Caveolin-1 Potentiates Estrogen Receptor α (ERα) Signaling

Estrogen receptor α (ERα) is a soluble protein that mediates the effects of the gonadal estrogens such as 17β-estradiol. Upon ligand binding, a cytoplasmic pool of ERα translocates to the nucleus, where it acts as a transcription factor, driving the expression of genes that contain estrogen-response elements. The activity of ERα is regulated by a number of proteins, including cytosolic chaperones and nuclear cofactors. Here, we show that caveolin-1 potentiates ERα-mediated signal transduction. Coexpression of caveolin-1 and ERα resulted in ligand-independent translocation of ERα to the nucleus as shown by both cell fractionation and immunofluorescence microscopic studies. Similarly, caveolin-1 augmented both ligand-independent and ligand-dependent ERα signaling as measured using a estrogen-response element-based luciferase reporter assay. Caveolin-1-mediated activation of ERα was sensitive to a well known ER antagonist, 4-hydroxytamoxifen. However, much higher concentrations of tamoxifen were required to mediate inhibition in the presence of caveolin-1. Interestingly, caveolin-1 expression also synergized with a constitutively active, ligand-independent ERα mutant, dramatically illustrating the potent stimulatory effect of caveolin-1 in this receptor system. Taken together, our results identify caveolin-1 as a new positive regulator of ERα signal transduction.

Caveolae are flask-shaped vesicular invaginations of the plasma membrane (1). So far, these structures have been implicated in three overlapping areas of cell physiology, i.e. endocytosis (2,3), cholesterol trafficking (4 -9), and signal transduction (reviewed in Ref. 10). To engage in these processes, caveolae have a protein and lipid composition that is distinct from the plasma membrane proper. More specifically, they are enriched in cholesterol, glycosphingolipids, and sphingomyelin as well as lipid-modified signaling proteins (10,11).
The principal coat proteins of caveolae are the caveolins. Thus far, three distinct mammalian caveolin genes have been identified, and their 20 -25-kDa gene products are broadly expressed in a variety of tissues and cell types (12)(13)(14)(15)(16). In addition to interacting with an array of integral membrane, lipid-modified, and soluble signaling molecules, the caveolins share the ability to self-oligomerize, to bind cholesterol, and to cross-link cell-surface gangliosides (10,(17)(18)(19)(20).
In general, caveolins bind to and inactivate signaling molecules. Such examples include, but are not limited to, the following: receptor tyrosine kinases (e.g. epidermal growth factor receptor and c-Neu) and their downstream targets (e.g. Ha-Ras, MEK1, and ERK2), serpentine receptors (e.g. endothelin receptor) and their attendant enzymes (e.g. various G␣ subunits, adenylyl cyclase, and protein kinase A), and regulated enzymes (e.g. endothelial nitric-oxide synthase). All these signaling components are inhibited by their interaction with caveolins (reviewed in Ref. 21).
The interaction of caveolin-1 with many of the proto-oncogene products described above has important consequences for cellular transformation and, perhaps, cancer. Several experimental lines of evidence support this hypothesis. First, caveolin-1 is down-regulated in a variety of oncogenically transformed cells (22). Second, when the caveolin-1 cDNA is reintroduced into Ras G12V -transformed NIH 3T3 cells, anchorage-independent growth is abrogated (23). Third, disruption of caveolae by antisense-mediated down-regulation of caveolin-1 protein expression in normal NIH 3T3 cells results in (i) hyperactivation of the p42/44 mitogen-activated protein kinase cascade, (ii) anchorage-independent growth, and (iii) tumor formation in nude mice (24). Fourth, pharmacological depletion of cellular cholesterol with a concomitant morphologic loss of caveolae also results in p42/44 mitogen-activated protein kinase activation (25). Finally, the caveolin-1 and -2 genes are co-localized to a known tumor suppressor locus in mice and humans (7q31.1/D7S522) (reviewed in Ref. 26).
We recently uncovered a reciprocal relationship between Neu tyrosine kinase activity and caveolin-1 expression in mammary adenocarcinomas (27). An increase in Neu kinase activity correlated with a decrease in caveolin-1 expression both in vitro and in vivo. Conversely, overexpression of caveolin-1 inhibited Neu kinase activity in vivo. As the c-Neu proto-oncogene is mutationally activated in human breast cancers, these results may have implications for understanding the functional role of caveolin expression in the prevention of mammary tumorigenesis.
In this report, we address the possible role of caveolin-1 in estrogen receptor (ER) 1 signal transduction, another major pathway that is thought to be involved in the development of human breast cancers. Here, we show that caveolin-1 re-expression in MCF-7 cells, an estrogen-dependent human breast cancer cell line, promotes nuclear translocation of ER␣. The possible implications of these findings for understanding ER␣ signaling and breast carcinogenesis are discussed.
Cell Culture-MCF-7 cells (ATCC/HTB-22) were obtained from the American Type Cell Collection and were propagated in Dulbecco's modified Eagle's medium, 10% donor bovine serum, 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin. Human embryonic kidney 293T cells were cultured in Dulbecco's modified Eagle's medium, 10% fetal bovine serum, 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin. Cells were seeded into tissue culture dishes containing phenol red-free Dulbecco's modified Eagle's medium supplemented with 10% charcoal/dextran-treated fetal bovine serum and cultured for at least 24 h prior to all experimental treatments.
Cell Fractionation-Thirty-six hours post-transfection, cells from one 60-mm diameter plate were harvested by gentle scraping into phosphate-buffered saline and collected by centrifugation at 1000 ϫ g. Cells were then subjected to hypotonic lysis in 10 mM Tris (pH 7.5) and 20 mM Na 2 MoO 7 , passed through a 26-gauge needle, and then sonicated. Cells were fractionated into cytoplasmic and nuclear fractions as we described in detail previously (24). Samples were brought to the same volume, and equal-volume aliquots from each fraction were separated by SDS-polyacrylamide gel electrophoresis and subjected to immunoblot analysis (35).
Immunoblotting-Proteins were separated by SDS-polyacrylamide gel electrophoresis under reducing conditions and then transferred to nitrocellulose. Protein bands were visualized by staining with Ponceau S. Blots were washed with Tris-buffered saline (10 mM Tris (pH 8.0) and 150 mM NaCl) and 0.05% Tween 20 and then placed in blocking solution (Tris-buffered saline, 0.05% Tween 20, 2% nonfat milk, and 1% bovine serum albumin) for 1 h. Blots were incubated for 1 h with primary antibodies, washed with Tris-buffered saline and 0.05% Tween 20, and incubated with horseradish peroxidase-conjugated secondary antibodies (Transduction Laboratories). Bound IgGs were visualized using an enhanced chemiluminescence detection system (Pierce) according to the manufacturer's protocol.
Immunolocalization Studies-Immunofluorescent labeling was performed as described previously (28). Briefly, cells were fixed in 2% paraformaldehyde and doubly immunostained with mouse anti-caveolin-1 IgG (cl 2234) and rabbit anti-ER␣ IgGs (H-184). Bound primary antibodies were visualized with fluorescein isothiocyanate-conjugated donkey anti-mouse and lissamine rhodamine-conjugated donkey antirabbit secondary antibodies (Jackson ImmunoResearch Laboratories, Inc.). Cells were viewed with an Olympus IX70 inverted microscope using a 60ϫ objective, and images were collected with a Photonics cooled CCD camera.
In Vivo ER␣ Signaling Assay-Twenty-four hours prior to transfec-tion, ϳ1 ϫ 10 5 cells were seeded into 12-well tissue culture plates and cultured in phenol red-free Dulbecco's modified Eagle's medium containing charcoal/dextran-treated fetal bovine serum. Cells were then transfected with 1 g of vector (pCB7) or vector containing caveolin cDNA (Cav-1/pCB7), 1 g of ERE2TK81LUC reporter, and 200 ng of pSV-␤-galactosidase by calcium phosphate precipitation. One microgram of the wild-type human ER␣ cDNA or a constitutively active (Y537S) ER␣ mutant cDNA was also cotransfected. Twelve hours after addition of calcium phosphate precipitates, cells were washed twice with phosphate-buffered saline and incubated for an additional 24 h in medium containing E 2 or an equivalent volume of vehicle (ethanol). In all experiments, the final concentration of ethanol was 0.1% (v/v). In antagonist studies, OHT dissolved in dimethyl sulfoxide was added from stock solutions such that the final concentration of solvent was 0.1% (v/v). Lysates were prepared 24 h after pharmacological treatment and assayed for luciferase and ␤-galactosidase activities. Results represent the mean Ϯ S.D. of luciferase activity normalized to ␤-galactosidase activity (n ϭ 3).
Co-immunoprecipitation Studies-Immunoprecipitation of ER␣ was performed essentially as we described previously for another transcription factor, C/EBP␤ (36). Briefly, cells were subjected to lysis in immunoprecipitation buffer (50 mM Tris (pH 7.5), 150 mM NaCl, 1% (v/v) Nonidet P-40, 0.5% (w/v) deoxycholate, 0.1% (w/v) SDS, and 0.1 mM Na 3 VO 4 supplemented with protease inhibitors). DNA was sheered by brief sonication on ice, and cellular debris was removed by centrifugation at 12,000 ϫ g for 10 min. Lysates were precleared by incubation with protein A-Sepharose for 1 h at 4°C and then transferred to fresh tubes containing 30 l of a 1:1 slurry of protein A-Sepharose and immunoprecipitation buffer. Ten micrograms of anti-ER␣ IgGs were added to the mixture. Following a 6-h incubation rotating at 4°C, immune complexes were collected by centrifugation, washed four times with 1 ml of immunoprecipitation buffer lacking Na 3 VO 4 and protease inhibitors, and disrupted by boiling in 1% SDS.

Caveolin-1 Expression Induces Ligand-independent Nuclear
Translocation of ER␣-Mammary epithelial cells normally express both caveolin-1 and -2, whereas many mammary adenocarcinoma cell lines such as MCF-7 show selective down-regulation of caveolin-1 (reviewed in Ref. 26). However, MCF-7 cells continue to express wild-type ER␣. Thus, we employed MCF-7 cells as a model system to study the effects of recombinant caveolin-1 expression on the behavior of endogenous ER␣. Fig. 1A (upper panel) shows that transient expression of caveolin-1 in MCF-7 cells resulted in a dramatic decrease in the cytoplasmic pool of ER␣. To ensure equal protein loading, the same blots were reprobed with antibodies against both membrane (caveolin-2) and cytosolic (GDI) proteins (Fig. 1A, lower panels).
To examine whether the decrease in the cytoplasmic pool of ER␣ was due to its degradation, total cellular proteins were recovered by lysis in 1% SDS; the amount of ER␣ was determined by immunoblot analysis. Fig. 1B shows that ER␣ levels were identical in caveolin-1 transfectants and mock-transfected control cells both in the presence and absence of estradiol. As expected, estradiol treatment decreased the steadystate levels of ER␣ expression, as ligand binding results in nuclear translocation and subsequent ubiquitination, followed by proteasomal degradation (37). Thus, the decrease in cytoplasmic ER␣ levels following caveolin-1 transfection is most likely due to enhanced nuclear translocation.
Next, we tested whether caveolin-1 expression can influence ER␣ translocation from the cytoplasm to the nucleus using subcellular fractionation techniques. In unstimulated control cells, ER␣ resided in both the nucleus and the cytoplasm; the cytoplasmic pool underwent translocation to the nucleus when stimulated with estradiol ( Fig. 2A, left panel) (38). Interestingly, in unstimulated cells, caveolin-1 caused a shift in the subcellular localization from the cytoplasm to the nucleus ( Fig.  2A, right panel). As a control, we verified that caveolin-1 expression did not alter the location of a known cytosolic protein, GDI (Fig. 2B). Thus, it appears that caveolin-1 expression can cause ligand-independent nuclear translocation of ER␣.
Similar results were obtained by immunofluorescence microscopy. In caveolin-1-transfected cells and in the absence of ligand, the cytoplasmic pool of endogenous ER␣ shifted to the nucleus (Fig. 3A). More important, this effect was cell autonomous, as adjacent untransfected cells did not show enhanced nuclear concentration of ER␣ (Fig. 3, A and C, compare caveolin-1 transfectants to the left with the corresponding untransfected cells to the right in the same field).
As a positive control for these studies, we also evaluated the ability of estradiol to induce nuclear translocation of ER␣ under identical conditions. Note that when cells were stimulated with estradiol, caveolin-1 expression did not influence nuclear translocation of ER␣; as expected, both caveolin-1-transfected and adjacent untransfected cells showed nuclear concentration of ER␣ (Fig. 3, B and D, compare caveolin-1 transfectants to the left with the corresponding untransfected cells to the right in the same field). These data independently confirm our results obtained via cell fractionation (Fig. 2).
Potentiation of ER␣ Signaling by Caveolin-1-Since caveolin-1 causes ligand-independent ER␣ translocation from the cytoplasm to the nucleus, we wondered whether caveolin-1 expression also results in ER␣-mediated transcriptional activation of estrogen-responsive genes. To evaluate this possibility, we employed an established luciferase-based reporter sys-tem that has been used extensively by other investigators to monitor ER␣-mediated signal transduction in vivo. This reporter contains two EREs linked 5Ј to a minimal thymidine kinase promoter that drives luciferase expression (33). Fig. 4 shows that when unstimulated cells were transfected with caveolin-1, an ϳ2-fold increase was observed in ERE reporter activity. In addition, when caveolin-1-transfected cells were stimulated with a range of estradiol concentrations (from 5 to 100 nM), caveolin-1 induced a dramatic increase (up to ϳ7-8-fold) in ER␣ reporter activity. Thus, caveolin-1 expression is sufficient to induce ligand-independent activation of ER␣, and caveolin-1 can potentiate ER␣ signaling in the presence of ligand.
We next evaluated whether caveolin-1-mediated potentiation of ER␣ signaling is sensitive to ER antagonists (Fig. 5). For this purpose, we treated cells with estradiol (10 nM) in the absence or presence of OHT (0.1-100 nM). Note that in cells transfected with vector alone, OHT had an IC 50 of Ͻ0.1 nM. In contrast, in cells transfected with caveolin-1, the IC 50 for OHT was increased Ͼ5-fold to ϳ0.5 nM. These results indicate that caveolin-1 expression can prevent OHT-mediated inhibition of ER␣ signaling in vivo. In addition, these data may have clinical implications for understanding the development of tamoxifen resistance in breast and ovarian cancer cells, as tamoxifen is routinely used in a variety of cancer chemotherapy regimens.
As caveolin-1 expression can potentiate both ligand-dependent and ligand-independent ER␣ signaling, we next determined if caveolin-1 can influence signal transduction mediated by a mutated, constitutively activated form of ER␣. For this purpose, we utilized a well characterized constitutively active mutant (ER␣ Y573S ) that is known to dramatically increase EREdependent transcription in the absence of ligand (39). Fig. 6 shows that caveolin-1 expression augmented ER␣ Y573S activation of estrogen signaling, resulting in an ϳ7-8-fold increase in transcription both in the presence and absence of estradiol. As compared with wild-type ER␣, in the absence of caveolin-1 (Fig.  6, see vector-alone controls (open bars)), ER␣ Y573S plus caveolin-1 resulted in an ϳ150-fold increase in ER␣ ligand-independ-

FIG. 1. Recombinant expression of caveolin-1 decreases the cytosolic level of endogenous ER␣ in MCF-7 cells.
A, cells were transfected with the caveolin-1 cDNA (Cav-1/pCB7) or with vector alone (pCB7). Membrane and cytosolic protein fractions were prepared by lysis in 10 mM Tris (pH 8.0), 150 mM NaCl, 1% Triton X-100, and 60 mM octyl glucoside containing protease inhibitors. As controls for equal protein loading, the blots were also probed with antibodies directed against an endogenous membrane protein (caveolin-2 (Cav-2)) and an endogenous cytosolic protein (GDI). Note that less cytosolic ER␣ was detected in caveolin-1 (Cav-1) transfectants as compared with vectoralone or mock-transfected cells. B, cells were transfected as described for A. Twenty-four hours prior to cell lysis, the cells were treated with 10 nM E 2 or vehicle alone. Total cellular ER␣ was collected by cellular lysis in 1% SDS and detected by immunoblot analysis. Note that caveolin-1 expression did not lower total ER␣ levels, nor did it detectably affect estradiol-mediated degradation of ER␣.

FIG. 2. Recombinant expression of caveolin-1 induces the nuclear translocation of endogenous ER␣.
A, cells were transfected with the indicated constructs and treated with E 2 (10 nM) or with vehicle alone for 30 min prior to cell lysis and fractionation into cytosolic (C) and nuclear (N) fractions. Note that in vehicle-treated control cells (vector-transfected), endogenous ER␣ distributed nearly equally in cytosolic and nuclear fractions. In contrast, following estradiol treatment, the protein was found predominantly in the nuclear fraction. Interestingly, in vehicle-treated caveolin-1 (Cav-1) transfectants, endogenous ER␣ resided predominantly in the nuclear fraction. B, Note that recombinant expression of caveolin-1 did not alter the cytoplasmic localization of GDI, either in the presence or absence of estradiol. ent signaling. These findings directly support our observation that caveolin-1 causes nuclear translocation and activation of wild-type ER␣ (Figs. 2-4).
Interaction of Caveolin-1 and ER␣ in Vivo-One possible mechanism by which caveolin-1 potentiates ER␣ signaling is through a direct or indirect interaction between caveolin-1 and ER␣ itself. Although caveolin-1 is an integral membrane pro-  (pCB7; open bars), an ER␣ expression vector, the ERE2TK81LUC reporter, and a ␤-galactosidase expression vector as described under "Experimental Procedures." Twelve hours post-transfection, cells were washed with phosphate-buffered saline and cultured for 24 h in medium containing the indicated concentrations of E 2 . Lysates were then prepared and assayed for luciferase and ␤-galactosidase activities. To correct for transfection efficiency, luciferase activity (raw light units) was divided by the corresponding ␤-galactosidase activity (absorbance at 574 nm). The resulting ratios were then expressed as fold stimulation relative to vehicle-treated, vector-transfected cells normalized to 1. Note that cotransfection with caveolin-1 stimulated ER␣ signaling activity ϳ2-fold in cells treated with vehicle alone. In addition, when caveolin-1-transfected cells were stimulated with a range of estradiol concentrations (from 5 to 100 nM), caveolin-1 induced a dramatic increase (up to ϳ7-8-fold) in ER␣ reporter activity. Data represent the mean Ϯ S.D. of luciferase activity normalized to ␤-galactosidase activity (n ϭ 3).

FIG. 5. Caveolin-1 expression can prevent tamoxifen-mediated inhibition of ER␣ signaling in vivo.
Cells were transfected and processed as described in the legend to Fig. 4. Twelve hours posttransfection, cells were washed and cultured further in medium containing E 2 (10 nM) and the indicated concentrations of OHT (0.1-100 nM). Twenty-four hours after pharmacological treatment, cells were subjected to lysis and assayed for luciferase and ␤-galactosidase activities. To correct for transfection efficiency, luciferase activity (raw light units) was divided by the corresponding ␤-galactosidase activity (absorbance at 574 nm). The resulting ratios were then expressed as the fold stimulation relative to vector-transfected cells without OHT treatment normalized to 1. Note that in cells transfected with vector alone, OHT had an IC 50 of Ͻ0.1 nM (open bars). In contrast, in cells transfected with caveolin-1, the IC 50 for OHT was increased Ͼ5-fold to ϳ0.5 nM (closed bars). These results indicate that caveolin-1 expression can prevent OHT-mediated inhibition of ER␣ signaling in vivo. Data represent the mean Ϯ S.D. of luciferase activity normalized to ␤-galactosidase activity (n ϭ 3). tein, a soluble cytoplasmic pool of caveolin-1 has been reported (7). This is consistent with the finding that caveolin-1 can move in and out of membranes (existing as a soluble protein) depending on the oxidation state of caveolin-bound membrane cholesterol.
To evaluate the potential interaction of caveolin-1 with ER␣, we cotransfected 293T cells with their corresponding cDNAs. We chose 293T cells for these studies as they do not express either caveolin-1 or ER␣ endogenously. Cell lysates were then prepared and subjected to immunoprecipitation with antibodies directed against ER␣. Fig. 7 shows that when cells were cotransfected with ER␣ and caveolin-1, caveolin-1 co-immunoprecipitated with antibodies directed against ER␣ (third lane). In contrast, when cells were transfected with the caveolin-1 cDNA alone, anti-ER␣ antibodies did not coprecipitate caveolin-1 (Fig. 7, first lane). These results indicate that the observed caveolin-1 co-immunoprecipitation with ER␣ is highly specific, as it was strictly dependent on ER␣ expression.
Possible Functional Significance of the Caveolin-1/ER␣ Interaction-Caveolins are known to interact with a diverse group of signaling molecules. However, it remains unknown whether caveolins influence steroid receptor signaling pathways. Here, we have provided several independent lines of evidence that suggest that caveolin-1 acts as a positive modulator of estrogen receptor signaling in vivo. (i) We found that caveolin-1 directly potentiated estrogen signaling by inducing translocation of ER␣ from the cytoplasm to the nucleus, even in the absence of ligand. (ii) Caveolin-1-driven ER␣ nuclear translocation resulted in increased transcription from an ERE-dependent reporter gene. (iii) Caveolin-1 conferred resistance to the anti-estrogen tamoxifen (with a Ͼ5-fold increase in IC 50 ). (iv) Caveolin-1 augmented the transcriptional activation of a constitutively active form of the estrogen receptor, ER␣ Y537S .
(v) Finally, we observed that caveolin-1 interacted with ER␣ in vivo, as evidence by co-immunoprecipitation studies.
The interaction of signaling molecules with caveolins is mediated largely by the caveolin scaffolding domain, a 20-aminoacyl residue membrane proximal domain (40). Using a phage display-based approach, we have previously defined a consensus caveolin-binding motif, i.e. XXXXX and XXXXXX, where is an aromatic residue and X is any residue (41). However, analysis of the protein sequence of ER␣ failed to identify a putative caveolin-binding motif.
We recently reported that G protein-coupled receptor kinase 2 interacts with caveolin-1 through a motif that diverges slightly from the above-described consensus, 63 LGYLLFRDF 71, where Leu substitutes for an aromatic amino acid (42). Interestingly, G␣ q has similar substitutions for aromatic residues in its caveolin-binding motif, but G␣ q still co-immunoprecipitates caveolin-1 (43). Thus, ER␣ may interact with caveolins through a divergent caveolin-binding motif or may be recognized by other caveolin domains that have been shown to interact with signaling molecules (44,45).
Finally, we consider the significance of the caveolin-1/ER␣ interaction. ER␣-expressing breast cancer cells show enhanced tumorigenicity; however, these ER␣-positive cells have less of a propensity to metastasize (reviewed in Refs. 46 and 47). One possibility is that caveolin-1 expression may help prevent metastasis by potentiating estrogen-mediated transcription in these cells. This view is consistent with the suggestion that caveolin-1 may function as a tumor suppressor gene whose expression is down-regulated during cell transformation (23,24,27,48). Ultimately, gene ablation studies with model animals will be needed to elucidate the exact physiologic role of the caveolins during the development of mammary adenocarcinomas.
FIG. 6. Caveolin-1 also potentiates signaling via a constitutively active form of ER␣. Cells were cotransfected with the wildtype (WT) or constitutively active (CA; Y537S) ER␣ cDNA and with vector alone (pCB7; open bars) or with the caveolin-1 cDNA (Cav-1/ pCB7; closed bars). Twelve hours post-transfection, the cells were washed with phosphate-buffered saline and cultured in vehicle alone (Ϫ) or with E 2 (ϩ; 10 nM) for 24 h. Cells were then subjected to lysis and assayed for luciferase and ␤-galactosidase activities. To correct for transfection efficiency, luciferase activity (raw light units) was divided by the corresponding ␤-galactosidase activity (absorbance at 574 nm). The resulting ratios were then expressed as fold stimulation relative to vehicle-treated, vector-transfected cells normalized to 1. Note that caveolin-1 expression augmented ER␣ Y573S activation of estrogen receptor signaling, resulting in an ϳ7-8-fold increase in transcription both in the presence and absence of estradiol. Values are plotted logarithmically on the ordinate. Data represent the mean Ϯ S.D. of luciferase activity normalized to ␤-galactosidase activity (n ϭ 3) .   FIG. 7. Interaction of caveolin-1 and ER␣ in vivo. To evaluate the potential interaction of caveolin-1 (Cav-1) with ER␣, we cotransfected 293T cells with the caveolin-1 and ER␣ cDNAs. Cell lysates were then prepared and subjected to immunoprecipitation with antibodies directed against ER␣. Proteins were separated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose. Membranes were then probed with anti-ER␣ IgGs (upper panel) and anti-caveolin-1 IgGs (lower panel). Note that when cells were cotransfected with ER␣ and caveolin-1, caveolin-1 co-immunoprecipitated with antibodies directed against ER␣ (third lane). In contrast, when cells were transfected with caveolin-1 alone, anti-ER␣ antibodies did not coprecipitate caveolin-1 (first lane). These results indicate that the observed caveolin-1 coimmunoprecipitation with ER␣ is highly specific, as it was strictly dependent of ER␣ expression.