Reciprocal Regulation of Neu Tyrosine Kinase Activity and Caveolin-1 Protein Expression in Vitro and in Vivo

Neu (c-erbB2) is a proto-oncogene product that encodes an epidermal growth factor-like receptor tyrosine kinase. Amplification of wild-type c-Neuand mutational activation of Neu (Neu T) have been implicated in oncogenic transformation of cultured fibroblasts and mammary tumorigenesis in vivo. Here, we examine the relationship between Neu tyrosine kinase activity and caveolin-1 protein expression in vitro and in vivo. Recent studies have suggested that caveolins may function as negative regulators of signal transduction. Our current results show that mutational activation of c-Neu down-regulates caveolin-1 protein expression, but not caveolin-2, in cultured NIH 3T3 and Rat 1 cells. Conversely, recombinant overexpression of caveolin-1 blocks Neu-mediated signal transduction in vivo. These results suggest a reciprocal relationship between c-Neu tyrosine kinase activity and caveolin-1 protein expression. We next analyzed a variety of caveolin-1 deletion mutants to map this caveolin-1-dependent inhibitory activity to a given region of the caveolin-1 molecule. Results from this mutational analysis show that this functional in vivo inhibitory activity is contained within caveolin-1 residues 32–95. In accordance with thesein vivo studies, a 20-amino acid peptide derived from this region (the caveolin-1 scaffolding domain) was sufficient to inhibit Neu autophosphorylation in an in vitro kinase assay. To further confirm or refute the relevance of our findings in vivo, we next examined the expression levels of caveolin-1 in mammary tumors derived from c-Neu transgenic mice. Our results indicate that dramatic reduction of caveolin-1 expression occurs in mammary tumors derived from c-Neu-expressing transgenic mice and other transgenic mice expressing downstream effectors of Neu-mediated signal transduction, such as Src and Ras. Taken together, our data suggest that a novel form of reciprocal negative regulation exists between c-Neu and caveolin-1.


matic reduction of caveolin-1 expression occurs in mammary tumors derived from c-Neu-expressing transgenic mice and other transgenic mice expressing downstream effectors of Neu-mediated signal transduction, such as Src and Ras. Taken together, our data suggest that a novel form of reciprocal negative regulation exists between c-Neu and caveolin-1.
Development of the normal breast and breast carcinogenesis involves a complex interplay among growth factors, steroids, activation of oncogenes, and inactivation of tumor suppressor genes (1). The Neu (c-erbB2) proto-oncogene encodes a tyrosine kinase receptor that is amplified and overexpressed in a significant proportion of human breast tumors (1,2). Increased amplification of the Neu oncogene directly correlates with poor clinical outcome for breast cancer patients in whom the cancer has not spread to the lymph nodes (3,4).
Neu is a member of a growth factor receptor family that includes the epidermal growth factor (EGF) 1 receptor (EGFR),and Neu is therefore also known as HER2, for human EGF receptor 2. The ligands for the Neu receptor are distinct from the structurally related EGFR (5); however, the EGFR is capable of dimerizing with Neu, and this heterodimer can form when only one member of the pair binds ligand (5). Both Neu and the EGFR stimulate proliferation of breast cancer cells, and overexpression of these two proteins is roughly correlated with progression of human breast cancer (6,7).
Oncogenic activation by Neu can occur through (i) overexpression, (ii) a point mutation within the transmembrane domain, or (iii) small deletions of the membrane-proximal region of the extracellular domain (8). A variety of mutations within the transmembrane domain of the coding sequence of the rat Neu gene (for example, Neu T) are highly oncogenic (8). Overexpression of Neu T in transgenic mice under the control of the MMTV long terminal repeat induces mammary adenocarcinomas with high frequency (9,10), and overexpression of wildtype Neu can also induce breast tumors (11). Several signaling pathways are activated by Neu, including activation of the Ras GTPase-activating protein and the tyrosine kinase c-Src.
Here, we examine the relationship between Neu tyrosine kinase activity and caveolin-1 protein expression using a variety of in vitro, in vivo, and whole animal approaches. Taken together, our results indicate (i) that increases in Neu tyrosine kinase activity down-regulate caveolin-1 protein expression and (ii) that caveolin-1 overexpression can inhibit Neu-mediated signal transduction. These results suggest that a novel form of reciprocal negative regulation exists between c-Neu and caveolin-1. This is the first in vivo demonstration that caveolin-1 expression can functionally suppress signal transduction from a growth factor receptor to the nucleus.
Cell Culture-NIH 3T3, Rat 1, and CHO cells were propagated in T-75 tissue culture flasks in Dulbecco's modified Eagle's medium supplemented with antibiotics and 10% serum.
Caveolin-1 Expression Vectors-Canine caveolin-1 (full-length and deletion mutants) was subcloned into the multiple cloning site (HindIII/ BamHI) of the vector pCB7 (containing the cytomegalovirus promoter and a hygro R marker; gift of J. Casanova, Massachusetts General Hospital) for expression in CHO cells.
Assay for Neu-mediated Signal Transduction-To measure Neu-mediated signal transduction, we employed the PathDetect TM Elk trans-Reporting System. This assay employs a fusion protein that contains the DNA-binding domain of GAL4 and the transactivation domain of Elk to induce expression of a luciferase reporter driven by an artificial promoter containing five GAL4-binding sites. Phosphorylation of the transactivation domain of Elk, in turn, activates transcription of the luciferase gene from the reporter plasmid. Experiments testing a plasmid encoding only the GAL4 DNA-binding domain demonstrated that luciferase expression is specifically dependent on activation of the Elk transactivation domain (data not shown).Transient transfections were performed using calcium phosphate precipitation. Briefly, 300,000 CHO cells were seeded in six-well plates 12-24 h before the transfection. Each point was transfected with 1 g of pSV, pSV-Neu wt, or pSV-Neu T; 1 g of pFR-Luc; 50 ng of pFA-Elk (as described by Stratagene); and 1 g of the indicated caveolin or empty vector. 12 h post-transfection, the cells were rinsed twice with phosphate-buffered saline and incubated in 1% fetal bovine serum for another 24 -36 h. The cells were then in lysed in 200 l of extraction buffer, 50 l of which was used to measure luciferase activity, as described (18). These assays were made possible through the use of a special CHO-derived cell line called GRC ϩ LR-73. Unlike parental CHO cells, GRC ϩ LR-73 cells are a non-transformed growth control revertant that has normal fibroblastic morphology, does not grow in suspension, requires high serum concentrations for growth, and undergoes synchronized growth arrest in low concentrations of serum (1-2%) without a loss of viability (16). Also, these cells have a much higher transfection efficiency (ϳ10-fold) than parental CHO cells.
Immunoblotting-Samples were separated by SDS-polyacrylamide gel electrophoresis and transferred to nitrocellulose. After transfer, nitrocellulose sheets were stained with Ponceau S to visualize protein bands and subjected to immunoblotting. For immunoblotting, incuba-tion conditions were as described by the manufacturer (Amersham Pharmacia Biotech), except we supplemented our blocking solution with both 1% bovine serum albumin and 2% nonfat dry milk (Carnation).
Northern Analysis-Total RNA was extracted and purified according to the manufacturer's instructions (QIAGEN Inc.). 10 g of total cellular RNA was denatured with formaldehyde and subjected to Northern blot analysis with 32 P-labeled probes for the mouse caveolin-1 mRNA (2.4 kilobases) and the human 18 S rRNA, as a control for equal loading.
Neu in Vitro Kinase Assay-A rabbit anti-peptide antibody directed against Neu C-terminal residues 1169 -1186 (C-18; Santa Cruz Biotechnology, Inc.) was used to immunoprecipitate Neu from CHO cells transiently transfected with the c-Neu cDNA. In vitro kinase assays were performed essentially as described previously (19). Briefly, immunoprecipitates were equilibrated with kinase reaction buffer (20 mM Hepes, pH 7.4, 5 mM MgCl 2 , and 1 mM MnCl 2 ), and the reaction was initiated by addition of 15 Ci of [␥-32 P]ATP. After 15 min of incubation at 25°C, the reaction was terminated by addition of 2ϫ SDS-polyacrylamide gel electrophoresis sample buffer and boiling for 2 min. Phosphorylated c-Neu was detected by autoradiography using an intensifying screen. Prior to initiating the reaction, immunoprecipitates were preincubated in kinase reaction buffer with the indicated caveolin-derived peptides for 1 h at 4°C. Peptides were dissolved in Me 2 SO, and 100ϫ stock solutions were prepared. Controls omitting peptide contained an equivalent volume of Me 2 SO, which did not exceed 1%.
Immunohistochemistry-Paraffin-embedded sections were stained as detailed in a protocol provided by Transduction Laboratories, Inc. Briefly, after re-hydration, slides were placed face up in an incubation container, and each section was covered with a solution of 1% SDS in Tris-buffered saline (100 mM Tris, pH 7.4, 138 mM NaCl, and 27 mM KCl). After a blocking step, sections were incubated with anti-caveolin-1 IgG (1:500 dilution in blocking buffer; Santa Cruz Biotechnology, Inc.) for 2 h at 37°C. After rinsing with Tris-buffered saline (3 ϫ 5 min each), sections were incubated with fluorescein isothiocyanate-conjugated goat anti-rabbit IgG (1:200 dilution in blocking buffer; Jackson ImmunoResearch Laboratories, Inc.) for 1 h at 37°C. After rinsing, coverslips were mounted with Slow-Fade anti-fade reagent, and slides were examined by fluorescence microscopy. Note that pretreatment of sections with 1% SDS has recently been shown to be an effective method for antigen retrieval, especially with anti-caveolin IgG (22).

Mutational Activation of c-Neu Down-regulates Caveolin-1, but Not Caveolin-2, in Cultured NIH 3T3 and Rat 1 Cells-A
variety of Neu constructions that were stably expressed in cultured fibroblasts (NIH 3T3 or Rat 1 cells) are shown in Fig.  1. Note that these include wild-type c-Neu, Neu T (a point mutation within the transmembrane domain), Neu T with a C-terminal deletion, and a series of N-terminal domain deletion mutants. All of these mutational changes are known to greatly increase their oncogenic potential and basal tyrosine kinase activity, as compared with wild-type c-Neu (14,15). However, Neu T with a C-terminal deletion lacks the ability to undergo autophosphorylation and does not form a stable complex with Shc, yet it is still more active than wild-type c-Neu.
Lysates were prepared from these cell lines and subjected to immunoblot analysis with caveolin-specific antibodies. Immunoblotting with mAb 2297, a caveolin-1-specific mAb probe, revealed that caveolin-1 protein expression was downregulated in all the mutants of Neu. However, overexpression of wild-type c-Neu did not affect the levels of caveolin-1 in Rat 1 cells and only moderately reduced caveolin-1 levels in NIH 3T3 cells (Fig. 2, A and B, upper panels). In striking contrast, immunoblotting with mAb 65, a caveolin-2-specific mAb probe, revealed that caveolin-2 levels remained relatively constant in all these cell lines (Fig. 2, A and B, lower panels). These results suggest that caveolin-1 is selectively downregulated in response to mutational activation of the c-Neu proto-oncogene.
Neu activates a number of distinct signaling cascades, including the p42/44 MAP kinase pathway. Thus, we tested the effects of a well characterized MEK1/2 inhibitor (PD 98059) on caveolin-1 protein expression in NIH 3T3 cells transformed with Neu T. Fig. 2C shows that treatment with PD 98059 increased caveolin-1 levels by ϳ2-3-fold, but caveolin-1 levels did not reach the wild-type levels observed with normal NIH 3T3 cells. This indicates that constitutive activation of the p42/44 MAP kinase pathway contributes partially to Neu-induced caveolin-1 down-regulation. In contrast to Neu T, treatment of Ras-transformed NIH 3T3 cells with the MEK inhibitor completely restored caveolin-1 to normal levels seen in untransformed NIH 3T3 cells (23).
What is the mechanism by which the caveolin-1 protein is down-regulated in response to Neu-mediated cell transformation? Northern blot analysis of total cellular RNA revealed that the caveolin-1 mRNA was dramatically reduced in the Neutransformed NIH 3T3 cells (Fig. 2D), in parallel with the reduction in the amount of caveolin-1 protein found in this cell line ( Fig. 2A). These studies clearly indicate that reductions in the steady-state level of caveolin-1 mRNA lead to the observed reduction in caveolin-1 protein expression. Similarly, v-Abl and activated Ha-Ras (G12V) have been shown to down-regulate both caveolin-1 mRNA and protein levels in NIH 3T3 cells (24).
As activation of the p42/44 MAP kinase cascade appears to contribute to the down-regulation of caveolin-1 expression Cell lysates were prepared and subjected to immunoblot analysis with caveolin-specific antibodies. Note that caveolin-1 (Cav-1) was downregulated in all the mutants of Neu tested, whereas caveolin-2 (Cav-2) levels remained relatively constant. Interestingly, parental Rat 1 cells expressed significantly lower levels of caveolin-1 than parental NIH 3T3 cells. C, treatment with PD 98059 only partially restored caveolin-1 to normal expression levels in NIH 3T3 cells. Ha-Ras (G12V)-transformed NIH 3T3 cells were used as a positive control for these experiments. In contrast to Neu T, treatment of Ras-transformed NIH 3T3 cells with the MEK inhibitor completely restored caveolin-1 to normal levels seen in untransformed NIH 3T3 cells. D, total cellular RNA (10 g) purified from NIH 3T3 cells was denatured with formaldehyde and subjected to Northern blot analysis with 32 P-labeled probes for caveolin-1 mRNA (2.4 kilobases) and 18 S rRNA, as a control. Note that the caveolin-1 mRNA was down-regulated in the mutants of Neu tested. WT, wild type; DHFR, dihydrofolate reductase. NIH 3T3 cells, whereas caveolin-2 levels remained relatively constant. Similarly, treatment of normal NIH 3T3 cells with an active cell-permeable derivative of ceramide (C 6 -D-erythro) was sufficient to selectively down-regulate caveolin-1 expression, whereas two inactive control derivatives (C 6 -D-erythro-(dihydro) and C 6 -D-threo) did not affect caveolin-1 levels (Fig. 3B). These results indicate that other activating stimuli that feed into the p42/44 MAP kinase cascade can down-regulate caveolin-1 protein expression.
Recombinant Overexpression of Caveolin-1 Blocks Neu-mediated Signal Transduction in Vivo-Why is caveolin-1 downregulated in response to oncogenic Neu? One possibility is that caveolin-1 functions as a negative regulator of Neu-mediated signal transduction. To test this hypothesis of "reciprocal reg-ulation," we employed an assay to measure Neu-mediated signal transduction from the plasma membrane to the nucleus. In this assay, transient coexpression of Neu with a luciferase reporter plasmid allowed us to measure activation of the transcription factor Elk. This luciferase reporter plasmid contains a promoter specifically responsive to Elk phosphorylation and activation (see "Experimental Procedures"). Elk is activated in response to a wide variety of signals, including activation of p42/44 MAP kinase, p38 MAP kinase, and Jun kinase cascades.
As expected, Neu T was ϳ3.5-fold more active than wild-type c-Neu in this assay system (Fig. 4). However, coexpression of caveolin-1 dramatically inhibited Neu-mediated signal transduction evoked by overexpression of either wild-type c-Neu or mutationally activated Neu T by ϳ10 -20-fold (Fig. 4). Importantly, the empty vector used to express caveolin-1 had no effect by itself. Additionally, overexpression of ␤-galactosidase had no inhibitory effect on wild-type c-Neu-or Neu T-mediated signaling (data not shown).
Defining a Region of Caveolin-1 That Inhibits Neu-mediated Signal Transduction by Deletion Mutagenesis-We next employed this assay system and a variety of caveolin-1 deletion mutants to map this caveolin-1-dependent inhibitory activity to a given region of the caveolin-1 molecule. The caveolin-1 deletion mutants that were used are shown schematically in Fig.  5A. Note that the full-length caveolin-1 molecule contains residues 1-178 (␣-isoform), the ␤-isoform lacks residues 1-31, ⌬C lacks residues 141-178 of the C-terminal domain, and ⌬N lacks residues 1-95 of the N-terminal domain. These constructs have all been previously characterized and are expressed to equivalent levels in transfected cells (25). Fig. 5B shows the results of this mutational analysis. Virtually identical results were obtained for both wild-type c-Neu and mutationally activated Neu T. Relative to wild-type full-length caveolin-1 (FL; ␣-iso-

FIG. 3. Expression of caveolin-1 and -2 in v-Src-transformed NIH 3T3 cells and normal NIH 3T3 cells treated with ceramide derivatives.
Cell lysates were prepared and subjected to immunoblot analysis with caveolin-specific antibodies. A, comparison of normal and v-Src-transformed NIH 3T3 cells; B, comparison of NIH 3T3 cells treated with various ceramide derivatives (as indicated) or left untreated. The effects of ceramide (C 6 -D-erythro) purchased from two different suppliers (Calbiochem and Matreya, Inc.) are shown. Cells were cultured for 48 h in the presence of a given ceramide derivative (40 M each) in normal medium (i.e. Dulbecco's modified Eagle's medium supplemented with antibiotics and 10% donor calf serum). Note that caveolin-1 protein expression was down-regulated in response to these other stimuli (v-Src or ceramide) that are known to activate the p42/44 MAP kinase cascade. Cav-1, caveolin-1; Cav-2, caveolin-2.

FIG. 4. Overexpression of caveolin-1 blocks Neu-mediated signal transduction to the nucleus.
Transient coexpression of Neu with a luciferase reporter plasmid allowed us to measure the transactivation of the transcription factor Elk, which is a principal substrate for MAP kinases (see "Experimental Procedures"). Note that coexpression with caveolin-1 dramatically inhibited Neu-mediated signal transduction evoked by overexpression of either wild-type (WT) c-Neu or mutationally activated Neu T by ϳ10 -20-fold. More important, the empty vector used to express caveolin-1 (pCB7) had no inhibitory effect by itself. Also, the vector (pSV2) used to express wild-type c-Neu and Neu T had no activity by itself in this assay system. RLU, relative light units; FL, full length. form), ⌬C was ϳ10 -20-fold more potent, and ⌬N was ϳ10-fold less potent; the ␤-isoform was almost as potent as wild-type full-length caveolin-1. From this mutational analysis, we can conclude that the in vivo inhibitory activity of caveolin-1 is contained within the N-terminal domain and, to a first approximation, within caveolin-1 residues 32-95.
Peptides Derived from the Caveolin Scaffolding Domain Inhibit Neu Tyrosine Kinase Activity in Vitro-One possible explanation for the ability of caveolin-1 to impede Neu-mediated signal transduction is that caveolin-1 interacts directly with the Neu tyrosine kinase. We decided to test this hypothesis by examining the effects of synthetic peptides derived from caveolin-1 on the in vitro kinase activity of c-Neu. Fig. 6A shows that the antibody used for these experiments efficiently immunoprecipitated recombinantly expressed c-Neu. The potential inhibitory activity of caveolin-1-derived peptides was evaluated using an in vitro kinase assay that measures the activity of immunopurified recombinant wild-type c-Neu (Fig. 6, B and C). Only the caveolin-1 peptide encoding caveolin-1 residues 61-101 or 82-101 showed significant inhibitory activity, whereas a peptide encoding an adjacent region of caveolin-1 (residues 53-81) had no effect at the same concentration. Note that residues 82-101 correspond to the scaffolding domain (SD) of caveolin-1. However, if the peptide encoding caveolin-1 residues 82-101 was divided into two shorter peptides (residues 84 -92 and 93-101), its inhibitory activity was abolished.
In this in vitro kinase assay, the caveolin-1 scaffolding domain inhibited the autophosphorylation of c-Neu with an IC 50 between 0.3 and 1.0 M (Fig. 6C). These values correspond well to those observed previously for inhibition of other kinases involved in signal transduction by the caveolin-1 scaffolding domain peptide (19,26,27).

Expression of Caveolin-1 in Normal Mouse Mammary Tissue and Neu-induced Mammary Tumors-
The Neu (c-erbB2) protooncogene is amplified and overexpressed in a significant proportion of human mammary tumors (1,2). Are our current findings relevant in whole animals and specifically to mammary tumorigenesis? To answer this question, we chose to evaluate the expression of caveolin-1 protein in normal mouse mammary tissue and in mammary tumors from an established mouse model of c-Neu-induced mammary tumorigenesis. Targeted overexpression is achieved by using the MMTV long terminal repeat to drive mammary-specific expression of a given oncogene. Using this strategy, mammary tumorigenesis can be induced by targeted overexpression of several different classes of oncogenes, including the c-Neu receptor tyrosine kinases (9), cytoplasmic tyrosine kinases (Src) (20), nuclear transcription factors (Myc) (21), and cytoplasmic GTPases (Ras) (28). Fig. 7 shows a Western blot analysis of the expression of caveolin-1 protein in 10 independent mammary tumors, each derived from a different MMTV-Neu transgenic mouse. Normal virgin mammary tissue served as positive controls for caveolin-1 expression. Immunoblotting with anti-tubulin antibodies wild-type caveolin-1 and a series of caveolin-1 deletions mutants. Note that the full-length caveolin-1 molecule contains residues 1-178 (␣isoform), the ␤-isoform lacks residues 1-31, ⌬C lacks residues 141-178 of the C-terminal domain, and ⌬N lacks residues 1-95 of the N-terminal domain. b, effect of transiently expressing caveolin-deletion mutants on Neu-mediated signal transduction. Note that relative to wild-type fulllength caveolin-1 (FL; ␣-isoform), ⌬C was ϳ10 -20-fold more potent, and ⌬N was ϳ10-fold less potent; the ␤-isoform was almost as potent as wild-type full-length caveolin-1. These differences were best visualized by representing the data as a log scale of relative inhibitory activity. Virtually identical results were obtained for both wild-type (WT) c-Neu (upper panel) and mutationally activated Neu T (lower panel). Cav-1␣, caveolin-1␣.
FIG . 5. Defining a region within caveolin-1 that blocks Neumediated signal transduction in vivo. A, diagram summarizing served as a control for equal loading. Note that loss or dramatic reduction of caveolin-1 expression occurred in mammary tumors derived from several independent c-Neu-expressing transgenic lines.

What About Downstream Effectors of the Neu Signaling
Pathway?-Src family tyrosine kinases and Ras are thought to be downstream components of the Neu signaling pathway and have also been implicated in mammary tumorigenesis. Thus, we next examined caveolin-1 expression in mammary tumors from mice transgenically expressing MMTV-Src and MMTV-Ras. Fig. 8 shows that as with MMTV-Neu tumors, caveolin-1 expression was also dramatically reduced or undetectable in these tumors as well. Also, virtually identical results were obtained with tumors from MMTV-Myc mice. The histopathological appearances of these 25 independent mammary tumors included four distinct types of mammary tumors: comedo, scirrhous, papillary, and acinar adenocarcinomas.
Immunolocalization of Caveolin-1 in Normal Mouse Mammary Tissue and Mammary Tumors-Mammary tissue consists of a complex array of cells, including the mammary epithelium, fibroblasts, adipocytes, endothelial cells, smooth muscle cells, and many other cell types. However, it remains unknown whether caveolin-1 is expressed in normal mammary epithelium. Caveolin-1 expression has been well documented in fibroblasts, adipocytes, endothelial cells, and smooth muscle cells. Thus, "down-regulation" of caveolin-1 expression observed by Western analysis may simply reflect tumor mass derived from mammary epithelium that normally lacks caveolin-1 expression.
To examine this possibility, we immunolocalized caveolin-1 within normal mammary tissue. Our results indicate that caveolin-1 is normally abundant in the mammary epithelial cells, as the alveolar ductal epithelia were quite brightly stained with an anti-caveolin-1 antibody probe (Fig. 9). We next examined the distribution of caveolin-1 expression in mammary tumors derived from all four transgenic mouse models of mammary tumorigenesis (Fig. 9). In all cases, little or no caveolin-1 staining of the tumor mass itself was observed. However, adjacent fibro-fatty connective tissue remained caveolin-1-positive, and the vasculature within the tumors (consisting of endothelial cells and smooth muscle cells) was also brightly stained. These results clearly show that caveolin-1 is abundantly expressed in normal mammary epithelium, but that cancerous transformed mammary tumor cells have lost caveolin-1 expression.
FIG. 6. Caveolin-1-derived peptides inhibit the in vitro tyrosine kinase activity of Neu. A rabbit anti-peptide antibody directed against Neu was used to immunoprecipitate Neu from CHO cells transiently transfected with the c-Neu cDNA. A, shown are the results from Western blot analysis confirming that Neu is effectively immunoprecipitated from cell lysates. Immunoprecipitates were probed with a monoclonal antibody that is specific for Neu (mAb 42; Transduction Laboratories, Inc.). As a negative control, immunoprecipitation was carried out with an irrelevant IgG and is shown for comparison. B, a variety of caveolin-1 (Cav-1)-derived peptides were evaluated for their effect on the autophosphorylation of the Neu receptor tyrosine kinase. Note that peptides encoding caveolin-1 residues 61-101 or residues 82-101 effectively inhibited Neu autophosphorylation in vitro. Other caveolin-derived peptides had no effect. All five peptides were added at a concentration of 10 M. C, a peptide encoding the caveolin-1 scaffolding domain (residues 82-101) was evaluated for its dose-dependent effect on the autophosphorylation of the Neu receptor tyrosine kinase. Note that a dose of ϳ1 M yielded an inhibition of ϳ 90%; autophosphorylation was completely inhibited at 3 M. pAb, polyclonal antibody; IP, immunoprecipitate; SD, scaffolding domain.

FIG. 7. Expression of caveolin-1 in normal mouse mammary tissue and Neu-induced mammary tumors.
Expression of the proto-oncogene wild-type c-Neu was transgenically targeted to mammary tissue epithelium using the mouse mammary tumor virus long terminal repeat. Normal virgin "breast" tissue served as normal controls. Numbers reflect independently derived tumors that appeared in several different lines of c-Neu-expressing transgenic mice. Note that in most cases, caveolin-1 (Cav-1) expression was absent or dramatically reduced compared with normal control samples. Blotting with ␣-tubulin antibodies was performed as a control for equal protein loading.

DISCUSSION
Caveolae are plasma membrane-attached vesicular organelles with a diameter of ϳ50 -100 nm (29,30). It has been proposed that caveolae play a pivotal role in a number of essential cellular functions, including signal transduction, lipid metabolism, cellular growth control, and apoptotic cell death. The principal protein components of caveolae are the caveolin family of proteins, termed caveolin-1, -2, and -3 (29,30). Caveolins provide a means for caveolin-interacting proteins to be concentrated within caveolae membranes. Caveolins interact directly with a number of caveolae-associated signaling molecules, such as Ha-Ras, heterotrimeric G-proteins, epidermal growth factor receptor, protein kinase C, Src family tyrosine kinases, and nitric-oxide synthase isoforms (19,26,27,31,32). In many of these cases, it has been documented that caveolin binding can effectively inhibit the enzymatic activity of these signaling molecules in vitro.
As caveolin-1 was the first caveolin family member to be identified, it has served as the prototype for the study of caveolins and caveolae function. Caveolin-1 is a 22-24-kDa integral membrane protein that consists of 178 amino acid residues. The central region of caveolin-1 contains a string of 33 hydrophobic amino acids that function as a membrane anchor, al-lowing caveolin-1 to assume a hairpin configuration with both N-and C-terminal domains facing the cytoplasm. Residues 61-100 of caveolin-1 function to direct the self-oligomerization of caveolin-1 and as a plasma membrane-bound scaffold to sequester specific caveolin-interacting proteins within caveolae membranes (19,37,38).
Here, we show that mutational activation of c-Neu downregulates caveolin-1 protein expression, but not caveolin-2, in Note that in all cases, little or no caveolin-1 staining of the tumor mass itself was observed. However, the vasculature within the tumors (consisting of endothelial cells and smooth muscle cells) remained caveolin-1-positive. Left panels, phase image; right panels, immunofluorescent (IF) staining. cultured NIH 3T3 and Rat 1 cells. A dramatic reduction of caveolin-1 expression also occurs in mammary tumors derived from c-Neu-expressing transgenic mice and other transgenic mice expressing downstream effectors of Neu-mediated signal transduction. Conversely, recombinant overexpression of caveolin-1 blocks Neu-mediated signal transduction in vivo. Results from this mutational analysis show that this in vivo inhibitory activity is contained within caveolin-1 residues 32-95. Furthermore, a specific peptide encoding the caveolin-1 scaffolding domain (residues 82-101) inhibits Neu tyrosine kinase activity in vitro. These results suggest a reciprocal relationship between c-Neu tyrosine kinase activity and caveolin-1 protein expression.
Our present results are consistent with a number of independent experimental observations. (i) Transformation of NIH 3T3 cells with oncogenic forms of Abl or Ras is sufficient to down-regulate caveolin-1 protein expression; however, caveolin-2 protein levels remain unchanged (13,24). (ii) A variety of receptor tyrosine kinases (platelet-derived growth factor receptor, insulin receptor, nerve growth factor receptor (trk), EGFR, and c-Neu) have been shown to co-fractionate with caveolin-1 in a variety of cell fractionation approaches (26, 39 -43). (iii) Caveolin-1 can interact directly with the EGFR in vitro, through the recognition of a conserved caveolin-binding motif within the kinase domain; this interaction is sufficient to inhibit the in vitro autophosphorylation of the purified EGFR kinase domain (26). However, the in vivo relevance of these results has remained unknown. In this study, we have demonstrated that caveolin-1 inhibits Neu-mediated signaling in vivo and in vitro. As Neu possesses a putative caveolin-1-binding motif within its kinase domain that is identical to that of the EGFR (DVWSYGVTVWEL in both c-Neu and the EGFR), it is not unreasonable to suspect a direct interaction between c-Neu and caveolin-1.
Of note, we observed an interesting difference between the effects of c-Neu overexpression in cultured cells and in transgenic mice. In cultured Rat 1 cells, overexpression of wild-type c-Neu was not sufficient to effectively down-regulate caveolin-1. In striking contrast, targeted expression of wild-type c-Neu in mouse mammary cells was sufficient to down-regulate caveolin-1 in mammary tumors. These results suggest that mammary cells in vivo may be more sensitive to the downregulation of caveolin-1. Alternatively, c-Neu may have undergone somatic mutations during the process of mammary tumorigenesis, as has been documented previously (8). In addition, these findings point out that, in some instances, results with cultured cells may not adequately reflect the in vivo situation. This is particularly relevant in the case of caveolin-1, as many cell lines that we have examined lack caveolin-1 expression, whereas caveolin-1 is well expressed in the corresponding tissue in vivo. In many instances, these cell lines were derived from the primary culture of tumors, suggesting that caveolin-1 expression may be very sensitive to down-regulation during cell transformation. However, more examples will be necessary before we can conclude that caveolin-1 down-regulation is a general response to any form of cell transformation.
As amplification or mutational activation of Neu occurs in humans and has been implicated in the development of human breast cancers, our current experimental observations may have implications for the future diagnosis or treatment of human breast cancers. For example, it will be important to examine the possible relationship among c-Neu amplification/activation, caveolin-1 down-regulation, and clinical outcome in patients with breast cancers.