点击显示 收起
1 From the Departments of Physiology, Clinical Nutrition, and Medicine, University of Kuopio, Kuopio, Finland.
2 Address reprint requests to JJ Ågren, Department of Physiology, University of Kuopio, PO Box 1627, 70211 Kuopio, Finland. E-mail: jyrki.agren{at}uku.fi.
ABSTRACT
Background: The affinity of intestinal fatty acid binding protein (FABP) for fatty acids is regulated by the polymorphism at codon 54 of the FABP2 gene (alanine-to-threonine shift). We found earlier that the threonine-encoding allele (Thr54) is associated with an increased postprandial lipemic response.
Objective: We studied the postprandial responses of individual fatty acids in subjects homozygous for the Thr54 or alanine-encoding allele (Ala54).
Design: Oral-fat-loading tests were performed in 8 subjects homozygous for Thr54 and in 7 subjects homozygous for Ala54.
Results: The postprandial responses of most of the 14–18-carbon fatty acids in chylomicron and VLDL triacylglycerols were significantly elevated in the Thr54 homozygotes whereas the relative increases in these fatty acids were not significantly different in both groups. The amounts of 20- and 22-carbon polyunsaturated fatty acids started to increase later than the amounts of shorter ones after the test meal, and the differences between the groups were mostly insignificant. The responses of chylomicron fatty acids correlated positively with postprandial insulin response in the Thr54 homozygotes and inversely in the Ala54 homozygotes. VLDL fatty acid responses correlated with fasting triacylglycerol concentrations in the Ala54 homozygotes but not in the Thr54 homozygotes.
Conclusion: The threonine-encoding allele of the FABP2 gene is associated with an increased postprandial response of 14–18-carbon fatty acids but not with changes in the relative amounts of individual fatty acids introduced to chylomicron triacylglycerols.
Key Words: Postprandial lipemia • triacylglycerol • insulin • fish oil • FABP2 gene • intestinal fatty acid–binding protein • polymorphism
INTRODUCTION
Several different fatty acid binding proteins (FABPs) have been identified in various tissues (1). Intestinal FABP (IFABP), encoded by the FABP2 gene, is expressed only in intestinal enterocytes (2). A polymorphism at codon 54 of the FABP2 gene, changing alanine to threonine, has been reported to increase the affinity of IFABP for long-chain fatty acids (3) and to increase triacylglycerol secretion by Caco-2 cells (4). The threonine-encoding allele (Thr54) is also related to impaired insulin action and increased fat oxidation (3, 5). Therefore, the threonine-containing protein may alter the processing of dietary fatty acids in a way that increases fat oxidation and decreases glucose oxidation, thus contributing to insulin resistance (3).
The role of IFABP in fat absorption is largely unknown, as is how the threonine-containing protein could modify this process. One possibility is that the greater affinity of the threonine-containing protein for long-chain fatty acids could cause increased postprandial lipemia. To test this hypothesis, we previously performed an oral-fat-loading test in Thr54 and Ala54 (alanine-encoding allele) homozygous subjects (6). Indeed, greater postprandial triacylglycerolemia was found in the Thr54 homozygotes. Furthermore, triacylglycerolemia was associated with the postprandial insulin response. The affinity of rat IFABP was shown to be much higher for palmitic than for arachidonic acid (7), and an increased affinity of threonine-containing protein may also change its preference for different fatty acids. This, in turn, could affect the metabolic fate of different fatty acids. To investigate the postprandial differences in fatty acid metabolism between the Ala54 and Thr54 homozygotes in more detail, the responses of individual fatty acids after an oral fat load were determined in the present study.
SUBJECTS AND METHODS
Subjects
Ten subjects homozygous for the Thr54 allele of the FABP2 gene were identified from 300 subjects screened in earlier studies (8, 9). They and 12 subjects homozygous for the Ala54 allele, selected on the basis of sex, age, body mass index, and triacylglycerol concentrations measured in the previous studies (8, 9), participated in the present study. Five subjects with high fasting plasma triacylglycerol concentrations (>2.0 mmol/L), 1 who had the apolipoprotein E phenotype E2/3, and 1 who did not follow study instructions were excluded from the study. The final study group consisted of 8 Thr54 homozygotes and 7 Ala54 homozygotes. The Ethics Committee of the University of Kuopio approved the study and all subjects gave written consent.
Oral-fat-loading test
The oral-fat-loading test started at 0730–0830 after a 12-h fast. Subjects were advised to not drink alcohol and to avoid strenuous exercise for 3 d before the test. After fasting blood samples were collected, subjects ate a small rice cake (7.6 g) with a piece of cheese (10 g, 2.3 g fat/70 kg body wt). Five minutes later, they drank a cream mixture (100 mL, 27.4 g fat/70 kg body wt) that included heptadecanoic acid (17:0; 4 g/70 kg body wt) and that otherwise had a fatty acid profile resembling that of the average Finnish diet; immediately after that they drank fish oil (20 mL, 18.6 g fat/70 kg body wt). The fatty acid content of the test meal is presented in Table 1. The total amount of fat given was 52 g/70 kg body wt. Blood samples were collected 0.5, 1, 2, 3, 4, 6, and 8 h after the test meal.
View this table:
TABLE 1.. Fatty acid content of the fat test meal
Laboratory measurements
To separate chylomicrons, plasma (1.8 mL) was overlaid with 1.6 mL sodium chloride solution (density: 1006 g/L) and ultracentrifuged with a TFT 45.6 rotor (Kontron Instruments, Milan, Italy) at 38000 x g (18000 rpm) and 16°C for 30 min. The top 1 mL was aspirated to remove the chylomicron fraction. The infranate was overlaid again with sodium chloride solution, and samples were ultracentrifuged to separate VLDL [160000 x g (37000 rpm) at 16°C for 16 h]. Lipoproteins were separated from fasting serum samples by ultracentrifugation (density: 1006 g/L) to remove VLDL and by precipitation of LDL.
Cholesterol and triacylglycerol concentrations in plasma or serum and in separated lipoprotein fractions were determined by enzymatic colorimetric methods using commercial kits (Boehringer Mannheim, Mannheim, Germany). Serum fatty acids were determined with a turbidometric method by using an automated instrument (Specific Clinical Analyser; Kone Ltd, Espoo, Finland). Plasma insulin was measured with a radioimmunoassay (Phadeseph Insulin RIA 100; Pharmacia Diagnostics, Uppsala, Sweden). Plasma glucose was analyzed by a glucose oxidase method (Glucose Auto&Stat, model GA-110; Daiichi, Kyoto, Japan).
Chylomicron and VLDL lipids were extracted with chloroform-methanol (2:1, by vol) (10). Triacylglycerols were separated by solid-phase extraction with an aminopropyl column (11) and transmethylated with 14% boron trifluoride in methanol. The fatty acid methyl esters were analyzed with a gas chromatograph (HP 5890 Series II; Hewlett-Packard Company, Waldbronn, Germany) equipped with an HP-FFAP capillary column (25 m, 0.20-mm internal diameter, 0.33-µm film thickness; Hewlett-Packard, Palo Alto, CA). Tripentadecanoin was used as an internal standard (Sigma Chemical Co, St Louis).
Statistics
The Mann-Whitney U test was used for the comparisons of groups and Spearman's correlation coefficient was used for correlation analyses (STATVIEW; Abacus Concepts, Inc, Berkeley, CA). Areas under the fatty acid and insulin response curves (above baseline) were calculated by the trapezoidal rule using a computer program (CANVAS; Deneba Systems, Inc, Miami).
RESULTS
There were no significant differences between the Ala54 and Thr54 homozygotes in age, body mass index, fasting plasma glucose and insulin, or serum lipid concentrations (Table 2). Nutrient intakes also did not differ significantly between the groups on the basis of 3-d food records (data not shown).
View this table:
TABLE 2.. Subject characteristics1
The postprandial increases in most fatty acids with a 14–18-carbon chain were greater in the Thr54 homozygotes than in the Ala54 homozygotes in both chylomicrons and VLDL (Table 3, Figure 1, and Figure 2). Peak concentrations were reached 4 h after the meal in both fractions in the Thr54 homozygotes, at 3 h in chylomicrons in the Ala54 homozygotes, and at 6 h in VLDL in the Ala54 homozygotes. The proportions of 14:0 and 16:0 increased and those of 18:1n-9, 18:2n-6, and 18:3n-3 decreased in chylomicrons from 2 to 8 h after the meal in both groups, except for a slight increase of 18:2n-6 in the Thr54 homozygotes from 6 to 8 h after the meal. In both groups, an early peak after 30 min was seen in most VLDL fatty acids.
View this table:
TABLE 3.. Area under the response curves of chylomicron and VLDL triacylglycerol fatty acids after a fat test meal1
FIGURE 1. . Mean (±SEM) responses of major saturated and monounsaturated chylomicron and VLDL triacylglycerol fatty acids after an oral fat load in persons homozygous for the alanine-encoding allele in codon 54 of intestinal fatty acid binding protein 2 (Ala54; ; n = 7) and in persons homozygous for the threonine-encoding allele (•; n = 8). Chylomicron and VLDL responses are on the same scale for each individual fatty acid. *Significantly different from Ala54 homozygotes, P < 0.05.
FIGURE 2. . Mean (±SEM) responses of major polyunsaturated chylomicron and VLDL triacylglycerol fatty acids after an oral fat load in persons homozygous for the alanine-encoding allele in codon 54 of intestinal fatty acid binding protein 2 (Ala54; ; n = 7) and in persons homozygous for the threonine-encoding allele (•; n = 8). *Significantly different from Ala54 homozygotes, P < 0.05.
The amounts of 20- and 22-carbon fatty acids started to increase later than those of shorter-chain fatty acids and were still high at the end of the study (Figure 2). In VLDL, the highest concentrations of 20- and 22-carbon fatty acids were seen 8 h after the meal whereas the amounts of most other fatty acids had returned to their fasting concentration or even below by this time. Thus, the calculated area under the response curve represents only a part of the total area (Table 3). However, the present data indicate clearly that the responses of 20- and 22-carbon fatty acids in the Ala54 and Thr54 homozygotes were quite similar, in contrast with the responses seen for the shorter-chain fatty acids. The ratio of 20:5n-3 to 22:6n-3 increased in chylomicrons during the follow-up period. The ratio was similar to that in the test meal (1.7) after 4 h (1.6 ± 0.5 and 1.4 ± 0.6 in the Ala54 and Thr54 homozygotes, respectively); the ratio increased to 2.7 ± 0.7 in both groups after 8 h.
The response of 17:0, given as an unesterified fatty acid, was small compared with the responses of fatty acids obtained in triacylglycerol form. In addition, it decreased in VLDL for the first 2 h after the meal and then started to increase.
The responses of major VLDL fatty acids correlated with fasting triacylglycerol concentrations in the Ala54 homozygotes but not in the Thr54 homozygotes (Table 4). The responses of chylomicron fatty acids showed an inverse correlation with insulin response in the Ala54 homozygotes whereas there was a positive correlation in the Thr54 homozygotes.
View this table:
TABLE 4.. Correlation of fasting triacylglycerol with VLDL fatty acid response and insulin response with chylomicron fatty acid response after the fat test meal1
DISCUSSION
Our study showed that the postprandial responses of most of the 14–18-carbon fatty acids in chylomicron and VLDL triacylglycerols were greater in the Thr54 homozygotes than in the Ala54 homozygotes, which is in accordance with our earlier finding of greater postprandial triacylglycerolemia in Thr54 homozygotes (6). However, no clear differences in the relative amounts of individual fatty acids were observed between the groups.
The responses of 14–18-carbon fatty acids in chylomicron triacylglycerols followed amounts in the test meal in both groups. This indicates that these fatty acids were absorbed and introduced to chylomicron triacylglycerols with approximately similar efficiency, which was not affected by the alanine-to-threonine substitution in IFABP. The relative amounts of 18:1n-9 and 18:2n-6 decreased slightly whereas those of 14:0 and 16:0 increased during the postprandial period in both groups. In addition to possible differences in the intestinal phase, this could have been caused by the preferential hydrolysis of triacylglycerols containing 18:1n-9 and 18:2n-6 by lipoprotein lipase, which would be consistent with the reported specificity of this enzyme (12). Another possibility could be the greater contribution of endogenous fatty acids to chylomicron formation in the early stage of lipid absorption.
As in chylomicrons, in VLDL triacylglycerols there was a quantitative difference between the Ala54 and Thr54 homozygotes but no difference in the relative increases in 14–18-carbon fatty acids. Increases in these fatty acids resembled their ratio in fasting VLDL, indicating reduced hydrolysis or increased secretion of VLDL and minor contributions of chylomicron remnants to this fraction. This agrees with the 10-fold higher plasma concentration of apolipoprotein B-100 than apolipoprotein B-48 in the Sf 60–400 fraction at all time points after an oral fat load reported by Björkegren et al (13).
The concentrations of 20- and 22-carbon fatty acids, mainly derived from fish oil, started to increase later and remained high longer than did those of shorter-chain fatty acids. In addition, their responses were similar in the Ala54 and Thr54 homozygotes. This indicates that the difference in the properties of IFABP (Ala54 compared with Thr54 homozygosity) does not affect the handling of these longer-chain fatty acids in the intestine. Some studies reported lower acute postprandial responses with fish oil than with saturated fat (14, 15), and it was suggested that n-3 fatty acids promote more efficient hydrolysis of triacylglycerol-rich lipoproteins (15). In contrast, our results showed that the 20- and 22-carbon fatty acids were absorbed, secreted, and cleared more slowly than were the shorter-chain fatty acids, resulting in a greater postprandial response than with the shorter-chain fatty acids. The most probable explanation for this finding is the resistance of these fatty acids to hydrolysis by both pancreatic lipase in intestine and endothelial lipoprotein lipase. The slower appearance of these longer-chain fatty acids in chylomicrons could affect their clearance even without the specific effect of lipoprotein lipase.
The responses of n-3 fatty acids differed from each other. The ratio of 20:5n-3 to 22:6n-3 in chylomicron triacylglycerols increased during the postprandial period in accordance with the in vitro studies of Ekström et al (16), indicating preferential hydrolysis of 22:6n-3 over 20:5n-3 by lipoprotein lipase. Faster liberation of 22:6n-3 from chylomicrons could also explain the greater response of 22:6n-3 seen in VLDL triacylglycerols. This difference in the distribution of 20:5n-3 and 22:6n-3 indicates that their increased amount in the VLDL fraction was mostly a consequence of their reesterification in the liver. The possibility of their direct increase via intestinally derived VLDL cannot be ruled out (17).
In this study, we tried to use 17:0 as a marker fatty acid because it is usually present in very small amounts in plasma lipids. Compared with its intake, its response was, however, clearly lower than that of other fatty acids in both chylomicrons and VLDL. The most plausible explanation is that 17:0 was given as an unesterified fatty acid; it may have been poorly absorbed or transported partly through the portal circulation without entering chylomicrons.
The greater responses of individual chylomicron fatty acids of the Thr54 homozygotes could be caused directly by the changed properties of IFABP. On the other hand, the threonine-encoding allele has been associated with insulin resistance, which in turn could increase postprandial lipemia through the weaker activation of lipoprotein lipase (18). We found a positive correlation in the Thr54 homozygotes and an inverse correlation in the Ala54 homozygotes between the fatty acid and insulin responses in chylomicrons. This suggests that a difference in insulin secretion or action between the groups may have contributed to the observed difference in lipemic response.
The responses of individual chylomicron fatty acids correlated with fasting triacylglycerols in the Ala54 homozygotes. This agrees with the results of several earlier studies that showed a correlation between fasting and postprandial triacylglycerol concentrations (19). This was explained by the competition between VLDL and chylomicron triacylglycerols for lipoprotein lipase (20). It was shown, however, that there are also high responders with normal fasting triacylglycerol but high insulin concentrations and hypertriacylglycerolemic subjects with low response (21). Although it is obvious that the concentrations of different triacylglycerol-rich lipoproteins are affected by their competition for hydrolyzing enzymes, it is possible that the activation of these enzymes by insulin plays a major role in the determination of postprandial, and possibly also fasting, triacylglycerol concentrations.
In summary, our study showed that the homozygous Thr54 genotype of the FABP2 gene is associated with an increase in postprandial response of 14–18-carbon fatty acids in chylomicrons and VLDL but does not alter the relative amounts of these fatty acids. The homozygous Thr54 genotype does not affect the response of 20- and 22-carbon polyunsaturated fatty acids, which appear later in chylomicrons than do the shorter-chain fatty acids. Further studies are needed to clarify the primary change in the handling of dietary fatty acids caused by the codon 54 polymorphism of the FABP2 gene.
REFERENCES