Flotillins/Cavatellins Are Differentially Expressed in Cells and Tissues and Form a Hetero-oligomeric Complex with Caveolins in Vivo

Caveolae are vesicular organelles that represent a subcompartment of the plasma membrane. Caveolins and flotillins are two families of mammalian caveolae-associated integral membrane proteins. However, it remains unknown whether flotillins interact with caveolin proteins to form a stable caveolar complex or if expression of flotillins can drive vesicle formation. Here, we examine the cell type and tissue-specific expression of the flotillin gene family. For this purpose, we generated a novel monoclonal antibody probe that recognizes only flotillin-1. A survey of cell and tissue types demonstrates that flotillins 1 and 2 have a complementary tissue distribution. At the cellular level, flotillin-2 was ubiquitously expressed, whereas flotillin-1 was most abundant in A498 kidney cells, muscle cell lines, and fibroblasts. Using three different models of cellular differentiation, we next examined the expression of flotillins 1 and 2. Taken together, our data suggest that the expression levels of flotillins 1 and 2 are independently regulated and does not strictly correlate with known expression patterns of caveolin family members. However, when caveolins and flotillins are co-expressed within the same cell, as in A498 cells, they form a stable hetero-oligomeric “caveolar complex.” In support of these observations, we show that heterologous expression of murine flotillin-1 in Sf21 insect cells using baculovirus-based vectors is sufficient to drive the formation of caveolae-like vesicles. These results suggest that flotillins may participate functionally in the formation of caveolae or caveolae-like vesicles in vivo. Thus, flotillin-1 represents a new integral membrane protein marker for the slightly larger caveolae-related domains (50–200 nm) that are observed in cell types that fail to express caveolin-1. As a consequence of these findings, we propose the term “cavatellins” be used (instead of flotillins) to describe this gene family.

Caveolin, a 21-24-kDa integral membrane protein, is a principal component of caveolae membranes in vivo (9 -13). Caveolin-rich membrane domains purified by either detergent-based or detergent-free methods are enriched in a variety of lipidmodified signaling molecules such as heterotrimeric G proteins, Src-family tyrosine kinases, Ha-Ras and Rap GTPases, and endothelial cell nitric-oxide synthase (1, 14 -24). Many of these signaling molecules interact in a regulated manner directly with caveolin (18,25,26).
However, caveolin is only the first member of a growing gene family of caveolin proteins; caveolin has been re-termed caveolin-1. Three different caveolin genes (Cav-1, Cav-2, and Cav-3) encoding four different subtypes of caveolin have been described thus far (2). There are two subtypes of caveolin-1 (Cav-1␣ and Cav-1␤) that differ in their respective translation initiation sites (27). The tissue distribution of caveolin-2 mRNA is extremely similar to caveolin-1 mRNA (5). In striking contrast, caveolin-3 mRNA and protein is expressed mainly in muscle tissue types (skeletal, cardiac, and smooth) (28 -30).
Recently, we have identified another family of integral membrane proteins that may contribute to the structural organization of caveolae membranes (31,32). Micro-sequence analysis of purified caveolin-rich membrane domains isolated from lung tissue revealed a novel ϳ45-kDa component of caveolae membranes termed flotillin (31). Molecular cloning of flotillin and analysis of the cDNA for this protein has provided new avenues by which to explore the structure and function of caveolae organelles. Interestingly, flotillin is a close homologue of ESA 1 (epidermal surface antigen (33)), and together they define a new "flotillin family" of caveolae-associated integral membrane proteins (flotillin-1 and flotillin-2/ESA)) (31). However, the study of flotillin-1 has been hampered by the lack of a flotillin-1 specific antibody probe. Previously, we generated a rabbit antipeptide antibody against murine flotillin-1. This anti-peptide antibody did not recognize flotillin-1 well in crude extracts, precluding a detailed analysis of the tissue distribution and expression patterns of the flotillin protein (31). These antipeptide antibodies also recognized another protein of ϳ27 kDa we termed f គlotillin c គross-r គeacting d គ eterminant (FCRD) (31). The identity and significance of flotillin cross-reacting determinant remains unknown.
Here, we examine the cell type and tissue-specific expression of the flotillins using a novel monoclonal antibody probe that recognizes only flotillin-1. A survey of cell and tissue types demonstrates that flotillins 1 and 2 have a complementary tissue distribution. In addition, co-immunoprecipitation experiments revealed that flotillins 1 and 2 are part of a stable hetero-oligomeric complex that contains both caveolins 1 and 2. However, protein levels of flotillin-1 and flotillin-2/ESA remain unchanged in response to oncogenic transformation of NIH 3T3 cells. Our data suggest that the expression of the flotillins can be independently regulated from that of caveolin-1 and caveolae formation. Also, we show that flotillin-1 represents a new marker protein for the slightly larger caveolae-related domains (50 -200 nm) that are observed in cell types that fail to express caveolin-1 and do not contain detectable caveolae.
Hybridoma Production-A monoclonal antibody to murine flotillin-1 was generated by multiple immunizations of Balb/c female mice with a fusion protein encoding the full-length flotillin protein. Mice showing the highest titer of anti-flotillin-1 immunoreactivity were used to create fusions with myeloma cells using standard protocols (35). Positive hybridomas were cloned twice by limiting dilution. These hybridomas were also screened against flotillin-2/ESA to prevent the selection of a cross-reacting monoclonal antibody. Positive hybridomas recognizing only flotillin-1, but not flotillin-2, were then injected into mice to produce ascites fluid. IgGs were purified by affinity chromatography on protein A-Sepharose. These antibodies were produced in collaboration with Drs. Roberto Campos-Gonzalez and John R. Glenney, Jr. (Transduction Laboratories).
Cell Culture-HEK-293T cells were the gift of Dr. Anthony J. Koleske (in Dr. David Baltimore's laboratory at MIT, Cambridge, MA) and were propagated in t75 tissue-culture flasks in Dulbecco's modified Eagle's medium supplemented with antibiotics and 10% fetal bovine serum, as we described previously (34). v-Abl-and Ha-Ras (G12V)transformed NIH 3T3 cells were as we described previously and were propagated in t75 tissue-culture flasks in Dulbecco's modified Eagle's medium supplemented with antibiotics and 10% donor bovine serum (34,36,37).
Tissue Western-Approximately 200 mg (wet weight) of various mouse tissues were lysed in immunoprecipitation buffer and homogenized on ice with a Polytron tissue grinder, as described (4). Equal amounts (100 g of protein) were loaded on an SDS-PAGE gel (12% acrylamide). After transfer to nitrocellulose, the blot was probed with antibodies directed against flotillin-1 and flotillin-2/ESA.
Immunoblotting-Samples were separated by SDS-PAGE (15% acrylamide) and transferred to nitrocellulose. After transfer, nitrocellulose sheets were stained with Ponceau S to visualize protein bands and subjected to immunoblotting. For immunoblotting, incubation conditions were as described by the manufacturer (Amersham Pharmacia Biotech), except we supplemented our blocking solution with both 1% bovine serum albumin and 1% nonfat dry milk (Carnation).
Synthesis of Immobilized Flotillin-derived Peptides for Epitope Mapping-Flotillin-derived polypeptides were synthesized directly onto an activated polymeric membrane by Research Genetics. The peptide chemistry was standard Fmoc with coupling mediated through butyl alcohol/DIC and Fmoc removal with 1:3 piperidine/N,N-dimethylformamide. For final peptide protecting group removal, the membrane was placed in a bath of 50:47.5:2.5:1.5:1 DCM/trifluoroacetic acid/thioanisole/EDT/anisole for 1 h and, finally, washed and dried. These sheets were probed by immunoblotting as described above.
Cell Culture Models of Skeletal and Neuronal Differentiation-C2C12-3 cells (39) were derived from a single colony of C2C12 cells (40) cultured at clonal density and display a more stable phenotype than the parental cell line. C2C12-3 myoblasts were cultured as described elsewhere (39). Briefly, proliferating C2C12-3 cells were cultured in high mitogen medium (Dulbecco's modified Eagle's medium containing 15% fetal bovine serum and 1% chicken embryo extract) and induced to differentiate at confluence in low mitogen medium (Dulbecco's modified Eagle's medium containing 3% horse serum). Overt differentiation was indicated by the assembly of multinucleated syncytia, which commenced 36 -48 h after the cells were switched to low mitogen media. PC 12 cells were grown in RPMI 1640 medium with 5% fetal bovine serum and 10% horse serum. PC 12 cells were differentiated for 1-4 days by culturing the cells in low serum medium (RPMI 1640 with 1% horse serum) with nerve growth factor (100 ng/ml) (41).
Baculovirus-based Expression of Murine Flotillin-1 in Sf21 Insect Cells-Spodoptera frugiperda (Sf21) cells were provided by Dr. Takashi Okamoto (Cleveland Clinic Foundation). Sf21 cells were grown in Excell 401 medium containing 10% fetal bovine serum and antibiotics (penicillin-streptomycin) at 27°C. The cDNA encoding murine flotillin-1 was subcloned into the multiple cloning site of a transfer plasmid vector, pBacPAK9. A mixture of 2 g of recombinant plasmid pBacPAK 9-flotillin-1 DNA and 1 g of purified engineered baculoviral vector DNA BacPAK 6 (Bsu36I digest) (CLONTECH) were transfected into insect Sf21 cells, as suggested by the manufacturer (42). Four days later, culture supernatants were removed and centrifuged at 1,000 rpm for 10 min. Clarified supernatants containing wild-type and recombinant baculoviruses were plaque-assayed on a monolayer of Sf21 cells. Occlusion negative plaques were picked and seeded onto 2.5 ϫ 10 6 cells. After a 3-day incubation, cells and culture supernatants were removed and centrifuged at 1,000 rpm for 10 min. The cell pellets were analyzed by immunoblotting analysis using anti-flotillin-1 antibodies or anti-myc tag mAb 9E10. Those plaques testing positive for the presence of flotillin-1 were selected for three rounds of plaque purification. The selected plaques with the highest yield of expression were used as recombinant baculovirus stock for producing protein by infecting insect Sf21 cells.
Transmission electron microscopy was performed as described previously by our laboratory. Samples were fixed with glutaraldehyde, postfixed with osmium tetroxide, and stained with uranyl acetate and lead citrate, as detailed in Sargiacomo et al. (14) and Lisanti et al. (15). Samples were examined under the Philips 410 TEM.

Generation and Epitope-mapping of a mAb Probe Specific for
Flotillin-1-A fusion protein carrying the full-length flotillin-1 protein was used to generate a flotillin-1-specific monoclonal antibody probe. This antibody does not recognize flotillin-2/ ESA (see "Experimental Procedures"). Thus, this antibody can be used in conjunction with other published antibodies to study the function and differential expression of distinct flotillin and caveolin family members. Fig. 1 demonstrates that this novel mAb probe recognizes both mammalian flotillin-1 (murine) and Drosophila flotillin-1 proteins. Immunoblotting with mAb 9E10 was also used in parallel to detect these C-terminal myc-tagged proteins.
As this flotillin-1 mAb reacts with both murine and Drosophila forms of flotillin-1, it must recognize an evolutionarily conserved amino acid epitope. Thus, we performed epitope mapping by generating a series of 48 overlapping peptides that are derived from the sequence of murine flotillin-1 (listed in Table  I). These peptides were synthesized as an immobilized array that can be probed by immunoblotting. Fig. 2 shows the results of this epitope-mapping strategy. The flotillin-1 mAb only reacted with a single peptide (number 38 in Table I). As expected based on its reactivity with both mammalian and Drosophila flotillin-1, this epitope is highly conserved but is divergent in flotillin-2/ESA. An alignment of this region of flotillin-1 and other flotillin-related proteins is shown in Fig. 2C.
Cell Type and Tissue-specific Expression of Flotillins 1 and 2-To establish the tissue distributions of flotillins 1 and 2, we prepared tissue extracts from a variety of different murine tissues (Fig. 3). Flotillin-1 is detected mainly in striated muscle                                                SISQVNHNKPLRTA tissues (heart, diaphragm, and psoas muscle), adipose tissue, and lung. In striking contrast, flotillin-2/ESA shows a much wider tissue distribution but is virtually absent in skeletal muscle (diaphragm and psoas muscle). Thus, flotillins 1 and 2 show a relative complementary tissue distribution.
To identify model cell systems to study flotillins 1 and 2, we examined the expression of flotillins in a variety of commonly used cell lines and primary cultured cells (Fig. 4). Note that flotillin-2/ESA is most widely expressed, whereas flotillin-1 shows a more restricted distribution. More specifically, flotillin-1 was most abundant in A498 kidney cells, muscle cell lines (smooth muscle cells, L6 skeletal myoblasts, and BC3H1 myoblasts) and fibroblasts.

Flotillins 1 and 2 Form a Stable Hetero-oligomeric Complex with Caveolins 1 and 2-Given that flotillins have been shown
to co-fractionate with caveolin-1 using three independent fractionation schemes used to purify caveolae membranes, we wondered whether they form a complex with caveolins (31). To address this issue, we performed a series of co-immunoprecipitation experiments using A498 cells; these cells co-express both flotillin 1 and 2 (See Fig. 4).
Lysates from A498 cells were prepared and subjected to immunoprecipitation with an anti-peptide antibody that recognizes the unique N terminus of caveolin-1 that is not found in other caveolin family members. These immunoprecipitates were then probed by Western blot analysis using monoclonal antibodies directed against caveolin-1 (mAb 2297), caveolin-2 (mAb 65), flotillin-1 (mAb 18), and flotillin-2/ESA (mAb 29). Fig. 5 demonstrates that the anti-peptide antibody directed against the unique N terminus of caveolin-1 can be used to co-immunoprecipitate caveolin-2, flotillin-1, and flotillin-2.
Thus, it appears that caveolins 1 and 2 and flotillins 1 and 2 form a stable complex in vivo.
Differential Expression of Flotillins and Caveolins in Oncogenically Transformed NIH 3T3 Cells-Caveolin-1 mRNA and protein expression are reduced or absent in NIH 3T3 cells transformed by a variety of activated oncogenes (such as v-Abl and Ha-Ras (G12V)); caveolae organelles are also missing from these transformed cells (36). However, it remains unknown whether flotillins 1 and 2 are down-regulated in response to oncogenic transformation. Fig. 6 shows that although caveolin-1 expression is dramat-  ically down-regulated in v-Abl and Ha-Ras (G12V)-transformed NIH 3T3 cells, the expression of flotillins 1 and 2 remains virtually unaffected in v-Abl and Ha-Ras (G12V)-transformed NIH 3T3 cells. The expression of caveolin-2 is shown for comparison. As we have previously demonstrated that these transformed cells do not contain detectable caveolae (36), it appears that co-expression of caveolin-2, flotillin-1, and flotillin-2/ESA is not sufficient to drive caveolae formation. Thus, flotillins 1 and 2 can be expressed within cells that lack morphologically distinguishable caveolae.
However, this does not preclude a role for the flotillins in the formation of caveolae (50 -100 nm) or the slightly larger caveo-lae-related domains (50 -200 nm) that are observed in cells that fail to express caveolin-1 and do not contain detectable caveolae (See below).
Flotillin-2/ESA Is Up-regulated during the Differentiation of C2C12 Skeletal Myoblasts in Vitro-As flotillins 1 and 2 were co-expressed in a variety of muscle cell lines (smooth muscle cells, L6 myoblasts, and BC3H1 myoblasts) and are expressed in muscle tissue types (heart, diaphragm, and psoas muscle), we assessed whether flotillins are induced during differentiation of C2C12 cells from myoblasts to myotubes. Fig. 7 shows that although flotillin-1 is undetectable in myoblasts and myotubes, flotillin-2/ESA is dramatically induced during this process of differentiation.
Expression of Flotillins 1 and 2 during PC12 Cell Differentiation-As flotillin-2/ESA was particularly abundant in brain, we next assessed the expression of flotillins during neuronal differentiation using PC12 cells as a model system. Although PC12 cells underwent morphological differentiation in response to treatment with nerve growth factor, little or no change in the expression of flotillins 1 or 2 was observed (Fig. 8).

Recombinant Expression of Flotillin-1 in Insect Cells Is Sufficient to Drive the Formation of Caveolae-sized Vesicles-Full-
length flotillin-1 was integrated into an engineered baculovirus genome via a recombinant transfer plasmid, as detailed under "Experimental Procedures." For these constructions, a myc epitope tag was placed at the C terminus with a polyhistidine tag following the myc tag (Fig. 9A).
Murine flotillin-1 was expressed very well in Sf21 insect cells using the baculovirus system. Fig. 9B shows that flotillin-1 migrated at its expected molecular weight, including the myc and polyhistidine tags. We and others have previously shown that these tags do not interfere with the targeting of recombinant flotillin-1 and caveolins or the ability of caveolin-1 expression to drive vesicle formation (5,13,18,27,29,31,32,(43)(44)(45). However, it remains unknown whether the expression of recombinant flotillin-1 is sufficient to drive vesicle formation.
Electron microscopic analysis of uninfected Sf 21 insect cells did not show the appearance of caveolae, as we have reported previously (44) (Fig. 9C). However, insect cells infected with murine flotillin-1 accumulated a uniform population of caveolae-like structures within their cytoplasm. This population of vesicles was relatively homogeneous in size and were the same  5. Caveolins 1 and 2 and flotillins 1 and 2 form a stable hetero-oligomeric complex in vivo. A498 cells were lysed and subjected to immunoprecipitations with anti-caveolin-1 pAb (N-2-21), which recognizes the unique extreme N terminus of caveolin-1. These immunoprecipitates were then probed by Western analysis with monoclonal antibodies directed against caveolin-1 (mAb 2297), caveolin-2 (mAb 65), flotillin-1 (mAb 18), and flotillin-2/ESA (mAb 29). Protein A-Sepharose beads alone were also incubated with lysates and processed in parallel as a negative control for nonspecific binding. Note that anti-caveolin-1 IgG can be used to specifically co-immunoprecipitate caveolin-2 and flotillins 1 and 2.
Thus, it appears that flotillins and caveolins form a stable hetero-oligomeric complex. WB, Western blot.
Similarly, we have previously reported that insect cells infected with either caveolin-1 or caveolin-3 accumulated a uniform population of caveolae-like structures within their cytoplasm. However, this population of vesicles was more homogeneous in size, with a diameter of ϳ50 -100 nm (44,46). In contrast, expression of caveolin-2 did not drive vesicle formation, despite evidence of infection such as viral particles (46). These results indicate that a transmembrane domain alone is not sufficient to drive vesicle formation in Sf21 insect cells.
Thus, in this functional assay system, flotillin-1 behaved more like caveolins 1 and 3 than another member of the caveolin gene family, namely caveolin-2. These results suggest that flotillin-1 may participate in the formation of caveolae-related vesicles that are of a larger diameter than caveolae. DISCUSSION Flotillins are a new family of caveolae-associated integral membrane proteins. They were first identified by microsequence analysis of purified caveolae. Subsequently, molecular cloning and data base searches revealed two flotillins, flotillin-1 and flotillin-2/ESA (31). Flotillin-2/ESA was independently identified as a epidermal surface antigen using a mAb probe directed against epidermal cell plasma membrane fractions. This antibody probe also was sufficient to detach cultured keratinocytes, suggesting that ESA may function in cellcell or cell substrate attachments. In collaboration with Lodish, Schnitzer, and colleagues (31), we have shown that both flotillins 1 and 2 are dramatically enriched within caveolae purified from adipocytes and endothelial cells (31). However, the rabbit anti-peptide antibodies that we generated against murine flotillin-1 did not recognize flotillin-1 well in crude extracts, precluding an analysis of its true tissue distribution and expression patterns. These anti-peptide antibodies also recognized another unknown protein of ϳ27 kDa, which we termed flotillin cross-reacting determinant.
Here, we have generated and characterized a novel monoclonal antibody probe that selectively recognizes murine flotillin-1. As this antibody is directed against an evolutionarily conserved epitope, this mAb probe also recognizes Drosophila flotillin-1. A survey of mammalian cell lines and murine tissue types reveals that flotillins 1 and 2 have distinct tissue distributions, and flotillin-2 appears to be more ubiquitously expressed than flotillin-1. Using A498 cells, we show that caveolins (Cav-1 and Cav-2) and flotillins form a stable heterooligomeric complex as seen by co-immunoprecipitation studies using an antibody directed against the unique N terminus of caveolin-1. This suggests that flotillins 1 and 2 may play some unknown structural role within caveolae membranes or caveolae-related domains.
Caveolin-1 mRNA and protein expression are down-regulated in response to transformation of NIH 3T3 cells by activated oncogenes such as v-Abl and Ha-Ras; caveolae organelles are also missing from these transformed cells (36). Conversely, recombinant expression of caveolin-1 in these cell lines is sufficient to restore caveolae formation. Interestingly, we find here that these transformed cells that lack morphologically detectable caveolae continue to express flotillins 1 and 2 and caveolin-2. These results indicate that expression of flotillins is not sufficient to generate mature invaginated caveolae.
Using two independent model cell systems, we also show that the expression of flotillins is independently regulated during myoblast differentiation and neuronal differentiation. This suggests that flotillins 1 and 2 may be functionally redundant or that their individual functions may be tailored to a given tissue and cell type. Taken together, our data suggest that the expression levels of flotillins 1 and 2 do not strictly correlate with known expression patterns of caveolin family members. However, when caveolins and flotillins are co-expressed within the same cell, as in A498 cells, they form a stable heterooligomeric caveolar complex. In this regard, the antibodies we have generated and characterized should provide an invaluable tool to elucidate the function of flotillin-1.
A number of investigators have purified caveolae from cells and tissues that lack apparent expression of caveolin (47)(48)(49)(50)(51). These membranes were purified based on their Triton insolubility and light buoyant density in sucrose gradients; they have also been purified based simply on their light buoyant density in the absence of detergents. These domains are ϳ50 -200 nm in diameter and have been termed Triton-insoluble complexes, detergent-resistant membranes, low-density membranes, and caveolae-related domains. Like caveolae, these caveolae-related microdomains are dramatically enriched in cholesterol, sphingolipids, and lipid-modified signaling molecules (47,48).
The existence of "caveolae-related domains" or caveolae-related domains that fail to contain caveolin has caused considerable confusion (47). However, this was at a time when only one caveolin gene was known to exist, i.e. caveolin (now termed caveolin-1). In addition, it has been shown that caveolin-1 and caveolae are down-regulated in response to cell transformation, whereas caveolin-2 levels remain constant (34). As a consequence, many commonly used cell lines lack caveolin-1 protein expression and visible caveolae, as they are immortalized or transformed. This suggests that caveolae may be more versatile structures than we have previously imagined. In addition, expression of caveolin-2 in the absence of caveolin-1 is apparently not sufficient to generate invaginated caveolae that are visible by transmission electron microscopy (34).
A similar situation also existed for caveolae purified from striated muscle, as cardiac myocytes and skeletal muscle myocytes fail to express caveolin-1; we now know that muscle cell caveolin contain caveolin-3 (28,29,52). Thus, if a cell does not express detectable amounts of caveolins-1, -2, or -3, this does not mean that they do not express a protein that fulfills an equivalent function, a functional homologue rather than a structural homologue. Other detergent-insoluble membrane proteins have recently been cloned such as VIP-36, VIP-17/ MAL, and the flotillin gene family (flotillin-1 and flotillin-2/ ESA), and one or more of them may represent functional homologues of the caveolins (31,47,53).
In support of this hypothesis, we now show here that heterologous expression of murine flotillin-1 is sufficient to drive the formation of caveolae-like vesicles. These vesicles have a relatively uniform diameter of ϳ50 -200 nm but are slightly larger than caveolae (ϳ50 -100 nm). These results suggest that flotillins may participate functionally in the formation of caveolae or caveolae-like vesicles in vivo. Thus, flotillin-1 represents a new integral membrane protein marker for the slightly larger detergent-resistant caveolae-related domains (50 -200 nm) that are observed in cell types which fail to express caveolin-1 and do not contain detectable caveolae. In light of this new functional data, we propose the more descriptive term "cavatellins 2 " be used to refer to this gene family