Ultrastructure of Early Lipid Accumulation in ApoE-Deficient Mice
Abstract—Apolipoprotein (apo) E–deficient mice develop severe hypercholesterolemia and have lesions that progress from fatty streaks to fibrous plaques distributed in lesion-prone areas throughout the aorta. Lesions develop in apoE-deficient mice on a regular chow diet and will occur faster on a diet higher in cholesterol. Examination of the aortas from these mice on a chow diet by high-resolution, freeze-etch electron microscopy demonstrated lipid retention in the intima by 3 weeks of age. Lipid was retained in the matrix as individual particles between 33 and 48 nm in diameter, aligned along the collagen fibrils and in aggregates consisting of lipid particles with average diameters of 33 and 68 nm. Larger particles seemed to have formed from fusion of smaller particles. Lipid retention was more widespread in 5- and 9-week-old mice. Monocyte attachment to endothelial cells was observed by electron microscopy at 5 weeks of age. The appearance of the intimal lipid was similar to that previously described in rabbit models and suggests that lipid interaction with matrix filaments and subsequent aggregation of lipid particles are critical first steps in the process of foam cell formation.
- Received April 28, 1998.
- Accepted July 31, 1998.
The creation of apoE-deficient mice by homologous recombination of embryonic stem cells has provided a murine model of atherosclerosis.1 2 The mice have severe hypercholesterolemia and have lesions that progress with age from fatty streaks to fibrous plaques distributed in lesion-prone areas. Lesions develop even on a low-cholesterol, low-fat chow diet but occur sooner and become larger on a high-cholesterol, high-fat, Western-type diet.
Two studies have provided detailed histological observations on the morphology of the development of atherosclerotic lesions in apoE-deficient mice. Nakashima et al3 studied animals between 6 and 40 weeks of age fed Western- and chow-type diets. These authors showed that these animals developed the full range of lesions from fatty streaks to fibrous plaques and, most important, that the lesions were similar in their inflammatory-fibroproliferative features to those seen in rabbit models, nonhuman primates, and humans. Reddick et al4 studied mice from 11 to 64 weeks of age fed solely a chow diet. Their morphological studies, at the light-microscope level of resolution, also showed that the progression of atherosclerotic lesions in mice lacking apoE paralleled that described in Watanabe heritable hyperlipidemic rabbits. These histological studies established that the light-microscopic appearance, cellular composition, and distribution of the lesions are typical of lesion development in animal models and similar in some respects to lesions in humans. However, these studies were at the light-microscope level of resolution, and the observations occurred after fatty streaks had begun to form. Given the importance of the apoE-knockout mouse model as an excellent system to study the pathogenesis and progression of atherosclerosis, the objective of the present study was to extend the structural observations to include high-resolution electron microscopy to early and prelesional periods. Electron microscope techniques, especially ultrarapid freeze-etch, was used to investigate apoE-deficient mice on a chow diet at 3, 5, and 9 weeks of age. This study was able to examine lipid retention in the intima before foam cells formed. We were able to show that as early as 3 weeks of age, apoE-deficient mice clearly accumulate lipoproteins within the matrix, in close association with collagen filaments of the intima. The configuration of the retained lipid is similar to that previously described in rabbit models5 6 and suggests that the lipid retention induced by the association of the lipid-rich particles with collagen and matrix filaments is a critical first step in the process of lipid aggregation and monocyte infiltration into the intima.
Animal and Tissue Preparation
Animals were housed at the Rockefeller University Laboratory Animal Research Center in rooms maintained on 12-hour light (7 am to 7 pm)/dark cycles. Diet (rodent chow) and water were provided ad libitum. ApoE-deficient mice were created by homologous recombination in ES cells in the laboratory of Dr Jan Breslow as previously described.1 The original apoE knockout was a C57BL6/129F1. Heterozygotes for the apoE knockout on this background were mated to generate littermate mice on the same F1 background that were either homozygous wild-type or homozygous knockout. ApoE deficiency was detected by elevated total serum cholesterol levels as well as genotyping at the mouse apoE locus as previously described.7 The mice were shipped overnight to the laboratory of Dr Joy Frank. Most of the mice were killed on the day of arrival; however, 2 apoE-deficient mice were kept with food and water for 2 days after arrival and then killed. The heart with the attached aorta was rapidly removed from the animal and placed in oxygenated, phosphate-buffered Ringers’ solution. Pieces of aorta were removed from regions of predilection for development of lesions, as indicated in Figure 1⇓. Tissue pieces from both the control and the apoE-deficient mice were treated similarly. Rectangular pieces of tissue (2×1 mm) were rapidly removed and placed endothelial side up on rectangles of gelatin, which rested on moist filter paper glued to aluminum support disks as previously described.6 8 The tissue was ultrarapidly frozen on a highly polished LN2-cooled copper block in a Life Cell CF 1N freezing device. Use and care of animals is in accordance with NIH guidelines.
Freeze-Fracture, Deep Etching, and Replication
The tissue was fractured on a Balzers 301 freeze-fracture apparatus. Frozen pieces of aorta were fractured superficially (<10 μm) to ensure good freezing and to limit the fracture plane to the intima. The tissue was fractured at −150°C and under vacuum of 1×10−7 mm Hg. For deep-etching, the specimen stage was warmed to −110°C and maintained at this temperature for 3 minutes, followed by 3 minutes at −100°C and 2 minutes at −95°C. Rotary shadowing and replica formation were performed as previously described.5 6 The tissue was digested with household bleach, rinsed in distilled water, and picked up on Formvar-coated grids. A total of 150 replicas were examined from control and experimental animals (10 replicas per animal; total animals=15), and each replica provided ≈500 μm2 of tissue for examination.
Thin-Section Electron Microscopy
Tissue pieces from the aorta were selected from the same locations as indicated in Figure 1⇑. For thin-section electron microscopy, the tissue was fixed in glutaraldehyde followed by OsO4. The segments were dehydrated in graded concentrations of ethanol and embedded in epoxy resin for ultrathin sectioning. Sections were stained with uranyl acetate and lead citrate and observed with a transmission electron microscope (JEOL 100CX).
Freeze-etch replicas from ultrarapidly frozen tissue were generated from normal and apoE-deficient mice (both male and female). Samples were taken from areas throughout the aortic arch, including the lesser curvature and the aortic root. In addition, several samples were taken from the descending thoracic and abdominal aortas. Fracture planes that were in the intima just beneath the endothelial cells and above the internal elastic lamina were examined extensively to determine the presence, configuration, and distribution of lipid particles. The intimas in both wild-type (control) and apoE-deficient mice consisted of a reticular network of branching filaments (size, 5.3 to 7.9 nm) presumed to be mainly proteoglycans. Within this matrix there were numerous collagen fibrils that had contact sites along their length with the matrix filaments.
Three-Week-Old ApoE-Deficient Mice
As early as 3 weeks of age, the first stage of lesion formation, ie, lipid retention in the intima, was present in the apoE-deficient mice but absent in all of the control mice. In discrete areas in the intimas of the apoE-deficient mice, pockets of aggregated lipid were seen enmeshed in the extracellular matrix filaments and in close proximity to collagen fibrils (Figure 2⇓). This observation was in sharp contrast to the normal control (wild-type) 3-week-old mice. The intimas of these normal 3-week-old animals consisted of the same network of extracellular matrix filaments and collagen fibrils but were free of any lipid deposition (Figure 3⇓). Shown in Figure 4⇓ are clusters of lipid particles of various sizes (33 to 66 nm). The larger aggregates contained particles with diameters >60 nm and appeared to result from the fusion of smaller lipid particles. On close inspection and at higher magnification, individual lipid particles could be seen associated with matrix filaments or aligned close to the collagen fibrils (Figures 5⇓ and 6⇓). The close proximity of the lipid particles to the collagen fibrils was a consistent finding (eg, Figure 6⇓). These patterns of lipid deposition were present in the subendothelial matrix throughout the aortic arch (areas 1 through 7 in Figure 1⇑), and they became less prevalent in the subendothelium farther down the aortic tree (area 8 in Figure 1⇑). However, even in the abdominal aorta, there were small accumulations of lipid in a few discrete areas in the subendothelial matrix (Figure 7⇓). In the ≈500 μm2 of intima that was examined in the control (wild-type) specimens, lipid deposition was never seen in the matrix. In the apoE-deficient mice, lipid deposition was seen in the subendothelial matrix in each of the 150 samples that was examined. There was variation in the size and number of lipid aggregate pools present in the intima from replica to replica.
Five-Week-Old ApoE-Deficient Mice
Monocytes were present on the endothelial surface of aortic tissue samples selected from areas 1 through 7, as illustrated in Figure 1⇑. Thin sections prepared for electron microscopy demonstrated monocytes in close association with the endothelial cells (data not shown). At 5 weeks of age, lipid deposition in the subendothelial matrix was similar to that contained in discrete areas of the intima in the 3-week-old mice (Figures 4⇑, 5⇑, and 6⇑). However, at 5 weeks of age, there were widespread areas of the subendothelial space that were “seeded” with individual lipid particles and that contained small aggregates of lipid (Figures 8⇓ and 9⇓).
Nine-Week-Old ApoE-Deficient Mice
The very early retention of lipid within the matrix present in 3- and 5-week-old mice was present throughout the subendothelial space in the 9-week-old animals. Lipid accumulation at all of the various stages of evolution, from individual lipoprotein particles to lipid associated with collagen to clusters of aggregated lipid, could be observed in a single replica. Most typical were large aggregates of lipid consisting of particles of various sizes (Figure 10⇓). The smaller particles (33 to 60 nm) appeared to be fusing, giving rise to the larger particles. Monocyte infiltration into the intima at this stage was frequently observed. For example, Figure 11⇓ illustrates, in a conventionally prepared thin-section electron photomicrograph, a typical monocyte interdigitated between the surface endothelial cells, with part of the monocyte within the subendothelial space. In close proximity to the entering monocyte and just beneath the endothelial cell, there is a large aggregate of lipid. The details of the structure of this lipid aggregate are poorly retained in the conventionally fixed tissue. Figure 10⇓, on the other hand, illustrates a similar subendothelial pool of lipid observed with freeze-fracture electron microscopy. This lipid aggregate immediately beneath the endothelium was typically seen in the 9-week-old apoE-deficient mice.
The lesions that form in apoE-deficient mice on either a chow- or Western-type diet have been shown to be very similar in their location and in their progression to those that develop in hyperlipidemic rabbits, including the cholesterol-fed and Watanabe heritable hyperlipidemic rabbits.3 4 Foam cells and fibrous plaques that developed in the apoE-deficient mice and the rabbit models have similar histological characteristics.3 4
The present freeze-etch morphology data show that lipid retention and aggregation occur in the matrix of apoE-deficient mice before foam cells are present. The configuration of the accumulated lipid is similar to that reported in hyperlipidemic rabbit models.6 8 This is a significant finding because subendothelial retention of atherogenic lipoproteins and their subsequent aggregation may be the central processes in atherogenesis.9 The retention of lipoproteins in the proteoglycan and collagen matrix filaments provides a microenvironment where lipid oxidation and lipid aggregation can occur. The evolution from single lipoprotein entry into the intima, to lipoprotein association with collagen filaments, and then to the aggregation of individual lipid particles into variously sized clusters of particles, was present in the intima in replicas generated from 3-week-old apoE-deficient mouse aortas (Figures 2⇑, 4⇑, 5⇑, and 6⇑) The early disposition of lipoproteins in the intima results in the formation of numerous pockets of aggregated lipid in the subendothelial space adjacent to the entering monocytes.
The mice at 3 weeks of age had just been weaned, and it is possible that this very early lipid retention was related to their high-fat nursing diet. However, an extensive study of normal control mice of the same age failed to reveal any lipid retention in the matrix of these animals (cf Figures 2⇑ and 3⇑). Subsequent sampling of the apoE-deficient mice at 5 and 9 weeks demonstrated the continued progression and more pervasive presence of lipid deposition in the intimas of these animals on a chow diet. It is interesting that Nakashima et al3 in their histological study noted that in the 6-week-old mice (the earliest time they examined), sporadic foam cells were present, suggesting to them that monocyte adhesion and chemotaxis may have occurred even earlier. The retention of lipid in the intimas of the 3-week-old mice and monocyte adhesion to the endothelial cells by 5 weeks of age confirm their hypothesis. Previous histological findings had indicated that lesions are present throughout the aorta in apoE-deficient mice.3 4 It was not surprising, then, to discover that there were small accumulations of lipid aggregates, even in the abdominal aortas of the 3-week-old mice.
The ability of freeze-etch microscopy to detect the earliest lipid-matrix association and to provide a quasi–3-dimensional image of the structures involved in lipid retention may help unravel the complex process of foam cell formation. Studies by Schwenke and Carew10 pointed to lipid retention as an important if not the key step in lesion development. Subsequent in vitro studies showed the affinity of lipoproteins for matrix proteoglycans and collagen fibrils.11 12 Studies with freeze-etch morphology produced the first clear images of the complex structure of the intima and captured the association of retained lipid with the matrix filaments, including collagen fibrils, first in rabbit models8 and now in the present study in the apoE-deficient mouse model. Still unclear are the mechanisms by which the matrix filaments and associated collagen fibrils bind lipoproteins or facilitate their transformation and/or oxidation. It is evident from the freeze-etch data that filaments (between 3.8 and 7.7 nm in diameter) extend from the collagen fibrils at regular intervals and at ≈90o are directly linked to LDL.6 Recent work supports this observation by presenting evidence that decorin, a small proteoglycan with a core protein of 45 kDa and a single dermatan sulfate–rich side chain, can link LDL with collagen I in vitro.13 These experiments showed that when decorin was first allowed to bind to the collagen, binding of LDL to the decorin-collagen complex was >10 fold greater than to collagen alone. Decorin has been shown to bind and modify the fibrillar structure of collagen I.14 15 In addition, both decorin and collagen I have been localized in primary atherosclerotic plaques.16 Fibronectin, another abundant matrix protein, has also been shown to bind lipoproteins in vitro. Recent data demonstrated that after selective removal of heparan sulfate proteoglycans, lipoprotein retention increases, and one possibility suggested by these data was that the increase in lipoprotein(a) binding involved newly accessible sites on fibronectin within the matrix.17
The loss of apoE in the mouse model appears to increase the rate at which lipid is retained in the intima and, as a result, markedly accelerates atherogenesis. The fact that even on a chow diet, 5-week-old mice accumulated lipid in the intima to a sufficient degree to initiate monocyte adhesion to the endothelium is striking. This study provides new data on the earliest stages of foam cell formation in a unique mouse model. Future studies that assess the effects of genetic manipulations on lipid retention in the matrix should provide greater insight into the complex processes that lead to atherogenesis.
This work was supported by National Institutes of Health (Bethesda, Md) Program Project grant HL-30568, Project 1 (to J.S.F.) and by grants HL32435 and HL33714 to J.L.B.
Zhang 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 EW, Breslow JL, Ross R. ApoE-deficient mice develop lesions of all phases of atherosclerosis throughout the arterial tree. Arterioscler Thromb. 1994;14:133–140.
Reddick RL, Zhang SH, Maeda N. Atherosclerosis in mice lacking apoE: evaluation of lesional development and progression. Arterioscler Thromb. 1994;14:141–147.
Nievelstein PFEM, Fogelman AM, Mottino G, Frank JS. Lipid accumulation in rabbit aortic intima 2 hours after bolus infusion of low-density lipoprotein: a deep-etch and immunolocalization study of ultrarapidly frozen tissue. Arterioscler Thromb. 1991;11:1795–1805.
Nievelstein-Post P, Mottino G, Fogelman AM, Frank JS. An ultrastructural study of lipoprotein accumulation in cardiac valves of the rabbit. Arterioscler Thromb. 1994;14:1151–1161.
Smith JD, Trogan E, Ginsberg M, Grigaux C, Tian J, Miyata M. Decreased atherosclerosis in mice deficient in both macrophage colony-stimulating factor (OP) and apolipoprotein E. Proc Natl Acad Sci U S A. 1995;92:8264–8268.
Frank JS, Fogelman AM. Ultrastructure of the intima in WHHL and cholesterol-fed rabbit aortas prepared by ultra-rapid freezing and freeze-etching. J Lipid Res. 1989;30:967–978.
Williams KJ, Tabas I. The response to retention hypothesis of early atherogenesis. Arterioscler Thromb. 1995;15:551–561.
Schwenke DC, Carew TE. Initiation of atherosclerotic lesions in cholesterol-fed rabbits: focal increases in arterial LDL concentration precede development of fatty streak lesions. Arterioscler Thromb. 1989;9:895–907.
Falone DJ, Salisbury BGJ. Fibronectin stimulates macrophage uptake of low density lipoproteins-heparin-collagen complexes. Arteriosclerosis. 1988;8:263–273.
Pentikainen MO, Oorni K, Lassila R, Kovanen PT. The proteoglycan decorin links low density lipoproteins with collagen type I. J Biol Chem. 1997;272:7633–7638.
Vogel KG, Paulsson M, Heinegard D. Specific inhibition of type I and type II collagen fibrillogenesis by the small proteoglycan of tendon. Biochem J. 1984;223:587–597.