Original Article
Obesity
OBESITY BIOLOGY AND INTEGRATED PHYSIOLOGY
SULF2 Strongly Prediposes to Fasting and Postprandial Triglycerides in Patients with Obesity and Type 2 Diabetes Mellitus H. Carlijne Hassing1, R. Preethi Surendran2, Bruno Derudas3, An Verrijken4, Sven M. Francque5, Hans L. Mooij1, Sophie J. Bernelot Moens1, Leen M.’t Hart6,7, Giel Nijpels8, Jacqueline M. Dekker9, Kevin Jon Williams10,11, Erik S. G. Stroes1, Luc F. Van Gaal4, Bart Staels3, Max Nieuwdorp1 and Geesje M. Dallinga-Thie1,2
Objective: Hepatic overexpression of sulfatase-2 (SULF2), a heparan sulfate remodeling enzyme, strongly contributes to high triglyceride (TG) levels in obese, type 2 diabetic (T2DM) db/db mice. Nevertheless, data in humans are lacking. Here, the association of human hepatic SULF2 expression and SULF2 gene variants with TG metabolism in patients with obesity and/or T2DM was investigated. Methods: Liver biopsies from 121 obese subjects were analyzed for relations between hepatic SULF2 mRNA levels and plasma TG. Associations between seven SULF2 tagSNPs and TG levels were assessed in 210 obese T2DM subjects with dyslipidemia. Replication of positive findings was performed in 1,316 independent obese T2DM patients. Postprandial TRL clearance was evaluated in 29 obese T2DM subjects stratified by SULF2 genotype. Results: Liver SULF2 expression was significantly associated with fasting plasma TG (r 5 0.271; P 5 0.003) in obese subjects. The SULF2 rs2281279(A>G) SNP was reproducibly associated with lower fasting plasma TG levels in obese T2DM subjects (P < 0.05). Carriership of the minor G allele was associated with lower levels of postprandial plasma TG (P < 0.05) and retinyl esters levels (P < 0.001). Conclusions: These findings implicate SULF2 as potential therapeutic target in the atherogenic dyslipidemia of obesity and T2DM. Obesity (2014) 22, 1309-1316. doi:10.1002/oby.20682
Introduction The prevalence of obesity, the metabolic syndrome, type 2 diabetes mellitus (T2DM) and their sequelae, particularly cardiovascular events, is increasing worldwide (1,2). The accelerated risk for atherosclerosis in these subjects results in large part from their atherogenic dyslipidemia, which includes increased fasting levels of very low density lipoprotein (VLDL) and its major component triglyceride (TG), in combination with impaired clearance of postprandial
triglyceride-rich lipoprotein (TRL) remnants (3-5). In particular, subendothelial retention of VLDL and particularly postprandial TRLremnants have been linked to cardiovascular events in these individuals (6-10). TRL metabolism involves a series of steps that culminate in the uptake of TRL-remnants by hepatocytes (8,11,12). During the past decades, hepatic heparan sulfate proteoglycans (HSPGs) (13-16) and in particular the syndecan-1 HSPG, have been implicated in TRL
1
Department of Vascular Medicine, Academic Medical Center, Amsterdam, The Netherlands 2 Department of Experimental Vascular Medicine, Academic Medical Center, Amsterdam, The Netherlands. Correspondence: G. M. Dallinga-Thie ([email protected]) 3 University of Lille 2; INSERM U1011; EGID; Institute Pasteur de Lille, Lille, France 4 Department of Endocrinology, Diabetology and Metabolism, Antwerp University Hospital, University of Antwerp, Antwerp, Belgium 5 Department of Gastroenterology and Hepatology, Antwerp University Hospital, University of Antwerp, Antwerp, Belgium 6 Department of Molecular Epidemiology, Leiden University Medical Center, Leiden, The Netherlands 7 Department of Molecular Cell Biology, Leiden University Medical Center, Leiden, The Netherlands 8 Department of General Practice, EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, The Netherlands 9 Department of Epidemiology and Biostatistics, EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, The Netherlands 10 Section of Endocrinology, Diabetes and Metabolism, Temple University School of Medicine, Philadelphia, Pennsylvania, USA 11 Department of Molecular and Clinical Medicine, Sahlgrenska Center for Cardiovascular and Metabolic Research, University of G€othenborg, Gothenburg, Sweden Funding agencies: This work was supported by the Dutch Heart Foundation (2010B242 to R.P.S.; CVON-GENIUS 2011-19 to S.J.B.M), the Netherlands Organization for Scientific Research (VENI grant 016.096.044; to M.N.) and by the National Institutes of Health (grants HL94277 and DK100851; to K.J.W.), FP6 Hepadip grant (L.V.G.), the FP7 Resolve grant (to M.N., B.S. and L.V.G.). Disclosure: The authors declared no conflict of interest. Author contributions: H. Carlijne Hassing and R. Preethi Surendran both contributed equally to this work. Additional Supporting Information may be found in the online version of this article. Received: 20 September 2013; Accepted: 9 December 2013; Published online 13 December 2013. doi:10.1002/oby.20682
www.obesityjournal.org
Obesity | VOLUME 22 | NUMBER 5 | MAY 2014
1309
Obesity
SULF2 and Triglycerides in Obesity Carlijne Hassing et al.
removal (17-21). The syndecan-1 HSPG consists of a single-pass transmembrane core protein onto which three heparan sulfate (HS) side-chains are covalently attached. These side-chains bind several important ligands, including TRLs, resulting in hepatic uptake and clearance of these particles. About 50 genes are involved in HSPG assembly and disassembly, affecting core protein expression, HS side-chain length, epimerization of glucuronyl residues, and sulfation patterns (22). It was recently shown that expression of glucosamine6-O-endosulfatase-2 (Sulf2 encoding for sulfatase 2, an HSPG remodeling enzyme) is substantially increased in livers of obese, T2DM db/db mice compared with lean controls (23). In these mice, hepatic Sulf2 mRNA expression was strongly and positively related to plasma TG levels and Sulf2 protein inhibited the ability of syndecan-1 to clear model remnants by cultured liver cells. Recently, we showed that suppression of hepatic Sulf2 expression in obese, T2DM db/db mice by targeted allele-specific antisense administration normalized fasting plasma TG levels and, most strikingly, eliminated postprandial hypertriglyceridemia (24). However, human data on the biology of hepatic SULF2 and its role in TG metabolism remain to be elucidated. In the present study, we sought to answer three questions. First, is there evidence in humans for dysregulation of hepatic SULF2 mRNA levels as part of the metabolic disturbances of obesity, particularly dyslipidemia? Second, are there SULF2 SNPs that associate with fasting dyslipidemia in cohorts of obese, T2DM subjects? Third, does any SULF2 SNP affect postprandial lipoprotein metabolism in obese subjects with T2DM, as assessed by oral fat tolerance studies? To address the first two questions, we used pre-existing cohorts, which have the advantage of large numbers of subjects, but the disadvantage that fasting—not postprandial—lipoprotein values are available. Because fat tolerance studies are laborious, they are practical only in a subset of individuals, not in population-wide cohorts. Our results implicate human SULF2 as an important participant in deranged TRL metabolism in obesity and T2DM.
(Agilent Technologies). Specific primers for SULF2 were designed using Primer3 software (50 -CCT TTG CCG TGT ACC TCA AT-30 and 50 -GCA CGT AGG AGC CGT TGT AT-30 ). mRNA levels were normalized to those of transcription factor IIb (TFIIb) by calculating the difference in cycle threshold (DCt).
SULF2 SNPs and fasting plasma TGs in cohorts of obese, T2DM patients Associations between SULF2 SNPs and plasma TG levels were assessed in participants in the Diabetes Atorvastatin Lipid Intervention (DALI) study and subsequently validated in the Diabetes Care System cohort, both described immediately below. All analysis were performed using the baseline data from either the DALI cohort or the Diabetes Care System Cohort.
Diabetes Atorvastatin Lipid Intervention (DALI) study This double-blind randomized placebo-controlled multicenter study evaluated the effect of atorvastatin 10 mg versus 80 mg on lipid metabolism in 217 unrelated Dutch men and women with T2DM and diabetic dyslipidemia (fasting TG levels between 1.5 and 6.0 mmol/l) (29). Male and post-menopausal female patients, aged 4575 years, with T2DM (HbA1c 0.8 (see Supporting Information Figure 1 for the genomic context of SULF2). Allele-specific primers were obtained from Applied Biosystems (Life Technology, Rotterdam, The Netherlands). Genotyping was performed using allelic discrimination with FAM and VIC as fluorophores. PCR conditions were denaturation for 10 min at 95 C, followed by 40 cycles (30 sec 92 C, 45 sec 60 C) run on a CFX PCR system (BioRad). Tagman PCR assay mix was obtained from Applied Biosystems.
Oral fat tolerance in obese T2DM patients To study postprandial TRL metabolism, we performed an oral fat tolerance test in a subset of n 5 29 obese T2DM subjects, stratified
www.obesityjournal.org
Original Article
Obesity
OBESITY BIOLOGY AND INTEGRATED PHYSIOLOGY
Biochemical analyses Total cholesterol, HDL-cholesterol, LDL-cholesterol, and TGs were measured by standard enzymatic methods on a Cobas Mira system (Roche Diagnostics, Basel, Switzerland). Glucose was assessed using the hexokinase method (Gluco-quant, Hitachi 917; Hitachi). Plasma insulin was measured by an immunoluminimetric assay (Immulite insuline) on the Immulite 2000 autoanalyzer (Diagnostic Products). HbA1c was measured by HPLC (Reagens Bio-Rad Laboratories, Veenendaal, The Netherlands) on a Variant II (Bio-Rad Laboratories). Homeostatic model assessment for insulin resistance for glucose handling (HOMA-IR) was calculated by the widely-used formula: (glucose [mmol/l] 3 insulin [mU/l]/22.5) (27). RE were analyzed in 200 ll plasma after extraction of REs using chloroform/methanol/water as described (33,34). In short, retinyl propionate (Sigma Chemicals; St. Louis), a form that does not occur in humans, was used as internal standard; methanol was used as mobile phase at a flow rate of 1 ml/ min and the effluent was monitored at 330 nm. A standard curve of retinyl palmitate in pooled plasma was used as reference. Peak heights were measured and used for calculations of the absolute RE values.
Statistical analyses Clinical parameters passed tests for normality and are expressed as mean 6 standard deviation unless stated otherwise. Effects of single SNPs on TG levels were examined using one way analysis of variances (ANOVA). Variables with skewed distribution were logtransformed before being used as continuous variables in statistical analyses. Effects of SNPs on baseline characteristics were analyzed by ANOVA for continuous variables and with chi-square for categorical variables. Postprandial TG and RE were calculated as total area under the curve (AUC) calculated by the trapezoid rule. Incremental area under the curve (iAUC) was obtained by subtracting the fasting plasma level from each postprandial time point. Comparisons between groups were analyzed using one-way ANOVA, followed by pair-wise comparison using the Kruskal–Wallis test for multiple comparisons. Two-sided probability values of G) from the DALI and Diabetes Care System cohorts. Exclusion criteria were a history of manifest coronary artery disease, the use of any lipid-lowering medication in the 8 weeks preceding the fat tolerance test, clinical signs of malabsorption (e.g., diarrhea), or exogenous insulin treatment. Participants were asked to refrain from alcohol intake the day before. Participants fasted overnight and then were admitted at 7:30 am to the ward. Cream (consisting of 40% fat [wt/vol] with a polyunsaturated fat to saturated fat ratio of 0.06, 0.001% cholesterol [wt/ vol]), and carbohydrates (refined sugar) were administered to achieve a dosage of 35 g fat and 50 g sugar per m2 body surface. This mixture was supplemented with 150 ml water and 100,000 IU of vitamin A (Retinyl palmitate, AMC Clinical Pharmacy). The cream drink was consumed within 10 min. Postprandial blood samples were drawn at 0, 1, 2, 3, 4, 5, and 7 hr from a peripheral vein into EDTA-containing tubes, which were placed on ice and protected from light. Plasma was separated within 30 min by centrifugation at 1,700g for 20 min at 4 C. Aliquots of plasma were frozen at 280 C for subsequent analysis of TG and retinyl esters (REs).
www.obesityjournal.org
Results Characteristic metabolic abnormalities in obese subjects correlate with disturbed hepatic SULF2 expression We searched for evidence in humans that hepatic SULF2 mRNA expression is dysregulated as a part of the metabolic disturbances of obesity, particularly dyslipidemia. We studied a large consecutively recruited cohort of obese patients (n 5 121) who had undergone a liver biopsy before bariatric surgery or dietary intervention. Main characteristics of this cohort are presented in Table 1. Liver SULF2 expression was significantly associated with fasting plasma TGs (r 5 0.271; P 5 0.003), fasting glucose (r 5 0.252; P 5 0.005) and HOMA-IR (r 5 0.186; P 5 0.043; Figure 1), and with anthropometric parameters reflected by waist (P < 0.002) and waist-to-hip ratio (P < 0.001; Table 2).
Common genetic variants in SULF2 and fasting plasma TG levels in two independent cohorts of obese T2DM subjects To investigate whether genetic variation in SULF2 is associated with fasting plasma TG levels, we selected seven tagSNPs: rs2281279;
Obesity | VOLUME 22 | NUMBER 5 | MAY 2014
1311
Obesity
SULF2 and Triglycerides in Obesity Carlijne Hassing et al.
TABLE 1 Characteristics of the obesity cohort
Parameters
N Female/male Age (yr) Weight (kg) BMI (kg/m2) Waist circumference (cm) Waist-to-hip ratio Cholesterol (mmol/l) HDL cholesterol (mmol/l) LDL cholesterol (mmol/l) TGs (mmol/l) Glucose (mmol/l) Insulin (mIU/l) HOMA-IR HbA1c (mmol/mol) AST (U/l) ALT (U/l)
121 83/38 45 6 12 111 6 23 38.7 6 6.7 117 6 13 0.97 6 0.10 5.3 6 1.0 1.29 6 0.36 3.2 6 0.9 1.7 6 0.9 4.6 (4.3-5.05) 14 (10-22) 2.94 (2.04-4.40) 38 (34-40) 28 (24-36) 40 (32-53)
Data are presented as mean 6 SD for normally distributed variables or as median (interquartile range) when distribution of variable is skewed. AST, aspartate aminotransferase; ALT, alanine aminotransferase. LDL cholesterol was calculated by the Friedewald formula (cholesterol 2 HDL cholesterol 2 [TGs/5]) HOMA-IR was calculated as (glucose [mmol/l] 3 insulin [mIU/l])/22.5 (27).
c.24941267A>G, rs6090717: c.416-3507T>C, rs6090714: c.5671 4691G>A; rs6094818:-100-7141T>C, rs13044051:c.416-8793G>C, rs6122615: c.10111307T>C, and rs2235734:2227148T>G; (Supporting Information Figure 1). All pairwise metrics of LD between rs2281279 and other SNPs were
Obesity
OBESITY BIOLOGY AND INTEGRATED PHYSIOLOGY
SULF2 Strongly Prediposes to Fasting and Postprandial Triglycerides in Patients with Obesity and Type 2 Diabetes Mellitus H. Carlijne Hassing1, R. Preethi Surendran2, Bruno Derudas3, An Verrijken4, Sven M. Francque5, Hans L. Mooij1, Sophie J. Bernelot Moens1, Leen M.’t Hart6,7, Giel Nijpels8, Jacqueline M. Dekker9, Kevin Jon Williams10,11, Erik S. G. Stroes1, Luc F. Van Gaal4, Bart Staels3, Max Nieuwdorp1 and Geesje M. Dallinga-Thie1,2
Objective: Hepatic overexpression of sulfatase-2 (SULF2), a heparan sulfate remodeling enzyme, strongly contributes to high triglyceride (TG) levels in obese, type 2 diabetic (T2DM) db/db mice. Nevertheless, data in humans are lacking. Here, the association of human hepatic SULF2 expression and SULF2 gene variants with TG metabolism in patients with obesity and/or T2DM was investigated. Methods: Liver biopsies from 121 obese subjects were analyzed for relations between hepatic SULF2 mRNA levels and plasma TG. Associations between seven SULF2 tagSNPs and TG levels were assessed in 210 obese T2DM subjects with dyslipidemia. Replication of positive findings was performed in 1,316 independent obese T2DM patients. Postprandial TRL clearance was evaluated in 29 obese T2DM subjects stratified by SULF2 genotype. Results: Liver SULF2 expression was significantly associated with fasting plasma TG (r 5 0.271; P 5 0.003) in obese subjects. The SULF2 rs2281279(A>G) SNP was reproducibly associated with lower fasting plasma TG levels in obese T2DM subjects (P < 0.05). Carriership of the minor G allele was associated with lower levels of postprandial plasma TG (P < 0.05) and retinyl esters levels (P < 0.001). Conclusions: These findings implicate SULF2 as potential therapeutic target in the atherogenic dyslipidemia of obesity and T2DM. Obesity (2014) 22, 1309-1316. doi:10.1002/oby.20682
Introduction The prevalence of obesity, the metabolic syndrome, type 2 diabetes mellitus (T2DM) and their sequelae, particularly cardiovascular events, is increasing worldwide (1,2). The accelerated risk for atherosclerosis in these subjects results in large part from their atherogenic dyslipidemia, which includes increased fasting levels of very low density lipoprotein (VLDL) and its major component triglyceride (TG), in combination with impaired clearance of postprandial
triglyceride-rich lipoprotein (TRL) remnants (3-5). In particular, subendothelial retention of VLDL and particularly postprandial TRLremnants have been linked to cardiovascular events in these individuals (6-10). TRL metabolism involves a series of steps that culminate in the uptake of TRL-remnants by hepatocytes (8,11,12). During the past decades, hepatic heparan sulfate proteoglycans (HSPGs) (13-16) and in particular the syndecan-1 HSPG, have been implicated in TRL
1
Department of Vascular Medicine, Academic Medical Center, Amsterdam, The Netherlands 2 Department of Experimental Vascular Medicine, Academic Medical Center, Amsterdam, The Netherlands. Correspondence: G. M. Dallinga-Thie ([email protected]) 3 University of Lille 2; INSERM U1011; EGID; Institute Pasteur de Lille, Lille, France 4 Department of Endocrinology, Diabetology and Metabolism, Antwerp University Hospital, University of Antwerp, Antwerp, Belgium 5 Department of Gastroenterology and Hepatology, Antwerp University Hospital, University of Antwerp, Antwerp, Belgium 6 Department of Molecular Epidemiology, Leiden University Medical Center, Leiden, The Netherlands 7 Department of Molecular Cell Biology, Leiden University Medical Center, Leiden, The Netherlands 8 Department of General Practice, EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, The Netherlands 9 Department of Epidemiology and Biostatistics, EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, The Netherlands 10 Section of Endocrinology, Diabetes and Metabolism, Temple University School of Medicine, Philadelphia, Pennsylvania, USA 11 Department of Molecular and Clinical Medicine, Sahlgrenska Center for Cardiovascular and Metabolic Research, University of G€othenborg, Gothenburg, Sweden Funding agencies: This work was supported by the Dutch Heart Foundation (2010B242 to R.P.S.; CVON-GENIUS 2011-19 to S.J.B.M), the Netherlands Organization for Scientific Research (VENI grant 016.096.044; to M.N.) and by the National Institutes of Health (grants HL94277 and DK100851; to K.J.W.), FP6 Hepadip grant (L.V.G.), the FP7 Resolve grant (to M.N., B.S. and L.V.G.). Disclosure: The authors declared no conflict of interest. Author contributions: H. Carlijne Hassing and R. Preethi Surendran both contributed equally to this work. Additional Supporting Information may be found in the online version of this article. Received: 20 September 2013; Accepted: 9 December 2013; Published online 13 December 2013. doi:10.1002/oby.20682
www.obesityjournal.org
Obesity | VOLUME 22 | NUMBER 5 | MAY 2014
1309
Obesity
SULF2 and Triglycerides in Obesity Carlijne Hassing et al.
removal (17-21). The syndecan-1 HSPG consists of a single-pass transmembrane core protein onto which three heparan sulfate (HS) side-chains are covalently attached. These side-chains bind several important ligands, including TRLs, resulting in hepatic uptake and clearance of these particles. About 50 genes are involved in HSPG assembly and disassembly, affecting core protein expression, HS side-chain length, epimerization of glucuronyl residues, and sulfation patterns (22). It was recently shown that expression of glucosamine6-O-endosulfatase-2 (Sulf2 encoding for sulfatase 2, an HSPG remodeling enzyme) is substantially increased in livers of obese, T2DM db/db mice compared with lean controls (23). In these mice, hepatic Sulf2 mRNA expression was strongly and positively related to plasma TG levels and Sulf2 protein inhibited the ability of syndecan-1 to clear model remnants by cultured liver cells. Recently, we showed that suppression of hepatic Sulf2 expression in obese, T2DM db/db mice by targeted allele-specific antisense administration normalized fasting plasma TG levels and, most strikingly, eliminated postprandial hypertriglyceridemia (24). However, human data on the biology of hepatic SULF2 and its role in TG metabolism remain to be elucidated. In the present study, we sought to answer three questions. First, is there evidence in humans for dysregulation of hepatic SULF2 mRNA levels as part of the metabolic disturbances of obesity, particularly dyslipidemia? Second, are there SULF2 SNPs that associate with fasting dyslipidemia in cohorts of obese, T2DM subjects? Third, does any SULF2 SNP affect postprandial lipoprotein metabolism in obese subjects with T2DM, as assessed by oral fat tolerance studies? To address the first two questions, we used pre-existing cohorts, which have the advantage of large numbers of subjects, but the disadvantage that fasting—not postprandial—lipoprotein values are available. Because fat tolerance studies are laborious, they are practical only in a subset of individuals, not in population-wide cohorts. Our results implicate human SULF2 as an important participant in deranged TRL metabolism in obesity and T2DM.
(Agilent Technologies). Specific primers for SULF2 were designed using Primer3 software (50 -CCT TTG CCG TGT ACC TCA AT-30 and 50 -GCA CGT AGG AGC CGT TGT AT-30 ). mRNA levels were normalized to those of transcription factor IIb (TFIIb) by calculating the difference in cycle threshold (DCt).
SULF2 SNPs and fasting plasma TGs in cohorts of obese, T2DM patients Associations between SULF2 SNPs and plasma TG levels were assessed in participants in the Diabetes Atorvastatin Lipid Intervention (DALI) study and subsequently validated in the Diabetes Care System cohort, both described immediately below. All analysis were performed using the baseline data from either the DALI cohort or the Diabetes Care System Cohort.
Diabetes Atorvastatin Lipid Intervention (DALI) study This double-blind randomized placebo-controlled multicenter study evaluated the effect of atorvastatin 10 mg versus 80 mg on lipid metabolism in 217 unrelated Dutch men and women with T2DM and diabetic dyslipidemia (fasting TG levels between 1.5 and 6.0 mmol/l) (29). Male and post-menopausal female patients, aged 4575 years, with T2DM (HbA1c 0.8 (see Supporting Information Figure 1 for the genomic context of SULF2). Allele-specific primers were obtained from Applied Biosystems (Life Technology, Rotterdam, The Netherlands). Genotyping was performed using allelic discrimination with FAM and VIC as fluorophores. PCR conditions were denaturation for 10 min at 95 C, followed by 40 cycles (30 sec 92 C, 45 sec 60 C) run on a CFX PCR system (BioRad). Tagman PCR assay mix was obtained from Applied Biosystems.
Oral fat tolerance in obese T2DM patients To study postprandial TRL metabolism, we performed an oral fat tolerance test in a subset of n 5 29 obese T2DM subjects, stratified
www.obesityjournal.org
Original Article
Obesity
OBESITY BIOLOGY AND INTEGRATED PHYSIOLOGY
Biochemical analyses Total cholesterol, HDL-cholesterol, LDL-cholesterol, and TGs were measured by standard enzymatic methods on a Cobas Mira system (Roche Diagnostics, Basel, Switzerland). Glucose was assessed using the hexokinase method (Gluco-quant, Hitachi 917; Hitachi). Plasma insulin was measured by an immunoluminimetric assay (Immulite insuline) on the Immulite 2000 autoanalyzer (Diagnostic Products). HbA1c was measured by HPLC (Reagens Bio-Rad Laboratories, Veenendaal, The Netherlands) on a Variant II (Bio-Rad Laboratories). Homeostatic model assessment for insulin resistance for glucose handling (HOMA-IR) was calculated by the widely-used formula: (glucose [mmol/l] 3 insulin [mU/l]/22.5) (27). RE were analyzed in 200 ll plasma after extraction of REs using chloroform/methanol/water as described (33,34). In short, retinyl propionate (Sigma Chemicals; St. Louis), a form that does not occur in humans, was used as internal standard; methanol was used as mobile phase at a flow rate of 1 ml/ min and the effluent was monitored at 330 nm. A standard curve of retinyl palmitate in pooled plasma was used as reference. Peak heights were measured and used for calculations of the absolute RE values.
Statistical analyses Clinical parameters passed tests for normality and are expressed as mean 6 standard deviation unless stated otherwise. Effects of single SNPs on TG levels were examined using one way analysis of variances (ANOVA). Variables with skewed distribution were logtransformed before being used as continuous variables in statistical analyses. Effects of SNPs on baseline characteristics were analyzed by ANOVA for continuous variables and with chi-square for categorical variables. Postprandial TG and RE were calculated as total area under the curve (AUC) calculated by the trapezoid rule. Incremental area under the curve (iAUC) was obtained by subtracting the fasting plasma level from each postprandial time point. Comparisons between groups were analyzed using one-way ANOVA, followed by pair-wise comparison using the Kruskal–Wallis test for multiple comparisons. Two-sided probability values of G) from the DALI and Diabetes Care System cohorts. Exclusion criteria were a history of manifest coronary artery disease, the use of any lipid-lowering medication in the 8 weeks preceding the fat tolerance test, clinical signs of malabsorption (e.g., diarrhea), or exogenous insulin treatment. Participants were asked to refrain from alcohol intake the day before. Participants fasted overnight and then were admitted at 7:30 am to the ward. Cream (consisting of 40% fat [wt/vol] with a polyunsaturated fat to saturated fat ratio of 0.06, 0.001% cholesterol [wt/ vol]), and carbohydrates (refined sugar) were administered to achieve a dosage of 35 g fat and 50 g sugar per m2 body surface. This mixture was supplemented with 150 ml water and 100,000 IU of vitamin A (Retinyl palmitate, AMC Clinical Pharmacy). The cream drink was consumed within 10 min. Postprandial blood samples were drawn at 0, 1, 2, 3, 4, 5, and 7 hr from a peripheral vein into EDTA-containing tubes, which were placed on ice and protected from light. Plasma was separated within 30 min by centrifugation at 1,700g for 20 min at 4 C. Aliquots of plasma were frozen at 280 C for subsequent analysis of TG and retinyl esters (REs).
www.obesityjournal.org
Results Characteristic metabolic abnormalities in obese subjects correlate with disturbed hepatic SULF2 expression We searched for evidence in humans that hepatic SULF2 mRNA expression is dysregulated as a part of the metabolic disturbances of obesity, particularly dyslipidemia. We studied a large consecutively recruited cohort of obese patients (n 5 121) who had undergone a liver biopsy before bariatric surgery or dietary intervention. Main characteristics of this cohort are presented in Table 1. Liver SULF2 expression was significantly associated with fasting plasma TGs (r 5 0.271; P 5 0.003), fasting glucose (r 5 0.252; P 5 0.005) and HOMA-IR (r 5 0.186; P 5 0.043; Figure 1), and with anthropometric parameters reflected by waist (P < 0.002) and waist-to-hip ratio (P < 0.001; Table 2).
Common genetic variants in SULF2 and fasting plasma TG levels in two independent cohorts of obese T2DM subjects To investigate whether genetic variation in SULF2 is associated with fasting plasma TG levels, we selected seven tagSNPs: rs2281279;
Obesity | VOLUME 22 | NUMBER 5 | MAY 2014
1311
Obesity
SULF2 and Triglycerides in Obesity Carlijne Hassing et al.
TABLE 1 Characteristics of the obesity cohort
Parameters
N Female/male Age (yr) Weight (kg) BMI (kg/m2) Waist circumference (cm) Waist-to-hip ratio Cholesterol (mmol/l) HDL cholesterol (mmol/l) LDL cholesterol (mmol/l) TGs (mmol/l) Glucose (mmol/l) Insulin (mIU/l) HOMA-IR HbA1c (mmol/mol) AST (U/l) ALT (U/l)
121 83/38 45 6 12 111 6 23 38.7 6 6.7 117 6 13 0.97 6 0.10 5.3 6 1.0 1.29 6 0.36 3.2 6 0.9 1.7 6 0.9 4.6 (4.3-5.05) 14 (10-22) 2.94 (2.04-4.40) 38 (34-40) 28 (24-36) 40 (32-53)
Data are presented as mean 6 SD for normally distributed variables or as median (interquartile range) when distribution of variable is skewed. AST, aspartate aminotransferase; ALT, alanine aminotransferase. LDL cholesterol was calculated by the Friedewald formula (cholesterol 2 HDL cholesterol 2 [TGs/5]) HOMA-IR was calculated as (glucose [mmol/l] 3 insulin [mIU/l])/22.5 (27).
c.24941267A>G, rs6090717: c.416-3507T>C, rs6090714: c.5671 4691G>A; rs6094818:-100-7141T>C, rs13044051:c.416-8793G>C, rs6122615: c.10111307T>C, and rs2235734:2227148T>G; (Supporting Information Figure 1). All pairwise metrics of LD between rs2281279 and other SNPs were
