HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: M400382-JLR200v1 46/7/1416    most recent Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yang, Q. Articles by Ordovas, J. M. Search for Related Content PubMed PubMed Citation Articles by Yang, Q. Articles by Ordovas, J. M. Social Bookmarking                      What's this?

Journal of Lipid Research, Vol. 46, 1416-1425, July 2005
Copyright © 2005 by American Society for Biochemistry and Molecular Biology

Genome-wide linkage analyses and candidate gene fine mapping for HDL₃ cholesterol: the Framingham Study

Qiong Yang^1,*, Chao-Qiang Lai, Laurence Parnell, L. Adrienne Cupples^*, Xian Adiconis, Yueping Zhu, Peter W. F. Wilson, David E. Housman^**, Amanda M. Shearman^**, Ralph B. D'Agostino and Jose M. Ordovas

^* Department of Biostatistics, Boston University, Boston, MA 02118
Department of Endocrinology, Nutrition, and Diabetes, Boston University, Boston, MA 02118
Department of Mathematics and Statistics, Boston University, Boston, MA 02118
Nutrition and Genomics Laboratory, Jean Mayer-United States Department of Agriculture Human Nutrition Research Center on Aging, Tufts University, Boston, MA 02111
^** Center for Cancer Research, Massachusetts Institute of Technology, Cambridge, MA 02139

Published, JLR Papers in Press, April 1, 2005. DOI 10.1194/jlr.M400382-JLR200

¹ To whom correspondence should be addressed. e-mail: qyang{at}bu.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
High density lipoprotein cholesterol (HDL-C) is inversely associated with coronary heart disease and has a genetic component; however, linkage to HDL-C is not conclusive. Subfractions of HDL, such as HDL₃-C, may be better phenotypes for linkage studies. Using HDL₃-C levels measured on 907 Framingham Heart Study subjects from 330 families around 1987, we conducted a genome-wide variance components linkage analysis with 401 microsatellite markers spaced 10 centimorgan (cM) apart. Nine candidate genes were identified and annotated using a bioinformatics approach in the region of the highest linkage peak. Twenty-eight single nucleotide polymorphisms (SNPs) were selected from these candidate genes, and linkage and family-based association fine mapping were conducted using these SNPs. The highest multipoint log-of-the-odds (LOD) score from the initial linkage analysis was 3.7 at 133 cM on chromosome 6. Linkage analyses with additional SNPs yielded the highest LOD score of 4.0 at 129 cM on chromosome 6. Family-based association analysis revealed that SNP rs2257104 in PLAGL1 at 143 cM was associated with multivariable adjusted HDL₃ (P = 0.03).

Further study of the linkage region and exploration of other variants in PLAGL1 are warranted to define the potential functional variants of HDL-C metabolism.

Supplementary key words high density lipoprotein • high density lipoprotein 3 • microsatellite marker • single nucleotide polymorphism • PLAGL1

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
The inverse association of high density lipoprotein cholesterol (HDL-C) concentrations and risk of coronary heart disease was initially observed in the 1960s and 1970s and is now well established (1). Several studies have provided significant evidence supporting strong familial aggregation of HDL-C levels, but inconsistent results have been reported on whether this is attributable to a gene of major effect (2–6).

HDL is a heterogeneous mixture of particles of different sizes, densities, and compositions (7, 8). Differential chemical precipitation separates HDL particles into two major subfractions: HDL₂ and HDL₃. As thoroughly reviewed by Wilson (9), a number of studies have found that lower concentrations of HDL₂-C and HDL₃-C are significantly associated with increased coronary heart disease risk, suggesting an important role of these two subfractions in coronary heart disease. However, there have been inconsistent findings across populations. To obtain a better understanding of the genetic influences on HDL levels, some studies have examined the more metabolically homogenous subfractions. A segregation study using families of patients who underwent coronary arteriography found that a major gene explains 34% of the total variation in HDL₃-C and 9% in HDL-C, but the variation in HDL₂-C is likely mostly the result of environmental factors (10). Using gradient gel electrophoresis to further differentiate three subclasses (HDL_3a, HDL_3b, and HDL_3c) within HDL₃ and two subclasses (HDL_2a and HDL_2b) within HDL₂, a genome-wide linkage study (11) was performed on the five subclasses, and significant linkage [log-of-the-odds (LOD) > 3] to HDL_2a and suggestive linkage (LOD > 2) to HDL_3a and HDL_2b was found. These studies suggest that the subfractions may provide better phenotypes than HDL-C levels for genetic studies.

Other investigators (12–18) have examined the association between HDL and HDL subfractions and candidate genes such as hepatic lipase, LPL, cholesteryl ester transfer protein (CETP), LCAT, and apolipoproteins A-I, A-II, A-IV, B, C-III, and E. Similar to the outcomes of the segregation studies, there is considerable heterogeneity regarding the significance of these associations in different populations.

To identify chromosome regions likely to contain quantitative trait loci (QTLs) affecting HDL₃-C concentration, we performed a genome-wide linkage study using 401 microsatellite markers and 330 extended families from the Framingham Heart Study. Because this analysis provided significant evidence of linkage to a region on chromosome 6, we identified a number of candidate genes within this promising region of the highest LOD score. A number of single nucleotide polymorphisms (SNPs) in these candidate genes were selected and genotyped to further map possible variants that may affect HDL₃-C concentration.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Subjects
The study subjects are members of the 330 largest extended families in the Framingham Heart Study. The selection criteria and study design of the Framingham Heart Study have been described in detail previously (19, 20). In brief, the study began in 1948 with the enrollment of 5,209 men and women, referred to as the original cohort, from Framingham, Massachusetts, who have undergone biennial examinations. Starting in 1971, 5,124 individuals, adult children of the original cohort, and their spouses, were recruited and are referred to as the offspring cohort. The members of the offspring cohort were examined every 4 years (except for an 8 year gap between the first and second examinations). Subjects in the Framingham Heart Study were ascertained without regard to any trait values. In the mid to late 1990s, 1,702 subjects from the 330 largest extended pedigrees were genotyped for a set of 401 microsatellite markers spaced 10 centimorgan (cM) apart.

The analysis reported here was conducted on a subset of 907 offspring subjects who had HDL₃-C levels measured at examination cycle 4 around 1987 and genotypic data of the CETP polymorphism. The sample included 729 sibling pairs, 37 half-sibling pairs, 41 avuncular pairs, and 489 first cousin pairs. CETP was genotyped on study participants who attended examinations 4 or 5 between 1987 and 1995 and for whom we had DNA. Participants may have missed both exams as a result of death, serious illness, moving out of Massachusetts, or other unknown reasons. Of those 1,089 subjects who attended examination 4 or 5, had DNA, and were members of the families that were included in the genome scan, 282 (23%) subjects were excluded because their CETP genotypes were not available.

All subjects provided informed consent before each clinic visit and were examined under the standard protocol for the Framingham Offspring Study approved by the Institutional Review Board at Boston Medical Center (Boston, MA).

Genotyping of microsatellite markers
DNA specimens were obtained from blood samples routinely collected during examinations of original and offspring subjects between 1987 and 1991 and between 1995 and 2000. DNA was extracted from the buffy coat of whole blood specimens using a Qiagen Blood and Cell Culture DNA Maxi Kit. A genome-wide scan with 401 microsatellite markers over the 22 chromosomes, at an average of 1 marker every 10 cM, was completed by Mammalian Genotyping Service (Marshfield, WI) in 2000. The marker screen set 9 and genotyping protocols are available at the website of the Center for Medical Genetics, Marshfield Medical Research Foundation (http://research.marshfieldclinic.org/genetics/).

Measurement of lipids
Fasting venous blood samples were collected, and plasma was separated from blood cells by centrifugation and immediately used for the measurement of lipids. HDL₃-C concentrations were measured as described by Gidez et al. (21).

Definition of traditional risk factors
Body mass index (BMI; in kg/m²) was calculated using the measurements of weight and height. Alcohol consumption was reported as the usual number of drinks (of comparable ethanol content) per week. Cigarette smoking (yes/no) was defined by whether a person had a history of regularly smoking cigarettes. Physical activity index was derived by taking a weighted average of the number of hours spent in five different energy expenditure categories (sleep, 1.0; sedentary, 1.1; slight, 1.5; moderate, 2.4; and heavy, 5.0). Medication treatment variables (yes/no; including anticholesterol therapy, ß-blockers, and oral contraceptives) were defined by whether a person took such treatment in the past year. Menopause status (yes/no) was defined by whether a woman's menstrual periods had stopped for 1 year. Estrogen therapy (yes/no) was defined by whether a woman was taking estrogen therapy after menopause.

Heritability and linkage analyses
Linkage information content at each locus was calculated in GENEHUNTER. Linkage information content at a locus for a pedigree is defined as how closely the exact allele identity-by-descendent sharing at this locus can be determined for every relative pair (22). Linkage information content at a locus for a sample of pedigrees is a sum of information content at this locus over all of the pedigrees in the sample. The higher the linkage information content, the better the power to detect linkage if the locus is close to the QTL.

To reduce the variability caused by known risk factors, we calculated the standardized residuals from sex-specific multiple linear regression models adjusted for the traditional risk factors. We further calculated normalized deviate scores transformed from ranked standardized residuals for heritability and linkage analyses to avoid potentially inflated false-positive rates attributable to nonnormality. The linear regressions were conducted separately for men and women to account for different associations and to further adjust for oral contraceptive medicine, estrogen therapy, and menopause in women. The heritability and linkage analyses were conducted using MERLIN (23), which is based on a variance components methodology. A measure of the evidence for linkage, the LOD score, is log base 10 of the likelihood ratio of the model with a QTL effect to that without such effect associated with the marker locus under evaluation.

Bioinformatic identification of candidate genes in the linkage region
We conducted a bioinformatic search of candidate genes within or slightly beyond the 2-LOD support interval of the linkage peak (125–150 cM or 6q22.33–6q24.3) on chromosome 6 (Fig. 1), the only chromosome on which significant linkage was found. MapViewer at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/) was used to obtain predicted genes. Similarity searches, using these predicted genes and genomic sequence as input, were performed with the BLAST algorithm (24) (http://www.ncbi.nlm.nih.gov/BLAST/).

View larger version (10K): [in this window] [in a new window]   Fig. 1. Multipoint log-of-the-odds (LOD) scores on chromosome 6 for high density lipoprotein cholesterol (HDL3-C) with existing microsatellite markers and additional single nucleotide polymorphisms (SNPs). cM, centimorgan.

SNP selection and genotyping
Twenty-eight SNPs in the candidate genes were selected from the public SNP database (http://www.ncbi.nlm.nih.gov/SNP) based on the following rationale. First, even distribution across a gene was sought with three, four, or five SNPs. Second, SNPs were selected based on their potential to be functional: altering protein sequence (particularly in functional domains), affecting mRNA splicing, or altering promoter activity. Third, attention was given to SNPs in regions of either protein-coding or noncoding sequence that is conserved in other species, notably the mouse. Fourth, SNPs must have a minor allele frequency of at least 5%. SNP genotyping was conducted using the ABI Prism SnapShot multiplex system (Applied Biosystems, Foster City, CA) and was described previously (30). The primers and probes used for genotyping are displayed in Table 1.

Linkage disequilibrium structure at the candidate genes
Each SNP was evaluated for deviation from Hardy-Weinberg equilibrium (HWE) by randomly selecting one subject from each family and comparing the genotype frequencies of those unrelated subjects with the expected frequencies under HWE in a Chi-square statistic of 2 degrees of freedom.

For each pair of SNPs within the same candidate gene or located less than 500 kb apart, a measure of the strength of linkage disequilibrium (LD), D' (32), was calculated using Haploview (http://www.broad.mit.edu/personal/jcbarret/haploview/).

Family-based association analyses with SNPs in the candidate genes
A score statistic that captures the covariance between marker genotype and trait was calculated using FBAT (33, 34) to test for association. The score statistic divided by its variance was distributed as standard normal (Z). This statistic conditions on parental genotypes to control for bias caused by potential population admixture. Population admixture can be characterized as differential marker allele and trait distribution among unobserved subpopulations. When parental genotypes were incomplete, the distribution of the score statistic was conditioned on the genotypes of available parents and the whole sibship. Because the association was tested with markers in the linkage region, an empirical variance of the score statistic that took into account the dependence among relatives was used in computing the conditional distribution (35).

In haplotype analyses using FBAT (36), multiple marker haplotypes instead of single marker alleles were used in the score statistic. For markers with ambiguous phase, the test statistic was a weighted average of the score statistics over all compatible phase configurations. Two types of hypothesis tests were performed: one was haplotype specific for a difference between a specific haplotype against all other haplotypes combined; the other was a global test for a difference among all of the haplotypes that generalized the score statistic to a Chi-square statistic. The latter test was robust to false positives that could arise from multiple testing using the first type of test.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Characteristics of the study subjects
Means and standard deviations for relevant anthropometric, behavioral, and biochemical risk factors are presented in Table 2. The mean HDL₃-C concentrations were 38.5 and 46.2 mg/dl in men and women, respectively. Men had higher values for BMI, percentage of cigarette smokers, alcohol consumption, physical activity index, percentage on anticholesterol treatment, and percentage using ß-blockers.

Identification of candidate genes in the linkage region
Nine genes that code for proteins that are candidates, either because of a likely direct role in regulating HDL₃-C level or a more indirect mechanism mediated via a role in some aspect of lipid biochemistry, were identified between 125 and 150 cM. These nine candidate genes are NCOA7, CTGF, VNN1, VNN3, VNN2, AIG-1, PLAGL1, LOC285746, and FLJ14735 (LRP11). Locations and annotations of those candidate genes are listed in Table 4.

In Fig. 2, we plotted the multipoint LOD scores as well as FBAT P values from association analyses with both physical and genetic locations on the same figure. Locations of the SNPs and microsatellite markers are also marked on the figure.

View larger version (22K): [in this window] [in a new window]   Fig. 2. Multipoint log-of-the-odds (LOD) scores, –log10 of FBAT P values in the 2-LOD region on chromosome 6 with both physical and genetic locations. The scale above the LOD score curves is the physical map. The position of each SNP or microsatellite marker is marked on the physical map by a line connecting the name of the marker and the position on the physical map.

View larger version (29K): [in this window] [in a new window]   Fig. 3. Linkage disequilibrium (LD) structure at candidate genes. The horizontal white bar at the top indicates the total physical length of the genomic region, and the small vertical black lines in the bar indicate the positions of the SNP markers. The SNP makers are numbered consecutively according to their physical locations, with their SNP name on top. The number in each square is D' x 100 between two markers, and the color indicates the significance of D' > 0. For squares without numbers, the pairwise D' is 1. Black or shades of gray indicate log-of-the-odds (LOD) 2, and white indicates LOD < 2. Only LD of pairs of SNPs within a candidate gene or within 500 kb apart are displayed.

Fine mapping added 28 SNPs in the candidate genes to the existing map
Compared with existing microsatellite markers, incorporating the 28 SNPs into the existing 10 cM map increased the linkage information content of genotype data in the region (Fig. 4). The largest increases occurred at 134 and 140 cM, and each was 0.1.

View larger version (14K): [in this window] [in a new window]   Fig. 4. Information content of chromosome 6 marker genotype data between 125 and 149 cM with microsatellite markers only and with both microsatellite and SNP markers.

Significant association (P < 0.05) was found with two SNPs, rs15960 (CTGF) and rs2257104 (PLAGL1), in family-based association analyses. The A allele of rs15960 was associated with lower levels of HDL₃-C (Z = –2.95, P = 0.0032) in crude analyses without adjusting for traditional risk factors. The association was not significant after adjusting for traditional risk factors (Z = –1.3, P = 0.2). The A allele of rs2257104 has an additive effect associated with lower levels of HDL₃-C (Z = –2.2, P = 0.026) in a fully adjusted model.

However, little of the linkage signal was explained by rs2257104. The proportion of total phenotypic variation explained by rs2257104 is 0.007. After adjusting for rs2257104, the difference in the multivariate LOD score at the SNP location is 0.04.

In haplotype analyses of SNPs within a candidate gene, significant global P values (P < 0.05) were only found with SNPs in PLAGL1 for fully adjusted HDL₃-C (Table 5) by assuming a dominant effect with each haplotype. There were six haplotypes with frequencies >0.01. The haplotype h6 containing C, A, C, and T alleles at rs1884087, rs2257104, rs2076684, and rs2064495, respectively, had a frequency of 0.02 and was associated with lower HDL₃-C levels (haplotype-specific P = 0.02). A borderline significant global P value (P = 0.07) was found with haplotypes in CTGT that contained rs15960 and rs928501, again only in the crude model.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The region of chromosome 6 in which we obtained the highest LOD is close to the region where a 4.64 LOD score at 144 cM was found for BMI measured at examination cycle 1 in the Framingham Heart Study (39). Our results were based on the HDL₃-C measured during examination cycle 4, 16 years after the time of examination cycle 1. For BMI measured at examination cycle 4, the highest LOD score on chromosome 6 was also obtained at the same region, but the value of the LOD scores was 1.43. BMI was used as a covariate in our analysis, so our linkage result on chromosome 6 is not likely to be confounded by BMI.

There has been a great deal of heterogeneity in the linkage findings of HDL-C and its subparticles. Significant findings (LOD > 3) were found on chromosomes 8 and 15 using 477 Mexican Americans of the San Antonio Family Heart Study (11), on chromosome 5 using 1,027 Caucasians of the Family Heart Study (6), on chromosome 12 using 534 pairs of siblings of the Quebec Family Study, on chromosome 10q11 using 1,109 individuals from 92 low HDL-C or hyperlipidemia families, and on chromosome 6 between 73 and 80 cM for increased HDL-C levels (38). It is difficult to explain the lack of replication across these studies. Multiple factors, including phenotype definition, ethnicity, gene-environment interaction, power, and false positives, could contribute to the lack of replication of significant results. Although there is no simple solution to all of these problems, fine mapping in the linkage region may prove an effective tool to discover the existence of any functional variants as the cost of genotyping declines.

The genetic linkage data for HDL₃-C was used to select a region on chromosome 6 from 6q22.33 to 6q24.3, and a bioinformatics analysis produced nine candidate genes. Each of these genes encodes a protein with the likelihood of playing a role in lipid homeostasis by physical interaction with lipid moieties, by regulating lipid biosynthesis or catabolism, by stearoyl hormone-dependent regulation of gene expression, or by association with manifestations of cardiovascular disease. By including the 28 SNPs in the nine candidate genes into the linkage analyses, the linkage information content of the marker data has been improved, thereby providing stronger information with which to evaluate linkage. Furthermore, the increase in the LOD score after incorporating the SNPs provides additional evidence that at least one QTL exists in this region.

Family-based association analyses revealed that two SNPs in two of the candidate genes were associated with HDL₃-C. However, only rs2257104 in PLGAL1 was significant after adjusting for all of the traditional risk factors. PLAGL1, or pleiomorphic adenoma gene-like 1, encodes a putative coactivator of hormone-dependent nuclear receptors with several C₂H₂-type zinc finger domains. SNP rs2257104 is located at 143 cM, which had a LOD score of 1.9 in the linkage analyses incorporating SNPs into the current map. For the SNPs near the maximum LOD score at 128 cM, we did not find any association for fully adjusted HDL₃-C. Although it is possible that the SNPs we typed near the location of the maximum LOD score are not in linkage disequilibrium with a QTL at that location, it is also possible that the QTL was not at the location of the maximum LOD score. Simulation studies (40) have shown that compared with a 10 cM map, further fine mapping using linkage analysis did not result in much reduction in the average location error of maximum LOD score from the QTL. Although only nine candidate genes in the 25 cM interval were investigated and there was considerable distance between these candidate genes, we feel that these nine genes represent the best candidates based on extensive bioinformatics analysis. Nonetheless, we may have missed genes that are located between or outside of these candidate genes but are associated with HDL₃-C. Increasing the density of SNP by genotyping more SNPs within and outside of the nine candidate genes in the linkage region would increase the likelihood of detecting a true association.

We noted that the A allele of rs2257104 resided on all three haplotypes (h2, h5, and h6) with negative Z statistics (Table 5), which was consistent with the result from single SNP association analyses of this marker. However, only the h6 haplotype was statistically significant compared with all others combined. None of the other SNPs in PLAGL1 had an allele that resided on haplotypes that corresponded only to positive or only to negative Z statistics. In sum, these results suggested that the significance of the global haplotype test is most likely attributable to rs2257104.

The estrogen receptor initiates transcription of select genes after binding of ligand and translocation of the receptor-ligand complex to the nucleus (41). Important to lipid physiology, it has been demonstrated that estrogen, via estrogen receptor and coactivators, promotes transcription of the ApoE (42) and ApoA1 genes (43–45). However, for the following reasons, we do not believe that the gene encoding the estrogen receptor (ESR1) is directly responsible for the genetic determinant of HDL₃-C levels: ESR1 maps to 154.8 cM (6q25.1) or 7 Mb beyond our 2-LOD interval; ESR1 is at a LOD score of 0.9, whereas the maximum LOD score seen here is 4.0 after incorporating the SNPs in the linkage analyses. Furthermore, a reanalysis of the linkage data in which we adjusted for either or both of the genotypes of two common variants in ESR1, PvuII and C1335G, generated a similar maximum LOD score (difference in LOD score of <0.17) on chromosome 6.

The study subjects are those with CETP polymorphism typed, because we were also interested in testing whether some of the known lipid candidate genes, such as CETP and hepatic lipase, could explain the linkage peaks. We did not find that adjusting for CETP or hepatic lipase changed the linkage signals much on chromosome 6, which suggested lack of evidence for interaction between QTLs on chromosome 6 and these two known lipid candidate genes. However, when we used all of the subjects with HDL₃-C without regard to whether they were typed for CETP genotype, the maximum LOD score decreased to 1.8 on chromosome 6 in the analyses with the microsatellite markers. Subjects with CETP genotyped were younger on average (48 vs. 50 years) than subjects with missing CETP, which may have resulted in a more homogeneous group to yield better linkage signals. Otherwise, there were no significant differences between those with CETP genotyping and those without.

Regulation of HDL metabolism, lipid homeostasis, and the determinants of cardiovascular health is complex and undoubtedly involves the functions and interactions of many different genes. On the long arm of chromosome 6 are several genes, which were not considered in this study, but whose roles in regulating cardiovascular health via lipid levels ought to be considered as candidates in affecting lipid metabolism. The LPA gene cluster, including APOARGC (apolipoprotein A-related gene C), LPA [lipoprotein Lp(a)], and possible pseudogenes LPAL1 and LPAL2, encodes genes in which certain variations are known to be risk factors for cardiovascular disease (46) but that are not related to the HDL₃-C trait presented in this study. In addition, this cluster maps to 6q26-q27, 10 Mb downstream of LETAL. Several IDDM (insulin-dependent diabetes mellitus) loci have been identified but not narrowed to a specific gene. IDDM5, IDDM8, and IDDM15, although mapping to 6q24-q27, 6q25-q27, and 6q21, respectively, are not considered further because the gene or genetic element responsible for the observed phenotype has not been discovered and because, unlike type II diabetes, which has been shown to have important links to cardiovascular disease via HDL, such a connection to type I diabetes has not been established.

A number of previous genome-wide linkage analyses of cardiovascular anthropometric and lipid phenotypes were based on the same genome scan used in this study. The phenotypes studied included triglycerides (47), BMI (39), waist circumference (48), and blood pressure (49, 50). Our finding of significant linkage on chromosome 6 may be subject to an inflated false-positive rate as a result of multiple testing. Because these anthropometric or lipid phenotypes may not be independent, however, a Bonferroni correction would be too conservative and is thus not suitable for correction of multiple testing here. As the LOD score increased in the fine mapping, the likelihood of a false positive is reduced for our study (51). Furthermore, the chromosome 6 location of the peak LOD score is close to the locations of peak LOD scores of BMI (39) that may be in the lipid metabolism pathway. Considering all of these factors, we believe that our results, although not completely conclusive, can still be a valuable reference for further research in HDL₃ or lipid metabolism in general.

In summary, a genome-wide scan of 401 microsatellite markers in 330 extended families of the Framingham Heart Study has revealed a promising and statistically significant linkage with HDL₃-C concentrations on chromosome 6 (multipoint LOD = 3.7 at 133 cM). None of the classical genes associated with HDL metabolism are located within this region. Bioinformatic analyses of the region between 125 and 150 cM suggested the presence of nine interesting genes in which we further genotyped 28 SNPs. Linkage analyses incorporating those SNPs into the current marker map increased the linkage information of marker data in that region and resulted in a multipoint LOD score of 4.0 at 129 cM. Family-based association analyses revealed that SNP rs2257104 in PLAGL1 was associated with fully adjusted HDL₃-C. Further study of variants in PLAGL1 and increased SNP density in the linkage region are warranted to more clearly define the potential functional variants.

	ACKNOWLEDGMENTS

 
The authors thank all of the individuals who participated in the Framingham Heart Study. This work was supported in part by Contracts N01-HC-25195 and 1-38038 and Grants HL-54776, P50-HL-63494, and RO1-HL-65230 from the National Heart, Lung, and Blood Institute and by Contracts 53-K06-5-10 and 58-1950-9-001 from the U. S. Department of Agriculture Research Service.

Manuscript received October 4, 2004 and in revised form March 2, 2005.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

1. Miller, N. E. 1987. Associations of high-density lipoprotein subclasses and apolipoproteins with ischemic heart disease and coronary atherosclerosis. Am. Heart J. 113: 589–597.[CrossRef][Medline]

2. Cupples, L. A., and R. H. Myers. 1993. Segregation analysis for high density lipoprotein in the Berkeley data. Genet. Epidemiol. 10: 629–634.[CrossRef][Medline]

3. Xu, J. F., V. L. Prenger, J. Coresh, and T. H. Beaty. 1993. Two-locus and bivariate segregation analysis of HDL-C and apo AI. Genet. Epidemiol. 10: 605–610.[CrossRef][Medline]

4. Williams, R. R., S. C. Hunt, P. N. Hopkins, L. L. Wu, S. J. Hasstedt, T. D. Berry, G. K. Barlow, B. M. Stults, M. C. Schumacher, and E. H. Ludwig. 1993. Genetic basis of familial dyslipidemia and hypertension: 15-year results from Utah. Am. J. Hypertens. 6 (Suppl. 1): 319–327.

5. Mahaney, M. C., J. Blangero, D. L. Rainwater, A. G. Comuzzie, J. L. VandeBerg, M. P. Stern, J. W. MacCluer, and J. E. Hixson. 1995. A major locus influencing plasma high-density lipoprotein cholesterol levels in the San Antonio Family Heart Study. Segregation and linkage analyses. Arterioscler. Thromb. Vasc. Biol. 15: 1730–1739.[Abstract/Free Full Text]

6. Peacock, J. M., D. K. Arnett, L. D. Atwood, R. H. Myers, H. Coon, S. S. Rich, M. A. Province, and G. Heiss. 2001. Genome scan for quantitative trait loci linked to high-density lipoprotein cholesterol: the NHLBI Family Heart Study. Arterioscler. Thromb. Vasc. Biol. 21: 1823–1828.[Abstract/Free Full Text]

7. Delalla, O. F., H. A. Elliot, and J. W. Gofman. 1954. Ultracentrifugal studies of high density serum lipoproteins in clinically healthy adults. Am. J. Physiol. 179: 333–337.[Free Full Text]

8. Handbook of Lipoprotein Testing. American Association of Clinical Chemistry, Washington, 1997.

9. Wilson, P. W. 1995. Relation of high-density lipoprotein subfractions and apolipoprotein E isoforms to coronary disease. Clin. Chem. 41: 165–169.[Abstract/Free Full Text]

10. Coresh, J., T. H. Beaty, V. L. Prenger, J. Xu, and P. O. Kwiterovich, Jr. 1995. Segregation analysis of HDL3-C levels in families of patients undergoing coronary arteriography at an early age. Arterioscler. Thromb. Vasc. Biol. 15: 1307–1313.[Abstract/Free Full Text]

11. Almasy, L., J. E. Hixson, D. L. Rainwater, S. Cole, J. T. Williams, M. C. Mahaney, J. L. VandeBerg, M. P. Stern, J. W. MacCluer, and J. Blangero. 1999. Human pedigree-based quantitative-trait-locus mapping: localization of two genes influencing HDL-cholesterol metabolism. Am. J. Hum. Genet. 64: 1686–1693.[CrossRef][Medline]

12. Gerdes, C., L. U. Gerdes, P. S. Hansen, and O. Faergeman. 1995. Polymorphisms in the lipoprotein lipase gene and their associations with plasma lipid concentrations in 40-year-old Danish men. Circulation. 92: 1765–1769.[Abstract/Free Full Text]

13. Minnich, A., G. DeLangavant, J. Lavigne, G. Roederer, S. Lussier-Cacan, and J. Davignon. 1995. GÆA substitution at position –75 of the apolipoprotein A-I gene promoter. Evidence against a direct effect on HDL cholesterol levels. Arterioscler. Thromb. Vasc. Biol. 15: 1740–1745.[Abstract/Free Full Text]

14. Kamboh, M. I., C. E. Aston, C. M. Nestlerode, A. E. McAllister, and R. F. Hamman. 1996. Haplotype analysis of two APOA1/MspI polymorphisms in relation to plasma levels of apo A-I and HDL-cholesterol. Atherosclerosis. 127: 255–262.[CrossRef][Medline]

15. McPherson, R., S. M. Grundy, R. Guerra, and J. C. Cohen. 1996. Allelic variation in the gene encoding the cholesteryl ester transfer protein is associated with variation in the plasma concentrations of cholesteryl ester transfer protein. J. Lipid Res. 37: 1743–1748.[Abstract]

16. Guerra, R., J. Wang, S. M. Grundy, and J. C. Cohen. 1997. A hepatic lipase (LIPC) allele associated with high plasma concentrations of high density lipoprotein cholesterol. Proc. Natl. Acad. Sci. USA. 94: 4532–4537.[Abstract/Free Full Text]

17. Devlin, C. M., V. L. Prenger, and M. Miller. 1998. Linkage of the apo CIII microsatellite with isolated low high-density lipoprotein cholesterol. Hum. Genet. 102: 273–281.[CrossRef][Medline]

18. Ordovas, J. M., L. A. Cupples, D. Corella, J. D. Otvos, D. Osgood, A. Martinez, C. Lahoz, O. Coltell, P. W. Wilson, and E. J. Schaefer. 2000. Association of cholesteryl ester transfer protein-TaqIB polymorphism with variations in lipoprotein subclasses and coronary heart disease risk: the Framingham Study. Arterioscler. Thromb. Vasc. Biol. 20: 1323–1329.[Abstract/Free Full Text]

19. Dawber, T. R., G. F. Meadors, and F. E. J. Moore. 1951. Epidemiological approaches to heart disease: the Framingham Study. Am. J. Public Health. 41: 279–286.

20. Kannel, W. B., M. Feinleib, P. M. McNamara, R. J. Garrison, and W. P. Castelli. 1979. An investigation of coronary heart disease in families. The Framingham Offspring Study. Am. J. Epidemiol. 110: 281–290.[Abstract/Free Full Text]

21. Gidez, L. I., G. J. Miller, M. Burstein, S. Slagle, and H. A. Eder. 1982. Separation and quantitation of subclasses of human plasma high density lipoproteins by a simple precipitation procedure. J. Lipid Res. 23: 1206–1223.[Abstract]

22. Kruglyak, L., M. J. Daly, M. P. Reeve-Daly, and E. S. Lander. 1996. Parametric and nonparametric linkage analysis: a unified multipoint approach. Am. J. Hum. Genet. 58: 1347–1363.[Medline]

23. Abecasis, G. R., S. S. Cherny, W. O. Cookson, and L. R. Cardon. 2002. Merlin—rapid analysis of dense genetic maps using sparse gene flow trees. Nat. Genet. 30: 97–101.[CrossRef][Medline]

24. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389–3402.[Abstract/Free Full Text]

25. Bateman, A., E. Birney, L. Cerruti, R. Durbin, L. Etwiller, S. R. Eddy, S. Griffiths-Jones, K. L. Howe, M. Marshall, and E. L. Sonnhammer. 2002. The Pfam protein families database. Nucleic Acids Res. 30: 276–280.[Abstract/Free Full Text]

26. Mulder, N. J., R. Apweiler, T. K. Attwood, A. Bairoch, D. Barrell, A. Bateman, D. Binns, M. Biswas, P. Bradley, P. Bork, et al. 2003. The InterPro Database, 2003 brings increased coverage and new features. Nucleic Acids Res. 31: 315–318.[Abstract/Free Full Text]

27. Krogh, A., B. Larsson, G. von Heijne, and E. L. Sonnhammer. 2001. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 305: 567–580.[CrossRef][Medline]

28. Emanuelsson, O., H. Nielsen, S. Brunak, and G. von Heijne. 2000. Predicting subcellular localization of proteins based on their N-terminal amino acid sequence. J. Mol. Biol. 300: 1005–1016.[CrossRef][Medline]

29. Nielsen, H., J. Engelbrecht, S. Brunak, and G. von Heijne. 1997. Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Eng. 10: 1–6.[Abstract/Free Full Text]

30. Tai, E. S., J. M. Ordovas, D. Corella, M. Deurenberg-Yap, E. Chan, X. Adiconis, S. K. Chew, L. M. Loh, and C. E. Tan. 2003. The TaqIB and –629C>A polymorphisms at the cholesteryl ester transfer protein locus: associations with lipid levels in a multiethnic population. The 1998 Singapore National Health Survey. Clin. Genet. 63: 19–30.[CrossRef][Medline]

31. Broman, K. W., J. C. Murray, V. C. Sheffield, R. L. White, and J. L. Weber. 1998. Comprehensive human genetic maps: individual and sex-specific variation in recombination. Am. J. Hum. Genet. 63: 861–869.[CrossRef][Medline]

32. Weiss, K. M., and A. G. Clark. 2002. Linkage disequilibrium and the mapping of complex human traits. Trends Genet. 18: 19–24.[CrossRef][Medline]

33. Rabinowitz, D., and N. Laird. 2000. A unified approach to adjusting association tests for population admixture with arbitrary pedigree structure and arbitrary missing marker information. Hum. Hered. 50: 211–223.[CrossRef][Medline]

34. Laird, N. M., S. Horvath, and X. Xu. 2000. Implementing a unified approach to family-based tests of association. Genet. Epidemiol. 19 (Suppl. 1): 36–42.[CrossRef]

35. Lake, S. L., D. Blacker, and N. M. Laird. 2000. Family-based tests of association in the presence of linkage. Am. J. Hum. Genet. 67: 1515–1525.[CrossRef][Medline]

36. Horvath, S., X. Xu, S. L. Lake, E. K. Silverman, S. T. Weiss, and N. M. Laird. 2004. Family-based tests for associating haplotypes with general phenotype data: application to asthma genetics. Genet. Epidemiol. 26: 61–69.[CrossRef][Medline]

37. Coon, H., M. F. Leppert, J. H. Eckfeldt, A. Oberman, R. H. Myers, J. M. Peacock, M. A. Province, P. N. Hopkins, and G. Heiss. 2001. Genome-wide linkage analysis of lipids in the Hypertension Genetic Epidemiology Network (HyperGEN) Blood Pressure Study. Arterioscler. Thromb. Vasc. Biol. 21: 1969–1976.[Abstract/Free Full Text]

38. Canizales-Quinteros, S., C. A. Aguilar-Salinas, E. Reyes-Rodriguez, L. Riba, M. Rodriguez-Torres, S. Ramirez-Jimenez, A. Huertas-Vazquez, V. Fragoso-Ontiveros, A. Zentella-Dehesa, J. L. Ventura-Gallegos, et al. 2003. Locus on chromosome 6p linked to elevated HDL cholesterol serum levels and to protection against premature atherosclerosis in a kindred with familial hypercholesterolemia. Circ. Res. 92: 569–576.[Abstract/Free Full Text]

39. Atwood, L. D., N. L. Heard-Costa, L. A. Cupples, C. E. Jaquish, P. W. Wilson, and R. B. D'Agostino. 2002. Genomewide linkage analysis of body mass index across 28 years of the Framingham Heart Study. Am. J. Hum. Genet. 71: 1044–1050.[CrossRef][Medline]

40. Atwood, L. D., and N. L. Heard-Costa. 2003. Limits of fine-mapping a quantitative trait. Genet. Epidemiol. 24: 99–106.[CrossRef][Medline]

41. Metzger, D., J. H. White, and P. Chambon. 1988. The human oestrogen receptor functions in yeast. Nature. 334: 31–36.[CrossRef][Medline]

42. Srivastava, R. A., N. Srivastava, M. Averna, R. C. Lin, K. S. Korach, D. B. Lubahn, and G. Schonfeld. 1997. Estrogen up-regulates apolipoprotein E (ApoE) gene expression by increasing ApoE mRNA in the translating pool via the estrogen receptor alpha-mediated pathway. J. Biol. Chem. 272: 33360–33366.[Abstract/Free Full Text]

43. Harnish, D. C., M. J. Evans, M. S. Scicchitano, R. A. Bhat, and S. K. Karathanasis. 1998. Estrogen regulation of the apolipoprotein AI gene promoter through transcription cofactor sharing. J. Biol. Chem. 273: 9270–9278.[Abstract/Free Full Text]

44. Lamon-Fava, S., J. M. Ordovas, and E. J. Schaefer. 1999. Estrogen increases apolipoprotein (apo) A-I secretion in Hep G2 cells by modulating transcription of the apo A-I gene promoter. Arterioscler. Thromb. Vasc. Biol. 19: 2960–2965.[Abstract/Free Full Text]

45. Zhang, X., J. J. Jiao, B. R. Bhavnani, and S. P. Tam. 2001. Regulation of human apolipoprotein A-I gene expression by equine estrogens. J. Lipid Res. 42: 1789–1800.[Abstract/Free Full Text]

46. Rainwater, D. L., and C. M. Kammerer. 1998. Lipoprotein Lp(a): effects of allelic variation at the LPA locus. J. Exp. Zool. 282: 54–61.[CrossRef][Medline]

47. Shearman, A. M., J. M. Ordovas, L. A. Cupples, E. J. Schaefer, M. D. Harmon, Y. Shao, J. D. Keen, A. L. DeStefano, O. Joost, P. W. Wilson, et al. 2000. Evidence for a gene influencing the TG/HDL-C ratio on chromosome 7q32.3-qter: a genome-wide scan in the Framingham study. Hum. Mol. Genet. 9: 1315–1320.[Abstract/Free Full Text]

48. Fox, C. S., N. L. Heard-Costa, P. W. Wilson, D. Levy, R. B. D'Agostino, Sr., and L. D. Atwood. 2004. Genome-wide linkage to chromosome 6 for waist circumference in the Framingham Heart Study. Diabetes. 53: 1399–1402.[Abstract/Free Full Text]

49. Levy, D., A. L. DeStefano, M. G. Larson, C. J. O'Donnell, R. P. Lifton, H. Gavras, L. A. Cupples, and R. H. Myers. 2000. Evidence for a gene influencing blood pressure on chromosome 17. Genome scan linkage results for longitudinal blood pressure phenotypes in subjects from the Framingham Heart Study. Hypertension. 36: 477–483.[Abstract/Free Full Text]

50. Meigs, J. B., C. I. Panhuysen, R. H. Myers, P. W. Wilson, and L. A. Cupples. 2002. A genome-wide scan for loci linked to plasma levels of glucose and HbA(1c) in a community-based sample of Caucasian pedigrees: the Framingham Offspring Study. Diabetes. 51: 833–840.[Abstract/Free Full Text]

51. Wiltshire, S., A. P. Morris, M. I. McCarthy, and L. R. Cardon. 2005. How useful is the fine-scale mapping of complex trait linkage peaks? Evaluating the impact of additional microsatellite genotyping on the posterior probability of linkage. Genet. Epidemiol. 28: 1–10.[CrossRef][Medline]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				B. Kaess, M. Fischer, A. Baessler, K. Stark, F. Huber, W. Kremer, H. R. Kalbitzer, H. Schunkert, G. Riegger, and C. Hengstenberg The lipoprotein subfraction profile: heritability and identification of quantitative trait loci J. Lipid Res., April 1, 2008; 49(4): 715 - 723. [Abstract] [Full Text] [PDF]

				R. Sherva, P. Yue, G. Schonfeld, and R. J. Neuman Evidence for a quantitative trait locus affecting low levels of apolipoprotein B and low density lipoprotein on chromosome 10 in Caucasian families J. Lipid Res., December 1, 2007; 48(12): 2632 - 2639. [Abstract] [Full Text] [PDF]

				H. Watanabe, S. Soderlund, A. Soro-Paavonen, A. Hiukka, E. Leinonen, C. Alagona, R. Salonen, T.-P. Tuomainen, C. Ehnholm, M. Jauhiainen, et al. Decreased High-Density Lipoprotein (HDL) Particle Size, Pre{beta}-, and Large HDL Subspecies Concentration in Finnish Low-HDL Families: Relationship With Intima-Media Thickness Arterioscler. Thromb. Vasc. Biol., April 1, 2006; 26(4): 897 - 902. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: M400382-JLR200v1 46/7/1416    most recent Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yang, Q. Articles by Ordovas, J. M. Search for Related Content PubMed PubMed Citation Articles by Yang, Q. Articles by Ordovas, J. M. Social Bookmarking