Endothelial lipase modulates susceptibility to atherosclerosis in apolipoprotein-E-deficient mice.

Endothelial lipase (EL) expression correlates inversely with circulating high density lipoprotein (HDL) cholesterol levels in genetic mouse models, and human genetic variation in this locus has been linked to differences in HDL cholesterol levels. These data suggest a role for EL in the development of atherosclerotic vascular disease. To investigate this possibility, LIPG-null alleles were bred onto the apoE knockout background, and the homozygous double knockout animals were characterized. Both apoE knockout and double knockout mice had low HDL cholesterol levels when compared with wild-type mice, but the HDL cholesterol levels of the double knockout mice were higher than those of apoE knockout mice. Atherogenic very low density lipoprotein and intermediate density lipoprotein/low density lipoprotein cholesterol levels of the double knockout mice were also greater than those of the apoE knockout animals. Despite this lipid profile, there was a significant approximately 70% decrease in atherosclerotic disease area in double knockout mice on a regular diet. Immunohistochemistry and protein blot studies revealed increased EL expression in the atherosclerotic aortas of the apoE knockout animals. An observed decrease in macrophage content in vessels lacking EL correlated with ex vivo vascular monocyte adhesion assays, suggesting that this protein can modulate monocyte adhesion and infiltration into diseased tissues. These data suggest that EL may have indirect atherogenic actions in vivo through its effect on circulating HDL cholesterol and direct atherogenic actions through vascular wall processes such as monocyte recruitment and cholesterol uptake.

It is known that members of the highly conserved lipase family of lipid metabolic enzymes affect vascular disease through regulation of high density lipoprotein cholesterol (HDL-C) 1 levels, as well as through local actions in the vessel wall. Regulation of HDL-C level is important since serum HDL level is widely accepted as an inverse correlate of coronary heart disease risk (1,2). Lipases have been shown to regulate HDL-C through effects on HDL particle size and metabolism (3). Further, lipoprotein lipase and hepatic lipase perform functions in the vessel wall that are independent of their catalytic activity. It is well known that lipases can form a molecular bridge between endothelial cells and lipoproteins or circulating macrophages through interaction with heparan sulfate proteoglycans (4). This nonenzymatic action of lipases can increase cellular lipoprotein uptake and monocyte adhesion. For instance, the local expression of lipoprotein lipase in the vascular wall has been implicated in the local trapping of lipid and acceleration of atherosclerosis through this bridging function (5).
To further investigate the genetic basis of HDL-C level and the modulation of atherosclerosis by lipases, recent studies have focused on a new member of the lipase gene family termed endothelial lipase (EL) (gene nomenclature LIPG, locus link I.D. 16891) (6,7). EL has been shown to be expressed by vascular cells, including endothelial cells (6,7), and to be highly regulated in these cells in vitro and in vivo (8). Recent gene manipulation and antibody blocking experiments have established EL to be a primary determinant of HDL levels in the mouse (9 -11). Further, association-based human genetic studies have provided evidence that variation in the LIPG locus is linked to differences in circulating HDL-C levels (10,12,13). Recently, it has been shown that EL directly mediates cellular lipoprotein uptake (14,15).
Because of its significant effect on HDL levels and the documented association of other lipase family members with atherosclerosis, it seemed possible that EL might also play a role in atherogenesis. Here we report studies confirming this hypothesis by showing that apoE knockout animals, which also are targeted in the LIPG locus, have decreased atherosclerosis. blot was performed for genotyping the LIPG locus (9). Between 14 and 16 mice were included in each experimental subgroup. ApoEϪ/Ϫ mice and LIPGϪ/Ϫ, apoEϪ/Ϫ mice were weaned at 4 weeks of age onto a high fat diet (0.15% cholesterol, 21% milk fat, 19.5% casein) or normal chow diet and maintained on the diet for 12 weeks. Wild-type (WT) and LIPGϪ/Ϫ mice were used as control groups. All animal experiments were conducted according to the Guidelines for Animal Experiments at the Stanford University School of Medicine.
Histological Analysis of Atherosclerotic Lesions-The apoEϪ/Ϫ and LIPGϪ/Ϫ, apoEϪ/Ϫ mice were euthanized at the age of 16 weeks (12 weeks on the chow diet or high fat diet), and the atherosclerotic lesions were analyzed as described previously (17). The aortic samples were fixed in 10% buffered formalin phosphate, embedded in OCT compound, and sectioned (10 m thickness). Five consecutive sections, spanning 550 m of the aortic root, were collected from each mouse and stained with Sudan III or Masson's trichrome. For quantitative analysis of atherosclerosis, the total lesion area of five sections from each mouse was measured with the NIH Image 1.61 software by modifying the method reported previously (18). The amount of aortic lesion formation in each animal was measured as the percentage of the lesion area per total area of the aorta (17).
Analysis of Plasma Lipids-Following an overnight fast, whole blood was collected by cardiac puncture into tubes containing 0.3 mg of EDTA. Plasma was collected by centrifugation at 8000 ϫ g for 5 min at 4°C. Total plasma cholesterol, triglyceride (TG), and phospholipid (PL) levels were measured as described (9). Separation of plasma lipoproteins by fast protein liquid chromatography (FPLC) was performed using two Superose 6 columns (Amersham Biosciences) as described (9). Separation of plasma lipoproteins by ultracentrifuge was performed using 1.5 ml of pooled plasma from 2-3 male mice, and the cholesterol, TG, and PL levels in each lipoprotein fraction were determined by biochemical assays (9,19).
Immunohistochemistry and Protein Blot Analysis-Immunohistochemical staining for macrophages in atherosclerotic lesions in the aortic root was performed using a MOMA-2 antibody (DAKO, Glostrup, Denmark) by the labeled streptavidin-biotin method as described previously (17). Immunohistochemical staining for T-cells and smooth muscle cells in atherosclerotic lesions was performed using an anti-CD3 antibody and an anti-␣-smooth muscle actin (DAKO) antibody, respectively. Quantitative analysis was measured as a percentage of the positive-stained area in the total atherosclerotic lesion area as described previously (17). Western blots and immunohistochemical analysis for EL in the vessel wall were performed using an anti-mouse EL antibody (generous gift from Dr. Daniel J. Rader, University of Pennsylvania, Philadelphia, PA).
Ex Vivo Adhesion Assay-Ex vivo adhesion assays were performed as described (20). Briefly, the murine thoracic aorta was isolated and fixed in an agarose-coated culture plate. Fluorescein-labeled cultured mouse monocyte/macrophage cells, WEHI78/24 or MH-S cell lines, were plated on the aortic strip, and the plate was gently shaken for 30 min. Some experiments were performed in the presence of heparin (1-5 units/ml) (Sigma). After washing the tissue to remove unbound cells, the number of binding cells per high power field, 10 arbitrary views per one tissue, was counted with fluorescence microscopy.
Statistical Analysis-Data are expressed as mean Ϯ S.E. An unpaired Student t test was used to detect significant differences when two groups were compared. One-way analysis of variance was used to compare the differences among three or four groups, with Bonferroni's test for post hoc analysis. Repeated measures analysis of variance was used to compare lipid fraction results obtained from FPLC. p values less than 0.05 were considered statistically significant.

Atherosclerosis Is Attenuated in Animals with Functional EL
Deletion-Male and female apoEϪ/Ϫ and LIPGϪ/Ϫ, apoEϪ/Ϫ mice were evaluated for development of atherosclerosis on normal chow and high fat dietary challenge. After 12 weeks on normal chow, at the age of 16 weeks, atherosclerotic lesion formation was evaluated by Sudan III staining of sections at the level of the aortic valve (Fig. 1). The extent of disease was quantified as the total lesion area determined by planimetry. Interestingly, the atherosclerotic lesions were markedly attenuated in double knockout mice when compared with apoEϪ/Ϫ mice. For animals on a normal chow diet, lesion size was significantly decreased by 71% in males (p Ͻ 0.001) and by 67% in females (p Ͻ 0.001) when double knockout mice were com-FIG. 1. Quantitation of atherosclerotic lesions in the aortic sinus of different genetic models. A, representative photographs of Sudan III-stained aortic root sections from apoEϪ/Ϫ and LIPGϪ/Ϫ, apoEϪ/Ϫ mice fed normal chow for 12 weeks. Sections were taken at the same level of the aortic valves (original magnification ϫ40). B, quantitative analysis of atherosclerotic lesion size in male and female apoEϪ/Ϫ and double knockout mice. The total lesion area of five sections in the aortic root from each mouse was quantified morphometrically as described under "Experimental Procedures." After 12 weeks on normal chow, the atherosclerotic lesion areas were significantly attenuated in LIPGϪ/Ϫ, apoEϪ/Ϫ when compared with apoEϪ/Ϫ mice (*, p Ͻ 0.001). C, representative photographs of Sudan III-stained aortic root sections from apoEϪ/Ϫ and LIPGϪ/Ϫ, apoEϪ/Ϫ mice fed a high fat diet for 12 weeks. Sections were taken at the same level of the aortic valves (original magnification ϫ40). D, quantitative analysis of atherosclerotic lesion size in male and female apoEϪ/Ϫ and double knockout mice. After 12 weeks on a high fat diet, the atherosclerotic lesion areas were significantly attenuated in male double knockout animals when compared with apoEϪ/Ϫ mice (*, p Ͻ 0.001). pared with apoEϪ/Ϫ mice (Fig. 1, A and B). Feeding mice with a high fat diet for 12 weeks accelerated atherosclerotic lesion formation in both apoEϪ/Ϫ and LIPGϪ/Ϫ, apoEϪ/Ϫ mice ( Fig.  1, C and D). As was the case with normal chow, formation of atherosclerosis was attenuated in male double knockout mice when compared with apoEϪ/Ϫ mice. Quantitative analysis showed a significant 24% decrease (p Ͻ 0.001) in the lesion area of male LIPGϪ/Ϫ, apoEϪ/Ϫ mice when compared with apoEϪ/Ϫ mice (Fig. 1D). However, the difference was smaller and not statistically significant between female apoEϪ/Ϫ and double knockout mice.

Serum Lipid Profiles Do Not Explain Attenuated Atherosclerosis in EL Knockout
Mice-To investigate the mechanism by which atherosclerosis was attenuated in the double knockout animals, lipid profiles were obtained from wild-type C57Bl/6J control (WT), single knockout apoEϪ/Ϫ male and female mice, LIPGϪ/Ϫ male and female mice, and double knockout LIPGϪ/Ϫ and apoEϪ/Ϫ male and female mice on normal chow and high fat diets. The previously described 2-fold increase in serum cholesterol of LIPGϪ/Ϫ animals over WT was confirmed in these studies for male and female animals on both diets (Table I) (9). The male LIPGϪ/Ϫ, apoEϪ/Ϫ mice on a high fat diet had a significantly higher cholesterol level than either single knockout alone, and female double knockout mice on high fat had higher but not significantly increased cholesterol levels when compared with either single knockout strain. There was no difference between cholesterol levels in the LIPGϪ/Ϫ, apoEϪ/Ϫ mice and the apoEϪ/Ϫ animals on regular diet.
TG levels were increased in female LIPGϪ/Ϫ animals on normal chow, as reported previously, and male animals also showed a significant increase in these experiments (Table I) (9). These differences disappeared on a high fat diet. Plasma TG levels were increased over WT in male and female apoEϪ/Ϫ animals on a regular diet, as expected. TG levels for the double knockout female mice on high fat were significantly higher than control and single knockout female mice but the same as seen on a normal chow diet. A different pattern was seen in male mice on the high fat diet. There was a decrease in serum TG in the apoEϪ/Ϫ animals, whereas TG levels in the LIPGϪ/Ϫ animals remained unchanged. On this atherogenic diet, there was no significant difference in TG levels between control, apoEϪ/Ϫ, and LIPGϪ/Ϫ male animals. Interestingly, the LIPGϪ/Ϫ, apoEϪ/Ϫ male animals on high fat continued to have significantly elevated TG levels.
Plasma PL levels tended to be higher in LIPGϪ/Ϫ animals. Both male and female LIPGϪ/Ϫ animals showed a signifi-cantly greater than 50% increase in plasma PL levels when compared with WT mice (Table I). LIPGϪ/Ϫ, apoEϪ/Ϫ animals showed modest but significant increases in PL levels when compared with LIPGϪ/Ϫ animals. These changes are likely due to the loss of the phospholipase activity of EL.
Fasting plasma from the different mouse lines was fractionated by gel filtration chromatography (FPLC) and cholesterol measured in individual fractions (Fig. 2, A-D). For both the WT and LIPGϪ/Ϫ lines, the majority of plasma cholesterol was found in the HDL fractions. The previously described 2-fold increase in cholesterol measured in HDL fractions was again observed for LIPGϪ/Ϫ animals when compared with WT animals on normal chow (Fig. 2, A and B) (9). This increase was less apparent for those animals on the high fat diet but was still significant in females. Also, significant increases in IDL/LDL cholesterol were found in LIPGϪ/Ϫ animals when compared with WT, for female mice on normal chow and both male and female mice on a high fat diet (Fig. 2, C and D). The cholesterol concentration in the HDL fractions was similar for the apoEϪ/Ϫ and LIPGϪ/Ϫ, apoEϪ/Ϫ animals, and in all cases, it was significantly lower than in the other two genotype groups. By FPLC analysis, only a minimal increase in HDL was seen in apoEϪ/Ϫ knockout mice also missing EL. By contrast, in the apoEϪ/Ϫ and LIPGϪ/Ϫ, apoEϪ/Ϫ animals, most of the cholesterol was distributed among larger lipoproteins in the VLDL and IDL/LDL range. Examining the VLDL fractions, male double knockout animals had significantly greater VLDL cholesterol than apoEϪ/Ϫ animals, on both chow and high fat diet, and female double knockout animals had greater mean VLDL cholesterol levels that approached significance.
To further characterize lipid metabolism in these genetic mouse models, lipoprotein fractions from pooled mouse plasma were separated by ultracentrifugation, and the cholesterol, TG, and PL content of each fraction was determined by biochemical assays. These studies also showed that LIPGϪ/Ϫ mice had 2-fold higher HDL-C than WT mice (Table II). HDL-C was found to be markedly decreased with loss of apoE, but this method found that EL deficiency partially restored the decreased HDL levels in apoEϪ/Ϫ mice. Also, a significant increase in LDL-C was observed in LIPGϪ/Ϫ when compared with WT. A prominent increase in VLDL/LDL triglyceride was observed in LIPGϪ/Ϫ, apoEϪ/Ϫ animals when compared with apoEϪ/Ϫ animals, consistent with the FPLC data showing that loss of EL was associated with increased TG-rich lipoproteins. HDL phospholipid was markedly increased in both the WT and apoEϪ/Ϫ mice missing EL. A smaller but significant increase in LDL phospholipid was observed in LIPGϪ/Ϫ animals when

TABLE II Lipid and lipoprotein profile in apoEϪ/Ϫ, LIPGϪ/Ϫ, LIPGϪ/Ϫ, apoEϪ/Ϫ, and wild-type mice on normal chow
The HDL, LDL, and VLDL fractions were obtained from 1.5 ml pooled plasma from 2-3 male mice by ultracentrifugation. The cholesterol (chol), TG, and PL levels in each lipoprotein fraction were determined by biochemical assay. A total of 16 -23 mice were included in each group. Data are shown as mean Ϯ S.E. (mg/dl). The abbreviations are as follows: WT, wild-type C57Bl/6; NS, not significant. WT  compared with WT, and there was a nonsignificant increase in LDL phospholipid in LIPGϪ/Ϫ, apoEϪ/Ϫ animals when compared with apoEϪ/Ϫ animals. These data suggest that HDL phospholipid is the most specific substrate for EL, and as a result of EL action, HDL catabolism produces a decrease in HDL-C levels. Moreover, EL may have weak enzymatic action on apoB-containing lipoproteins such as LDL.

EL Expression in the Vascular Wall May Mediate Local Effects on Atherosclerosis-Since the lipoprotein profile in
LIPGϪ/Ϫ, apoEϪ/Ϫ animals did not provide a clear mechanism for the observed decrease in atherosclerosis and because other lipases have been shown to have local effects in the vessel wall, we investigated vascular wall differences in the single and double knockout apoE mice. First, to determine whether EL was expressed in the apoE model atherosclerotic lesions, immunohistochemistry was performed using an antibody against mouse EL. In large vessels of WT mice, EL was predominantly expressed by luminal endothelial cells, although expression was also detected in the adventitia (Fig. 3, a and b).
In apoEϪ/Ϫ mice, expression of EL was detected in the atherosclerotic plaque as well as endothelial cells (Fig. 3, d and e). Most of the EL-positive area was rich in macrophages, suggesting that EL-positive cells in the plaque were mainly infiltrating macrophages (Fig. 3, g and h). Smooth muscle ␣-actin staining was barely detectable in the EL-positive area of the atheroma (data not shown). Specificity of the staining for EL was confirmed by immunohistological analysis of atherosclerotic lesions from LIPGϪ/Ϫ, apoEϪ/Ϫ mice. These studies suggested reduced macrophage infiltration when compared with apoEϪ/Ϫ animals (Fig. 3, i and j).
To compare EL protein levels in normal versus atherosclerotic vessels, homogenates of aortic tissue from WT and apoEϪ/Ϫ animals on normal chow or high fat diet were analyzed. Tissue samples were subjected to Western blotting using the same antibody as employed for immunohistochemistry (Fig. 4). This quantitative assay revealed that EL expression was increased by 2.1-fold in aortas from apoEϪ/Ϫ mice when compared with aortas from WT mice. Feeding mice with a high fat diet resulted in a modest (ϳ10%) increase in EL expression, in both the WT and the apoEϪ/Ϫ mice.
These results in the apoEϪ/Ϫ mouse model of atherosclerosis are consistent with a previous study that documented EL expression in infiltrating macrophages and smooth muscle cells in advanced lesions of atherosclerosis in postmortem human coronary arteries (21). Thus, the degree and extent of EL expression appears to be increased early in human atherosclerotic vascular disease.
To examine whether lesions in the apoEϪ/Ϫ and LIPGϪ/Ϫ, apoEϪ/Ϫ mice had different cellular composition, we performed immunohistochemical staining of atherosclerotic areas with antibodies chosen to specifically label different cell types. An antibody to MOMA-2 was employed to evaluate the area containing infiltrating macrophages. The percentage of the  versus g and h). Atherosclerotic lesions of LIPGϪ/Ϫ, apoEϪ/Ϫ mice were negative for EL expression (i) but positive for MOMA-2 (j). Panels c and f were negative controls in which the primary antibodies were replaced with nonspecific rabbit or rat immunoglobulins. Original magnifications were as follows: ϫ400 (a, c, d, f, g, i, and j) and ϫ1000 (b, e, and h). MOMA-2-stained area in the aortic sinus lesions was significantly smaller in LIPGϪ/Ϫ, apoEϪ/Ϫ mice, consistent with overall fewer macrophages in these lesions (Fig. 5). There was no significant difference in T-cell or smooth muscle cell composition of the lesions in the apoEϪ/Ϫ and the LIPGϪ/Ϫ, apoEϪ/Ϫ mice, as evaluated with immunohistochemistry with antibodies to CD3 and smooth muscle ␣-actin respectively (data not shown).
To investigate the mechanism for the observed difference in macrophage composition of lesions between EL-expressing and non-EL-expressing apoE mice, aortic strips were isolated from LIPGϪ/Ϫ, apoEϪ/Ϫ and apoEϪ/Ϫ mice and incubated with cultured mouse monocyte/macrophage cells (the WEHI78/24 cell line). An ex vivo adhesion assay revealed that monocyte/ macrophage binding to LIPGϪ/Ϫ, apoEϪ/Ϫ aortic strips was significantly less than to strips from apoEϪ/Ϫ mice (Fig. 6, A  and B). Similar results were obtained when MH-S monocyte cells were used (data not shown). Heparin treatment significantly diminished the augmented cell adhesion to aortic strips from apoEϪ/Ϫ mice but not to those from LIPGϪ/Ϫ, apoEϪ/Ϫ mice, suggesting that up-regulation of EL in the aorta may in part mediate the cell adhesion through heparan sulfate proteoglycans (Fig. 6B). These findings support the measured decrease in macrophage composition of EL-negative lesions and provide a possible mechanism by which EL might act directly to impact the development of vascular wall disease. Whether EL might modulate monocyte adherence and transit into the vessel wall through local lipid metabolism or through a direct bridging function will require further investigation.

DISCUSSION
To initiate studies investigating EL involvement in vascular disease, the previously described congenic strain of mice carrying a 129SV/J-null LIPG allele in the C57Bl/6 background was bred to a strain homozygous for null apoE alleles in the C57Bl/6 background (9,16,22,23). This model has been extensively characterized, both in terms of the altered lipoprotein and lipid profiles and in terms of the predisposition to atherosclerosis. Homozygous apoE knockout animals have a markedly elevated serum cholesterol with increased VLDL and IDL/ LDL cholesterol and reduced HDL-C. As a result of this highly atherogenic lipoprotein profile, apoEϪ/Ϫ animals develop spontaneously complex atherosclerotic lesions by 15 weeks of age that are similar to those in humans. The atherogenic potential of a broad range of lipid metabolism and inflammatory mediator genes has been investigated by evaluation of disease development in the apoEϪ/Ϫ model (24).
It was anticipated that LIPGϪ/Ϫ, apoEϪ/Ϫ animals might show increased HDL-C levels and a related decrease in the rate and extent of atherosclerotic disease when compared with the apoEϪ/Ϫ animals. Interestingly, the LIPGϪ/Ϫ, apoEϪ/Ϫ mice showed only a modest increase in HDL-C levels. There was also an increase in VLDL cholesterol in LIPGϪ/Ϫ, apoEϪ/Ϫ animals, and this increase was significant for the males. The increase in serum TG level was consistent with the increase in VLDL cholesterol in LIPGϪ/Ϫ, apoEϪ/Ϫ animals. The mechanism for this increase in VLDL is unclear, but the observation is consistent with the recent study that demonstrated a role for EL in the metabolism of apoB-containing lipoprotein particles such as VLDL and LDL (25). In this case, loss of EL in addition to apoE would result in further prolonged half-life of VLDL, accounting for the increases noted in the double knockout animals. A previous study has indicated that EL has some weak triglyceride lipase activity, particularly when lipoprotein substrates are abundant, as found in apoEϪ/Ϫ mice on a high fat diet (26). It is possible that EL triglyceride lipase activity accounted for part of the difference in TG level. Consistent with this hypothesis is the observation that LIPGϪ/Ϫ animals on a high fat diet showed increased LDL/IDL levels when compared with WT animals. Although loss of EL does not impair the metabolism of apoB-containing particles in the setting of a normal diet, challenge with high fat provides evidence for a decreased ability to clear LDL/IDL particles.
The lipid profile of the LIPGϪ/Ϫ, apoEϪ/Ϫ animals showed an increase in antiatherogenic HDL and an increase in atherogenic VLDL/LDL. This balanced increase in lipoprotein particle makeup was expected to be neutral with respect to the degree of atherosclerosis. It was thus surprising to find a 70% decrease in the vascular lesion area in the LIPGϪ/Ϫ, apoEϪ/Ϫ animals when compared with the apoEϪ/Ϫanimals. The observed effect is likely mediated in part by the observed modest increase in antiatherogenic HDL particle number. There was no evidence for other HDL-related mechanisms for the decrease in atherosclerosis. Other groups have shown changes in HDL particle size with EL function (10,27). In studies reported here, changes in HDL size were not apparent from FPLC curves. Also, the reported changes in particle size in EL-null animals are consistent with increased numbers of phospholipid-rich particles, which are known to have proatherogenic effects, and could not simply account for the observed histologic differences in atherosclerosis (28). A recent study showed that EL alters the composition of HDL to decrease binding capacity to scavenger receptor class B, type I and modulate scavenger receptor class B, type I-dependent cholesterol efflux (29). Overall, available data suggest that atherosclerosis is modulated by EL in the apoE knockout model and that only a portion of this effect is likely explained by the observed increase in HDL-C.
As is the case with other lipases, EL may regulate atherosclerosis through functions in the vessel wall, including lipoprotein uptake and monocyte trafficking. EL has been shown to regulate cellular lipoprotein uptake. Several studies have indicated that EL mediates binding and uptake of plasma lipoproteins via a process that is independent of its enzymatic activity, requires cellular heparan sulfate proteoglycans, and is regulated by ligand clustering (14,15). As a first step to investigate possible direct vascular wall functions of EL, we have documented increased EL expression in the diseased aorta of apoE knockout animals, and validated this observation in human atherosclerosis (21). Through studies investigating the cell-specific expression of EL in atherosclerotic plaque, we have shown that infiltrating macrophages are primarily responsible for this increased expression. Interestingly, these experiments also identified a decreased relative macrophage content in the combined EL-and apoE-negative lesions. Follow-up studies have documented a related decrease in monocyte adhesion to vascular wall tissue missing EL. Increased monocyte adhesion was documented in vitro and ex vivo using EL-overexpressing COS7 cells and EL-overexpressing vessels, and the EL-mediated adhesions were inhibited by heparin and heparinase I, suggesting that up-regulation of EL may enhance monocyte adhesion to the vessel wall by interaction with heparan sulfate proteoglycans (30). Taken together, these data strongly suggest a primary role for EL in mediating monocyte adhesion and uptake into the vessel wall in the setting of atherosclerotic disease.
Comparable studies performed with other lipases have provided results similar to those reported here for EL. Mice lacking both hepatic lipase and apoE had increased total plasma cholesterol, with the increase in cholesterol mainly due to VLDL and only a very modest increase in HDL cholesterol (28,31). Despite this increase in plasma cholesterol, hepatic lipase deficiency in apoEϪ/Ϫ mice resulted in significantly decreased aortic plaque sizes in female mice fed normal chow. The improved capacity to promote cholesterol efflux, together with the modest increase in HDL, were thought to overcome the increase in atherogenic lipoproteins. Such observations, in conjunction with those reported here for EL, argue that this family of proteins can modulate atherosclerosis through local vascular wall disease-related processes as well as regulation of circulating lipid levels.