The Genetics of Obesity the Genetics of Obesity Journal of Obesity

Journal of Obesity

The Genetics of Obesity The Genetics of Obesity Journal of Obesity

This is a focus issue published in “Journal of Obesity.” All articles are open access articles distributed under the Creative Commons At- tribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Editorial Board

DavidB.Allison,USA Xu Feng Huang, Australia Jonatan R. Ruiz, Sweden B. J. Ammori, UK Terr y Huang , USA Jordi Salas-Salvado,´ Spain Marco Anselmino, Italy Gianluca Iacobellis, Canada Francesco Saverio Papadia, Italy Molly S. Bray, USA Lauren E. Lissner, Sweden J. C. Seidell, The Netherlands Bernhard Breier, New Zealand Yannis Manios, Greece G. Silecchia, Italy Eliot Brinton, USA Claude Marcus, Sweden Laurence Tecott, USA Yvon Chagnon, Canada Ron F. Morrison, USA Serena Tonstad, Norway Karen Charlton, Australia Michael M. Murr, USA Paul Trayhurn, UK Eric Doucet, Canada Tomoo Okada, Japan Rob Martinus Van Dam, Singapore Pietro Forestieri, Italy Renato Pasquali, Italy Youfa Wang, USA Jayne Fulkerson, USA Mark A. Pereira, USA Aron Weller, Israel Jesus´ M. Garagorri, Spain Angelo Pietrobelli, Italy Aimin Xu, Hong Kong Tiﬀany L. Gary-Webb, USA R. Prager, Austria Jack A. Yanovski, USA Andras Hajnal, USA Denis Richard, Canada Alfredo Halpern, Brazil Robert J. Ross, Canada Contents

The Role of Leptin in Antipsychotic-Induced Weight Gain: Genetic and Non-Genetic Factors, Fabio Panariello, Gina Polsinelli, Carol Borlido, Marcellino Monda, and Vincenzo De Luca Volume 2012, Article ID 572848, 7 pages

High Levels of Cardiovascular Risk Factors among Pakistanis in Norway Compared to Pakistanis in Pakistan, Naeem Zahid, Haakon E. Meyer, Bernadette N. Kumar, Bjørgulf Claussen, and Akhtar Hussain Volume 2011, Article ID 163749, 5 pages

Mendelian Randomisation Study of Childhood BMI and Early Menarche, Hannah S. Mumby, Cathy E. Elks, Shengxu Li, Stephen J. Sharp, Kay-Tee Khaw, Robert N. Luben, Nicholas J. Wareham, Ruth J. F. Loos, and Ken K. Ong Volume 2011, Article ID 180729, 6 pages

An Obesity Risk SNP (rs17782313) near the MC4R Gene Is Associated with Cerebrocortical Insulin Resistance in Humans, Otto Tschritter, Axel Haupt, Hubert Preissl, Caroline Ketterer, Anita M. Hennige, Tina Sartorius, Fausto Machicao, Andreas Fritsche, and Hans-Ulrich Haring¨ Volume 2011, Article ID 283153, 4 pages

Sequence Analysis of the UCP1 Gene in a Severe Obese Population from Southern Italy, Giuseppe Labruna, Fabrizio Pasanisi, Giuliana Fortunato, Carmela Nardelli, Carmine Finelli, Eduardo Farinaro, Franco Contaldo, and Lucia Sacchetti Volume 2011, Article ID 269043, 4 pages

Genetics of Childhood Obesity, Jianhua Zhao and Struan F. A. Grant Volume 2011, Article ID 845148, 9 pages

Studies of Gene Variants Related to Inﬂammation, Oxidative Stress, Dyslipidemia, and Obesity: Implications for a Nutrigenetic Approach, Maira Ladeia R. Curti, Patri’cia Jacob, Maria Carolina Borges, Marcelo Macedo Rogero, and Sandra Roberta G. Ferreira Volume 2011, Article ID 497401, 31 pages

Variations in Adipokine Genes AdipoQ, Lep, and LepR Are Associated with Risk for Obesity-Related Metabolic Disease: The Modulatory Role of Gene-Nutrient Interactions, Jennifer Emily Enns, Carla G. Taylor, and Peter Zahradka Volume 2011, Article ID 168659, 17 pages

Associations of FTO and MC4R Variants with Obesity Traits in Indians and the Role of Rural/Urban Environment as a Possible Eﬀect Modiﬁer, A. E. Taylor, M. N. Sandeep, C. S. Janipalli, C. Giambartolomei, D. M. Evans, M. V. Kranthi Kumar, D. G. Vinay, P. Smitha, V. Gupta, M. Aruna, S. Kinra, R. M. Sullivan, L. Bowen, N. J. Timpson, G. Davey Smith, F. Dudbridge, D. Prabhakaran, Y. Ben-Shlomo, K. S. Reddy, S. Ebrahim, and G. R. Chandak Volume 2011, Article ID 307542, 7 pages

Can Thrifty Gene(s) or Predictive Fetal Programming for Thriftiness Lead to Obesity?, Ulfat Baig, Prajakta Belsare, Milind Watve, and Maithili Jog Volume 2011, Article ID 861049, 11 pages Diﬀerential Eﬀects of Calorie Restriction and Exercise on the Adipose Transcriptome in Diet-Induced Obese Mice, Karrie E. Wheatley, Leticia M. Nogueira, Susan N. Perkins, and Stephen D. Hursting Volume 2011, Article ID 265417, 13 pages

Microarray Evidences the Role of Pathologic Adipose Tissue in Insulin Resistance and Their Clinical Implications, Sandeep Kumar Mathur, Priyanka Jain, and Prashant Mathur Volume 2011, Article ID 587495, 16 pages

Genetic Variance in Uncoupling Protein 2 in Relation to Obesity, Type 2 Diabetes, and Related Metabolic Traits: Focus on the Functional −866G>A Promoter Variant (rs659366), Louise T. Dalgaard Volume 2011, Article ID 340241, 12 pages

Relationships of Adrenoceptor Polymorphisms with Obesity, Kazuko Masuo and Gavin W. Lambert Volume 2011, Article ID 609485, 10 pages

Interleukin-15, IL-15 Receptor-Alpha, and Obesity: Concordance of Laboratory Animal and Human Genetic Studies, LeBris S. Quinn and Barbara G. Anderson Volume 2011, Article ID 456347, 8 pages

Rs9939609 Variant of the Fat Mass and Obesity-Associated Gene and Trunk Obesity in Adolescents, Harald Mangge, Wilfried Renner, Gunter Almer, Daniel Weghuber, Reinhard Moller,¨ and Renate Horejsi Volume 2011, Article ID 186368, 4 pages

Gene by Sex Interaction for Measures of Obesity in the Framingham Heart Study, Ashlee M. Benjamin, Sunil Suchindran, Kaela Pearce, Jennifer Rowell, Lillian F. Lien, John R. Guyton, and Jeanette J. McCarthy Volume 2011, Article ID 329038, 8 pages Hindawi Publishing Corporation Journal of Obesity Volume 2012, Article ID 572848, 7 pages doi:10.1155/2012/572848

Review Article TheRoleofLeptininAntipsychotic-InducedWeightGain: Genetic and Non-Genetic Factors

Fabio Panariello,1 Gina Polsinelli,2 Carol Borlido,2 Marcellino Monda,3 andVincenzoDeLuca2, 4

1 Spedali Civili Brescia, Department Mental Health, 25123 Brescia, Italy 2 Centre for Addiction and Mental Health, room 30, 250 College street, Toronto, ON, Canada M5T 1R8 3 Dipartimento di Medicina Sperimentale, Seconda Universita` degli Studi di Napoli, 80131 Napoli, Italy 4 Department of Psychiatry, University of Toronto, Toronto, ON, Canada M5T 1R8

Correspondence should be addressed to Vincenzo De Luca, vincenzo [email protected]

Received 2 January 2011; Revised 3 October 2011; Accepted 13 October 2011

Academic Editor: Angelo Pietrobelli

Copyright © 2012 Fabio Panariello et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Schizophrenia is a chronic and disabling mental illness aﬀecting millions of people worldwide. A greater proportion of people with schizophrenia tends to be overweight. Antipsychotic medications have been considered the primary risk factor for obesity in schizophrenia, although the mechanisms by which they increase weight and produce metabolic disturbances are unclear. Several lines of research indicate that leptin could be a good candidate involved in pathways linking antipsychotic treatment and weight gain. Leptin is a circulating hormone released by adipocytes in response to increased fat deposition to regulate body weight, acting through receptors in the hypothalamus. In this work, we reviewed preclinical, clinical, and genetic data in order to infer the potential role played by leptin in antipsychotic-induced weight gain considering two main hypotheses: (1) leptin is an epiphenomenon of weight gain; (2) leptin is a consequence of antipsychotic-induced “leptin-resistance status,” causing weight gain.

1. Background The increased use of atypical antipsychotics over the last decade has raised concerns about their metabolic side effects, Schizophrenia has a worldwide prevalence of about one per- such as weight gain, diabetes, and dyslipidemia [4]. In ad- cent and has the potential for devastating emotional, phys- dition, medication-induced weight gain has been associated ical, and mental consequences [1]. Antipsychotic drugs are with a lower quality of life and noncompliance, which in- the first line of treatment for those with schizophrenia and creases the risk for relapse. The higher risk of cardiovascular other psychoses [2]. There are two classes of antipsychotic disease, leading to enhanced morbidity, mortality, significant medications used for treatment referred to as either typical or atypical. Typical antipsychotic drugs, like haloperidol, act economic cost, reduced quality of life, and lower compliance as high-affinity antagonists for dopamine 2 (D2) receptors to treatment can be linked to the obesitogenic and diabeto- with a possibility of extrapyramidal side effects. genic capacity of these drugs [5]. Second-generation antipsychotics (SGAs), like olanzap- However, the mechanisms of weight gain and dyslipi- ine and clozapine, have lower incidences of extrapyramidal demia are poorly understood, and various parts of the en- side effects than typical antipsychotics because they are po- docrine system are presumably involved in these side effects. tent antagonists of serotonin 2A (5-HT2A) rather than D2 Several lines of evidence indicate that antipsychotic drugs receptor antagonists, with a higher affinity for the former. elicit weight gain in some, but not all, individuals, suggesting These drugs also inhibit the G-protein-coupled receptors for a genetic predisposition [5]. Several hypotheses with regard several other biogenic amines including cholinergic, adrener- to the biological mechanism underlying differential genetic gic (α), histaminergic (H), and dopaminergic receptors [3]. liability have been suggested, mostly linked to genes which 2 Journal of Obesity regulate appetite and food intake caused by SGA-induced (NPY)/agoutirelated protein (AgRP) neurons and the pro- stimulation of 5-HT1A and 5-HT6 receptors, as well as α2 opiomelanocortin (POMC)/cocaine- and amphetamine-re- and muscarinic M3 receptors. Additionally, blocking recep- lated transcript (CART) neurons [25, 26]. These two groups tors such as the 5-HT2C, 5-HT1B, α1, H1 are known to result of neurons interact with “second-order” neurons in other in additional weight gain [6–9]. hypothalamic regions such as the paraventricular nucleus Not surprisingly, hyperphagia is associated with weight (PVN), dorsomedial hypothalamus (DMH), ventrome- gain in humans. Accordingly, it has been demonstrated that dial hypothalamus (VMH), and lateral hypothalamic area olanzapine-induced weight gain is caused by increased (LHA). caloric intake, rather than alterations in basal energy expend- Another key region implicated in research pertaining iture [10]. It has been suggested that antipsychotics interact to weight and energy homeostasis is the caudate putamen with the complex system of neurotransmitters, neuropep- (CPu). tides, and other modulators in brain neuronal circuits in- Dopamine (DA) neurons in the CPu have been shown volving the hypothalamus and brain stem where a neuropep- to play a role in maintaining food intake and hunger levels tidergic network mediates the actions of leptin and ghrelin [29–33]. Indeed, imaging studies have shown that severely to provoke disturbances in energy homeostasis, endocrine obese individuals have decreased striatal D2 receptor (D2R) alterations, and body weight (BW) control [11]. availability, and leptin-receptor-deficient obese rodents also Recent studies have shown that the hormones leptin and show decreased D2R binding in striatum [34, 35]. In con- ghrelin are crucial elements of the hypothalamic neurocir- trast, chronic food restriction showed greater striatal D2R cuitry. Leptin, a 16 kD peptide, is a cytokine-like molecule binding relative to ad libitum fed rats [35–37]. Genetic synthesized in white adipose tissue [12–14]. Leptin is a highly studies, although not always consistent, have reported that hydrophilic 167 amino acid protein and it is transported individuals carrying the Taq I A1 allele of the D2R gene, across the blood brain barrier by binding to the short form which was associated with decrease in D2R in striatum of the leptin receptor [14, 15] and is actively transported into by some investigators, are more vulnerable to addictive the hypothalamus, where it acts to limit food intake. It is a behaviors such as compulsive food intake and are more likely product of the obese (OB) gene located on chromosome 7 to be obese [38, 39]. Leptin receptors are also present in the (7q31.3). Receptors for leptin are widely expressed through- ventrotegmental area (VTA) and leptin targets dopaminergic out the central nervous system (CNS), but the major target and gammaaminobutyric acid neurons in this region critical of this hormone is the medial hypothalamus [16]. Although to brain reward circuits, inducing phosphorylation of signal the leptin receptor is present as different splice-variant iso- transducer and activator of transcription. Direct adminis- forms in the CNS, the form OBRb has the major role in tration of leptin in the VTA causes decreased food intake its metabolic action [12, 16]. The obese db/db mouse and [40, 41]. Zucker fa/fa rat represent naturally occurring “knockouts” In humans, DA metabolite concentrations in the CSF of the leptin receptor, that have helped to validate the im- decrease as leptin increases which could reflect inhibition of portance of CNS leptin action in energy homeostasis [17]. DA release by leptin. Indeed, preclinical studies have shown Exogenous leptin reduces appetite and feeding while leptin that short-term leptin treatment decreases both DA release deficiency (both mice and humans with mutations in the and concentration in NAc (nucleus accumbens) and leptin gene) causes extreme obesity and can lead to reproduc- hypothalamus in a dose-dependent manner [41–43]. How- tive problems, bone formation deficiency, and cardiovascular ever, it is postulated that leptin may exert at least part of its complications [18–22]. It appears more involved in long- influence through a pathway linked to H1-receptor, reducing term regulation of energy, being released into the circulatory food intake [44]. In fact leptin-induced food intake appears system as a function of energy stores. suppressed in H1 knockout mice [45, 46]. This data suggests that hypothalamic histamine is a modulator of leptin activity. In particular, leptin has multiple effects on energy ho- Hence, it is conceivable that antipsychotic drugs with a meostasis through activation of key hypothalamic nuclei and high affinity for H1-receptors could disarrange this pathway, peptides to regulate energy balance. Leptin directly activates inducing or exacerbateing a resistance to leptin action. proopiomelanocortin (POMC) cells in the arcuate nucleus Beside the very rare case of genetic leptin deficiency, the (ARC) to increase the release of melanocortin peptides in- vast majority of obese humans have high plasma leptin concluding the POMC product α-melanocyte-stimulating hor- α centrations related to the size of adipose tissue. However, mone ( -MSH). Melanocortin peptides inhibit food intake this elevated leptin signal does not induce expected responses and regulates metabolism via energy storage, insulin secre- (i.e., a reduction in food intake and an increase in energy tion, and gastrointestinal motility predominantly through expenditure), thus suggesting that most obese human sub- projections to MC4 receptor neurons [23–28]. jects are resistant to the effects of endogenous leptin [47]. Moreover, leptin also directly inhibits ARC, which pro- There is some evidence in regard of a “suppressor of cytokine duces agouti-related protein and neuropeptide Y. Hypotha- signaling-3” (SOCS-3) that could act as an inhibitor of leptin lamic ARC neurons—referred to as “first-order” neurons— signaling [48, 49]. High leptin levels could determine an receive signals from the blood via the median eminence, and increase of SOC-3 that may in turn lead to resistance to leptin from the cerebrospinal fluid (CSF) via the third ventricle. action [49]. The ARC contains two populations of neurons particu- Thus, leptin has been intensively investigated with larly relevant to feeding and satiety; the neuropeptide Y respect to its association with changes in weight and glucose Journal of Obesity 3 metabolism during treatment with various antipsychotics olanzapine in seven Japanese patients with schizophrenia. (APs). In that regard, our team reviewed longitudinal and They found that the SLLs were significantly increased, but cross-sectional studies which evaluated the effects of AP BW and BMI did not vary significantly after 6 months of on leptin concentration, in addition to literature assessing treatment [56]. Hosojima et al. (2006) recruited 13 patients genetic factors causing risk for fluctuations in leptin levels with schizophrenia in monotherapy with olanzapine for 4 and risk for weight gain in response to AP treatment. weeks [57]. As seen in previous studies, SLL increased from baseline to week 4. These changed in leptin levels appear to 2. Clozapine Treatment be quite rapid, and Wang et al. (2006) were the first to report these very early changes in leptin levels. They found a rapid Clozapine has been a major focus of this research. Bromel increment in the PLLs at the fourth hour after the beginning et al. (1998) [50] evaluated serum leptin levels (SLLs) in of the olanzapine treatment in 9 schizophrenic patients. This 12 patients (9 schizophrenic and 3 schizoaffective disorder increment stayed stable in week 2 [58]. patients) in clozapine treatment. In this study, only 8 patients Interestingly, Pena˜ et al. (2008) observed a significant were additionally treated with conventional neuroleptics and increment in SLL at week 8 but not in week 16 versus baseline other psychotropic drugs. The authors reported that after in a sample of 60 olanzapine-treated patients after a switch clozapine treatment, patients’ SLLs differed significantly from typical antipsychotics. They did, however, find a sig- from levels measured afterward (P<.0001), with concen- nificant weight gain (WG) and BMI increase that paralleled trations at least doubling in 8 clozapine-treated psychotic leptin levels at week 8 [59]. Popovic et al. (2007) evaluated patients, acutely evaluated (week 2) versus baseline, and no the modifications of SLL in 13 clozapine or risperidone- significant changes in more chronic treatment (up to the treated schizophrenic patients switched by conventional AP 10th week versus week 2). The net differences in BW and during 12 weeks [60]. A significant increase in SLL was body mass index (BMI) between baseline at week 2 revealed observed at week 12. However, leptin concentrations in 18 positive correlations to the relative increases in the SLL. patients treated with conventional AP were not different as Kivircik et al. (2003) found that the analysis of variance on compared with 20 healthy controls. In a longitudinal study, 19 clozapine-treated schizophrenic in- and outpatients who Eder et al. (2001) observed a concomitant significant increase completed 10 weeks of treatment did not reveal any signif- in SLL over baseline and olanzapine-induced WG over icant change in plasma leptin levels (PLLs), however, they baseline during an 8-week period in schizophrenic inpatients reported that the patient gained 5% of their BW, with a (n = 10), assigned to monotherapy with olanzapine and significant increase in BMI [51]. compared with healthy subjects [61]. Similarly, Atmaca et al. Theisen et al. (2005) investigated SLL and BMI in 12 (2007) evaluated schizophrenic patients (n = 21), enrolled clozapine treated patients with schizophrenia or schizoaffec- in olanzapine monotherapy during 6 weeks, and compared tive disorders over a 10-week drug assumption (8 patients with healthy controls (n = 21) [62]. These investigators were additionally treated with other AP, benzodiazepines, reported that leptin levels were increased from the baseline in and/or antidepressant). The investigators observed that both the patients, while no difference was observed in the control SLL and BMI increased significantly from baseline [52]. group during the same period. In addition, the mean change Long-term treatment with clozapine and conventional in BW in the olanzapine group correlated with the change in APs has been studied in a longitudinal study by Hagg et al. leptin levels and in BMI [62]. (2001) [53]. They compared 41 schizophrenic patients in The changes observed in leptin levels with regard to treatment with clozapine over approximatley 2.8 years with treatment with olanzapine have been seen as significantly 62 taking conventional over the course of about 8.7 years. varied in comparison to treatment with other APs. For ex- The authors reported an increase of PLL significantly associ- ample, a longitudinal study by Atmaca and collegues (2003), ated with clozapine treatment in both men and women, but compared 45 patients with schizophrenia, treated by mon- an increase in leptin levels with conventional AP treatment otherapy with either quetiapine (n = 15), olanzapine (n = only in men. 15), or haloperidol (n = 15) [63]. The patients were eval- Furthermore, Monteleone et al. (2002) found that a sig- uated at baseline and after 6 weeks of medication. These nificant increase in circulating leptin may be a predictive authors found a marked increase in leptin levels for the factor for weight gain after clozapine treatment [54]. olanzapine versus the quetiapine group and for the olanzapine versus the haloperidol group. In addition, Kraus and 3. Olanzapine Treatment colleagues (1999) measured PLLs, weight, and BMI at baseline and weekly (over 4 weeks) in patients with schizophrenia Studies on olanzapine treatment have some mixed findings, who received clozapine (n = 11), olanzapine (n = 8), or but an increase in leptin levels has generally been observed. treatment with haloperidol (n = 13), and in another group Graham et al. (2003) evaluated SLL levels at baseline, and of patients receiving no pharmacological treatment (n = 12) again after 12 weeks, in nine first-episode psychosis patients. [64]. They observed an early increase (at the end of the They did not observe any significant variation in SLL levels. first week) of PLL associated with olanzapine and clozapine- However, they did find significant variation in BW [55]. On induced WG. In contrast leptin levels, weight, and BMI the other hand, Murashita et al. (2005) compared meta- remained stable in patients receiving haloperidol or no bolic parameters before and after 6-month treatment with pharmacological treatment. 4 Journal of Obesity

4. Treatment with Other Atypical and scrutinized for a 5-HT2C receptor association. Data from Typical Antipsychotics the combined leptin and 5-HT2C −759 C/T genotype effect reported by Templeman et al. (2005) indicates that this ge- Zhang et al. (2003) evaluated the effects of risperidone and netic variability can account for over 25% of the variance in chlorpromazine on leptin, insulin secretion, and fat deposi- weight gain. The same team also showed that the 5-HT2C tion, in 46 schizophrenic patients [65]. These authors have polymorphism, like the leptin polymorphism, influences found a significant elevation in PLL after 10 weeks of med- leptin secretion [72]. On the other hand the −759T allele ication, with significant interactions with gender. In their showed a protective effect and resulted in higher plasma study, Herran et al. (2001) evaluated the effects on SLL of leptin prior to treatment. These findings suggest a pharma- long-term typical and atypical antipsychotic treatment (at cogenetic influence of leptin gene polymorphisms in antip- least 6 months) of 59 schizophrenic outpatients compared sychotics-induced weight gain. with 59 healthy controls [66]. Differences in leptin levels were Perez-Iglesias et al. (2010) studied patients (n = 205) significant in patients treated with atypical antipsychotics who received either haloperidol, olanzapine, risperidone, zi- with gender effect (levels were higher in females). The prasidone, aripiprazole, or quetiapine treatment (all antipsy- authors have also noted that SLL in patients correlated chotic treatments were adjusted to the lowest effective significantly with WG and showed a trend for an association dosage), and who were genotyped for rs7799039 LEP and with BMI gains. Haupt and colleagues have pointed out no rs1137101 LEPR, but no significant association with BMI or evidence of modulation of plasma leptin concentration by with antipsychotic-induced weight gain was found [73]. As typical or atypical antipsychotics (olanzapine, risperidone) well, the LEPR gene has been investigated by Moons et al. in 72 schizophrenic patients [67]. (2010) who have done an association study to examine possible association between LEPR polymorphism rs8179183 and 5. Preclinical Research several body parameters in 261 schizophrenia or schizoaffective patients treated typical or atypical antipsychotics with The findings in human studies are inconsistent with results no significant results [74]. Fernandez et al. (2010) conducted from animal models and in vitro studies. In their experiments a study in clozapine-treated patients who entered a trial to treating mature adipose cells from human mammary tissue assess the effect of metformin, an antidiabetic medication with clozapine, Hauner et al. (2004) did not find any var- [75]. Clinical and preclinical findings suggest that metformin iations in insulin-stimulated glucose transport or leptin affects leptin synthesis and serum levels and enhances leptin production [68]. Consistently, Cooper et al., (2008) using and insulin sensitivity. In this study, they first evaluated an animal model, performed a 20-day fixed-dosed treatment the frequency of the metabolic syndrome and obesity, using 1, 2, or 4 mg/kg/day olanzapine in female rats, and did anthropometric and biochemical variables before and after not find any significant differences across any of the cohorts randomization, and subsequently compared differences in in serum leptin levels at endpoint despite differences in BW among genotypes (LEP −2548 G/A and Q223R LEPR). weight gain [69]. On the other hand, Sondhi et al. (2006) No association was observed between the leptin system pol- found increases in serum leptin levels after 28 days of ymorphisms and the anthropometric variables during treat- clozapine treatment in rats [70]. ment with metformin or placebo. The QQ LEPR genotype has displayed significantly lower triglyceride levels at baseline 6. Genetic and Pharmacogenetic Studies and showed some of the expected response to metformin. In constrast, the GG LEP genotype has showed a significant To our knowledge, rare mutations in leptin gene (LEP) and increase in glucose after treatment with metformin [75]. Leptin Receptor gene (LEPR) cause morbid obesity in studies Gregoor et al. (2009) performed a cross-sectional analysis using both human subjects or animal models. Several studies to determine whether the LEPR Q223R polymorphism and have shown modest effects of rs7799039 LEP (−2548 G/A the LEP promoter −2548 G/A polymorphism are associated functional polymorphism) and rs1137101 LEPR variants on with obesity and atypical AP treatment [76]. No significant BMI, and some research has associated the same polymor- association was found between LEP promoter −2548 G/A phisms with antipsychotic-induced weight gain, with some polymorphism and obesity, while in females, but not in inconsistent results. males, the LEPR 223QR and LEPR 223RR genotypes were The first of these studies by Yang et al. (2007) examined associated with lower risk of obesity. Srivastava et al. (2008) the −2548 G/A functional polymorphism in the leptin gene found that the rs4731426 SNP, a variant in the leptin gene, promoter [71]. This promoter region polymorphism is was moderately associated with median weight gain and reported to influence the secretion of leptin which is asso- significantly associated with extreme weight gain in North ciated with obesity. The authors found that homozygosity India subjects in olanzapine treatment [77]. Leptin has been for this polymorphism was significantly associated with investigated also by Ellingrod et al. (2007) in a phar- antipsychotic-induced weight gain. The same polymorphism macogenetic association reanalysis of a longitudinal open (−2548 G/A polymorphism) was investigated by Templeman label fixed-dose trial of olanzapine response and adverse et al. (2005) in a 9-month study of neuroleptic-na¨ıve patients effects [78].Theenrolledsubjectsweregenotypedforthe with schizophrenia from Spain (n = 73). The authors found −2548 G/A polymorphism of the leptin gene and the Q223R that the polymorphism was associated with weight gain polymorphism of the leptin gene receptor. Genotypes were over the 9-month period. The same population has been not individually associated with olanzapine-induced weight Journal of Obesity 5 gain. Changes in weight from baseline increased significantly [2] K. T. Mueser and S. R. McGurk, “Schizophrenia,” The Lancet, in patients carrying at least one G allele at both candidate loci vol. 363, no. 9426, pp. 2063–2072, 2004. and high olanzapine plasma levels. [3] P. Seeman, “Antipsychotic drugs, dopamine receptors, and schizophrenia,” Clinical Neuroscience Research, vol. 1, no. 1-2, pp. 53–60, 2001. 7. Conclusions [4]D.W.Haupt,“Differential metabolic effects of antipsychotic ff treatments,” European Neuropsychopharmacology, vol. 16, sup- A majority of research has focused on the e ects of clozapine plement 3, pp. S149–S155, 2006. and olanzapine on SLL. The positive modulation of this [5] G. P. Reynolds and S. L. Kirk, “Metabolic side effects of an- parameter by the studied atypical antipsychotics has been tipsychotic drug treatment—pharmacological mechanisms,” observed as early as a few hours after the beginning of the Pharmacology and Therapeutics, vol. 125, no. 1, pp. 169–179, treatment, peaking between 6 and 10 weeks after, and re- 2010. maining elevated for a period of up to several months. With [6]M.Han,X.F.Huang,T.M.duBois,andC.Deng,“Theeffects respect to atypical AP treatment, a number of studies found of antipsychotic drugs administration on 5-HT1A receptor elevated leptin levels following atypical AP medications, in expression in the limbic system of the rat brain,” Neuroscience, some cases independent of BMI increase. vol. 164, no. 4, pp. 1754–1763, 2009. It is relevant to note that while weight gain with olanza- [7] R. Coccurello and A. Moles, “Potential mechanisms of atyp- pine and clozapine therapy predominantly occurs over the ical antipsychotic-induced metabolic derangement: clues for first 6 months of treatment plateauing after 6 months to 1 understanding obesity and novel drug design,” Pharmacology year of treatment, leptin changes do not parallel with weight and Therapeutics, vol. 127, no. 3, pp. 210–251, 2010. [8] C. Deng, K. Weston-Green, and X. F. Huang, “The role of his- changes during extended AP treatment as pointed out by taminergic H1 and H3 receptors in food intake: a mechanism Bromel et al. [50]. This pattern was replicated by Monteleone for atypical antipsychotic-induced weight gain?” Progress in et al. (2002) [54], and Theisen et al. (2005) [52]inlater Neuro-Psychopharmacology and Biological Psychiatry, vol. 34, studies. In contrast to olanzapine and clozapine, fluctuations no. 1, pp. 1–4, 2010. in SLL have not been seen in risperidone treatment, which [9] H. Blasco-Fontecilla, E. Baca-Garcia, and J. de Leon, “Do is consistent with the conventional neuroleptic treatment atypical antipsychotic drugs reduce the risk of ischemic heart causing less fluctuation in BW. disease and mortality? Possible role of 5-HT2A receptor Weight gain as a result of AP treatment is associated blockade,” Schizophrenia Research, vol. 119, no. 1–3, pp. 160– with significant physical and psychological morbidity, and 163, 2010. patients undergoing AP treatment have significant and valid [10] S. Choi, B. DiSilvio, J. Unangst, and J. D. Fernstrom, “Effect of concerns about weight gain and weight management. Sub- chronic infusion of olanzapine and clozapine on food intake stantial progress has recently been made in understanding and body weight gain in male and female rats,” Life Sciences, weight gain, energy expenditure rate, and changes in appetite vol. 81, no. 12, pp. 1024–1030, 2007. associated with AP treatment, including the identification [11] A. Esen-Danaci, A. Sarandol, F. Taneli, F. Yurtsever, and N. Ozlen,¨ “Effects of second generation antipsychotics on of markers of obesity risk caused by fluctuation in energy leptin and ghrelin,” Progress in Neuro-Psychopharmacology and homeostasis, metabolic variation, and energy intake and Biological Psychiatry, vol. 32, no. 6, pp. 1434–1438, 2008. expenditure during antipsychotic treatment. Among these [12] J.HarveyandM.L.J.Ashford,“LeptinintheCNS:muchmore markers, leptin could play an important role and is associated than a satiety signal,” Neuropharmacology, vol. 44, no. 7, pp. with weight gain in atypical antipsychotic treatment as seen 845–854, 2003. in the present review. [13] F. C. J. Starrenburg and J. P. A. M. Bogers, “How can antip- Although it is true that these changes can be explained sychotics cause diabetes mellitus? Insights based on receptor- by adiposity, and produced and released by adipocytes in binding profiles, humoral factors and transporter proteins,” response to increased fat deposition to regulate body weight, European Psychiatry, vol. 24, no. 3, pp. 164–170, 2009. there is intriguing evidence to support the hypothesis that [14] F. Zhang, Y. Chen, M. Heiman et al., “Leptin: structure, func- antipsychotics could induce or exacerbate a leptin-resistance tion and biology,” in Vitamins & Hormones,L.Gerald,Ed.,pp. status. Resistance to leptin could generate metabolic condi- 345–372, Academic Press, 2005. tions resulting in weight gain. Further research is still needed [15] G. Sweeney, “Leptin signalling,” Cellular Signalling, vol. 14, no. to clarify the relationship between leptin and antipsychotics. 8, pp. 655–663, 2002. The elucidation of these potential mechanisms could be [16] H. Funahashi, T. Yada, R. Suzuki et al., “Distribution, function, and properties of leptin receptors in the brain,” in In- important for quantification of patient risk during treatment ternational Review of Cytology,W.J.Kwang,Ed.,pp.1–27, with antipsychotics and can spur development of novel Academic Press, 2003. forms of treatment. [17] D. O. Yarnell, D. S. Knight, K. Hamilton, O. Tulp, and P. Tso, “Localization of leptin receptor immunoreactivity in the lean References and obese Zucker rat brain,” Brain Research, vol. 785, no. 1, pp. 80–90, 1998. [1] C. Rummel-Kluge, K. Komossa, S. Schwarz et al., “Head- [18] C. Buchanan, V.Mahesh, P.Zamorano, and D. Brann, “Central to-head comparisons of metabolic side effects of second nervous system effects of leptin,” Trends in Endocrinology and generation antipsychotics in the treatment of schizophrenia: a Metabolism, vol. 9, no. 4, pp. 146–150, 1998. systematic review and meta-analysis,” Schizophrenia Research, [19] A. Buison, F. Ordiz, and K. L. C. Jen, “Long-term effects of vol. 123, no. 2-3, pp. 225–233, 2010. exogenousleptinonbodyweightandfatinpost-obesefemale 6 Journal of Obesity

rats,” Physiology and Behavior, vol. 74, no. 3, pp. 321–328, [37] L. Kostal,´ P. Vyboh,C.J.Savory,M.Jur´ ani,´ L. Kub´ıkova,´ and P. 2001. Blazˇ´ıcek,ˇ “Influence of food restriction on dopamine recep- [20] G. A. Wittert, H. Turnbull, and P. Hope, “Exogenously tor densities, catecholamine concentrations and dopamine administered leptin leads to weight loss and increased physical turnover in chicken brain,” Neuroscience, vol. 94, no. 1, pp. activity in the marsupial Sminthopsis crassicaudata,” Physiol- 323–328, 1999. ogy and Behavior, vol. 85, no. 5, pp. 613–620, 2005. [38] N. D. Barnard, E. P. Noble, T. Ritchie et al., “D2 dopamine [21] J. Lopez-Soriano,´ N. Carbo,´ F. J. Lopez-Soriano,´ and J. M. receptor Taq1A polymorphism, body weight, and dietary Argiles,´ “Short-term effects of leptin on lipid metabolism in intake in type 2 diabetes,” Nutrition, vol. 25, no. 1, pp. 58–65, the rat,” FEBS Letters, vol. 431, no. 3, pp. 371–374, 1998. 2009. [22] S. P. Kalra, “Central leptin insufficiency syndrome: an interac- [39]Y.-J.Fang,G.N.Thomas,Z.-L.Xu,J.Q.Fang,J.A.J.H. tive etiology for obesity, metabolic and neural diseases and for Critchley, and B. Tomlinson, “An affected pedigree member designing new therapeutic interventions,” Peptides, vol. 29, no. analysis of linkage between the dopamine D2 receptor gene 1, pp. 127–138, 2008. TaqI polymorphism and obesity and hypertension,” Interna- [23] T. L. Breen, I. M. Conwell, and S. L. Wardlaw, “Effects of tional Journal of Cardiology, vol. 102, no. 1, pp. 111–116, 2005. fasting, leptin, and insulin on AGRP and POMC peptide [40] Z. Jovanovic and G. S. H. Yeo, “Central leptin signalling: release in the hypothalamus,” Brain Research, vol. 1032, no. 1- beyond the arcuate nucleus,” Autonomic Neuroscience: Basic 2, pp. 141–148, 2005. and Clinical, vol. 156, no. 1-2, pp. 8–14, 2010. [24] L. K. Malendowicz, M. Rucinski, A. S. Belloni et al., “Leptin [41] D. M. Opland, G. M. Leinninger, and J. M. G. Myers, “Mod- and the regulation of the hypothalamic-pituitary-adrenal ulation of the mesolimbic dopamine system by leptin,” Brain axis,” in International Review of Cytology,W.J.Kwang,Ed.,pp. Research C, vol. 1350, pp. 65–70, 2010. 63–102, Academic Press, 2007. [42] L. Brunetti, B. Michelotto, G. Orlando, and M. Vacca, “Leptin [25] J. E. Johansen, C. Broberger, C. Lavebratt et al., “Hypothalamic inhibits norepinephrine and dopamine release from rat hypo- CART and serum leptin levels are reduced in the anorectic thalamic neuronal endings,” EuropeanJournalofPharmacol- (anx/anx) mouse,” Molecular Brain Research, vol. 84, no. 1-2, ogy, vol. 372, no. 3, pp. 237–240, 1999. pp. 97–105, 2000. [43] U. Krugel,¨ T. Schraft, H. Kittner, W. Kiess, and P. Illes, [26] G. Orlando, L. Brunetti, C. Di Nisio et al., “Effects of cocaine- “Basal and feeding-evoked dopamine release in the rat nucleus and amphetamine-regulated transcript peptide, leptin and accumbens is depressed by leptin,” European Journal of orexins on hypothalamic serotonin release,” European Journal Pharmacology, vol. 482, no. 1–3, pp. 185–187, 2003. of Pharmacology, vol. 430, no. 2-3, pp. 269–272, 2001. [44] T. Morimoto, Y. Yamamoto, J. I. Mobarakeh et al., “Involve- [27] T. Shiraishi, Y. Oomura, K. Sasaki, and M. J. Wayner, “Effects ment of the histaminergic system in leptin-induced suppres- of leptin and orexin-A on food intake and feeding related sion of food intake,” Physiology and Behavior, vol. 67, no. 5, hypothalamic neurons,” Physiology and Behavior, vol. 71, no. pp. 679–683, 1999. 3-4, pp. 251–261, 2000. [45] T. Ishiguro, M. Iwase, M. Kanamaru, M. Izumizaki, Y. Ohshi- [28] T. Shiraishi, K. Sasaki, A. Niijima, and Y. Oomura, “Leptin ma, and I. Homma, “Impaired ventilation and metabolism effects on feeding-related hypothalamic and peripheral neu- response to hypoxia in histamine H1 receptor-knockout ronal activities in normal and obese rats,” Nutrition, vol. 15, mice,” Respiratory Physiology and Neurobiology, vol. 154, no. no. 7-8, pp. 576–579, 1999. 3, pp. 331–341, 2006. [29] N. S. Narayanan, D. J. Guarnieri, and R. J. DiLeone, “Metabolic [46] S. Ogawa, K. Yanai, T. Watanabe et al., “Histamine responses hormones, dopamine circuits, and feeding,” Frontiers in of large neostriatal interneurons in histamine H1 and H2 Neuroendocrinology, vol. 31, no. 1, pp. 104–112, 2010. receptor knock-out mice,” Brain Research Bulletin, vol. 78, no. [30] M. S. Szczypka, K. Kwok, M. D. Brot et al., “Dopamine pro- 4-5, pp. 189–194, 2009. duction in the caudate putamen restores feeding in dopamine- [47] Y. Zhang and P. J. Scarpace, “The role of leptin in leptin deficient mice,” Neuron, vol. 30, no. 3, pp. 819–828, 2001. resistance and obesity,” Physiology and Behavior, vol. 88, no. [31] R. D. Palmiter, “Is dopamine a physiologically relevant medi- 3, pp. 249–256, 2006. ator of feeding behavior?” Trends in Neurosciences, vol. 30, no. [48] S. Peralta, J. M. Carrascosa, N. Gallardo, M. Ros, and C. 8, pp. 375–381, 2007. Arribas, “Ageing increases SOCS-3 expression in rat hypotha- [32]J.Hao,S.CabezadeVaca,Y.Pan,andK.D.Carr,“Effects of lamus: effects of food restriction,” Biochemical and Biophysical central leptin infusion on the reward-potentiating effectofd- Research Communications, vol. 296, no. 2, pp. 425–428, 2002. amphetamine,” Brain Research, vol. 1087, no. 1, pp. 123–133, [49] C. Bjørbæk, J. K. Elmquist, J. D. Frantz, S. E. Shoelson, and 2006. J. S. Flier, “Identification of SOCS-3 as a potential mediator [33] C. M. Cannon, L. Abdallah, L. H. Tecott, M. J. During, and R. of central leptin resistance,” Molecular Cell,vol.1,no.4,pp. D. Palmiter, “Dysregulation of striatal dopamine signaling by 619–625, 1998. amphetamine inhibits feeding by hungry mice,” Neuron, vol. [50] T. Bromel, W. F. Blum, A. Ziegler et al., “Serum leptin 44, no. 3, pp. 509–520, 2004. levels increase rapidly after initiation of clozapine therapy,” [34] A. Hamdi, J. Porter, and P. Chandan, “Decreased striatal D2 Molecular Psychiatry, vol. 3, no. 1, pp. 76–80, 1998. dopamine receptors in obese Zucker rats: changes during [51] B. B. Kivircik, K. Alptekin, S. Çalis¸kan et al., “Effect of aging,” Brain Research, vol. 589, no. 2, pp. 338–340, 1992. clozapine on serum leptin, insulin levels, and body weight [35] N. D. Volkow, G.-J. Wang, F. Telang et al., “Low dopamine and composition in patients with schizophrenia,” Progress in striatal D2 receptors are associated with prefrontal metabolism Neuro-Psychopharmacology and Biological Psychiatry, vol. 27, in obese subjects: possible contributing factors,” NeuroImage, no. 5, pp. 795–799, 2003. vol. 42, no. 4, pp. 1537–1543, 2008. [52] F. M. Theisen, S. Gebhardt, T. Bromel et al., “A prospective [36] K. D. Carr, “Chronic food restriction: enhancing effects study of serum ghrelin levels in patients treated with clozap- on drug reward and striatal cell signaling,” Physiology and ine,” Journal of Neural Transmission, vol. 112, no. 10, pp. 1411– Behavior, vol. 91, no. 5, pp. 459–472, 2007. 1416, 2005. Journal of Obesity 7

[53] S. Hagg, S. Soderberg, B. Ahren, T. Olsson, and T. Mjorndal, induce weight gain or metabolic abnormalities but enhances “Leptin concentrations are increased in subjects treated with adiposity: implications for animal models of antipsychotic- clozapine or conventional antipsychotics,” Journal of Clinical induced weight gain,” Progress in Neuro-Psychopharmacology Psychiatry, vol. 62, no. 11, pp. 843–848, 2001. and Biological Psychiatry, vol. 32, no. 2, pp. 428–436, 2008. [54] P. Monteleone, M. Fabrazzo, A. Tortorella, S. La Pia, and M. [70] S. Sondhi, J. M. Castellano, V. Z. Chong et al., “cDNA array Maj, “Pronounced, early increase in circulating leptin predicts reveals increased expression of glucose-dependent insulin- lower weight gain during clozapine treatment,” Journal of otropic polypeptide following chronic clozapine treatment: Clinical Psychopharmacology, vol. 22, no. 4, pp. 424–426, 2002. role in atypical antipsychotic drug-induced adverse metabolic [55] K. A. Graham, D. O. Perkins, J. B. Harp et al., “Investigation effects,” Pharmacogenomics Journal, vol. 6, no. 2, pp. 131–140, of olanzapine-associated weight gain in first episode subjects,” 2006. Schizophrenia Research, vol. 60, no. 1, supplement 1, pp. 355– [71] D. Yang, H. Chen, X. Wang et al., “Association of poly- 356, 2003. morphisms of lptin gene with body weight and body sizes [56] M. Murashita, I. Kusumi, T. Inoue et al., “Olanzapine in- indexes in Chinese indigenous cattle,” Journal of Genetics and creases plasma ghrelin level in patients with schizophrenia,” Genomics, vol. 34, no. 5, pp. 400–405, 2007. Psychoneuroendocrinology, vol. 30, no. 1, pp. 106–110, 2005. [72] L. A. Templeman, G. P. Reynolds, B. Arranz, and L. San, [57] H. Hosojima, T. Togo, T. Odawara et al., “Early effects of olan- “Polymorphisms of the 5-HT2C receptor and leptin genes zapine on serum levels of ghrelin, adiponectin and leptin in are associated with antipsychotic drug-induced weight gain in patients with schizophrenia,” Journal of Psychopharmacology, Caucasian subjects with a first-episode psychosis,” Pharmaco- vol. 20, no. 1, pp. 75–79, 2006. genetics and Genomics, vol. 15, no. 4, pp. 195–200, 2005. [58] C.-J. Wang, Z.-J. Zhang, J. Sun et al., “Serum free fatty acids [73] R. Perez-Iglesias, I. Mata, J. A. Amado et al., “Effect of and glucose metabolism, insulin resistance in schizophrenia FTO, SH2B1, LEP, and LEPR polymorphisms on weight gain with chronic antipsychotics,” Biological Psychiatry, vol. 60, no. associated with antipsychotic treatment,” Journal of Clinical 12, pp. 1309–1313, 2006. Psychopharmacology, vol. 30, no. 6, pp. 661–666, 2010. [59] R. Pena,˜ D. Marquina, A. Serrano et al., “Frequency of abnor- [74] T. Moons, S. Claes, G. J. Martens et al., “Clock genes and body mal correlation between leptin and the body mass index dur- composition in patients with schizophrenia under treatment ing first and second generation antipsychotic drug treatment,” with antipsychotic drugs,” Schizophrenia Research, vol. 125, Schizophrenia Research, vol. 106, no. 2-3, pp. 315–319, 2008. no. 2-3, pp. 187–193, 2010. [60] V. Popovic, M. Doknic, N. Maric et al., “Changes in neu- [75] E. Fernandez, E. Carrizo, V. Fernandez et al., “Polymorphisms roendocrine and metabolic hormones induced by atypical of the LEP- and LEPR genes, metabolic profile after prolonged antipsychotics in normal-weight patients with schizophrenia,” clozapine administration and response to the antidiabetic Neuroendocrinology, vol. 85, no. 4, pp. 249–256, 2007. metformin,” Schizophrenia Research, vol. 121, no. 1–3, pp. [61] U. Eder, B. Mangweth, C. Ebenbichler et al., “Association of 213–217, 2010. olanzapine-induced weight gain with an increase in body fat,” [76] J. G. Gregoor, J. van der Weide, H. Mulder et al., “Poly- American Journal of Psychiatry, vol. 158, no. 10, pp. 1719–1722, morphisms of the LEP- and LEPR gene and obesity in 2001. patients using antipsychotic medication,” Journal of Clinical [62] M. Atmaca, E. Tezcan, and B. Ustundag, “Plasma nitric oxide Psychopharmacology, vol. 29, no. 1, pp. 21–25, 2009. and leptin values in patients with olanzapine-induced weight [77] V. Srivastava, S. N. Deshpande, V. L. Nimgaonkar, B. Lerer, gain,” Journal of Psychiatric Research, vol. 41, no. 1-2, pp. 74– and B. K. Thelma, “Genetic correlates of olanzapine-induced 79, 2007. weight gain in schizophrenia subjects from north India: role [63] M. Atmaca, M. Kuloglu, E. Tezcan, O. Gecici, and B. Ustundag, of metabolic pathway genes,” Pharmacogenomics,vol.9,no.8, “Weight gain, serum leptin and triglyceride levels in patients pp. 1055–1068, 2008. with schizophrenia on antipsychotic treatment with quetiap- [78] V. L. Ellingrod, J. R. Bishop, J. Moline, Y. C. Lin, and D. ine, olanzapine and haloperidol,” Schizophrenia Research, vol. D. Miller, “Leptin and leptin receptor gene polymorphisms 60, no. 1, pp. 99–100, 2003. and increases in body mass index (BMI) from olanzapine [64] T. Kraus, M. Haack, A. Schuld et al., “Body weight and leptin treatment in persons with schizophrenia,” Psychopharmacol- plasma levels during treatment with antipsychotic drugs,” ogy Bulletin, vol. 40, no. 1, pp. 57–62, 2007. American Journal of Psychiatry, vol. 156, no. 2, pp. 312–314, 1999. [65] Z. Zhang, Z. J. Yao, W. Liu et al., “Assessment of fat distribution by MRI, and changes in leptin and insulin associated with drug treatment of first episode schizophrenic patients,” Schizophrenia Research, vol. 60, no. 1, supplement 1, pp. 372– 372, 2003. [66] A. Herran, M. T. Garcia-Unzueta, J. A. Amado, M. T. De La Maza, C. Alvarez, and J. L. Vazquez-Barquero,´ “Effects of long- term treatment with antipsychotics on serum leptin levels,” British Journal of Psychiatry, vol. 179, pp. 59–62, 2001. [67] D. W. Haupt, J. A. Schweiger, A. K. Melson et al., “Plasma leptin in treated patients with schizophrenia,” Schizophrenia Research, vol. 60, no. 1, supplement 1, pp. 356–356, 2003. [68] H. Hauner, “The new concept of adipose tissue function,” Physiology and Behavior, vol. 83, no. 4, pp. 653–658, 2004. [69]G.D.Cooper,J.A.Harrold,J.C.G.Halford,andA.J. Goudie, “Chronic clozapine treatment in female rats does not Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 163749, 5 pages doi:10.1155/2011/163749

Research Article High Levels of Cardiovascular Risk Factors among Pakistanis in Norway Compared to Pakistanis in Pakistan

Naeem Zahid,1, 2 Haakon E. Meyer,2, 3 Bernadette N. Kumar,2 Bjørgulf Claussen,2 and Akhtar Hussain2

1 Department of Gastrointestinal Surgery, Akershus University Hospital, Akershus, 1478 Lørenskog, Norway 2 Department of General Practice and Community Medicine, University of Oslo, 0316 Oslo, Norway 3 Norwegian Institute of Public Health, 0403 Oslo, Norway

Correspondence should be addressed to Naeem Zahid, [email protected]

Received 22 November 2010; Accepted 3 May 2011

Academic Editor: Gianluca Iacobellis

Copyright © 2011 Naeem Zahid et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Objectives. Previous studies have shown that the Norwegian-Pakistanis had considerably higher prevalence for diabetes and obesity compared to Norwegians. We studied the additional risk of obesity, dyslipidemia, and hypertension among Pakistanis in Norway compared to Pakistanis living in Pakistan. Method. 770 Norwegian-Pakistani adults (53.9% men and 46.1% women) born in Pakistan from two surveys conducted in Norway between 2000 and 2002 were compared with a sample of 1230 individuals (29.1% men and 70.9% women) that participated in a survey in Pakistan in 2006. Results. Both populations had similar height, but Norwegian-Pakistanis had considerably higher mean weight. Of the Norwegian-Pakistanis, 56% of the males and 40% of the females had a BMI above 25 kg/m2, as opposed to 30% and 56% in Pakistan, for males and females, respectively. Norwegian- Pakistanis had higher total cholesterol. Conclusion. Obesity and an unfavourable lipid proﬁle were widely prevalent in both populations; the highest level was recorded amongst those living in Norway. The increased risk for obesity and dyslipidemia may be ascribed to change of lifestyle after migration.

1. Introduction concluded that the lack of undesirable weight gain was the reason for the lower risk [13]. The population of the The population in Pakistan has a high risk of diabetes and Norwegian capital has been increasingly diversified during coronary heart disease [1], and common risk factors related the last decades, and there is now a large Norwegian- to the two conditions are present at early ages [2]. This Pakistanis population in Oslo [15]. The prevalence of obesity elevated risk amongst Pakistanis is demonstrated both in and diabetes in Oslo has proven to be very high especially Pakistan and abroad [3–5]. Migrants from the Indian sub- amongst those from Pakistan [8, 16]. It has been suggested continent living in Europe and America have higher rates that migration from developing countries to developed of cardiovascular risk factors compared to the locals [6–10]. countries leads to these changes [12]. No study has so far Few articles have compared the prevalence of risk factors compared Pakistanis living in Norway with Pakistanis living in immigrants living in western societies with those still in Pakistan. It is therefore necessary to find out whether there living in their homesteads in the subcontinent [11–14]. Two is an elevated prevalence of risk factors for cardiovascular of these were comparisons of Indians living in London disease in Pakistanis residinginOslocomparedtothose with people living in India. Both showed that British- living in the country of origin. In this paper, we compare Indians had unfavourable cardiovascular risk factor profile two Pakistani populations, one in Norway and the other compared to those living in India [11, 12]. One team studied in Pakistan, for cardiovascular risk factors. Most Pakistanis Indians in Australia and their relatives in India and found living in Norway are from an area called Kharian, in Punjab, that women living in Australia had a more desirable risk and therefore we chose this area for an epidemiological profile compared to those still living in India. The researchers study. 2 Journal of Obesity

2. Methods 5.0 mmol/L were considered high. HDL levels below 0.9 and 1.0 mmol for males and females, respectively, were regarded 2.1. Norway. Data were obtained from two population- as low. based, cross-sectional surveys conducted in Oslo, Norway, In Norway, blood pressure was measured with an auto- between 2000 and 2002 with similar protocol. The Oslo matic device (DINAMAP, Criticon, Tampa, Fla, USA), while Health Study (HUBRO) was a collaboration between the a standard sphygmomanometer was used in Pakistan. Sys- Norwegian Institute of Public Health, the University of Oslo tolic pressure above 140 mmHg and diastolic pressure above and the Oslo Municipality (2000-01). All Oslo residents born 90 mmHg were classified as high. in 1924, 1925, 1940, 1941, 1955, 1960, and 1970 were invited Respondents in both countries were asked about their to a health survey. Of these 18,770 (46%) attended [17]. smoking habits. They were classified as either smokers, non- The second survey, the Oslo Immigrant Health Study, was smokers or previous smokers. conducted by the Norwegian Institute of Public Health and the University of Oslo in 2002. In this survey individuals born in Turkey, Iran, Pakistan Sri Lanka, or Vietnam between 2.5. Statistics. Mean values are presented with one standard 1942 and 1971 were invited to participate, and the response deviation. Student’s t-test was used to calculate P values rate for Pakistani immigrants was 31.7% [16]. In the when comparing two means. present analysis, we have only included participants born in Age adjusted prevalence and means using direct stan- Pakistan. Of these, ten subjects were excluded due to invalid dardization with averaged weight as the standard population height/weight measurements. A total of 770 participants are presented in brackets or below the tables were applicable. were included from the Norwegian material; 415 men, and 355 women were included in Norway while in Pakistan 358 men and 872 women were included. 3. Results Four hundred and fifteen Pakistani men and 355 women 2.2. Pakistan. Data was collected during spring and summer wereincludedinNorway,whileinPakistan358menand872 of 2006. Subjects were enrolled from 44 villages from this women were included. The mean age in Norway was 44.2 area, about 150 km from the capital Islamabad. This is years for males and 42.4 for females (Table 1). In Pakistan, primarily an agricultural community but has developed the mean age was 46.4 and 44.2 for males and females, rapidly in the recent decades due to migration to the west. All respectively. participants were 20 years or older and met after fasting for Both genders had similar height in Norway and Pakistan, 8–10 hours before the examinations. Verbal information was but their weight on the other hand was not similar. Pakistanis secured from all participants. Only subjects aged between living in Norway had significantly higher mean weight and 30 and 61 were included in the current analysis in order to BMI (Table 1). Being overweight and obese, in terms of match the Oslo sampling procedure. The methods for the having a BMI between 25 and 30 and above 30, was more survey in Pakistan are described in detail elsewhere [18]. commonly seen among Pakistanis in Norway (Table 2). More From the Pakistani material, 1230 subjects were included, than one-fifth of the Pakistani males in Norway were obese, 358 men and 872 women. while only 7% of the males in Pakistan had a BMI above 30. It was more common for males in Norway to have 2.3. Procedures for Anthropometry Measurements. In Norway, high waist girth and WHR compared to males in Pakistan. weight was measured on a beam scale. This professional scale High WHR was more frequently observed in Pakistani comes complete with an attached height rod, where both female subjects. Pakistani males in Norway had higher waist weight and height can be measured simultaneously. Waist circumference, as well as hip girth and waist-hip ratio and hip girths were measured with a steel measuring tape (WHR), compared to males in Pakistan. [19]. In Pakistan, the height was measured against a wall Women in Pakistan had higher systolic and diastolic where a measuring tape was attached. Weight was measured pressure compared to females in Norway (Table 3). Men in with an electronic scale. Waist and hip girths were measured Norway had higher systolic pressure compared to men in with a nonelastic plastic tape measure. Pakistan. Hypertension appeared to be more common in The participants in both countries were weighed wearing Pakistan, especially among females. light clothing. 2 Both males and females in Norway had higher total cho- NormalweightwasdefinedasBMIupto24.9kg/m, lesterol compared to their counterparts in Pakistan. Women and overweight was BMI between 25 and 29.9 kg/m2.Obesity ≥ 2 residing in Norway had higher HDL than women in Pakis- wasdefinedasaBMI 30 kg/m [20]. Waist above 80 cm tan. for women and 90 cm for men was labelled high. Waist-hip ff Systolic and diastolic blood pressure increased with ratio cuto s were set at 0.8 and 0.9 for females and males, increasing BMI for both genders in both Norway and Pak- respectively [21]. istan (Table 4). Not surprisingly, waist and WHR increased with BMI so did total cholesterol. With increasing BMI, HDL 2.4. Lipids, Blood Pressure, and Smoking Habits. Total choles- decreased in the Norwegian-Pakistanis but increased in those terol and HDL were analysed in both countries using living in Pakistan. The highest standardized beta coefficient enzymatic methods. Total cholesterol levels higher than was seen for waist girth, with a standardized beta value of Journal of Obesity 3

Table 1: Age and physical characteristics of the Pakistanis in Oslo and in Pakistan. Numbers are mean values with one SD in brackets. P values for diﬀerence between Pakistanis and Norwegian-Pakistanis. Age adjusted values below.

Males Females Norway Pakistan P value Norway Pakistan P value 44.2 (9.3) 46.4 (9.5) <.01 42.4 (8.6) 44.2 (8.6) Age <.01 45.2 45.3 42.9 43.9 170.5 (5.9) 170.6 (7.5) .84 156.4 (5.6) 156.6 (6.0) Height (cm) .69 170.5 170.7 156.5 156.6 79.8 (11.6) 67.1 (14.1) <.01 71.6 (12.9) 64.1 (14.2) Weight (kg) <.01 79.7 67.0 71.6 64.0 27.5 (3.5) 23.0 (4.5) <.01 29.4 (5.0) 26.1 (5.5) BMI <.01 27.5 23.0 29.4 26.1 94.4 (10.2) 86.9 (12.8) <.01 89.0 (12.0) 90.5 (12.9) Waist (cm) .05 94.7 86.6 89.2 90.5 100.4 (6.8) 94.3 (9.9) <.01 104.6 (9.8) 105.5 (13.2) Hip (cm) .21 100.5 94.2 104.6 105.5 0.94 (0.07) 0.92 (0.07) <.01 0.85 (0.08) 0.86 (0.07) Waist/Hip .08 0.94 0.92 0.85 0.86

Table 2: Prevalence of obesity (per 100) among Pakistanis and 4. Discussion Pakistanis residing in Oslo. Age adjusted values in brackets. We demonstrated high prevalence of obesity and cardiovas- Males Norway Pakistan P value cular risk factors in both populations, this is in line with BMI > 25 56 (56) 23 (23) <.01 earlier studies [11, 12]. Obesity, overweight, and having high levels of lipids were more common in Norway, while high BMI > 30 22 (22) 7 (7) <.01 blood pressure was seen more frequently in Pakistan. We High waist 63 (64) 36 (35) .01 believe that the two populations are comparable because High WHR 69 (70) 60 (59) .01 the majority of the Pakistanis living in Norway actually Females migrated from this particular area in Pakistan, an area called Kharian in the district of Gujrat. Therefore, it is reasonable to > < BMI 25 40 (40) 33 (33) .01 postulate that the populations are genetically and culturally BMI > 30 40 (41) 23 (23) <.01 comparable. The differences we observe between the two ff High waist 73 (73) 76 (76) .28 populations could therefore be due to the e ects of migration and changes in lifestyle from a low-income to a high-income High WHR 70 (71) 82 (82) <.01 country. High waist: above 90 for males and 80 for females. The difference in weight and BMI are of such a magni- High WHR: above 0.9 for males and 0.8 for females. P values for difference between Pakistanis and Norwegian-Pakistanis. tude that they cannot be explained by possible measuring error. This is particularly true for the males; the difference is more than 10 kg, whilst it is almost 8 kg in women. This difference is reflected in the BMI; increased BMI in Oslo approximately 0.8 for all groups. The standardized beta among the Pakistani population may only be explained by value for WHR was 0.48 and 0.52 for males in Norway and added weight in this population, since height remains the Pakistan, respectively. Females however had a considerably same in both populations. lower standardized beta value for WHR. Females in Pakistan The difference between the populations in waist and hip had the lowest. girth are also evident among the males. The men living in Smoking habits were unfortunately not recorded in all Norway have almost a waist of more than seven cm greater of those living in Pakistan. However, the results obtained than the males in Pakistan; similarly, the hip is also more than in Pakistan (n = 401) were similar to those in Norway. In six cm larger in the Norwegian-Pakistanis. The differences Pakistan, 40.5% of the males and 2.8% females were current between the females on the other hand are small. Some smokers; in Norway, 34% of the males and 3.8% of the women in Pakistan might have been reluctant to remove their females were smokers. None of the females in Pakistan said clothes for the measurement of the waist and hip girth even they were previous smokers, while 2.2% of the females in though same gender investigators did all the measurements. Norway said so. Among the males in Pakistan, 7.4% said they Women living in rural Pakistan might also have had a higher were previous smokers, of the Norwegian-Pakistanis, 18.5% number of pregnancies, which could have resulted in a higher were previous smokers. waist, hip, and waist-hip ratio. On the other hand, the WHR 4 Journal of Obesity

Table 3: Diﬀerences in clinical features among Pakistanis in Oslo and in Pakistan. Numbers are mean values with one SD in brackets. P values for diﬀerence between Pakistanis and Norwegian-Pakistanis. Age adjusted values below.

Males Females Norway Pakistan P value Norway Pakistan P value 128.8 (14.2) 123.7 (16.8) <.01 120.9 (18.2) 125.5 (17.1) SysBP <.01 129.3 123.4 121.3 125.5 78.0 (10.1) 79.4 (11.4) .07 71.3 (11.0) 80.8 (10.5) DiaBP <.01 78.3 79.3 71.5 80.8 18 20 .59 14 24 %Sys≥ 140 <.01 19 19 14 24 13 28 <.01 5 31 %Dia≥ 90 <.01 13 28 5 31 5.4 (1.0) 4.5 (1.0) <.01 5.1 (1.0) 4.7 (1.0) TotChol (mmol/L) <.01 5.5 4.5 5.1 4.7 1.1 (0.2) 1.0 (0.5) .38 1.2 (0.3) 1.0 (0.4) HDL (mmol/L) <.01 1.1 1.0 1.2 1.0

Table 4: Association between BMI and other clinical features. is low in Pakistan, and few patients have had their blood pressure measured [24, 25]. It is also important to note that Males Females the blood pressure was measured differently in the two pop- βPvalue βPvalue βPvalue βPvalue ulations. Dinamap was used in Norway, in Pakistan we used < < < < SysBP 0.19 .01 0.35 .01 0.15 .01 0.25 .01 a standard sphygmomanometer. The different measuring DiaBP 0.11 .02 0.35 <.01 0.07 .18 0.27 <.01 methods might have yielded different results. Cautiousness Waist 0.78 <.01 0.84 <.01 0.78 <.01 0.80 <.01 should therefore be applied when comparing blood pressure WHR 0.48 <.01 0.52 <.01 0.27 <.01 0.12 <.01 in the two populations and interpreting these results. High TotChol 0.01 .80 0.20 <.01 0.12 .01 0.10 .01 levels of hypertension have been reported earlier amongst HDL −0.16 <.01 0.11 .06 −0.24 .00 0.09 .02 Pakistanis [3, 4], although, some large studies have reported Standardized coefficients, only adjusted for age. considerably lower prevalence of hypertension [26, 27]. Comparison between most of these studies is hampered by diverging measuring techniques. did not increase as steeply with increasing BMI among the Smoking is common amongst Pakistani males both in females in Pakistan as it did amongst the Pakistani females in Norway and in Pakistan; women, however, are fortunately Norway. One study has shown that expatriate Indian women spared. This pattern of smoking has been demonstrated in in Australia did in fact have a better risk profile than their several studies that have looked at smoking habits amongst counterparts still living in India [13]. Pakistanis in Pakistan and abroad [2, 4, 5]. There are large differences in the levels of total cholesterol especially for the males. The difference between females in the two populations was also considerable. However, this is 5. Conclusion not surprising since obesity was highly increased among the Our data demonstrate differences in cardiovascular risk Pakistani population residing in Oslo. factors in these two populations, possibly as a consequence Several studies have showed that people from South Asia of migration and related changes in lifestyle. More research living in western societies have a relatively low level of physi- is needed on the modification of lifestyle and food habits cal activity [22, 23]. This might be the cause of the high level following migration. Nevertheless, Pakistanis living in Nor- of adiposity among the Pakistanis living in Norway. In addi- way have proven to have higher levels of diabetogenic and tion, higher consumption of unhealthy fatty foods, which cardiovascular risk factors and therefore should be treated as is available in Oslo due to privileged income and sedentary “high risk group” for both prevention and treatment. lifestyle, may have contributed to the observed conditions. There are some interesting differences in blood pressure between these two populations. The Pakistanis living in References Norway have lower pressures, except for the systolic pressure [1] M. Gupta, N. Singh, and S. Verma, “South Asians and cardio- in males. We do not have data on use of antihypertensive vascular risk: what clinicians should know,” Circulation,vol. drugs in Pakistan, but we find it reasonable to believe that 113, no. 25, pp. e924–e929, 2006. such medication might be less common than in Norway. Due [2]P.Joshi,S.Islam,P.Paisetal.,“Riskfactorsforearlymyocar- to low access to doctors, undiagnosed hypertension might be dial infarction in South Asians compared with individuals in more common in rural Pakistan compared to Oslo. Earlier other countries,” Journal of the American Medical Association, studies have showed that the awareness about hypertension vol. 297, no. 3, pp. 286–294, 2007. Journal of Obesity 5

[3] T. H. Jafar, F. H. Jafary, S. Jessani, and N. Chaturvedi, “Heart [18] N. Zahid, B. Claussen, and A. Hussain, “Diabetes and im- disease epidemic in Pakistan: women and men at equal risk,” paired glucose tolerance in a rural area in Pakistan and asso- American Heart Journal, vol. 150, no. 2, pp. 221–226, 2005. ciated risk factors,” Diabetes and Metabolic Syndrome: Clinical [4] T. H. Jafar, Z. Qadri, and N. Chaturvedi, “Coronary artery dis- Research and Reviews, vol. 2, no. 2, pp. 125–130, 2008. ease epidemic in Pakistan: more electrocardiographic evidence [19] A. J. Søgaard and R. Selmer, “The Oslo Health Study,” Septem- of ischaemia in women than in men,” Heart,vol.94,no.4,pp. ber 2006. 408–413, 2008. [20] WHO, “Physical Status: the use and interpretation of anthro- [5]F.P.Cappuccio,D.G.Cook,R.W.Atkinson,andP.Strazzullo, pometry,” Report of a WHO Expert Committee, WHO, “Prevalence, detection, and management of cardiovascular Geneva, Switzerland, 1995. risk factors in different ethnic groups in South London,” Heart, [21] K. G. M. M. Alberti, P. Zimmet, and J. Shaw, “Metabolic vol. 78, no. 6, pp. 555–563, 1997. syndrome—a new world-wide definition. A consensus state- [6]S.S.Anand,S.Yusuf,V.Vuksanetal.,“Differences in risk ment from the International Diabetes Federation,” Diabetic factors, atherosclerosis, and cardiovascular disease between Medicine, vol. 23, no. 5, pp. 469–480, 2006. ethnic groups in Canada: the Study of Health Assessment and [22] A. K. Jenum, S. A. Anderssen, K. I. Birkeland et al., “Promoting Risk in Ethnic groups (SHARE),” Lancet, vol. 356, no. 9226, physical activity in a low-income multiethnic district: effects pp. 279–284, 2000. of a community intervention study to reduce risk factors [7] P.M. McKeigue and M. G. Marmot, “Mortality from coronary for type 2 diabetes and cardiovascular disease: a community heart disease in Asian communities in London,” British intervention reducing inactivity,” Diabetes Care, vol. 29, no. 7, Medical Journal, vol. 297, no. 6653, p. 903, 1988. pp. 1605–1612, 2006. [8]A.K.Jenum,I.Holme,S.Graff-Iversen, and K. I. Birkeland, [23] R. Bhopal, C. Fischbacher, E. Vartiainen, N. Unwin, M. White, “Ethnicity and sex are strong determinants of diabetes in and G. Alberti, “Predicted and observed cardiovascular disease an urban Western society: implications for prevention,” in South Asians: application of FINRISK, Framingham and Diabetologia, vol. 48, no. 3, pp. 435–439, 2005. SCORE models to Newcastle Heart Project data,” Journal of [9]E.Flowers,C.Molina,A.Mathuretal.,“Prevalenceof Public Health, vol. 27, no. 1, pp. 93–100, 2005. metabolic syndrome in South Asians residing in the United [24]T.Ashfaq,Q.Anjum,H.Siddiqui,S.Shaikh,andE.A. States,” Metabolic Syndrome and Related Disorders,vol.8,no. Vohra, “Awareness of Hypertension among patients attending 5, pp. 417–423, 2010. Primary Health Care Centre and Outpatient Department of [10] S. N. Rajpathak, L. S. Gupta, E. N. Waddell et al., “Elevated risk tertiary care hospital of Karachi,” Journal of the Pakistan of type 2 diabetes and metabolic syndrome among Asians and Medical Association, vol. 57, no. 8, pp. 396–399, 2007. south Asians: results from the 2004 New York City HANES,” [25] K. Ahmad and T. H. Jafar, “Prevalence and determinants of Ethnicity & Disease, vol. 20, no. 3, pp. 225–230, 2010. blood pressure screening in Pakistan,” Journal of Hypertension, [11]J.V.Patel,A.Vyas,J.K.Cruickshanketal.,“Impactof vol. 23, no. 11, pp. 1979–1984, 2005. migration on coronary heart disease risk factors: comparison [26] A. S. Shera, F. Jawad, and A. Maqsood, “Prevalence of diabetes of Gujaratis in Britain and their contemporaries in villages of in Pakistan,” Diabetes Research and Clinical Practice, vol. 76, origin in India,” Atherosclerosis, vol. 185, no. 2, pp. 297–306, no. 2, pp. 219–222, 2007. 2006. [27] T. H. Jafar, A. S. Levey, F. H. Jafary et al., “Ethnic subgroup dif- [12] D. Bhatnagar, I. S. Anand, P. N. Durrington et al., “Coronary ferences in hypertension in Pakistan,” Journal of Hypertension, risk factors in people from the Indian subcontinent living in vol. 21, no. 5, pp. 905–912, 2003. West London and their siblings in India,” Lancet, vol. 345, no. 8947, pp. 405–409, 1995. [13] D. Mahajan and M. A. Bermingham, “Risk factors for coronary heart disease in two similar Indian population groups, one residing in India, and the other in Sydney, Australia,” European Journal of Clinical Nutrition, vol. 58, no. 5, pp. 751– 760, 2004. [14]S.U.B.Tennakoon,B.N.Kumar,D.B.Nugegoda,andH.E. Meyer, “Comparison of cardiovascular risk factors between Sri Lankans living in Kandy and Oslo,” BMC Public Health,vol. 10, no. 1, article 654, 2010. [15] K. Glenday, B. N. Kumar, A. Tverdal, and H. E. Meyer, “Cardiovascular disease risk factors among five major ethnic groups in Oslo, Norway: the Oslo Immigrant Health Study,” European Journal of Cardiovascular Prevention and Rehabilita- tion, vol. 13, no. 3, pp. 348–355, 2006. [16] B. N. Kumar, H. E. Meyer, M. Wandel, I. Dalen, and G. Holmboe-Ottesen, “Ethnic differences in obesity among immigrants from developing countries, in Oslo, Norway,” Inter- national Journal of Obesity, vol. 30, no. 4, pp. 684–690, 2006. [17] A. J. Søgaard, R. Selmer, E. Bjertness, and D. Thelle, “The Oslo Health Study: the impact of self-selection in a large, population-based survey,” International Journal for Equity in Health, vol. 3, no. 1, Article ID 3, 2004. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 180729, 6 pages doi:10.1155/2011/180729

Research Article Mendelian Randomisation Study of Childhood BMI and Early Menarche

Hannah S. Mumby,1 Cathy E. Elks,1 Shengxu Li,1 Stephen J. Sharp,1 Kay-Tee Khaw,2 Robert N. Luben,2 Nicholas J. Wareham,1 Ruth J. F. Loos,1 and Ken K. Ong1

1 MRC Epidemiology Unit, Institute of Metabolic Science, Addenbrooke’s Hospital, P.O. Box 285, Cambridge CB2 0QQ, UK 2 Department of Public Health and Primary Care, Institute of Public Health, University of Cambridge, Cambridge CB2 0SR, UK

Correspondence should be addressed to Ken K. Ong, [email protected]

Received 1 December 2010; Revised 4 April 2011; Accepted 7 April 2011

Academic Editor: Angelo Pietrobelli

Copyright © 2011 Hannah S. Mumby et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

To infer the causal association between childhood BMI and age at menarche, we performed a mendelian randomisation analysis using twelve established “BMI-increasing” genetic variants as an instrumental variable (IV) for higher BMI. In 8,156 women of European descent from the EPIC-Norfolk cohort, height was measured at age 39–77 years; age at menarche was self-recalled, as was body weight at age 20 years, and BMI at 20 was calculated as a proxy for childhood BMI. DNA was genotyped for twelve BMI- associated common variants (in/near FTO, MC4R, TMEM18, GNPDA2, KCTD15, NEGR1, BDNF, ETV5, MTCH2, SEC16B, FAIM2 and SH2B1), and for each individual a “BMI-increasing-allele-score” was calculated by summing the number of BMI-increasing alleles across all 12 loci. Using this BMI-increasing-allele-score as an instrumental variable for BMI, each 1 kg/m2 increase in childhood BMI was predicted to result in a 6.5% (95% CI: 4.6–8.5%) higher absolute risk of early menarche (before age 12 years). While mendelian randomisation analysis is dependent on a number of assumptions, our ﬁndings support a causal eﬀect of BMI on early menarche and suggests that increasing prevalence of childhood obesity will lead to similar trends in the prevalence of early menarche.

1. Introduction have slowed or even stopped since around 1950 [6]while the prevalence of childhood overweight and obesity has Early age at menarche, the onset of menstrual periods in increased since the 1980s [7]. It is possible, therefore, that girls, is associated with increased risks of adverse health the apparent association between higher BMI and earlier age outcomes such as breast, ovarian, and endometrial cancer, at menarche might be confounded by other factors such as hypertension, type 2 diabetes, and cardiovascular disease [1, diet or exposure to endocrine disruptors [8]. The association 2]. Earlier age at menarche is also associated with increased could also be explained by reverse causality as the progression risk for a number of psychosocial outcomes in adolescence of puberty in girls is accompanied by rapid gains in body including depression, eating disorders, substance abuse, weight and body fat [9]. sexual risk-taking and teenage pregnancy [3]. Mendelian randomisation, using robust genetic vari- It has been suggested that childhood BMI has a causal ants as “instrumental variables” [10], has been suggested effect on the risk for early menarche and there are a number as an approach to avoid the problems of confounding, of strongly plausible biological mechanisms [4, 5]. However, residual confounding and specificity that are experienced discordant secular trends in obesity and age at menarche have by traditional epidemiological studies [11]. For example, raised doubts about the causal nature of these associations. Mendelian randomisation studies have demonstrated the In developed countries, a long-term trend towards earlier causal effects of low-density lipoprotein (LDL) cholesterol menarche has been observed from the late 1800s to the on risk of myocardial infarction [12], apparent protective mid 1900s [6]. In many countries these trends appear to effects of high-density lipoprotein (HDL) cholesterol on 2 Journal of Obesity coronary heart disease [13], and lack of a causal effect of C- data, 10,136 had data on recalled age at menarche within the reactive protein (CRP) on ischemic cerebrovascular disease physiological range of 8–18 years, and of these 8,387 women and carotid intimamedia thickness [14]. Similar approaches, had data on their recalled body weight at age 20 years, and using BMI-increasing variants at FTO and MC4R have 8,156 also had data on measured height in order to calculate reported apparent causal effects of BMI on hypertension [15] BMI at age 20. All analyses were therefore based on 8,156 and markers of atherosclerosis [16]. In recent years, large women, who were slightly younger, taller and lighter at their scale genome wide association (GWA) studies have identified baseline visit compared to excluded women, but had very several common genetic variants that are robustly associated similar recalled weight and (calculated) BMI at age 20 and with increased BMI [17–20]. Furthermore, two large studies recalled age at menarche (Table 1). have reported that some of these genetic variants for higher BMI are also associated with earlier age at menarche [21, 22]. We therefore used a mendelian randomisation approach to 2.2. Statistical Analyses assess the likely causal nature of the observed association 2.2.1. Mendelian Randomisation. The outcome variable between higher BMI and risk of early menarche [11]. “early menarche” was defined as menarche before age 12 years, as previously described [25, 26]. As the association of interest was between prepubertal BMI and age at menarche, 2. Materials and Methods BMI at age 20 years was used as a proxy for prepubertal BMI. The directly observed increase in risk of early menarche per 2.1. The EPIC-Norfolk Study. The European Prospective 1kg/m2 change in BMI at age 20 was assessed by logistic Investigation into Cancer and Nutrition-(EPIC-) Norfolk regression (to estimate an odds ratio), and by binomial study is a large, predominantly ethnically homogenous, regression (to estimate an absolute increase in risk), with white European population-based cohort study, which is part adjustment for age at baseline. of a multicentre international study designed to investigate To represent an instrumental variable (IV) for higher the relationship between diet and chronic disease. The design BMI, a “BMI-increasing-allele-score” was created in EPIC- of the EPIC-Norfolk study has been described in detail Norfolk by summing the number of BMI-increasing alleles previously [23]. The EPIC-Norfolk study was approved by across all 12 loci in each person. The association between the the Norwich local research ethics committee and informed BMI-increasing-allele-score and BMI at age 20 was analysed consent was given by all participants. by linear regression, with adjustment for age at baseline. The Age at menarche in completed whole years and weight association between the BMI-increasing-allele-score and risk at age 20 years were ascertained by recall in the baseline of early menarche was analysed by logistic regression. questionnaire which women completed at age 39–77 years The IV-predicted risk for early menarche per 1 kg/m2 old. Weight at age 20 was recalled and adult height measured change in BMI and was calculated using the ivprobit at baseline by trained nurses. The measures were used to command in STATA with a maximum likelihood estimator calculate BMI at age 20 (recalled weight in kg divided by in order to calculate the predicted absolute risk probabilities. height squared in metres). This was used as a proxy for All analyses were conducted using STATA version 10.1 childhood BMI. Genotype information was available for StataCorp., College Station, TX). 12 variants in the first 12 loci that were identified in the first three waves of GWA studies for BMI [17–20]; these variants were rs91121980 (in/near to gene FTO); rs17782313 3. Results (MC4R), rs6548238 (TMEM18), rs10938397 (GNPDA2), rs368794 (KCTD15), rs32568958 (NEGR1), rs10838738 3.1. The BMI-Increasing-Allele-Score. Associations between (MTCH2), rs925946 (BDNF), rs7498665 (SH2B1), the individual variants and BMI at age 20 years in EPIC- rs10913469 (SEC16B), rs10938397 (FAIM2/BCDIN3D)and Norfolk women are shown in Table 2.All12variantsshowed rs7647305 (ETV5). Genotyping was performed by custom directionally consistent associations with BMI at age 20 TaqMan SNP Genotyping Assays (Applied Biosystems, as expected from the original reports [17–20]. The BMI- Warrington, UK) or (markers rs10938397 and rs10838738) increasing-allele-score ranged from 3 to 20 alleles and was Sequenom iPLEX Gold standard chemistry (Sequenom, San normally distributed (Figure 1). Because few women had a n = n = Diego, CA) as previously described [24]. Call rates were score below 6 ( 51) or above 17 ( 25) in Figure 1 >95% and each locus genotyped was under HW equilibrium these scores were collapsed into the categories: “less than or given α = 0.05. equal to 6” and “greater than or equal to 17” and imputed In the EPIC-Norfolk study, 10,957 women had DNA SNP counts were rounded to the nearest whole number for the figure. On average, each additional BMI-increasing allele available for genotyping. Of these, only 6,709 women had 2 complete genotype data on all 12 SNPs, however a further was associated with 0.12 kg/m higher BMI at age 20 (95% CI: P = . × −19 3,972 women had genotype data on at least 9 SNPs and for 0.10–0.15, 6 8 10 ), but showed no association with P = . these women we imputed genotype data on their missing adultheight( 3). (upto3)SNPsusingthemeannumberofBMI-increasing alleles at each SNP as the individual values (we excluded 276 3.2. Mendelian Randomisation. 1,766 (21.7%) EPIC-Norfolk women who lacked genotype data on more than 3 SNPs). women reported that their menarche occurred before age 12 Of the 10,681 women with complete or imputed genotype years (early menarche). Each 1 kg/m2 higher BMI at age 20 Journal of Obesity 3

Table 1: Characteristics of EPIC-Norfolk women included in the current study.

Included women Excluded women n Mean SD n Mean SD ∗P value Age at baseline visit (years) 8,156 58.2 9.2 2,801 59.8 9.3 <.001 Height at baseline visit (cm) 8,156 161.1 6.2 2,503 160.5 6.2 <.001 Weight at baseline visit (kg) 8,150 67.6 11.5 2,510 68.3 12.1 .004 Weight at age 20 (kg) 8,156 56.8 8.0 804 56.5 7.8 .2 BMI at age 20 (kg/m2) 8,156 21.9 2.8 559 21.7 2.6 .2 Age at menarche (years) 8,156 13.0 1.6 2,241 13.0 1.6 .04 Inclusion criteria were complete genotype data on at least 9 SNPs, available height measurement and recalled information on age at menarche between 8 to 18 years, and body weight at age 20 years. ∗P values for unpaired t-test.

Table 2: Associations between individual BMI-increasing variants and BMI at age 20 years in 8,156 EPIC-Norfolk women.

Nearby gene SNP Chromosome Position B∗ (kg/m2/allele) Lower CI Upper CI SEC16B rs10913469 1 176180142 0.28 0.17 0.38 TMEM18 rs6548238 2 624905 0.21 0.10 0.32 FTO rs1121980 16 52366748 0.20 0.11 0.29 FAIM2 rs7132908 12 48549415 0.15 0.06 0.23 BDNF rs925946 11 27623778 0.13 0.03 0.22 MC4R rs17782313 18 56002077 0.13 0.03 0.23 GNPDA2 rs10938397 4 45023455 0.10 0.01 0.19 SH2B1 rs7498665 16 28790742 0.09 0.00 0.18 NEGR1 rs2568958 1 72477137 0.08 −0.01 0.16 ETV5 rs7647305 3 187316992 0.08 −0.03 0.18 MTCH2 rs10838738 11 47619625 0.08 −0.02 0.17 KCTD15 rs368794 19 39012292 0.06 −0.03 0.15 ∗ B: regression coeﬃcient from additive genetic models for the previously reported BMI-increasing allele.

The BMI-increasing-allele-score was positively associ- 13.4 ated with the relative risk of early menarche (OR = 1.06 per 1200 allele, 95% CI =1.03–1.08, P = 5.8 × 10−6). When using the 13.2 1000 BMI-increasing-allele-score as an instrumental variable for BMI, each 1 kg/m2 increase in BMI at age 20 was predicted to 800 13 result in a 6.5% (95% CI: 4.6–8.5%) increase in the absolute 600 risk of early menarche. 12.8 We performed a sensitivity analysis by calculating a

Number of women 400 “weighted BMI-increasing-allele-score”, where the contri- 12.6 200 Mean age at menarche (years) bution of each genotype to the score was weighted by 0 12.4 its individual association with BMI at age 20. Using this 2 ≤6 7 8 9 10 11 12 13 14 15 16 ≥17 weighted BMI-increasing-allele-score, each 1 kg/m increase in BMI at age 20 was predicted to result in a 6.7% (95% CI: Number of BMI-increasing alleles 5.0–8.4%) higher absolute risk of early menarche. Figure 1: Histogram showing the distribution of the BMI- increasing-allele-score in EPIC-Norfolk women (n = 8, 156). Within each allele score category, the mean and 95% CI for age at 4. Discussion menarche are shown by circles and error bars. The trend line shows The results of this mendelian randomisation analysis infers a the inverse linear trend between mean age at menarche and allele causal eﬀect of higher BMI on increased risk of early menar- score category. che (at age <12 years), and that the observed association is unlikely to be explained by positive confounding or by reverse causality. It is not suggested that such genetic instrumental variable was directly associated with an 11% (95% CI: 9–14%) higher approaches can replace randomised controlled trials, but relative risk of early menarche, or in terms of absolute risk rather that they support the causal inference from obser- 1.7% higher (1.5 to 2.0%). vational studies. This is because mendelian randomisation 4 Journal of Obesity

Table 3: Longitudinal studies reporting the association between childhood BMI and subsequent age at menarche.

Reference Number of Mean age at Mean age at Findings (country) participants BMI assessment followup Mean ± SE BMI by age at menarche Ong et al. 2009 9mo P value 1,781 13 y <12 y 12-13 y >13 y (UK) [33] 19 mo for trend (i) 9 mo 17.5 ± 0.1 17. 3 ± 0.1 17.3 ± 0.1 .007 (ii) 19 mo 16.9 ± 0.1 16.7 ± 0.1 16.7 ± 0.1 .09 Age at menarche (mean, 95% CI) by BMI z-score 1 year before 7.7 y height “take-off” Buyken et al. 2009 87 (Interquartile 13 y P value (i) Lowest BMI quartile 12.9 y (12.4–13.4) (Germany) [34] range 6.5–8.8) for trend (ii) Quartiles 2 and 3 11.7 y (11.4–12.1) .03 (iii) Highest BMI quartile 12.4 y (11.9–12.8) Odds ratio (95% CI) for early menarche (by age 12 y) per +1 BMI 3.0 y z-score Lee et al. 2007 354 4.5 y 12 y (i) 3.0 y OR = 1.45 (1.10–1.93) (USA) [36] 6-7 y (ii) 4.5 y OR = 1.50 (1.14–1.97) (iii) 6-7 y OR = 1.85 (1.38–2.47) Must et al. 2005 12.0 y Age at menarche showed an inverse trend with premenarche BMI 307 15 y (USA) [35] (SD 1.2) Correlation coefficient = −0.10; P = .08. 8.7 y (SD 2) Odds ratio (95% CI) for early menarche (<12 y) per +1 BMI z-score Freedman et al. 771 Whites whites 17 y White girls OR = 2.0 (1.6–2.5) 2003 (USA) [31] 408 Blacks 8.9 y (SD 2) = blacks Black girls OR 2.1 (1.5–3.0) relies on several assumptions [11]. Firstly, that there is a observed association between BMI and early menarche. The reliable association between the genetic variant and the expo- mechanisms of action for these variants are yet unknown and sure, childhood BMI. While we did not have information on it is possible that some might indeed have direct effects on the childhood BMI in this study, the effect size of our 12-variant timing of menarche. Alternatively, the observed association BMI-increasing-allele-score on BMI at 20 years (+0.12 kg/m2 between childhood BMI and age at menarche might have per allele, or around 0.04 of an SD) is identical to that in been artificially diminished due to negative confounding or 9–15-year-old children with research clinic measurements due to our imprecise estimate of childhood BMI. In this case, in the European Youth Heart Study (0.04 SD per allele; the IV-predicated association may actually be closer to true n = 2, 042 children) [27], and is similar to the effect size of the causal association between childhood BMI and age size of a 10-variant score at age 9 years in the ALSPAC at menarche. study (BMI: 0.07 SD per allele; weight 0.05 SD per allele The lack of direct assessments of growth during child- [28]. We are therefore confident that our genetic score is hood was a limitation of our study. Recalled body weight a valid instrumental variable for childhood BMI. Secondly, at age 20 years was used to calculate a proxy measure of the genetic variants should not have pleiotropic effects on childhood BMI. However, other studies have demonstrated different biological processes, or be in linkage disequilibrium that recall of early adult weight in this age group is reliable with other genetic variants that might directly affect the [30] and that BMI at age 20 is well-correlated with childhood outcome [11]. We would expect that pleiotropic effects (i.e., BMI [31]. Furthermore in our IV analyses, BMI at 20 is only mediated by independent biological processes) on BMI and used in order to estimate the effect size of the genetic score on timing of menarche are infrequent among BMI variants. In childhood BMI, and the result was very similar to that found contrast, a recent very large study in 87,000 women reported in earlier childhood studies (discussed above). that individually 9 of the 12 BMI-increasing variants that Our inference of a causal relationship between higher we studied here showed significant associations with lower BMI and early menarche is supported by other sources age at menarche [29]. As most BMI-increasing variants are of evidence. One randomised control trial showed that associated with lower age at menarche, we consider that reduction in childhood obesity led to avoidance of early pleiotropy is unlikely and therefore our findings indicate menarche [32]. However, the lifestyle intervention used in a causal pathway linking higher BMI, or the growth and that trial lead to changes in fruit and vegetable consumption, developmental processes that lead to higher BMI, to earlier duration of physical activity and sedentary behaviour, and it menarche. is therefore not possible to specify any one causal factor. Most It is unclear why our IV-predicted association, based longitudinal studies, while small in numbers, consistently on the genetic score, was even stronger than the directly report that higher BMI during childhood is associated with Journal of Obesity 5 subsequent increased risk of early menarche [31, 33–36] References (Table 3). In those longitudinal studies, associations with earlier menarche were seen with childhood BMI at various [1] R. Lakshman, N. G. Forouhi, S. J. Sharp et al., “Early ages from 9 months through to 12 years. One study reported age at menarche associated with cardiovascular disease and mortality,” Journal of Clinical Endocrinology and Metabolism, that greater weight gain even in the first weeks and months of vol. 94, no. 12, pp. 4953–4960, 2009. life was associated with earlier age at menarche [33]. Further [2] P. Hartge, “Genetics of reproductive lifespan,” Nature Genetics, longitudinal studies are needed to identify the relative effects ff vol. 41, no. 6, pp. 637–638, 2009. of rapid weight during di erent childhood ages, up to and [3]M.S.Golub,G.W.Collman,P.M.D.Fosteretal.,“Public including the pubertal years, on the timing of menarche. health implications of altered puberty timing,” Pediatrics,vol. In a separate birth cohort study [37] we recently exam- 121, no. 3, pp. S218–S230, 2008. ined the relationship between the age at menarche genetic [4]P.B.Kaplowitz,“Linkbetweenbodyfatandthetimingof variant in LIN28B to measures of growth and weight gain puberty,” Pediatrics, vol. 121, supplement 3, pp. S208–S217, from birth to age 53 years. In that study, that the menarche- 2008. lowering allele was associated with higher BMI in women [5]M.L.Ahmed,K.K.Ong,andD.B.Dunger,“Childhood from ages 15–43 years, but not before age 15. That finding is obesity and the timing of puberty,” Trends in Endocrinology therefore consistent with our current analysis in supporting and Metabolism, vol. 20, no. 5, pp. 237–242, 2009. a causal direction of childhood BMI on puberty timing, [6] S.Y.Euling,M.E.Herman-Giddens,P.A.Leeetal.,“Examina- although puberty timing may in turn influence BMI beyond tion of US puberty-timing data from 1940 to 1994 for secular puberty. trends: panel findings,” Pediatrics, vol. 121, supplement 3, pp. Together, these findings indicate that the occurrence of S172–S191, 2008. early menarche is likely to increase as the prevalence of [7]P.Bundred,D.Kitchiner,andI.Buchan,“Prevalenceof overweight and obese children between 1989 and 1998: childhood overweight and obesity increases. An expert panel population based series of cross sectional studies,” BMJ,vol. assembled to assess the evidence for an ongoing secular trend 322, no. 7282, pp. 326–328, 2001. in age at menarche since 1940 failed to reach a consensus, ffi [8] E. Jacobson-Dickman and M. M. Lee, “The influence of although the majority concluded that there was su cient endocrine disruptors on pubertal timing,” Current Opinion in evidence for an ongoing decline [6]. The rapid reductions in Endocrinology, Diabetes and Obesity, vol. 16, no. 1, pp. 25–30, mean age at menarche from the late 1800s to the mid 1900s 2009. have been attributed to the avoidance of under-nutrition, [9] C. B. Jasik and R. H. Lustig, “Adolescent obesity and puberty: while the effects of overnutrition on further advancing age the “perfect storm”,” Annals of the New York Academy of at menarche may result in relatively slower yet continuing Sciences, vol. 1135, pp. 265–279, 2008. trends to earlier menarche [38]. [10] S. Greenland, “An introduction to instrumental variables for epidemiologists,” International Journal of Epidemiology,vol. 29, no. 6, p. 1102, 2000. 5. Conclusions [11]D.A.Lawlor,R.M.Harbord,J.A.C.Sterne,N.Timpson, ff and G. D. Smith, “Mendelian randomization: using genes as Our genetic IV analysis infers a strong causal e ect of instruments for making causal inferences in epidemiology,” higher childhood BMI on the risk of early menarche. Statistics in Medicine, vol. 27, no. 8, pp. 1133–1163, 2008. While mendelian randomisation analysis is dependent on [12] S. Kathiresan, “A PCSK9 missense variant associated with a a number of assumptions, our conclusions are supported reduced risk of early-onset myocardial infarction,” The New by the findings of longitudinal studies and suggest that England Journal of Medicine, vol. 358, no. 21, pp. 2299–2300, increasing trends in childhood obesity will lead to increasing 2008. prevalence of early menarche. [13] A. Thompson, E. di Angelantonio, N. Sarwar et al., “Associa- tion of cholesteryl ester transfer protein genotypes with CETP mass and activity, lipid levels, and coronary risk,” Journal of the Conflict of Interests American Medical Association, vol. 299, no. 23, pp. 2777–2788, 2008. The authors declare no conflict of interest. [14] J. Zacho, A. Tybjærg-Hansen, J. S. Jensen, P. Grande, H. Sillesen, and B. G. Nordestgaard, “Genetically elevated C- Acknowledgments reactive protein and ischemic vascular disease,” The New England Journal of Medicine, vol. 359, no. 18, pp. 1897–1908, TheauthorsaregratefultoalltheparticipantsinEPIC- 2008. Norfolk and in the studies included in the meta-analysis. [15] N. J. Timpson, R. Harbord, G. D. Smith, J. Zacho, A. They thank their colleagues in the EPIC-Norfolk study and Tybjærg-Hansen, and B. G. Nordestgaard, “Does greater adiposity increase blood pressure and hypertension risk?: MRC Epidemiology Unit for data collection, management, mendelian randomization using the FTO/MC4R genotype,” and genotyping. EPIC-Norfolk is supported by research Hypertension, vol. 54, no. 1, pp. 84–90, 2009. programme grant funding from Cancer Research UK and [16] M. Kivimaki,¨ G. D. Smith, N. J. Timpson et al., “Lifetime the Medical Research Council with additional support from body mass index and later atherosclerosis risk in young adults: the Stroke Association, British Heart Foundation, Research examining causal links using mendelian randomization in Into Ageing, Academy of Medical Sciences, and the Wellcome the cardiovascular risk in young finns study,” European Heart Trust. Journal, vol. 29, no. 20, pp. 2552–2560, 2008. 6 Journal of Obesity

[17]T.M.Frayling,N.J.Timpson,M.N.Weedonetal.,“A [32] J. E. Chavarro, K. E. Peterson, A. M. Sobol, J. L. Wiecha, and common variant in the FTO gene is associated with body mass S. L. Gortmaker, “Effects of a school-based obesity-prevention index and predisposes to childhood and adult obesity,” Science, intervention on menarche (United States),” Cancer Causes and vol. 316, no. 5826, pp. 889–894, 2007. Control, vol. 16, no. 10, pp. 1245–1252, 2005. [18] R. J. F. Loos, C. M. Lindgren, S. Li et al., “Common variants [33] K. K. Ong, P. Emmett, K. Northstone et al., “Infancy weight near MC4R are associated with fat mass, weight and risk of gain predicts childhood body fat and age at menarche in girls,” obesity,” Nature Genetics, vol. 40, no. 6, pp. 768–775, 2008. Journal of Clinical Endocrinology and Metabolism, vol. 94, no. [19] C. J. Willer, E. K. Speliotes, R. J. F. Loos et al., “Six new 5, pp. 1527–1532, 2009. loci associated with body mass index highlight a neuronal [34] A. E. Buyken, N. Karaolis-Danckert, and T. Remer, “Associa- influence on body weight regulation,” Nature Genetics, vol. 41, tion of prepubertal body composition in healthy girls and boys no. 1, pp. 25–34, 2009. with the timing of early and late pubertal markers,” American [20] G. Thorleifsson, G. B. Walters, D. F. Gudbjartsson et al., Journal of Clinical Nutrition, vol. 89, no. 1, pp. 221–230, 2009. “Genome-wide association yields new sequence variants at [35]A.Must,E.N.Naumova,S.M.Phillips,M.Blum,B.Dawson- seven loci that associate with measures of obesity,” Nature Hughes, and W. M. Rand, “Childhood overweight and mat- Genetics, vol. 41, no. 1, pp. 18–24, 2009. urational timing in the development of adult overweight and [21] P. Sulem, D. F. Gudbjartsson, T. Rafnar et al., “Genome- fatness: the newton girls study and its follow-up,” Pediatrics, wide association study identifies sequence variants on 6q21 vol. 116, no. 3, pp. 620–627, 2005. associated with age at menarche,” Nature Genetics, vol. 41, no. [36] J. M. Lee, D. Appugliese, N. Kaciroti, R. F. Corwyn, R. H. 6, pp. 734–738, 2009. Bradley, and J. C. Lumeng, “Weight status in young girls and [22] J. R. B. Perry, L. Stolk, N. Franceschini et al., “Meta-analysis of the onset of puberty,” Pediatrics, vol. 119, no. 3, pp. e624–e630, genome-wide association data identifies two loci influencing 2007. age at menarche,” Nature Genetics, vol. 41, no. 6, pp. 648–650, [37] K. K. Ong, C. E. Elks, A. K. Wills et al., “Associations between 2009. the pubertal timing-related variant in LIN28B and BMI vary [23] N. Day, S. Oakes, R. Luben et al., “EPIC-Norfolk: study across the life-course,” Journal of Clinical Endocrinology & design and characteristics of the cohort. European prospective Metabolism, vol. 96, pp. E125–E129, 2011. investigation of cancer,” British Journal of Cancer, vol. 80, [38] K. K. Ong, M. L. Ahmed, and D. B. Dunger, “Lessons from supplement 1, pp. 95–103, 1999. large population studies on timing and tempo of puberty [24] S. Li, J. H. Zhao, J. Luan et al., “Cumulative effects and (secular trends and relation to body size): the European trend,” predictive value of common obesity-susceptibility variants Molecular and Cellular Endocrinology, vol. 254-255, pp. 8–12, identified by genome-wide association studies,” American 2006. Journal of Clinical Nutrition, vol. 91, no. 1, pp. 184–190, 2010. [25] M. G. Frontini, S. R. Srinivasan, and G. S. Berenson, “Longitudinal changes in risk variables underlying metabolic syndrome X from childhood to young adulthood in female subjects with a history of early menarche: the Bogalusa heart study,” International Journal of Obesity, vol. 27, pp. 1398–1404, 2003. [26]K.E.Remsberg,E.W.Demerath,C.M.Schubert,WM.C. Chumlea, S. S. Sun, and R. M. Siervogel, “Early menarche and the development of cardiovascular disease risk factors in adolescent girls: the fels longitudinal study,” Journal of Clinical Endocrinology and Metabolism, vol. 90, no. 5, pp. 2718–2724, 2005. [27] M. den Hoed, U. Ekelund, S. Brage et al., “Genetic susceptibility to obesity and related traits in childhood and adolescence; influence of loci identified by genome-wide association studies,” Diabetes, vol. 59, no. 11, pp. 2980–2988, 2010. [28]C.E.Elks,R.J.F.Loos,S.J.Sharpetal.,“Geneticmarkers of adult obesity risk are associated with greater early infancy weight gain and growth,” PLoS Medicine,vol.7,no.5,Article ID e1000284, 2010. [29] C. E. Elks, J. R. Perry, P. Sulem et al., “Thirty new loci for age at menarche identified by a meta-analysis of genome-wide association studies,” Nature Genetics, vol. 42, no. 12, pp. 1077– 1085, 2010. [30] S. Kovalchik, “Validity of adult lifetime self-reported body weight,” Public Health Nutrition, vol. 12, no. 8, pp. 1072–1077, 2009. [31]D.S.Freedman,L.K.Khan,M.K.Serdula,W.H.Dietz,S.R. Srinivasan, and G. S. Berenson, “The relation of menarcheal age to obesity in childhood and adulthood: the Bogalusa heart study,” BMC Pediatrics, vol. 3, no. 1, article 3, 2003. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 283153, 4 pages doi:10.1155/2011/283153

Research Article An Obesity Risk SNP (rs17782313) near the MC4R Gene Is Associated with Cerebrocortical Insulin Resistance in Humans

Otto Tschritter,1 Axel Haupt,1, 2 Hubert Preissl,3, 4 Caroline Ketterer,1 Anita M. Hennige,1 Tina Sartorius,1 Fausto Machicao,1 Andreas Fritsche,1 and Hans-Ulrich Haring¨ 1

1 Department of Internal Medicine IV, University of Tübingen, Otfried-Müller-Strasse 10, 72076 Tübingen, Germany 2 Eli Lilly and Company, Lilly Deutschland GmbH, 61352 Bad Homburg, Germany 3 Institute of Medical Psychology and Behavioral Neurobiology, University of Tübingen, 72076 Tübingen, Germany 4 Department of Obstetrics and Gynecology, College of Medicine, University of Arkansas for Medical Sciences, Little Rock, AR 72205, USA

Correspondence should be addressed to Otto Tschritter, [email protected]

Received 30 November 2010; Revised 2 March 2011; Accepted 4 April 2011

Academic Editor: Jack A. Yanovski

Copyright © 2011 Otto Tschritter et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Activation of melanocortin-4 receptor (MC4R) by insulin sensitive neurons is a central mechanism in body weight regulation, and genetic variants in the MC4R gene (e.g., rs17782313) are associated with obesity. By using magnetoencephalography, we addressed whether rs17782313 affects the cerebrocortical insulin response. We measured the cerebrocortical insulin response by using magnetoencephalography in a hyperinsulinemic euglycemic clamp (versus placebo) in 51 nondiabetic humans (26 f/25 m, age 35 ± 3years,BMI28± 1kg/m2). The C-allele of rs17782313 was minor allele (frequency 23%), and the genotype distribution (TT 30, TC 19, CC 2) was in Hardy-Weinberg-Equilibrium. Insulin-stimulated cerebrocortical theta activity was decreased in the presence of the C-allele (TT 33 ± 16 fT; TC/CC −27 ± 20 fT; P = .023), and this effect remained significant after adjusting for BMI and peripheral insulin sensitivity (P = .047). Cerebrocortical theta activity was impaired in carriers of the obesity risk allele. Therefore, cerebral insulin resistance may contribute to the obesity effect of rs17782313.

1. Introduction status. While POMC derivates like alpha-MSH and beta- MSH stimulate melanocortin receptors, agouti-related pro- Melanocortin receptors (MC3R and MC4R) have been dem- tein (AgRP) is known to be a natural antagonist. As leptin onstrated in multiple brain regions including the hypothala- and insulin activate POMC neurons and suppress AgRP mus [1, 2] and represent critical components of a regulating neurons, both hormones contribute to the regulation of body system for body weight and energy homeostasis. Both dis- weight and energy homeostasis via melanocortin receptors, ruption of MC4R in mice [3] and mutations in the coding and knock-out of MC4R results in decreased action of leptin region of human MC4R result in a severely obese phenotype and insulin in the brain [9]. [4, 5]. Another relatively rare (2–4%) polymorphism in We previously established a method to measure acute the MC4R coding region has been reported to protect insulin responses in the human brain by combining magne- from obesity [6]. In recent genome-wide association studies toencephalography (MEG) and the hyperinsulinemic eug- (GWAS) also common genetic variants near the MC4R gene lycemic clamp technique [10]. In this study, we observed were associated with BMI [7], waist circumference, and that cerebral insulin resistance is associated with obesity in insulin resistance [8]. humans and therefore speculated that a decreased insulin Melanocortin receptors receive information from POMC response of the brain might contribute to obesity caused by and AgRP neurons about the nutritional and metabolic genetic alterations of MC4R. As rs17782313 had the strongest 2 Journal of Obesity

BMI signal in a GWAS study [7], we studied the effect of Table 1: Subject characteristics and effect of rs17782313 on obesity this single nucleotide polymorphism (SNP) on the insulin measures and peripheral insulin sensitivity. response of the brain. Genotype TT TC/CC P N (%) 30 (59%) 21 (41%) — 2. Methods Gender (F/M) 14/16 12/9 — 2.1. Human Subjects and Experimental Design. We deter- Age (years) 33 ± 238± 3.14 mined rs17782313 in the MC4R gene region in 51 subjects Weight (kg) 81.4 ± 3.483.9 ± 3.6.47∗ who were healthy by self-report and clinical examination BMI (kg/m2)27.1 ± 0.928.4 ± 1.0.62∗ and presented nondiabetic in an oral glucose tolerance test Body fat (%) 27.1 ± 1.630.5 ± 1.9.57∗ according to WHO/ADA criteria. Detailed characteristics of Waist circumference 92 ± 395± 3.34∗ these subjects are given in Table 1. (cm) Fasting plasma glucose 4.9 ± 0.15.1 ± 0.1.35∗ 2.2. Hyperinsulinemic Euglycemic Clamp and Saline Experi- (mmol/L) ment with Measurement of Cerebrocortical Activity by Magne- 2Hrglucose 6.1 ± 0.36.6 ± 0.4.51∗ toencephalography (MEG). To measure the insulin response (mmol/L)∗∗ of the brain, these subjects participated in an insulin and Insulin sensitivity index 0.074 ± 0.009 0.062 ± 0.010 .67∗ aplacebo(=saline) experiment in random order on two (µmol·kg/−1·pM−1)∗∗∗ ff di erent days approximately 1 to 2 weeks apart. Each M ± SE; ∗adjusted for age and gender; ∗∗OGTT; ∗∗∗Hyperinsulinemic experiment started at approximately 7.00 a.m. and consisted euglycemic clamp. of a 30-minute baseline period, and a 2-step hyperinsulinemic euglycemic clamp or saline infusion. To maintain blood glucose at baseline levels a standard hyperinsulinemic insulin experiment to correct for potential placebo effects euglycemic clamp protocol was followed. The details of the and daytime variation. The change of the investigated MEG clamp procedures and the neurophysiologic measurements parameters from basal to the second step of the clamp was performed by MEG have been described in [10]. Here we considered to indicate the total insulin effect and was used in used the change of spontaneous cortical beta and theta activ- further analyses. ity during insulin infusion (corrected for placebo derived Unless otherwise stated, data are means ± SE. Non- changes) to quantify the cerebrocortical response to insulin. normally distributed variables (Shapiro-Wilk W test) were Beta and theta activity were extracted from spontaneous logarithmically transformed. To adjust for covariates and cortical activity by using fast Fourier transformation. to identify independent relationships, we performed linear multiple regression analyses. In general, a P value of <.05 was 2.3. Analytical Procedures and Measurement of Body Fat. considered to indicate statistical significance. As we studied Plasma glucose was determined during the OGTT using the insulin-induced changes in beta and theta activity, correction glucose oxidase method (YSI, Yellow Springs Instruments, for multiple comparisons requires an alpha level of 0.025 Yellow Springs, CO, USA). Blood glucose was determined instead of 0.05 for the primary analysis of MEG data. Due in the clamp experiments by a HemoCue blood glucose to the very low number of CC carriers, determination of photometer (HemoCue AB, Aengelholm, Sweden). Plasma a per allele risk as suggested by another study [7]wasnot insulin levels were determined by microparticle enzyme possible. Therefore, TC and CC carriers were treated as one immunoassay (Abbott Laboratories, Tokyo, Japan). Body group and compared with wild-type carriers in all further composition was measured by bioelectrical impedance anal- analyses. For all statistical analyses, the software package JMP ysis (BIA-101A, RJL Systems, Detroit, Michigan, USA) and (SAS Institute, Cary, NC) was used. expressed as percent body fat. 2.6. Ethical Statement. The study protocol was approved by 2.4. Genotyping. For genotyping, DNA was isolated from the Ethical Committee of the University of Tubingen¨ and whole blood using a commercial DNA isolation kit (Nucle- informed written consent had been obtained from each oSpin; Macherey & Nagel, Duren,¨ Germany). The SNPs subject prior to the studies. The whole study protocol has were genotyped using the TaqMan assay (Applied Biosys- been designed in accordance with the declaration of Helsinki tems, Foster City, CA). The TaqMan genotyping reaction (1964). Furthermore, we certify that all applicable institutio- was amplified on a GeneAmp PCR system 7000 (Applied nal and governmental regulations concerning the ethical use Biosystems) (50◦Cfor2minutes,95◦Cfor10minutes, of human volunteers were followed during the research. followed by 40 cycles of 95◦C for 15 seconds and 60◦Cfor 1 minute), and fluorescence was detected on an ABI Prism 3. Results sequence detector (Applied Biosystems). Quality control was performed as reported in [11]. 3.1. Genotype Data and Body Weight Effect of rs17782313. The genotype distribution (TT 30, TC 19, and CC 2) was 2.5. Statistical Analysis. For statistical analyses, MEG data in Hardy-Weinberg-Equilibrium, and the minor C-allele of the saline experiment were subtracted from data of the had a frequency of 23%. Carriers of the rare allele were Journal of Obesity 3

300 adjust the cerebral effect of the polymorphism for obesity and for peripheral insulin resistance. As the SNP effect on 200 the cerebrocortical insulin response remained significant, our data suggest that rs17782313 directly alters the insulin 100 response of the brain. In contrast to the cerebrocortical effect, the differences 0 in obesity measures and metabolic parameters between the genotype group were not significant, though the effect size −100 seems comparable to previous reports on MC4R polymor- thetaactivity(fT) phisms [7, 8]. One explanation for the lack of significance Insulin-induced change in −200 in these parameters is that altered brain function is likely to be the primary effect of the MC4R SNP, while obesity and −300 metabolic alterations are secondary. Furthermore, BMI and TT TC CC waist circumference have limited accuracy in small sample rs17782313 sizes. Figure 1: Insulin effect on theta activity in the genotype groups: in We have previously demonstrated reduced insulin re- subjects carrying the rare allele of rs17782313 in the homozygous sponses of beta and theta activities in obese humans [10]. (TC) or the heterozygous form (CC), the effect of insulin on theta Notably, the effect of rs17782313 occurred only in the theta activity was significantly reduced as compared to carriers of the activity band, while there was no effect on beta activity. In wild-type allele (TT) (P = .023). The diamonds show the means contrast, we observed that variation in the FTO and IRS- of each group (horizontal line) and the confidence intervals of 1 gene is associated with a reduction of insulin-mediated ff the means (upper and lower point). The e ect of the genotype beta activity, while theta activity is not affected [10, 12]. remained significant after adjusting for BMI (P = .038) and also Furthermore, we reported that beta and theta activities after adjusting for BMI and peripheral insulin sensitivity (P = .047). are differently affected by aging [13] and suggested that these parameters may represent different brain structures and functions and their responses to insulin. Therefore, 2.5 kg heavier, displayed a slightly increased BMI, waist cir- these data suggest that suppressed cerebral insulin effects cumference, and body fat as compared to the wild-type by genetic variation in FTO and aging involve different carriers; however, all these effects failed statistical significance mechanisms than variation in the MC4R gene. (Table 1). According to the weight effect, peripheral insulin It is interesting to note, that theta activity is mainly sensitivity was reduced in CC carriers. generated in the hippocampus [14]. Furthermore, the hippocampus has been shown to contribute to food-related reward 3.2. Cerebrocortical Insulin Resistance in Carriers of in obese women [15]. Additionally, theta activity has been rs17782313. Regarding the effect of insulin on the cerebral attributed to locomotor activity and voluntary movements cortex, we observed a decreased effect on theta activity [16]. Therefore, a decreased response of theta activity to (Figure 1) in presence of the C-allele (TT 33 ± 16 fT, TC/CC insulin by the obesity risk allele of rs17782313 may indicate −27 ± 20 fT, P = .023). This association remained significant increased food reward and decreased locomotor activity as after adjusting for BMI (P = .038) and also after adjusting possible mechanisms that contribute to the development of for BMI and peripheral insulin sensitivity (P = .047). In obesity in these subjects. contrast to theta activity, there was no significant reduction of insulin-mediated beta activity by rs17782313 (TT 3.0±3.2 5. Conclusions fT, TC/CC −1.3 ± 4.1fT,P = .41). Genetic variation in MC4R leads to an impaired insulin 4. Discussion response in theta activity in the human brain. Though specific suggestions on underlying brain functions are specu- ff Our data demonstrate that the insulin response of the brain lative, our data support that these functions may be di erent measured by magnetoencephalography (MEG) is diminished from alterations observed with aging and variation in the in carriers of rs17782313. From a mechanistical point of FTO gene. view, these data support the idea that a decreased insulin response of the brain is a pathogenic factor that contributes Acknowledgments to the effect of this polymorphism on obesity. However, association data have some limitations, as they cannot definitely The authors gratefully acknowledge the superb technical prove the direction of an effect and provide only indirect assistance of Anna Teigeler, Heike Luz, and Gabi Walker. information on the underlying mechanisms. Here, it is clear This study was supported by the Deutsche Forschungsge- that the primary site of the obesity effect must be located in meinschaft (KFO 114 and FR 1561/4-1) and by the German thebrain,asMC4Rishighlyexpressedinthebrainbutnot Federal Ministry of Education and Research in the form of in peripheral tissues [1]. To get the best possible information grants to the German Center for Diabetes Research (DZD from our data, we performed multiple regression models to e.V.) and to the Kompetenznetzwerk Adipositas 01GI0849. 4 Journal of Obesity

References [16] E. Dzoljic, R. van Leeuwen, R. de Vries, and M. R. Dzoljic, “Vigilance and EEG power in rats: effects of potent inhibitors [1]K.G.Mountjoy,M.T.Mortrud,M.J.Low,R.B.Simerly, of the neuronal nitric oxide synthase,” Naunyn-Schmiedeberg’s and R. D. Cone, “Localization of the melanocortin-4 receptor Archives of Pharmacology, vol. 356, no. 1, pp. 56–61, 1997. (MC4-R) in neuroendocrine and autonomic control circuits in the brain,” Molecular Endocrinology, vol. 8, no. 10, pp. 1298– 1308, 1994. [2]L.Roselli-Rehfuss,K.G.Mountjoy,L.S.Robbinsetal., “Identification of a receptor for gamma melanotropin and other proopiomelanocortin peptides in the hypothalamus and limbic system,” Proceedings of the National Academy of Sciences of the United States of America, vol. 90, pp. 8856–8860, 1993. [3] D. Huszar, C. A. Lynch, V. Fairchild-Huntress et al., “Targeted disruption of the melanocortin-4 receptor results in obesity in mice,” Cell, vol. 88, no. 1, pp. 131–141, 1997. [4] G. S. Yeo, I. S. Farooqi, S. Aminian, D. J. Halsall, R. G. Stanhope, and S. O’Rahilly, “A frameshift mutation in MC4R associated with dominantly inherited human obesity,” Nature Genetics, vol. 20, no. 2, pp. 111–112, 1998. [5]C.Vaisse,K.Clement,B.Guy-Grand,andP.Froguel,“A frameshift mutation in human MC4R is associated with a dominant form of obesity,” Nature Genetics, vol. 20, no. 2, pp. 113–114, 1998. [6] E. H. Young, N. J. Wareham, S. Farooqi et al., “The V103I polymorphism of the MC4R gene and obesity: population based studies and meta-analysis of 29 563 individuals,” International Journal of Obesity, vol. 31, no. 9, pp. 1437–1441, 2007. [7] R. J. Loos, C. M. Lindgren, S. Li et al., “Common variants near MC4R are associated with fat mass, weight and risk of obesity,” Nature Genetics, vol. 40, no. 6, pp. 768–775, 2008. [8] J. C. Chambers, P. Elliott, D. Zabaneh et al., “Common genetic variation near MC4R is associated with waist circumference and insulin resistance,” Nature Genetics, vol. 40, no. 6, pp. 716– 718, 2008. [9]M.W.Schwartz,S.C.Woods,D.PorteJr.,R.J.Seeley,and D. G. Baskin, “Central nervous system control of food intake,” Nature, vol. 404, no. 6778, pp. 661–671, 2000. [10]O.Tschritter,H.Preissl,A.M.Hennigeetal.,“Thecerebro- cortical response to hyperinsulinemia is reduced in overweight humans: a magnetoencephalographic study,” Proceedings of the National Academy of Sciences of the United States of America, vol. 103, no. 32, pp. 12103–12108, 2006. [11] N. Stefan, F. Machicao, H. Staiger et al., “Polymorphisms in the gene encoding adiponectin receptor 1 are associated with insulin resistance and high liver fat,” Diabetologia, vol. 48, no. 11, pp. 2282–2291, 2005. [12]O.Tschritter,H.Preissl,Y.Yokoyama,F.Machicao,H.U. Haring, and A. Fritsche, “Variation in the FTO gene locus is associated with cerebrocortical insulin resistance in humans,” Diabetologia, vol. 50, no. 12, pp. 2602–2603, 2007. [13]O.Tschritter,A.M.Hennige,H.Preissletal.,“Insulineffects on beta and theta activity in the human brain are differentially affected by ageing,” Diabetologia, vol. 52, no. 1, pp. 169–171, 2009. [14] C. D. Tesche and J. Karhu, “Theta oscillations index human hippocampal activation during a working memory task,” Proceedings of the National Academy of Sciences of the United States of America, vol. 97, no. 2, pp. 919–924, 2000. [15] L. E. Stoeckel, R. E. Weller, E. W. Cook III, D. B. Twieg, R. C. Knowlton, and J. E. Cox, “Widespread reward-system activation in obese women in response to pictures of high- calorie foods,” Neuroimage, vol. 41, no. 2, pp. 636–647, 2008. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 269043, 4 pages doi:10.1155/2011/269043

Research Article Sequence Analysis of the UCP1 Gene in a Severe Obese Population from Southern Italy

Giuseppe Labruna,1 Fabrizio Pasanisi,2 Giuliana Fortunato,3, 4 Carmela Nardelli,3, 4 Carmine Finelli,5 Eduardo Farinaro,6 Franco Contaldo,2 and Lucia Sacchetti3, 4

1 Fondazione IRCCS SDN, Istituto di Ricerca Diagnostica e Nucleare, Via Gianturco 113, 80143 Naples, Italy 2 Centro Interuniversitario di Studi e Ricerche sull’Obesita` e Dipartimento di Medicina Clinica e Sperimentale, Universita` degli Studi di Napoli Federico II, Via Pansini 5, 80131 Naples, Italy 3 CEINGE Biotecnologie Avanzate S.C. a R.L., Via Gaetano Salvatore 486, 80145, Naples, Italy 4 Dipartimento di Biochimica e Biotecnologie Mediche, UniversitadegliStudidiNapoliFedericoII,` Via Pansini 5, Via Pansini 5, 80131 Naples, Italy 5 Fondazione Stella Maris Mediterraneo, Centro Disturbi del Comportamento Alimentare e del Peso “G. Gioia”, Chiaromonte, C/da S. Lucia, 85100, Chiaromonte, Potenza, Italy 6 Dipartimento di Scienze Mediche Preventive, Universita` degli Studi di Napoli Federico II, Via Pansini 5, 80131 Naples, Italy

Correspondence should be addressed to Lucia Sacchetti, [email protected]

Received 1 December 2010; Accepted 8 April 2011

Academic Editor: Francesco Saverio Papadia

Copyright © 2011 Giuseppe Labruna et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Brown adipose tissue, where Uncoupling Protein 1 (UCP1) activity uncouples mitochondrial respiration, is an important site of facultative energy expenditure. This tissue may normally function to prevent obesity. Our aim was to investigate by sequence analysis the presence of UCP1 gene variations that may be associated with obesity. We studied 100 severe obese adults (BMI > 40 kg/m2) and 100 normal-weight control subjects (BMI range = 19–24.9 kg/m2). We identiﬁed 7 variations in the promoter region, 4 in the intronic region and 4 in the exonic region. Globally, 72% of obese patients bore UCP1 polymorphisms. Among UCP1 variants, g.IVS4−208T>G SNP was associated with obesity (OR: 1.77; 95% CI = 1.26–2.50; P = .001). Further, obese patients bearing the g.−451C>T (CT+TT) or the g.940G>A (GA+AA) genotypes showed a higher BMI than not polymorphic obese patients (P = .008 and P = .043, resp.). In conclusion, UCP1 SNPs could represent “thrifty” factors that promote energy storage in prone subjects.

1. Introduction has been calculated that BAT malfunction could lead to a weight gain of 1-2 kg/yr [9]. These data suggest that BAT Brown adipose tissue (BAT) plays an important role in specific proteins, such as UCP1, could be involved in obesity energy expenditure [1]. Its thermogenic activity requires not onset so representing a possible target of pharmaceutical only the presence of a dense vascularisation and sympathetic interventions in this field [10, 11]. In the last years, UCP1 loss innervation, but also the expression of Uncoupling Protein 1 has been associated with obesity susceptibility in UCP1−/− (UCP1) [2, 3]. UCP1 is localized on the inner mitochondrial mice, particularly during aging and a high-fat diet [12, membrane where it uncouples oxidative metabolism from 13]. We previously described the association between the ATP synthesis, resulting in the dissipation of energy through variation −3826A>G in the UCP1 promoter and a severe the release of heat [4]. In humans, BAT exerts its function fatty liver steatosis during obesity [14].Theaimofthisstudy especially during the first years of life and decreases with age was to search for further gene alterations associated with [5]. However, several metabolic active depots of BAT have obese phenotype in the UCP1 gene (ENSG00000109424) by been recently demonstrated also in adult humans [6–8]. It sequence analysis. 2 Journal of Obesity

Table 1: General and biochemical characteristics of obese patients The mean value and the standard deviation (SD) were and control subjects. calculated for each investigated parameter. The Mann- Whitney test and/or χ2, when necessary, were used for Obese patients Control subjects between-group comparisons. Differences were considered n = n = ( 100) ( 100) significant at P level <.05. Linkage analysis was performed Females (%) 60 64 by using Haploview 4.0 software [15]. Binomial logistic Age (years) 32.1 ± 10.933.3 ± 8.1 regression analysis was used to investigate the association BMI∗ (kg/m2)47.9 ± 6.922.8 ± 2.1 between the biochemical and genetic characteristics (i.e., Adiponectin∗ (μg/mL) 31.6 ± 30.053.8 ± 38.6 glucose, total cholesterol and triacylglycerols concentrations ∗ and AST activity; g.−451C>T, g.940G>A, g.IVS4−208, and Leptin (ng/mL) 119.6 ± 72.421.9 ± 18.7 g.6537A>T polymorphisms) and the condition of being . ± . . ± . Resistin (ng/mL) 12 2 8 4127 7 9 obese, after adjustment for age and sex. ∗ . ± . . ± . Glucose (mmol/L) 4 9 0 845 0 4 Statistical analyses were carried out with the PASW Total cholesterol (mmol/L) 4.7 ± 1.15.0 ± 0.7 package for Windows (Ver.18; SPSS Inc. Headquarters, Triacylglycerols∗ (mmol/L) 1.5 ± 0.60.9 ± 0.3 Chicago, Ill). AST∗ (U/L) 26.5 ± 16.719.8 ± 5.6 ALT∗ (U/L) 39.8 ± 35.022.5 ± 12.9 3. Results and Discussion GGT∗ (U/L) 35.3 ± 26.017.4 ± 10.4 Adiponectin and leptin concentrations were statistically dif- Creatinine (mg/dL) 0.9 ± 0.20.7 ± 0.1 P<. ∗ ferent ( 001) between obese and control subjects (mean Statistically significant difference between obese and control subjects, P< ± ± μ . level SD respectively: adiponectin 31.6 30.0 g/mL ver- 001 at Mann-Whitney test. Biochemical parameters were measured by ± μ ± ± routine laboratory methods. Adipokines concentrations were measured by sus 53.8 38.6 g/mL; leptin 119.6 72.4 versus 21.9 ELISA assay (LINCO Research, Mo, USA). Values are expressed as mean ± 18.7 ng/mL). Higher concentrations or activities of glucose, SD. triacylglycerols, AST, ALT and GGT were measured in obese patients than in controls (P<.001) (Table 1). We identified 15 sequence variations in UCP1 gene (Table 2): 7 in the promoter region (3/7 described for the 2. Materials and Methods first time), 4 in the intronic regions (1/4 described for the first time) and 4 in the exonic regions (2 in the 5 UTR; 2 in We studied 200 age-matched unrelated Caucasian subjects the translated region). Globally, 72% of obese patients bore from Southern Italy: 100 adult severe obese patients (60% one or more UCP1 polymorphisms. female, mean ± SD: BMI = 47.9 ± 6.9 kg/m2;age= There were no differences in genotype frequencies 32.1 ± 10.9 years) and 100 unrelated adult normal-weight between obese and control subjects at level of the detected subjects (64% female, mean ± SD: BMI = 22.8 ± 2.1 kg/m2; SNPs, except for g.IVS4−208T>G polymorphism more age = 33.3 ± 8.1 years). The patients were recruited at the frequent in obese than in control subjects (P = .002). After a obesity outpatient clinic of the Department of Clinical and permutation test with 100000 permutations, the association Experimental Medicine, University of Naples Federico II, of the polymorphic allele with the obese phenotype remained Italy, from 2007 to 2008, whereas control subjects were statistically significant (P = .017). Subjects bearing this recruited at the Department of Preventive Medical Science polymorphism (TG or GG) were at high risk for obesity (OR: of the Federico II University Hospital. Clinical and bio- 1.774; 95% CI = 1.26–2.50, P = .001). At binomial logistic chemical data were obtained from each patient on their first regression analysis, the g.IVS4−208 (TG+GG) genotype admission. The general and biochemical characteristics of was confirmed to be statistically associated in our patients the studied populations are reported in Table 1.Allpatients with obesity independently of sex and age (OR: 22.0; 95% and controls gave their informed consent to the study, which CI = 5.6–87.1). This SNP did not alter the splicing site was carried out according to the Helsinki II Declaration. The nor the branch site [16, http://www.umd.be/HSF/], and research was also approved by the Ethics Committee of the the polymorphic allele did not change the ΔG of the School of Medicine, University of Naples Federico II. predicted mRNA secondary structure by mfold analysis Genomic DNA was extracted from whole blood (Nucl- (http://mfold.bioinfo.rpi.edu), suggesting that the stability eon BACC-II; Amersham Science Europe). UCP1 5 flank- of the polymorphic mRNA is the same as the wild-type. ing region, exons and intron-exon junction regions were The G allele may be a marker linked to other gene variants amplified by ten sets of primers (primers ID: RSA000984680, promotingenergystorageaswellasfataccumulationin RSA000984677, RSA000984675, RSA000984673, RSA00098 prone subjects. 4666, RSA000990288, RSA000990284, RSA000990283, RSA0 The novel UCP1 variants g.−637T>C, g.−206C>A, and 00990281, and RSA000990278 http://www.ncbi.nlm.nih.gov/ g.IVS2+174T>A, each of them present in a single obese sites/entrez). PCR products were sequenced on ABI Prism patient, were not associated with differences in clinical 3130 Genetic Analyzer (Applied Biosystems, Foster City, and/or biochemical parameters measured in the obese and CA). PCR conditions were 96◦Cfor5min;than94◦Cfor30 control populations. Among them, only the g.−206C>A sec, 60◦C for 45 sec and 72◦C for 45 sec, for 40 cycles; final occurred in a conserved region indentified by cisRED algo- extension at 72◦C for 10 min; final soak at 25◦C. rithm (http://www.cisred.org/)asacis-regulatory element Journal of Obesity 3

Table 2: UCP1 sequence variations and their frequencies in obese and control subjects.

Polymorphisms Obese patients Control subjects n = 100 n = 100 Position rs# wt HE HO wt HE HO g.−637T>C1 99 1 0 100 0 0 g.−451C>T rs36207410 82 16 2 86 14 0 g.−412A>C rs3811787 57 36 7 49 43 8 g.−372A>C rs1800660 97 3 0 97 3 0 g.−206C>A1 99 1 0 100 0 0 g.−56C>T rs3749539 91 9 0 90 10 0 g.−17C>G1 94 6 0 94 6 0 g.12A>C rs10011540 91 9 0 90 10 0 g.21G>A rs1800661 86 13 1 79 21 0 g.940G>A (p.A64T) rs45539933 92 8 0 91 9 0 g.IVS2+138C>T rs7688743 80 15 5 70 27 3 g.IVS2+174T>A1 99 1 0 100 0 0 g.IVS2+201T>G rs2071416 79 21 0 77 22 1 g.IVS4−208T>G2 rs1494808 45 44 11 69 23 8 g.6537A>T (p.M229L) rs2270565 89 11 0 87 13 0 1 New variants; 2More frequent polymorphism in obese patients (P = .002) than in controls. wt: wild-type homozygous subjects; HE: heterozygous and HO: homozygous subjects at level of the detected variant.

(craHsap157022), and we could hypothesize to alter the The g.12A>C polymorphism is located in the insulin interaction with transcriptional factors. response sequence (IRS). In in vitro experiments, the DNA Regarding the previously described UCP1 polymor- mutated C allele was demonstrated to reduce the transcrip- phisms, a higher mean BMI was observed in our obese tion of UCP1 by 40% respect to the wild-type allele. This patients bearing the g.−451C>T (CT+TT) than in not variation was hypothesized to impair the affinity of the polymorphic obese patients (resp., 52.6 ± 7.4 kg/m2 versus transcription factors for the consensus motif of IRS [18]. 47.0 ± 6.6 kg/m2, P = .008). Further, this SNP was previously indicated as contributing The amino acidic substitution p.M229L (g.6537A>T) in to hepatic lipid accumulation and altering insulin sensitivity the fifth helix of the protein is due to an A>T transversion in in Japanese individuals with Type II diabetes mellitus the 5th exon of the UCP1 gene [17]. Mori and colleagues [18] (NIDDM) [18]. In our population, the lack of association found a higher frequency of the Leu allele of the p.M229L of this SNP with any obesity-related phenotype could be (g.6537A>T) polymorphism in a Japanese obese population duetotheyoungermeanageofourpatientsrespectto with Type II diabetes, indicating this gene variation as those investigated by Fukuyama et al. [21] (32.1 years versus a diabetes-associated SNP, while other studies failed to 56.6 years, resp.) and to different ethnic background of the demonstrate such association [9, 19, 20]. In our study we studied groups. found that patients carrying the polymorphic allele for the The amino acidic substitution p.A64T (g.940G>A) in the p.M229L polymorphism showed a slightly higher mean BMI first matrix loop of the protein is due to a G>A transition in than the wild-type patients (50.6 kg/m2 versus 47.6 kg/m2, the 2nd exon of the UCP1 gene [17]. resp.) while no difference were found at level of glucose and Cha et al. [22] reported in a Korean female population insulin concentration or regarding the homeostatic model an association between the mutated allele and a higher blood assessment (HOMA) index (a measure of insulin sensitivity) pressure. In our population, polymorphic patients compared (data not shown). This difference could be due to the lower to wild-type patients showed a higher mean BMI (52.0 ± mean age of our studied subjects (32.1 years in our patients 6.4 kg/m2 versus 47.5 ± 6.9 kg/m2, P = .043) but only a trend versus 58.6 years in Mori et al. [18]), since Type II diabetes is toward a higher mean systolic blood pressure (130.0 mmHg more frequent in middle aged than in young adult patients. versus 124.4 mmHg, resp.). This difference does not raise Further, the haplotype investigation by Haploview soft- the statistically significant level probably due to the lower ware showed a significant linkage disequilibrium among the number of patients in our examined casistic. three SNPs g.−56C>T (a), g.12A>C (b) and g.940G>A (c) (a-b: log likelihood ratio, LOD = 27.5; r2 = 1; b-c and a- 4. Conclusions c: LOD = 22.6; r2 = 0.9); however no statistically significant association was observed between obesity and this haplotype, Functional activity of BAT has been recently demonstrated the frequency of this latter being the same in obese and in adult humans [6–8] and its amount is inversely related to control subjects (8.0% versus 9.0%, resp.). body fat percentage [23]. We do not have any information 4 Journal of Obesity in our patients about BAT amount. However, variations in [13] H. M. Feldmann, V. Golozoubova, B. Cannon, and J. Ned- the BAT marker UCP1 gene were present in most of our ergaard, “UCP1 ablation induces obesity and abolishes diet- obese patients. These variations could represent common induced thermogenesis in mice exempt from thermal stress by factors contributing to the development of obesity, partic- living at thermoneutrality,” Cell Metabolism,vol.9,no.2,pp. ularly, g.−451C>T, g.940G>A, and g.IVS4−208T>Gcould 203–209, 2009. represent “thrifty” factors that promote energy storage. The [14] G. Labruna, F. Pasanisi, C. Nardelli et al., “UCP1 -3826 AG+GG genotypes, adiponectin, and leptin/adiponectin ratio precise role in obesity of these variants should be investigated in severe obesity,” Journal of Endocrinological Investigation, vol. in a larger casistic. 32, no. 6, pp. 525–529, 2009. [15] J. C. Barrett, B. Fry, J. Maller, and M. J. Daly, “Haploview: Acknowledgments analysis and visualization of LD and haplotype maps,” Bioin- formatics, vol. 21, no. 2, pp. 263–265, 2005. The authors thank Jean Ann Gilder (Scientific Communica- [16] F. O. Desmet, D. Hamroun, M. Lalande, G. Collod-Beroud,´ M. tion srl) for text revision and editing. The work suppor- Claustres, and C. Beroud,´ “Human splicing finder: an online ted by grants Conv. CEINGE-Regione Campania (DGRC bioinformatics tool to predict splicing signals,” Nucleic Acids 1901/2009), Regione Campania LR n5/2005 and MIUR Research, vol. 37, no. 9, article e67, 2009. PRIN 2008, and Progetto di Ricerca Finalizzata RF-SDN- [17] J. Jimenez-Jim´ enez,´ R. Zardoya, A. Ledesma et al., “Evolu- tionarily distinct residues in the uncoupling protein UCP1 2007-635809 (Ministero del Lavoro, della Salute e delle are essential for its characteristic basal proton conductance,” Politiche Sociali). Journal of Molecular Biology, vol. 359, no. 4, pp. 1010–1022, 2006. References [18] H. Mori, H. Okazawa, K. Iwamoto, E. Maeda, M. Hashiramoto, and M. Kasuga, “A polymorphism in the [1] S. R. Farmer, “Molecular determinants of brown adipocyte 5’ untranslated region and a Met229−− >Leu variant in exon formation and function,” Genes and Development, vol. 22, no. 5 of the human UCP1 gene are associated with susceptibility 10, pp. 1269–1275, 2008. to type II diabetes mellitus,” Diabetologia, vol. 44, no. 3, pp. [2] M. C. Zingaretti, F. Crosta, A. Vitali et al., “The presence of 373–376, 2001. UCP1 demonstrates that metabolically active adipose tissue [19] A. Hamann, J. Tafel, B. Busing,¨ H. Munzberg,¨ A. Hinney, H. in the neck of adult humans truly represents brown adipose Mayer et al., “Analysis of the uncoupling protein-1 (UCP1) tissue,” The FASEB Journal, vol. 23, no. 9, pp. 3113–3120, 2009. gene in obese and lean subjects: identification of four amino [3] M. Rosenbaum and R. L. Leibel, “Adaptive thermogenesis in acid variants,” International Journal of Obesity, vol. 22, no. 9, humans,” International Journal of Obesity, vol. 34, pp. S47–S55, pp. 939–941, 1998. 2010. [20] K. S. Vimaleswaran, V. Radha, R. Deepa, and V. Mohan, [4] B. Cannon and J. Nedergaard, “Metabolic consequences of “Absence of association of metabolic syndrome with the presence or absence of the thermogenic capacity of PPARGC1A, PPARG and UCP1 gene polymorphisms in brown adipose tissue in mice (and probably in humans),” Asian Indians,” Metabolic Syndrome and Related Disorders, International Journal of Obesity, vol. 34, pp. S7–S16, 2010. vol. 5, no. 2, pp. 153–162, 2007. [5] A. Frontini and S. Cinti, “Distribution and development of [21] K. Fukuyama, T. Ohara, Y. Hirota et al., “Association of the - brown adipocytes in the murine and human adipose organ,” 112A>C polymorphism of the uncoupling protein 1 gene with Cell Metabolism, vol. 11, no. 4, pp. 253–256, 2010. insulin resistance in Japanese individuals with type 2 diabetes,” [6] S. Enerback,¨ “Brown adipose tissue in humans,” International Biochemical and Biophysical Research Communications, vol. Journal of Obesity, vol. 34, pp. S43–S46, 2010. 339, no. 4, pp. 1212–1216, 2006. [7] K. A. Virtanen, M. E. Lidell, J. Orava et al., “Functional brown [22] M. H. Cha, B. K. Kang, D. Suh, K. S. Kim, Y. Yang, and adipose tissue in healthy adults,” The New England Journal of Y. Yoon, “Association of UCP1 genetic polymorphisms with Medicine, vol. 360, no. 15, pp. 1518–1525, 2009. blood pressure among Korean female subjects,” Journal of [8] A. M. Cypess, S. Lehman, G. Williams et al., “Identification Korean Medical Science, vol. 23, no. 5, pp. 776–780, 2008. and importance of brown adipose tissue in adult humans,” The [23] W. D. van Marken Lichtenbelt, J. W. Vanhommerig, N. M. New England Journal of Medicine, vol. 360, no. 15, pp. 1509– Smulders et al., “Cold-activated brown adipose tissue in 1517, 2009. healthy men,” The New England Journal of Medicine, vol. 360, [9] S. A. Urhammer, M. Fridberg, T. I. Sørensen et al., “Studies no. 15, pp. 1500–1508, 2009. of genetic variability of the uncoupling protein 1 gene in Caucasian subjects with juvenile-onset obesity,” Journal of Clinical Endocrinology and Metabolism, vol. 82, no. 12, pp. 4069–4074, 1997. [10] S. Costford, A. Gowing, and M. E. Harper, “Mitochondrial uncoupling as a target in the treatment of obesity,” Current Opinion in Clinical Nutrition and Metabolic Care, vol. 10, no. 6, pp. 671–678, 2007. [11] J. Nedergaard and B. Cannon, “The changed metabolic world with human brown adipose tissue: therapeutic visions,” Cell Metabolism, vol. 11, no. 4, pp. 268–272, 2010. [12] Y. Kontani, Y. Wang, K. Kimura et al., “UCP1 deficiency increases susceptibility to diet-induced obesity with age,” Aging Cell, vol. 4, no. 3, pp. 147–155, 2005. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 845148, 9 pages doi:10.1155/2011/845148

Review Article Genetics of Childhood Obesity

Jianhua Zhao1 and Struan F. A. Grant1, 2, 3

1 Division of Human Genetics, The Children’s Hospital of Philadelphia, Philadelphia, PA 19104, USA 2 Department of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA 3 Center for Applied Genomics, Abramson Research Center, The Children’s Hospital of Philadelphia Research Institute, 34th and Civic Center Boulevard, Philadelphia, PA 19104, USA

Correspondence should be addressed to Struan F. A. Grant, [email protected]

Received 30 November 2010; Accepted 6 April 2011

Academic Editor: Andrew P. Hills

Copyright © 2011 J. Zhao and S. F. A. Grant. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Obesity is a major health problem and an immense economic burden on the health care systems both in the United States and the rest of the world. The prevalence of obesity in children and adults in the United States has increased dramatically over the past decade. Besides environmental factors, genetic factors are known to play an important role in the pathogenesis of obesity. Genome- wide association studies (GWAS) have revealed strongly associated genomic variants associated with most common disorders; indeed there is general consensus on these ﬁndings from generally positive replication outcomes by independent groups. To date, there have been only a few GWAS-related reports for childhood obesity speciﬁcally, with studies primarily uncovering loci in the adult setting instead. It is clear that a number of loci previously reported from GWAS analyses of adult BMI and/or obesity also play a role in childhood obesity.

1. Definition and Epidemiology of between energy intake and expenditure. In modern times, Childhood Obesity this excess in adipose tissue fuel storage is considered a disease; however, a better viewpoint would be that obesity Obesity is a major health problem in modern societies, with a is a survival advantage that has gone astray that is, what is prevalence of up to 25% in Western societies and an increas- now considered a disease was probably advantageous when ing incidence in children [1]. Obesity, plus the associated food was less available and a high level of energy expenditure insulin resistance [2, 3], is also considered a contributor to through physical activity was a way of life [12]. the major causes of death in the United States and is an The true prevalence of childhood obesity is difficult to important risk factor for type 2 diabetes (T2D), cardiovascu- empirically quantify as there is currently no internationally lar diseases (CVD), hypertension, and other chronic diseases. accepted definition; however, in general terms, childhood Approximately 70% of obese adolescents grow up to obesity is considered to have reached epidemic levels in become obese adults [4–6]. The main direct adverse effects developed countries. of childhood obesity include orthopedic complications, sleep Approximately 25% of children in the US are overweight apnea, and psychosocial disorders [7, 8]. Obesity present in and approximately 11% are obese. In the 10-year period adolescence has been shown to be associated with increased between the National Health and Nutrition Examination overall mortality in adults [9]; overweight children followed Survey (NHANES) II (1976–1980) and NHANES III (1988– up for 40 [10]and55years[11]weremorelikelytohave 1991), the prevalence of overweight children in the USA had CVD and digestive diseases, and to die from any cause as increased by approximately forty percent [1]. Examination compared with those who were lean. of historical standards for defining overweight in children Obesity is a complex disease that involves interactions from many countries tells us that the distribution of BMI between environmental and genetic factors. Excess in adipose is becoming increasingly skewed [13]. The lower part of the tissue mass can be seen as a disruption in the balance distribution has shifted relatively little whereas the upper 2 Journal of Obesity part has widened substantially. This finding suggests that fat-soluble vitamin supplementation and monitoring [26, many children may be more susceptible (genetically or 27]. socially) to influence by the changing environment. Metformin, used to treat T2D, has been used in insulin- Although the definition of obesity and overweight has resistant children and adolescents who are overweight, but changed over time [14, 15], it can be defined as an excess long-term efficacy and safety are unknown [28]. Addition- of body fat. The definition of childhood obesity continues to ally, surgical approaches to treat severe adolescent obesity are be problematic due to the fact that almost all definitions use being undertaken by several centers [29]. some variant of BMI (body mass index). A range of other For rare genetic and metabolic disorders, pharmacolog- methods are available which allow for accurate estimates of ical treatment may be useful. For example, recombinant total body fat; however, none of these are widely available leptin is useful in hereditary leptin deficiency. Octreotide and/or are easily applicable to the clinical situation. Body may be useful in hypothalamic obesity [30]. weight is reasonably well correlated with body fat but is also highly correlated with height, and children of the 3. Evidence for a Genetic Component in Obesity same weight but different heights can have widely differing amounts of adiposity, but in adults BMI correlates more The rising prevalence of obesity can be partly explained by strongly with more specific measurements of body fat, that environmental changes over the last 30 years, in particular is, BMI is useful for depicting overweight in the population the unlimited supply of convenient, highly calorific foods but is an imperfect approximation of excess adiposity [16]. together with a sedentary lifestyle. Despite these changes, In addition, the relation between BMI and body fat in there is also strong evidence for a genetic component to the children varies widely with age and with pubertal matura- risk of obesity [31, 32]; indeed, obesity is now considered a tion. This in itself makes BMI definitions of overweight for classic example of a complex multifactorial disease resulting children more complex than definitions for adults, which use from the interplay between behavioral, environmental and a single cutoff value for all ages. Definitions of overweight genetic factors which may influence individual responses to that use BMI-for-age can be based on a number of different diet and physical activity. standards that all give slightly different results, and all are A genetic component for obesity is reflected in prevalence essentially statistical not functional definitions. However, differences between racial groups, from 5% or less in Cau- useful percentile charts relating BMI to age have now been casian and Asian populations to 50% or more among Pima published in several countries [17]. The Center for Disease Indians and South Sea Island populations [33]. In addition, Control and Prevention defined overweight as at or above the the familial occurrences of obesity have been long noted with 95th percentile of BMI for age and “at risk for overweight” the concordance for fat mass among MZ twins reported to be as between 85th to 95th percentile of BMI for age [18, 19]. 70–90%, higher than the 35–45% concordance in DZ twins; European researchers classified overweight as at or above as such, the estimated heritability of BMI ranges from 30 to 85th percentile and obesity as at or above 95th percentile 70% [34–36]. of BMI [20]. A recent report from the Institute of Medicine has specifically used the term “obesity” to characterize BMI 4. Previous Genetic Studies in Obesity and ≥ 95th percentile in children and adolescents [21]. By the Need for GWAS Approaches late adolescence, these percentiles approach those used for adult definitions; the 95th percentile is then approximately Genome-wide linkage scans in families with the common 30 kg/m2 [8]. These statistical percentile definitions are now form of childhood obesity have yielded several loci, but the general guidelines for clinicians and others [19]. genes in these loci have yet to be elucidated. A number of families with rare pleiotropic obesity syndromes have 2. Therapeutic Options been studied by linkage analysis where chromosomal loci for Prader-Willi syndrome [37], Alstrom’s¨ syndrome [38], Data supporting the use of pharmacological therapy for and Bardet-Biedl syndrome [39–41]havebeenmapped pediatric overweight are limited and inconclusive [22]. but the underlying molecular mechanisms have yet to be Sibutramine has been studied in a randomized controlled determined. trial of severe obesity [23]. It has been shown to be efficacious Recent studies of genetic syndromes of obesity in rodents as compared with behavior therapy alone, but it may be have provided insights in to the underlying mechanisms that associated with side effects including increases in heart rate may play a role in energy homoeostasis. In recent years, and blood pressure [24]; recent clinical trial studies have research has begun to identify human disorders of energy concluded that subjects with preexisting cardiovascular con- balance that arise from defects in these or related genes ditions who were receiving long-term sibutramine treatment [42]. These mutations have been shown to result in morbid had an increased risk of nonfatal myocardial infarction and obesity in children without the developmental features that nonfatal stroke [25]; indeed, it was recently dropped from commonly accompany recognized syndromes of childhood further development based on the results from such clinical obesity. trials. The severely obese ob/ob mouse strain [43] inherits its Orlistat is approved for use in adolescence but its efficacy early-onset obesity autosomal recessively and weighs approx- has not yet been tested extensively in young patients. Orlistat imately three-times more than normal mice by maturity. is associated with gastrointestinal side effects and requires Zhang et al. [44] cloned and characterized the ob gene Journal of Obesity 3 which is expressed primarily in white adipose tissue as the regulatory elements, such as enhancers and silencers, and secreted protein, leptin, a mutation of which renders these genetic variants that disrupt those elements could equally mice leptin deficient. Administration of recombinant leptin confer susceptibility to complex disease. is known to reverse the phenotypic abnormalities in these The human genome and International HapMap projects mice entirely [45–47] while there is no effect in another have enabled the development of unprecedented technol- strain of severely obese mice, db/db, who instead have been ogy and tools to investigate the genetic basis of complex characterized to have a mutation in the leptin receptor gene, disease. The HapMap project, a large-scale effort aimed at which is primarily expressed in a different site, namely, understanding human sequence variation, has yielded new the hypothalamus [48]. In human studies, serum leptin insights into human genetic diversity that is essential for concentrations are widely recognized as being positively the rigorous study design needed to maximize the likelihood correlated with obesity-related traits [49]. that a genetic association study will be successful [59–61]. The behavioral and neuroendocrine effects of leptin Genome-wide genotyping of over 500,000 SNPs can now be could potentially be mediated through its actions at hypotha- readily achieved in an efficient and highly accurate manner lamic leptin receptors. Proopiomelanocortin (POMC) is [62, 63]. Since much of human diversity is due to single produced by the hypothalamus, which is subsequently base pair variations together with variations in copy number cleaved by prohormone convertases to yield peptides [64] throughout the genome, current advances in single-base (including α melanocyte stimulating hormone, αMSH) that extension (SBE) biochemistry and hybridization/detection to play a role in feeding behavior. Forty percent of POMC synthetic oligonucleotides now make it possible to accurately neurons express mRNA for the long form of the leptin genotype and quantitate allelic copy number [63, 65]. receptor, and POMC expression is positively regulated by There is now a revolution occurring in SNP genotyp- leptin [50]. Work in rodents has demonstrated that αMSH ing technology, with high-throughput genotyping methods acts as a suppressor of feeding behavior; recently, mutations allowing large volumes of SNPs (105-106)tobegenotyped in POMC associated with severe and early-onset obesity have in large cohorts of patients and controls, therefore enabling been described in two unrelated German subjects [51]. A large-scale GWAS in complex diseases. Already with this single patient with severe early-onset obesity was reported to technology compelling evidence for genetic variants involved have compound heterozygote mutations in the prohormone in type 1 diabetes [66–68], type 2 diabetes [68–72], age- convertase 1 (PC1) gene, a key component in the proteolytic related macular degeneration [73], inflammatory bowel processing of POMC [52, 53]. disease [74, 75], heart disease [76, 77], and breast cancer [78] One form of melanocortin receptor (MC4R) is highly has been described. expressed in areas of the hypothalamus involved in feeding; mice with disruption of the MC4R gene are severely obese 6. Findings from First GWAS [54]. More recently in humans, mutations in the MC4R gene Analyses of Obesity have been associated with obesity [55–58]. The MC4R gene is the first locus at which mutations are associated with In the past four years, many genetic loci have been implicated dominantly inherited morbid human obesity thus making for BMI from the outcomes of GWAS, primarily in adults. it the commonest genetic cause of human obesity described Insulin-induced gene 2 (INSIG2) was the first locus to before the era of GWAS. be reported by this method to have a role in obesity [79]but replication attempts have yielded inconsistent outcomes [80– 5. Genome Wide Association Studies 84]. A common genetic variant with modest relative risk (RR = ∼1.2), rs7566605, near the INSIG2 gene has been described Overall, linkage analysis studies conducted to date have to be associated with both adult and childhood obesity achieved only limited success in identifying genetic deter- from a GWAS employing 100,000 SNPs [79]. This variant, minants of obesity due to various reasons, importantly present in 10% of individuals, was subsequently replicated including the generic problem that the linkage analysis in four separate cohorts in the same study, including approach is generally poor in identifying common genetic individuals who were Caucasian, African American, and variants that have modest effects [59, 60]. Comparably, a children; however, three subsequent technical comments to generic problem with the candidate gene association studies Science [80–82] disagreed with this observation. is their general reliance on a suspected disease-causing The identification of the second locus, the fat mass- and gene(s) whose identity derives from a particular biological obesity-associated gene (FTO)[85], which has been more hypothesis on the pathogenesis of obesity. Thus, since robustly observed by others [86–89], including by us [90]. the pathophysiological mechanisms underlying obesity are Interestingly, its role in obesity pathogenesis was actually generally unknown, continued use of the hypothesis-driven made indirectly as a consequence of a GWAS of T2D [68, 71], candidate gene association approach is likely to identify only but it became quite clear that its primary influence is on BMI a relatively small fraction of the genetic risk factors for the determination which then in turn impairs glycemic control disease. [85]. However, the mechanism by which the variant in FTO The GWAS approach serves the critical need for a more influences the risk of obesity is largely unknown. comprehensive and unbiased strategy to identify causal Studies from both FTO knockout and FTO overexpres- genes related to obesity. It is also well established that sion mouse model support the fact that FTO is directly in noncoding regions of the genome there are important involved in the regulation of energy intake and metabolism 4 Journal of Obesity in mice, where the lack of FTO expression leads to lean- The positive results for FTO and MC4R come as no ness while enhanced expression of FTO leads to obesity surprise as we previously reported their association with the [91, 92]. CDC-defined 95th percentile of BMI, that is, obesity, in our A French sequencing effort in Caucasians (primarily pediatric cohort, but limited to ages 2–18 years old [90, 106]. adults) has reported a set of exonic mutations in FTO; One of the more notable results is the positive association however, due to the lack of significant difference in the with INSIG2. This association with pediatric BMI, albeit at frequencies of these variants between lean and obese indi- just the nominal level, contributes to the ongoing debate on viduals, this study was largely negative [93]. the relative contribution of INSIG2 in BMI determination. However, these nine loci only explain 1.12% of the total 7. Meta-Analyses variation for BMI. In addition, testing pair-wise interactions between the fifteen significant SNPs, none of the interac- Subsequent larger studies have uncovered eleven additional tion effects were significant suggesting that these loci act genes [94–96], again primarily in adults, firstly melanocortin additively on pediatric BMI [99]. As such, we do observe 4receptor(MC4R) from a multicenter meta-analysis [94], a cumulative effect but not as striking as reported by the then the GIANT consortium revealed six more genes GIANT consortium in their adult cohorts [96]. (transmembrane protein 18 (TMEM18), potassium channel A number of studies have found that body mass index tetramerisation domain containing 15 (KCTD15), glucosa- (BMI) in early life influences the risk of developing type 2 mine-6-phosphate deaminase 2 (GNPDA2), SH2B adaptor diabetes (T2D) later in life. Indeed, the same variant in IDE- protein 1 (SH2B1), mitochondrial carrier 2 (MTCH2), and HHEX that increases the risk of developing the disease later neuronal growth regulator 1 (NEGR1)) [96], five of which in life turns out to be also associated with increased BMI in were confirmed in the GWAS reported from Iceland (but childhood [100]. not GNPDA2 due to an unavailable proxy SNP), who also uncovered and reported loci on 1q25, 3q27 and 12q13 [95] 9. Loci Specifically Identified in Childhood and verified association with the brain-derived neurotrophic Obesity GWAS Analyses factor (BDNF)gene[97]. The latest GIANT meta-analysis revealed multiple new Two new loci for body-weight regulation were identified in a loci associated with body mass index in a study involving joint analysis of GWAS data for early-onset extreme obesity, a total of 249,796 individuals [98]. A total of 32 loci that is, BMI ≥ 99th, in French and German study groups reached genome wide significance, which included ten [101], namely, SDCCAG8 and TNKS/MSRA (Table 1). In known loci associated with BMI, four known loci associ- the discovery step, association was examined in a combined ated with weight and/or waist-hip ratio, namely, SEC16B, French and German sample of 1,138 extremely obese chil- TFAP2B, FAIM2, NRXN3, and eighteen new BMI loci, dren and adolescents and 1120 normal or underweight con- namely, RBJ-ADCY3-POMC, GPRC5B-IQCK, MAP2K5- trols with screening of 2,339,392 genotyped or imputed SNPs LBXCOR1, QPCTL-GIPR, TNNI3K, SLC39A8, FLJ35779- and testing ultimately 1,596,878 SNPs. In the replication HMGCR, LRRN6C, TMEM160-ZC3H4, FANCL, CADM2, cohort, all SNPs with strong evidence for association were PRKD1, LRP1B, PTBP2, MTIF3-GTF3A, ZNF608, RPL27A- genotyped in independent samples of 1,181 obese children TUB,andNUDT3-HMGA1. Besides association to SNPs, a and adolescents and 1,960 normal or underweight controls correlated copy number variation (CNV), that is, a 21 kb and in up to 715 nuclear families with at least one extremely deletion, was identified 50 kb upstream of GPRC5B. This obese offspring. However the two loci were, at most, only study also leveraged a pediatric cohort to lend further marginally associated with adult BMI in the latest GIANT support for their findings. study [98], suggesting their influence may be limited to extreme obesity in children. 8. Testing Adult-Discovered Loci in Children The biochemistry employed in the current genome wide SNP arrays allows also for the accurate genotyping and quan- As described above, a number of genetic determinants of titation of allelic CNV genome-wide [62, 63, 65]. Neuro- adult BMI have already been established through GWAS. logical disorders have proven the most challenging complex One obvious question is how do these loci operate in disease to address using genome wide SNP approaches, pri- childhood with respect to the pathogenesis of obesity? We marily as a consequence of the need for strict, uniform phe- have an ongoing GWAS of BMI variation in children so we notyping across very large, multicenter cohorts. However, are in position to query these SNPs in our dataset of in they have led the way in the uncovering of CNVs in common excess of 6,000 children with measures of BMI [99]. To date disorders such as autism [107–110], attention-deficit hyper- nine such loci have yielded evidence of association to BMI activity disorder [111], and schizophrenia [112–114]. in childhood, of which variants at the FTO locus yielded Genomic copy number variations (CNVs) have been the strongest association. With a similar magnitude of strongly implicated in subjects with extreme obesity and association to FTO was TMEM18 followed by GNPDA2.The coexisting developmental delay (Table 1). Two groups in remaining loci with evidence for association were INSIG2, the UK plus collaborators independently reported deletions MC4R, NEGR1, 1q25, BDNF,andKCTD15 (Table 1). This on chromosome 16p11.2 to be present at much higher in is very much in line with the observations made with the extreme obese cases than normal and obese individuals pediatric cohort utilized by Willer et al. [96]. [102, 103]. These deletions, estimated to range in size from Journal of Obesity 5

Table 1: Childhood obesity loci that have been identiﬁed to date and the route through which they were implicated.

Category Loci Citations Adult BMI GWAS loci also associated with childhood FTO, TMEM18, GNPDA2, INSIG2, MC4R, NEGR1, 1q25, [99] BMI/obesity in independent studies BDNF, KCTD15 Adult 2 type diabetes GWAS loci also associated with HHEX-IDE [100] childhood BMI/obesity GWAS of extreme childhood obesity—novel loci SDCCAG8, TNKS-MSRA [101] SH2B1, EDIL3, S1PR5, FOXP2, TBCA, ABCB5, ZPLD1, CNV analyses of childhood obesity—novel loci [102–105] KIF2B, ARL15, EPHA6-UNQ6114, OR4P4-OR4S2-OR4C6

220 kb to 1.7 Mb, encompass several genes. Bochukova et al. some 11q11, harboring the OR4P4, OR4S2,andOR4C6 [102] pointed out that the SH2B1 gene is within the deleted genes using a similar approach [105](Table 1). Indeed, as region that is common to all five cases studied. SH2B1 may higher and higher resolution genome wide scans are carried be the culprit as its role in leptin and insulin signaling and out, one would expect further reports of such findings. energy homeostasis is well described [102], plus common SNPs near SH2B1 locus have already been associated with 10. Other Ethnicities BMI in GWAS reports [96, 102]. To complement these previous CNV studies on extreme Studying populations of different ancestry will also help obesity, we addressed CNVs in common childhood obesity us to globally identify and understand the genetic and by examining children in the upper 5th percentile of BMI environmental factors associated with estimates of obesity, as but excluding any subject greater than 3 standard deviations variants found in populations of both African and Caucasian from the mean to reduce severe cases in the cohort [104] ancestry may represent more universally important genes (Table 1). We performed a whole-genome CNV survey of and pathways for subsequent diagnosis, prevention, and our cohort of European American (EA) childhood obesity treatment of obesity and its complications. In addition, a cases (n =∼ 1, 000) and lean controls (n = 2, 500) who cohort of African ancestry in many instances can aid in were genotyped with 550,000 SNP markers. We identified 34 refining the anticipated association(s) made in with the putative CNVR loci (15 deletions and 19 duplications) that GWAS approach due to lower LD in this ethnicity, for were exclusive to EA cases; however, three of the deletions example, the association of T2D with TCF7L2 [116]hasbeen proved to be false positives during the validation process refined utilizing a West African patient cohort [117]. with quantitative PCR (qPCR). Only 17 of these CNVR loci To date, most obesity GWAS reports have come from were unique to our cohort that is, not reported in controls investigations of populations of European origin. This is by the Database of Genomic Variants. Positive findings were partly due to the relatively low haplotypic complexity of evaluated in an independent African American (AA) cohort Caucasian genomes and partly to get around admixture (n =∼ 1, 500) of childhood obesity cases and lean controls concerns. Indeed, like many of the other replication efforts, (n =∼1, 500). Surprisingly, eight of these loci, that is, almost FTO shows the strongest association with BMI in our large half, also replicated exclusively in AA cases (6 deletions and European American pediatric cohort [98]. However, the 2 duplications). Replicated deletion loci consisted of EDIL3, role of the FTO locus in influencing BMI and obesity S1PR5, FOXP2, TBCA, ABCB5, and ZPLD1 while replicated predisposition in populations of African ancestry has been duplication loci consisted of KIF2B and ARL15. We also previously less clear [88, 118], but consensus is emerging observed evidence for a deletion at the EPHA6-UNQ6114 from large cohort studies, both in adults [119] and in our locus when the AA cohort was investigated as a discovery set. own pediatric cohort [90], that SNP rs3751812 captures The majority of genes harboring at the loci uncovered the FTO association with the trait in both ethnicities; this in this study have not been implicated in obesity previously. finding is comparable to similar outcomes working with loci However, the most notable finding is with ARL15, which was in asthma [120]andT2D[117]. recently uncovered in a GWAS of adiponectin levels, with the same risk allele also being associated with a higher risk of 11. Conclusions CVD and T2D [115]. We also evaluated large rare deletions present in <1% While these recently discovered loci unveil several new of individuals and >500 kb in size as set previously [104] biomolecular pathways not previously associated with obe- and did not observe excess of large rare deletions genome- sity, it is important to note that these well-established genetic wide. This is not unexpected given that previous reports only associations with obesity explain very little of the genetic found significance when including developmental delay sub- risk for this pediatric phenotype, suggesting the existence jects but not when severe early-onset obesity was evaluated of additional loci whose number and effect size remain alone [102, 103]. unknown. More recently, a novel common copy number variation These findings have left the genetics community to pon- for early-onset extreme obesity was reported on chromo- der how we are going to finally uncover the full repertoire of 6 Journal of Obesity the genetic component of given traits in order to explain the [14] K. M. Flegal, M. D. Carroll, C. L. Ogden, and C. L. Johnson, “missing heritability” [121]. Thus, it is clear that in addition “Prevalence and trends in obesity among US adults, 1999- to larger and larger cohorts combined in to meta-analyses, 2000,” Journal of the American Medical Association, vol. 288, new whole genome sequencing technologies will be a large no. 14, pp. 1723–1727, 2002. part of the solution. With new advances in sequencing, one [15] R. J. Kuczmarski and K. M. Flegal, “Criteria for definition of would expect further variants to be characterized in this overweight in transition: background and recommendations condition so collectively they could build up to a meaningful for the United States,” The American Journal of Clinical contribution to the missing heritably for this trait. Nutrition, vol. 72, no. 5, pp. 1074–1081, 2000. Taken together, the unbiased genome wide approach to [16] S. R. Daniels, P. R. Khoury, and J. A. Morrison, “The utility assess the entire genome has revealed genes that underpin the of body mass index as a measure of body fatness in children and adolescents: differences by race and gender,” Pediatrics, pathogenesis of childhood obesity. Further functional studies vol. 99, no. 6, pp. 804–807, 1997. will be needed to fully characterize the function of the genes [17] T. J. Cole, J. V. Freeman, and M. A. Preece, “Body mass index at these loci in relation to childhood obesity. reference curves for the UK, 1990,” Archives of Disease in Childhood, vol. 73, no. 1, pp. 25–29, 1995. References [18] K. M. Flegal, R. Wei, and C. Ogden, “Weight-for-stature compared with body mass index-for-age growth charts for [1] R. P. Troiano and K. M. Flegal, “Overweight children and the United States from the Centers for Disease Control and adolescents: description, epidemiology, and demographics,” Prevention,” The American Journal of Clinical Nutrition, vol. Pediatrics, vol. 101, no. 3, part 2, pp. 497–504, 1998. 75, no. 4, pp. 761–766, 2002. [2] G. M. Reaven, “Banting Lecture 1988. Role of insulin resistance in human disease. 1988,” Nutrition,vol.13,no.1, [19] J. H. Himes and W. H. Dietz, “Guidelines for overweight pp. 1595–1607, 1997. in adolescent preventive services: recommendations from [3] R. A. DeFronzo and E. Ferrannini, “Insulin resistance: a an expert committee. The expert committee on clinical multifaceted syndrome responsible for NIDDM, obesity, guidelines for overweight in adolescent preventive services,” hypertension, dyslipidemia, and atherosclerotic cardiovascu- The American Journal of Clinical Nutrition,vol.59,no.2,pp. lar disease,” Diabetes Care, vol. 14, no. 3, pp. 173–194, 1991. 307–316, 1994. [4] T. A. Nicklas, T. Baranowski, K. W. Cullen, and G. Berenson, [20] C.-E. Flodmark, I. Lissau, L. A. Moreno, A. Pietrobelli, and “Eating patterns, dietary quality and obesity,” Journalofthe K. Widhalm, “New insights into the field of children and American College of Nutrition, vol. 20, no. 6, pp. 599–608, adolescents’ obesity: the European perspective,” International 2001. Journal of Obesity, vol. 28, no. 10, pp. 1189–1196, 2004. [5]R.C.Whitaker,J.A.Wright,M.S.Pepe,K.D.Seidel,and [21]J.P.Koplan,C.T.Liverman,andV.I.Kraak,“Preventing W. H. Dietz, “Predicting obesity in young adulthood from childhood obesity: health in the balance: executive sum- childhood and parental obesity,” The New England Journal of mary,” Journal of the American Dietetic Association, vol. 105, Medicine, vol. 337, no. 13, pp. 869–873, 1997. no. 1, pp. 131–138, 2005. [6]T.J.Parsons,C.Power,S.Logan,andC.D.Summerbell, [22] J. A. Yanovski, “Intensive therapies for pediatric obesity,” “Childhood predictors of adult obesity: a systematic review,” Pediatric Clinics of North America, vol. 48, no. 4, pp. 1041– International Journal of Obesity, vol. 23, supplement 8, pp. 1053, 2001. S1–S107, 1999. [23] R. I. Berkowitz, K. Fujioka, S. R. Daniels et al., “Effects of [7] W. H. Dietz, “Health consequences of obesity in youth: sibutramine treatment in obese adolescents: a randomized childhood predictors of adult disease,” Pediatrics, vol. 101, trial,” Annals of Internal Medicine, vol. 145, no. 2, pp. 81–90, no. 3, pp. 518–525, 1998. 2006. [8] S. R. Daniels, D. K. Arnett, R. H. Eckel et al., “Overweight [24] R. I. Berkowitz, T. A. Wadden, A. M. Tershakovec, and J. in children and adolescents: pathophysiology, consequences, L. Cronquist, “Behavior therapy and sibutramine for the prevention, and treatment,” Circulation, vol. 111, no. 15, pp. treatment of adolescent obesity: a randomized controlled 1999–2012, 2005. trial,” Journal of the American Medical Association, vol. 289, [9] A. Must, “Does overweight in childhood have an impact on no. 14, pp. 1805–1812, 2003. adult health?” Nutrition Reviews, vol. 61, no. 4, pp. 139–142, [25] W. P. T. James, I. D. Caterson, W. Coutinho et al., “Effect of 2003. sibutramine on cardiovascular outcomes in overweight and [10] H. O. Mossberg, “40-year follow-up of overweight children,” obese subjects,” The New England Journal of Medicine, vol. The Lancet, vol. 2, no. 8661, pp. 491–493, 1989. 363, no. 10, pp. 905–917, 2010. [11]A.Must,P.F.Jacques,G.E.Dallal,C.J.Bajema,andW.H. ffi Dietz, “Long-term morbidity and mortality of overweight [26] J. R. McDu e, K. A. Calis, G. I. Uwaifo et al., “Three-month adolescents—afollow-upoftheHarvardGrowthStudyof tolerability of orlistat in adolescents with obesity-related 1922 to 1935,” The New England Journal of Medicine, vol. 327, comorbid conditions,” Obesity Research, vol. 10, no. 7, pp. no. 19, pp. 1350–1355, 1992. 642–650, 2002. [12] R. H. Eckel, “Obesity: a disease or a physiologic adaptation [27] J. R. McDuffie, K. A. Calis, S. L. Booth, G. I. Uwaifo, and J. A. for survival?” in Obesity Mechanisms and Clinical Manage- Yanovski, “Effects of orlistat on fat-soluble vitamins in obese ment, R. H. Eckel, Ed., pp. 3–30, Lippincott Williams & adolescents,” Pharmacotherapy, vol. 22, no. 7, pp. 814–822, Wilkins, Philadelphia, Pa, USA, 2003. 2002. [13] K. M. Flegal and R. P. Troiano, “Changes in the distribution [28] M. Freemark and D. Bursey, “The effects of metformin on of body mass index of adults and children in the US body mass index and glucose tolerance in obese adolescents population,” International Journal of Obesity, vol. 24, no. 7, with fasting hyperinsulinemia and a family history of type 2 pp. 807–818, 2000. diabetes,” Pediatrics, vol. 107, no. 4, p. E55, 2001. Journal of Obesity 7

[29] T. H. Inge, V. Garcia, S. Daniels et al., “A multidisciplinary [47] M. A. Pelleymounter, M. J. Cullen, M. B. Baker et al., “Effects approach to the adolescent bariatric surgical patient,” Journal of the obese gene product on body weight regulation in of Pediatric Surgery, vol. 39, no. 3, pp. 442–447, 2004. ob/ob mice,” Science, vol. 269, no. 5223, pp. 540–543, 1995. [30] R. H. Lustig, P. S. Hinds, K. Ringwald-Smith et al., “Octre- [48] S. C. Chua Jr., W. K. Chung, X. S. Wu-Peng et al., “Pheno- otide therapy of pediatric hypothalamic obesity: a types of mouse diabetes and rat fatty due to mutations in double-blind, placebo-controlled trial,” Journal of Clinical the OB (leptin) receptor,” Science, vol. 271, no. 5251, pp. Endocrinology and Metabolism, vol. 88, no. 6, pp. 2586–2592, 994–996, 1996. 2003. [49] R. V. Considine, M. K. Sinha, M. L. Heiman et al., “Serum [31] J. M. Friedman, “Modern science versus the stigma of immunoreactive-leptin concentrations in normal-weight obesity,” Nature Medicine, vol. 10, no. 6, pp. 563–569, 2004. and obese humans,” The New England Journal of Medicine, [32] H. N. Lyon and J. N. Hirschhorn, “Genetics of common vol. 334, no. 5, pp. 292–295, 1996. forms of obesity: a brief overview,” The American Journal of [50] C. C. Cheung, D. K. Clifton, and R. A. Steiner, “Proopiome- Clinical Nutrition, vol. 82, no. 1, supplement, pp. 215S–217S, lanocortin neurons are direct targets for leptin in the 2005. hypothalamus,” Endocrinology, vol. 138, no. 10, pp. 4489– [33]W.C.Knowler,D.J.Pettit,M.F.Saad,andP.H.Bennett, 4492, 1997. “Diabetes mellitus in the pima indians: incidence, risk [51] H. Krude, H. Biebermann, W. Luck, R. Horn, G. Brabant, and factors and pathogenesis,” Diabetes/Metabolism Reviews, vol. A. Gruters,¨ “Severe early-onset obesity, adrenal insufficiency 6, no. 1, pp. 1–27, 1990. and red hair pigmentation caused by POMC mutations in [34] J. Hebebrand, S. Friedel, N. Schauble,¨ F. Geller, and A. humans,” Nature Genetics, vol. 19, no. 2, pp. 155–157, 1998. Hinney, “Perspectives: molecular genetic research in human [52] R. S. Jackson, J. W. M. Creemers, S. Ohagi et al., “Obesity and obesity,” Obesity Reviews, vol. 4, no. 3, pp. 139–146, 2003. impaired prohormone processing associated with mutations [35] I. S. Farooqi and S. O’Rahilly, “New advances in the genetics in the human prohormone convertase 1 gene,” Nature of early onset obesity,” International Journal of Obesity, vol. Genetics, vol. 16, no. 3, pp. 303–306, 1997. 29, no. 10, pp. 1149–1152, 2005. [53] S. O’Rahilly, H. Gray, P. J. Humphreys et al., “Brief report: [36] C. G. Bell, A. J. Walley, and P. Froguel, “The genetics of impaired processing of prohormones associated with human obesity,” Nature Reviews Genetics, vol. 6, no. 3, pp. abnormalities of glucose homeostasis and adrenal function,” 221–234, 2005. The New England Journal of Medicine, vol. 333, no. 21, pp. [37] I. Kondo, J. Hamabe, K. Yamamoto, and N. Niikawa, 1386–1390, 1995. “Exclusion mapping of the Cohen syndrome gene from the [54] D. Huszar, C. A. Lynch, V. Fairchild-Huntress et al., “Targeted Prader-Willi syndrome locus,” Clinical Genetics, vol. 38, no. disruption of the melanocortin-4 receptor results in obesity 6, pp. 422–426, 1990. in mice,” Cell, vol. 88, no. 1, pp. 131–141, 1997. [38] I. M. Russell-Eggitt, P. T. Clayton, R. Coffey, A. Kriss, D. S. I. [55] G. S. H. Yeo, I. S. Farooqi, S. Aminian, D. J. Halsall, R. G. Taylor, and J. F. N. Taylor, “Alstrom syndrome: report of 22 Stanhope, and S. O’Rahilly, “A frameshift mutation in MC4R cases and literature review,” Ophthalmology, vol. 105, no. 7, associated with dominantly inherited human obesity,” pp. 1274–1280, 1998. Nature Genetics, vol. 20, no. 2, pp. 111–112, 1998. [39] P. L. Beales, A. M. Warner, G. A. Hitman, R. Thakker, and [56]C.Vaisse,K.Clement,B.Guy-Grand,andP.Froguel,“A F. A. Flinter, “Bardet-Biedl syndrome: a molecular and frameshift mutation in human MC4R is associated with a phenotypic study of 18 families,” Journal of Medical Genetics, dominant form of obesity,” Nature Genetics, vol. 20, no. 2, vol. 34, no. 2, pp. 92–98, 1997. pp. 113–114, 1998. [40] E. A. Bruford, R. Riise, P. W. Teague et al., “Linkage mapping [57] W. Gu, Z. Tu, P. W. Kleyn et al., “Identification and in 29 Bardet-Biedl syndrome families confirms loci in chro- functional analysis of novel human melanocortin-4 receptor mosomal regions 11q13, 15q22.3-q23, and 16q21,” Genomics, variants,” Diabetes, vol. 48, no. 3, pp. 635–639, 1999. vol. 41, no. 1, pp. 93–99, 1997. [58] A. Hinney, A. Schmidt, K. Nottebom et al., “Several muta- [41] T. L. Young, L. Penney, M. O. Woods et al., “A fifth locus for tions in the melanocortin-4 receptor gene including a non- Bardet-Biedl syndrome maps to chromosome 2q31,” Ameri- sense and a frameshift mutation associated with dominantly can Journal of Human Genetics, vol. 64, no. 3, pp. 900–904, inherited obesity in humans,” Journal of Clinical Endocrinol- 1999. ogy and Metabolism, vol. 84, no. 4, pp. 1483–1486, 1999. [42] I. S. Farooqi and S. O’Rahilly, “Recent advances in the [59] J. N. Hirschhorn and M. J. Daly, “Genome-wide association genetics of severe childhood obesity,” Archives of Disease in studies for common diseases and complex traits,” Nature Childhood, vol. 83, no. 1, pp. 31–34, 2000. Reviews Genetics, vol. 6, no. 2, pp. 95–108, 2005. [43] A. M. Ingalls, M. M. Dickie, and G. D. Snell, “Obese, a new [60]C.S.Carlson,M.A.Eberle,L.Kruglyak,andD.A.Nickerson, mutation in the house mouse,” The Journal of heredity, vol. “Mapping complex disease loci in whole-genome association 41, no. 12, pp. 317–318, 1950. studies,” Nature, vol. 429, no. 6990, pp. 446–452, 2004. [44] Y. Zhang, R. Proenca, M. Maffei, M. Barone, L. Leopold, and [61] The International HapMap Consortium, “A haplotype map J. M. Friedman, “Positional cloning of the mouse obese gene of the human genome,” Nature, vol. 437, no. 7063, pp. anditshumanhomologue,”Nature, vol. 372, no. 6505, pp. 1299–1320, 2005. 425–432, 1994. [62] D. Reich, N. Patterson, P. L. De Jager et al., “A whole- [45] J. L. Halaas, K. S. Gajiwala, M. Maffei et al., “Weight-reducing genome admixture scan finds a candidate locus for multiple effects of the plasma protein encoded by the obese gene,” sclerosis susceptibility,” Nature Genetics, vol. 37, no. 10, pp. Science, vol. 269, no. 5223, pp. 543–546, 1995. 1113–1118, 2005. [46] L. A. Campfield, F. J. Smith, Y. Guisez, R. Devos, and P. Burn, [63]F.J.Steemers,W.Chang,G.Lee,D.L.Barker,R.Shen, “Recombinant mouse OB protein: evidence for a peripheral and K. L. Gunderson, “Whole-genome genotyping with the signal linking adiposity and central neural networks,” single-base extension assay,” Nature Methods, vol. 3, no. 1, Science, vol. 269, no. 5223, pp. 546–549, 1995. pp. 31–33, 2006. 8 Journal of Obesity

[64] R. Redon, S. Ishikawa, K. R. Fitch et al., “Global variation in [81]C.Dina,D.Meyre,C.Samsonetal.,“Commenton“A copy number in the human genome,” Nature, vol. 444, no. common genetic variant is associated with adult and 7118, pp. 444–454, 2006. childhood obesity”,” Science, vol. 315, no. 5809, p. 187, 2007. [65] K. L. Gunderson, F. J. Steemers, G. Lee, L. G. Mendoza, and [82] D. Rosskopf, A. Bornhorst, C. Rimmbach et al., “Comment M. S. Chee, “A genome-wide scalable SNP genotyping assay on “A common genetic variant is associated with adult and using microarray technology,” Nature Genetics, vol. 37, no. 5, childhood obesity”,” Science, vol. 315, no. 5809, p. 187, 2007. pp. 549–554, 2005. [83] H. N. Lyon, V. Emilsson, A. Hinney et al., “The association [66]H.Hakonarson,S.F.A.Grant,J.P.Bradfieldetal.,“A of a SNP upstream of INSIG2 with body mass index is genome-wide association study identifies KIAA0350 as reproduced in several but not all cohorts,” PLoS Genetics, atype1diabetesgene,”Nature, vol. 448, no. 7153, pp. vol. 3, no. 4, article e61, 2007. 591–594, 2007. [84] K. Hotta, M. Nakamura, Y. Nakata et al., “INSIG2 gene [67] J. A. Todd, N. M. Walker, J. D. Cooper et al., “Robust rs7566605 polymorphism is associated with severe obesity associations of four new chromosome regions from genome- in Japanese,” Journal of Human Genetics,vol.53,no.9,pp. wide analyses of type 1 diabetes,” Nature Genetics, vol. 39, 857–862, 2008. no. 7, pp. 857–864, 2007. [85]T.M.Frayling,N.J.Timpson,M.N.Weedonetal.,“A common variant in the FTO gene is associated with body [68] P. R. Burton, D. G. Clayton, L. R. Cardon et al., “Genome- mass index and predisposes to childhood and adult obesity,” wide association study of 14,000 cases of seven common Science, vol. 316, no. 5826, pp. 889–894, 2007. diseases and 3,000 shared controls,” Nature, vol. 447, no. [86] A. Hinney, T. T. Nguyen, A. Scherag et al., “Genome Wide 7145, pp. 661–678, 2007. Association (GWA) study for early onset extreme obesity [69] R. Sladek, G. Rocheleau, J. Rung et al., “A genome-wide supports the role of fat mass and obesity associated gene association study identifies novel risk loci for type 2 (FTO) variants,” PLoS One, vol. 2, no. 12, Article ID e1361, diabetes,” Nature, vol. 445, no. 7130, pp. 881–885, 2007. 2007. [70]R.Saxena,B.F.Voight,V.Lyssenkoetal.,“Genome-wide [87] C. Dina, D. Meyre, S. Gallina et al., “Variation in FTO association analysis identifies loci for type 2 diabetes and contributes to childhood obesity and severe adult obesity,” triglyceride levels,” Science, vol. 316, no. 5829, pp. 1331–1336, Nature Genetics, vol. 39, no. 6, pp. 724–726, 2007. 2007. [88] A. Scuteri, S. Sanna, W. M. Chen et al., “Genome-wide [71] E. Zeggini, M. N. Weedon, C. M. Lindgren et al., “Replication association scan shows genetic variants in the FTO gene are of genome-wide association signals in UK samples reveals associated with obesity-related traits,” PLoS Genetics, vol. 3, risk loci for type 2 diabetes,” Science, vol. 316, no. 5829, pp. no. 7, article e115, 2007. 1336–1341, 2007. [89] K. A. Fawcett and I. Barroso, “The genetics of obesity: FTO [72] L. J. Scott, K. L. Mohlke, L. L. Bonnycastle et al., “A genome- leads the way,” Trends in Genetics, vol. 26, no. 6, pp. 266–274, wide association study of type 2 diabetes in finns detects 2010. multiple susceptibility variants,” Science, vol. 316, no. 5829, [90] S. F. A. Grant, M. Li, J. P.Bradfield et al., “Association analysis pp. 1341–1345, 2007. of the FTO gene with obesity in children of Caucasian and [73] R. J. Klein, C. Zeiss, E. Y. Chew et al., “Complement factor African ancestry reveals a common tagging SNP,” PLoS One, H polymorphism in age-related macular degeneration,” vol. 3, no. 3, Article ID e1746, 2008. Science, vol. 308, no. 5720, pp. 385–389, 2005. [91] C. Church, L. Moir, F. McMurray et al., “Overexpression of [74] R. H. Duerr, K. D. Taylor, S. R. Brant et al., “A genome-wide Fto leads to increased food intake and results in obesity,” association study identifies IL23R as an inflammatory bowel Nature Genetics, vol. 42, no. 12, pp. 1086–1092, 2010. disease gene,” Science, vol. 314, no. 5804, pp. 1461–1463, [92] J. Fischer, L. Koch, C. Emmerling et al., “Inactivation of the 2006. Fto gene protects from obesity,” Nature, vol. 458, no. 7240, [75] M. Imielinski, R. N. Baldassano, A. Griffiths et al., “Common pp. 894–898, 2009. variants at five new loci associated with early-onset [93] D. Meyre, K. Proulx, H. Kawagoe-Takaki et al., “Prevalence inflammatory bowel disease,” Nature Genetics, vol. 41, no. of loss-of-function FTO mutations in lean and obese 12, pp. 1335–1340, 2009. individuals,” Diabetes, vol. 59, no. 1, pp. 311–318, 2010. [94] R. J. F. Loos, C. M. Lindgren, S. Li et al., “Common variants [76] A. Helgadottir, G. Thorleifsson, A. Manolescu et al., “A near MC4R are associated with fat mass, weight and risk of common variant on chromosome 9p21 affects the risk obesity,” Nature Genetics, vol. 40, no. 6, pp. 768–775, 2008. of myocardial infarction,” Science, vol. 316, no. 5830, pp. [95] G. Thorleifsson, G. B. Walters, D. F. Gudbjartsson et al., 1491–1493, 2007. “Genome-wide association yields new sequence variants at [77] R. McPherson, A. Pertsemlidis, N. Kavaslar et al., “A common seven loci that associate with measures of obesity,” Nature allele on chromosome 9 associated with coronary heart Genetics, vol. 41, no. 1, pp. 18–24, 2009. disease,” Science, vol. 316, no. 5830, pp. 1488–1491, 2007. [96]C.J.Willer,E.K.Speliotes,R.J.F.Loosetal.,“Sixnew [78] D. F. Easton, K. A. Pooley, A. M. Dunning et al., “Genome- loci associated with body mass index highlight a neuronal wide association study identifies novel breast cancer suscep- influence on body weight regulation,” Nature Genetics, vol. tibility loci,” Nature, vol. 447, no. 7148, pp. 1087–1093, 2007. 41, no. 1, pp. 25–34, 2009. [79] A. Herbert, N. P. Gerry, M. B. McQueen et al., “A common [97] J. Gunstad, P. Schofield, R. H. Paul et al., “BDNF Val66Met genetic variant is associated with adult and childhood polymorphism is associated with body mass index in healthy obesity,” Science, vol. 312, no. 5771, pp. 279–283, 2006. adults,” Neuropsychobiology, vol. 53, no. 3, pp. 153–156, 2006. [80] R. J. Loos, I. Barroso, S. O’rahilly, and N. J. Wareham, [98] E. K. Speliotes, C. J. Willer, S. I. Berndt et al., “Association “Comment on “A common genetic variant is associated with analyses of 249,796 individuals reveal 18 new loci associated adult and childhood obesity”,” Science, vol. 315, no. 5809, p. with body mass index,” Nature Genetics, vol. 42, no. 11, pp. 187, 2007. 937–948, 2010. Journal of Obesity 9

[99] J. Zhao, J. P. Bradfield, M. Li et al., “The role of obesity- influence adiponectin levels,” PLoS Genetics, vol. 5, no. 12, associated loci identified in genome-wide association studies Article ID e1000768, 2009. in the determination of pediatric BMI,” Obesity, vol. 17, no. [116] S. F. A. Grant, G. Thorleifsson, I. Reynisdottir et al., “Variant 12, pp. 2254–2257, 2009. of transcription factor 7-like 2 (TCF7L2) gene confers risk of [100] J. Zhao, J. P. Bradfield, H. Zhang et al., “Examination of all type 2 diabetes,” Nature Genetics, vol. 38, no. 3, pp. 320–323, type 2 diabetes GWAS loci reveals HHEX-IDE as a locus 2006. influencing pediatric BMI,” Diabetes,vol.59,no.3,pp. [117] A. Helgason, S. Palsson,´ G. Thorleifsson et al., “Refining 751–755, 2010. the impact of TCF7L2 gene variants on type 2 diabetes [101] A. Scherag, C. Dina, A. Hinney et al., “Two new loci for and adaptive evolution,” Nature Genetics,vol.39,no.2,pp. body-weight regulation identified in a joint analysis of 218–225, 2007. genome-wide association studies for early-onset extreme [118] A. Adeyemo, G. Chen, J. Zhou et al., “FTO genetic variation obesity in French and German study groups,” PLoS Genetics, and association with obesity in West Africans and African vol. 6, no. 4, Article ID e1000916, 2010. Americans,” Diabetes, vol. 59, no. 6, pp. 1549–1554, 2010. [102] E. G. Bochukova, N. I. Huang, J. Keogh et al., “Large, rare [119] M. T. Hassanein, H. N. Lyon, T. T. Nguyen et al., “Fine chromosomal deletions associated with severe early-onset mapping of the association with obesity at the FTO locus obesity,” Nature, vol. 463, no. 7281, pp. 666–670, 2010. in African-derived populations,” Human Molecular Genetics, [103] R. G. Walters, S. Jacquemont, A. Valsesia et al., “A new highly vol. 19, no. 14, pp. 2907–2916, 2010. penetrant form of obesity due to deletions on chromosome [120] P. M. A. Sleiman, J. Flory, M. Imielinski et al., “Variants of 16p11.2,” Nature, vol. 463, no. 7281, pp. 671–675, 2010. DENND1B associated with asthma in children,” The New [104] J. T. Glessner, J. P. Bradfield, K. Wang et al., “A genome-wide England Journal of Medicine, vol. 362, no. 1, pp. 36–44, 2010. study reveals copy number variants exclusive to childhood [121] T. A. Manolio, F. S. Collins, N. J. Cox et al., “Finding the obesity cases,” American Journal of Human Genetics, vol. 87, missing heritability of complex diseases,” Nature, vol. 461, no. 5, pp. 661–666, 2010. no. 7265, pp. 747–753, 2009. [105] I. Jarick, C. I. Vogel, S. Scherag et al., “Novel common copy number variation for early onset extreme obesity on chromosome 11q11 identified by a genome-wide analysis,” Human Molecular Genetics, vol. 20, no. 4, pp. 840–852, 2011. [106] S. F. A. Grant, J. P. Bradfield, H. Zhang et al., “Investigation of the locus near MC4R with childhood obesity in Americans of European and African ancestry,” Obesity,vol.17,no.7, pp. 1461–1465, 2009. [107] J. Sebat, B. Lakshmi, D. Malhotra et al., “Strong association of de novo copy number mutations with autism,” Science, vol. 316, no. 5823, pp. 445–449, 2007. [108] C. R. Marshall, A. Noor, J. B. Vincent et al., “Structural variation of chromosomes in autism spectrum disorder,” American Journal of Human Genetics, vol. 82, no. 2, pp. 477–488, 2008. [109] L. A. Weiss, Y. Shen, J. M. Korn et al., “Association between microdeletion and microduplication at 16p11.2 and autism,” The New England Journal of Medicine, vol. 358, no. 7, pp. 667–675, 2008. [110] J. T. Glessner, K. Wang, G. Cai et al., “Autism genome-wide copy number variation reveals ubiquitin and neuronal genes,” Nature, vol. 459, no. 7246, pp. 569–573, 2009. [111] J. Elia, X. Gai, H. M. Xie et al., “Rare structural variants found in attention-deficit hyperactivity disorder are preferentially associated with neurodevelopmental genes,” Molecular Psychiatry, vol. 15, no. 6, pp. 637–646, 2010. [112] H. Stefansson, D. Rujescu, S. Cichon et al., “Large recurrent microdeletions associated with schizophrenia,” Nature, vol. 455, no. 7210, pp. 232–236, 2008. [113] T. Walsh, J. M. McClellan, S. E. McCarthy et al., “Rare structural variants disrupt multiple genes in neurodevelopmental pathways in schizophrenia,” Science, vol. 320, no. 5875, pp. 539–543, 2008. [114] J. T. Glessner, M. P. Reilly, C. E. Kim et al., “Strong synaptic transmission impact by copy number variations in schizophrenia,” Proceedings of the National Academy of Sciences of the United States of America, vol. 107, no. 23, pp. 10584–10589, 2010. [115] J. B. Richards, D. Waterworth, S. O’Rahilly et al., “A genome- wide association study reveals variants in ARL15 that Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 497401, 31 pages doi:10.1155/2011/497401

Review Article Studies of Gene Variants Related to Inﬂammation, Oxidative Stress, Dyslipidemia, and Obesity: Implications for a Nutrigenetic Approach

Maira Ladeia R. Curti, Patrıcia´ Jacob, Maria Carolina Borges, Marcelo Macedo Rogero, and Sandra Roberta G. Ferreira

Department of Nutrition, School of Public Health, University of S˜ao Paulo, Avenida Dr. Arnaldo, 715, 01246-904, S˜ao Paulo, SP, Brazil

Correspondence should be addressed to Sandra Roberta G. Ferreira, [email protected]

Received 25 November 2010; Revised 15 February 2011; Accepted 14 March 2011

Academic Editor: Jordi Salas Salvado

Copyright © 2011 Maira Ladeia R. Curti et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Obesity is currently considered a serious public health issue due to its strong impact on health, economy, and quality of life. It is considered a chronic low-grade inflammation state and is directly involved in the genesis of metabolic disturbances, such as insulin resistance and dyslipidemia, which are well-known risk factors for cardiovascular disease. Furthermore, there is evidence that genetic variation that predisposes to inflammation and metabolic disturbances could interact with environmental factors, such as diet, modulating individual susceptibility to developing these conditions. This paper aims to review the possible interactions between diet and single-nucleotide polymorphisms (SNPs) in genes implicated on the inflammatory response, lipoprotein metabolism, and oxidative status. Therefore, the impact of genetic variants of the peroxisome proliferator-activated receptor- (PPAR-)gamma, tumor necrosis factor-(TNF-)alpha, interleukin (IL)-1, IL-6, apolipoprotein (Apo) A1, Apo A2, Apo A5, Apo E, glutathione peroxidases 1, 2, and 4, and selenoprotein P exposed to variations on diet composition is described.

1. Introduction of life, increase in costs of medical care, and for premature mortality [3, 7]. Obesity is a global epidemic [1] and its prevalence is growing Excessive energy intake combined to low-energy expen- worldwide [2]. In 2007, prevalence in North America was diture induces lipid accumulation in adipose tissue but also 26.4% for men and 24.8% for women [3]. In developing in liver, muscle, and other internal organs, predisposing to countries like Brazil, overweight (body mass index (BMI) the development of insulin resistance (IR) and metabolic between 25 and 30 kg/m2) reaches approximately 50% of disturbances [8]. In parallel to impairment of insulin action adults and obesity (BMI ≥ 30 kg/m2) occurs on 16.9% of in tissues and organs, also insulin secretion is aﬀected by women and 12.4% of men [4]. One billion people in the diet. It is known that high saturated fatty acids (SFAs) intake world overweight and 300 million are considered obese. provokes a deleterious lipid proﬁle and that the elevated Estimates for 2030 are that 2 billion people will be overweight plasma-free fatty acid concentration induces apoptosis of β and 1.12 billion obese around the world [5]. This scenario cells, contributing to dysfunctional insulin secretion [9, 10]. mayreducelifeexpectancyforthiscentury[6], since obesity This condition of high cardiometabolic risk has been incrim- is a major condition triggering several diseases. Particularly inated in the mortality of people living in developed and in accumulation of visceral fat elevates the risk of type 2 developing countries [11–13]. Statistics indicates precocity diabetes (T2DM), dyslipidemia, and hypertension, which of cardiovascular disease, already reaching 4% of Americans contribute for cardiovascular diseases, reduction in quality adolescents [14] being more prevalent among girls. 2 Journal of Obesity

The cluster of cardiovascular risk factors led to the peroxisome proliferator-activated receptor gamma (PPAR-γ) description of the metabolic syndrome (MS), in which IR coactivator-1 a/b activation and adiponectin production, represents a link between hemodynamic and metabolic dis- which decreases the oxidation of glucose and fatty acids; (d) turbances [15]. Visceral fat accumulation, commonly present recruitment of immune cells like macrophages, neutrophils, in MS, contributes to a proinflammatory and pro-oxidant and bone-marrow-derived dendritic cells to white adipose state and to deterioration of glucose and lipid metabolism. tissue and muscle [32, 33]. High amounts of free fatty acids and several adipocytokines Reducing consumption of foods rich in SFA and increas- released from visceral adipose tissue to portal and systemic ing consumption of whole grains, fruits, vegetables, lean circulation are implicated in the genesis of IR [7, 16, 17]. meats and poultry, fish, low-fat dairy products, and oils Tumor necrosis factor alpha (TNF-α) is a proinflammatory containing oleic acid are expected to reduce the incidence adipocytokine whose production is increased in obese of metabolic diseases [10, 33]. In fact, this dietary pattern, humans and its neutralization resulted in amelioration of which has similarities with the Mediterranean diet, has been IR [18, 19]. In vitro studies, in both rodents and humans, associated with better metabolic and inflammatory profiles showed that TNF-α stimulates adipocyte lipolysis [19, 20] in some clinical trials [34]. through downregulation of the expression of the perilipin, The expression of genes is highly dependent on, and a lipid droplet-associated protein which is thought to mod- regulated by, nutrients and dietary bioactive compounds ulate the access of hormone-sensitive lipase to the surface of found in food. A variety of dietary components can alter the fat droplet [20]. Several evidences indicated that obesity gene expression, and thus significantly influence health. At may generate IR via inflammation [21, 22]. The relationship the same time, the genetic makeup of an individual may between obesity and subclinical inflammation was firstly coordinate its response to diet [35]. Investigations of gene- described by Hotamisligil in 2003 [8], who demonstrated a environment interactions have identified genetic polymor- positive correlation between adipose mass and the expression phisms associated with individual susceptibility to obesity, of the gene coding for TNF-α. Further, the link between inflammation, dyslipidemia, and oxidative stress. In this obesity and inflammation was reinforced by the findings of context, unbalanced diets may shift the balance between increased concentrations of other inflammatory biomarkers healthy and diseased conditions, increasing the risk of these in obese individuals, like interleukin 6 (IL-6) and interleukin metabolic and immune disturbances, particularly in genetic- 8(IL-8)[21, 23] and acute phase proteins such as C reactive predisposed subjects (Figure 1). protein (CRP) [6]. The assessment of the interactions between nutrients, In addition, augmented release of TNF-α,nonesterified dietary bioactive compounds, and genotypes may pave the fatty acids and angiotensinogen, found in visceral obesity, way for more targeted prevention strategies, and thereby increases oxidative stress, which furthers contribute to IR a better success in the prevention and treatment [36]. In [24]. In this context, it has been reported that angiotensin- this context, there is increasing evidence that supports a ogen-II-induced reactive oxygen species (ROS) upregulation role for genotype-nutrient interactions in obesity and its affects several parts of the intracellular insulin signalling associated disorders [37, 38]. The present study aimed to pathways [25]. In vitro, ROS impairs insulin receptor sub- review the relation between diet and SNPs that impact on in- strate-1 (IRS-1) phosphorylation and IRS-1-induced phos- flammation (PPAR-γ, Tumor Necrosis Factor alpha (TNF- phatidylinositol 3-kinase (PI3-kinase) activation in cultured α), interleukin (IL)-1, IL-6), oxidative status (glutathione adipocytes, leading to the impaired translocation of glucose peroxidases (GPX) 1, 2 and 4, and selenoprotein P) and transporter 4 (GLUT-4) into the membrane, resulting in IR lipoprotein metabolism (apolipoprotein (Apo) A1, Apo A2, [26]. Apo A5, Apo E). Modern lifestyle—characterized by high food intake and low physical activity—has been considered the main determinant of increased adiposity in mankind. Not only 2. Peroxisome Proliferator: Activated the amount of calories contributes to the deleterious effects Receptor Gamma of obesity, since certain dietary patterns are related to cardiovascular risk or protection. Morbidity and mortality The PPARγ is a nuclear hormone receptor that serves as a of Mediterranean populations are shown to be lower than master regulator of adipocytes-specific genes contributing those observed in population exposed to the typical Western to adipocytes differentiation, susceptibility to obesity, and diet, rich in SFA [27]. Animal studies have helped to insulin sensitivity. Furthermore, this induces the expression understand the role of dietary fat in disturbances of lipid of genes that promote entrance of lipids in the cells, and glucose metabolism [28, 29]. SFA have shown to synthesis triglycerides, fatty acid oxidation, and extracellular stimulate intracellular pathways, which result in proinflam- metabolism of lipids (reverse cholesterol transport through matory gene expression and/or IR [9, 10, 30, 31]. The regulation cholesterol-efflux from macrophages) [39–42]. underlying mechanisms for deterioration of glucose and Several SNPs variants related to PPARγ gene have been lipid metabolism include (a) accumulation of diacylglycerol identified, like 161C/T, 1431C/T, and 162L/V, but the most and ceramide; (b) activation of nuclear factor-kB (NFkB), prevalent is the Pro12Ala (substitution of proline to alanine protein kinase C, and mitogen-activated protein kinases, at codon 12 of this gene, rs1801282), which is associated and subsequent induction of inflammatory genes in white withT2DM,obesity,andotherclinicaldisorders[43– adipose tissue, immune cells, and myotubes; (c) decreased 47]. However, the prevalence of this polymorphism varies Journal of Obesity 3

Dietary exposure (e.g., western diet)

Resistant Susceptible genotype genotype

Obesity Inﬂammation Oxidative stress Homeostasis Insulin resistance Dyslipidemia

Figure 1: The association of gene variants with diﬀerential responses to diet and its impact on the shift between homeostasis and metabolic/immune disturbances.

according to the ethnic background of the population [43– be associated with BMI in postmenopausal Polish women, 46, 48–54]. although the Ala allele seems to predispose to a less favorable The relationship of SNP in the PPARγ gene and obesity lipid profile in this population [63]. In Sweden, an inter- or MS is controversial. A recent research showed that vention study included lifestyle modifications metformin or the Pro12Ala SNP was associated with obesity in Iranian placebo. Metformin and lifestyle groups, Ala12 carriers had individuals, and the presence of the Ala allele predicted a greater weight loss [64]. higher BMI [55]. In European descent Brazilian men [56], In addition to the association of Ala allele with reduced andinWhitemenfromItaly[57], the Ala allele interacts with risk of T2DM [65], its presence resulted in fat oxidation gender and contributes to the susceptibility to obesity. Also increase and higher satiety suggesting benefits in food in children of Greek origin, adiposity through measures of intake control [66]. Concordantly, a study conducted in skinfold triceps and subscapular was shown to be influenced Spanish obese woman showed that Pro12Ala SNP resulted by Pro12Ala in a gender specific manner [58]. in increased fat oxidation [67]. Nevertheless, a study with childhood and adult obesity Nutrients and bioactive compounds of food are able in the French Caucasian population showed that the SNP to interfere with the genome by highly complex forms. Pro12Ala was not associated with obesity, but is confirmed Studies have shown that effect of Pro12Ala may be mediated a contribution in the genetic risk for T2DM especially by the dietary fat [61], particularly by the proportion in obese subjects, in which this allele worsens IR and of dietary polyunsaturated fatty acid (PUFA) and SFA. A increases fasting insulin levels [59]. Another study among population-based cohort study, including 592 nondiabetic native Javaneses did not show an association between the Caucasian, showed that fasting insulin concentration was Pro12Ala polymorphism and T2DM; a weak association with negatively associated with the dietary PUFA-to-SFA ratio and obesity was found [38]. In Poland obese individuals with a strong interaction was evident between this ratio and the 10-year history of T2DM, there was no association between Pro12Ala polymorphism for both, BMI and fasting insulin, Pro12Ala SNP and body mass changes observed during the suggesting that when the dietary ratio is low, the BMI in disease course, neither differences in IR and incidence and Ala carriers is greater than that in Pro homozygotes, but progression of T2DM complications [60]. when the dietary ratio is high, the opposite is seen [68]. In In another study, fasting and postprandial triglycerides a Spanish population similar results were found; the intake and insulin levels as well as homeostasis model assessment of monounsaturated fatty acid (MUFA) mainly oleic acid, of insulin resistance (HOMA-IR) were significantly lower in contributed to the variance of the HOMA-IR. These findings the Ala12Ala group than in the Pro12Pro group after the suggest the existence of an interaction between Pro12Ala mixed meal [61]. Others confirmed an association between and dietary MUFA, such that obese people with the Ala-12 the Ala allele and reduced incidence of obesity in prepubertal allele have higher HOMA-IR values, especially if their MUFA children and found strong associations between this SNP and intake is low [69]. Studying the P12Ala with Glu27Glu percentage body fat [62]. Pro12Ala SNP does not seem to polymorphism of beta 2-adrenergic receptor, Rosado et al., 4 Journal of Obesity

[67], showed that individuals carrying both polymorphisms suchasCRPinsome[101, 102] but not all studies [103]. It Pro12Ala and Gln27Glu alleles had high PUFA intake and is noteworthy that this relationship was conditioned by age, greater body weight loss. Similar results were shown in gender, and ethnicity [101, 102]. A meta-analysis aiming to study with children. Pro12Ala may be associated with higher investigate whether −308G/A TNF-α gene variant was asso- insulin sensitivity and higher long-chain PUFAs, particularly ciated with MS phenotype found that carriers of the TNF2 n-3, levels in plasma phospholipids of 140 Italian normolipi- allele had higher systolic blood pressure and fasting insulin, demic obese children [70]. This gene-nutrient interaction which appeared to be dependent of age group and ethnic emphasizes the difficulty of examining the effect of common background,however,therewasnosignificantassociation polymorphisms in the absence of data about environment of TNF-α genotypes with plasma fasting glucose and leptin, exposures. T2DM and hypertension [104]. Concerning cardiovascular Selected studies on the main interactions of polymor- events, another meta-analysis failed to find an association phisms with dietary factors are in Table 1. of −308G/A polymorphism with ischemic heart disease and ischemic stroke, except for Asians in which carrying −308 variant seemed to be protective [105]. 3. Tumor Necrosis Factor Alpha Some studies reported that the presence of −308G/A TNF-α polymorphism may affect the individuals’ response TNF-α is a pleiotropic cytokine with a central role in to supplementation of dietary components such as fish oil inflammation. Obesity leads to an increased production of and vitamin E [71, 72]. In this context, Grimble et al. [71] TNF-α, which is a crucial mediator of IR states and other reported that supplementing fish oil (6 g/day for 12 weeks) related comorbidities [85–87]. in healthy men decreased TNF-α production by lipopoly- TNF-α production in whole blood cell culture of healthy saccharides-stimulated peripheral blood mononuclear cells individuals shows a great, though stable, variation, making it only in individuals with high TNF-α producer phenotype, possible to identify high and low producer phenotypes in the irrespective of genotype. Nevertheless, the suppressive effect population [88, 89]. This fact points to a substantial genetic of fish oil supplementation on TNF-α production was greater contribution to TNF-α regulation [90, 91]. The TNF gene is among individuals in the highest tertile of presupplementa- located on human chromosome 6p21.3 in the class III region tion with the TNF1/TNF2 genotype compared to the wild- of the major histocompatibility complex [92]. type individuals. Eight SNPs were described within the TNF promoter, In a randomized, double-blind, placebo-controlled trial, at positions −1031T/C, −863C/A, −857C/T, −575G/A, elderly men and women received vitamin E supplementation −376G/A, −308G/A, −244G/A, and −238G/A [91, 92]. (182 mg α-tocopherol) for one year. The ex vivo production Regulating the transcription of TNF gene could potentially of interleukin 1 β (IL-β), IL-6, and TNF-α was determined blunt the deleterious effects of dysregulated TNF-α synthesis, in whole blood samples incubated with lipopolysaccharide as seen in obesity-induced inflammation. As such, genetic (LPS) (either 1.0 or 0.01 mg/L) for 24 h. Although there were variations in TNF gene regulatory region may affect TNF-α no overall differences with respect to treatment group for any transcription and expression, influencing the development of of the cytokines, when genotype was taken into account, par- conditions associated with excessive TNF-α production, such ticipants with the A/A and A/G genotype at TNF −308G/A, as rheumatoid arthritis, inflammatory bowel disease and IR treated with vitamin E, had lower TNF-α production than [93]. those with the A allele treated with placebo [72]. Wilson et al. [94] identified the polymorphism located There is also some evidence that carrying TNF2 allele at position −308 in the TNF promoter (rs1800629), which may influence diet response in obese individuals. Obese defined the TNF1 (−308G) and TNF2 (−308A) alleles. The nondiabetic individuals were randomly allocated to 2 types presence of TNF2 allele seems to be associated with a greater of energy-restricted diets for 2 months: a diet lower in fat (LF: risk of diseases, mainly infectious and autoimmune diseases 1500 kcal/day, 52% carbohydrates, 20% proteins, and 27% [95]. However, results concerning the possible functional fats) and a diet lower in carbohydrates (LC: 1507 kcal/day, role of −308 TNF polymorphism are conflicting [96]. Gene 38% carbohydrates, 26% proteins, and 36% fats). The results reporter assays have been employed to investigate whether of this study indicated that obese subjects carrying the TNF2 the TNF 2 allele could influence TNF-α gene transcription. allele responded differently to the diets since, contrary to Some studies show significant differences in transcriptional the wild genotype individuals, no improvement in plasma activity between TNF1 and TNF2 alleles [97–99], while glucose, insulin, triglyceride, total cholesterol, low density others could not confirm these findings [99, 100]. Some lipoprotein (LDL), and blood pressure was observed. Despite of the variables that might explain the mixed results of the fact that both genotype groups have lost fat mass these experiments are differences in cell type used for following the interventions, it is important to emphasize that transfection, cell origin (human or nonhuman), type of such decrease in fat mass was slightly smaller in −308G/A stimulus (lipopolysaccharide, phorbol myristate acetate, and individuals (LF diet: 5.0% versus 6.5%; LC diet: 4.9% versus TNF-α), length of the promoter sequence used and the 7.3%) [73]. presence or absence of the 3 untranslated regions of the Another study [74] reported significant interactions mRNA (3UTR). between dietary fat intake and TNFA −308G/A polymor- Nonobese healthy individuals carrying TNF2 allele phism affecting obesity risk in a sub-Saharan women showed an increase in circulating inflammatory biomarkers population. When the dietary fat intake was 30 (% energy), Journal of Obesity 5 0.30). Obese people with the Ala allele have higher = ect is observed. Strong interaction was evident between ff Subjects with the Ala allele had(OR a lower risk for T2DM PUFA/SFA ratio inand serum the fasting insulin diet concentration. and Pro12Ala for both BMI Fasting and postprandial serum TG,HOMA-IR insulin were significantly levels lower and in thethan Ala12Ala in group Pro12Pro group after the metabolic tolerance test. In Pro homozygous, VAT was reduced to airrespective similar of degree the level ofacid polyunsaturated/saturated ratio fatty (P : Swhere ratio) consuming in low the P diet. Ala12 :mass, allele S whereas carriers ratio those diets who tended consumed high totended gain P to VAT : lose In VAT. S metformin ratio and diets lifestyleAla12 groups, carriers had greater weightWhen loss. the amount of SFA inamount the of diet PUFA, the is BMI greater of thanthe Ala the amount carriers of is PUFA high. is When greatere than the SFA, the opposite HOMA-IR values, especially if their MUFA intake is low. Cross-sectional Metabolic tolerance test: oral glucose tolerance test and oral metabolic tolerance test (mixed meal containing 51,6 kJ% fat, 29,6 kJ% carbohydrates, 11,9 kJ% protein, with a total of 4406 kJ. Randomized clinical trial treatment with metformin, troglitazone, or lifestyle modification versus placebo for T2DM prevention in high-risk individuals. Cross-sectional ariants of genes involved in inflammation and oxidant status. Pro12Pro: 85,8 Pro12Ala: 13,4% Ala12Ala: 0,8% Pro: 85% Ala: 15% Ala frequencies: European Americans: 11,7% African Americans: 4% Hispanics: 9% American Indians: 19,6% Asian: 8,6% Pro/Pro: 79,1% Ala/Ala: 2% ] 64 ] 68 ] 69 ] 61 538 subjects from southern Spain [ Total of 3,356 individuals from Diabetes Prevention Program, USA [ 592 nondiabetic Caucasian from UK [ 708 men from Germany [ 1: Summary of studies evaluating interaction between diet and v Table Pro12Ala (rs1801282) Gene Variant Population [reference] Frequency Design Main findings PPAR- gamma 6 Journal of Obesity when α production in α ect of fish oil supplementation on ff 308A allele had no improvement in plasma − 308A compared to the wild-type individuals. − production was greater among individuals in the 036). For individuals with the GA+AA genotype, the α . = P The suppressive e TNF- highest tertile of presupplementation TNF- Participants with the A/A andvitamin A/G E, genotype, had treated lower TNF- with Contrary to the wild genotype individuals,carrying obese the subjects When the dietary fat intakeodds was of 30 being to 35% obese of withlower energy, the than the TNFA of GA+AA that genotype with was GG.dietary However, fat increasing was intake of associated with aincrease significantly in faster obesity rate risk of in womengenotype with compared the with TNFA GA+AA those with( the GG genotype glucose, insulin, TG, TC and LDL-cpressure levels following and the blood intervention. carrying LPS-stimulated whole blood cell culture thanthe those A with allele treated with placebo. obesity OR was 3.02 and 9.12of for 35 total and dietary 40 fat (%E), intakes respectively,indicating compared that with individuals 30%E, carrying A allelesignificantly were more responsive to an increaseintake in in dietary their fat risk of beingweight. obese compared with normal -tocopherol/day α Clinical trial: supplementation with 6 g/day of fish oil (1.8 g DHA+EPA) for 12 weeks. Randomized, double-blind, placebo-controlled trial: supplementation with 182 mg for one year. Randomized clinical trial: ingestion of energy-restricted diets (1,500 kcal/day) for 2 months. Cross-sectional 1: Continued. Table GG: 68,0% GA: 30,0% AA: 2,0% GG: 72,5% AG: 21,5% AA: 6,0% GG: 75,5% GA/AA: 24,5% GG: 70,0% GA+AA: 30,0% ] 73 ± 68) and ] = 203) [ n 71 = ] ] n 70) women aged 72 74 111) [ = = n n 45 y [ 110) [ ± = n 18 Elderly men and women ( Healthy men aged 28 8yrs( obese ( Normal weight ( Obese nondiabetic individuals ( 308A (rs1800629) − Gene Variant Population [reference] Frequency Design Main findings TNF Journal of Obesity 7 γ 10% 032) in < . = P 01). . = P 01) and VLDL-c . 01), and postglucose . = = P 043). P . 03), slightly lower HDL-c . = = P P 174G/C SNP gives higher ability to increase − 002), higher fasting ( 058) than the ones with C allele. . . = = P P After a 3-year intervention withCC a individuals Mediterranean-style were diet predicted toreduction have in the body greatest weight. Athad baseline, the these highest individuals body weight and BMI. Individuals with the G alelle presentedconcentration twice of the serum TG ( The C allele was more frequently observed ( Presence of fat oxidation after a high fat load ( ( load-free fatty acids ( ( individuals with successful weight maintenance ( weight regain). The presence of thetogether Ala with allele the of C PPAR- alleleprotection of (OR the 0.19; IL-6 strengthens this Clinical trial: each participant was placed in one of three diets: low-fat diet; Mediterranean diet supplemented with virgin olive oil; Mediterranean diet supplemented with nuts. Metabolic tolerance test: plasma free fatty acids suppression was evaluated in the fasting state and 120 minutes after an oral glucosee tolerance test. Clinical trial: volunteers were enrolled in a 10-week dietary intervention programme with a balanced low-energy diet, followed by dietitians. They were contacted again 1 year after the end of this period. Metabolic tolerance test:high a fat load containing 95% energy from fat. 1: Continued. Table CC: 34% GC+CC: 66% C allele: 0,39 C allele: 40,2% — ] 77 ]. 75 ]. ] 76 78 32 healthy Caucasian origin subjects from Spain [ 737 individuals with high cardiovascular risk from Spain [ 67 obese subjects from Spain [ 722 obese subjects from 8 European centers [ 174G/C (rs1800795) 174G/C (rs1800795) and PPAR Pro12Ala (rs1801282) − − Gene Variant Population [reference] Frequency Design Main findings IL-6 8 Journal of Obesity 05 511); P<. − -carotene levels, β 00 for Pro/Pro, . = P 001 for Pro/Pro and 31C/T polymorphism with 001 for Leu/Leu). In − P<. P<. 00 for Pro/Pro and Pro/Leu, . = 05). P 005), and those levels were higher in 05 for Leu/Leu), erythrocyte Se P<. P<. P<. 00001 for Pro/Leu, -carotene levels, but, in those with low concentration ( At baseline, 100% of the subjectsafter were the Se supplementation, deficient, there and wasplasma an Se improvement concentration in ( Pro/Leu, IL-1 risk genotype individuals receivingexperienced the a formulation greater reduction infrom IL-1B LPS-stimulated gene peripheral expression blood mononuclearand cells in plasma CRP levels, beingby the the risk presence genotype of defined the followinghomozygous 3 for genotypes: the (1) common allele (C) at IL1B ( The association of the IL-1B Leu/Leu subjects compared with thosegenotype with ( the wild-type addition, the Pro/Pro group showed adamage decrease in after DNA Brazil nut consumptionbaseline compared ( with The prevalence of metabolic syndrome for11222 haplotype was significantly higher than foramong haplotype those 21111 with low DHA+EPA membrane content. P<. hypertension was weak in womenβ with high serum for Leu/Leu), and GPx activity ( (2) carrying 2 copies of the(+4845); less or common (3) allele carrying (T) one at copyallele IL1A of at the IL1A less (+4845) common plus atcommon least allele one (T) copy at of IL1B the (+3954). less the TT genotype increased the prevalence of hypertension. gofSe/day)for8 μ weeks. Randomized trial: consumed one Brazililian nut (290 Randomized placebo-controlled trial: supplementation with a botanical extract (1,200 mg of rose hips extract, 165 mg of blackberry powder, 330 mg of blueberry powder, and 40 mg of grapevine extract per day) for 12 weeks. Cross-sectional 1: Continued. Table Pro/Pro: 48,7% Pro/Leu: 37,8% Leu/Leu: 13,5% CC: 22,4% CT: 51,8% TT: 25,8% Haplotype 11222: 25% Cross-sectional 1, 120) = ]— n 81 ] 82 ] 16 yrs from the 80 ± 200 men and 425 ] = n 79 37 morbidly obese women from Brazil [ Japaneses aged 39–70 yrs ( Caucasian men and women aged 49 Healthy adults [ women) [ GOLDN Study ( [ 511T − 31T (IL-1B) Pro198Leu Haplotype (IL-1B):6054G/A, 3966C/T, 231T/C, 2511G/A and 21473G/C − +3954T IL-1A: +4845T IL-1B: Gene Variant Population [reference] Frequency DesignGPx1 Main findings IL-1A IL-1B Journal of Obesity 9 ;VAT: α erences in ff : tumor necrosis factor- α entation SeP concentration rleukin; LDL-c: low-density lipoprotein cholesterol; LPS: sterol; TG: triglycerides; TNF- ted to saturated fatty acid ratio; PUFA: polyunsaturated fatty acid; Both lymphocyte GPx1 protein concentrations and plasma GPx3 activity increased significantly inTT CC participants. but After not Se withdrawal, theresignificant was fall a in both lymphocyte GPx4concentration protein and GPx4 activity in TT,participants; but females not had in higher CC concentrations thanmales. did was associated with gender and genotypeand at postsupplementation SNP concentration with 24731 SNP 25191. Both SNPs and gender wereGPx3 associated activity, with plasma, di and erythrocytereductase thioredoxin 1 concentrations and lymphocyte glutathione peroxidase 1 and 4 activities and concentrations. Plasma Se, SeP, and GPx3 levelssupplementation. increased Presupplem after selenium gSe/day). g Se/day) as sodium μ μ A clinical trial: 6-week selenium supplementation (100 A clinical trial: 6-week selenium supplementation (100 selenite. 1: Continued. Table CC: 55% TT: 45% Ala234Thr in Caucasian: GG: 46% GA: 47% AA: 7%25191G/A in Caucasian: GG: 47,8% GA: 45,6% AA: 6,4% ] 83 icosapentaenoic acid plus docosahexaenoic acid; GOLDN Study: Genetics of Lipid-Lowering Drugs and Diet Network (GOLDN) Study; e ]. 84 40 nonsmokers subjects from United Kingdom [ 121 nonsmokers subjects from European, Indian, and Chinese ethnic origins [ and UTR Ala234Thr (rs3877899) 25191G/A (rs7579) in 3 Gene VariantGPx4 718 T/C (rs713041) Population [reference]SeP Frequency gene Design Main findings GPx: glutathione peroxidase; HDL-c high-density lipoprotein cholesterol; HOMA-IR: homeostasis model assessment of insulin resistance; IL: inte lipopolysaccharide; MUFA: monounsaturated fat acid; OR: odds ratio; PPAR:Se: peroxisome receptor; proliferator-activated selenium; P SeP: : selenoprotein P; S ratio: SFA: polyunsatura saturated fatty acid; SNP: single nucleotide polymorphism; T2DM: Type 2 diabetes mellitus; TC: total chole visceral adipose tissue; VLDL-c: very low-density lipoprotein cholesterol. BMI: body mass index; CRP: C-reactive protein; DHA+EPA: 10 Journal of Obesity the odds of being obese with the TNF-α GA+AA genotype components on obesity-induced inflammation according to was only 12% of that with GG, and at 35 (% energy) it was individuals’ TNF-α haplotype. 33%. However, increasing intake of dietary fat (% energy) Selected studies on the main interactions of polymor- was associated with a significantly faster rate of increase in phisms with dietary factors are in Table 1. obesity risk in women with the TNF-α GA+AA genotype compared with those with the GG genotype (P = .036). For individuals with the GG genotype, the obesity OR was 4. Interleukin-6 1.12 and 1.26 for total dietary fat intakes of 35 and 40 (%E), respectively, compared with 30 (% energy). On the other IL-6 is secreted by a wide variety of cells such as endothelial hand, for individuals with the GA+AA genotype, the obesity cells, keratinocytes, osteoblasts, myocytes, adipocytes, β OR was 3.02 and 9.12 for total dietary fat intakes of 35 pancreatic cells, monocytes, macrophages, and a number and 40 (%E), respectively, compared with 30%E, indicating of other tissues, including a few tumors. This cytokine is that individuals carrying A allele were significantly more essential in reducing the inflammatory process by promot- responsive to an increase in dietary fat intake in their risk ing the synthesis of anti-inflammatory cytokines and by of being obese compared with normal weight. negatively regulating inflammatory targets. Therefore, this Inflammatory status is characterized by the production protein has been classified as both a pro- and anti-inflam- of a wide range of mediators that work in a complex network. matory;atcertainlevelitactsasadefensemechanism Thus, studying only one gene is not sufficient to understand but in chronic inflammation it has rather proinflammatory the complex nutrient-gene interactions that take place in properties [123–125] obesity-induced inflammation. In this context, a case-control In humans, higher circulating IL-6 levels have been study evaluated the additive effect of polymorphisms in associated with obesity and visceral fat deposition [126– TNF-α,lymphotoxin-α (LTA), and IL-6 genes in MS. The 128], increased risk of impaired glucose tolerance, T2DM risk genotype, characterized by the combination of the LTA [126, 129–131], and high blood pressure [132, 133]. IL- rs915654 A allele, the TNF GG, and the IL-6 rs1800797 6 is a central mediator of the acute-phase response and a GG genotype were associated with an increased risk of MS primary determinant of hepatic production of CRP [134, when compared to noncarriers (OR 2.10, 95% P<.005). 135]. Visceral adipose tissue secretes about two to three times Interestingly, such association was affected by plasma fatty more IL-6 than subcutaneous tissue, secreting also other acid profile. Among carriers of the 3 risk genotypes with the molecules that stimulate further IL-6 expression. lowest 50th percentile of PUFA/SFA, MS risk was increased The associations between common variations in the IL-6 more than 4-fold compared with noncarriers (OR 4.40, P< gene and obesity have been examined in many studies [136– .005). Similarly, a low PUFA/SFA exacerbated the risk of fast- 141]. The human gene of IL-6 is located on the short arm ing hyperglycemia, high systolic blood pressure, abdominal of chromosome 7 (7p21). In the last years many different obesity, and high plasma complement component 3 (C3) SNPs in the promoter region of this gene were reported, levels in these individuals. Also, MS risk was more than 5- like −572G/C, −373A(n)T(n), −597G/A (rs10242595), and fold higher in the risk genotype carriers with the highest SFA −174G/C (rs1800795), the latter being the most prevalent levels compared with noncarriers (OR 5.20, P<.005) [106]. and of greatest biological importance [142]. Another TNF-α genetic variant that has received atten- In a population-based cross-sectional study in Sweden, tion in the context of MS and chronic diseases is the TNF SNP −597G/A (minor allele frequency = 29%) 3 of the IL6 238G/A polymorphism (rs 361525), which is located within gene was negatively associated with the primary outcome a TNF gene repressor site whose effect on transcription total body fat mass (BFM) (effect size −0.11 SD per A activity still remains unknown [92]. The association between allele, P = .02). When results from three cohorts were TNF 238G/A polymorphism and T2DM in case-control combined (n = 8, 927 Caucasians), allele A showed a studies has generated conflicting results. A recent meta- negative association with total BFM (effect size −0.05 SD per analysis indicated that the TNF 238G/A polymorphism did A allele, P<.0002). Furthermore, the allele A was associated not increase the risk of T2DM. with lower BMI (effect size −0.03, P<.001) and smaller It has been suggested that there is linkage disequilibrium regional BFM [143]. between the −238, −308, and −376 (rs1800750) polymor- As far as the SNP −174G/C (rs1800795) is concerned, phisms. The −376A allele preferentially segregates with the three genotypes G/G, G/C, and C/C seem to affect the IL- −308G and the −238A alleles in most Africans and in 6 transcription (116,119). Genotype G/C is more frequently Europeans [92, 122]. This fact has allowed the study of reported by most studies [142]. In a population-based study haplotypes, groups of SNPs that are inherited together in including 9,960 Americans, it was shown frequencies of this blocks. Using haplotype approaches to identify functional SNP of 0.57 in Caucasians and 0.93 in Afro-Americans [144]. loci is interesting since multiple polymorphic loci may have The presence of C allele has been related with a higher combined effects on expression [97].InthecaseofTNF-α risk of obesity, high BMI, high waist circumference (WC), gene, the haplotypes −238G, −308A, and −376G are associ- high leptin levels, and high BFM [145, 146]. Both studies ated with higher TNF-α production, while −238A, −308G, with a Swedish and Canadian population reported associa- and −376A are associated with lower TNF-α production, tion of the C allele with obesity indices [136]. Carriers of which may depend on the racial group [92]. Unfortunately, the CC genotype had lower energy expenditure and insulin at present, no study investigated the effectofdietorits sensitivity, hence implying a causative role in IR and obesity Journal of Obesity 11

[77, 147]. The C allele was associated with higher BMI in in a high cardiovascular risk population, CC individuals T2DM individuals, but not amongst healthy subjects [134]. with the −174G/C polymorphism were predicted to have Studies relate G allele with an increase production of IL-6 the greatest reduction in body weight. At baseline these [142]. The GG genotype in Spanish Caucasian was associated individuals had the highest body weight and BMI [76]. with T2DM [148]. In a Spanish population, the G allele Therefore, the body of evidence suggests that variation was associated with hyperglycaemia with a reduction in in IL-6 gene is associated with metabolic and cytokine mod- insulin sensitivity [77] and with lipid profile disturbance ulation, and subsequently may play an important role in [149]. Contradicting others findings, a recent study with 228 impaired glucose and lipid homeostasis, cardiometabolic risk patients with T2DM and 300 healthy controls in Tunisia, the and seems to influence how the individual responds to fat SNP −174G/C was not associated with T2DM or risk for intake. Results are still controversial and further studies on overweight (P = .86). Bonferroni correction showed that this issue are needed. the association of the SNP with T2DM susceptibility was Selected studies on the main interactions of polymor- restricted to overweight individuals and may be likely to be a phisms with dietary factors are in Table 1. random result [150]. On the other hand, it was described that the individuals with GG genotype lost weight significantly β after aerobic exercises training. This effectwasnotobserved 5. Interleukin-1 in heterozygous neither the homozygous CC individuals who Together with TNF-α and IL-6, IL-1β also plays a central role did not reduced the fat mass and insulin levels after the in the regulation of the immune responses and inflammatory physical activity [146]. In addition we observed a higher process. IL-1β regulates the production of a variety of incidence of G allele in subjects with normal weight (BMI < inflammatory mediators, such as IL-6, intercellular adhesion 25 kg/m2)[136]. molecule (ICAM)-1, and E-selectin [156, 157], and it has Whereas several studies indicate that C allele is associated been associated with metabolic disturbances present in obese with higher plasma levels of IL-6, particularly in inflamma- subjects such as dyslipidemia and IR [158, 159]. tory situations [151, 152], in study including obese men, this SNP was associated with variation of plasma CRP concentra- The IL-1 family of genes comprises 9 genes, including the tion after weight loss [153]. A study with 504 Spanish ado- IL-1A, IL-1B, and IL-1 receptor antagonist gene (IL-1RN), lescents shows no differences between genotypes observed which are all located on the long arm of human chromosome α β in anthropometric values, body composition measurements, 2. Of the gene products, IL-1 and IL-1 are agonists, and plasma markers concentration. Physical activity level dif- whereas IL-1 receptor antagonist (IL-1Ra) is a competitive fers between genotypes with individuals carrying the C allele antagonist of the IL-1 receptor and is, therefore, a negative polymorphism being significantly more active than GG ones. regulator of inflammation [160, 161]. The association between body fat mass and plasma glucose Current data indicate that some SNPs in IL-1B gene was influenced by this SNP. Those carrying the C allele of the are associated with increased levels of IL-1β and other mutation seem to have higher values of lipoproteins and CRP inflammatory markers such as CRP,suggesting a relationship as their percentage of BFM increases [154]. Nevertheless, between IL-1 gene variants and systemic inflammation. the C allele was more frequently observed (P = .032) In patients with established coronary heart disease, IL-1B in individuals with successful weight maintenance (<10% (+3954)T allele correlated with increased CRP levels, even weight regain). Moreover, the presence of the Ala allele of the after adjusting for smoking status, BMI, total cholesterol, PPAR-γ together with the C allele of the IL-6 strengthens this and the presence of DM [162]. In a larger study performed protection (OR 0.19; P = .043), suggesting a synergetic effect in 454 subjects undergoing coronary angiography, individ- of both variants on weight maintenance after a diet to lose uals homozygous for IL-1B (+3954)T allele had 2- to 3- weight [78]. fold higher CRP levels than wild-type individuals, which Fernandez–Real´ et al. [77],studyingthesameSNP, remained significant after adjusting for gender, smoking, and observed in healthy individuals that those with the G alelle age [163]. Another study reported that the presence of −511 presented concentration two times bigger of triglycerides T (rs1143627) decreased IL-1β production in LPS-stimulated than the ones with C allele, suggesting that the associated mononuclear cells and that homozygous subjects for T to a bigger secretion of IL-6 can predispose disturbs of genotype showed a decreased risk of myocardial infarction lipid metabolism. Berthier et al. reported that GG allele (OR 0.36) and stroke (OR 0.32) [164]. was more common among lean men, and those with lower Given the role of IL-1RA in modulating the biological plasma concentrations of insulin and glucose following an activity of IL-1, polymorphic variants in IL-1RN gene may oral glucose tolerance test [155]. affect IL-1RA production, influencing the inflammatory Recent findings show that individuals with certain response. In fact, there is evidence that minor alleles polymorphism can respond differently to diet than those of IL1RN 1018 (rs4251961) and 13888 (rs2232354) are without polymorphism. The association between variation associated with increased levels of inflammatory biomarkers, in fat oxidation rates among obese subjects and genotype was such as IL-1β,CRP,fibrinogen,IL-6,interferon-γ,and studied for 42 common SNPs. Subjects with GC phenotype α-2 macroglobulin (a hepatic acute phase protein). Also, of SNP −174 had higher ability to increase fat oxidation after these polymorphisms are associated with lower IL-1RA a high fat load (P = .01) [75]. In a Spanish study, after a 3- concentration and ex vivo cellular production in response to year intervention with a Mediterranean-style diet conducted an inflammatory stimulus [79, 165]. 12 Journal of Obesity

In cross-sectional study with a predominantly Caucasian in IL-1 gene may affectdietresponseinsomeproinflamma- population, adjusted odds ratio for MS was significantly tory conditions. greater in subjects carrying 6054G allele, while it was Selected studies on the main interactions of polymor- significantly lower in individuals carrying −511G and phisms with dietary factors are in Table 1. −1473G IL-1β allele. This study also provided information on the interaction between genotype and membrane fatty acid content, a surrogate of habitual dietary fatty 6. Apolipoprotein A1 acid intake, in determining MS risk. Among individuals Apo A1 is the major protein component (roughly 70%) in the lowest 50th percentile of total PUFA or total of high-density lipoproteins (HDL), which is synthesized n-3 PUFA—eicosapentaenoic acid + docosahexaenoic acid primarily in the liver and small intestine, and acts in (EPA+DHA)—carriers of 6054G had a higher prevalence the transport of cholesterol from the peripheral tissues to of MS when compared to wild-type individuals; such the liver. Apo A1 plays a crucial role in the regulation association was not found in the highest 50th percentile of of reverse cholesterol transport [167–169]. Upon secretion total PUFA or total EPA+DHA membrane content, which into plasma, Apo A1 is lapidated through ATP-binding suggests that EPA+DHA may be helpful in preventing MS cassette transporter-(ABC-)A1 that mediated efflux of free in 6054G genetic predisposed individuals. The authors also cholesterol and phospholipids, resulting in the formation of examined a haplotype consisting of 5 SNPs: 6054G/A, nascent disc shaped HDL particles [170]. 3966C/T, 231T/C, 2511G/A, and 21473G/C and found that More than a dozen functionally significant mutations of individuals with haplotype 11222 had higher prevalence of the Apo A1 gene have been described, including gene disrup- MS than haplotype 21111. Among those with low DHA+EPA tions, nonsense mutations, frameshifts, missense mutations, membrane content, there was significant haplotype asso- chromosomal aberrations or deletions, and inversion of ciation with MS, being OR of MS for haplotype 11222 the APOA1/APOA3/APOA4 gene cluster; these are typically significantly higher than for haplotype 21111. However, there associated with decreased plasma HDL concentration [171– was no significant association among individuals with high 174]. membrane DHA+EPA content [166]. Another study found an association between the preva- Overexpression of the human Apo A1 gene in mice lence of hypertension and the presence of −31C allele, which increased plasma HDL concentration and protected the ani- could be modified by plasma β-carotene levels. Male CC mals from the development of high-fat diet-(HFD-)induced carriers with high serum β-carotene levels had a significantly atherosclerosis [175]. Conversely, Apo A1 knockout mice lower OR (OR 0.25, P<.05) for hypertension relative to exhibited decreased plasma HDL concentration and devel- those with low serum β-carotene levels. The association of oped atherosclerotic lesions [176]. the IL-1B −31C/T polymorphism with hypertension was Apo A1 missense mutations are a type of nonsynony- weak in women with high β-carotene circulating levels, but, mous mutation that causes a single nucleotide change, in those with low β-carotene levels, the TT genotype clearly resulting in a codon that codes for a amino acid, which increased the prevalence of hypertension (OR 2.47, P<.05). turns the protein without function. This kind of structural The results were not adjusted to some known risk factors for variationhavealsobeenfoundand,althoughthereare hypertension as smoking/drinking habits, menopause, and exceptions [177], most of these do not change plasma diet [80]. HDL concentration. However, a well-known variant of the A randomized placebo-controlled trial was carried out Apo A1, called A-I Milano (apoA-IM), is associated with in healthy adults to evaluate the effect of 12-week sup- reducedHDLandLDLlevels[178, 179]. This polymorphism plementation with a formulated botanical extract rich in is a variant form of Apo A1 that contains a cysteine dehydroascorbate, anthocyanins, and all-trans-resveratrol replaced by arginine at amino acid 173 (R173C) [180]. (1200 mg/d of rose hips extract, 165 mg/d of blackberry Moreover, Calabresi et al. [181] showed that carriers of the powder, 330 mg/d of blueberry powder, and 40 mg/d of apoA-IM mutation have an increased postprandial lipemic grapevine extract) on inflammation in individuals with response. Another polymorphism called PstI (rs12721026) genetic variations that predispose to overexpression of IL- is located in the 30 flanking region of Apo A1 gene and 1β. Individuals were stratified by IL-1 genotype prior to has been associated with the development of cardiovascular randomization. They were classified as IL-1 positive geno- disease and decreased plasma HDL concentration [182, type if they had any of the following 3 genotypes: (1) 183]. homozygous for the common allele (C) at IL1B (−511); (2) A common G to A substitution in the promoter area carrying 2 copies of the less common allele (T) at IL1A at position −76 bp (−76G/A) of the Apo A1 gene has been (+4845); or (3) carrying one copy of the less common extensively studied and some researchers have observed that allele at IL1A (+4845) plus at least one copy of the less carriers of this polymorphism presented an increase in the common allele (T) at IL1B (+3954). IL-1 positive genotype promoter activity in vitro and plasma HDL concentration individuals receiving the formulation experienced a greater [184, 185]. Also, Marın´ et al. [186] observed that carriers of reduction in IL-1B gene expression and in plasma CRP levels this mutation have a greater postprandial increase in large [81]. triglyceride-rich lipoproteins and a smaller decline in LDL Current data is clearly insufficient to draw definite con- and Apo B, after the ingestion of a dietary fat load than clusions; however, it seems likely that some polymorphisms carriers of the G/G genotype. Journal of Obesity 13

The minor allele of MspI polymorphism, consisting of The effectiveness of a dietary portfolio consisting of soy a G to A transition at position −75 bp (rs799837), has protein (25 g/day) and soluble fiber mix (Fibregum—100% been associated in a number of studies with higher Apo made of gum acacia, which has 90% of soluble dietary fiber— A1 and HDL plasma levels [187], although others were 15 g/day), integrated in a low SFA diet, on blood lipids and unable to confirm this relation [188]. Moreover, such pol- the association between this diet and the Apo E, Apo A1, and ymorphism may contribute to variability in postprandial ABCG5/8 polymorphisms were investigated in hyperlipi- lipid metabolism and in the lipoprotein response to dietary demic individuals [192]. Significant decreases in plasma total changes in healthy subjects [107]. cholesterol and TG concentrations was observed; 51% of the Some dietary factors have been extensively studied and individuals had >20% reduction in plasma total cholesterol related to the mutation of Apo A1 gene. The total dietary fat concentration, and 77% had a reduction >20% in plasma and the type of fat are the main cited interactions resulting TG concentration. Among hypercholesterolemic individuals, in serum lipids alterations. Gomez et al. [107]investigated 14% had the ABCG8 (+52G/C) polymorphism, 65% had the whether the presence of the −76G/A SNP in the Apo A1 ABCG5 (+1950 C/G and G/G) polymorphism, 53.5% the gene interacts with diet to determine changes in LDL particle Apo A1 (−75G/A and A/A) polymorphism, and 23.3% had size and their susceptibility to oxidative modifications. In a the Apo E (ε4) polymorphism. The presence of ABCG5/8 second step, they examined these effects by analyzing the has been associated with reduction of absorption of plant contribution of the combination of the Apo A1 −76G/A sterols and cholesterol from the diet by effluxing these and Apo A4 Thr347Ser SNP. Each of 97 healthy volunteers sterols from the enterocyte back into the intestinal lumen consumed 3 types of diet for 4 weeks: a high-saturated fat and by facilitating efficient secretion of plant sterols and diet (38% fat, being 20% SFA), 12% monounsaturated fatty cholesterol from hepatocytes into the bile [193]. However, acids (MUFA), and 6% polyunsaturated (PUFA) followed independently of genotype, the combination of cholesterol- by a low-fat and high-carbohydrate (CHO) diets (30% fat, lowering foods in a low-saturated fat diet improved lipid 55% CHO) or a MUFA diet (38% fat, 22% MUFA). After profile. Another interesting study examined the effect of the consuming the CHO diet, there was a significant decrease in dietary fat saturation when −75G/A mutation was present. LDL size with respect to high-fat diets in GG homozygotes Fifty men and women were first fed with a saturated fat for the Apo A1 −76G/A SNP. However, LDL size did not diet (17% SFA) for 28 days, followed by a MUFA diet (22% differ in GA carriers among participants consuming the 3 MUFA) for 35 days and a PUFA diet (13% PUFA) for 35 days. diets. Carriers of the A allele for this polymorphism had The allele frequency for the A allele was 0.13 and individuals smaller LDL size as well as increased susceptibility to oxida- who carry this allele had higher cholesterol, LDL, and TG tion after the SFA diet than the GG homozygous. levels than those with G/G. In women with the A allele, the Ordovas et al. [108] examined whether dietary fat mod- PUFA diet was compared to the saturated-fat-diet-induced ulates the association between the −75G/A of Apo A1 poly- significant decreases in total and LDL concentrations. These morphism and plasma HDL concentration. Participants (755 results suggested that the G/A polymorphism appears to have men and 822 women) of the Framingham Offspring Study asmallbutsignificanteffect on plasma LDL responsiveness were divided into low (<4% of energy), medium (4–8% of to changes in dietary fat saturation specially in women [109]. energy), and high (>8% of energy) PUFA intake. Women Selected studies on the main interactions of Apo poly- who were carriers of the A allele submitted to high PUFA morphisms with dietary factors are in Table 2. intake increased plasma HDL concentration significantly, but in homozygous women for the G allele an inverse relation was found. This meant that when PUFA intake provided 7. Apolipoprotein A2 <4% of energy, women who were homozygous for the G allele had almost 14% higher plasma HDL concentration Synthesized in the liver, apolipoprotein A2 (Apo A2) is the than A allele carriers and when PUFA intake provided >8% second most abundant protein of HDL particles and can be of energy, plasma HDL concentration in carriers of the found with the Apo A1 in subfractions of HDL. However, A allele were 13% higher than those of G/G individuals. its function remains largely unknown [111]. Although some Thereby, a significant interaction in terms of plasma HDL initial studies reported an inverse relationship between concentration was observed between Apo A1 genotype and plasma Apo A2 concentration and cardiovascular risk [194, PUFA intake. No significant association was detected in men. 195], findings of subsequent studies did not reveal significant The soy protein has been associated with improvement associations or even suggest a proatherogenic role [196–198]. on lipid levels and the mechanisms by which it acts Several SNPs have been reported in Apo A2 gene, but include several metabolic pathways regulated by different only MSP-I and −265T/C have been associated with plasma transcription factors. Animals who consumed a long-term lipid concentration [113]. MSP-1 was analyzed in a study of of soy protein had a decrease in insulin/glucagon ratio, 1,102 individuals from the Pacific island of Kosrae, in which due mainly to a reduction in insulin secretion [189]and several other candidate genes for increased cardiovascular an increase in glucagon concentration, that leads to an risk were also tested. The carriers of the less frequent MSP-1 expression reduction in the transcriptional factor sterol allele had higher serum TG concentration and, interestingly, regulatory element-binding protein (SREBP)-1 [190], which reduced blood pressure [199]. is involved in fatty acid synthesis and triglyceride (TG) The most studied SNP in the literature is the −265 esterification [191]. T/C (rs5082) that may affect element D of the Apo A2 14 Journal of Obesity 05). LDL . 05). PUFA = ,003) and P P<. = 05). These . P = P 76G/A ( 045). 4% of energy, the G allele − . < = P ,001) in G/A women than in G/G as 05). However, when PUFA intake was = . P = er in GA carriers. Carriers of the A allele P 14% higher plasma HDL-c than did carriers ff ∼ 8%, HDL-c levels in the A allele carriers were 13% carriers had When PUFA intake was across After the participants consumed the CHOdecrease diet, in there LDL was size a with respecthomozygotes for to the high-fat carriers diets of in GG of the A allele ( for this polymorphism had smaller LDLincreased size susceptibility as to well oxidation as after thethan SFA the diet GG homozygous ( size did not di > higher than those of G/G subjects ( LDL-c decreases ( compared to SFA diet. The variabilityfrom in the LDL-c SFA response diet to theassociated PUFA with diet LDL-c in (55,1%), women waist was hipand ratio the (11,4%), G/A polymorphism (10%). interactions were not significant in men The A allele carriers had higherTG plasma levels CT, than LDL-c, the and G homozygousdiet-induced allele significantly ( greater CT ( Cross-sectional Clinical trial followed by a randomized crossoverThe subjects consumed three diets for four weeks: SFA diet, CHO diet or MUFA diet. Clinical Trial Subjects were first fed a SFA diet fordays, 28 followed by a MUFA diet for 35 days anddiet a for PUFA 35 days. G allele: 83,5% A allele: 16,5% G allele: 41,2% A allele: 58,8% G allele: 87% A allele: 13% ] 109 spring ff ] 108 ] 107 Study [ 97 subjects recruited among students of the University of Cordoba at age-range 18–49 years [ 1,577 subjects from the Framingham O 50 subjects voluntaries members of two urban religious communities with 47,1 medium age [ 2: Summary of studies evaluating interaction between diet and variants of genes involved in lipoprotein metabolism. Table 76G/A (rs 1799837) 75G/A (rs 670) − − Gene Variant Population [reference] Frequency Design Main findings Apo A1 Journal of Obesity 15 036). . = 05). CC P ect against 026) and . ff = P<. P ect BMI. When SFA ff 018), but not a low . = P 029). Not associated with T2DM. . = P 22 g/d), ( 316) SFA diet. CC genotype was 22 g/d), SNP is associated with BMI and . ≥ ≥ = P 22 g/d), the SNP does not a ≤ 22 g/d), ( ≤ No significant association with HDL-c. Whenis SFA low intake ( intake is high ( Marginally associated with CT levels ( In Mediterranean individuals, the CCassociated genotype with a was 6.8% greater BMIhigh-SFA in diet those ( consuming a Carriers of the C allele havepostprandial significantly increases lower in plasma total TG and chylomicron TG, suggesting a protective e cardiovascular disease. ( obesity: a mean increase ofgenotype 6.2% was BMI associated ( with higherall obesity populations only prevalence in in the high-SFA intake stratum. associated with higher obesity prevalenceAsian in Indians Chinese only and with a high-SFA intake ( waist-to-hip ratio ( CandSFA > 265T − Cross-sectional, follow-up (20 years), and case-control analyses. Case-control study intake on BMI and obesity. Clinical trial: subjects were given a fatty meal containing 1 g fat and 7cholesterol/kg mg weight and capsules containing 60,000 IU vitamin A. Postprandial lipemia was assessed during the 11 h following the meal. Cross-sectional study: analyzed gene-diet interactions between the APOA2 2: Continued. Table ered strongly among ff CC subjects between Framingham and GOLDN: 15%; in BPR study: 10,5%. In normal glycemia T: 63%, C: 62% In T2DM subjects T: 37%, C: 38%. CT+CC: 60% TT: 40% Frequency of CC subjects di Chinese, Malays and Asian Indians (1–15%). ] ] 110 ] 111 113 spring ff ] 112 Study (1,454 Whites), The GOLDN Study (1,078 Whites) and BPR Study (930 Hispanics of Caribbean origin) [ 3,462 subjects from Framingham O 3,093 French Caucasian subjects with T2DM [ 88 normolipidemic young men from Spain [ 4,602 subjects from two independent populations: –high–cardiovascular risk Mediterranean –multiethnic Asian population including Chinese, Malays and Asian Indians) [ 256T/C (rs 5082) − Gene Variant Population [reference] Frequency Design Main findings Apo A2 16 Journal of Obesity 1131C − 032) and 034). . . = = P P -tocopherol was α 012). Both . 1131T/C and PUFA = − 1131T/C and total fat P − t statistical significance for ,001) was found. The 032) and CT ( . = = 001) in determining fasting P . P 02). . 48). The monounsaturated fatty 1131T/C with total fat energy intake = 01). The size of VLDL increases and . . − = P = 031) risk when compared with TT erence in postprandial responses of total = . P ff P P = P 044). The minor allele carriers had lower HDL-c . erent between TT men and C carriers. TT subject had = ff 1131T/C was associated with higher fasting TG and RLP 1131T/C carriers. The PUFA Apo A5 interactions were P A significant interaction between Significant interactions between the TG and chylomicron TG between LFcarriers and had HF delayed meal. peak C timeTT of subject total and TG higher compared postprandial to responsewhen at compared HF meal to LF meal. no significant di Fasting total TG were higher inmen, TC+CC but men fasting than chylomicron TT TG weredi not significantly The 56C/G polymorphism was associated with( HDL-c acids intake showed the highes these interactions. was observed for TG ( Individual who had the C alleleTG, presented VLDL-C, higher and plasma LDL-c levels. Plasma increased in C allele carriers comparedT with allele homozygote carriers ( concentration ( specific for dietary n-6 fatty acids. − − minor allele carriers had a lower obesity ( subjects in the high fat intakeintake group, was but low not ( when fat LDL decreases while PUFA intake increased in the than those with common variant ( plasma TG, RLP, and particle size. Thenot same be results observed could in 56 C/Gwho polymorphism. consumed In individuals a high PUFA diet (6% of the energy), the polymorphisms were associated with TG orAssociations other of lipids. the intake was found ( in relation to the BMI ( overweight ( Cross-sectional Cross-sectional Clinical Trial The subjects were randomly assigned to consume one of two types of experimental enteral formulae (LF versus HF) with a seven-day interval. Cross-sectional Cross-sectional cient ffi 2: Continued. Table between the APOA5 1131T/C and 56C/G was T allele: 87,15% C allele: 12,85% C allele: 88,95% G allele: 11,05% T allele: 86,6% C allele: 13,4% C allele: 88,3% G allele: 11,7% TT: 46,9% TC: 36,8% CC: 16,3% 0,016 − R TT: 84,51% TC: 15,16% CC: 0,34% The pairwise LD coe ] 117 spring spring ff ff ] 116 ] ] ] 115 115 114 Study with 54,2 medium age [ Study with 50,45 medium age [ 2,280 subjects from the Framingham O 49 male subjects at age 28–55 years were recruited from volunteers who responded to an advertisement for a nutrition study conducted by the Clinical Nutrition Research Team at Yonsei University [ 2,148 subjects from the Framingham O 299 healthy male at age 20–75 of the 5th Framework Program, [ 1,020 of the Boston Puerto Rican Health Study at age 45–75 [ 1131T/C (rs 662799) 1131T/C (rs 662799) 56C/G (rs 3135506) − − Gene Variant Population [reference] Frequency Design Main findings Apo A5 Journal of Obesity 17 3, ε 046) . 2, 05). . ε 3 = ε = P 3/ P ε 023 for men . 43% ( 001) between = . ε P 4genotype 2/ = ε ε P 2andApo 3/ ε ε 3/ ε 01), but the Apo E . ect is dependent on dietary = ff P 4 genotype presented the ε 025) and saturated fat ( . 4/ ε = 043). The individuals who carried . P = P 009) when compared with men after they . et Network (GOLDN) Study; HDL-c: High density ee. 4 individuals showed the lowest concentration 2 allele showed an inverse association with = ff ε ε 034 for women). The Apo E was higher in 2 allele demonstrated a positive correlation P . ε = 001). There were positive associations between total 041). Sex and Apo E genotype determine the Apo E 01). 2 carriers presented the highest Apo E plasma levels, . . . 4 expressing individuals with no significant ε P ε = = = ect of co saturated fatty acids; RLP: Remnant-like particle; SFA: saturated terol. ff P P P highest serum CT and LDL-c( and lowest HDL-c and TG Individuals who have The ApoE serum CT concentration ( A significant correlation between CT andderived energy from intake total ( women ( displayed a stronger positive correlationLDL-c between level serum and percentage ofof energy saturated derived fat from ( intake carriers, the shift from the SFA todiets CHO decreased or the MUFA-rich Apo E( concentration in women consumed the SFA diet. In Apo and between alcohol consumption and serum( HDL-c level the Apo while Apo after the SFA, CHO, and MUFA diets ( Apo fat. and saturated fat from usual dietLDL-c, and and serum an total inverse and associations ( and plasma levels, but this e polyunsaturated fat, dietary fiber, and lipidoverall. fractions Associations were in the same direction for and interactions between diet and genotypelipids, group except in on those blood who expressed polymorphisms do not influence the cholesterol-raising e was observed. Carriers of the Apo ee/day for four ff ee free period of ff Cross-sectional Clinical Trial: the subjects were submitted to a four intervention periods: 1 and 3aco three weeks, 2 and 4 600 mL co weeks. Cross-sectional Clinical trial: subjects consumed for 28 days a SFA-rich diet. After this, they were randomly assigned to one of twosequences: diet the first one received a MUFA-rich diet for 28 days, followed for more 28 days with CHO-rich diet. The other group consumed CHO diet before the MUFA diet. 2: Continued. Table 3 4 3 4 3 4 3 3 2 ε ε ε ε ε ε ε ε ε 4 4 2 4 4 4 2 3 3 3 ε ε ε ε ε ε ε ε ε ε 3/ 3/ 2/ 3/ 3/ 3/ 2/ 2 3/ 3/ ε ε ε ε ε ε ε ε ε ε 4/ 2/ 2/ 4/ 2/ 2/ 2/ 2/ 4/ 3/ ε ε ε ε ε ε ε ε ε ε 2/ ε 58,6% 23,5% 2,3% 12,4% 2,6% 0,6% 29,8% 2,5% 64,4% 17,4% 57% 14,4% 3,0% 1,7% 0,8% 7% 7,4% 78,6% 11,9% 9,5% llitus; TG: triglycerides; VLDL-c: Very low density lipoprotein-choles ] ] 119 ] 118 120 ] 121 22,915 subjects at age 45–75 years from Norfolk arm of the European Prospective Investigation of Cancer (EPIC) [ 132 clinically healthy Caucasians subjects at age 40–69 years [ 121 subjects recruited by advertising in Gothenburg’s major newspaper at age-range 30–65 years [ 84 subjects from students at the University of Cordoba at age 21–55 years [ Apo E2 rs429358 (T) + rs7412 (T) Apo E3 rs429358 (T) + rs7412 (C) Apo E4 rs429358 (C) + rs7412 (C) Apo E2 rs429358 (T) + rs7412 (T) Apo E3 rs429358 (T) + rs7412 (C) Apo E4 rs429358 (C) + rs7412 (C) Gene VariantApo E Population [reference] Frequency Design Main findings lipoprotein-cholesterol; HF: High fat; LDL-c: Low density lipoprotein-cholesterol; LF: Low fat; MUFA: Monounsaturated fatty acids; PUFA: Polyun BMI: body mass index; BPR: Boston-Puerto Rican Study; CHO: carbohydrate; CT: total cholesterol; GOLDN Study: Genetics of Lipid Lowering Drugs and Di fatty acids; SNP: single nucleotide polymorphism; T2DM: type 2 diabetes me 18 Journal of Obesity promoter, reported to be functional in 2 independent studies, 8. Apolipoprotein A5 in which C allele was associated with decreased plasma Apo It was well known that the main genetic factor related A2 concentration [200, 201]. In agreement, a meta-analysis to the plasma triglyceride determination is apolipoprotein of data from 12,387 subjects found no association of this C3, however trying to better understand the mechanisms, SNP with T2DM [110]. The C allele was associated with an intensive investigation of DNA sequence around the waist circumference in men [200] and in white woman, but APOA1/APOC3/APOA4 gene cluster was conducted [203]. not in African-American woman [202]. In addition, plasma Therefore the available mice and human sequences (about Apo A2 concentration was significantly lower in carriers of 200000 bp) around this locus were completed by sequencing C allele, while postprandial Apo B100 in the large VLDL and compared, leading to the identification of evolutionary > particles (Sf 60) were lower only in those in homozygous highly conserved sequence that contained a putative lipid- P = . for the C allele ( 01) [200]. Discordantly, in other study binding apolipoprotein gene, named the Apo A5 [204, 205]. carriers of the C allele had significantly lower postprandial The human Apo A5 gene consists of 4 exons and codes 369 increases in plasma total and chylomicron TG, suggesting aminoacid protein, which is expressed almost exclusively in ff a protective cardiovascular e ect, which may be related the liver [204]. Apo A5 is located on triglyceride rich particles to the lowered risk of postprandial hypertriglyceridemia (chylomicrons and very low density lipoproteins—VLDL) [113]. and HDL particles. An inverse relationship between APO A gene-diet interaction influencing BMI and obesity has A5 and TG levels has been described in animal studies, in been strongly and consistently replicated in many studies. which knockout mice developed hypertriglyceridemia and Interactions between the −265T/C and SFA intake on transgenic mice overexpressing APO A5 reduces plasma TG BMI and obesity in 3,462 individuals from three American levels [204]. populations: Framingham Offspring Study: (FOS) (1,454 In comparison to other apolipoproteins, the plasma Whites), Genetics of Lipid Lowering Drugs and Diet Net- concentration of Apo A5 is low in human—about 100 μg/L work: (GOLDN) Study (1,078 Whites), and Boston-Puerto [206]. In addition, researches confirm that Apo A5 binds Rican: (BPR) Study (930 Hispanics of Caribbean origin). to and enhances the activity of lipoprotein lipase (LPL) Frequencies of CC individuals did not differ between FOS enzyme and, consequently, reduces triglyceride levels in and GOLDN (15% in both), but this was lower in the VLDL particles. Moreover, the treatment in mice with Apo BPR Study (10.5%). No significant association of the SNP A5 lead to a reduction of VLDL-triglyceride production rate, with plasma HDL-c concentration was found among all the but the concentration of the VLDL particles was the same as three populations. Also, statistically significant interactions in normal mice [206, 207]. These results confirmed that Apo between this SNP and SFA intake were detected: when SFA A5 plays a role in the LPL activation. intake is low (≤22 g/d), the −265T>C SNP did not affect To identify common polymorphisms, extensive sequenc- BMI. However, when SFA intake was high (≥22 g/d), this ing of the Apo A5 interval in humans has been performed. SNP is strongly associated with BMI and obesity: a mean A set of 4 common polymorphisms (SNPs1 through 4; also increase of 6.2% BMI (4.3%–7.9%; P<.05). Moreover, named 259T/C, IVS3+476G/A, −1131T/C, and −12,238T/C, the CC genotype was significantly associated with higher resp.) were first identified within the human Apo A5. obesity prevalence in all populations only in the high-SFA Statistical analysis indicated that the minor alleles SNPs 1 intake stratum [111]. Similar results were showed in a cross- through 3 formed a relatively common haplotype that is sectional study with 4,602 subjects from two independent found in approximately 15% of Whites [204, 208]. populations (a high-cardiovascular risk Mediterranean and Subsequently, through direct DNA sequencing of the a multiethnic Asian population including Chinese, Malays, gene in 116 hyperlipidemic individuals, 9 additional SNPs and Asian Indians). In this study, the frequency of CC were identified. One of the polymorphisms (−3A/G) was ff subjects di ered strongly among populations (1–15%) and found to be in strong linkage disequilibrium with the minor an interaction of Apo A2 with saturated fat on body alleles for SNPs 1 trough 3 and this haplotype was named weight was confirmed: in Mediterranean subjects, the CC APOA5∗2. In addition, a second common polymorphism genotype was associated with a 6.8% greater BMI in those was also identified, which results in a C to G nonsynonymous P = . P = consuming a high ( 018), but not a low ( substitution (56C/G) that changes codon 19 from serine to . 316) SFA diet. Likewise, the CC genotype was significantly tryptophan. Further haplotype analysis in Whites indicated associated with higher obesity prevalence in Chinese and that the minor allele of this polymorphism defines a third P = . ∗ Asian Indians, only those with a high SFA intake ( 036) common Apo A5 haplotype (APOA5 3), and this one was [112]. also found in almost 15% of Whites. The remaining 7 poly- Despite the evidences of the influence of the ApoA2 morphisms from this study were either uncommon or not −265T/C polymorphism on body-weight-related measures, obviously associated with plasma triglycerides concentration modulated by saturated fat consumption, in different pop- [208]. ulations, its role in the control of circulating lipoproteins Thus, polymorphism discovery and haplotype analysis in and cardiovascular risk profile is not completely under- Whites defined 3 common haplotypes in the Apo A5 interval stood. and provided detailed information for genetic association Selected studies on the main interactions of Apo poly- studies in humans. In these analyses, the −1131T/C allele morphisms with dietary factors are in Table 2. (SNP3-rs662799) was used as a marker to define APOA5 ∗2, Journal of Obesity 19 whereas the 56C/G allele (S19W-rs3135506) was used to associated with triglyceride or other lipids, but interaction of define APOA5 ∗3 [208]. the −1131T/C SNP with total fat energy intake was observed Carriers of the less frequent allele −1131T/C have higher for plasma triglyceride (P = .032) and total cholesterol concentrations of plasma triglycerides, both in fasting [209] (P = .034). Apo A5 56C/G interacted with total fat intake and postprandial [210] states, and total cholesterol, lower in association with systolic and diastolic blood pressure (P< blood HDL-c [211], smaller LDL particles [212], and are at .001). higher cardiovascular risk [213]. Moreover, carriers of allele As previously cited, Apo A5 mutations have been exten- 56C/G show an increase in triglyceride levels, independently sively related to the postprandial lipemic response. Kim et al. of the effects observed for the −1131T/C, as well as an [116] compared low-fat (LF) meal and high-fat (HF) meal increased risk of suffering from atherosclerosis [214]. on the postprandial lipemic responses according to the In this context, it is most commonly found in the −1131T/C polymorphism of the APOA5 gene. Fasting total literature studies that related −1131T/C and 56C/G SNPs triglycerides were higher in heterozygous for the C allele (TC) to dietary factors. Thus, Lai et al. [114]haveproposedto and homozygous for C allele (CC) men than homozygous investigate the interaction between Apo A5 gene variation for the T allele (TT) men, but fasting chylomicron TG were and dietary fat in determining plasma fasting triglycerides, not significantly different between TT men and C carriers. remnant-like particle (RLP) concentrations, and lipopro- TT individuals had no significant differences in postprandial tein particle size in 1001 men and 1147 women who responses of total TG and chylomicron TG and postprandial were Framingham Heart Study participants. They found mean changes of chylomicron TG between LF and HF meal. a significant gene-diet interactions between the −1131T/C On the other hand, C carriers had delayed peak time of total polymorphism and PUFA intake in determining fasting TG compared to TT individuals and higher postprandial triglyceride, RLP concentrations, and particle size. However, response and mean changes of chylomicron TG at HF meal these interactions were not observed for the 56C/G poly- compared to LF meal. The authors hypothesized that these morphism. The −1131C allele was associated with higher facts occur due to the limiting capacity to clear chylomicron fasting triglyceride and RLP concentrations (P<.01) only TG or hydrolyze TG on HF diet in TC and CC men, resulting in the individuals consuming a high-PUFA diet (>6% of in higher postprandial triglyceridemia. total energy). Similar interactions were found for the sizes Vitamin E also has been related to Apo A5 polymorph- of VLDL and LDL particles. Only in carriers of the −1131C isms since it is a lipophilic micronutrient that is also allele did the size of these particles increase (VLDL) or transported within lipoproteins. The effects of the Apo A5 decrease (LDL) as PUFA intake increased (P<.01). The −1131T/C gene variant according to vitamin E status (α- authors further analyzed the effects of w-6 and w-3 fatty tocopherol, γ-tocopherol, buccal mucosa cells total vitamin acids, and found that the interactions between PUFA and E, LDL α-tocopherol, and LDL γ-tocopherol) and lipid Apo A5 were specific for dietary n-6 fatty acids. profile (VLDL, HDL, intermediate density lipoprotein (IDL), Another research involving the Framingham Heart Study and LDL) were investigated. C allele carriers showed signif- participants concluded that Apo A5 gene variation modu- icantly higher TG, VLDL, and LDL concentrations, higher lates the effects of dietary fat intake on BMI and obesity cholesterol in VLDL and IDL, and higher plasma fatty risk. The authors investigated the interaction between the acids. Plasma α-tocopherol was increased significantly in C Apo A5 −1131T/C and 56C/G polymorphisms and the allele carriers compared with homozygote T allele carriers macronutrient intake (total fat, CHO, and protein) in their (P = .02), suggesting that higher plasma lipids in the TC relation to the BMI and obesity risk. They found a significant and CC genotypes were efficiently protected against lipid interaction between the −1131T/C SNP and total fat intake peroxidation by higher plasma α-tocopherol concentration for BMI. In individuals homozygous for the −1131T major [117]. allele, BMI increased as total fat intake increased. In contrast, Selected studies on the main interactions of Apo poly- this increase was not present in carriers of the −1131C minor morphisms with dietary factors in Table 2. allele. Significant interactions were found in determining obesity and overweight risks, since the Apo A5 −1131C minor allele carriers had a lower obesity risk (OR 0.61, P = 9. Apolipoprotein E .032) and overweight risk (OR 0.63, P = .031) compared to TT individuals in the high-fat intake group (≥30% of Apo E was discovered in 1970 as a component of triglyceride- energy) but not when fat intake was low (OR 1.16, P = rich lipoproteins [216]. It is an amphipathic 299 amino .47 and OR 1.15, P = .48) for obesity and overweight, acid glycoprotein of 34,145 KDa that is mainly secreted by respectively. When specific fatty acid groups were analyzed, hepatocytes [217], but expressed in the brain and liver [218]. MUFA showed the highest statistical significance for these The primary functional role of Apo E is to transport and interactions [215]. deliver lipids mainly through the LDL-c receptor pathway. A more recent study [115] that aimed to determine the A secondary proposed pathway involves the heparin sul- association of the same Apo A5 polymorphism previously phate proteoglycan (HSPG)/LDL-C receptor-related protein mentioned with plasma lipids and markers of MS, alone and pathway [219]. Apo E acting as a ligand for these receptors in interaction with total fat intake in Puerto Ricans (n = [220] plays a crucial role in determining the metabolic fate 802, 45–75 yrs) found interesting results. Apo A5 S19W was of plasma lipoproteins and consequently of cholesterol [220, associated with HDL (P<.05). Neither polymorphism was 221], while its accumulation on the surface of lipoproteins 20 Journal of Obesity can decrease the lipolysis rate of TG by lipase [222–224]. genotype. They concluded that sex and Apo E genotype Furthermore, Apo E as a component of HDL influences the determine Apo E circulating levels; however, this effect is cholesterol influx and efflux of cells [225–227]. dependent on dietary fat. Plasma Apo E isoforms have two kinds of polymor- Loktionov et al. [120] investigated the relationship phism, one genetically determined and one not. The former between Apo E genotype and the effects of a CHD-promot- polymorphism is the result of three alleles, called epsilon ing diet in free-living individuals. No relationship between (ε) alleles: ε2, ε3, and ε4 at a single gene locus and that dietary fats and cholesterol levels were observed in the differs in 112 e 158 position of the aminoacids (Apo E2: E2 or E3 group when analyzed separately. However, the Cys112-Cys158; Apo E3: Cys112-Arg158; Apo E4: Arg112- presence of the E4 allele resulted in positive associations Arg158). The relative frequency in white population of between total and saturated fat intake with total cholesterol Apo E2, Apo E3, and Apo E4 alleles is 0,08, 0,77, and and LDL. When expressing fat and saturated fat intake as 0,15, respectively [228]. This alleles can form six genotypes a percentage of total energy intake compared to absolute ranking in order of most to least common (ε33, ε34, ε23, amounts, this relationship was stronger. The study suggested ε44, ε24, and ε22) [229]. The ε33 genotype, being the most that individuals with the E4 allele may be more responsive common, is used as reference for all Apo-E-related func- to the dietary therapy. Calculations showed that cutting tions [230]. The nongenetically determined polymorphism saturated fat consumption in half resulted in a 20% decrease results from variable posttranslational sialyation of Apo E in LDL in those with the E4 allele compared to just a 5% and accounts for 10%–20% of plasma Apo E [219]. In decrease in those without. This speculation is supported by addition, Apo E isoforms are important determinants of the results of Dreon et al. [234] who found that that LDL- postprandial lipemia [231]. It has been demonstrated that lowering effects of a low-fat diet were greater in men with the Apo E2 homozygous subjects have the lowest affinity for E4 allele than those without. TRL remnant receptors, and this genotype is associated A prospective transversal cohort study including 22,915 with delayed postprandial clearance. A recent Brazilian study participants of the Norfolk arm of the European Prospective indicates that carriers of Apo E4 had a positive association Investigation of Cancer (EPIC Norfolk) investigated whether with higher total cholesterol (P<.001), LDL-C (P< blood lipid response to dietary fat and fibers varied according .001), total-cholesterol/HDL-C ratio (P<.001), LDL/HDL- to the Apo E gene locus. Significant (P<.001) differences Cratio(P<.001), lower HDL-C values (P<.001), in serum lipids according to genotype were found. Highest and higher risk to obesity (OR = 1.358, 95% CI = 1.019– total and LDL and lowest HDL and TG levels were detected 1.811) [232]. Furthermore, patients with MS, with genotypes in ε4/ε4 individuals. Total dietary fat and SFA were associated other than the Apo E3/3, are at greater risk of postprandial with total and LDL cholesterol concentrations and significant hypertriglyceridemia and hyperuricemia following the acute inverse associations between PUFA and dietary fiber and ingestion of a fat overload [233]. Several studies have shown lipid fractions were found. Associations were in the same interaction between the postprandial lipemic response and direction for ε2, ε3, and ε4 expressing individuals with no Apo E polymorphisms. Subjects who are carriers of the significant interactions between diet and genotype group on −219T allele, which is located in the promoter region of blood lipids, except in the 3% individuals expressing ε2/ε4 the Apo E gene, −219G/T, display lower levels of Apo E in (P<.05) in whom the associations were doubled [118]. plasma in both fasting and postprandial states, and this effect Strandhagen et al. [121] hypothesized that the increase was associated with a stronger postprandial response, and a of plasma CT and TG stimulated by the ingestion of the decrease in transcription activity [231]. coffee might be modulated by the Apo ε2/ε3/ε4. The subjects Moreno et al. [119] evaluated whether the quantity and were submitted to four periods of intervention: two of quality of dietary fat interacts with the Apo E genotype and them included three free weeks of coffee and the others sex modifying plasma Apo E levels in young healthy subjects. two included the ingestion of 600 mL/day for four weeks. Theparticipantsweresubjectedtothreedietaryperiods,each The cholesterol-raising effect of coffee was not influenced oneof4weeks.ThefirstwasaSFA-enricheddiet(38%fat, significantly by the Apo E allele carriers status, however being 20% SFA, 12% MUFA, and 6% PUFA), which was individuals with Apo ε2 polymorphism had significantly followed by a CHO rich diet (55% CHO, <30% fat, being lower total cholesterol concentration at baseline (P = .01). 10% SFA, 12% MUFA, and 6% PUFA) or a MUFA-rich diet Selected studies on the main interactions of Apo poly- (38% fat, being <10% SFA, 22% MUFA and 6% PUFA). Apo morphisms with dietary factors are in Table 2. E2 carriers have the highest Apo E levels, whereas Apo E4 individuals show the lowest concentration after the saturated fat, CHO, and MUFA diets. Women had significantly higher 10. Glutathione Peroxidases Apo E concentration than men only after the consumption of the high-saturated fat diet. The SFA-enriched diet increased GPx1 and GPx2 have antioxidant functions, protecting cells the Apo E plasma concentration when compared with the from oxidative stress damage. Knockout mice lacking both CHO-, and MUFA-rich diets in women, but not in men. GPx1 and 2 are more susceptible to an oxidative challenge In women, but not in men, the shift from the SFA- to [235]. Responses of transgenic mice lacking or overexpress- CHO- or MUFA-rich diets significantly decreased the Apo ing GPx1 have suggested roles for GPx1 in relation to E concentration in apoE3/2 and apoE3/3 subjects, whereas both reactive oxygen species and reactive nitrogen species no differences were observed in women with the apoE4/3 as well as a link to insulin-mediated effects [235]. GPx4 Journal of Obesity 21 appears to have a complex range of functions in protection their genes, as well as the factors involved in their synthesis, from oxidative stress, lipoxygenase metabolism, and sperm provide a focus for both mechanistic and genetic studies. function [236, 237] and may play a role in regulation of Selenoprotein synthesis and functional activity will result leukotriene biosynthesis which in turn regulates the biosyn- from the combined influences of the genetic information thesis of leukotrienes, thromboxanes, and prostaglandins, in the genes encoding the selenome and dietary Se intake. thus modulating inflammatory. [238]. Polymorphism in these genes could potentially influence the A recent research [82] investigated the associations be- Se intake required to achieve a particular level of functional tween glutathione peroxidase-1 Pro198Leu polymorphism, activity and thus individual requirements for Se [244]. selenium status, and DNA damage levels in obese women Any functional effect of a single nucleotide change is after consumption of Brazil nuts. Participants consumed likely to be relatively small and can be higher with a one Brazil nut, which provided approximately 290 age of combination of different SNP(s) in selenoprotein-related Se a day, for 8 weeks. At baseline, 100% of the subjects gene(s), possibly in combination with dietary Se intake. were Se deficient, and after the supplementation, there was When this is suboptimal, such an interaction may give rise an improvement in plasma Se (P<.001 for Pro/Pro and to differences in the ability of selenoproteins to function at Pro/Leu, P<.05 for Leu/Leu), erythrocyte Se (P = .00 for optimal capacity; thus, selenoprotein function in protection Pro/Pro and Pro/Leu, P<.05 for Leu/Leu), and GPx activity of cells from oxidative stress or inflammation could be (P = .00 for Pro/Pro, P<.00001 for Pro/Leu, P<.001 for compromised [244]. Leu/Leu). In addition, the Pro/Pro group showed a decrease The best characterized selenoproteins are the glutathione in DNA damage after Brazil nut consumption compared peroxidases (GPxs), the thioredoxin reductases (TRs), and with baseline (P<.005), and those levels were higher in deiodinases (IDIs), and selenoprotein P (SeP) [244]. SeP is an Leu/Leu individuals compared with those with the wild-type extracellular glycoprotein and is expressed most abundantly genotype (P<.05). in liver and at relatively high level in kidney and heart [245]. A study about the region of the GPX4 gene corre- However, the expression and regulation of SeP in adipose sponding to the 3UTR was scanned for mutations in a tissue has not been previously reported [246, 247]. SeP group of 66 volunteers Scotland. The data show a T/C represents the best Se marker, accounting for the majority variant at position 718, distribution was 34% CC, 25% TT, of circulating Se in blood and responding over a broad and 41% TC. Individuals of different genotypes exhibited rangetoSeintake[248]. In addition to the role as a significant differences in the plasma levels of 5-lipoxygenase selenium transporter, SeP has been postulated to possess total products in lymphocyte, with 718C showed increased antioxidant properties, for instance, SeP has the ability to levels of those products compared to 718T and 718T/C (36% bind epithelial cells and displays phospholipid hydroperoxide and 44% increases, resp.). The data suggest that the SNP 718 thiol peroxidase activity [248]. The overexpression of SeP1 has functional effects and support the hypothesis that GPX4 suppressed hydrogen peroxide-induced activation of JNK plays a regulatory role in leukotriene biosynthesis [239]. and p38 in retinal pigment epithelial cells [249]. Another study showed that the activation of GPx activity is required The same SNP 718 (rs713041) influenced the concentra- for the protective role of SeP against oxidative damage in tion of GPx4 and other selenoproteins in vivo. A selenium human endothelial cells and astrocytes [250, 251] indicating supplementation trial was carried out with both homozygote that SeP participates in the control of the cellular redox genotypes for this SNP. Blood samples were analyzed at balance. Since oxidative stress in adipose tissue has been baseline, after a 6 week supplementation with 100 g Seleniun implicated in obesity and insulin resistance, it has been (Se) as sodium selenite/day and during a 6 week washout hypothesized that antioxidant SeP may likely be involved in period. Both lymphocyte GPx1 protein concentrations and cellular mechanisms of oxidative stress -linked inflammation plasma GPx3 activity increased significantly after Se sup- and insulin resistance [252]. plementation in CC but not TT participants. After Se A study showed that SeP1 gene expression was signifi- withdrawal, there was a significant fall in both lymphocyte cantly reduced in adipose tissue of ob/ob and high-fat-diet- GPx4 protein concentrations and GPx4 activity in TT but induced obese mice, as well as in primary adipose cells not in CC participants; this effect was modulated by sex [83]. isolated from Zucker obese rats. Interestingly, SeP1 gene Selected studies on the main interactions of polymor- silencing resulted in the reduction in GPX activity and the phisms with dietary factors are in Table 1. upregulation of inflammatory cytokines MCP-1 and IL-6 in preadipocytes, leading to the inhibition of adipogenesis 11. Selenoprotein P and adipokine and lipogenic gene expression. SeP has an important role in adipocyte differentiation via modulating The physiological functions of Se result from its existence in a oxidative stress and inflammatory response [252]. number of selenoproteins in which Se is present as the amino A research studied two SNPs in the SeP gene, one in the acid selenocysteine (Sec). Se was first shown to be an essential coding region (position 24731, causing an Ala to Thr change) component of glutathione peroxidase and subsequently has and one in 3UTR (position 25191). Their frequency was been found in 25 mammalian selenoproteins [240–242]. similar in Caucasian, Chinese, and South Asian populations. In eukaryotes Se is incorporated into selenoproteins as the Prospectively genotyped volunteers were supplemented for amino acid selenocysteine in a process requiring a stem loop 6 wk with 100 ug sodium selenite/day. Blood samples were within region 3UTR of the mRNA [243]. These proteins and analyzed for plasma Se and selenoprotein biomarkers at 22 Journal of Obesity baseline, after supplementation, and during a washout [7]P.G.Kopelman,“Obesityasamedicalproblem,”Nature,vol. period. Plasma Se, SeP, and GPx3 levels increased with 404, no. 6778, pp. 635–643, 2000. supplementation. Baseline plasma Se content depended on [8] G. S. Hotamisligil, “Inflammatory pathways and insulin both SeP genotypes and BMI. Presupplementation SeP action,” International Journal of Obesity, vol. 27, supplement concentration was associated with gender and genotype at 3, pp. S53–S55, 2003. SNP 24731 and postsupplementation concentration with [9] G. Boden and G. I. Shulman, “Free fatty acids in obesity and SNP 25191. The data revealed two common functional SNPs type 2 diabetes: defining their role in the development of β within the human SePP gene that may predict behavior of insulin resistance and -cell dysfunction,” European Journal biomarkers of Se status and response to supplementation and of Clinical Investigation, vol. 32, supplement 3, pp. 14–23, 2002. thus susceptibility to disease [84]. Selected studies on the main interactions of polymor- [10] H. Ortsater,¨ “Arachidonic acid fights palmitate: new insights into fatty acid toxicity in β-cells,” Clinical Science, vol. 120, phisms with dietary factors are in Table 1. no. 5, pp. 179–181, 2011. Several SNPs have been identified in the genes for [11] S. G. Barroso, V. G. Abreu, and E. A. Francischetti, “The selenoproteins and some are linked to the occurrence of adipose tissue in the genesis of hypertension and atheroscle- diseases. However, the current knowledge is limited and the rotic cardiovascular disease. An emerging concept,” Arquivos further studies are needed to identify and assess the effect Brasileiros de Cardiologia, vol. 78, no. 6, pp. 618–630, 2002. of these SNPs in the nutritional requirements of Se, your [12] B. M. Popkin, “Nutrition in transition: the changing global metabolism, and susceptibility to diseases. nutrition challenge,” Asia Pacific Journal of Clinical Nutrition, vol. 10, pp. S13–S18, 2001. [13] WHO, Diet, Nutrition and the Prevention of Chronic Diseases. 12. Conclusion Report of a Joint WHO/FAO Expert Consultation,WHO Technical Report Series, 916, World Health Organization, There is evidence that genetic variation at genes involved in Geneva, Switzerland, 2003. the etiology of inflammation and dyslipidemia could interact [14] R. Sinha, G. Fisch, B. Teague et al., “Prevalence of impaired with environmental exposures, such as diet, to modulate glucose tolerance among children and adolescents with individuals’ susceptibility to developing these conditions. marked obesity,” The New England Journal of Medicine,vol. Present data support the notion that more tailored dietary 346, no. 11, pp. 802–810, 2002. recommendations may be helpful in the prevention of [15] G. M. Reaven, “Role of insulin resistance in human disease,” obesity-induced comorbidities. However, the relevance and Diabetes, vol. 37, no. 12, pp. 1595–1607, 1988. magnitude of nutrient-gene interaction needs to be further [16] G. Boden, “Fatty acid-induced inflammation and insulin elucidated. It is also important to emphasize that the study resistance in skeletal muscle and liver,” Current Diabetes of the combined effect of SNPs in different genes, as Reports, vol. 6, no. 3, pp. 177–181, 2006. well as haplotypes, may be promising since gene-nutrient [17] P. Arner, “Insulin resistance in type 2 diabetes: role of fatty interactions in obesity-induced metabolic disturbances are acids,” Diabetes/Metabolism Research and Reviews,vol.18,no. complex. 2, pp. S5–S9, 2002. [18]G.S.Hotamisligil,P.Arner,J.F.Caro,R.L.Atkinson,and B. M. Spiegelman, “Increased adipose tissue expression of References tumor necrosis factor-α in human obesity and insulin resistance,” The Journal of Clinical Investigation,vol.95,no.5,pp. [1] WHO, Obesity: Preventing and Managing the Global Epi- 2409–2415, 1995. demic, WHO Technical Report Series, 894, World Health [19] A. Green, S. B. Dobias, D. J. A. Walters, and A. R. Brasier, Organization, Geneva, Switzerland, 2003. “Tumor necrosis factor increases the rate of lipolysis in [2]C.L.Ogden,S.Z.Yanovski,M.D.Carroll,andK.M.Flegal, primary cultures of adipocytes without altering levels of “The epidemiology of obesity,” Gastroenterology, vol. 132, no. hormone-sensitive lipase,” Endocrinology, vol. 134, no. 6, pp. 6, pp. 2087–2102, 2007. 2581–2588, 1994. [3]D.A.Galuska,C.Gillespie,S.A.Kuester,A.H.Mokdad,M. [20] M. Ryden,´ E. Arvidsson, L. Blomqvist, L. Perbeck, A. E. Cogswell, and C. M. Philip, “State-specific prevalence of α obesity among adults—United States, 2007,” Morbidity and Dicker, and P. Arner, “Targets for TNF- -induced lipolysis Mortality Weekly Report, vol. 57, no. 28, pp. 765–768, 2008. in human adipocytes,” Biochemical and Biophysical Research [4] Instituto Brasileiro de Geografia e Estatistica (IBGE), Dire- Communications, vol. 318, no. 1, pp. 168–175, 2004. toria de Pesquisas, CoordenaçãdeTrabalhoeRendimento. [21]J.S.Yudkin,I.Juhan-Vague,E.Haweetal.,“Low-grade Pesquisa de Orçamentos Familiares 2008–2009, Instituto inflammation may play a role in the etiology of the metabolic Brasileiro de Geografia e Estatistica, Rio de Janeiro, Brazil, syndrome in patients with coronary heart disease: the 2010. HIFMECH study,” Metabolism, vol. 53, no. 7, pp. 852–857, [5]T.Kelly,W.Yang,C.S.Chen,K.Reynolds,andJ.He, 2004. “Global burden of obesity in 2005 and projections to 2030,” [22]D.J.Wexler,F.B.Hu,J.E.Manson,N.Rifai,andJ.B.Meigs, International Journal of Obesity, vol. 32, no. 9, pp. 1431–1437, “Mediating effects of inflammatory biomarkers on insulin 2008. resistance associated with obesity,” Obesity Research,vol.13, [6]A.deLorenzo,V.delGobbo,M.G.Premrov,M.Bigioni,F. no. 10, pp. 1772–1783, 2005. Galvano, and L. Di Renzo, “Normal-weight obese syndrome: [23] J. M. Fernandez-Real´ and W. Ricart, “Insulin resistance and early inflammation?” The American Journal of Clinical Nutri- chronic cardiovascular inflammatory syndrome,” Endocrine tion, vol. 85, no. 1, pp. 40–45, 2007. Reviews, vol. 24, no. 3, pp. 278–301, 2003. Journal of Obesity 23

[24] K. Ando and T. Fujita, “Metabolic syndrome and oxidative Gln27Glu β-adrenoceptor polymorphism,” The Journal of stress,” Free Radical Biology & Medicine,vol.47,no.3,pp. Nutrition, vol. 133, no. 8, pp. 2549–2554, 2003. 213–218, 2009. [39] C. W. Danawati, M. Nagata, H. Moriyama et al., “A pos- [25]T.M.Gwathmey,B.M.Westwood,N.T.Pirroetal.,“Nuclear sible association of Pro12Ala polymorphism in peroxisome angiotensin-(1-7) receptor is functionally coupled to the proliferator-activated receptor γ 2 gene with obesity in native formation of nitric oxide,” American Journal of Physiology, Javanese in Indonesia,” Diabetes/Metabolism Research and vol. 299, no. 5, pp. F983–F990, 2010. Reviews, vol. 21, no. 5, pp. 465–469, 2005. [26] T. Ogihara, T. Asano, H. Katagiri et al., “Oxidative stress [40] A. B. Jones, “Peroxisome proliferator-activated recep- induces insulin resistance by activating the nuclear factor- tor (PPAR) modulators: diabetes and beyond,” Medicinal κB pathway and disrupting normal subcellular distribution Research Reviews, vol. 21, no. 6, pp. 540–552, 2001. of phosphatidylinositol 3-kinase,” Diabetologia, vol. 47, no. [41]I.P.Torra,G.Chinetti,C.Duval,J.C.Fruchart,andB.Staels, 5, pp. 794–805, 2004. “Peroxisome proliferator-activated receptors: from transcrip- [27] F. Sofi, R. Abbate, G. F. Gensini, and A. Casini, “Accruing tional control to clinical practice,” Current Opinion in evidence on benefits of adherence to the Mediterranean diet Lipidology, vol. 12, no. 3, pp. 245–254, 2001. on health: an updated systematic review and meta-analysis,” [42] B. Aldhoon, H. Zamrazilova,´ I. Aldhoon Hainerovaetal.,´ The American Journal of Clinical Nutrition,vol.92,no.5,pp. “Role of the PPARα Leu162Val and PPARγ2 Pro12Ala gene 1189–1196, 2010. polymorphisms in weight change after 2.5-year follow-up in [28] B. Sishi, B. Loos, B. Ellis, W. Smith, E. F. Du Toit, and A.-M. Czech obese women,” Folia Biologica, vol. 56, no. 3, pp. 116– Engelbrecht, “Diet-induced obesity alters signalling pathways 123, 2010. and induces atrophy and apoptosis in skeletal muscle in a [43]S.S.Deeb,L.Fajas,M.Nemotoetal.,“APro12Alasubstitu- prediabetic rat model,” Experimental Physiology, vol. 96, no. tion in PPARγ2 associated with decreased receptor activity, 2, pp. 179–193, 2011. lower body mass index and improved insulin sensitivity,” [29]M.Figueras,M.Olivan,S.Busquets,F.J.Lopez-Soriano,´ Nature Genetics, vol. 20, no. 3, pp. 284–287, 1998. and J. M. Argiles,´ “Effects of eicosapentaenoic acid (EPA) [44] L. Frederiksen, K. Brødbæk, M. Fenger et al., “Comment: treatment on insulin sensitivity in an animal model of Studies of the Pro12Ala polymorphism of the PPAR-γ gene in diabetes: improvement of the inflammatory status,” Obesity, the Danish MONICA cohort: homozygosity of the Ala allele vol. 19, no. 2, pp. 362–369, 2011. confers a decreased risk of the insulin resistance syndrome,” [30] G. Riccardi, R. Giacco, and A. A. Rivellese, “Dietary fat, The Journal of Clinical Endocrinology and Metabolism, vol. 87, insulin sensitivity and the metabolic syndrome,” Clinical no. 8, pp. 3989–3992, 2002. Nutrition, vol. 23, no. 4, pp. 447–456, 2004. [45]H.Mori,H.Ikegami,Y.Kawaguchietal.,“ThePro12→ [31] R. C. Bunn, G. E. Cockrell, Y. Ou, K. M. Thrailkill, C. Ala substitution in PPAR-γ is associated with resistance K. Lumpkin Jr., and J. L. Fowlkes, “Palmitate and insulin to development of diabetes in the general population: synergistically induce IL-6 expression in human monocytes,” possible involvement in impairment of insulin secretion in Cardiovascular Diabetology, vol. 9, article 73, 2010. individuals with type 2 diabetes,” Diabetes,vol.50,no.4,pp. [32] J. A. Chavez and S. A. Summers, “Characterizing the effects 891–894, 2001. of saturated fatty acids on insulin signaling and ceramide [46] D. Altshuler, J. N. Hirschhorn, M. Klannemark et al., “The and diacylglycerol accumulation in 3T3-L1 adipocytes and common PPARγ Pro12Ala polymorphism is associated with C2C12 myotubes,” Archives of Biochemistry and Biophysics, decreased risk of type 2 diabetes,” Nature Genetics, vol. 26, vol. 419, no. 2, pp. 101–109, 2003. no. 1, pp. 76–80, 2000. [33] A. Kennedy, K. Martinez, C. C. Chuang, K. Lapoint, and M. [47]Y.Lu,X.Ye,Y.Caoetal.,“Geneticvariantsinperoxisome Mcintosh, “Saturated fatty acid-mediated inflammation and proliferator-activated receptor-γ and retinoid X receptor- insulin resistance in adipose tissue: mechanisms of action α gene and type 2 diabetes risk: a case-control study and implications,” The Journal of Nutrition, vol. 139, no. 1, of a Chinese han population,” Diabetes Technology and pp. 1–4, 2009. Therapeutics, vol. 13, no. 2, pp. 157–164, 2011. [34] K. Esposito, R. Marfella, M. Ciotola et al., “Effect of a [48] S. Masud and S. Ye, “Effect of the peroxisome proliferates Mediterranean-style diet on endothelial dysfunction and activated receptor-γ gene Pro12Ala variant on body mass markers of vascular inflammation in the metabolic syn- index: a meta-analysis,” Journal of Medical Genetics, vol. 40, drome: a randomized trial,” Journal of the American Medical no. 10, pp. 773–780, 2003. Association, vol. 292, no. 12, pp. 1440–1446, 2004. [49] B. A. Beamer, C.-J. Yen, R. E. Andersen et al., “Association of [35] R. Raqib and A. Cravioto, “Nutrition, immunology, and the Pro12Ala variant in the peroxisome proliferator-activated genetics: future perspectives,” Nutrition Reviews,vol.67,pp. receptor-γ2 gene with obesity in two Caucasian populations,” S227–S236, 2009. Diabetes, vol. 47, no. 11, pp. 1806–1808, 1998. [36] M. J. Gibney and E. R. Gibney, “Diet, genes and disease: [50] E. S. Tai, D. Corella, M. Deurenberg-Yap et al., “Differential implications for nutrition policy,” Proceedings of the Nutrition effects of the C1431T and Pro12Ala PPARγ gene variants Society, vol. 63, no. 3, pp. 491–500, 2004. on plasma lipids and diabetes risk in an Asian population,” [37] A. Memisoglu, F. B. Hu, S. E. Hankinson et al., “Interaction Journal of Lipid Research, vol. 45, no. 4, pp. 674–685, 2004. between a peroxisome proliferator-activated receptor γ gene [51]E.Y.Oh,K.M.Min,J.H.Chungetal.,“Significance polymorphism and dietary fat intake in relation to body of ProAla mutation in peroxisome proliferator-activated mass,” Human Molecular Genetics, vol. 12, no. 22, pp. 2923– receptor-γ2 in Korean diabetic and obese subjects,” The 2929, 2003. Journal of Clinical Endocrinology and Metabolism,vol.85,no. [38] J. A. Martınez,M.S.Corbal´ an,A.S´ anchez-Villegas,´ L. Forga, 5, pp. 1801–1804, 2000. A.Marti,andM.A.Martınez-Gonz´ alez,´ “Obesity risk is [52] Y. L. Muller, C. Bogardus, B. A. Beamer, A. R. Shuldiner, associated with carbohydrate intake in women carrying the and L. J. Baier, “A functional variant in the peroxisome 24 Journal of Obesity

proliferator—activated receptor γ2 promoter is associated the Korean population,” Diabetic Medicine,vol.22,no.9,pp. with predictors of obesity and type 2 diabetes in Pima 1161–1166, 2005. Indians,” Diabetes, vol. 52, no. 7, pp. 1864–1871, 2003. [66] E. L. Rosado, J. Bressan, M. F. Martins, P. R. Cecon, and J. A. [53] R. Rosmond, M. Chagnon, and C. Bouchard, “The Pro12Ala Martınez,´ “Polymorphism in the PPARgamma2 and beta2- PPARγ2 gene missense mutation is associated with obesity adrenergic genes and diet lipid effects on body composition, and insulin resistance in Swedish middle-aged men,” Dia- energy expenditure and eating behavior of obese women,” betes/Metabolism Research and Reviews,vol.19,no.2,pp. Appetite, vol. 49, no. 3, pp. 635–643, 2007. 159–163, 2003. [67] E. L. Rosado, J. Bressan, J. A. M. Hernandez,´ M. F. Martins, [54] J. L. Gonzalez´ Sanchez,M.SerranoR´ ıos,´ C. Fernandez´ and P. R. Cecon, “EffectofdietandPPARγ2andβ2- Perez,´ M. Laakso, and M. T. Martınez´ Larrad, “Effect of adrenergic receptor genes on energy metabolism and body the Pro12Ala polymorphism of the peroxisome proliferator- composition in obese women,” Nutricion Hospitalaria,vol. activated receptor γ-2 gene on adiposity, insulin sensitivity 21, no. 3, pp. 317–331, 2006. and lipid profile in the Spanish population,” European [68]J.Luan,P.O.Browne,A.H.Hardingetal.,“Evidencefor Journal of Endocrinology, vol. 147, no. 4, pp. 495–501, 2002. gene-nutrient interaction at the PPARγ locus,” Diabetes,vol. [55] H. Mirzaei, S. M. Akrami, T. Golmohammadi et al., 50, no. 3, pp. 686–689, 2001. “Polymorphism of Pro12Ala in the peroxisome proliferator- [69]F.Soriguer,S.Morcillo,F.Cardonaetal.,“Pro12Alapoly- activated receptor 2 gene in Iranian diabetic and obese morphism of the PPARG2 gene is associated with type subjects,” Metabolic Syndrome and Related Disorders,vol.7, 2 diabetes mellitus and peripheral insulin sensitivity in a no. 5, pp. 453–458, 2009. population with a high intake of oleic acid,” The Journal of [56] V. S. Mattevi, V. M. Zembrzuski, and M. H. Hutz, “Effects Nutrition, vol. 136, no. 9, pp. 2325–2330, 2006. of a PPARG gene variant on obesity characteristics in Brazil,” [70] S. Scaglioni, E. Verduci, M. Salvioni et al., “PPAR-γ2 Brazilian Journal of Medical and Biological Research,vol.40, Pro12Ala variant, insulin resistance and plasma long-chain no. 7, pp. 927–932, 2007. polyunsaturated fatty acids in childhood obesity,” Pediatric [57] E. Morini, V. Tassi, D. Capponi et al., “Interaction between Research, vol. 60, no. 4, pp. 485–489, 2006. PPARγ2 variants and gender on the modulation of body [71] R. F. Grimble, W. M. Howell, G. O’Reilly et al., “The ability weight,” Obesity, vol. 16, no. 6, pp. 1467–1470, 2008. of fish oil to suppress tumor necrosis factor α production [58] G. V. Dedoussis, N. Vidra, J. Butler et al., “Peroxisome by peripheral blood mononuclear cells in healthy men proliferator-activated receptor-γ (PPARγ) Pro12Ala poly- is associated with polymorphisms in genes that influence morphism and risk for pediatric obesity,” Clinical Chemistry tumor necrosis factor α production,” The American Journal and Laboratory Medicine, vol. 47, no. 9, pp. 1047–1050, 2009. of Clinical Nutrition, vol. 76, no. 2, pp. 454–459, 2002. [59] M. Ghoussaini, D. Meyre, S. Lobbens et al., “Implication of [72]S.E.Belisle,L.S.Leka,J.Delgado-Lista,P.F.Jacques,J.M. the Pro12Ala polymorphism of the PPAR-gamma 2 gene in Ordovas, and S. N. Meydani, “Polymorphisms at cytokine type 2 diabetes and obesity in the French population,” BMC genes may determine the effect of vitamin E on cytokine Medical Genetics, vol. 6, article 11, 2005. production in the elderly,” The Journal of Nutrition, vol. 139, [60] A. Stefanski, L. Majkowska, A. Ciechanowicz et al., “Lack of no. 10, pp. 1855–1860, 2009. association between the Pro12Ala polymorphism in PPAR- [73]D.A.deLuis,R.Aller,O.Izaola,M.G.Sagrado,andR. γ2 gene and body weight changes, insulin resistance and Conde, “Influence of G308A promoter variant of tumor chronic diabetic complications in obese patients with type necrosis factor-α gene on insulin resistance and weight loss 2diabetes,”Archives of Medical Research,vol.37,no.6,pp. secondary to two hypocaloric diets: a randomized clinical 736–743, 2006. trial,” Archives of Medical Research, vol. 40, no. 1, pp. 36–41, [61] U. Helwig, D. Rubin, J. Kiosz et al., “The minor allele of the 2009. PPARγ2 Pro12Ala polymorphism is associated with lower [74] Y. T. Joffe, L. van der Merwe, M. Carstens et al., “Tumor postprandial TAG and insulin levels in non-obese healthy necrosis factor-α gene -308 G/A polymorphism modulates men,” British Journal of Nutrition, vol. 97, no. 5, pp. 847–854, the relationship between dietary fat intake, serum lipids, and 2007. obesity risk in black south African women,” The Journal of [62] J. E. Cecil, P. Watt, C. N. Palmer, and M. Hetherington, Nutrition, vol. 140, no. 5, pp. 901–907, 2010. “Energy balance and food intake: the role of PPARγ gene [75] E. Corpeleijn, L. Petersen, C. Holst et al., “Obesity-related polymorphisms,” Physiology & Behavior,vol.88,no.3,pp. polymorphisms and their associations with the ability to 227–233, 2006. regulate fat oxidation in obese Europeans: the NUGENOB [63]A.Milewicz,U.Tworowska-Bardzinska,˜ K. Dunajska, D. study,” Obesity, vol. 18, no. 7, pp. 1369–1377, 2010. Jedrzejuk,ˆ and F. Lwow, “Relationship of PPARgamma2 pol- [76]C.Razquin,J.A.Martinez,M.A.Martinez-Gonzalez,J. ymorphism with obesity ana metabolic syndrome in post- Fernandez-Crehuet,´ J. M. Santos, and A. Marti, “A mediter- menopausal Polish women,” Experimental and Clinical ranean diet rich in virgin olive oil may reverse the effects of Endocrinology and Diabetes, vol. 117, no. 10, pp. 628–632, the-174g/c il6 gene variant on 3-year body weight change,” 2009. Molecular Nutrition & Food Research, vol. 54, supplement 1, [64] P. W. Franks, K. A. Jablonski, L. Delahanty et al., “The pp. S75–S82, 2010. Pro12Ala variant at the peroxisome proliferator-activated [77] J. M. Fernandez-Real,´ M. Broch, J. Vendrell, C. Richart, receptor γ gene and change in obesity-related traits in the and W. Ricart, “Interleukin-6 gene polymorphism and lipid Diabetes Prevention Program,” Diabetologia, vol. 50, no. 12, abnormalities in healthy subjects,” The Journal of Clinical pp. 2451–2460, 2007. Endocrinology and Metabolism, vol. 85, no. 3, pp. 1334–1339, [65]M.K.Moon,Y.M.Cho,H.S.Jungetal.,“Geneticpoly- 2000. morphisms in peroxisome proliferator-activated receptor γ [78] E. Goyenechea, M. D. Parra, and J. A. Martınez,´ “Weight are associated with type 2 diabetes mellitus and obesity in regain after slimming induced by an energy-restricted Journal of Obesity 25

diet depends on interleukin-6 and peroxisome-proliferator- [93] L. J. Abraham and K. M. Kroeger, “Impact of the -308 TNF activated-receptor-γ2 gene polymorphisms,” British Journal promoter polymorphism on the transcriptional regulation of Nutrition, vol. 96, no. 5, pp. 965–972, 2006. of the TNF gene: relevance to disease,” Journal of Leukocyte [79] A. P. Reiner, M. M. Wurfel, L. A. Lange et al., “Polymor- Biology, vol. 66, no. 4, pp. 562–566, 1999. phisms of the IL1-receptor antagonist gene (IL1RN) are [94]A.G.Wilson,F.S.diGiovine,A.I.F.Blakemore,andG. associated with multiple markers of systemic inflammation,” W. Duff, “Single base polymorphism in the human tumour Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 28, no. necrosis factor alpha (TNFα) gene detectable by NcoI 7, pp. 1407–1412, 2008. restriction of PCR product,” Human Molecular Genetics,vol. [80]A.Yanagisawa,K.Suzuki,A.Kimura,Y.Ito,N.Hamajima, 1, no. 5, p. 353, 1992. ff β and T. Inoue, “Possible protective e ect of serum -carotene [95] M. M. Elahi, K. Asotra, B. M. Matata, and S. S. Mastana, levels on the association between interleukin-1B C-31T “Tumor necrosis factor alpha -308 gene locus promoter polymorphism and hypertension in a Japanese population,” polymorphism: an analysis of association with health and Clinical Nutrition, vol. 28, no. 2, pp. 198–202, 2009. disease,” Biochimica et Biophysica Acta, vol. 1792, no. 3, pp. [81] K. Kornman, J. Rogus, H. Roh-Schmidt et al., “Interleukin-1 163–172, 2009. genotype-selective inhibition of inflammatory mediators by [96] A. J. P. Smith and S. E. Humphries, “Cytokine and cytokine a botanical: a nutrigenetics proof of concept,” Nutrition,vol. receptor gene polymorphisms and their functionality,” 23, no. 11-12, pp. 844–852, 2007. Cytokine and Growth Factor Reviews, vol. 20, no. 1, pp. 43– [82]C.Cominetti,M.C.deBortoli,E.Purgattoetal.,“Asso- 59, 2009. ciations between glutathione peroxidase-1 Pro198Leu poly- [97] K. M. Kroeger, K. S. Carville, and L. J. Abraham, “The - morphism, selenium status, and DNA damage levels in obese α ff womenafterconsumptionofBrazilnuts,”Nutrition. In press. 308 tumor necrosis factor- promoter polymorphism e ects [83] C. Meplan,´ L. K. Crosley, F. Nicol et al., “Functional effects transcription,” Molecular Immunology, vol. 34, no. 5, pp. 391– of a common single-nucleotide polymorphism (GPX4c718t) 399, 1997. in the glutathione peroxidase 4 gene: interaction with sex,” [98] A. G. Wilson, J. A. Symons, T. L. Mcdowell, H. O. Mcdevitt, ff ff The American Journal of Clinical Nutrition,vol.87,no.4,pp. and G. W. Du ,“E ects of a polymorphism in the human α 1019–1027, 2008. tumor necrosis factor promoter on transcriptional activa- [84] C. Meplan,´ L. K. Crosley, F. Nicol et al., “Genetic polymor- tion,” Proceedings of the National Academy of Sciences of the phisms in the human selenoprotein P gene determine the United States of America, vol. 94, no. 7, pp. 3195–3199, 1997. response of selenoprotein markers to selenium supplementa- [99]K.M.Kroeger,J.H.Steer,D.A.Joyce,andL.J.Abraham, tion in a gender-specific manner (the SELGEN study),” The “Effects of stimulus and cell type on the expression of FASEB Journal, vol. 21, no. 12, pp. 3063–3074, 2007. the -308 tumour necrosis factor promoter polymorphism,” [85]G.S.Hotamisligil,N.S.Shargill,andB.M.Spiegelman, Cytokine, vol. 12, no. 2, pp. 110–119, 2000. “Adipose expression of tumor necrosis factor-α:directrolein [100] B. M. N. Brinkman, D. Zuijdgeest, E. L. Kaijzel, F. C. obesity-linked insulin resistance,” Science, vol. 259, no. 5091, Breedveld, and C. L. Verweij, “Relevance of the tumor pp. 87–91, 1993. necrosis factor alpha (TNFα) -308 promoter polymorphism [86]C.N.Lumeng,S.M.Deyoung,andA.R.Saltiel,“Macropha- in TNFα gene regulation,” Journal of Inflammation, vol. 46, gesblockinsulinactioninadipocytesbyalteringexpression no. 1, pp. 32–41, 1996. of signaling and glucose transport proteins,” American [101] F. Araujo,A.C.Pereira,G.F.Mota,M.doR.Latorre,J.E.´ Journal of Physiology, vol. 292, no. 1, pp. E166–E174, 2007. Krieger, and A. J. Mansur, “The influence of tumor necrosis [87] G. R. Steinberg, “Inflammation in obesity is the common factor -308 and C-reactive protein G1059C gene variants on link between defects in fatty acid metabolism and insulin serum concentration of C-reactive protein: evidence for an resistance,” Cell Cycle, vol. 6, no. 8, pp. 888–894, 2007. age-dependent association,” Clinica Chimica Acta, vol. 349, [88] E. Louis, D. Franchimont, A. Piron et al., “Tumour necrosis no. 1-2, pp. 129–134, 2004. α factor (TNF) gene polymorphism influences TNF- pro- [102] H. M. Lakka, T. A. Lakka, T. Rankinen et al., “The TNF-α duction in lipopolysaccharide (LPS)-stimulated whole blood G-308A polymorphism is associated with C-reactive protein cell culture in healthy humans,” Clinical and Experimental levels: the HERITAGE Family Study,” Vascular Pharmacology, Immunology, vol. 113, no. 3, pp. 401–406, 1998. vol. 44, no. 5, pp. 377–383, 2006. [89]M.W.vanderLinden,T.W.J.Huizinga,D.J.Stoeken,A. [103] M. L. Gander, J. E. Fischer, F. E. Maly, and R. von Kanel,¨ Sturk, and R. G. J. Westendorp, “Determination of tumour “Effect of the G-308A polymorphism of the tumor necrosis necrosis factor-α and interleukin-10 production in a whole factor (TNF)-α gene promoter site on plasma levels of TNF- blood stimulation system: assessment of laboratory error and α and C-reactive protein in smokers: a cross-sectional study,” individual variation,” Journal of Immunological Methods,vol. BMC Cardiovascular Disorders, vol. 4, article 17, 2004. 218, no. 1-2, pp. 63–71, 1998. [90] J. P. Bayley, H. de Rooij, P. J. van den Elsen, T. W. J. [104] S. C. Sookoian, C. Gonzalez,andC.J.Pirola,“Meta-´ α Huizinga, and C. L. Verweij, “Functional analysis of linker- analysis on the G-308A tumor necrosis factor gene variant scan mutants spanning the -376, -308, -244, and -238 and phenotypes associated with the metabolic syndrome,” polymorphic sites of the TNF-α promoter,” Cytokine,vol.14, Obesity Research, vol. 13, no. 12, pp. 2122–2131, 2005. no. 6, pp. 316–323, 2001. [105] T. V. Pereira, M. Rudnicki, R. F. Franco, A. C. Pereira, and [91]J.P.Bayley,T.H.M.Ottenhoff,andC.L.Verweij,“Is J. E. Krieger, “Effect of the G-308A polymorphism of the there a future for TNF promoter polymorphisms?” Genes & tumor necrosis factor α gene on the risk of ischemic heart Immunity, vol. 5, no. 5, pp. 315–329, 2004. disease and ischemic stroke: a meta-analysis,” American Heart [92]A.H.HajeerandI.V.Hutchinson,“TNF-α gene poly- Journal, vol. 153, no. 5, pp. 821–830, 2007. morphism: clinical and biological implications,” Microscopy [106] C. M. Phillips, L. Goumidi, S. Bertrais et al., “Additive Research and Technique, vol. 50, no. 3, pp. 216–228, 2000. effect of polymorphisms in the IL-6, LTA, and TNF-α genes 26 Journal of Obesity

and plasma fatty acid level modulate risk for the metabolic British Journal of Nutrition, vol. 101, no. 12, pp. 1745–1752, syndrome and its components,” The Journal of Clinical 2009. Endocrinology and Metabolism, vol. 95, no. 3, pp. 1386–1394, [120] A. Loktionov, S. Scollen, N. McKeown, and S. A. Bingham, 2010. “Gene-nutrient interactions: dietary behaviour associated [107] P. Gomez, P. Perez-Martinez, C. Marin et al., “APOA1 and with high coronary heart disease risk particularly affects APOA4 gene polymorphisms influence the effects of dietary serum LDL cholesterol in apolipoprotein E ε4-carrying free- fat on LDL particle size and oxidation in healthy young living individuals,” British Journal of Nutrition, vol. 84, no. 6, adults,” The Journal of Nutrition, vol. 140, no. 4, pp. 773–778, pp. 885–890, 2000. 2010. [121] E. Strandhagen, H. Zetterberg, N. Aires et al., “The apolipo- [108] J. M. Ordovas, D. Corella, L. A. Cupples et al., “Polyun- protein E polymorphism and the cholesterol-raising effect of saturated fatty acids modulate the effects of the APOA1 G- coffee,” Lipids in Health and Disease, vol. 3, article 26, 2004. A polymorphism on HDL-cholesterol concentrations in a [122] J. C. Knight, I. Udalova, A. V. S. Hill et al., “A polymorphism sex-specific manner: the Framingham study,” The American that affects OCT-1 binding to the TNF promoter region is Journal of Clinical Nutrition, vol. 75, no. 1, pp. 38–46, 2002. associated with severe malaria,” Nature Genetics, vol. 22, no. [109] P. Mata, J. Lopez-Miranda, M. Pocovi et al., “Human apol- 2, pp. 145–150, 1999. ipoprotein A-I gene promoter mutation influences plasma [123] A. Steensberg, C. P. Fischer, C. Keller, K. Møller, and B. K. low density lipoprotein cholesterol response to dietary fat Pedersen, “IL-6 enhances plasma IL-1ra, IL-10, and cortisol saturation,” Atherosclerosis, vol. 137, no. 2, pp. 367–376, 1998. in humans,” American Journal of Physiology, vol. 285, no. 2, [110] K. Duesing, G. Charpentier, M. Marre et al., “Evaluating pp. E433–E437, 2003. the association of common APOA2 variants with type 2 [124] R. Starkie, S. R. Ostrowski, S. Jauffred, M. Febbraio, and B. diabetes,” BMC Medical Genetics, vol. 10, article 13, 2009. K. Pedersen, “Exercise and IL-6 infusion inhibit endotoxin- [111] D. Corella, G. Peloso, D. K. Arnett et al., “APOA2, dietary fat, induced TNF-alpha production in humans,” The FASEB and body mass index: replication of a gene-diet interaction Journal, vol. 17, no. 8, pp. 884–886, 2003. in 3 independent populations,” Archives of Internal Medicine, [125] L. Ghazouani, N. Abboud, S. B. H. Khalifa et al., “-174G>C vol. 169, no. 20, pp. 1897–1906, 2009. interleukin-6 gene polymorphism in Tunisian patients with [112] D. Corella, E. S. Tai, J. V. Sorlı´ et al., “Association between coronary artery disease,” Annals of Saudi Medicine, vol. 31, the APOA2 promoter polymorphism and body weight in no. 1, pp. 40–44, 2011. Mediterranean and Asian populations: replication of a gene- [126] L. Qi, C. Zhang, R. M. van Dam, and F. B. Hu, “Interleukin-6 saturated fat interaction,” International Journal of Obesity, genetic variability and adiposity: associations in two prospec- vol. 35, no. 5, pp. 666–675, 2011. tive cohorts and systematic review in 26,944 individuals,” The [113] J. Delgado-Lista, F. Perez-Jimenez, T. Tanaka et al., “An Journal of Clinical Endocrinology and Metabolism,vol.92,no. apolipoprotein A-II polymorphism (-265T/C, rs5082) reg- 9, pp. 3618–3625, 2007. ulates postprandial response to a saturated fat overload in [127] K. M. Rexrode, A. Pradhan, J. E. Manson, J. E. Buring, and healthy men,” The Journal of Nutrition, vol. 137, no. 9, pp. P. M. Ridker, “Relationship of total and abdominal adiposity 2024–2028, 2007. with CRP and IL-6 in women,” Annals of Epidemiology,vol. [114] C. Q. Lai, D. Corella, S. Demissie et al., “Dietary intake of n- 13, no. 10, pp. 674–682, 2003. 6 fatty acids modulates effect of apolipoprotein A5 gene on [128] B. Vozarova, C. Weyer, K. Hanson, P. A. Tataranni, C. plasma fasting triglycerides, remnant lipoprotein concentra- Bogardus, and R. E. Pratley, “Circulating interleukin-6 in tions, and lipoprotein particle size: the Framingham Heart relation to adiposity, insulin action, and insulin secretion,” Study,” Circulation, vol. 113, no. 17, pp. 2062–2070, 2006. Obesity Research, vol. 9, no. 7, pp. 414–417, 2001. [115] J. Mattei, S. Demissie, K. L. Tucker, and J. M. Ordovas, “Apol- [129] J. W. Stephens, S. J. Hurel, G. D. O. Lowe, A. Rumley, and ipoprotein A5 polymorphisms interact with total dietary fat S. E. Humphries, “Association between plasma IL-6, the IL6 intake in association with markers of metabolic syndrome in -174G>C gene variant and the metabolic syndrome in type Puerto Rican older adults,” The Journal of Nutrition, vol. 139, 2 diabetes mellitus,” Molecular Genetics and Metabolism,vol. no. 12, pp. 2301–2308, 2009. 90, no. 4, pp. 422–428, 2007. [116] J. Y. Kim, O. Y. Kim, S. J. Koh et al., “Comparison of low-fat [130] N. Sattar, C. G. Perry, and J. R. Petrie, “Type 2 diabetes as meal and high-fat meal on postprandial lipemic response in an inflammatory disorder,” British Journal of Diabetes and non-obese men according to the -1131T>C polymorphism Vascular Disease, vol. 3, no. 1, pp. 36–41, 2003. of the apolipoprotein A5 (APOA5) gene (randomized cross- [131] D. J. Freeman, J. Norrie, M. J. Caslake et al., “C-reactive over design),” Journal of the American College of Nutrition, protein is an independent predictor of risk for the develop- vol. 25, no. 4, pp. 340–347, 2006. ment of diabetes in the west of Scotland coronary prevention [117] I. Sundl, M. Guardiola, G. Khoschsorur et al., “Increased study,” Diabetes, vol. 51, no. 5, pp. 1596–1600, 2002. concentrations of circulating vitamin E in carriers of the [132] S. E. Humphries, P. J. Talmud, E. Hawe, M. Bolla, I. N. M. apolipoprotein A5 gene -1131T>C variant and associations Day, and G. J. Miller, “Apolipoprotein E4 and coronary heart with plasma lipids and lipid peroxidation,” Journal of Lipid disease in middle-aged men who smoke: a prospective study,” Research, vol. 48, no. 11, pp. 2506–2513, 2007. The Lancet, vol. 358, no. 9276, pp. 115–119, 2001. [118] K. Wu, R. Bowman, A. A. Welch et al., “Apolipoprotein E [133] J. M. Fernandez-Real, M. Vayreda, C. Richart et al., “Circulat- polymorphisms, dietary fat and fibre, and serum lipids: the ing interleukin 6 levels, blood pressure, and insulin sensitivity EPIC Norfolk study,” European Heart Journal, vol. 28, no. 23, in apparently healthy men and women,” The Journal of pp. 2930–2936, 2007. Clinical Endocrinology and Metabolism, vol. 86, no. 3, pp. [119] J. A. Moreno, F. Perez-Jimn´ ez,´ R. Moreno-Luna et al., “The 1154–1159, 2001. effect of apoE genotype and sex on ApoE plasma concen- [134] J. W. Stephens, S. J. Hurel, J. A. Cooper, J. Acharya, G. tration is determined by dietary fat in healthy subjects,” J.Miller,andS.E.Humphries,“Acommonfunctional Journal of Obesity 27

variant in the interleukin-6 gene is associated with increased associated with type-2 diabetes mellitus in Native Americans body mass index in subjects with type 2 diabetes mellitus,” and Caucasians,” Human Genetics, vol. 112, no. 4, pp. 409– Molecular Genetics and Metabolism, vol. 82, no. 2, pp. 180– 413, 2003. 186, 2004. [149] D. Fishman, G. Faulds, R. Jeffey et al., “The effect of novel [135] P. C. Heinrich, J. V. Castell, and T. Andus, “Interleukin-6 and polymorphisms in the interleukin-6 (IL-6) gene on IL-6 the acute phase response,” Biochemical Journal, vol. 265, no. transcription and plasma IL-6 levels, and an association 3, pp. 621–636, 1990. with systemic-onset juvenile chronic arthritis,” The Journal [136] M. T. Berthier, A. M. Paradis, A. Tchernof et al., “The of Clinical Investigation, vol. 102, no. 7, pp. 1369–1376, 1998. interleukin 6 -174G/C polymorphism is associated with [150] R. Bouhaha, T. Baroudi, H. Ennafaa et al., “Study of indices of obesity in men,” Journal of Human Genetics,vol. TNFα-308G/A and IL6 -174G/C polymorphisms in type 2 48, no. 1, pp. 14–19, 2003. diabetes and obesity risk in the Tunisian population,” Clinical [137] M. Cardellini, L. Perego, M. D’Adamo et al., “C-174G Biochemistry, vol. 43, no. 6, pp. 549–552, 2010. polymorphism in the promoter of the interleukin-6 gene is [151] D. J. Brull, H. E. Montgomery, J. Sanders et al., “Interleukin- associated with insulin resistance,” Diabetes Care, vol. 28, no. 6 gene -174G>C and -572G>C promoter polymorphisms 8, pp. 2007–2012, 2005. are strong predictors of plasma interleukin-6 levels after [138] H. Grallert, C. Huth, M. Kolz et al., “IL-6 promoter poly- coronary artery bypass surgery,” Arteriosclerosis, Thrombosis, morphisms and quantitative traits related to the metabolic and Vascular Biology, vol. 21, no. 9, pp. 1458–1463, 2001. syndrome in KORA S4,” Experimental Gerontology, vol. 41, [152] K. G. Jones, D. J. Brull, L. C. Brown et al., “Interleukin-6 no. 8, pp. 737–745, 2006. (IL-6) and the prognosis of abdominal aortic aneurysms,” [139] Y. H. Hamid, C. S. Rose, S. A. Urhammer et al., “Variations Circulation, vol. 103, no. 18, pp. 2260–2265, 2001. of the interleukin-6 promoter are associated with features of [153] C. Eklund, A. Nenonen, K. Kukkonen-Harjula et al., “Asso- the metabolic syndrome in Caucasian Danes,” Diabetologia, ciation of the IL6-174(G/C) polymorphism with C-reactive vol. 48, no. 2, pp. 251–260, 2005. protein concentration after weight loss in obese men,” [140] K. Klipstein-Grobusch, M. Mohlig,¨ J. Spranger et al., European Cytokine Network, vol. 17, no. 2, pp. 131–135, 2006. > “Interleukin-6 g.-174G C promoter polymorphism is asso- [154] A. Moleres, T. Rendo-Urteaga, C. Azcona et al., “IL6 gene ciated with obesity in the EPIC-potsdam study,” Obesity,vol. promoter polymorphism (-174G/C) influences the associ- 14, no. 1, pp. 14–18, 2006. ation between fat mass and cardiovascular risk factors,” [141] I. Wernstedt, A.-L. Eriksson, A. Berndtsson et al., “A common Journal of Physiology and Biochemistry, vol. 65, no. 4, pp. 405– polymorphism in the interleukin-6 gene promoter is associ- 413, 2009. ated with overweight,” International Journal of Obesity,vol. [155] M. T. Berthier, A. Houde, M. Cotˆ eetal.,“Impactofad-´ 28, no. 10, pp. 1272–1279, 2004. iponectin gene polymorphisms on plasma lipoprotein and [142] E. Goyenechea, M. D. Parra, and J. A. M. Hernandez,´ “Role of adiponectin concentrations of viscerally obese men,” Journal IL-6 and its -174G>C polymorphism in weight management of Lipid Research, vol. 46, no. 2, pp. 237–244, 2005. and in the metabolic comorbidities associated with obesi- [156] S. B. Ng, Y. H. Tan, and G. R. Guy, “Differential induction tyImplicacion´ de la IL-6 y su polimorfismo -174G>Cenel of the interleukin-6 gene by tumor necrosis factor and control del peso corporal y en las complicaciones metabolicas´ interleukin-1,” The Journal of Biological Chemistry, vol. 269, asociadas a la obesidad,” Anales del Sistema Sanitario de no. 29, pp. 19021–19027, 1994. Navarra, vol. 28, no. 3, pp. 357–366, 2005. [157] D. Zhang, S. L. Jiang, D. Rzewnicki, D. Samols, and I. [143] N. Andersson, L. Strandberg, S. Nilsson et al., “A variant ff near the interleukin-6 gene is associated with fat mass in Kushner, “The e ect of interleukin-1 on C-reactive protein Caucasian men,” International Journal of Obesity, vol. 34, no. expression in Hep3B cells is exerted at the transcriptional 6, pp. 1011–1019, 2010. level,” Biochemical Journal, vol. 310, no. 1, pp. 143–148, 1995. ff [144] H. Y. Huang, L. Thuita, P. Strickland, S. C. Hoffman, G. [158] W. Khovidhunkit, M. S. Kim, R. A. Memon et al., “E ects W. Comstock, and K. J. Helzlsouer, “Frequencies of single of infection and inflammation on lipid and lipoprotein nucleotide polymorphisms in genes regulating inflammatory metabolism: mechanisms and consequences to the host,” responses in a community-based population,” BMC Genetics, Journal of Lipid Research, vol. 45, no. 7, pp. 1169–1196, 2004. vol. 8, article 7, 2007. [159] J. Jager, T. Gremeaux,´ M. Cormont, Y. Le Marchand-Brustel, [145] C. Poitou, J.-M. Lacorte, M. Coupaye et al., “Relationship and J.-F. Tanti, “Interleukin-1β-induced insulin resistance between single nucleotide polymorphisms in leptin, IL6 and in adipocytes through down-regulation of insulin receptor adiponectin genes and their circulating product in morbidly substrate-1 expression,” Endocrinology, vol. 148, no. 1, pp. obese subjects before and after gastric banding surgery,” 241–251, 2007. Obesity Surgery, vol. 15, no. 1, pp. 11–23, 2005. [160] M. J. H. Nicklin, J. L. Barton, M. Nguyen, M. G. FitzGerald, [146] J. A. McKenzie, E. P. Weiss, I. A. Ghiu et al., “Influence of G. W. Duff, and K. Kornman, “A sequence-based map of the the interleukin-6 -174 G/C gene polymorphism on exercise nine genes of the human interleukin-1 cluster,” Genomics, training-induced changes in glucose tolerance indexes,” vol. 79, no. 5, pp. 718–725, 2002. Journal of Applied Physiology, vol. 97, no. 4, pp. 1338–1342, [161] K. S. Kornman, “Interleukin 1 genetics, inflammatory mech- 2004. anisms, and nutrigenetic opportunities to modulate diseases [147] A. Kubaszek, J. Pihlajamaki,¨ K. Punnonen, P. Karhapaä,¨ of aging,” The American Journal of Clinical Nutrition, vol. 83, I. Vauhkonen, and M. Laakso, “The C-174G promoter no. 2, pp. 475S–483S, 2006. polymorphism of the IL-6 gene affects energy expenditure [162] G. Latkovskis, N. Licis, and U. Kalnins, “C-reactive protein and insulin sensitivity,” Diabetes, vol. 52, no. 2, pp. 558–561, levels and common polymorphisms of the interleukin-1 gene 2003. cluster and interleukin-6 gene in patients with coronary heart [148] B. Vozarova, J. M. Fernandez-Real,´ W. C. Knowler et al., disease,” European Journal of Immunogenetics, vol. 31, no. 5, “The interleukin-6 (-174) G/C promoter polymorphism is pp. 207–213, 2004. 28 Journal of Obesity

[163] P. Berger, J. P. McConnell, M. Nunn et al., “C-reactive protein cholesterol in French Canadians,” Atherosclerosis, vol. 185, levels are influenced by common IL-1 gene variations,” no. 1, pp. 127–136, 2006. Cytokine, vol. 17, no. 4, pp. 171–174, 2002. [178] G. Franceschini, C. R. Sirtori, A. Capurso et al., “A-IMilano [164] L. Iacoviello, A. Di Castelnuovo, M. Gattone et al., “Poly- apoprotein. Decreased high density lipoprotein cholesterol morphisms of the interleukin-1β gene affect the risk of levels with significant lipoprotein modifications and without myocardial infarction and ischemic stroke at young age and clinical atherosclerosis in an Italian family,” The Journal of the response of mononuclear cells to stimulation in vitro,” Clinical Investigation, vol. 66, no. 5, pp. 892–900, 1980. Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 25, no. [179] P. Roma, R. E. Gregg, M. S. Meng et al., “In vivo metabolism 1, pp. 222–227, 2005. of a mutant form of apolipoprotein A-I, apo A-IMilano, associ- [165] S. Rafiq, K. Stevens, A. J. Hurst et al., “Common genetic vari- ated with familial hypoalphalipoproteinemia,” The Journal of ation in the gene encoding interleukin-1-receptor antagonist Clinical Investigation, vol. 91, no. 4, pp. 1445–1452, 1993. (IL-1RA) is associated with altered circulating IL-1RA levels,” [180] K. H. Weisgraber, S. C. Rall Jr., T. P.Bersot et al., “Apolipopro- Genes & Immunity, vol. 8, no. 4, pp. 344–351, 2007. tein A-IMilano. Detection of normal A-I in affected subjects [166] J. Shen, D. K. Arnett, J. M. Peacock et al., “Interleukin1β and evidence for a cysteine for arginine substitution in the genetic polymorphisms interact with polyunsaturated fatty variant A-I,” The Journal of Biological Chemistry, vol. 258, no. acids to modulate risk of the metabolic syndrome,” The 4, pp. 2508–2513, 1983. Journal of Nutrition, vol. 137, no. 8, pp. 1846–1851, 2007. [181] L. Calabresi, M. Cassinotti, G. Gianfranceschi et al., “In- [167] J. M. Kuyl and D. Mendelsohn, “Observed relationship creased postprandial lipemia in apo A-IMilano carriers,” Arte- between ratios HDL-cholesterol/total cholesterol and apol- riosclerosis and Thrombosis, vol. 13, no. 4, pp. 521–528, 1993. ipoprotein A1/apolipoprotein B,” Clinical Biochemistry,vol. [182] Q. Dan, Y. Huang, and Z. Peng, “DNA polymorphisms of the 25, no. 5, pp. 313–316, 1992. apolipoprotein AI-CIII-AIV gene cluster in coronary artery [168] P. Perez-Martinez, J. Lopez-Miranda, F. Perez-Jimenez, and J. disease,” Zhonghua Yi Xue Za Zhi, vol. 75, no. 10, pp. 584– M. Ordovas, “Influence of genetic factors in the modulation 637, 1995. of postprandial lipemia,” Atherosclerosis Supplements,vol.9, [183] N. A. Vavatsi, S. A. Kouidou, P. N. Geleris et al., “Increased no. 2, pp. 49–55, 2008. frequency of the rare PstI allele (P2) in a population of CAD [169] M. Sorci-Thomas, M. M. Prack, N. Dashti, F. Johnson, L. L. patients in Northern Greece,” Clinical Genetics, vol. 47, no. 1, Rudel, and D. L. Williams, “Differential effects of dietary fat pp. 22–26, 1995. on the tissue-specific expression of the apolipoprotein A-I [184] M. Jeenah, A. Kessling, N. Miller, and S. Humphries, “G to gene: relationship to plasma concentration of high density A substitution in the promoter region of the apolipoprotein lipoproteins,” Journal of Lipid Research,vol.30,no.9,pp. AI gene is associated with elevated serum apolipoprotein 1397–1403, 1989. AI and high density lipoprotein cholesterol concentrations,” [170] M. Marcil, A. Brooks-Wilson, S. M. Clee et al., “Mutations Molecular Biology and Medicine, vol. 7, no. 3, pp. 233–241, in the ABC1 gene in familial HDL deficiency with defective 1990. cholesterol efflux,” The Lancet, vol. 354, no. 9187, pp. 1341– [185] F. Pagani, A. Sidoli, G. A. Giudici, L. Barenghi, C. Vergani, 1346, 1999. and F. E. Baralle, “Human apolipoprotein A-I gene promoter [171] E. J. Schaefer, “Familial lipoprotein disorders and premature polymorphism: association with hyperalphalipoproteine- coronary artery disease,” The Medical Clinics of North mia,” Journal of Lipid Research, vol. 31, no. 8, pp. 1371–1377, America, vol. 78, no. 1, pp. 21–39, 1994. 1990. [172] L. Pisciotta, R. Miccoli, A. Cantafora et al., “Recurrent [186] C. Marın,J.L´ opez-Miranda,´ P. Gomez´ et al., “Effects of mutations of the apolipoprotein A-I gene in three kindreds the human apolipoprotein A-I promoter G-A mutation on with severe HDL deficiency,” Atherosclerosis, vol. 167, no. 2, postprandial lipoprotein metabolism,” The American Journal pp. 335–345, 2003. of Clinical Nutrition, vol. 76, no. 2, pp. 319–325, 2002. [173] K. Ikewaki, A. Matsunaga, H. Han et al., “A novel two [187] N. Saha, J. S. H. Tay, P. S. Low, and S. E. Humphries, nucleotide deletion in the apolipoprotein A-I gene, apoA- “Guanidine to adenine (G/A) substitution in the promoter I Shinbashi, associated with high density lipoprotein defi- region of the apolipoprotein AI gene is associated with ciency, corneal opacities, planar xanthomas, and premature elevated serum apolipoprotein AI levels in Chinese non- coronary artery disease,” Atherosclerosis, vol. 172, no. 1, pp. smokers,” Genetic Epidemiology, vol. 11, no. 3, pp. 255–264, 39–45, 2004. 1994. [174] H. Yokota, Y. Hashimoto, S. Okubo et al., “Apolipoprotein [188] M. Groenendijk, R. M. Cantor, T. W. A. de Bruin, and A-I deficiency with accumulated risk for CHD but no G. M. Dallinga-Thie, “The apoAI-CIII-AIV gene cluster,” symptoms of CHD,” Atherosclerosis, vol. 162, no. 2, pp. 399– Atherosclerosis, vol. 157, no. 1, pp. 1–11, 2001. 407, 2002. [189] L. Noriega-Lopez,A.R.Tovar,M.Gonzalez-Granilloetal.,´ [175] E. M. Rubin, R. M. Krauss, E. A. Spangler, J. G. Verstuyft, and “Pancreatic insulin secretion in rats fed a soy protein high S. M. Clift, “Inhibition of early atherogenesis in transgenic fat diet depends on the interaction between the amino acid mice by human apolipoprotein A1,” Nature, vol. 353, no. pattern and isoflavones,” The Journal of Biological Chemistry, 6341, pp. 265–267, 1991. vol. 282, no. 28, pp. 20657–20666, 2007. [176] P. N. Herbert, G. Assmann, J. A. M. Grotton, and [190] C. Ascencio, N. Torres, F. Isoard-Acosta, F. J. Gomez-P´ erez,´ D. S. Fredrickson, “Disorders of lipid and lipoprotein R. Hernandez-Pando,´ and A. R. Tovar, “Soy protein affects metabolism,” in Metabolic Basis of Inherited Disease,J.B. serum insulin and hepatic SREBP-1 mRNA and reduces fatty Stanbury, D. S. Wyngaarden, D. S. Fredrickson et al., Eds., liver in rats,” The Journal of Nutrition, vol. 134, no. 3, pp. 522– pp. 589–651, McGraw-Hill, New York, NY, USA, 1999. 529, 2004. [177] Z. Dastani, C. Dangoisse, B. Boucher et al., “A novel nonsense [191] J. D. Horton, J. L. Goldstein, and M. S. Brown, “SREBPs: apolipoprotein A-I mutation (apoA-I) causes low HDL activators of the complete program of cholesterol and Journal of Obesity 29

fatty acid synthesis in the liver,” The Journal of Clinical an early phase of liver regeneration,” The Journal of Biological Investigation, vol. 109, no. 9, pp. 1125–1131, 2002. Chemistry, vol. 276, no. 48, pp. 44512–44520, 2001. [192] N. Torres, M. Guevara-Cruz, J. Granados et al., “Reduction of [206] F. G. Schaap, P. C. N. Rensen, P. J. Voshol et al., “ApoAV serum lipids by soy protein and soluble fiber is not associated reduces plasma triglycerides by inhibiting very low density with the ABCG5/G8, apolipoprotein E, and apolipoprotein lipoprotein-triglycerides (VLDL-TG) production and stimu- A1 polymorphisms in a group of hyperlipidemic Mexican lating lipoprotein lipase-mediated VLDL-TG hydrolysis,” The subjects,” Nutrition Research, vol. 29, no. 10, pp. 728–735, JournalofBiologicalChemistry, vol. 279, no. 27, pp. 27941– 2009. 27947, 2004. [193] K. E. Berge, H. Tian, G. A. Graf et al., “Accumulation of [207] J. Fruchart-Najib, E. Bauge,´ L. S. Niculescu et al., “Mech- dietary cholesterol in sitosterolemia caused by mutations in anism of triglyceride lowering in mice expressing human adjacent ABC transporters,” Science, vol. 290, no. 5497, pp. apolipoprotein A5,” Biochemical and Biophysical Research 1771–1775, 2000. Communications, vol. 319, no. 2, pp. 397–404, 2004. [194] G. Fager, O. Wiklund, S. O. Olofsson, L. Wilhelmsen, and [208] L. A. Pennacchio, M. Olivier, J. A. Hubacek, R. M. Krauss, E. G. Bondjers, “Multivariate analyses of serum apolipoproteins M.Rubin,andJ.C.Cohen,“Twoindependentapolipopro- and risk factors in relation to acute myocardial infarction,” tien A5 haplotypes influence human plasma triglyceride Arteriosclerosis, vol. 1, no. 4, pp. 273–279, 1981. levels,” Human Molecular Genetics, vol. 11, no. 24, pp. 3031– [195] J. E. Buring, G. T. O’Connor, S. Z. Goldhaber et al., 3038, 2002. “Decreased HDL and HDL cholesterol, apo A-I and apo A- [209] B. E. Aouizerat, M. Kulkarni, D. Heilbron et al., “Genetic II, and increased risk of myocardial infarction,” Circulation, analysis of a polymorphism in the human apoA-V gene: ff vol. 85, no. 1, pp. 22–29, 1992. e ect on plasma lipids,” Journal of Lipid Research,vol.44,no. [196] P. M. Ridker, N. Rifai, N. R. Cook, G. Bradwin, and J. E. 6, pp. 1167–1173, 2003. Buring, “Non-HDL cholesterol, apolipoproteins A-I and B, [210] R. Moreno, F. Perez-Jimenez, C. Marin et al., “A single standard lipid measures, lipid ratios, and CRP as risk factors nucleotide polymorphism of the apolipoprotein A-V gene - > for cardiovascular disease in women,” Journal of the American 1131T C modulates postprandial lipoprotein metabolism,” Medical Association, vol. 294, no. 3, pp. 326–333, 2005. Atherosclerosis, vol. 189, no. 1, pp. 163–168, 2006. [211] T. Zhao and J. Zhao, “Association of the apolipoprotein A5 [197] A. Tailleux, P. Duriez, J. C. Fruchart, and V. Clavey, > “Apolipoprotein A-II, HDL metabolism and atherosclerosis,” gene -1131 T C polymorphism with fasting blood lipids: a Atherosclerosis, vol. 164, no. 1, pp. 1–13, 2002. meta-analysis in 37859 subjects,” BMC Medical Genetics,vol. 11, no. 1, article 120, pp. 1–12, 2010. [198] J. C. Escola-Gil,´ A. Marzal-Casacuberta, J. Julve-Gil et al., [212] M. A. Austin, P. J. Talmud, F. M. Farin et al., “Association “Human apolipoprotein A-II is a pro-atherogenic molecule of apolipoprotein A5 variants with LDL particle size and when it is expressed in transgenic mice at a level similar to triglyceride in Japanese Americans,” Biochimica et Biophysica that in humans: evidence of a potentially relevant species- Acta, vol. 1688, no. 1, pp. 1–9, 2004. specific interaction with diet,” Journal of Lipid Research,vol. 39, no. 2, pp. 457–462, 1998. [213] C. Szalai, M. Keszei, J. Duba et al., “Polymorphism in the promoter region of the apolipoprotein A5 gene is associated [199] Z. Han, S. C. Heath, D. Shmulewitz et al., “Candidate genes with an increased susceptibility for coronary artery disease,” involved in cardiovascular risk factors by a family-based Atherosclerosis, vol. 173, no. 1, pp. 109–114, 2004. association study on the island of Kosrae, Federated States of [214] J. Dallongeville, D. Cottel, M. Montaye, V. Codron, P. Micronesia,” American Journal of Medical Genetics, vol. 110, Amouyel, and N. Helbecque, “Impact of APOA5/A4/C3 no. 3, pp. 234–242, 2002. genetic polymorphisms on lipid variables and cardiovascular [200] F. M. Van’t Hooft, G. Ruotolo, S. Boquist, U. De Faire, disease risk in French men,” International Journal of Cardiol- G. Eggertsen, and A. Hamsten, “Human evidence that the ogy, vol. 106, no. 2, pp. 152–156, 2006. apolipoprotein A-II gene is implicated in visceral fat accu- [215] D. Corella, C. Q. Lai, S. Demissie et al., “APOA5 gene mulation and metabolism of triglyceride-rich lipoproteins,” variation modulates the effects of dietary fat intake on body Circulation, vol. 104, no. 11, pp. 1223–1228, 2001. mass index and obesity risk in the Framingham Heart Study,” [201] D. Takada, M. Emi, Y. Ezura et al., “Interaction between the Journal of Molecular Medicine, vol. 85, no. 2, pp. 119–128, LDL-receptor gene bearing a novel mutation and a variant in 2007. the apolipoprotein a-ii promoter: molecular study in a 1135- [216] R. W. Mahley, “Apolipoprotein E: cholesterol transport member familial hypercholesterolemia kindred,” Journal of protein with expanding role in cell biology,” Science, vol. 240, Human Genetics, vol. 47, no. 12, pp. 656–664, 2002. no. 4852, pp. 622–630, 1988. [202] C. Lara-Castro, G. R. Hunter, J. C. Lovejoy, B. A. Gower, and [217] S. C. Rall Jr., K. H. Weisgraber, and R. W. Mahley, “Human J. R. Fernandez,´ “Apolipoprotein A-II polymorphism and apolipoprotein E. The complete amino acid sequence,” The visceral adiposity in African-American and white women,” JournalofBiologicalChemistry, vol. 257, no. 8, pp. 4171– Obesity Research, vol. 13, no. 3, pp. 507–512, 2005. 4178, 1982. [203] P. J. Talmud and S. E. Humphries, “Genetic polymorphisms, [218] R. W. Mahley and S. C. Rall Jr., “Apolipoprotein E: far more lipoproteins and coronary artery disease risk,” Current than a lipid transport protein,” Annual Review of Genomics Opinion in Lipidology, vol. 12, no. 4, pp. 405–409, 2001. and Human Genetics, vol. 1, no. 2000, pp. 507–537, 2000. [204] L. A. Pennacchio, M. Olivier, J. A. Hubacek et al., “An apol- [219] R. W. Mahley and Z. S. Ji, “Remnant lipoprotein metabolism: ipoprotein influencing triglycerides in humans and mice key pathways involving cell-surface heparan sulfate proteo- revealed by comparative sequencing,” Science, vol. 294, no. glycans and apolipoprotein E,” Journal of Lipid Research,vol. 5540, pp. 169–173, 2001. 40, no. 1, pp. 1–16, 1999. [205] H. N. van der Vliet, M. G. Sammels, A. C. J. Leegwater et al., [220] K. Taira, H. Bujo, S. Hirayama et al., “LR11, a mosaic “Apolipoprotein A-V: a novel apolipoprotein associated with LDL receptor family member, mediates the uptake of 30 Journal of Obesity

apoE-rich lipoproteins in vitro,” Arteriosclerosis, Thrombosis, [234] D. M. Dreon, H. A. Fernstrom, B. Miller, and R. M. Krauss, and Vascular Biology, vol. 21, no. 9, pp. 1501–1506, 2001. “Apolipoprotein E isoform phenotype and LDL subclass [221] M. Xhignesse, S. Lussier-Cacan, C. F. Sing, A. M. Kessling, response to a reduced-fat diet,” Arteriosclerosis, Thrombosis, and J. Davignon, “Influences of common variants of apol- and Vascular Biology, vol. 15, no. 1, pp. 105–111, 1995. ipoprotein E on measures of lipid metabolism in a sample [235] F. F. Chu, R. S. Esworthy, P. G. Chu et al., “Bacteria-induced selected for health,” Arteriosclerosis and Thrombosis,vol.11, intestinal cancer in mice with disrupted Gpx1 and Gpx2 no. 4, pp. 1100–1110, 1991. genes,” Cancer Research, vol. 64, no. 3, pp. 962–968, 2004. [222] Y. Huang, X. Q. Liu, S. C. Rail Jr., and R. W. Mahley, [236] R. Brigelius-Flohe,´ “Tissue-specific functions of individual “Apolipoprotein E2 reduces the low density lipoprotein level glutathione peroxidases,” Free Radical Biology & Medicine, in transgenic mice by impairing lipoprotein lipase-mediated vol. 27, no. 9-10, pp. 951–965, 1999. lipolysis of triglyceride-rich lipoproteins,” The Journal of [237] Q. Ran, H. Liang, M. Gu et al., “Transgenic mice over- Biological Chemistry, vol. 273, no. 28, pp. 17483–17490, 1998. expressing glutathione peroxidase 4 are protected against [223] Y. Huang, X. Q. Liu, S. C. Rall Jr. et al., “Overexpression and oxidative stress-induced apoptosis,” The Journal of Biological accumulation of apolipoprotein E as a cause of hypertriglyc- Chemistry, vol. 279, no. 53, pp. 55137–55146, 2004. eridemia,” The Journal of Biological Chemistry, vol. 273, no. [238] H. Imai, K. Narashima, M. Arai, H. Sakamoto, N. Chiba, and 41, pp. 26388–26393, 1998. Y. Nakagawa, “Suppression of leukotriene formation in RBL- [224] P. C. N. Rensen and T. J. C. van Berkel, “Apolipoprotein 2H3 cells that overexpressed phospholipid hydroperoxide Eeffectively inhibits lipoprotein lipase-mediated lipolysis of glutathione peroxidase,” The Journal of Biological Chemistry, chylomicron-like triglyceride-rich lipid emulsions in vitro vol. 273, no. 4, pp. 1990–1997, 1998. and in vivo,” The Journal of Biological Chemistry, vol. 271, no. [239] S. Villette, J. A. M. Kyle, K. M. Brown et al., “A novel single 25, pp. 14791–14799, 1996. nucleotide polymorphism in the 3 untranslated region of [225] Y. Huang, A. von Eckardstein, S. Wu, N. Maeda, and G. Ass- human glutathione peroxidase 4 influences lipoxygenase mann, “A plasma lipoprotein containing only apolipoprotein metabolism,” Blood Cells, Molecules, & Diseases, vol. 29, no. Eandwithγ mobility on electrophoresis releases cholesterol 2, pp. 174–178, 2002. from cells,” Proceedings of the National Academy of Sciences of [240] T. C. Stadtman, “Selenocysteine,” Annual Review of Biochem- the United States of America, vol. 91, no. 5, pp. 1834–1838, istry, vol. 65, pp. 83–100, 1996. 1994. [241] D. L. Hatfield and V. N. Gladyshev, “How selenium has [226] D. Y. Hui and J. A. Harmony, “Inhibition of Ca2+ accumula- altered our understanding of the genetic code,” Molecular and tion in mitogen-activated lymphocytes: role of membrane- Cellular Biology, vol. 22, no. 11, pp. 3565–3576, 2002. bound plasma lipoproteins,” Proceedings of the National [242] G. V. Kryukov, S. Castellano, S. V. Novoselov et al., “Charac- Academy of Sciences of the United States of America, vol. 77, terization of mammalian selenoproteomes,” Science, vol. 300, no. 8, pp. 4764–4768, 1980. no. 5624, pp. 1439–1443, 2003. [227] T. A. Miettinen, H. Gylling, H. Vanhanen, and A. Ollus, [243] M. J. Berry, L. Banu, Y. Chen et al., “Recognition of UGA as a “Cholesterol absorption, elimination, and synthesis related to selenocysteine codon in type I deiodinase requires sequences LDL kinetics during varying fat intake in men with different in the 3’ untranslated region,” Nature, vol. 353, no. 6341, pp. apoprotein E phenotypes,” Arteriosclerosis and Thrombosis, 273–276, 1991. vol. 12, no. 9, pp. 1044–1052, 1992. [244] J. Hesketh, “Nutrigenomics and selenium: gene expression [228] A. Loktionov, H. Vorster, I. K. O’Neill et al., “Apolipoprotein patterns, physiological targets, and genetics,” Annual Review E and methylenetetrahydrofolate reductase genetic polymor- of Nutrition, vol. 28, pp. 157–177, 2008. phisms in relation to other risk factors for cardiovascu- [245] R. F. Burk and K. E. Hill, “Selenoprotein P. A selenium-rich lar disease in UK Caucasians and Black South Africans,” extracellular glycoprotein,” The Journal of Nutrition, vol. 124, Atherosclerosis, vol. 145, no. 1, pp. 125–135, 1999. no. 10, pp. 1891–1897, 1994. [229] J. Davignon, R. E. Gregg, and C. F. Sing, “Apolipoprotein [246] K. E. Hill, J. Zhou, W. J. McMahan et al., “Deletion of E polymorphism and atherosclerosis,” Arteriosclerosis,vol.8, selenoprotein P alters distribution of selenium in the mouse,” no. 1, pp. 1–21, 1988. The Journal of Biological Chemistry, vol. 278, no. 16, pp. [230] G. Utermann, K. H. Vogelberg, A. Steinmetz et al., “Polymor- 13640–13646, 2003. phism of apolipoprotein E. II. Genetics of hyperlipoproteine- [247] L. Schomburg, U. Schweizer, B. Holtmann, L. Flohe,´ M. mia type III,” Clinical Genetics, vol. 15, no. 1, pp. 37–62, 1979. Sendtner, and J. Kohrle,¨ “Gene disruption discloses role [231] J. A. Moreno, J. Lopez-Miranda,´ C. Marın´ et al., “The influ- of selenoprotein P in selenium delivery to target tissues,” ence of the apolipoprotein E gene promoter (-219G/T) poly- Biochemical Journal, vol. 370, no. 2, pp. 397–402, 2003. morphism on postprandial lipoprotein metabolism in young [248] R. F. Burk and K. E. Hill, “Selenoprotein P: an extracellular normolipemic males,” Journal of Lipid Research,vol.44,no. protein with unique physical characteristics and a role in 11, pp. 2059–2064, 2003. selenium homeostasis,” Annual Review of Nutrition, vol. 25, [232] R. O. Alvim, S. R. Freitas, N. E. Ferreira et al., “APOE pol- pp. 215–235, 2005. ymorphism is associated with lipid profile, but not with [249] Y. Kabuyama, K. Oshima, T. Kitamura et al., “Involvement arterial stiffness in the general population,” Lipids in Health of selenoprotein P in the regulation of redox balance and and Disease, vol. 9, article 128, pp. 1–7, 2010. myofibroblast viability in idiopathic pulmonary fibrosis,” [233] F. Cardona, S. Morcillo, M. Gonzalo-Marin, and F. J. Genes to Cells, vol. 12, no. 11, pp. 1235–1244, 2007. Tinahones, “The apolipoprotein E genotype predicts post- [250] H. Steinbrenner, L. Alili, E. Bilgic, H. Sies, and P. Brenneisen, prandial hypertriglyceridemia in patients with the metabolic “Involvement of selenoprotein P in protection of human syndrome,” The Journal of Clinical Endocrinology and astrocytes from oxidative damage,” Free Radical Biology & Metabolism, vol. 90, no. 5, pp. 2972–2975, 2005. Medicine, vol. 40, no. 9, pp. 1513–1523, 2006. Journal of Obesity 31

[251] H. Steinbrenner, E. Bilgic, L. Alili, H. Sies, and P. Brenneisen, “Selenoprotein P protects endothelial cells from oxidative damage by stimulation of glutathione peroxidase expression and activity,” Free Radical Research, vol. 40, no. 9, pp. 936– 943, 2006. [252] Y. Zhang and X. Chen, “Reducing selenoprotein P expression suppresses adipocyte diﬀerentiation as a result of increased preadipocyte inﬂammation,” American Journal of Physiology, vol. 300, no. 1, pp. E77–E85, 2011. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 168659, 17 pages doi:10.1155/2011/168659

Review Article Variations in Adipokine Genes AdipoQ, Lep,andLepR are Associated with Risk for Obesity-Related Metabolic Disease: The Modulatory Role of Gene-Nutrient Interactions

Jennifer Emily Enns,1 Carla G. Taylor,2 and Peter Zahradka3

1 Department of Physiology, University of Manitoba and Canadian Centre for Agri-Food Research in Health and Medicine, St. Boniface Hospital Research Centre, 351 Tache Ave, Winnipeg, Manitoba, Canada R2H 2A6 2 Departments of Human Nutritional Sciences and Physiology, University of Manitoba, 351 Tache Ave, Winnipeg, Manitoba, Canada R2H 2A6 3 Departments of Physiology and Human Nutritional Sciences, University of Manitoba and Canadian Centre for Agri-Food Research in Health and Medicine, St. Boniface Hospital Research Centre, 351 Tache Ave, Winnipeg, Manitoba, Canada R2H 2A6

Correspondence should be addressed to Peter Zahradka, [email protected]

Received 9 January 2011; Accepted 10 March 2011

Academic Editor: P. Trayhurn

Copyright © 2011 Jennifer Emily Enns et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Obesity rates are rapidly increasing worldwide and facilitate the development of many related disease states, such as cardiovascular disease, the metabolic syndrome, type 2 diabetes mellitus, and various types of cancer. Variation in metabolically important genes can have a great impact on a population’s susceptibility to becoming obese and/or developing related complications. The adipokines adiponectin and leptin, as well as the leptin receptor, are major players in the regulation of body energy homeostasis and fat storage. This paper summarizes the ﬁndings of single nucleotide polymorphisms in these three genes and their eﬀect on obesity and metabolic disease risk. Additionally, studies of gene-nutrient interactions involving adiponectin, leptin, and the leptin receptor are highlighted to emphasize the critical role of diet in susceptible populations.

1. Introduction The growing prevalence of obesity and obesity-related pathologies has spurred the search for greater insight into the Obesity is the result of an imbalance in energy homeostasis mechanisms that contribute to the development of obesity and is characterized by increased adipose tissue mass, chronic and its complications. low-grade inflammation, insulin resistance, and endothelial Adipose tissue plays multiple important roles in body dysfunction. Obesity is a major risk factor for type 2 diabetes weight regulation and energy homeostasis. Adipose func- mellitus (T2DM), cardiovascular disease (CVD), and several tions as an energy storage organ, storing fat primarily types of cancer [1–3], and lies at the core of a cluster of metabolic abnormalities defined as the metabolic syndrome, in the form of triglycerides and releasing free fatty acids which includes insulin resistance and hyperinsulinemia, as the body’s energy demands change. Adipose tissue is hypertension, impaired glucose tolerance, and T2DM [4]. also an active endocrine organ, secreting many cytokines, According to the National Health and Nutrition Examination chemokines, and hormone-like factors. These molecules, Survey (NHANES), 33.8% of American adults are obese which are produced and secreted primarily by adipocytes, (BMI ≥ 30) and 68.0% are considered overweight or obese are known as adipokines [7]. Adipokines constitute a diverse (BMI ≥ 25) [5]. Dietary choices (and overeating) in com- group of bioactive peptides with many and varied roles, bination with low physical activity are typically attributed including mediation of glucose and lipid metabolism, blood as the root cause of the rapid spread of the obesity epide- pressure regulation, and modulation of inflammation and mic in the modern world [6], but genetic factors are a immune function [8]. While over 100 adipokines have strong determinant of individual susceptibility to obesity. been identified [9], the specific functions of many of 2 Journal of Obesity these molecules are poorly understood. Regardless, it has 3.1. Adiponectin Levels and Obesity-Related Metabolic Disease been clearly established that in obesity, adipocytes undergo Parameters. It is estimated that a 30–70% variation in hypertrophy and become dysfunctional [10, 11]. As a result, normal circulating adiponectin levels can be attributed to the adipokine profile they express and secrete is altered, genetic factors [23–29]. A total of 42 SNPs in AdipoQ leading to a proinflammatory environment both locally and and its regulatory region with a minor allele frequency systemically, and contributing to the pathological effects of of >1.5% have been identified [30]. Table 1 lists AdipoQ obesity. SNPs studied in the last decade and their relation (if any) Since many critical metabolic functions are influenced by to adiponectin levels and other obesity-related metabolic adipokines, genetic variations that affect their efficacy may disease parameters. Since changes in adiponectin due to contribute to various pathophysiological states. For instance, genetic variation are of particular interest in this paper, the genetic variation in adipokine genes has been shown to table highlights the percent change in adiponectin levels modulate circulating adipokine levels and thus could pre- in those studies that reported this parameter. Four SNPs dispose carriers of single nucleotide polymorphisms (SNPs) (−11391 G > A, −11377 C > G, +45 T > G and +276 G > to developing obesity or other metabolic illnesses in which T) were analyzed with far greater frequency than any others adipokines play a prominent role, or alternatively, provide and will be the focus of the following discussion. them some protection against disease. Studying the impact The majority of studies analyzing the SNP −11391 G of such gene polymorphisms in human populations can > A found a favourable increase in circulating adiponectin provide insight into the roles specific adipokines play in levels in those subjects carrying the A allele. A recent meta- obesity and related pathologies. This paper will discuss the analysis determined that SNP −11391 G>A was associated association of SNPs in the protein-coding genes for two well- with adiponectin levels according to a dominant model with studied adipokines, adiponectin and leptin, as well as the A allele carriers (GA and AA genotypes) having higher leptin receptor, in the context of obesity-related metabolic adiponectin levels compared with GG carriers [29]. An disease. In addition, due to the profound effectthatdietcan in vitro study supports these data, reporting a biological have on weight gain and regulation, the recent literature on function of this SNP with the A allele enhancing AdipoQ nutrient-gene interaction studies involving adipokines and promoter activity [31].Despitetheprevalenceofthehigh dietary factors will be highlighted. adiponectin finding, most studies did not report any association with improved health in their subjects. Only one study, an analysis of a Caucasian population from Italy, showed 2. Methods a decrease in obesity-related risk factors (BMI, weight, and waist and hip circumference) [32]. Three studies had Articles for this review were identified using the PubMed/ conflicting findings. In a study of European children carrying Medline databases. Search terms included “single gene poly- the A allele, adiponectin levels were found to be higher, morphism”,“gene variant”,“adiponectin”,“AdipoQ”,“leptin”, similar to the situation in adults; however, the SNP showed “leptin receptor”, and “obesity” or “metabolic disease”. An an obesity-mediated detrimental association with fasting emphasis was placed on studies published in the last decade, serum insulin and HOMA-IR [33]. A second study involving but the search was not limited to a specific time interval. obese and morbidly obese French Caucasians found the SNP The articles were chosen by scanning the abstract to ensure to be associated with lower adiponectin levels, but similarly relevancy. Only studies in human populations and in English accompanied by lower insulin sensitivity and a higher risk of were included. T2DM [32]. In a third study, there was also an association between the GA genotype and risk of hyperglycemia in 3. Adiponectin a population of French Caucasians [34]. While it is clear that most carriers of the A allele have raised adiponectin Adiponectin is an important anti-inflammatory and insulin- levels and could expect to be protected from metabolic sensitizing hormone and promotes lipid oxidation in tissues disease, in certain populations the increase in adiponectin such as skeletal muscle and liver [12, 13]. Adiponectin observed in GA and AA carriers appears to be too small to also has direct antiatherosclerotic properties, as it strongly impart any appreciable metabolic advantage, where it fails to inhibits expression of adhesion molecules and growth factors counterbalance the metabolic damages of obesity. In fact, it [14]. Adiponectin serum levels are inversely correlated with may contribute to the increased risk for childhood obesity body fat percentage in obese subjects, as well as in those and related insulin resistance. afflicted with T2DM or coronary heart disease [15, 16]. The findings on the −11377 G > C SNP are inconsistent, Due to the protective nature of adiponectin in many types but the general trend links the G allele to various detrimental of cardiovascular and metabolic disease states, low serum conditions, including lower adiponectin levels [23, 42, 43, levels of this adipokine are thought to contribute to the 45, 48], risk for developing hypertension [43], and, in some pathogenesis of these conditions. Several excellent reviews cases, risk for developing colorectal cancer [46]. On the other on the various roles of adiponectin are available [17–20]. hand, the presence of the C allele has also been associated The adiponectin gene AdipoQ has been identified as a with higher BMI and obesity risk [31, 36], increased fasting susceptibility locus for the metabolic syndrome, T2DM and glucose levels and T2DM risk [35, 42]. For example, a CVD [21, 22]. AdipoQ is located on chromosome 3q27. The study that investigated genetic variations in adiponectin gene is 15.8 kb long and contains three exons. in individuals with metabolic syndrome found that SNP Journal of Obesity 3

Table 1: SNPs in the adiponectin gene AdipoQ.

Adiponectin SNP ID Position Parameter association Population level P value Reference (% change) No association with T2DM, rs860291 −12823 Pima Indians [24] BMI, or insulin sensitivity SNP associated with increased −11426 rs16861194 risk for gaining weight in Chinese (T2DM) [35] G > A diabetics SNP associated with fasting Swedish Caucasians plasma glucose in T2DM (T2DM/ impaired [36] patients and in those with glucose tolerance/ impaired glucose tolerance nondiabetic) G allele moderately associated French Caucasians [37] with T2DM A allele associated with higher −1391 Children of rs17300539 adn levels, higher BMI, and 13.39 6.00E-08 [33] G > A European origin obesity A allele carriers have lower weight, waist and hip [32] circumferences and BMI GA carriers had increased risk for becoming French Caucasians [34] hyperglycaemic/diabetic A allele associated with higher French Caucasians .0001 [23] adn levels Hispanic Americans A allele associated with higher and African 18.89 .0001 [27] adn levels Americans A allele associated with higher Caucasians [38] adn levels A allele associated with higher Caucasian women 36.93 .0006 [39] adn levels

A allele associated with higher French Caucasians .005 [31] adn levels in obese children (obese/lean) Caucasian and A allele associated with higher African American 29.41 .002 [40] adn adolescents A allele associated with higher Caucasians 19.05 .0005 [41] adn levels A associated with lower adn levels, lower insulin sensitivity, French Caucasians 32.01 .0003 [42] and higher risk of T2DM in (lean/obese) obese subjects C allele associated with higher −11377 rs266729 fasting plasma glucose levels Chinese (T2DM) [35] G > C in diabetics

C allele associated with severe French Caucasians [31] obesity (obese/lean) G allele associated with lower Chinese adn levels, higher risk of .0037 [43] (hypertensive) hypertension SNP associated with increase in plasma oxidative stress T2DM patients [44] markers 4 Journal of Obesity

Table 1: Continued.

Adiponectin SNP ID Position Parameter association Population level P value Reference (% change) G allele associated with lower adn levels, lower insulin French Caucasians 20.66 .008 [42] sensitivity, and higher risk of (lean/obese) T2DM in obese subjects G allele associated with European men with coronary stenoses and lower 26.92 .003 [45] CVD adn levels SNP associated with increased Czech patients [46] risk for colorectal cancer No association with adn levels Caucasian Italians [32] No association with colorectal UK [47] cancer risk

GG and CG associated with American CRC [48] lower CRC risk patients G associated with lower adn French Caucasians .0003 [23] levels Swedish Caucasians CC and CG genotypes had (T2DM/ impaired [36] higher BMI than GG glucose tolerance/ nondiabetic) SNP associated with lower −11365 18.36 .007 [49] plasma adn levels No association with T2DM, Pima Indians [24] BMI, or insulin sensitivity − 10677 SNP associated with lower adn Chinese .0027 [43] C > T levels (hypertensive) −10068 A allele associated with lower Hypertensive rs182052 .0001 [43] G > A adn levels Chinese

A allele associated with waist American Caucasian [50] circumference young adults Caucasian and G allele associated with higher African American 17.58 0.03 [40] adn adolescents −10066 G allele associated with higher Caucasian women 8.67 .01 [39] G > A adn A allele associated with higher rs16861209 −7734 C > A Caucasian women 22.68 .004 [39] adn rs822395 −4041 A > C No association with adn levels Caucasian Italians [32] −4034 CC associated with CVD risk [49] A allele associated with worse Caucasian Canadians −3971 G > A glucose tolerance and insulin [51] (nondiabetic) sensitivity, but not adn levels GG and TG genotypes were at rs2241766 +45 T > G Obese Iranians [52] higher risk for T2DM Both TG and GG genotypes were associated with Pregnant (<18 gestational T2DM, whereas weeks) Malaysian 19.92 .05 [53] among healthy participants, women the TT genotype had higher adnlevelsthanothergroups Journal of Obesity 5

Table 1: Continued. Adiponectin SNP ID Position Parameter association Population level P value Reference (% change) G allele associated with lower Nondiabetic Greek fasting insulin levels and lower [54] women HOMA-IR score G allele associated with higher TG, HOMA, fasting blood Chinese glucose, BMI and ALT, and (NAFLD/metabolic 28.68 .008 [55] lower adn levels; T allele syndrome) associated with lower body weight GG associated with T2DM Japanese [56] G allele associated with T2DM (lower insulin sensitivity), lower adn, higher blood Chinese (T2DM) 15.47 .01 [57] pressure, higher LDL and total cholesterol levels A allele associated with worse Caucasian Canadians glucose tolerance and insulin [51] (nondiabetic) sensitivity, but not adn levels G allele associated with BMI Hispanic Americans [58] and waist circumference GG carriers had higher risk of becoming hyperglycaemic/diabetic, French Caucasian [34] associated with increase in BMI and WHR over 3 years No diﬀerence in risk for Korean (diabetic/ [59] T2DM or IR nondiabetic) T allele and TG genotype associated with lower serum Caucasians 25.17 .0008 [60] adn, no association with IR GT genotype associated with Spanish [61] impaired glucose tolerance G allele conferred higher risk European/Canadian of developing T2DM than TT subjects with genotype, particularly when [62] impaired glucose combined with SNP +276 T tolerance allele T allele associated with lower Japanese [63] BMI and HOMA-IR (nondiabetic) In obese subjects, serum cholesterol and waist Swedish women circumference were lower in [64] (obese/lean) TG genotype than in TT genotype No association with adn levels Caucasian Italians [32] No association with risk for Caucasian Italians [65] coronary artery disease (T2DM) No association with T2DM, Pima Indians [24] BMI, or insulin sensitivity G allele associated with coronary artery disease in European Caucasians [66] T2DM patients G associated with higher adn French Caucasians .01 [23] levels 6 Journal of Obesity

Table 1: Continued. Adiponectin SNP ID Position Parameter association Population level P value Reference (% change) African American rs1501299 +276 G > T T allele associated with obesity [67] men GG associated with T2DM, higher insulin resistance, and Japanese 10.40 .01 [56] lower adn levels in subjects with higher BMI T allele associated with higher Caucasian women 4.46 .0031 [39] adn levels T allele associated with central Indigenous [68] obesity and hyperglycemia Taiwanese T allele associated with lower adn levels, diastolic blood Finnish men 33.58 .001 [69] pressure T allele associated with higher fasting insulin levels and Greek women higher HOMA-IR score, [54] (nondiabetic) possible association with body fat GG genotype associated with lower adn levels, impaired Spanish .015 [61] glucose tolerance SNP associated with higher rate of insulin resistance, Normolipidaemic [70] higher n-6/n-3 LCPUFA ratio obese children in plasma phospholipids T allele associated with severe French Caucasians [31] obesity, but not adn (obese/lean) TT genotype associated with lower CVD risk in diabetic American men 27.03 .0029 [71] patients, those without CVD (T2DM) had higher adn levels T allele is an important determinant of CAD and lower adn levels in patients Italian CAD patients [72] with early onset CAD (50 years of age or less) Caucasian and T allele associated with higher African American 4.95, 5.81 .05,.03 [40] adn adolescents G allele carriers had higher TG, higher small dense LDL, and smaller LDL particle size; Korean (nondiabetic) 18.90 .026 [73] GG had lower adn, higher HOMA-IR Japanese men No association with adn levels (hypertensive/ [74] or hypertension normotensive) No diﬀerence in allele frequencies between diabetic Korean (diabetic/ and nondiabetic, no diﬀerence [59] nondiabetic) in risk of T2DM or insulin resistance No association with T2DM, Pima Indians [24] BMI, or insulin sensitivity TT genotype associated with Caucasian Italians lower risk of coronary artery [65] (T2DM) disease in T2DM patients Journal of Obesity 7

Table 1: Continued. Adiponectin SNP ID Position Parameter association Population level P value Reference (% change) TT genotype associated with Caucasian Italians .032 [32] higher adn levels G allele, GT genotype associated with lower serum Caucasians 13.70 .00005 [60] adn, no association with insulin resistance T allele associated with lower Japanese [63] BMI and HOMA-IR (nondiabetic) T associated with higher adn French Caucasians .01 [23] levels T allele associated with higher rs1063538 +3228 C > T Caucasian women 24.97 .036 [39] adn levels No association with T2DM, rs1063538 +3286 Pima Indians [24] BMI, or insulin sensitivity G allele associated with higher +10211 T > G diabetes risk, higher BMI, and Asian Indians .007 [75] lower adn levels T allele associated with lower Chinese rs12495941 G > T .0001 [43] adn levels (hypertensive) rs3774261 A > G G allele associated with IR African Americans [76] Hispanic Americans rs1656943 C allele associated with higher T > C and African 12.62 .003 [27] (rs822387) adn levels Americans

−11377 G > C was a determinant of HOMA-IR, where expression of adiponectin. It is thus possible that SNPs with CC homozygotes had significantly lower HOMA-IR scores no apparent biological significance may have an effect on [77]. There is evidence that the presence of the minor G gene expression, although in this case it is likely that SNP allele decreases the affinity of the transcription factor Sp1 +45 is in linkage disequilibrium with some other functional to its binding site within the AdipoQ promoter [78], and genetic alterations, resulting in the difference in mRNA a recent study showed that this allele had altered DNA- expression of its two alleles. Other research has indicated that binding activity, leading to lower basal and inducible AdipoQ the +45 T > G SNP is in linkage disequilibrium with the promoter activity in mouse 3T3-L1 adipocytes [79]. The +276 G > T SNP and that the haplotype defined by the two mechanism by which the C allele might contribute to disease together is strongly associated with many components of the remains unclear. metabolic syndrome [61, 62, 81]. The silent +45 T > G SNP is strongly associated with The G allele of the +276 G > T SNP is primarily detrimental health effects including lower adiponectin levels associated with lower insulin sensitivity and increased T2DM [55, 57], higher BMI and lower insulin sensitivity [34, risk, lower adiponectin levels, and increased blood lipids. 55], higher risk for developing hyperglycemia and T2DM Conversely, many carriers of the T allele have higher [34, 52, 53, 56, 57, 62], and higher levels of blood lipids adiponectin levels and a lower BMI. Two notable exceptions (triglycerides, LDL cholesterol, and total cholesterol) [55, to this trend are the studies authored by Beebe-Dimmer et al. 57]. In contrast, the T allele generally appeared to afford [67] and Bouatia-Naji et al. [31], in which the presence of the the carrier some protection, being associated with higher T allele corresponded with severe obesity. The first of these adiponectin levels [53] and lower body weight [55, 63]. A occurred in African American men, and so the conflicting few studies obtained results that differed, and some found no results may be attributed to the racial composition of the significant associations with these parameters at all; however, populations studied. The second was a study of lean and the overall trends remained apparent even across such diverse obese French Caucasians. The authors suggest that while populations as Iranians, Japanese, and European Caucasians the higher adiponectin levels seen in the obese T allele [34, 52, 56]. The mechanisms by which these genetic varia- carriers may protect them against insulin resistance and tions exert effects on adiponectin levels and metabolic disease T2DM, hyperadiponectinemia may actually predispose these parameters have not been fully elucidated. Yang et al. [80] patients to weight gain due to the insulin-sensitizing effects showed that the silent +45 T > G mutation may alter RNA of adiponectin, which could promote lipid uptake and splicing or stability, suggesting an allele-specific differential storage [31, 82]. This mechanism may also explain why 8 Journal of Obesity some populations have higher adiponectin levels despite 4. Leptin being in an obese state. The T allele also appears to be an important determinant of CVD risk; however, this may Leptin regulates body weight and energy expenditure, and correlate directly with the tendency of T allele carriers to have plays important roles in the modulation of glucose and lipid increased adiponectin levels, as demonstrated in one study metabolism, angiogenesis, immunity, and blood pressure in which the diabetic participants with the TT genotype had homeostasis. Leptin is also a critical signalling molecule lower risk of CVD, while those without CVD had higher in the hypothalamus, where it influences appetite and adiponectin levels [71]. Several other studies identified the satiety. The circulating levels of leptin correlate directly T allele as a determinant of cardiovascular risk [65, 72]. with adipocyte number and size [87], thus, leptin levels are elevated in obesity and are thought to exacerbate many Overall, the magnitude of change in adiponectin levels ff varies considerably between studies. This is not surprising, of the negative e ects of weight gain, such as contributing since the studies examine populations that are very different to the local inflammatory response [88] and creating a in ethnicity, age, and health condition. However, it is worth positive feedback loop for feeding behaviour through leptin noting that several studies show quite substantial changes resistance [89]. More details on the many roles of leptin in in adiponectin levels and therefore provide convincing metabolic disease can be found in recent reviews [90–92]. evidence that seemingly minute variations in the genetic The study of leptin began when mice homozygous code can produce large changes in adipokine levels, and thus for single-gene mutations in the leptin gene (ob/ob)and significantly affect health status. the leptin receptor gene (db/db) were identified [93]. The absence of leptin or its receptor leads to uncontrolled eating, and mice with either defect become massively obese. 3.2. AdipoQ Nutrient-Gene Interactions. Subtle genetic vari- Treatment of ob/ob mice with leptin injections brings about ations can have a large impact on important obesity- a reduction in body weight to that of a normal mouse. related disease determinants, as demonstrated by many of For this reason, leptin was once believed to be the solution studies discussed above. Individuals with SNPs in genes to the Western world’s epidemic of obesity. Unfortunately, such as AdipoQ can be subject to greater sensitivity to it was soon determined that such monogenic mutations dietary factors, due to the critical role adiponectin plays occur very rarely in humans. In fact, severe obesity due to in maintaining metabolic balance. Several recent studies a single mutation in the leptin gene has been observed in have investigated the nutrient-gene interactions that take only 12 human cases in the entire world [94]. It is now clear place between dietary factors and AdipoQ SNPs. Perez-´ that multiple genes and gene variants are involved in the Martınez´ et al. [83] investigated the influence of dietary development of human obesity. SNPs in the human leptin fat on insulin resistance in C allele carriers of the −11377 gene (Lep) and the leptin receptor gene (LepR)canhave G > C SNP in Caucasian men and women. Only men a profound impact on body weight, insulin resistance and who were homozygous for the C allele had significantly other metabolic disease parameters. The literature describing lower IR after consuming monounsaturated (MUFA-) and the effects of these SNPs is summarized in Table 2. carbohydrate-rich diets than after consuming an SFA-rich diet. In a population of European Caucasian ancestry with 4.1. Leptin and Leptin Receptor Gene Variants: Risk for MUFA intake above the median, lower BMI and decreased Obesity-Related Metabolic Disease. The Lep gene is located obesity risk were observed in carriers of the −11391 A allele on chromosome 7q31 and encompasses approximately 20 kb. [84]. The +45, +276, and −11377 SNPs were examined It contains 3 exons, the first of which is noncoding. in 363 subjects with impaired fasting glucose or newly Its sequence is highly conserved and contains very little diagnosed type 2 diabetes following a dietary intervention reported variation. Only one leptin SNP, +19 G > A, has (replacement of cooked refined rice with whole grains and been investigated in detail for its effects on obesity-related an increase in vegetable intake) and regular walking for metabolic disease. A recent study found that the A allele 12 weeks without any medication. Fasting glucose levels was significantly associated with lower body weight, lower declined in all genotype groups of the +45 T > G SNP, and BMI, lower circulating leptin levels, and consequently, a TT homozygotes had increased adiponectin levels and lower lower risk for obesity in Caucasian and African-American HOMA-IR indexes [85]. In another study, obese Japanese women [100]. Another reported that the presence of the women with the +276 SNP were placed on a low-calorie diet A allele was linked to higher leptin levels and lower BMI for 8 weeks. At the study conclusion, those with a GT or in obese Caucasian females compared to GG homozygotes, TT genotype had a greater decrease in waist circumference, suggesting that carriers of this allele may experience better and those with the TT genotype in the +45 SNP had lower sensitivity to satiety signals via the leptin protein [99]. Several plasma triglycerides. In the same population, the participants other publications failed to find any significant associations with CG and GG genotypes at SNP −11377 enjoyed a between this SNP and BMI or blood lipid levels [96–98]. The greater decrease in systolic blood pressure and fasting plasma Lep +19 A > G variant lies within the first untranslated exon glucose than those with CC [86]. Each of these studies of the gene, and it is not known how such an alteration might used appropriate statistical testing to determine significant modify protein function. However, it is has been suggested interactions between the gene polymorphisms and metabolic that the Lep +19 A > G SNP is in disequilibrium with parameters. These findings correspond with the trends in promoter region variation that may have an effect on gene metabolic disease discussed in the previous section. transcription [100]. Journal of Obesity 9

Table 2: SNPs in the Leptin (Lep) and Leptin Receptor (LepR) Genes.

Amino acid Nucleotide Leptin level SNP ID Parameter association Population P value Reference change change (% change) Leptin (Lep) Yo ung a dults SNP associated with weight (Caucasians, rs4731427 and waist circumference in [95] African African Americans Americans) No association with BMI, WHR, fasting glucose & +19G > A [96] insulin, lipids and leptin levels No association with waist girth, plasma triglycerides, French HDL-cholesterol, glucose [97] Caucasian and systolic, and diastolic blood pressure No association with Italian waist-to-hip ratio, fasting Caucasian leptin, total cholesterol, [98] (obese/non- high-density lipoproteins, obese) triglycerides No genotype associated French with BMI, but A allele Caucasian associated with higher 18.93 .001 [99] (obese/non- leptin levels in obese obese) patients A allele in females associated with lower body African weight, BMI and plasma Americans and 6.68 .01 [100] leptin levels, lower risk of Caucasians obesity SNP associated with weight and waist circumference in Yo ung a dults African Americans and (Caucasians, rs17151919 [95] weight in Caucasians, waist African circumference in Caucasian Americans) women SNP associated with weight, waist circumference Yo ung a dults in African Americans and (Caucasians, rs28954369 [95] weight in Caucasians, waist African circumference in Caucasian Americans) women SNP associated with weight Yo ung a dults in Caucasians, waist (Caucasians, rs2167270 [95] circumference in Caucasian African women Americans) A allele signiﬁcantly rs7799039 G > A Caucasians [101] associated with BMI G allele associated with overweight, and with lower −2548 G > A 17.46 .05 [102] leptin concentrations in men

A allele not associated with Spanish [103] obesity Mediterranean 10 Journal of Obesity

Table 2: Continued. Amino acid Nucleotide Leptin level SNP ID Parameter association Population P value Reference change change (% change) Leptin Receptor (LepR) G allele associated with Asian Indians rs1137101 Gln223Arg A > G BMI, WHR, leptin levels, (diabetic/ .001 [104] and insulin levels nondiabetic) G allele associated with Caucasian higher BMI, fat mass, and women (post- 31.15 .0001 [105] serum leptin levels menopausal) Brazilians of GG genotype associated European [106] with BMI in nonsmokers descent (Caucasian) G allele associated with increased rates of obesity, Greek [107] higher BMI and % fat mass GG genotype had larger subcutaneous abdominal adipocyte size than AA, Pima Indians [108] however, no diﬀerence in overall adiposity G allele associated with Caucasians [109] insulin resistance GG phenotype associated Spanish [103] with lean phenotype Mediterranean GG and AG genotypes associated with increased risk of familial hypercholesterolemia, but Dutch [110] not obesity, insulin resistance or other lipid parameters Brazilians G allele associated with (obese/non- [111] increased rate of obesity obese) Belgian AA genotype was Caucasian associated with increased women [112] total abdominal fat (overweight and obese) No association with BMI, fasting insulin, HOMA-IR, Japanese [113] serum leptin, or soluble leptin receptor levels Danish post- G allele associated with menopausal [114] increased adiposity women GG genotype had lower Swedish men blood pressure compared (hypertensive/ [115] to AA normotensive) G allele associated with Czech (CRC [46] CRC risk patients) G allele associated with [116] higher fat mass and BMI A allele associated with Mexican higher insulin, leptin levels, 62.02 .001 [117] adolescents and body fat No association with BMI, WHR, fasting glucose & [96] insulin, lipids and leptin levels Journal of Obesity 11

Table 2: Continued. Amino acid Nucleotide Leptin level SNP ID Parameter association Population P value Reference change change (% change) Finnish AA genotype had greater (impaired [118] risk of developing T2DM glucose tolerance) G allele associated with BMI, change in BMI over [119] time T allele associated with overweight and fat mass in Ser(T)343- T > C women; C allele carriers [120] Ser(C) more responsive to weight loss on a low calorie diet C allele associated with fat +70 T > C [120] mass in women G allele associated with Lys656Asn G>C Caucasians [121] higher lean and fat mass No association with Greek [107] obesity, BMI, or % fat mass C allele associated with Belgian increased hip Caucasian circumference, total women [112] abdominal fat, and (overweight subcutaneous fat and obese) Belgian C allele associated with Caucasian higher fasting glucose and women [122] fasting insulin in (overweight postmenopausal women and obese) Swedish men No association with blood (hypertensive/ [115] pressure or BMI normotensive) No association with blood Mexican pressure, serum glucose, [117] adolescents insulin, or leptin levels No association with BMI, WHR, fasting glucose & Arg109Lys +5193 G > A [96] insulin, lipids and leptin levels Belgian AA genotype was associated Caucasian with higher leptin levels in women 14.99 .02 [112] postmenopausal women (overweight and obese) A allele positively Korean [123] associated with BMI No association with BMI, fasting insulin, HOMA-IR, Japanese [113] serum leptin, or soluble leptin receptor levels No association with Greek [107] obesity, BMI, or % fat mass Belgian A allele associated with Caucasian fasting insulin in women [122] postmenopausal women (overweight and obese) 12 Journal of Obesity

Table 2: Continued. Amino acid Nucleotide Leptin level SNP ID Parameter association Population P value Reference change change (% change) Finnish AA genotype had greater (impaired [118] risk of developing T2DM glucose tolerance) GG genotype had lower Swedish men blood pressure and lower (hypertensive/ [115] BMI compared to AA normotensive) SNP associated with change rs1045895 [119] in BMI over time

The LepR gene is found on chromosome 1p31, spans in individuals homozygous for the G allele than for the A about 100 kb, and contains 20 exons. Numerous analyses allele. These authors suggest the change in binding affinity of LepR SNPs have been published over the last decade, might also result in an evolutionarily significant tendency to as the role of these genetic variants as determinants of higher fat mass, contributing to the fixation of these alleles in adiposity and related conditions has become clear. The SNP human populations. Gln223Arg A > G has been studied extensively in a wide range of populations. The G allele is primarily associated 4.2. Gene-Nutrient Interactions with Lep and LepR. Most of with increased adiposity, BMI and percent fat mass, as well as the literature on gene-nutrient interactions involving leptin higher circulating insulin and leptin levels. Larger adipocyte or the leptin receptor focuses on fetal nutrition and leptin size has also been observed in individuals with the GG levels in breast-feeding mothers. Recent evidence suggests genotype. A couple of studies revealed varying results, with that early prenatal and postnatal nutrition has an impact on the G allele linked to a lean phenotype [103] and lower blood susceptibility to chronic disease later in life. Leptin in breast pressure [115], or no association whatsoever with weight- milk has been identified as a key protective factor against related parameters [96, 113]. Conversely, the A allele has several metabolic and physiological changes at an older age, also been found to be associated with total abdominal fat such as obesity and related medical complications [124]. mass, increased insulin and leptin levels, and higher risk of Several recent reviews on this topic are available [125, 126]. developing T2DM [117]. There is scant research on the interactions of leptin The Lys656Asn G > C SNP was analyzed in several differ- or leptin receptor gene SNPs with dietary factors. One ent populations, but no consistent trends were identified— study analyzed the effects of leptin receptor polymorphisms both alleles are associated with increased fat mass in different and PUFA consumption in relation to insulin resistance ethnic groups [112, 121], whereas the C allele has also been and metabolic syndrome [127]. The findings revealed that linked to increased fasting glucose and fasting insulin [122]. participants in the lowest median of plasma (n-3) PUFA Two studies also reported no association between this SNP and LCPUFA with the GG genotype of the rs3790433 and blood pressure, BMI, leptin levels, insulin and serum SNP were at higher risk for hyperinsulinemia and insulin glucose [115, 117]. resistance, whereas in individuals with the same genotype The A allele of the Arg109Lys G > A SNP is associated but high plasma (n-3) PUFA and LCPUFA the risk of with increased T2DM risk [118], higher leptin levels [112], developing hyperinsulinemia and insulin resistance was and higher BMI [123]. Several studies reported no associa- effectively eliminated. Moreover, a high-plasma (n-6) PUFA tion between this SNP and obesity-related disease parameters profile accentuated the risk of these same conditions in [96, 107, 113], while one study found the GG, but not AA, GG homozygotes. The study concluded that homozygous genotype to be linked to lower blood pressure and BMI [115]. carriers of this SNP may be predisposed to metabolic Functional data on these polymorphisms is scarce, and syndrome compared with the A allele carriers, especially if the mechanisms by which the genetic variation influences the plasma fatty acid profile was unfavourable. Another study metabolism and obesity are largely speculative at this point assessed the influence of the LepR Lys656Asn polymorphism in time. The increased fat mass and leptin levels in many on the leptin response secondary to a low fat or low of the LepR SNPcarrierspointstowardapossiblechange carbohydrate diet in obese people [128]. Leptin levels were in binding affinity for the leptin receptor. The receptor significantly lower in the Lys656/Lys656 cohort on a low fat functions in a dimeric form both in serum and at the cell diet than on a low carbohydrate diet. The Lys656/Lys656 surface. The genetic variations in the coding gene for the group also enjoyed a decrease in several obesity-related receptor may alter the ability of subunits to form dimers, disease parameters (BMI, weight, fat mass, blood pressure, leading to the phenotypic differences observed. Quinton et total cholesterol, triglycerides, blood insulin, and glucose) al. [105] measured the leptin-binding activity of the soluble on a low fat diet, while significant changes in only BMI, leptin receptor in study participants who were carriers of weight and fat mass were observed in Asn656 carriers on the Gln223Arg SNP and found significantly higher activity the same diet. These studies indicate that polymorphisms Journal of Obesity 13 in the leptin receptor genes can play an influential role in [2]E.E.Calle,C.Rodriguez,K.Walker-Thurmond,andM.J. the body’s physiological response to diet. There is a need for Thun,“Overweight,obesity,andmortalityfromcancerin more research in this important area. a prospectively studied cohort of U.S. Adults,” New England Journal of Medicine, vol. 348, no. 17, pp. 1625–1638, 2003. [3] C. Samanic, W. H. Chow, G. Gridley, B. Jarvholm, and J. F. 5. Concluding Remarks Fraumeni Jr., “Relation of body mass index to cancer risk in 362,552 Swedish men,” Cancer Causes and Control, vol. 17, In summary, the current literature demonstrates that SNPs no. 7, pp. 901–909, 2006. in the genes for adiponectin, leptin, and the leptin receptor [4] O. Ukkola and C. Bouchard, “Clustering of metabolic abnor- can to a great degree influence the carriers’ susceptibility malities in obese individuals: the role of genetic factors,” to obesity and related complications. While epidemiological Annals of Medicine, vol. 33, no. 2, pp. 79–90, 2001. studies have reported associations between adipokine levels [5] K. M. Flegal, M. D. Carroll, C. L. Ogden, and L. R. Curtin, and metabolic disease parameters, the evidence compiled “Prevalence and trends in obesity among US adults, 1999– here provides a compelling argument for the causality of 2008,” Journal of the American Medical Association, vol. 303, adipokine levels in disease. In many of the studies cited here, no. 3, pp. 235–241, 2010. changes in adipokine levels were observed in individuals [6]P.G.Kopelman,“Obesityasamedicalproblem,”Nature,vol. with gene polymorphisms, and thus, the change in adipokine 404, no. 6778, pp. 635–643, 2000. levels precedes the occurrence of pathological conditions. [7] P. Trayhurn and I. S. Wood, “Adipokines: inflammation and Varied downstream effects on parameters of health are the pleiotropic role of white adipose tissue,” British Journal of to be expected, since the adipokines examined here play Nutrition, vol. 92, no. 3, pp. 347–355, 2004. diverse roles in many organ systems, as recently reviewed by [8] F. Y. Li, K. K. Cheng, K. S. Lam, P. M. Vanhoutte, and A. Xu, DeClercq et al. [129]. The genetic variants included in this “Cross-talk between adipose tissue and vasculature: role of review by no means constitute an exhaustive list, and further adiponectin,” Acta Physiologica. In press. investigation of adipokine SNPs and their impact on human [9] H. Hauner, “Secretory factors from human adipose tissue ff and their functional role,” Proceedings of the Nutrition Society, health is warranted. Furthermore, population di erences vol. 64, no. 2, pp. 163–169, 2005. are also a factor in the strength of genetic determinants [10] M. Bluher, “Adipose tissue dysfunction in obesity,” Experi- of disease. The gene-nutrient interaction studies discussed mental and Clinical Endocrinology and Diabetes, vol. 117, no. emphasize the role of diet in the modification of risk for 6, pp. 241–250, 2009. developing metabolic disease. Future research in this field is [11] H. E. Bays, J. M. Gonzalez-Campoy,´ G. A. Bray et al., needed to confirm and expand the findings presented here. “Pathogenic potential of adipose tissue and metabolic consequences of adipocyte hypertrophy and increased visceral adiposity,” Expert Review of Cardiovascular Therapy,vol.6, Abbreviations no. 3, pp. 343–368, 2008. adn: Adiponectin [12] T. Yamauchi, J. Kamon, H. Waki et al., “The fat-derived BMI: Body mass index hormone adiponectin reverses insulin resistance associated CAD: Coronary artery disease with both lipoatrophy and obesity,” Nature Medicine,vol.7, CRC: Colorectal cancer no. 8, pp. 941–946, 2001. CVD: Cardiovascular disease [13] T. Yamauchi, J. Kamon, Y. Minokoshi et al., “Adiponectin FSI: Fasting serum insulin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase,” Nature Medicine, HOMA-IR: Homeostatic model assessment of insulin vol. 8, no. 11, pp. 1288–1295, 2002. resistance [14] T. Kadowaki and T. Yamauchi, “Adiponectin and adiponectin IR: Insulin resistance receptors,” Endocrine Reviews, vol. 26, no. 3, pp. 439–451, LCPUFA: Long chain polyunsaturated fatty acid 2005. LDL: Low density lipoprotein [15] J. J. Dıez´ and P. Iglesias, “The role of the novel adipocyte- MUFA: Monounsaturated fatty acid derived hormone adiponectin in human disease,” European PUFA: Polyunsaturated fatty acid Journal of Endocrinology, vol. 148, no. 3, pp. 293–300, 2003. SNP: Single nucleotide polymorphism [16] K. Hotta, T. Funahashi, Y. Arita et al., “Plasma concentrations T2DM: Type 2 diabetes mellitus of a novel, adipose-specific protein, adiponectin, in type 2 WHR: Waist-to-hip ratio. diabetic patients,” Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 20, no. 6, pp. 1595–1599, 2000. [17] P.Trayhurn and I. S. Wood, “Signalling role of adipose tissue: Acknowledgment adipokines and inflammation in obesity,” Biochemical Society Transactions, vol. 33, no. 5, pp. 1078–1081, 2005. This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada. [18] G. R. Hajer, T. W. Van Haeften, and F. L. J. Visseren, “Adipose tissue dysfunction in obesity, diabetes, and vascular diseases,” European Heart Journal, vol. 29, no. 24, pp. 2959–2971, 2008. References [19] E. Maury and S. M. Brichard, “Adipokine dysregulation, adipose tissue inflammation and metabolic syndrome,” [1] K. B. Schelbert, “Comorbidities of Obesity,” Primary Care, Molecular and Cellular Endocrinology, vol. 314, no. 1, pp. 1– vol. 36, no. 2, pp. 271–285, 2009. 16, 2010. 14 Journal of Obesity

[20] Y. Matsuzawa, “Adiponectin: a key player in obesity related [34] F. Fumeron, R. Aubert, A. Siddiq et al., “Adiponectin gene disorders,” Current Pharmaceutical Design, vol. 16, no. 17, pp. polymorphisms and adiponectin levels are independently 1896–1901, 2010. associated with the development of hyperglycemia during [21] N. Vionnet, E. H. Hani, S. Dupont et al., “Genomewide a 3-year period: the epidemiologic data on the insulin search for type 2 diabetes-susceptibility genes in French resistance syndrome prospective study,” Diabetes, vol. 53, no. whites: evidence for a novel susceptibility locus for early- 4, pp. 1150–1157, 2004. onset diabetes on chromosiome 3q27-qter and independent [35]M.Yang,C.C.Qui,W.Chen,L.L.Xu,M.Yu,andH. replication of a type 2-diabetes locus on chromosome 1q21- D. Xiang, “Identification of a regulatory single nucleotide q24,” American Journal of Human Genetics, vol. 67, no. 6, pp. polymorphism in the adiponectin (APM1) gene associated 1470–1480, 2000. with type 2 diabetes in Han nationality,” Biomedical and [22] S. Francke, M. Manraj, C. Lacquemant et al., “Agenome-wide Environmental Sciences, vol. 21, no. 6, pp. 454–459, 2008. ¨ scan for coronary heart disease suggests in Indo-Mauritians [36] H. F. Gu, A. Abulaiti, C.-G. Ostenson et al., “Single nucleotide a susceptibility locus on chromosome 16p13 and replicates polymorphisms in the proximal promoter region of the linkage with the metabolic syndrome on 3q27,” Human adiponectin (APM1) gene are associated with type 2 diabetes Molecular Genetics, vol. 10, no. 24, pp. 2751–2765, 2001. in Swedish Caucasians,” Diabetes,vol.53,supplement1,pp. S31–S35, 2004. [23]F.Vasseur,N.Helbecque,C.Dinaetal.,“Single-nucleotide [37] F. Gibson and P. Froguel, “Genetics of the APM1 locus and polymorphism haplotypes in the both proximal promoter its contribution to type 2 diabetes susceptibility in French and exon 3 of the APM1 gene modulate adipocyte-secreted Caucasians,” Diabetes, vol. 53, no. 11, pp. 2977–2983, 2004. adiponectin hormone levels and contribute to the genetic [38] P. Henneman, Y. S. Aulchenko, R. R. Frants et al., “Genetic risk for type 2 diabetes in French Caucasians,” Human architecture of plasma adiponectin overlaps with the genetics Molecular Genetics, vol. 11, no. 21, pp. 2607–2614, 2002. of metabolic syndrome-related traits,” Diabetes Care, vol. 33, [24] B. V. De Courten, R. L. Hanson, T. Funahashi et al., no. 4, pp. 908–913, 2010. “Common polymorphisms in the adiponectin gene ACDC [39]T.Kyriakou,L.J.Collins,N.J.Spencer-Jonesetal., are not associated with diabetes in Pima Indians,” Diabetes, “Adiponectin gene ADIPOQ SNP associations with serum vol. 54, no. 1, pp. 284–289, 2005. adiponectin in two female populations and effects of SNPs [25] L. M. Chuang, Y. F. Chiu, W. H. H. Sheu et al., “Biethnic on promoter activity,” Journal of Human Genetics, vol. 53, no. comparisons of autosomal genomic scan for loci linked to 8, pp. 718–727, 2008. plasma adiponectin in populations of chinese and Japanese [40] J. G. Woo, L. M. Dolan, R. Deka et al., “Interactions between origin,” Journal of Clinical Endocrinology and Metabolism,vol. noncontiguous haplotypes in the adiponectin gene ACDC 89, no. 11, pp. 5772–5778, 2004. are associated with plasma adiponectin,” Diabetes, vol. 55, no. [26] A. G. Comuzzie, T. Funahashi, G. Sonnenberg et al., “The 2, pp. 523–529, 2006. genetic basis of plasma variation in adiponectin, a global [41] M. F. Hivert, A. K. Manning, J. B. McAteer et al., “Common endophenotype for obesity and the metabolic syndrome,” variants in the adiponectin gene (ADIPOQ) associated with Journal of Clinical Endocrinology and Metabolism,vol.86,no. plasma adiponectin levels, type 2 diabetes, and diabetes- 9, pp. 4321–4325, 2001. related quantitative traits: the framingham offspring study,” [27] X. Guo, M. F. Saad, C. D. Langefeld et al., “Genome-wide Diabetes, vol. 57, no. 12, pp. 3353–3359, 2008. linkage of plasma adiponectin reveals a major locus on [42] F. Vasseur, N. Helbecque, S. Lobbens et al., “Hypoadiponecti- chromosome 3q distinct from the adiponectin structural naemia and high risk of type 2 diabetes are associated with gene: the IRAS Family Study,” Diabetes,vol.55,no.6,pp. adiponectin-encoding (ACDC) gene promoter variants in 1723–1730, 2006. morbid obesity: evidence for a role of ACDC in diabesity,” [28] R. S. Lindsay, T. Funahashi, J. Krakoff et al., “Genome-wide Diabetologia, vol. 48, no. 5, pp. 892–899, 2005. linkage analysis of serum adiponectin in the Pima Indian [43] K. L. Ong, M. Li, A. W. K. Tso et al., “Association of genetic population,” Diabetes, vol. 52, no. 9, pp. 2419–2425, 2003. variants in the adiponectin gene with adiponectin level and [29] C. Menzaghi, V. Trischitta, and A. Doria, “Genetic influences hypertension in Hong Kong Chinese,” European Journal of of adiponectin on insulin resistance, type 2 diabetes, and Endocrinology, vol. 163, no. 2, pp. 251–257, 2010. cardiovascular disease,” Diabetes, vol. 56, no. 5, pp. 1198– [44] S. L. Prior, D. R. Gable, J. A. Cooper et al., “Association 1209, 2007. between the adiponectin promoter rs266729 gene variant and oxidative stress in patients with diabetes mellitus,” [30] H. F. Gu, “Biomarkers of adiponectin: plasma protein European Heart Journal, vol. 30, no. 10, pp. 1263–1269, 2009. variation and genomic DNA polymorphisms,” Biomarker [45] G. Hoefle, A. Muendlein, C. H. Saely et al., “The -11377 Insights, vol. 4, pp. 123–133, 2009. C>G promoter variant of the adiponectin gene, prevalence [31]N.Bouatia-Naji,D.Meyre,S.Lobbensetal., of coronary atherosclerosis, and incidence of vascular events “ACDC/adiponectin polymorphisms are associated with in men,” Thrombosis and Haemostasis, vol. 97, no. 3, pp. 451– severe childhood and adult obesity,” Diabetes,vol.55,no.2, 457, 2007. pp. 545–550, 2006. [46] S. Pechlivanis, J. L. Bermejo, B. Pardini et al., “Genetic [32] C. Menzaghi, T. Ercolino, L. Salvemini et al., “Multigenic variation in adipokine genes and risk of colorectal cancer,” control of serum adiponectin levels: evidence for a role of the European Journal of Endocrinology, vol. 160, no. 6, pp. 933– APM1 gene and a locus on 14q13,” Physiological Genomics, 940, 2009. vol. 19, pp. 170–174, 2005. [47] L. G. Carvajal-Carmona, S. Spain, D. Kerr, R. Houlston, [33] A. Morandi, C. Maffeis, S. Lobbens et al., “Early detrimental J. B. Cazier, and I. Tomlinson, “Common variation at the metabolic outcomes of rs17300539-A Allele of ADIPOQ gene adiponectin locus is not associated with colorectal cancer risk despite higher adiponectinemia,” Obesity,vol.18,no.7,pp. in the UK,” Human Molecular Genetics, vol. 18, no. 10, pp. 1469–1473, 2010. 1889–1892, 2009. Journal of Obesity 15

[48] V. G. Kaklamani, K. B. Wisinski, M. Sadim et al., “Variants [61] J. L. Gonzalez-S´ anchez,´ C. A. Zabena, M. T. Martınez-Larrad´ of the adiponectin (ADIPOQ) and adiponectin receptor 1 et al., “An SNP in the adiponectin gene is associated with (ADIPOR1) genes and colorectal cancer risk,” Journal of the decreased serum adiponectin levels and risk for impaired American Medical Association, vol. 300, no. 13, pp. 1523– glucose tolerance,” Obesity Research, vol. 13, no. 5, pp. 807– 1531, 2008. 812, 2005. [49] LU. Qi, A. Doria, J. E. Manson et al., “Adiponectin genetic [62] J. Zacharova, J. -L. Chiasson, and M. Laakso, “The common variability, plasma adiponectin, and cardiovascular risk in polymorphisms (Single Nucleotide Polymorphism [SNP] patients with type 2 diabetes,” Diabetes,vol.55,no.5,pp. +45 and SNP +276) of the adiponectin gene predict the 1512–1516, 2006. conversion from impaired glucose tolerance to type 2 [50]C.L.Wassel,J.S.Pankow,D.R.JacobsJr.,M.W.Steﬀes, NA. diabetes: the STOP-NIDDM trial,” Diabetes, vol. 54, no. 3, Li, and P. J. Schreiner, “Variants in the Adiponectin Gene and pp. 893–899, 2005. Serum Adiponectin: The Coronary Artery Development in [63] K. Nakatani, K. Noma, J. Nishioka et al., “Adiponectin gene Young Adults (CARDIA) Study,” Obesity, vol. 18, no. 12, pp. variation associates with the increasing risk of type 2 diabetes 2333–2338, 2010. in non-diabetic Japanese subjects,” International Journal of [51] S.-M. Ruchat, R. J. F. Loos, T. Rankinen et al., “Associations Molecular Medicine, vol. 15, no. 1, pp. 173–177, 2005. between glucose tolerance, insulin sensitivity and insulin [64] O. Ukkola, E. Ravussin, P. Jacobson, L. Sjostr¨ om,¨ and C. secretion phenotypes and polymorphisms in adiponectin Bouchard, “Mutations in the adiponectin gene in lean and and adiponectin receptor genes in the Quebec Family Study,” obese subjects from the Swedish obese subjects cohort,” Diabetic Medicine, vol. 25, no. 4, pp. 400–406, 2008. Metabolism: Clinical and Experimental,vol.52,no.7,pp. [52] G. Mohammadzadeh and N. Zarghami, “Associations 881–884, 2003. between single-nucleotide polymorphisms of the [65] S. Bacci, C. Menzaghi, T. Ercolino et al., “The +276 G/T adiponectin gene, serum adiponectin levels and increased single nucleotide polymorphism of the adiponectin gene is risk of type 2 diabetes mellitus in Iranian obese individuals,” associated with coronary artery disease in type 2 diabetic Scandinavian Journal of Clinical and Laboratory Investigation, patients,” Diabetes Care, vol. 27, no. 8, pp. 2015–2020, 2004. vol. 69, no. 7, pp. 764–771, 2009. [66]C.Lacquemant,P.Froguel,S.Lobbens,P.Izzo,C.Dina,and [53] C. F. Low, E. R. Mohd Tohit, P. P. Chong, and F. Idris, J. Ruiz, “The adiponectin gene SNP+45 is associated with “Adiponectin SNP45TG is associated with gestational dia- coronary artery disease in Type 2 (non-insulin-dependent) betes mellitus,” Archives of Gynecology and Obstetrics.In diabetes mellitus,” Diabetic Medicine, vol. 21, no. 7, pp. 776– press. 781, 2004. [67] J. L. Beebe-Dimmer, K. A. Zuhlke, A. M. Ray, E. M. [54] L. Melistas, C. S. Mantzoros, M. Kontogianni, S. Lange, and K. A. Cooney, “Genetic variation in adiponectin Antonopoulou, J. M. Ordovas, and N. Yiannakouris, (ADIPOQ) and the type 1 receptor (ADIPOR1), obesity and “Association of the +45T>G and +276G>T polymorphisms prostate cancer in African Americans,” Prostate Cancer and in the adiponectin gene with insulin resistance in nondiabetic Prostatic Diseases, vol. 13, no. 4, pp. 362–368, 2010. Greek women,” European Journal of Endocrinology, vol. 161, no. 6, pp. 845–852, 2009. [68] MC Huang, TN Wang, KT Lee et al., “Adiponectin gene SNP276 variants and central obesity confer risks for hyper- [55] Z. L. Wang, B. Xia, U. Shrestha et al., “Correlation between glycemia in indigenous Taiwanese,” The Kaohsiung Journal of adiponectin polymorphisms and non-alcoholic fatty liver Medical Sciences, vol. 26, pp. 227–236, 2010. disease with or without metabolic syndrome in Chinese [69] F. Mousavinasab, T. Tahtinen,¨ J. Jokelainen et al., “Common population,” Journal of Endocrinological Investigation,vol.31, polymorphisms (single-nucleotide polymorphisms SNP+45 no. 12, pp. 1086–1091, 2008. and SNP+276) of the adiponectin gene regulate serum [56] K. Hara, P. Boutin, Y. Mori et al., “Genetic variation in the adiponectin concentrations and blood pressure in young gene encoding adiponectin is associated with an increased Finnish men,” Molecular Genetics and Metabolism, vol. 87, no. risk of type 2 diabetes in the Japanese population,” Diabetes, 2, pp. 147–151, 2006. vol. 51, no. 2, pp. 536–540, 2002. [70] E. Verduci, S. Scaglioni, C. Agostoni et al., “The relationship [57] L. L. Li, X. L. Kang, X. J. Ran et al., “Associations between of insulin resistance with SNP 276G>T at adiponectin gene 45T/G polymorphism of the adiponectin gene and plasma and plasma long-chain polyunsaturated fatty acids in obese adiponectin levels with type 2 diabetes,” Clinical and Exper- children,” Pediatric Research, vol. 66, no. 3, pp. 346–349, imental Pharmacology and Physiology, vol. 34, no. 12, pp. 2009. 1287–1290, 2007. [71]LU.Qi,T.Li,E.Rimmetal.,“The+276polymorphism [58] B. S. Sutton, S. Weinert, C. D. Langefeld et al., “Genetic of the APM1 gene, plasma adiponectin concentration, and analysis of adiponectin and obesity in Hispanic families: the cardiovascular risk in diabetic men,” Diabetes, vol. 54, no. 5, IRAS Family Study,” Human Genetics, vol. 117, no. 2-3, pp. pp. 1607–1610, 2005. 107–118, 2005. [72] E. Filippi, F. Sentinelli, S. Romeo et al., “The adiponectin [59] Y. Y. Lee, N. S. Lee, Y. M. Cho et al., “Genetic association gene SNP+276G>T associates with early-onset coronary study of adiponectin polymorphisms with risk of Type 2 artery disease and with lower levels of adiponectin in younger diabetes mellitus in Korean population,” Diabetic Medicine, coronary artery disease patients (age ≤50 years),” Journal of vol. 22, no. 5, pp. 569–575, 2005. Molecular Medicine, vol. 83, no. 9, pp. 711–719, 2005. [60] V. Mackevics, I. M. Heid, S. A. Wagner et al., “The [73] Y. Jang, J. H. Lee, J. S. Chae et al., “Association of the adiponectin gene is associated with adiponectin levels but 276G → T polymorphism of the adiponectin gene with not with characteristics of the insulin resistance syndrome cardiovascular disease risk factors in nondiabetic Koreans,” in healthy Caucasians,” European Journal of Human Genetics, American Journal of Clinical Nutrition, vol. 82, no. 4, pp. 760– vol. 14, no. 3, pp. 349–356, 2006. 767, 2005. 16 Journal of Obesity

[74] Y. Iwashima, T. Katsuya, K. Ishikawa et al., “Hypoad- [88] G. Fantuzzi, “Adipose tissue, adipokines, and inflammation,” iponectinemia is an independent risk factor for hyperten- Journal of Allergy and Clinical Immunology, vol. 115, no. 5, sion,” Hypertension, vol. 43, no. 6, pp. 1318–1323, 2004. pp. 911–920, 2005. [75] K. S. Vimaleswaran, V. Radha, K. Ramya et al., “A novel asso- [89] P. J. Scarpace and Y. Zhang, “Elevated leptin: consequence or ciation of a polymorphism in the first intron of adiponectin cause of obesity?” Frontiers in Bioscience, vol. 12, pp. 3531– gene with type 2 diabetes, obesity and hypoadiponectinemia 3544, 2007. in Asian Indians,” Human Genetics, vol. 123, no. 6, pp. 599– [90]T.A.Dardeno,S.H.Chou,H.S.Moon,J.P.Chamberland,C. 605, 2008. G. Fiorenza, and C. S. Mantzoros, “Leptin in human physi- [76] C. Specchia, K. Scott, P. Fortina, M. Devoto, and B. Falkner, ology and therapeutics,” Frontiers in Neuroendocrinology,vol. “Association of a polymorphic variant of the adiponectin 31, no. 3, pp. 377–393, 2010. gene with Insulin resistance in African Americans,” Clinical [91] A. Stofkova, “Leptin and adiponectin: from energy and and Translational Science, vol. 1, no. 3, pp. 194–199, 2008. metabolic dysbalance to inflammation and autoimmunity,” [77]J.F.Ferguson,C.M.Phillips,A.C.Tierneyetal.,“Gene- Endocrine Regulations, vol. 43, no. 4, pp. 157–168, 2009. nutrient interactions in the metabolic syndrome: single [92] J. P. Thaler and M. W. Schwartz, “Minireview: inflammation nucleotide polymorphisms in ADIPOQ and ADIPOR1 inter- and obesity pathogenesis: the hypothalamus heats up,” act with plasma saturated fatty acids to modulate insulin Endocrinology, vol. 151, no. 9, pp. 4109–4115, 2010. resistance,” American Journal of Clinical Nutrition, vol. 91, no. [93] D. L. Coleman, “Obese and diabetes: two mutant genes 3, pp. 794–801, 2010. causing diabetes-obesity syndromes in mice,” Diabetologia, [78]D.Zhang,J.Ma,K.Brismar,S.Efendic,andH.F.Gu, vol. 14, no. 3, pp. 141–148, 1978. “A single nucleotide polymorphism alters the sequence of [94] S. A. Ranadive and C. Vaisse, “Lessons from extreme SP1 binding site in the adiponectin promoter region and is human obesity: monogenic disorders,” Endocrinology and associated with diabetic nephropathy among type 1 diabetic Metabolism Clinics of North America, vol. 37, no. 3, pp. 733– patients in the Genetics of Kidneys in Diabetes Study,” 751, 2008. Journal of Diabetes and its Complications, vol. 23, no. 4, pp. [95] Y. Friedlander, G. Li, M. Fornage et al., “Candidate molecular 265–272, 2009. pathway genes related to appetite regulatory neural network, [79]H.Laumen,A.D.Saningong,I.M.Heidetal.,“Functional adipocyte homeostasis and obesity: results from the CARDIA characterization of promoter variants of the adiponectin Study,” Annals of Human Genetics, vol. 74, no. 5, pp. 387–398, gene complemented by epidemiological data,” Diabetes,vol. 2010. 58, no. 4, pp. 984–991, 2009. [96] N. Y. Souren, A. D. Paulussen, A. Steyls et al., “Common SNPs in LEP and LEPR associated with birth weight and [80]W.S.Yang,P.L.Tsou,W.J.Leeetal.,“Allele-specific type 2 diabetes-related metabolic risk factors in twins,” differential expression of a common adiponectin gene poly- International Journal of Obesity, vol. 32, no. 8, pp. 1233–1239, morphism related to obesity,” Journal of Molecular Medicine, 2008. vol. 81, no. 7, pp. 428–434, 2003. [97] A. Meirhaeghe, D. Cottel, P. Amouyel, and J. Dallongeville, [81] C. Menzaghi, T. Ercolino, R. D. Paola et al., “A haplotype at “Lack of association between certain candidate gene poly- the adiponectin locus is associated with obesity and other morphisms and the metabolic syndrome,” Molecular Genetics features of the insulin resistance syndrome,” Diabetes,vol.51, and Metabolism, vol. 86, no. 1-2, pp. 293–299, 2005. no. 7, pp. 2306–2312, 2002. [98] R. Lucantoni, E. Ponti, M. E. Berselli et al., “The A19G [82] F. Abbasi, J. W. Chu, C. Lamendola et al., “Discrimination polymorphism in the 5’ untranslated region of the human between obesity and insulin resistance in the relationship obese gene does not affect leptin levels in severely obese with adiponectin,” Diabetes, vol. 53, no. 3, pp. 585–590, 2004. patients,” Journal of Clinical Endocrinology and Metabolism, [83] P. Perez-Mart´ ınez,´ J. Lopez-Miranda,´ C. Cruz-Teno et al., vol. 85, no. 10, pp. 3589–3591, 2000. “Adiponectin gene variants are associated with insulin sen- [99] J. Hager, K. Clement, S. Francke et al., “A polymorphism in sitivity in response to dietary fat consumption in caucasian the 5’ untranslated region of the human ob gene is associated men,” Journal of Nutrition, vol. 138, no. 9, pp. 1609–1614, with low leptin levels,” International Journal of Obesity,vol. 2008. 22, no. 3, pp. 200–205, 1998. [84]D.Warodomwichit,J.Shen,D.K.Arnettetal.,“ADIPOQ [100] M. L. Hart Sailors, A. R. Folsom, C. M. Ballantyne et polymorphisms, monounsaturated fatty acids, and obesity al., “Genetic variation and decreased risk for obesity in risk: the GOLDN study,” Obesity, vol. 17, no. 3, pp. 511–517, the Atherosclerosis Risk in Communities Study,” Diabetes, 2009. Obesity and Metabolism, vol. 9, no. 4, pp. 548–557, 2007. [85] H. K. Chung, J. S. Chae, Y. Hyun et al., “Influence of [101] Y. Jiang, J. B. Wilk, I. Borecki et al., “Common variants in adiponectin gene polymorphisms on adiponectin level and the 5 region of the leptin gene are associated with body insulin resistance index in response to dietary intervention mass index in men from the National Heart, Lung, and Blood in overweight-obese patients with impaired fasting glucose Institute Family Heart Study,” American Journal of Human or newly diagnosed type 2 diabetes,” Diabetes Care, vol. 32, Genetics, vol. 75, no. 2, pp. 220–230, 2004. no. 4, pp. 552–558, 2009. [102] O. Mammes,` D. Betoulle, R. Aubert, B. Herbeth, G. Siest, and [86] K. Tsuzaki, K. Kotani, N. Nagai et al., “Adiponectin gene F. Fumeron, “Association of the G-2548A polymorphism in single-nucleotide polymorphisms and treatment response to the 5 region of the LEP gene with overweight,” Annals of obesity,” Journal of Endocrinological Investigation,vol.32,no. Human Genetics, vol. 64, no. 5, pp. 391–394, 2000. 5, pp. 395–400, 2009. [103] O. Portoles,J.V.Sorl´ ı,´ F. Frances´ et al., “Effect of genetic [87] T. Skurk, C. Alberti-Huber, C. Herder, and H. Hauner, variation in the leptin gene promoter and the leptin receptor “Relationship between adipocyte size and adipokine expres- gene on obesity risk in a population-based case-control study sion and secretion,” Journal of Clinical Endocrinology and in Spain,” European Journal of Epidemiology,vol.21,no.8,pp. Metabolism, vol. 92, no. 3, pp. 1023–1033, 2007. 605–612, 2006. Journal of Obesity 17

[104] D. Murugesan, T. Arunachalam, V. Ramamurthy, and S. Sub- [117] J. M. Guızar-Mendoza,N.Amador-Licona,S.E.Flores-´ ramanian, “Association of polymorphisms in leptin receptor Martınez,M.G.L´ opez-Cardona,´ R. Ahuatzin-Tremary,´ and gene with obesity and type 2 diabetes in the local population J. Sanchez-Corona,´ “Association analysis of the Gln223Arg of Coimbatore,” Indian Journal of Human Genetics, vol. 16, polymorphism in the human leptin receptor gene, and traits pp. 72–77, 2010. related to obesity in Mexican adolescents,” Journal of Human [105] N. D. Quinton, A. J. Lee, R. J. M. Ross, R. Eastell, and A. Hypertension, vol. 19, no. 5, pp. 341–346, 2005. I. F. Blakemore, “A single nucleotide polymorphism (SNP) [118] T. Salopuro, L. Pulkkinen, J. Lindstrom¨ et al., “Genetic in the leptin receptor is associated with BMI, fat mass and variation in leptin receptor gene is associated with type 2 leptin levels in postmenopausal Caucasian women,” Human diabetes and body weight: the Finnish Diabetes Prevention Genetics, vol. 108, no. 3, pp. 233–236, 2001. Study,” International Journal of Obesity, vol. 29, no. 10, pp. [106] V. S. Mattevi, V. M. Zembrzuski, and M. H. Hutz, “Asso- 1245–1251, 2005. ciation analysis of genes involved in the leptin-signaling [119] L. Gallicchio, H. H. Chang, D. K. Christo et al., “Single pathway with obesity in Brazil,” International Journal of nucleotide polymorphisms in obesity-related genes and all- Obesity, vol. 26, no. 9, pp. 1179–1185, 2002. cause and cause-specific mortality: a prospective cohort [107] N. Yiannakouris, M. Yannakoulia, L. Melistas, J. L. Chan, D. study,” BMC Medical Genetics, vol. 10, article no. 103, 2009. Klimis-Zacas, and C. S. Mantzoros, “The Q223R polymor- [120] O. Mammes,` R. Aubert, D. Betoulle et al., “LEPR gene phism of the leptin receptor gene is significantly associated polymorphisms: associations with overweight, fat mass and with obesity and predicts a small percentage of body response to diet in women,” European Journal of Clinical weight and body composition variability,” Journal of Clinical Investigation, vol. 31, no. 5, pp. 398–404, 2001. Endocrinology and Metabolism, vol. 86, no. 9, pp. 4434–4439, [121] Y. J. Liu, S. M. S. Rocha-Sanchez, P. Y. Liu et al., “Tests of 2001. linkage and/or association of the LEPR gene polymorphisms [108] N. Stefan, B. Vozarova, A. Del Parigi et al., “The Gln223Arg with obesity phenotypes in Caucasian nuclear families,” polymorphism of the leptin receptor in Pima Indians: Physiological Genomics, vol. 17, pp. 101–106, 2004. influence on energy expenditure, physical activity and lipid [122] M. Wauters, I. Mertens, T. Rankinen, M. Chagnon, C. metabolism,” International Journal of Obesity, vol. 26, no. 12, Bouchardt, and L. Van Gaal, “Leptin receptor gene poly- pp. 1629–1632, 2002. morphisms are associated with insulin in obese women with [109] K. C. Chiu, A. Chu, L. M. Chuang, and M. F. Saad, impaired glucose tolerance,” Journal of Clinical Endocrinology “Association of leptin receptor polymorphism with insulin and Metabolism, vol. 86, no. 7, pp. 3227–3232, 2001. resistance,” European Journal of Endocrinology, vol. 150, no. [123] K. S. Park, H. D. Shin, B. L. Park et al., “Polymorphisms in the 5, pp. 725–729, 2004. leptin receptor (LEPR)—putative association with obesity [110] G. M. Van Der Vleuten, L. A. Kluijtmans, A. Hijmans, H. J. and T2DM,” Journal of Human Genetics,vol.51,no.2,pp. Blom, A. F. H. Stalenhoef, and J. De Graaf, “The Gln223Arg 85–91, 2006. polymorphism in the leptin receptor is associated with [124] A. Palou, J. Sanchez,´ and C. Pico,´ “Nutrient-gene interactions familial combined hyperlipidemia,” International Journal of in early life programming: leptin in breast milk prevents Obesity, vol. 30, no. 6, pp. 892–898, 2006. obesity later on in life,” Advances in Experimental Medicine [111] S. F. Pimentel Duarte, E. A. Francischetti, V. Genelhu-Abreu and Biology, vol. 646, pp. 95–104, 2009. et al., “p.Q223R leptin receptor polymorphism associated [125] F. Savino and S. A. Liguori, “Update on breast milk hor- with obesity in Brazilian multiethnic subjects,” American mones: leptin, ghrelin and adiponectin,” Clinical Nutrition, Journal of Human Biology, vol. 18, no. 4, pp. 448–453, 2006. vol. 27, no. 1, pp. 42–47, 2008. [112] M. Wauters, I. Mertens, M. Chagnon et al., “Polymorphisms [126] S. G. Bouret, “Leptin, nutrition, and the programming of in the leptin receptor gene, body composition and fat hypothalamic feeding circuits,” Nestle Nutrition Workshop distribution in overweight and obese women,” International Series: Pediatric Program, vol. 65, pp. 25–39, 2010. Journal of Obesity, vol. 25, no. 5, pp. 714–720, 2001. [127] C. M. Phillips, L. Goumidi, S. Bertrais et al., “Leptin receptor [113] T. Ogawa, H. Hirose, Y. Yamamoto et al., “Relationships polymorphisms interact with polyunsaturated fatty acids to between serum soluble leptin receptor level and serum leptin augment risk of insulin resistance and metabolic syndrome and adiponectin levels, insulin resistance index, lipid profile, in adults,” Journal of Nutrition, vol. 140, no. 2, pp. 238–244, and leptin receptor gene polymorphisms in the Japanese 2010. population,” Metabolism, vol. 53, no. 7, pp. 879–885, 2004. [128] D. A. De Luis, R. Aller, O. Izaola, M. Gonzalez Sagrado, [114] U. L. Fairbrother, L. B. Tanko,´ A. J. Walley, C. Christiansen, P. and R. Conde, “Influence of Lys656Asn polymorphism of Froguel, and A. I. F. Blakemore, “Leptin receptor genotype leptin receptor gene on leptin response secondary to two at Gln223Arg is associated with body composition, BMD, hypocaloric diets: a randomized clinical trial,” Annals of and vertebral fracture in postmenopausal Danish women,” Nutrition and Metabolism, vol. 52, no. 3, pp. 209–214, 2008. Journal of Bone and Mineral Research, vol. 22, no. 4, pp. 544– [129] V. DeClercq, D. Stringer, R. Hunt, C. G. Taylor, and P. 550, 2007. Zahradka, “Adipokine production by adipose tissue: a novel [115] R. Rosmond, Y. C. Chagnon, G. Holm et al., “Hypertension target for treating metabolic syndrome and its sequelae,” in obesity and the leptin receptor gene locus,” Journal of in Metabolic Syndrome: Underlying Mechanisms and Drug Clinical Endocrinology and Metabolism,vol.85,no.9,pp. Therapies, M. Wang, Ed., pp. 73–131, John Wiley & Sons, 3126–3131, 2000. Hoboken, NJ, USA, 2011. [116] Y. C. Chagnon, W. K. Chung, L. Perusse,M.Chagnon,´ R. L. Leibel, and C. Bouchard, “Linkages and associations between the leptin receptor (LEPR) gene and human body composition in the Quebec Family Study,” International Journal of Obesity, vol. 23, no. 3, pp. 278–286, 1999. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 307542, 7 pages doi:10.1155/2011/307542

Research Article Associations of FTO and MC4R Variants with Obesity Traits in Indians and the Role of Rural/Urban Environment as a Possible Effect Modiﬁer

A. E. Taylor,1, 2 M. N. Sandeep,3 C. S. Janipalli,3 C. Giambartolomei,4, 5 D. M. Evans,1, 2 M. V. Kranthi Kumar,3 D. G. Vinay,3 P. Smi t ha , 3 V. Gupta, 6 M. Aruna,3 S. Kinra,4, 5 R. M. Sullivan,4 L. Bowen,4 N. J. Timpson,1, 2 G. Davey Smith,1, 2 F. Dudbridge, 4, 5 D. Prabhakaran,6, 7, 8 Y. Ben-Shlomo, 1 K. S. Reddy,8 S. Ebrahim,4, 5, 6 and G. R. Chandak3

1 School of Social and Community Medicine, University of Bristol, Canynge Hall, 39 Whatley Road, Bristol BS8 2PS, UK 2 MRC Centre for Causal Analyses in Translational Epidemiology, University of Bristol, Bristol BS8 2BN, UK 3 Centre for Cellular and Molecular Biology (CCMB), Council of Scientiﬁc and Industrial Research (CSIR), Hyderabad, India 4 Faculty of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, London WC1E 7HT, UK 5 Bloomsbury Centre for Genetic Epidemiology and Statistics, London WC1E 6BT, UK 6 South Asia Network for Chronic Disease, Public Health Foundation of India, New Delhi 110 016, India 7 Centre for Chronic Disease Control, New Delhi 110 016, India 8 Public Health Foundation of India, New Delhi 110 016, India

Correspondence should be addressed to A. E. Taylor, [email protected] and G. R. Chandak, [email protected]

Received 2 December 2010; Accepted 4 March 2011

Academic Editor: Yvon Chagnon

Copyright © 2011 A. E. Taylor et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Few studies have investigated the association between genetic variation and obesity traits in Indian populations or the role of environmental factors as modifiers of these relationships. In the context of rapid urbanisation, resulting in significant lifestyle changes, understanding the aetiology of obesity is important. We investigated associations of FTO and MC4R variants with obesity traits in 3390 sibling pairs from four Indian cities, most of whom were discordant for current dwelling (rural or urban). The FTO variant rs9939609 predicted increased weight (0.09 Z-scores, 95% CI: 0.03, 0.15) and BMI (0.08 Z-scores, 95% CI: 0.02, 0.14). The MC4R variant rs17782313 was weakly associated with weight and hip circumference (P<.05). There was some indication that the association between FTO and weight was stronger in urban than that in rural dwellers (P for interaction = .03), but no evidence for effect modification by diet or physical activity. Further studies are needed to investigate ways in which urban environment may modify genetic risk of obesity.

1. Introduction reported in several European populations [2, 7, 8]. These associations have been replicated in a UK Indian population A large number of common genetic variants have been found and a Sikh population in north India [9, 10]. In both studies, to be associated with obesity phenotypes in Europeans [1–3]. risk allele frequencies were higher than in Europeans. Variants in the FTO (fat mass and obesity-associated) gene The pathways by which these genetic variants contribute have demonstrated the strongest associations with obesity in to obesity are not yet understood but there is some evidence Europeans [1, 4, 5]. However, results from a recent study in that the effects of variants in FTO are modified by energy Indians from Pune (western India) suggest that FTO may imbalance [11–14]. Several studies have shown that increases be less strongly associated with obesity in South Asians [6]. in body mass index (BMI) per risk allele are lower in people Recently, association of genetic variants near the melano- who engage in high levels of physical activity compared to cortin 4 receptor (MC4R) gene with obesity traits has been less active people [11, 12, 15]. FTO has also been shown 2 Journal of Obesity to be associated with dietary fat intake and overall energy [31]. Weight was measured in light indoor clothing (with consumption [13, 16–18]. However, the modifying effects shoes removed) using a digital weighing scale with 100 g of these environmental factors have not been consistently accuracy (Model PS16, Beurer, Germany) and standing replicated [19–21]. This may reflect true population dif- height using a portable plastic stadiometer (Leicester height ferences, study sample size, or interstudy heterogeneity of measure, Chasmors Ltd, London) [24]. Waist and hip cir- measurement of these lifestyle factors. In addition, within cumferences were measured twice using a non-stretch metal- population variation of these environmental factors may not lic tape with a blank lead-in (Chasmors metallic tape, Chas- be large enough to detect effect modification. mors Ltd, London). Waist circumference was measured at the India is currently experiencing rapid urbanisation, which narrowest part of the abdomen between the ribs and the iliac is leading to the adoption of significant lifestyle changes [22]; crest and hip circumference at the widest part of the hips. in 2001, almost a third of the population lived in urban areas, Percentage body fat was calculated using standard formulae but it is estimated that by 2025, half the population will be [32] from triceps and subscapular skinfold measures, which urban dwellers [23]. We have shown previously that rural were taken three times using Holtain calipers. to urban migration is associated with dietary changes such as increased fat intake and reductions in levels of physical 2.2.2. Dwelling. Urban/rural status was defined according to activity [24]. These factors are likely to contribute to the current place of residence, since little difference was observed markedly higher levels of obesity and diabetes seen in urban between rural-urban migrants and whole life urban dwellers compared to rural populations [25]. Prevalence of obesity in terms of their lifestyle and cardiovascular risk factors. The > 2 (BMI 25 kg/m ) has been reported to be between 25 and rural-urban migrants had lived, on average, for two decades 42% in urban areas compared to 10–22% in rural areas [24, in an urban area, with 85% having migrated at least 10 years ffi 26–28]. Rural/urban living may, therefore, be a su ciently prior to the study [24]. strong exposure to demonstrate large interaction effects with genetic factors. The Indian Migration Study was set up to investigate 2.2.3. Lifestyle Factors. Both diet and physical activity were the impact of rural to urban migration on obesity and assessed by interview-administered questionnaires. Full diabetes [29]. Using data on 3390 sibling pairs, we aimed to details have been published previously [24]. Dietary fat replicate associations of key variants in FTO (rs9939609) and intake (g/day) was calculated from a food frequency ques- near MC4R (rs12970134 and rs17782313) with a range of tionnaire, which asked participants about consumption ff obesity phenotypes and investigate whether urban or rural (daily, weekly, monthly, or yearly) of 184 di erent food environment alters the strength of these associations in this items. Metabolic equivalent tasks (MET) scores were calcu- Indian population. We hypothesised that, if there is evidence lated from participants’ accounts of their activities in the of modification by environment, genetic effects would be previous month. Activity data were summarised as MET strongerinurbanthaninruraldwellers. hours per day, with 1 MET being the equivalent to the energy expended whilst sitting quietly. Time spent doing moderate to vigorous physical activity (MVPA) (defined 2. Material and Methods as: moderate 3–6 MET; vigorous > 6 MET) in a 24-hour 2.1. Study Population. The Indian Migration Study was nest- period, and the average daily MET scores for each individual ed within a cardiovascular risk factor surveillance system, from moderate to vigorous activities were calculated. Both which monitors risk factors in industry populations across dietary and physical activity measures showed acceptable several large cities in India [30]. Full details of the study re- validity. cruitment have been described previously [29]. Genetic data was available for 6942 individuals recruited from four cities 2.2.4. Genotyping. Blood samples were collected from all the in India (Lucknow, Nagpur, Hyderabad, and Bangalore), individuals using EDTA vacutainers, and plasma samples of which 6780 were sibling pairs and 162 were unrelated were stored at −80◦C for further use. All packed cell samples individuals (147 from cousins (one half of cousin pairs) were transported on dry ice to Centre for Cellular and and friends, and 15 from single members of a sibling pair). Molecular Biology, Hyderabad, India. Genomic DNA was The majority of sibling pairs (N = 1997 pairs) consisted isolated from all samples using salt precipitation method of a rural to urban migrant factory worker and a sibling and DNA samples were plated in 96 deep well storage who had remained in a rural area. There were also 1152 plates at a uniform concentration of 10 ng/λ.Eachplate urban-urban sibling pairs and 241 rural-rural sibling pairs. included 8 repeat samples (∼10%) as a quality control Ethical approval was obtained from the All India Institute measure. We used Sequenom-based Mass ARRAY assay tech- of Medical Sciences Ethics Committee, reference number nology to genotype three obesity-related single-nucleotide A-60/4/8/2004. Fieldwork commenced in March 2005 and polymorphisms (SNPs), one in FTO (rs9939609) and two ended in December 2007. near MC4R (rs12970134 and rs17782313) as part of a common multiplex pool of 28 SNPs, collated from 20 genes 2.2. Measurements associated with diabetes-related intermediate traits. The genotyping success rate was >95%, and results of duplicate 2.2.1. Anthropometry. All participants were invited to attend samples had >97% concordance indicating high genotyping for an examination at the factory they were recruited from accuracy. Journal of Obesity 3

2.3. Statistical Analyses. Genotype frequencies were cal- rs17782313 (C allele) were 0.33, 0.36, and 0.34, respectively, culated in unrelated subjects (a single randomly chosen in this study. MAFs in each of the four cities are shown in member from each family) (N = 3552) and tested for depar- Supplementary Tables S2–S5. There was no strong evidence ture from Hardy-Weinberg proportions using the exact test that genotype frequency differed by city for either of the implemented in PLINK (version 0.99p; (http://pngu.mgh MC4R SNPs (rs12970134 and rs17782313) (P>.09) but .harvard.edu/∼purcell/plink/)) [33]. evidence of genotype frequency difference was seen for the Association analysis was carried out using the orthogonal rs9939609 variant (P = .001). In the sample stratified by the family-based model of Fulker et al. [34]. This is a linear re- four cities there was evidence of a deviation from HWE for gression model in which the genetic effect is decomposed rs9939609 in the Bangalore sample (P = .0001). into between- and within-family effects, with inference Data from 3390 sibling pairs were included in the performed on the within-family effect. This family based main analyses. The characteristics of the study population method is robust to population stratification, an important stratified by sex and urban/rural location are shown in consideration since the data were collected from individuals Table 1. We observed strong evidence for differences in of at least two broad ethnic groups in four cities across India, obesity traits and lifestyle characteristics between individuals where high levels of population substructure exist [35]. in rural and urban settings. Further analyses stratified by Within this framework, multilevel models were fitted to BMI (≤25 kg/m2, >25 kg/m2) are presented in Supplemen- the data in which sibling pair was modelled as a random tary Table S6; these show strong associations between BMI effect and city of recruitment as the fixed effect. Analysis was and lifestyle characteristics and cardiovascular and metabolic restricted to full sib pairs (N = 3390 pairs), and additive outcomes. models were assumed; for each SNP the major allele was the reference, and an effect estimate was calculated per copy 3.1. Association of FTO and MC4R Variants with Obesity- of the minor allele. Associations between SNPs and age, Related Traits. We found evidence that the rs9939609 SNP sex adjusted Z scores of the quantitative phenotypes BMI, was associated with BMI and weight (Table 2). BMI increased WHR, percent body fat, weight, waist circumference, and hip on average by 0.08 Z-scores (95% CI 0.02, 0.14) and weight circumference were tested using a Wald test. We also tested by 0.09 Z-scores (95% CI 0.03, 0.15) per copy of the “A”allele. separately for interaction between the within families genetic There was no strong evidence of association of this SNP effect and sex and location (rural/urban) as effect modifiers with other obesity phenotypes such as waist circumference in the models. We previously found differences between the or WHR. When these analyses were repeated excluding the sexes for lipids, glucose, and blood pressure [24]. In addition, Bangalore sample, effect sizes were largely unaltered. we investigated whether there was evidence of interaction by Since the two MC4R related SNPs are in strong linkage dietary fat intake or physical activity measures (total MET, disequilibrium (r2 = 0.895 in the Gujarati Indian in Houston time spent on MVPA (min/day), MET from MVPA) (all in (GIH) population in HapMap 3) (http://hapmap.ncbi.nlm tertiles) by including these as interaction terms in separate .nih.gov/), only the results for rs17782313 are presented. The models. Mixed effects logistic regression was performed to results for rs12970134 are presented in Supplementary Tables investigate the association of these SNPs with obesity, which and are similar, although associations with obesity traits are was defined as BMI > 25 kg/m2 (in line with previously used slightly weaker than for rs17782313 (Supplementary Table cutoffs for obesity in Indians) [24]. S7). Each additional copy of the “C” allele at the rs17782313 We fitted all models in STATA (version 11.1) as the QTDT SNP was associated with a 0.06 Z-score (95% CI 0.001, software [36], widely used for the Fulker model, does not 0.12) increase in weight and a 0.06 Z-score (95% CI 0.01, implement effect modification. To verify our implemen- 0.12) increase in hip circumference. There was no strong tation, we compared the STATA results to QTDT for the statistical evidence for associations with other obesity-related analysis of main effects only and to MX [37]forthe traits, but effect sizes were in the expected direction (apart analysis of main effects and interactions (for full meth- from WHR). Associations of the FTO and MC4R SNPs with ods, see Supplementary Material available online at doi: obesity (BMI > 25 kg/m2) are shown in Supplementary Table 10.1155/2011/307542) and obtained similar results through- S8. There was no evidence that the FTO SNP was associated out. As another check, we repeated all analyses using the with obesity (OR 1.08, 95% CI (0.91, 1.28), P = .39) but UNPHASED software [38] and obtained similar results, some weak evidence that the MC4R SNP was associated with although in a number of cases UNPHASED had numerical increased odds of obesity (1.19, 95% CI (1.00, 1.40), P = difficulties in maximizing its likelihood. .05).

3. Results 3.2. Interaction Analyses between FTO and MC4R Variants and Demographic Characteristics. When we tested for inter- Genetic data was available for a total of 6942 individuals from action by sex, there was some weak evidence that genetic all four cities. In the whole sample of unrelated individuals effects of FTO on weight were stronger in females than in (N = 3552), there was no evidence for deviation from Hardy males (Effect difference for females compared to males: 0.12 Weinberg equilibrium (HWE) for any of the SNPs (P all Z-scores, 95% CI: −0.01, 0.26) (P = .07) (Supplementary > .3) (Supplementary Table S1). Minor allele frequencies Table S9). The same pattern was seen for the other obesity (MAFs) for rs9939609 (A allele), rs12970134 (A allele), and traits (apart from WHR) with FTO but with no strong 4 Journal of Obesity

Table 1: Characteristics of the study population.

All1 Males1 Females1 Urban Rural P2 Urban Rural P2 N 6780 2276 1649 2025 830 Sex (% Female) 42 Urbanisation (% urban) 63 Age (years) 40.7 (0.13) 42.9 (0.20) 39.5 (0.28) <.001 39.5 (0.20) 40.2 (0.40) .12 Height (cm) 160.3 (0.11) 165.8 (0.14) 165.6 (0.15) .52 153.1 (0.13) 152.4 (0.20) .001 Weight (kg) 61.1 (0.15) 66.7 (0.24) 59.5 (0.28) <.001 59.6 (0.27) 52.5 (0.39) <.001 BMI (kg/m2) 23.8 (0.05) 24.3 (0.08) 21.7 (0.09) <.001 25.4 (0.11) 22.6 (0.16) <.001 Waist circumference (cm) 82.3 (0.14) 87.9 (0.23) 80.5 (0.28) <.001 80.2 (0.24) 75.2 (0.38) <.001 Hip circumference (cm) 94.3 (0.11) 94.7 (0.16) 90.0 (0.19) <.001 98.0 (0.24) 92.8 (0.35) <.001 WHR 0.87 (0.001) 0.93 (0.001) 0.89 (0.002) <.001 0.82 (0.001) 0.81 (0.002) .001 Percentage Body fat (%) 26.8 (0.10) 25.9 (0.14) 20.8 (0.18) <.001 31.9 (0.17) 29.1 (0.26) <.001 % Diabetic 6.8 9.8 3.6 <.001 7.1 4.3 .005 Dietary Fat intake (g/day) 83.1 (0.43) 96.0 (0.80) 79.9 (0.92) <.001 79.6 (0.67) 62.7 (1.01) <.001 Daily average MET score 38.8 (0.06) 38.4 (0.08) 41.2 (0.14) <.001 37.5 (0.08) 38.4 (0.17) <.001 BMI: bodymass index; WHR: waist-hip ratio; MET: metabolic equivalent tasks. 1Data presented as mean (standard error) or percentage (binary variables). 2P value for diﬀerence between rural and urban samples from linear and logistic regression (for diabetes) with robust standard errors to account for sibling pairs.

Table 2: Associations of SNPs in FTO and MC4R genes with age, sex adjusted Z scores of obesity traits.

rs9939609 rs17782313 N Coeffa 95% CI PNCoeffa 95% CI P BMI 6170 0.08 (0.02, 0.14) .009 6240 0.04 (−0.01, 0.10) .14 WHR 6160 0.01 (−0.05, 0.07) .77 6230 −0.01 (−0.07, 0.05) .75 Waist circumference 6160 0.04 (−0.02, 0.11) .16 6230 0.04 (−0.02, 0.10) .22 Weight 6170 0.09 (0.03, 0.15) .003 6240 0.06 (0.001, 0.12) .045 Hip circumference 6168 0.05 (−0.01, 0.11) .08 6238 0.06 (0.01, 0.12) .03 Body fat 5968 0.02 (−0.04, 0.08) .57 6028 0.05 (−0.01, 0.11) .08 BMI, body mass index; WHR, waist-hip ratio aCoefficents represent SD change per minor allele. statistical support. No evidence for interaction with sex was rs17782313 with weight and hip circumference, but did not seen with rs17782313. find statistical evidence for associations with other obesity- The results of interaction analyses, looking at rural/urban related traits. This study is the first to investigate gene envi- dwelling as a potential effect modifier of genetic associations, ronment interactions for FTO and obesity traits in an Indian are shown in Figure 1 and Table 3. The association of FTO population. We found some evidence that the strength of the with weight appeared to be stronger in urban dwellers com- association between FTO and weight was stronger in urban pared to rural habitants, with urban living associated with an dwellers compared to rural dwellers suggesting that the increase of 0.15 Z scores (95% CI 0.01, 0.29) per minor allele effect of FTO on weight may be modified by environmental of rs9939609 more than rural dwelling. There was no strong differences between rural and urban living in India. evidence for any interactions between rural/urban dwelling In this population, the minor allele of rs9939609 was and obesity-related traits with SNPs near MC4R (Table 3 and associated with increases in both BMI and weight (effect sizes Supplementary Table S10). 0.08 and 0.09 Z-scores, resp.) but not with regional measures Neither physical activity measures (total MET, time spent of adiposity (waist circumference, hip circumference, and on MVPA and total MET from MVPA) nor dietary fat intake WHR) or with percentage body fat derived from skinfolds. had any influence on the association of the FTO variant with In a population from Pune, western India [6], the reported BMI and weight (Supplementary Table S11). association of BMI with the same FTO variant was 0.06 Z scores (95% CI: 0.01–0.10) which sits in the context of 4. Discussion known associations in European populations (0.1 Z scores (95% CI: 0.09, 0.12) from the meta-analysis of Frayling Our study has replicated associations of the FTO SNP et al.) [1]. Based on the observation that controlling for rs9939609 with BMI and weight and the MC4R SNP BMI did not completely abolish the association of rs9939609 Journal of Obesity 5

Table 3: Interactions between FTO and MC4R SNPs and rural/urban dwelling in associations with obesity traits.

rs9939609 rs17782313 Coeffa 95% CI P Coeffa 95% CI P BMI 0.07 (−0.07, 0.21) .30 −0.003 (−0.14, 0.13) .96 WHR 0.003 (−0.14, 0.15) .97 −0.03 (−0.17, 0.11) .67 Waist circumference 0.09 (−0.05, 0.23) .21 −0.03 (−0.17, 0.11) .65 Weight 0.15 (0.01, 0.29) .03 0.02 (−0.12, 0.16) .77 Hip circumference 0.11 (−0.03, 0.25) .12 −0.006 (−0.14, 0.13) .93 Body fat 0.06 (−0.08, 0.20) .37 −0.05 (−0.19, 0.09) .47 BMI: body mass index; WHR: waist-hip ratio. aCoefficients represent differences in age, sex adjusted SD scores per minor allele in urban compared to rural dwellers.

0.2 are probably measured with less error than circumference or skinfold measures. However, these results still need to be interpreted carefully since BMI does not reflect the actual fat 0.1 mass percent in Indians [43]. Only associations with weight and hip circumference were successfully replicated with the MC4R-related SNP 0 rs17782313, although effect sizes for other traits (apart from WHR) were in the expected direction. Allele frequencies were similar to those reported previously in Indian populations [9, −0.1 10]. These earlier studies, however, did find strong evidence Per minor allele change (SD score) for associations with the MC4R SNP rs12970134 (which is BMI WHR Waist Weight Hip Bodyfat in strong LD with rs17782313) and BMI, weight, WHR, Trait and waist circumference. Although BMI as a continuous trait did not reach significance, there was some evidence of Rural an association with obesity (BMI > 25 kg/m2). The effect Urban estimate from the quantitative trait analysis in this study Figure 1: Associations between FTO and obesity traits by rural/ur- (0.04 Z-score units) was of a similar magnitude to that found ban location. BMI: body mass index; WHR: waist-hip ratio. in a large meta-analysis of Europeans (0.05 Z-score units) [2] so the lack of evidence for association in this study may reflect lack of power in our study rather than a true null effect. with type 2 diabetes, Yajnik et al. argued that association of Our finding that the effect of FTO on weight was stronger FTO variants may be weaker in Indians than in Europeans in urban compared to rural dwellers is consistent with our [6]. Meta-analyses of genetic studies in Indian populations prior hypothesis that genetic effects would be stronger in the would be of value to assess whether differences in strength of urban environment, where the prevalence of environmental associations with FTO compared to Europeans exist. risk factors for obesity is much higher. Some caution The association of rs9939609 was limited to weight and must be taken when interpreting the interaction results, BMI, suggesting that the FTO gene may be related more to since multiple statistical tests have been performed; similar overall body mass than to regional patterns of adiposity in patterns were seen for BMI, waist, and hip circumferences, Indians. The lack of association with WHR has been reported although there was no statistical evidence for differences with previously for Indians [6, 39]. In Europeans, the associations these traits. The sibling pair design of the Indian Migration of most measures of adiposity with FTO have been found Study allows the impact of migration on health outcomes to to be of similar magnitudes [1]. However, in one study of be studied rather than whole life environmental differences Europeans, the association of waist circumference with FTO between urban and rural living since the sibling pairs have was found to attenuate following adjustment for BMI which shared early life environment. Although we included all is consistent with an association with general rather than types of sibling pairs (both discordant and concordant for regional fat mass [40]. Genetic variants specifically associated current dwelling), most rural dwellers had a rural to urban with regional adiposity measures (waist circumference and migrant sibling in the study, which suggests that the effect WHR) in Europeans have been published recently [3, 41, modification of FTO in this study is related to environmental 42], and it may be that these variants (or others yet to be factors postmigration that is in adulthood. discovered) have a greater influence on regional adiposity Urban dwellers in the Indian Migration Study (urban in Indians than FTO or MC4R. An alternative explanation nonmigrants and rural to urban migrants) had lower daily for the differences in associations between measures may average MET (metabolic equivalent tasks) scores and higher be the precision with which adiposity has been measured. levels of dietary fat intake than rural dwellers, factors which BMI and weight are both relatively easy to measure and have both been shown to be associated with larger effects of 6 Journal of Obesity

FTO on obesity traits [11, 12, 15, 16, 44]. We investigated diet index and predisposes to childhood and adult obesity,” Science, and physical activity separately as potential effect modifiers vol. 316, no. 5826, pp. 889–894, 2007. of FTO to see if a possible gene environment interaction [2]R.J.F.Loos,C.M.Lindgren,S.Lietal.,“Commonvariants by rural/urban dwelling might be attributable to either of near MC4R are associated with fat mass, weight and risk of these specific lifestyle characteristics, but were unable to obesity,” Nature Genetics, vol. 40, no. 6, pp. 768–775, 2008. find evidence of effect modification with these factors. It is [3]C.M.Lindgren,I.M.Heid,J.C.Randalletal.,“Genome-wide association scan meta-analysis identifies three loci influencing unlikely that the study was sufficiently powered to detect adiposity and fat distribution,” PLoS Genetics,vol.5,no.6, interactions with these variables, which are likely to have Article ID e1000508, 2009. been measured with lower precision. [4]B.K.Cornes,P.A.Lind,S.E.Medland,G.W.Montgomery,D. R. Nyholt, and N. G. Martin, “Replication of the association 5. Conclusions of common rs9939609 variant of FTO with increased BMI in an Australian adult twin population but no evidence for gene Genetic variants of obesity which have been found to be by environment (G x E) interaction,” International Journal of important in Europeans do play a role in determining Obesity, vol. 33, no. 1, pp. 75–79, 2009. [5] M. Hakanen, O. T. Raitakari, T. Lehtimaki¨ et al., “FTO obesity phenotypes in Indian populations, although our data genotype is associated with body mass index after the age of suggests that associations of FTO with regional adiposity seven years but not with energy intake or leisure-time physical may be weaker in Indians. The urban environment, which is activity,” Journal of Clinical Endocrinology and Metabolism,vol. already known to be an important contributing factor to the 94, no. 4, pp. 1281–1287, 2009. obesity epidemic [25], may also modify genetic susceptibility [6] C. S. Yajnik, C. S. Janipalli, S. Bhaskar et al., “FTO gene to obesity through interaction with FTO.Variantsnear variants are strongly associated with type 2 diabetes in South MC4R had relatively minor influence on these obesity- Asian Indians,” Diabetologia, vol. 52, no. 2, pp. 247–252, 2009. related phenotypes. Further dissection of the specific aspects [7] A. Tenesa, H. Campbell, E. Theodoratou et al., “Common of migration which increase risk of obesity and subsequent genetic variants at the MC4R locus are associated with obesity, cardiovascular disease is needed to target interventions. but not with dietary energy intake or colorectal cancer in the Scottish population,” International Journal of Obesity, vol. 33, no. 2, pp. 284–288, 2009. The Indian Migration Study Group Comprises [8]D.P.Zobel,C.H.Andreasen,N.Grarupetal.,“Variants near MC4R are associated with obesity and influence obesity- New Delhi: K. Srinath Reddy, Dorairaj Prabhakaran, Tulsi related quantitative traits in a population of middle-aged Patel, Lakshmy Ramakrishnan, Ruby Gupta, and Tanica Lyn- people: studies of 14,940 danes,” Diabetes,vol.58,no.3,pp. gdoh;Lucknow:R.C.Ahuja,R.K.Saran;Nagpur:Prashant 757–764, 2009. Joshi,andN.M.Thakre;Hyderabad:K.V.R.Sarma,S. [9] J. C. Chambers, P. Elliott, D. Zabaneh et al., “Common genetic MohanDas,R.KJain,S.S.Potnis,andS.Bhogadi;Banga- variation near MC4R is associated with waist circumference lore: Anura V. Kurpad, Mario Vaz, A. V. Barathi, and Murali and insulin resistance,” Nature Genetics, vol. 40, no. 6, pp. 716– 718, 2008. Mohan; Pune: Chittaranjan Yajnik; Bristol: George Davey [10] L. F. Been, S. K. Nath, S. K. Ralhan et al., “Replication of Smith and Yoav Ben-Shlomo; London School of Hygiene & association between a common variant near melanocortin- Tropical Medicine: Shah Ebrahim and Sanjay Kinra. 4 receptor gene and obesity-related traits in Asian sikhs,” Obesity, vol. 18, no. 2, pp. 425–429, 2010. Acknowledgments [11] C. H. Andreasen, K. L. Stender-Petersen, M. S. Mogensen et al., “Low physical activity accentuates the effect of the Theauthorsaregratefultofieldstaff conducting the mi- FTO rs9939609 polymorphism on body fat accumulation,” gration study and to the participants. The Indian Migration Diabetes, vol. 57, no. 1, pp. 95–101, 2008. [12] E. Rampersaud, B. D. Mitchell, T. I. Pollin et al., “Physical Study was funded by the Wellcome Trust (Grant no. activity and the association of common FTO gene variants GR070797MF). The genetic work was funded by a Project with body mass index and obesity,” Archives of Internal Grant from the Wellcome Trust (083541/Z/07/Z). Amy Tay- Medicine, vol. 168, no. 16, pp. 1791–1797, 2008. lor is funded by a Wellcome Trust 4-year Ph.D. Studentship [13] J. E. Cecil, R. Tavendale, P. Watt, M. M. Hetherington, and (Grant no. WT083431MA). The funder had no role in study C. N. A. Palmer, “An obesity-associated FTO gene variant and design, data collection, analysis, interpretation, writing the increased energy intake in children,” The New England Journal report, or the decision to submit the article for publication. of Medicine, vol. 359, no. 24, pp. 2558–2566, 2008. The researchers are all independent from the funding source. [14] J. Fischer, L. Koch, C. Emmerling et al., “Inactivation of the Dr. Timpson is funded by and Prof. Davey Smith works Fto gene protects from obesity,” Nature, vol. 458, no. 7240, pp. within the MRC Centre for Causal Analyses in Translation- 894–898, 2009. al Epidemiology, which is capacity funded by Grant no. [15] K. S. Vimaleswaran, S. Li, J. H. Zhao et al., “Physical activity G0600705. attenuates the body mass index-increasing influence of genetic variationintheFTOgene,”American Journal of Clinical Nutrition, vol. 90, no. 2, pp. 425–428, 2009. References [16] N. J. Timpson, P. M. Emmett, T. M. Frayling et al., “The fat mass- and obesity-associated locus and dietary intake in [1] T. M. Frayling, N. J. Timpson, M. N. Weedon et al., “A children,” American Journal of Clinical Nutrition, vol. 88, no. common variant in the FTO gene is associated with body mass 4, pp. 971–978, 2008. Journal of Obesity 7

[17] C. Church, L. Moir, F. McMurray et al., “Overexpression of Fto years,” British Journal of Nutrition, vol. 32, no. 1, pp. 79–97, leads to increased food intake and results in obesity,” Nature 1974. Genetics, vol. 42, no. 12, pp. 1086–1092, 2010. [33] S. Purcell, B. Neale, K. Todd-Brown et al., “PLINK: a tool set [18] M. Tanofsky-Kraff,J.C.Han,K.Anandalingametal.,“The for whole-genome association and population-based linkage FTO gene rs9939609 obesity-risk allele and loss of control over analyses,” American Journal of Human Genetics,vol.81,no.3, eating,” American Journal of Clinical Nutrition, vol. 90, no. 6, pp. 559–575, 2007. pp. 1483–1488, 2009. [34]D.W.Fulker,S.S.Cherny,P.C.Sham,andJ.K.Hewitt, [19] G. Liu, H. Zhu, V. Lagou et al., “FTO variant rs9939609 is “Combined linkage and association sib-pair analysis for associated with body mass index and waist circumference, but quantitative traits,” American Journal of Human Genetics,vol. not with energy intake or physical activity in European- and 64, no. 1, pp. 259–267, 1999. African-American youth,” BMC Medical Genetics, vol. 11, no. [35] D. Reich, K. Thangaraj, N. Patterson, A. L. Price, and L. Singh, 1, article 57, 2010. “Reconstructing Indian population history,” Nature, vol. 461, [20] T. J. Lappalainen, A.-M. Tolppanen, M. Kolehmainen et al., no. 7263, pp. 489–494, 2009. “The common variant in the FTO gene did not modify the [36] G.R.Abecasis,L.R.Cardon,andW.O.C.Cookson,“Ageneral effect of lifestyle changes on body weight: the finnish diabetes test of association for quantitative traits in nuclear families,” prevention study,” Obesity, vol. 17, no. 4, pp. 832–836, 2009. American Journal of Human Genetics, vol. 66, no. 1, pp. 279– [21] M. Hakanen, O. T. Raitakari, T. Lehtimaki¨ et al., “FTO 292, 2000. genotype is associated with body mass index after the age of [37] M. Neale, Statistical Modeling with Mx, Department of Human seven years but not with energy intake or leisure-time physical Genetics, MCV, Richmond, Va, USA, 1997. activity,” Journal of Clinical Endocrinology and Metabolism,vol. [38] F. Dudbridge, “Likelihood-based association analysis for 94, no. 4, pp. 1281–1287, 2009. nuclear families and unrelated subjects with missing genotype [22] K. S. Reddy, B. Shah, C. Varghese, and A. Ramadoss, data,” Human Heredity, vol. 66, no. 2, pp. 87–98, 2008. “Responding to the threat of chronic diseases in India,” Lancet, [39] D. Zabaneh and D. J. Balding, “A genome-wide association vol. 366, no. 9498, pp. 1744–1749, 2005. study of the metabolic syndrome in Indian Asian men,” PLoS [23] K. Yadav and A. Krishnan, “Changing patterns of diet, physical ONE, vol. 5, no. 8, Article ID e11961, 2010. activity and obesity among urban, rural and slum populations [40] S. I. I. Kring, C. Holst, E. Zimmermann et al., “FTO gene in North India,” Obesity Reviews, vol. 9, no. 5, pp. 400–408, associated fatness in relation to body fat distribution and 2008. metabolic traits throughout a broad range of fatness,” PLoS [24] S. Ebrahim, S. Kinra, L. Bowen et al., “The effect of rural- ONE, vol. 3, no. 8, Article ID e2958, 2008. to-urban migration on obesity and diabetes in india: a cross- [41] N. L. Heard-Costa, M. Carola Zillikens, K. L. Monda et al., sectional study,” PLoS Medicine,vol.7,no.4,ArticleID “NRXN3Is a novel locus for waist circumference: a genome- e1000268, 2010. wide association study from the CHARGE consortium,” PLoS [25] A. Misra and L. Khurana, “The metabolic syndrome in Genetics, vol. 5, no. 6, Article ID e1000539, 2009. South Asians: epidemiology, determinants, and prevention,” [42]I.M.Heid,A.U.Jackson,J.C.Randalletal.,“Meta- Metabolic Syndrome and Related Disorders,vol.7,no.6,pp. analysis identifies 13 new loci associated with waist-hip ratio 497–514, 2009. and reveals sexual dimorphism in the genetic basis of fat [26] M. Deepa, S. Farooq, M. Datta, R. Deepa, and V. Mohan, distribution,” Nature Genetics, vol. 42, no. 11, pp. 949–960, “Prevalence of metabolic syndrome using WHO, ATPIII and 2010. IDF definitions in Asian Indians: the Chennai Urban Rural [43]M.Deurenberg-Yap,S.K.Chew,andP.Deurenberg,“Elevated Epidemiology Study (CURES-34),” Diabetes/Metabolism body fat percentage and cardiovascular risks at low body mass Research and Reviews, vol. 23, no. 2, pp. 127–134, 2007. index levels among Singaporean Chinese, Malays and Indians,” [27] R. Gupta, P. C. Deedwania, A. Gupta, S. Rastogi, R. B. Panwar, Obesity Reviews, vol. 3, no. 3, pp. 209–215, 2002. and K. Kothari, “Prevalence of metabolic syndrome in an [44] S. Cauchi, F. Stutzmann, C. Cavalcanti-Proença et al., “Com- Indian urban population,” International Journal of Cardiology, bined effects of MC4R and FTO common genetic variants on vol. 97, no. 2, pp. 257–261, 2004. obesity in European general populations,” Journal of Molecular [28]C.Chow,M.Cardona,P.K.Rajuetal.,“Cardiovascular Medicine, vol. 87, no. 5, pp. 537–546, 2009. disease and risk factors among 345 adults in rural India-the Andhra Pradesh Rural Health Initiative,” International Journal of Cardiology, vol. 116, no. 2, pp. 180–185, 2007. [29] T. Lyngdoh, S. Kinra, Y. B. Shlomo et al., “Sib-recruitment for studying migration and its impact on obesity and diabetes,” Emerging Themes in Epidemiology, vol. 3, article 2, 2006. [30] K. S. Reddy, D. Prabhakaran, V. Chaturvedi et al., “Methods for establishing a surveillance system for cardiovascular diseases in Indian industrial populations,” Bulletin of the World Health Organization, vol. 84, no. 6, pp. 461–469, 2006. [31] S. Kinra, L. J. Bowen, T. Lyngdoh et al., “Sociodemographic patterning of non-communicable disease risk factors in rural India: a cross sectional study,” BMJ, vol. 341, Article ID c4974, 2010. [32] J. V. G. A. Durnin and J. Womersley, “Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 861049, 11 pages doi:10.1155/2011/861049

Research Article Can Thrifty Gene(s) or Predictive Fetal Programming for Thriftiness Lead to Obesity?

Ulfat Baig,1 Prajakta Belsare,1 Milind Watve,1, 2 and Maithili Jog3

1 Indian Institute of Science Education and Research, Pune 411021, India 2 Anujeeva Biosciences Pvt. Ltd., Pune 411030, India 3 Department of Biotechnology, Abasaheb Garware College, Pune 411004, India

Correspondence should be addressed to Milind Watve, [email protected]

Received 7 November 2010; Accepted 18 February 2011

Academic Editor: Yvon Chagnon

Copyright © 2011 Ulfat Baig et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Obesity and related disorders are thought to have their roots in metabolic “thriftiness” that evolved to combat periodic starvation. The association of low birth weight with obesity in later life caused a shift in the concept from thrifty gene to thrifty phenotype or anticipatory fetal programming. The assumption of thriftiness is implicit in obesity research. We examine here, with the help of a mathematical model, the conditions for evolution of thrifty genes or fetal programming for thriftiness. The model suggests that a thrifty gene cannot exist in a stable polymorphic state in a population. The conditions for evolution of thrifty fetal programming are restricted if the correlation between intrauterine and lifetime conditions is poor. Such a correlation is not observed in natural courses of famine. If there is fetal programming for thriftiness, it could have evolved in anticipation of social factors aﬀecting nutrition that can result in a positive correlation.

1. Introduction under conditions of better availability of nutrients and allowed reutilization under starvation. The thrifty gene was Obesity and related disorders are on the rise throughout under positive selection pressure in ancestral life when the globe at an alarming rate, the causes of which are not seasonal and climatic conditions resulted into fluctuating yet completely understood. On the one hand, individuals food availability. This concept soon became almost an differ substantially in their tendency to accumulate fat, axiom, although no such “thrifty” gene or set of genes have and there are strong familial tendencies suggesting that been convincingly demonstrated. Later, the observation that genetic predisposition may play a major role. On the individuals born small for gestational age had a greater other hand, it is obvious that no alleles can increase in probability to become obese and type 2 diabetic in later frequency in a population with a rate matching the rate life led to the concept of fetal programming [2–4]. This of spread of the obesity epidemic, indicating that genetics hypothesis states that if a fetus faces inadequate nutrition in alone does not explain the rise in obesity. The most common intrauterine life, the body is programmed to be “thrifty” as perception, therefore, is that of a interplay between genes an adaptation. There are two possible components of this and environment. According to the current paradigm, a adaptation. One relates to an immediate gain in terms of gene or a set of genes predisposing to obesity presumably survival during fetal and early infant life. The other is a evolved owing to a selective advantage in ancestral “feast and predictive adaptive response in anticipation of starvation in famine” environment and remained in polymorphic state in later life [5]. This distinction is important in understanding the population which is turning pathological in the modern the evolution of fetal programming. urban environment selectively affecting individuals with the Both the concepts of thrifty gene and thrifty phenotype gene(s). by fetal programming have recently faced serious criticism This line of thinking was first stated explicitly by Neel [1] on several grounds [6–11]. The main objections of the critics who suggested that a “thrifty” gene helped storage of fat areasfollows. 2 Journal of Obesity

(1) The original suggestion of Neel [1] was that in set in hunter-gatherer life which was effectively lifted when individuals prone to obesity and type 2 diabetes humans became free of predation. Subsequently, the obesity- (T2D), a “quick insulin trigger” ensures rapid glucose related genes started spreading by genetic drift. (ii) Corbett uptake which is then converted into fat. In the et al. [21] argued that today’s obese, diabetic, and PCOS- early 1960s, it was well known that the levels of prone genotype was the ancestral one that had better insulin are higher in prediabetics, and Neel’s original fertility in famine conditions. In the modern era of food proposal looked sound. However, insulin resistance security since 1800 AD, an insulin-sensitive genotype that was discovered soon, and it became clear that the has better fertility under conditions of food abundance “quick insulin trigger” is unlikely to work as believed started spreading. (iii) Moalem et al. [22] hypothesized that by Neel [1]. Fat cell-specific insulin receptor knock- high plasma glucose lowers the freezing point of blood outs fail to accumulate fat [12], raising the possibility which prevents formation of ice crystals in cells through that insulin resistance actually arrests lipogenesis. It supercooling, and this has been suggested as an adaptation is therefore not logical to call the diabetes-prone to the ice age. genotypes as “thrifty.” (b) Alternative explanations for fetal programming have (2) If thriftiness is due to lower metabolic rate conserving also been offered. (i) Hattersley and Tooke [23] argued more energy which gets stored in fat tissue, a lower that the association between low birth weight and insulin metabolic rate should be observed in people having resistance arises out of a reverse causation, that is, babies a predisposition to obesity. Consistent correlations with insulin resistance genotype are more likely to survive between birth weight and metabolic rate are not fetal undernourishment. (ii) Wells [11, 24] argued that there found in empirical data to demonstrate the supposed is maternal advantage in fetal programming in the form of thriftiness of low birth weight individuals [13, 14]. optimizing maternal inputs per fetus or bet hedging, that is, Studies using doubly labeled water have not consis- distribution of risk among offspring. tently found lower metabolic rates in people with (c) There are hypotheses that account for genetic as well sedentary lifestyle in the modern urban societies [15, ff 16]. as intrauterine e ects. (i) Watve and Yajnik [10] argued that insulin resistance is a socioecological adaptation that (3) Impaired fat oxidation rather than lower metabolic mediates two phenotypic transitions, namely, (i) transition rate appears to be the main contributor to obesity in reproductive strategy from “r”(largenumberofoffspring of developmentally stunted individuals [17, 18]. If with little investment in each) to “K” (smaller number of inability to reutilize stored fat is the major cause offspring with more investment in each) and (ii) transition of obesity, the stored fat is unlikely to help under from “soldier to diplomat,” that is, from a physically aggres- “famine” conditions, and this is a major blow to sive behavior to a socially manipulative one. According to this the thriftiness hypotheses. The doubly labeled water hypothesis, insulin resistance changes the differential budget studies also suggest that obesity is more a product of allocation to tissues. Since all tissues are not dependent hyperphagia than metabolic thriftiness [15, 16]. The on insulin for nutrient uptake, when insulin resistance known genetic mechanisms of obesity also work by develops, the uptake of insulin-dependent tissues reduces, interfering with appetite control rather than through and more nutrients become available for insulin independent metabolic thriftiness [19]. tissues. Muscles are insulin dependent, and brain is insulin (4) Evidence that obese people have a significantly better independent, and, therefore, insulin resistance results in chance of surviving famines is debatable [8]. There- disinvestment from muscles and increased investment in fore, it is doubtful whether obesity actually offered brain. Insulin resistance is likely to have evolved as a switch sufficient advantage during famines to get selected in in reproductive and sustenance strategies rather than an spite of the fact that obesity is associated with reduced adaptation to feast and famine. (ii) Extending this logic fecundity [20]. further, Rashidi et al. [25] explained why pancreatic beta (5) Obesity and insulin resistance is associated with a cells and those of females in particular are more susceptible number of changes in the different body systems and to oxidative damage. Under stress conditions, the release their functions as diverse as ovulation, spermato- of stress hormones produces insulin resistance and, owing genesis, innate immunity, wound healing, memory, to reactive oxygen species (ROS) preventing beta cells and cognitive brain functions [10]. The thriftiness from secreting insulin at the level required to maintain hypotheses focus on energy homeostasis alone and homeostasis, diverts glucose to insulin-independent tissues, offer no explanation as to why these diverse changes such as the brain and the fetus. They suggest that pancreatic are associated with it. beta cells lost part of their antioxidant defense in association with brain evolution, and lost even more in females when Realizing the limitations, inadequacies, and flaws of the placental mammals evolved. thrifty gene and fetal programming hypotheses, a number of The emergence of alternative hypotheses having different alternatives have been suggested. implications for the genetics of obesity has made it even (a) As alternatives to the thrifty gene hypothesis, (i) more critical that the traditional hypothesis is examined Speakman [8] postulated genetic drift rather than selection analytically. The concept of thriftiness has been mostly for obesity-related genes. An upper limit on obesity was discussed descriptively and qualitatively and almost never Journal of Obesity 3 subject to quantitative methods commonly used by evolu- is calculated as the sum of all years, with the assumption tionary geneticists to test or support any argument. We use a that in the birth year, the phenotype is best suited for simple mathematical model here to examine the conditions given conditions. For the rest of the lifetime (S), the fitness under which thrifty gene or fetal programming is likely fluctuates according to randomly fluctuating environmental to gain a selective advantage and evaluate how likely it conditions. Therefore, is for human ancestors to have evolved constitutive or S − programmable thriftiness. There are no empirical estimates = 1 × × − 1 Ltp S pf tf0 + 1 pf nf1 + S of the actual fitness contributions of the hypothetical genes (3) or phenotypes, and, therefore, a quantitatively predictive × pf × Ltg + 1 − pf × Ln . model cannot be attempted at this stage. The limited objective of our model is to conceptually examine whether a The first term in (3) denotes fitness contribution of birth thrifty gene or thrifty programming can evolve in principle, year, and the second term sums the fitness contribution of whether stable polymorphism in this character is possible, the rest of the lifespan. if yes, what are the necessary conditions for its evolutionary stability, and whether the thriftiness hypotheses adequately Solution: Considering the nonthrifty and thrifty genotypes explain the current obesity epidemic. alone, it can be seen that, at large pf, thrifty gene has an advantage over nonthrifty gene, and, at small pf, nonthrifty 2. The Model gene cannot be invaded by the thrifty one. The transition is at Ln = Ltg and this happens when To model the possible evolution of thriftiness, we consider pf tf1 − nf1 3 hypothetical genotypes, namely, a nonthrifty wild type = . (4) having no mechanism for thriftiness (n), a thrifty genotype 1 − pf nf0 − tf0 (tg) which is genetically programmed for thriftiness and a genotype with a capacity for fetal programming for It is important to note that the areas of advantage are decided thriftiness (tp). Taking a year as a natural time unit of not by the absolute fitness values but only by the ratios of the ff seasonality, we assume a simple dichotomy of years with di erences in fitness in feast and famine conditions. adequate food supply (feast) and those with inadequate food Equation (4) implies that the evolution of thrifty gene supply (famine). Famines are assumed to occur randomly needs a high frequency of famine. Famines with significant with a probability pf. mortality occur with a frequency of once in 100–150 years Fitness of an individual with nonthrifty genotype in [8]. If surviving famines is the major selective force for thrifty = feast conditions (nf1) is assumed to be greater than that of gene, with Pf 0.01, the advantage of thrifty over nonthrifty individual with thrifty genotype (tf1) because there is cost phenotype in famine conditions should be more than 100- ff associated with thriftiness. When there is a feast, nonthrifty times the relative loss su ered by thrifty gene in feast individuals will do better as they do not pay the cost for conditions. If the obesity-induced reduction in fecundity is being thrifty (nf1 > tf1). The cost of thriftiness is justifiable sizable, the advantage of thriftiness in famines should be based on the reproductive effects of obesity. We assume exceedingly large for thrifty genes to evolve. Such a large that thrifty individuals are fast to become obese in feast advantage should be highly evident and easily measurable, conditions, and there are multiple mechanisms by which but obese people have not been shown to survive famines obesity causes reduction in fecundity. On the one hand, significantly better than lean individuals [8]. obesity and insulin resistance are associated with ovulation Considering competition between nonthrifty and fetal disorders and are major risk factors for polycystic ovary programming genotypes, it can be seen that at higher pf and > = syndrome (PCOS) [26]. On the other, reduced fecundity lower S, Ltp Ln. The transition line resides at Ln Ltp. is also seen in obese women with regular cycles [20, 27]. For a given pf, this condition is met at Moreover, obesity also affects spermatogenesis in men [29], pf × tf0 − Ltg + 1 − pf × nf1 − Ln and the effects of obesity on male and female fertility become S = . × − (5) additive if a couple is obese [30]. pf Ln Ltg Fitness of an individual with nonthrifty genotype in Figure 1(a) shows the transition line demarcating the areas famine conditions (nf0) is assumed to be less than that of of selective advantage of Ln and Ltp when only these < individuals with thrifty genotype (tf0)(nf0 tf0). two genotypes compete. At very high pf,however,anall- The lifetime fitness of an individual is given as follows. time thrifty genotype might have an advantage over fetal For individuals with nonthrifty genotype, programming. Therefore, we need to examine the areas of Ln = pf × nf0 + 1 − pf × nf1. (1) advantage of Ltp and Ltg. Taking a similar approach as above, Ltg = Ltp when For individuals with thrifty genotype, pf × tf0 − Ltg + 1 − pf × nf1 − Ln Ltg = pf × tf0 + 1 − pf × tf1. (2) S = . (6) 1 − pf × Ltg − Ln For individuals with thrifty phenotype or the capacity for irreversible fetal programming, assuming no correla- Agraphofpf versus S (Figure 1(b)) shows that, at greater tion between birth and lifetime conditions, the total fitness probability of famine and greater longevity, thrifty gene(s) 4 Journal of Obesity

60 60

50 50

40 40

30 30

20 20

10 10

0 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Area of advantage of non-thrifty genotype Area of advantage of non-thrifty genotype Area of advantage of thrifty genotype Area of advantage of thrifty genotype Area of advantage of fetal programmer Area of advantage of fetal programmer (a) (b)

Figure 1: Parameter areas of advantage: (a) when only nonthrifty genotype and fetal programmer compete, (b) when only thrifty genotype and fetal programmer compete. For this and all other ﬁgures, (tf1 − nf1)/(nf0 − tf0) is maintained constant at 1.33.

60 can evolve. The two results together (Figure 2) imply that at low probabilities of famine, a nonthrifty gene has a net 50 advantage, and at high pf, thrifty gene would evolve leaving a very narrow area of advantage for fetal programming. The 40 area is narrower for long-lived species whereas for short-lived ones, the birth year advantage is large enough as compared to 30 lifetime, and, therefore, fetal programming has a much larger width of advantage. 20 The results so far are based on the assumption that seasonal or annual climatic variations are random with little 10 or no predictability. We now consider the eﬀects of such predictability. If there exists a correlation r between uterine 0 conditions and lifetime conditions, then it follows that the 0 0.2 0.4 0.6 0.8 1 lifetime probability of famine would be higher if birth was in a famine year. We can, therefore, write the expected Area of advantage of non-thrifty genotype probability of facing a famine if the birth year was a famine year as Area of advantage of thrifty genotype Area of advantage of fetal programmer pfr0 = 1 − pf × r + pf. (7)

Figure 2: Parameter areas of advantage when nonthrifty genotype, The probability of facing a famine if the birth year was a feast thrifty genotype, and fetal programmer compete simultaneously. year as Fetal programming can evolve for species with short lifespan. If the ﬀ life span is longer, fetal programming is unlikely to o er selective = − × r advantage over thrifty or nonthrifty genotypes except for a speciﬁc pfr1 pf pf ,(8) and very narrow range of pf. accordingly,

× − × = pf tf0 + 1 pf nf1 Ltp S (9) S − × × × − × − × × − × ( 1) pf pfr0 tf0 + 1 pfr0 tf1 + 1 pf pfr1 nf0 + 1 pfr1 nf1 . + S Journal of Obesity 5

Now, the new transition between areas of advantage of non thrifty and fetal programming (when Ln = Ltp) is obtained at

× − × S = pf tf0+ 1 pf nf1 Ln − pf × pfr0 × tf0 + 1 − pfr0 × tf1 + 1 − pf × pfr1 × nf0 + 1 − pfr1 × nf1 (10) × × − × − × × − × − pf pfr0 tf0 + 1 pfr0 tf1 + 1 pf pfr1 nf0 + 1 pfr1 nf1 . Ln − pf × pfr0 × tf0 + 1 − pfr0 × tf1 + 1 − pf × pfr1 × nf0 + 1 − pfr1 × nf1

Similarly, the transition between areas of advantage of thrifty and fetal programming is obtained at Ltg = Ltp. This condition is satisﬁed when

× − × S = pf tf0 + 1 pf nf1 Ltg − pf × pfr0 × tf0 + 1 − pfr0 × tf1 + 1 − pf × pfr1 × nf0 + 1 − pfr1 × nf1 (11) × × − × − × × − × − pf pfr0 tf0 + 1 pfr0 tf1 + 1 pf pfr1 nf0 + 1 pfr1 nf1 . Ltg − pf × pfr0 × tf0 + 1 − pfr0 × tf1 + 1 − pf × pfr1 × nf0 + 1 − pfr1 × nf1

Figure 3 shows that as the birth-time and lifetime correlation conditions, similarly no such correlational patterns in any increases, the area of advantage of fetal programmer widens. other climatic variables are reported, fetal programming However, in long-lived species, since the advantage to fetal is unlikely to have evolved in anticipation of drought or programmer in the absence of correlation is very small, even famine. a small negative correlation drives the fetal programmer to extinction (Figure 3(c)). It is most important for the evolu- 3. Discussion tion of fetal programming that a positive correlation between birth-time and lifetime conditions exists. The correlation The model suggests that a thrifty gene or a set of genes can need not be very high. It can be seen that at r = 0.3, fetal evolve only if the frequency of famines is very high and under programmer has an advantage over a wide range of pf. such conditions the thrifty allele(s) would not exist in a stable polymorphic state. Selection for the ability to program Climatic fluctuations from year to year are a complex the fetus is likely in a very narrow range of frequency of phenomenon, and since prediction of important climatic famines, and at lower or higher frequency of famines, fetal features, such as rainfall, has important implications, programming would not evolve. The parameter space for there have been serious attempts to detect temporal the evolution of fetal programming becomes very narrow patterns. However, temporal patterns are of little help in and highly specific for long-lived species owing to which weather prediction since there are no consistent time lapse evolution of fetal programming for thriftiness is highly correlations in rainfall or other parameters. Since India has unlikely. the largest population of diabetics and monsoon is the most The model results need to be interpreted carefully in the important determinant of food availability in this region, context of ancestral human ecology. One of the key questions it would be enlightening to see the patterns in the Indian is when in human history could selection for thriftiness, if monsoon. Based on the public-domain data on Indian any, have operated. We will examine three possible scenarios monsoon by the Indian Institute of Tropical Meteorology below. (ftp://www.tropmet.res.in/pub/data/rain/iitm-subdivrf.txt), (a) Selection during hunting-gathering stage: pale- we investigated whether birth year and subsequent year’s oarcheological data suggests that chronic starvation was rainfall has any positive correlation in the short or long run uncommon during hunter-gatherer stage [31]. Today’s for any of the 30 monsoon subdivisions in India. Table 1 hunter-gatherer societies do not seem to suffer starvation shows that rainfall in a given year is not correlated with that more frequently or more intensively than agricultural soci- of the subsequent year, subsequent 10 years, or 40 years eties [9]. Therefore, the assumption that hunter-gatherer cumulative. Over the 30 monsoon subdivisions, only 6 are societies suffered frequent starvation is not well supported, statistically significant using individual α level of 0.05, out but even if we assume hunter-gatherer societies to be prone of which 4 are negative, contrary to the expectation. Using to feast and famine selection, a number of other questions Bonferroni correction for significance level, applicable when remain unanswered. Since hominids were hunter-gatherers a large number of tests are being done together, none of the for the most part of human evolutionary history, selection correlations remain significant. As there is no detectable would have been prolonged, and we would expect alleles positive correlation between birth year and lifetime rainfall to have reached equilibrium frequencies. The model implies 6 Journal of Obesity

60 60

50 50

40 40

30 30

20 20

10 10

0 0 0 0.3 0.5 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1

0 0 0.2 0.4 0.6 0.8 1

Area of advantage of non-thrifty genotype Area of advantage of thrifty genotype Area of advantage of fetal programmer (c)

Figure 3: Parameter areas of advantage of the 3 genotypes when there is a correlation between birth-time and lifetime conditions: (a) r = 0.1, (b) r = 0.2, and (c) r =−0.05. With a small positive correlation, the advantage of fetal programmer increases substantially. However, even very weak negative correlation can drive fetal programmer to extinction when life expectancy is high. Selection for fetal programming, therefore, must be driven by factors that produce significant positive birth-time and lifetime correlations. that selection cannot result in stable polymorphism of thrifty the population at any given time will consist of thrifty, alleles. In the modern human society, there is considerable nonthrifty, and all time fit individuals. However, there is no variation in the tendency to become obese or diabetic. empirical evidence of heterozygote advantage so far. Stable Therefore, polymorphism with respect to genes predisposing polymorphism is also possible if there is negative frequency- to obesity and type 2 diabetes presumably exists. If there is no dependent selection. However, if fitness is decided by climatic negative frequency dependence or heterozygote advantage, conditions as assumed by the popular version of thrifty gene natural selection will be directional, resulting into the hypothesis, frequency dependence is unlikely. Therefore, fixation of the advantageous genotype. In that case at no selection during hunter-gatherer stage does not explain the value of pf, the thrifty and nonthrifty genes can coexist stably. prevalent polymorphism in predisposition to obesity. Theoretically, if a heterozygote of thrifty and nonthrifty (b) Selection after the beginning of agriculture: chronic alleles gets a dual advantage by expressing the right allele starvation due to famines became more serious and common in the right environment, the two alleles can coexist, and with the beginning of agriculture [7, 8]. Signs of chronic Journal of Obesity 7

Table 1: Correlations of annual rainfall to that of the subsequent year, short-term (10 years), and long-term (40 years) cumulative: ∗ indicate significance of individual correlations at α = 0.05 level. Since a large number of statistical tests are being performed, a Bonferroni correction to the significance level is applicable. After the Bonferroni correction, none of the correlations are significant.

S. No. Subdivision 1 year 10 years cumulative 40 years cumulative 1 Assam Meghalaya 0.047 0.042 −0.029 Nagaland Manipur 2 0.006 0.010 −0.078 Mizoram and Tripura Sub-Himalayan West 3 −0.061 −0.103 −0.023 Bengal 4 Gangetic West Bengal −0.009 0.096 −0.265∗ 5 Orissa −0.111 0.135 −0.096 6 Jharkhand −0.054 −0.042 −0.012 7 Bihar 0.070 −0.158 −0.048 8 East Uttar Pradesh 0.096 −0.106 −0.091 9 West Uttar Pradesh Plains −0.036 0.010 −0.066 10 Haryana −0.010 0.050 0.041 11 Punjab −0.050 0.099 0.081 12 Rajasthan 0.048 −0.200∗ −0.094 13 East Rajasthan 0.047 −0.034 −0.017 14 West Madhya Pradesh 0.052 0.114 0.010 15 East Madhya Pradesh −0.076 0.040 −0.034 16 Gujarat −0.007 −0.121 −0.110 17 Saurashtra and Kutch −0.037 −0.118 −0.152 18 Konkan and Goa 0.121 0.154 0.080 19 Madhya Maharashtra 0.227∗ 0.010 −0.163 20 Marathwada 0.171 0.043 −0.101 21 Vidarbha −0.079 0.026 −0.193∗ 22 Chhattisgarh 0.039 0.278∗ 0.010 23 Coastal Andhra Pradesh 0.082 0.023 −0.053 24 Telangana 0.141 0.110 0.086 25 Rayalaseema −0.104 −0.008 −0.054 26 Tamil Nadu 0.049 −0.139 −0.169 27 Coastal Karnataka −0.035 −0.042 0.079 28 North Interior Karnataka 0.132 0.086 −0.020 29 South Interior Karnataka −0.025 −0.087 −0.202∗ 30 Kerala 0.117 0.083 0.012

starvation on teeth, such as linear enamel hypoplasia, are community developed a high prevalence of diabetes and more common in early agricultural societies than in hunter- hypertension [33]. It is difficult to argue, therefore, that gatherer societies [32]. This is owing to the fact that crops are thrifty genes evolved after the advent of agriculture. highly seasonal in nature, and the failure of a single crop leads (c) Selection in modern times: intensive agricultural and to long-term food scarcity. Such long-term food shortages industrial societies are modern phenomena not more than are much less probable in hunter-gatherer life, particularly in a few hundred years old, and it is highly unlikely that this biodiversity rich areas. Therefore, if selection for thriftiness period could have brought about an evolutionary change. It started acting after the beginning of agriculture, there could can be seen from all the three possible scenarios that natural be transient polymorphism. A testable prediction of the selection for the hypothetical thrifty gene(s) is unable to hypothesis would then be that ethnic groups that took to explain the polymorphism observed today. agriculture earlier should show higher tendencies to become Since different geographic regions of the world differ obese and diabetic. This has not been rigorously tested in the climatic conditions, seasonality, and food availabil- with quantitative data. However, ethnic groups, such as ity, ethnic groups evolved in different areas should show the Australian aborigines, remained hunter gatherers until differential predisposition to obesity and related disorders. recently, and the recently urbanized individuals of this People evolved in arid or drought-prone areas could have 8 Journal of Obesity suffered more frequent famines and, therefore, should have is negative frequency-dependent selection in a Hawk and a greater tendency to develop obesity. Also, ethnic groups Dove like game [42]. An alternative view is that the human from harsh winter environments were unable to hunt, fish, or population today is not at a stable equilibrium proportion farm during the colder months and, thus, historically could of alleles but is undergoing drift [8] or selection [21]. have faced regular feast and famine conditions. Therefore, These alternative hypotheses try to explain polymorphism we may expect higher diabetic tendencies in them. Although as a transient state. Speakman [7] also tries to quantify substantial cross-ethnic differences are observed, the trends the drift dynamics; however, his calculation is based on are not as expected by the thriftiness hypotheses. People the assumption that human ancestors became free from from Caucasoid, Eskimo, and some of the Himalayan ethnic predators about 1.8 million years ago, and this estimate is groups that have faced harsh winter environments have a debatable [41]. Moreover, if the drift is an ongoing process, it would take a different direction in different populations considerably lower frequency of obesity and/or T2D [34– ff 38] whereas almost all ethnic groups of warmer habitats and the cross-ethnic di erences are expected to be random. Instead, we see that almost all tropical ethnic groups have have a high frequency on adopting a Western urban lifestyle high tendency to develop obesity and T2D on urbanization, [33, 39], contrary to the expectation. and such a generalizable pattern is not expected by the The thrifty phenotype or fetal programming hypothesis drift hypothesis. Also, as we know now, a large number of ff ff su ers from a di erent set of problems. Fetal programming genes are associated with obesity, and if drift has to operate ff can o er two types of potential advantages. Short-term independently on all the genes, it is highly unlikely that it will survival advantages in fetal and early infant life and long- result in a directional change. term predictive adaptive responses. Our model incorporates Moalem et al. [22] presumed that selective pressures both of the advantages separately. If the advantage is of changed substantially by 1800 AD. It has not been critically a short duration, it is difficult to explain why a lifetime questioned whether only about 200 years of selection is commitment to a particular metabolic state could have sufficient to generate the current levels of polymorphism. evolved. A number of developmental genes have age-specific Any data to test the dynamics of both the hypotheses using a expressions, and, therefore, any rigid lifelong programming predictive model are currently absent. for short-term advantage is a difficult proposition. If climatic The third possible explanation to the apparent poly- fluctuations were the main selective force, it should evolve morphism is that there is no real genetic polymorphism, metabolic flexibility rather than lifelong rigid programming. but individuals are programmed differently depending upon Metabolic programming of a lifelong duration based on environmental conditions faced in early life. Apart from the intrauterine conditions is unlikely to offer a fitness advantage thrifty phenotype hypothesis, only the Watve and Yajnik except in two sets of conditions. As the model suggests, if [10] hypothesis accounts for lifetime programming in a way a species has a very short life span, fetal programming for discussed below. adapting to the birth year conditions can be beneficial, since The cold adaptation hypothesis of Hattersley and Tooke the birth year itself is a substantial part of the total life span. [23] does not explain polymorphism at any of the above levels. If high blood glucose is adaptive in cold environ- Assuming one year to be the natural unit of seasonal cycles, ments, then ethnic groups who evolved in cold climates species with a life span of less than 3–5 years can be expected should undergo directional selection leading to fixation. to evolve lifelong fetal programming for thriftiness even Polymorphism in that case can only arise by cross-breeding though the adaptive advantage is of a short duration. For between ethnic groups coming from warmer and colder long-lived species, fetal programming is unlikely to evolve environments. unless there is a significant positive correlation between birth (2) Low birth weight effects: there is globally consistent conditions and lifelong conditions. Since such correlations and strong evidence that a small birth weight is associated are not seen in climatological data, there are problems in with altered metabolic and endocrine states in adult life [3, explaining evolution of lifelong thrifty programming. 41, 42]. Different interpretations of the birth weight effects Since the thriftiness hypotheses are inadequate or weak [23]havebeenoffered, and the birth weight association on several grounds, there is a need to rethink the paradigm itself cannot be considered a convincing evidence of fetal and consider alternative hypotheses seriously. A detailed programming. However, rat experiments with maternal comparative analysis of all the alternative hypotheses is nutrient deficiencies have also given strong evidence in beyond the scope of this paper. Most of them are new, and support of fetal programming [43]. Therefore, some fetal all their implications have not yet come forward. We will programming can be safely inferred. It can nevertheless only briefly evaluate the alternative hypotheses below to see be debated whether the programming is for thriftiness or whether they offer better explanations where the thriftiness for any other adaptation. Speakman’s [7, 8]driftygene hypotheses are weaker. hypothesis, Corbett et al. [21] hypothesis of reverse selection (1) Polymorphism: polymorphism in the predisposition in modern times, and Moalem et al.’s [22] cold adaptation to obesity can be potentially explained in three different ways. hypotheses are unable to explain the birth weight effect. Wells It is possible that the allelic composition in the population [44] and Watve and Yajnik [10]havedifferent interpretations is close to equilibrium, and there is stable polymorphism. of the effect of maternal nutrient limitation, but both of them Out of all alternative hypotheses, only the Watve-Yajnik adequately account for birth weight effects. Wells’ hypothesis hypothesis is able to predict stable polymorphism since there [44] of maternal advantage does not have a predictive Journal of Obesity 9 adaptation element and, therefore, does not explain lifetime fore, that fetal programming could have evolved in rigidity in the programming. Watve and Yajnik [10] argue anticipation of population- and social hierarchy-related that fetal undernourishment affects muscle mass more than factors than anticipation of famines or other climate-related brain mass and, therefore, a brain-dependent “diplomat” factors. lifestyle marked by insulin resistance is adaptive for low (4) Multiple effects: obesity and T2D are not only about birth weight individuals. Since muscle cells do not replicate energy homeostasis but also about changes in innate immu- in adult life, this early developmental limitation persists nity, sexual and reproductive function, vascular development throughout life, and, thus, lifetime programming could and function, skin architecture, wound healing and tissue evolve. At the same time, low birth weight is not a necessary regeneration, memory, cognitive functions, behavior and prerequisite for the Watve and Yajnik hypothesis to work. mechanisms of decision making, social relations, and social Any hypothesis exclusively based on fetal programming signaling. Therefore, any hypothesis about the origins of suffers from another problem in that although low birth obesity needs to account for all the network of changes weight individuals have a greater probability of developing adequately. Most hypotheses are too glucolipo-centric and, obesity and T2D, a large proportion of adult type 2 diabetics therefore, fall short of this criterion. The only possible are not born with low weights. Therefore, any hypothesis exception is the Watve-Yajnik hypothesis which has an exclusively dependent upon fetal conditions are inherently inherent expectation that a number of systems of the body inadequate. would be simultaneously involved [10, 40]. (3) Birth time versus lifetime correlation: the model Implications for genetics of obesity: a large body of predicts that for lifetime fetal programming to evolve, research has now focused on the genetics of obesity. An a positive correlation between intrauterine and lifetime increasing number of loci and mutants associated with conditions is necessary. Although climatic variations are obesity are being identified. However, there are certain unlikely to cause such correlations, there can be other causes internal paradoxes associated with these data. Studies prior of food deprivation apart from climatic factors. A high to the genomic era that were based on familial, twin-pair, population density can lead to increased competition, result- and adoption studies typically predicted a large heritable ing into undernourishment. Although population density component in obesity (reviewed by [49]). The genome-wide may oscillate, the oscillations typically span over several association studies, on the other hand, have identified a large generations so that an individual born at a high population number of associations, but together they explain a very density is expected to see high population density through small fraction of variance in obesity parameters [50–54], most of his life. Therefore, periodic food scarcity caused by leaving a large gap between the pregenomic and emerging population oscillations can result into positive correlations genomic picture. We are yet to understand the reasons for between birth year and lifetime nutritional conditions. In this discrepancy, but a most likely implication goes against all social species, the social hierarchy might also be related hypotheses that assume a gene or a set of genes for obesity. to differential access to food. An individual born smaller These hypotheses include Neel’s thrifty gene, Speakman’s and weaker is more likely to face social subordination, drifty gene, Corbett et al.’s reverse selection, and Moalem et and, therefore, an anticipatory fetal programming would be al.’s cold adaptation. On the other hand, fetal programming adaptive. These social causes can produce positive birth- has a promise for filling the gap between the familial studies lifetime correlations, and they are more likely candidates to and genome-wide association studies. The programming is select for fetal programming of lifelong duration. There exist likely to involve epigenetic mechanisms as well. Three of the a broad range of metabolic, endocrinological, behavioral, above hypotheses involve fetal programming, namely, the and cognitive adaptations that accompany social hierarchical classical thrifty phenotype hypothesis, maternal adaptation positions. Therefore, if fetal programming is in response to hypothesis by Wells [44], and behavioural switch hypothesis social hierarchies, we expect that it need not be restricted to by Watve and Yajnik [10]. Since our model has indicated diet and energy homeostasis, but it affects a large number the inadequacies of the thrifty phenotype hypothesis, the of systems of the body along with brain and behavior. This other two need to be considered more seriously. We suggest, indeed, is the case, and type 2 diabetes involves almost all here, that more than one type of metabolic programming systems of the body [10, 45]. Social subordination is an may be involved in obesity, and they may be induced important predisposing factor for type 2 diabetes [46], and in various stages of life. There are strong data for fetal the link between the two is elaborately explained by Belsare programming, but behavioral programming is also likely et al. [40]. to affect metabolism, since associations between neuroen- We suggest that on exposure to intrauterine malnourish- docrine mechanisms of behavior and metabolic states are ment, there is fetal programming in anticipation of physical known [40]. weakness and social subordination or anticipation of high The classical thriftiness family of hypotheses would population density rather than anticipation of famine. In expect genes associated with metabolic rates to be the obesity primate societies, it is known that social ranks of juveniles genes. The behavioural origins hypothesis, on the other and subadults are influenced by the ranks of their mothers hand, expects genes involved in sexual function, cognitive [47, 48]. If social rank influences preferential access to food abilities, immunity, regulation of aggression, and other resources, it is sufficient to produce the positive correlation behavioral traits to be associated with obesity and related between fetal and lifetime nutritional conditions needed for disorders. It would be useful to interpret the emerging the evolution of fetal programming. It appears logical, there- data on genome-wide associations and, perhaps, near future 10 Journal of Obesity studies on epigenetics of obesity in the light of the more [10] M. G. Watve and C. S. Yajnik, “Evolutionary origins of insulin promising hypotheses from the above. resistance: a behavioral switch hypothesis,” BMC Evolutionary We have shown with a simple theoretical model that the Biology, vol. 7, article 61, 2007. classical concepts of thrifty gene as well as fetal programming [11] J. C. K. Wells, “Flaws in the theory of predictive adaptive for thriftiness face a number of problems and are inadequate responses,” Trends in Endocrinology and Metabolism, vol. 18, to account for the observable trends in the modern epidemic no. 9, pp. 331–337, 2007. of obesity and related disorders. A number of alternative [12] N. Kloting¨ and M. Bluher,¨ “Extended longevity and insulin hypotheses have been suggested recently. It might be too signaling in adipose tissue,” Experimental Gerontology, vol. 40, no. 11, pp. 878–883, 2005. early to discard any of the hypotheses right away, but a [13] J. Eriksson, T. Forsen,´ J. Tuomilehto, C. Osmond, and D. comparative analysis shows the strengths and weaknesses Barker, “Size at birth, fat-free mass and resting metabolic rate of each of them. Since search for obesity-associated genes in adult life,” Hormone and Metabolic Research, vol. 34, no. 2, has so far given limited success in terms of explaining pp. 72–76, 2002. population variance in obesity parameters, the hypotheses [14] O. A. Kensara, S. A. Wooton, D. I. W. Phillips et al., involving phenotypic programming look more promising. “Substrate-energy metabolism and metabolic risk factors for In understanding obesity and related disorders, the need cardiovascular disease in relation to fetal growth and adult for a combined genomic-phenomic approach is increasingly body composition,” American Journal of Physiology, vol. 291, being felt [55, 56], and such an integration can be best no. 2, pp. E365–E371, 2006. achieved on the platform of an evolutionary insight into [15] K. R. Westerterp and J. R. Speakman, “Physical activity energy the phenomenon. Greater efforts are, therefore, needed to expenditure has not declined since the 1980s and matches understand the evolutionary background of obesity and energy expenditures of wild mammals,” International Journal related pathophysiological conditions. of Obesity, vol. 32, no. 8, pp. 1256–1263, 2008. [16] A. E. Black, W. A. Coward, T. J. Cole, and A. M. Prentice, “Human energy expenditure in affluent societies: an analysis Acknowledgments of 574 doubly-labelled water measurements,” European Jour- nal of Clinical Nutrition, vol. 50, no. 2, pp. 72–92, 1996. P. Belsare was supported by CSIR during the study. Discus- [17] F. Zurlo, S. Lillioja, A. E. D. Puente et al., “Low ratio of fat to sions with Sulochana Gadgil (IISc) and Ashwini Kulkarni carbohydrate oxidation as predictor of weight gain: study of (IITM) were useful. 24-h RQ,” American Journal of Physiology, vol. 259, no. 5, pp. E650–E657, 1990. [18] D. J. Hoffman, A. L. Sawaya, I. Verreschi, K. L. Tucker, and S. B. References Roberts, “Why are nutritionally stunted children at increased risk of obesity? Studies of metabolic rate and fat oxidation [1] J. V. Neel, “Diabetes mellitus: a “thrifty” genotype rendered in shantytown children from Sao Paulo, Brazil,” American detrimental by “progress”?” American Journal of Human Journal of Clinical Nutrition, vol. 72, no. 3, pp. 702–707, 2000. Genetics, vol. 14, pp. 353–362, 1962. [19] I. S. Farooqi and S. O’Rahilly, “Genetic factors in human [2]C.N.HalesandD.J.P.Barker,“Type2(non-insulin- obesity,” Obesity Reviews, vol. 8, no. 1, pp. 37–40, 2007. dependent) diabetes mellitus: the thrifty phenotype hypoth- [20]R.J.NormanandA.M.Clark,“Obesityandreproductive esis,” Diabetologia, vol. 35, no. 7, pp. 595–601, 1992. disorders: a review,” Reproduction, Fertility and Development, [3] D. J. P. Barker, “In utero programming of chronic disease,” vol. 10, no. 1, pp. 55–63, 1998. Clinical Science, vol. 95, no. 2, pp. 115–128, 1998. [21] S. J. Corbett, A. J. McMichael, and A. M. Prentice, “Type 2 [4] A. J. Drake and B. R. Walker, “The intergenerational effects diabetes, cardiovascular disease, and the evolutionary paradox of fetal programming: non-genomic mechanisms for the of the polycystic ovary syndrome: a fertility first hypothesis,” inheritance of low birth weight and cardiovascular risk,” American Journal of Human Biology, vol. 21, no. 5, pp. 587– Journal of Endocrinology, vol. 180, no. 1, pp. 1–16, 2004. 598, 2009. [5] P. D. Gluckman, M. A. Hanson, and H. G. Spencer, “Predictive [22] S. Moalem, K. B. Storey, M. E. Percy, M. C. Peros, and D. P. adaptive responses and human evolution,” Trends in Ecology Perl, “The sweet thing about type 1 diabetes: a cryoprotective and Evolution, vol. 20, no. 10, pp. 527–533, 2005. evolutionary adaptation,” Medical Hypotheses,vol.65,no.1, [6] J. R. Speakman, “Thrifty genes for obesity and the metabolic pp. 8–16, 2005. ff syndrome—time to call o the search?” Diabetes and Vascular [23] A. T. Hattersley and J. E. Tooke, “The fetal insulin hypothesis: Disease Research, vol. 3, no. 1, pp. 7–11, 2006. an alternative explanation of the association of low birth- [7] J. R. Speakman, “A non-adaptive scenario explaining the weight with diabetes and vascular disease,” The Lancet, vol. genetic predisposition to obesity: the “predation release” 353, no. 9166, pp. 1789–1792, 1999. hypothesis,” Cell Metabolism, vol. 6, no. 1, pp. 5–12, 2007. [24] J. C. K. Wells, “Thrift: a guide to thrifty genes, thrifty pheno- [8] J. R. Speakman, “Thrifty genes for obesity, an attractive but types and thrifty norms,” International Journal of Obesity, vol. flawed idea, and an alternative perspective: the ‘drifty gene’ 33, no. 12, pp. 1331–1338, 2009. hypothesis,” International Journal of Obesity, vol. 32, no. 11, [25] A. Rashidi, T. B. L. Kirkwood, and D. P. Shanley, “Metabolic pp. 1611–1617, 2008. evolution suggests an explanation for the weakness of antiox- [9] D. C. Benyshek and J. T. Watson, “Exploring the thrifty idant defences in beta-cells,” Mechanisms of Ageing and genotype’s food-shortage assumptions: a cross-cultural com- Development, vol. 130, no. 4, pp. 216–221, 2009. parison of ethnographic accounts of food security among [26] R. Pasquali, A. Gambineri, and U. Pagotto, “The impact of foraging and agricultural societies,” American Journal of obesity on reproduction in women with polycystic ovary Physical Anthropology, vol. 131, no. 1, pp. 120–126, 2006. syndrome,” BJOG, vol. 113, no. 10, pp. 1148–1159, 2006. Journal of Obesity 11

[27] D. C. Gesink Law, R. F. Maclehose, and M. P. Longnecker, organs: “Ominous Orchestra of Organs,”,” Clinical Medicine: “Obesity and time to pregnancy,” Human Reproduction, vol. Endocrinology and Diabetes, vol. 1, pp. 1–6, 2008. 22, no. 2, pp. 414–420, 2007. [46] E. E. Agardh, A. Ahlbom, T. Andersson et al., “Work stress and [28] N. Yilmaz, S. Kilic, M. Kanat-Pektas, C. Gulerman, and L. low sense of coherence is associated with type 2 diabetes in Mollamahmutoglu, “The relationship between obesity and middle-aged Swedish women,” Diabetes Care, vol. 26, no. 3, fecundity,” Journal of Women’s Health, vol. 18, no. 5, pp. 633– pp. 719–724, 2003. 636, 2009. [47] J. A. Johnson, “Dominance rank in juvenile olive baboons, [29] M. Sallmen,´ D. P. Sandler, J. A. Hoppin, A. Blair, and D. D. Papio anubis: the influence of gender, size, maternal rank and Baird, “Reduced fertility among overweight and obese men,” orphaning,” Animal Behaviour, vol. 35, no. 6, pp. 1694–1708, Epidemiology, vol. 17, no. 5, pp. 520–523, 2006. 1987. [30] C. H. Ramlau-Hansen, A. M. Thulstrup, E. A. Nohr, J. P. [48] K. E. Holekamp and L. Smale, “Dominance acquisition during Bonde, T. I. A. Sørensen, and J. Olsen, “Subfecundity in mammalian social development: the “inheritance” of maternal overweight and obese couples,” Human Reproduction, vol. 22, rank,” American Zoologist, vol. 31, no. 2, pp. 306–317, 1991. no. 6, pp. 1634–1637, 2007. [49] H. H. M. Maes, M. C. Neale, and L. J. Eaves, “Genetic and [31] M. Sahlins, Stone Age Economics, Tavistock Publications, environmental factors in relative body weight and human London, UK, 1974. adiposity,” Behavior Genetics, vol. 27, no. 4, pp. 325–351, 1997. [32] J. R. Lukacs and S. R. Walimbe, “Physiological stress in [50] T. Rankinen, A. Zuberi, Y. C. Chagnon et al., “The human prehistoric India: new data on localized hypoplasia of primary obesity gene map: the 2005 update,” Obesity,vol.14,no.4, canines linked to climate and subsistence change,” Journal of pp. 529–644, 2006. Archaeological Science, vol. 25, no. 6, pp. 571–585, 1998. [51] S. Li, J. H. Zhao, J. Luan et al., “Cumulative effects and [33] K. O’Dea, “Westernisation, insulin resistance and diabetes in predictive value of common obesity-susceptibility variants Australian Aborigines,” Medical Journal of Australia, vol. 155, identified by genome-wide association studies,” American no. 4, pp. 258–264, 1991. Journal of Clinical Nutrition, vol. 91, no. 1, pp. 184–190, 2010. [34] B. A. Swinburn, “The thrifty genotype hypothesis: how does [52] S. Li, J. H. Zhao, J. Luan et al., “Physical activity attenuates it look after 30 years?” Diabetic Medicine,vol.13,no.8,pp. the genetic predisposition to obesity in 20,000 men and 695–699, 1996. women from EPIC-Norfolk prospective population study,” [35] H. Beck-Nielsen, “General characteristics of the insulin resis- PLoS Medicine, vol. 7, no. 8, Article ID e1000332, 2010. tance syndrome prevalence and heritability,” Drugs, vol. 58, [53] G. Thorleifsson, G. B. Walters, D. F. Gudbjartsson et al., supplement 1, pp. 7–10, 1999. “Genome-wide association yields new sequence variants at [36] S. Vilbergsson, G. Sigurdsson, H. Sigvaldason, A. B. Hreidars- seven loci that associate with measures of obesity,” Nature son, and N. Sigfusson, “Prevalence and incidence of NIDDM Genetics, vol. 41, no. 1, pp. 18–24, 2009. in Iceland: evidence for stable incidence among males and [54] D. M. Mutch and K. Clement,´ “Unraveling the genetics of females 1967–1991—the Reykjavik Study,” Diabetic Medicine, human obesity,” PLoS Genetics, vol. 2, no. 12, pp. 1956–1963, vol. 14, no. 6, pp. 491–498, 1997. 2006. [37] G. J. Mouratoff, N. V. Carroll, and E. M. Scott, “Diabetes mel- [55] R. A. Hegele and R. L. Pollex, “Hypertriglyceridemia: phe- litus in Eskimos,” Journal of the American Medical Association, nomics and genomics,” Molecular and Cellular Biochemistry, vol. 199, no. 13, pp. 107–112, 1967. vol. 326, no. 1-2, pp. 35–43, 2009. [38] M. Vaz, T. T. Ukyab, R. Padmavathi et al., “Body fat topogra- [56] D. Houle, “Numbering the hairs on our heads: the shared phy in Indian and Tibetan males of low and normal body mass challenge and promise of phenomics,” Proceedings of the index,” Indian Journal of Physiology and Pharmacology, vol. 43, National Academy of Sciences of the United States of America, no. 2, pp. 179–185, 1999. vol. 107, no. 1, pp. 1793–1799, 2010. [39] J. S. Yudkin, “Non-insulin-dependent diabetes mellitus (NIDDM) in Asians in the UK,” Diabetic Medicine, vol. 13, no. 6, pp. S16–S18, 1996. [40] P. V. Belsare, M. G. Watve, S. S. Ghaskadbi, D. S. Bhat, C. S. Yajnik, and M. Jog, “Metabolic syndrome: aggression control mechanisms gone out of control,” Medical Hypotheses, vol. 74, no. 3, pp. 578–589, 2010. [41] C. H. D. Fall, C. E. Stein, K. Kumaran et al., “Size at birth, maternal weight, and type 2 diabetes in South India,” Diabetic Medicine, vol. 15, no. 3, pp. 220–227, 1998. [42] C. Li, M. S. Johnson, and M. I. Goran, “Effects of low birth weight on insulin resistance syndrome in Caucasian and African-American children,” Diabetes Care, vol. 24, no. 12, pp. 2035–2042, 2001. [43] J. F. Martin, C. S. Johnston, C. T. Han, and D. C. Benyshek, “Nutritional origins of insulin resistance: a rat model for diabetes-prone human populations,” Journal of Nutrition, vol. 130, no. 4, pp. 741–744, 2000. [44] J. C.K. Wells, “The thrifty phenotype as an adaptive maternal effect,” Biological Reviews, vol. 82, no. 1, pp. 143–172, 2007. [45] H. Koshiyama, Y. Ogawa, K. Tanaka, and I. Tanaka, “Dia- betesmellitusasdysfunction of interactions among all Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 265417, 13 pages doi:10.1155/2011/265417

Research Article Differential Effects of Calorie Restriction and Exercise on the Adipose Transcriptome in Diet-Induced Obese Mice

Karrie E. Wheatley,1, 2 Leticia M. Nogueira,2, 3, 4 Susan N. Perkins,1 and Stephen D. Hursting1, 2, 3

1 Department of Nutritional Sciences, University of Texas, Austin, TX 78712, USA 2 Department of Molecular Carcinogenesis, UT-MD Anderson Cancer Center, Smithville, TX 78957, USA 3 Institute for Cellular and Molecular Biology, University of Texas, Austin, TX 78712, USA 4 Cancer Prevention Fellowship Program, National Cancer Institute, Bethesda, MD 20852, USA

Correspondence should be addressed to Stephen D. Hursting, [email protected]

Received 9 December 2010; Accepted 1 March 2011

Academic Editor: P. Trayhurn

Copyright © 2011 Karrie E. Wheatley et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

We tested the hypothesis that obesity reversal by calorie restriction (CR) versus treadmill exercise (EX) differentially modulates adipose gene expression using 48 female C57BL/6 mice administered a diet-induced obesity (DIO) regimen for 8 weeks, then randomized to receive for 8 weeks either: (1) a control (AIN-76A) diet, fed ad libitum (DIO control); (2) a 30% CR regimen; (3) a treadmill EX regimen (with AIN-76A diet fed ad libitum); or (4) continuation of the DIO diet. Relative to the DIO controls, both CR and EX reduced adiposity by 35–40% and serum leptin levels by 80%, but only CR increased adiponectin and insulin sensitivity. Gene expression microarray analysis of visceral white adipose tissue revealed 209 genes responsive to both CR and EX, relative to the DIO group. However, CR uniquely altered expression of an additional 496 genes, whereas only 20 were uniquely affected by EX. Of the genes distinctly responsive to CR, 17 related to carbohydrate metabolism and glucose transport, including glucose transporter (GLUT) 4. Chromatin immunoprecipitation assays of the Glut4 promoter revealed that, relative to the DIO controls, CR significantly increased histone 4 acetylation, suggesting epigenetic regulation may underlie some of the differential effects of CR versus EX on the adipose transcriptome.

1. Introduction hyperinsulinemia [6]. In the insulin-resistant state, adipose tissue secretes adipokines and proinflammatory factors that More than two thirds of all adults in the USA are either reduce insulin sensitivity in peripheral tissues, thereby affect- overweight or obese [1]. Obesity is associated with an ing whole-body glucose homeostasis [7, 8]. Unfortunately, increased risk of developing several chronic diseases, includ- the mechanisms underlying these changes in insulin respon- ing atherosclerosis, type 2 diabetes and many types of cancer siveness in adipocytes are poorly understood. Furthermore, [2–4]. At the crux of obesity-related diseases is metabolic mechanism-based lifestyle strategies for effectively offsetting dysregulation characterized by insulin resistance and ele- obesity-induced insulin resistance are lacking. vated levels of circulating insulin, glucose, and several other Increased energy expenditure and decreased energy metabolic factors directly linked to excess adiposity. In the intake are the two most commonly recommended lifestyle context of low adiposity, insulin activates signaling through changes to reduce adiposity and restore insulin sensitivity the insulin receptor, resulting in translocation of the glucose [9]. Calorie restriction (CR) and exercise (EX) are both transporter 4 (Glut4) to the cell membrane to increase effective at improving insulin sensitivity and decreasing both glucose uptake into the adipocyte [5]. In contrast, high body weight and percent body fat [10, 11], although the levels of adiposity are marked by enlarged adipocytes which differential effects of these two antiobesity interventions on are unresponsive to insulin levels even under conditions of weight reduction, body composition, and chronic disease 2 Journal of Obesity risk are well established [9, 11]. There are also conflicting used as a feeding control for determining CR feed intake and reports within the human literature concerning the effective- to ensure that EX mice were not overeating to compensate ness of EX in improving body fat distribution and adipokine for increased energy expenditure. We have previously shown secretion, two key predictors of insulin resistance [10, 12– that switching DIO mice to the control (AIN-76A) diet 14]. Furthermore, studies on the beneficial effects of EX have maintains adiposity near the peak level achieved during focused mainly on molecular changes in the skeletal muscle the 8 weeks of DIO [19], and this was confirmed in the and liver, while considerably less is known about changes in current study. Since body weight and body composition data adipose tissue [11, 15, 16]. Therefore, the aim of this study on this DIO control diet were comparable to continuous was to compare the effect of CR and EX on visceral white DIO, the DIO control group was used as the comparator adipose tissue (VWAT) gene expression, along with changes for all analyses. This also provided control for changes in body composition and insulin resistance, in diet-induced in expression due to differences in diet composition/fat obese mice. We utilized an animal model of postmenopausal consumption, since the diets for the DIO control, CR, and obesity because postmenopausal women are especially at risk EX groups all based on the AIN-76A diet. The CR group for developing diseases associated with obesity such as type 2 consumed a modified diet (D0302702, administered in daily diabetes [17] and breast cancer, the second leading cause of aliquots) providing 30% fewer calories from carbohydrates cancer death in women [4]. compared to the control diet, with all other components In the present study, we show that, despite comparable being isonutrient when intake was limited to 70% of mean reductions in adiposity in obese mice, CR was more effec- kcal consumption of the diet control group. The EX group tive than EX at increasing adiponectin, improving insulin were run on a variable speed treadmill 5 days/wk on a 5% sensitivity, and altering the adipose transcriptome. Although grade, beginning with 10 min/day at 12 m/min. Time and both CR and EX qualitatively affected a shared set of intensity were increased gradually over the next two weeks genes related to metabolism, CR had a stronger quantitative until the EX group reached 40 min/day at a maximum rate of effect on these genes. Furthermore, CR induced a dramatic 20 m/min. The DIO control, continuous DIO and CR mice change in expression of an additional set of genes related to were all placed on the treadmill but not run. Body weights carbohydrate metabolism and transport in VWAT that was and feed intake were measured weekly. not observed in the EX mice. At the beginning of week 17, when the CR and EX mice achieved comparable reductions in adiposity relative to the 2. Materials and Methods DIO controls, mice were euthanized. In the morning the mice were killed, all mice received their respective dietary 2.1. Animal Study Design. All animal protocols were ap- or exercise treatment, followed by a 6-hr fast. Mice were proved by the University of Texas at Austin Institutional anesthetized with isofluorane for terminal blood collection Animal Care and Use Committee. To model the postmeno- via the retro-orbital venous plexus, and then killed by pausal state, 6-week-old ovariectomized C57BL/6 mice were cervical dislocation. Whole blood was allowed to clot at room used (Charles River Labs, Inc. Frederick, Md, USA). Ovariec- temperature for 30 min prior to centrifugation at 1000 ×g tomized mice exhibit characteristics of the postmenopausal for10min.Theserumwasremovedandstoredat−80◦Cfor state in humans: decreased levels of circulating estrogen, loss analyses. A 1-gram sample of VWAT was collected from each of bone mineral density, and cessation of estrous cycles [18]. mouse and stored at −80◦C until further analyses. Carcasses Upon arrival, mice had ad libitum access to water and chow were stored at −20◦C. Percent body fat and lean mass were diet and were on a 12 : 12 h light/dark cycle. determined using dual energy X-ray absorptiometry (DXA) To compare the effects of CR and EX on reversal of (GE Lunar Piximus II, Madison, WI, USA) as described obesity and insulin resistance, and other metabolic pertur- previously [20]. bations, 48 mice were singly housed upon receipt and put on To further characterize the effect of CR on the histone a diet-induced obesity (DIO) regimen for 8 wks consisting of code (which required different tissue processing procedures ad libitum access to a 60 kcal% fat diet (D12492; Research than the gene expression microarray analysis), an additional Diets, Inc, New Brunswick, NJ, USA), beginning one week group of 15 mice received the AIN-76A control diet (labeled after arrival. At week 9, the mice were randomized into the overweight mice), CR diet (labeled lean mice), or DIO following treatment groups (n = 12/group): (1) DIO control (labeled obese mice) for 8 weeks (n = 5/diet group). (AIN-76A diet fed ad libitum); (2) 30% CR; (3) treadmill Body composition on these mice was determined using exercise regimen, fed AIN-76A diet ad libitum (EX); or quantitative magnetic resonance (Echo Medical Systems, (4) continuation on the DIO regimen. In animal models, Houston, TX, USA). Animals were then killed after an 8- CR diet regimens, typically involving a 20–40 reduction in hr fast, serum collected as described above, and tissues carbohydrate calorie intake and designed to limit total energy (including VWAT, liver and mammary glands 4 and 9) intake while insuring adequate nutrition, represent the most were excised, formadehyde treated to crosslink proteins, potent dietary approach to prevent and/or reverse obesity and immediately flash frozen for analysis by the chromatin and inhibit tumor growth [7]. The DIO control and EX immunoprecipitation assay described below. groups were switched from the DIO regimen to a modified AIN-76A diet (D12450B, that is 10 kcal% fat and is the base 2.2. Glucose Tolerance Test. To determine the effects of CR diet of our CR regimen; Research Diets, New Brunswick, and EX on glucose regulation following weight loss, we NJ,USA)consumedadlibitum.TheDIOcontrolgroupwas conducted a glucose tolerance test (GTT) on the 48 mice Journal of Obesity 3 on study at week 15. GTT was performed after a 6-hour Slc2a4 primers: forward primer 5-CCCTTTAAGGCTCCA- fast by administration of 20% glucose (2 g/kg body weight TCTCC-3 and reverse primer 5-TGTGTGTATGCCCCG- IP). Blood samples were taken from the tail and analyzed AAGTA-3 (ABI). GAPDH was used as the internal control for glucose concentration using an Ascencia Elite XL 3901G for analysis of acetylation with the following primers: glucose analyzer (Bayer Corporation, Mishawaka, Ind). forward primer 5-CATGGCCTTCCGTGTTCCTA-3 and Glucose levels were determined at baseline, 15, 30, 60, and reverse primer 5-CCTGCTTCACCACCTTCTTGAT-3. 120 min after injection of glucose. For analysis of methylation, p16 was used as the internal control with the following primers: forward primer 5- 2.3. Serum Hormones. Leptin, insulin, and adiponectin were ACACTCCTTGCCTACCTGAA-3 and reverse primer 5- measured in serum collected at the terminal bleed, using CGAACTCGAGGAGAGCCATC-3. mouse adipokine LINCOplex Multiplex Assays (Millipore, Inc., Billerica, MA, USA) analyzed on a BioRad Bioplex 200 2.6. Statistics. Values are presented as mean ± standard analysis system (Biorad, Inc. Hercules, CA, USA). error (SE). One-way analysis of variance (ANOVA) followed by Tukey’s Honestly Significant Differences test was used 2.4. Gene Expression Microarray Analysis. Total RNA was to assess the effects of diet on mean weekly body weight isolated from VWAT tissues using an organic extraction at weeks 8 and 16, body composition data at week 16, and precipitation protocol with a DNAseI treatment step serum adipokine levels, and fasting glucose levels. Repeated (Asuragen Inc., Austin, TX, USA). Biotin-labeled targets measures analysis was used to evaluate glucose tolerance were prepared using modified MessageAmp-based protocols tests. For serum insulin, mRNA levels (as measured by (Ambion Inc., Austin, TX, USA) and hybridized to MOE qRT-PCR), and relative quantification of Glut 4 in ChIP 430A 2.0 arrays (Affymetrix, Santa Clara, CA, USA). The experiments means, were compared using Student’s t-test. arrays were scanned on an Affymetrix GeneChip Scanner For all tests SPSS software was used (SPSS Inc., Chicago, IL, 3000 7G. A summary of the image signal data, detection USA), and P ≤ .05 was considered statistically significant. calls, and gene annotations for every gene interrogated on the array was generated using Affymetrix Statistical Algorithm 3. Results MAS 5.0 (GCOS v1.3), with all arrays scaled to 500. Sample normalization was carried out using the Robust Multichip 3.1. Both CR and EX Decrease Adiposity, Insulin and Leptin Average (RMA) followed by multiple group analysis compar- Levels, but Only CR Increases Adiponectin and Restores Insulin ison using ANOVA. Pairwise comparisons were performed Sensitivity in DIO Mice. During the first 8 weeks, the DIO to identify expression fold differences with false discovery regimen increased mean body weight of the 48 mice on study rate (FDR) set at 0.05. Genes with expression differences from 20.3 ± 0.5 g to 30.7 ± 0.5 g, and % body fat to 52.3%. equal or greater than 2-fold compared to DIO controls, were As shown in Table 1, one week after randomization (week selected to be analyzed using the Database for Annotation, 9 of the study), the DIO control, EX, CR, and continuous Visualization and Integrated Discovery (DAVID; [21]). The DIO groups did not differ in body weight. However, by week resulting Gene Ontology (GO) analysis was used to identify 16 of the study, the DIO control group (30.8 ± 1.6 g) was genes relevant to the different effects of CR and EX in significantly heavier than the EX mice (26.0 ± 0.9 g) and reversing obesity, some of which were selected for further the CR mice (19.9 ± 0.5 g), but not the continuous DIO analysis. In the DAVID analysis, genes that were represented group (33.2 ± 1.5 g). The body weight data closely correlated more than once in the microarray output were filtered. with calorie intake (for weeks 9–16: 709 ± 10.0 kcal for DIO Some of the genes in the Gene Ontology analysis belonged controls; 556 ± 4.6 kcal for the EX mice; 413.0 ± 4.3 kcal for to more than one functional category and are tabulated the CR; and 722 ± 11.6 kcal for the continuous DIO group) accordingly. Expression changes were verified in VWAT from and % body fat (Table 1). Although the CR mice weighed a separate cohort of mice that underwent CR or EX following significantly less than the EX mice (primarily due to the DIO, as described above, using Taqman Gene Expression increase in lean mass in the EX group relative to the CR Assay (Applied Bioystems Inc., Carlsbad, CA, USA). Gene mice), there was no difference in percent body fat, with both expression data were normalized to the housekeeping gene groups exhibiting >25% reductions in % body fat compared β-actin. to DIO control mice. Achieving meaningful reductions in adiposity in obese mice via CR and EX was a goal of the 2.5. Chromatin Immunoprecipitation (ChIP) Assay. ChIP study design, given that percent body fat is associated with assays were performed per manufacturer’s instructions (Mil- insulin resistance and other key metabolic changes associated lipore). Briefly, proteins from VWAT were formaldehyde with DIO [22]. Since body weight, kcal consumption and crosslinked to DNA. After homogenization, lysis, and son- body composition data for mice on the DIO control diet were ication, proteins were incubated overnight with antibodies comparable to mice on the continuous DIO diet, the DIO to acetyl-histone H4 or trimethyl histone H4 (Millipore). control group was used as the comparator for all analyses. The DNA-protein complexes were washed, DNA was eluted, This allowed us to eliminate the possibility that any changes and crosslinking was reversed by heating to 65◦C overnight. observed in hormones and gene expression could have been DNA was purified using QIAGEN PCR purification kit due to differences in diet composition/fat consumption, (QIAGEN, Valencia, CA, USA). Quantitative, real-time PCR since the diets for the DIO control, CR, and EX groups was performed using SYBR Green (ABI) with the following were all based on the same AIN-76A diet composition. 4 Journal of Obesity

Table 1: Body composition after 8 weeks of DIO followed by 8 weeks of control diet, exercise, or calorie restriction.

Group Body weight week 9 (g) Body weight week 16 (g) Percent body fat (%) Lean mass (g) DIO Control 30.1 ± 0.9a 30.8. ± 0.6a 51.1 ± 3.8a 12.6 ± 0.2a Exercise 29.3 ± 0.6a 26.0 ± 0.9b 38.9 ± 2.7b 13.6 ± 0.2b Calorie Restriction 31.3± 0.6a 19.9 ± 0.5c 33.7 ± 1.4b 10.5 ± 0.1c ContinuousDIO∗ 30.3 ± 1.2a 33.2± 1.5a 57.3 ± 2.9a 11.8 ± 0.2d Data are presented mean ± SEM. Significant differences (P<.05) beween data within a column are indicated by different superscripts; n = 12/group.

As shown in Figure 1(a), this also allowed us to limit our both CR and EX, although CR had a stronger quantitative hormone and microarray analyses to 3 groups (DIO control, effect. CR, and EX), without the continuous DIO group, thus Reductions in adiposity are accompanied by lower levels increasing the number of mice per group analyzed within our of immune cell infiltrates into adipose tissue, which mediate budget constraints. the proinflammatory state associated with obesity [24]. At the end of the study we also measured circulating As expected, the reduced adiposity in CR and EX mice leptin and adiponectin levels, two adipokines that are pos- was associated with decreased expression of genes related itively and negatively correlated with adiposity, respectively to immune response (Figure 2(b)). These immune-related [23]. Consistent with decreased adiposity, leptin levels were genes also comprised the majority of the genes in the stress roughly 80% lower in the CR and EX mice (Figure 1(b)). response category, including downregulation of transcripts However, only CR increased adiponectin levels compared that code for chemokines that attract and are produced by to DIO control mice (Figure 1(c)), even though percent monocytes and macrophages, specifically Chemokine (C-C body fat in CR and EX mice did not statistically differ. The motif)ligand(Ccl)2,6,7,and9. higher levels of adiponectin observed in the CR mice were associated with decreased fasting insulin levels (Figure 1(d)), 3.3. Unique Transcriptional Changes in Response to CR or decreasedfastingglucoselevels(Figure 1(e)), and increased EX in VWAT. CR uniquely affected expression of 496 genes, insulin sensitivity as indicated by significantly lower blood whereas a mere 20 genes were responsive only to EX glucose levels at every time point following glucose challenge (Figure 2(a)). GO analysis of the genes uniquely responsive (Figure 1(e)). In contrast, the EX mice did not display to EX revealed that only the grouping of genes related increased insulin sensitivity or decreased fasting insulin levels to mitochondrial transport was significant. Specifically, compared to sedentary DIO. Taken together, these data uncoupling proteins Ucp1 and Ucp2 were both upregulated demonstrate that CR and EX differentially affected adipose by EX. Given the robust transcriptional response to CR, we tissue metabolism. focused our analysis on those genes whose expression was affected by CR but not EX (Table 2). GO analysis showed 3.2. Transcriptional Changes Common to CR and EX in that in every category of genes altered by both CR and EX, VWAT. Gene expression microarray analysis was performed CR impacted an additional set of genes unaffected by EX. on VWAT collected following the 8-wk weight-loss phase For example, in genes relating to cellular lipid metabolic pro- after DIO. Pairwise comparisons of DIO versus CR and DIO cesses, which was the largest subset of transcripts uniquely versus EX revealed that 725 transcripts were significantly altered by CR, soluble carrier family 27 (Slc27a1) and Acetyl- altered (±2.0 fold, P<.05, Figure 2(a)). Of those 725 Coenzyme A carboxylase alpha (Acaca) were upregulated. transcripts, 209 were common to CR and EX (Figure 2(a)), CR also uniquely increased expression of sterol regulatory possibly representing a suite of genes most sensitive to energy element binding transcription factor 1 (Srebp1), a master balance. GO analysis was used to categorize these genes regulator for lipid metabolism in adipocytes. With respect according to function and revealed that the majority of genes to immune response and stress response, CR resulted in a altered both by CR and EX were related to metabolic process, downregulation of gene expression, whereas expression of immune response, and stress response (Figure 2(b)). Within genes related to biosynthesis of steroids was upregulated. the metabolic process category, 24 of the genes were related In addition to affecting more genes in each functional to lipid metabolism, and overall the response of the genes category than EX, CR affected the transcription of genes to CR and EX was qualitatively similar. More specifically, in another category not modulated by EX, specifically 4 a number of genes involved in fatty acid synthesis and genes related to glucose transport. Complementing this transport were upregulated (Figure 2(c)). These included increase in transcription of glucose transport genes, CR stearoyl-CoA desaturase (Scd1), fatty acid synthase (Fasn), resulted in upregulation of another 14 genes related to carnitine palmitoyltransferase 1 (Cpt1), and elongation of carbohydrate metabolism processes. Taken together, these long chain fatty acids 3 and 6 (ELOVL3 and ELOVL6). In data are suggestive of increased glucose flux into the adipose addition, 9 genes related to glucose metabolism were affected tissue, which may underlie the enhanced insulin sensitivity by CR and EX, including pyruvate dehydrogenase E1 alpha observed in response to CR. 1(Pdha1), leptin (Lep), and glycerol phosphate dehydrogenase 2 (Gpd2). As in lipid metabolism, genes related to 3.4. Real-Time RT-PCR Confirmation of Microarray Results. carbohydrate metabolism were qualitatively responsive to Given that the DIO mice were on a high-fat diet, and the CR Journal of Obesity 5

Diet-induced obesity High-fat diet Ad libitum n = 45 1–8 weeks Weight-gain phase

DIO EX CR High fat diet Control diet 30% energy Ad libitum restriction Ad libitum n n = 15 + Exercise = 15 9–16 weeks n = 15 Weight-loss phase

(a) Animal Study Design for Microarray

20 50 b a

15 40 g/mL) µ a 30 a 10 20 b 5 Serum leptin (ng/mL) 10 b Serum adiponectin ( 0 0 DIO CR Ex DIO CR Ex (b) (c)

2 500 a 400 1.5

300 a, b 1 b 200

0.5 Serum insulin (ng/mL) Blood glucose (mg/dL) 100

0 0 DIO CR Ex 0 15 45 60 120 (minutes)

DIO CR Ex (d) (e)

Figure 1: Effect of calorie restriction or exercise in diet-induced obese mice on serum hormones and glucose tolerance. (a) Animal study design for gene expression microarray experiments. (b) Serum leptin levels, (c) serum adiponectin levels, and (d) serum insulin levels after 8 weeks of intervention, (n = 11 for DIO group; n = 10 for CR group; n = 10 for EX group). (e) Blood glucose concentrations during a glucose tolerance after 7 weeks of intervention. Data shown are mean ± SE. DIO (•), EX (), CR (), n = 12/group. Significance (P ≤ .05) between groups is denoted by different letters. 6 Journal of Obesity

CR EX Number of Percentage P Gene Ontology genes of input -value Metabolic process 79 44.13% .02 496 209 20 Immune response 17 9.5% <.001 Response to stress 15 8.38% .009 Steroid biosynthetic 6 3.35% <.001 process (b) DAVID analysis of genes common to CR and EX (a) Venn diagram comparing CR and EX CR EX CR EX

CR EX Number of Percentage Gene ontology P-value Psph Prps1 Ubd genes of input Scd2 Adam8 Gpnmb Response to stress 30 6.82% .007 Scd1 Clqb Tph2 Pdhb Soat1 C3arl Cellular lipid metabolic 30 6.82% <.001 Gsta4 Pla2g5 Gpr50 process Me1 C4b Cellular carbohydrate Elovl6 Ptx3 14 3.18% .007 Pygl Mmp3 process Cidea Ltc4s Immune response 7 1.59% .04 Egln3 Mrc2 Steroid biosynthetic 7 1.59% .004 Pank1 Lep process Rdh11 Lipa Mogat1 Mest Glucose transport 4 0.91% .02 Lss Dclk1 Gbel Egr2 Elovl3 Rrm2 (d) DAVID analysis of genes unique to CR Acss2 Cpt1a Cyp51 Evl Fdps Dck Gsta3 Prps1 Pdk1 Ctss Aacs Hp Fasn Lgmn Trub2 Clqc Acsm3 Adcy7 Sorl1 Capg Dhrs7 Pcolce2 Pmvk Mafb Acly Daglb Enpep Cyba Mvd Timp1 Pdha1 C5arl Cox8b Aadacl Ptges Ddha1 Tkt Idi1 Gpd2 Acaca Ddo

(c) Heat map for metabolic genes common to both CR and EX Figure 2: Effect of weight loss induced by calorie restriction or exercise on mRNA expression in VWAT. (a) Venn Diagram of genes differentially expressed by CR and EX compared to DIO controls. (b) Classification of genes targeted by both CR and EX. (c) Heat map of genes related to metabolic processes affected by both CR and EX. (d) Classification of genes targeted uniquely by CR (n = 6/group).

and EX consumed a low-fat diet, we were concerned that the responsive to CR and relating to carbohydrate metabolism observed differences in expression of metabolic genes might and transport (Slc2a4, Acly, and Sh2b) were selected for val- be due to differences in the macronutrient contents of the idation. RT-PCR analysis verified that Ucp1, but not Ucp2, diets and not energy balance per se. To address this concern, was significantly increased by EX only (Figure 3). Although confirmatory analysis of mRNA expression was done using according to the microarray analyses, Lep was reduced by the diet control mice as the reference group. A gene that was both CR and EX, RT-PCR analyses revealed that only CR sig- responsive to both CR and EX (Lep), two genes uniquely nificantly reduced Lep expression (Figure 3). All three genes responsive to EX (Ucp1 and Ucp2), and three genes uniquely relating to carbohydrate metabolism and transport, Acly, Journal of Obesity 7

1.5 7 4 6 5 3 1 ∗ 4 2 3 0.5 ∗ 2 1 1 Relative mRNA expression Relative mRNA expression 0 Relative mRNA expression 0 0 Diet control CR Ex Diet control CR Ex Diet control CR Ex (a) Lep (b) Ucp1 (c) Ucp2 ∗ 4 4 ∗ 3 ∗ 3 3 2 2 2 1 1 1 Relative mRNA expression Relative mRNA expression 0 0 Relative mRNA expression 0 Diet control CR Ex Diet control CR Ex Diet control CR Ex (d) Slc2a4 (e) Acly (f) Sh2b2

Figure 3: Conﬁrmation of microarray data analysis of mRNA expression in VWAT. Expression of mRNA transcripts in VWAT from DIO control, CR and EX mice (n = 7/group). (a) Leptin (Lep), (b) Uncoupling protein 1 (Ucp1), (c) Uncoupling protein 2 (Ucp2), (d) Solute carrier 2, family 4 (Slc2a4) (e) ATP-citrate lyase (Acly), and (f) SH2B adapter protein 2 (Sh2b2) (Data shown are mean ± SE, ∗(P ≤ .05)).

Slc2a4, and Sh2b2, were indeed significantly increased by CR Glut4mRNAlevelsintheVWAT(Figures4(b), 4(c),and only (Figure 3). Importantly, Slc2a4 codes for the insulin- 4(d)). Modifications to the histone code such as methylation, responsive glucose transporter, Glut4. Translocation of Glut4 which can result in decreased transcription [26] or acetyla- from the cytosol to the plasma membrane in response to tion, which can result in increased transcription [25]may insulin signaling is the rate-limiting step of glucose transport account for the differences in Glut4 transcription in VWAT, into the adipocyte. Furthermore, downregulation of Glut4 at so both forms of epigenetic alteration were assessed. There the messenger RNA and protein levels has been implicated in were no differences in trimethylation of histone 4 at the Glut4 obesity and insulin resistance. Although we lacked sufficient promoter (Figure 5(a)). However, CR significantly increased VWAT samples for an extensive protein analysis, due to the histone 4 acetylation at the Glut4 promoter compared to use of these tissues for genomic and other analyses, Western control mice (Figure 5(b)), which was associated with higher blot analyses for Glut4 protein expression on 3 VWAT levels of Glut4 mRNA and increased insulin sensitivity. samples/group showed similar trends as observed with the mRNA analyses. Specifically, the lowest Glut 4 protein 4. Discussion expression was observed in VWAT from a control mouse, the highest expression was in a CR sample, and the samples With over two thirds of American adults classified as from the exercise group were similar to the controls (data not overweight or obese [1], increased understanding of how best shown). Finally, increases in the enzyme ATP-citrate lyase to reverse the harmful effects of obesity is urgently needed. (Acly), which was also upregulated by CR but not EX, has Given the critical role of adipose tissue in regulating glucose recently been linked to increases in Glut4 mRNA levels [25]. homeostasis and other aspects of metabolism, analysis of the For these reasons, we focused our analyses on elucidating changes that occur in adipose tissue after weight loss could how CR resulted in increased transcription of Glut4. reveal novel targets for prevention or treatment of obesity- related diseases. To our knowledge, this is the first study to 3.5. CR Results in Acetylation of Histone 4 at the GLUT4 compare the effects of CR and EX (the two most commonly Promoter. Since regulation of acetylation of histones has recommended lifestyle modifications to prevent or reverse been shown to be nutrient dependent [25], we hypothesized obesity) on gene expression in adipose tissue in a model of that increased Glut4 mRNA expression in CR mice may be DIO. The direct comparison of these two obesity reversal the result of histone modifications at the Glut4 promoter. To interventions revealed the following novel findings: (1) CR test this hypothesis, we generated obese (DIO) and lean (CR) led to altered expression of more than 20 times the number mice (relative to overweight control mice consuming AIN- of genes in the adipose tissue than were uniquely affected 76A diet ad libitum) through an 8-week diet intervention. by EX; (2) alteration of expression of carbohydrate transport Body weight and percent body fat were positively associated genes (particularly GLUT4) was uniquely affected by CR and with fasting blood glucose levels and inversely related to correlated with the increased insulin sensitivity exhibited by 8 Journal of Obesity

DIO Overweight Lean High-fat diet Control diet 30% energy Ad libitum Ad libitum restriction 8weeks n = treatments n = 5 n = 5 5 Energy balance

(a) 45

35 200 a 30 b 25

Average body weight (g) 20 100 c 15 0246 Weeks on study

DIO Fasting serum glucose (mg/dL) Overweight 0 Lean DIO Overweight Lean (b) (c) 2 c

b 1

a Relative mRNA expression

0 DIO Overweight Lean (d)

Figure 4: Lean phenotype is associated with lower blood glucose levels and elevated levels of Glut4 mRNA relative to control mice. (a) Study design for chromosomal immunoprecipitation experiments. (b) Average body weight of mice on DIO, overweight (control AIN-76A) diet, or a lean (CR) regimen to generate obese, overweight or lean mice (n = 5/group). Data shown are mean ± SE. (c) Fasting blood glucose levels of mice after 8 weeks of respective dietary regimen (n = 5/group). Data shown are mean ± SE. Significance (P<.05) between groups is denoted by different letters. (d) Relative mRNA expression of Glut4 in VWAT (n = 5/group). Significance (P<.05) between groups is denoted by different letters.

CR; (3) upregulation of Glut 4 by CR may be explained in effects are less clear in obese individuals such as the DIO part by our finding that CR increased acetylation of histone mice used in this report. The relatively short intervention 4attheGlut4promoter. in our study may explain why EX was not as affective CR and EX both resulted in significant weight loss as CR at altering indices of insulin resistance. In other compared to sedentary DIO controls, which remained obese rodent studies that showed a significant affect of EX, the with a % body fat >50%. Although CR, and EX groups intervention was either more than 10 weeks long [10, 28, 29] displayed comparable levels of percent body fat at the end or the intervention period was longer than the period of of the intervention, only CR significantly improved insulin diet-induced obesity [15]. These differences in study design sensitivity. Exercise has been shown to increase insulin suggest that, in the short term, EX may not be as effective as sensitivity in mice and humans [27, 28], although these CR in restoring insulin sensitivity. Journal of Obesity 9

Table 2: Transcriptional changes in response to calorie restriction or exercise in visceral white adipose tissue.

Gene Symbol Gene title Fold change Cellular lipid metabolic process CR × DIO Slc27a1 Solute carrier family 27 (fatty acid transporter) 2.53 up Fads2 Fatty acid desaturase 2 2.18 up Ces3 Carboxylesterase 3 2.51 up Sult1a1 Sulfotransferase family 1A 2.22 up Ptges Prostaglandin E synthase 2.25 up Sgpp1 Sphingosine-1-phosphate phosphatase 1 2.73 down Echs1 Enoyl Coenzyme A hydratase 2.00 up Hsd11b1 Hydroxysteroid 11-beta dehydrogenase 1 3.28 up Apoc3 Apolipoprotein C-III 5.71 up Srebf1∗ Sterol regulatory element binding transcription factor 1 2.98 up Aldh1a7 Aldehyde dehydrogenase family 1 2.18 up Hpgd Hydroxyprostaglandin dehydrogenase 15 (NAD) 2.09 down Rdh11 Retinol dehydrogenase 11 4.59 up Rarres2 Retinoic acid receptor responder (tazarotene induced) 2 2.61 down Nsdhl∗ NAD(P) dependent steroid dehydrogenase-like 2.65 up Gpam Glycerol-3-phosphate acyltransferase 2.06 up Hmgcs1∗ 3-hydroxy-3-methylglutaryl-Coenzyme A synthase 1 3.01 down Abat 4-aminobutyrate aminotransferase 2.88 down Sorl1 Sortilin-related receptor 2.77 up Pip4k2a Phosphatidylinositol-5-phosphate 4-kinase 2.02 down Acaca Acetyl-Coenzyme A carboxylase alpha 4.60 up Tm7sf2 ∗ Transmembrane 7 superfamily member 2 3.69 up Sc5d∗ Sterol-C5-desaturase 2.11 up Fdft1∗ Farnesyl diphosphate farnesyl transferase 1 2.68 up Hsd17b12∗ Hydroxysteroid (17-beta) dehydrogenase 12 2.17 up Pcx∗ Pyruvate carboxylase 2.88 up Cellular carbohydrate metabolic process Fn3k Fructosamine 3 kinase 4.05 up Chst1∗ Carbohydrate (keratan sulfate Gal-6) sulfotransferase 1 2.00 up Dlat Dihydrolipoamide S-acetyltransferase 2.59 up Pkm2 Pyruvate kinase 2.37 up Pmm1 Phosphomannomutase 1 2.72 up Ppp1r1a Protein phosphatase 1 2.61 up Pgd Phosphogluconate dehydrogenase 2.40 up Agl Amylo-1 2.16 up Oxct1 3-oxoacid CoA transferase 1 2.17 down Pdk1 Pyruvate dehydrogenase kinase 3.07 up Gpd1 Glycerol-3-phosphate dehydrogenase 1 (soluble) 2.02 up Taldo1 Transaldolase 1 2.03 up Glucose transport Sh2b2∗ SH2B adaptor protein 2 3.26 up Slc2a4 Solute carrier family 2 (facilitated glucose transporter) 3.43 up Pcx∗ Pyruvate carboxylase 2.88 up Klf15 Kruppel-like factor 15 2.36 up Immune response CR × DIO Malt1 Mucosa associated lymphoid tissue Lymphoma translocation gene 1 2.55 down Bcl6∗ B-cell leukemia/lymphoma 6 2.90 down Clec7a∗ C-typelectindomainfamily7 2.89 down Cfb Complement factor B 2.53 down Cd55∗ CD55 antigen 2.25 Down Thy1 Thymus cell antigen 1 2.44 down 10 Journal of Obesity

Table 2: Continued. Gene Symbol Gene title Fold change Biosynthesis of Steroids Lss Lanosterol synthase 2.13 up Hmgcs1∗ 3-hydroxy-3-methylglutaryl-Coenzyme A synthase 1 3.01 down Tm7sf2 ∗ Transmembrane 7 superfamily member 2 3.69 up Sc5d∗ Sterol-C5-desaturase 2.11 up Fdft1∗ Farnesyl diphosphate farnesyl transferase 1 2.68 up Hsd17b12∗ Hydroxysteroid (17-beta) dehydrogenase 12 2.17 up Nsdhl∗ NAD(P) dependent steroid dehydrogenase-like 2.65 up Sh2b2∗ SH2B adaptor protein 2 3.26 up Stress response Thbs1 Thrombospondin 1 2.29 down Tfpi2 Tissue factor pathway inhibitor 2 4.32 down Gp1bb Glycoprotein Ib 2.06 up Taok3 TAO kinase 3 2.21 down Sod3 Superoxide dismutase 3 2.17 down Dusp10 Dual speciﬁcity phosphatase 10 2.62 down Adrb3 Adrenergic receptor 2.09 up F2r Coagulation factor II (thrombin) receptor 2.22 down Ccnd1 Cyclin D1 2.91 down Evl Ena-vasodilator stimulated phosphoprotein 2.32 down Ctsb Cathepsin B 2.00 down Ly86 Lymphocyteantigen86 2.65 down Fabp4 Fatty acid binding protein 4 2.03 up Rad50 RAD50 homolog (S. cerevisiae) 2.27 down Tsc22d2 TSC22 domain family 2 2.41 down Ptger3 Prostaglandin E receptor 3 (subtype EP3) 2.21 up Lcp1 Lymphocyte cytosolic protein 1 3.06 down Pros1 Protein S (alpha) 2.09 down Hspa12a Heat shock protein 12A 2.40 down Anxa2 Annexin A2 2.01 down Uhrf1 Ubiquitin-like 2.84 down Cdkn1a Cyclin-dependent kinase inhibitor 1A (P21) 2.28 down Srebf1∗ Sterol regulatory element binding transcription factor 1 2.98 up Chst1∗ Carbohydrate (keratan sulfate Gal-6) sulfotransferase 1 2.00 up Bcl6∗ B-cell leukemia/lymphoma 6 2.90 down Clec7a∗ C-type lectin domain family 7 2.89 down Cd55∗ CD55 antigen 2.25 down Exercise unique Mitochondrial Transport EX × DIO Ucp1 Uncoupling protein 8.94 up Ucp2 Uncoupling protein 2 2.09 down ∗ Genes represented in two diﬀerent categories.

CR has been shown to decrease expression of genes by our group in the mouse mammary gland [31] and the related to aging and tumorigenesis in multiple tissues other by Lu et al. in mouse skin [32]. In these reports CR [30]. However, there is a paucity of studies examining the and EX exhibited distinct effects on gene expression, with CR effect of CR on adipose tissue following weight loss. More impacting more than 4 times the number of genes than EX. importantly, there are no studies, to our knowledge, directly In the present study, we found that this differential impact comparing the effect of CR and EX on gene expression was more pronounced in adipose tissue, with CR affecting in adipose tissue. To our knowledge there are only two more than 20 times the number of genes altered by EX. Not microarray studies comparing CR to EX; one was performed only did CR induce a stronger quantitative effect than EX on Journal of Obesity 11

2 2.5 ∗ 2

1.5 1 1 Relative quantiﬁcation of

Relative quantiﬁcation of 0.5 immuno precipitated DNA immuno precipitated DNA

0 0 DIO Overweight Lean DIO Overweight Lean (a) H3K9 Trimethylation of Glut4 promoter (b) H4 Acethylation of Glut4 promoter

Figure 5: Calorie restriction increases histone H4 acetylation at the GLUT 4 promoter. (a) Relative quantiﬁcation of Glut4 DNA immunoprecipitated with anti-trimethyl H4 antibody. (b) Relative quantiﬁcation of Glut4 DNA immunoprecipitated with anti-acetyl H4 antibody (n = 3/group). Data shown are mean ± SE, ∗(P ≤ 0.05).

genes that were qualitatively similar in their response to both from the cytosol to the membrane is the rate-limiting step in CR and EX, but CR affected an additional 48 genes related to insulin-mediated glucose uptake in adipocytes and skeletal metabolism that were unaffected by EX. Of those genes, there muscle [37]. The importance of Glut4 function in adipose was an overall upregulation of genes related to carbohydrate tissue is underscored by the finding that overexpression of metabolism and glucose transport, including Glut4. Glut4 in adipocytes rescues insulin resistance in mice with We also found that DIO downregulates multiple genes muscle-specific knockout of Glut4 [38]. However, expression that play a role in lipid metabolism and upregulates a profile of Glut4 in the muscle does not compensate for lack of Glut4 of genes related to immune/inflammatory response [33, 34]. activity in adipose tissue [39], further implicating adipose Furthermore, many of the lipid metabolism genes shown to tissue as a key metabolic organ in the etiology of insulin be decreased by DIO were increased by CR and EX in our resistance. There is considerable evidence that Glut4 mRNA study [33]. Likewise, immune response genes that have been levels in adipose tissue decrease with obesity [40], and that shown to be increased by DIO were decreased after weight increases in Glut4 mRNA in adipose tissue can ameliorate loss induced by CR or EX [33]. Together these data support insulin resistance [41, 42]. Indeed, our finding that Glut4 previous findings that in the obese state, there is diminished mRNA levels were significantly increased by CR, but not fatty acid synthesis and transport, characteristic of insulin- by EX, and that this increase was associated with improved resistant adipose tissue rich in immune cell infiltrates. insulin sensitivity, supports this idea. Therefore, increased Importantly, our data show that these processes are sensitive transcription of Glut4 in VWAT during weight loss may be a to both CR and EX interventions. critical event in reversing insulin resistance. Many of the transcripts related to lipid and carbohydrate Studies into the transcriptional regulation of Glut4 in metabolism that were affected by both CR and EX in the skeletal muscle implicate a histone deacetylase (HDAC5) present study were also shown by Shankar et al. to be induced as a crucial mediator of changes to Glut4 mRNA levels by a high-carbohydrate diet [35]. Increased transcription of in response to exercise [43]. Raychaudhuri et al. have also these genes is consistent with increased uptake of glucose and described a series of histone modifications mediated by fatty acids into the adipose tissue. In the study by Shankar histone deacetylases and histone methyltransferases that et al. these transcriptional changes were measured in rats culminate in a metabolic knockdown of the Glut4 gene in fed a high-carbohydrate diet for 4 weeks, during which time the skeletal muscle of rats that had experienced intrauterine the rats gained weight and the adipocytes hypertrophied, growth restriction [44]. Collectively, these studies suggest whereas the mice in our study first underwent DIO but that transcriptional regulation of the Glut4 gene is highly then lost weight for 8 weeks before analysis. The similarities responsive to changes in energy balance. This led to our between the two studies are indicative of increased signaling hypothesis that Glut4 mRNA levels in adipose tissue could be through the insulin receptor/phosphatidylinositol 3-kinase subjected to similar transcriptional regulation. In support of (IR/PI3K) pathway that mediates glucose uptake and the this hypothesis, Wellen et al. recently discovered that during lipogenic effects of insulin in adipose tissue. adipocyte differentiation, levels of global histone acetylation Glucose uptake into adipose tissue is mediated by two are dependent on glucose availability [25]. More specifically, different Glut isoforms: Glut1 and Glut4. Glut1 mediates acetylation of histones 3 and 4 at the Glut4 promoter is basal uptake of glucose into adipocytes. Although others linked to increased Glut4 mRNA expression in response to have reported that Glut1 mRNA increases with obesity [36], higher concentrations of glucose during differentiation. Our we did not observe any changes in Glut1 mRNA expression in vivo ChIP data extend the in vitro findings and show that in the microarray. Translocation of the GLUT4 transporter increased acetylation of histone 4 at the Glut4 promoter, 12 Journal of Obesity which was associated with higher levels of Glut4 mRNA, the general population: are there differences between men occurred in lean mice that were highly insulin sensitive as and women? The MONICA/KORA Augsburg Cohort Study,” indicated by significantly decreased fasting glucose levels. American Journal of Clinical Nutrition, vol. 84, no. 3, pp. 483– Taken together, these data suggest that insulin-responsive 489, 2006. adipose tissue maintains H4 acetylation. This leads to [4]E.E.Calle,C.Rodriguez,K.Walker-Thurmond,andM.J. increased transcription of Glut4 to facilitate continued Thun, “Overweight, obesity, and mortality from cancer in glucose uptake. However, as adiposity increases so does a prospectively studied cohort of U.S. Adults,” New England insulin resistance [3]. Deregulation of signal transduction Journal of Medicine, vol. 348, no. 17, pp. 1625–1638, 2003. downstream of the insulin receptor results in decreased traf- [5] R. T. Watson and J. E. Pessin, “GLUT4 translocation: the last ficking of Glut4 to the cell membrane [45, 46] and a decline 200 nanometers,” Cellular Signalling, vol. 19, no. 11, pp. 2209– in glucose flux into the adipocyte [47]. According to the 2217, 2007. findings of Wellen et al., limited glucose availability results in [6]A.S.GreenbergandM.S.Obin,“Obesityandtheroleof diminished histone acetylation and decreased Glut4 mRNA adipose tissue in inflammation and metabolism,” American expression [25]. Therefore, in the context of obesity and Journal of Clinical Nutrition, vol. 83, no. 2, pp. 461S–465S, 2006. insulin resistance, the lower levels of Glut4 mRNA expressed in adipose tissue may be a consequence of decreased insulin- [7] S. Crowe, L. E. Wu, C. Economou et al., “Pigment epithelium- derived factor contributes to insulin resistance in obesity,” Cell mediated uptake of glucose that results in diminished Metabolism, vol. 10, no. 1, pp. 40–47, 2009. histone acetylation at the Glut4 promoter. Further analyses ff are required to determine if other modifications to the [8] P. Sartipy and D. J. Loskuto , “Monocyte chemoattractant protein 1 in obesity and insulin resistance,” Proceedings of the histone code at the Glut4 promoter may be contributing to National Academy of Sciences of the United States of America, transcriptional repression of Glut4 mRNA in obesity. vol. 100, no. 12, pp. 7265–7270, 2003. In conclusion, these findings show that obesity rever- [9] S. General, The Surgeon General’s Call to Action to Prevent and sal by CR versus EX results in many shared, but also Decrease Overweight and Obesity, U.S. Department of Health many differential, changes in the adipose transcriptome. In and Human Services, 2001. ff particular, CR has specific and significant e ects on the [10] E. P. Weiss, S. B. Racette, D. T. Villareal et al., “Improvements expression of key metabolic genes and pathways associated in glucose tolerance and insulin action induced by increasing with obesity-related disease. In addition, some of the effects energy expenditure or decreasing energy intake: a randomized of these antiobesity interventions on VWAT gene expression controlled trial,” American Journal of Clinical Nutrition, vol. and metabolism may result from chromatin remodeling, as 84, no. 5, pp. 1033–1042, 2006. illustrated by CR’s effect on histone acetylation of the GLUT4 [11] R. H. Coker, R. H. Williams, S. E. Yeo et al., “The impact of promoter. Taken together, these studies provide insights exercise training compared to caloric restriction on hepatic regarding new targets, including potential epigenetic-related and peripheral insulin resistance in obesity,” Journal of Clinical regulation of key metabolic genes, such as Glut4 acetylation, Endocrinology and Metabolism, vol. 94, no. 11, pp. 4258–4266, for preventing or treating obesity-related diseases. 2009. [12] I. Giannopoulou, L. L. Ploutz-Snyder, R. Carhart et al., “Exercise is required for visceral fat loss in postmenopausal Acknowledgments women with type 2 diabetes,” Journal of Clinical Endocrinology and Metabolism, vol. 90, no. 3, pp. 1511–1518, 2005. This research was supported in part by the National Institutes of Health through the UT-MD Anderson Cancer Center R25 [13] A. S. Ryan, B. J. Nicklas, D. M. Berman, and D. Elahi, “Adiponectin levels do not change with moderate dietary Cancer Prevention Research Training Program (CA009480, induced weight loss and exercise in obese postmenopausal to KEW), a Breast Cancer Research Foundation Grant (to women,” International Journal of Obesity,vol.27,no.9,pp. SDH), a Concept Award grant from the DOD Breast Can- 1066–1071, 2003. cer Research Program (BC053292 to SDH), and an Idea [14] B. J. Nicklas, X. Wang, T. You et al., “Effect of exercise intensity Grant from the American Institute for Cancer Research on abdominal fat loss during calorie restriction in overweight (AICR 08A049 to SDH). K. E. Wheatley and L.M. Nogueira and obese postmenopausal women: a randomized, controlled contributed equally to this work. trial,” American Journal of Clinical Nutrition,vol.89,no.4,pp. 1043–1052, 2009. References [15] A. D. Krisan, D. E. Collins, A. M. Crain et al., “Resistance training enhances components of the insulin signaling cascade [1] K. M. Flegal, M. D. Carroll, C. L. Ogden, and L. R. Curtin, in normal and high-fat-fed rodent skeletal muscle,” Journal of “Prevalence and trends in obesity among US adults, 1999– Applied Physiology, vol. 96, no. 5, pp. 1691–1700, 2004. 2008,” Journal of the American Medical Association, vol. 303, [16] K. Y. Lee, S. J. Kim, Y. S. Cha et al., “Effect of exercise on no. 3, pp. 235–241, 2010. hepatic gene expression in an obese mouse model using cDNA [2] T. Y. Li, J. S. Rana, J. E. Manson et al., “Obesity as compared microarrays,” Obesity, vol. 14, no. 8, pp. 1294–1302, 2006. with physical activity in predicting risk of coronary heart [17] C. C. Cowie, K. F. Rust, D. D. Byrd-Holt et al., “Prevalence disease in women,” Circulation, vol. 113, no. 4, pp. 499–506, of diabetes and impaired fasting glucose in adults in the 2006. U.S. population: National Health and Nutrition Examination [3] C. Meisinger, A. Doring,¨ B. Thorand, M. Heier, and H. Survey 1999–2002,” Diabetes Care, vol. 29, no. 6, pp. 1263– Lowel,¨ “Body fat distribution and risk of type 2 diabetes in 1268, 2006. Journal of Obesity 13

[18]S.Z.Haslam,J.R.Osuch,A.M.Raafat,andL.J.Hofseth, [33] R. C. Moraes, A. Blondet, K. Birkenkamp-Demtroeder et “Postmenopausal hormone replacement therapy: effects on al., “Study of the alteration of gene expression in adi- normal mammary gland in humans and in a mouse post- pose tissue of diet-induced obese mice by microarray and menopausal model,” Journal of Mammary Gland Biology and reverse transcription-polymerase chain reaction analyses,” Neoplasia, vol. 7, no. 1, pp. 93–105, 2002. Endocrinology, vol. 144, no. 11, pp. 4773–4782, 2003. [19] K. E. Wheatley, E. A. Williams, N. C. P. Smith et al., “Low- [34] R. S. Miller, K. G. Becker, V. Prabhu, and D. W. Cooke, carbohydrate diet versus caloric restriction: effects on weight “Adipocyte gene expression is altered in formerly obese mice loss, hormones, and colon tumor growth in obese mice,” and as a function of diet composition,” Journal of Nutrition, Nutrition and Cancer, vol. 60, no. 1, pp. 61–68, 2008. vol. 138, no. 6, pp. 1033–1038, 2008. [20] D. Berrigan, J. A. Lavigne, S. N. Perkins, T. R. Nagy, J. C. [35] K. Shankar, A. Harrell, P. Kang, R. Singhal, M. J. J. Ronis, and Barrett, and S. D. Hursting, “Phenotypic effects of calorie T. M. Badger, “Carbohydrate-responsive gene expression in restriction and insulin-like growth factor-1 treatment on body the adipose tissue of rats,” Endocrinology, vol. 151, no. 1, pp. composition and bone mineral density of C57BL/6 mice: 153–164, 2010. implications for cancer prevention,” In Vivo,vol.19,no.4,pp. [36] I. Talior, M. Yarkoni, N. Bashan, and H. Eldar-Finkelman, 667–674, 2005. “Increased glucose uptake promotes oxidative stress and PKC- [21] D. W. Huang, B. T. Sherman, and R. A. Lempicki, “Systematic δ activation in adipocytes of obese, insulin-resistant mice,” and integrative analysis of large gene lists using DAVID American Journal of Physiology, vol. 285, no. 2, pp. E295–E302, bioinformatics resources,” Nature Protocols,vol.4,no.1,pp. 2003. 44–57, 2009. [37] E. Karnieli and M. Armoni, “Transcriptional regulation of [22] M. E.` Piche,´ S. J. Weisnagel, L. Corneau, A. Nadeau, J. Berg- the insulin-responsive glucose transporter GLUT4 gene: from eron, and S. Lemieux, “Contribution of abdominal visceral physiology to pathology,” American Journal of Physiology, vol. obesity and insulin resistance to the cardiovascular risk profile 295, no. 1, pp. E38–E45, 2008. of postmenopausal women,” Diabetes, vol. 54, no. 3, pp. 770– [38] E. Carvalho, K. Kotani, O. D. Peroni, and B. B. Kahn, “Adipose- 777, 2005. specific overexpression of GLUT4 reverses insulin resistance [23] M. Matsubara, S. Maruoka, and S. Katayose, “Inverse relation- and diabetes in mice lacking GLUT4 selectively in muscle,” ship between plasma adiponectin and leptin concentrations American Journal of Physiology, vol. 289, no. 4, pp. E551–E561, in normal-weight and obese women,” European Journal of 2005. Endocrinology, vol. 147, no. 2, pp. 173–180, 2002. [39] E. D. Abel, O. Peroni, J. K. Kim et al., “Adipose-selective [24] J. M. Bruun, J. W. Helge, B. Richelsen, and B. Stallknecht, “Diet targeting of the GLUT4 gene impairs insulin action in muscle and exercise reduce low-grade inflammation and macrophage and liver,” Nature, vol. 409, no. 6821, pp. 729–733, 2001. infiltration in adipose tissue but not in skeletal muscle in [40] C. N. Lumeng, S. M. Deyoung, and A. R. Saltiel, “Macrophages severely obese subjects,” American Journal of Physiology, vol. block insulin action in adipocytes by altering expression of 290, no. 5, pp. E961–E967, 2006. signaling and glucose transport proteins,” American Journal of [25] K. E. Wellen, G. Hatzivassiliou, U. M. Sachdeva, T. V. Bui, J. R. Physiology, vol. 292, no. 1, pp. E166–E174, 2007. Cross, and C. B. Thompson, “ATP-citrate lyase links cellular [41] M. Furuta, Y. Yano, E. C. Gabazza et al., “Troglitazone metabolism to histone acetylation,” Science, vol. 324, no. 5930, improves GLUT4 expression in adipose tissue in an animal pp. 1076–1080, 2009. model of obese type 2 diabetes mellitus,” Diabetes Research and [26] K. Tateishi, Y. Okada, E. M. Kallin, and Y. Zhang, “Role of Clinical Practice, vol. 56, no. 3, pp. 159–171, 2002. Jhdm2a in regulating metabolic gene expression and obesity [42] E. Tozzo, L. Gnudi, and B. B. Kahn, “Amelioration of insulin resistance,” Nature, vol. 458, no. 7239, pp. 757–761, 2009. resistance in streptozotocin diabetic mice by transgenic over- [27] L. L. Frank, B. E. Sorensen, Y. Yasui et al., “Effects of exercise expression of GLUT4 driven by an adipose-specific promoter,” on metabolic risk variables in overweight postmenopausal Endocrinology, vol. 138, no. 4, pp. 1604–1611, 1997. women: a randomized clinical trial,” Obesity Research, vol. 13, [43] S. L. McGee, B. J. W. Van Denderen, K. F. Howlett et al., no. 3, pp. 615–625, 2005. “AMP-activated protein kinase regulates GLUT4 transcription [28] V. J. Vieira, R. J. Valentine, K. R. Wilund, N. Antao, T. Baynard, by phosphorylating histone deacetylase 5,” Diabetes, vol. 57, andJ.A.Woods,“Effects of exercise and low-fat diet on no. 4, pp. 860–867, 2008. adipose tissue inflammation and metabolic complications in [44] N. Raychaudhuri, S. Raychaudhuri, M. Thamotharan, and obese mice,” American Journal of Physiology, vol. 296, no. 5, S. U. Devaskar, “Histone code modifications repress glucose pp. E1164–E1171, 2009. transporter 4 expression in the intrauterine growth-restricted [29] Z. A. Zarins, M. L. Johnson, N. Faghihnia et al., “Training offspring,” Journal of Biological Chemistry, vol. 283, no. 20, pp. improves the response in glucose flux to exercise in post- 13611–13626, 2008. menopausal women,” Journal of Applied Physiology, vol. 107, [45] B. Emanuelli, P. Peraldi, C. Filloux et al., “SOCS-3 inhibits no. 1, pp. 90–97, 2009. insulin signaling and is up-regulated in response to tumor [30] W. R. Swindell, “Genes and gene expression modules asso- necrosis factor-α in the adipose tissue of obese mice,” Journal ciated with caloric restriction and aging in the laboratory of Biological Chemistry, vol. 276, no. 51, pp. 47944–47949, mouse,” BMC Genomics, vol. 10, article 585, 2009. 2001. [31] M. Padovani, J. A. Lavigne, G. V. R. Chandramouli et [46] J. E. Friedman, T. Ishizuka, S. Liu et al., “Reduced insulin al., “Distinct effects of calorie restriction and exercise on receptor signaling in the obese spontaneously hypertensive mammary gland gene expression in C57BL/6 mice,” Cancer Koletsky rat,” American Journal of Physiology, vol. 273, no. 5, Prevention Research, vol. 2, no. 12, pp. 1076–1087, 2009. pp. E1014–E1023, 1997. [32]J.Lu,L.Xie,J.Sylvesteretal.,“Different gene expression of [47] J. Shao, H. Yamashita, L. Qiao, and J. E. Friedman, “Decreased skin tissues between mice with weight controlled by either Akt kinase activity and insulin resistance C57BL/KsJ- calorie restriction or physical exercise,” Experimental Biology Lepr(db/db) mice,” Journal of Endocrinology, vol. 167, no. 1, and Medicine, vol. 232, no. 4, pp. 473–480, 2007. pp. 107–115, 2000. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 587495, 16 pages doi:10.1155/2011/587495

Review Article Microarray Evidences the Role of Pathologic Adipose Tissue in Insulin Resistance and Their Clinical Implications

Sandeep Kumar Mathur,1 Priyanka Jain,2 and Prashant Mathur3

1 Department of Endocrinology, S. M. S. Medical College, India 2 Institute of Genomics and Integrative Biology, Mall Road, New Delhi 110007, India 3 Department of Pharmacology, S. M. S. Medical College, J. L. Marg, Jaipur 302004, India

Correspondence should be addressed to Sandeep Kumar Mathur, drsandeepmathur@rediﬀmail.com

Received 1 December 2010; Accepted 21 February 2011

Academic Editor: Eliot Brinton

Copyright © 2011 Sandeep Kumar Mathur et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Clustering of insulin resistance and dysmetabolism with obesity is attributed to pathologic adipose tissue. The morphologic hallmarks of this pathology are adipocye hypertrophy and heightened inflammation. However, it’s underlying molecular mechanisms remains unknown. Study of gene function in metabolically active tissues like adipose tissue, skeletal muscle and liver is a promising strategy. Microarray is a powerful technique of assessment of gene function by measuring transcription of large number of genes in an array. This technique has several potential applications in understanding pathologic adipose tissue. They are: (1) transcriptomic differences between various depots of adipose tissue, adipose tissue from obese versus lean individuals, high insulin resistant versus low insulin resistance, brown versus white adipose tissue, (2) transcriptomic profiles of various stages of adipogenesis, (3) effect of diet, cytokines, adipokines, hormones, environmental toxins and drugs on transcriptomic profiles, (4) influence of adipokines on transcriptomic profiles in skeletal muscle, hepatocyte, adipose tissue etc., and (5) genetics of gene expression. The microarray evidences of molecular basis of obesity and insulin resistance are presented here. Despite the limitations, microarray has potential clinical applications in finding new molecular targets for treatment of insulin resistance and classification of adipose tissue based on future risk of insulin resistance syndrome.

1. Introduction Recently, the concept of “adiposopathy” has been, suggested to define this cluster of obesity, inflammation and The term insulin resistance means impairment in insulin metabolic/vascular complications as a formal disease [14– action at its target organs [1]. This biochemical defect is 16]. The basic assumption is that occurrence of diabetes commonly seen in subjects with visceral obesity, and it and other metabolic problems are related to both quantity clusters with modern life diseases like type-2 diabetes mel- and quality of fat and this fat has unique cellular and litus, hypertension, dylipidemia and atherosclerotic vascular molecular pathology. Its precise etiology is unknown but disease [2–4]. This association was initially demonstrated expected to have a genetic component and is complex, in clinical studies, and subsequently epidemiological studies that is, contributed by both hereditary and environmental done across the globe have generated convincing evidences factors [17, 18]. Some racial groups like Pima Indians and to support this association [5–8]. Increasing adipose tissue Asian Indians have strong genetic factor in their increased mass was found to be associated with altered fat metabolism susceptibility to develop diabetes [19–22]. Identification of and inflammation, thus contributing to diabetes and vascular this genetic component is pursued intensively and is on disease. This association is recognized clinically as metabolic lines of strategy recommended for any complex trait [23– syndrome. However, questions have been raised on existence 26]. The cornerstone of this complex genetic trait dissection of metabolic syndrome as a distinct disease because its is genomewide association of clinical phenotypes as well as pathophysiology is still not well defined [9–13]. gene function [27]. Microarray is a powerful technique of 2 Journal of Obesity assessment of expression of genes in a tissue at any time at (microRNA microarray). Presence of a large number of genomewide scale and has been applied in understanding spots on one slide (which could be of even genomewide role of adipose tissue in insulin resistance [28–31]. The scale for DNA chip) and their ability to hybridize with its clinical implications of these microarray evidences and the complementary sequence in the target sample with high insight gained into the understanding adiposopathy are precision gives this technology the power of genomewide presented here. study in single experiment. Because of these advantages this technology is presently the most frequently used tech- 2. Overview of the Concept of nique for the study complex molecular functions in a cell [68]. Pathologic Adipose Tissue and Its Role in Because DNA microarrays simultaneously measure ex- Insulin Resistance pression of thousands of genes in a given cell at a given point, therefore information gained from these studies can generate The hallmark pathologic feature of adiposopathy is adipocyte working hypothesis for molecular pathways essential to a hypertrophy [32–34] and visceral deposition [35–42]. The given biologic process. These transcriptomic profiles depend other features are adipose tissue macrophage [43, 44], on the cell type, genotype, and environment of the cell. possibly lymphocytic infiltration [45], poor angiogenesis Therefore it can be applied for the study of following [46, 47], intercellular matrix defects [48–50]and“crosstalk” aspects of adiposopathy [69]. First, to find the transcriptomic that is, fat deposition in other organs like liver, muscle, differences between various depots of adipose tissue, adipose pancreatic beta cells, and so forth, [51–53]. Adipose tissue tissue from obese versus lean individuals, high insulin is not only the largest endocrine organ of the body but has resistance versus low insulin resistance, and brown versus unique ability to adapt its size in response to calorie excess. white adipose tissue. Second, to see transcriptomic profiles The initial compensation is adipocyte hypertrophy but with of various stages of adipogenesis, that is, transcription continued fat deposition there is recruitment of committed control of adipogenesis. Third, to see effect of factors like mesothelial stem cell progenitors of preadipocytes [54]. diet, cytokines, adipokines, hormones, environmental toxins Their proliferation leads to formation of small adipocytes and drugs on transcriptomic profiles. Fourth, to see the which are metabolically efficient in storing excess fat. This influence of adipokines on transcriptomic profiles in skeletal adipocyte proliferation is called adipogenesis. Impairment muscle, hepatocyte, adipose tissue, and so forth. Fifth, to in adipogenesis leads to further adipocyte hypertrophy. find genetics of gene expression, that is, integration of These hypertrophied cells have imbalance in lipid synthesis gene expression phenotypes with genomewide association. (lipogenesis) and breakdown (lipolysis) and overexpress free Sixth, the assessment of adipose tissue transcription control fattyacidflux.ThisheightenedFFAfluxplayanimportant by study of genomewide gene methylation studies and role in insulin resistance and other metabolic defects of microRNA by microarray. obesity [55, 56]. The other potential applications of microarray are: study Increasing adipose tissue mass has also been found to of gene sequence variation by typing of SNPs and copy be associated with macrophage infiltration [57]. It is seen number variation by using SNP arrays. in both genetic and nutritional animal models of obesity as well as in human obesity. These macrophages secrete a large number of cytokines like TNF-α,IL-1,IL-6,andso forth, [58, 59]. These infiltrating macrophages are of two 4. Obesity and Inflammation: types M1and M2. The phenotypic switch from M2 to M1 The Microarray Evidences is possibly an important determinant of insulin resistance Several human and animal model studies have compared [60]. Recently CD11c+ macrophages were found to be the adipose tissue gene expression profiles of obese and lean marker of insulin resistance in human obesity [61]. This individuals [31, 70–75]. A large number of genes repre- inflammatory state of adipose tissue plays an important role senting various molecular functions and pathways were in insulin resistance [62, 63]. Adipocyte hypertrophy itself found to be differentially expressed in adipose tissue of was also found to be associated macrophage infiltration and obese subject. Most of these genes were found to be in the heightened cytokine/adipokine secretion and inflammation molecular pathways of adipose tissue inflammation, lipid [64, 65]. and carbohydrate metabolism, adipogenesis and cytoskeletal structure regulation and so forth. However, the most of the 3. Relevance and Power of Microarray: genes identified in any studies could not be replicated in The Evidences Required to Support Concept subsequent studies. It could be because of several reasons of Pathologic Adipose Tissue like genetic differences in study population, disease heterogeneity, and nonuniform physiologic condition at time of The term microarray represents an array of large number sampling and method of sampling. But it is important to of microscopic spots on a solid surface [66, 67]. The solid note that gene expression profiling was found to be altered surface could be a glass slide or nylon membrane or quartz with obesity in almost all studies. This suggests that increase wafer. These spots could be of complementary cDNA (cDNA in adipose tissue mass is not merely a quantitative change, microarray), oligonucleotide probes (high density oligonu- but it also alters its quality, therefore supports the concept of cleotide microarray), protein (protein array), or microRNA adiposopathy. Journal of Obesity 3

Increasing adipose tissue mass was found to be asso- from adipocytes incubated with conditioned medium, with ciated with upregulation of inflammation-related genes in little release by control adipocytes. Treatment with TNF- almost all studies [76]. When the stromal-vascular frac- alpha induced substantial increases in MMP1 (>100-fold) tion of adipose tissue was separated from adipocytes, the and MMP3 (27-fold) mRNA level as well as MMP1 and inflammation-related genes were upregulated in stromal MMP3 release in adipocytes, suggesting that this cytokine fraction [77, 78]. These findings suggest that the inflam- could contribute to the stimulation of MMP expression by matory state of adipose tissue is predominantly because macrophages, thus contributing to tissue remodeling during of infiltrating macrophages. These macrophages secrete a adipose tissue expansion in obesity. large number of inflammatory cytokines like TNF-α,IL-6, Though there are convincing evidences for the role of IL-1, and so forth, [58, 59]. In fact microarray evidences macrophages in adipose tissue inflammation, the precise also support the histology findings of increased macrophage cause of their infiltration in theadiposetissueisstillnot infiltration in adipose tissue in obesity. However it is worth known. However several hypotheses have been suggested. mentioning here that adipocytes, when cultured isolated They are as follow (1) increasing adipose tissue mass not even then showed upregulation of inflammation-related accompanied by increase in its vascularity leads to local genes. Though the relative contribution of adipocytes in hypoxia, which is the cause of macrophage infiltration [89– overall systemic inflammation of obesity is much less as 92], (2) obesity is associated with increased expression of compared with macrophages, but their precise role in chemokines like monocyte chemoattractant protein (MCP)- initiation of inflammatory seems to be very important. 1, MCP-3, macrophage inflammatory protein (MIP)-1α,and The current evidences suggest that adipocyte hypertrophy they play an important role in macrophage infiltration [93– is associated with increased release of TNF-α,aninflam- 97] (3) increasing adiposity is associated with elevated leptin matory state [79]. These cytokines plays an important levels and later increases adhesion molecule expression in role in lymphocyte and macrophage infiltration in the endothelial cells and monocyte diapedesis [98, 99], and (4) as adipose tissue and further augmentation of inflammation by adipocytes share several characters with immune cells, it has heightened cytokine release. A vicious cycle is generated as also been suggested that preadipocytes are transdifferentiated adipokines released from infiltrating macrophages impairs into macrophages [98]. adipogenesis, and augments free fatty acid/cytokine release There is also possible role of epigenetic factors in adipose from hypertrophied adipocytes. Diabetics in one study were tissue inflammation. Epigenetic modifications are chemical found to have not only much higher FFA level and insulin additions to DNA and histones that are associated with resistance than BMI-matched controls, but also showed changes in gene expression and are heritable but do not alter positive correlation between them [80]. The vicious cycle the primary DNA sequence. DNA methylation is one of the of augmented inflammation in adipose tissue could be the epigenetic modification and in mammals may be associated underlying mechanism of this finding. with gene silencing. The role of epigenetic factors in inflam- Some recent studies examined effect of inflammatory mation of obesity was suggested in transgenic mice overex- factors on adipocytes. Shah et al. [81] applied microarray pressing DNA methyltransferase (Dnmt3a). Gene expression mRNA profiling of human adipose during endotoxemia to levels of inflammatory cytokines such as tumor necrosis identify novel inflammation-induced adipose tissue genes. factor-alpha (TNF-alpha) and monocyte chemoattractant Several genes implicated inflammation, but not known to be protein-1 (MCP-1) were higher in Dnmt3a mice than in expressed in adipose tissue were upregulated in response to wild-type mice on a high-fat diet. This study suggests that induced inflammation. These findings suggest that induced increased expression of Dnmt3a in the adipose tissue may inflammation further augments inflammation of obesity. contribute to obesity-related inflammation. However there Obesity was found to be associated with recruitment of are no human studies to support these findings [100]. lymphocytes [82–84] and increased expression of inflammation-related genes from M1 macrophages [85–87]. These infiltrating cells play an important role in inflammation of 5. Adipogenesis Defects in Pathologic Adipose obesity. Tissue: The Role of Transcriptional Factors Apart from inducing inflammation, infiltrating macrophages also influence remodeling of adipose tissue. O’Hara The pathologic hallmark of adiposopathy is adipocyte hyper- et al. [88] treated human adipocytes in culture with macro- trophy and has been hypothesized to be due to defects in phage conditioned medium. Microarray analysis identified proliferation and differentiation of mesenchymal stem cells 1,088 genes differentially expressed in response to the and preadipocytes into mature adipocytes, a process called conditioned medium at both 4 and 24 h (754 upregulated, adipogenesis [101]. Adipogenesis is a complicated process 334 downregulated at 24 h); these included genes associated and had been extensively studied. Much progress has been with inflammation and macrophage infiltration. A cluster made in the last two decades in defining transcriptional of matrix metalloproteinase genes were highly upregulated events controlling the differentiation of mesenchymal stem at both time points, including MMP1, MMP3, MMP9, cells into adipocytes. A complex network of transcription MMP10, MMP12 and MMP19. At 4 and 24 h, MMP1 factors and cell-cycle regulators, in concert with specific was the most highly upregulated gene (>2,400-fold increase transcriptional coactivators and corepressors, respond to in mRNA at 24 h). ELISA measurements indicated that extracellular stimuli to activate or repress adipocyte differen- substantial quantities of MMP1 and MMP3 were released tiation [102]. PPAR-γ and C/EBP-α are the major regulatory 4 Journal of Obesity factors. Precise understanding of this process is expected of PPAR-γ expression, but it is not required for adipocyte to open avenues for discovery of new molecular targets for differentiation [135]. A nuclear receptor gene expression prevention of expansion of adipose tissue in obesity and its atlas during the differentiation of 3T3-L1 cells, assessed transformation from benign to pathologic tissue found in using qPCR, also showed the importance of other nuclear high insulin resistance conditions. receptors such as the Nr2f2 (COUP-TF2) in adipogenesis Microarray is one of the most powerful techniques for [124]. Fourthly, the role of Ebf1 (O/E-1), a helix-loop understanding the process of adipogenesis not only because helix transcription factor, was studied in adipocytes with of its ability to study molecular function at large scale microarray analysis of Ebf1 overexpression in NIH-3T3 cells in single experiment, but also because of its precision, [115]. Further experiments helped place Ebf1 within the sensitivity and specificity. Therefore, gene expression during known transcriptional cascade of adipogenesis [136]. By various stages of adipogenesis has been extensively studied the year 2000, it was shown that Gata2 and Gata3 are and the data so generated has been subjected to meta- specifically expressed in adipocyte precursors and their down analysis. There are more than 20 published reports on gene regulation sets the stage for terminal differentiation [137]. expression profiling during various stages of adipogenesis This type of expression profile could be confirmed later [103–125]. Most of these studies used different platforms on with microarray experiments. A role for transcriptional and had inherent problems like different probes, annotations coregulators in the control of energy homeostasis could be and probe sequences. Apart from the heterogeneity of shown by knockout of the corepressor Nrip1 (RIP140) in platforms these studies also had heterogeneity in terms of adipocytes [117]. model studied: in vitro murine cell lines in culture, animal Microarray studies have also identified a number of models and human mesenchymal stem cell/preadipocyte in enzymes regulating adipogenesis. For example, Xanthine vitro culture models. However, an exhaustive analysis and dehydrogenase (Xdh, XOR), Stearoyl-CoA desaturase (Scd1), comparison of commonly used microarray platforms by adipose triglyceride lipase Pnpla2 (ATGL), and so forth. a multicenter consortium (MAQC) showed—contrary to Apart from identification of various transcription factors and earlier reports [126, 127]—acceptable concordance between genes showing differential expression during adipogenesis, the platforms [128]; however, there is a necessity for careful the other applications of microarray in understanding this control of biological samples and close adherence to standard process are: (1) understanding the regulation of expression protocols. of various genes, (2) influence of perturbation of these genes A large number of genes identified by these microarray on adipocyte physiology, and (3) effects of various drugs, studies were found to be associated with adipocyte develop- nutrients and so forth. ment. Some of them were further studied from point of view Methylation of cytosine and formation of 5-methyl- of understanding molecular mechanisms of adipogenesis. cytosine are the only covalent DNA modification known Few of them are as follows: firstly, several members of in vertebrates [138]. This epigenetic modification regulates the Kruppel-like¨ factor family have been implicated in gene expression. Horri et al. recently studied role of DNA adipogenesis [129]. In microarray studies, Klf4 was shown methylation in control of insulin-induced adipogenesis in to function as an immediate early regulator of adipogenesis 3T3-L1 preadipocytes using a method called microarray- by inducing C/EBP-β and is required for adipogenesis based integrated analysis of methylation by isoschizomers [129]. Klf6, Klf15 and Klf5 have been found to play a role (MIAMI). The MIAMI revealed that Hpa II sites of exon 1 in in adipogenesis. Klf5 acts by transactivating PPAR-γ.The a Rho guanine nucleotide exchange factor 19 (ARHGEF19; available evidences suggest that these factors function by WGEF) gene were demethylated during adipocyte differenti- recruiting different coactivators or repressors [102]. Sec- ation of 3T3-L1 cells. This study suggests that adipogenesis ondly, microarray studies identified differential expression was regulated by WGEF expression through DNA methyla- of zinc finger containing transcription factor Egr2 (Krox20) tion change [139]. However, Ocada et al have shown that the during adipogenesis. The expression of Egr2 is activated diet-induced upregulation of leptin, Mest/Peg1, and sFRP5 very early after induction of adipogenesis. It stimulates gene expression in WAT during the development of obesity in adipogenesis at least in part through activating C/EBP-β by mice is not mediated directly by changes in DNA methylation binding to its promoter [130]. Thirdly, expression profiles [140]. of adipogenesis have identified orphan nuclear receptor MicroRNAs (miRNAs) are short noncoding RNA that Nr4a1 (Nur77) as an early gene. But, there are conflicting posttranscriptionally regulates gene expression. Some miR- reports on effects of its altered expression in adipogenesis NAs have been proposed to be associated with obesity. [131–133]. Another nuclear hormone receptor involved in Nakanishi et al. [141] found the upregulation of miR-335 in adipogenesis is Nr1h3 (LXRalpha). A broader role of Nr1h3 obesity using microarray analysis for miRNA. Furthermore, in regulation of metabolism in adipocytes was suggested miR-335 levels were closely correlated with expression levels and the effects of Wnt-signaling in adipocyte differentiation of adipocyte differentiation markers such as PPARgamma, were studied in timed series microarray experiments of 3T3- aP2, and FAS in 3T3-L1 adipocyte. These findings provide L1 cells and retroviral infected 3T3-L1 cells encoding Wnt1 the first evidence that the upregulated expressions of miR- [134]. It is known that liver X receptors (LXRs) regulate 335 in liver and WAT of obese mice might contribute to the cholesterol and fatty acid metabolism in liver tissue and pathophysiology of obesity. Sun et al. [142]alsoinvestigated macrophages. Recently it was also shown that activated whether micro RNA plays a role in adipogenesis. They Nr1h3 stimulate adipocyte differentiation through induction performed microarray analysis to detect miRNA expression Journal of Obesity 5 during 3T3-L1 preadipocyte differentiation. Several miR- (IL)-8, serum amyloid a (SAA), C1q receptor 1, and CXCL2 NAs, including let-7, were upregulated during 3T3-L1 adipo- also known as MIP-2 or macrophage inflammatory protein- genesis. Ectopic introduction of let-7 into 3T3-L1 cells inhib- 2. SAA is secreted by adipose tissue and is acute phase protein ited clonal expansion as well as terminal differentiation. The implicated in insulin resistance, adipose tissue macrophage mRNA encoding high-mobility group AT-hook 2 (HMGA2), infiltration, and cholesterol reverse transport [146–148]. a transcription factor that regulates growth and proliferation Therefore, SSA and other proteins could be important in other contexts, was inversely correlated with let-7 levels mediators of hypertrophied adipocyte in insulin resistance, during 3T3-L1 cell adipogenesis, and let-7 markedly reduced inflammation, and diabetes. TM4SF1 gene was also upregu- HMGA2 concentrations. Knockdown of HMGA2 inhibited lated in hypertrophied adipocytes. The precise function of 3T3-L1 differentiation. These results suggest that let-7 plays this protein is not known, but speculated to be related to an important role in adipocyte differentiation and that it inflammation [149]. This gene was also over-expressed in does so in part by targeting HMGA2, thereby regulating the visceral adipose tissue as well as among obese individuals transition from clonal expansion to terminal differentiation. [150]. It was mapped to the chromosomal region which was also found to be linked with BMI and fasting insulin level among Pima Indians [151]. Therefore these transcriptomic 6. Mechanism of Inflammation and High FFA findings support the concept that hypertrophied adipocyte Flux from Hypertrophied Adipocytes itself induces inflammatory changes in adipose tissue. Hypertrophied adipocytes also releases large amount of Adipose tissue is the major energy reservoir of the body. The saturated fatty acids. These saturated free fatty acids act as excess calories resulting from energy imbalance are predom- natural ligands for toll-like receptor 4 and activate inflam- inantly stored as triglycerides. About 95% of adipocyte mass mation in both adipocytes and infiltrating macrophages consists of triglycerides. Fat content of an adipocyte (in other via nuclear factor kappa B (NF-κB) [152]. In a recent words size of an adipocyte) depends on balance between study cDNA microarray analysis of saturated fatty-acid- fat deposition (by either nutrient transport or fat synthesis stimulated macrophages in vitro and obese adipose tissue by lipogenesis) and fat depletion (by either utilization in in vivo was done by the same group of investigators. They lipolysis or export out of cell). However this single cell fat find activating transcription factor (ATF) 3 (a member of the balancemustbedifferentiated from overall fat balance of ATF/cAMP response element-binding protein family of basic the body, which in addition also depends on recruitment leucine zipper-type transcription factors) as a target gene of of newer adipocytes by adipogenesis. When the process of saturated fatty acids/TLR4 signaling in macrophages in obese adipogenesis is insufficient size of an individual fat cell adipose tissue. Importantly, ATF3, when induced by satu- increases to accommodate excess calories deposited as lipids. rated fatty acids, can transcriptionally repress tumor necrosis The lipid deposition in an adipocyte is regulated by the factor-α production in macrophages in vitro. Overexpression nutrients, hormones and perilipin family of lipid droplet of ATF3 specifically in macrophages results in the marked coat proteins [143]. The process of lipogenesis is stimulated attenuation of proinflammatory M1 macrophage activation by excess of carbohydrates and inhibited by polyunsaturated in the adipose tissue. Therefore it is hypothesized that it fatty acids and fasting. It is hormonally regulated, stimulated could be a negative feedback mechanism that attenuates by insulin and inhibited by growth hormone and leptin. obesity-induced macrophage activation. Hence activation of SREB-1 and PPAR-γ are the major transcriptional regulators ATF3 in macrophages offers a novel therapeutic strategy to of this process and mediate the action of nutrients and prevent or treat obesity-inducedadiposetissueinflammation hormones [144]. The process of lipolysis is stimulated by [153]. fasting, catecholamines, and atrial natriuretic peptide and Some of the recent microarray studies have examined inhibited by insulin. It is mediated by hormone sensitive effects of some chemicals on process of lipogenesis and lipol- lipase (HSL) and adipose triglyceride lipase (ATGL). ysis. Dietary flavonoid phloretin enhances 3T3-L1 adipocyte The mechanism of adipose tissue lipid balance has been differentiation and adiponectin expression at least in part extensively studied from point of view of understanding through PPAR-γ activation [154]. In a recent microarray molecular basis of obesity, inflammation and high free study it was shown that phloretin positively regulates the fatty acid flux. Microarray has potential application of expression of numerous genes involved in lipogenesis and identifying molecular pathways underlying physiology of triglyceride storage, including GLUT4, ACSL1, PEPCK1, lipid balance of an adipocyte in health and disease. It is lipin-1 and perilipin (more than twofold). The expres- also of help in understanding effects of various nutrients, sion of several genes encoding adipokines, in addition to chemical, experimental drugs on fat storage and lipolysis by adiponectin and its receptor, was positively or negatively an adipocyte. regulated in a way that suggests a possible reduction in Gene expression profiling using microarray has been systemic insulin resistance and obesity-associated inflamma- applied in understanding differences between small and tion. Improvement of insulin sensitivity was also suggested hypertrophied adipocytes. Jernas et al. [145]comparedgene by the overexpression of genes associated with insulin expression profiles of hypertrophied adipocytes with smaller signal transduction, such as CAP, PDK1 and Akt2. Many adipocytes obtained from the same individual. 14 genes of these genes are PPARgamma targets, confirming the showed 4-fold upregulation in hypertrophied adipocytes. involvement of PPARgamma pathway in the phloretin effects Five of them were immune related: E-selectin, interleukin on adipocytes [155]. 6 Journal of Obesity

Docosahexaenoic acid (DHA) increases lipolysis and ubiquitin also were underexpressed in NASH. Genes that decreases lipogenesis. Wang et al. [156] treated human breast were overexpressed in NASH included complement compo- adipocytes with hSAA1 (recombinant human serum amyloid nent C3 and hepatocyte-derived fibrinogen-related protein, Aprotein)orDHAin vitro. They find that expression of potentially contributing to impaired insulin sensitivity. In peroxisome proliferator-activated receptor gamma and other conclusion, these studies provide evidence for a transcrip- lipogenic genes was decreased, whereas the expression of tional or pretranscriptional basis for impaired mitochondrial several lipolytic genes was increased. Glycerol release was function (attenuated capacity for the dismutation of reactive increased by both SAA and DHA treatments, suggesting that oxygen species) and diminished insulin sensitivity (increased they increased lipolytic activity in human adipocytes. The acute phase reactants) in patients with histologically progres- expression of perilipin, a lipid droplet-protective protein, sive NASH. was decreased, and hormone-sensitive lipase was increased Cheung et al. [168] studied differentially expressed micro by both of hSAA1 and DHA treatment. They speculated RNAs, their target in NASH and impact of one specific that the mechanism of lipolysis by DHA or SAA is at least differentially expressed microRNA, miR122. Total of 23 partially the result of increased expression of hormone- microRNAs were underexpressed or overexpressed. The sensitive lipase and decreased expression of perilipin. The predicted targets of these microRNAs are known to affect cell finding of this study also suggested that as DHA treatment proliferation, protein translation, apoptosis, inflammation, increased expression of hSAA1 in human adipocytes, the oxidative stress, and metabolism. Underexpression of miR- DHA-mediated reduction in expression of lipogenesis genes 122 potentially contributes to altered lipid metabolism and enhancement of lipolysis may be through the activity of implicated in the pathogenesis of NASH. hSAA1. Type-2 diabetes is associated with muscle insulin resistance and first-degree relatives of them also have high insulin resistance. There is increasing evidence for a link between 7. Skeletal Muscle and Hepatocyte insulin resistance and impaired mitochondrial oxidative Transcriptome in Insulin-Resistant States: phosphorylation (OXPHOS) in human skeletal muscle in Role of Cytokines and FFA vivo [169–173]. Several microarray-based studies of skeletal muscle have reported coordinated downregulation of Obesity and insulin resistance is associated with fatty liver OXPHOS genes (mitochondrial biogenesis) in patients with disease manifested clinically as increased triglyceride deposi- type 2 diabetes and high-risk individuals [174–177], that tion in liver (steatosis) and progressive necroinflammatory reduced expression of the genes encoding transcriptional liver disease (steatohepatitis, NASH) [157, 158]. Increased coactivator peroxisome proliferator-activated receptor γ FFA flux from adipose tissue in insulin-resistant state is asso- coactivator 1α (PGC-1α) and nuclear respiratory factor 1 ciated with hepatic fat deposition, impairment in mitochon- (NRF1) could play a key role in these transcriptional changes drial energy biogenesis and insulin resistance. It was recently [175–177]. suggested that rather than being an “innocent bystander”, However, it is an important question one need to answer liver steatosis may be the “guilty party” of the progression is whether, skeletal muscle insulin resistance is a primary to cirrhosis [159–161] and possibly in the development of genetic defect or secondary to high FFA flux and cytokines hepatocellular carcinoma (HCC) [162]. This can result in secreted from visceral adipose tissue. Frederiksen et al. end-stage liver disease requiring liver transplantation [163– [178] generated transcriptional profiling of myotubes from 165]. Therefore, understanding the molecular mechanisms patients with type 2 diabetes mellitus and obese control sub- underlying FLD is of major importance. Microarray has jects. They find that no single gene was differently expressed potential application in understanding molecular basis of after correction for multiple testing, and no biological fatty liver disease. pathway was differentially expressed using global pathway Chiappini et al. [166] studied global gene expression analysis. In particular, they found no evidence for differential in human liver steatosis. They find that mitochondrial expression of genes involved in mitochondrial oxidative alterations play a major role in the development of steatosis. metabolism. Consistently, there was no difference in mRNA Activation of inflammatory pathways is present at a very levels of genes known to mediate the transcriptional control early stage of steatosis, even if no morphological sign of of mitochondrial biogenesis (PPARGC1A and NRF1) or in inflammation is observed. mitochondrial mass between diabetic and control myotubes. Sreekumar et al. [167] compared genomewide transcrip- Hence the results of this study support the hypothesis tion profiles in liver biopsies from cirrhotic stage of NASH that impaired mitochondrial biogenesis is not a primary with cirrhosis caused by hepatitis C and healthy controls. defect in the sequence of events leading to insulin resistance Sixteen genes were uniquely differentially expressed (4 over- and type 2 diabetes. The other meaning of this study expressed and 12 underexpressed) in patients with cirrhotic- is that impaired mitochondrial oxidative phosphorylation stage NASH. Genes that were significantly underexpressed (OXPHOS) in skeletal muscle is possibly reversible with included genes important for maintaining mitochondrial control of hyperglycemia and dysmetabolism. Sreekumar et function (copper/zinc superoxide dismutase, aldehyde oxi- al. [174] studied gene expression profile in the skeletal muscle dase, and catalase). Glucose 6-phospatase, alcohol dehy- of patients with type 2 diabetes while not on treatment for drogenase, elongation factor-TU, methylglutaryl coenzyme 2 weeks and after 10 days of intensive insulin treatment. Of A (CoA), acyl CoA synthetase, oxoacyl CoA thiolase, and 6,451 genes surveyed, transcriptional patterns of 85 genes Journal of Obesity 7 showed alterations in the diabetic patients after withdrawal a weight loss program. The diet-induced improvement of of treatment, when compared with patterns in the nondia- insulin sensitivity is associated with changes in clusters of betic control subjects. Insulin treatment reduced the differ- genes rather than in single genes, with the sets of genes ence in patterns between diabetic and nondiabetic control being different in each phase of the program. The kinetics of subjects (improved) in all but 11 gene transcripts, which changes within adipose tissue and the interactions between included genes involved in structural and contractile func- themetabolicstateofadipocytesandtheactivationstate tions: growth and tissue development, stress response, and of macrophages appear critical for the understanding of energy metabolism. These improved transcripts included the beneficial effects on health resulting from a long-term genes involved in insulin signaling, transcription factors, dietary weight loss program in obese subjects. Clement´ et and mitochondrial maintenance. However, insulin treatment al. [183] studied changes in subcutaneous adipose tissue altered the transcription of 29 additional genes involved gene expression profiles after feeding with very low calorie in signal transduction; structural and contractile functions; diet for 28 days. They find that weight loss improves growth and tissue development; protein, fat, and energy the inflammatory profile of obese subjects through a metabolism. Therefore they identified several candidate decrease of proinflammatory factors and an increase of anti- genes playing role in muscle insulin resistance, complications inflammatory molecules. The genes are expressed mostly associated with poor glycemic control, and effects of insulin in the stromavascular fraction of adipose tissue, which is treatment in people with type 2 diabetes. shown to contain numerous macrophages. The beneficial Yang et al. [179] compared muscle transcription profiles effect of weight loss on obesity-related complications may of equally obese high- and low-insulin-resistant individuals. be associated with the modification of the inflammatory The authors concluded that 185 differentially expressed tran- profile in adipose tissue. Viguerie et al. [184] determined scripts, 20 per cent were true positives and some could gen- the levels of transcripts for 38 genes that are expressed in erate new hypotheses about the aetiology or pathophysiology adipose tissue and encode transcription factors, enzymes, of insulin-resistance. Furthermore, differentially expressed transporters and receptors known to play critical roles in the genes in chromosomal regions with linkage to diabetes and regulation of adipogenesis, mitochondrial respiration, lipid, insulin-resistance serve as new diabetes susceptible genes. and carbohydrate metabolism in two groups of 25 obese Nguyen et al. performed gene expression profiling on subjects following 10-week hypocaloric diet programmes skeletal muscle of insulin-resistant and insulin-sensitive sub- with either 20–25 or 40–45% of total energy derived from jects using microarrays. Microarray analysis of 19,000 genes fat being investigated. Levels of mRNA were measured by in skeletal muscle did not display a significant difference performing real-time RT-PCR on subcutaneous fat samples between insulin-resistant and insulin sensitive muscle. This obtained from the subjects before and after the diets. Ten was confirmed with real-time PCR. These results suggest genes were regulated by energy restriction; however, none that insulin resistance is not reflected by changes in the gene of the genes showed a significantly different response to the expression profile in skeletal muscle [180]. diets. Levels of peroxisome proliferator-activated receptor gamma coactivator 1alpha mRNA were increased, while the expression of the genes encoding leptin, osteonectin, 8. Effect of Caloric Restriction, phosphodiesterase 3B, hormone-sensitive lipase, receptor A Drugs, and Chemical Toxins on Adipose for natriuretic peptide, fatty acid translocase, lipoprotein Tissue Transcriptome lipase, uncoupling protein 2 and peroxisome proliferator- activated receptor gamma was decreased. In accordance with There are several reports in the literature studying the effect the comparable loss of fat mass produced by the two diets, of various factors such as nutrition, stress, and drugs, on gene this study shows that energy restriction and/or weight loss expression profile of adipose tissue in animal models, in vitro rather than the ratio of fat: carbohydrate in a low-energy cell lines and occasionally in human beings. diet is of importance in modifying the expression of genes in Dahlman et al. [181]investigatedtheeffect of different the human adipose tissue. Kolehmainen et al. [185]findthat low-energy diets on gene expression in human adipose tissue. weight reduction in subjects with metabolic syndrome was They find that macronutrients have a secondary role in associated with strong down regulation of genes regulating changes in adipocyte gene expression after energy-restricted extracellular matrix and cell death. diets. The most striking alteration after energy restriction is van Dijk et al. [186]investigatedtheeffect of a satu- a coordinated reduction in the expression of genes regulating rated fatty acid-(SFA-) and a monounsaturated fatty acid- the production of polyunsaturated fatty acids. (MUFA-) rich diet on insulin sensitivity, serum lipids, and Capel et al. [182] studied effect of calorie restriction, gene expression profiles of adipose tissue in subjects at weight stabilization on insulin sensitivity, and subcutaneous risk of metabolic syndrome. They find that consumption adipose tissue gene expression profile. They find that adipose of an SFA diet resulted in a proinflammatory “obesity- tissue macrophages and adipocytes show distinct patterns linked” gene expression profile, whereas consumption of of gene regulation and association with insulin sensitivity a MUFA diet caused a more anti-inflammatory profile. duringthevariousphasesofadietaryweightlossprogram. This suggests that replacement of dietary SFA with MUFA They concluded that macrophage and fat cell gene expres- could prevent adipose tissue inflammation and may reduce sions in adipose tissue are differentially regulated during the risk of inflammation-related diseases such as metabolic the calorie restriction and weight maintenance phases of syndrome. 8 Journal of Obesity

Dietary supplementation of arginine is the physiologic that the physiological processes controlled by these genes precursor of nitric oxide (NO), and its dietary supplemen- contribute to depot and gender-related differences in the tation has potential to reduce the fat mass. Fu et al. [187] metabolic complications of obesity. Dolinkovaetal.[´ 196] studied effect of arginine supplementation on fat mass and evaluated expression profile of genes potentially related to underlying molecular mechanisms in Zucker diabetic fatty metabolic complications of obesity in the whole adipose (ZDF) rat, a genetically obese animal model of type II tissue and isolated adipocytes from subcutaneous (SAT) diabetes mellitus. They find that arginine supplementation and visceral adipose tissue (VAT) from 12 nondiabetic increased adipose tissue expression of key genes respon- obese women and 12 lean women. They found increased sible for fatty acid and glucose oxidation: NO synthase- expression of specific proinflammatory and adipogenic genes 1 (145%), heme oxygenase-3 (789%), AMP-activated pro- and reduced expression of specific lipogenic and insulin tein kinase (123%), and peroxisome proliferator-activated signaling pathway genes in obese relative to lean women receptor gamma coactivator-1alpha (500%). Their findings with no preferable localization in SAT or VAT depot. The suggest that arginine treatment may provide a potentially gene expression significantly differed between adipocytes and novel and useful means to enhance NO synthesis and reduce adipose tissue but both contributed to the proinflammatory fat mass in obese subjects with type II diabetes mellitus. profile in obesity. They conclude that both SAT and VAT Arginine treatment may provide a potentially novel and exhibit alterations in the expression of specific genes possibly useful means to enhance NO synthesis and reduce fat mass contributing to proinflammatory and insulin resistance state in obese subjects with type-II diabetes mellitus. Another and consequently to metabolic complications of obesity. studybyJobgenetal.[188] also showed that high fat diet and arginine supplementation differentially regulate gene expression to affect energy-substrate oxidation, redox state, 10. Genetics of Gene Expression: fat accretion, and adipocyte differentiation in adipose tissue. Integration of Gene Expression, Genotype, Their findings provide a molecular mechanism to explain and Clinical Phenotype a beneficial effect of arginine on ameliorating diet-induced obesity in mammals. Currently there are two major genomic strategies for iden- Molecular mechanisms underlying metabolic effects of tification of genes underlying complex diseases. The first nutritional factors like calorie, fat, arginine, vitamin C, strategy is finding the gene sequence variants associated with grape anthocyanins and so forth, were understood in human clinical phenotype [197]. Genomewide association studies beings and animal models. There are isolated reports about using genomewide SNP markers with diabetes and related the effects of several factors like stress, dihydrotestosterone, phenotypes have been quite successful in identification of lipocalin 2, rosiglitazone, licorice flavonoids, Turmeric oleo- genomic regions having diabetes susceptibility genes [198]. resin, coca, topiramate, leptin, β-adrenoceptor agonists, and However, limitations of these studies do not provide any so forth, in animal models, cell lines and occasionally in information about function of these genes and the mecha- humans [189–193]. nisms involved in pathogenesis of disease. The other strategy is identification of network of coexpressed genes perturbed in the insulin-resistant state by gene expression profiling. 9. Molecular Basis of Differences However, gene expression profiling does not provide any between Visceral and Subcutaneous Fat: information on drivers of disease, and causal information. Association studies of gene expression phenotypes identify Transcriptional Evidences gene sequence variants regulating gene function and when Association of visceral obesity with increased expression combined with clinical information could infer causal asso- of cytokines/adipokines and hormones mediating insulin ciation between expression and disease trait [199]. Identifica- resistance suggest that these two depots of fat are not tion of gene networks that are perturbed by susceptibility loci mere different locations of same tissue, rather biologically for a given clinical trait could refine definition of the trait, different organs with different molecular functions. It is also identify disease subtypes, construct gene networks associated possible that they differ in their origin during embryonic life with the disease and identify key drivers of disease [200–205]. [194]. Microarray is an important and appropriate tool for Data on this type of investigations in human adiposopathy is studying differences in gene function between visceral and lacking. However, in segregating mouse population resulted subcutaneous adipose tissue. Several studies have compared in the identification of a macrophage-enriched network transcriptomes of these two tissues. Ramis et al. [195] supported as having a causal relationship with disease traits compared transcriptomes of visceral and subcutaneous associated with metabolic syndrome [206]. tissues and reconfirmed by RT-PCR differential expression of the two best candidates, carboxypeptidase E (CPE), and 11. Clinical Applications thrombospondin-1 (THBS1) (EST N72406). They find that both genes appeared to be strongly differentially expressed, There are several potential applications of gene expression having a higher expression in the visceral depot than in profiling in clinical practice like finding the new molecular the subcutaneous. For THBS1, the difference in expression targets of drug development based on molecular mechanisms between the depots was greater in women than in men. underlying insulin resistance, classification of insulin resis- The involvement of CPE and THBS1 in obesity suggests tance into subtypes, molecular diagnostic markers of future Journal of Obesity 9 insulin resistance in early life, and prediction of therapeutic hyperglycemia, and several drugs used to treat clinical response or adverse effects on the basis of transcription conditions like diabetes (metformin, glitazones, telmisar- profiles [207–209]. At present gene expression profiling has tan, etc.). The method of obtaining tissue also influences not percolated to clinical practice and is only a research tool. gene expression (needle versus open adipose tissue biopsy) Several molecular mechanisms identified by microarray like [214]. those underlying adipogenesis and inflammation mediated by adipose tissue macrophage infiltration have potential to 13. Conclusion serve new molecular target of drug discovery. PPAR-γ and itsagonistsarebestexampleofthisclinicalapplication. Despite several limitations microarray has proved a valuable Similarly, Integration of molecular profiling with genotype tool in understanding obesity insulin resistance relationship. and clinical traits and identification of DNA polymorphs There is need to standardize the technique and overcome which are drivers of insulin resistance associated transcrip- the problems of lack of validation and nonreplication. tion profiles has potential application in classification of Integration of expression profiling with clinical phenotypes adiposity into pathogenic and benign fat [210]. More drastic and genotyping has potential to open avenues for its clinical life-style modification and drug therapy to prevent insulin application like classification of obesity, molecular diagnosis resistance will be advised in individuals having pathologic and newer therapeutic options. adipose tissue even before they gain weight and develop insulin resistance. References

[1] S. Del Prato, R. C. Bonadonna, E. Bonora et al., “Character- 12. Limitations of Microarray ization of cellular defects of insulin action in type 2 (non- Despite power (i.e., genomewide scale) and multiple appli- insulin-dependent) diabetes mellitus,” Journal of Clinical Investigation, vol. 91, no. 2, pp. 484–494, 1993. cations in understanding molecular functions of cell (level of [2] G. M. Reaven, “Role of insulin resistance in human disease,” gene and protein expression, genomewide association with Diabetes, vol. 37, no. 12, pp. 1595–1607, 1988. SNP markers and copy number variants, RNA splicing, gene [3] G. M. Reaven, “Pathophysiology of insulin resistance in methylation, microRNA study, etc.) there are limitations human disease,” Physiological Reviews, vol. 75, no. 3, pp. 473– of microarray in understanding molecular basis of insulin 486, 1995. resistance. These limitations are both of microarray tech- [4] M. Koffler,L.C.Ramirez,andP.Raskin,“Insulinresistance nology itself and its application in understanding insulin and diabetes, mechanism and possible intervention,” Dia- resistance. betes Research and Clinical Practice, vol. 7, no. 2, pp. 83–98, In contrast to gene sequence, transcriptome depends on 1989. [5] J. Vague, “The degree of masculine differentiation of obe- cell type and its environmental conditions; therefore large ff sities: a factor determining predisposition to diabetes, amount of data is generated which is analyzed by di erent atherosclerosis, gout, and uric calculous disease,” The Ameri- methods. Also there is no standardized unit of estimation of can Journal of Clinical Nutrition, vol. 4, no. 1, pp. 20–34, 1956. gene expression; hence data is only presented as fold change. [6] E. S. Ford, “Prevalence of the metabolic syndrome defined There are no known uniform standard gene expression levels by the international diabetes federation among adults in the for every cell in the body underphysiologic conditions, so U.S,” Diabetes Care, vol. 28, no. 11, pp. 2745–2749, 2005. that true assessment of gene expression in pathophysiologic [7]R.J.Adams,S.Appleton,D.H.Wilsonetal.,“Population conditions is impossible [211]. More overdata generated comparison of two clinical approaches to the metabolic by using different platforms, arrays, and methods of nor- syndrome: implications of the new International Diabetes malization has produced inconsistent data for same type Federation consensus definition,” Diabetes Care, vol. 28, no. of experiment performed in different labs [128, 212, 213]. 11, pp. 2777–2779, 2005. [8] E. S. Ford, W. H. Giles, and A. H. Mokdad, “Increasing Most of the genes prioritized to be differentially expressed prevalence of the metabolic syndrome among U.S. adults,” in adipose tissue, muscle tissue, and so forth, in any study Diabetes Care, vol. 27, no. 10, pp. 2444–2449, 2004. were not replicated in subsequent study. Lastly molecular [9] G. Reaven, “The metabolic syndrome or the insulin resistance functions identified by transcriptomic profile could not be syndrome? Different names, different concepts, and different reconfirmed by other methods [213]. goals,” Endocrinology and Metabolism Clinics of North Amer- In addition to limitations of microarray technology ica, vol. 33, no. 2, pp. 283–303, 2004. itself, there are many more constrains in its application [10] R. Kahn, J. Buse, E. Ferrannini, and M. Stern, “The metabolic in understanding molecular functions of muscle, liver, and syndrome: time for a critical appraisal—joint statement adipose tissue in insulin resistance. Insulin resistance is from the American Diabetes Association and the European a continuous parameter and clinically identified on crite- Association for the Study of Diabetes,” Diabetes Care, vol. 28, no. 9, pp. 2289–2304, 2005. ria which have no pathophysiologic basis (i.e., metabolic [11] D. C. W. Lau, “Metabolic syndrome: perception or reality?” syndrome). Precise labeling of insulin resistance and its Current Atherosclerosis Reports, vol. 11, no. 4, pp. 264–271, ffi true control is di cult because it is function of adipose 2009. tissue mass and location (visceral or subcutaneous) and [12] S. H. Kim and G. M. Reaven, “The metabolic syndrome: one quality (good or bad fat). Moreover transciptome also step forward, two steps back,” Diabetes & Vascular Disease depends on the associated conditions like infection, stress, Research, vol. 1, no. 2, pp. 68–75, 2004. 10 Journal of Obesity

[13] G. M. Reaven, “The metabolic syndrome: Is this diagnosis [30] P.A. Permana, A. Del Parigi, and P.A. Tataranni, “Microarray necessary?” American Journal of Clinical Nutrition, vol. 83, gene expression profiling in obesity and insulin resistance,” no. 6, pp. 1237–1247, 2006. Nutrition, vol. 20, no. 1, pp. 134–138, 2004. [14]H.E.Bays,J.M.Gonzalez-Campoy,´ R. R. Henry et al., “Is [31] S. T. Nadler and A. D. Attie, “Please pass the chips: genomic adiposopathy (sick fat) an endocrine disease?” International insights into obesity and diabetes,” Journal of Nutrition,vol. Journal of Clinical Practice, vol. 62, no. 10, pp. 1474–1483, 131, no. 8, pp. 2078–2081, 2001. 2008. [32] C. Weyer, J. E. Foley, C. Bogardus, P. A. Tataranni, and R. E. [15] H. Bays and C. A. Dujovne, “Adiposopathy is a more rational Pratley, “Enlarged subcutaneous abdominal adipocyte size, treatment target for metabolic disease than obesity alone,” but not obesity itself, predicts type II diabetes independent Current Atherosclerosis Reports, vol. 8, no. 2, pp. 144–156, of insulin resistance,” Diabetologia, vol. 43, no. 12, pp. 1498– 2006. 1506, 2000. [16] H. Bays, “Adiposopathy: role of adipocyte factors in a new [33] B. G. Marques, D. B. Hausman, and R. J. Martin, “Association paradigm,” Expert Review of Cardiovascular Therapy,vol.3, of fat cell size and paracrine growth factors in development no. 2, pp. 187–189, 2005. of hyperplastic obesity,” American Journal of Physiology,vol. [17] C. H. Andreasen and G. Andersen, “Gene-environment 275, no. 6, pp. R1898–R1908, 1998. interactions and obesity—further aspects of genomewide [34]J.Smith,M.Al-Amri,P.Dorairaj,andA.Sniderman,“The association studies,” Nutrition, vol. 25, no. 10, pp. 998–1003, adipocyte life cycle hypothesis,” Clinical Science, vol. 110, no. 2009. 1, pp. 1–9, 2006. [18] J. Vockley, “Metabolism as a complex genetic trait, a [35] P. Arner, “Regional adipocity in man,” Journal of Endocrinol- systems biology approach: Implications for inborn errors ogy, vol. 155, no. 2, pp. 191–192, 1997. of metabolism and clinical diseases,” Journal of Inherited [36] I. Raz, R. Eldor, S. Cernea, and E. Shafrir, “Diabetes: Metabolic Disease, vol. 31, no. 5, pp. 619–629, 2008. insulin resistance and derangements in lipid metabolism. ff [19]L.O.Schulz,P.H.Bennett,E.Ravussinetal.,“E ects of Cure through intervention in fat transport and storage,” traditional and western environments on prevalence of type Diabetes/Metabolism Research and Reviews,vol.21,no.1,pp. 2 diabetes in Pima Indians in Mexico and the U.S,” Diabetes 3–14, 2005. Care, vol. 29, no. 8, pp. 1866–1871, 2006. [37] P. Arner, “Regional differences in protein production by [20]R.G.Nelson,M.L.Sievers,W.C.Knowleretal.,“Low human adipose tissue,” Biochemical Society Transactions,vol. incidence of fatal coronary heart disease in Pima Indians 29, no. 2, pp. 72–75, 2001. despite high prevalence of non-insulin-dependent diabetes,” [38]I.J.Bujalska,S.Kumar,andP.MStewart,“Doescentral Circulation, vol. 81, no. 3, pp. 987–995, 1990. obesity reflect ’Cushing’s disease of the omentum ’?” Lancet, [21] N. Abate, M. Chandalia, P. G. Snell, and S. M. Grundy, “Adi- vol. 349, no. 9060, pp. 1210–1213, 1997. pose tissue metabolites and insulin resistance in nondiabetic [39] M. C. Carr and J. D. Brunzell, “Abdominal obesity and Asian Indian men,” Journal of Clinical Endocrinology and dyslipidemia in the metabolic syndrome: Importance of type Metabolism, vol. 89, no. 6, pp. 2750–2755, 2004. 2 diabetes and familial combined hyperlipidemia in coronary [22] M. Chandalia, N. Abate, A. Garg, J. Stray-Gundersen, and artery disease risk,” Journal of Clinical Endocrinology and S. M. Grundy, “Relationship between generalized and upper Metabolism, vol. 89, no. 6, pp. 2601–2607, 2004. body obesity to insulin resistance in Asian Indian men,” ´ Journal of Clinical Endocrinology and Metabolism,vol.84,no. [40] J. P. Despres, “Intra-abdominal obesity: an untreated risk factor for Type 2 diabetes and cardiovascular disease,” Journal 7, pp. 2329–2335, 1999. [23] C. Dina, “New insights into the genetics of body weight,” of Endocrinological Investigation, vol. 29, no. 3, pp. 77–82, Current Opinion in Clinical Nutrition and Metabolic Care,vol. 2006. 11, no. 4, pp. 378–384, 2008. [41] R. N. Bergman, S. P. Kim, K. J. Catalano et al., “Why visceral [24] D. Pomp, D. Nehrenberg, and D. Estrada-Smith, “Complex fat is bad: mechanisms of the metabolic syndrome,” Obesity, genetics of obesity in mouse models,” Annual Review of vol. 14, supplement 1, pp. 16S–19S, 2006. Nutrition, vol. 28, pp. 331–345, 2008. [42] S. M. Grundy, “Obesity, metabolic syndrome, and car- [25] T. Joy and R. A. Hegele, “Genetics of Metabolic syndrome: is diovascular disease,” Journal of Clinical Endocrinology and there a role for phenomics?” Current Atherosclerosis Reports, Metabolism, vol. 89, no. 6, pp. 2595–2600, 2004. vol. 10, no. 3, pp. 201–208, 2008. [43] S. P. Weisberg, D. McCann, M. Desai, M. Rosenbaum, R. [26] A. G. Comuzzie, J. T. Williams, L. J. Martin, and J. Blangero, L. Leibel, and A. W. Ferrante, “Obesity is associated with “Searching for genes underlying normal variation in human macrophage accumulation in adipose tissue,” Journal of adiposity,” Journal of Molecular Medicine,vol.79,no.1,pp. Clinical Investigation, vol. 112, no. 12, pp. 1796–1808, 2003. 57–70, 2001. [44] H. Xu, G. T. Barnes, Q. Yang et al., “Chronic inflammation in [27] T. F. C. MacKay, E. A. Stone, and J. F. Ayroles, “The genetics of fat plays a crucial role in the development of obesity-related quantitative traits: challenges and prospects,” Nature Reviews insulin resistance,” Journal of Clinical Investigation, vol. 112, Genetics, vol. 10, no. 8, pp. 565–577, 2009. no. 12, pp. 1821–1830, 2003. [28] W. Cookson, L. Liang, G. Abecasis, M. Moffatt, and M. [45] U. Kintscher, M. Hartge, K. Hess et al., “T-lymphocyte Lathrop, “Mapping complex disease traits with global gene infiltration in visceral adipose tissue: a primary event in expression,” Nature Reviews Genetics, vol. 10, no. 3, pp. 184– adipose tissue inflammation and the development of obesity- 194, 2009. mediated insulin resistance,” Arteriosclerosis, Thrombosis, and [29] C. Buechler and A. Scha¨ffler, “Does global gene expression Vascular Biology, vol. 28, no. 7, pp. 1304–1310, 2008. analysis in type 2 diabetes provide an opportunity to identify [46] C. M. Smas and H. S. Sul, “Control of adipocyte differentia- highly promising drug targets?” Endocrine, Metabolic and tion,” Biochemical Journal, vol. 309, no. 3, pp. 697–710, 1995. Immune Disorders—Drug Targets, vol. 7, no. 4, pp. 250–258, [47] S. R. Farmer, “Transcriptional control of adipocyte forma- 2007. tion,” Cell Metabolism, vol. 4, no. 4, pp. 263–273, 2006. Journal of Obesity 11

[48] G. J. Hausman, J. T. Wright, and R. L. Richardson, “The [63] Wood I. S. and P. Trayhurn, “Adipokines and the signaling influence of extracellular matrix substrata on preadipocyte role of adipose tissue in inflammation and obesity,” Future development in serum-free cultures of stromal-vascular Lipidology, vol. 1, no. 1, pp. 81–89, 2006. cells,” Journal of Animal Science, vol. 74, no. 9, pp. 2117–2128, [64] L. Lionetti, M. P. Mollica, A. Lombardi, G. Cavaliere, G. 1996. Gifuni, and A. Barletta, “From chronic overnutrition to [49] J. E. Rundhaug, “Matrix metalloproteinases and angiogene- insulin resistance: the role of fat-storing capacity and inflam- sis,” Journal of cellular and molecular medicine,vol.9,no.2, mation,” Nutrition, Metabolism and Cardiovascular Diseases, pp. 267–285, 2005. vol. 19, no. 2, pp. 146–152, 2009. [50] J. Lilla, D. Stickens, and Z. Werb, “Metalloproteases and adi- [65]T.Suganami,J.Nishida,andY.Ogawa,“Aparacrineloop pogenesis: a weighty subject,” American Journal of Pathology, between adipocytes and macrophages aggravates inflamma- vol. 160, no. 5, pp. 1551–1554, 2002. tory changes: role of free fatty acids and tumor necrosis factor α [51] J. M. Argiles,´ J. Lopez-Soriano,´ V. Almendro, S. Busquets, ,” Arteriosclerosis, Thrombosis, and Vascular Biology,vol.25, andF.J.Lopez-Soriano,´ “Cross-talk between skeletal muscle no. 10, pp. 2062–2068, 2005. and adipose tissue: a link with obesity?” Medicinal Research [66] M. Chee, R. Yang, E. Hubbell et al., “Accessing genetic infor- Reviews, vol. 25, no. 1, pp. 49–65, 2005. mation with high-density DNA arrays,” Science, vol. 274, no. [52]T.J.Bartness,C.K.Song,H.Shi,R.R.Bowers,andM.T. 5287, pp. 610–614, 1996. ff Foster, “Brain-adipose tissue cross talk,” Proceedings of the [67] A. S. Ishkanian, C. A. Mallo ,S.K.Watsonetal.,“Atiling Nutrition Society, vol. 64, no. 1, pp. 53–64, 2005. resolution DNA microarray with complete coverage of the human genome,” Nature Genetics, vol. 36, no. 3, pp. 299–303, [53] P. Storz, H. Doppler,¨ A. Wernig, K. Pfizenmaier, and G. 2004. Muller,¨ “Cross-talk mechanisms in the development of insulin resistance of skeletal muscle cells. Palmitate rather [68] D. Gresham, M. J. Dunham, and D. Botstein, “Comparing than tumour necrosis factor inhibits insulin-dependent pro- whole genomes using DNA microarrays,” Nature Reviews tein kinase B (PKB)/AKT stimulation and glucose uptake,” Genetics, vol. 9, no. 4, pp. 291–302, 2008. European Journal of Biochemistry, vol. 266, no. 1, pp. 17–25, [69] Y. Kim and T. Park, “DNA microarrays to define and search 1999. for genes associated with obesity,” Biotechnology Journal,vol. 5, no. 1, pp. 99–112, 2010. [54] C. M. Smas and H. S. Sul, “Control of adipocyte differentiation,” Biochemical Journal, vol. 309, no. 3, pp. 697–710, 1995. [70] S. T. Nadler, J. P. Stoehr, K. L. Schueler, G. Tanimoto, B. S. Yandell, and A. D. Attie, “The expression of adipogenic genes [55]H.Bays,L.Mandarino,andR.A.DeFronzo,“Roleofthe is decreased in obesity and diabetes mellitus,” Proceedings adipocyte, free fatty acids, and ectopic fat in pathogenesis of of the National Academy of Sciences of the United States of Type 2 diabetes mellitus: peroxisomal proliferator-activated America, vol. 97, no. 21, pp. 11371–11376, 2000. receptor agonists provide a rational therapeutic approach,” [71] J. Gomez-Ambrosi,V.Catal´ an,´ A. Diez-Caballero et al., Journal of Clinical Endocrinology and Metabolism,vol.89,no. “Gene expression profile of omental adipose tissue in human 2, pp. 463–478, 2004. obesity,” The FASEB Journal, vol. 18, no. 1, pp. 215–217, 2004. [56] L. Heilbronn, S. R. Smith, and E. Ravussin, “Failure of fat [72] U. N. Das and A. A. Rao, “Gene expression profile in obesity cell proliferation, mitochondrial function and fat oxidation and type 2 diabetes mellitus,” Lipids in Health and Disease, results in ectopic fat storage, insulin resistance and type II vol. 6, article 35, 2007. diabetes mellitus,” International Journal of Obesity, vol. 28, [73]F.E.VonEyben,J.P.Kroustrup,J.F.Larsen,andJ.Celis, supplement 4, pp. S12–S21, 2004. “Comparison of gene expression in intra-abdominal and [57] S. P. Weisberg, D. McCann, M. Desai, M. Rosenbaum, R. subcutaneous fat: a study of men with morbid obesity and L. Leibel, and A. W. Ferrante Jr., “Obesity is associated nonobese men using microarray and proteomics,” Annals of with macrophage accumulation in adipose tissue,” Journal of the New York Academy of Sciences, vol. 1030, pp. 508–536, Clinical Investigation, vol. 112, no. 12, pp. 1796–1808, 2003. 2004. α [58]J.N.Fain,S.W.Bahouth,andA.K.Madan,“TNF release [74]A.Baranova,R.Collantes,S.J.Gowderetal.,“Obesity- by the nonfat cells of human adipose tissue,” International related differential gene expression in the visceral adipose Journal of Obesity, vol. 28, no. 4, pp. 616–622, 2004. tissue,” Obesity Surgery, vol. 15, no. 6, pp. 758–765, 2005. [59] J. N. Fain, D. S. Tichansky, and A. K. Madan, “Most of the [75] I. Dahlman and P. Arner, “Obesity and polymorphisms in interleukin 1 receptor antagonist, cathepsin S, macrophage genes regulating human adipose tissue,” International Journal migration inhibitory factor, nerve growth factor, and inter- of Obesity, vol. 31, no. 11, pp. 1629–1641, 2007. leukin 18 release by explants of human adipose tissue is by [76] Y. H. Lee, S. Nair, E. Rousseau et al., “Microarray pro- the non-fat cells, not by the adipocytes,” Metabolism,vol.55, filing of isolated abdominal subcutaneous adipocytes from no. 8, pp. 1113–1121, 2006. obese vs non-obese Pima Indians: increased expression of [60] D. Patsouris, P. P. Li, D. Thapar, J. Chapman, J. M. Olefsky, inflammation-related genes,” Diabetologia,vol.48,no.9,pp. and J. G. Neels, “Ablation of CD11c-positive cells normalizes 1776–1783, 2005. insulin sensitivity in obese insulin resistant animals,” Cell [77] S. Nair, Y. H. Lee, E. Rousseau et al., “Increased expression of Metabolism, vol. 8, no. 4, pp. 301–309, 2008. inflammation-related genes in cultured preadipocytes/stro- [61] J. M. Wentworth, G. Naselli, W. A. Brown et al., “Pro- mal vascular cells from obese compared with non-obese inflammatory CD11c+CD206+ adipose tissue macrophages Pima Indians,” Diabetologia, vol. 48, no. 9, pp. 1784–1788, are associated with insulin resistance in human obesity,” 2005. Diabetes, vol. 59, no. 7, pp. 1648–1656, 2010. [78] C. Darimont, O. Avanti, F. Blancher et al., “Contribution of [62] H. Bays and C. Ballantyne, “Adiposopathy: why do adiposity mesothelial cells in the expression of inflammatory-related and obesity cause metabolic disease?” Future Lipidology,vol. factors in omental adipose tissue of obese subjects,” Interna- 1, no. 4, pp. 389–420, 2006. tional Journal of Obesity, vol. 32, no. 1, pp. 112–120, 2008. 12 Journal of Obesity

[79] M. Jernas,J.Palming,K.Sj˚ oholm¨ et al., “Separation of [94] H. Sell and J. Eckel, “Monocyte chemotactic protein-1 and its human adipocytes by size: hypertrophic fat cells display role in insulin resistance,” Current Opinion in Lipidology,vol. distinct gene expression,” The FASEB Journal, vol. 20, no. 9, 18, no. 3, pp. 258–262, 2007. pp. 1540–1542, 2006. [95] P. Sartipy and D. J. Loskutoff, “Monocyte chemoattractant [80]S.K.Mathur,P.Chandra,S.Mishra,P.Ajmera,andP. protein 1 in obesity and insulin resistance,” Proceedings of the Sharma, “Type-2 diabetes related intermediate phenotypic National Academy of Sciences of the United States of America, traits in north Indian diabetics,” Indian Journal of Clinical vol. 100, no. 12, pp. 7265–7270, 2003. Biochemistry, vol. 22, no. 2, pp. 70–73, 2007. [96] A. Chen, S. Mumick, C. Zhang et al., “Diet induction of [81] R. Shah, Y. Lu, C. C. Hinkle et al., “Gene profiling of monocyte chemoattractant protein-1 and its impact on human adipose tissue during evoked inflammation in vivo,” obesity,” Obesity Research, vol. 13, no. 8, pp. 1311–1320, 2005. Diabetes, vol. 58, no. 10, pp. 2211–2219, 2009. [97] C. A. Curat, A. Miranville, C. Sengenes` et al., “From blood [82] U. Kintscher, M. Hartge, K. Hess et al., “T-lymphocyte monocytes to adipose tissue-resident macrophages: induc- infiltration in visceral adipose tissue: a primary event in tion of diapedesis by human mature adipocytes,” Diabetes, adipose tissue inflammation and the development of obesity- vol. 53, no. 5, pp. 1285–1292, 2004. mediated insulin resistance,” Arteriosclerosis, Thrombosis, and [98] J. Huber, F. W. Kiefer, M. Zeyda et al., “CC chemokine and Vascular Biology, vol. 28, no. 7, pp. 1304–1310, 2008. CC chemokine receptor profiles in visceral and subcutaneous [83] M. E. Rausch, S. Weisberg, P. Vardhana, and D. V. Tortoriello, adipose tissue are altered in human obesity,” Journal of “Obesity in C57BL/6J mice is characterized by adipose Clinical Endocrinology and Metabolism, vol. 93, no. 8, pp. tissue hypoxia and cytotoxic T-cell infiltration,” International 3215–3221, 2008. Journal of Obesity, vol. 32, no. 3, pp. 451–463, 2008. [99] M. L. Gruen, M. Hao, D. W. Piston, and A. H. Hasty, “Leptin [84] H. Wu, S. Ghosh, X. D. Perrard et al., “T-cell accumulation requires canonical migratory signaling pathways for induc- and regulated on activation, normal T cell expressed and tion of monocyte and macrophage chemotaxis,” American secreted upregulation in adipose tissue in obesity,” Circula- Journal of Physiology, vol. 293, no. 5, pp. C1481–C1488, 2007. tion, vol. 115, no. 8, pp. 1029–1038, 2007. [100] Y. Kamei, T. Suganami, T. Ehara et al., “Increased expression of DNA methyltransferase 3a in obese adipose tissue: studies [85]C.N.Lumeng,J.L.Bodzin,andA.R.Saltiel,“Obesity with transgenic mice,” Obesity, vol. 18, no. 2, pp. 314–321, induces a phenotypic switch in adipose tissue macrophage 2010. polarization,” Journal of Clinical Investigation, vol. 117, no. 1, [101] F. M. Gregoire, C. M. Smas, and H. S. Sul, “Understanding pp. 175–184, 2007. adipocyte differentiation,” Physiological Reviews,vol.78,no. [86] L. K. Heilbronn and L. V. Campbell, “Adipose tissue 3, pp. 783–809, 1998. macrophages, low grade inflammation and insulin resistance [102] S. R. Farmer, “Transcriptional control of adipocyte forma- in human obesity,” Current Pharmaceutical Design,vol.14, tion,” Cell Metabolism, vol. 4, no. 4, pp. 263–273, 2006. no. 12, pp. 1225–1230, 2008. [103] X. Guo and K. Liao, “Analysis of gene expression profile [87] J. M. Olefsky and C. K. Glass, “Macrophages, inflammation, during 3T3-L1 preadipocyte differentiation,” Gene, vol. 251, and insulin resistance,” Annual review of physiology,vol.72, no. 1, pp. 45–53, 2000. pp. 219–246, 2010. [104] A. Soukas, N. D. Socci, B. D. Saatkamp, S. Novelli, and J. M. [88]A.O’Hara,F.L.Lim,D.J.Mazzatti,andP.Trayhurn, Friedman, “Distinct transcriptional profiles of adipogenesis “Microarray analysis identifies matrix metalloproteinases in vivo and in vitro,” Journal of Biological Chemistry, vol. 276, (MMPs) as key genes whose expression is up-regulated in no. 36, pp. 34167–34174, 2001. human adipocytes by macrophage-conditioned medium,” [105] G. R. Burton, YU. Guan, R. Nagarajan, and R. E. McGe- Pflugers Archiv European Journal of Physiology, vol. 458, no. hee, “Microarray analysis of gene expression during early 6, pp. 1103–1114, 2009. adipocyte differentiation,” Gene, vol. 293, no. 1-2, pp. 21–31, [89] M. E. Rausch, S. Weisberg, P. Vardhana, and D. V. Tortoriello, 2002. “Obesity in C57BL/6J mice is characterized by adipose [106] G. R. Burton, R. Nagarajan, C. A. Peterson, and R. E. McGe- tissue hypoxia and cytotoxic T-cell infiltration,” International hee, “Microarray analysis of differentiation-specific gene Journal of Obesity, vol. 32, no. 3, pp. 451–463, 2008. expression during 3T3-L1 adipogenesis,” Gene, vol. 329, no. [90] P. Trayhurn, B. Wang, and I. S. Wood, “Hypoxia in adipose 1-2, pp. 167–185, 2004. tissue: a basis for the dysregulation of tissue function in [107] B. A. Jessen and G. J. Stevens, “Expression profiling during obesity?” British Journal of Nutrition, vol. 100, no. 2, pp. 227– adipocyte differentiation of 3T3-L1 fibroblasts,” Gene,vol. 235, 2008. 299, no. 1-2, pp. 95–100, 2002. [91] N. Hosogai, A. Fukuhara, K. Oshima et al., “Adipose [108] D. L. Gerhold, F. Liu, G. Jiang et al., “Gene expression profile tissue hypoxia in obesity and its impact on adipocytokine of adipocyte differentiation and its regulation by peroxisome dysregulation,” Diabetes, vol. 56, no. 4, pp. 901–911, 2007. proliferator-activated receptor-γ agonists,” Endocrinology, [92]B.Wang,I.S.Wood,andP.Trayhurn,“Dysregulationofthe vol. 143, no. 6, pp. 2106–2118, 2002. expression and secretion of inflammation-related adipokines [109] A. Hsiao, D. S. Worrall, J. M. Olefsky, and S. Subramaniam, by hypoxia in human adipocytes,” Pflugers Archiv European “Variance-modeled posterior inference of microarray data: Journal of Physiology, vol. 455, no. 3, pp. 479–492, 2007. detecting gene-expression changes in 3T3-L1 adipocytes,” [93] G. B. Di Gregorio, A. Yao-Borengasser, N. Rasouli et al., Bioinformatics, vol. 20, no. 17, pp. 3108–3127, 2004. “Expression of CD68 and macrophage chemoattractant [110] J. Gomez-Ambrosi,V.Catal´ an,´ A. Diez-Caballero et al., protein-1 genes in human adipose and muscle tissues: “Gene expression profile of omental adipose tissue in human association with cytokine expression, insulin resistance, and obesity,” The FASEB Journal, vol. 18, no. 1, pp. 215–217, 2004. reduction by pioglitazone,” Diabetes, vol. 54, no. 8, pp. 2305– [111] H. Hackl, T. R. Burkard, A. Sturn et al., “Molecular processes 2313, 2005. during fat cell development revealed by gene expression Journal of Obesity 13

profiling and functional annotation,” Genome Biology,vol.6, [127] M. Severgnini, S. Bicciato, E. Mangano et al., “Strategies for no. 13, article 108, 2005. comparing gene expression profiles from different microar- [112] J. Liu, S. M. DeYoung, M. Zhang, M. Zhang, A. Cheng, and A. ray platforms: application to a case-control experiment,” R. Saltiel, “Changes in integrin expression during adipocyte Analytical Biochemistry, vol. 353, no. 1, pp. 43–56, 2006. differentiation,” Cell Metabolism, vol. 2, no. 3, pp. 165–177, [128] L. Shi, L. H. Reid, W. D. Jones et al., “The MicroArray Quality 2005. Control (MAQC) project shows inter- and intraplatform [113] C. Pantoja, J. T. Huff, and K. R. Yamamoto, “Glucocorticoid reproducibility of gene expression measurements,” Nature signaling defines a novel commitment state during adipoge- Biotechnology, vol. 24, no. 9, pp. 1151–1161, 2006. nesis in vitro,” Molecular Biology of the Cell, vol. 19, no. 10, [129] K. Birsoy, Z. Chen, and J. Friedman, “Transcriptional regula- pp. 4032–4041, 2008. tion of adipogenesis by KLF4,” Cell Metabolism,vol.7,no.4, [114] J. S. Huo, R. C. McEachin, T. X. Cui et al., “Profiles pp. 339–347, 2008. of growth hormone (GH)-regulated genes reveal time- [130] Z. Chen, J. I. Torrens, A. Anand, B. M. Spiegelman, and J. dependent responses and identify a mechanism for regula- M. Friedman, “Krox20 stimulates adipogenesis via C/EBPβ- tion of activating transcription factor 3 by GH,” Journal of dependent and -independent mechanisms,” Cell Metabolism, Biological Chemistry, vol. 281, no. 7, pp. 4132–4141, 2006. vol. 1, no. 2, pp. 93–106, 2005. [115] P. Akerblad,˚ R. Mansson,˚ A. Lagergren et al., “Gene expression analysis suggests that EBF-1 and PPARγ2 induce adipo- [131] L. C. Chao, S. J. Bensinger, C. J. Villanueva, K. Wroblewski, and P. Tontonoz, “Inhibition of adipocyte differentiation by genesis of NIH-3T3 cells with similar efficiency and kinetics,” Nur77, Nurr1, and Nor1,” Molecular Endocrinology, vol. 22, Physiological Genomics, vol. 23, no. 2, pp. 206–216, 2005. no. 12, pp. 2596–2608, 2008. [116] A. Baudry, Z.-Z. Yang, and B. A. Hemmings, “PKBα is required for adipose differentiation of mouse embryonic [132] W. S. Au, V. A. Payne, S. O’Rahilly, and J. J. Rochford, “The fibroblast,” Journal of Cell Science, vol. 119, no. 5, pp. 889– NR4A family of orphan nuclear receptors are not required for 897, 2006. adipogenesis,” International Journal of Obesity, vol. 32, no. 2, [117] M. Christian, E. Kiskinis, D. Debevec, G. Leonardsson, pp. 388–392, 2008. R. White, and M. G. Parker, “RIP140-targeted repression [133] T. Fumoto, T. Yamaguchi, F. Hirose, and T. Osumi, “Orphan of gene expression in adipocytes,” Molecular and Cellular nuclear receptor Nur77 accelerates the initial phase of Biology, vol. 25, no. 21, pp. 9383–9391, 2005. adipocyte differentiation in 3T3-L1 cells by promoting [118] T. C. Otto, R. R. Bowers, and M. D. Lane, “BMP-4 treatment mitotic clonal expansion,” Journal of Biochemistry, vol. 141, of C3H10T1/2 stem cells blocks expression of MMP-3 and no. 2, pp. 181–192, 2007. MMP-13,” Biochemical and Biophysical Research Communi- [134] S. E. Ross, R. L. Erickson, I. Gerin et al., “Microarray cations, vol. 353, no. 4, pp. 1097–1104, 2007. analyses during adipogenesis: understanding the effects of [119] Y. H. Tseng, A. J. Butte, E. Kokkotou et al., “Prediction Wnt signaling on adipogenesis and the roles of liver X of preadipocyte differentiation by gene expression reveals receptor α in adipocyte metabolism,” Molecular and Cellular role of insulin receptor substrates and necdin,” Nature Cell Biology, vol. 22, no. 16, pp. 5989–5999, 2002. Biology, vol. 7, no. 6, pp. 601–611, 2005. [135] J. B. Seo, H. M. Moon, W. S. Kim et al., “Activated [120] J. A. Timmons, K. Wennmalm, O. Larsson et al., “Myogenic liver X receptors stimulate adipocyte differentiation through gene expression signature establishes that brown and white induction of peroxisome proliferator-activated receptor γ adipocytes originate from distinct cell lineages,” Proceedings expression,” Molecular and Cellular Biology,vol.24,no.8,pp. of the National Academy of Sciences of the United States of 3430–3444, 2004. America, vol. 104, no. 11, pp. 4401–4406, 2007. [136] M. A. Jimenez, P. Akerblad,M.Sigvardsson,andE.D.˚ [121] S. C. Hung, C. F. Chang, H. L. Ma, T. H. Chen, and L. Low- Rosen, “Critical role for Ebf1 and Ebf2 in the adipogenic Tone Ho, “Gene expression profiles of early adipogenesis in transcriptional cascade,” Molecular and Cellular Biology,vol. human mesenchymal stem cells,” Gene, vol. 340, no. 1, pp. 27, no. 2, pp. 743–757, 2007. 141–150, 2004. [137] Q. Tong, G. Dalgin, H. Xu, C. N. Ting, J. M. Leiden, and [122] I. Sekiya, B. L. Larson, J. T. Vuoristo, J. G. Cui, and D. J. ff G. S. Hotamisligil, “Function of GATA transcription factors Prockop, “Adipogenic di erentiation of human adult stem in preadipocyte-adipocyte transition,” Science, vol. 290, no. cells from bone marrow stroma (MSCs),” Journal of Bone and 5489, pp. 134–138, 2000. Mineral Research, vol. 19, no. 2, pp. 256–264, 2004. [138] A. Jeltsch, “Beyond Watson and Crick: DNA methylation [123] T. Nakamura, S. Shiojima, Y. Hirai et al., “Temporal gene and molecular enzymology of DNA methyltransferases,” expression changes during adipogenesis in human mes- ChemBioChem, vol. 3, no. 4, pp. 275–293, 2002. enchymal stem cells,” Biochemical and Biophysical Research Communications, vol. 303, no. 1, pp. 306–312, 2003. [139] T. Horii, S. Morita, M. Kimura, and I. Hatada, “Epigenetic ff [124] M. Scheideler, C. Elabd, L.-E. Zaragosi et al., “Comparative regulation fo adipocyte di eretiation by a Rho guanine transcriptomics of human multipotent stem cells during nucleotide exchange factor, WGEF,” PLoS One,vol.4,no.6, adipogenesis and osteoblastogenesis,” BMC Genomics,vol.9, Article ID e5809, 2009. article 340, 2008. [140] Y. Okada, H. Sakaue, T. Nagare, and M. Kasuga, “Diet- [125] S. Urs, C. Smith, B. Campbell et al., “Gene expression pro- induced up-regulation of gene expression in adipocytes filing in human preadipocytes and adipocytes by microarray without changes in DNA methylation,” Kobe Journal of analysis,” Journal of Nutrition, vol. 134, no. 4, pp. 762–770, Medical Sciences, vol. 54, no. 5, pp. E241–E249, 2008. 2004. [141] N. Nakanishi, Y. Nakagawa, N. Tokushige et al., “The up- [126] P. K. Tan, T. J. Downey, E. L. Spitznagel et al., “Evaluation of regulation of microRNA-335 is associated with lipid gene expression measurements from commercial microarray metabolism in liver and white adipose tissue of genetically platforms,” Nucleic Acids Research, vol. 31, no. 19, pp. 5676– obese mice,” Biochemical and Biophysical Research Commu- 5684, 2003. nications, vol. 385, no. 4, pp. 492–496, 2009. 14 Journal of Obesity

[142] T. Sun, M. Fu, A. L. Bookout, S. A. Kliewer, and D. J. Man- through down regulation of perilipin,” Journal of Nutritional gelsdorf, “MicroRNA let-7 regulates 3T3-L1 adipogenesis,” Biochemistry, vol. 21, no. 4, pp. 317–324, 2010. Molecular Endocrinology, vol. 23, no. 6, pp. 925–931, 2009. [157] A. E. Reid, “Nonalcoholic steatohepatitis,” Gastroenterology, [143] N. A. Ducharme and P. E. Bickel, “Lipid droplets in vol. 121, no. 3, pp. 710–723, 2001. lipogenesis and lipolysis,” Endocrinology, vol. 149, no. 3, pp. [158] A. M. Diehl, “Nonalcoholic steatohepatitis,” Seminars in Liver 942–949, 2008. Disease, vol. 19, no. 2, pp. 221–230, 1999. [144] S. Kersten, “Mechanisms of nutritional and hormonal regula- [159] E. E. Powell, W. G. E. Cooksley, R. Hanson, J. Searle, J. tion of lipogenesis,” EMBO Reports, vol. 2, no. 4, pp. 282–286, W. Halliday, and L. W. Powell, “The natural history of 2001. nonalcoholic steatohepatitis: a follow-up study of forty-two [145] M. Jernas,J.Palming,K.Sj˚ oholm¨ et al., “Separation of patients for up to 21 years,” Hepatology,vol.11,no.1,pp. human adipocytes by size: hypertrophic fat cells display 74–80, 1990. distinct gene expression,” FASEB Journal, vol. 20, no. 9, pp. [160] E. M. Brunt, C. G. Janney, A. M. Di Bisceglie, B. A. E832–E839, 2006. Neuschwander-Tetri, and B. R. Bacon, “Nonalcoholic steato- [146] K. Sjoholm,¨ J. Palming, L. E. Olofsson et al., “A microarray hepatitis: a proposal for grading and staging the histological search for genes predominantly expressed in human omental lesions,” American Journal of Gastroenterology, vol. 94, no. 9, adipocytes: adipose tissue as a major production site of serum pp. 2467–2474, 1999. amyloid A,” Journal of Clinical Endocrinology and Metabolism, [161] C. A. Matteoni, Z. M. Younossi, T. Gramlich, N. Boparai, vol. 90, no. 4, pp. 2233–2239, 2005. Y. C. Liu, and A. J. McCullough, “Nonalcoholic fatty liver [147] S. B. Su, W. Gong, JI. L. Gao et al., “A seven-transmembrane, disease: a spectrum of clinical and pathological severity,” G protein-coupled receptor, FPRL1, mediates the chemotac- Gastroenterology, vol. 116, no. 6, pp. 1413–1419, 1999. tic activity of serum amyloid A for human phagocytic cells,” [162] E. Bugianesi, N. Leone, E. Vanni et al., “Expanding the natu- Journal of Experimental Medicine, vol. 189, no. 2, pp. 395– ral history of nonalcoholic steatohepatitis: from cryptogenic 402, 1999. cirrhosis to hepatocellular carcinoma,” Gastroenterology,vol. [148] J. A. Stonik, A. T. Remaley, S. J. Demosky, E. B. Neufeld, A. 123, no. 1, pp. 134–140, 2002. Bocharov, and H. B. Brewer, “Serum Amyloid a promotes [163] S. H. Caldwell, D. H. Oelsner, J. C. Iezzoni, E. E. Hespenheide, ABCA1-dependent and ABCA1-independent lipid efflux E. H. Battle, and C. J. Driscoll, “Cryptogenic cirrhosis: from cells,” Biochemical and Biophysical Research Communi- clinical characterization and risk factors for underlying cations, vol. 321, no. 4, pp. 936–941, 2004. disease,” Hepatology, vol. 29, no. 3, pp. 664–669, 1999. [149] M. Cortes-Canteli,´ M. Wagner, W. Ansorge, and A. [164] A. Poonawala, S. P. Nair, and P. J. Thuluvath, “Prevalence of Perez-Castillo,´ “Microarray analysis supports a role for obesity and diabetes in patients with cryptogenic cirrhosis: a CCAAT/enhancer-binding protein-β in brain injury,” Journal case-control study,” Hepatology, vol. 32, no. 4, pp. 689–692, of Biological Chemistry, vol. 279, no. 14, pp. 14409–14417, 2000. 2004. [165] J. Ong, Z. M. Younossi, V.Reddy et al., “Cryptogenic cirrhosis [150] B. G. Gabrielsson, J. M. Johansson, M. Lonn¨ et al., “High and posttransplantation nonalcoholic fatty liver disease,” expression of complement components in omental adipose Liver Transplantation, vol. 7, no. 9, pp. 797–801, 2001. tissue in obese men,” Obesity Research, vol. 11, no. 6, pp. 699– [166] F. Chiappini, A. Barrier, R. Saffroy et al., “Exploration of 708, 2003. global gene expression in human liver steatosis by high- [151] R. E. Pratley, D. B. Thompson, M. Prochazka et al., “An density oligonucleotide microarray,” Laboratory Investiga- autosomal genomic scan for loci linked to prediabetic tion, vol. 86, no. 2, pp. 154–165, 2006. phenotypes in pima indians,” Journal of Clinical Investigation, [167] R. Sreekumar, B. Rosado, D. Rasmussen, and M. Charlton, vol. 101, no. 8, pp. 1757–1764, 1998. “Hepatic gene expression in histologically progressive nonal- [152] T. Suganami, K. Tanimoto-Koyama, J. Nishida et al., “Role coholic steatohepatitis,” Hepatology, vol. 38, no. 1, pp. 244– of the Toll-like receptor 4/NF-κBpathwayinsaturated 251, 2003. fatty acid-induced inflammatory changes in the interac- [168] O. Cheung, P. Puri, C. Eicken et al., “Nonalcoholic steato- tion between adipocytes and macrophages,” Arteriosclerosis, hepatitis is associated with altered hepatic MicroRNA expres- Thrombosis, and Vascular Biology, vol. 27, no. 1, pp. 84–91, sion,” Hepatology, vol. 48, no. 6, pp. 1810–1820, 2008. 2007. [169] D. E. Kelley, J. He, E. V. Menshikova, and V. B. Ritov, [153] T. Suganami, X. Yuan, Y. Shimoda et al., “Activating tran- “Dysfunction of mitochondria in human skeletal muscle in scription factor 3 constitutes a negative feedback mechanism type 2 diabetes,” Diabetes, vol. 51, no. 10, pp. 2944–2950, that attenuates saturated fatty Acid/Toll-like receptor 4 2002. signaling and macrophage activation in obese adipose tissue,” [170] V. B. Schrauwen-Hinderling, M. E. Kooi, M. K. C. Hesselink Circulation Research, vol. 105, no. 1, pp. 25–32, 2009. et al., “Impaired in vivo mitochondrial function but similar [154] M. Hassan, C. E. Yazidi, J. F. Landrier, D. Lairon, A. intramyocellular lipid content in patients with type 2 diabetes Margotat, and M. J. Amiot, “Phloretin enhances adipocyte mellitus and BMI-matched control subjects,” Diabetologia, differentiation and adiponectin expression in 3T3-L1 cells,” vol. 50, no. 1, pp. 113–120, 2007. Biochemical and Biophysical Research Communications,vol. [171] V. B. Ritov, E. V. Menshikova, J. He, R. E. Ferrell, B. H. 361, no. 1, pp. 208–213, 2007. Goodpaster, and D. E. Kelley, “Deficiency of subsarcolemmal [155] M. Hassan, C. E. Yazidi, C. Malezet-Desmoulins, M. J. mitochondria in obesity and type 2 diabetes,” Diabetes,vol. Amiot, and A. Margotat, “Gene expression profiling of 3T3- 54, no. 1, pp. 8–14, 2005. L1 adipocytes exposed to phloretin,” Journal of Nutritional [172] R. Boushel, E. Gnaiger, P. Schjerling, M. Skovbro, R. Biochemistry, vol. 21, no. 7, pp. 645–652, 2010. Kraunsøe, and F. Dela, “Patients with type 2 diabetes have [156] Y. C. Wang, W. H. Kuo, C. Y. Chen et al., “Docosahexaenoic normal mitochondrial function in skeletal muscle,” Dia- acid regulates serum amyloid A protein to promote lipolysis betologia, vol. 50, no. 4, pp. 790–796, 2007. Journal of Obesity 15

[173] M. Mogensen, K. Sahlin, M. Fernstrom¨ et al., “Mitochondrial [188] W. Jobgen, W. J. Fu, H. Gao et al., “High fat feeding and respiration is decreased in skeletal muscle of patients with dietary L-arginine supplementation differentially regulate type 2 diabetes,” Diabetes, vol. 56, no. 6, pp. 1592–1599, 2007. gene expression in rat white adipose tissue,” Amino Acids,vol. [174] R. Sreekumar, P. Halvatsiotis, J. C. Schimke, and K. Sreeku- 37, no. 1, pp. 187–198, 2009. maran Nair, “Gene expression profile in skeletal muscle of [189] C. Bolduc, M. Yoshioka, and J. St-Amand, “Transcriptomic type 2 diabetes and the effect of insulin treatment,” Diabetes, characterization of the long-term dihydrotestosterone effects vol. 51, no. 6, pp. 1913–1920, 2002. in adipose tissue,” Obesity, vol. 15, no. 5, pp. 1107–1132, [175] M. E. Patti, A. J. Butte, S. Crunkhorn et al., “Coordinated 2007. reduction of genes of oxidative metabolism in humans with [190] V.Catalan,´ J. Gomez-Ambrosi,´ A. Rodrıguez´ et al., “Increased insulin resistance and diabetes: potential role of PGC1 and adipose tissue expression of lipocalin-2 in obesity is NRF1,” Proceedings of the National Academy of Sciences of the related to inflammation and matrix metalloproteinase-2 United States of America, vol. 100, no. 14, pp. 8466–8471, and metalloproteinase-9 activities in humans,” Journal of 2003. Molecular Medicine, vol. 87, no. 8, pp. 803–813, 2009. [176] V. K. Mootha, C. M. Lindgren, K. F. Eriksson et al., “PGC- [191] J. Zhang, Y. Wu, Y. Zhang, D. Leroith, D. A. Bernlohr, 1α-responsive genes involved in oxidative phosphorylation and X. Chen, “The role of lipocalin 2 in the regulation of are coordinately downregulated in human diabetes,” Nature inflammation in adipocytes and macrophages,” Molecular Genetics, vol. 34, no. 3, pp. 267–273, 2003. Endocrinology, vol. 22, no. 6, pp. 1416–1426, 2008. [177] V. Skov, D. Glintborg, S. Knudsen et al., “Reduced expres- [192] F. Aoki, S. Honda, H. Kishida et al., “Suppression by sion of nuclear-encoded genes involved in mitochondrial licorice flavonoids of abdominal fat accumulation and body oxidative metabolism in skeletal muscle of insulin-resistant weight gain in high-fat diet-induced obese C57BL/6J mice,” women with polycystic ovary syndrome,” Diabetes, vol. 56, Bioscience, Biotechnology and Biochemistry,vol.71,no.1,pp. no. 9, pp. 2349–2355, 2007. 206–214, 2007. [178] C. M. Frederiksen, K. Højlund, L. Hansen et al., “Tran- [193] O. Seda,ˇ L. Sedovˇ a,´ O. Oliyarnyk et al., “Pharmacogenomics scriptional profiling of myotubes from patients with type of metabolic effects of rosiglitazone,” Pharmacogenomics,vol. 2 diabetes: no evidence for a primary defect in oxidative 9, no. 2, pp. 141–155, 2008. phosphorylation genes,” Diabetologia, vol. 51, no. 11, pp. [194] V. Van Harmelen, A. Dicker, M. Ryden´ et al., “Increased 2068–2077, 2008. lipolysis and decreased leptin production by human omental [179] X. Yang, R. Pratley, S. Tokraks, C. Bogardus, and P. Permana, as compared with subcutaneous preadipocytes,” Diabetes, “Microarray profiling of skeletal muscle tissues from equally vol. 51, no. 7, pp. 2029–2036, 2002. obese, non-diabetic insulin-sensitive and insulin-resistant [195] J. M. Ramis, N. L. W. Franssen-van Hal, E. Kramer et al., Pima Indians,” Diabetologia, vol. 45, no. 11, pp. 1584–1593, “Carboxypeptidase E and thrombospondin-1 are differently 2002. expressed in subcutaneous and visceral fat of obese subjects,” [180] L. L. Nguyen, A. D. Kriketos, D. P. Hancock, I. D. Caterson, Cellular and Molecular Life Sciences, vol. 59, no. 11, pp. 1960– and G. S. Denyer, “Insulin resistance does not influence gene 1971, 2002. expression in skeletal muscle,” Journal of Biochemistry and [196] M. Dolinkova,´ I. Dostalov´ a,´ Z. Lacinovaetal.,“The´ Molecular Biology, vol. 39, no. 4, pp. 457–463, 2006. endocrine profile of subcutaneous and visceral adipose tissue [181] I. Dahlman, K. Linder, E. A. Nordstrom¨ et al., “Changes in of obese patients,” Molecular and Cellular Endocrinology,vol. adipose tissue gene expression with energy-restricted diets in 291, no. 1-2, pp. 63–70, 2008. obese women,” American Journal of Clinical Nutrition,vol. [197] P. Sebastiani, N. Timofeev, D. A. Dworkis, T. T. Perls, and 81, no. 6, pp. 1275–1285, 2005. M. H. Steinberg, “Genome-wide association studies and the [182] F. Capel, E. Klimcˇakov´ a,´ N. Viguerie et al., “Macrophages and genetic dissection of complex traits,” American Journal of adipocytes in human obesity: adipose tissue gene expression Hematology, vol. 84, no. 8, pp. 504–515, 2009. and insulin sensitivity during calorie restriction and weight stabilization,” Diabetes, vol. 58, no. 7, pp. 1558–1567, 2009. [198] M. I. McCarthy and E. Zeggini, “Genome-wide association [183] K. Clement,´ N. Viguerie, C. Poitou et al., “Weight loss studies in type 2 diabetes,” Current Diabetes Reports,vol.9, regulates inflammation-related genes in white adipose tissue no. 2, pp. 164–171, 2009. of obese subjects,” FASEB Journal, vol. 18, no. 14, pp. 1657– [199] R. B. Brem, G. Yvert, R. Clinton, and L. Kruglyak, “Genetic 1669, 2004. dissection of transcriptional regulation in budding yeast,” [184] N. Viguerie, H. Vidal, P. Arner et al., “Adipose tissue gene Science, vol. 296, no. 5568, pp. 752–755, 2002. expression in obese subjects during low-fat and high-fat [200] V. G. Cheung and R. S. Spielman, “The genetics of variation hypocaloric diets,” Diabetologia, vol. 48, no. 1, pp. 123–131, in gene expression,” Nature Genetics, vol. 32, no. 5, pp. 522– 2005. 525, 2002. [185] M. Kolehmainen, T. Salopuro, U. S. Schwab et al., “Weight [201] V. G. Cheung, L. K. Conlin, T. M. Weber et al., “Natural vari- reduction modulates expression of genes involved in extra- ation in human gene expression assessed in lymphoblastoid cellular matrix and cell death: the GENOBIN study,” Interna- cells,” Nature Genetics, vol. 33, no. 3, pp. 422–425, 2003. tional Journal of Obesity, vol. 32, no. 2, pp. 292–303, 2008. [202] E. E. Schadt, S. A. Monks, T. A. Drake et al., “Genetics of gene [186] S. J. van Dijk, E. J. M. Feskens, M. B. Bos et al., “A expression surveyed in maize, mouse and man,” Nature,vol. saturated fatty acid-rich diet induces an obesity-linked 422, no. 6929, pp. 297–302, 2003. proinflammatory gene expression profile in adipose tissue of [203] R. C. Jansen and J. P. Nap, “Genetical genomics: the added subjects at risk of metabolic syndrome,” American Journal of value from segregation,” Trends in Genetics,vol.17,no.7,pp. Clinical Nutrition, vol. 90, no. 6, pp. 1656–1664, 2009. 388–391, 2001. [187] W. J. Fu, T. E. Haynes, R. Kohli et al., “Dietary L-arginine [204] M. Morloy, C. M. Molony, T. M. Weber et al., “Genetic anal- supplementation reduces fat mass in Zucker diabetic fatty ysis of genome-wide variation in human gene expression,” rats,” Journal of Nutrition, vol. 135, no. 4, pp. 714–721, 2005. Nature, vol. 430, no. 7001, pp. 743–747, 2004. 16 Journal of Obesity

[205] V. G. Cheung, R. S. Spielman, K. G. Ewens, T. M. Weber, M. Morley, and J. T. Burdick, “Mapping determinants of human gene expression by regional and genome-wide association,” Nature, vol. 437, no. 7063, pp. 1365–1369, 2005. [206] Y. Chen, J. Zhu, P. Y. Lum et al., “Variations in DNA elucidate molecular networks that cause disease,” Nature, vol. 452, no. 7186, pp. 429–435, 2008. [207] S. Wang, L. M. Sparks, H. Xie, F. L. Greenway, L. De Jonge, and S. R. Smith, “Subtyping obesity with microarrays: implications for the diagnosis and treatment of obesity,” International Journal of Obesity, vol. 33, no. 4, pp. 481–489, 2009. [208] C. Buechler and A. Scha¨ffler, “Does global gene expression analysis in type 2 diabetes provide an opportunity to identify highly promising drug targets?” Endocrine, Metabolic and Immune Disorders—Drug Targets, vol. 7, no. 4, pp. 250–258, 2007. [209] S. T. Nadler and A. D. Attie, “Please pass the chips: genomic insights into obesity and diabetes,” Journal of Nutrition,vol. 131, no. 8, pp. 2078–2081, 2001. [210] P. A. Permana, S. Nair, and Y. H. Lee, “Application of DNA microarray to the study of human adipose tissue/cells,” Methods in Molecular Biology, vol. 456, pp. 141–154, 2008. [211] D. Gresham, M. J. Dunham, and D. Botstein, “Comparing whole genomes using DNA microarrays,” Nature Reviews Genetics, vol. 9, no. 4, pp. 291–302, 2008. [212] R. Brandt, R. Grutzmann,¨ A. Bauer et al., “DNA microarray analysis of pancreatic malignancies,” Pancreatology,vol.4,no. 6, pp. 587–597, 2004. [213] G. L. G. Miklos and R. Maleszka, “Microarray reality checks in the context of a complex disease,” Nature Biotechnology, vol. 22, no. 5, pp. 615–621, 2004. [214] D. M. Mutch, J. Tordjman, V. Pelloux et al., “Needle and surgical biopsy techniques differentially affect adipose tissue gene expression profiles,” American Journal of Clinical Nutrition, vol. 89, no. 1, pp. 51–57, 2009. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 340241, 12 pages doi:10.1155/2011/340241

Review Article Genetic Variance in Uncoupling Protein 2 in Relation to Obesity, Type 2 Diabetes, and Related Metabolic Traits: Focus on the Functional −866G>A Promoter Variant (rs659366)

Louise T. Dalgaard

Department of Science, Systems and Models, Roskilde University, Universitetsvej 1, 4000 Roskilde, Denmark

Correspondence should be addressed to Louise T. Dalgaard, [email protected]

Received 1 December 2010; Accepted 21 February 2011

Academic Editor: R. Prager

Copyright © 2011 Louise T. Dalgaard. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Uncoupling proteins (UCPs) are mitochondrial proteins able to dissipate the proton gradient of the inner mitochondrial membrane when activated. This decreases ATP-generation through oxidation of fuels and may theoretically decrease energy expenditure leading to obesity. Evidence from Ucp(−/−) mice revealed a role of UCP2 in the pancreatic β-cell, because β-cells without UCP2 had increased glucose-stimulated insulin secretion. Thus, from being a candidate gene for obesity UCP2 became a valid candidate gene for type 2 diabetes mellitus. This prompted a series of studies of the human UCP2 and UCP3 genes with respect to obesity and diabetes. Of special interest was a promoter variant of UCP2 situated 866bp upstream of transcription initiation (−866G>A, rs659366). This variant changes promoter activity and has been associated with obesity and/or type 2 diabetes in several, although not all, studies. The aim of the current paper is to summarize current evidence of association of UCP2 genetic variation with obesity and type 2 diabetes, with focus on the −866G>A polymorphism.

1. Introduction 2. Physiological Functions of UCP2 and UCP3 Uncoupling protein 2 (UCP2) and uncoupling protein 3 UCP2 is ubiquitously expressed [1, 2]whereasUCP3is (UCP3) belong to a large family of mitochondrial transmem- found predominantly in skeletal muscle and brown adipose brane carriers. UCP2 was identified in 1997 based on its tissue [3, 4, 10], and their expression is both induced by fast- homology to the brown fat uncoupling protein (UCP, then ing, and peroxisome proliferators as well as hyperglycemia, renamed UCP1) [1, 2]. Shortly thereafter, UCP3 was cloned which indicates a role connected with the availability of fuel also based on homology to UCP1 and UCP2 [3, 4]. Later, substrates [11–14]. However, the upregulation in response more distantly related proteins were identified and named to thyroid hormone, cold, β3-adrenergic agonists, and high UCP4 and UCP5 (BMCP1) [5–7]. The physiological role of fat diets also suggests involvement in regulation of energy UCP1 is well established; it is responsible for nonshivering expenditure [15–17]. thermogenesis in brown fat, in which it induces proton leak Neither UCP2 nor UCP3 affects basal proton con- across the inner mitochondrial membrane [8, 9]. Now 14 ductance of the mitochondrial inner membrane [18–23]. years later, the physiological functions of UCP2 and UCP3 However, they do induce proton leak across the inner are still under debate, as is the role of genetic variation mitochondrial membrane when activated by, for example, in these. The aim of this paper is to recapitulate the fatty acids, superoxide, or free radical derived peroxidation currently published literature on human genetic variation products of membrane phospholipids [24, 25]. UCP2 and in the UCP2 genomic region concerning development of UCP3 may decrease the formation of superoxide and reactive obesity, type 2 diabetes, and related metabolic disorders oxygen species (ROS) by mild uncoupling of the respiratory with focus on the −866G>A promoter polymorphism chain, whose activity is increased under these circumstances (rs659366). [21, 24, 26](Figure 1). This is concordant with the induction 2 Journal of Obesity of UCP2 and UCP3 during cold, fasting and high fat feeding, since these conditions require lipid oxidation and thus high Glucose Fatty acids Glutamine activity of the respiratory chain [18]. On the other hand, it has recently been suggested that UCP2 restricts pyruvate −866G>A A+ efflux from the mitochondria and hence ensures availability UCP2 ↑ O − (rs659366) − 2 of substrates for the citric acid cycle, which would then G − explain the increase in glucose oxidation compared with (−/−) lipid oxidation in Ucp2 mouse embryonic fibroblasts ATP ↓ ROS ↓ [27, 28].WhetherthisproposedfunctionofUCP2isshared −? with UCP3 is not known, and this hypothesis requires more ? β-cell dysfunction Inflammation ↓ investigation as it is less supported by experimental evidence Aging ↓ Cancer ↓ as the theory of mild uncoupling. − Complications ↓ In the pancreatic β-cell, UCP2 is important for appro- Less efficient metabolism priate glucose-stimulated insulin secretion. Overexpression (antiobesity) of UCP2 inhibits glucose-stimulated insulin secretion in β Figure 1: Mechanism by which UCP2 activation may lead to pancreatic rat islets and INS-1 -cells [36–38], which is obesity and type diabetes. UCP2 upregulation by nutrients (glucose, well explained by the theory of mild uncoupling because lipids, fatty acids, glutamine (protein-rich diet)) increases UCP2 the resulting decrease in ATP-levels decreases closure of mRNA transcription or translation. UCP2 activity is increased by the ATP/ADP sensitive potassium channels, and therefore superoxide radicals. Increased UCP2 amount or activity causes β- decreases insulin secretion (Figure 1). Concordant with this, cell dysfunction [29] and may contribute to decreased metabolic Ucp2(−/−) mice have increased glucose-stimulated insulin efficiency by decreasing ATP-generation. UCP2 activation decreases secretion and higher pancreatic islet ATP levels and are oxidative stress and may therefore decrease aging [30], cancer protected against glucose-toxicity in β-cells [29, 39], and on progression [31], and inflammation [32]. Decreased ROS levels may also protect β-cell [33]. Decreased ATP-generation by UCP2 a high-fat diet they show increased insulin secretion and ffi decreased plasma triglyceride concentrations [40]. This is in upregulation may cause less e cient metabolism and protect against obesity, which will decrease demand for insulin secretion line with in vitro studies of Ucp2(−/−) islets of Langerhans, by the β-cell. which are resistant to palmitate-induced β-cell dysfunction [41]. UCP2 mRNA is upregulated in obese ob/ob mice, and ob/ob mice lacking UCP2 showed restored first-phase insulin secretion and reduced level of hyperglycaemia [29]. No effect resolve this issue. Overexpression of UCP3 in skeletal muscle of Ucp2 gene disruption on obesity was observed, even or UCP3 together with UCP2 in skeletal muscle have been upon a high-fat diet or on a background of genetic obesity reported to create a lean phenotype in mice [48, 49]. It [29, 32]; however, short-term inhibition of Ucp2 using is uncertain whether these data are reliable as it has been antisense oligonucleotides ameliorated insulin resistance and shown that overexpression of mitochondrial carriers may improved insulin secretion in a diet-induced mouse model lead to over-load of the inner mitochondrial membrane [42]. Recently, β-cell function of Ucp2(−/−) on an in-bred and artifactual data [20]. Thus, over-expressing UCP2 or C57Bl background was reported to be opposite to earlier UCP3 in mice may not be a good or reliable strategy for published reports, in that β-cells without UCP2 showed interrogating their physiological function. lower glucose-stimulated insulin secretion, but maintained Interestingly, decreased ROS due to partial uncoupling by higher levels of reactive oxygen species [43]. The reason UCP2 or UCP3 could represent a link to the “thinness and for these contradictory results are at the moment unknown longevity” phenomenon observed when diet restriction of rodents increases their life span by up to 50% [50](Figure 1). but may be explained by yet unidentified modifier genes (−/−) different from the initial mixed C57Bl and 129 background In fact, Ucp2 mice, having increased oxidative stress in to congenic back-crossed strains [44]. Moreover, because their mitochondria, live significantly shorter than WT litter two different theories exist for explaining the role of UCP2 mates [30], supporting the hypothesis that mitochondrial- in β-cells, it is difficult to extrapolate from mouse data to derived free radicals are involved in aging [51]. Recently, data obtained in humans. The specific contribution between it was shown that UCP2 mRNA levels were increased in hypersecretion of insulin versus the deleterious effect of ROS colon cancer samples, also suggesting a link between levels on β-cells in humans versus mice is unknown (Figure 1). of oxidative stress modulated by UCP2 and development of Similar to UCP2, UCP3 has recently been found to be cancer [52](Figure 1). expressed in pancreatic β-cells, where it also influenced insulin secretion [45], but the physiological function of 3. UCPs: Candidate Genes for UCP3 in β-cells is not known. Obesity and Type 2 Diabetes Disruption of the UCP3 gene in mice does not cause an obese phenotype, but levels of oxidative stress are increased Because UCP2 and UCP3 decrease mitochondrial membrane in skeletal muscle of Ucp3(−/−) mice [46, 47]. There may be potential and mediate proton leak [53], they are candidate compensatory effects when removing UCP3 or UCP2 from genes for obesity and type 2 diabetes. UCP2 and UCP3 are skeletal muscle (or β-cells) in which they are coexpressed; coexpressed in skeletal muscle, which contributes the most to generation of UCP2 and UCP3 double knockout mice may the basal metabolic rate [54]. Mutations reducing the activity Journal of Obesity 3

Genomic organization of the UCP3-UCP2 region. Chr. 11: 73.72 73.71 73.7 73.69 Mbp

10 kb

UCP3 UCP2

ATG TGAS TGAL ATG TGA

−55 C/T Tyr99Tyr −866 G/A Ala55Val 3UTR insertion Figure 2: Diagram of the UCP2-UCP3 genomic region with indications of common genetic variation. Genomic organization of the UCP2- UCP3 region on chromosome 11. ATG: start codon, TGA: stop codon. Bent arrows indicate reported transcription start sites (from [34]). UCP3 protein exists in a short and a long form due to alternative polyadenylation sites, indicated by TGAS and TGAL [35].

or expression of either protein could theoretically diminish rs660339), and a 45 bp insertion-deletion polymorphism energy expenditure by an increase in coupling of oxidative in the 3untranslated region (UTR) of the UCP2 gene phosphorylation, and thereby contribute to development (3UTR ins/del). In UCP3, there is one common and of obesity. Mutations in UCP2 regulatory regions causing well-studied polymorphism: a promoter variant, −55 C/T increased levels could cause or worsen decreased glucose- (rs1800849) (Table 1)[63, 65–69]. stimulated insulin secretion directly through a decreased ATP/ADP ratio in the pancreatic β-cell and promote development of type 2 diabetes. 5. Effects of the −866G>AVarianton The most consistent trait found in Ucp2(−/−) and Transcriptional Activity of /− Ucp3(− ) mice has been the increased levels of superoxide the UCP2 Promoter radicals and oxidative stress. Insulin resistance may be caused by increased intracellular ROS levels [55], which are The −866G>A polymorphism is situated in the proxi- influenced by the expression or activity of UCPs [56]. UCP2 mal promoter of UCP2 and putatively changes one or may also modulate the severity of low-grade inflammation more transcription factor binding sites [60, 70]. Several present in obesity and obesity-associated type 2 diabetes, studies determined whether the activity of the promoter because ROS levels generated by macrophages and other changes with genotype. In insulin producing cells, the β- immune cells are increased in Ucp2(−/−) mice [32]. This cell transcription factor PAX6 binds preferentially to the A- also points to an important role of UCP2 in atherosclerosis, allele, which increases reporter-gene activity of constructs since Ucp2(−/−) mice fed an atherogenic diet developed containing the A-allele [70, 71]. Sesti et al. (2003) showed more atherosclerosis [57]. Similarly, oxidative stress may decreased glucose-stimulated insulin secretion from isolated be causative for late diabetic complications [58], and as human islets having the GA-genotype vs. the GG-genotype modulators of mitochondrial ROS levels, UCP2 and UCP3 [72], suggesting that increased UCP2 mRNA from the A- may affect the severity of diabetic complications. allele translates into increased UCP2 protein, induced proton leak, decreased ATP/ADP ratio, and decreased glucose- stimulated insulin secretion in accordance with the pheno- 4. Human UCP2 and UCP3 Genetic Variation type of the Ucp2(−/−) mice. In adipocytes, the −866 A-allele was associated with both decreased [73]orincreased[74] UCP2 and UCP3 are the likely result of an ancestral gene- levels of adipose tissue UCP2 mRNA. However, reporter- duplication, because they are situated close to each other gene constructs with the −866 A-allele showed increased on chromosome 11q13 [64](Figure 2). Because UCP2 and activity in adipocytes [70], similar to findings in insulin- UCP3 are considered candidate genes for development producing cells. Thus, the minor A-allele directs higher rates of obesity and type 2 diabetes, they have been studied of transcription from the UCP2 promoter compared with the extensively. There is a low number of frequent genetic G-allele. variants, which have been investigated in a large number of studies (Table 1 and Figure 2), and most identified variants have been of low frequency and have therefore not been 6. UCP2 Genetic Variation in Relation so intensively studied. There are 3 common polymorphisms to Obesity in UCP2, which are well studied: a promoter variant, −866G>A (rs659366), a missense polymorphism in codon The frequent −866G>A polymorphism (rs659366) has been 55 changing an alanine to a valine (Codon 55 Ala/Val, extensively investigated for association with obesity and 4 Journal of Obesity

Table 1: Studied high frequency variants of the UCP2 and UCP3 genes.

Gene Variant Acc. number Approximate frequency (ref) UCP2 Promoter −1957G>A rs649446 29.0% (A-allele) [59] UCP2 Promoter −866G>A rs659366 37.0% (A-allele) [60] UCP2 Codon 55 Ala/Val rs660339 39.6% (Val) [61] UCP2 3UTR ins>del — 29.6% (ins-allele) [62] UCP3 Promoter −55C>T rs1800849 26.9% (T-allele) [63] UCP3 Exon 3 Tyr99Tyr rs1800006 30.0% (T-allele) [59] UCP3 Exon 5 Tyr210Tyr rs2075577 16.0% (T-allele) [59]

related subphenotypes. The AA genotype was initially shown homozygous genotype was associated with increased BMI to associate with a reduced risk of obesity among 596 and in South Indian females and increased serum leptin levels 791 white Europeans [74]—an observation that has been in British women [95]. However, in Danish subjects there replicated [75], but more studies report either increased was no association with obesity or weight gain over a 26- prevalence of the A-allele in obesity [76–78] or no association year followup [62]. The 3UTR45bpinsertioncouldexert at all [59, 60, 72, 79–88](Table 2). The total number of its effect through altered mRNA stability; however, there was subjects in the studies reporting no association with obesity no difference in UCP2 mRNA levels between genotypes in for the A-allele is above 14000 and by far outnumbers skeletal muscle from Pima Indians [97], but in vitro mRNA the initial observation, and the number of participants in stability assays showed that the insertion allele had less stable the three studies reporting association of the A-allele with mRNA [74]. obesity or increasing indices of adiposity is approximately 4000. Therefore, it is most likely that the −866 A-allele has a very modest effect if any on development of obesity, but in 7. Type 2 Diabetes and the Metabolic Syndrome order to evaluate, this a proper meta-analysis is necessary. with Regard to UCP2 Genetic Variation Assuming that a more subtle intermediary obesity- related phenotype is affected by the −866G>A polymor- Mar Gonzalez-Barroso et al. (2008) reported on two families phism, a number of observations have been made; among in which congenital hyperinsulinemia occurred and who 681 French type 2 diabetic patients, the variant was carried heterozygous mutations in UCP2 [98]. The two associated with elevated triglyceride and total cholesterol families each carried their own mutations, which segregated concentrations and increased risk of dyslipidaemia [90], with the disease and which changed amino acids conserved and in line with this, decreased HDL-cholesterol levels were between species. Functional studies of recombinant yeast reported among 658 Korean women [59]. Lack of association showed lower proton leak of the mutant UCP2s, and the with lipid levels has also been reported [72, 79, 80, 82]. mutants were not able to suppress insulin secretion in β- Carriers of the G-allele of the −866G>A polymorphism lost cells when over-expressed as opposed to wild-type UCP2. more weight than A-homozygotes in a study of diet-induced Thus, the phenotype of carriers of heterozygous null-alleles body fat reduction in 301 Korean women undergoing a very- of UCP2 were in fact very similar to the phenotype of low-calorie programme [92]. Finally, in 296 obese children, Ucp2(−/−) mice on mixed-strain genetic background [29], homozygosity of the A-allele was related to increased resting- but opposite the phenotype of Ucp2(−/−) mice in congenic energy expenditure, increased glucose oxidation rate, and lines [43]. However, it is not known how the hyperinsulinism lower lipid oxidation rate [89], and among 185 Pima Indians, associated with UCP2 null-mutations affects β-cells later in the −866G>A polymorphism was associated with increased life; oxidative stress is increased in Ucp2(−/−) mice, and over 24-hour energy expenditure [83]. time this is associated with declining β-cell function. On the Numerous studies do not support a functional impact other hand, Ucp2(−/−) mice do not become diabetic [43]. of the 3UTR insertion or the Ala55Val polymorphism in Thus, studying adult and aging carriers of the identified causing obesity or type II diabetes. Few association studies UCP2 mutations is likely to be very rewarding for elucidating have found differences in allele or genotype frequencies of the contribution of UCP2 towards maintenance of glucose the Ala55Val polymorphism between obese and/or type 2 tolerance in humans. diabetic subjects and control subjects [61, 93, 94]andthis Given that UCP2 null mutations cause hyperinsulinemia, variant is generally not considered to predispose to obesity the −866 A-allele, having increased transcriptional activity, or type 2 diabetes. The 3UTR insertion polymorphism would be expected to show association with decreased β- has been related to measures of energy expenditure or cell function and ultimately with type 2 diabetes. When increased BMI [83, 95, 96]. In heterozygous state, the examining measures of insulin secretion, the −866 A-allele 3UTR insertion has been associated with increased sleeping was associated with decreased glucose-stimulated insulin metabolic rate and 24-h energy expenditure and lower BMI secretion among 137 Japanese type 2 diabetic patients in Pima Indians, in agreement with a role of UCP2 in undergoing frequently sampled IVGTT [71]andalsoin controlling energy expenditure [97]. Moreover, the insertion isolated pancreatic islets from nondiabetic subjects [72]. Journal of Obesity 5

Table 2: Summary of association studies of the UCP2 promoter −866G>A (rs659366) polymorphism in relation to obesity and related metabolic traits.

nobese ncontrol Ethnic (Frequency (Frequency Phenotypes Reference population of A-allele of A-allele in %) in %) 340 (46.5) 256 (52.2) Caucasian Common G-allele predisposed to obesity Esterbauer et al. 2001 [74] 109 (31.2) 589 (38.2) Caucasian 749 (39.6) 816 (40.7) Not associated with obesity or BMI within groups Dalgaard et al. 2003 [60] 122 (28.2) 374 (29.0) Caucasian Not associated with obesity or BMI within groups Mancini et al. 2003 [79] 76 (34.9) Caucasian — 302 (32.1) Not associated with BMI within group Sesti et al. 2003 [72] — 565 (32.4) Caucasian Not associated with BMI in control or diabetic patients D’Adamo et al. 2004 [80] — 483 (33.6) — 134 Japanese Not associated with BMI, but with hypertension Ji et al. 2004 [81] 342 Caucasian 296 (37.0) — A-allele associated with decreased lipid oxidation Le Fur et al. 2004 [89] — 327 (34.6) Caucasian Not associated with BMI within group Bulotta et al. 2005 [82] 746 (28.6) Pima 864 (54.0) — Not associated with BMI within group. AA genotype Kovacs et al. 2005 [83] Indians 263 (55.5) — increased 24 hr EE Not associated with BMI within group. Associated with Korean — 658 Cha et al. 2007 [59] decreased HDL-levels Caucasian — 598 Not associated with BMI within group. A-allele associated Gable et al. 2007 [84]P 653 with decreased W/H-ratio and lower fasting p-insulin Filipino — 1755 (29.7) Not associated with BMI within group Marvelle et al. 2008 [85] A-allele associated with obesity and associated with increased Caucasian 375 (41.3) 2316 (35.8) risk of CHD and systolic BP. AA genotype associated with Dhamrait et al. 2004 [77] increased oxidative stress AA genotype signiﬁcantly associated with obesity and insulin Caucasian 192 (38.3) 170 (38.2) Ochoa et al. 2007 [76] resistance in children Caucasian 225 (39.6) 294 (38.9) AA genotype associated with various indices of obesity Kring et al. 2008 [78] Caucasian 277 188 Not associated with early-onset obesity Schauble¨ et al. 2003 [86] Not associated with BMI in type 2 diabetic patients, but AA Caucasian — 681 (36.9) genotype associated with increased triglyceride and Reis et al. 2004 [90] cholesterol levels 3784 Not associated with obesity, allele-frequencies not given for Various — Hsu et al. 2008 [87] (35.4–46.7) obese subjects GG genotype associated with obesity in children but Korean — 1469 (∼48) Jun et al. 2009 [75] protective in adults AA genotype decreased total cholesterol and decreased Caucasian — 507 Salopuro et al. 2009 [88]P LDL-cholesterol. Not associated with BMI within group A-allele associated with obesity and hyperinsulinemia (in Indian 200 (42.0) 240 (32.2) Srivastava et al. 2010 [91] obese subjects) PDenotes prospective study. Abbreviations: CHD: coronary heart disease; BP: blood pressure; EE: energy expenditure; BMI: body mass index; HDL: high density lipoprotein; W/H: waist to hip.

These observations are in accordance with the A-allele [71, 99], but also with the G-allele [100], whereas early directing increased UCP2 expression and causing decreased requirement for insulin treatment was observed in A-allele insulin secretion (but also lower ROS-levels). Decreased carriers [71, 90](Table 3). basal insulin secretion was initially reported among A-allele Observations of a lower disposition index in −866A carriers [74] but was contrasted by subsequent studies [60, carriers have been made [70, 72], although this could also 72, 80–82], which showed no association. Also, early onset be induced by changes in insulin sensitivity rather than of type 2 diabetes has been correlated with the A-allele insulin secretory capacity. It is possible that the −866 A 6 Journal of Obesity

Table 3: Summary of association or prospective studies of the UCP2 promoter −866G>A (rs659366) polymorphism in relation to type 2 diabetes and intermediary phenotype.

ndiabetes ncontrol Ethnic (Allele (Allele Phenotypes Reference population frequency frequency in %) in %) Krempler et al. 2002 Caucasian 201 (41.2) 391 (32.5) A-allele associated with type 2 diabetes increased disposition index [70] AA genotype decreased insulin sensitivity and was associated with D’Adamo et al. 2004 Caucasian 565 (32.4) 483 (33.6) type 2 diabetes [80] AA genotype increased risk of type 2 diabetes, especially combined Gable et al. 2006 Caucasian — 2595 (37.0) with obesity [99]P A-allele associated with decreased insulin secretion. Isolated islets of Caucasian — 302 (28.8) Sesti et al. 2003 [72] A-allele carriers had decreased in vitro insulin secretion G-allele associated with type 2 diabetes and increased adipose tissue Wang et al. 2004 Caucasian 131 (33.0) 118 (48.0) mRNA [73] Bulotta et al. 2005 Caucasian 746 (28.6) 327 (34.5) G-allele associated with type 2 diabetes [82] Lyssenko et al. 2005 Caucasian — 2216 (38.1) GG genotype increased risk of type 2 diabetes [100]P Indian 762 (35.0) 924 (41.0) G-allele associated with type 2 diabetes Rai et al. 2007 [101] Cheurfa et al. 2008 Caucasian — 3122 (36.7) GG genotype increased risk of MI in men [102]P AA genotype borderline associated with increased fasting insulin Esterbauer et al. Caucasian — 589 (38.2) levels 2001 [74] 864 (54.0) — Not associated with type 2 diabetes within group. AA genotype Kovacs et al. 2005 Pima Indian [83] 263 (55.5) — borderline associated with decreased insulin sensitivity 2198 Various 1584 Not associated with type 2 diabetes Hsu et al. 2008 [87] (35.4–46.7) Not associated with type 2 diabetes, but A-allele showed higher Sasahara et al. 2004 Japanese 413 (47.2) 172 (43.1) transcriptional activity and carriers had decreased AIR [71] — 235 (43.2) No association with changes fasting p-glucose or s-insulin in Dalgaard et al. 2003 Caucasian [60] 410 (34.5) glucose-tolerant subjects AA genotype decreased total cholesterol and decreased Salopuro et al. 2009 Caucasian — 507 LDL-cholesterol. [88]P Le Fur et al. 2004 Caucasian — 296 (37.0) No influence on insulin sensitivity [89] A-allele associated with increased type 2 diabetes risk, increased risk Dhamrait et al. 2004 Caucasian 375 (41.3) 2316 (35.8) of CAD and systolic BP, and increased oxidative stress [77] Palmer et al. 2009 Various — 901 (39.4) Diabetic A-allele carriers poor survival after MI [103]P 453 Stephens et al. 2008 Caucasian AA genotype associated with increased oxidative stress and CAD (33.0–36.0) [104] Caucasian — 227 (39.3) Diab. neuropathy lower in AA genotype Rudofsky et al. [105] GG genotype associated with low-grade inflammation, but not Labayen et al. 2009 Caucasian — 280 (39.3) insulin levels [106] Lapice et al. 2010 Caucasian — 383 (31.9) GG genotype associated with increased CRP [107] PDenotes prospective study. Disposition index: the product of Si and AIR. Abbreviations: Si: insulin sensitivity; AIR: acute insulin response; MI: myocardial infarct; LDL: low density lipoprotein; CRP: C-reactive protein; CAD: coronary artery disease. allele is involved in mediating decreased β-cell function that ROS-levels would be lower in A-carriers. However, since as well as decreased insulin sensitivity of adipose tissue, insulin resistance is associated with increases in oxidative which would be expected to translate into an increased risk stress [55], it is more likely that changes in disposition of type 2 diabetes. As the −866 A-allele was reported to index are due to differences in insulin secretion rather increase UCP2 mRNA expression [70, 71], it is expected than insulin resistance. In line with this, insulin resistance Journal of Obesity 7

(HOMA-IR) has been reported to be positively correlated patients. In another study of 280 children and adolescents with visceral adipose tissue UCP2 mRNA expression [80]. CRP was unaltered, but fibrinogen, complement C3 and Following the “mild uncoupling theory” it would be expected C4 were lower in AA-carriers [106]. Finally, Rudofsky et al. that increased UCP2 expression—as a possible consequence (2006, 2007) showed increased prevalence of the G-allele in of carrying the −866A-allele—would be associated with type 1 diabetic patients, whereas there was no association increased insulin sensitivity. However, experimental studies with microvascular complications [105, 108]. do not agree on the effect of −866G>A on insulin sensitivity. Using either hyperinsulinaemic-euglycaemic clamp or an intravenous glucose tolerance test in 39, 263, and 181 9. Possible Influence of Other SNPs in subjects, respectively, AA genotype carriers were less insulin the UCP2-UCP3 Genomic Region sensitive [70, 80, 83], whereas in a number of other studies The genomic region containing the UCP2 and UCP3 genes insulin resistance estimated using the HOMA index in 632 − > Japanese subjects [81], 363 French adolescents [76], and 302 were investigated for a total of 14 SNPs (including 866G A) ff − > spanning the UCP2 and UCP3 loci among 3,782 women Italian subjects [72]wasnota ected by the UCP2 866G A ff variant (Table 3). Clearly, more information is needed on of di erent ethnicities [87]. No single-SNP association the physiological effects of UCP2 on whole body insulin with type 2 diabetes was observed following correction sensitivity. for multiple testing; yet, haplotype analysis indicated an association with increased type 2 diabetes risk among 968 Association studies of type 2 diabetes have reported ff association of the −866A-allele with increased risk of type Caucasian women, and this e ect was further accentuated 2 diabetes in studies representing up to 1640 subjects [70, by overweight although no direct association with BMI was observed. The four-SNP haplotype in question was 77, 80, 99], whereas other studies report association of the − G-allele with type 2 diabetes backed by studies of more in high LD with the 866 A-allele, suggesting that as yet than 2700 subjects [73, 82, 100, 101], and a number of unidentified variation covered by the haplotype-spanned large studies report no association of this variant with type area may be responsible for the observed relationships of −866G>A with metabolic variables. The presence of other 2diabetes[83, 87, 90](Table 3). Prospective studies have ff shown that subjects carrying the AA genotype were more functional variants may also account for the di erence in likely to become type 2 diabetic, or had poor survival diabetes or obesity risk-allele reported by a number of studies following myocardial infarction [77, 99, 103], but the G- (Tables 2 and 3). allele has also been associated with increased risk of type 2 diabetes [100]. Thus, it is necessary to perform more studies 10. Conclusions and Perspectives as well as a proper meta-analysis to investigate the impact of this variant on type 2 diabetes. In acute studies using antisense oligonucleotides, UCP2 was involved in both insulin secretion and insulin action [42], whereas Ucp2(−/−) mice have not been reported to have 8. A Possible Role of −866G>A Variant and altered insulin sensitivity [29]. Studies of Ucp2(−/−) mouse Oxidative Stress in Cardiovascular Disease embryonic fibroblasts have shown that loss of Ucp2 results and Late Diabetic Complications in increased glycolysis and decreased fatty acid oxidation— suggesting that UCP2 regulates mitochondrial substrate Both increased risk of hypertension [81] as well as decreased usage to a greater extent than its original role as an uncoupler risk of dying following myocardial infarction [102, 103]has of respiratory chain activity from ATP synthesis [27, 28]. been reported to be associated with the −866A allele, whereas Absence of UCP2 causes oxidative stress and superoxide pro- plasma total antioxidant status, which is low when oxidative duction [32, 39], which is associated with insulin resistance stress is increased, has been shown to be decreased in AA [55]. However, a number of studies report association of the genotype carriers. Among 2,695 healthy Caucasian men, the high-expressing allele of the −866G/A variant with oxidative risk of coronary heart disease and elevated diastolic blood stress, which is at odds with phenotype data from Ucp2(−/−) pressure was increased in men homozygous for the −866A- mice. However, the widespread expression pattern makes allele while among 465 diabetic men, the A-allele was asso- possible a dual function in obesity (energy metabolism) and ciated with increased oxidative stress [77]—an observation type 2 diabetes (glucose metabolism). that was significantly accentuated by cigarette smoking [104]. With so many contrasting studies there is, a genuine need Thus, the functional A-allele, which mediates increased for a thorough meta-analysis of the impact of the −866G>A UCP2 mRNA levels, is associated with increased oxidative polymorphism in order to conclude whether it predisposes stress. This may be linked with the poor insulin secretion to obesity and/or type 2 diabetes. It is important to note associated with the AA-genotype, leading to increased levels that genome-wide association studies (GWAS) have not of plasma glucose and HbA1c [71], and perhaps oxidative identified SNPs in the UCP2-UCP3 locus as being associated stress; however, this mechanism is speculative and needs with obesity or type 2 diabetes [109, 110]. However, if the experimental validation. Also, low-grade inflammation has mechanism of action of the −866G>A SNP, as some studies been investigated in the context of the −866G>A poly- indicate, occurs predominantly in already obese and type morphism, where increased C-reactive protein (CRP) was 2 diabetic subjects to increase late-diabetic complications, associated with the GG-genotype in a study of 283 diabetic such as cardiovascular disease via changes in oxidative stress 8 Journal of Obesity levels [77, 103–105], then this polymorphism is unlikely to nervous system of humans and rodents, and respiration be identified through a GWAS strategy looking primarily at uncoupling activity in recombinant yeast,” Journal of Biologi- obesity or type 2 diabetes. Furthermore, early disease onset cal Chemistry, vol. 273, no. 51, pp. 34611–34615, 1998. and a more frequent requirement for insulin may be related [6] X. X. Yu, W. Mao, A. Zhong et al., “Characterization of to a reduced capacity of insulin secretion. It may well be that novel UCP5/BMCP1 isoforms and differential regulation of the major contribution of genetic variability in UCP2 lies in UCP4 and UCP5 expression through dietary or temperature mediating susceptibility towards complications. manipulation,” FASEB Journal, vol. 14, no. 11, pp. 1611– A main conclusion is that variation in the uncoupling 1618, 2000. protein 2 gene is not associated with major alterations of [7]W.Mao,X.X.Yu,A.Zhongetal.,“UCP4,anovel body weight or risk of type 2 diabetes. It is naturally more brain-specific mitochondrial protein that reduces membrane ffi potential in mammalian cells,” FEBS Letters, vol. 443, no. 3, di cult to estimate the contribution of UCP genes towards pp. 326–330, 1999. polygenic obesity and type 2 diabetes; however, although − > [8] V. Golozoubova, B. Cannon, and J. Nedergaard, “UCP1 is many studies do indicate association of the 866G Avariant essential for adaptive adrenergic nonshivering thermogene- with obesity and/or type 2 diabetes, the impact of this sis,” American Journal of Physiology, vol. 291, no. 2, pp. E350– single variant is low, as is the case with most predisposing E357, 2006. variants on polygenic traits. Therefore, large numbers of [9] V. Golozoubova, E. Hohtola, A. Matthias, A. Jacobsson, well-characterised study subjects must be investigated to B. Cannon, and J. Nedergaard, “Only UCP1 can mediate detectthetrueeffect of a given variant. adaptive nonshivering thermogenesis in the cold,” The FASEB Journal, vol. 15, no. 11, pp. 2048–2050, 2001. Sources [10]A.Vidal-Puig,G.Solanes,D.Grujic,J.S.Flier,andB.B. Lowell, “UCP3: an uncoupling protein homologue expressed Google Scholar and PubMed were searched for publications preferentially and abundantly in skeletal muscle and brown in English containing the words “uncoupling protein 2”, adipose tissue,” Biochemical and Biophysical Research Com- “uncoupling protein 3”, “−866G/A”, “−55C/T”, “rs659366”, munications, vol. 235, no. 1, pp. 79–82, 1997. “rs660339” “polymorphism”, “UCP2”, “UCP3”, “GWAS”, [11] S. Samec, J. Seydoux, and A. G. Dulloo, “Post-starvation “SNP”, and “proton leak”, alone or in combinations with gene expression of skeletal muscle uncoupling protein 2 “obesity” and “diabetes”. and uncoupling protein 3 in response to dietary fat levels and fatty acid composition: a link with insulin resistance,” Diabetes, vol. 48, no. 2, pp. 436–441, 1999. Abbreviations [12] S. Samec, J. Seydoux, and A. G. Dulloo, “Skeletal muscle UCP3 and UCP2 gene expression in response to inhibition UCP2: Uncoupling protein 2 of free fatty acid flux through mitochondrial β-oxidation,” UCP3: Uncoupling protein 3 Pflugers Archiv European Journal of Physiology, vol. 438, no. 4, SNP: Single-nucleotide polymorphism pp. 452–457, 1999. GWAS:Genomewideassociationstudy [13]H.Kageyama,A.Suga,M.Kashibaetal.,“Increaseduncou- ROS: Reactive oxygen species. pling protein-2 and -3 gene expressions in skeletal muscle of STZ-induced diabetic rats,” FEBS Letters, vol. 440, no. 3, pp. Acknowledgment 450–453, 1998. [14] J. E. P. Brown, S. Thomas, J. E. Digby, and S. J. Dunmore, Work in L.T. Dalgaards laboratory was supported by a “Glucose induces and leptin decreases expression of uncou- grant from the Danish Research Council for Technology and pling protein-2 mRNA in human islets,” FEBS Letters,vol. Production. 513, no. 2-3, pp. 189–192, 2002. [15]D.W.Gong,Y.He,M.Karas,andM.Reitman,“Uncoupling protein-3 is a mediator of thermogenesis regulated by thyroid References hormone, β3-adrenergic agonists, and leptin,” Journal of Biological Chemistry, vol. 272, no. 39, pp. 24129–24132, 1997. [1] R. E. Gimeno, M. Dembski, X. Weng et al., “Cloning and characterization of an uncoupling protein homolog: [16] J. Matsuda, K. Hosoda, H. Itoh et al., “Cloning of rat a potential molecular mediator of human thermogenesis,” uncoupling protein-3 and uncoupling protein-2 cDNAs: Diabetes, vol. 46, no. 5, pp. 900–906, 1997. their gene expression in rats fed high-fat diet,” FEBS Letters, [2] C. Fleury, M. Neverova, S. Collins et al., “Uncoupling vol. 418, no. 1-2, pp. 200–204, 1997. protein-2: a novel gene linked to obesity and hyperinsuline- [17] O. Boss, S. Samec, F. Kuhne¨ et al., “Uncoupling protein-3 mia,” Nature Genetics, vol. 15, no. 3, pp. 269–272, 1997. expression in rodent skeletal muscle is modulated by food [3] O. Boss, S. Samec, A. Paoloni-Giacobino et al., “Uncoupling intake but not by changes in environmental temperature,” protein-3: a new member of the mitochondrial carrier family Journal of Biological Chemistry, vol. 273, no. 1, pp. 5–8, 1998. with tissue-specific expression,” FEBS Letters, vol. 408, no. 1, [18] M. D. Brand and T. C. Esteves, “Physiological functions of the pp. 39–42, 1997. mitochondrial uncoupling proteins UCP2 and UCP3,” Cell [4] B. B. Lowell, “Uncoupling protein-3 (UCP3): a mitochon- Metabolism, vol. 2, no. 2, pp. 85–93, 2005. drial carrier in search of a function,” International Journal of [19] S. Cadenas, J. A. Buckingham, S. Samec et al., “UCP2 and Obesity, vol. 23, supplement 6, pp. S43–S45, 1999. UCP3 rise in starved rat skeletal muscle but mitochondrial [5]D.Sanchis,C.Fleury,N.Chomikietal.,“BMCP1,anovel proton conductance is unchanged,” FEBS Letters, vol. 462, no. mitochondrial carrier with high expression in the central 3, pp. 257–260, 1999. Journal of Obesity 9

[20] S. Cadenas, K. S. Echtay, J. A. Harper et al., “The basal proton [35]G.Solanes,A.Vidal-Puig,D.Grujic,J.S.Flier,andB.B. conductance of skeletal muscle mitochondria from trans- Lowell, “The human uncoupling protein-3 gene. Genomic genic mice overexpressing or lacking uncoupling protein-3,” structure, chromosomal localization, and genetic basis for JournalofBiologicalChemistry, vol. 277, no. 4, pp. 2773– short and long form transcripts,” Journal of Biological 2778, 2002. Chemistry, vol. 272, no. 41, pp. 25433–25436, 1997. [21] K. S. Echtay, D. Roussel, J. St-Plerre et al., “Superoxide [36]C.B.Chan,P.E.MacDonald,M.C.Saleh,D.C.Johns,E. activates mitochondrial uncoupling proteins,” Nature,vol. Marban,` and M. B. Wheeler, “Overexpression of uncoupling 415, no. 6867, pp. 96–99, 2002. protein 2 inhibits glucose-stimulated insulin secretion from [22] K. S. Echtay, E. Winkler, K. Frischmuth, and M. Klingenberg, rat islets,” Diabetes, vol. 48, no. 7, pp. 1482–1486, 1999. “Uncoupling proteins 2 and 3 are highly active H(+) [37] C. B. Chan, “Endogenous regulation of insulin secretion by transporters and highly nucleotide sensitive when activated UCP2,” Clinical Laboratory, vol. 48, no. 11-12, pp. 599–604, by coenzyme Q (ubiquinone),” Proceedings of the National 2002. Academy of Sciences of the United States of America, vol. 98, [38]C.B.Chan,D.DeLeo,J.W.Josephetal.,“Increased no. 4, pp. 1416–1421, 2001. uncoupling protein-2 levels in β-cells are associated with [23] M. Jaburek,M.Va˚ recha,ˇ R. E. Gimeno et al., “Transport func- impaired glucose-stimulated insulin secretion: mechanism of tion and regulation of mitochondrial uncoupling proteins 2 action,” Diabetes, vol. 50, no. 6, pp. 1302–1310, 2001. and 3,” JournalofBiologicalChemistry, vol. 274, no. 37, pp. [39] S. Krauss, C. Y. Zhang, L. Scorrano et al., “Superoxide- 26003–26007, 1999. mediated activation of uncoupling protein 2 causes pancre- [24]K.S.Echtay,T.C.Esteves,J.L.Pakayetal.,“Asignalling atic β cell dysfunction,” Journal of Clinical Investigation,vol. role for 4-hydroxy-2-nonenal in regulation of mitochondrial 112, no. 12, pp. 1831–1842, 2003. uncoupling,” EMBO Journal, vol. 22, no. 16, pp. 4103–4110, [40]J.W.Joseph,V.Koshkin,C.Y.Zhangetal.,“Uncoupling 2003. protein 2 knockout mice have enhanced insulin secretory [25]M.P.Murphy,K.S.Echtay,F.H.Blaikieetal.,“Superoxide capacity after a high-fat diet,” Diabetes, vol. 51, no. 11, pp. activates uncoupling proteins by generating carbon-centered 3211–3219, 2002. radicals and initiating lipid peroxidation: studies using a α [41]J.W.Joseph,V.Koshkin,M.C.Salehetal.,“Freefattyacid- mitochondria-targeted spin trap derived from -phenyl-N- induced β-cell defects are dependent on uncoupling protein tert-butylnitrone,” JournalofBiologicalChemistry, vol. 278, 2 expression,” JournalofBiologicalChemistry, vol. 279, no. 49, no. 49, pp. 48534–48545, 2003. pp. 51049–51056, 2004. [26] T. C. Esteves and M. D. Brand, “The reactions catalysed by [42] C. T. De Souza, E. P. Araujo,´ L. F. Stoppiglia et al., “Inhibition the mitochondrial uncoupling proteins UCP2 and UCP3,” of UCP2 expression reverses diet-induced diabetes mellitus Biochimica et Biophysica Acta, vol. 1709, no. 1, pp. 35–44, by effects on both insulin secretion and action,” FASEB 2005. Journal, vol. 21, no. 4, pp. 1153–1163, 2007. [27] C. Pecqueur, C. Alves-Guerra, D. Ricquier, and F. Bouillaud, [43]J.Pi,Y.Bai,K.W.Danieletal.,“Persistentoxidativestressdue “UCP2, a metabolic sensor coupling glucose oxidation to to absence of uncoupling protein 2 associated with impaired mitochondrial metabolism?” IUBMB Life,vol.61,no.7,pp. pancreatic β-cell function,” Endocrinology, vol. 150, no. 7, pp. 762–767, 2009. 3040–3048, 2009. [28] C. Pecqueur, T. Bui, C. Gelly et al., “Uncoupling protein- [44] R. N. Kulkarni, “Uncoupling modifier genes from uncou- 2 controls proliferation by promoting fatty acid oxidation β and limiting glycolysis-derived pyruvate utilization,” FASEB pling protein 2 in pancreatic -cells,” Endocrinology, vol. 150, Journal, vol. 22, no. 1, pp. 9–18, 2008. no. 7, pp. 2994–2996, 2009. [29] C. Y. Zhang, G. Baffy, P. Perret et al., “Uncoupling protein- [45] Y. Li, K. Maedler, and L. Haataja, “UCP-2 and UCP-3 ff 2 negatively regulates insulin secretion and is a major link proteins are di erentially regulated in pancreatic beta-cells,” between obesity, β cell dysfunction, and type 2 diabetes,” Cell, PLoS ONE, vol. 3, no. 1, Article ID e1397, 2008. vol. 105, no. 6, pp. 745–755, 2001. [46] A. J. Vidal-Puig, D. Grujic, C. Y. Zhang et al., “Energy [30] Z. B. Andrews and T. L. Horvath, “Uncoupling protein-2 metabolism in uncoupling protein 3 gene knockout mice,” regulateslifespaninmice,”American Journal of Physiology, JournalofBiologicalChemistry, vol. 275, no. 21, pp. 16258– vol. 296, no. 4, pp. E621–E627, 2009. 16266, 2000. [31]Z.Derdak,N.M.Mark,G.Beldi,S.C.Robson,J.R.Wands, [47]D.W.Gong,S.Monemdjou,O.Gavrilovaetal.,“Lackof and G. Baffy, “The mitochondrial uncoupling protein-2 obesity and normal response to fasting and thyroid hormone promotes chemoresistance in cancer cells,” Cancer Research, in mice lacking uncoupling protein-3,” Journal of Biological vol. 68, no. 8, pp. 2813–2819, 2008. Chemistry, vol. 275, no. 21, pp. 16251–16257, 2000. [32] D. Arsenijevic, H. Onuma, C. Pecqueur et al., “Disruption [48] J. C. Clapham, J. R. S. Arch, H. Chapman et al., “Mice overex- of the uncoupling protein-2 gene in mice reveals a role in pressing human uncoupling protein-3 in skeletal muscle are immunity and reactive oxygen species production,” Nature hyperphagic and lean,” Nature, vol. 406, no. 6794, pp. 415– Genetics, vol. 26, no. 4, pp. 435–439, 2000. 418, 2000. [33] J. Pi and S. Collins, “Reactive oxygen species and uncoupling [49] T. L. Horvath, S. Diano, S. Miyamoto et al., “Uncoupling protein 2 in pancreatic β-cell function,” Diabetes, Obesity and proteins-2 and 3 influence obesity and inflammation in Metabolism, vol. 12, supplement 2, pp. 141–148, 2010. transgenic mice,” International Journal of Obesity,vol.27,no. [34]H.Esterbauer,H.Oberkofler,F.Krempler,A.D.Strosberg, 4, pp. 433–442, 2003. and W. Patsch, “The uncoupling protein-3 gene is tran- [50] W. G. Sheldon, T. J. Bucci, R. W. Hart, and A. Turturro, “Age- scribed from tissue-specific promoters in humans but not in related neoplasia in a lifetime study of ad libitum-fed and rodents,” JournalofBiologicalChemistry, vol. 275, no. 46, pp. food- restricted B6C3F1 mice,” Toxicologic Pathology,vol.23, 36394–36399, 2000. no. 4, pp. 458–476, 1995. 10 Journal of Obesity

[51] M. D. Brand, J. A. Buckingham, T. C. Esteves et al., [65] L. T. Dalgaard, T. Hansen, S. A. Urhammer, T. Drivsholm, K. “Mitochondrial superoxide and aging: uncoupling-protein Borch-Johnsen, and O. Pedersen, “The uncoupling protein activity and superoxide production,” Biochemical Society 3-55 C → T variant is not associated with type II diabetes Symposium, vol. 71, pp. 203–213, 2004. mellitus in Danish subjects,” Diabetologia,vol.44,no.8,pp. [52]M.Horimoto,M.B.Resnick,T.A.Konkin,J.Routhier,J.R. 1065–1067, 2001. Wands, and G. Baffy, “Expression of uncoupling protein-2 in [66] P. G. Cassell, P. J. Saker, S. J. Huxtable et al., “Evidence that human colon cancer,” Clinical Cancer Research,vol.10,no. single nucleotide polymorphism in the uncoupling protein 18, part 1, pp. 6203–6207, 2004. 3 [UCP3] gene influences fat distribution in women of [53]M.A.Paulik,R.G.Buckholz,M.E.Lancasteretal., European and Asian origin,” Diabetologia, vol. 43, no. 12, pp. “Development of infrared imaging to measure thermogenesis 1558–1564, 2000. in cell culture: thermogenic effects of uncoupling protein-2, [67] T. Berentzen, L. T. Dalgaard, L. Petersen, O. Pedersen, and troglitazone, and β-adrenoceptor agonists,” Pharmaceutical T. I. A. Sørensen, “Interactions between physical activity and Research, vol. 15, no. 6, pp. 944–949, 1998. variants of the genes encoding uncoupling proteins -2 and [54] D. F. S. Rolfe and G. C. Brown, “Cellular energy utilization -3 in relation to body weight changes during a 10-y follow- and molecular origin of standard metabolic rate in mam- up,” International Journal of Obesity, vol. 29, no. 1, pp. 93–99, mals,” Physiological Reviews, vol. 77, no. 3, pp. 731–758, 1997. 2005. [55] N. Houstis, E. D. Rosen, and E. S. Lander, “Reactive oxygen [68] A. M. Brown, J. W. Dolan, S. M. Willi, W. T. Garvey, species have a causal role in multiple forms of insulin and G. Argyropoulos, “Endogenous mutations in human resistance,” Nature, vol. 440, no. 7086, pp. 944–948, 2006. uncoupling protein 3 alter its functional properties,” FEBS [56] E. Chevillotte, M. Giralt, B. Miroux, D. Ricquier, and F. Letters, vol. 464, no. 3, pp. 189–193, 1999. Villarroya, “Uncoupling protein-2 controls adiponectin gene [69] A. M. Brown, S. M. Willi, G. Argyropoulos, and W. T. expression in adipose tissue through the modulation of Garvey, “A novel missense mutation, R70W, in the human reactive oxygen species production,” Diabetes,vol.56,no.4, uncoupling protein 3 gene in a family with type 2 diabetes,” pp. 1042–1050, 2007. Human Mutation, vol. 13, p. 506, 1999. [57]J.Blanc,M.C.Alves-Guerra,B.Espositoetal.,“Protective [70] F. Krempler, H. Esterbauer, R. Weitgasser et al., “A functional role of uncoupling protein 2 in atherosclerosis,” Circulation, polymorphism in the promoter of UCP2 enhances obesity vol. 107, no. 3, pp. 388–390, 2003. risk but reduces type 2 diabetes risk in obese middle-aged [58] T. Nishikawa, D. Edelstein, X. L. Du et al., “Normalizing humans,” Diabetes, vol. 51, no. 11, pp. 3331–3335, 2002. mitochondrial superoxide production blocks three pathways [71] M. Sasahara, M. Nishi, H. Kawashima et al., “Uncoupling of hyperglycaemic damage,” Nature, vol. 404, no. 6779, pp. protein 2 promoter polymorphism -866G/A affects its 787–790, 2000. expression in beta-cells and modulates clinical profiles of [59] M. H. Cha, I. C. Kim, K. S. Kim, B. K. Kang, S. M. Choi, Japanese type 2 diabetic patients,” Diabetes,vol.53,no.2,pp. and Y. Yoon, “Association of UCP2 and UCP3 gene poly- 482–485, 2004. morphisms with serum high-density lipoprotein cholesterol [72]G.Sesti,M.Cardellini,M.A.Marinietal.,“Acommon among Korean women,” Metabolism, vol. 56, no. 6, pp. 806– polymorphism in the promoter of UCP2 contributes to the 813, 2007. variation in insulin secretion in glucose-tolerant subjects,” [60] L. T. Dalgaard, G. Andersen, L. H. Larsen et al., “Mutational Diabetes, vol. 52, no. 5, pp. 1280–1283, 2003. analysis of the UCP2 core promoter and relationships of [73]H.Wang,W.S.Chu,T.Lu,S.J.Hasstedt,P.A.Kern,andS. variants with obesity,” Obesity Research, vol. 11, no. 11, pp. C. Elbein, “Uncoupling protein-2 polymorphisms in type 2 1420–1427, 2003. diabetes, obesity, and insulin secretion,” American Journal of [61] S. A. Urhammer, L. T. Dalgaard, T. I. A. Sørensen et al., Physiology, vol. 286, no. 1, pp. E1–E7, 2004. “Mutational analysis of the coding region of the uncoupling [74] H. Esterbauer, C. Schneitler, H. Oberkofler et al., “A common protein 2 gene in obese NIDDM patients: impact of a com- polymorphism in the promoter of UCP2 is associated with mon amino acid polymorphism on juvenile and maturity decreased risk of obesity in middle-aged humans,” Nature onset forms of obesity and insulin resistance,” Diabetologia, Genetics, vol. 28, no. 2, pp. 178–183, 2001. vol. 40, no. 10, pp. 1227–1230, 1997. [75] H. S. Jun, I. K. Kim, H. J. Lee et al., “Effects of UCP2 and [62] L. T. Dalgaard, T. I. A. Sørensen, T. Andersen, T. Hansen, UCP3 variants on the manifestation of overweight in Korean and O. Pedersen, “An untranslated insertion variant in the children,” Obesity, vol. 17, no. 2, pp. 355–362, 2009. uncoupling protein 2 gene is not related to body mass index [76] M. C. Ochoa, J. L. Santos, C. Azcona et al., “Association and changes in body weight during a 26-year follow-up in between obesity and insulin resistance with UCP2-UCP3 Danish Caucasian men,” Diabetologia, vol. 42, no. 12, pp. gene variants in Spanish children and adolescents,” Molecular 1413–1416, 1999. Genetics and Metabolism, vol. 92, no. 4, pp. 351–358, 2007. [63] L. T. Dalgaard, T. I. A. Sørensen, T. Drivsholm et al., “A [77]S.S.Dhamrait,J.W.Stephens,J.A.Cooperetal.,“FT prevalent polymorphism in the promoter of the UCP3 gene Cardiovascular risk in healthy men and markers of oxidative and its relationship to body mass index and long term body stress in diabetic men are associated with common variation weight change in the Danish population,” Journal of Clinical in the gene for uncoupling protein 2,” European Heart Endocrinology and Metabolism, vol. 86, no. 3, pp. 1398–1402, Journal, vol. 25, no. 6, pp. 468–475, 2004. 2001. [78] S. I. Kring, L. H. Larsen, C. Holst et al., “Genotype- [64] C. Pecqueur, A. M. Cassard-Doulcier, S. Raimbault et al., phenotype associations in obesity dependent on definition of “Functional organization of the human uncoupling protein- the obesity phenotype,” Obesity Facts, vol. 1, no. 3, pp. 138– 2 gene, and juxtaposition to the uncoupling protein-3 gene,” 145, 2008. Biochemical and Biophysical Research Communications,vol. [79] F. P. Mancini, L. Sabatino, V. Colantuoni et al., “Vari- 255, no. 1, pp. 40–46, 1999. ants of uncoupling protein-2 gene and obesity: interaction Journal of Obesity 11

with peroxisome proliferator-activated receptorγ2,” Clinical obesity and Type II diabetes,” Diabetologia, vol. 44, no. 8, pp. Endocrinology, vol. 59, no. 6, pp. 817–822, 2003. 946–965, 2001. [80] M. D’Adamo, L. Perego, M. Cardellini et al., “The -866A/A [94] J.J.Jia,X. Zhang,C. R. Ge, andM.Jois,“The polymorphisms genotype in the promoter of the human uncoupling protein of UCP2 and UCP3 genes associated with fat metabolism, 2 gene is associated with insulin resistance and increased risk obesity and diabetes: etiology and pathophysiology,” Obesity of type 2 diabetes,” Diabetes, vol. 53, no. 7, pp. 1905–1910, Reviews, vol. 10, no. 5, pp. 519–526, 2009. 2004. [95] P. G. Cassell, M. Neverova, S. Janmohamed et al., “An [81]Q.Ji,H.Ikegami,T.Fujisawaetal.,“Acommonpoly- uncoupling protein 2 gene variant is associated with a raised morphism of uncoupling protein 2 gene is associated with body mass index but not Type II diabetes,” Diabetologia,vol. hypertension,” Journal of Hypertension,vol.22,no.1,pp.97– 42, no. 6, pp. 688–692, 1999. 102, 2004. [96] B. Buemann, B. Schierning, S. Toubro et al., “The association [82]A.Bulotta,O.Ludovico,A.Cocoetal.,“Thecommon- between the val/ala-55 polymorphism of the uncoupling 866G/A polymorphism in the promoter region of the UCP- protein 2 gene and exercise efficiency,” International Journal 2 gene is associated with reduced risk of type 2 diabetes in of Obesity, vol. 25, no. 4, pp. 467–471, 2001. Caucasians from Italy,” Journal of Clinical Endocrinology and [97]K.Walder,R.A.Norman,R.L.Hansonetal.,“Association Metabolism, vol. 90, no. 2, pp. 1176–1180, 2005. between uncoupling protein polymorphisms (UCP2-UCP3) [83] P. Kovacs, L. Ma, R. L. Hanson et al., “Genetic variation and energy metabolism/obesity in Pima Indians,” Human in UCP2 (uncoupling protein-2) is associated with energy Molecular Genetics, vol. 7, no. 9, pp. 1431–1435, 1998. metabolism in Pima Indians,” Diabetologia, vol. 48, no. 11, [98] M. M. Gonzalez-Barroso,´ I. Giurgea, F. Bouillaud et al., pp. 2292–2295, 2005. “Mutations in UCP2 in congenital hyperinsulinism reveal a [84] D. R. Gable, J. W. Stephens, S. S. Dhamrait, E. Hawe, and role for regulation of insulin secretion,” PLoS ONE,vol.3, S. E. Humphries, “European differences in the association no. 12, Article ID e3850, 2008. between the UCP2 -866G > A common gene variant and [99]D.R.Gable,J.W.Stephens,J.A.Cooper,G.J.Miller, markers of body mass and fasting plasma insulin,” Diabetes, andS.E.Humphries,“VariationintheUCP2-UCP3gene Obesity and Metabolism, vol. 9, no. 1, pp. 130–131, 2007. cluster predicts the development of type 2 diabetes in healthy [85]A.F.Marvelle,L.A.Lange,L.Qin,L.S.Adair,andK.L. middle-aged men,” Diabetes, vol. 55, no. 5, pp. 1504–1511, Mohlke, “Association of FTO with obesity-related traits in 2006. the cebu longitudinal health and nutrition survey (CLHNS) [100] V. Lyssenko, P. Almgren, D. Anevski et al., “Genetic predic- cohort,” Diabetes, vol. 57, no. 7, pp. 1987–1991, 2008. tion of future type 2 diabetes,” PLoS Medicine,vol.2,no.12, [86] N. Schauble,¨ F. Geller, W. Siegfried et al., “No evidence Article ID e345, 2005. for involvement of the promoter polymorphism -866 G/A [101] E. Rai, S. Sharma, A. Koul, A. K. Bhat, A. J. S. Bhanwer, of the UCP2 gene in childhood-onset obesity in humans,” and R. N. K. Bamezai, “Interaction between the UCP2- Experimental and Clinical Endocrinology and Diabetes,vol. 866G/A, mtDNA 10398G/A and PGC1α p.Thr394Thr and 111, no. 2, pp. 73–76, 2003. p.Gly482Ser polymorphisms in type 2 diabetes susceptibility [87] Y. H. Hsu, T. Niu, Y. Song, L. Tinker, L. H. Kuller, and S. Liu, in North Indian population,” Human Genetics, vol. 122, no. “Genetic variants in the UCP2-UCP3 gene cluster and risk of 5, pp. 535–540, 2007. diabetes in the women’s health initiative observational study,” [102] N. Cheurfa, D. Dubois-Laforgue, D. A. F. Ferrarezi et al., Diabetes, vol. 57, no. 4, pp. 1101–1107, 2008. “The common -866G > A variant in the promoter of UCP2 [88] T. Salopuro, L. Pulkkinen, J. Lindstrom¨ et al., “Variation in is associated with decreased risk of coronary artery disease in the UCP2 and UCP3 genes associates with abdominal obesity type 2 diabetic men,” Diabetes, vol. 57, no. 4, pp. 1063–1068, and serum lipids: the Finnish diabetes prevention study,” 2008. BMC Medical Genetics, vol. 10, Article ID 1471, p. 94, 2009. [103] B. R. Palmer, C. L. Devereaux, S. S. Dhamrait et al., “The [89] S. Le Fur, C. Le Stunff,C.DosSantos,andP.Bougneres,` “The common G-866A polymorphism of the UCP2 gene and common -866 G/A polymorphism in the promoter of uncou- survival in diabetic patients following myocardial infarction,” pling protein 2 is associated with increased carbohydrate and Cardiovascular Diabetology, vol. 8, Article ID 31, 2009. decreased lipid oxidation in juvenile obesity,” Diabetes,vol. [104] J. W. Stephens, S. S. Dhamrait, A. R. Mani et al., “Interaction 53, no. 1, pp. 235–239, 2004. between the uncoupling protein 2 -866G > Agenevariant [90]A.F.Reis,D.Dubois-Laforgue,C.Bellanne-Chantelot,´ J. and cigarette smoking to increase oxidative stress in subjects Timsit, and G. Velho, “A polymorphism in the promoter of with diabetes,” Nutrition, Metabolism and Cardiovascular UCP2 gene modulates lipid levels in patients with type 2 Diseases, vol. 18, no. 1, pp. 7–14, 2008. diabetes,” Molecular Genetics and Metabolism, vol. 82, no. 4, [105] G. Rudofsky, A. Schroedter, A. Schlotterer et al., “Functional pp. 339–344, 2004. polymorphisms of UCP2 and UCP3 are associated with a [91] N. Srivastava, J. Prakash, R. Lakhan, C. G. Agarwal, D. reduced prevalence of diabetic neuropathy in patients with C. Pant, and B. Mittal, “A common polymorphism in the type 1 diabetes,” Diabetes Care, vol. 29, no. 1, pp. 89–94, 2006. promoter of UCP2 is associated with obesity and hyper- [106] I. Labayen, F. B. Ortega, M. Sjostr¨ om,¨ T. K. Nilsson, L. A. insulenemia in northern Indians,” Molecular and Cellular Olsson, and J. R. Ruiz, “Association of common variants Biochemistry, vol. 337, no. 1-2, pp. 293–298, 2010. of UCP2 gene with low-grade inflammation in Swedish [92]Y.Yoon,B.L.Park,M.H.Chaetal.,“Effects of genetic children and adolescents; The European Youth Heart Study,” polymorphisms of UCP2 and UCP3 on very low calorie Pediatric Research, vol. 66, no. 3, pp. 350–354, 2009. diet-induced body fat reduction in Korean female subjects,” [107] E. Lapice, M. Pinelli, E. Pisu et al., “Uncoupling protein Biochemical and Biophysical Research Communications,vol. 2 G(-866)A polymorphism: a new gene polymorphism 359, no. 3, pp. 451–456, 2007. associated with C-reactive protein in type 2 diabetic patients [93] L. T. Dalgaard and O. Pedersen, “Uncoupling proteins: C-reactive protein in type 2 diabetic patients,” Cardiovascular functional characteristics and role in the pathogenesis of Diabetology, vol. 9, article 68, 2010. 12 Journal of Obesity

[108] G. Rudofsky, A. Schrodter,¨ O. E. Voron’ko et al., “Promoter polymorphisms of UCP1, UCP2, and UCP3 are not associated with diabetic microvascular complications in type 2 diabetes,” Hormone and Metabolic Research,vol.39,no.4,pp. 306–309, 2007. [109] B. F. Voight, L. J. Scott, V. Steinthorsdottir et al., “Twelve type 2 diabetes susceptibility loci identiﬁed through large- scale association analysis,” Nature Genetics,vol.42,no.7,pp. 579–589, 2010. [110] E. K. Speliotes, C. J. Willer, S. I. Berndt et al., “Association analyses of 249,796 individuals reveal 18 new loci associated with body mass index,” Nature Genetics, vol. 42, no. 11, pp. 937–948, 2010. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 609485, 10 pages doi:10.1155/2011/609485

Review Article Relationships of Adrenoceptor Polymorphisms with Obesity

Kazuko Masuo1, 2 and Gavin W. Lambert2

1 Nucleus Network, Ltd, Baker IDI Heart and Diabetes Research Institute, 89 Commercial Road, Melbourne, VIC 3004, Australia 2 Human Neurotransmitter Laboratory, Baker IDI Heart and Diabetes Research Institute, Melbourne, VIC 3004, Australia

Correspondence should be addressed to Kazuko Masuo, [email protected]

Received 24 November 2010; Accepted 7 February 2011

Academic Editor: Eric Doucet

Copyright © 2011 K. Masuo and G. W. Lambert. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Obesity, hypertension, and type 2 diabetes are rapidly growing public health problems. Heightened sympathetic nerve activity is a well-established observation in obesity, hypertension, and type 2 diabetes. Human obesity, hypertension, and diabetes have strong genetic as well as environmental determinants. Reduced energy expenditure and resting metabolic rate are predictive of weight gain, and the sympathetic nervous system participates in regulating energy balance through thermogenesis. The thermogenic effects of catecholamines in obesity are mainly mediated via the β2, and β3-adrenergic receptors in humans. Further, β2- adrenoceptors importantly influence vascular reactivity and may regulate blood pressure. β-adrenoceptor polymorphisms have also been associated with adrenoceptor desensitization, increased adiposity, insulin resistance, and enhanced sympathetic nervous activity. Many epidemiological studies have shown strong relationships between adrenoceptor polymorphisms and obesity, but the observations have been discordant. This paper will discuss the current topics involving the influence of the sympathetic nervous system and β2- and β3-adrenoceptor polymorphisms in obesity.

1. Introduction esis [12]. Recently, the important relationships of brown adipose tissue for energy expenditure [13–15] were argued, Obesity is a major and growing health problem. Impor- however a large part of the sympathetic nervous system- tantly, the presence of increased adiposity is associated with mediated energy expenditure takes place in skeletal muscle, elevated risk of development of cardiovascular and renal via the coupling of catecholamines with β2-adrenoceptors. complications [1–4]. Obesity is frequently associated with Catecholamines are also powerful regulators of lipolysis hypertension, diabetes, and metabolic syndrome [5–7], and and act via β1-, β2-, β3-(stimulatory), and α2-(inhibitory) sympathetic nervous activation is frequently observed in adrenoceptor subtypes in adipose tissue, where their role those conditions [8]. Thus, sympathetic nerve activation becomes especially important during both exercise and may play a major role in the onset and development of energy restriction, when increased need for fat as a fuel obesity, hypertension, and the development of the metabolic exists. Thus, β-adrenoceptors play important roles in energy syndrome as well as controlling to the cardiovascular compli- expenditure and control body weight [16–20]. cations evident in patients with hypertension, diabetes, and Recent evidence indicates that human obesity indeed has obesity [2, 4, 9]. a genetic component with [21–23] several epidemiological The sympathetic nervous system plays an important role and clinical studies indicating a strong linkage between β- in the regulation of energy expenditure. Reduced energy adrenoceptor polymorphisms and obesity or weight gain expenditure and resting metabolic rate are predictive of [24, 25]. Furthermore, heightened sympathetic nervous weight gain (obesity). Furthermore, blunted sympathetic system activity associated with β2- and β3-adrenoceptor nervous responses to energy intake have been observed in polymorphisms predicts subsequent weight gain and blood obese subjects with the metabolic syndrome and insulin pressure elevation in originally nonobese subjects [11, 24], resistance [10, 11]. The sympathetic nervous system par- and rebound weight gain after significant weight loss in obese ticipates in regulating energy balance through thermogen- subjects [26]. β2-adrenoceptor polymorphisms are related to 2 Journal of Obesity the onset of insulin resistance [27] and blunted responses of tension supports the hypothesis that heightened sympathetic sympathetic nerve activity to acute hyperinsulinemia [10, 11, nerve activity, through downregulation of β-adrenoceptor- 27–29]. These findings show that the genetic background, mediated thermogenesis, may facilitate the development of especially β2- and β3-adrenoceptor polymorphisms, are obesity in hypertension. Their results suggested that sympa- associated with sympathetic nervous system activation, and thetic nerve activity-induced hypertension may subsequently are important in the pathogenesis of obesity-related hyper- lead to the development of obesity. tension and insulin resistance. Many investigations regarding the relationships between β-adrenoceptor polymorphisms β and obesity have been analysed; however, the results are 3. Role of -Adrenoceptor discordant [30–32]. Polymorphisms in Obesity The sympathetic nervous system plays an important role in 2. Role of the Sympathetic Nervous the regulation of energy expenditure. A large part of the System in Obesity sympathetic nervous system-mediated energy expenditure takes place in skeletal muscle, via the coupling of cate- Many epidemiological and clinical studies have demon- cholamines with β2-adrenoceptors [54]. Catecholamines are strated a close relationship between sympathetic nervous also powerful regulators of lipolysis and act via β1-, β2-, β3- activation and insulin levels in obesity [33–37]. Several (stimulatory) and α2-(inhibitory) adrenoceptor subtypes in longitudinal studies have examined the effect of body weight adipose tissue, where their role becomes especially important changes (weight loss or weight gain) on sympathetic nervous during both exercise and energy restriction, when increased system activity and insulin sensitivity (fasting plasma insulin need for fat as a fuel exists. Stimulation of β-adrenergic levels and HOMA-IR). Elevated activity of the sympathetic receptors by the sympathetic nervous system is a significant nervous system and increased insulin levels during weight physiological modulator of pre- and postprandial energy gain [16, 24, 26, 38, 39] and reductions of sympathetic expenditure [18–20] and total daily energy expenditure nerve activity and insulin levels during weight loss [40– [16, 17, 50]. The subtypes of adrenoceptors on lipid 45]havebeenobserved.Inobesenormotensivesubjects,a and glucose metabolisms are summarized as following; reduction in body weight induced exerts a marked reduction α1-adrenoceptors, glycogenolysis and gluconeogenesis in in sympathetic activity owing to central sympathoinhibition adipose tissue and liver; α2-adrenoceptors, induction of due to the consequences of an increased insulin sensitivity glucagon release from pancreas; β1-adrenoceptors, lipolysis and a restoration of the baroreflex [45]. These studies have in adipose tissue; β2-adrenoceptors, glycogenolysis and glu- clearly shown heightened sympathetic nerve activity and coneogenesis in adipose tissue and liver; β3-adrenoceptors, insulin resistance are closely linked to weight gain and the lipolysis on adipose tissue. onset and maintenance of obesity. Recent studies show that β-adrenoceptors are poly- Landsberg et al. [46–48]andJuliusetal.[49]have morphic with single nucleotide polymorphisms exerting proposed hypotheses to explain the mechanism linking the functional consequences in terms of receptor activity and sympathetic nervous system and insulin resistance in obesity. regulation and hence perhaps may contributing to the The former proposes that hyperinsulinemia and insulin pathophysiology of obesity and hypertension [24, 25, 55–59]. resistance in obese subjects are all part of a response to limit On the other hand, there are few studies on the relationships further weight gain via stimulating sympathetic nervous between α-adrenoceptor polymorphisms and obesity. activity and thermogenesis [50], and the latter indicates that sympathoexcitation in the skeletal muscle vascular bed cause 3.1. β1-Adrenoceptor Polymorphisms (Table 1). The β1- neurogenic vasoconstriction and reduction in blood flow to adrenoceptor is predominantly expressed in cardiac muscle and consequently induces a state of insulin resistance myocytes and adipose tissue, where its activation leads to by lowering glucose delivery and uptake in hypertension increased heart rate and contractility and stimulation of and obesity. Masuo et al. in a series of longitudinal studies lipolysis, respectively. The β1-adrenoceptor is a candidate observed that heightened sympathetic activity was the prime gene for obesity because of its role in catecholamine- mover for future weight gain in originally nonobese, nor- mediated energy homeostasis. In obese individuals, the motensive subjects, and that insulin resistance was more an degree of weight loss during a very low calorie diet has ancillary factor [24, 51, 52]. In investigations examining the been shown to correlate with changes in β1-adrenoceptor effect of weight loss, reductions in plasma norepinephrine protein concentration in adipose tissue [65]. The two most followed by reductions in HOMA-IR as a marker of insulin common β1-adrenoceptor polymorphisms are Ser49Gly resistance were significantly greater in subjects experiencing and Arg389Gly, with relative allele frequencies of 0.85/0.15 significant weight loss compared to those without significant and 0.70/0.30 in Caucasian population, respectively. An weight loss [26, 40, 43]. These observations provide some investigation involving a population cohort of 761 women support for the hypothesis of Julius and colleagues. indicated that women carrying the Gly49 genotype had Valentini et al. [53] reported attenuation of hemody- greater elevation in BMI over 15 years compared to those namic and energy expenditure responses to isoproterenol with the Ser49 genotype [62]. Again, in Caucasian women infusion in hypertensive patients. Their findings that a gen- (n = 931), Dionne et al. [60] observed that the Gly389Arg, eralized decrease of β-adrenergic responsiveness in hyper- β1-adrenoceptor variant exhibited a strong relationships Journal of Obesity 3

Table 1: Summary of studies showing associations between on β1-adrenoceptor polymorphisms and obesity.

Authors Year Population Subjects Findings (reference)

Arg allele of Argt389Gly was associated with obesity (greater body weight Dionne et al. [60] 2002 Caucasian 931 women and BMI due to greater fat mass).

Children and The distributions of Ser49Gly and Arg389Gly were not diﬀerent between Tafel et al. [61] 2004 German adolescents lean and obese adolescents.

The combination of Gly49-Gly389 (Ser49Gly + Arg389Gly) was associated Linneetal.[´ 62] 2005 Scandinavian 761 women with long term of 15 years weight gain and the incidence of adult-onset overweight in women, but no eﬀect of Arg389Gly alone on obesity.

Danish- Arg389Gly polymorphism was not related with obesity, but minor inﬂuence Gjesing et al. [63] 2007 7,677 Caucasians on BP.

188 type 2 Nonen et al. [64] 2008 Japanese diabetic Ser49Gly, but not Arg389Gly, was associated with obesity. patients BP: blood pressure. with obesity. Conversely, Gjesing and colleagues found Arg16, whereas the Glu27 receptor is resistant to down that the distribution of the Arg389Gly polymorphism was regulation when compared with the Gln27 variant [77]. A similar in lean and obese subjects, suggesting that it has no number of clinical studies have investigated the impact of important influence on human obesity [63, 66]. Although these polymorphisms on vascular responsiveness [55, 78]. earlier small case-control studies demonstrated an increase Gratze et al. [79] found that young normotensive white men in the risk of hypertension in Arg389 homozygotes [67, 68], a homozygous for the Gly16 allele had higher blood pressure recently published study comprising 3981 normotensive and and lower peripheral vasodilation after infusion of the β2- 2,518 hypertensive patients failed to replicate this association agonist salbutamol. Similar results were obtained by Hoit et [63] (summarised in Table 1). Arner [67] reviewed that al. [80] using the agonist terbutaline. On the other hand, vol- Arg389Gly polymorphism in the β1-adrenoceptor, which unteers homozygous for Gly16 exhibited larger vasodilatory alters receptor function in transfected cell lines, and responses than did volunteers homozygous for Arg16 [81]. concluded that the SNP has no effect on lipolysis in human Conflicting results have also been published with regards to fat cells and is not associated with obesity. the effects of genetic variants on the sympathetic nervous system modulation of energy expenditure. Bell et al. [82] 3.2. β2-Adrenoceptor Polymorphisms (Table 2). The β2- reported that the response of resting energy expenditure to adrenoceptor is the dominant lipolytic receptor in white nonspecific β-adrenoceptor stimulation (with isoproterenol human adipose tissue [20, 55, 56] and in skeletal muscle infusion) was not different between the 3 genotypes of [19, 57]. Gln16Glu and an Arg164Ile variation in the Arg16Gly. Stob et al. [70] showed that individuals carrying β2-adrenoceptor cause marked variations in the lipoly- the Arg16Arg variant of the β2-adrenoceptor gene have a tic sensitivity of this receptor in human adipocytes. reduced thermogenic response to selective β2-adrenoceptor Multiple β2-adrenoceptor polymorphisms including hap- activation. lotypes, markedly influence β2-receptor function- and Associations of β2-adrenoceptor polymorphisms with catecholamine-induced lipolysis in fat cells [76]. These hap- obesity have been reported in many epidemiological studies lotypes may be important genetic factors behind impaired but results are also discordant (summarised in Table 2). lipolysis in obesity [25]. The β2-adrenoceptor also plays an important regulatory 3.3. β3-Adrenoceptor Polymorphisms (Table 3). The β3- role in the peripheral vasculature. Genetic polymorphisms adrenoceptor, which is mainly expressed in adipose tissue, of the β2-adrenoceptor have been associated with obesity, differs from the β2-adrenoceptor in two ways: it has a lower hypertension, and diabetes mellitus. The most common affinity for catecholamines, and it resists desensitisation (i.e., polymorphisms are Arg16Gly, with an allele frequency of downregulation). These characteristic differences might lead 0.40/0.60 and Gln27Glu, with an allele frequency of 0.55/0.45 to the different effects of catecholamine on β2-adrenoceptors in the Caucasian population. The Thr164Ile polymorphism and β3-adrenoceptors. β3-adrenoceptors stimulates the is rare, occurring in only 3 to 5% of the general (Caucasian) mobilization of lipids from the white adipose tissue and population. increases thermogenesis in brown adipose tissue. Cypess Studies of agonist stimulation in cultured cells demon- et al. and other investigators demonstrated that potential strate that Gly16 receptors have a greater reduction in roles of β3-adrenoceptor polymorphism (Trp64Arg) asso- numbers or enhanced downregulation when compared with ciated with potential role of uncoupling protein (UCP) 4 Journal of Obesity

Table 2: Summary of studies showing associations between β2-adrenoceptor polymorphisms and obesity.

Authors [reference] Year Population Subjects Findings

Caucasian women with Large et al. [55] 1997 Swedish Gln27Glu polymorphism was associated with obesity. wide range of obesity

Caucasian juvenile-onset Echwald et al. [58] 1998 Danes No association between Gln27Glu and obesity. obese men

Swedish-Caucasian men Gln27Glu polymorphism was associated with obesity only in Hellstrom¨ et al. [59] 1999 Swedish and women women, but not in men.

Caucasian with morbid Kortner et al. [69] 1999 German Gln27Glu polymorphism was not associated with obesity. obesity

The Quebec Family Caucasian men and Gln27Glu polymorphism was associated with obesity and 2000 Canadian Study [70] women hyperlipidemia.

12 pairs of twins, Gln27Glu polymorphism was associated with weight gain Ukkola et al. [56] 2001 USA Caucasians (obesity).

Subjects carrying Gln27 homozygous had an increased risk of obesity in men, but not in women. Further, men with Gln27 Meirhaeghe et al. 2000 French 1,195 subjects homozygous carried in addition the Arg16 allele, had more [71] signiﬁcant increase in body weight, BMI and waist-to-hip ratio (central obesity).

The HERITAGE Sedentary black and Gln27Glu polymorphism was associated with lower fat in obese 2003 Canada family study [72] white men white men.

Subjects carrying Gln27 homozygous had higher risk of obesity, 1,576 individuals Pereira et al. [25] 2003 Brazilian whereas those with Gly27 homozygous had increased risk of randomly selected hypertension.

1,354 women and 421 Common haplotypes of ADRB2 polymorphisms had recessive Jiao et al. [73] 2005 Scandinavian men eﬀects against excess body fat only in women, but not in men.

154 overweight/obese Gly16 allele was related to obesity and rebound weight gain in Masuo et al. [26] 2005 Japanese men weight-loss study.

160 nonobese, Gly16 allele was related to future weight gain, BP elevation and Masuo et al. [24, 27] 2005 Japanese normotensive men insulin resistance in originally nonobese, normotensive men.

329 normotensive men Gly16 and Glu27 alleles were related to obesity through Masuo et al. [28] 2006 Japanese with a wide range of blunted-leptin-mediated sympathetic activity. BMI

Kawaguchi et al. 55 overweight/obese 2006 Japanese Gly16 allele was related to further weight gain in obese subjects. [29] men

Thehaplotypeof5LC-Cys(19)Arg(16)Gln(27) was related to 642 overweight/obese Petrone et al. [74] 2006 European additional weight gain with increases of triglycerides and subjects LDL-cholesterol.

No consistent effect of ADRB2 haplotypes on obesity and Gjesing et al. [75] 2009 Danes 6,514 adults quantitative traits of body fatness. ADRB2: β2-adrenoceptors; BP: blood pressure. polymorphisms and brown adipose tissue in thermogenesis stores effectively [88–90]. Hoffstedt et al. [91]compared and resultant body weight in humans [13–15]. Decreased adrenergic regulation of lipolysis between omental and function of β3-adrenoceptor in white adipose tissue could subcutaneous adipocytes from 15 obese and 14 nonobese slow lipolysis and thereby cause the retention of lipids men. In their study, catecholamine-induced lipolysis was in adipocytes. Slow lipolysis may contribute strongly to markedly increased in omental adipocytes as compared to visceral obesity in human, and treatment of obese animal subcutaneous adipocytes in obese male subjects mainly due models with selective β3-adrenergic agonists reduces fat to an increase in β3-adrenoceptor function of visceral fat, Journal of Obesity 5

Table 3: Summary of studies showing associations between β3-adrenoceptor polymorphisms and obesity.

Authors [reference] Year Population Subjects Findings

Patients with morbid Subjects carrying β3-ADR polymorphisms has an increased Clement et al. [83] 1995 French obesity capacity to gain weight.

The Arg64 allele of Trp64Arg may predict diﬃculty in losing body 61 obese women with weight, lowering waist-to-hip ratio, and improving glycemic Sakane et al. [84] 1997 Japanese type 2 diabetes control and insulin resistance in obese patients with type 2 diabetes.

18 omental fat samples Trp64Arg polymorphism was associated with lower lipolytic Umekawa et al. [85] 1999 Japanese obtained during total activities. hysterectomy

553 Japanese Trp64Arg polymorphism might be a genetic risk factor for obesity Endo et al. [86] 2000 Japanese schoolchildren (291 boys in Japanese children. and 262 girls)

Arg64/Arg64, but not Trp64/Arg64, of the β-adrenergic receptor 1,685 (935 women and Oizumi et al. [87] 2001 Japanese polymorphism was associated with both obesity and type 2 750 men) diabetes in a large Japanese cohort.

160 nonobese, Trp64Arg polymorphism was related to BP elevations, but not to Masuo et al. [24] 2005 Japanese normotensive men weight gain in originally nonobese subjects.

Kawaguchi et al. 55 overweight/obese Trp64Arg polymorphism was related to further weight gain in 2006 Japanese [29] men originally obese subjects.

Trp64Arg polymorphism did not confer an increased risk of Danish- Gjesing et al. [63] 2007 7,605 obesity among Danes, although the variant is associated with type Caucasians 2 diabetes and quantitative traits related to type 2 diabetes.

Table 4: Confounding variables considered to cause the discrepancy of the relationships between β-adrenoceptor polymorphisms and phenotypes of obesity, hypertension, and diabetes.

Variables [reference number] Findings in the studies

In lean subjects, β2-AR polymorphisms linked to obesity and obesity-related hypertension, but in obese subjects β2- and β3-AR Polymorphisms related to obesity and obesity-related hypertension. Severity of obesity [23, 28, 29] Morbid obesity was linked with β3-AR polymorphisms, but Overweight or mild obesity was not associated with those.

Interaction between β1- and β2-AR polymorphisms with changes in BMI was observed in men Gender diﬀerences [71, 73] only, while in women an interaction between β1- and β3-AR polymorphisms was observed in a longitudinal over a 24-year period large cohort study.

Ethnic difference [30] Distributions of β-AR polymorphisms are different in 8 different ethnic population.

Haplotype [25, 73, 74, 76, 86, 93–97] Functions expressed of β-AR polymorphisms are diﬀerent due to the other β-AR polymorphisms. AR: adrenoceptor; BMI: body mass index.

in combination with a smaller increase in β1-adrenoceptor 3.4. Confounding Variables Aﬀecting the Relationships of β- function [91]. Recently, Eriksson et al. [76] observed that Adrenoceptor Polymorphisms with Obesity, Hypertension and Trp64Arg polymorphism in the β3-receptor, which associates Diabetes (Table 4). Tables 1–3 show the discordant contribu- with obesity, is accompanied by changes in lipolytic sensitiv- tions of β-adrenoceptor polymorphisms to obesity. Table 4 ity of the receptor in human adipocytes. Many epidemiolog- summarizes factors which might explain the discrepancy of ical studies have shown the strong relationships between β3- published data. Importantly, haplotypes of polymorphisms adrenoceptor polymorphisms (mainly Trp54Arg), obesity, have strong inﬂuence on β-adrenoceptor function in each metabolic syndrome, and hypertension [87–92](Table 3). polymorphism [25, 73, 86, 93–97]. 6 Journal of Obesity

4. Sympathetic Nervous System these observations and the recent demonstration of the Activity and β2- and β3-Adrenoceptor effectiveness of catheter based sympathetic renal denervation Polymorphisms in Obesity for the treatment of refractory hypertension [114, 115], it may be of importance to aim antihypertensive treatments or Many studies have examined the associations of the β2- anti-diabetic treatment not only at the reduction of raised or β3-adrenoceptor polymorphisms with obesity and blood blood pressure or blood glucose but also at the excessive ff pressure as mentioned above. A series of studies conducted sympathetic activation that may underpin these e ects. by Masuo et al. have included measurements of sympathetic nervous system activity [24, 26]. In a longitudinal study 6. Conclusions over 5 years, originally nonobese, normotensive subjects carrying the Gly16 allele of Arg16Gly, the combination of Established and emerging data emphasises the impor- β2-adrenoceptor polymorphisms and high plasma nore- tance of the sympathetic nervous system in obesity and pinephrine levels on entry were linked to weight gain obesity-related illness. Sympathetic nervous system activity and blood pressure elevations in addition to weight gain- and β-adrenoceptor polymorphisms (mainly β2- and β3- induced blood pressure elevations [24]. In a weight loss adrenoceptor polymorphisms) may contribute to the onset study over a 24-month period, the β2-adrenoceptor the and maintenance of obesity; however, the findings have Gly16 allele of Arg16Gly was associated with resistance been discordant. A better understanding of the pathogenesis to long term significant weight loss, and the Glu27 allele of obesity, including an understanding of adrenoceptor was linked to resistance to short-term weight loss [26]. polymorphisms and their impact on sympathetic nervous Nonobese normotensive men carrying the Gly16 allele of activity might help in the prevention of obesity and the Arg16Gly had a higher frequency of insulin resistance, as pharmacological treatment of obesity-related illness includ- indicated by elevation in the homeostasis model assessment ing hypertension and insulin resistance. for insulin resistance (HOMA) index. This deterioration in insulin resistance is generally observed in obesity and hypertension [27, 36, 44, 98]. These studies provide strong Disclosure evidence for the linkage between β2-adrenoceptor polymorphisms, heightened sympathetic nervous system activ- The laboratory of Dr. Lambert GW currently receives ity, obesity, hypertension, and the development of insulin research funding from private organisations including resistance. ARDIAN Inc, Allergan, Abbott (formerly Solvay) Pharma- ceuticals, and Scientific Intake. These organizations played no role in this paper. The investigators report no conflict of 5. Elevated Sympathetic Nervous Activity interests with regards to this paper. in Obesity Is a Risk Factor for Cardiovascular Complications and Renal Complications References

The increased risk of cardiovascular complications in obesity, [1] A. H. Mokdad, B. A. Bowman, E. S. Ford, F. Vinicor, J. especially associated with hypertension or type 2 diabetes, S. Marks, and J. P. Koplan, “The continuing epidemics of has been attributed to a variety of mechanisms, including obesity and diabetes in the United States,” The Journal of dyslipidemia, coagulation abnormalities, endothelial dys- the American Medical Association, vol. 286, no. 10, pp. 1195– function, chronic sympathetic nerve activation, and repeated 1200, 2001. occurrence of excessive hyperinsulinemia [99–102]. Sympa- [2] C. Heidemann, H. Boeing, T. Pischon, U. Nothlings,¨ H. G. thetic nerve hyperactivity leads to arterial blood pressure Joost, and M. B. Schulze, “Association of a diabetes risk score elevation [103–105], triggers arterial damage, and results in with risk of myocardial infarction, stroke, speciﬁc types of cardiovascular events. Recent studies and reviews reported cancer, and mortality: a prospective study in the European that sympathetic nerve stimulation contributes to the pro- Prospective Investigation into Cancer and Nutrition (EPIC)- gression of renal disease [106, 107]. Norepinephrine infusion Potsdam cohort,” European Journal of Epidemiology, vol. 24, no. 6, pp. 281–288, 2009. into the renal artery in dogs produced a reversible ischemic model of acute renal failure [108]. Another study demon- [3] K. M. Flegal, B. I. Graubard, D. F. Williamson, and M. H. Gail, “Cause-speciﬁc excess deaths associated with under- strated renal protection by β-adrenergic receptor blockade weight, overweight, and obesity,” TheJournaloftheAmerican in a nephrectomized rat without any BP changes [109]. Medical Association, vol. 298, no. 17, pp. 2028–2037, 2007. Plasma norepinephrine and heightened sympathetic nerve [4] E. Lambert, C. I. Sari, T. Dawood et al., “Sympathetic nervous activity may predict mortality and incident cardiovascular system activity is associated with obesity-induced subclinical events including renal injury in large cohort longitudinal organ damage in young adults,” Hypertension,vol.56,no.3, studies [110] and clinical studies [111, 112]. Renal injury pp. 351–358, 2010. also predicts the development of cardiovascular disease [5] F. Ramsey, A. Ussery-Hall, D. Garcia et al., “Prevalence of [105, 106, 110, 111]. There is consistent evidence that selected risk behaviors and chronic diseases: behavioral Risk elevated sympathetic nervous activity predicts mortality Factor Surveillance System (BRFSS), 39 steps communities, in cardiovascular disease such as in patients with heart United States, 2005,” Morbidity and Mortality Weekly Report. failure [111] and end-stage renal disease [112, 113]. Given Surveillance Summaries/CDC, vol. 57, no. 11, pp. 1–20, 2008. Journal of Obesity 7

[6]C.L.Ogden,C.D.Fryar,M.D.Carroll,andK.M.Flegal, [21] K. Masuo, H. Mikami, T. Ogihara, and M. L. Tuck, “Familial “Mean body weight, height, and body mass index, United hypertension, insulin, sympathetic activity, and blood pres- States 1960–2002,” Advance Data, no. 347, pp. 1–17, 2004. sure elevation,” Hypertension, vol. 32, no. 1, pp. 96–100, 1998. [7] A. H. Mokdad, M. K. Serdula, W. H. Dietz, B. A. Bowman, [22] K. Masuo, H. Mikami, T. Ogihara, and M. L. Tuck, “Familial J. S. Marks, and J. P. Koplan, “The spread of the obesity obesity, sympathetic activation and blood pressure level,” epidemic in the United States, 1991–1998,” The Journal of Blood Pressure, vol. 10, no. 4, pp. 199–204, 2001. the American Medical Association, vol. 282, no. 16, pp. 1519– [23] J. Cui, J. L. Hopper, and S. B. Harrap, “Genes and family 1522, 1999. environment explain correlations between blood pressure [8]G.W.Lambert,N.E.Straznicky,E.A.Lambert,J.B.Dixon, and body mass index,” Hypertension, vol. 40, no. 1, pp. 7–12, and M. P. Schlaich, “Sympathetic nervous activation in obe- 2002. sity and the metabolic syndrome—causes, consequences and [24] K. Masuo, T. Katsuya, Y. Fu, H. Rakugi, T. Ogihara, and therapeutic implications,” Pharmacology and Therapeutics, M. L. Tuck, “β2- and β3-adrenergic receptor polymorphisms vol. 126, no. 2, pp. 159–172, 2010. are related to the onset of weight gain and blood pressure [9] S. R. Preis, S. J. Hwang, S. Coady et al., “Trends in all-cause elevation over 5 years,” Circulation, vol. 111, no. 25, pp. 3429– and cardiovascular disease mortality among women and men 3434, 2005. with and without diabetes mellitus in the framingham heart [25] A. C. Pereira, M. S. Floriano, G. F. A. Mota et al., “β2adreno- study, 1950 to 2005,” Circulation, vol. 119, no. 13, pp. 1728– ceptor functional gene variants, obesity, and blood pressure 1735, 2009. level interactions in the general population,” Hypertension, [10] N. E. Straznicky, G. W. Lambert, K. Masuo et al., “Blunted vol. 42, no. 4, pp. 685–692, 2003. sympathetic neural response to oral glucose in obese subjects [26] K. Masuo, T. Katsuya, H. Kawaguchi et al., “Rebound with the insulin-resistant metabolic syndrome,” American weight gain as associated with high plasma norepinephrine Journal of Clinical Nutrition, vol. 89, no. 1, pp. 27–36, 2009. levels that are mediated through polymorphisms in the β2- [11] K. Masuo, H. Mikami, T. Ogihara, and M. L. Tuck, “Dif- adrenoceptor,” American Journal of Hypertension, vol. 18, no. ferences in insulin and sympathetic responses to glucose 11, pp. 1508–1516, 2005. ingestion due to family history of hypertension,” American [27] K. Masuo, T. Katsuya, Y. Fu, H. Rakugi, T. Ogihara, and M. Journal of Hypertension, vol. 9, no. 8, pp. 739–745, 1996. L. Tuck, “β2-adrenoceptor polymorphisms relate to insulin [12] M. Spraul, E. Ravussin, A. M. Fontvieille, R. Rising, D. E. resistance and sympathetic overactivity as early markers of Larson, and E. A. Anderson, “Reduced sympathetic nervous metabolic disease in nonobese, normotensive individuals,” activity. A potential mechanism predisposing to body weight American Journal of Hypertension, vol. 18, no. 7, pp. 1009– gain,” Journal of Clinical Investigation,vol.92,no.4,pp. 1014, 2005. 1730–1735, 1993. [28] K. Masuo, T. Katsuya, H. Kawaguchi et al., “β2-adrenoceptor [13] A. M. Cypess, S. Lehman, G. Williams et al., “Identification polymorphisms relate to obesity through blunted leptin- and importance of brown adipose tissue in adult humans,” mediated sympathetic activation,” American Journal of The New England Journal of Medicine, vol. 360, no. 15, pp. Hypertension, vol. 19, no. 10, pp. 1084–1091, 2006. 1509–1517, 2009. [29] H. Kawaguchi, K. Masuo, T. Katsuya et al., “β2- and β3- [14] J. A. Timmons and B. K. Pedersen, “The importance of adrenoceptor polymorphisms relate to subsequent weight brown adipose tissue,” The New England Journal of Medicine, gain and blood pressure elevation in obese normotensive vol. 361, no. 4, pp. 415–416, 2009. individuals,” Hypertension Research, vol. 29, no. 12, pp. 951– [15] H. S. Sacks, “The importance of brown adipose tissue,” The 959, 2006. New England Journal of Medicine, vol. 361, no. 4, p. 418, 2009. [30] T. J. Maxwell, M. M. Ameyaw, S. Pritchard et al., “Beta-2 [16] S. Iwashita, M. Tanida, N. Terui et al., “Direct measurement adrenergic receptor genotypes and haplotypes in different of renal sympathetic nervous activity in high-fat diet-related ethnic groups,” International Journal of Molecular Medicine, hypertensive rats,” Life Sciences, vol. 71, no. 5, pp. 537–546, vol. 16, no. 4, pp. 573–580, 2005. 2002. [31] N. E. Straznicky, N. Eikelis, E. A. Lambert, and M. D. Esler, [17] M. B. Monroe, D. R. Seals, L. F. Shapiro, C. Bell, D. Johnson, “Mediators of sympathetic activation in metabolic syndrome and P. Parker-Jones, “Direct evidence for tonic sympathetic obesity,” Current Hypertension Reports, vol. 10, no. 6, pp. 440– support of resting metabolic rate in healthy adult humans,” 447, 2008. American Journal of Physiology, vol. 280, no. 5, pp. E740– [32] J. L. Gonzalez´ Sanchez,´ A. M. Proenza, M. T. Mart´ınez Larrad E744, 2001. et al., “The glutamine 27 glutamic acid polymorphism of the [18]E.E.Blaak,M.A.vanBaak,K.P.G.Kempen,andW.H. β2-adrenoceptor gene is associated with abdominal obesity M. Saris, “Role of α-andβ-adrenoceptors in sympathetically and greater risk of impaired glucose tolerance in men but mediated thermogenesis,” American Journal of Physiology, not in women: a population-based study in Spain,” Clinical vol. 264, no. 1, part 1, pp. E11–E17, 1993. Endocrinology, vol. 59, no. 4, pp. 476–481, 2003. [19] E. Hagstrom-Toft,¨ S. Enoksson, E. Moberg, J. Bolinder, and [33]N.E.Straznicky,E.A.Lambert,G.W.Lambert,K.Masuo, P. Arner, “β-adrenergic regulation of lipolysis and blood M. D. Esler, and P. J. Nestel, “Effects of dietary weight loss on flow in human skeletal muscle in vivo,” American Journal of sympathetic activity and cardiac risk factors associated with Physiology, vol. 275, no. 6, pp. E909–E916, 1998. the metabolic syndrome,” Journal of Clinical Endocrinology & [20] S. Enoksson, M. Talbot, F. Rife, W. V. Tamborlane, R. S. Metabolism, vol. 90, no. 11, pp. 5998–6005, 2005. Sherwin, and S. Caprio, “Impaired in vivo stimulation of [34] L. Landsberg, “Insulin-mediated sympathetic stimulation: lipolysis in adipose tissue by selective β2-adrenergic agonist role in the pathogenesis of obesity-related hypertension (or, in obese adolescent girls,” Diabetes, vol. 49, no. 12, pp. 2149– how insulin affects blood pressure, and why),” Journal of 2153, 2000. Hypertension, vol. 19, no. 3, pp. 523–528, 2001. 8 Journal of Obesity

[35]E.A.Anderson,R.P.Hoffman, T. W. Balon, C. A. Sinkey, and [51] K. Masuo, H. Kawaguchi, H. Mikami, T. Ogihara, and M. L. A. L. Mark, “Hyperinsulinemia produces both sympathetic Tuck, “Serum uric acid and plasma norepinephrine concen- neural activation and vasodilation in normal humans,” trations predict subsequent weight gain and blood pressure Journal of Clinical Investigation, vol. 87, no. 6, pp. 2246–2252, elevation,” Hypertension, vol. 42, no. 4, pp. 474–480, 2003. 1991. [52] K. Masuo, H. Mikami, T. Ogihara, and M. L. Tuck, “Weight [36] E. Ferrannini, “Physiological and metabolic consequences of gain-induced blood pressure elevation,” Hypertension, vol. obesity,” Metabolism, vol. 44, no. 9, supplement 3, pp. 15–17, 35, no. 5, pp. 1135–1140, 2000. 1995. [53] M. Valentini, S. Julius, P. Palatini et al., “Attenuation [37] J.-P. Montani, V. Antic, Z. Yang, and A. Dulloo, “Pathways of haemodynamic, metabolic and energy expenditure from obesity to hypertension: from the perspective of a responses to isoproterenol in patients with hypertension,” vicious triangle,” International Journal of Obesity, vol. 26, Journal of Hypertension, vol. 22, no. 10, pp. 1999–2006, 2004. supplement 2, pp. S28–S38, 2002. [54] Y. T. Yang and M. A. McElligott, “Multiple actions of [38] C. L. Gentile, J. S. Orr, B. M. Davy, and K. P. Davy, “Modest β-adrenergic agonists on skeletal muscle and adipose tissue,” weight gain is associated with sympathetic neural activation Biochemical Journal, vol. 261, no. 1, pp. 1–10, 1989. in nonobese humans,” American Journal of Physiology, vol. [55] V. Large, L. Hellstrom, S. Reynisdottir et al., “Human beta-2 292, no. 5, pp. R1834–R1838, 2007. adrenoceptor gene polymorphisms are highly frequent [39]M.J.Barnes,K.Lapanowski,A.Conley,J.A.Rafols,K.L.C. in obesity and associate with altered adipocyte beta-2 Jen, and J. C. Dunbar, “High fat feeding is associated with adrenoceptor function,” Journal of Clinical Investigation, vol. increased blood pressure, sympathetic nerve activity and 100, no. 12, pp. 3005–3013, 1997. hypothalamic mu opioid receptors,” Brain Research Bulletin, [56] O. Ukkola, A. Tremblay, and C. Bouchard, “Beta-2 adrenergic vol. 61, no. 5, pp. 511–519, 2003. receptor variants are associated with subcutaneous fat accu- [40] K. Masuo, H. Mikami, T. Ogihara, and M. L. Tuck, mulation in response to long-term overfeeding,” Interna- “Weight reduction and pharmacologic treatment in obese tionalJournalofObesity, vol. 25, no. 11, pp. 1604–1608, 2001. hypertensives,” American Journal of Hypertension, vol. 14, no. [57] S. Enoksson, E. Hagstrom-Toft,¨ J. Nordahl et al., “Marked 6, part 1, pp. 530–538, 2001. reutilization of free fatty acids during activated lipolysis in [41] B. Andersson, M. Elam, B. G. Wallin, P. Bjorntorp, and O. K. human skeletal muscle,” Journal of Clinical Endocrinology & Andersson, “Effect of energy-restricted diet on sympathetic Metabolism, vol. 90, no. 2, pp. 1189–1195, 2005. muscle nerve activity in obese women,” Hypertension, vol. [58] S. M. Echwald, T. I. A. Sørensen, A. Tybjærg-Hansen, T. 18, no. 6, pp. 783–789, 1991. Andersen, and O. Pedersen, “Gln27Glu variant of the human β2-adrenoreceptor gene is not associated with early-onset [42] N. E. Straznicky, E. A. Lambert, G. W. Lambert, K. Masuo, obesity in Danish men,” Diabetes, vol. 47, no. 10, pp. M. D. Esler, and P. J. Nestel, “Effects of dietary weight loss on 1657–1658, 1998. sympathetic activity and cardiac risk factors associated with themetabolicsyndrome,”Journal of Clinical Endocrinology [59] L. Hellstrom,¨ V. Large, S. Reynisdottir, H. Wahrenberg, and P. ff ff β & Metabolism, vol. 90, no. 11, pp. 5998–6005, 2005. Arner, “The di erent e ects of a Gln27Glu 2-adrenoceptor gene polymorphism on obesity in males and in females,” [43] K. Masuo, H. Mikami, T. Ogihara, and M. L. Tuck, Journal of Internal Medicine, vol. 245, no. 3, pp. 253–259, “Differences in mechanisms between weight loss-sensitive 1999. and -resistant blood pressure reduction in obese subjects,” Hypertension Research, vol. 24, no. 4, pp. 371–376, 2001. [60] I. J. Dionne, M. J. Garant, A. A. Nolan et al., “Association between obesity and a polymorphism in the β -adrenoceptor [44] M. L. Tuck, J. R. Sowers, and L. Dornfeld, “Reductions in 1 gene (Gly389Arg ADRB1) in Caucasian women,” Interna- plasma catecholamines and blood pressure during weight tional Journal of Obesity, vol. 26, no. 5, pp. 633–639, 2002. loss in obese subjects,” Acta Endocrinologica, vol. 102, no. 2, [61] J. Tafel, I. Branscheid, B. Skwarna et al., “Variants in the pp. 252–257, 1983. human β1-, β2- and β3-adrenergic receptor genes are not [45] G. Grassi, G. Seravalle, M. Colombo et al., “Body weight ffi associated with morbid obesity in children and adolescents,” reduction, sympathetic nerve tra c, and arterial baroreflex Diabetes, Obesity & Metabolism, vol. 6, no. 6, pp. 452–455, in obese normotensive humans,” Circulation, vol. 97, no. 20, 2004. pp. 2037–2042, 1998. [62] Y. Linne,´ I. Dahlman, and J. Hoffstedt, “β1-adrenoceptor [46] L. Landsberg, “Diet, obesity and hypertension: an hypothesis gene polymorphism predicts long-term changes in body involving insulin, the sympathetic nervous system, and weight,” International Journal of Obesity,vol.29,no.5,pp. adaptive thermogenesis,” The Quarterly Journal of Medicine, 458–462, 2005. vol. 61, no. 236, pp. 1081–1090, 1986. [63] A. P. Gjesing, G. Andersen, A. Albrechtsen et al., “Studies of [47] L. Landsberg and J. B. Young, “Diet and the sympathetic associations between the Arg389Gly polymorphism of the nervous system: relationship to hypertension,” International β1-adrenergic receptor gene (ADRB1) and hypertension and Journal of Obesity, vol. 5, supplement 1, pp. 79–91, 1981. obesity in 7677 Danish white subjects,” Diabetic Medicine, [48] L. Landsberg and D. R. Krieger, “Obesity, metabolism, vol. 24, no. 4, pp. 392–397, 2007. and the sympathetic nervous system,” American Journal of [64] S. Nonen, I. Yamamoto, J. Liu et al., “Adrenergic β1receptor Hypertension, vol. 2, no. 3, pp. 125S–132S, 1989. polymorphism (Ser49gly) is associated with obesity in type [49] S. Julius, M. Valentini, and P. Palatini, “Overweight and II diabetic patients,” Biological & Pharmaceutical Bulletin, hypertension: a 2-way street?” Hypertension,vol.35,no.3, vol. 31, no. 2, pp. 295–298, 2008. pp. 807–813, 2000. [65] M. Rasmussen, A. Belza, T. Almdal et al., “Change in β1- [50] K. O’Dea, M. Esler, and P. Leonard, “Noradrenaline turnover adrenergic receptor protein concentration in adipose tissue during under- and over-eating in normal weight subjects,” correlates with diet-induced weight loss,” Clinical Science, Metabolism, vol. 31, no. 9, pp. 896–899, 1982. vol. 108, no. 4, pp. 323–329, 2005. Journal of Obesity 9

[66] M. Ryden, J. Hoffstedt, P. Eriksson, S. Bringman, and P. [81] K. Leineweber, “Beta-adrenergic receptor polymorphism in Arner, “The Arg389Gly β1-adrenergic receptor gene poly- human cardiovascular disease,” Annals of Medicine, vol. 36, morphism and human fat cell lipolysis,” International Journal supplement 1, pp. 64–69, 2004. of Obesity, vol. 25, no. 11, pp. 1599–1603, 2001. [82]C.Bell,N.R.Stob,andD.R.Seals,“Thermogenicre- [67] K. Bengtsson, O. Melander, M. Orho-Melander et al., “Poly- sponsiveness to nonspecific β-adrenergic stimulation is not morphism in the β1-adrenergic receptor gene and hyper- related to genetic variation in codon 16 of the β2-adrenergic tension,” Circulation, vol. 104, no. 2, pp. 187–190, 2001. receptor,” American Journal of Physiology, vol. 290, no. 4, pp. [68]L.M.Humma,B.J.Puckett,H.E.Richardsonetal.,“Effects E703–E707, 2006. of beta-adrenoceptor genetic polymorphisms on resting [83] K. Clement, C. Vaisse, B. S. J. Manning et al., “Genetic hemodynamics in patients undergoing diagnostic testing for variationintheβ3-adrenergic receptor and an increased ischemia,” American Journal of Cardiology,vol.88,no.9,pp. capacity to gain weight in patients with morbid obesity,” The 1034–1037, 2001. New England Journal of Medicine, vol. 333, no. 6, pp. 352–354, [69] B. Kortner, A. Wolf, D. Wendt, U. Beisiegel, and D. Evans, 1995. β “Lack of association between a human -2 adrenoceptor gene [84] N. Sakane, T. Yoshida, T. Umekawa, A. Kogure, Y. Takakura, polymorphism (gln27glu) and morbid obesity,” International andM.Kondo,“Effects of Trp64Arg mutation in the Journal of Obesity, vol. 23, no. 10, pp. 1099–1100, 1999. β3-adrenergic receptor gene on weight loss, body fat [70] N. R. Stob, D. R. Seals, J. Jensen et al., “Increased thermogenic distribution, glycemic control, and insulin resistance in responsiveness to intravenous β-adrenergic stimulation in obese type 2 diabetic patients,” Diabetes Care, vol. 20, no. 12, habitually exercising humans is not related to skeletal muscle pp. 1887–1890, 1997. β 2-adrenergic receptor density,” Experimental Physiology, [85] T. Umekawa, T. Yoshida, N. Sakane, A. Kogure, M. Kondo, vol. 92, no. 5, pp. 823–830, 2007. and H. Honjyo, “Trp64Arg mutation of β3-adrenoceptor [71] A. Meirhaeghe, N. Helbecque, D. Cottel, and P. Amouyel, gene deteriorates lipolysis induced by β3-adrenoceptor β “Impact of polymorphisms of the human 2-adrenoceptor agonist in human omental adipocytes,” Diabetes, vol. 48, no. gene on obesity in a French population,” International 1, pp. 117–120, 1999. Journal of Obesity, vol. 24, no. 3, pp. 382–387, 2000. [86] K. Endo, H. Yanagi, C. Hirano, H. Hamaguchi, S. Tsuchiya, [72] K. Bengtsson, M. Orho-Melander, O. Melander et al., β and S. Tomura, “Association of Trp64Arg polymorphism “ 2-adrenergic receptor gene variation and hypertension in of the β3-adrenergic receptor gene and no association of subjects with type 2 diabetes,” Hypertension,vol.37,no.5, Gln223Arg polymorphism of the leptin receptor gene in pp. 1303–1308, 2001. Japanese schoolchildren with obesity,” International Journal [73] H. Jiao, I. Dahlman, P. Eriksson, J. Kere, and P. Arner, “A of Obesity, vol. 24, no. 4, pp. 443–449, 2000. common β2-adrenoceptor gene haplotype protects against [87] T. Oizumi, M. Daimon, T. Saitoh et al., “Genotype Arg/Arg, obesity in Swedish women,” Obesity Research, vol. 13, no. 10, but not Trp/Arg, of the Trp64Arg polymorphism of the pp. 1645–1650, 2005. β β 3-adrenergic receptor is associated with type 2 diabetes and [74] A. Petrone, S. Zavarella, G. Iacobellis et al., “Association of 2 obesity in a large Japanese sample,” Diabetes Care, vol. 24, adrenergic receptor polymorphisms and related haplotypes no. 9, pp. 1579–1583, 2001. with triglyceride and LDL-cholesterol levels,” European Journal of Human Genetics, vol. 14, no. 1, pp. 94–100, 2006. [88] A. P. Gjesing, G. Andersen, K. S. Burgdorf et al., “Studies of the associations between functional β -adrenergic receptor [75] A. P. Gjesing, T. Sparsø, K. Borch-Johnsen et al., “No 2 variants and obesity, hypertension and type 2 diabetes consistent effect of ADRB2 haplotypes on obesity, in 7,808 white subjects,” Diabetologia, vol. 50, no. 3, pp. hypertension and quantitative traits of body fatness 563–568, 2007. and blood pressure among 6,514 adult danes,” PLoS One, vol. 4, no. 9, Article ID e7206, 2009. [89]N.Kurokawa,K.Nakai,S.Kameo,Z.M.Liu,andH.Satoh, β [76] P. Eriksson, I. Dahlman, M. Ryden, J. Hoffstedt,andP.Arner, “Association of BMI with the 3-adrenergic receptor gene “Relationship between β-2 adrenoceptor gene haplotypes polymorphism in Japanese: meta-analysis,” Obesity Research, and adipocyte lipolysis in women,” International Journal of vol. 9, no. 12, pp. 741–745, 2001. Obesity, vol. 28, no. 2, pp. 185–190, 2004. [90] P. Arner, “The β3-adrenergic receptor—a cause and cure of [77] S. A. Green, J. Turki, M. Innis, and S. B. Liggett, “Amino- obesity?” The New England Journal of Medicine, vol. 333, no. terminal polymorphisms of the human β2-adrenergic 6, pp. 382–383, 1995. receptor impart distinct agonist-promoted regulatory prop- [91] J. Hoffstedt, O. Poirier, A. Thorne¨ et al., “Polymorphism of erties,” Biochemistry, vol. 33, no. 32, pp. 9414–9419, 1994. the human β3-adrenoceptor gene forms a well-conserved [78] V. Dishy, G. G. Sofowora, H.-G. Xie et al., “The effect of haplotype that is associated with moderate obesity and common polymorphisms of the β2-adrenergic receptor on altered receptor function,” Diabetes,vol.48,no.1,pp. agonist-mediated vascular desensitization,” The New England 203–205, 1999. Journal of Medicine, vol. 345, no. 14, pp. 1030–1035, 2001. [92] J. Walston, K. Silver, C. Bogardus et al., “Time of onset [79] G. Gratze, J. Fortin, R. Labugger et al., “β-2 adrenergic of non-insulin-dependent diabetes mellitus and genetic receptor variants affect resting blood pressure and variationintheβ3-adrenergic-receptor gene,” The New agonist-induced vasodilation in young adult Caucasians,” England Journal of Medicine, vol. 333, no. 6, pp. 343–347, Hypertension, vol. 33, no. 6, pp. 1425–1430, 1999. 1995. [80] B. D. Hoit, D. P. Suresh, L. Craft, R. A. Walsh, and S. B. [93] D. L. Ellsworth, S. A. Coady, W. Chen, S. R. Srinivasan, Liggett, “β2-adrenergic receptor polymorphisms at amino E. Boerwinkle, and G. S. Berenson, “Interactive effects acid 16 differentially influence agonist-stimulated blood between polymorphisms in the β-adrenergic receptors and pressure and peripheral blood flow in normal individuals,” longitudinal changes in obesity,” Obesity Research, vol. 13, American Heart Journal, vol. 139, no. 3, pp. 537–542, 2000. no. 3, pp. 519–526, 2005. 10 Journal of Obesity

[94]I.C.Trombetta,L.T.Batalha,M.U.P.B.Rondonetal., [110] C. Zoccali, F. Mallamaci, S. Parlongo et al., “Plasma “Gly16+Glu27 β2-adrenoceptor polymorphisms cause in- norepinephrine predicts survival and incident cardiovascular creased forearm blood flow responses to mental stress and events in patients with end-stage renal disease,” Circulation, handgrip in humans,” Journal of Applied Physiology, vol. 98, vol. 105, no. 11, pp. 1354–1359, 2002. no. 3, pp. 787–794, 2005. [111] M. Petersson, P. Friberg, G. Eisenhofer, G. Lambert, and [95] A. Sandilands, G. Yeo, M. J. Brown, and K. M. O’Shaugh- B. Rundqvist, “Long-term outcome in relation to renal nessy, “Functional responses of human β1adrenoceptors sympathetic activity in patients with chronic heart failure,” with defined haplotypes for the common 389R >G and 49S > European Heart Journal, vol. 26, no. 9, pp. 906–913, 2005. G polymorphisms,” Pharmacogenetics,vol.14,no.6,pp. [112] K. Masuo, T. Katsuya, K. Sugimoto et al., “High plasma 343–349, 2004. norepinephrine levels associated with β2-adrenoceptor [96] M. Tomaszewski, N. J. R. Brain, F. J. Charchar et al., polymorphisms predict future renal damage in nonobese “Essential hypertension and β2-adrenergic receptor gene: normotensive individuals,” Hypertension Research, vol. 30, linkage and association analysis,” Hypertension, vol. 40, no. no. 6, pp. 503–511, 2007. 3, pp. 286–291, 2002. [113] A. Ksiazek and W. Załuska, “Sympathetic overactivity in [97] S. M. Herrmann, V. Nicaud, L. Tiret et al., “Polymorphisms uremia,” Journal of Renal Nutrition,vol.18,no.1,pp. of the β2-adrenoceptor (ADRB2) gene and essential 118–121, 2008. hypertension: the ECTIM and PEGASE studies,” Journal of [114] H. Krum, M. Schlaich, R. Whitbourn et al., “Catheter-based Hypertension, vol. 20, no. 2, pp. 229–235, 2002. renal sympathetic denervation for resistant hypertension: a [98] K. Masuo, H. Mikami, T. Ogihara, and M. L. Tuck, multicentre safety and proof-of-principle cohort study,” The “Sympathetic nerve hyperactivity precedes hyperinsulinemia Lancet, vol. 373, no. 9671, pp. 1275–1281, 2009. and blood pressure elevation in a young, nonobese Japanese [115] M. P. Schlaich, P. A. Sobotka, H. Krum, E. Lambert, and M. population,” American Journal of Hypertension, vol. 10, no. D. Esler, “Renal sympathetic-nerve ablation for uncontrolled 1, pp. 77–83, 1997. hypertension,” The New England Journal of Medicine, vol. [99] W. B. Kannel and D. L. McGee, “Diabetes and cardiovascular 361, no. 9, pp. 932–934, 2009. disease. The Framingham study,” TheJournaloftheAmerican Medical Association, vol. 241, no. 19, pp. 2035–2038, 1979. [100] V. Frighi, I. Stratton, R. Holman, D. Matthews, A. Neil, and R. Turner, “Hypertension in Diabetes Study (HDS): II. Increased risk of cardiovascular complications in hypertensive type 2 diabetic patients,” Journal of Hypertension, vol. 11, no. 3, pp. 319–325, 1993. [101] G. J. J. Tack, P. Smits, J. J. Willemsen, J. W. M. Lenders, T. Thien, and J. A. Lutterman, “Effects of insulin on vascular tone and sympathetic nervous system in NIDDM,” Diabetes, vol. 45, no. 1, pp. 15–22, 1996. [102] K. Masuo, H. Rakugi, T. Ogihara, M. D. Esler, and G. W. Lambert, “Cardiovascular and renal complications of type 2 diabetes in obesity: role of sympathetic nerve activity and insulin resistance,” Current Diabetes Reviews, vol. 6, no. 2, pp. 58–67, 2010. [103] S. Julius and S. Nesbitt, “Sympathetic overactivity in hypertension a moving target,” American Journal of Hypertension, vol. 9, no. 11, pp. 113s–120s, 1996. [104] G. Mancia, G. Grassi, C. Giannattasio, and G. Seravalle, “Sympathetic activation in the pathogenesis of hypertension and progression of organ damage,” Hypertension, vol. 34, no. 4, part 2, pp. 724–728, 1999. [105] M. Esler, G. Jennings, and G. Lambert, “Noradrenaline release and the pathophysiology of primary human hypertension,” American Journal of Hypertension, vol. 2, no. 3, part 2, pp. 140S–146S, 1989. [106] J. A. Joles and H. A. Koomans, “Causes and consequences of increased sympathetic activity in renal disease,” Hypertension, vol. 43, no. 4, pp. 699–706, 2004. [107] M. P. Schlaich, F. Socratous, S. Hennebry et al., “Sympathetic activation in chronic renal failure,” Journal of the American Society of Nephrology, vol. 20, no. 5, pp. 933–939, 2009. [108] R. E. Bulger, T. J. Burke, and R. E. Cronin, “Morphology of norepinephrine-induced acute renal failure in the dog,” Anatomical Record, vol. 214, no. 4, pp. 341–347, 1986. [109] K. Amann, A. Koch, J. Hofstetter et al., “Glomerulosclerosis and progression: effect of subantihypertensive doses of α and β blockers,” Kidney International,vol.60,no.4,pp. 1309–1323, 2001. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 456347, 8 pages doi:10.1155/2011/456347

Review Article Interleukin-15, IL-15 Receptor-Alpha, and Obesity: Concordance of Laboratory Animal and Human Genetic Studies

LeBris S. Quinn and Barbara G. Anderson

Geriatric Research, Education, and Clinical Center, VA Puget Sound Health Care System, Seattle Institute for Biomedical and Clinical Research, and Division of Gerontology and Geriatric Medicine, Department of Medicine, University of Washington, 1660 S. Columbian Way, Seattle, WA 98108, USA

Correspondence should be addressed to LeBris S. Quinn, [email protected]

Received 20 November 2010; Revised 18 January 2011; Accepted 27 January 2011

Academic Editor: P. Trayhurn

Copyright © 2011 L. S. Quinn and B. G. Anderson. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Interleukin-15 (IL-15) is a cytokine which inhibits lipid deposition in cultured adipocytes and decreases adipose tissue deposition in laboratory rodents. In human subjects, negative correlations between circulating IL-15 levels and both total and abdominal fat have been demonstrated. Deletions of IL15 in humans and mice are associated with obesity, while gain-of-function IL-15 overexpressing mice are resistant to diet-induced obesity. IL-15 is highly (but not exclusively) expressed at the mRNA level in skeletal muscle tissue, and the regulation of IL-15 translation and secretion is complex. Conﬂicting evidence exists concerning whether circulating IL-15 is released from skeletal muscle tissue in response to exercise or other physiological stimuli. The IL-15 receptor-alpha (IL-15Rα) subunit has a complex biochemistry, encoding both membrane-bound and soluble forms which can modulate IL-15 secretion and bioactivity. The gene encoding this receptor, IL15RA, resides on human chromosome 10p, a location linked to obesity and type-2 diabetes. Several single-nucleotide polymorphisms (SNPs) in human IL15RA and IL15 correlate with adiposity and markers of the metabolic syndrome. Genetic variation in IL15RA may modulate IL-15 bioavailability, which in turn regulates adiposity. Thus, IL-15 and the IL-15Rα may be novel targets for pharmacologic control of obesity in the human population.

1. Introduction human subjects will be reviewed. Additionally, the complex regulation of IL-15 expression and secretion will be summa- The increased incidence of obesity in both developed and rized. The specific receptor for IL-15, IL-15 receptor-alpha developing nations is a major public health problem [1]. (IL-15Rα), appears to regulate IL-15 secretion, stability, Altered energy balance resulting in obesity is believed to and activity [8, 9]. Several single-nucleotide polymorphisms be causal in the development of the metabolic syndrome (SNPs) in human IL15 and IL15RA have been described and type-2 diabetes mellitus, conditions whose incidence is which correlate with adiposity and markers of the metabolic also rising at alarming rates [2, 3]. The economic burden syndrome [10–12]. These findings suggest a model in which of obesity and diabetes in the U.S. is currently in excess of genetic variation in IL15RA could modulate IL-15 activity $200 billion annually, and it is expected to rise [1]. Diet and bioavailability, which in turn regulate adiposity. More- and lifestyle modifications are often unsuccessful in reduc- over, because IL-15 mRNA is highly expressed in skeletal ing obesity [4–6], and presently available pharmacologic muscle tissue, some authors have suggested that IL-15 may treatments are problematic due to serious adverse effects function as a muscle-derived endocrine factor, or “myokine”, [7]. Therefore, research involving novel pathways to control which can modulate body composition [13–16]. Evidence in adiposity is warranted. support of, and against, this hypothesis will be discussed. Here, evidence that interleukin-15 (IL-15) can inhibit When human genetic, cell culture, and laboratory animal adipose tissue deposition in both laboratory animals and studies are taken together, it is clear that the IL-15/IL-15Rα 2 Journal of Obesity axis regulates adiposity. Modulation of this axis should be concentrations in serum than in amniotic fluid, consistent explored a novel target for control of obesity in the human with literature indicating many cytokines do not cross population. the placental barrier. However, studies of whether gravid animals or pregnant women display increased circulating levels of IL-15 compared to nongravid or nonpregnant 2. Molecular Genetics and Expression of females are lacking. In any case, such an observation would IL-15 with Regard to Adipose Tissue not apply to males and does not preclude skeletal muscle or other tissues as additional sources of IL-15 in both IL-15 is a 14 kDa cytokine that was originally isolated genders. on the basis of its ability to support natural killer (NK) Two IL-15 mRNA isoforms are generated from a single T-lymphocyte proliferation [17]. IL-15 was subsequently IL15 gene in humans and mice [28, 29]. The two isoforms foundtobeexpressedbymonocytes,macrophages,and differ in the lengths of the signal peptides and have been other cell types involved in immunity and to regulate a designated long signal peptide (LSP) and short signal peptide variety of processes comprising both innate and adaptive (SSP)-IL-15 mRNA [28]. However, tissue expression and immunity reviewed in [18]. IL-15 exhibits both pro- and intracellular trafficking of these isoforms are dissimilar [28, anti-inflammatory actions in a variety of tissues and has 29]. SSP-IL-15 mRNA is highly transcribed in heart, and both positive anticancer effects via stimulation of NK cells, also expressed in thymus, testes, and appendix, whereas LSP- and deleterious effects such as involvement in inflammatory IL-15 mRNA is transcribed strongly in skeletal muscle and bowel disease (reviewed in [18, 19]). Inasmuch as the present placenta, and at lower levels in heart, lung, liver, thymus, paper is focused on the relationship of IL-15 to obesity, the and kidney [28]. SSP-IL-15 does not appear to be secreted, reader is referred to comprehensive reviews of the diverse and either functions intracellularly or is released following functions of IL-15 provided by Fehniger and Caligiuri [18] cell damage [28, 29]. LSP-IL-15 is secreted; however, the and Budagian et al. [19]. unusually long 48 amino acid signal peptide renders IL- IL-15 is structurally part of the 4-helix bundle protein 15 secretion extremely inefficient [28–31]. IL-15 protein family [17], whose members exhibit structural, but not expression is also regulated at the translational level [31–33]. necessarily sequence, homology [20]. This family of proteins LSP-IL-15 protein translation is impeded by multiple AUGs includes many cytokines which regulate the immune system (initiation codons) in the 5 untranslated region [31–33]. and also includes factors with actions outside of the immune Because of the inefficiency of IL-15 translation and secretion, system, such as IL-6, leptin, growth hormone, and erythro- correlations between IL-15 mRNA levels and secretion of poietin [20]. IL-15 is expressed at the mRNA level in a variety IL-15 protein are often weak (reviewed in [18, 33]). A of nonlymphoid tissues, with particularly high expression systematic comparison of IL-15 protein expression among in skeletal muscle and placenta [17]. IL-15 is also expressed tissues has not been conducted, and it is technically difficult abundantly in cardiac muscle, lung, liver, kidney, brain, and to demonstrate secretion from a specific tissue in vivo, pancreas [17]. In the placenta, IL-15 regulates a specific particularly in mouse models. subset of NK cells involved in endometrial decidualization Conflicting reports exist as to whether IL-15 is expressed [21]. by adipocytes. The mouse 3T3-L1 adipogenic cell line does In nonlymphoid tissues, IL-15 has been implicated in not express levels of IL-15 mRNA detectable by highly processes ranging from angiogenesis [22] to skeletal muscle sensitive real-time PCR at any stage of differentiation [14]. hypertrophy [23]. Conflicting reports [14, 24] exist regarding However, another study demonstrated primary pig adipocyte IL-15 mRNA expression in cultured adipocytes (reviewed cultures express IL-15 mRNA at low basal levels, which are below), and a systematic study of IL-15 expression in adipose upregulated following stimulation with interferon-γ [24]. tissue in vivo has not been conducted. Both humans and Whether IL-15 protein was produced or released into the laboratory mice exhibit detectable levels of IL-15 in the culture medium was not determined. Differences between circulation, (for example, [9, 10, 15, 25]), allowing for rodents and swine, and/or between adipogenic cell lines and the possibility that IL-15 can exert endocrine (as well as primary cultures, as well as between adipocytes in vivo and paracrine) effects on cell types which do not express IL- in vitro, are quite possible. 15 itself. However, the tissues from which circulating IL-15 In obesity, adipose tissue develops an inflammatory originates are unknown. Because of the high expression of environment due to infiltrating macrophages which are a IL-15 in skeletal muscle and evidence that other cytokines source of numerous proinflammatory cytokines [34, 35]. such as IL-6 are released from muscle following physical In a mouse strain highly susceptible to oxidative stress, activity, some investigators have suggested IL-15 functions as high dietary calcium in conjunction with an obesigenic a myokine which exerts positive effects on body composition diet significantly stimulated IL-15 mRNA expression in via an endocrine mechanism [13–16]. Conflicting evidence both visceral fat and skeletal muscle tissue [36], but it for IL-15 as a myokine is reviewed below. was not determined which cell types were responsible for As mentioned above, another tissue which exhibits high the upregulated IL-15 message. A systematic study of IL- expression of both IL-15 mRNA and protein is placenta [17, 15 mRNA and protein expression in various depots of fat 26, 27]. In a study comparing concentrations of numerous tissue, in different physiological conditions, has not been cytokines in amniotic fluid and sera of normal pregnant performed. Thus, whether adipocytes and/or adipose tissue women, Chow et al. [27] reported significantly higher IL-15 can express and secrete IL-15 protein in basal conditions or Journal of Obesity 3 in inflammatory challenges associated with obesity remains IL-15 Tg mice described above [40, 42], no effects of loss unclear. or overexpression of IL-15 on food intake were observed. Therefore, the effects of IL-15 on adipose tissue are not likely to be due to an indirect effect of modified energy 3. Effects of Interleukin-15 on Adipose Tissue intake. However, one report indicated chronic treatment of rats with IL-15 slightly inhibited intestinal absorption of The IL-15 gene (IL15; human accession number U14407) triaclglycerols specifically [43]. Lean body mass is unaffected is mapped to human chromosome 4q31 and the central by IL-15 [13, 40, 42], indicating IL-15 does not induce region of chromosome 8 in mice [37]. A recent genome-wide a cachectic state; indeed, recombinant IL-15 injection can survey of human copy number variations which correlated preventlossofskeletalmusclemassinrodentmodelsof with obesity revealed a large (2.1 Mb) deletion of a region cancer cachexia [44]. which included both IL15 and the gene encoding the Recombinant IL-15 has been administered by injection mitochondrial uncoupling protein UCP1 [38]. Since the role into rodent genetic obesity models. IL-15 injection inhibited of UCP1 in modulating energy balance is well described fat deposition in both wild-type and leptin-deficient obese [39], the potential contribution of loss of IL15 by this large (ob/ob)mice[45]. IL-15 administration to lean rats also deletion was not additionally considered. However, research inhibited fat deposition, but it was unable to inhibit fat findings from adipogenic cell cultures, laboratory animals, deposition in leptin receptor-deficient obese (fa/fa)Zucker and human subjects all suggest IL-15 may also function as an rats [45]. Obese, but not lean, rats exhibited significant antiobesigenic factor. decreases in adipose tissue expression of mRNA for two In a study of human subjects comprising a wide range of the subunits of the heterotrimeric IL-15 receptor, the of body mass indices (BMI), Nielsen et al. [15]found IL-2 receptor beta and gamma subunits (IL-2Rβ and IL- negative associations between plasma IL-15 concentrations 2Rγ), while expression of adipose tissue IL-15Rα mRNA and BMI (P<.001), total fat mass (P<.001), trunk fat was unchanged [45]. This observation suggests that adipose mass (P<.01), and limb fat mass (P<.05). Negative tissue of obese Zucker rats failed to respond to IL-15 because associations between muscle IL-15 mRNA and obesity the signaling subunits (IL-2Rβ and IL-2Rγ) of the IL-15 parameters were also observed in that study [15]. A similar receptor were downregulated in adipose tissue in this strain, finding was reported by Barra et al. [40] who observed obese and that the effect of IL-15 on adipose tissue is direct. human subjects exhibited lower circulating IL-15 levels than The direct effect of IL-15 on adipose tissue was confirmed lean subjects. However, Christiansen et al. [41]reported using adipogenic cell cultures derived from several mam- decreased circulating IL-15 concentrations following diet- malian species, including human [14, 40, 46]. Recombinant induced weight loss in obese human subjects. Two SNPs IL-15 administration inhibited preadipocyte differentiation in human IL15 (rs1589241 and rs1057972) are associated and lipid deposition in the immortalized mouse 3T3-L1 with various predictors of the metabolic syndrome, BMI, and cell line [14]. Moreover, in differentiated 3T3-L1 adipocytes, muscle strength [10, 11]. These SNPs are located in the 5 and IL-15 dose-dependently stimulated secretion of the insulin- 3 untranslated regions (UTRs) of the gene, suggesting they sensitizing and antiobesigenic factor adiponectin [14]. Sim- could modulate IL-15 expression. ilar results were observed by another laboratory using cul- Mice with targeted deletion of IL15 (IL-15KO mice) tured primary porcine adipocytes, in which IL-15 potently exhibit higher amounts of body fat than control mice stimulated lipolysis and modestly inhibited lipogenesis [46]. [40]. Conversely, transgenic mice which were engineered for Finally, Barra et al. [40] found that administration of IL-15 elevated circulating levels of IL-15 (IL-15 Tg mice), expressed to lipoaspirate-derived human adipocyte cultures inhibited from a skeletal muscle-specific promoter, exhibited lower lipid deposition. The molecular pathways mediating the levels of body fat than closely related controls, and were effects of IL-15 on adipose tissue have not been character- resistant to diet-induced obesity [42]. In the same study, ized in detail. One study [47] suggested IL-15 upregulates mice which expressed high intramuscular levels of IL-15 but expression of calcineurin mRNA, a factor which inhibits which did not exhibit elevated serum IL-15 levels showed no adipocyte differentiation. Further work is needed to confirm differences in adiposity compared to controls, suggesting IL- and expand this observation. 15 must be secreted into the circulation to exert its effects on adipose tissue [42]. Since this is an artificially constructed system, it can model, but does not prove, the hypothesis that 4. IL-15 as a Potential Myokine muscle-derived IL-15 acts as a myokine in the native state. IL-15 has also been introduced into wild-type laboratory In contrast to adipose tissue, IL-15 mRNA and protein rodents by injection of recombinant IL-15 protein [13, expression has been observed consistently in skeletal mus- 40], by adenoviral expression vectors [40], and by DNA cle and skeletal muscle-derived culture systems [23, 48– electrotransfer into skeletal muscle [15]. In these studies, 51], where it functions to modulate myofibrillar protein IL-15 administration reduced fat mass by as much as 30% dynamics [23, 44]. At the protein level, IL-15 has been in normal rodents and 10% in obese rodents. Importantly, immunolocalized to human skeletal muscle fibers in tissue inhibition of fat deposition was observed in the absence of an sections which contained few IL-15-positive infiltrating effect of IL-15 on food consumption [13, 40]. Additionally, in cells [48]. Nielsen et al. [49]detectedIL-15proteinby the studies of both IL-15KO mice and the gain-of-function immunohistochemistry and Western blotting in human 4 Journal of Obesity skeletal muscle with both type-1 and type-2 fiber dominance, a program of intensive (hypocaloric) dietary intervention, which did not correlate with the respective levels of IL- aerobic exercise, or the combination of diet and exercise. 15 mRNA expression in these muscles. Expression of IL- Compared to baseline, diet alone and the combination of diet 15 mRNA and biologically active IL-15 protein has been and exercise significantly decreased circulating IL-15 levels, detected in primary human myogenic cultures [48], human whereas aerobic exercise alone had no effect. However, given rhabodomyosarcoma-derived cell lines [50], and a rat clonal the extremely hypocaloric nature of the dietary intervention myogenic cell line [51]. IL-15 mRNA expression is low, but (600–800 kcal/day), it is possible some loss of muscle tissue detectable, in mouse C2C12 skeletal myogenic cultures at occurred, which was not examined. Beavers et al. [58] the myoblast stage, but it is induced about 10-fold upon similarly reported no effect of a 12-month program of differentiation [23]. Therefore, published evidence indicates combined aerobic, strength and balance intervention on IL-15 is expressed by skeletal muscle fibers themselves, not circulating IL-15 levels in frail elderly subjects. It is possible vascular, connective tissue, or lymphoid infiltrating cells that differences among these studies were due to the use present in muscle and in primary cultures. However, it of highly trained versus untrained, obese, and/or elderly is unclear if these observations translate to meaningful subjects and/or to the difference between aerobic versus contributions to circulating IL-15 levels by skeletal muscle resistance exercise. Therefore, evidence for the hypothesis tissue. that IL-15 is released following exercise is mixed and largely Several cytokines, most notably interleukins −6, −8, and negative. Moreover, an increase in circulating IL-15 levels −10, are released from skeletal muscle following physical after exercise is only circumstantial evidence of its derivation activity, resulting in large changes in plasma concentrations from skeletal muscle tissue. For example, although skeletal of such “myokines” [52, 53]. One study found that in human muscle expresses the myostatin-inhibiting factor follistatin, a skeletal muscle, IL-15 was the mostly highly expressed of recent study provided evidence that the postexercise rise in all cytokines measured at the mRNA level [52]. Since both circulating follistatin is mostly likely derived from the liver IL-15 and physical exercise have positive effects on body rather than from muscle tissue [59]. composition, it is reasonable to hypothesize IL-15 release The concept of IL-15 as a “myokine” does not necessarily following exercise may mediate some of the beneficial effects imply that it is released by exercise, rather it implies simply of physical activity [16, 54]. However, there are conflicting that the factor is derived from skeletal muscle tissue and is reports as to whether physical activity affects IL-15 expres- released in quantities sufficient to have an endocrine effect sion in skeletal muscle and/or increases circulating IL-15 on other cell types. Such release also does not necessarily levels. In a study of young untrained and 10-week-trained imply that expression of the purported myokine is specific human subjects, Riechman et al. [10] demonstrated an to skeletal muscle tissue. On the other hand, skeletal muscle increase in plasma IL-15 protein levels immediately following tissue comprises almost 50% of body mass and is specialized whole-body resistance exercise and speculated that IL-15 was for protein synthesis, if not secretion, so a small release released following exercise via microtears in muscle fibers. of a bioactive factor from each muscle could have a large However, there was no effect of training on the postexercise systemic effect. Sporadic reports of other inflammatory, release of IL-15 in that study, casting doubt upon muscle hormonal, or nutritional factors which affect muscle IL-15 damage as the mechanism. In contrast, a similar study using mRNA expression and circulating IL-15 protein levels have untrained human subjects reported no increase in muscle or been published. Using elderly human male subjects, Lambert circulating IL-15 protein levels at intervals from 6 to 48 hours et al. [60] administered the synthetic progestin megestrol following an intensive resistance exercise protocol which acetate at 800 mg/d for 12 wk, with or without testosterone involved only the quadriceps muscles; however, increases in (100 mg/wk), resistance training, or the combination of quadriceps muscle IL-15 mRNA expression were observed 24 resistance training and testosterone. Progestin ingestion, but hours after exercise [49]. Discrepancies between these studies no other treatment, caused highly significant increases in may be due to the differences in the intervals between exercise circulating IL-15 levels, but this treatment did not correlate and blood sampling. withchangesinmusclemassorbodycomposition.As Prior training or physical condition may modulate the mentioned above, dietary calcium significantly stimulated purported release of IL-15 from skeletal muscle tissue by IL-15 mRNA expression in murine skeletal muscle tissue exercise. Using strength-trained human subjects, Nieman (as well as visceral fat) in obesigenic conditions and oxida- et al. [53] observed no changes in muscle IL-15 mRNA tive stress, which was interpreted as an increase in anti- following two hours of intensive weight training. Similarly, inflammatory cytokine expression due to calcium-mediated Ostrowski et al. [55] observed no changes in plasma IL- inhibition of 1,25-dihydroxyvitamin D3 [36]. In myogenic 15 following 2 h of treadmill running by 2 male athletes, cell cultures, overexpression of an orphan nuclear hormone and Andersson et al. [56] found no acute effects of com- receptor which is highly expressed in muscle, retinoid- petition on IL-15 levels in elite female soccer athletes. In related orphan receptor gamma, upregulated both IL-15 and contrast, using sedentary postmenopausal women, Prestes myogenin mRNA, as well as expression of several genes et al. [57] reported an increase in plasma IL-15 levels 48 which regulate lipid and carbohydrate metabolism, insulin hours following the first session of resistance training but no sensitivity, and reactive oxygen species [61]. Cultured rat effect following 16 weeks of training. Finally, Christiansen myocytes upregulated IL-15 mRNA in response to both the et al. [41] studied chronic (not acute) plasma cytokine inflammatory factor interferon-γ and the anti-inflammatory levels in obese, physically inactive subjects subjected to cytokine IL-4 [51], while primary human myoblast cultures Journal of Obesity 5 increased expression of both intracellular and secreted IL- expresses all three IL-15 receptor subunits at the mRNA level 15 protein in response to several inflammatory mediators [45]. (interferon-γ, interleukin-1α, interleukin-1β,TNF-α,and Unexpectedly, mice in which the IL-15Rα is deleted lipopolysaccharide) [48]. These observations suggest muscle (IL15RαKO mice) are leaner, rather than fatter, than controls IL-15 expression is modulated by dietary, hormonal, and [68]. This observation suggests a complex role for this inflammatory status, but, as in the experiments dealing with receptor subunit in control of body fat, and is consistent with exercise, a clear understanding of the regulation of IL-15 the idea that the IL-15Rα can function in roles other than a expression by such factors in skeletal muscle tissue is lacking. membrane-bound receptor component. Increased age is an important predictor of obesity and The molecular genetics of IL-15Rα are complex. IL15RA the metabolic syndrome [3, 62]. Reports in animal models (human accession number U31628) is located on mouse on the effects of age on IL-15 expression in muscle and in chromosome 2 and human chromosome 10p [69]. Loci serum are conflicting. Pistilli et al. [63] found that IL-15 on human chromosome 10p have been strongly linked to mRNA was elevated in both slow and fast aging rat muscles both obesity and type-2 diabetes [70, 71]. Human and compared to young muscles. The same study found that IL- mouse IL15RA exhibit similar genomic structures ([67, 15 mRNA was elevated in atrophied slow soleus muscles of 69] and reviewed in [19]). IL15RA has an efficient signal young rats but not in the fast plantaris muscle. A similar sequence (exon 1), a ligand-binding domain (exon 2), a effect of aging and immobilization recovery on IL-15 mRNA transmembrane domain (exon 6), and, as mentioned, a in rat muscles was reported in another study [64]. However, fairly short cytoplasmic domain (exon 7). Exon 2 encodes given the complex regulation of IL-15 described above, it is a conserved protein-binding region known as a “sushi” unclear whether these changes in IL-15 mRNA expression domain, which is responsible for the binding affinity of IL- reflect similar changes in muscle IL-15 protein expression 15Rα for the IL-15 ligand [67, 69]. A number of IL-15Rα and secretion in these physiological states. Indeed, in mice, mRNA splice variants have been reported in both mouse both muscle and serum IL-15 protein levels were reported and human, some of which encode soluble forms which may to decline progressively with increasing age; however, this modulate IL-15 secretion and bioavailability [19]. Because decline was not accompanied by decreases in transcription both isoforms of IL-15 have inefficient signal sequences, of either SSP- or LSP-IL-15 mRNA [25]. Rather, while recent experimental evidence suggests IL-15 is brought to expression of the membrane-associated form of IL-15Rα the cell surface by intracellular association with either sIL- (mbIL-15Rα) did not change with age, an age-related decline 15Rα or mbIL-15Rα, both of which would include the in expression of muscle soluble IL-15Rα (sIL-15Rα)mRNA efficient exon-1-encoded IL-15Rα signal sequence [8]. Two in muscle tissue was demonstrated [25], suggesting decreased forms of sIL-15Rα have been characterized. One, which expression of this factor (reviewed below) could lead to age- arises by differential splicing, comprises only exons 1 and related declines in IL-15 secretion from muscle. In another 2, so does not contain a transmembrane domain [9]. IL-15 study [65], a similar decline in muscle IL-15 protein levels secretion as an IL-15/sIL-15 Rα complex is thereby facilitated (and muscle IL-15Rα mRNA expression) in ad lib-fed rats following intracellular binding of IL-15 and this truncated was described, but no declines in IL-15 protein levels or receptor isoform [8, 9]. Another form of sIL-15Rα arises by IL-15Rα mRNA expression were observed in long-lived, proteolytic cleavage of mbIL-15Rα (which can be complexed calorie-restricted rats [65]. Intriguingly, Gangemi et al. [66] to IL-15) by the matrix metalloproteinase TNF-α converting observed that while serum IL-15 concentrations exhibited enzyme (TACE) [72]. Release of this form of circulating a downward trend with age in unselected human subjects, IL-15/sIL-15Rα complex is dependent upon TACE activity, individuals with unusually long lifespans (95 to more than which in turn is correlated with obesity and insulin resistance 100 years) and still living independently had significantly in mice and humans [73, 74]. elevated circulating IL-15 levels, suggesting elevated IL- Uncomplexed IL-15 and the two types of IL-15/sIL- 15 levels conferred some protection against age-related 15Rα complexes appear to be in a dynamic equilibrium in illness. the circulation, and the two soluble receptor variants can compete for binding of IL-15 [9]. However, while both forms facilitate IL-15 secretion and can increase IL-15 half-life in 5. The Soluble IL-15 Receptor-Alpha as the circulation, the two variants have differential effects on a Modulator of IL-15 Secretion IL-15 bioactivity; additionally, the two types of IL-15/sIL- 15Rα complexes can differentially bind heterodimeric versus IL-15 signaling is transduced either through a heterodimeric heterotrimeric IL-15 receptors [8, 9]. Therefore, regulation receptor comprising IL-2Rβ and IL-2Rγ, or through a of the various forms of sIL15Rα may be an important heterotrimeric receptor comprising mbIL-15Rα plus IL-2Rβ element modulating IL-15 secretion and bioactivity. and IL-2Rγ [67]. The IL-2Rβ and IL-2Rγ subunits are In this regard, human genetic studies by three separate responsible for signal transduction, while mbIL-15Rα has laboratories have identified several single-nucleotide poly- only a short cytoplasmic region and functions primarily to morphisms (SNPs) in IL15RA which impact muscularity, confer high-affinity binding to the receptor complex [67]. fat deposition, and markers of the metabolic syndrome Cells can express heterodimeric and heterotrimeric IL-15 [13, 14, 19]. For example, a SNP (rs2228059) in exon 3 receptor complexes simultaneously or express IL-15Rα in correlates with serum triglyceride levels in males [11]. An A the absence of the other two subunits [67]. Adipose tissue to G variation in the exon 5/intron border (SNP accession 6 Journal of Obesity rs3136618) is associated with the so-called normal weight [2] S. E. Kahn, R. L. Hull, and K. M. Utzschneider, “Mechanisms obese/De Lorenzo syndrome, characterizing young women linking obesity to insulin resistance and type 2 diabetes,” with normal BMIs but greater than 30% body fat and the Nature, vol. 444, no. 7121, pp. 840–846, 2006. presence of several markers of metabolic syndrome [12]. In a [3] M. I. Harris, K. M. Flegal, C. C. Cowie et al., “Prevalence separate study of male subjects by a different group, this same of diabetes, impaired fasting glucose, and impaired glucose SNP correlated with muscularity [10]. This SNP is close to tolerance in U.S. adults: the Third National Health and Nutrition Examination Survey, 1988–1994,” Diabetes Care, the site of mbIL-15Rα cleavage by TACE [72]. Whether IL- ff α vol. 21, no. 4, pp. 518–524, 1998. 15 is di erentially complexed with sIL-15R variants during [4]M.Rosenbaum,H.R.Kissileff,L.E.S.Mayer,J.Hirsch,and obesity has not been studied, but the existence of multiple R. L. Leibel, “Energy intake in weight-reduced humans,” Brain SNPs in IL15RA which correlate with body composition Research, vol. 1350, pp. 95–102, 2010. suggest this is a fertile area for future investigation. [5]M.W.Schwartz,S.C.Woods,R.J.Seeley,G.S.Barsh,D.G. Baskin, and R. L. Leibel, “Is the energy homeostasis system inherently biased toward weight gain?” Diabetes, vol. 52, no. 6. Effects of IL-15 via the Central Nervous 2, pp. 232–238, 2003. System (CNS) [6] R. Dalle Grave, S. Calugi, E. Centis, R. Marzocchi, M. El Ghoch, and G. Marchesini, “Lifestyle modification in the IL-15 and IL-15 receptor subunits are expressed in several management of the metabolic syndrome: achievements and regions of the CNS [68, 75, 76]. IL-15 has been reported challenges,” Diabetes, Metabolic Syndrome and Obesity: Targets to modulate non-REM sleep [75] and to play a protective and Therapy, vol. 3, pp. 373–385, 2010. role in neuroinflammation [76]. Recently, IL15RαKO mice [7] A. Astrup, “Drug management of obesity—efficacy versus were reported to be hyperphagic and hyperactive, with the safety,” The New England Journal of Medicine, vol. 363, no. 3, suggestion that IL-15 may control hypothalamic pathways pp. 288–290, 2010. which regulate energy balance [68]. This area is out of [8] C. Bergamaschi, M. Rosati, R. Jalah et al., “Intracellular α the scope of the present review. However, IL-15 expression, interaction of interleukin-15 with its receptor during trafficking and effects in the CNS are an additional, relatively production leads to mutual stabilization and increased bioactivity,” Journal of Biological Chemistry, vol. 283, no. 7, pp. unexplored area of investigation which could impact the 4189–4199, 2008. balance of energy expenditure and energy intake. [9] E. Bulanova, V. Budagian, E. Duitman et al., “Soluble interleukin (IL)-15Rα is generated by alternative splicing or 7. Conclusions proteolytic cleavage and forms functional complexes with IL- 15,” Journal of Biological Chemistry, vol. 282, no. 18, pp. Findings from cell culture, laboratory animal, and human 13167–13179, 2007. studies indicate IL-15 regulates adipose tissue deposition by a [10] S. E. Riechman, G. Balasekaran, S. M. Roth, and R. E. direct mechanism. There are conflicting reports concerning Ferrell, “Association of interleukin-15 protein and interleukin- 15 receptor genetic variation with resistance exercise training the tissue origin of circulating IL-15 and whether it can act responses,” Journal of Applied Physiology,vol.97,no.6,pp. as a circulating, muscle-derived “myokine”. Whatever the 2214–2219, 2004. tissue of origin, recent evidence supports the hypothesis that [11]E.E.Pistilli,J.M.Devaney,H.Gordish-Dressmanetal., circulating IL-15 levels and IL-15 bioactivity are determined “Interleukin-15 and interleukin-15Rα SNPs and associations by differential association of IL-15 with sIL-15Rα variants. with muscle, bone, and predictors of the metabolic syndrome,” The concordance of basic science findings and human Cytokine, vol. 43, no. 1, pp. 45–53, 2008. genetic studies suggests the sIL-15Rα, which in turn regulates [12] L. Di Renzo, M. Bigioni, F. G. Bottini et al., “Normal weight IL-15, is an important factor influencing body composition obese syndrome: role of single nucleotide polymorphism of and insulin sensitivity. The mechanism of IL-15 action on IL-15Rα and MTHFR 677C → T genes in the relationship adipose tissue is unknown. Therefore, further investigation between body composition and resting metabolic rate,” Euro- into the complex IL-15/IL-15Rα axis is needed to determine pean Review for Medical and Pharmacological Sciences, vol. 10, no. 5, pp. 235–245, 2006. if this system can be exploited for regulation of adiposity. [13] N. Carbo,´ J. Lopez-Soriano,´ P. Costelli et al., “Interleukin-15 mediates reciprocal regulation of adipose and muscle mass: a Acknowledgments potential role in body weight control,” Biochimica et Biophysica Acta, vol. 1526, no. 1, pp. 17–24, 2001. This work is Supported by NIH Grant no. RO1AG024136 [14] L. S. Quinn, L. Strait-Bodey, B. G. Anderson, J. M. Argiles,´ and from the National Institute on Aging to LSQ, Seattle Institute P. J. Havel, “Interleukin-15 stimulates adiponectin secretion for Biomedical and Clinical Research and by use of resources by 3T3-L1 adipocytes: evidence for a skeletal muscle-to-fat and facilities at the VA Puget Sound Health Care System, signaling pathway,” Cell Biology International,vol.29,no.6, Seattle, WA. pp. 449–457, 2005. [15] A. R. Nielsen, P. Hojman, C. Erikstrup et al., “Association between interleukin-15 and obesity: interleukin-15 as a poten- References tial regulator of fat mass,” Journal of Clinical Endocrinology and Metabolism, vol. 93, no. 11, pp. 4486–4493, 2008. [1] J. O. Hill, H. R. Wyatt, G. W. Reed, and J. C. Peters, “Obesity [16] B. K. Pedersen, “The diseasome of physical inactivity - and and the environment: where do we go from here?” Science, vol. the role of myokines in muscle-fat cross talk,” Journal of 299, no. 5608, pp. 853–855, 2003. Physiology, vol. 587, no. 23, pp. 5559–5568, 2009. Journal of Obesity 7

[17] K. H. Grabstein, J. Eisenman, K. Shanebeck et al., “Cloning of [32] H. Nishimura, J. Washizu, N. Nakamura, A. Enomoto, and Y. a T cell growth factor that interacts with the β chain of the Yoshikai, “Translational efficiency is up-regulated by alterna- interleukin-2 receptor,” Science, vol. 264, no. 5161, pp. 965– tive exon in murine IL-15 mRNA,” Journal of Immunology, vol. 968, 1994. 160, no. 2, pp. 936–942, 1998. [18] T. A. Fehniger and M. A. Caligiuri, “Interleukin 15: biology [33] Y. Tagaya, R. N. Bamford, A. P. DeFilippis, and T. A. and relevance to human disease,” Blood, vol. 97, no. 1, pp. 14– Waldmann, “IL-15: a pleiotropic cytokine with diverse recep- 32, 2001. tor/signaling pathways whose expression is controlled at [19] V. Budagian, E. Bulanova, R. Paus, and S. Bulfone-Paus, multiple levels,” Immunity, vol. 4, no. 4, pp. 329–336, 1996. “IL-15/IL-15 receptor biology: a guided tour through an [34] S. P. Weisberg, D. McCann, M. Desai, M. Rosenbaum, R. expanding universe,” Cytokine & Growth Factor Reviews, vol. L. Leibel, and A. W. Ferrante Jr., “Obesity is associated 17, no. 4, pp. 259–280, 2006. with macrophage accumulation in adipose tissue,” Journal of [20] N. D. Horseman and L. Y. Yu-Lee, “Transcriptional regulation Clinical Investigation, vol. 112, no. 12, pp. 1796–1808, 2003. by the helix bundle peptide hormones: growth hormone, [35] H. Xu, G. T. Barnes, Q. Yang et al., “Chronic inflammation in prolactin, and hematopoietic cytokines,” Endocrine Reviews, fat plays a crucial role in the development of obesity-related vol. 15, no. 5, pp. 627–649, 1994. insulin resistance,” Journal of Clinical Investigation, vol. 112, [21]S.Verma,S.E.Hiby,Y.W.Loke,andA.King,“Humandecid- no. 12, pp. 1821–1830, 2003. ual natural killer cells express the receptor for and respond to [36] X. Sun and M. B. Zemel, “Calcium and 1,25-dihydroxyvitamin the cytokine interleukin 15,” Biology of Reproduction, vol. 62, D3 regulation of adipokine expression,” Obesity,vol.15,no.2, no. 4, pp. 959–968, 2000. pp. 340–348, 2007. [22] A. L. Angiolillo, H. Kanegane, C. Sgadari, G. H. Reaman, [37]D.M.Anderson,L.Johnson,M.B.Glaccumetal.,“Chromo- and G. Tosato, “Interleukin-15 promotes angiogenesis in vivo,” somal assignment and genomic structure of Il15,” Genomics, Biochemical and Biophysical Research Communications, vol. vol. 25, no. 3, pp. 701–706, 1995. 233, no. 1, pp. 231–237, 1997. [38]K.Wang,W.-D.Li,J.T.Glessner,S.F.A.Grant,H. [23] L. S. Quinn, B. G. Anderson, R. H. Drivdahl, B. Alvarez, and J. Hakonarson, and R. A. Price, “Large copy-number variations M. Argiles,´ “Overexpression of interleukin-15 induces skeletal are enriched in cases with moderate to extreme obesity,” muscle hypertrophy in vitro: implications for treatment of Diabetes, vol. 59, no. 10, pp. 2690–2694, 2010. muscle wasting disorders,” Experimental Cell Research, vol. [39] L. P. Kozak and R. Anunciado-Koza, “UCP1: its involvement 280, no. 1, pp. 55–63, 2002. and utility in obesity,” International Journal of Obesity, vol. 32, [24] K. M. Ajuwon, S. K. Jacobi, J. L. Kuske, and M. E. no. 7, pp. S32–S38, 2008. Spurlock, “Interleukin-6 and interleukin-15 are selectively [40] N. G. Barra, S. Reid, R. MacKenzie et al., “Interleukin-15 regulated by lipopolysaccharide and interferon-γ in primary contributes to the regulation of murine adipose tissue and pig adipocytes,” American Journal of Physiology, vol. 286, no. human adipocytes,” Obesity, vol. 18, no. 8, pp. 1601–1607, 3, pp. R547–R553, 2004. 2010. [25] L. S. Quinn, B. G. Anderson, L. Strait-Bodey, and T. Wolden- Hanson, “Serum and muscle interleukin-15 levels decrease in [41] T. Christiansen, S. K. Paulsen, J. M. Bruun, S. B. Pedersen, and aging mice: correlation with declines in soluble interleukin-15 B. Richelsen, “Exercise training versus diet-induced weight- receptor alpha expression,” Experimental Gerontology, vol. 45, loss on metabolic risk factors and inflammatory markers in no. 2, pp. 106–112, 2010. obese subjects: a 12-week randomized intervention study,” [26] B. Toth, T. Haufe, C. Scholz et al., “Placental interleukin- American Journal of Physiology, vol. 298, no. 4, pp. E824–E831, 15 expression in recurrent miscarriage,” American Journal of 2010. Reproductive Immunology, vol. 64, no. 6, pp. 402–410, 2010. [42] L. S. Quinn, B. G. Anderson, L. Strait-Bodey, A. M. Stroud, [27]S.S.W.Chow,M.E.Craig,C.A.Jonesetal.,“Differences andJ.M.Argues,´ “Oversecretion of interleukin-15 from skele- in amniotic fluid and maternal serum cytokine levels in early tal muscle reduces adiposity,” American Journal of Physiology, midtrimester women without evidence of infection,” Cytokine, vol. 296, no. 1, pp. E191–E202, 2009. vol. 44, no. 1, pp. 78–84, 2008. [43] V. Almendro, N. Carbo,´ S. Busquets et al., “Interleukin-15 [28] Y. Tagaya, G. Kurys, T. A. Thies et al., “Generation of secretable decreases lipid intestinal absorption,” International Journal of and nonsecretable interleukin 15 isoforms through alternate Molecular Medicine, vol. 15, no. 6, pp. 963–967, 2005. usage of signal peptides,” Proceedings of the National Academy [44] N. Carbo,´ J. Lopez-Soriano,´ P. Costelli et al., “Interleukin- of Sciences of the United States of America, vol. 94, no. 26, pp. 15 antagonizes muscle protein waste in tumour-bearing rats,” 14444–14449, 1997. British Journal of Cancer, vol. 83, no. 4, pp. 526–531, 2000. [29] A. Onu, T. Pohl, H. Krause, and S. Bulfone-Paus, “Regulation [45] B. Alvarez, N. Carbo,´ J. Lopez-Soriano´ et al., “Effects of of IL-15 secretion via the leader peptide of two IL-15 interleukin-15 (IL-15) on adipose tissue mass in rodent isoforms,” Journal of Immunology, vol. 158, no. 1, pp. 255–262, obesity models: evidence for direct IL-15 action on adipose 1997. tissue,” Biochimica et Biophysica Acta, vol. 1570, no. 1, pp. 33– [30] R. Meazza, A. Gaggero, F. Neglia et al., “Expression of two 37, 2002. interleukin-15 mRNA isoforms in human tumors does not [46] K. M. Ajuwon and M. E. Spurlock, “Direct regulation correlate with secretion: role of different signal peptides,” of lipolysis by interleukin-15 in primary pig adipocytes,” European Journal of Immunology, vol. 27, no. 5, pp. 1049–1054, American Journal of Physiology, vol. 287, no. 3, pp. R608–R611, 1997. 2004. [31] R. N. Bamford, A. P. DeFilippis, N. Azimi, G. Kurys, and T. [47] V. Almendro, G. Fuster, E. Ametller et al., “Interleukin- A. Waldmann, “The 5’ untranslated region, signal peptide, 15 increases calcineurin expression in 3T3-L1 cells: possible and the coding sequence of the carboxyl terminus of IL-15 involvement on in vivo adipocyte differentiation,” Interna- participate in its multifaceted translational control,” Journal of tional Journal of Molecular Medicine, vol. 24, no. 4, pp. 453– Immunology, vol. 160, no. 9, pp. 4418–4426, 1998. 458, 2009. 8 Journal of Obesity

[48] T. Sugiura, M. Harigai, Y. Kawaguchi et al., “Increased IL-15 [63] E. E. Pistilli, P. M. Siu, and S. E. Alway, “Interleukin-15 production of muscle cells in polymyositis and dermatomyosi- responses to aging and unloading-induced skeletal muscle tis,” International Immunology, vol. 14, no. 8, pp. 917–924, atrophy,” American Journal of Physiology, vol. 292, no. 4, pp. 2002. C1298–C1304, 2007. [49] A. R. Nielsen, R. Mounier, P. Plomgaard et al., “Expression [64]J.S.Pattison,L.C.Folk,R.W.Madsen,andF.W. of interleukin-15 in human skeletal muscle—effect of exercise Booth, “Selected contribution: identification of differentially and muscle fibre type composition,” Journal of Physiology, vol. expressed genes between young and old rat soleus muscle dur- 584, no. 1, pp. 305–312, 2007. ing recovery from immobilization-induced atrophy,” Journal [50] P. L. Lollini, G. Palmieri, C. De Giovanni et al., “Expression of Applied Physiology, vol. 95, no. 5, pp. 2171–2179, 2003. of interleukin 15 (IL-15) in human rhabdomyosarcoma, [65]E.Marzetti,C.S.Carter,S.E.Wohlgemuthetal.,“Changes osteosarcoma and Ewing’s sarcoma,” International Journal of in IL-15 expression and death-receptor apoptotic signaling Cancer, vol. 71, no. 5, pp. 732–736, 1997. in rat gastrocnemius muscle with aging and life-long calorie [51] T. Stegall and K. A. Krolick, “Myocytes respond to both restriction,” Mechanisms of Ageing and Development, vol. 130, interleukin-4 and interferon-γ: cytokine responsiveness with no. 4, pp. 272–280, 2009. the potential to influence the severity and course of experi- [66] S. Gangemi, G. Basile, D. Monti et al., “Age-related modifi- mental myasthenia gravis,” Clinical Immunology, vol. 94, no. cations in circulating IL-15 levels in humans,” Mediators of 2, pp. 133–139, 2000. Inflammation, vol. 2005, no. 4, pp. 245–247, 2005. [52]D.C.Nieman,J.M.Davis,D.A.Hensonetal.,“Carbohydrate [67] J. Eisenman, M. Ahdieh, C. Beers et al., “Interleukin-15 ingestion influences skeletal muscle cytokine mRNA and interactions with interleukin-15 receptor complexes: charac- plasma cytokine levels after a 3-h run,” Journal of Applied terization and species specificity,” Cytokine,vol.20,no.3,pp. Physiology, vol. 94, no. 5, pp. 1917–1925, 2003. 121–129, 2002. [53] D. C. Nieman, J. M. Davis, V. A. Brown et al., “Influence [68] Y. He, X. Wu, R. S. Khan et al., “IL-15 receptor deletion results of carbohydrate ingestion on immune changes after 2 h of in circadian changes of locomotor and metabolic activity,” intensive resistance training,” Journal of Applied Physiology, Journal of Molecular Neuroscience, vol. 41, no. 2, pp. 315–321, vol. 96, no. 4, pp. 1292–1298, 2004. 2010. [54] A. R. Nielsen and B. K. Pedersen, “The biological roles of [69] D. M. Anderson, S. Kumaki, M. Ahdieh et al., “Functional exercise-induced cytokines: IL-6, IL-8, and IL-15,” Applied characterization of the human interleukin-15 receptor α chain Physiology, Nutrition and Metabolism, vol. 32, no. 5, pp. 833– and close linkage of IL15RA and IL2RA genes,” Journal of 839, 2007. Biological Chemistry, vol. 270, no. 50, pp. 29862–29869, 1995. [55] K. Ostrowski, C. Hermann, A. Bangash, P. Schjerling, J. [70] J. Hager, C. Dina, S. Francke et al., “A genome-wide scan for N. Nielsen, and B. K. Pedersen, “A trauma-like elevation human obesity genes reveals a major susceptibility locus on of plasma cytokines in humans in response to treadmill chromosome 10,” Nature Genetics, vol. 20, no. 3, pp. 304–308, running,” Journal of Physiology, vol. 513, no. 3, pp. 889–894, 1998. 1998. [71] R. S. Lindsay, S. Kobes, W. C. Knowler, P. H. Bennett, and R. L. [56] H. Andersson, S. K. Bøhn, T. Raastad, G. Paulsen, R. Blomhoff, Hanson, “Genome-wide linkage analysis assessing parent-of- and F. Kadi, “Differences in the inflammatory plasma cytokine origin effects in the inheritance of type 2 diabetes and BMI in response following two elite female soccer games separated by Pima Indians,” Diabetes, vol. 50, no. 12, pp. 2850–2857, 2001. a 72-h recovery,” Scandinavian Journal of Medicine and Science [72] V. Budagian, E. Bulanova, Z. Orinska et al., “Natural in Sports, vol. 20, no. 5, pp. 740–747, 2010. soluble interleukin-15Rα is generated by cleavage that [57] J. Prestes, G. Shiguemoto, J. P. Botero et al., “Effects of involves the tumor necrosis factor-α-converting enzyme resistance training on resistin, leptin, cytokines, and muscle (TACE/ADAM17),” Journal of Biological Chemistry, vol. 279, force in elderly post-menopausal women,” Journal of Sports no. 39, pp. 40368–40375, 2004. Sciences, vol. 27, no. 14, pp. 1607–1615, 2009. [73] A. Monroy, S. Kamath, A. O. Chavez et al., “Impaired [58] K. M. Beavers, F. C. Hsu, S. Isom et al., “Long-term phys- regulation of the TNF-α converting enzyme/tissue inhibitor ical activity and inflammatory biomarkers in older adults,” of metalloproteinase 3 proteolytic system in skeletal muscle Medicine & Science in Sports & Exercise, vol. 42, no. 12, pp. of obese type 2 diabetic patients: a new mechanism of insulin 2189–2196, 2010. resistance in humans,” Diabetologia, vol. 52, no. 10, pp. 2169– [59] J. Hansen, C. Brandt, A. R. Nielsen et al., “Exercise induces a 2181, 2009. marked increase in plasma follistatin: evidence that follistatin [74] M. Serino, R. Menghini, L. Fiorentino et al., “Mice het- is a contraction-induced hepatokine,” Endocrinology, vol. 152, erozygous for tumor necrosis factor-α converting enzyme no. 1, pp. 164–171, 2011. are protected from obesity-induced insulin resistance and [60] C. P. Lambert, M. G. Flynn, D. H. Sullivan, and W. J. Evans, diabetes,” Diabetes, vol. 56, no. 10, pp. 2541–2546, 2007. “Effects of megestrol acetate on circulating interleukin-15 and [75]T.Kubota,R.A.Brown,J.Fang,andJ.M.Krueger, interleukin-18 concentrations in healthy elderly men,” The “Interleukin-15 and interleukin-2 enhance non-REM sleep in Journals of Gerontology Series A, vol. 59, no. 8, pp. 855–858, rabbits,” American Journal of Physiology, vol. 281, no. 3, pp. 2004. R1004–R1012, 2001. [61] S. Raichur, P. Lau, B. Staels, and G. E. O. Muscat, “Retinoid- [76] X. Wu, W. Pan, Y. He, H. Hsuchou, and A. J. Kastin, “Cere- related orphan receptor γ regulatesseveralgenesthatcontrol bral interleukin-15 shows upregulation and beneficial effects metabolism in skeletal muscle cells: links to modulation of in experimental autoimmune encephalomyelitis,” Journal of reactive oxygen species production,” Journal of Molecular Neuroimmunology, vol. 223, no. 1-2, pp. 65–72, 2010. Endocrinology, vol. 39, no. 1-2, pp. 29–44, 2007. [62] C. M. Kotz, C. J. Billington, and A. S. Levine, “Obesity and aging,” Clinics in Geriatric Medicine, vol. 15, no. 2, pp. 391– 312, 1999. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 186368, 4 pages doi:10.1155/2011/186368

Clinical Study Rs9939609 Variant of the Fat Mass and Obesity-Associated Gene and Trunk Obesity in Adolescents

Harald Mangge,1 Wilfried Renner,1 Gunter Almer,1 Daniel Weghuber,2 Reinhard Moller,¨ 3 and Renate Horejsi3

1 Clinical Institute of Medical and Chemical Laboratory Diagnostics, Medical University of Graz, 8036 Graz, Austria 2 Department of Pediatrics, Paracelsus Private Medical University Salzburg, 5020 Salzburg, Austria 3 Institute of Physiological Chemistry, Center of Physiological Medicine, Medical University of Graz, 8036 Graz, Austria

Correspondence should be addressed to Harald Mangge, [email protected]

Received 10 November 2010; Revised 15 December 2010; Accepted 21 December 2010

Academic Editor: Francesco Saverio Papadia

Copyright © 2011 Harald Mangge et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

A common T/A polymorphism (rs9939609) in the fat mass and obesity associated (FTO) gene was found associated with early- onset and severe obesity in both adults and children. However, recent observations failed to find associations of FTO with obesity. Toinvestigate the genetic background of early obesity, we analysed the single nucleotide polymorphism (SNP) rs9939609 of FTO in 371 styrian adolescents towards degree of obesity, subcutaneous adipose tissue (SAT)-distribution determined by lipometry, early metabolic and preatherosclerotic symptoms. The percentage of AA homozygotes for the rs9939609 SNP of FTO was significantly increased in the obese adolescents. Compared to the TT wildtype, AA homozygotes showed significantly elevated values of SAT thickness at the trunk-located lipometer measure points neck and frontal chest, body weight, body mass index, waist, and hip circumference. No associations were found with carotis communis intima media thickness, systolic, diastolic blood pressure, ultrasensitive C-reactive protein (US-CRP), homocystein, total cholesterol, triglycerides, HDL cholesterol, oxidized LDL, fasted glucose, insulin, HOMA-index, liver transaminases, uric acid, and adipokines like resistin, leptin, and adiponectin. Taken together, to the best of our knowledge we are the first to report that the rs9939609 FTO SNP is associated with trunk weighted obesity as early as in adolescence.

1. Introduction and fat cell lipolysis [5, 6]. The single-nucleotide polymorphisms (SNPs) rs1421085, rs17817449, and rs9939609, of Juvenile obesity usually leads to obesity in adulthood which FTO were reported to be linked to body mass index (BMI) causes life threatening sequels such as diabetes, cardiovascu- and obesity in both adults and children [7–9]. However, lar disease, hypertension, stroke, and cancer [1]. In foregoing Li et al. failed to find associations of FTO with obesity studies, we detected an increased carotid intima-media [10]. To clarify this issue, we examined the so far best thickness (IMT) paralleled by a subclinical inflammation described FTO gene variant (i.e., SNP rs9939609) [6, 8]for in obese adolescents and we provided the first evidence correlations with grade of obesity, SAT distribution, and that preatherosclerosis is associated with individual risk obesity-related metabolic and cardiovascular risk parameters profiles characterised by subcutaneous adipose tissue (SAT) in styrian adolescents. topography and altered biomarkers [2–4]. In this paper, we analysed the single nucleotide polymorphism (SNP) rs9939609 of the fat mass and obesity-associated (FTO) gene 2. Material and Methods in obese styrian adolescents and in normal weight controls of the same age and gender distribution. FTO is a gene located 2.1. Subjects. Study participants (obese persons and normal in chromosome region 16q12.2. Recently, it was brought into weight, age- and sex-matched controls) were from the connection with the central control of energy homeostasis STYrian Juvenile OBesity Study (STYJOBS), which is 2 Journal of Obesity

Table 1: Baseline characteristics of study participants (age range (Applied Biosystems, Applera International Inc., Austria 5to20years,n = 371). GmbH, Mahlerstrasse 13, A-1010 Vienna, Austria), and assays were performed according to the manufacturer’s in- Variable Normal weight (mean + 1 SD) Obese structions. End-point fluorescence was measured and flu- Individuals 103 268 orescence plate reader data were exported into an Excel Female/male 51/52 146/122 format, depicted, and analyzed as a scatter plot. Age (years) 14.0 ± 3.112.5 ± 3.1 Liver transaminases, creatinine, glucose and uric acid Body length (m) 1.6 ± 0.12 1.6 ± 0.15 were measured by routine laboratory methods on a Hitachi Body weight (kg) 54.0 ± 13.676.2 ± 26.0 917 chemical analyser, cholesterol and triglycerides by means BMI (kg/m2)20.0 ± 2.930.1 ± 6.2 of ECLIA (ElectroChemiLuminiscenceAssay) on an Elecsys BMI-SDS 0.3 ± 1.06.0 ± 2.6 2010 analyser (Roche Diagnostics Mannheim, Germany), BMI: body mass index. and plasma insulin by ELISA (Mercodia, Uppsala, Swe- BMI-SDS: body mass index standard deviation score. den). HOMA-IR (homeostatic model assessment-insulin resistance) was calculated as reported [11]. Lipoproteins were separated by a combined ultracentrifugation-precipi- Table 2: Occurrence of the rs9939609 FTO gene polymorphism tation method (β-quantification) and analysed as outlined within the experimental groups. elsewhere [12]. Total adiponectin and subfractions were FTO Normal weight Obese determined by Adiponectin (Multimeric) Enzyme-Linked rs9939609 adolescents adolescents P value ImmunoSorbent Assay (47-ADPH-9755) from Alpco Diag- genotypes (n = 103) (n = 268) nostics, leptin and resistin by ELISAs from Biovendor Labo- TT 31 (30.1%) 75 (27.9%) ratory Medicine, Inc. (Brno, Czech Republic), oxidized low TA 56 (54.3%) 118 (44.0%) dense lipoprotein (oxLDL) by Mercodia oxidized LDL Com- .015 (AA versus petitive ELISA, SE-754 50 Uppsala, Sweden, ultra sensitive- AA 16 (15.5%) 75 (27.9%) TT&TA) CRP with a particle-enhanced immunoturbidimetric assay (Tina-quant C-reactive protein latex ultrasensitive assay, A allele 0.427 0.500 .075 frequency Roche diagnostics), homocysteine by triple quadrupole mass spectrometry (Applied biosystems, API 2000 LC/MS/MS- system) using a 3.3 × 0.46 cm HPLC column (SUPELCO LC-CN). designed to investigate early stages of atherosclerosis and metabolic disorders in obese juveniles. STYJOBS is regis- tered at Clinical-Trials.gov (Identifier NCT00482924), where 2.3. Carotid Artery Ultrasound. The ultrasound protocol detailed infor-mation of the study is available. The inclusion involved scanning of the bulbous near the common carotid criterium for the obese probands was BMI >97th percentile artery (CCA) on both sides with a 12-to-5-MHz broad- if under 18 years of age, BMI >30 kg/m2 if over 18 years band linear transducer on an HDI 5000 (ATL, Bothell, of age. Exclusion criteria were endocrine diseases (e.g., Washington, DC, USA). The carotid IMT was assessed at hypothyreosis), infectious or any other chronic diseases. the far wall as the distance between the interface of the Further, STYJOBS participants aged above 20 years were lumen and intima and the interface between the media and excluded in this study. Controls were healthy age-matched adventitia. All diameters were measured during diastole to volunteers. All controls had to be normal weight (BMI avoid image blurring due to systolic arterial wall motion and around 50th percentile if under 18 years of age, BMI to minimize the influence of blood pressure [13]. <25 kg/m2 ifover18yearsofage)freeofinfectious,chronic, and endocrine diseases. 268 obese juveniles recruited from 2.4. Lipometry. Measurements of SAT thickness were per- July 2003 to December 2006 (mean age 12.5 ± 3.1 (SD) formed by means of a patented optical device (EU Pat.Nr. years) and 103 normal weight healthy controls of similar 0516251) on 15 anatomically well-defined body sites dis- age and gender distribution were investigated. The study was tributed from neck to calf on the right and left side of all approved by the ethical committee of the Medical University obese juveniles [14] and then averaged for both body sides. of Graz. At the time of blood collection, the probands were fasting. Blood samples were immediately centrifuged at 2.5. Statistics. Statistical analysis was performed by SPSS ◦ 3500 rpm at ambient temperature and stored at −80 Cuntil version 14. Kolmogorov-Smirnov test was used to examine analysis. for normal distribution. Means were compared by a two- tailed unpaired sample t-test or by Mann-Whitney Test, 2.2. Laboratory Analysis. Genomic DNA was isolated from depending on the distribution of the data. A value of P<.05 peripheral lymphocytes by standard methods and stored at was considered statistically significant. −20◦C. FTO genotypes were determined by 5-exonuclease assay (TaqMan, Applied Biosystems, Applera International 3. Results Inc., Austria GmbH, Mahlerstrasse 13, A-1010 Vienna, Aus- tria). Primer and probe sets were designed and manufactured The clinical characteristics of the study participants are using Applied Biosystems “Assay-by-Design” custom service summarized in Table 1. FTO genotypes did not deviate from Journal of Obesity 3

Table 3: Associations of clinical parameters with the rs9939609 FTO gene polymorphism.

Mean + 1 SD Wildtype Heterozygote Homozygote Body weight (kg) 65.9 ± 23 70.4 ± 26.474.2 ± 25∗∗ BMI (kg/m2)26.5 ± 727.7 ± 7.228.8 ± 7∗∗ BMI-SDS 4.2 ± 3.54.2 ± 3.45.1 ± 3.5∗ Waist circumference (cm) 85.4 ± 16 88.1 ± 16.890.4 ± 13∗ Hip circumference (cm) 99.7 + 16 102.9 ± 16.9 105.1 ± 16∗ SAT-T, neck (cm) 12.7 ± 6.513.9 ± 715.3 ± 5.8∗∗ SAT-T, front chest (cm) 16.7 ± 8.918.4 ± 8.820.4 ± 9.3∗∗ ∗P<.05, ∗∗P<.01, ∗∗∗P<.001, MannWhitney U-test (wildtype versus homozygote). BMI: Body mass index. BMI-SDS: body mass index standard deviation score.

Table 4: Clinical parameters and the rs9939609 FTO gene polymorphism within all experimental groups.

Obese adolescents Normal weight adolescents Wildtype Heterozygot Homozygote Wildtype Heterozygote Homozygote Mean + 1SD n = 75 n = 118 n = 75 n = 31 n = 56 n = 16 Body weight (kg) 72.1 ± 23.377.8 ± 27.777.9 ± 25.751.1 ± 12.754.8 ± 13.956.8 ± 13.5 BMI (kg/m2)29.3 ± 6.130.3 ± 6.330.6 ± 6.119.5 ± 2.820.2 ± 2.820.3 ± 3.3 BMI-SDS 5.9 ± 2.66.0 ± 2.66.2 ± 2.90.17 ± 0.90.4 ± 1.00.2 ± 1.0 Waist circumference (cm) 89.9 ± 14.892.2 ± 15.892.7 ± 11.867.9 ± 5.870.8 ± 6.371.6 ± 7.4 Hip circumference (cm) 102.8 ± 15.8 106.4 ± 16.5 106.7 ± 16.487.6 ± 8.988.5 ± 8.992.6 ± 11.1 SAT-T, neck (cm) 14.3. ± 5.715.4 ± 6.415.6 ± 5.15.3 ± 4.55.9 ± 4.55.6 ± 3.6 SAT-T, front chest (cm) 19.9. ± 7.420.8 ± 8.122.3 ± 9.45.4 ± 4.57.5 ± 5.48.2 ± 7.1 BMI: body mass index. BMI-SDS: body mass index standard deviation score. the Hardy-Weinberg equilibrium. Compared to the normal distribution. The percentage of AA homozygotes for the weight controls, the percentage of carriers of the FTO AA rs9939609 SNP of FTO was significantly increased in the genotype was significantly increased in the obese ado- obese adolescents indicating a relevance of this SNP in early lescents (Table 2). Heterozygous and wildtype genotypes phases of obesity. The observation that the SAT thicknesses of were approximately equally distributed within the groups the trunk-located lipometer measure points, neck and frontal (Table 2). chest, were significantly increased in homozygote rs9939609 Homozygote carriers showed significantly elevated values carriers indicates for the first time a link between trunk- for body weight, body mass index (BMI), BMI standard weighted SAT distribution and a specific genetic disposition. deviation scores (BMI-SDS), and waist, hip circumferences, On the other hand, no associations could be found and the thickness of the SAT was significantly increased at between the investigated SNP and cardiovascular risk param- the trunk-located lipometer measure points, neck and frontal eters like carotis communis IMT, systolic, diastolic blood chest (Table 3). Analysed separately in the obese and normal pressure, conventional laboratory-, metabolic-, inflamma- weight group, these variables showed a trend to higher levels tory biomarkers (e.g., Liver enzymes, fasted glucose, HOMA- in heterozygote and homozygote carriers; however no signifi- index, homocystein, lipids, oxidized LDL, and US-CRP), and cant difference was achieved (Table 4). Further, no significant adipokines such as adiponectin, resistin, and leptin. This elevations were found with carotis IMT, systolic, diastolic may be caused by the fact that these cardiovascular and blood pressure, US-CRP, homocystein, total cholesterol, metabolic risk parameters reflect a more common patho- triglycerides, HDL cholesterol, oxidized LDL, fasted glucose, logic phenotype. Genetic risk constellations, possibly more insulin, HOMA-index, liver transaminases, uric acid, and important for future clinical endpoints, may be present in the adipokines (Resistin, leptin, adiponectin, and subfractions) background without correlation to these markers but related either in the homozygote nor in the heterozygote carriers to trunk-weighted obesity. It will be interesting to follow up (not shown). this cohort of adolescents for development of overt metabolic and atherosclerotic disease symptoms later in life and to 4. Discussion study correlations of clinical end points with homozygote rs9939609 carriers. We analysed the single nucleotide polymorphism (SNP) Several studies reported a strong link between the rs9939609 of the FTO gene in obese Caucasian adolescents rs1421085, rs17817449, and rs9939609 SNPs of FTO and and in normal weight controls of the same age and gender body mass index (BMI), even in children [7–9]. On the other 4 Journal of Obesity hand, a recent study failed to find associations of FTO with [13] S. Kiechl and J. Willeit, “The natural course of atherosclerosis. obesity [10]. This may be influenced be the fact that these Part II: vascular remodeling,” Arteriosclerosis, Thrombosis, and authors investigated Chinese and not Csaucasion probands Vascular Biology, vol. 19, no. 6, pp. 1491–1498, 1999. [10]. Our observations in juveniles are in accordance with [14] R. Moller,E.Tafeit,T.R.Pieber,K.Sudi,andG.Reibnegger,¨ the observations of adults by Frayling et al. [8]. “Measurement of subcutaneous adipose tissue topography Taken together, to the best of our knowledge, we are the (SAT-Top) by means of a new optical device, LIPOMETER, ffi first to report that homozygosity for the rs9939609 FTO SNP and the evaluation of standard factor coe cients in healthy is critically associated with trunk-weighted obesity in obese subjects,” American Journal of Human Biology,vol.12,no.2, pp. 231–239, 2000. adolescents.

References

[1] D. W. Haslam and W. P. T. James, “Obesity,” Lancet, vol. 366, no. 9492, pp. 1197–1209, 2005. [2] H. Mangge, K. Schauenstein, L. Stroedter, A. Griesl, W. Maerz, and M. Borkenstein, “Low grade inflammation in juvenile obesity and type 1 diabetes associated with early signs of atherosclerosis,” Experimental and Clinical Endocrinology and Diabetes, vol. 112, no. 7, pp. 378–382, 2004. [3] S. Pilz, R. Horejsi, R. Moller¨ et al., “Early atherosclerosis in obese juveniles is associated with low serum levels of adiponectin,” Journal of Clinical Endocrinology and Metabolism, vol. 90, no. 8, pp. 4792–4796, 2005. [4]H.Mangge,G.Almer,M.Truschnig-Wilders,A.Schmidt, R. Gasser, and D. Fuchs, “Inflammation, adiponectin, obesity and cardiovascular risk,” Current Medicinal Chemistry,vol.17, no. 36, pp. 4511–4520, 2010. [5] R. Fredriksson, M. Hagglund,¨ P. K. Olszewski et al., “The obesity gene, FTO, is of ancient origin, up-regulated during food deprivation and expressed in neurons of feeding-related nuclei of the brain,” Endocrinology, vol. 149, no. 5, pp. 2062– 2071, 2008. [6] K. Wahlen, E. Sjolin,¨ and J. Hoffstedt, “The common rs9939609 gene variant of the fat mass- and obesity-associated gene FTO is related to fat cell lipolysis,” Journal of Lipid Research, vol. 49, no. 3, pp. 607–611, 2008. [7] C. Dina, D. Meyre, S. Gallina et al., “Variation in FTO contributes to childhood obesity and severe adult obesity,” Nature Genetics, vol. 39, no. 6, pp. 724–726, 2007. [8] T. M. Frayling, N. J. Timpson, M. N. Weedon et al., “A common variant in the FTO gene is associated with body mass index and predisposes to childhood and adult obesity,” Science, vol. 316, no. 5826, pp. 889–894, 2007. [9]R.A.Price,W.D.Li,andH.Zhao,“FTOgeneSNPsassociated with extreme obesity in cases, controls and extremely discordant sister pairs,” BMC Medical Genetics,vol.9,ArticleID4, 2008. [10] H. Li, Y. Wu, R. J. F. Loos et al., “Variants in the fat mass- and obesity-associated (FTO) gene are not associated with obesity in a Chinese Han population,” Diabetes, vol. 57, no. 1, pp. 264– 268, 2008. [11] D. R. Matthews, J. P.Hosker, and A. S. Rudenski, “Homeostasis model assessment: Insulin resistance and β-cell function from fasting plasma glucose and insulin concentrations in man,” Diabetologia, vol. 28, no. 7, pp. 412–419, 1985. [12] B. R. Winkelmann, W. Marz,¨ B. O. Boehm et al., “Rationale and design of the LURIC study—a resource for functional genomics, pharmacogenomics and long-term prognosis of cardiovascular disease,” Pharmacogenomics,vol.2,no.1,pp. S1–S73, 2001. Hindawi Publishing Corporation Journal of Obesity Volume 2011, Article ID 329038, 8 pages doi:10.1155/2011/329038

Research Article Gene by Sex Interaction for Measures of Obesity in the Framingham Heart Study

Ashlee M. Benjamin,1 Sunil Suchindran,1 Kaela Pearce,1 Jennifer Rowell,2 Lillian F. Lien,2 John R. Guyton,2 and Jeanette J. McCarthy1, 3

1 Duke Institute for Genome Sciences and Policy, Duke University, Durham, NC 27708, USA 2 Division of Endocrinology, Metabolism and Nutrition, Department of Medicine, Duke University Medical Center, Durham, NC 27710, USA 3 Department of Community and Family Medicine, Duke University Medical Center, Durham, NC 27710, USA

Correspondence should be addressed to Jeanette J. McCarthy, [email protected]

Received 14 September 2010; Revised 17 November 2010; Accepted 22 November 2010

Academic Editor: Yvon Chagnon

Copyright © 2011 Ashlee M. Benjamin et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Obesity is an increasingly prevalent and severe health concern with a substantial heritable component and marked sex differences. We sought to determine if the effect of genetic variants also differed by sex by performing a genome-wide association study modeling the effect of genotype-by-sex interaction on obesity phenotypes. Genotype data from individuals in the Framingham Heart Study Offspring cohort were analyzed across five exams. Although no variants showed genome-wide significant gene-by-sex interaction in any individual exam, four polymorphisms displayed a consistent BMI association (P-values .00186 to .00010) across all five exams. These variants were clustered downstream of LYPLAL1, which encodes a lipase/esterase expressed in adipose tissue, a locus previously identified as having sex-specific effects on central obesity. Primary effects in males were in the opposite direction from females and were replicated in Framingham Generation 3. Our data support a sex-influenced association between genetic variation at the LYPLAL1 locus and obesity-related traits.

1. Introduction Obesity is a heritable trait and recent genome-wide association studies have identified dozens of loci influencing Overweight and obesity present a major public health measures of adiposity [5–8]. Sex differences in the heritabil- challenge in the developed world and are a primary focus of ity of obesity-related traits have been noted as well in several preventive healthcare. Rates of both overall adiposity, mea- studies [9]. In addition, linkage analysis in both rodent sured by body mass index (BMI), as well as central (intra- models and humans have found evidence of sex-specific abdominal) adiposity, measured by waist circumference loci affecting obesity-related traits [10, 11]. Framingham (WC) or waist to hip ratio (WHR) have been steadily rising Heart Study investigators found widespread evidence for sex- during the past several decades, accompanied by increased specific effects of genetic loci on body mass index, identifying rates of diabetes mellitus, cardiovascular disease, and other several chromosomal regions with suggestive linkage to BMI morbidities [1]. In the United States, regional, racial, and in one sex, but not the other [11]. Indeed some effects were sex differences in adiposity have been noted, but the patterns only seen in sex-stratified analyses and were not at all evident are complex and changing over time [1]. According to U.S. inthecombinedcohortofmenandwomen.Morerecently, national health survey data, men on average have had a two genome-wide association study meta-analyses of WHR higher BMI than women, but since the mid 1990s the average examined their top loci for sex differences and identified sex- BMI in women has been higher than men [2]. Men also specific effects for several loci [8, 12]. tend to have larger abdominal girth than women, and this We sought evidence for significant differences in SNP disparity has persisted over time [3, 4]. effects on adiposity traits in men and women across the 2 Journal of Obesity genome by carrying out a genome-wide association study 2.2. Genotype Data and Quality Control. Genome-wide modeling gene by sex interaction for WHR, WC, and BMI genotypes and detailed clinical data have been made acces- in the population-based Framingham Heart Study. Genome- sible to the research community through the SHARe project wide association analysis of SNPs having main effects (as (SNP-Health Association Resource). The study protocol was opposed to gene by sex interaction) on obesity were reported approved by Duke University’s Institutional Review Board earlier in the Framingham Heart Study using 100 K SNPs, and the Framingham SHARe Data Access Committee. The but gene by sex interactions were not considered at that unfiltered genotype data contained 9215 individuals (all time [13]. Subsequently, the full genotype data (>500 K generations) genotyped for 549782 SNPs. This included SNPs) have been pooled with other studies and reported in 500568 SNPs from the Affymetrix 500 K mapping array and large meta-analyses, which found evidence of gene by sex 49214 SNPs from the Affymetrix 50 K supplemental array interaction for WHR but not BMI among the SNPs with (Affymetrix, Santa Clara, CA, USA). We used the toolset main effects [5, 8]. PLINK [16] to perform quality control. Individuals were excluded if genotyping rates were less than 97%. Markers 2. Materials and Methods were excluded if genotyping rates were less than 97%, minor allele frequencies were less than 0.05, or if Hardy- 2.1. Study Population. We conducted this research using data Weinberg P-values were less than .001. All SNP exclusions from the Framingham Heart Study, a population-based, lon- were made sequentially in the preceding order. Using this gitudinal study of families living in the town of Framingham, filtered data, we checked for Mendel errors using a 5% cutoff Massachusetts collected over three-generations beginning in per family, and a 10% cutoff per SNP (as defined in PLINK), 1948. An overview of the study is provided at the dbGap but none were detected. Individuals were also excluded if website (http://www.ncbi.nlm.nih.gov/sites/entrez?db=gap) the predicted sex based on X-chromosome genotypes did and detailed descriptions are available elsewhere [14, 15]. not match the recorded sex. Pairwise identity-by-descent Briefly, the original study (Generation 1) enrolled 5209 measures were calculated to detect replicated samples and individuals, primarily Caucasian, and it later added the unknown interfamilial relationships. We detected 4 identical offspring of the original cohort (Generation 2), and the twins and randomly selected one member of each pair for grandchildren (Generation 3) of the original cohort. Primary the analytic sample. After quality controls, the remaining analyses were carried out using data from the five first exams sample consisted of genotype data on 360811 SNPs, attaining of subjects in Generation 2, collected between 1971 and a genotyping rate of 99.5%. 1994. Obesity-related traits evaluated in this study included BMI, measured at exams 1, 2, 3, 4, and 5, WHR, measured 2.3. Statistical Analyses. Analysis of WHR and WC were at exams 4 and 5, and WC measured at exams 4 and 5. based on data obtained at exam 4 (n = 1330) and exam We limited our analyses to these exams due to a drop in 5(n = 984) of subjects from Generation 2. The gene by sample size at subsequent exams. Replication of genome sex GWAS was run on data from each exam separately. We wide association study (GWAS) results was sought in subjects ran the full model for both WHR and WC regressed on from Generation 3 (data collected in 2002–2005). BMI, age, age-squared, genotype, sex, and the genotype- Individuals with diabetes (n = 92, 94, 59, 27, 116, and 136 by-sex cross product. BMI was available at all exams, with for generation 2 exams 1, 2, 3, 4, 5 and generation 3 exam adequate sample sizes on the first five exams. Five separate 1, resp.) or thyroid disorder (n = 117, 94, 9, 36, 265, and 72 GWAS were run using the full model of BMI regressed for generation 2 exams 1, 2, 3, 4, 5 and generation 3 exam on age, age-squared, genotype, sex, and the genotype-by- 1, resp.) were removed because these diseases have an effect sex cross product—one each for exams 1, 2, 3, 4, and on both BMI and fat distribution. The data were further 5 of Generation 2. SNPs were evaluated for associations trimmed, excluding individuals with outlier trait values in an additive genetic model. A main effect GWAS was determined by taking the mean of the phenotype (indepen- also run for BMI across the five exams, using the model dently for each exam and each sex) and adding/subtracting specifications above without the cross product term. Sex- three standard deviations. Removal of outlier values in specific associations were tested using the full model of BMI the BMI GWAS data was performed with weight, height, regressed on age, age-squared, and genotype on each sex. and BMI. WC and hip circumference (HC) outliers were To account for relatedness, we used generalized estimating also eliminated in the waist phenotype GWAS. Finally, equations while accounting for sibling correlation in the we restricted our analysis to premenopausal women and Yags package [17]oftheR statistical language. The P- individuals under the age of 50 to enhance differences related values of the covariates were obtained via the Wald test to estrogen-mediated gene by sex interaction and to reduce using robust standard errors. The Framingham population as much as possible the age-related differences in association has been studied extensively, and evidence for considerable that may occur across exams. The total sample sizes for the population stratification has not been detected. To test this BMI GWAS after genotype quality control and trait outlier assumption, we estimated the inflation factor by dividing removal were 3150, 1991, 1630, 1330, 990, and 2872 for the median of the observed χ2 statistics for each GWAS, by generation 2 exams 1, 2, 3, 4, 5, and generation 3 exam 1, the expected median in the absence of stratification (0.456) respectively. The sample sizes for the waist phenotype GWAS [18, 19]. Also, adjusted for population stratification with the were 1330, 984, and 2872 for generation 2 exams 4, 5 and scores of the first 10 principal components, computed with generation 3 exam 1, respectively. Eigenstrat [20]. We defined genome-wide significance using Journal of Obesity 3 a Bonferroni cutoff of 1.4 × 10−7, which corrects for 360811 BMIGenebysexinteraction, tests. generation 2 under 50 6 Following genome-wide analysis, we annotated results 5 using the WGAViewer package [21], Ensembl [22], and the 4 UCSC genome browser. We generated plots using the Gap 3 package [23] of the R statistical language and Haploview (observed) 10 2

software [24]. log 1 To enrich for true positive associations, we took a − 0 strategy whereby associations that appeared in all exams 0 123456 − were considered to have a higher likelihood of being true log10 (expected) associations. We expected earlier exams to have greater power (a) due to larger sample sizes, but other factors, including ff decreased heritability with age [11]mayaffect the results as BMI main e ect, generation 2 under 50 well.Thisstrategyrequiredustomakesomedecisionsabout 6 what cutoff to use when comparing results across exams. We 5 took the consensus across exams of the top most significant 4 3

10, 100, 1000, and 10000 hits and found 0, 0, 4, and 105 SNPs, (observed) respectively, and focused on the four SNPs from the top 1000 10 2 consensus further. log 1 − 0 0 123456 − 3. Results log10 (expected)

Characteristics of the subjects from Generations 2 and 3 Exam 1 Exam 4 of the Framingham Study used in the current analyses are Exam 2 Exam 5 presented in Table 1, broken down by exam. For each exam, Exam 3 Reference we restricted our analyses to men and women <50 years of (b) age, resulting in a decrease in sample size over time, above ff and beyond the loss due to death or nonparticipation. Figure 1: QQ plots for gene by sex interaction (a) and main e ect (b) GWAS for body mass index (BMI) in Generation 2, exams 1, 2, 3, 4, and 5. 3.1. Genome-Wide Association Analysis of Gene by Sex Inter- action for WHR and WC and BMI. None of the gene-by- sex interaction GWAS revealed genome-wide significant loci. For BMI we noted marked heterogeneity in quantile-quantile by these findings as the LYPLAL1 locus has been reported as ff (QQ) plots between exams (Figure 1), which does not appear a sex-specific locus a ecting central adiposity in two prior to be a function of sample size (which decreases with exam). genome-wide association meta-analyses [8, 12]. The extent There is also some evidence of inflation in the QQ plots, of linkage disequilibrium (LD) surrounding the associated which was not alleviated after controlling for population SNPs in the region of LYPLAL1 was determined in the Hap stratification. In sex-stratified analysis, the inflation appeared Map phase 3 CEU population by identifying the farthest r2 > . to be restricted to men. The top 1000 hits from each exam SNP away in each direction that had 0 5foreach for each trait (ordered by the P-value of the gene by sex of the four SNPs. The LD block extends over 330 kb from interaction term) were extracted (Supplementary Tables S1, position 217,321,833 to 217,655,426, and encompasses the S2, and S3 available online on doi:10.1155/2011/329038), LYPLAL1 gene (Figure 2). The block does not include the and the intersection of those datasets was sought for each SNPs from Lindgren et al. [12] or Heid et al. [8], which are in trait. moderate linkage disequilibrium with each other and located For WHR, we identified 43 SNPs (28 unique loci) and for an additional 55 kb and 258 kb downstream of LYPLAL1, WC, we identified 43 SNPs (27 unique loci) appearing among respectively. the top 1000 in both exams 4 and 5 (Tables S2 and S3). When examining loci across these two traits, SNPs near SPOCK3, 3.2. Replication of LYPLAL1 SNP Association with BMI in OSTF1, RAB31,andRPF1 appear in the top 1000 consensus Framingham Generation 3 Subjects. We next sought to repli- for WC and WHR. SPOCK3 stands out as appearing among cate the observed association in subjects from Generation 3 the top 100 hits across both exams 4 and 5 for WC (P = of the Framingham Study. Again, we restricted our analyses 5.33×10−7 and P = 2.45×10−5) and WHR (P = 1.85×10−4 to those less than 50 years of age. A comparison of results by and P = 7.95 × 10−5). sex in the five exams of Generation 2 and in Generation 3 For BMI, only four SNPs appeared among the top are shown in Figure 3 for the top associated LYPLAL1 SNP. 1000 hits in all five exams. All four SNPs localized to the same The SNP by sex interaction for LYPLAL1 was significant in linkage disequilibrium block on chromosome 1, ∼100 kb all Generation 2 exams, but not significant in Generation downstream of LYPLAL1. Supplementary Table S3.6 shows 3 subjects. However, when stratified by sex, the minor the location, minor allele frequency, P values, and rank of allele showed a consistent increase in BMI in men across each SNP by sex interaction by exam. We were most intrigued generations (Figure 3). In contrast, in women the minor 4 Journal of Obesity

Table 1: Mean ± standard deviation for obesity-related traits in Framingham subjects <50 years old.

Generation 2, Generation 2, Generation 2, Generation 2, Generation 2, Generation 3, Population exam 1 exam 2 exam 3 exam 4 exam 5 exam 1 All 3150 1991 1630 1330 990 2872 Sample size (N) Men 1478 958 776 640 463 1388 Women 1672 1033 854 690 527 1484 All 33.71 ± 8.59 38.39 ± 6.70 40.46 ± 5.77 42.27 ± 5.11 43.68 ± 4.44 37.63 ± 7.27 Age . ± . . ± . . ± . . ± . . ± . . ± . (years) Men 33 81 8 63 38 38 6 88 40 47 5 92 42 14 5 23 43 56 4 65 37 79 7 30 Women 33.62 ± 8.56 38.39 ± 6.54 40.45 ± 5.63 42.40 ± 5.00 43.78 ± 4.24 37.47 ± 7.24 All 24.85 ± 3.77∗∗∗ 25.10 ± 3.79∗∗∗∗ 25.40 ± 4.08∗∗∗ 25.97 ± 4.37∗ 26.42 ± 4.48 26.23 ± 4.70 BMIa . ± . ∗ . ± . ∗∗ . ± . ∗∗ . ± . . ± . . ± . (kg/m2) Men 26 42 3 47 26 57 3 44 26 72 3 49 27 23 3 62 27 55 3 92 27 42 4 03 Women 23.48 ± 3.47∗∗∗ 23.74 ± 3.59∗∗∗ 24.20 ± 4.20∗∗ 24.80 ± 4.66∗ 25.42 ± 4.70 25.11 ± 4.99 All 167.62 ± 9.36∗∗∗∗ 168.59 ± 9.57∗∗∗ 169.66 ± 9.20∗ 169.67 ± 9.12∗ 169.56 ± 9.06 170.91 ± 9.18 Height . ± . ∗∗∗∗ . ± . ∗∗∗ . ± . ∗ . ± . ∗ . ± . . ± . (cm) Men 174 95 6 79 175 99 6 77 176 75 6 66 176 65 6 52 176 69 6 41 177 88 6 41 Women 161.15 ± 5.90∗∗∗∗∗ 161.73 ± 6.02∗∗∗∗ 163.23 ± 5.86∗ 163.20 ± 5.81 163.29 ± 5.85 164.39 ± 6.06 All 69.77 ± 14.52∗∗∗∗ 71.25 ± 14.62∗∗∗∗ 72.89 ± 14.97∗∗∗ 74.50 ± 15.55∗∗ 75.72 ± 15.79 76.32 ± 16.55 Weight . ± . ∗∗∗∗ . ± . ∗∗∗ . ± . ∗∗∗ . ± . ∗∗ . ± . . ± . (kg) Men 80 26 11 76 81 63 11 34 82 78 11 82 84 27 12 24 85 34 13 02 86 06 13 76 Women 60.50 ± 9.58∗∗∗∗ 61.63 ± 10.01∗∗∗∗ 63.90 ± 11.45∗∗∗ 65.45 ± 12.51∗ 67.27 ± 12.91 67.20 ± 13.47 All — — — 86.70 ± 13.92∗∗∗∗ 88.95 ± 13.30∗∗∗ 91.03 ± 13.33 b WC (cm) Men — — — 95.85 ± 9.93∗∗ 96.49 ± 10.34 96.50 ± 11.15 Women — — — 78.21 ± 11.50∗∗∗∗∗ 82.36 ± 12.04∗∗∗ 85.90 ± 13.18 All — — — 100.07 ± 9.08 101.37 ± 8.63 — c HC (cm) Men — — — 101.31 ± 7.18 101.83 ± 6.85 — Women — — — 98.92 ± 10.42 100.98 ± 9.93 — All — — — 0.86 ± 0.10 0.88 ± 0.09 — d WHR Men — — — 0.95 ± 0.06 0.95 ± 0.06 — Women — — — 0.79 ± 0.06 0.81 ± 0.07 — a Body mass index, bWaist circumference, cHip circumference, dWaist to hip ratio, ∗Significant difference between Generation 2 and Generation 3 after controlling for age and age-squared (∗P<.001, ∗∗P<1e − 5, ∗∗∗P<1e − 10, ∗∗∗∗P< 1e − 20, ∗∗∗∗∗P<1e − 50). Significance in age differences is not noted. allele was associated with lower BMI in Generation 2 but not Map CEU r2 = 1) and found a borderline significant gene- in Generation 3. by-sex interaction with WHR (P = .09; Supplement S4).

3.3. Association of LYPLAL1 SNPs with Obesity-Related Traits. 3.4. Genome-Wide Association Analysis of Gene Main Effects To understand the relationship between LYPLAL1 SNPs and for BMI. We also explored our cross-exam consensus obesity in greater detail, we examined the top SNP from the approach for detecting significant main effects for BMI, using present study (rs7552206) along with SNPs from the Lind- the same age-restricted datasets as the gene by sex interaction gren et al. [12] and Heid et al. [8] studies for association with analyses. As with our gene by sex interaction analyses, related phenotypes, including height, weight, WC, and WHR the QQ plots show marked heterogeneity between exams (Supplemental Table S4). The rs7552206 by sex interaction (Figure 1) and modest inflation, which was not accounted for forBMItrackedwithweightinallfiveexams,andwith by population stratification. Only one SNP, located ∼26 kb WC and HC in the two exams that had these data available. upstream of DUSP10 on chromosome 1, appeared among However, the waist and hip associations were completely or the top 1000 hits (Supplement S5) in all five exams of nearly completely attenuated when controlling for BMI. For Generation 2 and was borderline significant in Generation 3 rs2605100 (Lindgren et al. [12]), no compelling evidence subjects (Figure 4). Interestingly, this locus is approximately of gene by sex interaction in central adiposity was found. 2.4 Mb away from the gene by sex interaction LYPLAL1 SNPs. Heid et al. [8] independently found a female-biased WHR No SNPs from prior genome-wide association studies of association with LYPLAL1 (rs4846567), an SNP in moderate BMI showed up among our top 1000 consensus, including linkage disequilibrium with the Lindgren et al. SNP. We SNPs in the genes INSIG2, FTO [13, 25], and MC4R analyzed an available proxy for this SNP (rs2820446, Hap [26](Figure 4). Surprisingly, the SNPs identified with the Journal of Obesity 5

LYPLAL1 SNPs BMI GWAS SNPs Lindgren SNP Heid SNP

chr1

Genotyped SNPs 217500 k 217600 k 217700 k 217800 k Entrez genes NM 138794 LYPLAL1: lysophospholipase-like 1

LYPLAL1 SNPs BMI GWAS SNPs Heid SNP Lindgren SNP rs1524633 rs17049143 rs1455703 rs1847661 rs7533619 rs7536671 rs7536586 rs6699816 rs2049423 rs11118234 rs10863436 rs12402039 rs10746384 rs1474374 rs10157397 rs12122774 rs2605100 rs4846567

Block 1 (24 kb) Block 2 (7 kb) Block 3 (15 kb) 47 48 49 51 52 53 54 55 56 57 58 59 60 109 114 120 211 289

52 56 56 25 6 10 36 9 12 12 39 29 89 89 5 48

93 92 44 7 25 10 36 5 89 21 11 27 4 4

52 25 14 92 25 2 5 41 9 39 6 14 29 5 3

47 10 56 92 10 34 41 5 21 11 3 11 0 4

13 56 41 18 4 5 52 3 14 6 6 1

52 23 7 10 34 52 18 5 11 16 3

52 36 41 36 18 13 18 14 3 6 12

21 5 23 7 47 1 14 18 14 3

39 41 41 52 52 3 18 4 11

21 5 93 15 51 1 4 4

39 52 26 17 52 2 4

18 25 15 0 1

29 14 26 3 0

27 18 0 3

29 4 1

0 4

Figure 2: Linkage disequilibrium (shown as r2) in the region encompassing LYPLAL1, the consensus SNPs associated with body mass index (BMI) in our gene by sex interaction GWAS, and the sex-speciﬁc SNPs associated with waist to hip ratio (WHR) in recent GWAS meta- analyses.

consensus approach yielded more significant P values than Mean BMI by genotype -rs7552206 30 other loci. 0 0.05 0.06 0.04 27.5 0.01 0.13 0.02 4. Discussion 0.03 0.95 25 0 We carried out a genome-wide assessment of gene by sex 0.05 0.01 interaction for standard measures of obesity in men and 22.5 women less than 50 years of age in the Framingham Heart Study. We took advantage of longitudinal data from multiple 20 exams to identify loci showing consistent evidence of SNP Men Men Men Men by sex interaction across exams. Among the most prominent Men Men Women ∼ Women Women Women Women Women was a region 100 kb downstream of LYPLAL1, encoding the G2E1 G2E2 G2E3 G2E4 G2E5 G3E1 lysophospholipase-like 1 protein. We found evidence across five exams, spanning a 20-year time frame, of opposite effects CC CA of genetic variants in this region on BMI in men and women. AA An attempt to replicate this finding in a later generation of Framingham Heart Study subjects found a consistent, Figure 3: Mean body mass index (BMI) by genotype and sex across significant association in men, but not in women, possibly exams for the top associated SNP in LYPLAL1 (rs7552206) with indicating a male-specific association. Standard Error Bars and SNP P values. 6 Journal of Obesity

Ours is not the first study to link LYPLAL1 to obesity: Gene by sex interaction/main effect— significance across exams two other genome-wide association meta-analyses identified 6 this locus as having a sex-specific effect on WHR [8, 12]. 5 While neither SNP is in linkage disequilibrium with the 4 3 region identified in our study, the coincidental discovery 2 of two distinct regions near the LYPLAL1 locus associated 1 0 with obesity-related traits in a sex-specific fashion warrants further attention. Moreover, a prior linkage analysis of BMI in Generation 2 of the Framingham Heart Study identified rs1121980 rs2820446 rs2820446 rs7566605 rs17782313 rs6665466 rs2605100 rs7552206 FTO (BMI ME) rs2605100 a male-biased linkage for BMI in the vicinity of LYPLAL1 Heid (BMI GxS) MC4R (BMI ME) Heid (WHR GxS) INSIG2 (BMI ME) DUSP10 (BMI ME) Lindgren (BMI GxS) on chromosome 1q41 [11]. None of the other sex-specific LYPLAL1 (BMI GxS) Lindgren (WHR GxS) obsesity loci from Heid et al. [8] were found in our G2E1 G2E5 study. G2E2 G3E1 LYPLAL1 is a member of the lysophospholipase gene G2E3 P = .05 family (EC number 3.1.1.5). It was initially identified G2E4 as a gene on chromosome 1 found incidentally during Figure 4: Significance level of main effect (ME) and/or gene by sex investigation of a familial chromosomal translocation [27]. interaction (GxS) associations with body mass index (BMI) and/or ∼ It was named on the basis of 30% predicted amino acid waist to hip ratio (WHR) for various loci of interest. sequence homology with lysophospholipases I and II [28]. The sequence suggests an α/β hydrolase fold typically found in many lipases and esterases. LYPLAL1 was subsequently identified as one of 23 esterolytic/lipolytic proteins extracted from mouse adipose tissue. The presence of an active site of lipid metabolism is not well understood, but recent serine was determined by activity tagging with a fluorescent discoveries and conflicting opinions warrant further studies ff probe of broad specificity, resembling a single-chain car- on LYPLAL1 and its potential roles and sex-specific e ects in boxylic acid ester. Similar probes modeling triglyceride and lipid metabolism and obesity. cholesteryl ester did not tag LYPLAL1 [29]. LYPLAL1 protein Our analysis revealed marked heterogeneity of effects has not yet been isolated, however, and its substrate speci- across different exams of the study, both in gene by sex ficity is unknown. Along with the gene for adipocyte triglyc- interaction and main effect analyses, even among established eride lipase and several others related to lipolysis, LYPLAL1 loci from other genome-wide association studies of BMI. The mRNA was expressed more abundantly in abdominal consensus approach appears to be robust, identifying a locus subcutaneous adipose tissue from obese versus lean humans with strong prior evidence of gene by sex interaction for [30]. obesity-related traits. Using this approach, we also identified Given the minimal characterization of LYPLAL1,we a possible novel candidate locus for BMI, located ∼26 kb can only speculate about its sex-specific role in adiposity. upstream of DUSP10, encoding a dual specificity protein It might be involved in triglyceride synthesis or lipolysis, phosphatase. The DUSPs are a subclass of the protein similar to some of the proteins with which it is coexpressed tyrosine phosphatase gene superfamily that controls MAP [31]. If indeed it is a lysophospholipase, it might play a kinase function [37]. role along with autotaxin, a secreted phospholipase D, in Our study was carried out in the Framingham Heart regulating extracellular levels of lysophosphatidic acid in Study Offspring cohort, a longitudinal, population-based adipose tissue. Via specific G protein-coupled receptors, study. Although no loci reached genome-wide significance in lysophosphatidic acid has been shown to have varying effects gene by sex interaction analyses, the longitudinal nature of on adipocyte differentiation and growth [32–34]. Another the data allowed us to prioritize SNPs based on consistency possibility relates to the endocannabinoid system, which has of effect across exams. However, data on waist circumference been a recent pharmacologic target for investigative obesity were available only at two exams, limiting the effectiveness treatments. The monoglyceride, 2-arachidonoyl glycerol, as of our approach for these traits. Nonetheless, for BMI, this well as other esters or amides of long-chain polyunsaturated approach yielded a plausible candidate sex-specific locus fatty acids belong to a family of compounds that are natural and another sex-independent locus. Interestingly, in both ligands for cannabinoid receptors. These endogenous sig- of these cases, results from Generation 3 were not as naling molecules affect physiologic and behavioral processes significant as in Generation 2, possibly reflecting a cohort governing appetite and energy metabolism [31]. effect: Generation 2 subjects were enrolled nearly a decade or Interestingly lipolysis control has been shown to vary more prior to Generation 3 subjects. Generation 3 subjects bysexinsomestudies[35] but not others [36]. The were on average more overweight than Generation 2 subjects aforementioned study showing support for sex differences at comparable ages, consistent with temporal trends of in lipolysis suggests that women show greater sensitivity increasing overweight/obesity observed in other population- to lipolysis in abdominal subcutaneous fat. The authors based studies. These differences, driven in large part by argue that the differences in lipolysis sensitivity are due to changes in diet and physical activity over time, may impact the presence of fewer inhibitory alpha-adrenergic receptors the heritability over time and thus, the ability to detect in the abdominal subcutaneous adipose tissue. This area genetic effects. Journal of Obesity 7

5. Conclusions [6] C. J. Willer, E. K. Speliotes, R. J. F. Loos et al., “Six new loci associated with body mass index highlight a neuronal Few studies have systematically modeled gene by sex inter- influence on body weight regulation,” Nature Genetics, vol. 41, action for obesity-related traits on a genome-wide level. no. 1, pp. 25–34, 2009. We confirm in our study that SNPs in the vicinity of [7] G. Thorleifsson, G. B. Walters, D. F. Gudbjartsson et al., LYPLAL1 may exhibit sex-specific effects on obesity-related “Genome-wide association yields new sequence variants at traits. By utilizing a well-designed population-based study, seven loci that associate with measures of obesity,” Nature and taking advantage of longitudinal data, we were able to Genetics, vol. 41, no. 1, pp. 18–24, 2009. demonstrate this effect using a much smaller sample size than [8] I. M. Heid, A. U. Jackson, J. C. Randall et al., “Meta- analysis identifies 13 new loci associated with waist-hip ratio the original meta-analysis that identified this locus. This has and reveals sexual dimorphism in the genetic basis of fat implications for the design of GWAS, where large samples distribution,” Nature Genetics, vol. 42, no. 11, pp. 949–960, sizes are often sought sometimes at the expense of population 2010. homogeneity. We suggest that smaller epidemiologically [9] M. C. Zillikens, M. Yazdanpanah, L. M. Pardo et al., “Sex- sound population-based studies may be more powerful specific genetic effects influence variation in body composi- than larger heterogeneous metacohorts. We also highlight tion,” Diabetologia, vol. 51, no. 12, pp. 2233–2241, 2008. the importance of considering longitudinal robustness of [10] S. Sammalisto, T. Hiekkalinna, K. Schwander et al., “Genome- association within a cohort as another means of prioritizing wide linkage screen for stature and body mass index in 3.032 loci and reducing false positive associations. Future studies of families: evidence for sex- and population-specific genetic LYPLAL1 are needed to determine the basis of the apparent effects,” European Journal of Human Genetics, vol. 17, no. 2, sex-specific effect on obesity. pp. 258–266, 2009. [11] L. D. Atwood, N. L. Heard-Costa, C. S. Fox, C. E. Jaquish, and L. A. Cupples, “Sex and age specific effects of chromosomal Acknowledgments regions linked to body mass index in the Framingham Study,” BMC Genetics, vol. 7, article 7, 2006. The authors would like to thank the National Heart, [12] C. M. Lindgren, I. M. Heid, J. C. Randall et al., “Genome-wide Lung and Blood Institute (NHLBI) for creating the open- association scan meta-analysis identifies three loci influencing access Framingham SHARe resource, and all of the study adiposity and fat distribution,” PLoS Genetics, vol. 5, no. 6, participants and Framingham Heart Study investigators Article ID e1000508, 2009. who helped to create this tremendously valuable resource. [13] C. S. Fox, N. Heard-Costa, L. A. Cupples, J. Dupuis, R. S. The Framingham Heart Study is conducted and supported Vasan,andL.D.Atwood,“Genome-wideassociationtobody by the NHLBI in collaboration with Boston University. mass index and waist circumference: the Framingham Heart This manuscript was not prepared in collaboration with Study 100K project,” BMC Medical Genetics,vol.8,no.1, investigators of the Framingham Heart Study and does not article no. S18, 2007. necessarily reflect the opinions or views of the Framingham [14] D. R. Govindaraju, L. Adrienne Cupples, W. B. Kannel et Heart Study, Boston University, or the NHLBI. This work al., “Chapter 2 genetics of the Framingham Heart Study population,” Advances in Genetics, vol. 62, pp. 33–65, 2008. was supported in part by National Institutes of Health [15] L. A. Cupples, H. T. Arruda, E. J. Benjamin et al., “The Grants: RO1 HL085191 to J. McCarthy, PI. Jennifer Rowell Framingham Heart Study 100K SNP genome-wide association was supported by NIH training Grant T32-DK007012-31 study resource: overview of 17 phenotype working group (Feinglos, PI), and Ashlee Benjamin by training Grant T32 reports,” BMC Medical Genetics, vol. 8, no. 1, article no. S1, GM071340 (Harer, PI). 2007. [16] S. Purcell, B. Neale, K. Todd-Brown et al., “PLINK: a tool set References for whole-genome association and population-based linkage analyses,” American Journal of Human Genetics, vol. 81, no. 3, [1] Y. Wang and M. A. Beydoun, “The obesity epidemic in the pp. 559–575, 2007. United States—gender, age, socioeconomic, racial/ethnic, and [17] C. Vince, “yags: Yet Another GEE Solver. R package version geographic characteristics: a systematic review and meta- 4.0-2,” 2004. regression analysis,” Epidemiologic Reviews,vol.29,no.1,pp. [18] B. Devlin and K. Roeder, “Genomic control for association 6–28, 2007. studies,” Biometrics, vol. 55, no. 4, pp. 997–1004, 1999. [2] QI. Zhang and Y. Wang, “Trends in the association between [19] S. A. Bacanu, B. Devlin, and K. Roeder, “Association studies obesity and socioeconomic status in U.S. adults: 1971 to 2000,” for quantitative traits in structured populations,” Genetic Obesity Research, vol. 12, no. 10, pp. 1622–1632, 2004. Epidemiology, vol. 22, no. 1, pp. 78–93, 2002. [3]I.S.Okosun,S.T.Choi,J.M.Boltrietal.,“Trendsof [20] A. L. Price, N. J. Patterson, R. M. Plenge, M. E. Weinblatt, abdominal adiposity in white, black, and Mexican-American N. A. Shadick, and D. Reich, “Principal components analysis adults, 1988 to 2000,” Obesity Research,vol.11,no.8,pp. corrects for stratification in genome-wide association studies,” 1010–1017, 2003. Nature Genetics, vol. 38, no. 8, pp. 904–909, 2006. [4] I. S. Okosun, K. M.D. Chandra, A. Boev et al., “Abdominal [21] D. Ge, K. Zhang, A. C. Need et al., “WGA Viewer: software adiposity in U.S. adults: prevalence and trends, 1960-2000,” for genomic annotation of whole genome association studies,” Preventive Medicine, vol. 39, no. 1, pp. 197–206, 2004. Genome Research, vol. 18, no. 4, pp. 640–643, 2008. [5] E. K. Speliotes, C. J. Willer, S. I. Berndt et al., “Association [22] T. J. P. Hubbard, B. L. Aken, S. Ayling et al., “Ensembl 2009,” analyses of 249,796 individuals reveal 18 new loci associated Nucleic Acids Research, vol. 37, no. 1, pp. D690–D697, 2009. with body mass index,” Nature Genetics, vol. 42, no. 11, pp. [23] J. H. Zhao, “Gap: genetic analysis package,” Journal of 937–948, 2010. Statistical Software, vol. 23, no. 8, pp. 1–18, 2007. 8 Journal of Obesity

[24] J. C. Barrett, B. Fry, J. Maller, and M. J. Daly, “Haploview: analysis and visualization of LD and haplotype maps,” Bioin- formatics, vol. 21, no. 2, pp. 263–265, 2005. [25] A. Herbert, N. P. Gerry, M. B. McQueen et al., “A common genetic variant is associated with adult and childhood obesity,” Science, vol. 312, no. 5771, pp. 279–283, 2006. [26] F. Renstrom,F.Payne,A.Nordstr¨ om¨ et al., “Replication and extension of genome-wide association study results for obesity in 4923 adults from northern Sweden,” Human Molecular Genetics, vol. 18, no. 8, pp. 1489–1496, 2009. [27] D. David, J. Cardoso, B. Marques et al., “Molecular characterization of a familial translocation implicates disruption of HDAC9 and possible position effect on TGFβ2inthe pathogenesis of Peters’ anomaly,” Genomics,vol.81,no.5,pp. 489–503, 2003. [28] A. Wang and E. A. Dennis, “Mammalian lysophospholipases,” Biochimica et Biophysica Acta, vol. 1439, no. 1, pp. 1–16, 1999. [29] R. Birner-Gruenberger, H. Susani-Etzerodt, M. Waldhubert et al., “The lipolytic proteome of mouse adipose tissue,” Molecular and Cellular Proteomics, vol. 4, no. 11, pp. 1710– 1717, 2005. [30] G. R. Steinberg, B. E. Kemp, and M. J. Watt, “Adipocyte triglyceride lipase expression in human obesity,” American Journal of Physiology, vol. 293, no. 4, pp. E958–E964, 2007. [31] R. Zechner, P. C. Kienesberger, G. Haemmerle, R. Zim- mermann, and A. Lass, “Adipose triglyceride lipase and the lipolytic catabolism of cellular fat stores,” Journal of Lipid Research, vol. 50, no. 1, pp. 3–21, 2009. [32] C. Pages,` A. Girard, O. Jeanneton et al., “LPA as a paracrine mediator of adipocyte growth and function,” Annals of the New York Academy of Sciences, vol. 905, pp. 159–164, 2000. [33] L. A. van Meeteren and W. H. Moolenaar, “Regulation and biological activities of the autotaxin-LPA axis,” Progress in Lipid Research, vol. 46, no. 2, pp. 145–160, 2007. [34] M. F. Simon, D. Daviaud, J. P.Pradere` et al., “Lysophosphatidic acid inhibits adipocyte differentiation via lysophosphatidic acid 1 receptor-dependent down-regulation of peroxisome proliferator-activated receptor γ2,” Journal of Biological Chem- istry, vol. 280, no. 15, pp. 14656–14662, 2005. [35] C. M. Williams, “Lipid metabolism in women,” Proceedings of the Nutrition Society, vol. 63, no. 1, pp. 153–160, 2004. [36] J. Bulow,¨ K. Gjeraa, L. H. Enevoldsen, and L. Simonsen, “Lipid mobilization from human abdominal, subcutaneous adipose tissue is independent of sex during steady-state exercise,” Clinical Physiology and Functional Imaging,vol.26,no.4,pp. 205–211, 2006. [37] M. Camps, A. Nichols, and S. Arkinstall, “Dual specificity phosphatases: a gene family for control of MAP kinase function,” FASEB Journal, vol. 14, no. 1, pp. 6–16, 2000.