Familial combined hyperlipidemia (FCHL) is the most

Transcription

1 Linkage of a Candidate Gene Locus to Familial Combined Hyperlipidemia Lecithin:Cholesterol Acyltransferase on 16q Bradley E. Aouizerat, Hooman Allayee, Rita M. Cantor, Geesje M. Dallinga-Thie, Christopher D. Lanning, Tjerk W.A. de Bruin, Aldons J. Lusis, Jerome I. Rotter Abstract Familial combined hyperlipidemia (FCHL) is a common lipid disorder characterized by elevated levels of plasma cholesterol and triglycerides that is present in 10% to 20% of patients with premature coronary artery disease. To study the pathophysiological basis and genetics of FCHL, we previously reported recruitment of 18 large families. We now report linkage studies of 14 candidate genes selected for their potential involvement in the aspects of lipid and lipoprotein metabolism that are altered in FCHL. We used highly polymorphic markers linked to the candidate genes, and these markers were analyzed using several complementary, nonparametric statistical allele-sharing linkage methodologies. This current sample has been extended over the one in which we identified an association with the apolipoprotein (apo) AI-CIII-AIV gene cluster. We observed evidence for linkage of this region and FCHL (P 0.001), providing additional support for its involvement in FCHL. We also identified a new locus showing significant evidence of linkage to the disorder: the lecithin:cholesterol acyltransferase (LCAT) locus (P ) on chromosome 16. In addition, analysis of the manganese superoxide dismutase locus on chromosome 6 revealed a suggestive linkage result in this sample (P 0.006). Quantitative traits related to FCHL also provided some evidence of linkage to these regions. No evidence of linkage to the lipoprotein lipase gene, the microsomal triglyceride transfer protein gene, or several other genes involved in lipid metabolism was observed. The data suggest that the lecithin:cholesterol acyltransferase and apolipoprotein AI-CIII-AIV loci may act as modifying genes contributing to the expression of FCHL. (Arterioscler Thromb Vasc Biol. 1999;19: ) Key Words: familial combined hyperlipidemia lipid metabolism genetics lecithin:cholesterol acyltransferase Familial combined hyperlipidemia (FCHL) is the most common genetic lipid disorder observed in patients with coronary artery disease and their relatives, with a frequency of 1% in the US white and European populations. FCHL occurs in 10% to 20% of patients with premature coronary artery disease. 1 9 The disorder is characterized by elevated concentrations of plasma triglycerides and/or plasma cholesterol in probands and their affected relatives. Moreover, affected relatives exhibit multiple lipoprotein phenotypes, a predisposition to premature coronary artery disease, and a vertical pattern of transmission of hyperlipidemia. 1 Early studies suggested that the familial aggregation of FCHL was consistent with a dominant, mendelian trait, 1,8 9 but subsequent studies have suggested a more complex mode of inheritance. 5,7,9 12 One characteristic of FCHL is the variability in lipoprotein profile, referred to as multiple lipoprotein phenotype, seen in affected relatives; thus, affected individuals may exhibit a Fredrickson type IIa, IIb, or IV hyperlipidemia. 1,13 The fully expressed FCHL phenotype involves elevated levels of LDL, IDL, and VLDL. These particles have a precursor-product relationship, as the liver secretes VLDL, which then undergoes lipolysis to IDL and subsequently to LDL. The principal structural protein of these particles is apolipoprotein B (apob), an integral membrane protein required for VLDL packaging. Elevated levels of apob are a useful indicator of FCHL. 13 Other traits frequently associated with FCHL include small, dense LDL particles, low HDL levels, insulin resistance, and elevated postprandial levels of free fatty acids. 3,12,14 17 The molecular mechanisms contributing to FCHL are largely unknown. Because hypertriglyceridemia is often the earliest manifestation of FCHL, it has been proposed that the primary defect(s) involves triglyceride metabolism, with secondary effects on cholesterol metabolism. 1,18 Physiologi- Received November 19, 1998; revision accepted March 24, From the Department of Microbiology and Molecular Genetics, the Department of Medicine, and the Molecular Biology Institute (B.E.A., H.A., A.J.L.), University of California, Los Angeles; the Division of Medical Genetics, Departments of Medicine and Pediatrics and Steven Spielberg Pediatric Research Center (R.M.C., C.D.L., J.I.R.), Cedars-Sinai Research Institute, Los Angeles, Calif; the Department of Medicine, University Hospital (G.M.D.-T.), Utrecht, the Netherlands; and the Departments of Medicine and Endocrinology (T.W.A.d.B.), University Hospital, Maastricht, the Netherlands. Correspondence to Aldons J. Lusis, PhD, Department of Medicine, Division of Cardiology, CHS, UCLA, Los Angeles, CA jlusis@medicine.medsch.ucla.edu 1999 American Heart Association, Inc. Arterioscler Thromb Vasc Biol. is available at

2 Aouizerat et al Candidate Gene Locus Linked to FCHL 2731 cal studies have suggested that patients with FCHL have increased secretion of VLDL triglycerides and associated apob, raising the possibility that this is the primary metabolic defect. There is also substantial evidence that the FCHL phenotype results in part from defects in lipoprotein catabolism. In particular, about one third of FCHL patients exhibit decreased levels of lipoprotein lipase (LPL), the enzyme primarily responsible for lipolysis of chylomicrons and VLDL, 4,24,25 although only a small fraction of this decrease in LPL levels appears to be due to mutations of the LPL gene Furthermore, there is impaired removal of chylomicron remnants after an oral fat load, and this is associated with increased plasma apociii concentrations. 5,31,32 We and others have observed significant evidence for linkage (and association) between specific alleles of the apoai-ciii-aiv gene cluster and FCHL A compelling hypothesis for FCHL is that it involves 1 or more primary gene defects affecting VLDL synthesis. These defects then act in combination with various secondary variations, most likely those affecting lipoprotein catabolism, such as the decrease in LPL enzyme activity or variations at the apoai- CIII-AIV gene cluster. 13,32,35 This can explain the variable expression of the FCHL phenotype among family members and the complex mode of inheritance. A relatively small number of candidate genes for FCHL, such as apob, LPL, and the apoai-ciii-aiv cluster, have been previously examined, mostly by association studies. The results suggest that apob gene variations do not contribute significantly to the trait 36,37 and that the LPL and apoai-ciii- AIV genes are likely to play a role as secondary or modifier genes ,32,33,36,38 To further identify potential primary and secondary defects contributing to the expression of FCHL, we studied 14 separate candidate gene loci in 18 wellcharacterized FCHL families. Linkage results reveal a new locus that appears to contribute to the expression of this disorder, the lecithin:cholesterol acyltransferase (LCAT) gene locus on chromosome 16. Methods Ascertainment of FCHL Study Population Eighteen unrelated FCHL probands were recruited through the Lipid Clinic of the Utrecht University Hospital. The probands met the following minimum criteria: (1) a primary hyperlipidemia with varying phenotypic expression: plasma cholesterol 250 mg/dl ( 6.5 mmol/l) and/or triglycerides 200 mg/dl ( 2.3 mmol/l), in addition to apob concentrations exceeding the mean 2 SDs for age-adjusted levels; (2) at least 1 first-degree relative with a different hyperlipidemic phenotype; (3) a positive family history of premature coronary heart disease defined as myocardial infarction or cardiovascular disease before 60 years of age; and (4) absence of xanthomas. Exclusion criteria included diabetes, obesity, familial hypercholesterolemia (absence of tendon xanthomas), or type III hyperlipidemia (absence of apoe2/e2). All subjects gave informed consent, and the Human Investigation Review Committee of the University Hospital, Utrecht, approved the study protocol. Families were ascertained as previously described. 5,31 33 The clinical and biochemical characteristics of the study population (n 481), which comprised 18 probands, their hyperlipidemic and normolipidemic relatives, and spouses, are summarized in Table 1. Over 95% of the relatives and spouses of probands were collected without regard to their FCHL status. In the present study, DNA samples from 481 subjects of the original data set (n 582) 32,33 were available for molecular and linkage analysis. Hyperlipidemic family members (n 151) were assigned the FCHL phenotype based on plasma TABLE 1. Clinical Characteristics of the FCHL Families Variable Hyperlipidemic Individuals* Normolipidemic Individuals Spouse Controls Number Age, y Cholesterol, mg/dl LDL-cholesterol, mg/dl Triglycerides, mg/dl ApoB, mg/dl Values are expressed as mean SD. *The hyperlipidemic group includes the 18 probands and 133 hyperlipidemic relatives (n 151). cholesterol 250 mg/dl and/or plasma triglycerides 200 mg/dl and/or apob 75th percentile for age-matched controls. 15 Plasma apob concentrations were used as an additional marker to identify affected FCHL subjects, complementing the original definition by Goldstein et al, 1 who used 90th percentile cholesterol and 90th percentile triglyceride as cutoff points. Plasma apob concentrations reflect the combined effects of high hepatic apob secretion and impaired removal of apob-containing lipoproteins in FCHL. 17 Normolipidemic family members (n 176) exhibited plasma cholesterol 250 mg/dl, plasma triglycerides 200 mg/dl, and apob levels 75th percentile matched for age and sex. Additionally, normolipidemic relatives were younger than hyperlipidemic relatives and spouses. It is possible that a subset of these normolipidemic individuals have yet to express the FCHL phenotype. As previously reported, hyperlipidemic family members had higher plasma cholesterol, triglycerides, LDL cholesterol, and apob concentrations than did normolipidemic relatives and spouses. Genetic Markers for Candidate Genes Genetic markers appropriate for testing specific candidate loci have been previously reported for a number of the genes studied herein These include the apoai-ciii-aiv gene cluster, apoaii, apob, apoe, carboxyl ester lipase (CEL), cholesteryl ester transfer protein (CETP), hepatic lipase (HL), LCAT, LDL receptor (LDLR), LPL, manganese superoxide dismutase (MnSOD), microsomal triglyceride transfer protein (MTP), peroxisome proliferator-activated receptor (PPAR ), and the VLDL receptor (VLDLR) (Table 2). Four candidate loci (CEL, CETP, HL, and LCAT) had not previously been mapped in detail; in these cases, polymorphic genetic markers were selected by employing a radiation hybrid mapping strategy. 42,48,49,51 The genetic markers were then genotyped in 481 individuals of the 18 FCHL kindreds. DNA was isolated from EDTA-augmented blood by following standard procedures, and polymerase chain reaction amplification was conducted according to recommended protocols. 41 ApoE genotypes were determined as described previously. 17 Linkage Methods To assess the cosegregation of the FCHL phenotypes with genetic markers at the 14 candidate loci, we used several nonparametric linkage tests, ie, tests for which the mode of inheritance of the disease does not have to be hypothesized. 52,53 At each candidate gene locus, the markers were tested for linkage individually. Initially, this was done using only affected sibling pairs. Positive linkage was followed by a linkage analysis of the data from all available sibling pairs in the sample. When significant evidence of linkage was observed for the individual marker, multiple markers in the same chromosomal region were typed and combined into haplotypes to maximize the linkage information. The advantage of analyzing only affected sibling pairs for a disease with complex inheritance, such as FCHL, is that each analyzed individual is assumed to have 1 or more of the disease genes, and the results are not confounded by reduced penetrance. However, because multiple genes may contribute to a complex disorder such as FCHL, different combinations of such genes may

3 2732 Arterioscler Thromb Vasc Biol. November 1999 TABLE 2. Candidate Genes and Polymorphic Markers Tested for Linkage With FCHL Candidate Gene Marker(s) Genetic Distance, cm Heterozygosity Chromosome Reference ApoAI-CIII-AIV HC3 0* ApoAII AII 0* ApoB D2S ApoE E 2 /E 3 /E 4 0* CEL D9S CETP D16S HL D15S LCAT D16S LDLR LDLR 0* LPL LPL3 0* MnSOD D6S MTP MTP 0* PPAR D22S VLDLR VLDLR 0* Markers for CEL, CETP, LCAT, and MnSOD were ascertained through radiation hybrid mapping(42,48 50). Previously reported microsatellites were used for the remainder of the candidate genes. Estimated genetic distances in centimorgans (cm) between polymorphic microsatellite markers and candidate genes are listed. *Marker is located intragenic to the candidate gene. lead to clinical disease. This reduces the power of any linkage analysis, since in such a case not all affected pairs will demonstrate allele sharing at any given disease gene locus. Nevertheless, if any given locus contributes to the disease in a large proportion of the cases, affected sib pairs will share that gene in a greater proportion than expected by chance. Any 2 random sib pairs are expected to share a given allele 50% of the time. If there is linkage of the disease to a specific locus, then affected sib pairs are expected to exhibit a mean allele sharing significantly 0.5. We calculated the mean allele sharing at each locus of interest and tested each for a significant difference from the expected value of 0.5 by using the SIBPAL subprogram of the SAGE package. 54 When a candidate gene showed evidence for linkage, we extended the analyses to include 2 additional groups of sib pairs, 1 in which both sibs were unaffected (clinically concordant unaffected sibling pairs) and another in which 1 was affected and the other was not (clinically discordant sibling pairs). At linked markers, both clinically concordant affected and clinically concordant unaffected sib pairs are expected to demonstrate an increased sharing of marker alleles identical-by-descent, while clinically discordant pairs should exhibit decreased sharing below the expected value of Clinically concordant unaffected and clinically discordant sib pairs may provide weaker evidence for linkage than affected sib pairs, as some unaffected siblings may carry the disease gene but not express it owing to the lack of other genetic and/or environmental factors that contribute to the development of the disease (ie, reduced penetrance). Haplotype information can improve the significance level of a true linkage result by increasing the informativeness over that of single markers. Often, haplotyping allows the number of alleles shared identical-by-descent at a given marker to be assessed unambiguously. Because the result for LCAT was new and significant by our criterion, we typed an additional 3 markers and tested the haplotype for linkage to FCHL. To incorporate information from all pairs, we used the Haseman-Elston 52 linear regression method as programmed in the SIBPAL subroutine of SAGE. For each sibling pair, the squared difference in the disease status (affected pairs are coded as 1 and unaffected pairs are coded as zero) is regressed against their estimated proportion of haplotypes shared identical-by-descent. A significant negative regression, representing a greater proportion of shared haplotypes for those who are clinically concordant, is taken as evidence for linkage. In addition to FCHL, 4 related quantitative traits were assessed for linkage to the 14 candidate genes. The squared trait differences between sib pairs for levels of total plasma triglyceride, cholesterol, apob, and apociii were regressed against the number of marker alleles that they share identical-by-descent for each locus A significantly negative regression line was taken as evidence for linkage. We performed all quantitative analyses initially on the untransformed data; logarithmically transformed data were analyzed when the quantitative trait distribution was nonnormal. Because only candidate genes were tested in these analyses, a significance level of was chosen as the criterion for indicating evidence for linkage. Results Fourteen candidate loci were tested for linkage to FCHL in 18 FCHL families consisting of 481 individuals. The genes selected have been implicated in VLDL synthesis and secretion (apob, MTP), lipolysis (CEL, HL, LPL), lipoprotein processing and remodeling (apoai-ciii-aiv, CETP, LPL, HL, LCAT), HDL metabolism (apoaii, HL, LCAT, CETP, LPL), lipoprotein removal (apoe, LDLR, VLDLR), and cellular metabolism (PPAR, MnSOD). 4,25,56 64 Highly informative genetic markers (Table 2) either at or near the candidate loci were chosen on the basis of previous studies or after physical mapping of the candidate genes by radiation hybrid analysis (see Methods). By using a nonparametric sib-pair linkage methodology, 11 of the candidate gene loci did not exhibit significant evidence of linkage to the qualitative FCHL trait (P 0.05, Table 3). Although not meeting the formal statistical threshold of P 0.001, the MnSOD locus on chromosome 6 did yield suggestive evidence of linkage (P 0.006), with a mean allele sharing of 0.57 (data not shown). Two candidate loci, the apoai-ciii-aiv locus (chromosome 11) and the LCAT locus (chromosome 16), exhibited significant evidence of linkage (Table 3). Mean allele sharing in affected sibling pairs at the markers linked to apociii and LCAT was 0.62 and 0.59, respectively, signifi-

4 Aouizerat et al Candidate Gene Locus Linked to FCHL 2733 TABLE 3. Nonparameteric Sib-Pair Linkage Analyses for the Qualitative FCHL Trait: Proportions of Alleles Shared in Affected Sib Pairs Candidate Gene Number of Affected Sib Pairs Mean Allele Sharing ApoAII ApoB ApoCIII ApoE CEL CETP HL LCAT LDLR LPL MnSOD MTP PPAR VLDLR In the analysis, the presence of FCHL in a family member was defined as plasma cholesterol levels 250 mg/dl and/or plasma triglyceride levels 200 mg/dl and/or apob 75th percentile. cantly different from the expected value of 0.50 when there is no linkage to the region. We then tested the FCHL-associated traits of triglycerides, total cholesterol, apob, and apociii levels for linkage to the LCAT, apoai-ciii-aiv, and MnSOD chromosomal regions. Two of these markers demonstrated possible linkage to the constellation of FCHL-related traits: the LCAT locus on chromosome 16 was linked to the traits of cholesterol (P 0.03) and apociii levels (P 0.03), and the MnSOD locus on chromosome 6 was linked to the traits of cholesterol (P 0.02), apob levels (P 0.02), and apociii levels (P 0.03). To further characterize the LCAT region, 3 additional markers were typed, D16S514, D16S400, and D16S408, spanning, in addition to the original marker, a 12-cM distance. These markers were combined into a haplotype, thus allowing assessment of their inheritance as a group. As shown in the Figure, the allele sharing for the haplotype exhibited a Mean allele sharing of the 16q haplotype among sib pairs from 18 kindreds. Mean allele sharing proportions between concordant affected, discordant (where 1 sibling is affected and 1 is not), and concordant unaffected sib pairs are compared. These analyses included n 78, 57, and 46 for the 3 groups of sib pairs, respectively. Values for allele sharing are indicated in closed diamonds ( ) for the 16q haplotype and with a minus ( ) for no linkage (the null hypothesis). P pattern of excess sharing for both affected and unaffected concordant pairs as well as reduced sharing in clinically discordant pairs, providing additional support for linkage of the LCAT gene region to FCHL. Discordant disease status (1 affected and 1 unaffected) in the sib pair should result in decreased allele sharing, while the same disease status (either both affected or both unaffected) should result in increased sharing. The proportion of haplotypes shared identical-bydescent by affected pairs was 0.60, and the linear regression linkage analysis yielded a P value of Discussion The molecular and genetic mechanisms contributing to the pathogenesis and expression of FCHL have remained largely unknown since the original delineation of the disorder by Goldstein and colleagues. 1 The analyses presented herein have revealed a new locus showing significant evidence of linkage to FCHL: the LCAT locus on chromosome 16. In addition, the previously reported linkage of the apoai-ciii- AIV gene cluster with FCHL was supported in an extended sample by using a highly polymorphic genetic marker. The MnSOD locus on chromosome 6 gave possible evidence for linkage to FCHL as well. Several analyses of the data provide linkage evidence for the LCAT locus: first, an increased allele sharing in affected sib pairs; second, a pattern of decreased sharing in discordant sib pairs and increased sharing in unaffected sib pairs expected with linkage; and third, additional support when additional markers and sibling pairs were included. In addition, consistent with the findings of several other studies, our data provide no support for linkage to 5 distinct loci, the apoaii, apob, HL, LDLR, and LPL loci on chromosomes 1, 2, 15, 19, and 8, respectively. Furthermore, the MTP, VLDLR, and CETP loci on chromosomes 4, 9, and 16, respectively, do not appear to play an important role in the expression of FCHL in this population. Recently, Porkaa et al 13 reported that the 90th percentile cutoff values in the Finnish population for cholesterol and/or triglycerides combined represented the optimal criteria for the diagnosis of FCHL. This finding is in agreement with the original analysis presented by Goldstein et al 1 and with our own approach to FCHL phenotyping in Dutch whites. 32 We recognize, however, that the current approach to diagnosis should be considered as interim, pending identification of the predisposing variations in the actual susceptibility genes. Lecithin:Cholesterol Acyltransferase This enzyme is synthesized in the liver and circulates in the plasma as a component of HDL particles. The LCAT/CETP locus (the 2 genes are roughly 12 cm apart on chromosome 16) was shown previously to exhibit significant linkage with LDL particle size. 15,58,65 Our data suggest that LCAT rather than CETP is the primary locus responsible for the observed linkage, based on the markers spanning the region. In addition, there is biological evidence supporting such a conclusion. Subjects with severe combined hyperlipidemia show significantly increased cholesterol ester mass transfer from HDL to VLDL but have reduced or even absent net mass cholesterol ester transfer from HDL to LDL. 65 This phenomenon, which potentially contributes to small, dense LDL particle formation, is caused by the low affinity of CETP protein for the LDL from combined hyperlipidemia

5 2734 Arterioscler Thromb Vasc Biol. November 1999 subjects as well as increased concentrations of the most active cholesterol ester acceptor, VLDL. Guerin et al 66 concluded that the biochemical characteristics of HDL and VLDL lipoproteins in combined-hyperlipidemia subjects were sufficient explanation for the altered composition rather than the plasma protein concentration of CETP itself. Several lines of biological evidence support this linkage as well. Transgenic rabbits overexpressing LCAT exhibit elevated levels of VLDL and LDL, similar to the phenotype in FCHL. 67 In addition, fish-eye disease, a familial syndrome of LCAT deficiency, results in increased VLDL levels and hypertriglyceridemia, among other phenotypic abnormalities. 49 Thus, LCAT may act as a modifier of the FCHL phenotype. Alternatively, a gene closely linked to LCAT must be considered. Whether the LCAT gene is in fact responsible for the linkage requires further testing in population or familybased association (linkage disequilibrium) studies. Apolipoprotein AI-CIII-AIV Previous association and linkage studies have implicated the apoai-ciii-aiv gene cluster in FCHL Recently, we carried out an association study confirming the involvement of the apoai-ciii-aiv gene cluster in FCHL. 32 Three restriction fragment length polymorphisms revealed that a minor allele was associated with elevated plasma cholesterol, triglycerides, LDL cholesterol, apob, and apociii levels in FCHL probands and hyperlipidemic relatives versus controls. 33 In addition, nonparametric sib-pair linkage analysis revealed significant evidence of linkage between these restriction fragment length polymorphisms in the gene cluster and the FCHL phenotype. 32 Moreover, demonstration of 2 different susceptibility haplotypes in the gene cluster revealed a paradigm of complex genetic contribution to FCHL. 32 A similar pattern was observed in a recent study of Finnish FCHL families, although the results did not attain statistical significance in that sample. 68 The data reported here continue to support the previous linkage finding in our sample by using a more informative genetic marker (a tetranucleotide repeat within the apociii gene). Although there are fewer sib pairs available for analysis in this study compared with our previous study, 32 the stronger linkage result with the gene cluster is likely due to the dramatic increase in the heterozygosity index of the microsatellite marker compared with the restriction fragment length polymorphism haplotypes (0.95 and 0.41, respectively) described previously. 15 This finding is complemented by biochemical studies. Expression of the apociii gene has been shown to be a determinant of plasma VLDL triglyceride levels in mice. 69 In humans, the apociii gene affects plasma apociii concentrations, and the elevated apociii concentrations in FCHL subjects predict impaired postprandial lipemia. 31 Thus, the involvement of the apociii gene in FCHL expression is not only evident from the linkage results but biologically plausible as well. 15 Manganese Superoxide Dismutase Although the results of our candidate gene analysis with markers linked to the MnSOD gene did not reach the criterion set for linkage, positive linkage results with related phenotypes in other samples suggest that this gene locus may have a role in FCHL. Previously, we reported significant linkage of the MnSOD locus with LDL particle size in a set of families enriched for coronary artery disease but without marked hyperlipidemia. 58 We have also observed linkage of LDL particle size to this locus in FCHL families 15 and in a sample of subjects undergoing dietary intervention. 65 In the present study, the MnSOD gene was examined because the small, dense LDL trait is commonly associated with FCHL, and this gene has been correspondingly linked to LDL particle size. 12,14,15 The mechanisms by which MnSOD, a mitochondrial enzyme responsible for the dismutation of superoxides generated as a by-product of cellular respiration, may influence lipid metabolism are yet unclear. Genetic inactivation of the mitochondrial form of superoxide dismutase in mice results in abundant hepatic lipid accumulation, dilated cardiomyopathy, and early neonatal death. 70 MnSOD has been shown to influence cellular responses to tumor necrosis factor-, which is a potent modulator of LPL expression. 64 Reduced LPL activity is a frequent finding in FCHL patients. 4,71 It is also conceivable that an increased incorporation of oxidized lipids, which accumulate intracellularly as a consequence of reduced clearance of superoxides by MnSOD, may affect the catabolism of VLDL. Evidence for an Oligogenic Model Our current understanding of FCHL is that it represents a complex metabolic syndrome, characterized by hepatic and possibly intestinal hypersecretion of lipoproteins, which in turn generates increased flux through the lipolytic cascade and receptor-mediated uptake routes. Therefore, each of the pathways involved is vulnerable to gene modifications that reduce the maximal functional capacity of the pathway under conditions of high lipoprotein flux caused by hepatic hypersecretion. In our view, major gene effects are expected to influence the regulation of hepatic lipoprotein secretion or the level of substrate supply to lipoprotein-secreting organs. Such major gene effects are expected to be present in all or the vast majority of FCHL subjects. Modifier gene effects can, at least in theory, be expected at the level of lipoprotein secretion, lipolysis, processing, remodeling, or cellular uptake, provided that the gene(s) involved plays a functional role in lipoprotein metabolism in FCHL. It is unlikely that LCAT is the primary determinant of FCHL because it would not be expected to influence VLDL assembly or secretion. Several kinetic studies have revealed a 2-fold increase in VLDL production, implicating a defect in VLDL synthesis as the primary cause In addition, modifier loci, which may not significantly influence plasma lipid levels under normal circumstances, are likely to affect lipid levels under conditions of increased flux of VLDL in FCHL. The apoai-ciii-aiv gene cluster and, less frequently, the LPL locus appear to be 2 such modifier loci, as identified previously. 30,32 ApoAI-CIII-AIV and LCAT are presently considered response modifier genes. Although the present study has been performed on 1 of the largest collections of affected FCHL individuals, our ability to detect modifier gene effects was limited to those with at least an estimated 15% contribution (see below). Modifier genes are expected to be common in FCHL families, and combinations of modifier genes may prove to be frequent as well. Various combinations of modifier genes could have an impact on the expression of the multiple type hyperlipidemia, characteristic of FCHL.

6 Aouizerat et al Candidate Gene Locus Linked to FCHL 2735 Evidence from these analyses indicates that multiple genes contribute to the expression of FCHL. The FCHL genes, which can be detected by these linkage analyses, can be characterized by reference to the parameter known as lambda. 72 Lambda represents the FCHL risk to the siblings of an affected individual compared with that of the general population. For FCHL, lambda may be as large as 25, calculated by dividing the nearly 50% risk to sibs by the 1% to 2% risk to the general population. Hauser et al 73 have provided tables that can be used to estimate the minimum lambda of genes that can be identified in an affected sib-pair linkage analysis wherein the numbers of nuclear families and degrees of parental genotyping vary. Using P as the level of significance (a log of the odds score of 2.0) and this sample size of 100 affected sib pairs, Table 2 of their article indicates we should be able to identify an FCHL gene with a lambda of 2.0 or larger 80% of the time. By using a multiplicative model for the genetic contribution to FCHL, such a gene could have a contribution of as little as 15% to this disorder. In conclusion, we provide evidence herein, by linkage analysis in FCHL families, that 2 distinct loci, the apoai-ciii-aiv and LCAT loci on chromosomes 11 and 16, respectively, may contribute to the expression of FCHL. These results illustrate the genetic complexity of this disorder and contribute to our understanding of this most common dyslipidemia predisposing to premature coronary artery disease. Acknowledgments This study was supported by National Institutes of Health grant HL (to A.J.L. and J.I.R.), by a grant from the Dutch Heart Foundation (D ) (to T.W.A.d.B.), and by the Cedars-Sinai Board of Governors Chair in Medical Genetics (to J.I.R.). The results of the SIBPAL analysis were obtained by using the program package SAGE and were supported by US Public Health Services Resource grant P41 RR03655 from the Division of Resources. T.W.A.d.B. is a recipient of the PIONEER research grant of the Dutch Organization of Scientific Research (NWO). We thank Dr Richard Gregg, Bristol-Myers Squibb Co, for supplying the polymorphic genetic marker for MTP. We are grateful to Dr John Kane, University of California, San Francisco, for discussions and for suggesting PPAR as a candidate gene. We thank Margreet van Linde-Sibenius for excellent technical assistance. In addition, we would also like to thank the patients, relatives, and spouses for participating in this study. References 1. Goldstein JL, Schrott HG, Hazzard WR, Bierman EL, Motulsky AG. Hyperlipidemia in coronary artery disease, II: genetic analysis of lipid levels in 176 families and delineation of a new inherited disorder, combined hyperlipidemia. J Clin Invest. 1973;52: Rose HG, Kranz P, Weinstock M, Juliano J, Haft JI. Inheritance of combined hyperlipoproteinemia: evidence for a new lipoprotein phenotype. Am J Med. 1973;54: Brunzell JD, Albers JJ, Chait A, Grundy SM, Groszek E, McDonald GB. Plasma lipoproteins in familial combined hyperlipidemia and monogenic familial hypertriglyceridemia. J Lipid Res. 1983;24: Babirak SP, Brown G, Brunzell JD. Familial combined hyperlipidemia and abnormal lipoprotein lipase. Arterioscler Thromb. 1992;12: Castro Cabezas M, de Bruin TWA, Erkelens DW. Familial combined hyperlipidemia: Neth J Med. 1992;40: Kwiterovich PO. Genetic and molecular biology of familial combined hyperlipidemia. Curr Opin Lipidol. 1993;4: Lusis AJ, Ivandic B. Lipid metabolism. In: Rimoin DL, Conner JM, Pyeriz RE, eds. Principles and Practice of Medical Genetics. New York: Churchill Livingston; 1997: Nikkila EA, Aro A. Family study of serum lipids and lipoproteins in coronary heart disease. Lancet. 1973;1: Glueck CJ, Fallat R, Buncher CR, Tsang R, Steiner P. Familial combined hyperlipoproteinemia: studies in 91 adults and 95 children from 33 kindred. Metabolism. 1973;23: Williams WR, Lalouel JM. Complex segregation analysis of hyperlipidemia in a Seattle sample. Hum Hered. 1982;32: Cullen P, Farren B, Scott J, Farrall M. Complex segregation analysis provides evidence for a major gene acting on serum triglyceride levels in 55 British families with familial combined hyperlipidemia. Arterioscler Thromb. 1994;14: Jarvik GP, Brunzell JD, Austin MA, Krauss RM, Motulsky AG, Wijsman E. Genetic predictors of FCHL in four large pedigrees: influence of apob level major locus predicted genotype and LDL subclass phenotype. Arterioscler Thromb. 1994;14: Porkaa KVK, Nuotio I, Pajukanta P, Ehnholm C, Suurinkeroinen L, Syvanen M, Lehtimaki T, Lahdenkari AT, Lahdenpera S, Ylitalo K, Antikainen M, Perola M, Raitakari OT, Kovanen P, Viikari JSA, Peltonen L, Taskinen MR. Phenotype expression in familial combined hyperlipidemia. Atherosclerosis. 1997;133: Bredie SJH, Kiemeney LA, de Haan AFJ, Demacker PMN, Stalenhoef AFH. Inherited susceptibility determines the distribution of dense lowdensity lipoprotein subfraction profiles in familial combined hyperlipidemia. Am J Hum Genet. 1996;58: Allayee H, Aouizerat BE, Cantor RM, Dallinga-Thie GM, Krauss RM, Lanning CD, Rotter JI, Lusis AJ, de Bruin TWA. Families with familial combined hyperlipidemia and families enriched for coronary artery disease share genetic determinants for the atherogenic lipoprotein phenotype. Am J Hum Genet. 1998;63: Aitman TJ, Godsland IF, Farren B, Crook D, Wong HJ, Scott J. Defects of insulin action on fatty acid and carbohydrate metabolism in familial combined hyperlipidemia. Arterioscler Thromb Vasc Biol. 1997;17: Castro Cabezas M, de Bruin TWA, de Valk HW, Shoulders CC, Jansen H, Erkelens DW. Impaired fatty acid metabolism in familial combined hyperlipidemia: a mechanism associating hepatic apolipoprotein B overproduction and insulin resistance. J Clin Invest. 1993;92: Grundy SM, Chait A, Brunzell JD. Familial combined hyperlipidemia workshop. Arteriosclerosis. 1987;7: Chait A, Albers JJ, Brunzell JD. Very low density lipoprotein overproduction in genetic forms of hypertriglyceridemia. Eur J Clin Invest. 1980;110: Janus ED, Nicoll AM, Turner PR, Magill P, Lewis B. Kinetic bases of the primary hyperlipidemias: studies of apolipoprotein B turnover in genetically defined subjects. Eur J Clin Invest. 1980;10: Kissebah AH, Alfarsi S, Adams PW. Integrated regulation of very low density lipoprotein triglyceride and apolipoprotein B kinetics in man: normolipidemic subjects, familial hypertriglyceridemia and familial combined hyperlipidemia. Metabolism. 1981;30: Cortner TA, Coates PM, Bennett MJ, Cryer DR, Le N-A. Familial combined hyperlipidemia: use of stable isotopes to demonstrate overproduction of very low density apolipoprotein B by the liver. J Inherit Metab Dis. 1991;14: Venkatesan S, Cullen P, Pacy P, Halliday D, Scott J. Stable isotopes show a direct relation between VLDL apob overproduction and serum triglyceride levels and indicate a metabolically and biochemically coherent basis for familial combined hyperlipidemia. Arterioscler Thromb. 1993;13: Nevin DN, Brunzell JD, Deeb SS. The LPL gene in individuals with familial combined hyperlipidemia and decreased LPL activity. Arterioscler Thromb. 1994;14: Gagne E, Genest J, Zang H, Clarke LA, Hayden MR. Analysis of DNA changes in the LPL gene in patients with familial combined hyperlipidemia. Arterioscler Thromb. 1994;14: Reymer PW, Groenemeyer BE, Gagne E, Miao L, Appelman EE, Seidel JC, Kromhout D, Bijvoet SM, van de Oever K, de Bruin TWA. A frequently occurring mutation in the lipoprotein lipase gene (Asn291Ser) contributes to the expression of familial combined hyperlipidemia. Hum Mol Genet. 1995;4: Deeb SS, Nevin DN, Iwasaki L, Brunzell JD. Two novel apolipoprotein A-IV variants in individuals with familial combined hyperlipidemia and diminished levels of lipoprotein lipase activity. Hum Mutat. 1996;8: Yang WS, Nevin DN, Peng R, Brunzell JD, Deeb SS. A mutation in the promoter of the lipoprotein lipase (LPL) gene in a patient with familial combined hyperlipidemia and low LPL activity. Proc Natl Acad Sci USA. 1995;92: Hoffer MJ, Bredie SJ, Boomsma DI, Reymer PW, Kastelein JJ, de Knijff P, Demacker PN, Stalenhoef AF, Havekes LM, Frants RR. The lipoprotein lipase (Asn291Ser) mutation is associated with elevated lipid levels in families with familial combined hyperlipidemia. Atherosclerosis. 1996;119: Mailly F, Tugrul Y, Reyme PW, Bruin T, Seed M, Groenemeyer BF, Asplund-Carlson A, Vallance D, Winder AF, Miller GJ. A common

7 2736 Arterioscler Thromb Vasc Biol. November 1999 variant in the gene for lipoprotein lipase (Asp9Asn): functional implications and prevalence in normal and hyperlipidemic subjects. Arterioscler Thromb Vasc Biol. 1995;15: Castro Cabezas M, de Bruin TWA, Jansen H, Kock LAW, Kortland W, Erkelens DW. Impaired chylomicron remnant clearance in familial combined hyperlipidemia. Arterioscler Thromb Vasc Biol. 1993;13: Dallinga-Thie GM, van Linde-Sibenius Trip M, Rotter JI, Cantor RM, Bu XD, Lusis AJ, de Bruin TWA. Complex genetic contribution of the apoai-ciii-aiv gene cluster to familial combined hyperlipidemia. J Clin Invest. 1997;99: Dallinga-Thie GM, Bu XD, van Linde-Sibenius Trip M, Rotter JI, Lusis AJ, de Bruin TWA. Apolipoprotein AI-CIII-AIV gene cluster in familial combined hyperlipidemia: effects on LDL-cholesterol and apolipoproteins B and C-III. J Lipid Res. 1996;37: Wojciechowski AP, Farrall M, Cullen P, Wilson TM, Bayliss JD, Farren B, Griffin BA, Caslake MJ, Packard CJ, Shepard J. Familial combined hyperlipidemia linked to the apolipoprotein AI-CIII-AIV gene cluster on chromosome 11q23 q24. Nature. 1991;349: de Bruin TWA, Mailly F, van Barlingen HJJ, Fisher R, Castro Cabezas M, Talmud P, Dallinga-Thie GM, Humphries SE. Lipoprotein lipase gene mutations D9N and N291S in four pedigrees with familial combined hyperlipidemia. Eur J Clin Invest. 1996;26: Coresh J, Beaty TH, Kwiterovich PO Jr, Antonarakis SE. Pedigree and sib-pair linkage analysis suggest the apolipoprotein B gene is not the major gene influencing plasma apolipoprotein B levels. Am J Hum Genet. 1992;50: Pajukanta P, Porkka KV, Antikainen M, Taskinen MR, Perola M, Murtomaki-Repo S, Ehnholm S, Nuotio I, Suurinkeroinen L, Lahdenkari AT, Syvanen AC, Viikari JS, Ehnholm C, Peltonen L. No evidence of linkage between familial combined hyperlipidemia and genes encoding lipolytic enzymes in Finnish families. Arterioscler Thromb Vasc Biol. 1997;17: Aguilar-Salinas CA, Hugh P, Barret R, Pulai J, Zhu XL, Schonfeld G. A familial combined hyperlipidemic kindred with impaired apolipoprotein B catabolism: kinetics of apolipoprotein B during placebo and pravastatin therapy. Arterioscler Thromb Vasc Biol. 1997;17: Zuliani G, Hobbs HH. Tetranucleotide repeat polymorphism in the apolipoprotein CIII gene. Nucleic Acids Res. 1990;18: Weber JL, May PE. Abundant class of human DNA polymorphisms which can be typed using polymerase chain reaction. Am J Hum Genet. 1989;44: Cooperative Human Linkage Center. Available at: Accessed November Whitehead Institute for Biomedical Research. Available at: Accessed November Beckman JS, Tomfohrde T, Barnes RI, Williams M, Broux O, Richard I, Weissenbach J, Bowcock AM. A linkage map of human chromosome 15 with an average resolution of 2 cm and containing 55 polymorphic microsatellites. Hum Mol Genet. 1993;2: Zuliani G, Hobbs HH. Dinucleotide repeat polymorphism at the 3 end of the LDL receptor gene. Nucleic Acids Res. 1990;18: The Genome Database. Available at: Accessed November Sharp D, Ricci B, Kienzle B, Lin MC, Wetterau JR. Human microsomal triglyceride transfer protein large subunit gene structure. Biochemistry. 1994;33: Jokinen E, Sakai J, Yamamoto T, Hobbs HH. CGG triple repeat polymorphism in the VLDL receptor (VLDLR) gene. Hum Mol Genet. 1994;3: Sakai N, Yamashita S, Hirano K, Menju M, Arai T, Kobayashi K, Ishigama M, Yoshida Y, Hoshino T, Nakajima N, Kameda-Takemura K, Matsuzawa Y. Frequency of exon 15 missense mutation (442D:G) in cholesteryl ester transfer protein gene in hyperalphalipoproteinemic Japanese subjects. Atherosclerosis. 1995;114: Skretting G, Prydz H. An amino acid exchange in exon I of the human lecithin:cholesterol acyltransferase (LCAT) gene is associated with fish eye disease. Biochem Biophys Res Commun. 1992;182: Wetterau JR, Zilversmit DB. Purification and characterization of microsomal triglyceride and cholesteryl ester transfer protein from bovine liver microsomes. Chem Phys Lipids. 1985;38: Walter MA, Spillett DJ, Thomas P, Weissenbach J, Goodfellow PN. A method for constructing radiation hybrid maps of whole genomes. Nat Genet. 1994;7: Haseman JK, Elston RC. The investigation of linkage between a quantitative trait and a marker locus. Behav Genet. 1972;2: Amos CI, Elston RC, Wilson AF, Bailey-Wilson JE. A more powerful robust sib-pair test of linkage for quantitative traits. Genet Epidemiol. 1989;6: Elston RC. SAGE: Statistical Analysis for Genetic Epidemiology. release 3.1. Cleveland, Ohio: Rammelkamp Center for Education and Research; Tsao BP, Cantor RM, Kalunian KC, Chen C-J, Badsha H, Singh R, Wallace DJ, Kitridou RC, Chen S, Shen N, Song YW, Isenberg DA, Yu C-L, Hahn BH, Rotter JI. Evidence for linkage of a candidate chromosome 1 region to human systemic lupus erythematosus. J Clin Invest. 1997;99: Gordon DA, Jamil H, Sharp D, Mullaney D, Yao Z, Gregg RE, Wetterau J. Secretion of apolipoprotein B-containing lipoproteins from HeLa cells is dependent on expression of the microsomal triglyceride transfer protein and is regulated by lipid availability. Proc Natl Acad Sci U S A. 1994; 91: Waite M. Phospholipases. In: Vance DE, Vance J, eds. Biochemistry of Lipids, Lipoproteins and Membranes. Amsterdam, Netherlands: Elsevier Science BV; 1996: Rotter JI, Bu X-D, Cantor RM, Warden CH, Brown J, Gray RJ, Blanche PJ, Krauss RM, Lusis AJ. Multilocus genetic determinants of LDL particle size in coronary artery disease families. Am J Hum Genet. 1996;58: 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: Tall AR. Plasma lipid transfer proteins. J Lipid Res. 1986;27: Hedrick CC, Castellani LW, Warden CH, Puppione DL, Lusis AJ. Influence of mouse apolipoprotein A-II on plasma lipoproteins in transgenic mice. J Biol Chem. 1993;268: Castellani LW, Navab M, Lenten BJV, Hedrick CC, Hama SY, Gotto AM, Fogelman AM, Lusis AJ. Overexpression of apolipoprotein AII in transgenic mice converts high density lipoproteins to proinflammatory particles. J Clin Invest. 1997;100: Von Shacky C, Kiefl R, Marcus AJ, Broekman MJ, Kaminski WE. Dietary n-3 fatty acids accelerate catabolism of leukotriene B4 in human granulocytes. Biochim Biophys Acta. 1993;1166: Wong GHW, Elwell JH, Oberley LW, Goeddel DV. Manganese superoxide dismutase is essential for cellular resistance to cytotoxicity of tumor necrosis factor. Cell. 1988;58: Cantor RM, Blanche PJ, Rotter JI, Davis R, Lanning CD, Fernstrom HS, Rawlings RS, Holl LG, Orr JE, Scheuner MT, Wen PZ, Lusis AJ, Krauss RM. Linkage of low density lipoprotein (LDL) subclass phenotypes to candidate gene loci in sibling pairs on high-fat (HF) and low-fat (LF) diets. J Invest Med. 1998;46:117A. Abstract. 66. Guerin M, Bruckert E, Dolphin PJ, Chapman MJ. Absence of cholesteryl ester transfer protein-mediated cholesteryl ester mass transfer from highdensity lipoprotein to low-density lipoprotein particles is a major feature of combined hyperlipidemia. Eur J Clin Invest. 1996;26: Brousseau ME, Santamarina-Fojo S, Zech LA, Berard AM, Vaisman BL, Meyn SM, Powell D, Brewer HB Jr, Hoeg JM. Hyperalphalipoproteinemia in human lecithin:cholesterol acyltransferase transgenic rabbits: in vivo apolipoprotein A-I catabolism is delayed in a gene dosedependent manner. J Clin Invest. 1996;97: Tahvanainen E, Pajukanta P, Porkka K, Nieminen S, Ikavalko L, Nuotio I, Taskinen MR, Peltonen L, Ehnholm C. Haplotypes of the ApoA-I/C- III/A-IV gene cluster and familial combined hyperlipidemia. Arterioscler Thromb Vasc Biol. 1998;18: Melov S, Schneider JA, Day BJ, Hinerfeld D, Coskun P, Mirra S, Crapo JD, Wallace DC. A novel neurological phenotype in mice lacking mitochondrial manganese superoxide dismutase. Nat Genet. 1998;18: Weinstock PH, Bisgaier CL, Hayek T, Aalto-Setala K, Schayek E, Wu L, Scheffele P, Merkel M, Essenburg AD, Breslow JL. Decreased HDL cholesterol levels but normal lipid absorption, growth, and feeding behavior in apolipoprotein A-IV knockout mice. J Lipid Res. 1997;38: Babirak SP, Iverius PH, Fujimoto WY, Brunzell JD. Detection and characterization of the heterozygote state for lipoprotein lipase deficiency. Arteriosclerosis. 1989;9: Risch N. Assessing the role of HLA-linked and unlinked determinant of disease. Am J Hum Genet. 1987;49: Hauser ER, Boehnke M, Sun-Wei G, Risch N. Affected-sib-pair interval mapping and exclusion for complex genetic traits: sampling considerations. Genet Epidemiol. 1996;13: