Flotillin and Epidermal Surface Antigen Define a New Family of Caveolae-associated Integral Membrane Proteins
1997; Elsevier BV; Volume: 272; Issue: 21 Linguagem: Inglês
10.1074/jbc.272.21.13793
ISSN1083-351X
AutoresPerry E. Bickel, Philipp E. Scherer, Jan E. Schnitzer, Phil Oh, Michael P. Lisanti, Harvey F. Lodish,
Tópico(s)Lipid metabolism and disorders
ResumoCaveolae are plasmalemmal microdomains that are involved in vesicular trafficking and signal transduction. We have sought to identify novel integral membrane proteins of caveolae. Here we describe the identification and molecular cloning of flotillin. By several independent methods, flotillin behaves as a resident integral membrane protein component of caveolae. Furthermore, we have identified epidermal surface antigen both as a flotillin homologue and as a resident caveolar protein. Significantly, flotillin is a marker for the Triton-insoluble, buoyant membrane fraction in brain, where to date mRNA species for known caveolin gene family members have not been detected. Caveolae are plasmalemmal microdomains that are involved in vesicular trafficking and signal transduction. We have sought to identify novel integral membrane proteins of caveolae. Here we describe the identification and molecular cloning of flotillin. By several independent methods, flotillin behaves as a resident integral membrane protein component of caveolae. Furthermore, we have identified epidermal surface antigen both as a flotillin homologue and as a resident caveolar protein. Significantly, flotillin is a marker for the Triton-insoluble, buoyant membrane fraction in brain, where to date mRNA species for known caveolin gene family members have not been detected. Within mammalian cells, proteins are segregated into distinct organellar membrane compartments, and this localization influences the function of these proteins. One such class of compartments is the set of plasmalemmal microdomains known as caveolae (1Yamada E. J. Biophys. Biochem. Cytol. 1955; 1: 445-458Google Scholar, 2Severs N.J. J. Cell Sci. 1988; 90: 341-348Google Scholar). These “little caves” are 50–100 nm invaginations of the plasma membrane that have distinct morphological and biochemical properties (3Anderson R.G.W. Curr. Opin. Cell Biol. 1993; 5: 647-652Google Scholar, 4Travis J. Science. 1993; 262: 1208-1209Google Scholar). Caveolae are present to some degree in most cell types, but they are particularly abundant in endothelial cells, adipocytes, fibroblasts, smooth muscle cells, and type I pneumocytes (5Lisanti M.P. Scherer P.E. Tang Z.-L. Sargiacomo M. Trends Cell Biol. 1994; 4: 231-235Google Scholar, 6Lisanti M.P. Scherer P.E. Tang Z.-L. Kubler E. Koleske A.J. Sargiacomo M.S. Semin. Dev. Biol. 1995; 6: 47-58Google Scholar). Coating the cytoplasmic surface of caveolae are concentric filaments composed at least in part by the family of 20–24-kDa integral membrane caveolin proteins (7Rothberg K.G. Heuser J.E. Donzell W.C. Ying Y. Glenney J.R. Anderson R.G.W. Cell. 1992; 68: 673-682Google Scholar). In neurons plasma membrane invaginations that exhibit caveolae-like properties have been observed by electron microscopy (8Bouillot C. Prochiantz A. Rougon G. Allinquant B. J. Biol. Chem. 1996; 271: 7640-7644Google Scholar, 9Olive S. Dubois C. Schachner M. Rougon G. J. Neurochem. 1995; 65: 2307-2317Google Scholar), but mRNA species for known caveolin gene family members have not been detected. Caveolae have at least three functions (6Lisanti M.P. Scherer P.E. Tang Z.-L. Kubler E. Koleske A.J. Sargiacomo M.S. Semin. Dev. Biol. 1995; 6: 47-58Google Scholar). First, in endothelial cells, caveolae mediate the transcytosis of macromolecules from the vascular lumen to the sub-endothelial space (10Simionescu N. Simionescu M. Palade G.E. J. Cell Biol. 1975; 64: 586-607Google Scholar). In accord with this transport function, caveolae contain proteins that have been implicated in vesicular trafficking (11Schnitzer J.E. Liu J. Oh P. J. Biol. Chem. 1995; 270: 14399-14404Google Scholar). Moreover, in a cell-free system caveolae bud as vesicles from plasma membranes derived from endothelial cells in a time- and GTP-dependent manner, and in permeabilized cells GTP stimulates the endocytosis of cholera toxin B chain via caveolae (12Schnitzer J.E. Oh P. McIntosh D.P. Science. 1996; 274: 239-242Google Scholar). Second, caveolae are the sites of potocytosis, whereby small molecules are concentrated within caveolae by binding to glycosylphosphatidylinositol (GPI) 1The abbreviations used are: GPI, glycosylphosphatidylinositol; ESA, epidermal surface antigen; Mes, 4-morpholineethanesulfonic acid; PBS, phosphate-buffered saline; FCRD,flotillin cross-reactingdeterminant; PAGE, polyacrylamide gel electrophoresis; PCR, polymerase chain reaction; ORF, open reading frame; P, plasma membrane; V, caveolae. 1The abbreviations used are: GPI, glycosylphosphatidylinositol; ESA, epidermal surface antigen; Mes, 4-morpholineethanesulfonic acid; PBS, phosphate-buffered saline; FCRD,flotillin cross-reactingdeterminant; PAGE, polyacrylamide gel electrophoresis; PCR, polymerase chain reaction; ORF, open reading frame; P, plasma membrane; V, caveolae.-linked receptors and then traverse the plasma membrane into the cytoplasm via an unknown transporter (13Anderson R.G.W. Kamen B.A. Rothberg K.G. Lacey S.W. Science. 1992; 255: 410-411Google Scholar). Third, caveolae may participate in the relay of extracellular signals to the cell's interior by organizing signal transduction molecules (“the caveolae signaling hypothesis”) (5Lisanti M.P. Scherer P.E. Tang Z.-L. Sargiacomo M. Trends Cell Biol. 1994; 4: 231-235Google Scholar). Caveolin-rich membrane domains purified by either detergent-based or detergent-free methods are enriched in several distinct classes of signaling molecules, including Gα and Gβγ subunits of heterotrimeric GTP-binding proteins, Src-like tyrosine kinases, protein kinase Cα, and small GTP-binding proteins such as H-Ras and Rap GTPases (14Song K.S. Li S. Okamoto T. Quilliam L.A. Sargiacomo M. Lisanti M.P. J. Biol. Chem. 1996; 271: 9690-9697Google Scholar, 15Li S. Couet J. Lisanti M.P. J. Biol. Chem. 1996; 271: 29182-29190Google Scholar, 16Li S. Okamoto T. Chun M. Sargiacomo M. Casanova J.E. Hansen S.H. Nishimoto I. Lisanti M.P. J. Biol. Chem. 1995; 270: 15693-15701Google Scholar). Some of these signaling molecules physically interact with caveolin (15Li S. Couet J. Lisanti M.P. J. Biol. Chem. 1996; 271: 29182-29190Google Scholar). During their biosynthesis, caveolin-1 monomers associate to form homo-oligomers of ∼350 kDa (14–16 monomers per oligomer) (17Sargiacomo M. Scherer P.E. Tang Z.-L. Kubler E. Song K.S. Sanders M.C. Lisanti M.P. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9407-9411Google Scholar). These higher order structures of caveolin monomers may form a scaffold on which caveolin-interacting signaling molecules are organized or sequestered. These purified homo-oligomers also can undergo a second stage of oligomerization and assemble in vitro into structures that are similar in size to caveolae. However, it remains unknown whether other novel integral membrane proteins contribute to the structural organization of caveolae membranes in vivo. To systematically identify novel protein components of caveolae, we have purified membrane domains that are significantly enriched in caveolin from murine lung tissue (18Lisanti M.P. Tang Z.-T. Scherer P.E. Sargiacomo M. Methods Enzymol. 1995; 250: 655-668Google Scholar, 19Scherer P.E. Tang Z. Chun M. Sargiacomo M. Lodish H.F. Lisanti M.P. J. Biol. Chem. 1995; 270: 16395-16401Google Scholar). This method of purification relies upon the characteristic insolubility of caveolae domains in Triton X-100 at 4 °C and their characteristic buoyant density when subjected to sucrose density gradient centrifugation. A ∼45-kDa component of these purified caveolin-rich membranes was one of 10 predominant polypeptides easily detected by Ponceau S staining (Fig. 1B in Ref. 19). Microsequencing of this ∼45-kDa component has revealed several novel peptide sequences, as well as peptide sequences from epidermal surface antigen (ESA). ESA was identified and cloned as a keratinocyte cell surface protein (20Schroeder W.T. Stewart-Galetka S. Mandavilli S. Parry D.A.D. Goldsmith L. Duvic M. J. Biol. Chem. 1994; 269: 19983-19991Google Scholar). Due to the emerging importance of caveolae in vesicular trafficking and signal transduction, we proceeded to clone the cDNA corresponding to the novel caveolae protein, which we have named “flotillin.” 2As we report, “flotillin” is a specific marker for the set of proteins that float like a flotilla of ships in the Triton-insoluble, buoyant fraction (flotilla: Spanish, diminutive of flota, fleet; from Old French, flote; from Old Norse, floti, raft, fleet) from American Heritage Dictionary of the English Language (1975) American Heritage Publishing Co., Inc., New York, NY. 2As we report, “flotillin” is a specific marker for the set of proteins that float like a flotilla of ships in the Triton-insoluble, buoyant fraction (flotilla: Spanish, diminutive of flota, fleet; from Old French, flote; from Old Norse, floti, raft, fleet) from American Heritage Dictionary of the English Language (1975) American Heritage Publishing Co., Inc., New York, NY. Surprisingly, flotillin is a close homologue of ESA, and together they define a new family of integral membrane proteins in caveolae. Dulbecco's modified Eagle's tissue culture medium lacking methionine, cysteine, and glutamine was purchased from ICN Pharmaceuticals, Inc. The Express Protein Labeling Reagent, a mixture of [35S]methionine and -cysteine was purchased from DuPont NEN. 3T3-L1 mouse fibroblasts (American Type Culture Collection, Rockville, MD) were propagated and differentiated to adipocytes as described (21Frost S.C. Lane M.D. J. Biol. Chem. 1985; 260: 2646-2652Google Scholar). Poly(A)+ RNA was isolated from mouse tissues and from 3T3-L1 cells at preconfluence and at progressive stages of differentiation as described (22Baldini G. Hohl T. Lin H. Lodish H.F. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 5049-5052Google Scholar), with the exception that RNA used for construction of the adipocyte phage library was eluted from the oligo(dT)-cellulose column in 10 mmTris-HCl, pH 7.4, 1 mm EDTA, without SDS. Formaldehyde/agarose gel electrophoresis of poly(A)+ RNA, transfer to nylon membranes, and 32P labeling of DNA were performed as described (22Baldini G. Hohl T. Lin H. Lodish H.F. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 5049-5052Google Scholar). Hybridizations were carried out overnight at 42 °C in 50% formamide, 5 × SSC, 25 mm sodium phosphate, pH 7.0, 10 × Denhardt's solution, 5 mmEDTA, 1% SDS, and 0.1 mg/ml poly(A). The nylon membranes were washed sequentially in 2 × SSC, 0.1% SDS, and 0.1 × SSC, 0.1% SDS at 50 °C. Autoradiography was performed at −80 °C with an intensifying screen. Microsequencing of the ∼45-kDa region from mouse lung caveolin-rich membrane domains was performed as described (19Scherer P.E. Tang Z. Chun M. Sargiacomo M. Lodish H.F. Lisanti M.P. J. Biol. Chem. 1995; 270: 16395-16401Google Scholar). Sense and antisense degenerate primers were designed based upon the novel peptide sequences listed in Table Iand were synthesized by Research Genetics, Huntsville, AL. These sense and antisense primer pairs were used for nested PCR with mouse lung first-strand cDNA as the template. Mouse lung first-strand cDNA was prepared from 1 μg of mouse lung polyadenylated RNA, according to the manufacturer's instructions (Life Technologies, Inc.). A discrete 0.3-kilobase pair band was generated by hemi-nested PCR as follows. The first reaction used the sense primer “DLV” (GAYHTNGTNAAYATGGGNAT) and the antisense primer “AYQ” (TGRTANGCNARRTCNGCYTG) (note: where Y = C,T; H = A,C,T; N = G,A,C,T; R = G,A; D = G,A,T) with the following conditions: 1.5 mmMgCl2, 0.2 mm each dNTP, 100 pmol of each primer, and 2.5 units of Amplitaq (Perkins-Elmer) in a 50-μl reaction volume. The cycling parameters were as follows: 95 °C for 3 min, (95 °C for 1 min, 42 °C for 1 min, 72 °C for 2 min) times 5 cycles, (95 °C for 1 min, 50 °C for 1 min, 72 °C for 2 min) times 25 cycles, and 72 °C for 10 min. This first reaction did not produce a discrete band. Of this first reaction 1 μl was used as template for the hemi-nested second reaction with sense primer DLV and antisense primer “EVN” (RTTNACYTCDATRTCRTA) (see note above for explanation). The conditions of the second reaction were identical to the first, except for the amplification cycles (95 °C for 1 min, 55 °C for 1.5 min, 72 °C for 2 min times 30 cycles). The 0.3-kilobase pair PCR product was subcloned into pCR-Script (Stratagene, Inc.) according to the manufacturer's instructions, sequenced, and found to correspond to an open reading frame. Because amino acid residues encoded by DNA sequences internal to the PCR primers matched residues from the microsequenced peptides, we concluded that this cDNA corresponded to the protein originally isolated from caveolin-rich membrane domains. The identical fragment could also be amplified from 3T3-L1 adipocyte first-strand cDNA.Table IPeptide sequences from caveolin-rich membrane domains (∼45 kDa)SequenceIdentityResiduesKVASSDLVNMGIXVVXXTLKNovel (flotillin)133 –152KXXYDIEVNXXXAQADLAYQLNovel (flotillin)220 –240KXQLIMQAXANovel (flotillin)302 –311KXAFXEEVNESA174 –182KXAEAQLAYELQGAXEQQKESA184 –202Caveolin-rich membrane domains were purified from mouse lung tissue. After SDS-PAGE and transfer to nitrocellulose, the protein bands were excised and digested with Lys-C. The resulting peptides were then separated by HPLC and subjected to peptide sequencing, as described (30Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z.-L. HermanoskiVosatka A. Tu Y.-H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Google Scholar). The ∼45-kDa region was of particular interest because two proteins were identified, one novel and one that corresponded to ESA. An X in the sequence indicates a residue that was indeterminate by microsequencing. Open table in a new tab Caveolin-rich membrane domains were purified from mouse lung tissue. After SDS-PAGE and transfer to nitrocellulose, the protein bands were excised and digested with Lys-C. The resulting peptides were then separated by HPLC and subjected to peptide sequencing, as described (30Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z.-L. HermanoskiVosatka A. Tu Y.-H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Google Scholar). The ∼45-kDa region was of particular interest because two proteins were identified, one novel and one that corresponded to ESA. An X in the sequence indicates a residue that was indeterminate by microsequencing. Poly(A)+ RNA (5 μg) from 3T3-L1 adipocytes at day 8 of differentiation was used as the template to construct a λEXlox™ library according to the manufacturer's instructions (Novagen, Inc., Madison, WI). Oligo(dT) was used to prime the first-strand cDNA synthesis. Prior to ligation to the phage arms, the cDNA was size-fractionated (>1.5 kilobase pairs) using a potassium acetate gradient (5–20%) as described (23Aruffo A. Seed B. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 8573-8577Google Scholar). The resulting library represented ∼1 × 106 independent clones. The library was amplified once, and this singly amplified library was used for screening. The 3T3-L1 adipocyte cDNA library described above was screened with a digoxigenin-labeled cDNA probe that corresponded to the original, partial clone generated by PCR (nucleotide position 571–843, Fig. 1). Screening was performed according to the manufacturer's instructions (Boehringer Mannheim). Several positive clones were obtained, none of which extended further 5′ than base 296. To generate a cDNA probe that would correspond to the 5′-end of flotillin, the 3T3-L1 adipocyte cDNA library described above was used as template in nested PCR with a vector-based primer (T7 Gene-10: TGAGGTTGTAGAAGTTCCG) and two antisense flotillin primers (TGGTCATCATGAATATCC and AGCATCTCCTTGTTCTGG). After removal of vector sequence by restriction digestion, the resulting PCR fragment was then digoxigenin-labeled and used to screen the 3T3-L1 adipocyte library as above. A positive clone was plaque-purified. The plasmid containing the positive clone, pFlotillin-1, was excised from the phage arms via Cre-lox recombination as described in the manufacturer's protocol (Novagen, Inc.). This plasmid was then fully sequenced on both strands, by manual sequencing using Sequenase (U. S. Biochemical Corp.) and by automated sequencing using an Applied Biosystems 373 Sequencer. This clone extended to nucleotide position 146 in the 5′-untranslated region. Because the 5′ PCR-generated fragment extended farther upstream than pFlotillin-1, we confirmed this upstream sequence by amplifying the 5′-end of flotillin as above from an independently constructed library (random-primed, mouse brain λgt10 library from CLONTECH). This fragment contained an identical sequence to that amplified from the adipocyte library, and even began with the same base. Accordingly, the sequence represented in Fig. 1 contains sequence data from these PCR-generated 5′-ends and pFlotillin-1. cDNAs were transiently transfected into COS-7 cells (10-cm dishes) by the DEAE-dextran method (24Seed B. Aruffo A. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 3365-3369Google Scholar). 48 h after transfection, cells were labeled for 10 min in 3 ml of Dulbecco's modified Eagle's medium lacking methionine and cysteine and supplemented with 0.5 mCi (1000 Ci/mmol) of Express Protein Labeling Reagent. Thereafter, cells were washed 3 times with chase medium (Dulbecco's modified Eagle's medium containing unlabeled methionine and cysteine at 1 mm and cycloheximide at 300 μm). Cells were then incubated for 5, 10, 20, 30, and 60 min with 2 ml of chase medium. The cells were washed twice with cold PBS and then scraped into TNET/OG lysis buffer (1% Triton X-100, 60 mm octyl glucoside, 5 mm EDTA, 20 mm Tris, pH 8.0, 150 mm NaCl, 1 mmphenylmethylsulfonyl fluoride, and 1 μg/ml leupeptin). Insoluble debris were removed by centrifugation for 10 min at 15,000 ×g. The postnuclear cell lysates were then incubated for 30 min at 4 °C with Protein A-Sepharose (Pharmacia Biotech Inc.). The Protein A-Sepharose was removed by centrifugation, and fresh Protein A-Sepharose was added along with the corresponding antibody. Immunoprecipitation was performed for 3 h at 4 °C. Immunoprecipitates were then washed 5 times with lysis buffer (lacking octyl glucoside), released from the Protein A-Sepharose by boiling in Laemmli sample buffer, and analyzed by SDS-PAGE (25Laemmli U.K. Nature. 1970; 227: 680-685Google Scholar). The gel was prepared for fluorography by incubation with 1 m sodium salicylate for 20 min. The dried gel was exposed to Kodak X-OMAT AR film at −80 °C. Antibodies against a flotillin peptide were generated in rabbits by Research Genetics, Huntsville, AL. This peptide, KAQRDYELKKATYD (residues 211–224, indicated within Fig.1), was coupled to keyhole limpet hemocyanin and injected. The resulting antisera were affinity-purified over a CNBr-activated Sepharose column (Pharmacia) to which the immune peptide had been coupled according to standard methods (26Harlow E. Lane D. Antibodies: A Laboratory Manual. 1988; (and 536–537, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY): 313-315Google Scholar). Antibodies against a carboxyl-terminal mouse GDI-1 peptide (FENMRKKQNDVFGEADQ) were generated in rabbits by Research Genetics as described above. Anti-Myc epitope IgG (mAb 9E10) was obtained from Santa Cruz Biotech. Anti-caveolin-1 IgG (mAb 2297) and anti-ESA monoclonal antibody were obtained from Transduction Laboratories. Polyclonal anti-Acrp30 antiserum was prepared as described (27Scherer P.E. Williams S. Fogliano M. Baldini G. Lodish H.F. J. Biol. Chem. 1995; 270: 26746-26749Google Scholar). Extraction of 3T3-L1 adipocytes with sodium carbonate was performed according to a previously described protocol (28Fujiki Y. Hubbard A.L. Fowler S. Lazarow P.B. J. Cell Biol. 1982; 93: 97-102Google Scholar). Adipocytes were washed 3 times in PBS and once in a solution containing 150 mm NaCl. The cells were then scraped into 1 ml of 100 mm sodium bicarbonate buffer, pH 11.5, and homogenized by 5 strokes in a 2-ml Dounce homogenizer. After 30 min incubation at 4 °C, the homogenates were centrifuged at 100,000 rpm at 4 °C in a TLA 100.2 rotor (Beckman). The pellets were resuspended in 0.5 ml of 100 mmsodium bicarbonate buffer, pH 11.5, and pellet and supernatant fractions were immediately mixed with an equal volume of SDS-PAGE (Laemmli) sample buffer. Any DNA in the pellets was sheared by several passages through a 26-gauge needle to decrease the viscosity; these samples were then used directly for SDS-PAGE and Western blot analysis. Plates of 3T3-L1 cells at interval stages of adipogenesis were washed twice with cold PBS, lysed in TNET/OG lysis buffer, incubated on ice for 15 min, and then centrifuged for 10 min at 4 °C in a Beckman Microfuge at 14,000 rpm. Protein determinations were made on the post-nuclear supernatants using the BCA assay (Pierce). Caveolin-rich membrane domains were purified according to an established method (29Brown D. Rose J.K. Cell. 1992; 68: 533-544Google Scholar, 30Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z.-L. HermanoskiVosatka A. Tu Y.-H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Google Scholar, 31Sargiacomo M. Sudol M. Tang Z.-L. Lisanti M.P. J. Cell Biol. 1993; 122: 789-807Google Scholar) with minor modifications. Briefly, 3T3-L1 adipocytes were washed 3 times with ice-cold PBS, scraped into 2 ml of Mes-buffered saline (25 mm Mes, pH 6.5, 150 mm NaCl; MBS) with 1% Triton X-100, 1 mmphenylmethylsulfonyl fluoride, and 1 μg/ml leupeptin and homogenized by 10 strokes in a glass Dounce homogenizer. The extract was then adjusted to 40% sucrose by addition of 2 ml of an 80% sucrose solution in MBS, transferred to an ultracentrifuge tube, and overlaid with 8 ml of a linear 5–30% sucrose gradient in MBS containing phenylmethylsulfonyl fluoride and leupeptin at the above concentrations but lacking Triton X-100. The gradients were centrifuged for 16 h at 4 °C in a Beckman SW41 rotor at 39,000 rpm. Caveolin-enriched membranes fractionate as a sharp, light-scattering band at a density of ∼15–20% sucrose (18Lisanti M.P. Tang Z.-T. Scherer P.E. Sargiacomo M. Methods Enzymol. 1995; 250: 655-668Google Scholar). Equal volumes of each fraction were analyzed by SDS-PAGE and Western blotting. The pellet was brought to a volume equal to that of the fractions in Laemmli sample buffer and solubilized in a Dounce homogenizer prior to boiling. Caveolin-rich membrane domains from 3T3-L1 adipocytes were purified by means of a previously reported protocol (14Song K.S. Li S. Okamoto T. Quilliam L.A. Sargiacomo M. Lisanti M.P. J. Biol. Chem. 1996; 271: 9690-9697Google Scholar) with minor modifications. 3T3-L1 adipocytes were washed 3 times in PBS and then scraped into 2 ml of 500 mm sodium carbonate, pH 11.0. Cells were then homogenized on ice with (i) a Polytron tissue grinder on output setting 10 for 30 s (Kinematica GmbH, Brinkmann Instruments, Westbury, NY) and (ii) a sonicator on output setting 2 for 30 s and then setting 5 for 5 bursts (Branson Sonifier 250, Branson Ultrasonic Corp., Danbury, CT). The homogenate was then adjusted to 45% sucrose by addition of 2 ml of 90% sucrose prepared in MBS and placed at the bottom of an ultracentrifuge tube. A 5–35% discontinuous sucrose gradient that contained MBS and 250 mm sodium carbonate was formed above and centrifuged at 39,000 rpm for 16 h in an SW41 rotor (Beckman Instruments, Palo Alto, CA). Silica-based purification of plasma membranes and caveolae was performed as described (32Schnitzer J.E. McIntosh D.P. Dvorak A.M. Liu J. Oh P. Science. 1995; 269: 1435-1439Google Scholar). After SDS-PAGE, proteins were transferred to BA83 nitrocellulose (Schleicher & Schuell) using a Hoefer Semi-Phor apparatus. Nitrocellulose membranes were blocked in PBS or Tris-buffered saline with 0.1% Tween 20 and 5% non-fat dry milk. Primary and secondary antibodies were diluted in PBS or Tris-buffered saline with 0.1% Tween 20 and 1% bovine serum albumin. Bound antibodies were detected by enhanced chemiluminescence according to the manufacturer's instructions (DuPont NEN). Separation of proteins by SDS-PAGE (Laemmli) was performed according to standard protocols (25Laemmli U.K. Nature. 1970; 227: 680-685Google Scholar). Computer analysis of DNA and protein sequences was performed with the DNASTAR software package, BLAST search software (33Altschul S.F. Gish W. Miller W. Myers E.W. Lipman D.J. J. Mol. Biol. 1990; 215: 403-410Google Scholar), and MacPattern software (34Fuchs R. Comput. Appl. Biosci. 1994; 10: 171-178Google Scholar). Densitometry was performed with a Bio-Rad model GS-700 Imaging Densitometer and Molecular Analyst Version 2.1.1 software (Bio-Rad). Microsequence analysis of a ∼45-kDa component of purified caveolin-rich membrane domains revealed several novel peptide sequences (Table I). To clone the cDNA for this novel protein, we generated a probe by reverse transcription-PCR with degenerate primers that were designed in accordance with the microsequence data. With the resulting probe we then screened an appropriate full-length cDNA library. A ∼1.7-kilobase clone was isolated that contained an open reading frame that predicted a polypeptide of 428 amino acids with a molecular mass of 47 kDa. The most 5′-encoded methionine is indicated in Fig. 1 as the initiator methionine. However, the next methionine at amino acid position 11 has a more classic Kozak consensus sequence (G at position +4 and A or G at position −3) (35Kozak M. J. Biol. Chem. 1991; 266: 19867-19870Google Scholar). We have not yet defined experimentally which of these two ATG triplets serves as the start codon for the flotillin polypeptide. To show that the isolated cDNA clone could encode a protein, a PCR-generated construct starting with the latter methionine and tagged at the carboxyl terminus with the Myc epitope (pMVVmyc) was expressed in COS-7 cells. In a metabolic pulse-chase experiment with these transfected cells, a ∼50-kDa protein was immunoprecipitated with 9E10, a monoclonal antibody that recognizes the Myc epitope (data not shown). The cDNA for flotillin encodes a protein of 428 amino acids with a predicted molecular mass of 47 kDa, which closely matches the size of the ∼45-kDa protein that was microsequenced from caveolin-rich membrane domains. Flotillin contains two hydrophobic domains of 27 and 18 amino acids, respectively (boxed regions in Fig. 1). The first of these domains (from position 10 to 36) may represent an atypical signal peptide or perhaps a transmembrane domain. The second hydrophobic region (from position 134 to 151) is another potential transmembrane domain. Computer-assisted motif analysis (34Fuchs R. Comput. Appl. Biosci. 1994; 10: 171-178Google Scholar) reveals two potential tyrosine phosphorylation sites at positions 160 and 292. In addition, flotillin contains predicted sites for phosphorylation by protein kinase C (at positions 52, 150, 229, and 315) and protein kinase A (at position 222). However, there are no predicted N-linked glycosylation sites. The region of flotillin from 328 to 355 is predicted to form an α helix that may form a triple coiled coil with other flotillin monomers. 3E. Wolf, P. S. Kim, and B. Berger, submitted for publication. Beginning at position 47, flotillin demonstrates significant homology with ESA, for which both human and mouse cDNA clones have been reported (20Schroeder W.T. Stewart-Galetka S. Mandavilli S. Parry D.A.D. Goldsmith L. Duvic M. J. Biol. Chem. 1994; 269: 19983-19991Google Scholar, 36Cho Y.J. Chema D. Moskow J.J. Cho M. Schroeder W.T. Overbeek P. Buchberg A.M. Duvic M. Genomics. 1995; 27: 251-258Google Scholar) (Fig. 2 A). The cDNA for ESA was cloned from a human foreskin keratinocyte λgt11 cDNA library that was screened with a monoclonal antibody. This monoclonal antibody, ECS-1, had been raised against cultured human keratinocytes and was found to stain nucleated epidermal cells in an intercellular pattern. Moreover, the addition of ECS-1 to cultured mouse keratinocytes led to cell detachment, and this detachment was enhanced by complement. Based upon these data, ESA has been implicated in epidermal cell adhesion. Over the region of homology, there is ∼47% identity between flotillin and ESA at the amino acid level. Of the potential phosphorylation sites in flotillin, only the protein kinase C phosphorylation site at residue 150 and the tyrosine phosphorylation site at residue 152 are conserved in ESA. Flotillin also shares with ESA a repeat of A/G, E, A/G, E that recurs 6 times in the carboxyl half of the polypeptide (Fig. 2 A, shaded boxes). The significance of this repeat remains to be determined. The two potential membrane-spanning domains indicated in the flotillin sequence are not conserved in ESA. The first such domain (flotillin residues 10–36) is missing entirely from ESA because the initiator methionine for ESA corresponds to flotillin residue 46. The second such domain (flotillin residues 134–151 and ESA residues 89–106) contains two additional charged amino acids in ESA that make this region unlikely to be membrane spanning. Homology searches also reveal a surprising similarity between flotillin and two conceptual proteins from the cyanobacteriumSynechococcus PCC7942 (Fig. 2 B). The homology extends for nearly the entire lengths of flotillin and the ORF2 protein. Two A/G, E, A/G, E motifs are partially conserved between flotillin and the Synechococcus ORFs (shaded boxes). 24% of the residues in ORF2 protein are identical to their corresponding residues in flotillin. In the region of highest homology (residues 84–240 of flotillin) the amino acid sequence of flotillin is 30% identical to that of ORF2 and 58% identical or conservatively substituted. However, no such homologies are detected between flotillin and the 6-frame translation of the completely sequenced yeast
Referência(s)