Macrophage lipoprotein lipase modulates the development of atherosclerosis but not adiposity.

The role of macrophage lipoprotein lipase (LpL) in the development of atherosclerosis and adiposity was examined in macrophage LpL knockout (MLpLKO) mice. MLpLKO mice were generated using cre-loxP gene targeting. Loss of LpL in macrophages did not alter plasma LpL activity or lipoprotein levels. Incubation of apolipoprotein E (ApoE)-deficient β-VLDL with peritoneal macrophages from ApoE knockout mice lacking macrophage LpL (MLpLKO/ApoEKO) led to less cholesteryl ester formation than that found with ApoEKO macrophages. MLpLKO/ApoEKO macrophages had reduced intracellular triglyceride levels, with decreased CD36 and carnitine palmitoyltransferase-1 mRNA levels compared with ApoEKO macrophages, when incubated with VLDL. Although both MLpLKO/ApoEKO and ApoEKO mice developed comparable hypercholesterolemia in response to feeding with a Western-type diet for 12 weeks, atherosclerosis was less in MLpLKO/ApoEKO mice. Epididymal fat mass and gene expression levels associated with inflammation did not differ between the two groups. In conclusion, macrophage LpL plays an important role in the development of atherosclerosis but not adiposity.

Cre recombinase was expressed in macrophages under the control of the lysozyme promoter Lys-Cre (Jackson Laboratory) ( 22 ). LpL +/f mice carrying one copy of the Lys-Cre transgene were then mated with LpL +/f mice lacking Cre to generate MLpLKO mice and littermate wild-type (WT) controls [LpL f/f (fLpL), Lys-Cre (CRE), and LpL +/+ (WT)]. Thereafter, MLpLKO mice were crossbred with ApoE knockout mice (ApoEKO) to generate MLpLKO/ ApoEKO mice. Littermate ApoEKO mice were used as controls. Genotyping was performed via PCR using genomic DNA isolated from the tail tip ( 21,22 ). Three diets were used: i ) a normal chow diet containing 4.8% (w/w) fat and 25.1% (w/w) protein (CE-2, Japan CLEA); ii ) a Western-type diet (WTD) containing 21% (w/w) fat and 0.21% (w/w) cholesterol (D12079B, Research Diets Inc., NJ); and iii ) a high-fat diet (HFD) containing 23.6% (w/w) fat (D12451, Research Diets). The WTD was used to induce obesity and atherosclerotic lesions in MLpLKO/ApoEKO mice, and the HFD was utilized to induce obesity in MLpLKO mice. Animal care and experimental procedures were performed according to the regulations of Jichi Medical University.

Northern blot analysis
Total RNA was extracted with TRIZOL reagent (Invitrogen). Ten micrograms of RNA were subjected to electrophoresis on a 1% agarose gel containing formamide, and then transferred onto a nylon fi lter (Hybond N; Amersham Bioscience). A murine LpL cDNA probe was radiolabeled with [ ␣ -32 P]deoxy-CTP. Following a 2 h prehybridization period, Northern blots were hybridized with the probe in a Rapid-hyb buffer (Amersham Biosciences) for 1 h at 65°C ( 23 ).

Real-time PCR
One microgram of total RNA was reverse-transcribed with a high-capacity cDNA reverse transcriptase kit (Applied Biosystems). All reactions were done in triplicate, and relative amounts of mRNA were calculated using a standard curve or comparative CT method on a 7300 Real-Time PCR system (Applied Biosystems), according to the manufacturer's protocol. Mouse ␤ -actin mRNA was used as the invariant control. The primer-probe sets for real-time PCR are listed in Table 1 .

Plasma lipids, lipoproteins, and glucose
Blood was collected from mice that were fasted for 16 h. Total cholesterol (TC) and TG levels in plasma were determined enzymatically via kits (Determiner TC 555 and Determiner TG 555, Kyowa Medex, Tokyo). HPLC analyses of plasma were performed as previously described ( 24 ). Blood glucose was determined via a FreeStyle blood glucose monitoring system (NIPRO).

Peritoneal macrophages
Peritoneal macrophages were obtained three days after a 1 ml intraperitoneal injection of 3% thioglycollate broth. Macrophages were plated on 12-well plates, and cultured in Dulbecco's modifi ed Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS) and antibiotics for 3 h. Thereafter, cells were washed with PBS, and if not stated otherwise, the adherent macrophages were maintained in DMEM supplemented with 10% (v/v) FBS and antibiotics.

LpL activity
Postheparin plasma was analyzed for LpL activity as previously described ( 25 ). Fasted mice were injected with 300 U/kg of heparin intravenously, and plasma was obtained 10 min later and assayed in triplicate for LpL activity. LpL activity in tissue homogenates and in media of cultured macrophages treated with heparin (10 U/ml) was measured, as previously described by Hocquette et al. ( 26 ). FA, produce chylomicron remnants and LDL, and create surface lipids that transfer to HDL ( 9,10 ). We have previously demonstrated that mice overexpressing human LpL in a variety of tissues, such as the heart, skeletal muscle, adipose, and aorta, are protected against atherosclerosis by the reduction of atherogenic lipoproteins in the LDL receptor-defi cient background ( 11 ) or by the increase of HDL in the apolipoprotein E (ApoE)-defi cient background ( 12 ). Furthermore, pharmaceutical interventions that increase LpL expression reduce atherosclerosis ( 13,14 ). Thus, through its actions on circulating lipoproteins, LpL is generally viewed as an antiatherogenic enzyme. However, none of these actions are thought to involve macrophage LpL.
Macrophage LpL is thought to increase macrophage infl ammation capacity and promote atherosclerosis. LpL can act as a molecular bridge between proteoglycans and lipoprotein receptors, such as remnant-like particles and LDL receptors, to induce the retention and/ or uptake of atherogenic lipoproteins by cells within the vascular wall ( 15,16 ). Macrophage LpL disruption through the transplantation of bone marrow or fetal liver cells from LpL-defi cient mice into C57BL/6 mice decreases atherosclerotic lesion development ( 17,18 ). These techniques are powerful tools for the study of macrophage regulation of plasma lipoprotein metabolism and in vivo atherogenesis process ( 19 ). However, graft rejection or irradiation of recipient mice may infl uence the experimental outcomes. Indeed, the fi ndings regarding plasma lipoproteins in two macrophage LpL-defi cient mouse models generated by two different laboratories were dissimilar ( 17,18 ).
The role of LpL in adipose macrophage biology has not been explored. In one model, the LpL knockout mice rescued by a transgenic expression of LpL under the control of a muscle creatine kinase promoter had LpL defi ciency in both adipocytes and macrophages. These mice have diminished weight gain and fat mass on an ob/ob background ( 20 ). However, the specifi c role of macrophage LpL, which infi ltrates obese adipose tissue, remains unclear. Similarly, the importance of macrophage LpL in adipose infl ammation due to a high-fat diet is unknown.
In the present study, we generated macrophage LpL knockout (MLpLKO) mice using cre/loxP gene targeting and determined the role of macrophage LpL in plasma lipoprotein metabolism, atherosclerosis, and adiposity.

Mice and diets
Mice heterozygous for the fl oxed LpL allele (referred to as LpL +/f ; f denotes the fl oxed allele) on a 129/Sv background were generated as described elsewhere ( 21 ). LpL +/f mice, which were backcrossed four times into the C57B/6 background, were crossbred with transgenic mice on a C57B/6 background, in which with saline containing 4% (w/v) formalin and fi xed for more than 48 h in the same solution. The basal half of the heart was embedded in Tissue-Tek OCT compound (Sakura Finetek Co., Tokyo), and serial sections were embedded in Cryostat (6 m thick), as previously described ( 12 ). Four sections, each separated by 60 m, were used to evaluate the lesions; two at the end of the aortic sinus and two at the junctional site of sinus and ascending. Sections were stained with Oil red O and counterstained with hematoxylin. Cross-sections of aortic roots were incubated with primary antibodies for mouse MOMA-2 (1:600; Accurate Chemical and Scientifi c) overnight at 4°C to investigate macrophage infi ltration in the atherosclerotic lesion. After washing, sections were incubated with biotinylated antirat antibody for 1 h at room temperature, and then with avidinbiotin peroxidase complex (Vector Labs) for 30 min. Lastly, sections were developed with 3,3 ′ -diaminobenzidine tetrahydrochloride (DAB) (Sigma), and counterstained with hematoxylin.

Histology
The adipose tissue of mice fed a WTD for 12 weeks was fi xed with 10% neutral-buffered formalin, embedded in paraffi n, and sections were stained with hematoxylin and eosin. Sections were photographed and the average size of lipid droplet from each mouse was determined by analyzing 200 cells using ImageJ software (National Institutes of Health).

Statistics
Data were presented as means ± S.E. Student t-test was used to compare the mean values between two groups, and the one way ANOVA (ANOVA) was used for multiple comparisons. When ANOVA results were statistically signifi cant (i.e., P < 0.05), then individual comparisons were made with the Tukey posthoc test.

Characterization of MLpLKO mice
Peritoneal macrophages from MLpLKO mice had signifi cantly less LpL mRNA than WT, CRE, and fLpL mice ( Fig. 1A , B ). LpL activity in macrophages was also barely detectable in MLpLKO mice ( Fig. 1C ). In contrast, postheparin plasma LpL activity was not reduced in MLpLKO

Cholesteryl ester formation assay
Blood was collected from ApoEKO mice that were fasted for 16 h after being fed a WTD for over two weeks. ␤ -VLDL was isolated by density gradient ultracentrifugation. Following a 24 h incubation period in medium containing 5 mg/ml of lipoprotein-defi cient serum (LPDS), macrophages from both ApoEKO and MLpLKO/ApoEKO mice were incubated with either 5 or 10 g of ApoE-defi cient ␤ -VLDL, 5 mg/ml of LPDS, and [1-14 C] oleate-albumin complex at 37°C for 24 h. Cholesteryl ester formation was determined as previously described ( 12 ).

VLDL uptake
VLDL was isolated from plasma of normolipidemic volunteers fasted overnight. Macrophages were incubated with 100 g of VLDL in DMEM containing 5% FBS for 6 h in a Lab-Tek II Chamber slide system (Thermo Scientifi c Inc.). After washing with 4% paraformaldehyde, macrophages were stained with Oil red O. Macrophages were also incubated with 100 g of VLDL in 12 well plates for 6 h. Thereafter, cells were washed twice with PBS, and lipids were extracted with hexane/isopropanol (3:2, v/v). The organic phase was evaporated to dryness under fl owing nitrogen, redissolved in isopropanol, and then cholesterol and TG levels were enzymatically determined. Cellular proteins were dissolved in 0.1N NaOH and assessed with a BCA kit. Total RNA was also extracted, and the genes associated with intracellular FA metabolism, proinfl ammatory cytokines, and ER stress were quantifi ed by real-time PCR.

Diet-induced atherosclerosis
At 14 weeks of age, ApoEKO and MLpLKO/ApoEKO male mice were fed a WTD for 12 weeks. Blood was collected, and plasma lipids and lipoproteins were determined. Mice were euthanized after 12 weeks, and atherosclerosis was determined. Atheromatous plaques in the aorta were visualized by staining with Sudan IV, and the luminal side of the stained aorta was photographed ( 12 ). Image capture and analysis were performed with Adobe Photoshop 6 image analysis software. The extent of atherosclerosis was expressed as the percentage of surface area of the entire aorta covered by lesions and designated as the en face surface lesion area ( 27 ). The cross-sectional lesion area was evaluated according to the modifi ed method of Paigen et al. ( 28 ). In brief, the heart was perfused  ␤ -actin  CGATGCCCTGAGGCTCTTT  TGGATGCCACAGGATTCCA  CCAGCCTTCCTTCTT  CPT-1  GAACCCCAACATCCCCAAAC  TCCTGGCATTCTCCTGGAAT  CACCAGGCTACAGTGG  TLR4  AGCTTCAATGGTGCCATCATT  CCAGGTGCTGCAGCTCTTCT  TGAGTGCCAATTTCATGG  FABP4  CCGCAGACGACAGGAAGG  AGGGCCCCGCCATCT  AAGAGCATCATAACCC  TNF-␣  AGGGATGAGAAGTTCCCAAATG  TGTGAGGGTCTGGGCCATA  CCTCCCTCTCATCAGTT  CHOP  CATCCCCAGGAAACGAAGAG  GCTAGGGACGCAGGGTCAA  AAGAATCAAAAACCTTCACTACT  Arginase1 CATGGGCAACCTGTGTCCTT Cholesteryl ester formation in mouse peritoneal macrophages ␤ -VLDL was isolated from ApoEKO mice fed a WTD for over two weeks, and it was used to stimulate cholesteryl ester formation in peritoneal macrophages. Cholesteryl ester foam cell formation by MLpLKO/ApoEKO macrophages incubated with 5 g of ApoE-defi cient ␤ -VLDL was 31% less than that by ApoEKO macrophages, and the reduction with 10 g of ␤ -VLDL was 51% less ( Fig. 2 ).

Metabolic effects of macrophage LpL defi ciency
To investigate the consequences of macrophage LpL defi ciency on intracellular lipid metabolism, MLpLKO/ ApoEKO macrophages were incubated with VLDL isolated from normolipidemic human volunteers. Oil red O staining demonstrated that VLDL loading resulted in substantial amounts of intracellular lipid droplets in ApoEKO macrophages but not in MLpLKO/ApoEKO macrophages ( Fig. 3A ). At baseline, cholesterol and TG concentrations were comparable in macrophages from ApoEKO and ML-pLKO/ApoEKO mice ( Fig. 3B ). VLDL loading resulted in a greater increase of TG in ApoEKO than in MLpLKO/ ApoEKO macrophages ( P < 0.001). Cholesterol levels mice compared with control fLpL mice ( Fig. 1D ). The ApoEKO background did not infl uence the macrophage and plasma LpL. Like MLpLKO mice, MLpLKO/ApoEKO mice had much reduced macrophage LpL, but not plasma LpL. Furthermore, macrophage LpL defi ciency did not infl uence LpL activity in heart, muscle, adipose tissue, or lung in the ApoEKO mice ( Fig. 1E ).
Plasma lipids and lipoproteins were determined in MLpLKO mice fed a regular chow diet. Plasma TC, TG, lipoprotein, and glucose levels were not signifi cantly different between MLpLKO and fLpL mice ( Table 2 and Fig. 1F ). MLpLKO/ApoEKO mice had signifi cantly higher plasma TC levels than MLpLKO mice. There were no signifi cant differences in plasma TC and TG levels between MLpLKO/ApoEKO and ApoEKO mice ( Table 3 ). We also assessed plasma lipids in MLpLKO/ApoEKO and ApoEKO mice fed a WTD diet for 12 weeks. The WTD induced a 2.2-and 2.3-fold increase in plasma TC in ApoEKO and MLpLKO/ApoEKO mice, respectively, with no significant differences between the two groups. Therefore, macrophage LpL did not infl uence plasma lipoprotein metabolism in WT and ApoEKO mice, which have defective lipoprotein metabolism.  ( Fig. 4A ). The en face surface lesion areas of MLpLKO/ ApoEKO mice were also smaller than those of ApoEKO mice by 20.6% (13.5 ± 1.0% versus 17.0 ± 1.4%, P < 0.05) ( Fig. 4B ). Macrophage content in cross-sectional lesion areas stained with MOMA-2 was also decreased in MLpLKO/ ApoEKO mice in proportion to the reduction of atherosclerotic lesion area. The ratio of macrophage-infi ltrating area to atherosclerotic lesion area was not signifi cantly different between the two genotypes ( Fig. 4C ). Thus, LpLdefi cient macrophages prevent lesion development without altering macrophage content in ApoEKO mice.
Peritoneal macrophages were isolated from MLpLKO/ ApoEKO and ApoEKO mice fed a WTD for 12 weeks, and gene expression levels of infl ammation markers were assessed. Although there were no signifi cant differences in MCP-1, IL6, iNOS, or CD36, TNF ␣ was decreased by 50% ( P < 0.01) in MLpLKO/ApoEKO macrophages compared with ApoEKO macrophages ( Fig. 5A ). Moreover, mRNA was extracted from whole aorta in mice fed a WTD for 12 weeks, and gene expression levels were assessed. Although LpL was signifi cantly decreased in the aortas from ML-pLKO/ApoEKO mice, the expression levels of infl ammation markers did not differ between ApoEKO and MLpLKO/ApoEKO mice (Fig. 5B). So the effects of diet and/or its secondary actions in vivo affect macrophages differently than was found with in vitro VLDL treatment.

Diet-induced obesity
As there are substantial amounts of LpL in adipose tissue where macrophages infi ltrate during the development of obesity, we assessed the consequence of macrophage LpL defi ciency on adiposity. At 9 weeks of age, male MLpLKO mice were fed a HFD for 10 weeks and weighed weekly. Both MLpLKO and control fLpL mice gained weight at similar rate over the entire observation period ( Fig.6A ). Moreover, neither body weight nor epididymal fat mass was signifi cantly different between ApoEKO and MLpLKO/ ApoEKO mice fed a WTD for 12 weeks ( Fig. 6B ). The size of lipid droplet in adipocytes did not differ between MLpLKO/ ApoEKO and ApoEKO mice ( Fig.6C ).
On regular chow diet, LpL mRNA levels in adipose tissue did not differ between the ApoEKO and MLpLKO/ ApoEKO mice, despite the similar expression levels of were marginally but signifi cantly decreased in MLpLKO/ ApoEKO compared with ApoEKO macrophages ( P < 0.05). These observations indicate that the hydrolysis of VLDL via macrophage LpL can control intracellular TG accumulation even in the absence of ApoE.
We then assessed whether changes in lipid uptake altered the expressions of genes involved in intracellular FA metabolism, infl ammation, and ER stress. In the absence of exogenous lipid loading, gene expression of CD36, TLR4, FABP4, and carnitine palmitoyltransferase-1 (CPT-1) did not differ between ApoEKO and MLpLKO/ ApoEKO macrophages. VLDL loading approximately doubled CD36 mRNA in ApoEKO macrophages, but it led to no changes in MLpLKO/ApoEKO macrophages ( Fig. 3C ), which resulted in a 45% reduction of CD36 mRNA in MLpLKO/ApoEKO macrophages compared with ApoEKO macrophages ( P < 0.01) Although both macrophages had increases in CPT-1 in response to VLDL loading, ApoEKO macrophages had a 2.9-fold increase in CPT-1, whereas MLpLKO/ApoEKO macrophages had only a 1.5-fold increase ( P < 0.05). Unlike CD36 and CPT-1, VLDL loading did not lead to a signifi cant difference in TLR4, ACOX1, MCAD, or FABP4 between the two macrophages.
Infl ammation is well documented to alter atherosclerosis development. Surprisingly, VLDL loading significantly reduced TNF-␣ expression in both LpL-expressing and LpL-knockout macrophages; however, ApoEKO macrophages had lower TNF-␣ expression than MLpLKO/ ApoEKO macrophages. Lipid loading did not lead to ER stress; CHOP did not differ between the two types of macrophages irrespective of the presence of VLDL. These data complement recent studies by Spann et al. showing reduced infl ammation in foam cells from mice fed a high-cholesterol diet ( 29 ).

Diet-induced atherosclerosis
Following 12 weeks on a WTD, atherosclerosis was evaluated by two methods: cross-sectional analysis of aortic roots and the en face surface lesion area of the aorta. Cross-sectional lesion areas of MLpLKO/ApoEKO mice were signifi cantly smaller than those of ApoEKO mice by 19.3% (505 ± 24 × 10 3 m 2 versus 626 ± 30 × 10 3 m 2 , P = 0.01)  were not observed between the two types of mice. When mice were challenged with a WTD for 12 weeks, F4/80 and CD11c mRNA levels were markedly increased in both adipose tissues, but no signifi cant differences were found between the lines. In contrast, the WTD signifi cantly reduced CD206 in both types of adipose tissue, with no signifi cant difference between two groups.

DISCUSSION
In the present study, we generated a macrophage LpL knockout mouse using cre-loxP gene targeting. The following important physiological roles of macrophage LpL were observed in this mouse model: i ) macrophage LpL did not infl uence postheparin plasma LpL activity or lipoprotein metabolism; ii ) macrophage LpL defi ciency led to a small decrease in cholesterol ester foam cell formation and dietinduced atherosclerosis; iii ) intracellular TG accumulation was reduced in association with decreased CD36 expression; and iv ) a deletion of macrophage LpL did not infl uence LpL activity or infl ammation in adipose tissue.
Cre-loxP gene targeting resulted in a complete deletion of LpL in macrophages. MLpLKO mice demonstrated normal plasma LpL activity and lipoprotein profi les. Furthermore, MLpLKO/ApoEKO mice had a similar level of plasma lipids as ApoEKO mice, a model of defective lipoprotein metabolism, when fed a regular chow diet or WTD. macrophage markers, specifi cally CD68 and F4/80, in adipose tissue ( Fig. 6D ). CD11c, a marker of M1 macrophage, was increased by 4.6-fold in adipose tissue from MLpLKO/ ApoEKO mice compared with ApoEKO adipose tissue. Signifi cant differences in TNF-␣ or MCP-1 expression levels Fig. 2. Cholesteryl ester formation stimulated via ApoE-defi cient ␤ -VLDL. ␤ Ϫ VLDL was obtained from ApoEKO mice that had been fed a WTD for over two weeks. Thioglycollate-elicited peritoneal macrophages were prepared from ApoEKO (open bars, n = 10) and MLpLKO/ApoEKO mice (fi lled bars, n = 10). Following LPDS treatment for 24 h, cells were incubated with either 5 or 10 g of ApoE-defi cient ␤ -VLDL for 24 h, and then cholesteryl ester formation was determined. Values are expressed as means ± SE. * P < 0.05, ** P < 0.01. Values are expressed as means ± SE. * P < 0.05, ** P < 0.01, *** P < 0.001. specifi cally the use of transplanted LpL-knockout fetal liver cells and bone marrow versus cre-loxP to generate macrophage LpL defi ciency.
We did not fi nd differences in LpL in the lung. Some lung LpL is thought to be synthesized in other tissues and then trapped within the lung due to the high level expression of GPIHBP1. In fact, Young and his colleagues showed that mice with a total body deletion of LpL rescued with the expression only in the skeletal muscle had readily detectable LpL protein in the lung despite low LpL transcription levels ( 31 ).
ApoE is the most important ligand for receptor-mediated lipoprotein uptake. LpL also functions as a ligand to promote lipoprotein binding to the LDL receptor, LDL receptor-associated protein, and extracellular proteoglycans (32)(33)(34). In the present study, MLpLKO mice on an ApoEKO background were fed a WTD for 12 weeks to elucidate the in vivo consequences of macrophage LpL deficiency on atherosclerosis. Foam cell formation stimulated via ApoE-defi cient ␤ -VLDL was signifi cantly reduced in MLpLKO/ApoEKO macrophages compared with ApoEKO macrophages. This suggested that macrophage LpL plays an important role in the development of atherosclerosis. We found that the cross-sectional lesion areas of the aortic root were signifi cantly reduced by 19.3% in MLpL/ApoEKO versus ApoEKO mice, when challenged with a WTD for 12 weeks. The en face aortic lesion areas were also significantly decreased by 20.6% in MLpLKO/ApoEKO compared with ApoEKO mice. These observations are in agreement with the fi ndings of Babaev et al. and Eck et al., who demonstrated that macrophage LpLKO mice generated via bone marrow or fetal liver cell transplantation had less diet-induced atherosclerosis on a WT or LDL receptor knockout background ( 17,18,30 ). It should be noted that the consequences of a macrophage LpL defi ciency on attenuating atherosclerosis were more moderate in an ApoE or LDLR knockout background compared with WT.
VLDL loading resulted in a marked TG accumulation in ApoEKO macrophages, whereas this effect was completely prevented in LpL-defi cient macrophages. The accumulation of cholesterol from VLDL was very small in ApoEKO macrophages. VLDL-induced lipid accumulation occurs via three different mechanisms: i ) the uptake of whole VLDL particles; ii ) the uptake of VLDL remnants resulting from VLDL-TG lipolysis; and iii ) the uptake of FA produced by VLDL-TG lipolysis. Since cholesterol accumulation through VLDL loading was much smaller than that of TG in ApoEKO macrophages, the uptake of FA generated from VLDL lipolysis via LpL enzymatic activity could be the most important of the three mechanisms in the setting of ApoE defi ciency.
Surprisingly, VLDL loading reduced TNF-␣ expression in both ApoEKO and MLpLKO/ApoEKO macrophages. The effect of VLDL lipolysis on proinfl ammatory cytokine production is controversial ( 35,36 ). Saraswathi et al. reported that VLDL induces proinfl ammatory genes, including TNF-␣ and MCP-1, in macrophages ( 36 ), whereas Li et al. revealed no such effects of VLDL on TNF-␣ expression ( 35 ). The reason for this discrepancy is unclear; however, The effects of macrophage LpL on plasma lipoproteins are controversial ( 17,18,30 ). Eck et al. reported that macrophage LpL deletion via bone marrow transplantation (BMT) in WT mice reduced plasma cholesterol, whereas TG levels were increased eight weeks after BMT ( 18 ). Conversely, consistent with our fi ndings, Babaev et al. reported that macrophage LpL did not affect plasma LpL activity or lipoproteins in mice with a WT or LDL receptor knockout background ( 17,30 ). Furthermore, despite the previous studies demonstrating that macrophage LpL knockout mice generated with bone marrow or fetal liver cell transplantation have a reduced LpL activity in the heart, spleen, or lung, where macrophages were present, MLpLKO mice in the present study had normal LpL activity in these tissues. The reason for this discrepancy is unclear, but it might be related to differences in technologies, FA and greater expression of CD36 in macrophages appeared to facilitate FA uptake from VLDL. This would be expected to supply the cells with a lipid source of substrate for energy generation. Macrophage phagocytosis demands a large amount of ATP and plays an essential role in host defense. Chandak et al. have recently reported that a defi ciency in adipose triglyceride lipase, a rate limiting enzyme involved in the hydrolysis of lipid dropletassociated TG, in macrophages resulted in a decrease in CPT-1 expression and impaired phagocytosis ( 42 ). Others have also previously reported that macrophage LpL acts on TG-rich lipoproteins to produce FA, which is an important source of energy for phagocytosis ( 43 ). Thus, it is possible that reduced availability of FA leads to decreased macrophage phagocytosis in MLpLKO mice. Macrophage phagocytosis has been shown to enhance or suppress the development of atherosclerosis, depending on the context ( 44,45 ).
Obesity is associated with chronic infl ammation of adipose tissue, where macrophages are a major source of proinfl ammatory mediators ( 2,7 ). However, the role of the FA composition of TG from VLDL, unsaturated versus saturated FA ( 36 ), might infl uence proinfl ammatory cytokine production in macrophages. In addition, both the amount of lipoproteins and LpL and the location of the lipolysis (in medium or on the cell surface) might alter the relative exposure of the cells to LpL-produced lipids.
Gene expression analysis provided some insight into how LpL defi ciency alters macrophage function and in turn atherosclerosis. Exposure of ApoEKO macrophages to VLDL increased CD36 mRNA levels, but MLpLKO/ ApoEKO macrophages did not increase CD36 mRNA after VLDL treatment. CD36 has many functions, including as a scavenger receptor for oxidatively modifi ed LDL in macrophages, and some ( 37,38 ), but not all ( 39 ) studies suggest that loss of CD36 reduces atherosclerosis. CD36 is also a receptor/transporter for long-chain FA in adipose tissue, heart, and skeletal muscle ( 40 ). CD36 is downstream of peroxisome proliferator-activated receptors (PPAR), especially the anti-infl ammatory PPAR ␥ ( 41 ), and its increased expression in LpL-expressing macro phages is likely due to PPAR activation. Thus, LpL production of Values are expressed as means ± SE. * P < 0.05, ** P < 0.01, *** P < 0.001. creating remnant lipoproteins that can interact with cell surface lipoprotein receptors ( 46 ). Furthermore, our data demonstrate that macrophage LpL is not involved in the infl ammatory response of adipose to a high-fat diet. Our studies further illustrate the physiological functions of LpL in various tissues and add to the perception that reduced LpL in macrophages, but not muscle, may be desirable.
macrophage LpL on the development of obesity is still unclear. We found that body weight increased at a similar rate in fLpL and MLpLKO mice on a HFD for 10 weeks. Moreover, body and gonadal fat pad weights did not differ between ApoEKO and MLpLKO/ApoEKO mice after 12 weeks on a WTD. Despite the comparable macrophage infi ltration into adipose tissue, as determined by CD68 or F4/80 expression levels, LpL activity and its mRNA expression in adipose tissue did not differ between ApoEKO and MLpLKO/ApoEKO mice. MCP-1 and TNF-␣ expression levels were not lower in the adipose tissue of MLpLKO/ ApoEKO mice. These data suggest that macrophage LpL in adipose tissue is not a major modulator of adiposity.
In conclusion, we generated mice with a macrophage deletion of LpL and demonstrated a role of macrophage LpL in the development of atherosclerosis. Because the area occupied by macrophages in lesions from mice with or without macrophage LpL expression was identical, LpL effects on lesion size and macrophage content are commensurate. Studies in vitro suggest that the primary role of LpL is to affect macrophage lipid uptake, perhaps by