Evidence for a Gene Influencing High-Density Lipoprotein Cholesterol on Chromosome 4q31.2

【摘要】  Objective- A low level of plasma high-density lipoprotein cholesterol (HDL-C) is a major risk factor for coronary atherosclerosis. To identify novel genes regulating plasma HDL-C levels, we investigated 13 multigenerational French Canadian families with an average of 12 affected individuals per family for genome-wide signals, which we subsequently fine mapped.

Methods and Results- We genotyped a total of 362 individuals, including 151 affected subjects for 485 autosomal microsatellite markers. In parametric 2-point linkage analyses, the highest 2-point logarithm of odds (lod) score of 4.6 was observed with marker D4S424 on chromosome 4q31.21 (at 142 Mb). The multipoint analysis of this region resulted in a lod score of 3.8 and a lod -1 region of 12.2 cM, containing 40 known genes. The results were obtained by allowing for genetic heterogeneity among these extended pedigrees, and 50% of families were linked to this region with the highest single-pedigree lod score being 3.6. We further restricted the linked region from 12.2 to 2.9 cM (2.37 Mb) by genotyping 15 additional markers in the 3 families with the highest lod scores. We sequenced 4 genes with a likely role in lipid metabolism as well as 2 genes residing directly under the linkage peak but found no evidence for a causative variant. None of the genes residing in the significantly restricted 2.37-Mb region has been associated previously with HDL-C metabolism.

Conclusion- This study provides significant evidence for a gene influencing HDL-C on chromosome 4q31.21.

To identify novel genes regulating plasma HDL-C levels, we analyzed multigenerational French Canadian families for genome-wide signals, which we subsequently fine mapped. We identified a 2.37-Mb region on chromosome 4q31.21 likely containing a gene influencing HDL-C levels.

【关键词】  highdensity lipoprotein cholesterol family study complex trait coronary heart disease gene identification

Introduction

Plasma high-density lipoprotein cholesterol (HDL-C) is a quantitative trait in which complex interactions of several genes and environmental factors determine the plasma levels in the general population. A decreased level of plasma HDL-C is a major risk factor for coronary atherosclerosis. Furthermore, aggregation of other cardiovascular risk factors including features of the metabolic syndrome is typical in patients with low HDL-C. 1,2 The genetic component, estimated to determine 50% of plasma HDL-C variability, has been investigated heavily to identify genes regulating plasma HDL-C levels. 3 Previously, complex quantitative traits have been suggested to be caused by common sequence variants, each with a small to moderate phenotypic effect. 4,5 However, most recent studies suggest that rare and common variants confer the susceptibility to low plasma levels of HDL-C in the general population, 6,7 although the extent to which each group contributes to this susceptibility is currently not known.

HDL has a key role in reverse cholesterol transport (RCT), mobilizing cholesterol from the peripheral tissues to liver. This mechanism contributes to the cardioprotective effect of HDL. In RCT, the ATP-binding cassette transporter A1 (ABCA1) protein controls efflux of intracellular cholesterol to lipid-poor apolipoprotein A-I, the major apolipoprotein of HDL. Interestingly, ABCA1 mutations were discovered to cause Tangier disease, a rare recessive HDL deficiency, 8,9,10 and sequence variants in ABCA1 also appear to contribute to variations in plasma HDL-C levels in the general population. 6,7 Recent evidence further shows that 2 other ABC transporters, ABCG1 and ABCG4, mediate the efflux of cellular cholesterol to smaller and larger HDL subclasses, HDL2 and HDL3, which constitute the bulk of the plasma HDL. 11 Genes implicated in rare Mendelian forms of HDL deficiency include apolipoprotein A-I and lecithin cholesterol acyltransferase; 12,13 and genes implicated in candidate gene studies include apolipoproteins and the enzymes remodeling HDL. 14,15 The previous genome-wide scans have also identified 6 chromosomal loci for HDL-C: a locus on chromosome 8q in Mexican Americans and Finns; 16,17 loci on 15q and 9p in Mexican Americans; 16,18 a locus on 11q23 in 105 Utah coronary heart disease families; 19 and loci on 16q and 20q in Finns. 17,20 However, DNA sequence variants contributing to variation in plasma levels of HDL-C in the general population are largely unknown, especially regarding the prevalence of variants with major effects.

We performed a genome scan in 13 extended multigenerational French Canadian families, each exhibiting affected individuals in consecutive generations. Importantly, families with a defect in cellular cholesterol efflux or previously known ABCA1 mutations were excluded. This was done to minimize confounding factors and to identify genes with novel functions. We also anticipated that extended families with an average of 12 HDL-C-affected individuals per family could provide a powerful approach for the identification of genes contributing to the complex HDL-C trait, especially when allowing for heterogeneity and thus assuming that the same gene is causative within a multigenerational family but not necessarily between all families.

Materials and Methods

Subjects

A total of 13 multigenerational French Canadian families consisting of 362 genotyped family members were collected in the Cardiovascular Genetics Laboratory, McGill University Health Centre, Royal Victoria Hospital, Montreal, Canada. All subjects provided separate informed consent forms for plasma and DNA sampling, isolation, and storage. The research protocol was approved by the research ethics board of the McGill University Health Centre.

The selection criterion for probands was an HDL-C level less than the fifth age-/sex-specific population percentile based on the Lipid Research Clinics Population Studies Data Book, as described previously. 21 Exclusion criteria for the probands 10 mmol/L), cellular cholesterol efflux or phospholipid efflux defect in skin fibroblasts, or previously known ABCA1 mutations. All available living relatives were invited to participate in the study. Family members were sampled after a 12-hour fast and discontinuation of lipid-modifying medications for 4 weeks. Demographic and clinical information, medications, and lipoprotein profiles were determined on all subjects. In the 13 extended families, there were 151 subjects affected using the HDL-C less than or equal to the 10th age-/sex-specific population percentile. Table 1 shows the phenotypic characteristics of the family members.

TABLE 1. Characteristics of the 13 French Canadian Low HDL-C Families

Biochemical Measurements

Lipids and lipoproteins were measured using standardized techniques as described previously. 9,22 Low-density lipoprotein cholesterol (LDL-C) was calculated according to the Friedewald formula [LDL-C (mmol/L)=total cholesterol-(triglycerides/2.19+HDL-C)], 4.5 mmol/L. In this case, ultracentrifugation of plasma was used and lipoprotein lipid concentration measured directly. Cellular cholesterol efflux and phospholipid efflux assays have been performed on the probands as described previously. 9,22,23

Genotyping and Sequencing

The genome scan was executed using 485 autosomal microsatellite markers covering the human genome with an average marker density of 6 cM. DNA from 362 individuals in the 13 French Canadian HDL-C families was genotyped by deCODE Genetics, Reykjavik, Iceland, using the Applied Biosystems automated DNA sequencing system. The genome-wide scan marker sets by deCODE Genetics are based partly on the ABI JD Marker Linkage set and partly on the in-house designed and validated markers originally selected from the Marshfield genetic map. For fine mapping of chromosome 4q31.21, a total of 15 additional fine mapping markers with 0.50 were analyzed in 155 individuals of the 3 families with the highest logarithm of odds (lod) scores. Their genotyping was also performed by the deCODE Genetics using the Applied Biosystems automated DNA sequencing system.

Sequencing of the RAB33B, a member of the RAS oncogene family (RAB33B), TBDN100 transcriptional coactivator tubedown-100 (TBDN100), uncoupling protein 1 (mitochondrial, proton carrier; UCP1), ATP-binding cassette, subfamily E (OABP), member 1 (ABCE1), solute carrier family 7, (cationic amino acid transporter, y+ system) member 11 (SLC7A11), and protocadherin 18 (PCDH18) genes was performed using the dye termination method and the Applied Biosystems automated DNA sequencing system.

Statistical Analyses

We performed parametric 2-point linkage analyses of all 485 microsatellites using the Mendel program 24 under the assumption of a recessive mode of inheritance. The Mendel program was selected because it allowed us to keep these extended families intact in the 2-point analyses. For these analyses, subjects were coded as affected or unaffected based on the 10th age-/sex-specific percentiles for HDL-C. We performed the analyses of this complex trait by allowing for heterogeneity because ignoring locus heterogeneity significantly decreases the power to detect linkage. 25 We used a gene frequency of 8% based on an estimated 1% disease prevalence for low HDL-C and allowed for a 2% phenocopy rate to address the problem of incomplete penetrance.

We also conducted parametric multipoint analyses using the Location Score option of SimWalk2 (version 2.90) 26 on chromosomal regions 3.0) to further explore the linkage evidence obtained in the initial 2-point screening of all markers. Assuming again a recessive mode of inheritance, we incorporated the same parameters into the model as in the 2-point linkage analyses described above. The SimWalk2 Location Score option calculates lod scores under homogeneity ( =1.0) and by allowing for heterogeneity. The heterogeneity lod scores are maximized over a grid of possible values. The SimWalk2 program was also used to build the gene-specific haplotypes of the identified single nucleotide polymorphisms (SNPs).

Results

We performed a genome-wide scan of 13 extended multigenerational French Canadian families with 362 genotyped individuals (an average of 28 genotyped individuals per family) to identify genes regulating plasma HDL-C levels. These 13 families included 151 affected subjects, resulting in an average of 12 affected individuals per family. All of the families had 3 generations of which DNA was available for family members. The detailed subject and family characteristics are shown in Tables 1 and 2.

TABLE 2. Phenotypic Characteristics of the 13 French Canadian Families With Low HDL-C Shown Separately in the Affected individuals, Unaffected Individuals, Spouses, and Probands

A total of 485 autosomal microsatellite markers were genotyped and tested for linkage using a recessive mode of inheritance. In the initial screening, we performed a 2-point linkage analysis for each marker on each chromosome by allowing for heterogeneity. We were able to keep all these extended families intact in the linkage analysis by using the Mendel program and thus to obtain all information about vertical transmissions in the families. The results of the 2-point analyses are shown in Figure 1. The highest 2-point heterogeneity lod score of 4.6 was observed with marker D4S424 on chromosome 4q31.21 (at 142 Mb), and 50% of families were linked to this chromosomal region ( Figure 1 ). It is noteworthy that separately, 1 of the families with 86 genotyped individuals of which 39 were affected provided a lod score of 3.6 (with a posterior probability of linkage of 0.9998) with marker D4S424 using the Mendel program. The next most significant family resulted in a 2-point lod score of 1.8 (with a posterior probability of linkage of 0.9834) with marker D4S424. Peak markers 1.0 are also indicated in Figure 1. Markers on chromosomes 2, 12, 14, 15, 1.0 ( Figure 1 ).

Figure 1. Two-point linkage results of the genome-wide scan of the 13 French Canadian families with low HDL-C when allowing for heterogeneity. 1.0.

We performed the multipoint analyses using the Location Score option of SimWalk2 (version 2.90) only for the regions that showed significant evidence for linkage in the initial screening with the 2-point linkage analysis to further explore the interesting signals of potential linkage evidence by multipoint. The multipoint analysis of the 4q31.21 region resulted in a lod score of 3.8 and a 12.2-cM lod -1 region ( Figure 2 ). Separately, the same family that produced a 2-point lod score of 3.6 produced a multipoint lod score of 3.8. Again, 50% of the families were linked to the region in the multipoint analysis. It is worth noting that the average marker density for this linked region in the genome scan was 1 marker/9.4 cM, affecting the information content of the multipoint analysis. The description of the gene symbols and exact positions of the genes residing in the lod -1 region are shown in supplemental Table I (available online at http://atvb.ahajournals.org).

Figure 2. Multipoint analyses of the genome-wide scan and fine mapping for HDL-C on chromosome 4q31.21. The curve of 7 markers represents the multipoint analysis using all families, and the curve of 22 markers the multipoint analysis using a subset of 3 families. For a list of genes residing in the lod -1 region, see Table I.

We fine mapped the 12.2-cM region on 4q31.21 by genotyping a total of 15 additional microsatellite markers. In the initial scan, we had genotyped 3 markers (D4S1527, D4S424, and D4S2962) for this region. This strategy resulted in a marker density of 1 marker/0.68 cM. We genotyped the additional 15 markers in 155 individuals of the 3 families with the highest lod scores and performed the multipoint analysis again using the Location Score option of SimWalk2 (version 2.90) with the same parameters as in the initial multipoint analysis. As a result, the lod -1 region was further restricted from 12.2 to 2.9 cM ( 2.37 Mb; Figure 2 ). Of the 40 previous candidate genes, 10 reside within this restricted region of 2.37 Mb. Two of these genes, PCDH18 and SLC7A11, reside directly under the peak ( Figure 2 ). The appearance of 2 peaks in this lod -1 region is more likely to be related to the marker frequencies (ie, the information content of the microsatellites) rather than to biological processes because 9 of the 22 genotyped microsatellite markers have a relatively low heterozygosity value between 0.5 and 0.7. Furthermore, there is a gap of 3.6 cM between markers D4S1586 and D4S3008, which reside in close vicinity of the second peak ( Figure 2 ), 0.5) were found using the Marshfield, deCODE, or UCSC Human Genome Browser databases.

We sequenced probands of 2 linked families, including the proband of the family that produced the significant linkage signal alone, for the coding regions and exon-intron boundaries of all regional candidate genes with a probable role in lipid metabolism. Accordingly, the RAB33B, TBDN100, UCP1, and ABCE1 genes were sequenced. We also sequenced the PCDH18 and SLC7A11 genes that are located directly under the fine map peak ( Figure 2 ). None of the 17 identified variants in these 6 genes (Table II, available online at http://atvb.ahajournals.org) represent a missense or nonsense variant with obvious functional consequences, although the possibility of their regulatory role cannot be excluded without actual functional analyses. We investigated these 17 variants first in 5 affected individuals/family. The segregation of 10 of the 17 variants with the low HDL-C trait could not be ruled out by genotyping the 5 affected subjects per family (for rs numbers of the 10 variants, see Table II). Therefore, we then genotyped these 10 SNPs in all family members of the particular family/families. None of the SNPs segregated with the low HDL-C trait in these extended families, nor did we find significant evidence for linkage with any of the SNPs using the Mendel program because all 10 SNPs resulted in lod scores <1.1. Finally, no gene-specific haplotypes segregating with the low HDL-C trait were observed using the SimWalk2 program.

Discussion

We identified a significant signal of linkage (a lod score of 4.6) for low plasma HDL-C on chromosome 4q31.21 in French Canadian families with low HDL-C. These extended families with an average of 12 low HDL-C-affected family members and an average of 28 genotyped individuals per family are very powerful to test for linkage because of the large number of potentially informative meioses in several consecutive generations. In the linkage analyses, we allowed for heterogeneity because plasma HDL-C is a quantitative trait likely regulated by variants of multiple genes interacting with environmental factors. We hypothesized that the underlying segregating variants conferring susceptibility to low levels of plasma HDL-C in these extended multigenerational families may well differ between the families but less likely within the families. The latter assumption is supported by the fact that these families are French Canadian, a regional population group known to exhibit a higher degree of genetic homogeneity when compared with more mixed populations. 27 Approximately 50% of families were linked to chromosome 4q31.21. After observing significant evidence for linkage in the total study sample, we investigated the families separately. In 1 single family with 86 genotyped individuals and 39 HDL-C affected individuals, we observed a significant lod score of 3.6 for this chromosomal region on 4q31.21.

Six previous genome-wide scans have identified 6 chromosomal loci for HDL-C on 8q, 9p, 11q23, 15q, 16q, and 20q. 16-20 1.0 for 2 of these 6 loci: the loci on 15q and 16q. In addition, we also detected the region on 2q observed previously for HDL-C in the Utah and Finnish families. 28,29 However, the only significant signal in the present study was observed for 4q31.21.

To select the region for further fine mapping, we used the lod -1 criterion, a heuristic approach that can generally be considered as somewhat arbitrary when investigating complex traits because this 4q31.21 region provided the only significant signal in this genome scan, and furthermore, when analyzing the families separately, 1 extended family alone resulted in a significant lod score of 3.6. Together, these results suggest that a gene in the region may contribute significantly to the low HDL-C levels in a manner resembling a single gene disorder at least in this 1 family, resulting in a significant lod score separately. If the employment of the lod -1 does not result in gene identification within this region, we will extend the investigated region to lod -2 in future steps of the fine mapping.

Fine mapping of the 4q31.21 region with 15 additional microsatellites and a marker density of 1 marker/0.68 cM allowed us to further restrict the lod -1 region from 12.2 cM, including 40 genes to 2.9 cM ( 2.37 Mb), including 10 genes. However, as described above, we recognize that the application of the lod -1 strategy is somewhat arbitrary when investigating complex traits. Therefore, we sequenced all genes among the 40 genes that have a likely role in lipid metabolism (ie, TBDN100, RAB33B, UCP1, and ABCE1) but found no evidence for a variant cosegregating with low HDL-C in the 2 linked families examined. We also sequenced the 2 genes residing precisely under the fine map peak: PCDH18 and SLC7A11. The PCDH18 gene is a potential calcium-dependent cell-adhesion protein expressed in several tissues, and the SLC7A11 gene is a member of a heteromeric Na+-independent anionic amino acid transport system, highly specific for cystine and glutamate. Again, no evidence for a cosegregating variant was observed. In future studies, the 34 remaining regional genes need to be investigated. Finally, all SNPs identified in the linked families as well as the tag SNPs, provided by the HapMap Project and capturing most of the genetic variation in this region, could be tested for association with HDL-C in a Canadian population-based case-control study sample to identify the underlying gene. By using this strategy for the identification of the DNA sequence variants, we can also identify rare variants in addition to the common variants identified by the HapMap Project. This may be of importance because rare variants may have a role in individual families and because recent studies suggest that rare and common variants confer the susceptibility to low plasma levels of HDL-C in the general population. 6,7 Because none of the probands in this study had a cellular cholesterol efflux defect, the functional role of the underlying gene for HDL-C in this region is unlikely to be directly linked to cellular efflux of cholesterol.

Acknowledgments

This work was supported by grant MOP 15042 from the Canadian Institutes of Health Research. J.C.E. and M.M. are Chercheur-Boursiers of the Fonds de la Recherche en Sante du Quebec. P.P. is supported by National Institutes of Health grants HL-28481 and HL-70150 as well as by American Heart Association grant 0430180N. We thank the French Canadian families that are participating in the study. The authors would also like to thank J. Faith, M. Lemire, and J.C. Loredo-Osti for their advice and assistance.

【参考文献】
  Genest J Jr, Bard JM, Fruchart JC, Ordovas JM, Schaefer EJ. Familial hypoalphalipoproteinemia in premature coronary artery disease. Arterioscler Thromb Vasc Biol. 1993; 13: 1728-1737.

Weber M, McNicoll S, Marcil M, Connelly P, Lussier-Cacan S, Davignon J, Latour Y, Genest J Jr. Metabolic factors clustering, lipoprotein cholesterol, apolipoprotein B, lipoprotein (a) and apolipoprotein E phenotypes in premature artery disease in French Canadians. Can J Cardiol. 1997; 13: 253-260.

Prenger VL, Beaty TH, Kwiterovich PO. Genetic determination of high-density lipoprotein-cholesterol and apolipoprotein A-1 plasma levels in a family study of cardiac catheterization patients. Am J Hum Genet. 1992; 51: 1047-1057.

Lander ES. The new genomics: global views of biology. Science. 1996; 274: 536-539.

Collins FS, Guyer MS, Chakravarti A. Variations on a theme: cataloging human DNA sequence variation. Science. 1997; 278: 1580-1581.

Frikke-Schmidt R, Nordestgaard BG, Jensen GB, Tybjaerg-Hansen A. Genetic variation in ABC transporter A1 contributes to HDL cholesterol in the general population. J Clin Invest. 2004; 114: 1343-1353.

Cohen JC, Kiss RS, Pertsemlidis A, Marcel YL, McPherson R, Hobbs HH. Multiple rare alleles contribute to low plasma levels of HDL cholesterol. Science. 2004; 305: 869-872.

Bodzioch M, Orso E, Klucken J, Langmann T, Bottcher A, Diederich W, Drobnik W, Barlage S, Buchler C, Porsch-Ozcurumez M, Kaminski WE, Hahmann HW, Oette K, Rothe G, Aslanidis C, Lackner KJ, Schmitz G. The gene encoding ATP-binding cassette transporter 1 is mutated in Tangier disease. Nat Genet. 1999; 22: 347-351.

Brooks-Wilson A, Marcil M, Clee SM, Zhang LH, Roomp K, van Dam M, Yu L, Brewer C, Collins JA, Molhuizen HO, Loubser O, Ouelette BF, Fichter K, Ashbourne-Excoffon KJ, Sensen CW, Scherer S, Mott S, Denis M, Martindale D, Frohlich J, Morgan K, Koop B, Pimstone S, Kastelein JJ, Genest J Jr, Hayden MR. Mutations in ABC1 in Tangier disease and familial high-density lipoprotein deficiency. Nat Genet. 1999; 22: 336-345.

Rust S, Rosier M, Funke H, Real J, Amoura Z, Piette JC, Deleuze JF, Brewer HB, Duverger N, Denefle P, Assmann G. Tangier disease is caused by mutations in the gene encoding ATP-binding cassette transporter 1. Nat Genet. 1999; 22: 352-355.

Wang N, Lan D, Chen W, Matsuura F, Tall AR. ATP-binding cassette transporters G1 and G4 mediate cellular cholesterol efflux to high-density lipoproteins. Proc Natl Acad Sci U S A. 2004; 101: 9774-9779.

Assmann G, von Eckardstein A, Funke H. High density lipoproteins, reverse transport of cholesterol, and coronary artery disease. Insights from mutations. Circulation. 1993; 87: III28-III34.

Kuivenhoven JA, Pritchard H, Hill J, Frohlich J, Assmann G, Kastelein J. The molecular pathology of lecithin: cholesterol acyltransferase (LCAT) deficiency syndromes. J Lipid Res. 1997; 38: 191-205.

Cohen JC, Wang Z, Grundy SM, Stoesz MR, Guerra R. Variation at the hepatic lipase and apolipoprotein AI/CIII/AIV loci is a major cause of genetically determined variation in plasma HDL cholesterol levels. J Clin Invest. 1994; 94: 2377-2384.

Davignon J, Genest J Jr. Genetics of lipoprotein disorders. Endocrinol Metab Clin North Am. 1998; 27: 521-550.

Almasy L, Hixson JE, Rainwater DL, Cole S, Williams JT, Mahaney MC, VandeBerg JL, Stern MP, MacCluer JW, Blangero J. Human pedigree-based quantitative-trait-locus mapping: localization of two genes influencing HDL-cholesterol metabolism. Am J Hum Genet. 1999; 64: 1686-1693.

Soro A, Pajukanta P, Lilja HE, Ylitalo K, Hiekkalinna T, Perola M, Cantor RM, Viikari JS, Taskinen MR, Peltonen L. Genome scans provide evidence for low-HDL-C loci on chromosomes 8q23, 16q24.1-24.2, and 20q13.11 in Finnish families. Am J Hum Genet. 2002; 70: 1333-1340.

Arya R, Duggirala R, Almasy L, Rainwater DL, Mahaney MC, Cole S, Dyer TD, Williams K, Leach RJ, Hixson JE, MacCluer JW, O?Connell P, Stern MP, Blangero J. Linkage of high-density lipoprotein-cholesterol concentrations to a locus on chromosome 9p in Mexican Americans. Nat Genet. 2002; 30: 102-105.

Kort EN, Ballinger DG, Ding W, Hunt SC, Bowen BR, Abkevich V, Bulka K, Campbell B, Capener C, Gutin A, Harshman K, McDermott M, Thorne T, Wang H, Wardell B, Wong J, Hopkins PN, Skolnick M, Samuels M. Evidence of linkage of familial hypoalphalipoproteinemia to a novel locus on chromosome 11q23. Am J Hum Genet. 2000; 66: 1845-1856.

Pajukanta P, Allayee H, Krass KL, Kuraishy A, Soro A, Lilja HE, Mar R, Taskinen MR, Nuotio I, Laakso M, Rotter JI, de Bruin TW, Cantor RM, Lusis AJ, Peltonen L. Combined analysis of genome scans of Dutch and Finnish families reveals a susceptibility locus for high-density lipoprotein cholesterol on chromosome 16q. Am J Hum Genet. 2003; 72: 903-917.

NIH, Washington DC. Lipid Research Clinics population studies databook. [Book, volume, 1], 28-81, 1980. Department of health and human services, Public Health Service, Washington, DC. NIH publication 80-1527.

Marcil M, Yu L, Krimbou L, Boucher B, Oram JF, Cohn JS, Genest J Jr. Cellular cholesterol transport and efflux in fibroblasts are abnormal in subjects with familial HDL deficiency. Arterioscler Thromb Vasc Biol. 1999; 19: 159-169.

Marcil M, Bissonnette R, Vincent J, Krimbou L, Genest J Jr. Cellular phospholipid and cholesterol efflux in high-density lipoprotein deficiency. Circulation. 2003; 107: 1366-1371.

Lange K, Weeks D, Boehnke M. Programs for pedigree analysis: MENDEL, FISHER, and dGENE. Genet Epidemiol. 1988; 5: 471-472.

Schaid DJ, McDonnell SK, Thibodeau SN. Regression models for linkage heterogeneity applied to familial prostate cancer. Am J Hum Genet. 2001; 68: 1189-1196.

Sobel E, Lange K. Descent graphs in pedigree analysis: applications to haplotyping, location scores, and marker-sharing statistics. Am J Hum Genet. 1996; 58: 1323-1337.

Myerowitz R, Hogikyan ND. Different mutations in Ashkenazi Jews and non-Jewish French Canadians with Tay-Sachs disease. Science. 1986; 232: 1646-1648.

Elbein SC, Hasstedt SJ. Quantitative trait linkage analysis of lipid-related traits in familial type 2 diabetes: evidence for linkage of triglyceride levels to chromosome 19q. Diabetes. 2002; 51: 528-535.

Lilja HE, Suviolahti E, Soro-Paavonen A, Hiekkalinna T, Day A, Lange K, Sobel E, Taskinen MR, Peltonen L, Perola M, Pajukanta P. Locus for quantitative HDL-cholesterol on chromosome 10q in Finnish families with dyslipidemia. J Lipid Res. 2004; 45: 1876-1884.

作者单位：Departments of Medicine and Human Genetics (Z.D., J.C.E., J.G.), McGill University, Montreal, Canada; Department of Human Genetics (L.Q., C.P., P.P.), David Geffen School of Medicine, University of California, Los Angeles; Division of Cardiology (M.M.), McGill University, Montreal, Canada.