Differential Additive Effects of Endothelial Lipase and Scavenger Receptor-Class B Type I on High-Density Lipoprotein Metabolism in Knockout
http://www.100md.com
《动脉硬化血栓血管生物学》
From the Section of Endocrinology and Metabolism (K.M., L.C.), Departments of Medicine and Molecular & Cellular Biology, Baylor College of Medicine, and St. Luke’s Episcopal Hospital, Houston, Tex; Lawrence Berkeley National Laboratory (T.F.), Berkeley, Calif; and LipoScience (J.D.O.), Raleigh, NC.
Correspondence to Lawrence Chan, Division of Endocrinology & Metabolism, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030. E-mail lchan@bcm.tmc.edu
Abstract
Objective— Endothelial lipase (EL) is a vascular phospholipase that hydrolyzes high-density lipoprotein (HDL) as its preferred substrate. Scavenger receptor-class B type I (SR-BI) is an HDL receptor that mediates the selective uptake of cholesteryl ester. This study investigates the role of EL and SR-BI in the regulation of HDL metabolism in gene knockout mouse models.
Methods and Results— We cross-bred EL–/– and SR-BI–/– mice and generated single- and double-null mice. We used biochemical, molecular biology, and nuclear magnetic resonance methods to analyze HDL concentration, composition, and structure. We found that EL and SR-BI display additive effects on HDL with evident gene dosage effects, but their mechanisms to regulate HDL concentration and composition are different. Whereas the elevated HDL cholesterol level in EL–/– mice is associated with increased phospholipid content in HDL particles, SR-BI–/– mice display markedly enlarged HDL particles shifted to larger subclasses with a phospholipid content similar to that of wild-type mice. Furthermore, absence of EL is associated with a 40% to 50% inhibition and absence of SR-BI, a 90% inhibition of endogenous lecithin cholesterol:acyltransferase rate.
Conclusions— EL and SR-BI are major genetic determinants of HDL metabolism in vivo, each exercising independent and additive effects on HDL structure and function.
We created EL and SR-BI single- and double-null mice and compared their effects on HDL concentration, composition, and structure. We demonstrated that EL and SR-BI display additive effects on plasma HDL in vivo with evident gene dosage effects. The 2 genes use different mechanisms to regulate HDL concentration and composition.
Key Words: endothelial lipase ? scavenger receptor-class B type I ? high-density lipoprotein
Introduction
Numerous epidemiological studies have established a negative correlation between high-density lipoprotein (HDL) levels and atherosclerosis development.1–3 HDL concentration is largely determined genetically.4,5 Multiple genes are directly involved in the HDL metabolic pathways,5 although other genetic factors influencing HDL level exist that are poorly understood.
Endothelial lipase (EL) is a newly described member of the vascular lipase family, sharing sequence homology with lipoprotein lipase and hepatic lipase (HL).6,7 EL possesses predominantly phospholipase activity and uses HDL as its preferred substrate.8 We recently showed that absence of EL in mice leads to delayed HDL clearance and elevated HDL levels, and that a common single-nucleotide polymorphism in the human EL gene (581 C/T) is associated with significantly higher HDL cholesterol levels.9 Similar findings were also reported on EL-null mice created independently by Ishida et al.10
Scavenger receptor-class B type I (SR-BI) is the major HDL receptor responsible for its selective uptake in liver and steroidogenic tissues.11 Selective uptake, a process that involves the transfer of cholesteryl ester from the lipoprotein core into cells without internalization of the whole lipoprotein particle, accounts for up to 70% of the HDL delivery in the liver12 and 90% of the cholesterol used for steroid hormone production in the adrenal gland in rodents.13 Overexpression of SR-BI in mice by adenovirus-mediated gene transfer14 or by a transgenic approach15 led to a marked reduction of plasma HDL levels, whereas targeted disruption of SR-BI led to markedly increase plasma HDL cholesterol.16
In this study, we used knockout mouse models to examine gene–gene interaction between EL and SR-BI, because both play dominant roles in HDL metabolism. We found that the effect of loss of EL on HDL composition and subclass size distribution was different from that of loss of SR-BI, although their individual effects on HDL level were additive. The 2 genes exert independent influence on HDL size, composition, and lecithin cholesteryl acyltransferase (LCAT) activity in vivo. Our findings have significant implications for the genetic regulation of the concentration and metabolism of HDL.
Methods
Animals
EL–/– mice were created by gene targeting9 and backcrossed to C57BL/6 for 4 generations. Mice were weaned at 21 days and fed a standard laboratory chow (7001; Teklad, Madison, Wis). SR-BI+/– mice were purchased from Jackson Laboratory (Bar Harbor, Me)16 and bred to homozygosity in a mixed 129/C57BL/6J background. EL–/–/SR-BI–/– mice were obtained by crossing EL–/–/SR-BI–/+ breeding pairs. Genotypes of EL–/– was confirmed by polymerase chain reaction (PCR) using the wild-type allele-specific primers: EL142 (5'-AGCTTGAAGGGTGACTTGAG-3') and EL18 (5'-CCTTCATGATTGTTCTTCAC-3'), and the targeted allele-specific primers: EL142 and Neo5 (5'-CTATCGCCTTCTTGACGAGT-3'). SR-BI genotyping was performed by PCR16 using primers SR-1, 2, and 3. All PCR reactions were performed at annealing temperature of 58°C and 32 cycles.
Plasma Lipids, Lipoproteins, and Lipoprotein Particle Chemical Composition Analysis
Cholesterol, phospholipids, triglyceride, HDL cholesterol, total cholesterol, and free cholesterol concentrations were assayed using commercial kits (Wako Chemicals, Richmond, Ga). For fast protein liquid chromatography (FPLC) analysis, plasma pooled from 5 to 7 mice (200 μL) after a 5-hour fast was loaded onto 2 Superose-6 columns connected in series (Amersham Pharmacia Biotech). Fractions (0.5 mL) were collected: fractions 4 to 9, very-low-density lipoprotein (VLDL); 16 to 22, intermediate-density lipoprotein (IDL)/low-density lipoprotein (LDL); 26 to 38, HDL.
Chylomicron/VLDL, density <1.006; IDL/LDL, 1.006
Lipoprotein Nuclear Magnetic Resonance Analysis
Lipoprotein subclass profiles were measured on frozen plasma (100 μL) by proton nuclear magnetic resonance (NMR) spectroscopy at LipoScience, Inc (Raleigh, NC) as previously described.9 The designations of the NMR-derived HDL subclasses and their estimated diameter ranges are as follows: H1 (7.3 to 7.7 nm), H2 (7.8 to 8.4 nm), H3 (8.5 to 8.9 nm), H4 (9.0 to 10.0 nm), H5 (10.1 to 12.0 nm), H6 (12.1 to 13.5), and H7 (13.6 to 15.0). HDL subclass particle concentrations were expressed in particle numbers (nmol/L) as derived from raw NMR data.9 Mean HDL particle size was determined by weighting the relative percentage of each subclass by its diameter.
Nondenaturing Polyacrylamide Gel Electrophoresis
Nondenaturing 4% to 30% polyacrylamide gradient gels were used to analyze the size distribution of HDL as described by Nichols et al18 using HDL with 1.063
Immunoblotting, Northern Blotting, and Semi-Quantitative Reverse-Transcription PCR
Immunoblotting of SR-BI and phospholipid transfer protein (PLTP) was performed as described16 using 40 μg of total protein in lysis buffer containing 50 mmol/L Tris-HCl, pH 7.4, 1% NP-40, 0.25% sodium deoxycholate, 150 mmol/L NaCl, 1 mmol/L EDTA, 1 mmol/L PMSF, 1 μg/mL aprotinin, leupeptin, pepstatin; 1 mmol/L Na3VO4, and 1 mM NaF. Proteins were transferred to nitrocellulose membrane after electrophoresis and immunoblots were developed by chemiluminescence detection (Supersignal, Pierce). Antibodies used were: goat anti-human apolipoprotein A-I, E antibodies, and mouse anti-actin monoclonal antibody (Chemicon), and rabbit anti-SR-BI and anti-PLTP polyclonal antibody (Novas Biologicals, Littleton, Co). Anti–SR-BI antibody recognizes the 82-Kd SR-BI protein, and the anti-human PLTP antibody cross-reacts with the mouse PLTP protein of 75 Kd.19 Polyclonal anti-mouse LCAT antibody was a gift from Dr John Parks at Wake Forest University (Winston-Salem, NC). RNA was extracted by Trizol reagent (Invitrogen). For Northern blotting, 20 μg of total hepatic RNA was separated on 0.2 mol/L formaldehyde 1.2% agarose gel, transferred to nylon membrane, and hybridized with gene-specific cDNA probes as indicated. Membranes were developed using a phospho-imager (Cyclone; PerkinElmer Life Sciences). Quantitation was performed by densitometry scanning using -imager LabWorks image acquisition and analysis software (UVP Inc, Upland, Calif). For semi-quantitative reverse-transcription (RT)-PCR, PCR reaction was performed using the same amount of RT product with EL, SR-BI, and GAPDH, and monitored at PCR cycles from 24 to 36.
LCAT Activity Assay
Plasma LCAT activity against endogenous substrate was measured using lipoproteins in whole plasma (50 μL) as substrate as described.9,20 [1, 2(n)-3H]-cholesterol (1.0 mCi/mL) was from Amersham Pharmacia Biotech (Buckinghamshire, England). Briefly, plasma was pre-incubated with 10 μL of [1, 2(n)-3H]-cholesterol bovine serum albumin solution (5 μCi/mL) at 4°C for 4 hours before incubating at 37°C to start the reaction. Reaction was stopped by addition of 2 mL chloroform:methanol (2:1) after 30 minutes and the lower-phase chloroform was extracted. Thin-layer chromatography was then performed and the radiolabels quantified by scintillation counting (Beckman Instrument). Results were expressed as the percentage of cholesterol esterified.
Statistical Analysis
Values were expressed as mean±SD. The differences between different genotypes were calculated by the 2-tailed Student t test. P<0.05 is considered statistically significant.
Results
Differential Effects of EL and SR-BI on Plasma Lipid Level and Lipoprotein Composition
Corroborating and extending previous reports,9,10,16 we found that absence of EL or SR-BI led to a significant increase in plasma total cholesterol level. The effect of lack of SR-BI was greater than that resulting from a lack of EL, 242% versus 186%, respectively, compared with wild-type (Table). In both cases, the elevated total cholesterol was associated with increased HDL cholesterol level, which accounts for >75% of the increase in cholesterol in EL–/– and >65% in SR-BI–/– samples. The effects of EL and SR-BI on cholesterol and HDL cholesterol were additive. Plasma total cholesterol level of the EL–/–/SR-BI–/– mice was 335% of wild-type controls and their HDL cholesterol level, 367% of wild-type, significantly higher than that in either EL–/– or SR-BI–/– groups. Thus, both EL and SR-BI exert major influence on plasma cholesterol level, primarily through their effect on HDL. In contrast, phospholipid level was significantly elevated in EL–/– mice but remained unchanged in SR-BI–/–, with the EL–/–/SR-BI–/– mice displaying a phospholipid level comparable to that of the EL–/–. These data suggest that only EL, not SR-BI, directly modulates phospholipid concentration in vivo.
Plasma Lipid Levels of Different Genotypes of Mice
Our analysis of cholesterol and phospholipid levels in FPLC fractions reflected the changes in plasma concentrations (Figure 1A and 1C). Compared with wild-type, cholesterol concentration in HDL fractions increased progressively in EL–/–, SR–/–, and EL–/–/SR–/–, although the phospholipid levels were elevated only in mice lacking EL. There was a minimally higher triglyceride level in these mice compare to the wild-type in the VLDL fraction (Figure 1D) in this experiment, which is not evident in other analyses. Free cholesterol level in both SR–/– and EL–/–/SR–/– mice was markedly higher than that in either wild-type or EL–/– mice (Figure 1B), suggesting an increased free cholesterol content of HDL particles in mice lacking SR-BI. A consistent finding was that whereas the HDL peak of EL–/– mice largely overlapped with that of the wild-type at 34 fraction, it shifted to 28 in both SR–/– and EL–/–/SR–/–, indicating the presence of much larger HDL particles. Thus, SR-BI exerts a much greater influence on HDL particle size than does EL.
Figure 1. Lipid analysis of FPLC fractionation of mice in 4 genotypes: WT, EL–/–, SR–/–, and EL–/–/SR–/– on regular chow diet. Cholesterol (A), free cholesterol (B), phospholipid (C), and triglyceride (D) levels in FPLC fractions in pooled plasma samples from each group of mice (n=7 to 10).
We analyzed lipid composition of the lipoproteins isolated by density gradient ultracentrifugation. There was no significant difference in the lipid composition of the VLDL and IDL/LDL fractions (data not shown), but it differed markedly in HDL species from the 4 genotypes. As shown in Figure 2A and 2B, there was a progressive decrease in the protein content of the HDL from wild-type to EL–/– and to SR-BI–/– with or without EL inactivation. Although the relative amount of total (free plus esterified) cholesterol remained fairly constant, the ratio of free to esterified cholesterol changed, being highest in SR-BI–/– mice with or without concomitant EL inactivation (6-fold that of wild-type or EL–/– alone). The reduced level of esterified cholesterol and markedly higher free-to-esterified cholesterol ratio of these animals (Figure 2A and 2B) could be a result of reduced LCAT activity (vide infra).
Figure 2. Chemical composition analysis of HDL separated by density ultracentrifugation in 4 groups of mice using pooled samples from 7 to 9 mice each group. A, Chemical composition in 4 genotypes of mice as represented by the percentage of each component in the HDL particle. B, Schematic representation of the chemical composition of HDL particles in 4 groups of mice. Note that the percent of triglyceride is too small to allow it to be discernible in the figure. C, Immunoblot analysis of apoA-I and apoE in VLDL (V), LDL (L), and HDL (H) fractions separated by ultracentrifugation method in 4 different groups.
FPLC analysis revealed the presence of significantly larger HDL particles in SR-BI–/– and EL–/–/SR-BI–/– mice, which might have exhibited buoyant densities that moved to a lower density range. We therefore checked for the presence of apoA-I, the major HDL apolipoprotein, and apoE, known to be present in multiple lipoprotein species, in VLDL, LDL, and HDL fractions. As shown in Figure 2C, apoA-I was present only in the HDL of wild-type and EL–/– mice, whereas apoE was detected in all lipoprotein fractions. Interestingly, in SR-BI–/– and EL–/–/SR-BI–/– mice, we found apoA-I in both the HDL density and the LDL density fractions. The presence of apoA-I in the LDL density fraction of SR-BI–/– mice, whether they carried the EL–/– alleles, indicates that there was "spillover" of the large HDL species into the IDL/LDL density range during isolation by ultracentrifugation.
Additive Effects of EL and SR-BI on HDL Apolipoprotein Levels
Our results revealed that effects of EL and SR-BI on HDL concentration were additive. We next analyzed their effects on HDL-associated apolipoproteins (see http://www.atvb.ahajournals.org). Absence of EL and SR-BI increased the level of the HDL-associated apolipoproteins, apoA-I and apoE, in plasma (Figure IB, available online at http://atvb.ahajournals.org), as well as in the HDL fractions separated by FPLC (Figure IA). In EL–/– mice, apoE and apoA-I peak concentrations overlapped with those of wild-type, whereas in SR-BI–/– and EL–/–/SR-BI–/–, their peaks shifted toward larger particles. This distribution of the apolipoproteins in SR-BI–/– and EL–/–/SR-BI–/– mice suggests the existence of 2 distinct species of HDL particles that are enriched in either apoE or apoA-I.
Effects of EL and SR-BI on the Size Distribution of HDL as Determined by Nuclear Magnetic Resonance
NMR lipoprotein analysis provides sensitive measurement of lipoprotein particle size and concentration directly from unfractionated plasma. Figure 3A showed that there was a progressive increase in the mean diameter of HDL particles in the following order: wild-type, EL–/–, SR–/–, and EL–/–/SR–/–. Therefore, absence of SR-BI had a larger impact on HDL size than did absence of EL, corroborating our results from FPLC analysis (Figure 1). Loss of either EL or SR-BI had little effect on VLDL or LDL particle size (Figure 3A). HDL subclass showed markedly skewed distribution toward the larger particles in EL–/–, SR–/–, and EL–/–/SR–/– mice (Figure 3B and 3C; H1 is the smallest and H7 is the largest subclass). This was apparent whether we compared the absolute particle concentration (Figure 3B) or the relative (percent) distribution (Figure 3C) of individual HDL subclass. Interestingly, the effect of loss of EL on HDL particle concentration differs from that of loss of SR-BI (Figure 3B). The absence of EL significantly increased the absolute amount of HDL particles, whereas the loss of SR-BI greatly reduced HDL particle concentration, which occurred as a result of a marked reduction in small and medium particles (H1 to H4). These data indicate that the significant increase in the large particles (H5 to H7), which individually contained substantially more cholesterol, more than compensated for the reduction in HDL cholesterol that occurred as a result of the loss of the small/medium particles. Additionally, the effect of SR-BI on HDL particle concentration seemed to override the effect of EL in the EL–/–/SR–/– mice. These results provide direct evidence that EL and SB-BI are major determinants of HDL particle size and concentration.
Figure 3. NMR lipoprotein profile analysis and nondenaturing gradient gel electrophoresis (GGE) of HDL particles. NMR analysis of mean HDL, LDL, and VLDL particle size in 4 groups (n=7 to 9) (A), WT, EL–/–, SR–/–, and EL–/–/SR–/–; HDL subclass (B), H1 to H7, concentration in 4 groups, with H1 being the smallest particles in HDL; and HDL subclass size distribution (C) as represented by the percentage of small (H1, H2), medium (H3, H4), and large (H5, H6, and H7) particles. D, GGE of HDL particles isolated by gradient density ultracentrifugation in pooled plasma from 4 groups: WT, EL–/–, SR–/–, and EL–/–/SR–/– (STD, lipoprotein standard of known diameters). E, Size distribution on GGE as measured by densitometry.
We used nondenaturing gradient gel electrophoresis to corroborate our findings on HDL particle size. Results in Figure 3D indicate that rate of migration of HDL particles from EL–/– to SR–/– to EL–/–/SR–/– mice was significantly slower than that of wild-type animals, indicating larger lipoprotein particle sizes. Densitometry analysis of HDL sizes (Figure 3E) yielded information in agreement with results from NMR analysis, although there was a slight variation in the absolute value of mean HDL size, which might be caused by differences in the lipoprotein preparation.
Expression Level of SR-BI in EL-Deficient Mice
We investigated whether there might be feedback and/or compensatory changes in the level of expression of one gene when the other one is absent. By RT-PCR, we found that EL mRNA in the liver was present at a comparable level in SR-BI–/– as in wild-type mice. In contrast, EL mRNA level was increased in the adrenal gland of SR-BI–/– mice, whereas hepatic and adrenal SR-BI mRNA, as well as protein in EL–/– mice, was not different from that of wild-type animals (data not shown, available on request).
Effects of EL and SR-BI on Hepatic Lipase, Phospholipid Transfer Protein, and LCAT
HL, phospholipid transfer protein (PLTP), and LCAT are the other major factors involved in regulation of HDL metabolism. Hepatic HL mRNA was upregulated in EL–/– and downregulated in SR-BI–/– and EL–/–/SR-BI–/– (Figure 4A) mice, whereas hepatic PLTP protein level remained the same in the 4 groups (Figure 4B). Although both hepatic LCAT mRNA (Figure 4A) and plasma protein levels (Figure 4C) were increased in EL–/– mice, the LCAT endogenous esterification rate in these mice was significantly impaired (60%). LCAT protein level was not significantly changed in SR-BI–/– or EL–/–/SR-BI–/– mice, despite a reduced mRNA level in EL–/–/SR-BI–/– mice. Of note is the fact that in both genotypes, the LCAT endogenous esterification rate was markedly reduced (to <10%) in the face of an unchanged LCAT protein level (Figure 4D).
Figure 4. Northern blot analysis (A) of hepatic lipase and LCAT (n=3), Western blot analysis (B) of liver PLTP (n=2), plasma level (1 μL) of LCAT (n=3) (C), and endogenous esterification rate (D) of LCAT in WT, EL–/–, SR–/–, and EL–/–/SR–/– (n=7 to 9). Significant differences: **P<0.01 versus WT; P<0.01 versus EL–/–. Quantitation was expressed as the fold change relative to wild-type group after normalization to ?-actin levels.
Discussion
HDL is an atheroprotective lipoprotein whose metabolism is regulated by multiple factors, including vascular lipolytic enzymes, lipid transfer proteins, cell surface receptors, and transporters.5,21–23 Variations in HDL level in populations are, however, largely determined genetically. Both EL and SR-BI are major determinants of HDL levels, as revealed by transgenic and gene-targeted mouse experiments.9,10,14,16 In this study, we examined the interactions between the 2 genes with respect to their roles in HDL metabolism in vivo. In our study of mice with combinations of EL and SR-BI genotypes, we showed for the first time to our knowledge that EL and SR-BI have distinct and additive effects on HDL concentration, composition, and structure.
The major difference in lipoprotein modification resulting from the action of EL and SR-BI on HDL is on phospholipids. Consistent with the phospholipase function of EL, which mediates the hydrolysis and remodeling of circulating HDL,9,10 EL–/– mice exhibit an increase in plasma phospholipid level. Similarly increased cholesterol level in EL–/– mice is a result of a selective increase in the concentration of HDL particles, especially that of the larger particles (Figure 3B and 3C). HDL particles of EL–/– mice are larger than those of wild-type mice, but significantly smaller than those in SR-BI–/– mice. In contrast, the primary effect of absence of SR-BI on HDL is the markedly larger particle size, with concomitant change in cholesteryl ester content caused by reduced LCAT activity. Thus, despite a concomitant reduction in HDL particle concentration, the HDL cholesterol level in SR-BI–/– mice is increased even higher than that of EL–/– mice.
EL is unique among the vascular lipases in that it is the only lipase synthesized by endothelial cells, whereas both lipoprotein lipase and HL are synthesized by other cell types before they are transported and attached to the surface of the vascular endothelium.24 EL not only shares significant sequence homology with HL (41%) but also has overlapping phospholipase activity. Compared to HL, whose expression is restricted to liver, adrenal, and ovary, EL is also present in lung and placenta. Whereas EL is a specific phospholipase for HDL, HL has both phospholipase and triglyceride lipase activity, and its substrate preference appears to be much broader and includes IDL and chylomicron remnants in addition to HDL.8 Recently, it was reported that EL, like HL, also has lipolysis-independent ligand-bridging function to facilitate HDL uptake.25 Because of the overlapping substrate and lipase activity, the upregulation of HL in the EL–/– mice is likely a result of a compensatory mechanism, although the reason for its decreased expression in SR-BI–/– and EL–/–/SR-BI–/– is unclear.
We and others showed that EL is a major modulator of HDL metabolism,9,10 and others found that SR-BI is the major HDL receptor responsible for the selective uptake in liver and steroidogenic tissues.16 The lack of effect of absence of EL on SR-BI expression in the liver and vice versa indicate that the 2 genes do not directly feed-back on each other, and they participate in independent pathways of HDL metabolism. Their additive HDL-raising effect further supports this conclusion. The presence of EL transcripts in the adrenal gland, the major steroidogenic tissue in rodents, suggests a potential role of EL in steroidogenesis6 related to either its phospholipase activity or a catalysis-independent ligand function, as was recently reported.25 Therefore, despite their independent actions, the 2 genes clearly interact functionally because HDL is their common target.
During the process of reverse cholesterol transport,26 HDL particles go through a series of modifications by lipases and lipid transfer proteins. Two major neutral lipid transfer proteins are involved in the HDL remodeling process, cholesteryl ester transfer protein and PLTP. Mice lack cholesteryl ester transfer protein that transfers cholesteryl ester between HDL particle and VLDL, a key mechanism for the delivery of cholesterol to the liver through the VLDL–LDL pathway in humans.27 So, the lipoprotein alterations observed in mice may not extrapolate directly to those in humans. The major lipid transfer protein between VLDL and HDL in mice is PLTP. PLTP facilitates phospholipids transfer from VLDL particles to HDL, as well as between large and small HDL particles.23,28 It has been shown that PLTP activity is correlated with serum triglyceride level in humans,19 and the remodeling of HDL is enhanced by an increased triglyceride content.28 In the knockout mouse models we studied, the PLTP protein mass and triglyceride level in plasma are similar among the 4 genotypes, suggesting that differences in PLTP activity is not likely to account for the changes in HDL level and composition.
Cholesterol esterification by LCAT in the nascent HDL particle is an important step in the reverse cholesterol transport pathway.26 Small HDL particles, such as HDL3 in human plasma, are optimal substrates for LCAT.20 We previously showed that LCAT activity against the endogenous lipoprotein substrate was reduced in EL–/– mice.9 The markedly altered HDL size and composition associated with absence of SR-BI resulted in >90% inhibition of LCAT activity (Figure 4D). These data indicate that the changes in size and composition of HDL in the absence of EL or SR-BI have a profound effect on LCAT activity. Thus, loss of EL or SR-BI could interfere with reverse cholesterol transport through alteration in HDL structure and function and/or indirectly through inhibition of LCAT activity.
In summary, we found that EL and SR-BI directly affect HDL concentration and composition in vivo. The absence of either gene product individually elevates plasma HDL concentration, and when both are absent, the effects are additive. However, the effect of absence of EL on HDL composition and structure is distinct from that of lack of SR-BI. We conclude that EL and SR-BI are major determinants of HDL, and they exert their influence via distinct and independent actions on HDL metabolism.
Acknowledgments
Research was supported by National Institutes of Health grants HL-51586 (to L.C.), HL-18574 (to T.F.), and the Betty Rutherford Chair from St. Luke’s Episcopal Hospital and Baylor College of Medicine (to L.C.).
Received May 22, 2004; accepted November 1, 2004.
References
Miller GJ, Miller NE. Plasma-high-density-lipoprotein concentration and development of ischaemic heart-disease. Lancet. 1975; 1: 16–19.
Miller NE, Thelle DS, Forde OH, Mjos OD. The Tromso heart-study. High-density lipoprotein and coronary heart-disease: a prospective case-control study. Lancet. 1977; 1: 965–968.
Gordon T, Castelli WP, Hjortland MC, Kannel WB, Dawber TR. High density lipoprotein as a protective factor against coronary heart disease. The Framingham Study. Am J Med. 1977; 62: 707–714.
Heller DA, de Faire U, Pedersen NL, Dahlen G, McClearn GE. Genetic and environmental influences on serum lipid levels in twins. N Engl J Med. 1993; 328: 1150–1156.
Tall AR. Metabolic and genetic control of HDL cholesterol levels. J Intern Med. 1992; 231: 661–668.
Hirata K, Dichek HL, Cioffi JA, Choi SY, Leeper NJ, Quintana L, Kronmal GS, Cooper AD, Quertermous T. Cloning of a unique lipase from endothelial cells extends the lipase gene family. J Biol Chem. 1999; 274: 14170–14175.
Jaye M, Lynch KJ, Krawiec J, Marchadier D, Maugeais C, Doan K, South V, Amin D, Perrone M, Rader DJ. A novel endothelial-derived lipase that modulates HDL metabolism. Nat Genet. 1999; 21: 424–428.
McCoy MG, Sun GS, Marchadier D, Maugeais C, Glick JM, Rader DJ. Characterization of the lipolytic activity of endothelial lipase. J Lipid Res. 2002; 43: 921–929.
Ma K, Cilingiroglu M, Otvos JD, Ballantyne CM, Marian AJ, Chan L. Endothelial lipase is a major genetic determinant for high-density lipoprotein concentration, structure, and metabolism. Proc Natl Acad Sci U S A. 2003; 100: 2748–2753.
Ishida T, Choi S, Kundu RK, Hirata K, Rubin EM, Cooper AD, Quertermous T. Endothelial lipase is a major determinant of HDL level. J Clin Invest. 2003; 111: 347–355.
Acton S, Rigotti A, Landschulz KT, Xu S, Hobbs HH, Krieger M. Identification of scavenger receptor SR-BI as a high density lipoprotein receptor. Science. 1996; 271: 518–520.
Glass CK, Pittman RC, Keller GA, Steinberg D. Tissue sites of degradation of apoprotein A-I in the rat. J Biol Chem. 1983; 258: 7161–7167.
Gwynne JT, Mahaffee DD. Rat adrenal uptake and metabolism of high density lipoprotein cholesteryl ester. J Biol Chem. 1989; 264: 8141–8150.
Kozarsky KF, Donahee MH, Rigotti A, Iqbal SN, Edelman ER, Krieger M. Overexpression of the HDL receptor SR-BI alters plasma HDL and bile cholesterol levels. Nature. 1997; 387: 414–417.
Wang N, Arai T, Ji Y, Rinninger F, Tall AR. Liver-specific overexpression of scavenger receptor BI decreases levels of very low density lipoprotein ApoB, low density lipoprotein ApoB, and high density lipoprotein in transgenic mice. J Biol Chem. 1998; 273: 32920–32926.
Rigotti A, Trigatti BL, Penman M, Rayburn H, Herz J, Krieger M. A targeted mutation in the murine gene encoding the high density lipoprotein (HDL) receptor scavenger receptor class B type I reveals its key role in HDL metabolism. Proc Natl Acad Sci U S A. 1997; 94: 12610–12615.
Traber MG, Kayden HJ, Rindler MJ. Polarized secretion of newly synthesized lipoproteins by the Caco-2 human intestinal cell line. J Lipid Res. 1987; 28: 1350–1363.
Nichols AV, Krauss RM, Musliner TA. Nondenaturing polyacrylamide gradient gel electrophoresis. Methods Enzymol. 1986; 128: 417–431.
Huuskonen J, Ekstrom M, Tahvanainen E, Vainio A, Metso J, Pussinen P, Ehnholm C, Olkkonen VM, Jauhiainen M. Quantification of human plasma phospholipid transfer protein (PLTP): relationship between PLTP mass and phospholipid transfer activity. Atherosclerosis. 2000; 151: 451–461.
Barter PJ, Hopkins GJ, Gorjatschko L. Lipoprotein substrates for plasma cholesterol esterification. Influence of particle size and composition of the high density lipoprotein subfraction 3. Atherosclerosis. 1985; 58: 97–107.
Bodzioch M, Orso E, Klucken J, Langmann T, Bottcher A, Diederich W, Drobnik W, Barlage S, Buchler C, Porsch-Ozcurumez M, Kaminski WE, Hahmann HW, Oette K, Rothe G, Aslanidis C, Lackner KJ, Schmitz G. The gene encoding ATP-binding cassette transporter 1 is mutated in Tangier disease. Nat Genet. 1999; 22: 347–351.
Jauhiainen M, Metso J, Pahlman R, Blomqvist S, van Tol A, Ehnholm C. Human plasma phospholipid transfer protein causes high density lipoprotein conversion. J Biol Chem. 1993; 268: 4032–4036.
Jiang XC, Bruce C, Mar J, Lin M, Ji Y, Francone OL, Tall AR. Targeted mutation of plasma phospholipid transfer protein gene markedly reduces high-density lipoprotein levels. J Clin Invest. 1999; 103: 907–914.
Rader DJ, Jaye M. Endothelial lipase: a new member of the triglyceride lipase gene family. Curr Opin Lipidol. 2000; 11: 141–147.
Broedl UC, Maugeais C, Marchadier D, Glick JM, Rader DJ. Effects of nonlipolytic ligand function of endothelial lipase on HDL metabolism in vivo. J Biol Chem. 2003; 278: 40688–40693.
Fielding CJ, Fielding PE. Molecular physiology of reverse cholesterol transport. J Lipid Res. 1995; 36: 211–228.
Tollefson JH, Ravnik S, Albers JJ. Isolation and characterization of a phospholipid transfer protein (LTP- II) from human plasma. J Lipid Res. 1988; 29: 1593–1602.
Rye KA, Jauhiainen M, Barter PJ, Ehnholm C. Triglyceride-enrichment of high density lipoproteins enhances their remodelling by phospholipid transfer protein. J Lipid Res. 1998; 39: 613–622.(Ke Ma; Trudy Forte; James)
Correspondence to Lawrence Chan, Division of Endocrinology & Metabolism, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030. E-mail lchan@bcm.tmc.edu
Abstract
Objective— Endothelial lipase (EL) is a vascular phospholipase that hydrolyzes high-density lipoprotein (HDL) as its preferred substrate. Scavenger receptor-class B type I (SR-BI) is an HDL receptor that mediates the selective uptake of cholesteryl ester. This study investigates the role of EL and SR-BI in the regulation of HDL metabolism in gene knockout mouse models.
Methods and Results— We cross-bred EL–/– and SR-BI–/– mice and generated single- and double-null mice. We used biochemical, molecular biology, and nuclear magnetic resonance methods to analyze HDL concentration, composition, and structure. We found that EL and SR-BI display additive effects on HDL with evident gene dosage effects, but their mechanisms to regulate HDL concentration and composition are different. Whereas the elevated HDL cholesterol level in EL–/– mice is associated with increased phospholipid content in HDL particles, SR-BI–/– mice display markedly enlarged HDL particles shifted to larger subclasses with a phospholipid content similar to that of wild-type mice. Furthermore, absence of EL is associated with a 40% to 50% inhibition and absence of SR-BI, a 90% inhibition of endogenous lecithin cholesterol:acyltransferase rate.
Conclusions— EL and SR-BI are major genetic determinants of HDL metabolism in vivo, each exercising independent and additive effects on HDL structure and function.
We created EL and SR-BI single- and double-null mice and compared their effects on HDL concentration, composition, and structure. We demonstrated that EL and SR-BI display additive effects on plasma HDL in vivo with evident gene dosage effects. The 2 genes use different mechanisms to regulate HDL concentration and composition.
Key Words: endothelial lipase ? scavenger receptor-class B type I ? high-density lipoprotein
Introduction
Numerous epidemiological studies have established a negative correlation between high-density lipoprotein (HDL) levels and atherosclerosis development.1–3 HDL concentration is largely determined genetically.4,5 Multiple genes are directly involved in the HDL metabolic pathways,5 although other genetic factors influencing HDL level exist that are poorly understood.
Endothelial lipase (EL) is a newly described member of the vascular lipase family, sharing sequence homology with lipoprotein lipase and hepatic lipase (HL).6,7 EL possesses predominantly phospholipase activity and uses HDL as its preferred substrate.8 We recently showed that absence of EL in mice leads to delayed HDL clearance and elevated HDL levels, and that a common single-nucleotide polymorphism in the human EL gene (581 C/T) is associated with significantly higher HDL cholesterol levels.9 Similar findings were also reported on EL-null mice created independently by Ishida et al.10
Scavenger receptor-class B type I (SR-BI) is the major HDL receptor responsible for its selective uptake in liver and steroidogenic tissues.11 Selective uptake, a process that involves the transfer of cholesteryl ester from the lipoprotein core into cells without internalization of the whole lipoprotein particle, accounts for up to 70% of the HDL delivery in the liver12 and 90% of the cholesterol used for steroid hormone production in the adrenal gland in rodents.13 Overexpression of SR-BI in mice by adenovirus-mediated gene transfer14 or by a transgenic approach15 led to a marked reduction of plasma HDL levels, whereas targeted disruption of SR-BI led to markedly increase plasma HDL cholesterol.16
In this study, we used knockout mouse models to examine gene–gene interaction between EL and SR-BI, because both play dominant roles in HDL metabolism. We found that the effect of loss of EL on HDL composition and subclass size distribution was different from that of loss of SR-BI, although their individual effects on HDL level were additive. The 2 genes exert independent influence on HDL size, composition, and lecithin cholesteryl acyltransferase (LCAT) activity in vivo. Our findings have significant implications for the genetic regulation of the concentration and metabolism of HDL.
Methods
Animals
EL–/– mice were created by gene targeting9 and backcrossed to C57BL/6 for 4 generations. Mice were weaned at 21 days and fed a standard laboratory chow (7001; Teklad, Madison, Wis). SR-BI+/– mice were purchased from Jackson Laboratory (Bar Harbor, Me)16 and bred to homozygosity in a mixed 129/C57BL/6J background. EL–/–/SR-BI–/– mice were obtained by crossing EL–/–/SR-BI–/+ breeding pairs. Genotypes of EL–/– was confirmed by polymerase chain reaction (PCR) using the wild-type allele-specific primers: EL142 (5'-AGCTTGAAGGGTGACTTGAG-3') and EL18 (5'-CCTTCATGATTGTTCTTCAC-3'), and the targeted allele-specific primers: EL142 and Neo5 (5'-CTATCGCCTTCTTGACGAGT-3'). SR-BI genotyping was performed by PCR16 using primers SR-1, 2, and 3. All PCR reactions were performed at annealing temperature of 58°C and 32 cycles.
Plasma Lipids, Lipoproteins, and Lipoprotein Particle Chemical Composition Analysis
Cholesterol, phospholipids, triglyceride, HDL cholesterol, total cholesterol, and free cholesterol concentrations were assayed using commercial kits (Wako Chemicals, Richmond, Ga). For fast protein liquid chromatography (FPLC) analysis, plasma pooled from 5 to 7 mice (200 μL) after a 5-hour fast was loaded onto 2 Superose-6 columns connected in series (Amersham Pharmacia Biotech). Fractions (0.5 mL) were collected: fractions 4 to 9, very-low-density lipoprotein (VLDL); 16 to 22, intermediate-density lipoprotein (IDL)/low-density lipoprotein (LDL); 26 to 38, HDL.
Chylomicron/VLDL, density <1.006; IDL/LDL, 1.006
Lipoprotein Nuclear Magnetic Resonance Analysis
Lipoprotein subclass profiles were measured on frozen plasma (100 μL) by proton nuclear magnetic resonance (NMR) spectroscopy at LipoScience, Inc (Raleigh, NC) as previously described.9 The designations of the NMR-derived HDL subclasses and their estimated diameter ranges are as follows: H1 (7.3 to 7.7 nm), H2 (7.8 to 8.4 nm), H3 (8.5 to 8.9 nm), H4 (9.0 to 10.0 nm), H5 (10.1 to 12.0 nm), H6 (12.1 to 13.5), and H7 (13.6 to 15.0). HDL subclass particle concentrations were expressed in particle numbers (nmol/L) as derived from raw NMR data.9 Mean HDL particle size was determined by weighting the relative percentage of each subclass by its diameter.
Nondenaturing Polyacrylamide Gel Electrophoresis
Nondenaturing 4% to 30% polyacrylamide gradient gels were used to analyze the size distribution of HDL as described by Nichols et al18 using HDL with 1.063
Immunoblotting, Northern Blotting, and Semi-Quantitative Reverse-Transcription PCR
Immunoblotting of SR-BI and phospholipid transfer protein (PLTP) was performed as described16 using 40 μg of total protein in lysis buffer containing 50 mmol/L Tris-HCl, pH 7.4, 1% NP-40, 0.25% sodium deoxycholate, 150 mmol/L NaCl, 1 mmol/L EDTA, 1 mmol/L PMSF, 1 μg/mL aprotinin, leupeptin, pepstatin; 1 mmol/L Na3VO4, and 1 mM NaF. Proteins were transferred to nitrocellulose membrane after electrophoresis and immunoblots were developed by chemiluminescence detection (Supersignal, Pierce). Antibodies used were: goat anti-human apolipoprotein A-I, E antibodies, and mouse anti-actin monoclonal antibody (Chemicon), and rabbit anti-SR-BI and anti-PLTP polyclonal antibody (Novas Biologicals, Littleton, Co). Anti–SR-BI antibody recognizes the 82-Kd SR-BI protein, and the anti-human PLTP antibody cross-reacts with the mouse PLTP protein of 75 Kd.19 Polyclonal anti-mouse LCAT antibody was a gift from Dr John Parks at Wake Forest University (Winston-Salem, NC). RNA was extracted by Trizol reagent (Invitrogen). For Northern blotting, 20 μg of total hepatic RNA was separated on 0.2 mol/L formaldehyde 1.2% agarose gel, transferred to nylon membrane, and hybridized with gene-specific cDNA probes as indicated. Membranes were developed using a phospho-imager (Cyclone; PerkinElmer Life Sciences). Quantitation was performed by densitometry scanning using -imager LabWorks image acquisition and analysis software (UVP Inc, Upland, Calif). For semi-quantitative reverse-transcription (RT)-PCR, PCR reaction was performed using the same amount of RT product with EL, SR-BI, and GAPDH, and monitored at PCR cycles from 24 to 36.
LCAT Activity Assay
Plasma LCAT activity against endogenous substrate was measured using lipoproteins in whole plasma (50 μL) as substrate as described.9,20 [1, 2(n)-3H]-cholesterol (1.0 mCi/mL) was from Amersham Pharmacia Biotech (Buckinghamshire, England). Briefly, plasma was pre-incubated with 10 μL of [1, 2(n)-3H]-cholesterol bovine serum albumin solution (5 μCi/mL) at 4°C for 4 hours before incubating at 37°C to start the reaction. Reaction was stopped by addition of 2 mL chloroform:methanol (2:1) after 30 minutes and the lower-phase chloroform was extracted. Thin-layer chromatography was then performed and the radiolabels quantified by scintillation counting (Beckman Instrument). Results were expressed as the percentage of cholesterol esterified.
Statistical Analysis
Values were expressed as mean±SD. The differences between different genotypes were calculated by the 2-tailed Student t test. P<0.05 is considered statistically significant.
Results
Differential Effects of EL and SR-BI on Plasma Lipid Level and Lipoprotein Composition
Corroborating and extending previous reports,9,10,16 we found that absence of EL or SR-BI led to a significant increase in plasma total cholesterol level. The effect of lack of SR-BI was greater than that resulting from a lack of EL, 242% versus 186%, respectively, compared with wild-type (Table). In both cases, the elevated total cholesterol was associated with increased HDL cholesterol level, which accounts for >75% of the increase in cholesterol in EL–/– and >65% in SR-BI–/– samples. The effects of EL and SR-BI on cholesterol and HDL cholesterol were additive. Plasma total cholesterol level of the EL–/–/SR-BI–/– mice was 335% of wild-type controls and their HDL cholesterol level, 367% of wild-type, significantly higher than that in either EL–/– or SR-BI–/– groups. Thus, both EL and SR-BI exert major influence on plasma cholesterol level, primarily through their effect on HDL. In contrast, phospholipid level was significantly elevated in EL–/– mice but remained unchanged in SR-BI–/–, with the EL–/–/SR-BI–/– mice displaying a phospholipid level comparable to that of the EL–/–. These data suggest that only EL, not SR-BI, directly modulates phospholipid concentration in vivo.
Plasma Lipid Levels of Different Genotypes of Mice
Our analysis of cholesterol and phospholipid levels in FPLC fractions reflected the changes in plasma concentrations (Figure 1A and 1C). Compared with wild-type, cholesterol concentration in HDL fractions increased progressively in EL–/–, SR–/–, and EL–/–/SR–/–, although the phospholipid levels were elevated only in mice lacking EL. There was a minimally higher triglyceride level in these mice compare to the wild-type in the VLDL fraction (Figure 1D) in this experiment, which is not evident in other analyses. Free cholesterol level in both SR–/– and EL–/–/SR–/– mice was markedly higher than that in either wild-type or EL–/– mice (Figure 1B), suggesting an increased free cholesterol content of HDL particles in mice lacking SR-BI. A consistent finding was that whereas the HDL peak of EL–/– mice largely overlapped with that of the wild-type at 34 fraction, it shifted to 28 in both SR–/– and EL–/–/SR–/–, indicating the presence of much larger HDL particles. Thus, SR-BI exerts a much greater influence on HDL particle size than does EL.
Figure 1. Lipid analysis of FPLC fractionation of mice in 4 genotypes: WT, EL–/–, SR–/–, and EL–/–/SR–/– on regular chow diet. Cholesterol (A), free cholesterol (B), phospholipid (C), and triglyceride (D) levels in FPLC fractions in pooled plasma samples from each group of mice (n=7 to 10).
We analyzed lipid composition of the lipoproteins isolated by density gradient ultracentrifugation. There was no significant difference in the lipid composition of the VLDL and IDL/LDL fractions (data not shown), but it differed markedly in HDL species from the 4 genotypes. As shown in Figure 2A and 2B, there was a progressive decrease in the protein content of the HDL from wild-type to EL–/– and to SR-BI–/– with or without EL inactivation. Although the relative amount of total (free plus esterified) cholesterol remained fairly constant, the ratio of free to esterified cholesterol changed, being highest in SR-BI–/– mice with or without concomitant EL inactivation (6-fold that of wild-type or EL–/– alone). The reduced level of esterified cholesterol and markedly higher free-to-esterified cholesterol ratio of these animals (Figure 2A and 2B) could be a result of reduced LCAT activity (vide infra).
Figure 2. Chemical composition analysis of HDL separated by density ultracentrifugation in 4 groups of mice using pooled samples from 7 to 9 mice each group. A, Chemical composition in 4 genotypes of mice as represented by the percentage of each component in the HDL particle. B, Schematic representation of the chemical composition of HDL particles in 4 groups of mice. Note that the percent of triglyceride is too small to allow it to be discernible in the figure. C, Immunoblot analysis of apoA-I and apoE in VLDL (V), LDL (L), and HDL (H) fractions separated by ultracentrifugation method in 4 different groups.
FPLC analysis revealed the presence of significantly larger HDL particles in SR-BI–/– and EL–/–/SR-BI–/– mice, which might have exhibited buoyant densities that moved to a lower density range. We therefore checked for the presence of apoA-I, the major HDL apolipoprotein, and apoE, known to be present in multiple lipoprotein species, in VLDL, LDL, and HDL fractions. As shown in Figure 2C, apoA-I was present only in the HDL of wild-type and EL–/– mice, whereas apoE was detected in all lipoprotein fractions. Interestingly, in SR-BI–/– and EL–/–/SR-BI–/– mice, we found apoA-I in both the HDL density and the LDL density fractions. The presence of apoA-I in the LDL density fraction of SR-BI–/– mice, whether they carried the EL–/– alleles, indicates that there was "spillover" of the large HDL species into the IDL/LDL density range during isolation by ultracentrifugation.
Additive Effects of EL and SR-BI on HDL Apolipoprotein Levels
Our results revealed that effects of EL and SR-BI on HDL concentration were additive. We next analyzed their effects on HDL-associated apolipoproteins (see http://www.atvb.ahajournals.org). Absence of EL and SR-BI increased the level of the HDL-associated apolipoproteins, apoA-I and apoE, in plasma (Figure IB, available online at http://atvb.ahajournals.org), as well as in the HDL fractions separated by FPLC (Figure IA). In EL–/– mice, apoE and apoA-I peak concentrations overlapped with those of wild-type, whereas in SR-BI–/– and EL–/–/SR-BI–/–, their peaks shifted toward larger particles. This distribution of the apolipoproteins in SR-BI–/– and EL–/–/SR-BI–/– mice suggests the existence of 2 distinct species of HDL particles that are enriched in either apoE or apoA-I.
Effects of EL and SR-BI on the Size Distribution of HDL as Determined by Nuclear Magnetic Resonance
NMR lipoprotein analysis provides sensitive measurement of lipoprotein particle size and concentration directly from unfractionated plasma. Figure 3A showed that there was a progressive increase in the mean diameter of HDL particles in the following order: wild-type, EL–/–, SR–/–, and EL–/–/SR–/–. Therefore, absence of SR-BI had a larger impact on HDL size than did absence of EL, corroborating our results from FPLC analysis (Figure 1). Loss of either EL or SR-BI had little effect on VLDL or LDL particle size (Figure 3A). HDL subclass showed markedly skewed distribution toward the larger particles in EL–/–, SR–/–, and EL–/–/SR–/– mice (Figure 3B and 3C; H1 is the smallest and H7 is the largest subclass). This was apparent whether we compared the absolute particle concentration (Figure 3B) or the relative (percent) distribution (Figure 3C) of individual HDL subclass. Interestingly, the effect of loss of EL on HDL particle concentration differs from that of loss of SR-BI (Figure 3B). The absence of EL significantly increased the absolute amount of HDL particles, whereas the loss of SR-BI greatly reduced HDL particle concentration, which occurred as a result of a marked reduction in small and medium particles (H1 to H4). These data indicate that the significant increase in the large particles (H5 to H7), which individually contained substantially more cholesterol, more than compensated for the reduction in HDL cholesterol that occurred as a result of the loss of the small/medium particles. Additionally, the effect of SR-BI on HDL particle concentration seemed to override the effect of EL in the EL–/–/SR–/– mice. These results provide direct evidence that EL and SB-BI are major determinants of HDL particle size and concentration.
Figure 3. NMR lipoprotein profile analysis and nondenaturing gradient gel electrophoresis (GGE) of HDL particles. NMR analysis of mean HDL, LDL, and VLDL particle size in 4 groups (n=7 to 9) (A), WT, EL–/–, SR–/–, and EL–/–/SR–/–; HDL subclass (B), H1 to H7, concentration in 4 groups, with H1 being the smallest particles in HDL; and HDL subclass size distribution (C) as represented by the percentage of small (H1, H2), medium (H3, H4), and large (H5, H6, and H7) particles. D, GGE of HDL particles isolated by gradient density ultracentrifugation in pooled plasma from 4 groups: WT, EL–/–, SR–/–, and EL–/–/SR–/– (STD, lipoprotein standard of known diameters). E, Size distribution on GGE as measured by densitometry.
We used nondenaturing gradient gel electrophoresis to corroborate our findings on HDL particle size. Results in Figure 3D indicate that rate of migration of HDL particles from EL–/– to SR–/– to EL–/–/SR–/– mice was significantly slower than that of wild-type animals, indicating larger lipoprotein particle sizes. Densitometry analysis of HDL sizes (Figure 3E) yielded information in agreement with results from NMR analysis, although there was a slight variation in the absolute value of mean HDL size, which might be caused by differences in the lipoprotein preparation.
Expression Level of SR-BI in EL-Deficient Mice
We investigated whether there might be feedback and/or compensatory changes in the level of expression of one gene when the other one is absent. By RT-PCR, we found that EL mRNA in the liver was present at a comparable level in SR-BI–/– as in wild-type mice. In contrast, EL mRNA level was increased in the adrenal gland of SR-BI–/– mice, whereas hepatic and adrenal SR-BI mRNA, as well as protein in EL–/– mice, was not different from that of wild-type animals (data not shown, available on request).
Effects of EL and SR-BI on Hepatic Lipase, Phospholipid Transfer Protein, and LCAT
HL, phospholipid transfer protein (PLTP), and LCAT are the other major factors involved in regulation of HDL metabolism. Hepatic HL mRNA was upregulated in EL–/– and downregulated in SR-BI–/– and EL–/–/SR-BI–/– (Figure 4A) mice, whereas hepatic PLTP protein level remained the same in the 4 groups (Figure 4B). Although both hepatic LCAT mRNA (Figure 4A) and plasma protein levels (Figure 4C) were increased in EL–/– mice, the LCAT endogenous esterification rate in these mice was significantly impaired (60%). LCAT protein level was not significantly changed in SR-BI–/– or EL–/–/SR-BI–/– mice, despite a reduced mRNA level in EL–/–/SR-BI–/– mice. Of note is the fact that in both genotypes, the LCAT endogenous esterification rate was markedly reduced (to <10%) in the face of an unchanged LCAT protein level (Figure 4D).
Figure 4. Northern blot analysis (A) of hepatic lipase and LCAT (n=3), Western blot analysis (B) of liver PLTP (n=2), plasma level (1 μL) of LCAT (n=3) (C), and endogenous esterification rate (D) of LCAT in WT, EL–/–, SR–/–, and EL–/–/SR–/– (n=7 to 9). Significant differences: **P<0.01 versus WT; P<0.01 versus EL–/–. Quantitation was expressed as the fold change relative to wild-type group after normalization to ?-actin levels.
Discussion
HDL is an atheroprotective lipoprotein whose metabolism is regulated by multiple factors, including vascular lipolytic enzymes, lipid transfer proteins, cell surface receptors, and transporters.5,21–23 Variations in HDL level in populations are, however, largely determined genetically. Both EL and SR-BI are major determinants of HDL levels, as revealed by transgenic and gene-targeted mouse experiments.9,10,14,16 In this study, we examined the interactions between the 2 genes with respect to their roles in HDL metabolism in vivo. In our study of mice with combinations of EL and SR-BI genotypes, we showed for the first time to our knowledge that EL and SR-BI have distinct and additive effects on HDL concentration, composition, and structure.
The major difference in lipoprotein modification resulting from the action of EL and SR-BI on HDL is on phospholipids. Consistent with the phospholipase function of EL, which mediates the hydrolysis and remodeling of circulating HDL,9,10 EL–/– mice exhibit an increase in plasma phospholipid level. Similarly increased cholesterol level in EL–/– mice is a result of a selective increase in the concentration of HDL particles, especially that of the larger particles (Figure 3B and 3C). HDL particles of EL–/– mice are larger than those of wild-type mice, but significantly smaller than those in SR-BI–/– mice. In contrast, the primary effect of absence of SR-BI on HDL is the markedly larger particle size, with concomitant change in cholesteryl ester content caused by reduced LCAT activity. Thus, despite a concomitant reduction in HDL particle concentration, the HDL cholesterol level in SR-BI–/– mice is increased even higher than that of EL–/– mice.
EL is unique among the vascular lipases in that it is the only lipase synthesized by endothelial cells, whereas both lipoprotein lipase and HL are synthesized by other cell types before they are transported and attached to the surface of the vascular endothelium.24 EL not only shares significant sequence homology with HL (41%) but also has overlapping phospholipase activity. Compared to HL, whose expression is restricted to liver, adrenal, and ovary, EL is also present in lung and placenta. Whereas EL is a specific phospholipase for HDL, HL has both phospholipase and triglyceride lipase activity, and its substrate preference appears to be much broader and includes IDL and chylomicron remnants in addition to HDL.8 Recently, it was reported that EL, like HL, also has lipolysis-independent ligand-bridging function to facilitate HDL uptake.25 Because of the overlapping substrate and lipase activity, the upregulation of HL in the EL–/– mice is likely a result of a compensatory mechanism, although the reason for its decreased expression in SR-BI–/– and EL–/–/SR-BI–/– is unclear.
We and others showed that EL is a major modulator of HDL metabolism,9,10 and others found that SR-BI is the major HDL receptor responsible for the selective uptake in liver and steroidogenic tissues.16 The lack of effect of absence of EL on SR-BI expression in the liver and vice versa indicate that the 2 genes do not directly feed-back on each other, and they participate in independent pathways of HDL metabolism. Their additive HDL-raising effect further supports this conclusion. The presence of EL transcripts in the adrenal gland, the major steroidogenic tissue in rodents, suggests a potential role of EL in steroidogenesis6 related to either its phospholipase activity or a catalysis-independent ligand function, as was recently reported.25 Therefore, despite their independent actions, the 2 genes clearly interact functionally because HDL is their common target.
During the process of reverse cholesterol transport,26 HDL particles go through a series of modifications by lipases and lipid transfer proteins. Two major neutral lipid transfer proteins are involved in the HDL remodeling process, cholesteryl ester transfer protein and PLTP. Mice lack cholesteryl ester transfer protein that transfers cholesteryl ester between HDL particle and VLDL, a key mechanism for the delivery of cholesterol to the liver through the VLDL–LDL pathway in humans.27 So, the lipoprotein alterations observed in mice may not extrapolate directly to those in humans. The major lipid transfer protein between VLDL and HDL in mice is PLTP. PLTP facilitates phospholipids transfer from VLDL particles to HDL, as well as between large and small HDL particles.23,28 It has been shown that PLTP activity is correlated with serum triglyceride level in humans,19 and the remodeling of HDL is enhanced by an increased triglyceride content.28 In the knockout mouse models we studied, the PLTP protein mass and triglyceride level in plasma are similar among the 4 genotypes, suggesting that differences in PLTP activity is not likely to account for the changes in HDL level and composition.
Cholesterol esterification by LCAT in the nascent HDL particle is an important step in the reverse cholesterol transport pathway.26 Small HDL particles, such as HDL3 in human plasma, are optimal substrates for LCAT.20 We previously showed that LCAT activity against the endogenous lipoprotein substrate was reduced in EL–/– mice.9 The markedly altered HDL size and composition associated with absence of SR-BI resulted in >90% inhibition of LCAT activity (Figure 4D). These data indicate that the changes in size and composition of HDL in the absence of EL or SR-BI have a profound effect on LCAT activity. Thus, loss of EL or SR-BI could interfere with reverse cholesterol transport through alteration in HDL structure and function and/or indirectly through inhibition of LCAT activity.
In summary, we found that EL and SR-BI directly affect HDL concentration and composition in vivo. The absence of either gene product individually elevates plasma HDL concentration, and when both are absent, the effects are additive. However, the effect of absence of EL on HDL composition and structure is distinct from that of lack of SR-BI. We conclude that EL and SR-BI are major determinants of HDL, and they exert their influence via distinct and independent actions on HDL metabolism.
Acknowledgments
Research was supported by National Institutes of Health grants HL-51586 (to L.C.), HL-18574 (to T.F.), and the Betty Rutherford Chair from St. Luke’s Episcopal Hospital and Baylor College of Medicine (to L.C.).
Received May 22, 2004; accepted November 1, 2004.
References
Miller GJ, Miller NE. Plasma-high-density-lipoprotein concentration and development of ischaemic heart-disease. Lancet. 1975; 1: 16–19.
Miller NE, Thelle DS, Forde OH, Mjos OD. The Tromso heart-study. High-density lipoprotein and coronary heart-disease: a prospective case-control study. Lancet. 1977; 1: 965–968.
Gordon T, Castelli WP, Hjortland MC, Kannel WB, Dawber TR. High density lipoprotein as a protective factor against coronary heart disease. The Framingham Study. Am J Med. 1977; 62: 707–714.
Heller DA, de Faire U, Pedersen NL, Dahlen G, McClearn GE. Genetic and environmental influences on serum lipid levels in twins. N Engl J Med. 1993; 328: 1150–1156.
Tall AR. Metabolic and genetic control of HDL cholesterol levels. J Intern Med. 1992; 231: 661–668.
Hirata K, Dichek HL, Cioffi JA, Choi SY, Leeper NJ, Quintana L, Kronmal GS, Cooper AD, Quertermous T. Cloning of a unique lipase from endothelial cells extends the lipase gene family. J Biol Chem. 1999; 274: 14170–14175.
Jaye M, Lynch KJ, Krawiec J, Marchadier D, Maugeais C, Doan K, South V, Amin D, Perrone M, Rader DJ. A novel endothelial-derived lipase that modulates HDL metabolism. Nat Genet. 1999; 21: 424–428.
McCoy MG, Sun GS, Marchadier D, Maugeais C, Glick JM, Rader DJ. Characterization of the lipolytic activity of endothelial lipase. J Lipid Res. 2002; 43: 921–929.
Ma K, Cilingiroglu M, Otvos JD, Ballantyne CM, Marian AJ, Chan L. Endothelial lipase is a major genetic determinant for high-density lipoprotein concentration, structure, and metabolism. Proc Natl Acad Sci U S A. 2003; 100: 2748–2753.
Ishida T, Choi S, Kundu RK, Hirata K, Rubin EM, Cooper AD, Quertermous T. Endothelial lipase is a major determinant of HDL level. J Clin Invest. 2003; 111: 347–355.
Acton S, Rigotti A, Landschulz KT, Xu S, Hobbs HH, Krieger M. Identification of scavenger receptor SR-BI as a high density lipoprotein receptor. Science. 1996; 271: 518–520.
Glass CK, Pittman RC, Keller GA, Steinberg D. Tissue sites of degradation of apoprotein A-I in the rat. J Biol Chem. 1983; 258: 7161–7167.
Gwynne JT, Mahaffee DD. Rat adrenal uptake and metabolism of high density lipoprotein cholesteryl ester. J Biol Chem. 1989; 264: 8141–8150.
Kozarsky KF, Donahee MH, Rigotti A, Iqbal SN, Edelman ER, Krieger M. Overexpression of the HDL receptor SR-BI alters plasma HDL and bile cholesterol levels. Nature. 1997; 387: 414–417.
Wang N, Arai T, Ji Y, Rinninger F, Tall AR. Liver-specific overexpression of scavenger receptor BI decreases levels of very low density lipoprotein ApoB, low density lipoprotein ApoB, and high density lipoprotein in transgenic mice. J Biol Chem. 1998; 273: 32920–32926.
Rigotti A, Trigatti BL, Penman M, Rayburn H, Herz J, Krieger M. A targeted mutation in the murine gene encoding the high density lipoprotein (HDL) receptor scavenger receptor class B type I reveals its key role in HDL metabolism. Proc Natl Acad Sci U S A. 1997; 94: 12610–12615.
Traber MG, Kayden HJ, Rindler MJ. Polarized secretion of newly synthesized lipoproteins by the Caco-2 human intestinal cell line. J Lipid Res. 1987; 28: 1350–1363.
Nichols AV, Krauss RM, Musliner TA. Nondenaturing polyacrylamide gradient gel electrophoresis. Methods Enzymol. 1986; 128: 417–431.
Huuskonen J, Ekstrom M, Tahvanainen E, Vainio A, Metso J, Pussinen P, Ehnholm C, Olkkonen VM, Jauhiainen M. Quantification of human plasma phospholipid transfer protein (PLTP): relationship between PLTP mass and phospholipid transfer activity. Atherosclerosis. 2000; 151: 451–461.
Barter PJ, Hopkins GJ, Gorjatschko L. Lipoprotein substrates for plasma cholesterol esterification. Influence of particle size and composition of the high density lipoprotein subfraction 3. Atherosclerosis. 1985; 58: 97–107.
Bodzioch M, Orso E, Klucken J, Langmann T, Bottcher A, Diederich W, Drobnik W, Barlage S, Buchler C, Porsch-Ozcurumez M, Kaminski WE, Hahmann HW, Oette K, Rothe G, Aslanidis C, Lackner KJ, Schmitz G. The gene encoding ATP-binding cassette transporter 1 is mutated in Tangier disease. Nat Genet. 1999; 22: 347–351.
Jauhiainen M, Metso J, Pahlman R, Blomqvist S, van Tol A, Ehnholm C. Human plasma phospholipid transfer protein causes high density lipoprotein conversion. J Biol Chem. 1993; 268: 4032–4036.
Jiang XC, Bruce C, Mar J, Lin M, Ji Y, Francone OL, Tall AR. Targeted mutation of plasma phospholipid transfer protein gene markedly reduces high-density lipoprotein levels. J Clin Invest. 1999; 103: 907–914.
Rader DJ, Jaye M. Endothelial lipase: a new member of the triglyceride lipase gene family. Curr Opin Lipidol. 2000; 11: 141–147.
Broedl UC, Maugeais C, Marchadier D, Glick JM, Rader DJ. Effects of nonlipolytic ligand function of endothelial lipase on HDL metabolism in vivo. J Biol Chem. 2003; 278: 40688–40693.
Fielding CJ, Fielding PE. Molecular physiology of reverse cholesterol transport. J Lipid Res. 1995; 36: 211–228.
Tollefson JH, Ravnik S, Albers JJ. Isolation and characterization of a phospholipid transfer protein (LTP- II) from human plasma. J Lipid Res. 1988; 29: 1593–1602.
Rye KA, Jauhiainen M, Barter PJ, Ehnholm C. Triglyceride-enrichment of high density lipoproteins enhances their remodelling by phospholipid transfer protein. J Lipid Res. 1998; 39: 613–622.(Ke Ma; Trudy Forte; James)