|
| |||||||||||||||||||||||||||||||
Background Previous studies have suggested that the variant apolipoprotein (apo) allele apo A-IV-2 may influence the response of the plasma cholesterol concentration to dietary cholesterol.
Methods We measured plasma lipids and lipoproteins in 11 subjects who were heterozygous for the apo A-IV-2 allele (apo A-IV-1/2 heterozygotes) and a control group of 12 subjects who were homozygous for the common apo A-IV allele (apo A-IV-1/1 homozygotes) in an outpatient dietary-modification study. (Approximately one in seven persons in the United States is a heterozygote.) The subjects consumed a low-cholesterol diet (about 200 mg [0.5 mmol] of cholesterol per day) during a two-week run-in period; daily cholesterol intake was then increased to approximately 1100 mg (2.8 mmol) by the addition of four egg yolks per day.
Results The fat intake and the ratio of polyunsaturated to saturated fat were similar in the two groups throughout the study. After three weeks of egg intake, the mean plasma total cholesterol increased by 22 mg per deciliter (0.57 mmol per liter) in the apo A-IV-1/1 group, but by only 6 mg per deciliter (0.15 mmol per liter) in the apo A-IV-1/2 group (P = 0.05). The mean plasma low-density lipoprotein cholesterol increased by 19 mg per deciliter (0.49 mmol per liter) in the apo A-IV-1/1 group, but by only 1 mg per deciliter (0.03 mmol per liter) in the apo A-IV-1/2 group (P = 0.03). There were no changes in the plasma triglyceride or high-density lipoprotein cholesterol concentrations in either group.
Conclusions The apo A-IV-2 allele attenuates the hypercholesterolemic response to the short-term ingestion of a very-high-cholesterol diet and may partially account for the heterogeneous response to dietary cholesterol. However, cholesterol intake in this study was more than twice that of the general population; whether the apo A-IV-2 allele alters responsiveness at lower levels of cholesterol intake remains to be determined.
The plasma apolipoproteins are a family of lipid-binding proteins that regulate intravascular lipoprotein metabolism5. Genetic variants of human apolipoproteins have been identified,6 and it has become evident that apolipoprotein polymorphisms may profoundly affect many aspects of human lipoprotein metabolism, including the response to diet7. Human apolipoprotein A-IV (apo A-IV) is a 46-kd plasma apolipoprotein8 synthesized by the small intestine during fat absorption9; it circulates primarily unassociated with plasma lipoproteins9,10. Although the specific function of apo A-IV in human lipoprotein metabolism has not been established, several lines of evidence suggest that it may play a part in the assembly and intravascular metabolism of high-density lipoproteins (HDL)11,12 and the process of reverse cholesterol transport13.
A polymorphism in the apo A-IV gene codes for the substitution of histidine for glutamine at position 360 near the carboxyl terminus,14 which generates an isoform, apo A-IV-2. This isoform is one charge unit more basic than the common isoform, apo A-IV-1. Population screening has established that the apo A-IV-2 allele exists worldwide and has a gene frequency of 0.07 to 0.09 in Western countries15,16,17,18,19,20; approximately one in seven persons in the United States is heterozygous for the apo A-IV-2 allele (apo A-IV-1/2 heterozygotes). In some studies, apo A-IV-1/2 heterozygotes had higher plasma HDL cholesterol15,17 and lower triglyceride15,16,17 levels than those who were homozygous for the apo A-IV-1 allele (apo A-IV-1/1 homozygotes). However, other studies have found no association between the apo A-IV-2 allele and plasma lipids18,19,20. One explanation for this disparity is an interaction between the gene and diet that influences the effect of the apo A-IV-2 allele in persons consuming different diets.
These observations suggest that the apo A-IV-2 allele may influence the response of the plasma cholesterol concentration to dietary cholesterol. To test this hypothesis, we measured plasma lipids and lipoproteins in a group of apo A-IV-1/2 heterozygotes and a control group of apo A-IV-1/1 homozygotes while they followed a short-term, high-cholesterol diet.
Methods
Subjects
This study was reviewed and approved by the Clinical Research Practices Committee of the Bowman Gray School of Medicine. Informed consent was obtained from all study subjects. Two hundred twenty-six students, house officers, and employees of the medical school and North Carolina Baptist Hospital were screened for apo A-IV polymorphisms by isoelectric focusing and immunoblotting, and 26 apo-A-IV-1/2 heterozygotes were identified. From the screened cohort 11 healthy, apo-A-IV-1/2 heterozygotes and a control group of 12 apo-A-IV-1/1 homozygotes were recruited for the dietary-modification phase of the study. Exclusion criteria included hyperlipidemia, pregnancy, tobacco use, allergy to eggs, participation in high-intensity athletic activities, or the use of drugs known to affect lipid metabolism. The female subjects were matched for use of oral contraceptives. All of the subjects were within 10 percent of their ideal body weight. Data on age, body-mass index, plasma lipoprotein levels, sex, and apolipoprotein E (apo E) phenotype are shown in Table 1. There were no significant differences between the two groups at entry, with the exception of a higher mean plasma HDL cholesterol level in the apo A-IV-1/2 group (P = 0.04).
|
The study was conducted on an outpatient basis under the supervision of a clinical dietitian. Before the beginning of the study, the subjects were taught the proper manner of recording their daily dietary intake in a written diary and were given classroom instruction regarding the National Cholesterol Education Program Step I diet, which consists of 30 percent of daily calories from fat, with less than 10 percent of daily calories in the form of saturated fat, no more than 10 percent in the form of polyunsaturated fat, and less than 300 mg (0.77 mmol) of cholesterol per day.
During the base-line run-in period (weeks 1 and 2), the subjects followed the Step I diet. Then, during the high-cholesterol period (weeks 3 to 5), the subjects maintained the Step I diet but added four large egg yolks per day, provided in the form of two hard-boiled eggs and a single serving of a frozen "ice cream" mixture made from two extra-large pasteurized egg yolks blended with frozen strawberries or peaches. The eggs and frozen desserts were prepared and distributed daily during the week, and weekend portions were provided every Friday. No attempt was made to adjust the diets to compensate for the added fat and protein in the eggs. Throughout the study, compliance and food intake were monitored by daily dietary diaries.
Dietary Analysis
Dietary diaries were analyzed with diet-analysis and nutrition-evaluation software (Nutritionist IV, version 2.0, N2 Computing, Salem, Oreg.) to determine the mean daily intake of calories, cholesterol, and dietary fiber; the percentage of calories from fat, carbohydrate, protein, and alcohol; and the ratio of polyunsaturated to saturated fatty acids.
Measurement of Plasma Lipids and Lipoproteins
For lipid analysis, blood samples from fasting subjects were collected into Vacutainer tubes containing EDTA (Becton Dickinson, Rutherford, N.J.) by venipuncture. Plasma was separated by low-speed centrifugation and stored at 4 °C; samples were assayed within 24 hours of collection. Plasma lipid analyses were performed in the Lipid Analytic Laboratory of the medical school. Plasma total cholesterol and triglyceride levels were determined with a Technicon RA-1000 analyzer (Technicon, Tarrytown, N.Y.). HDL cholesterol was measured by heparin-manganese precipitation, and the LDL cholesterol level was calculated from the total cholesterol, triglyceride, and HDL values with the Friedewald formula21.
Analysis of Apolipoprotein Phenotype
Apo A-IV phenotypes were determined by isoelectric focusing and immunoblotting22. Five microls of plasma was focused in a 7.5 percent polyacrylamide gel containing 8 M urea and 2 percent ampholytes (pH, 4.5 to 5.4). Electrophoretic transfer of proteins from the gels to nitrocellulose membranes was performed at 100 V and 4 °C in 23 mM TRIS, 12.5 mM borate, and 0.4 mM EDTA (pH 8.3). Apo A-IV bands were visualized with a monospecific rabbit antihuman apo A-IV antiserum followed by goat antirabbit IgG-horseradish peroxidase conjugate and reaction with 4-chloronaphthol in the presence of hydrogen peroxide. Apo E phenotypes were similarly determined.
Statistical Analysis
The statistical significance of differences in entry measurements between the apo A-IV-1/1 and apo A-IV-1/2 groups and of differences in the change in plasma LDL cholesterol levels between subjects in the two groups who had the apo E 3/3 phenotype was determined by two-tailed unpaired Student's t-test. The statistical significance of differences between the study groups in the change in lipid and lipoprotein values from weeks 2 to 5 was determined by the Mann-Whitney rank-sum test. The relation between the change in LDL cholesterol levels and the initial HDL cholesterol levels was examined by regression analysis. The interaction of the apo A-IV and apo E alleles with the change in LDL cholesterol levels was analyzed by two-way analysis of variance.
Results
All of the subjects in both groups completed the study; there were no missing data. Analysis of the daily dietary diaries revealed that both groups adhered very well to the National Cholesterol Education Program diet and that there were no significant differences in the nutrient composition of the groups' diets during the run-in period (Table 2). Mean (±SD) cholesterol intake during this phase was 213 ±82 mg (0.55 ±0.21 mmol) per day in the apo A-IV-1/1 group and 197 ±78 mg (0.51 ±0.20 mmol) per day in the apo A-IV-1/2 group.
|
The high-cholesterol diet had a significant effect on plasma lipid and lipoprotein levels (Table 3 and Figure 1). After three weeks of eggs, the mean total cholesterol increased by 22 mg per deciliter (0.57 mmol per liter) in the apo A-IV-1/1 group, but only by 6 mg per deciliter (0.15 mmol per liter) in the apo A-IV-1/2 group (P = 0.05). The mean LDL cholesterol level increased by 19 mg per deciliter (0.49 mmol per liter) in the apo A-IV-1/1 group, but only by 1 mg per deciliter (0.03 mmol per liter) in the apo A-IV-1/2 group (P = 0.03). There was no significant change in plasma triglyceride or HDL cholesterol levels in either group during the egg phase. There was no correlation between the change in LDL cholesterol levels and initial HDL cholesterol levels in either group. The difference between the changes in LDL cholesterol in the study groups was significant (P = 0.03) after we allowed for the effects of the differences in apo E alleles; however, the differences in the mean changes in LDL cholesterol among the groups with different apo E alleles were not statistically significant (P = 0.69). Moreover, among the subjects with the apo E-3/3 phenotype, there was a significantly lower LDL response in the apo A-IV-1/2 subjects than the apo A-IV-1/1 subjects (P = 0.04).
|
|
Ingestion of a high-cholesterol diet increases plasma total and LDL cholesterol levels in most subjects. This increase is thought to involve the suppression of hepatic cholesterol synthesis23 and down-regulation of hepatic LDL receptors24 by the increased flux of dietary cholesterol reaching the liver in chylomicron remnants. All of the subjects in the apo A-IV-1/1 group had an increase in total and LDL cholesterol that was in accord with the hypercholesterolemic response predicted by the Keys equation for dietary cholesterol25 and with the results of previous studies in which equivalent amounts of egg yolk were ingested24,26,27.
In contrast, a majority of the subjects in the apo A-IV-1/2 group had no increase in either total or LDL cholesterol while following the high-cholesterol diet. Since the intake of fat and cholesterol was the same in both groups throughout the study, this attenuated response cannot be attributed to differences in diet. Body mass can affect the response to dietary cholesterol,28 but there was no difference in body-mass index between the two groups. Increased dietary responsiveness has been associated with higher plasma HDL cholesterol levels,29 but we found no correlation between the change in LDL cholesterol and initial HDL cholesterol levels in either group. Some investigators have reported that apo E alleles influence the absorption of cholesterol30 and the response to dietary cholesterol31; like others,32,33,34 however, we found no significant effect of apo E phenotype on the LDL cholesterol response. Although the apo A-IV gene is tightly linked to the genes for apolipoproteins A-I and C-III on chromosome 11,35 mutations in these genes are rare and have not been implicated in dietary responsiveness; it is therefore unlikely that the apo A-IV-2 allele is a marker for another genetic mutation that affects LDL metabolism. Finally, a substitution of glutamic acid for lysine at position 76 in baboon apo A-IV blunts the increase in HDL cholesterol caused by a diet high in cholesterol and saturated fat36. Thus, we conclude that the attenuated LDL response to dietary cholesterol in the apo A-IV-1/2 group was attributable to an effect of the apo A-IV-2 allele.
At present, we can only speculate on the mechanism of the allele effect. In humans, the gene for apo A-IV is expressed only in the enterocytes of the small intestine,35 and its synthesis is specifically stimulated by fat absorption9,37. In genetic disorders in which the assembly of chylomicrons is disrupted, such as hypobetalipoproteinemia and abetalipoproteinemia, plasma apo A-IV levels are less than 50 percent of normal8,10. Conversely, the complete absence of apo A-IV in patients with deletion of the genes for apolipoproteins A-I, C-III, and A-IV is associated with fat malabsorption and vitamin E deficiency38. These observations suggest that apo A-IV may have a regulatory role in intestinal lipid transport. If apo A-IV-2 reduced the efficiency of cholesterol absorption, it would limit the down-regulation of hepatic LDL receptors during feeding with eggs and blunt increases in plasma LDL cholesterol. In this regard, Kern studied a normocholesterolemic 88-year-old man who ate more than two dozen eggs a day39 and found that the man's fractional cholesterol absorption was only 18 percent, as compared with the normal range of 46 to 55 percent. We have determined that the man is heterozygous for the apo A-IV-2 allele (unpublished data).
The distinctive biophysical properties of the apo A-IV-2 protein22 suggest other possibilities for the allele effect. In the circulation, apo A-IV is rapidly displaced from the surface of nascent chylomicrons by HDL-associated C apolipoproteins40 .Goldberg et al. have postulated that this displacement reaction enhances the adsorption of apolipoprotein C-II (apo C-II) to the surface of chylomicrons41 and ensures maximal activation of lipoprotein lipase, the key enzyme in the intravascular metabolism of triglyceride-rich lipoproteins. Since apo A-IV-2 binds to lipoproteins with higher affinity than A-IV-1,22 it might instead impede the adsorption of apo C-II, thus inhibiting lipoprotein lipase activity and delaying the formation and hepatic clearance of chylomicron remnants. This in turn would decrease the amount of cholesterol reaching the liver in the postprandial state and cause less down-regulation of hepatic LDL receptors.
The apo A-IV-2 allele could also affect the function of apo A-IV in the process of reverse cholesterol transport42. Apo A-IV stimulates cellular cholesterol and phospholipid efflux13 and is found in high concentrations in discoidal HDL-precursor particles in peripheral lymph43. Since apo A-IV-2 binds lipid with higher affinity than A-IV-1, it could be more effective in inducing cholesterol efflux from peripheral cells. The more efficient activation of lecithin-cholesterol acyltransferase by apo A-IV-224 could also enhance the transformation of HDL-precursor particles into mature HDL after they enter the circulation. By augmenting the flux of cholesterol to the liver, either of these effects would maintain peripheral LDL receptors in an up-regulated state and blunt the LDL response to a dietary-cholesterol load.
Whatever the mechanism of action of the apo A-IV-2 allele, one implication of this study is that it may partly account for the heterogeneous response to dietary cholesterol. Indeed, Mata et al.44 recently reported that apo A-IV-1/2 heterozygotes had a significantly smaller decrease in LDL cholesterol on a National Cholesterol Education Program Step II diet than apo A-IV-1/1 homozygotes. It is also possible that heterozygotes can follow diets that are higher in cholesterol with less effect on their cholesterol levels. However, the initial responses to short-term dietary modification may not persist with prolonged feeding. Moreover, it is important to note that not only was there no association between apo A-IV phenotype and base-line plasma lipid levels in our subjects, but also that cholesterol intake in this study was more than twice that of the general population. Furthermore, other genes may affect dietary responsiveness, and some of these could modulate the effect of the apo A-IV-2 allele. Thus, dietary recommendations based on apo A-IV phenotypes would be premature.
Supported by a grant (M01-RR07122) to the General Clinical Research Center of the Bowman Gray School of Medicine, a grant-in-aid (901289) from the American Heart Association, a grant (HL30897) from the National Heart, Lung, and Blood Institute, and a Nutrition Research Initiative grant from the Bowman Gray School of Medicine.
We are indebted to Jennifer B. Jones and Kristin Robie for assistance with the analyses of apo A-IV phenotypes, John R. Crouse and James G. Terry for performing the analyses of apo E phenotypes, and Patty Woodard for assistance in the design and administration of the high-cholesterol diet.
Source Information
From the Department of Internal Medicine, Section of Gastroenterology, Bowman Gray School of Medicine, Winston-Salem, N.C.
Address reprint requests to Dr. Weinberg at the Bowman Gray School of Medicine, Medical Center Blvd., Winston-Salem, NC 27257.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |