Aortic Constriction Exacerbates Atherosclerosis and Induces Cardiac Dysfunction in Mice Lacking Apolipoprotein E
Despite considerable evidence suggesting that hypertension contributes to the development and progression of atherosclerosis, the causative links remain unclear. We have tested the effects of chronic hypertension induced by suprarenal aortic constriction on the development of atherosclerosis in apolipoprotein E–deficient (Apoe−/−) mice. Compared with a sham operation, narrowing the aortic luminal diameter by 33% increased blood pressure proximal to the constriction by ≈15 mm Hg, but the pressures distal to the constriction were unchanged. Kidney renin mRNA and plasma renin activity were also unaffected. Compared with plaque size after the sham operation, atherosclerotic plaque size in the aortic root 8 weeks after coarctation was increased to 245% and 152% in males and females, respectively. Aortic segments at the constriction were free of atherosclerotic deposits, but segments proximal to the constriction were dilated and had atherosclerotic lesions. Thrombi were present immediately below the constriction in Apoe−/− and wild-type vessels. Surprisingly, compared with wild-type mice, the Apoe−/− mice were more susceptible to the cardiac hypertrophy and dysfunction induced by pressure overload. Thus, aortic coarctation exacerbates atherosclerosis in vessels proximal to the constriction without a concomitant increase in the renin-angiotensin system. Our study also suggests that apolipoprotein E plays an important role in modulating cardiac hypertrophy.
Hypertension and atherosclerosis are 2 of the most important causes of morbidity and mortality in humans. Numerous clinical and epidemiological studies have identified systemic arterial hypertension as an independent and potent risk factor for the development of atherosclerotic disease. For example, the Framingham study found that diastolic blood pressure (BP), systolic BP, and pulse pressure are predictors of coronary heart disease risk.1,2⇓ The extent of atherosclerosis in the aorta and coronary arteries of subjects in the Pathobiological Determinants of Atherosclerosis in Youth (PDAY) study3 was greater in hypertensive compared with normotensive individuals. Similarly, an association between ultrasound-measured carotid artery intima/media thickness and systolic BP was demonstrated in the Atherosclerosis Risk in Communities (ARIC) Study.4
Animal models with combined genetic risks for atherosclerosis and hypertension have also shown that elevated BP accelerates atherogenesis. Thus, lack of endothelial NO synthase increases BP and exacerbates atherosclerosis in mice lacking apoE (Apoe−/− mice).5 Similarly, Dahl salt-sensitive hypertensive rats that overexpress the human cholesteryl ester transfer protein develop severe hyperlipidemia and atherosclerosis and show decreased survival.6 Nevertheless, current data do not allow an unequivocal distinction between the physical effects on the atherogenic process of the increased systemic BP and the effects on vasculature caused by concurrent changes in circulating vasoactive mediators, and causative links between hypertension and atherosclerosis have yet to be defined.
Surgical coarctation of the aorta in combination with high dietary cholesterol has been shown to markedly increase atherosclerosis in rabbits and monkeys.7–9⇓⇓ In the present study, we have applied aortic coarctation to alter the BPs of Apoe−/− mice with the aim of understanding the interaction between BP changes and a genetic form of hyperlipidemia. We find that aortic coarctation in mice accelerates atherosclerosis in the aorta proximal to the constriction. Organized thrombi are found immediately distal to the constriction independent of hyperlipidemia. A surprising new finding is that compared with wild-type mice, Apoe−/− mice are significantly more susceptible to the cardiac hypertrophy induced by chronic coarctation of the aorta.
Apoe−/− mice10 were backcrossed at least 7 generations to a C57BL/6 genetic background. C57BL/6 mice were used as the wild-type control group. Care and experimental procedures were in compliance with the Principles of Laboratory and Animal Care established by the National Society for Medical Research and were approved by the institutional committee.
Mice were anesthetized with isoflurane and placed in the supine position on warm pads under an operating microscope. The abdomen was opened, and the abdominal aorta at the suprarenal level was freed of the surrounding adventitial adipose tissue by gentle dissection. The aorta between the celiac and superior mesenteric arteries was constricted by tying a 6-0 silk suture ligature against a 28-gauge needle for males and 30-gauge needle for females to yield an ≈33% narrowing of the luminal diameter when the needle was removed. For the sham operations, 18-gauge needles and 20-gauge needles were used for males and females, respectively. Coarctation was applied when the mice were 2 months of age, and they were euthanized at 4 months of age for evaluation.
Echocardiograms were obtained on conscious mice with gentle restraint at 2, 4, 6, and 8 weeks after surgery. Two-dimensional guided M-mode echocardiography was performed by using HDI 5000 echocardiograph equipment (ATL) and a 7.5-MHz transducer.11 Heart rate was calculated by using the period between 2 consecutive end-diastolic points from pulsed Doppler echocardiography. Left ventricular (LV) mass was calculated as described.12
Intra-arterial BP Measurement
Mice were anesthetized with isoflurane, and the right carotid and right femoral artery were cannulated with flame-stretched PE 20 tubing. Catheters were connected to MLT1050 precision transducers and Power-Laboratory recording equipment (ADI Instruments). Approximately 90 minutes after the mice had recovered from the anesthesia, intracarotid and femoral artery BPs were measured simultaneously. An average of 10 measurements was taken as a mean value for each animal.
Histology and Lesion Measurements
Mice were euthanized at 4 months of age. Blood was collected from the heart into a tube containing EDTA (final concentration 5 mmol/L) for plasma lipid analysis.13 The heart and vascular tree were perfused by intracardiac infusion with 10 mL of 4% paraformaldehyde in PBS under physiological pressure. Heart segments that contain the aortic sinus were serially sectioned for morphological evaluation of atherosclerotic lesions.13 Thoracic aortas were excised, freed of adventitial fat, and examined under microscopy. Specimens were subsequently dried and weighed, and lipids were extracted by methanol/chloroform (1:1 [vol/vol]). Serial 10-μm-thick sections of the contiguous abdominal aorta were cut every 100-μm distance from 5 mm above the suture level to 5 mm below for evaluation. Paraffin-embedded tissue sections were stained with hematoxylin and eosin (H&E), with periodic acid-Schiff stain (PAS), or with trichrome. Aortic segments were also embedded in epoxy resin, and sections were stained with toluidine blue.
Renin mRNA and Plasma Activity
Total RNA was prepared from homogenized kidney by using ABI 6700 (ABI). Renin mRNA levels were determined by real-time quantitative reverse transcription–polymerase chain reaction with ABI 7700 (ABI) with the use of 5′-ACAGTATCCCAACAGGAGAGACAAG-3′ and 5′-GCACCCAGGACCCAGACA-3′ as primers and 5′-FAM-TGGCTCTC-CATGCCATGGACATCC-Tamra-3‘ as the detection probe. Amplification of β-actin mRNA was used as an internal standard with 5′-CTGCCTGACGGCCAAGTC-3′ and 5′-CAAGAAGGAAGGCTGG-AAAAGA-3′ as primers and 5′-TET-CACTATTGGCAACGAGCGGT-TCCG-Tamra-3′ as the detection probe. Plasma concentration of active renin was determined by radioimmunoassay as described.14
Data were analyzed by using a JMP software package (SAS Institute Inc). Means of different groups were compared by ANOVA. Time-dependent changes were analyzed by MANOVA with repeated measures.
Aortic Coarctation and Survival of Animals
Surgical coarctation (banding) of the abdominal aorta of 2-month-old Apoe−/− or wild-type mice yielded an approximately one third narrowing of the luminal diameter (see below). Aortas of the sham-operated mice had no luminal narrowing. All mice recovered from anesthesia, but ≈20% of sham-operated mice and 40% of the banded mice died within a week after surgery (most died within 1 or 2 days). Postmortem examination of some banded mice that died after 24 hours showed thrombi at or near the banded area. No additional losses of mice were seen 1 week after banding, except in the banded Apoe−/− group. Thus, 34 of 86 banded Apoe−/− mice died during the first week, and 9 died between 6 and 8 weeks after surgery. Some of the later deaths were by aortic dissection, and some mice showed signs of heart failure. All the mice included in the present study that were alive at 8 weeks after surgery appeared healthy. There was no significant difference in weight gain between the sham-operated and banded groups (Table 1). Compared with sham-operated hearts, the hearts of male and female Apoe−/− mice were significantly enlarged after banding and showed an ≈50% increase in the heart-to-body weight (HW/BW) ratio (P<0.0001). In contrast, the increase of the HW/BW ratio was 20% in the banded wild-type mice compared with the sham-operated mice (P<0.01, in genotype effects). Total plasma cholesterol and triglyceride levels were not affected by banding.
Increased BP Proximal to the Coarctation
Aortic banding significantly increased BP proximal to the constriction (Table 2). The differences between the carotid and femoral artery BPs of individual animals averaged 25±2 mm Hg (n=52) in the banded group versus 10±1 mm Hg (n=32) in the sham-operated group (P<0.0001). The increased pressure differentials resulted mainly from an increased carotid BP in the banded mice (126±3 mm Hg in banded mice compared with 111±2 mm Hg in sham-operated mice, P<0.001). The difference in the femoral artery BP between banded and sham-operated mice was not significant (100±2 versus 101±2 mm Hg, P=0.3). There were no statistically significant differences in BPs (carotid or femoral) between the wild-type and Apoe−/− mice or between the sexes. The HW/BW ratio of individual Apoe−/− animals was positively correlated with the differential BPs between their carotid and femoral arteries (R2=0.31, P=0.0001; n=68).
Short-term aortic constriction above the renal arteries can increase renin production in the kidneys, which can sense the reduced BP distal to the constriction.15 However, we found no difference in renin mRNA levels in the kidneys or in plasma renin activity of banded and sham-operated mice at 8 weeks after banding (Table 2), which is consistent with the absence of significant differences between the femoral arterial BPs of the banded and the sham-operated animals.
Vessels in the Vicinity of Constrictions
Cross sections of the abdominal aorta of the banded area were examined by light microscopy (Figure 1). The position of the banding applied to the aorta was identified by the presence of the suture (Figure 1D and 1E). The banding narrowed the aortic luminal area in the Apoe−/− and wild-type mice to 39±4% (n=11) relative to that in the sham-operated mice (100±5%, n=5). The intima contained 2 or 3 layers of small cuboidal cells, most likely representing proliferation of smooth muscle cells to accommodate the increased pressure (Figure 1E and 1F). Lipids were absent in the intima.
The aorta proximal to the banding site was dilated with a luminal diameter ≈30% larger than the diameter of the sham-operated aorta (Figure 1A). In 10 of the 12 banded Apoe−/− mice examined, the aortas immediately above the constriction contained atherosclerotic lesions of various sizes, ranging from foam cell accumulations to well-developed plaques (Figure 1B). In 4 mice, the medial layer was disrupted, and plaque material was deposited outside the vessel wall (Figure 1C). Plaques were not present further upstream (>500 μm) from the banding site. The presence of abdominal aortic lesions is rare in Apoe−/− mice at 4 months of age, and none of the 6 sham-operated Apoe−/− mice that were examined had plaques in this area.
Vessels immediately distal to the site of banding were slightly dilated, with the luminal diameter increased by ≈10% relative to the sham-operated vessels. In all 6 Apoe−/− vessels and in the 3 wild-type vessels evaluated, there were large thrombi attached at multiple locations to the aortic wall (Figure 1G). Smooth muscle cells within the medial layers near the thrombi were elongated and contained lipids, but foamy macrophages were not present (Figure 1H). Most of the thrombi were well organized and composed of fibrous materials, but occasionally, small aggregates of platelets were present on the surface (Figure 1I), suggesting that thrombus formation at this location was an ongoing process. However, even in the hyperlipidemic environment of Apoe−/−, lipids were not identified as a component of these thrombi. Thrombi in the same area were also found in the vessels of 3 of the 6 Apoe−/− mice 4 weeks after the banding but in none of the vessels of 9 mice 2 weeks after banding (not shown). None of the sham-operated mice developed thrombi near the site of banding.
Cardiac Hypertrophy and Other Pathological Complications.
We used echocardiography to monitor changes in the heart induced by aortic banding. Abdominal aortic banding of Apoe−/− mice induced a time-dependent increase in LV end-diastolic dimension (LVEDD; P<0.01 by MANOVA versus sham, Figure 2A) and LV mass (P<0.01, Figure 2B) and a trend toward a decrease in ejection fraction (P=0.066, Figure 2C). Hearts were hypertrophied by 4 weeks after banding and began to dilate significantly after 6 weeks. At 8 weeks, 5 of 41 banded Apoe−/− mice had an ejection fraction of <30%, indicating heart failure. In contrast, neither the LV diameter nor the heart mass was increased significantly in wild-type mice after banding. Thus, the Apoe−/− mice are more prone to develop cardiomegaly and heart failure in response to chronic cardiac overload induced by aortic constriction than are wild-type mice. Heart rate was not significantly influenced by sex, treatment, or genotype (Table 2).
Histological examination showed that the LVs and right ventricles were dilated in the 8-week banded Apoe−/− mice. Myocardial hypertrophy and cardiac damage, as evidenced by larger myocytes with enlarged nuclei and extensive interstitial and perivascular fibrosis, were present in the hearts of the banded Apoe−/− mice (Figure 3A and 3B). In addition, in 8 of 37 Apoe−/− mice, the left atrium was enlarged and firm and contained large organized thrombi (Figure 3C). Foam cells were found on the left atrial wall adjacent to the thrombus and, presumably, were the consequence of a disturbed blood flow in this area (Figure 3D). Most of the mice with atrial thrombi also showed evidence of pulmonary congestion (Figure 3E). The lungs contained intra-alveolar collections of enlarged macrophages that were filled with acidophilic needle-shaped crystals (Figure 3F) that are typical of acidophilic macrophage pneumonia. Acute inflammatory cells were absent. Bronchi and alveolar spaces contained PAS staining of the proteinaceous materials (Figure 3E). Areas of fibrosis were present in some animals. Neither the sham-operated Apoe−/− mice nor the banded wild-type mice had atrial thrombi or pulmonary congestion, implicating the detrimental effects of the Apoe mutation under these circumstances.
Increased Aortic Atherosclerosis Proximal to the Constriction
Lesions in the aortic arch were significantly increased in banded Apoe−/− mice compared with sham-operated mice (Figure 4A through 4D). All the thoracic aortas from banded mice (6 males and 5 females) contained at least 3 large plaques in the aortic arch, whereas no aortas from the sham-operated mice (6 males and 5 females) had >1 plaque. The cholesterol ester content of the excised thoracic aortas of banded Apoe−/− mice was 244±38% that of the sham-operated vessels per aorta or 166±16% per dry weight (n=11 and 9, respectively; P<0.001; effects of the sex of the animal were not significant).
The mean size of lesions within the aortic root of banded males (86 000±8000 μm2) and females (128 000±9000 μm2) were 245% and 152% those of sham-operated males (35 000±7000 μm2) and females (84 000±8000 μm2), respectively (Figure 4E, P<0.0001 for effects of banding). No qualitative changes in the plaque morphology were notable in association with the banding (not shown). The lesion size was not correlated with either HW/BW ratio or ejection fraction of the heart of the individual animals, suggesting that cardiac dysfunction and atherosclerosis are independent processes in this system.
We have used mechanical methods to increase the aortic BP in mice to study the relationships between hypertension, atherosclerosis, and Apoe genotype. Our data show that BP and atherosclerosis in the aorta proximal to the constriction were increased by banding the suprarenal abdominal aorta of Apoe−/− mice, whereas no changes occurred in the BP and atherosclerosis in the distal aorta. We also found that Apoe−/− mice are significantly more susceptible than are wild-type mice to the cardiac hypertrophy induced by chronic constriction of the aorta.
The effects of chronic aortic coarctation on atherosclerosis in Apoe−/− mice confirm the effects observed in Watanabe heritable hyperlipidemic rabbits,7 in rabbits fed high cholesterol diets,8 and in cynomolgus monkeys.9 All these experiments demonstrate a marked increase in atherosclerotic lesions in response to elevated BP. The suggested explanations include the possibility that an increased intimal permeability induced by hypertension might lead to an increased rate of entry of lipoproteins into the intima16,17⇓ or that the higher pressure could increase endothelial or medial damage, thereby triggering the initiation of plaque formation by recruiting proinflammatory leukocytes to the sites.18
Angiotensin II is a potent vasoactive mediator that has profound effects on atherosclerosis. Thus, mice that overexpress a human renin transgene and a human angiotensinogen transgene are hypertensive and develop larger lesions than do nontransgenic mice when these mice are fed a high cholesterol diet.19 Daugherty et al20 and Weiss et al21 have shown that systemic infusion of angiotensin II at supraphysiological levels of ≈0.7 mg/kg per day dramatically increased atherosclerosis in Apoe−/− mice in the relatively short time of 4 weeks. Although these 2 studies differ (in that Daugherty et al did not see any BP increase in their animals, whereas Weiss et al reported that the treatment increased BP by ≈40 mm Hg), a marked infiltration of inflammatory cells in plaques and in the adventitia was noted by both studies, which agree that this is indicative of the proatherogenic effects of angiotensin II.
One of the hypotheses explaining the mechanism by which banding of the abdominal aorta above the renal arteries induces hypertension postulates that the constriction causes a reduction in renal blood flow, which in turn activates the renin-angiotensin system.15 In support of this idea, studies have shown that circulating renin levels22 and renin mRNA in the renal afferent arterioles23 increase 3- to 4-fold after abdominal aortic coarctation. However, other studies have shown that circulating renin levels return to normal within several days after the surgery22 and that chronic coarctation does not alter mean BPs distal to the coarctation.24 Our data agree with and extend the latter observations by showing that renin mRNA levels in the kidney, plasma renin activity, and the femoral BP in Apoe−/− mice 8 weeks after banding are not different from those in sham-operated mice. Thus, it is very unlikely that systemic angiotensin II levels are chronically altered in the banded animals compared with the sham-operated animals. Nevertheless, coarctation induced a sustained increase in BP and enhanced lesion development.
Although the increased atherosclerosis present in the proximal aorta and its branches is similar to that described in other animal models, the presence of organized thrombi in the portion of the aorta just distal to the coarctation has not, to our knowledge, been described. The thrombi were present at similar frequencies in banded wild-type and Apoe−/− mice, suggesting that disturbed blood flow due to the aortic constriction, not hyperlipidemia, is the cause of the thrombus formation. Thromboembolism induced by the injection of agents such as lipopolysaccharide, collagen, and thrombin or stasis induced by carotid artery ligation has been used as a model of thrombosis in mice.25 Aortic coarctation offers another reproducible system in which to study the process of chronic thrombus formation in vessels in which blood flow is disturbed. Whether the chronic and ongoing process of thrombus formation distal to the coarctation is relevant to the development of clinical complications of atherosclerosis in humans remains to be determined.
The second location at which we found organized thrombi in Apoe−/− mice but not in wild-type mice was the left atrium. The mice with atrial thrombi were also likely to have pulmonary acidophilic pneumonia. These complications indicate severe cardiac dysfunction and have been reported in rodents with hypertrophic cardiomyopathy26,27⇓ or acute viral myocarditis.28,29⇓ There were occasional small atherosclerotic plaques in the small- to medium-sized vessels in some of the banded Apoe−/− mice. However, cardiac changes in these mice are unlikely to be attributable to changes in their intracardiac vasculature, because the degree of atherosclerosis in the smaller vessels of the banded mice was not different from that of the sham-operated mice in the histological sections that we analyzed. Thus, perivascular fibrosis in the banded mice appears to be independent of atherosclerosis in the intracardiac vessels.
Cardiac dysfunction was not observed in any wild-type mice despite the fact that some of them had BP increases that were more than those seen in Apoe−/− mice with cardiac dysfunction. This suggests that a lack of apoE may exacerbate cardiac hypertrophy and subsequent dysfunction. Although little is known about the role of apoE in heart function, our earlier work30 has shown that apoE is highly expressed in the mouse heart, whereas its expression is low in skeletal muscles. In addition, Hartley et al31 have reported significantly elevated HW/BW ratios of the 13-month-old Apoe−/− mice compared with the wild-type mice.31 Many circulatory problems, including severe aortic atherosclerosis, can contribute to the hemodynamic changes in old Apoe−/− mice. However, our present study showed that the hearts of 2-month-old Apoe−/− mice, although normal, have a significantly larger LVEDD and a smaller ejection fraction than are seen in wild-type C57BL/6 mice (Figure 2). Furthermore, our unpublished data (J. Knowles, N. Maeda, 2001) show that the HW/BW ratios of Apoe−/− mice at 4 months of age are ≈15% larger than those of the wild-type mice with a similar genetic background (6 to 8 generations of backcrossing to C57BL/6). The degrees of atherosclerosis at these ages are still minimal. Although we cannot completely eliminate the possibility that some residual contribution of strain 129 derived genes linked to the Apoe locus and that differing in the 2 strains is responsible for this phenotype, a more conservative suggestion is that apoE is important for cardiac energy metabolism and has direct effects in modulating cardiac hypertrophy. Further studies are necessary to elucidate the mechanisms whereby apoE may exert this function.
This study was supported by National Institutes of Health grants HL-62845 and HL-42630. The authors thank Dr H. Rockman for advice on transverse aortic constriction and Drs H.-S. Kim, J. Knowles, L. James, and O. Smithies for useful discussions.
Received November 9, 2001; revision accepted December 31, 2001.
- ↵Franklin SS, Larson MG, Khan SA, Wong ND, Leip EP, Kannel WB, Levy D. Does the relation of BP to coronary heart disease risk change with aging?: the Framingham Heart Study. Circulation. 2001; 103: 1245–1249.
- ↵McGill HCJr, Strong JP, Tracy RE, McMahan CA, Oalmann MC. Relation of a postmortem renal index of hypertension to atherosclerosis in youth: the Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Arterioscler Thromb Vasc Biol. 1995; 15: 2222–2228.
- ↵Herrera VL, Makrides SC, Xie HX, Adari H, Krauss RM, Ryan US, Ruiz-Opazo N. Spontaneous combined hyperlipidemia, coronary heart disease and decreased survival in Dahl salt-sensitive hypertensive rats transgenic for human cholesteryl ester transfer protein. Nat Med. 1999; 5: 1383–1389.
- ↵Piedrahita JA, Zhang SH, Hagaman JR, Oliver PM, Maeda N. Generation of mice carrying a mutant apolipoprotein E gene inactivated by gene targeting in embryonic stem cells. Proc Natl Acad Sci U S A. 1992; 89: 4471–4475.
- ↵Rockman HA, Ross RS, Harris AN, Knowlton KU, Steinhelper ME, Field L, Ross JJr, Chien KR. Segregation of atrial-specific and inducible expression of an ANF transgene in an in vivo murine model of cardiac hypertrophy. Proc Natl Acad Sci U S A. 1991; 88: 8277–8281.
- ↵Gardin JM, Siri FM, Kitsis RN, Edwards JG, Leinwand LA. Echocardiographic assessment of left ventricular mass and systolic function in mice. Circ Res. 1995; 76: 907–914.
- ↵Kim HS, Krege JH, Kluckman KD, Hagaman JR, Hodgin JB, Best CF, Jennette JC, Coffman TM, Maeda N, Smithies O. Genetic control of blood pressure and the angiotensinogen locus. Proc Natl Acad Sci U S A. 1995; 92: 2735–2739.
- ↵Goldblatt H, Kahn JR, Hanzel RF. Studies on experimental hypertension, IX: the effect on blood pressure of constriction of the abdominal aorta above and below the site of origin of both main renal arteries. J Exp Med. 1939; 69: 649–674.
- ↵Nicoletti A, Michel JB. Cardiac fibrosis and inflammation: interaction with hemodynamic and hormonal factors. Cardiovasc Res. 1999; 41: 532–543.
- ↵Weiss D, Kools JJ, Taylor WR. Angiotensin II-induced hypertension accelerates the development of atherosclerosis in apoE-deficient mice. Circulation. 2000; 103: 448–454.
- ↵Yagi S, Kramsch DM, Madoff IM, Hollander W. Plasma renin activity in hypertension associated with coarctation of the aorta. Am J Physiol. 1968; 215: 605–610.
- ↵Stanek KA, Coleman TG, Murphy WR. Overall hemodynamic pattern in coarctation of the abdominal aorta in conscious rats. Hypertension. 1987; 9: 611–618.
- ↵Muthuchamy M, Pieples K, Rethinasamy P, Hoit B, Grupp IL, Boivin GP, Wolska B, Evans C, Solaro RJ, Wieczorek DF. Mouse model of a familial hypertrophic cardiomyopathy mutation in alpha-tropomyosin manifests cardiac dysfunction. Circ Res. 1999; 85: 47–56.
- ↵Du XJ, Autelitano DJ, Dilley RJ, Wang B, Dart AM, Woodcock EA. Beta(2)-adrenergic receptor overexpression exacerbates development of heart failure after aortic stenosis. Circulation. 2000; 101: 71–77.
- ↵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 APOE3 allele enhances diet-induced hypercholesterolemia and atherosclerosis. J Biol Chem. 1997; 272: 17972–17980.
- ↵Hartley CJ, Reddy AK, Madala S, Martin-McNulty B, Vergona R, Sullivan ME, Halks-Miller M, Taffet GE, Michael LH, Entman ML, et al. Hemodynamic changes in apolipoprotein E-knockout mice. Am J Physiol. 2000; 279: H2326–H2334.