Increased hepatic steatosis and insulin resistance in mice lacking hepatic androgen receptor

Authors

  • Hung-Yun Lin,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    Search for more papers by this author
    • These authors contributed equally to this study.

  • I-Chen Yu,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    Search for more papers by this author
    • These authors contributed equally to this study.

  • Ruey-Shen Wang,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    2. Department of Gynecology and Obstetrics, Taipei Medical University, Taipei, Taiwan
    Search for more papers by this author
  • Yei-Tsung Chen,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    Search for more papers by this author
  • Ning-Chun Liu,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    Search for more papers by this author
  • Saleh Altuwaijri,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    Search for more papers by this author
  • Cheng-Lung Hsu,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    2. Division of Hematology-Oncology, Department of Internal Medicine, Chang Gung Memorial Hospital, Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Wen-Lung Ma,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    Search for more papers by this author
  • Jenny Jokinen,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    Search for more papers by this author
  • Janet D. Sparks,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    Search for more papers by this author
  • Shuyuan Yeh,

    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    Search for more papers by this author
  • Chawnshang Chang

    Corresponding author
    1. George Whipple Lab for Cancer Research, Departments of Pathology, Urology, and the Cancer Center, University of Rochester Medical Center, Rochester, NY
    • 601 Elmwood Avenue, Box 626, Rochester, NY 14642
    Search for more papers by this author
    • fax: (585) 756-4133


  • Potential conflict of interest: Nothing to report.

Abstract

Early studies demonstrated that whole-body androgen receptor (AR)–knockout mice with hypogonadism exhibit insulin resistance. However, details about the mechanisms underlying how androgen/AR signaling regulates insulin sensitivity in individual organs remain unclear. We therefore generated hepatic AR-knockout (H-AR−/y) mice and found that male H-AR−/y mice, but not female H-AR−/− mice, fed a high-fat diet developed hepatic steatosis and insulin resistance, and aging male H-AR−/y mice fed chow exhibited moderate hepatic steatosis. We hypothesized that increased hepatic steatosis in obese male H-AR−/y mice resulted from decreased fatty acid β-oxidation, increased de novo lipid synthesis arising from decreased PPARα, increased sterol regulatory element binding protein 1c, and associated changes in target gene expression. Reduced insulin sensitivity in fat-fed H-AR−/y mice was associated with decreased phosphoinositide-3 kinase activity and increased phosphenolpyruvate carboxykinase expression and correlated with increased protein-tyrosine phosphatase 1B expression. Conclusion: Together, our results suggest that hepatic AR may play a vital role in preventing the development of insulin resistance and hepatic steatosis. AR agonists that specifically target hepatic AR might be developed to provide a better strategy for treatment of metabolic syndrome in men. (HEPATOLOGY 2008.)

Nonalcoholic fatty liver disease (NAFLD) is a common hepatic disorder with abnormal liver function, hyperlipidemia, and steatohepatitis that can progress to cirrhosis.1 The estimated prevalence of NAFLD in the United States is approximately 20% but is as high as 75% in obese and type 2 diabetic patients.2, 3 Studies have suggested that ectopic triglyceride (TG) accumulation in the liver is a key element in the pathogenesis of insulin resistance,4 with increased endogenous glucose production in both lean and obese subjects.5, 6 Insulin resistance is an important component of type 2 diabetes and metabolic syndrome. Feeding rodents a high-fat diet (HFD) leads to hepatic steatosis, with impaired insulin signaling and hepatic insulin resistance.7, 8 Together, these results strongly suggest that hepatic lipid accumulation may be directly responsible for the subsequent development of hepatic insulin resistance and increased endogenous glucose production, key elements in the pathogenesis of type 2 diabetes. They also suggest that reducing hepatic lipid accumulation might be an effective therapeutic strategy for both NAFLD and type 2 diabetes.

Epidemiological evidence suggests that sex differences exist in type 2 diabetes, with a higher prevalence in men than in women,9 possibly because of differences in insulin sensitivity and regional deposition of body fat.10 Testosterone effects are mediated by the androgen receptor (AR).11, 12 The total and general AR-knockout mouse model (T-AR−/y) demonstrates progressively reduced insulin sensitivity and impaired glucose tolerance at advanced ages, similar to that observed in people with testosterone deficiency. The loss of the AR contributes to increased triglyceride content in skeletal muscle and liver in these animals, and serum leptin concentrations are elevated.13 One mechanism by which testosterone via AR might facilitate insulin sensitivity is by regulating the expression of insulin-dependent proteins, and dose-dependent effects of testosterone on insulin receptor substrate–1 (IRS-1) and glucose transporter–4 (GLUT-4) have been documented in cell models.14 Another potential mechanism involves AR-dependent changes in mitochondrial function, and decreased transcription of oxidative phosphorylation genes has been observed in skeletal muscle of insulin-resistant subjects. Decreased mitochondrial function leads to decreased lipid oxidation, ectopic TG accumulation, and, ultimately, to insulin resistance.15 In male subjects, testosterone level has been shown to correlate positively with mitochondrial capacity, as assessed by maximal aerobic capacity and expression of oxidative phosphorylation genes.16

Another potential mechanism involves protein tyrosine phosphatase 1B (PTP1B), a member of the protein tyrosine phosphatase family. PTP1B is a negative regulator of insulin action,17 attenuating insulin signaling through dephosphorylation of IR and IRS-1.18 PTP1B-knockout mice have increased insulin sensitivity and obesity resistance.19, 20 Recently, it was shown that androgen withdrawal in LNCaP (human prostate cancer cell line) cells increased expression of PTP1B with a corresponding increase in its tyrosine phosphatase activity.21

To determine whether AR is required and important in glucose and lipid homeostasis in the liver, we used a Cre-loxP strategy to specifically delete AR in the liver without impaired development of genital organs and subsequent hypogonadism. Our results demonstrate that male hepatic AR-knockout (H-AR−/y) mice have increased hepatic steatosis and insulin resistance on HFD feeding, with decreased β-oxidation and increased PTP1B expression. Results suggest that the AR plays an important role in hepatic glucose and lipid homeostasis.

Abbreviations

AR, androgen receptor; AR+/y, wild-type mice; H-AR−/y, hepatic AR-knockout mice; HFD, high-fat diet; NAFLD, nonalcoholic fatty liver disease; PEPCK, phosphenolpyruvate carboxykinase; PTP1B, protein tyrosine phosphatase 1B; SREBP-1c, sterol regulatory element binding protein 1c; TG, triglyceride.

Materials and Methods

Animals.

All animal procedures were approved by the animal care and use committee of the University of Rochester School of Medicine, in accordance with National Institutes of Health guidelines. Construction of targeting vectors and generation of the chimera founder mice have been described previously.13, 22 The floxAR mice were bred into a C57BL/6 background. Albumin is specifically expressed in the liver, and its promoter was used to drive Cre (Alb-Cre; C57BL/6; The Jackson Laboratory) and delete floxed AR fragments specifically in the liver. The H-AR−/y mice were genotyped by PCR, as described previously.13 Animals were housed in pathogen-free facilities, maintained on a 12-hour light/12-hour dark schedule (light on at 6:00 AM), and had access to standard laboratory chow (5010; PMI Lab Diet) and water ad libitum.

Histological Analysis.

Tissues were fixed in 4% (g/v) para-formaldehyde and embedded in paraffin, and sections were with hematoxylin/eosin. Images were acquired using an E800 microscope (Nikon) and a SPOT camera (Diagnostic Instruments) and were analyzed using SigmaScan Pro software (version 5.0; SPSS).

Analytical Procedures.

Fasting blood samples were taken from mice 14 hours after withdrawal of food. Blood samples designated as random-fed state were drawn 6 hours after introducing food into the cages of mice that had been subjected to a preceding 14-hour fast. Blood glucose concentration was measured using a glucometer (One Touch Ultra; Lifescan). Insulin and leptin levels were determined in duplicate 5-μL serum samples, using a mouse insulin and leptin enzyme-linked immunosorbent assay kit (Crystal Chem) according to the manufacturer's instruction. Serum adiponectin level was determined in duplicate in 20-μL serum samples using an enzyme-linked immunosorbent assay kit (eBioscience). For the glucose tolerance test (GTT), after a 14-hour fast, mice were given an oral bolus of D-glucose (2 g/kg body weight), and blood glucose concentration was measured in samples taken after 0, 30, 60, 90, and 120 minutes. Insulin tolerance tests (ITTs) were performed on 6-hour-fasting mice by intraperitoneal injection of 1 unit/kg body weight human insulin (Sigma Aldrich). Blood glucose concentrations were determined 0, 60, and 120 minutes after insulin administration. The triglyceride level in sera from fasting animals was determined using a GPO-Trinder assay (Sigma Aldrich). The serum-free fatty acid level in fasting animals was measured using a NEFA-Kit-U (Wako Pure Chemical). For determination of tissue triglyceride content, 50-100 mg of tissue was homogenized on ice in pH 7.3 extraction buffer (20 mmol/L Tris, 1 mmol/L β-mercaptoethanol, 1 mmol/L EDTA). After centrifugation, the glycerol content of the supernatants was determined using a GPO-Trinder assay (Sigma Aldrich) according to the manufacturer's instructions.

PI3k Activity by ELISA.

PI3k activity was determined using PI3k ELISA (Echelon Biosciences Inc.). This kit measures PI3k activity as a conversion of PI(4,5)P2 into PI(3,4,5)P3. We used this kit in conjunction with anti-p85 PI3k antibody (Upstate Biotechnology). Briefly, cells were washed 3 times with ice-cold buffer A and then lysed in buffer A containing 1% NP40 and 1 mM phenylmethylsulfonyl fluoride. Lysates were incubated on ice for 20 to 30 minutes and then cleared at 16,000g for 20 minutes. PI3k was immunoprecipitated by polyclonal anti-p85 PI3k antibody at 4°C for 1 hour. Binding of the coimmunoprecipitated p85/p110 PI3k complexes to protein A–agarose beads was carried out at 4°C for 1 hour, and the beads then were washed 3 times with buffer A containing 1% NP40, 3 times with buffer B, and twice with buffer C, as suggested by the manufacturer. The kinase reaction was carried out for 1 hour at room temperature in 4 mM MgCl2, 20 mM Tris-HCl (pH 7.4), and 20 mM NaCl supplemented with 25 μM ATP.

Mouse Primary Hepatocytes.

Mouse hepatocytes were obtained by nonrecirculating collagenase perfusion through the portal vein of male wild-type (AR+/y) or age-matched H-AR−/y mice. This standard procedure was carried out as described previously.23

Fatty Acid Oxidation.

Isolated hepatocytes were incubated for 1 hour in scintillation vials in Krebs buffer with 2% BSA, with isotope-labeled 1 mM palmitic acid. Radiolabeled CO2 was collected in center wells with Whatman filter paper and 300 μL of 1M methylbenzethonium hydroxide in methanol. At the end of the incubation, 300 μL of 5M H2SO4 was added to volatilize the remaining CO2, and the solution was incubated for another 30 minutes. The center wells were then placed in other scintillation vials, and 8 mL of aqueous scintillant was added and counted on a β counter.

Real-Time Quantitative PCR.

Total RNA was prepared from cells or tissues with Trizol (Initrogen) according to the manufacture's instructions. We used a real-time PCR method to quantify the mRNA as described previously.13

Statistical Analysis.

Data are means ± SEM. Differences between 2 groups were assessed by the unpaired 2-tailed Student's t test. Data involving more than 2 groups were assessed by 1-way ANOVA plus the Student-Newman-Keuls method (SigmaStat, SyStat).

Results

Generation of Hepatic ARKO Mice.

Using a Cre-loxP conditional knockout strategy, we mated female heterozygous floxed AR mice with male Alb-Cre mice to generate H-AR−/y and wild type AR (AR+/y; Alb Cre+) littermates. We detected Cre and floxed AR DNA fragments in tail genomic DNA of weaned 21-day-old H-AR−/y mice (Fig. 1A). Various tissues including epididymis, heart, kidney, muscle, spleen, testis, and adipose were harvested from 8-week-old H-AR−/y and AR+/y mice, and only liver tissue showed deletion of AR exon 2 with an 180-bp product from RT-PCR using primers for exons 1 and 3 (Fig. 1B). This finding suggests that H-AR−/y mice have selective disruption of AR specifically in the liver.

Figure 1.

Generation of mice with conditional knockout of AR in hepatocytes (H-AR−/y). (A) Identification and confirmation of H-AR−/y mice. Genomic DNA was isolated from tail snips and used as a template for PCR with primers “select” and “2-3” or Cre. The detailed method and primer sequences were described previously.13 The expression of floxed AR and Alb-Cre in the tail genomic DNA of H-AR−/y mice was confirmed by PCR. (B) RT-PCR of various tissues harvested from AR+/y and H-AR−/y mice. Only the mRNA from the livers of the H-AR−/y mice showed a knockout band (ko) and a wt AR band when primers exons 1 and 3 were used; other tissues had only a wt band. Data are representative images from the two experimental groups (n = 5).

Male H-AR−/y Mice But Not Female H-AR−/− Mice Were More Susceptible to Diet-Induced Obesity.

Previous findings demonstrated that T-AR−/y mice develop hepatic steatosis and insulin resistance at an advanced age. To determine if hepatic AR plays major roles in the development of obesity and obesity-associated hepatic steatosis and insulin resistance, we fed either a normal diet or an HFD to H-AR−/y mice and AR+/y mice. In mice fed the normal diet for 24 weeks, the growth curve was not affected by AR genotype among mice of the same sex (Fig. 2A; Supplementary Fig. 1A). Previous studies suggested that male T-AR−/y mice fed a standard chow diet exhibit an obese phenotype at advanced age. Consistent with these reports, after 8 weeks of HFD feeding, male H-AR−/y mice weighed 13% more than their AR+/y counterparts (Fig. 2B,C). In contrast, female mice on an HFD were resistant to weight gain (Supplementary Fig. 1B).

Figure 2.

Growth curve, cumulative weight gain, and body composition of male AR+/y and H-AR−/y mice fed an HFD. Mice were given an HFD for 8 weeks starting when they were 8-weeks-old. (A) Growth curves of animals fed normal chow. (B) Growth curves of animals fed an HFD. (C) Cumulative weight gain of the animals fed an HFD was measured over an 8-week period. (D) Adiposity index was calculated by fat (epididymal plus infrarenal)/weight ratio of male AR+/y and H-AR−/y mice. (E) There were no significant differences in testes weight of male AR+/y and H-AR−/y mice. Data are means ± SEMs of 6-7 mice per group (*P < 0.05 AR+/y [HFD] versus H-AR−/y [HFD]).

To further dissect why male H-AR−/y mice gained more weight while consuming the HFD, we examined the adiposity index by calculating percent body fat (infrarenal plus periepididymal fat pads) in lean and obese groups. At the end of the 8-week HFD feeding period, obese male H-AR−/y mice had significant increased adiposity compared with that of obese male AR+/y mice (Fig. 2D), but had similar testis weights (Fig. 2E).

Male H-AR−/y Mice But Not Female H-AR−/− Mice Developed Obesity-Induced Hepatic Steatosis, and Normal Chow–Fed Aging Male H-AR−/y Mice Also Developed Hepatic Steatosis.

Obesity is the most common risk factor for the development of hepatic steatosis. To determine whether increased weight gain in obese male H-AR−/y mice was associated with hepatic steatosis and hepatomegaly, we analyzed livers from obese male H-AR−/y and AR+/y mice (Fig. 3A, D). As expected, hepatic histology demonstrated increased macrovesicular steatosis in obese male H-AR−/y mice (Fig. 3E), whereas obese male AR+/y mice had only moderate steatosis (Fig. 3B), and female mice showed little steatosis (data not shown). Likewise, the livers of obese male H-AR−/y mice weighed 25% more than those of obese male AR+/y mice (Fig. 3G). Interestingly, hepatic histology demonstrated that aging male H-AR−/y mice fed normal chow also developed moderate hepatic steatosis at 40 weeks of age (Fig. 3F).

Figure 3.

TG accumulation in the livers of male AR+/y and H-AR−/y mice on HFD. Images of H&E staining of liver sections of (A) AR+/y mice fed chow, (B) AR+/y mice fed HFD, (C) 40-week-old AR+/y mice fed chow, (D) H-AR−/y mice fed chow, (E) H-AR−/y mice fed HFD, and (F) 40-week-old H-AR−/y mice fed chow. The clear vacuoles in the liver sections are identified by arrows in (B) and (E). Lipid vacuoles in obese male H-AR−/y mice fed HFD were increased in size and number. Data are representative images (n = 4-5). (G) Liver weight was increased in obese H-AR−/y mice. (H) Expression of genes involved in lipogenesis and oxidation determined by QPCR. Data are means ± SEMs of 3 independent experiments (*P < 0.05 AR+/y [HFD] versus H-AR−/y [HFD]; **P < 0.01 AR+/y [HFD] versus H-AR−/y [HFD]). (I) Liver TG content was increased in obese H-AR−/y mice. Data are means ± SEMs of 4-5 mice per group (**P < 0.01 AR+/y [HFD] versus H-AR−/y [HFD]). (J) TG production rate was higher in H-AR−/y mice. Data are means ± SEMs of 3 mice per group (*P < 0.05 AR+/y versus H-AR−/y). (K) Expression of Scd1 and Ppara in primary hepatocytes determined by QPCR. Primary hepatocytes from AR+/y and H-AR−/y mice were incubated with EtOH or DHT (1 nM) for 24 hours. Data are means ± SEMs of three independent experiments (*P < 0.05 AR+/y [EtOH] versus AR+/y [DHT]). (L) Rates of oxidation and expression of Acox in primary hepatocytes determined by QPCR. Primary hepatocytes from AR+/y and H-AR−/y mice were incubated with vehicle or clofibrate (0.5 mM) for 24 hours. Data are means ± SEMs of three independent experiments (**P < 0.01 H-AR−/y [vehicle] versus H-AR−/y [clofibrate]).

As early studies suggested that adipokines such as leptin and adiponectin are able to reverse hepatic steatosis in ob/ob mice and lipodystrophic mice, we therefore assayed leptin and adiponectin levels in male H-AR−/y mice with hepatic steatosis. The refed serum leptin level was significantly higher in obese male H-AR−/y mice, along with a reduced fasting serum adiponectin level in lean male H-AR−/y mice (Table 1). However, increased serum leptin was not sufficient to reduce hepatic steatosis and insulin resistance in obese male H-AR−/y mice (Table 1). Together, these results suggest that obese male H-AR−/y mice had abnormal adipokine profiles and systemic leptin resistance that might contribute to the development of obesity-induced hepatic steatosis.

Table 1. Metabolic Parameters During Fasting in AR+/y and H-AR–/y Mice Fed Chow or HFD at 16 Weeks Old
GroupAR+/yH-AR–/y
Chow DietHigh-Fat DietChow DietHigh-Fat Diet
  • Fasting refers to an overnight fast. Data are mean ± SEM of 6-7 mice per group.

  • *

    P < 0.05 AR+/y (HFD) versus H-AR-/y (HFD).

  • Leptin was assayed in fed animals.

  • P < 0.01 AR+/y (HFD) versus H-AR-/y (HFD).

  • §

    P < 0.05 AR+/y (chow) versus H-AR-/y (chow). P < 0.01 AR+/y (chow) versus H-AR-/y (chow). nd, not determined.

Triglyceride (mg/dL)20.1 ± 2.627.3 ± 3.330.7 ± 2.435.2 ± 2.6*
NEFA (mEq/L)0.86 ± 0.070.76 ± 0.110.96 ± 0.080.71 ± 0.13
Cholesterol (mg/dL)113.2 ± 3.4142.6 ± 5.1105.7 ± 8.3137.3 ± 10.6
Leptin (ng/mL)2.3 ± 0.29.6 ± 2.13.2 ± 0.115.1 ± 3.4
Adiponectin (μg/mL)2.13 ± 0.09nd1.42 ± 0.19§nd
IGF-1 (ng/mL)342.9 ± 33.6nd156.7 ± 18.6nd
GH (ng/mL)62 ± 8.1nd240 ± 36.4nd
Testosterone (ng/mL)3.34 ± 0.59nd3.49 ± 0.91nd

Obese Male H-AR−/y Mice Developed Hepatic Steatosis Via Reduced Lipid Oxidation and Increased De Novo Lipid Synthesis.

We first observed reduced fatty acid oxidation in isolated primary hepatocytes from H-AR−/y mice (Fig. 3L). Mechanistic dissection found that PPARα and sterol regulatory element binding protein 1c (SREBP-1c), 2 key genes involved in the modulation of fatty acid oxidation, were altered in the livers of obese male H-AR−/y mice (Fig. 3H). Increased SREBP-1c expression is consistent with the up-regulation of acetyl CoA carboxylase (ACC), the first committed step in fatty acid synthesis. With increased malonyl CoA levels leading to inhibition of carnitine palmitoyl transferase 1 (CPT-1), it then results in reduced mitochondrial fatty oxidation via reduction of free fatty acid transport from the cytosol to mitochondria. Reduced expression of PPARα in H-AR−/y hepatocytes consistent with decreased malonyl CoA decaboxylase (MCAD) (Fig. 3H) could also contribute to increased malonyl CoA levels.

Increased De Novo Fatty Acid Synthesis.

We noted increased ectopic TG accumulation in the livers of obese male H-AR−/y mice (Fig. 3I) and increased exportation of VLDL-TG from the livers of obese male H-AR−/y mice (Fig. 3J). SREBP-1c, a key gene in the regulation of fatty acid synthesis, and its major target gene, stearoyl-CoA desaturase 1 (SCD1), were both increased in H-AR−/y hepatocytes (Fig. 3H). These results suggest increased de novo fatty acid synthesis may be partially responsible for ectopic TG accumulation in liver.

Confirmation of Involvement of PPARα in Development of Hepatic Steatosis in Obese Male H-AR−/y Mice.

Both decreased PPARα and increased SREBP-1c might play key roles in the development of hepatic steatosis in obese male H-AR−/y mice. To support the involvement of AR in PPARα, we further confirmed these results in primary hepatocytes. The addition of DHT to primary hepatocytes increased PPARα expression only in cells from AR+/y mice (Fig. 3K), indicating that lack of ARs prevented androgen-induced PPARα. However, the addition of the PPARα agonist clofibrate resulted in the restoration of fatty acid oxidation and expression of ACOX in isolated H-AR−/y hepatocytes (Fig. 3L), indicating that the PPARα ligand was still able to activate fatty acid oxidation in an androgen-independent fashion.

Male H-AR−/y Mice But Not Female H-AR−/− Mice Developed Obesity-Induced Hepatic Insulin Resistance.

Hepatic insulin resistance in obese mice and humans is strongly associated with hepatic steatosis and hepatomegaly. Increased fat mass in male mice fed an HFD would be expected to influence systemic metabolic parameters, including insulin sensitivity. We measured fasting blood glucose and serum insulin levels and found no differences between lean male H-AR−/y and lean AR+/y mice (Fig. 4A-C). Similarly, we found no differences in glucose level in female mice of either AR genotype (Supplementary Fig. 2). However, HFD-fed male H-AR−/y mice had increased fasting, refed blood glucose, and serum insulin levels compared with those in male AR+/y counterparts (Fig. 4A-C). The calculated homeostasis model of insulin resistance (HOMA-IR) values of obese male H-AR−/y mice were 1.2-fold higher than those of their obese AR+/y counterparts (Fig. 4D). During insulin tolerance tests (ITTs) following intraperitoneal injection of insulin in male obese mice, insulin was less effective in reducing blood glucose in obese male H-AR−/y mice than in their obese male AR+/y counterparts (Fig. 5A). Likewise, AR deficiency decreased glucose tolerance during glucose tolerance tests (GTTs) following administration of an intraperitoneal glucose bolus with increased IR index (Table 2). Although clearance of blood glucose was reduced in obese male H-AR−/y and AR+/y mice (Fig. 5B,C), serum insulin level was higher in obese male H-AR−/y mice (Fig. 5D).

Figure 4.

Insulin sensitivity in obese male AR+/y and H-AR−/y mice. (A) Fasting serum glucose levels, (B) refed serum glucose levels, and (C) fasting serum insulin levels were measured in male AR+/y and H-AR−/y mice fed normal chow and following 8 weeks of HFD feeding. (D) HOMA-IR indices were higher among 24-week-old obese male H-AR−/y mice than among 24-week-old obese male AR+/y mice. Data are means ± SEMs of 4-5 mice per group (*P < 0.05 AR+/y [chow] versus AR+/y [HFD]; **P < 0.01 AR+/y [HFD] versus H-AR−/y [HFD]).

Figure 5.

Male H-AR−/y mice fed HFD had defects in glucose homeostasis. (A) ITTs were performed. Blood samples were collected, and glucose was measured at the times indicated. The percent reduction in glucose concentration in obese male H-AR−/y mice was reduced compared with that in obese male AR+/y mice. (B) IGTTs were performed in overnight fasted male AR+/y and H-AR−/y mice. Blood samples were collected, and glucose was measured at the times indicated. Glucose concentration in obese male H-AR−/y mice was higher than that in obese male AR+/y mice at 30 and 60 minutes. (C) IGTT areas under the curve were higher among obese male H-AR−/y mice than in obese male AR+/y mice. (D) Serum insulin concentration in obese H-AR+/y mice was higher after 30, 60, 90, and 120 minutes. Data are means ± SEMs of four mice per group (*P < 0.05 AR+/y [HFD] versus H-AR−/y [HFD]; **P < 0.01 AR+/y [HFD] versus H-AR−/y [HFD]).

Table 2. Intraperitoneal Glucose Tolerance Test (IGTT) and Insulin Resistance (IR) Index in AR+/y and H-AR–/y Mice Treated with Chow or HFD at 16 Weeks Old
Group (16 Weeks)GTT Area (cm2)IR Index
GlucoseInsulin
  1. AUC was calculated by multiplying the cumulative mean height of the glucose values (1 mg/mL = 1 cm) and insulin values (1 ng/mL = 1 cm), respectively, by time (60 minutes = 1 cm). IR index was calculated by the product of the AUC of glucose and insulin × 102. Data are mean ± SEM of three mice per group; *P < 0.001 AR+/y (HFD) versus H-AR–/y (HFD).

ChowAR+/y38.35 ± 4.331.02 ± 0.2239.32 ± 4.34
 H-AR–/y40.63 ± 5.581.14 ± 0.2646.31 ± 5.59
HFDAR+/y53.44 ± 5.761.61 ± 0.3486.03 ± 5.77
 H-AR–/y62.18 ± 6.532.96 ± 0.58184.05 ± 6.56*

Obesity-Induced Insulin Resistance in Obese Male H-AR−/y Mice Was Further Confirmed with Decreased PI3k Activity.

Insulin resistance in mice is manifested by the reduced capacity of insulin to activate the downstream kinase phosphoinositide-3 kinase (PI3k). To evaluate insulin action, we injected fasted male H-AR−/y and AR+/y mice with an intraperitoneal insulin bolus and harvested target tissues 5 minutes later to assess the effects of insulin on PI3k activity by immunoprecipitation with an anti-PI3k antibody followed by a PI3k ELISA. Glucose and insulin intolerance previously observed in obese male H-AR−/y male mice was associated with decreased insulin-stimulated PI3k activity in liver, with only a moderate reduction in skeletal muscle (Fig. 6A). Together, these results suggest that hepatic insulin resistance may precede the development of muscle insulin resistance following loss of hepatic AR.

Figure 6.

Reduced hepatic insulin signaling in obese male H-AR−/y mice. (A) Fasted mice were injected with an intraperitoneal bolus of insulin, and tissues were sampled 5 minutes later. PI3k activity was measured in p85-immunoprecipitates, and representative PIP3 levels using ELISA method are shown. Hepatic PI3k activity was reduced in lean and obese male H-AR−/y mice. Data are means ± SEMs; n = 3 (*P < 0.05 AR+/y [chow] versus AR+/y [HFD]; #P < 0.05 AR+/y [HFD] versus H-AR−/y [HFD]). (B) Expression of Hnf4, Pck1, and Ptp1b was determined by QPCR. Data are means ± SEMs of 3 independent experiments (*P < 0.05 AR+/y [HFD] versus H-AR−/y [HFD]; ***P < 0.001 AR+/y [HFD] versus H-AR−/y [HFD]). (C) Expression of Pck1 and Ptp1b in primary hepatocytes was determined by QPCR. Primary hepatocytes from AR+/y and H-AR−/y mice were incubated with ethanol or DHT (1 nM) for 24 hours. Data are means ± SEMs of 3 independent experiments (*P < 0.05 AR+/y [EtOH] versus AR+/y [DHT]). (D) PI3k activity and expression of Pck in primary hepatocytes determined by QPCR. Primary hepatocytes from AR+/y and H-AR−/y mice were incubated with vehicle or the PTP1B inhibitor RK-682 (0.1 μM) with or without insulin for 24 hours. PI3k activity and expression of insulin targeting the Pck gene was reduced on RK-682 treatment in H-AR−/y hepatocytes. Data are means ± SEMs of three independent experiments (**P < 0.01 H-AR−/y [vehicle, with insulin] versus H-AR−/y [RK-682, with insulin]).

Obese Male H-AR−/y Mice Developed Hepatic Insulin Resistance That Correlated with Increased PTP1B and PEPCK Expression.

To dissect the molecular mechanism of how obese male H-AR−/y mice develop hepatic insulin resistance, we assayed PTP1B expression in liver because PTP1B has been demonstrated to play vital roles in the regulation of insulin and glucose signaling. We found that hepatic PTP1B was highly expressed in obese male H-AR−/y mice compared with obese male AR+/y mice (Fig. 6B). This finding was confirmed by comparing primary hepatocytes isolated from H-AR−/y and AR+/y mice (Fig. 6C). Furthermore, expression of phosphenolpyruvate carboxykinase (PEPCK), a master regulator of hepatic glucose production and a major downstream target of insulin signaling, was higher because of increased PTP1B expression in the livers of obese male H-AR−/y mice (Fig. 6B,C). The addition of the PTP1B antagonist RK-682 into primary hepatocytes from H-AR−/y mice reversed the increased PEPCK expression and reduced PI3k activity (Fig. 6D). Taken together, our data suggest that loss of hepatic AR leads to hepatic insulin resistance via increased PTP1B and PEPCK expression.

Discussion

Androgen–AR signaling plays a pivotal role in reproductive organ development and prostate cancer. The role of the AR in metabolic disorders has not been previously appreciated. Although hepatic insulin resistance and steatosis are associated, a causal relationship between ectopic fat accumulation and insulin sensitivity has not been clearly established. We present evidence from the present study that supports the importance of the hepatic AR in the regulation of lipid and glucose homeostasis and demonstrates that hepatic deficiency of ARs significantly reduces glucose tolerance and insulin sensitivity and increases the accumulation of lipid in livers derived from obese male H-AR−/y mice.

The mechanisms leading to the development of hepatic steatosis are likely to be complex. Previous studies have shown that genes involved in lipogenesis are elevated in livers derived from ob/ob mice,24 and the transcriptional factor SREBP-1c was shown to contribute to the increased rate of lipogenesis in livers of these mice.24, 25 Hepatic expression of fatty acid synthase is increased in livers of ob/ob mice, and the fatty acid synthase gene is transcriptionally regulated by SREBP-1c.26 Indeed, SREBP-1c expression was markedly increased in the livers of obese male H-AR−/y mice; they showed significantly increased hepatic steatosis on HFD feeding. Aging male H-AR−/y mice spontaneously developed hepatic steatosis even when consuming normal chow. Our study demonstrates that under short-term HFD feeding conditions, hepatic steatosis was substantially increased in male H-AR−/y mice and was associated with increasing rates of lipogenesis, a contributor to TG accumulation. In addition to affecting de novo lipogenesis, hepatic knockout of the AR also affected β-oxidation. Lipogenesis and β-oxidation are inversely related because malonyl-CoA produced by ACC, the first committed enzyme in fatty acid synthesis, appears to regulate fatty acid oxidation by allosteric inhibition of CPT-1, the rate-limiting enzyme of β-oxidation.27 In rodents with NAFLD, suppression of ACC with antisense oligonucleotide significantly reduces hepatic malonyl-CoA level, lowers hepatic lipids, and improves insulin sensitivity.28 Consistent with previous findings, ACC content was higher in livers of obese male H-AR−/y mice, which likely led to reduced activity of CPT-1. Another transcriptional factor, PPARα, governs the expression of major enzymes of the β-oxidation pathway, as well as CPT-1 expression.29 PPARα expression was reduced in the livers of obese male H-AR−/y mice. However, the lower expression of PPARα in human liver than in rodent liver,30 as well the dominant-negative splice variant of PPARα in human liver,31 suggests a more modest role of PPARα in humans. Therefore, the coordinated modulation of gene expression in lipogenesis and oxidation was influenced in livers derived from obese males lacking ARs.

Interestingly, we found that the sex of the animals appears to influence the effectiveness of AR deletion in increasing body weight and hepatic fat accumulation. Not only did male H-AR−/y mice weigh more and have more hepatic steatosis with short-term HFD feeding, but a similar effect also was observed in male H-AR−/y mice fed standard chow at advanced age. Female H-AR−/− mice fed an HFD for 8 weeks, however, did not gain more weight than standard chow–fed female AR+/+ mice, suggesting that in females, deletion of hepatic ARs does not influence obesity induced by HFD and that female mice are resistant to diet-induced obesity. In comparison, only male H-AR−/y mice fed an HFD for 8 weeks, rather than standard chow for 8 weeks, accumulated more body weight, and had increased hepatic steatosis and insulin resistance compared with their similarly fed male AR+/y counterparts. If these sexually dimorphic differences in the responses to changes in hepatic androgen–AR signaling were to hold true in humans, different outcomes in males and females with hyperandrogenic or hypoandrogenic conditions might be expected. Indeed, hyperandrogenic conditions, such as polycystic ovarian syndrome, have been strongly associated with reduced glucose tolerance and insulin sensitivity in women,32, 33 whereas hypoandrogenism has been linked with insulin resistance16, 34 and adiposity16, 35 in men.

PTP1B is a well-characterized protein phosphatase and has been shown to function as a negative regulator of insulin signaling and the leptin-signaling pathways and is a mediator of resistance to both hormones.36, 37 Leptin improves insulin sensitivity by direct action on hepatocytes.38 Markedly elevated hepatic PTP1B expression impairs the ability of leptin to reduce serum glucose levels by inhibition of hepatic PEPCK activity.39 Hepatic PTP1B and serum leptin levels were increased in obese male H-AR−/y mice, whereas hepatic leptin signaling was impaired. Leptin resistance and consequently up-regulated leptin secretion are believed to play a pivotal role in obesity. Testosterone supplementation in hypogonadal men reduces leptin secretion by fat cells through an AR-mediated pathway,40 breaking the circuit of leptin resistance and obesity.41 The liver has long been implicated in the regulation of food intake and body weight, but the signaling molecules and pathways involved have not been fully established. One hypothesis is that an increase in glucose utilization can produce a satiety signal.42 Hepatic sensors detecting changes in energy expenditure trigger a signal in the vagal afferents acting centrally to inhibit food intake.42 The lipolytic effect of leptin in the liver, depleting TG content and increasing β-oxidation,43, 44 might provide an alternate satiety signal that could also be transmitted through vagal afferents.42 A recent study demonstrated induction of leptin receptor expression in the liver by leptin and by food deprivation, in association with a large increase in a short form of the leptin receptor that might bind and reduce the bioavailability of leptin to satiety centers.43

In H-AR−/y mice with loss of AR specifically in the liver, feeding an HFD also induced insulin resistance as was observed in T-AR−/y mice. Reduced PPARα expression and consequent reduced lipid oxidation with ectopic triglyceride accumulation in the liver were also observed in H-AR−/y mice fed an HFD. Reduced PPARα expression in AR−/y mice and H-AR−/y mice because of loss of the T-AR and increased PPARα expression in AR+/y hepatocytes with DHT treatment indicate AR directly regulates PPARα expression. Reduced PI3k activity only in liver not in skeletal muscle in H-AR−/y mice further proved that increased hepatic PTP1B levels reduce phosphorylation on tyrosine residues and consequently reduces PI3k activity specifically in the liver. The H-AR−/y mice model also suggests a plausible novel model of hepatic steatosis that might be a key element in hepatic insulin resistance.

In summary, our results have demonstrated that HFD feeding leading to hepatic steatosis and insulin resistance is associated with AR deficiency in obese male H-AR−/y mice. We found increased expression of PTP1B in the liver through an androgen-dependent mechanism that curtails the action of insulin and leptin. Our findings implicate the hepatic AR as a positive factor in preventing the development of hepatic steatosis and insulin resistance. Strategies aimed at increasing AR activity specifically in the liver through tissue-selective AR modulators could therefore improve both hepatic insulin and leptin sensitivity and improve both lipid and glucose homeostasis. Further understanding of the role of the AR in the development of insulin and leptin resistance in other tissues, such as brain and skeletal muscle, may provide additional insights into the pathophysiology of metabolic diseases.

Ancillary

Advertisement