J C E M
O N L I N E
Hot Topics in Translational Endocrinology—Endocrine Research
Circulating Angiopoietin-like Protein 8 Is Independently Associated With Fasting Plasma Glucose and Type 2 Diabetes Mellitus T. Ebert,* S. Kralisch,* A. Hoffmann, A. Bachmann, U. Lössner, J. Kratzsch, M. Blüher, M. Stumvoll, A. Tönjes,† and M. Fasshauer† Department of Endocrinology and Nephrology (T.E., S.K., A.H., A.B., U.L., M.B., M.S., A.T., M.F.) and Institute of Laboratory Medicine (J.K.), University of Leipzig, and Integrated Research and Treatment Center Adiposity Diseases (T.E., S.K., U.L., A.T., M.F.), Leipzig University Medical Center, 04103 Leipzig, Germany
Objective: Angiopoietin-like protein 8 (Angptl8) has recently been introduced as a novel adipokine/hepatokine that promotes pancreatic -cell proliferation and improves glucose tolerance in mouse models of insulin resistance. However, regulation of Angptl8 in human type 2 diabetes mellitus (T2DM) and renal dysfunction has not been determined. Research Design and Methods: Serum Angptl8 levels were quantified by ELISA in 62 patients with T2DM as compared with 58 nondiabetic subjects in vivo. Within both groups, about half of the patients were on chronic hemodialysis or had an estimated glomerular filtration rate above 50 mL/min/1.73 m2. Furthermore, we investigated the effect of insulin and differentiation on Angptl8 mRNA expression in 3T3-L1 adipocytes in vitro. Results: Median [interquartile range] serum Angptl8 levels were higher in patients with T2DM (1.19 [0.37] g/L) as compared with nondiabetic subjects (1.03 [0.47] g/L) (P ⫽ .005). Furthermore, the adipokine/hepatokine was significantly higher in women (1.21 [0.47] g/L) as compared with men (1.05 [0.44] g/L]) (P ⫽ .013). In multivariate analysis, fasting glucose and T2DM but not renal function remained independent and positive predictors of circulating Angptl8 even after adjustment for markers of obesity, lipid status, and inflammation (P ⬍ .05). Furthermore, Angptl8 mRNA expression was induced by insulin and during adipogenesis in 3T3-L1 adipocytes in vitro. Conclusions: Circulating Angptl8 is positively and independently associated with T2DM and fasting glucose in vivo. Furthermore, Angptl8 mRNA expression is induced by insulin and during adipogenesis in 3T3-L1 adipocytes in vitro. (J Clin Endocrinol Metab 99: E2510 –E2517, 2014)
he metabolic syndrome and its phenotype including visceral obesity, insulin resistance, type 2 diabetes mellitus (T2DM), hypertension, and dyslipidemia is an increasing global health burden. In the last 2 decades, various metabolic hormones have been shown to significantly influence obesity and associated complications. Among those, adipocyte- and hepatocyte-secreted proteins, socalled adipokines and hepatokines, provide a link between
T
obesity, the metabolic syndrome, and associated cardiovascular diseases. Thus, the adipokine leptin is an appetite-suppressive adipokine upregulated in obesity and cardiovascular disease in various animal models (1, 2). Moreover, the adipocyte- and hepatocyte-derived cytokine fibroblast growth factor 21 (FGF21) stimulates glucose uptake in adipocytes (3) and increases insulin sensitivity in rodents (4).
ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2014 by the Endocrine Society Received December 7, 2013. Accepted September 4, 2014. First Published Online October 17, 2014
* T.E. and S.K. equally contributed to this work. † A.T. and M.F. equally contributed to this work. Abbreviations: Angptl8, angiopoietin-like protein 8; BMI, body mass index; CD, chronic hemodialysis; eGFR, estimated glomerular filtration rate; FG, fasting glucose; FGF21, fibroblast growth factor 21; FI, fasting insulin; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance; hsIL-6, high-sensitivity IL-6; LDL, low-density lipoprotein; oGTT, oral glucose tolerance test; T2DM, type 2 diabetes mellitus; TG, triglyceride.
E2510
jcem.endojournals.org
J Clin Endocrinol Metab, December 2014, 99(12):E2510 –E2517
doi: 10.1210/jc.2013-4349
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2013-4349
Recently, angiopoietin-like protein 8 (Angptl8; also known as betatrophin, hepatocellular carcinoma-associated protein TD26, or lipasin) has been introduced as a novel adipokine/hepatokine that significantly and specifically promotes pancreatic -cell proliferation, expands -cell mass, and improves glucose tolerance in mouse models of insulin resistance (5). The authors suggested that Angptl8 expression is a compensatory mechanism to increase -cell proliferation in insulin resistance (5). Interestingly, serum levels of this adipokine/hepatokine are 2-fold increased in patients with T1DM as compared with age-matched healthy controls in a very recent study (6). Taking these interesting and novel findings into consideration, Angptl8 appears as a novel adipokine/hepatokine that improves glucose tolerance. In contrast to these elegant studies in animals (5) and patients with T1DM (6), regulation of Angptl8 in human T2DM and insulin resistance has not been investigated until now. Furthermore, no study has evaluated potential renal elimination and regulation of this beneficial adipokine/hepatokine during renal failure in humans so far. To address these issues, we determined circulating Angptl8 concentrations in 62 well-phenotyped patients with T2DM as compared with 58 nondiabetic subjects in vivo. Within both groups, about half of the patients were on chronic hemodialysis (CD) or had an estimated glomerular filtration rate (eGFR) above 50 mL/min/1.73 m2. Furthermore, we correlated Angptl8 to clinical and biochemical measures of renal function, indices of glucose metabolism, and lipid metabolism as well as inflammation. Moreover, we investigated the effects of hormones inducing insulin resistance, the insulin-sensitizing peroxisome proliferatoractivated receptor-␥ agonist troglitazone, as well as differentiation, on Angptl8 mRNA expression in 3T3-L1 adipocytes in vitro. Our primary hypotheses were that Angptl8 concentrations are increased in T2DM as compared with nondiabetic subjects and that adipocyte mRNA expression of this novel adipokine/hepatokine is induced by insulin in vitro. Furthermore, we hypothesized that Angptl8 serum levels would increase in patients on CD, indicating that Angptl8 is eliminated by the kidneys.
Subjects and Methods Angptl8 in T2DM and renal dysfunction as well as regulation by glucose in vivo Subjects For the cross-sectional study on Angptl8 in T2DM and renal dysfunction, 120 Caucasian men (n ⫽ 62) and women (n ⫽ 58) were recruited by the Department of Endocrinology and Ne-
jcem.endojournals.org
E2511
phrology, University of Leipzig. The study design has recently been described (7–10). In brief, 60 patients had an eGFR ⬎50 mL/min/1.73 m2 (controls) according to the Modification of Diet in Renal Disease formula (11), whereas 60 patients were on CD. T2DM in controls and patients on CD was defined as fasting blood glucose ⱖ126 mg/dL or use of insulin or oral hypoglycemic medications according to Ref. 12. Using these criteria, 30 of the 60 control and 32 of the 60 CD patients presented with T2DM. Body mass index (BMI) was calculated as weight divided by height squared. Waist-to-hip ratio was determined after waist and hip circumferences were assessed. Age ranged from 32 to 85 years and BMI from 18.7 to 46.1 kg/m2. Homeostasis model assessment of insulin resistance (HOMA-IR) was calculated as previously described (13). Patients with severe conditions including generalized inflammation or end-stage malignant diseases were excluded from the study. To better define dynamic regulation of Angptl8 by glucose, concentrations of the novel adipokine/hepatokine were measured in a separate cohort comprising 30 subjects with normal glucose tolerance during a 75-g oral glucose tolerance test (oGTT). The study design has recently been described (14). Both studies were approved by the local ethics committee, and all subjects gave written informed consent before taking part in the respective study.
Assays All blood samples were taken after a fasting period of at least 8 hours. In CD patients, blood was obtained just before hemodialysis started. Serum Angptl8 (Phoenix Pharmaceuticals), leptin (Mediagnost), and high-sensitivity IL-6 (hsIL-6) (R&D Systems) were determined by ELISA according to the manufacturers’ instructions. According to the manufacturer, the sensitivity of the Angptl8 ELISA was 0.01 g/L. The degree of precision of the ELISA system in terms of intra-assay and interassay coefficient of variance (percentage) was less than 7.5% and 10.5%, respectively. Spike recovery and linearity were in a range of 97% to 131% and 0.3 to 4.4 g/L, respectively. Furthermore, the ELISA was specific for human Angptl8 and did not crossreact with murine and rat Angptl8. Moreover, the antibody used in the ELISA kit specifically detected Angtpl8 at the expected size of about 22.5 kDa when unextracted and extracted human plasma samples were probed by Western blot analysis. In addition, quality controls were included in all ELISA measurements, and results were within the expected range. Furthermore, the adipokine/hepatokine was recently quantified by our group in a second cohort of patients with the same Angptl8 ELISA kit lot without any noticeable problems (15). Serum creatinine, fasting glucose (FG), fasting insulin (FI), and triglycerides (TGs) as well as total, high-density lipoprotein (HDL), and low-density lipoprotein (LDL) cholesterol, were measured in a certified laboratory by standard methods.
Effects of hormones, troglitazone, and differentiation on Angptl8 expression in 3T3-L1 cells in vitro Reagents and cell culture Isobutylmethylxanthine, IL-6, and dexamethasone were obtained from Calbiochem; insulin from Roche Molecular Biochemicals; and IL-1, GH, and troglitazone from Sigma-Aldrich. Cell culture reagents were purchased from GE Healthcare and
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
E2512
Ebert et al
Angptl8 and Type 2 Diabetes Mellitus
Life Technologies and oligonucleotides from MWG-Biotech. 3T3-L1 cells (American Type Culture Collection) were grown and differentiated into adipocytes as previously described (16). After 6 hours serum starvation, differentiated 3T3-L1 adipocytes were cultured in the presence or absence of insulin, IL-6, IL1, GH, and troglitazone under conditions further described in Figure 2. Concentrations of the respective hormones and troglitazone were used according to prior studies (17–19). Furthermore, Angptl8 expression was assessed during differentiation of 3T3-L1 cells.
Analysis of Angptl8 mRNA production Angptl8 mRNA synthesis was determined by quantitative real-time RT-PCR. In more detail, 1 g total RNA was reverse transcribed in a 20-L reverse transcription reaction using standard reagents (Life Technologies) and 1 L of each reverse transcription reaction was quantified on a LightCycler 480 real-time PCR 96-well thermocycler using LightCycler 480 Probes Master Mix (Roche Diagnostics GmbH) essentially as described (20). The following primer pairs were used: Angptl8, CACCAGCCTGTCGGAGATTC (sense) and GTCTCTGCTGGATCTGTCGC (antisense); and m36B4, AAGCGCGTCCTGGCATTGTCT (sense) and CCGCAGGGGCAGCAGTGGT (antisense). Angptl8 expression was calculated relative to m36B4, which was used as an internal control due to its resistance to hormonal regulation.
Statistical analysis A P value ⬍ .05 was considered as statistically significant in all analyses. SPSS software version 20.0 (IBM) was used for all statistical analyses of human in vivo data. Overall group differences between the 4 subgroups (control/T2DM⫺, control/T2DM⫹, CD/ T2DM⫺, and CD/T2DM⫹) were assessed by nonparametric Kruskal-Wallis test with Bonferroni post hoc analysis. Differences between nominal variables, eg, T2DM status, were assessed by nonparametric Mann-Whitney U test. Furthermore, a two-way ANOVA was conducted to further examine the effects of T2DM and CD status on circulating Angptl8. Univariate correlations were analyzed by nonparametric Spearman’s rank correlation method. Furthermore, multivariate linear regression analyses were performed. Before multivariate analysis and twoway ANOVA, distribution of continuous variables was tested for normality using the Shapiro-Wilk test and nonnormally distributed parameters were logarithmically transformed. In multivariate linear analysis, only parameters that correlated significantly with Angptl8 in univariate analysis were included. For covariates, eg, creatinine and eGFR, the parameter with the strongest univariate correlation was included in the multivariate model. Furthermore, gender was included in multivariate analysis. For in vitro experiments, GraphPad Prism version 6 (GraphPad Software) was used. Here, Angptl8 expression adjusted for m36B4 after hormone and troglitazone treatment or during differentiation was analyzed relative to untreated control cells and preadipocytes, respectively. Results are shown as mean ⫾ SEM. Differences between treatments were analyzed by unpaired Student’s t tests with prior logarithmic transformation.
J Clin Endocrinol Metab, December 2014, 99(12):E2510 –E2517
Results Angptl8 in T2DM and renal dysfunction in vivo Baseline characteristics of the total sample Table 1 summarizes the clinical characteristics of the 4 subgroups (control/T2DM⫺, control/T2DM⫹, CD/ T2DM⫺, and CD/T2DM⫹) studied. In the total sample, the median (interquartile range) serum Angptl8 level was 1.14 (0.42) g/L. Circulating Angptl8 was significantly different between the 4 subgroups (P ⬍ .001) (Table 1). Furthermore, Angptl8 concentrations were significantly higher in women (1.21 [0.47] g/L) as compared with men (1.05 [0.44] g/L]) (P ⫽ .013) in all patients. Notably, there was a significant interaction between T2DM and CD status on Angptl8 concentrations (F1,116 ⫽ 6.149, P ⫽ .015) in two-way ANOVA (Table 1). Angptl8 in T2DM When investigating circulating Angptl8 in relation to T2DM status, patients with T2DM had significantly higher Angptl8 concentrations (1.19 [0.37] g/L) as compared with nondiabetic subjects (1.03 [0.47] g/L) (P ⫽ .005) (Figure 1). Interestingly, Angptl8 serum levels remained significantly higher in patients with T2DM as compared with nondiabetic subjects even after adjustment for age, gender, and CD status (P ⬍ .001). Furthermore, the main effect of T2DM on Angptl8 was significant (F1,116 ⫽ 10.460, P ⫽ .002) in two-way ANOVA (Table 1). When investigating Angptl8 levels in nondiabetic control subjects not on CD in a separate analysis, median levels of the adipokine/hepatokine were not significantly different in subjects with normal glucose tolerance as compared with subjects with prediabetes as defined by a 75-g oGTT (data not shown). Angptl8 in CD When investigating circulating Angptl8 in relation to CD, Angptl8 serum levels were significantly lower in patients on CD (1.06 [0.40] g/L) as compared with subjects with an eGFR ⬎50 mL/min/1.73 m2 (1.23 [0.45] g/L) (P ⫽ .014). Interestingly, the main effect of CD status on circulating Angptl8 did also reach statistical significance (F1,116 ⫽ 7.465, P ⫽ .007) in two-way ANOVA (Table 1). Univariate correlations in the total sample Serum Angptl8 concentrations in all individuals were positively and significantly correlated with BMI, FG, FI, HOMA-IR, HDL cholesterol, eGFR, and leptin (P ⬍ .05) (Table 2). In contrast, Angptl8 was significantly and negatively correlated with age, serum creatinine, and hsIL-6 (P ⬍ .05) (Table 2).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2013-4349
Table 1.
jcem.endojournals.org
E2513
Baseline Characteristics of the Study Populationa Two-way ANOVA
n Angptl8, g/L Age, yd Gender (M/F) BMI, kg/m2d WHR SBP, mm Hg DBP, mm Hgd FG, mmol/Ld FI, pmol/Ld HOMA-IRd Cholesterol, mmol/Ld HDL cholesterol, mmol/Ld LDL cholesterol, mmol/Ld TG, mmol/Ld FFA, mmol/Ld Creatinine, mol/Ld GFR, mL/min/1.73 m2d Leptin, g/Ld hsIL-6, ng/Ld
Control/T2DMⴚ
Control/T2DMⴙ
CD/T2DMⴚ
CD/T2DMⴙ
30 1.06 (0.49) 63 (19) 11/19 28.2 (5.6) 0.88 (0.12) 125 (21) 77 (10) 5.1 (1.3) 45.1 (33.3) 1.3 (1.2) 5.3 (0.9) 1.4 (0.4) 3.5 (1.2) 1.1 (0.8) 0.5 (0.2) 76 (17) 77 (21) 16.9 (27.9) 1.87 (2.11)
30 1.34 (0.50)b 63 (16) 16/14 29.1 (5.2) 0.94 (0.10) 126 (20) 73 (15) 7.6 (3.2)b 47.9 (62.6) 2.7 (2.9) 4.9 (1.5) 1.2 (0.5) 2.9 (0.9)b 1.4 (0.9) 0.6 (0.4) 72 (22) 84 (35) 16.7 (26.4) 2.07 (1.25)
28 1.01 (0.42)c 59 (22) 15/13 25.2 (6.5)c 0.96 (0.18) 125 (38) 77 (20) 4.6 (1.2)c 28.2 (47.6) 0.8 (1.3)c 4.4 (1.2)b 1.0 (0.5)b 2.7 (0.9)b 1.6 (0.9)b 0.6 (0.5) 829 (431)b,c 6 (3)b,c 11.8 (37.9) 6.14 (4.91)b,c
32 1.11 (0.34)c 68 (14) 20/12 27.9 (6.6) 1.00 (0.14)b 120 (25) 70 (18) 5.2 (3.3)c 50.1 (91.6) 1.4 (3.1) 4.2 (1.3)b 1.0 (0.3)b,c 2.1 (1.4)b 1.8 (1.4)b 0.7 (0.5) 717 (221)b,c 7 (3)b,c 32.5 (67.5) 7.48 (7.30)b,c
PT2DM
PCD
PT2DMⴛCD
.002 .018
.007 .590
.015 .137
.092 ⬍.001 .586 .114 ⬍.001 .117 .004 .031 .093 ⬍.001 .003 .104 .563 .354 .127 .361
.001 .003 .178 .060 ⬍.001 .554 .147 ⬍.001 ⬍.001 ⬍.001 ⬍.001 .830 ⬍.001 ⬍.001 .266 ⬍.001
.428 .924 .569 .967 .154 .678 .992 .376 .807 .644 .837 .660 .960 .691 .070 .214
Abbreviations: DBP, diastolic blood pressure; FFA, free fatty acids; SBP, systolic blood pressure; WHR, waist-to-hip ratio. a Values for median (interquartile range) are shown. Parameters were analyzed by Kruskal-Wallis test followed by Bonferroni post hoc analysis. Exact P values for the effects of T2DM (PT2DM) and CD (PCD), as well as for interaction between T2DM and CD (PT2DM⫻CD), on the respective parameters were assessed by two-way ANOVA. b
P ⬍ 0.05 compared with control/T2DM⫺.
c
P ⬍ 0.05 compared with control/T2DM⫹.
d
Nonnormally distributed variables.
Multivariate regression analysis in the total sample Multiple linear regression analysis revealed that FG remained independently and positively associated with circulating Angptl8 after adjustment for age, gender, BMI,
HDL cholesterol, creatinine, and hsIL-6 (P ⬍ .001) (Table 3). Furthermore, gender (gender coefficients were coded such that females have a larger value as compared with males), BMI, and HDL cholesterol were positive and inTable 2.
Univariate Correlations With Serum Angptl8
b
Age, y BMI, kg/m2b WHR SBP, mm Hg DBP, mm Hgb FG, mmol/Lb FI, pmol/Lb HOMA-IRb Cholesterol, mmol/Lb HDL cholesterol, mmol/Lb LDL cholesterol, mmol/Lb TG, mmol/Lb FFA, mmol/Lb Creatinine, mol/Lb eGFR, mL/min/1.73 m2b Leptin, g/Lb hsIL-6, ng/Lb
Figure 1. Box-and-whisker plot of Angptl8 serum levels in patients with T2DM (n ⫽ 62) as compared with nondiabetic subjects (n ⫽ 58). P value was assessed by nonparametric Mann-Whitney U test.
r
P
⫺0.208 0.251 ⫺0.175 0.016 0.127 0.291 0.182 0.256 0.105 0.302 ⫺0.016 ⫺0.014 0.057 ⫺0.232 0.187 0.216 ⫺0.184
.023a .006a .055 .866 .166 .001a .047a .005a .254 .001a .867 .880 .533 .011a .041a .018a .044a
Abbreviations: DBP, diastolic blood pressure; FFA, free fatty acids; SBP, systolic blood pressure; WHR, waist-to-hip ratio. a
significant correlation as assessed by Spearman’s correlation method.
b
Nonnormally distributed variables.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
E2514
Ebert et al
Angptl8 and Type 2 Diabetes Mellitus
Table 3. Multivariate Regression Analysis With Angptl8 as Dependent Variablea Independent Variables c
Age Gender BMIc FGc HDL cholesterolc Creatininec hsIL-6c

P
⫺0.221 .194 .165 .397 .287 .111 ⫺.009
.006b .023b .049b ⬍.001b .002b .345 .935
a Multivariate regression analysis between Angptl8 (dependent variable) and the independent variables shown. Standardized coefficients and P values are given. Gender coefficients were coded such that females have a larger value compared with males. b
Significant correlation.
c
Nonnormally distributed variables were logarithmically transformed before testing.
dependent predictors of Angptl8 serum concentrations, whereas age was independently and negatively associated with the adipokine/hepatokine (P ⬍ .05) (Table 3). Interestingly, T2DM status remained significantly and positively associated with circulating Angptl8 when included in the multivariate model instead of FG ( ⫽ .374; P ⬍ .001). Univariate correlations in the subgroups Of the associations remaining significant in multivariate analyses in the total sample (n ⫽ 120; Table 3), Angptl8 was significantly associated with FG and HDL cholesterol in subjects not on CD, with age in non-T2DM, and with age, BMI, FG, and HDL cholesterol in T2DM in univariate analyses (data not shown). In contrast, Angptl8 was not significantly correlated with any of these parameters when CD patients were analyzed separately by univariate analysis (data not shown). When further segregating the study cohort into 4 subgroups (control/T2DM⫺, control/ T2DM⫹, CD/T2DM⫺, and CD/T2DM⫹), HDL cholesterol was positively correlated with Angptl8 in control/ T2DM⫹ and age was negatively associated with Angptl8 in CD/T2DM⫺ in univariate analyses (data not shown). Regulation of Angptl8 by glucose in vivo To determine the influence of glucose on Angptl8 levels in more detail, dynamic regulation of circulating Angptl8 was elucidated in a cohort comprising 30 subjects with normal glucose tolerance during an oGTT. Here, mean ⫾ SE Angptl8 was not significantly different at 30 minutes (0.52 ⫾ 0.17 g/L) and 120 minutes (0.63 ⫾ 0.21 g/L) after glucose exposure as compared with fasting baseline levels at 0 minutes (0.61 ⫾ 0.15 g/L) (gender-adjusted P for overall differences ⫽ .370) (Supplemental Figure 1).
J Clin Endocrinol Metab, December 2014, 99(12):E2510 –E2517
Effects of hormones, troglitazone, and differentiation on Angptl8 expression in 3T3-L1 cells in vitro Treatment of differentiated 3T3-L1 adipocytes with insulin for 16 hours stimulated Angptl8 expression in a dosedependent manner (all P ⬍ .01) (Figure 2A). Here, a significant 8-fold induction of Angptl8 mRNA synthesis was detectable at insulin concentrations as low as 1nM, and insulin increased Angptl8 expression up to 18.5-fold at 3.3nM (P ⬍ .01) (Figure 2A). Furthermore, 100nM insulin upregulated Angptl8 expression also in a time-dependent fashion with significant 30-fold stimulation detectable as early as 2 hours after insulin addition (P ⬍ .05) (Figure 2B). Moreover, maximal 76-fold induction was seen after 8 hours of insulin treatment (P ⬍ .01) (Figure 2B). Angptl8 mRNA production also significantly increased in a differentiation-dependent manner with 50-fold and 145-fold induction of the adipokine/hepatokine at days 6 and 9 of differentiation, respectively, as compared with undifferentiated controls (day 0, P ⬍ .05) (Figure 2C). In contrast, chronic treatment of differentiated 3T3-L1 adipocytes with IL-6, IL-1, GH, and troglitazone did not significantly influence Angptl8 expression (Figure 2D).
Discussion In the current study, we show that patients with T2DM have significantly higher Angptl8 serum levels as compared with nondiabetic subjects. Furthermore, T2DM and FG remain independent and positive predictors of circulating Angptl8 in multivariate analysis. Moreover, we demonstrate that insulin is a direct dose- and time-dependent stimulator of Angptl8 mRNA expression in differentiated 3T3-L1 adipocytes in vitro similar to convincing results presented by Ren and co-workers (21). In agreement with our findings, increased expression of the adipokine/hepatokine has also been shown in animal models of insulin resistance and T2DM. Thus, Angptl8 mRNA is upregulated 3- to 6-fold in the liver of S961treated animals and ob/ob and db/db mice (5). Because insulin significantly induces Angptl8 mRNA expression in differentiated 3T3-L1 adipocytes in vitro, it is tempting to speculate that hyperinsulinemia contributes to upregulation of the adipokine/hepatokine in T2DM. Notably, circulating Angptl8 is not affected by acute changes of glucose homeostasis during a 75-g oGTT in a cohort comprising of 30 subjects with normal glucose tolerance (Supplemental Figure 1). Furthermore, time-dependent induction of Angptl8 expression reaches a peak at 8 hours of insulin treatment, and the insulin effect on Angptl8 synthesis is almost completely lost 24 hours after insulin ad-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2013-4349
jcem.endojournals.org
E2515
physiological significance of increased Angptl8 levels in T2DM need to be addressed in future experiments. In our study, patients on CD have significantly lower Angptl8 levels as compared with subjects with an eGFR ⬎50 mL/min/1.73 m2. These data suggest that Angptl8 is probably not eliminated by the kidneys. In contrast to Angptl8, other adipokines and hepatokines including leptin (23), adiponectin (24), retinol-binding protein-4 (8), progranulin (25), adipocyte fatty acid-binding protein (26), FGF21 (10), and chemerin (9) are significantly elevated in endstage renal disease. Interestingly, circulating Angptl8 was found to be significantly increased after as comFigure 2. Regulation of Angptl8 mRNA by insulin, other hormones, troglitazone, and during pared with before hemodialysis in a adipogenesis. A and B, Differentiated 3T3–L1 cells were serum starved for 6 hours (A) or overnight (B) before various concentrations of insulin (A) were added for 16 hours or 100nM subset of the patients (n ⫽ 31, insulin (B) was added for the indicated periods of time. C, Angptl8 expression was determined in mean ⫾ SD level 2.35 ⫾ 1.52 g/L as undifferentiated 3T3–L1 cells (day 0), as well as 3, 6, and 9 days after induction of differentiation. D, compared with 1.13 ⫾ 0.56 g/L, After 5 hours serum starvation, differentiated 3T3–L1 adipocytes were cultured in the presence or P ⬍ .001). First, these results suggest absence of 30 ng/mL IL-6, 20 ng/mL IL-1, or 500 ng/mL GH for 24 hours and 10M troglitazone (Trog) for 16 hours. Extraction of total RNA and quantitative real-time RT-PCR that the adipokine/hepatokine is not determining mRNA levels normalized to 36B4 expression were performed as described in dialyzable. It is possible that hemoSubjects and Methods. Results are shown as mean ⫾ SEM. **, P ⬍ .01; *, P ⬍ .05 as compared concentration from the removal of with untreated control (Con) (A, B, and D) and before induction of differentiation (0 days, C) as assessed by unpaired Student’s t tests. fluid may have contributed to this change. Alternatively, Angptl8 secretion may be enhanced in tissues as dition. These findings suggest a rather subacute effect of a reaction to the hemodynamic or metabolic changes ininsulin on Angptl8 expression in differentiated 3T3-L1 duced by dialysis. Furthermore, based on our results, one cells. Clearly, hormonal networks influencing Angptl8 secould speculate that Angptl8 decreases in renal failure as rum levels need to be elucidated in more detail in future has been shown for other cytokines (27, 28). Clearly, adexperiments. The pathophysiological significance of Angptl8 up- ditional experiments need to be performed to elucidate by regulation in T2DM remains to be determined. It is inter- which mechanisms circulating Angptl8 is regulated during esting to note in this context that Angptl8 has recently renal failure and eliminated. Besides markers of glucose homeostasis and renal funcbeen introduced as a novel adipokine/hepatokine inducing insulin secretion from pancreatic -cells (5). In more de- tion, Angptl8 is directly and independently associated tail, injecting mice with plasmids encoding Angptl8 results with age, gender, BMI, and HDL cholesterol in the current in a striking 17-fold increase in -cell replication as com- report. To date, no study has determined regulation of this pared with control animals (5). Mechanistically, Angptl8- adipokine/hepatokine in T2DM. Only one study has asinjected mice exhibit increased expression of cell cycle reg- sessed circulating Angptl8 levels in 33 patients with ulators and cell cycle-related transcription factors in islets, T1DM revealing no correlation of Angptl8 with age, BMI, whereas cell cycle inhibitors decrease as compared with and HDL cholesterol (6). Clearly, the pathophysiological control mice (5). Taking these findings into consideration, significance of these associations needs to be assessed in it is tempting to speculate that Angptl8 upregulation in future studies. It is interesting to note in this context that T2DM is a compensatory mechanism to limit the meta- a nonsynonymous variant in the Angptl8 gene affects bolic consequences of insulin resistance and to improve plasma levels of LDL and HDL cholesterol without influglucose tolerance. Alternatively, Angptl8 resistance might encing TG (29). In the present study, Angptl8 serum levels were signifdevelop in T2DM similar to insulin and leptin resistance found in obesity (22). Clearly, these hypotheses and the icantly higher in female as compared with male subjects in
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
E2516
Ebert et al
Angptl8 and Type 2 Diabetes Mellitus
agreement with findings for other adipokines and hepatokines including adiponectin (30), leptin (31, 32), vaspin (7), and FGF21 (10). It needs to be determined whether an inhibitory effect of androgens on expression of Angptl8 exists similar to adiponectin and leptin (30, 33). It should be noted that Angptl8 concentrations in our study were higher as compared with circulating levels of this novel adipokine/hepatokine in a previous study (6). Different patient characteristics and ELISA systems might well explain the observed variations. Taken together, circulating Angptl8 is independently and positively associated with T2DM and FG in vivo. Furthermore, insulin induces Angptl8 mRNA expression in differentiated 3T3-L1 adipocytes in vitro. The physiological significance of these findings, as well as the factors contributing to Angptl8 regulation in humans, needs to be established in future experiments.
Acknowledgments Addressallcorrespondenceandrequestsforreprintsto:ThomasEbert, MD, Leipzig University Medical Center, Liebigstrasse 20, 04103 Leipzig, Germany. E-mail: [email protected]. This study was supported by grants to M.F. from the Deutsche Forschungsgemeinschaft (DFG) (SFB 1052/1, C06) and the Federal Ministry of Education and Research (BMBF), Germany (FKZ: 01EO1001 [Integrated Research and Treatment Center AdiposityDiseases, project K7–58]), and the Deutsche Hochdruckliga eV, as well as grants to S.K. and T.E. from the German Diabetes Association (DDG). Furthermore, T.E. was supported by the BMBF, Germany (FKZ: 01EO1001 [Integrated Research and Treatment Center AdiposityDiseases, MetaRot program]) and by grant of MSD Sharp and Dohme GmbH (MSD Stipendium 2013 Diabetologie). Moreover, A.T. was supported by grants from the DFG (SFB 1052/1, C01 and SPP 1629 TO 718/2–1), from the DDG, and from the Diabetes Hilfs- und Forschungsfonds Deutschland. Author contributions: T.E., S.K., A.T., and M.F. wrote the manuscript and researched data. A.H., A.B., and J.K., researched data and reviewed/edited the manuscript. U.L. researched data. M.B. and M.S. contributed to the discussion and reviewed/edited the manuscript. Dr Thomas Ebert is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Disclosure Summary: No conflict of interest is declared.
References 1. Considine RV, Sinha MK, Heiman ML, et al. Serum Immunoreactive-Leptin Concentrations in Normal-Weight and Obese Humans. N Engl J Med. 1996;334:292–295. 2. Hou N, Luo JD. Leptin and cardiovascular diseases. Clin Exp Pharmacol Physiol. 2011;38:905–913. 3. Kharitonenkov A, Shiyanova TL, Koester A, et al. FGF-21 as a novel metabolic regulator. J Clin Invest. 2005;115:1627–1635.
J Clin Endocrinol Metab, December 2014, 99(12):E2510 –E2517
4. Sarruf DA, Thaler JP, Morton GJ, et al. Fibroblast growth factor 21 action in the brain increases energy expenditure and insulin sensitivity in obese rats. Diabetes. 2010;59:1817–1824. 5. Yi P, Park JS, Melton DA. Betatrophin: a hormone that controls pancreatic  cell proliferation. Cell. 2013;153:747–758. 6. Espes D, Lau J, Carlsson PO. Increased circulating levels of betatrophin in individuals with long-standing type 1 diabetes. Diabetologia. 2014;57:50 –53. 7. Seeger J, Ziegelmeier M, Bachmann A, et al. Serum levels of the adipokine vaspin in relation to metabolic and renal parameters. J Clin Endocrinol Metab. 2008;93:247–251. 8. Ziegelmeier M, Bachmann A, Seeger J, et al. Serum levels of adipokine retinol-binding protein-4 in relation to renal function. Diabetes Care. 2007;30:2588 –2592. 9. Pfau D, Bachmann A, Lössner U, et al. Serum levels of the adipokine chemerin in relation to renal function. Diabetes Care. 2010;33:171– 173. 10. Stein S, Bachmann A, Lössner U, et al. Serum levels of the adipokine FGF21 depend on renal function. Diabetes Care. 2009;32:126 –128. 11. Levey AS, Bosch JP, Lewis JB, et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med. 1999;130:461– 470. 12. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2013;36:S67–S74. 13. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28:412– 419. 14. Mai M, Tönjes A, Kovacs P, Stumvoll M, Fiedler GM, Leichtle AB. Serum Levels of Acylcarnitines Are Altered in Prediabetic Conditions. PLoS One. 2013;8:e82459. 15. Ebert T, Wurst U, Kralisch S, Stumvoll M, Fasshauer M. The novel adipokine/hepatokine betatrophin is increased in gestational diabetes mellitus. Diabetes. 2014;63(Suppl 1):A212– A343; 1267-P. 16. Kralisch S, Sommer G, Weise S, et al. Interleukin-1 is a positive regulator of TIARP/STAMP2 gene and protein expression in adipocytes in vitro. FEBS Lett. 2009;583:1196 –1200. 17. Kralisch S, Klein J, Lossner U, et al. Isoproterenol, TNF␣, and insulin downregulate adipose triglyceride lipase in 3T3–L1 adipocytes. Mol Cell Endocrinol. 2005;240:43– 49. 18. Fasshauer M, Klein J, Neumann S, Eszlinger M, Paschke R. Hormonal regulation of adiponectin gene expression in 3T3–L1 adipocytes. Biochem Biophys Res Commun. 2002;290:1084 –1089. 19. Motoshima H, Wu X, Sinha MK, et al. Differential regulation of adiponectin secretion from cultured human omental and subcutaneous adipocytes: effects of insulin and rosiglitazone. J Clin Endocrinol Metab. 2002;87:5662–5667. 20. Kralisch S, Weise S, Sommer G, et al. Interleukin-1 induces the novel adipokine chemerin in adipocytes in vitro. Regul Pept. 2009; 154:102–106. 21. Zhang Y, Scarpace PJ. The role of leptin in leptin resistance and obesity. Physiol Behav. 2006;88:249 –256. 22. Ren G, Kim JY, Smas CM. Identification of RIFL, a novel adipocyteenriched insulin target gene with a role in lipid metabolism. Am J Physiol Endocrinol Metab. 2012;303:E334 –E351. 23. Merabet E, Dagogo-Jack S, Coyne DW, et al. Increased plasma leptin concentration in end-stage renal disease. J Clin Endocrinol Metab. 1997;82:847– 850. 24. Zoccali C, Mallamaci F, Tripepi G, et al. Adiponectin, metabolic risk factors, and cardiovascular events among patients with endstage renal disease. J Am Soc Nephrol. 2002;13:134 –141. 25. Richter J, Focke D, Ebert T, et al. Serum levels of the adipokine progranulin depend on renal function. Diabetes Care. 2013;36: 410 – 414. 26. Sommer G, Ziegelmeier M, Bachmann A, et al. Serum levels of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2013-4349
27.
28.
29.
30.
adipocyte fatty acid binding protein are increased in chronic haemodialysis. Clin Endocrinol (Oxf). 2008;69:901–905. Ebert T, Focke D, Petroff D, et al. Serum levels of the myokine irisin in relation to metabolic and renal function. Eur J Endocrinol. 2014; 170:501–506. Ebert T, Bachmann A, Lössner U, et al. Serum levels of angiopoietinrelated growth factor in diabetes mellitus and chronic hemodialysis. Metabolism. 2009;58:547–551. Quagliarini F, Wang Y, Kozlitina J, et al. Atypical angiopoietin-like protein that regulates ANGPTL3. Proc Natl Acad Sci U S A. 2012; 109:19751–19756. Nishizawa H, Shimomura I, Kishida K, et al. Androgens decrease
jcem.endojournals.org
E2517
plasma adiponectin, an insulin-sensitizing adipocyte-derived protein. Diabetes. 2002;51:2734 –2741. 31. Horn R, Geldszus R, Pötter E, von zur Mühlen A, Brabant G. Radioimmunoassay for the detection of leptin in human serum. Exp Clin Endocrinol Diabetes. 1996;104:454 – 458. 32. Ma Z, Gingerich RL, Santiago JV, Klein S, Smith CH, Landt M. Radioimmunoassay of leptin in human plasma. Clin Chem. 1996; 42:942–946. 33. Kapoor D, Clarke S, Stanworth R, Channer KS, Jones TH. The effect of testosterone replacement therapy on adipocytokines and C-reactive protein in hypogonadal men with type 2 diabetes. Eur J Endocrinol. 2007;156:595– 602.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
O N L I N E
Hot Topics in Translational Endocrinology—Endocrine Research
Circulating Angiopoietin-like Protein 8 Is Independently Associated With Fasting Plasma Glucose and Type 2 Diabetes Mellitus T. Ebert,* S. Kralisch,* A. Hoffmann, A. Bachmann, U. Lössner, J. Kratzsch, M. Blüher, M. Stumvoll, A. Tönjes,† and M. Fasshauer† Department of Endocrinology and Nephrology (T.E., S.K., A.H., A.B., U.L., M.B., M.S., A.T., M.F.) and Institute of Laboratory Medicine (J.K.), University of Leipzig, and Integrated Research and Treatment Center Adiposity Diseases (T.E., S.K., U.L., A.T., M.F.), Leipzig University Medical Center, 04103 Leipzig, Germany
Objective: Angiopoietin-like protein 8 (Angptl8) has recently been introduced as a novel adipokine/hepatokine that promotes pancreatic -cell proliferation and improves glucose tolerance in mouse models of insulin resistance. However, regulation of Angptl8 in human type 2 diabetes mellitus (T2DM) and renal dysfunction has not been determined. Research Design and Methods: Serum Angptl8 levels were quantified by ELISA in 62 patients with T2DM as compared with 58 nondiabetic subjects in vivo. Within both groups, about half of the patients were on chronic hemodialysis or had an estimated glomerular filtration rate above 50 mL/min/1.73 m2. Furthermore, we investigated the effect of insulin and differentiation on Angptl8 mRNA expression in 3T3-L1 adipocytes in vitro. Results: Median [interquartile range] serum Angptl8 levels were higher in patients with T2DM (1.19 [0.37] g/L) as compared with nondiabetic subjects (1.03 [0.47] g/L) (P ⫽ .005). Furthermore, the adipokine/hepatokine was significantly higher in women (1.21 [0.47] g/L) as compared with men (1.05 [0.44] g/L]) (P ⫽ .013). In multivariate analysis, fasting glucose and T2DM but not renal function remained independent and positive predictors of circulating Angptl8 even after adjustment for markers of obesity, lipid status, and inflammation (P ⬍ .05). Furthermore, Angptl8 mRNA expression was induced by insulin and during adipogenesis in 3T3-L1 adipocytes in vitro. Conclusions: Circulating Angptl8 is positively and independently associated with T2DM and fasting glucose in vivo. Furthermore, Angptl8 mRNA expression is induced by insulin and during adipogenesis in 3T3-L1 adipocytes in vitro. (J Clin Endocrinol Metab 99: E2510 –E2517, 2014)
he metabolic syndrome and its phenotype including visceral obesity, insulin resistance, type 2 diabetes mellitus (T2DM), hypertension, and dyslipidemia is an increasing global health burden. In the last 2 decades, various metabolic hormones have been shown to significantly influence obesity and associated complications. Among those, adipocyte- and hepatocyte-secreted proteins, socalled adipokines and hepatokines, provide a link between
T
obesity, the metabolic syndrome, and associated cardiovascular diseases. Thus, the adipokine leptin is an appetite-suppressive adipokine upregulated in obesity and cardiovascular disease in various animal models (1, 2). Moreover, the adipocyte- and hepatocyte-derived cytokine fibroblast growth factor 21 (FGF21) stimulates glucose uptake in adipocytes (3) and increases insulin sensitivity in rodents (4).
ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2014 by the Endocrine Society Received December 7, 2013. Accepted September 4, 2014. First Published Online October 17, 2014
* T.E. and S.K. equally contributed to this work. † A.T. and M.F. equally contributed to this work. Abbreviations: Angptl8, angiopoietin-like protein 8; BMI, body mass index; CD, chronic hemodialysis; eGFR, estimated glomerular filtration rate; FG, fasting glucose; FGF21, fibroblast growth factor 21; FI, fasting insulin; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance; hsIL-6, high-sensitivity IL-6; LDL, low-density lipoprotein; oGTT, oral glucose tolerance test; T2DM, type 2 diabetes mellitus; TG, triglyceride.
E2510
jcem.endojournals.org
J Clin Endocrinol Metab, December 2014, 99(12):E2510 –E2517
doi: 10.1210/jc.2013-4349
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2013-4349
Recently, angiopoietin-like protein 8 (Angptl8; also known as betatrophin, hepatocellular carcinoma-associated protein TD26, or lipasin) has been introduced as a novel adipokine/hepatokine that significantly and specifically promotes pancreatic -cell proliferation, expands -cell mass, and improves glucose tolerance in mouse models of insulin resistance (5). The authors suggested that Angptl8 expression is a compensatory mechanism to increase -cell proliferation in insulin resistance (5). Interestingly, serum levels of this adipokine/hepatokine are 2-fold increased in patients with T1DM as compared with age-matched healthy controls in a very recent study (6). Taking these interesting and novel findings into consideration, Angptl8 appears as a novel adipokine/hepatokine that improves glucose tolerance. In contrast to these elegant studies in animals (5) and patients with T1DM (6), regulation of Angptl8 in human T2DM and insulin resistance has not been investigated until now. Furthermore, no study has evaluated potential renal elimination and regulation of this beneficial adipokine/hepatokine during renal failure in humans so far. To address these issues, we determined circulating Angptl8 concentrations in 62 well-phenotyped patients with T2DM as compared with 58 nondiabetic subjects in vivo. Within both groups, about half of the patients were on chronic hemodialysis (CD) or had an estimated glomerular filtration rate (eGFR) above 50 mL/min/1.73 m2. Furthermore, we correlated Angptl8 to clinical and biochemical measures of renal function, indices of glucose metabolism, and lipid metabolism as well as inflammation. Moreover, we investigated the effects of hormones inducing insulin resistance, the insulin-sensitizing peroxisome proliferatoractivated receptor-␥ agonist troglitazone, as well as differentiation, on Angptl8 mRNA expression in 3T3-L1 adipocytes in vitro. Our primary hypotheses were that Angptl8 concentrations are increased in T2DM as compared with nondiabetic subjects and that adipocyte mRNA expression of this novel adipokine/hepatokine is induced by insulin in vitro. Furthermore, we hypothesized that Angptl8 serum levels would increase in patients on CD, indicating that Angptl8 is eliminated by the kidneys.
Subjects and Methods Angptl8 in T2DM and renal dysfunction as well as regulation by glucose in vivo Subjects For the cross-sectional study on Angptl8 in T2DM and renal dysfunction, 120 Caucasian men (n ⫽ 62) and women (n ⫽ 58) were recruited by the Department of Endocrinology and Ne-
jcem.endojournals.org
E2511
phrology, University of Leipzig. The study design has recently been described (7–10). In brief, 60 patients had an eGFR ⬎50 mL/min/1.73 m2 (controls) according to the Modification of Diet in Renal Disease formula (11), whereas 60 patients were on CD. T2DM in controls and patients on CD was defined as fasting blood glucose ⱖ126 mg/dL or use of insulin or oral hypoglycemic medications according to Ref. 12. Using these criteria, 30 of the 60 control and 32 of the 60 CD patients presented with T2DM. Body mass index (BMI) was calculated as weight divided by height squared. Waist-to-hip ratio was determined after waist and hip circumferences were assessed. Age ranged from 32 to 85 years and BMI from 18.7 to 46.1 kg/m2. Homeostasis model assessment of insulin resistance (HOMA-IR) was calculated as previously described (13). Patients with severe conditions including generalized inflammation or end-stage malignant diseases were excluded from the study. To better define dynamic regulation of Angptl8 by glucose, concentrations of the novel adipokine/hepatokine were measured in a separate cohort comprising 30 subjects with normal glucose tolerance during a 75-g oral glucose tolerance test (oGTT). The study design has recently been described (14). Both studies were approved by the local ethics committee, and all subjects gave written informed consent before taking part in the respective study.
Assays All blood samples were taken after a fasting period of at least 8 hours. In CD patients, blood was obtained just before hemodialysis started. Serum Angptl8 (Phoenix Pharmaceuticals), leptin (Mediagnost), and high-sensitivity IL-6 (hsIL-6) (R&D Systems) were determined by ELISA according to the manufacturers’ instructions. According to the manufacturer, the sensitivity of the Angptl8 ELISA was 0.01 g/L. The degree of precision of the ELISA system in terms of intra-assay and interassay coefficient of variance (percentage) was less than 7.5% and 10.5%, respectively. Spike recovery and linearity were in a range of 97% to 131% and 0.3 to 4.4 g/L, respectively. Furthermore, the ELISA was specific for human Angptl8 and did not crossreact with murine and rat Angptl8. Moreover, the antibody used in the ELISA kit specifically detected Angtpl8 at the expected size of about 22.5 kDa when unextracted and extracted human plasma samples were probed by Western blot analysis. In addition, quality controls were included in all ELISA measurements, and results were within the expected range. Furthermore, the adipokine/hepatokine was recently quantified by our group in a second cohort of patients with the same Angptl8 ELISA kit lot without any noticeable problems (15). Serum creatinine, fasting glucose (FG), fasting insulin (FI), and triglycerides (TGs) as well as total, high-density lipoprotein (HDL), and low-density lipoprotein (LDL) cholesterol, were measured in a certified laboratory by standard methods.
Effects of hormones, troglitazone, and differentiation on Angptl8 expression in 3T3-L1 cells in vitro Reagents and cell culture Isobutylmethylxanthine, IL-6, and dexamethasone were obtained from Calbiochem; insulin from Roche Molecular Biochemicals; and IL-1, GH, and troglitazone from Sigma-Aldrich. Cell culture reagents were purchased from GE Healthcare and
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
E2512
Ebert et al
Angptl8 and Type 2 Diabetes Mellitus
Life Technologies and oligonucleotides from MWG-Biotech. 3T3-L1 cells (American Type Culture Collection) were grown and differentiated into adipocytes as previously described (16). After 6 hours serum starvation, differentiated 3T3-L1 adipocytes were cultured in the presence or absence of insulin, IL-6, IL1, GH, and troglitazone under conditions further described in Figure 2. Concentrations of the respective hormones and troglitazone were used according to prior studies (17–19). Furthermore, Angptl8 expression was assessed during differentiation of 3T3-L1 cells.
Analysis of Angptl8 mRNA production Angptl8 mRNA synthesis was determined by quantitative real-time RT-PCR. In more detail, 1 g total RNA was reverse transcribed in a 20-L reverse transcription reaction using standard reagents (Life Technologies) and 1 L of each reverse transcription reaction was quantified on a LightCycler 480 real-time PCR 96-well thermocycler using LightCycler 480 Probes Master Mix (Roche Diagnostics GmbH) essentially as described (20). The following primer pairs were used: Angptl8, CACCAGCCTGTCGGAGATTC (sense) and GTCTCTGCTGGATCTGTCGC (antisense); and m36B4, AAGCGCGTCCTGGCATTGTCT (sense) and CCGCAGGGGCAGCAGTGGT (antisense). Angptl8 expression was calculated relative to m36B4, which was used as an internal control due to its resistance to hormonal regulation.
Statistical analysis A P value ⬍ .05 was considered as statistically significant in all analyses. SPSS software version 20.0 (IBM) was used for all statistical analyses of human in vivo data. Overall group differences between the 4 subgroups (control/T2DM⫺, control/T2DM⫹, CD/ T2DM⫺, and CD/T2DM⫹) were assessed by nonparametric Kruskal-Wallis test with Bonferroni post hoc analysis. Differences between nominal variables, eg, T2DM status, were assessed by nonparametric Mann-Whitney U test. Furthermore, a two-way ANOVA was conducted to further examine the effects of T2DM and CD status on circulating Angptl8. Univariate correlations were analyzed by nonparametric Spearman’s rank correlation method. Furthermore, multivariate linear regression analyses were performed. Before multivariate analysis and twoway ANOVA, distribution of continuous variables was tested for normality using the Shapiro-Wilk test and nonnormally distributed parameters were logarithmically transformed. In multivariate linear analysis, only parameters that correlated significantly with Angptl8 in univariate analysis were included. For covariates, eg, creatinine and eGFR, the parameter with the strongest univariate correlation was included in the multivariate model. Furthermore, gender was included in multivariate analysis. For in vitro experiments, GraphPad Prism version 6 (GraphPad Software) was used. Here, Angptl8 expression adjusted for m36B4 after hormone and troglitazone treatment or during differentiation was analyzed relative to untreated control cells and preadipocytes, respectively. Results are shown as mean ⫾ SEM. Differences between treatments were analyzed by unpaired Student’s t tests with prior logarithmic transformation.
J Clin Endocrinol Metab, December 2014, 99(12):E2510 –E2517
Results Angptl8 in T2DM and renal dysfunction in vivo Baseline characteristics of the total sample Table 1 summarizes the clinical characteristics of the 4 subgroups (control/T2DM⫺, control/T2DM⫹, CD/ T2DM⫺, and CD/T2DM⫹) studied. In the total sample, the median (interquartile range) serum Angptl8 level was 1.14 (0.42) g/L. Circulating Angptl8 was significantly different between the 4 subgroups (P ⬍ .001) (Table 1). Furthermore, Angptl8 concentrations were significantly higher in women (1.21 [0.47] g/L) as compared with men (1.05 [0.44] g/L]) (P ⫽ .013) in all patients. Notably, there was a significant interaction between T2DM and CD status on Angptl8 concentrations (F1,116 ⫽ 6.149, P ⫽ .015) in two-way ANOVA (Table 1). Angptl8 in T2DM When investigating circulating Angptl8 in relation to T2DM status, patients with T2DM had significantly higher Angptl8 concentrations (1.19 [0.37] g/L) as compared with nondiabetic subjects (1.03 [0.47] g/L) (P ⫽ .005) (Figure 1). Interestingly, Angptl8 serum levels remained significantly higher in patients with T2DM as compared with nondiabetic subjects even after adjustment for age, gender, and CD status (P ⬍ .001). Furthermore, the main effect of T2DM on Angptl8 was significant (F1,116 ⫽ 10.460, P ⫽ .002) in two-way ANOVA (Table 1). When investigating Angptl8 levels in nondiabetic control subjects not on CD in a separate analysis, median levels of the adipokine/hepatokine were not significantly different in subjects with normal glucose tolerance as compared with subjects with prediabetes as defined by a 75-g oGTT (data not shown). Angptl8 in CD When investigating circulating Angptl8 in relation to CD, Angptl8 serum levels were significantly lower in patients on CD (1.06 [0.40] g/L) as compared with subjects with an eGFR ⬎50 mL/min/1.73 m2 (1.23 [0.45] g/L) (P ⫽ .014). Interestingly, the main effect of CD status on circulating Angptl8 did also reach statistical significance (F1,116 ⫽ 7.465, P ⫽ .007) in two-way ANOVA (Table 1). Univariate correlations in the total sample Serum Angptl8 concentrations in all individuals were positively and significantly correlated with BMI, FG, FI, HOMA-IR, HDL cholesterol, eGFR, and leptin (P ⬍ .05) (Table 2). In contrast, Angptl8 was significantly and negatively correlated with age, serum creatinine, and hsIL-6 (P ⬍ .05) (Table 2).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2013-4349
Table 1.
jcem.endojournals.org
E2513
Baseline Characteristics of the Study Populationa Two-way ANOVA
n Angptl8, g/L Age, yd Gender (M/F) BMI, kg/m2d WHR SBP, mm Hg DBP, mm Hgd FG, mmol/Ld FI, pmol/Ld HOMA-IRd Cholesterol, mmol/Ld HDL cholesterol, mmol/Ld LDL cholesterol, mmol/Ld TG, mmol/Ld FFA, mmol/Ld Creatinine, mol/Ld GFR, mL/min/1.73 m2d Leptin, g/Ld hsIL-6, ng/Ld
Control/T2DMⴚ
Control/T2DMⴙ
CD/T2DMⴚ
CD/T2DMⴙ
30 1.06 (0.49) 63 (19) 11/19 28.2 (5.6) 0.88 (0.12) 125 (21) 77 (10) 5.1 (1.3) 45.1 (33.3) 1.3 (1.2) 5.3 (0.9) 1.4 (0.4) 3.5 (1.2) 1.1 (0.8) 0.5 (0.2) 76 (17) 77 (21) 16.9 (27.9) 1.87 (2.11)
30 1.34 (0.50)b 63 (16) 16/14 29.1 (5.2) 0.94 (0.10) 126 (20) 73 (15) 7.6 (3.2)b 47.9 (62.6) 2.7 (2.9) 4.9 (1.5) 1.2 (0.5) 2.9 (0.9)b 1.4 (0.9) 0.6 (0.4) 72 (22) 84 (35) 16.7 (26.4) 2.07 (1.25)
28 1.01 (0.42)c 59 (22) 15/13 25.2 (6.5)c 0.96 (0.18) 125 (38) 77 (20) 4.6 (1.2)c 28.2 (47.6) 0.8 (1.3)c 4.4 (1.2)b 1.0 (0.5)b 2.7 (0.9)b 1.6 (0.9)b 0.6 (0.5) 829 (431)b,c 6 (3)b,c 11.8 (37.9) 6.14 (4.91)b,c
32 1.11 (0.34)c 68 (14) 20/12 27.9 (6.6) 1.00 (0.14)b 120 (25) 70 (18) 5.2 (3.3)c 50.1 (91.6) 1.4 (3.1) 4.2 (1.3)b 1.0 (0.3)b,c 2.1 (1.4)b 1.8 (1.4)b 0.7 (0.5) 717 (221)b,c 7 (3)b,c 32.5 (67.5) 7.48 (7.30)b,c
PT2DM
PCD
PT2DMⴛCD
.002 .018
.007 .590
.015 .137
.092 ⬍.001 .586 .114 ⬍.001 .117 .004 .031 .093 ⬍.001 .003 .104 .563 .354 .127 .361
.001 .003 .178 .060 ⬍.001 .554 .147 ⬍.001 ⬍.001 ⬍.001 ⬍.001 .830 ⬍.001 ⬍.001 .266 ⬍.001
.428 .924 .569 .967 .154 .678 .992 .376 .807 .644 .837 .660 .960 .691 .070 .214
Abbreviations: DBP, diastolic blood pressure; FFA, free fatty acids; SBP, systolic blood pressure; WHR, waist-to-hip ratio. a Values for median (interquartile range) are shown. Parameters were analyzed by Kruskal-Wallis test followed by Bonferroni post hoc analysis. Exact P values for the effects of T2DM (PT2DM) and CD (PCD), as well as for interaction between T2DM and CD (PT2DM⫻CD), on the respective parameters were assessed by two-way ANOVA. b
P ⬍ 0.05 compared with control/T2DM⫺.
c
P ⬍ 0.05 compared with control/T2DM⫹.
d
Nonnormally distributed variables.
Multivariate regression analysis in the total sample Multiple linear regression analysis revealed that FG remained independently and positively associated with circulating Angptl8 after adjustment for age, gender, BMI,
HDL cholesterol, creatinine, and hsIL-6 (P ⬍ .001) (Table 3). Furthermore, gender (gender coefficients were coded such that females have a larger value as compared with males), BMI, and HDL cholesterol were positive and inTable 2.
Univariate Correlations With Serum Angptl8
b
Age, y BMI, kg/m2b WHR SBP, mm Hg DBP, mm Hgb FG, mmol/Lb FI, pmol/Lb HOMA-IRb Cholesterol, mmol/Lb HDL cholesterol, mmol/Lb LDL cholesterol, mmol/Lb TG, mmol/Lb FFA, mmol/Lb Creatinine, mol/Lb eGFR, mL/min/1.73 m2b Leptin, g/Lb hsIL-6, ng/Lb
Figure 1. Box-and-whisker plot of Angptl8 serum levels in patients with T2DM (n ⫽ 62) as compared with nondiabetic subjects (n ⫽ 58). P value was assessed by nonparametric Mann-Whitney U test.
r
P
⫺0.208 0.251 ⫺0.175 0.016 0.127 0.291 0.182 0.256 0.105 0.302 ⫺0.016 ⫺0.014 0.057 ⫺0.232 0.187 0.216 ⫺0.184
.023a .006a .055 .866 .166 .001a .047a .005a .254 .001a .867 .880 .533 .011a .041a .018a .044a
Abbreviations: DBP, diastolic blood pressure; FFA, free fatty acids; SBP, systolic blood pressure; WHR, waist-to-hip ratio. a
significant correlation as assessed by Spearman’s correlation method.
b
Nonnormally distributed variables.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
E2514
Ebert et al
Angptl8 and Type 2 Diabetes Mellitus
Table 3. Multivariate Regression Analysis With Angptl8 as Dependent Variablea Independent Variables c
Age Gender BMIc FGc HDL cholesterolc Creatininec hsIL-6c

P
⫺0.221 .194 .165 .397 .287 .111 ⫺.009
.006b .023b .049b ⬍.001b .002b .345 .935
a Multivariate regression analysis between Angptl8 (dependent variable) and the independent variables shown. Standardized coefficients and P values are given. Gender coefficients were coded such that females have a larger value compared with males. b
Significant correlation.
c
Nonnormally distributed variables were logarithmically transformed before testing.
dependent predictors of Angptl8 serum concentrations, whereas age was independently and negatively associated with the adipokine/hepatokine (P ⬍ .05) (Table 3). Interestingly, T2DM status remained significantly and positively associated with circulating Angptl8 when included in the multivariate model instead of FG ( ⫽ .374; P ⬍ .001). Univariate correlations in the subgroups Of the associations remaining significant in multivariate analyses in the total sample (n ⫽ 120; Table 3), Angptl8 was significantly associated with FG and HDL cholesterol in subjects not on CD, with age in non-T2DM, and with age, BMI, FG, and HDL cholesterol in T2DM in univariate analyses (data not shown). In contrast, Angptl8 was not significantly correlated with any of these parameters when CD patients were analyzed separately by univariate analysis (data not shown). When further segregating the study cohort into 4 subgroups (control/T2DM⫺, control/ T2DM⫹, CD/T2DM⫺, and CD/T2DM⫹), HDL cholesterol was positively correlated with Angptl8 in control/ T2DM⫹ and age was negatively associated with Angptl8 in CD/T2DM⫺ in univariate analyses (data not shown). Regulation of Angptl8 by glucose in vivo To determine the influence of glucose on Angptl8 levels in more detail, dynamic regulation of circulating Angptl8 was elucidated in a cohort comprising 30 subjects with normal glucose tolerance during an oGTT. Here, mean ⫾ SE Angptl8 was not significantly different at 30 minutes (0.52 ⫾ 0.17 g/L) and 120 minutes (0.63 ⫾ 0.21 g/L) after glucose exposure as compared with fasting baseline levels at 0 minutes (0.61 ⫾ 0.15 g/L) (gender-adjusted P for overall differences ⫽ .370) (Supplemental Figure 1).
J Clin Endocrinol Metab, December 2014, 99(12):E2510 –E2517
Effects of hormones, troglitazone, and differentiation on Angptl8 expression in 3T3-L1 cells in vitro Treatment of differentiated 3T3-L1 adipocytes with insulin for 16 hours stimulated Angptl8 expression in a dosedependent manner (all P ⬍ .01) (Figure 2A). Here, a significant 8-fold induction of Angptl8 mRNA synthesis was detectable at insulin concentrations as low as 1nM, and insulin increased Angptl8 expression up to 18.5-fold at 3.3nM (P ⬍ .01) (Figure 2A). Furthermore, 100nM insulin upregulated Angptl8 expression also in a time-dependent fashion with significant 30-fold stimulation detectable as early as 2 hours after insulin addition (P ⬍ .05) (Figure 2B). Moreover, maximal 76-fold induction was seen after 8 hours of insulin treatment (P ⬍ .01) (Figure 2B). Angptl8 mRNA production also significantly increased in a differentiation-dependent manner with 50-fold and 145-fold induction of the adipokine/hepatokine at days 6 and 9 of differentiation, respectively, as compared with undifferentiated controls (day 0, P ⬍ .05) (Figure 2C). In contrast, chronic treatment of differentiated 3T3-L1 adipocytes with IL-6, IL-1, GH, and troglitazone did not significantly influence Angptl8 expression (Figure 2D).
Discussion In the current study, we show that patients with T2DM have significantly higher Angptl8 serum levels as compared with nondiabetic subjects. Furthermore, T2DM and FG remain independent and positive predictors of circulating Angptl8 in multivariate analysis. Moreover, we demonstrate that insulin is a direct dose- and time-dependent stimulator of Angptl8 mRNA expression in differentiated 3T3-L1 adipocytes in vitro similar to convincing results presented by Ren and co-workers (21). In agreement with our findings, increased expression of the adipokine/hepatokine has also been shown in animal models of insulin resistance and T2DM. Thus, Angptl8 mRNA is upregulated 3- to 6-fold in the liver of S961treated animals and ob/ob and db/db mice (5). Because insulin significantly induces Angptl8 mRNA expression in differentiated 3T3-L1 adipocytes in vitro, it is tempting to speculate that hyperinsulinemia contributes to upregulation of the adipokine/hepatokine in T2DM. Notably, circulating Angptl8 is not affected by acute changes of glucose homeostasis during a 75-g oGTT in a cohort comprising of 30 subjects with normal glucose tolerance (Supplemental Figure 1). Furthermore, time-dependent induction of Angptl8 expression reaches a peak at 8 hours of insulin treatment, and the insulin effect on Angptl8 synthesis is almost completely lost 24 hours after insulin ad-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2013-4349
jcem.endojournals.org
E2515
physiological significance of increased Angptl8 levels in T2DM need to be addressed in future experiments. In our study, patients on CD have significantly lower Angptl8 levels as compared with subjects with an eGFR ⬎50 mL/min/1.73 m2. These data suggest that Angptl8 is probably not eliminated by the kidneys. In contrast to Angptl8, other adipokines and hepatokines including leptin (23), adiponectin (24), retinol-binding protein-4 (8), progranulin (25), adipocyte fatty acid-binding protein (26), FGF21 (10), and chemerin (9) are significantly elevated in endstage renal disease. Interestingly, circulating Angptl8 was found to be significantly increased after as comFigure 2. Regulation of Angptl8 mRNA by insulin, other hormones, troglitazone, and during pared with before hemodialysis in a adipogenesis. A and B, Differentiated 3T3–L1 cells were serum starved for 6 hours (A) or overnight (B) before various concentrations of insulin (A) were added for 16 hours or 100nM subset of the patients (n ⫽ 31, insulin (B) was added for the indicated periods of time. C, Angptl8 expression was determined in mean ⫾ SD level 2.35 ⫾ 1.52 g/L as undifferentiated 3T3–L1 cells (day 0), as well as 3, 6, and 9 days after induction of differentiation. D, compared with 1.13 ⫾ 0.56 g/L, After 5 hours serum starvation, differentiated 3T3–L1 adipocytes were cultured in the presence or P ⬍ .001). First, these results suggest absence of 30 ng/mL IL-6, 20 ng/mL IL-1, or 500 ng/mL GH for 24 hours and 10M troglitazone (Trog) for 16 hours. Extraction of total RNA and quantitative real-time RT-PCR that the adipokine/hepatokine is not determining mRNA levels normalized to 36B4 expression were performed as described in dialyzable. It is possible that hemoSubjects and Methods. Results are shown as mean ⫾ SEM. **, P ⬍ .01; *, P ⬍ .05 as compared concentration from the removal of with untreated control (Con) (A, B, and D) and before induction of differentiation (0 days, C) as assessed by unpaired Student’s t tests. fluid may have contributed to this change. Alternatively, Angptl8 secretion may be enhanced in tissues as dition. These findings suggest a rather subacute effect of a reaction to the hemodynamic or metabolic changes ininsulin on Angptl8 expression in differentiated 3T3-L1 duced by dialysis. Furthermore, based on our results, one cells. Clearly, hormonal networks influencing Angptl8 secould speculate that Angptl8 decreases in renal failure as rum levels need to be elucidated in more detail in future has been shown for other cytokines (27, 28). Clearly, adexperiments. The pathophysiological significance of Angptl8 up- ditional experiments need to be performed to elucidate by regulation in T2DM remains to be determined. It is inter- which mechanisms circulating Angptl8 is regulated during esting to note in this context that Angptl8 has recently renal failure and eliminated. Besides markers of glucose homeostasis and renal funcbeen introduced as a novel adipokine/hepatokine inducing insulin secretion from pancreatic -cells (5). In more de- tion, Angptl8 is directly and independently associated tail, injecting mice with plasmids encoding Angptl8 results with age, gender, BMI, and HDL cholesterol in the current in a striking 17-fold increase in -cell replication as com- report. To date, no study has determined regulation of this pared with control animals (5). Mechanistically, Angptl8- adipokine/hepatokine in T2DM. Only one study has asinjected mice exhibit increased expression of cell cycle reg- sessed circulating Angptl8 levels in 33 patients with ulators and cell cycle-related transcription factors in islets, T1DM revealing no correlation of Angptl8 with age, BMI, whereas cell cycle inhibitors decrease as compared with and HDL cholesterol (6). Clearly, the pathophysiological control mice (5). Taking these findings into consideration, significance of these associations needs to be assessed in it is tempting to speculate that Angptl8 upregulation in future studies. It is interesting to note in this context that T2DM is a compensatory mechanism to limit the meta- a nonsynonymous variant in the Angptl8 gene affects bolic consequences of insulin resistance and to improve plasma levels of LDL and HDL cholesterol without influglucose tolerance. Alternatively, Angptl8 resistance might encing TG (29). In the present study, Angptl8 serum levels were signifdevelop in T2DM similar to insulin and leptin resistance found in obesity (22). Clearly, these hypotheses and the icantly higher in female as compared with male subjects in
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
E2516
Ebert et al
Angptl8 and Type 2 Diabetes Mellitus
agreement with findings for other adipokines and hepatokines including adiponectin (30), leptin (31, 32), vaspin (7), and FGF21 (10). It needs to be determined whether an inhibitory effect of androgens on expression of Angptl8 exists similar to adiponectin and leptin (30, 33). It should be noted that Angptl8 concentrations in our study were higher as compared with circulating levels of this novel adipokine/hepatokine in a previous study (6). Different patient characteristics and ELISA systems might well explain the observed variations. Taken together, circulating Angptl8 is independently and positively associated with T2DM and FG in vivo. Furthermore, insulin induces Angptl8 mRNA expression in differentiated 3T3-L1 adipocytes in vitro. The physiological significance of these findings, as well as the factors contributing to Angptl8 regulation in humans, needs to be established in future experiments.
Acknowledgments Addressallcorrespondenceandrequestsforreprintsto:ThomasEbert, MD, Leipzig University Medical Center, Liebigstrasse 20, 04103 Leipzig, Germany. E-mail: [email protected]. This study was supported by grants to M.F. from the Deutsche Forschungsgemeinschaft (DFG) (SFB 1052/1, C06) and the Federal Ministry of Education and Research (BMBF), Germany (FKZ: 01EO1001 [Integrated Research and Treatment Center AdiposityDiseases, project K7–58]), and the Deutsche Hochdruckliga eV, as well as grants to S.K. and T.E. from the German Diabetes Association (DDG). Furthermore, T.E. was supported by the BMBF, Germany (FKZ: 01EO1001 [Integrated Research and Treatment Center AdiposityDiseases, MetaRot program]) and by grant of MSD Sharp and Dohme GmbH (MSD Stipendium 2013 Diabetologie). Moreover, A.T. was supported by grants from the DFG (SFB 1052/1, C01 and SPP 1629 TO 718/2–1), from the DDG, and from the Diabetes Hilfs- und Forschungsfonds Deutschland. Author contributions: T.E., S.K., A.T., and M.F. wrote the manuscript and researched data. A.H., A.B., and J.K., researched data and reviewed/edited the manuscript. U.L. researched data. M.B. and M.S. contributed to the discussion and reviewed/edited the manuscript. Dr Thomas Ebert is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Disclosure Summary: No conflict of interest is declared.
References 1. Considine RV, Sinha MK, Heiman ML, et al. Serum Immunoreactive-Leptin Concentrations in Normal-Weight and Obese Humans. N Engl J Med. 1996;334:292–295. 2. Hou N, Luo JD. Leptin and cardiovascular diseases. Clin Exp Pharmacol Physiol. 2011;38:905–913. 3. Kharitonenkov A, Shiyanova TL, Koester A, et al. FGF-21 as a novel metabolic regulator. J Clin Invest. 2005;115:1627–1635.
J Clin Endocrinol Metab, December 2014, 99(12):E2510 –E2517
4. Sarruf DA, Thaler JP, Morton GJ, et al. Fibroblast growth factor 21 action in the brain increases energy expenditure and insulin sensitivity in obese rats. Diabetes. 2010;59:1817–1824. 5. Yi P, Park JS, Melton DA. Betatrophin: a hormone that controls pancreatic  cell proliferation. Cell. 2013;153:747–758. 6. Espes D, Lau J, Carlsson PO. Increased circulating levels of betatrophin in individuals with long-standing type 1 diabetes. Diabetologia. 2014;57:50 –53. 7. Seeger J, Ziegelmeier M, Bachmann A, et al. Serum levels of the adipokine vaspin in relation to metabolic and renal parameters. J Clin Endocrinol Metab. 2008;93:247–251. 8. Ziegelmeier M, Bachmann A, Seeger J, et al. Serum levels of adipokine retinol-binding protein-4 in relation to renal function. Diabetes Care. 2007;30:2588 –2592. 9. Pfau D, Bachmann A, Lössner U, et al. Serum levels of the adipokine chemerin in relation to renal function. Diabetes Care. 2010;33:171– 173. 10. Stein S, Bachmann A, Lössner U, et al. Serum levels of the adipokine FGF21 depend on renal function. Diabetes Care. 2009;32:126 –128. 11. Levey AS, Bosch JP, Lewis JB, et al. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med. 1999;130:461– 470. 12. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2013;36:S67–S74. 13. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28:412– 419. 14. Mai M, Tönjes A, Kovacs P, Stumvoll M, Fiedler GM, Leichtle AB. Serum Levels of Acylcarnitines Are Altered in Prediabetic Conditions. PLoS One. 2013;8:e82459. 15. Ebert T, Wurst U, Kralisch S, Stumvoll M, Fasshauer M. The novel adipokine/hepatokine betatrophin is increased in gestational diabetes mellitus. Diabetes. 2014;63(Suppl 1):A212– A343; 1267-P. 16. Kralisch S, Sommer G, Weise S, et al. Interleukin-1 is a positive regulator of TIARP/STAMP2 gene and protein expression in adipocytes in vitro. FEBS Lett. 2009;583:1196 –1200. 17. Kralisch S, Klein J, Lossner U, et al. Isoproterenol, TNF␣, and insulin downregulate adipose triglyceride lipase in 3T3–L1 adipocytes. Mol Cell Endocrinol. 2005;240:43– 49. 18. Fasshauer M, Klein J, Neumann S, Eszlinger M, Paschke R. Hormonal regulation of adiponectin gene expression in 3T3–L1 adipocytes. Biochem Biophys Res Commun. 2002;290:1084 –1089. 19. Motoshima H, Wu X, Sinha MK, et al. Differential regulation of adiponectin secretion from cultured human omental and subcutaneous adipocytes: effects of insulin and rosiglitazone. J Clin Endocrinol Metab. 2002;87:5662–5667. 20. Kralisch S, Weise S, Sommer G, et al. Interleukin-1 induces the novel adipokine chemerin in adipocytes in vitro. Regul Pept. 2009; 154:102–106. 21. Zhang Y, Scarpace PJ. The role of leptin in leptin resistance and obesity. Physiol Behav. 2006;88:249 –256. 22. Ren G, Kim JY, Smas CM. Identification of RIFL, a novel adipocyteenriched insulin target gene with a role in lipid metabolism. Am J Physiol Endocrinol Metab. 2012;303:E334 –E351. 23. Merabet E, Dagogo-Jack S, Coyne DW, et al. Increased plasma leptin concentration in end-stage renal disease. J Clin Endocrinol Metab. 1997;82:847– 850. 24. Zoccali C, Mallamaci F, Tripepi G, et al. Adiponectin, metabolic risk factors, and cardiovascular events among patients with endstage renal disease. J Am Soc Nephrol. 2002;13:134 –141. 25. Richter J, Focke D, Ebert T, et al. Serum levels of the adipokine progranulin depend on renal function. Diabetes Care. 2013;36: 410 – 414. 26. Sommer G, Ziegelmeier M, Bachmann A, et al. Serum levels of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/jc.2013-4349
27.
28.
29.
30.
adipocyte fatty acid binding protein are increased in chronic haemodialysis. Clin Endocrinol (Oxf). 2008;69:901–905. Ebert T, Focke D, Petroff D, et al. Serum levels of the myokine irisin in relation to metabolic and renal function. Eur J Endocrinol. 2014; 170:501–506. Ebert T, Bachmann A, Lössner U, et al. Serum levels of angiopoietinrelated growth factor in diabetes mellitus and chronic hemodialysis. Metabolism. 2009;58:547–551. Quagliarini F, Wang Y, Kozlitina J, et al. Atypical angiopoietin-like protein that regulates ANGPTL3. Proc Natl Acad Sci U S A. 2012; 109:19751–19756. Nishizawa H, Shimomura I, Kishida K, et al. Androgens decrease
jcem.endojournals.org
E2517
plasma adiponectin, an insulin-sensitizing adipocyte-derived protein. Diabetes. 2002;51:2734 –2741. 31. Horn R, Geldszus R, Pötter E, von zur Mühlen A, Brabant G. Radioimmunoassay for the detection of leptin in human serum. Exp Clin Endocrinol Diabetes. 1996;104:454 – 458. 32. Ma Z, Gingerich RL, Santiago JV, Klein S, Smith CH, Landt M. Radioimmunoassay of leptin in human plasma. Clin Chem. 1996; 42:942–946. 33. Kapoor D, Clarke S, Stanworth R, Channer KS, Jones TH. The effect of testosterone replacement therapy on adipocytokines and C-reactive protein in hypogonadal men with type 2 diabetes. Eur J Endocrinol. 2007;156:595– 602.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 18 December 2014. at 12:45 For personal use only. No other uses without permission. . All rights reserved.
