Search

Caveolae are 50-100 nm plasma membrane invaginations, which function in cell signaling, in transcytosis, and in regulating cellular cholesterol homeostasis. These subcompartments of the plasma membrane are characterized by the presence of caveolin proteins. Recent studies have indicated that caveolae may be involved in the regulation of cellular cholesterol efflux to high-density lipoproteins (HDL), as well as selective cholesteryl ester uptake mediated by scavenger receptor class B type I (SR-BI). In the present studies, we show that caveolin-1 expression in HEK-293T cells has no effect on SR-BI-mediated cellular cholesterol efflux to reconstituted HDL. However, SR-BI-mediated selective cholesteryl ester uptake is significantly inhibited by caveolin-1. Interestingly, we also found that SR-BI, but not CD36, can induce the dramatic stabilization of the caveolin-1 protein, independently of its transcriptional control. On the other hand, caveolin-1 has little effect on SR-BI stability, but clearly increases CD36 stability. Since SR-BI expression has been shown to increase cellular cholesterol levels, we next examined the effect of cholesterol itself on caveolin-1 stabilization and localization. When cells were loaded with cholesterol, we observed the dramatic stabilization of caveolin-1 with significant clustering of caveolin-1 at the cell surface. In addition, a palmitoylation-deficient caveolin-1 mutant was still responsive to cholesterol-induced stabilization, indicating that palmitoylation of caveolin-1 is not required for the cholesterol-induced stabilization of caveolin-1. These results suggest an important role for cholesterol and SR-BI in the regulation of caveolin functioning, especially in cell types such as endothelial cells and macrophages, which can be dramatically affected by changes in their cholesterol content during the development of atherosclerosis.

The contribution of the amphipathic alpha-helices of apoA-I toward lipid efflux from human skin fibroblasts and macrophage was examined. Four apoA-I mutants were designed, each by deletion of a pair of predicted adjacent helices. Three mutants lacked two consecutive central alpha-helices [Delta(100-143), Delta(122-165), and Delta(144-186)], whereas the final mutant lacked the C-terminal domain [Delta(187-243)]. When compared to recombinant wild-type apoA-I and mutants with central domain deletions, Delta(187-243) exhibited a marked reduction in its ability to promote either cholesterol or phospholipid efflux from THP-1 macrophages. This mutant also demonstrated a decreased ability to bind lipids and to form lipoprotein complexes. In contrast, the four mutants and apoA-I equally supported cholesterol efflux from fibroblasts, albeit with a reduced capacity when compared to macrophages. Delta(187-243) bound poorly to the macrophage cell surface when compared to apoA-I, and competitive binding studies with the central domain and C-terminal deletions mutants showed that only Delta(187-243) did not compete effectively with [(125)I]apoA-I. Omission of PMA during cholesterol loading enhanced cholesterol efflux to both apoA-I (1.5-fold) and the C-terminal deletion mutant (2.5-fold). Inclusion of the Sandoz ACAT inhibitor (58-035) during loading and, in the absence of PMA, increased and equalized cholesterol efflux to apoA-I and Delta(187-243). Surprisingly, omission of PMA during cholesterol loading had minimal effects on the binding of apoA-I or Delta(187-243) to the THP-1 cell surface. Overall, these results show that cholesterol efflux from cells such as fibroblasts does not require any specific sequence between residues 100 and 243 of apoA-I. In contrast, optimal cholesterol efflux in macrophages requires binding of the C-terminal domain of apoA-I to a cell surface-binding site and the subsequent translocation of intracellular cholesterol to an efflux-competent pool.

A unique class of lipid-poor high-density lipoprotein, pre-beta1 HDL, has been identified and shown to have distinct functional characteristics associated with intravascular cholesterol transport. In this study we have characterized the structure/function properties of poorly lipidated HDL particles and the factors that mediate their conversion into multimolecular lipoprotein particles. Studies were undertaken with homogeneous recombinant HDL particles (LpA-I) containing apolipoprotein (apo) A-I and various amounts of palmitoyloleoylphosphatidylcholine (PC) and cholesterol. Complexation of apoA-I with small amounts of PC and cholesterol results in the formation of discrete lipoprotein structures that have a hydrated diameter of about 6 nm but contain only one molecule of apoA-I (Lp1A-I). While the molecular charge and alpha-helix content of apoA-I are unaffected by lipidation, the thermodynamic stability of the protein is reduced significantly (from 2.4 to 0.9 kcal/mol of apoA-I). Evaluation of apoA-I conformation by competitive radioimmunoassay with monoclonal antibodies shows that addition of small amounts of PC and cholesterol to apoA-I significantly increases the immunoreactivity of a number of domains over the entire molecule. Increasing the ratio of PC:apoA-I to 10:1 in the Lp1A-I complex is associated with increases in the alpha-helix content and stability of apoA-I. However, incorporation of 10-15 mol of PC destabilizes the Lp1A-I complex and promotes the formation of more thermodynamically stable (1.8 kcal/mol of apoA-I) bimolecular structures (Lp2A-I) that are approximately 8 nm in diameter. The formation of an Lp2A-I particle is associated with an increased immunoreactivity of most of the epitopes studied, with the exception of one central domain (residues 98-121), which becomes significantly less exposed. This structural change parallels a significant increase in the net negative charge on the complex. Characterization of the ability of these lipoproteins to act as substrates for lecithin:cholesterol acyltransferase (LCAT) shows that unstable Lp1A-I complexes stimulate a higher rate of cholesterol esterification by LCAT than the small but more stable Lp2A-I particles (Vmax values are 5.8 and 0.3 nmol of free cholesterol esterified/h, respectively). The ability of LCAT to interact with lipid-poor apoA-I suggests that LCAT does not need to bind to the lipid interface on an HDL particle but that LCAT may directly interact with apoA-I. The data suggests that lipid-poor HDL particles may be metabolically reactive particles because they are thermodynamically unstable.

We have investigated the association of apolipoprotein E (apoE) with the HepG2 cell surface (i.e. plasma membrane and extracellular matrix) using domain specific monoclonal antibodies against apoE. Growth in beta-D-xyloside decreased the incorporation of 35S into glycosaminoglycans by 31% and cell surface apoE by 45% with a concomitant increase in apoE secretion (4.3-fold), underlining the importance of glycosaminoglycan association of apoE. Heparinase (3-10 U/mL) or heparin (1 mg/mL) decreased apoE by 25 and 30.5%, respectively, which suggests that some apoE is associated with cell surface heparan sulfate proteoglycans. Chondroitinase ABC (1.5 U/mL) reduced cell surface apoE by 40%, indicating that a major pool of apoE is associated with chondroitin sulfate proteoglycans. Further enzymatic and displacement analysis suggested that cell surface apoE associates specifically with GAGs containing chondroitin-4-sulfates. 3H1, a monoclonal antibody that recognizes an epitope within the lipid-binding C-terminal domain of apoE, decreased binding of apoE to chondroitin sulfate proteoglycans in solid-phase assays by 77% and to heparan sulfate proteoglycans by 46%, suggesting that this region is of increased importance for binding to chondroitin sulfate proteoglycans. Previous studies with 3H1 demonstrated that apoE of the extracellular matrix is lipid-poor (Burgess, J. W., Gould, D. R., and Marcel, Y. L. (1998) J. Biol. Chem. 273, 5645-5654), but we show here that apoE on the remaining cell surface is lipid-associated. In summary, lipidated apoE associates with the HepG2 plasma membrane through interactions with chondroitin-4-sulfate containing GAGs and, to a lesser extent, HSPG.

We have studied the role of amphipathic alpha-helices in the ability of apoA-I to promote cholesterol efflux from human skin fibroblasts and activate lecithin:cholesterol acyltransferase (LCAT). Three apoA-I mutants were designed, each by deletion of a pair of predicted adjacent central alpha-helices [Delta(100-143), Delta(122-165), Delta(144-186)], and expressed in Escherichia coli. This strategy was used to minimize disruption of the predicted secondary structure of the resulting protein. These three central deletion mutants have been previously shown to be expressed as stable folded proteins but to exhibit altered phospholipid-binding properties. When recombined with phospholipids to form homogeneous LpA-I containing equivalent amounts of POPC and tested for their ability to promote diffusional cholesterol efflux from normal [3H]cholesterol-labeled fibroblasts, each mutant and the wild-type recombinant protein (Rec.-apoA-I) promoted cholesterol efflux with very similar rates at all the concentrations tested. These experiments showed that all LpA-I could acquire cellular cholesterol with similar affinity and binding capacity. However, when the cell-incubated LpA-I were incubated with purified LCAT, two mutants, Delta(122-165) and Delta(144-186), appeared incapable of activating the enzyme. To directly determine their ability to activate LCAT, each mutant and the control were recombined with equivalent amounts of cholesterol and phospholipid and incubated with the purified enzyme. The results show that whereas deletion of residues 100-143 has little effect on LCAT activation, deletion of residues 122-165 or 144-186 results in an inability of the mutants to promote cholesterol esterification. In conclusion, our results show that no specific sequence in the central domain of apoA-I is required for efficient diffusional cholesterol efflux from normal fibroblasts; however, residues 144-186 appear critical for optimum LCAT activation and cholesteryl ester accumulation. Since deletion of residues 144-186 also perturbs phospholipid association and prevents the formation of large LpA-I particles [Frank, P. G., Bergeron, J., Emmanuel, F., Lavigne, J. P., Sparks, D. L., Denèfle, P., Rassart, E., and Marcel, Y. L. (1997) Biochemistry 36, 1798-1806], the data show that this pair of alpha-helices plays an important role in the maturation of HDL. Sequence analysis of these apoA-I helices further identifies specific residues that appear essential to this activity.

In order to better understand the structure-function properties of apolipoprotein (apo) A-I, we have constructed and expressed three apoA-I mutants using a system previously described for the expression of human apolipoprotein A-I (Rec.-apoA-I). These mutants (corresponding to deletion of apoA-I residues 100-143, 122-165, 144-186) have been studied for their ability to form reconstituted apoA-I-containing lipoproteins (LpA-I) with POPC and DMPC, and for their structural and physical properties. Rec.- and native apoA-I can form homogeneous discoidal Lp2A-I over a wide range of POPC/apoA-I ratios [(20-130)/1] and exhibit sizes ranging from 9.5 to 10.5 nm. When recombined with varying POPC content [(20-130)/1, POPC/A-I)], the three mutants produce homogeneous discoidal Lp2A-I that contain a low POPC/A-I molar ratio [(20-40)/l for all mutants] and exhibit a nearly constant size [7.5-7.6 nm for delta (100-143) and 7.9-8.0 nm for the other two mutants]. Kinetics of association of these proteins with DMPC are similar for delta (100-143) and Rec.-apoA-I (t 1/2 of 4.0 and 4.4 min, respectively) but appear significantly reduced for delta (122-165) and delta (144-186) (t 1/2 of 7.5 and 6.9 min, respectively). While in the lipid-free form, all proteins have a similar thermodynamic stability with a very comparable free energy of unfolding (delta GD degree) for the alpha-helical structure, as determined by isothermal denaturation studies. delta-(100-143) has a significantly lower alpha-helical content (33%) as compared to the other proteins [40, 41, and 45% for Rec.-apoA-I. delta (122-165), and delta (144-186), respectively]. When associated to POPC, delta (122-165) and delta (144-186) have a higher alpha-helicity (63 and 63%) and an enhanced stability (2.5 and 2.3 kcal/mol, respectively) as compared to delta (100-143) (49% and 1.8 kcal/mol) and Rec.-apoA-I (52% and 1.9 kcal/mol). These results suggest that the amphipathic alpha-helices within residues 100-186 are directly involved in interactions with phospholipids. The helical region 100-121 appears to be more important to the stabilization of the lipid-apoprotein complex formed whereas helices within residues 122-186 appear to be critical to the initial rates of association of the apoprotein with DMPC. These data suggest that an important role of the central domain 100-186 may be to maintain the plasticity of apoA-I and its ability to form different classes of HDL particles. Therefore, it is likely that this region may also play an important role in the functional properties of this apoprotein.

Five series of reconstituted discoidal HDL (LpA-I) particles have been prepared, and their constituents, apolipoprotein A-I (apoA-I), 1-palmitoyl-2-oleoylphosphatidylcholine (POPC), unesterified cholesterol (UC), phosphatidylinositol (PI), or sphingomyelin (SM), have been systematically varied to elucidate the relationship between HDL composition and cholesterol efflux from non-cholesterol-loaded human skin fibroblasts. The physical properties, such as hydrodynamic diameters, alpha-helix contents, and surface potentials, of these LpA-I have been measured and related to the ability of the LpA-I to accept cellular cholesterol. The results show that for LpA-I particles containing 2, 3, or 4 apoA-I per particle, Lp4A-I are the best acceptors of cellular cholesterol, followed by Lp3A-I and then Lp2A-I particles. Discoidal Lp2A-I with variations in POPC content, from 121 to 266 mol/particle; show no difference in their abilities to promote cholesterol efflux. Similarly, inclusion of 7 and 15 mol of free cholesterol to Lp2A-I also does not affect their ability to accept cellular cholesterol. However, increasing the content of either PI or SM, up to 20 mol/particle, is associated with significantly increased abilities of the LpA-I to promote cholesterol efflux. The efflux of cellular cholesterol to discoidal LpA-I particles is independent of specific changes in apoA-I conformation and charge, but appears to be positively related to major changes in the size of the lipoprotein particle. The study suggests that in contrast to interlipoprotein cholesterol transfers, the efflux of cholesterol from cultured fibroblasts is less sensitive to factors that affect the frequency of molecular collisions and more dependent on the ability of an HDL particle to absorb and retain cholesterol molecules. Since SM and PI appear to modulate this adsorption/desorption of cholesterol to HDL, variations in the concentration of these lipids within HDL would be expected to affect plasma cholesterol homeostasis.

The function of albumin in the movement of cholesterol into and out of non-cholesterol-loaded fibroblasts has been investigated. Cholesterol efflux from cholesterol labeled normal human skin fibroblasts to fatty acid-free human serum albumin (HSA) is biphasic with a rapid first phase that plateaus at about 15 min followed by a nearly linear phase up to 90 min, the longest incubation in this study. Saturation of efflux is observed at about 10 mg of albumin/mL. Efflux is specific to albumin since other molecules, such as ovalbumin or gelatin, do not induce efflux. The ability of HSA to induce cellular cholesterol efflux is low compared to reconstituted discoidal lipoprotein A-I (LpA-I). HSA at 2 mg/mL produces a rate of cholesterol efflux similar to that of LpA-I at 45 micrograms of protein/mL; however, these concentrations are within the physiological range for both HSA and apolipoprotein A-I (apoA-I). The efflux to the medium containing both LpA-I and HSA is greater than that to each of them alone but does not show complete additivity, indicating a competition between HSA and LpA-I. The HSA-mediated cholesterol movement is bidirectional as demonstrated by the transfer of cholesterol from HSA-(3H)- cholesterol complexes to fibroblasts; moreover, the HSA-mediated transfer is much faster than that from cholesterol-containing LpA-I (0.8 versus 0.2 pmol (micrograms of cell protein)-1 (90 min)-1. However, the presence of either low-density lipoprotein (LDL) or LpA-I in the incubation medium significantly inhibits the transfer of cholesterol from HSA-(3H)-cholesterol complexes to fibroblasts, thus allowing the bidirectional transfer of cholesterol between HSA and cells to possibly operate as a net efflux. In conclusion, albumin plays a significant role in cholesterol transfer between cells and lipoproteins.

To define the conformation of apolipoprotein A-I in discoidal particles, the immunoreactivity of a series of epitopes distributed along the apoA-I sequence has been evaluated in lipid-free apoA-I and in lipid-bound form. To this end, reconstituted discoidal lipoproteins, here called LpA-I, and defined by number of apoA-I per particle (e.g., Lp2A-I), have been prepared with palmitoyloleoylphosphatidylcholine, cholesterol, and apoA-I. Four LpA-I have been obtained and studied: two in the Lp2A-I class, 7.8 and 9.6 nm in diameter, and two in the Lp3A-I class, 10.8 and 13.4 nm. The immunoreactivity of all the epitopes tested was significantly different in LpA-I particles compared to lipid-free apoA-I, demonstrating that binding to lipids produces a drastic change in apoA-I conformation. Specific domains in the primary sequence become highly exposed while others are masked. Although the variation in immunoreactivity of the epitopes between various LpA-I was not drastic, significant differences in the calculated ED50 values were observed for a number of antibodies in small versus large particles within each class (Lp2A-I or Lp3A-I), indicating that particle size can modulate apoA-I conformation. In addition, when the competition between pairs of mAbs was analyzed in order to understand the relative position of epitopes, highly significant differences were observed as a function of particle size within each class. In particular, the competition between mAbs recognizing epitopes in the central region of apoA-I was greater in the larger particles than in their small counterparts.(ABSTRACT TRUNCATED AT 250 WORDS)

To understand the role of different discoidal lipoproteins in cellular cholesterol efflux, defined discoidal lipoproteins containing 2, 3, or 4 apolipoproteins (apo) A-I per particle (Lp2A-I, Lp3A-I, and Lp4A-I) were prepared from mixtures of apoA-I and phospholipids with or without cholesterol. Each particle had a slow pre beta migration on agarose gel electrophoresis which further decreased as the number of apoA-I increased. Incubation of cholesterol-labeled human fibroblasts with the different LpA-I at an equimolar concentration in apoA-I showed that the best acceptors of cellular cholesterol were Lp4A-I, followed by Lp3A-I and Lp2A-I. Cholesterol efflux to these particles was positively correlated to the number of apoA-I, to the ratio of phospholipids to apoA-I, and to the size of particles, three interrelated parameters. To follow the subsequent movement of cellular cholesterol after it became associated with LpA-I, cholesterol- and apoA-I-labeled LpA-I were incubated with plasma which resulted in parallel modifications of each labels electrophoretic migration with time. However, [3H]cholesterol-labeled LpA-I transferred from pre beta to alpha migration with a precursor-product relationship while 125I-LpA-I progressively shifted from pre beta to alpha migration. The change in electrophoretic migration of 125I-LpA-I is independent of cholesterol and appears related only to a modification of apoA-I charge. Lp2A-I was fastest in changing its electrophoretic migration to alpha, followed by Lp3A-I and then Lp4A-I.(ABSTRACT TRUNCATED AT 250 WORDS)

Lipoprotein (a) [Lp(a)] was isolated from several donors and its apolipoprotein (a) [apo(a)] dissociated by a reductive treatment, generating the apo(a)-free form of Lp(a) [Lp(a--)] that contains apolipoprotein B (apo B) as its sole protein. Using anti-apo B monoclonal antibodies, the properties of apo B in Lp(a), Lp(a--), and autologous low-density lipoprotein (LDL) were compared. Marked differences in apo B immunoreactivity were found between these lipoproteins, due to the presence of apo(a) in Lp(a). Apo(a) enhanced the expression of two epitopes in the amino-terminal part of apo B while it diminished the immunoreactivity of three other epitopes in the LDL receptor binding domain. Accordingly, the binding of the lipoproteins to the LDL receptor was also decreased in the presence of apo(a). In a different experimental system, the incubation of antibodies that react with 27 distinct epitopes distributed along the whole length of apo B sequence with plastic-bound Lp(a) and Lp(a--) failed to reveal any epitope of apo B that is sterically hindered by the presence of apo(a). Our results demonstrate that the presence of apo(a) modified the organization and function of apo B in Lp(a) particles. The data presented indicate that most likely the modification is not due to a steric hindrance but that some more profound conformational changes are involved. We suggest that the formation of the disulfide bridge between apo B and apo(a) in Lp(a) alters the system of disulfide bonds present in apo B and thereby modifies apo B structure.