Oral Vanadium Enhances the Catabolic Effects of Central Leptin in Young Adult Rats
http://www.100md.com
《内分泌学杂志》
Geriatric Research, Education
Clinical Center, Department of Veterans Affairs Medical Center (J.W.), Gainesville, Florida 32608-1197
Department of Pharmacology and Therapeutics, University of Florida College of Medicine (J.W., M.K.M., P.J.S.), Gainesville, Florida 32608
Abstract
Recently, vanadium has been shown to enhance leptin signal transduction in vitro. We hypothesized that chronic oral administration of an organic vanadium complex would enhance both leptin signaling and physiological responsiveness in vivo. Three-month-old F344 x Brown Norway male rats were provided a solution containing escalating doses of vanadyl acetoacetonate (V), peaking at 60 mg/liter elemental vanadium in drinking water on the 11th d of V treatment. Although V treatment tended to suppress weight gain, absolute body weights did not significantly differ between groups after 62 d of treatment. At this point, a permanent cannula was placed into the left lateral ventricle of all animals. The cannula was connected to a sc minipump providing either 5 μg/d leptin or artificial cerebral spinal fluid (ACSF) control solution. This yielded four groups: C-ACSF, C-leptin, V-ACSF, and V-leptin. During the ensuing 26 d, weight gain was similar in C-ACSF and V-ACSF. As expected, leptin caused dramatic weight loss in C-leptin, but leptin-induced weight loss was 43% greater in V-leptin. V enhanced leptin-induced signal transducer and activator of transcription-3 phosphorylation in the hypothalamus, whereas V alone had no effect. V also augmented the leptin-induced increase in brown adipose tissue uncoupling protein-1. The effects of vanadium on responsiveness to a submaximal dose of leptin (0.25 μg/d) were also evaluated, yielding qualitatively similar results. These data demonstrate, for the first time, that chronic V administration enhances the weight-reducing effects of centrally administered leptin in young adult animals, and the mechanism appears to involve enhanced leptin signal transduction.
Introduction
THE ELEMENT VANADIUM is found in most living organisms, but its essentiality in mammals has not been established (1). Various forms of the element have long been known to effect insulin sensitivity. In 1985, vanadate (V5+) was first shown to have an insulin-like action (2). It subsequently became clear that vanadium acted more as an insulin-sensitizing agent, because it was only effective in the presence of at least small amounts of insulin. Although the insulin-sensitizing effect of vanadium compounds has been established in humans (3) and a variety of animal models (1, 4), the mechanism of this effect remains in question. However, the mechanism may involve the ability of the element to inhibit the dephosphorylation of key intracellular signaling molecules. Vanadate potently inhibits purified protein tyrosine phosphatases (PTPases) in vitro (5). In studies of mouse adipocyte cultures, vanadium was shown to increase insulin receptor substrate-1 (IRS-1) phosphorylation (6). Increased tyrosine phosphorylation of IRS-1 and other molecules is most likely secondary to a decrease in PTPase activity. Consistent with this, it was recently demonstrated that oral vanadium reduced the activity of PTP1B in skeletal muscle of fatty Zucker rats by 25% (7). PTP1B is a specific tyrosine phosphatase that inhibits both leptin and insulin signal transduction (8, 9). Given the overlap between leptin and insulin signal transduction (10) as well as the importance of tyrosine phosphorylation of second messengers in the leptin signaling cascade, we reasoned that vanadium would enhance leptin signaling and sensitivity. To test our hypothesis, we added an organic complex of vanadyl acetoacetonate (V) to the drinking water of young adult F344 x Brown Norway male rats and then measured the biochemical and physiological responses to chronic intracranial leptin infusion.
Materials and Methods
Animals
Three-month-old male Fischer 344 x Brown Norway rats were obtained from Harlan Sprague Dawley (Indianapolis, IN). Upon arrival, rats were examined and remained quarantined for 1 wk. Animals were individually caged with a 12-h light, 12-h dark cycle (lights on from 0700–1900 h). Animals were cared for in accordance with the principles of the National Institutes of Health Guide to the Care and Use of Experimental Animals.
Experimental design
Experiment 1.
Three-month-old F344 x Brown Norway male rats were provided a solution containing escalating doses of V, peaking at 60 mg/liter elemental vanadium/d on the 11th d of V treatment. The V solution contained hypotonic saline 0.045%) and was sweetened with 0.125% saccharine to offset taste aversion. Control (C) animals were maintained on tap water. Hypotonic saline and saccharine were not provided to control animals, because it was observed in a preliminary experiment that this significantly increased fluid intake over baseline. After 62 d of V treatment (in the V group), a permanent cannula was placed into the left lateral cerebroventricle of all animals. The cannula was connected to a sc minipump providing either 5 μg/d leptin or ACSF control solution. This yielded four groups: C-ACSF, C-leptin, V-ACSF, and V-leptin. Animals were killed 28 d after cannulation for tissue analysis.
Experiment 2.
A second experiment was conducted to determine the effects of V treatment on physiological responsiveness to a submaximal dose of leptin. A submaximal dose of leptin for central infusion was first determined in a dose-response experiment in which animals were given 0, 0.05, 0.25, 1.0, or 5 μg/d leptin, intracerebroventricularly (icv). It was determined that 0.25 μg/d caused intermediate physiological effects. Three-month-old F344 x Brown Norway male rats were again pretreated with V as in experiment 1, then implanted with an intracranial (left lateral ventricle) cannula connected to an sc minipump providing 0.25 μg/d leptin. In this experiment animals were killed on d 21 in hope of preventing complete lipopenia in the C-leptin group (which makes it difficult to evaluate the synergistic effects of vanadium, if any, on fat loss).
Glucose tolerance test
After an overnight fast, rats were administered 2 g/kg glucose ip. Blood was drawn from the tail at baseline and 30, 60, and 120 min after glucose injection. Animals were not anesthetized. At each interval, approximately 250 μl blood was drawn from the tip of the tail. A single drop of tail blood was used to measure glucose via a glucose meter (OneTouch SureStep, Johnson & Johnson, Inc., Milpitas, CA). Remaining blood was centrifuged at 1300 x g for 10 min to yield serum for assessment of insulin by RIA (Linco Research, Inc., St. Charles, MO). The area under the curve (AUC) for glucose was calculated by subtracting the baseline serum glucose level in controls. This correction was not practical for insulin AUC, because baseline levels were not similar in controls and vanadium-treated animals. Thus, AUC for insulin was calculated from the x-axis (y = 0).
Intracranial cannulation and leptin infusion
Rats were anesthetized with 60 mg/kg pentobarbital, and heads were prepared for surgery. Animals were placed into a stereotaxic frame, and a small incision (1.5 cm) was made over the midline of the skull to expose the landmarks of the cranium (bregma and lambda). A cannula (Durect Corp., Cupertino, CA) was placed into the lateral ventricle using the following coordinates: 1.3 mm posterior to bregma, 1.9 mm lateral to the midsaggital suture, and to a depth of 3.5 mm. The cannula was anchored to the skull using acrylic dental cement. A sc pocket on the dorsal surface was created using blunt dissection, and the osmotic mini pump (Durect Corp.) was inserted. These pumps (model 2004, ALZET; Durect Corp.) infuse 0.25 μl fluid/h for a minimum of 28 d and have a total capacity of 200 μl. Thus, in experiment 1 the pumps in the leptin group were filled with a solution containing leptin in ACSF (0.833 μg/μl) to provide 5 μg/d leptin. In experiment 2 (submaximal dose of leptin), this was reduced to 0.0417 μg/μl, providing 0.25 μg/d leptin. After filling the minipumps with ACSF or ACSF/leptin solution, they were incubated in sterile saline at 37 C for 36 h before implantation. A catheter tube was employed to connect the cannula to the osmotic minipump flow moderator. The incision for the minipump was then closed with sutures. Rats were kept warm during the manipulations and until fully recovered.
Tissue harvesting
Anesthetized rats were killed by cervical dislocation. Blood was collected by cardiac puncture, and serum was harvested by a 10-min centrifugation in serum separator tubes. The circulatory system was perfused with 20 ml cold saline. Perirenal (PWAT), retroperitoneal (RTWAT), and epididymal (EWAT) white adipose tissues; interscapular brown adipose tissue (iBAT); and hypothalami were excised, weighed, and immediately frozen in liquid nitrogen. The hypothalamus was removed by making an incision medial to piriform lobes, caudal to the optic chiasm, and anterior to the cerebral crus to a depth of 2–3 mm. Tissues were stored at –80 C until analysis.
Signal transducer and activator of transcription-3 (STAT3)/phosphorylated STAT3 (P-STAT3) assay
These methods were described in detail previously (11). Briefly, hypothalamus was sonicated in 10 mM Tris-HCl (pH 6.8), 2% sodium dodecyl sulfate, and 0.08 μg/ml okadaic acid plus protease inhibitors (phenylmethylsulfonylfluoride, benzamidine, and leupeptin). Sonicate was diluted and quantified for protein using a detergent-compatible Bradford assay. Samples were boiled and separated on a 7.5% agarose/Tris-HCl gel (Bio-Rad Laboratories, Hercules, CA) and electrotransferred to a nitrocellulose membrane. Immunoreactivity was assessed with an antibody specific to P-STAT3 (antibody kit from New England Biolabs, Beverly, MA). Immunoreactivity was visualized by chemiluminescence detection (Amersham Biosciences, Piscataway, NJ) and quantified by video densitometry (Bio-Rad Laboratories). After P-STAT3 quantification, membranes were stripped of antibody with Immunopure (Pierce Chemical Co., Rockford, IL), and immunoreactivity was reassessed using a total STAT3 antibody.
Leptin mRNA levels in WAT
RTWAT (300 mg/sample) was sonicated in guanidine buffer, phenol extracted, and isopropanol precipitated using a modification of the method of Chomczynski and Sacchi (12). Isolated RNA was resuspended in ribonuclease-free water and quantified by spectrophotometry. Integrity was verified using 1% agarose gels stained with ethidium bromide. For dot-blot analysis, 100 ng total RNA was immobilized by loading directly onto a nylon membrane in triplicate using a dot-blot apparatus (Bio-Rad Laboratories, Richmond, CA). Membranes were baked in a UV cross-linking apparatus. Membranes were then prehybridized in 10 ml QuickHyb (Stratagene, La Jolla, CA) for 30 min, followed by hybridization in the presence of a labeled probe for leptin mRNA and 100 μg salmon sperm DNA. The probe used to detect leptin mRNA was a 33-mer antisense oligonucleotide (5'-GGTCTGAGGCAGGGA CAGCTCTTGGAGAAGGC) end labeled using terminal deoxynucleotidyl transferase (Promega Corp., Madison, WI). We previously demonstrated by Northern analysis that this probe binds to a single mRNA species of 4.1 kb (11). After hybridization for 2 h at 65 C, the membranes were washed and exposed to a phosphorimaging screen for 72 h. The screen was then scanned using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) and analyzed by Image-Quant software (Molecular Dynamics). Data are expressed as mRNA per total RTWAT pad.
Serum leptin
Serum leptin was measured using a mouse leptin ELISA kit (Crystal Chemicals, Inc., Chicago, IL) on blood harvested at the time of death by cardiac puncture.
Statistical analysis
All data are expressed as the mean ± SEM. The level was set at 0.05 for all analyses. Comparisons of mean food intake and body weight gain in control vs. vanadium-treated were animals were made by a repeated measures, two-way ANOVA, with time and treatment as factors. After minipump implantation, food and body mass data were analyzed by three-way ANOVA, with time, leptin, and vanadium as factors (experiment 1). Two-way analyses (leptin x vanadium), in which we examined the change in body mass during the treatment period and cumulative food intake during the treatment period, were also performed (experiments 1 and 2). All end-point serum and tissue data were analyzed by a two-way analysis, with leptin and vanadium as factors. When only main effects were significant, relevant pairwise comparisons were made using the Bonferroni multiple comparison method, with the error rate corrected for the number of contrasts (13). When there was an interaction, factors were separated, and an additional one-way ANOVA was applied with a Bonferroni multiple comparison post hoc test. When separation of factors resulted in only two population means to compare, the one-way ANOVA was replaced by Student’s t test. PRISM software version 4.0 (GraphPad, Inc., San Diego, CA) was used for all statistical analysis and graphing, with the exception of the three-way ANOVA. QuickCalc (GraphPad; www.graphpad.com) was used for post hoc analysis of two-way ANOVAs. GraphPad QuickCalc uses the Bonferroni correction for multiple comparisons. The three-factor ANOVAs (body mass and food intake after minipump implantation in experiment 1) were performed with the assistance of the VassarStats statistical package developed by Dr. Richard Lowry (Vassar College, Poughkeepsie, NY), and verified with SPSS Base 13.0 (SPSS, Inc., Chicago, IL).
Results
Experiment 1: vanadium pretreatment
Food intake and body mass.
Three-month-old F344 x Brown Norway male rats were provided a solution containing escalating doses of V, peaking at 60 mg/liter elemental vanadium/d on the 11th d of V treatment. To reduce the risk of taste aversion, 0.125% saccharine and 0.45% NaCl were added to the V solution. Control animals were maintained on tap water, because it was determined that saccharine and NaCl significantly increase fluid consumption (in the absence of V). During the first 42 d of V treatment, food intake per day averaged 5.7% less in V-treated animals than in controls (19.13 ± 0.31 vs. 20.29 ± 0.41 g/d, respectively; P < 0.05). However, this modest effect disappeared after d 42 (mean food intake on d 42–60 was 18.90 ± 0.38 g/d in controls vs. 19.02 ± 0.36 g/d in V-treated animals). At the start of V treatment, body weight was similar in V and control groups. Vanadium had a slight suppressive effect on weight gain during 60 d of V treatment (86.6 ± 5.0 g gained in controls vs. 69.2 ± 5.2 g in V-treated animals; P < 0.05; Fig. 1), but absolute body weights were not significantly different between groups after the V treatment period (371.9 ± 5.4 g in control vs. 358.1 ± 6.9 g in V group; Fig. 1).
Fluid and vanadium intake.
During the period of escalating the dose of V, fluid intake was variable in the V group, first increasing and then decreasing, with a nadir of 79% of control intake on d 18 (not shown). However, fluid intake in V-treated animals returned to baseline after this adjustment period, and there was no difference in fluid intake between V and control animals during the 30 d before leptin administration. During this period (30 d before leptin administration), fluid intake averaged 16.8 ± 0.45 ml/d in the V group. This corresponds to 1.01 ± 0.027 mg elemental vanadium/rat·d, or 2.66 ± 0.078 mg/kg just before the leptin infusion.
Glucose tolerance test (GTT).
Approximately 6 wk after commencing V treatment, a GTT was performed. Serum glucose and insulin were measured at baseline and 30, 60, and 120 min after injecting glucose (2 g/kg) ip. Glucose clearance was enhanced in V-treated animals. The effect of V on serum glucose during the GTT was significant by repeated measures two-way ANOVA (Fig. 2). The serum glucose area under the curve was reduced by 24.3% in the V group (baseline set at time zero serum glucose in controls; Fig. 2). V also had a significant suppressive effect on fasting serum insulin (1.27 ± 0.15 vs. 0.758 ± 0.11 ng/ml; P < 0.05) and on insulin secretion during the GTT (total area under the curve, 194.2 min·ng/ml in controls vs. 98.1 min·ng/ml in V animals; baseline set at y = 0).
Leptin infusion
Body mass, food intake, and fluid intake.
Before minipump implantation, body mass was similar in the four groups, but tended to be lower in the V-treated animals, as described above (C-ACSF, 373.0 ± 8.2 g; C-leptin, 370.8 ± 7.6; V-ACSF, 357.6 ± 10.5; V-leptin, 358.5 ± 9.7). During the ensuing 26 d, weight gain was similar in C-ACSF and V-ACSF (6.7 ± 3.8 vs. 4.2 ± 2.1 g, respectively; Fig. 3). Although leptin caused significant weight loss in both V animals and controls, V-leptin lost 42.7% more weight over 26 d than C-leptin (97.5 vs. 68.3 g; P < 0.05; Fig. 3).
Twenty-four-hour food intake was similar in the four groups before minipump implantation (C-ACSF, 19.95 ± 0.76 g; C-leptin, 20.17 ± 0.74; V-ACSF, 19.21 ± 0.66; V-leptin, 18.88 ± 0.56 g; averaged over 48 h). Food intake remained similar in the two ACSF-treated groups (C-ACSF and V-ACSF) after minipump implantation (Fig. 4). Leptin caused significant reductions in feeding in both C-leptin and V-leptin groups over the ensuing 26 d (Fig. 4). Mean food intake decreased by 32.5% in C-leptin and by 34.8% in V-leptin (percent decrease is based on average intake during 26 d after pump implantation relative to baseline), but V did not significantly affect leptin-induced suppression of food intake (Fig. 4). Fluid intake was unaffected compared with pretreatment values in C-leptin and V-leptin (data not shown).
Serum leptin and insulin.
Twenty-eight days of central leptin infusion caused significant 87% and 70% decreases, respectively, in serum leptin in controls (nonvanadium treated) and vanadium-treated animals (Table 1). Surprisingly, vanadium also caused a significant (53%) reduction in serum leptin (Control-ACSF vs. V-ACSF; Table 1). This was unexpected, because intraabdominal adiposity and body mass were similar in C-ACSF and V-ACSF, and in vitro studies suggest that V enhances leptin expression (14). In addition, both vanadium and leptin caused significant reductions in fasting serum insulin (Table 1).
Intraabdominal adiposity and leptin expression in WAT.
Leptin caused a near-complete disappearance of intraabdominal adipose tissue in both C-leptin and V-leptin. On the average, less than 2% of the abdominal adiposity present in ACSF-infused animals could be removed from leptin-treated animals regardless of vanadium treatment. RTWAT and PWAT were completely absent from five of six C-leptin rats and in six of seven V-leptin rats (data not shown). A very small amount of EWAT remained in three of six C-leptin rats, but in none of the V-leptin rats (data not shown). Leptin expression in the RTWAT of C-ACSF and V-ACSF was similar (100 ± 14.4 and 76.65 ± 16.5 arbitrary units/mg RNA, respectively). Due to complete leptin-induced catabolism of RTWAT, leptin expression could not be evaluated in C-leptin and V-leptin.
STAT3 phosphorylation.
As expected, chronic intracranial leptin infusion caused a significant 35% increase in P(Y-705)-STAT3 immunoreactivity in the medial basal hypothalamus in C-leptin vs. C-ACSF (Fig. 5A). Although V alone did not affect P-STAT3 levels, V did enhance leptin-induced STAT3 phosphorylation (114% increase in V-leptin vs. V-ACSF). Hypothalamic P-STAT3 immunoreactivity was 70% higher in V-leptin vs. C-leptin (Fig. 5A; P < 0.05). Chronic leptin treatment caused a mild elevation in total STAT3 in the hypothalamus, whereas V had no effect (Fig. 5B).
Uncoupling protein-1 (UCP-1) in BAT.
Consistent with our previous findings, leptin caused an approximately 2-fold increase in UCP-1 levels per iBAT pad (C-ACSF vs. C-leptin, P < 0.05; Fig. 6). When expressed as UCP-1 per milligram of iBAT protein, leptin caused a 3.98-fold elevation in UCP-1 (C-ACSF vs. C-leptin, P < 0.05; data not shown). Vanadium caused a 2-fold elevation in UCP-1 per BAT pad (V-ACSF vs. C-ACSF, P < 0.05). The effects of vanadium and leptin on UCP-1 appeared to be additive, because iBAT UCP-1 was 2-fold greater in V-leptin than in rats treated with either vanadium alone or leptin alone (P < 0.05; Fig. 6).
Experiment 2
Leptin infusion dose-response.
The dose of leptin provided in the first experiment appears to evoke maximal physiological responses, thus rendering it difficult to fully evaluate the potential of V to enhance the responses to leptin. Therefore, we employed a submaximal dose of leptin in experiment 2. First, a preliminary experiment was conducted to determine the dose of leptin that would be provided at the end of V pretreatment. Rats were infused with 0.05, 0.25, 1, or 5 μg/d leptin for 20 d (Fig. 7A). The rats given 0.25 μg/d leptin lost significantly more weight than controls, but lost significantly less weight than rats given the 5 μg/d dose provided in experiment 1 (Fig. 7A). Cumulative food intake during the first 20 d of leptin infusion was significantly lower in rats centrally infused with 0.25 μg leptin/d vs. controls, but was significantly greater than that in rats given 5 μg leptin/d (Fig. 7B). Moreover, although 5 μg leptin/d caused complete catabolism of RTWAT and PWAT in this dose-response experiment, all rats given 0.25 μg leptin/d had a small amount of these depots remaining (mean mass RTWAT + PWAT, 25.7 ± 0.9 mg in 0.25 μg/d group; data not shown). The next lowest dose tested, 0.05 μg/d, did not have a significant effect on body weight (Fig. 7A), food intake (Fig. 7B), or central adiposity (data not shown), and the next higher dose, 1.0 μg/d, was too efficacious. Thus, it was determined that 0.25 μg/d was the minimum efficacious, submaximal dose for chronic icv leptin infusion among the doses tested. This is the dose that was chosen for the experiment to test the interaction of vanadium and a submaximal central dose of leptin.
Vanadium pretreatment.
In experiment 2, the vanadium pretreatment protocol and physiological effects were virtually identical with those in experiment 1. At the end of the vanadium pretreatment period, body weights were not significantly different in controls (318.4 ± 2.1 g) and V-treated animals (313.5 ± 3.8 g).
Leptin infusion
Body mass and food intake.
Changes in body mass during the lower-dose central leptin infusion were qualitatively similar to what was observed in the initial high-dose experiment. Leptin caused a significant reduction in body mass in both controls and vanadium-treated rats, with the V-leptin group losing 43.6% more weight than the C-leptin group during the 21-d treatment period (P < 0.01; Fig. 8A). The food intake data did differ somewhat from the high-dose group (experiment 1). First, the leptin-induced reduction in food intake took much longer to appear with the 0.25 μg/d dose. Although a reduction in food intake was readily apparent by d 4 in experiment 1, a reduction in food intake in the 0.25 μg/d-treated animals in experiment 2 did not appear until d 8–10 (data not shown). Another difference between experiments is that vanadium significantly enhanced the leptin-induced inhibition in feeding in experiment 2. The V-leptin group consumed significantly less food than the C-leptin group between the 8th and 20th d of leptin infusion (Fig. 8B). Vanadium alone, however, did not affect food intake (no difference in food intake between C-ACSF and V-ACSF; Fig. 8B). This suggests that vanadium, at the dose provided, does not have significant anorectic properties by itself; rather, it interacts with the effects of submaximal leptin to enhance both feeding inhibition and weight loss.
Intraabdominal adiposity.
Central leptin at a dose that was clearly submaximal with respect to reductions in body mass and food intake was still able to cause near-complete catabolism of intraabdominal WAT. However, four of six C-leptin rats had small strips of RTWAT and EWAT (averaging 120.2 ± 26.9 mg in these four rats), whereas zero of six V-leptin rats had visible RTWAT or EWAT. When the intraabdominal adiposity of V-leptin and C-leptin were compared categorically for the presence or absence of excisable fat depots, the difference between groups was statistically significant (Table 2).
Discussion
The primary purpose of this investigation was to determine whether vanadium could enhance leptin signaling and the physiological effects of chronic icv leptin infusion in young adult rats. First, we wanted to determine whether the dose and form of V we provided were able to influence in vivo insulin sensitivity. Consistent with the reported effects of vanadium on glucose homeostasis in diabetic models (4, 6), chronic V treatment improved glucose tolerance and lowered fasting serum insulin levels in our young adult F344 x Brown Norway male rats. However, the present results are unique in that they demonstrate that V can enhance markers for insulin sensitivity even in a nonobese, nondiabetic model.
As hypothesized, V did enhance the catabolic effect of chronic intracranial leptin infusion. V treatment resulted in significantly more leptin-induced weight loss without significantly affecting leptin-induced suppression of energy intake when leptin was infused at 5 μg/d. This suggests that V acted primarily by augmenting leptin’s effects on energy expenditure. However, when the leptin dose was lowered to 0.25 μg/d, vanadium enhanced both leptin-induced weight loss and feeding inhibition. Thus, it seems plausible that the apparent lack of an effect of vanadium on the feeding-inhibiting properties of leptin at the 5 μg/d icv dose was secondary to a basement effect, where compensatory survival (orexigenic) pathways limited additional feeding inhibition.
We hypothesized that the mechanism by which V may enhance leptin responsiveness would involve enhanced leptin signal transduction. As discussed in the introduction, this hypothesis was based in part upon the recently reported effects of V on insulin signal transduction (6) and the overlap between leptin and insulin signaling. Moreover, Kita et al. (15) demonstrated that pretreatment of Chinese hamster ovary cells expressing ObRb with 100 μM orthovanadate dramatically enhanced leptin-induced STAT3 and Janus kinase-2 (JAK2) phosphorylation. Given the critical role of STAT3 signaling with respect to leptin-induced suppression of energy intake and thermogenesis (16), we examined the effects of V on this pathway in vivo. As surmised, V amplified leptin’s effects on STAT3 phosphorylation in our present study. However, V alone had no effect on basal STAT3 phosphorylation. Our results suggest that in vivo, a P-STAT3-inducing stimulus (i.e. leptin infusion) must be concomitantly administered for V to exert an effect on this pathway.
If V enhances physiological leptin responsiveness by enhancing leptin signal transduction through STAT3 and possibly other pathways, what is the mechanism behind the enhanced signaling Although we did not address this question directly in the present study, it is plausible that the mechanism involves an inhibition of PTPase activity. Leptin signal transduction relies upon the tyrosine phosphorylation of not only the receptor itself, but also messengers, including JAK2 and STAT3. As such, PTPases inhibit leptin signal transduction. One such phosphatase is Src homology domain 2-containing phosphatase-2 (17). Prevention of Src homology domain 2-containing phosphatase-2 recruitment via a point mutation enhances the expression of genes under the control of STAT3 (17). Another tyrosine phosphatase believed to inhibit JAK2/STAT3 signaling through the leptin receptor is PTPase-1B (18). This phosphatase is of particular clinical interest because it is also an inhibitor of insulin signal transduction (19). PTPase-1B is believed to block JAK2 activation (20), inhibiting both the JAK2-STAT3 and IRS-phosphotidylinositol 3-kinase cascades. The latter pathway is involved in both insulin and leptin signaling. Transgenic mice lacking PTPase-1B are lean and resistant to diet-induced obesity and have enhanced insulin sensitivity (21). Similarly, mice treated with PTPase-1B antisense oligonucleotides have enhanced insulin sensitivity and reduced adiposity (19). Mice overexpressing PTPase-1B, in contrast, are insulin resistant (9). Moreover, overexpression of PTPase-1B in a hypothalamic cell line lowers basal JAK2 phosphorylation and blocks the normal leptin-induced up-regulation of suppressor of cytokine signaling-3 expression (22). Taken together, this evidence suggests that inhibiting the activity of PTPase-1B and related tyrosine phosphatases could be an effective strategy to enhance leptin sensitivity. As discussed in the introduction, V is an established inhibitor of PTPases and has been shown to do so in vivo (7).
Leptin signaling in the hypothalamus causes a negative shift in energy balance by reducing food intake and increasing energy expenditure (11). Given that V enhanced leptin-induced weight loss with an insignificant effect on food intake in experiment 1 (high dose of leptin), it follows that V probably enhanced leptin’s effects on energy expenditure. Therefore, it is not surprising that V augmented the leptin-induced increase in BAT UCP-1. However, the V-induced increase in basal UCP-1 was not expected. Because V (alone) did not affect leptin signaling in the hypothalamus, this is probably a leptin-independent increase in UCP-1. Perhaps V has some direct effect on BAT, possibly sensitizing it to norepinephrine via some affect on -adrenergic signaling. Consistent with this possibility, in in vitro preparations of rabbit ileum and guinea pig vas deferens, V enhanced the magnitude and duration of responsiveness of these tissues to epinephrine (23). A second possibility is that circulating V directly activated or sensitized BAT leptin receptors, which can also stimulate UCP-1 via a noradrenergic pathway (24).
Another unexpected finding was that V suppressed serum leptin levels without a significant effect on visceral adiposity. This was particularly surprising in light of a report by Wang et al. (14) suggesting that V promotes leptin expression and secretion by isolated white adipocytes. In the present study we could not detect a significant effect of V on leptin mRNA in RTWAT, although the tendency was toward lower leptin expression in V-treated animals. Unfortunately, we did not access total adiposity in this experiment. Rather, we report an index of abdominal adiposity that includes the sum of RTWAT, PWAT, and EWAT. Similar to the leptin expression results, this index tended to be lower in V-treated animals in experiment 1, but failed to reach statistical significance. However, in experiment 2 this trend of reduced adiposity with vanadium treatment alone (C-ACSF vs. V-ACSF) was not present despite reduced serum leptin levels after vanadium treatment. This raises a suspicion that vanadium has a suppressive effect on leptin synthesis or secretion that is independent of adiposity. Perhaps the mechanism behind this effect is related to vanadium’s effect on UCP-1 and is also mediated by -adrenergic signaling. Indeed, it is known that 3-receptor agonists curb leptin gene expression (25).
Although not the focus of this study, the steepness of the leptin dose-response icv infusion curve was unexpected. Although the 0.25 μg/d dose of leptin had intermediate effects on food intake and body weight, it caused a greater than 95% reduction in an index of intraabdominal adiposity (as assessed by the sum of RTWAT, PWAT, and EWAT). This contrasts sharply with the results using a 5-fold lower dose of leptin (0.05 μg/d), which had no demonstrable effect on food intake, body weight, or adiposity. This steep curve added to the challenge of selecting the optimal submaximal dose for testing vanadium’s effect on leptin responsiveness. Nonetheless, with the lower dose of leptin in experiment 2, small strips of intraabdominal fat remained in the majority of rats treated with leptin alone after 21 d of leptin infusion. No rats treated with vanadium plus leptin had visible intraabdominal white fat in experiment 2. Similarly, in experiment 1, there was complete disappearance of intraabdominal white fat in all but one V-leptin animal, whereas half the animals treated with leptin alone had excisable RTWAT and/or EWAT. This in conjunction with the body weight data from both experiments provides convincing evidence that V, at the dose provided, enhances the catabolic effects of leptin.
In summary, we have demonstrated for the first time that oral vanadium treatment enhances the physiological effects of leptin in young adult F344 x Brown Norway male rats. Moreover, we show that V enhances leptin-induced STAT3 phosphorylation in the hypothalamus. Although not definitively proven, we believe that vanadium’s effects on leptin-induced STAT3 phosphorylation contribute to the observed physiological effects. The observation that vanadium potentiates leptin-mediated weight loss in this highly leptin-responsive model is promising with respect to our ultimate objective: overcoming leptin resistance. Concomitant treatment with V or similar PTPase inhibitors may prove to be an effective method to restore leptin responsiveness. This could enable researchers and clinicians to harness the power of leptin to combat recalcitrant obesity.
Footnotes
This work was supported by the Medical Research Service of the Department of Veterans Affairs and National Institute of Aging Grant AG-26159.
First Published Online September 29, 2005
Abbreviations: ACSF, Artificial cerebral spinal fluid; AUC, area under the curve; C, control; EWAT, epididymal white adipose tissue; GTT, glucose tolerance test; iBAT, interscapular brown adipose tissue; icv, intracerebroventricularly; IRS, insulin receptor substrate; JAK, Janus kinase; P-STAT, phosphorylated STAT; PTPase, protein tyrosine phosphatase; PWAT, perirenal white adipose tissue; RTWAT, retroperitoneal white adipose tissue; STAT, signal transducer and activator of transcription; UCP-1, uncoupling protein-1; V, vanadyl acetoacetonate.
Accepted for publication September 21, 2005.
References
Brichard SM, Okitolonda W, Henquin JC 1988 Long term improvement of glucose homeostasis by vanadate treatment in diabetic rats. Endocrinology 123:2048–2053
Heyliger CE, Tahiliani AG, McNeill JH 1985 Effect of vanadate on elevated blood glucose and depressed cardiac performance of diabetic rats. Science 227:1474–1477
Boden G, Chen X, Ruiz J, van Rossum GD, Turco S 1996 Effects of vanadyl sulfate on carbohydrate and lipid metabolism in patients with non-insulin-dependent diabetes mellitus. Metabolism 45:1130–1135
Brichard SM, Pottier AM, Henquin JC 1989 Long term improvement of glucose homeostasis by vanadate in obese hyperinsulinemic fa/fa rats. Endocrinology 125:2510–2516
Pugazhenthi S, Tanha F, Dahl B, Khandelwal RL 1995 Decrease in protein tyrosine phosphatase activities in vanadate-treated obese Zucker (fa/fa) rat liver. Mol Cell Biochem 153:125–129
Takeshita S, Kawamura I, Yasuno T, Kimura C, Yamamoto T, Seki J, Tamura A, Sakurai H, Goto T 2001 Amelioration of insulin resistance in diabetic ob/ob mice by a new type of orally active insulin-mimetic vanadyl complex: bis(1-oxy-2-pyridinethiolato)oxovanadium(IV) with VO(S2O2) coordination mode. J Inorg Biochem 85:179–186
Mohammad A, Wang J, McNeill JH 2002 Bis(maltolato)oxovanadium(IV) inhibits the activity of PTP1B in Zucker rat skeletal muscle in vivo. Mol Cell Biochem 229:125–128
Zabolotny JM, Bence-Hanulec KK, Stricker-Krongrad A, Haj F, Wang Y, Minokoshi Y, Kim YB, Elmquist JK, Tartaglia LA, Kahn BB, Neel BG 2002 PTP1B regulates leptin signal transduction in vivo. Dev Cell 2:489–495
Zabolotny JM, Haj FG, Kim YB, Kim HJ, Shulman GI, Kim JK, Neel BG, Kahn BB 2004 Transgenic overexpression of protein-tyrosine phosphatase 1B in muscle causes insulin resistance, but overexpression with leukocyte antigen-related phosphatase does not additively impair insulin action. J Biol Chem 279:24844–24851
Schwartz MW, Woods SC, Porte Jr D, Seeley RJ, Baskin DG 2000 Central nervous system control of food intake. Nature 404:661–671
Li H, Matheny M, Nicolson M, Tumer N, Scarpace PJ 1997 Leptin gene expression increases with age independent of increasing adiposity in rats. Diabetes 46:2035–2039
Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156–159
Rao PV 1998 Statistical research methods in the life sciences. Pacific Grove: Brooks/Cole
Wang Y, Han J, Zang Y, Guo W 2002 Effects of vanadate on leptin production from isolated rat adipocytes. Biol Trace Elem Res 85:171–182
Kita A, Uotani S, Kuwahara H, Takahashi R, Oshima K, Yamasaki H, Mizuguchi H, Hayakawa T, Nagayama Y, Yamaguchi Y, Eguchi K 2003 Vanadate enhances leptin-induced activation of JAK/STAT pathway in CHO cells. Biochem Biophys Res Commun 302:805–809
Bates SH, Stearns WH, Dundon TA, Schubert M, Tso AW, Wang Y, Banks AS, Lavery HJ, Haq AK, Maratos-Flier E, Neel BG, Schwartz MW, Myers Jr MG 2003 STAT3 signalling is required for leptin regulation of energy balance but not reproduction. Nature 421:856–859
Carpenter LR, Farruggella TJ, Symes A, Karow ML, Yancopoulos GD, Stahl N 1998 Enhancing leptin response by preventing SH2-containing phosphatase 2 interaction with Ob receptor. Proc Natl Acad Sci USA 95:6061–6066
Brautigan DL, Pinault FM 1993 Serine phosphorylation of protein tyrosine phosphatase (PTP1B) in HeLa cells in response to analogues of cAMP or diacylglycerol plus okadaic acid. Mol Cell Biochem 127–128:121–129
Goldstein BJ 2001 Protein-tyrosine phosphatase 1B (PTP1B): a novel therapeutic target for type 2 diabetes mellitus, obesity and related states of insulin resistance. Curr Drug Targets Immune Endocr Metab Disord 1:265–275
Myers MP, Andersen JN, Cheng A, Tremblay ML, Horvath CM, Parisien JP, Salmeen A, Barford D, Tonks NK 2001 TYK2 and JAK2 are substrates of protein-tyrosine phosphatase 1B. J Biol Chem 276:47771–47774
Elchebly M, Payette P, Michaliszyn E, Cromlish W, Collins S, Loy AL, Normandin D, Cheng A, Himms-Hagen J, Chan CC, Ramachandran C, Gresser MJ, Tremblay ML, Kennedy BP 1999 Increased insulin sensitivity and obesity resistance in mice lacking the protein tyrosine phosphatase-1B gene. Science 283:1544–1548
Kaszubska W, Falls HD, Schaefer VG, Haasch D, Frost L, Hessler P, Kroeger PE, White DW, Jirousek MR, Trevillyan JM 2002 Protein tyrosine phosphatase 1B negatively regulates leptin signaling in a hypothalamic cell line. Mol Cell Endocrinol 195:109–118
Delamere NA, Williams RN 1986 Modulation by vanadate of the adrenergic characteristics of the iris, ileum, and vas deferens. Invest Ophthalmol Vis Sci 27:1336–1341
Commins SP, Marsh DJ, Thomas SA, Watson PM, Padgett MA, Palmiter R, Gettys TW 1999 Norepinephrine is required for leptin effects on gene expression in brown and white adipose tissue. Endocrinology 140:4772–4778
Li H, Matheny M, Scarpace PJ 1997 3-Adrenergic-mediated suppression of leptin gene expression in rats. Am J Physiol 272:E1031–E1036(Jared Wilsey, Michael K. Matheny and Phi)
Clinical Center, Department of Veterans Affairs Medical Center (J.W.), Gainesville, Florida 32608-1197
Department of Pharmacology and Therapeutics, University of Florida College of Medicine (J.W., M.K.M., P.J.S.), Gainesville, Florida 32608
Abstract
Recently, vanadium has been shown to enhance leptin signal transduction in vitro. We hypothesized that chronic oral administration of an organic vanadium complex would enhance both leptin signaling and physiological responsiveness in vivo. Three-month-old F344 x Brown Norway male rats were provided a solution containing escalating doses of vanadyl acetoacetonate (V), peaking at 60 mg/liter elemental vanadium in drinking water on the 11th d of V treatment. Although V treatment tended to suppress weight gain, absolute body weights did not significantly differ between groups after 62 d of treatment. At this point, a permanent cannula was placed into the left lateral ventricle of all animals. The cannula was connected to a sc minipump providing either 5 μg/d leptin or artificial cerebral spinal fluid (ACSF) control solution. This yielded four groups: C-ACSF, C-leptin, V-ACSF, and V-leptin. During the ensuing 26 d, weight gain was similar in C-ACSF and V-ACSF. As expected, leptin caused dramatic weight loss in C-leptin, but leptin-induced weight loss was 43% greater in V-leptin. V enhanced leptin-induced signal transducer and activator of transcription-3 phosphorylation in the hypothalamus, whereas V alone had no effect. V also augmented the leptin-induced increase in brown adipose tissue uncoupling protein-1. The effects of vanadium on responsiveness to a submaximal dose of leptin (0.25 μg/d) were also evaluated, yielding qualitatively similar results. These data demonstrate, for the first time, that chronic V administration enhances the weight-reducing effects of centrally administered leptin in young adult animals, and the mechanism appears to involve enhanced leptin signal transduction.
Introduction
THE ELEMENT VANADIUM is found in most living organisms, but its essentiality in mammals has not been established (1). Various forms of the element have long been known to effect insulin sensitivity. In 1985, vanadate (V5+) was first shown to have an insulin-like action (2). It subsequently became clear that vanadium acted more as an insulin-sensitizing agent, because it was only effective in the presence of at least small amounts of insulin. Although the insulin-sensitizing effect of vanadium compounds has been established in humans (3) and a variety of animal models (1, 4), the mechanism of this effect remains in question. However, the mechanism may involve the ability of the element to inhibit the dephosphorylation of key intracellular signaling molecules. Vanadate potently inhibits purified protein tyrosine phosphatases (PTPases) in vitro (5). In studies of mouse adipocyte cultures, vanadium was shown to increase insulin receptor substrate-1 (IRS-1) phosphorylation (6). Increased tyrosine phosphorylation of IRS-1 and other molecules is most likely secondary to a decrease in PTPase activity. Consistent with this, it was recently demonstrated that oral vanadium reduced the activity of PTP1B in skeletal muscle of fatty Zucker rats by 25% (7). PTP1B is a specific tyrosine phosphatase that inhibits both leptin and insulin signal transduction (8, 9). Given the overlap between leptin and insulin signal transduction (10) as well as the importance of tyrosine phosphorylation of second messengers in the leptin signaling cascade, we reasoned that vanadium would enhance leptin signaling and sensitivity. To test our hypothesis, we added an organic complex of vanadyl acetoacetonate (V) to the drinking water of young adult F344 x Brown Norway male rats and then measured the biochemical and physiological responses to chronic intracranial leptin infusion.
Materials and Methods
Animals
Three-month-old male Fischer 344 x Brown Norway rats were obtained from Harlan Sprague Dawley (Indianapolis, IN). Upon arrival, rats were examined and remained quarantined for 1 wk. Animals were individually caged with a 12-h light, 12-h dark cycle (lights on from 0700–1900 h). Animals were cared for in accordance with the principles of the National Institutes of Health Guide to the Care and Use of Experimental Animals.
Experimental design
Experiment 1.
Three-month-old F344 x Brown Norway male rats were provided a solution containing escalating doses of V, peaking at 60 mg/liter elemental vanadium/d on the 11th d of V treatment. The V solution contained hypotonic saline 0.045%) and was sweetened with 0.125% saccharine to offset taste aversion. Control (C) animals were maintained on tap water. Hypotonic saline and saccharine were not provided to control animals, because it was observed in a preliminary experiment that this significantly increased fluid intake over baseline. After 62 d of V treatment (in the V group), a permanent cannula was placed into the left lateral cerebroventricle of all animals. The cannula was connected to a sc minipump providing either 5 μg/d leptin or ACSF control solution. This yielded four groups: C-ACSF, C-leptin, V-ACSF, and V-leptin. Animals were killed 28 d after cannulation for tissue analysis.
Experiment 2.
A second experiment was conducted to determine the effects of V treatment on physiological responsiveness to a submaximal dose of leptin. A submaximal dose of leptin for central infusion was first determined in a dose-response experiment in which animals were given 0, 0.05, 0.25, 1.0, or 5 μg/d leptin, intracerebroventricularly (icv). It was determined that 0.25 μg/d caused intermediate physiological effects. Three-month-old F344 x Brown Norway male rats were again pretreated with V as in experiment 1, then implanted with an intracranial (left lateral ventricle) cannula connected to an sc minipump providing 0.25 μg/d leptin. In this experiment animals were killed on d 21 in hope of preventing complete lipopenia in the C-leptin group (which makes it difficult to evaluate the synergistic effects of vanadium, if any, on fat loss).
Glucose tolerance test
After an overnight fast, rats were administered 2 g/kg glucose ip. Blood was drawn from the tail at baseline and 30, 60, and 120 min after glucose injection. Animals were not anesthetized. At each interval, approximately 250 μl blood was drawn from the tip of the tail. A single drop of tail blood was used to measure glucose via a glucose meter (OneTouch SureStep, Johnson & Johnson, Inc., Milpitas, CA). Remaining blood was centrifuged at 1300 x g for 10 min to yield serum for assessment of insulin by RIA (Linco Research, Inc., St. Charles, MO). The area under the curve (AUC) for glucose was calculated by subtracting the baseline serum glucose level in controls. This correction was not practical for insulin AUC, because baseline levels were not similar in controls and vanadium-treated animals. Thus, AUC for insulin was calculated from the x-axis (y = 0).
Intracranial cannulation and leptin infusion
Rats were anesthetized with 60 mg/kg pentobarbital, and heads were prepared for surgery. Animals were placed into a stereotaxic frame, and a small incision (1.5 cm) was made over the midline of the skull to expose the landmarks of the cranium (bregma and lambda). A cannula (Durect Corp., Cupertino, CA) was placed into the lateral ventricle using the following coordinates: 1.3 mm posterior to bregma, 1.9 mm lateral to the midsaggital suture, and to a depth of 3.5 mm. The cannula was anchored to the skull using acrylic dental cement. A sc pocket on the dorsal surface was created using blunt dissection, and the osmotic mini pump (Durect Corp.) was inserted. These pumps (model 2004, ALZET; Durect Corp.) infuse 0.25 μl fluid/h for a minimum of 28 d and have a total capacity of 200 μl. Thus, in experiment 1 the pumps in the leptin group were filled with a solution containing leptin in ACSF (0.833 μg/μl) to provide 5 μg/d leptin. In experiment 2 (submaximal dose of leptin), this was reduced to 0.0417 μg/μl, providing 0.25 μg/d leptin. After filling the minipumps with ACSF or ACSF/leptin solution, they were incubated in sterile saline at 37 C for 36 h before implantation. A catheter tube was employed to connect the cannula to the osmotic minipump flow moderator. The incision for the minipump was then closed with sutures. Rats were kept warm during the manipulations and until fully recovered.
Tissue harvesting
Anesthetized rats were killed by cervical dislocation. Blood was collected by cardiac puncture, and serum was harvested by a 10-min centrifugation in serum separator tubes. The circulatory system was perfused with 20 ml cold saline. Perirenal (PWAT), retroperitoneal (RTWAT), and epididymal (EWAT) white adipose tissues; interscapular brown adipose tissue (iBAT); and hypothalami were excised, weighed, and immediately frozen in liquid nitrogen. The hypothalamus was removed by making an incision medial to piriform lobes, caudal to the optic chiasm, and anterior to the cerebral crus to a depth of 2–3 mm. Tissues were stored at –80 C until analysis.
Signal transducer and activator of transcription-3 (STAT3)/phosphorylated STAT3 (P-STAT3) assay
These methods were described in detail previously (11). Briefly, hypothalamus was sonicated in 10 mM Tris-HCl (pH 6.8), 2% sodium dodecyl sulfate, and 0.08 μg/ml okadaic acid plus protease inhibitors (phenylmethylsulfonylfluoride, benzamidine, and leupeptin). Sonicate was diluted and quantified for protein using a detergent-compatible Bradford assay. Samples were boiled and separated on a 7.5% agarose/Tris-HCl gel (Bio-Rad Laboratories, Hercules, CA) and electrotransferred to a nitrocellulose membrane. Immunoreactivity was assessed with an antibody specific to P-STAT3 (antibody kit from New England Biolabs, Beverly, MA). Immunoreactivity was visualized by chemiluminescence detection (Amersham Biosciences, Piscataway, NJ) and quantified by video densitometry (Bio-Rad Laboratories). After P-STAT3 quantification, membranes were stripped of antibody with Immunopure (Pierce Chemical Co., Rockford, IL), and immunoreactivity was reassessed using a total STAT3 antibody.
Leptin mRNA levels in WAT
RTWAT (300 mg/sample) was sonicated in guanidine buffer, phenol extracted, and isopropanol precipitated using a modification of the method of Chomczynski and Sacchi (12). Isolated RNA was resuspended in ribonuclease-free water and quantified by spectrophotometry. Integrity was verified using 1% agarose gels stained with ethidium bromide. For dot-blot analysis, 100 ng total RNA was immobilized by loading directly onto a nylon membrane in triplicate using a dot-blot apparatus (Bio-Rad Laboratories, Richmond, CA). Membranes were baked in a UV cross-linking apparatus. Membranes were then prehybridized in 10 ml QuickHyb (Stratagene, La Jolla, CA) for 30 min, followed by hybridization in the presence of a labeled probe for leptin mRNA and 100 μg salmon sperm DNA. The probe used to detect leptin mRNA was a 33-mer antisense oligonucleotide (5'-GGTCTGAGGCAGGGA CAGCTCTTGGAGAAGGC) end labeled using terminal deoxynucleotidyl transferase (Promega Corp., Madison, WI). We previously demonstrated by Northern analysis that this probe binds to a single mRNA species of 4.1 kb (11). After hybridization for 2 h at 65 C, the membranes were washed and exposed to a phosphorimaging screen for 72 h. The screen was then scanned using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) and analyzed by Image-Quant software (Molecular Dynamics). Data are expressed as mRNA per total RTWAT pad.
Serum leptin
Serum leptin was measured using a mouse leptin ELISA kit (Crystal Chemicals, Inc., Chicago, IL) on blood harvested at the time of death by cardiac puncture.
Statistical analysis
All data are expressed as the mean ± SEM. The level was set at 0.05 for all analyses. Comparisons of mean food intake and body weight gain in control vs. vanadium-treated were animals were made by a repeated measures, two-way ANOVA, with time and treatment as factors. After minipump implantation, food and body mass data were analyzed by three-way ANOVA, with time, leptin, and vanadium as factors (experiment 1). Two-way analyses (leptin x vanadium), in which we examined the change in body mass during the treatment period and cumulative food intake during the treatment period, were also performed (experiments 1 and 2). All end-point serum and tissue data were analyzed by a two-way analysis, with leptin and vanadium as factors. When only main effects were significant, relevant pairwise comparisons were made using the Bonferroni multiple comparison method, with the error rate corrected for the number of contrasts (13). When there was an interaction, factors were separated, and an additional one-way ANOVA was applied with a Bonferroni multiple comparison post hoc test. When separation of factors resulted in only two population means to compare, the one-way ANOVA was replaced by Student’s t test. PRISM software version 4.0 (GraphPad, Inc., San Diego, CA) was used for all statistical analysis and graphing, with the exception of the three-way ANOVA. QuickCalc (GraphPad; www.graphpad.com) was used for post hoc analysis of two-way ANOVAs. GraphPad QuickCalc uses the Bonferroni correction for multiple comparisons. The three-factor ANOVAs (body mass and food intake after minipump implantation in experiment 1) were performed with the assistance of the VassarStats statistical package developed by Dr. Richard Lowry (Vassar College, Poughkeepsie, NY), and verified with SPSS Base 13.0 (SPSS, Inc., Chicago, IL).
Results
Experiment 1: vanadium pretreatment
Food intake and body mass.
Three-month-old F344 x Brown Norway male rats were provided a solution containing escalating doses of V, peaking at 60 mg/liter elemental vanadium/d on the 11th d of V treatment. To reduce the risk of taste aversion, 0.125% saccharine and 0.45% NaCl were added to the V solution. Control animals were maintained on tap water, because it was determined that saccharine and NaCl significantly increase fluid consumption (in the absence of V). During the first 42 d of V treatment, food intake per day averaged 5.7% less in V-treated animals than in controls (19.13 ± 0.31 vs. 20.29 ± 0.41 g/d, respectively; P < 0.05). However, this modest effect disappeared after d 42 (mean food intake on d 42–60 was 18.90 ± 0.38 g/d in controls vs. 19.02 ± 0.36 g/d in V-treated animals). At the start of V treatment, body weight was similar in V and control groups. Vanadium had a slight suppressive effect on weight gain during 60 d of V treatment (86.6 ± 5.0 g gained in controls vs. 69.2 ± 5.2 g in V-treated animals; P < 0.05; Fig. 1), but absolute body weights were not significantly different between groups after the V treatment period (371.9 ± 5.4 g in control vs. 358.1 ± 6.9 g in V group; Fig. 1).
Fluid and vanadium intake.
During the period of escalating the dose of V, fluid intake was variable in the V group, first increasing and then decreasing, with a nadir of 79% of control intake on d 18 (not shown). However, fluid intake in V-treated animals returned to baseline after this adjustment period, and there was no difference in fluid intake between V and control animals during the 30 d before leptin administration. During this period (30 d before leptin administration), fluid intake averaged 16.8 ± 0.45 ml/d in the V group. This corresponds to 1.01 ± 0.027 mg elemental vanadium/rat·d, or 2.66 ± 0.078 mg/kg just before the leptin infusion.
Glucose tolerance test (GTT).
Approximately 6 wk after commencing V treatment, a GTT was performed. Serum glucose and insulin were measured at baseline and 30, 60, and 120 min after injecting glucose (2 g/kg) ip. Glucose clearance was enhanced in V-treated animals. The effect of V on serum glucose during the GTT was significant by repeated measures two-way ANOVA (Fig. 2). The serum glucose area under the curve was reduced by 24.3% in the V group (baseline set at time zero serum glucose in controls; Fig. 2). V also had a significant suppressive effect on fasting serum insulin (1.27 ± 0.15 vs. 0.758 ± 0.11 ng/ml; P < 0.05) and on insulin secretion during the GTT (total area under the curve, 194.2 min·ng/ml in controls vs. 98.1 min·ng/ml in V animals; baseline set at y = 0).
Leptin infusion
Body mass, food intake, and fluid intake.
Before minipump implantation, body mass was similar in the four groups, but tended to be lower in the V-treated animals, as described above (C-ACSF, 373.0 ± 8.2 g; C-leptin, 370.8 ± 7.6; V-ACSF, 357.6 ± 10.5; V-leptin, 358.5 ± 9.7). During the ensuing 26 d, weight gain was similar in C-ACSF and V-ACSF (6.7 ± 3.8 vs. 4.2 ± 2.1 g, respectively; Fig. 3). Although leptin caused significant weight loss in both V animals and controls, V-leptin lost 42.7% more weight over 26 d than C-leptin (97.5 vs. 68.3 g; P < 0.05; Fig. 3).
Twenty-four-hour food intake was similar in the four groups before minipump implantation (C-ACSF, 19.95 ± 0.76 g; C-leptin, 20.17 ± 0.74; V-ACSF, 19.21 ± 0.66; V-leptin, 18.88 ± 0.56 g; averaged over 48 h). Food intake remained similar in the two ACSF-treated groups (C-ACSF and V-ACSF) after minipump implantation (Fig. 4). Leptin caused significant reductions in feeding in both C-leptin and V-leptin groups over the ensuing 26 d (Fig. 4). Mean food intake decreased by 32.5% in C-leptin and by 34.8% in V-leptin (percent decrease is based on average intake during 26 d after pump implantation relative to baseline), but V did not significantly affect leptin-induced suppression of food intake (Fig. 4). Fluid intake was unaffected compared with pretreatment values in C-leptin and V-leptin (data not shown).
Serum leptin and insulin.
Twenty-eight days of central leptin infusion caused significant 87% and 70% decreases, respectively, in serum leptin in controls (nonvanadium treated) and vanadium-treated animals (Table 1). Surprisingly, vanadium also caused a significant (53%) reduction in serum leptin (Control-ACSF vs. V-ACSF; Table 1). This was unexpected, because intraabdominal adiposity and body mass were similar in C-ACSF and V-ACSF, and in vitro studies suggest that V enhances leptin expression (14). In addition, both vanadium and leptin caused significant reductions in fasting serum insulin (Table 1).
Intraabdominal adiposity and leptin expression in WAT.
Leptin caused a near-complete disappearance of intraabdominal adipose tissue in both C-leptin and V-leptin. On the average, less than 2% of the abdominal adiposity present in ACSF-infused animals could be removed from leptin-treated animals regardless of vanadium treatment. RTWAT and PWAT were completely absent from five of six C-leptin rats and in six of seven V-leptin rats (data not shown). A very small amount of EWAT remained in three of six C-leptin rats, but in none of the V-leptin rats (data not shown). Leptin expression in the RTWAT of C-ACSF and V-ACSF was similar (100 ± 14.4 and 76.65 ± 16.5 arbitrary units/mg RNA, respectively). Due to complete leptin-induced catabolism of RTWAT, leptin expression could not be evaluated in C-leptin and V-leptin.
STAT3 phosphorylation.
As expected, chronic intracranial leptin infusion caused a significant 35% increase in P(Y-705)-STAT3 immunoreactivity in the medial basal hypothalamus in C-leptin vs. C-ACSF (Fig. 5A). Although V alone did not affect P-STAT3 levels, V did enhance leptin-induced STAT3 phosphorylation (114% increase in V-leptin vs. V-ACSF). Hypothalamic P-STAT3 immunoreactivity was 70% higher in V-leptin vs. C-leptin (Fig. 5A; P < 0.05). Chronic leptin treatment caused a mild elevation in total STAT3 in the hypothalamus, whereas V had no effect (Fig. 5B).
Uncoupling protein-1 (UCP-1) in BAT.
Consistent with our previous findings, leptin caused an approximately 2-fold increase in UCP-1 levels per iBAT pad (C-ACSF vs. C-leptin, P < 0.05; Fig. 6). When expressed as UCP-1 per milligram of iBAT protein, leptin caused a 3.98-fold elevation in UCP-1 (C-ACSF vs. C-leptin, P < 0.05; data not shown). Vanadium caused a 2-fold elevation in UCP-1 per BAT pad (V-ACSF vs. C-ACSF, P < 0.05). The effects of vanadium and leptin on UCP-1 appeared to be additive, because iBAT UCP-1 was 2-fold greater in V-leptin than in rats treated with either vanadium alone or leptin alone (P < 0.05; Fig. 6).
Experiment 2
Leptin infusion dose-response.
The dose of leptin provided in the first experiment appears to evoke maximal physiological responses, thus rendering it difficult to fully evaluate the potential of V to enhance the responses to leptin. Therefore, we employed a submaximal dose of leptin in experiment 2. First, a preliminary experiment was conducted to determine the dose of leptin that would be provided at the end of V pretreatment. Rats were infused with 0.05, 0.25, 1, or 5 μg/d leptin for 20 d (Fig. 7A). The rats given 0.25 μg/d leptin lost significantly more weight than controls, but lost significantly less weight than rats given the 5 μg/d dose provided in experiment 1 (Fig. 7A). Cumulative food intake during the first 20 d of leptin infusion was significantly lower in rats centrally infused with 0.25 μg leptin/d vs. controls, but was significantly greater than that in rats given 5 μg leptin/d (Fig. 7B). Moreover, although 5 μg leptin/d caused complete catabolism of RTWAT and PWAT in this dose-response experiment, all rats given 0.25 μg leptin/d had a small amount of these depots remaining (mean mass RTWAT + PWAT, 25.7 ± 0.9 mg in 0.25 μg/d group; data not shown). The next lowest dose tested, 0.05 μg/d, did not have a significant effect on body weight (Fig. 7A), food intake (Fig. 7B), or central adiposity (data not shown), and the next higher dose, 1.0 μg/d, was too efficacious. Thus, it was determined that 0.25 μg/d was the minimum efficacious, submaximal dose for chronic icv leptin infusion among the doses tested. This is the dose that was chosen for the experiment to test the interaction of vanadium and a submaximal central dose of leptin.
Vanadium pretreatment.
In experiment 2, the vanadium pretreatment protocol and physiological effects were virtually identical with those in experiment 1. At the end of the vanadium pretreatment period, body weights were not significantly different in controls (318.4 ± 2.1 g) and V-treated animals (313.5 ± 3.8 g).
Leptin infusion
Body mass and food intake.
Changes in body mass during the lower-dose central leptin infusion were qualitatively similar to what was observed in the initial high-dose experiment. Leptin caused a significant reduction in body mass in both controls and vanadium-treated rats, with the V-leptin group losing 43.6% more weight than the C-leptin group during the 21-d treatment period (P < 0.01; Fig. 8A). The food intake data did differ somewhat from the high-dose group (experiment 1). First, the leptin-induced reduction in food intake took much longer to appear with the 0.25 μg/d dose. Although a reduction in food intake was readily apparent by d 4 in experiment 1, a reduction in food intake in the 0.25 μg/d-treated animals in experiment 2 did not appear until d 8–10 (data not shown). Another difference between experiments is that vanadium significantly enhanced the leptin-induced inhibition in feeding in experiment 2. The V-leptin group consumed significantly less food than the C-leptin group between the 8th and 20th d of leptin infusion (Fig. 8B). Vanadium alone, however, did not affect food intake (no difference in food intake between C-ACSF and V-ACSF; Fig. 8B). This suggests that vanadium, at the dose provided, does not have significant anorectic properties by itself; rather, it interacts with the effects of submaximal leptin to enhance both feeding inhibition and weight loss.
Intraabdominal adiposity.
Central leptin at a dose that was clearly submaximal with respect to reductions in body mass and food intake was still able to cause near-complete catabolism of intraabdominal WAT. However, four of six C-leptin rats had small strips of RTWAT and EWAT (averaging 120.2 ± 26.9 mg in these four rats), whereas zero of six V-leptin rats had visible RTWAT or EWAT. When the intraabdominal adiposity of V-leptin and C-leptin were compared categorically for the presence or absence of excisable fat depots, the difference between groups was statistically significant (Table 2).
Discussion
The primary purpose of this investigation was to determine whether vanadium could enhance leptin signaling and the physiological effects of chronic icv leptin infusion in young adult rats. First, we wanted to determine whether the dose and form of V we provided were able to influence in vivo insulin sensitivity. Consistent with the reported effects of vanadium on glucose homeostasis in diabetic models (4, 6), chronic V treatment improved glucose tolerance and lowered fasting serum insulin levels in our young adult F344 x Brown Norway male rats. However, the present results are unique in that they demonstrate that V can enhance markers for insulin sensitivity even in a nonobese, nondiabetic model.
As hypothesized, V did enhance the catabolic effect of chronic intracranial leptin infusion. V treatment resulted in significantly more leptin-induced weight loss without significantly affecting leptin-induced suppression of energy intake when leptin was infused at 5 μg/d. This suggests that V acted primarily by augmenting leptin’s effects on energy expenditure. However, when the leptin dose was lowered to 0.25 μg/d, vanadium enhanced both leptin-induced weight loss and feeding inhibition. Thus, it seems plausible that the apparent lack of an effect of vanadium on the feeding-inhibiting properties of leptin at the 5 μg/d icv dose was secondary to a basement effect, where compensatory survival (orexigenic) pathways limited additional feeding inhibition.
We hypothesized that the mechanism by which V may enhance leptin responsiveness would involve enhanced leptin signal transduction. As discussed in the introduction, this hypothesis was based in part upon the recently reported effects of V on insulin signal transduction (6) and the overlap between leptin and insulin signaling. Moreover, Kita et al. (15) demonstrated that pretreatment of Chinese hamster ovary cells expressing ObRb with 100 μM orthovanadate dramatically enhanced leptin-induced STAT3 and Janus kinase-2 (JAK2) phosphorylation. Given the critical role of STAT3 signaling with respect to leptin-induced suppression of energy intake and thermogenesis (16), we examined the effects of V on this pathway in vivo. As surmised, V amplified leptin’s effects on STAT3 phosphorylation in our present study. However, V alone had no effect on basal STAT3 phosphorylation. Our results suggest that in vivo, a P-STAT3-inducing stimulus (i.e. leptin infusion) must be concomitantly administered for V to exert an effect on this pathway.
If V enhances physiological leptin responsiveness by enhancing leptin signal transduction through STAT3 and possibly other pathways, what is the mechanism behind the enhanced signaling Although we did not address this question directly in the present study, it is plausible that the mechanism involves an inhibition of PTPase activity. Leptin signal transduction relies upon the tyrosine phosphorylation of not only the receptor itself, but also messengers, including JAK2 and STAT3. As such, PTPases inhibit leptin signal transduction. One such phosphatase is Src homology domain 2-containing phosphatase-2 (17). Prevention of Src homology domain 2-containing phosphatase-2 recruitment via a point mutation enhances the expression of genes under the control of STAT3 (17). Another tyrosine phosphatase believed to inhibit JAK2/STAT3 signaling through the leptin receptor is PTPase-1B (18). This phosphatase is of particular clinical interest because it is also an inhibitor of insulin signal transduction (19). PTPase-1B is believed to block JAK2 activation (20), inhibiting both the JAK2-STAT3 and IRS-phosphotidylinositol 3-kinase cascades. The latter pathway is involved in both insulin and leptin signaling. Transgenic mice lacking PTPase-1B are lean and resistant to diet-induced obesity and have enhanced insulin sensitivity (21). Similarly, mice treated with PTPase-1B antisense oligonucleotides have enhanced insulin sensitivity and reduced adiposity (19). Mice overexpressing PTPase-1B, in contrast, are insulin resistant (9). Moreover, overexpression of PTPase-1B in a hypothalamic cell line lowers basal JAK2 phosphorylation and blocks the normal leptin-induced up-regulation of suppressor of cytokine signaling-3 expression (22). Taken together, this evidence suggests that inhibiting the activity of PTPase-1B and related tyrosine phosphatases could be an effective strategy to enhance leptin sensitivity. As discussed in the introduction, V is an established inhibitor of PTPases and has been shown to do so in vivo (7).
Leptin signaling in the hypothalamus causes a negative shift in energy balance by reducing food intake and increasing energy expenditure (11). Given that V enhanced leptin-induced weight loss with an insignificant effect on food intake in experiment 1 (high dose of leptin), it follows that V probably enhanced leptin’s effects on energy expenditure. Therefore, it is not surprising that V augmented the leptin-induced increase in BAT UCP-1. However, the V-induced increase in basal UCP-1 was not expected. Because V (alone) did not affect leptin signaling in the hypothalamus, this is probably a leptin-independent increase in UCP-1. Perhaps V has some direct effect on BAT, possibly sensitizing it to norepinephrine via some affect on -adrenergic signaling. Consistent with this possibility, in in vitro preparations of rabbit ileum and guinea pig vas deferens, V enhanced the magnitude and duration of responsiveness of these tissues to epinephrine (23). A second possibility is that circulating V directly activated or sensitized BAT leptin receptors, which can also stimulate UCP-1 via a noradrenergic pathway (24).
Another unexpected finding was that V suppressed serum leptin levels without a significant effect on visceral adiposity. This was particularly surprising in light of a report by Wang et al. (14) suggesting that V promotes leptin expression and secretion by isolated white adipocytes. In the present study we could not detect a significant effect of V on leptin mRNA in RTWAT, although the tendency was toward lower leptin expression in V-treated animals. Unfortunately, we did not access total adiposity in this experiment. Rather, we report an index of abdominal adiposity that includes the sum of RTWAT, PWAT, and EWAT. Similar to the leptin expression results, this index tended to be lower in V-treated animals in experiment 1, but failed to reach statistical significance. However, in experiment 2 this trend of reduced adiposity with vanadium treatment alone (C-ACSF vs. V-ACSF) was not present despite reduced serum leptin levels after vanadium treatment. This raises a suspicion that vanadium has a suppressive effect on leptin synthesis or secretion that is independent of adiposity. Perhaps the mechanism behind this effect is related to vanadium’s effect on UCP-1 and is also mediated by -adrenergic signaling. Indeed, it is known that 3-receptor agonists curb leptin gene expression (25).
Although not the focus of this study, the steepness of the leptin dose-response icv infusion curve was unexpected. Although the 0.25 μg/d dose of leptin had intermediate effects on food intake and body weight, it caused a greater than 95% reduction in an index of intraabdominal adiposity (as assessed by the sum of RTWAT, PWAT, and EWAT). This contrasts sharply with the results using a 5-fold lower dose of leptin (0.05 μg/d), which had no demonstrable effect on food intake, body weight, or adiposity. This steep curve added to the challenge of selecting the optimal submaximal dose for testing vanadium’s effect on leptin responsiveness. Nonetheless, with the lower dose of leptin in experiment 2, small strips of intraabdominal fat remained in the majority of rats treated with leptin alone after 21 d of leptin infusion. No rats treated with vanadium plus leptin had visible intraabdominal white fat in experiment 2. Similarly, in experiment 1, there was complete disappearance of intraabdominal white fat in all but one V-leptin animal, whereas half the animals treated with leptin alone had excisable RTWAT and/or EWAT. This in conjunction with the body weight data from both experiments provides convincing evidence that V, at the dose provided, enhances the catabolic effects of leptin.
In summary, we have demonstrated for the first time that oral vanadium treatment enhances the physiological effects of leptin in young adult F344 x Brown Norway male rats. Moreover, we show that V enhances leptin-induced STAT3 phosphorylation in the hypothalamus. Although not definitively proven, we believe that vanadium’s effects on leptin-induced STAT3 phosphorylation contribute to the observed physiological effects. The observation that vanadium potentiates leptin-mediated weight loss in this highly leptin-responsive model is promising with respect to our ultimate objective: overcoming leptin resistance. Concomitant treatment with V or similar PTPase inhibitors may prove to be an effective method to restore leptin responsiveness. This could enable researchers and clinicians to harness the power of leptin to combat recalcitrant obesity.
Footnotes
This work was supported by the Medical Research Service of the Department of Veterans Affairs and National Institute of Aging Grant AG-26159.
First Published Online September 29, 2005
Abbreviations: ACSF, Artificial cerebral spinal fluid; AUC, area under the curve; C, control; EWAT, epididymal white adipose tissue; GTT, glucose tolerance test; iBAT, interscapular brown adipose tissue; icv, intracerebroventricularly; IRS, insulin receptor substrate; JAK, Janus kinase; P-STAT, phosphorylated STAT; PTPase, protein tyrosine phosphatase; PWAT, perirenal white adipose tissue; RTWAT, retroperitoneal white adipose tissue; STAT, signal transducer and activator of transcription; UCP-1, uncoupling protein-1; V, vanadyl acetoacetonate.
Accepted for publication September 21, 2005.
References
Brichard SM, Okitolonda W, Henquin JC 1988 Long term improvement of glucose homeostasis by vanadate treatment in diabetic rats. Endocrinology 123:2048–2053
Heyliger CE, Tahiliani AG, McNeill JH 1985 Effect of vanadate on elevated blood glucose and depressed cardiac performance of diabetic rats. Science 227:1474–1477
Boden G, Chen X, Ruiz J, van Rossum GD, Turco S 1996 Effects of vanadyl sulfate on carbohydrate and lipid metabolism in patients with non-insulin-dependent diabetes mellitus. Metabolism 45:1130–1135
Brichard SM, Pottier AM, Henquin JC 1989 Long term improvement of glucose homeostasis by vanadate in obese hyperinsulinemic fa/fa rats. Endocrinology 125:2510–2516
Pugazhenthi S, Tanha F, Dahl B, Khandelwal RL 1995 Decrease in protein tyrosine phosphatase activities in vanadate-treated obese Zucker (fa/fa) rat liver. Mol Cell Biochem 153:125–129
Takeshita S, Kawamura I, Yasuno T, Kimura C, Yamamoto T, Seki J, Tamura A, Sakurai H, Goto T 2001 Amelioration of insulin resistance in diabetic ob/ob mice by a new type of orally active insulin-mimetic vanadyl complex: bis(1-oxy-2-pyridinethiolato)oxovanadium(IV) with VO(S2O2) coordination mode. J Inorg Biochem 85:179–186
Mohammad A, Wang J, McNeill JH 2002 Bis(maltolato)oxovanadium(IV) inhibits the activity of PTP1B in Zucker rat skeletal muscle in vivo. Mol Cell Biochem 229:125–128
Zabolotny JM, Bence-Hanulec KK, Stricker-Krongrad A, Haj F, Wang Y, Minokoshi Y, Kim YB, Elmquist JK, Tartaglia LA, Kahn BB, Neel BG 2002 PTP1B regulates leptin signal transduction in vivo. Dev Cell 2:489–495
Zabolotny JM, Haj FG, Kim YB, Kim HJ, Shulman GI, Kim JK, Neel BG, Kahn BB 2004 Transgenic overexpression of protein-tyrosine phosphatase 1B in muscle causes insulin resistance, but overexpression with leukocyte antigen-related phosphatase does not additively impair insulin action. J Biol Chem 279:24844–24851
Schwartz MW, Woods SC, Porte Jr D, Seeley RJ, Baskin DG 2000 Central nervous system control of food intake. Nature 404:661–671
Li H, Matheny M, Nicolson M, Tumer N, Scarpace PJ 1997 Leptin gene expression increases with age independent of increasing adiposity in rats. Diabetes 46:2035–2039
Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156–159
Rao PV 1998 Statistical research methods in the life sciences. Pacific Grove: Brooks/Cole
Wang Y, Han J, Zang Y, Guo W 2002 Effects of vanadate on leptin production from isolated rat adipocytes. Biol Trace Elem Res 85:171–182
Kita A, Uotani S, Kuwahara H, Takahashi R, Oshima K, Yamasaki H, Mizuguchi H, Hayakawa T, Nagayama Y, Yamaguchi Y, Eguchi K 2003 Vanadate enhances leptin-induced activation of JAK/STAT pathway in CHO cells. Biochem Biophys Res Commun 302:805–809
Bates SH, Stearns WH, Dundon TA, Schubert M, Tso AW, Wang Y, Banks AS, Lavery HJ, Haq AK, Maratos-Flier E, Neel BG, Schwartz MW, Myers Jr MG 2003 STAT3 signalling is required for leptin regulation of energy balance but not reproduction. Nature 421:856–859
Carpenter LR, Farruggella TJ, Symes A, Karow ML, Yancopoulos GD, Stahl N 1998 Enhancing leptin response by preventing SH2-containing phosphatase 2 interaction with Ob receptor. Proc Natl Acad Sci USA 95:6061–6066
Brautigan DL, Pinault FM 1993 Serine phosphorylation of protein tyrosine phosphatase (PTP1B) in HeLa cells in response to analogues of cAMP or diacylglycerol plus okadaic acid. Mol Cell Biochem 127–128:121–129
Goldstein BJ 2001 Protein-tyrosine phosphatase 1B (PTP1B): a novel therapeutic target for type 2 diabetes mellitus, obesity and related states of insulin resistance. Curr Drug Targets Immune Endocr Metab Disord 1:265–275
Myers MP, Andersen JN, Cheng A, Tremblay ML, Horvath CM, Parisien JP, Salmeen A, Barford D, Tonks NK 2001 TYK2 and JAK2 are substrates of protein-tyrosine phosphatase 1B. J Biol Chem 276:47771–47774
Elchebly M, Payette P, Michaliszyn E, Cromlish W, Collins S, Loy AL, Normandin D, Cheng A, Himms-Hagen J, Chan CC, Ramachandran C, Gresser MJ, Tremblay ML, Kennedy BP 1999 Increased insulin sensitivity and obesity resistance in mice lacking the protein tyrosine phosphatase-1B gene. Science 283:1544–1548
Kaszubska W, Falls HD, Schaefer VG, Haasch D, Frost L, Hessler P, Kroeger PE, White DW, Jirousek MR, Trevillyan JM 2002 Protein tyrosine phosphatase 1B negatively regulates leptin signaling in a hypothalamic cell line. Mol Cell Endocrinol 195:109–118
Delamere NA, Williams RN 1986 Modulation by vanadate of the adrenergic characteristics of the iris, ileum, and vas deferens. Invest Ophthalmol Vis Sci 27:1336–1341
Commins SP, Marsh DJ, Thomas SA, Watson PM, Padgett MA, Palmiter R, Gettys TW 1999 Norepinephrine is required for leptin effects on gene expression in brown and white adipose tissue. Endocrinology 140:4772–4778
Li H, Matheny M, Scarpace PJ 1997 3-Adrenergic-mediated suppression of leptin gene expression in rats. Am J Physiol 272:E1031–E1036(Jared Wilsey, Michael K. Matheny and Phi)