Pantophysin Is a Phosphoprotein Component of Adipocyte Transport Vesicles and Associates with GLUT4-containing Vesicles*

Pantophysin, a protein related to the neuroendocrine-specific synaptophysin, recently has been identified in non-neuronal tissues. In the present study, Northern blots showed that pantophysin mRNA was abundant in adipose tissue and increased during adipogenesis of 3T3-L1 cells. Immunoblot analysis of subcellular fractions showed pantophysin present exclusively in membrane fractions and relatively evenly distributed in the plasma membrane and internal membrane fractions. Sucrose gradient ultracentrifugation demonstrated that pantophysin and GLUT4 exhibited overlapping distribution profiles. Furthermore, immunopurified GLUT4 vesicles contained pantophysin, and both GLUT4 and pantophysin were depleted from this vesicle population following treatment with insulin. Additionally, a subpopulation of immunopurified pantophysin vesicles contained insulin-responsive GLUT4. Consistent with the interaction of synaptophysin with vesicle-associated membrane protein 2 in neuroendocrine tissues, pantophysin associated with vesicle-associated membrane protein 2 in adipocytes. Furthermore, in [32P]orthophosphate-labeled cells, pantophysin was phosphorylated in the basal state. This phosphorylation was unchanged in response to insulin; however, insulin stimulated the phosphorylation of a 77-kDa protein associated with α-pantophysin immunoprecipitates. Although the functional role of pantophysin in vesicle trafficking is unclear, its presence on GLUT4 vesicles is consistent with the emerging role of solubleN-ethylmaleimide-sensitive protein receptor (SNARE) factor complex and related proteins in regulated vesicle transport in adipocytes. In addition, pantophysin may provide a marker for the analysis of other vesicles in adipocytes.

ingly, however, adipocytes secrete a variety of other proteins such as adipsin, complement factors C3 and B, the complement-related protein Acrp30, tumor necrosis factor-␣, leptin, and others (3)(4)(5)(6). Despite the relatively large volume of vesicle traffic in adipocytes, the pathways and molecular components governing exocytosis of these factors remain undescribed.
The transport and recycling of vesicles have been studied in depth using yeast, neuronal and neuroendocrine systems to identify molecular components involved in protein trafficking pathways, and molecular mechanisms of vesicle docking and fusion (7)(8)(9). With respect to the insulin-stimulated translocation of GLUT4 in adipocytes, several SNARE complex proteins recently have been shown to be functionally involved in the docking of GLUT4 vesicles. These include the v-SNARE VAMP2 and the t-SNAREs syntaxin-4 and synaptosome-associated protein-23 (10 -15). Potential modulators of SNARE complex formation also have been identified in adipocytes and these include several members of the Rab family of small GTP-binding proteins, members of the Munc18 family of proteins, and synaptotagmin-5 (16 -22). Adipocytes also express the N-ethylmaleimide-sensitive fusion protein and ␣-synaptosome-associated proteins involved in catalyzing vesicle membrane fusion events (23).
Another potential modulator of SNARE complex formation is the integral membrane protein synaptophysin. Synaptophysin is a neuroendocrine-specific protein and is characteristically associated with synaptic vesicles as well as small vesicles in these cell types (24,25). The in vivo function of synaptophysin is unclear; however, in vitro studies suggest that synaptophysin may play a role in regulating interactions of SNARE complex proteins (26). Homozygous deletion of the synaptophysin gene in mice caused no identifiable phenotype; synaptic vesicles appeared to traffic normally in these animals (27,28), suggesting that related proteins may function in the place of the deleted synaptophysin. Indeed, synaptophysin-related molecules have been described (29 -32).
Recently, the cDNA for pantophysin, a protein with sequence similarity to synaptophysin, was cloned from non-neuroendocrine cells (31,33). Pantophysin displays only a 43% overall amino acid sequence identity to synaptophysin. The primary structure of pantophysin features four putative transmembrane domains similar to those of synaptophysin; however, pantophysin differs from synaptophysin in the cytoplasmic Nand C-terminal domains. Pantophysin mRNA and protein have been shown to be expressed in a variety of tissues and cell lines. Immunofluorescent microscopy of cultured cells showed colocalization of endogenous pantophysin with secretory carrier membrane proteins and the v-SNARE cellubrevin (33). Based on these studies, pantophysin is proposed to be associated with small cytoplasmic transport vesicles; however, very little is known about the function of pantophysin in vesicle trafficking and fusion events.
The aim of the present investigation was to characterize pantophysin in adipocytes and to explore the possibility that pantophysin may be used as a marker and potential tool for analysis of transport vesicles in adipocytes. As a model of SNARE complex regulation in GLUT4 translocation emerges, the relation of endogenous pantophysin to GLUT4-containing vesicles and the involvement of pantophysin in the trafficking of GLUT4 is of interest. Moreover, the relation of pantophysin to the transport of other vesicles in adipocytes is of importance to the characterization of regulated vesicle trafficking processes in these cells.

EXPERIMENTAL PROCEDURES
Materials-All tissue culture reagents were purchased from Life Technologies, Inc. with the exception of calf serum, which was purchased from HyClone (Logan, UT). Human insulin was purchased from Roche Molecular Biochemicals, Protogel acrylamide/bis-acrylamide solution was from National Diagnostics (Atlanta, GA), bovine serum albumin was from Intergen Company (Purchase, NY), and all other chemicals were from Sigma or Fisher unless otherwise noted.
Cloning of Pantophysin mRNA-Human and rat synaptophysin and rat synaptoporin protein sequences as well as the translation product of the synaptophysin-related h-Sp1 (GenBank TM Accession number S72481) were aligned using the Jotun Hein algorithm in the DNASTAR MegAlign program. Degenerate oligonucleotide primers were designed based on peptides in two of the four highly conserved transmembrane regions of synaptophysin. The primer PAN5, 5Ј-GGI TT(C/T) (G/A)TI AA(A/G) GTI (C/T)TI (C/G)A(A/G) TGG-3Ј (bases in parenthesis represent degenerate bases; I, inosine), encodes the peptide GF(V/I)KVL(Q/ E)W located in the first transmembrane domain. The primer PAN3, 5Ј-(G/A)(T/A)A IAC (G/A)AA CCA I(A/G)I (G/A)TT ICC I(C/A)C CCA-3Ј encodes the peptide W(V/G)GN(L/A)WFV(F/Y) located in the fourth transmembrane domain. These primers were used in RT-PCR with mRNA isolated from differentiated 3T3-L1 adipocytes as described below. The reverse transcription was performed using superscript reverse transcriptase (Life Technologies, Inc.) according to manufacturer's instructions. The PCR cycling profile was: 94°C for 3 min, 45°C for 1 min, 72°C for 2 min followed by 29 cycles of 94°C for 1 min, 45°C for 1 min, and 72°C for 2 min. The 600-bp product of RT-PCR was subcloned into pCRII (Invitrogen, Carlsbad, CA) according to manufacturer's instructions for dideoxy chain termination sequence determination using T7 polymerase (Sequenase version 2.0, United States Biochemical, Cleveland, OH) (36). This 600-bp fragment was used as a hybridization probe to isolate full-length clones from a cDNA library made from differentiated 3T3-L1 adipocytes (17). Five independent clones were isolated, and their sequence was determined. To obtain a complete cDNA, 5Ј-rapid amplification of cDNA ends was performed using the CLONTECH 5Ј rapid amplification of cDNA ends kit according to manufacturer's instructions (CLONTECH, Palo Alto, CA).
SDS-PAGE, Antibodies, and Immunoblotting-Triton-soluble lysates and subcellular membrane fractions were adjusted to equal protein concentrations by dilution with HES (20 mM HEPES, pH 7.4; 250 mM sucrose; 1 mM EDTA; 10 g/ml aprotinin; 1 g/ml leupeptin; 2 mM PMSF) and solubilized in SDS-PAGE sample buffer. Equal amounts of protein were separated by SDS-PAGE, electrophoretically transferred to nitrocellulose membranes (Schleicher & Schuell), blocked in Trisbuffered saline, 0.5% Tween-20, 3% BSA and analyzed by immunoblotting with ␣-GLUT4, ␣-pantophysin, ␣-VAMP2, or ␣-insulin-responsive aminopeptidase (IRAP). Rabbit polyclonal antibodies were raised (Covance, Denver, PA) against synthetic peptides corresponding to the 12 C-terminal amino acids of human GLUT4, the 16 N-terminal amino acids of murine VAMP2, or the 14 N-terminal amino acids of VAMP3 and against GST fusion proteins coding for either the 26 C-terminal amino acids of murine pantophysin or the 80 N-terminal amino acids of murine IRAP. These antisera were used at 1:100 to 1:200 dilution and incubated for 60 min at room temperature. Membranes were washed in Tris-buffered saline, 0.5% Tween-20, and immunoreactive protein was detected by incubation with 125 I-labeled protein A (NEN Life Science Products) followed by autoradiography (BioMax MR, Eastman Kodak Co.) for 18 -48 h at Ϫ80°C with an intensifying screen. Intensities of bands on autoradiographs were measured using a Molecular Dynamics densitometer.
Immunoprecipitation-Fully differentiated 3T3-L1 adipocytes were starved overnight in DMEM, 0.1% BSA and incubated in the absence or presence of 100 nM insulin for 10 min. Cells were washed with PBS and lysed in harvest buffer (20 mM HEPES, pH 7.4; 1% Triton X-100; 150 mM NaCl; 10 g/ml aprotinin; 1 g/ml leupeptin; 2 mM PMSF). Tritonsoluble material was incubated with ␣-pantophysin overnight followed by the addition of protein A-Sepharose beads (Amersham Pharmacia Biotech). Immunoprecipitates were washed twice with harvest buffer, solubilized in SDS-PAGE sample buffer, and analyzed by SDS-PAGE and immunoblotting.
Subcellular Fractionation of 3T3-L1 Adipocytes-3T3-L1 adipocytes were starved overnight in DMEM, 0.1% BSA and incubated in the absence or presence of 100 nM insulin for 10 min. Cells were washed with PBS and harvested in HES buffer and homogenized using 26 strokes of a Dounce homogenizer. Following the removal of the fat layer, the supernatant from a 10-min 16,000 ϫ g centrifugation was centrifuged at 48,000 ϫ g for 30 min to obtain a high density microsomal (HDM) pellet, which was resuspended in HES buffer. The 48,000 ϫ g supernatant was centrifuged at 210,000 ϫ g for 50 min to obtain a low density microsomal (LDM) pellet, which was resuspended in HES buffer. The 16,000 ϫ g pellet was washed with HES buffer, resuspended in HES buffer, layered over a 1.12 M sucrose cushion, and centrifuged at 100,000 ϫ g for 60 min. The PM at the interface was removed, collected by centrifugation, and resuspended in HES buffer.
Endoglycosidase Treatment-Whole cell lysates or membrane protein samples were boiled in 0.5% SDS for 3 min. Samples were diluted and incubated overnight at 37°C in glycosidase buffer (PBS, 0.5% Triton X-100, 2 mM PMSF) containing two units of N-glycosidase F (Roche Molecular Biochemicals). Reactions were stopped by the addition of SDS-PAGE sample buffer. Samples were analyzed by immunoblotting with ␣-pantophysin.
Sucrose Gradient Centrifugation-Sucrose velocity and equilibrium density gradient centrifugations were carried out essentially as described (38). Briefly, 3T3-L1 adipocytes were homogenized, and LDM was prepared from control or insulin-treated cells as described above for subcellular fractionation. The LDM pellet was resuspended in buffer containing 10 mM HEPES, pH 7.4; 1 mM EDTA; 5% sucrose. For velocity gradient centrifugation, the resuspended LDM was layered on top of a 10 -30% continuous sucrose gradient and centrifuged at 4°C for 55 min at 150,000 ϫ g. For equilibrium density centrifugation, the resuspended LDM was layered onto a 10 -50% continuous sucrose gradient and centrifuged at 4°C for 16 h at 150,000 ϫ g. Fractions were collected from the bottom of the gradients and subjected to protein determination, SDS-PAGE, and immunoblotting.
Immunopurification of Vesicles-Fully differentiated 3T3-L1 adipocytes were starved overnight in DMEM, 0.1% BSA and incubated in the absence or presence of 100 nM insulin for 15 min prior to harvesting. Cells were washed with PBS, harvested in HES buffer, and homogenized using a Dounce homogenizer. Supernatants of a 10 min 16,000 ϫ g centrifugation were centrifuged at 48,000 ϫ g for 30 min. The fat cake was removed and the supernatant, containing LDMs, was adjusted to 150 mM NaCl. Following preclearing of the LDM fraction with protein A-Trisacryl beads (Pierce), samples were incubated with antibodies (non-immune sera, ␣-GLUT4, or ␣-pantophysin) pre-coupled to protein A-Trisacryl beads. Samples were washed extensively with PBS and solubilized in SDS-PAGE sample buffer. Following separation by SDS-PAGE, samples were analyzed by immunoblotting with ␣-GLUT4, ␣-pantophysin, or ␣-IRAP.
[ 32 P]Orthophosphate Labeling-Fully differentiated 3T3-L1 adipocytes were starved overnight in DMEM, 0.1% BSA. The cells were then incubated in phosphate-free DMEM containing 0.1 mCi/ml [ 32 P]orthophosphate for 2 h followed by incubation for 10 min in the absence or presence of 100 nM insulin. The cells were washed twice in ice-cold PBS and harvested in 50 mM HEPES, pH 7.4; 1% Nonidet P-40; 1 g/ml leupeptin; 1 g/ml aprotinin; 2 mM PMSF; 2 mM sodium orthovanadate; 100 mM NaF; 10 mM sodium pyrophosphate. The lysates were centrifuged for 10 min at 10,000 ϫ g, and immunoprecipitates with non-immune serum or ␣-pantophysin were prepared as described above. Samples were solubilized in SDS-PAGE sample buffer and separated by SDS-PAGE. The gels were dried and subjected to autoradiography.

RESULTS
Cloning of Pantophysin cDNA-Synaptophysin contains four membrane-spanning regions and cytoplasmic N and C termini (39). Human and rat synaptophysin, rat synaptoporin, and a synaptophysin-like protein h-Sp1 were aligned, and DNA sequences encoding the highly conserved transmembrane domains were chosen as targets for degenerate oligonucleotide primers in RT-PCR using mRNA isolated from fully differentiated 3T3-L1 adipocytes as template. The upstream primer encoded eight amino acids (GF(V/I)KVL(Q/E)W) located in the first transmembrane domain and the downstream primer encoded the peptide W(V/G)GN(L/A)WFV(F/Y) located in the fourth transmembrane domain. After RT-PCR, a fragment of the predicted size (600 bp) was obtained and was used as a hybridization probe to isolate a full-length clone from a cDNA library made from differentiated 3T3-L1 adipocytes (17). Five independent clones were isolated; their sequences were determined and shown to correspond to the previously identified protein pantophysin (31,33). In all clones isolated, the sequence at the 5Ј-end of the clone began with a methionine codon. To determine whether this methionine codon corresponded to the start of translation, the 5Ј-end of the clone was completed using 5Ј-rapid amplification of cDNA ends as described under "Experimental Procedures." The additional 5Јsequence (42 nucleotides) was analyzed and confirmed the methionine identified in the original clone as the translational start site. The sequence coding for the amino acids "MASKAN-MVRQRFSRLSQR" previously reported for the mouse pantophysin protein was not obtained (31). This difference may reflect a tissue-specific alternate transcriptional start site; however, the possibility cannot be excluded that the mRNA coding for these additional N-terminal amino acids is expressed at low levels in adipocytes.
Pantophysin mRNA and Protein Expression-To begin to characterize the distribution and function of pantophysin in adipocytes, differentiation-dependent expression of pantophysin mRNA was examined. Poly(A) ϩ RNA was isolated from 3T3-L1 cells at Day 0 (fibroblast stage) and Day 8 (fully-differentiated adipocytes) during differentiation, and Northern blot analysis was performed using the 32 P-labeled 600-bp fragment obtained by RT-PCR as probe. As indicated in Fig. 1A, the steady-state level of the 2.1-kilobase pantophysin mRNA increased 1.7-fold (Ϯ 0.3) between Day 0 and Day 8 during differentiation from fibroblasts to adipocytes. In addition, Northern blot analysis of poly(A) ϩ RNA prepared from various murine tissues indicated that pantophysin mRNA was abundantly expressed in adipose compared with the other tissues examined (Fig. 1B).
To assess the expression of pantophysin protein throughout differentiation, Triton-soluble protein lysates were immuno-blotted with a rabbit polyclonal antibody directed against the C-terminal 26 amino acids of murine pantophysin. The ␣-pantophysinrecognizedabroadbandthatexhibitedadifferentiationdependent alteration in electrophoretic mobility (predicted MW ϭ 28,000) ( Fig. 2A). The diffuse pattern of migration is consistent with N-linked glycosylation and variable processing. A similar observation has been made for synaptophysin, and both proteins contain a potential N-glycosylation site in the first intravesicular loop (33). To verify glycosylation of pantophysin, protein samples were treated with N-glycosidase F, which cleaves asparagine-linked glycans (40). Immunoblot analysis revealed that treatment with N-glycosidase F resulted in collapse of the broad pantophysin band to a single immunoreactive species migrating at the predicted molecular weight and that steady-state protein levels remained relatively constant throughout adipogenesis of 3T3-L1 cells (Fig. 2B).
Subcellular Distribution of Pantophysin in 3T3-L1 Adipocytes-To determine the subcellular distribution of pantophysin in adipocytes and whether this distribution was altered by insulin, PM, HDM, and LDM fractions were prepared by differential centrifugation of control or insulin-treated 3T3-L1 adipocytes. For precise measurement of pantophysin, each subcellular membrane fraction was treated with N-glycosidase F to produce the single migrating species of pantophysin. Immunoblots of these fractions using ␣-pantophysin showed that in the basal state pantophysin was localized to the PM, LDM, and HDM fractions. Pantophysin was not observed in the cytosolic fraction (data not shown). As shown in Fig. 3, insulin stimulated the redistribution of pantophysin from internal membrane fractions to the PM by a modest but significant 1.4-fold (p Ͻ 0.05, determined by Student's t test comparing control versus insulin-treated samples), much less than the insulinstimulated recruitment of GLUT4 to the PM (3-5-fold). Similar results for the insulin-stimulated redistribution of pantophysin were obtained from experiments using freshly isolated rat epididymal adipocytes (data not shown).
Equilibrium and Velocity Gradient Centrifugation-To further characterize pantophysin-containing vesicles, the LDM fraction of 3T3-L1 adipocytes was analyzed by velocity and equilibrium density gradient ultracentrifugation. For separation of the vesicles based on their relative size, LDM pellets from 3T3-L1 adipocytes were resuspended and separated on a continuous 10 -30% sucrose velocity gradient as described under "Experimental Procedures." Fractions from the velocity gradient were taken from the bottom to the top and subjected to protein determination and analyzed by immunoblotting for pantophysin and GLUT4. Pantophysin exhibited a broad distribution profile indicative of variably sized vesicles; GLUT4 showed a more distinct peak; however, considerable overlap between the two vesicle populations was observed (Fig. 4A). Further analysis of these vesicle populations by equilibrium density ultracentrifugation showed nearly identical distribution profiles for pantophysin and GLUT4, suggesting that pantophysin-and GLUT4-containing vesicles within the LDM are of similar buoyant densities (Fig. 4B).
The effect of insulin on the distribution of pantophysin-and GLUT4-containing vesicles was examined using equilibrium density centrifugation of the LDM fractions isolated from control or insulin-treated cells. This analysis indicated that insulin stimulated the redistribution of GLUT4 out of the LDM compartment as expected (Fig. 5A). In addition, insulin stimulated a redistribution of pantophysin out of the LDM (Fig. 5B). This is in agreement with the immunoblot analysis of subcellular fractions from 3T3-L1 adipocytes (Fig. 3); however, a more pronounced insulin effect on pantophysin redistribution is consistently observed with the gradient centrifugation analysis. Total protein content in the gradient fractions obtained from the LDM did not change in response to insulin (data not shown).
Immunopurification of GLUT4-and Pantophysin-containing Vesicles-Subcellular fractionation and sucrose gradient centrifugation experiments showed that pantophysin and GLUT4 were distributed in overlapping membrane compartments, and in addition to translocation of GLUT4, a fraction of pantophysin trafficked from internal membranes to the PM in response to insulin. To determine whether pantophysin and GLUT4 reside in the same vesicle populations, immunopurified vesicles (IPVs) were prepared from the LDM fraction using either ␣-GLUT4 or ␣-pantophysin, and the presence of GLUT4 and pantophysin in these vesicle preparations was detected by immunoblotting (Fig. 6). As expected, ␣-GLUT4 immunoblots of GLUT4 IPVs showed that GLUT4-containing vesicles are depleted from the LDM following stimulation with insulin and translocation of GLUT4 to the PM (Fig. 6, top). Vesicles purified using ␣-pantophysin also contained GLUT4, albeit much less than vesicles purified with ␣-GLUT4, and the GLUT4 detectable in these pantophysin IPVs also was depleted in response to insulin (Fig. 6, top). IRAP, which has been colocalized to GLUT4-containing vesicles (41)(42)(43), also was present on pantophysin IPVs and was depleted from this vesicle population in response to insulin (Fig. 6, middle). Consistent with the presence of GLUT4 in pantophysin IPVs, immunoblotting with ␣-pantophysin showed that pantophysin was associated with GLUT4 IPVs (Fig. 6, bottom). Quantitative immunoprecipitation of GLUT4 IPVs showed that only 34% (Ϯ 10%) of the total LDM-derived pantophysin was associated with the GLUT4 IPVs. Pantophysin also was present in immunopurified IRAP vesicles (data not shown). Following treatment with insulin, pantophysin was depleted from both GLUT4 and pantophysin IPVs.
Association of Pantophysin with VAMP2-Synaptophysin has been shown by co-immunoprecipitation and cross-linking experiments to associate with the v-SNARE VAMP2 (26,44,45). This interaction appears to occur through the cytoplasmic N terminus of VAMP2 (45). To determine whether VAMP2 associates with pantophysin, ␣-pantophysin immunoprecipitates were prepared from fully differentiated 3T3-L1 cells and immunoblotted with ␣-VAMP2 or ␣-VAMP3. As shown in Fig.  7, VAMP2 was associated with ␣-pantophysin but not with pre-immune immunoprecipitates. VAMP3 was not detected in these ␣-pantophysin immunoprecipitates. Pantophysin also was observed in ␣-VAMP2 immunoprecipitates (data not shown). In addition, these interactions did not appear to be affected by insulin.
Pantophysin and Insulin-mediated Translocation of FIG. 2. Expression of pantophysin protein during differentiation of 3T3-L1 fibroblasts to adipocytes. Cell extracts prepared at two day intervals during differentiation were incubated in the absence (A) or presence (B) of N-glycosidase F as described under "Experimental Procedures." Ten (A) or five g (B) of total protein were solubilized in SDS-PAGE sample buffer, separated by 10% SDS-PAGE, and electrophoretically transferred to nitrocellulose membranes for immunoblotting (IB) with ␣-pantophysin. Immunoreactive material was detected by 125 I-labeled protein A and subsequent autoradiography. The bracket indicates glycosylated pantophysin. The arrow indicates the 28-kDa pantophysin species.

FIG. 3. Subcellular distribution of pantophysin in 3T3-L1 adipocytes.
A, membrane fractions treated with N-glycosidase F and immunoblotted with ␣-pantophysin. B, untreated membrane fractions immunoblotted with ␣-GLUT4. Lower panels in A and B represent immunoreactive material quantified by densitometry and expressed as a percent of the total pantophysin or GLUT4 signal. Membrane fractions were prepared as described under "Experimental Procedures" from fully differentiated 3T3-L1 adipocytes incubated in DMEM, 0.1% BSA overnight and treated for 10 min in the absence (Ϫ) or presence (ϩ) of 100 nM insulin. Ten g of protein from each of these fractions were untreated (B) or treated with N-glycosidase F (A) as described under "Experimental Procedures." Samples were solubilized in SDS-PAGE sample buffer, separated on 10% SDS-PAGE, and electrophoretically transferred to nitrocellulose membranes for immunoblotting (IB) with either ␣-pantophysin or ␣-GLUT4. Immunoreactive material was detected by 125 I-labeled protein A and subsequent autoradiography. Results are representative of three or more independent experiments. Statistical analysis was performed using a Student's t test. A probability value of p Ͻ 0.05 was found for comparison of insulin-treated (Ins) versus control samples.
GLUT4 -The association of pantophysin with GLUT4 vesicles and its interaction with VAMP2 suggest a potential role for this protein in insulin-regulated translocation of GLUT4. Although the exact role of synaptophysin is unclear, the cytoplasmic C-terminal portion of the protein contains motifs that are suggestive of a possible role in the function of the protein (46). These motifs include potential endocytic signals and serine phosphorylation sites, and a tyrosine-rich domain (47)(48)(49)(50)(51). The C terminus of pantophysin lacks many of these motifs present in synaptophysin; however, to test for the involvement of the C terminus of pantophysin in GLUT4 translocation, streptolysin-O-permeabilized 3T3-L1 adipocytes were used. Permeabilized adipocytes pre-incubated with either GST or GST fused to the C-terminal cytoplasmic domain of pantophysin (GST-panto-CT) were stimulated with insulin, and translocation of GLUT4 to the PM was assessed by immunofluorescence staining of PM sheets. Incubation with GST-panto-CT (25 M) had no effect on GLUT4 translocation either in the basal state or in response to insulin (data not shown).
Phosphorylation of Pantophysin-Synaptophysin has been shown to be phosphorylated in vitro on serine residues by Ca 2ϩ /calmodulin-dependent protein kinase II and on tryosine residues by pp60c-src (49,51,52). To determine whether pantophysin is phosphorylated in adipocytes, 3T3-L1 adipocytes were labeled in vivo with [ 32 P]orthophosphate. Lysates from control and insulin-treated 32 P-labeled cells were prepared and subjected to immunoprecipitation with non-immune serum or ␣-pantophysin (Fig. 8). Pantophysin was phosphorylated in the basal state and this phosphorylation appeared to be unchanged in response to treatment of cells with insulin. However, insulin stimulated the association of a 77-kDa protein in ␣-pantophysin immunoprecipitates, suggesting either increased insulinmediated association of this phosphoprotein with pantophysin or increased phosphorylation in response to insulin or both. Immunoblotting of ␣-pantophysin precipitates with an ␣-phos-photyrosine revealed no detectable tyrosine phosphorylation of pantophysin or associated proteins (data not shown).

DISCUSSION
Adipocytes have a large volume of intracellular vesicle traffic and undergo regulated exocytosis. However, with the exception of the insulin-stimulated redistribution of GLUT4 from an intracellular vesicle pool to the PM, the process of vesicular transport in these cells is poorly understood. The present study begins to characterize pantophysin in 3T3-L1 adipocytes, and to address the hypothesis that this ubiquitously expressed integral membrane protein may be a useful marker and functional component of adipocyte transport vesicles.
Northern and Western blot analysis of 3T3-L1 adipocyte RNA and Triton-soluble lysates, respectively, revealed the expression of a 2.1-kilobase mRNA and protein of the predicted size, 28 kDa, in 3T3-L1 adipocytes. The steady-state levels of the pantophysin transcript show a modest increase during adipogenesis of 3T3-L1 cells. Moreover, adipose tissue exhibited the highest level of pantophysin mRNA on tissue Northern blots.
Immunoblotting of cellular lysates prepared during adipogenesis with ␣-pantophysin showed that the steady-state level of pantophysin protein remains relatively constant during differentiation. These data, coupled with the increase in pantophysin mRNA during adipogenesis, suggest potential differentiation-dependent changes in mRNA and/or protein turnover events. The broad band of ␣-pantophysin-reactive material is consistent with modification by glycosylation. The related protein synaptophysin is glycosylated in vivo (25), and pantophysin and synaptophysin both contain a putative N-glycosylation site in the first intravesicular loop of the proteins (33). Indeed, glycosidase treatment of total cell lysates and of adipocyte membrane fractions prepared by differential centrifugation showed that pantophysin was glycosylated in 3T3-L1 cells. Moreover, glycosylated pantophysin exhibited a differentiationdependent alteration of its migration pattern on SDS-PAGE suggesting differential processing of the oligosaccharide moiety.
Synaptophysin has been localized to synaptic vesicles and small vesicles in neuroendocrine tissue (24,25), and pantophy- sin has been localized to small cytoplasmic trafficking vesicles (33). In the present study, three biochemical analyses were used to study the distribution of pantophysin in 3T3-L1 adipocytes, and each showed an overlap of pantophysin and the insulin-sensitive glucose transporter GLUT4. First, subcellular fractionation of 3T3-L1 adipocytes by differential centrifugation was used to localize pantophysin in these cells; in the basal state pantophysin was localized to the PM, LDM, and HDM fractions but was undetectable in the cytoplasm, consistent with the presence of transmembrane domains. Immunoblotting of equal amounts of protein from these subcellular membrane fractions indicated that pantophysin appears to be relatively evenly distributed among these compartments.
Second, the distribution of pantophysin in cultured adipocytes was examined using sucrose gradient ultracentrifugation of LDM preparations. Fractions collected from both velocity and equilibrium density gradient centrifugation revealed that pantophysin and GLUT4 reside in vesicles of overlapping size and almost identical density. Similar experiments examining the distribution of synaptophysin in vesicle preparations of neuronal or neuroendocrine cells have shown a broad profile in both velocity and equilibrium gradient centrifugation (53,54). Interestingly, in this study pantophysin was associated with LDM-derived vesicles of distinct density that overlap entirely with GLUT4-containing vesicles.
Analysis by both subcellular fractionation and sucrose gradient centrifugation showed that pantophysin and GLUT4 reside in overlapping membrane compartments. To determine whether these proteins localized to the same vesicle population, a third method was employed. Immunopurification of LDMderived vesicles demonstrated that pantophysin and GLUT4 were present on the same vesicles; GLUT4 IPVs contained pantophysin, and pantophysin IPVs contained GLUT4. In addition, the three biochemical analyses employed here also showed that, as is the case for GLUT4, insulin stimulates a portion of pantophysin to redistribute from internal membranes to the PM. Subcellular fractionation experiments in this study showed as much as an 8-fold increase in PM-associated GLUT4 in response to insulin. Insulin also stimulated a redistribution of pantophysin from intracellular membrane compartments to the PM. However, the fraction of pantophysin moving to the PM (1.4-fold increase) following treatment with insulin was much less than that for GLUT4. Consistent with these observations, equilibrium density gradient ultracentifugation also showed a depletion of both GLUT4 and pantophysin from the LDM fraction in response to insulin. Furthermore, immunopurification of vesicles from the LDM fraction with ␣-GLUT4 and ␣-pantophysin showed depletion of both vesicle pools from this fraction in response to insulin. Insulin induced approximately a 66 Ϯ 10% depletion of GLUT4 and a 41 Ϯ 2% depletion of pantophysin associated with the GLUT4 IPVs. Pantophysin IPVs exhibited approximately a 21 Ϯ 8% and 17 Ϯ 5% depletion of GLUT4 and pantophysin, respectively, in response to insulin. Finally, only 34% of the total LDM-derived pantophysin is associated with GLUT4 IPVs in the same compartment. The above data suggest that a subpopulation of the total pantophysin is present on GLUT4 vesicles and that this population traffics out of the LDM in response to insulin.
Our findings are consistent with the hypothesis that pantophysin is a component on a variety of adipocyte transport vesicles and that GLUT4 is localized to only a subpopulation of pantophysin-containing vesicles. Thus, the translocation of pantophysin observed following insulin stimulation would be a response of only a subpopulation of pantophysin-containing vesicles and would not be expected to be as dramatic as that observed for the translocation of GLUT4. One also might spec-ulate that if pantophysin is present on all vesicles undergoing regulated transport, within the LDM for example, then insulinstimulated vesicle transport represents only a minor component of the total population of these vesicles and that this subpopulation is composed primarily of GLUT4-containing vesicles within this compartment.
The v-SNARE VAMP2 has been shown, in various systems, to be involved in targeting intracellular transport vesicles (7)(8)(9). In adipocytes, both cleavage of VAMP2 by clostridial neurotoxins and addition of the cytoplasmic domain of VAMP2 blocked GLUT4 translocation to the PM in response to insulin (11,12,55). In neuronal systems, an association between VAMP2 and synaptophysin has been observed by co-immunoprecipitation and cross-linking experiments (26,44,45). In isolated synaptosomes it appears there are two pools of VAMP2, one complexed with syntaxin-1 (forming the SNARE complex) and one complexed with synaptophysin. When complexed with synaptophysin, VAMP2 appears to be unavailable for interaction with the t-SNAREs syntaxin-1 and synaptosome-associated protein-25 (26). These data suggest that synaptophysin may play a role in regulation of SNARE complex formation. In the present study pantophysin was shown to associate with VAMP2 in ␣-pantophysin immunoprecipitates from adipocytes. This protein-protein interaction occurred despite the lack of significant overall sequence identity (43%) and many of the sequence motifs present in the C-terminal tail of synaptophysin (31,33). A low affinity interaction of synaptophysin with VAMP3 has been reported (26); however, VAMP3 was not detected in ␣-pantophysin immunoprecipitates. This difference may reflect the difference in experimental methods or in the primary structures of synaptophysin and pantophysin. At this point the exact nature of the protein-protein interaction of pantophysin with VAMP2 and the role this interaction may play in vesicle trafficking requires additional characterization.
The function of pantophysin in vesicle trafficking remains unclear. The in vivo function of the related synaptophysin also is unclear; however, in vitro studies suggest that synaptophysin may play a role in regulating interactions of SNARE complex proteins (see above). Furthermore, synaptophysin forms hexameric homo-oligomers in vitro and may be involved in formation of pores for vesicle fusion (56). Ablation of the synaptophysin gene in mice does not disrupt vesicle trafficking, and other proteins of the synaptophysin family may carry out the function of synaptophysin in its absence (27,28). Synaptophysin is detected only in neuroendocrine tissues, and in other cells types pantophysin may function in an analogous manner to synaptophysin. Along these lines, to attempt to disrupt a potential function of pantophysin in GLUT4 vesicle trafficking, permeabilized adipocytes were incubated with the C-terminal cytosolic domain of pantophysin expressed as a GST fusion protein. GST-panto-CT had no effect on the ability of insulin to stimulate GLUT4 translocation. It is possible that if pantophysin does participate in the trafficking of GLUT4 vesicles, this experimental approach may be unable to define such a role. Alternatively, perhaps higher concentrations of the fusion protein are required, and/or the N-terminal cytoplasmic domain plays a role independently or in conjunction with other regions of the protein.
Synaptophysin has been shown to be phosphorylated in vitro by several kinases (48,49,51,52). In this study, pantophysin appeared as a phosphoprotein in [ 32 P]orthophosphate-labeled adipocytes in the basal state, and this phosphorylation was unchanged in response to insulin. This labeling did not appear to be a result of tyrosine phosphorylation (data not shown) suggesting phosphorylation on serine/threonine residues.
Throughout the membrane spanning regions of the proteins, the primary structure of pantophysin is similar to that of synaptophysin. Synaptophysin, however, contains a C-terminal cytoplasmic region containing multiple copies of a tyrosinerich domain that is absent in pantophysin (33). The lack of tyrosine phosphorylation observed here is consistent with the absence of this tyrosine-rich domain in pantophysin. In addition to undergoing tyrosine phosphorylation, synaptophysin is phosphorylated in vitro on serine residues by calmodulin kinase II (51). It will be of interest to determine which kinase(s) phosphorylates pantophysin and whether its phosphorylation state is related to regulation of function. In addition, despite no observable change in phosphorylation of pantophysin in response to insulin, there was a 77-kDa phosphoprotein associated with pantophysin in ␣-pantophysin immunoprecipitates. This protein did not exhibit phosphorylation on tyrosine (data not shown). At present the identity of this phosphoprotein is unknown.
This study provides the first characterization of pantophysin in adipocytes. Pantophysin is an abundant protein in adipocytes and is present in cellular membrane compartments that overlap with those containing GLUT4. Furthermore, pantophysin is a resident protein in GLUT4-containing vesicles and this subpopulation of pantophysin-associated vesicles exhibits agonist-stimulated regulation. In addition to residing on GLUT4 vesicles, pantophysin may also be a useful marker for following other vesicle transport events, as well as for the enrichment and purification of adipocyte vesicles. Numerous questions regarding the molecular interactions and the specific function of pantophysin in regulated and constitutive exocytic pathways remain to be addressed.