Carboxyl Ester Lipase Expression in Macrophages Increases Cholesteryl Ester Accumulation and Promotes Atherosclerosis*

Carboxyl ester lipase (CEL, also called cholesterol esterase or bile salt-dependent lipase) is a lipolytic enzyme capable of hydrolyzing cholesteryl esters, triacylglycerols, and phospholipids in a trihydroxy bile salt-dependent manner but hydrolyzes ceramides and lysophospholipids via bile salt-independent mechanisms. Although CEL is synthesized predominantly in the pancreas, a low level of CEL expression was reported in human macrophages. This study used transgenic mice with macrophage CEL expression at levels comparable with that observed in human macrophages to explore the functional role and physiological significance of macrophage CEL expression. Peritoneal macrophages from CEL transgenic mice displayed a 4-fold increase in [3H]oleate incorporation into cholesteryl [3H]oleate compared with CEL-negative macrophages when the cells were incubated under basal conditions in vitro. When challenged with acetylated low density lipoprotein, cholesteryl ester accumulation was 2.5-fold higher in macrophages expressing the CEL transgene. The differences in cholesteryl ester accumulation were attributed to the lower levels of ceramide and lysophosphatidylcholine in CEL-expressing cells than in CEL-negative cells. CEL transgenic mice bred to an atherosclerosis susceptible apoE-/- background displayed an approximate 4-fold higher atherosclerotic lesion area than apoE-/- mice without the CEL transgene when both were fed a high fat/cholesterol diet. Plasma level of the atherogenic lysophosphatidylcholine was lower in the CEL transgenic mice, but plasma cholesterol level and lipoprotein profile were similar between the two groups. These studies documented that CEL expression in macrophages is pro-atherogenic and that the mechanism is because of its hydrolysis of ceramide and lysophosphatidylcholine in promoting cholesterol esterification and decreasing cholesterol efflux.

dase, capable of releasing fatty acyl groups from amide linkages such as those present in ceramides (3,4). The conformation of CEL stipulates that its hydrolytic activity on cholesteryl esters, triacylglycerols, and phospholipids is dependent on the presence of bile salt (1,2,5,6). However, CEL hydrolysis of lysophosphatidylcholine and ceramide does not have an absolute requirement for bile salt (2,3).
The CEL is synthesized predominantly in the pancreas, and is also found to be present in abundance in human milk. Accordingly, early investigations on its physiological functions have focused primarily in the digestive tract, where CEL is postulated to play a role in lipid nutrient absorption. Our recently completed study revealed that CEL plays a significant role in catalyzing cholesteryl ester absorption from the intestinal lumen (7) and promoting large chylomicron assembly and secretion from enterocytes in the digestive tract (8). Interestingly, CEL-mediated cholesteryl ester absorption is dependent on its cholesteryl ester hydrolytic activity, whereas its modulation of large chylomicron production is unrelated to its cholesteryl ester hydrolytic activity but dependent on its hydrolysis of ceramide generated during the lipid absorption process (8).
In addition to its presence in the digestive tract and its participation in the intestinal lipid absorption and transport pathway, CEL is also found to be present in human plasma and in the vessel wall (9). We have shown previously that CEL is synthesized by human macrophages and endothelial cells (10,11). The presence of CEL in the vasculature, and its increased synthesis by vascular cells after incubation with oxidized LDL (10,11), suggested a potential role of CEL in modulation of atherosclerosis. Cell culture studies have shown that CEL induces vascular smooth muscle cell proliferation (12), suggesting its contributory role toward atherosclerosis. In contrast, other in vitro studies showed that CEL reduces lysophospholipid content in oxidized LDL (9). The latter studies suggested that CEL expression in the vessel wall may be protective against atherosclerosis. Whether the presence of CEL in the vessel wall is protective or contributory to the atherosclerotic lesion development has not been clarified.
The paucity of information regarding the physiological role of the vascular CEL in atherosclerosis is because of species-specific differences in macrophage CEL expression. Unlike the human macrophages, mouse macrophages do not synthesize CEL (10). To circumvent this problem, we have produced transgenic mice expressing CEL in the vessel wall to evaluate its role in atherosclerosis. Results of this study showed that CEL expression in macrophages promotes cholesteryl ester synthesis and accumulation in response to modified LDL and increases atherosclerosis lesions in apoE-null mice.

EXPERIMENTAL PROCEDURES
Production and Screening of Transgenic Mice-The sheep visna virus long terminal repeat (LTR) was used as promoter to drive macrophage * This work was supported by National Institutes of Health Grant HL66246. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  CEL expression in transgenic mice. The LTR-apoE gene construct containing the visna virus LTR upstream of the human apoE3 gene was obtained from Dr. John Taylor (Gladstone Institute, San Francisco, CA) (13). A 2-kb cDNA fragment containing the entire coding sequence of rat CEL without the polyadenylation signal was obtained by SmaI digestion of the plasmid pSVL-CEL (14) and inserted into the unique BalI site located in the beginning of exon 2 of apoE to replace the coding exons and subsequent introns in the LTR-apoE chimeric gene. The resulting chimeric construct contains from the 5Ј to 3Ј direction: the visna virus LTR, the noncoding exon 1 sequence and intron 1 sequence of the human apoE gene, and the entire coding sequence of rat CEL. A schematic depicting the approach for construction of the LTR-CEL chimeric gene is presented in Fig. 1. The sequence of the chimeric gene was confirmed by dideoxy chain termination sequence analysis. The sequence predicts that the entire native rat CEL protein, with no additional residue from apoE, will be produced from the LTR promoter. The LTR-CEL chimeric gene was separated from vector sequence by digestion with ScaI and SalI, and then purified by agarose gel electrophoresis. Four thousand copies of the chimeric gene were microinjected into the pronuclei of fertilized mouse eggs obtained from superovulated female C57BL/6 mice. The injected eggs were surgically transferred to oviducts of surrogate females. Offspring were screened for integration of the transgene by PCR amplification of tail DNA with the upstream primer sequence: 5Ј-GTGGTGACTCTGTTGACATCTTC-3Ј and the downstream primer sequence: 5Ј-TGGCTTGTAGACATCGTTTCTTG-3Ј to yield a 131-bp product that overlaps the nucleotide sequence between residues 148 and 279 of rat CEL cDNA. Transgenic founder lines were maintained and propagated. The expression profile of the CEL transgene in the transgenic mice was determined by reverse transcriptase-PCR amplification of RNA isolated from various tissues using the above CEL primers, with GAPDH primers (forward: 5Ј-CATCAAGAAGGT-GGTGAAGC-3Ј and reverse: 5ЈGAGCTTGACAAAGTTGTC-3Ј) for amplification of GAPDH mRNA as control. Peritoneal macrophages were obtained from the progeny and used to determine CEL expression. Two lines were shown to display macrophage CEL expression levels similar to that observed in human monocyte-derived macrophages (10). Cholesteryl Ester Hydrolytic Activity in Mouse Macrophages-Wild type C57BL/6 and CEL transgenic mice were injected intraperitoneally with a 4% thioglycolate solution, and macrophages were harvested 3 days later by lavage of the peritoneal cavity with phosphate-buffered saline. The mouse peritoneal macrophages were cultured in 6-well plates at a density of 2 ϫ 10 6 cells/well with Dulbecco's modified Eagle's medium containing 10% fetal bovine serum. Nonadherent cells were removed after an overnight culture and the adhering cells were washed with phosphate-buffered saline. Conditioned media from the macrophages after 1 day of culture in the absence of serum were collected and concentrated to 300 l by centrifugation through an Ultrafree-4 centrifugal filter. An aliquot of each sample, usually 140 l, was used to measure cholesterol esterase activity, based on the hydrolysis of cholesteryl [ 14 C]oleate in the presence of 33 mM sodium cholate or sodium deoxycholate (15).
Lipoprotein Degradation and Lipid Accumulation by Macrophages-Human LDL was isolated from fresh plasma by preparative ultracentrifugal flotation in KBr solutions between the densities of 1.02 and 1.063 g/ml. Acetylated LDL was prepared by acetylation of LDL with acetic anhydride and radiolabeled with 125 I using the iodine monochloride method as described (16 (17).
Cholesterol esterification was assessed by incubating macrophages in serum-free medium containing 200 g/ml acLDL and 0.2 mM [ 3 H]oleate-albumin complex for 3 or 48 h at 37°C (16). The cells were then washed, and lipids were extracted for thin layer chromatography separation in petroleum ether:ethyl ether:acetic acid (300:60:1, v/v/v). The cholesteryl ester spots were scraped from the plates and subjected to liquid scintillation counting to determine the amount of [ 3 H]oleate incorporated into cholesteryl [ 3 H]oleate. In parallel experiments, lipids were extracted from macrophages after their incubation with unlabeled acLDL for 48 h. The mass of free cholesterol and cholesteryl ester was analyzed by gas chromatography using stigmastanol as the internal standard (16). Ceramide content in macrophages after incubation with acLDL was determined based on the amount of 32 P from [␥-32 P]ATP incorporated into ceramide in the presence of diacylglycerol kinase (18).
Cholesterol Efflux from Mouse Peritoneal Macrophages-Mouse peritoneal macrophages were obtained after 3 days of thioglycolate activation, washed with phosphate-buffered saline, and plated in 12-well dishes at a density of 2 ϫ 10 6 cells/ml. Non-adhering cells were removed after 2 h by exhaustive washing and macrophages attached to the dishes were then incubated with 200 g/ml [ 3 H]cholesteryl oleate-containing acLDL overnight at 37°C. The cells were washed twice with phosphatebuffered saline and then equilibrated in serum-free media for 6 h. One set of wells was used to measure the total amount of [ 3 H]cholesterol loaded into the macrophages. Equivalent wells were incubated with or without apoA-I in serum-free media for 6 h to induce cholesterol efflux. The amount of [ 3 H]cholesterol transferred from the cells to apoA-I in the incubation media was quantified by determining the radioactivity level in the media. Each treatment was performed in triplicates and the data are presented as percentage of total radiolabel loaded into the macrophages.
Effects of CEL Transgenic Expression on Plasma Lipids-The impact of macrophage CEL expression on plasma lipid levels and atherosclerosis was evaluated by cross-breeding the CEL-transgenic mice with the atherosclerosis-susceptible apoE-null mice. Female apoE Ϫ/Ϫ mice with or without macrophage-specific CEL transgenic expression were fed the Western-type high fat/cholesterol diet (TD88137, Harlan Teklad), which contains 21% fat, 0.15% cholesterol, and 19.5% casein by weight with no sodium cholate (19). After 8 weeks, fasting plasma was collected from the animals. Cholesterol distribution among the various classes of lipoproteins was analyzed by subjecting 150 l of plasma to fast phase liquid chromatography gel filtration on two Superose 6 HR columns in tandem (20), as described previously (21). Each fraction (0.5 ml) was collected for cholesterol determination by colorimetric assay (Wako Chemicals, Richmond, VA).
Total lysophosphatidylcholine concentration in plasma was assayed enzymatically as described previously by Kishimoto et al. (22). Briefly, 8 l of each plasma sample was preincubated for 5 min in 240 l of reagent A (0.1 M Tris-HCl, pH 8.0, 0.01% Triton X-100, 1 mM CaCl 2 , 3 mM N-ethyl-N-(2-hydroxy-3-sulfopropyl)-3-methylaniline sodium dihydrate; 10 kilounits/liter peroxidase, 0.1 kilounits/liter glycerophosphorylcholine phosphodiesterase, 1 kilounits/liter choline oxidase) in 96-well plates. The initial absorbance at 600 nm was recorded prior to starting the reaction with the addition of 80 l of reagent B (0.1 M Tris-HCl, pH 8.0, 0.01% Triton X-100, 5 mM 4-aminoantipyrine, and 1 kilounits/liter phospholipase B). The incubation was continued for 30 min at 37°C and the final absorbance at 600 nm was recorded. Total LPC concentration in each sample was calculated by comparing the change in absorbance between the standard and samples after subtraction of the background absorbance recorded prior to initiation of the reaction. The accuracy of this method was confirmed by comparing concentrations determined by this enzymatic assay with high performance liquid chromatography analysis of the pooled sample.
Atherosclerosis Studies-Female apoE Ϫ/Ϫ mice with or without the CEL transgene were maintained on the Western-type high fat/cholesterol diet for 8 weeks and then sacrificed. The heart and aorta were perfused with phosphate-buffered saline and then formalin fixed with a 4% neutral formalin solution. Fat tissues were carefully trimmed. The upper heart section was incubated in phosphate-buffered saline containing 30% sucrose for 18 h, embedded in OCT medium, frozen, and sectioned with a cryostat at 10-m intervals throughout the aortic sinus and the aortic arch as described (19,23). The sections were stained with oil red O and lesions were quantified according to the procedure of Paigen et al. (24).

RESULTS
Human monocyte-derived macrophages but not mouse macrophages express low level of CEL, detectable by reverse transcriptase-PCR amplification of cellular RNA and trihydroxy bile salt-specific enhancement of cholesteryl ester hydrolytic activity in the conditioned media (10). The physiological role of the macrophage-derived CEL is unknown. Therefore, this study generated transgenic mouse lines with macrophage expression of a rat CEL cDNA to address this issue. The visna virus LTR, which has previously been shown to promote low levels of human apoE expression in transgenic mouse macrophages (13), was used to direct CEL expression in macrophages. The insertion of the rat CEL cDNA into the 5Ј-untranslated region located in exon 2 of the human apoE disrupts apoE coding and effectively replaces expression of the apoE gene with a chimeric gene that encodes the rat CEL protein (Fig. 1A). Polymerase chain termination reaction amplification of tail DNA identified CEL transgenic mice (Fig. 1B), and reverse tran-scriptase-PCR amplification of RNA obtained from various tissues of these transgenic animals showed CEL expression in brain, heart, kidney, lung, and peritoneal macrophages (Fig. 1C). The expression profile of the CEL transgene is consistent with visna virus LTR-directed transgenic expression reported previously (25).
Peritoneal macrophages isolated from CEL transgenic mice showed elevated levels of cholate-induced cholesteryl [ 3 H]oleate hydrolytic activity in comparison with macrophages isolated from their nontransgenic littermates ( Fig. 2A). When the assays were performed with deoxycholate instead of sodium cholate, only low levels of cholesteryl ester hydrolytic activity were detected in macrophages from either control or CEL transgenic mice (Fig. 2B). Importantly, the increase in cholesteryl ester hydrolytic activity in the CEL transgenic macrophages was not observed when the assays were performed with deoxycholate instead of cholate (Fig. 2). The specific induction of cholesteryl ester hydrolytic activity with the trihydroxy bile salt cholate but not the dihydroxy bile salt deoxycholate confirmed the expression of the pancreatic type bile salt-stimulated CEL in macrophages of the transgenic mice. A transgenic mouse line in which their peritoneal macrophages displayed similar levels of cholate-specific cholesteryl ester hydrolytic activity as that observed previously in human macrophages (10) was selected for expansion and additional experimentation.
Initial experiments to examine the impact of CEL expression on macrophage lipid metabolism compared cholesterol esterification activity between wild type and CEL transgenic mice. As expected, significantly higher levels of [ 3 H]oleate incorporation into cholesteryl [ 3 H]oleate was observed with both wild type and CEL transgenic mouse macrophages when the cells were incubated for 3 h with acLDL in comparison to cells incubated without added lipoproteins. However, no difference in cholesterol esterification activity was observed between wild type and CEL transgenic macrophages under these experimental conditions, with both types of cells displaying similar levels of cholesteryl [ 3 H]oleate formation after a 3-h incubation period (Fig. 3, A and  B). Interestingly, when the cells were cultured with [ 3 H]oleate in basal media for 48 h, an approximate 4-fold higher level of cholesteryl [ 3 H]oleate was found to be accumulated in CEL transgenic macrophages than in wild type macrophages (Fig. 3C). In contrast, the level of cholesteryl [ 3 H]oleate accumulated in CEL transgenic macrophages after 48 h incubation with acLDL was only marginally higher than that observed in macrophages from nontransgenic wild type littermates (Fig.  3D). Importantly, when the total amount of cholesteryl ester accumulation was measured, incubation with 200 g/ml of acLDL for 48 h  resulted in a 2.5-fold higher cholesteryl ester mass in the CEL transgenic macrophages compared with that observed in the wild type macrophages (Fig. 4).
The difference in cholesteryl ester accumulation between wild type and CEL transgenic macrophages was not related to uptake and degradation of acLDL because incubation with various concentrations of 125 I-acLDL at 37°C revealed no difference in the amount of 125 I-acLDL internalized and degraded by macrophages isolated from wild type and CEL transgenic mice (data not shown). Additionally, the ACAT-specific inhibitor PD138142-0000 (Parke-Davis, Ann Arbor, MI), which has no effect on CEL activity in vitro (data not shown), was capable of inhibiting cholesteryl ester accumulation in both wild type and CEL transgenic macrophages in a concentration-dependent manner (Fig. 5). The latter observation indicated that ACAT was primarily responsible for cholesterol esterification in both wild type and CEL transgenic macrophages, and CEL expressed in macrophages does not participate directly in intracellular cholesterol esterification.
We also considered the possibility that CEL transgenic macrophages accumulated more cholesteryl esters than wild type cells in response to acLDL challenge because of their failure to hydrolyze cholesteryl esters associated with the incoming acLDL. To test this possibility, wild type and CEL transgenic macrophages were incubated with [ 3 h]cholesteryl oleate-containing acLDL. The amounts of non-esterified and esterified [ 3 H]cholesterol accumulated in the cells were determined after 30 min. Results showed the presence of higher levels of both non-esterified and esterified [ 3 H]cholesterol in CEL transgenic macrophages than in wild type macrophages (Fig. 6). The increased levels of both free and esterified [ 3 H]cholesterol in CEL transgenic macrophages, despite similar levels of acLDL uptake and degradation, suggested a more active futile cycle for cholesterol esterification and de-esterification in macrophages expressing CEL. These observations also implied reduced cholesterol efflux capability in CEL transgenic macrophages. This possibility was examined directly by pre-loading the cells with [ 3 H]cholesteryl oleatecontaining acLDL and then measured cholesterol efflux to apoA-I added to the incubation media. The results clearly demonstrated that significantly less [ 3 H]cholesterol was transferred to the extracellular acceptor from CEL transgenic macrophages than from wild type cells (Fig. 7).
One mechanism by which CEL expression may influence ACATmediated cholesterol esterification and apoA-I-mediated cholesterol efflux is via its ability to hydrolyze lysophospholipids. Previous studies have shown that LPC reduces the cholesterol level in the endoplasmic reticulum (26), thus reducing substrate availability for ACAT esterification. Accordingly, we compared LPC content in wild type and CEL transgenic macrophages after a 48-h incubation with acLDL. Results showed an approximate 40% reduction of LPC level in CEL transgenic macrophages compared with that observed in wild type macrophages (Fig. 8). These results are consistent with the interpretation that CEL expressed in macrophages is capable of LPC hydrolysis.
Carboxyl ester lipase is also a lipoamidase, capable of hydrolyzing ceramides even under low or no bile salt conditions (3,4). In view of previous reports showing that intracellular sphingolipid content can influence cholesterol esterification activity and cholesteryl ester accumulation (27), particularly ceramide inhibition of ACAT activity and promotion of cholesterol efflux has been noted in cultured cells (28,29), we ascertained the possibility that increased cholesteryl ester accumulation in the macrophages of CEL transgenic mice may be related to   alteration in ceramide concentrations in the cells. Accordingly, extracts were prepared from wild type and CEL transgenic macrophages after their incubation with 200 g/ml acLDL for 48 h at 37°C. Results revealed that acLDL-induced ceramide accumulation was ϳ2-fold lower in the CEL transgenic macrophages compared with that in wild type macrophages (Fig. 9).
The results showing increased cholesteryl ester accumulation in CEL transgenic macrophages compared with cells from nontransgenic wild type mice suggested that macrophage expression of CEL may promote atherosclerosis. To test this hypothesis, we cross-bred the CEL transgenic mice with the atherosclerosis-susceptible apoE Ϫ/Ϫ mice. The progenies were fed a Western-type high fat/cholesterol diet (no cholate added) to facilitate atherogenesis (19). Although apoE Ϫ/Ϫ mice with or without macrophage CEL transgenic expression displayed similar plasma cholesterol level and lipoprotein profile (Fig. 10), plasma LPC concentration was 3-fold lower in apoE Ϫ/Ϫ mice expressing the CEL transgene than in the nontransgenic apoE Ϫ/Ϫ mice (Fig. 11). Importantly, a significant increase of atherosclerosis lesions was observed in apoE Ϫ/Ϫ mice with CEL expression in macrophages compared with apoE Ϫ/Ϫ mice without the CEL transgene (Fig. 12).

DISCUSSION
Carboxyl ester lipase is an avid cholesteryl ester hydrolase. One of its functions in the gastrointestinal tract is to catalyze cholesteryl ester hydrolysis to facilitate its absorption (7,30). This enzyme is also synthesized in human macrophages but not in mouse macrophages (10). The current study used macrophages from CEL transgenic mice to show that CEL expression in macrophages actually increased instead of decreased acLDL-induced cholesteryl ester accumulation, implying that the role of CEL in macrophages is probably not related to its cholesteryl ester hydrolytic activity. This hypothesis is consistent with previous reports showing that CEL cDNA did not promote intracellular cholesteryl ester hydrolysis when transfected into hepatoma cells (31). The inability of the CEL transgene to decrease cholesteryl ester accumulation in macrophages is most likely because of the low level or absence of trihydroxy bile salts in the cytosolic compartment where cholesteryl ester droplets accumulate in macrophages. Thus, in this environment, CEL cannot function as a cholesteryl ester hydrolase (1, 2, 5, 6) and the cells continue to accumulate cholesteryl esters in response to acLDL.
The data of the current study revealed that macrophages expressing the CEL transgene actually displayed increased cholesterol esterification and accumulated more cholesteryl esters in response to acLDL in    comparison to macrophages lacking CEL expression. These observations suggested that CEL participates either directly as a cholesterol esterification enzyme in macrophages or indirectly via substrate delivery to an ACAT-sensitive pool. The possibility that CEL may serve as a cholesterol esterification enzyme was suggested previously from in vitro studies demonstrating that the incubation of CEL with cholesterol, cholesteryl esters, and fatty acids favors the de-esterification reaction in the presence of trihydroxy bile salt, but the reaction equilibrium favors the cholesterol esterification direction in the absence of trihydroxy bile salt (32). However, in the current study, we showed that cholesterol accumulation in CEL transgenic macrophages was inhibited by specific ACAT inhibitors. Thus, the increased cholesteryl ester accumulation in these cells is unlikely because of the cholesterol esterification activity of CEL and probably an indirect effect because of other enzymatic activities of CEL that may influence cholesterol delivery to the ACAT-sensitive pool.
Previous studies have shown that preincubation of macrophages with ceramide decreased whole cell cholesterol esterification activity by ϳ50% (27). The mechanism appears to be related to ceramide induction of endogenous sphingomyelin synthesis (27) and/or ceramide trapping of cholesterol in endosomal compartments (33), both of which would affect intracellular cholesterol trafficking and limit cholesterol accessi-bility for esterification by ACAT (27). Ceramide is also capable of increasing cholesterol efflux from cholesterol-loaded cells via stabilization of ABCA1 on the plasma membrane (29). In the current study, we showed that CEL expression decreased intracellular ceramide concentration with a concomitant increase in cholesteryl ester accumulation. Accordingly, CEL may be promoting cholesteryl ester accumulation via increased cholesterol translocation to the ACAT-accessible pool for esterification and/or reducing ABCA1-mediated cholesterol efflux from cells. The ability of CEL to promote cholesterol delivery to the ACAT-accessible pool may also be related to its ability to hydrolyze LPC, thereby decreasing its inhibition of cholesterol distribution to the endoplasmic reticulum (26) and promoting cholesterol efflux.
It is interesting to note that the difference between wild type and CEL transgenic macrophages in cholesteryl ester accumulation was less obvious when the assays were performed by measuring [ 3 H]oleate incorporation into cholesteryl [ 3 H]oleate in the presence of acLDL. However, the total amount of acLDL-induced cholesteryl ester accumulation remained notably different between wild type and CEL transgenic macrophages. These latter observations indicated that under cholesterol-loaded conditions, CEL may facilitate overall intracellular lipid turnover by supplying increasing amounts of endogenously derived fatty acids to compete with the exogenously supplied [ 3 H]oleate as substrates for cholesterol esterification. This interpretation is consistent with previous experiments demonstrating that CEL cDNA transfection into Chinese hamster ovary cells increased intracellular phospholipid turnover rates (34). In view of previous reports documenting a direct effect of ceramide in inhibiting phospholipid biosynthesis (35)(36)(37), the mechanism by which CEL increases phospholipid turnover may also be related to the reduction of steady state levels of ceramide in acLDLtreated macrophages. Taken together, these observations suggested that CEL may function as a lipoamidase and a lysophospholipase to promote intracellular lipid trafficking between different membrane compartments.
Results of the current study also revealed a pathophysiological role of CEL in macrophages. Its reduction of ceramide and LPC levels in macrophages that led to increased cholesteryl ester accumulation in response to atherogenic lipoproteins also resulted in increased athero-  . Plasma lysophosphatidylcholine concentrations in mice. Plasma was obtained from apoE Ϫ/Ϫ (apoE null) or apoE Ϫ/Ϫ mice with the CEL transgene (apoE null; CEL tg). Total lysophosphatidylcholine concentration in plasma was measured enzymatically. * denotes significance difference from the group without the CEL transgene at p Ͻ 0.05 (n ϭ 4). sclerosis lesion size in vivo. Interestingly, macrophage expression of CEL in vivo also resulted in decreased plasma LPC level in apoE Ϫ/Ϫ mice. This decrease can be attributed to a direct effect of CEL on hydrolysis of LPC in atherogenic lipoproteins (9). However, despite the decrease of this pro-atherogenic constituent (38), atherosclerosis was exacerbated in the CEL transgenic mice compared with their nontransgenic apoE Ϫ/Ϫ littermates (Fig. 10). These observations implied that the increased LPC level is not a prerequisite of atherogenesis and the increased atherosclerosis observed in apoE Ϫ/Ϫ mice compared with their wild type counterparts is not because of the elevated levels of LPC in the apoE Ϫ/Ϫ mice. In fact, results of the current study suggest that LPC may be anti-atherogenic under selected circumstances, such as in hypercholesterolemia where the cholesterol is esterified in macrophages to induce foam cells. Reports showing LPC promotion of cholesterol efflux from macrophage foam cells (39) and the ability of lipoprotein-associated phospholipase A 2 to reduce LDL degradation and foam cell formation in vitro (40) are supportive of this hypothesis.
In summary, our data documented that macrophage expression of CEL is pro-atherogenic, favoring cholesteryl ester accumulation and foam cell formation to increase atherosclerosis lesions in mice. The mechanism by which CEL expression in macrophages promotes atherosclerosis is likely because of its ability to promote intracellular cholesterol esterification by reducing ceramide and LPC levels in cholesterolloaded cells and lowering LPC levels in the circulation. These results, along with previous in vitro studies demonstrating CEL promotion of smooth muscle cell proliferation (12), offer a new therapeutic target to reduce the severity of vascular occlusive disease.