点击显示 收起
1 From the Lipid Research Center (A-MP, BF-B, YB, JR, AT, PC, and M-CV) and the Molecular Endocrinology and Oncology Laboratory Research Center (AT), CHUQ-CHUL Pavilion, Sainte-Foy, Canada; the Food Science and Nutrition Department, Laval University, Québec (A-MP, YB, JR, SL, HJ, BL, AT, and M-CV); and the Nutraceuticals and Functional Foods Institute, Sainte-Foy, Canada (A-MP, SL, HJ, BL, AT, and M-CV).
2 Supported in part by Valorisation-Recherche-Québec, the Canada Research Chair in Nutrition, Functional Food and Cardiovascular Health and the Fonds de la recherche en santé du Québec (FRSQ). YB is a recipient of a doctoral scholarship from the Canada Graduate Scholarship program. JR received a doctoral studentship from the Canadian Institutes of Health Research (CIHR). BL is the director of the Canada Research Chair in Nutrition, Functional Foods and Cardiovascular Health. AT is a recipient of a new investigator scholarship from CIHR. PC is a recipient of a clinical scholarship from the FRSQ. SL and M-CV are research scholars from FRSQ. 3 Address reprint requests to M-C Vohl, Lipid Research Center, CHUQ-CHUL, 2705 Laurier Boulevard, TR-93, Sainte-Foy, Canada, G1V 4G2. E-mail: marie-claude.vohl{at}crchul.ulaval.ca.
ABSTRACT
Background: Serum lipid responses to dietary modification are partly determined by genetic factors.
Objective: We tested whether plasma lipoprotein and lipid responsiveness to a modification in the dietary ratio of polyunsaturated to saturated fatty acids (P:S) is influenced by the peroxisome proliferator-activated receptor (PPAR) Leu162Val polymorphism in healthy men.
Design: Ten carriers of the V162 allele and 10 L162 homozygotes were matched according to age and body mass index (BMI). During the protocol, all subjects followed the National Cholesterol Education Program Step I diet, but intake of saturated and polyunsaturated fatty acids was adjusted to obtain a P:S of 0.3 for the first 4-wk period (low-P:S diet) and a P:S of 1.0 for the next 4-wk period (high-P:S diet).
Results: At screening, the PPAR Leu162Val polymorphism was not associated with anthropometric indexes or plasma lipoprotein and lipid concentrations. After the high-P:S diet, a significant gene-by-diet interaction was observed for changes in plasma total cholesterol, apolipoprotein (apo) A-I, and cholesterol concentrations in small LDL particles (P 0.05). Mean differences after the high-P:S diet were observed between genotype groups for plasma apo A-I concentrations (P < 0.05). Changes in BMI, waist circumference, and concentrations of triacylglycerol, phospholipid, and apo B did not differ significantly between groups.
Conclusion: The PPAR Leu162Val polymorphism may contribute to interindividual variability in plasma lipoprotein and lipid response after modification of the dietary P:S ratio.
Key Words: Peroxisome proliferator-activated receptor • PPAR • fatty acids • polyunsaturated fatty acids • gene-by-diet interaction • nutritional intervention • interindividual variability
INTRODUCTION
Common diseases such as cardiovascular diseases, type 2 diabetes, obesity, and some cancers result from the interaction between genetic and environmental factors, including diet. Regarding dietary intake, diets high in saturated fatty acids (SFAs) and low in polyunsaturated fatty acids (PUFAs) increase blood cholesterol concentrations (1, 2). However, studies have shown that when SFAs are replaced by unsaturated fatty acids, total plasma cholesterol concentrations are lowered (3). Moreover, the study by Judd et al (4) suggested that a moderate reduction in dietary fat intake combined with an increase in the ratio of polyunsaturated to saturated fatty acids (P:S) from 0.3 to 1.0 leads to significant beneficial changes in the lipoprotein and lipid profile. On the other hand, interindividual variations in the plasma lipoprotein and lipid response to dietary changes are generally considerable (5, 6). Many studies have shown that the heterogeneity in plasma lipoprotein and lipid responsiveness to change in dietary fat is partly explained by variation in genes whose products affect lipoprotein metabolism (6-8).
Peroxisome proliferator-activated receptors (PPARs) have emerged as one of the central regulators of gene-by-diet interactions (9). The PPAR family of nuclear receptors is composed of 3 members: , , and . PPAR is involved in glucose and lipid metabolism and thus may affect the etiology of dyslipidemia, atherosclerosis, obesity, insulin resistance, and type 2 diabetes. PPAR is highly expressed in tissues responsible for fatty acid catabolism, particularly the liver, kidney, heart, and muscle (10). This ligand-inducible transcription factor regulates the expression of genes involved in fatty acid oxidation, extracellular lipid metabolism (11), homeostasis (12), and inflammation (13). Ligands for PPAR include long-chain PUFAs, eicosanoids, peroxisome proliferators, nonsteroidal antiinflammatory drugs, and the fibric acid derivative class of hypolipidemic drugs (14). It has been well documented that fibrates lower plasma triacylglycerol concentrations and increase HDL cholesterol concentrations (15). A molecular scanning of the human PPAR gene revealed a leucine to valine substitution at codon 162 (L162V polymorphism) associated with alterations of the lipoprotein-lipid profile (16, 17). Accordingly, the less common allele (V162) has been associated with an increase in serum concentrations of total cholesterol (16, 18, 19), LDL cholesterol (17, 19), apolipoprotein (apo) B (17-19), HDL cholesterol (16), apo A-1 (16), and apo C-III (19), depending on the population studied. Long-chain PUFAs have also been shown to bind and activate PPAR (20). In this context, the objective of the present study was to examine whether plasma lipoprotein and lipid responsiveness to a modification in the P:S ratio is influenced by the PPAR L162V gene polymorphism in healthy men.
SUBJECTS AND METHODS
Subjects
Sixty-six healthy white men from the greater Quebec City metropolitan area were recruited through the media and screened for the presence of the PPAR L162V polymorphism. Carriers of the V162 allele were matched to L162 homozygotes according to age and body mass index (BMI). Twenty subjects were selected: 10 carriers and 10 noncarriers of the V162 allele. To be eligible, men had to be nonsmokers and free of any thyroid or metabolic disorders requiring treatment, such as diabetes, hypertension, severe dyslipidemia, and coronary heart disease. Moreover, to better follow the dietary modifications proposed in the study, the volunteers were asked to be in charge of their food purchases and meal preparation. Men were aged between 23 and 49 y, were nonsmokers, and were not using medications known to affect plasma lipoprotein or lipid concentrations. All subjects gave their written consent to participate in the study, which was approved by the ethics committees of Laval University and Laval University Hospital Research Center.
Study design and diets
During the screening period, the 20 selected subjects completed a 3-d food record (2 weekdays and 1 weekend day) in which they were required to provide information on their usual food intake. Individual dietary instructions were then given by a dietitian to achieve the National Cholesterol Education Program (NCEP) Step I diet guidelines (21). All subjects were asked to maintain this diet throughout the protocol. However, because of protocol specification, SFA and PUFA intakes did not meet the NCEP Step I recommendations. During the low-P:S diet, specific instructions were given for SFA and PUFA intakes to obtain a P:S of 0.3. This ratio was reached by adding 1 or 2 tablespoons of butter to the diet per day. The volunteers were asked to use butter as a spreading fat and for baking and cooking foods to avoid using vegetable oils. During the high-P:S diet, the subjects followed the same recommendations, but the intake of SFAs and PUFAs was adjusted to obtain a P:S of 1.0. To achieve that, the butter was replaced by 1 or 2 tablespoons of sunflower oil (La Maison Orphée, Québec, Canada) per day (depending on the intake of butter during the low-P:S diet). In addition, the subjects were asked to use nonhydrogenated (free of trans fat) margarine as a spreading fat and to avoid butter. Margarine contained 13.8% SFAs, 33.8% PUFAs, and 45% monounsaturated fatty acids (MUFAs) and was provided by Becel (Uniliver Canada, Toronto). To avoid changes in dietary cholesterol from the low-P:S diet, quail eggs, which are rich in cholesterol (76 mg) (22), were added to the diet during the high-P:S diet. The subjects were asked to abstain from consuming alcohol and to report any deviations during the protocol. The subjects received butter, sunflower oil, margarine, and quail eggs in sufficient quantities. Compliance was assessed from the return of food items (full and empty), and each product was weighed to determine the quantity consumed. Moreover, during each phase, a dietitian made a phone call follow-up to ensure that the participants were achieving the desired nutritional changes. During each phase of the protocol, the subjects received detailed written and oral instructions on their diet. All participants self-selected and prepared their foods during the protocol.
A dietitian administrated a validated food-frequency questionnaire (FFQ) at weeks 0, 4, and 8 to each participant (23). This FFQ is based on typical food items available in Quebec and contains 91 items; 27 items had between 1 and 3 subquestions. The subjects were asked how often they consumed each item per day, per week, per month, or none at all during the last month. Many examples of portion size were provided for a better estimation of the real portion consumed by the subject. Moreover, the estimated quantity consumed for butter, sunflower oil, and margarine was the quantity determined from the return of food items. The subjects completed a 3-d food record before beginning the study and then again at weeks 3 and 7. Dietary data included both foods and beverages consumed at home and outside. A dietitian provided instructions on how to complete the food record with some examples and a written copy of these examples. All foods and beverages consumed on 2 representative weekdays and 1 weekend day were weighed or estimated and recorded in food diaries. Data from the food record were used to validate the FFQ, and data from the FFQ were used for statistical analysis. To verify whether physical activity was maintained at a constant level, the subjects completed the Leisure Time Physical Activity Questionnaire (24) with the dietitian at weeks 0, 4, and 8. Nutrient intakes from the food record and the FFQ were calculated with Nutrition Data System for Research (NDS-R) software, version 4.03 (Nutrition Coordination Center, University of Minnesota, Minneapolis), and the Food and Nutrient Database 31, released in November 2000 (25).
Anthropometric measurements
Body weight, height, and waist girth were measured according to the procedures recommended by the Airlie Conference (26) and were taken at weeks 0, 4, and 8. BMI was calculated as weight per meter squared (kg/m2).
Blood variables
Blood samples were collected from an antecubital vein into evacuated tubes containing EDTA. A first blood sample was taken to determine the PPAR L162V genotype of each participant. Selected participants were asked to give another blood sample to eliminate any metabolic disorder. Afterward, each subject had 3 fasting blood samplings at weeks 0, 4, and 8. No alcohol consumption was allowed for 48 h before blood sampling. Blood samples were immediately centrifuged (2500 x g, 10 min, 4 °C), and the plasma was portioned and frozen for subsequent measurements of fasting plasma insulin and glucose concentrations and electrophoretic LDL properties.
Cholesterol and triacylglycerol concentrations were measured in plasma and lipoprotein fractions by using a Technicon RA-500 (Bayer Corporation Inc, Tarrytown, NY) (27). Plasma VLDL was isolated by ultracentrifugation (d < 1.006 g/mL), and the HDL fraction was obtained after precipitation of LDL in the infranatant fluid (d > 1.006 g/mL) with heparin-manganese chloride (28). LDL-cholesterol concentrations were estimated by use of the equation of Friedewald et al (29). The cholesterol content of the HDL2 and HDL3 subfractions was determined after further precipitation of HDL2 with dextran sulfate (30). Apo B and apo A-1 concentrations and lipoprotein fractions were measured in plasma by the rocket immunoelectrophoretic method of Laurell (31).
Nondenaturing 2% to 16% polyacrylamide gradient gel electrophoresis was used to characterize LDL peak particle size as previously described (32). Briefly, gels were prepared in batches. Aliquots of 3.5 mL of whole plasma samples taken at weeks 0, 4, and 8 and kept at –80 °C were mixed 1:1 with a sampling buffer containing 20% sucrose and 0.25% bromophenol blue and were loaded onto the gels. After a 15-min pre-run at 75 V, electrophoresis was performed at 150 V for 3 h. Gels were stained for 1 h with Sudan black (0.07%) and were stored in a 0.81% acetic acid, 4% methanol solution. Gels were analyzed by using Imagemaster 1-D Prime computer software (Amersham Pharmacia Biotech, Baie d'Urfé, Québec, Canada). LDL size was extrapolated from the relative migration of 4 plasma standards of known diameter. The estimated diameter of the major peak in each scan was identified as the LDL peak particle diameter. An integrated LDL diameter (LDL weight) was also computed by using the approach described previously (32). It was calculated as a continuous variable and was computed as the sum of the diameter of each LDL subclass multiplied by its relative area. Analysis of pooled plasma standards showed that measurement of LDL peak and integrated particle size was highly reproducible, with an interassay CV of 0.6%. The relative proportion of small LDL particles (<25.5 nm) was determined by computing the relative area of the densitometric scan <25.5 nm. The absolute concentration of cholesterol in the LDL subfraction with a diameter <25.5 nm was calculated by multiplying the relative proportion by the total plasma LDL-cholesterol concentrations as described earlier (32). A similar approach was used to assess the relative and absolute concentrations of cholesterol in particles of intermediate (25.5–26.0 nm) and large (>26.0 nm) size.
Phospholipids were measured by colorimetry by using a WAKO Phospholipids B (990-54009) kit (WAKO Chemicals GmbH, Neuss, Germany) with a Technicon RA-500 analyzer (Bayer Corporation Inc) as described by the manufacturer. Fasting plasma insulin was measured by radioimmunoassay (Linco Research, St Louis), and fasting plasma glucose was measured enzymatically (Sigma, St Louis).
DNA analysis
Genetic analyses were performed on genomic DNA isolated from human leukocytes. The PPAR L162V polymorphism was determined by the polymerase chain reaction–restriction fragment length polymorphism (PCR-RFLP) method, as previously described (17). PCR products were then digested with HinfI, electrophoresed through a 12%-acrylamide gel, stained with ethidium bromide, and visualized on an ultraviolet light box. The common three-allele (2/3/4) variation in the APOE gene was analyzed with the PCR-RFLP method described by Hixson and Vernier (33). We were unable to determine the APOE genotype of one carrier of the V162 allele.
Statistical analyses
Variables not normally distributed (triacylglycerol, fasting insulin, and VLDL cholesterol) were log10 transformed before analyses. Subject with the V/V (1 subject) and L/V (9 subjects) genotypes were combined and compared with the L/L homozygotes (10 subjects). Pearson correlation coefficients were computed to assess associations between the data from the 3-d food records and the FFQ. Two-tailed unpaired Student's t test was applied to compare means of each subject's characteristics at screening according to the PPAR L162V polymorphism. Fisher's exact test was used to compare the distribution of APOE alleles in the PPAR genotype groups. Differences in energy and nutrient intakes between phases were tested by using repeated-measured analysis of variance (general linear model) followed by Tukey's honestly significant difference test for pairwise comparisons between the 3 phases. A PROC MIXED procedure was performed to evaluate whether genotype interacted with diet. Finally, plasma lipoprotein and lipid changes in response to the high-P:S diet were compared between carriers of the V162 allele and L162 homozygotes by using the general linear model. Statistical analyses were performed with SAS statistical software, version 8.2 (SAS Institute Inc, Cary, NC). Statistical significance was defined as P 0.05.
RESULTS
All subjects successfully completed the study. V162 allele frequency was calculated in all 66 subjects screened and reached 10.6%. The frequency of the rare allele is established to be 6.2% in the healthy European population (16) and reached 6.6% in another sample of healthy men from the greater Quebec City area (17).
The subjects' characteristics at the beginning of the low-P:S diet (week 0) are presented in Table 1 according to PPAR L162V genotype. No significant differences were observed between the 2 genotype groups for anthropometric indexes or plasma lipoprotein and lipid concentrations. There were no homozygotes for the apo E2 and apo E4 alleles, but there were 2 apo E2 carriers in the L162 group.
View this table:
TABLE 1. Subjects' characteristics at screening according to the peroxisome proliferator-activated (PPAR) Leu162Val polymorphism
Data provided from the 3-d food records at screening and the FFQ correlated for energy (r = 0.46, P 0.05) and percentage of energy from fat (r = 0.77, P < 0.0001), SFAs (r = 0.50, P 0.005), MUFAs (r = 0.73, P 0.0005), and PUFAs (r = 0.43, P = 0.05). Nutrient intakes were similar in both genotype groups at screening and during the protocol (data not shown). As shown in Table 2, MUFAs were higher during the high-P:S diet. As expected, the P:S was higher during the high-P:S diet than during the low-P:S diet. Alcohol consumption decreased after screening and remained stable throughout the protocol. No changes in the relative intakes of energy, fat, carbohydrates, protein, total dietary fiber, and cholesterol were observed.
View this table:
TABLE 2. . Daily energy and nutrient intakes1
There were no significant differences in questionnaire-based physical activity throughout the study (data not shown). A slight weight loss (1.38 ± 0.30 kg) was observed during the protocol (week 0 to 8), but there was no significant difference between genotype groups (data not shown).
To test the potential interaction between the PPAR L162V polymorphism and the high-P:S diet on plasma lipoprotein and lipid concentrations, a PROC MIXED procedure was performed. As shown in Table 3, a significant genotype-by-diet interaction was observed for changes in plasma total cholesterol, apo A-1, and cholesterol concentrations in small LDL particles. Moreover, independently of genotype, diet had an effect on plasma HDL and HDL2.
View this table:
TABLE 3. . Metabolic variables at the end of each dietary period according to the peroxisome proliferator-activated (PPAR) Leu162Val polymorphism1
Plasma lipoprotein and lipid changes in response to the high-P:S diet were compared between carriers of the V162 allele and L162 homozygotes (Table 4). After the high-P:S diet, there was a significant difference between the genotype groups for apo A-1 concentrations. These results were unaffected by low-P:S diet measurements. There were also trends for plasma total cholesterol, apo B, and cholesterol concentrations in small LDL particles. Carriers and noncarriers of the PPAR L162V polymorphism showed opposite changes. Indeed, carriers of the V162 allele had lower apo A-1 concentrations after the high-P:S diet compared with slight but nonsignificant increases in these respective values among L162 homozygotes. Changes in BMI, waist circumference, fasting insulin and glucose, LDL cholesterol, HDL cholesterol, triacylglycerol, phospholipid subfractions, and apo B concentrations were not significantly different between groups.
View this table:
TABLE 4. . Changes in the metabolic profile after the high-P:S diet according to the peroxisome proliferator-activated receptor (PPAR) Leu162Val polymorphism1
DISCUSSION
The role of dietary fatty acids in regulating plasma lipoprotein and lipid concentrations is well documented (34). Compared with SFA intake, PUFA intake has been shown to lower LDL-cholesterol and HDL-cholesterol concentrations (3). However, the great interindividual lipoprotein and lipid response to such dietary changes is not well understood. Genetic factors are likely to play a significant role, because polymorphisms in several genes seem to modulate lipoprotein and lipid responses to dietary modifications (6). To the best of our knowledge, this study is the first to examine whether plasma lipoprotein and lipid responsiveness to a change in the P:S ratio is influenced by the PPAR L162V gene polymorphism in healthy men.
In vitro studies have shown that the V162 allele has a greater transactivation activity when treated with a PPAR agonist (16, 35). Previous studies showed the increasing effect of the L162V polymorphism on total cholesterol (16, 18, 19), LDL cholesterol (17, 19), HDL cholesterol (16), apo B (17-19), apo A-1(16), and apo C-III (19). This polymorphism was also found to be associated with decreased concentrations of fasting serum triacylglycerol among white subjects with normal glucose tolerance (36). Moreover, the V162 allele was associated with a decrease in BMI and percentage body fat in healthy adults (37). In contrast with these studies, our results showed no significant difference in screening anthropometric indexes and lipoprotein and lipid measurements between the L162V genotype groups. This discrepancy can be explained by the small sample size but more importantly by the study design, which included healthy men matched for age and BMI.
Two studies had tested the effect of the PPAR L162V polymorphism on changes in plasma lipid concentrations in fibrate-treated participants (16, 38). Only one of those studies reported a greater lowering of total cholesterol in carriers of the V162 allele when the subjects were treated with bezafibrate, a PPAR agonist (16). In the present study, carriers of the V162 allele had a cholesterol impoverishment of VLDL, LDL, HDL, and HDL3 after the high-P:S diet, which could explain the trend toward a greater decrease in total cholesterol concentrations in carriers of the V162 allele. However, it is important to mention that within each genotype group, the switch from a low to a high-P:S diet was not associated with a significant increase or decrease in total plasma cholesterol concentrations. The changes observed in apo A-1 concentrations were significantly different between the 2 genotype groups. Differences observed in apo A-1 responsiveness in the present study are divergent from the literature. Indeed, it is well documented that an increase of the P:S ratio decreases apo A-1 concentrations (39-42). In addition, in vitro experiments have shown that when cells are treated with a PPAR agonist, the V162 allele, compared with the L162 allele, shows enhanced transactivation activity (16, 35) that theoretically leads to increases in apo A-1 gene transcription (13). Considering that linoleic acid binds to PPAR like a fibric acid derivative, we expected a greater increase in apo A-1 in carriers of the V162 allele. At this time, we are unable to explain the result observed in healthy men.
Our investigation of the gene-by-diet interaction effects on the metabolic profile led us to several interesting observations. We found that diet interacts with the L162V polymorphism to modulate total cholesterol concentrations, plasma apo A-1 concentrations, and cholesterol in small LDL particles. These gene-by-diet interaction effects may then modulate plasma lipid concentrations and need further replication in different intervention studies and later in a meta-analysis. In addition, independently of genotype, diet had an effect on HDL and HDL2.
The APOE genotype had no effect on the metabolic response to a switch from a low- to a high-P:S diet. Thus, the APOE genotype effects were not similar to those previously observed in nutritional studies (43, 44). Two reasons can explain this discrepancy with published literature on APOE. First, the present study included a small number of healthy subjects, and, second, a selection bias resulting from recruitment on the basis of PPAR L162V genotype may have occurred.
Because the subjects in the present study were free-living and ate self-selected foods, it was difficult to control their diet and to replace only SFAs by PUFAs without affecting other nutrients. The consequence is that the energy content of the diet was 300 kcal lower during the low- and high-P:S diets than during the screening period. However, because this difference was not statistically significant and weight loss was similar in both genotype groups, we did not adjust for changes in energy intake in the statistical analyses. MUFA intake was higher during the high-P:S diet than during the low-P:S diet because sunflower oil and margarine contain more of these fatty acids than does butter.
Fatty acids and fatty acid–derived compounds are natural ligands for PPAR (45). Fatty acids containing a 16–20-carbon chain length with several double bonds bind optimally with PPAR (20). Thus, long-chain unsaturated fatty acids, such as linoleic acid, bind to PPAR with reasonable affinity (45). Despite the higher affinity of eicosanoid metabolites of the linoleic acid in vitro, it is unclear whether the concentration of this compound is sufficient to cause activation of PPAR in vivo (46). In the present study, subjects consumed 1 or 2 tablespoons of sunflower oil and a certain quantity of nonhydrogenated (free of trans fat) margarine during the high-P:S diet, which corresponds with a total linoleic acid intake of 19 g/d and a PUFA intake representing 10.5% of total energy. High daily intake of n–6 PUFAs is not currently recommended because these fatty acids can decrease HDL cholesterol and could be more easily oxidized by free radicals (47). Lipid peroxides were reported to lead to atherosclerosis and coronary heart disease (48, 49). Thus, an increase in the intake of PUFAs might have been sufficient to reach the intracellular concentrations required to activate PPAR, but might also have a deleterious effect on the subject's metabolic lipid profile. Other studies are necessary to determine the adequate dietary intake of PUFAs required to activate PPAR.
In conclusion, the present study provides evidence that the PPAR L162V polymorphism can contribute to interindividual changes in total cholesterol and apo A-1 observed in response to modification of the dietary P:S ratio. Allelic variability in the PPAR gene could partially explain differences in individual responses to diet. Our findings also underline the importance of taking into account the quality of fat in the diet as well as gene-by-diet interaction effects when dissecting the genetic architecture of complex traits such as plasma lipoprotein and lipid concentrations.
ACKNOWLEDGMENTS
We express our gratitude to the subjects for their excellent collaboration. We thank the staff of the CHUL Lipid Clinic and Amélie Charest and Chantal Bélanger for their technical assistance. We also thank Danielle Aubin for nursing assistance and Alain Houde for contributing to the laboratory work.
The contributions of the authors were as follow: A-MP reviewed the literature, recruited the subjects, coordinated the study, collected data, performed the statistical analyses, interpreted the data, and wrote the manuscript; BF-B assisted with the development of the research study protocol; YB and JR assisted with the data analysis and interpretation; SL, HJ, BL, AT, and PC were involved in the design of the study and laboratory measurements; M-CV was the principal investigator and designed the study, supervised the research, directed the data analysis and interpretation, and assisted with the manuscript preparation. All of the authors revised the manuscript. None of the authors had a personal or professional conflict of interest.
REFERENCES