Eliminating Atherogenesis in Mice by Switching Off Hepatic Lipoprotein Secretion
Background— LDL receptor–deficient “apolipoprotein (apo)-B100–only” mice (Ldlr−/−Apob100/100 have elevated LDL cholesterol levels on a chow diet and develop severe aortic atherosclerosis. We hypothesized that both the hypercholesterolemia and the susceptibility to atherosclerosis could be eliminated by switching off hepatic lipoprotein production.
Methods and Results— We bred Ldlr−/−Apob100/100 mice that were homozygous for a conditional allele for Mttp (the gene for microsomal triglyceride transfer protein) and the inducible Mx1-Cre transgene. In these animals, which we called “Reversa mice,” the hypercholesterolemia could be reversed, without modifying the diet or initiating a hypolipidemic drug, by the transient induction of Cre expression in the liver. After Cre induction, hepatic Mttp expression was virtually eliminated (as judged by quantitative real-time PCR), hepatic lipoprotein secretion was abolished (as judged by electron microscopy), and LDLs were virtually eliminated from the plasma. Intestinal lipoprotein production was unaffected. In mice fed a chow diet, Cre induction reduced plasma cholesterol levels from 233.9±46.0 to 37.2±6.5 mg/dL. In mice fed a high-fat diet, cholesterol levels fell from 525.7±32.2 to 100.6±14.3 mg/dL. The elimination of hepatic lipoprotein production completely prevented both the development of atherosclerosis and the changes in gene expression that accompany atherogenesis.
Conclusions— We developed mice in which hypercholesterolemia can be reversed with a genetic switch. These mice will be useful for understanding gene-expression changes that accompany the reversal of hypercholesterolemia and atherosclerosis.
Received October 22, 2002; accepted December 6, 2002.
Hypercholesterolemic mice are widely used in atherosclerosis research. Apolipoprotein (apo) E–deficient (Apoe−/−) mice have elevated plasma concentrations of apo-B48–containing remnant lipoproteins and develop advanced atherosclerosis by 4 to 6 months of age, even when fed a low-fat diet.1,2 LDL receptor–deficient (Ldlr−/−) mice have elevated levels of apo-B100–containing LDL in their plasma. However, Ldlr−/− mice have far lower total plasma cholesterol levels than Apoe−/− mice on a chow diet and develop only tiny atherosclerotic lesions.3 We recently demonstrated that “apo-B100–only” LDL receptor–deficient mice (Ldlr−/−Apob100/100)4 have higher LDL cholesterol levels than Ldlr−/− mice on a chow diet and develop severe atherosclerosis—even more severe than that of Apoe−/− mice.5
The hypercholesterolemic mouse models have been a boon for atherosclerosis research. Genetic studies with these mice have made it possible to define the impact of many genes on the development of atherosclerosis6–9 and also to document the gene-expression changes that accompany atherogenesis.10 However, experimental approaches with these mouse models are limited. For example, if one were interested in inducing the regression of atherosclerosis or in defining the effects of plasma cholesterol lowering on gene expression in the arterial wall, the existing mouse models would be of little use. Ldlr−/− and Apoe−/− mice remain hyperlipidemic after virtually all of the fat and cholesterol is removed from the diet,9 and neither animal model responds particularly well to the cholesterol-lowering effects of statins.11,12 Severe atherosclerotic lesions in Apoe−/− mice can be induced by feeding them a high-fat diet,1,2,13 but switching the mice to a low-fat chow diet does not cause lesion regression,14 because plasma lipid levels remain quite high.9
We reasoned that it would be useful to develop a hypercholesterolemic mouse in which the elevated levels of cholesterol in the plasma could be reversed with a simple genetic intervention. A potential strategy was suggested by a recent study15 in which Cre/loxP recombination techniques were used to switch off a conditional allele of microsomal triglyceride transfer protein (Mttp)16 in the liver, thereby preventing hepatic lipoprotein assembly and secretion. We predicted that it might be possible to produce mice with “reversible hypercholesterolemia” by breeding Ldlr−/−Apob100/100 mice that also were homozygous for a conditional Mttp allele and an inducible Cre transgene. In such an animal, we hypothesized that the induction of Cre would eliminate LDL from the plasma and that the resultant fall in cholesterol levels would reduce susceptibility to atherosclerosis. We further hypothesized that the reduction in plasma cholesterol levels would abort the arterial wall gene-expression perturbations that accompany atherogenesis. In the present studies, we bred Ldlr−/−Apob100/100 mice with the conditional Mttp allele and the Cre transgene and tested each of those hypotheses.
Genetically Modified Mice
Ldlr−/− mice3 homozygous for an apo-B100–only allele5,17 were bred with mice that were homozygous for a conditional (“floxed”) Mttp allele15 and the Mx1-Cre transgene18 (Mttpfl/flMx1-Cre+/+). Offspring were bred to produce mice that were homozygous for all 4 alleles (Ldlr−/−Apob100/100Mttpfl/flMx1-Cre+/+). Those mice, called “Reversa mice,” were intercrossed for 3 generations. All mice were bred in our lab, weaned at 28 days of age, and fed a chow diet containing 4.5% fat (Ralston Purina) or a high-fat diet containing 0.15% cholesterol and 21% fat (Research Diets).
Cre expression from the Mx1-Cre transgene was induced with intraperitoneal injections of polyinosinic-polycytidylic ribonucleic acid (pI-pC) (Sigma) (500 μg every other day for a total of 4 injections).18 The Mx1-Cre transgene is effective in causing near-complete recombination in liver and spleen.15,18 In the case of the Mttpfl allele, Cre deletes exon 1 of Mttp, inactivating the gene (producing an MttpΔ allele).15
Judging the Extent of Mttp Inactivation
To assess Cre-mediated recombination within Mttp in the liver, we analyzed SacI-cleaved liver genomic DNA with Southern blots.15 Mttp mRNA levels in the livers of untreated and pI-pC–treated Reversa mice were assessed by quantitative real-time polymerase chain reaction (PCR). Total liver RNA (5 ng), oligonucleotide primers (10 μmol/L), and the oligonucleotide probes (5 μmol/L) were mixed, and the PCR reactions were carried out with the PLATINUM Quantitative RT-PCR Thermoscript 1-step kit (Invitrogen Life Technologies). The Mttp oligonucleotide primers were 5′-ATGATCCTCTTGGCAGTGCTT-3′ and 5′-TGAGAGGCCA-GTTGTGTGAC-3′, and the probe was 5′-6FAM–TCTCTGCTTCTTCTCCTCCTACTCTGCTTC–TAMRA-3′ (PE Applied Biosystems). Mttp expression levels were normalized to expression levels for GAPDH (Rodent GAPDH Control Reagents, PE Applied Biosystems). The reverse transcriptase step was performed for 30 minutes at 60°C, followed by a heat-inactivation step for 5 minutes at 95°C; the real-time PCR reaction was performed for 40 cycles (95°C for 15 seconds, 60°C for 60 seconds) on an ABI Prism 7700 Sequence Detection System (PE Applied Biosystems).
Western blots of fresh plasma samples were used to judge the relative levels of apo-B100 in the plasma of untreated and pI-pC–treated mice. Mouse plasma (1.0 μL) was size-fractionated on a 4% polyacrylamide/SDS gel, and Western blots were performed with a rabbit antiserum against mouse apo-B and chemiluminescence detection reagents (Amersham).19
Lipid Measurements and Lipoprotein Analyses
Total plasma cholesterol and triglyceride concentrations were measured on fresh plasma with colorimetric assays (Spectrum Cholesterol Assay; Abbot Laboratories; and Triglyceride/GB Kit; Roche). The distribution of lipids within the plasma lipoproteins was determined by fractionating mouse plasma (350 μL pooled from 10 mice) by fast-performance liquid chromatography (FPLC) on a Superose 6 10/30 column.4,5 Mice were fasted for 4 hours before blood sampling, except when indicated otherwise.
Ultrastructural Analysis of Lipoprotein Assembly in Mouse Liver
Tissue samples were embedded in epoxy resin after perfusion fixation with 1.5% glutaraldehyde, 4% polyvinylpyrrolidone (molecular weight 10 000), and 0.05% calcium chloride in 100 mmol/L sodium cacodylate buffer (pH 7.4). Electron microscopy was performed on liver tissue that had been stained with the imidazole-buffered osmium tetroxide procedure.15,20,21
Analysis of Lipoprotein Particle Size
The d<1.052-g/mL fraction of plasma was prepared by ultracentrifugation, and lipoprotein particle diameters were determined by dynamic light scattering analysis with a Microtrac Series 150 Ultrafine particle analyzer fitted with a flexible conduit-sheathed probe tip (UPA-150; Microtrac).5 Data were plotted as the percentage of total particles in different particle size ranges.
Analysis of Mouse Atherosclerosis
Reversa mice were placed on a high-fat diet at 4 weeks of age. One group (n=6) was immediately treated with pI-pC, and the other group (n=8) was untreated. After 20 weeks on the high-fat diet, both groups of mice were euthanized. Mice were perfused with PBS followed by a fixative solution (4% paraformaldehyde, 5% sucrose, 20 mmol/L EDTA; pH 7.4). With the major branching vessels attached, the aorta was opened longitudinally from the iliac arteries to the aortic root. Then, all branching vessels were removed, and the aorta was pinned out flat on a black wax surface. Images of the pinned-out aortas were recorded with a Polaroid digital microscope camera. Each image was analyzed with Adobe Photoshop 5.5 (Adobe Systems) with gray and white threshold to define lesions, and the percentage of the surface covered by lesions was calculated.5 Lesions were quantified by a trained observer blinded to treatment group.
Analysis of Arterial Gene Expression
Reversa mice [either nontreated (n=5) or pI-pC–treated (n=4)] were weaned onto a high-fat diet. After 20 weeks, the mice were perfused with PBS, and the thoracic aorta was removed. The thoracic aorta was then separated into 2 segments (the aortic arch and the lower thoracic aorta) by cutting the aorta 3 mm distal to the origin of the left subclavian artery. RNA from the pooled aortic tissues was prepared with the RNeasy kit (Qiagen). To eliminate the possibility of DNA contamination, each RNA sample was treated with DNase (2.0 U, DNA-free, Ambion) in the presence of RNase inhibitor (RNaseOut, Invitrogen Life Technologies). Reverse transcription of RNA was performed with an oligo(dT), and the expression levels of CD68, vascular cell adhesion molecule-1 (VCAM-1), monocyte chemoattractant protein-1 (MCP-1), and intercellular adhesion molecule-1 (ICAM-1) were determined by quantitative real-time PCR on an ABI Prism 7700 Sequence Detection System (Applied Biosystems), as previously described.10 Expression levels were normalized to the expression of cyclophilin A.10
Data are expressed as mean±SEM. Measurements were analyzed with Primer of Biostatistics Software (Version 3.0, McGraw Hill, 1992). A Student’s 2-tailed t test was used to compare lipid levels before and after treatment.
Inactivating Mttp in Reversa Mice With pI-pC
Treatment of Reversa mice with pI-pC was effective in inducing recombination within the Mttp in the liver. Southern blotting revealed that the vast majority of liver genomic DNA had undergone recombination, converting the Mttpfl allele to a MttpΔ allele (Figure 1A). Small amounts of recombination were observed in tail DNA. The pI-pC treatment nearly abolished Mttp mRNA levels, as judged by quantitative real-time PCR (Figure 1B), and Western blotting revealed a striking reduction in plasma apo-B100 levels (Figure 1C). The presence of some apo-B100 in the plasma of the pI-pC–treated mice was expected. Apo-B100 is produced by the intestines of Reversa mice, and the extent of Mttp inactivation in that tissue is minimal.15
Ultrastructural examination of liver samples revealed that VLDL production was abolished in the pI-pC–treated Reversa mice (Figure 2). Hepatocytes from untreated Reversa mice (Figure 2A) contained numerous lipoprotein particles in the Golgi stacks and vesicles. In the pI-pC–treated mice, VLDL particles were absent, although there were a few small, irregularly shaped lipid-staining particles in the Golgi stacks (Figure 2B). Of note, the inactivation of Mttp was virtually complete in Reversa mice. By electron microscopy, we could not identify any VLDL-producing hepatocytes in pI-pC–treated mice. Light-level histological studies showed an increased amount of neutral lipids in the livers of the pI-pC–treated mice (not shown).
Plasma Lipid Levels in Reversa Mice Fall After Induction of Cre Expression
In mice on a chow diet, the plasma cholesterol levels fell from 233.9±46 to 37.2±6.5 mg/dL (n=15; P<0.001) after pI-pC treatment (Figure 3A), and the triglycerides fell from 70.2±7.88 to 16.3±8.4 mg/dL (n=12; P<0.001) (Figure 3B). In mice on the high-fat diet, plasma cholesterol levels fell from 525.7±32.24 to 100.6±14.33 mg/dL (n=16; P<0.001) (Figure 3A), and triglyceride levels fell from 110.33±6.45 to 23.4±4.56 mg/dL (n=14; P<0.001) (Figure 3B). The plasma lipid levels reached their nadirs ≈8 to 10 days after the first pI-pC injection (Figure 3, C and D).
The distribution of lipids within the plasma lipoproteins was assessed by FPLC fractionation studies (Figure 4). In the plasma of chow-fed Reversa mice, there was a large IDL/LDL cholesterol peak. This peak was virtually eliminated by pI-pC treatment (Figure 4A). On the high-fat diet, the cholesterol levels were higher, but the findings were similar (Figure 4B). On both diets, HDL cholesterol levels were lower in the pI-pC–treated mice. The pI-pC treatment caused a striking reduction in the triglyceride contents of VLDL, IDL, and LDL on both diets (Figure 4, C and D).
Spectrum of Lipoprotein Particle Size Shifts After Induction of Cre Expression
In chow-fed mice under nonfasting conditions, most of the lipoproteins in the plasma of the Reversa mice were in the LDL size range (≈20 nm) (Figure 5). The same was the case for Ldlr−/−Apob100/100 controls. After treatment of the Reversa mice with pI-pC, the lipoproteins that remained in nonfasting plasma had a bimodal size distribution. A sizable percentage of the particles were immense, with diameters >150 nm. A second peak in the particle size distribution occurred at ≈33 nm. We suspect that the larger particles represented nascent chylomicrons, because that peak was completely absent after an overnight fast (not shown). The smaller particles, which at ≈33 nm were significantly larger than LDL, were probably chylomicron remnants.
We strongly suspect that the chylomicron particles were also present in the plasma of untreated Reversa mice (no pI-pC). However, the large particles represented a very tiny percentage of the total particles in those animals and therefore did not appear on the particle distribution plot. In the pI-pC–treated mice, the elimination of hepatic lipoproteins unmasked the presence of the very large intestinal lipoproteins.
Prevention of Atherosclerosis by Induction of Cre Expression
To determine whether Cre-mediated inactivation of Mttp prevents atherosclerosis, Reversa mice (n=6) were treated at 4 weeks of age with pI-pC and then placed on a high-fat diet. Littermate controls (n=8) did not receive pI-pC. After 20 weeks, atherosclerotic lesions covered 5.2±0.8% of the surface of the aorta in the untreated mice. Lesions were prominent in the aortic arch, but only tiny lesions were present in the remainder of the aorta, primarily at vessel branch points (Figure 6A). In contrast, pI-pC–treated mice had no detectable lesions (Figure 6B).
Arterial Wall Gene Expression in Reversa Mice
We suspected that untreated Reversa mice (no pI-pC) would have greater expression of CD68 (a macrophage marker) in the aorta than the pI-pC–treated mice. We also suspected that the untreated mice would have higher mRNA levels for VCAM-1, MCP-1, and ICAM-1. Indeed, this proved to be the case (Figure 7). Expression levels of CD68, ICAM-1, and MCP-1 in the aortic arch were much higher in untreated mice than in pI-pC–treated mice. CD68, VCAM-1, and MCP-1 expression levels were much lower in the lower thoracic aorta, which contained far fewer lesions. ICAM-1 expression in the aortic arch was also greater in untreated mice than in pI-pC–treated mice (Figure 7), although the relative difference was less than with the other genes. ICAM-1 expression seems to be more dependent on shear stress than on plasma lipid levels.22
In this study, we produced Reversa mice, a hypercholesterolemic mouse model in which high plasma cholesterol levels can be eliminated, without hypolipidemic drugs or changing the diet, by flipping a genetic switch. At baseline, Reversa mice had high levels of LDL in their plasma, very similar to the lipoprotein profile of most humans with coronary artery disease.23 After induction of Mx1-Cre expression with pI-pC, hepatic Mttp mRNA levels fell precipitously, lipoprotein production in the liver was abolished, and LDL cholesterol in the plasma was virtually eliminated. The striking change in LDL cholesterol levels can be achieved without affecting the overall nutritional status of the mice. The pI-pC injections do not cause significant levels of recombination in intestinal enterocytes.15 After pI-pC, Reversa mice appeared to be healthy and exhibited normal levels of vitality. They consumed normal amounts of food (H. Lieu, unpublished observations) and had very large chylomicrons in their plasma.
In an earlier study, we demonstrated that Ldlr−/−Apob100/100 mice have high LDL cholesterol levels and develop severe atherosclerosis after 40 weeks on a chow diet, with ≈10% to 15% of the surface of the aorta covered with sudanophilic lesions.5 In the present study, we sought to determine whether the blockade of hepatic lipoprotein secretion would eliminate the heightened susceptibility of those mice to atherosclerosis. We quantified atherosclerosis in pI-pC–treated Reversa mice and untreated controls, choosing a high-fat diet instead of a chow diet so that the study could be completed in a shorter period of time. Our findings were unequivocal: untreated Reversa mice developed severe atherosclerotic lesions after 20 weeks on the diet, whereas the pI-pC–treated mice had no lesions. Thus, hepatic lipoproteins are clearly the culprits in promoting atherogenesis in Reversa mice, not chylomicrons or chylomicron remnants. Mttp is not expressed in the cells of the arterial wall, so we can be confident that the elimination of atherosclerotic lesions was a result of the change in the plasma lipoprotein profile rather than a direct effect of Mttp inactivation in the arterial wall.
The elimination of hepatic lipoprotein assembly and secretion in Reversa mice was associated with a reduction in the plasma levels of HDL cholesterol. A similar decrease in HDL cholesterol was observed when chylomicron assembly and secretion were eliminated with genetic approaches.20 The reduction in HDL cholesterol levels in the setting of reduced VLDL or chylomicron secretion is not surprising, because the biogenesis of mature HDL depends on the delivery of surface lipids (phospholipids and free cholesterol) from the triglyceride-rich lipoproteins to nascent apo-AI–containing particles.24 It is conceivable that the low HDL cholesterol levels in pI-pC–treated Reversa mice are proatherogenic, but if so, that effect is dramatically and unequivocally trumped by the antiatherogenic effect of the nearly complete absence of LDL cholesterol.
Reversa mice should expand experimental possibilities in atherosclerosis research, for at least 2 reasons. First, because the change in plasma cholesterol levels in Reversa mice is so striking, we suspect that this model will make it possible to study atherosclerosis regression in the mouse. We would be surprised if the elimination of LDL cholesterol did not cause regression of atherosclerotic lesions, as judged either by Sudan IV staining or by the quantification of cholesteryl esters in aortic lesions.5 Perhaps more interesting, however, will be to determine whether and to what extent the elimination of plasma LDL cholesterol causes the disappearance of free cholesterol crystals, macrophages, and the fibrous tissue within lesions. Second, because plasma cholesterol levels in Reversa mice can be titrated (by adjusting the dose of pI-pC), it will be possible to design better control groups for certain types of atherosclerosis experiments, making it possible to sharpen certain experimental conclusions. For example, in assessing the impact of certain antioxidants or certain dietary fatty acids on atherogenesis, it has occasionally been difficult to distinguish between the direct effects of the antioxidants or fatty acids on the development of lesions from the indirect effects of those agents on plasma cholesterol levels. If one were to perform those experiments in Reversa mice, it would be possible to titrate pI-pC dose so as to obtain groups of experimental animals that were matched for plasma cholesterol levels.
In the present study, we demonstrate that inhibition of hepatic lipoprotein secretion lowered the expression of several genes in the arterial wall that are associated with atherogenesis. The expression of CD68, VCAM-1, and MCP-1 in aortic tissue was strikingly lower in pI-pC–treated Reversa mice than in untreated controls. In part, this difference undoubtedly reflects different numbers of macrophages within the arterial wall tissue. However, the higher expression levels in the untreated Reversa mice could be in part a result of activation of the macrophages by the high levels of apo-B100–containing lipoproteins (and oxidized lipoproteins25) within the arterial wall. In a recent publication, Trogan et al10 used laser-capture microdissection techniques to demonstrate that the expression of CD68, VCAM-1, and MCP-1 in arterial wall macrophages was increased by activating macrophages with bacterial lipopolysaccharides. It would be interesting to use the same experimental techniques to determine whether activated macrophages in atherosclerotic lesions could be deactivated by lowering plasma LDL cholesterol levels. If so, would the macrophage deactivation require a threshold level of cholesterol lowering? Also, would the macrophage deactivation occur immediately or only several months later? The existence of Reversa mice now makes it practical to answer these questions.
Reversa mice can now be ordered from The Jackson Laboratory (http://jaxmice.jax.org/jaxmicedb/html/gemm_target.shtml).
This work was supported in part by National Heart, Lung, and Blood Institute (NHLBI) grant 41633 (to S.G.Y.), National Institute of Environmental Health Sciences center grant 09106 (R.L.W), NHLBI grant HL-61814 (to E.A.F.), a grant from the University of California Tobacco-Related Disease Program, and a National Institutes of Health Institutional Training grant to the University of California, San Francisco (to H.D.L). We thank B. Young for assistance with graphics and G. Howard and S. Ordway for helpful comments on the manuscript.
Reddick RL, Zhang SH, Maeda N. Atherosclerosis in mice lacking apo E: evaluation of lesional development and progression. Arterioscler Thromb. 1994; 14: 141–147.
Nakashima Y, Plump AS, Raines EW, et al. ApoE-deficient mice develop lesions of all phases of atherosclerosis throughout the arterial tree. Arterioscler Thromb. 1994; 14: 133–140.
Plump AS, Scott CJ, Breslow JL. Human apolipoprotein A-I gene expression increases high density lipoprotein and suppresses atherosclerosis in the apolipoprotein E-deficient mouse. Proc Natl Acad Sci U S A. 1994; 91: 9607–9611.
Ishibashi S, Herz J, Maeda N, et al. The two-receptor model of lipoprotein clearance: tests of the hypothesis in “knockout” mice lacking the low density lipoprotein receptor, apolipoprotein E, or both proteins. Proc Natl Acad Sci U S A. 1994; 91: 4431–4435.
Dansky HM, Charlton SA, Harper MM, et al. T and B lymphocytes play a minor role in atherosclerotic plaque formation in the apolipoprotein E-deficient mouse. Proc Natl Acad Sci U S A. 1997; 94: 4642–4646.
Breslow JL, Plump A, Dammerman M. New mouse models of lipoprotein disorders and atherosclerosis. In: Fuster V, Ross R, Topol EJ, eds. Atherosclerosis and Coronary Artery Disease. Vol 1. Philadelphia, Pa: Lippincott-Raven; 1996: 363–378.
Trogan E, Choudhury RP, Dansky HM, et al. Laser capture microdissection analysis of gene expression in macrophages from atherosclerotic lesions of apolipoprotein E-deficient mice. Proc Natl Acad Sci U S A. 2002; 99: 2234–2239.
Bisgaier CL, Essenburg AD, Auerbach BJ, et al. Attenuation of plasma low density lipoprotein cholesterol by select 3-hydroxy-3-methylglutaryl coenzyme A reductase in mice devoid of low density lipoprotein receptors. J Lipid Res. 1997; 38: 2502–2515.
Farese RV Jr, Véniant MM, Cham CM, et al. Phenotypic analysis of mice expressing exclusively apolipoprotein B48 or apolipoprotein B100. Proc Natl Acad Sci U S A. 1996; 93: 6393–6398.
Kühn R, Schwenk F, Aguet M, et al. Inducible gene targeting in mice. Science. 1995; 269: 1427–1429.
McCormick SPA, Ng JK, Véniant M, et al. Transgenic mice that overexpress mouse apolipoprotein B: evidence that the DNA sequences controlling intestinal expression of the apolipoprotein B gene are distant from the structural gene. J Biol Chem. 1996; 271: 11963–11970.
Hamilton RL, Wong JS, Cham CM, et al. Chylomicron-sized lipid particles are formed in the setting of apolipoprotein B deficiency. J Lipid Res. 1998; 39: 1543–1557.
Nakashima Y, Raines EW, Plump AS, et al. Upregulation of VCAM-1 and ICAM-1 at atherosclerosis-prone sites on the endothelium in the apoE-deficient mouse. Arterioscler Thromb Vasc Biol. 1998; 18: 842–851.
Havel RJ, Kane JP. Introduction: structure and metabolism of plasma lipoproteins. In: Scriver CR, Beaudet AL, Sly WS, et al., eds. The Metabolic and Molecular Bases of Inherited Disease. Vol 2. 8th ed. New York, NY: McGraw-Hill; 2001: 2705–2716.