Genetic Epidemiology 8:269-275 (1991)
Linkage Analysis of Low-Density Lipoprotein Subclass Phenotypes and the Apolipoprotein B Gene Michael LaBelle, Melissa A. Austin, Edward Rubin, and Ronald M. Krauss Molecular Medicine Research Program, Research Medicine and Radiation Biophysics Division, Lawrence Berkeley Laboratory, University of California, Berkeley (M.L.B., E. R., R. M. K.); Department of Epidemiology, School of Public Health and Community Medicine, University of Washington, Seattle (M.A.A.). A common heritable phenotype has recently been identified which is characterized by a relative abundance of small, dense low-density lipoproteins (LDL), and mild elevations of plasma triglycerides and reductions in plasma high-density lipoproteins (HDL) cholesterol. This phenotype, designated LDL subclass phenotype B, has been associated with up to a three-fold increase in coronary disease risk. Complex segregation analysis in two large family studies has demonstrated that LDL subclass phenotype B is influenced by an allele at a single genetic locus with a population frequency of 0.25-0.3, and autosomal dominant inheritance, but with full penetrance only in males age 20 and over and in postmenopausal women. Since apolipoprotein B (apoB) is the principal protein component of LDL, linkage analysis was used to investigate possible linkage between the phenotyope B phenotype and the apoB gene, using a variable number of tandem repeats site located 0.5 kb from the 3' end of the apoB gene. In 6 informative families including only family members in the penetrant classes, a total LOD score of - 7.49 was found at a recombination fraction of 0.001. Thus, under the assumptions of the single gene model, it is unlikely that the apoB locus controls LDL subclass phenotype B. Key words: candidate gene, coronary heart disease risk, electrophoresis, genetics, polymerase chain reaction
INTRODUCTION Low-density lipoprotein (LDL) particles are the major carriers of cholesterol in humans. Using analytic and density gradient ultracentrifugation and gradient gel elecReceived for publication March 25, 1991; revision accepted July 19, 1991. Address reprint requests to Michael LaBelle, Ph.D., Lawrence Berkeley Laboratory 74- 157, Univeristy of California, 1 Cyclotron Road, Berkeley, CA 94720.
0 1991 Wiley-Liss, Inc.
270
LaBelle et al.
trophoresis, subclasses of LDL particles have been identified and characterized [Fisher et al., 1972; Krauss and Burke, 1982). More recently, two distinct LDL subclass phenotypes, denoted A and B, have been described in individual study subjects based on gradient gel electrophoresis [Austin et al., 1988al. LDL subclass phenotype A is characterized by a predominance of large, buoyant LDL particles, generally with a In contrast, subjects with LDL subclass peak particle diameter greater than 255 phenotype B have a predominance of small, dense LDL particles, with a diameter usually less than or equal to 255 [Austin et al., 1988al. In studies to date, 85%-90% of study subjects have one of these two LDL subclass phenotypes, with the remainder having an intermediate phenotype. LDL subclass phenotype B is present in approximately 30% of the general population [Austin et al., 1988bl. In a case-control study of myocardial infarction survivors, LDL subclass phenotype B was associated with a threefold increase in risk of myocardial infarction [Austin et al., 1988al. A population-based investigation of primarily healthy families also demonstrated that LDL subclass phenotype B was inherited consistent with the presence of a single major genetic locus [Austin et al., 1988bl. Using complex segregation analysis, the model providing the best fit to the family data included a single major locus with dominant mode of inheritance, an allele frequency of 25% for the proposed phenotype B allele, and reduced penetrance in young males and premenopausal females [Austin et al., 1988bl. In both of these studies, LDL subclass phenotype B was associated with an atherogenic lipoprotein profile, including increased plasma triglyceride and apoB levels and decreased HDL cholesterol levels [Austin et al., 1990aI. Thus a predominance of small, dense LDL is a marker for a complex lipoprotein profile associated with increased cardiovascular disease risk. The protein component of LDL particles consists of a single, large (550 kd) protein, apolipoprotein (apo) B, which accounts for at least 20% of the LDL particle mass [Goldstein and Brown, 19771. The apoB protein is encoded by a 43 kilobase (kb) gene on chromosome 2 [Ludwig et al., 19871 and is highly polymorphic [Boerwinkle et al., 19881. Restriction fragment length DNA polymorphisms (RFLPs) of apoB have also been associated with risk of coronary heart disease [Rajput-Williams et al., 1988; Hegele et al., 1986; Genest et al., 1990; Fried1 et al., 19901. Because apoB is the most likely structural protein to play a role in determining the properties and levels of LDL subclasses, the apoB gene is a reasonable candidate for controlling LDL subclass phenotypes. The present study used linkage analysis to determine if variation at the aopB locus is responsible for LDL subclass phenotypes. Variation at the aopB locus was characterized based on a region approximately 0.5 kb from the 3' end of the aopB gene that contains a variable number of tandem repeats (VNTR) [Boenvinkle et al., 19881. Fourteen different size alleles, containing up to 52 repeats of the basic 15 bp unit, have been found in this region [Ludwig et al., 19891.
A.
MATERIALS AND METHODS
Six nuclear families including 32 individual study subjects were recruited for the study, and the pedigrees are shown in Figure 1. The families were specifically chosen to be informative for linkage analysis based on two criteria. First, families were selected in which one parent had LDL subclass phenotype A and the other had LDL subclass phenotype B. Based on complex segregation analysis, males aged 20 years and over
LDL Subclass Phenotypes and Apo B Gene
m
Fmily
LN
26
25
(53.53)
Pr-?
Family MA
JJimn 40
34
271
32
Fmiily ME
2x
27
24
(53.45)
7-e
(S3.53)
(53.45)
Fanlily
22
FamilyCL
Fig. 1. Inheritance of the VNTR region near the apoB gene and LDL subclass phenotypes in six informative families. Subjects with the LDL subclass A phenotype are represented by open symbols, while subjects with the LDL subclass B phenotype are represented by the solid symbols. The age of each family member is given, and the numbers in parentheses indicate the number of 15 bp repeats found in the VNTR region on each chromosome.
and postmenopausal women have full penetrance for the proposed phenotype B allele [Austin et al., 1988bl. Thus the second criteria was that all family members included in the analysis were in these age and gender categories. The families were not selected for the presence of coronary heart disease or lipid disorders. For the LDL subclass determinations, 30 ml of whole blood was drawn from each study subject into a final concentration of 0.15 mg K2EDTA/ml, and plasma was
272
LaBelle et al.
immediately obtained by centrifugation. Gradient gel electrophoresis was performed on plasma using 2%- 16% polyacrylamide gradient gels (Pharmacia), as previously described [Krauss and Burke, 1982; Nichols et al., 19861. The resulting scans were used to classify study subjects as having either LDL subclass phenotype A or B. For analysis of inheritance of the VNTR size alleles, 40 ml of whole blood from each subject was drawn into acid citrate dextrose. DNA was isolated from the buffycoat cells by the method of Bell et al. [1981J. The polymerase chain reaction (PCR) technique was used to amplify the VNTR region of the apoB gene [Boenvinkle et al., 19881. The oligonucleotide primers for the PCR were 19 nucleotides long and were chosen to flank the region known to contain the VNTR near the apoB translational stop codon. The sequence of the primer on the 5‘ side of the VNTR region was 5’-ACGGAGAAATTATGGAGGG-3’, and the sequence on the 3’ side was 5‘TGGCAAATACAATTCCTGA-3’. The PCR was performed in a reaction volume of 30 p1using 0.1 pg genomic DNA, a 2 mM final concentration of each primer, and a concentration of 200 mM for each dNTP. Amplification of the VNTR region was by a modification of the method of Boenvinkle et al. [1988], in which the amplification mix was incubated at 94°C for 3 minutes before addition of 0.9 units of a 1:lO dilution into sterile water of TAQ DNA polymerase (obtained from Perkin-Elmer-Cetus or United States Biochemical Corporation). Annealing and extension (26 cycles) were performed for 6 minutes at 61°C and for 1 minute at 95”C, respectively. The amplification mixture was then electrophoresed at 125 V for 2 to 3 hours in 2% agarose gels and the DNA stained with ethidium bromide. The amplified DNA from parents and offspring from each nuclear family were run on the same gel. The VNTR alleles were defined based on a linear relationship between the estimated number of base pairs, ranging from 612 to 930, and the number of 15 base pair repeats. In the subjects studied here, 13 VNTR alleles were identified, ranging from 33 to 55 repeats. Linkage analysis was performed using the computer program LIPED [Ott, 1985; Morton et al., 19831. Lod scores were calculated over a range of recombination values from 0.001 to 0.5. For the apoB VNTR, codominant inheritance and equal allele frequencies were assumed. Although the distribution of alleles has been shown to be bimodel [Boenvinkle et al., 1988; Ludwig et al., 1989; Fried1 et al., 19901, lod scores were virtually unchanged when frequency estimates were incorporated into the analysis. Based on the best-fitting model from complex segregation analysis of primarily healthy families [Austin et al., 1988b], the LDL subclass phenotype B was assumed to have a dominant mode of inheritance, an allele frequency of 0.25, and reduced penetrance in males under age 20 years (0.39) and premenopausal women (0.30). RESULTS
Figure 1 shows the segregation of both LDL subclass phenotypes and the apoB VNTR alleles in each of six informative nuclear families. As expected from the complex segregation analysis [Austin et al., 1988b], LDL subclass phenotype B segregates consistent with the presence of a dominant mode of inheritance. The apoB VNTR alleles can also be seen to segregate in a codominant fashion. The results of the linkage analysis between LDL subclass phenotypes and the apoB VNTR are given in Table I. In four of the families (ME, CL, MY, and MA), lod scores are consistently negative over a range of recombination fractions. In the two
LDL Subclass Phenotypes and Apo B Gene
273
TABLE I. Lod Scores for Linkage Between LDL Subclass Phenotype B and the VNTR Region Immediately 3’ lipoprotein B Gene Family
0.001
0.100
ME LN CL HN MY MA Sum
2.098 0.300 - 2.398 0.300 -2.398 -1.199 -7.493
-0.441 0.215 - 0.444 0.214 -0.444 0.530 -0.370
Recombination fraction (6) 0.200 0.300
0.400
0.500
0.000 0.000 O.OO0 0.000 0.000 0.000 0.000
-0.193 0.134 -0.194 0.133 -0.194 0.525
- 0.075
0.064 -0.076 0.064 -0.076 0.359
-0.018 0.017 -0.018 0.017 -0.018 0.130
0.21 I
0.260
0.110
other families (LN and HN), lod scores were slightly higher than zero. The total lod score at a recombination fraction of 0.001 was - 7.49, providing evidence against linkage between the LDL subclass phenotypes and the apoB VNTR in these families. DISCUSSION
Because apolipoprotein B is the primary protein component of LDL particles, the present study was performed to test the hypothesis that the proposed locus controlling LDL subclass phenotypes is genetically linked to the apoB locus on chromosome 2. That is, a candidate gene approach was used to determine if variation at the apoB locus is responsible for the inheritance of LDL subclass phenotypes. Based on lod score analysis from six informative nuclear families, linkage between the proposed LDL subclass phenotype locus and apoB is unlikely. Two important assumptions have been made in performing the linkage analysis in this study. First, the VNTR region as a DNA marker is approximately 0.5 kb from the 3’ end of the apoB gene on chromosome 2 and thus is not contained within the apoB gene itself. However, because this region is so close to the apoB locus, recombination between the marker and the apoB gene is unlikely. In addition, Fried1 et al. [ 19901 have demonstrated linkage disquilibrium between this hypervariable region and an EcoRI RFLP site within the apo B gene. Thus the results of linkage analysis can be interpreted to reflect variation at the apoB locus. Second, the segregation of LDL subclass phenotypes is assumed to be controlled by a major genetic locus and that phenotype B is inherited as a dominant trait. This assumption is based on the results of complex segregation analysis of a sample of 61 nuclear families from a community-based study [Austin et al., 1988bl. Although complex segregation analysis is known to have a number of limitations and cannot exclude heterogeneity of genetic loci responsible for the LDL subclass B phenotype, a recent study based on a sample of families with familial comined hyperlipidemia also demonstrated a similar mode of inheritance and penetrance values [Austin et al., 1990bl. Thus the single locus inheritance of LDL subclass phenotypes appears to be a reasonable model. However, it is important to note that this model assumes full penetrance (1 .O) for LDL subclass phenotype B among males aged 20 years and over and among postmenopausal women. When lower penetrance values are assumed, less information about recombinants and negative linkage is available. The present analysis was repeated using
274
LaBelle et al.
a value of 0.8 for the penetrance of phenotype B. Although the sign of the lod scores remained the same in each family, the overall lod score was 0.051 at a recombination fraction of 0.001. Thus no firm conclusion about lack of linkage can be drawn if less than full penetrance is assumed. Because LDL subclass phenotypes are associated with a constellation of other lipid and lipoprotein measures [Austin et al., 1990a1, numerous other genes known to have a role in lipoprotein metabolism could be involved in determining the phenotypes. These candidate genes include structural genes for other apolipoproteins such as the AI-CIII-AIV locus, lipoprotein receptors such as the LDL receptor, or enzymes such as lipoprotein lipase [Lusis, 19881. However, given the complexity of lipoprotein metabolism and the small proportion of mapped genes in the genome, it is also possible that as yet unidentified gene(s) control LDL subclass phenotypes. In conclusion, the present study has provided evidence that the apo B locus on chromosome 2 is not linked to the proposed gene controlling LDL subclass phenotypes, under the assumptions of the single gene model. Thus the apo B gene can likely be excluded as a candidate locus controlling LDL subclass phenotypes. ACKNOWLEDGMENTS
The authors wish to thank Patsy Nishina for preparation of the genomic DNA samples, Laura Glines Holl for gradient gel electrophoresis, Adelie Cavanaugh for sample collection, and Sunwei Guo for performing the LIPED computer runs. This research was supported by NIH program project grant HL-18574 from the National Heart, Lung and Blood Institute, by NIH First Independent Research Support and Transition Award HL-38760, and by the National Dairy Research and Promotion Board. The work was conducted at the Lawrence Berkeley Laboratory (Department of Energy contract DE-AC03-76sF00098 to the University of California). REFERENCES Austin MA, Breslow JL, Hennekens CH, Buring JE, Williett WC, Krauss RM (1988a): Low-density lipoprotein subclass patterns and risk of myocardial infarction. JAMA 260:1917-1921. Austin MA, King M-C, Vranizan KM, Newman B, Krauss RM (1988b): Inheritance of low-density lipoprotein subclass patterns: results of complex segregation analysis. Am J Hum Genet 43:838-846. Austin MA, King M-C, Vranizan KM, Krauss RM (1990a): Atherogenic lipoprotein phenotype: A proposed genetic marker for coronary heart disease risk. Circulation 82:495-506. Austin MA, Brunzell JD, Fitch WL, Krauss RM (1990b): Inheritance of low density lipoprotein subclass patterns in familial combined hyperlipidemia. Arteriosclerosis 10520-530. Bell GI, Karam JH, Rutter WJ (1981): Polymorphic DNA region adjacent to the 5’ end of the human insulin gene. Proc Natl Acad Sci USA 785759-5763. Boerwinkle E, Xiong WJ, Fourest E, Chan L (1988): Rapid typing of tandemly repeated hypervariable loci by the polymerase chain reaction: Application to the apolipoprotein B 3’ hypervariable region. Proc Natl Acad Sci USA 85:212-216. Fisher WR, Hammond MG, Warmke GL (1972): Measurements of the molecular weight variability of plasma low density lipoproteins among normals and subjects with hyper-P-lipoproteinemia: Demonstration of macromolecular heterogeneity. Biochemistry 1 1519-529. Fried1 W, Ludwig EH, Paulweber 9, Sandhofer F, McCarthy BJ (1990): Hypervariability in a minisatellite 3‘ of the apolipoprotein B gene in patients with coronary heart disease compared with normal controls. J Lipid Res 3 1:659-665. Genest JJ, Ordovas JM, McNamara JR, Robbins AM, Meade T, Cohn SD, Salem DN, Wilson PWF,
LDL Subclass Phenotypes and Apo B Gene
275
Masharani U, Frossard PM, Schaefer EJ (1990): DNA polymorphisms of the apolipoprotein B gene in patients with premature coronary artery disease. Atherosclerosis 82:7-17. Goldstein JL, Brown MS (1977): The low-density lipoprotein pathway and its relationship to atherosclerosis. Annu Rev Biochem 46897-930. Hegele RA, Huang L-S, Herbert PN, Blum CB, Buring JB, Hennekens CH, Breslow JL (1986): Apolipoprotin B gene DNA polymorphisms associated with myocardial infarction. N Engl J Med 315: 1509- 1515. Krauss RM, Burke DJ (1982): Identification of multiple subclasses of plasma low density lipoproteins in normal humans. J Lipid Res 23:97-104. Ludwig EH, Blackhart BD, Pierotti VR, Caiati L. Fortier C , Knott T, Scott J, Mahley RW, LevyWilson B, McCarthy BJ (1987): DNA sequence of the human apolipoprotein B gene. DNA 6:363-372. Ludwig EH, Fried W, McCarthy, EJ (1989): High-resolution analysis of a hypervariable region in the human apolipoprotein B gene. Am J Hum Genet 45:458-464. Lusis AJ (1988): Genetic factors affecting blood lipoproteins: The candidate gene approach. J Lipid Res 29~397-429. Morton NE, Rao DC, Lalouel J-M (1983): Methods in genetic epidemiology. In Klingberg MA (ed): “Contributions to Genetic Epidemiology,” vol4. Basel, Karger: pp 103-107. Nichols AV, Krauss RM, Musliner TA (1986): Nondenaturing polyacrylamide gradient gel electrophoresis. In Segrest JP, Albers JJ (eds): “Methods in Enzymology, vol 128: Plasma Lipoproteins, part A,” Orlando, Academic Press, pp 417-431. Ott J (1985): “Analysis of Human Genetic Linkage.” Baltimore, MD: John Hopkins University Press, pp 22-80. Rajput-Williams J , Knott TJ, Wallis SC, Sweetman P, Yarnell J, Cox N, Bell GI, Niller NE, Scott J (1988): Variation of apolipoprotein-B gene is associated with obesity, high blood cholesterol levels, and increased risk of coronary heart disease. Lancet 2: 1442- 1446.
Edited by G. P. Vogler
Linkage Analysis of Low-Density Lipoprotein Subclass Phenotypes and the Apolipoprotein B Gene Michael LaBelle, Melissa A. Austin, Edward Rubin, and Ronald M. Krauss Molecular Medicine Research Program, Research Medicine and Radiation Biophysics Division, Lawrence Berkeley Laboratory, University of California, Berkeley (M.L.B., E. R., R. M. K.); Department of Epidemiology, School of Public Health and Community Medicine, University of Washington, Seattle (M.A.A.). A common heritable phenotype has recently been identified which is characterized by a relative abundance of small, dense low-density lipoproteins (LDL), and mild elevations of plasma triglycerides and reductions in plasma high-density lipoproteins (HDL) cholesterol. This phenotype, designated LDL subclass phenotype B, has been associated with up to a three-fold increase in coronary disease risk. Complex segregation analysis in two large family studies has demonstrated that LDL subclass phenotype B is influenced by an allele at a single genetic locus with a population frequency of 0.25-0.3, and autosomal dominant inheritance, but with full penetrance only in males age 20 and over and in postmenopausal women. Since apolipoprotein B (apoB) is the principal protein component of LDL, linkage analysis was used to investigate possible linkage between the phenotyope B phenotype and the apoB gene, using a variable number of tandem repeats site located 0.5 kb from the 3' end of the apoB gene. In 6 informative families including only family members in the penetrant classes, a total LOD score of - 7.49 was found at a recombination fraction of 0.001. Thus, under the assumptions of the single gene model, it is unlikely that the apoB locus controls LDL subclass phenotype B. Key words: candidate gene, coronary heart disease risk, electrophoresis, genetics, polymerase chain reaction
INTRODUCTION Low-density lipoprotein (LDL) particles are the major carriers of cholesterol in humans. Using analytic and density gradient ultracentrifugation and gradient gel elecReceived for publication March 25, 1991; revision accepted July 19, 1991. Address reprint requests to Michael LaBelle, Ph.D., Lawrence Berkeley Laboratory 74- 157, Univeristy of California, 1 Cyclotron Road, Berkeley, CA 94720.
0 1991 Wiley-Liss, Inc.
270
LaBelle et al.
trophoresis, subclasses of LDL particles have been identified and characterized [Fisher et al., 1972; Krauss and Burke, 1982). More recently, two distinct LDL subclass phenotypes, denoted A and B, have been described in individual study subjects based on gradient gel electrophoresis [Austin et al., 1988al. LDL subclass phenotype A is characterized by a predominance of large, buoyant LDL particles, generally with a In contrast, subjects with LDL subclass peak particle diameter greater than 255 phenotype B have a predominance of small, dense LDL particles, with a diameter usually less than or equal to 255 [Austin et al., 1988al. In studies to date, 85%-90% of study subjects have one of these two LDL subclass phenotypes, with the remainder having an intermediate phenotype. LDL subclass phenotype B is present in approximately 30% of the general population [Austin et al., 1988bl. In a case-control study of myocardial infarction survivors, LDL subclass phenotype B was associated with a threefold increase in risk of myocardial infarction [Austin et al., 1988al. A population-based investigation of primarily healthy families also demonstrated that LDL subclass phenotype B was inherited consistent with the presence of a single major genetic locus [Austin et al., 1988bl. Using complex segregation analysis, the model providing the best fit to the family data included a single major locus with dominant mode of inheritance, an allele frequency of 25% for the proposed phenotype B allele, and reduced penetrance in young males and premenopausal females [Austin et al., 1988bl. In both of these studies, LDL subclass phenotype B was associated with an atherogenic lipoprotein profile, including increased plasma triglyceride and apoB levels and decreased HDL cholesterol levels [Austin et al., 1990aI. Thus a predominance of small, dense LDL is a marker for a complex lipoprotein profile associated with increased cardiovascular disease risk. The protein component of LDL particles consists of a single, large (550 kd) protein, apolipoprotein (apo) B, which accounts for at least 20% of the LDL particle mass [Goldstein and Brown, 19771. The apoB protein is encoded by a 43 kilobase (kb) gene on chromosome 2 [Ludwig et al., 19871 and is highly polymorphic [Boerwinkle et al., 19881. Restriction fragment length DNA polymorphisms (RFLPs) of apoB have also been associated with risk of coronary heart disease [Rajput-Williams et al., 1988; Hegele et al., 1986; Genest et al., 1990; Fried1 et al., 19901. Because apoB is the most likely structural protein to play a role in determining the properties and levels of LDL subclasses, the apoB gene is a reasonable candidate for controlling LDL subclass phenotypes. The present study used linkage analysis to determine if variation at the aopB locus is responsible for LDL subclass phenotypes. Variation at the aopB locus was characterized based on a region approximately 0.5 kb from the 3' end of the aopB gene that contains a variable number of tandem repeats (VNTR) [Boenvinkle et al., 19881. Fourteen different size alleles, containing up to 52 repeats of the basic 15 bp unit, have been found in this region [Ludwig et al., 19891.
A.
MATERIALS AND METHODS
Six nuclear families including 32 individual study subjects were recruited for the study, and the pedigrees are shown in Figure 1. The families were specifically chosen to be informative for linkage analysis based on two criteria. First, families were selected in which one parent had LDL subclass phenotype A and the other had LDL subclass phenotype B. Based on complex segregation analysis, males aged 20 years and over
LDL Subclass Phenotypes and Apo B Gene
m
Fmily
LN
26
25
(53.53)
Pr-?
Family MA
JJimn 40
34
271
32
Fmiily ME
2x
27
24
(53.45)
7-e
(S3.53)
(53.45)
Fanlily
22
FamilyCL
Fig. 1. Inheritance of the VNTR region near the apoB gene and LDL subclass phenotypes in six informative families. Subjects with the LDL subclass A phenotype are represented by open symbols, while subjects with the LDL subclass B phenotype are represented by the solid symbols. The age of each family member is given, and the numbers in parentheses indicate the number of 15 bp repeats found in the VNTR region on each chromosome.
and postmenopausal women have full penetrance for the proposed phenotype B allele [Austin et al., 1988bl. Thus the second criteria was that all family members included in the analysis were in these age and gender categories. The families were not selected for the presence of coronary heart disease or lipid disorders. For the LDL subclass determinations, 30 ml of whole blood was drawn from each study subject into a final concentration of 0.15 mg K2EDTA/ml, and plasma was
272
LaBelle et al.
immediately obtained by centrifugation. Gradient gel electrophoresis was performed on plasma using 2%- 16% polyacrylamide gradient gels (Pharmacia), as previously described [Krauss and Burke, 1982; Nichols et al., 19861. The resulting scans were used to classify study subjects as having either LDL subclass phenotype A or B. For analysis of inheritance of the VNTR size alleles, 40 ml of whole blood from each subject was drawn into acid citrate dextrose. DNA was isolated from the buffycoat cells by the method of Bell et al. [1981J. The polymerase chain reaction (PCR) technique was used to amplify the VNTR region of the apoB gene [Boenvinkle et al., 19881. The oligonucleotide primers for the PCR were 19 nucleotides long and were chosen to flank the region known to contain the VNTR near the apoB translational stop codon. The sequence of the primer on the 5‘ side of the VNTR region was 5’-ACGGAGAAATTATGGAGGG-3’, and the sequence on the 3’ side was 5‘TGGCAAATACAATTCCTGA-3’. The PCR was performed in a reaction volume of 30 p1using 0.1 pg genomic DNA, a 2 mM final concentration of each primer, and a concentration of 200 mM for each dNTP. Amplification of the VNTR region was by a modification of the method of Boenvinkle et al. [1988], in which the amplification mix was incubated at 94°C for 3 minutes before addition of 0.9 units of a 1:lO dilution into sterile water of TAQ DNA polymerase (obtained from Perkin-Elmer-Cetus or United States Biochemical Corporation). Annealing and extension (26 cycles) were performed for 6 minutes at 61°C and for 1 minute at 95”C, respectively. The amplification mixture was then electrophoresed at 125 V for 2 to 3 hours in 2% agarose gels and the DNA stained with ethidium bromide. The amplified DNA from parents and offspring from each nuclear family were run on the same gel. The VNTR alleles were defined based on a linear relationship between the estimated number of base pairs, ranging from 612 to 930, and the number of 15 base pair repeats. In the subjects studied here, 13 VNTR alleles were identified, ranging from 33 to 55 repeats. Linkage analysis was performed using the computer program LIPED [Ott, 1985; Morton et al., 19831. Lod scores were calculated over a range of recombination values from 0.001 to 0.5. For the apoB VNTR, codominant inheritance and equal allele frequencies were assumed. Although the distribution of alleles has been shown to be bimodel [Boenvinkle et al., 1988; Ludwig et al., 1989; Fried1 et al., 19901, lod scores were virtually unchanged when frequency estimates were incorporated into the analysis. Based on the best-fitting model from complex segregation analysis of primarily healthy families [Austin et al., 1988b], the LDL subclass phenotype B was assumed to have a dominant mode of inheritance, an allele frequency of 0.25, and reduced penetrance in males under age 20 years (0.39) and premenopausal women (0.30). RESULTS
Figure 1 shows the segregation of both LDL subclass phenotypes and the apoB VNTR alleles in each of six informative nuclear families. As expected from the complex segregation analysis [Austin et al., 1988b], LDL subclass phenotype B segregates consistent with the presence of a dominant mode of inheritance. The apoB VNTR alleles can also be seen to segregate in a codominant fashion. The results of the linkage analysis between LDL subclass phenotypes and the apoB VNTR are given in Table I. In four of the families (ME, CL, MY, and MA), lod scores are consistently negative over a range of recombination fractions. In the two
LDL Subclass Phenotypes and Apo B Gene
273
TABLE I. Lod Scores for Linkage Between LDL Subclass Phenotype B and the VNTR Region Immediately 3’ lipoprotein B Gene Family
0.001
0.100
ME LN CL HN MY MA Sum
2.098 0.300 - 2.398 0.300 -2.398 -1.199 -7.493
-0.441 0.215 - 0.444 0.214 -0.444 0.530 -0.370
Recombination fraction (6) 0.200 0.300
0.400
0.500
0.000 0.000 O.OO0 0.000 0.000 0.000 0.000
-0.193 0.134 -0.194 0.133 -0.194 0.525
- 0.075
0.064 -0.076 0.064 -0.076 0.359
-0.018 0.017 -0.018 0.017 -0.018 0.130
0.21 I
0.260
0.110
other families (LN and HN), lod scores were slightly higher than zero. The total lod score at a recombination fraction of 0.001 was - 7.49, providing evidence against linkage between the LDL subclass phenotypes and the apoB VNTR in these families. DISCUSSION
Because apolipoprotein B is the primary protein component of LDL particles, the present study was performed to test the hypothesis that the proposed locus controlling LDL subclass phenotypes is genetically linked to the apoB locus on chromosome 2. That is, a candidate gene approach was used to determine if variation at the apoB locus is responsible for the inheritance of LDL subclass phenotypes. Based on lod score analysis from six informative nuclear families, linkage between the proposed LDL subclass phenotype locus and apoB is unlikely. Two important assumptions have been made in performing the linkage analysis in this study. First, the VNTR region as a DNA marker is approximately 0.5 kb from the 3’ end of the apoB gene on chromosome 2 and thus is not contained within the apoB gene itself. However, because this region is so close to the apoB locus, recombination between the marker and the apoB gene is unlikely. In addition, Fried1 et al. [ 19901 have demonstrated linkage disquilibrium between this hypervariable region and an EcoRI RFLP site within the apo B gene. Thus the results of linkage analysis can be interpreted to reflect variation at the apoB locus. Second, the segregation of LDL subclass phenotypes is assumed to be controlled by a major genetic locus and that phenotype B is inherited as a dominant trait. This assumption is based on the results of complex segregation analysis of a sample of 61 nuclear families from a community-based study [Austin et al., 1988bl. Although complex segregation analysis is known to have a number of limitations and cannot exclude heterogeneity of genetic loci responsible for the LDL subclass B phenotype, a recent study based on a sample of families with familial comined hyperlipidemia also demonstrated a similar mode of inheritance and penetrance values [Austin et al., 1990bl. Thus the single locus inheritance of LDL subclass phenotypes appears to be a reasonable model. However, it is important to note that this model assumes full penetrance (1 .O) for LDL subclass phenotype B among males aged 20 years and over and among postmenopausal women. When lower penetrance values are assumed, less information about recombinants and negative linkage is available. The present analysis was repeated using
274
LaBelle et al.
a value of 0.8 for the penetrance of phenotype B. Although the sign of the lod scores remained the same in each family, the overall lod score was 0.051 at a recombination fraction of 0.001. Thus no firm conclusion about lack of linkage can be drawn if less than full penetrance is assumed. Because LDL subclass phenotypes are associated with a constellation of other lipid and lipoprotein measures [Austin et al., 1990a1, numerous other genes known to have a role in lipoprotein metabolism could be involved in determining the phenotypes. These candidate genes include structural genes for other apolipoproteins such as the AI-CIII-AIV locus, lipoprotein receptors such as the LDL receptor, or enzymes such as lipoprotein lipase [Lusis, 19881. However, given the complexity of lipoprotein metabolism and the small proportion of mapped genes in the genome, it is also possible that as yet unidentified gene(s) control LDL subclass phenotypes. In conclusion, the present study has provided evidence that the apo B locus on chromosome 2 is not linked to the proposed gene controlling LDL subclass phenotypes, under the assumptions of the single gene model. Thus the apo B gene can likely be excluded as a candidate locus controlling LDL subclass phenotypes. ACKNOWLEDGMENTS
The authors wish to thank Patsy Nishina for preparation of the genomic DNA samples, Laura Glines Holl for gradient gel electrophoresis, Adelie Cavanaugh for sample collection, and Sunwei Guo for performing the LIPED computer runs. This research was supported by NIH program project grant HL-18574 from the National Heart, Lung and Blood Institute, by NIH First Independent Research Support and Transition Award HL-38760, and by the National Dairy Research and Promotion Board. The work was conducted at the Lawrence Berkeley Laboratory (Department of Energy contract DE-AC03-76sF00098 to the University of California). REFERENCES Austin MA, Breslow JL, Hennekens CH, Buring JE, Williett WC, Krauss RM (1988a): Low-density lipoprotein subclass patterns and risk of myocardial infarction. JAMA 260:1917-1921. Austin MA, King M-C, Vranizan KM, Newman B, Krauss RM (1988b): Inheritance of low-density lipoprotein subclass patterns: results of complex segregation analysis. Am J Hum Genet 43:838-846. Austin MA, King M-C, Vranizan KM, Krauss RM (1990a): Atherogenic lipoprotein phenotype: A proposed genetic marker for coronary heart disease risk. Circulation 82:495-506. Austin MA, Brunzell JD, Fitch WL, Krauss RM (1990b): Inheritance of low density lipoprotein subclass patterns in familial combined hyperlipidemia. Arteriosclerosis 10520-530. Bell GI, Karam JH, Rutter WJ (1981): Polymorphic DNA region adjacent to the 5’ end of the human insulin gene. Proc Natl Acad Sci USA 785759-5763. Boerwinkle E, Xiong WJ, Fourest E, Chan L (1988): Rapid typing of tandemly repeated hypervariable loci by the polymerase chain reaction: Application to the apolipoprotein B 3’ hypervariable region. Proc Natl Acad Sci USA 85:212-216. Fisher WR, Hammond MG, Warmke GL (1972): Measurements of the molecular weight variability of plasma low density lipoproteins among normals and subjects with hyper-P-lipoproteinemia: Demonstration of macromolecular heterogeneity. Biochemistry 1 1519-529. Fried1 W, Ludwig EH, Paulweber 9, Sandhofer F, McCarthy BJ (1990): Hypervariability in a minisatellite 3‘ of the apolipoprotein B gene in patients with coronary heart disease compared with normal controls. J Lipid Res 3 1:659-665. Genest JJ, Ordovas JM, McNamara JR, Robbins AM, Meade T, Cohn SD, Salem DN, Wilson PWF,
LDL Subclass Phenotypes and Apo B Gene
275
Masharani U, Frossard PM, Schaefer EJ (1990): DNA polymorphisms of the apolipoprotein B gene in patients with premature coronary artery disease. Atherosclerosis 82:7-17. Goldstein JL, Brown MS (1977): The low-density lipoprotein pathway and its relationship to atherosclerosis. Annu Rev Biochem 46897-930. Hegele RA, Huang L-S, Herbert PN, Blum CB, Buring JB, Hennekens CH, Breslow JL (1986): Apolipoprotin B gene DNA polymorphisms associated with myocardial infarction. N Engl J Med 315: 1509- 1515. Krauss RM, Burke DJ (1982): Identification of multiple subclasses of plasma low density lipoproteins in normal humans. J Lipid Res 23:97-104. Ludwig EH, Blackhart BD, Pierotti VR, Caiati L. Fortier C , Knott T, Scott J, Mahley RW, LevyWilson B, McCarthy BJ (1987): DNA sequence of the human apolipoprotein B gene. DNA 6:363-372. Ludwig EH, Fried W, McCarthy, EJ (1989): High-resolution analysis of a hypervariable region in the human apolipoprotein B gene. Am J Hum Genet 45:458-464. Lusis AJ (1988): Genetic factors affecting blood lipoproteins: The candidate gene approach. J Lipid Res 29~397-429. Morton NE, Rao DC, Lalouel J-M (1983): Methods in genetic epidemiology. In Klingberg MA (ed): “Contributions to Genetic Epidemiology,” vol4. Basel, Karger: pp 103-107. Nichols AV, Krauss RM, Musliner TA (1986): Nondenaturing polyacrylamide gradient gel electrophoresis. In Segrest JP, Albers JJ (eds): “Methods in Enzymology, vol 128: Plasma Lipoproteins, part A,” Orlando, Academic Press, pp 417-431. Ott J (1985): “Analysis of Human Genetic Linkage.” Baltimore, MD: John Hopkins University Press, pp 22-80. Rajput-Williams J , Knott TJ, Wallis SC, Sweetman P, Yarnell J, Cox N, Bell GI, Niller NE, Scott J (1988): Variation of apolipoprotein-B gene is associated with obesity, high blood cholesterol levels, and increased risk of coronary heart disease. Lancet 2: 1442- 1446.
Edited by G. P. Vogler
