Pathophysiology/Complications ORIGINAL ARTICLE

Low Plasma Adiponectin Levels Are Associated With Increased Hepatic Activity In Vivo

JOCHEN G. SCHNEIDER, MD PETER P. NAWROTH, MD epatic lipase functions as a lipolytic MAXIMILIAN VON EYNATTEN, MD KLAUS A. DUGI, MD that hydrolyzes triglycer- STEPHAN SCHIEKOFER, MD H ides and phospholipids in lipopro- teins of intermediate and high density. It is thereby involved in the formation of small, dense LDL and represents a major determinant of the plasma HDL choles- OBJECTIVE — Hepatic lipase plays a key role in hydrolyzing and phospholipids terol concentration (1,2). Hepatic lipase present in circulating plasma . Plasma hepatic lipase activity is known to be regulated activity is regulated by several hormonal by several hormonal and metabolic factors, but hepatic lipase responsiveness to insulin is still and metabolic factors (3). In insulin resis- controversial. Hypoadiponectinemia is known to be associated with insulin resistance, diabetes, and obesity. These conditions are often characterized by high plasma and low HDL tance, most studies show increased he- levels, and they have been shown to be associated with high plasma hepatic lipase patic lipase activity, although the exact activity. We therefore raised the question whether adiponectin may be associated with plasma regulation of hepatic lipase in insulin re- hepatic lipase activity in vivo. sistance is still controversial (4). Adi- ponectin is a member of a class of RESEARCH DESIGN AND METHODS — We measured plasma adiponectin and post- bioactive substances known as adipocyto- heparin hepatic lipase activity in 206 nondiabetic men and in a second group of 110 patients with kines (5). It is related to tumor necrosis type 2 diabetes. The correlation of these parameters with markers of insulin resistance and factor-␣ expression (6) and has the ability systemic inflammation was investigated. to suppress tumor necrosis factor-␣– induced activation of nuclear transcrip- RESULTS — In nondiabetic patients, adiponectin levels were significantly inversely corre- ␬ lated with plasma hepatic lipase activity (r ϭϪ0.4, P Ͻ 0.01). These results were confirmed in tion factor B (7), thereby demonstrating the group of patients with type 2 diabetes (r ϭϪ0.32, P ϭ 0.004). Multivariate analysis revealed anti-inflammatory potential. Recent stud- that adiponectin was the strongest factor influencing hepatic lipase activity. The association was ies suggest associations between low lev- independent of age, sex, BMI, plasma triglycerides, insulin, HDL cholesterol, and high- els of adiponectin and a profile of sensitivity C-reactive protein and accounted for ϳ10 and 12% of the variation in hepatic lipase low HDL cholesterol and increased activity in the two different patient cohorts, respectively. plasma triglyceride levels (8). Further- more, adiponectin plasma levels are in- CONCLUSIONS — These results demonstrate for the first time a significant inverse associ- versely correlated with BMI (9), insulin ation between adiponectin and postheparin plasma hepatic lipase activity that is independent of resistance, and type 2 diabetes (10). Low other factors such as markers of insulin resistance or inflammation. Therefore, adiponectin, HDL cholesterol levels, hypertriglyceride- rather than insulin, may represent an important factor contributing to the regulation of hepatic lipase activity in both nondiabetic individuals and patients with type 2 diabetes. The effect of mia, insulin resistance, type 2 diabetes, adiponectin on hepatic lipase activity may also help to explain the HDL cholesterol–elevating and abdominal obesity have also been action of adiponectin. consistently shown to be associated with postheparin hepatic lipase activity (11– Diabetes Care 28:2181–2186, 2005 13). In addition, the liver, as the main expression site for hepatic lipase, is also a major target of adiponectin action (14). These data suggest a direct or indirect re- lationship between hepatic lipase activity ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● and adiponectin by an as yet unknown mechanism. We therefore investigated in From the Department of Medicine I (Endocrinology and Metabolism), Ruprecht-Karls-University of Heidel- berg, Heidelberg, Germany. two different groups of patients whether Address correspondence and reprint requests to Jochen G. Schneider, MD, Washington University School adiponectin is associated with plasma of Medicine, Department of Endocrinology, Metabolism and Lipid Research, Campus Box 8127, St. Louis, hepatic lipase activity and whether this MO 63110. E-mail: [email protected]. relationship is affected by systemic in- Received for publication 7 March 2005 and accepted in revised form 26 May 2005. Abbreviations: CAD, coronary artery disease; CRP, C-reactive protein; ELISA, enzyme-linked immu- flammatory activity, insulin resistance, or nosorbent assay; hs-CRP, high-sensitivity CRP; HOMA, homeostasis model assessment; PPAR, peroxisome overt diabetes. proliferator–activated receptor; SREBP, sterol regulatory element–binding protein. A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion RESEARCH DESIGN AND factors for many substances. METHODS — Two different cohorts © 2005 by the American Diabetes Association. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby of patients were studied. 1) A total of 206 marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. nondiabetic men were recruited from the

DIABETES CARE, VOLUME 28, NUMBER 9, SEPTEMBER 2005 2181 Adiponectin and hepatic lipase activity

University Hospital Heidelberg, Depart- ultracentrifugation in a Beckman LM-8 tion were Ͻ5.0%. Plasma glucose was ment of Medicine. All individuals in this ultracentrifuge in 100-␮l volumes with a measured by a glucose oxidase method. group had diagnosed or suspected coro- VT-51.2 rotor (Beckman Coulter). The Serum insulin immunoreactivity was de- nary artery disease (CAD) (nondiabetic atherogenic index was calculated by the termined from frozen serum by ELISA CAD subjects) and underwent elective formula: (TC – HDL cholesterol)/HDL (CIS Bio International, Gif-Sur-Yvette, coronary angiography. The overt diagno- cholesterol. France). Plasma concentrations of CRP in sis of diabetes, according to American Di- a highly sensitive assay (high-sensitivity abetes Association criteria (fasting plasma Hepatic lipase CRP [hs-CRP]) were determined by glucose Ͼ126 mg/dl) (15), was an exclu- After an overnight fast, venous blood ELISA (Dade Behring, Cupertino, CA). sion criterion. 2) A total of 110 patients samples were drawn into EDTA tubes be- (82 male and 28 female subjects) with fore and 10 min after intravenous injec- Statistical analyses known type 2 diabetes were recruited tion of 60 IU heparin (Braun Melsungen, Statistical analyses were performed with from the Diabetes Outpatient Clinic of the Melsungen, Germany) per kg body wt. SPSS for Windows (version 12.0; SPSS, University Hospital Heidelberg. The dia- The samples were immediately chilled to Chicago, IL). Spearman correlation coef- betic patients enrolled in the study were 4°C, centrifuged, and stored at Ϫ80°C ficients were used to describe the associ- treated by either dietary intervention or until assayed. Postheparin hepatic lipase ation between adiponectin and other oral antidiabetic drugs (sulfonylurea activity was determined with a triolein– continuous variables of interest. Linear drugs, metformin, acarbose, glinides, or phosphatidylcholine emulsion as de- regression was used to control for poten- combinations); 29 patients (4 female and scribed previously (16). Selective tially confounding variables. The model, 25 male subjects) were treated with insu- measurement of hepatic lipase was based fitted for hepatic lipase activity as a de- lin and were omitted from the homeosta- on the inactivation of LPL by 1.0 mol/l pendent variable, included age, sex, BMI, sis model assessment [HOMA] calculation). NaCl. The samples were quantitated in plasma triglyceride concentration, HDL Treatment with subcutaneous or in- duplicate, and postheparin plasma from cholesterol concentration, hs-CRP, and travenous heparin in the previous 72 h, pooled normal control subjects was used plasma adiponectin concentration as in- severe kidney or liver disease, treatment to correct for interassay variation. The in- dependent variables to demonstrate the with drugs known to affect adiponectin tra-assay coefficient was 7.8%, and the in- relative contribution of each of these vari- plasma levels (such as peroxisome prolif- terassay coefficient was 11.4%. ables to the outcome variable. Variables erator–activated receptor [PPAR]-␥ ago- that were not normally distributed, such nists), and a fasting triglyceride level Blood variables as plasma adiponectin, triglyceride, and Ͼ11.4 mmol/l (1,000 mg/dl), suggesting In the morning after an overnight fast, ve- hs-CRP concentrations, were log trans- secondary lipid disorders, were exclusion nous blood was sampled for the measure- formed before they were entered into the criteria in both nondiabetic subjects with ment of plasma concentrations of statistical evaluation to better approxi- CAD and type 2 diabetic subjects. The adiponectin, glucose, insulin, and C-reac- mate normal distributions. Results are ex- study was approved by the Internal Ethics tive protein (CRP). Adiponectin samples pressed as means Ϯ SE. A P value Ͻ0.05 Committee of Heidelberg University, and were quantitated in duplicate by enzyme- was considered statistically significant. each patient gave informed consent. The linked immunosorbent assay (ELISA) (B- patients were advised not to consume al- Bridge International, San Jose, CA), and RESULTS — The clinical characteris- cohol during the study period to avoid plasma from four normal control subjects tics of the subjects studied are reported in abnormal changes in plasma triglyceride was used for interassay variation. Both the Table 1. Hepatic lipase activity was signif- levels caused by excessive consumption intra- and interassay coefficients of varia- icantly higher in male versus female dia- of alcohol. Patients were also advised to maintain a standard diet composed of ϳ25% protein, 15% fat, and 60% carbo- Table 1—Anthropometric and biochemical variables of the study subjects hydrate during the study period. The pa- tients were requested to not eat or drink Nondiabetic subjects Diabetic subjects from 10 P.M. on the day before the study visit. In the diabetic patients, antidiabetic n 206 110 medication was maintained at a constant Age (years) 60.9 Ϯ 10.4 55.9 Ϯ 9.6 dose throughout the study period, and BMI (kg/m²) 27.3 Ϯ 3.2 28.6 Ϯ 4.1 patients were instructed not to take their Total cholesterol (mmol/l) 5.3 Ϯ 1.1 5.7 Ϯ 1.4 morning dose of antidiabetic medication LDL cholesterol (mmol/l) 3.6 Ϯ 1.0 3.8 Ϯ 1.1 on study days. Concomitant medications HDL cholesterol (mmol/l) 1.1 Ϯ 0.3 1.0 Ϯ 0.4 were also maintained at a constant dose. VLDL cholesterol (mmol/l)* 0.6 Ϯ 0.4 1.0 Ϯ 1.4 Triglycerides (mmol/l)* 1.5 Ϯ 0.8 2.3 Ϯ 2.2 Analysis of /lipoproteins Glucose (mg/dl) 96.6 Ϯ 13.0 159.4 Ϯ 51.0 Total cholesterol, HDL cholesterol, and Insulin (␮U/ml)† 21.2 Ϯ 8.0 36.9 Ϯ 26.0 triglyceride concentrations were enzy- HOMA-IR† 4.9 Ϯ 1.6 13.1 Ϯ 7.9 matically determined with a Synchron Adiponectin (␮g/ml)* 5.9 Ϯ 4.4 5.7 Ϯ 4.4 LX-20 (Beckman Coulter, Munich, Ger- Data are means or *medians Ϯ SE. Diabetes was defined by history or by a fasting plasma glucose Ͼ126 many). LDL and VLDL were separated by mg/dl. †Subjects not being treated with insulin. HOMA-IR, HOMA of insulin resistance.

2182 DIABETES CARE, VOLUME 28, NUMBER 9, SEPTEMBER 2005 Schneider and Associates

Table 2—Spearman correlation coefficients with plasma adiponectin and hepatic lipase betic patients as presented in Fig. 1B. activity When sex was entered in the linear model as an independent variable, it predicted ϭ Nondiabetic subjects Diabetic subjects hepatic lipase activity significantly (T Ϫ2.831, P ϭ 0.005). When both adi- rPvalue RPvalue ponectin and sex were included in the analysis, adiponectin remained a signifi- Plasma adiponectin ϳ Age Ϫ0.1 0.2 Ϫ0.2 0.2 cant predictor ( 10%) of hepatic lipase BMI Ϫ2.976 0.004 Ϫ2.3 0.025 activity (Table 3). Total cholesterol 0.08 0.22 0.07 0.4 The inverse correlation between he- LDL cholesterol 0.072 0.3 0.13 0.18 patic lipase activity and HDL cholesterol HDL cholesterol 0.364 0.001 0.4 0.001 and the positive correlation between adi- Triglycerides Ϫ0.194 0.003 Ϫ0.285 0.003 ponectin and HDL cholesterol were tested VLDL cholesterol Ϫ0.196 0.005 Ϫ0.297 0.002 in a multivariate regression analysis with Insulin Ϫ0.322 0.001 Ϫ0.343 0.001 HDL cholesterol as the dependent vari- HOMA Ϫ0.316 0.001 Ϫ0.13 0.9 able and BMI, age, total cholesterol, adi- hs-CRP Ϫ0.219 0.001 Ϫ0.1 0.5 ponectin, insulin, HOMA, and hepatic Atherogenic index Ϫ0.222 0.001 Ϫ0.307 0.001 lipase activity as independent variables. Hepatic lipase This analysis revealed that adiponectin Age Ϫ0.07 0.2 Ϫ0.21 0.1 and BMI were the strongest predictors of BMI 0.072 0.27 Ϫ0.06 0.5 HDL cholesterol in the nondiabetic sub- jects with CAD (adiponectin T ϭ 5.33, Total cholesterol 0.02 0.7 0.07 0.5 Ͻ ϭϪ ϭ LDL cholesterol 0.04 0.5 0.08 0.3 P 0.001; BMI T 2.978, P 0.003, HDL cholesterol Ϫ0.2 0.02 Ϫ0.2 0.04 respectively). Adiponectin alone was the strongest predictor of HDL cholesterol in Triglycerides 0.08 0.2 0.134 0.1 ϭ ϭ VLDL cholesterol 0.07 0.3 0.224 0.02 the diabetic patients (T 3.352, P Insulin 0.125 0.1 0.116 0.2 0.002). When sex was included in the HOMA 0.1 0.2 0.11 0.4 analysis of the diabetic patients, both adi- Ϫ ponectin and sex were the strongest pre- hs-CRP 0.05 0.40 0.07 0.5 ϭ Atherogenic index 0.14 0.03 0.22 0.02 dictors of HDL cholesterol (T 2.505, P Ͻ 0.05 and T ϭ 2.105, P Ͻ 0.05, re- spectively), with hepatic lipase activity betic patients (283.5 Ϯ 112 vs. 220.1 Ϯ cients obtained for hepatic lipase activity being collinear to adiponectin in both co- Ϫ Ϫ 74 nmol ml 1 min 1, P Ͻ 0.05), as are also displayed in Table 2. Hepatic horts (r ϭ 0.34 and 0.23), and to female expected. In the nondiabetic CAD group, lipase activity was found to be signifi- sex in diabetes (r ϭϪ0.26). Finally, adi- the hepatic lipase activity was 272.7 Ϯ cantly inversely associated with HDL cho- ponectin was inversely and hepatic lipase Ϫ Ϫ 110 nmol ml 1 min 1 on average and lesterol in both groups. In both activity positively correlated with the higher in patients with more severe CAD nondiabetic subjects with CAD and type 2 atherogenic index. (severe score Ͼ1) (286.02 Ϯ 108 nmol diabetic subjects, adiponectin levels were Ϫ Ϫ ml 1 min 1). strongly correlated with postheparin CONCLUSIONS — This is the first Plasma adiponectin levels were signif- plasma hepatic lipase activity (Fig. 1). In study showing an independent inverse icantly higher in female compared with multivariate analyses, including markers correlation between adiponectin and he- male diabetic patients (8.26 Ϯ 5.3 vs. of systemic inflammation (hs-CRP), BMI, patic lipase activity in two different pa- 4.23 Ϯ 3.3 ␮g/ml, P Ͻ 0.001). In the age, insulin, and HDL cholesterol as fac- tient cohorts. Our findings may help to nondiabetic subjects with CAD, the tors known to influence hepatic lipase ac- explain observations made in several pre- median plasma adiponectin level was tivity, adiponectin was the only vious studies. These previous studies 5.9 Ϯ 4.4 ␮g/ml. In the presence of CAD independent predictor for hepatic lipase have demonstrated correlations either (severe score Ͼ1), adiponectin levels activity in the nondiabetic patients with between elevated plasma hepatic lipase were significantly lower (4.67 Ϯ 3.5 CAD (Table 3). Regression analyses re- activity (11–13,17) or hypoadiponectine- ␮g/ml [n ϭ 166] vs. 6.39 Ϯ 4.9 ␮g/ml vealed that plasma adiponectin levels ac- mia (8–10,18) with parameters of the in- [n ϭ 40], P ϭ 0.003). counted for 12% of the variation in sulin resistance (metabolic) syndrome, Table 2 shows the association be- hepatic lipase activity in nondiabetic sub- such as obesity, type 2 diabetes, hypertri- tween adiponectin and selected variables jects with CAD. glyceridemia, and low HDL cholesterol. for both nondiabetic subjects with CAD In the male and female patients of the Hepatic lipase is regulated predominantly and diabetic subjects. As expected, we diabetic study cohort analyzed separately, by cell cholesterol content, steroids, and found significant inverse associations be- adiponectin levels were significantly in- thyroid hormones (3). The hepatic lipase tween adiponectin and BMI, plasma tri- versely correlated with plasma hepatic regulatory interactions with insulin are, glycerides, and VLDL cholesterol and a lipase activity (r ϭϪ0.25, P ϭ 0.027 and however, not clear-cut, and hepatic lipase positive correlation with HDL cholesterol r ϭϪ0.37, P Ͻ 0.05, respectively), cor- responsiveness to insulin is controversial. in both cohorts. The correlation coeffi- responding to the correlation for all dia- Most studies in patients with chronic hy-

DIABETES CARE, VOLUME 28, NUMBER 9, SEPTEMBER 2005 2183 Adiponectin and hepatic lipase activity

normal homeostasis in energy metabo- lism (26), is also a major target organ for the action of adiponectin (14). Our data of a significant association of plasma adi- ponectin and postheparin hepatic lipase activity in two independent cohorts are the first evidence to support such a role of adiponectin. Although we did not obtain a direct measurement of fat mass, the ob- servation of an inverse relationship be- tween adiponectin and hepatic lipase activity does complement the findings by Cnop et al. (18). These authors suggested that adiponectin concentrations are de- termined by visceral fat mass and that hy- poadiponectinemia in the setting of central adiposity may lead to increased hepatic lipase activity, which could in turn contribute to low HDL cholesterol levels. The exact mechanism for the inverse correlation between adiponectin and hepatic lipase activity in vivo, however, remains to be studied. Potential explana- tions include decreased hepatic lipase promoter activity mediated by cholesterol depletion of the cell. This hypothesis sug- gests a role for a sterol regulatory element– binding protein (SREBP) in controlling hepatic lipase expression with a modulating role of adiponectin. The fact that an inverse relation between the cell cholesterol content and the levels of hepatic lipase mRNA and activity has been described (3) supports this theory. In addition, incubation with mevinolin, a blocker of cholesterol synthe- sis, induced a stimulation of both hepatic lipase transcripts and activity in vitro, an effect reversed by mevalonate (27). Accord- ingly, Botma et al. (28) very recently re- ported SREBP-mediated inhibition of an Figure 1—A: Linear regression line of hepatic lipase activity as a dependent variable curve with upstream stimulatory factor–stimulated 95% prediction interval in nondiabetic patients. B: Linear regression line of hepatic lipase activity hepatic lipase . Adiponec- as a dependent variable curve with 95% prediction interval in the diabetic patients. tin, on the other hand, has been shown to modulate hepatic lipogenesis by the re- duction of SREBP-1c expression (29), perinsulinemia show higher hepatic patic lipase in obesity and insulin resis- which in turn could reduce transcription lipase activities than in control subjects tance. Recently, it has been speculated of the hepatic lipase gene by the above- (11,13,17,19), whereas others report ei- that adiponectin, an adipokine known to described mechanism. Alternatively, the ther reduced hepatic lipase activity in be related to obesity, insulin resistance, association between adiponectin and he- type 1 diabetic patients (20) or upregu- and even CAD (18,23–25), may be such a patic lipase activity could be mediated lated hepatic lipase activity by insulin factor and that reduced adiponectin con- through the PPAR␥. The recent finding of treatment in general (21,22). Therefore, centrations may mediate plasma hepatic a partial normalization of increased he- hepatic lipase activity appears to be not lipase activity. Support for this hypothesis patic lipase activity in insulin-resistant clearly upregulated by insulin, and the can be derived from the fact that both hy- hamsters by a PPAR␥ agonist (4) supports precise mechanism that links hepatic poadiponectinemia and high hepatic this hypothesis. Adiponectin has been lipase activity to insulin resistance re- lipase activity occur in the setting of in- demonstrated to upregulate PPAR␥ mains unclear. It was suggested that, in- creased abdominal fat mass (18) and that mRNA in adipocytes in vitro (30), and stead of insulin per se, a secondary factor the liver, as the main source of hepatic circulating adiponectin was associated might contribute to the regulation of he- lipase and primary organ maintaining with increased expression of PPAR␥ tran-

2184 DIABETES CARE, VOLUME 28, NUMBER 9, SEPTEMBER 2005 Schneider and Associates

Table 3—Multiple regression analysis result of variables with significant effect on hepatic zone. Diabetes 53:2893–2900, 2004 lipase activity 5. Maeda K, Okubo K, Shimomura I, Fu- nahashi T, Matsuzawa Y, Matsubara K: cDNA cloning and expression of a novel Nondiabetic subjects Diabetic subjects adipose specific collagen-like factor, Independent variable* T ␤ T ␤ apM1 (AdiPose most abundant gene tran- script 1). Biochem Biophys Res Commun Adiponectin Ϫ3.5* Ϫ0.344* Ϫ3.5† Ϫ0.307† 221:286–289, 1996 BMI Ϫ0.314 Ϫ0.27 Ϫ1.49 Ϫ0.134 6. Kern PA, Di Gregorio GB, Lu T, Rassouli HDL 0.726 0.221 Ϫ0.646 Ϫ0.066 N, Ranganathan G: Adiponectin expres- Age Ϫ1.168 Ϫ0.096 Ϫ2.58‡ Ϫ0.227‡ sion from human adipose tissue: relation Insulin 1.1 0.104 0.03 0.003 to obesity, insulin resistance, and tumor necrosis factor-alpha expression. Diabetes Index 1.707 0.089 Ϫ0.273 0.03 Ϫ Ϫ 52:1779–1785, 2003 Sex NA NA 1.85 0.323 7. Ouchi N, Kihara S, Arita Y, Okamoto Y, The dependent variable is hepatic lipase (nanomoles per milliliter per minute). r ϭ 0.347; r2 ϭ 0.121. ␤ is Maeda K, Kuriyama H, Hotta K, Nishida the standardized coefficient, and T represents the estimated coefficient, divided by its own standard error. M, Takahashi M, Muraguchi M, Ohmoto T values below Ϫ2 or above 2 are considered as useful predictors in the model. *P Ͻ 0.01; †P ϭ 0.001; Ͻ Y, Nakamura T, Yamashita S, Funahashi ‡P 0.05. T, Matsuzawa Y: Adiponectin, an adipo- cyte-derived plasma protein, inhibits en- dothelial NF-␬B signaling through a scripts in another study (31). Even re- turn could contribute to a low HDL cho- cAMP-dependent pathway. Circulation duced expression of inflammatory lesterol level in vivo, a critical feature of the 102:1296–1301, 2000 cytokines by the anti-inflammatory action dyslipidemia in the metabolic syndrome. 8. Matsubara M, Maruoka S, Katayose S: De- of adiponectin could play a role, because In summary, we demonstrate for the creased plasma adiponectin concentra- both adiponectin and hepatic lipase are first time that low plasma levels of adi- tions in women with dyslipidemia. J Clin regulated by interleukin-1 on a mRNA ponectin are significantly correlated with Endocrinol Metab 87:2764–2769, 2002 9. Arita Y, Kihara S, Ouchi N, Takahashi M, level (32), (33). Finally, a direct effect of elevated hepatic lipase activity in two in- Maeda K, Miyagawa J, Hotta K, Shimo- adiponectin on hepatic lipase transcrip- dependent cohorts, nondiabetic subjects mura I, Nakamura T, Miyaoka K, tion is possible, mediated by the direct with CAD and patients with type 2 diabe- Kuriyama H, Nishida M, Yamashita S, action of adiponectin on its receptor in tes. This relationship is independent of Okubo K, Matsubara K, Muraguchi M, liver (14) by a so far unknown mecha- systemic inflammatory or metabolic Ohmoto Y, Funahashi T, Matsuzawa Y: nism. These speculations about the exact markers and may help to explain the Paradoxical decrease of an adipose-spe- mechanism linking adiponectin with he- HDL-elevating action of adiponectin. cific protein, adiponectin, in obesity. Bio- patic lipase merit further investigations. chem Biophys Res Commun 257:79–83, The importance of HDL cholesterol in its 1999 meaning as a prognostic factor for the in- Acknowledgments— This work was sup- 10. Hotta K, Funahashi T, Bodkin NL, Ort- meyer HK, Arita Y, Hansen BC, Matsu- dividual atherosclerotic risk is indisput- ported by research grants from Pfizer (CT0981-00SP08 and ATV-D-01-006G). J.S. zawa Y: Circulating concentrations of the able (34). Although we did not compare was supported by the National Institutes of adipocyte protein adiponectin are de- our two study groups, the data obtained Health Grant 5T32DK07296-23. creased in parallel with reduced insulin from our study correspond to previous sensitivity during the progression to type reports suggesting that adiponectin levels 2 diabetes in rhesus monkeys. Diabetes are lower in diabetic male versus female References 50:1126–1133, 2001 subjects and lower in the presence of CAD 1. Zambon A, Austin MA, Brown BG, Ho- 11. Laakso M, Sarlund H, Ehnholm C, Vouti- (10,25). Low plasma adiponectin has also kanson JE, Brunzell JD: Effect of hepatic lainen E, Aro A, Pyorala K: Relationship been shown to be associated with low lipase on LDL in normal men and those between postheparin plasma and HDL cholesterol levels (8,35), but the with coronary artery disease. Arterioscler high-density cholesterol in Thromb 13:147–153, 1993 different types of diabetes. Diabetologia mechanism by which adiponectin influ- 2. Cohen JC, Vega GL, Grundy SM: Hepatic 30:703–706, 1987 ences plasma lipoproteins has not been lipase: new insights from genetic and met- 12. Deeb SS, Zambon A, Carr MC, Ayyobi AF, elucidated. Hepatic lipase is a major de- abolic studies. Curr Opin Lipidol 10:259– Brunzell JD: Hepatic lipase and dyslipide- terminant of HDL cholesterol concentra- 267, 1999 mia: interactions among genetic variants, tion, converting HDL subclasses by 3. Perret B, Mabile L, Martinez L, Terce F, obesity, gender, and diet. J Lipid Res 44: hydrolyzing phospholipids and triglycer- Barbaras R, Collet X: Hepatic lipase: struc- 1279–1286, 2003 ides (36,37) and/or mediating the uptake ture/function relationship, synthesis, and 13. Sibley SD, Palmer JP, Hirsch IB, Brunzell of HDL cholesterol directly or indirectly regulation. J Lipid Res 43:1163–1169, JD: Visceral obesity, hepatic lipase activ- by lipoprotein receptors or scavenger re- 2002 ity, and dyslipidemia in type 1 diabetes. ceptors (38,39). Because of these func- 4. Lewis GF, Murdoch S, Uffelman K, Na- J Clin Endocrinol Metab 88:3379–3384, ples M, Szeto L, Albers A, Adeli K, Brun- 2003 tions, hepatic lipase activity is inversely zell JD: Hepatic lipase mRNA, protein, 14. Yamauchi T, Kamon J, Ito Y, Tsuchida A, correlated with HDL cholesterol in and plasma enzyme activity is increased in Yokomizo T, Kita S, Sugiyama T, Miya- plasma. Our findings demonstrate that the insulin-resistant, fructose-fed Syrian gishi M, Hara K, Tsunoda M, Murakami hypoadiponectinemia is associated with golden hamster and is partially normal- K, Ohteki T, Uchida S, Takekawa S, Waki increased hepatic lipase activity, which in ized by the insulin sensitizer rosiglita- H, Tsuno NH, Shibata Y, Terauchi Y,

DIABETES CARE, VOLUME 28, NUMBER 9, SEPTEMBER 2005 2185 Adiponectin and hepatic lipase activity

Froguel P, Tobe K, Koyasu S, Taira K, K, Matsuzawa Y, Pratley RE, Tataranni Cabo R, Lane MA, Roth GS, Ingram DK: Kitamura T, Shimizu T, Nagai R, Kad- PA: Hypoadiponectinemia in obesity and Circulating adiponectin levels increase in owaki T: Cloning of adiponectin recep- type 2 diabetes: close association with in- rats on caloric restriction: the potential for tors that mediate antidiabetic metabolic sulin resistance and hyperinsulinemia. insulin sensitization. Exp Gerontol 39: effects. Nature 423:762–769, 2003 J Clin Endocrinol Metab 86:1930–1935, 1049–1059, 2004 15. Expert Committee on the Diagnosis and 2001 32. Feingold KR, Memon RA, Moser AH, Shi- Classification of Diabetes Mellitus: Report 24. Kumada M, Kihara S, Sumitsuji S, genaga JK, Grunfeld C: Endotoxin and in- of the Expert Committee on the Diagnosis Kawamoto T, Matsumoto S, Ouchi N, terleukin-1 decrease hepatic lipase mRNA and Classification of Diabetes Mellitus. Arita Y, Okamoto Y, Shimomura I, levels. 142:379–387, 1999 Diabetes Care 20:1183–1197, 1997 Hiraoka H, Nakamura T, Funahashi T, 33. Lihn AS, Bruun JM, He G, Pedersen SB, 16. Iverius PH, Brunzell JD: Human adipose Matsuzawa Y: Association of hypoadi- Jensen PF, Richelsen B: Lower expression tissue : changes with ponectinemia with coronary artery dis- of adiponectin mRNA in visceral adipose feeding and relation to postheparin ease in men. Arterioscler Thromb Vasc Biol tissue in lean and obese subjects. Mol Cell plasma enzyme. Am J Physiol 249:E107– 23:85–89, 2003 Endocrinol 219:9–15, 2004 E114, 1985 25. Hotta K, Funahashi T, Arita Y, Takahashi 34. Boden WE: High-density lipoprotein cho- 17. Cominacini L, Garbin U, Davoli A, Cam- M, Matsuda M, Okamoto Y, Iwahashi H, lesterol as an independent risk factor in pagnola M, De Santis A, Pasini C, Pas- Kuriyama H, Ouchi N, Maeda K, Nishida cardiovascular disease: assessing the data torino AM, Bosello O: High-density M, Kihara S, Sakai N, Nakajima T, Hase- from Framingham to the Veterans Affairs lipoprotein cholesterol concentrations gawa K, Muraguchi M, Ohmoto Y, Naka- High-Density Lipoprotein Intervention and postheparin hepatic and lipoprotein mura T, Yamashita S, Hanafusa T, Trial. Am J Cardiol 86:19L–22L, 2000 lipases in obesity: relationships with Matsuzawa Y: Plasma concentrations of a 35. von Eynatten M, Schneider JG, Humpert plasma insulin levels. Ann Nutr Metab 37: novel, adipose-specific protein, adi- PM, Rudofsky G, Schmidt N, Barosch P, 175–184, 1993 ponectin, in type 2 diabetic patients. Ar- Hamann A, Morcos M, Kreuzer J, Bier- 18. Cnop M, Havel PJ, Utzschneider KM, Carr terioscler Thromb Vasc Biol 20:1595–1599, haus A, Nawroth PP, Dugi KA: Decreased DB, Sinha MK, Boyko EJ, Retzlaff BM, 2000 plasma lipoprotein lipase in hypoadi- Knopp RH, Brunzell JD, Kahn SE: Rela- 26. Semenkovich CF: Fatty acid metabolism ponectinemia: an association indepen- tionship of adiponectin to body fat distri- and vascular disease. Trends Cardiovasc dent of systemic inflammation and insulin bution, insulin sensitivity and plasma Med 14:72–76, 2004 resistance. Diabetes Care 27:2925–2929, lipoproteins: evidence for independent 27. Busch SJ, Barnhart RL, Martin GA, Flana- 2004 roles of age and sex. Diabetologia 46:459– gan MA, Jackson RL: Differential regula- 36. Zambon A, Deeb SS, Hokanson JE, Brown 469, 2003 tion of hepatic triglyceride lipase and 19. Purnell JQ, Kahn SE, Albers JJ, Nevin DN, 3-hydroxy-3-methylglutaryl-CoA reduc- BG, Brunzell JD: Common variants in the Brunzell JD, Schwartz RS: Effect of weight tase gene expression in a human hepa- promoter of the hepatic lipase gene are loss with reduction of intra-abdominal fat toma cell line, HepG2. J Biol Chem 265: associated with lower levels of hepatic on lipid metabolism in older men. J Clin 22474–22479, 1990 lipase activity, buoyant LDL, and higher Endocrinol Metab 85:977–982, 2000 28. Botma GJ, van Deursen D, Vieira D, van HDL2 cholesterol. Arterioscler Thromb 20. Caixas A, Perez A, Payes A, Otal C, Car- Hoek M, Jansen H, Verhoeven AJM: Ste- Vasc Biol 18:1723–1729, 1998 reras G, Ordonez-Llanos J, Reviriego J, rol-regulatory-element binding protein 37. Cheung MC, Sibley SD, Palmer JP, Oram Anderson JH, de Leiva A: Effects of a inhibits upstream stimulatory factor- JF, Brunzell JD: Lipoprotein lipase and short-acting insulin analog (insulin lis- stimulated hepatic lipase gene expression. hepatic lipase: their relationship with pro) versus regular insulin on lipid me- Atherosclerosis 179:61–67, 2005 HDL subspecies Lp(A-I) and Lp(A-I,A-II). tabolism in insulin-dependent diabetes 29. Shklyaev S, Aslanidi G, Tennant M, Prima J Lipid Res 44:1552–1558, 2003 mellitus. Metabolism 47:371–376, 1998 V, Kohlbrenner E, Kroutov V, Campbell- 38. Lambert G, Chase MB, Dugi K, Bensa- 21. Baynes C, Henderson AD, Richmond W, Thompson M, Crawford J, Shek EW, doun A, Brewer HB Jr, Santamarina-Fojo Johnston DG, Elkeles RS: The response of Scarpace PJ, Zolotukhin S: Sustained pe- S: Hepatic lipase promotes the selective hepatic lipase and serum lipoproteins to ripheral expression of transgene adi- uptake of high density lipoprotein-cho- acute hyperinsulinaemia in type 2 diabe- ponectin offsets the development of diet- lesteryl esters via the scavenger receptor tes. Eur J Clin Invest 22:341–346, 1992 induced obesity in rats. Proc Natl Acad Sci B1. J Lipid Res 40:1294–1303, 1999 22. Ruotolo G, Parlavecchia M, Taskinen MR, USA100:14217–14222, 2003 39. Collet X, Tall AR, Serajuddin H, Guen- Galimberti G, Zoppo A, Le NA, Ragogna 30. Ajuwon KM, Spurlock ME: Adiponectin douzi K, Royer L, Oliveira H, Barbaras R, F, Micossi P, Pozza G: Normalization of inhibits LPS-induced NF␬B activation Jiang XC, Francone OL: Remodeling of lipoprotein composition by intraperito- and IL-6 production, and increases HDL by CETP in vivo and by CETP and neal insulin in IDDM: role of increased PPAR␥2 expression in adipocytes. Am J hepatic lipase in vitro results in enhanced hepatic lipase activity. Diabetes Care 17: Physiol Regul Integr Comp Physiol 288: uptake of HDL CE by cells expressing 6–12, 1994 1220–1225, 2005 scavenger receptor B-I. J Lipid Res 23. Weyer C, Funahashi T, Tanaka S, Hotta 31. Zhu M, Miura J, Lu LX, Bernier M, De- 40:1185–1193, 1999

2186 DIABETES CARE, VOLUME 28, NUMBER 9, SEPTEMBER 2005