Identification of Insulin-Like Growth Factor-I Receptor (IGF-IR) Gene Promoter-Binding Proteins in Estrogen Receptor (ER)-Positive and ER-Depleted Breast Cancer Cells

The insulin-like growth factor I receptor (IGF-IR) has been implicated in the etiology of breast cancer. Overexpression of the IGF-IR gene is a typical feature of most primary breast cancers, whereas low IGF-IR levels are seen at advanced stages. Hence, evaluation of IGF-IR levels might be important for assessing prognosis. In the present study, we employed a proteomic approach based on DNA affinity chromatography followed either by mass spectroscopy (MS) or Western blot analysis to identify transcription factors that may associate with the IGF-IR promoter in estrogen receptor (ER)-positive and ER-depleted breast cancer cells. A biotinylated IGF-IR promoter fragment was bound to streptavidin magnetic beads and incubated with nuclear extracts of breast cancer cells. IGF-IR promoter-binding proteins were eluted with high salt and analyzed by MS and Western blots. Among the proteins that were found to bind to the IGF-IR promoter we identified zinc finger transcription factors Sp1 and KLF6, ER-α, p53, c-jun, and poly (ADP-ribosylation) polymerase. Furthermore, chromatin immune-precipitation (ChIP) analysis confirmed the direct in vivo binding of some of these transcription factors to IGF-IR promoter DNA. The functional relevance of binding data was assessed by cotransfection experiments with specific expression vectors along with an IGF-IR promoter reporter. In summary, we identified nuclear proteins that are potentially responsible for the differential expression of the IGF-IR gene in ER-positive and ER-depleted breast cancer cells.


Introduction
The insulin-like growth factors, IGF-I and IGF-II, are mitogenic polypeptides with important roles in cell growth and differentiation [1]. The biological activities of IGF-I and IGF-II are mediated by the IGF-I receptor (IGF-IR) and are modulated by a family of IGF-binding proteins (IGFBPs) that control ligand bioavailability [2][3][4]. The IGF-IR is essential for normal cell cycle progression and has a crucial role in tumor growth and metastasis development. Ligand-mediated activation of the IGF-IR tyrosine kinase domain leads to mitogenic and antiapoptotic effects in multiple cells and cancer-derived cell lines [5]. In the specific context of breast tumorigenesis, the IGF-IR is overexpressed in most primary tumors [4]. In addition, and consistent with its potent antiapoptotic role, IGF-IR-null cells do not undergo transformation when exposed to different oncogenes [6,7]. On the other hand, inhibition of IGF-IR signaling by IGF-IR antibodies, antisense oligonucleotides, IGF-BPs, or specific IGF-IR kinase inhibitors results in reduced cell proliferation and cancer progression in vivo and in vitro [8][9][10][11][12][13][14]. Furthermore, epidemiological studies revealed that high levels of circulating IGF-I are linked to an increased risk of developing breast cancer in premenopausal women [15,16].
Regulation of IGF-IR gene expression is mainly attained at the transcription level. The IGF-IR promoter is a TATA-less, CCAAT-less, highly GC-rich, 'initiator'-type of promoter. IGF-IR gene transcription is dependent on a number of stimulatory zinc-finger nuclear proteins, including Sp1 [17] and KLF6 [18]. In addition, IGF-IR gene transcription is negatively regulated by several tumor suppressors, including BRCA1, p53/p63/p73, the von Hippel-Lindau protein (VHL), and the Wilms' protein-1 (WT1) [19][20][21][22][23][24][25]. Interactions between stimulatory and inhibitory transcription factors play an important role in IGF-IR gene regulation and, therefore, were postulated to have a major impact on the proliferative status of the cell. The molecular mechanisms and specific transcription factors responsible for regulating IGF-IR gene expression in breast cancer cells, however, have not yet been identified.
The IGF-I and estrogen signaling systems were shown to act in a synergistic fashion in breast epithelial cells [25]. Estrogens regulate IGF-I signaling and the expression of several members of the IGF system [25][26][27][28][29][30][31]. Moreover, activation of estrogen receptor- (ER by estrogens induces a physical interaction between ER and IGF-IR [32] that results in activation and phosphorylation of IGF-IR and downstream signaling molecules [33][34][35][36]. The aim of this study was to identify the collections of IGF-IR promoter-binding transcription factors in ER-positive and ER-depleted breast cancer cells. Using DNA affinity chromatography, mass spectroscopy (MS), and Western blot analyses, we identified a series of known and previously unidentified transcription factors that specifically bind to the IGF-IR promoter in either cell type. The ability of selected proteins to bind and transactivate the IGF-IR promoter was confirmed by chromatin immunoprecipitation (ChIP) and transient transfection assays. In addition, we identified a number of non-DNA sequence-specific nuclear proteins that are probably involved in IGF-IR gene regulation. Identification of differentially expressed IGF-IR promoter-binding and non-binding transcription factors may help elucidate the mechanisms responsible for the differential expression of the IGF-IR gene in ER-positive and ERdepleted breast cancer cells.

PCR and DNA Affinity Chromatography of the IGF-IR Promoter
For DNA affinity chromatography, a 511-bp human proximal IGF-IR promoter fragment extending from nucleotides −458 to +53 was labeled using a 5'-biotinylated antisense primer. This fragment includes most of the proximal 5'-flanking region and comprises the 'initiator' motif from which transcription starts in vivo (Nucleotide 1) ( Figure 1).
Primer sequences were derived from genomic IGF-IR clone #R73P2 [38] as follows: sense, 5'-CTTTCCAGCCGCGCTGTTGTTG-3'; antisense, 5'-(Biotin)-GGTAAACAAGAGCCCCAGCCTC-3'. PCR was performed using the TermalAce TM DNA Polymerase reagent (InVitrogen, Carlsbad, CA, USA). The biotinylated PCR product was bound to streptavidin magnetic beads (Dynabeads ® M-270 Streptavidin; Dynak Biotech ASA, Oslo, Norway), and incubated with nuclear extracts of MCF7 or C4.12.5 cells in the presence of double stranded poly dI-dC (Amersham Pharmacia Biotech, Piscataway, NJ, USA). Following high-salt elution, specifically-bound proteins were resolved on 10% SDS-PAGE gels and visualized by silver staining (Figure 2). Gel slices were then submitted for MS analysis at the Smoler Proteomics Center (Technion-Israel Institute of Technology, Haifa, Israel) or analyzed by Western blotting as described below.  [5]. The initiator (INR) element is denoted by an arrow. The coding region, starting with the AUG codon, is shown in gray. The translation start site is denoted by a dashed arrow. The 5'-flanking region is denoted by a black bar, and the 5'-untranslated region (UTR) is represented by a dotted bar. The location of primers (P#1 and P#2), employed to amplify the proximal promoter is indicated. (B) PCR amplification of the human IGF-IR promoter. The PCR reaction was performed using a biotinylated antisense primer, as described under Materials and Methods section, and 4 L of the biotinylated PCR product was loaded on a 1% agarose gel. A PCR reaction without DNA served as negative control. A 100-bp DNA ladder was used as a M.W. marker.

Figure 2.
Silver staining of IGF-IR promoter-bound proteins identified by DNA affinity chromatography. Nuclear extracts of MCF7 cells were prepared as described under Materials and Methods Section. Proteins bound to IGF-IR promoter DNA (10 μg) were electrophoresed through 10% SDS-PAGE, fixed, and stained with silver (Bio-Rad, Hercules, CA, USA). Lane 1, M.W. marker; lane 2, starting material (MCF7 nuclear extract, 4.2 g); lane 3, unbound material (MCF7 nuclear extract that did not bind to DNA, 4.2 g); lane 4, eluted material bound to Biotin-520 bp (MCF7 nuclear proteins that bound to IGF-IR promoter, 10 g); and lane 5, negative control (MCF7 nuclear proteins bound to strepavidin magnetic beads, 10 g).

Gel Slices Proteolysis and Mass Spectroscopy Analysis
The gel slices containing the specific DNA-bound proteins were reduced with 10 mM DTT, incubated at 60 °C for 30 min, carboxyamidomethylated with 40 mM iodoacetamide at room temperature for 30 min, and trypsinized overnight at 37 °C [modified trypsin (Promega)] at a 1:100 enzyme-to-substrate ratio. The tryptic peptides were resolved by reverse-phase chromatography on 0.075 × 200-mm fused silica capillaries (J&W, Wilmington, DE, USA) packed with Reprosil reversed phase material (Dr Maisch GmbH, Germany). The peptides were eluted with a linear 90 min gradient from 5 to 45% acetonitrile containing 0.1% formic acid, followed by a wash of 95% acetonitrile for 15 min. The flow rate was 0.25 μl/min and the peptides were electrosprayed into an ion-trap mass spectrometer (Orbitrap, Thermo Fisher Scientific Inc., USA). MS was performed in a positive mode using repetitively full MS scans followed by collision induced dissociation (CID) of the 7 most dominant ions selected from the first MS scan. The MS data was clustered and analyzed using the Sequest software (Finnigan, San Jose, CA, USA) and Pep-Miner [39] searching against International Protein Index (IPI) protein IDs (part of the NR-NCBI database).

Cell Fractionation
Whole cell extracts were prepared by lysing cells in a buffer containing protease inhibitors, as described previously [20]. To prepare cytosolic and nuclear extracts, cells were washed with ice-cold phosphate-buffered saline (PBS), centrifuged, and the pellet was resuspended in buffer A [10 mM Hepes (pH 7.9), 1.5 mM MgCl 2 , 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, 2 g/mL aprotinin, 1 mM ortovanadate, 2 g/mL leupeptin, 4 mM pyrophosphate, 2 g/mL pepstatin, and 1 mM PMSF] and incubated for 30 min on ice. IGEPAL (CA-630, Sigma-Aldrich Co.) was then added to a final concentration of 3%, vortexed, and centrifuged (14,000 rpm, 30 sec), after which the supernatant containing the cytosolic fraction was saved at −70 °C. Buffer C [20 mM Hepes, 420 mM NaCl, 1 mM EDTA, 1 mM EGTA, and protease inhibitors] was added to the nuclei-containing pellet, incubated for 30 min at room temperature, and spun at 14,000 rpm for 5 min. The supernatant containing the nuclear extract was frozen at −70 °C. Protein concentrations were determined with the Bradford reagent (Bio-Rad, Hercules, CA, USA) using bovine serum albumin (BSA) as a standard.

Western Immunoblots
Samples were electrophoresed through 12%, 10%, or 7.5% SDS-PAGE gels, followed by blotting of the proteins onto nitrocellulose membranes. After blocking with either skim milk or 3% BSA, the blots were incubated overnight with the antibodies listed below, washed and incubated with the appropriate horseradish peroxidase (HRP)-conjugated secondary antibody. Antibodies against BRCA1

Chromatin Immunoprecipitation Assays
The ChIP assays were performed as described previously [40]. Briefly, MCF7 cells were crosslinked with 1% formaldehyde at 37 °C for 10 min. Cells were rinsed with ice-cold PBS and centrifuged. The pellet was then resuspended in lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl, pH 8.1 and protease inhibitor mixture: 1 mM PMSF, 1 g/mL aprotinin and 1 g/ml pepstatin A) and sonicated for 3 min. The supernatant fraction was immunoprecipitated with anti-PARP, anti-cjun, anti-Sp1, anti-KLF6, anti-HA, and NRS or NMS for 18 h at 4 °C. The next day, Protein A/G plus agarose beads (SC-20003, Santa Cruz Biotechnology) was added for one hour at 4 °C. Immunoprecipitates were electrophoresed and immunoblotted as described under Western immunoblots. For PCR analysis of antibody-immunoprecipitated chromatin, a set of primers encompassing the IGF-IR proximal promoter region, sense, 3'-CCAGCCGCGCTGTTGTTG-5', antisense: 3-CCAGCCGCGCTGTTGTTG-5, and Thermolace TM DNA polymerase kit were used.

DNA Transfections
For transient cotransfection experiments an IGF-IR promoter luciferase reporter construct including 476 bp of 5'-flanking and 640 bp of 5'-untranslated regions of the IGF-IR gene [p(−476/+640) luciferase (LUC)] was employed [41,42]. An expression vector encoding c-jun (RSV-c-jun) was provided by Dr. Lili Vardimon (Tel Aviv University). An HA-HMGA1 expression vector (in pcDNA3) was provided by Dr. G. Manfioletti (University of Trieste, Italy) [43]. A KLF6 expression vector (pCI-neo-KLF6) was provided by Dr. Scott L. Friedman (Mount Sinai School of Medicine, New York, USA). The E2F1 expression vector, pcDNA-I-E2F-1, was previously described [44]. Cells were seeded in six-well plates the day before transfection and cotransfected with 1 g of the IGF-IR promoter reporter along with 1g of the E2F1 expression plasmid (or empty pCDNA3), or 2 g of the HA-HMGA1 (or empty pCDNA3), or 2 g of the pCI-neo-KLF6 (or empty pCI-neo) using the Jet-PEI (Polyplus, Illkirch, France) transfection reagent. Cells were harvested 48 h after transfection, and luciferase activity was measured as previously described [41]. Promoter activities were expressed as luciferase values normalized to protein concentrations. Protein content was determined using the Bradford reagent (Bio-Rad, Hercules, CA, USA) using BSA as a standard.

ER and IGF-IR Expression in MCF7 and C4.12.5 Cells
The involvement of the IGF-IR in breast cancer development has been well established. To address the potential regulation of IGF-IR expression during breast cancer progression, we measured in initial experiments the levels of total and phospho-IGF-IR (pIGF-IR) in syngeneic cell lines representing different stages of the disease. For this purpose, we employed the human breast cancer-derived MCF7 (ER-positive) and MCF7-derived C4.12.5 (ER-depleted) cell lines. C4.12.5 cells were generated by clonal selection of MCF7 cells that were maintained in estrogen-free conditions for nine months [37]. Although C4.12.5 cells exhibited some variability in ER expression, the passages used in the present study displayed consistently low levels of ER (~59% of the ER levels in MCF7 cells). Western blots using anti-IGF-IR -subunit and anti-pIGF-IR antibodies showed that both total IGF-IR and pIGF-IR levels were significantly reduced in C4.12.5, in comparison to, MCF7 cells (76% for IGF-IR and 30% for pIGF-IR) ( Figure 3). These results corroborate previous observations [40,45]. Figure 3. IGF-IR and ER expression in MCF7 and C4.12.5 breast cancer cell lines. MCF7 and C4.12.5 cells were lysed as described under Materials and Methods. Cellular extracts (80 μg) were electrophoresed through 10% SDS-PAGE, transferred to nitrocellulose membranes, and blotted with anti-total IGF-IR-, anti-pIGF-IR, and anti-ER. Blots were then blotted with a tubulin antibody as a loading control.

DNA Affinity Chromatography
The MCF7 and C4.12.5 cell lines were employed to identify the sets of IGF-IR promoter-binding proteins that may be involved in IGF-IR transcription and that are responsible for the differential expression of the gene at different stages of the disease. The rationale for these studies was the fact that regulation at the transcription level has been identified as a major control step in IGF-IR regulation [46]. To this end, we used a proteomic approach based on DNA affinity chromatography followed either by MS or Western blot analyses using antibodies against previously identified or novel IGF-IR promoterbinding proteins. The basis for selecting this specific fragment was the fact that previous studies have shown that this region is responsible for most of the promoter activity. Nuclear extracts of MCF7 and C4.12.5 cells were incubated with the PCR-amplified, biotin-labeled IGF-IR proximal promoter DNA probe, after which DNA-protein complexes were adsorbed to streptavidin beads. Bound proteins were eluted with high-salt buffer and concentrated by centrifugation at reduced pressure. The eluted material was electrophoresed through 10% SDS-PAGE and proteins were visualized by silver staining (Figure 2). A number of bands were seen in the eluted material (lane 4) whereas no proteins were detected in eluates of lysates incubated with strepavidin magnetic beads alone (lane 5). Silver stained gels were cut into slices and the specific DNA-bound proteins were analyzed by MS.

MS Analysis
MS analysis of DNA affinity chromatography-purified proteins identified a number of previously reported and several novel IGF-IR promoter binding proteins. The proteins identified were categorized using the Gi (gene index) identification tool, coverage (percentage of amino acids covered by the identified peptide), and source ( Table 1). The specific proteins bound to the IGF-IR promoter were classified based on cellular localization and biological function ( Table 2). Most of the proteins were localized in the nucleus, but some of them were found also in the cytoplasm as well as in the plasma membrane. Furthermore, Table 2 illustrates the variety of biological processes associated with the identified IGF-IR promoter-binding transcription factors in ER-positive and ER-depleted breast cancer cells. Out of 91 proteins identified by MS, 6 correspond to cytoskeleton-associated proteins, 6 are involved in transcription, regulation of nucleobase, nucleoside and nucleic acid metabolism, 2 in nuclear stability, chromatin structure, cycle control, and gene expression, 20 in DNA repair, breaking, replication, and cell death, 24 in RNA splicing and processing, and translation, and 33 in other functions, including proliferation, apoptosis, and proteosomal degradation.

Identification of IGF-IR Promoter-Binding Transcription Factors by Western Blots
Western blot analysis of DNA affinity chromatography-purified proteins identified a number of previously reported as well as novel IGF-IR promoter binding proteins. The data are summarized in Figure 4 and Tables 3 and 4. Consistent with their role as positive regulators of the IGF-IR gene, the binding of zinc finger proteins Sp1 and KLF6 to the IGF-IR promoter was stronger in MCF7 than in C4.12.5 cells. Similarly, ER- binding to the IGF-IR promoter was stronger in MCF7 than in C4.12.5 cells. On the other hand, a number of transcription factors, including AR, Smad2/3, E2F1, cjun, TCF1A, TCF3A, GSK3, and p21 appear to bind to the promoter region only in the C4.12.5 cell line. p53, c-fos, and PARP bind to IGF-IR promoter elements in both cell lines with similar affinity. No DNA binding was detected in either cell line using antibodies against BRCA1, TOPOII, IRS-1, MET, p120, Foxo1, p73, lamin A/C, c-Myc, p63, cyclin D, cyclin G, histone1, and BRCA2.     AR Androgen receptor (AR), a zinc finger transcription factor of the nuclear receptor superfamily. SP1 Transcription factor of the zinc finger family. Foxo1 Transcription factor of the forkhead family.

ER
Estrogen receptor (ER), a DNA binding transcription factor of the nuclear receptor superfamily.
c-Myc Nuclear proteins involved in cell proliferation, differentiation and neoplastic disease. E2F1 Involved in cell cycle progression, apoptosis and oncogenesis. LEF1 TCFA1 TCF3A Members of the high mobility group (HMG) DNA binding protein family of transcription factors.

GSK3
Involved in protein synthesis, cell adhesion, proliferation, and differentiation. c-fos c-jun pc-jun Fos and Jun protooncogenes.

Cyclin G Cyclin D
Regulatory subunits of Cdc2 p34 and related cyclin-dependent kinases (Cdks).

Histone1
Basic and water soluble nuclear proteins. p21 Tumor suppressor protein p21/Waf1/Cip1 acts as an inhibitor of cell cycle progression.

Cellular Distribution of IGF-IR Promoter-Binding Proteins
Given that differences in binding could be due to differential expression of the transcription factors in both cell lines, we next investigated the subcellular distribution and expression of IGF-IR promoterbinding transcription factors. For this purpose, MCF7 and C4.12.5 lysates were fractionated into cytoplasmic and nuclear fractions and the subcellular distribution of the different proteins was examined by Western blots (Figure 5 and Tables 3 and 4). The cytosolic marker MET [47] was used to check for cytoplasmic contamination of the nuclear fraction, and it was found to be expressed at very low levels in both cell lines. A number of proteins, including PARP (200% of levels in MCF7 cells), lamin (240%), Smad 2/3 (260%), and c-jun (1000%) had higher expression levels in C4.12.5 than in MCF7 cells. Other proteins, such as BRCA1 (30% of levels in MCF7), AR (64%), Sp1 (79%), Foxo I (50%), p73 (3%), TCF3A (0.2%), ER (59%), TCF1 (71.7%), pc-jun (50%), cyclin G (13%), and histone (74%) were expressed at lower levels in C4.12.5 compared to MCF7 cells. No differences in expression between the cell lines were detected in p120, c-fos, WT1, KLF6, pSmad2/3, c-Myc, p63, E2F, LEF1, p53, GSK3, and p21. We were unable to detect expression of -catenin, p300, TLE1, APC, and BRCA2 in neither cell line. Transcription factor WT1 was present in both cell lines, though its low molecular weight form appears to be specific to the MCF7 cell line.

ChIP Assays
To corroborate results of DNA chromatography we measured the binding of selected proteins identified by MS or by Western blots using ChIP assays. The following proteins were chosen for ChIP assays: PARP (detected by MS and Western blots in both cell lines), Sp1 (positive control, detected in both cell lines), c-jun, (detected only in C4.12.5) and HMGA1 (detected by MS only in MCF7). Briefly, cells were lysed, sonicated for 3 min, and immunoprecipitated with the indicated antibodies (or normal mouse or rabbit sera) for 18 h ( Figure 6A). In addition, MCF7 cells transfected with HA-HMGA1 (or empty pCDNA3 vector) were also immunoprecipitated with an HA monoclonal antibody ( Figure 6B). For PCR analysis of immunoprecipitatd chromatin, a set of primers encompassing the human IGF-IR promoter was employed as described under Materials and Methods section. The input bands represent the amplified PCR product in the absence of antibodies (Figures 6A and 6B). The results of ChIP analyses confirmed that PARP, c-jun, Sp1, KLF6 and HMGA1 bound in vivo directly to the IGF-IR promoter and, thus, corroborate the results of MS and Western blot analyses. Figure 6. ChIP assays of transcription factor binding to IGF-IR promoter DNA. (A). MCF7 cells were cross-linked with formaldehyde, lysed, sonicated, and immunoprecipitated with PARP, c-jun, Sp1, or KLF6 antibodies, or normal mouse or rabbit sera, followed by PCR amplification of precipitated chromatin using primers encompassing the IGF-IR promoter. The position of the 773 bp-amplified fragments is indicated. The input bands represent the amplified PCR product in the absence of antibodies. The insets indicate the endogenous levels of expression of the various proteins as measured by Western blots (WB) with specific antibodies or NRS or NMS as negative controls. IP, immunoprecipitation. (B). MCF7 cells were transfected with an expression vector encoding HA-HMGA1 (or empty pcDNA3 vector), after which the cells were cross linked with formaldehyde, lysed, sonicated and immunoprecipitated with an HA antibody, followed by PCR amplification of IGF-IR promoter DNA. HA-HMGA1 immunoprecipitated protein was detected by Western blots using an anti-HA antibody (inset).

Regulation of IGF-IR Promoter Activity by c-jun, HMGA1, KLF6 and E2F1
To evaluate the potential functional implications of our findings we next assessed the ability of selected IGF-IR promoter-binding proteins to activate the promoter. For this purpose cotransfection experiments were performed in MCF7 and C4.12.5 cells using expression vectors encoding c-jun, HMGA1, KLF6 and E2F1, along with an IGF-IR promoter-luciferase reporter plasmid, p(−476/+640)LUC, which contains most of the proximal region of the IGF-IR promoter. The results of cotransfection experiments are presented in Figures 7A-D. Results obtained indicate that transcription factors c-jun, KLF6, and E2F1 induced a significant increase in IGF-IR promoter activity in both cells lines. Specifically, c-jun enhanced IGF-IR promoter activity by 445 ± 12.4% in MCF7 and by 209  3.1% in C4.12.5 cells (versus empty vector), KLF6 enhanced promoter acitiviy by 237 ± 9.7% in MCF7 and by 159  7.1% in C4.12.5 cells, and E2F1 stimulated promoter activity by 631.5 ± 44.5% in MCF7 and by 1127  22.2% in C4.12.5 cells. On the other hand, HMGA1 had a minimal effect (131 ± 4.3% increase in MCF7 and 80.3  1.7% reduction in C4.12.5 cells).

Discussion
Breast cancer is one of the main causes of cancer-related mortality among women in economically developed regions of the world. Care of patients with breast cancer is complicated by the wide differences in the natural history of the disease, as manifested by a remarkable variability in response to treatment and survival. Over 100 potential biomarkers were reported in various studies to be differentially expressed between normal and cancerous cells and tissues. Cancer biomarkers include, among others, ER, progesterone receptor, HER2, cathepsin D, cyclin E, CA 15-3, and CA 27.29. Currently, only a few of those biomarkers are used in clinical practice. High-throughput genomic and proteomic techniques developed over the last decade offer the potential to better define the biological nature of the disease process, both for risk assessment and therapy purposes [48].
There is conflicting evidence regarding the importance of IGF-IR expression for diagnosis and prognosis of breast cancer. In different studies, IGF-IR expression was detected in 39% to 93% of the cases [49]. Whereas some studies reported that IGF-IR levels are indicative of early recurrence of the disease (i.e., poor prognosis), others considered high IGF-IR values as an indication of good prognosis [4,[50][51][52][53][54]. These disparities may be due to a number of factors, including selection of patients or the diverse approaches used to detect IGF-IR, including immunoblots of membrane fractions and/or whole tissues, ligand binding assays, etc. Immunohistochemical studies revealed high IGF-IR and IRS-1 levels in control breast tissues and in well-differentiated breast carcinoma, whereas very low levels were seen in poorly differentiated cancers [3]. These results, suggesting that progression to advanced-stage disease is associated with a reduction in IGF-IR expression, indicate that IGF-IR status can be used as a marker for tumor evaluation [3]. Notably, no correlation was seen between insulin receptor levels and the differentiation stage of the tumors.
In the present study we employed a novel and reliable DNA affinity chromatography protocol in order to identify IGF-IR promoter-binding proteins in breast cancer cells. This method allowed us to detect known and new specific potential biomarkers in ER-positive and ER-depleted cells, which may reflect early and advanced stages of the disease. Our analyses identified twenty four proteins that bind to the IGF-IR promoter only in MCF7 cells whereas, on the other hand, C4.12.5 cells contain nineteen proteins that did not exhibit IGF-IR promoter binding activity in MCF7 cells. The results obtained suggest that the specific binding of different DNA-binding transcription factors and non-DNA sequence specific-binding nuclear proteins (including Smad2/3, LEF1, c-jun, TCF1A, TCF3A, GSK3and others) in the MCF7 and C4.12.5 cell lines could reflect differences in transcriptional regulation of the IGF-IR gene at different stages of the disease. Interestingly, sixty two proteins were shown to bind (either directly or indirectly) to IGF-IR promoter elements in both cell lines. As shown in Tables 2 and 4, the list of proteins identified in this study includes proteins involved in apoptosis, protein synthesis, cell adhesion, proliferation, differentiation, nuclear stability, chromatin structure, DNA repair, recombination, replication, tumor suppression, and oncogenesis. In this context, it has been previously shown that the regulatory region of the IGF-IR gene constitutes a molecular target for a number of transcription factors with tumor suppressor or stimulatory activity [46]. Under normal physiological conditions, IGF-IR expression is tightly regulated by complex interactions involving positively acting (Sp1, KLF-6, ER) and negatively acting (p53, p73, BRCA1, ATM, WT1, VHL) transcription factors [17,18,[22][23][24]40,45,[55][56][57]. Elevated IGF-IR levels were postulated to be the result of loss-of-function mutations of tumor suppressor genes in cancer cells [58]. The approach employed in this study allowed us to identify several previously reported IGF-IR promoter-binding proteins, including Sp1, KLF6, and WT1. Moreover, results of MS and Western blots confirmed bioinformatic predictions on transcription factor binding sites (e.g., WT1, ER, E2F, Sp1, etc.), obtained using the TRANSFAC and AliBaba2 softwares (http://www.biobase-international.com and http://www.gene-regulation.com) (data not shown). Furthermore, the fact that we were able to detect differences in promoter binding between the two cell lines, despite the lack of differences in protein expression, indicates that the DNA affinity chromatography approach used was highly specific and sensitive. Given that differences in binding could be due to differences in the availability of these proteins, we confirmed that the proteins detected in the eluates were truly present in the nuclear fractions. Several proteins, including KLF6, Smad2/3, GSK3, c-jun, and p21, were present in both cytosolic and nuclear fractions. A number of transcription factors, including BRCA1, p120, Foxo1, p73, c-myc, p63, cyclin D and G, were undetectable in eluates of both cell lines. A potential explanation for these results is the fact that the expression of these transcription factors was very low in the nuclear fractions [59] or, alternatively, that they are engaged in multi-protein complexes and, therefore, cannot bind to IGF-IR promoter sequences.
The capacity of a number of proteins identified by MS and/or Western blots (PARP, Sp1, c-jun, HMGA1) to directly bind to IGF-IR promoter DNA in vivo was confirmed by ChIP assays. Furthermore, the functional relevance of binding data was addressed by coexpression experiments which revealed that c-jun, KLF6, and E2F1 significantly enhanced IGF-IR promoter activity. Differences in the capacity of these factors to stimulate IGF-IR promoter activity was seen between both cell lines. The luciferase assays corroborate the results obtained by Western blots and ChIP assays. We have previously shown using EMSA and ChIP that both Sp1 and ER bind to the IGF-IR promoter and stimulate gene transcription, whereas no ER binding was seen in C4 cells [60]. In addition, we have reported that Sp1 increased IGF-IR promoter activity with a 5.8-fold higher potency in MC7 than in C4 cells [60]. Taken together, there is a remarkable consistency between the various technological approaches used to identify IGF-IR promoter-binding transcription factors.

Conclusions
In summary, using DNA affinity chromatography we identified a collection of known and novel IGF-IR promoter-binding transcription factors in breast cancer cell lines. These proteins may constitute potential biomarkers characteristic of ER-positive or ER-negative tumors. The potential clinical relevance of our results need to be confirmed by future studies on biopsies from patients at early and advanced stages, and need to be correlated with conventional prognostic factors such as tumor size, lymph nodes status, histological grade, and ER status.