Caveolin Isoforms Differ in Their N-terminal Protein Sequence and Subcellular Distribution. IDENTIFICATION AND EPITOPE MAPPING OF AN ISOFORM-SPECIFIC MONOCLONAL ANTIBODY PROBE
1995; Elsevier BV; Volume: 270; Issue: 27 Linguagem: Inglês
10.1074/jbc.270.27.16395
ISSN1083-351X
AutoresPhilipp E. Scherer, Zhao Lan Tang, Miyoung Chun, Massimo Sargiacomo, Harvey F. Lodish, Michael P. Lisanti,
Tópico(s)Erythrocyte Function and Pathophysiology
ResumoCaveolin, an integral membrane protein, is a principal component of caveolae membranes in vivo. Two isoforms of caveolin have been identified: a slower migrating 24-kDa species (α-isoform) and a faster migrating 21-kDa species (β-isoform). Little is known about how these isoforms differ, either structurally or functionally. Here we have begun to study the differences between these two isoforms. Microsequencing of caveolin reveals that both isoforms contain internal caveolin residues 47-77. In a second independent approach, we recombinantly expressed caveolin in a caveolin-negative cell line (FRT cells). Stable transfection of FRT cells with the full-length caveolin cDNA resulted in the expression of both caveolin isoforms, indicating that they can be derived from a single cDNA. Using extracts from caveolin-expressing FRT cells, we fortuitously identified a monoclonal antibody that recognizes only the α-isoform of caveolin. Epitope mapping of this monoclonal antibody reveals that it recognizes an epitope within the extreme N terminus of caveolin, specifically residues 1-21. These results suggest that α- and β-isoforms of caveolin differ in their N-terminal protein sequences. To independently evaluate this possibility, we placed an epitope tag at either the extreme N or C terminus of full-length caveolin. Results of these “tagging” experiments clearly demonstrate that (i) both isoforms of caveolin contain a complete C terminus and (ii) that the α-isoform contains a complete N terminus while the β-isoform lacks N-terminal-specific protein sequences. Mutational analysis reveals that these two isoforms apparently derive from the use of two alternate start sites: methionine at position 1 and an internal methionine at position 32. This would explain the ~3-kDa difference in their apparent migration in SDS-polyacrylamide electrophoresis gels. In addition, using isoform-specific antibody probes we show that caveolin isoforms may assume a distinct but overlapping subcellular distribution by confocal immunofluorescence microscopy. We discuss the possible implications of these differences between α- and β-caveolin. Caveolin, an integral membrane protein, is a principal component of caveolae membranes in vivo. Two isoforms of caveolin have been identified: a slower migrating 24-kDa species (α-isoform) and a faster migrating 21-kDa species (β-isoform). Little is known about how these isoforms differ, either structurally or functionally. Here we have begun to study the differences between these two isoforms. Microsequencing of caveolin reveals that both isoforms contain internal caveolin residues 47-77. In a second independent approach, we recombinantly expressed caveolin in a caveolin-negative cell line (FRT cells). Stable transfection of FRT cells with the full-length caveolin cDNA resulted in the expression of both caveolin isoforms, indicating that they can be derived from a single cDNA. Using extracts from caveolin-expressing FRT cells, we fortuitously identified a monoclonal antibody that recognizes only the α-isoform of caveolin. Epitope mapping of this monoclonal antibody reveals that it recognizes an epitope within the extreme N terminus of caveolin, specifically residues 1-21. These results suggest that α- and β-isoforms of caveolin differ in their N-terminal protein sequences. To independently evaluate this possibility, we placed an epitope tag at either the extreme N or C terminus of full-length caveolin. Results of these “tagging” experiments clearly demonstrate that (i) both isoforms of caveolin contain a complete C terminus and (ii) that the α-isoform contains a complete N terminus while the β-isoform lacks N-terminal-specific protein sequences. Mutational analysis reveals that these two isoforms apparently derive from the use of two alternate start sites: methionine at position 1 and an internal methionine at position 32. This would explain the ~3-kDa difference in their apparent migration in SDS-polyacrylamide electrophoresis gels. In addition, using isoform-specific antibody probes we show that caveolin isoforms may assume a distinct but overlapping subcellular distribution by confocal immunofluorescence microscopy. We discuss the possible implications of these differences between α- and β-caveolin. Caveolae are small flask-shaped invaginations located at or near the plasma membrane(1Severs N.J. J. Cell Sci. 1988; 90: 341-348Crossref PubMed Google Scholar, 2Anderson R.G.W. Curr. Opin. Cell Biol. 1993; 5: 647-652Crossref PubMed Scopus (171) Google Scholar). They are thought to exist in most cell types, although they are most abundant in endothelial cells, type I pneumocytes, adipocytes, fibroblasts, and smooth muscle cells (reviewed in Refs. 3Lisanti M.P. Scherer P. Tang Z.-L. Sargiacomo M. Trends Cell Biol. 1994; 4: 231-235Abstract Full Text PDF PubMed Scopus (590) Google Scholar and 4Lisanti M.P. Scherer P.E. Tang Z.-L. Kubler E. Koleske A.J. Sargiacomo M.S. Semin. Dev. Biol. 1995; 6: 47-58Crossref Scopus (31) Google Scholar). Functionally, caveolae are involved in the uptake of small molecules such as folate (5Kamen B.A. Smith A.K. Anderson R.G.W. J. Clin. Invest. 1991; 87: 1442Crossref PubMed Scopus (106) Google Scholar) and the transport of macromolecules across capillary endothelial cells, including modified atherogenic low density lipoproteins(6Vasile E. Simionescu M. Simionescu N. J. Cell Biol. 1983; 96: 1677-1689Crossref PubMed Scopus (245) Google Scholar, 7Snelting-Havinga I. Mommaas M. Van Hinsbergh V. Daha M. Daems W. Vermeer B. Eur. J. Cell Biol. 1989; 1989: 27-36Google Scholar). Recently, we and others have proposed that caveolae may also participate in a subset of transmembrane signaling events, such as G-protein-coupled signaling(3Lisanti M.P. Scherer P. Tang Z.-L. Sargiacomo M. Trends Cell Biol. 1994; 4: 231-235Abstract Full Text PDF PubMed Scopus (590) Google Scholar, 8Sargiacomo M. Sudol M. Tang Z.-L. Lisanti M.P. J. Cell Biol. 1993; 122: 789-807Crossref PubMed Scopus (863) Google Scholar, 9Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z.-L. Hermanoski-Vosatka A. Tu Y.-H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Crossref PubMed Scopus (815) Google Scholar, 10Chang W.J. Ying Y. Rothberg K. Hooper N. Turner A. Gambliel H. De Gunzburg J. Mumby S. Gilman A. Anderson R.G.W. J. Cell Biol. 1994; 126: 127-138Crossref PubMed Scopus (311) Google Scholar, 11Shenoy-Scaria A.M. Dietzen D.J. Kwong J. Link D.C. Lublin D.M. J. Cell Biol. 1994; 126: 353-363Crossref PubMed Scopus (343) Google Scholar). Caveolin, a 21-24-kDa integral membrane protein, has been identified as a principal component of caveolae membranes in vivo(12Rothberg K.G. Heuser J.E. Donzell W.C. Ying Y. Glenney J.R. Anderson R.G.W. Cell. 1992; 68: 673-682Abstract Full Text PDF PubMed Scopus (1868) Google Scholar). However, caveolin was first identified as a major v-src substrate in Rous sarcoma virus-transformed chick embryo fibroblasts (13Glenney J.R. Zokas L. J. Cell Biol. 1989; 108: 2401-2408Crossref PubMed Scopus (360) Google Scholar). Both cell transformation and tyrosine phosphorylation of caveolin are dependent on membrane attachment of v-src(14Glenney J.R. J. Biol. Chem. 1989; 264: 20163-20166Abstract Full Text PDF PubMed Google Scholar), suggesting that caveolin may represent a critical substrate for cellular transformation. In support of this view, we have recently observed that both caveolin expression and caveolae are lost during cell transformation by activated oncogenes other than v-src (v-abl, bcr-abl, middle T antigen, and activated ras)(15Koleske A.J. Baltimore D. Lisanti M.P. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 1381-1385Crossref PubMed Scopus (472) Google Scholar). These results support the hypothesis that caveolin may represent a candidate tumor suppressor protein(15Koleske A.J. Baltimore D. Lisanti M.P. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 1381-1385Crossref PubMed Scopus (472) Google Scholar). Indeed, Krev-1, a Ras-related transformation suppressor protein, is concentrated in purified caveolin-rich membrane domains (9Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z.-L. Hermanoski-Vosatka A. Tu Y.-H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Crossref PubMed Scopus (815) Google Scholar) and purified caveolae(10Chang W.J. Ying Y. Rothberg K. Hooper N. Turner A. Gambliel H. De Gunzburg J. Mumby S. Gilman A. Anderson R.G.W. J. Cell Biol. 1994; 126: 127-138Crossref PubMed Scopus (311) Google Scholar). Two major isoforms of caveolin are known to exist: a slower migrating 24-kDa species and a faster migrating 21-kDa species. For simplicity, we will designate them as α- and β-isoforms of caveolin, respectively. Little is known about how these isoforms differ. This may be important for understanding the role of caveolin and caveolae in normal and transformed cells. Caveolin is the product of a single gene(16Glenney J.R. FEBS Lett. 1992; 314: 45-48Crossref PubMed Scopus (189) Google Scholar). Furthermore, as caveolin mRNA is a single species, it is unlikely that these two isoforms arise from differential mRNA splicing(9Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z.-L. Hermanoski-Vosatka A. Tu Y.-H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Crossref PubMed Scopus (815) Google Scholar, 16Glenney J.R. FEBS Lett. 1992; 314: 45-48Crossref PubMed Scopus (189) Google Scholar, 17Scherer P.E. Lisanti M.P. Baldini G. Sargiacomo M. Corley-Mastick C. Lodish H.F. J. Cell Biol. 1994; 127: 1233-1243Crossref PubMed Scopus (356) Google Scholar). As both isoforms are present immediately after caveolin synthesis, it does not appear that there is a precursor-product relationship between them(18Lisanti M.P. Tang Z.-L. Sargiacomo M. J. Cell Biol. 1993; 123: 595-604Crossref PubMed Scopus (160) Google Scholar, 19Kurzchalia T. Dupree P. Parton R.G. Kellner R. Virta H. Lehnert M. Simons K. J. Cell Biol. 1992; 118: 1003-1014Crossref PubMed Scopus (464) Google Scholar). Caveolin is constitutively phosphorylated on serine residues(8Sargiacomo M. Sudol M. Tang Z.-L. Lisanti M.P. J. Cell Biol. 1993; 122: 789-807Crossref PubMed Scopus (863) Google Scholar, 14Glenney J.R. J. Biol. Chem. 1989; 264: 20163-20166Abstract Full Text PDF PubMed Google Scholar, 20Sargiacomo M. Scherer P.E. Tang Z.-L. Casanova J.E. Lisanti M.P. Oncogene. 1994; 9: 2589-2595PubMed Google Scholar). Interestingly, only the β-isoform is phosphorylated in vivo, while both forms are capable of undergoing serine phosphorylation in vitro(17Scherer P.E. Lisanti M.P. Baldini G. Sargiacomo M. Corley-Mastick C. Lodish H.F. J. Cell Biol. 1994; 127: 1233-1243Crossref PubMed Scopus (356) Google Scholar). These observations point to a functional difference that appears to be recognized by a specific serine kinase in vivo. Importantly, these observations demonstrate that the β-isoform is not an artifactual degradation product of the α-isoform, as only the faster migrating β-isoform is selectively phosphorylated in vivo. Here, we have begun to study the differences between these two caveolin isoforms. Over 20 different monoclonal antibodies have been generated against native chicken caveolin by Glenney and co-workers(13Glenney J.R. Zokas L. J. Cell Biol. 1989; 108: 2401-2408Crossref PubMed Scopus (360) Google Scholar, 14Glenney J.R. J. Biol. Chem. 1989; 264: 20163-20166Abstract Full Text PDF PubMed Google Scholar); only a few of these are reactive with canine and human caveolin, despite the fact that caveolin is highly conserved from chicken to man (over 86% identical(16Glenney J.R. FEBS Lett. 1992; 314: 45-48Crossref PubMed Scopus (189) Google Scholar)). Among these cross-reactive antibodies we have fortuitously identified an α-isoform-specific monoclonal antibody. Identification and epitope mapping of this monoclonal antibody has provided both structural and functional information regarding the differences between α- and β-caveolin. In addition, using this isoform-specific mAb, we show by confocal fluorescence microscopy that α-caveolin may assume a distinct subcellular distribution. This antibody, which is commercially available, should prove to be a powerful molecular probe for studying the function of caveolin isoforms. Monoclonal antibodies directed against full-length caveolin were the generous gift J. R. Glenney (Transduction Laboratories, Lexington, KY). The monoclonal antibody 9E10 was provided by The Harvard Monoclonal Antibody Facility (Cambridge, MA). A variety of other reagents were purchased commercially: fetal bovine serum (JRH Biosciences), prestained protein markers (Life Technologies, Inc.), Lab-Tek chamber slides (Nunc Inc., Naperville, IL), normal goat and donkey IgGs, fluorescein isothiocyanate-conjugated goat anti-mouse antibody, and lissamine rhodamine B sulfonyl chloride-conjugated donkey anti-rabbit antibody (Jackson Immunoresearch Laboratory, West Grove, PA); SlowFade™ anti-fade reagent (Molecular Probes). Caveolin-rich domains were purified from murine lung tissue, as described(9Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z.-L. Hermanoski-Vosatka A. Tu Y.-H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Crossref PubMed Scopus (815) Google Scholar, 21Lisanti M.P. Tang Z.-T. Scherer P. Sargiacomo M. Methods Enzymol. 1995; 250: 655-668Crossref PubMed Scopus (117) Google Scholar). To further enrich for caveolin, these domains were resuspended on ice in Mes1( 1The abbreviations used are: Mes4-morpholineethanesulfonic acidMDCKMadin-Darby canine kidney cellsPCRpolymerase chain reactionmAbmonoclonal antibodyPAGEpolyacrylamide gel electrophoresisPBSphosphate-buffered saline.) -buffered saline (25 mM Mes, pH 6.5, 0.15 M NaCl) containing 60 mM octyl glucoside, adjusted to 40% sucrose, and refloated using the same type of bottom-loaded sucrose density gradients (a 5-30% linear sucrose gradient) used for initial purification. The single light-scattering band floating at the 15-20% sucrose region was collected, diluted 1:3 with Mes-buffered saline alone and recovered by centrifugation in the Microfuge (14,000 × g for 15 min at 4°C). This additional purification step eliminated other 21-24-kDa proteins that were previously found to co-purify with caveolin. Caveolin isoforms and proteins co-purifying with caveolin under these conditions were identified by microsequence analysis. Microsequencing was performed as we described previously(9Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z.-L. Hermanoski-Vosatka A. Tu Y.-H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Crossref PubMed Scopus (815) Google Scholar). 4-morpholineethanesulfonic acid Madin-Darby canine kidney cells polymerase chain reaction monoclonal antibody polyacrylamide gel electrophoresis phosphate-buffered saline. FRT and MDCK cells were propagated as described previously(8Sargiacomo M. Sudol M. Tang Z.-L. Lisanti M.P. J. Cell Biol. 1993; 122: 789-807Crossref PubMed Scopus (863) Google Scholar, 22Zurzolo C. van't Hof W. van Meer G. Rodriguez-Boulan E. EMBO J. 1994; 13: 42-53Crossref PubMed Scopus (140) Google Scholar). Cells were subjected to steady-state metabolic labeling with 32S-labeled amino acids (methionine and cysteine) as described(23Lisanti M.P. Le Bivic A. Sargiacomo M. Rodriguez-Boulan E. J. Cell Biol. 1989; 109: 2117-2127Crossref PubMed Scopus (100) Google Scholar). The expression levels of a given transfected antigen was increased by an overnight incubation with normal medium containing 10 mM sodium butyrate (21Lisanti M.P. Tang Z.-T. Scherer P. Sargiacomo M. Methods Enzymol. 1995; 250: 655-668Crossref PubMed Scopus (117) Google Scholar, 24Lisanti M.P. Caras I.W. Davitz M.A. Rodriguez-Boulan E. J. Cell Biol. 1989; 109: 2145-2156Crossref PubMed Scopus (375) Google Scholar). Wild type full-length caveolin and epitope-tagged forms of caveolin were subcloned into the multiple cloning site (HindIII/BamHI) of the vector pCB7 (containing the hygroR marker; gift of J. Casanova, MGH) for expression in FRT or MDCK cells, respectively. In order to recombinantly express epitope-tagged forms of caveolin in MDCK cells, we incorporated the myc epitope tag into the N terminus (MEQKLISEEDLNGG-caveolin) or the C terminus (caveolin-GGEQKLISEEDLN) of the cloned canine caveolin cDNA using PCR primers. We placed GG as a spacer between the epitope and the caveolin coding sequences, as has been suggested previously(19Kurzchalia T. Dupree P. Parton R.G. Kellner R. Virta H. Lehnert M. Simons K. J. Cell Biol. 1992; 118: 1003-1014Crossref PubMed Scopus (464) Google Scholar, 25Kolodziej P.A. Young R.A. Methods Enzymol. 1991; 194: 508-519Crossref PubMed Scopus (423) Google Scholar). Correct placement of the epitope tag and caveolin coding sequences were verified by double-stranded DNA sequencing. FRT or MDCK cells were stably transfected using a modification of the calcium-phosphate precipitation procedure(21Lisanti M.P. Tang Z.-T. Scherer P. Sargiacomo M. Methods Enzymol. 1995; 250: 655-668Crossref PubMed Scopus (117) Google Scholar, 24Lisanti M.P. Caras I.W. Davitz M.A. Rodriguez-Boulan E. J. Cell Biol. 1989; 109: 2145-2156Crossref PubMed Scopus (375) Google Scholar). After selection in media supplemented with 400 μg/ml hygromycin B, resistant colonies were picked by trypsinization using cloning rings. Individual clones were screened by immunofluorescence for recombinant expression of caveolin. Wild type full-length caveolin expressed in FRT cells was detected using anti-caveolin IgG (mAb 2297 or 2234). Epitope-tagged forms of caveolin expressed in MDCK cells were detected using monoclonal antibody, 9E10, that recognizes the myc epitope (EQKLISEEDLN). Methionine residues at either position 1 or at position 32 were mutated to valine by changing ATG to GTG. Single base mutations (A to G) were introduced into the carboxyl-terminally myc-tagged canine caveolin cDNA by PCR site-directed mutagenesis(26Landt O. Grunert H.-P. Hahn U. Gene (Amst.). 1990; 96: 125-128Crossref PubMed Scopus (639) Google Scholar). After subcloning into the pCB7 vector, the corresponding constructs were transiently transfected into Cos7 cells by the DEAE-dextran method(27Seed B. Aruffo A. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 3365-3369Crossref PubMed Scopus (789) Google Scholar). 48 h post-transfection, cells were scraped into lysis buffer (20 mM Tris, pH 8.0, 150 mM NaCl, 1% Triton X-100, 60 mM octyl glucoside, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride). Insoluble material was removed by centrifugation at 15,000 × g for 10 min, and the supernatant was diluted with an equal volume of SDS-PAGE sample buffer containing 100 mM NaOH and boiled for 5 min. The equivalent of 5% of a 10-cm dish of Cos7 cells was analyzed by SDS-PAGE (14% acrylamide) followed by Western blotting with mAb 9E10. Full-length caveolin (residues 1-178), each caveolin subdomain (N terminus (residues 1-101), the transmembrane domain (residues 102-134), and the C terminus (residues 135-178)), and deletion mutants of the N-terminal domain (residues 1-21, 1-41, 1-61, 1-81, 25-58, 61-101, 71-101, 76-101, 81-101) were amplified by PCR and separately subcloned into the multiple cloning site of the vector pGEX-4T-1 to obtain GST-fusion proteins when expressed in a suitable Escherichia coli strain (BL21, lacking lon and ompT proteases; Novagen, Inc.). The exact reading frame and caveolin coding sequences were verified by double-stranded DNA sequencing. GST-fusion proteins were purified by affinity chromatography using glutathione agarose, as described previously(28Frangioni J.V. Neel B.G. Anal. Biochem. 1993; 210: 179-187Crossref PubMed Scopus (832) Google Scholar). A peptide encoding the entire canine caveolin C-terminal domain (residues 135-178) was synthesized, purified, and used to immunize rabbits with RIBI's adjuvant following a standard protocol(29Sawin K. Mitchison T. Wordeman L. J. Cell Sci. 1992; 101: 303-313PubMed Google Scholar). Test bleeds were taken at 0, 5, 7, 9, and 16 weeks and monitored for antibody titer by immunoprecipitation of metabolically labeled MDCK lysates. Anti-caveolin IgGs were purified using protein A-Sepharose as described by the manufacturer. FRT cells (grown in Lab-Tek chamber slides at a subconfluent density) were washed three times with PBS and fixed for 45 min in PBS containing 3% paraformaldehyde, 10 mM NaIO4, and 70 mM lysine-HCl. Fixed cells were rinsed with PBS and treated with 100 mM NH4Cl in PBS for 10 min to quench free aldehyde groups. Cells were then permeabilized with 0.1% Triton X-100 for 10 min, either at room temperature or on ice, and washed with PBS, four times, 10 min each. The cells were then successively incubated with PBS, 2% bovine serum albumin containing: (i) 50 μg/ml each of normal goat and donkey IgGs, (ii) a 1:300 dilution of anti-caveolin mAb 2234 and 40 μg/ml anti-caveolin C-terminal-specific polyclonal IgG, and (iii) fluorescein isothiocyanate-conjugated goat anti-mouse antibody (5 μg/ml) and lissamine rhodamine B sulfonyl chloride-conjugated donkey anti-rabbit antibody (5 μg/ml). The first incubation was 30 min, while primary and secondary antibody reactions were 60 min each. Cells were washed three time with PBS between incubations. Slides were mounted with SlowFade anti-fade reagent and observed under a Bio-Rad MR600 confocal fluorescence microscope. Two major isoforms of caveolin have been identified by one- or two-dimensional gel electrophoresis of cell extracts, including MDCK cells (Fig. 1A). In order to investigate the difference between these two isoforms, we partially purified caveolin from murine lung tissue, a caveolin-rich tissue source, and subjected these isoforms to microsequence analysis (Fig. 1B). N-terminal microsequencing revealed that the N terminus of each isoform is blocked. We next performed internal microsequencing with the endoproteinase Lys-C. Sequencing of these peptides, detailed in Table I, reveals that both caveolin isoforms contain internal caveolin residues 47-77. In addition, α-caveolin appears to contain a complete N-terminal domain (caveolin residues 5-26; Table I), and β-caveolin appears to contain a complete C-terminal domain (caveolin residues 165-176; Table I). Thus, the faster migrating β-isoform might still lack a complete N-terminal domain.TABLE IInternal microsequencing of caveolin isoforms Open table in a new tab FRT cells fail to express detectable levels of caveolin mRNA or protein, although they contain caveolae-like structures(8Sargiacomo M. Sudol M. Tang Z.-L. Lisanti M.P. J. Cell Biol. 1993; 122: 789-807Crossref PubMed Scopus (863) Google Scholar, 22Zurzolo C. van't Hof W. van Meer G. Rodriguez-Boulan E. EMBO J. 1994; 13: 42-53Crossref PubMed Scopus (140) Google Scholar). As a consequence, we stably expressed full-length caveolin in FRT cells to study its properties. Our results indicate that both caveolin isoforms derive from a single cDNA. This conclusion is consistent with studies demonstrating a single species of caveolin mRNA (9Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z.-L. Hermanoski-Vosatka A. Tu Y.-H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Crossref PubMed Scopus (815) Google Scholar, 16Glenney J.R. FEBS Lett. 1992; 314: 45-48Crossref PubMed Scopus (189) Google Scholar, 17Scherer P.E. Lisanti M.P. Baldini G. Sargiacomo M. Corley-Mastick C. Lodish H.F. J. Cell Biol. 1994; 127: 1233-1243Crossref PubMed Scopus (356) Google Scholar) and that in vitro translation produces both isoforms(19Kurzchalia T. Dupree P. Parton R.G. Kellner R. Virta H. Lehnert M. Simons K. J. Cell Biol. 1992; 118: 1003-1014Crossref PubMed Scopus (464) Google Scholar). However, these authors did not rule out the possibility that this is an artifact of the in vitro translation system. To examine this possibility, we derived stable cell lines. Fig. 2 shows FRT cells stably transfected with the full-length canine caveolin cDNA. Immunoblotting with monoclonal antibody 2297 reveals both α- and β-caveolin in FRT transfectants (Fig. 2A). However, we noticed that immunoblotting of the same cell lysate with another monoclonal antibody, 2234, only revealed α-caveolin (Fig. 2B). These results suggest that monoclonal 2234 is α-specific and recognizes an epitope that is absent from β-caveolin. Thus, monoclonal antibody 2234 could be used as a molecular probe to structurally and functionally distinguish between α- and β-caveolin. We next used epitope mapping to understand why monoclonal 2234 might only recognize α-caveolin. As the substrate for antibody binding, we expressed full-length caveolin and portions of caveolin as GST-fusion proteins. After purification, fusion proteins were subjected to immunoblotting with either monoclonal antibody (2234 or 2297) (Fig. 3A). Based on their differential immunoreactivity, mAb 2234 recognizes an epitope within caveolin residues 1-21, while mAb 2297 recognizes an epitope within caveolin residues 61-71 (summarized schematically in Fig. 3B). As mAb 2297 identifies both caveolin isoforms, these results are consistent with the above microsequence analysis demonstrating that both isoforms contain caveolin residues 47-77. In addition, these results directly show that these two isoforms differ by a discrete N-terminal epitope. This might reflect the addition of a post-translational modification that renders β-caveolin nonreactive to mAb 2234 or, more likely, the absence of N-terminal protein sequence. To monitor the completeness of the N- or C-terminal ends of caveolin independently of mAb 2234, we next expressed epitope-tagged forms of caveolin in MDCK cells. Using PCR, we placed a myc epitope tag (EQKLISEEDLN) at either the extreme N or C terminus of full-length caveolin (Fig. 4). Caveolin isoforms can then be visualized after recombinant expression using a monoclonal antibody (9E10) that is directed against the myc epitope. Consistent with previous topology studies demonstrating that both the N- and C-terminal domains of caveolin face the cytoplasm(30Dupree P. Parton R.G. Raposo G. Kurzchalia T.V. Simons K. EMBO J. 1993; 12: 1597-1605Crossref PubMed Scopus (403) Google Scholar, 31Dietzen D.J. Hastings W.R. Lublin D.M. J. Biol. Chem. 1995; 270: 6838-6842Abstract Full Text Full Text PDF PubMed Scopus (289) Google Scholar), immunofluorescent detection of either N- or C-terminally tagged caveolin with mAb 9E10 required detergent permeabilization (not shown). Expression of caveolin containing a C-terminal tag yielded both caveolin isoforms when visualized with an antibody that recognizes the myc epitope (Fig. 4, A and B). These results clearly demonstrate that both isoforms contain a complete C terminus. In contrast, N-terminally tagged caveolin yielded only the α-isoform when visualized using an antibody that recognizes the myc epitope. However, both isoforms were expressed in these cells transfected with N-terminally tagged caveolin, as visualized with mAb 2297 (Fig. 4, A and B). These results independently demonstrate that (i) caveolin isoforms derive from a single cDNA and (ii) that the α-isoform contains a complete N terminus while the β-isoform lacks N-terminal-specific protein sequences, as it fails to contain the myc epitope. These studies directly support our results from epitope mapping of mAb 2234. Although we have determined that caveolin isoforms differ in N-terminal protein sequence, it is unclear how these two isoforms are generated. One possibility is that α-caveolin undergoes rapid co-translational or post-translational proteolytic processing by a specific aminopeptidase or an endopeptidase to generate β-caveolin. This would be supported if a precursor-product relationship existed between α- and β-caveolin. Although we cannot completely exclude this possibility, it appears unlikely as both caveolin isoforms are present immediately after caveolin synthesis(8Sargiacomo M. Sudol M. Tang Z.-L. Lisanti M.P. J. Cell Biol. 1993; 122: 789-807Crossref PubMed Scopus (863) Google Scholar, 19Kurzchalia T. Dupree P. Parton R.G. Kellner R. Virta H. Lehnert M. Simons K. J. Cell Biol. 1992; 118: 1003-1014Crossref PubMed Scopus (464) Google Scholar). Another possibility is alternate initiation of translation. Caveolin contains a second methionine residue at position 32 that is conserved in every caveolin cDNA sequenced to date (see Ref. 32 for an alignment). This methionine might function as an internal start site during initiation of translation. This would yield two caveolin isoforms immediately after synthesis that differ by ~3 kDa, and the smaller isoform would lack specific N-terminal protein sequences such as residues 1-21, as we observe. Alternate initiation of translation from a single mRNA transcript is used to generate two isoforms of other proteins, the progesterone receptor and the yeast gene MOD5(33Conneely O. Maxwell B. Toft D. Schrader W. O'Malley B. Biochem. Biophys. Res. Commun. 1987; 149: 493-501Crossref PubMed Scopus (186) Google Scholar, 34Slusher L. Gillman E. Martin N. Hopper A. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 9789-9793Crossref PubMed Scopus (105) Google Scholar). The use of a given methionine as a translational start site is determined in part by surrounding mRNA sequences, known as the Kozak sequence(35Kozak M. J. Biol. Chem. 1991; 266: 19867-19870Abstract Full Text PDF PubMed Google Scholar). Kozak analysis of sequences surrounding methionine 32 reveals that this residue could act as an internal start site. The methionine at position 32 in caveolin matches the Kozak consensus sequence for translation initiation even more closely than methionine at position 1 (Fig. 5), especially at the two most critical nucleotide positions. Fig. 4C shows that methionine 32 acts as an internal start site to generate caveolin isoforms. For these experiments, we used C-terminally myc-tagged caveolin and recombinant expression in COS cells. To evaluate the relative importance of methionine 32 in generating caveolin isoforms, we mutated the meth
Referência(s)