The interrelationship between insulin and leptin resistance in young Fischer 344 (F344) rats was studied. Young F344 and Sprague-Dawley (SD) rats were fed regular chow. F344 animals had two- to threefold higher insulin and triglyceride concentrations and increased stores of triglycerides within liver and muscle. F344 animals gained more body fat. Both acyl-CoA oxidase (ACO) and carnitine palmitoyltransferase I gene expression were 20–50% less in F344 animals than in age-matched SD animals. Peroxisome proliferator-activated receptor-α gene expression was reduced in 70-day-old F344 animals. Finally, resistin gene expression was similar in 70-day-old SD and F344 animals. Resistin gene expression increased fivefold in F344 animals and twofold in SD animals from 70 to 130 days, without a change in insulin sensitivity. We conclude that young F344 animals have both insulin and leptin resistance, which may lead to diminished fatty oxidation and accumulation of triglycerides in insulin-sensitive target tissues. We did not detect a role for resistin in the etiology of insulin resistance in F344 animals.
- Fischer 344 rats
- acyl-coenzyme A oxidase
- carnitine palmitoyltransferase I
- peroxisome proliferator-activated receptor-α
insulin and leptin, intimately involved with intermediary metabolism and body weight homeostasis, appear to be physiologically linked. Insulin and leptin are secreted in the peripheral circulation by pancreatic islets and adipocytes, respectively, and they cross the blood-brain barrier (19). Both hormones have receptors within appetite centers in the hypothalamus and inhibit feeding (22). Finally, the secretion of both hormones appears to be regulated by glucose and the generation of intracellular energy (14). A close relationship also exists between insulin and leptin resistance. Nearly all individuals with type 2 diabetes are markedly insulin resistant, and the majority of them are obese and leptin resistant (2,17). Most animal models of insulin resistance and type 2 diabetes mellitus have defects either in leptin secretion (e.g., as in the ob/ob mouse) or in leptin sensitivity (e.g., as in the db/db mouse, Zuckerfa/fa rat, diet-induced obesity; see Ref.10). Morbid obesity and glucose intolerance are noted early in the life span of the animals that display these monogenetic mutations; therefore, they do not adequately represent the majority of human type 2 diabetes subjects in whom these diseases develop later in life.
The Fischer 344 (F344) rat has been used as a model of aging-induced insulin resistance (3). This animal strain was believed to become insulin resistant with aging, but it did not appear to gain appreciable amounts of weight or of body fat. However, recent studies in our laboratory disclosed that F344 animals gained more body fat than did a commonly studied insulin-sensitive animal strain [Sprague-Dawley (SD); see Ref. 15]. Not only did F344 rats ingest more calories per gram body weight but also they stored more calories as weight and they converted ingested calories into fat more efficiently than did the SD rats. The propensity of the F344 rats to gain body fat occurred at young ages (70 days), and it occurred despite higher fasting and meal-induced concentrations of leptin. These studies suggested to us that the F344 animals were resistant to the two main biological responses to leptin, namely the inhibition of appetite and the enhancement of energy expenditure.
Therefore, it appears that the F344 strain may be a good model of human diabetes mellitus. These F344 animals have leptin resistance without developing morbid obesity, and they develop diabetes late in their life span. The purpose of the present study was to define more rigorously the insulin sensitivity of young Fischer rats and to study the interrelationship between insulin resistance and leptin resistance. Several studies have suggested that physiological and cellular mechanisms that lead to the accumulation of lipids in insulin-sensitive target tissues (16, 18, 5) and in the pancreatic islets (12) may produce a clinical syndrome of insulin resistance and declining insulin secretory capacity; this syndrome is typical of type 2 diabetes mellitus. Leptin reduces intracellular accumulation of lipids in liver, muscle, and in the pancreatic islet (1, 8, 13,20). Therefore, leptin resistance may exaggerate the lipid accumulation in adipose and nonadipose tissue and worsen insulin resistance.
We have measured indexes of lipid accumulation within insulin-sensitive target tissues and the gene expression of factors and enzymes involved with mitochondrial and peroxisomal fatty acid oxidation in the leptin-resistant F344 strain. Furthermore, we tested the hypothesis that the newly discovered resistin gene links obesity and insulin resistance (21). We have found that the F344 animals are both leptin and insulin resistant. We have also found that this strain accumulates lipid excessively in liver and muscle and that the gene expression of factors and enzymes involved with fatty acid oxidation are reduced. However, we did not find that resistin gene expression played a role in the insulin resistance observed in F344 animals.
Animals and diets.
All animals were humanely treated, and the experimental protocols were reviewed and accepted by the Institutional Animal Care and Use Committee at Virginia Commonwealth University. Sixty-day-old SD and F344 rats were purchased from Harlan Sprague-Dawley (Indianapolis, IN) and housed in individual cages under controlled lighting conditions on a natural dark-light cycle (1800–0600). The animals were allowed free access to food and water. All animals were fed a regular chow diet (Purina).
Serum and tissue measurements.
Seventy- and 130-day-old SD (n = 10 at each age) and F344 (n = 10 at each age) animals were fasted overnight. One-half of the animals was killed (fasted), and the other one-half of the animals was fed a regular powdered chow (total 8 g) for 3 h and then killed (refed). Therefore, there were five animals in four experimental groups (70-day-old SD and F344 animals, fasted; 130-day-old SD and F344 animals, refed). Invariably, the animals ingested the entire 8 g of pelleted chow that were supplied. All animals were anaesthetized with isoflurane before they were decapitated. Blood was collected and saved on ice. After being spun in a centrifuge, the serum was isolated and saved in a −80°C freezer. Liver, abdominal muscle, and the epididymal fat pad were quickly dissected and weighed, and then they were frozen in liquid nitrogen and stored in a −80°C freezer.
Serum glucose was measured by an automated colorimetric glucose oxidase system (Vitros 700 system; Johnson and Johnson). Serum leptin and insulin were measured by RIA kits (Linco Research, St. Charles, MO). The limit of sensitivity and linearity for the rat leptin assay was 0.5 and 50 ng/ml, respectively. The intra-assay variation for the leptin assay at 1.6 ng/ml (Quality Control 1) was <2%. Tissue triglycerides and nonesterified fatty acids in serum were measured by kits described by the supplier (Wako Chemicals, Richmond, VA). In brief, 100 mg of liver or muscle were homogenized by a saw-tooth generator (OMNI TH) in 2 ml of homogenizing buffer [18 mM Tris · HCl (pH 7.5), 300 mMd-mannitol, and 5 mM EGTA]. The solution was centrifuged (3,500 rpm, 20 min) to remove debris. Isopropyl alcohol (5 ml) was added to 0.2 ml of supernatant (experimental sample) or standard solutions (0–2,000 mg/dl). After vigorous shaking for 10 min, the solutions were centrifuged (3,000 rpm, 10 min), and 3.0 ml of color reagent solution of triglyceride E (model no. 432–40201; Wako) were added to 0.3 ml of supernatant. The absorbance at 600 nm was measured by a spectrophotometer.
Intravenous glucose tolerance test.
Seventy- and 130-day old SD and F344 animals were anesthetized with isoflurane, and a catheter was inserted in the external jugular vein, as previously described (11). The infusion cannulas were tunneled subcutaneously to the back of the neck. The catheter was infused with heparin to prevent clot formation at the intravenous tip of the catheter, and the external end of the catheter was plugged. The animals were allowed to recover for 2 days. On the 3rd day, the animals were fasted overnight. The next morning, 0.5 ml of blood (for fasting glucose and insulin levels) was withdrawn through the catheter, and then 0.55 g/kg of glucose was infused through the catheter over a 1-min period, as previously described (4). After 5, 10, 15, 20, and 30 min, 0.5 ml of whole blood was again withdrawn through the catheter. Serum glucose and insulin were measured at the indicated times.
Gene expression by quantitative RT-PCR.
RNA was extracted from liver, muscle, and adipose tissue by the TRIzol Reagent method, as described by the suppliers (GIBCO-BRL, Grand Island, NY). cDNA synthesis was performed with an oligo(dT) primer and Superscript reverse transcriptase, as described by the supplier (GIBCO-BRL). Duplex PCR was performed in a solution with primer pairs from a gene of interest and a housekeeping gene. The sense and anti-sense primers for each gene are as follows: peroxisome proliferator-activated receptor-α (PPARα), sense 5′-AAGCCATCTTCACGATGCTG-3′ and anti-sense 5′-TCAGAGGTCCCTGAACAGTG-3′ (25); acyl-CoA oxidase (ACO), sense 5′-GCCCTCAGCTATGGTATTAC-3′ and anti-sense 5′-AGGAACTGCTCTCACAATGC-3′ (25); carnitine palmitoyltransferase I (CPT I), sense 5′-TATGTGAGGATGCTGCTTCC-3′ and anti-sense 5′-CTCGGAGAGCTAAGCTTGTC-3′ (25); resistin, sense 5′-GGGAGTTGTGCCCTGCT-3′ and anti-sense 5′-CAGCACTCGGAGGGCAA-3′ (9); enolase, sense 5′-TTCTCAAGATCCATGCCAGG-3′ and anti-sense 5′-GCGTTCGCACCAAACTTAGA-3′; β-actin, sense 5′-GTGACGAGGCCCAGAGCAAGAG-3′ and anti-sense 5′-AGGGGCCGGACTCATCGTACTC-3′.
PCR conditions were carefully chosen to optimize amplification of the experimental and housekeeping cDNA and to limit the amplification of nonspecific cDNA. As shown in Fig. 1, amplification of each duplex cDNA was relatively linear within a range of cycle numbers. The range of linearity depended on the gene of interest and the conditions of PCR. When comparing gene expression between experimental groups, cycle number was chosen so that the amplification of the cDNA pair was well within the linear portion of the cycle number-quantity of DNA curve. The conditions of each duplex PCR are as follows: PPARα-enolase, 33 cycles of melt at 92°C for 45 s, anneal at 57°C for 45 s, extend at 72°C for 60 s; ACO-actin, 22 cycles of melt 92°C 45 s, anneal at 55.5°C for 45 s, and extend at 72°C for 60 s; CPT I-actin, 24 cycles of melt at 92°C for 45 s, anneal at 55.5°C for 45 s, and extend at 72°C for 60 s; resistin-actin, 24 cycles of melt 92°C for 45 s, anneal at 54°C for 45 s, and extend at 72°C for 60 s.
When we compared the gene expression between experimental groups, we carefully prepared the master mixes. All PCR solutions were pipetted at the same time, were amplified in the same run in the thermocycler (MJ Research thermocycler), and were analyzed on the same agarose gel. Solutions for PCR differed only by the addition of various cDNAs.
Forty percent of the PCR reaction (total volume = 50 μl) was analyzed on a 1.5% agarose gel with ethidium bromide. The gel was placed in a Luminescent Image Analyzer (LAS-1000; Fujifilm), and the densities (light arbitrary units/mm2) of the amplified DNA were assessed by the Advanced Image Data Analyzer software (AIDA version 2.0). The abundance of the expressed gene was calculated by dividing the density of the amplified experimental gene by the density of the amplified housekeeping gene in the same sample.
We performed a two-way ANOVA (GraphPad Prism Version 3.0; GraphPad Software) to compare the effects of the fed state (fasting vs. refeeding) or strain (SD vs. F344) on various indexes of weight, body composition, and insulin sensitivity. We performed an unpairedt-test (GraphPad Prism version 3.0) when an experimental parameter in one strain of animal was compared with the other strain of animal.
Seventy-day-old SD and F344 animals were fed regular rat chow ad libitum for 60 days. Indexes of weight and body composition are shown before (Table 1) and after (Table2) the 60-day feeding period. In one-half of the animals, measurements were obtained in animals fasted overnight (fast). In the other one-half of the animals, measurements were obtained in animals that were fasted overnight and then refed regular chow for 3 h (refed). Age-matched SD rats are larger than F344 rats, and both strains gain ∼1.5-fold their body weight in 60 days when eating regular chow. However, the F344 animals are fatter than the SD animals. At 60 days, the epididymal fat pad is ∼15% larger in F344 animals compared with SD animals (Table 1). In the 60-day feeding period, the fat pad weight in SD animals increased by ∼1.9-fold, whereas the fat pad weight in F344 animals increased between 2.2- and 2.7-fold (Table 2). As previously demonstrated in F344 animals by our laboratory, epididymal fat pad weight correlates well with total body adiposity, as measured by dual-energy X-ray absorptiometry scanning (15). Fat storage in nonadipocyte tissue was also higher in F344 animals than in SD animals. Liver triglyceride content was higher in 70-day-old F344 animals than in age-matched SD animals (Table 1); after 60 days, the muscle and liver triglyceride content increased fractionally more in F344 animals than in SD animals. Muscle and liver triglyceride content were both higher in 130-day-old F344 animals than in SD animals (Table 2).
Also shown in Tables 1 and 2 are indexes of insulin and leptin sensitivity. Serum glucose concentrations were appropriately higher in refed animals compared with fasted animals. However, fasting glucose concentrations were minimally higher in 70-day-old F344 animals than in SD animals (Table 1) but were no different in 130-day-old animals (Table 2). However, the differences in insulin concentration were marked. Insulin concentrations in both fasted and refed F344 animals were approximately two- to threefold higher than in fasted and refed SD animals in both 70 (Table 1)- and 130 (Table 2)-day-old animals. Fasting serum triglyceride concentrations were higher in 70-day-old (Table 1) and 130-day-old (Table 2) F344 animals than in SD animals. In contrast to SD animals, the triglyceride concentration in F344 animals increased markedly with refeeding (Tables 1 and 2). In addition, fasting triglycerides in F344 animals increased over the 60-day experiment, whereas fasting triglycerides in SD animals did not change (compare Tables 1 and 2). Fasting free fatty acids were minimally elevated in 70-day-old SD animals compared with age-matched F344 animals (Table 1). There was no detectable difference in fasting free fatty acid between 130-day-old SD and F344 animals (Table 2). Serum free fatty acid decreased postprandially in both strains because of insulin-mediated inhibition of lipolysis. The serum leptin concentrations were not detectable in the young SD animals, which had small fat stores. This finding is consistent with our previous in vivo data (15). Serum leptin concentrations in 70-day-old F344 animals were measurable, and they increased by ∼100% with refeeding (Table 1). After 60 days, the leptin concentration in F344 animals continued to a rise in fasted animals, and the concentration was responsive to refeeding. Note that, despite the high concentrations of this appetite-suppressing hormone, F344 animals still gained more triglyceride stores in both adipose and in nonadipose tissue than in SD animals. This observation is consistent with leptin resistance in the F344 animals.
To better characterize insulin sensitivity, intravenous glucose tolerance tests were performed in 70- and 130-day-old animals from both strains. The results are shown in Fig. 2. Fasting and peak glucose concentrations were similar in 70-day-old SD and F344 animals (Fig. 2, top left). Glucose concentrations at 15 and 20 min were greater in 70-day-old SD than in age-matched F344 animals (Fig. 2, top left). The area under the curve (AUC) for glucose was ∼50% higher (P < 0.05) in SD animals than in F344 animals (3,117.2 ± 94 vs. 2,124.5 ± 342 mg/dl, respectively). The insulin response to an intravenous bolus of glucose was markedly different. Fasting insulin was approximately threefold higher in the 70-day-old F344 rats than in age-matched SD rats (Fig. 2, bottom left). Peak insulin concentrations were approximately fivefold higher in the F344 than in the SD animals, and the insulin AUC was ∼2.7-fold higher (P < 0.001) in the F344 animals than in the SD animals (66.8 ± 3.4 vs. 24.5 ± 1.9 ng/ml, respectively). The ratio of the insulin AUC to glucose AUC in 70-day-old F344 animals vs. SD animals was 33 ± 4 vs. 7.7 ± 0.54 × 10−3 (P < 0.001). Therefore, to produce similar serum glucose concentrations after an intravenous glucose infusion, F344 animals require markedly higher insulin concentrations than do SD animals. This finding is consistent with insulin resistance in the F344 animals, a state that has not been recognized previously in these young animals. There were no statistical differences in the glucose (Fig. 2, top right) and insulin (Fig. 2, bottom right) responses to intravenous glucose in 130-day old animals compared with 70-day-old animals in both strains.
A mechanistic link between the leptin and insulin resistance observed in F344 animals may involve dysregulation of enzymes involved in fatty acid oxidation. Therefore, we compared the relative gene expression of transcription factors and enzymes involved with fatty acid oxidation in the livers from fasted SD and F344 animals. The gene expression of the transcription factor PPARα was ∼1.8-fold greater in 70-day-old SD rats than in age-matched F344 rats (Fig.3). PPARα stimulates the gene expression of several enzymes involved with fatty acid oxidation. These enzymes include the rate-limiting enzyme ACO in peroxisomal lipid oxidation and the protein CPT I, which transports fatty acids for oxidation in mitochondria. The ACO gene expression was somewhat less in 70-day-old F344 animals than in SD animals. The gene expression of CPT I in 70-day-old SD animals was 1.5-fold greater than the CPT I gene expression in age-matched F344 animals. In the 130-day-old animals, PPARα gene expression was greater in the F344 animals than in the SD animals. However, the gene expression of ACO and CPT I in 130-day-old SD animals was more than two times as great as the gene expression of ACO and CPT I in age-matched F344 animals.
The effects of fasting and refeeding on gene expression of proteins involved in fatty acid oxidation were assessed in both 70-day-old (Table 3) and 130-day-old (Table4) animals. The gene expression of PPARα in fasted animals was ∼1.5- to 1.75-fold greater than the gene expression of PPARα in refed animals in both strains and ages of rat. However, the ratio of fasted to refed ACO and CPT I gene expression was significantly greater in SD animals than in age-matched F344 animals.
Finally, recent studies in experimental animals have shown that resistin gene expression increases with adiposity and with insulin resistance (21). We therefore measured the gene expression of resistin in the insulin- and leptin-resistant F344 strain and the more insulin- and leptin-sensitive SD strain. As shown in Fig. 4,left, resistin gene expression was slightly lower in fasted, 70-day-old F344 animals than in fasted, 70-day-old SD animals. We did not detect a statistically significant difference in resistin gene expression between fasted and refed 70-day old animals. After 2 mo, the resistin gene expression increased by ∼1.6-fold (P < 0.001) in SD animals. By contrast, in F344 animals, the resistin gene expression increased by ∼5.0-fold (P < 0.0001) in the 130-day-old animals than in the 70-day-old animals. In 130-day-old animals, resistin gene expression in the F344 strain was significantly greater than the resistin gene expression in the SD strain in the refed state. In the fasted state, there was not a statistically significant difference in resistin gene expression (P = 0.07).
In the present study, we have characterized an inbred rodent strain that we believe represents a good model for typical type 2 diabetes in humans. We have found that young F344 animals with normal glucose tolerance have evidence of insulin resistance (Fig. 2), leptin resistance (Tables 1 and 2), dyslipidemia (fasting and postprandial hypertriglyceridemia; Tables 1 and 2), and excess triglyceride accumulation in insulin-sensitive target tissues such as liver and muscle (Tables 1 and 2). These constellations of findings are consistent with a metabolic syndrome that is observed in young animals with normal glucose tolerance, and they probably represent a genetic risk for diabetes mellitus that occurs later in life.
We also investigated the link between insulin resistance and the propensity for obesity in F344 animals. Because leptin has been shown to stimulate fatty acid oxidation and reduce triglyceride storage in nonadipose tissue (23), we hypothesized that leptin resistance in F344 animals would diminish fatty acid oxidation and enhance triglyceride storage in nonadipose tissue. Although we did not directly measure fatty acid oxidation, we did measure the gene expression of several key regulatory proteins in fatty acid oxidation. We have demonstrated that the hepatic gene expression of ACO and CPT I is consistently less in the leptin- and insulin-resistant F344 animals than in age-matched SD animals (Fig. 3). ACO and CPT I are critical regulatory proteins in the fatty acid oxidation pathway in peroxisomes and mitochondria, respectively. The gene expression of both proteins is regulated, in part, by the transcription factor PPARα. Consistent with this observation, we have found that the gene expression of PPARα is less in 70-day-old F344 rats than in age-matched SD rats. Diminished PPARα gene expression in the leptin-resistant F344 animal supports the findings by Zhou et al. (25) that leptin normally stimulates PPARα gene expression. However, we did not find that PPARα gene expression was less in the 130-day-old F344 rats than in SD rats. The reason for this apparent discrepancy is unknown. However, despite increased PPARα gene expression, 130-day-old F344 animals had diminished downstream expression of regulatory proteins in fatty acid oxidation. Two reasons might account for an increase in PPARα but a decrease in ACO and CPT I gene expression. First, PPARα transcription activity not only depends on the amount of the gene product but also on the binding of its endogenous ligand, purportedly a fatty acid. PPARα transcriptional activity could be reduced in the 130-day-old F344 animals if the intrahepatic fatty acid composition was different and the activated PPARα gene transcribed less well than age-matched SD animals. Second, PPARα is not the only transcriptional activator of ACO and CPT I; other PPARα-independent transcriptional inhibitors may play a role in the gene expression of ACO and CPT I in F344 animals.
We believe that the reduction in gene expression of ACO and CPT I in F344 animals probably reduced hepatic fatty acid oxidation and increased the propensity to store fatty acids as triglycerides. We found that liver contains more triglycerides in F344 animals than in SD animals (Tables 1 and 2). Muscle contains more triglycerides in 130-day old F344 animals than in age-matched SD animals (Table 2). The enhanced triglyceride storage in nonadipose tissue resulted in insulin resistance (Fig. 2) and its associated disturbance of hyperinsulinemia and hypertriglyceridemia (Tables 1 and 2).
The response of the gene expression of PPARα and its cognate genes to fasting and refeeding was appropriate; the gene expression was higher with fasting and was diminished with refeeding, as previously demonstrated (7). However, the fractional decrease in response to refeeding was not as great in F344 animals as in the SD animals (see Tables 3 and 4). This phenomenon may be an example of what Kelley and Mandarino (6) have referred to as “metabolic inflexibility.” In the lean, insulin-sensitive SD animals, energy production during fasting conditions might rely predominantly on lipid oxidation, with a rise in gene expression of proteins mediating lipid oxidation. With refeeding, the gene expression of proteins involved in lipid oxidation decreases as insulin-mediated glucose production likely becomes predominant in the SD animals. In contrast, the obesity-prone, insulin-resistant F344 animals likely manifest less lipid oxidation with a fall in gene expression of proteins involved in lipid oxidation during fasting conditions. However, lipid oxidation most likely plays a greater role during insulin-stimulated refeeding in F344 animals, as the fractional decline in gene expression of proteins mediating lipid oxidation do not decline to the same degree as in SD animals. The failure to augment lipid oxidation during fasting conditions likely is a mechanism leading to lipid accumulation within insulin-sensitive target tissues, whereas the failure to fully suppress lipid oxidation during refeeding may be a manifestation of the relative resistance to insulin.
Finally, we have examined the role of resistin in the etiology of insulin resistance in F344 animals. Resistin gene expression and protein secretion are increased in both genetic and diet-induced models of obesity and diabetes (21). Furthermore, resistin appears to antagonize the effects of insulin both in vivo and in vitro (21). Such an antagonism might provide a link between obesity and insulin resistance. In vivo administration of resistin antiserum to diet-induced obese mice improved blood glucose levels and insulin sensitivity (21). Conversely, administration of recombinant resistin impairs glucose tolerance and insulin action in normal mice. Resistin's role in obesity and insulin resistance has not been confirmed by other investigators (24). We have hypothesized that resistin gene expression in the insulin-resistant, obesity-prone F344 rats would be higher than in SD rats. Our results do not support this hypothesis. The resistin gene expression in insulin-resistant, 70-day-old F344 animals was less than in age-matched SD animals (Fig. 4). At 70 days, the difference in the amount of stored fat in the epididymal fat pad (Table1) or in total body adiposity (15) differs only slightly between the two strains. After 60 days of a low-fat diet, F344 rats gain markedly more fat, without much change in insulin resistance, as measured by the intravenous glucose tolerance test. Despite almost equal glucose tolerance between 70- and 130-day-old F344 animals, the expression of the resistin gene increases dramatically. From these observations, we must conclude that resistin gene expression is not involved in the etiology of insulin resistance in F344 animals. Resistin gene expression does increase proportionally to the gain in total body fat, but animals do not become progressively more insulin resistant in this 2-mo period. Resistin gene expression is another marker for total body adiposity in both SD and F344 strains.
This work is supported by the Veterans Administration Merit Review (J. R. Levy).
Address for reprint requests and other correspondence: J. R. Levy, McGuire Veterans Administration Medical Center 111-P, 1201 Broad Rock Blvd., Richmond, VA 23249 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society