Portal de Periódicos da CAPES

Cellugyrin and Synaptogyrin Facilitate Targeting of Synaptophysin to a Ubiquitous Synaptic Vesicle-sized Compartment in PC12 Cells

2003; Elsevier BV; Volume: 278; Issue: 48 Linguagem: Inglês

10.1074/jbc.m304174200

1083-351X

Gabriel M. Belfort, Konstantin V. Kandror,

Erythrocyte Function and Pathophysiology

Cellugyrin represents a ubiquitously expressed four-transmembrane domain protein that is closely related to synaptic vesicle protein synaptogyrin and, more remotely, to synaptophysin. We report here that, in PC12 cells, cellugyrin is localized in synaptic-like microvesicles (SLMVs), along with synaptogyrin and synaptophysin. Upon overexpression of synaptophysin in PC12 cells, it is localized in rapidly sedimenting membranes and practically is not delivered to the SLMVs. On the contrary, the efficiency of the SLMV targeting of exogenously expressed cellugyrin and synaptogyrin is high. Moreover, expression of cellugyrin (or synaptogyrin) in PC12 cells dramatically and specifically increases SLMV targeting of endogenous synaptophysin. Finally, we utilized the SLMV purification scheme on a series of non-neuroendocrine cell types including the mouse fibroblast cell line 3T3-L1, the Chinese hamster ovary cell line CHO-K1, and the monkey kidney epithelial cell line COS7 and found that a cellugyrin-positive microvesicular compartment was present in all cell types tested. We suggest that synaptic vesicles have evolved from cellugyrin-positive ubiquitous microvesicles and that neuroendocrine SLMVs represent a step along that pathway of evolution. Cellugyrin represents a ubiquitously expressed four-transmembrane domain protein that is closely related to synaptic vesicle protein synaptogyrin and, more remotely, to synaptophysin. We report here that, in PC12 cells, cellugyrin is localized in synaptic-like microvesicles (SLMVs), along with synaptogyrin and synaptophysin. Upon overexpression of synaptophysin in PC12 cells, it is localized in rapidly sedimenting membranes and practically is not delivered to the SLMVs. On the contrary, the efficiency of the SLMV targeting of exogenously expressed cellugyrin and synaptogyrin is high. Moreover, expression of cellugyrin (or synaptogyrin) in PC12 cells dramatically and specifically increases SLMV targeting of endogenous synaptophysin. Finally, we utilized the SLMV purification scheme on a series of non-neuroendocrine cell types including the mouse fibroblast cell line 3T3-L1, the Chinese hamster ovary cell line CHO-K1, and the monkey kidney epithelial cell line COS7 and found that a cellugyrin-positive microvesicular compartment was present in all cell types tested. We suggest that synaptic vesicles have evolved from cellugyrin-positive ubiquitous microvesicles and that neuroendocrine SLMVs represent a step along that pathway of evolution. Elucidating the mechanisms that determine the localization of proteins in specific sub-cellular compartments is critical to understanding how the cell organizes functionally distinct organelles. Because of their small size (∼50 nm), synaptic vesicles (SVs) 1The abbreviations used are: SVsynaptic vesicleSLMVsynaptic-like microvesicleCHOChinese hamster ovaryGFPgreen fluorescent proteinDMEMDulbecco's modified Eagle's mediumPBSphosphate-buffered salinePNSpostnuclear supernatantS2second supernatantPVDFpolyvinylidene difluoride.1The abbreviations used are: SVsynaptic vesicleSLMVsynaptic-like microvesicleCHOChinese hamster ovaryGFPgreen fluorescent proteinDMEMDulbecco's modified Eagle's mediumPBSphosphate-buffered salinePNSpostnuclear supernatantS2second supernatantPVDFpolyvinylidene difluoride. and their neuroendocrine counterpart, synaptic-like microvesicles (SLMVs), present an opportunity for studying a compartment with a relatively small number of proteins (1Fernandez-Chacon R. Sudhof T.C. Annu. Rev. Physiol. 1999; 61: 753-776Crossref PubMed Scopus (153) Google Scholar). The neuroendocrine cell line, PC12, has been shown to generate SLMVs that are similar to SVs in size, density, and many marker proteins (2Navone F. Jahn R. Di Gioia G. Stukenbrok H. Greengard P. De Camilli P. J. Cell Biol. 1986; 103: 2511-2527Crossref PubMed Scopus (557) Google Scholar, 3Lowe A.W. Madeddu L. Kelly R.B. J. Cell Biol. 1988; 106: 51-59Crossref PubMed Scopus (146) Google Scholar, 4Cameron P.L. Sudhof T.C. Jahn R. DeCamilli P. J. Cell Biol. 1991; 115: 151-164Crossref PubMed Scopus (237) Google Scholar, 5Clift-O'Grady L. Linstedt A.D. Lowe A.W. Grote E. Kelly R.B. J. Cell Biol. 1990; 110: 1693-1703Crossref PubMed Scopus (146) Google Scholar). Thus PC12 cells have been used as an in vitro model for studying the biogenesis of SVs/SLMVs. Using this cell type, targeting signals have been identified in the amino acid sequence of the synaptic vesicle proteins VAMPII/synaptobrevin (6Grote E. Hao J.C. Bennet M.K. Kelly R.B. Cell. 1995; 81: 581-589Abstract Full Text PDF PubMed Scopus (144) Google Scholar) and synaptotagmin (7Blagoveshchenskaya A.D. Hewitt E.W. Cutler D.F. Mol. Biol. Cell. 1999; 10: 3979-3990Crossref PubMed Scopus (59) Google Scholar). These studies, however, have not identified a universal targeting signal. synaptic vesicle synaptic-like microvesicle Chinese hamster ovary green fluorescent protein Dulbecco's modified Eagle's medium phosphate-buffered saline postnuclear supernatant second supernatant polyvinylidene difluoride. synaptic vesicle synaptic-like microvesicle Chinese hamster ovary green fluorescent protein Dulbecco's modified Eagle's medium phosphate-buffered saline postnuclear supernatant second supernatant polyvinylidene difluoride. In addition to the work done with PC12 cells, several groups have studied the sub-cellular localization of synaptic vesicle proteins exogenously expressed in non-neuroendocrine cell lines. Transfection of cDNA encoding the four-transmembrane protein, synaptophysin, into CHO cells leads to the expression of intact protein. However, whether synaptophysin is segregated away from early endosomal markers in a microvesicular compartment, as is the case in PC12 cells, has been the subject of much debate. Several groups have presented data that indicate that synaptophysin is co-localized in endosomes with the transferrin and low density lipoprotein receptors, both of which are excluded from PC12 SLMV (4Cameron P.L. Sudhof T.C. Jahn R. DeCamilli P. J. Cell Biol. 1991; 115: 151-164Crossref PubMed Scopus (237) Google Scholar, 8Johnston P.A. Cameron P.L. Stukenbrok H. Jahn R. De Camilli P. Sudhof T.C. EMBO J. 1989; 8: 2863-2872Crossref PubMed Scopus (106) Google Scholar, 9Linstedt A.D. Kelly R.B. Neuron. 1991; 7: 309-317Abstract Full Text PDF PubMed Scopus (100) Google Scholar, 10Feany M.B. Yee A.G. Delvy M.L. Buckley K.M. J. Cell Biol. 1993; 123: 575-584Crossref PubMed Scopus (61) Google Scholar). Targeting of ectopically expressed synaptophysin to microvesicles has been demonstrated in other non-neuronal cell types, including human breast (MCF-7) and liver (PLC) carcinoma cell lines, as well as rat smooth muscle (RVF-SMC) and bovine mammary gland cell lines (11Leube R.E. Wiedenmann B. Franke W.W. Cell. 1989; 59: 433-446Abstract Full Text PDF PubMed Scopus (68) Google Scholar, 12Leube R.E. Leimer U. Grund C. Franke W.W. Harth N. Wiedenmann B. J. Cell Biol. 1994; 127: 1589-1601Crossref PubMed Scopus (38) Google Scholar). In addition, Leube and co-workers (13Haass N.K. Kartenbeck M.A. Leube R.E. J. Cell Biol. 1996; 134: 731-746Crossref PubMed Scopus (68) Google Scholar) have shown that overexpressed pantophysin, a non-neuronal paralog of synaptophysin, was localized to microvesicles in PLC cells. These latter studies contain strong electron microscopic evidence suggesting the presence of a microvesicular compartment in non-neuronal cells. However, definitive biochemical studies using an endogenous marker for this compartment have so far been lacking. Synaptophysin and pantophysin belong to the family of tetraspan vesicle membrane proteins. In addition to the physins, this family includes the gyrins and the secretory carrier-associated membrane proteins (14Hubner K. Windoffer R. Hutter H. Leube R.E. Int. Rev. Cytol. 2002; 214: 103-159Crossref PubMed Scopus (47) Google Scholar). Cellugyrin/synaptogyrin 2 is a non-neuronal four-transmembrane paralog of the synaptic vesicle protein, synaptogyrin 1 (15Janz R. Sudhof T.C. J. Biol. Chem. 1998; 273: 2851-2857Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar, 16Kedra D. Pan H.-Q. Seroussi E. Fransson I. Guilbaud C. Collins J.E. Dunham I. Blennow E. Roe B.A. Piehl F. Dumanski J.P. Hum. Genet. 1998; 103: 131-141Crossref PubMed Scopus (46) Google Scholar). Synaptogyrin 1 has been shown to participate in Ca2+-dependent exocytosis and synaptic plasticity (17Janz R. Sudhof T.C. Hammer R.E. Unni V. Siegelbaum S.A. Bolshakov V.Y. Neuron. 1999; 24: 687-700Abstract Full Text Full Text PDF PubMed Scopus (264) Google Scholar, 18Sugita S. Janz R. Sudhof T.C. J. Biol. Chem. 1999; 274: 18893-18901Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar). Despite these findings it is generally agreed that a function for the large family of tetraspan vesicle membrane proteins has yet to be elucidated (14Hubner K. Windoffer R. Hutter H. Leube R.E. Int. Rev. Cytol. 2002; 214: 103-159Crossref PubMed Scopus (47) Google Scholar). Previously, our laboratory has shown that cellugyrin participates in the trafficking of the insulin responsive glucose transporter (Glut4) in adipose cells (19Kupriyanova T.A. Kandror K.V. J. Biol. Chem. 2000; 275: 36263-36268Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar). Specifically, cellugyrin populates those vesicles shuttling Glut4 between sorting and recycling endosomes but is excluded from the larger Glut4 vesicles that translocate to the plasma membrane in response to insulin (20Kupriyanova T.A. Kandror V. Kandror K.V. J. Biol. Chem. 2002; 277: 9133-9138Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). Cellugyrin-positive vesicles are present in nonhomogenized adipocytes and are formed in vitro in a cytosol-, ATP-, and time-dependent manner, suggesting that these vesicles are not a product of homogenization (21Xu Z. Kandror K.V. J. Biol. Chem. 2002; 277: 47972-47975Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar). In the present study, we have examined the sub-cellular localization of cellugyrin in diverse cell types. We show that cellugyrin is a marker of ubiquitous microvesicles present in all species tested indistinguishable based on velocity sedimentation properties from the SLMVs in PC12 cells. Furthermore, initiation of these studies led to the observation that cellugyrin, and its neuronal paralog, synaptogyrin, enhances the efficiency of synaptophysin targeting to SLMVs in PC12 cells in a robust and specific fashion. At the same time, ectopically expressed synaptophysin did not alter targeting of cellugyrin to SLMVs. These findings may shed light on the mechanism of SV biogenesis from a ubiquitous vesicular carrier. We suggest that the production of microvesicles is an evolutionarily ancient mechanism that is co-opted by neurons for the production of synaptic vesicles. Thus, microvesicle targeting of exogenously expressed synaptic vesicle proteins in non-neuronal systems represents an interaction between the neuronal paralog and the ubiquitous system of sorting such proteins. According to this model cellugyrin, and probably other tetraspan vesicle membrane proteins, may provide a core set of component proteins required for a minimal vesicle that can serve as a matrix for assembly of specialized vesicular carriers in differentiated cells. Materials—pBJ1-Cg-GFP, which expresses mouse cellugyrin with a C-terminal green fluorescent protein tag, was the kind gift of Dr. Phillip Reay (Oxford, United Kingdom). pCMV-Cgyr, which expresses rat cellugyrin, pCMV5-p38, which expresses rat synaptophysin, and pCMV5-p29, which expresses rat synaptogyrin, were all kind gifts from Dr. T. C. Südhof (University of Texas at South Western). pcDNA3.1His(+) and pcDNA3.1/myc-His(+) were purchased from Invitrogen. Three anti-cellugyrin antibodies were used: 1) monoclonal antibody used for immunostaining, immunoadsorption, and vesicle decoration/shift analysis (19Kupriyanova T.A. Kandror K.V. J. Biol. Chem. 2000; 275: 36263-36268Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar); 2) monoclonal antibody from BD Biosciences used for Western blotting; and 3) affinity-purified rabbit polyclonal antibody raised against the C-terminal 16 amino acids of rat cellugyrin used for Western blot analysis (21Xu Z. Kandror K.V. J. Biol. Chem. 2002; 277: 47972-47975Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar). Anti-transferrin receptor monoclonal (Zymed Laboratories Inc.) was used for Western blot analysis. Anti-insulin responsive amino peptidase rabbit polyclonal was the kind gift of Dr. P. Pilch (Boston University School of Medicine). Two different anti-synaptophysin monoclonal antibodies were used for immunoadsorption (Chemicon), vesicle decoration/shift analysis (Chemicon), and Western blot analysis (Chemicon and BD Biosciences). Anti-SV2 monoclonal was the kind gift of Dr. K. Buckley (Harvard Medical School). Also used for Western blotting were anti-synaptogyrin polyclonal (Synaptic Systems), anti-synaptobrevin/VAMPII monoclonal (Synaptic Systems), anti-GFP monoclonal (Santa Cruz Biotechnology, Inc.), anti-tubulin-β monoclonal (Sigma), and anti-HisG monoclonal (Invitrogen). The antibody used in each experiment is identified in the figure legends. Cell Culture—PC12 cells, a pheochromocytoma cell line, were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 5% fetal bovine serum, 2 mm l-glutamine, 50 μg/ml gentamicin, 100 units/ml penicillin, and 100 μg/ml streptomycin in 5% CO2 at 37 °C. Chinese hamster ovary cells (CHO-K1) were grown in DMEM supplemented with 10% fetal bovine serum, 2 mm l-glutamine, 40 μg/ml proline, 50 μg/ml gentamicin, 100 units/ml penicillin, and 100 μg/ml streptomycin in 5% CO2 at 37 °C. Monkey kidney cells (COS7) were grown in DMEM supplemented with 10% fetal bovine serum, 2 mm l-glutamine, 50 μg/ml gentamicin, 100 units/ml penicillin, and 100 μg/ml streptomycin in 5% CO2 at 37 °C. Mouse fibroblasts (3T3-L1) were grown in DMEM supplemented with 10% fetal calf serum, 2 mm l-glutamine, 100 units/ml penicillin, and 100 μg/ml streptomycin in 10% CO2 at 37 °C. Isolation and Fractionation Procedures—Confluent 15- or 10-cm plates were rinsed once with PBS. Cells were lifted with cell dissociation media (Sigma) for PC12 cells, trypsin/EDTA (Invitrogen) for COS7 and CHO-K1 cells, or by scraping for 3T3-L1 cells. Cells were then pelleted at 400 × g for 7 min. Cells were resuspended in Buffer A (150 mm NaCl, 10 mm HEPES, pH 7.4, 1 mm EGTA, 0.1 mm MgCl2) with protease inhibitors (1 μm aprotinin, 5 μm benzamidine, 2 μm leupeptin, 1 μm pepstatin, 1 μm phenylmethylsulfonyl fluoride). Resuspended cells were homogenized with 11 strokes through a ball bearing cell cracker (European Molecular Biology Laboratory) and centrifuged at 300 × g for 5 min to generate a postnuclear supernatant (PNS). A second supernatant (S2) that contains only microvesicles was generated by centrifugation of the PNS at 27,000 × g (15,000 rpm) for 35 min in a Ti42.2 rotor (Beckman) (5Clift-O'Grady L. Linstedt A.D. Lowe A.W. Grote E. Kelly R.B. J. Cell Biol. 1990; 110: 1693-1703Crossref PubMed Scopus (146) Google Scholar). Velocity gradient fractionation was performed by layering 500–1000 μg of PNS or S2 onto one of two types of gradients; the majority of experiments were performed using 4.6 ml of 10–25% glycerol in Buffer A (v/v) gradients centrifuged for 55 min. Also used was a 4.2-ml 5–25% glycerol in Buffer A (v/v) gradient with a 400-μl 50% sucrose in Buffer A (w/v) pad centrifuged for 1 h (5Clift-O'Grady L. Linstedt A.D. Lowe A.W. Grote E. Kelly R.B. J. Cell Biol. 1990; 110: 1693-1703Crossref PubMed Scopus (146) Google Scholar). All velocity gradients were centrifuged in an SW55 rotor (Beckman) at 280,000 × g (48,000 rpm). Fractionation was from the bottom via a peristaltic pump. Density gradient fractionation was performed by layering 500–1000 μg of S2 onto a 4.6-ml 10–50% sucrose in Buffer A (w/v) gradient. These gradients were centrifuged at 280,000 × g (48,000 rpm) in an SW55 rotor (Beckman) for 16 h. Fractionation was from the bottom via a peristaltic pump. Vesicle decoration/shift analysis was performed by incubating 1 mg of S2 with 2.5 μg of purified nonspecific mouse IgG, anti-cellugyrin monoclonal, or anti-synaptophysin monoclonal antibodies, together with 2 μg of nanogold conjugated goat anti-mouse Fab fragments (Nanoprobes). The mixture was rotated at 4 °C for 2 h. The total volume of decorated vesicles was then loaded on 4.6-ml linear 10–30% sucrose in Buffer A (w/v). Gradients were centrifuged at 280,000 × g (48,000 rpm) in an SW55 rotor (Beckman) for 1 h 15 min. Fractionation was from the bottom via a peristaltic pump. Immunofluorescence—PC12 cells transfected with pBJ1-Cg-GFP were lifted and grown on coverslips coated with poly-l-lysine over night. Next, cells were fixed with 4% paraformaldehyde in PBS for 30 min, then washed with PBS, permeabilized with 0.2% Triton X-100 for 5 min, blocked with 4% donkey serum, and stained with monoclonal anti-synaptophysin or anti-synaptogyrin antibodies followed by Rhodamine RedX-conjugated donkey anti-mouse IgG (Jackson ImmunoResearch). Each incubation with antibody lasted for 60 min at room temperature and was followed by six quick rinses with PBS. A SlowFade-Light Antifade kit (Molecular Probes) was used for mounting cells on slides. Staining was examined by confocal laser scanning microscopy (LSM510; Carl Zeiss Inc.). Gel Electrophoresis and Immunoblotting—Proteins were separated by SDS-PAGE according to Laemmli (22Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (206631) Google Scholar) and transferred to a PVDF membrane in 25 mm Tris, 192 mm glycine. Following transfer, the membrane was blocked with 10% nonfat dry milk in PBS for 1 h at 25 °C and probed with specific antibodies overnight. The membranes were washed three times with phosphate-buffered saline with 0.05% Tween and incubated with horseradish peroxidase-labeled secondary antibody for 1 h at room temperature. After three more washes the membranes were incubated in ECL reagent (PerkinElmer Life Sciences) for 1 min and then exposed to an Eastman Kodak Co. 440 image station. Data analysis was performed with Kodak 1D image analysis software. Statistical Analysis—The student's unpaired two-tailed t test was used to evaluate the statistical significance of the differences in targeting efficiencies. Targeting efficiency is defined as the amount of a specific protein in the S2 fraction normalized by the total amount of the same protein in the PNS. Cellugyrin Is Localized in Synaptic-like Microvesicles in PC12 Cells—We have examined cellugyrin compartmentalization in PC12 cells. After homogenization, a PNS was obtained by centrifugation for 5 min at 300 × g. Separation of the PNS on two different types of glycerol velocity gradient followed by Western blot analysis of gradient fractions revealed a peak of synaptic-like microvesicles marked by synaptophysin and synaptogyrin in fractions 7–13 (Fig. 1, A and B) (4Cameron P.L. Sudhof T.C. Jahn R. DeCamilli P. J. Cell Biol. 1991; 115: 151-164Crossref PubMed Scopus (237) Google Scholar, 5Clift-O'Grady L. Linstedt A.D. Lowe A.W. Grote E. Kelly R.B. J. Cell Biol. 1990; 110: 1693-1703Crossref PubMed Scopus (146) Google Scholar). Probing of these membranes with anti-cellugyrin antibodies showed that cellugyrin co-localizes with the synaptic vesicle markers in this gradient. In addition, exogenously expressed mouse cellugyrin with a C-terminal green fluorescent protein tag (mouse cellugyrin-GFP) co-localized with the synaptic vesicle markers. It is important to note that the peak of small vesicles is clearly distinct from the endosomal marker the transferrin receptor and the insulin responsive amino peptidase, the latter having been characterized as a marker of a larger vesicular carrier (23Thoidis G. Kandror K.V. Traffic. 2001; 2: 577-587Crossref PubMed Scopus (20) Google Scholar). In the following experiments, PC12 SLMVs were separated from heavier membrane compartments recovered at the bottom of the glycerol velocity gradient (Fig. 1A) by centrifugation for 35 min at 27,000 × g as described previously (4Cameron P.L. Sudhof T.C. Jahn R. DeCamilli P. J. Cell Biol. 1991; 115: 151-164Crossref PubMed Scopus (237) Google Scholar, 5Clift-O'Grady L. Linstedt A.D. Lowe A.W. Grote E. Kelly R.B. J. Cell Biol. 1990; 110: 1693-1703Crossref PubMed Scopus (146) Google Scholar). This high speed supernatant (S2) was layered onto a 10–50% sucrose density gradient, and the gradient fractions were analyzed by Western blot analysis (Fig. 2). Again, cellugyrin co-localized with synaptic vesicle markers (SV2, synaptophysin, and synaptobrevin) in the SLMV. Immunoadsorptions of intact vesicles from S2 were carried out to determine whether cellugyrin is localized in the same compartment with synaptic vesicle markers (Fig. 3). S2 was incubated with magnetic beads coated with monoclonal antibodies and the postabsorptive supernatants and SDS elutions were subjected to Western blot analysis. Immunoadsorption with cellugyrin antibodies precipitated all of the synaptophysin, synaptogyrin, and VAMPII signal from S2. Furthermore, immunoadsorption with synaptophysin antibodies precipitated the entire endogenous cellugyrin signal from S2. Control immunoadsorption with nonspecific mouse IgG did not precipitate any of the cellugyrin or synaptic vesicle markers signals from S2. Thus immunoadsorption experiments confirm the results from velocity and density gradient fractionation and directly suggest co-localization of cellugyrin and synaptic vesicle markers in small vesicles. Despite the results obtained from immunoadsorption, it is still possible that different vesicular compartments exist in PC12 cells. For example, two putative vesicular populations may have similar velocity and density gradient characteristics but contain different sets of component proteins. If one imagined the protein content of hypothetical vesicle A to be composed of 99% synaptophysin and 1% cellugyrin and the protein content of hypothetical vesicle B to be composed of 99% cellugyrin and 1% synaptophysin, then these populations would remain indistinguishable by immunoadsorption, because both vesicle populations do contain both proteins and would likely be precipitated by either antibody. To determine the presence of different vesicle populations, we turned to a method known as vesicle decoration/shift analysis (20Kupriyanova T.A. Kandror V. Kandror K.V. J. Biol. Chem. 2002; 277: 9133-9138Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). Briefly, vesicles were decorated by a monoclonal antibody against a cytoplasmic epitope of one of the vesicle proteins. Additional mass is added to the monoclonal antibody by the co-incubation of nanogold conjugated anti-mouse Fab fragments. The decorated vesicles were layered onto a 10–30% sucrose velocity gradient, and the gradient fractions were analyzed by Western blot analysis (Fig. 4). In Fig. 4A, the baseline sedimentation of this vesicular population is determined after incubation with nonspecific IgG. In Fig. 4B, decoration with anti-cellugyrin antibody has resulted in a shift of the peak of vesicular markers by four fractions toward the bottom of the gradient. Most significantly the alignment of cellugyrin with the rest of synaptic vesicle markers is maintained. In Fig. 4C, decoration with anti-synaptophysin antibody has resulted in a shift of the peak of vesicular markers by eight fractions toward the bottom of the gradient suggesting that synaptophysin is more abundant in the SLMVs than cellugyrin. In this case, the peak of cellugyrin also co-localizes with the synaptic vesicle markers. Results of this experiment directly demonstrate that synaptophysin and cellugyrin are present, although in different amounts, in a single type of vesicle population. Immunofluorescence staining of PC12 cells was used as an independent method for examining the sub-cellular distribution of cellugyrin and synaptophysin (Fig. 5). PC12 cells transfected with mouse cellugyrin-GFP were fixed, permeabilized, and probed with rat-specific anti-cellugyrin antibody. This showed co-localization of the endogenous (rat) and the exogenous (mouse) proteins (data not shown). Next, we probed similarly transfected cells with antibodies against synaptophysin and synaptogyrin. The staining in Fig. 4 shows that transfected cellugyrin co-localizes with these synaptic markers in PC12 cells. Cellugyrin and Synaptogyrin Specifically Increase the Targeting Efficiency of Synaptophysin to SLMV—To examine the contribution of cellugyrin to the biogenesis of microvesicles, we decided to test the effects of increased cellugyrin expression. Varying doses of either His- or Myc/His-tagged cellugyrin were transfected into PC12 cells, followed by isolation of PNS and S2, which were then subjected to SDS-PAGE and Western blot analysis (Fig. 6, A and B). Based on Figs. 1 and 2, as well as previously published reports, the expression level of the construct in the PNS was taken as a measure of total exogenous protein, whereas the expression level in S2 was taken as a measure of protein in microvesicles (6Grote E. Hao J.C. Bennet M.K. Kelly R.B. Cell. 1995; 81: 581-589Abstract Full Text PDF PubMed Scopus (144) Google Scholar). Note that PNS and S2 differ only very slightly (5–10%) in protein concentration, and thus loading equal protein in these experiments is tantamount to loading equal starting material. This experiment led to the discovery that transfection of cellugyrin leads to a dose-dependent increase of synaptophysin in the microvesicle fraction (S2) (p < 0.0005) without changing its total expression level. To ensure that the increased synaptophysin levels found in S2 were in fact in SLMV, we subjected S2 fractions from empty vector and His-cellugyrin-transfected cells to glycerol gradient centrifugation (Fig. 7). This experiment revealed that the increased levels of synaptophysin seen in the S2 fraction of cellugyrin cells are in a vesicular compartment with the same sedimentation characteristics as SLMV. Next, we compared the levels of synaptophysin in S2 from cells transfected with His-cellugyrin or His-synaptophysin (Fig. 8, top panel). The transfection efficiency of His-synaptophysin and His-cellugyrin was similar as their signals in the PNS are equivalent. In agreement with data shown in Figs. 6 and 7, we found that transfection of cellugyrin dramatically increases the amount of endogenous synaptophysin in S2, whereas transfection of synaptophysin does not dramatically change the level of cellugyrin in this fraction. A high level of expression of His-synaptophysin only leads to a moderate increase of this protein in S2, so that the total synaptophysin signal in S2 was much lower than in cells transfected with His-cellugyrin. In other words, synaptophysin SLMV targeting efficiency is increased more by cellugyrin expression than by expression of synaptophysin itself. This is consistent with data that show that the majority of synaptophysin in non-transfected PC12 cells is located in compartments that are larger than the SLMV (Fig. 1, A and B) and may represent donor membranes from which SLMVs originate (24Desnos C. Clift-O'Grady L. Kelly R.B. J. Cell Biol. 1995; 130: 1041-1049Crossref PubMed Scopus (61) Google Scholar, 25Faundez V. Horng J.T. Kelly R.B. J. Cell Biol. 1997; 138: 505-515Crossref PubMed Scopus (88) Google Scholar, 26Shi G. Faundez V. Roos J. Dell'Angelica E.C. Kelly R.B. J. Cell Biol. 1998; 143: 947-955Crossref PubMed Scopus (83) Google Scholar). Next we wanted to know whether the effect of cellugyrin on synaptophysin is a general function of the gyrin class of proteins or if it is specific to cellugyrin. To test this, rat synaptogyrin was transfected into PC12 cells after which PNS and S2 were subjected to Western blot analysis (Fig. 9). This figure demonstrates that increased targeting efficiency of synaptophysin is equally inducible by cellugyrin or synaptogyrin. We suggest, therefore, that targeting of synaptophysin to SLMV requires the presence of cellugyrin or synaptogyrin that radically increases the efficiency of synaptophysin targeting into SLMV. Cellugyrin Is Present in a Synaptic Vesicle-sized Compartment in All Cell Types Tested—Because cellugyrin is expressed in all tissues (15Janz R. Sudhof T.C. J. Biol. Chem. 1998; 273: 2851-2857Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar), we explored the compartmentalization of cellugyrin in diverse cell types (Fig. 10). Sub-cellular fractionation was performed on the mouse fibroblast cell line 3T3-L1, the Chinese hamster ovary cell line CHO-K1, and the monkey kidney epithelial cell line COS7. A high speed supernatant (S2) from each cell type, prepared in the same way as for PC12 cells, was separated in a 10–25% glycerol gradient. Fractions from these gradients were subjected to Western blot analysis. A clear cellugyrin peak is present in the same location as the SLMV of PC12. Furthermore, His-cellugyrin transfected into COS7 cells also sedimented to the same location on the glycerol gradient. Together, these data suggest that cellugyrin is present in microvesicles in all cell types. Synaptic vesicles were once considered to be the smallest type of vesicular carrier and an adaptation unique to the nervous system. However, Johnston et al. (8Johnston P.A. Cameron P.L. Stukenbrok H. Jahn R. De Camilli P. Sudhof T.C. EMBO J. 1989; 8: 2863-2872Crossref PubMed Scopus (106) Google Scholar) reported that exogenously expressed synaptophysin targets to microvesicles in CHO cells. The fact that a non-neuronal cell type was capable of sorting a synaptic vesicle protein to a sub-cellular compartment similar in size to synaptic vesicles suggested that synaptic vesicles evolved from this ubiquitous microvesicle (8Johnston P.A. Cameron P.L. Stukenbrok H. Jahn R. De Camilli P. Sudhof T.C. EMBO J. 1989; 8: 2863-2872Crossref PubMed Scopus (106) Google Scholar). Subsequently, studies using the electron microscope have identified pantophysin, a non-neuronal paralog of synaptophysin, as a marker of ubiquitous microvesicles of similar size (<90 nm) (13Haass N.K. Kartenbeck M.A. Leube R.E. J. Cell Biol. 1996; 134: 731-746Crossref PubMed Scopus (68) Google Scholar). However, ubiquitous microvesicles have never been isolated and characterized biochemically using a non-overexpressed endogenous marker protein. Moreover, it is not clear if these vesicles are present in non-transfected cells or represent an artifact of protein overexpression. The data presented in this paper provide the first biochemical evidence for the existence of a ubiquitous microvesicle compartment using cellugyrin as an endogenous marker protein. Furthermore, the suggested evolutionary relationship between ubiquitous microvesicles and synaptic vesicles (8Johnston P.A. Cameron P.L. Stukenbrok H. Jahn R. De Camilli P. Sudhof T.C. EMBO J. 1989; 8: 2863-2872Crossref PubMed Scopus (106) Google Scholar) gains weight from the observation that the expression level of cellugyrin, a non-neuronal protein, has functional effects on the synaptic vesicle protein synaptophysin. Specifically, the expression level of cellugyrin has a direct effect on the targeting efficiency of synaptophysin to SLMVs. The use of PC12 cell SLMVs as a model system for the biosynthesis of synaptic vesicles has been a matter of some controversy. On the one hand, this model has been useful for elucidating SLMV targeting sequences for synaptic vesicle proteins, such as synaptobrevin and synaptotagmin (6Grote E. Hao J.C. Bennet M.K. Kelly R.B. Cell. 1995; 81: 581-589Abstract Full Text PDF PubMed Scopus (144) Google Scholar, 7Blagoveshchenskaya A.D. Hewitt E.W. Cutler D.F. Mol. Biol. Cell. 1999; 10: 3979-3990Crossref PubMed Scopus (59) Google Scholar). It is assumed that SLMV targeting sequences in PC12 cells would correspond to synaptic vesicle targeting sequences in neurons. On the other hand, two classes of non-neuronal proteins also target to SLMVs, the first being non-neuronal paralogs of synaptic vesicle proteins, like cellubrevin/VAMPIII (6Grote E. Hao J.C. Bennet M.K. Kelly R.B. Cell. 1995; 81: 581-589Abstract Full Text PDF PubMed Scopus (144) Google Scholar, 27Chilcote T.J. Galli T. Mundigl O. Edelmann L. McPherson P.S. Takei K. De Camilli P. J. Cell Biol. 1995; 129: 219-231Crossref PubMed Scopus (117) Google Scholar) and cellugyrin as reported here. The second class of non-neuronal proteins that target to PC12 SLMVs have no neuronal counterpart. Examples of this class of protein are tyrosinase and p-selectin, which are recognized by the AP3 adaptor, part of the SLMV budding machinery (7Blagoveshchenskaya A.D. Hewitt E.W. Cutler D.F. Mol. Biol. Cell. 1999; 10: 3979-3990Crossref PubMed Scopus (59) Google Scholar, 28Blagoveshchenskaya A.D. Hewitt E.W. Cutler D.F. J. Cell Biol. 1999; 145: 1419-1433Crossref PubMed Scopus (40) Google Scholar, 29Blagoveshchenskaya A.D. Cutler D.F. Mol. Biol. Cell. 2000; 11: 1801-1814Crossref PubMed Scopus (21) Google Scholar). A possibility exists that the SLMVs in PC12 cells represents a mixture of real synaptic vesicles and ubiquitous microvesicles. To address this question we performed velocity and density gradient centrifugation in an attempt to identify two distinct vesicular pools based on physical characteristics. The results of these experiments indicate that if two pools exist, they are inseparable based on physical parameters (Figs. 1 and 2). Next, we used immunoadsorption to examine whether cellugyrin is co-localized in PC12 SLMVs with the synaptic vesicle marker proteins. Immunoadsorption of cellugyrin containing membranes precipitated all of the synaptic vesicle markers and immunoadsorption of synaptophysin containing membranes precipitated all of the cellugyrin (Fig. 3). Finally, vesicle decoration/shift analysis provides a method for directly determining whether there are distinguishable populations of vesicles. Co-localization of cellugyrin with the synaptic vesicle markers, synaptophysin, synaptogyrin, and synaptobrevin (Fig. 4) in these experiments strongly suggest that that there is only one pool of microvesicles in PC12 cells, namely SLMVs, that contain both neuronal and non-neuronal proteins. We then turned to examine the targeting of different proteins into this compartment. Targeting efficiency in this paper was defined as the ratio of a given protein in microvesicles (S2) to the total protein (PNS). We believe that it is more accurate to normalize data by the total amount of starting material because of reported interactions among the various proteins that might be exaggerated or alternatively bypassed by manipulation of the system (30Edelmann L. Hanson P.I. Chapman E.R. Jahn R. EMBO J. 1995; 14: 224-231Crossref PubMed Scopus (387) Google Scholar). The "targeting index" used for VAMP-TAg constructs (6Grote E. Hao J.C. Bennet M.K. Kelly R.B. Cell. 1995; 81: 581-589Abstract Full Text PDF PubMed Scopus (144) Google Scholar), which normalizes all VAMP-TAg values by dividing by synaptophysin content (as a measure of vesicle number), cannot be used for cellugyrin, because cellugyrin directly affects the amount of synaptophysin in small vesicles. We do not yet know whether the increase in synaptophysin represents increased copy number per vesicle or whether the transfection of cellugyrin increases the number of vesicles. Irrespectively, it is clear that the amount of cellugyrin and/or synaptogyrin in a cell can have an effect on the targeting of synaptophysin to microvesicles. This observation may shed light on the longstanding controversy regarding synaptophysin targeting in non-neuronal cells. Although some groups claim that ectopically expressed synaptophysin in CHO cells is found in larger endosomal vesicles (4Cameron P.L. Sudhof T.C. Jahn R. DeCamilli P. J. Cell Biol. 1991; 115: 151-164Crossref PubMed Scopus (237) Google Scholar, 9Linstedt A.D. Kelly R.B. Neuron. 1991; 7: 309-317Abstract Full Text PDF PubMed Scopus (100) Google Scholar), others have argued that the same protein can be found in microvesicles in the hepatocellular carcinoma cell line PLC cells (11Leube R.E. Wiedenmann B. Franke W.W. Cell. 1989; 59: 433-446Abstract Full Text PDF PubMed Scopus (68) Google Scholar, 12Leube R.E. Leimer U. Grund C. Franke W.W. Harth N. Wiedenmann B. J. Cell Biol. 1994; 127: 1589-1601Crossref PubMed Scopus (38) Google Scholar). We suggest that these two cell lines express different levels of cellugyrin and thus have different capacities for packaging synaptophysin into microvesicles. In other words, the high level of exogenous synaptophysin targeting to microvesicles in PLC cells may be a direct result of the high level of cellugyrin expression in hepatocytes. Finally, we examined the sub-cellular distribution of cellugyrin in a series of non-neuroendocrine cells. Using the same isolation technique used for PC12 SLMVs microvesicles from mouse fibroblasts (3T3-L1), hamster ovary cells (CHO-K1), and monkey kidney cells (COS7) were isolated. Western analysis of fractions from glycerol velocity gradient centrifugation of these diverse samples indicates that there is a microvesicular pool with the same sedimentation characteristics as PC12 SLMVs in all of these cell types with cellugyrin as a marker protein. Given that all cell types tested contain a cellugyrin-positive microvesicular compartment and that cellugyrin expression levels have a direct impact on synaptophysin targeting efficiency to PC12 cell SLMVs, we propose that cellugyrin-positive small vesicles represent an evolutionary predecessor to synaptic vesicles. One can further speculate that PC12 SLMVs represent an intermediate stage on the evolution from ubiquitous microvesicles to synaptic vesicles. From this perspective, the targeting of non-neuronal proteins to this compartment represents an ancient mechanism, whereas the targeting of neuronal proteins represents the evolutionarily more recent utilization and adaptation of those ancient mechanisms. We thank Dr. T. Südhof (University of Texas at South Western) and Dr. P. Reay (Oxford, United Kingdom) for providing critical reagents.