Molecular Mechanism for Changes in Proteoglycan Binding on Compositional Changes of the Core and the Surface of Low-Density Lipoprotein–Cont
http://www.100md.com
《动脉硬化血栓血管生物学》
From the Wallenberg Laboratory for Cardiovascular Research (C.F., M.G., J.B.), The Sahlgrenska Academy at G?teborg University, G?teborg, Sweden; the Gladstone Institute of Cardiovascular Disease (R.E.P., L.A.), Cardiovascular Research Institute and the Department of Pathology (R.E.P.), University of California, San Francisco; and the Department of Poultry Science (R.L.W.), Texas A&M University, Kleberg Center, College Station, TX.
Correspondence to Dr Jan Borén, Wallenberg Laboratory, Sahlgrenska University Hospital, S-413 45 G?teborg, Sweden. E-mail Jan.Boren@wlab.gu.se
Abstract
Objective— The aim of this study was to investigate the molecular mechanism for changes in proteoglycan binding and LDL receptor affinity on two compositional changes in LDL that have been associated with atherosclerosis: cholesterol enrichment of the core and modification by secretory group IIA phospholipase A2 (sPLA2) of the surface.
Methods and Results— Transgenic mice expressing recombinant apolipoprotein (apo) B and sPLA2 were generated. Recombinant LDL were isolated and tested for their proteoglycan and LDL receptor-binding activity. The results show site A (residues 3148-3158) in apoB100 becomes functional in sPLA2-modified LDL and that site A acts cooperatively with site B (residues 3359-3369), the primary proteoglycan-binding site in native LDL, in the binding of sPLA2-modified LDL to proteoglycans. Our results also show that cholesterol enrichment of LDL is associated with increased affinity for proteoglycans and for the LDL receptor. This mechanism is likely mediated by a conformational change of site B and is independent of site A in apoB100.
Conclusion— Site A in apoB100 becomes functional in sPLA2-modified LDL and acts cooperatively with site B resulting in increased proteoglycan-binding activity. The increased binding for proteoglycans of cholesterol-enriched LDL is solely dependent on site B.
Key Words: lipoproteins ? apolipoprotein B ? proteoglycans ? atherosclerosis ? transgenic mice ? phospholipase A2
Introduction
High levels of plasma low-density lipoproteins (LDL) are a major risk factor for coronary disease and are the source of most of the cholesterol that accumulates in the arterial wall. Subendothelial retention of LDL is a key pathogenic process in atherosclerosis,1,2 and several lines of evidence suggest that intramural retention of atherogenic lipoproteins involves the extracellular matrix, chiefly proteoglycans.1 Subsequently, a series of biological responses to this retained material leads to an inflammatory response that promotes lesion formation and increased retention of atherogenic lipoproteins.1
Proteoglycans contain long carbohydrate side-chains of glycosaminoglycans, which are covalently attached to a core protein by a glycosidic linkage. The glycosaminoglycans consist of repeating disaccharide units, all bearing negatively charged groups, usually sulfate or carboxylic acid groups. In vitro, LDL binds with high affinity to many proteoglycans found in the artery wall. The interaction between LDL and proteoglycans involves basic amino acids in apolipoprotein (apo) B100, the protein moiety of LDL, that interact with the negatively charged glycosaminoglycans3–5 or bridging molecules such as apoE or lipoprotein lipase.1
Isolation of large fragments of apoB100 from different regions characterized by concentrations of positive clusters indicated that up to 8 specific regions in delipidated apoB100 bind proteoglycans.6–8 Two of these sites (site A at residues 3148 to 3158 and site B at residues 3359 to 3369) have been proposed to act cooperatively in the association with proteoglycans.9 In the human sequence, a disulfide link between Cys-3167 and Cys-3297 has been postulated to bring the two sites into close proximity.6 However, these cysteine residues are not conserved in other species.7 We have earlier shown that LDL in which the basic amino acid residues of site B have been mutated to neutral amino acid residues (RK3359-3369SA LDL) fail to interact with proteoglycans.8 This result indicates that site B is the primary proteoglycan-binding site in apoB100 when the protein is associated with native LDL.8 However, the conformation of apoB100 on the surface of the LDL particle is likely dependent on the composition of the core lipids, the surface phospholipid content, and the diameter of the LDL particle. Thus, other binding sites may become functional in modified LDL.
An elevated plasma level of secretory group II phospholipase A2 (sPLA2) is a strong independent risk factor for coronary heart disease.10 Whether this is associated with the proatherogenic mechanisms of sPLA2 activity in plasma or with its actions in the arterial wall remains to be elucidated. In blood plasma, sPLA2 can modify the circulating lipoproteins, inducing the formation of small dense LDL particles, which are associated with increased risk of cardiovascular disease. Lipolysis of LDL by sPLA2 alters the expression of apoB100 epitopes on the LDL particle,11 and sPLA2-modified LDL display increased interaction with glycosaminoglycans.12 The plasma concentration of sPLA2 is increased in systemic inflammations, such as rheumatoid arthritis (RA),13 which is known to be associated with increased risk of coronary heart disease, and LDL from RA patients has increased binding affinities to glycosaminoglycans.13
Willner et al recently showed that the enrichment of cholesterol of LDL is another important risk factor for atherosclerosis; triglyceride-rich apoB-containing lipoproteins are not as atherogenic as those containing cholesteryl esters.14 The authors proposed that the inflammatory response in atherosclerosis occurs primarily in response to toxicity from cholesterol accumulation in lesions.14 However, it has also been shown that the triglyceride content of LDL are reciprocally related to the number of exposed free lysine amino groups of apoB100.15 This could indicate that the increased atherogenicity of cholesteryl ester-rich LDL could, at least partly, be linked to an increased affinity for artery wall proteoglycans.
In this study, we tested the hypothesis that site A is important for proteoglycan-binding in sPLA2-modified LDL. Here we show that site A acts cooperatively with site B to increase the affinity of sPLA2-modified LDL for artery wall proteoglycans. We also show that cholesterol enrichment of the LDL particle is associated with increased affinity for proteoglycans. This mechanism is likely mediated by a conformational change of site B and is independent of site A.
Methods
Chemicals for SDS-PAGE were from Bio-Rad (Hercules, Calif). All other chemicals were from Sigma (Stockholm, Sweden). Horseradish peroxidase-conjugated polyclonal antibody against human apoB was from the Binding Site (Birmingham, UK); 1-Step Turbo TMB-ELISA was from Pierce (Rockford, Ill).
Generation of Transgenic Mice Expressing RK3148-3158SQ ApoB100-Containing LDL
The 95-kb apoB P1 plasmid p15816 was prepared and modified by RecA-assisted restriction endonuclease cleavage.17 See online Methods, available at http://atvb.ahajournals.org. The resulting P1 plasmid encodes RK3148-3158SQ apoB100 in which the basic residues arginine and lysine in site A are replaced with serine and glutamine, respectively. The plasmid also contains a CAA to CTA mutation in codon 2153 that effectively abolishes the formation of apoB48 in mouse livers.17 The structure of human apoB100 containing the mutation is largely preserved when expressed in the mouse,18 and the mutation does not affect proteoglycan binding or interaction with the LDL receptor.7 P1 DNA was prepared and microinjected into fertilized mouse eggs (FVB/N) and back-crossed into C57BL/6J Bom. The two founders with the highest levels of plasma apoB were selected for breeding and further analysis.
Isolation of LDL
LDL (density=1.02 to 1.05 g/mL) were isolated from human apoB100 transgenic expressing RK3148-3158SQ LDL, RK3359-3369SA LDL19 or recombinant control LDL19 by sequential ultracentrifugation and dialyzed against 150 mmol/L NaCl and 0.01% EDTA, with pH 7.4.19 Mouse apoE and apoB were removed by immunoaffinity chromatography with rabbit polyclonal antibodies.19 Lion LDL was isolated identically19 from lion plasma pooled from 8 lions of different ages and sexes.
Competition of Mouse apoE With Chemically Modified Human apoE
ApoB-containing lipoproteins (density=1.02 to 1.05 g/mL) isolated from transgenic LDL receptor-deficient (Ldlr-/- or Ldlr+/+) mice were incubated with a 200-fold molar excess of cyclohexanedione-modified apoE from human very-LDL as described by Sk?lén et al.2
Characterization of Isolated Lipoproteins
The binding of LDL to biglycan or decorin was assessed with solid-phase assays as described in detail by Sk?lén et al.2 Competitive LDL receptor-binding assays were performed as described,19 and the amount of unlabeled lipoproteins needed to complete 50% with iodine (I) 125-labeled LDL after a 3-hour incubation at 4°C was determined from an exponential decay curve-fitting model.19 Lipoprotein particle diameters were determined by dynamic light-scattering analysis with a Microtrac Series 250 Ultrafine particle analyzer fitted with a flexible conduit-sheathed probe tip (UPA-250; Microtrac, Clearwater, Fla).21
Lipolysis of Recombinant and Lion LDL With PLA2
LDL (1 mg/mL) was incubated for 3 hours in 100 mmol/L HEPES, 5 mmol/L CaCl2, 2 mmol/L MgCl2, 140 mmol/L NaCl, and 10 μmol/L butylated hydroxytoluene, with pH 7.4, with or without 50 ng/mL PLA2 (from bee venom) in the presence of 2% (weight/volume) bovine serum albumin.22 Lipolysis was stopped by adding EDTA to achieve a final concentration of 10 mmol/L as described.22 Samples of the reaction were separated by fast performance liquid chromatography gel filtration (KTA Explorer; Amersham Pharmacia) with a Superose 6 HR 10/30 column. Aliquots of 200 μL were injected onto the column and separated with PBS buffer, with pH 7.4, at a flow rate of 0.2 mL/min. The fractions corresponding to non-aggregated LDL were subjected to binding studies.
Data Analysis
Curve-fit analysis was performed in GraphPad Prism version 3.00 for Windows using a nonlinear curve-fit assuming one-site binding (San Diego, Calif). Statistical analysis was performed using SigmaStat for Windows version 2.03 (SPSS, Chicago, Ill).
Results
Generation and Characterization of Recombinant RK3148–3158SQ LDL
To evaluate the importance of site A (residues 3148-3158) in the interaction between native apoB100-containing LDL and proteoglycans, we generated transgenic mice expressing human recombinant LDL in which site A in apoB100 was mutated by substituting serines and glutamines for arginines and lysines, respectively (RK3148-3158SQ). Recombinant control LDL, RK3148-3158SQ LDL, and RK3359-3369SA LDL19 were isolated from mouse plasma by ultracentrifugation, and lipoproteins containing endogenous mouse apoE and apoB were removed by immunoaffinity chromatography; the removal of all endogenous apoB and apoE was confirmed by Western blot analysis (not shown). In RK3359-3369SA LDL, the arginine residues were converted to serines and the lysine residues were changed to alanines in site B of apoB100.19 Previously, we showed that site B (residues 3359-3369) is the combined proteoglycan binding and LDL receptor binding site in apoB100 and that RK3359-3369SA LDL do not interact with proteoglycans or the LDL receptor.7
The isolated recombinant control LDL, RK3148-3158SQ LDL, and RK3359-3369SA LDL had almost identical lipid compositions (data not shown) and particle diameters (27.1±0.86, 26.7±0.20, and 26.8±0.85 nm, respectively; mean±SD).
To determine the ability of the recombinant LDL to interact with proteoglycans, we subjected recombinant control LDL, RK3148-3158SQ LDL, and RK3359-3369SA LDL to solid-phase assays with biglycan (Figure 1A) and decorin (not shown). In two independent experiments, recombinant control LDL and recombinant RK3148-3158SQ LDL bound biglycan with high affinity, whereas recombinant RK3359-369SA LDL displayed severely impaired binding (Figure 1A).
Figure 1. Solid-phase assays of the binding of LDL to biglycan (A) and competitive receptor binding assay of recombinant LDL (B). The abilities of recombinant control LDL (?), RK3148-3158SQ LDL (), and RK3359-3369SA LDL () to bind biglycan or compete with 125I-labeled human plasma LDL (2 μg/mL) for binding to LDL receptors on normal human fibroblasts were determined. The recombinant lipoproteins were isolated from 20 mice, and endogenous apoE and apoB were removed by affinity chromatography. The results represent mean values (±SD) from 2 independent experiments, each performed with freshly isolated LDL (n=4 in each experiment) (A) or the mean of 2 independent experiments, each performed in duplicate (B).
To investigate the importance of site A in the interaction with the LDL receptor, we isolated recombinant control LDL, RK3148-3158SQ LDL, and RK3359-3369SA LDL and tested them in an in vitro competitive receptor-binding assay (Figure 1B). Recombinant control LDL and RK3148-3158SQ LDL had almost identical receptor binding (ED50 2.1 and 2.3 μg/mL, respectively), whereas recombinant RK3359-3369SA LDL displayed defective receptor binding (ED50 >20 μg/mL). Taken together, these results indicate that site A in native LDL is not involved in binding to proteoglycans or to the LDL receptor.
Solid-Phase Assays of Recombinant LDL Isolated From sPLA2 Transgenic Mice
Next, we tested the hypothesis that site A can modulate the proteoglycan binding activity of modified LDL. Recombinant control LDL, RK3148-3158SQ LDL, and RK3359-3369SA LDL were isolated by ultracentrifugation from human apoB transgenic mice and from doubly transgenic mice expressing human apoB and sPLA2.23 After removal of endogenous mouse apoE-containing and apoB-containing lipoproteins by immunoaffinity chromatography, the isolated recombinant LDL were subjected to solid-phase assays with biglycan (Figure 2A) and decorin (not shown). In two independent experiments, recombinant control LDL from sPLA2 transgenic mice displayed increased binding, whereas recombinant RK3148-3158SQ LDL from sPLA2 transgenic mice and sPLA2 nontransgenic littermate controls had almost identical binding (Figure 2A). These results indicate that site A is important for the increased interaction of sPLA2-modified LDL with glycosaminoglycans. However, RK3359-3369SA LDL did not interact with biglycan, regardless of whether it was isolated from sPLA2 transgenic mice or not (Figure 2B). These results indicate that site A in sPLA2-modified LDL modulates the binding of the LDL particle to proteoglycans but cannot substitute for site B, which is essential for the interaction with proteoglycans.
Figure 2. Binding of recombinant LDL isolated from sPLA2 transgenic mice to biglycan. The abilities of recombinant control LDL (?) and RK3148-3158SQ LDL () and RK3359-3369SA LDL () isolated from human apoB transgenic mice (?, , ) and from apoBxsPLA2 transgenic mice (, , ) to bind biglycan were determined. Endogenous mouse apoB and apoE were removed from all lipoproteins by immunoaffinity chromatography. The results represent mean values (±SD) from two independent experiments, each performed with freshly isolated LDL from 15 mice of each group (n=4 in each experiment).
Characterization of sPLA2-Modified Recombinant LDL
sPLA2 hydrolyzes glyceroacyl phospholipids to form nonesterified fatty acids and lysophospholipids. Therefore, we analyzed the phospholipid composition of recombinant LDL isolated from human apoB transgenic mice and from apoB transgenic mice with mice expressing sPLA2 (apoBxsPLA2 transgenic mice). The sPLA2-modifed LDL contained more lysophosphatidylcholine than nonmodified recombinant LDL (Table 1). We also analyzed the content of total cholesterol, free cholesterol, cholesteryl esters, and triglycerides. No differences were found in the content of these lipids in recombinant LDL isolated from apoB transgenic mice or from apoBxsPLA2 transgenic mice (Table 2). Neither the phospholipid composition (Table 1) nor the content of the other lipids (Table 2) was affected by the type of apoB100 (control or the two mutations, ie, RK3148-3158SQ or RK3359-3369SA) in LDL isolated from apoB transgenic mice or from apoBxsPLA2 transgenic mice (data not shown).
TABLE 1. Phospholipid Composition of Recombinant LDL Isolated From Human apoB Transgenic Mice and From Double-Transgenic apoBxsPLA2 Mice
TABLE 2. Lipid Composition of Recombinant LDL Isolated From Human apoB Transgenic Mice and From Double-Transgenic apoBxsPLA2 Mice
Dynamic light-scattering analysis of recombinant control LDL isolated from human apoB transgenic mice and from apoBxsPLA2 transgenic mice showed that recombinant LDL from the singly human apoB transgenic mice was significantly larger (26.7±0.42 versus 25.8±0.38 nm;, mean ± SD; P<0.05). Taken together, these results show that in vivo modification of recombinant LDL in sPLA2 transgenic mice induces the formation of smaller LDL enriched in lysophosphatidylcholine. Thus, it is possible that the conformational change of apoB in LDL modified with sPLA2 in vivo is caused by the formation of smaller LDL and/or by changes in the phospholipids composition of the LDL particle.
Solid-Phase Assays of Recombinant LDL Isolated From
sPLA2 Transgenic Mice by Size-Exclusion Chromatography
To discriminate between these possibilities, recombinant control LDL and RK3148-3158SQ LDL were isolated from human apoB transgenic mice and from apoBxsPLA2 transgenic mice by a combination of ultracentrifugation and size-exclusion chromatography and subjected to solid-phase assays with biglycan. The isolated LDL had almost identical size (26.4±0.23 and 26.3±0.15 nm). In two independent experiments, recombinant control LDL from sPLA2 transgenic mice displayed increased binding, whereas recombinant RK3148-3158SQ LDL from sPLA2 transgenic mice and sPLA2 nontransgenic littermate controls had almost identical binding. The Kd for control LDL and RK3148-3158SQ LDL isolated from apoB transgenic mice were 16.3 and 17.1, respectively, and from apoBxsPLA2 transgenic mice 10.3 and 16.6, respectively. The results also showed that the diameter of sPLA2-modified and non-modified recombinant control LDL were inversely correlated to the binding affinity to biglycan; smaller LDL had higher binding affinity to biglycan than larger LDL (data not shown). Taken together, the results show that the conformational change of apoB in sPLA2-modified LDL is caused by the formation of smaller LDL and by changes in the phospholipid composition of the LDL particle.
To investigate if in vitro digestion and in vivo modification of LDL with PLA2 give similar results, recombinant control LDL, RK3359-3369SA LDL, and RK3148-3158SQ LDL were isolated and digested in vitro with PLA2 for 3 hours. Samples of the reaction were separated by size-exclusion chromatography, and non-aggregated LDL was isolated and subjected to binding studies. This was performed because lipolysis of LDL with PLA2 in vitro induces particle aggregation.22 The results from these studies showed that in vitro modification of recombinant LDL with PLA2 gave similar results as in vivo modification of LDL, ie, binding affinity increased on PLA2 digestion and site A was required for the increased interaction of sPLA2-modified LDL with glycosaminoglycans (data not shown). However, the intra-assay variations (ie, imprecision) and the inter-assay variations (ie, irreproducibility) were much more pronounced when analyzing the binding of LDL digested with PLA2 in vitro compared with LDL modified in vivo.
Binding of Lion LDL to Biglycan and to the LDL Receptor
Our data indicate that site A becomes exposed and functional on sPLA2 modification and that site A and site B act cooperatively in binding to proteoglycans. In the human sequence, site A and site B are separated by 202 amino acids, but a disulfide link between Cys-3167 and Cys-3297 has been postulated to bring the two binding sites into close proximity.10 However, these cysteine residues are not conserved in other species;6 but because of a naturally occurring deletion between site A and site B in several carnivores, including lions, the binding sites are separated by 76 amino acids only in these species.24 Therefore, we investigated the proteoglycan-binding activity of lion LDL with and without PLA2 digestion.
Lion LDL and human LDL were isolated and digested in vitro with PLA2 for 3 hours and non-aggregated LDL was isolated by a combination of ultracentrifugation and size-exclusion chromatography. The PLA2 modification induced the formation of smaller LDL enriched in lysophosphatidylcholine. The diameter of lion LDL and human LDL decreased from 27.2±0.33 to 26.1±0.25 nm and from 27.5±0.39 to 26.3±0.15 nm, respectively. The content of lysophosphatidylcholine in PLA2-digested LDL increased 2.4-fold±0.3 and 2.8-fold±0.6, respectively.
As shown by SDS-PAGE, lion apoB migrated faster than human apoB100, and the LDL did not contain any apolipoprotein other than apoB (Figure 3A). In solid-phase assays, lion LDL bound with higher affinity than human LDL to biglycan; however, PLA2 modification did not increase the binding of lion LDL to biglycan, as was seen in human LDL (Figure 4A). The Kd for lion and human LDL were 10.2 and 15.4, respectively, and for PLA2-modified lion and human LDL 10.4 and 11.3, respectively. The binding capacity of lion and human LDL was similar (Bmax 0.71 and 0.66, respectively), indicating that the deletion between sites A and B increases the affinity for proteoglycans but not the binding capacity. In competitive LDL receptor-binding studies with human 125I-labeled LDL, lion and human LDL had almost identical receptor binding (ED50 1.9 and 1.8 μg/mL, respectively). These results are consistent with the hypothesis that site A is functional on lion LDL with or without PLA2 modification.
Figure 3. Coomassie staining and Western analysis of apoB-containing LDL. A, Lion and human plasma LDL (density=1.02 to 1.05 g/mL) were isolated by sequential ultracentrifugation. Five micrograms of LDL were separated by SDS-PAGE with 3% to 15% gels. Arrows indicate molecular mass markers. B, Recombinant LDL (density=1.02 to 1.05 g/mL) was isolated from LDLR-/- mice (HuApoBxLDLR-/-) or LDLR+/+ mice (HuApoB), and 2 μg were analyzed on SDS-PAGE with 3% to 15% gels. C, Endogenous apoB was removed by immunoaffinity chromatography, and 0.5 μg of recombinant control LDL (density=1.02 to 1.05 g/mL) from LDLR-/- mice (lane 1) or LDLR+/+ mice (lane 2) was analyzed by Western blots with polyclonal antibodies against mouse apoE. Lane 3 shows recombinant control LDL (0.5 μg) from LDLR-/- mice analyzed by Western blots with polyclonal antibodies against mouse apoE (upper panel) and human apoE (lower panel) after incubation with CHD-modified human apoE (HuApoE competition). The CHD-modified samples were incubated with an equal volume of 1 mol/L hydroxylamine, 0.3 mol/L mannitol, at pH 7.0, at 37°C for 16 hours to reverse the modification of the arginine residues before sodium dodecyl sulfate-polyacrylamide gel electrophoresis.
Figure 4. Solid-phase assays of lion LDL or recombinant LDL isolated from Ldlr+/+ and Ldlr-/- mice to biglycan binding to biglycan. A, Lion () and human (?) plasma LDL (density=1.02 to 1.05 g/mL) were isolated by sequential ultracentrifugation and incubated with ( and ) or without (? and ) 50 ng/mL of PLA2 (from bee venom). To prevent overlap in the figure, and have been moved by -0.5 and +0.5 U on the X-axis, respectively. B, The abilities of recombinant control LDL (?) and RK3148-3158SQ LDL () isolated from Ldlr+/+ ( and ) and Ldlr-/- (? and ) mice to biglycan were determined. Endogenous apoE was replaced with cyclohexanedione-modified human apoE, and endogenous mouse apoB was removed by immunoaffinity chromatography. The results represent mean values (±SD) from two independent experiments, each performed with freshly isolated LDL from 15 mice of each group (n=4 in each experiment).
Solid-Phase Assays of Cholesterol-Enriched ApoB100 Recombinant LDL With Biglycan
To analyze if cholesterol enrichment of LDL is associated with increased affinity for proteoglycans, human apoB transgenic mice expressing recombinant control LDL, RK3148-3158SQ LDL, or RK3359-3369SA LDL were bred onto Ldlr-/- background. Recombinant LDL isolated from Ldlr-/- mice is enriched with cholesteryl esters. The total cholesterol to triglyceride (in mmol/L) ratios in recombinant LDL isolated from human apoBxLdlr+/+ transgenic mice and apoBxLdlr-/- transgenic mice are 1.7 and 4.2, respectively.
Recombinant control LDL and RK3148-3158SQ LDL from Ldlr+/+ and Ldlr-/- mice were isolated by ultracentrifugation, and endogenous mouse apoB and apoE were removed. LDL isolated from Ldlr-/- mice is greatly enriched in endogenous apoE (Figure 3B), and immunoaffinity techniques fail to remove all endogenous apoE from LDL isolated from Ldlr-/- mice.1,2 Competition of mouse apoE with CHD-modified human apoE efficiently removes the endogenous apoE from the recombinant LDL (Figure 3C), and the CHD-modified apoE fail to facilitate an indirect binding between the recombinant LDL and glycosaminoglycans.1,2 The competition of mouse apoE with chemically modified human apoE did not affect the cholesterol or triglyceride content of the LDL particles. The recovery of these lipids were 94%±8% and 92%±11% after competition with CHD-modified apoE, respectively (mean±SD, n=5).
The purified recombinant control LDL and RK3148-3158SQ LDL from Ldlr+/+ and Ldlr-/- mice were then subjected to size-exclusion chromatography to isolate LDL particles with almost identical size (diameters: 26.8±0.84 and 26.7±0.21 nm versus 26.9±1.87 and 26.8±0.42 nm, respectively; mean±SD). In two independent experiments, recombinant control LDL and RK3148-3158SQ LDL isolated from Ldlr-/- mice bound biglycan and decorin (not shown) with higher affinity than recombinant control LDL and RK3148-3158SQ LDL isolated from Ldlr+/+ mice (Figure 4B). Within each genotype, recombinant control LDL and RK3148-3158SQ LDL displayed almost identical binding. The Kd for control LDL and RK3148-3158SQ LDL isolated from Ldlr+/+ were 17.0 and 16.7, respectively, and from Ldlr-/- mice 9.1 and 8.2, respectively. The binding capacity of recombinant control LDL and RK3148-3158SQ LDL was almost identical, both when isolated from Ldlr+/+ and from Ldlr-/- mice (Bmax were 0.51, 0.47, 0.51, and 0.48, respectively). Control experiments showed that RK3359-3369SA LDL from both Ldlr+/+ and Ldlr-/- mice had severely impaired binding (data not shown). Competitive receptor-binding studies with human 125I-labeled LDL showed that cholesterol-enriched recombinant control LDL from Ldlr-/- mice bound with higher affinity to the LDL receptor than that from Ldlr+/+ mice (ED50 1.7 versus 2.5 μg/mL).
These findings indicate that cholesterol enrichment of the LDL particle is associated with increased affinity for proteoglycans and for the LDL receptor. This mechanism is likely mediated by a conformational change of site B and is independent of site A in apoB100.
Discussion
The aim of this study was to investigate the molecular mechanism for changes in proteoglycan binding and LDL receptor affinity on two compositional changes in LDL: modification by secretory group IIA phospholipase A2 (sPLA2) of the surface and cholesterol-enrichment of the core.
Elevated plasma sPLA2 activity has been shown to be an important risk factor for coronary heart disease.8 Thus, clarification of the proatherogenic mechanisms by which sPLA2 acts in the plasma and in the artery wall is important to our understanding of the pathogenesis of the disease. To investigate the molecular mechanism for how circulating sPLA2 induces the increased binding affinities of LDL to glycosaminoglycans,13 we crossed our human apoB transgenic mice with mice expressing sPLA2.23 The serum activity of sPLA2 in these transgenic mice is elevated approximately 8-fold compared with nontransgenic littermates, but there is no evidence of systemic inflammation in the transgenic mice.23 Here we show that in vivo modification of recombinant LDL in sPLA2 transgenic mice induces the formation of slightly smaller LDL enriched with lysophosphatidylcholine. These results are in agreement with data of Grass et al23 and with results from patients with rheumatoid arthritis who have increased plasma levels of sPLA2.13 Taken together, our results are consistent with the hypothesis that sPLA2 modification of LDL in the circulation leads to exposure of site A, resulting in increased proteoglycan binding. Furthermore, the conformational change of apoB in sPLA2-modified LDL is caused by the formation of smaller LDL and by changes in the phospholipid composition of the LDL particle.
However, it is conceivable that some of the observed effects could result from indirect secondary consequences. For example, sPLA2 modification might cause LDL with low PG binding to be cleared, leaving behind a high-affinity subpopulation. It is also possible that sPLA2 digestion in vivo allows a second enzyme to further alter the LDL, eg, sphingomyelinase, which has been reported to readily digest sPLA2-treated LDL.25 However, because a gentle PLA2 digestion of recombinant LDL in vitro gives similar results, and because isolated LDL from apoBxsPLA2 transgenic mice contain abundant sphingomyelin, these alternative explanations seems unlikely.
In some experiments, LDL was digested with PLA2 in vitro. These incubations were performed for 3 hours only. The reason for this was to achieve similar modification of LDL as in patients with increased plasma levels of sPLA2.13 Prolonged digestion of LDL with PLA2 in vitro results in hydrolysis of nearly all of the phosphatidylcholines, in pronounced aggregation, and in the presence of glycosaminoglycans fusion of LDL,22 ie, the modifications of LDL that are more likely to occur within the extracellular matrix of the arterial wall.
The findings that lion LDL, with a 126-amino acid residue deletion between site A and site B, has increased affinity for artery wall proteoglycans and that PLA2 modification of lion LDL did not increase the binding of lion LDL to biglycan are consistent with the hypothesis that site A on lion LDL is also exposed with or without sPLA2 modification. However, studies with cross-linking or some other direct proof would be needed to draw this conclusion. It is also possible that other basic sequences or even other apolipoproteins might play a role in the unusual properties of lion LDL. However, the latter possibility seems unlikely because no apolipoproteins other than apoB was detected in isolated lion LDL (Figure 3A). The fact that two independent deletions have been found in apoB100 between site A and site B might indicate that these mutations provide an evolutionary advantage. In addition to the 126-amino acid deletion in several carnivores, including lions,24 an 84-amino acid deletion has been found in elephants.24 A tentative hypothesis is that an enhanced binding of LDL to matrix might promote wound healing, as has been proposed, although not proven, for Lp(a).26
Our results show that cholesterol-enriched LDL has increased affinity for glycosaminoglycans and the LDL receptor. This mechanism is likely mediated by a conformational change of site B and is, in contrast to sPLA2-modified LDL, independent of site A. The finding that cholesterol-enriched LDL have increased affinity for artery wall proteoglycans offers an additional explanation for why the atherogenicity of apoB-containing lipoproteins correlates with their cholesteryl ester content.
The binding studies with LDL isolated from Ldlr-/- and Ldlr+/+ mice were performed with LDL of almost identical diameters, and the results are consistent with the triglyceride-dependent decrease in affinity for the LDL receptor reported in hypertriglyceridemic subjects.27 However, our results are not consistent with the finding that in the absence of changes in the size of the particle, the relative content of neutral lipid in the core of LDL does not significantly affect apoB conformation or its affinity for the LDL receptor.28 The reasons for these differences are unclear but could reflect differences in experimental design, because the secondary structure of apoB in LDL strongly depends on the physical state of the lipoprotein core.29
In this study, we tested the hypothesis that site A is important for proteoglycan binding of sPLA2-modified LDL. The results confirm that site B is the principal proteoglycan-binding site in native apoB100-containing LDL and show that site A acts cooperatively with site B in the association with proteoglycans in sPLA2-modified LDL. The results also show that cholesterol enrichment of the LDL particle is associated with increased affinity for proteoglycans and for the LDL receptor. This mechanism is likely mediated by a conformational change of site B and is independent of site A in apoB100. Taken together, the results show that LDL affinity for proteoglycans can be modified by size-dependent effects and size-independent compositional effects of either the surface or the core.
Acknowledgments
The authors thank K. Sk?lén, E. Bj?rk, L. Lindgren, M. Heyden, B. Gabrielli, and K. Arnold for excellent technical assistance, S. Ordway and G. Howard for editorial assistance, and Dr Torsten M?ller and Karin Magnusson at K?lmorden Zoo for lion plasma.
This work was supported by the Swedish Medical Research Council, The Swedish Foundation for Strategic Research, and The Swedish Heart–Lung Foundation.
References
Williams KJ, Tabas I. The response-to-retention hypothesis of early atherogenesis. Arterioscler Thromb Vasc Biol. 1995; 15: 551–561.
Skalen K, Gustafsson M, Rydberg EK, Hulten LM, Wiklund O, Innerarity TL, Boren J. Subendothelial retention of atherogenic lipoproteins in early atherosclerosis. Nature. 2002; 417: 750–754.
Camejo G, Olofsson SO, Lopez F, Carlsson P, Bondjers G. Identification of Apo B-100 segments mediating the interaction of low density lipoproteins with arterial proteoglycans. Arteriosclerosis. 1988; 8: 368–377.
Weisgraber KH, Rall SC, Jr. Human apolipoprotein B-100 heparin-binding sites. J Biol Chem. 1987; 262: 11097–11103.
Hirose N, Blankenship DT, Krivanek MA, Jackson RL, Cardin AD. Isolation and characterization of four heparin-binding cyanogen bromide peptides of human plasma apolipoprotein B. Biochemistry. 1987; 26: 5505–5512.
Chan L. Apolipoprotein B, the major protein component of triglyceride-rich and low density lipoproteins. J Biol Chem. 1992; 267: 25621–25624.
Law A, Scott J. A cross-species comparison of the apolipoprotein B domain that binds to the LDL receptor. J Lipid Res. 1990; 31: 1109–1120.
Boren J, Olin K, Lee I, Chait A, Wight T, Innerarity T. Identification of the principal proteoglycan binding site in LDL: A single point mutation in apolipoprotein B100 severely affects proteoglycan interaction without affecting LDL receptor binding. J Clin Invest. 1998; 101: 2658–2664.
Olsson U, Camejo G, Hurt-Camejo E, Elfsber K, Wiklund O, Bondjers G. Possible functional interactions of apolipoprotein B-100 segments that associate with cell proteoglycans and the ApoB/E receptor. Arterioscler Thromb Vasc Biol. 1997; 17: 149–155.
Kugiyama K, Ota Y, Takazoe K, Moriyama Y, Kawano H, Miyao Y, Sakamoto T, Soejima H, Ogawa H, Doi H, Sugiyama S, Yasue H. Circulating levels of secretory type II phospholipase A(2) predict coronary events in patients with coronary artery disease. Circulation. 1999; 100: 1280–1284.
Kleinman Y, Krul ES, Burnes M, Aronson W, Pfleger B, Schonfeld G. Lipolysis of LDL with phospholipase A2 alters the expression of selected apoB-100 epitopes and the interaction of LDL with cells. J Lipid Res. 1988; 29: 729–743.
Sartipy P, Camejo G, Svensson L, Hurt-Camejo E. Phospholipase A(2) modification of low density lipoproteins forms small high density particles with increased affinity for proteoglycans and glycosaminoglycans. J Biol Chem. 1999; 274: 25913–25920.
Hurt-Camejo E, Paredes S, Masana L, Camejo G, Sartipy P, Rosengren B, Pedreno J, Vallve JC, Benito P, Wiklund O. Elevated levels of small, low-density lipoprotein with high affinity for arterial matrix components in patients with rheumatoid arthritis: possible contribution of phospholipase A2 to this atherogenic profile. Arthritis Rheum. 2001; 44: 2761–2767.
Willner EL, Tow B, Buhman KK, Wilson M, Sanan DA, Rudel LL, Farese RV, Jr. Deficiency of acyl CoA:cholesterol acyltransferase 2 prevents atherosclerosis in apolipoprotein E-deficient mice. Proc Natl Acad Sci U S A. 2003; 100: 1262–1267.
Aviram M, Lund-Katz S, Phillips MC, Chait A. The influence of the triglyceride content of low density lipoprotein on the interaction of apolipoprotein B-100 with cells. J Biol Chem. 1988; 263: 16842–16848.
Linton MF, Farese RV, Jr., Chiesa G, Grass DS, Chin P, Hammer RE, Hobbs HH, Young SG. Transgenic mice expressing high plasma concentrations of human apolipoprotein B100 and lipoprotein(a). J Clin Invest. 1993; 92: 3029–3037.
Boren J, Lee I, Callow MJ, Rubin EM, Innerarity TL. A simple and efficient method for making site-directed mutants, deletions, and fusions of large DNA such as P1 and BAC clones. Genome Res. 1996; 6: 1123–1130.
Wang X, Chauhan V, Nguyen AT, Schultz J, Davignon J, Young SG, Boren J, Innerarity TL, Rutai H, Milne RW. Immunochemical evidence that human apoB differs when expressed in rodent versus human cells. J Lipid Res. 2003; 44: 547–553.
Boren J, Lee I, Zhu W, Arnold K, Taylor S, Innerarity TL. Identification of the low density lipoprotein receptor-binding site in apolipoprotein B100 and the modulation of its binding activity by the carboxyl terminus in familial defective apoB100. J Clin Invest. 1998; 101: 1084–1093.
Flood C, Gustafsson M, Richardson PE, Harvey SC, Segrest JP, Boren J. Identification of the proteoglycan binding site in apolipoprotein B48. J Biol Chem. 2002; 277: 32228–32233.
Walzem RL, Davis PA, Hansen RJ. Overfeeding increases very low density lipoprotein diameter and causes the appearance of a unique lipoprotein particle in association with failed yolk deposition. J. Lipid Res. 1994; 35: 1354–1366.
Hakala JK, Oorni K, Ala-Korpela M, Kovanen PT. Lipolytic modification of LDL by phospholipase A2 induces particle aggregation in the absence and fusion in the presence of heparin. Arterioscler Thromb Vasc Biol. 1999; 19: 1276–1283.
Grass DS, Felkner RH, Chiang MY, Wallace RE, Nevalainen TJ, Bennett CF, Swanson ME. Expression of human group II PLA2 in transgenic mice results in epidermal hyperplasia in the absence of inflammatory infiltrate. J Clin Invest. 1996; 97: 2233–2241.
Amrine-Madsen H, Koepfli KP, Wayne RK, Springer MS. A new phylogenetic marker, apolipoprotein B, provides compelling evidence for eutherian relationships. Mol Phylogenet Evol. 2003; 28: 225–240.
Schissel SL, Keesler GA, Schuchman EH, Williams KJ, Tabas I. The cellular trafficking and zinc dependence of secretory and lysosomal sphingomyelinase, two products of the acid sphingomyelinase gene. J Biol Chem. 1998; 273: 18250–18259.
Yano Y, Shimokawa K, Okada Y, Noma A. Immunolocalization of lipoprotein(a) in wounded tissues. J Histochem Cytochem. 1997; 45: 559–568.
McKeone BJ, Patsch JR, Pownall HJ. Plasma triglycerides determine low density lipoprotein composition, physical properties, and cell-specific binding in cultured cells. J Clin Invest. 1993; 91: 1926–1933.
Galeano NF, Milne R, Marcel YL, Walsh MT, Levy E, Ngu’yen TD, Gleeson A, Arad Y, Witte L, Al-Haideri M, Rumsey SC, Deckelbaum RJ. Apoprotein B structure and receptor recognition of triglyceride-rich low density lipoprotein (LDL) is modified in small LDL but not in triglyceride-rich LDL of normal size. J Biol Chem. 1994; 269: 511–519.
Coronado-Gray A, van Antwerpen R. The physical state of the LDL core influences the conformation of apolipoprotein B-100 on the lipoprotein surface. FEBS Lett. 2003; 533: 21–24.(Christofer Flood; Maria G)
Correspondence to Dr Jan Borén, Wallenberg Laboratory, Sahlgrenska University Hospital, S-413 45 G?teborg, Sweden. E-mail Jan.Boren@wlab.gu.se
Abstract
Objective— The aim of this study was to investigate the molecular mechanism for changes in proteoglycan binding and LDL receptor affinity on two compositional changes in LDL that have been associated with atherosclerosis: cholesterol enrichment of the core and modification by secretory group IIA phospholipase A2 (sPLA2) of the surface.
Methods and Results— Transgenic mice expressing recombinant apolipoprotein (apo) B and sPLA2 were generated. Recombinant LDL were isolated and tested for their proteoglycan and LDL receptor-binding activity. The results show site A (residues 3148-3158) in apoB100 becomes functional in sPLA2-modified LDL and that site A acts cooperatively with site B (residues 3359-3369), the primary proteoglycan-binding site in native LDL, in the binding of sPLA2-modified LDL to proteoglycans. Our results also show that cholesterol enrichment of LDL is associated with increased affinity for proteoglycans and for the LDL receptor. This mechanism is likely mediated by a conformational change of site B and is independent of site A in apoB100.
Conclusion— Site A in apoB100 becomes functional in sPLA2-modified LDL and acts cooperatively with site B resulting in increased proteoglycan-binding activity. The increased binding for proteoglycans of cholesterol-enriched LDL is solely dependent on site B.
Key Words: lipoproteins ? apolipoprotein B ? proteoglycans ? atherosclerosis ? transgenic mice ? phospholipase A2
Introduction
High levels of plasma low-density lipoproteins (LDL) are a major risk factor for coronary disease and are the source of most of the cholesterol that accumulates in the arterial wall. Subendothelial retention of LDL is a key pathogenic process in atherosclerosis,1,2 and several lines of evidence suggest that intramural retention of atherogenic lipoproteins involves the extracellular matrix, chiefly proteoglycans.1 Subsequently, a series of biological responses to this retained material leads to an inflammatory response that promotes lesion formation and increased retention of atherogenic lipoproteins.1
Proteoglycans contain long carbohydrate side-chains of glycosaminoglycans, which are covalently attached to a core protein by a glycosidic linkage. The glycosaminoglycans consist of repeating disaccharide units, all bearing negatively charged groups, usually sulfate or carboxylic acid groups. In vitro, LDL binds with high affinity to many proteoglycans found in the artery wall. The interaction between LDL and proteoglycans involves basic amino acids in apolipoprotein (apo) B100, the protein moiety of LDL, that interact with the negatively charged glycosaminoglycans3–5 or bridging molecules such as apoE or lipoprotein lipase.1
Isolation of large fragments of apoB100 from different regions characterized by concentrations of positive clusters indicated that up to 8 specific regions in delipidated apoB100 bind proteoglycans.6–8 Two of these sites (site A at residues 3148 to 3158 and site B at residues 3359 to 3369) have been proposed to act cooperatively in the association with proteoglycans.9 In the human sequence, a disulfide link between Cys-3167 and Cys-3297 has been postulated to bring the two sites into close proximity.6 However, these cysteine residues are not conserved in other species.7 We have earlier shown that LDL in which the basic amino acid residues of site B have been mutated to neutral amino acid residues (RK3359-3369SA LDL) fail to interact with proteoglycans.8 This result indicates that site B is the primary proteoglycan-binding site in apoB100 when the protein is associated with native LDL.8 However, the conformation of apoB100 on the surface of the LDL particle is likely dependent on the composition of the core lipids, the surface phospholipid content, and the diameter of the LDL particle. Thus, other binding sites may become functional in modified LDL.
An elevated plasma level of secretory group II phospholipase A2 (sPLA2) is a strong independent risk factor for coronary heart disease.10 Whether this is associated with the proatherogenic mechanisms of sPLA2 activity in plasma or with its actions in the arterial wall remains to be elucidated. In blood plasma, sPLA2 can modify the circulating lipoproteins, inducing the formation of small dense LDL particles, which are associated with increased risk of cardiovascular disease. Lipolysis of LDL by sPLA2 alters the expression of apoB100 epitopes on the LDL particle,11 and sPLA2-modified LDL display increased interaction with glycosaminoglycans.12 The plasma concentration of sPLA2 is increased in systemic inflammations, such as rheumatoid arthritis (RA),13 which is known to be associated with increased risk of coronary heart disease, and LDL from RA patients has increased binding affinities to glycosaminoglycans.13
Willner et al recently showed that the enrichment of cholesterol of LDL is another important risk factor for atherosclerosis; triglyceride-rich apoB-containing lipoproteins are not as atherogenic as those containing cholesteryl esters.14 The authors proposed that the inflammatory response in atherosclerosis occurs primarily in response to toxicity from cholesterol accumulation in lesions.14 However, it has also been shown that the triglyceride content of LDL are reciprocally related to the number of exposed free lysine amino groups of apoB100.15 This could indicate that the increased atherogenicity of cholesteryl ester-rich LDL could, at least partly, be linked to an increased affinity for artery wall proteoglycans.
In this study, we tested the hypothesis that site A is important for proteoglycan-binding in sPLA2-modified LDL. Here we show that site A acts cooperatively with site B to increase the affinity of sPLA2-modified LDL for artery wall proteoglycans. We also show that cholesterol enrichment of the LDL particle is associated with increased affinity for proteoglycans. This mechanism is likely mediated by a conformational change of site B and is independent of site A.
Methods
Chemicals for SDS-PAGE were from Bio-Rad (Hercules, Calif). All other chemicals were from Sigma (Stockholm, Sweden). Horseradish peroxidase-conjugated polyclonal antibody against human apoB was from the Binding Site (Birmingham, UK); 1-Step Turbo TMB-ELISA was from Pierce (Rockford, Ill).
Generation of Transgenic Mice Expressing RK3148-3158SQ ApoB100-Containing LDL
The 95-kb apoB P1 plasmid p15816 was prepared and modified by RecA-assisted restriction endonuclease cleavage.17 See online Methods, available at http://atvb.ahajournals.org. The resulting P1 plasmid encodes RK3148-3158SQ apoB100 in which the basic residues arginine and lysine in site A are replaced with serine and glutamine, respectively. The plasmid also contains a CAA to CTA mutation in codon 2153 that effectively abolishes the formation of apoB48 in mouse livers.17 The structure of human apoB100 containing the mutation is largely preserved when expressed in the mouse,18 and the mutation does not affect proteoglycan binding or interaction with the LDL receptor.7 P1 DNA was prepared and microinjected into fertilized mouse eggs (FVB/N) and back-crossed into C57BL/6J Bom. The two founders with the highest levels of plasma apoB were selected for breeding and further analysis.
Isolation of LDL
LDL (density=1.02 to 1.05 g/mL) were isolated from human apoB100 transgenic expressing RK3148-3158SQ LDL, RK3359-3369SA LDL19 or recombinant control LDL19 by sequential ultracentrifugation and dialyzed against 150 mmol/L NaCl and 0.01% EDTA, with pH 7.4.19 Mouse apoE and apoB were removed by immunoaffinity chromatography with rabbit polyclonal antibodies.19 Lion LDL was isolated identically19 from lion plasma pooled from 8 lions of different ages and sexes.
Competition of Mouse apoE With Chemically Modified Human apoE
ApoB-containing lipoproteins (density=1.02 to 1.05 g/mL) isolated from transgenic LDL receptor-deficient (Ldlr-/- or Ldlr+/+) mice were incubated with a 200-fold molar excess of cyclohexanedione-modified apoE from human very-LDL as described by Sk?lén et al.2
Characterization of Isolated Lipoproteins
The binding of LDL to biglycan or decorin was assessed with solid-phase assays as described in detail by Sk?lén et al.2 Competitive LDL receptor-binding assays were performed as described,19 and the amount of unlabeled lipoproteins needed to complete 50% with iodine (I) 125-labeled LDL after a 3-hour incubation at 4°C was determined from an exponential decay curve-fitting model.19 Lipoprotein particle diameters were determined by dynamic light-scattering analysis with a Microtrac Series 250 Ultrafine particle analyzer fitted with a flexible conduit-sheathed probe tip (UPA-250; Microtrac, Clearwater, Fla).21
Lipolysis of Recombinant and Lion LDL With PLA2
LDL (1 mg/mL) was incubated for 3 hours in 100 mmol/L HEPES, 5 mmol/L CaCl2, 2 mmol/L MgCl2, 140 mmol/L NaCl, and 10 μmol/L butylated hydroxytoluene, with pH 7.4, with or without 50 ng/mL PLA2 (from bee venom) in the presence of 2% (weight/volume) bovine serum albumin.22 Lipolysis was stopped by adding EDTA to achieve a final concentration of 10 mmol/L as described.22 Samples of the reaction were separated by fast performance liquid chromatography gel filtration (KTA Explorer; Amersham Pharmacia) with a Superose 6 HR 10/30 column. Aliquots of 200 μL were injected onto the column and separated with PBS buffer, with pH 7.4, at a flow rate of 0.2 mL/min. The fractions corresponding to non-aggregated LDL were subjected to binding studies.
Data Analysis
Curve-fit analysis was performed in GraphPad Prism version 3.00 for Windows using a nonlinear curve-fit assuming one-site binding (San Diego, Calif). Statistical analysis was performed using SigmaStat for Windows version 2.03 (SPSS, Chicago, Ill).
Results
Generation and Characterization of Recombinant RK3148–3158SQ LDL
To evaluate the importance of site A (residues 3148-3158) in the interaction between native apoB100-containing LDL and proteoglycans, we generated transgenic mice expressing human recombinant LDL in which site A in apoB100 was mutated by substituting serines and glutamines for arginines and lysines, respectively (RK3148-3158SQ). Recombinant control LDL, RK3148-3158SQ LDL, and RK3359-3369SA LDL19 were isolated from mouse plasma by ultracentrifugation, and lipoproteins containing endogenous mouse apoE and apoB were removed by immunoaffinity chromatography; the removal of all endogenous apoB and apoE was confirmed by Western blot analysis (not shown). In RK3359-3369SA LDL, the arginine residues were converted to serines and the lysine residues were changed to alanines in site B of apoB100.19 Previously, we showed that site B (residues 3359-3369) is the combined proteoglycan binding and LDL receptor binding site in apoB100 and that RK3359-3369SA LDL do not interact with proteoglycans or the LDL receptor.7
The isolated recombinant control LDL, RK3148-3158SQ LDL, and RK3359-3369SA LDL had almost identical lipid compositions (data not shown) and particle diameters (27.1±0.86, 26.7±0.20, and 26.8±0.85 nm, respectively; mean±SD).
To determine the ability of the recombinant LDL to interact with proteoglycans, we subjected recombinant control LDL, RK3148-3158SQ LDL, and RK3359-3369SA LDL to solid-phase assays with biglycan (Figure 1A) and decorin (not shown). In two independent experiments, recombinant control LDL and recombinant RK3148-3158SQ LDL bound biglycan with high affinity, whereas recombinant RK3359-369SA LDL displayed severely impaired binding (Figure 1A).
Figure 1. Solid-phase assays of the binding of LDL to biglycan (A) and competitive receptor binding assay of recombinant LDL (B). The abilities of recombinant control LDL (?), RK3148-3158SQ LDL (), and RK3359-3369SA LDL () to bind biglycan or compete with 125I-labeled human plasma LDL (2 μg/mL) for binding to LDL receptors on normal human fibroblasts were determined. The recombinant lipoproteins were isolated from 20 mice, and endogenous apoE and apoB were removed by affinity chromatography. The results represent mean values (±SD) from 2 independent experiments, each performed with freshly isolated LDL (n=4 in each experiment) (A) or the mean of 2 independent experiments, each performed in duplicate (B).
To investigate the importance of site A in the interaction with the LDL receptor, we isolated recombinant control LDL, RK3148-3158SQ LDL, and RK3359-3369SA LDL and tested them in an in vitro competitive receptor-binding assay (Figure 1B). Recombinant control LDL and RK3148-3158SQ LDL had almost identical receptor binding (ED50 2.1 and 2.3 μg/mL, respectively), whereas recombinant RK3359-3369SA LDL displayed defective receptor binding (ED50 >20 μg/mL). Taken together, these results indicate that site A in native LDL is not involved in binding to proteoglycans or to the LDL receptor.
Solid-Phase Assays of Recombinant LDL Isolated From sPLA2 Transgenic Mice
Next, we tested the hypothesis that site A can modulate the proteoglycan binding activity of modified LDL. Recombinant control LDL, RK3148-3158SQ LDL, and RK3359-3369SA LDL were isolated by ultracentrifugation from human apoB transgenic mice and from doubly transgenic mice expressing human apoB and sPLA2.23 After removal of endogenous mouse apoE-containing and apoB-containing lipoproteins by immunoaffinity chromatography, the isolated recombinant LDL were subjected to solid-phase assays with biglycan (Figure 2A) and decorin (not shown). In two independent experiments, recombinant control LDL from sPLA2 transgenic mice displayed increased binding, whereas recombinant RK3148-3158SQ LDL from sPLA2 transgenic mice and sPLA2 nontransgenic littermate controls had almost identical binding (Figure 2A). These results indicate that site A is important for the increased interaction of sPLA2-modified LDL with glycosaminoglycans. However, RK3359-3369SA LDL did not interact with biglycan, regardless of whether it was isolated from sPLA2 transgenic mice or not (Figure 2B). These results indicate that site A in sPLA2-modified LDL modulates the binding of the LDL particle to proteoglycans but cannot substitute for site B, which is essential for the interaction with proteoglycans.
Figure 2. Binding of recombinant LDL isolated from sPLA2 transgenic mice to biglycan. The abilities of recombinant control LDL (?) and RK3148-3158SQ LDL () and RK3359-3369SA LDL () isolated from human apoB transgenic mice (?, , ) and from apoBxsPLA2 transgenic mice (, , ) to bind biglycan were determined. Endogenous mouse apoB and apoE were removed from all lipoproteins by immunoaffinity chromatography. The results represent mean values (±SD) from two independent experiments, each performed with freshly isolated LDL from 15 mice of each group (n=4 in each experiment).
Characterization of sPLA2-Modified Recombinant LDL
sPLA2 hydrolyzes glyceroacyl phospholipids to form nonesterified fatty acids and lysophospholipids. Therefore, we analyzed the phospholipid composition of recombinant LDL isolated from human apoB transgenic mice and from apoB transgenic mice with mice expressing sPLA2 (apoBxsPLA2 transgenic mice). The sPLA2-modifed LDL contained more lysophosphatidylcholine than nonmodified recombinant LDL (Table 1). We also analyzed the content of total cholesterol, free cholesterol, cholesteryl esters, and triglycerides. No differences were found in the content of these lipids in recombinant LDL isolated from apoB transgenic mice or from apoBxsPLA2 transgenic mice (Table 2). Neither the phospholipid composition (Table 1) nor the content of the other lipids (Table 2) was affected by the type of apoB100 (control or the two mutations, ie, RK3148-3158SQ or RK3359-3369SA) in LDL isolated from apoB transgenic mice or from apoBxsPLA2 transgenic mice (data not shown).
TABLE 1. Phospholipid Composition of Recombinant LDL Isolated From Human apoB Transgenic Mice and From Double-Transgenic apoBxsPLA2 Mice
TABLE 2. Lipid Composition of Recombinant LDL Isolated From Human apoB Transgenic Mice and From Double-Transgenic apoBxsPLA2 Mice
Dynamic light-scattering analysis of recombinant control LDL isolated from human apoB transgenic mice and from apoBxsPLA2 transgenic mice showed that recombinant LDL from the singly human apoB transgenic mice was significantly larger (26.7±0.42 versus 25.8±0.38 nm;, mean ± SD; P<0.05). Taken together, these results show that in vivo modification of recombinant LDL in sPLA2 transgenic mice induces the formation of smaller LDL enriched in lysophosphatidylcholine. Thus, it is possible that the conformational change of apoB in LDL modified with sPLA2 in vivo is caused by the formation of smaller LDL and/or by changes in the phospholipids composition of the LDL particle.
Solid-Phase Assays of Recombinant LDL Isolated From
sPLA2 Transgenic Mice by Size-Exclusion Chromatography
To discriminate between these possibilities, recombinant control LDL and RK3148-3158SQ LDL were isolated from human apoB transgenic mice and from apoBxsPLA2 transgenic mice by a combination of ultracentrifugation and size-exclusion chromatography and subjected to solid-phase assays with biglycan. The isolated LDL had almost identical size (26.4±0.23 and 26.3±0.15 nm). In two independent experiments, recombinant control LDL from sPLA2 transgenic mice displayed increased binding, whereas recombinant RK3148-3158SQ LDL from sPLA2 transgenic mice and sPLA2 nontransgenic littermate controls had almost identical binding. The Kd for control LDL and RK3148-3158SQ LDL isolated from apoB transgenic mice were 16.3 and 17.1, respectively, and from apoBxsPLA2 transgenic mice 10.3 and 16.6, respectively. The results also showed that the diameter of sPLA2-modified and non-modified recombinant control LDL were inversely correlated to the binding affinity to biglycan; smaller LDL had higher binding affinity to biglycan than larger LDL (data not shown). Taken together, the results show that the conformational change of apoB in sPLA2-modified LDL is caused by the formation of smaller LDL and by changes in the phospholipid composition of the LDL particle.
To investigate if in vitro digestion and in vivo modification of LDL with PLA2 give similar results, recombinant control LDL, RK3359-3369SA LDL, and RK3148-3158SQ LDL were isolated and digested in vitro with PLA2 for 3 hours. Samples of the reaction were separated by size-exclusion chromatography, and non-aggregated LDL was isolated and subjected to binding studies. This was performed because lipolysis of LDL with PLA2 in vitro induces particle aggregation.22 The results from these studies showed that in vitro modification of recombinant LDL with PLA2 gave similar results as in vivo modification of LDL, ie, binding affinity increased on PLA2 digestion and site A was required for the increased interaction of sPLA2-modified LDL with glycosaminoglycans (data not shown). However, the intra-assay variations (ie, imprecision) and the inter-assay variations (ie, irreproducibility) were much more pronounced when analyzing the binding of LDL digested with PLA2 in vitro compared with LDL modified in vivo.
Binding of Lion LDL to Biglycan and to the LDL Receptor
Our data indicate that site A becomes exposed and functional on sPLA2 modification and that site A and site B act cooperatively in binding to proteoglycans. In the human sequence, site A and site B are separated by 202 amino acids, but a disulfide link between Cys-3167 and Cys-3297 has been postulated to bring the two binding sites into close proximity.10 However, these cysteine residues are not conserved in other species;6 but because of a naturally occurring deletion between site A and site B in several carnivores, including lions, the binding sites are separated by 76 amino acids only in these species.24 Therefore, we investigated the proteoglycan-binding activity of lion LDL with and without PLA2 digestion.
Lion LDL and human LDL were isolated and digested in vitro with PLA2 for 3 hours and non-aggregated LDL was isolated by a combination of ultracentrifugation and size-exclusion chromatography. The PLA2 modification induced the formation of smaller LDL enriched in lysophosphatidylcholine. The diameter of lion LDL and human LDL decreased from 27.2±0.33 to 26.1±0.25 nm and from 27.5±0.39 to 26.3±0.15 nm, respectively. The content of lysophosphatidylcholine in PLA2-digested LDL increased 2.4-fold±0.3 and 2.8-fold±0.6, respectively.
As shown by SDS-PAGE, lion apoB migrated faster than human apoB100, and the LDL did not contain any apolipoprotein other than apoB (Figure 3A). In solid-phase assays, lion LDL bound with higher affinity than human LDL to biglycan; however, PLA2 modification did not increase the binding of lion LDL to biglycan, as was seen in human LDL (Figure 4A). The Kd for lion and human LDL were 10.2 and 15.4, respectively, and for PLA2-modified lion and human LDL 10.4 and 11.3, respectively. The binding capacity of lion and human LDL was similar (Bmax 0.71 and 0.66, respectively), indicating that the deletion between sites A and B increases the affinity for proteoglycans but not the binding capacity. In competitive LDL receptor-binding studies with human 125I-labeled LDL, lion and human LDL had almost identical receptor binding (ED50 1.9 and 1.8 μg/mL, respectively). These results are consistent with the hypothesis that site A is functional on lion LDL with or without PLA2 modification.
Figure 3. Coomassie staining and Western analysis of apoB-containing LDL. A, Lion and human plasma LDL (density=1.02 to 1.05 g/mL) were isolated by sequential ultracentrifugation. Five micrograms of LDL were separated by SDS-PAGE with 3% to 15% gels. Arrows indicate molecular mass markers. B, Recombinant LDL (density=1.02 to 1.05 g/mL) was isolated from LDLR-/- mice (HuApoBxLDLR-/-) or LDLR+/+ mice (HuApoB), and 2 μg were analyzed on SDS-PAGE with 3% to 15% gels. C, Endogenous apoB was removed by immunoaffinity chromatography, and 0.5 μg of recombinant control LDL (density=1.02 to 1.05 g/mL) from LDLR-/- mice (lane 1) or LDLR+/+ mice (lane 2) was analyzed by Western blots with polyclonal antibodies against mouse apoE. Lane 3 shows recombinant control LDL (0.5 μg) from LDLR-/- mice analyzed by Western blots with polyclonal antibodies against mouse apoE (upper panel) and human apoE (lower panel) after incubation with CHD-modified human apoE (HuApoE competition). The CHD-modified samples were incubated with an equal volume of 1 mol/L hydroxylamine, 0.3 mol/L mannitol, at pH 7.0, at 37°C for 16 hours to reverse the modification of the arginine residues before sodium dodecyl sulfate-polyacrylamide gel electrophoresis.
Figure 4. Solid-phase assays of lion LDL or recombinant LDL isolated from Ldlr+/+ and Ldlr-/- mice to biglycan binding to biglycan. A, Lion () and human (?) plasma LDL (density=1.02 to 1.05 g/mL) were isolated by sequential ultracentrifugation and incubated with ( and ) or without (? and ) 50 ng/mL of PLA2 (from bee venom). To prevent overlap in the figure, and have been moved by -0.5 and +0.5 U on the X-axis, respectively. B, The abilities of recombinant control LDL (?) and RK3148-3158SQ LDL () isolated from Ldlr+/+ ( and ) and Ldlr-/- (? and ) mice to biglycan were determined. Endogenous apoE was replaced with cyclohexanedione-modified human apoE, and endogenous mouse apoB was removed by immunoaffinity chromatography. The results represent mean values (±SD) from two independent experiments, each performed with freshly isolated LDL from 15 mice of each group (n=4 in each experiment).
Solid-Phase Assays of Cholesterol-Enriched ApoB100 Recombinant LDL With Biglycan
To analyze if cholesterol enrichment of LDL is associated with increased affinity for proteoglycans, human apoB transgenic mice expressing recombinant control LDL, RK3148-3158SQ LDL, or RK3359-3369SA LDL were bred onto Ldlr-/- background. Recombinant LDL isolated from Ldlr-/- mice is enriched with cholesteryl esters. The total cholesterol to triglyceride (in mmol/L) ratios in recombinant LDL isolated from human apoBxLdlr+/+ transgenic mice and apoBxLdlr-/- transgenic mice are 1.7 and 4.2, respectively.
Recombinant control LDL and RK3148-3158SQ LDL from Ldlr+/+ and Ldlr-/- mice were isolated by ultracentrifugation, and endogenous mouse apoB and apoE were removed. LDL isolated from Ldlr-/- mice is greatly enriched in endogenous apoE (Figure 3B), and immunoaffinity techniques fail to remove all endogenous apoE from LDL isolated from Ldlr-/- mice.1,2 Competition of mouse apoE with CHD-modified human apoE efficiently removes the endogenous apoE from the recombinant LDL (Figure 3C), and the CHD-modified apoE fail to facilitate an indirect binding between the recombinant LDL and glycosaminoglycans.1,2 The competition of mouse apoE with chemically modified human apoE did not affect the cholesterol or triglyceride content of the LDL particles. The recovery of these lipids were 94%±8% and 92%±11% after competition with CHD-modified apoE, respectively (mean±SD, n=5).
The purified recombinant control LDL and RK3148-3158SQ LDL from Ldlr+/+ and Ldlr-/- mice were then subjected to size-exclusion chromatography to isolate LDL particles with almost identical size (diameters: 26.8±0.84 and 26.7±0.21 nm versus 26.9±1.87 and 26.8±0.42 nm, respectively; mean±SD). In two independent experiments, recombinant control LDL and RK3148-3158SQ LDL isolated from Ldlr-/- mice bound biglycan and decorin (not shown) with higher affinity than recombinant control LDL and RK3148-3158SQ LDL isolated from Ldlr+/+ mice (Figure 4B). Within each genotype, recombinant control LDL and RK3148-3158SQ LDL displayed almost identical binding. The Kd for control LDL and RK3148-3158SQ LDL isolated from Ldlr+/+ were 17.0 and 16.7, respectively, and from Ldlr-/- mice 9.1 and 8.2, respectively. The binding capacity of recombinant control LDL and RK3148-3158SQ LDL was almost identical, both when isolated from Ldlr+/+ and from Ldlr-/- mice (Bmax were 0.51, 0.47, 0.51, and 0.48, respectively). Control experiments showed that RK3359-3369SA LDL from both Ldlr+/+ and Ldlr-/- mice had severely impaired binding (data not shown). Competitive receptor-binding studies with human 125I-labeled LDL showed that cholesterol-enriched recombinant control LDL from Ldlr-/- mice bound with higher affinity to the LDL receptor than that from Ldlr+/+ mice (ED50 1.7 versus 2.5 μg/mL).
These findings indicate that cholesterol enrichment of the LDL particle is associated with increased affinity for proteoglycans and for the LDL receptor. This mechanism is likely mediated by a conformational change of site B and is independent of site A in apoB100.
Discussion
The aim of this study was to investigate the molecular mechanism for changes in proteoglycan binding and LDL receptor affinity on two compositional changes in LDL: modification by secretory group IIA phospholipase A2 (sPLA2) of the surface and cholesterol-enrichment of the core.
Elevated plasma sPLA2 activity has been shown to be an important risk factor for coronary heart disease.8 Thus, clarification of the proatherogenic mechanisms by which sPLA2 acts in the plasma and in the artery wall is important to our understanding of the pathogenesis of the disease. To investigate the molecular mechanism for how circulating sPLA2 induces the increased binding affinities of LDL to glycosaminoglycans,13 we crossed our human apoB transgenic mice with mice expressing sPLA2.23 The serum activity of sPLA2 in these transgenic mice is elevated approximately 8-fold compared with nontransgenic littermates, but there is no evidence of systemic inflammation in the transgenic mice.23 Here we show that in vivo modification of recombinant LDL in sPLA2 transgenic mice induces the formation of slightly smaller LDL enriched with lysophosphatidylcholine. These results are in agreement with data of Grass et al23 and with results from patients with rheumatoid arthritis who have increased plasma levels of sPLA2.13 Taken together, our results are consistent with the hypothesis that sPLA2 modification of LDL in the circulation leads to exposure of site A, resulting in increased proteoglycan binding. Furthermore, the conformational change of apoB in sPLA2-modified LDL is caused by the formation of smaller LDL and by changes in the phospholipid composition of the LDL particle.
However, it is conceivable that some of the observed effects could result from indirect secondary consequences. For example, sPLA2 modification might cause LDL with low PG binding to be cleared, leaving behind a high-affinity subpopulation. It is also possible that sPLA2 digestion in vivo allows a second enzyme to further alter the LDL, eg, sphingomyelinase, which has been reported to readily digest sPLA2-treated LDL.25 However, because a gentle PLA2 digestion of recombinant LDL in vitro gives similar results, and because isolated LDL from apoBxsPLA2 transgenic mice contain abundant sphingomyelin, these alternative explanations seems unlikely.
In some experiments, LDL was digested with PLA2 in vitro. These incubations were performed for 3 hours only. The reason for this was to achieve similar modification of LDL as in patients with increased plasma levels of sPLA2.13 Prolonged digestion of LDL with PLA2 in vitro results in hydrolysis of nearly all of the phosphatidylcholines, in pronounced aggregation, and in the presence of glycosaminoglycans fusion of LDL,22 ie, the modifications of LDL that are more likely to occur within the extracellular matrix of the arterial wall.
The findings that lion LDL, with a 126-amino acid residue deletion between site A and site B, has increased affinity for artery wall proteoglycans and that PLA2 modification of lion LDL did not increase the binding of lion LDL to biglycan are consistent with the hypothesis that site A on lion LDL is also exposed with or without sPLA2 modification. However, studies with cross-linking or some other direct proof would be needed to draw this conclusion. It is also possible that other basic sequences or even other apolipoproteins might play a role in the unusual properties of lion LDL. However, the latter possibility seems unlikely because no apolipoproteins other than apoB was detected in isolated lion LDL (Figure 3A). The fact that two independent deletions have been found in apoB100 between site A and site B might indicate that these mutations provide an evolutionary advantage. In addition to the 126-amino acid deletion in several carnivores, including lions,24 an 84-amino acid deletion has been found in elephants.24 A tentative hypothesis is that an enhanced binding of LDL to matrix might promote wound healing, as has been proposed, although not proven, for Lp(a).26
Our results show that cholesterol-enriched LDL has increased affinity for glycosaminoglycans and the LDL receptor. This mechanism is likely mediated by a conformational change of site B and is, in contrast to sPLA2-modified LDL, independent of site A. The finding that cholesterol-enriched LDL have increased affinity for artery wall proteoglycans offers an additional explanation for why the atherogenicity of apoB-containing lipoproteins correlates with their cholesteryl ester content.
The binding studies with LDL isolated from Ldlr-/- and Ldlr+/+ mice were performed with LDL of almost identical diameters, and the results are consistent with the triglyceride-dependent decrease in affinity for the LDL receptor reported in hypertriglyceridemic subjects.27 However, our results are not consistent with the finding that in the absence of changes in the size of the particle, the relative content of neutral lipid in the core of LDL does not significantly affect apoB conformation or its affinity for the LDL receptor.28 The reasons for these differences are unclear but could reflect differences in experimental design, because the secondary structure of apoB in LDL strongly depends on the physical state of the lipoprotein core.29
In this study, we tested the hypothesis that site A is important for proteoglycan binding of sPLA2-modified LDL. The results confirm that site B is the principal proteoglycan-binding site in native apoB100-containing LDL and show that site A acts cooperatively with site B in the association with proteoglycans in sPLA2-modified LDL. The results also show that cholesterol enrichment of the LDL particle is associated with increased affinity for proteoglycans and for the LDL receptor. This mechanism is likely mediated by a conformational change of site B and is independent of site A in apoB100. Taken together, the results show that LDL affinity for proteoglycans can be modified by size-dependent effects and size-independent compositional effects of either the surface or the core.
Acknowledgments
The authors thank K. Sk?lén, E. Bj?rk, L. Lindgren, M. Heyden, B. Gabrielli, and K. Arnold for excellent technical assistance, S. Ordway and G. Howard for editorial assistance, and Dr Torsten M?ller and Karin Magnusson at K?lmorden Zoo for lion plasma.
This work was supported by the Swedish Medical Research Council, The Swedish Foundation for Strategic Research, and The Swedish Heart–Lung Foundation.
References
Williams KJ, Tabas I. The response-to-retention hypothesis of early atherogenesis. Arterioscler Thromb Vasc Biol. 1995; 15: 551–561.
Skalen K, Gustafsson M, Rydberg EK, Hulten LM, Wiklund O, Innerarity TL, Boren J. Subendothelial retention of atherogenic lipoproteins in early atherosclerosis. Nature. 2002; 417: 750–754.
Camejo G, Olofsson SO, Lopez F, Carlsson P, Bondjers G. Identification of Apo B-100 segments mediating the interaction of low density lipoproteins with arterial proteoglycans. Arteriosclerosis. 1988; 8: 368–377.
Weisgraber KH, Rall SC, Jr. Human apolipoprotein B-100 heparin-binding sites. J Biol Chem. 1987; 262: 11097–11103.
Hirose N, Blankenship DT, Krivanek MA, Jackson RL, Cardin AD. Isolation and characterization of four heparin-binding cyanogen bromide peptides of human plasma apolipoprotein B. Biochemistry. 1987; 26: 5505–5512.
Chan L. Apolipoprotein B, the major protein component of triglyceride-rich and low density lipoproteins. J Biol Chem. 1992; 267: 25621–25624.
Law A, Scott J. A cross-species comparison of the apolipoprotein B domain that binds to the LDL receptor. J Lipid Res. 1990; 31: 1109–1120.
Boren J, Olin K, Lee I, Chait A, Wight T, Innerarity T. Identification of the principal proteoglycan binding site in LDL: A single point mutation in apolipoprotein B100 severely affects proteoglycan interaction without affecting LDL receptor binding. J Clin Invest. 1998; 101: 2658–2664.
Olsson U, Camejo G, Hurt-Camejo E, Elfsber K, Wiklund O, Bondjers G. Possible functional interactions of apolipoprotein B-100 segments that associate with cell proteoglycans and the ApoB/E receptor. Arterioscler Thromb Vasc Biol. 1997; 17: 149–155.
Kugiyama K, Ota Y, Takazoe K, Moriyama Y, Kawano H, Miyao Y, Sakamoto T, Soejima H, Ogawa H, Doi H, Sugiyama S, Yasue H. Circulating levels of secretory type II phospholipase A(2) predict coronary events in patients with coronary artery disease. Circulation. 1999; 100: 1280–1284.
Kleinman Y, Krul ES, Burnes M, Aronson W, Pfleger B, Schonfeld G. Lipolysis of LDL with phospholipase A2 alters the expression of selected apoB-100 epitopes and the interaction of LDL with cells. J Lipid Res. 1988; 29: 729–743.
Sartipy P, Camejo G, Svensson L, Hurt-Camejo E. Phospholipase A(2) modification of low density lipoproteins forms small high density particles with increased affinity for proteoglycans and glycosaminoglycans. J Biol Chem. 1999; 274: 25913–25920.
Hurt-Camejo E, Paredes S, Masana L, Camejo G, Sartipy P, Rosengren B, Pedreno J, Vallve JC, Benito P, Wiklund O. Elevated levels of small, low-density lipoprotein with high affinity for arterial matrix components in patients with rheumatoid arthritis: possible contribution of phospholipase A2 to this atherogenic profile. Arthritis Rheum. 2001; 44: 2761–2767.
Willner EL, Tow B, Buhman KK, Wilson M, Sanan DA, Rudel LL, Farese RV, Jr. Deficiency of acyl CoA:cholesterol acyltransferase 2 prevents atherosclerosis in apolipoprotein E-deficient mice. Proc Natl Acad Sci U S A. 2003; 100: 1262–1267.
Aviram M, Lund-Katz S, Phillips MC, Chait A. The influence of the triglyceride content of low density lipoprotein on the interaction of apolipoprotein B-100 with cells. J Biol Chem. 1988; 263: 16842–16848.
Linton MF, Farese RV, Jr., Chiesa G, Grass DS, Chin P, Hammer RE, Hobbs HH, Young SG. Transgenic mice expressing high plasma concentrations of human apolipoprotein B100 and lipoprotein(a). J Clin Invest. 1993; 92: 3029–3037.
Boren J, Lee I, Callow MJ, Rubin EM, Innerarity TL. A simple and efficient method for making site-directed mutants, deletions, and fusions of large DNA such as P1 and BAC clones. Genome Res. 1996; 6: 1123–1130.
Wang X, Chauhan V, Nguyen AT, Schultz J, Davignon J, Young SG, Boren J, Innerarity TL, Rutai H, Milne RW. Immunochemical evidence that human apoB differs when expressed in rodent versus human cells. J Lipid Res. 2003; 44: 547–553.
Boren J, Lee I, Zhu W, Arnold K, Taylor S, Innerarity TL. Identification of the low density lipoprotein receptor-binding site in apolipoprotein B100 and the modulation of its binding activity by the carboxyl terminus in familial defective apoB100. J Clin Invest. 1998; 101: 1084–1093.
Flood C, Gustafsson M, Richardson PE, Harvey SC, Segrest JP, Boren J. Identification of the proteoglycan binding site in apolipoprotein B48. J Biol Chem. 2002; 277: 32228–32233.
Walzem RL, Davis PA, Hansen RJ. Overfeeding increases very low density lipoprotein diameter and causes the appearance of a unique lipoprotein particle in association with failed yolk deposition. J. Lipid Res. 1994; 35: 1354–1366.
Hakala JK, Oorni K, Ala-Korpela M, Kovanen PT. Lipolytic modification of LDL by phospholipase A2 induces particle aggregation in the absence and fusion in the presence of heparin. Arterioscler Thromb Vasc Biol. 1999; 19: 1276–1283.
Grass DS, Felkner RH, Chiang MY, Wallace RE, Nevalainen TJ, Bennett CF, Swanson ME. Expression of human group II PLA2 in transgenic mice results in epidermal hyperplasia in the absence of inflammatory infiltrate. J Clin Invest. 1996; 97: 2233–2241.
Amrine-Madsen H, Koepfli KP, Wayne RK, Springer MS. A new phylogenetic marker, apolipoprotein B, provides compelling evidence for eutherian relationships. Mol Phylogenet Evol. 2003; 28: 225–240.
Schissel SL, Keesler GA, Schuchman EH, Williams KJ, Tabas I. The cellular trafficking and zinc dependence of secretory and lysosomal sphingomyelinase, two products of the acid sphingomyelinase gene. J Biol Chem. 1998; 273: 18250–18259.
Yano Y, Shimokawa K, Okada Y, Noma A. Immunolocalization of lipoprotein(a) in wounded tissues. J Histochem Cytochem. 1997; 45: 559–568.
McKeone BJ, Patsch JR, Pownall HJ. Plasma triglycerides determine low density lipoprotein composition, physical properties, and cell-specific binding in cultured cells. J Clin Invest. 1993; 91: 1926–1933.
Galeano NF, Milne R, Marcel YL, Walsh MT, Levy E, Ngu’yen TD, Gleeson A, Arad Y, Witte L, Al-Haideri M, Rumsey SC, Deckelbaum RJ. Apoprotein B structure and receptor recognition of triglyceride-rich low density lipoprotein (LDL) is modified in small LDL but not in triglyceride-rich LDL of normal size. J Biol Chem. 1994; 269: 511–519.
Coronado-Gray A, van Antwerpen R. The physical state of the LDL core influences the conformation of apolipoprotein B-100 on the lipoprotein surface. FEBS Lett. 2003; 533: 21–24.(Christofer Flood; Maria G)