Semicarbazide-Sensitive Amine Oxidase/Vascular Adhesion Protein-1 Activity Exerts an Antidiabetic Action in Goto-Kakizaki Rats

The 2-[1,2-³H]-d-deoxyglucose (26 Ci/mmol) was from NEN Life Science Products (Boston, MA) and the [¹⁴C]Benzylamine (59 Ci/mmol) was from Amersham Pharmacia Biotech (Buckinghamshire, U.K.). Purified porcine insulin was a kind gift from Eli Lilly (Indianapolis, IN). Semicarbazide hydrochloride, benzylamine hydrochloride, sodium orthovanadate, and other chemicals were purchased from Sigma Aldrich (St. Louis, MO). Ketamine was obtained from Mérieux (Imalgene, Mérieux, France). Collagenase type I was obtained from Worthington (Lakewood, NJ) and collagenase P from Roche Diagnostics (Basel, Switzerland). The osmotic minipumps used in chronic studies were from Alza (Mountainview, CA). All electrophoresis reagents and molecular weight markers were obtained from Bio-Rad (Hercules, CA). Enhanced chemiluminescence reagents (super signal substrate) were from Amersham (Arlington Heights, IL). Anti-GLUT4 antibody (OSCRX) was produced from rabbit as previously reported (15). Rabbit polyclonal antibodies against rat β₁-integrin were kindly given by Dr. C. Enrich (Universitat de Barcelona, Barcelona, Spain). Anti-SSAO/VAP-1 was a kind gift from Dr. S. Jalkanen (University of Turku, Turku, Finland), and monoclonal antiphosphotyrosine antibody 4G10 was from Upstate Biotechnology. Human recombinant SSAO was provided by BioTie (Turku, Finland). Antibody against amylase was from Sigma Aldrich. Antibody against islet amyloid polypeptide (IAPP) was kindly given by Dr. A. Clark (Radcliffe Infirmary, Oxford, U.K.).

Male Wistar rats weighting 180–220 g were purchased from Harlan (Interfauna Ibèrica S.A., Sant Felin de Codines, Spain). GK weight-matched diabetic rats were purchased from M&B Animal Models (Ry, Denmark). The animals were housed in animal quarters at 22°C with a 12-h light/dark cycle and were fed ad libitum. All procedures used were approved by the animal ethical committee of the University of Barcelona.

Unanesthetized Wistar or GK diabetic rats were injected intravenously into the tail vein with vehicle (PBS, pH 7.4) or with a dose of 7 μmol/kg body wt benzylamine and/or 20 μmol/kg sodium orthovanadate. An oral glucose tolerance test (2 g/kg body wt) was carried out 15 min after benzylamine and vanadate injection. Blood samples were collected before (0 min) and 30, 45, 60, 90, and 120 min after glucose administration. Plasma was obtained for the determination of glucose and insulin concentrations. Food was withdrawn at 8:00 a.m. to ensure a 3-h fast before the beginning of the study.

Osmotic minipumps containing benzylamine (84 μmol · kg body wt⁻¹ · day⁻¹) were implanted subcutaneously in GK diabetic rats anesthetized by ketamine hydrochloride (95 mg/kg) and xylasine (10 mg/kg). Animals that did not receive benzylamine were sham operated. A single intraperitoneal injection of vanadate (25 μmol/kg body wt) or the vehicle (PBS) was administered daily at 0900. Glycemia was measured on arterio-venous blood collected from the tail vessels at 0900, before the administration of vanadate for 2 weeks. Insulin concentrations were determined before and after treatment.

Adipocytes were isolated from epididymal and perirenal fat pads in nondiabetic and GK diabetic male rats (180–220 g), and 2-d-deoxyglucose uptake measurements were performed as reported (10,12). Isolated fat cells from control or treated rats were disrupted for total membrane preparation by hypo-osmotic lysis in a 20 mmol/l HES buffer and an antiprotease cocktail as reported (10). Protein concentrations were determined by the Bradford method (16) with γ-globulin as a standard. Isolated strips of soleus muscles were incubated, and 2-d-deoxyglucose uptake measurements were performed as reported (17). Adipose tissue explants used for coincubation experiments with soleus muscle were prepared by dissecting and cutting epididymal and perirenal fat pads into small pieces that were further rinsed with Krebs-Ringer buffer before use.

Pancreata from Wistar and GK diabetic rats were enzymatically digested with collagenase (Roche Diagnostics) using the procedure of pancreatic duct cannulation and density gradient purification (Histopaque; Sigma Diagnostics) as described (18). Islets were separated from the remaining exocrine tissue by handpicking under a stereomicroscope.

Batches of six freshly isolated islets were incubated in a shaking water bath for 90 min at 37°C in 1.0 ml bicarbonate buffered medium (pH 7.4) containing 0.5% BSA, d-glucose (at 5.5 or 16.7 mmol/l), and test substances as appropriate: sodium orthovanadate (100 μ/mmol/l), benzylamine hydrochloride (100 μ/mmol/l), or semicarbazide hydrochloride (1 mmol/l). For the first 10 min of incubation, the vials containing the medium with the islets were gassed with 95% O₂ 5% CO₂. At the end of the incubation period the supernatants were collected and stored at −20°C until insulin determination by radioimmunoassay (CIS Biointernational, Gyf-Sur-Yvette, France).

The radiochemical determination of amine oxidase activity was performed basically as described by Fowler and Tipton (19), with slight modifications (11).

The glucose concentration in plasma was determined by the glucose oxidase method (MPR-three glucose/GOD-PAP Method; Boehringer-Mannheim) in glucose tolerance experiments and with a rapid glucose analyser (Accutrend Sensor Comfort; Roche) in chronic treatments. Plasma immunoreactive insulin concentration was determined with a Sensitive Rat insulin RIA kit (Linco Research, St. Charles, MO). SDS polyacrylamide gel electrophoresis was performed on membrane proteins following Laemmli (20). Proteins were transferred to Immobilon, and immunoblotting was performed as reported (15).

The ⁵¹V nuclear magnetic resonance (NMR) spectra for vanadate and peroxovanadate detection were recorded on a Varian Unity 300-MHz spectrometer. Field frequency stabilization was achieved by disolving samples with D₂O. Spectral widths of 380 ppm (78.84 MHz), a 90^o pulse angle, and an accumulation time of 0.5 s were used. The chemical shifts are reported relative to the external reference standard VOCl₃ (assigned to 0 ppm, where ppm represents the chemical shift of an atom in a defined chemical environment and is directly proportional to the resonance frequency of the atom in that environment). Exponential line-broadening (10 Hz) was applied before Fourier transformation.

Rat adipocyte suspension was incubated at 37°C with the indicated compounds at 100 μmol/l during 45 min. At the end of the incubation adipocytes were homogenized in 10 mmol/l Tris buffer containing protease inhibitors (pH 7.4). After a 15,000g centrifugation at 4°C for 5 min, the fat cake was discarded. The homogenates, after desalting step by Sepharose-G25 column, were conserved at −80°C until use for phosphatase activity determination. Phosphatase activity measurement was estimated by the potency of the homogenates (20 μg proteins) to hydrolyze p-nitrophenyl phosphate (5 mmol/l pNPP) in 10 mmol/l Tris-HCl, pH 7.5, in the presence of 2 mmol/l dithiothreitol. The extent of hydrolysis was determined by spectroscopy at 405 nm after 1 h of incubation at 37°C, as described by Kremerskothen and Barnekow (21).

Insulin and glucose responses during the glucose tolerance test were calculated as the incremental plasma values integrated over a period of 120 min after the injection of glucose. Areas under curve of insulin and glucose responses were calculated using the Graph Prism program. Data were presented as mean ± SE, and unpaired Student’s t test was used to compare two groups. When experimental series involved more than two groups, statistical analysis was done by one-way ANOVA and further post hoc Dunnett’s or Tukey’s t tests.

To determine whether SSAO/VAP-1 is a drug target for type 2 diabetes, we selected the GK diabetic rats. These animals are hyperinsulinemic, nonobese, and hyperglycemic, and they show defective insulin secretion and peripheral insulin resistance but normal insulin-stimulated glucose transport in adipocytes (22–25). Adipose tissue from GK diabetic rats showed normal SSAO activity (162 ± 12 and 173 ± 19 nmol benzylamine · min⁻¹ · mg protein⁻¹ in control and GK rats), and the combination of 0.1 mmol/l benzylamine and 0.1 mmol/l vanadate markedly stimulated glucose transport in isolated adipocytes from GK rats (Fig. 1A). In fact, benzylamine/vanadate-stimulated and maximally insulin-induced glucose transport was similar in control Wistar and in GK rats (Fig. 1A), the latter in agreement with previous data (25). Next, GK diabetic rats were injected via tail vein with vehicle or a dose of 7 μmol/kg body wt benzylamine and/or 10 or 20 μmol/kg sodium orthovanadate. An oral glucose tolerance test (2 g/kg body wt) was carried out 15 min after drug injection (Fig. 1B). The acute intravenous administration of benzylamine and vanadate enhanced glucose tolerance in GK rats (30 and 47% increase after benzylamine/vanadate 10 μmol/kg and benzylamine/vanadate 20 μmol/kg, respectively) (Fig. 1B and C). In fact, administration of benzylamine and 20 μmol/kg vanadate ameliorated a glucose tolerance test of GK diabetic rats to values approaching those of nondiabetic rats (Fig. 1C). In these conditions, administration of benzylamine or vanadate alone did not alter glucose tolerance (Fig. 1B and C). The improvement in glucose tolerance induced by benzylamine and vanadate was also detected in nondiabetic Wistar rats (10 and data not shown).

Next, we tested the effects of the chronic administration of these compounds in GK diabetic rats. To this end, GK diabetic rats were implanted subcutaneously with osmotic minipumps releasing benzylamine (84 μmol · kg⁻¹ · day⁻¹) or were sham operated. Benzylamine is stable for 2 weeks in implanted osmotic minipumps as based on its capacity to further stimulate glucose transport in isolated adipocytes after this period (data not shown). A group of animals was also subjected to daily intraperitoneal injection of vanadate (25 μmol/kg) for 2 weeks. While treatment with vanadate caused a moderate and unsignificant reduction of glycemia, the combination of benzylamine and vanadate during 1 week significantly reduced the hyperglycemia to levels close to those of the nondiabetic group (Fig. 2A). Administration of benzylamine plus vanadate for 2 weeks did not alter SSAO/VAP-1 or GLUT4 protein expression in adipose cells (Fig. 2B). The normalization of glycemia caused by chronic treatment with both benzylamine and vanadate took place in the absence of changes in plasma insulin concentrations (data not shown).

To determine whether the antidiabetic effects caused by the chronic combined treatment with benzylamine and vanadate were a consequence of stimulation of peripheral glucose utilization, we analyzed glucose transport in adipose cells obtained from rats subjected to chronic treatment. A 2-week treatment with benzylamine and vanadate stimulated basal glucose transport and GLUT4 abundance at the plasma membrane in isolated adipocytes (Fig. 2C). We also detected a statistically significant correlation between the abundance of GLUT4 at the cell surface and basal glucose transport in adipocytes from untreated and benzylamine/vanadate-treated diabetic rats (Fig. 2C). These effects on glucose transport were not detected in adipose cells obtained from rats treated with benzylamine or vanadate alone (data not shown).

Chronic treatment with benzylamine and vanadate also had an impact on muscle insulin responsiveness. Indeed, incubation of soleus muscle preparations from GK rats showed normal basal glucose transport and insulin resistance (Fig. 2D), which is in keeping with previous observations (21,26). Chronic treatment with benzylamine/vanadate reversed the insulin response to normal levels (Fig. 2D). No effects were observed on muscle glucose transport after chronic administration with vanadate alone (Fig. 2D). Chronic administration of benzylamine/vanadate did not alter adipose mRNA levels of leptin, tumor necrosis factor-α, adiponectin, or resistin or plasma concentrations of free fatty acids (data not shown). These data indicate that the effects of chronic benzylamine/vanadate on glycemia and muscle insulin responsiveness are unrelated to changes on adipocytokines gene expression or on circulating free fatty acids.

To determine the basis for the improvement in glucose tolerance detected in nondiabetic and diabetic rats after acute treatment with benzylamine/vanadate, nondiabetic rats were acutely treated with benzylamine (7 μmol/kg body wt) and/or vanadate (20 μmol/kg) or with insulin (10 units/kg), and 30 min later, soleus muscles were incubated and glucose transport measured. Previous in vivo insulin administration caused a 75% stimulation of muscle glucose transport, and previous benzylamine/vanadate treatment also caused a marked enhancement of glucose transport (40% increase) (Fig. 3A). Next, soleus muscles were incubated in the presence of different combinations of 0.1 mmol/l benzylamine and/or 0.1 mmol/l vanadate in the absence or presence of adipose tissue explants (Fig. 3B). In the absence of adipose tissue, muscle glucose transport remained unaffected by benzylamine/vanadate, which is in keeping with the fact that SSAO activity is very low in rat skeletal muscle; however, in the presence of adipose tissue explants, muscles underwent a marked stimulation of glucose transport by benzylamine/vanadate (Fig. 3B), and this effect was inhibited by the SSAO inhibitor semicarbazide. The presence of adipose tissue explants did not alter basal or insulin-stimulated glucose transport in the incubated soleus muscle (data not shown).

In a further set of experiments, soleus muscles were incubated in the presence of benzylamine/vanadate and human recombinant SSAO/VAP-1 (Fig. 3C). As mentioned before, benzylamine/vanadate caused no effect on muscle glucose transport; however, the presence of SSAO/VAP-1 caused the stimulation of muscle glucose transport that was comparable with the effect of a supramaximal concentration of insulin and was also inhibited by semicarbazide (Fig. 3C). In all, these data indicate that the in vivo treatment with the combination of benzylamine/vanadate stimulates muscle glucose transport, and this is a consequence of signals that depend on SSAO activity and that may originate in adipose tissue.

In a further step, we tested the possibility that insulin secretion was responsible for some of the effects of the acute treatment with benzylamine/vanadate on glucose tolerance detected in GK diabetic rats. Thus, we measured plasma insulin concentrations at different times after the glucose tolerance test in nondiabetic and GK diabetic rats injected with PBS or in GK rats administered with benzylamine/vanadate. GK diabetic rats showed a marked deficiency in insulin secretion (Fig. 4A); in contrast, previous administration of benzylamine/vanadate showed a normalization of the profile of insulin concentrations after a glucose challenge (Fig. 4A and B). The acute treatment with vanadate alone showed no effect on plasma insulin concentrations (Fig. 4B). Also, no effect of benzylamine/vanadate on plasma insulin concentrations was detected in nondiabetic rats (12). To test whether benzylamine/vanadate acted directly on pancreatic islets, isolated islets from nondiabetic Wistar and GK diabetic rats were incubated in the presence of benzylamine and/or vanadate, and the response to glucose was evaluated. Insulin secretion measured at 16.7 and 5.5 mmol/l glucose was defective in pancreatic islets from GK diabetic rats compared with those from controls (values of insulin secretion at 16.7 mmol/l glucose were 81 ± 12 and 50 ± 5 μU islet/90 min, and at 5.5 mmol/l glucose was 51 ± 8 and 34 ± 4 μU islet/90 min in control and GK rats, respectively); in keeping with the in vivo observations, benzylamine/vanadate markedly enhanced insulin secretion in pancreatic islets from GK diabetic rats but not from nondiabetic rats (Fig. 4C and D). In addition, the effect of benzylamine/vanadate on insulin secretion was blocked by semicarbazide (Fig. 4D). In keeping with these observations, we detected a similar abundance of SSAO/VAP-1 protein in extracts from isolated pancreatic islets compared with extracts from total pancreas; under these conditions, IAPP (a marker of the pancreatic islets) was highly enriched in pancreatic islets compared with total pancreas, whereas amylase abundance (a marker of exocrine pancreas) was very low in pancreatic islets (Fig. 4E). The abundance of SSAO/VAP-1 protein was much higher in adipose tissue than in pancreatic islets (Fig. 4E). In addition, SSAO activity was nearly 300-fold greater in extracts from adipocytes than in pancreatic islets (data not shown). In all, our results show that endocrine islets express SSAO/VAP-1 protein but at lower levels than in adipose cells.

As to the nature of the molecules that promotes the effects on adipose cells, skeletal muscle, and pancreatic islets in response to benzylamine/vanadate, we have ruled out the involvement of changes in gene expression of adipocytokines and, at least, for skeletal muscle they only require the presence of benzylamine, vanadate, and the enzymatic activity of human recombinant SSAO. Next, we tested the possible formation of peroxovanadium compounds, based on the fact that SSAO catalysis generates hydrogen peroxide, and this is known to react with vanadate, generating peroxovanadium. Vanadate was detected by NMR assays as a number of peaks at approximately −560 and −580 ppm (Fig. 5A), which was independent of the presence of human recombinant SSAO (data not shown). Under our assay conditions, the presence of benzylamine caused no NMR signals (data not shown). Incubation of vanadate and hydrogen peroxide generated monoperoxovanadium (identified as a peak at −630 ppm) and triperoxovanadium (identified as a peak at −730 ppm) compounds detected by NMR (Fig. 5A) (27). Furthermore, the incubation of vanadate, benzylamine, and human recombinant SSAO also caused a substantial formation of monoperoxovanadium and triperoxovanadium compounds in NMR assays (Fig. 5A). Further support for the in vivo generation of peroxovanadium complexes was the observation that the combination of benzylamine and vanadate caused a very potent stimulation of protein tyrosine phosphorylation in extracts from isolated rat adipocytes. The effect of benzylamine/vanadate was much greater than that caused by insulin; it was not observed in the presence of benzylamine or vanadate alone, and it was blocked by semicarbazide (Fig. 5B). The effect of benzylamine/vanadate was similar to that caused by peroxovanadate (Fig. 5B). This is consistent with the powerful inhibitory effect of peroxovanadium on protein tyrosine phosphatases (28,29). Moreover, we have found that the incubation of adipocytes with benzylamine and vanadate caused a marked inhibition of protein tyrosine phosphatase activity, whereas incubation with benzylamine or vanadate alone had no effect on protein tyrosine phosphatase (Fig. 5C). The effect of benzylamine/vanadate on protein tyrosine phosphatase was similar to the effect caused by 100 μmol/l peroxovanadate (Fig. 5C).

In this study, we have demonstrated that the utilization of the SSAO/VAP-1 substrate benzylamine in combination with ineffective doses of vanadate has very remarkable pharmacological properties in the treatment of type 2 diabetes. Thus, both acute and chronic treatment with benzylamine and vanadate stimulates glucose utilization in adipose cells and in skeletal muscle and reverses muscle insulin resistance in GK diabetic rats. In addition, benzylamine and vanadate also acutely ameliorate insulin secretion in GK diabetic rats. Therefore, treatment with benzylamine and vanadate shows two types of antidiabetic effects: 1) metabolic effects in adipose and muscle cells that lead to enhanced glucose utilization and 2) stimulatory effects on insulin secretion. This type of double action is ideal in the treatment of type 2 diabetes, which is characterized by insulin resistance affecting peripheral tissues and defective insulin secretion.

The biological effects of benzylamine/vanadate require an intact SSAO activity. This is an important consideration because benzylamine may also be metabolized via monoamino oxidase activity (14). The experimental evidence in support of the view that SSAO activity is essential for the effects of benzylamine/vanadate is as follows: 1) the effects of benzylamine/vanadate on glucose transport in isolated rat adipocytes are blocked with semicarbazide but not with pargyline (5,10); 2) the acute effects of benzylamine/vanadate on glucose disposal in nondiabetic and streptozotocin-induced diabetic rats are blocked in rats previously treated with semicarbazide (12); 3) the effects of benzylamine/vanadate on skeletal muscle glucose transport (in coincubation studies with adipose tissue explants) is prevented in the presence of semicarbazide (this study); 4) the effect of benzylamine/vanadate on insulin secretion in isolated rat pancreatic islets is also prevented in the presence of semicarbazide (this study); and 5) the effect of benzylamine/vanadate on protein phosphorylation on tyrosine residues and activation of IRS-1 and -3 in adipocytes is prevented in the presence of semicarbazide (10 and this study).

In this study we only used benzylamine as a SSAO/VAP-1 substrate. However, current available evidence indicates that other SSAO substrates also show similar biological effects, which correlates with their capacity of amines to be oxidized by SSAO activity (13). The evidence for this is as follows: other substrates of SSAO, such as tyramine, methylamine, n-decylamine, β-phenylethylamine, histamine, or n-acetylputrescine stimulate glucose transport in combination with vanadate in isolated rat adipocytes (10) or alone in isolated human adipocytes (13); 2) chronic incubation in the presence of benzylamine, tyramine, methylamine, or β-phenylethylamine stimulated adipose cell differentiation in 3T3 F442A preadipocytes (14,30); 3) the acute administration of tyramine (another SSAO substrate) enhances glucose utilization as assessed by glucose tolerance test in streptozotocin-induced diabetic rats, and this effect was blocked in rats chronically treated with semicarbazide, which shows completely inactive SSAO activity (31); and 4) the chronic administration for 2 weeks of tyramine together with vanadate enhanced glucose utilization as assessed by glucose tolerance test in streptozotocin-induced diabetic rats (31).

Diabetic GK rats show hyperglycemia, muscle insulin resistance, and defective insulin secretion despite normal circulating concentrations of insulin (22–25). Chronic treatment of diabetic GK rats with benzylamine/vanadate caused amelioration of plasma glucose, enhanced glucose transport in adipocytes, and improved insulin responsiveness in skeletal muscle, which occurred in the absence of alterations in circulating insulin. It is likely that amelioration of plasma glucose by benzylamine/vanadate is due to enhanced glucose transport in adipocytes and in skeletal muscle caused by benzylamine/vanadate together with the improvement in the capacity of muscle to respond to insulin. It is also likely that the amelioration in glucose-induced insulin secretion is involved in the effects of benzylamine/vanadate on glucose homeostasis. Benzylamine/vanadate did not alter plasma insulin concentrations in diabetic GK rats, which was not surprising because they were already normal under untreated conditions.

We have also demonstrated the nature of the mechanisms implicated in the metabolic effects associated with the utilization of SSAO/VAP-1 substrates, such as benzylamine, in combination with vanadate. Benzylamine metabolism catalyzed by SSAO generates hydrogen peroxide, and the combination of hydrogen peroxide and vanadate generates peroxovanadium compounds, a process that does not require enzymatic catalysis. This is the explanation by which the presence of benzylamine, human recombinant SSAO and vanadate produces peroxovanadium compounds. The generation of peroxovanadium compounds is relevant because they inhibit protein tyrosine phosphatase activity. In consequence, it is likely that the in vivo treatment with benzylamine and vanadate results in the production of peroxovanadium compounds. This type of compounds combination (SSAO substrate and vanadate) permits the generation of protein tyrosine phosphatase inhibitors in key sites such as adipose tissue or pancreatic islets, which can display a variety of local effects and which can also act in neighbor tissues such as skeletal muscle. The approach of generating inhibitors of protein tyrosine phosphatase activity is sensible, because protein tyrosine phosphatases are implicated in the dephosphorylation of signaling molecules that terminate insulin signaling. In addition, alteration in protein tyrosine phosphatase activity may contribute to insulin-resistant states. The expression of specific protein tyrosine phosphatases, including leukocyte antigen-related phosphatase (LAR), protein tyrosine phosphatase-1B (PTP1B), and src-homology phosphatase-2 (SHP2) are increased in muscle and adipose tissue from obese subjects or obese rodents (32–34). Enhanced PTP1B and SHP2 have also been detected in muscle from diabetic rats (32,35). Overexpression of the LAR in muscle causes insulin resistance characterized by hyperinsulinemia and reduced whole-body glucose disposal and muscle glucose uptake (36). Similarly, overexpression of PTP1B in adipose tissue impairs insulin-stimulated glucose transport (37,38). This suggests that increased LAR or PTP1B activity may contribute to the pathogenesis of insulin resistance. Evidence indicating that mice lacking the PTP1B gene show increased insulin sensitivity and resistance to the development of obesity induced by a high-fat diet (39) has provided additional support to the view that inhibitors of protein tyrosine phosphatase activity may have beneficial properties.

Our data indicate that SSAO/VAP-1 is a drug target for type 2 diabetes and that the combination of SSAO substrates and low doses of vanadate efficiently normalizes glycemia and reverses muscle insulin resistance in GK diabetic rats. Moreover, our results are consistent with a model (Fig. 6) in which SSAO/VAP-1 activity, present in adipocyte cell surface or in pancreatic islets, generates hydrogen peroxide, and this promotes the formation of peroxovanadium complexes in the presence of vanadate. The peroxovanadium generated inhibits protein tyrosine phosphatase activity, which causes activation of protein tyrosine phosphorylation in adipose cells (Fig. 6). We propose that the peroxovanadium generated in the plasma membrane from adipose cells also reaches muscle fibers found in the neighborhood. This cascade of events is instrumental to acutely stimulate glucose transport and lipogenesis in adipose cells, glucose transport in skeletal muscle, and glucose-stimulated insulin secretion in pancreatic islets from GK rats (Fig. 6). This proposal is in keeping with the potent insulin-mimicking activity of peroxovanadate previously detected in adipocytes and in skeletal muscle (40–45). Our model also agrees with previous observations indicating that peroxovanadate regulates insulin secretion (46).

This study was supported by research grants from the Dirección General de Investigación Científica y Técnica (PM98/0197), the Ministerio de Ciencia y Tecnología (SAF 2002-02125), Grant 1999SGR 00039 from Generalitat de Catalunya, the Fondo de Investigaciones Sanitarias (grant 00/0125), the European Commission (Quality of Life grant QLG-CT-1999-00295), “Accords INSERM/CSIC,” “Actions Integrées Franco-Espagnoles PICASSO,” and COST B17 Action and Fundació Marató de TV3 (300720). A.A. is a recipient of a predoctoral fellowship from the Universitat de Barcelona. L.M. was a recipient from a Marie Curie postdoctoral fellowship from the European Union (ERBFMBICT985088).

We thank Drs. Miguel Feliz and Francisco Cárdenas (NMR facilities, Serveis Científic-Tècnics from University of Barcelona) for excellent assistance in NMR assays. We also thank Robin Rycroft for his editorial support, Judith Garcia for technical assistance, and Dr. Roser Casamitjana for help in insulin determinations.