An Exonic Peroxisome Proliferator-Activated Receptor ...

J Nutrigenet Nutrigenomics DOI: 10.1159/000337352 Received: October 6, 2011 Accepted: February 14, 2012 Published online: $ $ $

© 2012 S. Karger AG, Basel 1661–6499/12/0000–0000$38.00/0 www.karger.com/jnn

1

Original Paper

An Exonic Peroxisome Proliferator-Activated Receptor-␥ Coactivator-1␣ Variation May Mediate the Resting Energy Expenditure through a Potential Regulatory Role on Important Gene Expression in This Pathway Khadijeh Mirzaei a, b Arash Hossein-nezhad a Solaleh Emamgholipour a Hasti Ansar a Mahtab Khosrofar a Ali Tootee a Soudabeh Alatab a a Bionanotechnology

Group, Endocrinology and Metabolism Research Institute, and b Department of Nutrition and Biochemistry, School of Public Health and Institute of Public Health Research, Tehran University of Medical Sciences, Tehran, Iran

Key Words PPARGC1A variation ⴢ Resting energy expenditure ⴢ Gene expression ⴢ MAPK ⴢ UCP2

Abstract Background/Aims: We studied peroxisome proliferator-activated receptor- coactivator-1 (PPARGC1A) gene variations at the 23815227–23815706 positions and examined their possible correlation with obesity-related conditions and resting energy expenditure (REE). We investigated the expression of PPARGC1A, mitogen-activated protein kinase (MAPK) and uncoupling protein 2 (UCP2), which play key roles in cellular energy expenditure, in a cellular model consisting of peripheral blood mononuclear cells, and compared them with various genotypes of the PPARGC1A gene. Methods: In total, 100 normal-weight and 129 obese subjects participated in the current study. All subjects were assessed for REE and body composition. We sequenced the PPARGC1A gene. Real-time PCR was used for determining the PPARGC1A, MAPK, and UCP2 gene expression. Results: There were significant differences in terms of body mass index, fat mass, low-density lipoprotein, insulin levels, REE/kg body weight, and REE/lean body mass among rs17574213 genotypes. There were significant differences in total cholesterol and low-density lipoprotein cholesterol levels among the various genotypes of Gly482Ser (rs8192678) and rs3755863. The relative PPARGC1A, MAPK, and UCP2 gene expressions had similar trends in the two studied SNPs, and the expression level of these genes was lowest in the TT genotype of Arash Hossein-nezhad, MD, PhD

JNN337352.indd 1

Endocrinology and Metabolism Research Institute Tehran University of Medical Sciences, Shariati Hospital North Kargar Ave., 5th Floor, IR–Tehran 14114 (Iran) Tel. +98 218 8822 0378, E-Mail ahosseinnezhad @ sina.tums.ac.ir

16.05.2012 08:47:19

J Nutrigenet Nutrigenomics DOI: 10.1159/000337352

© 2012 S. Karger AG, Basel www.karger.com/jnn

2

Mirzaei et al.: Energy Expenditure Regulation via PPARGC1A Variation in Obesity

Gly482Ser and rs3755863 and highest in the CC genotype of Gly482Ser and rs3755863. Conclusions: Our findings suggest that PPARGC1A variations may influence PPARGC1A expression and the coordinating regulators of downstream targets in energy homeostasis. Further study is needed to shed some light on this process. Copyright © 2012 S. Karger AG, Basel

Introduction

Energy homeostasis requires a balance between energy intake, storage, and expenditure. Any disturbance in this caloric equation can lead to metabolic malfunctions. For instance, obesity, as the leading cause of a significant number of complications and several pathologic conditions, is caused by defects in the regulation of energy homeostasis [1]. Peroxisome proliferator-activated receptor- coactivator-1 (PPARGC1A) plays a key role in the development of obesity and altered energy metabolism [2, 3], with a crucial impact on several aspects of cellular energy metabolic pathways such as mitochondrial biogenesis, cellular respiration, and the regulation of adaptive thermogenesis [4]. PPARGC1A has been mapped to the chromosome 4p15.1–2 [5] and spans over 13 exons [6]. It has been demonstrated that this chromosomal region has a strong relationship with obesity-related complications and contributes to the development of obesity [7]. For example, a linkage between this chromosomal region and an increased body mass index (BMI) in women from Utah pedigree studies [8], abdominal subcutaneous fat in the Quebec Family Study [9], and obesity indices in Mexican-Americans [10] has been documented. Therefore, it is plausible that the dysfunction of this protein contributes to the development of obesity and the consequent metabolic syndrome [6]. Since PPARGC1A activity may also be regulated by alternative splicing [7], it is very likely that the exon 8 sequence plays a key role in splice variants that significantly affect the function of the protein and alternative splicing [11]. The common coding single nucleotide polymorphisms (SNPs) rs17574213, rs8192678, rs3755863, rs58772979, rs17847360, and rs34074379 are located in exon 8, and specifically Gly482Ser results in amino acid changes. It is possible that these SNPs are associated with obesity and some obesity-related complications in different populations [12, 13]. The Gly482Ser SNP appears to be functional as transfection assays have demonstrated that it affects the protein’s efficiency [14], and carriers of various alleles of this SNP demonstrated significant differences in PGC-1a expression in several studied tissues [15]. Considering the key role of PPARGC1A in metabolism and energy expenditure, the regulation of its expression and activity seems to be of considerable importance [7]. The study of gene expressions in this pathway might shed some light on the potential role of PPARGC1A in metabolism and energy expenditure. There are three conserved Thr and Ser residues within the negative regulatory domain that are phosphorylated by p38 mitogen-activated protein kinase (MAPK) [16]. A MAPK-sensitive repressor mechanism is involved in the activation of some receptors by PPARGC1A [17]. This mechanism is also involved in the regulation of an important pathway at a multipartite response element within the promoter region of the gene encoding the uncoupling protein (UCP) 1 [18], which is one of the most important genes in energy homeostasis. Moreover, the association between gene variations and the levels of peripheral blood mtDNA has been clearly demonstrated and reported previously by Choi et al. [14]. We designed the current study in order to elucidate the potential underlying mechanism of the role of PPARGC1A variations and gene expressions in the regulation of cellular energy homeostasis in peripheral mononuclear cells as well as their correlation with resting energy expenditure (REE) in a clinical setting. To be precise, we investigated PPARGC1A gene variations at the 23815227–23815706 positions and their possible correlation with obesity-related

JNN337352.indd 2

16.05.2012 08:47:26



3


conditions, REE, and the homeostatic model assessment (HOMA) and quantitative insulin sensitivity check (QUICKI) indices. Moreover, we studied the expression of PGC-1, MAPK, and UCP2, which play key roles in cellular energy expenditure, in a cellular model in peripheral blood mononuclear cells (PBMC) and compared their expression in various genotypes of the PPARGC1A gene. Subjects and Methods Study Population A total of 229 participants were recruited for this case-control study. Of the 229 participants, 47 (20.52%) had a normal BMI (18.5 ^ BMI ! 25), 53 (23.14%) were overweight (25 ^ BMI ! 30), 48 (20.96%) were class I obese (BMI 30–34.9), and 81 (35.37%) of the participants were class II and III obese (BMI 635). The proportion of men was 11.79% (n = 27) and that of women was 88.2% (n = 202). The study had the approval of the Local Ethics Committee of the Endocrinology and Metabolism Research Center of Tehran University of Medical Sciences. All participants were recruited from a nutrition clinic of the Shariati Hospital’s outpatient clinic. The registered patients in the clinic were enrolled in our study according to inclusion and exclusion criteria. Therefore, every obese subject registered in the nutrition clinic who met the entrance requirements for study enrolment was invited to participate in this study. Patients were chosen according to our defined inclusion criteria, which were a BMI 630, age 22–52 years, absence of any acute or chronic inflammatory disease, no history of hypertension, no alcohol or drug abuse, and not being pregnant. Exclusion criteria included a history of any condition affecting inflammatory markers such as known cardiovascular disease, thyroid diseases, malignancies, current smoking, diabetes mellitus, sustained hypertension, heart failure, acute or chronic infections, and hepatic or renal diseases. Subjects maintained their usual diet and refrained from vigorous exercise for 2 days prior to the study. All participants gave written informed consent before any study procedure was performed. It is noteworthy that there were no significant differences in terms of gender and age between the cases and the control group. Hypertriglyceridemia was defined as a triglyceride level of 1150 mg/dl [19]. Biochemical Parameters and Hormonal Assay All baseline blood samples were obtained between 8:00 and 10:00 a.m. following an overnight fasting. The serum was centrifuged, aliquoted, and stored at –80ºC. All samples were analyzed by means of a single assay. All measurements were performed at the Endocrinology and Metabolism Research Institute (EMRI) laboratory of the Shariati Hospital. The GOD-PAP method was used for the measurement of fasting serum glucose, and triglyceride levels were measured by the GPO-PAP method. Total cholesterol levels were measured by the enzymatic endpoint method, and direct high-density lipoprotein (HDL) cholesterol levels were measured using an enzymatic clearance assay. All measurements were done with the use of a Randox Laboratories kit (Hitachi 902). Serum hypersensitivity C-reactive protein (hs-CRP), a wellknown marker of inflammation, was measured with the use of an immunoturbidimetric assay (highsensitivity assay, Hitachi 902). HOMA and QUICKI Calculations Insulin resistance (IR) was calculated using the HOMA model. The HOMA IR was calculated according to the following equation: HOMA IR = [fasting plasma glucose (mmol/l) ! fasting plasma insulin (mIU/l)]/22.5 [20]. QUICKI was calculated based on: ISQUICKI = 1/[log (fasting insulin) + log (fasting glucose)] [21]. REE Measurements REE measurements were performed by expert nutritionists using a standard protocol. Body composition of all cases was accurately assessed and reported with the use of the body composition analyzer BC418MA-Tanita before REE measurement. Height was measured with the use of Seca 200 Girth Measuring Tape, and the established protocol was followed (without shoes; heels together; subject’s heels, buttocks, shoulders, and head touching the vertical wall surface; and with a horizontal line of sight). REE was measured using FitmateTM calorimeter (Cosmed Company, Rome, Italy). We used the same procedure as described previously [22].

JNN337352.indd 3

16.05.2012 08:47:26

J Nutrigenet Nutrigenomics © 2012 S. Karger AG, Basel www.karger.com/jnn

DOI: 10.1159/000337352

4


Subjects were asked to fast and to remain in a resting state for 12 h before the test and to abstain from smoking for 61 h before testing, although the ideal interval would be 12 h to make sure that the body was in a resting and post-absorptive state. Patients were instructed to lie in a supine position on a mattress for 15 min, and then the measurement was performed for a period of 20 min. However, the first 5 min were not included, and only the last 15 min were used to calculate the REE. Complete Body Composition Analysis We assessed the body composition of all cases with the use of the body composition analyzer BC418MA-Tanita (UK). This equipment is designed to send out a very weak electric current to measure the impedance (electrical resistance) of the body. Therefore, in principle, subjects were barefoot when they were assessed using this device. Moreover, since impedance fluctuates in accordance with the distribution of the body fluid, we followed all of the following instructions for an accurate measurement. To prevent a possible discrepancy in measured values, we avoided taking measurements after vigorous exercise and waited until the subject was sufficiently rested. To prevent inaccurately low body fat percentage measurements and other measurement errors, we always held both arms straight down when taking measurements. As changes in body water distribution and body temperature can have a major impact on measurement outcomes, the measurements were performed in the morning in a fasting and empty bladder condition to get more accurate results. In all cases, we ensured that the subjects’ arms were not touching her/his side and that the inner thighs were not touching each other during measurements; if necessary, we placed a dry towel between their arms and sides and/or between their thighs. Also, we made sure the soles of the feet were free from excessive dirt or crust as this may also act as a barrier for the mild current. We were aware that false results might be reported after excessive food/fluid intake or after periods of intense exercise. All participants enjoyed healthy lives with a regular lifestyle. We did not take measurements while using transmitters, such as mobile phones, as they might affect the readings. The device calculates the body fat percentage, fat mass, and fat-free mass and predicts the muscle mass on the basis of data obtained by dual-energy X-ray absorptiometry using bioelectrical impedance analysis [23]. DNA Extraction and Sequencing of the Gene The extraction of genomic DNA from blood samples was carried out with the use of the FlexiGene DNA kit (QIAGEN GmbH, Germany), according to the manufacturer’s protocol. After sampling, DNA of blood samples was extracted. The extracted DNA was stored at 4 ° C up to 1 week before sequencing was performed. We sequenced 479 bp of the PPARGC1A gene from 23815227 to 23815706 positions, which were amplified using specific primers. The primer sequences were as follows: forward primer 5ⴕ-CAG TTT TTC AGG CCT TGT CA-3ⴕ and reverse primer 5ⴕ-GCT GTT TTC TGC TGC AA-3ⴕ. We sequenced the PPARGC1A gene using the ABI PRISM 3730 automated sequencer (Applied Biosystems, Foster City, Calif., USA). This region was sequenced for the accurate detection of SNPs and the exact assignment of genotypes of rs17574213, Gly482Ser, and rs3755863. For confirming the sequence results, the reverse strand of the studied gene was sequenced.

PBMC Preparation PBMC were obtained from venous blood samples (12 ml, heparin tube). Monocytes were isolated from peripheral blood by Ficoll density gradient centrifugation. To summarize, the blood was diluted to a 1:1 solution with RPMI 1640 supplemented with Glutamax-I, 25 mM HEPES (Cambrex Bio Science, Verviers, Belgium) and appropriate antibiotics (100 U/ml penicillin and 100 mg/ml streptomycin). Twentyfour milliliters of the solution was laid on top of a 15-ml high-density solution (LymphoprepTM; AxisShield PoCAS, Oslo, Norway) and was centrifuged at a speed of 1,500 rpm for 20 min. The middle layer, which consisted of PBMC, was separated and washed twice. RNA Extraction and Quantitative Real-Time RT-PCR At every turn, 106 cells were harvested, and mRNA extraction was performed using the High-Pure RNA isolation kit (Roche Diagnostics). PPARGC1A, MAPK, PPARg, UCP, and -actin genes were quantified by means of quantitative real-time RT-PCR as described in detail elsewhere. In summary, 0.5–1 mg total RNA was reverse transcribed using the RevertAid First Strand cDNA synthesis kit (Fermentase). For subsequent RT-PCR amplification, a maximum of 2 l of each cDNA sample was used per 20-l PCR mix. PCR reactions were performed either in triplicate wells or in an ABI StepOne real-time PCR system (Ap-

JNN337352.indd 4

16.05.2012 08:47:27



5


plied Biosystems). Reagents needed for the PCR reaction were purchased from Applied Biosystems, and each sample was run in triplicates in a volume of 20 l for up to 40 cycles using standard real-time cycling conditions. The Ct values were normalized to a relative standard curve run at the same plate as the samples. -actin has been frequently considered as a constitutive housekeeping gene for RT-PCR, and has been used to normalize changes in specific gene expressions. The relative quantification of the target gene expression was normalized to the expression of a housekeeping gene (-actin) for each sample. Data was reported as fold changes in mRNA levels compared with the internal control. Primers right: 5ⴕ-CAA GCC AAA CCA ACA ACT TTA TCT CT-3ⴕ and left: 5ⴕ-CAC ACT TAA GGT GCG TTC AAT AGT C-3ⴕ were used to amplify the product of the PPARGC1A locus. Primers right: 5ⴕ-GCG CTA CAC CAA CCT CTC GT-3ⴕ and left: 5ⴕ-CAC GGT GCA GAA CGT TAG CTG-3ⴕ were used to amplify the product of the MAPK locus. Primers right: 5ⴕ-TCT ACA ATG GGC TGG TGG C-3ⴕ and left: 5ⴕ-TGT ATC TCG TCT TGA CCA C-3ⴕ were used to amplify the product of the UCP2 locus. Primers right: 5ⴕ-TCT TTG ATG TCA CGC ACG ATT T-3ⴕ and left: 5ⴕ-GGA CCT GAC GGA CTA CCT CA-3ⴕ were used to amplify the product of the -actin gene. Statistical Analyses Overall group differences were assessed by ANOVA. The 2 test was used for comparing the frequency of variables between groups. Results were reported as target PPARGC1A, MAPK, and UCP2 quantity divided by -actin quantity. For sample normalization, -actin was used as a housekeeping gene. Dependent variables in the genotype groups were compared using ANOVA. We used the logistic regression model to evaluate the relationship between studied genetic variation and obesity independent of other variables. In addition, in the univariate model, BMI and gender were entered as fixed factors, and the effects of covariates on dependent variables were investigated. In all statistical analyses, a significant difference was detected when p ! 0.05. Statistical analyses were performed using SPSS version 16.0 (Chicago, Ill., USA).

Results

Study Population Characteristics Study population characteristics, body dimensions, and laboratory measurements in each group are summarized in table 1 according to the standard definition of obesity, which is based on fat percentage (135% in women and 125% in men) and defined by the World Health Organization (WHO). Accordingly, of all participants, 100 (43.66%) were considered normal weight and 129 (56.33%) were categorized as obese. We found no statistically significant differences in terms of sex distribution and mean age between the different groups of participants. As shown in table 1, the study groups showed significant differences in terms of BMI, basal metabolic rate predicted, fat proportion, fat mass, fat-free mass, visceral fat mass, waist and hip circumference, fasting serum glucose, triglyceride, low-density lipoprotein (LDL) cholesterol, hs-CRP, fasting insulin, resting metabolic rate (RMR), and RMR/kg body weight. We did not detect any significant difference in HDL cholesterol levels between the two different groups. PPARGC1A Gene Sequencing and the Frequency of Studied Polymorphisms We sequenced the 23815227–23815706 positions on the PPARGC1A gene; there were 6 SNPs in the target amplicon. The frequencies of the genotypes in all studied SNPs in our population were in Hardy-Weinberg equilibrium. The frequencies of the genotypes of SNP rs17574213 were 93.1% homozygous major allele (GG) and 6.9% heterozygous allele (GA). Regarding Gly482Ser, the frequencies of the genotypes were 46.28% homozygous major allele (CC), 43.66% heterozygous allele (CT), and 10.04% homozygous minor allele (TT). The frequencies of the genotypes of SNP rs3755863 were 36.68% homozygous major allele (CC), 44.97% heterozygous allele (CT), and 18.34% homozygous minor allele (TT). Tables 2–4 demonstrate the mean values for each phenotype according to the genotype, and the geno-

JNN337352.indd 5

16.05.2012 08:47:27

J Nutrigenet Nutrigenomics © 2012 S. Karger AG, Basel www.karger.com/jnn

DOI: 10.1159/000337352

6


Table 1. Characteristics of participants in the two groups with respect to obesity

Characteristic

Age, years BMI BMR predicted, kcal/24 h Fat percent, % Fat mass, kg Fat-free mass, kg Visceral fat, kg WC, cm HC, cm Fasting serum glucose, mg/dl TG, mg/dl T-chol, mg/dl HDL, mg/dl LDL, mg/dl hs-CRP, mg/l Insulin, lU/ml RMR, kcal/24 h/kg RMR/kg, kcal/24 h/kg

Participants

p value*

obese1 (n = 129)

normal weight (n = 100)

38.93811.75 33.4984.69 1,567.298322.29 40.5485.44 35.428688.99 51.324488.45 9.4183.11 103.12811.39 118.71812.64 106.31833.91 136.26856.06 179.6835.34 42.8810.27 103.81824.94 4.8485.94 14.3689.88 1,633.588312.19 19.3483.06

37.84811.31 24.3083.48 1,347.38513.82 26.3986.29 17.38085.33 48.39811.15 4.0182.66 81.8789.99 96.3688.88 93.38816.38 101.34844.68 166.33829.93 44.23810.23 92.58823.36 1.2181.16 8.6984.47 1,514.058347.63 21.7783.7

0.3