Main

Alterations of the lipoprotein-lipid profile are associated with an increased risk of developing coronary heart disease, the leading cause of death in westernized societies. Many lipid-lowering agents have been developed for the treatment of dyslipidemia. In this regard, the clinical benefits of gemfibrozil, a fibrate agent, have been reported in both primary1 and secondary2 prevention trials. Although fibrates have been used in clinical practice for more then three decades, their molecular mechanism of action has just recently been elucidated3, and it is now recognized that their effects are mediated by the activation of a specific nuclear receptor termed peroxisome proliferator-activated receptor alpha (PPARα). The activation of PPARα by fibrates causes the transactivation of PPARα-responsive genes, which include those encoding proteins that control lipid metabolism.3

Recently, a molecular scanning of the human PPARα has revealed a L162V polymorphism associated with alterations of the lipoprotein-lipid profile.4,5 The frequency for the rare allele is established to be 0.062 in the healthy European population5 and reaches 0.128 in French-Canadians.4 To demonstrate the functional consequences of this polymorphism, transient transfection assays in Hepa-1 and HepG2 cell lines have been performed.5,6 Both studies have shown enhanced transactivation activity in cells containing vectors expressing PPARα-V162 allele, compared with L162 allele, when treated with the PPARα ligand WY-14,643. In this context, it was interesting to verify whether the L162V polymorphism in the PPARα gene can modulate the plasma lipoprotein/lipid response to gemfibrozil.

MATERIALS AND METHODS

Subjects

Subjects were asymptomatic, nondiabetic volunteers who had to fulfill the following criteria. Men had to be between 25 and 55 years of age and willing to participate to a 6-month intervention program in which they were asked to follow the National Cholesterol Education Program Phase 1 dietary guidelines with or without gemfibrozil. Experience from the Helsinki Heart Study revealed that the effect of gemfibrozil was largely confined to overweight subjects.7 Men in the present study were weight stable obese (27 and 40 kg/m2) and characterized by a dyslipidemic state (1.7 mmol/L ≥ TG ≤ 5.7 mmol/L; high-density lipoprotein cholesterol (HDL-C) ≤ 1.2 mmol/L and total plasma cholesterol < 6.7 mmol/L). The study was approved by the Medical Ethics Committee of Laval University. All subjects gave their informed written consent to participate in this study.

Study design

After having completed their baseline measurements, 71 subjects were selected to participate to the study and were randomly assigned to either receiving a placebo or gemfibrozil 600 mg bid. Dietary recommendations were given by a registered dietician on a voluntary basis for the duration (6 months) of the study. Drug safety was assessed every 4 weeks by the physician in charge of the project. Subjects were tested at baseline and at the end of the 6-month intervention protocol. During that period, there were six dropouts, and genotype information was not available for two subjects. We thus ended-up with a total of 63 subjects who completed the trial.

Lipids and lipoproteins

Fasting blood samples were collected and plasma lipid, lipoprotein, and apolipoprotein (apo) levels were measured as previously described.8 HDL2 was precipitated from the HDL fraction with dextran sulfate.9 The cholesterol content of the supernatant fraction (HDL3) was determined, and HDL2-C levels were derived by subtracting HDL3-C from total HDL-C concentrations. Nondenaturing 4% to 30% polyacrylamide gel electrophoresis was performed for the measurement of the average HDL size using whole plasma kept at −80°C as recently described.10 The same plasma were also used to determine the low-density lipoprotein (LDL) peak particle size measured by electrophoresis on a 2% to 16% polyacrylamide gradient gel as previously described.11,12

DNA analysis

The L162V polymorphism does not alter any restriction site. A mismatch polymerase chain reaction method previously described was used to genotype all the subjects participating in the study.4

Statistical analyses

Variables not normally distributed were log10 transformed before analysis. The difference in response between genotype groups was assessed by a two-tailed unpaired Student's t-test. To evaluate whether the PPARα-L162V genotype may interact with gemfibrozil treatment, we performed an ANOVA in which the interaction term was included (two-way factorial ANOVA). The source of variation in lipoprotein-lipid profile was computed using the type III sum of squares. This sum of squares applies to unbalanced study designs and quantifies the effects of an independent variable after adjustment for all other variables included in the model. All statistical analyses were performed using the SAS package (SAS Institute, Cary, NC), and a statistically significant difference was defined as P < 0.05.

RESULTS

Among the 63 men who completed the study, 31 were assigned to the placebo group and 32 to the gemfibrozil group. The effects of the 6-month-intervention program on plasma lipoprotein/lipid profile have been published elsewhere.8 In the present study, the placebo and gemfibrozil groups were further subdivided on the basis of the PPARα-L162V genotype (Fig. 1).

Fig 1
figure 1

Distribution of subjects into the intervention program. Men were randomly assigned to either receiving a placebo or gemfibrozil and then divided according to PPARα-L162V genotype.

Full size image

The baseline lipoprotein-lipid profile according to the PPARα-L162V genotype is presented in Table 1. For these analyses, subjects in the placebo and gemfibrozil groups were combined because medication had not started. No statistically significant difference was observed between the two genotype groups for baseline lipid and lipoprotein concentrations. However, there was a tendency toward higher LDL-C levels among carriers of the V162 allele compared with L162-homozygotes (HMZ) (P = 0.08).

Table 1 Baseline lipoprotein-lipid concentrations between carriers and noncarriers of the PPARα-L162V polymorphism
Full size table

The plasma lipoprotein-lipid changes of the 6-month gemfibrozil therapy was compared between carriers and noncarriers of the PPARα-L162V polymorphism (Table 2). There was no statistically significant difference between the two genotype groups for all lipoprotein-lipid changes except for HDL2-C levels. Indeed, V162 carriers exhibited a 50% increase in HDL2-C concentrations compared with a 5.5% increase among L162-HMZ (Fig. 2). The same trend was observed with changes in HDL-C levels although it did not reach statistical significance (P = 0.08). These results were the same after adjustment for either baseline measurements or changes observed in body weight (data not shown). To corroborate the difference observed between the two genotype groups on HDL2-C changes with gemfibrozil, we analyzed the average HDL size change with the therapy. Although not statistically significant, carriers of the V162 allele had a more favorable HDL size response compared with L162-HMZ (Table 2). There was also a tendency toward a greater decrease in LDL-apo B for those carrying the V162 allele (P = 0.09). This trend might be explained by a greater reduction in LDL-C and a greater increase in LDL size observed in V162 allele carriers although the difference did not reach statistical significance.

Table 2 Changes in plasma lipid and lipoprotein levels in gemfibrozil-treated participants according to PPARα-L162V genotypes
Full size table
Fig 2
figure 2

Changes in HDL-C and HDL-C subfractions in gemfibrozil-treated participants according to PPARα-L162V genotype (N = 26 L162-HMZ and 6 V162 carriers). Each dot represents the change observed for one individual within his respective group. Relative change is indicated at the bottom of the graph. The bar chart illustrates mean HDL changes in the different groups.

Full size image

When individual HDL-C responses to gemfibrozil were plotted (Fig. 3), two phenomena were noted. First, there was a large interindividual variation in HDL-C changes with gemfibrozil therapy. Second, there was a cluster of carriers of the V162-allele on the right side of the graph, indicating a greater increase of HDL-C levels among these individuals. In these individuals, the apparent difference in plasma HDL-C levels was mainly explained by changes in the HDL2-C subfraction as no difference was observed between the genotype groups regarding the change in the HDL3-C subfraction (Fig. 2). Subsequent analyses performed in men treated with gemfibrozil revealed that after the 6-month-intervention program, V162 carriers were characterized by higher levels of HDL2-C compared with L162-HMZ (Fig. 4).

Fig 3
figure 3

Individual changes in HDL-C levels among gemfibrozil-treated subjects (N = 32). Each bar represents the after-before difference in HDL-cholesterol levels observed in response to gemfibrozil therapy. Individual responses of L162-homozygotes are illustrated by the gray bars, whereas the white bars show the responses of carriers of the V162 allele.

Full size image
Fig 4
figure 4

Mean HDL2-C values before and after gemfibrozil treatment according to PPARα-L162V genotype (N = 26 L162-HMZ and 6 V162 carriers). Values are mean ± SE.

Full size image

To test the potential interaction between the PPARα-L162V polymorphism and gemfibrozil treatment on plasma lipoprotein-lipid concentrations, an analysis of variance was performed for each lipoprotein-lipid variable. The effects of the genotype and the treatment, as well as the potential interaction between these two independent variables, are presented in Table 3. As expected, the treatment (placebo versus gemfibrozil) had a statistically significant impact on plasma lipoprotein-lipid changes. On the other hand, the PPARα-L162V genotype by itself did not have a significant impact on plasma lipoprotein-lipid responses. However, a significant genotype-by-treatment interaction was observed for changes in plasma HDL2-C levels. This interaction explained 7.0% of the total variance of the change in plasma HDL2-C concentrations. This finding suggests that the PPARα-L162V polymorphism may influence plasma HDL2-C responsiveness to gemfibrozil therapy. Additionally, it also implies that men carrying the PPARα-V162 allele will experience a higher increase in HDL2-C levels when treated with fibrates.

Table 3 Effects of PPARα-L162V polymorphism, the treatment (placebo vs. gemfibrozil) and their interaction on plasma lipid-lipoprotein response to the intervention program
Full size table

DISCUSSION

The documented interindividual variation in the response to fibrate therapy is clinically relevant. In this report, we demonstrated that a naturally occurring variation, L162V, in the PPARα gene is associated with greater responsiveness of HDL2-C levels after a 6-month-therapy with gemfibrozil. Before this study, Flavell et al.5 had reported a greater lowering effect of bezafibrate on total cholesterol and non-HDL-C in V162 carriers. In the present study, there was a trend toward higher increase in HDL-C levels after the administration of gemfibrozil, but it did not reach statistical significance. Taken together, these results suggest potentially greater benefits of fibrate treatment among individuals carrying the PPARα-V162 allele. This could be of great interest as previous work suggests that the PPARα-L162V polymorphism seems to have a deleterious impact on plasma lipoprotein-lipid levels.4,5,13

PPARα regulates the expression of gene encoding proteins that control lipoprotein metabolism. In vitro studies performed on hepatic cells (HepG2 and Hepa-1) have demonstrated higher transactivation activity in PPARα carrying the V162 allele when treated with a PPARα agonist.5,6 Differences observed in HDL2-C responsiveness in the present study could then be explained by a greater transcriptional regulation of PPARα target genes in subjects carrying the V162 allele. Several proteins controlling HDL metabolism are regulated at the gene level by PPARα activators such as fibrates. In fact, fibrate therapy has been shown to induce overexpression of apo A-I and apo A-II genes, leading to an increase in plasma HDL-C levels.14,15 Recently, other proteins such as ATP-binding cassette transporter 1 and CD36- and LIMPII-analogous 1/scavenger receptor class B type 1 involved in the reverse-cholesterol-transport pathway have been shown to be up-regulated by PPARα activators in macrophages.16,17 Furthermore, it has been demonstrated in mice that fibrate treatment increases phospholipid transfer protein gene expression through a PPARα-dependent mechanism, which accounts for a marked enlargement of HDL particles.18 The functional consequence of the PPARα-L162V polymorphism could result in higher transcriptional regulation of genes controlling HDL metabolism and then explain the relationship between the PPARα-L162V polymorphism and the HDL2-C response to gemfibrozil observed in the present study.

This study reported associations that make biological sense and study genetic variations that affect the gene product in a physiologically meaningful way. However, certain limitations may give some uncertainty about our results. First, dealing with the low frequency of the polymorphism and with a small sample size impaired our ability to recognize association with a smaller effect. Second, considering the multiple comparisons made in the study, some of the significant findings may have occurred by chance. Consequently, further studies with a larger sample size are warranted to determine the interest of the PPARα L162V polymorphism in the management of dyslipidemia.

In summary, we have demonstrated that subjects carrying the PPARα-V162 allele showed higher increases in HDL2-C levels in response to gemfibrozil therapy. This finding could be of clinical relevance for these individuals as a low concentration of the HDL2 subfraction seems to be more closely related to the incidence of ischemic heart disease.19,20 We speculate that the greater HDL2-C increase among gemfibrozil-treated men carrying the PPARα-L162V polymorphism is mediated by a more pronounced transcriptional regulation of genes controlling HDL metabolism.