI3K-mTOR mediate parallel pathways. Mol Biol Cell 1999;10:1693–703.[Abstract/Free Full Text]
  23. Trevino JG, Summy JM, Lesslie DP, et al. Inhibition of Src expression and activity inhibits tumor progression and metastasis of human pancreatic adenocarcinima cells in an orthotopic nude mouse model. Am J Pathol 2006;168:962–72.[Abstract/Free Full Text]
  24. Slyke PV, Coll ML, Master Z, Kim H, Filmus J, Dumont DJ. Dok-R mediates attenuation of epidermal growth factor-dependent mitogen-activated protein kinase and Akt activation through processive recruitment of c-Src and Csk. Mol Cell Biol 2005;25:3831–41.[Abstract/Free Full Text]
  25. Maroni P, Bendinelli P, Matteucci E, Desiderio MA. HGF induces CXCR4 and CXCL12-mediated tumor invasion through Ets1 and NF-{kappa}B. Carcinogenesis 2007;28:267–79.[Abstract/Free Full Text]
  26. Guo Y, Pakneshan P, Gladu J, Slack A, Szyf M, Rabbani SA. Regulation of DNA methylation in human breast cancer. J Biol Chem 2002;277:41571–9.[Abstract/Free Full Text]
  27. González L, Agulló-Ortuño MT, García-Martínez JM, et al. Role of c-Src in human MCF7 breast cancer cell tumorigenesis. J Biol Chem 2006;281:20851–64.[Abstract/Free Full Text]
  28. Rucci N, Recchia I, Angelucci A, et al. Inhibition of protein kinase c-Src reduces the incidence of breast cancer metastasis and increases survival in mice: implications for therapy. J Pharmacol Exp Ther 2006;318:161–72.[Abstract/Free Full Text]
  29. Stoica GE, Franke TF, Wellstein A, et al. Estradiol rapidly activates Akt via the ErbB2 signalling pathway. Mol Endocrinol 2003;17:818–30.[Abstract/Free Full Text]
  30. Mazzone M, Comoglio PM. The Met pathway: master switch and drug target in cancer progression. FASEB J 2006;20:1611–21.[Abstract/Free Full Text]
  31. Desiderio M.A. Hepatocyte growth factor in invasive growth of carcinomas. Cell Mol Life Sci 2007;64:1341–54.[Medline]
  32. Perkins ND. The Rel/NF-{kappa}B family: friend and foe. Trends Biochem Sci 2000;25:434–40.[CrossRef][Medline]
  33. Helbig G, Christopherson KW II, Bhat-Nakshatri P, et al. NF-{kappa}B promotes breast cancer cell migration and metastasis by inducing the expression of the chemokine receptor CXCR4. J Biol Chem 2003;278:21631–8.[Abstract/Free Full Text]
  34. Bhowmick NA, Neilson EG, Moses HL. Stromal fibroblasts in cancer initiation and progression. Nature 2004;432:332–7.[CrossRef][Medline]
  35. Abounader R, Laterra J. Scatter factor/hepatocyte growth factor in brain tumor growth and angiogenesis. Neuro-oncol 2005;7:436–51.[Abstract]
  36. Vogelstein B, Kinzler KW. Cancer genes and the pathways they control. Nat Med 2004;10:789–99.[CrossRef][Medline]
  37. Ducasse M, Brown MA. Epigenetic aberrations and cancer. Mol Cancer 2006;5:60–70.[Medline]
  38. Boyer B, Bourgeois Y, Poupon MF. Src kinase contributes to the metastatic spread of carcinoma cells. Oncogene 2002;21:2347–56.[CrossRef][Medline]
  39. Myoui A, Mishimura R, Williams PJ, et al. c-Src tyrosine kinase activity is associated with tumor colonization in bone and lung in an animal model of human breast cancer metastasis. Cancer Res 2003;63:5028–33.[Abstract/Free Full Text]
  40. Irby RB, Yeatman TJ. Increased Src activity disrupts cadherin/catenin-mediated homotypic adhesion in human colon cancer and transformed rodent cells. Cancer Res 2002;62:2669–74.[Abstract/Free Full Text]
  41. Kermorgant S, Zicha D, Parker PJ. Protein kinase C controls microtubule-based traffic but nonproteasomal degradation of c-Met. J Biol Chem 2003;278:28921–9.[Abstract/Free Full Text]
  42. Rayala SK, den Hollander P, Balasenthil S, et al. Hepatocyte growth factor-regulated tyrosine kinase substrate (HRS) interacts with PELP1 and activates MAPK. J Biol Chem 2006;281:4395–403.[Abstract/Free Full Text]
  43. Yan D, Guo L, Wang Y. Requirement of dendritic Akt degradation by the ubiquitin-proteasome system for neuronal polarity. J Cell Biol 2006;174:415–24.[Abstract/Free Full Text]
  44. Zeng L, Xiao Q, Margariti A, et al. HDAC3 is crucial in shear- and VEGF-induced stem cell differentiation toward endothelial cells. J Cell Biol 2007;174:1059–69.
  45. Hu J, Colburn NH. Histone deacetylase inhibition down-regulates cyclin D1 transcription by inhibiting nuclear factor-kB/p65 DNA binding. Mol Cancer Res 2005;3:100–9.[Abstract/Free Full Text]
  46. Perkins ND. Integrating cell-signalling pathways with NF-{kappa}B and IKK function. Nat Rev 2007;8:49–62.
  47. Karin M, Cao Y, G