Chapter 4. Haplotypes in the Phospholipid Transfer Protein Gene are Associated with Obesity-Related Phenotypes: The Québec Family Study

Yohan Bossé, Luigi Bouchard, Jean-Pierre Després, Claude Bouchard, Louis Pérusse, Marie-Claude Vohl

La protéine de transfert des phospholipides (PLTP) pourrait jouer un rôle dans la régulation du poids corporel. L’objectif était d’étudié l’association entre des polymorphismes du gène de la PLTP et des phénotypes d’adiposité. Deux variations introniques, localisées dans les intron 1 (-c.-87G>A) et 12 (c.1175+68T>G), ont été génotypées chez 811 sujets participant à l’Étude des familles de Québec. Des tests d’association familiale ont été réalisés pour chacun des polymorphismes ainsi que pour les haplotypes. L’allèle A de l’intron 1 était associé avec des moyennes phénotypiques plus élevées pour le poids, l’indice de masse corporelle, la circonférence de taille et la masse maigre. Pour les analyses d’haplotype, la transmission de l’haplotype AT était associée positivement avec les phénotypes d’adiposité, alors que l’haplotype GT semblait protecteur contre l’obésité. Le séquençage du promoteur et des parties codantes du gène n’a révélé aucune mutation pouvant expliquer ces résultats.

Haplotypes in the Phospholipid Transfer Protein Gene are Associated with Obesity-Related Phenotypes: The Québec Family Study

Yohan Bossé1,2, Luigi Bouchard3, Jean-Pierre Després2,4, Claude Bouchard5, Louis Pérusse3 and Marie-Claude Vohl1,2

1- Lipid Research Center, CHUL Research Center, Québec, Canada; 2- Department of Food Sciences and Nutrition, Laval University, Québec, Canada; 3- Division of Kinesiology, Department of Social and Preventive Medicine, Laval University, Québec, Canada; 4- The Quebec Heart Institute, Québec, Canada; 5- Pennington Biomedical Research Center, Baton Rouge, Louisiana, USA.

Running title: PLTP and obesity

Address all correspondence to:

Marie-Claude Vohl, Ph.D., Lipid Research Center, CHUL Research Center, TR-93, 2705 Laurier Blvd, Sainte-Foy, Québec, G1V 4G2, Canada. Phone: (418) 656-4141 extension 48280, Fax: (418) 654-2145, E-mail: marie-claude.vohl@crchul.ulaval.ca

Abstract

Some line of evidence suggested that the phospholipid transfer protein (PLTP) may play a role in body fat regulation. We thus investigated the association between PLTP genetic variants and obesity-related phenotypes. Two intronic variants, one in intron 1 (c.-87G>A ) and the other in intron 12 (c.1175+68T>G), were genotyped in 811 participants of the Québec Family Study. Ten obesity-related phenotypes were under study, including body-mass index (BMI), obesity (a dichotomous trait with a threshold of BMI ≥ 30 kg/m2), weight and waist circumference as well as percentage of fat, fat mass and fat-free mass assessed by the hydrostatic weighing technique and total, visceral and subcutaneous abdominal adipose tissue areas assessed by computed tomography. Single marker and haplotype tests of association in family-based studies were performed using the FBAT program. The SNP located in intron 1 showed significant association with obesity, weight, BMI, waist circumference and fat-free mass (p < 0.05). The low frequency allele (A allele) was associated with greater trait values suggesting that the transmission of this allele is associated with an increased risk of being obese. For haplotype analyses, significant associations were observed with obesity, waist circumference, percentage of fat and fat-free mass (p < 0.05). The transmission of the AT haplotype (frequency = 0.180) was positively associated with obesity-related phenotypes whereas the GT haplotype (frequency = 0.468) seemed to be protective against obesity. By sequencing the promotor and the coding regions of the PLTP gene, we were unable to identify a mutation that could provide functional meaning to the results. Considering the number and the relevance of candidate genes surrounding the PLPT locus, it is unclear whether PLTP itself is responsible for the association or the effect is mediated by a second gene allele in linkage disequilibrium with the marker locus.

Key words: PLTP, FBAT, haplotype, tests of association, obesity.

Introduction

The phospholipid transfer protein (PLTP), also referred to lipid transfer protein 2, belongs to the lipopolysaccharide binding/lipid transfer protein family, together with cholesteryl ester transfer protein (CETP), lipopolysaccharide-binding protein (LBP) and bactericidal/permeability-increasing protein (BPI). The PLTP gene was mapped to chromosome 20q12-q13.11 and encoded a mature protein of 476 residues2. The PLPT mRNA transcipt is detectable in a wide variety of tissues including pancreas, lung, kidney, heart, liver, skeletal muscle and brain3. The messenger is also highly expressed in subcutaneous and visceral adipose tissues with a depot-specific difference4,5. The predicted model structure of PLTP consists of two lipid-binding pockets characterized by apolar residues, with a N-terminal pocket critical for PLTP transfer activity and a C-terminal pocket involved in lipid binding6,7.

The initial physiological function ascribed to plasma PLTP was one of transfer of phospholipids from triglyceride-rich lipoproteins to high-density lipoproteins (HDL) during lipolysis8. Since then, animal and human studies have suggested that plasma PLTP level is an important factor in lipoprotein/lipid metabolism and atherosclerosis development9,10. More recently, PLTP activity has been shown to be positively and independently related to coronary artery disease11. Despite these great physiological insights, the role of PLTP in human metabolism and particularly the function of peripheral PLTP is still limited. Recently, some lines of evidences suggested that PLTP might play a role in the regulation of body fat content. First, the mRNA levels and the activity of PLTP have been consistently associated with obesity4,12-15. This tight relationship between PLTP and obesity is not fully understood. Kaser et al.12 have proposed that the increased synthesis of the protein may be the result of the enlarged mass of adipose tissue. This has been supported by Murdoch et al.16 who reported that PLTP activity is decreased following a diet-induced weight loss. Secondly, the inactivation of the PLTP gene in Caenorhabditis elegans by RNA interference (RNAi) cause increase in fat storage, suggesting that loss-of-function mutations in mammalian homologue could underlie obesity17. Third, two independent genome-wide scans provided significant evidence of linkage with obesity-related phenotypes within the PLTP region18,19. In addition, several mouse studies have suggested that genes influencing body fatness in mice reside on chromosome 2, a region homologous to human chromosome 20q20-22. Finally, it has been shown that PLPT facilitates the production of triglyceride-rich apoB-containing lipoproteins23,24 and facilitates as well the transport of lipids from cells25, which are central functions of lipid homeostasis. Based on these observations, we suspect that PLTP itself could modulate the level of adiposity. To test this hypothesis, we investigated the association between PLTP genetic variants and obesity-related phenotypes.

Materials and Methods

Subjects

Subjects were from the Québec Family Study (QFS) which is an ongoing project of French-Canadian families representing a mixture of random sampling and assortment through obese proband. This project was specifically design to understand the genetic basis of obesity and its comorbidities. Details of recruitment procedures have been published26. Only adults, 18 years and older, were considered for the present study. Table 1 presents the characteristics of the subjects. The Laval University Medical Ethics Committee approved the study, and all subjects provided written informed consent.

Anthropometry , body composition and fat distribution measurements

Body weight, height, and waist circumference were measured following standardized procedures27. Body density was measured by the hydrostatic weighing technique28. Pulmonary residual volume was assessed before immersion in the hydrostatic tank, using the helium dilution technique of Meneely and Kaltreider29. Percentage of body fat, fat mass and fat-free mass were derived from body density using the Siri equation30. Finally, a cross-sectional abdominal scan was performed by computed tomography using a Somatom DRH scanner (Siemens, Erlanger, Germany) to quantify the adipose tissue areas between L4 and L5 vertebra as described in detail elsewhere31.

SNPs selections and genotyping

To explore the possible involvement of PLTP in obesity-related phenotypes, we selected and genotyped two variants previously reported in dbSNP (rs394643 and rs553359) (Figure 1). The dbSNPs were chosen based on the number of chomosomes tested and their average estimated heterozygosity. They were then defined as c.-87G>A (denotes an intronic G to A substitution located 87 nucleotides downstream of the start codon) and c.1175+68T>G (denotes an intronic T to G substitution located 68 nucleotides upstream of the last nucleotide of exon 12) according to the nomenclature recommendations32. A total of 898 and 893 subjects were genotyped for the c.-87G>A and c.1175+68T>G polymorphisms, respectively. Genotyping was performed using a mini-sequencing assay33. PCR primers [forward (f), reverse (r)] and minisequencing (ms) primers were as follow: rs553359 (324 base pairs), f-5’-GGTCAGTAACATCCTCCTC-3’, r-5’-GACCCATTTGTTCATCTCTC-3’, ms-5’- AGGTATCACTGTACTTTAAGC-3’ rs394643 (365 base pairs), f-5’-CACGAGGGAACTGGGAACG-3’, r-5’-CGCCTTACCCAGCTCCAG-3’, ms-5’-GACGTCCAACCATAAGTGGG-3’. PCR conditions were as follow: In final volume of 6 μl, 20 ng of genomic DNA were added to a mixture containing a final concentration of dNTP (Amersham Pharmacia Biotech Inc.), 30μM each; Taq DNA polymerase (QUIAGEN), 0.3 U; buffer 1X [10 X: TRIS-HCl, KCL, (NH4)2SO4 and 15 mM MgCl2; pH 8.7 (20°C)]; MgCl2, 2.25 mM; flanking primers, 50 nM each. Following a 5-min denaturation step at 95°C, 30 PCR amplification cycles were performed as follow: denaturation at 95°C, 20 sec; annealing 60°C, 1 min; for 10 cycles and denaturation at 95°C, 20 sec; annealing at 57°C, 1 min; for the remaining 20 cycles. In the same well, the PCR mixture dNTP’s was digested using Shrimp Alkaline Phosphatase (USB), 0.2 U (final volume: 11 μl) for 15 min at 37°C follow by 20 min at 80°C. Mini-sequencing assay was performed in a final volume of 16 μl (in the same well); dTTP/ddNTP mix (dTTP, ddATP, ddCTP and ddGTP) for rs553359 and dGTP/ddNTP mix (dGTP, ddATP, ddTTP and ddCTP) for rs 394643 (dNTP and ddNTP are from Amersham Pharmacia Biotech Inc.), 1.56 μM each; IRDye tag primer, 3.125 nM (LICOR); Thermosequenase (USB), 0,3 U; 0.6 X buffer (10X: Tris-HCl, 260 mM, MgCl2, 65 mM, pH 9.5) were added to microplates. Following 2 min denaturation step at 95°C, 30 PCR amplification cycles were performed as follow: denaturation at 95°C, 10 sec; annealing at 60°C, 30 sec; extension at 72°C, 5 sec. Detection was done on a LICOR automated sequencer model 4200.

Sequencing

The promotor and the coding regions of the PLTP gene were sequenced in 19 subjects with different genotypes for the intronic variants and with different degrees of obesity. All exons and exon-intron splicing boundaries were amplified from genomic DNA by use of specific primers derived from the 5’ and 3’ ends of intronic sequence. We also sequenced up to 230 base pairs located downstream relative to the first transcriptional initiation site which is responsible for the full promotor activity34. Because of the particular genomic structure of the PLTP gene, characterized by small intron, some exons were amplified within the same fragment. Table 2 presents the specific primers of each fragment with their product size. All primers were designed using the Primer 3.0 software available on the Whitehead Insitute/MIT Center for Genome Research server (http://www-genome.wi.mit.edu/cgi-bin/primer/primer3.cgi). Amplification was performed by polymerase chain reaction using the thermal cycler, model PTC-200 (MJ Research, Watertown, MA). The cycler was programmed at 95oC for 5 min followed by 35 cycles of the following 3 min: 1 min at 95oC for denaturation, 1 min at annealing temperature, and 1 min at 72oC for elongation. The program was then completed with 10 min at 72oC. The annealing temperature was optimized for each pairs of primers by performing a 53oC to 67oC gradient assay using stock DNA (see Table 2). PCR conditions were as follow: reaction volume was 50 μL including 0.2 μL of AmpliTaq®DNA polymerase (Perkin-Elmer Cetus), 5 μL of 10X PCR buffer and 2.5 mM of MgCl2 as recommended by the manufacturer, 0.2 mM of dNTPs, 8.4 μL of each primer at a final concentration of 7.5 μM and 6 μL of genomic DNA at a final concentration of 20 ng/μL. PCR products were purified by the ABI ethanol-EDTA precipitation protocol, collected using a Beckman-Coulter Allegra 6R centrifuge, and resuspended in a 50% HiDi-formamide solution. Sequence reactions were performed using the BigDyeTH Terminator v3.1 kit and samples were run on ABI Prism® 3730/XL DNA Analyzer automated sequencers (Applied Biosystems, Foster City, CA). The sequences were then assembled and analyzed using the Staden preGAP4 and GAP4 programs35.

Linkage disequilibrium

Prior to calculate linkage disequilibrium, haplotype frequencies were estimated using the EH+ program available at this address: http://www.iop.kcl.ac.uk/IoP/Departments/PsychMed/GEpiBSt/software.shtml. Thereafter, these parameters were used to calculate pairwise measures of linkage disequilibrium (D’) using the 2LD program available at the same address.

Statistical analysis

Hardy-Weinberg equilibrium for genotype frequencies was evaluated using a χ2 test. We used the FBAT program to test association with either single SNPs or haplotypes and obesity-related phenotypes36 (http://www.biostat.harvard.edu/~fbat/default.html). The FBAT program performed family-based test of association that is efficient and robust to population admixture, phenotype distribution and ascertainment based on phenotype. It can also handle missing parental genotypes and/or missing phase in both offspring and parents for haplotype analysis. The approach holds as well for multi-locus and multi-allelic markers. The haplotype test is ideal for candidate gene studies with tightly linked markers (no recombination between the markers). Because of the high proportion of obese individuals in our cohort, the affection status was set as follows: 2 for affected (BMI ≥ 30 kg/m2), 1 for unaffected (BMI < 30 kg/m2) and 0 for unknown. These values allowed unaffected subjects to contribute to the analyses when association is tested with the dichotomous phenotype37. The minimum number of informative families necessary to compute the test statistics was set to 10. We first tested association with the global test (mode = multi-allelic) for intron 1 and intron 12 polymorphisms with the dichotomous and the quantitative phenotypes. For the global test, the FBAT program gives a χ2 statistic and its one-sided corresponding p-value. To know the effect of the transmited allele on the traits values, an univariate FBAT test was performed for each allele. This test provided a Z-statistic with the corresponding p-value. A positive Z-statistic is indicative of a high risk allele and a negative Z-statistic is indicative of a protective allele. This univariate FBAT statistic (Z-statistic) was also used to make inference regarding the effect of haplotypes of the PLTP gene on obesity-related phenotypes.

Results

A total of 811 subjects above 18 years old have been genotyped for the c.-87G>A and the c.1175+68T>G polymorphisms. For the c.-87G>A polymorphism, 215 subjects were homozygotes GG, 413 were heterozygotes GA and 183 were homozygotes AA. For the c.1175+68T>G polymorphism, 332 subjects were homozygotes TT, 382 were heterozygotes TG and 97 were homozygotes GG. The genotype distribution of both polymorphisms was in Hardy-Weinberg equilibrium (c.-87G>A, χ2 = 0.33, p=0.85; c.1175+68T>G, χ2 = 0.65, p=0.72). The two variants were in linkage disequilibrium with a D’ coefficient of 0.69 (p < 0.001).

Associations between single SNPs and the obesity-related phenotypes were tested using a family-based association test. The results of these analyses are presented in Table 3. The χ2 statistics with its one df corresponding p-value are presented for each obesity-related phenotypes. In addition, the effect of the low frequency allele (allele A for the c.-87G>A polymorphism and allele G for the c.1175+68T>G polymorphism) on the trait values are indicate, based on the Z-statistic provided by the univariate test performed for each allele. Of interest to note, since bi-allelic markers are under study, the p-value associated with the Z-statistic is the same to the one calculated by the χ2 statistic (the two tests are equivalent). The SNP located in intron 1 (c.-87G>A) showed significant association with obesity, weight, BMI, waist circumference and fat-free mass. The Z-statistics for allele A were positive for every obesity-related phenotypes suggesting that the transmission of this allele increases the risk of being obese. On the other hand, no associations were observed between the c.1175+68T>G polymorphism and phenotypes under study.

Haplotype association tests for family-based studies were also performed (Figure 2). For each haplotype, the Z-statistic is presented for the dichotomous (obesity) and the nine quantitative phenotypes. The haplotype frequencies were as follow: GT = 0.468, AG = 0.286, AT = 0.180 and GG = 0.066. The AT haplotype is significantly and positively associated with the dichotomous obesity phenotype, waist circumference, percent body fat and fat-free mass (p < 0.05), suggesting a high risk haplotype. The same trend (p < 0.1) is also observed for BMI, weight, fat mass and subcutaneous adipose tissue areas. On the other hand, the transmission of the GT haplotype seems to be protective against obesity but only the waist circumference phenotype reached statistical significance (p < 0.05).

In an attempt to identify a functional mutation in the PLTP gene that could explain the observed results, we sequenced the coding and the promotor regions of the gene. In addition to the two genotyped polymorphisms (rs394643 and rs553359), we identified three SNPs (Figure 1): 1- two heterozygous subjects for a C>G substitution in intron 1 (c.-601C>G) of the gene which was already in dbSNP (rs2294213), 2- one heterozygous subject carries a C>T substitution in intron 2 (C.100+42C>T), and 3- one heterozygous subject carries a G>A substitution in exon 6 (c.537G>A) which is a synonymous change. These three SNPs are located at position 84449, 83707 and 80249 in the DNA genomic sequence AL008726 (GenBank accession number), respectively. This attempt decreases the likelihood that a functional mutation in the PLTP gene, in linkage disequilibrium with the tested markers, is responsible for the significant association observed.

Discussion

In this study, we performed family-based tests of association between PLTP genetic variants and obesity-related phenotypes. Two SNP polymorphisms were genotyped, one located in intron 1 (c.-87G>A) and the second in intron 12 (c.1175+68T>G). Single marker association tests revealed significant associations between SNP in intron 1 and several indices of adiposity, including BMI (both as a quantitative trait and as a dichotomous trait with a threshold of BMI ≥ 30 kg/m2), weight and waist circumference. In these analyses the transmission of the low frequency allele (A allele) was associated with increased trait values. Despite of being in linkage disequilibrium with the SNP in intron 1, the SNP in intron 12 provided no evidence of association with any obesity-related phenotypes. Since multimarker haplotypes are likely to yield more genetic information than the study of a single marker, we also performed haplotype association tests for family-based study. Again significant associations were observed for some haplotypes. Indeed, the transmission of the AT haplotype increases the likelihood of being obese, whereas the transmission of the GT haplotype seems to be protective.

It is recognized that a significant association test could be the result of a functional variant in the gene in linkage disequilibrium with the tested marker. We thus verified this hypothesis by sequencing the promotor and the coding regions of the PLTP gene in subjects with different genotypes for the intronic variants and with different degrees of obesity. This attempt was undertaken to identify a mutation that could provide functional meaning to the results observed with intronic variants. Three additional SNPs were identified, the first one located in the 5’ untranslated region (c.-601C>G), the second one in intron 2 (c.100+42C>T) and the last one in exon 6 (c.537G>A) (Figure 1). None of these SNPs changed the amino acid sequence of the protein and their low frequencies make them unlikely to be the genetic variants responsible for the association. We thus ended up with a limited explanation. We might first suspect the intronic markers tested to be functional. Recently, it has been demonstrated that a silent substitution in intron 11 of the lamin A gene causes a rare disorder, called Hutchinson-Gilford Progeria Syndrome, that result in premature ageing and shortened lifespan38. Intronic SNPs have also been associated with rheumatoid arthritis39 and myocardial infarction40. Accordingly, if a silent mutation could produce such severe phenotypes, more subtle effect from this type of mutation might also be expected. However, the burden of the proof remains without functional studies evaluating the effect of the intronic variants on the protein products.

Different mechanisms could explain the association between PLTP and obesity-related phenotypes. First, the messenger and the activity of PLTP have been consistently related to obesity4,12-15. However, these studies used correlations to investigate the relationship and cannot determine a cause and effect. Although, it has been suggested that PLTP activity is influenced directly by body weight12,16, we might also suspect that PLTP itself is responsible for this tight relationship. Important functions governing lipid homeostasis have been ascribed to PLTP. In addition to its well established role in mediating the transfer of phospholipids between triglyceride rich lipoproteins and HDL in the intravascular compartment8, the hepatic PLTP in mice has been shown to play a major role in regulating the secretion of apoB-containing lipoproteins24. Furthermore, peripheral PLTP is also known to enhance cellular lipids efflux. Indeed, Wolfbauer et al.41 have demonstrated that PLTP increases cholesterol and phospholipid efflux from cholesterol-loaded human fibroblasts. Subsequently, the same group reported similar observations in murine macrophages and hamster kidney cells and shown that PLTP mediates its effect via the ATP-binding cassette transporter A1 (ABCA1) pathway25. This process might play an important role in enhancing flux of lipids from tissues. Given the expression of ABCA142 and PLTP4 in the adipose tissue, such transport mechanism could prevent lipid storage into adipocytes, which is the long-term process leading to obesity.

On the other hand, it is possible that the PLTP allele significantly associated with obesity-related phenotypes in our study is not responsible for the effect. In fact, the association may arise because of a second gene allele in linkage disequilibrium with the PLPT marker locus. A first glance of genes located in proximity of the PLTP gene revealed multiple candidate genes43. First, two genes, the melanocortin 3 receptor (MC3R) and the adenosine deaminase (ADA), located approximately 10.3 Mb and 1.3 Mb from the PLTP gene, respectively, have been significantly associated with obesity-related phenotypes44,45. The agouti signaling protein (ASIP), is also located within the relevant region of human chromosome 20. The ASIP is the human orthologue of the mouse agouti gene, which is a single-gene mutation model of obesity46. Finally, three candidate genes located in the vicinity of the PLTP locus namely the CCAAT enhancer binding protein (CEBPB), the protein tyrosine phosphatase-1B (PTPN1), and the growth hormone releasing hormone (GHRH) have been revealed as obesity-modifying genes in knockout and transgenic experiments47-49. Accordingly, considering the number of interesting candidate genes potentially implicated in body weight and fatness nearby the PLTP locus, it is highly plausible that the significant associations observed in the present study are due to an allele in a second gene in linkage disequilibrium with the marker locus.

In conclusion, we reported for the first time significant associations between PLTP genetic variants and obesity-related phenotypes. The associations were carry out using both single locus and haplotypes analyses based on family study. By sequencing all the exons and the promotor region of the gene, we were unable to identify a mutation that could explain the significant associations with the intronic variants. Although, some evidence suggested that PLTP itself may be responsible for the association, the number and the relevance of candidate genes surrounding the PLTP locus highly suggested that an allele in a second gene in linkage disequilibrium with the PLTP markers is responsible for the association. Further studies will be required to elucidate this uncertainty.

Acknowledgments

This study was supported by the Canadian Institutes of Health Research (MOP-13960). The authors would like to express their gratitude to the subjects for their excellent collaboration and to the staff of the Physical Activity Sciences Laboratory for their contribution to the study. Y. Bossé is the recipient of a studentship from the Canadian Institutes of Health Research. L. Bouchard is recipient of a studentship from the "Fonds pour la formation de chercheurs et l'aide à la recherche (FCAR) et le Fonds de la recherche en santé du Québec (FRSQ)". M.C. Vohl is a research scholar of the FRSQ. J.P. Després is chair professor of human nutrition, lipidology and prevention of cardiovascular disease supported by Provigo and Pfizer Canada. C. Bouchard is partially funded by the George A. Bray Chair in Nutrition.

References

1. Whitmore TE, Day JR, Albers JJ. Localization of the human phospholipid transfer protein gene to chromosome 20q12-q13.1. Genomics . 1995;28:599-600.

2. Albers JJ, Tu AY, Wolfbauer G, et al. Molecular biology of phospholipid transfer protein. Curr Opin Lipidol . 1996;7:88-93.

3. Day JR, Albers JJ, Lofton-Day CE, et al. Complete cDNA encoding human phospholipid transfer protein from human endothelial cells. J Biol Chem . 1994;269:9388-91.

4. Dusserre E, Moulin P, Vidal H. Differences in mRNA expression of the proteins secreted by the adipocytes in human subcutaneous and visceral adipose tissues. Biochim Biophys Acta . 2000;1500:88-96.

5. Linder K, Arner P, Flores-Morales A, et al. Differentially expressed genes in visceral or subcutaneous adipose tissue of obese men and women. J Lipid Res . 2003.

6. Huuskonen J, Wohlfahrt G, Jauhiainen M, et al. Structure and phospholipid transfer activity of human PLTP: analysis by molecular modeling and site-directed mutagenesis. J Lipid Res . 1999;40:1123-30.

7. Ponsin G, Qu SJ, Fan HZ, et al. Structural and functional determinants of human plasma phospholipid transfer protein activity as revealed by site-directed mutagenesis of charged amino acids. Biochemistry . 2003;42:4444-51.

8. Tall AR, Krumholz S, Olivecrona T, et al. Plasma phospholipid transfer protein enhances transfer and exchange of phospholipids between very low density lipoproteins and high density lipoproteins during lipolysis. J Lipid Res . 1985;26:842-51.

9. van Tol A. Phospholipid transfer protein. Curr Opin Lipidol . 2002;13:135-9.

10. Huuskonen J, Ehnholm C. Phospholipid transfer protein in lipid metabolism. Curr Opin Lipidol . 2000;11:285-9.

11. Schlitt A, Bickel C, Thumma P, et al. High Plasma Phospholipid Transfer Protein Levels as a Risk Factor for Coronary Artery Disease. Arterioscler Thromb Vasc Biol . 2003;23:1857-1862.

12. Kaser S, Sandhofer A, Foger B, et al. Influence of obesity and insulin sensitivity on phospholipid transfer protein activity. Diabetologia . 2001;44:1111-7.

13. Cheung MC, Knopp RH, Retzlaff B, et al. Association of plasma phospholipid transfer protein activity with IDL and buoyant LDL: impact of gender and adiposity. Biochim Biophys Acta . 2002;1587:53-9.

14. Dullaart RP, Sluiter WJ, Dikkeschei LD, et al. Effect of adiposity on plasma lipid transfer protein activities: a possible link between insulin resistance and high density lipoprotein metabolism. Eur J Clin Invest . 1994;24:188-94.

15. Murdoch SJ, Carr MC, Hokanson JE, et al. PLTP activity in premenopausal women. Relationship with lipoprotein lipase, HDL, LDL, body fat, and insulin resistance. J Lipid Res . 2000;41:237-44.

16. Murdoch SJ, Kahn SE, Albers JJ, et al. PLTP activity decreases with weight loss: changes in PLTP are associated with changes in subcutaneous fat and FFA but not IAF or insulin sensitivity. J Lipid Res . 2003;44:1705-12.

17. Ashrafi K, Chang FY, Watts JL, et al. Genome-wide RNAi analysis of Caenorhabditis elegans fat regulatory genes. Nature . 2003;421:268-72.

18. Lee JH, Reed DR, Li WD, et al. Genome scan for human obesity and linkage to markers in 20q13. Am J Hum Genet . 1999;64:196-209.

19. Norman RA, Tataranni PA, Pratley R, et al. Autosomal genomic scan for loci linked to obesity and energy metabolism in Pima Indians. Am J Hum Genet . 1998;62:659-68.

20. Lembertas AV, Perusse L, Chagnon YC, et al. Identification of an obesity quantitative trait locus on mouse chromosome 2 and evidence of linkage to body fat and insulin on the human homologous region 20q. J Clin Invest . 1997;100:1240-7.

21. Pomp D. Genetic dissection of obesity in polygenic animal models. Behav Genet . 1997;27:285-306.

22. Mehrabian M, Wen PZ, Fisler J, et al. Genetic loci controlling body fat, lipoprotein metabolism, and insulin levels in a multifactorial mouse model. J Clin Invest . 1998;101:2485-96.

23. van Haperen R, van Tol A, van Gent T, et al. Increased risk of atherosclerosis by elevated plasma levels of phospholipid transfer protein. J Biol Chem . 2002;277:48938-43.

24. Jiang XC, Qin S, Qiao C, et al. Apolipoprotein B secretion and atherosclerosis are decreased in mice with phospholipid-transfer protein deficiency. Nat Med . 2001;7:847-52.

25. Oram JF, Wolfbauer G, Vaughan AM, et al. Phospholipid transfer protein interacts with and stabilizes ABCA1 and enhances cholesterol efflux from cells. J Biol Chem . 2003; 278:52379-85.

26. Bouchard C. Genetic epidemiology, association, and sib-pair linkage: Results from the Québec Family Study. In: Bray GA, Ryan DH, eds. Molecular and genetic aspects of obesity . Baton Rouge, LA: Louisiana State University Press; 1996:470-81.

27. The Airlie (VA) Consensus Conference. In: Lohman T, Roche A, Martorel R, eds. Standardization of anthropometric measurements. Champaign, IL: Human Kinetics; 1988:39-80.

28. Behnke AR, Wilmore JH. Evaluation and regulation of body build and composition . Englewood Cliffs, NJ: Prentice-Hall; 1974.

29. Meneely GR, Kaltreider NL. Volume of the lung determined by helium dilution. J Clin Invest . 1949;28:129-139.

30. Siri WE. The gross composition of the body. Adv Biol Med Phys . 1956;4:239-280.

31. Despres JP, Prud'homme D, Pouliot MC, et al. Estimation of deep abdominal adipose-tissue accumulation from simple anthropometric measurements in men. Am J Clin Nutr . 1991;54:471-7.

32. den Dunnen JT, Antonarakis SE. Nomenclature for the description of human sequence variations. Hum Genet . 2001;109:121-4.

33. Sun X, Ding H, Hung K, et al. A new MALDI-TOF based mini-sequencing assay for genotyping of SNPS. Nucleic Acids Res . 2000;28:E68.

34. Tu AY, Wolfbauer G, Albers JJ. Functional characterization of the promoter region of the human phospholipid transfer protein gene. Biochem Biophys Res Commun . 1995;217:705-11.

35. Bonfield JK, Smith K, Staden R. A new DNA sequence assembly program. Nucleic Acids Res . 1995;23:4992-9.

36. Laird NM, Horvath S, Xu X. Implementing a unified approach to family-based tests of association. Genet Epidemiol . 2000;19 Suppl 1:S36-42.

37. Lunetta KL, Faraone SV, Biederman J, et al. Family-based tests of association and linkage that use unaffected sibs, covariates, and interactions. Am J Hum Genet . 2000;66:605-14.

38. Eriksson M, Brown WT, Gordon LB, et al. Recurrent de novo point mutations in lamin A cause Hutchinson-Gilford progeria syndrome. Nature . 2003;423:293-8.

39. Tokuhiro S, Yamada R, Chang X, et al. An intronic SNP in a RUNX1 binding site of SLC22A4, encoding an organic cation transporter, is associated with rheumatoid arthritis. Nat Genet . 2003.

40. Ozaki K, Ohnishi Y, Iida A, et al. Functional SNPs in the lymphotoxin-alpha gene that are associated with susceptibility to myocardial infarction. Nat Genet . 2002;32:650-4.

41. Wolfbauer G, Albers JJ, Oram JF. Phospholipid transfer protein enhances removal of cellular cholesterol and phospholipids by high-density lipoprotein apolipoproteins. Biochim Biophys Acta . 1999;1439:65-76.

42. Langmann T, Klucken J, Reil M, et al. Molecular cloning of the human ATP-binding cassette transporter 1 (hABC1): evidence for sterol-dependent regulation in macrophages. Biochem Biophys Res Commun . 1999;257:29-33.

43. Chagnon YC, Rankinen T, Snyder EE, et al. The human obesity gene map: the 2002 update. Obes Res . 2003;11:313-67.

44. Boucher N, Lanouette CM, Larose M, et al. A +2138InsCAGACC polymorphism of the melanocortin receptor 3 gene is associated in human with fat level and partitioning in interaction with body corpulence. Mol Med . 2002;8:158-65.

45. Bottini E, Gloria-Bottini F. Adenosine deaminase and body mass index in non-insulin-dependent diabetes mellitus. Metabolism . 1999;48:949-51.

46. Wilson BD, Ollmann MM, Kang L, et al. Structure and function of ASP, the human homolog of the mouse agouti gene. Hum Mol Genet . 1995;4:223-30.

47. Tanaka T, Yoshida N, Kishimoto T, et al. Defective adipocyte differentiation in mice lacking the C/EBPbeta and/or C/EBPdelta gene. Embo J . 1997;16:7432-43.

48. Elchebly M, Payette P, Michaliszyn E, et al. Increased insulin sensitivity and obesity resistance in mice lacking the protein tyrosine phosphatase-1B gene. Science . 1999;283:1544-8.

49. Cai A, Hyde JF. The human growth hormone-releasing hormone transgenic mouse as a model of modest obesity: differential changes in leptin receptor (OBR) gene expression in the anterior pituitary and hypothalamus after fasting and OBR localization in somatotrophs. Endocrinology . 1999;140:3609-14.

Table 1. Characteristics of the Subjects.

Table 2. PCR primers for genomic amplification of PLTP promotor and exons.

Table 3. Global tests of association between SNPs in the PLTP gene and obesity–related phenotypes.

Z-statistic of the low frequency allele, positive Z-statistics are indicative of a high risk allele and negative values are indicative of a protective allele.

§Obesity is a dichotomous trait based on BMI criteria.

Figure 1. Genomic organization of the PLTP gene. The location of the genotyped variants are indicated with black arrows. The additional variants identified by sequencing are indicated with grey arrows. The 16 exons are shown as vertical bars whose width corresponds to their base-pairs length. Coding regions are in black and untranslated regions are in grey. Values in parentheses are the frequencies of the rare allele.

Figure 2. Haplotype-specific association tests in the PLTP gene and obesity related-phenotypes. The test was conducted on 10 phenotypes indicated above each graph. The phenotype labelled “obesity” is a dichotomous trait based on BMI above and below 30 kg/m2. Each bar represents the Z-statistic for one of the four haplotypes constructed with the two polymorphisms (c.-87G>A and c.1175+68T>G). A positive Z-statistics are indicative of a high risk haplotype and negative values are indicative of a protective haplotype. The haplotypes frequencies were as follow: GT = 0.468, AG = 0.286, AT = 0.180, GG = 0.066. AT, adipose tissue. *p < 0.1, †p < 0.05.