SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. REFERENCES

Mounting evidence suggests that the endocannabinoid system regulates energy metabolism through direct effects on peripheral tissues as well as central effects that regulate appetite. Here we examined the effect of cannabinoid receptor 1 (CB1) signaling on insulin action in fat cells. We examined effects of the natural CB1 agonist, 2-Arachidonoylglycerol (2-AG), and the synthetic CB1 antagonist, SR141716, on insulin action in cultured adipocytes. We used translocation of glucose transporter GLUT4 to plasma membrane (PM) as a measure of insulin action. 2-AG activation of the CB1 receptor promoted insulin sensitivity whereas antagonism by SR141716 reduced insulin sensitivity. Neither drug affected GLUT4 translocation in the absence of insulin or with high doses of insulin. Consistent with these results we found that insulin-stimulated phosphorylation of the protein kinase Akt was increased by 2-AG, attenuated by SR141716, and unaffected in the absence of insulin or by addition of high-dose insulin. These data provide a functional and molecular link between the CB1 receptor and insulin sensitivity, because insulin-stimulated phosphorylation of Akt is required for GLUT4 translocation to the PM. The sensitizing effects of 2-AG were abrogated by SR141716 and Pertussis toxin, indicating that the effects are mediated by CB1 receptor. Importantly, neither 2-AG nor SR141716 alone or in combination with maximal dose of insulin had effects on GLUT4 translocation and Akt phosphorylation. These data are consistent with a model in which the endocannabinoid system sets the sensitivity of the insulin response in adipocytes rather than directly regulating the redistribution of GLUT4 or Akt phosphorylation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. REFERENCES

Insulin has a prominent role in the regulation of glucose homeostasis and energy metabolism. Insulin resistance is a component of the metabolic syndrome, associated with cardiovascular disease, and is a prelude to type 2 diabetes mellitus. Obesity is a significant risk factor for insulin resistance and the prevalence of insulin resistance will rise with the increase in the prevalence of obesity (1). Consequently, it is important to understand factors that regulate insulin sensitivity.

The view of the role of adipose tissue in metabolism is changing from one limited to the storage of triglycerides to one in which adipose tissue has a more global role in regulation of nutrient metabolism exerted through the effects of secreted molecules with endocrine functions (adipokines) (2,3). Because of the central role of adipose tissue in the regulation of energy metabolism, understanding modulation of insulin sensitivity in adipocytes is especially relevant.

The endocannabinoid receptor 1 (CB1) is expressed in central nervous system and peripheral tissues, particularly those involved in insulin action: liver, muscle, pancreas, and adipose tissue. There is mounting evidence that the endocannabinoid system has a role in regulating metabolism (4,5). In the brain, the endocannabinoid system functions in the regulation of appetite (6). Recent data suggest that a disruption in endocannabinoid signaling may be involved in obesity (e.g., 4,7,8,9,10). There is evidence that prolonged inhibition of CB1 receptor with a selective antagonist, SR141716, may promote weight loss and improve insulin sensitivity by directly affecting peripheral energy metabolism (reviewed in 4,11). Adipose tissue expresses CB1 receptor and therefore might be one of the main targets for the insulin-sensitizing effects of CB1 antagonist SR141716 (12,13).

Here we report on our studies the effect of CB1 receptor on insulin sensitivity of cultured mouse adipocytes. We found that activation of the CB1 receptor by the endocannabinoid, 2-Arachidonoylglycerol (2-AG), promotes insulin sensitivity, whereas antagonism of CB1 receptor with SR141716 reduces sensitivity to insulin.

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. REFERENCES

Antibodies and reagents

SR141716 was provided by Sanofi-Aventis and 2-AG was purchased from Biomol International (Plymouth Meeting, PA). Stocks of SR141716 and 2-AG were made in dimethylsulfoxide. Dimethylsulfoxide was used as a vehicle for the control conditions.

Mouse antihemagglutinin epitope monoclonal antibody (HA.11; Covance, Berkley, CA) was purified from ascites (14). Goat antimouse antibody labeled with Cy3 was purchased from Jackson Immunolabs. (West Grove, PA). Rabbit polyclonal antibody for CB1 receptor was purchased from Cayman Chemical (Ann Arbor, MI), and insulin receptor β chain (C-19) from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit polyclonal antibodies against phospho-Akt (Ser473, Thr308), and a pan Akt antibody were purchased from Cell Signaling Technology (Beverly, MA). An anti-GLUT4 antibody was a gift from Sam Cushman (National Institute of Health, Bethesda, MD).

Adipocyte cell culture

The 3T3-L1 adipocytes were maintained in Dulbecco's modified Eagle's medium containing 10% calf serum and differentiated as described (15). Treatment of adipocytes with the drugs began at day 5 of differentiation, a time we routinely use for the analysis of insulin action and GLUT4 behavior (16). For most experiments cells were incubated for a total of 48 h with the drugs (starting at day 5 of differentiation) before assessing insulin sensitivity. During the 48 h incubation with drugs, the medium was replaced every 24 h with fresh medium containing drugs. The drugs were maintained in the medium during the 30 min stimulation with insulin. For some experiments, on day 5 of differentiation drugs were added to media 20 min prior to and during the 30 min insulin stimulation (a total of 50 min).

3T3-L1 adipocytes stably expressing, by retroviral infection, a HA-GLUT4 reporter were used for most studies. The HA-GLUT4 construct has an HA epitope tag inserted in the first exofacial loop of GLUT4 and this reporter has been extensively used in studies of GLUT4 trafficking (e.g., 16,17). The amount of HA-GLUT4 in the plasma membrane (PM) is determined by indirect immunofluorescence (IF) of fixed cells (see below). We also studied the behavior of a second GLUT4 reporter, HA-GLUT4-GFP, which in addition to the HA epitope also has a GFP fused to carboxyl terminus of GLUT4 (15). In these studies, the reporter is transiently expressed in adipocytes by electroporation. This reporter has been extensively used in the quantitative analyses of GLUT4 trafficking as well (e.g., 15,18,19).

Translocation assay

On the day of the experiment, cells were incubated in serum-free medium for 150 min. By convention, at the end of this incubation the cells are considered to be in the basal or unstimulated state. For insulin stimulation, following the preincubation with serum-free media, cells were incubated with insulin and the drugs for 30 min at 37 °C. At the end of the incubations, cells were fixed with 3.7% formaldehyde for 7 min.

PM HA-GLUT4 was determined as previously described (16). In each experiment, all images of the same fluorophore were collected with the same exposure. Data were collected from fields of confluent cells, typically 10 fields of cells per condition. The average fluorescence intensity per field, corrected for background (fluorescence determined from confluent fields of cells without primary antibody), was determined for each condition. In each experiment, the average fluorescence intensity/field of the various treatment conditions was normalized to the value of the 10 nmol/l insulin vehicle control. Total cellular HA-GLUT4 (intracellular and surface) was measured by indirect IF of cells permeabilized with 250 μg/ml saponin in PBS with 5% calf serum. For HA-GLUT4-GFP, GFP fluorescence is a measure of total expression of the construct, and the amount of HA-GLUT4-GFP on the surface of intact cells is determined with a fluorescent labeled antibody against the HA epitope, as discussed for HA-GLUT4. The microscopy data were collected and analyzed as previously described (15,20).

Western blot

Proteins were resolved in sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to nitrocellulose membranes blocked with 5% casein for 1 h, and incubated with primary antibodies overnight at 4 °C. Horseradish peroxidase-labeled secondary antibody was visualized by enhanced chemiluminescence (ECL system; GE Healthcare, Freiburg, Germany). Quantitative analysis of western blot images was performed using the Scion software (Scion, Frederick, MD).

Pertussis toxin

To further evaluate the role of CB1 signaling on insulin action and GLUT4 translocation, cells were incubated with Pertussis toxin for 14 h, starting the night before the experiment. On the day of the experiment, cells were incubated with serum-free media containing Pertussis toxin for 150 min and at the end of that incubation cells were incubated for 20 min with 2-AG, 2-AG and Pertussis toxin, Pertussis toxin alone, or dimethylsulfoxide as a vehicle control. Finally, different concentrations of insulin were added to the media for an additional 30 min.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. REFERENCES

CB1 modulates insulin-stimulated redistribution of GLUT4 to the PM

To test the hypothesis that CB1 receptor is a modulator of insulin action in adipocytes, we determined the effects of CB1 agonist, 2-AG (13.2 μmol/l), and CB1 antagonist, SR141716 (200 nmol/l), on insulin-induced translocation of the GLUT4 glucose transporter to the PM. In adipose tissue, insulin regulates glucose uptake by acutely recruiting the GLUT4 glucose transporter from intracellular storage sites to the PM (21,22). This is one of the best-described actions of insulin at a cellular level. The concentrations of the 2-AG and SR141716 were chosen based on previous studies in cultured cells (23,24). Cells were maintained in 2-AG or SR141716 for a total of 48 h with a change of media after 24 h. Following the 48 h incubation, insulin-induced recruitment of HA-GLUT4 to the PM was assayed (15). Changes in PM GLUT4 can be used to monitor insulin action. The amount of HA-GLUT4 in the PM was determined by quantitative anti-HA epitope IF microscopy of intact cells.

In basal unstimulated adipocytes, little HA-GLUT4 was in the PM and the insulin-induced redistribution of HA-GLUT4 to the PM was readily apparent in anti-HA IF of intact adipocytes (Figure 1a). The increase in HA-GLUT4 in the PM of adipocytes stimulated with 170 nmol/l insulin corresponds to an approximate sevenfold translocation (Figure 1b). Based upon qualitative analysis of the IF, the 48 h incubation with SR141716 did not affect the amount of HA-GLUT4 in the PM (Figure 1a). Similarly, 2-AG did not affect HA-GLUT4 in the PM of basal adipocytes; however, there was a clear increase in HA-GLUT4 in the PM of the 2-AG-treated adipocytes at 1 nmol/l compared to both control and SR141716-treated adipocytes (Figure 1a). Qualitatively there was no effect of 2-AG on HA-GLUT4 in the PM of adipocytes stimulated with 10 or 170 nmol/l insulin.

image

Figure 1. CB1 modulates insulin-stimulated redistribution of HA-GLUT4 to the plasma membrane of adipocytes. (a) Anti-HA indirect immunofluorescence (IF) of intact cells detects HA-GLUT4 in the plasma membrane. Day 5 adipocytes were treated with 13.2 μmol/l 2-AG, 200 nmol/l SR141716, or 0.1% dimethylsulfoxide (DMSO) (vehicle) for 48 h, with a change of medium after 24 h, before the 30 min stimulation with the noted doses of insulin. The compounds or DMSO were included in the incubation with insulin. All the images were collected and printed with the same settings so that the intensities of the fluorescence in the different conditions are proportional to the relative amounts of HA-GLUT4 in the plasma membrane. (b) Quantification of anti-HA IF from cells treated as in a. The data are averages of 10 experiments ± s.e.m. Results from the individual experiments were normalized to the control 10 nmol/l insulin value. *P < 0.01, two-tailed paired t- test. (c) Anti-HA IF of saponin permeabilized adipocytes. Cells were treated with drugs as in a. Basal cells were fixed, permeabilized, and stained with anti-HA to reveal intracellular HA-GLUT4. (d) Total HA-GLUT4 was determined by quantification of data from experiments in b. Data are means ± s.e.m. of three experiments. Data were normalized to the mean total HA-GLUT4 of control adipocytes (DMSO treated).

Download figure to PowerPoint

Quantitative analysis of the effect of insulin on HA-GLUT4 in the PM revealed a significant increase of HA-GLUT4 in the PM of 2-AG treated adipocytes stimulated with 1 nmol/l insulin compared to control adipocytes (59% increase, P < 0.01) (Figure 1b). The effect of 10 nmol/l insulin was also potentiated in the 2-AG treated adipocytes, albeit to a lesser degree (25% increase, P < 0.01). There were no differences in PM HA-GLUT4 between control and 2-AG treated adipocytes in either basal or 170 nmol/l insulin-stimulated conditions. Treatment with SR141716 reduced HA-GLUT4 translocation to PM at both 1 nmol/l and 10 nmol/l insulin compared to control adipocytes (P < 0.01) (Figure 1b). As was the case for 2-AG, in basal and 170 nmol/l insulin-stimulated adipocytes, there were no differences in PM HA-GLUT4. We examined the effects of three different doses of 2-AG on insulin action. The effect of 2-AG on insulin sensitivity was maximum at 13.2 μmol/l as a tenfold lower dose had a smaller effect on insulin sensitivity and a fivefold increase in 2-AG concentration did not promote a greater effect on insulin sensitivity than that induced by 13.2 μmol/l (not shown).

One explanation for the changes in the amounts of HA-GLUT4 in the PM is that the drugs affect the expression of the HA-GLUT4 (e.g., 2-AG might increase and SR141716 might decrease the amount of HA-GLUT4 expressed in cells). To investigate this possibility, we measured total HA-GLUT4 in permeabilized adipocytes by quantitative IF (Figure 1c, d). There were no differences among the different conditions in the average HA-GLUT4 expressed per cell. Furthermore, examination of HA-GLUT4 in permeabilized cells did not reveal any differences in the intracellular localization of HA-GLUT4 among the different treatments. In all cases, the intracellular HA-GLUT4 was distributed between the perinuclear region of cells and punctate structures throughout the cytosol (Figure 1c).

Analysis of the behavior of HA-GLUT4-GFP is an alternative method for directly assessing the effect of insulin on GLUT4 distribution between the PM and the interior of adipocytes. In this procedure, the redistribution of HA-GLUT4-GFP to the PM is measured by a ratiometric method in individual cells transiently expressing the reporter (15). An important difference between these methods is that in one case stably expressing cells are used (HA-GLUT4) and in the other transiently expressing cells are used (HA-GLUT4-GFP). To confirm that the results of Figure 1 are not confounded by using adipocytes stably expressing the HA-GLUT4 reporter, we examined the effects of 2-AG and SR171416 on the behavior of HA-GLUT4-GFP as well. In agreement with the studies using HA-GLUT4, 48 h exposure to 2-AG increased GLUT4 translocation to the PM stimulated by 1 nmol/l insulin, whereas exposure to SR141716 decreased the redistribution (Figure 2). Similarly, PM HA-GLUT4-GFP did not vary among the different conditions in basal adipocytes or adipocytes stimulated with 170 nmol/l insulin. By using a second method for detecting the effect of insulin of GLUT4 redistribution to the PM, these data confirm that modulation of CB1 receptor affects insulin-stimulated recruitment of GLUT4 to the PM: the CB1 agonist, 2-AG, increases sensitivity of adipocytes to insulin, and the CB1 antagonist, SR141716, reduces the response to insulin.

image

Figure 2. CB1 modulates insulin-stimulated redistribution of HA-GLUT4-GFP to the plasma membrane. Cells were treated as in Figure 1a. The data are the averages of five experiments ± s.e.m. normalized to the averages of the control 10 nmol/l insulin of the same experiment. *P < 0.05, two-tailed paired t -test.

Download figure to PowerPoint

SR141716 blocks the insulin-sensitizing effect of 2-AG

To test the hypothesis that 2-AG and SR141716 act through the CB1 receptor to modulate insulin sensitivity, we first established that the cells express CB1 and that the various treatments do not alter CB1 expression (Figure 3a). We next determined the effect of coincubation of cells with 2-AG and SR141716 (Figure 3b). Coincubation of adipocytes with 2-AG and SR141716 normalized the insulin response supporting the hypothesis that both 2-AG and SR141716 modulate insulin sensitivity through the CB1 receptor.

image

Figure 3. CB1 antagonist SR141716 blocks the insulin-sensitizing effect of CB1 agonist 2-AG. (a) Western immunoblot of CB1 in adipocytes treated for 48 h with 200 nmol/l SR141716, 13.2 μmol/l 2-AG, or dimethylsulfoxide (control). Extracts were made from unstimulated adipocytes and adipocytes stimulated with the noted doses of insulin for 30 min. Actin was measured as a sample loading control. (b) Cells were treated with 13.2 μmol/l 2-AG, 200 nmol/l SR141716, both drugs, or DMSO (control) for 48 h. Insulin stimulation was for 30 min. HA-GLUT4 in the plasma membrane was determined by anti-HA IF. The data from the individual experiments were normalized to the HA-GLUT4 value measured in control cells at 10 nmol/l insulin. The values are the averages of four experiments ± s.e.m.

Download figure to PowerPoint

GLUT4 and insulin receptor expression are not altered by 2-AG or SR141716

The expression of GLUT4 and sensitivity of GLUT4 trafficking to insulin develops during differentiation of preadipocytes to adipocytes and therefore these characteristics are markers of the differentiated phenotype. One possibility for the alteration in insulin sensitivity induced by the CB1 agonist and antagonist is that they affect differentiation, or more accurately, maintenance of the differentiated state because the cells are first exposed to the drugs at Day 5 of differentiation, a time at which GLUT4 is expressed and insulin is fully capable of regulating trafficking (25). The 48-h incubations with 2-AG or SR141716 did not alter the expression of endogenous GLUT4, indicating that the drugs were not markedly altering the differentiated state of the cells (Figure 4). In addition, the treatments did not alter the expression of the insulin receptor supporting the conclusion that the effects of the CB1 agonist and antagonist are neither due to marked alterations in differentiation nor due to changes in the insulin receptor expression (Figure 4). Furthermore, modulation of CB1 activity did not alter the amount of GLUT4 or insulin receptor following activation of the receptor by insulin indicating that the effects of 2-AG and SR141716 were not due to changes in the stability of the insulin receptor after activation (Figure 4).

image

Figure 4. CB1 antagonist SR141716 and agonist 2-AG do not alter expression on GLUT4 or insulin receptor in adipocytes. Western immunoblot of CB1 in adipocytes treated for 48 h with 200 nmol/l SR141716, 13.2 μmol/l 2-AG, or dimethylsulfoxide (control). Extracts were made from unstimulated adipocytes and adipocytes stimulated with the noted doses of insulin for 30 min. Actin was measured as a sample loading control.

Download figure to PowerPoint

Insulin-stimulated phosphorylation of Akt is potentiated by 2-AG

Insulin signals to GLUT4 in part through activation of the serine/threonine kinase Akt (26,27). Akt activation is critical for the redistribution of GLUT4 to the PM, therefore, the altered insulin sensitivity induced by the CB1 agonist and antagonists could be due to alteration in Akt activation. To investigate this possibility, we examined Akt in 2-AG and SR141716-treated adipocytes. Neither 2-AG nor SR141716 altered the total amount of Akt expressed in cells (Figure 5). To determine the effects of 2-AG and SR141716 on Akt activation, we examined insulin-stimulated Akt phosphorylation on Ser473. This phosphorylation is required for activation of Akt and is routinely used as a monitor of Akt activation. Neither SR141716 nor 2-AG induced a detectable change in Akt phosphorylation of basal adipocytes (Figure 5). However, Akt phosphorylation in adipocytes stimulated with 1 nmol/l insulin was increased in 2-AG-treated cells and reduced by SR141716 (Figure 5). These data are consistent with the potentiation of the effect of 1 nmol/l insulin on GLUT4 redistribution by 2-AG (Figures 1 and 2). Akt phosphorylation in 10 nmol/l insulin-stimulated conditions was not significantly altered by 2-AG or SR141716 (Figure 5). In addition to Ser473, insulin stimulates the phosphorylation of Akt on Thr308. Consistent with 2-AG acting as an insulin sensitizer, 2-AG increased Akt phosphorylation at the Thr308 site stimulated by 1 nmol/l insulin (not shown).

image

Figure 5. CB1 receptor modulates insulin-induced phosphorylation of Akt. (a) Western immunoblot of lysates from adipocytes treated for 48 h with 200 nmol/l SR171416, 13.2 μmol/l 2-AG, or dimethylsulfoxide (control). Insulin treatments were for 30 min. Phospho-Ser473 and total Akt were measured from the same extracts. Actin was measured on both blots as a sample loading control. (b) Quantification of Ser473 phosphorylation. The data are the averages of five experiments ± s.e.m. The band intensities were normalized to actin of the corresponding lane and the signal for Ser473 of each lane was normalized to total Akt of the same lane. The Ser473 or total Akt ratios from each condition were normalized to the value from the control condition at 10 nmol/l of insulin. Results are average of values from four experiments. *P < 0.05. **P = 0.06.

Download figure to PowerPoint

Acute modulation of CB1 receptor affects insulin sensitivity

To investigate whether these drugs acutely alter insulin action, we determined the effects of 50 min treatment (20 min preincubation with 2-AG, SR141716 or dimethylsulfoxide followed by 30 min incubation with insulin and the drugs) on insulin-induced redistribution of HA-GLUT4 to the PM of adipocytes. The results of the acute treatments had similar effect on GLUT4 as did the 48 h treatments (Figure 6a). As was the case with the longer exposure to 2-AG (Figure 1), the maximum effect of 2-AG on insulin sensitivity measured by GLUT4 recruitment to the PM was at 13.2 μmol/l (not shown).

image

Figure 6. Acute modulation of CB1 receptor activity alters insulin response of adipocytes. (a) Day 5 adipocytes were treated with 13.2 μmol/l 2-AG, 200 nmol/l SR141716 or dimethylsulfoxide (DMSO) (vehicle) for 20 min before the 30 min stimulation with the noted doses of insulin. The drugs were included in the incubation with insulin. The data are averages of three experiments. Each bar is the mean ± s.e.m. plasma membrane HA-GLUT4 of each condition normalized to the mean of the control 10 nmol/l insulin of the same experiment. *P < 0.05. **P = 0.09. (b) Western immunoblot of lysates from adipocytes treated as in a. Phospho-Ser473 Akt and total Akt were measured from the same extracts. Actin was measured on both blots as a sample loading control. (c) Quantification of Ser473 phosphorylation. The band intensities were normalized to actin of the corresponding lane and the signal for Ser473 of each lane was normalized to total Akt of the same lane. The Ser473 or total Akt ratios from each condition were normalized to the value from the control condition at 10 nmol/l of insulin. Results are average of values from four experiments. *P < 0.05.

Download figure to PowerPoint

The 50 min exposure to 2-AG induced Akt phosphorylation in 1 nmol/l and 10 nmol/l insulin-stimulated adipocytes, but not in unstimulated condition, in agreement with the effects on HA-GLUT4 distribution (Figure 6b, c). The 50 min incubation with SR141716 did not significantly alter Akt phosphorylation in any condition.

Pertussis toxin abolished the insulin-sensitizing effect of 2-AG in adipocytes

CB1 is a Gi protein-coupled receptor (28). Pertussis toxin interrupts Gi protein-coupled receptor signaling by catalyzing the ADP ribosylation of Gαi, locking it in the GDP-bound state, and rendering it inactive. To investigate whether CB1 modulates insulin sensitivity through Gαi, we determined the effect of Pertussis toxin on the effect of 2-AG on insulin action. Adipocytes were preincubated with Pertussis toxin for 14 h and the effects of acute treatments with 2-AG (50 min) on insulin action were examined. Pertussis toxin treatment reduced the stimulatory effect of 2-AG on GLUT4 translocation at 1 nmol/l insulin supporting the hypothesis that 2-AG modulates insulin sensitivity through the Gαi signaling (Figure 7).

image

Figure 7. Pertussis toxin abolishes the effect of CB1 agonist on insulin response. Day 5 adipocytes were incubated in Pertussis toxin 100 ng/ml over night and subsequently preincubated with dimethylsulfoxide (vehicle), 13.2 μmol/l 2-AG alone, Pertussis toxin alone, or 13.2 μmol/l 2-AG and Pertussis toxin combined for 20 min before the 30 min stimulation with the noted doses of insulin. The drug and Pertussis toxin were also included in the incubation with insulin. The data are averages of four experiments. Each bar is the mean ± s.e.m. plasma membrane HA-GLUT4 of each condition normalized to the mean of the control 10 nmol/l insulin of the same experiment. *P < 0.05.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. REFERENCES

We have shown that 2-AG, a CB1 agonist, enhances the sensitivity of adipocytes to insulin, whereas the CB1 antagonist, SR141716, reduces insulin sensitivity. One of the primary acute effects of insulin is to stimulate glucose uptake into adipocytes and muscle cells, a process essential for disposal of dietary glucose (21). The regulation of glucose uptake is achieved through the recruitment of the GLUT4 glucose transporter from intracellular compartments to the PM, therefore, analysis of GLUT4 translocation to the membrane is a measure of a key physiologic function of insulin. In our studies we used two validated methods to study GLUT4 translocation: (i) 3T3-L1 adipocytes stably expressing HA-GLUT4 and (ii) 3T3-L1 adipocytes transiently expressing HA-GLUT4-GFP. The results of both approaches support the conclusion that 2-AG and SR141716 modulate insulin-stimulated translocation of GLUT4 to the PM. Our findings that CB1 agonists and antagonist have opposite effects on insulin sensitivity, coupled with the results that the stimulatory effect of 2-AG is abrogated by cotreatment of adipocytes with the CB1 antagonist SR141716 and by Pertussis toxin, argue that the 2-AG effect is mediated by the CB1 receptor.

Neither 2-AG nor SR141716 alone affected GLUT4 in the PM in basal conditions (no insulin) nor did either drug significantly alter the effect of a dose of insulin (170 nmol/l) that induces the maximum redistribution of GLUT4 to the PM. The effects of the drugs were most evident at 1 nmol/l insulin, a dose of insulin that stimulates about half maximum redistribution of GLUT4 to the PM of cultured 3T3-L1 adipocytes. These data support the hypothesis that the CB1 endocannabinoid system modulates the sensitivity of insulin response rather than directly activating or inhibiting the signal transduction cascade that results in redistribution of GLUT4 to the PM. Although it is difficult to extrapolate the effects of various doses of insulin on cultured adipocytes to insulin effects in situ, it is of note that the endocannabinoid system exerts effects on insulin concentrations that are near the physiologic range (postprandial serum insulin levels in mouse ∼0.6 nmol/l), and therefore the CB1 modulation of insulin sensitivity detected in cultured adipocytes may have clinical relevance.

Our finding that CB1 modulates insulin sensitivity is in general agreement with a recent report (29). In this study, the authors found that a 24 h treatment of day 9 adipocytes with anandamide resulted in a 60% increase in glucose uptake stimulated by 100 nmol/l insulin and a smaller effect after a 48 h exposure consistent with our findings of an increased GLUT4 translocation induced by 2-AG. Gasperi et al. (29) also reported that the stimulatory effect of anandamide was blocked by SR141716 in agreement with our results (Figure 3). One difference in our findings is that we did not find an effect of 2-AG at higher insulin concentrations (170 nmol/l). In the other study, the authors only examined the effect of anandamide on 100 nmol/l insulin and therefore we do not know whether the anandamide would enhance lower insulin doses (29). Regardless, the results of our study and the study of Gasperi et al. are complementary and in agreement with the proposal that CB1 has a role in determining insulin sensitivity of adipocytes.

We did not measure glucose uptake because it has been shown that in 3T3-L1 adipocytes about half of the insulin-regulated glucose uptake is through the GLUT1 glucose transporter (30), whereas in primary adipocytes, GLUT4 is the predominant glucose isoform expressed and glucose transport through GLUT1 does not contribute greatly to the effects of insulin on glucose flux (31). Thus, the high level of GLUT1 expression in 3T3-L1 adipocytes confounds interpretation of glucose uptake data and changes in GLUT4 behavior assayed by glucose uptake might be obscured by the GLUT1.

It has also been reported that activation of the endocannabinoid system in human adipocytes promotes GLUT4 translocation and glucose uptake independent of insulin (32). This is in contrast to our findings and there are several possible explanations for the differences in results. In that study, the authors used a synthetic CB1 agonist, WIN 55,212 rather than a native endocannabinoid. It is known that the different endocannabinoids, 2-AG and anandamide, can differentially activate G proteins coupled to CB1 (reviewed in 33) and it is therefore possible that the differences are due to use of different agonist. The authors did not detect additive effects of WIN 55,212, and insulin although only a single-high dose of insulin (1 μmol/l) was used and it is possible that activation of CB1 in human adipocytes would potentiate the effects of more physiologic doses of insulin.

How does CB1 modulate insulin action? CB1 receptor is a 7-transmembrane Gi/o protein-coupled receptor negatively coupled to adenylyl cyclase (11,34,35). Insulin receptor signals to GLUT4 through phosphoinositol 3-kinase (PI3 K) and Akt. Previous studies in other cell types have demonstrated a coupling of CB1 receptor to the PI3 K/Akt signaling (36,37,38,39) establishing precedence for cross talk between the CB1 and the PI3K/Akt signaling pathways. Our study demonstrates that in cultured adipocytes 2-AG does not activate Akt independently but rather sensitizes adipocytes to insulin-mediated activation of Akt, whereas the CB1 antagonist, SR141716, reduces the sensitivity of adipocytes to insulin-mediated activation of Akt. The fact that Pertussis toxin blocks the 2-AG effect provides additional evidence that the 2-AG effect is coupled to insulin sensitivity through Gαi, a known mediator of the CB1 receptor signaling.

The effect of SR141716 to blunt insulin response in the absence of 2-AG indicates that it does not act like a neutral antagonist. One explanation for this effect is that SR141716 blocks the action of endogenously produced endocannabinoid. Alternatively, SR141716 might signal through the CB1 receptor as an inverse agonist. As discussed in a review by Pertwee, in almost every study of SR141716 it has been noted that the drug has an “opposite” effect to endocannabinoids (40). For example, the reverse activity of SR141716 has been observed in studies of the effect of SR141716 on locomotor activity, memory, nociception, emesis, intestinal transit, spontaneous firing of neurons, and so on (40). Thus, the effects of SR141716 on insulin sensitivity that we observed in this study are consistent with previous reports of SR141716 being an inverse agonist.

There is evidence for dysregulation of peripheral endocannabinoid system in obese humans. Patients with diabetes and obesity have enhanced endocannabinoid levels in their adipose tissue and circulating 2-AG was positively correlated with the amount of body fat, visceral fat mass, and fasting insulin levels in obese human subjects (41,42). The observations that elevated endocannabinoids are correlated with obesity and insulin resistance and that antagonism of CB1 increases the expression of the insulin-sensitizing adipokine adiponectin may seem to suggest that activation of CB1 is a negative modulator of insulin action. However, our results suggest the contrary. It has been shown in a study of primary rodent adipocytes that activation of CB1 receptor stimulated lipogenesis (9). Those results together with our data suggest that activation of CB1 in adipocytes promotes the insulin effect that include lipid accumulation and energy storage. Thus, the increased endocannabinoid levels observed in human obesity may constitute an initial compensatory mechanism to improve insulin function and to overcome insulin resistance associated with obesity. This compensatory mechanism may increase the efficiency of the system initially but upon prolonged activation of CB1 (e.g., as would be the case in chronically overweight individuals) the increased sensitivity of Akt to insulin activation might lead to hyperactivation of Akt and subsequent downregulation of signaling downstream to Akt. Subsequently, one may predict that chronic treatment with CB1 antagonist SR141716 in obesity might have a protective effect on this chain of events.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. REFERENCES

This work was funded by a grant from Sanofi-Aventis. We thank the members of the McGraw lab, specifically Drs Gonzalez, Blot, and Xiong, for critical reading of the manuscript and for many helpful discussions and Tuan Daniel Chuang and Abraham Cespedes for expert technical assistance. The fluorescence microscopy was performed in the Weill Cornell Medical College Optical Microscopy Core Facility.

Disclosure

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. REFERENCES

The work was funded by Sanofi-Aventis, maker of SR141716. The authors disclose this apparent conflict.

REFERENCES

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Disclosure
  9. REFERENCES