Quantitative Assessment of Aortic Atherosclerosis in APOE*3 Leiden Transgenic Mice and Its Relationship to Serum Cholesterol Exposure
Transgenic mice overexpressing the human dysfunctional apolipoprotein E variant, APOE*3 Leiden, develop hyperlipidemia and are highly susceptible to diet-induced atherosclerosis. In the present study, we investigated the effects of diet composition and feeding period on serum cholesterol exposure and the amount of atherosclerosis in the aortic sinus in these mice, using quantitative image analysis. On each of the three diets tested—a low-fat diet, a high-saturated-fat/cholesterol diet, and a high saturated-fat/high-cholesterol/0.5%-cholate diet—transgenic animals showed a marked hyperlipidemia compared with nontransgenic littermates. Measurement of the atherosclerotic lesion areas in cross sections of the aortic sinus in animals exposed to these three diets for up to 6 months showed a 5 to 10 times greater lesion area in transgenic mice compared with nontransgenic controls. Highly significant positive correlations were found between the log-transformed data on lesion area and serum cholesterol exposure (r=.82 to .85 for the 1-, 2-, and 3-month treatment groups), indicating that the hyperlipidemia is likely to be a major determinant in lesion formation. On the basis of these findings, we suggest that the APOE*3 Leiden mouse represents a promising model for intervention studies with hypolipidemic and antiatherosclerotic drugs.
- Received November 14, 1995.
- Revision received February 7, 1996.
Apolipoprotein E (apoE) is a major regulator of plasma lipoprotein metabolism, functioning as a ligand for the receptor-mediated uptake of chylomicron and VLDL remnants, as well as apoE-rich HDL.1 2 The importance of apoE is well illustrated in familial dyslipoproteinemia, whose underlying cause is one of several mutations in the APOE gene, resulting in the production of dysfunctional apoE or its absence from plasma.3 One of those variants, APOE*3 Leiden, is a tandem duplication of codons 120 through 126 in the apoE gene, and the presence of a single allele for this mutation results in the expression of familial dyslipoproteinemia.4 5 6 The dominance of the APOE*3 Leiden trait in expressing this disease is probably due to the high affinity of the dysfunctional apoE3 Leiden over wild-type apoE for triglyceride-rich lipoproteins.7 Recently we have described the development of a transgenic mouse carrying a transgene containing the APOE*3 Leiden and APOC1 genes plus the downstream regulatory elements for enhanced liver expression.8 APOE*3 Leiden mice were shown to develop hyperlipidemia, which tendency could be greatly enhanced by feeding a diet enriched in saturated fat, cholesterol, and sodium cholate. In a subsequent study9 an increased susceptibility for diet-induced atherosclerosis was also demonstrated. This mouse model of hyperlipidemia and premature atherosclerosis is of interest for investigating the influence of environmental and genetic factors on disease progression and for testing candidate hypolipidemic and antiatherosclerotic drugs. The present report describes a detailed study on the effect of diet composition and duration of exposure to these diets on hyperlipidemia and on the lesion area in the aortic root using color video image analysis.
All animal procedures were approved by the institutional committees on animal experimentation of TNO-Prevention and Health and SmithKline Beecham. Transgenic mice expressing human APOE*3 Leiden and APOC1 genes (line 2), were generated as described by van den Maagdenberg et al.8 Microinjection of the transgene was performed in eggs from (C57BL/6JxCBA/J) F1 females that had been mated to males of the same genetic background, and transgenic founder mice were bred with C57BL/6J mice to establish a transgenic strain. Subsequent generations were produced by mating male transgene carriers (identified with an ELISA for human apoE9 ) with C57BL/6J females. Nontransgenic littermates were used routinely as controls. In the present study 132 females (66 transgenic and 66 nontransgenic) from the F7 generation were used. At the start of the study, animals aged 8 to 10 weeks were allocated randomly to one of the experimental groups on the basis of age and litter and were housed in groups of up to 8. Animals had free access to water and food.
Before the study, animals were kept on standard rat/mouse chow (SRM-A, Hope Farms). During the experimental period, animals were fed one of three semisynthetic diets containing sucrose as the main energy source. These diets were essentially composed as described by Nishina et al10 11 and were a low-fat/low-cholesterol control diet (LFC); a cocoa butter (15%) and cholesterol (0.25%)–enriched high-fat/cholesterol diet (HFC); and a cocoa butter (15%), cholesterol (1%), and cholate (0.5%)–enriched high-fat/high-cholesterol/cholate diet (HFC/0.5% cholate) and were all formulated by Hope Farms. The composition is shown in the Table⇓. Experimental diets were fed for up to 3 months (LFC and HFC/0.5% cholate diets) or 6 months (HFC diet), and all diets were well tolerated by the animals. Body weights were not significantly different among animals on any of the diets at any of the time points tested. None of the animals was lost during the study.
Lipid and Lipoprotein Analysis
Blood samples were taken from the tail under light diethylether anesthesia, after an overnight fast. Total serum cholesterol and triglyceride concentrations (without free glycerol) were measured enzymatically with commercial test kit 236691 from Boehringer Mannheim GmbH and test kit 337-B from Sigma, respectively. For a more detailed analysis of the serum lipoprotein profiles, Superose 6B column chromatography was used. Equal volumes of serum from all the animals in a group were pooled and 200-μL aliquots analyzed on a 25-mL Superose 6B column (Pharmacia AB) eluted at a constant rate of 0.5 mL/min with PBS buffer (pH 7.4) using a high-performance liquid chromatography pump system. Cholesterol concentrations in the lipoprotein fractions were analyzed as described above.
Perfusion of Hearts and Aortas
Mice were killed by sodium pentobarbital (Nembutal) anesthesia, and the thorax was opened. The heart and vascular tree were perfused in situ with oxygenated Krebs' Ringer bicarbonate buffer at 37°C under a pressure of ≈120 cm of water through a cannula positioned into the left ventricle and an outlet created by cutting the lower vena cava. After 30 minutes, the buffer was replaced by neutral-buffered formalin (3.7% formaldehyde, Formal-fix, Shandon Scientific Ltd at 37°C), and the perfusion fixation continued for a further 30 minutes. Finally, the hearts and aortas were dissected out, cleaned of extravascular fat, and stored in neutral-buffered formalin until they were processed.
Tissue Preparation and Sectioning of the Aortic Root
The hearts were bisected just below the atria, and the base of the hearts plus aortic root were taken for analysis. The tissue was washed three times over a period of 8 hours in fresh OCT liquid (Bayer Diagnostics) and then left overnight in OCT liquid. The following day, hearts were placed in fresh OCT liquid on a cryostat chuck (Bright Instrument Company Ltd), with the aorta facing the chuck and frozen using dry ice. The hearts were then sectioned perpendicular to the axis of the aorta, starting within the heart and working in the direction of the aortic arch as described by Paigen et al.12 Once the aortic root was identified by the appearance of aortic valve leaflets, alternate 10-μm sections were taken and mounted on gelatinized slides. Sections were air dried for 1 hour and rinsed briefly in 60% isopropyl alcohol. Sections were then stained with oil red O, counterstained with Mayer's hematoxylin stain, coverslipped with glycerol gelatin, and sealed with nail varnish. These sections were used for quantification of aortic atherosclerosis, as detailed below.
In some animals the perfused-fixed aortas were used to visualize and document atherosclerosis over the whole length of the blood vessel. To this end, aortas were cleaned of extravascular fat, stained for lipids with oil red O, and studied under a stereomicroscope (Olympus, SZH-10) fitted with a 3-chip color video camera (model HV-C10, Hitachi) and a video color printer (Mavigraph, Sony) for documentation. Aortas were also opened longitudinally to study lesion morphology en face.
Quantification of Atherosclerosis in Sections of the Aortic Root
Up to 40 sections of the aortic root per animal were imaged using an Olympus BH-2 microscope equipped with a 4× objective, neutral density (ND-6) and blue (KB-4) filters, and a video camera (Hitachi, HV-C10). Full color 24-bit images were acquired by using a PC fitted with a frame-grabbing board (MFG/3M/V, Data Cell, Ltd) and running Optimas software (version 4.02, Bioscan Inc). The images were all captured under identical lighting, microscope, camera, and PC conditions and were stored on optical discs (Panasonic) in TIFF files. Quantification of atherosclerotic lesion areas in these cross sections (from here on referred to as lesion area) was performed with the Optimas software. We initially selected threshold values (red, green, and blue) that discriminated between lesion and nonlesion areas by using the lesions in ≈12 histological sections. Then, as we continued with the area measurements, we verified that the thresholds we had selected were suitable for each section by having the image analysis program lay a mask over the areas within our thresholds. The threshold values we had selected were considered to be satisfactory because the mask covered all of the lesioned areas. Extra vascular fat and other nonlesion areas also identified by the threshold setting were edited out before lesion areas were computed. Absolute values for the surface area of lesions were obtained by calibration of the software using the image of the grid on a hemocytometer slide. After the initial analysis of all sections in some animals (see “Results”), we analyzed routinely the first 10 sections in the direction of the aortic arch from the point where all three aortic valve leaflets first appeared.
Estimation of Vascular Exposure to Serum Cholesterol
To investigate the possible relationship between lesion area and exposure of the arterial wall to increased concentrations of plasma cholesterol, cholesterol exposures were calculated for each animal. Exposures were defined as the areas under the curve (AUCs) in serum cholesterol versus time plots. All animals used in the study were bled only once. Bleeding was done at the time the animals were killed. To calculate the cholesterol exposure, we made best estimates of serum cholesterol concentration for the preceding month(s) for each individual animal. For this calculation, it was assumed that an animal with a relatively high serum cholesterol at the time of death, eg, 3 months in comparison with the other animals in its experimental group killed at 3 months would be likely to have had a similarly high relative serum cholesterol concentration at 1 and 2 months. As mean serum cholesterol concentrations of groups of animals on the same diet at 1 and 2 months were available (from the animals that were killed at 1 and 2 months for lesion analysis), 1- and 2-month serum cholesterol concentrations in the 3-month group could be estimated for each individual animal, and plots of serum cholesterol against period on diet for each animal constructed. From those plots, the AUCs for each animal were calculated.
Serum lipid data sets were analyzed and, as some showed a nongaussian distribution, comparisons between groups were routinely made by a nonparametric Mann-Whitney test, using RS/1 statistical software (BBN Software Products Corp).
The statistical significance of differences in lesion area between control and transgenic mice was determined by a two-tailed unpaired Student t test.
To ascertain whether there was any relationship between the cholesterol exposures and lesion areas, data were analyzed after the possible differences between diets and periods of killing were considered. The ANOVA technique requires that data be normally distributed and that variability be similar across groups. A plot of SD versus arithmetic mean derived for each subgroup of six animals strongly indicated that for both cholesterol exposure and lesion area, variability increased as the mean value increased. On a logarithmic scale, however, this relationship disappeared; hence both variables were transformed to logarithms. Residual plots performed after the ANOVA further validated the above two assumptions.
To explore the possible relationships between the log10 (lesion area) data and log10 (cholesterol exposure) period and diet, a full model, with all possible interaction terms was first fitted and an ANOVA performed. Since all the interactions were highly insignificant (P>>.05), the data were then fitted, with the main effects being diet and period and the covariate log10 (cholesterol exposure). In this analysis, the effect of diet was insignificant, and hence diet was removed from the model and the data from all the dietary groups were combined. All these tests were performed without consideration of the data in the 6-month (period 6) group, because these animals were all fed the same diet.
An ANOVA was again performed to investigate the interaction between log10 (cholesterol exposure) and period. The data from period 6 were included for this analysis. Here, the interaction term would test for the parallelism of the slopes of the linear associations for the periods, if any. The interaction was not statistically significant at the 5% level of significance (P>.05), so it was dropped from the model. This result justified considering equal slopes for the four periods. The ANOVA was then repeated without the interaction term. Here, the log10 (cholesterol exposure) term indicated a statistical significance (P=.0001), and there were statistically significant differences between the four periods (P=.0230). The significant effect for period signifies different intercepts for the regression lines for the four periods. Pairwise tests between periods showed that the 1-month group was significantly different from the 2-, 3-, and 6-month groups (P=.0094, P=.0075, and P=.0094, respectively). However, there was no statistical significance for the differences among periods 2, 3, and 6 months. Therefore, the 2-, 3-, and 6-month groups were pooled and tested against period 1. The results of this ANOVA again showed significant difference (P=.0036), thus justifying fitting lines with the same slope but with different intercepts for the 1-month group and the combined 2-, 3-, and 6-month group.
Effects of Different Diets on Plasma Lipoprotein Concentrations
The effects of a switch from normal mouse diet to one of three semisynthetic diets on serum cholesterol and triglyceride concentrations in female APOE*3 Leiden mice and their nontransgenic control littermates are shown in Figs 1 and 2⇓⇓,⇓⇓ respectively. Experimental diets were chosen to simulate a Western-type diet (HFC) or a more severely atherogenic diet (HFC/0.5% cholate) and compared with the control diet (LFC). All diets contained sucrose as the major source of energy and were based on formulations developed by Nishina et al10 11 to limit the side effects (liver damage and gallstone formation) seen with traditional atherogenic mouse diets of chow supplemented with fat, cholesterol, and bile salts. The effects of the experimental diets in the present study were investigated for a period of up to 3 months (diets LFC and HFC/0.5% cholate) or 6 months (diet HFC).
Compared with a standard mouse diet, all three semisynthetic diets increased serum cholesterol concentrations in transgenic mice as well as control animals, while serum triglyceride concentrations were decreased rather than increased on the HFC and HFC/0.5% cholate diets. On all three semisynthetic diets and at all time points tested, serum lipids were manyfold (2 to 13 times) higher in the APOE*3 Leiden mice than in the corresponding control groups, differences being the most prominent on the HFC/0.5% cholate diet (compare Fig 1A and 1B⇑⇑ for cholesterol and Fig 2A and 2B⇑⇑ for triglycerides). All differences in serum lipids between APOE*3 Leiden and nontransgenic control mice were highly statistically significant.
The distribution of lipids among the different lipoprotein classes was analyzed by Superose 6B column chromatography on pools of serum from animals in the same experimental group. Lipoprotein profiles of animals after 2 months of feeding these experimental diets are shown in Fig 3⇓. Control animals on the LFC and HFC diet carry most of their serum cholesterol in the HDL fraction. On the same diets, transgenic mice showed a marked increase in cholesterol associated with VLDL/LDL-sized fractions, which accounted for more than 50% of total plasma cholesterol in these animals. The shifts in the profiles were more pronounced in animals on the HFC diet than on the LFC diet. On the HFC/0.5% cholate diet, the differences between the control and transgenic animals were even more outspoken. In control animals, this diet raised cholesterol associated with the VLDL/LDL-sized fraction severalfold but in the transgenic animals, increases were massive. Although fast protein liquid chromatography analysis of hyperlipidemic serum does not result in a complete separation of HDL from VLDL/LDL, the profiles in Fig 3⇓ (at least for the LFC and HFC diets) indicate that HDL concentrations in transgenic animals on these diets are similar to or higher than those of control animals on the same diet. Similar lipoprotein distribution profiles as shown in Fig 3⇓ were found in the corresponding 1-, 3-, and 6-month groups (data not shown).
Effect of Different Diets on the Development of Atherosclerosis in the Aortic Sinus
On all diets tested, APOE*3 Leiden transgenic mice showed enhanced aortic atherosclerosis compared with nontransgenic controls. Dramatic differences were seen in lesion development between transgenic animals fed the HFC/0.5% cholate, HFC, and LFC diets. Atherosclerosis in control animals was minimal except in those fed the HFC/0.5% cholate diet for 2 and 3 months. Photomicrographs of various stages of lesion development in cross sections of the aortic root are shown in Fig 4A through 4F⇓. Using the nomenclature used by Qiao et al,13 we found that early lesions in all groups were mainly of the type I (oil red O–positive plaques related to aortic valves, ie, on valve attachments, valve cusps, and valve residuals) although type II lesions (oil red O–positive lesions on the free aortic wall) were also seen. Early type I lesions consisted of small, sometimes intensely oil red O–stained plaques (arrows 1 and 3 in Fig 4A and 4B⇓⇓, APOE*3 Leiden mice on the HFC diet for 2 months). These early lesions are rather superficial and contain a single or double layer of lipid-laden foam cells. Early type II lesions are very similar in appearance. An example of an early type II lesion is also seen in Fig 4A⇓ (arrow 2).
The early types I and II plaques develop into more extensive lesions in which the entire arterial wall within the cusp is covered (Fig 4C and 4D⇑⇑, APOE*3 Leiden mice on the HFC diet for 3 months). Those lesions are raised and rich in lipid-laden foam cells underlying an apparently intact layer of endothelial cells (arrow 4 in Fig 4D⇑). Interestingly, extensive lesion development can sometimes be seen in a single cusp, while the other cusps in the same section are totally unaffected (see Fig 4C⇑).
In transgenic mice exposed to the atherogenic HFC/0.5% cholate diet for 3 months, even more complex lesions are seen (Fig 4E and 4F⇑⇑). These lesions often have a core that stains intensively with oil red O and a cap of several layers of spindle-shaped cells (Fig 4F⇑). Extracellular lipid dominates in the core (arrow 6 in Fig 4F⇑), but lipid-laden foam cells are also present, both in the core and among the spindle-shaped cells in the cap. The core of some of these more complex lesions contains areas that stain purple with hematoxylin, suggesting the presence of calcification (Qiao et al,13 arrow 7 in Fig 4F⇑). Many of the complex lesions are very friable, as plaques easily shear (eg, the white area to the left of arrow 7 in Fig 4F⇑) and even detach from the aortic wall during sectioning. In none of the animals, even after 3 months on the HFC/0.5% cholate diet, were lesions seen in the proximal coronary arteries (arrow 9 in Fig 4E⇑), but lesions were present in discrete areas in the ascending aorta, aortic arch, and descending aorta. Fig 5A and 5B⇓⇓ shows the presence of these lesions in the inner curvature of the aortic arch and branch points with the carotid arteries. Most of these lesions are rather superficial and represent early fatty streaks, although raised lesions, eg, at the branch point of carotid arteries, were also seen. No lesions were present in nontransgenic littermates exposed to the same dietary regimen (Fig 5C⇓) compared with APOE*3 Leiden transgenics (Fig 5D⇓).
Quantification of Lesion Area
Atherosclerotic lesions were quantified in cross sections of the aortic roots of all animals using video image analysis, as described in “Methods.” Initially, all sections from the animals on the HFC/0.5% cholate diet were imaged and analyzed. As shown in Fig 6⇓, APOE*3 Leiden mice on this diet developed lesions reasonably uniformly over the whole of this part of the aorta. On the basis of these results, it was decided to analyze all the other animals in the study using sections 1 through 10 only. The results of this analysis are shown in Fig 7⇓.
Massive differences were seen in lesion areas between animals on LFC, HFC, and HFC/0.5% cholate diets (note the difference in scale of the y axis in Fig 7A through 7C⇑). On the control LFC diet, atherosclerotic lesions were small and variable. Only at 2 months were differences between nontransgenic and transgenic animals significantly different. On the HFC diet, a clear time-dependent increase in lesion area was seen in the transgenic mice, and differences with control animals were large (5 to 10 times) and statistically significant (except for the 3-month time point, at which interanimal variation in the transgenic group was particularly large). On the HFC/0.5% cholate diet, a time-dependent increase in lesion area was again observed, especially in transgenic mice, the average levels being about 10 times higher than in the corresponding groups on the HFC diet. Interanimal variability in all groups was appreciable.
To determine the relationship between plasma cholesterol levels and lesion area in APOE*3 Leiden mice, cholesterol exposures were calculated for each animal, as described in “Methods” and plotted against the mean area of plaque measured. As discussed in that section, data required a log10 transformation and were analyzed by ANOVA. The results of this analysis are given in Fig 8⇓. Impressive correlations were found in each of the 1-, 2-, and 3-month subgroups (r=.82 to .85), and pairwise testing indicated that the data of the 2-, 3-, and 6-month groups could be combined, whereas those of the 1-month group showed a somewhat different relationship. Assuming a linear log/log relationship, the data could be best described by two closely parallel regression lines, one for the 1-month group (y=1.32+2.34x, r=.82) and one for the combined 2, 3-, and 6-month groups (y=0.72+2.34x, r=.84). (The inability to fit the 1-month data on the same regression line as the combined 2-, 3-, and 6-month groups may be due to an underestimation of the cholesterol exposure in the former group, as serum cholesterol may not increase linearly during the first month, as assumed in the AUC calculation, but faster.) These interesting results indicate that in APOE*3 transgenic mice plasma cholesterol concentrations and the duration of hyperlipidemia are major determinants of plaque formation.
Previously, we have shown that transgenic mice carrying the dysfunctional human APOE*3 Leiden gene develop hyperlipidemia and premature atherosclerosis.8 9 In the present study, we performed a quantitative analysis of the effects of duration of exposure to different diets on the development of atherosclerotic lesions in these animals. Compared with nontransgenic littermates on the same regimen, APOE*3 Leiden mice show a 5 to 10 times increased lesion area when fed an HFC or HFC/0.5% cholate diet. On both these diets, a clear phenotype is also seen in the hypercholesterolemic response, with serum cholesterol concentrations being increased 5 to 10 times those of their nontransgenic littermates and mainly confined to cholesterol associated with the VLDL/LDL-sized lipoprotein fractions. A clear phenotype was also seen in VLDL/LDL cholesterol concentrations in mice on the LFC diet, but the hyperlipidemia was mild compared with the HFC and HFC/0.5% cholate diets, and this LFC dietary regimen did not result in a consistent significantly increased lesion area over the 3-month period.
Morphological analysis of the lesions showed a pattern seen in many other transgenic mice models with increased susceptibility for atherosclerosis. Early aortic lesions were seen in the valve cusps and valve attachments, and these lesions progressed into raised complex lesions with an extracellular lipid core covered by a cap after longer exposure to atherogenic diets. On the HFC/0.5% cholate diet, distinct fatty streak–like lesions were also seen in the inner curvature of the aortic arch and at branch points of major arteries in the APOE*3 Leiden transgenics but not in their nontransgenic littermates.
Comparison of the phenotype in this transgenic mouse model with that of related models is of interest. Several groups, including ours, have generated mice in which the ApoE gene has been silenced by gene targeting, and these ApoE knockout mice were shown to develop severe hyperlipidemia and atherosclerosis, even when fed a normal chow diet.14 15 16 17 Heterozygous ApoE knockout mice, when fed an atherogenic diet, were also shown to have hyperlipidemia and an increased susceptibility to atherosclerosis, demonstrating that the subnormal expression of the ApoE gene also leads to a (mild) phenotype.16 On the basis of in-house experience with all three models, APOE*3 Leiden mice are more susceptible to atherosclerosis than the heterozygous ApoE knockout mouse. For comparison, we previously showed16 that female heterozygous ApoE knockout mice on the HFC/0.5% cholate diet for 3 months developed lesions with a surface area of 61 000 μm2 per cross section, compared with 270 000 μm2 in APOE*3 Leiden mice, found in the present study. However, APOE*3 Leiden mice were less susceptible than the homozygous ApoE knockout mouse, which developed lesions even on a normal chow diet. In contrast to ApoE knockout mice, APOE*3 Leiden mice have the ability to synthesize functional endogenous apoE, and although this apoE may not be very effective as a ligand on triglyceride-rich lipoproteins for uptake via lipoprotein receptors (presumably due to competition with the dysfunctional APOE*3 Leiden6 7 ), other functions may not be affected. For instance, lipid-laden macrophages synthesize large quantities of apoE,18 and this process may be associated indirectly or directly with cholesterol efflux from these cells.18 19 20 21 This process presumably is still functional in APOE*3 Leiden mice but not in ApoE knockout mice. In this context, two other murine models, the bone marrow–transplanted ApoE knockout mice22 and mice carrying an apoE transgene under control of a promoter that stimulates expression in cells of the arterial wall,23 are of major interest. In the first model, restoration of apoE synthesis in monocyte-derived macrophages by bone marrow transplantation decreases the susceptibility of these mice for atherosclerosis. However, this effect may not be solely due to stimulated cholesterol efflux from arterial macrophages, as this intervention also decreases serum lipids, presumably due to apoE synthesis by liver-fixed macrophages (Kupffer cells) and consequently stimulates hepatic uptake of VLDL and chylomicron remnants. In the second model, a tissue-specific overexpression of apoE in arterial tissue was induced by introduction of an apoE transgene under the control of an H2 Ld promoter. These transgenic mice showed a decreased susceptibility for diet-induced atherosclerosis compared with nontransgenic controls but no differences in serum lipids, supporting the concept of a direct protective effect of locally synthesized apoE on arterial lesion development. We therefore postulate that the less severe atherogenic phenotype of APOE*3 Leiden mice compared with homozygous ApoE knockout mice is at least partly due to its unaltered ability to synthesize functional apoE in extrahepatic cells, including vascular macrophages.
After our initial report on the APOE*3 Leiden mouse,8 Fazio et al24 described another transgenic mouse model, overexpressing a human dysfunctional human apoE gene (APO*E Arg 112, Cys 142). These mice also developed hyperlipidemia and an increased susceptibility for diet-induced atherosclerosis. Whether the phenotype of those mice is similar or differs from that found in the APOE*3 Leiden remains to be seen.
The data collected in the present study allowed us to address the question of whether the level of hyperlipidemia and the period of exposure of the vascular tree to the hyperlipidemia are indeed important determinants for lesion area. Previously, Nishina et al11 have shown in C57BL/6J mice on various lipid-rich diets that the lesion area is positively correlated with the ratio of VLDL+LDL to HDL cholesterol and negatively with HDL cholesterol concentrations. However, no correlation was found between lesion area and plasma cholesterol and/or triglycerides. Studies in WHHL rabbits25 have also failed to show a clear relationship between serum cholesterol and the severity of aortic atherosclerosis. In contrast, in the present study, a strong positive correlation was observed between the vascular exposure to plasma cholesterol in APOE*3 Leiden transgenics and lesion area in the aortic root. As differences in serum cholesterol between animals on different diets are mainly localized in the VLDL to LDL fraction, it is tempting to speculate that this fraction is responsible for the observed correlation. It should be noted that the lesion areas found in the present study in APOE*3 Leiden transgenics span a far greater range (0 to 450 000 μm2) than in the studies of Nishina et al (0 to 4000 μm2), underlining the significance of the present finding. Because HDL cholesterol concentrations were not specifically measured in the present study and are difficult to determine from Superose 6B profiles with dominant VLDL/LDL contributions, a possible negative relation between lesion area and exposure to HDL cholesterol could not be assessed.
Transgenic mice with a susceptibility for diet-induced hyperlipidemia and atherosclerosis are potentially useful small-animal models for the testing of hypolipidemic and antiatherosclerotic agents. In light of the excellent correlation between cholesterol exposure and lesion area, we postulate that the APOE*3 Leiden mice may be particularly useful in that respect. Studies along this line are presently in progress.
We wish to thank H. Belinda van 't Hof for expert technical assistance and gratefully acknowledge Amit Bhattacharyya for support in the statistical analysis. We also thank Dr Nigel Toseland for advice and for reading the manuscript.
Havel RJ, Chow YS, Windler E, Kotite L, Guo LSS. Isoprotein specificity in the hepatic uptake of apolipoprotein E and the pathophysiology of familial dysbetalipoproteinemia. Proc Natl Acad Sci U S A. 1980;77:4349-4353.
Mahley RW. Apolipoprotein E: cholesterol transport protein with expanding role in cell biology. Science. 1988;240:622-630.
Wardell MR, Weisgraber KH, Havekes LM, Rall SC. Apolipoprotein E3 Leiden contains a seven–amino acid insertion that is a tandem repeat of residues 121 to 127. J Biol Chem. 1989;264:21205-21210.
Fazio SY, Horie KH, Weisgraber KH, Havekes LM, Rall SC. Preferential association of apolipoprotein E Leiden with very low density lipoproteins of human plasma. J Lipid Res. 1993;34:447-453.
van den Maagdenberg AMJM, Hofker MH, Krimpenfort PJA, de Bruijn I, van Vlijmen B, van der Boom H, Havekes LM, Frants RR. Transgenic mice carrying the apolipoprotein E3-Leiden gene exhibit hyperlipoproteinemia. J Biol Chem. 1993;268:10540-10545.
van Vlijmen BJM, van den Maagdenberg AMJM, Gijbels MJJ, van der Boom H, HogenEsch H, Frants RR, Hofker MH, Havekes LM. Diet-induced hyperlipoproteinemia and atherosclerosis in apolipoprotein E3-Leiden transgenic mice. J Clin Invest. 1994;93:1403-1410.
Nishina PM, Verstuyft J, Paigen B. Synthetic low and high fat diets for the study of atherosclerosis in the mouse. J Lipid Res. 1990;31:859-869.
Nishina PM, Lowe S, Verstuyft J, Naggert JK, Kuypers FA, Paigen B. Effects of dietary fats from animal and plant sources on diet-induced fatty streak lesions in C57Bl/6J mice. J Lipid Res. 1993;34:1413-1422.
Qiao J-H, Xie P-Z, Fishbein MC, Kreuzer J, Drake TA, Demer LL, Lusis AJ. Pathology of atheromatous lesions in inbred and genetically engineered mice. Arterioscler Thromb. 1994;14:1480-1497.
Zhand S, Reddick RL, Piedrahita JA, Maeda N. Spontaneous hypercholesterolemia and arterial lesions in mice lacking apolipoprotein E. Science. 1992;258:468-471.
Nakashima Y, Plump AS, Raines EWR, Breslow JL, Ross R. Apo E-deficient mice develop lesions in all phases of atherosclerosis throughout the arterial tree. Arterioscler Thromb. 1994;14:133-140.
Basu SK, Goldstein JL, Brown MS. Independent pathways for secretion of cholesterol and apolipoprotein E by macrophages. Science. 1983;219:871-873.
Rosenfeld ME, Butler S, Ord VA, Lipton BA, Dyer CA, Curtiss LK, Palinski W, Witztum JL. Abundant expression of apolipoprotein E by macrophages in human and rabbit atherosclerotic lesions. Arterioscler Thromb. 1993;13:1382-1389.
Schmitz G, Robenek H, Krause R, Schurek A, Niemann R. Ca2+ antagonists and ACAT inhibitors promote cholesterol efflux from macrophages by different mechanisms, I: characterization of cellular lipid metabolism. Arteriosclerosis. 1988;8:46-56.
Kruth HS, Skarlatos SI, Gaynor PM, Gamble W. Production of cholesterol-enriched nascent high density lipoproteins by human monocyte-derived macrophages is a mechanism that contributes to macrophage cholesterol efflux. J Biol Chem. 1994;269:24511-24518.
Linton RF, Atkinson JB, Fazio S. Prevention of atherosclerosis in apolipoprotein E-deficient mice by bone marrow transplantation. Science. 1995;267:1034-1037.
Shimano H, Ohsuga JO, Shimada M, Namba Y, Gotoda T, Harada K, Katsuki M, Yazaki Y, Yamada N. Inhibition of diet-induced atheroma formation in transgenic mice expressing apolipoprotein E in the arterial wall. J Clin Invest. 1995;95:469-476.
Fazio S, Sanan DA, Lee Y-L, Ji Z-S, Mahley RW, Rall SC. Susceptibility to diet-induced atherosclerosis in transgenic mice expressing a dysfunctional human apolipoprotein E (Arg 112, Cys 142). Arterioscler Thromb. 1994;14:1873-1879.
Shiomi M, Ito T, Shiraishi M, Watanabe Y. Inheritability of atherosclerosis and the role of lipoproteins as risk factors in the development of atherosclerosis in WHHL rabbits: risk factors related to coronary atherosclerosis are different from those related to aortic atherosclerosis. Atherosclerosis. 1992;96:43-52.