Cholesteryl Ester Transfer Protein Mediates Selective Uptake of High Density Lipoprotein Cholesteryl Esters by Human Adipose Tissue*

We have determined the role of cholesteryl ester transfer protein (CETP) in selective uptake of high density lipoprotein (HDL)-derived cholesteryl esters (CE) by human adipose tissue, using organ culture or colla-genase-digested adipocytes. Incubation of the fresh tis- sue fragments with HDL containing [ 3 H]CE or 125 I-ap-oprotein (apo) A-I resulted in significant uptake of HDL- CE-derived label. Addition of recombinant CETP (rCETP) increased CE uptake in a dose-response fash- ion. In contrast, little association of 125 I-apoA-I with adipose tissue was noted, and addition of rCETP did not alter apoA-I uptake or degradation. Incubation of adipose tissue with TP2, an anti-CETP monoclonal anti- body, which inhibits neutral lipid transfer, markedly reduced selective uptake of HDL-CE. Studies using hu- man adipocytes isolated by collagenase digestion also demonstrated selective uptake of HDL-CE and enhance- ment of this process by rCETP. To confirm that the association of HDL-CE-derived radioactivity with adi- pose tissue was not due to neutral lipid exchange between adipocytes and HDL, we measured changes in HDL composition following incubation of HDL and rCETP with isolated adipocytes. A decrease in HDL-CE concentration in the medium was observed, an effect which was markedly attenuated when incubations were carried out in the presence of monoclonal antibody TP2. Furthermore, the decrease in of recombinant human CETP (19). In some studies, Fab fragments (20 (cid:109) g/ml) of anti-CETP monoclonal antibodies were added 1 or 2 h prior to the addition of the labeled lipoprotein.

Uptake of HDL 1 -derived CE by adipocytes has been well documented and, in the obese state, plasma HDL-CE may be reduced as a consequence (1)(2)(3). The uptake of CE by adipose tissue is disproportionately greater than the uptake of HDL apolipoprotein (1,4,5). Thus, CE are transferred from HDL to adipocytes by a nonendocytotic process, known as selective cholesterol uptake (6 -9). The selective uptake of CE by adipose cells resembles in many ways the processes described in vivo for rat liver, ovary, and adrenal (6) and in vitro for adrenal cells (7,10) and rat luteal cells (11). While many similarities in selective cholesteryl ester uptake have been noted in several tissues, the physiological consequences vary with cellular processing. It is likely that HDL can deliver cholesterol in the esterified form to adipocytes, where it is hydrolyzed to free cholesterol, which can then be stored in the oil droplet, equilibrate with the membrane, or efflux from the cell over time (12,13).
The role of CETP in reverse cholesterol transport has been well characterized (14). CETP is highly expressed in mammalian adipose tissue (15). There is also evidence of CETP activity associated with adipocyte plasma membranes (16). We have demonstrated that immunoreactive CETP is present on the adipocyte plasma membrane and that CETP gene expression is greatest in very small fat cells (10 -50 m), which are likely to be in a trophic phase and thus to have a particular requirement for lipoprotein cholesterol (17). This led us to test the hypothesis that human adipocyte CETP functions to facilitate uptake of HDL cholesteryl esters. Our findings demonstrate a novel function for CETP in human adipocyte cholesterol metabolism.

EXPERIMENTAL PROCEDURES
HDL Isolation and Labeling-Plasma was collected from healthy normolipemic donors. HDL (d ϭ 1.063-1.21 g/ml) were isolated by sequential ultracentrifugation (Beckman 55.2 Ti rotor, 40 Lipoprotein Labeling with 125 I-Purified human apoA-I was iodinated using 125 I-IODO-BEADs (32). The beads were equilibrated in phosphate buffer (0.3 M NaPO 4 , pH 7.4) and then incubated with 20 l of 125 I (2 mCi of NaI) in 70 l of phosphate buffer for 5 min at room temperature. 150 g of apoA-I were added to the mixture and incubated at room temperature for 45 min with agitation. Unincorporated label was removed by passing the mixture through a desalting column (Excellulose GF5, Pierce) which had been prewashed in 0.1% bovine serum albumin in PBS followed by PBS. Labeled apoA-1 was equilibrated with 35 mg of HDL (previously isolated) at 37°C for 4 h with gentle agitation. The density of the mixture (d ϭ 1.21 g/ml) was adjusted with KBr and then centrifuged (Beckman 55.2 Ti rotor, 40,000 rpm, 40 h, 8°C).
Lipoprotein Labeling with 3 H and 125 I-To obtain doubly labeled HDL with 3 H in esterified cholesterol and 125 I in apoA-I, the [ 3 H]HDL was then equilibrated with labeled apoA-I as described above.
The HDL-containing fraction was dialyzed against PBS overnight and electrophoretic mobility and purity were verified by Lipogel agarose gel electrophoresis (Paragon®, Beckman Instruments) and 8 -25% gradient SDS gel electrophoresis (Pharmacia Biotech, Uppsala, Sweden). Lipoprotein mass was determined by measuring protein content by the method of Lowry et al. (18).
Adipose Tissue Culture and Incubation-Subcutaneous adipose tissue was obtained from healthy normolipemic subjects undergoing reduction mammoplasty procedures for cosmetic purposes. The study was approved by the Institutional Review Board of the Ottawa Civic Hospital, and written informed consent was obtained from all subjects. Adipose tissue fragments were prepared and maintained in organ culture using standard procedures as described previously (17) and incubated (37°C, 5% CO 2 ), with constant shaking, for indicated times (1-8 h) in culture medium containing labeled HDL (50 g/ml). To determine the dose-response effect of CETP on selective uptake of HDL-CE, incubations were performed with and without the addition of physiologic quantitites (0 -2.4 g/ml) of recombinant human CETP (19). In some studies, Fab fragments (20 g/ml) of anti-CETP monoclonal antibodies were added 1 or 2 h prior to the addition of the labeled lipoprotein.
Primary Human Adipocyte Culture and Incubation-For some experiments, small fragments of tissue were incubated with collagenase (0.5 mg/ml) in Dulbecco's modified Eagle's/F-12 medium containing 10% bovine serum albumin for 1 h at 37°C. Adipocytes were separated from undigested tissue by filtering through a fine nylon mesh and washed three times with PBS containing bovine serum albumin (20). HDL incubation experiments (1-18 h) were performed as described above.
Lipoprotein Uptake Measurement-At the end of the incubation period, the medium containing labeled lipoproteins was removed and stored at 4°C and the tissue was washed three times in ice cold PBS. The adipose tissue was then homogenized in homogenization buffer (20 mM Tris buffer, pH 7.4 containing 1 mM EDTA and 1 mM ␤-mercaptoethanol) to isolate the membrane fraction, or in lysis buffer (0.02 M sodium phosphate, pH 7.5, 0.2 mM NaCl, 2% Triton X-100, 1% sodium deoxycholate, 0.2% SDS, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride) for cytosolic analysis. The homogenate in lysis buffer was centrifuged (Sorval RT 6,000D, 2,000 rpm, 20 min, 4°C) to separate the lipophilic layer (oil droplet) from the remaining homogenate (infranatant). The infranatant (cytosol and membrane) was removed.
[ 3 H]Cholesteryl Ester Uptake-After homogenization in lysis buffer, lipids from lipophilic layer (oil droplet) were extracted as described by Folch et al. (21). The mass of total core lipid was determined by weighing lipid extract. Then, extracted lipids were redissolved in chloroform, 3 H radioactivity was counted, and the amount of CE was determined with respect to initial specific activity. Results were expressed as nanograms of HDL protein/mg of total core lipid.
Membrane Lipid Analysis-The infranatant from tissue homogenized in homogenization buffer was ultracentrifuged (Beckman 100.4 rotor, 35,000 rpm, 8°C, 1 h) to separate the membrane fraction (pellet) from the cytosolic fraction (supernatant). The pellet was dissolved into 1 ml of 0.1 N NaOH at room temperature, overnight. Lipids were extracted as described by Bligh and Dyer (22), and mass of FC and PL was determined by gas liquid chromatography (23) and HPTLC analysis (24) as described previously with some modifications. Thin layer chromatography was developed in hexane/diethylether/formic acid, 80: 20:2, and washed in 100% heptane. Results were expressed as percent of control (HDL prior to incubation). Protein was determined by the method of Lowry et al. (18).
HDL Analysis-The 3 H counts remaining in HDL in the medium following incubation were determined. The medium was adjusted to d ϭ 1.21 g/ml with KBr, and HDL were isolated by sequential ultracentrifugation (Beckman 45.6Ti rotor, 40,000rpm, 4°C, 48 h). The 3 H-associated radioactivity was counted. Lipids were extracted and mass was determined as described above by HPTLC and gas liquid chromatography. Results were expressed as nanograms of lipid/g of protein. Protein was determined by Lowry assay (18).

125
I-ApoA-1 Uptake-The infranatant (cytosol ϩ membrane) removed after homogenization in lysis buffer was counted for 125 I radio-activity using a gamma counter. In another experiment, the infranatant was ultracentrifuged to separate the membrane fractions (pellet) from the cytosolic fraction (supernatant). Proteins from these fractions were precipitated by trichloroacetic acid methods (25) and 125 I-associated radioactivity was counted. ApoA-I was quantitated with respect to initial specific activity. Results are expressed as nanograms of HDL protein/ng of total core lipid determined as described above.
HDL-ApoA-1 and ApoA-II Composition-ApoA-I and ApoA-II were quantified by immunoelectrodiffusion in agarose gel using the Hydragel A-I/B and A-II kit (Sebia R , France).
Statistics and Calculations-Data are expressed as mean Ϯ S.E. Significant of differences was examined using Student's t test for unpaired data. Cellular uptake of HDL tracers is shown as apparent HDL particle uptake, as indicated by the cell content of each tracer (6). This apparent uptake is expressed in terms of HDL protein, to compare uptake of both tracers on the same basis.

RESULTS
The time course pattern of uptake of HDL by human adipose tissue is illustrated in Fig. 1. The uptake of cholesteryl ester from HDL was most rapid over the 1st h but continued to increase over 4 h ( Fig. 1), suggesting that the adipocyte membrane capacity as a cholesteryl ester acceptor is not limiting. This experiment was repeated with HDL labeled with 125 I-apoA-I to determine if the tissue-associated radioactivity was due to selective uptake of HDL-derived CE or to particle uptake. In contrast to the pattern of a continuous increase in [ 3 H]CE uptake with time, little 125 I-apoA-I became associated with adipose tissue within 30 min of incubation and the level of cell-associated 125 I-apoA-I declined by 2 h and remained constant thereafter (Fig. 1). These experiments confirm the phenomenon of selective uptake of HDL-derived CE by human adipose tissue, in accord with previous reports (1,4,5).
Human adipocytes synthesize and secrete CETP (15). To determine whether the observed selective uptake of HDL-CE could be further enhanced by exogenous CETP, 4-h incubations were performed following addition to the medium of physiological plasma concentrations (0.6, 1.2, 2.4 g/ml) (19) of recombinant human CETP (rCETP). As shown in Fig. 2, rCETP had a linear dose-response enhancing effect on the selective uptake of cholesteryl ester from HDL. In the absence of exogenous rCETP, there was a smaller but nonetheless significant uptake of HDL-derived cholesteryl ester by adipose tissue demonstrating that this is a physiological process, which is likely to be mediated, in part, by adipocyte CETP. In contrast to its effects on CE, increasing amounts of rCETP did not increase cellassociated 125 I-apoA-I, demonstrating that the observed uptake of [ 3 H]CE by adipose tissue was not due to particle uptake (Fig. 2).
The effect of exogenous rCETP (2.4 g/ml of medium) on the time course pattern of selective uptake is shown in Fig. 3. Using adipose tissue fragments obtained from another normal subject, we found that in the presence of rCETP, selective CE uptake was more rapid and continuous and did not plateau at 4 h. These data confirm that adipocyte membrane capacity as a cholesteryl ester acceptor is not limiting in presence of rCETP. Results of experiments using adipose tissue fragment from different patients (Figs. 1-3) demonstrate individual variation in the extent of selective uptake of HDL-CE possibly due to differences in adipocyte size and adipose tissue matrix composition.
Monoclonal antibodies (mAbs) against different epitopes of the CETP molecule (26,27) were used to verify that the observed HDL-CE uptake was mediated by CETP. To eliminate steric hindrance artifact, we prepared Fab fragments for these studies. The anti-CETP mAb, TP2, which reacts with an epitope in the carboxyl end of the molecule and which inhibits neutral lipid transfer (27), had no effect on cell-associated 125 I-apoA-I after a 4 h incubation (data not shown). However, TP2 reduced selective uptake of HDL-CE by human adipose tissue by 70 Ϯ 3.5% (p ϭ 0.03). When studies were carried out in the presence of rCETP, TP2 reduced selective uptake by 54 Ϯ 1.4% (p ϭ 0.05) (Fig. 4) In contrast, the anti-CETP mAb, TP20, which reacts with an epitope in the amino-terminal of CETP and does not alter neutral lipid transfer (27), did not alter selective uptake (Fig. 4) suggesting that the neutral lipid transfer activity of CETP is essential to this process. The observed CETP-mediated uptake was not due to CETP associated with labeled HDL since removal of trace amounts of CETP from HDL by TP2 immunoaffinity did not affect selective uptake (data not shown). It is thus likely that CETP endogenously synthesized by adipocytes mediates selective uptake of HDL-CE by adipose tissue, accounting for the TP2-inhibitable selective uptake of HDL-CE. Although the neutralizing mAb, TP2, reduced selective uptake, this effect was not enhanced when the concentration of TP2 was doubled, indicating that the levels added were sufficient to block most, but not all, HDL-CE uptake. Thus, additional mechanisms, not involving CETP, may account for a portion of the selective uptake of HDL-derived CE by human adipose tissue.
Adipose tissue contains adipocytes and other cell types including endothelial cells, fibroblasts and macrophages. To ascertain that these other adipose tissue constituents were not responsible for selective uptake, we studied the uptake of HDL-CE and its stimulation in presence of rCETP using isolated human adipocytes. When studies were carried out with singly ([ 3 H]CE) or doubly labeled ([ 3 H]CE and 125 I-apoA-I) HDL, results were similar to experiments performed with adipose tissue fragments, indicating an increase in CE uptake over time and association of 125 I-apoA-I with adipocytes after 1-h incubation followed by a slow decline in cell associated apoA-I counts over 18 h (Fig. 5). These data were confirmed by quantification of apoA-I protein associated with adipocyte cytosol and membrane. Our results indicate that when HDL is incubated with adipocytes or adipose tissue, a small amount of apoA-I associates with the adipocyte membrane before being released or degraded. Trichloroacetic acid precipitation of 125 I-apoA-I derived counts revealed that only 20 Ϯ 5% of the small fraction of apoA-I which became associated with adipocytes To verify that the apoA-I labeling protocol did not alter the HDL composition, this experiment was performed with doubly labeled HDL with 125 I-apoA-I and [ 3 H]CE. Comparison of singly and doubly labeled HDL revealed that there was no effect of the labeling procedure on HDL particle composition (PL, FC, CE, apoA-I, and apoA-II). Nor was there an effect of the labeling procedure on the time-course pattern of HDL-CE uptake (data not shown). For studies with doubly labeled HDL particles, we measured only cell associated, 125 I-apoA-I derived radioactivity since ␥ counts could be determined without interference of ␤ counts (Fig. 5).
The phenomenon of selective uptake of HDL-CE was clearly demonstrable in studies with isolated adipocytes (Fig. 6). Compared with above data (Figs. 2 and 3), albeit from different subjects, it appears that the effect of rCETP on selective uptake of HDL-CE may be greater in isolated adipocytes than in adipose tissue fragments (10-fold versus 2-5-fold increase in CE uptake, respectively). The removal of the adipose tissue matrix in primary adipocyte cultures may explain the enhanced effect of rCETP on selective uptake of HDL-CE.
We also measured changes in HDL lipid and protein composition and adipocyte membrane cholesterol before and after incubation with the cells. After 4 h of incubation of HDL with primary adipocytes, the CE content of HDL decreased by 37% (p ϭ 0.01). In contrast, the free cholesterol mass increased 9-fold (p ϭ 0.0001), whereas the PL content (Fig. 7) and ratio of apoA-I/apoA-II (data not shown) were not affected. No change in HDL-TG content was detectable by enzymatic assay (data not shown).
When HDL was incubated with adipose tissue in the absence of added rCETP, adipocyte plasma membrane cholesterol decreased by 30% after 4 h of incubation (Fig. 9), likely representing efflux, to HDL, of adipocyte membrane cholesterol (endogenous membrane cholesterol as well as free cholesterol derived from CE acquired from HDL). In the presence of rCETP (2.4 g/ml), the cholesterol content of the membrane was restored to baseline (Fig. 8). This likely reflects the balance of increased CE uptake mediated by rCETP, CE hydrolysis, transfer of FC to both membrane and core lipid compartments and continued cholesterol efflux to HDL (Fig. 8). Addition of rCETP (2.4 g/ml) to the medium containing HDL incubated with adipocytes did not alter the FC and PL composition of HDL but induced a further and significant 23% decrease in HDL-CE compared with a control without addition of rCETP (p ϭ 0.023) (Fig. 7). This effect was markedly attenuated when incubations were carried out in the presence of the neutralizing mAb, TP2 (data not shown).

DISCUSSION
Adipose tissue is composed of adipocytes and adipocyte precursors interspaced in an abundant capillary layer, held together by connective tissue. Freshly isolated human fat cells studied ex vivo maintain the properties of in vivo cells (3,28) in terms of HDL binding and cholesterol uptake. Adipocytes have very limited capacity for cholesterol synthesis and are highly dependent on lipoproteins as a source of cholesterol to maintain a relatively fixed ratio of cholesterol to triglyceride in the core lipid droplet. We have defined a novel role for CETP in medi-  1-18 h), adipocytes were washed and homogenized, and lipids were extracted to measure 3 H radioactivity from singly labeled HDL (OE). In another experiment, at the end of incubations periods, adipocytes doubly labeled with [ 3 H]cholesteryl oleate and 125 I-apoA-I were washed and homogenized in homogenization buffer. The infranatant (membrane and cytosol) was ultracentrifuged to separate the membrane fraction (pellet) (E) from the cytosolic fraction (supernatant) (q). Proteins from these fractions were trichloroacetic acid-precipitated and 125 I-apoA-I derived counts measured. Results are expressed as nanograms of HDL protein/mg of total core lipid. Each data point represents the mean Ϯ S.E. for three samples. ating the selective uptake of HDL-derived CE by human adipocytes. Adipocytes synthesize and secrete CETP and selectively acquire CE from HDL. These studies demonstrate that selective uptake of HDL-CE by human adipocytes can be further enhanced in a dose-response fashion by addition of rCETP and blocked by incubation with the neutralizing mAb, TP2.
The preferential uptake of HDL-cholesteryl ester is not unique to adipocytes and has been observed in cultured adrenal cells (7,10) and other steroidogenic cells. Similarly, the majority of HDL-CE clearance in the rat does not involve particle uptake (6,29,30). In HepG2 cells, CETP has been shown to mediate the transfer of HDL-CE to newly secreted lipoproteins which are then retaken up by the liver (31). Thus, in human liver, there is no evidence that CETP mediates direct cellular uptake of cholesteryl esters but CETP does, nonetheless, play an important role in an indirect and quantitatively important pathway for the clearance of HDL-CE (29). However, in adipocytes and steroidogenic cells which do not secrete apoB lipoproteins, more direct mechanisms for selective uptake of HDL-CE are likely to be operative. In transfected COS cells, SR-B1, a member of the scavenger receptor superfamily, has been shown to mediate the selective uptake of HDL-CE (32). Immunoreactive SR-B1 is present in high concentrations in adrenal, ovary and testes and may mediate selective uptake of HDL-CE by these tissues (32). Although adipocytes have been reported to express SR-B1 mRNA in some (33,34) but not other studies (35), they appear to express no (33,35), or negligible amounts of immunoreactive SR-B1 protein (34). Further studies are required to determine whether human adipocytes express SR-B1 and its role if any in adipocyte cholesterol homeostasis.
The role of CETP in reverse cholesterol transport has been well characterized (14). CETP is highly expressed in mammalian adipose tissue (15). We have demonstrated that human adipose tissue, in organ culture, synthesizes and secretes CETP (17) and that immunoreactive CETP is present on the plasma membrane. There is also evidence of neutral lipid transfer activity associated with adipocyte plasma membranes and, interestingly, this has been demonstrated in species which lack immunoreactive CETP or neutral lipid transfer activity in plasma (16). Adipocyte membrane cholesterol is a modifiable pool of cholesterol and cholesterol concentrations vary in different metabolic states (17). This study demonstrates that adipocyte membrane cholesterol is decreased following a 4 h incubation with HDL and that this decrease in membrane cholesterol can be prevented by addition of rCETP to the media (Fig. 8). In other studies, we have reported that CETP mRNA abundance is significantly correlated with membrane cholesterol (17), suggesting that there is a pool of membrane cholesterol which regulates CETP gene expression. The requirement for exogeneous lipoprotein derived cholesterol would be expected to be greatest for immature lipid poor adipocytes and, consistent with this hypothesis, we have shown that CETP mRNA abundance is greater in very small fat cells (10 -50 m in diameter) in comparison to mature adipocytes (50 -200 m) (17).
In the present experiments, following incubation of HDL with adipose tissue, we noted a significant decrease in HDL-CE (Fig. 7), consistent with the observed uptake of [ 3 H]CE-derived counts by adipose tissue. In addition, the ratio of FC to PL increased in HDL and decreased in the adipocyte membrane suggesting net movement of free cholesterol to HDL in accord with previous studies showing net efflux of adipocyte free cholesterol to HDL acceptor particles (36). The transfer of free cholesterol between cells and HDL particles is likely to be bidirectional, involving diffusion of cholesterol in the aqueous space between the plasma membrane and the lipoprotein surface (37)(38)(39)(40). When these incubations were carried out in the presence of rCETP, HDL was further depleted of CE, and adipocyte membrane cholesterol was restored to base line. These changes in HDL and adipocyte membrane lipid composition are likely to reflect the balance of CE uptake which is facilitated by CETP, CE hydolysis, transfer of free cholesterol to the core lipid droplet and plasma membrane, and efflux of free cholesterol to HDL. Following incubation of HDL with adipose tissue either in the presence or absence of rCETP, no change in the ratio of apoA-I to apoA-II in HDL or in the concentration of adipocyte membrane-associated apoA-I was noted.
Preferential cholesteryl ester uptake by adipocytes and steroidogenic cells implies that HDL can deliver cholesterol without undergoing catabolic degradation through a lysosomal process. It is likely that there are several independent regulatory mechanisms at various stages between the delivery of HDL-CE to the adipocyte plasma membrane and subsequent uptake and intracellular metabolism. Little is known regarding cholesterol exchange between core and membrane cholesterol. Despite a disproportionate uptake of HDL core lipid, the process may not be completely independent of HDL-apolipoprotein interaction and cholesteryl ester transfer to the adipocyte may involve a specific apolipoprotein. In previous studies, using isolated human adipocytes, a correlation between HDL binding and cholesteryl ester uptake was demonstrated (4). The nonendocytotic mechanism for selective uptake may be mediated by an interaction with a cell surface lipoprotein binding domain. Cultured adipocytes actively synthesize and secrete apoE and lipoprotein lipase (41) as well as heparin sulfate proteoglycans (HSPG) which bind lipoprotein lipase and apoE (42). Adipocytes express various receptors involved in lipoprotein catabolism, including the LDL-receptor (43), LDL-related receptor protein (44), and very low density lipoprotein-receptor (45). Little is known about the interaction between these molecules and the adipocyte membrane. The fact that the effect of rCETP on selective uptake is greater when HDL is incubated with isolated adipocytes as compared with adipose tissue fragments suggests that the collagenous matrix of adipose tissue may have impeded the effect of exogenous rCETP on selective uptake. Other components of the matrix surrounding adipocytes may, however, mediate lipoprotein interactions. In other experiments (data not shown) disruption of the adipose tissue HSPG matrix with heparinases, reduced selective uptake of FIG. 8. Changes in membrane lipid composition following incubation with HDL. Isolated adipocytes (same patient illustrated in Fig. 7) were incubated for 4 h in culture medium containing HDL (50 g/ml) labeled with [ 3 H]cholesteryl oleate in presence or absence of rCETP. Adipocytes labeled with [ 3 H]cholesteryl oleate were homogenized. The infranatant (membrane and cytosol) was ultracentrifuged to separate the membrane fraction (pellet) from the cytosolic fraction. Lipids were extracted as described and mass was determined by gas liquid chromatography for FC and HPTLC for PL. Base-line adipocyte membrane composition is indicated as Ⅺ and composition following incubation with HDL in the presence of rCETP as f and in the absence of rCETP as u. Results obtained from three experiments are shown and are expressed as percent of control. a, p Ͻ 0.05. HDL-CE, suggesting that HSPG may facilitate the interaction of HDL with the adipocyte plasma membrane. Further studies are under way to determine whether CETP-mediated uptake of HDL-derived CE by adipocytes is dependent upon interaction of apoA-I or apoE with the adipocyte membrane or HSPG complex.
These data demonstrate that CETP plays a novel and important role in the selective uptake of CE from HDL by human adipocytes and suggest that this pathway may be of quantitative physiological significance in HDL remodeling and adipocyte cholesterol accumulation. Obesity is commonly associated with hypoalphalipoproteinemia even in the absence of hypertriglyceridemia and it is possible that CETP-mediated clearance of HDL-CE by adipose tissue may contribute to this phenotype.