Cellugyrin Induces Biogenesis of Synaptic-like Microvesicles in PC12 Cells

The four-transmembrane domain proteins synaptophysin and synaptogyrin represent the major constituents of synaptic vesicles. Our previous studies in PC12 cells demonstrated that synaptogyrin or its nonneuronal paralog cellugyrin targets efficiently to synaptic-like microvesicles (SLMVs) and dramatically increases the synaptophysin content of SLMVs (Belfort, G. M., and Kandror, K. V. (2003) J. Biol. Chem. 278, 47971–47978). Here, we explored the mechanism of these phenomena and found that ectopic expression of cellugyrin increases the number of SLMVs in PC12 cells. Mutagenesis studies revealed that cellugyrin's hydrophilic cytoplasmic domains are not involved in vesicle biogenesis, whereas small conserved hydrophobic hairpins in the first luminal loop and the carboxyl terminus of cellugyrin were found to be critical for the formation of SLMVs. In addition, the length but not the primary sequence of the second luminal loop was essential for SLMV biogenesis. We suggest that changing the length of this loop similar to disruption of the short hydrophobic hairpins alters the position of the vicinal transmembrane domains that may be crucial for protein function. The four-transmembrane domain proteins synaptophysin and synaptogyrin represent the major constituents of synaptic vesicles. Our previous studies in PC12 cells demonstrated that synaptogyrin or its nonneuronal paralog cellugyrin targets efficiently to synaptic-like microvesicles (SLMVs) and dramatically increases the synaptophysin content of SLMVs (Belfort, G. M., and Kandror, K. V. (2003) J. Biol. Chem. 278, 47971–47978). Here, we explored the mechanism of these phenomena and found that ectopic expression of cellugyrin increases the number of SLMVs in PC12 cells. Mutagenesis studies revealed that cellugyrin's hydrophilic cytoplasmic domains are not involved in vesicle biogenesis, whereas small conserved hydrophobic hairpins in the first luminal loop and the carboxyl terminus of cellugyrin were found to be critical for the formation of SLMVs. In addition, the length but not the primary sequence of the second luminal loop was essential for SLMV biogenesis. We suggest that changing the length of this loop similar to disruption of the short hydrophobic hairpins alters the position of the vicinal transmembrane domains that may be crucial for protein function. Understanding the mechanisms by which a cell sequesters specific proteins in unique subcellular compartments is fundamental to understanding how the cell maintains distinct organelles. For the past 25 years, the PC12 neuroendocrine cell line has provided a convenient system for studying the biogenesis of synaptic like microvesicles (SLMVs) 1The abbreviations used are: SLMV, synaptic-like microvesicle; PNS, postnuclear supernatant(s); PAS, postabsorptive supernatant(s); TE, Triton elution; SE, SDS elution; TM, transmembrane; ER, endoplasmic reticulum. (1Johnston P. Cameron P.L. Stukenbrok H. Jahn R. De Camilli P. Sudhof T.C. EMBO J. 1989; 8: 2863-2872Crossref PubMed Scopus (109) Google Scholar). SLMVs were chosen as a model organelle, because, like neuronal synaptic vesicles, their small size (∼50 nm) restricts the number of proteins present in each vesicle. Based on the estimates of 3 × 106 daltons of protein/synaptic vesicle and 50,000 daltons/average protein, ∼60 proteins could be contained in a vesicle of this size (2Jahn R. Sudhof T.C. J. Neurochem. 1993; 61: 12-21Crossref PubMed Scopus (85) Google Scholar). Two main methods have been used to study the biogenesis of SLMVs. Several groups have studied cell-free or perforated cell vesicle budding systems in order to identify cytosolic factors critical for vesicle budding. Others studying specific proteins that target to SLMVs have identified regions in their primary amino acid sequences that are required for targeting to that compartment. Two main pathways of SLMV biogenesis have emerged from in vitro studies. First, vesicles can bud from endosomes in a brefeldin A-sensitive mechanism that requires the AP3 adaptor protein and the ARF1 and Rab4 GTPases (3Faundez V. Horng J.T. Kelly R.B. J. Cell Biol. 1997; 138: 505-515Crossref PubMed Scopus (88) Google Scholar, 4Lichtenstein Y. Desnos C. Faundez V. Kelly R.B. Clift-O'Grady L. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 11223-11228Crossref PubMed Scopus (53) Google Scholar, 5Faundez V. Horng J.T. Kelly R.B. Cell. 1998; 93: 423-432Abstract Full Text Full Text PDF PubMed Scopus (247) Google Scholar, 6de Wit H. Lichtenstein Y. Kelly R.B. Geuze H.J. Klumperman J. van der Sluijs P. Mol. Biol. Cell. 2001; 12: 3703-3715Crossref PubMed Scopus (58) Google Scholar). By the second pathway, SLMVs can bud from the plasma membrane in a brefeldin A-insensitive mechanism that involves the AP2 adapter, clathrin, and dynamin (7Shi G. Faundez V. Roos J. Dell'Angelica E.C. Kelly R.B. J. Cell Biol. 1998; 143: 947-955Crossref PubMed Scopus (84) Google Scholar). Further work with perforated PC12 cells demonstrated that endophilin I, a dynamin-binding protein, facilitates SLMV budding, probably by modifying the membrane lipid composition (8Schmidt A. Wolde M. Thiele C. Fest W. Kratzin H. Podtelejnikov V. Witke W. Huttner W.B. Soling H.D. Nature. 1999; 401: 133-141Crossref PubMed Scopus (449) Google Scholar). Two categories of proteins are present in SLMVs in PC12 cells: neuronal synaptic vesicle proteins and other proteins that are not found in brain synaptic vesicles. Of those proteins belonging to the former category, synaptophysin, synaptobrevin/VAMPII, the vesicular acetylcholine transporter, and synaptotagmin I have all been at least partially analyzed for SLMV targeting sequences (9Leube R.E. J. Cell Sci. 1995; 108: 883-894Crossref PubMed Google Scholar, 10Linstedt D. Kelly R.B. J. Physiol. 1991; 85: 90-96Google Scholar, 11Grote E. Hao J.C. Bennett M.K. Kelly R.B. Cell. 1995; 81: 581-589Abstract Full Text PDF PubMed Scopus (145) Google Scholar, 12Varoqui H. Erickson J.D. J. Biol. Chem. 1998; 273: 9094-9098Abstract Full Text Full Text PDF PubMed Scopus (61) Google Scholar, 13Blagoveshchenskaya D. Hewitt E.W. Cutler D.F. Mol. Biol. Cell. 1999; 10: 3979-3990Crossref PubMed Scopus (60) Google Scholar). The group of nonsynaptic vesicle proteins localized in SLMVs can be further divided into the nonneuronal paralogs of synaptic vesicle proteins (cellubrevin, cellugyrin, and pantophysin) (14Chilcote 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 (118) Google Scholar, 15Belfort G.M. Kandror K.V. J. Biol. Chem. 2003; 278: 47971-47978Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar, 16Haass N.K. Kartenbeck M. Leube R.E. J. Cell Biol. 1996; 134: 731-746Crossref PubMed Scopus (68) Google Scholar) and those proteins that are unrelated to synaptic vesicle proteins (tyrosinase and P-selectin) (13Blagoveshchenskaya D. Hewitt E.W. Cutler D.F. Mol. Biol. Cell. 1999; 10: 3979-3990Crossref PubMed Scopus (60) Google Scholar, 17Norcott J.P. Solari R. Cutler D.F. J. Cell Biol. 1996; 134: 1229-1240Crossref PubMed Scopus (52) Google Scholar). These studies, which examined targeting motifs, failed to identify a universal sequence applicable to all SLMV proteins. The one commonality among all of the targeting motifs identified so far is that they are all located in cytoplasmic domains. The question then arises as to what mechanisms maintain the structural integrity of SLMVs, such that, despite the complexity of their trafficking pathways in the cells and apparent heterogeneity of targeting signals, these vesicles maintain uniform and specific protein composition? Undoubtedly, adaptor proteins provide a certain level of specificity upon vesicle budding from the donor membranes. However, given the fact that at least two adaptors, AP-2 and AP-3, with different binding specificities, are involved in the formation of vesicles with the same protein composition from different compartments, we speculate that additional factors should participate in the formation of SLMVs. Cellugyrin (also referred to as synaptogyrin 2) is the nonneuronal paralog of the synaptic vesicle protein, synaptogyrin 1. Both cellugyrin and synaptogyrin 1 target to SLMVs in PC12 cells, and both do so with a much higher efficiency than any other protein tested so far (15Belfort G.M. Kandror K.V. J. Biol. Chem. 2003; 278: 47971-47978Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar). Increasing the level of cellugyrin or synaptogyrin 1 in PC12 cells also causes a concomitant increase of endogenous synaptophysin in SLMVs. Here, we report that cellugyrin expression is sufficient to induce the biogenesis of SLMVs, as measured by two independent methods. Furthermore, despite the finding that cellugyrin and synaptogyrin are capable of targeting very efficiently to microvesicles in all cell types tested, the capacity to facilitate targeting of synaptophysin to microvesicles was not replicated in a nonneuronal cell line (COS7 cells). These results suggest that the capacity of cellugyrin to facilitate the targeting of synaptophysin is not an artifact of overexpression but rather represents an epiphenomenon of cellugyrin targeting to microvesicles that requires cell type-specific factors. In support of this latter hypothesis, molecular dissection of cellugyrin via recombinant DNA techniques revealed that the targeting of cellugyrin mutants to SLMVs can be uncoupled from cellugyrin's effects on synaptophysin targeting. These studies demonstrated that conserved hydrophobic sequences in the carboxyl-terminal domain and the first luminal loop, in addition to the lengths, but not the primary sequences, of the second luminal loop and the cytoplasmic loop are key determinants for efficient SLMV targeting. Furthermore, inefficiencies in the targeting of these mutants to SLMVs yielded an even more in the of synaptophysin targeting. We that the efficient targeting of cellugyrin and synaptogyrin 1 to SLMVs and their to facilitate the targeting of synaptophysin is on their capacity to increase the number of SLMVs and that this is by the regions identified in targeting studies. which synaptophysin, which cellugyrin, and which synaptogyrin were all from T. C. of which a acid was the of of and were from was from which the was from was from and was from and were from Cell was from was from DNA cellugyrin with an was (15Belfort G.M. Kandror K.V. J. Biol. Chem. 2003; 278: 47971-47978Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar). I on how the used in these studies were DNA of from 1 and from 1 and from 1 and from 1 and from 1 and from 1 and and and in a Cell and cell was in with and in at cells, a cell were in with and in at were with to the were with of and a to of 3 of The efficiency of was cell were by cells in the of for and of PC12 or were with were from the with cell for PC12 cells or for in were at × for cells were in 1 with 1 1 cells were with a cell and at × for to a postnuclear that microvesicles and cytosolic proteins was by of the at × for in a L. D. W. E. Kelly R.B. J. Cell Biol. PubMed Scopus Google Scholar). was by of to a in were at × in an for were from the via a was by 1 of with of or with of was at for The of vesicles to was then on a in were at × in an for 1 and were from the via a and of or were with were from the with in were at × for cells were in with cells were with a cell and at × for The × represents and the × represents the The × was with a of Triton in A. was tested in with the in PC12 cells and found to be and cell were as M. R.E. M. J.D. J. 1999; PubMed Google Scholar). cells were with of was and then for protein from and cells were to the same protein was were for with of or synaptophysin and with A. of or were in from or cells. referred to as postabsorptive was The were with A. Triton with Triton in was used to the from the SDS with was used to the and its from the Nature. PubMed Scopus Google Scholar). of the at and PNS, PAS, TE, and was in a In addition, PAS, TE, and were to cells with which for cellugyrin with a carboxyl a first luminal loop, and a carboxyl-terminal were and on with were then with in for with with Triton for with and with and or and by and with primary and for at and was by with were with to the The was used for was examined by and were by to and to in 25 Nature. PubMed Scopus Google Scholar). the membrane was with in for 1 at 25 and with specific The were with with and with for 1 at more the were in for 1 and then to an was with of cellugyrin was by the on the on the at J. J. Mol. Biol. PubMed Scopus Google Scholar). was used to the of the in targeting efficiency is as the of a specific protein in the to the of the same protein in the was used to the of the in the Triton from and cells. are as the with synaptophysin the with Cellugyrin the of SLMVs in the order to further the of cellugyrin, we explored the of cellugyrin expression on the of the In a previous (15Belfort G.M. Kandror K.V. J. Biol. Chem. 2003; 278: 47971-47978Abstract Full Text Full Text PDF PubMed Scopus (16) Google we that the cellugyrin in the SLMV on vesicle in was we the vesicles in that an increase in protein vesicle should increase the of the vesicles from and cells were on and for at × The of cellugyrin and synaptophysin in both and cells is and that both endogenous and cellugyrin are in the cellugyrin These that the vesicle from cells are the same as those from cells. that the vesicles are the same size and we that cellugyrin increases the number of vesicles cell and that these vesicles the of proteins. We tested this two independent we used a in involves the addition of an of a to a the SLMVs. the the to the cytoplasmic of the target protein in a are in the to more to each the vesicles are on to a and with a number of the target protein than vesicles with a they are by more specific and in more In we the of vesicles from and cells synaptophysin as the and as the In the of vesicles from both and cells with a and both to 1 and These suggest that vesicles from cells and the same number of synaptophysin that the increase in synaptophysin in the is in fact by a number of vesicles. In order to the by lipid were and or with or were in was by of the protein of the different so that they were a was and into two one for with and the other for with The with vesicular proteins were with Triton and were then with of the revealed increase in the lipid content in these vesicles and that the vesicles are the same size and this that expression of cellugyrin to an increase in the number of SLMVs. a of on for PC12 is a for microvesicles present in all cell types tested, the cell line (15Belfort G.M. Kandror K.V. J. Biol. Chem. 2003; 278: 47971-47978Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar). In order to or not can their effects in nonneuronal cells, we with in cells that synaptophysin in cells represents a which is with results of and Kelly Kelly R.B. 1991; Full Text PDF PubMed Scopus Google Scholar). expression of DNA of all proteins were similar to one in as as in PC12 cells. In with results (15Belfort G.M. Kandror K.V. J. Biol. Chem. 2003; 278: 47971-47978Abstract Full Text Full Text PDF PubMed Scopus (16) Google a of cellugyrin and synaptogyrin were found in the of vesicles in the × of both and PC12 cells, that proteins of the can facilitate formation of microvesicles in different cells. was not efficiently to microvesicles in cell In addition, cellugyrin and synaptogyrin not facilitate the targeting of synaptophysin to the in cells, as they do in PC12 cells. These results the that the of synaptophysin targeting by cellugyrin or synaptogyrin 1 represents an artifact of protein these suggest that synaptophysin targeting to the requires a or factors that are in certain cell of of of the cellugyrin protein was recombinant DNA to identify domains for the formation of SLMVs as as cellugyrin's on synaptophysin targeting in PC12 cells (15Belfort G.M. Kandror K.V. J. Biol. Chem. 2003; 278: 47971-47978Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar). We to the molecular dissection of this acid protein number on the domain and by and which that both the amino and carboxyl are located in the with one cytosolic loop and and two luminal and and and R. Sudhof T.C. J. Biol. Chem. 1998; 273: Full Text Full Text PDF PubMed Scopus Google Scholar). In targeting studies, the level of a given protein in a postnuclear for referred to as was to represent the level of whereas the level of the protein in a × for referred to as was as a of the protein in the L. D. W. E. Kelly R.B. J. Cell Biol. PubMed Scopus Google Scholar). mutants analyzed were present at than of We divided of the primary amino acid sequence into amino and carboxyl 2) loop regions loop 1 and and the cytoplasmic and transmembrane domains. studies with synaptogyrin 1 and the C. synaptogyrin that the carboxyl of those proteins were required for in and H. Mol. Biol. Cell. 2001; 12: PubMed Scopus Google Scholar). we to on the terminus of of the terminus by with a yielded an not However, we were to by a and In with the protein with the same this a in vesicle efficiency and a in its on synaptophysin targeting efficiency of function. SLMV targeting efficiency is as the of a given protein divided by the of the same protein as by efficiency for all cellugyrin mutants was to cellugyrin with the same efficiency of synaptophysin for cells with cellugyrin mutants was to synaptophysin targeting efficiency of cellugyrin with the same a in targeting efficiency in with cellugyrin with the same for the number of in which targeting efficiency was for that identified the carboxyl terminus as a domain required for cellugyrin we a of a sufficient number of amino so that the protein was of amino and by not a that in from the protein not we suggest that amino are essential for efficient SLMV formation and synaptophysin targeting. of the carboxyl terminus is hydrophobic and is conserved in all synaptogyrin paralogs and and from and of the amino terminus of the protein revealed that the first amino can be of not of the amino terminus or of amino which to in of the protein not in to the first amino is with the fact that cellugyrin is despite additional in with and We studied the luminal and cytoplasmic by the order of a or acid in each loop, the regions from transmembrane domains. were with of on cellugyrin function. The first luminal loop, amino was from the sequence to and In with cellugyrin, the protein a in vesicle efficiency and a in its on synaptophysin targeting efficiency of amino was were which amino were to 2) and The protein was in both and so was not The protein was even more but the protein from cellugyrin the of this to amino with amino more crucial than the first in of of or with not the of the protein not these suggest that hydrophobic amino in this are more critical for cellugyrin than hydrophilic amino is with the fact that synaptogyrin and 3 from and and the C. of synaptogyrin all have a short hydrophobic that and of cellugyrin's second luminal loop to on the of the protein In order to further the of this we a acid and The protein was but with cellugyrin, a in vesicle efficiency and a in its on synaptophysin targeting efficiency and cytoplasmic loop was to the of the was in a acid was into the cytoplasmic loop in with the in the second luminal In the the was with the first luminal loop to a latter was partially but was on of the to the SLMV was was in the even the was In order to this we on PC12 cells with the of expression by cells contained of the with and revealed that the the the other cells with and that the not with an SLMV the results by that cellugyrin with synaptophysin in previous studies (15Belfort G.M. Kandror K.V. J. Biol. Chem. 2003; 278: 47971-47978Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar). these suggest that the and the and then target to SLMVs, which to of the is not in the and not with PC12 cells were on and with in were then by in Triton in for were then with and for 1 at in in were then for 1 at in and in or and in all were with and for 1 at in and in The cells were with and the were with by the by the were on with the were on a we of each transmembrane domain with the with of all that this have of and In each the was to be of a transmembrane of the were The one protein, of the luminal of was from cellugyrin not the of this that the of and were and that is the specific sequences of the other transmembrane domain regions that are required for protein ectopic expression of cellugyrin or synaptogyrin 1 in PC12 cells, a of both proteins is to SLMVs. both proteins are capable of the synaptophysin content in SLMVs (15Belfort G.M. Kandror K.V. J. Biol. Chem. 2003; 278: 47971-47978Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar). that synaptogyrin 1 and synaptophysin are to for of the protein in synaptic vesicles, we how the SLMVs to this increase in revealed that ectopic expression of cellugyrin not the vesicle size in the SLMV (15Belfort G.M. Kandror K.V. J. Biol. Chem. 2003; 278: 47971-47978Abstract Full Text Full Text PDF PubMed Scopus (16) Google Scholar). the of the vesicles from cells with cellugyrin is similar to the of vesicles from cells Furthermore, we found that the number of synaptophysin vesicle in cells is the same as in cells a that can be by that are more vesicles The from the lipid provide that such an is In these studies vesicle with revealed at least a increase in the lipid content of SLMVs from cells with the In the of a in vesicle size or we that the lipid content represents vesicles. We recombinant DNA in order to those regions of cellugyrin that are for vesicle biogenesis and the increase in synaptophysin targeting. we found that of the hydrophilic regions of cellugyrin that are to the are for vesicle that cytoplasmic such as adaptor may not be for the formation of microvesicles but may be involved in and to vesicle the same two conserved short hydrophobic one in the first luminal loop and in the carboxyl terminus of cellugyrin, key in this that is critical to cellugyrin is the length than the primary of the second luminal We that changing the length of this loop, similar to or of the short hydrophobic alters the position of the vicinal transmembrane domains that may be crucial for protein the same the primary sequence of the cytoplasmic loop not to be critical for cellugyrin whereas the length of that loop is a more than the length of the second luminal report on the C. synaptogyrin that an in the cytoplasmic loop was critical for targeting to H. Mol. Biol. Cell. 2001; 12: PubMed Scopus Google Scholar). is present in cellugyrin's cytoplasmic loop as but targeting of the protein to SLMVs was despite the position of that amino acid is that targeting to in C. requires different molecular mechanisms than targeting to SLMVs. The on the luminal in with the cytoplasmic loop may suggest a for cellugyrin in the of membrane which is to represent a crucial for vesicle biogenesis R. PubMed Scopus Google Scholar). the luminal of the transmembrane domains may be required to be than the cytosolic to one cellugyrin may an of the the that may facilitate the budding from the donor model with the of an lipid (8Schmidt A. Wolde M. Thiele C. Fest W. Kratzin H. Podtelejnikov V. Witke W. Huttner W.B. Soling H.D. Nature. 1999; 401: 133-141Crossref PubMed Scopus (449) Google which is to to the vesicle. the luminal are to be 3 as as cellugyrin's luminal the synaptophysin may not have a and not induce membrane and vesicle The of the loop domains in the by transmembrane domain proteins is by the fact that, whereas the sequences of the of all vesicle proteins are the of these regions are conserved each and K. R. H. Leube R.E. PubMed Scopus Google Scholar). of cellugyrin and synaptophysin are whereas cellugyrin and the in this cytoplasmic loop, are and not Mutagenesis to the of cellugyrin mutants in SLMV targeting and of synaptophysin targeting and is that despite the different of each in the primary the in SLMV targeting and of synaptophysin targeting are similar all a in SLMV targeting with a in synaptophysin that as the targeting capacity of cellugyrin is by the of synaptophysin is more model for these results by that the mechanism for synaptophysin into the vesicle on efficient vesicle biogenesis. M. D. Mol. Biol. Cell. 2003; PubMed Scopus Google Scholar) that synaptophysin facilitates targeting of to synaptic vesicles in finding may the in the of synaptic vesicle synaptogyrin synaptophysin, in to the vesicle. These studies may also on a report that the of a in the synaptogyrin 1 with R. C. C. S. Biol. Full Text Full Text PDF PubMed Scopus Google Scholar). We may speculate that this may be to a in synaptic vesicle biogenesis In we have identified regions in the cellugyrin We suggest that these regions are critical for the of the transmembrane domains that, in induce membrane which to the formation of SLMVs in PC12 cells. Furthermore, on studies in cells, we suggest that the of cellugyrin in vesicles may not on the cell further of this with proteins on targeting Further studies be the of the cellugyrin and the cytoplasmic budding We T. C. for the of the