Harmful Effects of Increased LDLR Expression in Mice With Human APOE*4 But Not APOE*3
http://www.100md.com
《动脉硬化血栓血管生物学》
From the Curriculum in Genetics and Molecular Biology (S.I.M., C.K., N.M.) and Department of Pathology and Laboratory Medicine (M.K.A., N.M.), University of North Carolina, Chapel Hill, and Department of Pathology (L.L.-F., J.S.P.), Wake Forest University School of Medicine, Winston-Salem, NC.
Correspondence to Nobuyo Maeda, Department of Pathology and Laboratory Medicine, University of North Carolina, Chapel Hill, NC 27599. E-mail nobuyo@med.unc.edu
Abstract
Objective— Increased expression of the low-density lipoprotein receptor (LDLR) is generally considered beneficial for reducing plasma cholesterol and atherosclerosis, and its downregulation has been thought to explain the association between apolipoprotein (apo) E4 and increased risk of coronary heart disease in humans.
Methods and Results— Contrary to this hypothesis, doubling Ldlr expression caused severe atherosclerosis with marked accumulation of cholesterol-rich, apoE-poor remnants in mice with human apoE4, but not apoE3, when the animals were fed a Western-type diet. The increased Ldlr expression enhanced in vivo clearance of exogenously introduced remnants in mice with apoE4 only when the remnants were already enriched with apoE4. The rates of nascent lipoprotein production were the same. The adverse effects of increased LDLR suggest a possibility that the receptor can trap apoE4, reducing its availability for the transfer to nascent lipoproteins needed for their rapid clearance, thereby increasing the production of apoE-poor remnants that are slowly cleared. The lower affinity for the LDLR of apoE3 compared with apoE4 could then explain why increased receptor expression had no adverse effects with apoE3.
Conclusions— Our results emphasize the occurrence of important and unexpected interactions between APOE genotype, LDLR expression, and diet.
Key Words: apolipoprotein E isoforms ? atherosclerosis ? genetic interaction ? lipid metabolism ? postprandiol hypercholesterolemia
Introduction
Apolipoprotein E (apoE) plays a central role in the clearance of atherogenic lipoprotein particles from the circulation.1 The APOE gene in humans is polymorphic with 3 common alleles, APOE*2, APOE*3, and APOE*4, which code for apoE2, apoE3, and apoE4. These isoforms differ by the amino acids at positions 112 and 158, where apoE2 has Cys at both sites, apoE4 has Arg at both sites, and apoE3 has Cys-112 and Arg-158. There is a well-established association between the APOE polymorphism and the risk for vascular diseases; individuals with APOE*4 allele have increased plasma cholesterol and an increased risk of atherosclerosis.2,3 Although these increases are modest, they have a large impact on the overall human population, because 25% carry 1 or 2 APOE*4 alleles. However, this association is paradoxical when one considers that apoE4 binds to the LDL receptor (LDLR) with equal or a slightly greater affinity than apoE3, the most common isoform.4–6 Additionally, individuals homozygous for apoE2, which binds to the LDLR with much less affinity than either apoE3 or apoE4, have low plasma cholesterol and are generally protected from atherosclerosis, except for the 5% to 10% of apoE2 homozygotes who develop type III hyperlipoproteinemia.1 The present explanation of this paradox is that the high affinity of apoE4 for the receptor leads to increased apoE-mediated cholesterol uptake followed subsequently by downregulation of the LDLR gene. This then leads to reduced apoB100-mediated uptake of LDL, accumulation of LDL cholesterol, and the vascular problems.2,7,8 Conversely, the low affinity of apoE2 is thought to lead to upregulation of LDLR. Although this explanation seems reasonable given that the loss of function of even one LDLR allele leads to markedly elevated plasma LDL and premature atherosclerosis,9,10 it has not been proven.
In the present study, we find that, contrary to the expectations of this hypothesis, increased Ldlr expression in mice with human APOE*4 causes severe atherosclerosis with marked elevation of plasma cholesterol when they are fed a Western-type diet. Mice with APOE*3, on the other hand, are not harmed by the increase in Ldlr expression. Based on these studies, we propose an alternative mechanism that the increased amount of LDLR can trap apoE and deplete the pool of apoE transferable to nascent lipoproteins.
Methods
Mice heterozygous for targeted replacement of the mouse Ldlr gene with the human LDLR minigene (Apoe+/+Ldlrh/+)11 were bred to mice homozygous for replacement of the mouse apoE gene with either the human APOE*3 or APOE*4 allele (Apoe3/3Ldlr+/+ and Apoe4/4Ldlr+/+).6,12 The experimental animals were mostly littermates generated by crossing Apoe3/3Ldlr+/+ (3m) with Apoe3/3Ldlrh/+ (3h) and Apoe4/4Ldlr+/+ (4m) with Apoe4/4Ldlrh/+ (4h), respectively. Mice with human APOE*4 and lacking LDLR (4KO) were generated by crossing Apoe4/4 and Ldlr-/- mice (see Table 1 for the definition of the shorthand designation of the mice used in this study). All mice were hybrids between 129 and C57BL/6, having approximately one fourth to one eighth of their genome from 129 and the remaining three fourths to seven eighths from C57BL/6. Twelve- to 36-week-old mice of both sexes were used for experiments that were conducted under protocols approved by the Institutional Animal Care and Use Committees. Littermates were used in each experiment as much as possible. Mice were fed either normal mouse chow (NC) containing 4.5% (wt/wt) fat and 0.022% (wt/wt) cholesterol (Prolab RMH 3000, Agway Inc) or a high-fat Western-type diet (HFW) containing 21% (wt/wt) fat and 0.2% (wt/wt) cholesterol (TD88137; Teklad). Experimental procedures can be found online at http://www.atvb.ahajournals.org.
TABLE 1. Keys to the Shorthand Definition of Mice Described in This Study
Results
Increased Ldlr expression causes hypercholesterolemia in mice with human APOE*4 We replaced the endogenous mouse Ldlr with a minigene coding for human LDLR (Ldlrh) that produces an mRNA with an increased half-life11 and introduced 1 copy of this Ldlrh allele into mice expressing solely apoE3 (Apoe3/3) or apoE4 (Apoe4/4).6,12 When mice were fed NC, the increase in Ldlr expression significantly lowered the plasma lipids in both Apoe3/3Ldlrh/+ (3h) and Apoe4/4Ldlrh/+ (4h) mice, relative to the Apoe3/3Ldlr+/+ (3m) or Apoe4/4Ldlr+/+ (4m) mice (Table 2). The Ldlr genotype had highly significant effects on total cholesterol (TC), triglyceride (TG), and HDL cholesterol (HDL-C) (P<0.0001). The effects of the Apoe genotype on TC and HDL-C were not significant, but females with apoE4 tended to have higher TG than those with apoE3 whereas males with apoE3 tended to have higher TG than those with apoE4 (P=0.0002 for Apoe, sex interaction). All classes of plasma lipoproteins including HDL were reduced in mice with increased LDLR, as assayed by FPLC analysis (Figure 1A, ).
TABLE 2. Plasma TC, TG, and HDL-C
Figure 1. Plasma lipoproteins. A, FPLC of plasma lipoproteins of mice fed NC ( and HFW (?). Plasma was pooled from at least 6 male mice of each genotype. Fractions containing VLDL, LDL, and HDL are indicated. B, SDS polyacrylamide gel electrophoresis. Plasma was collected from mice fed HFW and separated by ultracentrifugation. Lanes 1 through 7 are density fractions d<1.006, 1.006
Feeding a HFW increased the plasma TC and the HDL-C levels of all of the mice (Table 2). Surprisingly, however, the increase in plasma TC was much greater in the 4h mice than in 4m mice (120±11 versus 32±5 mg/dL, P<0.0001). This increase resulted mainly from a dramatic accumulation of non-HDL particles that elute in the VLDL region during fast performance liquid chromatography (FPLC, Figure 1A, ?). In contrast, the 3h mice on HFW showed only a small increase in non-HDL particles, and they had significantly lower cholesterol levels than the 3m mice primarily because of reduced HDL. Thus, the Ldlr genotype has markedly different effects on the response to HFW in mice with apoE4 compared with those with apoE3.
The 4h remnants were mostly in very low to intermediate density fractions (d<1.02 g/mL) by ultracentrifugation and were enriched in TC but poor in TG, with a TC/TG ratio of 5.3 compared with 0.6, 0.6, and 1.2 in 4m, 3m, and 3h remnants, respectively. The apolipoprotein compositions of the VLDL fraction also differed significantly (Figure 1B). Densitometric analysis of at least 4 different preparations of the VLDL fractions showed that the 4h remnants had a marked reduction of apoE4 (x0.4) but increased apoB48 (x6.7) and apoAIV (x4.7) compared with 4m remnants. ApoE3 in the 3h remnants was also reduced (x0.5) compared with 3m remnants, but the increase in apoB48 (x3) was less prominent. The 4h remnant fraction contains an average of 4.5 times more apoB proteins than that of the 3h mice and had a smaller apoE/apoB ratio (8±1% relative to that in 4m) compared with 36±4% in 3h fraction (P<0.005). Thus, the remnants of the 4h mice are cholesterol-rich, TG-poor, and apoE-poor compared with those of the 3h mice.
The enrichment of apoAIV in the 4h remnants suggests that they are mainly from intestine-derived chylomicrons. Consistent with this possibility, plasma cholesterol levels declined steadily in fasting 4h mice but not in 4m mice (Figure 2A) and remnant particles were reduced in 4h mice fasted for 18 hours or longer. There were no significant differences in the TC and TG content in the livers of these mice. The relative amounts of apoE protein estimated by Western blot analysis indicate that the elevated LDLR expression increases liver-associated apoE by 30% but reduces plasma apoE by 60% regardless of their Apoe genotype.
Figure 2. Effects of increased LDLR expression on mice with human apoE4 on HFW. A, Fasting effects on plasma cholesterol levels in male 4h (?, n=8) and 4m (, n=8) mice on HFW. Error bars are SEM. P<0.0001 by MANOVA. *P<0.001 and P<0.05 between genotypes by 2-tailed Student’s t test. B, Liver mRNA levels for the mouse Ldlr (black bars) and for the human Ldlr (white bars) expressed as percent of the Ldlr gene expression in 4m mice on NC. Numbers of animals (all males) are shown in the bottom of the bars. Error bars are SEM. P<0.05 for diet effect, P<0.0005 for Ldlr genotype effect by 2-way ANOVA. The effect of Apoe genotype is not significant.
The level of mouse Ldlr mRNA in the liver of 4h mice with a single copy of the gene was 61% of that in 4m mice with 2 copies (100%, Figure 2B). The human Ldlr message was 155% (total, 216%). The regulatory machinery of the Ldlrh allele is intact, because HFW downregulated the mouse and human Ldlr messages equally to 60% of their levels in mice fed NC. The total Ldlr mRNA level in 4h mice fed HFW was 2-fold higher than in 4m mice on HFW and a trace higher than in 4m mice fed NC. Thus, the marked hypercholesterolemia in the 4h mice fed HFW occurs despite their having high levels of LDLR expression. Ldlr expression is similar in the livers of 4h and 3h mice on HFW, indicating that the hypercholesterolemia in the 4h mice, but not the 3h mice, is not attributable to any differences in the diet-induced downregulation of Ldlr expression in the liver.
In Vivo Lipoprotein Metabolism in 4h Mice
To estimate the production of VLDL, we injected Triton WR1339 into 4h and 4m mice fed the HFW to inhibit lipolysis and the uptake of TG-rich particles. The initial rate of TG accumulation in plasma was not different between the 4h and 4m mice (Figure 3A). Plasma TC increased steadily and equally in both mice. They also responded similarly to the fat loading. The rates of plasma TG increase after fat loading in 4h mice treated with Triton WR1339 (604±33 mg/dL per h, n=4) was similar to those in 4m mice (656±37 mg/dL per h, n=4). Thus, increased LDLR expression seems to have no effect on the chylomicron production. These data suggest that the dramatic accumulation of remnants in the 4h mice compared with 4m mice is not attributable to oversecretion of TG-rich particles.
Figure 3. In vivo lipoprotein metabolism. A, Triglyceride (left) and cholesterol (right) secretion of 4m ( and 4h (?) mice after inhibition of lipolysis with Triton WR1339. The overall rate of TG secretion was not significantly different between the genotypes by MANOVA. B, Clearance of remnants. Radioactivity remaining in the plasma after injection of I125-labeled VLDL obtained from 4h remnants into HFW-fed 4h males (n=10) and 4m males (n=9) was not significantly different at any time points except at 10 minutes (left, P<0.05). Clearance of the 131I radiolabeled apoE4-rich 4KO remnants (right) in 4h males (n=5) was significantly faster than in 4m males (n=5, P<0.0001 by MANOVA, and P<0.01 at all times up to 120 minutes).
To examine the ability of mice to clear cholesterol-enriched remnant particles, we isolated remnants (d<1.02) from HFW-fed 4h mice, radiolabeled the particles with 125I, and injected them into the jugular vein of 4h and 4m mice fed the HFW diet. Despite the increased Ldlr expression in the 4h mice, removal of the apoE-poor remnants from their plasma (0.04 pools/min) was not different from that in the 4m mice (0.04 pools/min, Figure 3B, left). Similar results were obtained with remnants (d<1.006) isolated from apoE-deficient mice (not shown). To test whether the 4h mice having increased LDLR are able to clear apoE4-enriched remnants, we isolated lipoproteins (d<1.006) from mice expressing apoE4 but lacking LDLR (4KO) that are markedly enriched with apoE but not with apoAIV (C. Knouff and N. Maeda, unpublished data). The plasma decay of 131I-radiolabeled 4KO remnants (Figure 3B, right) was significantly faster in the 4h mice (0.07 pools/min) than in the 4m mice (0.03 pools/min, P<0.0001). Thus, the 4h mice with genetically increased LDLR expression clear remnant lipoproteins at an enhanced rate as long as these particles are enriched in apoE.
Taken together, these data indicate that the marked accumulation of remnant particles in the 4h mice is neither because they have increased secretion of nascent TG-rich lipoproteins nor because they have reduced clearance of apoE-poor remnants compared with 4m mice. The inference from these data is that the conversion of large TG-rich particles to smaller cholesterol-rich remnants is enhanced in the 4h mice.
Severe Atherosclerosis in 4h But Not 3h Mice on HFW Diet
Accumulation of remnant particles and reduction of HDL-cholesterol in HFW-fed 4h mice is a high-risk profile for atherosclerosis, even though the average plasma total cholesterol of these mice is only marginally elevated (200 mg/dL). We therefore evaluated the development of atherosclerosis in mice fed HFW containing 21% fat and 0.5% cholesterol for 3 months. No plaques were found in any of the 3m mice (5 females) or 4m mice (8 females and 5 males) on HFW. Similarly, none of the 7 female 3h mice on HFW developed plaques (Figure 4A). In contrast, all of the 15 female and 7 male 4h mice on HFW developed significant plaques at the aortic sinus area (Figures 4B through 4D) with average plaque sizes of 59±15x103 μm2 in females and 22±7x103 μm2 in males. Although the numbers and sizes of plaques varied in individual animals, most of the mice had mature complex plaques with fibrous caps, necrotic lipid cores, cholesterol clefts, and calcifications (Figures 4C and 4D). Thus, mice having human APOE*4 and a moderately increased amount of LDLR develop significant atherosclerosis when fed a diet similar in composition to that of Western societies.
Figure 4. Representative atherosclerotic plaques in the aortic sinus of mice fed HFW for 3 months. A, 3h female; B, 4h female; C, 4h male; and D, 4h female. Frozen sections were stained with Sudan IVB for lipids and counterstained with hematoxylin. Black arrowheads indicate fibrous caps. Yellow and blue arrows designate calcification and cholesterol clefts, respectively. L indicates lumen; M, media; and A, adventitia. Scale bars=500 μm in a and b and 100 μm in c and d.
Discussion
Our clear demonstration of hypercholesterolemia and atherosclerosis in mice that have human apoE4, but not apoE3, replicates in mice fed a Western diet the paradoxical association between APOE polymorphisms and the risk of atherosclerosis in humans. However, contrary to the currently accepted hypothesis that downregulation of LDLR in individuals with an APOE*4 allele is the cause of their high plasma cholesterol levels,2,7,8 we find that these isoform-specific effects are only present when LDLR is increased.
How can increased LDLR expression ever be harmful? We suggest that this is because the LDLR, under some circumstances, traps sufficient apoE and the supply becomes inadequate to process a high dietary intake of lipids. Exchange of apoE onto nascent triglyceride-rich lipoproteins is a necessary prerequisite for their internalization via LDLR.13–17 The overall process that we envision is illustrated in Figure 5. ApoE4 may be particularly susceptible to this trapping because of its strong affinity for the LDLR. Transient particles that fail to acquire apoE4 are excellent substrates for lipases,1 and the resulting enhanced lipolysis will increase cholesterol-rich apoE-poor remnant particles in plasma. We suggest that apoE3 is less susceptible to trapping than apoE4, perhaps because of its somewhat lower affinity for the LDLR.5,6 Consequently, the 3h mice have sufficient apoE3 to process nascent lipoproteins for rapid internalization. This differential transfer of apoE3 and apoE4 to lipoproteins can also explain our previous observation that in vivo clearance of exogenously introduced remnants is significantly faster in 3m mice than in 4m mice.6
Figure 5. ApoE trapping by the LDLR. Triglyceride-rich chylomicrons (yellow) secreted by the intestine into lymph are remodeled to transient particles (orange) mainly in capillaries and in the space of Disse in the liver. This initial capturing phase is very rapid and does not depend on the LDLR; it probably involves heparan sulfate proteoglycans that facilitate lipolysis and apolipoprotein exchange. Enrichment with apoE is a requirement for the fast internalization of the transient particles primarily via LDLR-mediated endocytosis. Otherwise, the particles are additionally processed to cholesterol-enriched remnants (red) that are slowly cleared from the plasma. The lipid composition of the remnants in plasma depends on both Apoe genotype and the amount of LDLR. The processing of liver-derived VLDL particles is likely to be similar to that illustrated for chylomicrons.
Previously we reported that mice expressing solely apoE2 (2m) show features typical of type III hyperlipoproteinemia in humans18 but that increased LDLR expression in these mice (2h) completely ameliorates their hyperlipoproteinemia.11 According to our hypothesis, apoE2 with its very low affinity for the LDLR should be virtually free from trapping and should therefore be efficiently transferred to transient TG-rich particles. Although such an apoE2 enrichment of TG-rich particles is likely to increase their LDLR-mediated internalization, it will also severely inhibit lipolysis1 and could account for the prominent accumulation of TG-rich remnants seen in the circulation of the 2m mice.18 In the 2h mice, however, high LDLR expression tips the balance toward more internalization and lowers their plasma cholesterol.11
Clearly, additional studies are necessary to refine and test our proposed apoE trapping by the LDLR. For example, we do not know whether it occurs on the cell surface, as illustrated in Figure 5, or during intracellular trafficking.19,20 The word trapping should not be taken too literally; difference in the interaction between apoE and LDLR or their subsequent processing may result not only from differences in binding affinities but also from other properties influenced by the specific amino acids that differ among 3 isoforms. Interactions of apoE with other molecules, such as proteoglycans and hepatic lipase, that are also expressed on the basolateral microvilli of hepatocytes may also influence the apoE interaction with LDLR in an isoform-specific fashion. Published studies have shown that newly synthesized apoE is incompletely secreted and partially degraded in HepG2 cells in culture21 and a significant portion of apoE synthesized by macrophages undergoes rapid cellular degradation in a non-lysosomal compartment in a sterol-regulated manner.22 Although the role of LDLR in these processes has not been addressed, LDLR is known to bind to newly synthesized apoE in macrophages and limits its secretion.23 LDLR expression in macrophages could also contribute to atherogenesis in an apoE isoform–specific fashion. For example, a differential effect on cholesterol homeostasis in macrophages by apoE isoforms with apoE4 being least effective in promoting cholesterol efflux from macrophage has been reported.24 Additionally, Linton et al25 have shown that C57BL/6 mice receiving Ldlr-/- marrow developed 63% smaller lesions than mice receiving Ldlr+/+ marrow after dietary atherogenic stimuli, demonstrating that macrophage LDLR affects the rate of foam cell formation under conditions of dietary stress.
Some comments are required on the relevance of our findings in mice to the effects of different APOE genotypes in humans. We note that 4h mice preferentially accumulate apoB48-containing lipoproteins of an intermediate density, whereas humans with the APOE*4 allele mainly have elevated levels of apoB100-containing LDL.2 This is not incompatible with our hypothesis, which predicts that trapping of apoE4 by the LDLR will hinder enrichment of VLDL remnants with apoE4, thereby leading to an increase in their conversion to LDL. Because the clearance of LDL particles mediated by binding apoB100 to the receptor is much slower than apoE-mediated VLDL clearance,13,15 we expect that the plasma cholesterol levels in individuals with apoE4 will be increased. Supporting this explanation, an increased conversion of VLDL to smaller remnants and a relative decrease in direct removal of VLDL in APOE*4 homozygotes compared with APOE*3 subjects have been demonstrated.26
Our finding that hypercholesterolemia is seen only when the 4h mice are fed HFW diet is also consistent with observations that human subjects carrying the APOE*4 allele are more responsive than others to LDL cholesterol–lowering by diet.27,28 In addition, some though not all studies have found prolonged postprandial lipemia in normolipidemic subjects that carry APOE*4.29–32 Bergeron and Havel29 have proposed that prolonged residence times of chylomicron and VLDL remnants in persons with APOE*4 raise the concentration of LDL by increasing the amount of VLDL converted to LDL. We additionally note that an apoE5 variant with lysine in place of glutamic acid at position 3 is also associated with hyperlipidemia and atherosclerosis and it has a twice-normal LDLR binding activity.33 Finally, some although not all studies have shown that the cholesterol-lowering effects of statins, thought to be primarily mediated by increased LDLR, are apoE isoform–dependent. In these studies, individuals with the APOE*3/4 and APOE*4/4 genotypes had significantly smaller LDL cholesterol reductions in response to statin treatment than those with the APOE*3/3 genotype.34,35 Clinical studies clearly indicate that statin therapies reduce the risk of cardiovascular disease in humans, including those with apoE4, and no serious adverse effects on plasma lipids have been reported.35,36 Nevertheless, our observations suggest the need for additional studies of the interaction between the cholesterol-lowering effect of statins and genetic variations.
In conclusion, our studies demonstrate that, contrary to the presently accepted downregulation of LDLR hypothesis, increased LDLR has harmful effects in Western diet–fed mice expressing human apoE4 and causes marked accumulation of apoE-poor lipoprotein remnants in plasma and severe atherosclerosis. The alternative mechanism of apoE-trapping by LDLR explains our observations and offers a plausible explanation why apoE4, which has a greater affinity for the LDLR than apoE3, is associated with higher plasma cholesterol and a greater risk of atherosclerosis in humans. Our unexpected findings in mice predict important interactions between APOE genotype, LDLR expression, and diet.
Acknowledgments
This work was supported by grants from NIH (HL42360 to N.M., HL54176 and HL49373 to J.S.P, and HL07115 to L.L.F). The authors thank Dr Tom Smith for the jugular vein injection during the lipoprotein clearance experiments, Drs Greg Shelness, Leighton James, and Oliver Smithies for discussion, and Jennifer Altenburg and Ellen Young for technical assistance.
References
Mahley RW, Rall SC Jr. Type III hyperlipoproteinemia (dysbetalipoproteinemia): the role of apolipoprotein E in normal and abnormal lipoprotein metabolism. In: Scriver AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inheritable Disease. New York: MacGraw Hill; 2001: 2835–2913.
Davignon J, Gregg RE, Sing CF. Apolipoprotein E polymorphism and atherosclerosis. Arteriosclerosis. 1988; 8: 1–21.
Boerwinkle E, Utermann G. Simultaneous effects of the apolipoprotein E polymorphism on apolipoprotein E, apolipoprotein B, and cholesterol metabolism. Am J Hum Genet. 1988; 42: 104–112.
Weisgraber KH. Apolipoprotein E: structure-function relationships. Adv Protein Chem. 1994; 45: 249–302.
Bohnet K, Pillot T, Visvikis S, Sabolovic N, Siest G. Apolipoprotein (apo) E genotype and apoE concentration determine binding of normal very low density lipoproteins to HepG2 cell surface receptors. J Lipid Res. 1996; 37: 1316–1324.
Knouff C, Hinsdale ME, Mezdour H, Altenburg M, Watanabe M., Quarfordt SH, Sullivan PM, Maeda N. Apo E structure determines VLDL clearance and atherosclerosis risk in mice. J Clin Invest. 1999; 103: 1579–1586.
Gregg RE, Zech LA, Schaefer EJ, Stark D, Wilson D, Brewer Jr. HB. Abnormal in vivo metabolism of apolipoprotein E4 in humans. J. Clin Invest. 1986; 78: 815–821.
Weintraub MS, Eisenberg S, Breslow JL. Dietary fat clearance in normal subjects is regulated by genetic variation in apolipoprotein E. J Clin Invest. 1987; 80: 1571–1577.
Brown MS, Goldstein JL. A receptor-mediated pathway for cholesterol homeostasis. Science. 1986; 232: 34–47.
Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia. In: Scriver AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inheritable Disease. New York: MacGraw Hill; 2001: 2863–2913.
Knouff C, Malloy S, Wilder J, Altenburg MK, Maeda N. Doubling expression of the low density lipoprotein receptor by truncation of the 3'-untranslated region sequence ameliorates type III hyperlipoproteinemia in mice expressing the human apoE2 isoform. J Biol Chem. 2001; 276: 3856–3862.
Sullivan PM, Mezdour H, Aratani Y, Knouff C, Najib J, Reddick RL, Quarfordt SH, Maeda N. Targeted replacement of the mouse apolipoprotein E gene with the common human APOE*3 allele enhances diet-induced hypercholesterolemia and atherosclerosis. J Biol Chem. 1997; 272: 17972–17980.
Cooper AD. Hepatic uptake of chylomicron remnants. J Lipid Res. 1997; 38: 2173–2192.
Herz J, Qiu SQ, Oesterle A, DeSilva HV, Shafi S, Havel RJ. Initial hepatic removal of chylomicron remnants is unaffected but endocytosis is delayed in mice lacking the low density lipoprotein receptor. Proc Natl Acad Sci U S A. 1995; 92: 4611–4615.
Mortimer BC, Beveridge DJ, Martins IJ, Redgrave TG. Intracellular localization and metabolism of chylomicron remnants in the livers of low density lipoprotein receptor-deficient mice and apoE-deficient mice: evidence for slow metabolism via an alternative apoE-dependent pathway. J Biol Chem. 1995; 270: 28767–28776.
Mahley RW, Ji ZS. Remnant lipoprotein metabolism: key pathways involving cell-surface heparan sulfate proteoglycans and apolipoprotein E. J Lipid Res. 1999; 40: 1–16.
Bradley WA, Gianturco SH. ApoE is necessary and sufficient for the binding of large triglyceride-rich lipoproteins to the LDLR: apoB is unnecessary. J Lipid Res. 1986; 27: 40–48.
Sullivan PM, Mezdour H, Quarfordt SH, Maeda N. Type III hyperlipoproteinemia and spontaneous atherosclerosis in mice resulting from gene replacement of mouse Apoe with human APOE*2.5 J Clin Invest. 1998; 102: 130–135.
Brown MS, Herz J, Goldstein JL. LDL-receptor structure: calcium cages, acid baths and recycling receptors. Nature. 1997; 388: 629–630.
Swift LL, Farkas MH, Major AS, Valyi-Nagy K, Linton MF, Fazio S. A recycling pathway for resecretion of internalized apolipoprotein E in liver cells. J Biol Chem. 2001; 276: 22965–22970.
Schmitt M, Grand-Perret T. Regulated turnover of a cell surface-associated pool of newly synthesized apolipoprotein E in HepG2 cells. J Lipid Res. 1999; 40: 39–49.
Duan H, Lin CY, Mazzone T. Degradation of macrophage ApoE in a nonlysosomal compartment: regulation by sterols. J Biol Chem. 1997; 272: 31156–1162.
Zhao Y, Mazzone T. LDL receptor binds newly synthesized apoE in macrophages: a precursor pool for apoE secretion. J Lipid Res. 1999; 40: 1029–1035.
Cullen P, Cignarella A, Brennhausen B, Mohr S, Assmann G, von Eckardstein A. Phenotype-dependent differences in apolipoprotein E metabolism and in cholesterol homeostasis in human monocyte-derived macrophages. J Clin Invest. 1998; 101: 1670–1677.
Linton MF, Babaev VR, Gleaves LA, Fazio S. A direct role for the macrophage low density lipoprotein receptor in atherosclerotic lesion formation. J Biol Chem. 1999; 274: 19204–19210.
Demant T, Bedford D, Packard CJ, Shepherd J. Influence of apolipoprotein E polymorphism on apolipoprotein B-100 metabolism in normolipemic subjects. J Clin Invest. 1991; 88: 1490–1501.
Lopez-Miranda J, Ordovas JM, Mata P, Lichtenstein AH, Clevidence B, Judd JT, Schaefer EJ. Effect of apolipoprotein E phenotype on diet-induced lowering of plasma low density lipoprotein cholesterol. J Lipid Res. 1994; 35: 1965–1975.
Dreon DM, Fernstrom HA, Miller B, Krauss RM. Apolipoprotein E isoform phenotype and LDL subclass response to a reduced-fat diet. Arterioscler Thromb Vasc Biol. 1995; 15: 105–111.
Bergeron N, Havel RJ. Prolonged postprandial responses of lipids and apolipoproteins in triglyceride-rich lipoproteins of individuals expressing an apolipoprotein epsilon 4 allele. J Clin Invest. 1996; 97: 65–72.
Superko HR, Haskell WL. The effect of apolipoprotein E isoform difference on postprandial lipoprotein in patients matched for triglycerides, LDL-cholesterol, and HDL-cholesterol. Artery. 1991; 18: 315–325.
Boerwinkle E, Brown S, Sharrett AR, Heiss G, Patsch W. Apolipoprotein E polymorphism influences postprandial retinyl palmitate but not triglyceride concentrations. Am J Hum Genet. 1994; 54: 341–360.
Dallongeville J, Tiret L, Visvikis S, O’Reilly DS, Saava M, Tsitouris G, Rosseneu M, DeBacker G, Humphries SE, Beisiegel U. Effect of apo E phenotype on plasma postprandial triglyceride levels in young male adults with and without a familial history of myocardial infarction: the EARS II study.European Atherosclerosis Research Study. Atherosclerosis. 1999; 145: 381–388.
Dong LM, Yamamura T, Yamamoto A. Enhanced binding activity of an apolipoprotein E mutant, APOE5, to LDLRs on human fibroblasts. Biochem Biophys Res Commun. 1990; 168: 409–414.
Ordovas JM, Lopez-Miranda J, Perez-Jimenez F, Rodriguez C, Park JS, Cole T, Schaefer EJ. Effect of apolipoprotein E and A-IV phenotypes on the low density lipoprotein response to HMG CoA reductase inhibitor therapy. Atherosclerosis. 1995; 113: 157–166.
Ballantyne CM, Herd JA, Stein EA, Ferlic LL, Dunn JK, Gotto AM Jr, Marian AJ. Apolipoprotein E genotypes and response of plasma lipids and progression-regression of coronary atherosclerosis to lipid-lowering drug therapy. J Am Coll Cardiol. 2000; 36: 1572–1578.
Gerdes LU, Gerdes 2C, Kervinen K, Savolainen M, Klausen IC, Hansen PS, Kesaniemi YA, Faergeman O. The apolipoprotein epsilon4 allele determines prognosis and the effect on prognosis of simvastatin in survivors of myocardial infarction: a substudy of the Scandinavian simvastatin survival study. Circulation. 2000; 101: 1366–1371.(Sudi I. Malloy*; Michael )
Correspondence to Nobuyo Maeda, Department of Pathology and Laboratory Medicine, University of North Carolina, Chapel Hill, NC 27599. E-mail nobuyo@med.unc.edu
Abstract
Objective— Increased expression of the low-density lipoprotein receptor (LDLR) is generally considered beneficial for reducing plasma cholesterol and atherosclerosis, and its downregulation has been thought to explain the association between apolipoprotein (apo) E4 and increased risk of coronary heart disease in humans.
Methods and Results— Contrary to this hypothesis, doubling Ldlr expression caused severe atherosclerosis with marked accumulation of cholesterol-rich, apoE-poor remnants in mice with human apoE4, but not apoE3, when the animals were fed a Western-type diet. The increased Ldlr expression enhanced in vivo clearance of exogenously introduced remnants in mice with apoE4 only when the remnants were already enriched with apoE4. The rates of nascent lipoprotein production were the same. The adverse effects of increased LDLR suggest a possibility that the receptor can trap apoE4, reducing its availability for the transfer to nascent lipoproteins needed for their rapid clearance, thereby increasing the production of apoE-poor remnants that are slowly cleared. The lower affinity for the LDLR of apoE3 compared with apoE4 could then explain why increased receptor expression had no adverse effects with apoE3.
Conclusions— Our results emphasize the occurrence of important and unexpected interactions between APOE genotype, LDLR expression, and diet.
Key Words: apolipoprotein E isoforms ? atherosclerosis ? genetic interaction ? lipid metabolism ? postprandiol hypercholesterolemia
Introduction
Apolipoprotein E (apoE) plays a central role in the clearance of atherogenic lipoprotein particles from the circulation.1 The APOE gene in humans is polymorphic with 3 common alleles, APOE*2, APOE*3, and APOE*4, which code for apoE2, apoE3, and apoE4. These isoforms differ by the amino acids at positions 112 and 158, where apoE2 has Cys at both sites, apoE4 has Arg at both sites, and apoE3 has Cys-112 and Arg-158. There is a well-established association between the APOE polymorphism and the risk for vascular diseases; individuals with APOE*4 allele have increased plasma cholesterol and an increased risk of atherosclerosis.2,3 Although these increases are modest, they have a large impact on the overall human population, because 25% carry 1 or 2 APOE*4 alleles. However, this association is paradoxical when one considers that apoE4 binds to the LDL receptor (LDLR) with equal or a slightly greater affinity than apoE3, the most common isoform.4–6 Additionally, individuals homozygous for apoE2, which binds to the LDLR with much less affinity than either apoE3 or apoE4, have low plasma cholesterol and are generally protected from atherosclerosis, except for the 5% to 10% of apoE2 homozygotes who develop type III hyperlipoproteinemia.1 The present explanation of this paradox is that the high affinity of apoE4 for the receptor leads to increased apoE-mediated cholesterol uptake followed subsequently by downregulation of the LDLR gene. This then leads to reduced apoB100-mediated uptake of LDL, accumulation of LDL cholesterol, and the vascular problems.2,7,8 Conversely, the low affinity of apoE2 is thought to lead to upregulation of LDLR. Although this explanation seems reasonable given that the loss of function of even one LDLR allele leads to markedly elevated plasma LDL and premature atherosclerosis,9,10 it has not been proven.
In the present study, we find that, contrary to the expectations of this hypothesis, increased Ldlr expression in mice with human APOE*4 causes severe atherosclerosis with marked elevation of plasma cholesterol when they are fed a Western-type diet. Mice with APOE*3, on the other hand, are not harmed by the increase in Ldlr expression. Based on these studies, we propose an alternative mechanism that the increased amount of LDLR can trap apoE and deplete the pool of apoE transferable to nascent lipoproteins.
Methods
Mice heterozygous for targeted replacement of the mouse Ldlr gene with the human LDLR minigene (Apoe+/+Ldlrh/+)11 were bred to mice homozygous for replacement of the mouse apoE gene with either the human APOE*3 or APOE*4 allele (Apoe3/3Ldlr+/+ and Apoe4/4Ldlr+/+).6,12 The experimental animals were mostly littermates generated by crossing Apoe3/3Ldlr+/+ (3m) with Apoe3/3Ldlrh/+ (3h) and Apoe4/4Ldlr+/+ (4m) with Apoe4/4Ldlrh/+ (4h), respectively. Mice with human APOE*4 and lacking LDLR (4KO) were generated by crossing Apoe4/4 and Ldlr-/- mice (see Table 1 for the definition of the shorthand designation of the mice used in this study). All mice were hybrids between 129 and C57BL/6, having approximately one fourth to one eighth of their genome from 129 and the remaining three fourths to seven eighths from C57BL/6. Twelve- to 36-week-old mice of both sexes were used for experiments that were conducted under protocols approved by the Institutional Animal Care and Use Committees. Littermates were used in each experiment as much as possible. Mice were fed either normal mouse chow (NC) containing 4.5% (wt/wt) fat and 0.022% (wt/wt) cholesterol (Prolab RMH 3000, Agway Inc) or a high-fat Western-type diet (HFW) containing 21% (wt/wt) fat and 0.2% (wt/wt) cholesterol (TD88137; Teklad). Experimental procedures can be found online at http://www.atvb.ahajournals.org.
TABLE 1. Keys to the Shorthand Definition of Mice Described in This Study
Results
Increased Ldlr expression causes hypercholesterolemia in mice with human APOE*4 We replaced the endogenous mouse Ldlr with a minigene coding for human LDLR (Ldlrh) that produces an mRNA with an increased half-life11 and introduced 1 copy of this Ldlrh allele into mice expressing solely apoE3 (Apoe3/3) or apoE4 (Apoe4/4).6,12 When mice were fed NC, the increase in Ldlr expression significantly lowered the plasma lipids in both Apoe3/3Ldlrh/+ (3h) and Apoe4/4Ldlrh/+ (4h) mice, relative to the Apoe3/3Ldlr+/+ (3m) or Apoe4/4Ldlr+/+ (4m) mice (Table 2). The Ldlr genotype had highly significant effects on total cholesterol (TC), triglyceride (TG), and HDL cholesterol (HDL-C) (P<0.0001). The effects of the Apoe genotype on TC and HDL-C were not significant, but females with apoE4 tended to have higher TG than those with apoE3 whereas males with apoE3 tended to have higher TG than those with apoE4 (P=0.0002 for Apoe, sex interaction). All classes of plasma lipoproteins including HDL were reduced in mice with increased LDLR, as assayed by FPLC analysis (Figure 1A, ).
TABLE 2. Plasma TC, TG, and HDL-C
Figure 1. Plasma lipoproteins. A, FPLC of plasma lipoproteins of mice fed NC ( and HFW (?). Plasma was pooled from at least 6 male mice of each genotype. Fractions containing VLDL, LDL, and HDL are indicated. B, SDS polyacrylamide gel electrophoresis. Plasma was collected from mice fed HFW and separated by ultracentrifugation. Lanes 1 through 7 are density fractions d<1.006, 1.006
Feeding a HFW increased the plasma TC and the HDL-C levels of all of the mice (Table 2). Surprisingly, however, the increase in plasma TC was much greater in the 4h mice than in 4m mice (120±11 versus 32±5 mg/dL, P<0.0001). This increase resulted mainly from a dramatic accumulation of non-HDL particles that elute in the VLDL region during fast performance liquid chromatography (FPLC, Figure 1A, ?). In contrast, the 3h mice on HFW showed only a small increase in non-HDL particles, and they had significantly lower cholesterol levels than the 3m mice primarily because of reduced HDL. Thus, the Ldlr genotype has markedly different effects on the response to HFW in mice with apoE4 compared with those with apoE3.
The 4h remnants were mostly in very low to intermediate density fractions (d<1.02 g/mL) by ultracentrifugation and were enriched in TC but poor in TG, with a TC/TG ratio of 5.3 compared with 0.6, 0.6, and 1.2 in 4m, 3m, and 3h remnants, respectively. The apolipoprotein compositions of the VLDL fraction also differed significantly (Figure 1B). Densitometric analysis of at least 4 different preparations of the VLDL fractions showed that the 4h remnants had a marked reduction of apoE4 (x0.4) but increased apoB48 (x6.7) and apoAIV (x4.7) compared with 4m remnants. ApoE3 in the 3h remnants was also reduced (x0.5) compared with 3m remnants, but the increase in apoB48 (x3) was less prominent. The 4h remnant fraction contains an average of 4.5 times more apoB proteins than that of the 3h mice and had a smaller apoE/apoB ratio (8±1% relative to that in 4m) compared with 36±4% in 3h fraction (P<0.005). Thus, the remnants of the 4h mice are cholesterol-rich, TG-poor, and apoE-poor compared with those of the 3h mice.
The enrichment of apoAIV in the 4h remnants suggests that they are mainly from intestine-derived chylomicrons. Consistent with this possibility, plasma cholesterol levels declined steadily in fasting 4h mice but not in 4m mice (Figure 2A) and remnant particles were reduced in 4h mice fasted for 18 hours or longer. There were no significant differences in the TC and TG content in the livers of these mice. The relative amounts of apoE protein estimated by Western blot analysis indicate that the elevated LDLR expression increases liver-associated apoE by 30% but reduces plasma apoE by 60% regardless of their Apoe genotype.
Figure 2. Effects of increased LDLR expression on mice with human apoE4 on HFW. A, Fasting effects on plasma cholesterol levels in male 4h (?, n=8) and 4m (, n=8) mice on HFW. Error bars are SEM. P<0.0001 by MANOVA. *P<0.001 and P<0.05 between genotypes by 2-tailed Student’s t test. B, Liver mRNA levels for the mouse Ldlr (black bars) and for the human Ldlr (white bars) expressed as percent of the Ldlr gene expression in 4m mice on NC. Numbers of animals (all males) are shown in the bottom of the bars. Error bars are SEM. P<0.05 for diet effect, P<0.0005 for Ldlr genotype effect by 2-way ANOVA. The effect of Apoe genotype is not significant.
The level of mouse Ldlr mRNA in the liver of 4h mice with a single copy of the gene was 61% of that in 4m mice with 2 copies (100%, Figure 2B). The human Ldlr message was 155% (total, 216%). The regulatory machinery of the Ldlrh allele is intact, because HFW downregulated the mouse and human Ldlr messages equally to 60% of their levels in mice fed NC. The total Ldlr mRNA level in 4h mice fed HFW was 2-fold higher than in 4m mice on HFW and a trace higher than in 4m mice fed NC. Thus, the marked hypercholesterolemia in the 4h mice fed HFW occurs despite their having high levels of LDLR expression. Ldlr expression is similar in the livers of 4h and 3h mice on HFW, indicating that the hypercholesterolemia in the 4h mice, but not the 3h mice, is not attributable to any differences in the diet-induced downregulation of Ldlr expression in the liver.
In Vivo Lipoprotein Metabolism in 4h Mice
To estimate the production of VLDL, we injected Triton WR1339 into 4h and 4m mice fed the HFW to inhibit lipolysis and the uptake of TG-rich particles. The initial rate of TG accumulation in plasma was not different between the 4h and 4m mice (Figure 3A). Plasma TC increased steadily and equally in both mice. They also responded similarly to the fat loading. The rates of plasma TG increase after fat loading in 4h mice treated with Triton WR1339 (604±33 mg/dL per h, n=4) was similar to those in 4m mice (656±37 mg/dL per h, n=4). Thus, increased LDLR expression seems to have no effect on the chylomicron production. These data suggest that the dramatic accumulation of remnants in the 4h mice compared with 4m mice is not attributable to oversecretion of TG-rich particles.
Figure 3. In vivo lipoprotein metabolism. A, Triglyceride (left) and cholesterol (right) secretion of 4m ( and 4h (?) mice after inhibition of lipolysis with Triton WR1339. The overall rate of TG secretion was not significantly different between the genotypes by MANOVA. B, Clearance of remnants. Radioactivity remaining in the plasma after injection of I125-labeled VLDL obtained from 4h remnants into HFW-fed 4h males (n=10) and 4m males (n=9) was not significantly different at any time points except at 10 minutes (left, P<0.05). Clearance of the 131I radiolabeled apoE4-rich 4KO remnants (right) in 4h males (n=5) was significantly faster than in 4m males (n=5, P<0.0001 by MANOVA, and P<0.01 at all times up to 120 minutes).
To examine the ability of mice to clear cholesterol-enriched remnant particles, we isolated remnants (d<1.02) from HFW-fed 4h mice, radiolabeled the particles with 125I, and injected them into the jugular vein of 4h and 4m mice fed the HFW diet. Despite the increased Ldlr expression in the 4h mice, removal of the apoE-poor remnants from their plasma (0.04 pools/min) was not different from that in the 4m mice (0.04 pools/min, Figure 3B, left). Similar results were obtained with remnants (d<1.006) isolated from apoE-deficient mice (not shown). To test whether the 4h mice having increased LDLR are able to clear apoE4-enriched remnants, we isolated lipoproteins (d<1.006) from mice expressing apoE4 but lacking LDLR (4KO) that are markedly enriched with apoE but not with apoAIV (C. Knouff and N. Maeda, unpublished data). The plasma decay of 131I-radiolabeled 4KO remnants (Figure 3B, right) was significantly faster in the 4h mice (0.07 pools/min) than in the 4m mice (0.03 pools/min, P<0.0001). Thus, the 4h mice with genetically increased LDLR expression clear remnant lipoproteins at an enhanced rate as long as these particles are enriched in apoE.
Taken together, these data indicate that the marked accumulation of remnant particles in the 4h mice is neither because they have increased secretion of nascent TG-rich lipoproteins nor because they have reduced clearance of apoE-poor remnants compared with 4m mice. The inference from these data is that the conversion of large TG-rich particles to smaller cholesterol-rich remnants is enhanced in the 4h mice.
Severe Atherosclerosis in 4h But Not 3h Mice on HFW Diet
Accumulation of remnant particles and reduction of HDL-cholesterol in HFW-fed 4h mice is a high-risk profile for atherosclerosis, even though the average plasma total cholesterol of these mice is only marginally elevated (200 mg/dL). We therefore evaluated the development of atherosclerosis in mice fed HFW containing 21% fat and 0.5% cholesterol for 3 months. No plaques were found in any of the 3m mice (5 females) or 4m mice (8 females and 5 males) on HFW. Similarly, none of the 7 female 3h mice on HFW developed plaques (Figure 4A). In contrast, all of the 15 female and 7 male 4h mice on HFW developed significant plaques at the aortic sinus area (Figures 4B through 4D) with average plaque sizes of 59±15x103 μm2 in females and 22±7x103 μm2 in males. Although the numbers and sizes of plaques varied in individual animals, most of the mice had mature complex plaques with fibrous caps, necrotic lipid cores, cholesterol clefts, and calcifications (Figures 4C and 4D). Thus, mice having human APOE*4 and a moderately increased amount of LDLR develop significant atherosclerosis when fed a diet similar in composition to that of Western societies.
Figure 4. Representative atherosclerotic plaques in the aortic sinus of mice fed HFW for 3 months. A, 3h female; B, 4h female; C, 4h male; and D, 4h female. Frozen sections were stained with Sudan IVB for lipids and counterstained with hematoxylin. Black arrowheads indicate fibrous caps. Yellow and blue arrows designate calcification and cholesterol clefts, respectively. L indicates lumen; M, media; and A, adventitia. Scale bars=500 μm in a and b and 100 μm in c and d.
Discussion
Our clear demonstration of hypercholesterolemia and atherosclerosis in mice that have human apoE4, but not apoE3, replicates in mice fed a Western diet the paradoxical association between APOE polymorphisms and the risk of atherosclerosis in humans. However, contrary to the currently accepted hypothesis that downregulation of LDLR in individuals with an APOE*4 allele is the cause of their high plasma cholesterol levels,2,7,8 we find that these isoform-specific effects are only present when LDLR is increased.
How can increased LDLR expression ever be harmful? We suggest that this is because the LDLR, under some circumstances, traps sufficient apoE and the supply becomes inadequate to process a high dietary intake of lipids. Exchange of apoE onto nascent triglyceride-rich lipoproteins is a necessary prerequisite for their internalization via LDLR.13–17 The overall process that we envision is illustrated in Figure 5. ApoE4 may be particularly susceptible to this trapping because of its strong affinity for the LDLR. Transient particles that fail to acquire apoE4 are excellent substrates for lipases,1 and the resulting enhanced lipolysis will increase cholesterol-rich apoE-poor remnant particles in plasma. We suggest that apoE3 is less susceptible to trapping than apoE4, perhaps because of its somewhat lower affinity for the LDLR.5,6 Consequently, the 3h mice have sufficient apoE3 to process nascent lipoproteins for rapid internalization. This differential transfer of apoE3 and apoE4 to lipoproteins can also explain our previous observation that in vivo clearance of exogenously introduced remnants is significantly faster in 3m mice than in 4m mice.6
Figure 5. ApoE trapping by the LDLR. Triglyceride-rich chylomicrons (yellow) secreted by the intestine into lymph are remodeled to transient particles (orange) mainly in capillaries and in the space of Disse in the liver. This initial capturing phase is very rapid and does not depend on the LDLR; it probably involves heparan sulfate proteoglycans that facilitate lipolysis and apolipoprotein exchange. Enrichment with apoE is a requirement for the fast internalization of the transient particles primarily via LDLR-mediated endocytosis. Otherwise, the particles are additionally processed to cholesterol-enriched remnants (red) that are slowly cleared from the plasma. The lipid composition of the remnants in plasma depends on both Apoe genotype and the amount of LDLR. The processing of liver-derived VLDL particles is likely to be similar to that illustrated for chylomicrons.
Previously we reported that mice expressing solely apoE2 (2m) show features typical of type III hyperlipoproteinemia in humans18 but that increased LDLR expression in these mice (2h) completely ameliorates their hyperlipoproteinemia.11 According to our hypothesis, apoE2 with its very low affinity for the LDLR should be virtually free from trapping and should therefore be efficiently transferred to transient TG-rich particles. Although such an apoE2 enrichment of TG-rich particles is likely to increase their LDLR-mediated internalization, it will also severely inhibit lipolysis1 and could account for the prominent accumulation of TG-rich remnants seen in the circulation of the 2m mice.18 In the 2h mice, however, high LDLR expression tips the balance toward more internalization and lowers their plasma cholesterol.11
Clearly, additional studies are necessary to refine and test our proposed apoE trapping by the LDLR. For example, we do not know whether it occurs on the cell surface, as illustrated in Figure 5, or during intracellular trafficking.19,20 The word trapping should not be taken too literally; difference in the interaction between apoE and LDLR or their subsequent processing may result not only from differences in binding affinities but also from other properties influenced by the specific amino acids that differ among 3 isoforms. Interactions of apoE with other molecules, such as proteoglycans and hepatic lipase, that are also expressed on the basolateral microvilli of hepatocytes may also influence the apoE interaction with LDLR in an isoform-specific fashion. Published studies have shown that newly synthesized apoE is incompletely secreted and partially degraded in HepG2 cells in culture21 and a significant portion of apoE synthesized by macrophages undergoes rapid cellular degradation in a non-lysosomal compartment in a sterol-regulated manner.22 Although the role of LDLR in these processes has not been addressed, LDLR is known to bind to newly synthesized apoE in macrophages and limits its secretion.23 LDLR expression in macrophages could also contribute to atherogenesis in an apoE isoform–specific fashion. For example, a differential effect on cholesterol homeostasis in macrophages by apoE isoforms with apoE4 being least effective in promoting cholesterol efflux from macrophage has been reported.24 Additionally, Linton et al25 have shown that C57BL/6 mice receiving Ldlr-/- marrow developed 63% smaller lesions than mice receiving Ldlr+/+ marrow after dietary atherogenic stimuli, demonstrating that macrophage LDLR affects the rate of foam cell formation under conditions of dietary stress.
Some comments are required on the relevance of our findings in mice to the effects of different APOE genotypes in humans. We note that 4h mice preferentially accumulate apoB48-containing lipoproteins of an intermediate density, whereas humans with the APOE*4 allele mainly have elevated levels of apoB100-containing LDL.2 This is not incompatible with our hypothesis, which predicts that trapping of apoE4 by the LDLR will hinder enrichment of VLDL remnants with apoE4, thereby leading to an increase in their conversion to LDL. Because the clearance of LDL particles mediated by binding apoB100 to the receptor is much slower than apoE-mediated VLDL clearance,13,15 we expect that the plasma cholesterol levels in individuals with apoE4 will be increased. Supporting this explanation, an increased conversion of VLDL to smaller remnants and a relative decrease in direct removal of VLDL in APOE*4 homozygotes compared with APOE*3 subjects have been demonstrated.26
Our finding that hypercholesterolemia is seen only when the 4h mice are fed HFW diet is also consistent with observations that human subjects carrying the APOE*4 allele are more responsive than others to LDL cholesterol–lowering by diet.27,28 In addition, some though not all studies have found prolonged postprandial lipemia in normolipidemic subjects that carry APOE*4.29–32 Bergeron and Havel29 have proposed that prolonged residence times of chylomicron and VLDL remnants in persons with APOE*4 raise the concentration of LDL by increasing the amount of VLDL converted to LDL. We additionally note that an apoE5 variant with lysine in place of glutamic acid at position 3 is also associated with hyperlipidemia and atherosclerosis and it has a twice-normal LDLR binding activity.33 Finally, some although not all studies have shown that the cholesterol-lowering effects of statins, thought to be primarily mediated by increased LDLR, are apoE isoform–dependent. In these studies, individuals with the APOE*3/4 and APOE*4/4 genotypes had significantly smaller LDL cholesterol reductions in response to statin treatment than those with the APOE*3/3 genotype.34,35 Clinical studies clearly indicate that statin therapies reduce the risk of cardiovascular disease in humans, including those with apoE4, and no serious adverse effects on plasma lipids have been reported.35,36 Nevertheless, our observations suggest the need for additional studies of the interaction between the cholesterol-lowering effect of statins and genetic variations.
In conclusion, our studies demonstrate that, contrary to the presently accepted downregulation of LDLR hypothesis, increased LDLR has harmful effects in Western diet–fed mice expressing human apoE4 and causes marked accumulation of apoE-poor lipoprotein remnants in plasma and severe atherosclerosis. The alternative mechanism of apoE-trapping by LDLR explains our observations and offers a plausible explanation why apoE4, which has a greater affinity for the LDLR than apoE3, is associated with higher plasma cholesterol and a greater risk of atherosclerosis in humans. Our unexpected findings in mice predict important interactions between APOE genotype, LDLR expression, and diet.
Acknowledgments
This work was supported by grants from NIH (HL42360 to N.M., HL54176 and HL49373 to J.S.P, and HL07115 to L.L.F). The authors thank Dr Tom Smith for the jugular vein injection during the lipoprotein clearance experiments, Drs Greg Shelness, Leighton James, and Oliver Smithies for discussion, and Jennifer Altenburg and Ellen Young for technical assistance.
References
Mahley RW, Rall SC Jr. Type III hyperlipoproteinemia (dysbetalipoproteinemia): the role of apolipoprotein E in normal and abnormal lipoprotein metabolism. In: Scriver AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inheritable Disease. New York: MacGraw Hill; 2001: 2835–2913.
Davignon J, Gregg RE, Sing CF. Apolipoprotein E polymorphism and atherosclerosis. Arteriosclerosis. 1988; 8: 1–21.
Boerwinkle E, Utermann G. Simultaneous effects of the apolipoprotein E polymorphism on apolipoprotein E, apolipoprotein B, and cholesterol metabolism. Am J Hum Genet. 1988; 42: 104–112.
Weisgraber KH. Apolipoprotein E: structure-function relationships. Adv Protein Chem. 1994; 45: 249–302.
Bohnet K, Pillot T, Visvikis S, Sabolovic N, Siest G. Apolipoprotein (apo) E genotype and apoE concentration determine binding of normal very low density lipoproteins to HepG2 cell surface receptors. J Lipid Res. 1996; 37: 1316–1324.
Knouff C, Hinsdale ME, Mezdour H, Altenburg M, Watanabe M., Quarfordt SH, Sullivan PM, Maeda N. Apo E structure determines VLDL clearance and atherosclerosis risk in mice. J Clin Invest. 1999; 103: 1579–1586.
Gregg RE, Zech LA, Schaefer EJ, Stark D, Wilson D, Brewer Jr. HB. Abnormal in vivo metabolism of apolipoprotein E4 in humans. J. Clin Invest. 1986; 78: 815–821.
Weintraub MS, Eisenberg S, Breslow JL. Dietary fat clearance in normal subjects is regulated by genetic variation in apolipoprotein E. J Clin Invest. 1987; 80: 1571–1577.
Brown MS, Goldstein JL. A receptor-mediated pathway for cholesterol homeostasis. Science. 1986; 232: 34–47.
Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia. In: Scriver AL, Sly WS, Valle D, eds. The Metabolic and Molecular Bases of Inheritable Disease. New York: MacGraw Hill; 2001: 2863–2913.
Knouff C, Malloy S, Wilder J, Altenburg MK, Maeda N. Doubling expression of the low density lipoprotein receptor by truncation of the 3'-untranslated region sequence ameliorates type III hyperlipoproteinemia in mice expressing the human apoE2 isoform. J Biol Chem. 2001; 276: 3856–3862.
Sullivan PM, Mezdour H, Aratani Y, Knouff C, Najib J, Reddick RL, Quarfordt SH, Maeda N. Targeted replacement of the mouse apolipoprotein E gene with the common human APOE*3 allele enhances diet-induced hypercholesterolemia and atherosclerosis. J Biol Chem. 1997; 272: 17972–17980.
Cooper AD. Hepatic uptake of chylomicron remnants. J Lipid Res. 1997; 38: 2173–2192.
Herz J, Qiu SQ, Oesterle A, DeSilva HV, Shafi S, Havel RJ. Initial hepatic removal of chylomicron remnants is unaffected but endocytosis is delayed in mice lacking the low density lipoprotein receptor. Proc Natl Acad Sci U S A. 1995; 92: 4611–4615.
Mortimer BC, Beveridge DJ, Martins IJ, Redgrave TG. Intracellular localization and metabolism of chylomicron remnants in the livers of low density lipoprotein receptor-deficient mice and apoE-deficient mice: evidence for slow metabolism via an alternative apoE-dependent pathway. J Biol Chem. 1995; 270: 28767–28776.
Mahley RW, Ji ZS. Remnant lipoprotein metabolism: key pathways involving cell-surface heparan sulfate proteoglycans and apolipoprotein E. J Lipid Res. 1999; 40: 1–16.
Bradley WA, Gianturco SH. ApoE is necessary and sufficient for the binding of large triglyceride-rich lipoproteins to the LDLR: apoB is unnecessary. J Lipid Res. 1986; 27: 40–48.
Sullivan PM, Mezdour H, Quarfordt SH, Maeda N. Type III hyperlipoproteinemia and spontaneous atherosclerosis in mice resulting from gene replacement of mouse Apoe with human APOE*2.5 J Clin Invest. 1998; 102: 130–135.
Brown MS, Herz J, Goldstein JL. LDL-receptor structure: calcium cages, acid baths and recycling receptors. Nature. 1997; 388: 629–630.
Swift LL, Farkas MH, Major AS, Valyi-Nagy K, Linton MF, Fazio S. A recycling pathway for resecretion of internalized apolipoprotein E in liver cells. J Biol Chem. 2001; 276: 22965–22970.
Schmitt M, Grand-Perret T. Regulated turnover of a cell surface-associated pool of newly synthesized apolipoprotein E in HepG2 cells. J Lipid Res. 1999; 40: 39–49.
Duan H, Lin CY, Mazzone T. Degradation of macrophage ApoE in a nonlysosomal compartment: regulation by sterols. J Biol Chem. 1997; 272: 31156–1162.
Zhao Y, Mazzone T. LDL receptor binds newly synthesized apoE in macrophages: a precursor pool for apoE secretion. J Lipid Res. 1999; 40: 1029–1035.
Cullen P, Cignarella A, Brennhausen B, Mohr S, Assmann G, von Eckardstein A. Phenotype-dependent differences in apolipoprotein E metabolism and in cholesterol homeostasis in human monocyte-derived macrophages. J Clin Invest. 1998; 101: 1670–1677.
Linton MF, Babaev VR, Gleaves LA, Fazio S. A direct role for the macrophage low density lipoprotein receptor in atherosclerotic lesion formation. J Biol Chem. 1999; 274: 19204–19210.
Demant T, Bedford D, Packard CJ, Shepherd J. Influence of apolipoprotein E polymorphism on apolipoprotein B-100 metabolism in normolipemic subjects. J Clin Invest. 1991; 88: 1490–1501.
Lopez-Miranda J, Ordovas JM, Mata P, Lichtenstein AH, Clevidence B, Judd JT, Schaefer EJ. Effect of apolipoprotein E phenotype on diet-induced lowering of plasma low density lipoprotein cholesterol. J Lipid Res. 1994; 35: 1965–1975.
Dreon DM, Fernstrom HA, Miller B, Krauss RM. Apolipoprotein E isoform phenotype and LDL subclass response to a reduced-fat diet. Arterioscler Thromb Vasc Biol. 1995; 15: 105–111.
Bergeron N, Havel RJ. Prolonged postprandial responses of lipids and apolipoproteins in triglyceride-rich lipoproteins of individuals expressing an apolipoprotein epsilon 4 allele. J Clin Invest. 1996; 97: 65–72.
Superko HR, Haskell WL. The effect of apolipoprotein E isoform difference on postprandial lipoprotein in patients matched for triglycerides, LDL-cholesterol, and HDL-cholesterol. Artery. 1991; 18: 315–325.
Boerwinkle E, Brown S, Sharrett AR, Heiss G, Patsch W. Apolipoprotein E polymorphism influences postprandial retinyl palmitate but not triglyceride concentrations. Am J Hum Genet. 1994; 54: 341–360.
Dallongeville J, Tiret L, Visvikis S, O’Reilly DS, Saava M, Tsitouris G, Rosseneu M, DeBacker G, Humphries SE, Beisiegel U. Effect of apo E phenotype on plasma postprandial triglyceride levels in young male adults with and without a familial history of myocardial infarction: the EARS II study.European Atherosclerosis Research Study. Atherosclerosis. 1999; 145: 381–388.
Dong LM, Yamamura T, Yamamoto A. Enhanced binding activity of an apolipoprotein E mutant, APOE5, to LDLRs on human fibroblasts. Biochem Biophys Res Commun. 1990; 168: 409–414.
Ordovas JM, Lopez-Miranda J, Perez-Jimenez F, Rodriguez C, Park JS, Cole T, Schaefer EJ. Effect of apolipoprotein E and A-IV phenotypes on the low density lipoprotein response to HMG CoA reductase inhibitor therapy. Atherosclerosis. 1995; 113: 157–166.
Ballantyne CM, Herd JA, Stein EA, Ferlic LL, Dunn JK, Gotto AM Jr, Marian AJ. Apolipoprotein E genotypes and response of plasma lipids and progression-regression of coronary atherosclerosis to lipid-lowering drug therapy. J Am Coll Cardiol. 2000; 36: 1572–1578.
Gerdes LU, Gerdes 2C, Kervinen K, Savolainen M, Klausen IC, Hansen PS, Kesaniemi YA, Faergeman O. The apolipoprotein epsilon4 allele determines prognosis and the effect on prognosis of simvastatin in survivors of myocardial infarction: a substudy of the Scandinavian simvastatin survival study. Circulation. 2000; 101: 1366–1371.(Sudi I. Malloy*; Michael )