Role of Semicarbazide-sensitive Amine Oxidase on Glucose Transport and GLUT4 Recruitment to the Cell Surface in Adipose Cells*

The previous characterization of an abundant population of non-adrenergic imidazoline-I 2 binding sites in adipocytes and the recent demonstration of the inter-play between these binding sites and amine oxidases led us to analyze the amine oxidase activity in membranes from isolated rat adipocytes. Adipocyte membranes had substantial levels of semicarbazide-sensitive amine oxidase (SSAO). SSAO activity and immunoreactive SSAO protein were maximal in plasma membranes, and they were also detectable in intracellular membranes. Vesicle immunoisolation analysis indicated that GLUT4-con-taining vesicles from rat adipocytes contain substantial levels of SSAO activity and immunoreactive SSAO protein. Immunotitration of intracellular GLUT4 vesicles indicated that GLUT4 and SSAO colocalize in an endosomal compartment in rat adipocytes. SSAO activity was also at 30 min at Intracellular membranes were incubated with beads overnight at 4 °C (50 m g of membranes and 20 m l of beads). The beads were spun down, the super- natant was taken for later analysis, and the beads were washed five times in phosphate-buffered saline. The adsorbed material was used directly to determine amine oxidase activity; in some experiments, the adsorbed material was eluted with electrophoresis sample buffer M Tris-HCl, 20% glycerol, 2% SDS, 5% b -mercaptoethanol, 6.8), incubated for 5 min at 95 °C, cooled, and microcentrifuged before subject- ing the extracts to sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Electrophoresis and Immunoblot Analysis— SDS-polyacrylamide gel electrophoresis was performed on membrane proteins following Lae- mmli (38). Proteins were transferred to Immobilon in buffer consisting of 20% methanol, 200 m M glycine, 25 m M Tris, pH 8.3. After transfer, the filters were blocked with 4% fish gelatin, 0.02% sodium azide in phosphate-buffered saline for 1 h at 37 °C, and then incubated with antibodies in 1% nonfat dry milk, 0.02% sodium azide in phosphate-buffered saline. Transfer was confirmed by Coomassie Blue staining of the gel after the electroblot. Detection of the immune complex with the rabbit polyclonal antibodies was accomplished using 125 I-protein A for 4 h at room temperature or using the ECL Western blot detection system (Amersham). The autoradiograms were quantified using scan- ning densitometry. Immunoblots were performed under conditions in which autoradiographic detection was in the linear response range. Amine Oxidase Activity— Amine oxidase activity was determined radiochemically following the procedure by Fowler and Tipton (39). The reaction was performed in 200 m l of 0.2 M phosphate buffer at 37 °C in the presence of radioactive benzylamine (50 or 100 m M ; 50 mCi/mmol) or tyramine (50 or 100 m M ; 50 mCi/mmol) at pH 7.4 for 60 min unless otherwise stated. Reactions were carried out at 37 °C in a final volume of 225 m l of 50 m M potassium phosphate buffer, pH 7.2. Reactions were stopped by the addition of 50 m l of 4 N HCl, and the products were extracted into toluene/ethyl acetate 1:1 (v/v) containing 0.6% (w/v) 2,5-diphenyloxazole before liquid scin- tillation counting. Blank values were measured in assays stopped im-mediately after the addition of cellular extracts and represented less than 1% of total radioactivity added in each incubation mixture. Time course assays were performed to ensure that initial rates of the reaction were precisely determined; proportionality to enzyme concentration was established in each case. Protein concentrations were determined by the Bradford method (40) with g -globulin as a standard. by isolated rat adipocytes. Rat adipocytes were incubated for 45 min in incubation medium without or with 100 n M insulin, 100 m M benzylamine, 100 m M sodium orthovanadate, 1 m M semicarbazide, or combination of them. Subsequently, 2-deoxyglucose uptake was meas- ured over a 10-min period (see “Materials and Methods”). Results are mean 6 S.E. of four distinct experiments. The effects of the combination of benzylamine and vanadate on 2-deoxyglucose uptake were observed in 28 independent experiments, and the results indicated that ben- zylamine and vanadate caused a 6-fold stimulation of 2-deoxyglucose uptake, whereas insulin caused a 11.5-fold stimulation of transport.

Insulin stimulates glucose transport in adipose tissue and cardiac and skeletal muscle by promoting glucose transporter translocation from an intracellular locus to the cell surface (1)(2)(3). Muscle and fat express two isoforms of glucose transporters named GLUT4 and GLUT1. The latter is found mainly in the plasma membrane but also in the interior of the cell, and insulin causes its redistribution to the plasma membrane in adipocytes and cardiac myocytes (1)(2)(3)(4). In contrast, GLUT4 is excluded from the plasma membrane and instead localizes to an intracellular storage pool in the basal state, and insulin causes a recruitment of GLUT4-containing vesicles to the cell surface (1)(2)(3)(4). Since the translocation of GLUT4 is greater than GLUT1, GLUT4 accounts for most of the insulin-stimulated glucose transport in adipose and muscle cells (1)(2)(3)(4).
In the absence of insulin, GLUT4 is concentrated in membrane vesicles, which can be separated from cellular microsomes by velocity gradient centrifugation (5). To delineate the intracellular trafficking pathway of GLUT4, it is important to identify the proteins that colocalize with GLUT4 in the same vesicles. The mannose 6-phosphate/insulin-like growth factor II receptor and secretory carrier membrane proteins (SCAMPs) 1 have been detected in GLUT4 vesicles obtained from rat adipocytes, cardiomyocytes, and skeletal muscle (4, 6 -9), which is consistent with the hypothesis that GLUT4 moves through the endosomal recycling compartment. Phosphatidylinositol 4-kinase and the aminopeptidase gp160 (or vps 165) have also been detected in GLUT4 vesicles (10 -13), although their role in GLUT4 trafficking remains unknown. GTP-binding proteins as well as vesicle SNAREs have also been detected in GLUT4 vesicles. Thus, rab4 was initially detected in GLUT4 vesicles derived from adipocytes and from skeletal muscle (14 -16), and the overexpression of rab4 in adipocytes or in Xenopus oocytes reduces the presence of GLUT4 at the cell surface and diminishes glucose transport, substantiating a role of rab4 in GLUT4 trafficking (17,18). The vesicle SNARE proteins vesicle-associated membrane protein 2 and cellubrevin have also been detected in GLUT4 vesicles derived from rat adipocytes and rat skeletal muscle (9,19,20), and cellubrevin has also been found in GLUT4 vesicles derived from rat cardiomyocytes (9). Although there are some discrepancies, available data supports the view that vesicle-associated membrane protein 2 and cellubrevin function as critical proteins in GLUT4 trafficking (21)(22)(23)(24).
Previous reports have substantiated a large population of imidazoline-I 2 binding sites in human, rat, and hamster adipose tissue (25)(26)(27). In addition, several studies have demonstrated that imidazoline-I 2 sites correspond to amine oxidases (28 -31). These results together with the observation that GLUT4 vesicles contain several proteins as yet unidentified (3,9) compelled us to characterize the amine oxidases present in adipocytes. As a result, we have identified a semicarbazidesensitive amine oxidase (SSAO, E.C. 1.4.3.6) as another protein found in GLUT4 vesicles from rat adipocytes, 3T3-L1 adipocytes, and rat skeletal muscle, and we demonstrate a stimulatory role of SSAO on glucose transport and GLUT4 translocation to the cell surface in adipose cells. The identification of SSAO as a protein found in GLUT4 vesicles is in keeping with a recent report by Morris et al. (32) MATERIALS AND METHODS Immobilon polyvinylidene difluoride was obtained from Millipore (Bedford, MA). ␥-globulin, goat anti-mouse IgG, semicarbazide hydrochloride, pargyline, clorgyline, tyramine, benzylamine, and most commonly used chemicals were from Sigma. Purified porcine insulin was a kind gift from Eli Lilly (Indianapolis, IN). All electrophoresis reagents and molecular weight markers were obtained from Bio-Rad. Anti-GLUT4 antibody (OSCRX) was produced from rabbit after immunization with a peptide corresponding to the last 15 amino acids of the carboxyl terminus (33). Anti-SSAO antibody was produced from rabbit after immunization with the membrane-associated SSAO purified from bovine lung. Rabbit polyclonal antibodies against rat ␤1-integrin were kindly given by Dr. C. Enrich (Universitat de Barcelona).

Materials
Hexose Transport-Adipocytes were isolated from the epididymal fat pads of male Wistar rats (180 -220 g) by digestion in Krebs-Ringer buffer containing 15 mM sodium bicarbonate, 10 mM Hepes, 2 mM sodium pyruvate, bovine serum albumin (3.5% w/v), and 1.5 mg/ml collagenase. After digestion for 35-45 min at 37°C under shaking, isolated fat cells were filtered and washed three times in the same buffer without collagenase (KRBH buffer). After a preincubation period of 45 min at 37°C, each vial containing 400 l of cell suspension in KRBH containing the tested drugs (added in 4 l of suitable dilutions to obtain the final concentrations) received an isotopic dilution of 2-deoxy-D-[ 3 H]glucose, giving a final concentration of 0.1 mM equivalent to approximately 1,300,000 dpm/vial. Assays were further incubated for 10 min and then stopped with 100 l of 100 M cytochalasin B. 200-l aliquots of the cell suspension were centrifuged as described by Olefsky (34) in microtubes containing dinonyl phthalate. After centrifugation, the cells (upper part of the tubes) were placed in scintillation vials, and the incorporated radioactivity was counted. The extracellular 2-deoxyglucose present in the cell fraction was determined with adipocytes previously stopped with cytochalasin B; it did not exceed 1% of the maximum 2-deoxyglucose transport in the presence of insulin, as previously reported (35).
Subcellular Fractionation of Membranes from Adipocytes and Skeletal Muscle-Fat cell suspensions were incubated in KRBH buffer in the absence or presence of 100 nM insulin for 30 min. Cells were homogenized with a Potter-Elvejheim Teflon pestle, and subcellular membrane fractions were prepared as described previously (36). 3T3-L1 fibroblasts obtained from the American Type Culture Collection (Rockville, MD) were cultured in Dulbecco's modified Eagle's medium containing high glucose and L-glutamine and supplemented with 10% calf serum. Cells were maintained and passaged as preconfluent cultures at 37°C in a 5% CO 2 -humidified incubator. Two days postconfluence (day 0), differentiation was induced with methylisobutylxanthine (0.5 mM), dexamethasone (0.25 M), and insulin (5 g/ml) in Dulbecco's modified Eagle's medium containing high glucose, L-glutamine, and 10% fetal bovine serum. After 2 days, the methylisobutylxanthine and dexamethasone were removed, and insulin was maintained for 2 additional days. On day 4, and thereafter, Dulbecco's modified Eagle's medium and 10% fetal bovine serum was replaced every 2 days. Before each experiment, cell monolayers were incubated in serum-free Dulbecco's modified Eagle's medium for 2 h. Cells were used for experimentation between days 8 and 14. Subcellular fractionation of 3T3-L1 membranes was performed as in rat adipocytes. Microsomes from rat skeletal muscle were prepared as described (37). Crude extracts were prepared from white adipocytes or from rat liver homogenates as previously reported (25,28) and assayed either for amine oxidase activity or [ 3 H]idazoxan binding as described (28).
Protocols of Vesicle Immunoisolation-Protein A-purified 1F8 antibody was coupled to acrylamide beads (Reacti-gel GF 2000; Pierce) at a concentration of 1 mg of antibody/ml of resin according to the manufacturer's instructions. Before use, the beads were washed with phosphatebuffered saline (134 mM NaCl, 2.6 mM KCl, 6.4 mM Na 2 HPO 4 , 1.46 mM KH 2 PO 4 , pH 7.4) for at least 30 min at room temperature. Intracellular membranes were incubated with beads overnight at 4°C (50 g of membranes and 20 l of beads). The beads were spun down, the supernatant was taken for later analysis, and the beads were washed five times in phosphate-buffered saline. The adsorbed material was used directly to determine amine oxidase activity; in some experiments, the adsorbed material was eluted with electrophoresis sample buffer (0.1 M Tris-HCl, 20% glycerol, 2% SDS, 5% ␤-mercaptoethanol, pH 6.8), incubated for 5 min at 95°C, cooled, and microcentrifuged before subjecting the extracts to sodium dodecyl sulfate-polyacrylamide gel electrophoresis.
Electrophoresis and Immunoblot Analysis-SDS-polyacrylamide gel electrophoresis was performed on membrane proteins following Laemmli (38). Proteins were transferred to Immobilon in buffer consisting of 20% methanol, 200 mM glycine, 25 mM Tris, pH 8.3. After transfer, the filters were blocked with 4% fish gelatin, 0.02% sodium azide in phosphate-buffered saline for 1 h at 37°C, and then incubated with antibodies in 1% nonfat dry milk, 0.02% sodium azide in phosphatebuffered saline. Transfer was confirmed by Coomassie Blue staining of the gel after the electroblot. Detection of the immune complex with the rabbit polyclonal antibodies was accomplished using 125 I-protein A for 4 h at room temperature or using the ECL Western blot detection system (Amersham). The autoradiograms were quantified using scanning densitometry. Immunoblots were performed under conditions in which autoradiographic detection was in the linear response range.
Determination of Semicarbazide-sensitive Amine Oxidase Activity-Amine oxidase activity was determined radiochemically following the procedure by Fowler and Tipton (39). The reaction was performed in 200 l of 0.2 M phosphate buffer at 37°C in the presence of radioactive benzylamine (50 or 100 M; 50 mCi/mmol) or tyramine (50 or 100 M; 50 mCi/mmol) at pH 7.4 for 60 min unless otherwise stated. Reactions were carried out at 37°C in a final volume of 225 l of 50 mM potassium phosphate buffer, pH 7.2. Reactions were stopped by the addition of 50 l of 4N HCl, and the products were extracted into toluene/ethyl acetate 1:1 (v/v) containing 0.6% (w/v) 2,5-diphenyloxazole before liquid scintillation counting. Blank values were measured in assays stopped immediately after the addition of cellular extracts and represented less than 1% of total radioactivity added in each incubation mixture. Time course assays were performed to ensure that initial rates of the reaction were precisely determined; proportionality to enzyme concentration was established in each case. Protein concentrations were determined by the Bradford method (40) with ␥-globulin as a standard.

Membranes from Adipocytes and Skeletal Muscle Express
Semicarbazide-sensitive Amine Oxidase Activity-Preliminary studies indicated the existence in crude membranes from rat adipocytes of substantial levels of [ 3 H]idazoxan binding to imidazoline-I 2 sites, characterized by an apparent dissociation constant (K d ) of 12 Ϯ 2 nM and a maximal binding capacity (Bmax) of 390 Ϯ 112 fmol/mg of protein (n ϭ 7). This is in agreement with previous reports showing a large number of imidazoline-I 2 binding sites in membranes from white adipocytes (25)(26)(27). Because amine oxidases have been reported to display imidazoline-I 2 binding sites (28 -31), we initially determined the type of amine oxidase present in crude extracts from rat adipocytes, and this was compared with crude membrane preparations derived from rat liver. This was done by assaying the oxidation of tyramine, a substrate than can be oxidized by monoamine oxidases A and B (41) and by SSAO in the absence or presence of pargyline (a preferential irreversible inhibitor of MAOs) or semicarbazide (an inhibitor of SSAO). Results shown in Fig. 1 clearly indicate that liver extracts express MAO activity since all tyramine oxidation is blocked by pargyline and does not express SSAO activity since no inhibition occurs in the presence of semicarbazide. On the other hand, adipocyte extracts contain both MAO and SSAO; complete inhibition of tyramine oxidation is only obtained by the addition of semicarbazide and pargyline (Fig. 1). Previous studies have also reported the activity of SSAO in crude extracts from adipose tissue (42,43).
Further studies were performed comparing the oxidation of tyramine to that of benzylamine (a preferential substrate of SSAO) but also oxidized by MAO-B (42) in light microsomes from isolated rat adipocytes; they indicated the presence of a high SSAO activity in light microsomes. This was substantiated by (a) tyramine oxidative activity largely inhibited (89% decrease) by 10 Ϫ4 M semicarbazide ( Fig. 2A) and (b) benzylamine oxidative activity totally blocked by semicarbazide (Fig. 2B). In contrast, the tyramine oxidation measured in mitochondrial fractions from rat adipocytes was barely inhibited by semicarbazide (less than 9% inhibition at 10 Ϫ3 M inhibitor) but was totally blocked by the MAO inhibitor clorgyline (90% inhibition at 10 Ϫ4 M) (Fig. 2C). Of note, clorgyline blocked tyramine oxidation by only 10%, and it did not alter benzylamine oxidation in light microsomes from rat adipocytes ( Fig. 2A). This indicates that most of benzylamine and tyramine oxidation by light microsomes is not due to MAO activity. Furthermore, light microsomes from rat adipocytes contain SSAO, which can be assayed by using benzylamine as a specific substrate or by evaluating the fraction of tyramine oxidation that is semicarbazide-sensitive. Substantial semicarbazidesensitive amine oxidase activity was also detected in light microsomes obtained from 3T3-L1 adipocytes or from intracellular membranes obtained from rat skeletal muscle (data not shown).
Distribution of Semicarbazide-sensitive Amine Oxidase in Isolated Rat Adipocytes-Subcellular fractionation of membranes from isolated rat adipocytes revealed high semicarbazide-sensitive amine oxidase activity, assayed as the oxidation of benzylamine or the oxidation of tyramine sensitive to semicarbazide, in plasma membrane preparations ( Fig. 3A and data not shown). SSAO activity in plasma membrane was near 4-fold greater than in light microsomes (Fig. 3A). Low monoamine oxidase activity was detected both in plasma membrane and in light microsomes (data not shown). The utilization of a polyclonal antibody raised against SSAO purified from bovine lung microsomes detected a single band in adipocyte mem-branes showing an electrophoretic mobility similar to that observed in bovine lung microsomes (Fig. 3B). Analysis of immunoreactive SSAO protein indicated that it was more abundant in plasma membrane than in light microsomes (Fig. 3B). Incubation of adipocytes for 30 min in the presence of 100 nM insulin did not alter SSAO activity in light microsomes or in plasma membrane preparations (Fig. 3A). No alteration in the abundance of immunoreactive SSAO protein was detected after incubation with insulin in plasma membranes or in light microsomes (Fig. 3B).
Intracellular GLUT4-containing Vesicles Contain Semicarbazide-sensitive Amine Oxidase Activity-Based on the pres- ence of SSAO activity in light microsomes from adipose cells and skeletal muscle, we next examined whether semicarbazide-sensitive amine oxidase activity colocalized with intracellular GLUT4-containing vesicles. In a first step, we fractionated intracellular membranes obtained from rat skeletal muscle in sucrose gradient. These membranes are highly enriched in intracellular markers such as GLUT4, SCAMPs, cellubrevin, or vesicle-associated membrane protein 2 and are free from plasma membrane markers (9,37,44). Analysis of fractions obtained from sucrose gradient centrifugation demonstrated a parallelism between the profile of GLUT4 abundance and the activity of SSAO present in the different membrane fractions (Fig. 4).
More direct evidence was obtained in vesicle immunoisolation assays. Quantitative vesicle immunoisolation analysis using monoclonal antibody 1F8, specific against GLUT4, and coupled to acrylic beads was performed in light microsomes from rat adipocytes. This method of vesicle immunoisolation adsorbed 70 -90% of total GLUT4 from the fractions in a specific manner (Fig. 5A). Under these conditions, GLUT4 vesicles specifically contain immunoreactive SSAO protein, which accounted for by 18 -24% of total SSAO present in light microsome membranes (Fig. 5A). Furthermore, GLUT4 vesicles con-tain active SSAO. Thus, whereas the material nonspecifically bound to IgG beads contained very low levels of SSAO activity, 1F8 antibody brought a substantial level of SSAO activity (Fig.  5B). Near 25% of total SSAO activity present in light microsomes was found in GLUT4-containing vesicles; based on this and that GLUT4-containing vesicles represent around 3% of total protein in light microsome membranes (45), we conclude that there is a substantial enrichment in specific SSAO activity in these vesicles in comparison with the total light microsomal fraction. A specific immunoadsorption of SSAO activity was also found in light microsomes obtained from 3T3-L1 adipocytes and in intracellular membranes obtained from rat skeletal muscle (data not shown).
To further analyze the precise degree of colocalization between SSAO and GLUT4 in intracellular membranes, we performed immunotitration experiments using increasing amounts of 1F8 (bound to acrylamide) to adsorb GLUT4 vesicles in adipocyte light microsomes and determined the amount of recovered GLUT4, SCAMPs, and SSAO. We found that about seven times less 1F8 antibody was required to reach a halfsaturating degree of GLUT4 adsorption than was the case for SCAMPs (Fig. 6). Thus, when using 8 g of 1F8 antibody, 70% of all GLUT4 was immunoadsorbed, and only 20% of SCAMP37 was immunoprecipitated (data not shown). This is in keeping with previous observations in low density microsomes obtained from isolated rat cardiomyocytes (4). Immunotitration assays revealed that SSAO colocalized with GLUT4 vesicles only when maximal amounts of 1F8 antibody were used (20 g) and, in consequence, when values of GLUT4 vesicle immunoisolation were maximal (80 -90% of total GLUT4 present in light microsome fractions) (Fig. 6). These observations highly suggest that that in isolated rat adipocytes, only a fraction of GLUT4-containing vesicles, probably those in an endosomal compartment, contain SSAO.
Benzylamine and Vanadate Stimulate in a Synergic Manner Glucose Transport in Isolated Rat Adipocytes-Based on the high abundance of SSAO activity in rat adipocytes and its presence in GLUT4 vesicles, we next studied whether SSAO activity might exert a regulatory role on glucose transport. To this end, we initially studied the effect of benzylamine, a preferential substrate of SSAO, on 2-deoxyglucose uptake by iso- lated rat adipocytes (Fig. 7). Adipose cells showed a low rate of 2-deoxyglucose uptake under basal conditions, and insulin caused a 10-fold stimulation of hexose transport (Fig. 7). Under these conditions no stimulatory effect of benzylamine (up to 1 mM) was detected (Fig. 7 and data not shown). In parallel, we also studied the combination of benzylamine (100 M) and a very low concentration of vanadate (100 M), which alone did not stimulate 2-deoxyglucose uptake. Under these conditions, benzylamine and vanadate caused a near 6-fold stimulation of glucose transport (Fig. 7). This effect was due to the presence of adipose cells, and incubation medium alone showed no amine oxidase activity (data not shown). Furthermore, the benzylamine-vanadate-stimulated 2-deoxyglucose uptake was dependent on the activity of SSAO, since incubation in the presence of the SSAO inhibitor semicarbazide (1 mM) abolished the stimulation of 2-deoxyglucose uptake by benzylamine and vanadate (Fig. 7). Semicarbazide did not block insulin-stimulated glucose transport (Fig. 7). The effect of the combination of benzylamine and vanadate on glucose transport was not due to a stimulatory effect of vanadate on SSAO activity as determined by using either adipocyte light microsomes or purified bovine lung SSAO (data not shown).
The addition of catalase to the incubation medium did not modify basal or insulin-stimulated glucose transport (Fig. 8). However, the addition of catalase completely abolished the stimulation of 2-deoxyglucose uptake caused by the combination of benzylamine and vanadate (Fig. 8). Further support to the idea that benzylamine stimulates glucose transport as a result of the production of hydrogen peroxide came from experiments in which cells were incubated in the presence of hydrogen peroxide and vanadate. 0.5 mM hydrogen peroxide stimulated 3-fold glucose transport, and it caused a synergistic effect in combination with 100 M vanadate (Fig. 8). Furthermore, lower concentrations of hydrogen peroxide such as 0.1 mM did not stimulate glucose transport, but combination with 100 M vanadate again stimulated glucose transport in a synergistic manner (data not shown).
Benzylamine and Vanadate Recruit GLUT4 Glucose Carriers to the Plasma Membrane in Adipose Cells-Next, we determined whether the stimulation of glucose transport induced by the combination of benzylamine and vanadate is a consequence of alterations in the abundance of GLUT4 glucose carriers in FIG. 5. SSAO activity in GLUT4-containing vesicles from isolated rat adipocytes. Light microsomes from isolated rat adipocytes (100 g) were incubated with acrylic beads coupled to monoclonal anti-GLUT4 antibody 1F8 (ϩ) or to an irrelevant antibody (Ϫ). After incubation, the adsorbed (pellet) and the nonadsorbed fractions (supernatant) were either electrophoresed and immunoblotted to determine the abundance of GLUT4 and SSAO (panel A) or subjected to assay of SSAO activity (benzylamine oxidation as described, panel B). Because the amount of protein bound to the acrylic beads is too low to measure, the enzyme activity was not expressed as specific activity. Results are representative of four different experiments.
FIG. 6. Immunotitration of GLUT4, SCAMPs, and SSAO in intracellular membranes from isolated rat adipocytes. Light microsomes from rat adipocytes were immunoadsorbed with increasing amounts of 1F8 acrylic beads (corresponding to the 1F8 quantities indicated). GLUT4, SCAMPs, and SSAO were detected in the adsorbed fractions by Western blotting. Results are representative of four different experiments.

FIG. 7.
Effect of benzylamine and vanadate on glucose transport by isolated rat adipocytes. Rat adipocytes were incubated for 45 min in incubation medium without or with 100 nM insulin, 100 M benzylamine, 100 M sodium orthovanadate, 1 mM semicarbazide, or combination of them. Subsequently, 2-deoxyglucose uptake was measured over a 10-min period (see "Materials and Methods"). Results are mean Ϯ S.E. of four distinct experiments. The effects of the combination of benzylamine and vanadate on 2-deoxyglucose uptake were observed in 28 independent experiments, and the results indicated that benzylamine and vanadate caused a 6-fold stimulation of 2-deoxyglucose uptake, whereas insulin caused a 11.5-fold stimulation of transport. the cell surface from adipose cells. To this end, isolated rat adipocytes were incubated in the absence or in the presence of insulin, benzylamine (100 M), vanadate (100 M), or the combination of benzylamine and vanadate, and 45 min later, cells were homogenized and subjected to subcellular membrane fractionation. Insulin treatment caused a 3.3-fold increase in the abundance of GLUT4 in plasma membrane preparations (Fig.  9). Under these conditions, incubation of cells with benzylamine or vanadate alone did not alter the abundance of GLUT4 in plasma membranes ( Fig. 9 and data not shown). However, the incubation of cells with benzylamine and vanadate caused a 2.3-fold increase in the presence of GLUT4 in plasma membrane fractions (Fig. 9). The effects of insulin or the combination benzylamine and vanadate on GLUT4 abundance in plasma membranes were specific, and the plasma membrane marker ␤ 1 -integrin remained unaltered after incubation with these agents (Fig. 9). The recruitment of GLUT4 to the plasma membranes promoted by insulin or the combination of benzylamine and vanadate was concomitant with a reduction of GLUT4 abundance in light microsomes (Fig. 9). This is consistent with the view that benzylamine and vanadate translocate GLUT4 to the plasma membrane in isolated rat adipocytes.

DISCUSSION
Semicarbazide-sensitive amine oxidase activities are found in blood plasma and associated with membranes in many mammalian tissues such as in smooth muscle from rat aorta (46,47), endothelial cells (48), chondrocytes in rat articular cartilage (49), adipocytes from white and brown adipose tissue (42,43), pig dental pulp (50), bovine eye (51), and bovine lung (52,53). Recent cloning analysis indicates that the sequences of three different semicarbazide-sensitive amine oxidases such as bovine serum amine oxidase, human placental amine oxidase, and rat SSAO show a high similarity (ranging from 78 to 84%) and therefore they can be considered as species counterparts (32,54,55). In our study we provide evidence for a high SSAO activity in membranes from isolated rat adipocytes and in microsomes from 3T3-L1 adipocytes and rat skeletal muscle. Furthermore, the maximal SSAO activity is detected in plasma membranes from rat adipocytes, and this is a reflection of a greater abundance of SSAO in this cellular compartment.
An important finding of our study is that SSAO colocalizes with GLUT4 in intracellular vesicles, and this seems to be a rather general property of GLUT4-expressing tissues, since this colocalization was found in rat adipocytes, 3T3-L1 adipocytes, and rat skeletal muscle. We have found that the colocalization of GLUT4 and SSAO is partial, so nearly 18 -24% of total intracellular SSAO is present in the intracellular GLUT4 membrane population purified from isolated rat adipocytes or from 3T3-L1 adipocytes. In addition, immunotitration experiments revealed that the colocalization of SSAO and GLUT4 was only detectable when maximal amounts of anti-GLUT4 antibody were used, suggesting that only a fraction of GLUT4 vesicles, probably those in an endosomal compartment, contain SSAO. Based on these results as well as on observations in 3T3-L1 adipocytes, skeletal muscle, and cardiomyocytes, supporting the view that GLUT4 is present in an endosomal compartment as well as in an exocytic compartment (4,9,56), we propose that SSAO colocalizes with intracellular GLUT4 in an endosomal population rather than in the specific exocytic/storage compartment. According to this view, we propose that SSAO and GLUT4 would colocalize in an endosomal population, and then they would undergo a differential sorting process to distinct membrane compartments. Furthermore, we have found that insulin does not modify the cellular distribution of SSAO or its enzyme activity in adipose cells, and this suggests that the traffic of the specific endosomal compartment in which GLUT4 and SSAO colocalize is not affected by insulin.
Here we have found that benzylamine, a substrate of SSAO activity, stimulates glucose transport in the presence of very low concentrations of vanadate in isolated rat adipocytes, which by themselves are incapable of stimulating glucose transport. This stimulatory effect of benzylamine and vanadate indeed depends on SSAO activity, and it is completely abolished in the presence of the specific inhibitor of SSAO, semicarbazide. Benzylamine by itself does not enhance glucose transport (not even at millimolar concentrations), and vanadate is not acting by stimulating SSAO activity. In addition, the effect of benzylamine is also mimicked by other amines that are substrates of SSAO, such as tyramine. Thus, tyramine and vanadate synergistically stimulate glucose transport in iso-  3 and 7) before they were subjected to subcellular membrane fractionation. Fifteen g of plasma membrane (PM, lanes [5][6][7][8] or light microsomes (LDM, lanes 1-4) obtained from the different experimental groups were laid on gels. The distribution of GLUT4 or ␤1-integrin was determined by immunoblot analysis by using specific antibodies (see "Materials and Methods"). Representative autoradiograms are shown. lated rat adipocytes in a way that is also inhibitable by semicarbazide. 2 Both tyramine and benzylamine incubated in the presence of vanadate stimulate glucose transport to a similar extent that represents 50 -60% of the maximal effect of insulin.
We have also found that benzylamine promotes in combination with vanadate the translocation of GLUT4 to the plasma membrane in isolated rat adipocytes. Again, this effect is also observed for other substrates of SSAO such as tyramine, and the stimulation of GLUT4 abundance in the plasma membrane is around 50 -60% of the maximal effect caused by incubation in the presence of insulin. Based on this, it is very likely that the mechanism by which glucose transport is markedly enhanced by the combination of benzylamine and vanadate entails the recruitment of GLUT4 glucose transporters to the cell surface.
The precise physiological roles of the amine oxidases sensitive to semicarbazide as well as the endogenous substrates of SSAO remain speculative. In this regard, we have found that substrates of SSAO stimulate, in concert with vanadate, glucose transport and GLUT4 glucose transporter translocation in isolated rat adipocytes. Based on these observations, we propose that SSAO participates under in vivo conditions in the regulation of glucose disposal in insulin-sensitive tissues. In this connection, it will be important to determine (a) whether endogenous SSAO substrates exert stimulatory effects on glucose disposal in peripheral tissues in vivo and if so (b) whether these substrates play a regulatory role on glucose transport in peripheral tissues.
Another important issue is the understanding of the signals triggered by the combination of SSAO substrates and vanadate that lead to activation of glucose transport and translocation of GLUT4 carriers in rat adipocytes. As a consequence of the catalysis of SSAO, the amine substrate is converted into the aldehyde, and this is concomitant to the production of ammonia and H 2 O 2 . Several pieces of information indicate that hydrogen peroxide production is a crucial event in the effects of SSAO substrates on glucose transport. (a) Different SSAO substrates such as benzylamine or tyramine exert quantitatively similar stimulatory effects on glucose transport and GLUT4 translocation, (b) the effects of the combination of benzylamine and vanadate on glucose transport are blocked by the addition of catalase, and (c) the combination of hydrogen peroxide and vanadate also show synergistic effects on glucose transport. In this regard, it has been proposed that the participation of SSAO activity in prostaglandin biosynthesis (57) or on the influence of nitric oxide in the vascular smooth muscle (58) might be a consequence of the production of H 2 O 2 . Further supporting this view, hydrogen peroxide has been previously implicated in insulin action; thus, the effects of insulin on glucose transport in rat adipocytes, skeletal muscle, or cardiomyocytes can be mimicked by H 2 O 2 (59 -61), the activity of the insulin receptor kinase can be stimulated by hydrogen peroxide, suggesting a modulatory role for H 2 O 2 (62), and insulin has been reported to stimulate H 2 O 2 production in rat adipocytes (63). In any case, SSAO activity does not seem to play a crucial role on insulin-stimulated glucose transport, since inhibition by semicarbazide does not block insulin-stimulated 2-deoxyglucose uptake in isolated rat adipocytes. In summary, we favor the view that the combination of benzylamine and vanadate stimulates glucose transport in adipose cells by a mechanism that depends on hydrogen peroxide formation perhaps via its oxidative capacity. However, this does not rule out the possibility that SSAO substrates modulate glucose transport via vanadate-independent pathways.
The observation that SSAO colocalizes with GLUT4 in intracellular vesicles provides further support to the view that SSAO activity, through H 2 O 2 production, plays a regulatory role in the traffic of GLUT4. Based on the fact that most SSAO is in cell surface, it is likely that amines provided exogenously will be metabolized at the plasma membrane in the adipocyte. However, it remains to be determined whether SSAO-dependent oxidative deamination and products formation critical for regulation of glucose transport occurs at the cell surface or alternatively by intracellular SSAO. This requires further experimental evidence.