IQGAP1 Binds to Estrogen Receptor-α and Modulates Its Function*

Background: IQGAP1 is a scaffold protein that modulates diverse signaling pathways. Results: IQGAP1 binds to estrogen receptor-α (ERα) and ERβ, and knockdown of IQGAP1 impairs 17β-estradiol (E2)-stimulated transcriptional activation in human breast epithelial cells. Conclusion: IQGAP1 modulates ERα function and is required for maximal E2 function. Significance: IQGAP1 might be a therapeutic target for patients with breast carcinoma. The estrogen receptor (ER) is a steroid hormone receptor that acts as a transcription factor, modulating genes that regulate a vast range of cellular functions. IQGAP1 interacts with several signaling proteins, cytoskeletal components, and transmembrane receptors, thereby serving as a scaffold to integrate signaling pathways. Both ERα and IQGAP1 contribute to breast cancer. In this study, we report that IQGAP1 binds ERα and ERβ. In vitro analysis with pure proteins revealed a direct interaction between IQGAP1 and ERα. Investigation with multiple short fragments of each protein showed that ERα binds to the IQ domain of IQGAP1, whereas the hinge region of ERα is responsible for binding IQGAP1. In addition, IQGAP1 and ERα co-immunoprecipitated from cells, and the association was modulated by estradiol. The interaction has functional effects. Knockdown of endogenous IQGAP1 attenuated the ability of estradiol to induce transcription of the estrogen-responsive genes pS2, progesterone receptor, and cyclin D1. These data reveal that IQGAP1 binds to ERα and modulates its transcriptional function, suggesting that IQGAP1 might be a target for therapy in patients with breast carcinoma.

tion-2 (region E). Notwithstanding these similarities, the receptors differ in both their tissue distribution and functions (1,2). 17␤-Estradiol (E 2 ) binding to the ligand-binding domain of ER␣ can produce distinct effects in the target cells by stimulating two pathways. These are the "classic" pathway in the nucleus that enhances gene transcription and extranuclear actions that are initiated from ER␣ in the plasma membrane and result in rapid actions of E 2 by activating signaling pathways (3). In the classic pathway, E 2 induces a change in ER␣ conformation, resulting in dimerization of the receptor. In the nucleus, the E 2 -ER␣ complex binds either directly to estrogen response elements in the promoters of target genes or indirectly through protein-protein interactions with other transcription factor complexes like Fos-Jun (activator protein-1 response elements) and influences transcription of genes that lack estrogen response elements (2,4). Multiple ER␣ target genes have been identified, including activator protein-1 (5), specific protein-1 (6), progesterone receptor (PR) (7), and pS2/trefoil factor 1 (8).
Distinct co-regulatory proteins are recruited to ER␣ and modulate ER␣ function by serving as co-repressors or co-activators (2). Accumulating evidence reveals that ER␣ associates with interconnected networks of proteins that maintain the structure and function of the receptor and influence estrogenresponsive gene expression (9). One of the proteins that interact with ER␣ is the Ca 2ϩ signaling protein calmodulin. Prior work from our laboratory characterized the interaction between ER␣ and calmodulin. We demonstrated that calmodulin binds directly to ER␣ in a Ca 2ϩ -regulated manner (10). Calmodulin binding promotes the stability of ER␣ by sequestering it away from the ubiquitin-proteasome pathway (11). Interestingly, calmodulin is also necessary for E 2 -stimulated transcriptional activation of ER␣ (12). Thus, specific cellular proteins can influence ER␣ degradation and transcriptional activity.
IQGAP1 is a ubiquitously expressed scaffold protein that associates with a wide repertoire of binding partners. Almost 100 proteins have been identified that interact with IQGAP1 either directly or in a multiprotein complex (13,14). These range from signaling proteins and small GTPases to cytoskeletal components and kinases (13)(14)(15). IQGAP1 also binds to and regulates the function of selected growth factor receptors, such as human epidermal growth factor receptor-2 (16), epi-dermal growth factor receptor (17), fibroblast growth factor receptor (18) and vascular endothelial growth factor receptor-2 (19).
ER␣ target gene expression results from the coordinated action of ER␣ and its co-regulators (4). Most of these co-regulatory proteins contain an LXXLL motif (L, leucine; X, any amino acid) that interacts with the ligand-binding domain of ER␣ (20). Inspection of the amino acid sequence of human IQGAP1 revealed three LXXLL motifs, raising the possibility that IQGAP1 might bind ER␣. In this study, we demonstrated that IQGAP1 and ER␣ associate in vitro and co-immunoprecipitate from cells. Binding was regulated by E 2 , and the ability of E 2 to induce transcriptional activation was impaired in cells in which IQGAP1 was specifically knocked down by siRNA.

EXPERIMENTAL PROCEDURES
Materials-MCF-7, T47D, and HEK293 cells were obtained from the American Type Culture Collection. All reagents for tissue culture were bought from Invitrogen. Protein A-Sepharose and glutathione-Sepharose were purchased from GE Healthcare. Anti-FLAG affinity gel was from Sigma-Aldrich. PVDF membranes were purchased from Millipore Corp. Anti-ER␣ polyclonal and monoclonal antibodies, anti-ER␤ polyclonal antibodies, anti-FLAG monoclonal antibody, rabbit IgG, and mouse IgG were obtained from Santa Cruz Biotechnology. Anti-IQGAP1 polyclonal antibodies have been characterized previously (21). Blocking buffer and infrared dye-conjugated (IRDye) antibodies, both anti-mouse and anti-rabbit, were obtained from LI-COR Biosciences. Recombinant human ER␣ and ER␤ were purchased from Invitrogen.
Preparation of Fusion Proteins-GST-IQGAP1 was expressed in Escherichia coli and isolated using glutathione-Sepharose chromatography essentially as described previously (21). Where indicated, IQGAP1 was further purified by cleaving GST using tobacco etch virus (22). GST-ER␣ was expressed and isolated as described for GST-IQGAP1. The size and purity of the GST proteins were evaluated by SDS-PAGE and Coomassie staining. All proteins were at least 90% pure.
In Vitro Binding Assays-Purified untagged IQGAP1 (1 g) was incubated with 4 g of GST alone or 4 g of GST-ER␣ on glutathione-Sepharose beads in 500 l of Buffer A (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, and 1% Triton X-100) with protease and phosphatase inhibitors (Thermo Scientific) and 1 mM phenylmethylsulfonyl fluoride (PMSF) (Buffer B) for 3 h at 4°C. After washing the beads five times with Buffer A, samples were resolved by SDS-PAGE. The gel was cut at the 100-kDa region. The top part of the gel was transferred to PVDF, blocked with Blocking Buffer (LI-COR Biosciences) for 1 h at 22°C, and then probed with anti-IQGAP1 polyclonal antibodies overnight at 4°C. The membrane was incubated with IRDye-conjugated anti-rabbit antibody for 1 h, and antigen-antibody complexes were detected using the Odyssey imaging system (LI-COR Biosciences). The lower portion of the gel containing GST and GST-ER␣ was stained with Coomassie Blue.
For GST-IQGAP1 pulldown, recombinant human ER␣ (1 g) was incubated with 1 g of GST or 1 g of GST-IQGAP1 on glutathione-Sepharose beads in Buffer B for 3 h at 4°C. After washing the beads five times with Buffer A, samples were resolved by SDS-PAGE, and the gel was cut at both the 37-and 100-kDa regions. The top and bottom portions of the gel containing GST-IQGAP1 and GST, respectively, were stained with Coomassie Blue. The middle portion of the gel was transferred to PVDF and processed for Western blotting with anti-ER␣ monoclonal antibody and IRDye-conjugated anti-mouse antibody. GST-IQGAP1 pulldown of recombinant human ER␤ was performed essentially as described for ER␣ except that the blots were probed with anti-ER␤ antibody.
Cell Culture and Transfection-HEK293 and MCF-7 cells were maintained in DMEM, and T47D cells were maintained in Roswell Park Memorial Institute (RPMI) 1640 medium. Both media were supplemented with 10% FBS and 1% penicillin/ streptomycin. HEK293 cells were transfected using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Transient overexpression of IQGAP1 increased the amount of the protein in cell lysates by 1.95 Ϯ 0.23-fold (mean Ϯ S.E., n ϭ 5).
GST Pulldown Assays-MCF-7 and T47D cells were plated in 10-cm dishes. When they reached 90% confluence, cells were washed twice with ice-cold PBS (155.6 mM NaCl, 1 mM KH 2 PO 4 , and 2.9 mM Na 2 HPO 4 , pH 7.4) and lysed with 500 l of Buffer B. Lysates were processed by sonication for 10 s with a Model 100 Dismembrator (Fischer Scientific), and insoluble material was pelleted by centrifugation for 10 min. The resultant supernatants were precleared by incubating with glutathione-Sepharose beads for 1 h at 4°C. Equal amounts of protein lysate were incubated with 4 g of GST, GST-ER␣, or GST-IQGAP1 for 3 h at 4°C. After centrifugation, samples were washed five times with Buffer A and separated by SDS-PAGE. The gel was cut at the 100-and 37-kDa regions. For GST-IQGAP1 pulldown, the top and bottom portions containing GST-IQGAP1 and GST proteins, respectively, were stained with Coomassie Blue. The middle part of the gel was transferred to PVDF. For GST-ER␣ pulldown, the gel was cut at 100 kDa. The bottom portion of the gel was stained with Coomassie, and the top portion was transferred to PVDF. After transferring to PVDF, samples were processed by Western blotting.
Immunoprecipitation-HEK293 cells were plated in 10-cm dishes to obtain 80% confluence. After 24 h, each plate was transfected with 5 g of myc-tagged IQGAP1 and either 5 g of myc-tagged ER␣ or FLAG-tagged ER␤ plasmids. The following day, cells were washed with ice-cold PBS and lysed with 500 l of Buffer B. Lysates were subjected to sonication for 10 s, and insoluble material was precipitated by centrifugation at 20,000 ϫ g for 10 min at 4°C. Supernatants were precleared with glutathione-Sepharose beads for 1 h. Equal amounts of protein lysate were incubated with protein A-Sepharose beads and either rabbit IgG, anti-IQGAP1 polyclonal antibodies, anti-ER␣ polyclonal antibodies, or anti-FLAG affinity gel for 3 h at 4°C. Samples were washed five times with Buffer A, resolved by SDS-PAGE, and processed by Western blotting.
Paraformaldehyde cross-linking was performed using a modification of a procedure described previously (26). MCF-7 and T47D cells were grown to 90% confluence and incubated in 1 ml of 0.2% paraformaldehyde in PBS for 10 or 20 min at 22°C. The cross-linking reaction was terminated by adding 1.25 M glycine to a final concentration of 125 mM. After 5 min, cells were washed with PBS and lysed with Buffer B. Cell lysates were processed by immunoprecipitation with rabbit IgG or anti-ER␣ polyclonal antibodies using the protocol described above. E 2 Stimulation-HEK293 cells were cultured in phenol redfree medium with 10% charcoal-stripped FBS. After 24 h, cells were transfected with both IQGAP1 and ER␣ plasmids as described above. The following day, the medium was replaced with fresh medium containing 100 nM E 2 or vehicle (ethanol). After 15 min, 30 min, 60 min, 120 min or 360 min, E 2 stimulation was stopped by washing cells with PBS and lysing them with Buffer B. Immunoprecipitation with anti-IQGAP1 polyclonal antibodies and Western blotting were performed as described above.
Quantitative RT-PCR-To measure pS2 hnRNA, MCF-7 cells were incubated in phenol red-free medium for 24 h. Then vehicle (EtOH) or E 2 (to obtain a final concentration of 100 nM) was added to the medium. After 6 h, cells were harvested, and total RNA was isolated from the cells using TRIzol (Invitrogen). 1 g of RNA was reverse transcribed to cDNA using a High Capacity cDNA Reverse Transcriptase kit (Applied Biosystems) according to the manufacturer's instructions. RT-PCR was performed on a StepOnePlus Real Time PCR system (Applied Biosystems) using SYBR Green PCR Master Mix (Applied Biosystems) and 200 nM forward and reverse primers. The primers used were: pS2 hnRNA, forward primer, 5-TTG-GAGAAGGAAGCTGGATGG-3 (start position 3997, within the intron); reverse primer, 5-ACCACAATTCTGTCTTTCA-CGG-3 (start position 4126, within the second exon); and ␤-actin, forward primer, 5-TGCGTGACATTAAGGAGAAG-3; reverse primer, 5-GCTCGTAGCTCTTCTCCA-3. RT-PCR enzyme activation was initiated for 10 min at 95°C and then amplified by 40 cycles (15 s at 95°C and 1 min at 60°C). All samples were assayed in triplicate, and ␤-actin was used as an internal control. Results were analyzed using the ⌬⌬CT method with StepOnePlus software (Applied Biosystems).
Measurement of PR and cyclin D1 hnRNA was performed as described for pS2 with the following modifications. MCF-7 cells were incubated in phenol red-free medium for 72 h. Vehicle (EtOH) or E 2 (final concentration of 10 nM) was added to the medium for 4 h after which cells were harvested and processed as described in the preceding paragraph. The primers used were: cyclin D1, forward primer, 5-GGATGCTGGAGGTCT-GCGA-3; reverse primer, 5-AGAGGCCACGAACATGCAAG-3; PR, forward primer, 5-CCTCGGACACCTTGCCTGAA-3; reverse primer, 5-CGCCAACAGAGTGTCCAAGAC-3.
Cell Proliferation Assays-Cell proliferation was evaluated by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide dye method performed essentially as previously described (27). Briefly, MCF-siIQ8, MCF-siIQ15, and MCF-7 cells expressing control siRNA were plated in 96-well, flat bottom tissue culture plates. After incubating for 24 h at 37°C in 5% CO 2 , the medium was changed to phenol red-free DMEM. E 2 (final concentration of 50 nM) or an equal volume of EtOH (vehicle) was added, and cells were incubated for another 72 h. Then 15 l of Dye Solution (Promega 2015-09-14) was added, and the cells were incubated for 4 h. 100 l of Solubilization/Stop Solution (Promega 2016-04-21) was added to each well, and the absorbances were determined at 570 nm using a SYNERGY4 multilabel counter (BioTek). The reference wavelength was 700 nm. All of the assays were performed in triplicate.
Miscellaneous Methods-Statistical analysis was performed by Student's t test with Prism 6 (GraphPad). Western blot images were quantified with Image Studio 2.0 (LI-COR Biosciences) according to the manufacturer's instructions. Protein concentrations were determined with the DC Protein Assay (Bio-Rad).

IQGAP1 and ER␣ Bind Directly in Vitro-Analysis was per-
formed in vitro with pure proteins to ascertain whether IQGAP1 and ER␣ interact. Pure ER␣ was incubated with a GST fusion protein of full-length IQGAP1, and complexes were isolated with glutathione-Sepharose beads. Western blotting showed that ER␣ bound to IQGAP1 (Fig. 1A). By contrast, no ER␣ binding was detected to GST alone, validating the specificity of the association with IQGAP1. The interaction was confirmed by pulldown with GST-ER␣. IQGAP1 was cleaved from GST, and the purified protein was incubated with a GST fusion protein of full-length ER␣. Complexes were isolated, and Western blotting was performed. IQGAP1 bound GST-ER␣ (Fig.  1B). No IQGAP1 was detected with GST alone. Coomassie staining showed the expression of GST-tagged IQGAP1 and ER␣ proteins (Fig. 1).
ER␣ and IQGAP1 Interact in Cells-To determine whether IQGAP1 and ER␣ interact in a normal cell milieu, we used cultured human breast epithelial cell lines that express ER␣. MCF-7 cells were lysed and incubated with GST-IQGAP1. Endogenous ER␣ bound to GST-IQGAP1 ( Fig. 2A, left panel). The specificity of binding was confirmed by the absence of ER␣ from the samples incubated with GST alone. Essentially identical results were obtained with a second ER␣-containing human breast epithelial cell line, T47D ( Fig. 2A, right panel).
A reciprocal analysis was conducted with GST-ER␣ to investigate whether endogenous IQGAP1 binds to ER␣. MCF-7 and T47D cell lysates were incubated with GST-ER␣. Western blotting documented that IQGAP1 in both cell lines bound specifically to GST-ER␣ (Fig. 2B). Coomassie staining confirmed the expression of the GST-tagged proteins (Fig. 2).
IQGAP1 and ER␣ Co-immunoprecipitate-The interaction between IQGAP1 and ER␣ in intact cells was evaluated by immunoprecipitation. HEK293 cells were transiently co-trans-fected with ER␣ and IQGAP1. Immunoprecipitation with anti-ER␣ antibodies revealed that IQGAP1 binds to ER␣ in cells (Fig.  3A). No IQGAP1 was detected in the samples precipitated with rabbit IgG, validating the specificity of the interaction. Western blotting confirmed that HEK293 cells did not express ER␣, and neither ER␣ nor IQGAP1 was detected in anti-ER␣ immuno-   Complexes were isolated and washed as described under "Experimental Procedures." The samples were separated by SDS-PAGE, and the gel was cut at 100 kDa region. The lower portion of the gel was transferred to PVDF, and the blot was probed with anti-ER␣ monoclonal antibody (WB; upper panels). The upper part of the gel was stained with Coomassie Blue (lower panels). Aliquots of lysates not subjected to GST pulldown were processed in parallel (Lysate). B, 4 g of GST-ER␣ (ER␣) or 4 g of GST alone bound to glutathione-Sepharose was incubated with lysates from MCF-7 (left panel) or T47D (right panel) breast epithelial cells. Complexes were isolated and washed as described under "Experimental Procedures." The samples were separated by SDS-PAGE, and the gel was cut at the 100-kDa region. The top part of the gel was transferred to PVDF membrane, and the blot was probed with anti-IQGAP1 polyclonal antibodies (WB; upper panels). The bottom part of the gel was stained with Coomassie Blue (lower panels). Aliquots of lysates not subjected to GST pulldown were processed in parallel (Lysate). All data are representative of at least three independent experiments. WB, Western blot.
The transfection was repeated, and lysates were immunoprecipitated with anti-IQGAP1 antibody. ER␣ co-immunoprecipitated with IQGAP1 (Fig. 3B). No ER␣ was detected in samples precipitated with IgG or from the cells immunoprecipitated with anti-IQGAP1 antibody but not transfected with ER␣. These data confirm the specificity of the interaction. We used both MCF-7 and T47D cells to examine whether endogenous IQGAP1 binds to endogenous ER␣. Cells were exposed to a cross-linker, paraformaldehyde, before lysis. Formaldehyde, which has short spacer arms, is widely used to stabilize protein-protein interactions. Immunoprecipitation with anti-ER␣ antibody revealed that IQGAP1 co-immunoprecipitated with ER␣ from MCF-7 (Fig. 3C, left panel) and T47D cells (Fig. 3C, right panel). Very little protein was seen in sam-ples precipitated with IgG, indicating that endogenous ER␣ binds to endogenous IQGAP1 in both MCF-7 and T47D cells.
ER␣ Binds to the IQ Region of IQGAP1-To determine the region of IQGAP1 to which ER␣ binds, selected portions of IQGAP1 (Fig. 4A) were labeled with [ 35 S]methionine in a reticulocyte lysate. The constructs were incubated with GST-ER␣ and isolated with glutathione-Sepharose beads. After washing, samples were resolved by SDS-PAGE, transferred to PVDF, and processed by autoradiography. As shown with both pure proteins and cell lysates (Figs. 1B and 2B), full-length IQGAP1 expressed with the TNT system bound to GST-ER␣ (Fig. 4B, top panel). Analysis with smaller fragments revealed that the N-terminal half of IQGAP1 interacted with GST-ER␣,  . Each IQGAP1 construct was also incubated with GST alone (middle panel). Complexes were isolated with glutathione-Sepharose beads and washed. Samples were resolved by SDS-PAGE and transferred to PVDF, and blots were processed by autoradiography. In addition, an aliquot of each [ 35 S]methionine-labeled TNT product (equivalent to 5% of the amount that was used in the pulldown) was resolved by SDS-PAGE, transferred to PVDF, and processed by autoradiography (Input; bottom panel). The data are representative of three independent experiments.

IQGAP1 Binds Estrogen Receptors
but the C-terminal half of IQGAP1 did not bind. Specificity of binding to GST-ER␣ was confirmed by the absence of bands from the samples incubated with GST alone (Fig. 4B, middle  panel). The level of expression of each TNT product was approximately equivalent (Fig. 4B, bottom panel).
Several proteins associate with the IQ region of IQGAP1 (13,14). To ascertain whether ER␣ binds to the IQ region, a fragment comprising amino acids 717-916 of IQGAP1 was labeled with [ 35 S]methionine and incubated with GST-ER␣. Autoradiography demonstrated that the IQ region of IQGAP1 bound to ER␣ (Fig. 4B, top panel). Consistent with these findings, deletion of the IQ domain (the construct is termed IQGAP1⌬IQ) abrogated the binding of IQGAP1 to ER␣ (Fig. 4B, top panel). These data reveal that the IQ region of IQGAP1 is necessary and sufficient for ER␣ binding.
IQGAP1 Binds to the Hinge Region of ER␣-The TNT system was also used to identify the region of ER␣ with which IQGAP1 interacts. [ 35 S]Methionine-labeled full-length ER␣ and a series of ER␣ fragments and deletion mutants (Fig. 5A) were incubated with GST-IQGAP1. The complexes were isolated with glutathione-Sepharose beads and processed by SDS-PAGE and autoradiography. As observed with pure proteins and cell lysates ( Figs. 1 and 2), full-length ER␣ expressed by the TNT system bound to GST-IQGAP1 (Fig. 5B, top panel). Although the N-terminal portion of ER␣ bound to IQGAP1, the C-terminal portion (ER␣(300 -595)) did not bind. Extending the C-terminal portion proximally by adding the hinge region and DBD (ER␣(180 -595)) restored binding equivalent to that of fulllength ER␣ (Fig. 5B, top panel). Similarly, ER␣(180 -353), which comprises the DBD, hinge, and the proximal portion of the C-terminal of ER␣, bound IQGAP1 to essentially the same extent as full-length ER␣. Input data reveal that the amounts of each ER␣ fragment were essentially the same (Fig. 5B, bottom  panel).
To further narrow the binding region, we deleted selected portions of ER␣ from the DBD and hinge regions. Analysis of the deletion constructs revealed that removing amino acids 185-240 did not impair binding to IQGAP1 (Fig. 5C). By contrast, ER␣ constructs that lacked portions of the hinge region (namely ER␣⌬240 -270 and ER␣⌬270 -311) exhibited weak binding, and removal of amino acids 240 -311 abrogated binding. Collectively, these data strongly suggest that IQGAP1 binds to ER␣ between amino acids 240 and 311, and the whole hinge region of ER␣ is needed for effective binding. E 2 Attenuates the Binding between IQGAP1 and ER␣-The possible effect of E 2 on the interaction between IQGAP1 and ER␣ was investigated. HEK293 cells transiently transfected with IQGAP1 and ER␣ were stimulated with E 2 for different time intervals, and the association between the two proteins was examined by co-immunoprecipitation. Consistent with prior reports (10), we observed a decrease in the total ER␣ in lysates of cells treated with E 2 that became statistically significant at 360 min (Fig. 6, A and B). As anticipated, E 2 did not significantly change the amount of IQGAP1 in the lysates or in the immunoprecipitates. By contrast, the amount of ER␣ that co-immunoprecipitated with IQGAP1 decreased in a time-dependent manner (Fig. 6, A and C). To correct for the reduction of total ER␣ produced by E 2 , we quantified both the total amount of ER␣ in the cells and the amount of ER␣ in anti-IQGAP1 immunoprecipitates. After correction for total ER␣ in lysates, the amount of ER␣ that co-immunoprecipitated with IQGAP1 from E 2 -stimulated cells was lower than that from control cells (Fig. 6D). Reduction was seen at the earliest time point (15 min) and became statistically significant at 60 min, and at 360 min of E 2 treatment, the amount of ER␣ that coimmunoprecipitated with IQGAP1 was reduced by 58 Ϯ 6.5% (mean Ϯ S.E., n ϭ 3). These data reveal that E 2 modulates the interaction between IQGAP1 and ER␣ in cells.
Knockdown of IQGAP1 Impairs ER␣ Function-We evaluated the possible role of IQGAP1 in ER␣ function. ER␣-positive MCF-7 cell lines were chosen because they express high levels of IQGAP1 (25,27). We used two different MCF-7 cell lines with stable IQGAP1 knockdown by siRNA. These cells, termed MCF-siIQ8 and MCF-siIQ15, had IQGAP1 protein levels 65 Ϯ 8 and 66 Ϯ 4% (mean Ϯ S.E., n ϭ 4), respectively, lower than control cells (Fig. 7A, left panel, and B). E 2 -induced transcriptional activation of pS2 was measured. E 2 produced a 7.2 Ϯ 0.6-fold (mean Ϯ S.E., n ϭ 3) increase in pS2 hnRNA in MCF-7 cells stably expressing control siRNA (Fig. 7C). The ability of E 2 to stimulate pS2 hnRNA was significantly attenuated when the amount of IQGAP1 was reduced. E 2 stimulated pS2 hnRNA expression in MCF-siIQ8 and MCF-siIQ15 cells by only 2.7 Ϯ 0.2-and 3.8 Ϯ 0.5-fold (mean Ϯ S.E.), respectively (Fig. 7C). Note that knockdown of IQGAP1 did not significantly reduce the amount of ER␣ in the MCF-7 cell lines (Fig. 7A). Moreover, the level of the ER␣ in cells incubated with E 2 was not significantly altered by IQGAP1 knockdown (Fig. 7A, right panel). E 2 stimulation reduced the amount of ER␣ as observed in Fig. 6A and described previously (10). Collectively, these data validate that IQGAP1 is required for maximal E 2 -stimulated transcriptional activation of pS2 in MCF-7 cells.
To extend these findings, we examined the effects of IQGAP1 on the ability of ER␣ to activate two other endogenous estrogen-responsive genes, namely PR and cyclin D1. Analysis revealed that E 2 stimulation of PR hnRNA was significantly impaired in both MCF-siIQ8 and MCF-siIQ15 cells (Fig. 7D). Similarly, reducing the amount of IQGAP1 in MCF-7 human breast epithelial cells decreased stimulation of cyclin D1 hnRNA by E 2 (Fig. 7E). These results support our observations with pS2 and confirm that endogenous IQGAP1 contributes to normal transcriptional function of ER␣.
In addition to promoting transcription of selected genes, ER␣ mediates an increase in cell proliferation (1). We evaluated the possible role of IQGAP1 in this process. The ability of E 2 to enhance proliferation of MCF-7 control cells was compared with cells with IQGAP1 knockdown, namely MCF-siIQ8 and MCF-siIQ15 cells. Reducing IQGAP1 levels in MCF-7 cells abrogated the stimulation of cell proliferation by E 2 (Fig. 7F).
ER␤ Binds to IQGAP1-We investigated the possible interaction between IQGAP1 and ER␤, which is another member of the ER family with a domain structure similar to that of ER␣ (Fig. 8A). Pure ER␤ was incubated with a GST fusion protein of full-length IQGAP1, and complexes were isolated with glutathione-Sepharose beads. Western blotting showed that ER␤ bound to IQGAP1 (Fig. 8B). By contrast, no ER␤ binding was detected to GST alone, validating the specificity of the association with IQGAP1. Coomassie staining showed GST-IQGAP1 and GST (Fig. 8B). ER␤, synthesized with TNT and labeled with [ 35 S]methionine, bound specifically to GST-IQGAP1 (Fig. 8C). Thus, IQGAP1 and ER␤ interact directly.
Immunoprecipitation was performed to ascertain whether IQGAP1 and ER␤ associate in cells. HEK293 cells were transiently transfected with IQGAP1 and FLAG-tagged ER␤. Immunoprecipitation with anti-FLAG antibodies revealed that IQGAP1 bound to ER␤ in cells (Fig. 9A). Analogous experi-ments were performed in which cells were transfected and lysates were immunoprecipitated with anti-IQGAP1 antibodies. ER␤ co-immunoprecipitated with IQGAP1 (Fig. 9B). Neither IQGAP1 nor ER␤ was seen in samples precipitated with IgG nor was ER␤ seen in samples obtained from cells transfected with empty vector (Fig. 9, A and B). The blots were also probed with anti-ER␤ antibodies, which confirmed that the bands detected with anti-FLAG antibodies are ER␤ (data not shown). Collectively, these data reveal that ER␤ and IQGAP1 interact in cells.

DISCUSSION
IQGAP1 interacts with numerous proteins, thereby coordinating the assembly of multiprotein complexes and modulating diverse signaling pathways (13). For example, IQGAP1 binds to several components of the epidermal growth factor receptor-Raf-MAPK cascade, regulating activation of ERK (13,17,28). Work from a number of groups has demonstrated that IQGAP1 associates with selected transmembrane receptors and participates in their normal function (13,14). Here we show for the first time that IQGAP1 also interacts directly with nuclear receptors. We demonstrate that IQGAP1 binds to both ER␣ and ER␤ and influences ER␣ transcriptional function.
Based on the presence of three LXXLL motifs located at amino acid residues 347-351, 975-979, and 1646 -1650 in IQGAP1, we predicted that ER␣ would bind one or more of these regions. The data did not validate this hypothesis. ER␣ did not associate with the C-terminal half of IQGAP1, which contains two of the LXXLL regions. Although ER␣ bound the N-terminal half of IQGAP1, which contains the other LXXLL motif, amino acids 717-916 of IQGAP1 were sufficient for binding ER␣, indicating that the LXXLL motif at amino acids 347-351 does not mediate binding. Moreover, the observation that deletion of amino acids 763-864 from IQGAP1 eliminated its association with ER␣ confirms that the IQ region of IQGAP1

IQGAP1 Binds Estrogen Receptors
is the site to which ER␣ binds. Several other proteins, including calmodulin (29,30), myosin light chain (31), Rap1 (32), human epidermal growth factor receptor-2 (16), epidermal growth factor receptor (17), and S100B (33), also bind to the IQ region of IQGAP1. Protein binding to IQGAP1 may be regulated. For example, Ca 2ϩ modulates the interaction of calmodulin with the IQ region of IQGAP1 (21,29). Similarly, Zn 2ϩ promotes the association of S100B with the IQ motif of IQGAP1 (33). In addition, although there are no published data, it is likely that the different proteins bind the IQ domains of IQGAP1 with different affinities. Therefore, a cell may have several distinct IQGAP1 complexes with different proteins associated with the IQ region depending on signaling inputs, subcellular location, cell type, and other factors. An important finding in our study is that E 2 regulated the association between IQGAP1 and ER␣. The effects of ligand binding on the interaction of receptors with IQGAP1 are diverse. For example, EGF has no effect on the association between IQGAP1 and epidermal growth factor receptor (17). By contrast, stimulation of Madin-Darby bovine kidney cells with FGF2 promotes binding of fibroblast growth factor recep- FIGURE 7. Knockdown of IQGAP1 alters ER␣ function. A, MCF-7 cells were stably transfected with siRNA against Renilla luciferase (Control) or two distinct siRNAs directed against different regions of IQGAP1 (siIQ8 and siIQ15). Where indicated, cells were cultured in phenol red-free medium for 24 h followed by incubation with vehicle (EtOH) or 100 nM E 2 for 6 h (right panel). Cells were lysed, and equal amounts of protein lysate were analyzed by Western blotting. B, the amount of IQGAP1 in the Western blot was quantified with Image Studio 2.0 (LI-COR Biosciences) and corrected for the amount of actin in the corresponding sample. Data represent the means Ϯ S.E. (error bars) of three independent experiments with control cells set as 1. ***, p Ͻ 0.001. C, MCF-7 cells stably expressing control siRNA, siIQ8, or siIQ15 were incubated in phenol red-free medium for 24 h followed by 6-h incubation with vehicle (EtOH; black bars) or 100 nM E 2 (white bars). Total RNA was extracted, and quantitative RT-PCR analysis was performed to measure pS2 hnRNA. The amount of hnRNA in each sample was corrected for ␤-actin hnRNA in the same sample. Vehicle-treated cells were set as 1. The data represent the means Ϯ S.E. (error bars) of three independent experiments, each performed in triplicate. *, p Ͻ 0.05; **, p Ͻ 0.01 compared with the control cell line treated with E 2 . D, MCF-7 cells expressing control siRNA, siIQ8, or siIQ15 were incubated in phenol red-free medium with vehicle (EtOH; black bars) or 10 nM E 2 (white bars) for 4 h. Quantitative RT-PCR analysis was performed to measure PR hnRNA, and samples were analyzed as described for pS2. The data represent the means Ϯ S.E. (error bars) of three independent experiments for control and siIQ8 cells and two independent experiments for siIQ15 cells. *, p Ͻ 0.05 compared with the control cell line treated with E 2 . E, control and sIQ8 cell lines were incubated in phenol red-free medium with vehicle (EtOH; black bars) or 10 nM E 2 (white bars) for 4 h, and cyclin D1 hnRNA was measured by quantitative RT-PCR. Samples were analyzed as described for pS2. A representative experiment of three independent determinations is shown. F, cell proliferation was measured with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay as described under "Experimental Procedures." Data are representative of three independent experimental determinations, each performed in triplicate. Complexes were isolated and washed as described under "Experimental Procedures." The samples were separated by SDS-PAGE, and the gel was cut at both the 100-and 37-kDa regions. The portion of the gel between 37 and 100 kDa was transferred to PVDF, and the blot was probed with anti-ER␤ monoclonal antibody (WB; top panel). The top and bottom parts of the gel containing GST-IQGAP1 and GST, respectively, were stained with Coomassie Blue. Input is pure ER␤. C, [ 35 S]methionine-labeled ER␤ was incubated with 4 g of GST-IQGAP1 or GST alone. Complexes were isolated with glutathione-Sepharose beads and washed. Samples were resolved by SDS-PAGE and transferred to PVDF, and blots were processed by autoradiography. In addition, the [ 35 S]methionine-labeled product (equivalent to 5% of the amount that was used for the pulldown) was processed as described above (Input). Data are representative of at least two independent experiments. WB, Western blot.  tor 1 to IQGAP1 (18). Similarly, VEGF enhances the interaction of vascular endothelial growth factor receptor-2 with IQGAP1 in human umbilical vein endothelial cells (19), and transforming growth factor ␤-1 (TGF-␤1) increases the binding of IQGAP1 to TGF-␤ receptor II (34). In contrast to these data, we observed that E 2 significantly attenuated the interaction between IQGAP1 and ER␣ in cells. To our knowledge, this is the first description of a ligand that reduces receptor binding to IQGAP1.

IQGAP1
Prior publications have identified the participation of IQGAP1 in transcriptional activation. ␤-Catenin is an integral component of the cadherin cell adhesion complex, and it co-activates transcription factors in the nucleus (35). IQGAP1 binds ␤-catenin and regulates its function (36,37). Overexpression of IQGAP1 significantly enhances ␤-catenin-mediated transcriptional co-activation in SW480 colon carcinoma cells (37). Although IQGAP1 was not detected in the nucleus, IQGAP1 overexpression increased the amount of ␤-catenin in the nucleus. Similarly, IQGAP1 regulates the nuclear translocation of nuclear factor of activated T cells (NFAT) and alters transcription of NFAT target genes (38). More recent studies have revealed that IQGAP1 regulates nuclear translocation and transcriptional activation of both nuclear factor-erythroid-related factor 2 (Nrf2) (39) and Dishevelled (40). Analogous to our observations with ER␣ transcriptional activation, knockdown of IQGAP1 impairs the transcriptional activation of target genes by Nrf2 and Dishevelled (39,40). By contrast, CD8 ϩ T cells isolated from IQGAP1-deficient mice exhibit increased expression of interferon-␥, an established target gene of NFAT (38). Thus, IQGAP1 exerts different effects on different gene transcription pathways.
For ER␣, IQGAP1 is necessary for E 2 to maximally stimulate transcription of the pS2, PR, and cyclin D1 genes. Knockdown of IQGAP1 by two distinct siRNAs directed against different regions of IQGAP1 significantly reduced the ability of E 2 to stimulate transcription of pS2 and PR. This finding initially appears to contradict the observation that E 2 attenuated the interaction between IQGAP1 and ER␣ in cells. The knockdown experiments imply that IQGAP1 is required for ER␣ to enhance pS2 transcription, whereas the co-immunoprecipitation experiments revealed that E 2 reduced the interaction of ER␣ with IQGAP1 at the same time point (6 h) at which pS2 transcription was analyzed. Taken together, these data suggest that a direct interaction with IQGAP1 in the nucleus is not required for ER␣ to promote pS2 gene transcription. Although IQGAP1 has been reported in the nucleus, the pool is small (Ϸ5% of total IQGAP1) (41). Moreover, IQGAP1 accumulates in the nucleus at the G 1 /S phase of the cell cycle and appears to contribute to cell cycle progression. Although not impossible, it is unlikely that nuclear IQGAP1 makes a substantial contribution to ER␣ function. There are several possible ways in which IQGAP1 may modulate ER␣ function. 1) The binding may alter ER␣ degradation by the ubiquitin-proteasome system. The transcriptional activity of ER␣ is linked to receptor turnover. Inhibition of ER␣ degradation leads to its stabilization but also to loss of its transcriptional activity (42). However, there is no evidence that IQGAP1 alters ER␣ degradation as we observed that reducing the intracellular amount of IQGAP1 in MCF-7 cells did not significantly change the amount of ER␣. Therefore, changing ER␣ degradation is probably not the mechanism by which IQGAP1 influences ER␣ transcription. 2) Numerous cofactors associate with ER␣ in interconnected networks of proteins (9). IQGAP1 is a scaffold that integrates signaling pathways and may coordinate the assembly of a multiprotein complex to regulate ER␣ function. 3) IQGAP1 may alter the ability of ER␣ to bind other proteins. IQGAP1 binds to the hinge (D) region of ER␣. Initially believed to serve exclusively as a flexible linker between the DBD and ligand-binding domain, evidence accumulated over the last decade has identified important functional roles for the hinge region of ER␣. For example, a recent study reveals that selective mutations in the D domain of ER␣ alter its nuclear translocation and its ability to activate its target genes activator protein-1, pS2, and specific protein-1 (43). One of the mutant ER␣ constructs generated in that study can translocate to the nucleus and bind to DNA in the presence of E 2 but is unable to interact with the necessary tethering factors to activate activator protein-1 and specific protein-1 response elements. The hinge region is implicated in interconnections with some co-regulator molecules (42). It is possible that IQGAP1 sterically hinders the interaction of one (or more) co-regulators, thereby changing pS2, PR, and cyclin D1 transcription. 4) In addition, several post-translational modifications, including phosphorylation, acetylation, methylation, and ubiquitination, occur in the hinge region of ER␣ (44). These covalent additions, which affect receptor activity, could be modified by IQGAP1 binding. 5) A nuclear localization signal is contained in the hinge region, and IQGAP1 could conceivably regulate the nuclear translocation of ER␣ as it does for ␤-catenin, NFAT, and Nrf2 (37-39). 6) Finally, IQGAP1 is known to dimerize (22). Although the dimerization domain of ER␣ is outside the hinge region, IQGAP1 could enhance ER␣ dimerization, analogous to that produced by calmodulin. Detailed structural analysis reveals that binding of calmodulin to the hinge region of ER␣ promotes dimerization of ER␣ (45). These possible mechanisms are not mutually exclusive, and more than one may be responsible for the effects we observed with IQGAP1. Additional work is required to dissect these possibilities and elucidate the molecular mechanism(s) by which IQGAP1 regulates ER␣ transcriptional activation.
Association of IQGAP1 with members of the steroid receptor family is not confined to ER␣. We observed a direct in vitro interaction of pure ER␤ with pure IQGAP1. Importantly, the two proteins co-immunoprecipitated from cell lysates, revealing that they interact in cells. Further studies are needed to determine whether the interaction with IQGAP1 alters ER␤ function. Moreover, several other nuclear receptors, including the androgen receptor, glucocorticoid receptor, progesterone receptor, thyroid receptor, and vitamin D receptor, have domain structures similar to that of ER (46). We are examining the possible interactions of these receptors with IQGAP1, and if binding is detected, we plan to investigate whether IQGAP1 has any effects on receptor function.
Our data identify a previously unrecognized interaction between the nuclear receptor ER␣ and IQGAP1. Experimental evidence supports the notion that IQGAP1 modulates ER␣ transcriptional activation of pS2, PR, and cyclin D1. These observations enhance our comprehension of both ER␣ signaling and IQGAP1 function. Importantly, a large body of data implicates both ER␣ (47) and IQGAP1 (27,48) in breast carcinoma. It seems feasible that therapeutic agents directed toward the interaction between ER␣ and IQGAP1 might be useful in the management of patients with breast carcinoma.