Hepatic-dependent and -independent Insulin Actions Are Impaired in the Obese Zucker Rat Model


  • Ricardo A. Afonso,

    1. Department of Biochemistry, Faculty of Medical Sciences, New University of Lisbon, Lisbon, Portugal
    2. Department of Physiology, Faculty of Medical Sciences, New University of Lisbon, Lisbon, Portugal
    Search for more papers by this author
  • Rogerio T. Ribeiro,

    1. Department of Physiology, Faculty of Medical Sciences, New University of Lisbon, Lisbon, Portugal
    Search for more papers by this author
  • Ana B. Fernandes,

    1. Department of Physiology, Faculty of Medical Sciences, New University of Lisbon, Lisbon, Portugal
    Search for more papers by this author
  • Rita S. Patarrão,

    1. Department of Biochemistry, Faculty of Medical Sciences, New University of Lisbon, Lisbon, Portugal
    2. Department of Physiology, Faculty of Medical Sciences, New University of Lisbon, Lisbon, Portugal
    Search for more papers by this author
  • M. Paula Macedo

    Corresponding author
    1. Department of Physiology, Faculty of Medical Sciences, New University of Lisbon, Lisbon, Portugal
    2. Portuguese Diabetes Association, Lisbon, Portugal.
    Search for more papers by this author

Department of Physiology, Faculty of Medical Sciences, New University of Lisbon, Campo Martires da Patria 130, 1169-056 Lisbon, Portugal. E-mail: mpmacedo.biot@fcm.unl.pt


Objective: Whole-body insulin sensitivity (IS) depends on a hepatic pathway, involving parasympathetic activation and hepatic nitric oxide (NO) production. Both atropine and N-monomethyl-l-arginine (l-NMMA, NO synthase inhibitor) induce insulin resistance (IR). IR is associated with obesity. Because NO action was shown to be impaired in animal models of obesity, such as the obese Zucker rat (OZR), we tested the hypothesis that the hepatic-dependent pathway is diminished in OZR, resulting in IR.

Research Methods and Procedures: Lean Zucker rats (LZRs) were used as OZR controls. IS was evaluated in terms of glucose disposal [milligrams of glucose per kilogram of body weight (bw)]. Two groups were submitted to two protocols. First, a control clamp was followed by a post-atropine (3 mg/kg intravenously) clamp. Second, after the control clamp, l-NMMA (0.73 mg/kg intraportally) was given, and a second clamp was performed. Hepatic-dependent IS was assessed by subtracting the response after atropine or l-NMMA from the basal response.

Results: In the first protocol, basal IS was lower in OZR than in LZR (OZR, 73.7 ± 14.2; LZR, 289.2 ± 24.7 mg glucose/kg bw; p < 0.001), and atropine decreased IS in the same proportion for both groups (OZR, 41.3 ± 8.0%; LZR, 40.1 ± 6.5%). Equally, in the second protocol, OZR presented lower IS (OZR, 79.3 ± 1.6; LZR, 287.4 ± 22.7 mg glucose/kg bw; p < 0.001). l-NMMA induced IS inhibition in both groups (OZR, 48.3 ± 6.6%; LZR, 46.4 ± 4.1%), similar to that after atropine.

Discussion: We show that the IR in OZR is due to similar impairment of both hepatic-dependent and -independent components of insulin action, suggesting the existence of a defect common to both pathways.


A number of studies indicate a central role of the liver in the process of glucose uptake by extrahepatic tissues (1, 2, 3, 4, 5), in which hepatic parasympathetic nerves (HPNs)1 were found to be involved (3, 5) and where a putative humoral factor is released from the liver to potentiate peripheral glucose uptake (2, 4). Surgical hepatic parasympathetic denervation performed in healthy animals decreases insulin effectiveness substantially, and this condition is not further aggravated by intraportal (IPV) administration of atropine, either in cats or rats (3). Such observation places the HPNs as important players in the process of peripheral insulin action. This parasympathetic activation increases in the postprandial state, declining progressively with fasting (6). Feeding stimulates the release of acetylcholine from HPNs, which activates M1 muscarinic receptors (3), leading to the production of nitric oxide (NO) in the liver and subsequent release of a humoral factor, the hepatic insulin-sensitizing substance (HISS), which potentiates insulin action selectively in the skeletal muscle (7, 8). Hepatic but not peripheral NO is essential in the insulin-sensitizing action of the hepatic-dependent pathway. This consideration is supported by the finding that IPV but not intravenous (IV) administration of submaximal doses of two competitive NO synthase inhibitors [Nω-nitro-l-arginine methyl ester (L-NAME) and N-monomethyl-l-arginine (l-NMMA)] induces insulin resistance (IR) of similar magnitude to that caused by hepatic denervation or atropine administration (7). Furthermore, it was shown that a submaximal dose of the NO donor 3-morpholinosydnomine (SIN-1) was capable of reversing IR caused by hepatic NO synthase inhibition but only if administered IPV (7, 8).

Therefore, peripheral insulin action can be divided into two components: one dependent on the hepatic parasympathetic stimulation and NO production in the liver (hepatic pathway), which accounts for ∼50% of the peripheral glucose disposal effect of insulin (3, 7), and another which is independent of the hepatic pathway (i.e., insulin action per se).

The association among obesity, IR, and the onset of type 2 diabetes is well documented. Because obesity is associated with parasympathetic (9) and NO (10) dysfunctions, we used the obese Zucker rat (OZR) as an animal model of obesity, which is also known as Zucker Fatty (fa/fa genotype). OZR is an autosomal recessive genetic model of obesity (10), characterized by a defective leptin receptor that may account for the severe adiposity and hyperphagia observed in these animals (11, 12). OZR is often characterized as being insulin resistant (12, 13).

In the present study, we proposed to characterize the hepatic-dependent pathway and its involvement in the IR observed in obesity. We tested the hypothesis that this component of insulin action is impaired in the OZR and consequently contributing to the state of IR in these animals.

Research Methods and Procedures

Animal Subjects

Lean Zucker rats (LZRs, n = 16) were used as controls of the obese animals (OZR, n = 16). All rats were 9-week-old males, purchased from Charles River Laboratories (Barcelona, Spain) and used 1 to 2 weeks after arrival. All animals were housed in air-conditioned quarters and subjected to a 12-hour light/dark cycle (8 am to 8 pm). Rats had free access to food (chow pellets Panlab A04, Charles River Laboratories, Spain) and tap water until the day before the experiment. These studies were carried out in compliance with both the Laboratory Animal Care Guidelines of the European Union (86/609/CEE) and the U.S. NIH guidelines.

Presurgical Procedure

On the day before the experiment, rats were subjected to an 18-hour fast. The free access to water was maintained. At 8 am of the day of the experiment, animals were allowed access to food during a period of 1 hour to ensure that they had eaten by the time the experiment started. Afterwards, animals were anesthetized using sodium pentobarbital (65 mg/kg intraperitoneal).

Surgical Procedure

During the experiment, anesthesia was maintained by continuous IV infusion of sodium pentobarbital [0.5 mL/h per 100 grams body weight (bw)], using a B. Braun Perfusor automatic pump (Lisboa, Portugal) and through a polyethylene PE50 tubing inserted in the venous side of the arterial-venous shunt (see description below). Body temperature was maintained at 37.0 ± 0.5 °C by means of a Homeothermic Blanket Control Unit (Harvard Apparatus Inc., Holliston, MA). Surgical protocol was as previously described (14). Tracheotomy was performed to allow spontaneous respiration throughout the experiment. The right femoral artery and internal jugular vein were cannulated and the arterial-venous shunt inserted. The arterial-venous shunt is an external shunt, connecting the femoral artery to the internal jugular vein and so allowing blood circulation through a connecting silicon sleeve. The arterial-venous shunt was primed with a 200 IU/mL sodium heparin solution. Multiple arterial blood samples were collected by puncture into the arterial side of the shunt. Drugs and anesthetic infusions were made through the venous side. Because this shunt was connected to a pressure transducer, blood pressure was also monitored; mean arterial pressure was obtained by clamping the venous side of the loop. In the animals receiving l-NMMA, a laparotomy was performed; after, the portal vein was cannulated by insertion of an Abbocath 24G Optiva IV (Johnson & Johnson, Pomezia, Italy) catheter, connected to PE50 tubing, through which drugs could be infused directly into the liver. A 30-minute period (minimum) was allowed for stabilization.

Assessing Insulin Sensitivity (IS): Rapid IS Test (RIST)

At 5-minute intervals, 25 μl of arterial blood samples were collected with a microsyringe (YSI Model 1501 Syringepet; YSI Inc., Yellow Springs, OH), and glycemia was immediately determined through the enzymatic glucose oxidase method, using a glucose analyzer, YSI Model 1500 Sport. Three stable glycemic values were obtained, and the average of these values is referred to as glycemia baseline. RIST was then started. The RIST technique, performed as described by Lautt et al. (14), is a euglycemic clamp, which starts with a 5-minute insulin infusion (50 mU/kg IV; t = 0 minutes). To avoid hypoglycemic insulin effects, 1 minute after beginning the insulin infusion (t = 1 minute), d-glucose infusion (100 mg/mL IV) was also started at a rate of 5 mg/kg per minute. Every 2 minutes, starting at t = 1 minute, arterial blood samples were collected (25 μL), and glycemia was determined. The euglycemia was maintained at the baseline level throughout the assay by adjusting the glucose infusion rate, according to the last arterial glycemia determined, using a variable infusion pump (B. Braun Perfusor). The RIST is considered to be finished when no further glucose infusion is needed to maintain baseline glycemia. The total amount of glucose infused during the RIST is referred to as the RIST index (milligrams of glucose per kilogram bw), and it is a measure of IS. After the end of the RIST, rats were allowed at least 30 minutes of rest for stabilization before another RIST could be performed. The RIST is very useful for assessing IS because it presents a high reproducibility and because it is possible to perform up to four consecutive control RISTs (14). After an initial (control) RIST, drug administrations were done, after which a new glucose baseline was determined, and a postdrug RIST was performed.

Experimental Protocols

RIST Reproducibility Controls

Basal arterial glycemia was determined, and a control RIST was performed. Sixty minutes afterwards, a second control RIST was done (OZR, n = 5; LZR, n = 4). Before each RIST, blood samples were collected for plasma insulin determination (Rat Insulin RIA Kit; Linco Research, St. Louis, MO).

Blockade of the Parasympathetic Nerves

Initial arterial glycemia baseline was determined, and a control RIST was performed. Atropine (3 mg/kg, 5 minutes IV) was infused, and a 30-minute period was given for animal stabilization. A new glycemia baseline was determined, and another RIST (post-atropine RIST) was performed (OZR, n = 5; LZR, n = 6).

Inhibition of the Hepatic NO Synthase

After a control RIST was performed, l-NMMA (0.73 mg/kg, 0.4 mL bolus IPV) was administered. Maximal metabolic effect of l-NMMA was achieved after 1 hour (Afonso RA and Macedo MP, unpublished data), and a subsequent RIST (post l-NMMA RIST) was performed (OZR, n = 6; LZR, n = 6).

Processing and Analysis of Samples

Blockade of the Parasympathetic Nerves

Characterization of the hepatic-dependent component of insulin action consisted of quantification of the hepatic-dependent component. The post-atropine RIST index represents the hepatic-independent component of insulin action (after HPN blockade). By subtracting the post-atropine RIST index from the control RIST, the hepatic-dependent component of insulin action can be quantified. Characterization of the hepatic-dependent component of insulin action also consisted of a dynamic profile of the hepatic-dependent component. By plotting the glucose infusion rate (milligrams of glucose per kilogram per minute) vs. time (minutes), the dynamic profile of the RIST is obtained. By subtracting post-atropine RIST from control RIST curves, the hepatic-dependent component's dynamic profile can also be determined.

Inhibition of the Hepatic NO Synthase

The difference between control RIST index and the post-l-NMMA RIST index quantifies the hepatic NO (hepatic pathway) contribution to insulin action.

Statistical Analysis

Data expressed as means ± standard error. The significance of the differences was calculated through two-tailed Student's t tests. At p < 0.05, differences were accepted as statistically significant. The curves representing the dynamic profile of the RIST as time functions were obtained from glucose infusion rates in 0.1-minute intervals.


Sodium pentobarbital was supplied by Eutasil-Sanofi Veterinária (Lisboa, Portugal); sodium heparin and saline by B. Braun; insulin (regular Humulin) by Lilly (Lisboa, Portugal); and d-glucose, atropine (atropine sulfate), and l-NMMA by Sigma-Aldrich Chemical (Madrid, Spain). All solutions were prepared in saline.


The obese rats (OZR) were significantly heavier (336.9 ± 15.0 grams, n = 16) than their controls, LZR (287.0 ± 9.1 grams, n = 16; p < 0.01), with a mean bw difference of 14.8% between the two groups. Resting arterial blood pressure was similar in LZR and OZR (104.2 ± 2.4 and 109.5 ± 4.2 mm Hg, respectively). Neither atropine (3 mg/kg IV) nor l-NMMA (0.73 mg/kg IPV), at the time of the RIST, induced significant changes in the arterial blood pressure of both animal groups (LZR, 101.7 ± 3.7 mm Hg; OZR, 113.8 ± 1 mm Hg) at the given doses. There was no significant difference in the arterial glucose levels between LZR (109.6 ± 1.7 mg/dL) and OZR (113.0 ± 1.8 mg/dL) during the course of the experiment.

In the control experiments, the first and second control RIST indices were similar, both in OZR (70.8 ± 4.9 vs. 62.0 ± 5.9 mg glucose/kg bw; n = 5; not significant) and LZR (279.0 ± 45.5 vs. 261.9 ± 41.6; n = 4; not significant), showing that the initial RIST did not affect IS. The RIST did not affect plasma insulin levels either (OZR, 18.6 ± 2.0 vs. 14.9 ± 1.7 ng/mL; LZR, 1.1 ± 0.1 vs. 1.0 ± 0.1 ng/mL), although they were significantly higher in OZR (p < 0.001).

Blockade of the Parasympathetic Nerves

OZR showed a lower control RIST index than the lean animals (OZR, 73.7 ± 14.2 mg glucose/kg bw, n = 5; LZR, 289.2 ± 24.7 mg glucose/kg bw, n = 6; p < 0.001). The hepatic-independent component of insulin action (i.e., the post-atropine RIST index) was lower in OZR (39.3 ± 3.5 mg glucose/kg bw, n = 5) than in LZR (173.3 ± 20.5 mg glucose/kg bw, n = 6; p < 0.001). Similarly, the hepatic-dependent component, obtained by subtraction of the post-atropine RIST index from the control RIST index, was significantly lower in OZR (34.4 ± 12.8 mg glucose/kg bw, n = 5) than in LZR (115.9 ± 19.4 mg glucose/kg bw, n = 6; p < 0.01). These results are summarized in Figure 1A.

Figure 1.

Effect of atropine on IS (hepatic pathway inhibition). (A) RIST indices for the HPN-independent components (black bars, LZR, 173.3 ± 20.5 mg glucose/kg bw, n = 6; OZR, 39.3 ± 3.5 mg glucose/kg bw, n = 5) and HPN-dependent components (white bars, LZR, 115.9 ± 19.4 mg glucose/kg bw; OZR, 34.4 ± 12.8 mg glucose/kg bw) of insulin action. The sum of both components represents the control RIST index for each strain. (B) Relative contribution of hepatic (HPN)-dependent (in white) and HPN-independent (in black) components to total insulin action in LZR (hepatic-dependent, 40.1 ± 6.5%; hepatic-independent, 59.9 ± 6.5%; n = 6) and OZR (hepatic-dependent, 41.3 ± 8.0%; hepatic-independent, 58.8 ± 8.0%; n = 5). Values are means ± standard error. (**) p < 0.01. (***) p < 0.001.

Interestingly, the relative contribution of the hepatic-dependent component to total insulin action, given by the percentage of inhibition achieved after atropine administration (Figure 1B), was similar between OZR (41.3 ± 8.0%) and LZR (40.1 ± 6.5%). The hepatic-independent contribution was also identical between the two groups (OZR, 58.8 ± 8.0%; LZR, 59.9 ± 6.5%).

The mean profile curves, obtained in the basal state (control RIST), after atropine administration (post-atropine RIST), and for the hepatic-dependent component, are shown in Figure 2. All parameters (peak, peak time, and duration) for the control and post-atropine RIST profiles are diminished in the OZR when compared with LZR (p < 0.001). The duration of the hepatic-dependent component action was significantly shorter in OZR than in LZR (p < 0.01) due to an earlier offset in OZR (p < 0.01) because there is no significant difference in the onset time between the two groups. Table 1 summarizes these parameters.

Figure 2.

Dynamic profiles of RISTs for LZRs (n = 6, A and B) and OZRs (n = 5, C and D). (A) Control (bold line) and postatropine (hepatic-independent component, regular line) RIST profiles of LZR. (B) Dynamic profile of the hepatic-dependent component of insulin action in LZR. (C) Control (bold line) and post-atropine (regular line) RIST profiles of OZR. (D) Dynamic profile of the hepatic-dependent component of insulin action in OZR.

Table 1.  Parameters of the dynamic profiles of control RIST, post-atropine RIST, and hepatic-dependent action
 Lean ratsObese rats
  • *

    p< 0.001;

  • p < 0.01 (obese vs. lean rats).

Control RIST  
 Peak (mg glucose/kg per minute)14.4 ± 1.65.4 ± 0.5*
 Peak time (min)15.2 ± 0.87.2 ± 1.1*
 Duration (min)39.5 ± 1.921.7 ± 2.6*
Post-atropine RIST  
 Peak (mg glucose/kg per minute)9.6 ± 1.34.1 ± 0.3*
 Peak time (min)12.7 ± 1.87.0 ± 1.0*
 Duration (min)35.2 ± 2.916.2 ± 1.7*
Hepatic-dependent component  
 Onset (min)6.4 ± 1.37.7 ± 2.0
 Peak (mg glucose/kg per minute)6.0 ± 1.13.4 ± 0.9
 Peak time (min)16.0 ± 2.111.9 ± 2.4
 Offset (min)33.8 ± 3.321.4 ± 2.6
 Duration (min)27.4 ± 4.013.8 ± 1.8

Inhibition of the Hepatic NO Synthase

The data obtained for the l-NMMA protocol were identical to that obtained in the atropine protocol. Control RIST index in LZR rats (287.4 ± 22.7 mg glucose/kg bw; n = 6) was higher than in OZR (79.3 ± 1.6 mg glucose/kg bw; n = 6; p < 0.001). After l-NMMA administration, i.e., after inhibiting hepatic NO synthesis, there was a significant decrease in the RIST index, both in LZR (156.1 ± 19.2 mg glucose/kg bw, n = 6; p < 0.001) and OZR (41.3 ± 5.6 mg glucose/kg bw, n = 6; p < 0.001). Similarly, the hepatic pathway-dependent component of total insulin action, calculated by subtraction of the post-l-NMMA RIST index from the control RIST index, was higher in LZR (131.2 ± 10.8 mg glucose/kg bw) than in OZR (38.0 ± 5.0 mg glucose/kg bw; p < 0.001). These results are presented in Figure 3A.

Figure 3.

Effect of hepatic NO synthase inhibition on IS. (A) Control RIST indices (LZR, 287.4 ± 22.7 mg glucose/kg bw, n = 6; OZR, 79.3 ± 1.6 mg glucose/kg bw, n = 6) refer to the total insulin action and can be divided into two components: the hepatic (NO)-dependent component (white bar, LZR, 131.2 ± 10.8 mg glucose/kg bw; OZR, 38.0 ± 5.0 mg glucose/kg bw) and the hepatic-independent component, obtained after l-NMMA administration (black bars, LZR, 156.1 ± 19.2 mg glucose/kg bw; OZR, 41.3 ± 5.6 mg glucose/kg bw). (B) Partial contribution to total insulin action of the hepatic-pathway (white bars) in LZR (46.4 ± 4.1%, n = 6) and OZR (48.3 ± 6.6%, n = 6). The black bars represent the partial contribution of the hepatic-independent component (LZR, 53.6 ± 4.1%; OZR, 51.7 ± 6.6%). Values are means ± standard error. (***) p < 0.001 (OZR vs. LZR).

The proportion of IS inhibition, induced by l-NMMA administration, calculated from the post-l-NMMA RIST indices and representative of the hepatic-dependent component, was similar between both strains: 46.4 ± 4.1% (LZR) and 48.3 ± 6.6% (OZR). The hepatic-independent component contribution was also identical in LZR (53.6 ± 4.1%) and OZR (51.7 ± 6.6%) (Figure 3B).


Our aim was to characterize the hepatic-dependent and -independent components of whole-body insulin action in an obesity animal model, the Zucker rat (OZR). The results reported herein suggest that the IR observed in the OZR is due to the impairment of both of these components, which are decreased exactly to the same extent. The metabolic defect responsible for the severe IR observed in the OZR is located downstream from HPN stimulation and hepatic NO synthesis and seems to be common to both components of insulin action.

Methodological Considerations

To assess the IS, we used the modified euglycemic clamp RIST because it is the test that better fits within the experimental protocol design for this study. It avoids induction of hypoglycemia and does not alter the circulating catecholamines or glucagon levels (14). Moreover, both insulinemia and glycemia always return to basal levels after each RIST (14). The RIST technique can be reproduced several times in each experiment (up to four consecutive RISTs), without causing significant interference with the animal's physiological conditions, while simultaneously maintaining the high sensitivity of the test (14). Furthermore, our results show that the RIST is a reproducible tool for measuring IS also in the Zucker rat model because there was no change between the first and second control RISTs (control experiments). Results obtained using the RIST are similar to those obtained using the insulin tolerance test, although somewhat different from those obtained with the hyperinsulinemic euglycemic clamp (15). This may be related to the vagal impairment induced by the hyperinsulinemic euglycemic clamp technique (16), making it hard to detect the hepatic-dependent component of insulin action, which requires hepatic parasympathetic stimulation. Most methods used to assess IS are based on the steady-state glucose and insulin concentrations achieved after an overnight fast. However, the hepatic-dependent component of peripheral insulin action is maximal in the postprandial state. By refeeding the animals after an overnight fast, as it was done, we ensured that hepatic-dependent insulin action was maximal at the time of the test.

IR in the Obese Rats (OZR)

Our results show that total peripheral IS is significantly diminished in OZR when compared with their lean litter mates (LZR). OZRs present a 70% to 75% deficit of IS compared with the LZR, which is in accordance with reports of marked IR in obese subjects, both rats and humans (12, 13, 17, 18, 19, 20). In the fasted state, the glucose disposal is due only to the insulin action per se (hepatic-independent component) because the hepatic pathway contribution for the overall insulin action is absent in this state. Therefore, because most studies concerning the assessment of the OZR IS in vivo were performed in the fasted state (17), they only show the impairment of the hepatic-independent component, which we also confirmed in the present report.

The skeletal muscle still represents the majority of glucose uptake in OZR, although the weight ratio of fat to skeletal muscle is increased in these animals (13), and because this tissue is primarily affected by IR in the OZR model (13, 18), abnormalities at this level will reflect in the overall insulin-stimulated glucose disposal. The deficiency in insulin action in OZR is also supported by several in vitro studies, in which multiple defects in the insulin-stimulated glucose uptake in the skeletal muscle of these animals have been described and can lead to the observed hepatic-independent IR. These reports include the impairment of insulin responsiveness due to a decreased binding to the plasma membrane (21, 22) and to postreceptor abnormalities. These include excessive serine/threonine phosphorylation of the insulin receptor (23), leading to interruption of the insulin signaling, or the impairment of phosphatidylinositol 3 kinase activity (24), which is associated with reduced glucose transporter (GLUT) 4 translocation (18, 25, 26, 27, 28).

As we mentioned above, the normal peripheral IS depends also on the action of the hepatic pathway that starts with the meal-induced activation of the HPNs, followed by stimulation of the hepatic M1 muscarinic receptors and NO synthesis. These events lead to the release of HISS, the liver-derived factor that acts selectively in the skeletal muscle, enhancing insulin action. The impairment of the hepatic pathway results in the decrease of total insulin action, as previously observed in several other animal models of IR, such as the spontaneously hypertensive rat, sucrose-fed rats, aging rats, adult offspring of fetal alcohol exposure, and liver disease induced by chronic bile duct ligation (6, 29, 30). In all these models, the impairment of the total IS was due only to a deficient hepatic-dependent component, whereas the hepatic-independent component remained unchanged (6, 29, 30). On the contrary, the OZR present an impairment of the hepatic-dependent component, which is accompanied by a proportional decrease of the hepatic-independent component of insulin action, as we report herein.

The parasympathetic dysfunction is usually associated with the obesity condition (9) and could explain the observed hepatic-dependent IR. However, although such parasympathetic dysfunction could contribute to the impairment of the hepatic-dependent insulin action in the obese animals, it fails to explain why the hepatic-independent component is also affected. Some authors have suggested that a decrease of the parasympathetic tone can lead to the impairment of insulin secretion (31), but this may be a controversial issue because the atropine-induced decrease in insulin secretion seems to be attenuated in obese subjects (31). Moreover, OZR have been described as hyperinsulinemic (13). On the other hand, even when what is evaluated is the insulin action through the administration of exogenous insulin, we observed an impaired glucose disposal, suggesting that the major defect in OZR does not seem to concern the insulin production, but rather the insulin peripheral action.

Hepatic NO synthesis is also an important intermediate step of the pathway initiated with the HPN stimulation and that culminates with the synthesis of HISS (hepatic-dependent pathway). However, impairment of the hepatic NO synthase activity is unlikely to account for the observed hepatic-dependent IR. Although NO synthase activity is known to be significantly altered in several tissues in the OZR, such as the gastric fundus, central nervous system, and skeletal muscle (10, 19), there are no reports of NO synthase activity deficiency in the liver of the OZR. Even if we extrapolate these studies and consider that the hepatic NO synthase activity is also decreased in the OZR, such impairment would affect only the hepatic-dependent component and not the hepatic-independent component, which we observed to be equally impaired.

The similar impairment of both hepatic-dependent and -independent components of insulin action in OZR suggests that the deficiency of insulin action lies downstream from insulin secretion, HPN stimulation, and NO synthesis, probably at a metabolic site common to hepatic-dependent and -independent components. Alternatively, insulin action may be impaired both in the ability to stimulate glucose uptake and to stimulate HISS release.

Impairment of Both Components of Insulin Action in the OZR: the Same Metabolic Deficiency

The OZR is an animal model characterized by deficiencies at the leptin receptor (11, 12, 13). Under physiological conditions, leptin inhibits insulin secretion (32) and stimulates endothelial NO release in what seems to be a compensatory mechanism to counteract the pressor effect of sympathoexcitation induced by leptin (11). In obesity, however, the leptin action deficiency results in decreased production of NO, which could only explain the impairment of the hepatic-dependent insulin sensitization observed, as discussed before. On the other hand, leptin metabolism deficiencies cannot account for the decrease in the hepatic-independent component because the major effect of a deficient leptin action, as it occurs in OZR, is the over-production of insulin, which may be responsible for the observed hyperinsulinemia. Furthermore, others show that leptin has no direct effect on glucose uptake in rat skeletal muscle (33). Therefore, although the role of leptin in obesity and IR is worthy of study, deficiencies at the level of its metabolism by themselves do not explain the observed IR (hepatic-dependent plus -independent) in OZR.

As mentioned, the skeletal muscle is the tissue primarily affected by IR in OZR. Indeed, deficiencies concerning both the insulin receptor, which leads to a deficient binding of insulin, and the GLUT4 translocation have been reported in the skeletal muscle of OZR (18, 21, 23, 25, 26, 27, 28). These deficiencies may affect not only the insulin action per se (hepatic-independent component) but also the hepatic-dependent component because this pathway enhances the glucose uptake, mainly in the skeletal muscle. Hypothesizing that HISS may exert its action primarily by interacting with the insulin receptor in the cell membrane, resulting in the initiation of the signal transduction pathway or in its own receptor requiring GLUT4 translocation to the cell surface to promote glucose uptake, seem to be two essential steps of the insulin intracellular signaling cascade where the confluence between the two pathways can occur. Therefore, because insulin binding and GLUT4 translocation are known to be impaired in OZR, neither insulin nor HISS are capable of inducing proper glucose uptake; thus, the impairment of the hepatic-dependent and -independent action is similar because it results from the same defect. Further studies are required to elucidate these questions.

In conclusion, our results confirm that the obesity animal model OZR is considerably insulin resistant in comparison with its lean litter mates (LZR). Furthermore, we show for the first time that in the OZR, both the hepatic-dependent and -independent components of insulin action are decreased. The fact that both components are decreased to the same extent suggests a defect in a pathway common to both components, giving us an indication regarding the hepatic-dependent component and its role in the control of glucose uptake in the skeletal muscle that should be pursued through further studies.


We acknowledge the intellectual support provided by Wayne Lautt, Graça Morais, Mota-Carmo, and Maria Guarino. From the animal care facilities, we acknowledge Ana Santos and Mariana. This work was supported by the Foundation for Science and Technology (Grants POCTI/SAU/14009/1998 and POCTI/NSE/42397/2001), by the Portuguese Diabetes Association, by the Portuguese Endocrinology, Diabetes, and Metabolism Society (SPEDM) grant for the study of obesity, and by Foundation for Science and Technology Ph.D. Fellowship SFRH/BD/9082/2002 (to R.A.).


  • 1

    Non-standard abbreviations: HPN, hepatic parasympathetic nerve; IPV, intraportal; NO, nitric oxide; HISS, hepatic insulin-sensitizing substance; IV, intravenous(ly); l-NMMA, N-monomethyl-L-arginine; IR, insulin resistance; OZR, obese Zucker rat; LZR, Lean Zucker rat; bw, body weight; IS, insulin sensitivity; RIST, rapid IS test; GLUT, glucose transporter.

  • The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.