| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Atherosclerosis and Lipoproteins |
From the National Human Genome Research Institute (S.-H.H.J.), National Institutes of Health, Baltimore, Md; the Laboratory of Statistical Genetics (S.-H.H.J.) and the Laboratory of Biochemical Genetics and Metabolism (Z.H., J.D.S.), Rockefeller University, New York, NY; and the Department of Preventive Medicine (L.C., K.L.), Northwestern University, Chicago, Ill.
Correspondence to Kiang Liu, PhD, Department of Preventive Medicine, Northwestern University Medical School, Suite 1102, 680 N Lake Shore Dr, Chicago, IL 60611. E-mail kiangliu{at}nwu.edu
| Abstract |
|---|
|
|
|---|
0.003) mg/dL in all 5
exams. HDL cholesterol and apolipoprotein A-I levels were
similar among the 3 genotypes. Our serial cross-sectional
analyses indicated that the TT genotype was associated
with higher levels of total cholesterol, LDL
cholesterol, triglycerides, and apoB in young
black men. The broad effect of this polymorphism on several
atherogenic traits suggests that the MTP gene could be influential
in atherosclerosis.
Key Words: polymorphism microsomal triglyceride transfer proteins lipoproteins apolipoproteins
| Introduction |
|---|
|
|
|---|
Rare mutations in mendelian disorders can lead to extreme lipid levels. For example, mutations in the LDL receptor gene can lead to very high cholesterol levels13 14 ; rare mutations in the microsomal triglyceride transfer protein (MTP) gene can cause abetalipoproteinemia, which has very low levels of apoB15 ; and an apoC-II mutation can lead to severe hypertriglyceridemia due to apoC-II deficiency.16 However, these rare mutant alleles cannot explain the normal variation of cholesterol, triglycerides, and apoB in the population at large. Although the apoE genotypes have been repeatedly shown to be associated with LDL cholesterol (LDL-C) levels, this gene accounts for only a small proportion of LDL-C variance.17 The association between the apoE genotypes and triglyceride levels is controversial.18 19 20 A large proportion of the genetic factors influencing LDL-C, triglyceride, and apoB variation remains undetected.
MTP, which exists in the endoplasmic reticulum (ER), plays a key role
in the early stage of lipoprotein assembly, most likely by transferring
lipids to a nascent apoB molecule as it enters the lumen of the
ER.21 When MTP activity is inhibited, the secretion of
apoB is dramatically reduced.22 23 The human MTP gene is
55 kb in length and is located on chromosome 4. Recently, a common
polymorphism (-493 G/T) in the promoter of the MTP gene was
reported to have an association with LDL-C and LDL
triglycerides in healthy, middle-aged, white Swedish
men.24 Two linkage studies also found evidence of linkage
between the MTP gene and LDL-C25 and
triglyceride26 levels. Functional studies have
suggested that this polymorphism (-493 G/T) may regulate the
transcriptional activity by influencing allele-specific binding of
nuclear proteins.24
We conducted a study with a large cohort from the Coronary Artery Risk Development in Young Adults (CARDIA) Study to investigate the prevalence and effect of this polymorphism on lipid levels in 5 consecutive exams in healthy young African American men. In this cohort, DNA samples were available from 586 African American men who had complete lipid measurements from 5 consecutive exams (year 0, 1985 to 1986; year 2, 1987 to 1988; year 5, 1990 to 1991; year 7, 1992 to 1993; and year 10, 1995 to 1996).
| Methods |
|---|
|
|
|---|
Subjects
The details of the CARDIA Study are described
elsewhere.27 In brief, the CARDIA Study is a
multicenter, longitudinal study on lifestyle and evolution of
cardiovascular disease risk factors in young adults
aged 18 to 30 years at initial examination (1985 to 1986). Participants
were randomly recruited from the total community or from selected
census tracts in the community for the centers in Birmingham, Chicago,
and Minneapolis. For the center in Oakland, participants were randomly
recruited from the Kaiser-Permanente health plan membership. Study
subjects received examinations and questionnaires every 2 to 3 years.
To test the present hypothesis, we used the data from black men in
the CARDIA cohort. Five hundred eighty-six black men who had
participated in all 5 examinations over a 10-year follow-up had
available DNA samples. Various individuals were further excluded in the
analyses because they did not fast 8 hours before drawing the
blood (17 individuals in year 0, 59 in year 2, 53 in year 5, 64 in year
7, and 46 in year 10).
Lipoprotein Measurements
Venous blood was drawn after a 12-hour fast. Total
cholesterol and total triglyceride levels were
enzymatically determined.28 LDL-C was estimated by using
the Friedewald equation: LDL-C=total
cholesterol-HDL-C-(triglyceride/5), where
HDL-C indicates HDL cholesterol.29 Subjects
with triglyceride levels
400 mg/dL did not have
calculable LDL-C; thus, their LDL-C levels were not included in the
analyses. HDL-C was measured enzymatically after dextran
sulfatemagnesium precipitation of apoB-containing
particles.30 31 ApoA-I32 and
apoB33 were analyzed by radioimmunoassay. Lipid
data for year 2 were systematically elevated because of laboratory
drift, and we performed analyses on data with and without
adjustment for this effect.
Assay of MTP Polymorphism
DNA was isolated from whole blood samples. The method for
detecting the -493 G/T polymorphism was adapted from Karpe et
al.24 A 109-bp DNA product, encompassing the -493
site, was generated by polymerase chain reaction with a 5' mismatched
primer (5'-GGA TTT AAA TTT AAA CTG TTA ATT CAT ATC AC) and a 3' primer
(5'-AGT TTC ACA CAT AAG GAC AAT CAT CTA). The polymerase chain reaction
was performed in 11 µL of 10 mmol/L Tris-HCl (pH 9.0 at 25°C),
50 mmol/L KCl, 4 mmol/L MgCl2, 0.1%
Triton X-100, 240 µmol/L dNTPs, 1 µL genomic DNA, 12.5 pmol of
each primer, and 0.5 U of Taq polymerase. The reaction was
carried out first by denaturing at 94°C for 4 minutes, then by 35
rounds of denaturing for 30 seconds at 94°C, annealing at 55°C for
45 seconds, and elongating for 30 seconds at 72°C, and then by a
final elongation step of 72°C for 7 minutes. HphI
(0.5 µL, 2.5 U) plus 1 µL of 10x Buffer 4 (New England
Biolabs) was subsequently added into the mix. After 4 hours of
incubation at 37°C, the product was run on 4% agarose gel. A
"G" at position -493 yielded bands of 89 and 20 bp, whereas a
"T" at position -493 yielded a band of 109 bp.
Measurement of Other Covariates
Several factors can influence lipid levels, and they have been
collected in the CARDIA Study. We also investigated the distribution of
demographic characteristics among 3 genotypes. These data
included age, body mass index (BMI), education, alcohol
consumption,34 smoking history,34 total
physical activity scores,35 calories,36
dietary total saturated fatty acids and polyunsaturated fatty
acids,36 37 and Keys scores.37 38 The details
of the measurements of the demographic data were described elsewhere.
In brief, age, years of education, alcohol intake, and smoking history
were ascertained by a self-administered questionnaire. Body weight was
measured while subjects wore light clothing, and height was measured
without shoes. BMI was calculated as weight (in kilograms) divided by
height squared (square meters). Physical activity was measured by use
of a physical activity history questionnaire, and a total physical
activity score was based on moderate and intense exercise in the
previous year. Dietary intake data were obtained from a detailed
diet-history questionnaire. Total calories and fat were estimated from
the diet-history questionnaire. The Keys score was calculated by using
the following formula: 1.35(2S-P)+1.5Z, where S is the percentage of
dietary calories from saturated fatty acids, P is the percentage of
dietary calories from polyunsaturated fatty acids, and Z is the square
root of dietary cholesterol in milligrams per 1000
kilocalories. This formula, derived by Keys et al,9 was
based on data from a serial of metabolic ward studies to
estimate the combined effect of these dietary factors on serum
cholesterol.
Statistical Methods
Allele frequencies were estimated by direct gene counting.
Hardy-Weinberg equilibrium was tested by the
x2 test. We first used ANCOVA to test
for the overall significant differences of lipids and baseline
characteristics among the 3 genotypes, with adjustment for the
differences in data from the 4 participating centers. Second,
pairwise comparisons between any 2 of the 3 genotypes were
tested by using the Scheffé method to adjust for multiple
comparisons to determine the mode of the genetic effect. Third,
covariates including age, BMI, alcohol consumption, smoking history,
total physical activity scores, and Keys scores were further taken into
account in evaluating genetic effects. Because the data were available
for 5 consecutive exams, we conducted the above statistical tests for
each examination to see whether the genetic effect was
consistently significant throughout the follow-up period.
To reduce skewness and kurtosis, triglyceride levels were logarithmically transformed, but untransformed means are presented in the tables. SAS software (SAS Institute Inc) was used for all statistical analyses. A value of P<0.05 (by 2-tailed test) was considered statistically significant.
| Results |
|---|
|
|
|---|
|
Allele Frequency
DNA samples were available from 586 individuals, and 579
individuals were successfully genotyped. There were 304, 236,
and 39 individuals with genotypes GG, GT, and TT (Table 1
), respectively, and the T allele frequency was 0.27. The
distribution of genotypes was not significantly different from
the expectation under the Hardy-Weinberg equilibrium
(P=0.75). The allele and genotype frequencies in
African American men were similar to those in healthy middle-aged white
men.24
Effects of MTP Genotypes on Lipids and
Apolipoproteins
Mean total cholesterol, LDL-C, and apoB levels in the
TT genotype group were consistently higher than those
in either the GT or GG genotype groups in any of the 5 exams,
except for LDL-C in year 10 (Table 2
;
note that apoB levels were available only for the first 2 exams). The
mean plasma levels were very similar between GT and GG
genotypes for these 3 lipids. Thus, the T allele fits a
recessive model, and the data in the GT and GG genotypes were
pooled in the following analyses. Table 3
shows pooled data for total
cholesterol, LDL-C, and apoB, adjusting for the center
effect. The difference between the TT and GG+GT genotypes was
significant (or marginally significant) in the first 4 exams but not
significant in the year-10 examination. By and large, the results
changed only slightly (Table 4
) after
additional adjustment for age and BMI. When more covariates (smoking,
alcohol drinking, physical activity, and Keys scores) were taken into
account, the results showed patterns similar to those in Table 4
(data not shown). The results remained the same when we adjusted for
laboratory drift in year 2. All values from year 2 in the present
study do not account for laboratory drift.
|
|
|
For triglyceride levels, which have a large biological
variation, the TT genotype still had consistently
higher mean levels than did the other 2 genotypes (Table 2
). However, the T allele does not fit a recessive model. A
significant genetic effect was noticed in all exams (Tables 2
and 5
) regardless of adjustment for the
center effect, age, and BMI, although an additive genetic effect was
not observed. When smoking history, daily alcohol intake, and physical
activity were also taken into account, the results remained similar
(data not shown). There were no differences in HDL-C or apoA-I levels
among the 3 genotypes (Table 2
).
|
| Discussion |
|---|
|
|
|---|
Although epidemiological studies have demonstrated an inverse relation between HDL-C and triglyceride levels,43 our results did not show any significant difference in HDL-C levels among genotypes. Whether triglyceride levels in the TT genotype were not high enough to influence HDL-C levels or whether an unknown metabolic adjustment in the TT genotype led to the lack of an inverse relation between triglyceride and HDL-C requires further investigation.
MTP expression can greatly enhance the process of apoB-containing
lipoprotein formation and also decrease apoB degradation in the ER
lumen.21 MTP appears to control the number of
apoB-containing lipoprotein particles secreted rather than their lipid
composition.21 Karpe et al24 have performed
expression studies showing that the T allele has an almost 2-fold
higher transcriptional activity than does the G allele. This might
suggest that individuals with the TT genotype have a higher
production rate of apoB-containing lipoproteins. It would be a
reasonable hypothesis to test for increasing levels of components in
apoB-containing lipoproteins in individuals with the TT
genotype. Furthermore, given only 1 apoB molecule in each VLDL
and LDL particle, we would expect to see a compatible increase in apoB
and in lipids carried mainly by apoB-containing particles. Indeed, our
results support this hypothesis. ApoB and LDL-C were
9% (in year 0)
and 11% (in year 2) higher in the TT genotype than in the
GG+GT genotype (Table 3
). This finding offers further
support for the association between this polymorphism and lipid
profiles.
It should be noted that the increase of apoB-related lipids in the TT individuals is not as striking as what was observed in the expression studies, suggesting that the TT genotype could have 2-fold higher transcriptional activity than the GG genotype.24 The apoB gene is constitutively expressed44 ; thus, the rate of synthesis of apoB remains constant. ApoB is rapidly degraded in the ER before secretion if a nascent apoB molecule does not acquire sufficient lipid.45 46 Therefore, the secretion of apoB and apoB-containing lipoprotein from the liver depends on the balance between degradation and lipoprotein assembly.44 47 Several studies have indicated that MTP transfers lipid to apoB and stabilizes nascent apoB polypeptides in the ER.22 48 49 Thus, although the T allele has 2-fold higher transcriptional activity in transfected cells, the availability of apoB polypeptides in the ER could restrict the production of apoB-containing lipoprotein. Accordingly, the increased levels of plasma apoB-containing lipoprotein could be disproportional to the increased levels of MTP. Alternatively, there might be posttranscriptional modification to metabolize overproduced apoB-containing lipoprotein in the TT individuals. However, further studies are necessary to clarify this point.
Abetalipoproteinemia is a rare autosomal recessive disease caused by mutations in the MTP gene leading to undetectable MTP activity.50 51 Subjects have only trace levels of apoB-containing lipoproteins. Therefore, MTP activity should reflect levels of apoB-containing lipoproteins. In the present study, the TT individuals had a higher level of apoB-containing lipoproteins, which is in concert with the expression study by Karpe et al.24 However, Karpe et al reported that the TT genotype was associated with lower levels of LDL-C (mean±SD 2.90±0.59 mmol/L) and LDL triglycerides (0.23±0.04 mmol/L) than the GT (3.70±0.98 mmol/L [LDL-C] and 0.33±0.12 mmol/L [LDL triglycerides]) and GG (3.74±0.79 [LDL-C] and 0.32±0.10 mmol/L [LDL triglycerides]) genotypes. The explanation of the opposite findings between our study and theirs is not straightforward. There are some potentially important differences between the 2 studies, such as sample sizes, race, age, BMI, and other environmental factors. Our sample size was >3-fold larger than theirs (579 versus 184). Our population consisted of young black men, and their study consisted of middle-aged white men. In addition, our population had a higher BMI than theirs. Other environmental factors, such as diet, could also contribute to the discrepancies between these 2 studies. Unfortunately, no dietary information was provided by Karpe et al. Although the TT individuals were almost always associated with a higher level of apoB-containing lipoproteins in our 5 consecutive measurements, total cholesterol and LDL-C levels from the year-10 examination were similar among the 3 genotypes. Whether the substantial decrease of cholesterol in the TT genotype in the year-10 examination will continue and result in an opposite association between the TT genotype and lipid levels or whether it was merely due to a random variation requires further follow-up. Herrmann et al52 reported no association between another polymorphism at -400 (A/T) and lipid profiles in 728 white men aged 25 to 64 years. Karpe et al also found this -400 A/T polymorphism, but they reported that this -400 polymorphism was not a functional polymorphism. Furthermore, these 2 polymorphisms are not in complete linkage disequilibrium.24 If a tested locus is not in complete linkage disequilibrium with a causative locus, the power to detect association under this circumstance will be substantially reduced.53 Therefore, the nonsignificant results from the study of Herrmann et al might be due to this situation. On the other hand, whether the study of Herrmann et al can be explained by the loss of a lipid-raising effect in the older TT individuals remains unanswered. Future studies to investigate the age and racial effects and other possible modifying factors, such as another functional mutation in linkage disequilibrium with this polymorphism, may answer this question.
In summary, the present study found that the TT genotype in young African American men was associated with a higher mean level of apoB-related lipids. The T allele appears to be recessive in raising total cholesterol, LDL-C, and apoB levels, given that the GG and GT genotypes had similar mean levels. The genetic effect of the -493 G/T MTP polymorphism on several atherogenic lipids suggests that this polymorphism could have profound effects on cardiovascular risks and could be important in understanding the genetics of atherosclerosis.
| Acknowledgments |
|---|
Received October 8, 1999; accepted December 29, 1999.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Gastaldi, S. Diziere, C. Defoort, H. Portugal, D. Lairon, M. Darmon, and R. Planells Sex-specific association of fatty acid binding protein 2 and microsomal triacylglycerol transfer protein variants with response to dietary lipid changes in the 3-mo Medi-RIVAGE primary intervention study Am. J. Clinical Nutrition, December 1, 2007; 86(6): 1633 - 1641. [Abstract] [Full Text] [PDF] |
||||
![]() |
X. Pan and M. M. Hussain Diurnal Regulation of Microsomal Triglyceride Transfer Protein and Plasma Lipid Levels J. Biol. Chem., August 24, 2007; 282(34): 24707 - 24719. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. M. Morgan, H. M. Krumholz, R. P. Lifton, and J. A. Spertus Nonvalidation of Reported Genetic Risk Factors for Acute Coronary Syndrome in a Large-Scale Replication Study JAMA, April 11, 2007; 297(14): 1551 - 1561. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Lundahl, C. Skoglund-Andersson, M. Caslake, D. Bedford, P. Stewart, A. Hamsten, C. J. Packard, and F. Karpe Microsomal triglyceride transfer protein -493T variant reduces IDL plus LDL apoB production and the plasma concentration of large LDL particles Am J Physiol Endocrinol Metab, April 1, 2006; 290(4): E739 - E745. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Stan, M. Lambert, E. Delvin, G. Paradis, J. O'Loughlin, J. A. Hanley, and E. Levy Intestinal fatty acid binding protein and microsomal triglyceride transfer protein polymorphisms in French-Canadian youth J. Lipid Res., February 1, 2005; 46(2): 320 - 327. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Ledmyr, A. D. McMahon, E. Ehrenborg, L. B. Nielsen, M. Neville, H. Lithell, P. W. MacFarlane, C. J. Packard, F. Karpe, and on behalf of the WOSCOPS executive; The Microsomal Triglyceride Transfer Protein Gene-493T Variant Lowers Cholesterol But Increases the Risk of Coronary Heart Disease Circulation, May 18, 2004; 109(19): 2279 - 2284. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Phillips, K. Mullan, D. Owens, and G.H. Tomkin Microsomal triglyceride transfer protein polymorphisms and lipoprotein levels in type 2 diabetes QJM, April 1, 2004; 97(4): 211 - 218. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. J. Geesaman, E. Benson, S. J. Brewster, L. M. Kunkel, H. Blanche, G. Thomas, T. T. Perls, M. J. Daly, and A. A. Puca Haplotype-based identification of a microsomal transfer protein marker associated with the human lifespan PNAS, November 25, 2003; 100(24): 14115 - 14120. [Abstract] [Full Text] [PDF] |
||||
![]() |
S.-H. H. Juo, L. Colangelo, Z. Han, J. D. Smith, and K. Liu Confirmation of the Microsomal Triglyceride Transfer Protein Genetic Effect on Lipids in Young African American Men From the CARDIA Study Arterioscler. Thromb. Vasc. Biol., May 1, 2003; 23(5): 912 - 913. [Full Text] [PDF] |
||||
![]() |
M. C. Lin, E.-J. Wang, C. Lee, K.T. Chin, D. Liu, J.-F. Chiu, and H.-f. Kung Garlic Inhibits Microsomal Triglyceride Transfer Protein Gene Expression in Human Liver and Intestinal Cell Lines and in Rat Intestine J. Nutr., June 1, 2002; 132(6): 1165 - 1168. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Ledmyr, F. Karpe, B. Lundahl, M. McKinnon, C. Skoglund-Andersson, and E. Ehrenborg Variants of the microsomal triglyceride transfer protein gene are associated with plasma cholesterol levels and body mass index J. Lipid Res., January 1, 2002; 43(1): 51 - 58. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2000 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |